diff --git a/llvm/lib/Target/RISCV/RISCVInsertVSETVLI.cpp b/llvm/lib/Target/RISCV/RISCVInsertVSETVLI.cpp
index 870e393b40411..0f952384549ed 100644
--- a/llvm/lib/Target/RISCV/RISCVInsertVSETVLI.cpp
+++ b/llvm/lib/Target/RISCV/RISCVInsertVSETVLI.cpp
@@ -40,6 +40,12 @@ using namespace llvm;
 STATISTIC(NumInsertedVSETVL, "Number of VSETVL inst inserted");
 STATISTIC(NumCoalescedVSETVL, "Number of VSETVL inst coalesced");
 
+static cl::opt<bool> EnsureWholeVectorRegisterMoveValidVTYPE(
+    DEBUG_TYPE "-whole-vector-register-move-valid-vtype", cl::Hidden,
+    cl::desc("Insert vsetvlis before vmvNr.vs to ensure vtype is valid and "
+             "vill is cleared"),
+    cl::init(true));
+
 namespace {
 
 /// Given a virtual register \p Reg, return the corresponding VNInfo for it.
@@ -195,6 +201,14 @@ static bool hasUndefinedPassthru(const MachineInstr &MI) {
   return UseMO.getReg() == RISCV::NoRegister || UseMO.isUndef();
 }
 
+/// Return true if \p MI is a copy that will be lowered to one or more vmvNr.vs.
+static bool isVectorCopy(const TargetRegisterInfo *TRI,
+                         const MachineInstr &MI) {
+  return MI.isCopy() && MI.getOperand(0).getReg().isPhysical() &&
+         RISCVRegisterInfo::isRVVRegClass(
+             TRI->getMinimalPhysRegClass(MI.getOperand(0).getReg()));
+}
+
 /// Which subfields of VL or VTYPE have values we need to preserve?
 struct DemandedFields {
   // Some unknown property of VL is used.  If demanded, must preserve entire
@@ -221,10 +235,13 @@ struct DemandedFields {
   bool SEWLMULRatio = false;
   bool TailPolicy = false;
   bool MaskPolicy = false;
+  // If this is true, we demand that VTYPE is set to some legal state, i.e. that
+  // vill is unset.
+  bool VILL = false;
 
   // Return true if any part of VTYPE was used
   bool usedVTYPE() const {
-    return SEW || LMUL || SEWLMULRatio || TailPolicy || MaskPolicy;
+    return SEW || LMUL || SEWLMULRatio || TailPolicy || MaskPolicy || VILL;
   }
 
   // Return true if any property of VL was used
@@ -239,6 +256,7 @@ struct DemandedFields {
     SEWLMULRatio = true;
     TailPolicy = true;
     MaskPolicy = true;
+    VILL = true;
   }
 
   // Mark all VL properties as demanded
@@ -263,6 +281,7 @@ struct DemandedFields {
     SEWLMULRatio |= B.SEWLMULRatio;
     TailPolicy |= B.TailPolicy;
     MaskPolicy |= B.MaskPolicy;
+    VILL |= B.VILL;
   }
 
 #if !defined(NDEBUG) || defined(LLVM_ENABLE_DUMP)
@@ -308,7 +327,8 @@ struct DemandedFields {
     OS << ", ";
     OS << "SEWLMULRatio=" << SEWLMULRatio << ", ";
     OS << "TailPolicy=" << TailPolicy << ", ";
-    OS << "MaskPolicy=" << MaskPolicy;
+    OS << "MaskPolicy=" << MaskPolicy << ", ";
+    OS << "VILL=" << VILL;
     OS << "}";
   }
 #endif
@@ -503,6 +523,21 @@ DemandedFields getDemanded(const MachineInstr &MI, const RISCVSubtarget *ST) {
     }
   }
 
+  // In §32.16.6, whole vector register moves have a dependency on SEW. At the
+  // MIR level though we don't encode the element type, and it gives the same
+  // result whatever the SEW may be.
+  //
+  // However it does need valid SEW, i.e. vill must be cleared. The entry to a
+  // function, calls and inline assembly may all set it, so make sure we clear
+  // it for whole register copies. Do this by leaving VILL demanded.
+  if (isVectorCopy(ST->getRegisterInfo(), MI)) {
+    Res.LMUL = DemandedFields::LMULNone;
+    Res.SEW = DemandedFields::SEWNone;
+    Res.SEWLMULRatio = false;
+    Res.TailPolicy = false;
+    Res.MaskPolicy = false;
+  }
+
   return Res;
 }
 
@@ -1208,6 +1243,18 @@ static VSETVLIInfo adjustIncoming(VSETVLIInfo PrevInfo, VSETVLIInfo NewInfo,
 // legal for MI, but may not be the state requested by MI.
 void RISCVInsertVSETVLI::transferBefore(VSETVLIInfo &Info,
                                         const MachineInstr &MI) const {
+  if (isVectorCopy(ST->getRegisterInfo(), MI) &&
+      (Info.isUnknown() || !Info.isValid() || Info.hasSEWLMULRatioOnly())) {
+    // Use an arbitrary but valid AVL and VTYPE so vill will be cleared. It may
+    // be coalesced into another vsetvli since we won't demand any fields.
+    VSETVLIInfo NewInfo; // Need a new VSETVLIInfo to clear SEWLMULRatioOnly
+    NewInfo.setAVLImm(1);
+    NewInfo.setVTYPE(RISCVII::VLMUL::LMUL_1, /*sew*/ 8, /*ta*/ true,
+                     /*ma*/ true);
+    Info = NewInfo;
+    return;
+  }
+
   if (!RISCVII::hasSEWOp(MI.getDesc().TSFlags))
     return;
 
@@ -1296,7 +1343,8 @@ bool RISCVInsertVSETVLI::computeVLVTYPEChanges(const MachineBasicBlock &MBB,
   for (const MachineInstr &MI : MBB) {
     transferBefore(Info, MI);
 
-    if (isVectorConfigInstr(MI) || RISCVII::hasSEWOp(MI.getDesc().TSFlags))
+    if (isVectorConfigInstr(MI) || RISCVII::hasSEWOp(MI.getDesc().TSFlags) ||
+        isVectorCopy(ST->getRegisterInfo(), MI))
       HadVectorOp = true;
 
     transferAfter(Info, MI);
@@ -1426,6 +1474,16 @@ void RISCVInsertVSETVLI::emitVSETVLIs(MachineBasicBlock &MBB) {
       PrefixTransparent = false;
     }
 
+    if (EnsureWholeVectorRegisterMoveValidVTYPE &&
+        isVectorCopy(ST->getRegisterInfo(), MI)) {
+      if (!PrevInfo.isCompatible(DemandedFields::all(), CurInfo, LIS)) {
+        insertVSETVLI(MBB, MI, MI.getDebugLoc(), CurInfo, PrevInfo);
+        PrefixTransparent = false;
+      }
+      MI.addOperand(MachineOperand::CreateReg(RISCV::VTYPE, /*isDef*/ false,
+                                              /*isImp*/ true));
+    }
+
     uint64_t TSFlags = MI.getDesc().TSFlags;
     if (RISCVII::hasSEWOp(TSFlags)) {
       if (!PrevInfo.isCompatible(DemandedFields::all(), CurInfo, LIS)) {
diff --git a/llvm/test/CodeGen/RISCV/inline-asm-v-constraint.ll b/llvm/test/CodeGen/RISCV/inline-asm-v-constraint.ll
index c04e4fea7b2c2..64957ed6e4ba6 100644
--- a/llvm/test/CodeGen/RISCV/inline-asm-v-constraint.ll
+++ b/llvm/test/CodeGen/RISCV/inline-asm-v-constraint.ll
@@ -45,6 +45,7 @@ define <vscale x 1 x i8> @constraint_vd(<vscale x 1 x i8> %0, <vscale x 1 x i8>
 define <vscale x 1 x i1> @constraint_vm(<vscale x 1 x i1> %0, <vscale x 1 x i1> %1) nounwind {
 ; RV32I-LABEL: constraint_vm:
 ; RV32I:       # %bb.0:
+; RV32I-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; RV32I-NEXT:    vmv1r.v v9, v0
 ; RV32I-NEXT:    vmv1r.v v0, v8
 ; RV32I-NEXT:    #APP
@@ -54,6 +55,7 @@ define <vscale x 1 x i1> @constraint_vm(<vscale x 1 x i1> %0, <vscale x 1 x i1>
 ;
 ; RV64I-LABEL: constraint_vm:
 ; RV64I:       # %bb.0:
+; RV64I-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; RV64I-NEXT:    vmv1r.v v9, v0
 ; RV64I-NEXT:    vmv1r.v v0, v8
 ; RV64I-NEXT:    #APP
diff --git a/llvm/test/CodeGen/RISCV/rvv/abs-vp.ll b/llvm/test/CodeGen/RISCV/rvv/abs-vp.ll
index 163d9145bc362..ee0016ec080e2 100644
--- a/llvm/test/CodeGen/RISCV/rvv/abs-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/abs-vp.ll
@@ -567,6 +567,7 @@ define <vscale x 16 x i64> @vp_abs_nxv16i64(<vscale x 16 x i64> %va, <vscale x 1
 ; CHECK-NEXT:    slli a1, a1, 4
 ; CHECK-NEXT:    sub sp, sp, a1
 ; CHECK-NEXT:    .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x10, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 16 * vlenb
+; CHECK-NEXT:    vsetvli a1, zero, e8, mf4, ta, ma
 ; CHECK-NEXT:    vmv1r.v v24, v0
 ; CHECK-NEXT:    csrr a1, vlenb
 ; CHECK-NEXT:    slli a1, a1, 3
@@ -576,7 +577,6 @@ define <vscale x 16 x i64> @vp_abs_nxv16i64(<vscale x 16 x i64> %va, <vscale x 1
 ; CHECK-NEXT:    csrr a1, vlenb
 ; CHECK-NEXT:    srli a2, a1, 3
 ; CHECK-NEXT:    sub a3, a0, a1
-; CHECK-NEXT:    vsetvli a4, zero, e8, mf4, ta, ma
 ; CHECK-NEXT:    vslidedown.vx v0, v0, a2
 ; CHECK-NEXT:    sltu a2, a0, a3
 ; CHECK-NEXT:    addi a2, a2, -1
diff --git a/llvm/test/CodeGen/RISCV/rvv/bitreverse-vp.ll b/llvm/test/CodeGen/RISCV/rvv/bitreverse-vp.ll
index 66a1178cddb66..10d24927d9b78 100644
--- a/llvm/test/CodeGen/RISCV/rvv/bitreverse-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/bitreverse-vp.ll
@@ -3075,6 +3075,7 @@ define <vscale x 64 x i16> @vp_bitreverse_nxv64i16(<vscale x 64 x i16> %va, <vsc
 ; CHECK-NEXT:    slli a1, a1, 4
 ; CHECK-NEXT:    sub sp, sp, a1
 ; CHECK-NEXT:    .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x10, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 16 * vlenb
+; CHECK-NEXT:    vsetvli a1, zero, e8, m1, ta, ma
 ; CHECK-NEXT:    vmv1r.v v24, v0
 ; CHECK-NEXT:    csrr a1, vlenb
 ; CHECK-NEXT:    slli a1, a1, 3
@@ -3086,7 +3087,6 @@ define <vscale x 64 x i16> @vp_bitreverse_nxv64i16(<vscale x 64 x i16> %va, <vsc
 ; CHECK-NEXT:    lui a2, 3
 ; CHECK-NEXT:    srli a4, a3, 1
 ; CHECK-NEXT:    slli a3, a3, 2
-; CHECK-NEXT:    vsetvli a5, zero, e8, m1, ta, ma
 ; CHECK-NEXT:    vslidedown.vx v0, v0, a4
 ; CHECK-NEXT:    sub a4, a0, a3
 ; CHECK-NEXT:    sltu a5, a0, a4
@@ -3158,11 +3158,11 @@ define <vscale x 64 x i16> @vp_bitreverse_nxv64i16(<vscale x 64 x i16> %va, <vsc
 ;
 ; CHECK-ZVBB-LABEL: vp_bitreverse_nxv64i16:
 ; CHECK-ZVBB:       # %bb.0:
+; CHECK-ZVBB-NEXT:    vsetvli a1, zero, e8, m1, ta, ma
 ; CHECK-ZVBB-NEXT:    vmv1r.v v24, v0
 ; CHECK-ZVBB-NEXT:    csrr a1, vlenb
 ; CHECK-ZVBB-NEXT:    srli a2, a1, 1
 ; CHECK-ZVBB-NEXT:    slli a1, a1, 2
-; CHECK-ZVBB-NEXT:    vsetvli a3, zero, e8, m1, ta, ma
 ; CHECK-ZVBB-NEXT:    vslidedown.vx v0, v0, a2
 ; CHECK-ZVBB-NEXT:    sub a2, a0, a1
 ; CHECK-ZVBB-NEXT:    sltu a3, a0, a2
diff --git a/llvm/test/CodeGen/RISCV/rvv/bswap-vp.ll b/llvm/test/CodeGen/RISCV/rvv/bswap-vp.ll
index 1c95ec8fafd4f..0dc1d0c32ac44 100644
--- a/llvm/test/CodeGen/RISCV/rvv/bswap-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/bswap-vp.ll
@@ -1584,6 +1584,7 @@ define <vscale x 64 x i16> @vp_bswap_nxv64i16(<vscale x 64 x i16> %va, <vscale x
 ; CHECK-NEXT:    slli a1, a1, 4
 ; CHECK-NEXT:    sub sp, sp, a1
 ; CHECK-NEXT:    .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x10, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 16 * vlenb
+; CHECK-NEXT:    vsetvli a1, zero, e8, m1, ta, ma
 ; CHECK-NEXT:    vmv1r.v v24, v0
 ; CHECK-NEXT:    csrr a1, vlenb
 ; CHECK-NEXT:    slli a1, a1, 3
@@ -1593,7 +1594,6 @@ define <vscale x 64 x i16> @vp_bswap_nxv64i16(<vscale x 64 x i16> %va, <vscale x
 ; CHECK-NEXT:    csrr a1, vlenb
 ; CHECK-NEXT:    srli a2, a1, 1
 ; CHECK-NEXT:    slli a1, a1, 2
-; CHECK-NEXT:    vsetvli a3, zero, e8, m1, ta, ma
 ; CHECK-NEXT:    vslidedown.vx v0, v0, a2
 ; CHECK-NEXT:    sub a2, a0, a1
 ; CHECK-NEXT:    sltu a3, a0, a2
@@ -1631,11 +1631,11 @@ define <vscale x 64 x i16> @vp_bswap_nxv64i16(<vscale x 64 x i16> %va, <vscale x
 ;
 ; CHECK-ZVKB-LABEL: vp_bswap_nxv64i16:
 ; CHECK-ZVKB:       # %bb.0:
+; CHECK-ZVKB-NEXT:    vsetvli a1, zero, e8, m1, ta, ma
 ; CHECK-ZVKB-NEXT:    vmv1r.v v24, v0
 ; CHECK-ZVKB-NEXT:    csrr a1, vlenb
 ; CHECK-ZVKB-NEXT:    srli a2, a1, 1
 ; CHECK-ZVKB-NEXT:    slli a1, a1, 2
-; CHECK-ZVKB-NEXT:    vsetvli a3, zero, e8, m1, ta, ma
 ; CHECK-ZVKB-NEXT:    vslidedown.vx v0, v0, a2
 ; CHECK-ZVKB-NEXT:    sub a2, a0, a1
 ; CHECK-ZVKB-NEXT:    sltu a3, a0, a2
diff --git a/llvm/test/CodeGen/RISCV/rvv/calling-conv-fastcc.ll b/llvm/test/CodeGen/RISCV/rvv/calling-conv-fastcc.ll
index a4e5ab661c528..15f6ca600cb37 100644
--- a/llvm/test/CodeGen/RISCV/rvv/calling-conv-fastcc.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/calling-conv-fastcc.ll
@@ -336,6 +336,7 @@ define fastcc <vscale x 32 x i32> @ret_nxv32i32_call_nxv32i32_nxv32i32_i32(<vsca
 ; RV32-NEXT:    add a1, a3, a1
 ; RV32-NEXT:    li a3, 2
 ; RV32-NEXT:    vs8r.v v16, (a1)
+; RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; RV32-NEXT:    vmv8r.v v8, v0
 ; RV32-NEXT:    vmv8r.v v16, v24
 ; RV32-NEXT:    call ext2
@@ -374,6 +375,7 @@ define fastcc <vscale x 32 x i32> @ret_nxv32i32_call_nxv32i32_nxv32i32_i32(<vsca
 ; RV64-NEXT:    add a1, a3, a1
 ; RV64-NEXT:    li a3, 2
 ; RV64-NEXT:    vs8r.v v16, (a1)
+; RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; RV64-NEXT:    vmv8r.v v8, v0
 ; RV64-NEXT:    vmv8r.v v16, v24
 ; RV64-NEXT:    call ext2
@@ -451,6 +453,7 @@ define fastcc <vscale x 32 x i32> @ret_nxv32i32_call_nxv32i32_nxv32i32_nxv32i32_
 ; RV32-NEXT:    add a1, sp, a1
 ; RV32-NEXT:    addi a1, a1, 128
 ; RV32-NEXT:    vl8r.v v8, (a1) # Unknown-size Folded Reload
+; RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; RV32-NEXT:    vmv8r.v v16, v0
 ; RV32-NEXT:    call ext3
 ; RV32-NEXT:    addi sp, s0, -144
@@ -523,6 +526,7 @@ define fastcc <vscale x 32 x i32> @ret_nxv32i32_call_nxv32i32_nxv32i32_nxv32i32_
 ; RV64-NEXT:    add a1, sp, a1
 ; RV64-NEXT:    addi a1, a1, 128
 ; RV64-NEXT:    vl8r.v v8, (a1) # Unknown-size Folded Reload
+; RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; RV64-NEXT:    vmv8r.v v16, v0
 ; RV64-NEXT:    call ext3
 ; RV64-NEXT:    addi sp, s0, -144
diff --git a/llvm/test/CodeGen/RISCV/rvv/calling-conv.ll b/llvm/test/CodeGen/RISCV/rvv/calling-conv.ll
index 9b27116fef7ca..2e181e0914c88 100644
--- a/llvm/test/CodeGen/RISCV/rvv/calling-conv.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/calling-conv.ll
@@ -103,6 +103,7 @@ define target("riscv.vector.tuple", <vscale x 16 x i8>, 2) @caller_tuple_return(
 ; RV32-NEXT:    sw ra, 12(sp) # 4-byte Folded Spill
 ; RV32-NEXT:    .cfi_offset ra, -4
 ; RV32-NEXT:    call callee_tuple_return
+; RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; RV32-NEXT:    vmv2r.v v6, v8
 ; RV32-NEXT:    vmv2r.v v8, v10
 ; RV32-NEXT:    vmv2r.v v10, v6
@@ -119,6 +120,7 @@ define target("riscv.vector.tuple", <vscale x 16 x i8>, 2) @caller_tuple_return(
 ; RV64-NEXT:    sd ra, 8(sp) # 8-byte Folded Spill
 ; RV64-NEXT:    .cfi_offset ra, -8
 ; RV64-NEXT:    call callee_tuple_return
+; RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; RV64-NEXT:    vmv2r.v v6, v8
 ; RV64-NEXT:    vmv2r.v v8, v10
 ; RV64-NEXT:    vmv2r.v v10, v6
@@ -144,6 +146,7 @@ define void @caller_tuple_argument(target("riscv.vector.tuple", <vscale x 16 x i
 ; RV32-NEXT:    .cfi_def_cfa_offset 16
 ; RV32-NEXT:    sw ra, 12(sp) # 4-byte Folded Spill
 ; RV32-NEXT:    .cfi_offset ra, -4
+; RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; RV32-NEXT:    vmv2r.v v6, v8
 ; RV32-NEXT:    vmv2r.v v8, v10
 ; RV32-NEXT:    vmv2r.v v10, v6
@@ -160,6 +163,7 @@ define void @caller_tuple_argument(target("riscv.vector.tuple", <vscale x 16 x i
 ; RV64-NEXT:    .cfi_def_cfa_offset 16
 ; RV64-NEXT:    sd ra, 8(sp) # 8-byte Folded Spill
 ; RV64-NEXT:    .cfi_offset ra, -8
+; RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; RV64-NEXT:    vmv2r.v v6, v8
 ; RV64-NEXT:    vmv2r.v v8, v10
 ; RV64-NEXT:    vmv2r.v v10, v6
diff --git a/llvm/test/CodeGen/RISCV/rvv/ceil-vp.ll b/llvm/test/CodeGen/RISCV/rvv/ceil-vp.ll
index 7d0b0118a7272..a81f0ac982f56 100644
--- a/llvm/test/CodeGen/RISCV/rvv/ceil-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/ceil-vp.ll
@@ -117,8 +117,8 @@ declare <vscale x 4 x bfloat> @llvm.vp.ceil.nxv4bf16(<vscale x 4 x bfloat>, <vsc
 define <vscale x 4 x bfloat> @vp_ceil_vv_nxv4bf16(<vscale x 4 x bfloat> %va, <vscale x 4 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vp_ceil_vv_nxv4bf16:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v9, v0
 ; CHECK-NEXT:    vsetvli a1, zero, e16, m1, ta, ma
+; CHECK-NEXT:    vmv1r.v v9, v0
 ; CHECK-NEXT:    vfwcvtbf16.f.f.v v10, v8
 ; CHECK-NEXT:    lui a1, 307200
 ; CHECK-NEXT:    vsetvli zero, a0, e32, m2, ta, ma
@@ -169,8 +169,8 @@ declare <vscale x 8 x bfloat> @llvm.vp.ceil.nxv8bf16(<vscale x 8 x bfloat>, <vsc
 define <vscale x 8 x bfloat> @vp_ceil_vv_nxv8bf16(<vscale x 8 x bfloat> %va, <vscale x 8 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vp_ceil_vv_nxv8bf16:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vsetvli a1, zero, e16, m2, ta, ma
+; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vfwcvtbf16.f.f.v v12, v8
 ; CHECK-NEXT:    lui a1, 307200
 ; CHECK-NEXT:    vsetvli zero, a0, e32, m4, ta, ma
@@ -221,8 +221,8 @@ declare <vscale x 16 x bfloat> @llvm.vp.ceil.nxv16bf16(<vscale x 16 x bfloat>, <
 define <vscale x 16 x bfloat> @vp_ceil_vv_nxv16bf16(<vscale x 16 x bfloat> %va, <vscale x 16 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vp_ceil_vv_nxv16bf16:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v12, v0
 ; CHECK-NEXT:    vsetvli a1, zero, e16, m4, ta, ma
+; CHECK-NEXT:    vmv1r.v v12, v0
 ; CHECK-NEXT:    vfwcvtbf16.f.f.v v16, v8
 ; CHECK-NEXT:    lui a1, 307200
 ; CHECK-NEXT:    vsetvli zero, a0, e32, m8, ta, ma
@@ -279,9 +279,9 @@ define <vscale x 32 x bfloat> @vp_ceil_vv_nxv32bf16(<vscale x 32 x bfloat> %va,
 ; CHECK-NEXT:    slli a1, a1, 3
 ; CHECK-NEXT:    sub sp, sp, a1
 ; CHECK-NEXT:    .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x08, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 8 * vlenb
+; CHECK-NEXT:    vsetvli a1, zero, e16, m4, ta, ma
 ; CHECK-NEXT:    vmv1r.v v7, v0
 ; CHECK-NEXT:    csrr a2, vlenb
-; CHECK-NEXT:    vsetvli a1, zero, e16, m4, ta, ma
 ; CHECK-NEXT:    vfwcvtbf16.f.f.v v24, v12
 ; CHECK-NEXT:    lui a3, 307200
 ; CHECK-NEXT:    slli a1, a2, 1
@@ -582,8 +582,8 @@ define <vscale x 4 x half> @vp_ceil_vv_nxv4f16(<vscale x 4 x half> %va, <vscale
 ;
 ; ZVFHMIN-LABEL: vp_ceil_vv_nxv4f16:
 ; ZVFHMIN:       # %bb.0:
-; ZVFHMIN-NEXT:    vmv1r.v v9, v0
 ; ZVFHMIN-NEXT:    vsetvli a1, zero, e16, m1, ta, ma
+; ZVFHMIN-NEXT:    vmv1r.v v9, v0
 ; ZVFHMIN-NEXT:    vfwcvt.f.f.v v10, v8
 ; ZVFHMIN-NEXT:    lui a1, 307200
 ; ZVFHMIN-NEXT:    vsetvli zero, a0, e32, m2, ta, ma
@@ -649,10 +649,10 @@ declare <vscale x 8 x half> @llvm.vp.ceil.nxv8f16(<vscale x 8 x half>, <vscale x
 define <vscale x 8 x half> @vp_ceil_vv_nxv8f16(<vscale x 8 x half> %va, <vscale x 8 x i1> %m, i32 zeroext %evl) {
 ; ZVFH-LABEL: vp_ceil_vv_nxv8f16:
 ; ZVFH:       # %bb.0:
-; ZVFH-NEXT:    vmv1r.v v10, v0
-; ZVFH-NEXT:    lui a1, %hi(.LCPI18_0)
-; ZVFH-NEXT:    flh fa5, %lo(.LCPI18_0)(a1)
 ; ZVFH-NEXT:    vsetvli zero, a0, e16, m2, ta, ma
+; ZVFH-NEXT:    vmv1r.v v10, v0
+; ZVFH-NEXT:    lui a0, %hi(.LCPI18_0)
+; ZVFH-NEXT:    flh fa5, %lo(.LCPI18_0)(a0)
 ; ZVFH-NEXT:    vfabs.v v12, v8, v0.t
 ; ZVFH-NEXT:    vsetvli zero, zero, e16, m2, ta, mu
 ; ZVFH-NEXT:    vmflt.vf v10, v12, fa5, v0.t
@@ -668,8 +668,8 @@ define <vscale x 8 x half> @vp_ceil_vv_nxv8f16(<vscale x 8 x half> %va, <vscale
 ;
 ; ZVFHMIN-LABEL: vp_ceil_vv_nxv8f16:
 ; ZVFHMIN:       # %bb.0:
-; ZVFHMIN-NEXT:    vmv1r.v v10, v0
 ; ZVFHMIN-NEXT:    vsetvli a1, zero, e16, m2, ta, ma
+; ZVFHMIN-NEXT:    vmv1r.v v10, v0
 ; ZVFHMIN-NEXT:    vfwcvt.f.f.v v12, v8
 ; ZVFHMIN-NEXT:    lui a1, 307200
 ; ZVFHMIN-NEXT:    vsetvli zero, a0, e32, m4, ta, ma
@@ -735,10 +735,10 @@ declare <vscale x 16 x half> @llvm.vp.ceil.nxv16f16(<vscale x 16 x half>, <vscal
 define <vscale x 16 x half> @vp_ceil_vv_nxv16f16(<vscale x 16 x half> %va, <vscale x 16 x i1> %m, i32 zeroext %evl) {
 ; ZVFH-LABEL: vp_ceil_vv_nxv16f16:
 ; ZVFH:       # %bb.0:
-; ZVFH-NEXT:    vmv1r.v v12, v0
-; ZVFH-NEXT:    lui a1, %hi(.LCPI20_0)
-; ZVFH-NEXT:    flh fa5, %lo(.LCPI20_0)(a1)
 ; ZVFH-NEXT:    vsetvli zero, a0, e16, m4, ta, ma
+; ZVFH-NEXT:    vmv1r.v v12, v0
+; ZVFH-NEXT:    lui a0, %hi(.LCPI20_0)
+; ZVFH-NEXT:    flh fa5, %lo(.LCPI20_0)(a0)
 ; ZVFH-NEXT:    vfabs.v v16, v8, v0.t
 ; ZVFH-NEXT:    vsetvli zero, zero, e16, m4, ta, mu
 ; ZVFH-NEXT:    vmflt.vf v12, v16, fa5, v0.t
@@ -754,8 +754,8 @@ define <vscale x 16 x half> @vp_ceil_vv_nxv16f16(<vscale x 16 x half> %va, <vsca
 ;
 ; ZVFHMIN-LABEL: vp_ceil_vv_nxv16f16:
 ; ZVFHMIN:       # %bb.0:
-; ZVFHMIN-NEXT:    vmv1r.v v12, v0
 ; ZVFHMIN-NEXT:    vsetvli a1, zero, e16, m4, ta, ma
+; ZVFHMIN-NEXT:    vmv1r.v v12, v0
 ; ZVFHMIN-NEXT:    vfwcvt.f.f.v v16, v8
 ; ZVFHMIN-NEXT:    lui a1, 307200
 ; ZVFHMIN-NEXT:    vsetvli zero, a0, e32, m8, ta, ma
@@ -821,10 +821,10 @@ declare <vscale x 32 x half> @llvm.vp.ceil.nxv32f16(<vscale x 32 x half>, <vscal
 define <vscale x 32 x half> @vp_ceil_vv_nxv32f16(<vscale x 32 x half> %va, <vscale x 32 x i1> %m, i32 zeroext %evl) {
 ; ZVFH-LABEL: vp_ceil_vv_nxv32f16:
 ; ZVFH:       # %bb.0:
-; ZVFH-NEXT:    vmv1r.v v16, v0
-; ZVFH-NEXT:    lui a1, %hi(.LCPI22_0)
-; ZVFH-NEXT:    flh fa5, %lo(.LCPI22_0)(a1)
 ; ZVFH-NEXT:    vsetvli zero, a0, e16, m8, ta, ma
+; ZVFH-NEXT:    vmv1r.v v16, v0
+; ZVFH-NEXT:    lui a0, %hi(.LCPI22_0)
+; ZVFH-NEXT:    flh fa5, %lo(.LCPI22_0)(a0)
 ; ZVFH-NEXT:    vfabs.v v24, v8, v0.t
 ; ZVFH-NEXT:    vsetvli zero, zero, e16, m8, ta, mu
 ; ZVFH-NEXT:    vmflt.vf v16, v24, fa5, v0.t
@@ -846,9 +846,9 @@ define <vscale x 32 x half> @vp_ceil_vv_nxv32f16(<vscale x 32 x half> %va, <vsca
 ; ZVFHMIN-NEXT:    slli a1, a1, 3
 ; ZVFHMIN-NEXT:    sub sp, sp, a1
 ; ZVFHMIN-NEXT:    .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x08, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 8 * vlenb
+; ZVFHMIN-NEXT:    vsetvli a1, zero, e16, m4, ta, ma
 ; ZVFHMIN-NEXT:    vmv1r.v v7, v0
 ; ZVFHMIN-NEXT:    csrr a2, vlenb
-; ZVFHMIN-NEXT:    vsetvli a1, zero, e16, m4, ta, ma
 ; ZVFHMIN-NEXT:    vfwcvt.f.f.v v24, v12
 ; ZVFHMIN-NEXT:    lui a3, 307200
 ; ZVFHMIN-NEXT:    slli a1, a2, 1
@@ -1068,8 +1068,8 @@ declare <vscale x 4 x float> @llvm.vp.ceil.nxv4f32(<vscale x 4 x float>, <vscale
 define <vscale x 4 x float> @vp_ceil_vv_nxv4f32(<vscale x 4 x float> %va, <vscale x 4 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vp_ceil_vv_nxv4f32:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e32, m2, ta, ma
+; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vfabs.v v12, v8, v0.t
 ; CHECK-NEXT:    lui a0, 307200
 ; CHECK-NEXT:    fmv.w.x fa5, a0
@@ -1112,8 +1112,8 @@ declare <vscale x 8 x float> @llvm.vp.ceil.nxv8f32(<vscale x 8 x float>, <vscale
 define <vscale x 8 x float> @vp_ceil_vv_nxv8f32(<vscale x 8 x float> %va, <vscale x 8 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vp_ceil_vv_nxv8f32:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v12, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e32, m4, ta, ma
+; CHECK-NEXT:    vmv1r.v v12, v0
 ; CHECK-NEXT:    vfabs.v v16, v8, v0.t
 ; CHECK-NEXT:    lui a0, 307200
 ; CHECK-NEXT:    fmv.w.x fa5, a0
@@ -1156,8 +1156,8 @@ declare <vscale x 16 x float> @llvm.vp.ceil.nxv16f32(<vscale x 16 x float>, <vsc
 define <vscale x 16 x float> @vp_ceil_vv_nxv16f32(<vscale x 16 x float> %va, <vscale x 16 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vp_ceil_vv_nxv16f32:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v16, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e32, m8, ta, ma
+; CHECK-NEXT:    vmv1r.v v16, v0
 ; CHECK-NEXT:    vfabs.v v24, v8, v0.t
 ; CHECK-NEXT:    lui a0, 307200
 ; CHECK-NEXT:    fmv.w.x fa5, a0
@@ -1242,10 +1242,10 @@ declare <vscale x 2 x double> @llvm.vp.ceil.nxv2f64(<vscale x 2 x double>, <vsca
 define <vscale x 2 x double> @vp_ceil_vv_nxv2f64(<vscale x 2 x double> %va, <vscale x 2 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vp_ceil_vv_nxv2f64:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v10, v0
-; CHECK-NEXT:    lui a1, %hi(.LCPI36_0)
-; CHECK-NEXT:    fld fa5, %lo(.LCPI36_0)(a1)
 ; CHECK-NEXT:    vsetvli zero, a0, e64, m2, ta, ma
+; CHECK-NEXT:    vmv1r.v v10, v0
+; CHECK-NEXT:    lui a0, %hi(.LCPI36_0)
+; CHECK-NEXT:    fld fa5, %lo(.LCPI36_0)(a0)
 ; CHECK-NEXT:    vfabs.v v12, v8, v0.t
 ; CHECK-NEXT:    vsetvli zero, zero, e64, m2, ta, mu
 ; CHECK-NEXT:    vmflt.vf v10, v12, fa5, v0.t
@@ -1286,10 +1286,10 @@ declare <vscale x 4 x double> @llvm.vp.ceil.nxv4f64(<vscale x 4 x double>, <vsca
 define <vscale x 4 x double> @vp_ceil_vv_nxv4f64(<vscale x 4 x double> %va, <vscale x 4 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vp_ceil_vv_nxv4f64:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v12, v0
-; CHECK-NEXT:    lui a1, %hi(.LCPI38_0)
-; CHECK-NEXT:    fld fa5, %lo(.LCPI38_0)(a1)
 ; CHECK-NEXT:    vsetvli zero, a0, e64, m4, ta, ma
+; CHECK-NEXT:    vmv1r.v v12, v0
+; CHECK-NEXT:    lui a0, %hi(.LCPI38_0)
+; CHECK-NEXT:    fld fa5, %lo(.LCPI38_0)(a0)
 ; CHECK-NEXT:    vfabs.v v16, v8, v0.t
 ; CHECK-NEXT:    vsetvli zero, zero, e64, m4, ta, mu
 ; CHECK-NEXT:    vmflt.vf v12, v16, fa5, v0.t
@@ -1330,10 +1330,10 @@ declare <vscale x 7 x double> @llvm.vp.ceil.nxv7f64(<vscale x 7 x double>, <vsca
 define <vscale x 7 x double> @vp_ceil_vv_nxv7f64(<vscale x 7 x double> %va, <vscale x 7 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vp_ceil_vv_nxv7f64:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v16, v0
-; CHECK-NEXT:    lui a1, %hi(.LCPI40_0)
-; CHECK-NEXT:    fld fa5, %lo(.LCPI40_0)(a1)
 ; CHECK-NEXT:    vsetvli zero, a0, e64, m8, ta, ma
+; CHECK-NEXT:    vmv1r.v v16, v0
+; CHECK-NEXT:    lui a0, %hi(.LCPI40_0)
+; CHECK-NEXT:    fld fa5, %lo(.LCPI40_0)(a0)
 ; CHECK-NEXT:    vfabs.v v24, v8, v0.t
 ; CHECK-NEXT:    vsetvli zero, zero, e64, m8, ta, mu
 ; CHECK-NEXT:    vmflt.vf v16, v24, fa5, v0.t
@@ -1374,10 +1374,10 @@ declare <vscale x 8 x double> @llvm.vp.ceil.nxv8f64(<vscale x 8 x double>, <vsca
 define <vscale x 8 x double> @vp_ceil_vv_nxv8f64(<vscale x 8 x double> %va, <vscale x 8 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vp_ceil_vv_nxv8f64:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v16, v0
-; CHECK-NEXT:    lui a1, %hi(.LCPI42_0)
-; CHECK-NEXT:    fld fa5, %lo(.LCPI42_0)(a1)
 ; CHECK-NEXT:    vsetvli zero, a0, e64, m8, ta, ma
+; CHECK-NEXT:    vmv1r.v v16, v0
+; CHECK-NEXT:    lui a0, %hi(.LCPI42_0)
+; CHECK-NEXT:    fld fa5, %lo(.LCPI42_0)(a0)
 ; CHECK-NEXT:    vfabs.v v24, v8, v0.t
 ; CHECK-NEXT:    vsetvli zero, zero, e64, m8, ta, mu
 ; CHECK-NEXT:    vmflt.vf v16, v24, fa5, v0.t
@@ -1425,13 +1425,13 @@ define <vscale x 16 x double> @vp_ceil_vv_nxv16f64(<vscale x 16 x double> %va, <
 ; CHECK-NEXT:    slli a1, a1, 3
 ; CHECK-NEXT:    sub sp, sp, a1
 ; CHECK-NEXT:    .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x08, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 8 * vlenb
+; CHECK-NEXT:    vsetvli a1, zero, e8, mf4, ta, ma
 ; CHECK-NEXT:    vmv1r.v v7, v0
 ; CHECK-NEXT:    csrr a1, vlenb
 ; CHECK-NEXT:    lui a2, %hi(.LCPI44_0)
 ; CHECK-NEXT:    srli a3, a1, 3
 ; CHECK-NEXT:    fld fa5, %lo(.LCPI44_0)(a2)
 ; CHECK-NEXT:    sub a2, a0, a1
-; CHECK-NEXT:    vsetvli a4, zero, e8, mf4, ta, ma
 ; CHECK-NEXT:    vslidedown.vx v6, v0, a3
 ; CHECK-NEXT:    sltu a3, a0, a2
 ; CHECK-NEXT:    addi a3, a3, -1
diff --git a/llvm/test/CodeGen/RISCV/rvv/compressstore.ll b/llvm/test/CodeGen/RISCV/rvv/compressstore.ll
index a407cd048ffe3..08bf82d4bc796 100644
--- a/llvm/test/CodeGen/RISCV/rvv/compressstore.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/compressstore.ll
@@ -197,9 +197,9 @@ entry:
 define void @test_compresstore_v256i8(ptr %p, <256 x i1> %mask, <256 x i8> %data) {
 ; RV64-LABEL: test_compresstore_v256i8:
 ; RV64:       # %bb.0: # %entry
+; RV64-NEXT:    vsetivli zero, 1, e64, m1, ta, ma
 ; RV64-NEXT:    vmv1r.v v7, v8
 ; RV64-NEXT:    li a2, 128
-; RV64-NEXT:    vsetivli zero, 1, e64, m1, ta, ma
 ; RV64-NEXT:    vslidedown.vi v9, v0, 1
 ; RV64-NEXT:    vmv.x.s a3, v0
 ; RV64-NEXT:    vsetvli zero, a2, e8, m8, ta, ma
@@ -230,9 +230,9 @@ define void @test_compresstore_v256i8(ptr %p, <256 x i1> %mask, <256 x i8> %data
 ; RV32-NEXT:    slli a2, a2, 3
 ; RV32-NEXT:    sub sp, sp, a2
 ; RV32-NEXT:    .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x08, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 8 * vlenb
+; RV32-NEXT:    vsetivli zero, 1, e64, m1, ta, ma
 ; RV32-NEXT:    vmv8r.v v24, v16
 ; RV32-NEXT:    li a2, 128
-; RV32-NEXT:    vsetivli zero, 1, e64, m1, ta, ma
 ; RV32-NEXT:    vslidedown.vi v9, v0, 1
 ; RV32-NEXT:    li a3, 32
 ; RV32-NEXT:    vmv.x.s a4, v0
diff --git a/llvm/test/CodeGen/RISCV/rvv/constant-folding-crash.ll b/llvm/test/CodeGen/RISCV/rvv/constant-folding-crash.ll
index ad176df71397e..f6c26bbba89fe 100644
--- a/llvm/test/CodeGen/RISCV/rvv/constant-folding-crash.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/constant-folding-crash.ll
@@ -18,11 +18,11 @@
 define void @constant_folding_crash(ptr %v54, <4 x ptr> %lanes.a, <4 x ptr> %lanes.b, <4 x i1> %sel) {
 ; RV32-LABEL: constant_folding_crash:
 ; RV32:       # %bb.0: # %entry
+; RV32-NEXT:    vsetivli zero, 4, e8, mf4, ta, ma
 ; RV32-NEXT:    vmv1r.v v10, v0
 ; RV32-NEXT:    lw a0, 8(a0)
 ; RV32-NEXT:    andi a0, a0, 1
 ; RV32-NEXT:    seqz a0, a0
-; RV32-NEXT:    vsetivli zero, 4, e8, mf4, ta, ma
 ; RV32-NEXT:    vmv.v.x v11, a0
 ; RV32-NEXT:    vmsne.vi v0, v11, 0
 ; RV32-NEXT:    vsetvli zero, zero, e32, m1, ta, ma
@@ -43,11 +43,11 @@ define void @constant_folding_crash(ptr %v54, <4 x ptr> %lanes.a, <4 x ptr> %lan
 ;
 ; RV64-LABEL: constant_folding_crash:
 ; RV64:       # %bb.0: # %entry
+; RV64-NEXT:    vsetivli zero, 4, e8, mf4, ta, ma
 ; RV64-NEXT:    vmv1r.v v12, v0
 ; RV64-NEXT:    ld a0, 8(a0)
 ; RV64-NEXT:    andi a0, a0, 1
 ; RV64-NEXT:    seqz a0, a0
-; RV64-NEXT:    vsetivli zero, 4, e8, mf4, ta, ma
 ; RV64-NEXT:    vmv.v.x v13, a0
 ; RV64-NEXT:    vmsne.vi v0, v13, 0
 ; RV64-NEXT:    vsetvli zero, zero, e64, m2, ta, ma
diff --git a/llvm/test/CodeGen/RISCV/rvv/ctlz-vp.ll b/llvm/test/CodeGen/RISCV/rvv/ctlz-vp.ll
index f56a792fdef6a..ce4bc48dff042 100644
--- a/llvm/test/CodeGen/RISCV/rvv/ctlz-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/ctlz-vp.ll
@@ -1235,13 +1235,13 @@ declare <vscale x 16 x i64> @llvm.vp.ctlz.nxv16i64(<vscale x 16 x i64>, i1 immar
 define <vscale x 16 x i64> @vp_ctlz_nxv16i64(<vscale x 16 x i64> %va, <vscale x 16 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vp_ctlz_nxv16i64:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli a1, zero, e8, mf4, ta, ma
 ; CHECK-NEXT:    vmv1r.v v24, v0
 ; CHECK-NEXT:    csrr a1, vlenb
 ; CHECK-NEXT:    fsrmi a4, 1
 ; CHECK-NEXT:    li a2, 52
 ; CHECK-NEXT:    srli a3, a1, 3
 ; CHECK-NEXT:    sub a5, a0, a1
-; CHECK-NEXT:    vsetvli a6, zero, e8, mf4, ta, ma
 ; CHECK-NEXT:    vslidedown.vx v0, v0, a3
 ; CHECK-NEXT:    sltu a3, a0, a5
 ; CHECK-NEXT:    addi a3, a3, -1
@@ -1270,11 +1270,11 @@ define <vscale x 16 x i64> @vp_ctlz_nxv16i64(<vscale x 16 x i64> %va, <vscale x
 ;
 ; CHECK-ZVBB-LABEL: vp_ctlz_nxv16i64:
 ; CHECK-ZVBB:       # %bb.0:
+; CHECK-ZVBB-NEXT:    vsetvli a1, zero, e8, mf4, ta, ma
 ; CHECK-ZVBB-NEXT:    vmv1r.v v24, v0
 ; CHECK-ZVBB-NEXT:    csrr a1, vlenb
 ; CHECK-ZVBB-NEXT:    srli a2, a1, 3
 ; CHECK-ZVBB-NEXT:    sub a3, a0, a1
-; CHECK-ZVBB-NEXT:    vsetvli a4, zero, e8, mf4, ta, ma
 ; CHECK-ZVBB-NEXT:    vslidedown.vx v0, v0, a2
 ; CHECK-ZVBB-NEXT:    sltu a2, a0, a3
 ; CHECK-ZVBB-NEXT:    addi a2, a2, -1
@@ -2465,12 +2465,12 @@ define <vscale x 8 x i64> @vp_ctlz_zero_undef_nxv8i64_unmasked(<vscale x 8 x i64
 define <vscale x 16 x i64> @vp_ctlz_zero_undef_nxv16i64(<vscale x 16 x i64> %va, <vscale x 16 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vp_ctlz_zero_undef_nxv16i64:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli a1, zero, e8, mf4, ta, ma
 ; CHECK-NEXT:    vmv1r.v v24, v0
 ; CHECK-NEXT:    csrr a1, vlenb
 ; CHECK-NEXT:    fsrmi a3, 1
 ; CHECK-NEXT:    srli a2, a1, 3
 ; CHECK-NEXT:    sub a4, a0, a1
-; CHECK-NEXT:    vsetvli a5, zero, e8, mf4, ta, ma
 ; CHECK-NEXT:    vslidedown.vx v0, v0, a2
 ; CHECK-NEXT:    sltu a2, a0, a4
 ; CHECK-NEXT:    addi a2, a2, -1
@@ -2497,11 +2497,11 @@ define <vscale x 16 x i64> @vp_ctlz_zero_undef_nxv16i64(<vscale x 16 x i64> %va,
 ;
 ; CHECK-ZVBB-LABEL: vp_ctlz_zero_undef_nxv16i64:
 ; CHECK-ZVBB:       # %bb.0:
+; CHECK-ZVBB-NEXT:    vsetvli a1, zero, e8, mf4, ta, ma
 ; CHECK-ZVBB-NEXT:    vmv1r.v v24, v0
 ; CHECK-ZVBB-NEXT:    csrr a1, vlenb
 ; CHECK-ZVBB-NEXT:    srli a2, a1, 3
 ; CHECK-ZVBB-NEXT:    sub a3, a0, a1
-; CHECK-ZVBB-NEXT:    vsetvli a4, zero, e8, mf4, ta, ma
 ; CHECK-ZVBB-NEXT:    vslidedown.vx v0, v0, a2
 ; CHECK-ZVBB-NEXT:    sltu a2, a0, a3
 ; CHECK-ZVBB-NEXT:    addi a2, a2, -1
diff --git a/llvm/test/CodeGen/RISCV/rvv/ctpop-vp.ll b/llvm/test/CodeGen/RISCV/rvv/ctpop-vp.ll
index 9e75dc9dccffd..52ddd9ab2f832 100644
--- a/llvm/test/CodeGen/RISCV/rvv/ctpop-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/ctpop-vp.ll
@@ -2022,6 +2022,7 @@ define <vscale x 16 x i64> @vp_ctpop_nxv16i64(<vscale x 16 x i64> %va, <vscale x
 ; RV32-NEXT:    mul a1, a1, a2
 ; RV32-NEXT:    sub sp, sp, a1
 ; RV32-NEXT:    .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x30, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 48 * vlenb
+; RV32-NEXT:    vsetvli a1, zero, e8, mf4, ta, ma
 ; RV32-NEXT:    vmv1r.v v7, v0
 ; RV32-NEXT:    csrr a1, vlenb
 ; RV32-NEXT:    li a2, 24
@@ -2032,7 +2033,6 @@ define <vscale x 16 x i64> @vp_ctpop_nxv16i64(<vscale x 16 x i64> %va, <vscale x
 ; RV32-NEXT:    csrr a1, vlenb
 ; RV32-NEXT:    lui a2, 349525
 ; RV32-NEXT:    srli a3, a1, 3
-; RV32-NEXT:    vsetvli a4, zero, e8, mf4, ta, ma
 ; RV32-NEXT:    vslidedown.vx v0, v0, a3
 ; RV32-NEXT:    sub a3, a0, a1
 ; RV32-NEXT:    addi a2, a2, 1365
@@ -2294,11 +2294,11 @@ define <vscale x 16 x i64> @vp_ctpop_nxv16i64(<vscale x 16 x i64> %va, <vscale x
 ;
 ; CHECK-ZVBB-LABEL: vp_ctpop_nxv16i64:
 ; CHECK-ZVBB:       # %bb.0:
+; CHECK-ZVBB-NEXT:    vsetvli a1, zero, e8, mf4, ta, ma
 ; CHECK-ZVBB-NEXT:    vmv1r.v v24, v0
 ; CHECK-ZVBB-NEXT:    csrr a1, vlenb
 ; CHECK-ZVBB-NEXT:    srli a2, a1, 3
 ; CHECK-ZVBB-NEXT:    sub a3, a0, a1
-; CHECK-ZVBB-NEXT:    vsetvli a4, zero, e8, mf4, ta, ma
 ; CHECK-ZVBB-NEXT:    vslidedown.vx v0, v0, a2
 ; CHECK-ZVBB-NEXT:    sltu a2, a0, a3
 ; CHECK-ZVBB-NEXT:    addi a2, a2, -1
diff --git a/llvm/test/CodeGen/RISCV/rvv/cttz-vp.ll b/llvm/test/CodeGen/RISCV/rvv/cttz-vp.ll
index 9e6295b664417..766717d92a749 100644
--- a/llvm/test/CodeGen/RISCV/rvv/cttz-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/cttz-vp.ll
@@ -2246,6 +2246,7 @@ define <vscale x 16 x i64> @vp_cttz_nxv16i64(<vscale x 16 x i64> %va, <vscale x
 ; RV32-NEXT:    mul a1, a1, a2
 ; RV32-NEXT:    sub sp, sp, a1
 ; RV32-NEXT:    .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x38, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 56 * vlenb
+; RV32-NEXT:    vsetvli a1, zero, e8, mf4, ta, ma
 ; RV32-NEXT:    vmv1r.v v24, v0
 ; RV32-NEXT:    csrr a1, vlenb
 ; RV32-NEXT:    slli a1, a1, 5
@@ -2256,7 +2257,6 @@ define <vscale x 16 x i64> @vp_cttz_nxv16i64(<vscale x 16 x i64> %va, <vscale x
 ; RV32-NEXT:    li a2, 1
 ; RV32-NEXT:    srli a3, a1, 3
 ; RV32-NEXT:    sub a4, a0, a1
-; RV32-NEXT:    vsetvli a5, zero, e8, mf4, ta, ma
 ; RV32-NEXT:    vslidedown.vx v0, v0, a3
 ; RV32-NEXT:    sltu a3, a0, a4
 ; RV32-NEXT:    addi a3, a3, -1
@@ -2499,6 +2499,7 @@ define <vscale x 16 x i64> @vp_cttz_nxv16i64(<vscale x 16 x i64> %va, <vscale x
 ; RV64-NEXT:    slli a1, a1, 4
 ; RV64-NEXT:    sub sp, sp, a1
 ; RV64-NEXT:    .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x10, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 16 * vlenb
+; RV64-NEXT:    vsetvli a1, zero, e8, mf4, ta, ma
 ; RV64-NEXT:    vmv1r.v v24, v0
 ; RV64-NEXT:    csrr a1, vlenb
 ; RV64-NEXT:    slli a1, a1, 3
@@ -2517,7 +2518,6 @@ define <vscale x 16 x i64> @vp_cttz_nxv16i64(<vscale x 16 x i64> %va, <vscale x
 ; RV64-NEXT:    addiw a4, a4, 819
 ; RV64-NEXT:    addiw a5, a5, -241
 ; RV64-NEXT:    addiw t1, a6, 257
-; RV64-NEXT:    vsetvli a6, zero, e8, mf4, ta, ma
 ; RV64-NEXT:    vslidedown.vx v0, v0, a7
 ; RV64-NEXT:    slli a7, a3, 32
 ; RV64-NEXT:    add a7, a3, a7
@@ -2586,11 +2586,11 @@ define <vscale x 16 x i64> @vp_cttz_nxv16i64(<vscale x 16 x i64> %va, <vscale x
 ;
 ; CHECK-ZVBB-LABEL: vp_cttz_nxv16i64:
 ; CHECK-ZVBB:       # %bb.0:
+; CHECK-ZVBB-NEXT:    vsetvli a1, zero, e8, mf4, ta, ma
 ; CHECK-ZVBB-NEXT:    vmv1r.v v24, v0
 ; CHECK-ZVBB-NEXT:    csrr a1, vlenb
 ; CHECK-ZVBB-NEXT:    srli a2, a1, 3
 ; CHECK-ZVBB-NEXT:    sub a3, a0, a1
-; CHECK-ZVBB-NEXT:    vsetvli a4, zero, e8, mf4, ta, ma
 ; CHECK-ZVBB-NEXT:    vslidedown.vx v0, v0, a2
 ; CHECK-ZVBB-NEXT:    sltu a2, a0, a3
 ; CHECK-ZVBB-NEXT:    addi a2, a2, -1
@@ -4002,6 +4002,7 @@ define <vscale x 16 x i64> @vp_cttz_zero_undef_nxv16i64(<vscale x 16 x i64> %va,
 ; CHECK-NEXT:    slli a1, a1, 4
 ; CHECK-NEXT:    sub sp, sp, a1
 ; CHECK-NEXT:    .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x10, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 16 * vlenb
+; CHECK-NEXT:    vsetvli a1, zero, e8, mf4, ta, ma
 ; CHECK-NEXT:    vmv1r.v v24, v0
 ; CHECK-NEXT:    csrr a1, vlenb
 ; CHECK-NEXT:    slli a1, a1, 3
@@ -4012,7 +4013,6 @@ define <vscale x 16 x i64> @vp_cttz_zero_undef_nxv16i64(<vscale x 16 x i64> %va,
 ; CHECK-NEXT:    fsrmi a3, 1
 ; CHECK-NEXT:    srli a2, a1, 3
 ; CHECK-NEXT:    sub a4, a0, a1
-; CHECK-NEXT:    vsetvli a5, zero, e8, mf4, ta, ma
 ; CHECK-NEXT:    vslidedown.vx v0, v0, a2
 ; CHECK-NEXT:    sltu a2, a0, a4
 ; CHECK-NEXT:    addi a2, a2, -1
@@ -4057,11 +4057,11 @@ define <vscale x 16 x i64> @vp_cttz_zero_undef_nxv16i64(<vscale x 16 x i64> %va,
 ;
 ; CHECK-ZVBB-LABEL: vp_cttz_zero_undef_nxv16i64:
 ; CHECK-ZVBB:       # %bb.0:
+; CHECK-ZVBB-NEXT:    vsetvli a1, zero, e8, mf4, ta, ma
 ; CHECK-ZVBB-NEXT:    vmv1r.v v24, v0
 ; CHECK-ZVBB-NEXT:    csrr a1, vlenb
 ; CHECK-ZVBB-NEXT:    srli a2, a1, 3
 ; CHECK-ZVBB-NEXT:    sub a3, a0, a1
-; CHECK-ZVBB-NEXT:    vsetvli a4, zero, e8, mf4, ta, ma
 ; CHECK-ZVBB-NEXT:    vslidedown.vx v0, v0, a2
 ; CHECK-ZVBB-NEXT:    sltu a2, a0, a3
 ; CHECK-ZVBB-NEXT:    addi a2, a2, -1
diff --git a/llvm/test/CodeGen/RISCV/rvv/expandload.ll b/llvm/test/CodeGen/RISCV/rvv/expandload.ll
index b32d85bb1943a..a35cf14203f78 100644
--- a/llvm/test/CodeGen/RISCV/rvv/expandload.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/expandload.ll
@@ -227,9 +227,9 @@ define <256 x i8> @test_expandload_v256i8(ptr %base, <256 x i1> %mask, <256 x i8
 ; CHECK-RV32-NEXT:    add a2, sp, a2
 ; CHECK-RV32-NEXT:    addi a2, a2, 16
 ; CHECK-RV32-NEXT:    vs8r.v v16, (a2) # Unknown-size Folded Spill
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e64, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv1r.v v7, v8
 ; CHECK-RV32-NEXT:    li a2, 128
-; CHECK-RV32-NEXT:    vsetivli zero, 1, e64, m1, ta, ma
 ; CHECK-RV32-NEXT:    vslidedown.vi v9, v0, 1
 ; CHECK-RV32-NEXT:    li a3, 32
 ; CHECK-RV32-NEXT:    vmv.x.s a4, v0
@@ -338,9 +338,9 @@ define <256 x i8> @test_expandload_v256i8(ptr %base, <256 x i1> %mask, <256 x i8
 ; CHECK-RV64-NEXT:    add a2, sp, a2
 ; CHECK-RV64-NEXT:    addi a2, a2, 16
 ; CHECK-RV64-NEXT:    vs8r.v v16, (a2) # Unknown-size Folded Spill
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e64, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv1r.v v7, v8
 ; CHECK-RV64-NEXT:    li a2, 128
-; CHECK-RV64-NEXT:    vsetivli zero, 1, e64, m1, ta, ma
 ; CHECK-RV64-NEXT:    vslidedown.vi v9, v0, 1
 ; CHECK-RV64-NEXT:    vmv.x.s a3, v0
 ; CHECK-RV64-NEXT:    vsetvli zero, a2, e8, m8, ta, ma
@@ -1626,8 +1626,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    bgez a1, .LBB61_30
 ; CHECK-RV32-NEXT:  .LBB61_29: # %cond.load109
 ; CHECK-RV32-NEXT:    lbu a1, 0(a0)
-; CHECK-RV32-NEXT:    vmv8r.v v16, v8
 ; CHECK-RV32-NEXT:    vsetivli zero, 29, e8, m1, tu, ma
+; CHECK-RV32-NEXT:    vmv8r.v v16, v8
 ; CHECK-RV32-NEXT:    vmv.s.x v9, a1
 ; CHECK-RV32-NEXT:    vslideup.vi v8, v9, 28
 ; CHECK-RV32-NEXT:    addi a0, a0, 1
@@ -1639,8 +1639,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    bgez a2, .LBB61_32
 ; CHECK-RV32-NEXT:  # %bb.31: # %cond.load113
 ; CHECK-RV32-NEXT:    lbu a2, 0(a0)
-; CHECK-RV32-NEXT:    vmv8r.v v16, v8
 ; CHECK-RV32-NEXT:    vsetivli zero, 30, e8, m1, tu, ma
+; CHECK-RV32-NEXT:    vmv8r.v v16, v8
 ; CHECK-RV32-NEXT:    vmv.s.x v9, a2
 ; CHECK-RV32-NEXT:    vslideup.vi v8, v9, 29
 ; CHECK-RV32-NEXT:    addi a0, a0, 1
@@ -1786,9 +1786,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    bgez a3, .LBB61_66
 ; CHECK-RV32-NEXT:  .LBB61_65: # %cond.load241
 ; CHECK-RV32-NEXT:    lbu a3, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v16, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v9, a3
 ; CHECK-RV32-NEXT:    li a3, 62
 ; CHECK-RV32-NEXT:    li a4, 61
@@ -1939,9 +1938,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    bgez a2, .LBB61_100
 ; CHECK-RV32-NEXT:  .LBB61_99: # %cond.load369
 ; CHECK-RV32-NEXT:    lbu a2, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v24, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v10, a2
 ; CHECK-RV32-NEXT:    li a2, 94
 ; CHECK-RV32-NEXT:    li a4, 93
@@ -2092,9 +2090,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    bgez a3, .LBB61_134
 ; CHECK-RV32-NEXT:  .LBB61_133: # %cond.load497
 ; CHECK-RV32-NEXT:    lbu a3, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v16, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v10, a3
 ; CHECK-RV32-NEXT:    li a3, 126
 ; CHECK-RV32-NEXT:    li a4, 125
@@ -2245,9 +2242,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    bgez a2, .LBB61_168
 ; CHECK-RV32-NEXT:  .LBB61_167: # %cond.load625
 ; CHECK-RV32-NEXT:    lbu a2, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v24, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v12, a2
 ; CHECK-RV32-NEXT:    li a2, 158
 ; CHECK-RV32-NEXT:    li a4, 157
@@ -2398,9 +2394,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    bgez a3, .LBB61_202
 ; CHECK-RV32-NEXT:  .LBB61_201: # %cond.load753
 ; CHECK-RV32-NEXT:    lbu a3, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v16, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v12, a3
 ; CHECK-RV32-NEXT:    li a3, 190
 ; CHECK-RV32-NEXT:    li a4, 189
@@ -2551,9 +2546,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    bgez a2, .LBB61_236
 ; CHECK-RV32-NEXT:  .LBB61_235: # %cond.load881
 ; CHECK-RV32-NEXT:    lbu a2, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v24, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v12, a2
 ; CHECK-RV32-NEXT:    li a2, 222
 ; CHECK-RV32-NEXT:    li a4, 221
@@ -2704,9 +2698,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    bgez a3, .LBB61_270
 ; CHECK-RV32-NEXT:  .LBB61_269: # %cond.load1009
 ; CHECK-RV32-NEXT:    lbu a3, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v16, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v12, a3
 ; CHECK-RV32-NEXT:    li a3, 254
 ; CHECK-RV32-NEXT:    li a4, 253
@@ -3907,10 +3900,9 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_2
 ; CHECK-RV32-NEXT:  .LBB61_545: # %cond.load1
 ; CHECK-RV32-NEXT:    lbu a1, 0(a0)
+; CHECK-RV32-NEXT:    vsetivli zero, 2, e8, m1, tu, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v16, v8
-; CHECK-RV32-NEXT:    vsetvli zero, zero, e8, mf8, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v9, a1
-; CHECK-RV32-NEXT:    vsetivli zero, 2, e8, m1, tu, ma
 ; CHECK-RV32-NEXT:    vslideup.vi v8, v9, 1
 ; CHECK-RV32-NEXT:    addi a0, a0, 1
 ; CHECK-RV32-NEXT:    vmv1r.v v16, v8
@@ -3920,8 +3912,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_3
 ; CHECK-RV32-NEXT:  .LBB61_546: # %cond.load5
 ; CHECK-RV32-NEXT:    lbu a1, 0(a0)
-; CHECK-RV32-NEXT:    vmv8r.v v16, v8
 ; CHECK-RV32-NEXT:    vsetivli zero, 3, e8, m1, tu, ma
+; CHECK-RV32-NEXT:    vmv8r.v v16, v8
 ; CHECK-RV32-NEXT:    vmv.s.x v9, a1
 ; CHECK-RV32-NEXT:    vslideup.vi v8, v9, 2
 ; CHECK-RV32-NEXT:    addi a0, a0, 1
@@ -3932,8 +3924,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_4
 ; CHECK-RV32-NEXT:  .LBB61_547: # %cond.load9
 ; CHECK-RV32-NEXT:    lbu a1, 0(a0)
-; CHECK-RV32-NEXT:    vmv8r.v v16, v8
 ; CHECK-RV32-NEXT:    vsetivli zero, 4, e8, m1, tu, ma
+; CHECK-RV32-NEXT:    vmv8r.v v16, v8
 ; CHECK-RV32-NEXT:    vmv.s.x v9, a1
 ; CHECK-RV32-NEXT:    vslideup.vi v8, v9, 3
 ; CHECK-RV32-NEXT:    addi a0, a0, 1
@@ -3944,8 +3936,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_5
 ; CHECK-RV32-NEXT:  .LBB61_548: # %cond.load13
 ; CHECK-RV32-NEXT:    lbu a1, 0(a0)
-; CHECK-RV32-NEXT:    vmv8r.v v16, v8
 ; CHECK-RV32-NEXT:    vsetivli zero, 5, e8, m1, tu, ma
+; CHECK-RV32-NEXT:    vmv8r.v v16, v8
 ; CHECK-RV32-NEXT:    vmv.s.x v9, a1
 ; CHECK-RV32-NEXT:    vslideup.vi v8, v9, 4
 ; CHECK-RV32-NEXT:    addi a0, a0, 1
@@ -3956,8 +3948,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_6
 ; CHECK-RV32-NEXT:  .LBB61_549: # %cond.load17
 ; CHECK-RV32-NEXT:    lbu a1, 0(a0)
-; CHECK-RV32-NEXT:    vmv8r.v v16, v8
 ; CHECK-RV32-NEXT:    vsetivli zero, 6, e8, m1, tu, ma
+; CHECK-RV32-NEXT:    vmv8r.v v16, v8
 ; CHECK-RV32-NEXT:    vmv.s.x v9, a1
 ; CHECK-RV32-NEXT:    vslideup.vi v8, v9, 5
 ; CHECK-RV32-NEXT:    addi a0, a0, 1
@@ -3968,8 +3960,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_7
 ; CHECK-RV32-NEXT:  .LBB61_550: # %cond.load21
 ; CHECK-RV32-NEXT:    lbu a1, 0(a0)
-; CHECK-RV32-NEXT:    vmv8r.v v16, v8
 ; CHECK-RV32-NEXT:    vsetivli zero, 7, e8, m1, tu, ma
+; CHECK-RV32-NEXT:    vmv8r.v v16, v8
 ; CHECK-RV32-NEXT:    vmv.s.x v9, a1
 ; CHECK-RV32-NEXT:    vslideup.vi v8, v9, 6
 ; CHECK-RV32-NEXT:    addi a0, a0, 1
@@ -3980,8 +3972,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_8
 ; CHECK-RV32-NEXT:  .LBB61_551: # %cond.load25
 ; CHECK-RV32-NEXT:    lbu a1, 0(a0)
-; CHECK-RV32-NEXT:    vmv8r.v v16, v8
 ; CHECK-RV32-NEXT:    vsetivli zero, 8, e8, m1, tu, ma
+; CHECK-RV32-NEXT:    vmv8r.v v16, v8
 ; CHECK-RV32-NEXT:    vmv.s.x v9, a1
 ; CHECK-RV32-NEXT:    vslideup.vi v8, v9, 7
 ; CHECK-RV32-NEXT:    addi a0, a0, 1
@@ -3992,8 +3984,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_9
 ; CHECK-RV32-NEXT:  .LBB61_552: # %cond.load29
 ; CHECK-RV32-NEXT:    lbu a1, 0(a0)
-; CHECK-RV32-NEXT:    vmv8r.v v16, v8
 ; CHECK-RV32-NEXT:    vsetivli zero, 9, e8, m1, tu, ma
+; CHECK-RV32-NEXT:    vmv8r.v v16, v8
 ; CHECK-RV32-NEXT:    vmv.s.x v9, a1
 ; CHECK-RV32-NEXT:    vslideup.vi v8, v9, 8
 ; CHECK-RV32-NEXT:    addi a0, a0, 1
@@ -4004,8 +3996,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_10
 ; CHECK-RV32-NEXT:  .LBB61_553: # %cond.load33
 ; CHECK-RV32-NEXT:    lbu a1, 0(a0)
-; CHECK-RV32-NEXT:    vmv8r.v v16, v8
 ; CHECK-RV32-NEXT:    vsetivli zero, 10, e8, m1, tu, ma
+; CHECK-RV32-NEXT:    vmv8r.v v16, v8
 ; CHECK-RV32-NEXT:    vmv.s.x v9, a1
 ; CHECK-RV32-NEXT:    vslideup.vi v8, v9, 9
 ; CHECK-RV32-NEXT:    addi a0, a0, 1
@@ -4016,8 +4008,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_11
 ; CHECK-RV32-NEXT:  .LBB61_554: # %cond.load37
 ; CHECK-RV32-NEXT:    lbu a1, 0(a0)
-; CHECK-RV32-NEXT:    vmv8r.v v16, v8
 ; CHECK-RV32-NEXT:    vsetivli zero, 11, e8, m1, tu, ma
+; CHECK-RV32-NEXT:    vmv8r.v v16, v8
 ; CHECK-RV32-NEXT:    vmv.s.x v9, a1
 ; CHECK-RV32-NEXT:    vslideup.vi v8, v9, 10
 ; CHECK-RV32-NEXT:    addi a0, a0, 1
@@ -4028,8 +4020,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_12
 ; CHECK-RV32-NEXT:  .LBB61_555: # %cond.load41
 ; CHECK-RV32-NEXT:    lbu a1, 0(a0)
-; CHECK-RV32-NEXT:    vmv8r.v v16, v8
 ; CHECK-RV32-NEXT:    vsetivli zero, 12, e8, m1, tu, ma
+; CHECK-RV32-NEXT:    vmv8r.v v16, v8
 ; CHECK-RV32-NEXT:    vmv.s.x v9, a1
 ; CHECK-RV32-NEXT:    vslideup.vi v8, v9, 11
 ; CHECK-RV32-NEXT:    addi a0, a0, 1
@@ -4040,8 +4032,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_13
 ; CHECK-RV32-NEXT:  .LBB61_556: # %cond.load45
 ; CHECK-RV32-NEXT:    lbu a1, 0(a0)
-; CHECK-RV32-NEXT:    vmv8r.v v16, v8
 ; CHECK-RV32-NEXT:    vsetivli zero, 13, e8, m1, tu, ma
+; CHECK-RV32-NEXT:    vmv8r.v v16, v8
 ; CHECK-RV32-NEXT:    vmv.s.x v9, a1
 ; CHECK-RV32-NEXT:    vslideup.vi v8, v9, 12
 ; CHECK-RV32-NEXT:    addi a0, a0, 1
@@ -4052,8 +4044,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_14
 ; CHECK-RV32-NEXT:  .LBB61_557: # %cond.load49
 ; CHECK-RV32-NEXT:    lbu a1, 0(a0)
-; CHECK-RV32-NEXT:    vmv8r.v v16, v8
 ; CHECK-RV32-NEXT:    vsetivli zero, 14, e8, m1, tu, ma
+; CHECK-RV32-NEXT:    vmv8r.v v16, v8
 ; CHECK-RV32-NEXT:    vmv.s.x v9, a1
 ; CHECK-RV32-NEXT:    vslideup.vi v8, v9, 13
 ; CHECK-RV32-NEXT:    addi a0, a0, 1
@@ -4064,8 +4056,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_15
 ; CHECK-RV32-NEXT:  .LBB61_558: # %cond.load53
 ; CHECK-RV32-NEXT:    lbu a1, 0(a0)
-; CHECK-RV32-NEXT:    vmv8r.v v16, v8
 ; CHECK-RV32-NEXT:    vsetivli zero, 15, e8, m1, tu, ma
+; CHECK-RV32-NEXT:    vmv8r.v v16, v8
 ; CHECK-RV32-NEXT:    vmv.s.x v9, a1
 ; CHECK-RV32-NEXT:    vslideup.vi v8, v9, 14
 ; CHECK-RV32-NEXT:    addi a0, a0, 1
@@ -4076,8 +4068,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_16
 ; CHECK-RV32-NEXT:  .LBB61_559: # %cond.load57
 ; CHECK-RV32-NEXT:    lbu a1, 0(a0)
-; CHECK-RV32-NEXT:    vmv8r.v v16, v8
 ; CHECK-RV32-NEXT:    vsetivli zero, 16, e8, m1, tu, ma
+; CHECK-RV32-NEXT:    vmv8r.v v16, v8
 ; CHECK-RV32-NEXT:    vmv.s.x v9, a1
 ; CHECK-RV32-NEXT:    vslideup.vi v8, v9, 15
 ; CHECK-RV32-NEXT:    addi a0, a0, 1
@@ -4088,8 +4080,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_17
 ; CHECK-RV32-NEXT:  .LBB61_560: # %cond.load61
 ; CHECK-RV32-NEXT:    lbu a1, 0(a0)
-; CHECK-RV32-NEXT:    vmv8r.v v16, v8
 ; CHECK-RV32-NEXT:    vsetivli zero, 17, e8, m1, tu, ma
+; CHECK-RV32-NEXT:    vmv8r.v v16, v8
 ; CHECK-RV32-NEXT:    vmv.s.x v9, a1
 ; CHECK-RV32-NEXT:    vslideup.vi v8, v9, 16
 ; CHECK-RV32-NEXT:    addi a0, a0, 1
@@ -4100,8 +4092,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_18
 ; CHECK-RV32-NEXT:  .LBB61_561: # %cond.load65
 ; CHECK-RV32-NEXT:    lbu a1, 0(a0)
-; CHECK-RV32-NEXT:    vmv8r.v v16, v8
 ; CHECK-RV32-NEXT:    vsetivli zero, 18, e8, m1, tu, ma
+; CHECK-RV32-NEXT:    vmv8r.v v16, v8
 ; CHECK-RV32-NEXT:    vmv.s.x v9, a1
 ; CHECK-RV32-NEXT:    vslideup.vi v8, v9, 17
 ; CHECK-RV32-NEXT:    addi a0, a0, 1
@@ -4112,8 +4104,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_19
 ; CHECK-RV32-NEXT:  .LBB61_562: # %cond.load69
 ; CHECK-RV32-NEXT:    lbu a1, 0(a0)
-; CHECK-RV32-NEXT:    vmv8r.v v16, v8
 ; CHECK-RV32-NEXT:    vsetivli zero, 19, e8, m1, tu, ma
+; CHECK-RV32-NEXT:    vmv8r.v v16, v8
 ; CHECK-RV32-NEXT:    vmv.s.x v9, a1
 ; CHECK-RV32-NEXT:    vslideup.vi v8, v9, 18
 ; CHECK-RV32-NEXT:    addi a0, a0, 1
@@ -4124,8 +4116,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_20
 ; CHECK-RV32-NEXT:  .LBB61_563: # %cond.load73
 ; CHECK-RV32-NEXT:    lbu a1, 0(a0)
-; CHECK-RV32-NEXT:    vmv8r.v v16, v8
 ; CHECK-RV32-NEXT:    vsetivli zero, 20, e8, m1, tu, ma
+; CHECK-RV32-NEXT:    vmv8r.v v16, v8
 ; CHECK-RV32-NEXT:    vmv.s.x v9, a1
 ; CHECK-RV32-NEXT:    vslideup.vi v8, v9, 19
 ; CHECK-RV32-NEXT:    addi a0, a0, 1
@@ -4136,8 +4128,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_21
 ; CHECK-RV32-NEXT:  .LBB61_564: # %cond.load77
 ; CHECK-RV32-NEXT:    lbu a1, 0(a0)
-; CHECK-RV32-NEXT:    vmv8r.v v16, v8
 ; CHECK-RV32-NEXT:    vsetivli zero, 21, e8, m1, tu, ma
+; CHECK-RV32-NEXT:    vmv8r.v v16, v8
 ; CHECK-RV32-NEXT:    vmv.s.x v9, a1
 ; CHECK-RV32-NEXT:    vslideup.vi v8, v9, 20
 ; CHECK-RV32-NEXT:    addi a0, a0, 1
@@ -4148,8 +4140,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_22
 ; CHECK-RV32-NEXT:  .LBB61_565: # %cond.load81
 ; CHECK-RV32-NEXT:    lbu a1, 0(a0)
-; CHECK-RV32-NEXT:    vmv8r.v v16, v8
 ; CHECK-RV32-NEXT:    vsetivli zero, 22, e8, m1, tu, ma
+; CHECK-RV32-NEXT:    vmv8r.v v16, v8
 ; CHECK-RV32-NEXT:    vmv.s.x v9, a1
 ; CHECK-RV32-NEXT:    vslideup.vi v8, v9, 21
 ; CHECK-RV32-NEXT:    addi a0, a0, 1
@@ -4160,8 +4152,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_23
 ; CHECK-RV32-NEXT:  .LBB61_566: # %cond.load85
 ; CHECK-RV32-NEXT:    lbu a1, 0(a0)
-; CHECK-RV32-NEXT:    vmv8r.v v16, v8
 ; CHECK-RV32-NEXT:    vsetivli zero, 23, e8, m1, tu, ma
+; CHECK-RV32-NEXT:    vmv8r.v v16, v8
 ; CHECK-RV32-NEXT:    vmv.s.x v9, a1
 ; CHECK-RV32-NEXT:    vslideup.vi v8, v9, 22
 ; CHECK-RV32-NEXT:    addi a0, a0, 1
@@ -4172,8 +4164,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_24
 ; CHECK-RV32-NEXT:  .LBB61_567: # %cond.load89
 ; CHECK-RV32-NEXT:    lbu a1, 0(a0)
-; CHECK-RV32-NEXT:    vmv8r.v v16, v8
 ; CHECK-RV32-NEXT:    vsetivli zero, 24, e8, m1, tu, ma
+; CHECK-RV32-NEXT:    vmv8r.v v16, v8
 ; CHECK-RV32-NEXT:    vmv.s.x v9, a1
 ; CHECK-RV32-NEXT:    vslideup.vi v8, v9, 23
 ; CHECK-RV32-NEXT:    addi a0, a0, 1
@@ -4184,8 +4176,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_25
 ; CHECK-RV32-NEXT:  .LBB61_568: # %cond.load93
 ; CHECK-RV32-NEXT:    lbu a1, 0(a0)
-; CHECK-RV32-NEXT:    vmv8r.v v16, v8
 ; CHECK-RV32-NEXT:    vsetivli zero, 25, e8, m1, tu, ma
+; CHECK-RV32-NEXT:    vmv8r.v v16, v8
 ; CHECK-RV32-NEXT:    vmv.s.x v9, a1
 ; CHECK-RV32-NEXT:    vslideup.vi v8, v9, 24
 ; CHECK-RV32-NEXT:    addi a0, a0, 1
@@ -4196,8 +4188,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_26
 ; CHECK-RV32-NEXT:  .LBB61_569: # %cond.load97
 ; CHECK-RV32-NEXT:    lbu a1, 0(a0)
-; CHECK-RV32-NEXT:    vmv8r.v v16, v8
 ; CHECK-RV32-NEXT:    vsetivli zero, 26, e8, m1, tu, ma
+; CHECK-RV32-NEXT:    vmv8r.v v16, v8
 ; CHECK-RV32-NEXT:    vmv.s.x v9, a1
 ; CHECK-RV32-NEXT:    vslideup.vi v8, v9, 25
 ; CHECK-RV32-NEXT:    addi a0, a0, 1
@@ -4208,8 +4200,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_27
 ; CHECK-RV32-NEXT:  .LBB61_570: # %cond.load101
 ; CHECK-RV32-NEXT:    lbu a1, 0(a0)
-; CHECK-RV32-NEXT:    vmv8r.v v16, v8
 ; CHECK-RV32-NEXT:    vsetivli zero, 27, e8, m1, tu, ma
+; CHECK-RV32-NEXT:    vmv8r.v v16, v8
 ; CHECK-RV32-NEXT:    vmv.s.x v9, a1
 ; CHECK-RV32-NEXT:    vslideup.vi v8, v9, 26
 ; CHECK-RV32-NEXT:    addi a0, a0, 1
@@ -4220,8 +4212,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_28
 ; CHECK-RV32-NEXT:  .LBB61_571: # %cond.load105
 ; CHECK-RV32-NEXT:    lbu a1, 0(a0)
-; CHECK-RV32-NEXT:    vmv8r.v v16, v8
 ; CHECK-RV32-NEXT:    vsetivli zero, 28, e8, m1, tu, ma
+; CHECK-RV32-NEXT:    vmv8r.v v16, v8
 ; CHECK-RV32-NEXT:    vmv.s.x v9, a1
 ; CHECK-RV32-NEXT:    vslideup.vi v8, v9, 27
 ; CHECK-RV32-NEXT:    addi a0, a0, 1
@@ -4247,9 +4239,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_36
 ; CHECK-RV32-NEXT:  .LBB61_573: # %cond.load125
 ; CHECK-RV32-NEXT:    lbu a3, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v16, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v9, a3
 ; CHECK-RV32-NEXT:    li a3, 33
 ; CHECK-RV32-NEXT:    li a4, 32
@@ -4263,9 +4254,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_37
 ; CHECK-RV32-NEXT:  .LBB61_574: # %cond.load129
 ; CHECK-RV32-NEXT:    lbu a3, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v16, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v9, a3
 ; CHECK-RV32-NEXT:    li a3, 34
 ; CHECK-RV32-NEXT:    li a4, 33
@@ -4279,9 +4269,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_38
 ; CHECK-RV32-NEXT:  .LBB61_575: # %cond.load133
 ; CHECK-RV32-NEXT:    lbu a3, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v16, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v9, a3
 ; CHECK-RV32-NEXT:    li a3, 35
 ; CHECK-RV32-NEXT:    li a4, 34
@@ -4295,9 +4284,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_39
 ; CHECK-RV32-NEXT:  .LBB61_576: # %cond.load137
 ; CHECK-RV32-NEXT:    lbu a3, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v16, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v9, a3
 ; CHECK-RV32-NEXT:    li a3, 36
 ; CHECK-RV32-NEXT:    li a4, 35
@@ -4311,9 +4299,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_40
 ; CHECK-RV32-NEXT:  .LBB61_577: # %cond.load141
 ; CHECK-RV32-NEXT:    lbu a3, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v16, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v9, a3
 ; CHECK-RV32-NEXT:    li a3, 37
 ; CHECK-RV32-NEXT:    li a4, 36
@@ -4327,9 +4314,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_41
 ; CHECK-RV32-NEXT:  .LBB61_578: # %cond.load145
 ; CHECK-RV32-NEXT:    lbu a3, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v16, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v9, a3
 ; CHECK-RV32-NEXT:    li a3, 38
 ; CHECK-RV32-NEXT:    li a4, 37
@@ -4343,9 +4329,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_42
 ; CHECK-RV32-NEXT:  .LBB61_579: # %cond.load149
 ; CHECK-RV32-NEXT:    lbu a3, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v16, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v9, a3
 ; CHECK-RV32-NEXT:    li a3, 39
 ; CHECK-RV32-NEXT:    li a4, 38
@@ -4359,9 +4344,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_43
 ; CHECK-RV32-NEXT:  .LBB61_580: # %cond.load153
 ; CHECK-RV32-NEXT:    lbu a3, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v16, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v9, a3
 ; CHECK-RV32-NEXT:    li a3, 40
 ; CHECK-RV32-NEXT:    li a4, 39
@@ -4375,9 +4359,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_44
 ; CHECK-RV32-NEXT:  .LBB61_581: # %cond.load157
 ; CHECK-RV32-NEXT:    lbu a3, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v16, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v9, a3
 ; CHECK-RV32-NEXT:    li a3, 41
 ; CHECK-RV32-NEXT:    li a4, 40
@@ -4391,9 +4374,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_45
 ; CHECK-RV32-NEXT:  .LBB61_582: # %cond.load161
 ; CHECK-RV32-NEXT:    lbu a3, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v16, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v9, a3
 ; CHECK-RV32-NEXT:    li a3, 42
 ; CHECK-RV32-NEXT:    li a4, 41
@@ -4407,9 +4389,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_46
 ; CHECK-RV32-NEXT:  .LBB61_583: # %cond.load165
 ; CHECK-RV32-NEXT:    lbu a3, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v16, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v9, a3
 ; CHECK-RV32-NEXT:    li a3, 43
 ; CHECK-RV32-NEXT:    li a4, 42
@@ -4423,9 +4404,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_47
 ; CHECK-RV32-NEXT:  .LBB61_584: # %cond.load169
 ; CHECK-RV32-NEXT:    lbu a3, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v16, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v9, a3
 ; CHECK-RV32-NEXT:    li a3, 44
 ; CHECK-RV32-NEXT:    li a4, 43
@@ -4439,9 +4419,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_48
 ; CHECK-RV32-NEXT:  .LBB61_585: # %cond.load173
 ; CHECK-RV32-NEXT:    lbu a3, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v16, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v9, a3
 ; CHECK-RV32-NEXT:    li a3, 45
 ; CHECK-RV32-NEXT:    li a4, 44
@@ -4455,9 +4434,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_49
 ; CHECK-RV32-NEXT:  .LBB61_586: # %cond.load177
 ; CHECK-RV32-NEXT:    lbu a3, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v16, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v9, a3
 ; CHECK-RV32-NEXT:    li a3, 46
 ; CHECK-RV32-NEXT:    li a4, 45
@@ -4471,9 +4449,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_50
 ; CHECK-RV32-NEXT:  .LBB61_587: # %cond.load181
 ; CHECK-RV32-NEXT:    lbu a3, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v16, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v9, a3
 ; CHECK-RV32-NEXT:    li a3, 47
 ; CHECK-RV32-NEXT:    li a4, 46
@@ -4487,9 +4464,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_51
 ; CHECK-RV32-NEXT:  .LBB61_588: # %cond.load185
 ; CHECK-RV32-NEXT:    lbu a3, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v16, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v9, a3
 ; CHECK-RV32-NEXT:    li a3, 48
 ; CHECK-RV32-NEXT:    li a4, 47
@@ -4503,9 +4479,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_52
 ; CHECK-RV32-NEXT:  .LBB61_589: # %cond.load189
 ; CHECK-RV32-NEXT:    lbu a3, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v16, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v9, a3
 ; CHECK-RV32-NEXT:    li a3, 49
 ; CHECK-RV32-NEXT:    li a4, 48
@@ -4519,9 +4494,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_53
 ; CHECK-RV32-NEXT:  .LBB61_590: # %cond.load193
 ; CHECK-RV32-NEXT:    lbu a3, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v16, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v9, a3
 ; CHECK-RV32-NEXT:    li a3, 50
 ; CHECK-RV32-NEXT:    li a4, 49
@@ -4535,9 +4509,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_54
 ; CHECK-RV32-NEXT:  .LBB61_591: # %cond.load197
 ; CHECK-RV32-NEXT:    lbu a3, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v16, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v9, a3
 ; CHECK-RV32-NEXT:    li a3, 51
 ; CHECK-RV32-NEXT:    li a4, 50
@@ -4551,9 +4524,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_55
 ; CHECK-RV32-NEXT:  .LBB61_592: # %cond.load201
 ; CHECK-RV32-NEXT:    lbu a3, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v16, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v9, a3
 ; CHECK-RV32-NEXT:    li a3, 52
 ; CHECK-RV32-NEXT:    li a4, 51
@@ -4567,9 +4539,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_56
 ; CHECK-RV32-NEXT:  .LBB61_593: # %cond.load205
 ; CHECK-RV32-NEXT:    lbu a3, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v16, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v9, a3
 ; CHECK-RV32-NEXT:    li a3, 53
 ; CHECK-RV32-NEXT:    li a4, 52
@@ -4583,9 +4554,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_57
 ; CHECK-RV32-NEXT:  .LBB61_594: # %cond.load209
 ; CHECK-RV32-NEXT:    lbu a3, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v16, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v9, a3
 ; CHECK-RV32-NEXT:    li a3, 54
 ; CHECK-RV32-NEXT:    li a4, 53
@@ -4599,9 +4569,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_58
 ; CHECK-RV32-NEXT:  .LBB61_595: # %cond.load213
 ; CHECK-RV32-NEXT:    lbu a3, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v16, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v9, a3
 ; CHECK-RV32-NEXT:    li a3, 55
 ; CHECK-RV32-NEXT:    li a4, 54
@@ -4615,9 +4584,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_59
 ; CHECK-RV32-NEXT:  .LBB61_596: # %cond.load217
 ; CHECK-RV32-NEXT:    lbu a3, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v16, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v9, a3
 ; CHECK-RV32-NEXT:    li a3, 56
 ; CHECK-RV32-NEXT:    li a4, 55
@@ -4631,9 +4599,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_60
 ; CHECK-RV32-NEXT:  .LBB61_597: # %cond.load221
 ; CHECK-RV32-NEXT:    lbu a3, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v16, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v9, a3
 ; CHECK-RV32-NEXT:    li a3, 57
 ; CHECK-RV32-NEXT:    li a4, 56
@@ -4647,9 +4614,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_61
 ; CHECK-RV32-NEXT:  .LBB61_598: # %cond.load225
 ; CHECK-RV32-NEXT:    lbu a3, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v16, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v9, a3
 ; CHECK-RV32-NEXT:    li a3, 58
 ; CHECK-RV32-NEXT:    li a4, 57
@@ -4663,9 +4629,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_62
 ; CHECK-RV32-NEXT:  .LBB61_599: # %cond.load229
 ; CHECK-RV32-NEXT:    lbu a3, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v16, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v9, a3
 ; CHECK-RV32-NEXT:    li a3, 59
 ; CHECK-RV32-NEXT:    li a4, 58
@@ -4679,9 +4644,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_63
 ; CHECK-RV32-NEXT:  .LBB61_600: # %cond.load233
 ; CHECK-RV32-NEXT:    lbu a3, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v16, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v9, a3
 ; CHECK-RV32-NEXT:    li a3, 60
 ; CHECK-RV32-NEXT:    li a4, 59
@@ -4695,9 +4659,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_64
 ; CHECK-RV32-NEXT:  .LBB61_601: # %cond.load237
 ; CHECK-RV32-NEXT:    lbu a3, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v16, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v9, a3
 ; CHECK-RV32-NEXT:    li a3, 61
 ; CHECK-RV32-NEXT:    li a4, 60
@@ -4727,9 +4690,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_70
 ; CHECK-RV32-NEXT:  .LBB61_603: # %cond.load253
 ; CHECK-RV32-NEXT:    lbu a2, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v24, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v10, a2
 ; CHECK-RV32-NEXT:    li a2, 65
 ; CHECK-RV32-NEXT:    li a4, 64
@@ -4743,9 +4705,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_71
 ; CHECK-RV32-NEXT:  .LBB61_604: # %cond.load257
 ; CHECK-RV32-NEXT:    lbu a2, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v24, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v10, a2
 ; CHECK-RV32-NEXT:    li a2, 66
 ; CHECK-RV32-NEXT:    li a4, 65
@@ -4759,9 +4720,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_72
 ; CHECK-RV32-NEXT:  .LBB61_605: # %cond.load261
 ; CHECK-RV32-NEXT:    lbu a2, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v24, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v10, a2
 ; CHECK-RV32-NEXT:    li a2, 67
 ; CHECK-RV32-NEXT:    li a4, 66
@@ -4775,9 +4735,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_73
 ; CHECK-RV32-NEXT:  .LBB61_606: # %cond.load265
 ; CHECK-RV32-NEXT:    lbu a2, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v24, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v10, a2
 ; CHECK-RV32-NEXT:    li a2, 68
 ; CHECK-RV32-NEXT:    li a4, 67
@@ -4791,9 +4750,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_74
 ; CHECK-RV32-NEXT:  .LBB61_607: # %cond.load269
 ; CHECK-RV32-NEXT:    lbu a2, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v24, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v10, a2
 ; CHECK-RV32-NEXT:    li a2, 69
 ; CHECK-RV32-NEXT:    li a4, 68
@@ -4807,9 +4765,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_75
 ; CHECK-RV32-NEXT:  .LBB61_608: # %cond.load273
 ; CHECK-RV32-NEXT:    lbu a2, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v24, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v10, a2
 ; CHECK-RV32-NEXT:    li a2, 70
 ; CHECK-RV32-NEXT:    li a4, 69
@@ -4823,9 +4780,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_76
 ; CHECK-RV32-NEXT:  .LBB61_609: # %cond.load277
 ; CHECK-RV32-NEXT:    lbu a2, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v24, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v10, a2
 ; CHECK-RV32-NEXT:    li a2, 71
 ; CHECK-RV32-NEXT:    li a4, 70
@@ -4839,9 +4795,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_77
 ; CHECK-RV32-NEXT:  .LBB61_610: # %cond.load281
 ; CHECK-RV32-NEXT:    lbu a2, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v24, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v10, a2
 ; CHECK-RV32-NEXT:    li a2, 72
 ; CHECK-RV32-NEXT:    li a4, 71
@@ -4855,9 +4810,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_78
 ; CHECK-RV32-NEXT:  .LBB61_611: # %cond.load285
 ; CHECK-RV32-NEXT:    lbu a2, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v24, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v10, a2
 ; CHECK-RV32-NEXT:    li a2, 73
 ; CHECK-RV32-NEXT:    li a4, 72
@@ -4871,9 +4825,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_79
 ; CHECK-RV32-NEXT:  .LBB61_612: # %cond.load289
 ; CHECK-RV32-NEXT:    lbu a2, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v24, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v10, a2
 ; CHECK-RV32-NEXT:    li a2, 74
 ; CHECK-RV32-NEXT:    li a4, 73
@@ -4887,9 +4840,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_80
 ; CHECK-RV32-NEXT:  .LBB61_613: # %cond.load293
 ; CHECK-RV32-NEXT:    lbu a2, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v24, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v10, a2
 ; CHECK-RV32-NEXT:    li a2, 75
 ; CHECK-RV32-NEXT:    li a4, 74
@@ -4903,9 +4855,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_81
 ; CHECK-RV32-NEXT:  .LBB61_614: # %cond.load297
 ; CHECK-RV32-NEXT:    lbu a2, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v24, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v10, a2
 ; CHECK-RV32-NEXT:    li a2, 76
 ; CHECK-RV32-NEXT:    li a4, 75
@@ -4919,9 +4870,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_82
 ; CHECK-RV32-NEXT:  .LBB61_615: # %cond.load301
 ; CHECK-RV32-NEXT:    lbu a2, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v24, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v10, a2
 ; CHECK-RV32-NEXT:    li a2, 77
 ; CHECK-RV32-NEXT:    li a4, 76
@@ -4935,9 +4885,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_83
 ; CHECK-RV32-NEXT:  .LBB61_616: # %cond.load305
 ; CHECK-RV32-NEXT:    lbu a2, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v24, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v10, a2
 ; CHECK-RV32-NEXT:    li a2, 78
 ; CHECK-RV32-NEXT:    li a4, 77
@@ -4951,9 +4900,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_84
 ; CHECK-RV32-NEXT:  .LBB61_617: # %cond.load309
 ; CHECK-RV32-NEXT:    lbu a2, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v24, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v10, a2
 ; CHECK-RV32-NEXT:    li a2, 79
 ; CHECK-RV32-NEXT:    li a4, 78
@@ -4967,9 +4915,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_85
 ; CHECK-RV32-NEXT:  .LBB61_618: # %cond.load313
 ; CHECK-RV32-NEXT:    lbu a2, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v24, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v10, a2
 ; CHECK-RV32-NEXT:    li a2, 80
 ; CHECK-RV32-NEXT:    li a4, 79
@@ -4983,9 +4930,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_86
 ; CHECK-RV32-NEXT:  .LBB61_619: # %cond.load317
 ; CHECK-RV32-NEXT:    lbu a2, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v24, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v10, a2
 ; CHECK-RV32-NEXT:    li a2, 81
 ; CHECK-RV32-NEXT:    li a4, 80
@@ -4999,9 +4945,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_87
 ; CHECK-RV32-NEXT:  .LBB61_620: # %cond.load321
 ; CHECK-RV32-NEXT:    lbu a2, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v24, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v10, a2
 ; CHECK-RV32-NEXT:    li a2, 82
 ; CHECK-RV32-NEXT:    li a4, 81
@@ -5015,9 +4960,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_88
 ; CHECK-RV32-NEXT:  .LBB61_621: # %cond.load325
 ; CHECK-RV32-NEXT:    lbu a2, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v24, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v10, a2
 ; CHECK-RV32-NEXT:    li a2, 83
 ; CHECK-RV32-NEXT:    li a4, 82
@@ -5031,9 +4975,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_89
 ; CHECK-RV32-NEXT:  .LBB61_622: # %cond.load329
 ; CHECK-RV32-NEXT:    lbu a2, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v24, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v10, a2
 ; CHECK-RV32-NEXT:    li a2, 84
 ; CHECK-RV32-NEXT:    li a4, 83
@@ -5047,9 +4990,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_90
 ; CHECK-RV32-NEXT:  .LBB61_623: # %cond.load333
 ; CHECK-RV32-NEXT:    lbu a2, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v24, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v10, a2
 ; CHECK-RV32-NEXT:    li a2, 85
 ; CHECK-RV32-NEXT:    li a4, 84
@@ -5063,9 +5005,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_91
 ; CHECK-RV32-NEXT:  .LBB61_624: # %cond.load337
 ; CHECK-RV32-NEXT:    lbu a2, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v24, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v10, a2
 ; CHECK-RV32-NEXT:    li a2, 86
 ; CHECK-RV32-NEXT:    li a4, 85
@@ -5079,9 +5020,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_92
 ; CHECK-RV32-NEXT:  .LBB61_625: # %cond.load341
 ; CHECK-RV32-NEXT:    lbu a2, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v24, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v10, a2
 ; CHECK-RV32-NEXT:    li a2, 87
 ; CHECK-RV32-NEXT:    li a4, 86
@@ -5095,9 +5035,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_93
 ; CHECK-RV32-NEXT:  .LBB61_626: # %cond.load345
 ; CHECK-RV32-NEXT:    lbu a2, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v24, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v10, a2
 ; CHECK-RV32-NEXT:    li a2, 88
 ; CHECK-RV32-NEXT:    li a4, 87
@@ -5111,9 +5050,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_94
 ; CHECK-RV32-NEXT:  .LBB61_627: # %cond.load349
 ; CHECK-RV32-NEXT:    lbu a2, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v24, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v10, a2
 ; CHECK-RV32-NEXT:    li a2, 89
 ; CHECK-RV32-NEXT:    li a4, 88
@@ -5127,9 +5065,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_95
 ; CHECK-RV32-NEXT:  .LBB61_628: # %cond.load353
 ; CHECK-RV32-NEXT:    lbu a2, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v24, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v10, a2
 ; CHECK-RV32-NEXT:    li a2, 90
 ; CHECK-RV32-NEXT:    li a4, 89
@@ -5143,9 +5080,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_96
 ; CHECK-RV32-NEXT:  .LBB61_629: # %cond.load357
 ; CHECK-RV32-NEXT:    lbu a2, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v24, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v10, a2
 ; CHECK-RV32-NEXT:    li a2, 91
 ; CHECK-RV32-NEXT:    li a4, 90
@@ -5159,9 +5095,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_97
 ; CHECK-RV32-NEXT:  .LBB61_630: # %cond.load361
 ; CHECK-RV32-NEXT:    lbu a2, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v24, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v10, a2
 ; CHECK-RV32-NEXT:    li a2, 92
 ; CHECK-RV32-NEXT:    li a4, 91
@@ -5175,9 +5110,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_98
 ; CHECK-RV32-NEXT:  .LBB61_631: # %cond.load365
 ; CHECK-RV32-NEXT:    lbu a2, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v24, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v10, a2
 ; CHECK-RV32-NEXT:    li a2, 93
 ; CHECK-RV32-NEXT:    li a4, 92
@@ -5207,9 +5141,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_104
 ; CHECK-RV32-NEXT:  .LBB61_633: # %cond.load381
 ; CHECK-RV32-NEXT:    lbu a3, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v16, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v10, a3
 ; CHECK-RV32-NEXT:    li a3, 97
 ; CHECK-RV32-NEXT:    li a4, 96
@@ -5223,9 +5156,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_105
 ; CHECK-RV32-NEXT:  .LBB61_634: # %cond.load385
 ; CHECK-RV32-NEXT:    lbu a3, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v16, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v10, a3
 ; CHECK-RV32-NEXT:    li a3, 98
 ; CHECK-RV32-NEXT:    li a4, 97
@@ -5239,9 +5171,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_106
 ; CHECK-RV32-NEXT:  .LBB61_635: # %cond.load389
 ; CHECK-RV32-NEXT:    lbu a3, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v16, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v10, a3
 ; CHECK-RV32-NEXT:    li a3, 99
 ; CHECK-RV32-NEXT:    li a4, 98
@@ -5255,9 +5186,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_107
 ; CHECK-RV32-NEXT:  .LBB61_636: # %cond.load393
 ; CHECK-RV32-NEXT:    lbu a3, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v16, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v10, a3
 ; CHECK-RV32-NEXT:    li a3, 100
 ; CHECK-RV32-NEXT:    li a4, 99
@@ -5271,9 +5201,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_108
 ; CHECK-RV32-NEXT:  .LBB61_637: # %cond.load397
 ; CHECK-RV32-NEXT:    lbu a3, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v16, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v10, a3
 ; CHECK-RV32-NEXT:    li a3, 101
 ; CHECK-RV32-NEXT:    li a4, 100
@@ -5287,9 +5216,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_109
 ; CHECK-RV32-NEXT:  .LBB61_638: # %cond.load401
 ; CHECK-RV32-NEXT:    lbu a3, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v16, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v10, a3
 ; CHECK-RV32-NEXT:    li a3, 102
 ; CHECK-RV32-NEXT:    li a4, 101
@@ -5303,9 +5231,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_110
 ; CHECK-RV32-NEXT:  .LBB61_639: # %cond.load405
 ; CHECK-RV32-NEXT:    lbu a3, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v16, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v10, a3
 ; CHECK-RV32-NEXT:    li a3, 103
 ; CHECK-RV32-NEXT:    li a4, 102
@@ -5319,9 +5246,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_111
 ; CHECK-RV32-NEXT:  .LBB61_640: # %cond.load409
 ; CHECK-RV32-NEXT:    lbu a3, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v16, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v10, a3
 ; CHECK-RV32-NEXT:    li a3, 104
 ; CHECK-RV32-NEXT:    li a4, 103
@@ -5335,9 +5261,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_112
 ; CHECK-RV32-NEXT:  .LBB61_641: # %cond.load413
 ; CHECK-RV32-NEXT:    lbu a3, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v16, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v10, a3
 ; CHECK-RV32-NEXT:    li a3, 105
 ; CHECK-RV32-NEXT:    li a4, 104
@@ -5351,9 +5276,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_113
 ; CHECK-RV32-NEXT:  .LBB61_642: # %cond.load417
 ; CHECK-RV32-NEXT:    lbu a3, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v16, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v10, a3
 ; CHECK-RV32-NEXT:    li a3, 106
 ; CHECK-RV32-NEXT:    li a4, 105
@@ -5367,9 +5291,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_114
 ; CHECK-RV32-NEXT:  .LBB61_643: # %cond.load421
 ; CHECK-RV32-NEXT:    lbu a3, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v16, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v10, a3
 ; CHECK-RV32-NEXT:    li a3, 107
 ; CHECK-RV32-NEXT:    li a4, 106
@@ -5383,9 +5306,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_115
 ; CHECK-RV32-NEXT:  .LBB61_644: # %cond.load425
 ; CHECK-RV32-NEXT:    lbu a3, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v16, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v10, a3
 ; CHECK-RV32-NEXT:    li a3, 108
 ; CHECK-RV32-NEXT:    li a4, 107
@@ -5399,9 +5321,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_116
 ; CHECK-RV32-NEXT:  .LBB61_645: # %cond.load429
 ; CHECK-RV32-NEXT:    lbu a3, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v16, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v10, a3
 ; CHECK-RV32-NEXT:    li a3, 109
 ; CHECK-RV32-NEXT:    li a4, 108
@@ -5415,9 +5336,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_117
 ; CHECK-RV32-NEXT:  .LBB61_646: # %cond.load433
 ; CHECK-RV32-NEXT:    lbu a3, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v16, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v10, a3
 ; CHECK-RV32-NEXT:    li a3, 110
 ; CHECK-RV32-NEXT:    li a4, 109
@@ -5431,9 +5351,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_118
 ; CHECK-RV32-NEXT:  .LBB61_647: # %cond.load437
 ; CHECK-RV32-NEXT:    lbu a3, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v16, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v10, a3
 ; CHECK-RV32-NEXT:    li a3, 111
 ; CHECK-RV32-NEXT:    li a4, 110
@@ -5447,9 +5366,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_119
 ; CHECK-RV32-NEXT:  .LBB61_648: # %cond.load441
 ; CHECK-RV32-NEXT:    lbu a3, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v16, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v10, a3
 ; CHECK-RV32-NEXT:    li a3, 112
 ; CHECK-RV32-NEXT:    li a4, 111
@@ -5463,9 +5381,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_120
 ; CHECK-RV32-NEXT:  .LBB61_649: # %cond.load445
 ; CHECK-RV32-NEXT:    lbu a3, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v16, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v10, a3
 ; CHECK-RV32-NEXT:    li a3, 113
 ; CHECK-RV32-NEXT:    li a4, 112
@@ -5479,9 +5396,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_121
 ; CHECK-RV32-NEXT:  .LBB61_650: # %cond.load449
 ; CHECK-RV32-NEXT:    lbu a3, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v16, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v10, a3
 ; CHECK-RV32-NEXT:    li a3, 114
 ; CHECK-RV32-NEXT:    li a4, 113
@@ -5495,9 +5411,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_122
 ; CHECK-RV32-NEXT:  .LBB61_651: # %cond.load453
 ; CHECK-RV32-NEXT:    lbu a3, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v16, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v10, a3
 ; CHECK-RV32-NEXT:    li a3, 115
 ; CHECK-RV32-NEXT:    li a4, 114
@@ -5511,9 +5426,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_123
 ; CHECK-RV32-NEXT:  .LBB61_652: # %cond.load457
 ; CHECK-RV32-NEXT:    lbu a3, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v16, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v10, a3
 ; CHECK-RV32-NEXT:    li a3, 116
 ; CHECK-RV32-NEXT:    li a4, 115
@@ -5527,9 +5441,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_124
 ; CHECK-RV32-NEXT:  .LBB61_653: # %cond.load461
 ; CHECK-RV32-NEXT:    lbu a3, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v16, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v10, a3
 ; CHECK-RV32-NEXT:    li a3, 117
 ; CHECK-RV32-NEXT:    li a4, 116
@@ -5543,9 +5456,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_125
 ; CHECK-RV32-NEXT:  .LBB61_654: # %cond.load465
 ; CHECK-RV32-NEXT:    lbu a3, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v16, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v10, a3
 ; CHECK-RV32-NEXT:    li a3, 118
 ; CHECK-RV32-NEXT:    li a4, 117
@@ -5559,9 +5471,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_126
 ; CHECK-RV32-NEXT:  .LBB61_655: # %cond.load469
 ; CHECK-RV32-NEXT:    lbu a3, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v16, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v10, a3
 ; CHECK-RV32-NEXT:    li a3, 119
 ; CHECK-RV32-NEXT:    li a4, 118
@@ -5575,9 +5486,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_127
 ; CHECK-RV32-NEXT:  .LBB61_656: # %cond.load473
 ; CHECK-RV32-NEXT:    lbu a3, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v16, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v10, a3
 ; CHECK-RV32-NEXT:    li a3, 120
 ; CHECK-RV32-NEXT:    li a4, 119
@@ -5591,9 +5501,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_128
 ; CHECK-RV32-NEXT:  .LBB61_657: # %cond.load477
 ; CHECK-RV32-NEXT:    lbu a3, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v16, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v10, a3
 ; CHECK-RV32-NEXT:    li a3, 121
 ; CHECK-RV32-NEXT:    li a4, 120
@@ -5607,9 +5516,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_129
 ; CHECK-RV32-NEXT:  .LBB61_658: # %cond.load481
 ; CHECK-RV32-NEXT:    lbu a3, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v16, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v10, a3
 ; CHECK-RV32-NEXT:    li a3, 122
 ; CHECK-RV32-NEXT:    li a4, 121
@@ -5623,9 +5531,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_130
 ; CHECK-RV32-NEXT:  .LBB61_659: # %cond.load485
 ; CHECK-RV32-NEXT:    lbu a3, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v16, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v10, a3
 ; CHECK-RV32-NEXT:    li a3, 123
 ; CHECK-RV32-NEXT:    li a4, 122
@@ -5639,9 +5546,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_131
 ; CHECK-RV32-NEXT:  .LBB61_660: # %cond.load489
 ; CHECK-RV32-NEXT:    lbu a3, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v16, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v10, a3
 ; CHECK-RV32-NEXT:    li a3, 124
 ; CHECK-RV32-NEXT:    li a4, 123
@@ -5655,9 +5561,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_132
 ; CHECK-RV32-NEXT:  .LBB61_661: # %cond.load493
 ; CHECK-RV32-NEXT:    lbu a3, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v16, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v10, a3
 ; CHECK-RV32-NEXT:    li a3, 125
 ; CHECK-RV32-NEXT:    li a4, 124
@@ -5687,9 +5592,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_138
 ; CHECK-RV32-NEXT:  .LBB61_663: # %cond.load509
 ; CHECK-RV32-NEXT:    lbu a2, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v24, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v12, a2
 ; CHECK-RV32-NEXT:    li a2, 129
 ; CHECK-RV32-NEXT:    li a4, 128
@@ -5703,9 +5607,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_139
 ; CHECK-RV32-NEXT:  .LBB61_664: # %cond.load513
 ; CHECK-RV32-NEXT:    lbu a2, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v24, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v12, a2
 ; CHECK-RV32-NEXT:    li a2, 130
 ; CHECK-RV32-NEXT:    li a4, 129
@@ -5719,9 +5622,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_140
 ; CHECK-RV32-NEXT:  .LBB61_665: # %cond.load517
 ; CHECK-RV32-NEXT:    lbu a2, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v24, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v12, a2
 ; CHECK-RV32-NEXT:    li a2, 131
 ; CHECK-RV32-NEXT:    li a4, 130
@@ -5735,9 +5637,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_141
 ; CHECK-RV32-NEXT:  .LBB61_666: # %cond.load521
 ; CHECK-RV32-NEXT:    lbu a2, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v24, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v12, a2
 ; CHECK-RV32-NEXT:    li a2, 132
 ; CHECK-RV32-NEXT:    li a4, 131
@@ -5751,9 +5652,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_142
 ; CHECK-RV32-NEXT:  .LBB61_667: # %cond.load525
 ; CHECK-RV32-NEXT:    lbu a2, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v24, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v12, a2
 ; CHECK-RV32-NEXT:    li a2, 133
 ; CHECK-RV32-NEXT:    li a4, 132
@@ -5767,9 +5667,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_143
 ; CHECK-RV32-NEXT:  .LBB61_668: # %cond.load529
 ; CHECK-RV32-NEXT:    lbu a2, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v24, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v12, a2
 ; CHECK-RV32-NEXT:    li a2, 134
 ; CHECK-RV32-NEXT:    li a4, 133
@@ -5783,9 +5682,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_144
 ; CHECK-RV32-NEXT:  .LBB61_669: # %cond.load533
 ; CHECK-RV32-NEXT:    lbu a2, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v24, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v12, a2
 ; CHECK-RV32-NEXT:    li a2, 135
 ; CHECK-RV32-NEXT:    li a4, 134
@@ -5799,9 +5697,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_145
 ; CHECK-RV32-NEXT:  .LBB61_670: # %cond.load537
 ; CHECK-RV32-NEXT:    lbu a2, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v24, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v12, a2
 ; CHECK-RV32-NEXT:    li a2, 136
 ; CHECK-RV32-NEXT:    li a4, 135
@@ -5815,9 +5712,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_146
 ; CHECK-RV32-NEXT:  .LBB61_671: # %cond.load541
 ; CHECK-RV32-NEXT:    lbu a2, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v24, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v12, a2
 ; CHECK-RV32-NEXT:    li a2, 137
 ; CHECK-RV32-NEXT:    li a4, 136
@@ -5831,9 +5727,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_147
 ; CHECK-RV32-NEXT:  .LBB61_672: # %cond.load545
 ; CHECK-RV32-NEXT:    lbu a2, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v24, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v12, a2
 ; CHECK-RV32-NEXT:    li a2, 138
 ; CHECK-RV32-NEXT:    li a4, 137
@@ -5847,9 +5742,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_148
 ; CHECK-RV32-NEXT:  .LBB61_673: # %cond.load549
 ; CHECK-RV32-NEXT:    lbu a2, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v24, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v12, a2
 ; CHECK-RV32-NEXT:    li a2, 139
 ; CHECK-RV32-NEXT:    li a4, 138
@@ -5863,9 +5757,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_149
 ; CHECK-RV32-NEXT:  .LBB61_674: # %cond.load553
 ; CHECK-RV32-NEXT:    lbu a2, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v24, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v12, a2
 ; CHECK-RV32-NEXT:    li a2, 140
 ; CHECK-RV32-NEXT:    li a4, 139
@@ -5879,9 +5772,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_150
 ; CHECK-RV32-NEXT:  .LBB61_675: # %cond.load557
 ; CHECK-RV32-NEXT:    lbu a2, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v24, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v12, a2
 ; CHECK-RV32-NEXT:    li a2, 141
 ; CHECK-RV32-NEXT:    li a4, 140
@@ -5895,9 +5787,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_151
 ; CHECK-RV32-NEXT:  .LBB61_676: # %cond.load561
 ; CHECK-RV32-NEXT:    lbu a2, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v24, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v12, a2
 ; CHECK-RV32-NEXT:    li a2, 142
 ; CHECK-RV32-NEXT:    li a4, 141
@@ -5911,9 +5802,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_152
 ; CHECK-RV32-NEXT:  .LBB61_677: # %cond.load565
 ; CHECK-RV32-NEXT:    lbu a2, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v24, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v12, a2
 ; CHECK-RV32-NEXT:    li a2, 143
 ; CHECK-RV32-NEXT:    li a4, 142
@@ -5927,9 +5817,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_153
 ; CHECK-RV32-NEXT:  .LBB61_678: # %cond.load569
 ; CHECK-RV32-NEXT:    lbu a2, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v24, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v12, a2
 ; CHECK-RV32-NEXT:    li a2, 144
 ; CHECK-RV32-NEXT:    li a4, 143
@@ -5943,9 +5832,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_154
 ; CHECK-RV32-NEXT:  .LBB61_679: # %cond.load573
 ; CHECK-RV32-NEXT:    lbu a2, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v24, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v12, a2
 ; CHECK-RV32-NEXT:    li a2, 145
 ; CHECK-RV32-NEXT:    li a4, 144
@@ -5959,9 +5847,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_155
 ; CHECK-RV32-NEXT:  .LBB61_680: # %cond.load577
 ; CHECK-RV32-NEXT:    lbu a2, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v24, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v12, a2
 ; CHECK-RV32-NEXT:    li a2, 146
 ; CHECK-RV32-NEXT:    li a4, 145
@@ -5975,9 +5862,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_156
 ; CHECK-RV32-NEXT:  .LBB61_681: # %cond.load581
 ; CHECK-RV32-NEXT:    lbu a2, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v24, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v12, a2
 ; CHECK-RV32-NEXT:    li a2, 147
 ; CHECK-RV32-NEXT:    li a4, 146
@@ -5991,9 +5877,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_157
 ; CHECK-RV32-NEXT:  .LBB61_682: # %cond.load585
 ; CHECK-RV32-NEXT:    lbu a2, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v24, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v12, a2
 ; CHECK-RV32-NEXT:    li a2, 148
 ; CHECK-RV32-NEXT:    li a4, 147
@@ -6007,9 +5892,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_158
 ; CHECK-RV32-NEXT:  .LBB61_683: # %cond.load589
 ; CHECK-RV32-NEXT:    lbu a2, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v24, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v12, a2
 ; CHECK-RV32-NEXT:    li a2, 149
 ; CHECK-RV32-NEXT:    li a4, 148
@@ -6023,9 +5907,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_159
 ; CHECK-RV32-NEXT:  .LBB61_684: # %cond.load593
 ; CHECK-RV32-NEXT:    lbu a2, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v24, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v12, a2
 ; CHECK-RV32-NEXT:    li a2, 150
 ; CHECK-RV32-NEXT:    li a4, 149
@@ -6039,9 +5922,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_160
 ; CHECK-RV32-NEXT:  .LBB61_685: # %cond.load597
 ; CHECK-RV32-NEXT:    lbu a2, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v24, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v12, a2
 ; CHECK-RV32-NEXT:    li a2, 151
 ; CHECK-RV32-NEXT:    li a4, 150
@@ -6055,9 +5937,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_161
 ; CHECK-RV32-NEXT:  .LBB61_686: # %cond.load601
 ; CHECK-RV32-NEXT:    lbu a2, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v24, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v12, a2
 ; CHECK-RV32-NEXT:    li a2, 152
 ; CHECK-RV32-NEXT:    li a4, 151
@@ -6071,9 +5952,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_162
 ; CHECK-RV32-NEXT:  .LBB61_687: # %cond.load605
 ; CHECK-RV32-NEXT:    lbu a2, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v24, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v12, a2
 ; CHECK-RV32-NEXT:    li a2, 153
 ; CHECK-RV32-NEXT:    li a4, 152
@@ -6087,9 +5967,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_163
 ; CHECK-RV32-NEXT:  .LBB61_688: # %cond.load609
 ; CHECK-RV32-NEXT:    lbu a2, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v24, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v12, a2
 ; CHECK-RV32-NEXT:    li a2, 154
 ; CHECK-RV32-NEXT:    li a4, 153
@@ -6103,9 +5982,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_164
 ; CHECK-RV32-NEXT:  .LBB61_689: # %cond.load613
 ; CHECK-RV32-NEXT:    lbu a2, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v24, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v12, a2
 ; CHECK-RV32-NEXT:    li a2, 155
 ; CHECK-RV32-NEXT:    li a4, 154
@@ -6119,9 +5997,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_165
 ; CHECK-RV32-NEXT:  .LBB61_690: # %cond.load617
 ; CHECK-RV32-NEXT:    lbu a2, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v24, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v12, a2
 ; CHECK-RV32-NEXT:    li a2, 156
 ; CHECK-RV32-NEXT:    li a4, 155
@@ -6135,9 +6012,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_166
 ; CHECK-RV32-NEXT:  .LBB61_691: # %cond.load621
 ; CHECK-RV32-NEXT:    lbu a2, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v24, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v12, a2
 ; CHECK-RV32-NEXT:    li a2, 157
 ; CHECK-RV32-NEXT:    li a4, 156
@@ -6167,9 +6043,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_172
 ; CHECK-RV32-NEXT:  .LBB61_693: # %cond.load637
 ; CHECK-RV32-NEXT:    lbu a3, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v16, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v12, a3
 ; CHECK-RV32-NEXT:    li a3, 161
 ; CHECK-RV32-NEXT:    li a4, 160
@@ -6183,9 +6058,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_173
 ; CHECK-RV32-NEXT:  .LBB61_694: # %cond.load641
 ; CHECK-RV32-NEXT:    lbu a3, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v16, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v12, a3
 ; CHECK-RV32-NEXT:    li a3, 162
 ; CHECK-RV32-NEXT:    li a4, 161
@@ -6199,9 +6073,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_174
 ; CHECK-RV32-NEXT:  .LBB61_695: # %cond.load645
 ; CHECK-RV32-NEXT:    lbu a3, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v16, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v12, a3
 ; CHECK-RV32-NEXT:    li a3, 163
 ; CHECK-RV32-NEXT:    li a4, 162
@@ -6215,9 +6088,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_175
 ; CHECK-RV32-NEXT:  .LBB61_696: # %cond.load649
 ; CHECK-RV32-NEXT:    lbu a3, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v16, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v12, a3
 ; CHECK-RV32-NEXT:    li a3, 164
 ; CHECK-RV32-NEXT:    li a4, 163
@@ -6231,9 +6103,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_176
 ; CHECK-RV32-NEXT:  .LBB61_697: # %cond.load653
 ; CHECK-RV32-NEXT:    lbu a3, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v16, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v12, a3
 ; CHECK-RV32-NEXT:    li a3, 165
 ; CHECK-RV32-NEXT:    li a4, 164
@@ -6247,9 +6118,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_177
 ; CHECK-RV32-NEXT:  .LBB61_698: # %cond.load657
 ; CHECK-RV32-NEXT:    lbu a3, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v16, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v12, a3
 ; CHECK-RV32-NEXT:    li a3, 166
 ; CHECK-RV32-NEXT:    li a4, 165
@@ -6263,9 +6133,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_178
 ; CHECK-RV32-NEXT:  .LBB61_699: # %cond.load661
 ; CHECK-RV32-NEXT:    lbu a3, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v16, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v12, a3
 ; CHECK-RV32-NEXT:    li a3, 167
 ; CHECK-RV32-NEXT:    li a4, 166
@@ -6279,9 +6148,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_179
 ; CHECK-RV32-NEXT:  .LBB61_700: # %cond.load665
 ; CHECK-RV32-NEXT:    lbu a3, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v16, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v12, a3
 ; CHECK-RV32-NEXT:    li a3, 168
 ; CHECK-RV32-NEXT:    li a4, 167
@@ -6295,9 +6163,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_180
 ; CHECK-RV32-NEXT:  .LBB61_701: # %cond.load669
 ; CHECK-RV32-NEXT:    lbu a3, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v16, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v12, a3
 ; CHECK-RV32-NEXT:    li a3, 169
 ; CHECK-RV32-NEXT:    li a4, 168
@@ -6311,9 +6178,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_181
 ; CHECK-RV32-NEXT:  .LBB61_702: # %cond.load673
 ; CHECK-RV32-NEXT:    lbu a3, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v16, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v12, a3
 ; CHECK-RV32-NEXT:    li a3, 170
 ; CHECK-RV32-NEXT:    li a4, 169
@@ -6327,9 +6193,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_182
 ; CHECK-RV32-NEXT:  .LBB61_703: # %cond.load677
 ; CHECK-RV32-NEXT:    lbu a3, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v16, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v12, a3
 ; CHECK-RV32-NEXT:    li a3, 171
 ; CHECK-RV32-NEXT:    li a4, 170
@@ -6343,9 +6208,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_183
 ; CHECK-RV32-NEXT:  .LBB61_704: # %cond.load681
 ; CHECK-RV32-NEXT:    lbu a3, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v16, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v12, a3
 ; CHECK-RV32-NEXT:    li a3, 172
 ; CHECK-RV32-NEXT:    li a4, 171
@@ -6359,9 +6223,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_184
 ; CHECK-RV32-NEXT:  .LBB61_705: # %cond.load685
 ; CHECK-RV32-NEXT:    lbu a3, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v16, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v12, a3
 ; CHECK-RV32-NEXT:    li a3, 173
 ; CHECK-RV32-NEXT:    li a4, 172
@@ -6375,9 +6238,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_185
 ; CHECK-RV32-NEXT:  .LBB61_706: # %cond.load689
 ; CHECK-RV32-NEXT:    lbu a3, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v16, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v12, a3
 ; CHECK-RV32-NEXT:    li a3, 174
 ; CHECK-RV32-NEXT:    li a4, 173
@@ -6391,9 +6253,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_186
 ; CHECK-RV32-NEXT:  .LBB61_707: # %cond.load693
 ; CHECK-RV32-NEXT:    lbu a3, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v16, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v12, a3
 ; CHECK-RV32-NEXT:    li a3, 175
 ; CHECK-RV32-NEXT:    li a4, 174
@@ -6407,9 +6268,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_187
 ; CHECK-RV32-NEXT:  .LBB61_708: # %cond.load697
 ; CHECK-RV32-NEXT:    lbu a3, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v16, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v12, a3
 ; CHECK-RV32-NEXT:    li a3, 176
 ; CHECK-RV32-NEXT:    li a4, 175
@@ -6423,9 +6283,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_188
 ; CHECK-RV32-NEXT:  .LBB61_709: # %cond.load701
 ; CHECK-RV32-NEXT:    lbu a3, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v16, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v12, a3
 ; CHECK-RV32-NEXT:    li a3, 177
 ; CHECK-RV32-NEXT:    li a4, 176
@@ -6439,9 +6298,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_189
 ; CHECK-RV32-NEXT:  .LBB61_710: # %cond.load705
 ; CHECK-RV32-NEXT:    lbu a3, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v16, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v12, a3
 ; CHECK-RV32-NEXT:    li a3, 178
 ; CHECK-RV32-NEXT:    li a4, 177
@@ -6455,9 +6313,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_190
 ; CHECK-RV32-NEXT:  .LBB61_711: # %cond.load709
 ; CHECK-RV32-NEXT:    lbu a3, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v16, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v12, a3
 ; CHECK-RV32-NEXT:    li a3, 179
 ; CHECK-RV32-NEXT:    li a4, 178
@@ -6471,9 +6328,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_191
 ; CHECK-RV32-NEXT:  .LBB61_712: # %cond.load713
 ; CHECK-RV32-NEXT:    lbu a3, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v16, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v12, a3
 ; CHECK-RV32-NEXT:    li a3, 180
 ; CHECK-RV32-NEXT:    li a4, 179
@@ -6487,9 +6343,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_192
 ; CHECK-RV32-NEXT:  .LBB61_713: # %cond.load717
 ; CHECK-RV32-NEXT:    lbu a3, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v16, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v12, a3
 ; CHECK-RV32-NEXT:    li a3, 181
 ; CHECK-RV32-NEXT:    li a4, 180
@@ -6503,9 +6358,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_193
 ; CHECK-RV32-NEXT:  .LBB61_714: # %cond.load721
 ; CHECK-RV32-NEXT:    lbu a3, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v16, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v12, a3
 ; CHECK-RV32-NEXT:    li a3, 182
 ; CHECK-RV32-NEXT:    li a4, 181
@@ -6519,9 +6373,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_194
 ; CHECK-RV32-NEXT:  .LBB61_715: # %cond.load725
 ; CHECK-RV32-NEXT:    lbu a3, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v16, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v12, a3
 ; CHECK-RV32-NEXT:    li a3, 183
 ; CHECK-RV32-NEXT:    li a4, 182
@@ -6535,9 +6388,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_195
 ; CHECK-RV32-NEXT:  .LBB61_716: # %cond.load729
 ; CHECK-RV32-NEXT:    lbu a3, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v16, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v12, a3
 ; CHECK-RV32-NEXT:    li a3, 184
 ; CHECK-RV32-NEXT:    li a4, 183
@@ -6551,9 +6403,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_196
 ; CHECK-RV32-NEXT:  .LBB61_717: # %cond.load733
 ; CHECK-RV32-NEXT:    lbu a3, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v16, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v12, a3
 ; CHECK-RV32-NEXT:    li a3, 185
 ; CHECK-RV32-NEXT:    li a4, 184
@@ -6567,9 +6418,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_197
 ; CHECK-RV32-NEXT:  .LBB61_718: # %cond.load737
 ; CHECK-RV32-NEXT:    lbu a3, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v16, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v12, a3
 ; CHECK-RV32-NEXT:    li a3, 186
 ; CHECK-RV32-NEXT:    li a4, 185
@@ -6583,9 +6433,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_198
 ; CHECK-RV32-NEXT:  .LBB61_719: # %cond.load741
 ; CHECK-RV32-NEXT:    lbu a3, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v16, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v12, a3
 ; CHECK-RV32-NEXT:    li a3, 187
 ; CHECK-RV32-NEXT:    li a4, 186
@@ -6599,9 +6448,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_199
 ; CHECK-RV32-NEXT:  .LBB61_720: # %cond.load745
 ; CHECK-RV32-NEXT:    lbu a3, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v16, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v12, a3
 ; CHECK-RV32-NEXT:    li a3, 188
 ; CHECK-RV32-NEXT:    li a4, 187
@@ -6615,9 +6463,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_200
 ; CHECK-RV32-NEXT:  .LBB61_721: # %cond.load749
 ; CHECK-RV32-NEXT:    lbu a3, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v16, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v12, a3
 ; CHECK-RV32-NEXT:    li a3, 189
 ; CHECK-RV32-NEXT:    li a4, 188
@@ -6647,9 +6494,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_206
 ; CHECK-RV32-NEXT:  .LBB61_723: # %cond.load765
 ; CHECK-RV32-NEXT:    lbu a2, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v24, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v12, a2
 ; CHECK-RV32-NEXT:    li a2, 193
 ; CHECK-RV32-NEXT:    li a4, 192
@@ -6663,9 +6509,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_207
 ; CHECK-RV32-NEXT:  .LBB61_724: # %cond.load769
 ; CHECK-RV32-NEXT:    lbu a2, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v24, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v12, a2
 ; CHECK-RV32-NEXT:    li a2, 194
 ; CHECK-RV32-NEXT:    li a4, 193
@@ -6679,9 +6524,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_208
 ; CHECK-RV32-NEXT:  .LBB61_725: # %cond.load773
 ; CHECK-RV32-NEXT:    lbu a2, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v24, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v12, a2
 ; CHECK-RV32-NEXT:    li a2, 195
 ; CHECK-RV32-NEXT:    li a4, 194
@@ -6695,9 +6539,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_209
 ; CHECK-RV32-NEXT:  .LBB61_726: # %cond.load777
 ; CHECK-RV32-NEXT:    lbu a2, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v24, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v12, a2
 ; CHECK-RV32-NEXT:    li a2, 196
 ; CHECK-RV32-NEXT:    li a4, 195
@@ -6711,9 +6554,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_210
 ; CHECK-RV32-NEXT:  .LBB61_727: # %cond.load781
 ; CHECK-RV32-NEXT:    lbu a2, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v24, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v12, a2
 ; CHECK-RV32-NEXT:    li a2, 197
 ; CHECK-RV32-NEXT:    li a4, 196
@@ -6727,9 +6569,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_211
 ; CHECK-RV32-NEXT:  .LBB61_728: # %cond.load785
 ; CHECK-RV32-NEXT:    lbu a2, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v24, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v12, a2
 ; CHECK-RV32-NEXT:    li a2, 198
 ; CHECK-RV32-NEXT:    li a4, 197
@@ -6743,9 +6584,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_212
 ; CHECK-RV32-NEXT:  .LBB61_729: # %cond.load789
 ; CHECK-RV32-NEXT:    lbu a2, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v24, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v12, a2
 ; CHECK-RV32-NEXT:    li a2, 199
 ; CHECK-RV32-NEXT:    li a4, 198
@@ -6759,9 +6599,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_213
 ; CHECK-RV32-NEXT:  .LBB61_730: # %cond.load793
 ; CHECK-RV32-NEXT:    lbu a2, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v24, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v12, a2
 ; CHECK-RV32-NEXT:    li a2, 200
 ; CHECK-RV32-NEXT:    li a4, 199
@@ -6775,9 +6614,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_214
 ; CHECK-RV32-NEXT:  .LBB61_731: # %cond.load797
 ; CHECK-RV32-NEXT:    lbu a2, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v24, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v12, a2
 ; CHECK-RV32-NEXT:    li a2, 201
 ; CHECK-RV32-NEXT:    li a4, 200
@@ -6791,9 +6629,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_215
 ; CHECK-RV32-NEXT:  .LBB61_732: # %cond.load801
 ; CHECK-RV32-NEXT:    lbu a2, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v24, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v12, a2
 ; CHECK-RV32-NEXT:    li a2, 202
 ; CHECK-RV32-NEXT:    li a4, 201
@@ -6807,9 +6644,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_216
 ; CHECK-RV32-NEXT:  .LBB61_733: # %cond.load805
 ; CHECK-RV32-NEXT:    lbu a2, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v24, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v12, a2
 ; CHECK-RV32-NEXT:    li a2, 203
 ; CHECK-RV32-NEXT:    li a4, 202
@@ -6823,9 +6659,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_217
 ; CHECK-RV32-NEXT:  .LBB61_734: # %cond.load809
 ; CHECK-RV32-NEXT:    lbu a2, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v24, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v12, a2
 ; CHECK-RV32-NEXT:    li a2, 204
 ; CHECK-RV32-NEXT:    li a4, 203
@@ -6839,9 +6674,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_218
 ; CHECK-RV32-NEXT:  .LBB61_735: # %cond.load813
 ; CHECK-RV32-NEXT:    lbu a2, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v24, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v12, a2
 ; CHECK-RV32-NEXT:    li a2, 205
 ; CHECK-RV32-NEXT:    li a4, 204
@@ -6855,9 +6689,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_219
 ; CHECK-RV32-NEXT:  .LBB61_736: # %cond.load817
 ; CHECK-RV32-NEXT:    lbu a2, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v24, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v12, a2
 ; CHECK-RV32-NEXT:    li a2, 206
 ; CHECK-RV32-NEXT:    li a4, 205
@@ -6871,9 +6704,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_220
 ; CHECK-RV32-NEXT:  .LBB61_737: # %cond.load821
 ; CHECK-RV32-NEXT:    lbu a2, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v24, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v12, a2
 ; CHECK-RV32-NEXT:    li a2, 207
 ; CHECK-RV32-NEXT:    li a4, 206
@@ -6887,9 +6719,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_221
 ; CHECK-RV32-NEXT:  .LBB61_738: # %cond.load825
 ; CHECK-RV32-NEXT:    lbu a2, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v24, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v12, a2
 ; CHECK-RV32-NEXT:    li a2, 208
 ; CHECK-RV32-NEXT:    li a4, 207
@@ -6903,9 +6734,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_222
 ; CHECK-RV32-NEXT:  .LBB61_739: # %cond.load829
 ; CHECK-RV32-NEXT:    lbu a2, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v24, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v12, a2
 ; CHECK-RV32-NEXT:    li a2, 209
 ; CHECK-RV32-NEXT:    li a4, 208
@@ -6919,9 +6749,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_223
 ; CHECK-RV32-NEXT:  .LBB61_740: # %cond.load833
 ; CHECK-RV32-NEXT:    lbu a2, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v24, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v12, a2
 ; CHECK-RV32-NEXT:    li a2, 210
 ; CHECK-RV32-NEXT:    li a4, 209
@@ -6935,9 +6764,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_224
 ; CHECK-RV32-NEXT:  .LBB61_741: # %cond.load837
 ; CHECK-RV32-NEXT:    lbu a2, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v24, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v12, a2
 ; CHECK-RV32-NEXT:    li a2, 211
 ; CHECK-RV32-NEXT:    li a4, 210
@@ -6951,9 +6779,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_225
 ; CHECK-RV32-NEXT:  .LBB61_742: # %cond.load841
 ; CHECK-RV32-NEXT:    lbu a2, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v24, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v12, a2
 ; CHECK-RV32-NEXT:    li a2, 212
 ; CHECK-RV32-NEXT:    li a4, 211
@@ -6967,9 +6794,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_226
 ; CHECK-RV32-NEXT:  .LBB61_743: # %cond.load845
 ; CHECK-RV32-NEXT:    lbu a2, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v24, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v12, a2
 ; CHECK-RV32-NEXT:    li a2, 213
 ; CHECK-RV32-NEXT:    li a4, 212
@@ -6983,9 +6809,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_227
 ; CHECK-RV32-NEXT:  .LBB61_744: # %cond.load849
 ; CHECK-RV32-NEXT:    lbu a2, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v24, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v12, a2
 ; CHECK-RV32-NEXT:    li a2, 214
 ; CHECK-RV32-NEXT:    li a4, 213
@@ -6999,9 +6824,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_228
 ; CHECK-RV32-NEXT:  .LBB61_745: # %cond.load853
 ; CHECK-RV32-NEXT:    lbu a2, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v24, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v12, a2
 ; CHECK-RV32-NEXT:    li a2, 215
 ; CHECK-RV32-NEXT:    li a4, 214
@@ -7015,9 +6839,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_229
 ; CHECK-RV32-NEXT:  .LBB61_746: # %cond.load857
 ; CHECK-RV32-NEXT:    lbu a2, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v24, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v12, a2
 ; CHECK-RV32-NEXT:    li a2, 216
 ; CHECK-RV32-NEXT:    li a4, 215
@@ -7031,9 +6854,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_230
 ; CHECK-RV32-NEXT:  .LBB61_747: # %cond.load861
 ; CHECK-RV32-NEXT:    lbu a2, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v24, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v12, a2
 ; CHECK-RV32-NEXT:    li a2, 217
 ; CHECK-RV32-NEXT:    li a4, 216
@@ -7047,9 +6869,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_231
 ; CHECK-RV32-NEXT:  .LBB61_748: # %cond.load865
 ; CHECK-RV32-NEXT:    lbu a2, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v24, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v12, a2
 ; CHECK-RV32-NEXT:    li a2, 218
 ; CHECK-RV32-NEXT:    li a4, 217
@@ -7063,9 +6884,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_232
 ; CHECK-RV32-NEXT:  .LBB61_749: # %cond.load869
 ; CHECK-RV32-NEXT:    lbu a2, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v24, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v12, a2
 ; CHECK-RV32-NEXT:    li a2, 219
 ; CHECK-RV32-NEXT:    li a4, 218
@@ -7079,9 +6899,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_233
 ; CHECK-RV32-NEXT:  .LBB61_750: # %cond.load873
 ; CHECK-RV32-NEXT:    lbu a2, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v24, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v12, a2
 ; CHECK-RV32-NEXT:    li a2, 220
 ; CHECK-RV32-NEXT:    li a4, 219
@@ -7095,9 +6914,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_234
 ; CHECK-RV32-NEXT:  .LBB61_751: # %cond.load877
 ; CHECK-RV32-NEXT:    lbu a2, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v24, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v12, a2
 ; CHECK-RV32-NEXT:    li a2, 221
 ; CHECK-RV32-NEXT:    li a4, 220
@@ -7127,9 +6945,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_240
 ; CHECK-RV32-NEXT:  .LBB61_753: # %cond.load893
 ; CHECK-RV32-NEXT:    lbu a3, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v16, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v12, a3
 ; CHECK-RV32-NEXT:    li a3, 225
 ; CHECK-RV32-NEXT:    li a4, 224
@@ -7143,9 +6960,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_241
 ; CHECK-RV32-NEXT:  .LBB61_754: # %cond.load897
 ; CHECK-RV32-NEXT:    lbu a3, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v16, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v12, a3
 ; CHECK-RV32-NEXT:    li a3, 226
 ; CHECK-RV32-NEXT:    li a4, 225
@@ -7159,9 +6975,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_242
 ; CHECK-RV32-NEXT:  .LBB61_755: # %cond.load901
 ; CHECK-RV32-NEXT:    lbu a3, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v16, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v12, a3
 ; CHECK-RV32-NEXT:    li a3, 227
 ; CHECK-RV32-NEXT:    li a4, 226
@@ -7175,9 +6990,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_243
 ; CHECK-RV32-NEXT:  .LBB61_756: # %cond.load905
 ; CHECK-RV32-NEXT:    lbu a3, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v16, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v12, a3
 ; CHECK-RV32-NEXT:    li a3, 228
 ; CHECK-RV32-NEXT:    li a4, 227
@@ -7191,9 +7005,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_244
 ; CHECK-RV32-NEXT:  .LBB61_757: # %cond.load909
 ; CHECK-RV32-NEXT:    lbu a3, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v16, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v12, a3
 ; CHECK-RV32-NEXT:    li a3, 229
 ; CHECK-RV32-NEXT:    li a4, 228
@@ -7207,9 +7020,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_245
 ; CHECK-RV32-NEXT:  .LBB61_758: # %cond.load913
 ; CHECK-RV32-NEXT:    lbu a3, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v16, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v12, a3
 ; CHECK-RV32-NEXT:    li a3, 230
 ; CHECK-RV32-NEXT:    li a4, 229
@@ -7223,9 +7035,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_246
 ; CHECK-RV32-NEXT:  .LBB61_759: # %cond.load917
 ; CHECK-RV32-NEXT:    lbu a3, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v16, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v12, a3
 ; CHECK-RV32-NEXT:    li a3, 231
 ; CHECK-RV32-NEXT:    li a4, 230
@@ -7239,9 +7050,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_247
 ; CHECK-RV32-NEXT:  .LBB61_760: # %cond.load921
 ; CHECK-RV32-NEXT:    lbu a3, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v16, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v12, a3
 ; CHECK-RV32-NEXT:    li a3, 232
 ; CHECK-RV32-NEXT:    li a4, 231
@@ -7255,9 +7065,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_248
 ; CHECK-RV32-NEXT:  .LBB61_761: # %cond.load925
 ; CHECK-RV32-NEXT:    lbu a3, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v16, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v12, a3
 ; CHECK-RV32-NEXT:    li a3, 233
 ; CHECK-RV32-NEXT:    li a4, 232
@@ -7271,9 +7080,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_249
 ; CHECK-RV32-NEXT:  .LBB61_762: # %cond.load929
 ; CHECK-RV32-NEXT:    lbu a3, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v16, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v12, a3
 ; CHECK-RV32-NEXT:    li a3, 234
 ; CHECK-RV32-NEXT:    li a4, 233
@@ -7287,9 +7095,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_250
 ; CHECK-RV32-NEXT:  .LBB61_763: # %cond.load933
 ; CHECK-RV32-NEXT:    lbu a3, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v16, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v12, a3
 ; CHECK-RV32-NEXT:    li a3, 235
 ; CHECK-RV32-NEXT:    li a4, 234
@@ -7303,9 +7110,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_251
 ; CHECK-RV32-NEXT:  .LBB61_764: # %cond.load937
 ; CHECK-RV32-NEXT:    lbu a3, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v16, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v12, a3
 ; CHECK-RV32-NEXT:    li a3, 236
 ; CHECK-RV32-NEXT:    li a4, 235
@@ -7319,9 +7125,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_252
 ; CHECK-RV32-NEXT:  .LBB61_765: # %cond.load941
 ; CHECK-RV32-NEXT:    lbu a3, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v16, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v12, a3
 ; CHECK-RV32-NEXT:    li a3, 237
 ; CHECK-RV32-NEXT:    li a4, 236
@@ -7335,9 +7140,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_253
 ; CHECK-RV32-NEXT:  .LBB61_766: # %cond.load945
 ; CHECK-RV32-NEXT:    lbu a3, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v16, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v12, a3
 ; CHECK-RV32-NEXT:    li a3, 238
 ; CHECK-RV32-NEXT:    li a4, 237
@@ -7351,9 +7155,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_254
 ; CHECK-RV32-NEXT:  .LBB61_767: # %cond.load949
 ; CHECK-RV32-NEXT:    lbu a3, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v16, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v12, a3
 ; CHECK-RV32-NEXT:    li a3, 239
 ; CHECK-RV32-NEXT:    li a4, 238
@@ -7367,9 +7170,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_255
 ; CHECK-RV32-NEXT:  .LBB61_768: # %cond.load953
 ; CHECK-RV32-NEXT:    lbu a3, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v16, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v12, a3
 ; CHECK-RV32-NEXT:    li a3, 240
 ; CHECK-RV32-NEXT:    li a4, 239
@@ -7383,9 +7185,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_256
 ; CHECK-RV32-NEXT:  .LBB61_769: # %cond.load957
 ; CHECK-RV32-NEXT:    lbu a3, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v16, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v12, a3
 ; CHECK-RV32-NEXT:    li a3, 241
 ; CHECK-RV32-NEXT:    li a4, 240
@@ -7399,9 +7200,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_257
 ; CHECK-RV32-NEXT:  .LBB61_770: # %cond.load961
 ; CHECK-RV32-NEXT:    lbu a3, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v16, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v12, a3
 ; CHECK-RV32-NEXT:    li a3, 242
 ; CHECK-RV32-NEXT:    li a4, 241
@@ -7415,9 +7215,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_258
 ; CHECK-RV32-NEXT:  .LBB61_771: # %cond.load965
 ; CHECK-RV32-NEXT:    lbu a3, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v16, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v12, a3
 ; CHECK-RV32-NEXT:    li a3, 243
 ; CHECK-RV32-NEXT:    li a4, 242
@@ -7431,9 +7230,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_259
 ; CHECK-RV32-NEXT:  .LBB61_772: # %cond.load969
 ; CHECK-RV32-NEXT:    lbu a3, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v16, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v12, a3
 ; CHECK-RV32-NEXT:    li a3, 244
 ; CHECK-RV32-NEXT:    li a4, 243
@@ -7447,9 +7245,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_260
 ; CHECK-RV32-NEXT:  .LBB61_773: # %cond.load973
 ; CHECK-RV32-NEXT:    lbu a3, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v16, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v12, a3
 ; CHECK-RV32-NEXT:    li a3, 245
 ; CHECK-RV32-NEXT:    li a4, 244
@@ -7463,9 +7260,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_261
 ; CHECK-RV32-NEXT:  .LBB61_774: # %cond.load977
 ; CHECK-RV32-NEXT:    lbu a3, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v16, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v12, a3
 ; CHECK-RV32-NEXT:    li a3, 246
 ; CHECK-RV32-NEXT:    li a4, 245
@@ -7479,9 +7275,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_262
 ; CHECK-RV32-NEXT:  .LBB61_775: # %cond.load981
 ; CHECK-RV32-NEXT:    lbu a3, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v16, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v12, a3
 ; CHECK-RV32-NEXT:    li a3, 247
 ; CHECK-RV32-NEXT:    li a4, 246
@@ -7495,9 +7290,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_263
 ; CHECK-RV32-NEXT:  .LBB61_776: # %cond.load985
 ; CHECK-RV32-NEXT:    lbu a3, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v16, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v12, a3
 ; CHECK-RV32-NEXT:    li a3, 248
 ; CHECK-RV32-NEXT:    li a4, 247
@@ -7511,9 +7305,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_264
 ; CHECK-RV32-NEXT:  .LBB61_777: # %cond.load989
 ; CHECK-RV32-NEXT:    lbu a3, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v16, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v12, a3
 ; CHECK-RV32-NEXT:    li a3, 249
 ; CHECK-RV32-NEXT:    li a4, 248
@@ -7527,9 +7320,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_265
 ; CHECK-RV32-NEXT:  .LBB61_778: # %cond.load993
 ; CHECK-RV32-NEXT:    lbu a3, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v16, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v12, a3
 ; CHECK-RV32-NEXT:    li a3, 250
 ; CHECK-RV32-NEXT:    li a4, 249
@@ -7543,9 +7335,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_266
 ; CHECK-RV32-NEXT:  .LBB61_779: # %cond.load997
 ; CHECK-RV32-NEXT:    lbu a3, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v16, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v12, a3
 ; CHECK-RV32-NEXT:    li a3, 251
 ; CHECK-RV32-NEXT:    li a4, 250
@@ -7559,9 +7350,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_267
 ; CHECK-RV32-NEXT:  .LBB61_780: # %cond.load1001
 ; CHECK-RV32-NEXT:    lbu a3, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v16, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v12, a3
 ; CHECK-RV32-NEXT:    li a3, 252
 ; CHECK-RV32-NEXT:    li a4, 251
@@ -7575,9 +7365,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV32-NEXT:    j .LBB61_268
 ; CHECK-RV32-NEXT:  .LBB61_781: # %cond.load1005
 ; CHECK-RV32-NEXT:    lbu a3, 0(a0)
-; CHECK-RV32-NEXT:    li a4, 512
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv8r.v v16, v8
-; CHECK-RV32-NEXT:    vsetvli zero, a4, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv.s.x v12, a3
 ; CHECK-RV32-NEXT:    li a3, 253
 ; CHECK-RV32-NEXT:    li a4, 252
@@ -10998,9 +10787,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    bgez a1, .LBB61_63
 ; CHECK-RV64-NEXT:  .LBB61_62: # %cond.load241
 ; CHECK-RV64-NEXT:    lbu a1, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v9, a1
 ; CHECK-RV64-NEXT:    li a1, 62
 ; CHECK-RV64-NEXT:    li a3, 61
@@ -11279,9 +11067,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    bgez a2, .LBB61_129
 ; CHECK-RV64-NEXT:  .LBB61_128: # %cond.load497
 ; CHECK-RV64-NEXT:    lbu a2, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v10, a2
 ; CHECK-RV64-NEXT:    li a2, 126
 ; CHECK-RV64-NEXT:    li a3, 125
@@ -11560,9 +11347,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    bgez a1, .LBB61_195
 ; CHECK-RV64-NEXT:  .LBB61_194: # %cond.load753
 ; CHECK-RV64-NEXT:    lbu a1, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v12, a1
 ; CHECK-RV64-NEXT:    li a1, 190
 ; CHECK-RV64-NEXT:    li a3, 189
@@ -11841,9 +11627,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    bgez a2, .LBB61_261
 ; CHECK-RV64-NEXT:  .LBB61_260: # %cond.load1009
 ; CHECK-RV64-NEXT:    lbu a2, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v12, a2
 ; CHECK-RV64-NEXT:    li a2, 254
 ; CHECK-RV64-NEXT:    li a3, 253
@@ -12968,10 +12753,9 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_2
 ; CHECK-RV64-NEXT:  .LBB61_528: # %cond.load1
 ; CHECK-RV64-NEXT:    lbu a1, 0(a0)
+; CHECK-RV64-NEXT:    vsetivli zero, 2, e8, m1, tu, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, zero, e8, mf8, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v9, a1
-; CHECK-RV64-NEXT:    vsetivli zero, 2, e8, m1, tu, ma
 ; CHECK-RV64-NEXT:    vslideup.vi v8, v9, 1
 ; CHECK-RV64-NEXT:    addi a0, a0, 1
 ; CHECK-RV64-NEXT:    vmv1r.v v16, v8
@@ -12981,8 +12765,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_3
 ; CHECK-RV64-NEXT:  .LBB61_529: # %cond.load5
 ; CHECK-RV64-NEXT:    lbu a1, 0(a0)
-; CHECK-RV64-NEXT:    vmv8r.v v16, v8
 ; CHECK-RV64-NEXT:    vsetivli zero, 3, e8, m1, tu, ma
+; CHECK-RV64-NEXT:    vmv8r.v v16, v8
 ; CHECK-RV64-NEXT:    vmv.s.x v9, a1
 ; CHECK-RV64-NEXT:    vslideup.vi v8, v9, 2
 ; CHECK-RV64-NEXT:    addi a0, a0, 1
@@ -12993,8 +12777,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_4
 ; CHECK-RV64-NEXT:  .LBB61_530: # %cond.load9
 ; CHECK-RV64-NEXT:    lbu a1, 0(a0)
-; CHECK-RV64-NEXT:    vmv8r.v v16, v8
 ; CHECK-RV64-NEXT:    vsetivli zero, 4, e8, m1, tu, ma
+; CHECK-RV64-NEXT:    vmv8r.v v16, v8
 ; CHECK-RV64-NEXT:    vmv.s.x v9, a1
 ; CHECK-RV64-NEXT:    vslideup.vi v8, v9, 3
 ; CHECK-RV64-NEXT:    addi a0, a0, 1
@@ -13005,8 +12789,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_5
 ; CHECK-RV64-NEXT:  .LBB61_531: # %cond.load13
 ; CHECK-RV64-NEXT:    lbu a1, 0(a0)
-; CHECK-RV64-NEXT:    vmv8r.v v16, v8
 ; CHECK-RV64-NEXT:    vsetivli zero, 5, e8, m1, tu, ma
+; CHECK-RV64-NEXT:    vmv8r.v v16, v8
 ; CHECK-RV64-NEXT:    vmv.s.x v9, a1
 ; CHECK-RV64-NEXT:    vslideup.vi v8, v9, 4
 ; CHECK-RV64-NEXT:    addi a0, a0, 1
@@ -13017,8 +12801,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_6
 ; CHECK-RV64-NEXT:  .LBB61_532: # %cond.load17
 ; CHECK-RV64-NEXT:    lbu a1, 0(a0)
-; CHECK-RV64-NEXT:    vmv8r.v v16, v8
 ; CHECK-RV64-NEXT:    vsetivli zero, 6, e8, m1, tu, ma
+; CHECK-RV64-NEXT:    vmv8r.v v16, v8
 ; CHECK-RV64-NEXT:    vmv.s.x v9, a1
 ; CHECK-RV64-NEXT:    vslideup.vi v8, v9, 5
 ; CHECK-RV64-NEXT:    addi a0, a0, 1
@@ -13029,8 +12813,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_7
 ; CHECK-RV64-NEXT:  .LBB61_533: # %cond.load21
 ; CHECK-RV64-NEXT:    lbu a1, 0(a0)
-; CHECK-RV64-NEXT:    vmv8r.v v16, v8
 ; CHECK-RV64-NEXT:    vsetivli zero, 7, e8, m1, tu, ma
+; CHECK-RV64-NEXT:    vmv8r.v v16, v8
 ; CHECK-RV64-NEXT:    vmv.s.x v9, a1
 ; CHECK-RV64-NEXT:    vslideup.vi v8, v9, 6
 ; CHECK-RV64-NEXT:    addi a0, a0, 1
@@ -13041,8 +12825,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_8
 ; CHECK-RV64-NEXT:  .LBB61_534: # %cond.load25
 ; CHECK-RV64-NEXT:    lbu a1, 0(a0)
-; CHECK-RV64-NEXT:    vmv8r.v v16, v8
 ; CHECK-RV64-NEXT:    vsetivli zero, 8, e8, m1, tu, ma
+; CHECK-RV64-NEXT:    vmv8r.v v16, v8
 ; CHECK-RV64-NEXT:    vmv.s.x v9, a1
 ; CHECK-RV64-NEXT:    vslideup.vi v8, v9, 7
 ; CHECK-RV64-NEXT:    addi a0, a0, 1
@@ -13053,8 +12837,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_9
 ; CHECK-RV64-NEXT:  .LBB61_535: # %cond.load29
 ; CHECK-RV64-NEXT:    lbu a1, 0(a0)
-; CHECK-RV64-NEXT:    vmv8r.v v16, v8
 ; CHECK-RV64-NEXT:    vsetivli zero, 9, e8, m1, tu, ma
+; CHECK-RV64-NEXT:    vmv8r.v v16, v8
 ; CHECK-RV64-NEXT:    vmv.s.x v9, a1
 ; CHECK-RV64-NEXT:    vslideup.vi v8, v9, 8
 ; CHECK-RV64-NEXT:    addi a0, a0, 1
@@ -13065,8 +12849,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_10
 ; CHECK-RV64-NEXT:  .LBB61_536: # %cond.load33
 ; CHECK-RV64-NEXT:    lbu a1, 0(a0)
-; CHECK-RV64-NEXT:    vmv8r.v v16, v8
 ; CHECK-RV64-NEXT:    vsetivli zero, 10, e8, m1, tu, ma
+; CHECK-RV64-NEXT:    vmv8r.v v16, v8
 ; CHECK-RV64-NEXT:    vmv.s.x v9, a1
 ; CHECK-RV64-NEXT:    vslideup.vi v8, v9, 9
 ; CHECK-RV64-NEXT:    addi a0, a0, 1
@@ -13077,8 +12861,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_11
 ; CHECK-RV64-NEXT:  .LBB61_537: # %cond.load37
 ; CHECK-RV64-NEXT:    lbu a1, 0(a0)
-; CHECK-RV64-NEXT:    vmv8r.v v16, v8
 ; CHECK-RV64-NEXT:    vsetivli zero, 11, e8, m1, tu, ma
+; CHECK-RV64-NEXT:    vmv8r.v v16, v8
 ; CHECK-RV64-NEXT:    vmv.s.x v9, a1
 ; CHECK-RV64-NEXT:    vslideup.vi v8, v9, 10
 ; CHECK-RV64-NEXT:    addi a0, a0, 1
@@ -13089,8 +12873,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_12
 ; CHECK-RV64-NEXT:  .LBB61_538: # %cond.load41
 ; CHECK-RV64-NEXT:    lbu a1, 0(a0)
-; CHECK-RV64-NEXT:    vmv8r.v v16, v8
 ; CHECK-RV64-NEXT:    vsetivli zero, 12, e8, m1, tu, ma
+; CHECK-RV64-NEXT:    vmv8r.v v16, v8
 ; CHECK-RV64-NEXT:    vmv.s.x v9, a1
 ; CHECK-RV64-NEXT:    vslideup.vi v8, v9, 11
 ; CHECK-RV64-NEXT:    addi a0, a0, 1
@@ -13101,8 +12885,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_13
 ; CHECK-RV64-NEXT:  .LBB61_539: # %cond.load45
 ; CHECK-RV64-NEXT:    lbu a1, 0(a0)
-; CHECK-RV64-NEXT:    vmv8r.v v16, v8
 ; CHECK-RV64-NEXT:    vsetivli zero, 13, e8, m1, tu, ma
+; CHECK-RV64-NEXT:    vmv8r.v v16, v8
 ; CHECK-RV64-NEXT:    vmv.s.x v9, a1
 ; CHECK-RV64-NEXT:    vslideup.vi v8, v9, 12
 ; CHECK-RV64-NEXT:    addi a0, a0, 1
@@ -13113,8 +12897,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_14
 ; CHECK-RV64-NEXT:  .LBB61_540: # %cond.load49
 ; CHECK-RV64-NEXT:    lbu a1, 0(a0)
-; CHECK-RV64-NEXT:    vmv8r.v v16, v8
 ; CHECK-RV64-NEXT:    vsetivli zero, 14, e8, m1, tu, ma
+; CHECK-RV64-NEXT:    vmv8r.v v16, v8
 ; CHECK-RV64-NEXT:    vmv.s.x v9, a1
 ; CHECK-RV64-NEXT:    vslideup.vi v8, v9, 13
 ; CHECK-RV64-NEXT:    addi a0, a0, 1
@@ -13125,8 +12909,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_15
 ; CHECK-RV64-NEXT:  .LBB61_541: # %cond.load53
 ; CHECK-RV64-NEXT:    lbu a1, 0(a0)
-; CHECK-RV64-NEXT:    vmv8r.v v16, v8
 ; CHECK-RV64-NEXT:    vsetivli zero, 15, e8, m1, tu, ma
+; CHECK-RV64-NEXT:    vmv8r.v v16, v8
 ; CHECK-RV64-NEXT:    vmv.s.x v9, a1
 ; CHECK-RV64-NEXT:    vslideup.vi v8, v9, 14
 ; CHECK-RV64-NEXT:    addi a0, a0, 1
@@ -13137,8 +12921,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_16
 ; CHECK-RV64-NEXT:  .LBB61_542: # %cond.load57
 ; CHECK-RV64-NEXT:    lbu a1, 0(a0)
-; CHECK-RV64-NEXT:    vmv8r.v v16, v8
 ; CHECK-RV64-NEXT:    vsetivli zero, 16, e8, m1, tu, ma
+; CHECK-RV64-NEXT:    vmv8r.v v16, v8
 ; CHECK-RV64-NEXT:    vmv.s.x v9, a1
 ; CHECK-RV64-NEXT:    vslideup.vi v8, v9, 15
 ; CHECK-RV64-NEXT:    addi a0, a0, 1
@@ -13149,8 +12933,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_17
 ; CHECK-RV64-NEXT:  .LBB61_543: # %cond.load61
 ; CHECK-RV64-NEXT:    lbu a1, 0(a0)
-; CHECK-RV64-NEXT:    vmv8r.v v16, v8
 ; CHECK-RV64-NEXT:    vsetivli zero, 17, e8, m1, tu, ma
+; CHECK-RV64-NEXT:    vmv8r.v v16, v8
 ; CHECK-RV64-NEXT:    vmv.s.x v9, a1
 ; CHECK-RV64-NEXT:    vslideup.vi v8, v9, 16
 ; CHECK-RV64-NEXT:    addi a0, a0, 1
@@ -13161,8 +12945,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_18
 ; CHECK-RV64-NEXT:  .LBB61_544: # %cond.load65
 ; CHECK-RV64-NEXT:    lbu a1, 0(a0)
-; CHECK-RV64-NEXT:    vmv8r.v v16, v8
 ; CHECK-RV64-NEXT:    vsetivli zero, 18, e8, m1, tu, ma
+; CHECK-RV64-NEXT:    vmv8r.v v16, v8
 ; CHECK-RV64-NEXT:    vmv.s.x v9, a1
 ; CHECK-RV64-NEXT:    vslideup.vi v8, v9, 17
 ; CHECK-RV64-NEXT:    addi a0, a0, 1
@@ -13173,8 +12957,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_19
 ; CHECK-RV64-NEXT:  .LBB61_545: # %cond.load69
 ; CHECK-RV64-NEXT:    lbu a1, 0(a0)
-; CHECK-RV64-NEXT:    vmv8r.v v16, v8
 ; CHECK-RV64-NEXT:    vsetivli zero, 19, e8, m1, tu, ma
+; CHECK-RV64-NEXT:    vmv8r.v v16, v8
 ; CHECK-RV64-NEXT:    vmv.s.x v9, a1
 ; CHECK-RV64-NEXT:    vslideup.vi v8, v9, 18
 ; CHECK-RV64-NEXT:    addi a0, a0, 1
@@ -13185,8 +12969,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_20
 ; CHECK-RV64-NEXT:  .LBB61_546: # %cond.load73
 ; CHECK-RV64-NEXT:    lbu a1, 0(a0)
-; CHECK-RV64-NEXT:    vmv8r.v v16, v8
 ; CHECK-RV64-NEXT:    vsetivli zero, 20, e8, m1, tu, ma
+; CHECK-RV64-NEXT:    vmv8r.v v16, v8
 ; CHECK-RV64-NEXT:    vmv.s.x v9, a1
 ; CHECK-RV64-NEXT:    vslideup.vi v8, v9, 19
 ; CHECK-RV64-NEXT:    addi a0, a0, 1
@@ -13197,8 +12981,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_21
 ; CHECK-RV64-NEXT:  .LBB61_547: # %cond.load77
 ; CHECK-RV64-NEXT:    lbu a1, 0(a0)
-; CHECK-RV64-NEXT:    vmv8r.v v16, v8
 ; CHECK-RV64-NEXT:    vsetivli zero, 21, e8, m1, tu, ma
+; CHECK-RV64-NEXT:    vmv8r.v v16, v8
 ; CHECK-RV64-NEXT:    vmv.s.x v9, a1
 ; CHECK-RV64-NEXT:    vslideup.vi v8, v9, 20
 ; CHECK-RV64-NEXT:    addi a0, a0, 1
@@ -13209,8 +12993,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_22
 ; CHECK-RV64-NEXT:  .LBB61_548: # %cond.load81
 ; CHECK-RV64-NEXT:    lbu a1, 0(a0)
-; CHECK-RV64-NEXT:    vmv8r.v v16, v8
 ; CHECK-RV64-NEXT:    vsetivli zero, 22, e8, m1, tu, ma
+; CHECK-RV64-NEXT:    vmv8r.v v16, v8
 ; CHECK-RV64-NEXT:    vmv.s.x v9, a1
 ; CHECK-RV64-NEXT:    vslideup.vi v8, v9, 21
 ; CHECK-RV64-NEXT:    addi a0, a0, 1
@@ -13221,8 +13005,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_23
 ; CHECK-RV64-NEXT:  .LBB61_549: # %cond.load85
 ; CHECK-RV64-NEXT:    lbu a1, 0(a0)
-; CHECK-RV64-NEXT:    vmv8r.v v16, v8
 ; CHECK-RV64-NEXT:    vsetivli zero, 23, e8, m1, tu, ma
+; CHECK-RV64-NEXT:    vmv8r.v v16, v8
 ; CHECK-RV64-NEXT:    vmv.s.x v9, a1
 ; CHECK-RV64-NEXT:    vslideup.vi v8, v9, 22
 ; CHECK-RV64-NEXT:    addi a0, a0, 1
@@ -13233,8 +13017,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_24
 ; CHECK-RV64-NEXT:  .LBB61_550: # %cond.load89
 ; CHECK-RV64-NEXT:    lbu a1, 0(a0)
-; CHECK-RV64-NEXT:    vmv8r.v v16, v8
 ; CHECK-RV64-NEXT:    vsetivli zero, 24, e8, m1, tu, ma
+; CHECK-RV64-NEXT:    vmv8r.v v16, v8
 ; CHECK-RV64-NEXT:    vmv.s.x v9, a1
 ; CHECK-RV64-NEXT:    vslideup.vi v8, v9, 23
 ; CHECK-RV64-NEXT:    addi a0, a0, 1
@@ -13245,8 +13029,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_25
 ; CHECK-RV64-NEXT:  .LBB61_551: # %cond.load93
 ; CHECK-RV64-NEXT:    lbu a1, 0(a0)
-; CHECK-RV64-NEXT:    vmv8r.v v16, v8
 ; CHECK-RV64-NEXT:    vsetivli zero, 25, e8, m1, tu, ma
+; CHECK-RV64-NEXT:    vmv8r.v v16, v8
 ; CHECK-RV64-NEXT:    vmv.s.x v9, a1
 ; CHECK-RV64-NEXT:    vslideup.vi v8, v9, 24
 ; CHECK-RV64-NEXT:    addi a0, a0, 1
@@ -13257,8 +13041,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_26
 ; CHECK-RV64-NEXT:  .LBB61_552: # %cond.load97
 ; CHECK-RV64-NEXT:    lbu a1, 0(a0)
-; CHECK-RV64-NEXT:    vmv8r.v v16, v8
 ; CHECK-RV64-NEXT:    vsetivli zero, 26, e8, m1, tu, ma
+; CHECK-RV64-NEXT:    vmv8r.v v16, v8
 ; CHECK-RV64-NEXT:    vmv.s.x v9, a1
 ; CHECK-RV64-NEXT:    vslideup.vi v8, v9, 25
 ; CHECK-RV64-NEXT:    addi a0, a0, 1
@@ -13269,8 +13053,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_27
 ; CHECK-RV64-NEXT:  .LBB61_553: # %cond.load101
 ; CHECK-RV64-NEXT:    lbu a1, 0(a0)
-; CHECK-RV64-NEXT:    vmv8r.v v16, v8
 ; CHECK-RV64-NEXT:    vsetivli zero, 27, e8, m1, tu, ma
+; CHECK-RV64-NEXT:    vmv8r.v v16, v8
 ; CHECK-RV64-NEXT:    vmv.s.x v9, a1
 ; CHECK-RV64-NEXT:    vslideup.vi v8, v9, 26
 ; CHECK-RV64-NEXT:    addi a0, a0, 1
@@ -13281,8 +13065,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_28
 ; CHECK-RV64-NEXT:  .LBB61_554: # %cond.load105
 ; CHECK-RV64-NEXT:    lbu a1, 0(a0)
-; CHECK-RV64-NEXT:    vmv8r.v v16, v8
 ; CHECK-RV64-NEXT:    vsetivli zero, 28, e8, m1, tu, ma
+; CHECK-RV64-NEXT:    vmv8r.v v16, v8
 ; CHECK-RV64-NEXT:    vmv.s.x v9, a1
 ; CHECK-RV64-NEXT:    vslideup.vi v8, v9, 27
 ; CHECK-RV64-NEXT:    addi a0, a0, 1
@@ -13293,8 +13077,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_29
 ; CHECK-RV64-NEXT:  .LBB61_555: # %cond.load109
 ; CHECK-RV64-NEXT:    lbu a1, 0(a0)
-; CHECK-RV64-NEXT:    vmv8r.v v16, v8
 ; CHECK-RV64-NEXT:    vsetivli zero, 29, e8, m1, tu, ma
+; CHECK-RV64-NEXT:    vmv8r.v v16, v8
 ; CHECK-RV64-NEXT:    vmv.s.x v9, a1
 ; CHECK-RV64-NEXT:    vslideup.vi v8, v9, 28
 ; CHECK-RV64-NEXT:    addi a0, a0, 1
@@ -13305,8 +13089,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_30
 ; CHECK-RV64-NEXT:  .LBB61_556: # %cond.load113
 ; CHECK-RV64-NEXT:    lbu a1, 0(a0)
-; CHECK-RV64-NEXT:    vmv8r.v v16, v8
 ; CHECK-RV64-NEXT:    vsetivli zero, 30, e8, m1, tu, ma
+; CHECK-RV64-NEXT:    vmv8r.v v16, v8
 ; CHECK-RV64-NEXT:    vmv.s.x v9, a1
 ; CHECK-RV64-NEXT:    vslideup.vi v8, v9, 29
 ; CHECK-RV64-NEXT:    addi a0, a0, 1
@@ -13317,8 +13101,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_31
 ; CHECK-RV64-NEXT:  .LBB61_557: # %cond.load117
 ; CHECK-RV64-NEXT:    lbu a1, 0(a0)
-; CHECK-RV64-NEXT:    vmv8r.v v16, v8
 ; CHECK-RV64-NEXT:    vsetivli zero, 31, e8, m1, tu, ma
+; CHECK-RV64-NEXT:    vmv8r.v v16, v8
 ; CHECK-RV64-NEXT:    vmv.s.x v9, a1
 ; CHECK-RV64-NEXT:    vslideup.vi v8, v9, 30
 ; CHECK-RV64-NEXT:    addi a0, a0, 1
@@ -13329,9 +13113,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_32
 ; CHECK-RV64-NEXT:  .LBB61_558: # %cond.load121
 ; CHECK-RV64-NEXT:    lbu a1, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v9, a1
 ; CHECK-RV64-NEXT:    li a1, 32
 ; CHECK-RV64-NEXT:    vsetvli zero, a1, e8, m1, tu, ma
@@ -13344,9 +13127,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_33
 ; CHECK-RV64-NEXT:  .LBB61_559: # %cond.load125
 ; CHECK-RV64-NEXT:    lbu a1, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v9, a1
 ; CHECK-RV64-NEXT:    li a1, 33
 ; CHECK-RV64-NEXT:    li a3, 32
@@ -13360,9 +13142,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_34
 ; CHECK-RV64-NEXT:  .LBB61_560: # %cond.load129
 ; CHECK-RV64-NEXT:    lbu a1, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v9, a1
 ; CHECK-RV64-NEXT:    li a1, 34
 ; CHECK-RV64-NEXT:    li a3, 33
@@ -13376,9 +13157,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_35
 ; CHECK-RV64-NEXT:  .LBB61_561: # %cond.load133
 ; CHECK-RV64-NEXT:    lbu a1, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v9, a1
 ; CHECK-RV64-NEXT:    li a1, 35
 ; CHECK-RV64-NEXT:    li a3, 34
@@ -13392,9 +13172,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_36
 ; CHECK-RV64-NEXT:  .LBB61_562: # %cond.load137
 ; CHECK-RV64-NEXT:    lbu a1, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v9, a1
 ; CHECK-RV64-NEXT:    li a1, 36
 ; CHECK-RV64-NEXT:    li a3, 35
@@ -13408,9 +13187,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_37
 ; CHECK-RV64-NEXT:  .LBB61_563: # %cond.load141
 ; CHECK-RV64-NEXT:    lbu a1, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v9, a1
 ; CHECK-RV64-NEXT:    li a1, 37
 ; CHECK-RV64-NEXT:    li a3, 36
@@ -13424,9 +13202,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_38
 ; CHECK-RV64-NEXT:  .LBB61_564: # %cond.load145
 ; CHECK-RV64-NEXT:    lbu a1, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v9, a1
 ; CHECK-RV64-NEXT:    li a1, 38
 ; CHECK-RV64-NEXT:    li a3, 37
@@ -13440,9 +13217,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_39
 ; CHECK-RV64-NEXT:  .LBB61_565: # %cond.load149
 ; CHECK-RV64-NEXT:    lbu a1, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v9, a1
 ; CHECK-RV64-NEXT:    li a1, 39
 ; CHECK-RV64-NEXT:    li a3, 38
@@ -13456,9 +13232,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_40
 ; CHECK-RV64-NEXT:  .LBB61_566: # %cond.load153
 ; CHECK-RV64-NEXT:    lbu a1, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v9, a1
 ; CHECK-RV64-NEXT:    li a1, 40
 ; CHECK-RV64-NEXT:    li a3, 39
@@ -13472,9 +13247,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_41
 ; CHECK-RV64-NEXT:  .LBB61_567: # %cond.load157
 ; CHECK-RV64-NEXT:    lbu a1, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v9, a1
 ; CHECK-RV64-NEXT:    li a1, 41
 ; CHECK-RV64-NEXT:    li a3, 40
@@ -13488,9 +13262,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_42
 ; CHECK-RV64-NEXT:  .LBB61_568: # %cond.load161
 ; CHECK-RV64-NEXT:    lbu a1, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v9, a1
 ; CHECK-RV64-NEXT:    li a1, 42
 ; CHECK-RV64-NEXT:    li a3, 41
@@ -13504,9 +13277,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_43
 ; CHECK-RV64-NEXT:  .LBB61_569: # %cond.load165
 ; CHECK-RV64-NEXT:    lbu a1, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v9, a1
 ; CHECK-RV64-NEXT:    li a1, 43
 ; CHECK-RV64-NEXT:    li a3, 42
@@ -13520,9 +13292,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_44
 ; CHECK-RV64-NEXT:  .LBB61_570: # %cond.load169
 ; CHECK-RV64-NEXT:    lbu a1, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v9, a1
 ; CHECK-RV64-NEXT:    li a1, 44
 ; CHECK-RV64-NEXT:    li a3, 43
@@ -13536,9 +13307,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_45
 ; CHECK-RV64-NEXT:  .LBB61_571: # %cond.load173
 ; CHECK-RV64-NEXT:    lbu a1, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v9, a1
 ; CHECK-RV64-NEXT:    li a1, 45
 ; CHECK-RV64-NEXT:    li a3, 44
@@ -13552,9 +13322,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_46
 ; CHECK-RV64-NEXT:  .LBB61_572: # %cond.load177
 ; CHECK-RV64-NEXT:    lbu a1, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v9, a1
 ; CHECK-RV64-NEXT:    li a1, 46
 ; CHECK-RV64-NEXT:    li a3, 45
@@ -13568,9 +13337,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_47
 ; CHECK-RV64-NEXT:  .LBB61_573: # %cond.load181
 ; CHECK-RV64-NEXT:    lbu a1, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v9, a1
 ; CHECK-RV64-NEXT:    li a1, 47
 ; CHECK-RV64-NEXT:    li a3, 46
@@ -13584,9 +13352,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_48
 ; CHECK-RV64-NEXT:  .LBB61_574: # %cond.load185
 ; CHECK-RV64-NEXT:    lbu a1, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v9, a1
 ; CHECK-RV64-NEXT:    li a1, 48
 ; CHECK-RV64-NEXT:    li a3, 47
@@ -13600,9 +13367,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_49
 ; CHECK-RV64-NEXT:  .LBB61_575: # %cond.load189
 ; CHECK-RV64-NEXT:    lbu a1, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v9, a1
 ; CHECK-RV64-NEXT:    li a1, 49
 ; CHECK-RV64-NEXT:    li a3, 48
@@ -13616,9 +13382,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_50
 ; CHECK-RV64-NEXT:  .LBB61_576: # %cond.load193
 ; CHECK-RV64-NEXT:    lbu a1, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v9, a1
 ; CHECK-RV64-NEXT:    li a1, 50
 ; CHECK-RV64-NEXT:    li a3, 49
@@ -13632,9 +13397,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_51
 ; CHECK-RV64-NEXT:  .LBB61_577: # %cond.load197
 ; CHECK-RV64-NEXT:    lbu a1, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v9, a1
 ; CHECK-RV64-NEXT:    li a1, 51
 ; CHECK-RV64-NEXT:    li a3, 50
@@ -13648,9 +13412,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_52
 ; CHECK-RV64-NEXT:  .LBB61_578: # %cond.load201
 ; CHECK-RV64-NEXT:    lbu a1, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v9, a1
 ; CHECK-RV64-NEXT:    li a1, 52
 ; CHECK-RV64-NEXT:    li a3, 51
@@ -13664,9 +13427,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_53
 ; CHECK-RV64-NEXT:  .LBB61_579: # %cond.load205
 ; CHECK-RV64-NEXT:    lbu a1, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v9, a1
 ; CHECK-RV64-NEXT:    li a1, 53
 ; CHECK-RV64-NEXT:    li a3, 52
@@ -13680,9 +13442,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_54
 ; CHECK-RV64-NEXT:  .LBB61_580: # %cond.load209
 ; CHECK-RV64-NEXT:    lbu a1, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v9, a1
 ; CHECK-RV64-NEXT:    li a1, 54
 ; CHECK-RV64-NEXT:    li a3, 53
@@ -13696,9 +13457,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_55
 ; CHECK-RV64-NEXT:  .LBB61_581: # %cond.load213
 ; CHECK-RV64-NEXT:    lbu a1, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v9, a1
 ; CHECK-RV64-NEXT:    li a1, 55
 ; CHECK-RV64-NEXT:    li a3, 54
@@ -13712,9 +13472,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_56
 ; CHECK-RV64-NEXT:  .LBB61_582: # %cond.load217
 ; CHECK-RV64-NEXT:    lbu a1, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v9, a1
 ; CHECK-RV64-NEXT:    li a1, 56
 ; CHECK-RV64-NEXT:    li a3, 55
@@ -13728,9 +13487,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_57
 ; CHECK-RV64-NEXT:  .LBB61_583: # %cond.load221
 ; CHECK-RV64-NEXT:    lbu a1, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v9, a1
 ; CHECK-RV64-NEXT:    li a1, 57
 ; CHECK-RV64-NEXT:    li a3, 56
@@ -13744,9 +13502,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_58
 ; CHECK-RV64-NEXT:  .LBB61_584: # %cond.load225
 ; CHECK-RV64-NEXT:    lbu a1, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v9, a1
 ; CHECK-RV64-NEXT:    li a1, 58
 ; CHECK-RV64-NEXT:    li a3, 57
@@ -13760,9 +13517,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_59
 ; CHECK-RV64-NEXT:  .LBB61_585: # %cond.load229
 ; CHECK-RV64-NEXT:    lbu a1, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v9, a1
 ; CHECK-RV64-NEXT:    li a1, 59
 ; CHECK-RV64-NEXT:    li a3, 58
@@ -13776,9 +13532,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_60
 ; CHECK-RV64-NEXT:  .LBB61_586: # %cond.load233
 ; CHECK-RV64-NEXT:    lbu a1, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v9, a1
 ; CHECK-RV64-NEXT:    li a1, 60
 ; CHECK-RV64-NEXT:    li a3, 59
@@ -13792,9 +13547,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_61
 ; CHECK-RV64-NEXT:  .LBB61_587: # %cond.load237
 ; CHECK-RV64-NEXT:    lbu a1, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v9, a1
 ; CHECK-RV64-NEXT:    li a1, 61
 ; CHECK-RV64-NEXT:    li a3, 60
@@ -13824,9 +13578,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_67
 ; CHECK-RV64-NEXT:  .LBB61_589: # %cond.load253
 ; CHECK-RV64-NEXT:    lbu a2, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v10, a2
 ; CHECK-RV64-NEXT:    li a2, 65
 ; CHECK-RV64-NEXT:    li a3, 64
@@ -13840,9 +13593,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_68
 ; CHECK-RV64-NEXT:  .LBB61_590: # %cond.load257
 ; CHECK-RV64-NEXT:    lbu a2, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v10, a2
 ; CHECK-RV64-NEXT:    li a2, 66
 ; CHECK-RV64-NEXT:    li a3, 65
@@ -13856,9 +13608,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_69
 ; CHECK-RV64-NEXT:  .LBB61_591: # %cond.load261
 ; CHECK-RV64-NEXT:    lbu a2, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v10, a2
 ; CHECK-RV64-NEXT:    li a2, 67
 ; CHECK-RV64-NEXT:    li a3, 66
@@ -13872,9 +13623,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_70
 ; CHECK-RV64-NEXT:  .LBB61_592: # %cond.load265
 ; CHECK-RV64-NEXT:    lbu a2, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v10, a2
 ; CHECK-RV64-NEXT:    li a2, 68
 ; CHECK-RV64-NEXT:    li a3, 67
@@ -13888,9 +13638,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_71
 ; CHECK-RV64-NEXT:  .LBB61_593: # %cond.load269
 ; CHECK-RV64-NEXT:    lbu a2, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v10, a2
 ; CHECK-RV64-NEXT:    li a2, 69
 ; CHECK-RV64-NEXT:    li a3, 68
@@ -13904,9 +13653,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_72
 ; CHECK-RV64-NEXT:  .LBB61_594: # %cond.load273
 ; CHECK-RV64-NEXT:    lbu a2, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v10, a2
 ; CHECK-RV64-NEXT:    li a2, 70
 ; CHECK-RV64-NEXT:    li a3, 69
@@ -13920,9 +13668,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_73
 ; CHECK-RV64-NEXT:  .LBB61_595: # %cond.load277
 ; CHECK-RV64-NEXT:    lbu a2, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v10, a2
 ; CHECK-RV64-NEXT:    li a2, 71
 ; CHECK-RV64-NEXT:    li a3, 70
@@ -13936,9 +13683,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_74
 ; CHECK-RV64-NEXT:  .LBB61_596: # %cond.load281
 ; CHECK-RV64-NEXT:    lbu a2, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v10, a2
 ; CHECK-RV64-NEXT:    li a2, 72
 ; CHECK-RV64-NEXT:    li a3, 71
@@ -13952,9 +13698,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_75
 ; CHECK-RV64-NEXT:  .LBB61_597: # %cond.load285
 ; CHECK-RV64-NEXT:    lbu a2, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v10, a2
 ; CHECK-RV64-NEXT:    li a2, 73
 ; CHECK-RV64-NEXT:    li a3, 72
@@ -13968,9 +13713,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_76
 ; CHECK-RV64-NEXT:  .LBB61_598: # %cond.load289
 ; CHECK-RV64-NEXT:    lbu a2, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v10, a2
 ; CHECK-RV64-NEXT:    li a2, 74
 ; CHECK-RV64-NEXT:    li a3, 73
@@ -13984,9 +13728,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_77
 ; CHECK-RV64-NEXT:  .LBB61_599: # %cond.load293
 ; CHECK-RV64-NEXT:    lbu a2, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v10, a2
 ; CHECK-RV64-NEXT:    li a2, 75
 ; CHECK-RV64-NEXT:    li a3, 74
@@ -14000,9 +13743,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_78
 ; CHECK-RV64-NEXT:  .LBB61_600: # %cond.load297
 ; CHECK-RV64-NEXT:    lbu a2, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v10, a2
 ; CHECK-RV64-NEXT:    li a2, 76
 ; CHECK-RV64-NEXT:    li a3, 75
@@ -14016,9 +13758,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_79
 ; CHECK-RV64-NEXT:  .LBB61_601: # %cond.load301
 ; CHECK-RV64-NEXT:    lbu a2, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v10, a2
 ; CHECK-RV64-NEXT:    li a2, 77
 ; CHECK-RV64-NEXT:    li a3, 76
@@ -14032,9 +13773,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_80
 ; CHECK-RV64-NEXT:  .LBB61_602: # %cond.load305
 ; CHECK-RV64-NEXT:    lbu a2, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v10, a2
 ; CHECK-RV64-NEXT:    li a2, 78
 ; CHECK-RV64-NEXT:    li a3, 77
@@ -14048,9 +13788,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_81
 ; CHECK-RV64-NEXT:  .LBB61_603: # %cond.load309
 ; CHECK-RV64-NEXT:    lbu a2, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v10, a2
 ; CHECK-RV64-NEXT:    li a2, 79
 ; CHECK-RV64-NEXT:    li a3, 78
@@ -14064,9 +13803,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_82
 ; CHECK-RV64-NEXT:  .LBB61_604: # %cond.load313
 ; CHECK-RV64-NEXT:    lbu a2, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v10, a2
 ; CHECK-RV64-NEXT:    li a2, 80
 ; CHECK-RV64-NEXT:    li a3, 79
@@ -14080,9 +13818,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_83
 ; CHECK-RV64-NEXT:  .LBB61_605: # %cond.load317
 ; CHECK-RV64-NEXT:    lbu a2, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v10, a2
 ; CHECK-RV64-NEXT:    li a2, 81
 ; CHECK-RV64-NEXT:    li a3, 80
@@ -14096,9 +13833,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_84
 ; CHECK-RV64-NEXT:  .LBB61_606: # %cond.load321
 ; CHECK-RV64-NEXT:    lbu a2, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v10, a2
 ; CHECK-RV64-NEXT:    li a2, 82
 ; CHECK-RV64-NEXT:    li a3, 81
@@ -14112,9 +13848,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_85
 ; CHECK-RV64-NEXT:  .LBB61_607: # %cond.load325
 ; CHECK-RV64-NEXT:    lbu a2, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v10, a2
 ; CHECK-RV64-NEXT:    li a2, 83
 ; CHECK-RV64-NEXT:    li a3, 82
@@ -14128,9 +13863,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_86
 ; CHECK-RV64-NEXT:  .LBB61_608: # %cond.load329
 ; CHECK-RV64-NEXT:    lbu a2, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v10, a2
 ; CHECK-RV64-NEXT:    li a2, 84
 ; CHECK-RV64-NEXT:    li a3, 83
@@ -14144,9 +13878,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_87
 ; CHECK-RV64-NEXT:  .LBB61_609: # %cond.load333
 ; CHECK-RV64-NEXT:    lbu a2, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v10, a2
 ; CHECK-RV64-NEXT:    li a2, 85
 ; CHECK-RV64-NEXT:    li a3, 84
@@ -14160,9 +13893,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_88
 ; CHECK-RV64-NEXT:  .LBB61_610: # %cond.load337
 ; CHECK-RV64-NEXT:    lbu a2, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v10, a2
 ; CHECK-RV64-NEXT:    li a2, 86
 ; CHECK-RV64-NEXT:    li a3, 85
@@ -14176,9 +13908,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_89
 ; CHECK-RV64-NEXT:  .LBB61_611: # %cond.load341
 ; CHECK-RV64-NEXT:    lbu a2, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v10, a2
 ; CHECK-RV64-NEXT:    li a2, 87
 ; CHECK-RV64-NEXT:    li a3, 86
@@ -14192,9 +13923,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_90
 ; CHECK-RV64-NEXT:  .LBB61_612: # %cond.load345
 ; CHECK-RV64-NEXT:    lbu a2, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v10, a2
 ; CHECK-RV64-NEXT:    li a2, 88
 ; CHECK-RV64-NEXT:    li a3, 87
@@ -14208,9 +13938,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_91
 ; CHECK-RV64-NEXT:  .LBB61_613: # %cond.load349
 ; CHECK-RV64-NEXT:    lbu a2, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v10, a2
 ; CHECK-RV64-NEXT:    li a2, 89
 ; CHECK-RV64-NEXT:    li a3, 88
@@ -14224,9 +13953,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_92
 ; CHECK-RV64-NEXT:  .LBB61_614: # %cond.load353
 ; CHECK-RV64-NEXT:    lbu a2, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v10, a2
 ; CHECK-RV64-NEXT:    li a2, 90
 ; CHECK-RV64-NEXT:    li a3, 89
@@ -14240,9 +13968,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_93
 ; CHECK-RV64-NEXT:  .LBB61_615: # %cond.load357
 ; CHECK-RV64-NEXT:    lbu a2, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v10, a2
 ; CHECK-RV64-NEXT:    li a2, 91
 ; CHECK-RV64-NEXT:    li a3, 90
@@ -14256,9 +13983,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_94
 ; CHECK-RV64-NEXT:  .LBB61_616: # %cond.load361
 ; CHECK-RV64-NEXT:    lbu a2, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v10, a2
 ; CHECK-RV64-NEXT:    li a2, 92
 ; CHECK-RV64-NEXT:    li a3, 91
@@ -14272,9 +13998,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_95
 ; CHECK-RV64-NEXT:  .LBB61_617: # %cond.load365
 ; CHECK-RV64-NEXT:    lbu a2, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v10, a2
 ; CHECK-RV64-NEXT:    li a2, 93
 ; CHECK-RV64-NEXT:    li a3, 92
@@ -14288,9 +14013,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_96
 ; CHECK-RV64-NEXT:  .LBB61_618: # %cond.load369
 ; CHECK-RV64-NEXT:    lbu a2, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v10, a2
 ; CHECK-RV64-NEXT:    li a2, 94
 ; CHECK-RV64-NEXT:    li a3, 93
@@ -14304,9 +14028,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_97
 ; CHECK-RV64-NEXT:  .LBB61_619: # %cond.load373
 ; CHECK-RV64-NEXT:    lbu a2, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v10, a2
 ; CHECK-RV64-NEXT:    li a2, 95
 ; CHECK-RV64-NEXT:    li a3, 94
@@ -14320,9 +14043,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_98
 ; CHECK-RV64-NEXT:  .LBB61_620: # %cond.load377
 ; CHECK-RV64-NEXT:    lbu a2, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v10, a2
 ; CHECK-RV64-NEXT:    li a2, 96
 ; CHECK-RV64-NEXT:    li a3, 95
@@ -14336,9 +14058,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_99
 ; CHECK-RV64-NEXT:  .LBB61_621: # %cond.load381
 ; CHECK-RV64-NEXT:    lbu a2, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v10, a2
 ; CHECK-RV64-NEXT:    li a2, 97
 ; CHECK-RV64-NEXT:    li a3, 96
@@ -14352,9 +14073,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_100
 ; CHECK-RV64-NEXT:  .LBB61_622: # %cond.load385
 ; CHECK-RV64-NEXT:    lbu a2, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v10, a2
 ; CHECK-RV64-NEXT:    li a2, 98
 ; CHECK-RV64-NEXT:    li a3, 97
@@ -14368,9 +14088,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_101
 ; CHECK-RV64-NEXT:  .LBB61_623: # %cond.load389
 ; CHECK-RV64-NEXT:    lbu a2, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v10, a2
 ; CHECK-RV64-NEXT:    li a2, 99
 ; CHECK-RV64-NEXT:    li a3, 98
@@ -14384,9 +14103,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_102
 ; CHECK-RV64-NEXT:  .LBB61_624: # %cond.load393
 ; CHECK-RV64-NEXT:    lbu a2, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v10, a2
 ; CHECK-RV64-NEXT:    li a2, 100
 ; CHECK-RV64-NEXT:    li a3, 99
@@ -14400,9 +14118,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_103
 ; CHECK-RV64-NEXT:  .LBB61_625: # %cond.load397
 ; CHECK-RV64-NEXT:    lbu a2, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v10, a2
 ; CHECK-RV64-NEXT:    li a2, 101
 ; CHECK-RV64-NEXT:    li a3, 100
@@ -14416,9 +14133,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_104
 ; CHECK-RV64-NEXT:  .LBB61_626: # %cond.load401
 ; CHECK-RV64-NEXT:    lbu a2, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v10, a2
 ; CHECK-RV64-NEXT:    li a2, 102
 ; CHECK-RV64-NEXT:    li a3, 101
@@ -14432,9 +14148,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_105
 ; CHECK-RV64-NEXT:  .LBB61_627: # %cond.load405
 ; CHECK-RV64-NEXT:    lbu a2, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v10, a2
 ; CHECK-RV64-NEXT:    li a2, 103
 ; CHECK-RV64-NEXT:    li a3, 102
@@ -14448,9 +14163,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_106
 ; CHECK-RV64-NEXT:  .LBB61_628: # %cond.load409
 ; CHECK-RV64-NEXT:    lbu a2, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v10, a2
 ; CHECK-RV64-NEXT:    li a2, 104
 ; CHECK-RV64-NEXT:    li a3, 103
@@ -14464,9 +14178,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_107
 ; CHECK-RV64-NEXT:  .LBB61_629: # %cond.load413
 ; CHECK-RV64-NEXT:    lbu a2, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v10, a2
 ; CHECK-RV64-NEXT:    li a2, 105
 ; CHECK-RV64-NEXT:    li a3, 104
@@ -14480,9 +14193,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_108
 ; CHECK-RV64-NEXT:  .LBB61_630: # %cond.load417
 ; CHECK-RV64-NEXT:    lbu a2, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v10, a2
 ; CHECK-RV64-NEXT:    li a2, 106
 ; CHECK-RV64-NEXT:    li a3, 105
@@ -14496,9 +14208,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_109
 ; CHECK-RV64-NEXT:  .LBB61_631: # %cond.load421
 ; CHECK-RV64-NEXT:    lbu a2, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v10, a2
 ; CHECK-RV64-NEXT:    li a2, 107
 ; CHECK-RV64-NEXT:    li a3, 106
@@ -14512,9 +14223,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_110
 ; CHECK-RV64-NEXT:  .LBB61_632: # %cond.load425
 ; CHECK-RV64-NEXT:    lbu a2, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v10, a2
 ; CHECK-RV64-NEXT:    li a2, 108
 ; CHECK-RV64-NEXT:    li a3, 107
@@ -14528,9 +14238,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_111
 ; CHECK-RV64-NEXT:  .LBB61_633: # %cond.load429
 ; CHECK-RV64-NEXT:    lbu a2, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v10, a2
 ; CHECK-RV64-NEXT:    li a2, 109
 ; CHECK-RV64-NEXT:    li a3, 108
@@ -14544,9 +14253,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_112
 ; CHECK-RV64-NEXT:  .LBB61_634: # %cond.load433
 ; CHECK-RV64-NEXT:    lbu a2, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v10, a2
 ; CHECK-RV64-NEXT:    li a2, 110
 ; CHECK-RV64-NEXT:    li a3, 109
@@ -14560,9 +14268,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_113
 ; CHECK-RV64-NEXT:  .LBB61_635: # %cond.load437
 ; CHECK-RV64-NEXT:    lbu a2, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v10, a2
 ; CHECK-RV64-NEXT:    li a2, 111
 ; CHECK-RV64-NEXT:    li a3, 110
@@ -14576,9 +14283,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_114
 ; CHECK-RV64-NEXT:  .LBB61_636: # %cond.load441
 ; CHECK-RV64-NEXT:    lbu a2, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v10, a2
 ; CHECK-RV64-NEXT:    li a2, 112
 ; CHECK-RV64-NEXT:    li a3, 111
@@ -14592,9 +14298,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_115
 ; CHECK-RV64-NEXT:  .LBB61_637: # %cond.load445
 ; CHECK-RV64-NEXT:    lbu a2, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v10, a2
 ; CHECK-RV64-NEXT:    li a2, 113
 ; CHECK-RV64-NEXT:    li a3, 112
@@ -14608,9 +14313,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_116
 ; CHECK-RV64-NEXT:  .LBB61_638: # %cond.load449
 ; CHECK-RV64-NEXT:    lbu a2, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v10, a2
 ; CHECK-RV64-NEXT:    li a2, 114
 ; CHECK-RV64-NEXT:    li a3, 113
@@ -14624,9 +14328,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_117
 ; CHECK-RV64-NEXT:  .LBB61_639: # %cond.load453
 ; CHECK-RV64-NEXT:    lbu a2, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v10, a2
 ; CHECK-RV64-NEXT:    li a2, 115
 ; CHECK-RV64-NEXT:    li a3, 114
@@ -14640,9 +14343,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_118
 ; CHECK-RV64-NEXT:  .LBB61_640: # %cond.load457
 ; CHECK-RV64-NEXT:    lbu a2, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v10, a2
 ; CHECK-RV64-NEXT:    li a2, 116
 ; CHECK-RV64-NEXT:    li a3, 115
@@ -14656,9 +14358,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_119
 ; CHECK-RV64-NEXT:  .LBB61_641: # %cond.load461
 ; CHECK-RV64-NEXT:    lbu a2, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v10, a2
 ; CHECK-RV64-NEXT:    li a2, 117
 ; CHECK-RV64-NEXT:    li a3, 116
@@ -14672,9 +14373,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_120
 ; CHECK-RV64-NEXT:  .LBB61_642: # %cond.load465
 ; CHECK-RV64-NEXT:    lbu a2, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v10, a2
 ; CHECK-RV64-NEXT:    li a2, 118
 ; CHECK-RV64-NEXT:    li a3, 117
@@ -14688,9 +14388,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_121
 ; CHECK-RV64-NEXT:  .LBB61_643: # %cond.load469
 ; CHECK-RV64-NEXT:    lbu a2, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v10, a2
 ; CHECK-RV64-NEXT:    li a2, 119
 ; CHECK-RV64-NEXT:    li a3, 118
@@ -14704,9 +14403,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_122
 ; CHECK-RV64-NEXT:  .LBB61_644: # %cond.load473
 ; CHECK-RV64-NEXT:    lbu a2, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v10, a2
 ; CHECK-RV64-NEXT:    li a2, 120
 ; CHECK-RV64-NEXT:    li a3, 119
@@ -14720,9 +14418,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_123
 ; CHECK-RV64-NEXT:  .LBB61_645: # %cond.load477
 ; CHECK-RV64-NEXT:    lbu a2, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v10, a2
 ; CHECK-RV64-NEXT:    li a2, 121
 ; CHECK-RV64-NEXT:    li a3, 120
@@ -14736,9 +14433,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_124
 ; CHECK-RV64-NEXT:  .LBB61_646: # %cond.load481
 ; CHECK-RV64-NEXT:    lbu a2, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v10, a2
 ; CHECK-RV64-NEXT:    li a2, 122
 ; CHECK-RV64-NEXT:    li a3, 121
@@ -14752,9 +14448,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_125
 ; CHECK-RV64-NEXT:  .LBB61_647: # %cond.load485
 ; CHECK-RV64-NEXT:    lbu a2, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v10, a2
 ; CHECK-RV64-NEXT:    li a2, 123
 ; CHECK-RV64-NEXT:    li a3, 122
@@ -14768,9 +14463,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_126
 ; CHECK-RV64-NEXT:  .LBB61_648: # %cond.load489
 ; CHECK-RV64-NEXT:    lbu a2, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v10, a2
 ; CHECK-RV64-NEXT:    li a2, 124
 ; CHECK-RV64-NEXT:    li a3, 123
@@ -14784,9 +14478,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_127
 ; CHECK-RV64-NEXT:  .LBB61_649: # %cond.load493
 ; CHECK-RV64-NEXT:    lbu a2, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v10, a2
 ; CHECK-RV64-NEXT:    li a2, 125
 ; CHECK-RV64-NEXT:    li a3, 124
@@ -14816,9 +14509,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_133
 ; CHECK-RV64-NEXT:  .LBB61_651: # %cond.load509
 ; CHECK-RV64-NEXT:    lbu a1, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v12, a1
 ; CHECK-RV64-NEXT:    li a1, 129
 ; CHECK-RV64-NEXT:    li a3, 128
@@ -14832,9 +14524,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_134
 ; CHECK-RV64-NEXT:  .LBB61_652: # %cond.load513
 ; CHECK-RV64-NEXT:    lbu a1, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v12, a1
 ; CHECK-RV64-NEXT:    li a1, 130
 ; CHECK-RV64-NEXT:    li a3, 129
@@ -14848,9 +14539,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_135
 ; CHECK-RV64-NEXT:  .LBB61_653: # %cond.load517
 ; CHECK-RV64-NEXT:    lbu a1, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v12, a1
 ; CHECK-RV64-NEXT:    li a1, 131
 ; CHECK-RV64-NEXT:    li a3, 130
@@ -14864,9 +14554,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_136
 ; CHECK-RV64-NEXT:  .LBB61_654: # %cond.load521
 ; CHECK-RV64-NEXT:    lbu a1, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v12, a1
 ; CHECK-RV64-NEXT:    li a1, 132
 ; CHECK-RV64-NEXT:    li a3, 131
@@ -14880,9 +14569,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_137
 ; CHECK-RV64-NEXT:  .LBB61_655: # %cond.load525
 ; CHECK-RV64-NEXT:    lbu a1, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v12, a1
 ; CHECK-RV64-NEXT:    li a1, 133
 ; CHECK-RV64-NEXT:    li a3, 132
@@ -14896,9 +14584,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_138
 ; CHECK-RV64-NEXT:  .LBB61_656: # %cond.load529
 ; CHECK-RV64-NEXT:    lbu a1, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v12, a1
 ; CHECK-RV64-NEXT:    li a1, 134
 ; CHECK-RV64-NEXT:    li a3, 133
@@ -14912,9 +14599,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_139
 ; CHECK-RV64-NEXT:  .LBB61_657: # %cond.load533
 ; CHECK-RV64-NEXT:    lbu a1, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v12, a1
 ; CHECK-RV64-NEXT:    li a1, 135
 ; CHECK-RV64-NEXT:    li a3, 134
@@ -14928,9 +14614,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_140
 ; CHECK-RV64-NEXT:  .LBB61_658: # %cond.load537
 ; CHECK-RV64-NEXT:    lbu a1, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v12, a1
 ; CHECK-RV64-NEXT:    li a1, 136
 ; CHECK-RV64-NEXT:    li a3, 135
@@ -14944,9 +14629,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_141
 ; CHECK-RV64-NEXT:  .LBB61_659: # %cond.load541
 ; CHECK-RV64-NEXT:    lbu a1, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v12, a1
 ; CHECK-RV64-NEXT:    li a1, 137
 ; CHECK-RV64-NEXT:    li a3, 136
@@ -14960,9 +14644,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_142
 ; CHECK-RV64-NEXT:  .LBB61_660: # %cond.load545
 ; CHECK-RV64-NEXT:    lbu a1, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v12, a1
 ; CHECK-RV64-NEXT:    li a1, 138
 ; CHECK-RV64-NEXT:    li a3, 137
@@ -14976,9 +14659,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_143
 ; CHECK-RV64-NEXT:  .LBB61_661: # %cond.load549
 ; CHECK-RV64-NEXT:    lbu a1, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v12, a1
 ; CHECK-RV64-NEXT:    li a1, 139
 ; CHECK-RV64-NEXT:    li a3, 138
@@ -14992,9 +14674,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_144
 ; CHECK-RV64-NEXT:  .LBB61_662: # %cond.load553
 ; CHECK-RV64-NEXT:    lbu a1, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v12, a1
 ; CHECK-RV64-NEXT:    li a1, 140
 ; CHECK-RV64-NEXT:    li a3, 139
@@ -15008,9 +14689,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_145
 ; CHECK-RV64-NEXT:  .LBB61_663: # %cond.load557
 ; CHECK-RV64-NEXT:    lbu a1, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v12, a1
 ; CHECK-RV64-NEXT:    li a1, 141
 ; CHECK-RV64-NEXT:    li a3, 140
@@ -15024,9 +14704,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_146
 ; CHECK-RV64-NEXT:  .LBB61_664: # %cond.load561
 ; CHECK-RV64-NEXT:    lbu a1, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v12, a1
 ; CHECK-RV64-NEXT:    li a1, 142
 ; CHECK-RV64-NEXT:    li a3, 141
@@ -15040,9 +14719,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_147
 ; CHECK-RV64-NEXT:  .LBB61_665: # %cond.load565
 ; CHECK-RV64-NEXT:    lbu a1, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v12, a1
 ; CHECK-RV64-NEXT:    li a1, 143
 ; CHECK-RV64-NEXT:    li a3, 142
@@ -15056,9 +14734,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_148
 ; CHECK-RV64-NEXT:  .LBB61_666: # %cond.load569
 ; CHECK-RV64-NEXT:    lbu a1, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v12, a1
 ; CHECK-RV64-NEXT:    li a1, 144
 ; CHECK-RV64-NEXT:    li a3, 143
@@ -15072,9 +14749,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_149
 ; CHECK-RV64-NEXT:  .LBB61_667: # %cond.load573
 ; CHECK-RV64-NEXT:    lbu a1, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v12, a1
 ; CHECK-RV64-NEXT:    li a1, 145
 ; CHECK-RV64-NEXT:    li a3, 144
@@ -15088,9 +14764,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_150
 ; CHECK-RV64-NEXT:  .LBB61_668: # %cond.load577
 ; CHECK-RV64-NEXT:    lbu a1, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v12, a1
 ; CHECK-RV64-NEXT:    li a1, 146
 ; CHECK-RV64-NEXT:    li a3, 145
@@ -15104,9 +14779,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_151
 ; CHECK-RV64-NEXT:  .LBB61_669: # %cond.load581
 ; CHECK-RV64-NEXT:    lbu a1, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v12, a1
 ; CHECK-RV64-NEXT:    li a1, 147
 ; CHECK-RV64-NEXT:    li a3, 146
@@ -15120,9 +14794,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_152
 ; CHECK-RV64-NEXT:  .LBB61_670: # %cond.load585
 ; CHECK-RV64-NEXT:    lbu a1, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v12, a1
 ; CHECK-RV64-NEXT:    li a1, 148
 ; CHECK-RV64-NEXT:    li a3, 147
@@ -15136,9 +14809,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_153
 ; CHECK-RV64-NEXT:  .LBB61_671: # %cond.load589
 ; CHECK-RV64-NEXT:    lbu a1, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v12, a1
 ; CHECK-RV64-NEXT:    li a1, 149
 ; CHECK-RV64-NEXT:    li a3, 148
@@ -15152,9 +14824,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_154
 ; CHECK-RV64-NEXT:  .LBB61_672: # %cond.load593
 ; CHECK-RV64-NEXT:    lbu a1, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v12, a1
 ; CHECK-RV64-NEXT:    li a1, 150
 ; CHECK-RV64-NEXT:    li a3, 149
@@ -15168,9 +14839,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_155
 ; CHECK-RV64-NEXT:  .LBB61_673: # %cond.load597
 ; CHECK-RV64-NEXT:    lbu a1, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v12, a1
 ; CHECK-RV64-NEXT:    li a1, 151
 ; CHECK-RV64-NEXT:    li a3, 150
@@ -15184,9 +14854,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_156
 ; CHECK-RV64-NEXT:  .LBB61_674: # %cond.load601
 ; CHECK-RV64-NEXT:    lbu a1, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v12, a1
 ; CHECK-RV64-NEXT:    li a1, 152
 ; CHECK-RV64-NEXT:    li a3, 151
@@ -15200,9 +14869,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_157
 ; CHECK-RV64-NEXT:  .LBB61_675: # %cond.load605
 ; CHECK-RV64-NEXT:    lbu a1, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v12, a1
 ; CHECK-RV64-NEXT:    li a1, 153
 ; CHECK-RV64-NEXT:    li a3, 152
@@ -15216,9 +14884,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_158
 ; CHECK-RV64-NEXT:  .LBB61_676: # %cond.load609
 ; CHECK-RV64-NEXT:    lbu a1, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v12, a1
 ; CHECK-RV64-NEXT:    li a1, 154
 ; CHECK-RV64-NEXT:    li a3, 153
@@ -15232,9 +14899,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_159
 ; CHECK-RV64-NEXT:  .LBB61_677: # %cond.load613
 ; CHECK-RV64-NEXT:    lbu a1, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v12, a1
 ; CHECK-RV64-NEXT:    li a1, 155
 ; CHECK-RV64-NEXT:    li a3, 154
@@ -15248,9 +14914,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_160
 ; CHECK-RV64-NEXT:  .LBB61_678: # %cond.load617
 ; CHECK-RV64-NEXT:    lbu a1, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v12, a1
 ; CHECK-RV64-NEXT:    li a1, 156
 ; CHECK-RV64-NEXT:    li a3, 155
@@ -15264,9 +14929,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_161
 ; CHECK-RV64-NEXT:  .LBB61_679: # %cond.load621
 ; CHECK-RV64-NEXT:    lbu a1, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v12, a1
 ; CHECK-RV64-NEXT:    li a1, 157
 ; CHECK-RV64-NEXT:    li a3, 156
@@ -15280,9 +14944,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_162
 ; CHECK-RV64-NEXT:  .LBB61_680: # %cond.load625
 ; CHECK-RV64-NEXT:    lbu a1, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v12, a1
 ; CHECK-RV64-NEXT:    li a1, 158
 ; CHECK-RV64-NEXT:    li a3, 157
@@ -15296,9 +14959,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_163
 ; CHECK-RV64-NEXT:  .LBB61_681: # %cond.load629
 ; CHECK-RV64-NEXT:    lbu a1, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v12, a1
 ; CHECK-RV64-NEXT:    li a1, 159
 ; CHECK-RV64-NEXT:    li a3, 158
@@ -15312,9 +14974,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_164
 ; CHECK-RV64-NEXT:  .LBB61_682: # %cond.load633
 ; CHECK-RV64-NEXT:    lbu a1, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v12, a1
 ; CHECK-RV64-NEXT:    li a1, 160
 ; CHECK-RV64-NEXT:    li a3, 159
@@ -15328,9 +14989,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_165
 ; CHECK-RV64-NEXT:  .LBB61_683: # %cond.load637
 ; CHECK-RV64-NEXT:    lbu a1, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v12, a1
 ; CHECK-RV64-NEXT:    li a1, 161
 ; CHECK-RV64-NEXT:    li a3, 160
@@ -15344,9 +15004,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_166
 ; CHECK-RV64-NEXT:  .LBB61_684: # %cond.load641
 ; CHECK-RV64-NEXT:    lbu a1, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v12, a1
 ; CHECK-RV64-NEXT:    li a1, 162
 ; CHECK-RV64-NEXT:    li a3, 161
@@ -15360,9 +15019,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_167
 ; CHECK-RV64-NEXT:  .LBB61_685: # %cond.load645
 ; CHECK-RV64-NEXT:    lbu a1, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v12, a1
 ; CHECK-RV64-NEXT:    li a1, 163
 ; CHECK-RV64-NEXT:    li a3, 162
@@ -15376,9 +15034,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_168
 ; CHECK-RV64-NEXT:  .LBB61_686: # %cond.load649
 ; CHECK-RV64-NEXT:    lbu a1, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v12, a1
 ; CHECK-RV64-NEXT:    li a1, 164
 ; CHECK-RV64-NEXT:    li a3, 163
@@ -15392,9 +15049,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_169
 ; CHECK-RV64-NEXT:  .LBB61_687: # %cond.load653
 ; CHECK-RV64-NEXT:    lbu a1, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v12, a1
 ; CHECK-RV64-NEXT:    li a1, 165
 ; CHECK-RV64-NEXT:    li a3, 164
@@ -15408,9 +15064,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_170
 ; CHECK-RV64-NEXT:  .LBB61_688: # %cond.load657
 ; CHECK-RV64-NEXT:    lbu a1, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v12, a1
 ; CHECK-RV64-NEXT:    li a1, 166
 ; CHECK-RV64-NEXT:    li a3, 165
@@ -15424,9 +15079,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_171
 ; CHECK-RV64-NEXT:  .LBB61_689: # %cond.load661
 ; CHECK-RV64-NEXT:    lbu a1, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v12, a1
 ; CHECK-RV64-NEXT:    li a1, 167
 ; CHECK-RV64-NEXT:    li a3, 166
@@ -15440,9 +15094,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_172
 ; CHECK-RV64-NEXT:  .LBB61_690: # %cond.load665
 ; CHECK-RV64-NEXT:    lbu a1, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v12, a1
 ; CHECK-RV64-NEXT:    li a1, 168
 ; CHECK-RV64-NEXT:    li a3, 167
@@ -15456,9 +15109,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_173
 ; CHECK-RV64-NEXT:  .LBB61_691: # %cond.load669
 ; CHECK-RV64-NEXT:    lbu a1, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v12, a1
 ; CHECK-RV64-NEXT:    li a1, 169
 ; CHECK-RV64-NEXT:    li a3, 168
@@ -15472,9 +15124,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_174
 ; CHECK-RV64-NEXT:  .LBB61_692: # %cond.load673
 ; CHECK-RV64-NEXT:    lbu a1, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v12, a1
 ; CHECK-RV64-NEXT:    li a1, 170
 ; CHECK-RV64-NEXT:    li a3, 169
@@ -15488,9 +15139,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_175
 ; CHECK-RV64-NEXT:  .LBB61_693: # %cond.load677
 ; CHECK-RV64-NEXT:    lbu a1, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v12, a1
 ; CHECK-RV64-NEXT:    li a1, 171
 ; CHECK-RV64-NEXT:    li a3, 170
@@ -15504,9 +15154,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_176
 ; CHECK-RV64-NEXT:  .LBB61_694: # %cond.load681
 ; CHECK-RV64-NEXT:    lbu a1, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v12, a1
 ; CHECK-RV64-NEXT:    li a1, 172
 ; CHECK-RV64-NEXT:    li a3, 171
@@ -15520,9 +15169,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_177
 ; CHECK-RV64-NEXT:  .LBB61_695: # %cond.load685
 ; CHECK-RV64-NEXT:    lbu a1, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v12, a1
 ; CHECK-RV64-NEXT:    li a1, 173
 ; CHECK-RV64-NEXT:    li a3, 172
@@ -15536,9 +15184,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_178
 ; CHECK-RV64-NEXT:  .LBB61_696: # %cond.load689
 ; CHECK-RV64-NEXT:    lbu a1, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v12, a1
 ; CHECK-RV64-NEXT:    li a1, 174
 ; CHECK-RV64-NEXT:    li a3, 173
@@ -15552,9 +15199,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_179
 ; CHECK-RV64-NEXT:  .LBB61_697: # %cond.load693
 ; CHECK-RV64-NEXT:    lbu a1, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v12, a1
 ; CHECK-RV64-NEXT:    li a1, 175
 ; CHECK-RV64-NEXT:    li a3, 174
@@ -15568,9 +15214,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_180
 ; CHECK-RV64-NEXT:  .LBB61_698: # %cond.load697
 ; CHECK-RV64-NEXT:    lbu a1, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v12, a1
 ; CHECK-RV64-NEXT:    li a1, 176
 ; CHECK-RV64-NEXT:    li a3, 175
@@ -15584,9 +15229,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_181
 ; CHECK-RV64-NEXT:  .LBB61_699: # %cond.load701
 ; CHECK-RV64-NEXT:    lbu a1, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v12, a1
 ; CHECK-RV64-NEXT:    li a1, 177
 ; CHECK-RV64-NEXT:    li a3, 176
@@ -15600,9 +15244,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_182
 ; CHECK-RV64-NEXT:  .LBB61_700: # %cond.load705
 ; CHECK-RV64-NEXT:    lbu a1, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v12, a1
 ; CHECK-RV64-NEXT:    li a1, 178
 ; CHECK-RV64-NEXT:    li a3, 177
@@ -15616,9 +15259,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_183
 ; CHECK-RV64-NEXT:  .LBB61_701: # %cond.load709
 ; CHECK-RV64-NEXT:    lbu a1, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v12, a1
 ; CHECK-RV64-NEXT:    li a1, 179
 ; CHECK-RV64-NEXT:    li a3, 178
@@ -15632,9 +15274,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_184
 ; CHECK-RV64-NEXT:  .LBB61_702: # %cond.load713
 ; CHECK-RV64-NEXT:    lbu a1, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v12, a1
 ; CHECK-RV64-NEXT:    li a1, 180
 ; CHECK-RV64-NEXT:    li a3, 179
@@ -15648,9 +15289,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_185
 ; CHECK-RV64-NEXT:  .LBB61_703: # %cond.load717
 ; CHECK-RV64-NEXT:    lbu a1, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v12, a1
 ; CHECK-RV64-NEXT:    li a1, 181
 ; CHECK-RV64-NEXT:    li a3, 180
@@ -15664,9 +15304,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_186
 ; CHECK-RV64-NEXT:  .LBB61_704: # %cond.load721
 ; CHECK-RV64-NEXT:    lbu a1, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v12, a1
 ; CHECK-RV64-NEXT:    li a1, 182
 ; CHECK-RV64-NEXT:    li a3, 181
@@ -15680,9 +15319,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_187
 ; CHECK-RV64-NEXT:  .LBB61_705: # %cond.load725
 ; CHECK-RV64-NEXT:    lbu a1, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v12, a1
 ; CHECK-RV64-NEXT:    li a1, 183
 ; CHECK-RV64-NEXT:    li a3, 182
@@ -15696,9 +15334,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_188
 ; CHECK-RV64-NEXT:  .LBB61_706: # %cond.load729
 ; CHECK-RV64-NEXT:    lbu a1, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v12, a1
 ; CHECK-RV64-NEXT:    li a1, 184
 ; CHECK-RV64-NEXT:    li a3, 183
@@ -15712,9 +15349,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_189
 ; CHECK-RV64-NEXT:  .LBB61_707: # %cond.load733
 ; CHECK-RV64-NEXT:    lbu a1, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v12, a1
 ; CHECK-RV64-NEXT:    li a1, 185
 ; CHECK-RV64-NEXT:    li a3, 184
@@ -15728,9 +15364,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_190
 ; CHECK-RV64-NEXT:  .LBB61_708: # %cond.load737
 ; CHECK-RV64-NEXT:    lbu a1, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v12, a1
 ; CHECK-RV64-NEXT:    li a1, 186
 ; CHECK-RV64-NEXT:    li a3, 185
@@ -15744,9 +15379,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_191
 ; CHECK-RV64-NEXT:  .LBB61_709: # %cond.load741
 ; CHECK-RV64-NEXT:    lbu a1, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v12, a1
 ; CHECK-RV64-NEXT:    li a1, 187
 ; CHECK-RV64-NEXT:    li a3, 186
@@ -15760,9 +15394,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_192
 ; CHECK-RV64-NEXT:  .LBB61_710: # %cond.load745
 ; CHECK-RV64-NEXT:    lbu a1, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v12, a1
 ; CHECK-RV64-NEXT:    li a1, 188
 ; CHECK-RV64-NEXT:    li a3, 187
@@ -15776,9 +15409,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_193
 ; CHECK-RV64-NEXT:  .LBB61_711: # %cond.load749
 ; CHECK-RV64-NEXT:    lbu a1, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v12, a1
 ; CHECK-RV64-NEXT:    li a1, 189
 ; CHECK-RV64-NEXT:    li a3, 188
@@ -15808,9 +15440,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_199
 ; CHECK-RV64-NEXT:  .LBB61_713: # %cond.load765
 ; CHECK-RV64-NEXT:    lbu a2, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v12, a2
 ; CHECK-RV64-NEXT:    li a2, 193
 ; CHECK-RV64-NEXT:    li a3, 192
@@ -15824,9 +15455,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_200
 ; CHECK-RV64-NEXT:  .LBB61_714: # %cond.load769
 ; CHECK-RV64-NEXT:    lbu a2, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v12, a2
 ; CHECK-RV64-NEXT:    li a2, 194
 ; CHECK-RV64-NEXT:    li a3, 193
@@ -15840,9 +15470,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_201
 ; CHECK-RV64-NEXT:  .LBB61_715: # %cond.load773
 ; CHECK-RV64-NEXT:    lbu a2, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v12, a2
 ; CHECK-RV64-NEXT:    li a2, 195
 ; CHECK-RV64-NEXT:    li a3, 194
@@ -15856,9 +15485,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_202
 ; CHECK-RV64-NEXT:  .LBB61_716: # %cond.load777
 ; CHECK-RV64-NEXT:    lbu a2, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v12, a2
 ; CHECK-RV64-NEXT:    li a2, 196
 ; CHECK-RV64-NEXT:    li a3, 195
@@ -15872,9 +15500,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_203
 ; CHECK-RV64-NEXT:  .LBB61_717: # %cond.load781
 ; CHECK-RV64-NEXT:    lbu a2, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v12, a2
 ; CHECK-RV64-NEXT:    li a2, 197
 ; CHECK-RV64-NEXT:    li a3, 196
@@ -15888,9 +15515,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_204
 ; CHECK-RV64-NEXT:  .LBB61_718: # %cond.load785
 ; CHECK-RV64-NEXT:    lbu a2, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v12, a2
 ; CHECK-RV64-NEXT:    li a2, 198
 ; CHECK-RV64-NEXT:    li a3, 197
@@ -15904,9 +15530,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_205
 ; CHECK-RV64-NEXT:  .LBB61_719: # %cond.load789
 ; CHECK-RV64-NEXT:    lbu a2, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v12, a2
 ; CHECK-RV64-NEXT:    li a2, 199
 ; CHECK-RV64-NEXT:    li a3, 198
@@ -15920,9 +15545,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_206
 ; CHECK-RV64-NEXT:  .LBB61_720: # %cond.load793
 ; CHECK-RV64-NEXT:    lbu a2, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v12, a2
 ; CHECK-RV64-NEXT:    li a2, 200
 ; CHECK-RV64-NEXT:    li a3, 199
@@ -15936,9 +15560,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_207
 ; CHECK-RV64-NEXT:  .LBB61_721: # %cond.load797
 ; CHECK-RV64-NEXT:    lbu a2, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v12, a2
 ; CHECK-RV64-NEXT:    li a2, 201
 ; CHECK-RV64-NEXT:    li a3, 200
@@ -15952,9 +15575,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_208
 ; CHECK-RV64-NEXT:  .LBB61_722: # %cond.load801
 ; CHECK-RV64-NEXT:    lbu a2, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v12, a2
 ; CHECK-RV64-NEXT:    li a2, 202
 ; CHECK-RV64-NEXT:    li a3, 201
@@ -15968,9 +15590,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_209
 ; CHECK-RV64-NEXT:  .LBB61_723: # %cond.load805
 ; CHECK-RV64-NEXT:    lbu a2, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v12, a2
 ; CHECK-RV64-NEXT:    li a2, 203
 ; CHECK-RV64-NEXT:    li a3, 202
@@ -15984,9 +15605,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_210
 ; CHECK-RV64-NEXT:  .LBB61_724: # %cond.load809
 ; CHECK-RV64-NEXT:    lbu a2, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v12, a2
 ; CHECK-RV64-NEXT:    li a2, 204
 ; CHECK-RV64-NEXT:    li a3, 203
@@ -16000,9 +15620,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_211
 ; CHECK-RV64-NEXT:  .LBB61_725: # %cond.load813
 ; CHECK-RV64-NEXT:    lbu a2, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v12, a2
 ; CHECK-RV64-NEXT:    li a2, 205
 ; CHECK-RV64-NEXT:    li a3, 204
@@ -16016,9 +15635,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_212
 ; CHECK-RV64-NEXT:  .LBB61_726: # %cond.load817
 ; CHECK-RV64-NEXT:    lbu a2, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v12, a2
 ; CHECK-RV64-NEXT:    li a2, 206
 ; CHECK-RV64-NEXT:    li a3, 205
@@ -16032,9 +15650,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_213
 ; CHECK-RV64-NEXT:  .LBB61_727: # %cond.load821
 ; CHECK-RV64-NEXT:    lbu a2, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v12, a2
 ; CHECK-RV64-NEXT:    li a2, 207
 ; CHECK-RV64-NEXT:    li a3, 206
@@ -16048,9 +15665,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_214
 ; CHECK-RV64-NEXT:  .LBB61_728: # %cond.load825
 ; CHECK-RV64-NEXT:    lbu a2, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v12, a2
 ; CHECK-RV64-NEXT:    li a2, 208
 ; CHECK-RV64-NEXT:    li a3, 207
@@ -16064,9 +15680,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_215
 ; CHECK-RV64-NEXT:  .LBB61_729: # %cond.load829
 ; CHECK-RV64-NEXT:    lbu a2, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v12, a2
 ; CHECK-RV64-NEXT:    li a2, 209
 ; CHECK-RV64-NEXT:    li a3, 208
@@ -16080,9 +15695,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_216
 ; CHECK-RV64-NEXT:  .LBB61_730: # %cond.load833
 ; CHECK-RV64-NEXT:    lbu a2, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v12, a2
 ; CHECK-RV64-NEXT:    li a2, 210
 ; CHECK-RV64-NEXT:    li a3, 209
@@ -16096,9 +15710,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_217
 ; CHECK-RV64-NEXT:  .LBB61_731: # %cond.load837
 ; CHECK-RV64-NEXT:    lbu a2, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v12, a2
 ; CHECK-RV64-NEXT:    li a2, 211
 ; CHECK-RV64-NEXT:    li a3, 210
@@ -16112,9 +15725,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_218
 ; CHECK-RV64-NEXT:  .LBB61_732: # %cond.load841
 ; CHECK-RV64-NEXT:    lbu a2, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v12, a2
 ; CHECK-RV64-NEXT:    li a2, 212
 ; CHECK-RV64-NEXT:    li a3, 211
@@ -16128,9 +15740,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_219
 ; CHECK-RV64-NEXT:  .LBB61_733: # %cond.load845
 ; CHECK-RV64-NEXT:    lbu a2, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v12, a2
 ; CHECK-RV64-NEXT:    li a2, 213
 ; CHECK-RV64-NEXT:    li a3, 212
@@ -16144,9 +15755,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_220
 ; CHECK-RV64-NEXT:  .LBB61_734: # %cond.load849
 ; CHECK-RV64-NEXT:    lbu a2, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v12, a2
 ; CHECK-RV64-NEXT:    li a2, 214
 ; CHECK-RV64-NEXT:    li a3, 213
@@ -16160,9 +15770,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_221
 ; CHECK-RV64-NEXT:  .LBB61_735: # %cond.load853
 ; CHECK-RV64-NEXT:    lbu a2, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v12, a2
 ; CHECK-RV64-NEXT:    li a2, 215
 ; CHECK-RV64-NEXT:    li a3, 214
@@ -16176,9 +15785,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_222
 ; CHECK-RV64-NEXT:  .LBB61_736: # %cond.load857
 ; CHECK-RV64-NEXT:    lbu a2, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v12, a2
 ; CHECK-RV64-NEXT:    li a2, 216
 ; CHECK-RV64-NEXT:    li a3, 215
@@ -16192,9 +15800,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_223
 ; CHECK-RV64-NEXT:  .LBB61_737: # %cond.load861
 ; CHECK-RV64-NEXT:    lbu a2, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v12, a2
 ; CHECK-RV64-NEXT:    li a2, 217
 ; CHECK-RV64-NEXT:    li a3, 216
@@ -16208,9 +15815,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_224
 ; CHECK-RV64-NEXT:  .LBB61_738: # %cond.load865
 ; CHECK-RV64-NEXT:    lbu a2, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v12, a2
 ; CHECK-RV64-NEXT:    li a2, 218
 ; CHECK-RV64-NEXT:    li a3, 217
@@ -16224,9 +15830,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_225
 ; CHECK-RV64-NEXT:  .LBB61_739: # %cond.load869
 ; CHECK-RV64-NEXT:    lbu a2, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v12, a2
 ; CHECK-RV64-NEXT:    li a2, 219
 ; CHECK-RV64-NEXT:    li a3, 218
@@ -16240,9 +15845,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_226
 ; CHECK-RV64-NEXT:  .LBB61_740: # %cond.load873
 ; CHECK-RV64-NEXT:    lbu a2, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v12, a2
 ; CHECK-RV64-NEXT:    li a2, 220
 ; CHECK-RV64-NEXT:    li a3, 219
@@ -16256,9 +15860,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_227
 ; CHECK-RV64-NEXT:  .LBB61_741: # %cond.load877
 ; CHECK-RV64-NEXT:    lbu a2, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v12, a2
 ; CHECK-RV64-NEXT:    li a2, 221
 ; CHECK-RV64-NEXT:    li a3, 220
@@ -16272,9 +15875,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_228
 ; CHECK-RV64-NEXT:  .LBB61_742: # %cond.load881
 ; CHECK-RV64-NEXT:    lbu a2, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v12, a2
 ; CHECK-RV64-NEXT:    li a2, 222
 ; CHECK-RV64-NEXT:    li a3, 221
@@ -16288,9 +15890,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_229
 ; CHECK-RV64-NEXT:  .LBB61_743: # %cond.load885
 ; CHECK-RV64-NEXT:    lbu a2, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v12, a2
 ; CHECK-RV64-NEXT:    li a2, 223
 ; CHECK-RV64-NEXT:    li a3, 222
@@ -16304,9 +15905,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_230
 ; CHECK-RV64-NEXT:  .LBB61_744: # %cond.load889
 ; CHECK-RV64-NEXT:    lbu a2, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v12, a2
 ; CHECK-RV64-NEXT:    li a2, 224
 ; CHECK-RV64-NEXT:    li a3, 223
@@ -16320,9 +15920,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_231
 ; CHECK-RV64-NEXT:  .LBB61_745: # %cond.load893
 ; CHECK-RV64-NEXT:    lbu a2, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v12, a2
 ; CHECK-RV64-NEXT:    li a2, 225
 ; CHECK-RV64-NEXT:    li a3, 224
@@ -16336,9 +15935,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_232
 ; CHECK-RV64-NEXT:  .LBB61_746: # %cond.load897
 ; CHECK-RV64-NEXT:    lbu a2, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v12, a2
 ; CHECK-RV64-NEXT:    li a2, 226
 ; CHECK-RV64-NEXT:    li a3, 225
@@ -16352,9 +15950,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_233
 ; CHECK-RV64-NEXT:  .LBB61_747: # %cond.load901
 ; CHECK-RV64-NEXT:    lbu a2, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v12, a2
 ; CHECK-RV64-NEXT:    li a2, 227
 ; CHECK-RV64-NEXT:    li a3, 226
@@ -16368,9 +15965,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_234
 ; CHECK-RV64-NEXT:  .LBB61_748: # %cond.load905
 ; CHECK-RV64-NEXT:    lbu a2, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v12, a2
 ; CHECK-RV64-NEXT:    li a2, 228
 ; CHECK-RV64-NEXT:    li a3, 227
@@ -16384,9 +15980,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_235
 ; CHECK-RV64-NEXT:  .LBB61_749: # %cond.load909
 ; CHECK-RV64-NEXT:    lbu a2, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v12, a2
 ; CHECK-RV64-NEXT:    li a2, 229
 ; CHECK-RV64-NEXT:    li a3, 228
@@ -16400,9 +15995,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_236
 ; CHECK-RV64-NEXT:  .LBB61_750: # %cond.load913
 ; CHECK-RV64-NEXT:    lbu a2, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v12, a2
 ; CHECK-RV64-NEXT:    li a2, 230
 ; CHECK-RV64-NEXT:    li a3, 229
@@ -16416,9 +16010,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_237
 ; CHECK-RV64-NEXT:  .LBB61_751: # %cond.load917
 ; CHECK-RV64-NEXT:    lbu a2, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v12, a2
 ; CHECK-RV64-NEXT:    li a2, 231
 ; CHECK-RV64-NEXT:    li a3, 230
@@ -16432,9 +16025,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_238
 ; CHECK-RV64-NEXT:  .LBB61_752: # %cond.load921
 ; CHECK-RV64-NEXT:    lbu a2, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v12, a2
 ; CHECK-RV64-NEXT:    li a2, 232
 ; CHECK-RV64-NEXT:    li a3, 231
@@ -16448,9 +16040,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_239
 ; CHECK-RV64-NEXT:  .LBB61_753: # %cond.load925
 ; CHECK-RV64-NEXT:    lbu a2, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v12, a2
 ; CHECK-RV64-NEXT:    li a2, 233
 ; CHECK-RV64-NEXT:    li a3, 232
@@ -16464,9 +16055,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_240
 ; CHECK-RV64-NEXT:  .LBB61_754: # %cond.load929
 ; CHECK-RV64-NEXT:    lbu a2, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v12, a2
 ; CHECK-RV64-NEXT:    li a2, 234
 ; CHECK-RV64-NEXT:    li a3, 233
@@ -16480,9 +16070,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_241
 ; CHECK-RV64-NEXT:  .LBB61_755: # %cond.load933
 ; CHECK-RV64-NEXT:    lbu a2, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v12, a2
 ; CHECK-RV64-NEXT:    li a2, 235
 ; CHECK-RV64-NEXT:    li a3, 234
@@ -16496,9 +16085,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_242
 ; CHECK-RV64-NEXT:  .LBB61_756: # %cond.load937
 ; CHECK-RV64-NEXT:    lbu a2, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v12, a2
 ; CHECK-RV64-NEXT:    li a2, 236
 ; CHECK-RV64-NEXT:    li a3, 235
@@ -16512,9 +16100,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_243
 ; CHECK-RV64-NEXT:  .LBB61_757: # %cond.load941
 ; CHECK-RV64-NEXT:    lbu a2, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v12, a2
 ; CHECK-RV64-NEXT:    li a2, 237
 ; CHECK-RV64-NEXT:    li a3, 236
@@ -16528,9 +16115,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_244
 ; CHECK-RV64-NEXT:  .LBB61_758: # %cond.load945
 ; CHECK-RV64-NEXT:    lbu a2, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v12, a2
 ; CHECK-RV64-NEXT:    li a2, 238
 ; CHECK-RV64-NEXT:    li a3, 237
@@ -16544,9 +16130,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_245
 ; CHECK-RV64-NEXT:  .LBB61_759: # %cond.load949
 ; CHECK-RV64-NEXT:    lbu a2, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v12, a2
 ; CHECK-RV64-NEXT:    li a2, 239
 ; CHECK-RV64-NEXT:    li a3, 238
@@ -16560,9 +16145,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_246
 ; CHECK-RV64-NEXT:  .LBB61_760: # %cond.load953
 ; CHECK-RV64-NEXT:    lbu a2, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v12, a2
 ; CHECK-RV64-NEXT:    li a2, 240
 ; CHECK-RV64-NEXT:    li a3, 239
@@ -16576,9 +16160,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_247
 ; CHECK-RV64-NEXT:  .LBB61_761: # %cond.load957
 ; CHECK-RV64-NEXT:    lbu a2, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v12, a2
 ; CHECK-RV64-NEXT:    li a2, 241
 ; CHECK-RV64-NEXT:    li a3, 240
@@ -16592,9 +16175,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_248
 ; CHECK-RV64-NEXT:  .LBB61_762: # %cond.load961
 ; CHECK-RV64-NEXT:    lbu a2, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v12, a2
 ; CHECK-RV64-NEXT:    li a2, 242
 ; CHECK-RV64-NEXT:    li a3, 241
@@ -16608,9 +16190,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_249
 ; CHECK-RV64-NEXT:  .LBB61_763: # %cond.load965
 ; CHECK-RV64-NEXT:    lbu a2, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v12, a2
 ; CHECK-RV64-NEXT:    li a2, 243
 ; CHECK-RV64-NEXT:    li a3, 242
@@ -16624,9 +16205,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_250
 ; CHECK-RV64-NEXT:  .LBB61_764: # %cond.load969
 ; CHECK-RV64-NEXT:    lbu a2, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v12, a2
 ; CHECK-RV64-NEXT:    li a2, 244
 ; CHECK-RV64-NEXT:    li a3, 243
@@ -16640,9 +16220,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_251
 ; CHECK-RV64-NEXT:  .LBB61_765: # %cond.load973
 ; CHECK-RV64-NEXT:    lbu a2, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v12, a2
 ; CHECK-RV64-NEXT:    li a2, 245
 ; CHECK-RV64-NEXT:    li a3, 244
@@ -16656,9 +16235,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_252
 ; CHECK-RV64-NEXT:  .LBB61_766: # %cond.load977
 ; CHECK-RV64-NEXT:    lbu a2, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v12, a2
 ; CHECK-RV64-NEXT:    li a2, 246
 ; CHECK-RV64-NEXT:    li a3, 245
@@ -16672,9 +16250,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_253
 ; CHECK-RV64-NEXT:  .LBB61_767: # %cond.load981
 ; CHECK-RV64-NEXT:    lbu a2, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v12, a2
 ; CHECK-RV64-NEXT:    li a2, 247
 ; CHECK-RV64-NEXT:    li a3, 246
@@ -16688,9 +16265,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_254
 ; CHECK-RV64-NEXT:  .LBB61_768: # %cond.load985
 ; CHECK-RV64-NEXT:    lbu a2, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v12, a2
 ; CHECK-RV64-NEXT:    li a2, 248
 ; CHECK-RV64-NEXT:    li a3, 247
@@ -16704,9 +16280,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_255
 ; CHECK-RV64-NEXT:  .LBB61_769: # %cond.load989
 ; CHECK-RV64-NEXT:    lbu a2, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v12, a2
 ; CHECK-RV64-NEXT:    li a2, 249
 ; CHECK-RV64-NEXT:    li a3, 248
@@ -16720,9 +16295,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_256
 ; CHECK-RV64-NEXT:  .LBB61_770: # %cond.load993
 ; CHECK-RV64-NEXT:    lbu a2, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v12, a2
 ; CHECK-RV64-NEXT:    li a2, 250
 ; CHECK-RV64-NEXT:    li a3, 249
@@ -16736,9 +16310,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_257
 ; CHECK-RV64-NEXT:  .LBB61_771: # %cond.load997
 ; CHECK-RV64-NEXT:    lbu a2, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v12, a2
 ; CHECK-RV64-NEXT:    li a2, 251
 ; CHECK-RV64-NEXT:    li a3, 250
@@ -16752,9 +16325,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_258
 ; CHECK-RV64-NEXT:  .LBB61_772: # %cond.load1001
 ; CHECK-RV64-NEXT:    lbu a2, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v12, a2
 ; CHECK-RV64-NEXT:    li a2, 252
 ; CHECK-RV64-NEXT:    li a3, 251
@@ -16768,9 +16340,8 @@ define <512 x i8> @test_expandload_v512i8_vlen512(ptr %base, <512 x i1> %mask, <
 ; CHECK-RV64-NEXT:    j .LBB61_259
 ; CHECK-RV64-NEXT:  .LBB61_773: # %cond.load1005
 ; CHECK-RV64-NEXT:    lbu a2, 0(a0)
-; CHECK-RV64-NEXT:    li a3, 512
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv8r.v v16, v8
-; CHECK-RV64-NEXT:    vsetvli zero, a3, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv.s.x v12, a2
 ; CHECK-RV64-NEXT:    li a2, 253
 ; CHECK-RV64-NEXT:    li a3, 252
diff --git a/llvm/test/CodeGen/RISCV/rvv/extract-subvector.ll b/llvm/test/CodeGen/RISCV/rvv/extract-subvector.ll
index 869478a1efa78..83637e4a71d45 100644
--- a/llvm/test/CodeGen/RISCV/rvv/extract-subvector.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/extract-subvector.ll
@@ -13,6 +13,7 @@ define <vscale x 4 x i32> @extract_nxv8i32_nxv4i32_0(<vscale x 8 x i32> %vec) {
 define <vscale x 4 x i32> @extract_nxv8i32_nxv4i32_4(<vscale x 8 x i32> %vec) {
 ; CHECK-LABEL: extract_nxv8i32_nxv4i32_4:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-NEXT:    vmv2r.v v8, v10
 ; CHECK-NEXT:    ret
   %c = call <vscale x 4 x i32> @llvm.vector.extract.nxv4i32.nxv8i32(<vscale x 8 x i32> %vec, i64 4)
@@ -30,6 +31,7 @@ define <vscale x 2 x i32> @extract_nxv8i32_nxv2i32_0(<vscale x 8 x i32> %vec) {
 define <vscale x 2 x i32> @extract_nxv8i32_nxv2i32_2(<vscale x 8 x i32> %vec) {
 ; CHECK-LABEL: extract_nxv8i32_nxv2i32_2:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    ret
   %c = call <vscale x 2 x i32> @llvm.vector.extract.nxv2i32.nxv8i32(<vscale x 8 x i32> %vec, i64 2)
@@ -39,6 +41,7 @@ define <vscale x 2 x i32> @extract_nxv8i32_nxv2i32_2(<vscale x 8 x i32> %vec) {
 define <vscale x 2 x i32> @extract_nxv8i32_nxv2i32_4(<vscale x 8 x i32> %vec) {
 ; CHECK-LABEL: extract_nxv8i32_nxv2i32_4:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-NEXT:    vmv1r.v v8, v10
 ; CHECK-NEXT:    ret
   %c = call <vscale x 2 x i32> @llvm.vector.extract.nxv2i32.nxv8i32(<vscale x 8 x i32> %vec, i64 4)
@@ -48,6 +51,7 @@ define <vscale x 2 x i32> @extract_nxv8i32_nxv2i32_4(<vscale x 8 x i32> %vec) {
 define <vscale x 2 x i32> @extract_nxv8i32_nxv2i32_6(<vscale x 8 x i32> %vec) {
 ; CHECK-LABEL: extract_nxv8i32_nxv2i32_6:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-NEXT:    vmv1r.v v8, v11
 ; CHECK-NEXT:    ret
   %c = call <vscale x 2 x i32> @llvm.vector.extract.nxv2i32.nxv8i32(<vscale x 8 x i32> %vec, i64 6)
@@ -65,6 +69,7 @@ define <vscale x 8 x i32> @extract_nxv16i32_nxv8i32_0(<vscale x 16 x i32> %vec)
 define <vscale x 8 x i32> @extract_nxv16i32_nxv8i32_8(<vscale x 16 x i32> %vec) {
 ; CHECK-LABEL: extract_nxv16i32_nxv8i32_8:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-NEXT:    vmv4r.v v8, v12
 ; CHECK-NEXT:    ret
   %c = call <vscale x 8 x i32> @llvm.vector.extract.nxv8i32.nxv16i32(<vscale x 16 x i32> %vec, i64 8)
@@ -82,6 +87,7 @@ define <vscale x 4 x i32> @extract_nxv16i32_nxv4i32_0(<vscale x 16 x i32> %vec)
 define <vscale x 4 x i32> @extract_nxv16i32_nxv4i32_4(<vscale x 16 x i32> %vec) {
 ; CHECK-LABEL: extract_nxv16i32_nxv4i32_4:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-NEXT:    vmv2r.v v8, v10
 ; CHECK-NEXT:    ret
   %c = call <vscale x 4 x i32> @llvm.vector.extract.nxv4i32.nxv16i32(<vscale x 16 x i32> %vec, i64 4)
@@ -91,6 +97,7 @@ define <vscale x 4 x i32> @extract_nxv16i32_nxv4i32_4(<vscale x 16 x i32> %vec)
 define <vscale x 4 x i32> @extract_nxv16i32_nxv4i32_8(<vscale x 16 x i32> %vec) {
 ; CHECK-LABEL: extract_nxv16i32_nxv4i32_8:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-NEXT:    vmv2r.v v8, v12
 ; CHECK-NEXT:    ret
   %c = call <vscale x 4 x i32> @llvm.vector.extract.nxv4i32.nxv16i32(<vscale x 16 x i32> %vec, i64 8)
@@ -100,6 +107,7 @@ define <vscale x 4 x i32> @extract_nxv16i32_nxv4i32_8(<vscale x 16 x i32> %vec)
 define <vscale x 4 x i32> @extract_nxv16i32_nxv4i32_12(<vscale x 16 x i32> %vec) {
 ; CHECK-LABEL: extract_nxv16i32_nxv4i32_12:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-NEXT:    vmv2r.v v8, v14
 ; CHECK-NEXT:    ret
   %c = call <vscale x 4 x i32> @llvm.vector.extract.nxv4i32.nxv16i32(<vscale x 16 x i32> %vec, i64 12)
@@ -117,6 +125,7 @@ define <vscale x 2 x i32> @extract_nxv16i32_nxv2i32_0(<vscale x 16 x i32> %vec)
 define <vscale x 2 x i32> @extract_nxv16i32_nxv2i32_2(<vscale x 16 x i32> %vec) {
 ; CHECK-LABEL: extract_nxv16i32_nxv2i32_2:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    ret
   %c = call <vscale x 2 x i32> @llvm.vector.extract.nxv2i32.nxv16i32(<vscale x 16 x i32> %vec, i64 2)
@@ -126,6 +135,7 @@ define <vscale x 2 x i32> @extract_nxv16i32_nxv2i32_2(<vscale x 16 x i32> %vec)
 define <vscale x 2 x i32> @extract_nxv16i32_nxv2i32_4(<vscale x 16 x i32> %vec) {
 ; CHECK-LABEL: extract_nxv16i32_nxv2i32_4:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-NEXT:    vmv1r.v v8, v10
 ; CHECK-NEXT:    ret
   %c = call <vscale x 2 x i32> @llvm.vector.extract.nxv2i32.nxv16i32(<vscale x 16 x i32> %vec, i64 4)
@@ -135,6 +145,7 @@ define <vscale x 2 x i32> @extract_nxv16i32_nxv2i32_4(<vscale x 16 x i32> %vec)
 define <vscale x 2 x i32> @extract_nxv16i32_nxv2i32_6(<vscale x 16 x i32> %vec) {
 ; CHECK-LABEL: extract_nxv16i32_nxv2i32_6:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-NEXT:    vmv1r.v v8, v11
 ; CHECK-NEXT:    ret
   %c = call <vscale x 2 x i32> @llvm.vector.extract.nxv2i32.nxv16i32(<vscale x 16 x i32> %vec, i64 6)
@@ -144,6 +155,7 @@ define <vscale x 2 x i32> @extract_nxv16i32_nxv2i32_6(<vscale x 16 x i32> %vec)
 define <vscale x 2 x i32> @extract_nxv16i32_nxv2i32_8(<vscale x 16 x i32> %vec) {
 ; CHECK-LABEL: extract_nxv16i32_nxv2i32_8:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-NEXT:    vmv1r.v v8, v12
 ; CHECK-NEXT:    ret
   %c = call <vscale x 2 x i32> @llvm.vector.extract.nxv2i32.nxv16i32(<vscale x 16 x i32> %vec, i64 8)
@@ -153,6 +165,7 @@ define <vscale x 2 x i32> @extract_nxv16i32_nxv2i32_8(<vscale x 16 x i32> %vec)
 define <vscale x 2 x i32> @extract_nxv16i32_nxv2i32_10(<vscale x 16 x i32> %vec) {
 ; CHECK-LABEL: extract_nxv16i32_nxv2i32_10:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-NEXT:    vmv1r.v v8, v13
 ; CHECK-NEXT:    ret
   %c = call <vscale x 2 x i32> @llvm.vector.extract.nxv2i32.nxv16i32(<vscale x 16 x i32> %vec, i64 10)
@@ -162,6 +175,7 @@ define <vscale x 2 x i32> @extract_nxv16i32_nxv2i32_10(<vscale x 16 x i32> %vec)
 define <vscale x 2 x i32> @extract_nxv16i32_nxv2i32_12(<vscale x 16 x i32> %vec) {
 ; CHECK-LABEL: extract_nxv16i32_nxv2i32_12:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-NEXT:    vmv1r.v v8, v14
 ; CHECK-NEXT:    ret
   %c = call <vscale x 2 x i32> @llvm.vector.extract.nxv2i32.nxv16i32(<vscale x 16 x i32> %vec, i64 12)
@@ -171,6 +185,7 @@ define <vscale x 2 x i32> @extract_nxv16i32_nxv2i32_12(<vscale x 16 x i32> %vec)
 define <vscale x 2 x i32> @extract_nxv16i32_nxv2i32_14(<vscale x 16 x i32> %vec) {
 ; CHECK-LABEL: extract_nxv16i32_nxv2i32_14:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-NEXT:    vmv1r.v v8, v15
 ; CHECK-NEXT:    ret
   %c = call <vscale x 2 x i32> @llvm.vector.extract.nxv2i32.nxv16i32(<vscale x 16 x i32> %vec, i64 14)
@@ -224,6 +239,7 @@ define <vscale x 1 x i32> @extract_nxv16i32_nxv1i32_15(<vscale x 16 x i32> %vec)
 define <vscale x 1 x i32> @extract_nxv16i32_nxv1i32_2(<vscale x 16 x i32> %vec) {
 ; CHECK-LABEL: extract_nxv16i32_nxv1i32_2:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    ret
   %c = call <vscale x 1 x i32> @llvm.vector.extract.nxv1i32.nxv16i32(<vscale x 16 x i32> %vec, i64 2)
@@ -287,6 +303,7 @@ define <vscale x 2 x i8> @extract_nxv32i8_nxv2i8_6(<vscale x 32 x i8> %vec) {
 define <vscale x 2 x i8> @extract_nxv32i8_nxv2i8_8(<vscale x 32 x i8> %vec) {
 ; CHECK-LABEL: extract_nxv32i8_nxv2i8_8:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    ret
   %c = call <vscale x 2 x i8> @llvm.vector.extract.nxv2i8.nxv32i8(<vscale x 32 x i8> %vec, i64 8)
@@ -357,6 +374,7 @@ define <vscale x 2 x half> @extract_nxv2f16_nxv16f16_2(<vscale x 16 x half> %vec
 define <vscale x 2 x half> @extract_nxv2f16_nxv16f16_4(<vscale x 16 x half> %vec) {
 ; CHECK-LABEL: extract_nxv2f16_nxv16f16_4:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    ret
   %c = call <vscale x 2 x half> @llvm.vector.extract.nxv2f16.nxv16f16(<vscale x 16 x half> %vec, i64 4)
@@ -504,6 +522,7 @@ define <vscale x 2 x bfloat> @extract_nxv2bf16_nxv16bf16_2(<vscale x 16 x bfloat
 define <vscale x 2 x bfloat> @extract_nxv2bf16_nxv16bf16_4(<vscale x 16 x bfloat> %vec) {
 ; CHECK-LABEL: extract_nxv2bf16_nxv16bf16_4:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    ret
   %c = call <vscale x 2 x bfloat> @llvm.vector.extract.nxv2bf16.nxv16bf16(<vscale x 16 x bfloat> %vec, i64 4)
diff --git a/llvm/test/CodeGen/RISCV/rvv/fixed-vector-i8-index-cornercase.ll b/llvm/test/CodeGen/RISCV/rvv/fixed-vector-i8-index-cornercase.ll
index ce83e2d8a6220..1a5ca429b531f 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vector-i8-index-cornercase.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vector-i8-index-cornercase.ll
@@ -16,10 +16,10 @@ define <512 x i8> @single_source(<512 x i8> %a) {
 ; CHECK-NEXT:    addi s0, sp, 1536
 ; CHECK-NEXT:    .cfi_def_cfa s0, 0
 ; CHECK-NEXT:    andi sp, sp, -512
+; CHECK-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-NEXT:    vmv8r.v v16, v8
 ; CHECK-NEXT:    li a0, 512
 ; CHECK-NEXT:    addi a1, sp, 512
-; CHECK-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-NEXT:    vmv.x.s a2, v16
 ; CHECK-NEXT:    vslidedown.vi v24, v16, 5
 ; CHECK-NEXT:    li a3, 432
@@ -104,10 +104,10 @@ define <512 x i8> @two_source(<512 x i8> %a, <512 x i8> %b) {
 ; CHECK-NEXT:    addi s0, sp, 1536
 ; CHECK-NEXT:    .cfi_def_cfa s0, 0
 ; CHECK-NEXT:    andi sp, sp, -512
+; CHECK-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-NEXT:    vmv8r.v v24, v8
 ; CHECK-NEXT:    li a0, 512
 ; CHECK-NEXT:    addi a1, sp, 512
-; CHECK-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-NEXT:    vslidedown.vi v0, v24, 5
 ; CHECK-NEXT:    vmv.x.s a2, v24
 ; CHECK-NEXT:    li a3, 432
diff --git a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-bitreverse-vp.ll b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-bitreverse-vp.ll
index 3eb5d36b4896a..5ea4924468595 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-bitreverse-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-bitreverse-vp.ll
@@ -1659,6 +1659,7 @@ define <15 x i64> @vp_bitreverse_v15i64(<15 x i64> %va, <15 x i1> %m, i32 zeroex
 ; RV32-NEXT:    mul a1, a1, a2
 ; RV32-NEXT:    sub sp, sp, a1
 ; RV32-NEXT:    .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x30, 0x22, 0x11, 0x18, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 48 + 24 * vlenb
+; RV32-NEXT:    vsetvli zero, a0, e64, m8, ta, ma
 ; RV32-NEXT:    vmv8r.v v24, v8
 ; RV32-NEXT:    lui a2, 1044480
 ; RV32-NEXT:    lui a3, 61681
@@ -1677,7 +1678,6 @@ define <15 x i64> @vp_bitreverse_v15i64(<15 x i64> %va, <15 x i1> %m, i32 zeroex
 ; RV32-NEXT:    sw a3, 36(sp)
 ; RV32-NEXT:    addi a3, sp, 16
 ; RV32-NEXT:    addi a4, a5, 1365
-; RV32-NEXT:    vsetvli zero, a0, e64, m8, ta, ma
 ; RV32-NEXT:    vsll.vx v16, v8, a1, v0.t
 ; RV32-NEXT:    addi a5, a6, -256
 ; RV32-NEXT:    sw a4, 24(sp)
@@ -2055,6 +2055,7 @@ define <16 x i64> @vp_bitreverse_v16i64(<16 x i64> %va, <16 x i1> %m, i32 zeroex
 ; RV32-NEXT:    mul a1, a1, a2
 ; RV32-NEXT:    sub sp, sp, a1
 ; RV32-NEXT:    .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x30, 0x22, 0x11, 0x18, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 48 + 24 * vlenb
+; RV32-NEXT:    vsetvli zero, a0, e64, m8, ta, ma
 ; RV32-NEXT:    vmv8r.v v24, v8
 ; RV32-NEXT:    lui a2, 1044480
 ; RV32-NEXT:    lui a3, 61681
@@ -2073,7 +2074,6 @@ define <16 x i64> @vp_bitreverse_v16i64(<16 x i64> %va, <16 x i1> %m, i32 zeroex
 ; RV32-NEXT:    sw a3, 36(sp)
 ; RV32-NEXT:    addi a3, sp, 16
 ; RV32-NEXT:    addi a4, a5, 1365
-; RV32-NEXT:    vsetvli zero, a0, e64, m8, ta, ma
 ; RV32-NEXT:    vsll.vx v16, v8, a1, v0.t
 ; RV32-NEXT:    addi a5, a6, -256
 ; RV32-NEXT:    sw a4, 24(sp)
diff --git a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-calling-conv-fastcc.ll b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-calling-conv-fastcc.ll
index ee953a66a004f..60a9948198c8f 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-calling-conv-fastcc.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-calling-conv-fastcc.ll
@@ -180,6 +180,7 @@ define fastcc <32 x i32> @ret_v32i32_call_v32i32_v32i32_i32(<32 x i32> %x, <32 x
 ; CHECK-NEXT:    .cfi_def_cfa_offset 16
 ; CHECK-NEXT:    sd ra, 8(sp) # 8-byte Folded Spill
 ; CHECK-NEXT:    .cfi_offset ra, -8
+; CHECK-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-NEXT:    vmv8r.v v24, v8
 ; CHECK-NEXT:    li a1, 2
 ; CHECK-NEXT:    vmv8r.v v8, v16
diff --git a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-calling-conv.ll b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-calling-conv.ll
index 73e148edbe2d6..f42b4a3a26aad 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-calling-conv.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-calling-conv.ll
@@ -180,6 +180,7 @@ define <32 x i32> @ret_v32i32_call_v32i32_v32i32_i32(<32 x i32> %x, <32 x i32> %
 ; CHECK-NEXT:    .cfi_def_cfa_offset 16
 ; CHECK-NEXT:    sd ra, 8(sp) # 8-byte Folded Spill
 ; CHECK-NEXT:    .cfi_offset ra, -8
+; CHECK-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-NEXT:    vmv8r.v v24, v8
 ; CHECK-NEXT:    li a1, 2
 ; CHECK-NEXT:    vmv8r.v v8, v16
diff --git a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-ceil-vp.ll b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-ceil-vp.ll
index 511242aa677c2..5fe55583f314c 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-ceil-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-ceil-vp.ll
@@ -194,8 +194,8 @@ define <8 x half> @vp_ceil_v8f16(<8 x half> %va, <8 x i1> %m, i32 zeroext %evl)
 ;
 ; ZVFHMIN-LABEL: vp_ceil_v8f16:
 ; ZVFHMIN:       # %bb.0:
-; ZVFHMIN-NEXT:    vmv1r.v v9, v0
 ; ZVFHMIN-NEXT:    vsetivli zero, 8, e16, m1, ta, ma
+; ZVFHMIN-NEXT:    vmv1r.v v9, v0
 ; ZVFHMIN-NEXT:    vfwcvt.f.f.v v10, v8
 ; ZVFHMIN-NEXT:    lui a1, 307200
 ; ZVFHMIN-NEXT:    vsetvli zero, a0, e32, m2, ta, ma
@@ -261,10 +261,10 @@ declare <16 x half> @llvm.vp.ceil.v16f16(<16 x half>, <16 x i1>, i32)
 define <16 x half> @vp_ceil_v16f16(<16 x half> %va, <16 x i1> %m, i32 zeroext %evl) {
 ; ZVFH-LABEL: vp_ceil_v16f16:
 ; ZVFH:       # %bb.0:
-; ZVFH-NEXT:    vmv1r.v v10, v0
-; ZVFH-NEXT:    lui a1, %hi(.LCPI6_0)
-; ZVFH-NEXT:    flh fa5, %lo(.LCPI6_0)(a1)
 ; ZVFH-NEXT:    vsetvli zero, a0, e16, m2, ta, ma
+; ZVFH-NEXT:    vmv1r.v v10, v0
+; ZVFH-NEXT:    lui a0, %hi(.LCPI6_0)
+; ZVFH-NEXT:    flh fa5, %lo(.LCPI6_0)(a0)
 ; ZVFH-NEXT:    vfabs.v v12, v8, v0.t
 ; ZVFH-NEXT:    vsetvli zero, zero, e16, m2, ta, mu
 ; ZVFH-NEXT:    vmflt.vf v10, v12, fa5, v0.t
@@ -280,8 +280,8 @@ define <16 x half> @vp_ceil_v16f16(<16 x half> %va, <16 x i1> %m, i32 zeroext %e
 ;
 ; ZVFHMIN-LABEL: vp_ceil_v16f16:
 ; ZVFHMIN:       # %bb.0:
-; ZVFHMIN-NEXT:    vmv1r.v v10, v0
 ; ZVFHMIN-NEXT:    vsetivli zero, 16, e16, m2, ta, ma
+; ZVFHMIN-NEXT:    vmv1r.v v10, v0
 ; ZVFHMIN-NEXT:    vfwcvt.f.f.v v12, v8
 ; ZVFHMIN-NEXT:    lui a1, 307200
 ; ZVFHMIN-NEXT:    vsetvli zero, a0, e32, m4, ta, ma
@@ -431,8 +431,8 @@ declare <8 x float> @llvm.vp.ceil.v8f32(<8 x float>, <8 x i1>, i32)
 define <8 x float> @vp_ceil_v8f32(<8 x float> %va, <8 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vp_ceil_v8f32:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e32, m2, ta, ma
+; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vfabs.v v12, v8, v0.t
 ; CHECK-NEXT:    lui a0, 307200
 ; CHECK-NEXT:    fmv.w.x fa5, a0
@@ -475,8 +475,8 @@ declare <16 x float> @llvm.vp.ceil.v16f32(<16 x float>, <16 x i1>, i32)
 define <16 x float> @vp_ceil_v16f32(<16 x float> %va, <16 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vp_ceil_v16f32:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v12, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e32, m4, ta, ma
+; CHECK-NEXT:    vmv1r.v v12, v0
 ; CHECK-NEXT:    vfabs.v v16, v8, v0.t
 ; CHECK-NEXT:    lui a0, 307200
 ; CHECK-NEXT:    fmv.w.x fa5, a0
@@ -561,10 +561,10 @@ declare <4 x double> @llvm.vp.ceil.v4f64(<4 x double>, <4 x i1>, i32)
 define <4 x double> @vp_ceil_v4f64(<4 x double> %va, <4 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vp_ceil_v4f64:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v10, v0
-; CHECK-NEXT:    lui a1, %hi(.LCPI18_0)
-; CHECK-NEXT:    fld fa5, %lo(.LCPI18_0)(a1)
 ; CHECK-NEXT:    vsetvli zero, a0, e64, m2, ta, ma
+; CHECK-NEXT:    vmv1r.v v10, v0
+; CHECK-NEXT:    lui a0, %hi(.LCPI18_0)
+; CHECK-NEXT:    fld fa5, %lo(.LCPI18_0)(a0)
 ; CHECK-NEXT:    vfabs.v v12, v8, v0.t
 ; CHECK-NEXT:    vsetvli zero, zero, e64, m2, ta, mu
 ; CHECK-NEXT:    vmflt.vf v10, v12, fa5, v0.t
@@ -605,10 +605,10 @@ declare <8 x double> @llvm.vp.ceil.v8f64(<8 x double>, <8 x i1>, i32)
 define <8 x double> @vp_ceil_v8f64(<8 x double> %va, <8 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vp_ceil_v8f64:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v12, v0
-; CHECK-NEXT:    lui a1, %hi(.LCPI20_0)
-; CHECK-NEXT:    fld fa5, %lo(.LCPI20_0)(a1)
 ; CHECK-NEXT:    vsetvli zero, a0, e64, m4, ta, ma
+; CHECK-NEXT:    vmv1r.v v12, v0
+; CHECK-NEXT:    lui a0, %hi(.LCPI20_0)
+; CHECK-NEXT:    fld fa5, %lo(.LCPI20_0)(a0)
 ; CHECK-NEXT:    vfabs.v v16, v8, v0.t
 ; CHECK-NEXT:    vsetvli zero, zero, e64, m4, ta, mu
 ; CHECK-NEXT:    vmflt.vf v12, v16, fa5, v0.t
@@ -649,10 +649,10 @@ declare <15 x double> @llvm.vp.ceil.v15f64(<15 x double>, <15 x i1>, i32)
 define <15 x double> @vp_ceil_v15f64(<15 x double> %va, <15 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vp_ceil_v15f64:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v16, v0
-; CHECK-NEXT:    lui a1, %hi(.LCPI22_0)
-; CHECK-NEXT:    fld fa5, %lo(.LCPI22_0)(a1)
 ; CHECK-NEXT:    vsetvli zero, a0, e64, m8, ta, ma
+; CHECK-NEXT:    vmv1r.v v16, v0
+; CHECK-NEXT:    lui a0, %hi(.LCPI22_0)
+; CHECK-NEXT:    fld fa5, %lo(.LCPI22_0)(a0)
 ; CHECK-NEXT:    vfabs.v v24, v8, v0.t
 ; CHECK-NEXT:    vsetvli zero, zero, e64, m8, ta, mu
 ; CHECK-NEXT:    vmflt.vf v16, v24, fa5, v0.t
@@ -693,10 +693,10 @@ declare <16 x double> @llvm.vp.ceil.v16f64(<16 x double>, <16 x i1>, i32)
 define <16 x double> @vp_ceil_v16f64(<16 x double> %va, <16 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vp_ceil_v16f64:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v16, v0
-; CHECK-NEXT:    lui a1, %hi(.LCPI24_0)
-; CHECK-NEXT:    fld fa5, %lo(.LCPI24_0)(a1)
 ; CHECK-NEXT:    vsetvli zero, a0, e64, m8, ta, ma
+; CHECK-NEXT:    vmv1r.v v16, v0
+; CHECK-NEXT:    lui a0, %hi(.LCPI24_0)
+; CHECK-NEXT:    fld fa5, %lo(.LCPI24_0)(a0)
 ; CHECK-NEXT:    vfabs.v v24, v8, v0.t
 ; CHECK-NEXT:    vsetvli zero, zero, e64, m8, ta, mu
 ; CHECK-NEXT:    vmflt.vf v16, v24, fa5, v0.t
@@ -743,6 +743,7 @@ define <32 x double> @vp_ceil_v32f64(<32 x double> %va, <32 x i1> %m, i32 zeroex
 ; CHECK-NEXT:    slli a1, a1, 4
 ; CHECK-NEXT:    sub sp, sp, a1
 ; CHECK-NEXT:    .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x10, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 16 * vlenb
+; CHECK-NEXT:    vsetivli zero, 2, e8, mf4, ta, ma
 ; CHECK-NEXT:    vmv1r.v v25, v0
 ; CHECK-NEXT:    csrr a1, vlenb
 ; CHECK-NEXT:    slli a1, a1, 3
@@ -750,7 +751,6 @@ define <32 x double> @vp_ceil_v32f64(<32 x double> %va, <32 x i1> %m, i32 zeroex
 ; CHECK-NEXT:    addi a1, a1, 16
 ; CHECK-NEXT:    vs8r.v v16, (a1) # Unknown-size Folded Spill
 ; CHECK-NEXT:    li a2, 16
-; CHECK-NEXT:    vsetivli zero, 2, e8, mf4, ta, ma
 ; CHECK-NEXT:    vslidedown.vi v24, v0, 2
 ; CHECK-NEXT:    mv a1, a0
 ; CHECK-NEXT:    bltu a0, a2, .LBB26_2
diff --git a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-ctpop-vp.ll b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-ctpop-vp.ll
index 5e73e6df9170c..a5a1061842427 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-ctpop-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-ctpop-vp.ll
@@ -1796,6 +1796,7 @@ define <32 x i64> @vp_ctpop_v32i64_unmasked(<32 x i64> %va, i32 zeroext %evl) {
 ; RV32-NEXT:    mul a1, a1, a2
 ; RV32-NEXT:    sub sp, sp, a1
 ; RV32-NEXT:    .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x30, 0x22, 0x11, 0x18, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 48 + 24 * vlenb
+; RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; RV32-NEXT:    vmv8r.v v24, v16
 ; RV32-NEXT:    lui a1, 349525
 ; RV32-NEXT:    lui a2, 209715
diff --git a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-floor-vp.ll b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-floor-vp.ll
index 02e99ea513e69..49255320c40a6 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-floor-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-floor-vp.ll
@@ -194,8 +194,8 @@ define <8 x half> @vp_floor_v8f16(<8 x half> %va, <8 x i1> %m, i32 zeroext %evl)
 ;
 ; ZVFHMIN-LABEL: vp_floor_v8f16:
 ; ZVFHMIN:       # %bb.0:
-; ZVFHMIN-NEXT:    vmv1r.v v9, v0
 ; ZVFHMIN-NEXT:    vsetivli zero, 8, e16, m1, ta, ma
+; ZVFHMIN-NEXT:    vmv1r.v v9, v0
 ; ZVFHMIN-NEXT:    vfwcvt.f.f.v v10, v8
 ; ZVFHMIN-NEXT:    lui a1, 307200
 ; ZVFHMIN-NEXT:    vsetvli zero, a0, e32, m2, ta, ma
@@ -261,10 +261,10 @@ declare <16 x half> @llvm.vp.floor.v16f16(<16 x half>, <16 x i1>, i32)
 define <16 x half> @vp_floor_v16f16(<16 x half> %va, <16 x i1> %m, i32 zeroext %evl) {
 ; ZVFH-LABEL: vp_floor_v16f16:
 ; ZVFH:       # %bb.0:
-; ZVFH-NEXT:    vmv1r.v v10, v0
-; ZVFH-NEXT:    lui a1, %hi(.LCPI6_0)
-; ZVFH-NEXT:    flh fa5, %lo(.LCPI6_0)(a1)
 ; ZVFH-NEXT:    vsetvli zero, a0, e16, m2, ta, ma
+; ZVFH-NEXT:    vmv1r.v v10, v0
+; ZVFH-NEXT:    lui a0, %hi(.LCPI6_0)
+; ZVFH-NEXT:    flh fa5, %lo(.LCPI6_0)(a0)
 ; ZVFH-NEXT:    vfabs.v v12, v8, v0.t
 ; ZVFH-NEXT:    vsetvli zero, zero, e16, m2, ta, mu
 ; ZVFH-NEXT:    vmflt.vf v10, v12, fa5, v0.t
@@ -280,8 +280,8 @@ define <16 x half> @vp_floor_v16f16(<16 x half> %va, <16 x i1> %m, i32 zeroext %
 ;
 ; ZVFHMIN-LABEL: vp_floor_v16f16:
 ; ZVFHMIN:       # %bb.0:
-; ZVFHMIN-NEXT:    vmv1r.v v10, v0
 ; ZVFHMIN-NEXT:    vsetivli zero, 16, e16, m2, ta, ma
+; ZVFHMIN-NEXT:    vmv1r.v v10, v0
 ; ZVFHMIN-NEXT:    vfwcvt.f.f.v v12, v8
 ; ZVFHMIN-NEXT:    lui a1, 307200
 ; ZVFHMIN-NEXT:    vsetvli zero, a0, e32, m4, ta, ma
@@ -431,8 +431,8 @@ declare <8 x float> @llvm.vp.floor.v8f32(<8 x float>, <8 x i1>, i32)
 define <8 x float> @vp_floor_v8f32(<8 x float> %va, <8 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vp_floor_v8f32:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e32, m2, ta, ma
+; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vfabs.v v12, v8, v0.t
 ; CHECK-NEXT:    lui a0, 307200
 ; CHECK-NEXT:    fmv.w.x fa5, a0
@@ -475,8 +475,8 @@ declare <16 x float> @llvm.vp.floor.v16f32(<16 x float>, <16 x i1>, i32)
 define <16 x float> @vp_floor_v16f32(<16 x float> %va, <16 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vp_floor_v16f32:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v12, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e32, m4, ta, ma
+; CHECK-NEXT:    vmv1r.v v12, v0
 ; CHECK-NEXT:    vfabs.v v16, v8, v0.t
 ; CHECK-NEXT:    lui a0, 307200
 ; CHECK-NEXT:    fmv.w.x fa5, a0
@@ -561,10 +561,10 @@ declare <4 x double> @llvm.vp.floor.v4f64(<4 x double>, <4 x i1>, i32)
 define <4 x double> @vp_floor_v4f64(<4 x double> %va, <4 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vp_floor_v4f64:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v10, v0
-; CHECK-NEXT:    lui a1, %hi(.LCPI18_0)
-; CHECK-NEXT:    fld fa5, %lo(.LCPI18_0)(a1)
 ; CHECK-NEXT:    vsetvli zero, a0, e64, m2, ta, ma
+; CHECK-NEXT:    vmv1r.v v10, v0
+; CHECK-NEXT:    lui a0, %hi(.LCPI18_0)
+; CHECK-NEXT:    fld fa5, %lo(.LCPI18_0)(a0)
 ; CHECK-NEXT:    vfabs.v v12, v8, v0.t
 ; CHECK-NEXT:    vsetvli zero, zero, e64, m2, ta, mu
 ; CHECK-NEXT:    vmflt.vf v10, v12, fa5, v0.t
@@ -605,10 +605,10 @@ declare <8 x double> @llvm.vp.floor.v8f64(<8 x double>, <8 x i1>, i32)
 define <8 x double> @vp_floor_v8f64(<8 x double> %va, <8 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vp_floor_v8f64:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v12, v0
-; CHECK-NEXT:    lui a1, %hi(.LCPI20_0)
-; CHECK-NEXT:    fld fa5, %lo(.LCPI20_0)(a1)
 ; CHECK-NEXT:    vsetvli zero, a0, e64, m4, ta, ma
+; CHECK-NEXT:    vmv1r.v v12, v0
+; CHECK-NEXT:    lui a0, %hi(.LCPI20_0)
+; CHECK-NEXT:    fld fa5, %lo(.LCPI20_0)(a0)
 ; CHECK-NEXT:    vfabs.v v16, v8, v0.t
 ; CHECK-NEXT:    vsetvli zero, zero, e64, m4, ta, mu
 ; CHECK-NEXT:    vmflt.vf v12, v16, fa5, v0.t
@@ -649,10 +649,10 @@ declare <15 x double> @llvm.vp.floor.v15f64(<15 x double>, <15 x i1>, i32)
 define <15 x double> @vp_floor_v15f64(<15 x double> %va, <15 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vp_floor_v15f64:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v16, v0
-; CHECK-NEXT:    lui a1, %hi(.LCPI22_0)
-; CHECK-NEXT:    fld fa5, %lo(.LCPI22_0)(a1)
 ; CHECK-NEXT:    vsetvli zero, a0, e64, m8, ta, ma
+; CHECK-NEXT:    vmv1r.v v16, v0
+; CHECK-NEXT:    lui a0, %hi(.LCPI22_0)
+; CHECK-NEXT:    fld fa5, %lo(.LCPI22_0)(a0)
 ; CHECK-NEXT:    vfabs.v v24, v8, v0.t
 ; CHECK-NEXT:    vsetvli zero, zero, e64, m8, ta, mu
 ; CHECK-NEXT:    vmflt.vf v16, v24, fa5, v0.t
@@ -693,10 +693,10 @@ declare <16 x double> @llvm.vp.floor.v16f64(<16 x double>, <16 x i1>, i32)
 define <16 x double> @vp_floor_v16f64(<16 x double> %va, <16 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vp_floor_v16f64:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v16, v0
-; CHECK-NEXT:    lui a1, %hi(.LCPI24_0)
-; CHECK-NEXT:    fld fa5, %lo(.LCPI24_0)(a1)
 ; CHECK-NEXT:    vsetvli zero, a0, e64, m8, ta, ma
+; CHECK-NEXT:    vmv1r.v v16, v0
+; CHECK-NEXT:    lui a0, %hi(.LCPI24_0)
+; CHECK-NEXT:    fld fa5, %lo(.LCPI24_0)(a0)
 ; CHECK-NEXT:    vfabs.v v24, v8, v0.t
 ; CHECK-NEXT:    vsetvli zero, zero, e64, m8, ta, mu
 ; CHECK-NEXT:    vmflt.vf v16, v24, fa5, v0.t
@@ -743,6 +743,7 @@ define <32 x double> @vp_floor_v32f64(<32 x double> %va, <32 x i1> %m, i32 zeroe
 ; CHECK-NEXT:    slli a1, a1, 4
 ; CHECK-NEXT:    sub sp, sp, a1
 ; CHECK-NEXT:    .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x10, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 16 * vlenb
+; CHECK-NEXT:    vsetivli zero, 2, e8, mf4, ta, ma
 ; CHECK-NEXT:    vmv1r.v v25, v0
 ; CHECK-NEXT:    csrr a1, vlenb
 ; CHECK-NEXT:    slli a1, a1, 3
@@ -750,7 +751,6 @@ define <32 x double> @vp_floor_v32f64(<32 x double> %va, <32 x i1> %m, i32 zeroe
 ; CHECK-NEXT:    addi a1, a1, 16
 ; CHECK-NEXT:    vs8r.v v16, (a1) # Unknown-size Folded Spill
 ; CHECK-NEXT:    li a2, 16
-; CHECK-NEXT:    vsetivli zero, 2, e8, mf4, ta, ma
 ; CHECK-NEXT:    vslidedown.vi v24, v0, 2
 ; CHECK-NEXT:    mv a1, a0
 ; CHECK-NEXT:    bltu a0, a2, .LBB26_2
diff --git a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-fmaximum-vp.ll b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-fmaximum-vp.ll
index f43934afc370d..11f92555f56cd 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-fmaximum-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-fmaximum-vp.ll
@@ -13,8 +13,8 @@ declare <2 x half> @llvm.vp.maximum.v2f16(<2 x half>, <2 x half>, <2 x i1>, i32)
 define <2 x half> @vfmax_vv_v2f16(<2 x half> %va, <2 x half> %vb, <2 x i1> %m, i32 zeroext %evl) {
 ; ZVFH-LABEL: vfmax_vv_v2f16:
 ; ZVFH:       # %bb.0:
-; ZVFH-NEXT:    vmv1r.v v10, v0
 ; ZVFH-NEXT:    vsetvli zero, a0, e16, mf4, ta, ma
+; ZVFH-NEXT:    vmv1r.v v10, v0
 ; ZVFH-NEXT:    vmfeq.vv v0, v8, v8, v0.t
 ; ZVFH-NEXT:    vmerge.vvm v11, v8, v9, v0
 ; ZVFH-NEXT:    vmv1r.v v0, v10
@@ -26,8 +26,8 @@ define <2 x half> @vfmax_vv_v2f16(<2 x half> %va, <2 x half> %vb, <2 x i1> %m, i
 ;
 ; ZVFHMIN-LABEL: vfmax_vv_v2f16:
 ; ZVFHMIN:       # %bb.0:
-; ZVFHMIN-NEXT:    vmv1r.v v10, v0
 ; ZVFHMIN-NEXT:    vsetivli zero, 2, e16, mf4, ta, ma
+; ZVFHMIN-NEXT:    vmv1r.v v10, v0
 ; ZVFHMIN-NEXT:    vfwcvt.f.f.v v11, v8
 ; ZVFHMIN-NEXT:    vsetvli zero, a0, e32, mf2, ta, ma
 ; ZVFHMIN-NEXT:    vmfeq.vv v0, v11, v11, v0.t
@@ -83,8 +83,8 @@ declare <4 x half> @llvm.vp.maximum.v4f16(<4 x half>, <4 x half>, <4 x i1>, i32)
 define <4 x half> @vfmax_vv_v4f16(<4 x half> %va, <4 x half> %vb, <4 x i1> %m, i32 zeroext %evl) {
 ; ZVFH-LABEL: vfmax_vv_v4f16:
 ; ZVFH:       # %bb.0:
-; ZVFH-NEXT:    vmv1r.v v10, v0
 ; ZVFH-NEXT:    vsetvli zero, a0, e16, mf2, ta, ma
+; ZVFH-NEXT:    vmv1r.v v10, v0
 ; ZVFH-NEXT:    vmfeq.vv v0, v8, v8, v0.t
 ; ZVFH-NEXT:    vmerge.vvm v11, v8, v9, v0
 ; ZVFH-NEXT:    vmv1r.v v0, v10
@@ -96,8 +96,8 @@ define <4 x half> @vfmax_vv_v4f16(<4 x half> %va, <4 x half> %vb, <4 x i1> %m, i
 ;
 ; ZVFHMIN-LABEL: vfmax_vv_v4f16:
 ; ZVFHMIN:       # %bb.0:
-; ZVFHMIN-NEXT:    vmv1r.v v10, v0
 ; ZVFHMIN-NEXT:    vsetivli zero, 4, e16, mf2, ta, ma
+; ZVFHMIN-NEXT:    vmv1r.v v10, v0
 ; ZVFHMIN-NEXT:    vfwcvt.f.f.v v11, v8
 ; ZVFHMIN-NEXT:    vsetvli zero, a0, e32, m1, ta, ma
 ; ZVFHMIN-NEXT:    vmfeq.vv v0, v11, v11, v0.t
@@ -153,8 +153,8 @@ declare <8 x half> @llvm.vp.maximum.v8f16(<8 x half>, <8 x half>, <8 x i1>, i32)
 define <8 x half> @vfmax_vv_v8f16(<8 x half> %va, <8 x half> %vb, <8 x i1> %m, i32 zeroext %evl) {
 ; ZVFH-LABEL: vfmax_vv_v8f16:
 ; ZVFH:       # %bb.0:
-; ZVFH-NEXT:    vmv1r.v v10, v0
 ; ZVFH-NEXT:    vsetvli zero, a0, e16, m1, ta, ma
+; ZVFH-NEXT:    vmv1r.v v10, v0
 ; ZVFH-NEXT:    vmfeq.vv v0, v8, v8, v0.t
 ; ZVFH-NEXT:    vmerge.vvm v11, v8, v9, v0
 ; ZVFH-NEXT:    vmv1r.v v0, v10
@@ -166,8 +166,8 @@ define <8 x half> @vfmax_vv_v8f16(<8 x half> %va, <8 x half> %vb, <8 x i1> %m, i
 ;
 ; ZVFHMIN-LABEL: vfmax_vv_v8f16:
 ; ZVFHMIN:       # %bb.0:
-; ZVFHMIN-NEXT:    vmv1r.v v10, v0
 ; ZVFHMIN-NEXT:    vsetivli zero, 8, e16, m1, ta, ma
+; ZVFHMIN-NEXT:    vmv1r.v v10, v0
 ; ZVFHMIN-NEXT:    vfwcvt.f.f.v v12, v8
 ; ZVFHMIN-NEXT:    vsetvli zero, a0, e32, m2, ta, ma
 ; ZVFHMIN-NEXT:    vmfeq.vv v8, v12, v12, v0.t
@@ -225,8 +225,8 @@ declare <16 x half> @llvm.vp.maximum.v16f16(<16 x half>, <16 x half>, <16 x i1>,
 define <16 x half> @vfmax_vv_v16f16(<16 x half> %va, <16 x half> %vb, <16 x i1> %m, i32 zeroext %evl) {
 ; ZVFH-LABEL: vfmax_vv_v16f16:
 ; ZVFH:       # %bb.0:
-; ZVFH-NEXT:    vmv1r.v v12, v0
 ; ZVFH-NEXT:    vsetvli zero, a0, e16, m2, ta, ma
+; ZVFH-NEXT:    vmv1r.v v12, v0
 ; ZVFH-NEXT:    vmfeq.vv v13, v8, v8, v0.t
 ; ZVFH-NEXT:    vmv1r.v v0, v13
 ; ZVFH-NEXT:    vmerge.vvm v14, v8, v10, v0
@@ -240,8 +240,8 @@ define <16 x half> @vfmax_vv_v16f16(<16 x half> %va, <16 x half> %vb, <16 x i1>
 ;
 ; ZVFHMIN-LABEL: vfmax_vv_v16f16:
 ; ZVFHMIN:       # %bb.0:
-; ZVFHMIN-NEXT:    vmv1r.v v12, v0
 ; ZVFHMIN-NEXT:    vsetivli zero, 16, e16, m2, ta, ma
+; ZVFHMIN-NEXT:    vmv1r.v v12, v0
 ; ZVFHMIN-NEXT:    vfwcvt.f.f.v v16, v8
 ; ZVFHMIN-NEXT:    vsetvli zero, a0, e32, m4, ta, ma
 ; ZVFHMIN-NEXT:    vmfeq.vv v8, v16, v16, v0.t
@@ -299,8 +299,8 @@ declare <2 x float> @llvm.vp.maximum.v2f32(<2 x float>, <2 x float>, <2 x i1>, i
 define <2 x float> @vfmax_vv_v2f32(<2 x float> %va, <2 x float> %vb, <2 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vfmax_vv_v2f32:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e32, mf2, ta, ma
+; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmfeq.vv v0, v8, v8, v0.t
 ; CHECK-NEXT:    vmerge.vvm v11, v8, v9, v0
 ; CHECK-NEXT:    vmv1r.v v0, v10
@@ -332,8 +332,8 @@ declare <4 x float> @llvm.vp.maximum.v4f32(<4 x float>, <4 x float>, <4 x i1>, i
 define <4 x float> @vfmax_vv_v4f32(<4 x float> %va, <4 x float> %vb, <4 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vfmax_vv_v4f32:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e32, m1, ta, ma
+; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmfeq.vv v0, v8, v8, v0.t
 ; CHECK-NEXT:    vmerge.vvm v11, v8, v9, v0
 ; CHECK-NEXT:    vmv1r.v v0, v10
@@ -365,8 +365,8 @@ declare <8 x float> @llvm.vp.maximum.v8f32(<8 x float>, <8 x float>, <8 x i1>, i
 define <8 x float> @vfmax_vv_v8f32(<8 x float> %va, <8 x float> %vb, <8 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vfmax_vv_v8f32:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v12, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e32, m2, ta, ma
+; CHECK-NEXT:    vmv1r.v v12, v0
 ; CHECK-NEXT:    vmfeq.vv v13, v8, v8, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v13
 ; CHECK-NEXT:    vmerge.vvm v14, v8, v10, v0
@@ -400,8 +400,8 @@ declare <16 x float> @llvm.vp.maximum.v16f32(<16 x float>, <16 x float>, <16 x i
 define <16 x float> @vfmax_vv_v16f32(<16 x float> %va, <16 x float> %vb, <16 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vfmax_vv_v16f32:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v16, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e32, m4, ta, ma
+; CHECK-NEXT:    vmv1r.v v16, v0
 ; CHECK-NEXT:    vmfeq.vv v17, v8, v8, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v17
 ; CHECK-NEXT:    vmerge.vvm v20, v8, v12, v0
@@ -435,8 +435,8 @@ declare <2 x double> @llvm.vp.maximum.v2f64(<2 x double>, <2 x double>, <2 x i1>
 define <2 x double> @vfmax_vv_v2f64(<2 x double> %va, <2 x double> %vb, <2 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vfmax_vv_v2f64:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e64, m1, ta, ma
+; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmfeq.vv v0, v8, v8, v0.t
 ; CHECK-NEXT:    vmerge.vvm v11, v8, v9, v0
 ; CHECK-NEXT:    vmv1r.v v0, v10
@@ -468,8 +468,8 @@ declare <4 x double> @llvm.vp.maximum.v4f64(<4 x double>, <4 x double>, <4 x i1>
 define <4 x double> @vfmax_vv_v4f64(<4 x double> %va, <4 x double> %vb, <4 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vfmax_vv_v4f64:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v12, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e64, m2, ta, ma
+; CHECK-NEXT:    vmv1r.v v12, v0
 ; CHECK-NEXT:    vmfeq.vv v13, v8, v8, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v13
 ; CHECK-NEXT:    vmerge.vvm v14, v8, v10, v0
@@ -503,8 +503,8 @@ declare <8 x double> @llvm.vp.maximum.v8f64(<8 x double>, <8 x double>, <8 x i1>
 define <8 x double> @vfmax_vv_v8f64(<8 x double> %va, <8 x double> %vb, <8 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vfmax_vv_v8f64:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v16, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e64, m4, ta, ma
+; CHECK-NEXT:    vmv1r.v v16, v0
 ; CHECK-NEXT:    vmfeq.vv v17, v8, v8, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v17
 ; CHECK-NEXT:    vmerge.vvm v20, v8, v12, v0
@@ -544,8 +544,8 @@ define <16 x double> @vfmax_vv_v16f64(<16 x double> %va, <16 x double> %vb, <16
 ; CHECK-NEXT:    slli a1, a1, 3
 ; CHECK-NEXT:    sub sp, sp, a1
 ; CHECK-NEXT:    .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x08, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 8 * vlenb
-; CHECK-NEXT:    vmv1r.v v7, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e64, m8, ta, ma
+; CHECK-NEXT:    vmv1r.v v7, v0
 ; CHECK-NEXT:    vmfeq.vv v25, v8, v8, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v25
 ; CHECK-NEXT:    vmerge.vvm v24, v8, v16, v0
@@ -595,6 +595,7 @@ define <32 x double> @vfmax_vv_v32f64(<32 x double> %va, <32 x double> %vb, <32
 ; CHECK-NEXT:    slli a1, a1, 5
 ; CHECK-NEXT:    sub sp, sp, a1
 ; CHECK-NEXT:    .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x20, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 32 * vlenb
+; CHECK-NEXT:    vsetivli zero, 16, e64, m8, ta, ma
 ; CHECK-NEXT:    vmv1r.v v25, v0
 ; CHECK-NEXT:    csrr a1, vlenb
 ; CHECK-NEXT:    slli a1, a1, 4
@@ -608,7 +609,6 @@ define <32 x double> @vfmax_vv_v32f64(<32 x double> %va, <32 x double> %vb, <32
 ; CHECK-NEXT:    addi a1, a1, 16
 ; CHECK-NEXT:    vs8r.v v8, (a1) # Unknown-size Folded Spill
 ; CHECK-NEXT:    addi a1, a0, 128
-; CHECK-NEXT:    vsetivli zero, 16, e64, m8, ta, ma
 ; CHECK-NEXT:    vle64.v v16, (a1)
 ; CHECK-NEXT:    csrr a1, vlenb
 ; CHECK-NEXT:    slli a1, a1, 3
diff --git a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-fminimum-vp.ll b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-fminimum-vp.ll
index 7067cc21ab56d..3fb586b67a21b 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-fminimum-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-fminimum-vp.ll
@@ -13,8 +13,8 @@ declare <2 x half> @llvm.vp.minimum.v2f16(<2 x half>, <2 x half>, <2 x i1>, i32)
 define <2 x half> @vfmin_vv_v2f16(<2 x half> %va, <2 x half> %vb, <2 x i1> %m, i32 zeroext %evl) {
 ; ZVFH-LABEL: vfmin_vv_v2f16:
 ; ZVFH:       # %bb.0:
-; ZVFH-NEXT:    vmv1r.v v10, v0
 ; ZVFH-NEXT:    vsetvli zero, a0, e16, mf4, ta, ma
+; ZVFH-NEXT:    vmv1r.v v10, v0
 ; ZVFH-NEXT:    vmfeq.vv v0, v8, v8, v0.t
 ; ZVFH-NEXT:    vmerge.vvm v11, v8, v9, v0
 ; ZVFH-NEXT:    vmv1r.v v0, v10
@@ -26,8 +26,8 @@ define <2 x half> @vfmin_vv_v2f16(<2 x half> %va, <2 x half> %vb, <2 x i1> %m, i
 ;
 ; ZVFHMIN-LABEL: vfmin_vv_v2f16:
 ; ZVFHMIN:       # %bb.0:
-; ZVFHMIN-NEXT:    vmv1r.v v10, v0
 ; ZVFHMIN-NEXT:    vsetivli zero, 2, e16, mf4, ta, ma
+; ZVFHMIN-NEXT:    vmv1r.v v10, v0
 ; ZVFHMIN-NEXT:    vfwcvt.f.f.v v11, v8
 ; ZVFHMIN-NEXT:    vsetvli zero, a0, e32, mf2, ta, ma
 ; ZVFHMIN-NEXT:    vmfeq.vv v0, v11, v11, v0.t
@@ -83,8 +83,8 @@ declare <4 x half> @llvm.vp.minimum.v4f16(<4 x half>, <4 x half>, <4 x i1>, i32)
 define <4 x half> @vfmin_vv_v4f16(<4 x half> %va, <4 x half> %vb, <4 x i1> %m, i32 zeroext %evl) {
 ; ZVFH-LABEL: vfmin_vv_v4f16:
 ; ZVFH:       # %bb.0:
-; ZVFH-NEXT:    vmv1r.v v10, v0
 ; ZVFH-NEXT:    vsetvli zero, a0, e16, mf2, ta, ma
+; ZVFH-NEXT:    vmv1r.v v10, v0
 ; ZVFH-NEXT:    vmfeq.vv v0, v8, v8, v0.t
 ; ZVFH-NEXT:    vmerge.vvm v11, v8, v9, v0
 ; ZVFH-NEXT:    vmv1r.v v0, v10
@@ -96,8 +96,8 @@ define <4 x half> @vfmin_vv_v4f16(<4 x half> %va, <4 x half> %vb, <4 x i1> %m, i
 ;
 ; ZVFHMIN-LABEL: vfmin_vv_v4f16:
 ; ZVFHMIN:       # %bb.0:
-; ZVFHMIN-NEXT:    vmv1r.v v10, v0
 ; ZVFHMIN-NEXT:    vsetivli zero, 4, e16, mf2, ta, ma
+; ZVFHMIN-NEXT:    vmv1r.v v10, v0
 ; ZVFHMIN-NEXT:    vfwcvt.f.f.v v11, v8
 ; ZVFHMIN-NEXT:    vsetvli zero, a0, e32, m1, ta, ma
 ; ZVFHMIN-NEXT:    vmfeq.vv v0, v11, v11, v0.t
@@ -153,8 +153,8 @@ declare <8 x half> @llvm.vp.minimum.v8f16(<8 x half>, <8 x half>, <8 x i1>, i32)
 define <8 x half> @vfmin_vv_v8f16(<8 x half> %va, <8 x half> %vb, <8 x i1> %m, i32 zeroext %evl) {
 ; ZVFH-LABEL: vfmin_vv_v8f16:
 ; ZVFH:       # %bb.0:
-; ZVFH-NEXT:    vmv1r.v v10, v0
 ; ZVFH-NEXT:    vsetvli zero, a0, e16, m1, ta, ma
+; ZVFH-NEXT:    vmv1r.v v10, v0
 ; ZVFH-NEXT:    vmfeq.vv v0, v8, v8, v0.t
 ; ZVFH-NEXT:    vmerge.vvm v11, v8, v9, v0
 ; ZVFH-NEXT:    vmv1r.v v0, v10
@@ -166,8 +166,8 @@ define <8 x half> @vfmin_vv_v8f16(<8 x half> %va, <8 x half> %vb, <8 x i1> %m, i
 ;
 ; ZVFHMIN-LABEL: vfmin_vv_v8f16:
 ; ZVFHMIN:       # %bb.0:
-; ZVFHMIN-NEXT:    vmv1r.v v10, v0
 ; ZVFHMIN-NEXT:    vsetivli zero, 8, e16, m1, ta, ma
+; ZVFHMIN-NEXT:    vmv1r.v v10, v0
 ; ZVFHMIN-NEXT:    vfwcvt.f.f.v v12, v8
 ; ZVFHMIN-NEXT:    vsetvli zero, a0, e32, m2, ta, ma
 ; ZVFHMIN-NEXT:    vmfeq.vv v8, v12, v12, v0.t
@@ -225,8 +225,8 @@ declare <16 x half> @llvm.vp.minimum.v16f16(<16 x half>, <16 x half>, <16 x i1>,
 define <16 x half> @vfmin_vv_v16f16(<16 x half> %va, <16 x half> %vb, <16 x i1> %m, i32 zeroext %evl) {
 ; ZVFH-LABEL: vfmin_vv_v16f16:
 ; ZVFH:       # %bb.0:
-; ZVFH-NEXT:    vmv1r.v v12, v0
 ; ZVFH-NEXT:    vsetvli zero, a0, e16, m2, ta, ma
+; ZVFH-NEXT:    vmv1r.v v12, v0
 ; ZVFH-NEXT:    vmfeq.vv v13, v8, v8, v0.t
 ; ZVFH-NEXT:    vmv1r.v v0, v13
 ; ZVFH-NEXT:    vmerge.vvm v14, v8, v10, v0
@@ -240,8 +240,8 @@ define <16 x half> @vfmin_vv_v16f16(<16 x half> %va, <16 x half> %vb, <16 x i1>
 ;
 ; ZVFHMIN-LABEL: vfmin_vv_v16f16:
 ; ZVFHMIN:       # %bb.0:
-; ZVFHMIN-NEXT:    vmv1r.v v12, v0
 ; ZVFHMIN-NEXT:    vsetivli zero, 16, e16, m2, ta, ma
+; ZVFHMIN-NEXT:    vmv1r.v v12, v0
 ; ZVFHMIN-NEXT:    vfwcvt.f.f.v v16, v8
 ; ZVFHMIN-NEXT:    vsetvli zero, a0, e32, m4, ta, ma
 ; ZVFHMIN-NEXT:    vmfeq.vv v8, v16, v16, v0.t
@@ -299,8 +299,8 @@ declare <2 x float> @llvm.vp.minimum.v2f32(<2 x float>, <2 x float>, <2 x i1>, i
 define <2 x float> @vfmin_vv_v2f32(<2 x float> %va, <2 x float> %vb, <2 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vfmin_vv_v2f32:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e32, mf2, ta, ma
+; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmfeq.vv v0, v8, v8, v0.t
 ; CHECK-NEXT:    vmerge.vvm v11, v8, v9, v0
 ; CHECK-NEXT:    vmv1r.v v0, v10
@@ -332,8 +332,8 @@ declare <4 x float> @llvm.vp.minimum.v4f32(<4 x float>, <4 x float>, <4 x i1>, i
 define <4 x float> @vfmin_vv_v4f32(<4 x float> %va, <4 x float> %vb, <4 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vfmin_vv_v4f32:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e32, m1, ta, ma
+; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmfeq.vv v0, v8, v8, v0.t
 ; CHECK-NEXT:    vmerge.vvm v11, v8, v9, v0
 ; CHECK-NEXT:    vmv1r.v v0, v10
@@ -365,8 +365,8 @@ declare <8 x float> @llvm.vp.minimum.v8f32(<8 x float>, <8 x float>, <8 x i1>, i
 define <8 x float> @vfmin_vv_v8f32(<8 x float> %va, <8 x float> %vb, <8 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vfmin_vv_v8f32:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v12, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e32, m2, ta, ma
+; CHECK-NEXT:    vmv1r.v v12, v0
 ; CHECK-NEXT:    vmfeq.vv v13, v8, v8, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v13
 ; CHECK-NEXT:    vmerge.vvm v14, v8, v10, v0
@@ -400,8 +400,8 @@ declare <16 x float> @llvm.vp.minimum.v16f32(<16 x float>, <16 x float>, <16 x i
 define <16 x float> @vfmin_vv_v16f32(<16 x float> %va, <16 x float> %vb, <16 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vfmin_vv_v16f32:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v16, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e32, m4, ta, ma
+; CHECK-NEXT:    vmv1r.v v16, v0
 ; CHECK-NEXT:    vmfeq.vv v17, v8, v8, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v17
 ; CHECK-NEXT:    vmerge.vvm v20, v8, v12, v0
@@ -435,8 +435,8 @@ declare <2 x double> @llvm.vp.minimum.v2f64(<2 x double>, <2 x double>, <2 x i1>
 define <2 x double> @vfmin_vv_v2f64(<2 x double> %va, <2 x double> %vb, <2 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vfmin_vv_v2f64:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e64, m1, ta, ma
+; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmfeq.vv v0, v8, v8, v0.t
 ; CHECK-NEXT:    vmerge.vvm v11, v8, v9, v0
 ; CHECK-NEXT:    vmv1r.v v0, v10
@@ -468,8 +468,8 @@ declare <4 x double> @llvm.vp.minimum.v4f64(<4 x double>, <4 x double>, <4 x i1>
 define <4 x double> @vfmin_vv_v4f64(<4 x double> %va, <4 x double> %vb, <4 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vfmin_vv_v4f64:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v12, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e64, m2, ta, ma
+; CHECK-NEXT:    vmv1r.v v12, v0
 ; CHECK-NEXT:    vmfeq.vv v13, v8, v8, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v13
 ; CHECK-NEXT:    vmerge.vvm v14, v8, v10, v0
@@ -503,8 +503,8 @@ declare <8 x double> @llvm.vp.minimum.v8f64(<8 x double>, <8 x double>, <8 x i1>
 define <8 x double> @vfmin_vv_v8f64(<8 x double> %va, <8 x double> %vb, <8 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vfmin_vv_v8f64:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v16, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e64, m4, ta, ma
+; CHECK-NEXT:    vmv1r.v v16, v0
 ; CHECK-NEXT:    vmfeq.vv v17, v8, v8, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v17
 ; CHECK-NEXT:    vmerge.vvm v20, v8, v12, v0
@@ -544,8 +544,8 @@ define <16 x double> @vfmin_vv_v16f64(<16 x double> %va, <16 x double> %vb, <16
 ; CHECK-NEXT:    slli a1, a1, 3
 ; CHECK-NEXT:    sub sp, sp, a1
 ; CHECK-NEXT:    .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x08, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 8 * vlenb
-; CHECK-NEXT:    vmv1r.v v7, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e64, m8, ta, ma
+; CHECK-NEXT:    vmv1r.v v7, v0
 ; CHECK-NEXT:    vmfeq.vv v25, v8, v8, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v25
 ; CHECK-NEXT:    vmerge.vvm v24, v8, v16, v0
@@ -595,6 +595,7 @@ define <32 x double> @vfmin_vv_v32f64(<32 x double> %va, <32 x double> %vb, <32
 ; CHECK-NEXT:    slli a1, a1, 5
 ; CHECK-NEXT:    sub sp, sp, a1
 ; CHECK-NEXT:    .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x20, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 32 * vlenb
+; CHECK-NEXT:    vsetivli zero, 16, e64, m8, ta, ma
 ; CHECK-NEXT:    vmv1r.v v25, v0
 ; CHECK-NEXT:    csrr a1, vlenb
 ; CHECK-NEXT:    slli a1, a1, 4
@@ -608,7 +609,6 @@ define <32 x double> @vfmin_vv_v32f64(<32 x double> %va, <32 x double> %vb, <32
 ; CHECK-NEXT:    addi a1, a1, 16
 ; CHECK-NEXT:    vs8r.v v8, (a1) # Unknown-size Folded Spill
 ; CHECK-NEXT:    addi a1, a0, 128
-; CHECK-NEXT:    vsetivli zero, 16, e64, m8, ta, ma
 ; CHECK-NEXT:    vle64.v v16, (a1)
 ; CHECK-NEXT:    csrr a1, vlenb
 ; CHECK-NEXT:    slli a1, a1, 3
diff --git a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-fp-interleave.ll b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-fp-interleave.ll
index 97e458e70565c..99aaecf4c6843 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-fp-interleave.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-fp-interleave.ll
@@ -38,8 +38,8 @@ define <4 x float> @interleave_v2f32(<2 x float> %x, <2 x float> %y) {
 define <4 x double> @interleave_v2f64(<2 x double> %x, <2 x double> %y) {
 ; V128-LABEL: interleave_v2f64:
 ; V128:       # %bb.0:
-; V128-NEXT:    vmv1r.v v12, v9
 ; V128-NEXT:    vsetivli zero, 4, e16, mf2, ta, ma
+; V128-NEXT:    vmv1r.v v12, v9
 ; V128-NEXT:    vid.v v9
 ; V128-NEXT:    vmv.v.i v0, 10
 ; V128-NEXT:    vsrl.vi v14, v9, 1
diff --git a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-fptrunc-vp.ll b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-fptrunc-vp.ll
index e64c7c87132ee..582706e4dfa18 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-fptrunc-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-fptrunc-vp.ll
@@ -97,9 +97,9 @@ declare <32 x float> @llvm.vp.fptrunc.v32f64.v32f32(<32 x double>, <32 x i1>, i3
 define <32 x float> @vfptrunc_v32f32_v32f64(<32 x double> %a, <32 x i1> %m, i32 zeroext %vl) {
 ; CHECK-LABEL: vfptrunc_v32f32_v32f64:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetivli zero, 2, e8, mf4, ta, ma
 ; CHECK-NEXT:    vmv8r.v v24, v8
 ; CHECK-NEXT:    li a2, 16
-; CHECK-NEXT:    vsetivli zero, 2, e8, mf4, ta, ma
 ; CHECK-NEXT:    vslidedown.vi v12, v0, 2
 ; CHECK-NEXT:    mv a1, a0
 ; CHECK-NEXT:    bltu a0, a2, .LBB7_2
diff --git a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-fshr-fshl-vp.ll b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-fshr-fshl-vp.ll
index a68dc11f3d21e..fc7cd94ca3de8 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-fshr-fshl-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-fshr-fshl-vp.ll
@@ -712,8 +712,8 @@ define <16 x i64> @fshl_v16i64(<16 x i64> %a, <16 x i64> %b, <16 x i64> %c, <16
 ; CHECK-NEXT:    .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x08, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 8 * vlenb
 ; CHECK-NEXT:    addi a2, sp, 16
 ; CHECK-NEXT:    vs8r.v v16, (a2) # Unknown-size Folded Spill
-; CHECK-NEXT:    vmv8r.v v16, v8
 ; CHECK-NEXT:    vsetivli zero, 16, e64, m8, ta, ma
+; CHECK-NEXT:    vmv8r.v v16, v8
 ; CHECK-NEXT:    vle64.v v24, (a0)
 ; CHECK-NEXT:    li a0, 63
 ; CHECK-NEXT:    vsetvli zero, a1, e64, m8, ta, ma
diff --git a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-insert-subvector.ll b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-insert-subvector.ll
index 1fbc8dfd688c4..62e7e3b109902 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-insert-subvector.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-insert-subvector.ll
@@ -133,6 +133,7 @@ define <vscale x 2 x i32> @insert_nxv8i32_v4i32_0(<vscale x 2 x i32> %vec, <4 x
 ;
 ; VLS-LABEL: insert_nxv8i32_v4i32_0:
 ; VLS:       # %bb.0:
+; VLS-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; VLS-NEXT:    vmv1r.v v8, v9
 ; VLS-NEXT:    ret
   %v = call <vscale x 2 x i32> @llvm.vector.insert.nxv2i32.v4i32(<vscale x 2 x i32> %vec, <4 x i32> %subvec, i64 0)
@@ -143,6 +144,7 @@ define <vscale x 2 x i32> @insert_nxv8i32_v4i32_0(<vscale x 2 x i32> %vec, <4 x
 define <4 x i32> @insert_v4i32_v4i32_0(<4 x i32> %vec, <4 x i32> %subvec) {
 ; CHECK-LABEL: insert_v4i32_v4i32_0:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    ret
   %v = call <4 x i32> @llvm.vector.insert.v4i32.v4i32(<4 x i32> %vec, <4 x i32> %subvec, i64 0)
diff --git a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-int-interleave.ll b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-int-interleave.ll
index a8eb1f97fd1a2..7500198f14002 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-int-interleave.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-int-interleave.ll
@@ -51,8 +51,8 @@ define <4 x i32> @interleave_v2i32(<2 x i32> %x, <2 x i32> %y) {
 define <4 x i64> @interleave_v2i64(<2 x i64> %x, <2 x i64> %y) {
 ; V128-LABEL: interleave_v2i64:
 ; V128:       # %bb.0:
-; V128-NEXT:    vmv1r.v v12, v9
 ; V128-NEXT:    vsetivli zero, 4, e16, mf2, ta, ma
+; V128-NEXT:    vmv1r.v v12, v9
 ; V128-NEXT:    vid.v v9
 ; V128-NEXT:    vmv.v.i v0, 10
 ; V128-NEXT:    vsrl.vi v14, v9, 1
diff --git a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-masked-gather.ll b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-masked-gather.ll
index 6cc3f7e76797b..141d54cf585f2 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-masked-gather.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-masked-gather.ll
@@ -556,11 +556,13 @@ define <4 x i8> @mgather_truemask_v4i8(<4 x ptr> %ptrs, <4 x i8> %passthru) {
 define <4 x i8> @mgather_falsemask_v4i8(<4 x ptr> %ptrs, <4 x i8> %passthru) {
 ; RV32-LABEL: mgather_falsemask_v4i8:
 ; RV32:       # %bb.0:
+; RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; RV32-NEXT:    vmv1r.v v8, v9
 ; RV32-NEXT:    ret
 ;
 ; RV64V-LABEL: mgather_falsemask_v4i8:
 ; RV64V:       # %bb.0:
+; RV64V-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; RV64V-NEXT:    vmv1r.v v8, v10
 ; RV64V-NEXT:    ret
 ;
@@ -733,13 +735,13 @@ define <8 x i8> @mgather_baseidx_v8i8(ptr %base, <8 x i8> %idxs, <8 x i1> %m, <8
 ; RV64ZVE32F-NEXT:    andi a2, a1, 4
 ; RV64ZVE32F-NEXT:    vsetivli zero, 2, e8, mf4, ta, ma
 ; RV64ZVE32F-NEXT:    vslidedown.vi v8, v8, 2
-; RV64ZVE32F-NEXT:    bnez a2, .LBB12_12
+; RV64ZVE32F-NEXT:    bnez a2, .LBB12_14
 ; RV64ZVE32F-NEXT:  # %bb.5: # %else5
 ; RV64ZVE32F-NEXT:    andi a2, a1, 8
-; RV64ZVE32F-NEXT:    bnez a2, .LBB12_13
+; RV64ZVE32F-NEXT:    bnez a2, .LBB12_15
 ; RV64ZVE32F-NEXT:  .LBB12_6: # %else8
 ; RV64ZVE32F-NEXT:    andi a2, a1, 16
-; RV64ZVE32F-NEXT:    bnez a2, .LBB12_14
+; RV64ZVE32F-NEXT:    bnez a2, .LBB12_16
 ; RV64ZVE32F-NEXT:  .LBB12_7: # %else11
 ; RV64ZVE32F-NEXT:    andi a2, a1, 32
 ; RV64ZVE32F-NEXT:    beqz a2, .LBB12_9
@@ -756,14 +758,31 @@ define <8 x i8> @mgather_baseidx_v8i8(ptr %base, <8 x i8> %idxs, <8 x i1> %m, <8
 ; RV64ZVE32F-NEXT:    andi a2, a1, 64
 ; RV64ZVE32F-NEXT:    vsetivli zero, 2, e8, mf4, ta, ma
 ; RV64ZVE32F-NEXT:    vslidedown.vi v8, v10, 2
-; RV64ZVE32F-NEXT:    bnez a2, .LBB12_15
-; RV64ZVE32F-NEXT:  # %bb.10: # %else17
+; RV64ZVE32F-NEXT:    beqz a2, .LBB12_11
+; RV64ZVE32F-NEXT:  # %bb.10: # %cond.load16
+; RV64ZVE32F-NEXT:    vmv.x.s a2, v8
+; RV64ZVE32F-NEXT:    add a2, a0, a2
+; RV64ZVE32F-NEXT:    lbu a2, 0(a2)
+; RV64ZVE32F-NEXT:    vmv.s.x v10, a2
+; RV64ZVE32F-NEXT:    vsetivli zero, 7, e8, mf2, tu, ma
+; RV64ZVE32F-NEXT:    vslideup.vi v9, v10, 6
+; RV64ZVE32F-NEXT:  .LBB12_11: # %else17
 ; RV64ZVE32F-NEXT:    andi a1, a1, -128
-; RV64ZVE32F-NEXT:    bnez a1, .LBB12_16
-; RV64ZVE32F-NEXT:  .LBB12_11: # %else20
+; RV64ZVE32F-NEXT:    beqz a1, .LBB12_13
+; RV64ZVE32F-NEXT:  # %bb.12: # %cond.load19
+; RV64ZVE32F-NEXT:    vsetivli zero, 1, e8, mf4, ta, ma
+; RV64ZVE32F-NEXT:    vslidedown.vi v8, v8, 1
+; RV64ZVE32F-NEXT:    vmv.x.s a1, v8
+; RV64ZVE32F-NEXT:    add a0, a0, a1
+; RV64ZVE32F-NEXT:    lbu a0, 0(a0)
+; RV64ZVE32F-NEXT:    vmv.s.x v8, a0
+; RV64ZVE32F-NEXT:    vsetivli zero, 8, e8, mf2, ta, ma
+; RV64ZVE32F-NEXT:    vslideup.vi v9, v8, 7
+; RV64ZVE32F-NEXT:  .LBB12_13: # %else20
+; RV64ZVE32F-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; RV64ZVE32F-NEXT:    vmv1r.v v8, v9
 ; RV64ZVE32F-NEXT:    ret
-; RV64ZVE32F-NEXT:  .LBB12_12: # %cond.load4
+; RV64ZVE32F-NEXT:  .LBB12_14: # %cond.load4
 ; RV64ZVE32F-NEXT:    vmv.x.s a2, v8
 ; RV64ZVE32F-NEXT:    add a2, a0, a2
 ; RV64ZVE32F-NEXT:    lbu a2, 0(a2)
@@ -772,7 +791,7 @@ define <8 x i8> @mgather_baseidx_v8i8(ptr %base, <8 x i8> %idxs, <8 x i1> %m, <8
 ; RV64ZVE32F-NEXT:    vslideup.vi v9, v11, 2
 ; RV64ZVE32F-NEXT:    andi a2, a1, 8
 ; RV64ZVE32F-NEXT:    beqz a2, .LBB12_6
-; RV64ZVE32F-NEXT:  .LBB12_13: # %cond.load7
+; RV64ZVE32F-NEXT:  .LBB12_15: # %cond.load7
 ; RV64ZVE32F-NEXT:    vsetivli zero, 1, e8, mf4, ta, ma
 ; RV64ZVE32F-NEXT:    vslidedown.vi v8, v8, 1
 ; RV64ZVE32F-NEXT:    vmv.x.s a2, v8
@@ -783,7 +802,7 @@ define <8 x i8> @mgather_baseidx_v8i8(ptr %base, <8 x i8> %idxs, <8 x i1> %m, <8
 ; RV64ZVE32F-NEXT:    vslideup.vi v9, v8, 3
 ; RV64ZVE32F-NEXT:    andi a2, a1, 16
 ; RV64ZVE32F-NEXT:    beqz a2, .LBB12_7
-; RV64ZVE32F-NEXT:  .LBB12_14: # %cond.load10
+; RV64ZVE32F-NEXT:  .LBB12_16: # %cond.load10
 ; RV64ZVE32F-NEXT:    vsetivli zero, 5, e8, mf2, tu, ma
 ; RV64ZVE32F-NEXT:    vmv.x.s a2, v10
 ; RV64ZVE32F-NEXT:    add a2, a0, a2
@@ -793,26 +812,6 @@ define <8 x i8> @mgather_baseidx_v8i8(ptr %base, <8 x i8> %idxs, <8 x i1> %m, <8
 ; RV64ZVE32F-NEXT:    andi a2, a1, 32
 ; RV64ZVE32F-NEXT:    bnez a2, .LBB12_8
 ; RV64ZVE32F-NEXT:    j .LBB12_9
-; RV64ZVE32F-NEXT:  .LBB12_15: # %cond.load16
-; RV64ZVE32F-NEXT:    vmv.x.s a2, v8
-; RV64ZVE32F-NEXT:    add a2, a0, a2
-; RV64ZVE32F-NEXT:    lbu a2, 0(a2)
-; RV64ZVE32F-NEXT:    vmv.s.x v10, a2
-; RV64ZVE32F-NEXT:    vsetivli zero, 7, e8, mf2, tu, ma
-; RV64ZVE32F-NEXT:    vslideup.vi v9, v10, 6
-; RV64ZVE32F-NEXT:    andi a1, a1, -128
-; RV64ZVE32F-NEXT:    beqz a1, .LBB12_11
-; RV64ZVE32F-NEXT:  .LBB12_16: # %cond.load19
-; RV64ZVE32F-NEXT:    vsetivli zero, 1, e8, mf4, ta, ma
-; RV64ZVE32F-NEXT:    vslidedown.vi v8, v8, 1
-; RV64ZVE32F-NEXT:    vmv.x.s a1, v8
-; RV64ZVE32F-NEXT:    add a0, a0, a1
-; RV64ZVE32F-NEXT:    lbu a0, 0(a0)
-; RV64ZVE32F-NEXT:    vmv.s.x v8, a0
-; RV64ZVE32F-NEXT:    vsetivli zero, 8, e8, mf2, ta, ma
-; RV64ZVE32F-NEXT:    vslideup.vi v9, v8, 7
-; RV64ZVE32F-NEXT:    vmv1r.v v8, v9
-; RV64ZVE32F-NEXT:    ret
   %ptrs = getelementptr inbounds i8, ptr %base, <8 x i8> %idxs
   %v = call <8 x i8> @llvm.masked.gather.v8i8.v8p0(<8 x ptr> %ptrs, i32 1, <8 x i1> %m, <8 x i8> %passthru)
   ret <8 x i8> %v
@@ -1253,11 +1252,13 @@ define <4 x i16> @mgather_truemask_v4i16(<4 x ptr> %ptrs, <4 x i16> %passthru) {
 define <4 x i16> @mgather_falsemask_v4i16(<4 x ptr> %ptrs, <4 x i16> %passthru) {
 ; RV32-LABEL: mgather_falsemask_v4i16:
 ; RV32:       # %bb.0:
+; RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; RV32-NEXT:    vmv1r.v v8, v9
 ; RV32-NEXT:    ret
 ;
 ; RV64V-LABEL: mgather_falsemask_v4i16:
 ; RV64V:       # %bb.0:
+; RV64V-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; RV64V-NEXT:    vmv1r.v v8, v10
 ; RV64V-NEXT:    ret
 ;
@@ -1435,13 +1436,13 @@ define <8 x i16> @mgather_baseidx_v8i8_v8i16(ptr %base, <8 x i8> %idxs, <8 x i1>
 ; RV64ZVE32F-NEXT:    andi a2, a1, 4
 ; RV64ZVE32F-NEXT:    vsetivli zero, 2, e8, mf4, ta, ma
 ; RV64ZVE32F-NEXT:    vslidedown.vi v8, v8, 2
-; RV64ZVE32F-NEXT:    bnez a2, .LBB23_12
+; RV64ZVE32F-NEXT:    bnez a2, .LBB23_14
 ; RV64ZVE32F-NEXT:  # %bb.5: # %else5
 ; RV64ZVE32F-NEXT:    andi a2, a1, 8
-; RV64ZVE32F-NEXT:    bnez a2, .LBB23_13
+; RV64ZVE32F-NEXT:    bnez a2, .LBB23_15
 ; RV64ZVE32F-NEXT:  .LBB23_6: # %else8
 ; RV64ZVE32F-NEXT:    andi a2, a1, 16
-; RV64ZVE32F-NEXT:    bnez a2, .LBB23_14
+; RV64ZVE32F-NEXT:    bnez a2, .LBB23_16
 ; RV64ZVE32F-NEXT:  .LBB23_7: # %else11
 ; RV64ZVE32F-NEXT:    andi a2, a1, 32
 ; RV64ZVE32F-NEXT:    beqz a2, .LBB23_9
@@ -1460,14 +1461,35 @@ define <8 x i16> @mgather_baseidx_v8i8_v8i16(ptr %base, <8 x i8> %idxs, <8 x i1>
 ; RV64ZVE32F-NEXT:    andi a2, a1, 64
 ; RV64ZVE32F-NEXT:    vsetivli zero, 2, e8, mf4, ta, ma
 ; RV64ZVE32F-NEXT:    vslidedown.vi v8, v10, 2
-; RV64ZVE32F-NEXT:    bnez a2, .LBB23_15
-; RV64ZVE32F-NEXT:  # %bb.10: # %else17
+; RV64ZVE32F-NEXT:    beqz a2, .LBB23_11
+; RV64ZVE32F-NEXT:  # %bb.10: # %cond.load16
+; RV64ZVE32F-NEXT:    vmv.x.s a2, v8
+; RV64ZVE32F-NEXT:    slli a2, a2, 1
+; RV64ZVE32F-NEXT:    add a2, a0, a2
+; RV64ZVE32F-NEXT:    lh a2, 0(a2)
+; RV64ZVE32F-NEXT:    vsetvli zero, zero, e16, mf2, ta, ma
+; RV64ZVE32F-NEXT:    vmv.s.x v10, a2
+; RV64ZVE32F-NEXT:    vsetivli zero, 7, e16, m1, tu, ma
+; RV64ZVE32F-NEXT:    vslideup.vi v9, v10, 6
+; RV64ZVE32F-NEXT:  .LBB23_11: # %else17
 ; RV64ZVE32F-NEXT:    andi a1, a1, -128
-; RV64ZVE32F-NEXT:    bnez a1, .LBB23_16
-; RV64ZVE32F-NEXT:  .LBB23_11: # %else20
+; RV64ZVE32F-NEXT:    beqz a1, .LBB23_13
+; RV64ZVE32F-NEXT:  # %bb.12: # %cond.load19
+; RV64ZVE32F-NEXT:    vsetivli zero, 1, e8, mf4, ta, ma
+; RV64ZVE32F-NEXT:    vslidedown.vi v8, v8, 1
+; RV64ZVE32F-NEXT:    vmv.x.s a1, v8
+; RV64ZVE32F-NEXT:    slli a1, a1, 1
+; RV64ZVE32F-NEXT:    add a0, a0, a1
+; RV64ZVE32F-NEXT:    lh a0, 0(a0)
+; RV64ZVE32F-NEXT:    vsetvli zero, zero, e16, mf2, ta, ma
+; RV64ZVE32F-NEXT:    vmv.s.x v8, a0
+; RV64ZVE32F-NEXT:    vsetivli zero, 8, e16, m1, ta, ma
+; RV64ZVE32F-NEXT:    vslideup.vi v9, v8, 7
+; RV64ZVE32F-NEXT:  .LBB23_13: # %else20
+; RV64ZVE32F-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; RV64ZVE32F-NEXT:    vmv1r.v v8, v9
 ; RV64ZVE32F-NEXT:    ret
-; RV64ZVE32F-NEXT:  .LBB23_12: # %cond.load4
+; RV64ZVE32F-NEXT:  .LBB23_14: # %cond.load4
 ; RV64ZVE32F-NEXT:    vmv.x.s a2, v8
 ; RV64ZVE32F-NEXT:    slli a2, a2, 1
 ; RV64ZVE32F-NEXT:    add a2, a0, a2
@@ -1478,7 +1500,7 @@ define <8 x i16> @mgather_baseidx_v8i8_v8i16(ptr %base, <8 x i8> %idxs, <8 x i1>
 ; RV64ZVE32F-NEXT:    vslideup.vi v9, v11, 2
 ; RV64ZVE32F-NEXT:    andi a2, a1, 8
 ; RV64ZVE32F-NEXT:    beqz a2, .LBB23_6
-; RV64ZVE32F-NEXT:  .LBB23_13: # %cond.load7
+; RV64ZVE32F-NEXT:  .LBB23_15: # %cond.load7
 ; RV64ZVE32F-NEXT:    vsetivli zero, 1, e8, mf4, ta, ma
 ; RV64ZVE32F-NEXT:    vslidedown.vi v8, v8, 1
 ; RV64ZVE32F-NEXT:    vmv.x.s a2, v8
@@ -1491,7 +1513,7 @@ define <8 x i16> @mgather_baseidx_v8i8_v8i16(ptr %base, <8 x i8> %idxs, <8 x i1>
 ; RV64ZVE32F-NEXT:    vslideup.vi v9, v8, 3
 ; RV64ZVE32F-NEXT:    andi a2, a1, 16
 ; RV64ZVE32F-NEXT:    beqz a2, .LBB23_7
-; RV64ZVE32F-NEXT:  .LBB23_14: # %cond.load10
+; RV64ZVE32F-NEXT:  .LBB23_16: # %cond.load10
 ; RV64ZVE32F-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; RV64ZVE32F-NEXT:    vmv.x.s a2, v10
 ; RV64ZVE32F-NEXT:    slli a2, a2, 1
@@ -1504,30 +1526,6 @@ define <8 x i16> @mgather_baseidx_v8i8_v8i16(ptr %base, <8 x i8> %idxs, <8 x i1>
 ; RV64ZVE32F-NEXT:    andi a2, a1, 32
 ; RV64ZVE32F-NEXT:    bnez a2, .LBB23_8
 ; RV64ZVE32F-NEXT:    j .LBB23_9
-; RV64ZVE32F-NEXT:  .LBB23_15: # %cond.load16
-; RV64ZVE32F-NEXT:    vmv.x.s a2, v8
-; RV64ZVE32F-NEXT:    slli a2, a2, 1
-; RV64ZVE32F-NEXT:    add a2, a0, a2
-; RV64ZVE32F-NEXT:    lh a2, 0(a2)
-; RV64ZVE32F-NEXT:    vsetvli zero, zero, e16, mf2, ta, ma
-; RV64ZVE32F-NEXT:    vmv.s.x v10, a2
-; RV64ZVE32F-NEXT:    vsetivli zero, 7, e16, m1, tu, ma
-; RV64ZVE32F-NEXT:    vslideup.vi v9, v10, 6
-; RV64ZVE32F-NEXT:    andi a1, a1, -128
-; RV64ZVE32F-NEXT:    beqz a1, .LBB23_11
-; RV64ZVE32F-NEXT:  .LBB23_16: # %cond.load19
-; RV64ZVE32F-NEXT:    vsetivli zero, 1, e8, mf4, ta, ma
-; RV64ZVE32F-NEXT:    vslidedown.vi v8, v8, 1
-; RV64ZVE32F-NEXT:    vmv.x.s a1, v8
-; RV64ZVE32F-NEXT:    slli a1, a1, 1
-; RV64ZVE32F-NEXT:    add a0, a0, a1
-; RV64ZVE32F-NEXT:    lh a0, 0(a0)
-; RV64ZVE32F-NEXT:    vsetvli zero, zero, e16, mf2, ta, ma
-; RV64ZVE32F-NEXT:    vmv.s.x v8, a0
-; RV64ZVE32F-NEXT:    vsetivli zero, 8, e16, m1, ta, ma
-; RV64ZVE32F-NEXT:    vslideup.vi v9, v8, 7
-; RV64ZVE32F-NEXT:    vmv1r.v v8, v9
-; RV64ZVE32F-NEXT:    ret
   %ptrs = getelementptr inbounds i16, ptr %base, <8 x i8> %idxs
   %v = call <8 x i16> @llvm.masked.gather.v8i16.v8p0(<8 x ptr> %ptrs, i32 2, <8 x i1> %m, <8 x i16> %passthru)
   ret <8 x i16> %v
@@ -1587,13 +1585,13 @@ define <8 x i16> @mgather_baseidx_sext_v8i8_v8i16(ptr %base, <8 x i8> %idxs, <8
 ; RV64ZVE32F-NEXT:    andi a2, a1, 4
 ; RV64ZVE32F-NEXT:    vsetivli zero, 2, e8, mf4, ta, ma
 ; RV64ZVE32F-NEXT:    vslidedown.vi v8, v8, 2
-; RV64ZVE32F-NEXT:    bnez a2, .LBB24_12
+; RV64ZVE32F-NEXT:    bnez a2, .LBB24_14
 ; RV64ZVE32F-NEXT:  # %bb.5: # %else5
 ; RV64ZVE32F-NEXT:    andi a2, a1, 8
-; RV64ZVE32F-NEXT:    bnez a2, .LBB24_13
+; RV64ZVE32F-NEXT:    bnez a2, .LBB24_15
 ; RV64ZVE32F-NEXT:  .LBB24_6: # %else8
 ; RV64ZVE32F-NEXT:    andi a2, a1, 16
-; RV64ZVE32F-NEXT:    bnez a2, .LBB24_14
+; RV64ZVE32F-NEXT:    bnez a2, .LBB24_16
 ; RV64ZVE32F-NEXT:  .LBB24_7: # %else11
 ; RV64ZVE32F-NEXT:    andi a2, a1, 32
 ; RV64ZVE32F-NEXT:    beqz a2, .LBB24_9
@@ -1612,14 +1610,35 @@ define <8 x i16> @mgather_baseidx_sext_v8i8_v8i16(ptr %base, <8 x i8> %idxs, <8
 ; RV64ZVE32F-NEXT:    andi a2, a1, 64
 ; RV64ZVE32F-NEXT:    vsetivli zero, 2, e8, mf4, ta, ma
 ; RV64ZVE32F-NEXT:    vslidedown.vi v8, v10, 2
-; RV64ZVE32F-NEXT:    bnez a2, .LBB24_15
-; RV64ZVE32F-NEXT:  # %bb.10: # %else17
+; RV64ZVE32F-NEXT:    beqz a2, .LBB24_11
+; RV64ZVE32F-NEXT:  # %bb.10: # %cond.load16
+; RV64ZVE32F-NEXT:    vmv.x.s a2, v8
+; RV64ZVE32F-NEXT:    slli a2, a2, 1
+; RV64ZVE32F-NEXT:    add a2, a0, a2
+; RV64ZVE32F-NEXT:    lh a2, 0(a2)
+; RV64ZVE32F-NEXT:    vsetvli zero, zero, e16, mf2, ta, ma
+; RV64ZVE32F-NEXT:    vmv.s.x v10, a2
+; RV64ZVE32F-NEXT:    vsetivli zero, 7, e16, m1, tu, ma
+; RV64ZVE32F-NEXT:    vslideup.vi v9, v10, 6
+; RV64ZVE32F-NEXT:  .LBB24_11: # %else17
 ; RV64ZVE32F-NEXT:    andi a1, a1, -128
-; RV64ZVE32F-NEXT:    bnez a1, .LBB24_16
-; RV64ZVE32F-NEXT:  .LBB24_11: # %else20
+; RV64ZVE32F-NEXT:    beqz a1, .LBB24_13
+; RV64ZVE32F-NEXT:  # %bb.12: # %cond.load19
+; RV64ZVE32F-NEXT:    vsetivli zero, 1, e8, mf4, ta, ma
+; RV64ZVE32F-NEXT:    vslidedown.vi v8, v8, 1
+; RV64ZVE32F-NEXT:    vmv.x.s a1, v8
+; RV64ZVE32F-NEXT:    slli a1, a1, 1
+; RV64ZVE32F-NEXT:    add a0, a0, a1
+; RV64ZVE32F-NEXT:    lh a0, 0(a0)
+; RV64ZVE32F-NEXT:    vsetvli zero, zero, e16, mf2, ta, ma
+; RV64ZVE32F-NEXT:    vmv.s.x v8, a0
+; RV64ZVE32F-NEXT:    vsetivli zero, 8, e16, m1, ta, ma
+; RV64ZVE32F-NEXT:    vslideup.vi v9, v8, 7
+; RV64ZVE32F-NEXT:  .LBB24_13: # %else20
+; RV64ZVE32F-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; RV64ZVE32F-NEXT:    vmv1r.v v8, v9
 ; RV64ZVE32F-NEXT:    ret
-; RV64ZVE32F-NEXT:  .LBB24_12: # %cond.load4
+; RV64ZVE32F-NEXT:  .LBB24_14: # %cond.load4
 ; RV64ZVE32F-NEXT:    vmv.x.s a2, v8
 ; RV64ZVE32F-NEXT:    slli a2, a2, 1
 ; RV64ZVE32F-NEXT:    add a2, a0, a2
@@ -1630,7 +1649,7 @@ define <8 x i16> @mgather_baseidx_sext_v8i8_v8i16(ptr %base, <8 x i8> %idxs, <8
 ; RV64ZVE32F-NEXT:    vslideup.vi v9, v11, 2
 ; RV64ZVE32F-NEXT:    andi a2, a1, 8
 ; RV64ZVE32F-NEXT:    beqz a2, .LBB24_6
-; RV64ZVE32F-NEXT:  .LBB24_13: # %cond.load7
+; RV64ZVE32F-NEXT:  .LBB24_15: # %cond.load7
 ; RV64ZVE32F-NEXT:    vsetivli zero, 1, e8, mf4, ta, ma
 ; RV64ZVE32F-NEXT:    vslidedown.vi v8, v8, 1
 ; RV64ZVE32F-NEXT:    vmv.x.s a2, v8
@@ -1643,7 +1662,7 @@ define <8 x i16> @mgather_baseidx_sext_v8i8_v8i16(ptr %base, <8 x i8> %idxs, <8
 ; RV64ZVE32F-NEXT:    vslideup.vi v9, v8, 3
 ; RV64ZVE32F-NEXT:    andi a2, a1, 16
 ; RV64ZVE32F-NEXT:    beqz a2, .LBB24_7
-; RV64ZVE32F-NEXT:  .LBB24_14: # %cond.load10
+; RV64ZVE32F-NEXT:  .LBB24_16: # %cond.load10
 ; RV64ZVE32F-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; RV64ZVE32F-NEXT:    vmv.x.s a2, v10
 ; RV64ZVE32F-NEXT:    slli a2, a2, 1
@@ -1656,30 +1675,6 @@ define <8 x i16> @mgather_baseidx_sext_v8i8_v8i16(ptr %base, <8 x i8> %idxs, <8
 ; RV64ZVE32F-NEXT:    andi a2, a1, 32
 ; RV64ZVE32F-NEXT:    bnez a2, .LBB24_8
 ; RV64ZVE32F-NEXT:    j .LBB24_9
-; RV64ZVE32F-NEXT:  .LBB24_15: # %cond.load16
-; RV64ZVE32F-NEXT:    vmv.x.s a2, v8
-; RV64ZVE32F-NEXT:    slli a2, a2, 1
-; RV64ZVE32F-NEXT:    add a2, a0, a2
-; RV64ZVE32F-NEXT:    lh a2, 0(a2)
-; RV64ZVE32F-NEXT:    vsetvli zero, zero, e16, mf2, ta, ma
-; RV64ZVE32F-NEXT:    vmv.s.x v10, a2
-; RV64ZVE32F-NEXT:    vsetivli zero, 7, e16, m1, tu, ma
-; RV64ZVE32F-NEXT:    vslideup.vi v9, v10, 6
-; RV64ZVE32F-NEXT:    andi a1, a1, -128
-; RV64ZVE32F-NEXT:    beqz a1, .LBB24_11
-; RV64ZVE32F-NEXT:  .LBB24_16: # %cond.load19
-; RV64ZVE32F-NEXT:    vsetivli zero, 1, e8, mf4, ta, ma
-; RV64ZVE32F-NEXT:    vslidedown.vi v8, v8, 1
-; RV64ZVE32F-NEXT:    vmv.x.s a1, v8
-; RV64ZVE32F-NEXT:    slli a1, a1, 1
-; RV64ZVE32F-NEXT:    add a0, a0, a1
-; RV64ZVE32F-NEXT:    lh a0, 0(a0)
-; RV64ZVE32F-NEXT:    vsetvli zero, zero, e16, mf2, ta, ma
-; RV64ZVE32F-NEXT:    vmv.s.x v8, a0
-; RV64ZVE32F-NEXT:    vsetivli zero, 8, e16, m1, ta, ma
-; RV64ZVE32F-NEXT:    vslideup.vi v9, v8, 7
-; RV64ZVE32F-NEXT:    vmv1r.v v8, v9
-; RV64ZVE32F-NEXT:    ret
   %eidxs = sext <8 x i8> %idxs to <8 x i16>
   %ptrs = getelementptr inbounds i16, ptr %base, <8 x i16> %eidxs
   %v = call <8 x i16> @llvm.masked.gather.v8i16.v8p0(<8 x ptr> %ptrs, i32 2, <8 x i1> %m, <8 x i16> %passthru)
@@ -1740,13 +1735,13 @@ define <8 x i16> @mgather_baseidx_zext_v8i8_v8i16(ptr %base, <8 x i8> %idxs, <8
 ; RV64ZVE32F-NEXT:    andi a2, a1, 4
 ; RV64ZVE32F-NEXT:    vsetivli zero, 2, e8, mf4, ta, ma
 ; RV64ZVE32F-NEXT:    vslidedown.vi v8, v8, 2
-; RV64ZVE32F-NEXT:    bnez a2, .LBB25_12
+; RV64ZVE32F-NEXT:    bnez a2, .LBB25_14
 ; RV64ZVE32F-NEXT:  # %bb.5: # %else5
 ; RV64ZVE32F-NEXT:    andi a2, a1, 8
-; RV64ZVE32F-NEXT:    bnez a2, .LBB25_13
+; RV64ZVE32F-NEXT:    bnez a2, .LBB25_15
 ; RV64ZVE32F-NEXT:  .LBB25_6: # %else8
 ; RV64ZVE32F-NEXT:    andi a2, a1, 16
-; RV64ZVE32F-NEXT:    bnez a2, .LBB25_14
+; RV64ZVE32F-NEXT:    bnez a2, .LBB25_16
 ; RV64ZVE32F-NEXT:  .LBB25_7: # %else11
 ; RV64ZVE32F-NEXT:    andi a2, a1, 32
 ; RV64ZVE32F-NEXT:    beqz a2, .LBB25_9
@@ -1766,14 +1761,37 @@ define <8 x i16> @mgather_baseidx_zext_v8i8_v8i16(ptr %base, <8 x i8> %idxs, <8
 ; RV64ZVE32F-NEXT:    andi a2, a1, 64
 ; RV64ZVE32F-NEXT:    vsetivli zero, 2, e8, mf4, ta, ma
 ; RV64ZVE32F-NEXT:    vslidedown.vi v8, v10, 2
-; RV64ZVE32F-NEXT:    bnez a2, .LBB25_15
-; RV64ZVE32F-NEXT:  # %bb.10: # %else17
+; RV64ZVE32F-NEXT:    beqz a2, .LBB25_11
+; RV64ZVE32F-NEXT:  # %bb.10: # %cond.load16
+; RV64ZVE32F-NEXT:    vmv.x.s a2, v8
+; RV64ZVE32F-NEXT:    andi a2, a2, 255
+; RV64ZVE32F-NEXT:    slli a2, a2, 1
+; RV64ZVE32F-NEXT:    add a2, a0, a2
+; RV64ZVE32F-NEXT:    lh a2, 0(a2)
+; RV64ZVE32F-NEXT:    vsetvli zero, zero, e16, mf2, ta, ma
+; RV64ZVE32F-NEXT:    vmv.s.x v10, a2
+; RV64ZVE32F-NEXT:    vsetivli zero, 7, e16, m1, tu, ma
+; RV64ZVE32F-NEXT:    vslideup.vi v9, v10, 6
+; RV64ZVE32F-NEXT:  .LBB25_11: # %else17
 ; RV64ZVE32F-NEXT:    andi a1, a1, -128
-; RV64ZVE32F-NEXT:    bnez a1, .LBB25_16
-; RV64ZVE32F-NEXT:  .LBB25_11: # %else20
+; RV64ZVE32F-NEXT:    beqz a1, .LBB25_13
+; RV64ZVE32F-NEXT:  # %bb.12: # %cond.load19
+; RV64ZVE32F-NEXT:    vsetivli zero, 1, e8, mf4, ta, ma
+; RV64ZVE32F-NEXT:    vslidedown.vi v8, v8, 1
+; RV64ZVE32F-NEXT:    vmv.x.s a1, v8
+; RV64ZVE32F-NEXT:    andi a1, a1, 255
+; RV64ZVE32F-NEXT:    slli a1, a1, 1
+; RV64ZVE32F-NEXT:    add a0, a0, a1
+; RV64ZVE32F-NEXT:    lh a0, 0(a0)
+; RV64ZVE32F-NEXT:    vsetvli zero, zero, e16, mf2, ta, ma
+; RV64ZVE32F-NEXT:    vmv.s.x v8, a0
+; RV64ZVE32F-NEXT:    vsetivli zero, 8, e16, m1, ta, ma
+; RV64ZVE32F-NEXT:    vslideup.vi v9, v8, 7
+; RV64ZVE32F-NEXT:  .LBB25_13: # %else20
+; RV64ZVE32F-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; RV64ZVE32F-NEXT:    vmv1r.v v8, v9
 ; RV64ZVE32F-NEXT:    ret
-; RV64ZVE32F-NEXT:  .LBB25_12: # %cond.load4
+; RV64ZVE32F-NEXT:  .LBB25_14: # %cond.load4
 ; RV64ZVE32F-NEXT:    vmv.x.s a2, v8
 ; RV64ZVE32F-NEXT:    andi a2, a2, 255
 ; RV64ZVE32F-NEXT:    slli a2, a2, 1
@@ -1785,7 +1803,7 @@ define <8 x i16> @mgather_baseidx_zext_v8i8_v8i16(ptr %base, <8 x i8> %idxs, <8
 ; RV64ZVE32F-NEXT:    vslideup.vi v9, v11, 2
 ; RV64ZVE32F-NEXT:    andi a2, a1, 8
 ; RV64ZVE32F-NEXT:    beqz a2, .LBB25_6
-; RV64ZVE32F-NEXT:  .LBB25_13: # %cond.load7
+; RV64ZVE32F-NEXT:  .LBB25_15: # %cond.load7
 ; RV64ZVE32F-NEXT:    vsetivli zero, 1, e8, mf4, ta, ma
 ; RV64ZVE32F-NEXT:    vslidedown.vi v8, v8, 1
 ; RV64ZVE32F-NEXT:    vmv.x.s a2, v8
@@ -1799,7 +1817,7 @@ define <8 x i16> @mgather_baseidx_zext_v8i8_v8i16(ptr %base, <8 x i8> %idxs, <8
 ; RV64ZVE32F-NEXT:    vslideup.vi v9, v8, 3
 ; RV64ZVE32F-NEXT:    andi a2, a1, 16
 ; RV64ZVE32F-NEXT:    beqz a2, .LBB25_7
-; RV64ZVE32F-NEXT:  .LBB25_14: # %cond.load10
+; RV64ZVE32F-NEXT:  .LBB25_16: # %cond.load10
 ; RV64ZVE32F-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; RV64ZVE32F-NEXT:    vmv.x.s a2, v10
 ; RV64ZVE32F-NEXT:    andi a2, a2, 255
@@ -1813,32 +1831,6 @@ define <8 x i16> @mgather_baseidx_zext_v8i8_v8i16(ptr %base, <8 x i8> %idxs, <8
 ; RV64ZVE32F-NEXT:    andi a2, a1, 32
 ; RV64ZVE32F-NEXT:    bnez a2, .LBB25_8
 ; RV64ZVE32F-NEXT:    j .LBB25_9
-; RV64ZVE32F-NEXT:  .LBB25_15: # %cond.load16
-; RV64ZVE32F-NEXT:    vmv.x.s a2, v8
-; RV64ZVE32F-NEXT:    andi a2, a2, 255
-; RV64ZVE32F-NEXT:    slli a2, a2, 1
-; RV64ZVE32F-NEXT:    add a2, a0, a2
-; RV64ZVE32F-NEXT:    lh a2, 0(a2)
-; RV64ZVE32F-NEXT:    vsetvli zero, zero, e16, mf2, ta, ma
-; RV64ZVE32F-NEXT:    vmv.s.x v10, a2
-; RV64ZVE32F-NEXT:    vsetivli zero, 7, e16, m1, tu, ma
-; RV64ZVE32F-NEXT:    vslideup.vi v9, v10, 6
-; RV64ZVE32F-NEXT:    andi a1, a1, -128
-; RV64ZVE32F-NEXT:    beqz a1, .LBB25_11
-; RV64ZVE32F-NEXT:  .LBB25_16: # %cond.load19
-; RV64ZVE32F-NEXT:    vsetivli zero, 1, e8, mf4, ta, ma
-; RV64ZVE32F-NEXT:    vslidedown.vi v8, v8, 1
-; RV64ZVE32F-NEXT:    vmv.x.s a1, v8
-; RV64ZVE32F-NEXT:    andi a1, a1, 255
-; RV64ZVE32F-NEXT:    slli a1, a1, 1
-; RV64ZVE32F-NEXT:    add a0, a0, a1
-; RV64ZVE32F-NEXT:    lh a0, 0(a0)
-; RV64ZVE32F-NEXT:    vsetvli zero, zero, e16, mf2, ta, ma
-; RV64ZVE32F-NEXT:    vmv.s.x v8, a0
-; RV64ZVE32F-NEXT:    vsetivli zero, 8, e16, m1, ta, ma
-; RV64ZVE32F-NEXT:    vslideup.vi v9, v8, 7
-; RV64ZVE32F-NEXT:    vmv1r.v v8, v9
-; RV64ZVE32F-NEXT:    ret
   %eidxs = zext <8 x i8> %idxs to <8 x i16>
   %ptrs = getelementptr inbounds i16, ptr %base, <8 x i16> %eidxs
   %v = call <8 x i16> @llvm.masked.gather.v8i16.v8p0(<8 x ptr> %ptrs, i32 2, <8 x i1> %m, <8 x i16> %passthru)
@@ -1896,13 +1888,13 @@ define <8 x i16> @mgather_baseidx_v8i16(ptr %base, <8 x i16> %idxs, <8 x i1> %m,
 ; RV64ZVE32F-NEXT:    andi a2, a1, 4
 ; RV64ZVE32F-NEXT:    vsetivli zero, 2, e16, mf2, ta, ma
 ; RV64ZVE32F-NEXT:    vslidedown.vi v8, v8, 2
-; RV64ZVE32F-NEXT:    bnez a2, .LBB26_12
+; RV64ZVE32F-NEXT:    bnez a2, .LBB26_14
 ; RV64ZVE32F-NEXT:  # %bb.5: # %else5
 ; RV64ZVE32F-NEXT:    andi a2, a1, 8
-; RV64ZVE32F-NEXT:    bnez a2, .LBB26_13
+; RV64ZVE32F-NEXT:    bnez a2, .LBB26_15
 ; RV64ZVE32F-NEXT:  .LBB26_6: # %else8
 ; RV64ZVE32F-NEXT:    andi a2, a1, 16
-; RV64ZVE32F-NEXT:    bnez a2, .LBB26_14
+; RV64ZVE32F-NEXT:    bnez a2, .LBB26_16
 ; RV64ZVE32F-NEXT:  .LBB26_7: # %else11
 ; RV64ZVE32F-NEXT:    andi a2, a1, 32
 ; RV64ZVE32F-NEXT:    beqz a2, .LBB26_9
@@ -1920,14 +1912,33 @@ define <8 x i16> @mgather_baseidx_v8i16(ptr %base, <8 x i16> %idxs, <8 x i1> %m,
 ; RV64ZVE32F-NEXT:    andi a2, a1, 64
 ; RV64ZVE32F-NEXT:    vsetivli zero, 2, e16, mf2, ta, ma
 ; RV64ZVE32F-NEXT:    vslidedown.vi v8, v10, 2
-; RV64ZVE32F-NEXT:    bnez a2, .LBB26_15
-; RV64ZVE32F-NEXT:  # %bb.10: # %else17
+; RV64ZVE32F-NEXT:    beqz a2, .LBB26_11
+; RV64ZVE32F-NEXT:  # %bb.10: # %cond.load16
+; RV64ZVE32F-NEXT:    vmv.x.s a2, v8
+; RV64ZVE32F-NEXT:    slli a2, a2, 1
+; RV64ZVE32F-NEXT:    add a2, a0, a2
+; RV64ZVE32F-NEXT:    lh a2, 0(a2)
+; RV64ZVE32F-NEXT:    vmv.s.x v10, a2
+; RV64ZVE32F-NEXT:    vsetivli zero, 7, e16, m1, tu, ma
+; RV64ZVE32F-NEXT:    vslideup.vi v9, v10, 6
+; RV64ZVE32F-NEXT:  .LBB26_11: # %else17
 ; RV64ZVE32F-NEXT:    andi a1, a1, -128
-; RV64ZVE32F-NEXT:    bnez a1, .LBB26_16
-; RV64ZVE32F-NEXT:  .LBB26_11: # %else20
+; RV64ZVE32F-NEXT:    beqz a1, .LBB26_13
+; RV64ZVE32F-NEXT:  # %bb.12: # %cond.load19
+; RV64ZVE32F-NEXT:    vsetivli zero, 1, e16, mf2, ta, ma
+; RV64ZVE32F-NEXT:    vslidedown.vi v8, v8, 1
+; RV64ZVE32F-NEXT:    vmv.x.s a1, v8
+; RV64ZVE32F-NEXT:    slli a1, a1, 1
+; RV64ZVE32F-NEXT:    add a0, a0, a1
+; RV64ZVE32F-NEXT:    lh a0, 0(a0)
+; RV64ZVE32F-NEXT:    vmv.s.x v8, a0
+; RV64ZVE32F-NEXT:    vsetivli zero, 8, e16, m1, ta, ma
+; RV64ZVE32F-NEXT:    vslideup.vi v9, v8, 7
+; RV64ZVE32F-NEXT:  .LBB26_13: # %else20
+; RV64ZVE32F-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; RV64ZVE32F-NEXT:    vmv1r.v v8, v9
 ; RV64ZVE32F-NEXT:    ret
-; RV64ZVE32F-NEXT:  .LBB26_12: # %cond.load4
+; RV64ZVE32F-NEXT:  .LBB26_14: # %cond.load4
 ; RV64ZVE32F-NEXT:    vmv.x.s a2, v8
 ; RV64ZVE32F-NEXT:    slli a2, a2, 1
 ; RV64ZVE32F-NEXT:    add a2, a0, a2
@@ -1937,7 +1948,7 @@ define <8 x i16> @mgather_baseidx_v8i16(ptr %base, <8 x i16> %idxs, <8 x i1> %m,
 ; RV64ZVE32F-NEXT:    vslideup.vi v9, v11, 2
 ; RV64ZVE32F-NEXT:    andi a2, a1, 8
 ; RV64ZVE32F-NEXT:    beqz a2, .LBB26_6
-; RV64ZVE32F-NEXT:  .LBB26_13: # %cond.load7
+; RV64ZVE32F-NEXT:  .LBB26_15: # %cond.load7
 ; RV64ZVE32F-NEXT:    vsetivli zero, 1, e16, mf2, ta, ma
 ; RV64ZVE32F-NEXT:    vslidedown.vi v8, v8, 1
 ; RV64ZVE32F-NEXT:    vmv.x.s a2, v8
@@ -1949,7 +1960,7 @@ define <8 x i16> @mgather_baseidx_v8i16(ptr %base, <8 x i16> %idxs, <8 x i1> %m,
 ; RV64ZVE32F-NEXT:    vslideup.vi v9, v8, 3
 ; RV64ZVE32F-NEXT:    andi a2, a1, 16
 ; RV64ZVE32F-NEXT:    beqz a2, .LBB26_7
-; RV64ZVE32F-NEXT:  .LBB26_14: # %cond.load10
+; RV64ZVE32F-NEXT:  .LBB26_16: # %cond.load10
 ; RV64ZVE32F-NEXT:    vsetivli zero, 5, e16, m1, tu, ma
 ; RV64ZVE32F-NEXT:    vmv.x.s a2, v10
 ; RV64ZVE32F-NEXT:    slli a2, a2, 1
@@ -1960,28 +1971,6 @@ define <8 x i16> @mgather_baseidx_v8i16(ptr %base, <8 x i16> %idxs, <8 x i1> %m,
 ; RV64ZVE32F-NEXT:    andi a2, a1, 32
 ; RV64ZVE32F-NEXT:    bnez a2, .LBB26_8
 ; RV64ZVE32F-NEXT:    j .LBB26_9
-; RV64ZVE32F-NEXT:  .LBB26_15: # %cond.load16
-; RV64ZVE32F-NEXT:    vmv.x.s a2, v8
-; RV64ZVE32F-NEXT:    slli a2, a2, 1
-; RV64ZVE32F-NEXT:    add a2, a0, a2
-; RV64ZVE32F-NEXT:    lh a2, 0(a2)
-; RV64ZVE32F-NEXT:    vmv.s.x v10, a2
-; RV64ZVE32F-NEXT:    vsetivli zero, 7, e16, m1, tu, ma
-; RV64ZVE32F-NEXT:    vslideup.vi v9, v10, 6
-; RV64ZVE32F-NEXT:    andi a1, a1, -128
-; RV64ZVE32F-NEXT:    beqz a1, .LBB26_11
-; RV64ZVE32F-NEXT:  .LBB26_16: # %cond.load19
-; RV64ZVE32F-NEXT:    vsetivli zero, 1, e16, mf2, ta, ma
-; RV64ZVE32F-NEXT:    vslidedown.vi v8, v8, 1
-; RV64ZVE32F-NEXT:    vmv.x.s a1, v8
-; RV64ZVE32F-NEXT:    slli a1, a1, 1
-; RV64ZVE32F-NEXT:    add a0, a0, a1
-; RV64ZVE32F-NEXT:    lh a0, 0(a0)
-; RV64ZVE32F-NEXT:    vmv.s.x v8, a0
-; RV64ZVE32F-NEXT:    vsetivli zero, 8, e16, m1, ta, ma
-; RV64ZVE32F-NEXT:    vslideup.vi v9, v8, 7
-; RV64ZVE32F-NEXT:    vmv1r.v v8, v9
-; RV64ZVE32F-NEXT:    ret
   %ptrs = getelementptr inbounds i16, ptr %base, <8 x i16> %idxs
   %v = call <8 x i16> @llvm.masked.gather.v8i16.v8p0(<8 x ptr> %ptrs, i32 2, <8 x i1> %m, <8 x i16> %passthru)
   ret <8 x i16> %v
@@ -2311,11 +2300,13 @@ define <4 x i32> @mgather_truemask_v4i32(<4 x ptr> %ptrs, <4 x i32> %passthru) {
 define <4 x i32> @mgather_falsemask_v4i32(<4 x ptr> %ptrs, <4 x i32> %passthru) {
 ; RV32-LABEL: mgather_falsemask_v4i32:
 ; RV32:       # %bb.0:
+; RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; RV32-NEXT:    vmv1r.v v8, v9
 ; RV32-NEXT:    ret
 ;
 ; RV64V-LABEL: mgather_falsemask_v4i32:
 ; RV64V:       # %bb.0:
+; RV64V-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; RV64V-NEXT:    vmv1r.v v8, v10
 ; RV64V-NEXT:    ret
 ;
@@ -2492,13 +2483,13 @@ define <8 x i32> @mgather_baseidx_v8i8_v8i32(ptr %base, <8 x i8> %idxs, <8 x i1>
 ; RV64ZVE32F-NEXT:    andi a2, a1, 4
 ; RV64ZVE32F-NEXT:    vsetivli zero, 2, e8, mf4, ta, ma
 ; RV64ZVE32F-NEXT:    vslidedown.vi v8, v8, 2
-; RV64ZVE32F-NEXT:    bnez a2, .LBB35_12
+; RV64ZVE32F-NEXT:    bnez a2, .LBB35_14
 ; RV64ZVE32F-NEXT:  # %bb.5: # %else5
 ; RV64ZVE32F-NEXT:    andi a2, a1, 8
-; RV64ZVE32F-NEXT:    bnez a2, .LBB35_13
+; RV64ZVE32F-NEXT:    bnez a2, .LBB35_15
 ; RV64ZVE32F-NEXT:  .LBB35_6: # %else8
 ; RV64ZVE32F-NEXT:    andi a2, a1, 16
-; RV64ZVE32F-NEXT:    bnez a2, .LBB35_14
+; RV64ZVE32F-NEXT:    bnez a2, .LBB35_16
 ; RV64ZVE32F-NEXT:  .LBB35_7: # %else11
 ; RV64ZVE32F-NEXT:    andi a2, a1, 32
 ; RV64ZVE32F-NEXT:    beqz a2, .LBB35_9
@@ -2517,14 +2508,35 @@ define <8 x i32> @mgather_baseidx_v8i8_v8i32(ptr %base, <8 x i8> %idxs, <8 x i1>
 ; RV64ZVE32F-NEXT:    andi a2, a1, 64
 ; RV64ZVE32F-NEXT:    vsetivli zero, 2, e8, mf4, ta, ma
 ; RV64ZVE32F-NEXT:    vslidedown.vi v8, v9, 2
-; RV64ZVE32F-NEXT:    bnez a2, .LBB35_15
-; RV64ZVE32F-NEXT:  # %bb.10: # %else17
+; RV64ZVE32F-NEXT:    beqz a2, .LBB35_11
+; RV64ZVE32F-NEXT:  # %bb.10: # %cond.load16
+; RV64ZVE32F-NEXT:    vmv.x.s a2, v8
+; RV64ZVE32F-NEXT:    slli a2, a2, 2
+; RV64ZVE32F-NEXT:    add a2, a0, a2
+; RV64ZVE32F-NEXT:    lw a2, 0(a2)
+; RV64ZVE32F-NEXT:    vsetvli zero, zero, e32, m1, ta, ma
+; RV64ZVE32F-NEXT:    vmv.s.x v12, a2
+; RV64ZVE32F-NEXT:    vsetivli zero, 7, e32, m2, tu, ma
+; RV64ZVE32F-NEXT:    vslideup.vi v10, v12, 6
+; RV64ZVE32F-NEXT:  .LBB35_11: # %else17
 ; RV64ZVE32F-NEXT:    andi a1, a1, -128
-; RV64ZVE32F-NEXT:    bnez a1, .LBB35_16
-; RV64ZVE32F-NEXT:  .LBB35_11: # %else20
+; RV64ZVE32F-NEXT:    beqz a1, .LBB35_13
+; RV64ZVE32F-NEXT:  # %bb.12: # %cond.load19
+; RV64ZVE32F-NEXT:    vsetivli zero, 1, e8, mf4, ta, ma
+; RV64ZVE32F-NEXT:    vslidedown.vi v8, v8, 1
+; RV64ZVE32F-NEXT:    vmv.x.s a1, v8
+; RV64ZVE32F-NEXT:    slli a1, a1, 2
+; RV64ZVE32F-NEXT:    add a0, a0, a1
+; RV64ZVE32F-NEXT:    lw a0, 0(a0)
+; RV64ZVE32F-NEXT:    vsetvli zero, zero, e32, m1, ta, ma
+; RV64ZVE32F-NEXT:    vmv.s.x v8, a0
+; RV64ZVE32F-NEXT:    vsetivli zero, 8, e32, m2, ta, ma
+; RV64ZVE32F-NEXT:    vslideup.vi v10, v8, 7
+; RV64ZVE32F-NEXT:  .LBB35_13: # %else20
+; RV64ZVE32F-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; RV64ZVE32F-NEXT:    vmv2r.v v8, v10
 ; RV64ZVE32F-NEXT:    ret
-; RV64ZVE32F-NEXT:  .LBB35_12: # %cond.load4
+; RV64ZVE32F-NEXT:  .LBB35_14: # %cond.load4
 ; RV64ZVE32F-NEXT:    vmv.x.s a2, v8
 ; RV64ZVE32F-NEXT:    slli a2, a2, 2
 ; RV64ZVE32F-NEXT:    add a2, a0, a2
@@ -2535,7 +2547,7 @@ define <8 x i32> @mgather_baseidx_v8i8_v8i32(ptr %base, <8 x i8> %idxs, <8 x i1>
 ; RV64ZVE32F-NEXT:    vslideup.vi v10, v12, 2
 ; RV64ZVE32F-NEXT:    andi a2, a1, 8
 ; RV64ZVE32F-NEXT:    beqz a2, .LBB35_6
-; RV64ZVE32F-NEXT:  .LBB35_13: # %cond.load7
+; RV64ZVE32F-NEXT:  .LBB35_15: # %cond.load7
 ; RV64ZVE32F-NEXT:    vsetivli zero, 1, e8, mf4, ta, ma
 ; RV64ZVE32F-NEXT:    vslidedown.vi v8, v8, 1
 ; RV64ZVE32F-NEXT:    vmv.x.s a2, v8
@@ -2548,7 +2560,7 @@ define <8 x i32> @mgather_baseidx_v8i8_v8i32(ptr %base, <8 x i8> %idxs, <8 x i1>
 ; RV64ZVE32F-NEXT:    vslideup.vi v10, v8, 3
 ; RV64ZVE32F-NEXT:    andi a2, a1, 16
 ; RV64ZVE32F-NEXT:    beqz a2, .LBB35_7
-; RV64ZVE32F-NEXT:  .LBB35_14: # %cond.load10
+; RV64ZVE32F-NEXT:  .LBB35_16: # %cond.load10
 ; RV64ZVE32F-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; RV64ZVE32F-NEXT:    vmv.x.s a2, v9
 ; RV64ZVE32F-NEXT:    slli a2, a2, 2
@@ -2561,30 +2573,6 @@ define <8 x i32> @mgather_baseidx_v8i8_v8i32(ptr %base, <8 x i8> %idxs, <8 x i1>
 ; RV64ZVE32F-NEXT:    andi a2, a1, 32
 ; RV64ZVE32F-NEXT:    bnez a2, .LBB35_8
 ; RV64ZVE32F-NEXT:    j .LBB35_9
-; RV64ZVE32F-NEXT:  .LBB35_15: # %cond.load16
-; RV64ZVE32F-NEXT:    vmv.x.s a2, v8
-; RV64ZVE32F-NEXT:    slli a2, a2, 2
-; RV64ZVE32F-NEXT:    add a2, a0, a2
-; RV64ZVE32F-NEXT:    lw a2, 0(a2)
-; RV64ZVE32F-NEXT:    vsetvli zero, zero, e32, m1, ta, ma
-; RV64ZVE32F-NEXT:    vmv.s.x v12, a2
-; RV64ZVE32F-NEXT:    vsetivli zero, 7, e32, m2, tu, ma
-; RV64ZVE32F-NEXT:    vslideup.vi v10, v12, 6
-; RV64ZVE32F-NEXT:    andi a1, a1, -128
-; RV64ZVE32F-NEXT:    beqz a1, .LBB35_11
-; RV64ZVE32F-NEXT:  .LBB35_16: # %cond.load19
-; RV64ZVE32F-NEXT:    vsetivli zero, 1, e8, mf4, ta, ma
-; RV64ZVE32F-NEXT:    vslidedown.vi v8, v8, 1
-; RV64ZVE32F-NEXT:    vmv.x.s a1, v8
-; RV64ZVE32F-NEXT:    slli a1, a1, 2
-; RV64ZVE32F-NEXT:    add a0, a0, a1
-; RV64ZVE32F-NEXT:    lw a0, 0(a0)
-; RV64ZVE32F-NEXT:    vsetvli zero, zero, e32, m1, ta, ma
-; RV64ZVE32F-NEXT:    vmv.s.x v8, a0
-; RV64ZVE32F-NEXT:    vsetivli zero, 8, e32, m2, ta, ma
-; RV64ZVE32F-NEXT:    vslideup.vi v10, v8, 7
-; RV64ZVE32F-NEXT:    vmv2r.v v8, v10
-; RV64ZVE32F-NEXT:    ret
   %ptrs = getelementptr inbounds i32, ptr %base, <8 x i8> %idxs
   %v = call <8 x i32> @llvm.masked.gather.v8i32.v8p0(<8 x ptr> %ptrs, i32 4, <8 x i1> %m, <8 x i32> %passthru)
   ret <8 x i32> %v
@@ -2643,13 +2631,13 @@ define <8 x i32> @mgather_baseidx_sext_v8i8_v8i32(ptr %base, <8 x i8> %idxs, <8
 ; RV64ZVE32F-NEXT:    andi a2, a1, 4
 ; RV64ZVE32F-NEXT:    vsetivli zero, 2, e8, mf4, ta, ma
 ; RV64ZVE32F-NEXT:    vslidedown.vi v8, v8, 2
-; RV64ZVE32F-NEXT:    bnez a2, .LBB36_12
+; RV64ZVE32F-NEXT:    bnez a2, .LBB36_14
 ; RV64ZVE32F-NEXT:  # %bb.5: # %else5
 ; RV64ZVE32F-NEXT:    andi a2, a1, 8
-; RV64ZVE32F-NEXT:    bnez a2, .LBB36_13
+; RV64ZVE32F-NEXT:    bnez a2, .LBB36_15
 ; RV64ZVE32F-NEXT:  .LBB36_6: # %else8
 ; RV64ZVE32F-NEXT:    andi a2, a1, 16
-; RV64ZVE32F-NEXT:    bnez a2, .LBB36_14
+; RV64ZVE32F-NEXT:    bnez a2, .LBB36_16
 ; RV64ZVE32F-NEXT:  .LBB36_7: # %else11
 ; RV64ZVE32F-NEXT:    andi a2, a1, 32
 ; RV64ZVE32F-NEXT:    beqz a2, .LBB36_9
@@ -2668,14 +2656,35 @@ define <8 x i32> @mgather_baseidx_sext_v8i8_v8i32(ptr %base, <8 x i8> %idxs, <8
 ; RV64ZVE32F-NEXT:    andi a2, a1, 64
 ; RV64ZVE32F-NEXT:    vsetivli zero, 2, e8, mf4, ta, ma
 ; RV64ZVE32F-NEXT:    vslidedown.vi v8, v9, 2
-; RV64ZVE32F-NEXT:    bnez a2, .LBB36_15
-; RV64ZVE32F-NEXT:  # %bb.10: # %else17
+; RV64ZVE32F-NEXT:    beqz a2, .LBB36_11
+; RV64ZVE32F-NEXT:  # %bb.10: # %cond.load16
+; RV64ZVE32F-NEXT:    vmv.x.s a2, v8
+; RV64ZVE32F-NEXT:    slli a2, a2, 2
+; RV64ZVE32F-NEXT:    add a2, a0, a2
+; RV64ZVE32F-NEXT:    lw a2, 0(a2)
+; RV64ZVE32F-NEXT:    vsetvli zero, zero, e32, m1, ta, ma
+; RV64ZVE32F-NEXT:    vmv.s.x v12, a2
+; RV64ZVE32F-NEXT:    vsetivli zero, 7, e32, m2, tu, ma
+; RV64ZVE32F-NEXT:    vslideup.vi v10, v12, 6
+; RV64ZVE32F-NEXT:  .LBB36_11: # %else17
 ; RV64ZVE32F-NEXT:    andi a1, a1, -128
-; RV64ZVE32F-NEXT:    bnez a1, .LBB36_16
-; RV64ZVE32F-NEXT:  .LBB36_11: # %else20
+; RV64ZVE32F-NEXT:    beqz a1, .LBB36_13
+; RV64ZVE32F-NEXT:  # %bb.12: # %cond.load19
+; RV64ZVE32F-NEXT:    vsetivli zero, 1, e8, mf4, ta, ma
+; RV64ZVE32F-NEXT:    vslidedown.vi v8, v8, 1
+; RV64ZVE32F-NEXT:    vmv.x.s a1, v8
+; RV64ZVE32F-NEXT:    slli a1, a1, 2
+; RV64ZVE32F-NEXT:    add a0, a0, a1
+; RV64ZVE32F-NEXT:    lw a0, 0(a0)
+; RV64ZVE32F-NEXT:    vsetvli zero, zero, e32, m1, ta, ma
+; RV64ZVE32F-NEXT:    vmv.s.x v8, a0
+; RV64ZVE32F-NEXT:    vsetivli zero, 8, e32, m2, ta, ma
+; RV64ZVE32F-NEXT:    vslideup.vi v10, v8, 7
+; RV64ZVE32F-NEXT:  .LBB36_13: # %else20
+; RV64ZVE32F-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; RV64ZVE32F-NEXT:    vmv2r.v v8, v10
 ; RV64ZVE32F-NEXT:    ret
-; RV64ZVE32F-NEXT:  .LBB36_12: # %cond.load4
+; RV64ZVE32F-NEXT:  .LBB36_14: # %cond.load4
 ; RV64ZVE32F-NEXT:    vmv.x.s a2, v8
 ; RV64ZVE32F-NEXT:    slli a2, a2, 2
 ; RV64ZVE32F-NEXT:    add a2, a0, a2
@@ -2686,7 +2695,7 @@ define <8 x i32> @mgather_baseidx_sext_v8i8_v8i32(ptr %base, <8 x i8> %idxs, <8
 ; RV64ZVE32F-NEXT:    vslideup.vi v10, v12, 2
 ; RV64ZVE32F-NEXT:    andi a2, a1, 8
 ; RV64ZVE32F-NEXT:    beqz a2, .LBB36_6
-; RV64ZVE32F-NEXT:  .LBB36_13: # %cond.load7
+; RV64ZVE32F-NEXT:  .LBB36_15: # %cond.load7
 ; RV64ZVE32F-NEXT:    vsetivli zero, 1, e8, mf4, ta, ma
 ; RV64ZVE32F-NEXT:    vslidedown.vi v8, v8, 1
 ; RV64ZVE32F-NEXT:    vmv.x.s a2, v8
@@ -2699,7 +2708,7 @@ define <8 x i32> @mgather_baseidx_sext_v8i8_v8i32(ptr %base, <8 x i8> %idxs, <8
 ; RV64ZVE32F-NEXT:    vslideup.vi v10, v8, 3
 ; RV64ZVE32F-NEXT:    andi a2, a1, 16
 ; RV64ZVE32F-NEXT:    beqz a2, .LBB36_7
-; RV64ZVE32F-NEXT:  .LBB36_14: # %cond.load10
+; RV64ZVE32F-NEXT:  .LBB36_16: # %cond.load10
 ; RV64ZVE32F-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; RV64ZVE32F-NEXT:    vmv.x.s a2, v9
 ; RV64ZVE32F-NEXT:    slli a2, a2, 2
@@ -2712,30 +2721,6 @@ define <8 x i32> @mgather_baseidx_sext_v8i8_v8i32(ptr %base, <8 x i8> %idxs, <8
 ; RV64ZVE32F-NEXT:    andi a2, a1, 32
 ; RV64ZVE32F-NEXT:    bnez a2, .LBB36_8
 ; RV64ZVE32F-NEXT:    j .LBB36_9
-; RV64ZVE32F-NEXT:  .LBB36_15: # %cond.load16
-; RV64ZVE32F-NEXT:    vmv.x.s a2, v8
-; RV64ZVE32F-NEXT:    slli a2, a2, 2
-; RV64ZVE32F-NEXT:    add a2, a0, a2
-; RV64ZVE32F-NEXT:    lw a2, 0(a2)
-; RV64ZVE32F-NEXT:    vsetvli zero, zero, e32, m1, ta, ma
-; RV64ZVE32F-NEXT:    vmv.s.x v12, a2
-; RV64ZVE32F-NEXT:    vsetivli zero, 7, e32, m2, tu, ma
-; RV64ZVE32F-NEXT:    vslideup.vi v10, v12, 6
-; RV64ZVE32F-NEXT:    andi a1, a1, -128
-; RV64ZVE32F-NEXT:    beqz a1, .LBB36_11
-; RV64ZVE32F-NEXT:  .LBB36_16: # %cond.load19
-; RV64ZVE32F-NEXT:    vsetivli zero, 1, e8, mf4, ta, ma
-; RV64ZVE32F-NEXT:    vslidedown.vi v8, v8, 1
-; RV64ZVE32F-NEXT:    vmv.x.s a1, v8
-; RV64ZVE32F-NEXT:    slli a1, a1, 2
-; RV64ZVE32F-NEXT:    add a0, a0, a1
-; RV64ZVE32F-NEXT:    lw a0, 0(a0)
-; RV64ZVE32F-NEXT:    vsetvli zero, zero, e32, m1, ta, ma
-; RV64ZVE32F-NEXT:    vmv.s.x v8, a0
-; RV64ZVE32F-NEXT:    vsetivli zero, 8, e32, m2, ta, ma
-; RV64ZVE32F-NEXT:    vslideup.vi v10, v8, 7
-; RV64ZVE32F-NEXT:    vmv2r.v v8, v10
-; RV64ZVE32F-NEXT:    ret
   %eidxs = sext <8 x i8> %idxs to <8 x i32>
   %ptrs = getelementptr inbounds i32, ptr %base, <8 x i32> %eidxs
   %v = call <8 x i32> @llvm.masked.gather.v8i32.v8p0(<8 x ptr> %ptrs, i32 4, <8 x i1> %m, <8 x i32> %passthru)
@@ -2798,13 +2783,13 @@ define <8 x i32> @mgather_baseidx_zext_v8i8_v8i32(ptr %base, <8 x i8> %idxs, <8
 ; RV64ZVE32F-NEXT:    andi a2, a1, 4
 ; RV64ZVE32F-NEXT:    vsetivli zero, 2, e8, mf4, ta, ma
 ; RV64ZVE32F-NEXT:    vslidedown.vi v8, v8, 2
-; RV64ZVE32F-NEXT:    bnez a2, .LBB37_12
+; RV64ZVE32F-NEXT:    bnez a2, .LBB37_14
 ; RV64ZVE32F-NEXT:  # %bb.5: # %else5
 ; RV64ZVE32F-NEXT:    andi a2, a1, 8
-; RV64ZVE32F-NEXT:    bnez a2, .LBB37_13
+; RV64ZVE32F-NEXT:    bnez a2, .LBB37_15
 ; RV64ZVE32F-NEXT:  .LBB37_6: # %else8
 ; RV64ZVE32F-NEXT:    andi a2, a1, 16
-; RV64ZVE32F-NEXT:    bnez a2, .LBB37_14
+; RV64ZVE32F-NEXT:    bnez a2, .LBB37_16
 ; RV64ZVE32F-NEXT:  .LBB37_7: # %else11
 ; RV64ZVE32F-NEXT:    andi a2, a1, 32
 ; RV64ZVE32F-NEXT:    beqz a2, .LBB37_9
@@ -2824,14 +2809,37 @@ define <8 x i32> @mgather_baseidx_zext_v8i8_v8i32(ptr %base, <8 x i8> %idxs, <8
 ; RV64ZVE32F-NEXT:    andi a2, a1, 64
 ; RV64ZVE32F-NEXT:    vsetivli zero, 2, e8, mf4, ta, ma
 ; RV64ZVE32F-NEXT:    vslidedown.vi v8, v9, 2
-; RV64ZVE32F-NEXT:    bnez a2, .LBB37_15
-; RV64ZVE32F-NEXT:  # %bb.10: # %else17
+; RV64ZVE32F-NEXT:    beqz a2, .LBB37_11
+; RV64ZVE32F-NEXT:  # %bb.10: # %cond.load16
+; RV64ZVE32F-NEXT:    vmv.x.s a2, v8
+; RV64ZVE32F-NEXT:    andi a2, a2, 255
+; RV64ZVE32F-NEXT:    slli a2, a2, 2
+; RV64ZVE32F-NEXT:    add a2, a0, a2
+; RV64ZVE32F-NEXT:    lw a2, 0(a2)
+; RV64ZVE32F-NEXT:    vsetvli zero, zero, e32, m1, ta, ma
+; RV64ZVE32F-NEXT:    vmv.s.x v12, a2
+; RV64ZVE32F-NEXT:    vsetivli zero, 7, e32, m2, tu, ma
+; RV64ZVE32F-NEXT:    vslideup.vi v10, v12, 6
+; RV64ZVE32F-NEXT:  .LBB37_11: # %else17
 ; RV64ZVE32F-NEXT:    andi a1, a1, -128
-; RV64ZVE32F-NEXT:    bnez a1, .LBB37_16
-; RV64ZVE32F-NEXT:  .LBB37_11: # %else20
+; RV64ZVE32F-NEXT:    beqz a1, .LBB37_13
+; RV64ZVE32F-NEXT:  # %bb.12: # %cond.load19
+; RV64ZVE32F-NEXT:    vsetivli zero, 1, e8, mf4, ta, ma
+; RV64ZVE32F-NEXT:    vslidedown.vi v8, v8, 1
+; RV64ZVE32F-NEXT:    vmv.x.s a1, v8
+; RV64ZVE32F-NEXT:    andi a1, a1, 255
+; RV64ZVE32F-NEXT:    slli a1, a1, 2
+; RV64ZVE32F-NEXT:    add a0, a0, a1
+; RV64ZVE32F-NEXT:    lw a0, 0(a0)
+; RV64ZVE32F-NEXT:    vsetvli zero, zero, e32, m1, ta, ma
+; RV64ZVE32F-NEXT:    vmv.s.x v8, a0
+; RV64ZVE32F-NEXT:    vsetivli zero, 8, e32, m2, ta, ma
+; RV64ZVE32F-NEXT:    vslideup.vi v10, v8, 7
+; RV64ZVE32F-NEXT:  .LBB37_13: # %else20
+; RV64ZVE32F-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; RV64ZVE32F-NEXT:    vmv2r.v v8, v10
 ; RV64ZVE32F-NEXT:    ret
-; RV64ZVE32F-NEXT:  .LBB37_12: # %cond.load4
+; RV64ZVE32F-NEXT:  .LBB37_14: # %cond.load4
 ; RV64ZVE32F-NEXT:    vmv.x.s a2, v8
 ; RV64ZVE32F-NEXT:    andi a2, a2, 255
 ; RV64ZVE32F-NEXT:    slli a2, a2, 2
@@ -2843,7 +2851,7 @@ define <8 x i32> @mgather_baseidx_zext_v8i8_v8i32(ptr %base, <8 x i8> %idxs, <8
 ; RV64ZVE32F-NEXT:    vslideup.vi v10, v12, 2
 ; RV64ZVE32F-NEXT:    andi a2, a1, 8
 ; RV64ZVE32F-NEXT:    beqz a2, .LBB37_6
-; RV64ZVE32F-NEXT:  .LBB37_13: # %cond.load7
+; RV64ZVE32F-NEXT:  .LBB37_15: # %cond.load7
 ; RV64ZVE32F-NEXT:    vsetivli zero, 1, e8, mf4, ta, ma
 ; RV64ZVE32F-NEXT:    vslidedown.vi v8, v8, 1
 ; RV64ZVE32F-NEXT:    vmv.x.s a2, v8
@@ -2857,7 +2865,7 @@ define <8 x i32> @mgather_baseidx_zext_v8i8_v8i32(ptr %base, <8 x i8> %idxs, <8
 ; RV64ZVE32F-NEXT:    vslideup.vi v10, v8, 3
 ; RV64ZVE32F-NEXT:    andi a2, a1, 16
 ; RV64ZVE32F-NEXT:    beqz a2, .LBB37_7
-; RV64ZVE32F-NEXT:  .LBB37_14: # %cond.load10
+; RV64ZVE32F-NEXT:  .LBB37_16: # %cond.load10
 ; RV64ZVE32F-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; RV64ZVE32F-NEXT:    vmv.x.s a2, v9
 ; RV64ZVE32F-NEXT:    andi a2, a2, 255
@@ -2871,32 +2879,6 @@ define <8 x i32> @mgather_baseidx_zext_v8i8_v8i32(ptr %base, <8 x i8> %idxs, <8
 ; RV64ZVE32F-NEXT:    andi a2, a1, 32
 ; RV64ZVE32F-NEXT:    bnez a2, .LBB37_8
 ; RV64ZVE32F-NEXT:    j .LBB37_9
-; RV64ZVE32F-NEXT:  .LBB37_15: # %cond.load16
-; RV64ZVE32F-NEXT:    vmv.x.s a2, v8
-; RV64ZVE32F-NEXT:    andi a2, a2, 255
-; RV64ZVE32F-NEXT:    slli a2, a2, 2
-; RV64ZVE32F-NEXT:    add a2, a0, a2
-; RV64ZVE32F-NEXT:    lw a2, 0(a2)
-; RV64ZVE32F-NEXT:    vsetvli zero, zero, e32, m1, ta, ma
-; RV64ZVE32F-NEXT:    vmv.s.x v12, a2
-; RV64ZVE32F-NEXT:    vsetivli zero, 7, e32, m2, tu, ma
-; RV64ZVE32F-NEXT:    vslideup.vi v10, v12, 6
-; RV64ZVE32F-NEXT:    andi a1, a1, -128
-; RV64ZVE32F-NEXT:    beqz a1, .LBB37_11
-; RV64ZVE32F-NEXT:  .LBB37_16: # %cond.load19
-; RV64ZVE32F-NEXT:    vsetivli zero, 1, e8, mf4, ta, ma
-; RV64ZVE32F-NEXT:    vslidedown.vi v8, v8, 1
-; RV64ZVE32F-NEXT:    vmv.x.s a1, v8
-; RV64ZVE32F-NEXT:    andi a1, a1, 255
-; RV64ZVE32F-NEXT:    slli a1, a1, 2
-; RV64ZVE32F-NEXT:    add a0, a0, a1
-; RV64ZVE32F-NEXT:    lw a0, 0(a0)
-; RV64ZVE32F-NEXT:    vsetvli zero, zero, e32, m1, ta, ma
-; RV64ZVE32F-NEXT:    vmv.s.x v8, a0
-; RV64ZVE32F-NEXT:    vsetivli zero, 8, e32, m2, ta, ma
-; RV64ZVE32F-NEXT:    vslideup.vi v10, v8, 7
-; RV64ZVE32F-NEXT:    vmv2r.v v8, v10
-; RV64ZVE32F-NEXT:    ret
   %eidxs = zext <8 x i8> %idxs to <8 x i32>
   %ptrs = getelementptr inbounds i32, ptr %base, <8 x i32> %eidxs
   %v = call <8 x i32> @llvm.masked.gather.v8i32.v8p0(<8 x ptr> %ptrs, i32 4, <8 x i1> %m, <8 x i32> %passthru)
@@ -2957,13 +2939,13 @@ define <8 x i32> @mgather_baseidx_v8i16_v8i32(ptr %base, <8 x i16> %idxs, <8 x i
 ; RV64ZVE32F-NEXT:    andi a2, a1, 4
 ; RV64ZVE32F-NEXT:    vsetivli zero, 2, e16, mf2, ta, ma
 ; RV64ZVE32F-NEXT:    vslidedown.vi v8, v8, 2
-; RV64ZVE32F-NEXT:    bnez a2, .LBB38_12
+; RV64ZVE32F-NEXT:    bnez a2, .LBB38_14
 ; RV64ZVE32F-NEXT:  # %bb.5: # %else5
 ; RV64ZVE32F-NEXT:    andi a2, a1, 8
-; RV64ZVE32F-NEXT:    bnez a2, .LBB38_13
+; RV64ZVE32F-NEXT:    bnez a2, .LBB38_15
 ; RV64ZVE32F-NEXT:  .LBB38_6: # %else8
 ; RV64ZVE32F-NEXT:    andi a2, a1, 16
-; RV64ZVE32F-NEXT:    bnez a2, .LBB38_14
+; RV64ZVE32F-NEXT:    bnez a2, .LBB38_16
 ; RV64ZVE32F-NEXT:  .LBB38_7: # %else11
 ; RV64ZVE32F-NEXT:    andi a2, a1, 32
 ; RV64ZVE32F-NEXT:    beqz a2, .LBB38_9
@@ -2982,14 +2964,35 @@ define <8 x i32> @mgather_baseidx_v8i16_v8i32(ptr %base, <8 x i16> %idxs, <8 x i
 ; RV64ZVE32F-NEXT:    andi a2, a1, 64
 ; RV64ZVE32F-NEXT:    vsetivli zero, 2, e16, mf2, ta, ma
 ; RV64ZVE32F-NEXT:    vslidedown.vi v8, v9, 2
-; RV64ZVE32F-NEXT:    bnez a2, .LBB38_15
-; RV64ZVE32F-NEXT:  # %bb.10: # %else17
+; RV64ZVE32F-NEXT:    beqz a2, .LBB38_11
+; RV64ZVE32F-NEXT:  # %bb.10: # %cond.load16
+; RV64ZVE32F-NEXT:    vmv.x.s a2, v8
+; RV64ZVE32F-NEXT:    slli a2, a2, 2
+; RV64ZVE32F-NEXT:    add a2, a0, a2
+; RV64ZVE32F-NEXT:    lw a2, 0(a2)
+; RV64ZVE32F-NEXT:    vsetvli zero, zero, e32, m1, ta, ma
+; RV64ZVE32F-NEXT:    vmv.s.x v12, a2
+; RV64ZVE32F-NEXT:    vsetivli zero, 7, e32, m2, tu, ma
+; RV64ZVE32F-NEXT:    vslideup.vi v10, v12, 6
+; RV64ZVE32F-NEXT:  .LBB38_11: # %else17
 ; RV64ZVE32F-NEXT:    andi a1, a1, -128
-; RV64ZVE32F-NEXT:    bnez a1, .LBB38_16
-; RV64ZVE32F-NEXT:  .LBB38_11: # %else20
+; RV64ZVE32F-NEXT:    beqz a1, .LBB38_13
+; RV64ZVE32F-NEXT:  # %bb.12: # %cond.load19
+; RV64ZVE32F-NEXT:    vsetivli zero, 1, e16, mf2, ta, ma
+; RV64ZVE32F-NEXT:    vslidedown.vi v8, v8, 1
+; RV64ZVE32F-NEXT:    vmv.x.s a1, v8
+; RV64ZVE32F-NEXT:    slli a1, a1, 2
+; RV64ZVE32F-NEXT:    add a0, a0, a1
+; RV64ZVE32F-NEXT:    lw a0, 0(a0)
+; RV64ZVE32F-NEXT:    vsetvli zero, zero, e32, m1, ta, ma
+; RV64ZVE32F-NEXT:    vmv.s.x v8, a0
+; RV64ZVE32F-NEXT:    vsetivli zero, 8, e32, m2, ta, ma
+; RV64ZVE32F-NEXT:    vslideup.vi v10, v8, 7
+; RV64ZVE32F-NEXT:  .LBB38_13: # %else20
+; RV64ZVE32F-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; RV64ZVE32F-NEXT:    vmv2r.v v8, v10
 ; RV64ZVE32F-NEXT:    ret
-; RV64ZVE32F-NEXT:  .LBB38_12: # %cond.load4
+; RV64ZVE32F-NEXT:  .LBB38_14: # %cond.load4
 ; RV64ZVE32F-NEXT:    vmv.x.s a2, v8
 ; RV64ZVE32F-NEXT:    slli a2, a2, 2
 ; RV64ZVE32F-NEXT:    add a2, a0, a2
@@ -3000,7 +3003,7 @@ define <8 x i32> @mgather_baseidx_v8i16_v8i32(ptr %base, <8 x i16> %idxs, <8 x i
 ; RV64ZVE32F-NEXT:    vslideup.vi v10, v12, 2
 ; RV64ZVE32F-NEXT:    andi a2, a1, 8
 ; RV64ZVE32F-NEXT:    beqz a2, .LBB38_6
-; RV64ZVE32F-NEXT:  .LBB38_13: # %cond.load7
+; RV64ZVE32F-NEXT:  .LBB38_15: # %cond.load7
 ; RV64ZVE32F-NEXT:    vsetivli zero, 1, e16, mf2, ta, ma
 ; RV64ZVE32F-NEXT:    vslidedown.vi v8, v8, 1
 ; RV64ZVE32F-NEXT:    vmv.x.s a2, v8
@@ -3013,7 +3016,7 @@ define <8 x i32> @mgather_baseidx_v8i16_v8i32(ptr %base, <8 x i16> %idxs, <8 x i
 ; RV64ZVE32F-NEXT:    vslideup.vi v10, v8, 3
 ; RV64ZVE32F-NEXT:    andi a2, a1, 16
 ; RV64ZVE32F-NEXT:    beqz a2, .LBB38_7
-; RV64ZVE32F-NEXT:  .LBB38_14: # %cond.load10
+; RV64ZVE32F-NEXT:  .LBB38_16: # %cond.load10
 ; RV64ZVE32F-NEXT:    vsetivli zero, 1, e16, m1, ta, ma
 ; RV64ZVE32F-NEXT:    vmv.x.s a2, v9
 ; RV64ZVE32F-NEXT:    slli a2, a2, 2
@@ -3026,30 +3029,6 @@ define <8 x i32> @mgather_baseidx_v8i16_v8i32(ptr %base, <8 x i16> %idxs, <8 x i
 ; RV64ZVE32F-NEXT:    andi a2, a1, 32
 ; RV64ZVE32F-NEXT:    bnez a2, .LBB38_8
 ; RV64ZVE32F-NEXT:    j .LBB38_9
-; RV64ZVE32F-NEXT:  .LBB38_15: # %cond.load16
-; RV64ZVE32F-NEXT:    vmv.x.s a2, v8
-; RV64ZVE32F-NEXT:    slli a2, a2, 2
-; RV64ZVE32F-NEXT:    add a2, a0, a2
-; RV64ZVE32F-NEXT:    lw a2, 0(a2)
-; RV64ZVE32F-NEXT:    vsetvli zero, zero, e32, m1, ta, ma
-; RV64ZVE32F-NEXT:    vmv.s.x v12, a2
-; RV64ZVE32F-NEXT:    vsetivli zero, 7, e32, m2, tu, ma
-; RV64ZVE32F-NEXT:    vslideup.vi v10, v12, 6
-; RV64ZVE32F-NEXT:    andi a1, a1, -128
-; RV64ZVE32F-NEXT:    beqz a1, .LBB38_11
-; RV64ZVE32F-NEXT:  .LBB38_16: # %cond.load19
-; RV64ZVE32F-NEXT:    vsetivli zero, 1, e16, mf2, ta, ma
-; RV64ZVE32F-NEXT:    vslidedown.vi v8, v8, 1
-; RV64ZVE32F-NEXT:    vmv.x.s a1, v8
-; RV64ZVE32F-NEXT:    slli a1, a1, 2
-; RV64ZVE32F-NEXT:    add a0, a0, a1
-; RV64ZVE32F-NEXT:    lw a0, 0(a0)
-; RV64ZVE32F-NEXT:    vsetvli zero, zero, e32, m1, ta, ma
-; RV64ZVE32F-NEXT:    vmv.s.x v8, a0
-; RV64ZVE32F-NEXT:    vsetivli zero, 8, e32, m2, ta, ma
-; RV64ZVE32F-NEXT:    vslideup.vi v10, v8, 7
-; RV64ZVE32F-NEXT:    vmv2r.v v8, v10
-; RV64ZVE32F-NEXT:    ret
   %ptrs = getelementptr inbounds i32, ptr %base, <8 x i16> %idxs
   %v = call <8 x i32> @llvm.masked.gather.v8i32.v8p0(<8 x ptr> %ptrs, i32 4, <8 x i1> %m, <8 x i32> %passthru)
   ret <8 x i32> %v
@@ -3109,13 +3088,13 @@ define <8 x i32> @mgather_baseidx_sext_v8i16_v8i32(ptr %base, <8 x i16> %idxs, <
 ; RV64ZVE32F-NEXT:    andi a2, a1, 4
 ; RV64ZVE32F-NEXT:    vsetivli zero, 2, e16, mf2, ta, ma
 ; RV64ZVE32F-NEXT:    vslidedown.vi v8, v8, 2
-; RV64ZVE32F-NEXT:    bnez a2, .LBB39_12
+; RV64ZVE32F-NEXT:    bnez a2, .LBB39_14
 ; RV64ZVE32F-NEXT:  # %bb.5: # %else5
 ; RV64ZVE32F-NEXT:    andi a2, a1, 8
-; RV64ZVE32F-NEXT:    bnez a2, .LBB39_13
+; RV64ZVE32F-NEXT:    bnez a2, .LBB39_15
 ; RV64ZVE32F-NEXT:  .LBB39_6: # %else8
 ; RV64ZVE32F-NEXT:    andi a2, a1, 16
-; RV64ZVE32F-NEXT:    bnez a2, .LBB39_14
+; RV64ZVE32F-NEXT:    bnez a2, .LBB39_16
 ; RV64ZVE32F-NEXT:  .LBB39_7: # %else11
 ; RV64ZVE32F-NEXT:    andi a2, a1, 32
 ; RV64ZVE32F-NEXT:    beqz a2, .LBB39_9
@@ -3134,14 +3113,35 @@ define <8 x i32> @mgather_baseidx_sext_v8i16_v8i32(ptr %base, <8 x i16> %idxs, <
 ; RV64ZVE32F-NEXT:    andi a2, a1, 64
 ; RV64ZVE32F-NEXT:    vsetivli zero, 2, e16, mf2, ta, ma
 ; RV64ZVE32F-NEXT:    vslidedown.vi v8, v9, 2
-; RV64ZVE32F-NEXT:    bnez a2, .LBB39_15
-; RV64ZVE32F-NEXT:  # %bb.10: # %else17
+; RV64ZVE32F-NEXT:    beqz a2, .LBB39_11
+; RV64ZVE32F-NEXT:  # %bb.10: # %cond.load16
+; RV64ZVE32F-NEXT:    vmv.x.s a2, v8
+; RV64ZVE32F-NEXT:    slli a2, a2, 2
+; RV64ZVE32F-NEXT:    add a2, a0, a2
+; RV64ZVE32F-NEXT:    lw a2, 0(a2)
+; RV64ZVE32F-NEXT:    vsetvli zero, zero, e32, m1, ta, ma
+; RV64ZVE32F-NEXT:    vmv.s.x v12, a2
+; RV64ZVE32F-NEXT:    vsetivli zero, 7, e32, m2, tu, ma
+; RV64ZVE32F-NEXT:    vslideup.vi v10, v12, 6
+; RV64ZVE32F-NEXT:  .LBB39_11: # %else17
 ; RV64ZVE32F-NEXT:    andi a1, a1, -128
-; RV64ZVE32F-NEXT:    bnez a1, .LBB39_16
-; RV64ZVE32F-NEXT:  .LBB39_11: # %else20
+; RV64ZVE32F-NEXT:    beqz a1, .LBB39_13
+; RV64ZVE32F-NEXT:  # %bb.12: # %cond.load19
+; RV64ZVE32F-NEXT:    vsetivli zero, 1, e16, mf2, ta, ma
+; RV64ZVE32F-NEXT:    vslidedown.vi v8, v8, 1
+; RV64ZVE32F-NEXT:    vmv.x.s a1, v8
+; RV64ZVE32F-NEXT:    slli a1, a1, 2
+; RV64ZVE32F-NEXT:    add a0, a0, a1
+; RV64ZVE32F-NEXT:    lw a0, 0(a0)
+; RV64ZVE32F-NEXT:    vsetvli zero, zero, e32, m1, ta, ma
+; RV64ZVE32F-NEXT:    vmv.s.x v8, a0
+; RV64ZVE32F-NEXT:    vsetivli zero, 8, e32, m2, ta, ma
+; RV64ZVE32F-NEXT:    vslideup.vi v10, v8, 7
+; RV64ZVE32F-NEXT:  .LBB39_13: # %else20
+; RV64ZVE32F-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; RV64ZVE32F-NEXT:    vmv2r.v v8, v10
 ; RV64ZVE32F-NEXT:    ret
-; RV64ZVE32F-NEXT:  .LBB39_12: # %cond.load4
+; RV64ZVE32F-NEXT:  .LBB39_14: # %cond.load4
 ; RV64ZVE32F-NEXT:    vmv.x.s a2, v8
 ; RV64ZVE32F-NEXT:    slli a2, a2, 2
 ; RV64ZVE32F-NEXT:    add a2, a0, a2
@@ -3152,7 +3152,7 @@ define <8 x i32> @mgather_baseidx_sext_v8i16_v8i32(ptr %base, <8 x i16> %idxs, <
 ; RV64ZVE32F-NEXT:    vslideup.vi v10, v12, 2
 ; RV64ZVE32F-NEXT:    andi a2, a1, 8
 ; RV64ZVE32F-NEXT:    beqz a2, .LBB39_6
-; RV64ZVE32F-NEXT:  .LBB39_13: # %cond.load7
+; RV64ZVE32F-NEXT:  .LBB39_15: # %cond.load7
 ; RV64ZVE32F-NEXT:    vsetivli zero, 1, e16, mf2, ta, ma
 ; RV64ZVE32F-NEXT:    vslidedown.vi v8, v8, 1
 ; RV64ZVE32F-NEXT:    vmv.x.s a2, v8
@@ -3165,7 +3165,7 @@ define <8 x i32> @mgather_baseidx_sext_v8i16_v8i32(ptr %base, <8 x i16> %idxs, <
 ; RV64ZVE32F-NEXT:    vslideup.vi v10, v8, 3
 ; RV64ZVE32F-NEXT:    andi a2, a1, 16
 ; RV64ZVE32F-NEXT:    beqz a2, .LBB39_7
-; RV64ZVE32F-NEXT:  .LBB39_14: # %cond.load10
+; RV64ZVE32F-NEXT:  .LBB39_16: # %cond.load10
 ; RV64ZVE32F-NEXT:    vsetivli zero, 1, e16, m1, ta, ma
 ; RV64ZVE32F-NEXT:    vmv.x.s a2, v9
 ; RV64ZVE32F-NEXT:    slli a2, a2, 2
@@ -3178,30 +3178,6 @@ define <8 x i32> @mgather_baseidx_sext_v8i16_v8i32(ptr %base, <8 x i16> %idxs, <
 ; RV64ZVE32F-NEXT:    andi a2, a1, 32
 ; RV64ZVE32F-NEXT:    bnez a2, .LBB39_8
 ; RV64ZVE32F-NEXT:    j .LBB39_9
-; RV64ZVE32F-NEXT:  .LBB39_15: # %cond.load16
-; RV64ZVE32F-NEXT:    vmv.x.s a2, v8
-; RV64ZVE32F-NEXT:    slli a2, a2, 2
-; RV64ZVE32F-NEXT:    add a2, a0, a2
-; RV64ZVE32F-NEXT:    lw a2, 0(a2)
-; RV64ZVE32F-NEXT:    vsetvli zero, zero, e32, m1, ta, ma
-; RV64ZVE32F-NEXT:    vmv.s.x v12, a2
-; RV64ZVE32F-NEXT:    vsetivli zero, 7, e32, m2, tu, ma
-; RV64ZVE32F-NEXT:    vslideup.vi v10, v12, 6
-; RV64ZVE32F-NEXT:    andi a1, a1, -128
-; RV64ZVE32F-NEXT:    beqz a1, .LBB39_11
-; RV64ZVE32F-NEXT:  .LBB39_16: # %cond.load19
-; RV64ZVE32F-NEXT:    vsetivli zero, 1, e16, mf2, ta, ma
-; RV64ZVE32F-NEXT:    vslidedown.vi v8, v8, 1
-; RV64ZVE32F-NEXT:    vmv.x.s a1, v8
-; RV64ZVE32F-NEXT:    slli a1, a1, 2
-; RV64ZVE32F-NEXT:    add a0, a0, a1
-; RV64ZVE32F-NEXT:    lw a0, 0(a0)
-; RV64ZVE32F-NEXT:    vsetvli zero, zero, e32, m1, ta, ma
-; RV64ZVE32F-NEXT:    vmv.s.x v8, a0
-; RV64ZVE32F-NEXT:    vsetivli zero, 8, e32, m2, ta, ma
-; RV64ZVE32F-NEXT:    vslideup.vi v10, v8, 7
-; RV64ZVE32F-NEXT:    vmv2r.v v8, v10
-; RV64ZVE32F-NEXT:    ret
   %eidxs = sext <8 x i16> %idxs to <8 x i32>
   %ptrs = getelementptr inbounds i32, ptr %base, <8 x i32> %eidxs
   %v = call <8 x i32> @llvm.masked.gather.v8i32.v8p0(<8 x ptr> %ptrs, i32 4, <8 x i1> %m, <8 x i32> %passthru)
@@ -3265,13 +3241,13 @@ define <8 x i32> @mgather_baseidx_zext_v8i16_v8i32(ptr %base, <8 x i16> %idxs, <
 ; RV64ZVE32F-NEXT:    andi a3, a2, 4
 ; RV64ZVE32F-NEXT:    vsetivli zero, 2, e16, mf2, ta, ma
 ; RV64ZVE32F-NEXT:    vslidedown.vi v8, v8, 2
-; RV64ZVE32F-NEXT:    bnez a3, .LBB40_12
+; RV64ZVE32F-NEXT:    bnez a3, .LBB40_14
 ; RV64ZVE32F-NEXT:  # %bb.5: # %else5
 ; RV64ZVE32F-NEXT:    andi a3, a2, 8
-; RV64ZVE32F-NEXT:    bnez a3, .LBB40_13
+; RV64ZVE32F-NEXT:    bnez a3, .LBB40_15
 ; RV64ZVE32F-NEXT:  .LBB40_6: # %else8
 ; RV64ZVE32F-NEXT:    andi a3, a2, 16
-; RV64ZVE32F-NEXT:    bnez a3, .LBB40_14
+; RV64ZVE32F-NEXT:    bnez a3, .LBB40_16
 ; RV64ZVE32F-NEXT:  .LBB40_7: # %else11
 ; RV64ZVE32F-NEXT:    andi a3, a2, 32
 ; RV64ZVE32F-NEXT:    beqz a3, .LBB40_9
@@ -3291,14 +3267,37 @@ define <8 x i32> @mgather_baseidx_zext_v8i16_v8i32(ptr %base, <8 x i16> %idxs, <
 ; RV64ZVE32F-NEXT:    andi a3, a2, 64
 ; RV64ZVE32F-NEXT:    vsetivli zero, 2, e16, mf2, ta, ma
 ; RV64ZVE32F-NEXT:    vslidedown.vi v8, v9, 2
-; RV64ZVE32F-NEXT:    bnez a3, .LBB40_15
-; RV64ZVE32F-NEXT:  # %bb.10: # %else17
+; RV64ZVE32F-NEXT:    beqz a3, .LBB40_11
+; RV64ZVE32F-NEXT:  # %bb.10: # %cond.load16
+; RV64ZVE32F-NEXT:    vmv.x.s a3, v8
+; RV64ZVE32F-NEXT:    and a3, a3, a1
+; RV64ZVE32F-NEXT:    slli a3, a3, 2
+; RV64ZVE32F-NEXT:    add a3, a0, a3
+; RV64ZVE32F-NEXT:    lw a3, 0(a3)
+; RV64ZVE32F-NEXT:    vsetvli zero, zero, e32, m1, ta, ma
+; RV64ZVE32F-NEXT:    vmv.s.x v12, a3
+; RV64ZVE32F-NEXT:    vsetivli zero, 7, e32, m2, tu, ma
+; RV64ZVE32F-NEXT:    vslideup.vi v10, v12, 6
+; RV64ZVE32F-NEXT:  .LBB40_11: # %else17
 ; RV64ZVE32F-NEXT:    andi a2, a2, -128
-; RV64ZVE32F-NEXT:    bnez a2, .LBB40_16
-; RV64ZVE32F-NEXT:  .LBB40_11: # %else20
+; RV64ZVE32F-NEXT:    beqz a2, .LBB40_13
+; RV64ZVE32F-NEXT:  # %bb.12: # %cond.load19
+; RV64ZVE32F-NEXT:    vsetivli zero, 1, e16, mf2, ta, ma
+; RV64ZVE32F-NEXT:    vslidedown.vi v8, v8, 1
+; RV64ZVE32F-NEXT:    vmv.x.s a2, v8
+; RV64ZVE32F-NEXT:    and a1, a2, a1
+; RV64ZVE32F-NEXT:    slli a1, a1, 2
+; RV64ZVE32F-NEXT:    add a0, a0, a1
+; RV64ZVE32F-NEXT:    lw a0, 0(a0)
+; RV64ZVE32F-NEXT:    vsetvli zero, zero, e32, m1, ta, ma
+; RV64ZVE32F-NEXT:    vmv.s.x v8, a0
+; RV64ZVE32F-NEXT:    vsetivli zero, 8, e32, m2, ta, ma
+; RV64ZVE32F-NEXT:    vslideup.vi v10, v8, 7
+; RV64ZVE32F-NEXT:  .LBB40_13: # %else20
+; RV64ZVE32F-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; RV64ZVE32F-NEXT:    vmv2r.v v8, v10
 ; RV64ZVE32F-NEXT:    ret
-; RV64ZVE32F-NEXT:  .LBB40_12: # %cond.load4
+; RV64ZVE32F-NEXT:  .LBB40_14: # %cond.load4
 ; RV64ZVE32F-NEXT:    vmv.x.s a3, v8
 ; RV64ZVE32F-NEXT:    and a3, a3, a1
 ; RV64ZVE32F-NEXT:    slli a3, a3, 2
@@ -3310,7 +3309,7 @@ define <8 x i32> @mgather_baseidx_zext_v8i16_v8i32(ptr %base, <8 x i16> %idxs, <
 ; RV64ZVE32F-NEXT:    vslideup.vi v10, v12, 2
 ; RV64ZVE32F-NEXT:    andi a3, a2, 8
 ; RV64ZVE32F-NEXT:    beqz a3, .LBB40_6
-; RV64ZVE32F-NEXT:  .LBB40_13: # %cond.load7
+; RV64ZVE32F-NEXT:  .LBB40_15: # %cond.load7
 ; RV64ZVE32F-NEXT:    vsetivli zero, 1, e16, mf2, ta, ma
 ; RV64ZVE32F-NEXT:    vslidedown.vi v8, v8, 1
 ; RV64ZVE32F-NEXT:    vmv.x.s a3, v8
@@ -3324,7 +3323,7 @@ define <8 x i32> @mgather_baseidx_zext_v8i16_v8i32(ptr %base, <8 x i16> %idxs, <
 ; RV64ZVE32F-NEXT:    vslideup.vi v10, v8, 3
 ; RV64ZVE32F-NEXT:    andi a3, a2, 16
 ; RV64ZVE32F-NEXT:    beqz a3, .LBB40_7
-; RV64ZVE32F-NEXT:  .LBB40_14: # %cond.load10
+; RV64ZVE32F-NEXT:  .LBB40_16: # %cond.load10
 ; RV64ZVE32F-NEXT:    vsetivli zero, 1, e16, m1, ta, ma
 ; RV64ZVE32F-NEXT:    vmv.x.s a3, v9
 ; RV64ZVE32F-NEXT:    and a3, a3, a1
@@ -3338,32 +3337,6 @@ define <8 x i32> @mgather_baseidx_zext_v8i16_v8i32(ptr %base, <8 x i16> %idxs, <
 ; RV64ZVE32F-NEXT:    andi a3, a2, 32
 ; RV64ZVE32F-NEXT:    bnez a3, .LBB40_8
 ; RV64ZVE32F-NEXT:    j .LBB40_9
-; RV64ZVE32F-NEXT:  .LBB40_15: # %cond.load16
-; RV64ZVE32F-NEXT:    vmv.x.s a3, v8
-; RV64ZVE32F-NEXT:    and a3, a3, a1
-; RV64ZVE32F-NEXT:    slli a3, a3, 2
-; RV64ZVE32F-NEXT:    add a3, a0, a3
-; RV64ZVE32F-NEXT:    lw a3, 0(a3)
-; RV64ZVE32F-NEXT:    vsetvli zero, zero, e32, m1, ta, ma
-; RV64ZVE32F-NEXT:    vmv.s.x v12, a3
-; RV64ZVE32F-NEXT:    vsetivli zero, 7, e32, m2, tu, ma
-; RV64ZVE32F-NEXT:    vslideup.vi v10, v12, 6
-; RV64ZVE32F-NEXT:    andi a2, a2, -128
-; RV64ZVE32F-NEXT:    beqz a2, .LBB40_11
-; RV64ZVE32F-NEXT:  .LBB40_16: # %cond.load19
-; RV64ZVE32F-NEXT:    vsetivli zero, 1, e16, mf2, ta, ma
-; RV64ZVE32F-NEXT:    vslidedown.vi v8, v8, 1
-; RV64ZVE32F-NEXT:    vmv.x.s a2, v8
-; RV64ZVE32F-NEXT:    and a1, a2, a1
-; RV64ZVE32F-NEXT:    slli a1, a1, 2
-; RV64ZVE32F-NEXT:    add a0, a0, a1
-; RV64ZVE32F-NEXT:    lw a0, 0(a0)
-; RV64ZVE32F-NEXT:    vsetvli zero, zero, e32, m1, ta, ma
-; RV64ZVE32F-NEXT:    vmv.s.x v8, a0
-; RV64ZVE32F-NEXT:    vsetivli zero, 8, e32, m2, ta, ma
-; RV64ZVE32F-NEXT:    vslideup.vi v10, v8, 7
-; RV64ZVE32F-NEXT:    vmv2r.v v8, v10
-; RV64ZVE32F-NEXT:    ret
   %eidxs = zext <8 x i16> %idxs to <8 x i32>
   %ptrs = getelementptr inbounds i32, ptr %base, <8 x i32> %eidxs
   %v = call <8 x i32> @llvm.masked.gather.v8i32.v8p0(<8 x ptr> %ptrs, i32 4, <8 x i1> %m, <8 x i32> %passthru)
@@ -3420,13 +3393,13 @@ define <8 x i32> @mgather_baseidx_v8i32(ptr %base, <8 x i32> %idxs, <8 x i1> %m,
 ; RV64ZVE32F-NEXT:    andi a2, a1, 4
 ; RV64ZVE32F-NEXT:    vsetivli zero, 2, e32, m1, ta, ma
 ; RV64ZVE32F-NEXT:    vslidedown.vi v8, v8, 2
-; RV64ZVE32F-NEXT:    bnez a2, .LBB41_12
+; RV64ZVE32F-NEXT:    bnez a2, .LBB41_14
 ; RV64ZVE32F-NEXT:  # %bb.5: # %else5
 ; RV64ZVE32F-NEXT:    andi a2, a1, 8
-; RV64ZVE32F-NEXT:    bnez a2, .LBB41_13
+; RV64ZVE32F-NEXT:    bnez a2, .LBB41_15
 ; RV64ZVE32F-NEXT:  .LBB41_6: # %else8
 ; RV64ZVE32F-NEXT:    andi a2, a1, 16
-; RV64ZVE32F-NEXT:    bnez a2, .LBB41_14
+; RV64ZVE32F-NEXT:    bnez a2, .LBB41_16
 ; RV64ZVE32F-NEXT:  .LBB41_7: # %else11
 ; RV64ZVE32F-NEXT:    andi a2, a1, 32
 ; RV64ZVE32F-NEXT:    beqz a2, .LBB41_9
@@ -3444,14 +3417,33 @@ define <8 x i32> @mgather_baseidx_v8i32(ptr %base, <8 x i32> %idxs, <8 x i1> %m,
 ; RV64ZVE32F-NEXT:    andi a2, a1, 64
 ; RV64ZVE32F-NEXT:    vsetivli zero, 2, e32, m1, ta, ma
 ; RV64ZVE32F-NEXT:    vslidedown.vi v8, v12, 2
-; RV64ZVE32F-NEXT:    bnez a2, .LBB41_15
-; RV64ZVE32F-NEXT:  # %bb.10: # %else17
+; RV64ZVE32F-NEXT:    beqz a2, .LBB41_11
+; RV64ZVE32F-NEXT:  # %bb.10: # %cond.load16
+; RV64ZVE32F-NEXT:    vmv.x.s a2, v8
+; RV64ZVE32F-NEXT:    slli a2, a2, 2
+; RV64ZVE32F-NEXT:    add a2, a0, a2
+; RV64ZVE32F-NEXT:    lw a2, 0(a2)
+; RV64ZVE32F-NEXT:    vmv.s.x v12, a2
+; RV64ZVE32F-NEXT:    vsetivli zero, 7, e32, m2, tu, ma
+; RV64ZVE32F-NEXT:    vslideup.vi v10, v12, 6
+; RV64ZVE32F-NEXT:  .LBB41_11: # %else17
 ; RV64ZVE32F-NEXT:    andi a1, a1, -128
-; RV64ZVE32F-NEXT:    bnez a1, .LBB41_16
-; RV64ZVE32F-NEXT:  .LBB41_11: # %else20
+; RV64ZVE32F-NEXT:    beqz a1, .LBB41_13
+; RV64ZVE32F-NEXT:  # %bb.12: # %cond.load19
+; RV64ZVE32F-NEXT:    vsetivli zero, 1, e32, m1, ta, ma
+; RV64ZVE32F-NEXT:    vslidedown.vi v8, v8, 1
+; RV64ZVE32F-NEXT:    vmv.x.s a1, v8
+; RV64ZVE32F-NEXT:    slli a1, a1, 2
+; RV64ZVE32F-NEXT:    add a0, a0, a1
+; RV64ZVE32F-NEXT:    lw a0, 0(a0)
+; RV64ZVE32F-NEXT:    vmv.s.x v8, a0
+; RV64ZVE32F-NEXT:    vsetivli zero, 8, e32, m2, ta, ma
+; RV64ZVE32F-NEXT:    vslideup.vi v10, v8, 7
+; RV64ZVE32F-NEXT:  .LBB41_13: # %else20
+; RV64ZVE32F-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; RV64ZVE32F-NEXT:    vmv2r.v v8, v10
 ; RV64ZVE32F-NEXT:    ret
-; RV64ZVE32F-NEXT:  .LBB41_12: # %cond.load4
+; RV64ZVE32F-NEXT:  .LBB41_14: # %cond.load4
 ; RV64ZVE32F-NEXT:    vmv.x.s a2, v8
 ; RV64ZVE32F-NEXT:    slli a2, a2, 2
 ; RV64ZVE32F-NEXT:    add a2, a0, a2
@@ -3461,7 +3453,7 @@ define <8 x i32> @mgather_baseidx_v8i32(ptr %base, <8 x i32> %idxs, <8 x i1> %m,
 ; RV64ZVE32F-NEXT:    vslideup.vi v10, v9, 2
 ; RV64ZVE32F-NEXT:    andi a2, a1, 8
 ; RV64ZVE32F-NEXT:    beqz a2, .LBB41_6
-; RV64ZVE32F-NEXT:  .LBB41_13: # %cond.load7
+; RV64ZVE32F-NEXT:  .LBB41_15: # %cond.load7
 ; RV64ZVE32F-NEXT:    vsetivli zero, 4, e32, m1, tu, ma
 ; RV64ZVE32F-NEXT:    vslidedown.vi v8, v8, 1
 ; RV64ZVE32F-NEXT:    vmv.x.s a2, v8
@@ -3472,7 +3464,7 @@ define <8 x i32> @mgather_baseidx_v8i32(ptr %base, <8 x i32> %idxs, <8 x i1> %m,
 ; RV64ZVE32F-NEXT:    vslideup.vi v10, v8, 3
 ; RV64ZVE32F-NEXT:    andi a2, a1, 16
 ; RV64ZVE32F-NEXT:    beqz a2, .LBB41_7
-; RV64ZVE32F-NEXT:  .LBB41_14: # %cond.load10
+; RV64ZVE32F-NEXT:  .LBB41_16: # %cond.load10
 ; RV64ZVE32F-NEXT:    vsetivli zero, 5, e32, m2, tu, ma
 ; RV64ZVE32F-NEXT:    vmv.x.s a2, v12
 ; RV64ZVE32F-NEXT:    slli a2, a2, 2
@@ -3483,28 +3475,6 @@ define <8 x i32> @mgather_baseidx_v8i32(ptr %base, <8 x i32> %idxs, <8 x i1> %m,
 ; RV64ZVE32F-NEXT:    andi a2, a1, 32
 ; RV64ZVE32F-NEXT:    bnez a2, .LBB41_8
 ; RV64ZVE32F-NEXT:    j .LBB41_9
-; RV64ZVE32F-NEXT:  .LBB41_15: # %cond.load16
-; RV64ZVE32F-NEXT:    vmv.x.s a2, v8
-; RV64ZVE32F-NEXT:    slli a2, a2, 2
-; RV64ZVE32F-NEXT:    add a2, a0, a2
-; RV64ZVE32F-NEXT:    lw a2, 0(a2)
-; RV64ZVE32F-NEXT:    vmv.s.x v12, a2
-; RV64ZVE32F-NEXT:    vsetivli zero, 7, e32, m2, tu, ma
-; RV64ZVE32F-NEXT:    vslideup.vi v10, v12, 6
-; RV64ZVE32F-NEXT:    andi a1, a1, -128
-; RV64ZVE32F-NEXT:    beqz a1, .LBB41_11
-; RV64ZVE32F-NEXT:  .LBB41_16: # %cond.load19
-; RV64ZVE32F-NEXT:    vsetivli zero, 1, e32, m1, ta, ma
-; RV64ZVE32F-NEXT:    vslidedown.vi v8, v8, 1
-; RV64ZVE32F-NEXT:    vmv.x.s a1, v8
-; RV64ZVE32F-NEXT:    slli a1, a1, 2
-; RV64ZVE32F-NEXT:    add a0, a0, a1
-; RV64ZVE32F-NEXT:    lw a0, 0(a0)
-; RV64ZVE32F-NEXT:    vmv.s.x v8, a0
-; RV64ZVE32F-NEXT:    vsetivli zero, 8, e32, m2, ta, ma
-; RV64ZVE32F-NEXT:    vslideup.vi v10, v8, 7
-; RV64ZVE32F-NEXT:    vmv2r.v v8, v10
-; RV64ZVE32F-NEXT:    ret
   %ptrs = getelementptr inbounds i32, ptr %base, <8 x i32> %idxs
   %v = call <8 x i32> @llvm.masked.gather.v8i32.v8p0(<8 x ptr> %ptrs, i32 4, <8 x i1> %m, <8 x i32> %passthru)
   ret <8 x i32> %v
@@ -3822,11 +3792,13 @@ define <4 x i64> @mgather_truemask_v4i64(<4 x ptr> %ptrs, <4 x i64> %passthru) {
 define <4 x i64> @mgather_falsemask_v4i64(<4 x ptr> %ptrs, <4 x i64> %passthru) {
 ; RV32V-LABEL: mgather_falsemask_v4i64:
 ; RV32V:       # %bb.0:
+; RV32V-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; RV32V-NEXT:    vmv2r.v v8, v10
 ; RV32V-NEXT:    ret
 ;
 ; RV64V-LABEL: mgather_falsemask_v4i64:
 ; RV64V:       # %bb.0:
+; RV64V-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; RV64V-NEXT:    vmv2r.v v8, v10
 ; RV64V-NEXT:    ret
 ;
@@ -7113,11 +7085,13 @@ define <4 x bfloat> @mgather_truemask_v4bf16(<4 x ptr> %ptrs, <4 x bfloat> %pass
 define <4 x bfloat> @mgather_falsemask_v4bf16(<4 x ptr> %ptrs, <4 x bfloat> %passthru) {
 ; RV32-LABEL: mgather_falsemask_v4bf16:
 ; RV32:       # %bb.0:
+; RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; RV32-NEXT:    vmv1r.v v8, v9
 ; RV32-NEXT:    ret
 ;
 ; RV64V-LABEL: mgather_falsemask_v4bf16:
 ; RV64V:       # %bb.0:
+; RV64V-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; RV64V-NEXT:    vmv1r.v v8, v10
 ; RV64V-NEXT:    ret
 ;
@@ -7295,13 +7269,13 @@ define <8 x bfloat> @mgather_baseidx_v8i8_v8bf16(ptr %base, <8 x i8> %idxs, <8 x
 ; RV64ZVE32F-NEXT:    andi a2, a1, 4
 ; RV64ZVE32F-NEXT:    vsetivli zero, 2, e8, mf4, ta, ma
 ; RV64ZVE32F-NEXT:    vslidedown.vi v8, v8, 2
-; RV64ZVE32F-NEXT:    bnez a2, .LBB64_12
+; RV64ZVE32F-NEXT:    bnez a2, .LBB64_14
 ; RV64ZVE32F-NEXT:  # %bb.5: # %else5
 ; RV64ZVE32F-NEXT:    andi a2, a1, 8
-; RV64ZVE32F-NEXT:    bnez a2, .LBB64_13
+; RV64ZVE32F-NEXT:    bnez a2, .LBB64_15
 ; RV64ZVE32F-NEXT:  .LBB64_6: # %else8
 ; RV64ZVE32F-NEXT:    andi a2, a1, 16
-; RV64ZVE32F-NEXT:    bnez a2, .LBB64_14
+; RV64ZVE32F-NEXT:    bnez a2, .LBB64_16
 ; RV64ZVE32F-NEXT:  .LBB64_7: # %else11
 ; RV64ZVE32F-NEXT:    andi a2, a1, 32
 ; RV64ZVE32F-NEXT:    beqz a2, .LBB64_9
@@ -7320,14 +7294,35 @@ define <8 x bfloat> @mgather_baseidx_v8i8_v8bf16(ptr %base, <8 x i8> %idxs, <8 x
 ; RV64ZVE32F-NEXT:    andi a2, a1, 64
 ; RV64ZVE32F-NEXT:    vsetivli zero, 2, e8, mf4, ta, ma
 ; RV64ZVE32F-NEXT:    vslidedown.vi v8, v10, 2
-; RV64ZVE32F-NEXT:    bnez a2, .LBB64_15
-; RV64ZVE32F-NEXT:  # %bb.10: # %else17
+; RV64ZVE32F-NEXT:    beqz a2, .LBB64_11
+; RV64ZVE32F-NEXT:  # %bb.10: # %cond.load16
+; RV64ZVE32F-NEXT:    vmv.x.s a2, v8
+; RV64ZVE32F-NEXT:    slli a2, a2, 1
+; RV64ZVE32F-NEXT:    add a2, a0, a2
+; RV64ZVE32F-NEXT:    lh a2, 0(a2)
+; RV64ZVE32F-NEXT:    vsetvli zero, zero, e16, mf2, ta, ma
+; RV64ZVE32F-NEXT:    vmv.s.x v10, a2
+; RV64ZVE32F-NEXT:    vsetivli zero, 7, e16, m1, tu, ma
+; RV64ZVE32F-NEXT:    vslideup.vi v9, v10, 6
+; RV64ZVE32F-NEXT:  .LBB64_11: # %else17
 ; RV64ZVE32F-NEXT:    andi a1, a1, -128
-; RV64ZVE32F-NEXT:    bnez a1, .LBB64_16
-; RV64ZVE32F-NEXT:  .LBB64_11: # %else20
+; RV64ZVE32F-NEXT:    beqz a1, .LBB64_13
+; RV64ZVE32F-NEXT:  # %bb.12: # %cond.load19
+; RV64ZVE32F-NEXT:    vsetivli zero, 1, e8, mf4, ta, ma
+; RV64ZVE32F-NEXT:    vslidedown.vi v8, v8, 1
+; RV64ZVE32F-NEXT:    vmv.x.s a1, v8
+; RV64ZVE32F-NEXT:    slli a1, a1, 1
+; RV64ZVE32F-NEXT:    add a0, a0, a1
+; RV64ZVE32F-NEXT:    lh a0, 0(a0)
+; RV64ZVE32F-NEXT:    vsetvli zero, zero, e16, mf2, ta, ma
+; RV64ZVE32F-NEXT:    vmv.s.x v8, a0
+; RV64ZVE32F-NEXT:    vsetivli zero, 8, e16, m1, ta, ma
+; RV64ZVE32F-NEXT:    vslideup.vi v9, v8, 7
+; RV64ZVE32F-NEXT:  .LBB64_13: # %else20
+; RV64ZVE32F-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; RV64ZVE32F-NEXT:    vmv1r.v v8, v9
 ; RV64ZVE32F-NEXT:    ret
-; RV64ZVE32F-NEXT:  .LBB64_12: # %cond.load4
+; RV64ZVE32F-NEXT:  .LBB64_14: # %cond.load4
 ; RV64ZVE32F-NEXT:    vmv.x.s a2, v8
 ; RV64ZVE32F-NEXT:    slli a2, a2, 1
 ; RV64ZVE32F-NEXT:    add a2, a0, a2
@@ -7338,7 +7333,7 @@ define <8 x bfloat> @mgather_baseidx_v8i8_v8bf16(ptr %base, <8 x i8> %idxs, <8 x
 ; RV64ZVE32F-NEXT:    vslideup.vi v9, v11, 2
 ; RV64ZVE32F-NEXT:    andi a2, a1, 8
 ; RV64ZVE32F-NEXT:    beqz a2, .LBB64_6
-; RV64ZVE32F-NEXT:  .LBB64_13: # %cond.load7
+; RV64ZVE32F-NEXT:  .LBB64_15: # %cond.load7
 ; RV64ZVE32F-NEXT:    vsetivli zero, 1, e8, mf4, ta, ma
 ; RV64ZVE32F-NEXT:    vslidedown.vi v8, v8, 1
 ; RV64ZVE32F-NEXT:    vmv.x.s a2, v8
@@ -7351,7 +7346,7 @@ define <8 x bfloat> @mgather_baseidx_v8i8_v8bf16(ptr %base, <8 x i8> %idxs, <8 x
 ; RV64ZVE32F-NEXT:    vslideup.vi v9, v8, 3
 ; RV64ZVE32F-NEXT:    andi a2, a1, 16
 ; RV64ZVE32F-NEXT:    beqz a2, .LBB64_7
-; RV64ZVE32F-NEXT:  .LBB64_14: # %cond.load10
+; RV64ZVE32F-NEXT:  .LBB64_16: # %cond.load10
 ; RV64ZVE32F-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; RV64ZVE32F-NEXT:    vmv.x.s a2, v10
 ; RV64ZVE32F-NEXT:    slli a2, a2, 1
@@ -7364,30 +7359,6 @@ define <8 x bfloat> @mgather_baseidx_v8i8_v8bf16(ptr %base, <8 x i8> %idxs, <8 x
 ; RV64ZVE32F-NEXT:    andi a2, a1, 32
 ; RV64ZVE32F-NEXT:    bnez a2, .LBB64_8
 ; RV64ZVE32F-NEXT:    j .LBB64_9
-; RV64ZVE32F-NEXT:  .LBB64_15: # %cond.load16
-; RV64ZVE32F-NEXT:    vmv.x.s a2, v8
-; RV64ZVE32F-NEXT:    slli a2, a2, 1
-; RV64ZVE32F-NEXT:    add a2, a0, a2
-; RV64ZVE32F-NEXT:    lh a2, 0(a2)
-; RV64ZVE32F-NEXT:    vsetvli zero, zero, e16, mf2, ta, ma
-; RV64ZVE32F-NEXT:    vmv.s.x v10, a2
-; RV64ZVE32F-NEXT:    vsetivli zero, 7, e16, m1, tu, ma
-; RV64ZVE32F-NEXT:    vslideup.vi v9, v10, 6
-; RV64ZVE32F-NEXT:    andi a1, a1, -128
-; RV64ZVE32F-NEXT:    beqz a1, .LBB64_11
-; RV64ZVE32F-NEXT:  .LBB64_16: # %cond.load19
-; RV64ZVE32F-NEXT:    vsetivli zero, 1, e8, mf4, ta, ma
-; RV64ZVE32F-NEXT:    vslidedown.vi v8, v8, 1
-; RV64ZVE32F-NEXT:    vmv.x.s a1, v8
-; RV64ZVE32F-NEXT:    slli a1, a1, 1
-; RV64ZVE32F-NEXT:    add a0, a0, a1
-; RV64ZVE32F-NEXT:    lh a0, 0(a0)
-; RV64ZVE32F-NEXT:    vsetvli zero, zero, e16, mf2, ta, ma
-; RV64ZVE32F-NEXT:    vmv.s.x v8, a0
-; RV64ZVE32F-NEXT:    vsetivli zero, 8, e16, m1, ta, ma
-; RV64ZVE32F-NEXT:    vslideup.vi v9, v8, 7
-; RV64ZVE32F-NEXT:    vmv1r.v v8, v9
-; RV64ZVE32F-NEXT:    ret
   %ptrs = getelementptr inbounds bfloat, ptr %base, <8 x i8> %idxs
   %v = call <8 x bfloat> @llvm.masked.gather.v8bf16.v8p0(<8 x ptr> %ptrs, i32 2, <8 x i1> %m, <8 x bfloat> %passthru)
   ret <8 x bfloat> %v
@@ -7447,13 +7418,13 @@ define <8 x bfloat> @mgather_baseidx_sext_v8i8_v8bf16(ptr %base, <8 x i8> %idxs,
 ; RV64ZVE32F-NEXT:    andi a2, a1, 4
 ; RV64ZVE32F-NEXT:    vsetivli zero, 2, e8, mf4, ta, ma
 ; RV64ZVE32F-NEXT:    vslidedown.vi v8, v8, 2
-; RV64ZVE32F-NEXT:    bnez a2, .LBB65_12
+; RV64ZVE32F-NEXT:    bnez a2, .LBB65_14
 ; RV64ZVE32F-NEXT:  # %bb.5: # %else5
 ; RV64ZVE32F-NEXT:    andi a2, a1, 8
-; RV64ZVE32F-NEXT:    bnez a2, .LBB65_13
+; RV64ZVE32F-NEXT:    bnez a2, .LBB65_15
 ; RV64ZVE32F-NEXT:  .LBB65_6: # %else8
 ; RV64ZVE32F-NEXT:    andi a2, a1, 16
-; RV64ZVE32F-NEXT:    bnez a2, .LBB65_14
+; RV64ZVE32F-NEXT:    bnez a2, .LBB65_16
 ; RV64ZVE32F-NEXT:  .LBB65_7: # %else11
 ; RV64ZVE32F-NEXT:    andi a2, a1, 32
 ; RV64ZVE32F-NEXT:    beqz a2, .LBB65_9
@@ -7472,14 +7443,35 @@ define <8 x bfloat> @mgather_baseidx_sext_v8i8_v8bf16(ptr %base, <8 x i8> %idxs,
 ; RV64ZVE32F-NEXT:    andi a2, a1, 64
 ; RV64ZVE32F-NEXT:    vsetivli zero, 2, e8, mf4, ta, ma
 ; RV64ZVE32F-NEXT:    vslidedown.vi v8, v10, 2
-; RV64ZVE32F-NEXT:    bnez a2, .LBB65_15
-; RV64ZVE32F-NEXT:  # %bb.10: # %else17
+; RV64ZVE32F-NEXT:    beqz a2, .LBB65_11
+; RV64ZVE32F-NEXT:  # %bb.10: # %cond.load16
+; RV64ZVE32F-NEXT:    vmv.x.s a2, v8
+; RV64ZVE32F-NEXT:    slli a2, a2, 1
+; RV64ZVE32F-NEXT:    add a2, a0, a2
+; RV64ZVE32F-NEXT:    lh a2, 0(a2)
+; RV64ZVE32F-NEXT:    vsetvli zero, zero, e16, mf2, ta, ma
+; RV64ZVE32F-NEXT:    vmv.s.x v10, a2
+; RV64ZVE32F-NEXT:    vsetivli zero, 7, e16, m1, tu, ma
+; RV64ZVE32F-NEXT:    vslideup.vi v9, v10, 6
+; RV64ZVE32F-NEXT:  .LBB65_11: # %else17
 ; RV64ZVE32F-NEXT:    andi a1, a1, -128
-; RV64ZVE32F-NEXT:    bnez a1, .LBB65_16
-; RV64ZVE32F-NEXT:  .LBB65_11: # %else20
+; RV64ZVE32F-NEXT:    beqz a1, .LBB65_13
+; RV64ZVE32F-NEXT:  # %bb.12: # %cond.load19
+; RV64ZVE32F-NEXT:    vsetivli zero, 1, e8, mf4, ta, ma
+; RV64ZVE32F-NEXT:    vslidedown.vi v8, v8, 1
+; RV64ZVE32F-NEXT:    vmv.x.s a1, v8
+; RV64ZVE32F-NEXT:    slli a1, a1, 1
+; RV64ZVE32F-NEXT:    add a0, a0, a1
+; RV64ZVE32F-NEXT:    lh a0, 0(a0)
+; RV64ZVE32F-NEXT:    vsetvli zero, zero, e16, mf2, ta, ma
+; RV64ZVE32F-NEXT:    vmv.s.x v8, a0
+; RV64ZVE32F-NEXT:    vsetivli zero, 8, e16, m1, ta, ma
+; RV64ZVE32F-NEXT:    vslideup.vi v9, v8, 7
+; RV64ZVE32F-NEXT:  .LBB65_13: # %else20
+; RV64ZVE32F-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; RV64ZVE32F-NEXT:    vmv1r.v v8, v9
 ; RV64ZVE32F-NEXT:    ret
-; RV64ZVE32F-NEXT:  .LBB65_12: # %cond.load4
+; RV64ZVE32F-NEXT:  .LBB65_14: # %cond.load4
 ; RV64ZVE32F-NEXT:    vmv.x.s a2, v8
 ; RV64ZVE32F-NEXT:    slli a2, a2, 1
 ; RV64ZVE32F-NEXT:    add a2, a0, a2
@@ -7490,7 +7482,7 @@ define <8 x bfloat> @mgather_baseidx_sext_v8i8_v8bf16(ptr %base, <8 x i8> %idxs,
 ; RV64ZVE32F-NEXT:    vslideup.vi v9, v11, 2
 ; RV64ZVE32F-NEXT:    andi a2, a1, 8
 ; RV64ZVE32F-NEXT:    beqz a2, .LBB65_6
-; RV64ZVE32F-NEXT:  .LBB65_13: # %cond.load7
+; RV64ZVE32F-NEXT:  .LBB65_15: # %cond.load7
 ; RV64ZVE32F-NEXT:    vsetivli zero, 1, e8, mf4, ta, ma
 ; RV64ZVE32F-NEXT:    vslidedown.vi v8, v8, 1
 ; RV64ZVE32F-NEXT:    vmv.x.s a2, v8
@@ -7503,7 +7495,7 @@ define <8 x bfloat> @mgather_baseidx_sext_v8i8_v8bf16(ptr %base, <8 x i8> %idxs,
 ; RV64ZVE32F-NEXT:    vslideup.vi v9, v8, 3
 ; RV64ZVE32F-NEXT:    andi a2, a1, 16
 ; RV64ZVE32F-NEXT:    beqz a2, .LBB65_7
-; RV64ZVE32F-NEXT:  .LBB65_14: # %cond.load10
+; RV64ZVE32F-NEXT:  .LBB65_16: # %cond.load10
 ; RV64ZVE32F-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; RV64ZVE32F-NEXT:    vmv.x.s a2, v10
 ; RV64ZVE32F-NEXT:    slli a2, a2, 1
@@ -7516,30 +7508,6 @@ define <8 x bfloat> @mgather_baseidx_sext_v8i8_v8bf16(ptr %base, <8 x i8> %idxs,
 ; RV64ZVE32F-NEXT:    andi a2, a1, 32
 ; RV64ZVE32F-NEXT:    bnez a2, .LBB65_8
 ; RV64ZVE32F-NEXT:    j .LBB65_9
-; RV64ZVE32F-NEXT:  .LBB65_15: # %cond.load16
-; RV64ZVE32F-NEXT:    vmv.x.s a2, v8
-; RV64ZVE32F-NEXT:    slli a2, a2, 1
-; RV64ZVE32F-NEXT:    add a2, a0, a2
-; RV64ZVE32F-NEXT:    lh a2, 0(a2)
-; RV64ZVE32F-NEXT:    vsetvli zero, zero, e16, mf2, ta, ma
-; RV64ZVE32F-NEXT:    vmv.s.x v10, a2
-; RV64ZVE32F-NEXT:    vsetivli zero, 7, e16, m1, tu, ma
-; RV64ZVE32F-NEXT:    vslideup.vi v9, v10, 6
-; RV64ZVE32F-NEXT:    andi a1, a1, -128
-; RV64ZVE32F-NEXT:    beqz a1, .LBB65_11
-; RV64ZVE32F-NEXT:  .LBB65_16: # %cond.load19
-; RV64ZVE32F-NEXT:    vsetivli zero, 1, e8, mf4, ta, ma
-; RV64ZVE32F-NEXT:    vslidedown.vi v8, v8, 1
-; RV64ZVE32F-NEXT:    vmv.x.s a1, v8
-; RV64ZVE32F-NEXT:    slli a1, a1, 1
-; RV64ZVE32F-NEXT:    add a0, a0, a1
-; RV64ZVE32F-NEXT:    lh a0, 0(a0)
-; RV64ZVE32F-NEXT:    vsetvli zero, zero, e16, mf2, ta, ma
-; RV64ZVE32F-NEXT:    vmv.s.x v8, a0
-; RV64ZVE32F-NEXT:    vsetivli zero, 8, e16, m1, ta, ma
-; RV64ZVE32F-NEXT:    vslideup.vi v9, v8, 7
-; RV64ZVE32F-NEXT:    vmv1r.v v8, v9
-; RV64ZVE32F-NEXT:    ret
   %eidxs = sext <8 x i8> %idxs to <8 x i16>
   %ptrs = getelementptr inbounds bfloat, ptr %base, <8 x i16> %eidxs
   %v = call <8 x bfloat> @llvm.masked.gather.v8bf16.v8p0(<8 x ptr> %ptrs, i32 2, <8 x i1> %m, <8 x bfloat> %passthru)
@@ -7600,13 +7568,13 @@ define <8 x bfloat> @mgather_baseidx_zext_v8i8_v8bf16(ptr %base, <8 x i8> %idxs,
 ; RV64ZVE32F-NEXT:    andi a2, a1, 4
 ; RV64ZVE32F-NEXT:    vsetivli zero, 2, e8, mf4, ta, ma
 ; RV64ZVE32F-NEXT:    vslidedown.vi v8, v8, 2
-; RV64ZVE32F-NEXT:    bnez a2, .LBB66_12
+; RV64ZVE32F-NEXT:    bnez a2, .LBB66_14
 ; RV64ZVE32F-NEXT:  # %bb.5: # %else5
 ; RV64ZVE32F-NEXT:    andi a2, a1, 8
-; RV64ZVE32F-NEXT:    bnez a2, .LBB66_13
+; RV64ZVE32F-NEXT:    bnez a2, .LBB66_15
 ; RV64ZVE32F-NEXT:  .LBB66_6: # %else8
 ; RV64ZVE32F-NEXT:    andi a2, a1, 16
-; RV64ZVE32F-NEXT:    bnez a2, .LBB66_14
+; RV64ZVE32F-NEXT:    bnez a2, .LBB66_16
 ; RV64ZVE32F-NEXT:  .LBB66_7: # %else11
 ; RV64ZVE32F-NEXT:    andi a2, a1, 32
 ; RV64ZVE32F-NEXT:    beqz a2, .LBB66_9
@@ -7626,14 +7594,37 @@ define <8 x bfloat> @mgather_baseidx_zext_v8i8_v8bf16(ptr %base, <8 x i8> %idxs,
 ; RV64ZVE32F-NEXT:    andi a2, a1, 64
 ; RV64ZVE32F-NEXT:    vsetivli zero, 2, e8, mf4, ta, ma
 ; RV64ZVE32F-NEXT:    vslidedown.vi v8, v10, 2
-; RV64ZVE32F-NEXT:    bnez a2, .LBB66_15
-; RV64ZVE32F-NEXT:  # %bb.10: # %else17
+; RV64ZVE32F-NEXT:    beqz a2, .LBB66_11
+; RV64ZVE32F-NEXT:  # %bb.10: # %cond.load16
+; RV64ZVE32F-NEXT:    vmv.x.s a2, v8
+; RV64ZVE32F-NEXT:    andi a2, a2, 255
+; RV64ZVE32F-NEXT:    slli a2, a2, 1
+; RV64ZVE32F-NEXT:    add a2, a0, a2
+; RV64ZVE32F-NEXT:    lh a2, 0(a2)
+; RV64ZVE32F-NEXT:    vsetvli zero, zero, e16, mf2, ta, ma
+; RV64ZVE32F-NEXT:    vmv.s.x v10, a2
+; RV64ZVE32F-NEXT:    vsetivli zero, 7, e16, m1, tu, ma
+; RV64ZVE32F-NEXT:    vslideup.vi v9, v10, 6
+; RV64ZVE32F-NEXT:  .LBB66_11: # %else17
 ; RV64ZVE32F-NEXT:    andi a1, a1, -128
-; RV64ZVE32F-NEXT:    bnez a1, .LBB66_16
-; RV64ZVE32F-NEXT:  .LBB66_11: # %else20
+; RV64ZVE32F-NEXT:    beqz a1, .LBB66_13
+; RV64ZVE32F-NEXT:  # %bb.12: # %cond.load19
+; RV64ZVE32F-NEXT:    vsetivli zero, 1, e8, mf4, ta, ma
+; RV64ZVE32F-NEXT:    vslidedown.vi v8, v8, 1
+; RV64ZVE32F-NEXT:    vmv.x.s a1, v8
+; RV64ZVE32F-NEXT:    andi a1, a1, 255
+; RV64ZVE32F-NEXT:    slli a1, a1, 1
+; RV64ZVE32F-NEXT:    add a0, a0, a1
+; RV64ZVE32F-NEXT:    lh a0, 0(a0)
+; RV64ZVE32F-NEXT:    vsetvli zero, zero, e16, mf2, ta, ma
+; RV64ZVE32F-NEXT:    vmv.s.x v8, a0
+; RV64ZVE32F-NEXT:    vsetivli zero, 8, e16, m1, ta, ma
+; RV64ZVE32F-NEXT:    vslideup.vi v9, v8, 7
+; RV64ZVE32F-NEXT:  .LBB66_13: # %else20
+; RV64ZVE32F-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; RV64ZVE32F-NEXT:    vmv1r.v v8, v9
 ; RV64ZVE32F-NEXT:    ret
-; RV64ZVE32F-NEXT:  .LBB66_12: # %cond.load4
+; RV64ZVE32F-NEXT:  .LBB66_14: # %cond.load4
 ; RV64ZVE32F-NEXT:    vmv.x.s a2, v8
 ; RV64ZVE32F-NEXT:    andi a2, a2, 255
 ; RV64ZVE32F-NEXT:    slli a2, a2, 1
@@ -7645,7 +7636,7 @@ define <8 x bfloat> @mgather_baseidx_zext_v8i8_v8bf16(ptr %base, <8 x i8> %idxs,
 ; RV64ZVE32F-NEXT:    vslideup.vi v9, v11, 2
 ; RV64ZVE32F-NEXT:    andi a2, a1, 8
 ; RV64ZVE32F-NEXT:    beqz a2, .LBB66_6
-; RV64ZVE32F-NEXT:  .LBB66_13: # %cond.load7
+; RV64ZVE32F-NEXT:  .LBB66_15: # %cond.load7
 ; RV64ZVE32F-NEXT:    vsetivli zero, 1, e8, mf4, ta, ma
 ; RV64ZVE32F-NEXT:    vslidedown.vi v8, v8, 1
 ; RV64ZVE32F-NEXT:    vmv.x.s a2, v8
@@ -7658,47 +7649,21 @@ define <8 x bfloat> @mgather_baseidx_zext_v8i8_v8bf16(ptr %base, <8 x i8> %idxs,
 ; RV64ZVE32F-NEXT:    vsetivli zero, 4, e16, m1, tu, ma
 ; RV64ZVE32F-NEXT:    vslideup.vi v9, v8, 3
 ; RV64ZVE32F-NEXT:    andi a2, a1, 16
-; RV64ZVE32F-NEXT:    beqz a2, .LBB66_7
-; RV64ZVE32F-NEXT:  .LBB66_14: # %cond.load10
-; RV64ZVE32F-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
-; RV64ZVE32F-NEXT:    vmv.x.s a2, v10
-; RV64ZVE32F-NEXT:    andi a2, a2, 255
-; RV64ZVE32F-NEXT:    slli a2, a2, 1
-; RV64ZVE32F-NEXT:    add a2, a0, a2
-; RV64ZVE32F-NEXT:    lh a2, 0(a2)
-; RV64ZVE32F-NEXT:    vsetvli zero, zero, e16, m2, ta, ma
-; RV64ZVE32F-NEXT:    vmv.s.x v8, a2
-; RV64ZVE32F-NEXT:    vsetivli zero, 5, e16, m1, tu, ma
-; RV64ZVE32F-NEXT:    vslideup.vi v9, v8, 4
-; RV64ZVE32F-NEXT:    andi a2, a1, 32
-; RV64ZVE32F-NEXT:    bnez a2, .LBB66_8
-; RV64ZVE32F-NEXT:    j .LBB66_9
-; RV64ZVE32F-NEXT:  .LBB66_15: # %cond.load16
-; RV64ZVE32F-NEXT:    vmv.x.s a2, v8
-; RV64ZVE32F-NEXT:    andi a2, a2, 255
-; RV64ZVE32F-NEXT:    slli a2, a2, 1
-; RV64ZVE32F-NEXT:    add a2, a0, a2
-; RV64ZVE32F-NEXT:    lh a2, 0(a2)
-; RV64ZVE32F-NEXT:    vsetvli zero, zero, e16, mf2, ta, ma
-; RV64ZVE32F-NEXT:    vmv.s.x v10, a2
-; RV64ZVE32F-NEXT:    vsetivli zero, 7, e16, m1, tu, ma
-; RV64ZVE32F-NEXT:    vslideup.vi v9, v10, 6
-; RV64ZVE32F-NEXT:    andi a1, a1, -128
-; RV64ZVE32F-NEXT:    beqz a1, .LBB66_11
-; RV64ZVE32F-NEXT:  .LBB66_16: # %cond.load19
-; RV64ZVE32F-NEXT:    vsetivli zero, 1, e8, mf4, ta, ma
-; RV64ZVE32F-NEXT:    vslidedown.vi v8, v8, 1
-; RV64ZVE32F-NEXT:    vmv.x.s a1, v8
-; RV64ZVE32F-NEXT:    andi a1, a1, 255
-; RV64ZVE32F-NEXT:    slli a1, a1, 1
-; RV64ZVE32F-NEXT:    add a0, a0, a1
-; RV64ZVE32F-NEXT:    lh a0, 0(a0)
-; RV64ZVE32F-NEXT:    vsetvli zero, zero, e16, mf2, ta, ma
-; RV64ZVE32F-NEXT:    vmv.s.x v8, a0
-; RV64ZVE32F-NEXT:    vsetivli zero, 8, e16, m1, ta, ma
-; RV64ZVE32F-NEXT:    vslideup.vi v9, v8, 7
-; RV64ZVE32F-NEXT:    vmv1r.v v8, v9
-; RV64ZVE32F-NEXT:    ret
+; RV64ZVE32F-NEXT:    beqz a2, .LBB66_7
+; RV64ZVE32F-NEXT:  .LBB66_16: # %cond.load10
+; RV64ZVE32F-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
+; RV64ZVE32F-NEXT:    vmv.x.s a2, v10
+; RV64ZVE32F-NEXT:    andi a2, a2, 255
+; RV64ZVE32F-NEXT:    slli a2, a2, 1
+; RV64ZVE32F-NEXT:    add a2, a0, a2
+; RV64ZVE32F-NEXT:    lh a2, 0(a2)
+; RV64ZVE32F-NEXT:    vsetvli zero, zero, e16, m2, ta, ma
+; RV64ZVE32F-NEXT:    vmv.s.x v8, a2
+; RV64ZVE32F-NEXT:    vsetivli zero, 5, e16, m1, tu, ma
+; RV64ZVE32F-NEXT:    vslideup.vi v9, v8, 4
+; RV64ZVE32F-NEXT:    andi a2, a1, 32
+; RV64ZVE32F-NEXT:    bnez a2, .LBB66_8
+; RV64ZVE32F-NEXT:    j .LBB66_9
   %eidxs = zext <8 x i8> %idxs to <8 x i16>
   %ptrs = getelementptr inbounds bfloat, ptr %base, <8 x i16> %eidxs
   %v = call <8 x bfloat> @llvm.masked.gather.v8bf16.v8p0(<8 x ptr> %ptrs, i32 2, <8 x i1> %m, <8 x bfloat> %passthru)
@@ -7756,13 +7721,13 @@ define <8 x bfloat> @mgather_baseidx_v8bf16(ptr %base, <8 x i16> %idxs, <8 x i1>
 ; RV64ZVE32F-NEXT:    andi a2, a1, 4
 ; RV64ZVE32F-NEXT:    vsetivli zero, 2, e16, mf2, ta, ma
 ; RV64ZVE32F-NEXT:    vslidedown.vi v8, v8, 2
-; RV64ZVE32F-NEXT:    bnez a2, .LBB67_12
+; RV64ZVE32F-NEXT:    bnez a2, .LBB67_14
 ; RV64ZVE32F-NEXT:  # %bb.5: # %else5
 ; RV64ZVE32F-NEXT:    andi a2, a1, 8
-; RV64ZVE32F-NEXT:    bnez a2, .LBB67_13
+; RV64ZVE32F-NEXT:    bnez a2, .LBB67_15
 ; RV64ZVE32F-NEXT:  .LBB67_6: # %else8
 ; RV64ZVE32F-NEXT:    andi a2, a1, 16
-; RV64ZVE32F-NEXT:    bnez a2, .LBB67_14
+; RV64ZVE32F-NEXT:    bnez a2, .LBB67_16
 ; RV64ZVE32F-NEXT:  .LBB67_7: # %else11
 ; RV64ZVE32F-NEXT:    andi a2, a1, 32
 ; RV64ZVE32F-NEXT:    beqz a2, .LBB67_9
@@ -7780,14 +7745,33 @@ define <8 x bfloat> @mgather_baseidx_v8bf16(ptr %base, <8 x i16> %idxs, <8 x i1>
 ; RV64ZVE32F-NEXT:    andi a2, a1, 64
 ; RV64ZVE32F-NEXT:    vsetivli zero, 2, e16, mf2, ta, ma
 ; RV64ZVE32F-NEXT:    vslidedown.vi v8, v10, 2
-; RV64ZVE32F-NEXT:    bnez a2, .LBB67_15
-; RV64ZVE32F-NEXT:  # %bb.10: # %else17
+; RV64ZVE32F-NEXT:    beqz a2, .LBB67_11
+; RV64ZVE32F-NEXT:  # %bb.10: # %cond.load16
+; RV64ZVE32F-NEXT:    vmv.x.s a2, v8
+; RV64ZVE32F-NEXT:    slli a2, a2, 1
+; RV64ZVE32F-NEXT:    add a2, a0, a2
+; RV64ZVE32F-NEXT:    lh a2, 0(a2)
+; RV64ZVE32F-NEXT:    vmv.s.x v10, a2
+; RV64ZVE32F-NEXT:    vsetivli zero, 7, e16, m1, tu, ma
+; RV64ZVE32F-NEXT:    vslideup.vi v9, v10, 6
+; RV64ZVE32F-NEXT:  .LBB67_11: # %else17
 ; RV64ZVE32F-NEXT:    andi a1, a1, -128
-; RV64ZVE32F-NEXT:    bnez a1, .LBB67_16
-; RV64ZVE32F-NEXT:  .LBB67_11: # %else20
+; RV64ZVE32F-NEXT:    beqz a1, .LBB67_13
+; RV64ZVE32F-NEXT:  # %bb.12: # %cond.load19
+; RV64ZVE32F-NEXT:    vsetivli zero, 1, e16, mf2, ta, ma
+; RV64ZVE32F-NEXT:    vslidedown.vi v8, v8, 1
+; RV64ZVE32F-NEXT:    vmv.x.s a1, v8
+; RV64ZVE32F-NEXT:    slli a1, a1, 1
+; RV64ZVE32F-NEXT:    add a0, a0, a1
+; RV64ZVE32F-NEXT:    lh a0, 0(a0)
+; RV64ZVE32F-NEXT:    vmv.s.x v8, a0
+; RV64ZVE32F-NEXT:    vsetivli zero, 8, e16, m1, ta, ma
+; RV64ZVE32F-NEXT:    vslideup.vi v9, v8, 7
+; RV64ZVE32F-NEXT:  .LBB67_13: # %else20
+; RV64ZVE32F-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; RV64ZVE32F-NEXT:    vmv1r.v v8, v9
 ; RV64ZVE32F-NEXT:    ret
-; RV64ZVE32F-NEXT:  .LBB67_12: # %cond.load4
+; RV64ZVE32F-NEXT:  .LBB67_14: # %cond.load4
 ; RV64ZVE32F-NEXT:    vmv.x.s a2, v8
 ; RV64ZVE32F-NEXT:    slli a2, a2, 1
 ; RV64ZVE32F-NEXT:    add a2, a0, a2
@@ -7797,7 +7781,7 @@ define <8 x bfloat> @mgather_baseidx_v8bf16(ptr %base, <8 x i16> %idxs, <8 x i1>
 ; RV64ZVE32F-NEXT:    vslideup.vi v9, v11, 2
 ; RV64ZVE32F-NEXT:    andi a2, a1, 8
 ; RV64ZVE32F-NEXT:    beqz a2, .LBB67_6
-; RV64ZVE32F-NEXT:  .LBB67_13: # %cond.load7
+; RV64ZVE32F-NEXT:  .LBB67_15: # %cond.load7
 ; RV64ZVE32F-NEXT:    vsetivli zero, 1, e16, mf2, ta, ma
 ; RV64ZVE32F-NEXT:    vslidedown.vi v8, v8, 1
 ; RV64ZVE32F-NEXT:    vmv.x.s a2, v8
@@ -7809,7 +7793,7 @@ define <8 x bfloat> @mgather_baseidx_v8bf16(ptr %base, <8 x i16> %idxs, <8 x i1>
 ; RV64ZVE32F-NEXT:    vslideup.vi v9, v8, 3
 ; RV64ZVE32F-NEXT:    andi a2, a1, 16
 ; RV64ZVE32F-NEXT:    beqz a2, .LBB67_7
-; RV64ZVE32F-NEXT:  .LBB67_14: # %cond.load10
+; RV64ZVE32F-NEXT:  .LBB67_16: # %cond.load10
 ; RV64ZVE32F-NEXT:    vsetivli zero, 5, e16, m1, tu, ma
 ; RV64ZVE32F-NEXT:    vmv.x.s a2, v10
 ; RV64ZVE32F-NEXT:    slli a2, a2, 1
@@ -7820,28 +7804,6 @@ define <8 x bfloat> @mgather_baseidx_v8bf16(ptr %base, <8 x i16> %idxs, <8 x i1>
 ; RV64ZVE32F-NEXT:    andi a2, a1, 32
 ; RV64ZVE32F-NEXT:    bnez a2, .LBB67_8
 ; RV64ZVE32F-NEXT:    j .LBB67_9
-; RV64ZVE32F-NEXT:  .LBB67_15: # %cond.load16
-; RV64ZVE32F-NEXT:    vmv.x.s a2, v8
-; RV64ZVE32F-NEXT:    slli a2, a2, 1
-; RV64ZVE32F-NEXT:    add a2, a0, a2
-; RV64ZVE32F-NEXT:    lh a2, 0(a2)
-; RV64ZVE32F-NEXT:    vmv.s.x v10, a2
-; RV64ZVE32F-NEXT:    vsetivli zero, 7, e16, m1, tu, ma
-; RV64ZVE32F-NEXT:    vslideup.vi v9, v10, 6
-; RV64ZVE32F-NEXT:    andi a1, a1, -128
-; RV64ZVE32F-NEXT:    beqz a1, .LBB67_11
-; RV64ZVE32F-NEXT:  .LBB67_16: # %cond.load19
-; RV64ZVE32F-NEXT:    vsetivli zero, 1, e16, mf2, ta, ma
-; RV64ZVE32F-NEXT:    vslidedown.vi v8, v8, 1
-; RV64ZVE32F-NEXT:    vmv.x.s a1, v8
-; RV64ZVE32F-NEXT:    slli a1, a1, 1
-; RV64ZVE32F-NEXT:    add a0, a0, a1
-; RV64ZVE32F-NEXT:    lh a0, 0(a0)
-; RV64ZVE32F-NEXT:    vmv.s.x v8, a0
-; RV64ZVE32F-NEXT:    vsetivli zero, 8, e16, m1, ta, ma
-; RV64ZVE32F-NEXT:    vslideup.vi v9, v8, 7
-; RV64ZVE32F-NEXT:    vmv1r.v v8, v9
-; RV64ZVE32F-NEXT:    ret
   %ptrs = getelementptr inbounds bfloat, ptr %base, <8 x i16> %idxs
   %v = call <8 x bfloat> @llvm.masked.gather.v8bf16.v8p0(<8 x ptr> %ptrs, i32 2, <8 x i1> %m, <8 x bfloat> %passthru)
   ret <8 x bfloat> %v
@@ -8135,11 +8097,13 @@ define <4 x half> @mgather_truemask_v4f16(<4 x ptr> %ptrs, <4 x half> %passthru)
 define <4 x half> @mgather_falsemask_v4f16(<4 x ptr> %ptrs, <4 x half> %passthru) {
 ; RV32-LABEL: mgather_falsemask_v4f16:
 ; RV32:       # %bb.0:
+; RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; RV32-NEXT:    vmv1r.v v8, v9
 ; RV32-NEXT:    ret
 ;
 ; RV64V-LABEL: mgather_falsemask_v4f16:
 ; RV64V:       # %bb.0:
+; RV64V-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; RV64V-NEXT:    vmv1r.v v8, v10
 ; RV64V-NEXT:    ret
 ;
@@ -8410,13 +8374,13 @@ define <8 x half> @mgather_baseidx_v8i8_v8f16(ptr %base, <8 x i8> %idxs, <8 x i1
 ; RV64ZVE32F-ZVFH-NEXT:    andi a2, a1, 4
 ; RV64ZVE32F-ZVFH-NEXT:    vsetivli zero, 2, e8, mf4, ta, ma
 ; RV64ZVE32F-ZVFH-NEXT:    vslidedown.vi v8, v8, 2
-; RV64ZVE32F-ZVFH-NEXT:    bnez a2, .LBB74_12
+; RV64ZVE32F-ZVFH-NEXT:    bnez a2, .LBB74_14
 ; RV64ZVE32F-ZVFH-NEXT:  # %bb.5: # %else5
 ; RV64ZVE32F-ZVFH-NEXT:    andi a2, a1, 8
-; RV64ZVE32F-ZVFH-NEXT:    bnez a2, .LBB74_13
+; RV64ZVE32F-ZVFH-NEXT:    bnez a2, .LBB74_15
 ; RV64ZVE32F-ZVFH-NEXT:  .LBB74_6: # %else8
 ; RV64ZVE32F-ZVFH-NEXT:    andi a2, a1, 16
-; RV64ZVE32F-ZVFH-NEXT:    bnez a2, .LBB74_14
+; RV64ZVE32F-ZVFH-NEXT:    bnez a2, .LBB74_16
 ; RV64ZVE32F-ZVFH-NEXT:  .LBB74_7: # %else11
 ; RV64ZVE32F-ZVFH-NEXT:    andi a2, a1, 32
 ; RV64ZVE32F-ZVFH-NEXT:    beqz a2, .LBB74_9
@@ -8435,14 +8399,35 @@ define <8 x half> @mgather_baseidx_v8i8_v8f16(ptr %base, <8 x i8> %idxs, <8 x i1
 ; RV64ZVE32F-ZVFH-NEXT:    andi a2, a1, 64
 ; RV64ZVE32F-ZVFH-NEXT:    vsetivli zero, 2, e8, mf4, ta, ma
 ; RV64ZVE32F-ZVFH-NEXT:    vslidedown.vi v8, v10, 2
-; RV64ZVE32F-ZVFH-NEXT:    bnez a2, .LBB74_15
-; RV64ZVE32F-ZVFH-NEXT:  # %bb.10: # %else17
+; RV64ZVE32F-ZVFH-NEXT:    beqz a2, .LBB74_11
+; RV64ZVE32F-ZVFH-NEXT:  # %bb.10: # %cond.load16
+; RV64ZVE32F-ZVFH-NEXT:    vmv.x.s a2, v8
+; RV64ZVE32F-ZVFH-NEXT:    slli a2, a2, 1
+; RV64ZVE32F-ZVFH-NEXT:    add a2, a0, a2
+; RV64ZVE32F-ZVFH-NEXT:    flh fa5, 0(a2)
+; RV64ZVE32F-ZVFH-NEXT:    vsetvli zero, zero, e16, mf2, ta, ma
+; RV64ZVE32F-ZVFH-NEXT:    vfmv.s.f v10, fa5
+; RV64ZVE32F-ZVFH-NEXT:    vsetivli zero, 7, e16, m1, tu, ma
+; RV64ZVE32F-ZVFH-NEXT:    vslideup.vi v9, v10, 6
+; RV64ZVE32F-ZVFH-NEXT:  .LBB74_11: # %else17
 ; RV64ZVE32F-ZVFH-NEXT:    andi a1, a1, -128
-; RV64ZVE32F-ZVFH-NEXT:    bnez a1, .LBB74_16
-; RV64ZVE32F-ZVFH-NEXT:  .LBB74_11: # %else20
+; RV64ZVE32F-ZVFH-NEXT:    beqz a1, .LBB74_13
+; RV64ZVE32F-ZVFH-NEXT:  # %bb.12: # %cond.load19
+; RV64ZVE32F-ZVFH-NEXT:    vsetivli zero, 1, e8, mf4, ta, ma
+; RV64ZVE32F-ZVFH-NEXT:    vslidedown.vi v8, v8, 1
+; RV64ZVE32F-ZVFH-NEXT:    vmv.x.s a1, v8
+; RV64ZVE32F-ZVFH-NEXT:    slli a1, a1, 1
+; RV64ZVE32F-ZVFH-NEXT:    add a0, a0, a1
+; RV64ZVE32F-ZVFH-NEXT:    flh fa5, 0(a0)
+; RV64ZVE32F-ZVFH-NEXT:    vsetvli zero, zero, e16, mf2, ta, ma
+; RV64ZVE32F-ZVFH-NEXT:    vfmv.s.f v8, fa5
+; RV64ZVE32F-ZVFH-NEXT:    vsetivli zero, 8, e16, m1, ta, ma
+; RV64ZVE32F-ZVFH-NEXT:    vslideup.vi v9, v8, 7
+; RV64ZVE32F-ZVFH-NEXT:  .LBB74_13: # %else20
+; RV64ZVE32F-ZVFH-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; RV64ZVE32F-ZVFH-NEXT:    vmv1r.v v8, v9
 ; RV64ZVE32F-ZVFH-NEXT:    ret
-; RV64ZVE32F-ZVFH-NEXT:  .LBB74_12: # %cond.load4
+; RV64ZVE32F-ZVFH-NEXT:  .LBB74_14: # %cond.load4
 ; RV64ZVE32F-ZVFH-NEXT:    vmv.x.s a2, v8
 ; RV64ZVE32F-ZVFH-NEXT:    slli a2, a2, 1
 ; RV64ZVE32F-ZVFH-NEXT:    add a2, a0, a2
@@ -8453,7 +8438,7 @@ define <8 x half> @mgather_baseidx_v8i8_v8f16(ptr %base, <8 x i8> %idxs, <8 x i1
 ; RV64ZVE32F-ZVFH-NEXT:    vslideup.vi v9, v11, 2
 ; RV64ZVE32F-ZVFH-NEXT:    andi a2, a1, 8
 ; RV64ZVE32F-ZVFH-NEXT:    beqz a2, .LBB74_6
-; RV64ZVE32F-ZVFH-NEXT:  .LBB74_13: # %cond.load7
+; RV64ZVE32F-ZVFH-NEXT:  .LBB74_15: # %cond.load7
 ; RV64ZVE32F-ZVFH-NEXT:    vsetivli zero, 1, e8, mf4, ta, ma
 ; RV64ZVE32F-ZVFH-NEXT:    vslidedown.vi v8, v8, 1
 ; RV64ZVE32F-ZVFH-NEXT:    vmv.x.s a2, v8
@@ -8466,7 +8451,7 @@ define <8 x half> @mgather_baseidx_v8i8_v8f16(ptr %base, <8 x i8> %idxs, <8 x i1
 ; RV64ZVE32F-ZVFH-NEXT:    vslideup.vi v9, v8, 3
 ; RV64ZVE32F-ZVFH-NEXT:    andi a2, a1, 16
 ; RV64ZVE32F-ZVFH-NEXT:    beqz a2, .LBB74_7
-; RV64ZVE32F-ZVFH-NEXT:  .LBB74_14: # %cond.load10
+; RV64ZVE32F-ZVFH-NEXT:  .LBB74_16: # %cond.load10
 ; RV64ZVE32F-ZVFH-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; RV64ZVE32F-ZVFH-NEXT:    vmv.x.s a2, v10
 ; RV64ZVE32F-ZVFH-NEXT:    slli a2, a2, 1
@@ -8479,30 +8464,6 @@ define <8 x half> @mgather_baseidx_v8i8_v8f16(ptr %base, <8 x i8> %idxs, <8 x i1
 ; RV64ZVE32F-ZVFH-NEXT:    andi a2, a1, 32
 ; RV64ZVE32F-ZVFH-NEXT:    bnez a2, .LBB74_8
 ; RV64ZVE32F-ZVFH-NEXT:    j .LBB74_9
-; RV64ZVE32F-ZVFH-NEXT:  .LBB74_15: # %cond.load16
-; RV64ZVE32F-ZVFH-NEXT:    vmv.x.s a2, v8
-; RV64ZVE32F-ZVFH-NEXT:    slli a2, a2, 1
-; RV64ZVE32F-ZVFH-NEXT:    add a2, a0, a2
-; RV64ZVE32F-ZVFH-NEXT:    flh fa5, 0(a2)
-; RV64ZVE32F-ZVFH-NEXT:    vsetvli zero, zero, e16, mf2, ta, ma
-; RV64ZVE32F-ZVFH-NEXT:    vfmv.s.f v10, fa5
-; RV64ZVE32F-ZVFH-NEXT:    vsetivli zero, 7, e16, m1, tu, ma
-; RV64ZVE32F-ZVFH-NEXT:    vslideup.vi v9, v10, 6
-; RV64ZVE32F-ZVFH-NEXT:    andi a1, a1, -128
-; RV64ZVE32F-ZVFH-NEXT:    beqz a1, .LBB74_11
-; RV64ZVE32F-ZVFH-NEXT:  .LBB74_16: # %cond.load19
-; RV64ZVE32F-ZVFH-NEXT:    vsetivli zero, 1, e8, mf4, ta, ma
-; RV64ZVE32F-ZVFH-NEXT:    vslidedown.vi v8, v8, 1
-; RV64ZVE32F-ZVFH-NEXT:    vmv.x.s a1, v8
-; RV64ZVE32F-ZVFH-NEXT:    slli a1, a1, 1
-; RV64ZVE32F-ZVFH-NEXT:    add a0, a0, a1
-; RV64ZVE32F-ZVFH-NEXT:    flh fa5, 0(a0)
-; RV64ZVE32F-ZVFH-NEXT:    vsetvli zero, zero, e16, mf2, ta, ma
-; RV64ZVE32F-ZVFH-NEXT:    vfmv.s.f v8, fa5
-; RV64ZVE32F-ZVFH-NEXT:    vsetivli zero, 8, e16, m1, ta, ma
-; RV64ZVE32F-ZVFH-NEXT:    vslideup.vi v9, v8, 7
-; RV64ZVE32F-ZVFH-NEXT:    vmv1r.v v8, v9
-; RV64ZVE32F-ZVFH-NEXT:    ret
 ;
 ; RV64ZVE32F-ZVFHMIN-LABEL: mgather_baseidx_v8i8_v8f16:
 ; RV64ZVE32F-ZVFHMIN:       # %bb.0:
@@ -8537,13 +8498,13 @@ define <8 x half> @mgather_baseidx_v8i8_v8f16(ptr %base, <8 x i8> %idxs, <8 x i1
 ; RV64ZVE32F-ZVFHMIN-NEXT:    andi a2, a1, 4
 ; RV64ZVE32F-ZVFHMIN-NEXT:    vsetivli zero, 2, e8, mf4, ta, ma
 ; RV64ZVE32F-ZVFHMIN-NEXT:    vslidedown.vi v8, v8, 2
-; RV64ZVE32F-ZVFHMIN-NEXT:    bnez a2, .LBB74_12
+; RV64ZVE32F-ZVFHMIN-NEXT:    bnez a2, .LBB74_14
 ; RV64ZVE32F-ZVFHMIN-NEXT:  # %bb.5: # %else5
 ; RV64ZVE32F-ZVFHMIN-NEXT:    andi a2, a1, 8
-; RV64ZVE32F-ZVFHMIN-NEXT:    bnez a2, .LBB74_13
+; RV64ZVE32F-ZVFHMIN-NEXT:    bnez a2, .LBB74_15
 ; RV64ZVE32F-ZVFHMIN-NEXT:  .LBB74_6: # %else8
 ; RV64ZVE32F-ZVFHMIN-NEXT:    andi a2, a1, 16
-; RV64ZVE32F-ZVFHMIN-NEXT:    bnez a2, .LBB74_14
+; RV64ZVE32F-ZVFHMIN-NEXT:    bnez a2, .LBB74_16
 ; RV64ZVE32F-ZVFHMIN-NEXT:  .LBB74_7: # %else11
 ; RV64ZVE32F-ZVFHMIN-NEXT:    andi a2, a1, 32
 ; RV64ZVE32F-ZVFHMIN-NEXT:    beqz a2, .LBB74_9
@@ -8562,14 +8523,35 @@ define <8 x half> @mgather_baseidx_v8i8_v8f16(ptr %base, <8 x i8> %idxs, <8 x i1
 ; RV64ZVE32F-ZVFHMIN-NEXT:    andi a2, a1, 64
 ; RV64ZVE32F-ZVFHMIN-NEXT:    vsetivli zero, 2, e8, mf4, ta, ma
 ; RV64ZVE32F-ZVFHMIN-NEXT:    vslidedown.vi v8, v10, 2
-; RV64ZVE32F-ZVFHMIN-NEXT:    bnez a2, .LBB74_15
-; RV64ZVE32F-ZVFHMIN-NEXT:  # %bb.10: # %else17
+; RV64ZVE32F-ZVFHMIN-NEXT:    beqz a2, .LBB74_11
+; RV64ZVE32F-ZVFHMIN-NEXT:  # %bb.10: # %cond.load16
+; RV64ZVE32F-ZVFHMIN-NEXT:    vmv.x.s a2, v8
+; RV64ZVE32F-ZVFHMIN-NEXT:    slli a2, a2, 1
+; RV64ZVE32F-ZVFHMIN-NEXT:    add a2, a0, a2
+; RV64ZVE32F-ZVFHMIN-NEXT:    lh a2, 0(a2)
+; RV64ZVE32F-ZVFHMIN-NEXT:    vsetvli zero, zero, e16, mf2, ta, ma
+; RV64ZVE32F-ZVFHMIN-NEXT:    vmv.s.x v10, a2
+; RV64ZVE32F-ZVFHMIN-NEXT:    vsetivli zero, 7, e16, m1, tu, ma
+; RV64ZVE32F-ZVFHMIN-NEXT:    vslideup.vi v9, v10, 6
+; RV64ZVE32F-ZVFHMIN-NEXT:  .LBB74_11: # %else17
 ; RV64ZVE32F-ZVFHMIN-NEXT:    andi a1, a1, -128
-; RV64ZVE32F-ZVFHMIN-NEXT:    bnez a1, .LBB74_16
-; RV64ZVE32F-ZVFHMIN-NEXT:  .LBB74_11: # %else20
+; RV64ZVE32F-ZVFHMIN-NEXT:    beqz a1, .LBB74_13
+; RV64ZVE32F-ZVFHMIN-NEXT:  # %bb.12: # %cond.load19
+; RV64ZVE32F-ZVFHMIN-NEXT:    vsetivli zero, 1, e8, mf4, ta, ma
+; RV64ZVE32F-ZVFHMIN-NEXT:    vslidedown.vi v8, v8, 1
+; RV64ZVE32F-ZVFHMIN-NEXT:    vmv.x.s a1, v8
+; RV64ZVE32F-ZVFHMIN-NEXT:    slli a1, a1, 1
+; RV64ZVE32F-ZVFHMIN-NEXT:    add a0, a0, a1
+; RV64ZVE32F-ZVFHMIN-NEXT:    lh a0, 0(a0)
+; RV64ZVE32F-ZVFHMIN-NEXT:    vsetvli zero, zero, e16, mf2, ta, ma
+; RV64ZVE32F-ZVFHMIN-NEXT:    vmv.s.x v8, a0
+; RV64ZVE32F-ZVFHMIN-NEXT:    vsetivli zero, 8, e16, m1, ta, ma
+; RV64ZVE32F-ZVFHMIN-NEXT:    vslideup.vi v9, v8, 7
+; RV64ZVE32F-ZVFHMIN-NEXT:  .LBB74_13: # %else20
+; RV64ZVE32F-ZVFHMIN-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; RV64ZVE32F-ZVFHMIN-NEXT:    vmv1r.v v8, v9
 ; RV64ZVE32F-ZVFHMIN-NEXT:    ret
-; RV64ZVE32F-ZVFHMIN-NEXT:  .LBB74_12: # %cond.load4
+; RV64ZVE32F-ZVFHMIN-NEXT:  .LBB74_14: # %cond.load4
 ; RV64ZVE32F-ZVFHMIN-NEXT:    vmv.x.s a2, v8
 ; RV64ZVE32F-ZVFHMIN-NEXT:    slli a2, a2, 1
 ; RV64ZVE32F-ZVFHMIN-NEXT:    add a2, a0, a2
@@ -8580,7 +8562,7 @@ define <8 x half> @mgather_baseidx_v8i8_v8f16(ptr %base, <8 x i8> %idxs, <8 x i1
 ; RV64ZVE32F-ZVFHMIN-NEXT:    vslideup.vi v9, v11, 2
 ; RV64ZVE32F-ZVFHMIN-NEXT:    andi a2, a1, 8
 ; RV64ZVE32F-ZVFHMIN-NEXT:    beqz a2, .LBB74_6
-; RV64ZVE32F-ZVFHMIN-NEXT:  .LBB74_13: # %cond.load7
+; RV64ZVE32F-ZVFHMIN-NEXT:  .LBB74_15: # %cond.load7
 ; RV64ZVE32F-ZVFHMIN-NEXT:    vsetivli zero, 1, e8, mf4, ta, ma
 ; RV64ZVE32F-ZVFHMIN-NEXT:    vslidedown.vi v8, v8, 1
 ; RV64ZVE32F-ZVFHMIN-NEXT:    vmv.x.s a2, v8
@@ -8593,7 +8575,7 @@ define <8 x half> @mgather_baseidx_v8i8_v8f16(ptr %base, <8 x i8> %idxs, <8 x i1
 ; RV64ZVE32F-ZVFHMIN-NEXT:    vslideup.vi v9, v8, 3
 ; RV64ZVE32F-ZVFHMIN-NEXT:    andi a2, a1, 16
 ; RV64ZVE32F-ZVFHMIN-NEXT:    beqz a2, .LBB74_7
-; RV64ZVE32F-ZVFHMIN-NEXT:  .LBB74_14: # %cond.load10
+; RV64ZVE32F-ZVFHMIN-NEXT:  .LBB74_16: # %cond.load10
 ; RV64ZVE32F-ZVFHMIN-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; RV64ZVE32F-ZVFHMIN-NEXT:    vmv.x.s a2, v10
 ; RV64ZVE32F-ZVFHMIN-NEXT:    slli a2, a2, 1
@@ -8606,30 +8588,6 @@ define <8 x half> @mgather_baseidx_v8i8_v8f16(ptr %base, <8 x i8> %idxs, <8 x i1
 ; RV64ZVE32F-ZVFHMIN-NEXT:    andi a2, a1, 32
 ; RV64ZVE32F-ZVFHMIN-NEXT:    bnez a2, .LBB74_8
 ; RV64ZVE32F-ZVFHMIN-NEXT:    j .LBB74_9
-; RV64ZVE32F-ZVFHMIN-NEXT:  .LBB74_15: # %cond.load16
-; RV64ZVE32F-ZVFHMIN-NEXT:    vmv.x.s a2, v8
-; RV64ZVE32F-ZVFHMIN-NEXT:    slli a2, a2, 1
-; RV64ZVE32F-ZVFHMIN-NEXT:    add a2, a0, a2
-; RV64ZVE32F-ZVFHMIN-NEXT:    lh a2, 0(a2)
-; RV64ZVE32F-ZVFHMIN-NEXT:    vsetvli zero, zero, e16, mf2, ta, ma
-; RV64ZVE32F-ZVFHMIN-NEXT:    vmv.s.x v10, a2
-; RV64ZVE32F-ZVFHMIN-NEXT:    vsetivli zero, 7, e16, m1, tu, ma
-; RV64ZVE32F-ZVFHMIN-NEXT:    vslideup.vi v9, v10, 6
-; RV64ZVE32F-ZVFHMIN-NEXT:    andi a1, a1, -128
-; RV64ZVE32F-ZVFHMIN-NEXT:    beqz a1, .LBB74_11
-; RV64ZVE32F-ZVFHMIN-NEXT:  .LBB74_16: # %cond.load19
-; RV64ZVE32F-ZVFHMIN-NEXT:    vsetivli zero, 1, e8, mf4, ta, ma
-; RV64ZVE32F-ZVFHMIN-NEXT:    vslidedown.vi v8, v8, 1
-; RV64ZVE32F-ZVFHMIN-NEXT:    vmv.x.s a1, v8
-; RV64ZVE32F-ZVFHMIN-NEXT:    slli a1, a1, 1
-; RV64ZVE32F-ZVFHMIN-NEXT:    add a0, a0, a1
-; RV64ZVE32F-ZVFHMIN-NEXT:    lh a0, 0(a0)
-; RV64ZVE32F-ZVFHMIN-NEXT:    vsetvli zero, zero, e16, mf2, ta, ma
-; RV64ZVE32F-ZVFHMIN-NEXT:    vmv.s.x v8, a0
-; RV64ZVE32F-ZVFHMIN-NEXT:    vsetivli zero, 8, e16, m1, ta, ma
-; RV64ZVE32F-ZVFHMIN-NEXT:    vslideup.vi v9, v8, 7
-; RV64ZVE32F-ZVFHMIN-NEXT:    vmv1r.v v8, v9
-; RV64ZVE32F-ZVFHMIN-NEXT:    ret
   %ptrs = getelementptr inbounds half, ptr %base, <8 x i8> %idxs
   %v = call <8 x half> @llvm.masked.gather.v8f16.v8p0(<8 x ptr> %ptrs, i32 2, <8 x i1> %m, <8 x half> %passthru)
   ret <8 x half> %v
@@ -8689,13 +8647,13 @@ define <8 x half> @mgather_baseidx_sext_v8i8_v8f16(ptr %base, <8 x i8> %idxs, <8
 ; RV64ZVE32F-ZVFH-NEXT:    andi a2, a1, 4
 ; RV64ZVE32F-ZVFH-NEXT:    vsetivli zero, 2, e8, mf4, ta, ma
 ; RV64ZVE32F-ZVFH-NEXT:    vslidedown.vi v8, v8, 2
-; RV64ZVE32F-ZVFH-NEXT:    bnez a2, .LBB75_12
+; RV64ZVE32F-ZVFH-NEXT:    bnez a2, .LBB75_14
 ; RV64ZVE32F-ZVFH-NEXT:  # %bb.5: # %else5
 ; RV64ZVE32F-ZVFH-NEXT:    andi a2, a1, 8
-; RV64ZVE32F-ZVFH-NEXT:    bnez a2, .LBB75_13
+; RV64ZVE32F-ZVFH-NEXT:    bnez a2, .LBB75_15
 ; RV64ZVE32F-ZVFH-NEXT:  .LBB75_6: # %else8
 ; RV64ZVE32F-ZVFH-NEXT:    andi a2, a1, 16
-; RV64ZVE32F-ZVFH-NEXT:    bnez a2, .LBB75_14
+; RV64ZVE32F-ZVFH-NEXT:    bnez a2, .LBB75_16
 ; RV64ZVE32F-ZVFH-NEXT:  .LBB75_7: # %else11
 ; RV64ZVE32F-ZVFH-NEXT:    andi a2, a1, 32
 ; RV64ZVE32F-ZVFH-NEXT:    beqz a2, .LBB75_9
@@ -8714,14 +8672,35 @@ define <8 x half> @mgather_baseidx_sext_v8i8_v8f16(ptr %base, <8 x i8> %idxs, <8
 ; RV64ZVE32F-ZVFH-NEXT:    andi a2, a1, 64
 ; RV64ZVE32F-ZVFH-NEXT:    vsetivli zero, 2, e8, mf4, ta, ma
 ; RV64ZVE32F-ZVFH-NEXT:    vslidedown.vi v8, v10, 2
-; RV64ZVE32F-ZVFH-NEXT:    bnez a2, .LBB75_15
-; RV64ZVE32F-ZVFH-NEXT:  # %bb.10: # %else17
+; RV64ZVE32F-ZVFH-NEXT:    beqz a2, .LBB75_11
+; RV64ZVE32F-ZVFH-NEXT:  # %bb.10: # %cond.load16
+; RV64ZVE32F-ZVFH-NEXT:    vmv.x.s a2, v8
+; RV64ZVE32F-ZVFH-NEXT:    slli a2, a2, 1
+; RV64ZVE32F-ZVFH-NEXT:    add a2, a0, a2
+; RV64ZVE32F-ZVFH-NEXT:    flh fa5, 0(a2)
+; RV64ZVE32F-ZVFH-NEXT:    vsetvli zero, zero, e16, mf2, ta, ma
+; RV64ZVE32F-ZVFH-NEXT:    vfmv.s.f v10, fa5
+; RV64ZVE32F-ZVFH-NEXT:    vsetivli zero, 7, e16, m1, tu, ma
+; RV64ZVE32F-ZVFH-NEXT:    vslideup.vi v9, v10, 6
+; RV64ZVE32F-ZVFH-NEXT:  .LBB75_11: # %else17
 ; RV64ZVE32F-ZVFH-NEXT:    andi a1, a1, -128
-; RV64ZVE32F-ZVFH-NEXT:    bnez a1, .LBB75_16
-; RV64ZVE32F-ZVFH-NEXT:  .LBB75_11: # %else20
+; RV64ZVE32F-ZVFH-NEXT:    beqz a1, .LBB75_13
+; RV64ZVE32F-ZVFH-NEXT:  # %bb.12: # %cond.load19
+; RV64ZVE32F-ZVFH-NEXT:    vsetivli zero, 1, e8, mf4, ta, ma
+; RV64ZVE32F-ZVFH-NEXT:    vslidedown.vi v8, v8, 1
+; RV64ZVE32F-ZVFH-NEXT:    vmv.x.s a1, v8
+; RV64ZVE32F-ZVFH-NEXT:    slli a1, a1, 1
+; RV64ZVE32F-ZVFH-NEXT:    add a0, a0, a1
+; RV64ZVE32F-ZVFH-NEXT:    flh fa5, 0(a0)
+; RV64ZVE32F-ZVFH-NEXT:    vsetvli zero, zero, e16, mf2, ta, ma
+; RV64ZVE32F-ZVFH-NEXT:    vfmv.s.f v8, fa5
+; RV64ZVE32F-ZVFH-NEXT:    vsetivli zero, 8, e16, m1, ta, ma
+; RV64ZVE32F-ZVFH-NEXT:    vslideup.vi v9, v8, 7
+; RV64ZVE32F-ZVFH-NEXT:  .LBB75_13: # %else20
+; RV64ZVE32F-ZVFH-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; RV64ZVE32F-ZVFH-NEXT:    vmv1r.v v8, v9
 ; RV64ZVE32F-ZVFH-NEXT:    ret
-; RV64ZVE32F-ZVFH-NEXT:  .LBB75_12: # %cond.load4
+; RV64ZVE32F-ZVFH-NEXT:  .LBB75_14: # %cond.load4
 ; RV64ZVE32F-ZVFH-NEXT:    vmv.x.s a2, v8
 ; RV64ZVE32F-ZVFH-NEXT:    slli a2, a2, 1
 ; RV64ZVE32F-ZVFH-NEXT:    add a2, a0, a2
@@ -8732,7 +8711,7 @@ define <8 x half> @mgather_baseidx_sext_v8i8_v8f16(ptr %base, <8 x i8> %idxs, <8
 ; RV64ZVE32F-ZVFH-NEXT:    vslideup.vi v9, v11, 2
 ; RV64ZVE32F-ZVFH-NEXT:    andi a2, a1, 8
 ; RV64ZVE32F-ZVFH-NEXT:    beqz a2, .LBB75_6
-; RV64ZVE32F-ZVFH-NEXT:  .LBB75_13: # %cond.load7
+; RV64ZVE32F-ZVFH-NEXT:  .LBB75_15: # %cond.load7
 ; RV64ZVE32F-ZVFH-NEXT:    vsetivli zero, 1, e8, mf4, ta, ma
 ; RV64ZVE32F-ZVFH-NEXT:    vslidedown.vi v8, v8, 1
 ; RV64ZVE32F-ZVFH-NEXT:    vmv.x.s a2, v8
@@ -8745,7 +8724,7 @@ define <8 x half> @mgather_baseidx_sext_v8i8_v8f16(ptr %base, <8 x i8> %idxs, <8
 ; RV64ZVE32F-ZVFH-NEXT:    vslideup.vi v9, v8, 3
 ; RV64ZVE32F-ZVFH-NEXT:    andi a2, a1, 16
 ; RV64ZVE32F-ZVFH-NEXT:    beqz a2, .LBB75_7
-; RV64ZVE32F-ZVFH-NEXT:  .LBB75_14: # %cond.load10
+; RV64ZVE32F-ZVFH-NEXT:  .LBB75_16: # %cond.load10
 ; RV64ZVE32F-ZVFH-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; RV64ZVE32F-ZVFH-NEXT:    vmv.x.s a2, v10
 ; RV64ZVE32F-ZVFH-NEXT:    slli a2, a2, 1
@@ -8758,30 +8737,6 @@ define <8 x half> @mgather_baseidx_sext_v8i8_v8f16(ptr %base, <8 x i8> %idxs, <8
 ; RV64ZVE32F-ZVFH-NEXT:    andi a2, a1, 32
 ; RV64ZVE32F-ZVFH-NEXT:    bnez a2, .LBB75_8
 ; RV64ZVE32F-ZVFH-NEXT:    j .LBB75_9
-; RV64ZVE32F-ZVFH-NEXT:  .LBB75_15: # %cond.load16
-; RV64ZVE32F-ZVFH-NEXT:    vmv.x.s a2, v8
-; RV64ZVE32F-ZVFH-NEXT:    slli a2, a2, 1
-; RV64ZVE32F-ZVFH-NEXT:    add a2, a0, a2
-; RV64ZVE32F-ZVFH-NEXT:    flh fa5, 0(a2)
-; RV64ZVE32F-ZVFH-NEXT:    vsetvli zero, zero, e16, mf2, ta, ma
-; RV64ZVE32F-ZVFH-NEXT:    vfmv.s.f v10, fa5
-; RV64ZVE32F-ZVFH-NEXT:    vsetivli zero, 7, e16, m1, tu, ma
-; RV64ZVE32F-ZVFH-NEXT:    vslideup.vi v9, v10, 6
-; RV64ZVE32F-ZVFH-NEXT:    andi a1, a1, -128
-; RV64ZVE32F-ZVFH-NEXT:    beqz a1, .LBB75_11
-; RV64ZVE32F-ZVFH-NEXT:  .LBB75_16: # %cond.load19
-; RV64ZVE32F-ZVFH-NEXT:    vsetivli zero, 1, e8, mf4, ta, ma
-; RV64ZVE32F-ZVFH-NEXT:    vslidedown.vi v8, v8, 1
-; RV64ZVE32F-ZVFH-NEXT:    vmv.x.s a1, v8
-; RV64ZVE32F-ZVFH-NEXT:    slli a1, a1, 1
-; RV64ZVE32F-ZVFH-NEXT:    add a0, a0, a1
-; RV64ZVE32F-ZVFH-NEXT:    flh fa5, 0(a0)
-; RV64ZVE32F-ZVFH-NEXT:    vsetvli zero, zero, e16, mf2, ta, ma
-; RV64ZVE32F-ZVFH-NEXT:    vfmv.s.f v8, fa5
-; RV64ZVE32F-ZVFH-NEXT:    vsetivli zero, 8, e16, m1, ta, ma
-; RV64ZVE32F-ZVFH-NEXT:    vslideup.vi v9, v8, 7
-; RV64ZVE32F-ZVFH-NEXT:    vmv1r.v v8, v9
-; RV64ZVE32F-ZVFH-NEXT:    ret
 ;
 ; RV64ZVE32F-ZVFHMIN-LABEL: mgather_baseidx_sext_v8i8_v8f16:
 ; RV64ZVE32F-ZVFHMIN:       # %bb.0:
@@ -8816,13 +8771,13 @@ define <8 x half> @mgather_baseidx_sext_v8i8_v8f16(ptr %base, <8 x i8> %idxs, <8
 ; RV64ZVE32F-ZVFHMIN-NEXT:    andi a2, a1, 4
 ; RV64ZVE32F-ZVFHMIN-NEXT:    vsetivli zero, 2, e8, mf4, ta, ma
 ; RV64ZVE32F-ZVFHMIN-NEXT:    vslidedown.vi v8, v8, 2
-; RV64ZVE32F-ZVFHMIN-NEXT:    bnez a2, .LBB75_12
+; RV64ZVE32F-ZVFHMIN-NEXT:    bnez a2, .LBB75_14
 ; RV64ZVE32F-ZVFHMIN-NEXT:  # %bb.5: # %else5
 ; RV64ZVE32F-ZVFHMIN-NEXT:    andi a2, a1, 8
-; RV64ZVE32F-ZVFHMIN-NEXT:    bnez a2, .LBB75_13
+; RV64ZVE32F-ZVFHMIN-NEXT:    bnez a2, .LBB75_15
 ; RV64ZVE32F-ZVFHMIN-NEXT:  .LBB75_6: # %else8
 ; RV64ZVE32F-ZVFHMIN-NEXT:    andi a2, a1, 16
-; RV64ZVE32F-ZVFHMIN-NEXT:    bnez a2, .LBB75_14
+; RV64ZVE32F-ZVFHMIN-NEXT:    bnez a2, .LBB75_16
 ; RV64ZVE32F-ZVFHMIN-NEXT:  .LBB75_7: # %else11
 ; RV64ZVE32F-ZVFHMIN-NEXT:    andi a2, a1, 32
 ; RV64ZVE32F-ZVFHMIN-NEXT:    beqz a2, .LBB75_9
@@ -8841,14 +8796,35 @@ define <8 x half> @mgather_baseidx_sext_v8i8_v8f16(ptr %base, <8 x i8> %idxs, <8
 ; RV64ZVE32F-ZVFHMIN-NEXT:    andi a2, a1, 64
 ; RV64ZVE32F-ZVFHMIN-NEXT:    vsetivli zero, 2, e8, mf4, ta, ma
 ; RV64ZVE32F-ZVFHMIN-NEXT:    vslidedown.vi v8, v10, 2
-; RV64ZVE32F-ZVFHMIN-NEXT:    bnez a2, .LBB75_15
-; RV64ZVE32F-ZVFHMIN-NEXT:  # %bb.10: # %else17
+; RV64ZVE32F-ZVFHMIN-NEXT:    beqz a2, .LBB75_11
+; RV64ZVE32F-ZVFHMIN-NEXT:  # %bb.10: # %cond.load16
+; RV64ZVE32F-ZVFHMIN-NEXT:    vmv.x.s a2, v8
+; RV64ZVE32F-ZVFHMIN-NEXT:    slli a2, a2, 1
+; RV64ZVE32F-ZVFHMIN-NEXT:    add a2, a0, a2
+; RV64ZVE32F-ZVFHMIN-NEXT:    lh a2, 0(a2)
+; RV64ZVE32F-ZVFHMIN-NEXT:    vsetvli zero, zero, e16, mf2, ta, ma
+; RV64ZVE32F-ZVFHMIN-NEXT:    vmv.s.x v10, a2
+; RV64ZVE32F-ZVFHMIN-NEXT:    vsetivli zero, 7, e16, m1, tu, ma
+; RV64ZVE32F-ZVFHMIN-NEXT:    vslideup.vi v9, v10, 6
+; RV64ZVE32F-ZVFHMIN-NEXT:  .LBB75_11: # %else17
 ; RV64ZVE32F-ZVFHMIN-NEXT:    andi a1, a1, -128
-; RV64ZVE32F-ZVFHMIN-NEXT:    bnez a1, .LBB75_16
-; RV64ZVE32F-ZVFHMIN-NEXT:  .LBB75_11: # %else20
+; RV64ZVE32F-ZVFHMIN-NEXT:    beqz a1, .LBB75_13
+; RV64ZVE32F-ZVFHMIN-NEXT:  # %bb.12: # %cond.load19
+; RV64ZVE32F-ZVFHMIN-NEXT:    vsetivli zero, 1, e8, mf4, ta, ma
+; RV64ZVE32F-ZVFHMIN-NEXT:    vslidedown.vi v8, v8, 1
+; RV64ZVE32F-ZVFHMIN-NEXT:    vmv.x.s a1, v8
+; RV64ZVE32F-ZVFHMIN-NEXT:    slli a1, a1, 1
+; RV64ZVE32F-ZVFHMIN-NEXT:    add a0, a0, a1
+; RV64ZVE32F-ZVFHMIN-NEXT:    lh a0, 0(a0)
+; RV64ZVE32F-ZVFHMIN-NEXT:    vsetvli zero, zero, e16, mf2, ta, ma
+; RV64ZVE32F-ZVFHMIN-NEXT:    vmv.s.x v8, a0
+; RV64ZVE32F-ZVFHMIN-NEXT:    vsetivli zero, 8, e16, m1, ta, ma
+; RV64ZVE32F-ZVFHMIN-NEXT:    vslideup.vi v9, v8, 7
+; RV64ZVE32F-ZVFHMIN-NEXT:  .LBB75_13: # %else20
+; RV64ZVE32F-ZVFHMIN-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; RV64ZVE32F-ZVFHMIN-NEXT:    vmv1r.v v8, v9
 ; RV64ZVE32F-ZVFHMIN-NEXT:    ret
-; RV64ZVE32F-ZVFHMIN-NEXT:  .LBB75_12: # %cond.load4
+; RV64ZVE32F-ZVFHMIN-NEXT:  .LBB75_14: # %cond.load4
 ; RV64ZVE32F-ZVFHMIN-NEXT:    vmv.x.s a2, v8
 ; RV64ZVE32F-ZVFHMIN-NEXT:    slli a2, a2, 1
 ; RV64ZVE32F-ZVFHMIN-NEXT:    add a2, a0, a2
@@ -8859,7 +8835,7 @@ define <8 x half> @mgather_baseidx_sext_v8i8_v8f16(ptr %base, <8 x i8> %idxs, <8
 ; RV64ZVE32F-ZVFHMIN-NEXT:    vslideup.vi v9, v11, 2
 ; RV64ZVE32F-ZVFHMIN-NEXT:    andi a2, a1, 8
 ; RV64ZVE32F-ZVFHMIN-NEXT:    beqz a2, .LBB75_6
-; RV64ZVE32F-ZVFHMIN-NEXT:  .LBB75_13: # %cond.load7
+; RV64ZVE32F-ZVFHMIN-NEXT:  .LBB75_15: # %cond.load7
 ; RV64ZVE32F-ZVFHMIN-NEXT:    vsetivli zero, 1, e8, mf4, ta, ma
 ; RV64ZVE32F-ZVFHMIN-NEXT:    vslidedown.vi v8, v8, 1
 ; RV64ZVE32F-ZVFHMIN-NEXT:    vmv.x.s a2, v8
@@ -8872,7 +8848,7 @@ define <8 x half> @mgather_baseidx_sext_v8i8_v8f16(ptr %base, <8 x i8> %idxs, <8
 ; RV64ZVE32F-ZVFHMIN-NEXT:    vslideup.vi v9, v8, 3
 ; RV64ZVE32F-ZVFHMIN-NEXT:    andi a2, a1, 16
 ; RV64ZVE32F-ZVFHMIN-NEXT:    beqz a2, .LBB75_7
-; RV64ZVE32F-ZVFHMIN-NEXT:  .LBB75_14: # %cond.load10
+; RV64ZVE32F-ZVFHMIN-NEXT:  .LBB75_16: # %cond.load10
 ; RV64ZVE32F-ZVFHMIN-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; RV64ZVE32F-ZVFHMIN-NEXT:    vmv.x.s a2, v10
 ; RV64ZVE32F-ZVFHMIN-NEXT:    slli a2, a2, 1
@@ -8885,30 +8861,6 @@ define <8 x half> @mgather_baseidx_sext_v8i8_v8f16(ptr %base, <8 x i8> %idxs, <8
 ; RV64ZVE32F-ZVFHMIN-NEXT:    andi a2, a1, 32
 ; RV64ZVE32F-ZVFHMIN-NEXT:    bnez a2, .LBB75_8
 ; RV64ZVE32F-ZVFHMIN-NEXT:    j .LBB75_9
-; RV64ZVE32F-ZVFHMIN-NEXT:  .LBB75_15: # %cond.load16
-; RV64ZVE32F-ZVFHMIN-NEXT:    vmv.x.s a2, v8
-; RV64ZVE32F-ZVFHMIN-NEXT:    slli a2, a2, 1
-; RV64ZVE32F-ZVFHMIN-NEXT:    add a2, a0, a2
-; RV64ZVE32F-ZVFHMIN-NEXT:    lh a2, 0(a2)
-; RV64ZVE32F-ZVFHMIN-NEXT:    vsetvli zero, zero, e16, mf2, ta, ma
-; RV64ZVE32F-ZVFHMIN-NEXT:    vmv.s.x v10, a2
-; RV64ZVE32F-ZVFHMIN-NEXT:    vsetivli zero, 7, e16, m1, tu, ma
-; RV64ZVE32F-ZVFHMIN-NEXT:    vslideup.vi v9, v10, 6
-; RV64ZVE32F-ZVFHMIN-NEXT:    andi a1, a1, -128
-; RV64ZVE32F-ZVFHMIN-NEXT:    beqz a1, .LBB75_11
-; RV64ZVE32F-ZVFHMIN-NEXT:  .LBB75_16: # %cond.load19
-; RV64ZVE32F-ZVFHMIN-NEXT:    vsetivli zero, 1, e8, mf4, ta, ma
-; RV64ZVE32F-ZVFHMIN-NEXT:    vslidedown.vi v8, v8, 1
-; RV64ZVE32F-ZVFHMIN-NEXT:    vmv.x.s a1, v8
-; RV64ZVE32F-ZVFHMIN-NEXT:    slli a1, a1, 1
-; RV64ZVE32F-ZVFHMIN-NEXT:    add a0, a0, a1
-; RV64ZVE32F-ZVFHMIN-NEXT:    lh a0, 0(a0)
-; RV64ZVE32F-ZVFHMIN-NEXT:    vsetvli zero, zero, e16, mf2, ta, ma
-; RV64ZVE32F-ZVFHMIN-NEXT:    vmv.s.x v8, a0
-; RV64ZVE32F-ZVFHMIN-NEXT:    vsetivli zero, 8, e16, m1, ta, ma
-; RV64ZVE32F-ZVFHMIN-NEXT:    vslideup.vi v9, v8, 7
-; RV64ZVE32F-ZVFHMIN-NEXT:    vmv1r.v v8, v9
-; RV64ZVE32F-ZVFHMIN-NEXT:    ret
   %eidxs = sext <8 x i8> %idxs to <8 x i16>
   %ptrs = getelementptr inbounds half, ptr %base, <8 x i16> %eidxs
   %v = call <8 x half> @llvm.masked.gather.v8f16.v8p0(<8 x ptr> %ptrs, i32 2, <8 x i1> %m, <8 x half> %passthru)
@@ -8969,13 +8921,13 @@ define <8 x half> @mgather_baseidx_zext_v8i8_v8f16(ptr %base, <8 x i8> %idxs, <8
 ; RV64ZVE32F-ZVFH-NEXT:    andi a2, a1, 4
 ; RV64ZVE32F-ZVFH-NEXT:    vsetivli zero, 2, e8, mf4, ta, ma
 ; RV64ZVE32F-ZVFH-NEXT:    vslidedown.vi v8, v8, 2
-; RV64ZVE32F-ZVFH-NEXT:    bnez a2, .LBB76_12
+; RV64ZVE32F-ZVFH-NEXT:    bnez a2, .LBB76_14
 ; RV64ZVE32F-ZVFH-NEXT:  # %bb.5: # %else5
 ; RV64ZVE32F-ZVFH-NEXT:    andi a2, a1, 8
-; RV64ZVE32F-ZVFH-NEXT:    bnez a2, .LBB76_13
+; RV64ZVE32F-ZVFH-NEXT:    bnez a2, .LBB76_15
 ; RV64ZVE32F-ZVFH-NEXT:  .LBB76_6: # %else8
 ; RV64ZVE32F-ZVFH-NEXT:    andi a2, a1, 16
-; RV64ZVE32F-ZVFH-NEXT:    bnez a2, .LBB76_14
+; RV64ZVE32F-ZVFH-NEXT:    bnez a2, .LBB76_16
 ; RV64ZVE32F-ZVFH-NEXT:  .LBB76_7: # %else11
 ; RV64ZVE32F-ZVFH-NEXT:    andi a2, a1, 32
 ; RV64ZVE32F-ZVFH-NEXT:    beqz a2, .LBB76_9
@@ -8995,14 +8947,37 @@ define <8 x half> @mgather_baseidx_zext_v8i8_v8f16(ptr %base, <8 x i8> %idxs, <8
 ; RV64ZVE32F-ZVFH-NEXT:    andi a2, a1, 64
 ; RV64ZVE32F-ZVFH-NEXT:    vsetivli zero, 2, e8, mf4, ta, ma
 ; RV64ZVE32F-ZVFH-NEXT:    vslidedown.vi v8, v10, 2
-; RV64ZVE32F-ZVFH-NEXT:    bnez a2, .LBB76_15
-; RV64ZVE32F-ZVFH-NEXT:  # %bb.10: # %else17
+; RV64ZVE32F-ZVFH-NEXT:    beqz a2, .LBB76_11
+; RV64ZVE32F-ZVFH-NEXT:  # %bb.10: # %cond.load16
+; RV64ZVE32F-ZVFH-NEXT:    vmv.x.s a2, v8
+; RV64ZVE32F-ZVFH-NEXT:    andi a2, a2, 255
+; RV64ZVE32F-ZVFH-NEXT:    slli a2, a2, 1
+; RV64ZVE32F-ZVFH-NEXT:    add a2, a0, a2
+; RV64ZVE32F-ZVFH-NEXT:    flh fa5, 0(a2)
+; RV64ZVE32F-ZVFH-NEXT:    vsetvli zero, zero, e16, mf2, ta, ma
+; RV64ZVE32F-ZVFH-NEXT:    vfmv.s.f v10, fa5
+; RV64ZVE32F-ZVFH-NEXT:    vsetivli zero, 7, e16, m1, tu, ma
+; RV64ZVE32F-ZVFH-NEXT:    vslideup.vi v9, v10, 6
+; RV64ZVE32F-ZVFH-NEXT:  .LBB76_11: # %else17
 ; RV64ZVE32F-ZVFH-NEXT:    andi a1, a1, -128
-; RV64ZVE32F-ZVFH-NEXT:    bnez a1, .LBB76_16
-; RV64ZVE32F-ZVFH-NEXT:  .LBB76_11: # %else20
+; RV64ZVE32F-ZVFH-NEXT:    beqz a1, .LBB76_13
+; RV64ZVE32F-ZVFH-NEXT:  # %bb.12: # %cond.load19
+; RV64ZVE32F-ZVFH-NEXT:    vsetivli zero, 1, e8, mf4, ta, ma
+; RV64ZVE32F-ZVFH-NEXT:    vslidedown.vi v8, v8, 1
+; RV64ZVE32F-ZVFH-NEXT:    vmv.x.s a1, v8
+; RV64ZVE32F-ZVFH-NEXT:    andi a1, a1, 255
+; RV64ZVE32F-ZVFH-NEXT:    slli a1, a1, 1
+; RV64ZVE32F-ZVFH-NEXT:    add a0, a0, a1
+; RV64ZVE32F-ZVFH-NEXT:    flh fa5, 0(a0)
+; RV64ZVE32F-ZVFH-NEXT:    vsetvli zero, zero, e16, mf2, ta, ma
+; RV64ZVE32F-ZVFH-NEXT:    vfmv.s.f v8, fa5
+; RV64ZVE32F-ZVFH-NEXT:    vsetivli zero, 8, e16, m1, ta, ma
+; RV64ZVE32F-ZVFH-NEXT:    vslideup.vi v9, v8, 7
+; RV64ZVE32F-ZVFH-NEXT:  .LBB76_13: # %else20
+; RV64ZVE32F-ZVFH-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; RV64ZVE32F-ZVFH-NEXT:    vmv1r.v v8, v9
 ; RV64ZVE32F-ZVFH-NEXT:    ret
-; RV64ZVE32F-ZVFH-NEXT:  .LBB76_12: # %cond.load4
+; RV64ZVE32F-ZVFH-NEXT:  .LBB76_14: # %cond.load4
 ; RV64ZVE32F-ZVFH-NEXT:    vmv.x.s a2, v8
 ; RV64ZVE32F-ZVFH-NEXT:    andi a2, a2, 255
 ; RV64ZVE32F-ZVFH-NEXT:    slli a2, a2, 1
@@ -9014,7 +8989,7 @@ define <8 x half> @mgather_baseidx_zext_v8i8_v8f16(ptr %base, <8 x i8> %idxs, <8
 ; RV64ZVE32F-ZVFH-NEXT:    vslideup.vi v9, v11, 2
 ; RV64ZVE32F-ZVFH-NEXT:    andi a2, a1, 8
 ; RV64ZVE32F-ZVFH-NEXT:    beqz a2, .LBB76_6
-; RV64ZVE32F-ZVFH-NEXT:  .LBB76_13: # %cond.load7
+; RV64ZVE32F-ZVFH-NEXT:  .LBB76_15: # %cond.load7
 ; RV64ZVE32F-ZVFH-NEXT:    vsetivli zero, 1, e8, mf4, ta, ma
 ; RV64ZVE32F-ZVFH-NEXT:    vslidedown.vi v8, v8, 1
 ; RV64ZVE32F-ZVFH-NEXT:    vmv.x.s a2, v8
@@ -9028,7 +9003,7 @@ define <8 x half> @mgather_baseidx_zext_v8i8_v8f16(ptr %base, <8 x i8> %idxs, <8
 ; RV64ZVE32F-ZVFH-NEXT:    vslideup.vi v9, v8, 3
 ; RV64ZVE32F-ZVFH-NEXT:    andi a2, a1, 16
 ; RV64ZVE32F-ZVFH-NEXT:    beqz a2, .LBB76_7
-; RV64ZVE32F-ZVFH-NEXT:  .LBB76_14: # %cond.load10
+; RV64ZVE32F-ZVFH-NEXT:  .LBB76_16: # %cond.load10
 ; RV64ZVE32F-ZVFH-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; RV64ZVE32F-ZVFH-NEXT:    vmv.x.s a2, v10
 ; RV64ZVE32F-ZVFH-NEXT:    andi a2, a2, 255
@@ -9042,32 +9017,6 @@ define <8 x half> @mgather_baseidx_zext_v8i8_v8f16(ptr %base, <8 x i8> %idxs, <8
 ; RV64ZVE32F-ZVFH-NEXT:    andi a2, a1, 32
 ; RV64ZVE32F-ZVFH-NEXT:    bnez a2, .LBB76_8
 ; RV64ZVE32F-ZVFH-NEXT:    j .LBB76_9
-; RV64ZVE32F-ZVFH-NEXT:  .LBB76_15: # %cond.load16
-; RV64ZVE32F-ZVFH-NEXT:    vmv.x.s a2, v8
-; RV64ZVE32F-ZVFH-NEXT:    andi a2, a2, 255
-; RV64ZVE32F-ZVFH-NEXT:    slli a2, a2, 1
-; RV64ZVE32F-ZVFH-NEXT:    add a2, a0, a2
-; RV64ZVE32F-ZVFH-NEXT:    flh fa5, 0(a2)
-; RV64ZVE32F-ZVFH-NEXT:    vsetvli zero, zero, e16, mf2, ta, ma
-; RV64ZVE32F-ZVFH-NEXT:    vfmv.s.f v10, fa5
-; RV64ZVE32F-ZVFH-NEXT:    vsetivli zero, 7, e16, m1, tu, ma
-; RV64ZVE32F-ZVFH-NEXT:    vslideup.vi v9, v10, 6
-; RV64ZVE32F-ZVFH-NEXT:    andi a1, a1, -128
-; RV64ZVE32F-ZVFH-NEXT:    beqz a1, .LBB76_11
-; RV64ZVE32F-ZVFH-NEXT:  .LBB76_16: # %cond.load19
-; RV64ZVE32F-ZVFH-NEXT:    vsetivli zero, 1, e8, mf4, ta, ma
-; RV64ZVE32F-ZVFH-NEXT:    vslidedown.vi v8, v8, 1
-; RV64ZVE32F-ZVFH-NEXT:    vmv.x.s a1, v8
-; RV64ZVE32F-ZVFH-NEXT:    andi a1, a1, 255
-; RV64ZVE32F-ZVFH-NEXT:    slli a1, a1, 1
-; RV64ZVE32F-ZVFH-NEXT:    add a0, a0, a1
-; RV64ZVE32F-ZVFH-NEXT:    flh fa5, 0(a0)
-; RV64ZVE32F-ZVFH-NEXT:    vsetvli zero, zero, e16, mf2, ta, ma
-; RV64ZVE32F-ZVFH-NEXT:    vfmv.s.f v8, fa5
-; RV64ZVE32F-ZVFH-NEXT:    vsetivli zero, 8, e16, m1, ta, ma
-; RV64ZVE32F-ZVFH-NEXT:    vslideup.vi v9, v8, 7
-; RV64ZVE32F-ZVFH-NEXT:    vmv1r.v v8, v9
-; RV64ZVE32F-ZVFH-NEXT:    ret
 ;
 ; RV64ZVE32F-ZVFHMIN-LABEL: mgather_baseidx_zext_v8i8_v8f16:
 ; RV64ZVE32F-ZVFHMIN:       # %bb.0:
@@ -9104,13 +9053,13 @@ define <8 x half> @mgather_baseidx_zext_v8i8_v8f16(ptr %base, <8 x i8> %idxs, <8
 ; RV64ZVE32F-ZVFHMIN-NEXT:    andi a2, a1, 4
 ; RV64ZVE32F-ZVFHMIN-NEXT:    vsetivli zero, 2, e8, mf4, ta, ma
 ; RV64ZVE32F-ZVFHMIN-NEXT:    vslidedown.vi v8, v8, 2
-; RV64ZVE32F-ZVFHMIN-NEXT:    bnez a2, .LBB76_12
+; RV64ZVE32F-ZVFHMIN-NEXT:    bnez a2, .LBB76_14
 ; RV64ZVE32F-ZVFHMIN-NEXT:  # %bb.5: # %else5
 ; RV64ZVE32F-ZVFHMIN-NEXT:    andi a2, a1, 8
-; RV64ZVE32F-ZVFHMIN-NEXT:    bnez a2, .LBB76_13
+; RV64ZVE32F-ZVFHMIN-NEXT:    bnez a2, .LBB76_15
 ; RV64ZVE32F-ZVFHMIN-NEXT:  .LBB76_6: # %else8
 ; RV64ZVE32F-ZVFHMIN-NEXT:    andi a2, a1, 16
-; RV64ZVE32F-ZVFHMIN-NEXT:    bnez a2, .LBB76_14
+; RV64ZVE32F-ZVFHMIN-NEXT:    bnez a2, .LBB76_16
 ; RV64ZVE32F-ZVFHMIN-NEXT:  .LBB76_7: # %else11
 ; RV64ZVE32F-ZVFHMIN-NEXT:    andi a2, a1, 32
 ; RV64ZVE32F-ZVFHMIN-NEXT:    beqz a2, .LBB76_9
@@ -9130,14 +9079,37 @@ define <8 x half> @mgather_baseidx_zext_v8i8_v8f16(ptr %base, <8 x i8> %idxs, <8
 ; RV64ZVE32F-ZVFHMIN-NEXT:    andi a2, a1, 64
 ; RV64ZVE32F-ZVFHMIN-NEXT:    vsetivli zero, 2, e8, mf4, ta, ma
 ; RV64ZVE32F-ZVFHMIN-NEXT:    vslidedown.vi v8, v10, 2
-; RV64ZVE32F-ZVFHMIN-NEXT:    bnez a2, .LBB76_15
-; RV64ZVE32F-ZVFHMIN-NEXT:  # %bb.10: # %else17
+; RV64ZVE32F-ZVFHMIN-NEXT:    beqz a2, .LBB76_11
+; RV64ZVE32F-ZVFHMIN-NEXT:  # %bb.10: # %cond.load16
+; RV64ZVE32F-ZVFHMIN-NEXT:    vmv.x.s a2, v8
+; RV64ZVE32F-ZVFHMIN-NEXT:    andi a2, a2, 255
+; RV64ZVE32F-ZVFHMIN-NEXT:    slli a2, a2, 1
+; RV64ZVE32F-ZVFHMIN-NEXT:    add a2, a0, a2
+; RV64ZVE32F-ZVFHMIN-NEXT:    lh a2, 0(a2)
+; RV64ZVE32F-ZVFHMIN-NEXT:    vsetvli zero, zero, e16, mf2, ta, ma
+; RV64ZVE32F-ZVFHMIN-NEXT:    vmv.s.x v10, a2
+; RV64ZVE32F-ZVFHMIN-NEXT:    vsetivli zero, 7, e16, m1, tu, ma
+; RV64ZVE32F-ZVFHMIN-NEXT:    vslideup.vi v9, v10, 6
+; RV64ZVE32F-ZVFHMIN-NEXT:  .LBB76_11: # %else17
 ; RV64ZVE32F-ZVFHMIN-NEXT:    andi a1, a1, -128
-; RV64ZVE32F-ZVFHMIN-NEXT:    bnez a1, .LBB76_16
-; RV64ZVE32F-ZVFHMIN-NEXT:  .LBB76_11: # %else20
+; RV64ZVE32F-ZVFHMIN-NEXT:    beqz a1, .LBB76_13
+; RV64ZVE32F-ZVFHMIN-NEXT:  # %bb.12: # %cond.load19
+; RV64ZVE32F-ZVFHMIN-NEXT:    vsetivli zero, 1, e8, mf4, ta, ma
+; RV64ZVE32F-ZVFHMIN-NEXT:    vslidedown.vi v8, v8, 1
+; RV64ZVE32F-ZVFHMIN-NEXT:    vmv.x.s a1, v8
+; RV64ZVE32F-ZVFHMIN-NEXT:    andi a1, a1, 255
+; RV64ZVE32F-ZVFHMIN-NEXT:    slli a1, a1, 1
+; RV64ZVE32F-ZVFHMIN-NEXT:    add a0, a0, a1
+; RV64ZVE32F-ZVFHMIN-NEXT:    lh a0, 0(a0)
+; RV64ZVE32F-ZVFHMIN-NEXT:    vsetvli zero, zero, e16, mf2, ta, ma
+; RV64ZVE32F-ZVFHMIN-NEXT:    vmv.s.x v8, a0
+; RV64ZVE32F-ZVFHMIN-NEXT:    vsetivli zero, 8, e16, m1, ta, ma
+; RV64ZVE32F-ZVFHMIN-NEXT:    vslideup.vi v9, v8, 7
+; RV64ZVE32F-ZVFHMIN-NEXT:  .LBB76_13: # %else20
+; RV64ZVE32F-ZVFHMIN-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; RV64ZVE32F-ZVFHMIN-NEXT:    vmv1r.v v8, v9
 ; RV64ZVE32F-ZVFHMIN-NEXT:    ret
-; RV64ZVE32F-ZVFHMIN-NEXT:  .LBB76_12: # %cond.load4
+; RV64ZVE32F-ZVFHMIN-NEXT:  .LBB76_14: # %cond.load4
 ; RV64ZVE32F-ZVFHMIN-NEXT:    vmv.x.s a2, v8
 ; RV64ZVE32F-ZVFHMIN-NEXT:    andi a2, a2, 255
 ; RV64ZVE32F-ZVFHMIN-NEXT:    slli a2, a2, 1
@@ -9149,7 +9121,7 @@ define <8 x half> @mgather_baseidx_zext_v8i8_v8f16(ptr %base, <8 x i8> %idxs, <8
 ; RV64ZVE32F-ZVFHMIN-NEXT:    vslideup.vi v9, v11, 2
 ; RV64ZVE32F-ZVFHMIN-NEXT:    andi a2, a1, 8
 ; RV64ZVE32F-ZVFHMIN-NEXT:    beqz a2, .LBB76_6
-; RV64ZVE32F-ZVFHMIN-NEXT:  .LBB76_13: # %cond.load7
+; RV64ZVE32F-ZVFHMIN-NEXT:  .LBB76_15: # %cond.load7
 ; RV64ZVE32F-ZVFHMIN-NEXT:    vsetivli zero, 1, e8, mf4, ta, ma
 ; RV64ZVE32F-ZVFHMIN-NEXT:    vslidedown.vi v8, v8, 1
 ; RV64ZVE32F-ZVFHMIN-NEXT:    vmv.x.s a2, v8
@@ -9163,7 +9135,7 @@ define <8 x half> @mgather_baseidx_zext_v8i8_v8f16(ptr %base, <8 x i8> %idxs, <8
 ; RV64ZVE32F-ZVFHMIN-NEXT:    vslideup.vi v9, v8, 3
 ; RV64ZVE32F-ZVFHMIN-NEXT:    andi a2, a1, 16
 ; RV64ZVE32F-ZVFHMIN-NEXT:    beqz a2, .LBB76_7
-; RV64ZVE32F-ZVFHMIN-NEXT:  .LBB76_14: # %cond.load10
+; RV64ZVE32F-ZVFHMIN-NEXT:  .LBB76_16: # %cond.load10
 ; RV64ZVE32F-ZVFHMIN-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; RV64ZVE32F-ZVFHMIN-NEXT:    vmv.x.s a2, v10
 ; RV64ZVE32F-ZVFHMIN-NEXT:    andi a2, a2, 255
@@ -9177,32 +9149,6 @@ define <8 x half> @mgather_baseidx_zext_v8i8_v8f16(ptr %base, <8 x i8> %idxs, <8
 ; RV64ZVE32F-ZVFHMIN-NEXT:    andi a2, a1, 32
 ; RV64ZVE32F-ZVFHMIN-NEXT:    bnez a2, .LBB76_8
 ; RV64ZVE32F-ZVFHMIN-NEXT:    j .LBB76_9
-; RV64ZVE32F-ZVFHMIN-NEXT:  .LBB76_15: # %cond.load16
-; RV64ZVE32F-ZVFHMIN-NEXT:    vmv.x.s a2, v8
-; RV64ZVE32F-ZVFHMIN-NEXT:    andi a2, a2, 255
-; RV64ZVE32F-ZVFHMIN-NEXT:    slli a2, a2, 1
-; RV64ZVE32F-ZVFHMIN-NEXT:    add a2, a0, a2
-; RV64ZVE32F-ZVFHMIN-NEXT:    lh a2, 0(a2)
-; RV64ZVE32F-ZVFHMIN-NEXT:    vsetvli zero, zero, e16, mf2, ta, ma
-; RV64ZVE32F-ZVFHMIN-NEXT:    vmv.s.x v10, a2
-; RV64ZVE32F-ZVFHMIN-NEXT:    vsetivli zero, 7, e16, m1, tu, ma
-; RV64ZVE32F-ZVFHMIN-NEXT:    vslideup.vi v9, v10, 6
-; RV64ZVE32F-ZVFHMIN-NEXT:    andi a1, a1, -128
-; RV64ZVE32F-ZVFHMIN-NEXT:    beqz a1, .LBB76_11
-; RV64ZVE32F-ZVFHMIN-NEXT:  .LBB76_16: # %cond.load19
-; RV64ZVE32F-ZVFHMIN-NEXT:    vsetivli zero, 1, e8, mf4, ta, ma
-; RV64ZVE32F-ZVFHMIN-NEXT:    vslidedown.vi v8, v8, 1
-; RV64ZVE32F-ZVFHMIN-NEXT:    vmv.x.s a1, v8
-; RV64ZVE32F-ZVFHMIN-NEXT:    andi a1, a1, 255
-; RV64ZVE32F-ZVFHMIN-NEXT:    slli a1, a1, 1
-; RV64ZVE32F-ZVFHMIN-NEXT:    add a0, a0, a1
-; RV64ZVE32F-ZVFHMIN-NEXT:    lh a0, 0(a0)
-; RV64ZVE32F-ZVFHMIN-NEXT:    vsetvli zero, zero, e16, mf2, ta, ma
-; RV64ZVE32F-ZVFHMIN-NEXT:    vmv.s.x v8, a0
-; RV64ZVE32F-ZVFHMIN-NEXT:    vsetivli zero, 8, e16, m1, ta, ma
-; RV64ZVE32F-ZVFHMIN-NEXT:    vslideup.vi v9, v8, 7
-; RV64ZVE32F-ZVFHMIN-NEXT:    vmv1r.v v8, v9
-; RV64ZVE32F-ZVFHMIN-NEXT:    ret
   %eidxs = zext <8 x i8> %idxs to <8 x i16>
   %ptrs = getelementptr inbounds half, ptr %base, <8 x i16> %eidxs
   %v = call <8 x half> @llvm.masked.gather.v8f16.v8p0(<8 x ptr> %ptrs, i32 2, <8 x i1> %m, <8 x half> %passthru)
@@ -9260,13 +9206,13 @@ define <8 x half> @mgather_baseidx_v8f16(ptr %base, <8 x i16> %idxs, <8 x i1> %m
 ; RV64ZVE32F-ZVFH-NEXT:    andi a2, a1, 4
 ; RV64ZVE32F-ZVFH-NEXT:    vsetivli zero, 2, e16, mf2, ta, ma
 ; RV64ZVE32F-ZVFH-NEXT:    vslidedown.vi v8, v8, 2
-; RV64ZVE32F-ZVFH-NEXT:    bnez a2, .LBB77_12
+; RV64ZVE32F-ZVFH-NEXT:    bnez a2, .LBB77_14
 ; RV64ZVE32F-ZVFH-NEXT:  # %bb.5: # %else5
 ; RV64ZVE32F-ZVFH-NEXT:    andi a2, a1, 8
-; RV64ZVE32F-ZVFH-NEXT:    bnez a2, .LBB77_13
+; RV64ZVE32F-ZVFH-NEXT:    bnez a2, .LBB77_15
 ; RV64ZVE32F-ZVFH-NEXT:  .LBB77_6: # %else8
 ; RV64ZVE32F-ZVFH-NEXT:    andi a2, a1, 16
-; RV64ZVE32F-ZVFH-NEXT:    bnez a2, .LBB77_14
+; RV64ZVE32F-ZVFH-NEXT:    bnez a2, .LBB77_16
 ; RV64ZVE32F-ZVFH-NEXT:  .LBB77_7: # %else11
 ; RV64ZVE32F-ZVFH-NEXT:    andi a2, a1, 32
 ; RV64ZVE32F-ZVFH-NEXT:    beqz a2, .LBB77_9
@@ -9284,14 +9230,33 @@ define <8 x half> @mgather_baseidx_v8f16(ptr %base, <8 x i16> %idxs, <8 x i1> %m
 ; RV64ZVE32F-ZVFH-NEXT:    andi a2, a1, 64
 ; RV64ZVE32F-ZVFH-NEXT:    vsetivli zero, 2, e16, mf2, ta, ma
 ; RV64ZVE32F-ZVFH-NEXT:    vslidedown.vi v8, v10, 2
-; RV64ZVE32F-ZVFH-NEXT:    bnez a2, .LBB77_15
-; RV64ZVE32F-ZVFH-NEXT:  # %bb.10: # %else17
+; RV64ZVE32F-ZVFH-NEXT:    beqz a2, .LBB77_11
+; RV64ZVE32F-ZVFH-NEXT:  # %bb.10: # %cond.load16
+; RV64ZVE32F-ZVFH-NEXT:    vmv.x.s a2, v8
+; RV64ZVE32F-ZVFH-NEXT:    slli a2, a2, 1
+; RV64ZVE32F-ZVFH-NEXT:    add a2, a0, a2
+; RV64ZVE32F-ZVFH-NEXT:    flh fa5, 0(a2)
+; RV64ZVE32F-ZVFH-NEXT:    vfmv.s.f v10, fa5
+; RV64ZVE32F-ZVFH-NEXT:    vsetivli zero, 7, e16, m1, tu, ma
+; RV64ZVE32F-ZVFH-NEXT:    vslideup.vi v9, v10, 6
+; RV64ZVE32F-ZVFH-NEXT:  .LBB77_11: # %else17
 ; RV64ZVE32F-ZVFH-NEXT:    andi a1, a1, -128
-; RV64ZVE32F-ZVFH-NEXT:    bnez a1, .LBB77_16
-; RV64ZVE32F-ZVFH-NEXT:  .LBB77_11: # %else20
+; RV64ZVE32F-ZVFH-NEXT:    beqz a1, .LBB77_13
+; RV64ZVE32F-ZVFH-NEXT:  # %bb.12: # %cond.load19
+; RV64ZVE32F-ZVFH-NEXT:    vsetivli zero, 1, e16, mf2, ta, ma
+; RV64ZVE32F-ZVFH-NEXT:    vslidedown.vi v8, v8, 1
+; RV64ZVE32F-ZVFH-NEXT:    vmv.x.s a1, v8
+; RV64ZVE32F-ZVFH-NEXT:    slli a1, a1, 1
+; RV64ZVE32F-ZVFH-NEXT:    add a0, a0, a1
+; RV64ZVE32F-ZVFH-NEXT:    flh fa5, 0(a0)
+; RV64ZVE32F-ZVFH-NEXT:    vfmv.s.f v8, fa5
+; RV64ZVE32F-ZVFH-NEXT:    vsetivli zero, 8, e16, m1, ta, ma
+; RV64ZVE32F-ZVFH-NEXT:    vslideup.vi v9, v8, 7
+; RV64ZVE32F-ZVFH-NEXT:  .LBB77_13: # %else20
+; RV64ZVE32F-ZVFH-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; RV64ZVE32F-ZVFH-NEXT:    vmv1r.v v8, v9
 ; RV64ZVE32F-ZVFH-NEXT:    ret
-; RV64ZVE32F-ZVFH-NEXT:  .LBB77_12: # %cond.load4
+; RV64ZVE32F-ZVFH-NEXT:  .LBB77_14: # %cond.load4
 ; RV64ZVE32F-ZVFH-NEXT:    vmv.x.s a2, v8
 ; RV64ZVE32F-ZVFH-NEXT:    slli a2, a2, 1
 ; RV64ZVE32F-ZVFH-NEXT:    add a2, a0, a2
@@ -9301,7 +9266,7 @@ define <8 x half> @mgather_baseidx_v8f16(ptr %base, <8 x i16> %idxs, <8 x i1> %m
 ; RV64ZVE32F-ZVFH-NEXT:    vslideup.vi v9, v11, 2
 ; RV64ZVE32F-ZVFH-NEXT:    andi a2, a1, 8
 ; RV64ZVE32F-ZVFH-NEXT:    beqz a2, .LBB77_6
-; RV64ZVE32F-ZVFH-NEXT:  .LBB77_13: # %cond.load7
+; RV64ZVE32F-ZVFH-NEXT:  .LBB77_15: # %cond.load7
 ; RV64ZVE32F-ZVFH-NEXT:    vsetivli zero, 1, e16, mf2, ta, ma
 ; RV64ZVE32F-ZVFH-NEXT:    vslidedown.vi v8, v8, 1
 ; RV64ZVE32F-ZVFH-NEXT:    vmv.x.s a2, v8
@@ -9313,7 +9278,7 @@ define <8 x half> @mgather_baseidx_v8f16(ptr %base, <8 x i16> %idxs, <8 x i1> %m
 ; RV64ZVE32F-ZVFH-NEXT:    vslideup.vi v9, v8, 3
 ; RV64ZVE32F-ZVFH-NEXT:    andi a2, a1, 16
 ; RV64ZVE32F-ZVFH-NEXT:    beqz a2, .LBB77_7
-; RV64ZVE32F-ZVFH-NEXT:  .LBB77_14: # %cond.load10
+; RV64ZVE32F-ZVFH-NEXT:  .LBB77_16: # %cond.load10
 ; RV64ZVE32F-ZVFH-NEXT:    vsetivli zero, 5, e16, m1, tu, ma
 ; RV64ZVE32F-ZVFH-NEXT:    vmv.x.s a2, v10
 ; RV64ZVE32F-ZVFH-NEXT:    slli a2, a2, 1
@@ -9324,28 +9289,6 @@ define <8 x half> @mgather_baseidx_v8f16(ptr %base, <8 x i16> %idxs, <8 x i1> %m
 ; RV64ZVE32F-ZVFH-NEXT:    andi a2, a1, 32
 ; RV64ZVE32F-ZVFH-NEXT:    bnez a2, .LBB77_8
 ; RV64ZVE32F-ZVFH-NEXT:    j .LBB77_9
-; RV64ZVE32F-ZVFH-NEXT:  .LBB77_15: # %cond.load16
-; RV64ZVE32F-ZVFH-NEXT:    vmv.x.s a2, v8
-; RV64ZVE32F-ZVFH-NEXT:    slli a2, a2, 1
-; RV64ZVE32F-ZVFH-NEXT:    add a2, a0, a2
-; RV64ZVE32F-ZVFH-NEXT:    flh fa5, 0(a2)
-; RV64ZVE32F-ZVFH-NEXT:    vfmv.s.f v10, fa5
-; RV64ZVE32F-ZVFH-NEXT:    vsetivli zero, 7, e16, m1, tu, ma
-; RV64ZVE32F-ZVFH-NEXT:    vslideup.vi v9, v10, 6
-; RV64ZVE32F-ZVFH-NEXT:    andi a1, a1, -128
-; RV64ZVE32F-ZVFH-NEXT:    beqz a1, .LBB77_11
-; RV64ZVE32F-ZVFH-NEXT:  .LBB77_16: # %cond.load19
-; RV64ZVE32F-ZVFH-NEXT:    vsetivli zero, 1, e16, mf2, ta, ma
-; RV64ZVE32F-ZVFH-NEXT:    vslidedown.vi v8, v8, 1
-; RV64ZVE32F-ZVFH-NEXT:    vmv.x.s a1, v8
-; RV64ZVE32F-ZVFH-NEXT:    slli a1, a1, 1
-; RV64ZVE32F-ZVFH-NEXT:    add a0, a0, a1
-; RV64ZVE32F-ZVFH-NEXT:    flh fa5, 0(a0)
-; RV64ZVE32F-ZVFH-NEXT:    vfmv.s.f v8, fa5
-; RV64ZVE32F-ZVFH-NEXT:    vsetivli zero, 8, e16, m1, ta, ma
-; RV64ZVE32F-ZVFH-NEXT:    vslideup.vi v9, v8, 7
-; RV64ZVE32F-ZVFH-NEXT:    vmv1r.v v8, v9
-; RV64ZVE32F-ZVFH-NEXT:    ret
 ;
 ; RV64ZVE32F-ZVFHMIN-LABEL: mgather_baseidx_v8f16:
 ; RV64ZVE32F-ZVFHMIN:       # %bb.0:
@@ -9379,13 +9322,13 @@ define <8 x half> @mgather_baseidx_v8f16(ptr %base, <8 x i16> %idxs, <8 x i1> %m
 ; RV64ZVE32F-ZVFHMIN-NEXT:    andi a2, a1, 4
 ; RV64ZVE32F-ZVFHMIN-NEXT:    vsetivli zero, 2, e16, mf2, ta, ma
 ; RV64ZVE32F-ZVFHMIN-NEXT:    vslidedown.vi v8, v8, 2
-; RV64ZVE32F-ZVFHMIN-NEXT:    bnez a2, .LBB77_12
+; RV64ZVE32F-ZVFHMIN-NEXT:    bnez a2, .LBB77_14
 ; RV64ZVE32F-ZVFHMIN-NEXT:  # %bb.5: # %else5
 ; RV64ZVE32F-ZVFHMIN-NEXT:    andi a2, a1, 8
-; RV64ZVE32F-ZVFHMIN-NEXT:    bnez a2, .LBB77_13
+; RV64ZVE32F-ZVFHMIN-NEXT:    bnez a2, .LBB77_15
 ; RV64ZVE32F-ZVFHMIN-NEXT:  .LBB77_6: # %else8
 ; RV64ZVE32F-ZVFHMIN-NEXT:    andi a2, a1, 16
-; RV64ZVE32F-ZVFHMIN-NEXT:    bnez a2, .LBB77_14
+; RV64ZVE32F-ZVFHMIN-NEXT:    bnez a2, .LBB77_16
 ; RV64ZVE32F-ZVFHMIN-NEXT:  .LBB77_7: # %else11
 ; RV64ZVE32F-ZVFHMIN-NEXT:    andi a2, a1, 32
 ; RV64ZVE32F-ZVFHMIN-NEXT:    beqz a2, .LBB77_9
@@ -9403,14 +9346,33 @@ define <8 x half> @mgather_baseidx_v8f16(ptr %base, <8 x i16> %idxs, <8 x i1> %m
 ; RV64ZVE32F-ZVFHMIN-NEXT:    andi a2, a1, 64
 ; RV64ZVE32F-ZVFHMIN-NEXT:    vsetivli zero, 2, e16, mf2, ta, ma
 ; RV64ZVE32F-ZVFHMIN-NEXT:    vslidedown.vi v8, v10, 2
-; RV64ZVE32F-ZVFHMIN-NEXT:    bnez a2, .LBB77_15
-; RV64ZVE32F-ZVFHMIN-NEXT:  # %bb.10: # %else17
+; RV64ZVE32F-ZVFHMIN-NEXT:    beqz a2, .LBB77_11
+; RV64ZVE32F-ZVFHMIN-NEXT:  # %bb.10: # %cond.load16
+; RV64ZVE32F-ZVFHMIN-NEXT:    vmv.x.s a2, v8
+; RV64ZVE32F-ZVFHMIN-NEXT:    slli a2, a2, 1
+; RV64ZVE32F-ZVFHMIN-NEXT:    add a2, a0, a2
+; RV64ZVE32F-ZVFHMIN-NEXT:    lh a2, 0(a2)
+; RV64ZVE32F-ZVFHMIN-NEXT:    vmv.s.x v10, a2
+; RV64ZVE32F-ZVFHMIN-NEXT:    vsetivli zero, 7, e16, m1, tu, ma
+; RV64ZVE32F-ZVFHMIN-NEXT:    vslideup.vi v9, v10, 6
+; RV64ZVE32F-ZVFHMIN-NEXT:  .LBB77_11: # %else17
 ; RV64ZVE32F-ZVFHMIN-NEXT:    andi a1, a1, -128
-; RV64ZVE32F-ZVFHMIN-NEXT:    bnez a1, .LBB77_16
-; RV64ZVE32F-ZVFHMIN-NEXT:  .LBB77_11: # %else20
+; RV64ZVE32F-ZVFHMIN-NEXT:    beqz a1, .LBB77_13
+; RV64ZVE32F-ZVFHMIN-NEXT:  # %bb.12: # %cond.load19
+; RV64ZVE32F-ZVFHMIN-NEXT:    vsetivli zero, 1, e16, mf2, ta, ma
+; RV64ZVE32F-ZVFHMIN-NEXT:    vslidedown.vi v8, v8, 1
+; RV64ZVE32F-ZVFHMIN-NEXT:    vmv.x.s a1, v8
+; RV64ZVE32F-ZVFHMIN-NEXT:    slli a1, a1, 1
+; RV64ZVE32F-ZVFHMIN-NEXT:    add a0, a0, a1
+; RV64ZVE32F-ZVFHMIN-NEXT:    lh a0, 0(a0)
+; RV64ZVE32F-ZVFHMIN-NEXT:    vmv.s.x v8, a0
+; RV64ZVE32F-ZVFHMIN-NEXT:    vsetivli zero, 8, e16, m1, ta, ma
+; RV64ZVE32F-ZVFHMIN-NEXT:    vslideup.vi v9, v8, 7
+; RV64ZVE32F-ZVFHMIN-NEXT:  .LBB77_13: # %else20
+; RV64ZVE32F-ZVFHMIN-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; RV64ZVE32F-ZVFHMIN-NEXT:    vmv1r.v v8, v9
 ; RV64ZVE32F-ZVFHMIN-NEXT:    ret
-; RV64ZVE32F-ZVFHMIN-NEXT:  .LBB77_12: # %cond.load4
+; RV64ZVE32F-ZVFHMIN-NEXT:  .LBB77_14: # %cond.load4
 ; RV64ZVE32F-ZVFHMIN-NEXT:    vmv.x.s a2, v8
 ; RV64ZVE32F-ZVFHMIN-NEXT:    slli a2, a2, 1
 ; RV64ZVE32F-ZVFHMIN-NEXT:    add a2, a0, a2
@@ -9420,7 +9382,7 @@ define <8 x half> @mgather_baseidx_v8f16(ptr %base, <8 x i16> %idxs, <8 x i1> %m
 ; RV64ZVE32F-ZVFHMIN-NEXT:    vslideup.vi v9, v11, 2
 ; RV64ZVE32F-ZVFHMIN-NEXT:    andi a2, a1, 8
 ; RV64ZVE32F-ZVFHMIN-NEXT:    beqz a2, .LBB77_6
-; RV64ZVE32F-ZVFHMIN-NEXT:  .LBB77_13: # %cond.load7
+; RV64ZVE32F-ZVFHMIN-NEXT:  .LBB77_15: # %cond.load7
 ; RV64ZVE32F-ZVFHMIN-NEXT:    vsetivli zero, 1, e16, mf2, ta, ma
 ; RV64ZVE32F-ZVFHMIN-NEXT:    vslidedown.vi v8, v8, 1
 ; RV64ZVE32F-ZVFHMIN-NEXT:    vmv.x.s a2, v8
@@ -9432,7 +9394,7 @@ define <8 x half> @mgather_baseidx_v8f16(ptr %base, <8 x i16> %idxs, <8 x i1> %m
 ; RV64ZVE32F-ZVFHMIN-NEXT:    vslideup.vi v9, v8, 3
 ; RV64ZVE32F-ZVFHMIN-NEXT:    andi a2, a1, 16
 ; RV64ZVE32F-ZVFHMIN-NEXT:    beqz a2, .LBB77_7
-; RV64ZVE32F-ZVFHMIN-NEXT:  .LBB77_14: # %cond.load10
+; RV64ZVE32F-ZVFHMIN-NEXT:  .LBB77_16: # %cond.load10
 ; RV64ZVE32F-ZVFHMIN-NEXT:    vsetivli zero, 5, e16, m1, tu, ma
 ; RV64ZVE32F-ZVFHMIN-NEXT:    vmv.x.s a2, v10
 ; RV64ZVE32F-ZVFHMIN-NEXT:    slli a2, a2, 1
@@ -9443,28 +9405,6 @@ define <8 x half> @mgather_baseidx_v8f16(ptr %base, <8 x i16> %idxs, <8 x i1> %m
 ; RV64ZVE32F-ZVFHMIN-NEXT:    andi a2, a1, 32
 ; RV64ZVE32F-ZVFHMIN-NEXT:    bnez a2, .LBB77_8
 ; RV64ZVE32F-ZVFHMIN-NEXT:    j .LBB77_9
-; RV64ZVE32F-ZVFHMIN-NEXT:  .LBB77_15: # %cond.load16
-; RV64ZVE32F-ZVFHMIN-NEXT:    vmv.x.s a2, v8
-; RV64ZVE32F-ZVFHMIN-NEXT:    slli a2, a2, 1
-; RV64ZVE32F-ZVFHMIN-NEXT:    add a2, a0, a2
-; RV64ZVE32F-ZVFHMIN-NEXT:    lh a2, 0(a2)
-; RV64ZVE32F-ZVFHMIN-NEXT:    vmv.s.x v10, a2
-; RV64ZVE32F-ZVFHMIN-NEXT:    vsetivli zero, 7, e16, m1, tu, ma
-; RV64ZVE32F-ZVFHMIN-NEXT:    vslideup.vi v9, v10, 6
-; RV64ZVE32F-ZVFHMIN-NEXT:    andi a1, a1, -128
-; RV64ZVE32F-ZVFHMIN-NEXT:    beqz a1, .LBB77_11
-; RV64ZVE32F-ZVFHMIN-NEXT:  .LBB77_16: # %cond.load19
-; RV64ZVE32F-ZVFHMIN-NEXT:    vsetivli zero, 1, e16, mf2, ta, ma
-; RV64ZVE32F-ZVFHMIN-NEXT:    vslidedown.vi v8, v8, 1
-; RV64ZVE32F-ZVFHMIN-NEXT:    vmv.x.s a1, v8
-; RV64ZVE32F-ZVFHMIN-NEXT:    slli a1, a1, 1
-; RV64ZVE32F-ZVFHMIN-NEXT:    add a0, a0, a1
-; RV64ZVE32F-ZVFHMIN-NEXT:    lh a0, 0(a0)
-; RV64ZVE32F-ZVFHMIN-NEXT:    vmv.s.x v8, a0
-; RV64ZVE32F-ZVFHMIN-NEXT:    vsetivli zero, 8, e16, m1, ta, ma
-; RV64ZVE32F-ZVFHMIN-NEXT:    vslideup.vi v9, v8, 7
-; RV64ZVE32F-ZVFHMIN-NEXT:    vmv1r.v v8, v9
-; RV64ZVE32F-ZVFHMIN-NEXT:    ret
   %ptrs = getelementptr inbounds half, ptr %base, <8 x i16> %idxs
   %v = call <8 x half> @llvm.masked.gather.v8f16.v8p0(<8 x ptr> %ptrs, i32 2, <8 x i1> %m, <8 x half> %passthru)
   ret <8 x half> %v
@@ -9666,11 +9606,13 @@ define <4 x float> @mgather_truemask_v4f32(<4 x ptr> %ptrs, <4 x float> %passthr
 define <4 x float> @mgather_falsemask_v4f32(<4 x ptr> %ptrs, <4 x float> %passthru) {
 ; RV32-LABEL: mgather_falsemask_v4f32:
 ; RV32:       # %bb.0:
+; RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; RV32-NEXT:    vmv1r.v v8, v9
 ; RV32-NEXT:    ret
 ;
 ; RV64V-LABEL: mgather_falsemask_v4f32:
 ; RV64V:       # %bb.0:
+; RV64V-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; RV64V-NEXT:    vmv1r.v v8, v10
 ; RV64V-NEXT:    ret
 ;
@@ -9847,13 +9789,13 @@ define <8 x float> @mgather_baseidx_v8i8_v8f32(ptr %base, <8 x i8> %idxs, <8 x i
 ; RV64ZVE32F-NEXT:    andi a2, a1, 4
 ; RV64ZVE32F-NEXT:    vsetivli zero, 2, e8, mf4, ta, ma
 ; RV64ZVE32F-NEXT:    vslidedown.vi v8, v8, 2
-; RV64ZVE32F-NEXT:    bnez a2, .LBB84_12
+; RV64ZVE32F-NEXT:    bnez a2, .LBB84_14
 ; RV64ZVE32F-NEXT:  # %bb.5: # %else5
 ; RV64ZVE32F-NEXT:    andi a2, a1, 8
-; RV64ZVE32F-NEXT:    bnez a2, .LBB84_13
+; RV64ZVE32F-NEXT:    bnez a2, .LBB84_15
 ; RV64ZVE32F-NEXT:  .LBB84_6: # %else8
 ; RV64ZVE32F-NEXT:    andi a2, a1, 16
-; RV64ZVE32F-NEXT:    bnez a2, .LBB84_14
+; RV64ZVE32F-NEXT:    bnez a2, .LBB84_16
 ; RV64ZVE32F-NEXT:  .LBB84_7: # %else11
 ; RV64ZVE32F-NEXT:    andi a2, a1, 32
 ; RV64ZVE32F-NEXT:    beqz a2, .LBB84_9
@@ -9872,14 +9814,35 @@ define <8 x float> @mgather_baseidx_v8i8_v8f32(ptr %base, <8 x i8> %idxs, <8 x i
 ; RV64ZVE32F-NEXT:    andi a2, a1, 64
 ; RV64ZVE32F-NEXT:    vsetivli zero, 2, e8, mf4, ta, ma
 ; RV64ZVE32F-NEXT:    vslidedown.vi v8, v9, 2
-; RV64ZVE32F-NEXT:    bnez a2, .LBB84_15
-; RV64ZVE32F-NEXT:  # %bb.10: # %else17
+; RV64ZVE32F-NEXT:    beqz a2, .LBB84_11
+; RV64ZVE32F-NEXT:  # %bb.10: # %cond.load16
+; RV64ZVE32F-NEXT:    vmv.x.s a2, v8
+; RV64ZVE32F-NEXT:    slli a2, a2, 2
+; RV64ZVE32F-NEXT:    add a2, a0, a2
+; RV64ZVE32F-NEXT:    flw fa5, 0(a2)
+; RV64ZVE32F-NEXT:    vsetvli zero, zero, e32, m1, ta, ma
+; RV64ZVE32F-NEXT:    vfmv.s.f v12, fa5
+; RV64ZVE32F-NEXT:    vsetivli zero, 7, e32, m2, tu, ma
+; RV64ZVE32F-NEXT:    vslideup.vi v10, v12, 6
+; RV64ZVE32F-NEXT:  .LBB84_11: # %else17
 ; RV64ZVE32F-NEXT:    andi a1, a1, -128
-; RV64ZVE32F-NEXT:    bnez a1, .LBB84_16
-; RV64ZVE32F-NEXT:  .LBB84_11: # %else20
+; RV64ZVE32F-NEXT:    beqz a1, .LBB84_13
+; RV64ZVE32F-NEXT:  # %bb.12: # %cond.load19
+; RV64ZVE32F-NEXT:    vsetivli zero, 1, e8, mf4, ta, ma
+; RV64ZVE32F-NEXT:    vslidedown.vi v8, v8, 1
+; RV64ZVE32F-NEXT:    vmv.x.s a1, v8
+; RV64ZVE32F-NEXT:    slli a1, a1, 2
+; RV64ZVE32F-NEXT:    add a0, a0, a1
+; RV64ZVE32F-NEXT:    flw fa5, 0(a0)
+; RV64ZVE32F-NEXT:    vsetvli zero, zero, e32, m1, ta, ma
+; RV64ZVE32F-NEXT:    vfmv.s.f v8, fa5
+; RV64ZVE32F-NEXT:    vsetivli zero, 8, e32, m2, ta, ma
+; RV64ZVE32F-NEXT:    vslideup.vi v10, v8, 7
+; RV64ZVE32F-NEXT:  .LBB84_13: # %else20
+; RV64ZVE32F-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; RV64ZVE32F-NEXT:    vmv2r.v v8, v10
 ; RV64ZVE32F-NEXT:    ret
-; RV64ZVE32F-NEXT:  .LBB84_12: # %cond.load4
+; RV64ZVE32F-NEXT:  .LBB84_14: # %cond.load4
 ; RV64ZVE32F-NEXT:    vmv.x.s a2, v8
 ; RV64ZVE32F-NEXT:    slli a2, a2, 2
 ; RV64ZVE32F-NEXT:    add a2, a0, a2
@@ -9890,7 +9853,7 @@ define <8 x float> @mgather_baseidx_v8i8_v8f32(ptr %base, <8 x i8> %idxs, <8 x i
 ; RV64ZVE32F-NEXT:    vslideup.vi v10, v12, 2
 ; RV64ZVE32F-NEXT:    andi a2, a1, 8
 ; RV64ZVE32F-NEXT:    beqz a2, .LBB84_6
-; RV64ZVE32F-NEXT:  .LBB84_13: # %cond.load7
+; RV64ZVE32F-NEXT:  .LBB84_15: # %cond.load7
 ; RV64ZVE32F-NEXT:    vsetivli zero, 1, e8, mf4, ta, ma
 ; RV64ZVE32F-NEXT:    vslidedown.vi v8, v8, 1
 ; RV64ZVE32F-NEXT:    vmv.x.s a2, v8
@@ -9903,7 +9866,7 @@ define <8 x float> @mgather_baseidx_v8i8_v8f32(ptr %base, <8 x i8> %idxs, <8 x i
 ; RV64ZVE32F-NEXT:    vslideup.vi v10, v8, 3
 ; RV64ZVE32F-NEXT:    andi a2, a1, 16
 ; RV64ZVE32F-NEXT:    beqz a2, .LBB84_7
-; RV64ZVE32F-NEXT:  .LBB84_14: # %cond.load10
+; RV64ZVE32F-NEXT:  .LBB84_16: # %cond.load10
 ; RV64ZVE32F-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; RV64ZVE32F-NEXT:    vmv.x.s a2, v9
 ; RV64ZVE32F-NEXT:    slli a2, a2, 2
@@ -9916,30 +9879,6 @@ define <8 x float> @mgather_baseidx_v8i8_v8f32(ptr %base, <8 x i8> %idxs, <8 x i
 ; RV64ZVE32F-NEXT:    andi a2, a1, 32
 ; RV64ZVE32F-NEXT:    bnez a2, .LBB84_8
 ; RV64ZVE32F-NEXT:    j .LBB84_9
-; RV64ZVE32F-NEXT:  .LBB84_15: # %cond.load16
-; RV64ZVE32F-NEXT:    vmv.x.s a2, v8
-; RV64ZVE32F-NEXT:    slli a2, a2, 2
-; RV64ZVE32F-NEXT:    add a2, a0, a2
-; RV64ZVE32F-NEXT:    flw fa5, 0(a2)
-; RV64ZVE32F-NEXT:    vsetvli zero, zero, e32, m1, ta, ma
-; RV64ZVE32F-NEXT:    vfmv.s.f v12, fa5
-; RV64ZVE32F-NEXT:    vsetivli zero, 7, e32, m2, tu, ma
-; RV64ZVE32F-NEXT:    vslideup.vi v10, v12, 6
-; RV64ZVE32F-NEXT:    andi a1, a1, -128
-; RV64ZVE32F-NEXT:    beqz a1, .LBB84_11
-; RV64ZVE32F-NEXT:  .LBB84_16: # %cond.load19
-; RV64ZVE32F-NEXT:    vsetivli zero, 1, e8, mf4, ta, ma
-; RV64ZVE32F-NEXT:    vslidedown.vi v8, v8, 1
-; RV64ZVE32F-NEXT:    vmv.x.s a1, v8
-; RV64ZVE32F-NEXT:    slli a1, a1, 2
-; RV64ZVE32F-NEXT:    add a0, a0, a1
-; RV64ZVE32F-NEXT:    flw fa5, 0(a0)
-; RV64ZVE32F-NEXT:    vsetvli zero, zero, e32, m1, ta, ma
-; RV64ZVE32F-NEXT:    vfmv.s.f v8, fa5
-; RV64ZVE32F-NEXT:    vsetivli zero, 8, e32, m2, ta, ma
-; RV64ZVE32F-NEXT:    vslideup.vi v10, v8, 7
-; RV64ZVE32F-NEXT:    vmv2r.v v8, v10
-; RV64ZVE32F-NEXT:    ret
   %ptrs = getelementptr inbounds float, ptr %base, <8 x i8> %idxs
   %v = call <8 x float> @llvm.masked.gather.v8f32.v8p0(<8 x ptr> %ptrs, i32 4, <8 x i1> %m, <8 x float> %passthru)
   ret <8 x float> %v
@@ -9998,13 +9937,13 @@ define <8 x float> @mgather_baseidx_sext_v8i8_v8f32(ptr %base, <8 x i8> %idxs, <
 ; RV64ZVE32F-NEXT:    andi a2, a1, 4
 ; RV64ZVE32F-NEXT:    vsetivli zero, 2, e8, mf4, ta, ma
 ; RV64ZVE32F-NEXT:    vslidedown.vi v8, v8, 2
-; RV64ZVE32F-NEXT:    bnez a2, .LBB85_12
+; RV64ZVE32F-NEXT:    bnez a2, .LBB85_14
 ; RV64ZVE32F-NEXT:  # %bb.5: # %else5
 ; RV64ZVE32F-NEXT:    andi a2, a1, 8
-; RV64ZVE32F-NEXT:    bnez a2, .LBB85_13
+; RV64ZVE32F-NEXT:    bnez a2, .LBB85_15
 ; RV64ZVE32F-NEXT:  .LBB85_6: # %else8
 ; RV64ZVE32F-NEXT:    andi a2, a1, 16
-; RV64ZVE32F-NEXT:    bnez a2, .LBB85_14
+; RV64ZVE32F-NEXT:    bnez a2, .LBB85_16
 ; RV64ZVE32F-NEXT:  .LBB85_7: # %else11
 ; RV64ZVE32F-NEXT:    andi a2, a1, 32
 ; RV64ZVE32F-NEXT:    beqz a2, .LBB85_9
@@ -10023,74 +9962,71 @@ define <8 x float> @mgather_baseidx_sext_v8i8_v8f32(ptr %base, <8 x i8> %idxs, <
 ; RV64ZVE32F-NEXT:    andi a2, a1, 64
 ; RV64ZVE32F-NEXT:    vsetivli zero, 2, e8, mf4, ta, ma
 ; RV64ZVE32F-NEXT:    vslidedown.vi v8, v9, 2
-; RV64ZVE32F-NEXT:    bnez a2, .LBB85_15
-; RV64ZVE32F-NEXT:  # %bb.10: # %else17
-; RV64ZVE32F-NEXT:    andi a1, a1, -128
-; RV64ZVE32F-NEXT:    bnez a1, .LBB85_16
-; RV64ZVE32F-NEXT:  .LBB85_11: # %else20
-; RV64ZVE32F-NEXT:    vmv2r.v v8, v10
-; RV64ZVE32F-NEXT:    ret
-; RV64ZVE32F-NEXT:  .LBB85_12: # %cond.load4
+; RV64ZVE32F-NEXT:    beqz a2, .LBB85_11
+; RV64ZVE32F-NEXT:  # %bb.10: # %cond.load16
 ; RV64ZVE32F-NEXT:    vmv.x.s a2, v8
 ; RV64ZVE32F-NEXT:    slli a2, a2, 2
 ; RV64ZVE32F-NEXT:    add a2, a0, a2
 ; RV64ZVE32F-NEXT:    flw fa5, 0(a2)
 ; RV64ZVE32F-NEXT:    vsetvli zero, zero, e32, m1, ta, ma
 ; RV64ZVE32F-NEXT:    vfmv.s.f v12, fa5
-; RV64ZVE32F-NEXT:    vsetivli zero, 3, e32, m1, tu, ma
-; RV64ZVE32F-NEXT:    vslideup.vi v10, v12, 2
-; RV64ZVE32F-NEXT:    andi a2, a1, 8
-; RV64ZVE32F-NEXT:    beqz a2, .LBB85_6
-; RV64ZVE32F-NEXT:  .LBB85_13: # %cond.load7
+; RV64ZVE32F-NEXT:    vsetivli zero, 7, e32, m2, tu, ma
+; RV64ZVE32F-NEXT:    vslideup.vi v10, v12, 6
+; RV64ZVE32F-NEXT:  .LBB85_11: # %else17
+; RV64ZVE32F-NEXT:    andi a1, a1, -128
+; RV64ZVE32F-NEXT:    beqz a1, .LBB85_13
+; RV64ZVE32F-NEXT:  # %bb.12: # %cond.load19
 ; RV64ZVE32F-NEXT:    vsetivli zero, 1, e8, mf4, ta, ma
 ; RV64ZVE32F-NEXT:    vslidedown.vi v8, v8, 1
-; RV64ZVE32F-NEXT:    vmv.x.s a2, v8
-; RV64ZVE32F-NEXT:    slli a2, a2, 2
-; RV64ZVE32F-NEXT:    add a2, a0, a2
-; RV64ZVE32F-NEXT:    flw fa5, 0(a2)
+; RV64ZVE32F-NEXT:    vmv.x.s a1, v8
+; RV64ZVE32F-NEXT:    slli a1, a1, 2
+; RV64ZVE32F-NEXT:    add a0, a0, a1
+; RV64ZVE32F-NEXT:    flw fa5, 0(a0)
 ; RV64ZVE32F-NEXT:    vsetvli zero, zero, e32, m1, ta, ma
 ; RV64ZVE32F-NEXT:    vfmv.s.f v8, fa5
-; RV64ZVE32F-NEXT:    vsetivli zero, 4, e32, m1, tu, ma
-; RV64ZVE32F-NEXT:    vslideup.vi v10, v8, 3
-; RV64ZVE32F-NEXT:    andi a2, a1, 16
-; RV64ZVE32F-NEXT:    beqz a2, .LBB85_7
-; RV64ZVE32F-NEXT:  .LBB85_14: # %cond.load10
+; RV64ZVE32F-NEXT:    vsetivli zero, 8, e32, m2, ta, ma
+; RV64ZVE32F-NEXT:    vslideup.vi v10, v8, 7
+; RV64ZVE32F-NEXT:  .LBB85_13: # %else20
 ; RV64ZVE32F-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
-; RV64ZVE32F-NEXT:    vmv.x.s a2, v9
-; RV64ZVE32F-NEXT:    slli a2, a2, 2
-; RV64ZVE32F-NEXT:    add a2, a0, a2
-; RV64ZVE32F-NEXT:    flw fa5, 0(a2)
-; RV64ZVE32F-NEXT:    vsetvli zero, zero, e32, m4, ta, ma
-; RV64ZVE32F-NEXT:    vfmv.s.f v12, fa5
-; RV64ZVE32F-NEXT:    vsetivli zero, 5, e32, m2, tu, ma
-; RV64ZVE32F-NEXT:    vslideup.vi v10, v12, 4
-; RV64ZVE32F-NEXT:    andi a2, a1, 32
-; RV64ZVE32F-NEXT:    bnez a2, .LBB85_8
-; RV64ZVE32F-NEXT:    j .LBB85_9
-; RV64ZVE32F-NEXT:  .LBB85_15: # %cond.load16
+; RV64ZVE32F-NEXT:    vmv2r.v v8, v10
+; RV64ZVE32F-NEXT:    ret
+; RV64ZVE32F-NEXT:  .LBB85_14: # %cond.load4
 ; RV64ZVE32F-NEXT:    vmv.x.s a2, v8
 ; RV64ZVE32F-NEXT:    slli a2, a2, 2
 ; RV64ZVE32F-NEXT:    add a2, a0, a2
 ; RV64ZVE32F-NEXT:    flw fa5, 0(a2)
 ; RV64ZVE32F-NEXT:    vsetvli zero, zero, e32, m1, ta, ma
 ; RV64ZVE32F-NEXT:    vfmv.s.f v12, fa5
-; RV64ZVE32F-NEXT:    vsetivli zero, 7, e32, m2, tu, ma
-; RV64ZVE32F-NEXT:    vslideup.vi v10, v12, 6
-; RV64ZVE32F-NEXT:    andi a1, a1, -128
-; RV64ZVE32F-NEXT:    beqz a1, .LBB85_11
-; RV64ZVE32F-NEXT:  .LBB85_16: # %cond.load19
+; RV64ZVE32F-NEXT:    vsetivli zero, 3, e32, m1, tu, ma
+; RV64ZVE32F-NEXT:    vslideup.vi v10, v12, 2
+; RV64ZVE32F-NEXT:    andi a2, a1, 8
+; RV64ZVE32F-NEXT:    beqz a2, .LBB85_6
+; RV64ZVE32F-NEXT:  .LBB85_15: # %cond.load7
 ; RV64ZVE32F-NEXT:    vsetivli zero, 1, e8, mf4, ta, ma
 ; RV64ZVE32F-NEXT:    vslidedown.vi v8, v8, 1
-; RV64ZVE32F-NEXT:    vmv.x.s a1, v8
-; RV64ZVE32F-NEXT:    slli a1, a1, 2
-; RV64ZVE32F-NEXT:    add a0, a0, a1
-; RV64ZVE32F-NEXT:    flw fa5, 0(a0)
+; RV64ZVE32F-NEXT:    vmv.x.s a2, v8
+; RV64ZVE32F-NEXT:    slli a2, a2, 2
+; RV64ZVE32F-NEXT:    add a2, a0, a2
+; RV64ZVE32F-NEXT:    flw fa5, 0(a2)
 ; RV64ZVE32F-NEXT:    vsetvli zero, zero, e32, m1, ta, ma
 ; RV64ZVE32F-NEXT:    vfmv.s.f v8, fa5
-; RV64ZVE32F-NEXT:    vsetivli zero, 8, e32, m2, ta, ma
-; RV64ZVE32F-NEXT:    vslideup.vi v10, v8, 7
-; RV64ZVE32F-NEXT:    vmv2r.v v8, v10
-; RV64ZVE32F-NEXT:    ret
+; RV64ZVE32F-NEXT:    vsetivli zero, 4, e32, m1, tu, ma
+; RV64ZVE32F-NEXT:    vslideup.vi v10, v8, 3
+; RV64ZVE32F-NEXT:    andi a2, a1, 16
+; RV64ZVE32F-NEXT:    beqz a2, .LBB85_7
+; RV64ZVE32F-NEXT:  .LBB85_16: # %cond.load10
+; RV64ZVE32F-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
+; RV64ZVE32F-NEXT:    vmv.x.s a2, v9
+; RV64ZVE32F-NEXT:    slli a2, a2, 2
+; RV64ZVE32F-NEXT:    add a2, a0, a2
+; RV64ZVE32F-NEXT:    flw fa5, 0(a2)
+; RV64ZVE32F-NEXT:    vsetvli zero, zero, e32, m4, ta, ma
+; RV64ZVE32F-NEXT:    vfmv.s.f v12, fa5
+; RV64ZVE32F-NEXT:    vsetivli zero, 5, e32, m2, tu, ma
+; RV64ZVE32F-NEXT:    vslideup.vi v10, v12, 4
+; RV64ZVE32F-NEXT:    andi a2, a1, 32
+; RV64ZVE32F-NEXT:    bnez a2, .LBB85_8
+; RV64ZVE32F-NEXT:    j .LBB85_9
   %eidxs = sext <8 x i8> %idxs to <8 x i32>
   %ptrs = getelementptr inbounds float, ptr %base, <8 x i32> %eidxs
   %v = call <8 x float> @llvm.masked.gather.v8f32.v8p0(<8 x ptr> %ptrs, i32 4, <8 x i1> %m, <8 x float> %passthru)
@@ -10153,13 +10089,13 @@ define <8 x float> @mgather_baseidx_zext_v8i8_v8f32(ptr %base, <8 x i8> %idxs, <
 ; RV64ZVE32F-NEXT:    andi a2, a1, 4
 ; RV64ZVE32F-NEXT:    vsetivli zero, 2, e8, mf4, ta, ma
 ; RV64ZVE32F-NEXT:    vslidedown.vi v8, v8, 2
-; RV64ZVE32F-NEXT:    bnez a2, .LBB86_12
+; RV64ZVE32F-NEXT:    bnez a2, .LBB86_14
 ; RV64ZVE32F-NEXT:  # %bb.5: # %else5
 ; RV64ZVE32F-NEXT:    andi a2, a1, 8
-; RV64ZVE32F-NEXT:    bnez a2, .LBB86_13
+; RV64ZVE32F-NEXT:    bnez a2, .LBB86_15
 ; RV64ZVE32F-NEXT:  .LBB86_6: # %else8
 ; RV64ZVE32F-NEXT:    andi a2, a1, 16
-; RV64ZVE32F-NEXT:    bnez a2, .LBB86_14
+; RV64ZVE32F-NEXT:    bnez a2, .LBB86_16
 ; RV64ZVE32F-NEXT:  .LBB86_7: # %else11
 ; RV64ZVE32F-NEXT:    andi a2, a1, 32
 ; RV64ZVE32F-NEXT:    beqz a2, .LBB86_9
@@ -10179,14 +10115,37 @@ define <8 x float> @mgather_baseidx_zext_v8i8_v8f32(ptr %base, <8 x i8> %idxs, <
 ; RV64ZVE32F-NEXT:    andi a2, a1, 64
 ; RV64ZVE32F-NEXT:    vsetivli zero, 2, e8, mf4, ta, ma
 ; RV64ZVE32F-NEXT:    vslidedown.vi v8, v9, 2
-; RV64ZVE32F-NEXT:    bnez a2, .LBB86_15
-; RV64ZVE32F-NEXT:  # %bb.10: # %else17
+; RV64ZVE32F-NEXT:    beqz a2, .LBB86_11
+; RV64ZVE32F-NEXT:  # %bb.10: # %cond.load16
+; RV64ZVE32F-NEXT:    vmv.x.s a2, v8
+; RV64ZVE32F-NEXT:    andi a2, a2, 255
+; RV64ZVE32F-NEXT:    slli a2, a2, 2
+; RV64ZVE32F-NEXT:    add a2, a0, a2
+; RV64ZVE32F-NEXT:    flw fa5, 0(a2)
+; RV64ZVE32F-NEXT:    vsetvli zero, zero, e32, m1, ta, ma
+; RV64ZVE32F-NEXT:    vfmv.s.f v12, fa5
+; RV64ZVE32F-NEXT:    vsetivli zero, 7, e32, m2, tu, ma
+; RV64ZVE32F-NEXT:    vslideup.vi v10, v12, 6
+; RV64ZVE32F-NEXT:  .LBB86_11: # %else17
 ; RV64ZVE32F-NEXT:    andi a1, a1, -128
-; RV64ZVE32F-NEXT:    bnez a1, .LBB86_16
-; RV64ZVE32F-NEXT:  .LBB86_11: # %else20
+; RV64ZVE32F-NEXT:    beqz a1, .LBB86_13
+; RV64ZVE32F-NEXT:  # %bb.12: # %cond.load19
+; RV64ZVE32F-NEXT:    vsetivli zero, 1, e8, mf4, ta, ma
+; RV64ZVE32F-NEXT:    vslidedown.vi v8, v8, 1
+; RV64ZVE32F-NEXT:    vmv.x.s a1, v8
+; RV64ZVE32F-NEXT:    andi a1, a1, 255
+; RV64ZVE32F-NEXT:    slli a1, a1, 2
+; RV64ZVE32F-NEXT:    add a0, a0, a1
+; RV64ZVE32F-NEXT:    flw fa5, 0(a0)
+; RV64ZVE32F-NEXT:    vsetvli zero, zero, e32, m1, ta, ma
+; RV64ZVE32F-NEXT:    vfmv.s.f v8, fa5
+; RV64ZVE32F-NEXT:    vsetivli zero, 8, e32, m2, ta, ma
+; RV64ZVE32F-NEXT:    vslideup.vi v10, v8, 7
+; RV64ZVE32F-NEXT:  .LBB86_13: # %else20
+; RV64ZVE32F-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; RV64ZVE32F-NEXT:    vmv2r.v v8, v10
 ; RV64ZVE32F-NEXT:    ret
-; RV64ZVE32F-NEXT:  .LBB86_12: # %cond.load4
+; RV64ZVE32F-NEXT:  .LBB86_14: # %cond.load4
 ; RV64ZVE32F-NEXT:    vmv.x.s a2, v8
 ; RV64ZVE32F-NEXT:    andi a2, a2, 255
 ; RV64ZVE32F-NEXT:    slli a2, a2, 2
@@ -10198,7 +10157,7 @@ define <8 x float> @mgather_baseidx_zext_v8i8_v8f32(ptr %base, <8 x i8> %idxs, <
 ; RV64ZVE32F-NEXT:    vslideup.vi v10, v12, 2
 ; RV64ZVE32F-NEXT:    andi a2, a1, 8
 ; RV64ZVE32F-NEXT:    beqz a2, .LBB86_6
-; RV64ZVE32F-NEXT:  .LBB86_13: # %cond.load7
+; RV64ZVE32F-NEXT:  .LBB86_15: # %cond.load7
 ; RV64ZVE32F-NEXT:    vsetivli zero, 1, e8, mf4, ta, ma
 ; RV64ZVE32F-NEXT:    vslidedown.vi v8, v8, 1
 ; RV64ZVE32F-NEXT:    vmv.x.s a2, v8
@@ -10212,7 +10171,7 @@ define <8 x float> @mgather_baseidx_zext_v8i8_v8f32(ptr %base, <8 x i8> %idxs, <
 ; RV64ZVE32F-NEXT:    vslideup.vi v10, v8, 3
 ; RV64ZVE32F-NEXT:    andi a2, a1, 16
 ; RV64ZVE32F-NEXT:    beqz a2, .LBB86_7
-; RV64ZVE32F-NEXT:  .LBB86_14: # %cond.load10
+; RV64ZVE32F-NEXT:  .LBB86_16: # %cond.load10
 ; RV64ZVE32F-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; RV64ZVE32F-NEXT:    vmv.x.s a2, v9
 ; RV64ZVE32F-NEXT:    andi a2, a2, 255
@@ -10226,32 +10185,6 @@ define <8 x float> @mgather_baseidx_zext_v8i8_v8f32(ptr %base, <8 x i8> %idxs, <
 ; RV64ZVE32F-NEXT:    andi a2, a1, 32
 ; RV64ZVE32F-NEXT:    bnez a2, .LBB86_8
 ; RV64ZVE32F-NEXT:    j .LBB86_9
-; RV64ZVE32F-NEXT:  .LBB86_15: # %cond.load16
-; RV64ZVE32F-NEXT:    vmv.x.s a2, v8
-; RV64ZVE32F-NEXT:    andi a2, a2, 255
-; RV64ZVE32F-NEXT:    slli a2, a2, 2
-; RV64ZVE32F-NEXT:    add a2, a0, a2
-; RV64ZVE32F-NEXT:    flw fa5, 0(a2)
-; RV64ZVE32F-NEXT:    vsetvli zero, zero, e32, m1, ta, ma
-; RV64ZVE32F-NEXT:    vfmv.s.f v12, fa5
-; RV64ZVE32F-NEXT:    vsetivli zero, 7, e32, m2, tu, ma
-; RV64ZVE32F-NEXT:    vslideup.vi v10, v12, 6
-; RV64ZVE32F-NEXT:    andi a1, a1, -128
-; RV64ZVE32F-NEXT:    beqz a1, .LBB86_11
-; RV64ZVE32F-NEXT:  .LBB86_16: # %cond.load19
-; RV64ZVE32F-NEXT:    vsetivli zero, 1, e8, mf4, ta, ma
-; RV64ZVE32F-NEXT:    vslidedown.vi v8, v8, 1
-; RV64ZVE32F-NEXT:    vmv.x.s a1, v8
-; RV64ZVE32F-NEXT:    andi a1, a1, 255
-; RV64ZVE32F-NEXT:    slli a1, a1, 2
-; RV64ZVE32F-NEXT:    add a0, a0, a1
-; RV64ZVE32F-NEXT:    flw fa5, 0(a0)
-; RV64ZVE32F-NEXT:    vsetvli zero, zero, e32, m1, ta, ma
-; RV64ZVE32F-NEXT:    vfmv.s.f v8, fa5
-; RV64ZVE32F-NEXT:    vsetivli zero, 8, e32, m2, ta, ma
-; RV64ZVE32F-NEXT:    vslideup.vi v10, v8, 7
-; RV64ZVE32F-NEXT:    vmv2r.v v8, v10
-; RV64ZVE32F-NEXT:    ret
   %eidxs = zext <8 x i8> %idxs to <8 x i32>
   %ptrs = getelementptr inbounds float, ptr %base, <8 x i32> %eidxs
   %v = call <8 x float> @llvm.masked.gather.v8f32.v8p0(<8 x ptr> %ptrs, i32 4, <8 x i1> %m, <8 x float> %passthru)
@@ -10312,13 +10245,13 @@ define <8 x float> @mgather_baseidx_v8i16_v8f32(ptr %base, <8 x i16> %idxs, <8 x
 ; RV64ZVE32F-NEXT:    andi a2, a1, 4
 ; RV64ZVE32F-NEXT:    vsetivli zero, 2, e16, mf2, ta, ma
 ; RV64ZVE32F-NEXT:    vslidedown.vi v8, v8, 2
-; RV64ZVE32F-NEXT:    bnez a2, .LBB87_12
+; RV64ZVE32F-NEXT:    bnez a2, .LBB87_14
 ; RV64ZVE32F-NEXT:  # %bb.5: # %else5
 ; RV64ZVE32F-NEXT:    andi a2, a1, 8
-; RV64ZVE32F-NEXT:    bnez a2, .LBB87_13
+; RV64ZVE32F-NEXT:    bnez a2, .LBB87_15
 ; RV64ZVE32F-NEXT:  .LBB87_6: # %else8
 ; RV64ZVE32F-NEXT:    andi a2, a1, 16
-; RV64ZVE32F-NEXT:    bnez a2, .LBB87_14
+; RV64ZVE32F-NEXT:    bnez a2, .LBB87_16
 ; RV64ZVE32F-NEXT:  .LBB87_7: # %else11
 ; RV64ZVE32F-NEXT:    andi a2, a1, 32
 ; RV64ZVE32F-NEXT:    beqz a2, .LBB87_9
@@ -10337,14 +10270,35 @@ define <8 x float> @mgather_baseidx_v8i16_v8f32(ptr %base, <8 x i16> %idxs, <8 x
 ; RV64ZVE32F-NEXT:    andi a2, a1, 64
 ; RV64ZVE32F-NEXT:    vsetivli zero, 2, e16, mf2, ta, ma
 ; RV64ZVE32F-NEXT:    vslidedown.vi v8, v9, 2
-; RV64ZVE32F-NEXT:    bnez a2, .LBB87_15
-; RV64ZVE32F-NEXT:  # %bb.10: # %else17
+; RV64ZVE32F-NEXT:    beqz a2, .LBB87_11
+; RV64ZVE32F-NEXT:  # %bb.10: # %cond.load16
+; RV64ZVE32F-NEXT:    vmv.x.s a2, v8
+; RV64ZVE32F-NEXT:    slli a2, a2, 2
+; RV64ZVE32F-NEXT:    add a2, a0, a2
+; RV64ZVE32F-NEXT:    flw fa5, 0(a2)
+; RV64ZVE32F-NEXT:    vsetvli zero, zero, e32, m1, ta, ma
+; RV64ZVE32F-NEXT:    vfmv.s.f v12, fa5
+; RV64ZVE32F-NEXT:    vsetivli zero, 7, e32, m2, tu, ma
+; RV64ZVE32F-NEXT:    vslideup.vi v10, v12, 6
+; RV64ZVE32F-NEXT:  .LBB87_11: # %else17
 ; RV64ZVE32F-NEXT:    andi a1, a1, -128
-; RV64ZVE32F-NEXT:    bnez a1, .LBB87_16
-; RV64ZVE32F-NEXT:  .LBB87_11: # %else20
+; RV64ZVE32F-NEXT:    beqz a1, .LBB87_13
+; RV64ZVE32F-NEXT:  # %bb.12: # %cond.load19
+; RV64ZVE32F-NEXT:    vsetivli zero, 1, e16, mf2, ta, ma
+; RV64ZVE32F-NEXT:    vslidedown.vi v8, v8, 1
+; RV64ZVE32F-NEXT:    vmv.x.s a1, v8
+; RV64ZVE32F-NEXT:    slli a1, a1, 2
+; RV64ZVE32F-NEXT:    add a0, a0, a1
+; RV64ZVE32F-NEXT:    flw fa5, 0(a0)
+; RV64ZVE32F-NEXT:    vsetvli zero, zero, e32, m1, ta, ma
+; RV64ZVE32F-NEXT:    vfmv.s.f v8, fa5
+; RV64ZVE32F-NEXT:    vsetivli zero, 8, e32, m2, ta, ma
+; RV64ZVE32F-NEXT:    vslideup.vi v10, v8, 7
+; RV64ZVE32F-NEXT:  .LBB87_13: # %else20
+; RV64ZVE32F-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; RV64ZVE32F-NEXT:    vmv2r.v v8, v10
 ; RV64ZVE32F-NEXT:    ret
-; RV64ZVE32F-NEXT:  .LBB87_12: # %cond.load4
+; RV64ZVE32F-NEXT:  .LBB87_14: # %cond.load4
 ; RV64ZVE32F-NEXT:    vmv.x.s a2, v8
 ; RV64ZVE32F-NEXT:    slli a2, a2, 2
 ; RV64ZVE32F-NEXT:    add a2, a0, a2
@@ -10355,7 +10309,7 @@ define <8 x float> @mgather_baseidx_v8i16_v8f32(ptr %base, <8 x i16> %idxs, <8 x
 ; RV64ZVE32F-NEXT:    vslideup.vi v10, v12, 2
 ; RV64ZVE32F-NEXT:    andi a2, a1, 8
 ; RV64ZVE32F-NEXT:    beqz a2, .LBB87_6
-; RV64ZVE32F-NEXT:  .LBB87_13: # %cond.load7
+; RV64ZVE32F-NEXT:  .LBB87_15: # %cond.load7
 ; RV64ZVE32F-NEXT:    vsetivli zero, 1, e16, mf2, ta, ma
 ; RV64ZVE32F-NEXT:    vslidedown.vi v8, v8, 1
 ; RV64ZVE32F-NEXT:    vmv.x.s a2, v8
@@ -10368,7 +10322,7 @@ define <8 x float> @mgather_baseidx_v8i16_v8f32(ptr %base, <8 x i16> %idxs, <8 x
 ; RV64ZVE32F-NEXT:    vslideup.vi v10, v8, 3
 ; RV64ZVE32F-NEXT:    andi a2, a1, 16
 ; RV64ZVE32F-NEXT:    beqz a2, .LBB87_7
-; RV64ZVE32F-NEXT:  .LBB87_14: # %cond.load10
+; RV64ZVE32F-NEXT:  .LBB87_16: # %cond.load10
 ; RV64ZVE32F-NEXT:    vsetivli zero, 1, e16, m1, ta, ma
 ; RV64ZVE32F-NEXT:    vmv.x.s a2, v9
 ; RV64ZVE32F-NEXT:    slli a2, a2, 2
@@ -10381,30 +10335,6 @@ define <8 x float> @mgather_baseidx_v8i16_v8f32(ptr %base, <8 x i16> %idxs, <8 x
 ; RV64ZVE32F-NEXT:    andi a2, a1, 32
 ; RV64ZVE32F-NEXT:    bnez a2, .LBB87_8
 ; RV64ZVE32F-NEXT:    j .LBB87_9
-; RV64ZVE32F-NEXT:  .LBB87_15: # %cond.load16
-; RV64ZVE32F-NEXT:    vmv.x.s a2, v8
-; RV64ZVE32F-NEXT:    slli a2, a2, 2
-; RV64ZVE32F-NEXT:    add a2, a0, a2
-; RV64ZVE32F-NEXT:    flw fa5, 0(a2)
-; RV64ZVE32F-NEXT:    vsetvli zero, zero, e32, m1, ta, ma
-; RV64ZVE32F-NEXT:    vfmv.s.f v12, fa5
-; RV64ZVE32F-NEXT:    vsetivli zero, 7, e32, m2, tu, ma
-; RV64ZVE32F-NEXT:    vslideup.vi v10, v12, 6
-; RV64ZVE32F-NEXT:    andi a1, a1, -128
-; RV64ZVE32F-NEXT:    beqz a1, .LBB87_11
-; RV64ZVE32F-NEXT:  .LBB87_16: # %cond.load19
-; RV64ZVE32F-NEXT:    vsetivli zero, 1, e16, mf2, ta, ma
-; RV64ZVE32F-NEXT:    vslidedown.vi v8, v8, 1
-; RV64ZVE32F-NEXT:    vmv.x.s a1, v8
-; RV64ZVE32F-NEXT:    slli a1, a1, 2
-; RV64ZVE32F-NEXT:    add a0, a0, a1
-; RV64ZVE32F-NEXT:    flw fa5, 0(a0)
-; RV64ZVE32F-NEXT:    vsetvli zero, zero, e32, m1, ta, ma
-; RV64ZVE32F-NEXT:    vfmv.s.f v8, fa5
-; RV64ZVE32F-NEXT:    vsetivli zero, 8, e32, m2, ta, ma
-; RV64ZVE32F-NEXT:    vslideup.vi v10, v8, 7
-; RV64ZVE32F-NEXT:    vmv2r.v v8, v10
-; RV64ZVE32F-NEXT:    ret
   %ptrs = getelementptr inbounds float, ptr %base, <8 x i16> %idxs
   %v = call <8 x float> @llvm.masked.gather.v8f32.v8p0(<8 x ptr> %ptrs, i32 4, <8 x i1> %m, <8 x float> %passthru)
   ret <8 x float> %v
@@ -10464,13 +10394,13 @@ define <8 x float> @mgather_baseidx_sext_v8i16_v8f32(ptr %base, <8 x i16> %idxs,
 ; RV64ZVE32F-NEXT:    andi a2, a1, 4
 ; RV64ZVE32F-NEXT:    vsetivli zero, 2, e16, mf2, ta, ma
 ; RV64ZVE32F-NEXT:    vslidedown.vi v8, v8, 2
-; RV64ZVE32F-NEXT:    bnez a2, .LBB88_12
+; RV64ZVE32F-NEXT:    bnez a2, .LBB88_14
 ; RV64ZVE32F-NEXT:  # %bb.5: # %else5
 ; RV64ZVE32F-NEXT:    andi a2, a1, 8
-; RV64ZVE32F-NEXT:    bnez a2, .LBB88_13
+; RV64ZVE32F-NEXT:    bnez a2, .LBB88_15
 ; RV64ZVE32F-NEXT:  .LBB88_6: # %else8
 ; RV64ZVE32F-NEXT:    andi a2, a1, 16
-; RV64ZVE32F-NEXT:    bnez a2, .LBB88_14
+; RV64ZVE32F-NEXT:    bnez a2, .LBB88_16
 ; RV64ZVE32F-NEXT:  .LBB88_7: # %else11
 ; RV64ZVE32F-NEXT:    andi a2, a1, 32
 ; RV64ZVE32F-NEXT:    beqz a2, .LBB88_9
@@ -10489,14 +10419,35 @@ define <8 x float> @mgather_baseidx_sext_v8i16_v8f32(ptr %base, <8 x i16> %idxs,
 ; RV64ZVE32F-NEXT:    andi a2, a1, 64
 ; RV64ZVE32F-NEXT:    vsetivli zero, 2, e16, mf2, ta, ma
 ; RV64ZVE32F-NEXT:    vslidedown.vi v8, v9, 2
-; RV64ZVE32F-NEXT:    bnez a2, .LBB88_15
-; RV64ZVE32F-NEXT:  # %bb.10: # %else17
+; RV64ZVE32F-NEXT:    beqz a2, .LBB88_11
+; RV64ZVE32F-NEXT:  # %bb.10: # %cond.load16
+; RV64ZVE32F-NEXT:    vmv.x.s a2, v8
+; RV64ZVE32F-NEXT:    slli a2, a2, 2
+; RV64ZVE32F-NEXT:    add a2, a0, a2
+; RV64ZVE32F-NEXT:    flw fa5, 0(a2)
+; RV64ZVE32F-NEXT:    vsetvli zero, zero, e32, m1, ta, ma
+; RV64ZVE32F-NEXT:    vfmv.s.f v12, fa5
+; RV64ZVE32F-NEXT:    vsetivli zero, 7, e32, m2, tu, ma
+; RV64ZVE32F-NEXT:    vslideup.vi v10, v12, 6
+; RV64ZVE32F-NEXT:  .LBB88_11: # %else17
 ; RV64ZVE32F-NEXT:    andi a1, a1, -128
-; RV64ZVE32F-NEXT:    bnez a1, .LBB88_16
-; RV64ZVE32F-NEXT:  .LBB88_11: # %else20
+; RV64ZVE32F-NEXT:    beqz a1, .LBB88_13
+; RV64ZVE32F-NEXT:  # %bb.12: # %cond.load19
+; RV64ZVE32F-NEXT:    vsetivli zero, 1, e16, mf2, ta, ma
+; RV64ZVE32F-NEXT:    vslidedown.vi v8, v8, 1
+; RV64ZVE32F-NEXT:    vmv.x.s a1, v8
+; RV64ZVE32F-NEXT:    slli a1, a1, 2
+; RV64ZVE32F-NEXT:    add a0, a0, a1
+; RV64ZVE32F-NEXT:    flw fa5, 0(a0)
+; RV64ZVE32F-NEXT:    vsetvli zero, zero, e32, m1, ta, ma
+; RV64ZVE32F-NEXT:    vfmv.s.f v8, fa5
+; RV64ZVE32F-NEXT:    vsetivli zero, 8, e32, m2, ta, ma
+; RV64ZVE32F-NEXT:    vslideup.vi v10, v8, 7
+; RV64ZVE32F-NEXT:  .LBB88_13: # %else20
+; RV64ZVE32F-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; RV64ZVE32F-NEXT:    vmv2r.v v8, v10
 ; RV64ZVE32F-NEXT:    ret
-; RV64ZVE32F-NEXT:  .LBB88_12: # %cond.load4
+; RV64ZVE32F-NEXT:  .LBB88_14: # %cond.load4
 ; RV64ZVE32F-NEXT:    vmv.x.s a2, v8
 ; RV64ZVE32F-NEXT:    slli a2, a2, 2
 ; RV64ZVE32F-NEXT:    add a2, a0, a2
@@ -10507,7 +10458,7 @@ define <8 x float> @mgather_baseidx_sext_v8i16_v8f32(ptr %base, <8 x i16> %idxs,
 ; RV64ZVE32F-NEXT:    vslideup.vi v10, v12, 2
 ; RV64ZVE32F-NEXT:    andi a2, a1, 8
 ; RV64ZVE32F-NEXT:    beqz a2, .LBB88_6
-; RV64ZVE32F-NEXT:  .LBB88_13: # %cond.load7
+; RV64ZVE32F-NEXT:  .LBB88_15: # %cond.load7
 ; RV64ZVE32F-NEXT:    vsetivli zero, 1, e16, mf2, ta, ma
 ; RV64ZVE32F-NEXT:    vslidedown.vi v8, v8, 1
 ; RV64ZVE32F-NEXT:    vmv.x.s a2, v8
@@ -10520,7 +10471,7 @@ define <8 x float> @mgather_baseidx_sext_v8i16_v8f32(ptr %base, <8 x i16> %idxs,
 ; RV64ZVE32F-NEXT:    vslideup.vi v10, v8, 3
 ; RV64ZVE32F-NEXT:    andi a2, a1, 16
 ; RV64ZVE32F-NEXT:    beqz a2, .LBB88_7
-; RV64ZVE32F-NEXT:  .LBB88_14: # %cond.load10
+; RV64ZVE32F-NEXT:  .LBB88_16: # %cond.load10
 ; RV64ZVE32F-NEXT:    vsetivli zero, 1, e16, m1, ta, ma
 ; RV64ZVE32F-NEXT:    vmv.x.s a2, v9
 ; RV64ZVE32F-NEXT:    slli a2, a2, 2
@@ -10533,30 +10484,6 @@ define <8 x float> @mgather_baseidx_sext_v8i16_v8f32(ptr %base, <8 x i16> %idxs,
 ; RV64ZVE32F-NEXT:    andi a2, a1, 32
 ; RV64ZVE32F-NEXT:    bnez a2, .LBB88_8
 ; RV64ZVE32F-NEXT:    j .LBB88_9
-; RV64ZVE32F-NEXT:  .LBB88_15: # %cond.load16
-; RV64ZVE32F-NEXT:    vmv.x.s a2, v8
-; RV64ZVE32F-NEXT:    slli a2, a2, 2
-; RV64ZVE32F-NEXT:    add a2, a0, a2
-; RV64ZVE32F-NEXT:    flw fa5, 0(a2)
-; RV64ZVE32F-NEXT:    vsetvli zero, zero, e32, m1, ta, ma
-; RV64ZVE32F-NEXT:    vfmv.s.f v12, fa5
-; RV64ZVE32F-NEXT:    vsetivli zero, 7, e32, m2, tu, ma
-; RV64ZVE32F-NEXT:    vslideup.vi v10, v12, 6
-; RV64ZVE32F-NEXT:    andi a1, a1, -128
-; RV64ZVE32F-NEXT:    beqz a1, .LBB88_11
-; RV64ZVE32F-NEXT:  .LBB88_16: # %cond.load19
-; RV64ZVE32F-NEXT:    vsetivli zero, 1, e16, mf2, ta, ma
-; RV64ZVE32F-NEXT:    vslidedown.vi v8, v8, 1
-; RV64ZVE32F-NEXT:    vmv.x.s a1, v8
-; RV64ZVE32F-NEXT:    slli a1, a1, 2
-; RV64ZVE32F-NEXT:    add a0, a0, a1
-; RV64ZVE32F-NEXT:    flw fa5, 0(a0)
-; RV64ZVE32F-NEXT:    vsetvli zero, zero, e32, m1, ta, ma
-; RV64ZVE32F-NEXT:    vfmv.s.f v8, fa5
-; RV64ZVE32F-NEXT:    vsetivli zero, 8, e32, m2, ta, ma
-; RV64ZVE32F-NEXT:    vslideup.vi v10, v8, 7
-; RV64ZVE32F-NEXT:    vmv2r.v v8, v10
-; RV64ZVE32F-NEXT:    ret
   %eidxs = sext <8 x i16> %idxs to <8 x i32>
   %ptrs = getelementptr inbounds float, ptr %base, <8 x i32> %eidxs
   %v = call <8 x float> @llvm.masked.gather.v8f32.v8p0(<8 x ptr> %ptrs, i32 4, <8 x i1> %m, <8 x float> %passthru)
@@ -10620,13 +10547,13 @@ define <8 x float> @mgather_baseidx_zext_v8i16_v8f32(ptr %base, <8 x i16> %idxs,
 ; RV64ZVE32F-NEXT:    andi a3, a2, 4
 ; RV64ZVE32F-NEXT:    vsetivli zero, 2, e16, mf2, ta, ma
 ; RV64ZVE32F-NEXT:    vslidedown.vi v8, v8, 2
-; RV64ZVE32F-NEXT:    bnez a3, .LBB89_12
+; RV64ZVE32F-NEXT:    bnez a3, .LBB89_14
 ; RV64ZVE32F-NEXT:  # %bb.5: # %else5
 ; RV64ZVE32F-NEXT:    andi a3, a2, 8
-; RV64ZVE32F-NEXT:    bnez a3, .LBB89_13
+; RV64ZVE32F-NEXT:    bnez a3, .LBB89_15
 ; RV64ZVE32F-NEXT:  .LBB89_6: # %else8
 ; RV64ZVE32F-NEXT:    andi a3, a2, 16
-; RV64ZVE32F-NEXT:    bnez a3, .LBB89_14
+; RV64ZVE32F-NEXT:    bnez a3, .LBB89_16
 ; RV64ZVE32F-NEXT:  .LBB89_7: # %else11
 ; RV64ZVE32F-NEXT:    andi a3, a2, 32
 ; RV64ZVE32F-NEXT:    beqz a3, .LBB89_9
@@ -10646,14 +10573,37 @@ define <8 x float> @mgather_baseidx_zext_v8i16_v8f32(ptr %base, <8 x i16> %idxs,
 ; RV64ZVE32F-NEXT:    andi a3, a2, 64
 ; RV64ZVE32F-NEXT:    vsetivli zero, 2, e16, mf2, ta, ma
 ; RV64ZVE32F-NEXT:    vslidedown.vi v8, v9, 2
-; RV64ZVE32F-NEXT:    bnez a3, .LBB89_15
-; RV64ZVE32F-NEXT:  # %bb.10: # %else17
+; RV64ZVE32F-NEXT:    beqz a3, .LBB89_11
+; RV64ZVE32F-NEXT:  # %bb.10: # %cond.load16
+; RV64ZVE32F-NEXT:    vmv.x.s a3, v8
+; RV64ZVE32F-NEXT:    and a3, a3, a1
+; RV64ZVE32F-NEXT:    slli a3, a3, 2
+; RV64ZVE32F-NEXT:    add a3, a0, a3
+; RV64ZVE32F-NEXT:    flw fa5, 0(a3)
+; RV64ZVE32F-NEXT:    vsetvli zero, zero, e32, m1, ta, ma
+; RV64ZVE32F-NEXT:    vfmv.s.f v12, fa5
+; RV64ZVE32F-NEXT:    vsetivli zero, 7, e32, m2, tu, ma
+; RV64ZVE32F-NEXT:    vslideup.vi v10, v12, 6
+; RV64ZVE32F-NEXT:  .LBB89_11: # %else17
 ; RV64ZVE32F-NEXT:    andi a2, a2, -128
-; RV64ZVE32F-NEXT:    bnez a2, .LBB89_16
-; RV64ZVE32F-NEXT:  .LBB89_11: # %else20
+; RV64ZVE32F-NEXT:    beqz a2, .LBB89_13
+; RV64ZVE32F-NEXT:  # %bb.12: # %cond.load19
+; RV64ZVE32F-NEXT:    vsetivli zero, 1, e16, mf2, ta, ma
+; RV64ZVE32F-NEXT:    vslidedown.vi v8, v8, 1
+; RV64ZVE32F-NEXT:    vmv.x.s a2, v8
+; RV64ZVE32F-NEXT:    and a1, a2, a1
+; RV64ZVE32F-NEXT:    slli a1, a1, 2
+; RV64ZVE32F-NEXT:    add a0, a0, a1
+; RV64ZVE32F-NEXT:    flw fa5, 0(a0)
+; RV64ZVE32F-NEXT:    vsetvli zero, zero, e32, m1, ta, ma
+; RV64ZVE32F-NEXT:    vfmv.s.f v8, fa5
+; RV64ZVE32F-NEXT:    vsetivli zero, 8, e32, m2, ta, ma
+; RV64ZVE32F-NEXT:    vslideup.vi v10, v8, 7
+; RV64ZVE32F-NEXT:  .LBB89_13: # %else20
+; RV64ZVE32F-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; RV64ZVE32F-NEXT:    vmv2r.v v8, v10
 ; RV64ZVE32F-NEXT:    ret
-; RV64ZVE32F-NEXT:  .LBB89_12: # %cond.load4
+; RV64ZVE32F-NEXT:  .LBB89_14: # %cond.load4
 ; RV64ZVE32F-NEXT:    vmv.x.s a3, v8
 ; RV64ZVE32F-NEXT:    and a3, a3, a1
 ; RV64ZVE32F-NEXT:    slli a3, a3, 2
@@ -10665,7 +10615,7 @@ define <8 x float> @mgather_baseidx_zext_v8i16_v8f32(ptr %base, <8 x i16> %idxs,
 ; RV64ZVE32F-NEXT:    vslideup.vi v10, v12, 2
 ; RV64ZVE32F-NEXT:    andi a3, a2, 8
 ; RV64ZVE32F-NEXT:    beqz a3, .LBB89_6
-; RV64ZVE32F-NEXT:  .LBB89_13: # %cond.load7
+; RV64ZVE32F-NEXT:  .LBB89_15: # %cond.load7
 ; RV64ZVE32F-NEXT:    vsetivli zero, 1, e16, mf2, ta, ma
 ; RV64ZVE32F-NEXT:    vslidedown.vi v8, v8, 1
 ; RV64ZVE32F-NEXT:    vmv.x.s a3, v8
@@ -10679,7 +10629,7 @@ define <8 x float> @mgather_baseidx_zext_v8i16_v8f32(ptr %base, <8 x i16> %idxs,
 ; RV64ZVE32F-NEXT:    vslideup.vi v10, v8, 3
 ; RV64ZVE32F-NEXT:    andi a3, a2, 16
 ; RV64ZVE32F-NEXT:    beqz a3, .LBB89_7
-; RV64ZVE32F-NEXT:  .LBB89_14: # %cond.load10
+; RV64ZVE32F-NEXT:  .LBB89_16: # %cond.load10
 ; RV64ZVE32F-NEXT:    vsetivli zero, 1, e16, m1, ta, ma
 ; RV64ZVE32F-NEXT:    vmv.x.s a3, v9
 ; RV64ZVE32F-NEXT:    and a3, a3, a1
@@ -10693,32 +10643,6 @@ define <8 x float> @mgather_baseidx_zext_v8i16_v8f32(ptr %base, <8 x i16> %idxs,
 ; RV64ZVE32F-NEXT:    andi a3, a2, 32
 ; RV64ZVE32F-NEXT:    bnez a3, .LBB89_8
 ; RV64ZVE32F-NEXT:    j .LBB89_9
-; RV64ZVE32F-NEXT:  .LBB89_15: # %cond.load16
-; RV64ZVE32F-NEXT:    vmv.x.s a3, v8
-; RV64ZVE32F-NEXT:    and a3, a3, a1
-; RV64ZVE32F-NEXT:    slli a3, a3, 2
-; RV64ZVE32F-NEXT:    add a3, a0, a3
-; RV64ZVE32F-NEXT:    flw fa5, 0(a3)
-; RV64ZVE32F-NEXT:    vsetvli zero, zero, e32, m1, ta, ma
-; RV64ZVE32F-NEXT:    vfmv.s.f v12, fa5
-; RV64ZVE32F-NEXT:    vsetivli zero, 7, e32, m2, tu, ma
-; RV64ZVE32F-NEXT:    vslideup.vi v10, v12, 6
-; RV64ZVE32F-NEXT:    andi a2, a2, -128
-; RV64ZVE32F-NEXT:    beqz a2, .LBB89_11
-; RV64ZVE32F-NEXT:  .LBB89_16: # %cond.load19
-; RV64ZVE32F-NEXT:    vsetivli zero, 1, e16, mf2, ta, ma
-; RV64ZVE32F-NEXT:    vslidedown.vi v8, v8, 1
-; RV64ZVE32F-NEXT:    vmv.x.s a2, v8
-; RV64ZVE32F-NEXT:    and a1, a2, a1
-; RV64ZVE32F-NEXT:    slli a1, a1, 2
-; RV64ZVE32F-NEXT:    add a0, a0, a1
-; RV64ZVE32F-NEXT:    flw fa5, 0(a0)
-; RV64ZVE32F-NEXT:    vsetvli zero, zero, e32, m1, ta, ma
-; RV64ZVE32F-NEXT:    vfmv.s.f v8, fa5
-; RV64ZVE32F-NEXT:    vsetivli zero, 8, e32, m2, ta, ma
-; RV64ZVE32F-NEXT:    vslideup.vi v10, v8, 7
-; RV64ZVE32F-NEXT:    vmv2r.v v8, v10
-; RV64ZVE32F-NEXT:    ret
   %eidxs = zext <8 x i16> %idxs to <8 x i32>
   %ptrs = getelementptr inbounds float, ptr %base, <8 x i32> %eidxs
   %v = call <8 x float> @llvm.masked.gather.v8f32.v8p0(<8 x ptr> %ptrs, i32 4, <8 x i1> %m, <8 x float> %passthru)
@@ -10775,13 +10699,13 @@ define <8 x float> @mgather_baseidx_v8f32(ptr %base, <8 x i32> %idxs, <8 x i1> %
 ; RV64ZVE32F-NEXT:    andi a2, a1, 4
 ; RV64ZVE32F-NEXT:    vsetivli zero, 2, e32, m1, ta, ma
 ; RV64ZVE32F-NEXT:    vslidedown.vi v8, v8, 2
-; RV64ZVE32F-NEXT:    bnez a2, .LBB90_12
+; RV64ZVE32F-NEXT:    bnez a2, .LBB90_14
 ; RV64ZVE32F-NEXT:  # %bb.5: # %else5
 ; RV64ZVE32F-NEXT:    andi a2, a1, 8
-; RV64ZVE32F-NEXT:    bnez a2, .LBB90_13
+; RV64ZVE32F-NEXT:    bnez a2, .LBB90_15
 ; RV64ZVE32F-NEXT:  .LBB90_6: # %else8
 ; RV64ZVE32F-NEXT:    andi a2, a1, 16
-; RV64ZVE32F-NEXT:    bnez a2, .LBB90_14
+; RV64ZVE32F-NEXT:    bnez a2, .LBB90_16
 ; RV64ZVE32F-NEXT:  .LBB90_7: # %else11
 ; RV64ZVE32F-NEXT:    andi a2, a1, 32
 ; RV64ZVE32F-NEXT:    beqz a2, .LBB90_9
@@ -10799,14 +10723,33 @@ define <8 x float> @mgather_baseidx_v8f32(ptr %base, <8 x i32> %idxs, <8 x i1> %
 ; RV64ZVE32F-NEXT:    andi a2, a1, 64
 ; RV64ZVE32F-NEXT:    vsetivli zero, 2, e32, m1, ta, ma
 ; RV64ZVE32F-NEXT:    vslidedown.vi v8, v12, 2
-; RV64ZVE32F-NEXT:    bnez a2, .LBB90_15
-; RV64ZVE32F-NEXT:  # %bb.10: # %else17
+; RV64ZVE32F-NEXT:    beqz a2, .LBB90_11
+; RV64ZVE32F-NEXT:  # %bb.10: # %cond.load16
+; RV64ZVE32F-NEXT:    vmv.x.s a2, v8
+; RV64ZVE32F-NEXT:    slli a2, a2, 2
+; RV64ZVE32F-NEXT:    add a2, a0, a2
+; RV64ZVE32F-NEXT:    flw fa5, 0(a2)
+; RV64ZVE32F-NEXT:    vfmv.s.f v12, fa5
+; RV64ZVE32F-NEXT:    vsetivli zero, 7, e32, m2, tu, ma
+; RV64ZVE32F-NEXT:    vslideup.vi v10, v12, 6
+; RV64ZVE32F-NEXT:  .LBB90_11: # %else17
 ; RV64ZVE32F-NEXT:    andi a1, a1, -128
-; RV64ZVE32F-NEXT:    bnez a1, .LBB90_16
-; RV64ZVE32F-NEXT:  .LBB90_11: # %else20
+; RV64ZVE32F-NEXT:    beqz a1, .LBB90_13
+; RV64ZVE32F-NEXT:  # %bb.12: # %cond.load19
+; RV64ZVE32F-NEXT:    vsetivli zero, 1, e32, m1, ta, ma
+; RV64ZVE32F-NEXT:    vslidedown.vi v8, v8, 1
+; RV64ZVE32F-NEXT:    vmv.x.s a1, v8
+; RV64ZVE32F-NEXT:    slli a1, a1, 2
+; RV64ZVE32F-NEXT:    add a0, a0, a1
+; RV64ZVE32F-NEXT:    flw fa5, 0(a0)
+; RV64ZVE32F-NEXT:    vfmv.s.f v8, fa5
+; RV64ZVE32F-NEXT:    vsetivli zero, 8, e32, m2, ta, ma
+; RV64ZVE32F-NEXT:    vslideup.vi v10, v8, 7
+; RV64ZVE32F-NEXT:  .LBB90_13: # %else20
+; RV64ZVE32F-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; RV64ZVE32F-NEXT:    vmv2r.v v8, v10
 ; RV64ZVE32F-NEXT:    ret
-; RV64ZVE32F-NEXT:  .LBB90_12: # %cond.load4
+; RV64ZVE32F-NEXT:  .LBB90_14: # %cond.load4
 ; RV64ZVE32F-NEXT:    vmv.x.s a2, v8
 ; RV64ZVE32F-NEXT:    slli a2, a2, 2
 ; RV64ZVE32F-NEXT:    add a2, a0, a2
@@ -10816,7 +10759,7 @@ define <8 x float> @mgather_baseidx_v8f32(ptr %base, <8 x i32> %idxs, <8 x i1> %
 ; RV64ZVE32F-NEXT:    vslideup.vi v10, v9, 2
 ; RV64ZVE32F-NEXT:    andi a2, a1, 8
 ; RV64ZVE32F-NEXT:    beqz a2, .LBB90_6
-; RV64ZVE32F-NEXT:  .LBB90_13: # %cond.load7
+; RV64ZVE32F-NEXT:  .LBB90_15: # %cond.load7
 ; RV64ZVE32F-NEXT:    vsetivli zero, 4, e32, m1, tu, ma
 ; RV64ZVE32F-NEXT:    vslidedown.vi v8, v8, 1
 ; RV64ZVE32F-NEXT:    vmv.x.s a2, v8
@@ -10827,7 +10770,7 @@ define <8 x float> @mgather_baseidx_v8f32(ptr %base, <8 x i32> %idxs, <8 x i1> %
 ; RV64ZVE32F-NEXT:    vslideup.vi v10, v8, 3
 ; RV64ZVE32F-NEXT:    andi a2, a1, 16
 ; RV64ZVE32F-NEXT:    beqz a2, .LBB90_7
-; RV64ZVE32F-NEXT:  .LBB90_14: # %cond.load10
+; RV64ZVE32F-NEXT:  .LBB90_16: # %cond.load10
 ; RV64ZVE32F-NEXT:    vsetivli zero, 5, e32, m2, tu, ma
 ; RV64ZVE32F-NEXT:    vmv.x.s a2, v12
 ; RV64ZVE32F-NEXT:    slli a2, a2, 2
@@ -10838,28 +10781,6 @@ define <8 x float> @mgather_baseidx_v8f32(ptr %base, <8 x i32> %idxs, <8 x i1> %
 ; RV64ZVE32F-NEXT:    andi a2, a1, 32
 ; RV64ZVE32F-NEXT:    bnez a2, .LBB90_8
 ; RV64ZVE32F-NEXT:    j .LBB90_9
-; RV64ZVE32F-NEXT:  .LBB90_15: # %cond.load16
-; RV64ZVE32F-NEXT:    vmv.x.s a2, v8
-; RV64ZVE32F-NEXT:    slli a2, a2, 2
-; RV64ZVE32F-NEXT:    add a2, a0, a2
-; RV64ZVE32F-NEXT:    flw fa5, 0(a2)
-; RV64ZVE32F-NEXT:    vfmv.s.f v12, fa5
-; RV64ZVE32F-NEXT:    vsetivli zero, 7, e32, m2, tu, ma
-; RV64ZVE32F-NEXT:    vslideup.vi v10, v12, 6
-; RV64ZVE32F-NEXT:    andi a1, a1, -128
-; RV64ZVE32F-NEXT:    beqz a1, .LBB90_11
-; RV64ZVE32F-NEXT:  .LBB90_16: # %cond.load19
-; RV64ZVE32F-NEXT:    vsetivli zero, 1, e32, m1, ta, ma
-; RV64ZVE32F-NEXT:    vslidedown.vi v8, v8, 1
-; RV64ZVE32F-NEXT:    vmv.x.s a1, v8
-; RV64ZVE32F-NEXT:    slli a1, a1, 2
-; RV64ZVE32F-NEXT:    add a0, a0, a1
-; RV64ZVE32F-NEXT:    flw fa5, 0(a0)
-; RV64ZVE32F-NEXT:    vfmv.s.f v8, fa5
-; RV64ZVE32F-NEXT:    vsetivli zero, 8, e32, m2, ta, ma
-; RV64ZVE32F-NEXT:    vslideup.vi v10, v8, 7
-; RV64ZVE32F-NEXT:    vmv2r.v v8, v10
-; RV64ZVE32F-NEXT:    ret
   %ptrs = getelementptr inbounds float, ptr %base, <8 x i32> %idxs
   %v = call <8 x float> @llvm.masked.gather.v8f32.v8p0(<8 x ptr> %ptrs, i32 4, <8 x i1> %m, <8 x float> %passthru)
   ret <8 x float> %v
@@ -11135,11 +11056,13 @@ define <4 x double> @mgather_truemask_v4f64(<4 x ptr> %ptrs, <4 x double> %passt
 define <4 x double> @mgather_falsemask_v4f64(<4 x ptr> %ptrs, <4 x double> %passthru) {
 ; RV32V-LABEL: mgather_falsemask_v4f64:
 ; RV32V:       # %bb.0:
+; RV32V-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; RV32V-NEXT:    vmv2r.v v8, v10
 ; RV32V-NEXT:    ret
 ;
 ; RV64V-LABEL: mgather_falsemask_v4f64:
 ; RV64V:       # %bb.0:
+; RV64V-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; RV64V-NEXT:    vmv2r.v v8, v10
 ; RV64V-NEXT:    ret
 ;
@@ -13700,6 +13623,7 @@ define <16 x i8> @mgather_baseidx_v16i8(ptr %base, <16 x i8> %idxs, <16 x i1> %m
 ; RV64ZVE32F-NEXT:    vsetivli zero, 16, e8, m1, ta, ma
 ; RV64ZVE32F-NEXT:    vslideup.vi v9, v8, 15
 ; RV64ZVE32F-NEXT:  .LBB107_24: # %else44
+; RV64ZVE32F-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; RV64ZVE32F-NEXT:    vmv1r.v v8, v9
 ; RV64ZVE32F-NEXT:    ret
 ; RV64ZVE32F-NEXT:  .LBB107_25: # %cond.load4
@@ -14086,6 +14010,7 @@ define <32 x i8> @mgather_baseidx_v32i8(ptr %base, <32 x i8> %idxs, <32 x i1> %m
 ; RV64ZVE32F-NEXT:    vsetvli zero, a1, e8, m2, ta, ma
 ; RV64ZVE32F-NEXT:    vslideup.vi v10, v8, 31
 ; RV64ZVE32F-NEXT:  .LBB108_48: # %else92
+; RV64ZVE32F-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; RV64ZVE32F-NEXT:    vmv2r.v v8, v10
 ; RV64ZVE32F-NEXT:    ret
 ; RV64ZVE32F-NEXT:  .LBB108_49: # %cond.load4
diff --git a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-masked-load-int.ll b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-masked-load-int.ll
index e0cf39c75da24..69903d77084bf 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-masked-load-int.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-masked-load-int.ll
@@ -318,6 +318,7 @@ define <128 x i16> @masked_load_v128i16(ptr %a, <128 x i1> %mask) {
 define <256 x i8> @masked_load_v256i8(ptr %a, <256 x i1> %mask) {
 ; CHECK-LABEL: masked_load_v256i8:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-NEXT:    vmv1r.v v16, v8
 ; CHECK-NEXT:    li a1, 128
 ; CHECK-NEXT:    vsetvli zero, a1, e8, m8, ta, ma
diff --git a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-nearbyint-vp.ll b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-nearbyint-vp.ll
index 46c2033d28b38..80a9143d1ad8b 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-nearbyint-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-nearbyint-vp.ll
@@ -135,10 +135,10 @@ declare <16 x half> @llvm.vp.nearbyint.v16f16(<16 x half>, <16 x i1>, i32)
 define <16 x half> @vp_nearbyint_v16f16(<16 x half> %va, <16 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vp_nearbyint_v16f16:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v10, v0
-; CHECK-NEXT:    lui a1, %hi(.LCPI6_0)
-; CHECK-NEXT:    flh fa5, %lo(.LCPI6_0)(a1)
 ; CHECK-NEXT:    vsetvli zero, a0, e16, m2, ta, ma
+; CHECK-NEXT:    vmv1r.v v10, v0
+; CHECK-NEXT:    lui a0, %hi(.LCPI6_0)
+; CHECK-NEXT:    flh fa5, %lo(.LCPI6_0)(a0)
 ; CHECK-NEXT:    vfabs.v v12, v8, v0.t
 ; CHECK-NEXT:    vsetvli zero, zero, e16, m2, ta, mu
 ; CHECK-NEXT:    vmflt.vf v10, v12, fa5, v0.t
@@ -263,8 +263,8 @@ declare <8 x float> @llvm.vp.nearbyint.v8f32(<8 x float>, <8 x i1>, i32)
 define <8 x float> @vp_nearbyint_v8f32(<8 x float> %va, <8 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vp_nearbyint_v8f32:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e32, m2, ta, ma
+; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vfabs.v v12, v8, v0.t
 ; CHECK-NEXT:    lui a0, 307200
 ; CHECK-NEXT:    fmv.w.x fa5, a0
@@ -307,8 +307,8 @@ declare <16 x float> @llvm.vp.nearbyint.v16f32(<16 x float>, <16 x i1>, i32)
 define <16 x float> @vp_nearbyint_v16f32(<16 x float> %va, <16 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vp_nearbyint_v16f32:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v12, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e32, m4, ta, ma
+; CHECK-NEXT:    vmv1r.v v12, v0
 ; CHECK-NEXT:    vfabs.v v16, v8, v0.t
 ; CHECK-NEXT:    lui a0, 307200
 ; CHECK-NEXT:    fmv.w.x fa5, a0
@@ -393,10 +393,10 @@ declare <4 x double> @llvm.vp.nearbyint.v4f64(<4 x double>, <4 x i1>, i32)
 define <4 x double> @vp_nearbyint_v4f64(<4 x double> %va, <4 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vp_nearbyint_v4f64:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v10, v0
-; CHECK-NEXT:    lui a1, %hi(.LCPI18_0)
-; CHECK-NEXT:    fld fa5, %lo(.LCPI18_0)(a1)
 ; CHECK-NEXT:    vsetvli zero, a0, e64, m2, ta, ma
+; CHECK-NEXT:    vmv1r.v v10, v0
+; CHECK-NEXT:    lui a0, %hi(.LCPI18_0)
+; CHECK-NEXT:    fld fa5, %lo(.LCPI18_0)(a0)
 ; CHECK-NEXT:    vfabs.v v12, v8, v0.t
 ; CHECK-NEXT:    vsetvli zero, zero, e64, m2, ta, mu
 ; CHECK-NEXT:    vmflt.vf v10, v12, fa5, v0.t
@@ -437,10 +437,10 @@ declare <8 x double> @llvm.vp.nearbyint.v8f64(<8 x double>, <8 x i1>, i32)
 define <8 x double> @vp_nearbyint_v8f64(<8 x double> %va, <8 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vp_nearbyint_v8f64:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v12, v0
-; CHECK-NEXT:    lui a1, %hi(.LCPI20_0)
-; CHECK-NEXT:    fld fa5, %lo(.LCPI20_0)(a1)
 ; CHECK-NEXT:    vsetvli zero, a0, e64, m4, ta, ma
+; CHECK-NEXT:    vmv1r.v v12, v0
+; CHECK-NEXT:    lui a0, %hi(.LCPI20_0)
+; CHECK-NEXT:    fld fa5, %lo(.LCPI20_0)(a0)
 ; CHECK-NEXT:    vfabs.v v16, v8, v0.t
 ; CHECK-NEXT:    vsetvli zero, zero, e64, m4, ta, mu
 ; CHECK-NEXT:    vmflt.vf v12, v16, fa5, v0.t
@@ -481,10 +481,10 @@ declare <15 x double> @llvm.vp.nearbyint.v15f64(<15 x double>, <15 x i1>, i32)
 define <15 x double> @vp_nearbyint_v15f64(<15 x double> %va, <15 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vp_nearbyint_v15f64:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v16, v0
-; CHECK-NEXT:    lui a1, %hi(.LCPI22_0)
-; CHECK-NEXT:    fld fa5, %lo(.LCPI22_0)(a1)
 ; CHECK-NEXT:    vsetvli zero, a0, e64, m8, ta, ma
+; CHECK-NEXT:    vmv1r.v v16, v0
+; CHECK-NEXT:    lui a0, %hi(.LCPI22_0)
+; CHECK-NEXT:    fld fa5, %lo(.LCPI22_0)(a0)
 ; CHECK-NEXT:    vfabs.v v24, v8, v0.t
 ; CHECK-NEXT:    vsetvli zero, zero, e64, m8, ta, mu
 ; CHECK-NEXT:    vmflt.vf v16, v24, fa5, v0.t
@@ -525,10 +525,10 @@ declare <16 x double> @llvm.vp.nearbyint.v16f64(<16 x double>, <16 x i1>, i32)
 define <16 x double> @vp_nearbyint_v16f64(<16 x double> %va, <16 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vp_nearbyint_v16f64:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v16, v0
-; CHECK-NEXT:    lui a1, %hi(.LCPI24_0)
-; CHECK-NEXT:    fld fa5, %lo(.LCPI24_0)(a1)
 ; CHECK-NEXT:    vsetvli zero, a0, e64, m8, ta, ma
+; CHECK-NEXT:    vmv1r.v v16, v0
+; CHECK-NEXT:    lui a0, %hi(.LCPI24_0)
+; CHECK-NEXT:    fld fa5, %lo(.LCPI24_0)(a0)
 ; CHECK-NEXT:    vfabs.v v24, v8, v0.t
 ; CHECK-NEXT:    vsetvli zero, zero, e64, m8, ta, mu
 ; CHECK-NEXT:    vmflt.vf v16, v24, fa5, v0.t
@@ -569,9 +569,9 @@ declare <32 x double> @llvm.vp.nearbyint.v32f64(<32 x double>, <32 x i1>, i32)
 define <32 x double> @vp_nearbyint_v32f64(<32 x double> %va, <32 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vp_nearbyint_v32f64:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetivli zero, 2, e8, mf4, ta, ma
 ; CHECK-NEXT:    vmv1r.v v6, v0
 ; CHECK-NEXT:    li a2, 16
-; CHECK-NEXT:    vsetivli zero, 2, e8, mf4, ta, ma
 ; CHECK-NEXT:    vslidedown.vi v7, v0, 2
 ; CHECK-NEXT:    mv a1, a0
 ; CHECK-NEXT:    bltu a0, a2, .LBB26_2
diff --git a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-reduction-mask-vp.ll b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-reduction-mask-vp.ll
index ad358d7320240..276f6b077931b 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-reduction-mask-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-reduction-mask-vp.ll
@@ -23,9 +23,9 @@ declare i1 @llvm.vp.reduce.or.v1i1(i1, <1 x i1>, <1 x i1>, i32)
 define zeroext i1 @vpreduce_or_v1i1(i1 zeroext %s, <1 x i1> %v, <1 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vpreduce_or_v1i1:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a1, e8, mf8, ta, ma
 ; CHECK-NEXT:    vmv1r.v v9, v0
 ; CHECK-NEXT:    vmv1r.v v0, v8
-; CHECK-NEXT:    vsetvli zero, a1, e8, mf8, ta, ma
 ; CHECK-NEXT:    vcpop.m a1, v9, v0.t
 ; CHECK-NEXT:    snez a1, a1
 ; CHECK-NEXT:    or a0, a1, a0
@@ -39,9 +39,9 @@ declare i1 @llvm.vp.reduce.xor.v1i1(i1, <1 x i1>, <1 x i1>, i32)
 define zeroext i1 @vpreduce_xor_v1i1(i1 zeroext %s, <1 x i1> %v, <1 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vpreduce_xor_v1i1:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a1, e8, mf8, ta, ma
 ; CHECK-NEXT:    vmv1r.v v9, v0
 ; CHECK-NEXT:    vmv1r.v v0, v8
-; CHECK-NEXT:    vsetvli zero, a1, e8, mf8, ta, ma
 ; CHECK-NEXT:    vcpop.m a1, v9, v0.t
 ; CHECK-NEXT:    andi a1, a1, 1
 ; CHECK-NEXT:    xor a0, a1, a0
@@ -71,9 +71,9 @@ declare i1 @llvm.vp.reduce.or.v2i1(i1, <2 x i1>, <2 x i1>, i32)
 define zeroext i1 @vpreduce_or_v2i1(i1 zeroext %s, <2 x i1> %v, <2 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vpreduce_or_v2i1:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a1, e8, mf8, ta, ma
 ; CHECK-NEXT:    vmv1r.v v9, v0
 ; CHECK-NEXT:    vmv1r.v v0, v8
-; CHECK-NEXT:    vsetvli zero, a1, e8, mf8, ta, ma
 ; CHECK-NEXT:    vcpop.m a1, v9, v0.t
 ; CHECK-NEXT:    snez a1, a1
 ; CHECK-NEXT:    or a0, a1, a0
@@ -87,9 +87,9 @@ declare i1 @llvm.vp.reduce.xor.v2i1(i1, <2 x i1>, <2 x i1>, i32)
 define zeroext i1 @vpreduce_xor_v2i1(i1 zeroext %s, <2 x i1> %v, <2 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vpreduce_xor_v2i1:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a1, e8, mf8, ta, ma
 ; CHECK-NEXT:    vmv1r.v v9, v0
 ; CHECK-NEXT:    vmv1r.v v0, v8
-; CHECK-NEXT:    vsetvli zero, a1, e8, mf8, ta, ma
 ; CHECK-NEXT:    vcpop.m a1, v9, v0.t
 ; CHECK-NEXT:    andi a1, a1, 1
 ; CHECK-NEXT:    xor a0, a1, a0
@@ -119,9 +119,9 @@ declare i1 @llvm.vp.reduce.or.v4i1(i1, <4 x i1>, <4 x i1>, i32)
 define zeroext i1 @vpreduce_or_v4i1(i1 zeroext %s, <4 x i1> %v, <4 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vpreduce_or_v4i1:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a1, e8, mf4, ta, ma
 ; CHECK-NEXT:    vmv1r.v v9, v0
 ; CHECK-NEXT:    vmv1r.v v0, v8
-; CHECK-NEXT:    vsetvli zero, a1, e8, mf4, ta, ma
 ; CHECK-NEXT:    vcpop.m a1, v9, v0.t
 ; CHECK-NEXT:    snez a1, a1
 ; CHECK-NEXT:    or a0, a1, a0
@@ -135,9 +135,9 @@ declare i1 @llvm.vp.reduce.xor.v4i1(i1, <4 x i1>, <4 x i1>, i32)
 define zeroext i1 @vpreduce_xor_v4i1(i1 zeroext %s, <4 x i1> %v, <4 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vpreduce_xor_v4i1:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a1, e8, mf4, ta, ma
 ; CHECK-NEXT:    vmv1r.v v9, v0
 ; CHECK-NEXT:    vmv1r.v v0, v8
-; CHECK-NEXT:    vsetvli zero, a1, e8, mf4, ta, ma
 ; CHECK-NEXT:    vcpop.m a1, v9, v0.t
 ; CHECK-NEXT:    andi a1, a1, 1
 ; CHECK-NEXT:    xor a0, a1, a0
@@ -167,9 +167,9 @@ declare i1 @llvm.vp.reduce.or.v8i1(i1, <8 x i1>, <8 x i1>, i32)
 define zeroext i1 @vpreduce_or_v8i1(i1 zeroext %s, <8 x i1> %v, <8 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vpreduce_or_v8i1:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a1, e8, mf2, ta, ma
 ; CHECK-NEXT:    vmv1r.v v9, v0
 ; CHECK-NEXT:    vmv1r.v v0, v8
-; CHECK-NEXT:    vsetvli zero, a1, e8, mf2, ta, ma
 ; CHECK-NEXT:    vcpop.m a1, v9, v0.t
 ; CHECK-NEXT:    snez a1, a1
 ; CHECK-NEXT:    or a0, a1, a0
@@ -183,9 +183,9 @@ declare i1 @llvm.vp.reduce.xor.v8i1(i1, <8 x i1>, <8 x i1>, i32)
 define zeroext i1 @vpreduce_xor_v8i1(i1 zeroext %s, <8 x i1> %v, <8 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vpreduce_xor_v8i1:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a1, e8, mf2, ta, ma
 ; CHECK-NEXT:    vmv1r.v v9, v0
 ; CHECK-NEXT:    vmv1r.v v0, v8
-; CHECK-NEXT:    vsetvli zero, a1, e8, mf2, ta, ma
 ; CHECK-NEXT:    vcpop.m a1, v9, v0.t
 ; CHECK-NEXT:    andi a1, a1, 1
 ; CHECK-NEXT:    xor a0, a1, a0
@@ -231,6 +231,7 @@ declare i1 @llvm.vp.reduce.and.v256i1(i1, <256 x i1>, <256 x i1>, i32)
 define zeroext i1 @vpreduce_and_v256i1(i1 zeroext %s, <256 x i1> %v, <256 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vpreduce_and_v256i1:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-NEXT:    vmv1r.v v11, v9
 ; CHECK-NEXT:    vmv1r.v v9, v0
 ; CHECK-NEXT:    li a3, 128
@@ -265,9 +266,9 @@ declare i1 @llvm.vp.reduce.or.v16i1(i1, <16 x i1>, <16 x i1>, i32)
 define zeroext i1 @vpreduce_or_v16i1(i1 zeroext %s, <16 x i1> %v, <16 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vpreduce_or_v16i1:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a1, e8, m1, ta, ma
 ; CHECK-NEXT:    vmv1r.v v9, v0
 ; CHECK-NEXT:    vmv1r.v v0, v8
-; CHECK-NEXT:    vsetvli zero, a1, e8, m1, ta, ma
 ; CHECK-NEXT:    vcpop.m a1, v9, v0.t
 ; CHECK-NEXT:    snez a1, a1
 ; CHECK-NEXT:    or a0, a1, a0
@@ -281,9 +282,9 @@ declare i1 @llvm.vp.reduce.xor.v16i1(i1, <16 x i1>, <16 x i1>, i32)
 define zeroext i1 @vpreduce_xor_v16i1(i1 zeroext %s, <16 x i1> %v, <16 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vpreduce_xor_v16i1:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a1, e8, m1, ta, ma
 ; CHECK-NEXT:    vmv1r.v v9, v0
 ; CHECK-NEXT:    vmv1r.v v0, v8
-; CHECK-NEXT:    vsetvli zero, a1, e8, m1, ta, ma
 ; CHECK-NEXT:    vcpop.m a1, v9, v0.t
 ; CHECK-NEXT:    andi a1, a1, 1
 ; CHECK-NEXT:    xor a0, a1, a0
@@ -297,9 +298,9 @@ declare i1 @llvm.vp.reduce.add.v1i1(i1, <1 x i1>, <1 x i1>, i32)
 define zeroext i1 @vpreduce_add_v1i1(i1 zeroext %s, <1 x i1> %v, <1 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vpreduce_add_v1i1:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a1, e8, mf8, ta, ma
 ; CHECK-NEXT:    vmv1r.v v9, v0
 ; CHECK-NEXT:    vmv1r.v v0, v8
-; CHECK-NEXT:    vsetvli zero, a1, e8, mf8, ta, ma
 ; CHECK-NEXT:    vcpop.m a1, v9, v0.t
 ; CHECK-NEXT:    andi a1, a1, 1
 ; CHECK-NEXT:    xor a0, a1, a0
@@ -313,9 +314,9 @@ declare i1 @llvm.vp.reduce.add.v2i1(i1, <2 x i1>, <2 x i1>, i32)
 define zeroext i1 @vpreduce_add_v2i1(i1 zeroext %s, <2 x i1> %v, <2 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vpreduce_add_v2i1:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a1, e8, mf8, ta, ma
 ; CHECK-NEXT:    vmv1r.v v9, v0
 ; CHECK-NEXT:    vmv1r.v v0, v8
-; CHECK-NEXT:    vsetvli zero, a1, e8, mf8, ta, ma
 ; CHECK-NEXT:    vcpop.m a1, v9, v0.t
 ; CHECK-NEXT:    andi a1, a1, 1
 ; CHECK-NEXT:    xor a0, a1, a0
@@ -329,9 +330,9 @@ declare i1 @llvm.vp.reduce.add.v4i1(i1, <4 x i1>, <4 x i1>, i32)
 define zeroext i1 @vpreduce_add_v4i1(i1 zeroext %s, <4 x i1> %v, <4 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vpreduce_add_v4i1:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a1, e8, mf4, ta, ma
 ; CHECK-NEXT:    vmv1r.v v9, v0
 ; CHECK-NEXT:    vmv1r.v v0, v8
-; CHECK-NEXT:    vsetvli zero, a1, e8, mf4, ta, ma
 ; CHECK-NEXT:    vcpop.m a1, v9, v0.t
 ; CHECK-NEXT:    andi a1, a1, 1
 ; CHECK-NEXT:    xor a0, a1, a0
@@ -345,9 +346,9 @@ declare i1 @llvm.vp.reduce.add.v8i1(i1, <8 x i1>, <8 x i1>, i32)
 define zeroext i1 @vpreduce_add_v8i1(i1 zeroext %s, <8 x i1> %v, <8 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vpreduce_add_v8i1:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a1, e8, mf2, ta, ma
 ; CHECK-NEXT:    vmv1r.v v9, v0
 ; CHECK-NEXT:    vmv1r.v v0, v8
-; CHECK-NEXT:    vsetvli zero, a1, e8, mf2, ta, ma
 ; CHECK-NEXT:    vcpop.m a1, v9, v0.t
 ; CHECK-NEXT:    andi a1, a1, 1
 ; CHECK-NEXT:    xor a0, a1, a0
@@ -361,9 +362,9 @@ declare i1 @llvm.vp.reduce.add.v16i1(i1, <16 x i1>, <16 x i1>, i32)
 define zeroext i1 @vpreduce_add_v16i1(i1 zeroext %s, <16 x i1> %v, <16 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vpreduce_add_v16i1:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a1, e8, m1, ta, ma
 ; CHECK-NEXT:    vmv1r.v v9, v0
 ; CHECK-NEXT:    vmv1r.v v0, v8
-; CHECK-NEXT:    vsetvli zero, a1, e8, m1, ta, ma
 ; CHECK-NEXT:    vcpop.m a1, v9, v0.t
 ; CHECK-NEXT:    andi a1, a1, 1
 ; CHECK-NEXT:    xor a0, a1, a0
@@ -489,9 +490,9 @@ declare i1 @llvm.vp.reduce.smin.v1i1(i1, <1 x i1>, <1 x i1>, i32)
 define zeroext i1 @vpreduce_smin_v1i1(i1 zeroext %s, <1 x i1> %v, <1 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vpreduce_smin_v1i1:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a1, e8, mf8, ta, ma
 ; CHECK-NEXT:    vmv1r.v v9, v0
 ; CHECK-NEXT:    vmv1r.v v0, v8
-; CHECK-NEXT:    vsetvli zero, a1, e8, mf8, ta, ma
 ; CHECK-NEXT:    vcpop.m a1, v9, v0.t
 ; CHECK-NEXT:    snez a1, a1
 ; CHECK-NEXT:    or a0, a1, a0
@@ -505,9 +506,9 @@ declare i1 @llvm.vp.reduce.smin.v2i1(i1, <2 x i1>, <2 x i1>, i32)
 define zeroext i1 @vpreduce_smin_v2i1(i1 zeroext %s, <2 x i1> %v, <2 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vpreduce_smin_v2i1:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a1, e8, mf8, ta, ma
 ; CHECK-NEXT:    vmv1r.v v9, v0
 ; CHECK-NEXT:    vmv1r.v v0, v8
-; CHECK-NEXT:    vsetvli zero, a1, e8, mf8, ta, ma
 ; CHECK-NEXT:    vcpop.m a1, v9, v0.t
 ; CHECK-NEXT:    snez a1, a1
 ; CHECK-NEXT:    or a0, a1, a0
@@ -521,9 +522,9 @@ declare i1 @llvm.vp.reduce.smin.v4i1(i1, <4 x i1>, <4 x i1>, i32)
 define zeroext i1 @vpreduce_smin_v4i1(i1 zeroext %s, <4 x i1> %v, <4 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vpreduce_smin_v4i1:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a1, e8, mf4, ta, ma
 ; CHECK-NEXT:    vmv1r.v v9, v0
 ; CHECK-NEXT:    vmv1r.v v0, v8
-; CHECK-NEXT:    vsetvli zero, a1, e8, mf4, ta, ma
 ; CHECK-NEXT:    vcpop.m a1, v9, v0.t
 ; CHECK-NEXT:    snez a1, a1
 ; CHECK-NEXT:    or a0, a1, a0
@@ -537,9 +538,9 @@ declare i1 @llvm.vp.reduce.smin.v8i1(i1, <8 x i1>, <8 x i1>, i32)
 define zeroext i1 @vpreduce_smin_v8i1(i1 zeroext %s, <8 x i1> %v, <8 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vpreduce_smin_v8i1:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a1, e8, mf2, ta, ma
 ; CHECK-NEXT:    vmv1r.v v9, v0
 ; CHECK-NEXT:    vmv1r.v v0, v8
-; CHECK-NEXT:    vsetvli zero, a1, e8, mf2, ta, ma
 ; CHECK-NEXT:    vcpop.m a1, v9, v0.t
 ; CHECK-NEXT:    snez a1, a1
 ; CHECK-NEXT:    or a0, a1, a0
@@ -553,9 +554,9 @@ declare i1 @llvm.vp.reduce.smin.v16i1(i1, <16 x i1>, <16 x i1>, i32)
 define zeroext i1 @vpreduce_smin_v16i1(i1 zeroext %s, <16 x i1> %v, <16 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vpreduce_smin_v16i1:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a1, e8, m1, ta, ma
 ; CHECK-NEXT:    vmv1r.v v9, v0
 ; CHECK-NEXT:    vmv1r.v v0, v8
-; CHECK-NEXT:    vsetvli zero, a1, e8, m1, ta, ma
 ; CHECK-NEXT:    vcpop.m a1, v9, v0.t
 ; CHECK-NEXT:    snez a1, a1
 ; CHECK-NEXT:    or a0, a1, a0
@@ -569,9 +570,9 @@ declare i1 @llvm.vp.reduce.smin.v32i1(i1, <32 x i1>, <32 x i1>, i32)
 define zeroext i1 @vpreduce_smin_v32i1(i1 zeroext %s, <32 x i1> %v, <32 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vpreduce_smin_v32i1:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a1, e8, m2, ta, ma
 ; CHECK-NEXT:    vmv1r.v v9, v0
 ; CHECK-NEXT:    vmv1r.v v0, v8
-; CHECK-NEXT:    vsetvli zero, a1, e8, m2, ta, ma
 ; CHECK-NEXT:    vcpop.m a1, v9, v0.t
 ; CHECK-NEXT:    snez a1, a1
 ; CHECK-NEXT:    or a0, a1, a0
@@ -585,9 +586,9 @@ declare i1 @llvm.vp.reduce.smin.v64i1(i1, <64 x i1>, <64 x i1>, i32)
 define zeroext i1 @vpreduce_smin_v64i1(i1 zeroext %s, <64 x i1> %v, <64 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vpreduce_smin_v64i1:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a1, e8, m4, ta, ma
 ; CHECK-NEXT:    vmv1r.v v9, v0
 ; CHECK-NEXT:    vmv1r.v v0, v8
-; CHECK-NEXT:    vsetvli zero, a1, e8, m4, ta, ma
 ; CHECK-NEXT:    vcpop.m a1, v9, v0.t
 ; CHECK-NEXT:    snez a1, a1
 ; CHECK-NEXT:    or a0, a1, a0
@@ -601,9 +602,9 @@ declare i1 @llvm.vp.reduce.umax.v1i1(i1, <1 x i1>, <1 x i1>, i32)
 define zeroext i1 @vpreduce_umax_v1i1(i1 zeroext %s, <1 x i1> %v, <1 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vpreduce_umax_v1i1:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a1, e8, mf8, ta, ma
 ; CHECK-NEXT:    vmv1r.v v9, v0
 ; CHECK-NEXT:    vmv1r.v v0, v8
-; CHECK-NEXT:    vsetvli zero, a1, e8, mf8, ta, ma
 ; CHECK-NEXT:    vcpop.m a1, v9, v0.t
 ; CHECK-NEXT:    snez a1, a1
 ; CHECK-NEXT:    or a0, a1, a0
@@ -617,9 +618,9 @@ declare i1 @llvm.vp.reduce.umax.v2i1(i1, <2 x i1>, <2 x i1>, i32)
 define zeroext i1 @vpreduce_umax_v2i1(i1 zeroext %s, <2 x i1> %v, <2 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vpreduce_umax_v2i1:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a1, e8, mf8, ta, ma
 ; CHECK-NEXT:    vmv1r.v v9, v0
 ; CHECK-NEXT:    vmv1r.v v0, v8
-; CHECK-NEXT:    vsetvli zero, a1, e8, mf8, ta, ma
 ; CHECK-NEXT:    vcpop.m a1, v9, v0.t
 ; CHECK-NEXT:    snez a1, a1
 ; CHECK-NEXT:    or a0, a1, a0
@@ -633,9 +634,9 @@ declare i1 @llvm.vp.reduce.umax.v4i1(i1, <4 x i1>, <4 x i1>, i32)
 define zeroext i1 @vpreduce_umax_v4i1(i1 zeroext %s, <4 x i1> %v, <4 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vpreduce_umax_v4i1:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a1, e8, mf4, ta, ma
 ; CHECK-NEXT:    vmv1r.v v9, v0
 ; CHECK-NEXT:    vmv1r.v v0, v8
-; CHECK-NEXT:    vsetvli zero, a1, e8, mf4, ta, ma
 ; CHECK-NEXT:    vcpop.m a1, v9, v0.t
 ; CHECK-NEXT:    snez a1, a1
 ; CHECK-NEXT:    or a0, a1, a0
@@ -649,9 +650,9 @@ declare i1 @llvm.vp.reduce.umax.v8i1(i1, <8 x i1>, <8 x i1>, i32)
 define zeroext i1 @vpreduce_umax_v8i1(i1 zeroext %s, <8 x i1> %v, <8 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vpreduce_umax_v8i1:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a1, e8, mf2, ta, ma
 ; CHECK-NEXT:    vmv1r.v v9, v0
 ; CHECK-NEXT:    vmv1r.v v0, v8
-; CHECK-NEXT:    vsetvli zero, a1, e8, mf2, ta, ma
 ; CHECK-NEXT:    vcpop.m a1, v9, v0.t
 ; CHECK-NEXT:    snez a1, a1
 ; CHECK-NEXT:    or a0, a1, a0
@@ -665,9 +666,9 @@ declare i1 @llvm.vp.reduce.umax.v16i1(i1, <16 x i1>, <16 x i1>, i32)
 define zeroext i1 @vpreduce_umax_v16i1(i1 zeroext %s, <16 x i1> %v, <16 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vpreduce_umax_v16i1:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a1, e8, m1, ta, ma
 ; CHECK-NEXT:    vmv1r.v v9, v0
 ; CHECK-NEXT:    vmv1r.v v0, v8
-; CHECK-NEXT:    vsetvli zero, a1, e8, m1, ta, ma
 ; CHECK-NEXT:    vcpop.m a1, v9, v0.t
 ; CHECK-NEXT:    snez a1, a1
 ; CHECK-NEXT:    or a0, a1, a0
@@ -681,9 +682,9 @@ declare i1 @llvm.vp.reduce.umax.v32i1(i1, <32 x i1>, <32 x i1>, i32)
 define zeroext i1 @vpreduce_umax_v32i1(i1 zeroext %s, <32 x i1> %v, <32 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vpreduce_umax_v32i1:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a1, e8, m2, ta, ma
 ; CHECK-NEXT:    vmv1r.v v9, v0
 ; CHECK-NEXT:    vmv1r.v v0, v8
-; CHECK-NEXT:    vsetvli zero, a1, e8, m2, ta, ma
 ; CHECK-NEXT:    vcpop.m a1, v9, v0.t
 ; CHECK-NEXT:    snez a1, a1
 ; CHECK-NEXT:    or a0, a1, a0
@@ -697,9 +698,9 @@ declare i1 @llvm.vp.reduce.umax.v64i1(i1, <64 x i1>, <64 x i1>, i32)
 define zeroext i1 @vpreduce_umax_v64i1(i1 zeroext %s, <64 x i1> %v, <64 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vpreduce_umax_v64i1:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a1, e8, m4, ta, ma
 ; CHECK-NEXT:    vmv1r.v v9, v0
 ; CHECK-NEXT:    vmv1r.v v0, v8
-; CHECK-NEXT:    vsetvli zero, a1, e8, m4, ta, ma
 ; CHECK-NEXT:    vcpop.m a1, v9, v0.t
 ; CHECK-NEXT:    snez a1, a1
 ; CHECK-NEXT:    or a0, a1, a0
diff --git a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-rint-vp.ll b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-rint-vp.ll
index b8617fda3aa7e..266772d36ee9c 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-rint-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-rint-vp.ll
@@ -123,10 +123,10 @@ declare <16 x half> @llvm.vp.rint.v16f16(<16 x half>, <16 x i1>, i32)
 define <16 x half> @vp_rint_v16f16(<16 x half> %va, <16 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vp_rint_v16f16:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v10, v0
-; CHECK-NEXT:    lui a1, %hi(.LCPI6_0)
-; CHECK-NEXT:    flh fa5, %lo(.LCPI6_0)(a1)
 ; CHECK-NEXT:    vsetvli zero, a0, e16, m2, ta, ma
+; CHECK-NEXT:    vmv1r.v v10, v0
+; CHECK-NEXT:    lui a0, %hi(.LCPI6_0)
+; CHECK-NEXT:    flh fa5, %lo(.LCPI6_0)(a0)
 ; CHECK-NEXT:    vfabs.v v12, v8, v0.t
 ; CHECK-NEXT:    vsetvli zero, zero, e16, m2, ta, mu
 ; CHECK-NEXT:    vmflt.vf v10, v12, fa5, v0.t
@@ -239,8 +239,8 @@ declare <8 x float> @llvm.vp.rint.v8f32(<8 x float>, <8 x i1>, i32)
 define <8 x float> @vp_rint_v8f32(<8 x float> %va, <8 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vp_rint_v8f32:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e32, m2, ta, ma
+; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vfabs.v v12, v8, v0.t
 ; CHECK-NEXT:    lui a0, 307200
 ; CHECK-NEXT:    fmv.w.x fa5, a0
@@ -279,8 +279,8 @@ declare <16 x float> @llvm.vp.rint.v16f32(<16 x float>, <16 x i1>, i32)
 define <16 x float> @vp_rint_v16f32(<16 x float> %va, <16 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vp_rint_v16f32:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v12, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e32, m4, ta, ma
+; CHECK-NEXT:    vmv1r.v v12, v0
 ; CHECK-NEXT:    vfabs.v v16, v8, v0.t
 ; CHECK-NEXT:    lui a0, 307200
 ; CHECK-NEXT:    fmv.w.x fa5, a0
@@ -357,10 +357,10 @@ declare <4 x double> @llvm.vp.rint.v4f64(<4 x double>, <4 x i1>, i32)
 define <4 x double> @vp_rint_v4f64(<4 x double> %va, <4 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vp_rint_v4f64:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v10, v0
-; CHECK-NEXT:    lui a1, %hi(.LCPI18_0)
-; CHECK-NEXT:    fld fa5, %lo(.LCPI18_0)(a1)
 ; CHECK-NEXT:    vsetvli zero, a0, e64, m2, ta, ma
+; CHECK-NEXT:    vmv1r.v v10, v0
+; CHECK-NEXT:    lui a0, %hi(.LCPI18_0)
+; CHECK-NEXT:    fld fa5, %lo(.LCPI18_0)(a0)
 ; CHECK-NEXT:    vfabs.v v12, v8, v0.t
 ; CHECK-NEXT:    vsetvli zero, zero, e64, m2, ta, mu
 ; CHECK-NEXT:    vmflt.vf v10, v12, fa5, v0.t
@@ -397,10 +397,10 @@ declare <8 x double> @llvm.vp.rint.v8f64(<8 x double>, <8 x i1>, i32)
 define <8 x double> @vp_rint_v8f64(<8 x double> %va, <8 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vp_rint_v8f64:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v12, v0
-; CHECK-NEXT:    lui a1, %hi(.LCPI20_0)
-; CHECK-NEXT:    fld fa5, %lo(.LCPI20_0)(a1)
 ; CHECK-NEXT:    vsetvli zero, a0, e64, m4, ta, ma
+; CHECK-NEXT:    vmv1r.v v12, v0
+; CHECK-NEXT:    lui a0, %hi(.LCPI20_0)
+; CHECK-NEXT:    fld fa5, %lo(.LCPI20_0)(a0)
 ; CHECK-NEXT:    vfabs.v v16, v8, v0.t
 ; CHECK-NEXT:    vsetvli zero, zero, e64, m4, ta, mu
 ; CHECK-NEXT:    vmflt.vf v12, v16, fa5, v0.t
@@ -437,10 +437,10 @@ declare <15 x double> @llvm.vp.rint.v15f64(<15 x double>, <15 x i1>, i32)
 define <15 x double> @vp_rint_v15f64(<15 x double> %va, <15 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vp_rint_v15f64:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v16, v0
-; CHECK-NEXT:    lui a1, %hi(.LCPI22_0)
-; CHECK-NEXT:    fld fa5, %lo(.LCPI22_0)(a1)
 ; CHECK-NEXT:    vsetvli zero, a0, e64, m8, ta, ma
+; CHECK-NEXT:    vmv1r.v v16, v0
+; CHECK-NEXT:    lui a0, %hi(.LCPI22_0)
+; CHECK-NEXT:    fld fa5, %lo(.LCPI22_0)(a0)
 ; CHECK-NEXT:    vfabs.v v24, v8, v0.t
 ; CHECK-NEXT:    vsetvli zero, zero, e64, m8, ta, mu
 ; CHECK-NEXT:    vmflt.vf v16, v24, fa5, v0.t
@@ -477,10 +477,10 @@ declare <16 x double> @llvm.vp.rint.v16f64(<16 x double>, <16 x i1>, i32)
 define <16 x double> @vp_rint_v16f64(<16 x double> %va, <16 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vp_rint_v16f64:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v16, v0
-; CHECK-NEXT:    lui a1, %hi(.LCPI24_0)
-; CHECK-NEXT:    fld fa5, %lo(.LCPI24_0)(a1)
 ; CHECK-NEXT:    vsetvli zero, a0, e64, m8, ta, ma
+; CHECK-NEXT:    vmv1r.v v16, v0
+; CHECK-NEXT:    lui a0, %hi(.LCPI24_0)
+; CHECK-NEXT:    fld fa5, %lo(.LCPI24_0)(a0)
 ; CHECK-NEXT:    vfabs.v v24, v8, v0.t
 ; CHECK-NEXT:    vsetvli zero, zero, e64, m8, ta, mu
 ; CHECK-NEXT:    vmflt.vf v16, v24, fa5, v0.t
@@ -517,9 +517,9 @@ declare <32 x double> @llvm.vp.rint.v32f64(<32 x double>, <32 x i1>, i32)
 define <32 x double> @vp_rint_v32f64(<32 x double> %va, <32 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vp_rint_v32f64:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetivli zero, 2, e8, mf4, ta, ma
 ; CHECK-NEXT:    vmv1r.v v6, v0
 ; CHECK-NEXT:    li a2, 16
-; CHECK-NEXT:    vsetivli zero, 2, e8, mf4, ta, ma
 ; CHECK-NEXT:    vslidedown.vi v7, v0, 2
 ; CHECK-NEXT:    mv a1, a0
 ; CHECK-NEXT:    bltu a0, a2, .LBB26_2
diff --git a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-round-vp.ll b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-round-vp.ll
index 820a05e3d6042..232a8a4827cb1 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-round-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-round-vp.ll
@@ -194,8 +194,8 @@ define <8 x half> @vp_round_v8f16(<8 x half> %va, <8 x i1> %m, i32 zeroext %evl)
 ;
 ; ZVFHMIN-LABEL: vp_round_v8f16:
 ; ZVFHMIN:       # %bb.0:
-; ZVFHMIN-NEXT:    vmv1r.v v9, v0
 ; ZVFHMIN-NEXT:    vsetivli zero, 8, e16, m1, ta, ma
+; ZVFHMIN-NEXT:    vmv1r.v v9, v0
 ; ZVFHMIN-NEXT:    vfwcvt.f.f.v v10, v8
 ; ZVFHMIN-NEXT:    lui a1, 307200
 ; ZVFHMIN-NEXT:    vsetvli zero, a0, e32, m2, ta, ma
@@ -261,10 +261,10 @@ declare <16 x half> @llvm.vp.round.v16f16(<16 x half>, <16 x i1>, i32)
 define <16 x half> @vp_round_v16f16(<16 x half> %va, <16 x i1> %m, i32 zeroext %evl) {
 ; ZVFH-LABEL: vp_round_v16f16:
 ; ZVFH:       # %bb.0:
-; ZVFH-NEXT:    vmv1r.v v10, v0
-; ZVFH-NEXT:    lui a1, %hi(.LCPI6_0)
-; ZVFH-NEXT:    flh fa5, %lo(.LCPI6_0)(a1)
 ; ZVFH-NEXT:    vsetvli zero, a0, e16, m2, ta, ma
+; ZVFH-NEXT:    vmv1r.v v10, v0
+; ZVFH-NEXT:    lui a0, %hi(.LCPI6_0)
+; ZVFH-NEXT:    flh fa5, %lo(.LCPI6_0)(a0)
 ; ZVFH-NEXT:    vfabs.v v12, v8, v0.t
 ; ZVFH-NEXT:    vsetvli zero, zero, e16, m2, ta, mu
 ; ZVFH-NEXT:    vmflt.vf v10, v12, fa5, v0.t
@@ -280,8 +280,8 @@ define <16 x half> @vp_round_v16f16(<16 x half> %va, <16 x i1> %m, i32 zeroext %
 ;
 ; ZVFHMIN-LABEL: vp_round_v16f16:
 ; ZVFHMIN:       # %bb.0:
-; ZVFHMIN-NEXT:    vmv1r.v v10, v0
 ; ZVFHMIN-NEXT:    vsetivli zero, 16, e16, m2, ta, ma
+; ZVFHMIN-NEXT:    vmv1r.v v10, v0
 ; ZVFHMIN-NEXT:    vfwcvt.f.f.v v12, v8
 ; ZVFHMIN-NEXT:    lui a1, 307200
 ; ZVFHMIN-NEXT:    vsetvli zero, a0, e32, m4, ta, ma
@@ -431,8 +431,8 @@ declare <8 x float> @llvm.vp.round.v8f32(<8 x float>, <8 x i1>, i32)
 define <8 x float> @vp_round_v8f32(<8 x float> %va, <8 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vp_round_v8f32:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e32, m2, ta, ma
+; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vfabs.v v12, v8, v0.t
 ; CHECK-NEXT:    lui a0, 307200
 ; CHECK-NEXT:    fmv.w.x fa5, a0
@@ -475,8 +475,8 @@ declare <16 x float> @llvm.vp.round.v16f32(<16 x float>, <16 x i1>, i32)
 define <16 x float> @vp_round_v16f32(<16 x float> %va, <16 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vp_round_v16f32:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v12, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e32, m4, ta, ma
+; CHECK-NEXT:    vmv1r.v v12, v0
 ; CHECK-NEXT:    vfabs.v v16, v8, v0.t
 ; CHECK-NEXT:    lui a0, 307200
 ; CHECK-NEXT:    fmv.w.x fa5, a0
@@ -561,10 +561,10 @@ declare <4 x double> @llvm.vp.round.v4f64(<4 x double>, <4 x i1>, i32)
 define <4 x double> @vp_round_v4f64(<4 x double> %va, <4 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vp_round_v4f64:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v10, v0
-; CHECK-NEXT:    lui a1, %hi(.LCPI18_0)
-; CHECK-NEXT:    fld fa5, %lo(.LCPI18_0)(a1)
 ; CHECK-NEXT:    vsetvli zero, a0, e64, m2, ta, ma
+; CHECK-NEXT:    vmv1r.v v10, v0
+; CHECK-NEXT:    lui a0, %hi(.LCPI18_0)
+; CHECK-NEXT:    fld fa5, %lo(.LCPI18_0)(a0)
 ; CHECK-NEXT:    vfabs.v v12, v8, v0.t
 ; CHECK-NEXT:    vsetvli zero, zero, e64, m2, ta, mu
 ; CHECK-NEXT:    vmflt.vf v10, v12, fa5, v0.t
@@ -605,10 +605,10 @@ declare <8 x double> @llvm.vp.round.v8f64(<8 x double>, <8 x i1>, i32)
 define <8 x double> @vp_round_v8f64(<8 x double> %va, <8 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vp_round_v8f64:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v12, v0
-; CHECK-NEXT:    lui a1, %hi(.LCPI20_0)
-; CHECK-NEXT:    fld fa5, %lo(.LCPI20_0)(a1)
 ; CHECK-NEXT:    vsetvli zero, a0, e64, m4, ta, ma
+; CHECK-NEXT:    vmv1r.v v12, v0
+; CHECK-NEXT:    lui a0, %hi(.LCPI20_0)
+; CHECK-NEXT:    fld fa5, %lo(.LCPI20_0)(a0)
 ; CHECK-NEXT:    vfabs.v v16, v8, v0.t
 ; CHECK-NEXT:    vsetvli zero, zero, e64, m4, ta, mu
 ; CHECK-NEXT:    vmflt.vf v12, v16, fa5, v0.t
@@ -649,10 +649,10 @@ declare <15 x double> @llvm.vp.round.v15f64(<15 x double>, <15 x i1>, i32)
 define <15 x double> @vp_round_v15f64(<15 x double> %va, <15 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vp_round_v15f64:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v16, v0
-; CHECK-NEXT:    lui a1, %hi(.LCPI22_0)
-; CHECK-NEXT:    fld fa5, %lo(.LCPI22_0)(a1)
 ; CHECK-NEXT:    vsetvli zero, a0, e64, m8, ta, ma
+; CHECK-NEXT:    vmv1r.v v16, v0
+; CHECK-NEXT:    lui a0, %hi(.LCPI22_0)
+; CHECK-NEXT:    fld fa5, %lo(.LCPI22_0)(a0)
 ; CHECK-NEXT:    vfabs.v v24, v8, v0.t
 ; CHECK-NEXT:    vsetvli zero, zero, e64, m8, ta, mu
 ; CHECK-NEXT:    vmflt.vf v16, v24, fa5, v0.t
@@ -693,10 +693,10 @@ declare <16 x double> @llvm.vp.round.v16f64(<16 x double>, <16 x i1>, i32)
 define <16 x double> @vp_round_v16f64(<16 x double> %va, <16 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vp_round_v16f64:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v16, v0
-; CHECK-NEXT:    lui a1, %hi(.LCPI24_0)
-; CHECK-NEXT:    fld fa5, %lo(.LCPI24_0)(a1)
 ; CHECK-NEXT:    vsetvli zero, a0, e64, m8, ta, ma
+; CHECK-NEXT:    vmv1r.v v16, v0
+; CHECK-NEXT:    lui a0, %hi(.LCPI24_0)
+; CHECK-NEXT:    fld fa5, %lo(.LCPI24_0)(a0)
 ; CHECK-NEXT:    vfabs.v v24, v8, v0.t
 ; CHECK-NEXT:    vsetvli zero, zero, e64, m8, ta, mu
 ; CHECK-NEXT:    vmflt.vf v16, v24, fa5, v0.t
@@ -743,6 +743,7 @@ define <32 x double> @vp_round_v32f64(<32 x double> %va, <32 x i1> %m, i32 zeroe
 ; CHECK-NEXT:    slli a1, a1, 4
 ; CHECK-NEXT:    sub sp, sp, a1
 ; CHECK-NEXT:    .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x10, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 16 * vlenb
+; CHECK-NEXT:    vsetivli zero, 2, e8, mf4, ta, ma
 ; CHECK-NEXT:    vmv1r.v v25, v0
 ; CHECK-NEXT:    csrr a1, vlenb
 ; CHECK-NEXT:    slli a1, a1, 3
@@ -750,7 +751,6 @@ define <32 x double> @vp_round_v32f64(<32 x double> %va, <32 x i1> %m, i32 zeroe
 ; CHECK-NEXT:    addi a1, a1, 16
 ; CHECK-NEXT:    vs8r.v v16, (a1) # Unknown-size Folded Spill
 ; CHECK-NEXT:    li a2, 16
-; CHECK-NEXT:    vsetivli zero, 2, e8, mf4, ta, ma
 ; CHECK-NEXT:    vslidedown.vi v24, v0, 2
 ; CHECK-NEXT:    mv a1, a0
 ; CHECK-NEXT:    bltu a0, a2, .LBB26_2
diff --git a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-roundeven-vp.ll b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-roundeven-vp.ll
index 8391c7939180a..7c80c037403c2 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-roundeven-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-roundeven-vp.ll
@@ -194,8 +194,8 @@ define <8 x half> @vp_roundeven_v8f16(<8 x half> %va, <8 x i1> %m, i32 zeroext %
 ;
 ; ZVFHMIN-LABEL: vp_roundeven_v8f16:
 ; ZVFHMIN:       # %bb.0:
-; ZVFHMIN-NEXT:    vmv1r.v v9, v0
 ; ZVFHMIN-NEXT:    vsetivli zero, 8, e16, m1, ta, ma
+; ZVFHMIN-NEXT:    vmv1r.v v9, v0
 ; ZVFHMIN-NEXT:    vfwcvt.f.f.v v10, v8
 ; ZVFHMIN-NEXT:    lui a1, 307200
 ; ZVFHMIN-NEXT:    vsetvli zero, a0, e32, m2, ta, ma
@@ -261,10 +261,10 @@ declare <16 x half> @llvm.vp.roundeven.v16f16(<16 x half>, <16 x i1>, i32)
 define <16 x half> @vp_roundeven_v16f16(<16 x half> %va, <16 x i1> %m, i32 zeroext %evl) {
 ; ZVFH-LABEL: vp_roundeven_v16f16:
 ; ZVFH:       # %bb.0:
-; ZVFH-NEXT:    vmv1r.v v10, v0
-; ZVFH-NEXT:    lui a1, %hi(.LCPI6_0)
-; ZVFH-NEXT:    flh fa5, %lo(.LCPI6_0)(a1)
 ; ZVFH-NEXT:    vsetvli zero, a0, e16, m2, ta, ma
+; ZVFH-NEXT:    vmv1r.v v10, v0
+; ZVFH-NEXT:    lui a0, %hi(.LCPI6_0)
+; ZVFH-NEXT:    flh fa5, %lo(.LCPI6_0)(a0)
 ; ZVFH-NEXT:    vfabs.v v12, v8, v0.t
 ; ZVFH-NEXT:    vsetvli zero, zero, e16, m2, ta, mu
 ; ZVFH-NEXT:    vmflt.vf v10, v12, fa5, v0.t
@@ -280,8 +280,8 @@ define <16 x half> @vp_roundeven_v16f16(<16 x half> %va, <16 x i1> %m, i32 zeroe
 ;
 ; ZVFHMIN-LABEL: vp_roundeven_v16f16:
 ; ZVFHMIN:       # %bb.0:
-; ZVFHMIN-NEXT:    vmv1r.v v10, v0
 ; ZVFHMIN-NEXT:    vsetivli zero, 16, e16, m2, ta, ma
+; ZVFHMIN-NEXT:    vmv1r.v v10, v0
 ; ZVFHMIN-NEXT:    vfwcvt.f.f.v v12, v8
 ; ZVFHMIN-NEXT:    lui a1, 307200
 ; ZVFHMIN-NEXT:    vsetvli zero, a0, e32, m4, ta, ma
@@ -431,8 +431,8 @@ declare <8 x float> @llvm.vp.roundeven.v8f32(<8 x float>, <8 x i1>, i32)
 define <8 x float> @vp_roundeven_v8f32(<8 x float> %va, <8 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vp_roundeven_v8f32:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e32, m2, ta, ma
+; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vfabs.v v12, v8, v0.t
 ; CHECK-NEXT:    lui a0, 307200
 ; CHECK-NEXT:    fmv.w.x fa5, a0
@@ -475,8 +475,8 @@ declare <16 x float> @llvm.vp.roundeven.v16f32(<16 x float>, <16 x i1>, i32)
 define <16 x float> @vp_roundeven_v16f32(<16 x float> %va, <16 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vp_roundeven_v16f32:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v12, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e32, m4, ta, ma
+; CHECK-NEXT:    vmv1r.v v12, v0
 ; CHECK-NEXT:    vfabs.v v16, v8, v0.t
 ; CHECK-NEXT:    lui a0, 307200
 ; CHECK-NEXT:    fmv.w.x fa5, a0
@@ -561,10 +561,10 @@ declare <4 x double> @llvm.vp.roundeven.v4f64(<4 x double>, <4 x i1>, i32)
 define <4 x double> @vp_roundeven_v4f64(<4 x double> %va, <4 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vp_roundeven_v4f64:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v10, v0
-; CHECK-NEXT:    lui a1, %hi(.LCPI18_0)
-; CHECK-NEXT:    fld fa5, %lo(.LCPI18_0)(a1)
 ; CHECK-NEXT:    vsetvli zero, a0, e64, m2, ta, ma
+; CHECK-NEXT:    vmv1r.v v10, v0
+; CHECK-NEXT:    lui a0, %hi(.LCPI18_0)
+; CHECK-NEXT:    fld fa5, %lo(.LCPI18_0)(a0)
 ; CHECK-NEXT:    vfabs.v v12, v8, v0.t
 ; CHECK-NEXT:    vsetvli zero, zero, e64, m2, ta, mu
 ; CHECK-NEXT:    vmflt.vf v10, v12, fa5, v0.t
@@ -605,10 +605,10 @@ declare <8 x double> @llvm.vp.roundeven.v8f64(<8 x double>, <8 x i1>, i32)
 define <8 x double> @vp_roundeven_v8f64(<8 x double> %va, <8 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vp_roundeven_v8f64:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v12, v0
-; CHECK-NEXT:    lui a1, %hi(.LCPI20_0)
-; CHECK-NEXT:    fld fa5, %lo(.LCPI20_0)(a1)
 ; CHECK-NEXT:    vsetvli zero, a0, e64, m4, ta, ma
+; CHECK-NEXT:    vmv1r.v v12, v0
+; CHECK-NEXT:    lui a0, %hi(.LCPI20_0)
+; CHECK-NEXT:    fld fa5, %lo(.LCPI20_0)(a0)
 ; CHECK-NEXT:    vfabs.v v16, v8, v0.t
 ; CHECK-NEXT:    vsetvli zero, zero, e64, m4, ta, mu
 ; CHECK-NEXT:    vmflt.vf v12, v16, fa5, v0.t
@@ -649,10 +649,10 @@ declare <15 x double> @llvm.vp.roundeven.v15f64(<15 x double>, <15 x i1>, i32)
 define <15 x double> @vp_roundeven_v15f64(<15 x double> %va, <15 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vp_roundeven_v15f64:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v16, v0
-; CHECK-NEXT:    lui a1, %hi(.LCPI22_0)
-; CHECK-NEXT:    fld fa5, %lo(.LCPI22_0)(a1)
 ; CHECK-NEXT:    vsetvli zero, a0, e64, m8, ta, ma
+; CHECK-NEXT:    vmv1r.v v16, v0
+; CHECK-NEXT:    lui a0, %hi(.LCPI22_0)
+; CHECK-NEXT:    fld fa5, %lo(.LCPI22_0)(a0)
 ; CHECK-NEXT:    vfabs.v v24, v8, v0.t
 ; CHECK-NEXT:    vsetvli zero, zero, e64, m8, ta, mu
 ; CHECK-NEXT:    vmflt.vf v16, v24, fa5, v0.t
@@ -693,10 +693,10 @@ declare <16 x double> @llvm.vp.roundeven.v16f64(<16 x double>, <16 x i1>, i32)
 define <16 x double> @vp_roundeven_v16f64(<16 x double> %va, <16 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vp_roundeven_v16f64:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v16, v0
-; CHECK-NEXT:    lui a1, %hi(.LCPI24_0)
-; CHECK-NEXT:    fld fa5, %lo(.LCPI24_0)(a1)
 ; CHECK-NEXT:    vsetvli zero, a0, e64, m8, ta, ma
+; CHECK-NEXT:    vmv1r.v v16, v0
+; CHECK-NEXT:    lui a0, %hi(.LCPI24_0)
+; CHECK-NEXT:    fld fa5, %lo(.LCPI24_0)(a0)
 ; CHECK-NEXT:    vfabs.v v24, v8, v0.t
 ; CHECK-NEXT:    vsetvli zero, zero, e64, m8, ta, mu
 ; CHECK-NEXT:    vmflt.vf v16, v24, fa5, v0.t
@@ -743,6 +743,7 @@ define <32 x double> @vp_roundeven_v32f64(<32 x double> %va, <32 x i1> %m, i32 z
 ; CHECK-NEXT:    slli a1, a1, 4
 ; CHECK-NEXT:    sub sp, sp, a1
 ; CHECK-NEXT:    .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x10, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 16 * vlenb
+; CHECK-NEXT:    vsetivli zero, 2, e8, mf4, ta, ma
 ; CHECK-NEXT:    vmv1r.v v25, v0
 ; CHECK-NEXT:    csrr a1, vlenb
 ; CHECK-NEXT:    slli a1, a1, 3
@@ -750,7 +751,6 @@ define <32 x double> @vp_roundeven_v32f64(<32 x double> %va, <32 x i1> %m, i32 z
 ; CHECK-NEXT:    addi a1, a1, 16
 ; CHECK-NEXT:    vs8r.v v16, (a1) # Unknown-size Folded Spill
 ; CHECK-NEXT:    li a2, 16
-; CHECK-NEXT:    vsetivli zero, 2, e8, mf4, ta, ma
 ; CHECK-NEXT:    vslidedown.vi v24, v0, 2
 ; CHECK-NEXT:    mv a1, a0
 ; CHECK-NEXT:    bltu a0, a2, .LBB26_2
diff --git a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-roundtozero-vp.ll b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-roundtozero-vp.ll
index 8c38d24460265..65a4725267cd3 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-roundtozero-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-roundtozero-vp.ll
@@ -194,8 +194,8 @@ define <8 x half> @vp_roundtozero_v8f16(<8 x half> %va, <8 x i1> %m, i32 zeroext
 ;
 ; ZVFHMIN-LABEL: vp_roundtozero_v8f16:
 ; ZVFHMIN:       # %bb.0:
-; ZVFHMIN-NEXT:    vmv1r.v v9, v0
 ; ZVFHMIN-NEXT:    vsetivli zero, 8, e16, m1, ta, ma
+; ZVFHMIN-NEXT:    vmv1r.v v9, v0
 ; ZVFHMIN-NEXT:    vfwcvt.f.f.v v10, v8
 ; ZVFHMIN-NEXT:    lui a1, 307200
 ; ZVFHMIN-NEXT:    vsetvli zero, a0, e32, m2, ta, ma
@@ -261,10 +261,10 @@ declare <16 x half> @llvm.vp.roundtozero.v16f16(<16 x half>, <16 x i1>, i32)
 define <16 x half> @vp_roundtozero_v16f16(<16 x half> %va, <16 x i1> %m, i32 zeroext %evl) {
 ; ZVFH-LABEL: vp_roundtozero_v16f16:
 ; ZVFH:       # %bb.0:
-; ZVFH-NEXT:    vmv1r.v v10, v0
-; ZVFH-NEXT:    lui a1, %hi(.LCPI6_0)
-; ZVFH-NEXT:    flh fa5, %lo(.LCPI6_0)(a1)
 ; ZVFH-NEXT:    vsetvli zero, a0, e16, m2, ta, ma
+; ZVFH-NEXT:    vmv1r.v v10, v0
+; ZVFH-NEXT:    lui a0, %hi(.LCPI6_0)
+; ZVFH-NEXT:    flh fa5, %lo(.LCPI6_0)(a0)
 ; ZVFH-NEXT:    vfabs.v v12, v8, v0.t
 ; ZVFH-NEXT:    vsetvli zero, zero, e16, m2, ta, mu
 ; ZVFH-NEXT:    vmflt.vf v10, v12, fa5, v0.t
@@ -280,8 +280,8 @@ define <16 x half> @vp_roundtozero_v16f16(<16 x half> %va, <16 x i1> %m, i32 zer
 ;
 ; ZVFHMIN-LABEL: vp_roundtozero_v16f16:
 ; ZVFHMIN:       # %bb.0:
-; ZVFHMIN-NEXT:    vmv1r.v v10, v0
 ; ZVFHMIN-NEXT:    vsetivli zero, 16, e16, m2, ta, ma
+; ZVFHMIN-NEXT:    vmv1r.v v10, v0
 ; ZVFHMIN-NEXT:    vfwcvt.f.f.v v12, v8
 ; ZVFHMIN-NEXT:    lui a1, 307200
 ; ZVFHMIN-NEXT:    vsetvli zero, a0, e32, m4, ta, ma
@@ -431,8 +431,8 @@ declare <8 x float> @llvm.vp.roundtozero.v8f32(<8 x float>, <8 x i1>, i32)
 define <8 x float> @vp_roundtozero_v8f32(<8 x float> %va, <8 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vp_roundtozero_v8f32:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e32, m2, ta, ma
+; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vfabs.v v12, v8, v0.t
 ; CHECK-NEXT:    lui a0, 307200
 ; CHECK-NEXT:    fmv.w.x fa5, a0
@@ -475,8 +475,8 @@ declare <16 x float> @llvm.vp.roundtozero.v16f32(<16 x float>, <16 x i1>, i32)
 define <16 x float> @vp_roundtozero_v16f32(<16 x float> %va, <16 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vp_roundtozero_v16f32:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v12, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e32, m4, ta, ma
+; CHECK-NEXT:    vmv1r.v v12, v0
 ; CHECK-NEXT:    vfabs.v v16, v8, v0.t
 ; CHECK-NEXT:    lui a0, 307200
 ; CHECK-NEXT:    fmv.w.x fa5, a0
@@ -561,10 +561,10 @@ declare <4 x double> @llvm.vp.roundtozero.v4f64(<4 x double>, <4 x i1>, i32)
 define <4 x double> @vp_roundtozero_v4f64(<4 x double> %va, <4 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vp_roundtozero_v4f64:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v10, v0
-; CHECK-NEXT:    lui a1, %hi(.LCPI18_0)
-; CHECK-NEXT:    fld fa5, %lo(.LCPI18_0)(a1)
 ; CHECK-NEXT:    vsetvli zero, a0, e64, m2, ta, ma
+; CHECK-NEXT:    vmv1r.v v10, v0
+; CHECK-NEXT:    lui a0, %hi(.LCPI18_0)
+; CHECK-NEXT:    fld fa5, %lo(.LCPI18_0)(a0)
 ; CHECK-NEXT:    vfabs.v v12, v8, v0.t
 ; CHECK-NEXT:    vsetvli zero, zero, e64, m2, ta, mu
 ; CHECK-NEXT:    vmflt.vf v10, v12, fa5, v0.t
@@ -605,10 +605,10 @@ declare <8 x double> @llvm.vp.roundtozero.v8f64(<8 x double>, <8 x i1>, i32)
 define <8 x double> @vp_roundtozero_v8f64(<8 x double> %va, <8 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vp_roundtozero_v8f64:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v12, v0
-; CHECK-NEXT:    lui a1, %hi(.LCPI20_0)
-; CHECK-NEXT:    fld fa5, %lo(.LCPI20_0)(a1)
 ; CHECK-NEXT:    vsetvli zero, a0, e64, m4, ta, ma
+; CHECK-NEXT:    vmv1r.v v12, v0
+; CHECK-NEXT:    lui a0, %hi(.LCPI20_0)
+; CHECK-NEXT:    fld fa5, %lo(.LCPI20_0)(a0)
 ; CHECK-NEXT:    vfabs.v v16, v8, v0.t
 ; CHECK-NEXT:    vsetvli zero, zero, e64, m4, ta, mu
 ; CHECK-NEXT:    vmflt.vf v12, v16, fa5, v0.t
@@ -649,10 +649,10 @@ declare <15 x double> @llvm.vp.roundtozero.v15f64(<15 x double>, <15 x i1>, i32)
 define <15 x double> @vp_roundtozero_v15f64(<15 x double> %va, <15 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vp_roundtozero_v15f64:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v16, v0
-; CHECK-NEXT:    lui a1, %hi(.LCPI22_0)
-; CHECK-NEXT:    fld fa5, %lo(.LCPI22_0)(a1)
 ; CHECK-NEXT:    vsetvli zero, a0, e64, m8, ta, ma
+; CHECK-NEXT:    vmv1r.v v16, v0
+; CHECK-NEXT:    lui a0, %hi(.LCPI22_0)
+; CHECK-NEXT:    fld fa5, %lo(.LCPI22_0)(a0)
 ; CHECK-NEXT:    vfabs.v v24, v8, v0.t
 ; CHECK-NEXT:    vsetvli zero, zero, e64, m8, ta, mu
 ; CHECK-NEXT:    vmflt.vf v16, v24, fa5, v0.t
@@ -693,10 +693,10 @@ declare <16 x double> @llvm.vp.roundtozero.v16f64(<16 x double>, <16 x i1>, i32)
 define <16 x double> @vp_roundtozero_v16f64(<16 x double> %va, <16 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vp_roundtozero_v16f64:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v16, v0
-; CHECK-NEXT:    lui a1, %hi(.LCPI24_0)
-; CHECK-NEXT:    fld fa5, %lo(.LCPI24_0)(a1)
 ; CHECK-NEXT:    vsetvli zero, a0, e64, m8, ta, ma
+; CHECK-NEXT:    vmv1r.v v16, v0
+; CHECK-NEXT:    lui a0, %hi(.LCPI24_0)
+; CHECK-NEXT:    fld fa5, %lo(.LCPI24_0)(a0)
 ; CHECK-NEXT:    vfabs.v v24, v8, v0.t
 ; CHECK-NEXT:    vsetvli zero, zero, e64, m8, ta, mu
 ; CHECK-NEXT:    vmflt.vf v16, v24, fa5, v0.t
@@ -743,6 +743,7 @@ define <32 x double> @vp_roundtozero_v32f64(<32 x double> %va, <32 x i1> %m, i32
 ; CHECK-NEXT:    slli a1, a1, 4
 ; CHECK-NEXT:    sub sp, sp, a1
 ; CHECK-NEXT:    .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x10, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 16 * vlenb
+; CHECK-NEXT:    vsetivli zero, 2, e8, mf4, ta, ma
 ; CHECK-NEXT:    vmv1r.v v25, v0
 ; CHECK-NEXT:    csrr a1, vlenb
 ; CHECK-NEXT:    slli a1, a1, 3
@@ -750,7 +751,6 @@ define <32 x double> @vp_roundtozero_v32f64(<32 x double> %va, <32 x i1> %m, i32
 ; CHECK-NEXT:    addi a1, a1, 16
 ; CHECK-NEXT:    vs8r.v v16, (a1) # Unknown-size Folded Spill
 ; CHECK-NEXT:    li a2, 16
-; CHECK-NEXT:    vsetivli zero, 2, e8, mf4, ta, ma
 ; CHECK-NEXT:    vslidedown.vi v24, v0, 2
 ; CHECK-NEXT:    mv a1, a0
 ; CHECK-NEXT:    bltu a0, a2, .LBB26_2
diff --git a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-setcc-int-vp.ll b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-setcc-int-vp.ll
index d52c42891fcc3..69d6ffa9f300c 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-setcc-int-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-setcc-int-vp.ll
@@ -598,6 +598,7 @@ define <256 x i1> @icmp_eq_vv_v256i8(<256 x i8> %va, <256 x i8> %vb, <256 x i1>
 ; CHECK-NEXT:    slli a1, a1, 4
 ; CHECK-NEXT:    sub sp, sp, a1
 ; CHECK-NEXT:    .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x10, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 16 * vlenb
+; CHECK-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-NEXT:    vmv1r.v v7, v0
 ; CHECK-NEXT:    csrr a1, vlenb
 ; CHECK-NEXT:    slli a1, a1, 3
@@ -648,6 +649,7 @@ define <256 x i1> @icmp_eq_vv_v256i8(<256 x i8> %va, <256 x i8> %vb, <256 x i1>
 define <256 x i1> @icmp_eq_vx_v256i8(<256 x i8> %va, i8 %b, <256 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: icmp_eq_vx_v256i8:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-NEXT:    vmv1r.v v24, v0
 ; CHECK-NEXT:    li a3, 128
 ; CHECK-NEXT:    vsetvli zero, a3, e8, m8, ta, ma
@@ -677,6 +679,7 @@ define <256 x i1> @icmp_eq_vx_v256i8(<256 x i8> %va, i8 %b, <256 x i1> %m, i32 z
 define <256 x i1> @icmp_eq_vx_swap_v256i8(<256 x i8> %va, i8 %b, <256 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: icmp_eq_vx_swap_v256i8:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-NEXT:    vmv1r.v v24, v0
 ; CHECK-NEXT:    li a3, 128
 ; CHECK-NEXT:    vsetvli zero, a3, e8, m8, ta, ma
diff --git a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-shuffle-concat.ll b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-shuffle-concat.ll
index 38026bb591f79..f2353e7d028bd 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-shuffle-concat.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-shuffle-concat.ll
@@ -8,8 +8,8 @@
 define <8 x i32> @concat_2xv4i32(<4 x i32> %a, <4 x i32> %b) {
 ; VLA-LABEL: concat_2xv4i32:
 ; VLA:       # %bb.0:
-; VLA-NEXT:    vmv1r.v v10, v9
 ; VLA-NEXT:    vsetivli zero, 8, e32, m2, ta, ma
+; VLA-NEXT:    vmv1r.v v10, v9
 ; VLA-NEXT:    vslideup.vi v8, v10, 4
 ; VLA-NEXT:    ret
 ;
@@ -32,9 +32,9 @@ define <8 x i32> @concat_4xv2i32(<2 x i32> %a, <2 x i32> %b, <2 x i32> %c, <2 x
 ;
 ; VLS-LABEL: concat_4xv2i32:
 ; VLS:       # %bb.0:
+; VLS-NEXT:    vsetivli zero, 4, e32, m1, ta, ma
 ; VLS-NEXT:    vmv1r.v v13, v10
 ; VLS-NEXT:    vmv1r.v v12, v8
-; VLS-NEXT:    vsetivli zero, 4, e32, m1, ta, ma
 ; VLS-NEXT:    vslideup.vi v13, v11, 2
 ; VLS-NEXT:    vslideup.vi v12, v9, 2
 ; VLS-NEXT:    vmv2r.v v8, v12
@@ -62,9 +62,9 @@ define <8 x i32> @concat_8xv1i32(<1 x i32> %a, <1 x i32> %b, <1 x i32> %c, <1 x
 ;
 ; VLS-LABEL: concat_8xv1i32:
 ; VLS:       # %bb.0:
+; VLS-NEXT:    vsetivli zero, 2, e32, mf2, ta, ma
 ; VLS-NEXT:    vmv1r.v v17, v12
 ; VLS-NEXT:    vmv1r.v v16, v8
-; VLS-NEXT:    vsetivli zero, 2, e32, mf2, ta, ma
 ; VLS-NEXT:    vslideup.vi v14, v15, 1
 ; VLS-NEXT:    vslideup.vi v17, v13, 1
 ; VLS-NEXT:    vsetivli zero, 4, e32, m1, ta, ma
@@ -89,8 +89,8 @@ define <8 x i32> @concat_8xv1i32(<1 x i32> %a, <1 x i32> %b, <1 x i32> %c, <1 x
 define <16 x i32> @concat_2xv8i32(<8 x i32> %a, <8 x i32> %b) {
 ; VLA-LABEL: concat_2xv8i32:
 ; VLA:       # %bb.0:
-; VLA-NEXT:    vmv2r.v v12, v10
 ; VLA-NEXT:    vsetivli zero, 16, e32, m4, ta, ma
+; VLA-NEXT:    vmv2r.v v12, v10
 ; VLA-NEXT:    vslideup.vi v8, v12, 8
 ; VLA-NEXT:    ret
 ;
@@ -104,10 +104,10 @@ define <16 x i32> @concat_2xv8i32(<8 x i32> %a, <8 x i32> %b) {
 define <16 x i32> @concat_4xv4i32(<4 x i32> %a, <4 x i32> %b, <4 x i32> %c, <4 x i32> %d) {
 ; VLA-LABEL: concat_4xv4i32:
 ; VLA:       # %bb.0:
+; VLA-NEXT:    vsetivli zero, 8, e32, m2, ta, ma
 ; VLA-NEXT:    vmv1r.v v14, v11
 ; VLA-NEXT:    vmv1r.v v12, v10
 ; VLA-NEXT:    vmv1r.v v10, v9
-; VLA-NEXT:    vsetivli zero, 8, e32, m2, ta, ma
 ; VLA-NEXT:    vslideup.vi v12, v14, 4
 ; VLA-NEXT:    vslideup.vi v8, v10, 4
 ; VLA-NEXT:    vsetivli zero, 16, e32, m4, ta, ma
@@ -140,11 +140,11 @@ define <16 x i32> @concat_8xv2i32(<2 x i32> %a, <2 x i32> %b, <2 x i32> %c, <2 x
 ;
 ; VLS-LABEL: concat_8xv2i32:
 ; VLS:       # %bb.0:
+; VLS-NEXT:    vsetivli zero, 4, e32, m1, ta, ma
 ; VLS-NEXT:    vmv1r.v v19, v14
 ; VLS-NEXT:    vmv1r.v v18, v12
 ; VLS-NEXT:    vmv1r.v v17, v10
 ; VLS-NEXT:    vmv1r.v v16, v8
-; VLS-NEXT:    vsetivli zero, 4, e32, m1, ta, ma
 ; VLS-NEXT:    vslideup.vi v19, v15, 2
 ; VLS-NEXT:    vslideup.vi v18, v13, 2
 ; VLS-NEXT:    vslideup.vi v17, v11, 2
@@ -164,6 +164,7 @@ define <16 x i32> @concat_8xv2i32(<2 x i32> %a, <2 x i32> %b, <2 x i32> %c, <2 x
 define <32 x i32> @concat_2xv16i32(<16 x i32> %a, <16 x i32> %b) {
 ; VLA-LABEL: concat_2xv16i32:
 ; VLA:       # %bb.0:
+; VLA-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; VLA-NEXT:    vmv4r.v v16, v12
 ; VLA-NEXT:    li a0, 32
 ; VLA-NEXT:    vsetvli zero, a0, e32, m8, ta, ma
@@ -180,11 +181,11 @@ define <32 x i32> @concat_2xv16i32(<16 x i32> %a, <16 x i32> %b) {
 define <32 x i32> @concat_4xv8i32(<8 x i32> %a, <8 x i32> %b, <8 x i32> %c, <8 x i32> %d) {
 ; VLA-LABEL: concat_4xv8i32:
 ; VLA:       # %bb.0:
+; VLA-NEXT:    vsetivli zero, 16, e32, m4, ta, ma
 ; VLA-NEXT:    vmv2r.v v20, v14
 ; VLA-NEXT:    vmv2r.v v16, v12
 ; VLA-NEXT:    vmv2r.v v12, v10
 ; VLA-NEXT:    li a0, 32
-; VLA-NEXT:    vsetivli zero, 16, e32, m4, ta, ma
 ; VLA-NEXT:    vslideup.vi v16, v20, 8
 ; VLA-NEXT:    vslideup.vi v8, v12, 8
 ; VLA-NEXT:    vsetvli zero, a0, e32, m8, ta, ma
@@ -203,6 +204,7 @@ define <32 x i32> @concat_4xv8i32(<8 x i32> %a, <8 x i32> %b, <8 x i32> %c, <8 x
 define <32 x i32> @concat_8xv4i32(<4 x i32> %a, <4 x i32> %b, <4 x i32> %c, <4 x i32> %d, <4 x i32> %e, <4 x i32> %f, <4 x i32> %g, <4 x i32> %h) {
 ; VLA-LABEL: concat_8xv4i32:
 ; VLA:       # %bb.0:
+; VLA-NEXT:    vsetivli zero, 8, e32, m2, ta, ma
 ; VLA-NEXT:    vmv1r.v v18, v15
 ; VLA-NEXT:    vmv1r.v v20, v14
 ; VLA-NEXT:    vmv1r.v v14, v13
@@ -211,7 +213,6 @@ define <32 x i32> @concat_8xv4i32(<4 x i32> %a, <4 x i32> %b, <4 x i32> %c, <4 x
 ; VLA-NEXT:    vmv1r.v v12, v10
 ; VLA-NEXT:    vmv1r.v v10, v9
 ; VLA-NEXT:    li a0, 32
-; VLA-NEXT:    vsetivli zero, 8, e32, m2, ta, ma
 ; VLA-NEXT:    vslideup.vi v20, v18, 4
 ; VLA-NEXT:    vslideup.vi v16, v14, 4
 ; VLA-NEXT:    vslideup.vi v12, v22, 4
diff --git a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-shuffle-exact-vlen.ll b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-shuffle-exact-vlen.ll
index d461fa8378cff..cadee8acf27d6 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-shuffle-exact-vlen.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-shuffle-exact-vlen.ll
@@ -108,6 +108,7 @@ define <4 x i64> @m2_splat_into_identity(<4 x i64> %v1) vscale_range(2,2) {
 define <4 x i64> @m2_broadcast_i128(<4 x i64> %v1) vscale_range(2,2) {
 ; CHECK-LABEL: m2_broadcast_i128:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-NEXT:    vmv1r.v v9, v8
 ; CHECK-NEXT:    ret
   %res = shufflevector <4 x i64> %v1, <4 x i64> poison, <4 x i32> <i32 0, i32 1, i32 0, i32 1>
@@ -117,6 +118,7 @@ define <4 x i64> @m2_broadcast_i128(<4 x i64> %v1) vscale_range(2,2) {
 define <8 x i64> @m4_broadcast_i128(<8 x i64> %v1) vscale_range(2,2) {
 ; CHECK-LABEL: m4_broadcast_i128:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-NEXT:    vmv1r.v v9, v8
 ; CHECK-NEXT:    vmv1r.v v10, v8
 ; CHECK-NEXT:    vmv1r.v v11, v8
diff --git a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-shuffle-reverse.ll b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-shuffle-reverse.ll
index 407535831aeda..f7647ff38c8a0 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-shuffle-reverse.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-shuffle-reverse.ll
@@ -966,9 +966,9 @@ define <16 x i8> @reverse_v16i8_2(<8 x i8> %a, <8 x i8> %b) {
 define <32 x i8> @reverse_v32i8_2(<16 x i8> %a, <16 x i8> %b) {
 ; CHECK-LABEL: reverse_v32i8_2:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli a0, zero, e16, m2, ta, ma
 ; CHECK-NEXT:    vmv1r.v v10, v9
 ; CHECK-NEXT:    csrr a0, vlenb
-; CHECK-NEXT:    vsetvli a1, zero, e16, m2, ta, ma
 ; CHECK-NEXT:    vid.v v12
 ; CHECK-NEXT:    addi a1, a0, -1
 ; CHECK-NEXT:    vrsub.vx v12, v12, a1
@@ -1035,9 +1035,9 @@ define <8 x i16> @reverse_v8i16_2(<4 x i16> %a, <4 x i16> %b) {
 define <16 x i16> @reverse_v16i16_2(<8 x i16> %a, <8 x i16> %b) {
 ; CHECK-LABEL: reverse_v16i16_2:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli a0, zero, e16, m1, ta, ma
 ; CHECK-NEXT:    vmv1r.v v10, v9
 ; CHECK-NEXT:    csrr a0, vlenb
-; CHECK-NEXT:    vsetvli a1, zero, e16, m1, ta, ma
 ; CHECK-NEXT:    vid.v v9
 ; CHECK-NEXT:    srli a1, a0, 1
 ; CHECK-NEXT:    addi a1, a1, -1
@@ -1060,9 +1060,9 @@ define <16 x i16> @reverse_v16i16_2(<8 x i16> %a, <8 x i16> %b) {
 define <32 x i16> @reverse_v32i16_2(<16 x i16> %a, <16 x i16> %b) {
 ; CHECK-LABEL: reverse_v32i16_2:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli a0, zero, e16, m1, ta, ma
 ; CHECK-NEXT:    vmv2r.v v12, v10
 ; CHECK-NEXT:    csrr a0, vlenb
-; CHECK-NEXT:    vsetvli a1, zero, e16, m1, ta, ma
 ; CHECK-NEXT:    vid.v v10
 ; CHECK-NEXT:    lui a1, 16
 ; CHECK-NEXT:    addi a1, a1, -1
@@ -1116,9 +1116,9 @@ define <4 x i32> @reverse_v4i32_2(<2 x i32> %a, < 2 x i32> %b) {
 define <8 x i32> @reverse_v8i32_2(<4 x i32> %a, <4 x i32> %b) {
 ; CHECK-LABEL: reverse_v8i32_2:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli a0, zero, e32, m1, ta, ma
 ; CHECK-NEXT:    vmv1r.v v10, v9
 ; CHECK-NEXT:    csrr a0, vlenb
-; CHECK-NEXT:    vsetvli a1, zero, e32, m1, ta, ma
 ; CHECK-NEXT:    vid.v v9
 ; CHECK-NEXT:    srli a1, a0, 2
 ; CHECK-NEXT:    addi a1, a1, -1
@@ -1142,9 +1142,9 @@ define <8 x i32> @reverse_v8i32_2(<4 x i32> %a, <4 x i32> %b) {
 define <16 x i32> @reverse_v16i32_2(<8 x i32> %a, <8 x i32> %b) {
 ; CHECK-LABEL: reverse_v16i32_2:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli a0, zero, e32, m1, ta, ma
 ; CHECK-NEXT:    vmv2r.v v12, v10
 ; CHECK-NEXT:    csrr a0, vlenb
-; CHECK-NEXT:    vsetvli a1, zero, e32, m1, ta, ma
 ; CHECK-NEXT:    vid.v v10
 ; CHECK-NEXT:    srli a1, a0, 2
 ; CHECK-NEXT:    addi a1, a1, -1
@@ -1170,9 +1170,9 @@ define <16 x i32> @reverse_v16i32_2(<8 x i32> %a, <8 x i32> %b) {
 define <32 x i32> @reverse_v32i32_2(<16 x i32> %a, <16 x i32> %b) {
 ; CHECK-LABEL: reverse_v32i32_2:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli a0, zero, e32, m1, ta, ma
 ; CHECK-NEXT:    vmv4r.v v16, v12
 ; CHECK-NEXT:    csrr a0, vlenb
-; CHECK-NEXT:    vsetvli a1, zero, e32, m1, ta, ma
 ; CHECK-NEXT:    vid.v v12
 ; CHECK-NEXT:    srli a1, a0, 2
 ; CHECK-NEXT:    addi a1, a1, -1
@@ -1219,9 +1219,9 @@ define <4 x i64> @reverse_v4i64_2(<2 x i64> %a, < 2 x i64> %b) {
 define <8 x i64> @reverse_v8i64_2(<4 x i64> %a, <4 x i64> %b) {
 ; CHECK-LABEL: reverse_v8i64_2:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli a0, zero, e64, m1, ta, ma
 ; CHECK-NEXT:    vmv2r.v v12, v10
 ; CHECK-NEXT:    csrr a0, vlenb
-; CHECK-NEXT:    vsetvli a1, zero, e64, m1, ta, ma
 ; CHECK-NEXT:    vid.v v10
 ; CHECK-NEXT:    srli a1, a0, 3
 ; CHECK-NEXT:    addi a1, a1, -1
@@ -1289,9 +1289,9 @@ define <8 x half> @reverse_v8f16_2(<4 x half> %a, <4 x half> %b) {
 define <16 x half> @reverse_v16f16_2(<8 x half> %a, <8 x half> %b) {
 ; CHECK-LABEL: reverse_v16f16_2:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli a0, zero, e16, m1, ta, ma
 ; CHECK-NEXT:    vmv1r.v v10, v9
 ; CHECK-NEXT:    csrr a0, vlenb
-; CHECK-NEXT:    vsetvli a1, zero, e16, m1, ta, ma
 ; CHECK-NEXT:    vid.v v9
 ; CHECK-NEXT:    srli a1, a0, 1
 ; CHECK-NEXT:    addi a1, a1, -1
@@ -1361,9 +1361,9 @@ define <4 x float> @reverse_v4f32_2(<2 x float> %a, <2 x float> %b) {
 define <8 x float> @reverse_v8f32_2(<4 x float> %a, <4 x float> %b) {
 ; CHECK-LABEL: reverse_v8f32_2:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli a0, zero, e32, m1, ta, ma
 ; CHECK-NEXT:    vmv1r.v v10, v9
 ; CHECK-NEXT:    csrr a0, vlenb
-; CHECK-NEXT:    vsetvli a1, zero, e32, m1, ta, ma
 ; CHECK-NEXT:    vid.v v9
 ; CHECK-NEXT:    srli a1, a0, 2
 ; CHECK-NEXT:    addi a1, a1, -1
@@ -1387,9 +1387,9 @@ define <8 x float> @reverse_v8f32_2(<4 x float> %a, <4 x float> %b) {
 define <16 x float> @reverse_v16f32_2(<8 x float> %a, <8 x float> %b) {
 ; CHECK-LABEL: reverse_v16f32_2:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli a0, zero, e32, m1, ta, ma
 ; CHECK-NEXT:    vmv2r.v v12, v10
 ; CHECK-NEXT:    csrr a0, vlenb
-; CHECK-NEXT:    vsetvli a1, zero, e32, m1, ta, ma
 ; CHECK-NEXT:    vid.v v10
 ; CHECK-NEXT:    srli a1, a0, 2
 ; CHECK-NEXT:    addi a1, a1, -1
@@ -1430,9 +1430,9 @@ define <4 x double> @reverse_v4f64_2(<2 x double> %a, < 2 x double> %b) {
 define <8 x double> @reverse_v8f64_2(<4 x double> %a, <4 x double> %b) {
 ; CHECK-LABEL: reverse_v8f64_2:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli a0, zero, e64, m1, ta, ma
 ; CHECK-NEXT:    vmv2r.v v12, v10
 ; CHECK-NEXT:    csrr a0, vlenb
-; CHECK-NEXT:    vsetvli a1, zero, e64, m1, ta, ma
 ; CHECK-NEXT:    vid.v v10
 ; CHECK-NEXT:    srli a1, a0, 3
 ; CHECK-NEXT:    addi a1, a1, -1
diff --git a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-shuffle-vslide1up.ll b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-shuffle-vslide1up.ll
index c37c3a9ee0ea0..49f6acf9ba8c9 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-shuffle-vslide1up.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-shuffle-vslide1up.ll
@@ -415,8 +415,8 @@ define <4 x i8> @vslide1up_4xi8_neg_incorrect_insert3(<4 x i8> %v, i8 %b) {
 define <2 x i8> @vslide1up_4xi8_neg_length_changing(<4 x i8> %v, i8 %b) {
 ; CHECK-LABEL: vslide1up_4xi8_neg_length_changing:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetivli zero, 1, e8, m1, tu, ma
 ; CHECK-NEXT:    vmv1r.v v9, v8
-; CHECK-NEXT:    vsetivli zero, 4, e8, m1, tu, ma
 ; CHECK-NEXT:    vmv.s.x v9, a0
 ; CHECK-NEXT:    vsetivli zero, 2, e8, mf8, ta, ma
 ; CHECK-NEXT:    vslideup.vi v9, v8, 1
diff --git a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-strided-load-store-asm.ll b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-strided-load-store-asm.ll
index 1a08c613ca36a..29d9a8a9b060c 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-strided-load-store-asm.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-strided-load-store-asm.ll
@@ -62,8 +62,8 @@ define void @gather_masked(ptr noalias nocapture %A, ptr noalias nocapture reado
 ; CHECK-NEXT:    li a4, 5
 ; CHECK-NEXT:  .LBB1_1: # %vector.body
 ; CHECK-NEXT:    # =>This Inner Loop Header: Depth=1
-; CHECK-NEXT:    vmv1r.v v9, v8
 ; CHECK-NEXT:    vsetvli zero, a3, e8, m1, ta, mu
+; CHECK-NEXT:    vmv1r.v v9, v8
 ; CHECK-NEXT:    vlse8.v v9, (a1), a4, v0.t
 ; CHECK-NEXT:    vle8.v v10, (a0)
 ; CHECK-NEXT:    vadd.vv v9, v10, v9
diff --git a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-strided-vpload.ll b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-strided-vpload.ll
index 1c2c90478a1f7..4b7f82f94f5e4 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-strided-vpload.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-strided-vpload.ll
@@ -542,6 +542,7 @@ declare <3 x double> @llvm.experimental.vp.strided.load.v3f64.p0.i32(ptr, i32, <
 define <32 x double> @strided_vpload_v32f64(ptr %ptr, i32 signext %stride, <32 x i1> %m, i32 zeroext %evl) nounwind {
 ; CHECK-LABEL: strided_vpload_v32f64:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-NEXT:    vmv1r.v v9, v0
 ; CHECK-NEXT:    li a4, 16
 ; CHECK-NEXT:    mv a3, a2
@@ -598,6 +599,7 @@ declare <32 x double> @llvm.experimental.vp.strided.load.v32f64.p0.i32(ptr, i32,
 define <33 x double> @strided_load_v33f64(ptr %ptr, i64 %stride, <33 x i1> %mask, i32 zeroext %evl) {
 ; CHECK-RV32-LABEL: strided_load_v33f64:
 ; CHECK-RV32:       # %bb.0:
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv1r.v v8, v0
 ; CHECK-RV32-NEXT:    li a5, 32
 ; CHECK-RV32-NEXT:    mv a3, a4
@@ -648,6 +650,7 @@ define <33 x double> @strided_load_v33f64(ptr %ptr, i64 %stride, <33 x i1> %mask
 ;
 ; CHECK-RV64-LABEL: strided_load_v33f64:
 ; CHECK-RV64:       # %bb.0:
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv1r.v v8, v0
 ; CHECK-RV64-NEXT:    li a5, 32
 ; CHECK-RV64-NEXT:    mv a4, a3
diff --git a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-trunc-vp.ll b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-trunc-vp.ll
index 12893ec55cda7..a91dee1cb245f 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-trunc-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-trunc-vp.ll
@@ -53,9 +53,9 @@ declare <128 x i7> @llvm.vp.trunc.v128i7.v128i16(<128 x i16>, <128 x i1>, i32)
 define <128 x i7> @vtrunc_v128i7_v128i16(<128 x i16> %a, <128 x i1> %m, i32 zeroext %vl) {
 ; CHECK-LABEL: vtrunc_v128i7_v128i16:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetivli zero, 8, e8, m1, ta, ma
 ; CHECK-NEXT:    vmv8r.v v24, v8
 ; CHECK-NEXT:    li a1, 64
-; CHECK-NEXT:    vsetivli zero, 8, e8, m1, ta, ma
 ; CHECK-NEXT:    vslidedown.vi v12, v0, 8
 ; CHECK-NEXT:    mv a2, a0
 ; CHECK-NEXT:    bltu a0, a1, .LBB4_2
@@ -231,6 +231,7 @@ define <128 x i32> @vtrunc_v128i32_v128i64(<128 x i64> %a, <128 x i1> %m, i32 ze
 ; CHECK-NEXT:    mul a2, a2, a3
 ; CHECK-NEXT:    sub sp, sp, a2
 ; CHECK-NEXT:    .cfi_escape 0x0f, 0x0e, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0xc8, 0x00, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 72 * vlenb
+; CHECK-NEXT:    vsetivli zero, 8, e8, m1, ta, ma
 ; CHECK-NEXT:    vmv1r.v v7, v0
 ; CHECK-NEXT:    csrr a2, vlenb
 ; CHECK-NEXT:    li a3, 24
@@ -243,7 +244,6 @@ define <128 x i32> @vtrunc_v128i32_v128i64(<128 x i64> %a, <128 x i1> %m, i32 ze
 ; CHECK-NEXT:    add a2, sp, a2
 ; CHECK-NEXT:    addi a2, a2, 16
 ; CHECK-NEXT:    vs8r.v v8, (a2) # Unknown-size Folded Spill
-; CHECK-NEXT:    vsetivli zero, 8, e8, m1, ta, ma
 ; CHECK-NEXT:    vslidedown.vi v6, v0, 8
 ; CHECK-NEXT:    addi a2, a1, 512
 ; CHECK-NEXT:    addi a3, a1, 640
@@ -541,9 +541,9 @@ declare <32 x i32> @llvm.vp.trunc.v32i32.v32i64(<32 x i64>, <32 x i1>, i32)
 define <32 x i32> @vtrunc_v32i32_v32i64(<32 x i64> %a, <32 x i1> %m, i32 zeroext %vl) {
 ; CHECK-LABEL: vtrunc_v32i32_v32i64:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetivli zero, 2, e8, mf4, ta, ma
 ; CHECK-NEXT:    vmv8r.v v24, v8
 ; CHECK-NEXT:    li a2, 16
-; CHECK-NEXT:    vsetivli zero, 2, e8, mf4, ta, ma
 ; CHECK-NEXT:    vslidedown.vi v12, v0, 2
 ; CHECK-NEXT:    mv a1, a0
 ; CHECK-NEXT:    bltu a0, a2, .LBB17_2
diff --git a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-unaligned.ll b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-unaligned.ll
index db03dc3d5ab1e..6d9f69f436fc4 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-unaligned.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-unaligned.ll
@@ -80,14 +80,8 @@ define <2 x i16> @mgather_v2i16_align1(<2 x ptr> %ptrs, <2 x i1> %m, <2 x i16> %
 ; RV32-SLOW-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; RV32-SLOW-NEXT:    vmv.x.s a0, v0
 ; RV32-SLOW-NEXT:    andi a1, a0, 1
-; RV32-SLOW-NEXT:    bnez a1, .LBB4_3
-; RV32-SLOW-NEXT:  # %bb.1: # %else
-; RV32-SLOW-NEXT:    andi a0, a0, 2
-; RV32-SLOW-NEXT:    bnez a0, .LBB4_4
-; RV32-SLOW-NEXT:  .LBB4_2: # %else2
-; RV32-SLOW-NEXT:    vmv1r.v v8, v9
-; RV32-SLOW-NEXT:    ret
-; RV32-SLOW-NEXT:  .LBB4_3: # %cond.load
+; RV32-SLOW-NEXT:    beqz a1, .LBB4_2
+; RV32-SLOW-NEXT:  # %bb.1: # %cond.load
 ; RV32-SLOW-NEXT:    vsetvli zero, zero, e32, m4, ta, ma
 ; RV32-SLOW-NEXT:    vmv.x.s a1, v8
 ; RV32-SLOW-NEXT:    lbu a2, 1(a1)
@@ -96,9 +90,10 @@ define <2 x i16> @mgather_v2i16_align1(<2 x ptr> %ptrs, <2 x i1> %m, <2 x i16> %
 ; RV32-SLOW-NEXT:    or a1, a2, a1
 ; RV32-SLOW-NEXT:    vsetvli zero, zero, e16, m2, tu, ma
 ; RV32-SLOW-NEXT:    vmv.s.x v9, a1
+; RV32-SLOW-NEXT:  .LBB4_2: # %else
 ; RV32-SLOW-NEXT:    andi a0, a0, 2
-; RV32-SLOW-NEXT:    beqz a0, .LBB4_2
-; RV32-SLOW-NEXT:  .LBB4_4: # %cond.load1
+; RV32-SLOW-NEXT:    beqz a0, .LBB4_4
+; RV32-SLOW-NEXT:  # %bb.3: # %cond.load1
 ; RV32-SLOW-NEXT:    vsetivli zero, 1, e32, mf2, ta, ma
 ; RV32-SLOW-NEXT:    vslidedown.vi v8, v8, 1
 ; RV32-SLOW-NEXT:    vmv.x.s a0, v8
@@ -109,6 +104,8 @@ define <2 x i16> @mgather_v2i16_align1(<2 x ptr> %ptrs, <2 x i1> %m, <2 x i16> %
 ; RV32-SLOW-NEXT:    vmv.s.x v8, a0
 ; RV32-SLOW-NEXT:    vsetivli zero, 2, e16, mf4, ta, ma
 ; RV32-SLOW-NEXT:    vslideup.vi v9, v8, 1
+; RV32-SLOW-NEXT:  .LBB4_4: # %else2
+; RV32-SLOW-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; RV32-SLOW-NEXT:    vmv1r.v v8, v9
 ; RV32-SLOW-NEXT:    ret
 ;
@@ -117,14 +114,8 @@ define <2 x i16> @mgather_v2i16_align1(<2 x ptr> %ptrs, <2 x i1> %m, <2 x i16> %
 ; RV64-SLOW-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; RV64-SLOW-NEXT:    vmv.x.s a0, v0
 ; RV64-SLOW-NEXT:    andi a1, a0, 1
-; RV64-SLOW-NEXT:    bnez a1, .LBB4_3
-; RV64-SLOW-NEXT:  # %bb.1: # %else
-; RV64-SLOW-NEXT:    andi a0, a0, 2
-; RV64-SLOW-NEXT:    bnez a0, .LBB4_4
-; RV64-SLOW-NEXT:  .LBB4_2: # %else2
-; RV64-SLOW-NEXT:    vmv1r.v v8, v9
-; RV64-SLOW-NEXT:    ret
-; RV64-SLOW-NEXT:  .LBB4_3: # %cond.load
+; RV64-SLOW-NEXT:    beqz a1, .LBB4_2
+; RV64-SLOW-NEXT:  # %bb.1: # %cond.load
 ; RV64-SLOW-NEXT:    vsetvli zero, zero, e64, m8, ta, ma
 ; RV64-SLOW-NEXT:    vmv.x.s a1, v8
 ; RV64-SLOW-NEXT:    lbu a2, 1(a1)
@@ -133,9 +124,10 @@ define <2 x i16> @mgather_v2i16_align1(<2 x ptr> %ptrs, <2 x i1> %m, <2 x i16> %
 ; RV64-SLOW-NEXT:    or a1, a2, a1
 ; RV64-SLOW-NEXT:    vsetvli zero, zero, e16, m2, tu, ma
 ; RV64-SLOW-NEXT:    vmv.s.x v9, a1
+; RV64-SLOW-NEXT:  .LBB4_2: # %else
 ; RV64-SLOW-NEXT:    andi a0, a0, 2
-; RV64-SLOW-NEXT:    beqz a0, .LBB4_2
-; RV64-SLOW-NEXT:  .LBB4_4: # %cond.load1
+; RV64-SLOW-NEXT:    beqz a0, .LBB4_4
+; RV64-SLOW-NEXT:  # %bb.3: # %cond.load1
 ; RV64-SLOW-NEXT:    vsetivli zero, 1, e64, m1, ta, ma
 ; RV64-SLOW-NEXT:    vslidedown.vi v8, v8, 1
 ; RV64-SLOW-NEXT:    vmv.x.s a0, v8
@@ -146,6 +138,8 @@ define <2 x i16> @mgather_v2i16_align1(<2 x ptr> %ptrs, <2 x i1> %m, <2 x i16> %
 ; RV64-SLOW-NEXT:    vmv.s.x v8, a0
 ; RV64-SLOW-NEXT:    vsetivli zero, 2, e16, mf4, ta, ma
 ; RV64-SLOW-NEXT:    vslideup.vi v9, v8, 1
+; RV64-SLOW-NEXT:  .LBB4_4: # %else2
+; RV64-SLOW-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; RV64-SLOW-NEXT:    vmv1r.v v8, v9
 ; RV64-SLOW-NEXT:    ret
 ;
@@ -174,23 +168,18 @@ define <2 x i64> @mgather_v2i64_align4(<2 x ptr> %ptrs, <2 x i1> %m, <2 x i64> %
 ; RV32-SLOW-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; RV32-SLOW-NEXT:    vmv.x.s a0, v0
 ; RV32-SLOW-NEXT:    andi a1, a0, 1
-; RV32-SLOW-NEXT:    bnez a1, .LBB5_3
-; RV32-SLOW-NEXT:  # %bb.1: # %else
-; RV32-SLOW-NEXT:    andi a0, a0, 2
-; RV32-SLOW-NEXT:    bnez a0, .LBB5_4
-; RV32-SLOW-NEXT:  .LBB5_2: # %else2
-; RV32-SLOW-NEXT:    vmv1r.v v8, v9
-; RV32-SLOW-NEXT:    ret
-; RV32-SLOW-NEXT:  .LBB5_3: # %cond.load
+; RV32-SLOW-NEXT:    beqz a1, .LBB5_2
+; RV32-SLOW-NEXT:  # %bb.1: # %cond.load
 ; RV32-SLOW-NEXT:    vsetivli zero, 2, e32, m1, tu, ma
 ; RV32-SLOW-NEXT:    vmv.x.s a1, v8
 ; RV32-SLOW-NEXT:    lw a2, 0(a1)
 ; RV32-SLOW-NEXT:    lw a1, 4(a1)
 ; RV32-SLOW-NEXT:    vslide1down.vx v9, v9, a2
 ; RV32-SLOW-NEXT:    vslide1down.vx v9, v9, a1
+; RV32-SLOW-NEXT:  .LBB5_2: # %else
 ; RV32-SLOW-NEXT:    andi a0, a0, 2
-; RV32-SLOW-NEXT:    beqz a0, .LBB5_2
-; RV32-SLOW-NEXT:  .LBB5_4: # %cond.load1
+; RV32-SLOW-NEXT:    beqz a0, .LBB5_4
+; RV32-SLOW-NEXT:  # %bb.3: # %cond.load1
 ; RV32-SLOW-NEXT:    vsetivli zero, 1, e32, mf2, ta, ma
 ; RV32-SLOW-NEXT:    vslidedown.vi v8, v8, 1
 ; RV32-SLOW-NEXT:    vmv.x.s a0, v8
@@ -201,6 +190,8 @@ define <2 x i64> @mgather_v2i64_align4(<2 x ptr> %ptrs, <2 x i1> %m, <2 x i64> %
 ; RV32-SLOW-NEXT:    vslide1down.vx v8, v8, a0
 ; RV32-SLOW-NEXT:    vsetivli zero, 2, e64, m1, ta, ma
 ; RV32-SLOW-NEXT:    vslideup.vi v9, v8, 1
+; RV32-SLOW-NEXT:  .LBB5_4: # %else2
+; RV32-SLOW-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; RV32-SLOW-NEXT:    vmv1r.v v8, v9
 ; RV32-SLOW-NEXT:    ret
 ;
@@ -209,14 +200,8 @@ define <2 x i64> @mgather_v2i64_align4(<2 x ptr> %ptrs, <2 x i1> %m, <2 x i64> %
 ; RV64-SLOW-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; RV64-SLOW-NEXT:    vmv.x.s a0, v0
 ; RV64-SLOW-NEXT:    andi a1, a0, 1
-; RV64-SLOW-NEXT:    bnez a1, .LBB5_3
-; RV64-SLOW-NEXT:  # %bb.1: # %else
-; RV64-SLOW-NEXT:    andi a0, a0, 2
-; RV64-SLOW-NEXT:    bnez a0, .LBB5_4
-; RV64-SLOW-NEXT:  .LBB5_2: # %else2
-; RV64-SLOW-NEXT:    vmv1r.v v8, v9
-; RV64-SLOW-NEXT:    ret
-; RV64-SLOW-NEXT:  .LBB5_3: # %cond.load
+; RV64-SLOW-NEXT:    beqz a1, .LBB5_2
+; RV64-SLOW-NEXT:  # %bb.1: # %cond.load
 ; RV64-SLOW-NEXT:    vsetvli zero, zero, e64, m8, tu, ma
 ; RV64-SLOW-NEXT:    vmv.x.s a1, v8
 ; RV64-SLOW-NEXT:    lwu a2, 4(a1)
@@ -224,9 +209,10 @@ define <2 x i64> @mgather_v2i64_align4(<2 x ptr> %ptrs, <2 x i1> %m, <2 x i64> %
 ; RV64-SLOW-NEXT:    slli a2, a2, 32
 ; RV64-SLOW-NEXT:    or a1, a2, a1
 ; RV64-SLOW-NEXT:    vmv.s.x v9, a1
+; RV64-SLOW-NEXT:  .LBB5_2: # %else
 ; RV64-SLOW-NEXT:    andi a0, a0, 2
-; RV64-SLOW-NEXT:    beqz a0, .LBB5_2
-; RV64-SLOW-NEXT:  .LBB5_4: # %cond.load1
+; RV64-SLOW-NEXT:    beqz a0, .LBB5_4
+; RV64-SLOW-NEXT:  # %bb.3: # %cond.load1
 ; RV64-SLOW-NEXT:    vsetivli zero, 2, e64, m1, ta, ma
 ; RV64-SLOW-NEXT:    vslidedown.vi v8, v8, 1
 ; RV64-SLOW-NEXT:    vmv.x.s a0, v8
@@ -236,6 +222,8 @@ define <2 x i64> @mgather_v2i64_align4(<2 x ptr> %ptrs, <2 x i1> %m, <2 x i64> %
 ; RV64-SLOW-NEXT:    or a0, a1, a0
 ; RV64-SLOW-NEXT:    vmv.s.x v8, a0
 ; RV64-SLOW-NEXT:    vslideup.vi v9, v8, 1
+; RV64-SLOW-NEXT:  .LBB5_4: # %else2
+; RV64-SLOW-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; RV64-SLOW-NEXT:    vmv1r.v v8, v9
 ; RV64-SLOW-NEXT:    ret
 ;
diff --git a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vadd-vp.ll b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vadd-vp.ll
index 5be1a771eb279..7ee8179acfdb9 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vadd-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vadd-vp.ll
@@ -363,6 +363,7 @@ declare <256 x i8> @llvm.vp.add.v258i8(<256 x i8>, <256 x i8>, <256 x i1>, i32)
 define <256 x i8> @vadd_vi_v258i8(<256 x i8> %va, <256 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vadd_vi_v258i8:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-NEXT:    vmv1r.v v24, v0
 ; CHECK-NEXT:    li a2, 128
 ; CHECK-NEXT:    vsetvli zero, a2, e8, m8, ta, ma
diff --git a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vmax-vp.ll b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vmax-vp.ll
index ac48542ca9ebb..fec54b36042fa 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vmax-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vmax-vp.ll
@@ -267,6 +267,7 @@ declare <256 x i8> @llvm.vp.smax.v258i8(<256 x i8>, <256 x i8>, <256 x i1>, i32)
 define <256 x i8> @vmax_vx_v258i8(<256 x i8> %va, i8 %b, <256 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vmax_vx_v258i8:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-NEXT:    vmv1r.v v24, v0
 ; CHECK-NEXT:    li a3, 128
 ; CHECK-NEXT:    vsetvli zero, a3, e8, m8, ta, ma
diff --git a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vmaxu-vp.ll b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vmaxu-vp.ll
index 794eef6ed40b2..7ca0dbd9adffc 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vmaxu-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vmaxu-vp.ll
@@ -266,6 +266,7 @@ declare <256 x i8> @llvm.vp.umax.v258i8(<256 x i8>, <256 x i8>, <256 x i1>, i32)
 define <256 x i8> @vmaxu_vx_v258i8(<256 x i8> %va, i8 %b, <256 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vmaxu_vx_v258i8:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-NEXT:    vmv1r.v v24, v0
 ; CHECK-NEXT:    li a3, 128
 ; CHECK-NEXT:    vsetvli zero, a3, e8, m8, ta, ma
diff --git a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vmin-vp.ll b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vmin-vp.ll
index 34011f6bd8acd..ea75742ca6e43 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vmin-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vmin-vp.ll
@@ -267,6 +267,7 @@ declare <256 x i8> @llvm.vp.smin.v258i8(<256 x i8>, <256 x i8>, <256 x i1>, i32)
 define <256 x i8> @vmin_vx_v258i8(<256 x i8> %va, i8 %b, <256 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vmin_vx_v258i8:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-NEXT:    vmv1r.v v24, v0
 ; CHECK-NEXT:    li a3, 128
 ; CHECK-NEXT:    vsetvli zero, a3, e8, m8, ta, ma
diff --git a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vminu-vp.ll b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vminu-vp.ll
index 79e72b7d9cac9..f4f54db64018d 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vminu-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vminu-vp.ll
@@ -266,6 +266,7 @@ declare <256 x i8> @llvm.vp.umin.v258i8(<256 x i8>, <256 x i8>, <256 x i1>, i32)
 define <256 x i8> @vminu_vx_v258i8(<256 x i8> %va, i8 %b, <256 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vminu_vx_v258i8:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-NEXT:    vmv1r.v v24, v0
 ; CHECK-NEXT:    li a3, 128
 ; CHECK-NEXT:    vsetvli zero, a3, e8, m8, ta, ma
diff --git a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vpgather.ll b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vpgather.ll
index 24e75cde2ce91..df9ff0fc39a7e 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vpgather.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vpgather.ll
@@ -2617,8 +2617,8 @@ define <32 x double> @vpgather_baseidx_zext_v32i32_v32f64(ptr %base, <32 x i32>
 define <32 x double> @vpgather_baseidx_v32f64(ptr %base, <32 x i64> %idxs, <32 x i1> %m, i32 zeroext %evl) {
 ; RV32-LABEL: vpgather_baseidx_v32f64:
 ; RV32:       # %bb.0:
-; RV32-NEXT:    vmv1r.v v7, v0
 ; RV32-NEXT:    vsetivli zero, 16, e32, m4, ta, ma
+; RV32-NEXT:    vmv1r.v v7, v0
 ; RV32-NEXT:    vnsrl.wi v24, v16, 0
 ; RV32-NEXT:    vnsrl.wi v16, v8, 0
 ; RV32-NEXT:    li a2, 32
diff --git a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vpload.ll b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vpload.ll
index 71f497e4c7be4..6c9989775f790 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vpload.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vpload.ll
@@ -394,6 +394,7 @@ declare <33 x double> @llvm.vp.load.v33f64.p0(ptr, <33 x i1>, i32)
 define <33 x double> @vpload_v33f64(ptr %ptr, <33 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vpload_v33f64:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-NEXT:    vmv1r.v v8, v0
 ; CHECK-NEXT:    li a4, 32
 ; CHECK-NEXT:    mv a3, a2
diff --git a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vpmerge.ll b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vpmerge.ll
index a11c2b6bca12e..a53d33e6120d5 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vpmerge.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vpmerge.ll
@@ -1181,9 +1181,9 @@ define <32 x double> @vpmerge_vv_v32f64(<32 x double> %va, <32 x double> %vb, <3
 ; CHECK-NEXT:    .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x08, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 8 * vlenb
 ; CHECK-NEXT:    addi a1, sp, 16
 ; CHECK-NEXT:    vs8r.v v16, (a1) # Unknown-size Folded Spill
+; CHECK-NEXT:    vsetivli zero, 16, e64, m8, ta, ma
 ; CHECK-NEXT:    vmv8r.v v16, v8
 ; CHECK-NEXT:    addi a1, a0, 128
-; CHECK-NEXT:    vsetivli zero, 16, e64, m8, ta, ma
 ; CHECK-NEXT:    vle64.v v24, (a1)
 ; CHECK-NEXT:    vle64.v v8, (a0)
 ; CHECK-NEXT:    li a1, 16
diff --git a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vsadd-vp.ll b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vsadd-vp.ll
index 888fc79f0122d..7afd31fdd663c 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vsadd-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vsadd-vp.ll
@@ -372,6 +372,7 @@ declare <256 x i8> @llvm.vp.sadd.sat.v258i8(<256 x i8>, <256 x i8>, <256 x i1>,
 define <256 x i8> @vsadd_vi_v258i8(<256 x i8> %va, <256 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vsadd_vi_v258i8:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-NEXT:    vmv1r.v v24, v0
 ; CHECK-NEXT:    li a2, 128
 ; CHECK-NEXT:    vsetvli zero, a2, e8, m8, ta, ma
diff --git a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vsaddu-vp.ll b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vsaddu-vp.ll
index e1d57ae1e6741..f61b112fd8024 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vsaddu-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vsaddu-vp.ll
@@ -368,6 +368,7 @@ declare <256 x i8> @llvm.vp.uadd.sat.v258i8(<256 x i8>, <256 x i8>, <256 x i1>,
 define <256 x i8> @vsaddu_vi_v258i8(<256 x i8> %va, <256 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vsaddu_vi_v258i8:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-NEXT:    vmv1r.v v24, v0
 ; CHECK-NEXT:    li a2, 128
 ; CHECK-NEXT:    vsetvli zero, a2, e8, m8, ta, ma
diff --git a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vselect-vp.ll b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vselect-vp.ll
index 1d8af4c46cc07..5d407caf71514 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vselect-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vselect-vp.ll
@@ -163,6 +163,7 @@ define <256 x i8> @select_v256i8(<256 x i1> %a, <256 x i8> %b, <256 x i8> %c, i3
 ; CHECK-NEXT:    .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x08, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 8 * vlenb
 ; CHECK-NEXT:    addi a2, sp, 16
 ; CHECK-NEXT:    vs8r.v v16, (a2) # Unknown-size Folded Spill
+; CHECK-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-NEXT:    vmv1r.v v6, v8
 ; CHECK-NEXT:    vmv1r.v v7, v0
 ; CHECK-NEXT:    li a2, 128
diff --git a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vssub-vp.ll b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vssub-vp.ll
index 8fad3db55f9bc..6ddf2e464750e 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vssub-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vssub-vp.ll
@@ -384,6 +384,7 @@ declare <256 x i8> @llvm.vp.ssub.sat.v258i8(<256 x i8>, <256 x i8>, <256 x i1>,
 define <256 x i8> @vssub_vi_v258i8(<256 x i8> %va, <256 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vssub_vi_v258i8:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-NEXT:    vmv1r.v v24, v0
 ; CHECK-NEXT:    li a2, 128
 ; CHECK-NEXT:    addi a3, a1, -128
diff --git a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vssubu-vp.ll b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vssubu-vp.ll
index ca35aa6c4a94c..c403593894794 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vssubu-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-vssubu-vp.ll
@@ -379,6 +379,7 @@ declare <256 x i8> @llvm.vp.usub.sat.v258i8(<256 x i8>, <256 x i8>, <256 x i1>,
 define <256 x i8> @vssubu_vi_v258i8(<256 x i8> %va, <256 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vssubu_vi_v258i8:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-NEXT:    vmv1r.v v24, v0
 ; CHECK-NEXT:    li a2, 128
 ; CHECK-NEXT:    addi a3, a1, -128
diff --git a/llvm/test/CodeGen/RISCV/rvv/floor-vp.ll b/llvm/test/CodeGen/RISCV/rvv/floor-vp.ll
index e6dfe5e78cdb4..74a00c655d526 100644
--- a/llvm/test/CodeGen/RISCV/rvv/floor-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/floor-vp.ll
@@ -117,8 +117,8 @@ declare <vscale x 4 x bfloat> @llvm.vp.floor.nxv4bf16(<vscale x 4 x bfloat>, <vs
 define <vscale x 4 x bfloat> @vp_floor_nxv4bf16(<vscale x 4 x bfloat> %va, <vscale x 4 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vp_floor_nxv4bf16:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v9, v0
 ; CHECK-NEXT:    vsetvli a1, zero, e16, m1, ta, ma
+; CHECK-NEXT:    vmv1r.v v9, v0
 ; CHECK-NEXT:    vfwcvtbf16.f.f.v v10, v8
 ; CHECK-NEXT:    lui a1, 307200
 ; CHECK-NEXT:    vsetvli zero, a0, e32, m2, ta, ma
@@ -169,8 +169,8 @@ declare <vscale x 8 x bfloat> @llvm.vp.floor.nxv8bf16(<vscale x 8 x bfloat>, <vs
 define <vscale x 8 x bfloat> @vp_floor_nxv8bf16(<vscale x 8 x bfloat> %va, <vscale x 8 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vp_floor_nxv8bf16:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vsetvli a1, zero, e16, m2, ta, ma
+; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vfwcvtbf16.f.f.v v12, v8
 ; CHECK-NEXT:    lui a1, 307200
 ; CHECK-NEXT:    vsetvli zero, a0, e32, m4, ta, ma
@@ -221,8 +221,8 @@ declare <vscale x 16 x bfloat> @llvm.vp.floor.nxv16bf16(<vscale x 16 x bfloat>,
 define <vscale x 16 x bfloat> @vp_floor_nxv16bf16(<vscale x 16 x bfloat> %va, <vscale x 16 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vp_floor_nxv16bf16:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v12, v0
 ; CHECK-NEXT:    vsetvli a1, zero, e16, m4, ta, ma
+; CHECK-NEXT:    vmv1r.v v12, v0
 ; CHECK-NEXT:    vfwcvtbf16.f.f.v v16, v8
 ; CHECK-NEXT:    lui a1, 307200
 ; CHECK-NEXT:    vsetvli zero, a0, e32, m8, ta, ma
@@ -279,9 +279,9 @@ define <vscale x 32 x bfloat> @vp_floor_nxv32bf16(<vscale x 32 x bfloat> %va, <v
 ; CHECK-NEXT:    slli a1, a1, 3
 ; CHECK-NEXT:    sub sp, sp, a1
 ; CHECK-NEXT:    .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x08, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 8 * vlenb
+; CHECK-NEXT:    vsetvli a1, zero, e16, m4, ta, ma
 ; CHECK-NEXT:    vmv1r.v v7, v0
 ; CHECK-NEXT:    csrr a2, vlenb
-; CHECK-NEXT:    vsetvli a1, zero, e16, m4, ta, ma
 ; CHECK-NEXT:    vfwcvtbf16.f.f.v v24, v12
 ; CHECK-NEXT:    lui a3, 307200
 ; CHECK-NEXT:    slli a1, a2, 1
@@ -582,8 +582,8 @@ define <vscale x 4 x half> @vp_floor_nxv4f16(<vscale x 4 x half> %va, <vscale x
 ;
 ; ZVFHMIN-LABEL: vp_floor_nxv4f16:
 ; ZVFHMIN:       # %bb.0:
-; ZVFHMIN-NEXT:    vmv1r.v v9, v0
 ; ZVFHMIN-NEXT:    vsetvli a1, zero, e16, m1, ta, ma
+; ZVFHMIN-NEXT:    vmv1r.v v9, v0
 ; ZVFHMIN-NEXT:    vfwcvt.f.f.v v10, v8
 ; ZVFHMIN-NEXT:    lui a1, 307200
 ; ZVFHMIN-NEXT:    vsetvli zero, a0, e32, m2, ta, ma
@@ -649,10 +649,10 @@ declare <vscale x 8 x half> @llvm.vp.floor.nxv8f16(<vscale x 8 x half>, <vscale
 define <vscale x 8 x half> @vp_floor_nxv8f16(<vscale x 8 x half> %va, <vscale x 8 x i1> %m, i32 zeroext %evl) {
 ; ZVFH-LABEL: vp_floor_nxv8f16:
 ; ZVFH:       # %bb.0:
-; ZVFH-NEXT:    vmv1r.v v10, v0
-; ZVFH-NEXT:    lui a1, %hi(.LCPI18_0)
-; ZVFH-NEXT:    flh fa5, %lo(.LCPI18_0)(a1)
 ; ZVFH-NEXT:    vsetvli zero, a0, e16, m2, ta, ma
+; ZVFH-NEXT:    vmv1r.v v10, v0
+; ZVFH-NEXT:    lui a0, %hi(.LCPI18_0)
+; ZVFH-NEXT:    flh fa5, %lo(.LCPI18_0)(a0)
 ; ZVFH-NEXT:    vfabs.v v12, v8, v0.t
 ; ZVFH-NEXT:    vsetvli zero, zero, e16, m2, ta, mu
 ; ZVFH-NEXT:    vmflt.vf v10, v12, fa5, v0.t
@@ -668,8 +668,8 @@ define <vscale x 8 x half> @vp_floor_nxv8f16(<vscale x 8 x half> %va, <vscale x
 ;
 ; ZVFHMIN-LABEL: vp_floor_nxv8f16:
 ; ZVFHMIN:       # %bb.0:
-; ZVFHMIN-NEXT:    vmv1r.v v10, v0
 ; ZVFHMIN-NEXT:    vsetvli a1, zero, e16, m2, ta, ma
+; ZVFHMIN-NEXT:    vmv1r.v v10, v0
 ; ZVFHMIN-NEXT:    vfwcvt.f.f.v v12, v8
 ; ZVFHMIN-NEXT:    lui a1, 307200
 ; ZVFHMIN-NEXT:    vsetvli zero, a0, e32, m4, ta, ma
@@ -735,10 +735,10 @@ declare <vscale x 16 x half> @llvm.vp.floor.nxv16f16(<vscale x 16 x half>, <vsca
 define <vscale x 16 x half> @vp_floor_nxv16f16(<vscale x 16 x half> %va, <vscale x 16 x i1> %m, i32 zeroext %evl) {
 ; ZVFH-LABEL: vp_floor_nxv16f16:
 ; ZVFH:       # %bb.0:
-; ZVFH-NEXT:    vmv1r.v v12, v0
-; ZVFH-NEXT:    lui a1, %hi(.LCPI20_0)
-; ZVFH-NEXT:    flh fa5, %lo(.LCPI20_0)(a1)
 ; ZVFH-NEXT:    vsetvli zero, a0, e16, m4, ta, ma
+; ZVFH-NEXT:    vmv1r.v v12, v0
+; ZVFH-NEXT:    lui a0, %hi(.LCPI20_0)
+; ZVFH-NEXT:    flh fa5, %lo(.LCPI20_0)(a0)
 ; ZVFH-NEXT:    vfabs.v v16, v8, v0.t
 ; ZVFH-NEXT:    vsetvli zero, zero, e16, m4, ta, mu
 ; ZVFH-NEXT:    vmflt.vf v12, v16, fa5, v0.t
@@ -754,8 +754,8 @@ define <vscale x 16 x half> @vp_floor_nxv16f16(<vscale x 16 x half> %va, <vscale
 ;
 ; ZVFHMIN-LABEL: vp_floor_nxv16f16:
 ; ZVFHMIN:       # %bb.0:
-; ZVFHMIN-NEXT:    vmv1r.v v12, v0
 ; ZVFHMIN-NEXT:    vsetvli a1, zero, e16, m4, ta, ma
+; ZVFHMIN-NEXT:    vmv1r.v v12, v0
 ; ZVFHMIN-NEXT:    vfwcvt.f.f.v v16, v8
 ; ZVFHMIN-NEXT:    lui a1, 307200
 ; ZVFHMIN-NEXT:    vsetvli zero, a0, e32, m8, ta, ma
@@ -821,10 +821,10 @@ declare <vscale x 32 x half> @llvm.vp.floor.nxv32f16(<vscale x 32 x half>, <vsca
 define <vscale x 32 x half> @vp_floor_nxv32f16(<vscale x 32 x half> %va, <vscale x 32 x i1> %m, i32 zeroext %evl) {
 ; ZVFH-LABEL: vp_floor_nxv32f16:
 ; ZVFH:       # %bb.0:
-; ZVFH-NEXT:    vmv1r.v v16, v0
-; ZVFH-NEXT:    lui a1, %hi(.LCPI22_0)
-; ZVFH-NEXT:    flh fa5, %lo(.LCPI22_0)(a1)
 ; ZVFH-NEXT:    vsetvli zero, a0, e16, m8, ta, ma
+; ZVFH-NEXT:    vmv1r.v v16, v0
+; ZVFH-NEXT:    lui a0, %hi(.LCPI22_0)
+; ZVFH-NEXT:    flh fa5, %lo(.LCPI22_0)(a0)
 ; ZVFH-NEXT:    vfabs.v v24, v8, v0.t
 ; ZVFH-NEXT:    vsetvli zero, zero, e16, m8, ta, mu
 ; ZVFH-NEXT:    vmflt.vf v16, v24, fa5, v0.t
@@ -846,9 +846,9 @@ define <vscale x 32 x half> @vp_floor_nxv32f16(<vscale x 32 x half> %va, <vscale
 ; ZVFHMIN-NEXT:    slli a1, a1, 3
 ; ZVFHMIN-NEXT:    sub sp, sp, a1
 ; ZVFHMIN-NEXT:    .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x08, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 8 * vlenb
+; ZVFHMIN-NEXT:    vsetvli a1, zero, e16, m4, ta, ma
 ; ZVFHMIN-NEXT:    vmv1r.v v7, v0
 ; ZVFHMIN-NEXT:    csrr a2, vlenb
-; ZVFHMIN-NEXT:    vsetvli a1, zero, e16, m4, ta, ma
 ; ZVFHMIN-NEXT:    vfwcvt.f.f.v v24, v12
 ; ZVFHMIN-NEXT:    lui a3, 307200
 ; ZVFHMIN-NEXT:    slli a1, a2, 1
@@ -1068,8 +1068,8 @@ declare <vscale x 4 x float> @llvm.vp.floor.nxv4f32(<vscale x 4 x float>, <vscal
 define <vscale x 4 x float> @vp_floor_nxv4f32(<vscale x 4 x float> %va, <vscale x 4 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vp_floor_nxv4f32:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e32, m2, ta, ma
+; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vfabs.v v12, v8, v0.t
 ; CHECK-NEXT:    lui a0, 307200
 ; CHECK-NEXT:    fmv.w.x fa5, a0
@@ -1112,8 +1112,8 @@ declare <vscale x 8 x float> @llvm.vp.floor.nxv8f32(<vscale x 8 x float>, <vscal
 define <vscale x 8 x float> @vp_floor_nxv8f32(<vscale x 8 x float> %va, <vscale x 8 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vp_floor_nxv8f32:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v12, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e32, m4, ta, ma
+; CHECK-NEXT:    vmv1r.v v12, v0
 ; CHECK-NEXT:    vfabs.v v16, v8, v0.t
 ; CHECK-NEXT:    lui a0, 307200
 ; CHECK-NEXT:    fmv.w.x fa5, a0
@@ -1156,8 +1156,8 @@ declare <vscale x 16 x float> @llvm.vp.floor.nxv16f32(<vscale x 16 x float>, <vs
 define <vscale x 16 x float> @vp_floor_nxv16f32(<vscale x 16 x float> %va, <vscale x 16 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vp_floor_nxv16f32:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v16, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e32, m8, ta, ma
+; CHECK-NEXT:    vmv1r.v v16, v0
 ; CHECK-NEXT:    vfabs.v v24, v8, v0.t
 ; CHECK-NEXT:    lui a0, 307200
 ; CHECK-NEXT:    fmv.w.x fa5, a0
@@ -1242,10 +1242,10 @@ declare <vscale x 2 x double> @llvm.vp.floor.nxv2f64(<vscale x 2 x double>, <vsc
 define <vscale x 2 x double> @vp_floor_nxv2f64(<vscale x 2 x double> %va, <vscale x 2 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vp_floor_nxv2f64:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v10, v0
-; CHECK-NEXT:    lui a1, %hi(.LCPI36_0)
-; CHECK-NEXT:    fld fa5, %lo(.LCPI36_0)(a1)
 ; CHECK-NEXT:    vsetvli zero, a0, e64, m2, ta, ma
+; CHECK-NEXT:    vmv1r.v v10, v0
+; CHECK-NEXT:    lui a0, %hi(.LCPI36_0)
+; CHECK-NEXT:    fld fa5, %lo(.LCPI36_0)(a0)
 ; CHECK-NEXT:    vfabs.v v12, v8, v0.t
 ; CHECK-NEXT:    vsetvli zero, zero, e64, m2, ta, mu
 ; CHECK-NEXT:    vmflt.vf v10, v12, fa5, v0.t
@@ -1286,10 +1286,10 @@ declare <vscale x 4 x double> @llvm.vp.floor.nxv4f64(<vscale x 4 x double>, <vsc
 define <vscale x 4 x double> @vp_floor_nxv4f64(<vscale x 4 x double> %va, <vscale x 4 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vp_floor_nxv4f64:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v12, v0
-; CHECK-NEXT:    lui a1, %hi(.LCPI38_0)
-; CHECK-NEXT:    fld fa5, %lo(.LCPI38_0)(a1)
 ; CHECK-NEXT:    vsetvli zero, a0, e64, m4, ta, ma
+; CHECK-NEXT:    vmv1r.v v12, v0
+; CHECK-NEXT:    lui a0, %hi(.LCPI38_0)
+; CHECK-NEXT:    fld fa5, %lo(.LCPI38_0)(a0)
 ; CHECK-NEXT:    vfabs.v v16, v8, v0.t
 ; CHECK-NEXT:    vsetvli zero, zero, e64, m4, ta, mu
 ; CHECK-NEXT:    vmflt.vf v12, v16, fa5, v0.t
@@ -1330,10 +1330,10 @@ declare <vscale x 7 x double> @llvm.vp.floor.nxv7f64(<vscale x 7 x double>, <vsc
 define <vscale x 7 x double> @vp_floor_nxv7f64(<vscale x 7 x double> %va, <vscale x 7 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vp_floor_nxv7f64:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v16, v0
-; CHECK-NEXT:    lui a1, %hi(.LCPI40_0)
-; CHECK-NEXT:    fld fa5, %lo(.LCPI40_0)(a1)
 ; CHECK-NEXT:    vsetvli zero, a0, e64, m8, ta, ma
+; CHECK-NEXT:    vmv1r.v v16, v0
+; CHECK-NEXT:    lui a0, %hi(.LCPI40_0)
+; CHECK-NEXT:    fld fa5, %lo(.LCPI40_0)(a0)
 ; CHECK-NEXT:    vfabs.v v24, v8, v0.t
 ; CHECK-NEXT:    vsetvli zero, zero, e64, m8, ta, mu
 ; CHECK-NEXT:    vmflt.vf v16, v24, fa5, v0.t
@@ -1374,10 +1374,10 @@ declare <vscale x 8 x double> @llvm.vp.floor.nxv8f64(<vscale x 8 x double>, <vsc
 define <vscale x 8 x double> @vp_floor_nxv8f64(<vscale x 8 x double> %va, <vscale x 8 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vp_floor_nxv8f64:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v16, v0
-; CHECK-NEXT:    lui a1, %hi(.LCPI42_0)
-; CHECK-NEXT:    fld fa5, %lo(.LCPI42_0)(a1)
 ; CHECK-NEXT:    vsetvli zero, a0, e64, m8, ta, ma
+; CHECK-NEXT:    vmv1r.v v16, v0
+; CHECK-NEXT:    lui a0, %hi(.LCPI42_0)
+; CHECK-NEXT:    fld fa5, %lo(.LCPI42_0)(a0)
 ; CHECK-NEXT:    vfabs.v v24, v8, v0.t
 ; CHECK-NEXT:    vsetvli zero, zero, e64, m8, ta, mu
 ; CHECK-NEXT:    vmflt.vf v16, v24, fa5, v0.t
@@ -1425,13 +1425,13 @@ define <vscale x 16 x double> @vp_floor_nxv16f64(<vscale x 16 x double> %va, <vs
 ; CHECK-NEXT:    slli a1, a1, 3
 ; CHECK-NEXT:    sub sp, sp, a1
 ; CHECK-NEXT:    .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x08, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 8 * vlenb
+; CHECK-NEXT:    vsetvli a1, zero, e8, mf4, ta, ma
 ; CHECK-NEXT:    vmv1r.v v7, v0
 ; CHECK-NEXT:    csrr a1, vlenb
 ; CHECK-NEXT:    lui a2, %hi(.LCPI44_0)
 ; CHECK-NEXT:    srli a3, a1, 3
 ; CHECK-NEXT:    fld fa5, %lo(.LCPI44_0)(a2)
 ; CHECK-NEXT:    sub a2, a0, a1
-; CHECK-NEXT:    vsetvli a4, zero, e8, mf4, ta, ma
 ; CHECK-NEXT:    vslidedown.vx v6, v0, a3
 ; CHECK-NEXT:    sltu a3, a0, a2
 ; CHECK-NEXT:    addi a3, a3, -1
diff --git a/llvm/test/CodeGen/RISCV/rvv/fmaximum-sdnode.ll b/llvm/test/CodeGen/RISCV/rvv/fmaximum-sdnode.ll
index a1cdbd4be2579..8b527fb152d68 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fmaximum-sdnode.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fmaximum-sdnode.ll
@@ -157,6 +157,7 @@ define <vscale x 32 x bfloat> @vfmax_nxv32bf16_vv(<vscale x 32 x bfloat> %a, <vs
 ; ZVFH-NEXT:    slli a0, a0, 1
 ; ZVFH-NEXT:    add a0, a0, a1
 ; ZVFH-NEXT:    sub sp, sp, a0
+; ZVFH-NEXT:    vsetvli a0, zero, e16, m4, ta, ma
 ; ZVFH-NEXT:    vmv8r.v v24, v16
 ; ZVFH-NEXT:    csrr a0, vlenb
 ; ZVFH-NEXT:    slli a0, a0, 3
@@ -164,7 +165,6 @@ define <vscale x 32 x bfloat> @vfmax_nxv32bf16_vv(<vscale x 32 x bfloat> %a, <vs
 ; ZVFH-NEXT:    addi a0, a0, 16
 ; ZVFH-NEXT:    vs8r.v v16, (a0) # Unknown-size Folded Spill
 ; ZVFH-NEXT:    vmv8r.v v0, v8
-; ZVFH-NEXT:    vsetvli a0, zero, e16, m4, ta, ma
 ; ZVFH-NEXT:    vfwcvtbf16.f.f.v v16, v24
 ; ZVFH-NEXT:    vfwcvtbf16.f.f.v v8, v0
 ; ZVFH-NEXT:    vsetvli zero, zero, e32, m8, ta, ma
@@ -232,6 +232,7 @@ define <vscale x 32 x bfloat> @vfmax_nxv32bf16_vv(<vscale x 32 x bfloat> %a, <vs
 ; ZVFHMIN-NEXT:    li a1, 24
 ; ZVFHMIN-NEXT:    mul a0, a0, a1
 ; ZVFHMIN-NEXT:    sub sp, sp, a0
+; ZVFHMIN-NEXT:    vsetvli a0, zero, e16, m4, ta, ma
 ; ZVFHMIN-NEXT:    vmv8r.v v24, v16
 ; ZVFHMIN-NEXT:    csrr a0, vlenb
 ; ZVFHMIN-NEXT:    slli a0, a0, 3
@@ -239,7 +240,6 @@ define <vscale x 32 x bfloat> @vfmax_nxv32bf16_vv(<vscale x 32 x bfloat> %a, <vs
 ; ZVFHMIN-NEXT:    addi a0, a0, 16
 ; ZVFHMIN-NEXT:    vs8r.v v16, (a0) # Unknown-size Folded Spill
 ; ZVFHMIN-NEXT:    vmv8r.v v0, v8
-; ZVFHMIN-NEXT:    vsetvli a0, zero, e16, m4, ta, ma
 ; ZVFHMIN-NEXT:    vfwcvtbf16.f.f.v v16, v24
 ; ZVFHMIN-NEXT:    vfwcvtbf16.f.f.v v8, v0
 ; ZVFHMIN-NEXT:    vsetvli zero, zero, e32, m8, ta, ma
@@ -501,6 +501,7 @@ define <vscale x 32 x half> @vfmax_nxv32f16_vv(<vscale x 32 x half> %a, <vscale
 ; ZVFHMIN-NEXT:    li a1, 24
 ; ZVFHMIN-NEXT:    mul a0, a0, a1
 ; ZVFHMIN-NEXT:    sub sp, sp, a0
+; ZVFHMIN-NEXT:    vsetvli a0, zero, e16, m4, ta, ma
 ; ZVFHMIN-NEXT:    vmv8r.v v24, v16
 ; ZVFHMIN-NEXT:    csrr a0, vlenb
 ; ZVFHMIN-NEXT:    slli a0, a0, 3
@@ -508,7 +509,6 @@ define <vscale x 32 x half> @vfmax_nxv32f16_vv(<vscale x 32 x half> %a, <vscale
 ; ZVFHMIN-NEXT:    addi a0, a0, 16
 ; ZVFHMIN-NEXT:    vs8r.v v16, (a0) # Unknown-size Folded Spill
 ; ZVFHMIN-NEXT:    vmv8r.v v0, v8
-; ZVFHMIN-NEXT:    vsetvli a0, zero, e16, m4, ta, ma
 ; ZVFHMIN-NEXT:    vfwcvt.f.f.v v16, v24
 ; ZVFHMIN-NEXT:    vfwcvt.f.f.v v8, v0
 ; ZVFHMIN-NEXT:    vsetvli zero, zero, e32, m8, ta, ma
diff --git a/llvm/test/CodeGen/RISCV/rvv/fmaximum-vp.ll b/llvm/test/CodeGen/RISCV/rvv/fmaximum-vp.ll
index 33e793a691b81..7649d60def111 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fmaximum-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fmaximum-vp.ll
@@ -19,8 +19,8 @@ declare <vscale x 1 x bfloat> @llvm.vp.maximum.nxv1bf16(<vscale x 1 x bfloat>, <
 define <vscale x 1 x bfloat> @vfmax_vv_nxv1bf16(<vscale x 1 x bfloat> %va, <vscale x 1 x bfloat> %vb, <vscale x 1 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vfmax_vv_nxv1bf16:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vsetvli a1, zero, e16, mf4, ta, ma
+; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vfwcvtbf16.f.f.v v11, v8
 ; CHECK-NEXT:    vsetvli zero, a0, e32, mf2, ta, ma
 ; CHECK-NEXT:    vmfeq.vv v0, v11, v11, v0.t
@@ -66,8 +66,8 @@ declare <vscale x 2 x bfloat> @llvm.vp.maximum.nxv2bf16(<vscale x 2 x bfloat>, <
 define <vscale x 2 x bfloat> @vfmax_vv_nxv2bf16(<vscale x 2 x bfloat> %va, <vscale x 2 x bfloat> %vb, <vscale x 2 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vfmax_vv_nxv2bf16:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vsetvli a1, zero, e16, mf2, ta, ma
+; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vfwcvtbf16.f.f.v v11, v8
 ; CHECK-NEXT:    vsetvli zero, a0, e32, m1, ta, ma
 ; CHECK-NEXT:    vmfeq.vv v0, v11, v11, v0.t
@@ -113,8 +113,8 @@ declare <vscale x 4 x bfloat> @llvm.vp.maximum.nxv4bf16(<vscale x 4 x bfloat>, <
 define <vscale x 4 x bfloat> @vfmax_vv_nxv4bf16(<vscale x 4 x bfloat> %va, <vscale x 4 x bfloat> %vb, <vscale x 4 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vfmax_vv_nxv4bf16:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vsetvli a1, zero, e16, m1, ta, ma
+; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vfwcvtbf16.f.f.v v12, v8
 ; CHECK-NEXT:    vsetvli zero, a0, e32, m2, ta, ma
 ; CHECK-NEXT:    vmfeq.vv v8, v12, v12, v0.t
@@ -162,8 +162,8 @@ declare <vscale x 8 x bfloat> @llvm.vp.maximum.nxv8bf16(<vscale x 8 x bfloat>, <
 define <vscale x 8 x bfloat> @vfmax_vv_nxv8bf16(<vscale x 8 x bfloat> %va, <vscale x 8 x bfloat> %vb, <vscale x 8 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vfmax_vv_nxv8bf16:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v12, v0
 ; CHECK-NEXT:    vsetvli a1, zero, e16, m2, ta, ma
+; CHECK-NEXT:    vmv1r.v v12, v0
 ; CHECK-NEXT:    vfwcvtbf16.f.f.v v16, v8
 ; CHECK-NEXT:    vsetvli zero, a0, e32, m4, ta, ma
 ; CHECK-NEXT:    vmfeq.vv v8, v16, v16, v0.t
@@ -217,8 +217,8 @@ define <vscale x 16 x bfloat> @vfmax_vv_nxv16bf16(<vscale x 16 x bfloat> %va, <v
 ; CHECK-NEXT:    slli a1, a1, 3
 ; CHECK-NEXT:    sub sp, sp, a1
 ; CHECK-NEXT:    .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x08, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 8 * vlenb
-; CHECK-NEXT:    vmv1r.v v7, v0
 ; CHECK-NEXT:    vsetvli a1, zero, e16, m4, ta, ma
+; CHECK-NEXT:    vmv1r.v v7, v0
 ; CHECK-NEXT:    vfwcvtbf16.f.f.v v24, v8
 ; CHECK-NEXT:    vsetvli zero, a0, e32, m8, ta, ma
 ; CHECK-NEXT:    vmfeq.vv v8, v24, v24, v0.t
@@ -569,8 +569,8 @@ declare <vscale x 1 x half> @llvm.vp.maximum.nxv1f16(<vscale x 1 x half>, <vscal
 define <vscale x 1 x half> @vfmax_vv_nxv1f16(<vscale x 1 x half> %va, <vscale x 1 x half> %vb, <vscale x 1 x i1> %m, i32 zeroext %evl) {
 ; ZVFH-LABEL: vfmax_vv_nxv1f16:
 ; ZVFH:       # %bb.0:
-; ZVFH-NEXT:    vmv1r.v v10, v0
 ; ZVFH-NEXT:    vsetvli zero, a0, e16, mf4, ta, ma
+; ZVFH-NEXT:    vmv1r.v v10, v0
 ; ZVFH-NEXT:    vmfeq.vv v0, v8, v8, v0.t
 ; ZVFH-NEXT:    vmerge.vvm v11, v8, v9, v0
 ; ZVFH-NEXT:    vmv1r.v v0, v10
@@ -582,8 +582,8 @@ define <vscale x 1 x half> @vfmax_vv_nxv1f16(<vscale x 1 x half> %va, <vscale x
 ;
 ; ZVFHMIN-LABEL: vfmax_vv_nxv1f16:
 ; ZVFHMIN:       # %bb.0:
-; ZVFHMIN-NEXT:    vmv1r.v v10, v0
 ; ZVFHMIN-NEXT:    vsetvli a1, zero, e16, mf4, ta, ma
+; ZVFHMIN-NEXT:    vmv1r.v v10, v0
 ; ZVFHMIN-NEXT:    vfwcvt.f.f.v v11, v8
 ; ZVFHMIN-NEXT:    vsetvli zero, a0, e32, mf2, ta, ma
 ; ZVFHMIN-NEXT:    vmfeq.vv v0, v11, v11, v0.t
@@ -639,8 +639,8 @@ declare <vscale x 2 x half> @llvm.vp.maximum.nxv2f16(<vscale x 2 x half>, <vscal
 define <vscale x 2 x half> @vfmax_vv_nxv2f16(<vscale x 2 x half> %va, <vscale x 2 x half> %vb, <vscale x 2 x i1> %m, i32 zeroext %evl) {
 ; ZVFH-LABEL: vfmax_vv_nxv2f16:
 ; ZVFH:       # %bb.0:
-; ZVFH-NEXT:    vmv1r.v v10, v0
 ; ZVFH-NEXT:    vsetvli zero, a0, e16, mf2, ta, ma
+; ZVFH-NEXT:    vmv1r.v v10, v0
 ; ZVFH-NEXT:    vmfeq.vv v0, v8, v8, v0.t
 ; ZVFH-NEXT:    vmerge.vvm v11, v8, v9, v0
 ; ZVFH-NEXT:    vmv1r.v v0, v10
@@ -652,8 +652,8 @@ define <vscale x 2 x half> @vfmax_vv_nxv2f16(<vscale x 2 x half> %va, <vscale x
 ;
 ; ZVFHMIN-LABEL: vfmax_vv_nxv2f16:
 ; ZVFHMIN:       # %bb.0:
-; ZVFHMIN-NEXT:    vmv1r.v v10, v0
 ; ZVFHMIN-NEXT:    vsetvli a1, zero, e16, mf2, ta, ma
+; ZVFHMIN-NEXT:    vmv1r.v v10, v0
 ; ZVFHMIN-NEXT:    vfwcvt.f.f.v v11, v8
 ; ZVFHMIN-NEXT:    vsetvli zero, a0, e32, m1, ta, ma
 ; ZVFHMIN-NEXT:    vmfeq.vv v0, v11, v11, v0.t
@@ -709,8 +709,8 @@ declare <vscale x 4 x half> @llvm.vp.maximum.nxv4f16(<vscale x 4 x half>, <vscal
 define <vscale x 4 x half> @vfmax_vv_nxv4f16(<vscale x 4 x half> %va, <vscale x 4 x half> %vb, <vscale x 4 x i1> %m, i32 zeroext %evl) {
 ; ZVFH-LABEL: vfmax_vv_nxv4f16:
 ; ZVFH:       # %bb.0:
-; ZVFH-NEXT:    vmv1r.v v10, v0
 ; ZVFH-NEXT:    vsetvli zero, a0, e16, m1, ta, ma
+; ZVFH-NEXT:    vmv1r.v v10, v0
 ; ZVFH-NEXT:    vmfeq.vv v0, v8, v8, v0.t
 ; ZVFH-NEXT:    vmerge.vvm v11, v8, v9, v0
 ; ZVFH-NEXT:    vmv1r.v v0, v10
@@ -722,8 +722,8 @@ define <vscale x 4 x half> @vfmax_vv_nxv4f16(<vscale x 4 x half> %va, <vscale x
 ;
 ; ZVFHMIN-LABEL: vfmax_vv_nxv4f16:
 ; ZVFHMIN:       # %bb.0:
-; ZVFHMIN-NEXT:    vmv1r.v v10, v0
 ; ZVFHMIN-NEXT:    vsetvli a1, zero, e16, m1, ta, ma
+; ZVFHMIN-NEXT:    vmv1r.v v10, v0
 ; ZVFHMIN-NEXT:    vfwcvt.f.f.v v12, v8
 ; ZVFHMIN-NEXT:    vsetvli zero, a0, e32, m2, ta, ma
 ; ZVFHMIN-NEXT:    vmfeq.vv v8, v12, v12, v0.t
@@ -781,8 +781,8 @@ declare <vscale x 8 x half> @llvm.vp.maximum.nxv8f16(<vscale x 8 x half>, <vscal
 define <vscale x 8 x half> @vfmax_vv_nxv8f16(<vscale x 8 x half> %va, <vscale x 8 x half> %vb, <vscale x 8 x i1> %m, i32 zeroext %evl) {
 ; ZVFH-LABEL: vfmax_vv_nxv8f16:
 ; ZVFH:       # %bb.0:
-; ZVFH-NEXT:    vmv1r.v v12, v0
 ; ZVFH-NEXT:    vsetvli zero, a0, e16, m2, ta, ma
+; ZVFH-NEXT:    vmv1r.v v12, v0
 ; ZVFH-NEXT:    vmfeq.vv v13, v8, v8, v0.t
 ; ZVFH-NEXT:    vmv1r.v v0, v13
 ; ZVFH-NEXT:    vmerge.vvm v14, v8, v10, v0
@@ -796,8 +796,8 @@ define <vscale x 8 x half> @vfmax_vv_nxv8f16(<vscale x 8 x half> %va, <vscale x
 ;
 ; ZVFHMIN-LABEL: vfmax_vv_nxv8f16:
 ; ZVFHMIN:       # %bb.0:
-; ZVFHMIN-NEXT:    vmv1r.v v12, v0
 ; ZVFHMIN-NEXT:    vsetvli a1, zero, e16, m2, ta, ma
+; ZVFHMIN-NEXT:    vmv1r.v v12, v0
 ; ZVFHMIN-NEXT:    vfwcvt.f.f.v v16, v8
 ; ZVFHMIN-NEXT:    vsetvli zero, a0, e32, m4, ta, ma
 ; ZVFHMIN-NEXT:    vmfeq.vv v8, v16, v16, v0.t
@@ -855,8 +855,8 @@ declare <vscale x 16 x half> @llvm.vp.maximum.nxv16f16(<vscale x 16 x half>, <vs
 define <vscale x 16 x half> @vfmax_vv_nxv16f16(<vscale x 16 x half> %va, <vscale x 16 x half> %vb, <vscale x 16 x i1> %m, i32 zeroext %evl) {
 ; ZVFH-LABEL: vfmax_vv_nxv16f16:
 ; ZVFH:       # %bb.0:
-; ZVFH-NEXT:    vmv1r.v v16, v0
 ; ZVFH-NEXT:    vsetvli zero, a0, e16, m4, ta, ma
+; ZVFH-NEXT:    vmv1r.v v16, v0
 ; ZVFH-NEXT:    vmfeq.vv v17, v8, v8, v0.t
 ; ZVFH-NEXT:    vmv1r.v v0, v17
 ; ZVFH-NEXT:    vmerge.vvm v20, v8, v12, v0
@@ -876,8 +876,8 @@ define <vscale x 16 x half> @vfmax_vv_nxv16f16(<vscale x 16 x half> %va, <vscale
 ; ZVFHMIN-NEXT:    slli a1, a1, 3
 ; ZVFHMIN-NEXT:    sub sp, sp, a1
 ; ZVFHMIN-NEXT:    .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x08, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 8 * vlenb
-; ZVFHMIN-NEXT:    vmv1r.v v7, v0
 ; ZVFHMIN-NEXT:    vsetvli a1, zero, e16, m4, ta, ma
+; ZVFHMIN-NEXT:    vmv1r.v v7, v0
 ; ZVFHMIN-NEXT:    vfwcvt.f.f.v v24, v8
 ; ZVFHMIN-NEXT:    vsetvli zero, a0, e32, m8, ta, ma
 ; ZVFHMIN-NEXT:    vmfeq.vv v8, v24, v24, v0.t
@@ -964,8 +964,8 @@ define <vscale x 32 x half> @vfmax_vv_nxv32f16(<vscale x 32 x half> %va, <vscale
 ; ZVFH-NEXT:    slli a1, a1, 3
 ; ZVFH-NEXT:    sub sp, sp, a1
 ; ZVFH-NEXT:    .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x08, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 8 * vlenb
-; ZVFH-NEXT:    vmv1r.v v7, v0
 ; ZVFH-NEXT:    vsetvli zero, a0, e16, m8, ta, ma
+; ZVFH-NEXT:    vmv1r.v v7, v0
 ; ZVFH-NEXT:    vmfeq.vv v25, v8, v8, v0.t
 ; ZVFH-NEXT:    vmv1r.v v0, v25
 ; ZVFH-NEXT:    vmerge.vvm v24, v8, v16, v0
@@ -1280,8 +1280,8 @@ declare <vscale x 1 x float> @llvm.vp.maximum.nxv1f32(<vscale x 1 x float>, <vsc
 define <vscale x 1 x float> @vfmax_vv_nxv1f32(<vscale x 1 x float> %va, <vscale x 1 x float> %vb, <vscale x 1 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vfmax_vv_nxv1f32:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e32, mf2, ta, ma
+; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmfeq.vv v0, v8, v8, v0.t
 ; CHECK-NEXT:    vmerge.vvm v11, v8, v9, v0
 ; CHECK-NEXT:    vmv1r.v v0, v10
@@ -1313,8 +1313,8 @@ declare <vscale x 2 x float> @llvm.vp.maximum.nxv2f32(<vscale x 2 x float>, <vsc
 define <vscale x 2 x float> @vfmax_vv_nxv2f32(<vscale x 2 x float> %va, <vscale x 2 x float> %vb, <vscale x 2 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vfmax_vv_nxv2f32:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e32, m1, ta, ma
+; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmfeq.vv v0, v8, v8, v0.t
 ; CHECK-NEXT:    vmerge.vvm v11, v8, v9, v0
 ; CHECK-NEXT:    vmv1r.v v0, v10
@@ -1346,8 +1346,8 @@ declare <vscale x 4 x float> @llvm.vp.maximum.nxv4f32(<vscale x 4 x float>, <vsc
 define <vscale x 4 x float> @vfmax_vv_nxv4f32(<vscale x 4 x float> %va, <vscale x 4 x float> %vb, <vscale x 4 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vfmax_vv_nxv4f32:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v12, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e32, m2, ta, ma
+; CHECK-NEXT:    vmv1r.v v12, v0
 ; CHECK-NEXT:    vmfeq.vv v13, v8, v8, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v13
 ; CHECK-NEXT:    vmerge.vvm v14, v8, v10, v0
@@ -1381,8 +1381,8 @@ declare <vscale x 8 x float> @llvm.vp.maximum.nxv8f32(<vscale x 8 x float>, <vsc
 define <vscale x 8 x float> @vfmax_vv_nxv8f32(<vscale x 8 x float> %va, <vscale x 8 x float> %vb, <vscale x 8 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vfmax_vv_nxv8f32:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v16, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e32, m4, ta, ma
+; CHECK-NEXT:    vmv1r.v v16, v0
 ; CHECK-NEXT:    vmfeq.vv v17, v8, v8, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v17
 ; CHECK-NEXT:    vmerge.vvm v20, v8, v12, v0
@@ -1416,8 +1416,8 @@ declare <vscale x 1 x double> @llvm.vp.maximum.nxv1f64(<vscale x 1 x double>, <v
 define <vscale x 1 x double> @vfmax_vv_nxv1f64(<vscale x 1 x double> %va, <vscale x 1 x double> %vb, <vscale x 1 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vfmax_vv_nxv1f64:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e64, m1, ta, ma
+; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmfeq.vv v0, v8, v8, v0.t
 ; CHECK-NEXT:    vmerge.vvm v11, v8, v9, v0
 ; CHECK-NEXT:    vmv1r.v v0, v10
@@ -1449,8 +1449,8 @@ declare <vscale x 2 x double> @llvm.vp.maximum.nxv2f64(<vscale x 2 x double>, <v
 define <vscale x 2 x double> @vfmax_vv_nxv2f64(<vscale x 2 x double> %va, <vscale x 2 x double> %vb, <vscale x 2 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vfmax_vv_nxv2f64:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v12, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e64, m2, ta, ma
+; CHECK-NEXT:    vmv1r.v v12, v0
 ; CHECK-NEXT:    vmfeq.vv v13, v8, v8, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v13
 ; CHECK-NEXT:    vmerge.vvm v14, v8, v10, v0
@@ -1484,8 +1484,8 @@ declare <vscale x 4 x double> @llvm.vp.maximum.nxv4f64(<vscale x 4 x double>, <v
 define <vscale x 4 x double> @vfmax_vv_nxv4f64(<vscale x 4 x double> %va, <vscale x 4 x double> %vb, <vscale x 4 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vfmax_vv_nxv4f64:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v16, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e64, m4, ta, ma
+; CHECK-NEXT:    vmv1r.v v16, v0
 ; CHECK-NEXT:    vmfeq.vv v17, v8, v8, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v17
 ; CHECK-NEXT:    vmerge.vvm v20, v8, v12, v0
@@ -1525,8 +1525,8 @@ define <vscale x 8 x double> @vfmax_vv_nxv8f64(<vscale x 8 x double> %va, <vscal
 ; CHECK-NEXT:    slli a1, a1, 3
 ; CHECK-NEXT:    sub sp, sp, a1
 ; CHECK-NEXT:    .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x08, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 8 * vlenb
-; CHECK-NEXT:    vmv1r.v v7, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e64, m8, ta, ma
+; CHECK-NEXT:    vmv1r.v v7, v0
 ; CHECK-NEXT:    vmfeq.vv v25, v8, v8, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v25
 ; CHECK-NEXT:    vmerge.vvm v24, v8, v16, v0
@@ -1577,6 +1577,7 @@ define <vscale x 16 x double> @vfmax_vv_nxv16f64(<vscale x 16 x double> %va, <vs
 ; CHECK-NEXT:    mul a1, a1, a3
 ; CHECK-NEXT:    sub sp, sp, a1
 ; CHECK-NEXT:    .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x18, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 24 * vlenb
+; CHECK-NEXT:    vsetvli a1, zero, e8, mf4, ta, ma
 ; CHECK-NEXT:    vmv1r.v v7, v0
 ; CHECK-NEXT:    addi a1, sp, 16
 ; CHECK-NEXT:    vs8r.v v16, (a1) # Unknown-size Folded Spill
@@ -1588,7 +1589,6 @@ define <vscale x 16 x double> @vfmax_vv_nxv16f64(<vscale x 16 x double> %va, <vs
 ; CHECK-NEXT:    csrr a1, vlenb
 ; CHECK-NEXT:    slli a3, a1, 3
 ; CHECK-NEXT:    srli a4, a1, 3
-; CHECK-NEXT:    vsetvli a5, zero, e8, mf4, ta, ma
 ; CHECK-NEXT:    vslidedown.vx v6, v0, a4
 ; CHECK-NEXT:    sub a4, a2, a1
 ; CHECK-NEXT:    add a3, a0, a3
diff --git a/llvm/test/CodeGen/RISCV/rvv/fminimum-sdnode.ll b/llvm/test/CodeGen/RISCV/rvv/fminimum-sdnode.ll
index d41da7b6a2af9..16c119dd87fe0 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fminimum-sdnode.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fminimum-sdnode.ll
@@ -157,6 +157,7 @@ define <vscale x 32 x bfloat> @vfmin_nxv32bf16_vv(<vscale x 32 x bfloat> %a, <vs
 ; ZVFH-NEXT:    slli a0, a0, 1
 ; ZVFH-NEXT:    add a0, a0, a1
 ; ZVFH-NEXT:    sub sp, sp, a0
+; ZVFH-NEXT:    vsetvli a0, zero, e16, m4, ta, ma
 ; ZVFH-NEXT:    vmv8r.v v24, v16
 ; ZVFH-NEXT:    csrr a0, vlenb
 ; ZVFH-NEXT:    slli a0, a0, 3
@@ -164,7 +165,6 @@ define <vscale x 32 x bfloat> @vfmin_nxv32bf16_vv(<vscale x 32 x bfloat> %a, <vs
 ; ZVFH-NEXT:    addi a0, a0, 16
 ; ZVFH-NEXT:    vs8r.v v16, (a0) # Unknown-size Folded Spill
 ; ZVFH-NEXT:    vmv8r.v v0, v8
-; ZVFH-NEXT:    vsetvli a0, zero, e16, m4, ta, ma
 ; ZVFH-NEXT:    vfwcvtbf16.f.f.v v16, v24
 ; ZVFH-NEXT:    vfwcvtbf16.f.f.v v8, v0
 ; ZVFH-NEXT:    vsetvli zero, zero, e32, m8, ta, ma
@@ -232,6 +232,7 @@ define <vscale x 32 x bfloat> @vfmin_nxv32bf16_vv(<vscale x 32 x bfloat> %a, <vs
 ; ZVFHMIN-NEXT:    li a1, 24
 ; ZVFHMIN-NEXT:    mul a0, a0, a1
 ; ZVFHMIN-NEXT:    sub sp, sp, a0
+; ZVFHMIN-NEXT:    vsetvli a0, zero, e16, m4, ta, ma
 ; ZVFHMIN-NEXT:    vmv8r.v v24, v16
 ; ZVFHMIN-NEXT:    csrr a0, vlenb
 ; ZVFHMIN-NEXT:    slli a0, a0, 3
@@ -239,7 +240,6 @@ define <vscale x 32 x bfloat> @vfmin_nxv32bf16_vv(<vscale x 32 x bfloat> %a, <vs
 ; ZVFHMIN-NEXT:    addi a0, a0, 16
 ; ZVFHMIN-NEXT:    vs8r.v v16, (a0) # Unknown-size Folded Spill
 ; ZVFHMIN-NEXT:    vmv8r.v v0, v8
-; ZVFHMIN-NEXT:    vsetvli a0, zero, e16, m4, ta, ma
 ; ZVFHMIN-NEXT:    vfwcvtbf16.f.f.v v16, v24
 ; ZVFHMIN-NEXT:    vfwcvtbf16.f.f.v v8, v0
 ; ZVFHMIN-NEXT:    vsetvli zero, zero, e32, m8, ta, ma
@@ -501,6 +501,7 @@ define <vscale x 32 x half> @vfmin_nxv32f16_vv(<vscale x 32 x half> %a, <vscale
 ; ZVFHMIN-NEXT:    li a1, 24
 ; ZVFHMIN-NEXT:    mul a0, a0, a1
 ; ZVFHMIN-NEXT:    sub sp, sp, a0
+; ZVFHMIN-NEXT:    vsetvli a0, zero, e16, m4, ta, ma
 ; ZVFHMIN-NEXT:    vmv8r.v v24, v16
 ; ZVFHMIN-NEXT:    csrr a0, vlenb
 ; ZVFHMIN-NEXT:    slli a0, a0, 3
@@ -508,7 +509,6 @@ define <vscale x 32 x half> @vfmin_nxv32f16_vv(<vscale x 32 x half> %a, <vscale
 ; ZVFHMIN-NEXT:    addi a0, a0, 16
 ; ZVFHMIN-NEXT:    vs8r.v v16, (a0) # Unknown-size Folded Spill
 ; ZVFHMIN-NEXT:    vmv8r.v v0, v8
-; ZVFHMIN-NEXT:    vsetvli a0, zero, e16, m4, ta, ma
 ; ZVFHMIN-NEXT:    vfwcvt.f.f.v v16, v24
 ; ZVFHMIN-NEXT:    vfwcvt.f.f.v v8, v0
 ; ZVFHMIN-NEXT:    vsetvli zero, zero, e32, m8, ta, ma
diff --git a/llvm/test/CodeGen/RISCV/rvv/fminimum-vp.ll b/llvm/test/CodeGen/RISCV/rvv/fminimum-vp.ll
index ef6f33de4ce63..8e448fcda9c5d 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fminimum-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fminimum-vp.ll
@@ -19,8 +19,8 @@ declare <vscale x 1 x bfloat> @llvm.vp.minimum.nxv1bf16(<vscale x 1 x bfloat>, <
 define <vscale x 1 x bfloat> @vfmin_vv_nxv1bf16(<vscale x 1 x bfloat> %va, <vscale x 1 x bfloat> %vb, <vscale x 1 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vfmin_vv_nxv1bf16:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vsetvli a1, zero, e16, mf4, ta, ma
+; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vfwcvtbf16.f.f.v v11, v8
 ; CHECK-NEXT:    vsetvli zero, a0, e32, mf2, ta, ma
 ; CHECK-NEXT:    vmfeq.vv v0, v11, v11, v0.t
@@ -66,8 +66,8 @@ declare <vscale x 2 x bfloat> @llvm.vp.minimum.nxv2bf16(<vscale x 2 x bfloat>, <
 define <vscale x 2 x bfloat> @vfmin_vv_nxv2bf16(<vscale x 2 x bfloat> %va, <vscale x 2 x bfloat> %vb, <vscale x 2 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vfmin_vv_nxv2bf16:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vsetvli a1, zero, e16, mf2, ta, ma
+; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vfwcvtbf16.f.f.v v11, v8
 ; CHECK-NEXT:    vsetvli zero, a0, e32, m1, ta, ma
 ; CHECK-NEXT:    vmfeq.vv v0, v11, v11, v0.t
@@ -113,8 +113,8 @@ declare <vscale x 4 x bfloat> @llvm.vp.minimum.nxv4bf16(<vscale x 4 x bfloat>, <
 define <vscale x 4 x bfloat> @vfmin_vv_nxv4bf16(<vscale x 4 x bfloat> %va, <vscale x 4 x bfloat> %vb, <vscale x 4 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vfmin_vv_nxv4bf16:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vsetvli a1, zero, e16, m1, ta, ma
+; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vfwcvtbf16.f.f.v v12, v8
 ; CHECK-NEXT:    vsetvli zero, a0, e32, m2, ta, ma
 ; CHECK-NEXT:    vmfeq.vv v8, v12, v12, v0.t
@@ -162,8 +162,8 @@ declare <vscale x 8 x bfloat> @llvm.vp.minimum.nxv8bf16(<vscale x 8 x bfloat>, <
 define <vscale x 8 x bfloat> @vfmin_vv_nxv8bf16(<vscale x 8 x bfloat> %va, <vscale x 8 x bfloat> %vb, <vscale x 8 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vfmin_vv_nxv8bf16:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v12, v0
 ; CHECK-NEXT:    vsetvli a1, zero, e16, m2, ta, ma
+; CHECK-NEXT:    vmv1r.v v12, v0
 ; CHECK-NEXT:    vfwcvtbf16.f.f.v v16, v8
 ; CHECK-NEXT:    vsetvli zero, a0, e32, m4, ta, ma
 ; CHECK-NEXT:    vmfeq.vv v8, v16, v16, v0.t
@@ -217,8 +217,8 @@ define <vscale x 16 x bfloat> @vfmin_vv_nxv16bf16(<vscale x 16 x bfloat> %va, <v
 ; CHECK-NEXT:    slli a1, a1, 3
 ; CHECK-NEXT:    sub sp, sp, a1
 ; CHECK-NEXT:    .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x08, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 8 * vlenb
-; CHECK-NEXT:    vmv1r.v v7, v0
 ; CHECK-NEXT:    vsetvli a1, zero, e16, m4, ta, ma
+; CHECK-NEXT:    vmv1r.v v7, v0
 ; CHECK-NEXT:    vfwcvtbf16.f.f.v v24, v8
 ; CHECK-NEXT:    vsetvli zero, a0, e32, m8, ta, ma
 ; CHECK-NEXT:    vmfeq.vv v8, v24, v24, v0.t
@@ -569,8 +569,8 @@ declare <vscale x 1 x half> @llvm.vp.minimum.nxv1f16(<vscale x 1 x half>, <vscal
 define <vscale x 1 x half> @vfmin_vv_nxv1f16(<vscale x 1 x half> %va, <vscale x 1 x half> %vb, <vscale x 1 x i1> %m, i32 zeroext %evl) {
 ; ZVFH-LABEL: vfmin_vv_nxv1f16:
 ; ZVFH:       # %bb.0:
-; ZVFH-NEXT:    vmv1r.v v10, v0
 ; ZVFH-NEXT:    vsetvli zero, a0, e16, mf4, ta, ma
+; ZVFH-NEXT:    vmv1r.v v10, v0
 ; ZVFH-NEXT:    vmfeq.vv v0, v8, v8, v0.t
 ; ZVFH-NEXT:    vmerge.vvm v11, v8, v9, v0
 ; ZVFH-NEXT:    vmv1r.v v0, v10
@@ -582,8 +582,8 @@ define <vscale x 1 x half> @vfmin_vv_nxv1f16(<vscale x 1 x half> %va, <vscale x
 ;
 ; ZVFHMIN-LABEL: vfmin_vv_nxv1f16:
 ; ZVFHMIN:       # %bb.0:
-; ZVFHMIN-NEXT:    vmv1r.v v10, v0
 ; ZVFHMIN-NEXT:    vsetvli a1, zero, e16, mf4, ta, ma
+; ZVFHMIN-NEXT:    vmv1r.v v10, v0
 ; ZVFHMIN-NEXT:    vfwcvt.f.f.v v11, v8
 ; ZVFHMIN-NEXT:    vsetvli zero, a0, e32, mf2, ta, ma
 ; ZVFHMIN-NEXT:    vmfeq.vv v0, v11, v11, v0.t
@@ -639,8 +639,8 @@ declare <vscale x 2 x half> @llvm.vp.minimum.nxv2f16(<vscale x 2 x half>, <vscal
 define <vscale x 2 x half> @vfmin_vv_nxv2f16(<vscale x 2 x half> %va, <vscale x 2 x half> %vb, <vscale x 2 x i1> %m, i32 zeroext %evl) {
 ; ZVFH-LABEL: vfmin_vv_nxv2f16:
 ; ZVFH:       # %bb.0:
-; ZVFH-NEXT:    vmv1r.v v10, v0
 ; ZVFH-NEXT:    vsetvli zero, a0, e16, mf2, ta, ma
+; ZVFH-NEXT:    vmv1r.v v10, v0
 ; ZVFH-NEXT:    vmfeq.vv v0, v8, v8, v0.t
 ; ZVFH-NEXT:    vmerge.vvm v11, v8, v9, v0
 ; ZVFH-NEXT:    vmv1r.v v0, v10
@@ -652,8 +652,8 @@ define <vscale x 2 x half> @vfmin_vv_nxv2f16(<vscale x 2 x half> %va, <vscale x
 ;
 ; ZVFHMIN-LABEL: vfmin_vv_nxv2f16:
 ; ZVFHMIN:       # %bb.0:
-; ZVFHMIN-NEXT:    vmv1r.v v10, v0
 ; ZVFHMIN-NEXT:    vsetvli a1, zero, e16, mf2, ta, ma
+; ZVFHMIN-NEXT:    vmv1r.v v10, v0
 ; ZVFHMIN-NEXT:    vfwcvt.f.f.v v11, v8
 ; ZVFHMIN-NEXT:    vsetvli zero, a0, e32, m1, ta, ma
 ; ZVFHMIN-NEXT:    vmfeq.vv v0, v11, v11, v0.t
@@ -709,8 +709,8 @@ declare <vscale x 4 x half> @llvm.vp.minimum.nxv4f16(<vscale x 4 x half>, <vscal
 define <vscale x 4 x half> @vfmin_vv_nxv4f16(<vscale x 4 x half> %va, <vscale x 4 x half> %vb, <vscale x 4 x i1> %m, i32 zeroext %evl) {
 ; ZVFH-LABEL: vfmin_vv_nxv4f16:
 ; ZVFH:       # %bb.0:
-; ZVFH-NEXT:    vmv1r.v v10, v0
 ; ZVFH-NEXT:    vsetvli zero, a0, e16, m1, ta, ma
+; ZVFH-NEXT:    vmv1r.v v10, v0
 ; ZVFH-NEXT:    vmfeq.vv v0, v8, v8, v0.t
 ; ZVFH-NEXT:    vmerge.vvm v11, v8, v9, v0
 ; ZVFH-NEXT:    vmv1r.v v0, v10
@@ -722,8 +722,8 @@ define <vscale x 4 x half> @vfmin_vv_nxv4f16(<vscale x 4 x half> %va, <vscale x
 ;
 ; ZVFHMIN-LABEL: vfmin_vv_nxv4f16:
 ; ZVFHMIN:       # %bb.0:
-; ZVFHMIN-NEXT:    vmv1r.v v10, v0
 ; ZVFHMIN-NEXT:    vsetvli a1, zero, e16, m1, ta, ma
+; ZVFHMIN-NEXT:    vmv1r.v v10, v0
 ; ZVFHMIN-NEXT:    vfwcvt.f.f.v v12, v8
 ; ZVFHMIN-NEXT:    vsetvli zero, a0, e32, m2, ta, ma
 ; ZVFHMIN-NEXT:    vmfeq.vv v8, v12, v12, v0.t
@@ -781,8 +781,8 @@ declare <vscale x 8 x half> @llvm.vp.minimum.nxv8f16(<vscale x 8 x half>, <vscal
 define <vscale x 8 x half> @vfmin_vv_nxv8f16(<vscale x 8 x half> %va, <vscale x 8 x half> %vb, <vscale x 8 x i1> %m, i32 zeroext %evl) {
 ; ZVFH-LABEL: vfmin_vv_nxv8f16:
 ; ZVFH:       # %bb.0:
-; ZVFH-NEXT:    vmv1r.v v12, v0
 ; ZVFH-NEXT:    vsetvli zero, a0, e16, m2, ta, ma
+; ZVFH-NEXT:    vmv1r.v v12, v0
 ; ZVFH-NEXT:    vmfeq.vv v13, v8, v8, v0.t
 ; ZVFH-NEXT:    vmv1r.v v0, v13
 ; ZVFH-NEXT:    vmerge.vvm v14, v8, v10, v0
@@ -796,8 +796,8 @@ define <vscale x 8 x half> @vfmin_vv_nxv8f16(<vscale x 8 x half> %va, <vscale x
 ;
 ; ZVFHMIN-LABEL: vfmin_vv_nxv8f16:
 ; ZVFHMIN:       # %bb.0:
-; ZVFHMIN-NEXT:    vmv1r.v v12, v0
 ; ZVFHMIN-NEXT:    vsetvli a1, zero, e16, m2, ta, ma
+; ZVFHMIN-NEXT:    vmv1r.v v12, v0
 ; ZVFHMIN-NEXT:    vfwcvt.f.f.v v16, v8
 ; ZVFHMIN-NEXT:    vsetvli zero, a0, e32, m4, ta, ma
 ; ZVFHMIN-NEXT:    vmfeq.vv v8, v16, v16, v0.t
@@ -855,8 +855,8 @@ declare <vscale x 16 x half> @llvm.vp.minimum.nxv16f16(<vscale x 16 x half>, <vs
 define <vscale x 16 x half> @vfmin_vv_nxv16f16(<vscale x 16 x half> %va, <vscale x 16 x half> %vb, <vscale x 16 x i1> %m, i32 zeroext %evl) {
 ; ZVFH-LABEL: vfmin_vv_nxv16f16:
 ; ZVFH:       # %bb.0:
-; ZVFH-NEXT:    vmv1r.v v16, v0
 ; ZVFH-NEXT:    vsetvli zero, a0, e16, m4, ta, ma
+; ZVFH-NEXT:    vmv1r.v v16, v0
 ; ZVFH-NEXT:    vmfeq.vv v17, v8, v8, v0.t
 ; ZVFH-NEXT:    vmv1r.v v0, v17
 ; ZVFH-NEXT:    vmerge.vvm v20, v8, v12, v0
@@ -876,8 +876,8 @@ define <vscale x 16 x half> @vfmin_vv_nxv16f16(<vscale x 16 x half> %va, <vscale
 ; ZVFHMIN-NEXT:    slli a1, a1, 3
 ; ZVFHMIN-NEXT:    sub sp, sp, a1
 ; ZVFHMIN-NEXT:    .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x08, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 8 * vlenb
-; ZVFHMIN-NEXT:    vmv1r.v v7, v0
 ; ZVFHMIN-NEXT:    vsetvli a1, zero, e16, m4, ta, ma
+; ZVFHMIN-NEXT:    vmv1r.v v7, v0
 ; ZVFHMIN-NEXT:    vfwcvt.f.f.v v24, v8
 ; ZVFHMIN-NEXT:    vsetvli zero, a0, e32, m8, ta, ma
 ; ZVFHMIN-NEXT:    vmfeq.vv v8, v24, v24, v0.t
@@ -964,8 +964,8 @@ define <vscale x 32 x half> @vfmin_vv_nxv32f16(<vscale x 32 x half> %va, <vscale
 ; ZVFH-NEXT:    slli a1, a1, 3
 ; ZVFH-NEXT:    sub sp, sp, a1
 ; ZVFH-NEXT:    .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x08, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 8 * vlenb
-; ZVFH-NEXT:    vmv1r.v v7, v0
 ; ZVFH-NEXT:    vsetvli zero, a0, e16, m8, ta, ma
+; ZVFH-NEXT:    vmv1r.v v7, v0
 ; ZVFH-NEXT:    vmfeq.vv v25, v8, v8, v0.t
 ; ZVFH-NEXT:    vmv1r.v v0, v25
 ; ZVFH-NEXT:    vmerge.vvm v24, v8, v16, v0
@@ -1280,8 +1280,8 @@ declare <vscale x 1 x float> @llvm.vp.minimum.nxv1f32(<vscale x 1 x float>, <vsc
 define <vscale x 1 x float> @vfmin_vv_nxv1f32(<vscale x 1 x float> %va, <vscale x 1 x float> %vb, <vscale x 1 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vfmin_vv_nxv1f32:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e32, mf2, ta, ma
+; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmfeq.vv v0, v8, v8, v0.t
 ; CHECK-NEXT:    vmerge.vvm v11, v8, v9, v0
 ; CHECK-NEXT:    vmv1r.v v0, v10
@@ -1313,8 +1313,8 @@ declare <vscale x 2 x float> @llvm.vp.minimum.nxv2f32(<vscale x 2 x float>, <vsc
 define <vscale x 2 x float> @vfmin_vv_nxv2f32(<vscale x 2 x float> %va, <vscale x 2 x float> %vb, <vscale x 2 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vfmin_vv_nxv2f32:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e32, m1, ta, ma
+; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmfeq.vv v0, v8, v8, v0.t
 ; CHECK-NEXT:    vmerge.vvm v11, v8, v9, v0
 ; CHECK-NEXT:    vmv1r.v v0, v10
@@ -1346,8 +1346,8 @@ declare <vscale x 4 x float> @llvm.vp.minimum.nxv4f32(<vscale x 4 x float>, <vsc
 define <vscale x 4 x float> @vfmin_vv_nxv4f32(<vscale x 4 x float> %va, <vscale x 4 x float> %vb, <vscale x 4 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vfmin_vv_nxv4f32:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v12, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e32, m2, ta, ma
+; CHECK-NEXT:    vmv1r.v v12, v0
 ; CHECK-NEXT:    vmfeq.vv v13, v8, v8, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v13
 ; CHECK-NEXT:    vmerge.vvm v14, v8, v10, v0
@@ -1381,8 +1381,8 @@ declare <vscale x 8 x float> @llvm.vp.minimum.nxv8f32(<vscale x 8 x float>, <vsc
 define <vscale x 8 x float> @vfmin_vv_nxv8f32(<vscale x 8 x float> %va, <vscale x 8 x float> %vb, <vscale x 8 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vfmin_vv_nxv8f32:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v16, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e32, m4, ta, ma
+; CHECK-NEXT:    vmv1r.v v16, v0
 ; CHECK-NEXT:    vmfeq.vv v17, v8, v8, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v17
 ; CHECK-NEXT:    vmerge.vvm v20, v8, v12, v0
@@ -1416,8 +1416,8 @@ declare <vscale x 1 x double> @llvm.vp.minimum.nxv1f64(<vscale x 1 x double>, <v
 define <vscale x 1 x double> @vfmin_vv_nxv1f64(<vscale x 1 x double> %va, <vscale x 1 x double> %vb, <vscale x 1 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vfmin_vv_nxv1f64:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e64, m1, ta, ma
+; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmfeq.vv v0, v8, v8, v0.t
 ; CHECK-NEXT:    vmerge.vvm v11, v8, v9, v0
 ; CHECK-NEXT:    vmv1r.v v0, v10
@@ -1449,8 +1449,8 @@ declare <vscale x 2 x double> @llvm.vp.minimum.nxv2f64(<vscale x 2 x double>, <v
 define <vscale x 2 x double> @vfmin_vv_nxv2f64(<vscale x 2 x double> %va, <vscale x 2 x double> %vb, <vscale x 2 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vfmin_vv_nxv2f64:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v12, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e64, m2, ta, ma
+; CHECK-NEXT:    vmv1r.v v12, v0
 ; CHECK-NEXT:    vmfeq.vv v13, v8, v8, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v13
 ; CHECK-NEXT:    vmerge.vvm v14, v8, v10, v0
@@ -1484,8 +1484,8 @@ declare <vscale x 4 x double> @llvm.vp.minimum.nxv4f64(<vscale x 4 x double>, <v
 define <vscale x 4 x double> @vfmin_vv_nxv4f64(<vscale x 4 x double> %va, <vscale x 4 x double> %vb, <vscale x 4 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vfmin_vv_nxv4f64:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v16, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e64, m4, ta, ma
+; CHECK-NEXT:    vmv1r.v v16, v0
 ; CHECK-NEXT:    vmfeq.vv v17, v8, v8, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v17
 ; CHECK-NEXT:    vmerge.vvm v20, v8, v12, v0
@@ -1525,8 +1525,8 @@ define <vscale x 8 x double> @vfmin_vv_nxv8f64(<vscale x 8 x double> %va, <vscal
 ; CHECK-NEXT:    slli a1, a1, 3
 ; CHECK-NEXT:    sub sp, sp, a1
 ; CHECK-NEXT:    .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x08, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 8 * vlenb
-; CHECK-NEXT:    vmv1r.v v7, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e64, m8, ta, ma
+; CHECK-NEXT:    vmv1r.v v7, v0
 ; CHECK-NEXT:    vmfeq.vv v25, v8, v8, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v25
 ; CHECK-NEXT:    vmerge.vvm v24, v8, v16, v0
@@ -1577,6 +1577,7 @@ define <vscale x 16 x double> @vfmin_vv_nxv16f64(<vscale x 16 x double> %va, <vs
 ; CHECK-NEXT:    mul a1, a1, a3
 ; CHECK-NEXT:    sub sp, sp, a1
 ; CHECK-NEXT:    .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x18, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 24 * vlenb
+; CHECK-NEXT:    vsetvli a1, zero, e8, mf4, ta, ma
 ; CHECK-NEXT:    vmv1r.v v7, v0
 ; CHECK-NEXT:    addi a1, sp, 16
 ; CHECK-NEXT:    vs8r.v v16, (a1) # Unknown-size Folded Spill
@@ -1588,7 +1589,6 @@ define <vscale x 16 x double> @vfmin_vv_nxv16f64(<vscale x 16 x double> %va, <vs
 ; CHECK-NEXT:    csrr a1, vlenb
 ; CHECK-NEXT:    slli a3, a1, 3
 ; CHECK-NEXT:    srli a4, a1, 3
-; CHECK-NEXT:    vsetvli a5, zero, e8, mf4, ta, ma
 ; CHECK-NEXT:    vslidedown.vx v6, v0, a4
 ; CHECK-NEXT:    sub a4, a2, a1
 ; CHECK-NEXT:    add a3, a0, a3
diff --git a/llvm/test/CodeGen/RISCV/rvv/fold-scalar-load-crash.ll b/llvm/test/CodeGen/RISCV/rvv/fold-scalar-load-crash.ll
index 025874a1a74e2..5e657a93ec0d6 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fold-scalar-load-crash.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fold-scalar-load-crash.ll
@@ -18,8 +18,8 @@ define i32 @test(i32 %size, ptr %add.ptr, i64 %const) {
 ; RV32-NEXT:  .LBB0_1: # %for.body
 ; RV32-NEXT:    # =>This Inner Loop Header: Depth=1
 ; RV32-NEXT:    th.lrb a0, a1, a0, 0
+; RV32-NEXT:    vsetivli zero, 1, e8, m1, tu, ma
 ; RV32-NEXT:    vmv1r.v v9, v8
-; RV32-NEXT:    vsetivli zero, 8, e8, m1, tu, ma
 ; RV32-NEXT:    vmv.s.x v9, a0
 ; RV32-NEXT:    vsetivli zero, 8, e8, mf2, ta, ma
 ; RV32-NEXT:    vmseq.vi v9, v9, 0
@@ -45,8 +45,8 @@ define i32 @test(i32 %size, ptr %add.ptr, i64 %const) {
 ; RV64-NEXT:    # =>This Inner Loop Header: Depth=1
 ; RV64-NEXT:    sext.w a0, a0
 ; RV64-NEXT:    th.lrb a0, a1, a0, 0
+; RV64-NEXT:    vsetivli zero, 1, e8, m1, tu, ma
 ; RV64-NEXT:    vmv1r.v v9, v8
-; RV64-NEXT:    vsetivli zero, 8, e8, m1, tu, ma
 ; RV64-NEXT:    vmv.s.x v9, a0
 ; RV64-NEXT:    vsetivli zero, 8, e8, mf2, ta, ma
 ; RV64-NEXT:    vmseq.vi v9, v9, 0
diff --git a/llvm/test/CodeGen/RISCV/rvv/fshr-fshl-vp.ll b/llvm/test/CodeGen/RISCV/rvv/fshr-fshl-vp.ll
index c7e3c8cb51982..b569efc7447da 100644
--- a/llvm/test/CodeGen/RISCV/rvv/fshr-fshl-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/fshr-fshl-vp.ll
@@ -703,10 +703,10 @@ define <vscale x 16 x i32> @fshl_v16i32(<vscale x 16 x i32> %a, <vscale x 16 x i
 ; CHECK-NEXT:    .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x08, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 8 * vlenb
 ; CHECK-NEXT:    addi a2, sp, 16
 ; CHECK-NEXT:    vs8r.v v16, (a2) # Unknown-size Folded Spill
+; CHECK-NEXT:    vsetvli zero, a1, e32, m8, ta, ma
 ; CHECK-NEXT:    vmv8r.v v16, v8
 ; CHECK-NEXT:    vl8re32.v v24, (a0)
 ; CHECK-NEXT:    li a0, 31
-; CHECK-NEXT:    vsetvli zero, a1, e32, m8, ta, ma
 ; CHECK-NEXT:    vand.vx v8, v24, a0, v0.t
 ; CHECK-NEXT:    vsll.vv v8, v16, v8, v0.t
 ; CHECK-NEXT:    vnot.v v16, v24, v0.t
@@ -882,10 +882,10 @@ define <vscale x 7 x i64> @fshl_v7i64(<vscale x 7 x i64> %a, <vscale x 7 x i64>
 ; CHECK-NEXT:    .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x08, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 8 * vlenb
 ; CHECK-NEXT:    addi a2, sp, 16
 ; CHECK-NEXT:    vs8r.v v16, (a2) # Unknown-size Folded Spill
+; CHECK-NEXT:    vsetvli zero, a1, e64, m8, ta, ma
 ; CHECK-NEXT:    vmv8r.v v16, v8
 ; CHECK-NEXT:    vl8re64.v v24, (a0)
 ; CHECK-NEXT:    li a0, 63
-; CHECK-NEXT:    vsetvli zero, a1, e64, m8, ta, ma
 ; CHECK-NEXT:    vand.vx v8, v24, a0, v0.t
 ; CHECK-NEXT:    vsll.vv v8, v16, v8, v0.t
 ; CHECK-NEXT:    vnot.v v16, v24, v0.t
@@ -953,10 +953,10 @@ define <vscale x 8 x i64> @fshl_v8i64(<vscale x 8 x i64> %a, <vscale x 8 x i64>
 ; CHECK-NEXT:    .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x08, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 8 * vlenb
 ; CHECK-NEXT:    addi a2, sp, 16
 ; CHECK-NEXT:    vs8r.v v16, (a2) # Unknown-size Folded Spill
+; CHECK-NEXT:    vsetvli zero, a1, e64, m8, ta, ma
 ; CHECK-NEXT:    vmv8r.v v16, v8
 ; CHECK-NEXT:    vl8re64.v v24, (a0)
 ; CHECK-NEXT:    li a0, 63
-; CHECK-NEXT:    vsetvli zero, a1, e64, m8, ta, ma
 ; CHECK-NEXT:    vand.vx v8, v24, a0, v0.t
 ; CHECK-NEXT:    vsll.vv v8, v16, v8, v0.t
 ; CHECK-NEXT:    vnot.v v16, v24, v0.t
@@ -988,6 +988,7 @@ define <vscale x 16 x i64> @fshr_v16i64(<vscale x 16 x i64> %a, <vscale x 16 x i
 ; CHECK-NEXT:    mul a1, a1, a3
 ; CHECK-NEXT:    sub sp, sp, a1
 ; CHECK-NEXT:    .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x30, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 48 * vlenb
+; CHECK-NEXT:    vsetvli a1, zero, e8, mf4, ta, ma
 ; CHECK-NEXT:    vmv1r.v v24, v0
 ; CHECK-NEXT:    csrr a1, vlenb
 ; CHECK-NEXT:    li a3, 24
@@ -1022,7 +1023,6 @@ define <vscale x 16 x i64> @fshr_v16i64(<vscale x 16 x i64> %a, <vscale x 16 x i
 ; CHECK-NEXT:    add a3, sp, a3
 ; CHECK-NEXT:    addi a3, a3, 16
 ; CHECK-NEXT:    vs8r.v v16, (a3) # Unknown-size Folded Spill
-; CHECK-NEXT:    vsetvli a3, zero, e8, mf4, ta, ma
 ; CHECK-NEXT:    vslidedown.vx v0, v0, a6
 ; CHECK-NEXT:    li a3, 63
 ; CHECK-NEXT:    csrr a6, vlenb
@@ -1173,6 +1173,7 @@ define <vscale x 16 x i64> @fshl_v16i64(<vscale x 16 x i64> %a, <vscale x 16 x i
 ; CHECK-NEXT:    mul a1, a1, a3
 ; CHECK-NEXT:    sub sp, sp, a1
 ; CHECK-NEXT:    .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x28, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 40 * vlenb
+; CHECK-NEXT:    vsetvli a1, zero, e8, mf4, ta, ma
 ; CHECK-NEXT:    vmv1r.v v24, v0
 ; CHECK-NEXT:    csrr a1, vlenb
 ; CHECK-NEXT:    slli a1, a1, 5
@@ -1189,7 +1190,6 @@ define <vscale x 16 x i64> @fshl_v16i64(<vscale x 16 x i64> %a, <vscale x 16 x i
 ; CHECK-NEXT:    slli a5, a3, 3
 ; CHECK-NEXT:    srli a1, a3, 3
 ; CHECK-NEXT:    sub a6, a4, a3
-; CHECK-NEXT:    vsetvli a7, zero, e8, mf4, ta, ma
 ; CHECK-NEXT:    vslidedown.vx v0, v0, a1
 ; CHECK-NEXT:    add a1, a2, a5
 ; CHECK-NEXT:    vl8re64.v v8, (a1)
diff --git a/llvm/test/CodeGen/RISCV/rvv/inline-asm.ll b/llvm/test/CodeGen/RISCV/rvv/inline-asm.ll
index 967a58b45a599..daacc853963fc 100644
--- a/llvm/test/CodeGen/RISCV/rvv/inline-asm.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/inline-asm.ll
@@ -365,11 +365,13 @@ entry:
 define <vscale x 4 x i8> @test_specify_reg_mf2(<vscale x 4 x i8> %in, <vscale x 4 x i8> %in2) nounwind {
 ; CHECK-LABEL: test_specify_reg_mf2:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-NEXT:    vmv1r.v v2, v9
 ; CHECK-NEXT:    vmv1r.v v1, v8
 ; CHECK-NEXT:    #APP
 ; CHECK-NEXT:    vadd.vv v0, v1, v2
 ; CHECK-NEXT:    #NO_APP
+; CHECK-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-NEXT:    vmv1r.v v8, v0
 ; CHECK-NEXT:    ret
 entry:
@@ -380,11 +382,13 @@ entry:
 define <vscale x 8 x i8> @test_specify_reg_m1(<vscale x 8 x i8> %in, <vscale x 8 x i8> %in2) nounwind {
 ; CHECK-LABEL: test_specify_reg_m1:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-NEXT:    vmv1r.v v2, v9
 ; CHECK-NEXT:    vmv1r.v v1, v8
 ; CHECK-NEXT:    #APP
 ; CHECK-NEXT:    vadd.vv v0, v1, v2
 ; CHECK-NEXT:    #NO_APP
+; CHECK-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-NEXT:    vmv1r.v v8, v0
 ; CHECK-NEXT:    ret
 entry:
@@ -395,11 +399,13 @@ entry:
 define <vscale x 16 x i8> @test_specify_reg_m2(<vscale x 16 x i8> %in, <vscale x 16 x i8> %in2) nounwind {
 ; CHECK-LABEL: test_specify_reg_m2:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-NEXT:    vmv2r.v v4, v10
 ; CHECK-NEXT:    vmv2r.v v2, v8
 ; CHECK-NEXT:    #APP
 ; CHECK-NEXT:    vadd.vv v0, v2, v4
 ; CHECK-NEXT:    #NO_APP
+; CHECK-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-NEXT:    vmv2r.v v8, v0
 ; CHECK-NEXT:    ret
 entry:
@@ -410,6 +416,7 @@ entry:
 define <vscale x 1 x i1> @test_specify_reg_mask(<vscale x 1 x i1> %in, <vscale x 1 x i1> %in2) nounwind {
 ; CHECK-LABEL: test_specify_reg_mask:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-NEXT:    vmv1r.v v2, v8
 ; CHECK-NEXT:    vmv1r.v v1, v0
 ; CHECK-NEXT:    #APP
diff --git a/llvm/test/CodeGen/RISCV/rvv/insert-subvector.ll b/llvm/test/CodeGen/RISCV/rvv/insert-subvector.ll
index 8925a9e0cee32..ca9cec921b3cd 100644
--- a/llvm/test/CodeGen/RISCV/rvv/insert-subvector.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/insert-subvector.ll
@@ -5,6 +5,7 @@
 define <vscale x 8 x i32> @insert_nxv8i32_nxv4i32_0(<vscale x 8 x i32> %vec, <vscale x 4 x i32> %subvec) {
 ; CHECK-LABEL: insert_nxv8i32_nxv4i32_0:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-NEXT:    vmv2r.v v8, v12
 ; CHECK-NEXT:    ret
   %v = call <vscale x 8 x i32> @llvm.vector.insert.nxv4i32.nxv8i32(<vscale x 8 x i32> %vec, <vscale x 4 x i32> %subvec, i64 0)
@@ -14,6 +15,7 @@ define <vscale x 8 x i32> @insert_nxv8i32_nxv4i32_0(<vscale x 8 x i32> %vec, <vs
 define <vscale x 8 x i32> @insert_nxv8i32_nxv4i32_4(<vscale x 8 x i32> %vec, <vscale x 4 x i32> %subvec) {
 ; CHECK-LABEL: insert_nxv8i32_nxv4i32_4:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-NEXT:    vmv2r.v v10, v12
 ; CHECK-NEXT:    ret
   %v = call <vscale x 8 x i32> @llvm.vector.insert.nxv4i32.nxv8i32(<vscale x 8 x i32> %vec, <vscale x 4 x i32> %subvec, i64 4)
@@ -23,6 +25,7 @@ define <vscale x 8 x i32> @insert_nxv8i32_nxv4i32_4(<vscale x 8 x i32> %vec, <vs
 define <vscale x 8 x i32> @insert_nxv8i32_nxv2i32_0(<vscale x 8 x i32> %vec, <vscale x 2 x i32> %subvec) {
 ; CHECK-LABEL: insert_nxv8i32_nxv2i32_0:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-NEXT:    vmv1r.v v8, v12
 ; CHECK-NEXT:    ret
   %v = call <vscale x 8 x i32> @llvm.vector.insert.nxv2i32.nxv8i32(<vscale x 8 x i32> %vec, <vscale x 2 x i32> %subvec, i64 0)
@@ -32,6 +35,7 @@ define <vscale x 8 x i32> @insert_nxv8i32_nxv2i32_0(<vscale x 8 x i32> %vec, <vs
 define <vscale x 8 x i32> @insert_nxv8i32_nxv2i32_2(<vscale x 8 x i32> %vec, <vscale x 2 x i32> %subvec) {
 ; CHECK-LABEL: insert_nxv8i32_nxv2i32_2:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-NEXT:    vmv1r.v v9, v12
 ; CHECK-NEXT:    ret
   %v = call <vscale x 8 x i32> @llvm.vector.insert.nxv2i32.nxv8i32(<vscale x 8 x i32> %vec, <vscale x 2 x i32> %subvec, i64 2)
@@ -41,6 +45,7 @@ define <vscale x 8 x i32> @insert_nxv8i32_nxv2i32_2(<vscale x 8 x i32> %vec, <vs
 define <vscale x 8 x i32> @insert_nxv8i32_nxv2i32_4(<vscale x 8 x i32> %vec, <vscale x 2 x i32> %subvec) {
 ; CHECK-LABEL: insert_nxv8i32_nxv2i32_4:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-NEXT:    vmv1r.v v10, v12
 ; CHECK-NEXT:    ret
   %v = call <vscale x 8 x i32> @llvm.vector.insert.nxv2i32.nxv8i32(<vscale x 8 x i32> %vec, <vscale x 2 x i32> %subvec, i64 4)
@@ -50,6 +55,7 @@ define <vscale x 8 x i32> @insert_nxv8i32_nxv2i32_4(<vscale x 8 x i32> %vec, <vs
 define <vscale x 8 x i32> @insert_nxv8i32_nxv2i32_6(<vscale x 8 x i32> %vec, <vscale x 2 x i32> %subvec) {
 ; CHECK-LABEL: insert_nxv8i32_nxv2i32_6:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-NEXT:    vmv1r.v v11, v12
 ; CHECK-NEXT:    ret
   %v = call <vscale x 8 x i32> @llvm.vector.insert.nxv2i32.nxv8i32(<vscale x 8 x i32> %vec, <vscale x 2 x i32> %subvec, i64 6)
@@ -86,6 +92,7 @@ define <vscale x 4 x i8> @insert_nxv1i8_nxv4i8_3(<vscale x 4 x i8> %vec, <vscale
 define <vscale x 16 x i32> @insert_nxv16i32_nxv8i32_0(<vscale x 16 x i32> %vec, <vscale x 8 x i32> %subvec) {
 ; CHECK-LABEL: insert_nxv16i32_nxv8i32_0:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-NEXT:    vmv4r.v v8, v16
 ; CHECK-NEXT:    ret
   %v = call <vscale x 16 x i32> @llvm.vector.insert.nxv8i32.nxv16i32(<vscale x 16 x i32> %vec, <vscale x 8 x i32> %subvec, i64 0)
@@ -95,6 +102,7 @@ define <vscale x 16 x i32> @insert_nxv16i32_nxv8i32_0(<vscale x 16 x i32> %vec,
 define <vscale x 16 x i32> @insert_nxv16i32_nxv8i32_8(<vscale x 16 x i32> %vec, <vscale x 8 x i32> %subvec) {
 ; CHECK-LABEL: insert_nxv16i32_nxv8i32_8:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-NEXT:    vmv4r.v v12, v16
 ; CHECK-NEXT:    ret
   %v = call <vscale x 16 x i32> @llvm.vector.insert.nxv8i32.nxv16i32(<vscale x 16 x i32> %vec, <vscale x 8 x i32> %subvec, i64 8)
@@ -104,6 +112,7 @@ define <vscale x 16 x i32> @insert_nxv16i32_nxv8i32_8(<vscale x 16 x i32> %vec,
 define <vscale x 16 x i32> @insert_nxv16i32_nxv4i32_0(<vscale x 16 x i32> %vec, <vscale x 4 x i32> %subvec) {
 ; CHECK-LABEL: insert_nxv16i32_nxv4i32_0:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-NEXT:    vmv2r.v v8, v16
 ; CHECK-NEXT:    ret
   %v = call <vscale x 16 x i32> @llvm.vector.insert.nxv4i32.nxv16i32(<vscale x 16 x i32> %vec, <vscale x 4 x i32> %subvec, i64 0)
@@ -113,6 +122,7 @@ define <vscale x 16 x i32> @insert_nxv16i32_nxv4i32_0(<vscale x 16 x i32> %vec,
 define <vscale x 16 x i32> @insert_nxv16i32_nxv4i32_4(<vscale x 16 x i32> %vec, <vscale x 4 x i32> %subvec) {
 ; CHECK-LABEL: insert_nxv16i32_nxv4i32_4:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-NEXT:    vmv2r.v v10, v16
 ; CHECK-NEXT:    ret
   %v = call <vscale x 16 x i32> @llvm.vector.insert.nxv4i32.nxv16i32(<vscale x 16 x i32> %vec, <vscale x 4 x i32> %subvec, i64 4)
@@ -122,6 +132,7 @@ define <vscale x 16 x i32> @insert_nxv16i32_nxv4i32_4(<vscale x 16 x i32> %vec,
 define <vscale x 16 x i32> @insert_nxv16i32_nxv4i32_8(<vscale x 16 x i32> %vec, <vscale x 4 x i32> %subvec) {
 ; CHECK-LABEL: insert_nxv16i32_nxv4i32_8:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-NEXT:    vmv2r.v v12, v16
 ; CHECK-NEXT:    ret
   %v = call <vscale x 16 x i32> @llvm.vector.insert.nxv4i32.nxv16i32(<vscale x 16 x i32> %vec, <vscale x 4 x i32> %subvec, i64 8)
@@ -131,6 +142,7 @@ define <vscale x 16 x i32> @insert_nxv16i32_nxv4i32_8(<vscale x 16 x i32> %vec,
 define <vscale x 16 x i32> @insert_nxv16i32_nxv4i32_12(<vscale x 16 x i32> %vec, <vscale x 4 x i32> %subvec) {
 ; CHECK-LABEL: insert_nxv16i32_nxv4i32_12:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-NEXT:    vmv2r.v v14, v16
 ; CHECK-NEXT:    ret
   %v = call <vscale x 16 x i32> @llvm.vector.insert.nxv4i32.nxv16i32(<vscale x 16 x i32> %vec, <vscale x 4 x i32> %subvec, i64 12)
@@ -140,6 +152,7 @@ define <vscale x 16 x i32> @insert_nxv16i32_nxv4i32_12(<vscale x 16 x i32> %vec,
 define <vscale x 16 x i32> @insert_nxv16i32_nxv2i32_0(<vscale x 16 x i32> %vec, <vscale x 2 x i32> %subvec) {
 ; CHECK-LABEL: insert_nxv16i32_nxv2i32_0:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-NEXT:    vmv1r.v v8, v16
 ; CHECK-NEXT:    ret
   %v = call <vscale x 16 x i32> @llvm.vector.insert.nxv2i32.nxv16i32(<vscale x 16 x i32> %vec, <vscale x 2 x i32> %subvec, i64 0)
@@ -149,6 +162,7 @@ define <vscale x 16 x i32> @insert_nxv16i32_nxv2i32_0(<vscale x 16 x i32> %vec,
 define <vscale x 16 x i32> @insert_nxv16i32_nxv2i32_2(<vscale x 16 x i32> %vec, <vscale x 2 x i32> %subvec) {
 ; CHECK-LABEL: insert_nxv16i32_nxv2i32_2:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-NEXT:    vmv1r.v v9, v16
 ; CHECK-NEXT:    ret
   %v = call <vscale x 16 x i32> @llvm.vector.insert.nxv2i32.nxv16i32(<vscale x 16 x i32> %vec, <vscale x 2 x i32> %subvec, i64 2)
@@ -158,6 +172,7 @@ define <vscale x 16 x i32> @insert_nxv16i32_nxv2i32_2(<vscale x 16 x i32> %vec,
 define <vscale x 16 x i32> @insert_nxv16i32_nxv2i32_4(<vscale x 16 x i32> %vec, <vscale x 2 x i32> %subvec) {
 ; CHECK-LABEL: insert_nxv16i32_nxv2i32_4:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-NEXT:    vmv1r.v v10, v16
 ; CHECK-NEXT:    ret
   %v = call <vscale x 16 x i32> @llvm.vector.insert.nxv2i32.nxv16i32(<vscale x 16 x i32> %vec, <vscale x 2 x i32> %subvec, i64 4)
@@ -167,6 +182,7 @@ define <vscale x 16 x i32> @insert_nxv16i32_nxv2i32_4(<vscale x 16 x i32> %vec,
 define <vscale x 16 x i32> @insert_nxv16i32_nxv2i32_6(<vscale x 16 x i32> %vec, <vscale x 2 x i32> %subvec) {
 ; CHECK-LABEL: insert_nxv16i32_nxv2i32_6:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-NEXT:    vmv1r.v v11, v16
 ; CHECK-NEXT:    ret
   %v = call <vscale x 16 x i32> @llvm.vector.insert.nxv2i32.nxv16i32(<vscale x 16 x i32> %vec, <vscale x 2 x i32> %subvec, i64 6)
@@ -176,6 +192,7 @@ define <vscale x 16 x i32> @insert_nxv16i32_nxv2i32_6(<vscale x 16 x i32> %vec,
 define <vscale x 16 x i32> @insert_nxv16i32_nxv2i32_8(<vscale x 16 x i32> %vec, <vscale x 2 x i32> %subvec) {
 ; CHECK-LABEL: insert_nxv16i32_nxv2i32_8:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-NEXT:    vmv1r.v v12, v16
 ; CHECK-NEXT:    ret
   %v = call <vscale x 16 x i32> @llvm.vector.insert.nxv2i32.nxv16i32(<vscale x 16 x i32> %vec, <vscale x 2 x i32> %subvec, i64 8)
@@ -185,6 +202,7 @@ define <vscale x 16 x i32> @insert_nxv16i32_nxv2i32_8(<vscale x 16 x i32> %vec,
 define <vscale x 16 x i32> @insert_nxv16i32_nxv2i32_10(<vscale x 16 x i32> %vec, <vscale x 2 x i32> %subvec) {
 ; CHECK-LABEL: insert_nxv16i32_nxv2i32_10:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-NEXT:    vmv1r.v v13, v16
 ; CHECK-NEXT:    ret
   %v = call <vscale x 16 x i32> @llvm.vector.insert.nxv2i32.nxv16i32(<vscale x 16 x i32> %vec, <vscale x 2 x i32> %subvec, i64 10)
@@ -194,6 +212,7 @@ define <vscale x 16 x i32> @insert_nxv16i32_nxv2i32_10(<vscale x 16 x i32> %vec,
 define <vscale x 16 x i32> @insert_nxv16i32_nxv2i32_12(<vscale x 16 x i32> %vec, <vscale x 2 x i32> %subvec) {
 ; CHECK-LABEL: insert_nxv16i32_nxv2i32_12:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-NEXT:    vmv1r.v v14, v16
 ; CHECK-NEXT:    ret
   %v = call <vscale x 16 x i32> @llvm.vector.insert.nxv2i32.nxv16i32(<vscale x 16 x i32> %vec, <vscale x 2 x i32> %subvec, i64 12)
@@ -203,6 +222,7 @@ define <vscale x 16 x i32> @insert_nxv16i32_nxv2i32_12(<vscale x 16 x i32> %vec,
 define <vscale x 16 x i32> @insert_nxv16i32_nxv2i32_14(<vscale x 16 x i32> %vec, <vscale x 2 x i32> %subvec) {
 ; CHECK-LABEL: insert_nxv16i32_nxv2i32_14:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-NEXT:    vmv1r.v v15, v16
 ; CHECK-NEXT:    ret
   %v = call <vscale x 16 x i32> @llvm.vector.insert.nxv2i32.nxv16i32(<vscale x 16 x i32> %vec, <vscale x 2 x i32> %subvec, i64 14)
@@ -512,6 +532,7 @@ define <vscale x 2 x i64> @insert_nxv2i64_nxv3i64(<3 x i64> %sv) #0 {
 define <vscale x 8 x i32> @insert_insert_combine(<2 x i32> %subvec) {
 ; CHECK-LABEL: insert_insert_combine:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-NEXT:    vmv1r.v v10, v8
 ; CHECK-NEXT:    ret
   %inner = call <vscale x 4 x i32> @llvm.vector.insert.nxv4i32.v2i32(<vscale x 4 x i32> undef, <2 x i32> %subvec, i64 0)
@@ -524,6 +545,7 @@ define <vscale x 8 x i32> @insert_insert_combine(<2 x i32> %subvec) {
 define <vscale x 8 x i32> @insert_insert_combine2(<vscale x 2 x i32> %subvec) {
 ; CHECK-LABEL: insert_insert_combine2:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-NEXT:    vmv1r.v v10, v8
 ; CHECK-NEXT:    ret
   %inner = call <vscale x 4 x i32> @llvm.vector.insert.nxv2i32.nxv4i32(<vscale x 4 x i32> undef, <vscale x 2 x i32> %subvec, i64 0)
diff --git a/llvm/test/CodeGen/RISCV/rvv/llrint-vp.ll b/llvm/test/CodeGen/RISCV/rvv/llrint-vp.ll
index ffb9bf76fb4fa..166dba6a56524 100644
--- a/llvm/test/CodeGen/RISCV/rvv/llrint-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/llrint-vp.ll
@@ -55,11 +55,11 @@ declare <vscale x 8 x i64> @llvm.vp.llrint.nxv8i64.nxv8f32(<vscale x 8 x float>,
 define <vscale x 16 x i64> @llrint_nxv16i64_nxv16f32(<vscale x 16 x float> %x, <vscale x 16 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: llrint_nxv16i64_nxv16f32:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli a1, zero, e8, mf4, ta, ma
 ; CHECK-NEXT:    vmv1r.v v24, v0
 ; CHECK-NEXT:    csrr a1, vlenb
 ; CHECK-NEXT:    srli a2, a1, 3
 ; CHECK-NEXT:    sub a3, a0, a1
-; CHECK-NEXT:    vsetvli a4, zero, e8, mf4, ta, ma
 ; CHECK-NEXT:    vslidedown.vx v0, v0, a2
 ; CHECK-NEXT:    sltu a2, a0, a3
 ; CHECK-NEXT:    addi a2, a2, -1
diff --git a/llvm/test/CodeGen/RISCV/rvv/lrint-vp.ll b/llvm/test/CodeGen/RISCV/rvv/lrint-vp.ll
index 9991bbc9725ba..21045b69a8b5d 100644
--- a/llvm/test/CodeGen/RISCV/rvv/lrint-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/lrint-vp.ll
@@ -117,11 +117,11 @@ define <vscale x 16 x iXLen> @lrint_nxv16f32(<vscale x 16 x float> %x, <vscale x
 ;
 ; RV64-i64-LABEL: lrint_nxv16f32:
 ; RV64-i64:       # %bb.0:
+; RV64-i64-NEXT:    vsetvli a1, zero, e8, mf4, ta, ma
 ; RV64-i64-NEXT:    vmv1r.v v24, v0
 ; RV64-i64-NEXT:    csrr a1, vlenb
 ; RV64-i64-NEXT:    srli a2, a1, 3
 ; RV64-i64-NEXT:    sub a3, a0, a1
-; RV64-i64-NEXT:    vsetvli a4, zero, e8, mf4, ta, ma
 ; RV64-i64-NEXT:    vslidedown.vx v0, v0, a2
 ; RV64-i64-NEXT:    sltu a2, a0, a3
 ; RV64-i64-NEXT:    addi a2, a2, -1
diff --git a/llvm/test/CodeGen/RISCV/rvv/masked-tama.ll b/llvm/test/CodeGen/RISCV/rvv/masked-tama.ll
index 420597b009f33..cd678c9740721 100644
--- a/llvm/test/CodeGen/RISCV/rvv/masked-tama.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/masked-tama.ll
@@ -1288,9 +1288,9 @@ declare <vscale x 1 x i8> @llvm.riscv.viota.mask.nxv1i8(
 define <vscale x 1 x i8> @intrinsic_viota_mask_m_nxv1i8_nxv1i1(<vscale x 1 x i1> %0, <vscale x 1 x i1> %1, iXLen %2) nounwind {
 ; CHECK-LABEL: intrinsic_viota_mask_m_nxv1i8_nxv1i1:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e8, mf8, ta, ma
 ; CHECK-NEXT:    vmv1r.v v9, v0
 ; CHECK-NEXT:    vmv1r.v v0, v8
-; CHECK-NEXT:    vsetvli zero, a0, e8, mf8, ta, ma
 ; CHECK-NEXT:    viota.m v8, v9, v0.t
 ; CHECK-NEXT:    ret
 entry:
@@ -1312,9 +1312,9 @@ declare <vscale x 1 x i1> @llvm.riscv.vmsbf.mask.nxv1i1(
 define <vscale x 1 x i1> @intrinsic_vmsbf_mask_m_nxv1i1_nxv1i1(<vscale x 1 x i1> %0, <vscale x 1 x i1> %1, iXLen %2) nounwind {
 ; CHECK-LABEL: intrinsic_vmsbf_mask_m_nxv1i1_nxv1i1:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e8, mf8, ta, ma
 ; CHECK-NEXT:    vmv1r.v v9, v0
 ; CHECK-NEXT:    vmv1r.v v0, v8
-; CHECK-NEXT:    vsetvli zero, a0, e8, mf8, ta, ma
 ; CHECK-NEXT:    vmsbf.m v8, v9, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v8
 ; CHECK-NEXT:    ret
@@ -1443,9 +1443,9 @@ declare <vscale x 64 x i1> @llvm.riscv.vmsbf.mask.nxv64i1(
 define <vscale x 64 x i1> @intrinsic_vmsbf_mask_m_nxv64i1_nxv64i1(<vscale x 64 x i1> %0, <vscale x 64 x i1> %1, iXLen %2) nounwind {
 ; CHECK-LABEL: intrinsic_vmsbf_mask_m_nxv64i1_nxv64i1:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e8, m8, ta, ma
 ; CHECK-NEXT:    vmv1r.v v9, v0
 ; CHECK-NEXT:    vmv1r.v v0, v8
-; CHECK-NEXT:    vsetvli zero, a0, e8, m8, ta, ma
 ; CHECK-NEXT:    vmsbf.m v8, v9, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v8
 ; CHECK-NEXT:    ret
diff --git a/llvm/test/CodeGen/RISCV/rvv/mgather-sdnode.ll b/llvm/test/CodeGen/RISCV/rvv/mgather-sdnode.ll
index a3eddbcc2baed..9ee2324f615dd 100644
--- a/llvm/test/CodeGen/RISCV/rvv/mgather-sdnode.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/mgather-sdnode.ll
@@ -221,11 +221,13 @@ define <vscale x 4 x i8> @mgather_truemask_nxv4i8(<vscale x 4 x ptr> %ptrs, <vsc
 define <vscale x 4 x i8> @mgather_falsemask_nxv4i8(<vscale x 4 x ptr> %ptrs, <vscale x 4 x i8> %passthru) {
 ; RV32-LABEL: mgather_falsemask_nxv4i8:
 ; RV32:       # %bb.0:
+; RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; RV32-NEXT:    vmv1r.v v8, v10
 ; RV32-NEXT:    ret
 ;
 ; RV64-LABEL: mgather_falsemask_nxv4i8:
 ; RV64:       # %bb.0:
+; RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; RV64-NEXT:    vmv1r.v v8, v12
 ; RV64-NEXT:    ret
   %v = call <vscale x 4 x i8> @llvm.masked.gather.nxv4i8.nxv4p0(<vscale x 4 x ptr> %ptrs, i32 1, <vscale x 4 x i1> zeroinitializer, <vscale x 4 x i8> %passthru)
@@ -442,11 +444,13 @@ define <vscale x 4 x i16> @mgather_truemask_nxv4i16(<vscale x 4 x ptr> %ptrs, <v
 define <vscale x 4 x i16> @mgather_falsemask_nxv4i16(<vscale x 4 x ptr> %ptrs, <vscale x 4 x i16> %passthru) {
 ; RV32-LABEL: mgather_falsemask_nxv4i16:
 ; RV32:       # %bb.0:
+; RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; RV32-NEXT:    vmv1r.v v8, v10
 ; RV32-NEXT:    ret
 ;
 ; RV64-LABEL: mgather_falsemask_nxv4i16:
 ; RV64:       # %bb.0:
+; RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; RV64-NEXT:    vmv1r.v v8, v12
 ; RV64-NEXT:    ret
   %v = call <vscale x 4 x i16> @llvm.masked.gather.nxv4i16.nxv4p0(<vscale x 4 x ptr> %ptrs, i32 2, <vscale x 4 x i1> zeroinitializer, <vscale x 4 x i16> %passthru)
@@ -686,11 +690,13 @@ define <vscale x 4 x i32> @mgather_truemask_nxv4i32(<vscale x 4 x ptr> %ptrs, <v
 define <vscale x 4 x i32> @mgather_falsemask_nxv4i32(<vscale x 4 x ptr> %ptrs, <vscale x 4 x i32> %passthru) {
 ; RV32-LABEL: mgather_falsemask_nxv4i32:
 ; RV32:       # %bb.0:
+; RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; RV32-NEXT:    vmv2r.v v8, v10
 ; RV32-NEXT:    ret
 ;
 ; RV64-LABEL: mgather_falsemask_nxv4i32:
 ; RV64:       # %bb.0:
+; RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; RV64-NEXT:    vmv2r.v v8, v12
 ; RV64-NEXT:    ret
   %v = call <vscale x 4 x i32> @llvm.masked.gather.nxv4i32.nxv4p0(<vscale x 4 x ptr> %ptrs, i32 4, <vscale x 4 x i1> zeroinitializer, <vscale x 4 x i32> %passthru)
@@ -949,6 +955,7 @@ define <vscale x 4 x i64> @mgather_truemask_nxv4i64(<vscale x 4 x ptr> %ptrs, <v
 define <vscale x 4 x i64> @mgather_falsemask_nxv4i64(<vscale x 4 x ptr> %ptrs, <vscale x 4 x i64> %passthru) {
 ; CHECK-LABEL: mgather_falsemask_nxv4i64:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-NEXT:    vmv4r.v v8, v12
 ; CHECK-NEXT:    ret
   %v = call <vscale x 4 x i64> @llvm.masked.gather.nxv4i64.nxv4p0(<vscale x 4 x ptr> %ptrs, i32 8, <vscale x 4 x i1> zeroinitializer, <vscale x 4 x i64> %passthru)
@@ -1232,12 +1239,12 @@ define void @mgather_nxv16i64(<vscale x 8 x ptr> %ptrs0, <vscale x 8 x ptr> %ptr
 ; RV64-NEXT:    .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x08, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 8 * vlenb
 ; RV64-NEXT:    addi a3, sp, 16
 ; RV64-NEXT:    vs8r.v v16, (a3) # Unknown-size Folded Spill
+; RV64-NEXT:    vsetvli a3, zero, e8, mf4, ta, ma
 ; RV64-NEXT:    vmv8r.v v16, v8
 ; RV64-NEXT:    vl8re64.v v24, (a0)
 ; RV64-NEXT:    csrr a0, vlenb
 ; RV64-NEXT:    vl8re64.v v8, (a1)
 ; RV64-NEXT:    srli a1, a0, 3
-; RV64-NEXT:    vsetvli a3, zero, e8, mf4, ta, ma
 ; RV64-NEXT:    vslidedown.vx v7, v0, a1
 ; RV64-NEXT:    vsetvli a1, zero, e64, m8, ta, mu
 ; RV64-NEXT:    vluxei64.v v24, (zero), v16, v0.t
@@ -1348,11 +1355,13 @@ define <vscale x 4 x bfloat> @mgather_truemask_nxv4bf16(<vscale x 4 x ptr> %ptrs
 define <vscale x 4 x bfloat> @mgather_falsemask_nxv4bf16(<vscale x 4 x ptr> %ptrs, <vscale x 4 x bfloat> %passthru) {
 ; RV32-LABEL: mgather_falsemask_nxv4bf16:
 ; RV32:       # %bb.0:
+; RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; RV32-NEXT:    vmv1r.v v8, v10
 ; RV32-NEXT:    ret
 ;
 ; RV64-LABEL: mgather_falsemask_nxv4bf16:
 ; RV64:       # %bb.0:
+; RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; RV64-NEXT:    vmv1r.v v8, v12
 ; RV64-NEXT:    ret
   %v = call <vscale x 4 x bfloat> @llvm.masked.gather.nxv4bf16.nxv4p0(<vscale x 4 x ptr> %ptrs, i32 2, <vscale x 4 x i1> zeroinitializer, <vscale x 4 x bfloat> %passthru)
@@ -1549,11 +1558,13 @@ define <vscale x 4 x half> @mgather_truemask_nxv4f16(<vscale x 4 x ptr> %ptrs, <
 define <vscale x 4 x half> @mgather_falsemask_nxv4f16(<vscale x 4 x ptr> %ptrs, <vscale x 4 x half> %passthru) {
 ; RV32-LABEL: mgather_falsemask_nxv4f16:
 ; RV32:       # %bb.0:
+; RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; RV32-NEXT:    vmv1r.v v8, v10
 ; RV32-NEXT:    ret
 ;
 ; RV64-LABEL: mgather_falsemask_nxv4f16:
 ; RV64:       # %bb.0:
+; RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; RV64-NEXT:    vmv1r.v v8, v12
 ; RV64-NEXT:    ret
   %v = call <vscale x 4 x half> @llvm.masked.gather.nxv4f16.nxv4p0(<vscale x 4 x ptr> %ptrs, i32 2, <vscale x 4 x i1> zeroinitializer, <vscale x 4 x half> %passthru)
@@ -1749,11 +1760,13 @@ define <vscale x 4 x float> @mgather_truemask_nxv4f32(<vscale x 4 x ptr> %ptrs,
 define <vscale x 4 x float> @mgather_falsemask_nxv4f32(<vscale x 4 x ptr> %ptrs, <vscale x 4 x float> %passthru) {
 ; RV32-LABEL: mgather_falsemask_nxv4f32:
 ; RV32:       # %bb.0:
+; RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; RV32-NEXT:    vmv2r.v v8, v10
 ; RV32-NEXT:    ret
 ;
 ; RV64-LABEL: mgather_falsemask_nxv4f32:
 ; RV64:       # %bb.0:
+; RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; RV64-NEXT:    vmv2r.v v8, v12
 ; RV64-NEXT:    ret
   %v = call <vscale x 4 x float> @llvm.masked.gather.nxv4f32.nxv4p0(<vscale x 4 x ptr> %ptrs, i32 4, <vscale x 4 x i1> zeroinitializer, <vscale x 4 x float> %passthru)
@@ -2012,6 +2025,7 @@ define <vscale x 4 x double> @mgather_truemask_nxv4f64(<vscale x 4 x ptr> %ptrs,
 define <vscale x 4 x double> @mgather_falsemask_nxv4f64(<vscale x 4 x ptr> %ptrs, <vscale x 4 x double> %passthru) {
 ; CHECK-LABEL: mgather_falsemask_nxv4f64:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-NEXT:    vmv4r.v v8, v12
 ; CHECK-NEXT:    ret
   %v = call <vscale x 4 x double> @llvm.masked.gather.nxv4f64.nxv4p0(<vscale x 4 x ptr> %ptrs, i32 8, <vscale x 4 x i1> zeroinitializer, <vscale x 4 x double> %passthru)
@@ -2317,8 +2331,8 @@ define <vscale x 32 x i8> @mgather_baseidx_nxv32i8(ptr %base, <vscale x 32 x i8>
 ;
 ; RV64-LABEL: mgather_baseidx_nxv32i8:
 ; RV64:       # %bb.0:
-; RV64-NEXT:    vmv1r.v v16, v0
 ; RV64-NEXT:    vsetvli a1, zero, e64, m8, ta, ma
+; RV64-NEXT:    vmv1r.v v16, v0
 ; RV64-NEXT:    vsext.vf8 v24, v8
 ; RV64-NEXT:    csrr a1, vlenb
 ; RV64-NEXT:    vsetvli zero, zero, e8, m1, ta, mu
diff --git a/llvm/test/CodeGen/RISCV/rvv/mscatter-sdnode.ll b/llvm/test/CodeGen/RISCV/rvv/mscatter-sdnode.ll
index 72c251ce985cb..77a1f508d2218 100644
--- a/llvm/test/CodeGen/RISCV/rvv/mscatter-sdnode.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/mscatter-sdnode.ll
@@ -2009,11 +2009,11 @@ define void @mscatter_baseidx_nxv16i16_nxv16f64(<vscale x 8 x double> %val0, <vs
 ; RV32-NEXT:    .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x08, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 8 * vlenb
 ; RV32-NEXT:    addi a2, sp, 16
 ; RV32-NEXT:    vs8r.v v16, (a2) # Unknown-size Folded Spill
+; RV32-NEXT:    vsetvli a2, zero, e8, mf4, ta, ma
 ; RV32-NEXT:    vmv8r.v v16, v8
 ; RV32-NEXT:    vl4re16.v v8, (a1)
 ; RV32-NEXT:    csrr a1, vlenb
 ; RV32-NEXT:    srli a1, a1, 3
-; RV32-NEXT:    vsetvli a2, zero, e8, mf4, ta, ma
 ; RV32-NEXT:    vslidedown.vx v7, v0, a1
 ; RV32-NEXT:    vsetvli a1, zero, e32, m8, ta, ma
 ; RV32-NEXT:    vsext.vf2 v24, v8
diff --git a/llvm/test/CodeGen/RISCV/rvv/named-vector-shuffle-reverse.ll b/llvm/test/CodeGen/RISCV/rvv/named-vector-shuffle-reverse.ll
index c49a55319a3c4..9298e8b520bd9 100644
--- a/llvm/test/CodeGen/RISCV/rvv/named-vector-shuffle-reverse.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/named-vector-shuffle-reverse.ll
@@ -1169,10 +1169,10 @@ define <vscale x 32 x i8> @reverse_nxv32i8(<vscale x 32 x i8> %a) {
 define <vscale x 64 x i8> @reverse_nxv64i8(<vscale x 64 x i8> %a) {
 ; RV32-BITS-UNKNOWN-LABEL: reverse_nxv64i8:
 ; RV32-BITS-UNKNOWN:       # %bb.0:
+; RV32-BITS-UNKNOWN-NEXT:    vsetvli a0, zero, e16, m2, ta, ma
 ; RV32-BITS-UNKNOWN-NEXT:    vmv8r.v v16, v8
 ; RV32-BITS-UNKNOWN-NEXT:    csrr a0, vlenb
 ; RV32-BITS-UNKNOWN-NEXT:    addi a0, a0, -1
-; RV32-BITS-UNKNOWN-NEXT:    vsetvli a1, zero, e16, m2, ta, ma
 ; RV32-BITS-UNKNOWN-NEXT:    vid.v v8
 ; RV32-BITS-UNKNOWN-NEXT:    vrsub.vx v24, v8, a0
 ; RV32-BITS-UNKNOWN-NEXT:    vsetvli zero, zero, e8, m1, ta, ma
@@ -1188,10 +1188,10 @@ define <vscale x 64 x i8> @reverse_nxv64i8(<vscale x 64 x i8> %a) {
 ;
 ; RV32-BITS-256-LABEL: reverse_nxv64i8:
 ; RV32-BITS-256:       # %bb.0:
+; RV32-BITS-256-NEXT:    vsetvli a0, zero, e8, m1, ta, ma
 ; RV32-BITS-256-NEXT:    vmv8r.v v16, v8
 ; RV32-BITS-256-NEXT:    csrr a0, vlenb
 ; RV32-BITS-256-NEXT:    addi a0, a0, -1
-; RV32-BITS-256-NEXT:    vsetvli a1, zero, e8, m1, ta, ma
 ; RV32-BITS-256-NEXT:    vid.v v8
 ; RV32-BITS-256-NEXT:    vrsub.vx v24, v8, a0
 ; RV32-BITS-256-NEXT:    vrgather.vv v15, v16, v24
@@ -1206,10 +1206,10 @@ define <vscale x 64 x i8> @reverse_nxv64i8(<vscale x 64 x i8> %a) {
 ;
 ; RV32-BITS-512-LABEL: reverse_nxv64i8:
 ; RV32-BITS-512:       # %bb.0:
+; RV32-BITS-512-NEXT:    vsetvli a0, zero, e8, m1, ta, ma
 ; RV32-BITS-512-NEXT:    vmv8r.v v16, v8
 ; RV32-BITS-512-NEXT:    csrr a0, vlenb
 ; RV32-BITS-512-NEXT:    addi a0, a0, -1
-; RV32-BITS-512-NEXT:    vsetvli a1, zero, e8, m1, ta, ma
 ; RV32-BITS-512-NEXT:    vid.v v8
 ; RV32-BITS-512-NEXT:    vrsub.vx v24, v8, a0
 ; RV32-BITS-512-NEXT:    vrgather.vv v15, v16, v24
@@ -1224,10 +1224,10 @@ define <vscale x 64 x i8> @reverse_nxv64i8(<vscale x 64 x i8> %a) {
 ;
 ; RV64-BITS-UNKNOWN-LABEL: reverse_nxv64i8:
 ; RV64-BITS-UNKNOWN:       # %bb.0:
+; RV64-BITS-UNKNOWN-NEXT:    vsetvli a0, zero, e16, m2, ta, ma
 ; RV64-BITS-UNKNOWN-NEXT:    vmv8r.v v16, v8
 ; RV64-BITS-UNKNOWN-NEXT:    csrr a0, vlenb
 ; RV64-BITS-UNKNOWN-NEXT:    addi a0, a0, -1
-; RV64-BITS-UNKNOWN-NEXT:    vsetvli a1, zero, e16, m2, ta, ma
 ; RV64-BITS-UNKNOWN-NEXT:    vid.v v8
 ; RV64-BITS-UNKNOWN-NEXT:    vrsub.vx v24, v8, a0
 ; RV64-BITS-UNKNOWN-NEXT:    vsetvli zero, zero, e8, m1, ta, ma
@@ -1243,10 +1243,10 @@ define <vscale x 64 x i8> @reverse_nxv64i8(<vscale x 64 x i8> %a) {
 ;
 ; RV64-BITS-256-LABEL: reverse_nxv64i8:
 ; RV64-BITS-256:       # %bb.0:
+; RV64-BITS-256-NEXT:    vsetvli a0, zero, e8, m1, ta, ma
 ; RV64-BITS-256-NEXT:    vmv8r.v v16, v8
 ; RV64-BITS-256-NEXT:    csrr a0, vlenb
 ; RV64-BITS-256-NEXT:    addi a0, a0, -1
-; RV64-BITS-256-NEXT:    vsetvli a1, zero, e8, m1, ta, ma
 ; RV64-BITS-256-NEXT:    vid.v v8
 ; RV64-BITS-256-NEXT:    vrsub.vx v24, v8, a0
 ; RV64-BITS-256-NEXT:    vrgather.vv v15, v16, v24
@@ -1261,10 +1261,10 @@ define <vscale x 64 x i8> @reverse_nxv64i8(<vscale x 64 x i8> %a) {
 ;
 ; RV64-BITS-512-LABEL: reverse_nxv64i8:
 ; RV64-BITS-512:       # %bb.0:
+; RV64-BITS-512-NEXT:    vsetvli a0, zero, e8, m1, ta, ma
 ; RV64-BITS-512-NEXT:    vmv8r.v v16, v8
 ; RV64-BITS-512-NEXT:    csrr a0, vlenb
 ; RV64-BITS-512-NEXT:    addi a0, a0, -1
-; RV64-BITS-512-NEXT:    vsetvli a1, zero, e8, m1, ta, ma
 ; RV64-BITS-512-NEXT:    vid.v v8
 ; RV64-BITS-512-NEXT:    vrsub.vx v24, v8, a0
 ; RV64-BITS-512-NEXT:    vrgather.vv v15, v16, v24
@@ -1367,11 +1367,11 @@ define <vscale x 16 x i16> @reverse_nxv16i16(<vscale x 16 x i16> %a) {
 define <vscale x 32 x i16> @reverse_nxv32i16(<vscale x 32 x i16> %a) {
 ; CHECK-LABEL: reverse_nxv32i16:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli a0, zero, e16, m1, ta, ma
 ; CHECK-NEXT:    vmv8r.v v16, v8
 ; CHECK-NEXT:    csrr a0, vlenb
 ; CHECK-NEXT:    srli a0, a0, 1
 ; CHECK-NEXT:    addi a0, a0, -1
-; CHECK-NEXT:    vsetvli a1, zero, e16, m1, ta, ma
 ; CHECK-NEXT:    vid.v v8
 ; CHECK-NEXT:    vrsub.vx v24, v8, a0
 ; CHECK-NEXT:    vrgather.vv v15, v16, v24
@@ -1458,11 +1458,11 @@ define <vscale x 8 x i32> @reverse_nxv8i32(<vscale x 8 x i32> %a) {
 define <vscale x 16 x i32> @reverse_nxv16i32(<vscale x 16 x i32> %a) {
 ; CHECK-LABEL: reverse_nxv16i32:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli a0, zero, e32, m1, ta, ma
 ; CHECK-NEXT:    vmv8r.v v16, v8
 ; CHECK-NEXT:    csrr a0, vlenb
 ; CHECK-NEXT:    srli a0, a0, 2
 ; CHECK-NEXT:    addi a0, a0, -1
-; CHECK-NEXT:    vsetvli a1, zero, e32, m1, ta, ma
 ; CHECK-NEXT:    vid.v v8
 ; CHECK-NEXT:    vrsub.vx v24, v8, a0
 ; CHECK-NEXT:    vrgather.vv v15, v16, v24
@@ -1533,11 +1533,11 @@ define <vscale x 4 x i64> @reverse_nxv4i64(<vscale x 4 x i64> %a) {
 define <vscale x 8 x i64> @reverse_nxv8i64(<vscale x 8 x i64> %a) {
 ; CHECK-LABEL: reverse_nxv8i64:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli a0, zero, e64, m1, ta, ma
 ; CHECK-NEXT:    vmv8r.v v16, v8
 ; CHECK-NEXT:    csrr a0, vlenb
 ; CHECK-NEXT:    srli a0, a0, 3
 ; CHECK-NEXT:    addi a0, a0, -1
-; CHECK-NEXT:    vsetvli a1, zero, e64, m1, ta, ma
 ; CHECK-NEXT:    vid.v v8
 ; CHECK-NEXT:    vrsub.vx v24, v8, a0
 ; CHECK-NEXT:    vrgather.vv v15, v16, v24
@@ -1644,11 +1644,11 @@ define <vscale x 16 x bfloat> @reverse_nxv16bf16(<vscale x 16 x bfloat> %a) {
 define <vscale x 32 x bfloat> @reverse_nxv32bf16(<vscale x 32 x bfloat> %a) {
 ; CHECK-LABEL: reverse_nxv32bf16:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli a0, zero, e16, m1, ta, ma
 ; CHECK-NEXT:    vmv8r.v v16, v8
 ; CHECK-NEXT:    csrr a0, vlenb
 ; CHECK-NEXT:    srli a0, a0, 1
 ; CHECK-NEXT:    addi a0, a0, -1
-; CHECK-NEXT:    vsetvli a1, zero, e16, m1, ta, ma
 ; CHECK-NEXT:    vid.v v8
 ; CHECK-NEXT:    vrsub.vx v24, v8, a0
 ; CHECK-NEXT:    vrgather.vv v15, v16, v24
@@ -1751,11 +1751,11 @@ define <vscale x 16 x half> @reverse_nxv16f16(<vscale x 16 x half> %a) {
 define <vscale x 32 x half> @reverse_nxv32f16(<vscale x 32 x half> %a) {
 ; CHECK-LABEL: reverse_nxv32f16:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli a0, zero, e16, m1, ta, ma
 ; CHECK-NEXT:    vmv8r.v v16, v8
 ; CHECK-NEXT:    csrr a0, vlenb
 ; CHECK-NEXT:    srli a0, a0, 1
 ; CHECK-NEXT:    addi a0, a0, -1
-; CHECK-NEXT:    vsetvli a1, zero, e16, m1, ta, ma
 ; CHECK-NEXT:    vid.v v8
 ; CHECK-NEXT:    vrsub.vx v24, v8, a0
 ; CHECK-NEXT:    vrgather.vv v15, v16, v24
@@ -1842,11 +1842,11 @@ define <vscale x 8 x float> @reverse_nxv8f32(<vscale x 8 x float> %a) {
 define <vscale x 16 x float> @reverse_nxv16f32(<vscale x 16 x float> %a) {
 ; CHECK-LABEL: reverse_nxv16f32:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli a0, zero, e32, m1, ta, ma
 ; CHECK-NEXT:    vmv8r.v v16, v8
 ; CHECK-NEXT:    csrr a0, vlenb
 ; CHECK-NEXT:    srli a0, a0, 2
 ; CHECK-NEXT:    addi a0, a0, -1
-; CHECK-NEXT:    vsetvli a1, zero, e32, m1, ta, ma
 ; CHECK-NEXT:    vid.v v8
 ; CHECK-NEXT:    vrsub.vx v24, v8, a0
 ; CHECK-NEXT:    vrgather.vv v15, v16, v24
@@ -1917,11 +1917,11 @@ define <vscale x 4 x double> @reverse_nxv4f64(<vscale x 4 x double> %a) {
 define <vscale x 8 x double> @reverse_nxv8f64(<vscale x 8 x double> %a) {
 ; CHECK-LABEL: reverse_nxv8f64:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli a0, zero, e64, m1, ta, ma
 ; CHECK-NEXT:    vmv8r.v v16, v8
 ; CHECK-NEXT:    csrr a0, vlenb
 ; CHECK-NEXT:    srli a0, a0, 3
 ; CHECK-NEXT:    addi a0, a0, -1
-; CHECK-NEXT:    vsetvli a1, zero, e64, m1, ta, ma
 ; CHECK-NEXT:    vid.v v8
 ; CHECK-NEXT:    vrsub.vx v24, v8, a0
 ; CHECK-NEXT:    vrgather.vv v15, v16, v24
diff --git a/llvm/test/CodeGen/RISCV/rvv/nearbyint-vp.ll b/llvm/test/CodeGen/RISCV/rvv/nearbyint-vp.ll
index 94fce80ad3b8e..9aa26e59c6a03 100644
--- a/llvm/test/CodeGen/RISCV/rvv/nearbyint-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/nearbyint-vp.ll
@@ -117,8 +117,8 @@ declare <vscale x 4 x bfloat> @llvm.vp.nearbyint.nxv4bf16(<vscale x 4 x bfloat>,
 define <vscale x 4 x bfloat> @vp_nearbyint_nxv4bf16(<vscale x 4 x bfloat> %va, <vscale x 4 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vp_nearbyint_nxv4bf16:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v9, v0
 ; CHECK-NEXT:    vsetvli a1, zero, e16, m1, ta, ma
+; CHECK-NEXT:    vmv1r.v v9, v0
 ; CHECK-NEXT:    vfwcvtbf16.f.f.v v10, v8
 ; CHECK-NEXT:    lui a1, 307200
 ; CHECK-NEXT:    vsetvli zero, a0, e32, m2, ta, ma
@@ -169,8 +169,8 @@ declare <vscale x 8 x bfloat> @llvm.vp.nearbyint.nxv8bf16(<vscale x 8 x bfloat>,
 define <vscale x 8 x bfloat> @vp_nearbyint_nxv8bf16(<vscale x 8 x bfloat> %va, <vscale x 8 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vp_nearbyint_nxv8bf16:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vsetvli a1, zero, e16, m2, ta, ma
+; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vfwcvtbf16.f.f.v v12, v8
 ; CHECK-NEXT:    lui a1, 307200
 ; CHECK-NEXT:    vsetvli zero, a0, e32, m4, ta, ma
@@ -221,8 +221,8 @@ declare <vscale x 16 x bfloat> @llvm.vp.nearbyint.nxv16bf16(<vscale x 16 x bfloa
 define <vscale x 16 x bfloat> @vp_nearbyint_nxv16bf16(<vscale x 16 x bfloat> %va, <vscale x 16 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vp_nearbyint_nxv16bf16:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v12, v0
 ; CHECK-NEXT:    vsetvli a1, zero, e16, m4, ta, ma
+; CHECK-NEXT:    vmv1r.v v12, v0
 ; CHECK-NEXT:    vfwcvtbf16.f.f.v v16, v8
 ; CHECK-NEXT:    lui a1, 307200
 ; CHECK-NEXT:    vsetvli zero, a0, e32, m8, ta, ma
@@ -273,9 +273,9 @@ declare <vscale x 32 x bfloat> @llvm.vp.nearbyint.nxv32bf16(<vscale x 32 x bfloa
 define <vscale x 32 x bfloat> @vp_nearbyint_nxv32bf16(<vscale x 32 x bfloat> %va, <vscale x 32 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vp_nearbyint_nxv32bf16:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli a1, zero, e16, m4, ta, ma
 ; CHECK-NEXT:    vmv1r.v v7, v0
 ; CHECK-NEXT:    csrr a2, vlenb
-; CHECK-NEXT:    vsetvli a1, zero, e16, m4, ta, ma
 ; CHECK-NEXT:    vfwcvtbf16.f.f.v v24, v12
 ; CHECK-NEXT:    lui a3, 307200
 ; CHECK-NEXT:    slli a1, a2, 1
@@ -566,8 +566,8 @@ define <vscale x 4 x half> @vp_nearbyint_nxv4f16(<vscale x 4 x half> %va, <vscal
 ;
 ; ZVFHMIN-LABEL: vp_nearbyint_nxv4f16:
 ; ZVFHMIN:       # %bb.0:
-; ZVFHMIN-NEXT:    vmv1r.v v9, v0
 ; ZVFHMIN-NEXT:    vsetvli a1, zero, e16, m1, ta, ma
+; ZVFHMIN-NEXT:    vmv1r.v v9, v0
 ; ZVFHMIN-NEXT:    vfwcvt.f.f.v v10, v8
 ; ZVFHMIN-NEXT:    lui a1, 307200
 ; ZVFHMIN-NEXT:    vsetvli zero, a0, e32, m2, ta, ma
@@ -633,10 +633,10 @@ declare <vscale x 8 x half> @llvm.vp.nearbyint.nxv8f16(<vscale x 8 x half>, <vsc
 define <vscale x 8 x half> @vp_nearbyint_nxv8f16(<vscale x 8 x half> %va, <vscale x 8 x i1> %m, i32 zeroext %evl) {
 ; ZVFH-LABEL: vp_nearbyint_nxv8f16:
 ; ZVFH:       # %bb.0:
-; ZVFH-NEXT:    vmv1r.v v10, v0
-; ZVFH-NEXT:    lui a1, %hi(.LCPI18_0)
-; ZVFH-NEXT:    flh fa5, %lo(.LCPI18_0)(a1)
 ; ZVFH-NEXT:    vsetvli zero, a0, e16, m2, ta, ma
+; ZVFH-NEXT:    vmv1r.v v10, v0
+; ZVFH-NEXT:    lui a0, %hi(.LCPI18_0)
+; ZVFH-NEXT:    flh fa5, %lo(.LCPI18_0)(a0)
 ; ZVFH-NEXT:    vfabs.v v12, v8, v0.t
 ; ZVFH-NEXT:    vsetvli zero, zero, e16, m2, ta, mu
 ; ZVFH-NEXT:    vmflt.vf v10, v12, fa5, v0.t
@@ -652,8 +652,8 @@ define <vscale x 8 x half> @vp_nearbyint_nxv8f16(<vscale x 8 x half> %va, <vscal
 ;
 ; ZVFHMIN-LABEL: vp_nearbyint_nxv8f16:
 ; ZVFHMIN:       # %bb.0:
-; ZVFHMIN-NEXT:    vmv1r.v v10, v0
 ; ZVFHMIN-NEXT:    vsetvli a1, zero, e16, m2, ta, ma
+; ZVFHMIN-NEXT:    vmv1r.v v10, v0
 ; ZVFHMIN-NEXT:    vfwcvt.f.f.v v12, v8
 ; ZVFHMIN-NEXT:    lui a1, 307200
 ; ZVFHMIN-NEXT:    vsetvli zero, a0, e32, m4, ta, ma
@@ -719,10 +719,10 @@ declare <vscale x 16 x half> @llvm.vp.nearbyint.nxv16f16(<vscale x 16 x half>, <
 define <vscale x 16 x half> @vp_nearbyint_nxv16f16(<vscale x 16 x half> %va, <vscale x 16 x i1> %m, i32 zeroext %evl) {
 ; ZVFH-LABEL: vp_nearbyint_nxv16f16:
 ; ZVFH:       # %bb.0:
-; ZVFH-NEXT:    vmv1r.v v12, v0
-; ZVFH-NEXT:    lui a1, %hi(.LCPI20_0)
-; ZVFH-NEXT:    flh fa5, %lo(.LCPI20_0)(a1)
 ; ZVFH-NEXT:    vsetvli zero, a0, e16, m4, ta, ma
+; ZVFH-NEXT:    vmv1r.v v12, v0
+; ZVFH-NEXT:    lui a0, %hi(.LCPI20_0)
+; ZVFH-NEXT:    flh fa5, %lo(.LCPI20_0)(a0)
 ; ZVFH-NEXT:    vfabs.v v16, v8, v0.t
 ; ZVFH-NEXT:    vsetvli zero, zero, e16, m4, ta, mu
 ; ZVFH-NEXT:    vmflt.vf v12, v16, fa5, v0.t
@@ -738,8 +738,8 @@ define <vscale x 16 x half> @vp_nearbyint_nxv16f16(<vscale x 16 x half> %va, <vs
 ;
 ; ZVFHMIN-LABEL: vp_nearbyint_nxv16f16:
 ; ZVFHMIN:       # %bb.0:
-; ZVFHMIN-NEXT:    vmv1r.v v12, v0
 ; ZVFHMIN-NEXT:    vsetvli a1, zero, e16, m4, ta, ma
+; ZVFHMIN-NEXT:    vmv1r.v v12, v0
 ; ZVFHMIN-NEXT:    vfwcvt.f.f.v v16, v8
 ; ZVFHMIN-NEXT:    lui a1, 307200
 ; ZVFHMIN-NEXT:    vsetvli zero, a0, e32, m8, ta, ma
@@ -805,10 +805,10 @@ declare <vscale x 32 x half> @llvm.vp.nearbyint.nxv32f16(<vscale x 32 x half>, <
 define <vscale x 32 x half> @vp_nearbyint_nxv32f16(<vscale x 32 x half> %va, <vscale x 32 x i1> %m, i32 zeroext %evl) {
 ; ZVFH-LABEL: vp_nearbyint_nxv32f16:
 ; ZVFH:       # %bb.0:
-; ZVFH-NEXT:    vmv1r.v v16, v0
-; ZVFH-NEXT:    lui a1, %hi(.LCPI22_0)
-; ZVFH-NEXT:    flh fa5, %lo(.LCPI22_0)(a1)
 ; ZVFH-NEXT:    vsetvli zero, a0, e16, m8, ta, ma
+; ZVFH-NEXT:    vmv1r.v v16, v0
+; ZVFH-NEXT:    lui a0, %hi(.LCPI22_0)
+; ZVFH-NEXT:    flh fa5, %lo(.LCPI22_0)(a0)
 ; ZVFH-NEXT:    vfabs.v v24, v8, v0.t
 ; ZVFH-NEXT:    vsetvli zero, zero, e16, m8, ta, mu
 ; ZVFH-NEXT:    vmflt.vf v16, v24, fa5, v0.t
@@ -824,9 +824,9 @@ define <vscale x 32 x half> @vp_nearbyint_nxv32f16(<vscale x 32 x half> %va, <vs
 ;
 ; ZVFHMIN-LABEL: vp_nearbyint_nxv32f16:
 ; ZVFHMIN:       # %bb.0:
+; ZVFHMIN-NEXT:    vsetvli a1, zero, e16, m4, ta, ma
 ; ZVFHMIN-NEXT:    vmv1r.v v7, v0
 ; ZVFHMIN-NEXT:    csrr a2, vlenb
-; ZVFHMIN-NEXT:    vsetvli a1, zero, e16, m4, ta, ma
 ; ZVFHMIN-NEXT:    vfwcvt.f.f.v v24, v12
 ; ZVFHMIN-NEXT:    lui a3, 307200
 ; ZVFHMIN-NEXT:    slli a1, a2, 1
@@ -1036,8 +1036,8 @@ declare <vscale x 4 x float> @llvm.vp.nearbyint.nxv4f32(<vscale x 4 x float>, <v
 define <vscale x 4 x float> @vp_nearbyint_nxv4f32(<vscale x 4 x float> %va, <vscale x 4 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vp_nearbyint_nxv4f32:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e32, m2, ta, ma
+; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vfabs.v v12, v8, v0.t
 ; CHECK-NEXT:    lui a0, 307200
 ; CHECK-NEXT:    fmv.w.x fa5, a0
@@ -1080,8 +1080,8 @@ declare <vscale x 8 x float> @llvm.vp.nearbyint.nxv8f32(<vscale x 8 x float>, <v
 define <vscale x 8 x float> @vp_nearbyint_nxv8f32(<vscale x 8 x float> %va, <vscale x 8 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vp_nearbyint_nxv8f32:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v12, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e32, m4, ta, ma
+; CHECK-NEXT:    vmv1r.v v12, v0
 ; CHECK-NEXT:    vfabs.v v16, v8, v0.t
 ; CHECK-NEXT:    lui a0, 307200
 ; CHECK-NEXT:    fmv.w.x fa5, a0
@@ -1124,8 +1124,8 @@ declare <vscale x 16 x float> @llvm.vp.nearbyint.nxv16f32(<vscale x 16 x float>,
 define <vscale x 16 x float> @vp_nearbyint_nxv16f32(<vscale x 16 x float> %va, <vscale x 16 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vp_nearbyint_nxv16f32:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v16, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e32, m8, ta, ma
+; CHECK-NEXT:    vmv1r.v v16, v0
 ; CHECK-NEXT:    vfabs.v v24, v8, v0.t
 ; CHECK-NEXT:    lui a0, 307200
 ; CHECK-NEXT:    fmv.w.x fa5, a0
@@ -1210,10 +1210,10 @@ declare <vscale x 2 x double> @llvm.vp.nearbyint.nxv2f64(<vscale x 2 x double>,
 define <vscale x 2 x double> @vp_nearbyint_nxv2f64(<vscale x 2 x double> %va, <vscale x 2 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vp_nearbyint_nxv2f64:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v10, v0
-; CHECK-NEXT:    lui a1, %hi(.LCPI36_0)
-; CHECK-NEXT:    fld fa5, %lo(.LCPI36_0)(a1)
 ; CHECK-NEXT:    vsetvli zero, a0, e64, m2, ta, ma
+; CHECK-NEXT:    vmv1r.v v10, v0
+; CHECK-NEXT:    lui a0, %hi(.LCPI36_0)
+; CHECK-NEXT:    fld fa5, %lo(.LCPI36_0)(a0)
 ; CHECK-NEXT:    vfabs.v v12, v8, v0.t
 ; CHECK-NEXT:    vsetvli zero, zero, e64, m2, ta, mu
 ; CHECK-NEXT:    vmflt.vf v10, v12, fa5, v0.t
@@ -1254,10 +1254,10 @@ declare <vscale x 4 x double> @llvm.vp.nearbyint.nxv4f64(<vscale x 4 x double>,
 define <vscale x 4 x double> @vp_nearbyint_nxv4f64(<vscale x 4 x double> %va, <vscale x 4 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vp_nearbyint_nxv4f64:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v12, v0
-; CHECK-NEXT:    lui a1, %hi(.LCPI38_0)
-; CHECK-NEXT:    fld fa5, %lo(.LCPI38_0)(a1)
 ; CHECK-NEXT:    vsetvli zero, a0, e64, m4, ta, ma
+; CHECK-NEXT:    vmv1r.v v12, v0
+; CHECK-NEXT:    lui a0, %hi(.LCPI38_0)
+; CHECK-NEXT:    fld fa5, %lo(.LCPI38_0)(a0)
 ; CHECK-NEXT:    vfabs.v v16, v8, v0.t
 ; CHECK-NEXT:    vsetvli zero, zero, e64, m4, ta, mu
 ; CHECK-NEXT:    vmflt.vf v12, v16, fa5, v0.t
@@ -1298,10 +1298,10 @@ declare <vscale x 7 x double> @llvm.vp.nearbyint.nxv7f64(<vscale x 7 x double>,
 define <vscale x 7 x double> @vp_nearbyint_nxv7f64(<vscale x 7 x double> %va, <vscale x 7 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vp_nearbyint_nxv7f64:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v16, v0
-; CHECK-NEXT:    lui a1, %hi(.LCPI40_0)
-; CHECK-NEXT:    fld fa5, %lo(.LCPI40_0)(a1)
 ; CHECK-NEXT:    vsetvli zero, a0, e64, m8, ta, ma
+; CHECK-NEXT:    vmv1r.v v16, v0
+; CHECK-NEXT:    lui a0, %hi(.LCPI40_0)
+; CHECK-NEXT:    fld fa5, %lo(.LCPI40_0)(a0)
 ; CHECK-NEXT:    vfabs.v v24, v8, v0.t
 ; CHECK-NEXT:    vsetvli zero, zero, e64, m8, ta, mu
 ; CHECK-NEXT:    vmflt.vf v16, v24, fa5, v0.t
@@ -1342,10 +1342,10 @@ declare <vscale x 8 x double> @llvm.vp.nearbyint.nxv8f64(<vscale x 8 x double>,
 define <vscale x 8 x double> @vp_nearbyint_nxv8f64(<vscale x 8 x double> %va, <vscale x 8 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vp_nearbyint_nxv8f64:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v16, v0
-; CHECK-NEXT:    lui a1, %hi(.LCPI42_0)
-; CHECK-NEXT:    fld fa5, %lo(.LCPI42_0)(a1)
 ; CHECK-NEXT:    vsetvli zero, a0, e64, m8, ta, ma
+; CHECK-NEXT:    vmv1r.v v16, v0
+; CHECK-NEXT:    lui a0, %hi(.LCPI42_0)
+; CHECK-NEXT:    fld fa5, %lo(.LCPI42_0)(a0)
 ; CHECK-NEXT:    vfabs.v v24, v8, v0.t
 ; CHECK-NEXT:    vsetvli zero, zero, e64, m8, ta, mu
 ; CHECK-NEXT:    vmflt.vf v16, v24, fa5, v0.t
@@ -1387,13 +1387,13 @@ declare <vscale x 16 x double> @llvm.vp.nearbyint.nxv16f64(<vscale x 16 x double
 define <vscale x 16 x double> @vp_nearbyint_nxv16f64(<vscale x 16 x double> %va, <vscale x 16 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vp_nearbyint_nxv16f64:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli a1, zero, e8, mf4, ta, ma
 ; CHECK-NEXT:    vmv1r.v v7, v0
 ; CHECK-NEXT:    csrr a1, vlenb
 ; CHECK-NEXT:    lui a2, %hi(.LCPI44_0)
 ; CHECK-NEXT:    srli a3, a1, 3
 ; CHECK-NEXT:    fld fa5, %lo(.LCPI44_0)(a2)
 ; CHECK-NEXT:    sub a2, a0, a1
-; CHECK-NEXT:    vsetvli a4, zero, e8, mf4, ta, ma
 ; CHECK-NEXT:    vslidedown.vx v6, v0, a3
 ; CHECK-NEXT:    sltu a3, a0, a2
 ; CHECK-NEXT:    addi a3, a3, -1
diff --git a/llvm/test/CodeGen/RISCV/rvv/pr88576.ll b/llvm/test/CodeGen/RISCV/rvv/pr88576.ll
index 37c67b9ff2f6a..dd7debd3ab046 100644
--- a/llvm/test/CodeGen/RISCV/rvv/pr88576.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/pr88576.ll
@@ -23,10 +23,10 @@ define i1 @foo(<vscale x 16 x i8> %x, i64 %y) {
 ; CHECK-NEXT:    slli a2, a2, 4
 ; CHECK-NEXT:    sub sp, sp, a2
 ; CHECK-NEXT:    andi sp, sp, -64
+; CHECK-NEXT:    vsetvli a2, zero, e8, m8, ta, ma
 ; CHECK-NEXT:    vmv1r.v v0, v9
 ; CHECK-NEXT:    addi a2, sp, 64
 ; CHECK-NEXT:    slli a1, a1, 3
-; CHECK-NEXT:    vsetvli a3, zero, e8, m8, ta, ma
 ; CHECK-NEXT:    vmv.v.i v16, 0
 ; CHECK-NEXT:    add a0, a2, a0
 ; CHECK-NEXT:    add a1, a2, a1
@@ -53,8 +53,8 @@ define i1 @foo(<vscale x 16 x i8> %x, i64 %y) {
 define i8 @bar(<vscale x 128 x i1> %x, i64 %y) {
 ; CHECK-LABEL: bar:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v1, v8
 ; CHECK-NEXT:    vsetivli zero, 1, e8, m2, ta, ma
+; CHECK-NEXT:    vmv1r.v v1, v8
 ; CHECK-NEXT:    vslidedown.vx v8, v0, a0
 ; CHECK-NEXT:    vmv.x.s a0, v8
 ; CHECK-NEXT:    ret
diff --git a/llvm/test/CodeGen/RISCV/rvv/rint-vp.ll b/llvm/test/CodeGen/RISCV/rvv/rint-vp.ll
index 2a69dd31118bd..70ea1bc78d2e5 100644
--- a/llvm/test/CodeGen/RISCV/rvv/rint-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/rint-vp.ll
@@ -109,8 +109,8 @@ declare <vscale x 4 x bfloat> @llvm.vp.rint.nxv4bf16(<vscale x 4 x bfloat>, <vsc
 define <vscale x 4 x bfloat> @vp_rint_nxv4bf16(<vscale x 4 x bfloat> %va, <vscale x 4 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vp_rint_nxv4bf16:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v9, v0
 ; CHECK-NEXT:    vsetvli a1, zero, e16, m1, ta, ma
+; CHECK-NEXT:    vmv1r.v v9, v0
 ; CHECK-NEXT:    vfwcvtbf16.f.f.v v10, v8
 ; CHECK-NEXT:    lui a1, 307200
 ; CHECK-NEXT:    vsetvli zero, a0, e32, m2, ta, ma
@@ -157,8 +157,8 @@ declare <vscale x 8 x bfloat> @llvm.vp.rint.nxv8bf16(<vscale x 8 x bfloat>, <vsc
 define <vscale x 8 x bfloat> @vp_rint_nxv8bf16(<vscale x 8 x bfloat> %va, <vscale x 8 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vp_rint_nxv8bf16:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vsetvli a1, zero, e16, m2, ta, ma
+; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vfwcvtbf16.f.f.v v12, v8
 ; CHECK-NEXT:    lui a1, 307200
 ; CHECK-NEXT:    vsetvli zero, a0, e32, m4, ta, ma
@@ -205,8 +205,8 @@ declare <vscale x 16 x bfloat> @llvm.vp.rint.nxv16bf16(<vscale x 16 x bfloat>, <
 define <vscale x 16 x bfloat> @vp_rint_nxv16bf16(<vscale x 16 x bfloat> %va, <vscale x 16 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vp_rint_nxv16bf16:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v12, v0
 ; CHECK-NEXT:    vsetvli a1, zero, e16, m4, ta, ma
+; CHECK-NEXT:    vmv1r.v v12, v0
 ; CHECK-NEXT:    vfwcvtbf16.f.f.v v16, v8
 ; CHECK-NEXT:    lui a1, 307200
 ; CHECK-NEXT:    vsetvli zero, a0, e32, m8, ta, ma
@@ -259,9 +259,9 @@ define <vscale x 32 x bfloat> @vp_rint_nxv32bf16(<vscale x 32 x bfloat> %va, <vs
 ; CHECK-NEXT:    slli a1, a1, 3
 ; CHECK-NEXT:    sub sp, sp, a1
 ; CHECK-NEXT:    .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x08, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 8 * vlenb
+; CHECK-NEXT:    vsetvli a1, zero, e16, m4, ta, ma
 ; CHECK-NEXT:    vmv1r.v v7, v0
 ; CHECK-NEXT:    csrr a2, vlenb
-; CHECK-NEXT:    vsetvli a1, zero, e16, m4, ta, ma
 ; CHECK-NEXT:    vfwcvtbf16.f.f.v v24, v12
 ; CHECK-NEXT:    lui a3, 307200
 ; CHECK-NEXT:    slli a1, a2, 1
@@ -535,8 +535,8 @@ define <vscale x 4 x half> @vp_rint_nxv4f16(<vscale x 4 x half> %va, <vscale x 4
 ;
 ; ZVFHMIN-LABEL: vp_rint_nxv4f16:
 ; ZVFHMIN:       # %bb.0:
-; ZVFHMIN-NEXT:    vmv1r.v v9, v0
 ; ZVFHMIN-NEXT:    vsetvli a1, zero, e16, m1, ta, ma
+; ZVFHMIN-NEXT:    vmv1r.v v9, v0
 ; ZVFHMIN-NEXT:    vfwcvt.f.f.v v10, v8
 ; ZVFHMIN-NEXT:    lui a1, 307200
 ; ZVFHMIN-NEXT:    vsetvli zero, a0, e32, m2, ta, ma
@@ -596,10 +596,10 @@ declare <vscale x 8 x half> @llvm.vp.rint.nxv8f16(<vscale x 8 x half>, <vscale x
 define <vscale x 8 x half> @vp_rint_nxv8f16(<vscale x 8 x half> %va, <vscale x 8 x i1> %m, i32 zeroext %evl) {
 ; ZVFH-LABEL: vp_rint_nxv8f16:
 ; ZVFH:       # %bb.0:
-; ZVFH-NEXT:    vmv1r.v v10, v0
-; ZVFH-NEXT:    lui a1, %hi(.LCPI18_0)
-; ZVFH-NEXT:    flh fa5, %lo(.LCPI18_0)(a1)
 ; ZVFH-NEXT:    vsetvli zero, a0, e16, m2, ta, ma
+; ZVFH-NEXT:    vmv1r.v v10, v0
+; ZVFH-NEXT:    lui a0, %hi(.LCPI18_0)
+; ZVFH-NEXT:    flh fa5, %lo(.LCPI18_0)(a0)
 ; ZVFH-NEXT:    vfabs.v v12, v8, v0.t
 ; ZVFH-NEXT:    vsetvli zero, zero, e16, m2, ta, mu
 ; ZVFH-NEXT:    vmflt.vf v10, v12, fa5, v0.t
@@ -613,8 +613,8 @@ define <vscale x 8 x half> @vp_rint_nxv8f16(<vscale x 8 x half> %va, <vscale x 8
 ;
 ; ZVFHMIN-LABEL: vp_rint_nxv8f16:
 ; ZVFHMIN:       # %bb.0:
-; ZVFHMIN-NEXT:    vmv1r.v v10, v0
 ; ZVFHMIN-NEXT:    vsetvli a1, zero, e16, m2, ta, ma
+; ZVFHMIN-NEXT:    vmv1r.v v10, v0
 ; ZVFHMIN-NEXT:    vfwcvt.f.f.v v12, v8
 ; ZVFHMIN-NEXT:    lui a1, 307200
 ; ZVFHMIN-NEXT:    vsetvli zero, a0, e32, m4, ta, ma
@@ -674,10 +674,10 @@ declare <vscale x 16 x half> @llvm.vp.rint.nxv16f16(<vscale x 16 x half>, <vscal
 define <vscale x 16 x half> @vp_rint_nxv16f16(<vscale x 16 x half> %va, <vscale x 16 x i1> %m, i32 zeroext %evl) {
 ; ZVFH-LABEL: vp_rint_nxv16f16:
 ; ZVFH:       # %bb.0:
-; ZVFH-NEXT:    vmv1r.v v12, v0
-; ZVFH-NEXT:    lui a1, %hi(.LCPI20_0)
-; ZVFH-NEXT:    flh fa5, %lo(.LCPI20_0)(a1)
 ; ZVFH-NEXT:    vsetvli zero, a0, e16, m4, ta, ma
+; ZVFH-NEXT:    vmv1r.v v12, v0
+; ZVFH-NEXT:    lui a0, %hi(.LCPI20_0)
+; ZVFH-NEXT:    flh fa5, %lo(.LCPI20_0)(a0)
 ; ZVFH-NEXT:    vfabs.v v16, v8, v0.t
 ; ZVFH-NEXT:    vsetvli zero, zero, e16, m4, ta, mu
 ; ZVFH-NEXT:    vmflt.vf v12, v16, fa5, v0.t
@@ -691,8 +691,8 @@ define <vscale x 16 x half> @vp_rint_nxv16f16(<vscale x 16 x half> %va, <vscale
 ;
 ; ZVFHMIN-LABEL: vp_rint_nxv16f16:
 ; ZVFHMIN:       # %bb.0:
-; ZVFHMIN-NEXT:    vmv1r.v v12, v0
 ; ZVFHMIN-NEXT:    vsetvli a1, zero, e16, m4, ta, ma
+; ZVFHMIN-NEXT:    vmv1r.v v12, v0
 ; ZVFHMIN-NEXT:    vfwcvt.f.f.v v16, v8
 ; ZVFHMIN-NEXT:    lui a1, 307200
 ; ZVFHMIN-NEXT:    vsetvli zero, a0, e32, m8, ta, ma
@@ -752,10 +752,10 @@ declare <vscale x 32 x half> @llvm.vp.rint.nxv32f16(<vscale x 32 x half>, <vscal
 define <vscale x 32 x half> @vp_rint_nxv32f16(<vscale x 32 x half> %va, <vscale x 32 x i1> %m, i32 zeroext %evl) {
 ; ZVFH-LABEL: vp_rint_nxv32f16:
 ; ZVFH:       # %bb.0:
-; ZVFH-NEXT:    vmv1r.v v16, v0
-; ZVFH-NEXT:    lui a1, %hi(.LCPI22_0)
-; ZVFH-NEXT:    flh fa5, %lo(.LCPI22_0)(a1)
 ; ZVFH-NEXT:    vsetvli zero, a0, e16, m8, ta, ma
+; ZVFH-NEXT:    vmv1r.v v16, v0
+; ZVFH-NEXT:    lui a0, %hi(.LCPI22_0)
+; ZVFH-NEXT:    flh fa5, %lo(.LCPI22_0)(a0)
 ; ZVFH-NEXT:    vfabs.v v24, v8, v0.t
 ; ZVFH-NEXT:    vsetvli zero, zero, e16, m8, ta, mu
 ; ZVFH-NEXT:    vmflt.vf v16, v24, fa5, v0.t
@@ -775,9 +775,9 @@ define <vscale x 32 x half> @vp_rint_nxv32f16(<vscale x 32 x half> %va, <vscale
 ; ZVFHMIN-NEXT:    slli a1, a1, 3
 ; ZVFHMIN-NEXT:    sub sp, sp, a1
 ; ZVFHMIN-NEXT:    .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x08, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 8 * vlenb
+; ZVFHMIN-NEXT:    vsetvli a1, zero, e16, m4, ta, ma
 ; ZVFHMIN-NEXT:    vmv1r.v v7, v0
 ; ZVFHMIN-NEXT:    csrr a2, vlenb
-; ZVFHMIN-NEXT:    vsetvli a1, zero, e16, m4, ta, ma
 ; ZVFHMIN-NEXT:    vfwcvt.f.f.v v24, v12
 ; ZVFHMIN-NEXT:    lui a3, 307200
 ; ZVFHMIN-NEXT:    slli a1, a2, 1
@@ -978,8 +978,8 @@ declare <vscale x 4 x float> @llvm.vp.rint.nxv4f32(<vscale x 4 x float>, <vscale
 define <vscale x 4 x float> @vp_rint_nxv4f32(<vscale x 4 x float> %va, <vscale x 4 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vp_rint_nxv4f32:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e32, m2, ta, ma
+; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vfabs.v v12, v8, v0.t
 ; CHECK-NEXT:    lui a0, 307200
 ; CHECK-NEXT:    fmv.w.x fa5, a0
@@ -1018,8 +1018,8 @@ declare <vscale x 8 x float> @llvm.vp.rint.nxv8f32(<vscale x 8 x float>, <vscale
 define <vscale x 8 x float> @vp_rint_nxv8f32(<vscale x 8 x float> %va, <vscale x 8 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vp_rint_nxv8f32:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v12, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e32, m4, ta, ma
+; CHECK-NEXT:    vmv1r.v v12, v0
 ; CHECK-NEXT:    vfabs.v v16, v8, v0.t
 ; CHECK-NEXT:    lui a0, 307200
 ; CHECK-NEXT:    fmv.w.x fa5, a0
@@ -1058,8 +1058,8 @@ declare <vscale x 16 x float> @llvm.vp.rint.nxv16f32(<vscale x 16 x float>, <vsc
 define <vscale x 16 x float> @vp_rint_nxv16f32(<vscale x 16 x float> %va, <vscale x 16 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vp_rint_nxv16f32:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v16, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e32, m8, ta, ma
+; CHECK-NEXT:    vmv1r.v v16, v0
 ; CHECK-NEXT:    vfabs.v v24, v8, v0.t
 ; CHECK-NEXT:    lui a0, 307200
 ; CHECK-NEXT:    fmv.w.x fa5, a0
@@ -1136,10 +1136,10 @@ declare <vscale x 2 x double> @llvm.vp.rint.nxv2f64(<vscale x 2 x double>, <vsca
 define <vscale x 2 x double> @vp_rint_nxv2f64(<vscale x 2 x double> %va, <vscale x 2 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vp_rint_nxv2f64:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v10, v0
-; CHECK-NEXT:    lui a1, %hi(.LCPI36_0)
-; CHECK-NEXT:    fld fa5, %lo(.LCPI36_0)(a1)
 ; CHECK-NEXT:    vsetvli zero, a0, e64, m2, ta, ma
+; CHECK-NEXT:    vmv1r.v v10, v0
+; CHECK-NEXT:    lui a0, %hi(.LCPI36_0)
+; CHECK-NEXT:    fld fa5, %lo(.LCPI36_0)(a0)
 ; CHECK-NEXT:    vfabs.v v12, v8, v0.t
 ; CHECK-NEXT:    vsetvli zero, zero, e64, m2, ta, mu
 ; CHECK-NEXT:    vmflt.vf v10, v12, fa5, v0.t
@@ -1176,10 +1176,10 @@ declare <vscale x 4 x double> @llvm.vp.rint.nxv4f64(<vscale x 4 x double>, <vsca
 define <vscale x 4 x double> @vp_rint_nxv4f64(<vscale x 4 x double> %va, <vscale x 4 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vp_rint_nxv4f64:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v12, v0
-; CHECK-NEXT:    lui a1, %hi(.LCPI38_0)
-; CHECK-NEXT:    fld fa5, %lo(.LCPI38_0)(a1)
 ; CHECK-NEXT:    vsetvli zero, a0, e64, m4, ta, ma
+; CHECK-NEXT:    vmv1r.v v12, v0
+; CHECK-NEXT:    lui a0, %hi(.LCPI38_0)
+; CHECK-NEXT:    fld fa5, %lo(.LCPI38_0)(a0)
 ; CHECK-NEXT:    vfabs.v v16, v8, v0.t
 ; CHECK-NEXT:    vsetvli zero, zero, e64, m4, ta, mu
 ; CHECK-NEXT:    vmflt.vf v12, v16, fa5, v0.t
@@ -1216,10 +1216,10 @@ declare <vscale x 7 x double> @llvm.vp.rint.nxv7f64(<vscale x 7 x double>, <vsca
 define <vscale x 7 x double> @vp_rint_nxv7f64(<vscale x 7 x double> %va, <vscale x 7 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vp_rint_nxv7f64:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v16, v0
-; CHECK-NEXT:    lui a1, %hi(.LCPI40_0)
-; CHECK-NEXT:    fld fa5, %lo(.LCPI40_0)(a1)
 ; CHECK-NEXT:    vsetvli zero, a0, e64, m8, ta, ma
+; CHECK-NEXT:    vmv1r.v v16, v0
+; CHECK-NEXT:    lui a0, %hi(.LCPI40_0)
+; CHECK-NEXT:    fld fa5, %lo(.LCPI40_0)(a0)
 ; CHECK-NEXT:    vfabs.v v24, v8, v0.t
 ; CHECK-NEXT:    vsetvli zero, zero, e64, m8, ta, mu
 ; CHECK-NEXT:    vmflt.vf v16, v24, fa5, v0.t
@@ -1256,10 +1256,10 @@ declare <vscale x 8 x double> @llvm.vp.rint.nxv8f64(<vscale x 8 x double>, <vsca
 define <vscale x 8 x double> @vp_rint_nxv8f64(<vscale x 8 x double> %va, <vscale x 8 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vp_rint_nxv8f64:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v16, v0
-; CHECK-NEXT:    lui a1, %hi(.LCPI42_0)
-; CHECK-NEXT:    fld fa5, %lo(.LCPI42_0)(a1)
 ; CHECK-NEXT:    vsetvli zero, a0, e64, m8, ta, ma
+; CHECK-NEXT:    vmv1r.v v16, v0
+; CHECK-NEXT:    lui a0, %hi(.LCPI42_0)
+; CHECK-NEXT:    fld fa5, %lo(.LCPI42_0)(a0)
 ; CHECK-NEXT:    vfabs.v v24, v8, v0.t
 ; CHECK-NEXT:    vsetvli zero, zero, e64, m8, ta, mu
 ; CHECK-NEXT:    vmflt.vf v16, v24, fa5, v0.t
@@ -1303,13 +1303,13 @@ define <vscale x 16 x double> @vp_rint_nxv16f64(<vscale x 16 x double> %va, <vsc
 ; CHECK-NEXT:    slli a1, a1, 3
 ; CHECK-NEXT:    sub sp, sp, a1
 ; CHECK-NEXT:    .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x08, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 8 * vlenb
+; CHECK-NEXT:    vsetvli a1, zero, e8, mf4, ta, ma
 ; CHECK-NEXT:    vmv1r.v v7, v0
 ; CHECK-NEXT:    csrr a1, vlenb
 ; CHECK-NEXT:    lui a2, %hi(.LCPI44_0)
 ; CHECK-NEXT:    srli a3, a1, 3
 ; CHECK-NEXT:    fld fa5, %lo(.LCPI44_0)(a2)
 ; CHECK-NEXT:    sub a2, a0, a1
-; CHECK-NEXT:    vsetvli a4, zero, e8, mf4, ta, ma
 ; CHECK-NEXT:    vslidedown.vx v6, v0, a3
 ; CHECK-NEXT:    sltu a3, a0, a2
 ; CHECK-NEXT:    addi a3, a3, -1
diff --git a/llvm/test/CodeGen/RISCV/rvv/round-vp.ll b/llvm/test/CodeGen/RISCV/rvv/round-vp.ll
index 8a10e75333ad0..4e15f13570583 100644
--- a/llvm/test/CodeGen/RISCV/rvv/round-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/round-vp.ll
@@ -117,8 +117,8 @@ declare <vscale x 4 x bfloat> @llvm.vp.round.nxv4bf16(<vscale x 4 x bfloat>, <vs
 define <vscale x 4 x bfloat> @vp_round_nxv4bf16(<vscale x 4 x bfloat> %va, <vscale x 4 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vp_round_nxv4bf16:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v9, v0
 ; CHECK-NEXT:    vsetvli a1, zero, e16, m1, ta, ma
+; CHECK-NEXT:    vmv1r.v v9, v0
 ; CHECK-NEXT:    vfwcvtbf16.f.f.v v10, v8
 ; CHECK-NEXT:    lui a1, 307200
 ; CHECK-NEXT:    vsetvli zero, a0, e32, m2, ta, ma
@@ -169,8 +169,8 @@ declare <vscale x 8 x bfloat> @llvm.vp.round.nxv8bf16(<vscale x 8 x bfloat>, <vs
 define <vscale x 8 x bfloat> @vp_round_nxv8bf16(<vscale x 8 x bfloat> %va, <vscale x 8 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vp_round_nxv8bf16:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vsetvli a1, zero, e16, m2, ta, ma
+; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vfwcvtbf16.f.f.v v12, v8
 ; CHECK-NEXT:    lui a1, 307200
 ; CHECK-NEXT:    vsetvli zero, a0, e32, m4, ta, ma
@@ -221,8 +221,8 @@ declare <vscale x 16 x bfloat> @llvm.vp.round.nxv16bf16(<vscale x 16 x bfloat>,
 define <vscale x 16 x bfloat> @vp_round_nxv16bf16(<vscale x 16 x bfloat> %va, <vscale x 16 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vp_round_nxv16bf16:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v12, v0
 ; CHECK-NEXT:    vsetvli a1, zero, e16, m4, ta, ma
+; CHECK-NEXT:    vmv1r.v v12, v0
 ; CHECK-NEXT:    vfwcvtbf16.f.f.v v16, v8
 ; CHECK-NEXT:    lui a1, 307200
 ; CHECK-NEXT:    vsetvli zero, a0, e32, m8, ta, ma
@@ -279,9 +279,9 @@ define <vscale x 32 x bfloat> @vp_round_nxv32bf16(<vscale x 32 x bfloat> %va, <v
 ; CHECK-NEXT:    slli a1, a1, 3
 ; CHECK-NEXT:    sub sp, sp, a1
 ; CHECK-NEXT:    .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x08, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 8 * vlenb
+; CHECK-NEXT:    vsetvli a1, zero, e16, m4, ta, ma
 ; CHECK-NEXT:    vmv1r.v v7, v0
 ; CHECK-NEXT:    csrr a2, vlenb
-; CHECK-NEXT:    vsetvli a1, zero, e16, m4, ta, ma
 ; CHECK-NEXT:    vfwcvtbf16.f.f.v v24, v12
 ; CHECK-NEXT:    lui a3, 307200
 ; CHECK-NEXT:    slli a1, a2, 1
@@ -582,8 +582,8 @@ define <vscale x 4 x half> @vp_round_nxv4f16(<vscale x 4 x half> %va, <vscale x
 ;
 ; ZVFHMIN-LABEL: vp_round_nxv4f16:
 ; ZVFHMIN:       # %bb.0:
-; ZVFHMIN-NEXT:    vmv1r.v v9, v0
 ; ZVFHMIN-NEXT:    vsetvli a1, zero, e16, m1, ta, ma
+; ZVFHMIN-NEXT:    vmv1r.v v9, v0
 ; ZVFHMIN-NEXT:    vfwcvt.f.f.v v10, v8
 ; ZVFHMIN-NEXT:    lui a1, 307200
 ; ZVFHMIN-NEXT:    vsetvli zero, a0, e32, m2, ta, ma
@@ -649,10 +649,10 @@ declare <vscale x 8 x half> @llvm.vp.round.nxv8f16(<vscale x 8 x half>, <vscale
 define <vscale x 8 x half> @vp_round_nxv8f16(<vscale x 8 x half> %va, <vscale x 8 x i1> %m, i32 zeroext %evl) {
 ; ZVFH-LABEL: vp_round_nxv8f16:
 ; ZVFH:       # %bb.0:
-; ZVFH-NEXT:    vmv1r.v v10, v0
-; ZVFH-NEXT:    lui a1, %hi(.LCPI18_0)
-; ZVFH-NEXT:    flh fa5, %lo(.LCPI18_0)(a1)
 ; ZVFH-NEXT:    vsetvli zero, a0, e16, m2, ta, ma
+; ZVFH-NEXT:    vmv1r.v v10, v0
+; ZVFH-NEXT:    lui a0, %hi(.LCPI18_0)
+; ZVFH-NEXT:    flh fa5, %lo(.LCPI18_0)(a0)
 ; ZVFH-NEXT:    vfabs.v v12, v8, v0.t
 ; ZVFH-NEXT:    vsetvli zero, zero, e16, m2, ta, mu
 ; ZVFH-NEXT:    vmflt.vf v10, v12, fa5, v0.t
@@ -668,8 +668,8 @@ define <vscale x 8 x half> @vp_round_nxv8f16(<vscale x 8 x half> %va, <vscale x
 ;
 ; ZVFHMIN-LABEL: vp_round_nxv8f16:
 ; ZVFHMIN:       # %bb.0:
-; ZVFHMIN-NEXT:    vmv1r.v v10, v0
 ; ZVFHMIN-NEXT:    vsetvli a1, zero, e16, m2, ta, ma
+; ZVFHMIN-NEXT:    vmv1r.v v10, v0
 ; ZVFHMIN-NEXT:    vfwcvt.f.f.v v12, v8
 ; ZVFHMIN-NEXT:    lui a1, 307200
 ; ZVFHMIN-NEXT:    vsetvli zero, a0, e32, m4, ta, ma
@@ -735,10 +735,10 @@ declare <vscale x 16 x half> @llvm.vp.round.nxv16f16(<vscale x 16 x half>, <vsca
 define <vscale x 16 x half> @vp_round_nxv16f16(<vscale x 16 x half> %va, <vscale x 16 x i1> %m, i32 zeroext %evl) {
 ; ZVFH-LABEL: vp_round_nxv16f16:
 ; ZVFH:       # %bb.0:
-; ZVFH-NEXT:    vmv1r.v v12, v0
-; ZVFH-NEXT:    lui a1, %hi(.LCPI20_0)
-; ZVFH-NEXT:    flh fa5, %lo(.LCPI20_0)(a1)
 ; ZVFH-NEXT:    vsetvli zero, a0, e16, m4, ta, ma
+; ZVFH-NEXT:    vmv1r.v v12, v0
+; ZVFH-NEXT:    lui a0, %hi(.LCPI20_0)
+; ZVFH-NEXT:    flh fa5, %lo(.LCPI20_0)(a0)
 ; ZVFH-NEXT:    vfabs.v v16, v8, v0.t
 ; ZVFH-NEXT:    vsetvli zero, zero, e16, m4, ta, mu
 ; ZVFH-NEXT:    vmflt.vf v12, v16, fa5, v0.t
@@ -754,8 +754,8 @@ define <vscale x 16 x half> @vp_round_nxv16f16(<vscale x 16 x half> %va, <vscale
 ;
 ; ZVFHMIN-LABEL: vp_round_nxv16f16:
 ; ZVFHMIN:       # %bb.0:
-; ZVFHMIN-NEXT:    vmv1r.v v12, v0
 ; ZVFHMIN-NEXT:    vsetvli a1, zero, e16, m4, ta, ma
+; ZVFHMIN-NEXT:    vmv1r.v v12, v0
 ; ZVFHMIN-NEXT:    vfwcvt.f.f.v v16, v8
 ; ZVFHMIN-NEXT:    lui a1, 307200
 ; ZVFHMIN-NEXT:    vsetvli zero, a0, e32, m8, ta, ma
@@ -821,10 +821,10 @@ declare <vscale x 32 x half> @llvm.vp.round.nxv32f16(<vscale x 32 x half>, <vsca
 define <vscale x 32 x half> @vp_round_nxv32f16(<vscale x 32 x half> %va, <vscale x 32 x i1> %m, i32 zeroext %evl) {
 ; ZVFH-LABEL: vp_round_nxv32f16:
 ; ZVFH:       # %bb.0:
-; ZVFH-NEXT:    vmv1r.v v16, v0
-; ZVFH-NEXT:    lui a1, %hi(.LCPI22_0)
-; ZVFH-NEXT:    flh fa5, %lo(.LCPI22_0)(a1)
 ; ZVFH-NEXT:    vsetvli zero, a0, e16, m8, ta, ma
+; ZVFH-NEXT:    vmv1r.v v16, v0
+; ZVFH-NEXT:    lui a0, %hi(.LCPI22_0)
+; ZVFH-NEXT:    flh fa5, %lo(.LCPI22_0)(a0)
 ; ZVFH-NEXT:    vfabs.v v24, v8, v0.t
 ; ZVFH-NEXT:    vsetvli zero, zero, e16, m8, ta, mu
 ; ZVFH-NEXT:    vmflt.vf v16, v24, fa5, v0.t
@@ -846,9 +846,9 @@ define <vscale x 32 x half> @vp_round_nxv32f16(<vscale x 32 x half> %va, <vscale
 ; ZVFHMIN-NEXT:    slli a1, a1, 3
 ; ZVFHMIN-NEXT:    sub sp, sp, a1
 ; ZVFHMIN-NEXT:    .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x08, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 8 * vlenb
+; ZVFHMIN-NEXT:    vsetvli a1, zero, e16, m4, ta, ma
 ; ZVFHMIN-NEXT:    vmv1r.v v7, v0
 ; ZVFHMIN-NEXT:    csrr a2, vlenb
-; ZVFHMIN-NEXT:    vsetvli a1, zero, e16, m4, ta, ma
 ; ZVFHMIN-NEXT:    vfwcvt.f.f.v v24, v12
 ; ZVFHMIN-NEXT:    lui a3, 307200
 ; ZVFHMIN-NEXT:    slli a1, a2, 1
@@ -1068,8 +1068,8 @@ declare <vscale x 4 x float> @llvm.vp.round.nxv4f32(<vscale x 4 x float>, <vscal
 define <vscale x 4 x float> @vp_round_nxv4f32(<vscale x 4 x float> %va, <vscale x 4 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vp_round_nxv4f32:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e32, m2, ta, ma
+; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vfabs.v v12, v8, v0.t
 ; CHECK-NEXT:    lui a0, 307200
 ; CHECK-NEXT:    fmv.w.x fa5, a0
@@ -1112,8 +1112,8 @@ declare <vscale x 8 x float> @llvm.vp.round.nxv8f32(<vscale x 8 x float>, <vscal
 define <vscale x 8 x float> @vp_round_nxv8f32(<vscale x 8 x float> %va, <vscale x 8 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vp_round_nxv8f32:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v12, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e32, m4, ta, ma
+; CHECK-NEXT:    vmv1r.v v12, v0
 ; CHECK-NEXT:    vfabs.v v16, v8, v0.t
 ; CHECK-NEXT:    lui a0, 307200
 ; CHECK-NEXT:    fmv.w.x fa5, a0
@@ -1156,8 +1156,8 @@ declare <vscale x 16 x float> @llvm.vp.round.nxv16f32(<vscale x 16 x float>, <vs
 define <vscale x 16 x float> @vp_round_nxv16f32(<vscale x 16 x float> %va, <vscale x 16 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vp_round_nxv16f32:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v16, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e32, m8, ta, ma
+; CHECK-NEXT:    vmv1r.v v16, v0
 ; CHECK-NEXT:    vfabs.v v24, v8, v0.t
 ; CHECK-NEXT:    lui a0, 307200
 ; CHECK-NEXT:    fmv.w.x fa5, a0
@@ -1242,10 +1242,10 @@ declare <vscale x 2 x double> @llvm.vp.round.nxv2f64(<vscale x 2 x double>, <vsc
 define <vscale x 2 x double> @vp_round_nxv2f64(<vscale x 2 x double> %va, <vscale x 2 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vp_round_nxv2f64:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v10, v0
-; CHECK-NEXT:    lui a1, %hi(.LCPI36_0)
-; CHECK-NEXT:    fld fa5, %lo(.LCPI36_0)(a1)
 ; CHECK-NEXT:    vsetvli zero, a0, e64, m2, ta, ma
+; CHECK-NEXT:    vmv1r.v v10, v0
+; CHECK-NEXT:    lui a0, %hi(.LCPI36_0)
+; CHECK-NEXT:    fld fa5, %lo(.LCPI36_0)(a0)
 ; CHECK-NEXT:    vfabs.v v12, v8, v0.t
 ; CHECK-NEXT:    vsetvli zero, zero, e64, m2, ta, mu
 ; CHECK-NEXT:    vmflt.vf v10, v12, fa5, v0.t
@@ -1286,10 +1286,10 @@ declare <vscale x 4 x double> @llvm.vp.round.nxv4f64(<vscale x 4 x double>, <vsc
 define <vscale x 4 x double> @vp_round_nxv4f64(<vscale x 4 x double> %va, <vscale x 4 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vp_round_nxv4f64:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v12, v0
-; CHECK-NEXT:    lui a1, %hi(.LCPI38_0)
-; CHECK-NEXT:    fld fa5, %lo(.LCPI38_0)(a1)
 ; CHECK-NEXT:    vsetvli zero, a0, e64, m4, ta, ma
+; CHECK-NEXT:    vmv1r.v v12, v0
+; CHECK-NEXT:    lui a0, %hi(.LCPI38_0)
+; CHECK-NEXT:    fld fa5, %lo(.LCPI38_0)(a0)
 ; CHECK-NEXT:    vfabs.v v16, v8, v0.t
 ; CHECK-NEXT:    vsetvli zero, zero, e64, m4, ta, mu
 ; CHECK-NEXT:    vmflt.vf v12, v16, fa5, v0.t
@@ -1330,10 +1330,10 @@ declare <vscale x 7 x double> @llvm.vp.round.nxv7f64(<vscale x 7 x double>, <vsc
 define <vscale x 7 x double> @vp_round_nxv7f64(<vscale x 7 x double> %va, <vscale x 7 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vp_round_nxv7f64:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v16, v0
-; CHECK-NEXT:    lui a1, %hi(.LCPI40_0)
-; CHECK-NEXT:    fld fa5, %lo(.LCPI40_0)(a1)
 ; CHECK-NEXT:    vsetvli zero, a0, e64, m8, ta, ma
+; CHECK-NEXT:    vmv1r.v v16, v0
+; CHECK-NEXT:    lui a0, %hi(.LCPI40_0)
+; CHECK-NEXT:    fld fa5, %lo(.LCPI40_0)(a0)
 ; CHECK-NEXT:    vfabs.v v24, v8, v0.t
 ; CHECK-NEXT:    vsetvli zero, zero, e64, m8, ta, mu
 ; CHECK-NEXT:    vmflt.vf v16, v24, fa5, v0.t
@@ -1374,10 +1374,10 @@ declare <vscale x 8 x double> @llvm.vp.round.nxv8f64(<vscale x 8 x double>, <vsc
 define <vscale x 8 x double> @vp_round_nxv8f64(<vscale x 8 x double> %va, <vscale x 8 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vp_round_nxv8f64:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v16, v0
-; CHECK-NEXT:    lui a1, %hi(.LCPI42_0)
-; CHECK-NEXT:    fld fa5, %lo(.LCPI42_0)(a1)
 ; CHECK-NEXT:    vsetvli zero, a0, e64, m8, ta, ma
+; CHECK-NEXT:    vmv1r.v v16, v0
+; CHECK-NEXT:    lui a0, %hi(.LCPI42_0)
+; CHECK-NEXT:    fld fa5, %lo(.LCPI42_0)(a0)
 ; CHECK-NEXT:    vfabs.v v24, v8, v0.t
 ; CHECK-NEXT:    vsetvli zero, zero, e64, m8, ta, mu
 ; CHECK-NEXT:    vmflt.vf v16, v24, fa5, v0.t
@@ -1425,13 +1425,13 @@ define <vscale x 16 x double> @vp_round_nxv16f64(<vscale x 16 x double> %va, <vs
 ; CHECK-NEXT:    slli a1, a1, 3
 ; CHECK-NEXT:    sub sp, sp, a1
 ; CHECK-NEXT:    .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x08, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 8 * vlenb
+; CHECK-NEXT:    vsetvli a1, zero, e8, mf4, ta, ma
 ; CHECK-NEXT:    vmv1r.v v7, v0
 ; CHECK-NEXT:    csrr a1, vlenb
 ; CHECK-NEXT:    lui a2, %hi(.LCPI44_0)
 ; CHECK-NEXT:    srli a3, a1, 3
 ; CHECK-NEXT:    fld fa5, %lo(.LCPI44_0)(a2)
 ; CHECK-NEXT:    sub a2, a0, a1
-; CHECK-NEXT:    vsetvli a4, zero, e8, mf4, ta, ma
 ; CHECK-NEXT:    vslidedown.vx v6, v0, a3
 ; CHECK-NEXT:    sltu a3, a0, a2
 ; CHECK-NEXT:    addi a3, a3, -1
diff --git a/llvm/test/CodeGen/RISCV/rvv/roundeven-vp.ll b/llvm/test/CodeGen/RISCV/rvv/roundeven-vp.ll
index 4cd909e4b0a63..f1510a8c53181 100644
--- a/llvm/test/CodeGen/RISCV/rvv/roundeven-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/roundeven-vp.ll
@@ -117,8 +117,8 @@ declare <vscale x 4 x bfloat> @llvm.vp.roundeven.nxv4bf16(<vscale x 4 x bfloat>,
 define <vscale x 4 x bfloat> @vp_roundeven_nxv4bf16(<vscale x 4 x bfloat> %va, <vscale x 4 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vp_roundeven_nxv4bf16:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v9, v0
 ; CHECK-NEXT:    vsetvli a1, zero, e16, m1, ta, ma
+; CHECK-NEXT:    vmv1r.v v9, v0
 ; CHECK-NEXT:    vfwcvtbf16.f.f.v v10, v8
 ; CHECK-NEXT:    lui a1, 307200
 ; CHECK-NEXT:    vsetvli zero, a0, e32, m2, ta, ma
@@ -169,8 +169,8 @@ declare <vscale x 8 x bfloat> @llvm.vp.roundeven.nxv8bf16(<vscale x 8 x bfloat>,
 define <vscale x 8 x bfloat> @vp_roundeven_nxv8bf16(<vscale x 8 x bfloat> %va, <vscale x 8 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vp_roundeven_nxv8bf16:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vsetvli a1, zero, e16, m2, ta, ma
+; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vfwcvtbf16.f.f.v v12, v8
 ; CHECK-NEXT:    lui a1, 307200
 ; CHECK-NEXT:    vsetvli zero, a0, e32, m4, ta, ma
@@ -221,8 +221,8 @@ declare <vscale x 16 x bfloat> @llvm.vp.roundeven.nxv16bf16(<vscale x 16 x bfloa
 define <vscale x 16 x bfloat> @vp_roundeven_nxv16bf16(<vscale x 16 x bfloat> %va, <vscale x 16 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vp_roundeven_nxv16bf16:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v12, v0
 ; CHECK-NEXT:    vsetvli a1, zero, e16, m4, ta, ma
+; CHECK-NEXT:    vmv1r.v v12, v0
 ; CHECK-NEXT:    vfwcvtbf16.f.f.v v16, v8
 ; CHECK-NEXT:    lui a1, 307200
 ; CHECK-NEXT:    vsetvli zero, a0, e32, m8, ta, ma
@@ -279,9 +279,9 @@ define <vscale x 32 x bfloat> @vp_roundeven_nxv32bf16(<vscale x 32 x bfloat> %va
 ; CHECK-NEXT:    slli a1, a1, 3
 ; CHECK-NEXT:    sub sp, sp, a1
 ; CHECK-NEXT:    .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x08, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 8 * vlenb
+; CHECK-NEXT:    vsetvli a1, zero, e16, m4, ta, ma
 ; CHECK-NEXT:    vmv1r.v v7, v0
 ; CHECK-NEXT:    csrr a2, vlenb
-; CHECK-NEXT:    vsetvli a1, zero, e16, m4, ta, ma
 ; CHECK-NEXT:    vfwcvtbf16.f.f.v v24, v12
 ; CHECK-NEXT:    lui a3, 307200
 ; CHECK-NEXT:    slli a1, a2, 1
@@ -582,8 +582,8 @@ define <vscale x 4 x half> @vp_roundeven_nxv4f16(<vscale x 4 x half> %va, <vscal
 ;
 ; ZVFHMIN-LABEL: vp_roundeven_nxv4f16:
 ; ZVFHMIN:       # %bb.0:
-; ZVFHMIN-NEXT:    vmv1r.v v9, v0
 ; ZVFHMIN-NEXT:    vsetvli a1, zero, e16, m1, ta, ma
+; ZVFHMIN-NEXT:    vmv1r.v v9, v0
 ; ZVFHMIN-NEXT:    vfwcvt.f.f.v v10, v8
 ; ZVFHMIN-NEXT:    lui a1, 307200
 ; ZVFHMIN-NEXT:    vsetvli zero, a0, e32, m2, ta, ma
@@ -649,10 +649,10 @@ declare <vscale x 8 x half> @llvm.vp.roundeven.nxv8f16(<vscale x 8 x half>, <vsc
 define <vscale x 8 x half> @vp_roundeven_nxv8f16(<vscale x 8 x half> %va, <vscale x 8 x i1> %m, i32 zeroext %evl) {
 ; ZVFH-LABEL: vp_roundeven_nxv8f16:
 ; ZVFH:       # %bb.0:
-; ZVFH-NEXT:    vmv1r.v v10, v0
-; ZVFH-NEXT:    lui a1, %hi(.LCPI18_0)
-; ZVFH-NEXT:    flh fa5, %lo(.LCPI18_0)(a1)
 ; ZVFH-NEXT:    vsetvli zero, a0, e16, m2, ta, ma
+; ZVFH-NEXT:    vmv1r.v v10, v0
+; ZVFH-NEXT:    lui a0, %hi(.LCPI18_0)
+; ZVFH-NEXT:    flh fa5, %lo(.LCPI18_0)(a0)
 ; ZVFH-NEXT:    vfabs.v v12, v8, v0.t
 ; ZVFH-NEXT:    vsetvli zero, zero, e16, m2, ta, mu
 ; ZVFH-NEXT:    vmflt.vf v10, v12, fa5, v0.t
@@ -668,8 +668,8 @@ define <vscale x 8 x half> @vp_roundeven_nxv8f16(<vscale x 8 x half> %va, <vscal
 ;
 ; ZVFHMIN-LABEL: vp_roundeven_nxv8f16:
 ; ZVFHMIN:       # %bb.0:
-; ZVFHMIN-NEXT:    vmv1r.v v10, v0
 ; ZVFHMIN-NEXT:    vsetvli a1, zero, e16, m2, ta, ma
+; ZVFHMIN-NEXT:    vmv1r.v v10, v0
 ; ZVFHMIN-NEXT:    vfwcvt.f.f.v v12, v8
 ; ZVFHMIN-NEXT:    lui a1, 307200
 ; ZVFHMIN-NEXT:    vsetvli zero, a0, e32, m4, ta, ma
@@ -735,10 +735,10 @@ declare <vscale x 16 x half> @llvm.vp.roundeven.nxv16f16(<vscale x 16 x half>, <
 define <vscale x 16 x half> @vp_roundeven_nxv16f16(<vscale x 16 x half> %va, <vscale x 16 x i1> %m, i32 zeroext %evl) {
 ; ZVFH-LABEL: vp_roundeven_nxv16f16:
 ; ZVFH:       # %bb.0:
-; ZVFH-NEXT:    vmv1r.v v12, v0
-; ZVFH-NEXT:    lui a1, %hi(.LCPI20_0)
-; ZVFH-NEXT:    flh fa5, %lo(.LCPI20_0)(a1)
 ; ZVFH-NEXT:    vsetvli zero, a0, e16, m4, ta, ma
+; ZVFH-NEXT:    vmv1r.v v12, v0
+; ZVFH-NEXT:    lui a0, %hi(.LCPI20_0)
+; ZVFH-NEXT:    flh fa5, %lo(.LCPI20_0)(a0)
 ; ZVFH-NEXT:    vfabs.v v16, v8, v0.t
 ; ZVFH-NEXT:    vsetvli zero, zero, e16, m4, ta, mu
 ; ZVFH-NEXT:    vmflt.vf v12, v16, fa5, v0.t
@@ -754,8 +754,8 @@ define <vscale x 16 x half> @vp_roundeven_nxv16f16(<vscale x 16 x half> %va, <vs
 ;
 ; ZVFHMIN-LABEL: vp_roundeven_nxv16f16:
 ; ZVFHMIN:       # %bb.0:
-; ZVFHMIN-NEXT:    vmv1r.v v12, v0
 ; ZVFHMIN-NEXT:    vsetvli a1, zero, e16, m4, ta, ma
+; ZVFHMIN-NEXT:    vmv1r.v v12, v0
 ; ZVFHMIN-NEXT:    vfwcvt.f.f.v v16, v8
 ; ZVFHMIN-NEXT:    lui a1, 307200
 ; ZVFHMIN-NEXT:    vsetvli zero, a0, e32, m8, ta, ma
@@ -821,10 +821,10 @@ declare <vscale x 32 x half> @llvm.vp.roundeven.nxv32f16(<vscale x 32 x half>, <
 define <vscale x 32 x half> @vp_roundeven_nxv32f16(<vscale x 32 x half> %va, <vscale x 32 x i1> %m, i32 zeroext %evl) {
 ; ZVFH-LABEL: vp_roundeven_nxv32f16:
 ; ZVFH:       # %bb.0:
-; ZVFH-NEXT:    vmv1r.v v16, v0
-; ZVFH-NEXT:    lui a1, %hi(.LCPI22_0)
-; ZVFH-NEXT:    flh fa5, %lo(.LCPI22_0)(a1)
 ; ZVFH-NEXT:    vsetvli zero, a0, e16, m8, ta, ma
+; ZVFH-NEXT:    vmv1r.v v16, v0
+; ZVFH-NEXT:    lui a0, %hi(.LCPI22_0)
+; ZVFH-NEXT:    flh fa5, %lo(.LCPI22_0)(a0)
 ; ZVFH-NEXT:    vfabs.v v24, v8, v0.t
 ; ZVFH-NEXT:    vsetvli zero, zero, e16, m8, ta, mu
 ; ZVFH-NEXT:    vmflt.vf v16, v24, fa5, v0.t
@@ -846,9 +846,9 @@ define <vscale x 32 x half> @vp_roundeven_nxv32f16(<vscale x 32 x half> %va, <vs
 ; ZVFHMIN-NEXT:    slli a1, a1, 3
 ; ZVFHMIN-NEXT:    sub sp, sp, a1
 ; ZVFHMIN-NEXT:    .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x08, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 8 * vlenb
+; ZVFHMIN-NEXT:    vsetvli a1, zero, e16, m4, ta, ma
 ; ZVFHMIN-NEXT:    vmv1r.v v7, v0
 ; ZVFHMIN-NEXT:    csrr a2, vlenb
-; ZVFHMIN-NEXT:    vsetvli a1, zero, e16, m4, ta, ma
 ; ZVFHMIN-NEXT:    vfwcvt.f.f.v v24, v12
 ; ZVFHMIN-NEXT:    lui a3, 307200
 ; ZVFHMIN-NEXT:    slli a1, a2, 1
@@ -1068,8 +1068,8 @@ declare <vscale x 4 x float> @llvm.vp.roundeven.nxv4f32(<vscale x 4 x float>, <v
 define <vscale x 4 x float> @vp_roundeven_nxv4f32(<vscale x 4 x float> %va, <vscale x 4 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vp_roundeven_nxv4f32:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e32, m2, ta, ma
+; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vfabs.v v12, v8, v0.t
 ; CHECK-NEXT:    lui a0, 307200
 ; CHECK-NEXT:    fmv.w.x fa5, a0
@@ -1112,8 +1112,8 @@ declare <vscale x 8 x float> @llvm.vp.roundeven.nxv8f32(<vscale x 8 x float>, <v
 define <vscale x 8 x float> @vp_roundeven_nxv8f32(<vscale x 8 x float> %va, <vscale x 8 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vp_roundeven_nxv8f32:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v12, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e32, m4, ta, ma
+; CHECK-NEXT:    vmv1r.v v12, v0
 ; CHECK-NEXT:    vfabs.v v16, v8, v0.t
 ; CHECK-NEXT:    lui a0, 307200
 ; CHECK-NEXT:    fmv.w.x fa5, a0
@@ -1156,8 +1156,8 @@ declare <vscale x 16 x float> @llvm.vp.roundeven.nxv16f32(<vscale x 16 x float>,
 define <vscale x 16 x float> @vp_roundeven_nxv16f32(<vscale x 16 x float> %va, <vscale x 16 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vp_roundeven_nxv16f32:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v16, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e32, m8, ta, ma
+; CHECK-NEXT:    vmv1r.v v16, v0
 ; CHECK-NEXT:    vfabs.v v24, v8, v0.t
 ; CHECK-NEXT:    lui a0, 307200
 ; CHECK-NEXT:    fmv.w.x fa5, a0
@@ -1242,10 +1242,10 @@ declare <vscale x 2 x double> @llvm.vp.roundeven.nxv2f64(<vscale x 2 x double>,
 define <vscale x 2 x double> @vp_roundeven_nxv2f64(<vscale x 2 x double> %va, <vscale x 2 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vp_roundeven_nxv2f64:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v10, v0
-; CHECK-NEXT:    lui a1, %hi(.LCPI36_0)
-; CHECK-NEXT:    fld fa5, %lo(.LCPI36_0)(a1)
 ; CHECK-NEXT:    vsetvli zero, a0, e64, m2, ta, ma
+; CHECK-NEXT:    vmv1r.v v10, v0
+; CHECK-NEXT:    lui a0, %hi(.LCPI36_0)
+; CHECK-NEXT:    fld fa5, %lo(.LCPI36_0)(a0)
 ; CHECK-NEXT:    vfabs.v v12, v8, v0.t
 ; CHECK-NEXT:    vsetvli zero, zero, e64, m2, ta, mu
 ; CHECK-NEXT:    vmflt.vf v10, v12, fa5, v0.t
@@ -1286,10 +1286,10 @@ declare <vscale x 4 x double> @llvm.vp.roundeven.nxv4f64(<vscale x 4 x double>,
 define <vscale x 4 x double> @vp_roundeven_nxv4f64(<vscale x 4 x double> %va, <vscale x 4 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vp_roundeven_nxv4f64:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v12, v0
-; CHECK-NEXT:    lui a1, %hi(.LCPI38_0)
-; CHECK-NEXT:    fld fa5, %lo(.LCPI38_0)(a1)
 ; CHECK-NEXT:    vsetvli zero, a0, e64, m4, ta, ma
+; CHECK-NEXT:    vmv1r.v v12, v0
+; CHECK-NEXT:    lui a0, %hi(.LCPI38_0)
+; CHECK-NEXT:    fld fa5, %lo(.LCPI38_0)(a0)
 ; CHECK-NEXT:    vfabs.v v16, v8, v0.t
 ; CHECK-NEXT:    vsetvli zero, zero, e64, m4, ta, mu
 ; CHECK-NEXT:    vmflt.vf v12, v16, fa5, v0.t
@@ -1330,10 +1330,10 @@ declare <vscale x 7 x double> @llvm.vp.roundeven.nxv7f64(<vscale x 7 x double>,
 define <vscale x 7 x double> @vp_roundeven_nxv7f64(<vscale x 7 x double> %va, <vscale x 7 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vp_roundeven_nxv7f64:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v16, v0
-; CHECK-NEXT:    lui a1, %hi(.LCPI40_0)
-; CHECK-NEXT:    fld fa5, %lo(.LCPI40_0)(a1)
 ; CHECK-NEXT:    vsetvli zero, a0, e64, m8, ta, ma
+; CHECK-NEXT:    vmv1r.v v16, v0
+; CHECK-NEXT:    lui a0, %hi(.LCPI40_0)
+; CHECK-NEXT:    fld fa5, %lo(.LCPI40_0)(a0)
 ; CHECK-NEXT:    vfabs.v v24, v8, v0.t
 ; CHECK-NEXT:    vsetvli zero, zero, e64, m8, ta, mu
 ; CHECK-NEXT:    vmflt.vf v16, v24, fa5, v0.t
@@ -1374,10 +1374,10 @@ declare <vscale x 8 x double> @llvm.vp.roundeven.nxv8f64(<vscale x 8 x double>,
 define <vscale x 8 x double> @vp_roundeven_nxv8f64(<vscale x 8 x double> %va, <vscale x 8 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vp_roundeven_nxv8f64:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v16, v0
-; CHECK-NEXT:    lui a1, %hi(.LCPI42_0)
-; CHECK-NEXT:    fld fa5, %lo(.LCPI42_0)(a1)
 ; CHECK-NEXT:    vsetvli zero, a0, e64, m8, ta, ma
+; CHECK-NEXT:    vmv1r.v v16, v0
+; CHECK-NEXT:    lui a0, %hi(.LCPI42_0)
+; CHECK-NEXT:    fld fa5, %lo(.LCPI42_0)(a0)
 ; CHECK-NEXT:    vfabs.v v24, v8, v0.t
 ; CHECK-NEXT:    vsetvli zero, zero, e64, m8, ta, mu
 ; CHECK-NEXT:    vmflt.vf v16, v24, fa5, v0.t
@@ -1425,13 +1425,13 @@ define <vscale x 16 x double> @vp_roundeven_nxv16f64(<vscale x 16 x double> %va,
 ; CHECK-NEXT:    slli a1, a1, 3
 ; CHECK-NEXT:    sub sp, sp, a1
 ; CHECK-NEXT:    .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x08, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 8 * vlenb
+; CHECK-NEXT:    vsetvli a1, zero, e8, mf4, ta, ma
 ; CHECK-NEXT:    vmv1r.v v7, v0
 ; CHECK-NEXT:    csrr a1, vlenb
 ; CHECK-NEXT:    lui a2, %hi(.LCPI44_0)
 ; CHECK-NEXT:    srli a3, a1, 3
 ; CHECK-NEXT:    fld fa5, %lo(.LCPI44_0)(a2)
 ; CHECK-NEXT:    sub a2, a0, a1
-; CHECK-NEXT:    vsetvli a4, zero, e8, mf4, ta, ma
 ; CHECK-NEXT:    vslidedown.vx v6, v0, a3
 ; CHECK-NEXT:    sltu a3, a0, a2
 ; CHECK-NEXT:    addi a3, a3, -1
diff --git a/llvm/test/CodeGen/RISCV/rvv/roundtozero-vp.ll b/llvm/test/CodeGen/RISCV/rvv/roundtozero-vp.ll
index 96c821a76ae84..dd7db58ccdf34 100644
--- a/llvm/test/CodeGen/RISCV/rvv/roundtozero-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/roundtozero-vp.ll
@@ -117,8 +117,8 @@ declare <vscale x 4 x bfloat> @llvm.vp.roundtozero.nxv4bf16(<vscale x 4 x bfloat
 define <vscale x 4 x bfloat> @vp_roundtozero_nxv4bf16(<vscale x 4 x bfloat> %va, <vscale x 4 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vp_roundtozero_nxv4bf16:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v9, v0
 ; CHECK-NEXT:    vsetvli a1, zero, e16, m1, ta, ma
+; CHECK-NEXT:    vmv1r.v v9, v0
 ; CHECK-NEXT:    vfwcvtbf16.f.f.v v10, v8
 ; CHECK-NEXT:    lui a1, 307200
 ; CHECK-NEXT:    vsetvli zero, a0, e32, m2, ta, ma
@@ -169,8 +169,8 @@ declare <vscale x 8 x bfloat> @llvm.vp.roundtozero.nxv8bf16(<vscale x 8 x bfloat
 define <vscale x 8 x bfloat> @vp_roundtozero_nxv8bf16(<vscale x 8 x bfloat> %va, <vscale x 8 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vp_roundtozero_nxv8bf16:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vsetvli a1, zero, e16, m2, ta, ma
+; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vfwcvtbf16.f.f.v v12, v8
 ; CHECK-NEXT:    lui a1, 307200
 ; CHECK-NEXT:    vsetvli zero, a0, e32, m4, ta, ma
@@ -221,8 +221,8 @@ declare <vscale x 16 x bfloat> @llvm.vp.roundtozero.nxv16bf16(<vscale x 16 x bfl
 define <vscale x 16 x bfloat> @vp_roundtozero_nxv16bf16(<vscale x 16 x bfloat> %va, <vscale x 16 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vp_roundtozero_nxv16bf16:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v12, v0
 ; CHECK-NEXT:    vsetvli a1, zero, e16, m4, ta, ma
+; CHECK-NEXT:    vmv1r.v v12, v0
 ; CHECK-NEXT:    vfwcvtbf16.f.f.v v16, v8
 ; CHECK-NEXT:    lui a1, 307200
 ; CHECK-NEXT:    vsetvli zero, a0, e32, m8, ta, ma
@@ -279,9 +279,9 @@ define <vscale x 32 x bfloat> @vp_roundtozero_nxv32bf16(<vscale x 32 x bfloat> %
 ; CHECK-NEXT:    slli a1, a1, 3
 ; CHECK-NEXT:    sub sp, sp, a1
 ; CHECK-NEXT:    .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x08, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 8 * vlenb
+; CHECK-NEXT:    vsetvli a1, zero, e16, m4, ta, ma
 ; CHECK-NEXT:    vmv1r.v v7, v0
 ; CHECK-NEXT:    csrr a2, vlenb
-; CHECK-NEXT:    vsetvli a1, zero, e16, m4, ta, ma
 ; CHECK-NEXT:    vfwcvtbf16.f.f.v v24, v12
 ; CHECK-NEXT:    lui a3, 307200
 ; CHECK-NEXT:    slli a1, a2, 1
@@ -582,8 +582,8 @@ define <vscale x 4 x half> @vp_roundtozero_nxv4f16(<vscale x 4 x half> %va, <vsc
 ;
 ; ZVFHMIN-LABEL: vp_roundtozero_nxv4f16:
 ; ZVFHMIN:       # %bb.0:
-; ZVFHMIN-NEXT:    vmv1r.v v9, v0
 ; ZVFHMIN-NEXT:    vsetvli a1, zero, e16, m1, ta, ma
+; ZVFHMIN-NEXT:    vmv1r.v v9, v0
 ; ZVFHMIN-NEXT:    vfwcvt.f.f.v v10, v8
 ; ZVFHMIN-NEXT:    lui a1, 307200
 ; ZVFHMIN-NEXT:    vsetvli zero, a0, e32, m2, ta, ma
@@ -649,10 +649,10 @@ declare <vscale x 8 x half> @llvm.vp.roundtozero.nxv8f16(<vscale x 8 x half>, <v
 define <vscale x 8 x half> @vp_roundtozero_nxv8f16(<vscale x 8 x half> %va, <vscale x 8 x i1> %m, i32 zeroext %evl) {
 ; ZVFH-LABEL: vp_roundtozero_nxv8f16:
 ; ZVFH:       # %bb.0:
-; ZVFH-NEXT:    vmv1r.v v10, v0
-; ZVFH-NEXT:    lui a1, %hi(.LCPI18_0)
-; ZVFH-NEXT:    flh fa5, %lo(.LCPI18_0)(a1)
 ; ZVFH-NEXT:    vsetvli zero, a0, e16, m2, ta, ma
+; ZVFH-NEXT:    vmv1r.v v10, v0
+; ZVFH-NEXT:    lui a0, %hi(.LCPI18_0)
+; ZVFH-NEXT:    flh fa5, %lo(.LCPI18_0)(a0)
 ; ZVFH-NEXT:    vfabs.v v12, v8, v0.t
 ; ZVFH-NEXT:    vsetvli zero, zero, e16, m2, ta, mu
 ; ZVFH-NEXT:    vmflt.vf v10, v12, fa5, v0.t
@@ -668,8 +668,8 @@ define <vscale x 8 x half> @vp_roundtozero_nxv8f16(<vscale x 8 x half> %va, <vsc
 ;
 ; ZVFHMIN-LABEL: vp_roundtozero_nxv8f16:
 ; ZVFHMIN:       # %bb.0:
-; ZVFHMIN-NEXT:    vmv1r.v v10, v0
 ; ZVFHMIN-NEXT:    vsetvli a1, zero, e16, m2, ta, ma
+; ZVFHMIN-NEXT:    vmv1r.v v10, v0
 ; ZVFHMIN-NEXT:    vfwcvt.f.f.v v12, v8
 ; ZVFHMIN-NEXT:    lui a1, 307200
 ; ZVFHMIN-NEXT:    vsetvli zero, a0, e32, m4, ta, ma
@@ -735,10 +735,10 @@ declare <vscale x 16 x half> @llvm.vp.roundtozero.nxv16f16(<vscale x 16 x half>,
 define <vscale x 16 x half> @vp_roundtozero_nxv16f16(<vscale x 16 x half> %va, <vscale x 16 x i1> %m, i32 zeroext %evl) {
 ; ZVFH-LABEL: vp_roundtozero_nxv16f16:
 ; ZVFH:       # %bb.0:
-; ZVFH-NEXT:    vmv1r.v v12, v0
-; ZVFH-NEXT:    lui a1, %hi(.LCPI20_0)
-; ZVFH-NEXT:    flh fa5, %lo(.LCPI20_0)(a1)
 ; ZVFH-NEXT:    vsetvli zero, a0, e16, m4, ta, ma
+; ZVFH-NEXT:    vmv1r.v v12, v0
+; ZVFH-NEXT:    lui a0, %hi(.LCPI20_0)
+; ZVFH-NEXT:    flh fa5, %lo(.LCPI20_0)(a0)
 ; ZVFH-NEXT:    vfabs.v v16, v8, v0.t
 ; ZVFH-NEXT:    vsetvli zero, zero, e16, m4, ta, mu
 ; ZVFH-NEXT:    vmflt.vf v12, v16, fa5, v0.t
@@ -754,8 +754,8 @@ define <vscale x 16 x half> @vp_roundtozero_nxv16f16(<vscale x 16 x half> %va, <
 ;
 ; ZVFHMIN-LABEL: vp_roundtozero_nxv16f16:
 ; ZVFHMIN:       # %bb.0:
-; ZVFHMIN-NEXT:    vmv1r.v v12, v0
 ; ZVFHMIN-NEXT:    vsetvli a1, zero, e16, m4, ta, ma
+; ZVFHMIN-NEXT:    vmv1r.v v12, v0
 ; ZVFHMIN-NEXT:    vfwcvt.f.f.v v16, v8
 ; ZVFHMIN-NEXT:    lui a1, 307200
 ; ZVFHMIN-NEXT:    vsetvli zero, a0, e32, m8, ta, ma
@@ -821,10 +821,10 @@ declare <vscale x 32 x half> @llvm.vp.roundtozero.nxv32f16(<vscale x 32 x half>,
 define <vscale x 32 x half> @vp_roundtozero_nxv32f16(<vscale x 32 x half> %va, <vscale x 32 x i1> %m, i32 zeroext %evl) {
 ; ZVFH-LABEL: vp_roundtozero_nxv32f16:
 ; ZVFH:       # %bb.0:
-; ZVFH-NEXT:    vmv1r.v v16, v0
-; ZVFH-NEXT:    lui a1, %hi(.LCPI22_0)
-; ZVFH-NEXT:    flh fa5, %lo(.LCPI22_0)(a1)
 ; ZVFH-NEXT:    vsetvli zero, a0, e16, m8, ta, ma
+; ZVFH-NEXT:    vmv1r.v v16, v0
+; ZVFH-NEXT:    lui a0, %hi(.LCPI22_0)
+; ZVFH-NEXT:    flh fa5, %lo(.LCPI22_0)(a0)
 ; ZVFH-NEXT:    vfabs.v v24, v8, v0.t
 ; ZVFH-NEXT:    vsetvli zero, zero, e16, m8, ta, mu
 ; ZVFH-NEXT:    vmflt.vf v16, v24, fa5, v0.t
@@ -846,9 +846,9 @@ define <vscale x 32 x half> @vp_roundtozero_nxv32f16(<vscale x 32 x half> %va, <
 ; ZVFHMIN-NEXT:    slli a1, a1, 3
 ; ZVFHMIN-NEXT:    sub sp, sp, a1
 ; ZVFHMIN-NEXT:    .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x08, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 8 * vlenb
+; ZVFHMIN-NEXT:    vsetvli a1, zero, e16, m4, ta, ma
 ; ZVFHMIN-NEXT:    vmv1r.v v7, v0
 ; ZVFHMIN-NEXT:    csrr a2, vlenb
-; ZVFHMIN-NEXT:    vsetvli a1, zero, e16, m4, ta, ma
 ; ZVFHMIN-NEXT:    vfwcvt.f.f.v v24, v12
 ; ZVFHMIN-NEXT:    lui a3, 307200
 ; ZVFHMIN-NEXT:    slli a1, a2, 1
@@ -1068,8 +1068,8 @@ declare <vscale x 4 x float> @llvm.vp.roundtozero.nxv4f32(<vscale x 4 x float>,
 define <vscale x 4 x float> @vp_roundtozero_nxv4f32(<vscale x 4 x float> %va, <vscale x 4 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vp_roundtozero_nxv4f32:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e32, m2, ta, ma
+; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vfabs.v v12, v8, v0.t
 ; CHECK-NEXT:    lui a0, 307200
 ; CHECK-NEXT:    fmv.w.x fa5, a0
@@ -1112,8 +1112,8 @@ declare <vscale x 8 x float> @llvm.vp.roundtozero.nxv8f32(<vscale x 8 x float>,
 define <vscale x 8 x float> @vp_roundtozero_nxv8f32(<vscale x 8 x float> %va, <vscale x 8 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vp_roundtozero_nxv8f32:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v12, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e32, m4, ta, ma
+; CHECK-NEXT:    vmv1r.v v12, v0
 ; CHECK-NEXT:    vfabs.v v16, v8, v0.t
 ; CHECK-NEXT:    lui a0, 307200
 ; CHECK-NEXT:    fmv.w.x fa5, a0
@@ -1156,8 +1156,8 @@ declare <vscale x 16 x float> @llvm.vp.roundtozero.nxv16f32(<vscale x 16 x float
 define <vscale x 16 x float> @vp_roundtozero_nxv16f32(<vscale x 16 x float> %va, <vscale x 16 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vp_roundtozero_nxv16f32:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v16, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e32, m8, ta, ma
+; CHECK-NEXT:    vmv1r.v v16, v0
 ; CHECK-NEXT:    vfabs.v v24, v8, v0.t
 ; CHECK-NEXT:    lui a0, 307200
 ; CHECK-NEXT:    fmv.w.x fa5, a0
@@ -1242,10 +1242,10 @@ declare <vscale x 2 x double> @llvm.vp.roundtozero.nxv2f64(<vscale x 2 x double>
 define <vscale x 2 x double> @vp_roundtozero_nxv2f64(<vscale x 2 x double> %va, <vscale x 2 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vp_roundtozero_nxv2f64:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v10, v0
-; CHECK-NEXT:    lui a1, %hi(.LCPI36_0)
-; CHECK-NEXT:    fld fa5, %lo(.LCPI36_0)(a1)
 ; CHECK-NEXT:    vsetvli zero, a0, e64, m2, ta, ma
+; CHECK-NEXT:    vmv1r.v v10, v0
+; CHECK-NEXT:    lui a0, %hi(.LCPI36_0)
+; CHECK-NEXT:    fld fa5, %lo(.LCPI36_0)(a0)
 ; CHECK-NEXT:    vfabs.v v12, v8, v0.t
 ; CHECK-NEXT:    vsetvli zero, zero, e64, m2, ta, mu
 ; CHECK-NEXT:    vmflt.vf v10, v12, fa5, v0.t
@@ -1286,10 +1286,10 @@ declare <vscale x 4 x double> @llvm.vp.roundtozero.nxv4f64(<vscale x 4 x double>
 define <vscale x 4 x double> @vp_roundtozero_nxv4f64(<vscale x 4 x double> %va, <vscale x 4 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vp_roundtozero_nxv4f64:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v12, v0
-; CHECK-NEXT:    lui a1, %hi(.LCPI38_0)
-; CHECK-NEXT:    fld fa5, %lo(.LCPI38_0)(a1)
 ; CHECK-NEXT:    vsetvli zero, a0, e64, m4, ta, ma
+; CHECK-NEXT:    vmv1r.v v12, v0
+; CHECK-NEXT:    lui a0, %hi(.LCPI38_0)
+; CHECK-NEXT:    fld fa5, %lo(.LCPI38_0)(a0)
 ; CHECK-NEXT:    vfabs.v v16, v8, v0.t
 ; CHECK-NEXT:    vsetvli zero, zero, e64, m4, ta, mu
 ; CHECK-NEXT:    vmflt.vf v12, v16, fa5, v0.t
@@ -1330,10 +1330,10 @@ declare <vscale x 7 x double> @llvm.vp.roundtozero.nxv7f64(<vscale x 7 x double>
 define <vscale x 7 x double> @vp_roundtozero_nxv7f64(<vscale x 7 x double> %va, <vscale x 7 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vp_roundtozero_nxv7f64:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v16, v0
-; CHECK-NEXT:    lui a1, %hi(.LCPI40_0)
-; CHECK-NEXT:    fld fa5, %lo(.LCPI40_0)(a1)
 ; CHECK-NEXT:    vsetvli zero, a0, e64, m8, ta, ma
+; CHECK-NEXT:    vmv1r.v v16, v0
+; CHECK-NEXT:    lui a0, %hi(.LCPI40_0)
+; CHECK-NEXT:    fld fa5, %lo(.LCPI40_0)(a0)
 ; CHECK-NEXT:    vfabs.v v24, v8, v0.t
 ; CHECK-NEXT:    vsetvli zero, zero, e64, m8, ta, mu
 ; CHECK-NEXT:    vmflt.vf v16, v24, fa5, v0.t
@@ -1374,10 +1374,10 @@ declare <vscale x 8 x double> @llvm.vp.roundtozero.nxv8f64(<vscale x 8 x double>
 define <vscale x 8 x double> @vp_roundtozero_nxv8f64(<vscale x 8 x double> %va, <vscale x 8 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vp_roundtozero_nxv8f64:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v16, v0
-; CHECK-NEXT:    lui a1, %hi(.LCPI42_0)
-; CHECK-NEXT:    fld fa5, %lo(.LCPI42_0)(a1)
 ; CHECK-NEXT:    vsetvli zero, a0, e64, m8, ta, ma
+; CHECK-NEXT:    vmv1r.v v16, v0
+; CHECK-NEXT:    lui a0, %hi(.LCPI42_0)
+; CHECK-NEXT:    fld fa5, %lo(.LCPI42_0)(a0)
 ; CHECK-NEXT:    vfabs.v v24, v8, v0.t
 ; CHECK-NEXT:    vsetvli zero, zero, e64, m8, ta, mu
 ; CHECK-NEXT:    vmflt.vf v16, v24, fa5, v0.t
@@ -1425,13 +1425,13 @@ define <vscale x 16 x double> @vp_roundtozero_nxv16f64(<vscale x 16 x double> %v
 ; CHECK-NEXT:    slli a1, a1, 3
 ; CHECK-NEXT:    sub sp, sp, a1
 ; CHECK-NEXT:    .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x08, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 8 * vlenb
+; CHECK-NEXT:    vsetvli a1, zero, e8, mf4, ta, ma
 ; CHECK-NEXT:    vmv1r.v v7, v0
 ; CHECK-NEXT:    csrr a1, vlenb
 ; CHECK-NEXT:    lui a2, %hi(.LCPI44_0)
 ; CHECK-NEXT:    srli a3, a1, 3
 ; CHECK-NEXT:    fld fa5, %lo(.LCPI44_0)(a2)
 ; CHECK-NEXT:    sub a2, a0, a1
-; CHECK-NEXT:    vsetvli a4, zero, e8, mf4, ta, ma
 ; CHECK-NEXT:    vslidedown.vx v6, v0, a3
 ; CHECK-NEXT:    sltu a3, a0, a2
 ; CHECK-NEXT:    addi a3, a3, -1
diff --git a/llvm/test/CodeGen/RISCV/rvv/rv32-spill-vector-csr.ll b/llvm/test/CodeGen/RISCV/rvv/rv32-spill-vector-csr.ll
index aef160049106b..f636ab9ebd0ce 100644
--- a/llvm/test/CodeGen/RISCV/rvv/rv32-spill-vector-csr.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/rv32-spill-vector-csr.ll
@@ -17,6 +17,7 @@ define <vscale x 1 x double> @foo(<vscale x 1 x double> %a, <vscale x 1 x double
 ; SPILL-O0-NEXT:    slli a1, a1, 1
 ; SPILL-O0-NEXT:    sub sp, sp, a1
 ; SPILL-O0-NEXT:    sw a0, 8(sp) # 4-byte Folded Spill
+; SPILL-O0-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; SPILL-O0-NEXT:    vmv1r.v v10, v9
 ; SPILL-O0-NEXT:    vmv1r.v v9, v8
 ; SPILL-O0-NEXT:    csrr a1, vlenb
diff --git a/llvm/test/CodeGen/RISCV/rvv/rv32-spill-zvlsseg.ll b/llvm/test/CodeGen/RISCV/rvv/rv32-spill-zvlsseg.ll
index b34952b64f09e..adb15f02e33a4 100644
--- a/llvm/test/CodeGen/RISCV/rvv/rv32-spill-zvlsseg.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/rv32-spill-zvlsseg.ll
@@ -15,6 +15,7 @@ define <vscale x 1 x i32> @spill_zvlsseg_nxv1i32(ptr %base, i32 %vl) nounwind {
 ; SPILL-O0-NEXT:    # implicit-def: $v8_v9
 ; SPILL-O0-NEXT:    vsetvli zero, a1, e32, mf2, tu, ma
 ; SPILL-O0-NEXT:    vlseg2e32.v v8, (a0)
+; SPILL-O0-NEXT:    vsetvli zero, a1, e32, mf2, tu, ma
 ; SPILL-O0-NEXT:    vmv1r.v v8, v9
 ; SPILL-O0-NEXT:    addi a0, sp, 16
 ; SPILL-O0-NEXT:    vs1r.v v8, (a0) # Unknown-size Folded Spill
@@ -90,6 +91,7 @@ define <vscale x 2 x i32> @spill_zvlsseg_nxv2i32(ptr %base, i32 %vl) nounwind {
 ; SPILL-O0-NEXT:    # implicit-def: $v8_v9
 ; SPILL-O0-NEXT:    vsetvli zero, a1, e32, m1, tu, ma
 ; SPILL-O0-NEXT:    vlseg2e32.v v8, (a0)
+; SPILL-O0-NEXT:    vsetvli zero, a1, e32, m1, tu, ma
 ; SPILL-O0-NEXT:    vmv1r.v v8, v9
 ; SPILL-O0-NEXT:    addi a0, sp, 16
 ; SPILL-O0-NEXT:    vs1r.v v8, (a0) # Unknown-size Folded Spill
@@ -166,6 +168,7 @@ define <vscale x 4 x i32> @spill_zvlsseg_nxv4i32(ptr %base, i32 %vl) nounwind {
 ; SPILL-O0-NEXT:    # implicit-def: $v8m2_v10m2
 ; SPILL-O0-NEXT:    vsetvli zero, a1, e32, m2, tu, ma
 ; SPILL-O0-NEXT:    vlseg2e32.v v8, (a0)
+; SPILL-O0-NEXT:    vsetvli zero, a1, e32, m2, tu, ma
 ; SPILL-O0-NEXT:    vmv2r.v v8, v10
 ; SPILL-O0-NEXT:    addi a0, sp, 16
 ; SPILL-O0-NEXT:    vs2r.v v8, (a0) # Unknown-size Folded Spill
@@ -246,6 +249,7 @@ define <vscale x 8 x i32> @spill_zvlsseg_nxv8i32(ptr %base, i32 %vl) nounwind {
 ; SPILL-O0-NEXT:    # implicit-def: $v8m4_v12m4
 ; SPILL-O0-NEXT:    vsetvli zero, a1, e32, m4, tu, ma
 ; SPILL-O0-NEXT:    vlseg2e32.v v8, (a0)
+; SPILL-O0-NEXT:    vsetvli zero, a1, e32, m4, tu, ma
 ; SPILL-O0-NEXT:    vmv4r.v v8, v12
 ; SPILL-O0-NEXT:    addi a0, sp, 16
 ; SPILL-O0-NEXT:    vs4r.v v8, (a0) # Unknown-size Folded Spill
@@ -326,6 +330,7 @@ define <vscale x 4 x i32> @spill_zvlsseg3_nxv4i32(ptr %base, i32 %vl) nounwind {
 ; SPILL-O0-NEXT:    # implicit-def: $v8m2_v10m2_v12m2
 ; SPILL-O0-NEXT:    vsetvli zero, a1, e32, m2, tu, ma
 ; SPILL-O0-NEXT:    vlseg3e32.v v8, (a0)
+; SPILL-O0-NEXT:    vsetvli zero, a1, e32, m2, tu, ma
 ; SPILL-O0-NEXT:    vmv2r.v v8, v10
 ; SPILL-O0-NEXT:    addi a0, sp, 16
 ; SPILL-O0-NEXT:    vs2r.v v8, (a0) # Unknown-size Folded Spill
diff --git a/llvm/test/CodeGen/RISCV/rvv/rv64-spill-vector-csr.ll b/llvm/test/CodeGen/RISCV/rvv/rv64-spill-vector-csr.ll
index c7c44fb0e1215..2cd80ef79bd82 100644
--- a/llvm/test/CodeGen/RISCV/rvv/rv64-spill-vector-csr.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/rv64-spill-vector-csr.ll
@@ -20,6 +20,7 @@ define <vscale x 1 x double> @foo(<vscale x 1 x double> %a, <vscale x 1 x double
 ; SPILL-O0-NEXT:    slli a1, a1, 1
 ; SPILL-O0-NEXT:    sub sp, sp, a1
 ; SPILL-O0-NEXT:    sd a0, 16(sp) # 8-byte Folded Spill
+; SPILL-O0-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; SPILL-O0-NEXT:    vmv1r.v v10, v9
 ; SPILL-O0-NEXT:    vmv1r.v v9, v8
 ; SPILL-O0-NEXT:    csrr a1, vlenb
diff --git a/llvm/test/CodeGen/RISCV/rvv/rv64-spill-zvlsseg.ll b/llvm/test/CodeGen/RISCV/rvv/rv64-spill-zvlsseg.ll
index 361adb55ef12f..ff0f1d7748668 100644
--- a/llvm/test/CodeGen/RISCV/rvv/rv64-spill-zvlsseg.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/rv64-spill-zvlsseg.ll
@@ -15,6 +15,7 @@ define <vscale x 1 x i32> @spill_zvlsseg_nxv1i32(ptr %base, i64 %vl) nounwind {
 ; SPILL-O0-NEXT:    # implicit-def: $v8_v9
 ; SPILL-O0-NEXT:    vsetvli zero, a1, e32, mf2, tu, ma
 ; SPILL-O0-NEXT:    vlseg2e32.v v8, (a0)
+; SPILL-O0-NEXT:    vsetvli zero, a1, e32, mf2, tu, ma
 ; SPILL-O0-NEXT:    vmv1r.v v8, v9
 ; SPILL-O0-NEXT:    addi a0, sp, 16
 ; SPILL-O0-NEXT:    vs1r.v v8, (a0) # Unknown-size Folded Spill
@@ -90,6 +91,7 @@ define <vscale x 2 x i32> @spill_zvlsseg_nxv2i32(ptr %base, i64 %vl) nounwind {
 ; SPILL-O0-NEXT:    # implicit-def: $v8_v9
 ; SPILL-O0-NEXT:    vsetvli zero, a1, e32, m1, tu, ma
 ; SPILL-O0-NEXT:    vlseg2e32.v v8, (a0)
+; SPILL-O0-NEXT:    vsetvli zero, a1, e32, m1, tu, ma
 ; SPILL-O0-NEXT:    vmv1r.v v8, v9
 ; SPILL-O0-NEXT:    addi a0, sp, 16
 ; SPILL-O0-NEXT:    vs1r.v v8, (a0) # Unknown-size Folded Spill
@@ -166,6 +168,7 @@ define <vscale x 4 x i32> @spill_zvlsseg_nxv4i32(ptr %base, i64 %vl) nounwind {
 ; SPILL-O0-NEXT:    # implicit-def: $v8m2_v10m2
 ; SPILL-O0-NEXT:    vsetvli zero, a1, e32, m2, tu, ma
 ; SPILL-O0-NEXT:    vlseg2e32.v v8, (a0)
+; SPILL-O0-NEXT:    vsetvli zero, a1, e32, m2, tu, ma
 ; SPILL-O0-NEXT:    vmv2r.v v8, v10
 ; SPILL-O0-NEXT:    addi a0, sp, 16
 ; SPILL-O0-NEXT:    vs2r.v v8, (a0) # Unknown-size Folded Spill
@@ -246,6 +249,7 @@ define <vscale x 8 x i32> @spill_zvlsseg_nxv8i32(ptr %base, i64 %vl) nounwind {
 ; SPILL-O0-NEXT:    # implicit-def: $v8m4_v12m4
 ; SPILL-O0-NEXT:    vsetvli zero, a1, e32, m4, tu, ma
 ; SPILL-O0-NEXT:    vlseg2e32.v v8, (a0)
+; SPILL-O0-NEXT:    vsetvli zero, a1, e32, m4, tu, ma
 ; SPILL-O0-NEXT:    vmv4r.v v8, v12
 ; SPILL-O0-NEXT:    addi a0, sp, 16
 ; SPILL-O0-NEXT:    vs4r.v v8, (a0) # Unknown-size Folded Spill
@@ -326,6 +330,7 @@ define <vscale x 4 x i32> @spill_zvlsseg3_nxv4i32(ptr %base, i64 %vl) nounwind {
 ; SPILL-O0-NEXT:    # implicit-def: $v8m2_v10m2_v12m2
 ; SPILL-O0-NEXT:    vsetvli zero, a1, e32, m2, tu, ma
 ; SPILL-O0-NEXT:    vlseg3e32.v v8, (a0)
+; SPILL-O0-NEXT:    vsetvli zero, a1, e32, m2, tu, ma
 ; SPILL-O0-NEXT:    vmv2r.v v8, v10
 ; SPILL-O0-NEXT:    addi a0, sp, 16
 ; SPILL-O0-NEXT:    vs2r.v v8, (a0) # Unknown-size Folded Spill
diff --git a/llvm/test/CodeGen/RISCV/rvv/rvv-args-by-mem.ll b/llvm/test/CodeGen/RISCV/rvv/rvv-args-by-mem.ll
index b27ba14e85c83..53d1666c30e96 100644
--- a/llvm/test/CodeGen/RISCV/rvv/rvv-args-by-mem.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/rvv-args-by-mem.ll
@@ -47,6 +47,7 @@ define <vscale x 16 x i32> @foo(i32 %0, i32 %1, i32 %2, i32 %3, i32 %4, i32 %5,
 ; CHECK-NEXT:    vs8r.v v8, (t1)
 ; CHECK-NEXT:    sd t1, 0(sp)
 ; CHECK-NEXT:    sd t0, 8(sp)
+; CHECK-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-NEXT:    vmv8r.v v16, v8
 ; CHECK-NEXT:    call bar
 ; CHECK-NEXT:    addi sp, sp, 16
diff --git a/llvm/test/CodeGen/RISCV/rvv/rvv-peephole-vmerge-vops.ll b/llvm/test/CodeGen/RISCV/rvv/rvv-peephole-vmerge-vops.ll
index 23ebfade6f6b0..d329979857a6b 100644
--- a/llvm/test/CodeGen/RISCV/rvv/rvv-peephole-vmerge-vops.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/rvv-peephole-vmerge-vops.ll
@@ -941,8 +941,8 @@ declare <vscale x 2 x i32> @llvm.riscv.vredsum.nxv2i32.nxv2i32(
 define <vscale x 2 x i32> @vredsum(<vscale x 2 x i32> %passthru, <vscale x 2 x i32> %x, <vscale x 2 x i32> %y, <vscale x 2 x i1> %m, i64 %vl) {
 ; CHECK-LABEL: vredsum:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v11, v8
 ; CHECK-NEXT:    vsetvli zero, a0, e32, m1, tu, ma
+; CHECK-NEXT:    vmv1r.v v11, v8
 ; CHECK-NEXT:    vredsum.vs v11, v9, v10
 ; CHECK-NEXT:    vmerge.vvm v8, v8, v11, v0
 ; CHECK-NEXT:    ret
@@ -965,8 +965,8 @@ define <vscale x 2 x float> @vfredusum(<vscale x 2 x float> %passthru, <vscale x
 ; CHECK-LABEL: vfredusum:
 ; CHECK:       # %bb.0:
 ; CHECK-NEXT:    fsrmi a1, 0
-; CHECK-NEXT:    vmv1r.v v11, v8
 ; CHECK-NEXT:    vsetvli zero, a0, e32, m1, tu, ma
+; CHECK-NEXT:    vmv1r.v v11, v8
 ; CHECK-NEXT:    vfredusum.vs v11, v9, v10
 ; CHECK-NEXT:    vmerge.vvm v8, v8, v11, v0
 ; CHECK-NEXT:    fsrm a1
@@ -1016,8 +1016,8 @@ define <vscale x 2 x float> @vfredusum_allones_mask(<vscale x 2 x float> %passth
 define <vscale x 2 x i32> @unfoldable_vredsum_allones_mask_diff_vl(<vscale x 2 x i32> %passthru, <vscale x 2 x i32> %x, <vscale x 2 x i32> %y) {
 ; CHECK-LABEL: unfoldable_vredsum_allones_mask_diff_vl:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v11, v8
 ; CHECK-NEXT:    vsetvli a0, zero, e32, m1, tu, ma
+; CHECK-NEXT:    vmv1r.v v11, v8
 ; CHECK-NEXT:    vredsum.vs v11, v9, v10
 ; CHECK-NEXT:    vsetivli zero, 1, e32, m1, tu, ma
 ; CHECK-NEXT:    vmv.v.v v8, v11
diff --git a/llvm/test/CodeGen/RISCV/rvv/setcc-fp-vp.ll b/llvm/test/CodeGen/RISCV/rvv/setcc-fp-vp.ll
index 6c11e9413525e..70b53841bff4c 100644
--- a/llvm/test/CodeGen/RISCV/rvv/setcc-fp-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/setcc-fp-vp.ll
@@ -1473,6 +1473,7 @@ define <vscale x 64 x i1> @fcmp_oeq_vv_nxv64bf16(<vscale x 64 x bfloat> %va, <vs
 ; CHECK-NEXT:    add a1, a1, a3
 ; CHECK-NEXT:    sub sp, sp, a1
 ; CHECK-NEXT:    .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x29, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 41 * vlenb
+; CHECK-NEXT:    vsetvli a1, zero, e8, m1, ta, ma
 ; CHECK-NEXT:    vmv1r.v v24, v0
 ; CHECK-NEXT:    csrr a1, vlenb
 ; CHECK-NEXT:    mv a3, a1
@@ -1504,7 +1505,6 @@ define <vscale x 64 x i1> @fcmp_oeq_vv_nxv64bf16(<vscale x 64 x bfloat> %va, <vs
 ; CHECK-NEXT:    add t0, sp, t0
 ; CHECK-NEXT:    addi t0, t0, 16
 ; CHECK-NEXT:    vs1r.v v24, (t0) # Unknown-size Folded Spill
-; CHECK-NEXT:    vsetvli t0, zero, e8, m1, ta, ma
 ; CHECK-NEXT:    vslidedown.vx v25, v24, a1
 ; CHECK-NEXT:    vsetvli t0, zero, e8, mf2, ta, ma
 ; CHECK-NEXT:    vslidedown.vx v24, v25, a3
@@ -3720,6 +3720,7 @@ define <vscale x 64 x i1> @fcmp_oeq_vv_nxv64f16(<vscale x 64 x half> %va, <vscal
 ; ZVFH-NEXT:    slli a1, a1, 4
 ; ZVFH-NEXT:    sub sp, sp, a1
 ; ZVFH-NEXT:    .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x10, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 16 * vlenb
+; ZVFH-NEXT:    vsetvli a1, zero, e8, m1, ta, ma
 ; ZVFH-NEXT:    vmv1r.v v24, v0
 ; ZVFH-NEXT:    csrr a1, vlenb
 ; ZVFH-NEXT:    slli a1, a1, 3
@@ -3738,7 +3739,6 @@ define <vscale x 64 x i1> @fcmp_oeq_vv_nxv64f16(<vscale x 64 x half> %va, <vscal
 ; ZVFH-NEXT:    vl8re16.v v0, (a0)
 ; ZVFH-NEXT:    addi a0, sp, 16
 ; ZVFH-NEXT:    vs8r.v v0, (a0) # Unknown-size Folded Spill
-; ZVFH-NEXT:    vsetvli a0, zero, e8, m1, ta, ma
 ; ZVFH-NEXT:    vslidedown.vx v0, v24, a1
 ; ZVFH-NEXT:    and a4, a4, a5
 ; ZVFH-NEXT:    vsetvli zero, a4, e16, m8, ta, ma
@@ -3781,6 +3781,7 @@ define <vscale x 64 x i1> @fcmp_oeq_vv_nxv64f16(<vscale x 64 x half> %va, <vscal
 ; ZVFHMIN-NEXT:    add a1, a1, a3
 ; ZVFHMIN-NEXT:    sub sp, sp, a1
 ; ZVFHMIN-NEXT:    .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x29, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 41 * vlenb
+; ZVFHMIN-NEXT:    vsetvli a1, zero, e8, m1, ta, ma
 ; ZVFHMIN-NEXT:    vmv1r.v v24, v0
 ; ZVFHMIN-NEXT:    csrr a1, vlenb
 ; ZVFHMIN-NEXT:    mv a3, a1
@@ -3812,7 +3813,6 @@ define <vscale x 64 x i1> @fcmp_oeq_vv_nxv64f16(<vscale x 64 x half> %va, <vscal
 ; ZVFHMIN-NEXT:    add t0, sp, t0
 ; ZVFHMIN-NEXT:    addi t0, t0, 16
 ; ZVFHMIN-NEXT:    vs1r.v v24, (t0) # Unknown-size Folded Spill
-; ZVFHMIN-NEXT:    vsetvli t0, zero, e8, m1, ta, ma
 ; ZVFHMIN-NEXT:    vslidedown.vx v25, v24, a1
 ; ZVFHMIN-NEXT:    vsetvli t0, zero, e8, mf2, ta, ma
 ; ZVFHMIN-NEXT:    vslidedown.vx v24, v25, a3
diff --git a/llvm/test/CodeGen/RISCV/rvv/setcc-int-vp.ll b/llvm/test/CodeGen/RISCV/rvv/setcc-int-vp.ll
index e8099c2d08a5f..2392cb77c753d 100644
--- a/llvm/test/CodeGen/RISCV/rvv/setcc-int-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/setcc-int-vp.ll
@@ -1092,6 +1092,7 @@ define <vscale x 128 x i1> @icmp_eq_vv_nxv128i8(<vscale x 128 x i8> %va, <vscale
 ; CHECK-NEXT:    slli a1, a1, 4
 ; CHECK-NEXT:    sub sp, sp, a1
 ; CHECK-NEXT:    .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x10, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 16 * vlenb
+; CHECK-NEXT:    vsetvli a1, zero, e8, m8, ta, ma
 ; CHECK-NEXT:    vmv1r.v v7, v0
 ; CHECK-NEXT:    csrr a1, vlenb
 ; CHECK-NEXT:    slli a1, a1, 3
@@ -1099,7 +1100,6 @@ define <vscale x 128 x i1> @icmp_eq_vv_nxv128i8(<vscale x 128 x i8> %va, <vscale
 ; CHECK-NEXT:    addi a1, a1, 16
 ; CHECK-NEXT:    vs8r.v v8, (a1) # Unknown-size Folded Spill
 ; CHECK-NEXT:    csrr a1, vlenb
-; CHECK-NEXT:    vsetvli a4, zero, e8, m8, ta, ma
 ; CHECK-NEXT:    vlm.v v0, (a2)
 ; CHECK-NEXT:    slli a1, a1, 3
 ; CHECK-NEXT:    add a2, a0, a1
@@ -1143,8 +1143,8 @@ define <vscale x 128 x i1> @icmp_eq_vv_nxv128i8(<vscale x 128 x i8> %va, <vscale
 define <vscale x 128 x i1> @icmp_eq_vx_nxv128i8(<vscale x 128 x i8> %va, i8 %b, <vscale x 128 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: icmp_eq_vx_nxv128i8:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v24, v0
 ; CHECK-NEXT:    vsetvli a3, zero, e8, m8, ta, ma
+; CHECK-NEXT:    vmv1r.v v24, v0
 ; CHECK-NEXT:    vlm.v v0, (a1)
 ; CHECK-NEXT:    csrr a1, vlenb
 ; CHECK-NEXT:    slli a1, a1, 3
@@ -1173,8 +1173,8 @@ define <vscale x 128 x i1> @icmp_eq_vx_nxv128i8(<vscale x 128 x i8> %va, i8 %b,
 define <vscale x 128 x i1> @icmp_eq_vx_swap_nxv128i8(<vscale x 128 x i8> %va, i8 %b, <vscale x 128 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: icmp_eq_vx_swap_nxv128i8:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v24, v0
 ; CHECK-NEXT:    vsetvli a3, zero, e8, m8, ta, ma
+; CHECK-NEXT:    vmv1r.v v24, v0
 ; CHECK-NEXT:    vlm.v v0, (a1)
 ; CHECK-NEXT:    csrr a1, vlenb
 ; CHECK-NEXT:    slli a1, a1, 3
@@ -2244,6 +2244,7 @@ define <vscale x 32 x i1> @icmp_eq_vv_nxv32i32(<vscale x 32 x i32> %va, <vscale
 ; CHECK-NEXT:    slli a1, a1, 4
 ; CHECK-NEXT:    sub sp, sp, a1
 ; CHECK-NEXT:    .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x10, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 16 * vlenb
+; CHECK-NEXT:    vsetvli a1, zero, e8, mf2, ta, ma
 ; CHECK-NEXT:    vmv1r.v v24, v0
 ; CHECK-NEXT:    csrr a1, vlenb
 ; CHECK-NEXT:    slli a1, a1, 3
@@ -2262,7 +2263,6 @@ define <vscale x 32 x i1> @icmp_eq_vv_nxv32i32(<vscale x 32 x i32> %va, <vscale
 ; CHECK-NEXT:    vl8re32.v v0, (a0)
 ; CHECK-NEXT:    addi a0, sp, 16
 ; CHECK-NEXT:    vs8r.v v0, (a0) # Unknown-size Folded Spill
-; CHECK-NEXT:    vsetvli a0, zero, e8, mf2, ta, ma
 ; CHECK-NEXT:    vslidedown.vx v0, v24, a1
 ; CHECK-NEXT:    and a4, a4, a5
 ; CHECK-NEXT:    vsetvli zero, a4, e32, m8, ta, ma
@@ -2299,11 +2299,11 @@ define <vscale x 32 x i1> @icmp_eq_vv_nxv32i32(<vscale x 32 x i32> %va, <vscale
 define <vscale x 32 x i1> @icmp_eq_vx_nxv32i32(<vscale x 32 x i32> %va, i32 %b, <vscale x 32 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: icmp_eq_vx_nxv32i32:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli a2, zero, e8, mf2, ta, ma
 ; CHECK-NEXT:    vmv1r.v v24, v0
 ; CHECK-NEXT:    csrr a3, vlenb
 ; CHECK-NEXT:    srli a2, a3, 2
 ; CHECK-NEXT:    slli a3, a3, 1
-; CHECK-NEXT:    vsetvli a4, zero, e8, mf2, ta, ma
 ; CHECK-NEXT:    vslidedown.vx v0, v0, a2
 ; CHECK-NEXT:    sub a4, a1, a3
 ; CHECK-NEXT:    sltu a5, a1, a4
@@ -2332,11 +2332,11 @@ define <vscale x 32 x i1> @icmp_eq_vx_nxv32i32(<vscale x 32 x i32> %va, i32 %b,
 define <vscale x 32 x i1> @icmp_eq_vx_swap_nxv32i32(<vscale x 32 x i32> %va, i32 %b, <vscale x 32 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: icmp_eq_vx_swap_nxv32i32:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli a2, zero, e8, mf2, ta, ma
 ; CHECK-NEXT:    vmv1r.v v24, v0
 ; CHECK-NEXT:    csrr a3, vlenb
 ; CHECK-NEXT:    srli a2, a3, 2
 ; CHECK-NEXT:    slli a3, a3, 1
-; CHECK-NEXT:    vsetvli a4, zero, e8, mf2, ta, ma
 ; CHECK-NEXT:    vslidedown.vx v0, v0, a2
 ; CHECK-NEXT:    sub a4, a1, a3
 ; CHECK-NEXT:    sltu a5, a1, a4
diff --git a/llvm/test/CodeGen/RISCV/rvv/sink-splat-operands.ll b/llvm/test/CodeGen/RISCV/rvv/sink-splat-operands.ll
index 197ba085c0359..75a5eea1cb409 100644
--- a/llvm/test/CodeGen/RISCV/rvv/sink-splat-operands.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/sink-splat-operands.ll
@@ -4865,10 +4865,10 @@ declare <4 x i1> @llvm.vp.icmp.v4i32(<4 x i32>, <4 x i32>, metadata, <4 x i1>, i
 define void @sink_splat_vp_icmp(ptr nocapture %x, i32 signext %y, <4 x i1> %m, i32 zeroext %vl) {
 ; CHECK-LABEL: sink_splat_vp_icmp:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetivli zero, 4, e32, m1, ta, ma
 ; CHECK-NEXT:    vmv1r.v v8, v0
 ; CHECK-NEXT:    lui a3, 1
 ; CHECK-NEXT:    add a3, a0, a3
-; CHECK-NEXT:    vsetivli zero, 4, e32, m1, ta, ma
 ; CHECK-NEXT:    vmv.v.i v9, 0
 ; CHECK-NEXT:  .LBB102_1: # %vector.body
 ; CHECK-NEXT:    # =>This Inner Loop Header: Depth=1
@@ -4906,10 +4906,10 @@ declare <4 x i1> @llvm.vp.fcmp.v4f32(<4 x float>, <4 x float>, metadata, <4 x i1
 define void @sink_splat_vp_fcmp(ptr nocapture %x, float %y, <4 x i1> %m, i32 zeroext %vl) {
 ; CHECK-LABEL: sink_splat_vp_fcmp:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetivli zero, 4, e32, m1, ta, ma
 ; CHECK-NEXT:    vmv1r.v v8, v0
 ; CHECK-NEXT:    lui a2, 1
 ; CHECK-NEXT:    add a2, a0, a2
-; CHECK-NEXT:    vsetivli zero, 4, e32, m1, ta, ma
 ; CHECK-NEXT:    vmv.v.i v9, 0
 ; CHECK-NEXT:  .LBB103_1: # %vector.body
 ; CHECK-NEXT:    # =>This Inner Loop Header: Depth=1
diff --git a/llvm/test/CodeGen/RISCV/rvv/strided-vpload.ll b/llvm/test/CodeGen/RISCV/rvv/strided-vpload.ll
index f8315de324e42..ecd098edb30ae 100644
--- a/llvm/test/CodeGen/RISCV/rvv/strided-vpload.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/strided-vpload.ll
@@ -663,6 +663,7 @@ declare <vscale x 3 x double> @llvm.experimental.vp.strided.load.nxv3f64.p0.i32(
 define <vscale x 16 x double> @strided_load_nxv16f64(ptr %ptr, i64 %stride, <vscale x 16 x i1> %mask, i32 zeroext %evl) {
 ; CHECK-RV32-LABEL: strided_load_nxv16f64:
 ; CHECK-RV32:       # %bb.0:
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv1r.v v9, v0
 ; CHECK-RV32-NEXT:    csrr a4, vlenb
 ; CHECK-RV32-NEXT:    sub a2, a3, a4
@@ -688,6 +689,7 @@ define <vscale x 16 x double> @strided_load_nxv16f64(ptr %ptr, i64 %stride, <vsc
 ;
 ; CHECK-RV64-LABEL: strided_load_nxv16f64:
 ; CHECK-RV64:       # %bb.0:
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv1r.v v9, v0
 ; CHECK-RV64-NEXT:    csrr a4, vlenb
 ; CHECK-RV64-NEXT:    sub a3, a2, a4
@@ -765,6 +767,7 @@ declare <vscale x 16 x double> @llvm.experimental.vp.strided.load.nxv16f64.p0.i6
 define <vscale x 16 x double> @strided_load_nxv17f64(ptr %ptr, i64 %stride, <vscale x 17 x i1> %mask, i32 zeroext %evl, ptr %hi_ptr) {
 ; CHECK-RV32-LABEL: strided_load_nxv17f64:
 ; CHECK-RV32:       # %bb.0:
+; CHECK-RV32-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV32-NEXT:    vmv1r.v v8, v0
 ; CHECK-RV32-NEXT:    csrr a2, vlenb
 ; CHECK-RV32-NEXT:    slli a7, a2, 1
@@ -812,6 +815,7 @@ define <vscale x 16 x double> @strided_load_nxv17f64(ptr %ptr, i64 %stride, <vsc
 ;
 ; CHECK-RV64-LABEL: strided_load_nxv17f64:
 ; CHECK-RV64:       # %bb.0:
+; CHECK-RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-RV64-NEXT:    vmv1r.v v8, v0
 ; CHECK-RV64-NEXT:    csrr a4, vlenb
 ; CHECK-RV64-NEXT:    slli a7, a4, 1
diff --git a/llvm/test/CodeGen/RISCV/rvv/strided-vpstore.ll b/llvm/test/CodeGen/RISCV/rvv/strided-vpstore.ll
index 98ec99bcfea33..9dcee7e5cb7d1 100644
--- a/llvm/test/CodeGen/RISCV/rvv/strided-vpstore.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/strided-vpstore.ll
@@ -615,6 +615,7 @@ define void @strided_store_nxv17f64(<vscale x 17 x double> %v, ptr %ptr, i32 sig
 ; CHECK-NEXT:    slli a4, a4, 3
 ; CHECK-NEXT:    sub sp, sp, a4
 ; CHECK-NEXT:    .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x08, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 8 * vlenb
+; CHECK-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-NEXT:    vmv1r.v v24, v0
 ; CHECK-NEXT:    addi a4, sp, 16
 ; CHECK-NEXT:    vs8r.v v16, (a4) # Unknown-size Folded Spill
diff --git a/llvm/test/CodeGen/RISCV/rvv/undef-earlyclobber-chain.ll b/llvm/test/CodeGen/RISCV/rvv/undef-earlyclobber-chain.ll
index ab13c78da05e8..c9f9a79733003 100644
--- a/llvm/test/CodeGen/RISCV/rvv/undef-earlyclobber-chain.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/undef-earlyclobber-chain.ll
@@ -158,8 +158,8 @@ declare <vscale x 8 x i8> @llvm.riscv.vrgatherei16.vv.nxv8i8.i64(<vscale x 8 x i
 define void @repeat_shuffle(<2 x double> %v, ptr noalias %q) {
 ; CHECK-LABEL: repeat_shuffle:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv2r.v v10, v8
 ; CHECK-NEXT:    vsetivli zero, 4, e64, m2, ta, ma
+; CHECK-NEXT:    vmv2r.v v10, v8
 ; CHECK-NEXT:    vslideup.vi v10, v8, 2
 ; CHECK-NEXT:    vse64.v v10, (a0)
 ; CHECK-NEXT:    ret
diff --git a/llvm/test/CodeGen/RISCV/rvv/vadd-vp.ll b/llvm/test/CodeGen/RISCV/rvv/vadd-vp.ll
index ebd550013ec78..fee6799e992f3 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vadd-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vadd-vp.ll
@@ -565,8 +565,8 @@ declare <vscale x 128 x i8> @llvm.vp.add.nxv128i8(<vscale x 128 x i8>, <vscale x
 define <vscale x 128 x i8> @vadd_vi_nxv128i8(<vscale x 128 x i8> %va, <vscale x 128 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vadd_vi_nxv128i8:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v24, v0
 ; CHECK-NEXT:    vsetvli a2, zero, e8, m8, ta, ma
+; CHECK-NEXT:    vmv1r.v v24, v0
 ; CHECK-NEXT:    vlm.v v0, (a0)
 ; CHECK-NEXT:    csrr a0, vlenb
 ; CHECK-NEXT:    slli a0, a0, 3
@@ -1343,11 +1343,11 @@ declare <vscale x 32 x i32> @llvm.vp.add.nxv32i32(<vscale x 32 x i32>, <vscale x
 define <vscale x 32 x i32> @vadd_vi_nxv32i32(<vscale x 32 x i32> %va, <vscale x 32 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vadd_vi_nxv32i32:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli a1, zero, e8, mf2, ta, ma
 ; CHECK-NEXT:    vmv1r.v v24, v0
 ; CHECK-NEXT:    csrr a1, vlenb
 ; CHECK-NEXT:    srli a2, a1, 2
 ; CHECK-NEXT:    slli a1, a1, 1
-; CHECK-NEXT:    vsetvli a3, zero, e8, mf2, ta, ma
 ; CHECK-NEXT:    vslidedown.vx v0, v0, a2
 ; CHECK-NEXT:    sub a2, a0, a1
 ; CHECK-NEXT:    sltu a3, a0, a2
@@ -1399,11 +1399,11 @@ declare i32 @llvm.vscale.i32()
 define <vscale x 32 x i32> @vadd_vi_nxv32i32_evl_nx8(<vscale x 32 x i32> %va, <vscale x 32 x i1> %m) {
 ; CHECK-LABEL: vadd_vi_nxv32i32_evl_nx8:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli a0, zero, e8, mf2, ta, ma
 ; CHECK-NEXT:    vmv1r.v v24, v0
 ; CHECK-NEXT:    csrr a0, vlenb
 ; CHECK-NEXT:    srli a2, a0, 2
 ; CHECK-NEXT:    slli a1, a0, 1
-; CHECK-NEXT:    vsetvli a3, zero, e8, mf2, ta, ma
 ; CHECK-NEXT:    vslidedown.vx v0, v0, a2
 ; CHECK-NEXT:    sub a2, a0, a1
 ; CHECK-NEXT:    sltu a3, a0, a2
diff --git a/llvm/test/CodeGen/RISCV/rvv/vcpop.ll b/llvm/test/CodeGen/RISCV/rvv/vcpop.ll
index e59a9174b03d9..6b35e4767b239 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vcpop.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vcpop.ll
@@ -43,9 +43,9 @@ declare iXLen @llvm.riscv.vcpop.mask.iXLen.nxv1i1(
 define iXLen @intrinsic_vcpop_mask_m_nxv1i1(<vscale x 1 x i1> %0, <vscale x 1 x i1> %1, iXLen %2) nounwind {
 ; CHECK-LABEL: intrinsic_vcpop_mask_m_nxv1i1:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e8, mf8, ta, ma
 ; CHECK-NEXT:    vmv1r.v v9, v0
 ; CHECK-NEXT:    vmv1r.v v0, v8
-; CHECK-NEXT:    vsetvli zero, a0, e8, mf8, ta, ma
 ; CHECK-NEXT:    vcpop.m a0, v9, v0.t
 ; CHECK-NEXT:    ret
 entry:
@@ -97,9 +97,9 @@ declare iXLen @llvm.riscv.vcpop.mask.iXLen.nxv2i1(
 define iXLen @intrinsic_vcpop_mask_m_nxv2i1(<vscale x 2 x i1> %0, <vscale x 2 x i1> %1, iXLen %2) nounwind {
 ; CHECK-LABEL: intrinsic_vcpop_mask_m_nxv2i1:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e8, mf4, ta, ma
 ; CHECK-NEXT:    vmv1r.v v9, v0
 ; CHECK-NEXT:    vmv1r.v v0, v8
-; CHECK-NEXT:    vsetvli zero, a0, e8, mf4, ta, ma
 ; CHECK-NEXT:    vcpop.m a0, v9, v0.t
 ; CHECK-NEXT:    ret
 entry:
@@ -137,9 +137,9 @@ declare iXLen @llvm.riscv.vcpop.mask.iXLen.nxv4i1(
 define iXLen @intrinsic_vcpop_mask_m_nxv4i1(<vscale x 4 x i1> %0, <vscale x 4 x i1> %1, iXLen %2) nounwind {
 ; CHECK-LABEL: intrinsic_vcpop_mask_m_nxv4i1:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e8, mf2, ta, ma
 ; CHECK-NEXT:    vmv1r.v v9, v0
 ; CHECK-NEXT:    vmv1r.v v0, v8
-; CHECK-NEXT:    vsetvli zero, a0, e8, mf2, ta, ma
 ; CHECK-NEXT:    vcpop.m a0, v9, v0.t
 ; CHECK-NEXT:    ret
 entry:
@@ -177,9 +177,9 @@ declare iXLen @llvm.riscv.vcpop.mask.iXLen.nxv8i1(
 define iXLen @intrinsic_vcpop_mask_m_nxv8i1(<vscale x 8 x i1> %0, <vscale x 8 x i1> %1, iXLen %2) nounwind {
 ; CHECK-LABEL: intrinsic_vcpop_mask_m_nxv8i1:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e8, m1, ta, ma
 ; CHECK-NEXT:    vmv1r.v v9, v0
 ; CHECK-NEXT:    vmv1r.v v0, v8
-; CHECK-NEXT:    vsetvli zero, a0, e8, m1, ta, ma
 ; CHECK-NEXT:    vcpop.m a0, v9, v0.t
 ; CHECK-NEXT:    ret
 entry:
@@ -217,9 +217,9 @@ declare iXLen @llvm.riscv.vcpop.mask.iXLen.nxv16i1(
 define iXLen @intrinsic_vcpop_mask_m_nxv16i1(<vscale x 16 x i1> %0, <vscale x 16 x i1> %1, iXLen %2) nounwind {
 ; CHECK-LABEL: intrinsic_vcpop_mask_m_nxv16i1:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e8, m2, ta, ma
 ; CHECK-NEXT:    vmv1r.v v9, v0
 ; CHECK-NEXT:    vmv1r.v v0, v8
-; CHECK-NEXT:    vsetvli zero, a0, e8, m2, ta, ma
 ; CHECK-NEXT:    vcpop.m a0, v9, v0.t
 ; CHECK-NEXT:    ret
 entry:
@@ -257,9 +257,9 @@ declare iXLen @llvm.riscv.vcpop.mask.iXLen.nxv32i1(
 define iXLen @intrinsic_vcpop_mask_m_nxv32i1(<vscale x 32 x i1> %0, <vscale x 32 x i1> %1, iXLen %2) nounwind {
 ; CHECK-LABEL: intrinsic_vcpop_mask_m_nxv32i1:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e8, m4, ta, ma
 ; CHECK-NEXT:    vmv1r.v v9, v0
 ; CHECK-NEXT:    vmv1r.v v0, v8
-; CHECK-NEXT:    vsetvli zero, a0, e8, m4, ta, ma
 ; CHECK-NEXT:    vcpop.m a0, v9, v0.t
 ; CHECK-NEXT:    ret
 entry:
@@ -297,9 +297,9 @@ declare iXLen @llvm.riscv.vcpop.mask.iXLen.nxv64i1(
 define iXLen @intrinsic_vcpop_mask_m_nxv64i1(<vscale x 64 x i1> %0, <vscale x 64 x i1> %1, iXLen %2) nounwind {
 ; CHECK-LABEL: intrinsic_vcpop_mask_m_nxv64i1:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e8, m8, ta, ma
 ; CHECK-NEXT:    vmv1r.v v9, v0
 ; CHECK-NEXT:    vmv1r.v v0, v8
-; CHECK-NEXT:    vsetvli zero, a0, e8, m8, ta, ma
 ; CHECK-NEXT:    vcpop.m a0, v9, v0.t
 ; CHECK-NEXT:    ret
 entry:
diff --git a/llvm/test/CodeGen/RISCV/rvv/vector-deinterleave-fixed.ll b/llvm/test/CodeGen/RISCV/rvv/vector-deinterleave-fixed.ll
index 6de846b2582da..2fc5b40a89afa 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vector-deinterleave-fixed.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vector-deinterleave-fixed.ll
@@ -7,8 +7,8 @@
 define {<16 x i1>, <16 x i1>} @vector_deinterleave_v16i1_v32i1(<32 x i1> %vec) {
 ; CHECK-LABEL: vector_deinterleave_v16i1_v32i1:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v8, v0
 ; CHECK-NEXT:    vsetivli zero, 2, e8, mf4, ta, ma
+; CHECK-NEXT:    vmv1r.v v8, v0
 ; CHECK-NEXT:    vslidedown.vi v0, v0, 2
 ; CHECK-NEXT:    vsetivli zero, 16, e8, m1, ta, ma
 ; CHECK-NEXT:    vmv.v.i v10, 0
diff --git a/llvm/test/CodeGen/RISCV/rvv/vector-deinterleave.ll b/llvm/test/CodeGen/RISCV/rvv/vector-deinterleave.ll
index 99743066c79a8..2291475cef014 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vector-deinterleave.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vector-deinterleave.ll
@@ -133,8 +133,8 @@ ret {<vscale x 64 x i1>, <vscale x 64 x i1>} %retval
 define {<vscale x 64 x i8>, <vscale x 64 x i8>} @vector_deinterleave_nxv64i8_nxv128i8(<vscale x 128 x i8> %vec) {
 ; CHECK-LABEL: vector_deinterleave_nxv64i8_nxv128i8:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv8r.v v24, v8
 ; CHECK-NEXT:    vsetvli a0, zero, e8, m4, ta, ma
+; CHECK-NEXT:    vmv8r.v v24, v8
 ; CHECK-NEXT:    vnsrl.wi v8, v24, 0
 ; CHECK-NEXT:    vnsrl.wi v0, v24, 8
 ; CHECK-NEXT:    vnsrl.wi v12, v16, 0
@@ -148,8 +148,8 @@ ret {<vscale x 64 x i8>, <vscale x 64 x i8>} %retval
 define {<vscale x 32 x i16>, <vscale x 32 x i16>} @vector_deinterleave_nxv32i16_nxv64i16(<vscale x 64 x i16> %vec) {
 ; CHECK-LABEL: vector_deinterleave_nxv32i16_nxv64i16:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv8r.v v24, v8
 ; CHECK-NEXT:    vsetvli a0, zero, e16, m4, ta, ma
+; CHECK-NEXT:    vmv8r.v v24, v8
 ; CHECK-NEXT:    vnsrl.wi v8, v24, 0
 ; CHECK-NEXT:    vnsrl.wi v0, v24, 16
 ; CHECK-NEXT:    vnsrl.wi v12, v16, 0
@@ -163,9 +163,9 @@ ret {<vscale x 32 x i16>, <vscale x 32 x i16>} %retval
 define {<vscale x 16 x i32>, <vscale x 16 x i32>} @vector_deinterleave_nxv16i32_nxvv32i32(<vscale x 32 x i32> %vec) {
 ; CHECK-LABEL: vector_deinterleave_nxv16i32_nxvv32i32:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli a0, zero, e32, m4, ta, ma
 ; CHECK-NEXT:    vmv8r.v v24, v16
 ; CHECK-NEXT:    li a0, 32
-; CHECK-NEXT:    vsetvli a1, zero, e32, m4, ta, ma
 ; CHECK-NEXT:    vnsrl.wx v20, v24, a0
 ; CHECK-NEXT:    vnsrl.wx v16, v8, a0
 ; CHECK-NEXT:    vnsrl.wi v0, v8, 0
@@ -374,8 +374,8 @@ declare {<vscale x 2 x double>, <vscale x 2 x double>} @llvm.vector.deinterleave
 define {<vscale x 32 x bfloat>, <vscale x 32 x bfloat>} @vector_deinterleave_nxv32bf16_nxv64bf16(<vscale x 64 x bfloat> %vec) {
 ; CHECK-LABEL: vector_deinterleave_nxv32bf16_nxv64bf16:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv8r.v v24, v8
 ; CHECK-NEXT:    vsetvli a0, zero, e16, m4, ta, ma
+; CHECK-NEXT:    vmv8r.v v24, v8
 ; CHECK-NEXT:    vnsrl.wi v8, v24, 0
 ; CHECK-NEXT:    vnsrl.wi v0, v24, 16
 ; CHECK-NEXT:    vnsrl.wi v12, v16, 0
@@ -389,8 +389,8 @@ ret {<vscale x 32 x bfloat>, <vscale x 32 x bfloat>} %retval
 define {<vscale x 32 x half>, <vscale x 32 x half>} @vector_deinterleave_nxv32f16_nxv64f16(<vscale x 64 x half> %vec) {
 ; CHECK-LABEL: vector_deinterleave_nxv32f16_nxv64f16:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv8r.v v24, v8
 ; CHECK-NEXT:    vsetvli a0, zero, e16, m4, ta, ma
+; CHECK-NEXT:    vmv8r.v v24, v8
 ; CHECK-NEXT:    vnsrl.wi v8, v24, 0
 ; CHECK-NEXT:    vnsrl.wi v0, v24, 16
 ; CHECK-NEXT:    vnsrl.wi v12, v16, 0
@@ -404,9 +404,9 @@ ret {<vscale x 32 x half>, <vscale x 32 x half>} %retval
 define {<vscale x 16 x float>, <vscale x 16 x float>} @vector_deinterleave_nxv16f32_nxv32f32(<vscale x 32 x float> %vec) {
 ; CHECK-LABEL: vector_deinterleave_nxv16f32_nxv32f32:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli a0, zero, e32, m4, ta, ma
 ; CHECK-NEXT:    vmv8r.v v24, v16
 ; CHECK-NEXT:    li a0, 32
-; CHECK-NEXT:    vsetvli a1, zero, e32, m4, ta, ma
 ; CHECK-NEXT:    vnsrl.wx v20, v24, a0
 ; CHECK-NEXT:    vnsrl.wx v16, v8, a0
 ; CHECK-NEXT:    vnsrl.wi v0, v8, 0
diff --git a/llvm/test/CodeGen/RISCV/rvv/vector-interleave-fixed.ll b/llvm/test/CodeGen/RISCV/rvv/vector-interleave-fixed.ll
index 7b0ac01918b9b..08aa02c7e869a 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vector-interleave-fixed.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vector-interleave-fixed.ll
@@ -91,10 +91,10 @@ define <8 x i32> @vector_interleave_v8i32_v4i32(<4 x i32> %a, <4 x i32> %b) {
 define <4 x i64> @vector_interleave_v4i64_v2i64(<2 x i64> %a, <2 x i64> %b) {
 ; CHECK-LABEL: vector_interleave_v4i64_v2i64:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetivli zero, 4, e64, m2, ta, ma
 ; CHECK-NEXT:    vmv1r.v v10, v9
 ; CHECK-NEXT:    lui a0, 12304
 ; CHECK-NEXT:    addi a0, a0, 512
-; CHECK-NEXT:    vsetivli zero, 4, e64, m2, ta, ma
 ; CHECK-NEXT:    vslideup.vi v8, v10, 2
 ; CHECK-NEXT:    vmv.s.x v10, a0
 ; CHECK-NEXT:    vsetvli zero, zero, e16, mf2, ta, ma
@@ -106,10 +106,10 @@ define <4 x i64> @vector_interleave_v4i64_v2i64(<2 x i64> %a, <2 x i64> %b) {
 ;
 ; ZVBB-LABEL: vector_interleave_v4i64_v2i64:
 ; ZVBB:       # %bb.0:
+; ZVBB-NEXT:    vsetivli zero, 4, e64, m2, ta, ma
 ; ZVBB-NEXT:    vmv1r.v v10, v9
 ; ZVBB-NEXT:    lui a0, 12304
 ; ZVBB-NEXT:    addi a0, a0, 512
-; ZVBB-NEXT:    vsetivli zero, 4, e64, m2, ta, ma
 ; ZVBB-NEXT:    vslideup.vi v8, v10, 2
 ; ZVBB-NEXT:    vmv.s.x v10, a0
 ; ZVBB-NEXT:    vsetvli zero, zero, e16, mf2, ta, ma
@@ -239,10 +239,10 @@ define <8 x float> @vector_interleave_v8f32_v4f32(<4 x float> %a, <4 x float> %b
 define <4 x double> @vector_interleave_v4f64_v2f64(<2 x double> %a, <2 x double> %b) {
 ; CHECK-LABEL: vector_interleave_v4f64_v2f64:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetivli zero, 4, e64, m2, ta, ma
 ; CHECK-NEXT:    vmv1r.v v10, v9
 ; CHECK-NEXT:    lui a0, 12304
 ; CHECK-NEXT:    addi a0, a0, 512
-; CHECK-NEXT:    vsetivli zero, 4, e64, m2, ta, ma
 ; CHECK-NEXT:    vslideup.vi v8, v10, 2
 ; CHECK-NEXT:    vmv.s.x v10, a0
 ; CHECK-NEXT:    vsetvli zero, zero, e16, mf2, ta, ma
@@ -254,10 +254,10 @@ define <4 x double> @vector_interleave_v4f64_v2f64(<2 x double> %a, <2 x double>
 ;
 ; ZVBB-LABEL: vector_interleave_v4f64_v2f64:
 ; ZVBB:       # %bb.0:
+; ZVBB-NEXT:    vsetivli zero, 4, e64, m2, ta, ma
 ; ZVBB-NEXT:    vmv1r.v v10, v9
 ; ZVBB-NEXT:    lui a0, 12304
 ; ZVBB-NEXT:    addi a0, a0, 512
-; ZVBB-NEXT:    vsetivli zero, 4, e64, m2, ta, ma
 ; ZVBB-NEXT:    vslideup.vi v8, v10, 2
 ; ZVBB-NEXT:    vmv.s.x v10, a0
 ; ZVBB-NEXT:    vsetvli zero, zero, e16, mf2, ta, ma
diff --git a/llvm/test/CodeGen/RISCV/rvv/vector-interleave-store.ll b/llvm/test/CodeGen/RISCV/rvv/vector-interleave-store.ll
index bc203e215d878..9b78f31d399d9 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vector-interleave-store.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vector-interleave-store.ll
@@ -9,9 +9,9 @@
 define void @vector_interleave_store_nxv32i1_nxv16i1(<vscale x 16 x i1> %a, <vscale x 16 x i1> %b, ptr %p) {
 ; CHECK-LABEL: vector_interleave_store_nxv32i1_nxv16i1:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli a1, zero, e8, m2, ta, ma
 ; CHECK-NEXT:    vmv1r.v v9, v0
 ; CHECK-NEXT:    vmv1r.v v0, v8
-; CHECK-NEXT:    vsetvli a1, zero, e8, m2, ta, ma
 ; CHECK-NEXT:    vmv.v.i v10, 0
 ; CHECK-NEXT:    li a1, -1
 ; CHECK-NEXT:    csrr a2, vlenb
diff --git a/llvm/test/CodeGen/RISCV/rvv/vector-interleave.ll b/llvm/test/CodeGen/RISCV/rvv/vector-interleave.ll
index 26e9afcb1d109..864acb320d8fe 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vector-interleave.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vector-interleave.ll
@@ -11,9 +11,9 @@
 define <vscale x 32 x i1> @vector_interleave_nxv32i1_nxv16i1(<vscale x 16 x i1> %a, <vscale x 16 x i1> %b) {
 ; CHECK-LABEL: vector_interleave_nxv32i1_nxv16i1:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli a0, zero, e8, m2, ta, ma
 ; CHECK-NEXT:    vmv1r.v v9, v0
 ; CHECK-NEXT:    vmv1r.v v0, v8
-; CHECK-NEXT:    vsetvli a0, zero, e8, m2, ta, ma
 ; CHECK-NEXT:    vmv.v.i v10, 0
 ; CHECK-NEXT:    li a0, -1
 ; CHECK-NEXT:    csrr a1, vlenb
@@ -32,9 +32,9 @@ define <vscale x 32 x i1> @vector_interleave_nxv32i1_nxv16i1(<vscale x 16 x i1>
 ;
 ; ZVBB-LABEL: vector_interleave_nxv32i1_nxv16i1:
 ; ZVBB:       # %bb.0:
+; ZVBB-NEXT:    vsetvli a0, zero, e8, m2, ta, mu
 ; ZVBB-NEXT:    vmv1r.v v9, v0
 ; ZVBB-NEXT:    vmv1r.v v0, v8
-; ZVBB-NEXT:    vsetvli a0, zero, e8, m2, ta, mu
 ; ZVBB-NEXT:    vmv.v.i v10, 0
 ; ZVBB-NEXT:    li a0, 1
 ; ZVBB-NEXT:    csrr a1, vlenb
@@ -160,9 +160,9 @@ declare <vscale x 4 x i64> @llvm.vector.interleave2.nxv4i64(<vscale x 2 x i64>,
 define <vscale x 128 x i1> @vector_interleave_nxv128i1_nxv64i1(<vscale x 64 x i1> %a, <vscale x 64 x i1> %b) {
 ; CHECK-LABEL: vector_interleave_nxv128i1_nxv64i1:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli a0, zero, e8, m8, ta, ma
 ; CHECK-NEXT:    vmv1r.v v9, v0
 ; CHECK-NEXT:    vmv1r.v v0, v8
-; CHECK-NEXT:    vsetvli a0, zero, e8, m8, ta, ma
 ; CHECK-NEXT:    vmv.v.i v24, 0
 ; CHECK-NEXT:    li a0, -1
 ; CHECK-NEXT:    vmerge.vim v16, v24, 1, v0
@@ -203,8 +203,8 @@ define <vscale x 128 x i1> @vector_interleave_nxv128i1_nxv64i1(<vscale x 64 x i1
 define <vscale x 128 x i8> @vector_interleave_nxv128i8_nxv64i8(<vscale x 64 x i8> %a, <vscale x 64 x i8> %b) {
 ; CHECK-LABEL: vector_interleave_nxv128i8_nxv64i8:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv8r.v v24, v8
 ; CHECK-NEXT:    vsetvli a0, zero, e8, m4, ta, ma
+; CHECK-NEXT:    vmv8r.v v24, v8
 ; CHECK-NEXT:    vwaddu.vv v8, v24, v16
 ; CHECK-NEXT:    li a0, -1
 ; CHECK-NEXT:    vwaddu.vv v0, v28, v20
@@ -215,8 +215,8 @@ define <vscale x 128 x i8> @vector_interleave_nxv128i8_nxv64i8(<vscale x 64 x i8
 ;
 ; ZVBB-LABEL: vector_interleave_nxv128i8_nxv64i8:
 ; ZVBB:       # %bb.0:
-; ZVBB-NEXT:    vmv8r.v v24, v8
 ; ZVBB-NEXT:    vsetvli a0, zero, e8, m4, ta, ma
+; ZVBB-NEXT:    vmv8r.v v24, v8
 ; ZVBB-NEXT:    vwsll.vi v8, v16, 8
 ; ZVBB-NEXT:    vwsll.vi v0, v20, 8
 ; ZVBB-NEXT:    vwaddu.wv v8, v8, v24
@@ -230,8 +230,8 @@ define <vscale x 128 x i8> @vector_interleave_nxv128i8_nxv64i8(<vscale x 64 x i8
 define <vscale x 64 x i16> @vector_interleave_nxv64i16_nxv32i16(<vscale x 32 x i16> %a, <vscale x 32 x i16> %b) {
 ; CHECK-LABEL: vector_interleave_nxv64i16_nxv32i16:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv8r.v v24, v8
 ; CHECK-NEXT:    vsetvli a0, zero, e16, m4, ta, ma
+; CHECK-NEXT:    vmv8r.v v24, v8
 ; CHECK-NEXT:    vwaddu.vv v8, v24, v16
 ; CHECK-NEXT:    li a0, -1
 ; CHECK-NEXT:    vwaddu.vv v0, v28, v20
@@ -242,8 +242,8 @@ define <vscale x 64 x i16> @vector_interleave_nxv64i16_nxv32i16(<vscale x 32 x i
 ;
 ; ZVBB-LABEL: vector_interleave_nxv64i16_nxv32i16:
 ; ZVBB:       # %bb.0:
-; ZVBB-NEXT:    vmv8r.v v24, v8
 ; ZVBB-NEXT:    vsetvli a0, zero, e16, m4, ta, ma
+; ZVBB-NEXT:    vmv8r.v v24, v8
 ; ZVBB-NEXT:    vwsll.vi v8, v16, 16
 ; ZVBB-NEXT:    vwsll.vi v0, v20, 16
 ; ZVBB-NEXT:    vwaddu.wv v8, v8, v24
@@ -257,8 +257,8 @@ define <vscale x 64 x i16> @vector_interleave_nxv64i16_nxv32i16(<vscale x 32 x i
 define <vscale x 32 x i32> @vector_interleave_nxv32i32_nxv16i32(<vscale x 16 x i32> %a, <vscale x 16 x i32> %b) {
 ; CHECK-LABEL: vector_interleave_nxv32i32_nxv16i32:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv8r.v v24, v8
 ; CHECK-NEXT:    vsetvli a0, zero, e32, m4, ta, ma
+; CHECK-NEXT:    vmv8r.v v24, v8
 ; CHECK-NEXT:    vwaddu.vv v8, v24, v16
 ; CHECK-NEXT:    li a0, -1
 ; CHECK-NEXT:    vwaddu.vv v0, v28, v20
@@ -269,9 +269,9 @@ define <vscale x 32 x i32> @vector_interleave_nxv32i32_nxv16i32(<vscale x 16 x i
 ;
 ; ZVBB-LABEL: vector_interleave_nxv32i32_nxv16i32:
 ; ZVBB:       # %bb.0:
+; ZVBB-NEXT:    vsetvli a0, zero, e32, m4, ta, ma
 ; ZVBB-NEXT:    vmv8r.v v24, v8
 ; ZVBB-NEXT:    li a0, 32
-; ZVBB-NEXT:    vsetvli a1, zero, e32, m4, ta, ma
 ; ZVBB-NEXT:    vwsll.vx v8, v16, a0
 ; ZVBB-NEXT:    vwsll.vx v0, v20, a0
 ; ZVBB-NEXT:    vwaddu.wv v8, v8, v24
@@ -575,8 +575,8 @@ declare <vscale x 4 x double> @llvm.vector.interleave2.nxv4f64(<vscale x 2 x dou
 define <vscale x 64 x bfloat> @vector_interleave_nxv64bf16_nxv32bf16(<vscale x 32 x bfloat> %a, <vscale x 32 x bfloat> %b) {
 ; CHECK-LABEL: vector_interleave_nxv64bf16_nxv32bf16:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv8r.v v24, v8
 ; CHECK-NEXT:    vsetvli a0, zero, e16, m4, ta, ma
+; CHECK-NEXT:    vmv8r.v v24, v8
 ; CHECK-NEXT:    vwaddu.vv v8, v24, v16
 ; CHECK-NEXT:    li a0, -1
 ; CHECK-NEXT:    vwaddu.vv v0, v28, v20
@@ -587,8 +587,8 @@ define <vscale x 64 x bfloat> @vector_interleave_nxv64bf16_nxv32bf16(<vscale x 3
 ;
 ; ZVBB-LABEL: vector_interleave_nxv64bf16_nxv32bf16:
 ; ZVBB:       # %bb.0:
-; ZVBB-NEXT:    vmv8r.v v24, v8
 ; ZVBB-NEXT:    vsetvli a0, zero, e16, m4, ta, ma
+; ZVBB-NEXT:    vmv8r.v v24, v8
 ; ZVBB-NEXT:    vwsll.vi v8, v16, 16
 ; ZVBB-NEXT:    vwsll.vi v0, v20, 16
 ; ZVBB-NEXT:    vwaddu.wv v8, v8, v24
@@ -602,8 +602,8 @@ define <vscale x 64 x bfloat> @vector_interleave_nxv64bf16_nxv32bf16(<vscale x 3
 define <vscale x 64 x half> @vector_interleave_nxv64f16_nxv32f16(<vscale x 32 x half> %a, <vscale x 32 x half> %b) {
 ; CHECK-LABEL: vector_interleave_nxv64f16_nxv32f16:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv8r.v v24, v8
 ; CHECK-NEXT:    vsetvli a0, zero, e16, m4, ta, ma
+; CHECK-NEXT:    vmv8r.v v24, v8
 ; CHECK-NEXT:    vwaddu.vv v8, v24, v16
 ; CHECK-NEXT:    li a0, -1
 ; CHECK-NEXT:    vwaddu.vv v0, v28, v20
@@ -614,8 +614,8 @@ define <vscale x 64 x half> @vector_interleave_nxv64f16_nxv32f16(<vscale x 32 x
 ;
 ; ZVBB-LABEL: vector_interleave_nxv64f16_nxv32f16:
 ; ZVBB:       # %bb.0:
-; ZVBB-NEXT:    vmv8r.v v24, v8
 ; ZVBB-NEXT:    vsetvli a0, zero, e16, m4, ta, ma
+; ZVBB-NEXT:    vmv8r.v v24, v8
 ; ZVBB-NEXT:    vwsll.vi v8, v16, 16
 ; ZVBB-NEXT:    vwsll.vi v0, v20, 16
 ; ZVBB-NEXT:    vwaddu.wv v8, v8, v24
@@ -629,8 +629,8 @@ define <vscale x 64 x half> @vector_interleave_nxv64f16_nxv32f16(<vscale x 32 x
 define <vscale x 32 x float> @vector_interleave_nxv32f32_nxv16f32(<vscale x 16 x float> %a, <vscale x 16 x float> %b) {
 ; CHECK-LABEL: vector_interleave_nxv32f32_nxv16f32:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv8r.v v24, v8
 ; CHECK-NEXT:    vsetvli a0, zero, e32, m4, ta, ma
+; CHECK-NEXT:    vmv8r.v v24, v8
 ; CHECK-NEXT:    vwaddu.vv v8, v24, v16
 ; CHECK-NEXT:    li a0, -1
 ; CHECK-NEXT:    vwaddu.vv v0, v28, v20
@@ -641,9 +641,9 @@ define <vscale x 32 x float> @vector_interleave_nxv32f32_nxv16f32(<vscale x 16 x
 ;
 ; ZVBB-LABEL: vector_interleave_nxv32f32_nxv16f32:
 ; ZVBB:       # %bb.0:
+; ZVBB-NEXT:    vsetvli a0, zero, e32, m4, ta, ma
 ; ZVBB-NEXT:    vmv8r.v v24, v8
 ; ZVBB-NEXT:    li a0, 32
-; ZVBB-NEXT:    vsetvli a1, zero, e32, m4, ta, ma
 ; ZVBB-NEXT:    vwsll.vx v8, v16, a0
 ; ZVBB-NEXT:    vwsll.vx v0, v20, a0
 ; ZVBB-NEXT:    vwaddu.wv v8, v8, v24
diff --git a/llvm/test/CodeGen/RISCV/rvv/vector-reassociations.ll b/llvm/test/CodeGen/RISCV/rvv/vector-reassociations.ll
index 79bd60d1702f3..6435c1c14e061 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vector-reassociations.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vector-reassociations.ll
@@ -120,8 +120,8 @@ entry:
 define <vscale x 1 x i8> @vadd_vv_passthru(<vscale x 1 x i8> %0, <vscale x 1 x i8> %1, i32 %2) nounwind {
 ; CHECK-LABEL: vadd_vv_passthru:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v10, v8
 ; CHECK-NEXT:    vsetvli zero, a0, e8, mf8, tu, ma
+; CHECK-NEXT:    vmv1r.v v10, v8
 ; CHECK-NEXT:    vadd.vv v10, v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v8
 ; CHECK-NEXT:    vadd.vv v9, v8, v8
@@ -152,8 +152,8 @@ entry:
 define <vscale x 1 x i8> @vadd_vv_passthru_negative(<vscale x 1 x i8> %0, <vscale x 1 x i8> %1, i32 %2) nounwind {
 ; CHECK-LABEL: vadd_vv_passthru_negative:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v10, v8
 ; CHECK-NEXT:    vsetvli zero, a0, e8, mf8, tu, ma
+; CHECK-NEXT:    vmv1r.v v10, v8
 ; CHECK-NEXT:    vadd.vv v10, v8, v9
 ; CHECK-NEXT:    vadd.vv v9, v8, v10
 ; CHECK-NEXT:    vadd.vv v8, v8, v9
@@ -183,8 +183,8 @@ entry:
 define <vscale x 1 x i8> @vadd_vv_mask(<vscale x 1 x i8> %0, <vscale x 1 x i8> %1, i32 %2, <vscale x 1 x i1> %m) nounwind {
 ; CHECK-LABEL: vadd_vv_mask:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v10, v8
 ; CHECK-NEXT:    vsetvli zero, a0, e8, mf8, ta, mu
+; CHECK-NEXT:    vmv1r.v v10, v8
 ; CHECK-NEXT:    vadd.vv v10, v8, v9, v0.t
 ; CHECK-NEXT:    vmv1r.v v9, v8
 ; CHECK-NEXT:    vadd.vv v9, v8, v8, v0.t
@@ -218,8 +218,8 @@ entry:
 define <vscale x 1 x i8> @vadd_vv_mask_negative(<vscale x 1 x i8> %0, <vscale x 1 x i8> %1, i32 %2, <vscale x 1 x i1> %m, <vscale x 1 x i1> %m2) nounwind {
 ; CHECK-LABEL: vadd_vv_mask_negative:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v11, v8
 ; CHECK-NEXT:    vsetvli zero, a0, e8, mf8, ta, mu
+; CHECK-NEXT:    vmv1r.v v11, v8
 ; CHECK-NEXT:    vadd.vv v11, v8, v9, v0.t
 ; CHECK-NEXT:    vmv1r.v v9, v8
 ; CHECK-NEXT:    vadd.vv v9, v8, v11, v0.t
diff --git a/llvm/test/CodeGen/RISCV/rvv/vector-splice.ll b/llvm/test/CodeGen/RISCV/rvv/vector-splice.ll
index 6a72043ca7e8e..90d798b167cfc 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vector-splice.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vector-splice.ll
@@ -11,9 +11,9 @@ declare <vscale x 1 x i1> @llvm.vector.splice.nxv1i1(<vscale x 1 x i1>, <vscale
 define <vscale x 1 x i1> @splice_nxv1i1_offset_negone(<vscale x 1 x i1> %a, <vscale x 1 x i1> %b) #0 {
 ; CHECK-LABEL: splice_nxv1i1_offset_negone:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli a0, zero, e8, mf8, ta, ma
 ; CHECK-NEXT:    vmv1r.v v9, v0
 ; CHECK-NEXT:    vmv1r.v v0, v8
-; CHECK-NEXT:    vsetvli a0, zero, e8, mf8, ta, ma
 ; CHECK-NEXT:    vmv.v.i v8, 0
 ; CHECK-NEXT:    csrr a0, vlenb
 ; CHECK-NEXT:    vmerge.vim v10, v8, 1, v0
@@ -33,9 +33,9 @@ define <vscale x 1 x i1> @splice_nxv1i1_offset_negone(<vscale x 1 x i1> %a, <vsc
 define <vscale x 1 x i1> @splice_nxv1i1_offset_max(<vscale x 1 x i1> %a, <vscale x 1 x i1> %b) #0 {
 ; CHECK-LABEL: splice_nxv1i1_offset_max:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli a0, zero, e8, mf8, ta, ma
 ; CHECK-NEXT:    vmv1r.v v9, v0
 ; CHECK-NEXT:    vmv1r.v v0, v8
-; CHECK-NEXT:    vsetvli a0, zero, e8, mf8, ta, ma
 ; CHECK-NEXT:    vmv.v.i v8, 0
 ; CHECK-NEXT:    csrr a0, vlenb
 ; CHECK-NEXT:    vmerge.vim v10, v8, 1, v0
@@ -59,9 +59,9 @@ declare <vscale x 2 x i1> @llvm.vector.splice.nxv2i1(<vscale x 2 x i1>, <vscale
 define <vscale x 2 x i1> @splice_nxv2i1_offset_negone(<vscale x 2 x i1> %a, <vscale x 2 x i1> %b) #0 {
 ; CHECK-LABEL: splice_nxv2i1_offset_negone:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli a0, zero, e8, mf4, ta, ma
 ; CHECK-NEXT:    vmv1r.v v9, v0
 ; CHECK-NEXT:    vmv1r.v v0, v8
-; CHECK-NEXT:    vsetvli a0, zero, e8, mf4, ta, ma
 ; CHECK-NEXT:    vmv.v.i v8, 0
 ; CHECK-NEXT:    csrr a0, vlenb
 ; CHECK-NEXT:    vmerge.vim v10, v8, 1, v0
@@ -81,9 +81,9 @@ define <vscale x 2 x i1> @splice_nxv2i1_offset_negone(<vscale x 2 x i1> %a, <vsc
 define <vscale x 2 x i1> @splice_nxv2i1_offset_max(<vscale x 2 x i1> %a, <vscale x 2 x i1> %b) #0 {
 ; CHECK-LABEL: splice_nxv2i1_offset_max:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli a0, zero, e8, mf4, ta, ma
 ; CHECK-NEXT:    vmv1r.v v9, v0
 ; CHECK-NEXT:    vmv1r.v v0, v8
-; CHECK-NEXT:    vsetvli a0, zero, e8, mf4, ta, ma
 ; CHECK-NEXT:    vmv.v.i v8, 0
 ; CHECK-NEXT:    csrr a0, vlenb
 ; CHECK-NEXT:    vmerge.vim v10, v8, 1, v0
@@ -107,9 +107,9 @@ declare <vscale x 4 x i1> @llvm.vector.splice.nxv4i1(<vscale x 4 x i1>, <vscale
 define <vscale x 4 x i1> @splice_nxv4i1_offset_negone(<vscale x 4 x i1> %a, <vscale x 4 x i1> %b) #0 {
 ; CHECK-LABEL: splice_nxv4i1_offset_negone:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli a0, zero, e8, mf2, ta, ma
 ; CHECK-NEXT:    vmv1r.v v9, v0
 ; CHECK-NEXT:    vmv1r.v v0, v8
-; CHECK-NEXT:    vsetvli a0, zero, e8, mf2, ta, ma
 ; CHECK-NEXT:    vmv.v.i v8, 0
 ; CHECK-NEXT:    csrr a0, vlenb
 ; CHECK-NEXT:    vmerge.vim v10, v8, 1, v0
@@ -129,9 +129,9 @@ define <vscale x 4 x i1> @splice_nxv4i1_offset_negone(<vscale x 4 x i1> %a, <vsc
 define <vscale x 4 x i1> @splice_nxv4i1_offset_max(<vscale x 4 x i1> %a, <vscale x 4 x i1> %b) #0 {
 ; CHECK-LABEL: splice_nxv4i1_offset_max:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli a0, zero, e8, mf2, ta, ma
 ; CHECK-NEXT:    vmv1r.v v9, v0
 ; CHECK-NEXT:    vmv1r.v v0, v8
-; CHECK-NEXT:    vsetvli a0, zero, e8, mf2, ta, ma
 ; CHECK-NEXT:    vmv.v.i v8, 0
 ; CHECK-NEXT:    csrr a0, vlenb
 ; CHECK-NEXT:    vmerge.vim v10, v8, 1, v0
@@ -155,9 +155,9 @@ declare <vscale x 8 x i1> @llvm.vector.splice.nxv8i1(<vscale x 8 x i1>, <vscale
 define <vscale x 8 x i1> @splice_nxv8i1_offset_negone(<vscale x 8 x i1> %a, <vscale x 8 x i1> %b) #0 {
 ; CHECK-LABEL: splice_nxv8i1_offset_negone:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli a0, zero, e8, m1, ta, ma
 ; CHECK-NEXT:    vmv1r.v v9, v0
 ; CHECK-NEXT:    vmv1r.v v0, v8
-; CHECK-NEXT:    vsetvli a0, zero, e8, m1, ta, ma
 ; CHECK-NEXT:    vmv.v.i v8, 0
 ; CHECK-NEXT:    csrr a0, vlenb
 ; CHECK-NEXT:    vmerge.vim v10, v8, 1, v0
@@ -176,9 +176,9 @@ define <vscale x 8 x i1> @splice_nxv8i1_offset_negone(<vscale x 8 x i1> %a, <vsc
 define <vscale x 8 x i1> @splice_nxv8i1_offset_max(<vscale x 8 x i1> %a, <vscale x 8 x i1> %b) #0 {
 ; CHECK-LABEL: splice_nxv8i1_offset_max:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli a0, zero, e8, m1, ta, ma
 ; CHECK-NEXT:    vmv1r.v v9, v0
 ; CHECK-NEXT:    vmv1r.v v0, v8
-; CHECK-NEXT:    vsetvli a0, zero, e8, m1, ta, ma
 ; CHECK-NEXT:    vmv.v.i v8, 0
 ; CHECK-NEXT:    csrr a0, vlenb
 ; CHECK-NEXT:    vmerge.vim v10, v8, 1, v0
@@ -201,9 +201,9 @@ declare <vscale x 16 x i1> @llvm.vector.splice.nxv16i1(<vscale x 16 x i1>, <vsca
 define <vscale x 16 x i1> @splice_nxv16i1_offset_negone(<vscale x 16 x i1> %a, <vscale x 16 x i1> %b) #0 {
 ; CHECK-LABEL: splice_nxv16i1_offset_negone:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli a0, zero, e8, m2, ta, ma
 ; CHECK-NEXT:    vmv1r.v v9, v0
 ; CHECK-NEXT:    vmv1r.v v0, v8
-; CHECK-NEXT:    vsetvli a0, zero, e8, m2, ta, ma
 ; CHECK-NEXT:    vmv.v.i v10, 0
 ; CHECK-NEXT:    csrr a0, vlenb
 ; CHECK-NEXT:    vmerge.vim v12, v10, 1, v0
@@ -223,9 +223,9 @@ define <vscale x 16 x i1> @splice_nxv16i1_offset_negone(<vscale x 16 x i1> %a, <
 define <vscale x 16 x i1> @splice_nxv16i1_offset_max(<vscale x 16 x i1> %a, <vscale x 16 x i1> %b) #0 {
 ; CHECK-LABEL: splice_nxv16i1_offset_max:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli a0, zero, e8, m2, ta, ma
 ; CHECK-NEXT:    vmv1r.v v9, v0
 ; CHECK-NEXT:    vmv1r.v v0, v8
-; CHECK-NEXT:    vsetvli a0, zero, e8, m2, ta, ma
 ; CHECK-NEXT:    vmv.v.i v10, 0
 ; CHECK-NEXT:    csrr a0, vlenb
 ; CHECK-NEXT:    vmerge.vim v12, v10, 1, v0
@@ -249,9 +249,9 @@ declare <vscale x 32 x i1> @llvm.vector.splice.nxv32i1(<vscale x 32 x i1>, <vsca
 define <vscale x 32 x i1> @splice_nxv32i1_offset_negone(<vscale x 32 x i1> %a, <vscale x 32 x i1> %b) #0 {
 ; CHECK-LABEL: splice_nxv32i1_offset_negone:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli a0, zero, e8, m4, ta, ma
 ; CHECK-NEXT:    vmv1r.v v9, v0
 ; CHECK-NEXT:    vmv1r.v v0, v8
-; CHECK-NEXT:    vsetvli a0, zero, e8, m4, ta, ma
 ; CHECK-NEXT:    vmv.v.i v12, 0
 ; CHECK-NEXT:    csrr a0, vlenb
 ; CHECK-NEXT:    vmerge.vim v16, v12, 1, v0
@@ -296,9 +296,9 @@ declare <vscale x 64 x i1> @llvm.vector.splice.nxv64i1(<vscale x 64 x i1>, <vsca
 define <vscale x 64 x i1> @splice_nxv64i1_offset_negone(<vscale x 64 x i1> %a, <vscale x 64 x i1> %b) #0 {
 ; CHECK-LABEL: splice_nxv64i1_offset_negone:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli a0, zero, e8, m8, ta, ma
 ; CHECK-NEXT:    vmv1r.v v9, v0
 ; CHECK-NEXT:    vmv1r.v v0, v8
-; CHECK-NEXT:    vsetvli a0, zero, e8, m8, ta, ma
 ; CHECK-NEXT:    vmv.v.i v24, 0
 ; CHECK-NEXT:    csrr a0, vlenb
 ; CHECK-NEXT:    vmerge.vim v16, v24, 1, v0
diff --git a/llvm/test/CodeGen/RISCV/rvv/vfabs-vp.ll b/llvm/test/CodeGen/RISCV/rvv/vfabs-vp.ll
index 2c92a5da8eecb..8f9f9c4256c8f 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vfabs-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vfabs-vp.ll
@@ -462,11 +462,11 @@ declare <vscale x 16 x double> @llvm.vp.fabs.nxv16f64(<vscale x 16 x double>, <v
 define <vscale x 16 x double> @vfabs_vv_nxv16f64(<vscale x 16 x double> %va, <vscale x 16 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vfabs_vv_nxv16f64:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli a1, zero, e8, mf4, ta, ma
 ; CHECK-NEXT:    vmv1r.v v24, v0
 ; CHECK-NEXT:    csrr a1, vlenb
 ; CHECK-NEXT:    srli a2, a1, 3
 ; CHECK-NEXT:    sub a3, a0, a1
-; CHECK-NEXT:    vsetvli a4, zero, e8, mf4, ta, ma
 ; CHECK-NEXT:    vslidedown.vx v0, v0, a2
 ; CHECK-NEXT:    sltu a2, a0, a3
 ; CHECK-NEXT:    addi a2, a2, -1
diff --git a/llvm/test/CodeGen/RISCV/rvv/vfadd-vp.ll b/llvm/test/CodeGen/RISCV/rvv/vfadd-vp.ll
index 1953cfd2a0169..87bc9f27d6dc9 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vfadd-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vfadd-vp.ll
@@ -411,11 +411,11 @@ define <vscale x 32 x bfloat> @vfadd_vv_nxv32bf16(<vscale x 32 x bfloat> %va, <v
 ; CHECK-NEXT:    slli a1, a1, 3
 ; CHECK-NEXT:    sub sp, sp, a1
 ; CHECK-NEXT:    .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x08, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 8 * vlenb
+; CHECK-NEXT:    vsetvli a1, zero, e16, m4, ta, ma
 ; CHECK-NEXT:    vmv1r.v v7, v0
 ; CHECK-NEXT:    csrr a2, vlenb
 ; CHECK-NEXT:    addi a1, sp, 16
 ; CHECK-NEXT:    vs8r.v v16, (a1) # Unknown-size Folded Spill
-; CHECK-NEXT:    vsetvli a1, zero, e16, m4, ta, ma
 ; CHECK-NEXT:    vfwcvtbf16.f.f.v v24, v20
 ; CHECK-NEXT:    slli a1, a2, 1
 ; CHECK-NEXT:    srli a2, a2, 2
@@ -518,12 +518,12 @@ define <vscale x 32 x bfloat> @vfadd_vf_nxv32bf16(<vscale x 32 x bfloat> %va, bf
 ; CHECK-NEXT:    add a1, a2, a1
 ; CHECK-NEXT:    sub sp, sp, a1
 ; CHECK-NEXT:    .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x11, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 17 * vlenb
+; CHECK-NEXT:    vsetvli a1, zero, e16, m4, ta, ma
 ; CHECK-NEXT:    vmv8r.v v16, v8
 ; CHECK-NEXT:    fmv.x.h a1, fa0
 ; CHECK-NEXT:    csrr a2, vlenb
 ; CHECK-NEXT:    addi a3, sp, 16
 ; CHECK-NEXT:    vs8r.v v8, (a3) # Unknown-size Folded Spill
-; CHECK-NEXT:    vsetvli a3, zero, e16, m4, ta, ma
 ; CHECK-NEXT:    vfwcvtbf16.f.f.v v8, v20
 ; CHECK-NEXT:    vsetvli a3, zero, e16, m8, ta, ma
 ; CHECK-NEXT:    vmv.v.x v16, a1
@@ -604,10 +604,10 @@ define <vscale x 32 x bfloat> @vfadd_vf_nxv32bf16_unmasked(<vscale x 32 x bfloat
 ; CHECK-NEXT:    slli a1, a1, 4
 ; CHECK-NEXT:    sub sp, sp, a1
 ; CHECK-NEXT:    .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x10, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 16 * vlenb
+; CHECK-NEXT:    vsetvli a1, zero, e8, m4, ta, ma
 ; CHECK-NEXT:    vmv8r.v v16, v8
 ; CHECK-NEXT:    fmv.x.h a1, fa0
 ; CHECK-NEXT:    csrr a2, vlenb
-; CHECK-NEXT:    vsetvli a3, zero, e8, m4, ta, ma
 ; CHECK-NEXT:    vmset.m v7
 ; CHECK-NEXT:    addi a3, sp, 16
 ; CHECK-NEXT:    vs8r.v v8, (a3) # Unknown-size Folded Spill
@@ -1205,11 +1205,11 @@ define <vscale x 32 x half> @vfadd_vv_nxv32f16(<vscale x 32 x half> %va, <vscale
 ; ZVFHMIN-NEXT:    slli a1, a1, 3
 ; ZVFHMIN-NEXT:    sub sp, sp, a1
 ; ZVFHMIN-NEXT:    .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x08, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 8 * vlenb
+; ZVFHMIN-NEXT:    vsetvli a1, zero, e16, m4, ta, ma
 ; ZVFHMIN-NEXT:    vmv1r.v v7, v0
 ; ZVFHMIN-NEXT:    csrr a2, vlenb
 ; ZVFHMIN-NEXT:    addi a1, sp, 16
 ; ZVFHMIN-NEXT:    vs8r.v v16, (a1) # Unknown-size Folded Spill
-; ZVFHMIN-NEXT:    vsetvli a1, zero, e16, m4, ta, ma
 ; ZVFHMIN-NEXT:    vfwcvt.f.f.v v24, v20
 ; ZVFHMIN-NEXT:    slli a1, a2, 1
 ; ZVFHMIN-NEXT:    srli a2, a2, 2
@@ -1324,12 +1324,12 @@ define <vscale x 32 x half> @vfadd_vf_nxv32f16(<vscale x 32 x half> %va, half %b
 ; ZVFHMIN-NEXT:    add a1, a2, a1
 ; ZVFHMIN-NEXT:    sub sp, sp, a1
 ; ZVFHMIN-NEXT:    .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x11, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 17 * vlenb
+; ZVFHMIN-NEXT:    vsetvli a1, zero, e16, m4, ta, ma
 ; ZVFHMIN-NEXT:    vmv8r.v v16, v8
 ; ZVFHMIN-NEXT:    fmv.x.h a1, fa0
 ; ZVFHMIN-NEXT:    csrr a2, vlenb
 ; ZVFHMIN-NEXT:    addi a3, sp, 16
 ; ZVFHMIN-NEXT:    vs8r.v v8, (a3) # Unknown-size Folded Spill
-; ZVFHMIN-NEXT:    vsetvli a3, zero, e16, m4, ta, ma
 ; ZVFHMIN-NEXT:    vfwcvt.f.f.v v8, v20
 ; ZVFHMIN-NEXT:    vsetvli a3, zero, e16, m8, ta, ma
 ; ZVFHMIN-NEXT:    vmv.v.x v16, a1
@@ -1416,10 +1416,10 @@ define <vscale x 32 x half> @vfadd_vf_nxv32f16_unmasked(<vscale x 32 x half> %va
 ; ZVFHMIN-NEXT:    slli a1, a1, 4
 ; ZVFHMIN-NEXT:    sub sp, sp, a1
 ; ZVFHMIN-NEXT:    .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x10, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 16 * vlenb
+; ZVFHMIN-NEXT:    vsetvli a1, zero, e8, m4, ta, ma
 ; ZVFHMIN-NEXT:    vmv8r.v v16, v8
 ; ZVFHMIN-NEXT:    fmv.x.h a1, fa0
 ; ZVFHMIN-NEXT:    csrr a2, vlenb
-; ZVFHMIN-NEXT:    vsetvli a3, zero, e8, m4, ta, ma
 ; ZVFHMIN-NEXT:    vmset.m v7
 ; ZVFHMIN-NEXT:    addi a3, sp, 16
 ; ZVFHMIN-NEXT:    vs8r.v v8, (a3) # Unknown-size Folded Spill
diff --git a/llvm/test/CodeGen/RISCV/rvv/vfdiv-vp.ll b/llvm/test/CodeGen/RISCV/rvv/vfdiv-vp.ll
index ccd286b7ee5fd..061af454aa8ba 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vfdiv-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vfdiv-vp.ll
@@ -373,11 +373,11 @@ define <vscale x 32 x bfloat> @vfdiv_vv_nxv32bf16(<vscale x 32 x bfloat> %va, <v
 ; CHECK-NEXT:    slli a1, a1, 3
 ; CHECK-NEXT:    sub sp, sp, a1
 ; CHECK-NEXT:    .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x08, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 8 * vlenb
+; CHECK-NEXT:    vsetvli a1, zero, e16, m4, ta, ma
 ; CHECK-NEXT:    vmv1r.v v7, v0
 ; CHECK-NEXT:    csrr a2, vlenb
 ; CHECK-NEXT:    addi a1, sp, 16
 ; CHECK-NEXT:    vs8r.v v16, (a1) # Unknown-size Folded Spill
-; CHECK-NEXT:    vsetvli a1, zero, e16, m4, ta, ma
 ; CHECK-NEXT:    vfwcvtbf16.f.f.v v24, v20
 ; CHECK-NEXT:    slli a1, a2, 1
 ; CHECK-NEXT:    srli a2, a2, 2
@@ -480,12 +480,12 @@ define <vscale x 32 x bfloat> @vfdiv_vf_nxv32bf16(<vscale x 32 x bfloat> %va, bf
 ; CHECK-NEXT:    add a1, a2, a1
 ; CHECK-NEXT:    sub sp, sp, a1
 ; CHECK-NEXT:    .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x11, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 17 * vlenb
+; CHECK-NEXT:    vsetvli a1, zero, e16, m4, ta, ma
 ; CHECK-NEXT:    vmv8r.v v16, v8
 ; CHECK-NEXT:    fmv.x.h a1, fa0
 ; CHECK-NEXT:    csrr a2, vlenb
 ; CHECK-NEXT:    addi a3, sp, 16
 ; CHECK-NEXT:    vs8r.v v8, (a3) # Unknown-size Folded Spill
-; CHECK-NEXT:    vsetvli a3, zero, e16, m4, ta, ma
 ; CHECK-NEXT:    vfwcvtbf16.f.f.v v8, v20
 ; CHECK-NEXT:    vsetvli a3, zero, e16, m8, ta, ma
 ; CHECK-NEXT:    vmv.v.x v16, a1
@@ -566,10 +566,10 @@ define <vscale x 32 x bfloat> @vfdiv_vf_nxv32bf16_unmasked(<vscale x 32 x bfloat
 ; CHECK-NEXT:    slli a1, a1, 4
 ; CHECK-NEXT:    sub sp, sp, a1
 ; CHECK-NEXT:    .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x10, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 16 * vlenb
+; CHECK-NEXT:    vsetvli a1, zero, e8, m4, ta, ma
 ; CHECK-NEXT:    vmv8r.v v16, v8
 ; CHECK-NEXT:    fmv.x.h a1, fa0
 ; CHECK-NEXT:    csrr a2, vlenb
-; CHECK-NEXT:    vsetvli a3, zero, e8, m4, ta, ma
 ; CHECK-NEXT:    vmset.m v7
 ; CHECK-NEXT:    addi a3, sp, 16
 ; CHECK-NEXT:    vs8r.v v8, (a3) # Unknown-size Folded Spill
@@ -1117,11 +1117,11 @@ define <vscale x 32 x half> @vfdiv_vv_nxv32f16(<vscale x 32 x half> %va, <vscale
 ; ZVFHMIN-NEXT:    slli a1, a1, 3
 ; ZVFHMIN-NEXT:    sub sp, sp, a1
 ; ZVFHMIN-NEXT:    .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x08, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 8 * vlenb
+; ZVFHMIN-NEXT:    vsetvli a1, zero, e16, m4, ta, ma
 ; ZVFHMIN-NEXT:    vmv1r.v v7, v0
 ; ZVFHMIN-NEXT:    csrr a2, vlenb
 ; ZVFHMIN-NEXT:    addi a1, sp, 16
 ; ZVFHMIN-NEXT:    vs8r.v v16, (a1) # Unknown-size Folded Spill
-; ZVFHMIN-NEXT:    vsetvli a1, zero, e16, m4, ta, ma
 ; ZVFHMIN-NEXT:    vfwcvt.f.f.v v24, v20
 ; ZVFHMIN-NEXT:    slli a1, a2, 1
 ; ZVFHMIN-NEXT:    srli a2, a2, 2
@@ -1236,12 +1236,12 @@ define <vscale x 32 x half> @vfdiv_vf_nxv32f16(<vscale x 32 x half> %va, half %b
 ; ZVFHMIN-NEXT:    add a1, a2, a1
 ; ZVFHMIN-NEXT:    sub sp, sp, a1
 ; ZVFHMIN-NEXT:    .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x11, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 17 * vlenb
+; ZVFHMIN-NEXT:    vsetvli a1, zero, e16, m4, ta, ma
 ; ZVFHMIN-NEXT:    vmv8r.v v16, v8
 ; ZVFHMIN-NEXT:    fmv.x.h a1, fa0
 ; ZVFHMIN-NEXT:    csrr a2, vlenb
 ; ZVFHMIN-NEXT:    addi a3, sp, 16
 ; ZVFHMIN-NEXT:    vs8r.v v8, (a3) # Unknown-size Folded Spill
-; ZVFHMIN-NEXT:    vsetvli a3, zero, e16, m4, ta, ma
 ; ZVFHMIN-NEXT:    vfwcvt.f.f.v v8, v20
 ; ZVFHMIN-NEXT:    vsetvli a3, zero, e16, m8, ta, ma
 ; ZVFHMIN-NEXT:    vmv.v.x v16, a1
@@ -1328,10 +1328,10 @@ define <vscale x 32 x half> @vfdiv_vf_nxv32f16_unmasked(<vscale x 32 x half> %va
 ; ZVFHMIN-NEXT:    slli a1, a1, 4
 ; ZVFHMIN-NEXT:    sub sp, sp, a1
 ; ZVFHMIN-NEXT:    .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x10, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 16 * vlenb
+; ZVFHMIN-NEXT:    vsetvli a1, zero, e8, m4, ta, ma
 ; ZVFHMIN-NEXT:    vmv8r.v v16, v8
 ; ZVFHMIN-NEXT:    fmv.x.h a1, fa0
 ; ZVFHMIN-NEXT:    csrr a2, vlenb
-; ZVFHMIN-NEXT:    vsetvli a3, zero, e8, m4, ta, ma
 ; ZVFHMIN-NEXT:    vmset.m v7
 ; ZVFHMIN-NEXT:    addi a3, sp, 16
 ; ZVFHMIN-NEXT:    vs8r.v v8, (a3) # Unknown-size Folded Spill
diff --git a/llvm/test/CodeGen/RISCV/rvv/vfirst.ll b/llvm/test/CodeGen/RISCV/rvv/vfirst.ll
index eafd605c6110e..c510121ee3ebe 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vfirst.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vfirst.ll
@@ -43,9 +43,9 @@ declare iXLen @llvm.riscv.vfirst.mask.iXLen.nxv1i1(
 define iXLen @intrinsic_vfirst_mask_m_nxv1i1(<vscale x 1 x i1> %0, <vscale x 1 x i1> %1, iXLen %2) nounwind {
 ; CHECK-LABEL: intrinsic_vfirst_mask_m_nxv1i1:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e8, mf8, ta, ma
 ; CHECK-NEXT:    vmv1r.v v9, v0
 ; CHECK-NEXT:    vmv1r.v v0, v8
-; CHECK-NEXT:    vsetvli zero, a0, e8, mf8, ta, ma
 ; CHECK-NEXT:    vfirst.m a0, v9, v0.t
 ; CHECK-NEXT:    ret
 entry:
@@ -97,9 +97,9 @@ declare iXLen @llvm.riscv.vfirst.mask.iXLen.nxv2i1(
 define iXLen @intrinsic_vfirst_mask_m_nxv2i1(<vscale x 2 x i1> %0, <vscale x 2 x i1> %1, iXLen %2) nounwind {
 ; CHECK-LABEL: intrinsic_vfirst_mask_m_nxv2i1:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e8, mf4, ta, ma
 ; CHECK-NEXT:    vmv1r.v v9, v0
 ; CHECK-NEXT:    vmv1r.v v0, v8
-; CHECK-NEXT:    vsetvli zero, a0, e8, mf4, ta, ma
 ; CHECK-NEXT:    vfirst.m a0, v9, v0.t
 ; CHECK-NEXT:    ret
 entry:
@@ -137,9 +137,9 @@ declare iXLen @llvm.riscv.vfirst.mask.iXLen.nxv4i1(
 define iXLen @intrinsic_vfirst_mask_m_nxv4i1(<vscale x 4 x i1> %0, <vscale x 4 x i1> %1, iXLen %2) nounwind {
 ; CHECK-LABEL: intrinsic_vfirst_mask_m_nxv4i1:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e8, mf2, ta, ma
 ; CHECK-NEXT:    vmv1r.v v9, v0
 ; CHECK-NEXT:    vmv1r.v v0, v8
-; CHECK-NEXT:    vsetvli zero, a0, e8, mf2, ta, ma
 ; CHECK-NEXT:    vfirst.m a0, v9, v0.t
 ; CHECK-NEXT:    ret
 entry:
@@ -177,9 +177,9 @@ declare iXLen @llvm.riscv.vfirst.mask.iXLen.nxv8i1(
 define iXLen @intrinsic_vfirst_mask_m_nxv8i1(<vscale x 8 x i1> %0, <vscale x 8 x i1> %1, iXLen %2) nounwind {
 ; CHECK-LABEL: intrinsic_vfirst_mask_m_nxv8i1:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e8, m1, ta, ma
 ; CHECK-NEXT:    vmv1r.v v9, v0
 ; CHECK-NEXT:    vmv1r.v v0, v8
-; CHECK-NEXT:    vsetvli zero, a0, e8, m1, ta, ma
 ; CHECK-NEXT:    vfirst.m a0, v9, v0.t
 ; CHECK-NEXT:    ret
 entry:
@@ -217,9 +217,9 @@ declare iXLen @llvm.riscv.vfirst.mask.iXLen.nxv16i1(
 define iXLen @intrinsic_vfirst_mask_m_nxv16i1(<vscale x 16 x i1> %0, <vscale x 16 x i1> %1, iXLen %2) nounwind {
 ; CHECK-LABEL: intrinsic_vfirst_mask_m_nxv16i1:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e8, m2, ta, ma
 ; CHECK-NEXT:    vmv1r.v v9, v0
 ; CHECK-NEXT:    vmv1r.v v0, v8
-; CHECK-NEXT:    vsetvli zero, a0, e8, m2, ta, ma
 ; CHECK-NEXT:    vfirst.m a0, v9, v0.t
 ; CHECK-NEXT:    ret
 entry:
@@ -257,9 +257,9 @@ declare iXLen @llvm.riscv.vfirst.mask.iXLen.nxv32i1(
 define iXLen @intrinsic_vfirst_mask_m_nxv32i1(<vscale x 32 x i1> %0, <vscale x 32 x i1> %1, iXLen %2) nounwind {
 ; CHECK-LABEL: intrinsic_vfirst_mask_m_nxv32i1:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e8, m4, ta, ma
 ; CHECK-NEXT:    vmv1r.v v9, v0
 ; CHECK-NEXT:    vmv1r.v v0, v8
-; CHECK-NEXT:    vsetvli zero, a0, e8, m4, ta, ma
 ; CHECK-NEXT:    vfirst.m a0, v9, v0.t
 ; CHECK-NEXT:    ret
 entry:
@@ -297,9 +297,9 @@ declare iXLen @llvm.riscv.vfirst.mask.iXLen.nxv64i1(
 define iXLen @intrinsic_vfirst_mask_m_nxv64i1(<vscale x 64 x i1> %0, <vscale x 64 x i1> %1, iXLen %2) nounwind {
 ; CHECK-LABEL: intrinsic_vfirst_mask_m_nxv64i1:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e8, m8, ta, ma
 ; CHECK-NEXT:    vmv1r.v v9, v0
 ; CHECK-NEXT:    vmv1r.v v0, v8
-; CHECK-NEXT:    vsetvli zero, a0, e8, m8, ta, ma
 ; CHECK-NEXT:    vfirst.m a0, v9, v0.t
 ; CHECK-NEXT:    ret
 entry:
diff --git a/llvm/test/CodeGen/RISCV/rvv/vfma-vp.ll b/llvm/test/CodeGen/RISCV/rvv/vfma-vp.ll
index fd518d9be786d..7ca1983e8b32c 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vfma-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vfma-vp.ll
@@ -628,6 +628,7 @@ define <vscale x 32 x bfloat> @vfma_vv_nxv32bf16(<vscale x 32 x bfloat> %va, <vs
 ; CHECK-NEXT:    add a2, a2, a3
 ; CHECK-NEXT:    sub sp, sp, a2
 ; CHECK-NEXT:    .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x29, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 41 * vlenb
+; CHECK-NEXT:    vsetvli a2, zero, e8, mf2, ta, ma
 ; CHECK-NEXT:    vmv1r.v v24, v0
 ; CHECK-NEXT:    vl8re16.v v0, (a0)
 ; CHECK-NEXT:    csrr a2, vlenb
@@ -639,7 +640,6 @@ define <vscale x 32 x bfloat> @vfma_vv_nxv32bf16(<vscale x 32 x bfloat> %va, <vs
 ; CHECK-NEXT:    add a4, sp, a4
 ; CHECK-NEXT:    addi a4, a4, 16
 ; CHECK-NEXT:    vs1r.v v24, (a4) # Unknown-size Folded Spill
-; CHECK-NEXT:    vsetvli a4, zero, e8, mf2, ta, ma
 ; CHECK-NEXT:    vslidedown.vx v24, v24, a2
 ; CHECK-NEXT:    csrr a2, vlenb
 ; CHECK-NEXT:    slli a2, a2, 3
@@ -2193,6 +2193,7 @@ define <vscale x 32 x half> @vfma_vv_nxv32f16(<vscale x 32 x half> %va, <vscale
 ; ZVFHMIN-NEXT:    add a2, a2, a3
 ; ZVFHMIN-NEXT:    sub sp, sp, a2
 ; ZVFHMIN-NEXT:    .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x29, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 41 * vlenb
+; ZVFHMIN-NEXT:    vsetvli a2, zero, e8, mf2, ta, ma
 ; ZVFHMIN-NEXT:    vmv1r.v v24, v0
 ; ZVFHMIN-NEXT:    vl8re16.v v0, (a0)
 ; ZVFHMIN-NEXT:    csrr a2, vlenb
@@ -2204,7 +2205,6 @@ define <vscale x 32 x half> @vfma_vv_nxv32f16(<vscale x 32 x half> %va, <vscale
 ; ZVFHMIN-NEXT:    add a4, sp, a4
 ; ZVFHMIN-NEXT:    addi a4, a4, 16
 ; ZVFHMIN-NEXT:    vs1r.v v24, (a4) # Unknown-size Folded Spill
-; ZVFHMIN-NEXT:    vsetvli a4, zero, e8, mf2, ta, ma
 ; ZVFHMIN-NEXT:    vslidedown.vx v24, v24, a2
 ; ZVFHMIN-NEXT:    csrr a2, vlenb
 ; ZVFHMIN-NEXT:    slli a2, a2, 3
@@ -3662,6 +3662,7 @@ define <vscale x 16 x double> @vfma_vv_nxv16f64(<vscale x 16 x double> %va, <vsc
 ; CHECK-NEXT:    add a1, a1, a3
 ; CHECK-NEXT:    sub sp, sp, a1
 ; CHECK-NEXT:    .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x30, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 48 * vlenb
+; CHECK-NEXT:    vsetvli a1, zero, e8, mf4, ta, ma
 ; CHECK-NEXT:    vmv1r.v v7, v0
 ; CHECK-NEXT:    csrr a1, vlenb
 ; CHECK-NEXT:    slli a1, a1, 3
@@ -3708,7 +3709,6 @@ define <vscale x 16 x double> @vfma_vv_nxv16f64(<vscale x 16 x double> %va, <vsc
 ; CHECK-NEXT:    vl8re64.v v8, (a0)
 ; CHECK-NEXT:    addi a0, sp, 16
 ; CHECK-NEXT:    vs8r.v v8, (a0) # Unknown-size Folded Spill
-; CHECK-NEXT:    vsetvli a0, zero, e8, mf4, ta, ma
 ; CHECK-NEXT:    vslidedown.vx v0, v0, a3
 ; CHECK-NEXT:    and a0, a7, a6
 ; CHECK-NEXT:    csrr a2, vlenb
@@ -7796,9 +7796,9 @@ define <vscale x 16 x half> @vfnmadd_vv_nxv16f16(<vscale x 16 x half> %va, <vsca
 ;
 ; ZVFHMIN-LABEL: vfnmadd_vv_nxv16f16:
 ; ZVFHMIN:       # %bb.0:
+; ZVFHMIN-NEXT:    vsetvli zero, a0, e16, m4, ta, ma
 ; ZVFHMIN-NEXT:    vmv4r.v v4, v8
 ; ZVFHMIN-NEXT:    lui a1, 8
-; ZVFHMIN-NEXT:    vsetvli zero, a0, e16, m4, ta, ma
 ; ZVFHMIN-NEXT:    vxor.vx v12, v12, a1, v0.t
 ; ZVFHMIN-NEXT:    vxor.vx v16, v16, a1, v0.t
 ; ZVFHMIN-NEXT:    vsetvli a1, zero, e16, m4, ta, ma
@@ -8152,9 +8152,9 @@ define <vscale x 16 x half> @vfnmadd_vf_nxv16f16_neg_splat_commute(<vscale x 16
 ;
 ; ZVFHMIN-LABEL: vfnmadd_vf_nxv16f16_neg_splat_commute:
 ; ZVFHMIN:       # %bb.0:
+; ZVFHMIN-NEXT:    vsetvli a1, zero, e16, m4, ta, ma
 ; ZVFHMIN-NEXT:    vmv4r.v v4, v8
 ; ZVFHMIN-NEXT:    fmv.x.h a1, fa0
-; ZVFHMIN-NEXT:    vsetvli a2, zero, e16, m4, ta, ma
 ; ZVFHMIN-NEXT:    vmv.v.x v16, a1
 ; ZVFHMIN-NEXT:    lui a1, 8
 ; ZVFHMIN-NEXT:    vsetvli zero, a0, e16, m4, ta, ma
@@ -8253,9 +8253,9 @@ define <vscale x 16 x half> @vfnmsub_vv_nxv16f16(<vscale x 16 x half> %va, <vsca
 ;
 ; ZVFHMIN-LABEL: vfnmsub_vv_nxv16f16:
 ; ZVFHMIN:       # %bb.0:
+; ZVFHMIN-NEXT:    vsetvli zero, a0, e16, m4, ta, ma
 ; ZVFHMIN-NEXT:    vmv4r.v v4, v8
 ; ZVFHMIN-NEXT:    lui a1, 8
-; ZVFHMIN-NEXT:    vsetvli zero, a0, e16, m4, ta, ma
 ; ZVFHMIN-NEXT:    vxor.vx v12, v12, a1, v0.t
 ; ZVFHMIN-NEXT:    vxor.vx v16, v16, a1, v0.t
 ; ZVFHMIN-NEXT:    vsetvli a1, zero, e16, m4, ta, ma
@@ -8553,9 +8553,9 @@ define <vscale x 16 x half> @vfnmsub_vf_nxv16f16_neg_splat(<vscale x 16 x half>
 ;
 ; ZVFHMIN-LABEL: vfnmsub_vf_nxv16f16_neg_splat:
 ; ZVFHMIN:       # %bb.0:
+; ZVFHMIN-NEXT:    vsetvli a1, zero, e16, m4, ta, ma
 ; ZVFHMIN-NEXT:    vmv4r.v v4, v8
 ; ZVFHMIN-NEXT:    fmv.x.h a1, fa0
-; ZVFHMIN-NEXT:    vsetvli a2, zero, e16, m4, ta, ma
 ; ZVFHMIN-NEXT:    vmv.v.x v16, a1
 ; ZVFHMIN-NEXT:    lui a1, 8
 ; ZVFHMIN-NEXT:    vfwcvt.f.f.v v24, v12
@@ -8712,11 +8712,11 @@ define <vscale x 32 x half> @vfmsub_vv_nxv32f16(<vscale x 32 x half> %va, <vscal
 ; ZVFHMIN-NEXT:    add a2, sp, a2
 ; ZVFHMIN-NEXT:    addi a2, a2, 16
 ; ZVFHMIN-NEXT:    vs8r.v v16, (a2) # Unknown-size Folded Spill
+; ZVFHMIN-NEXT:    vsetvli zero, a1, e16, m8, ta, ma
 ; ZVFHMIN-NEXT:    vmv8r.v v24, v8
 ; ZVFHMIN-NEXT:    vl8re16.v v8, (a0)
 ; ZVFHMIN-NEXT:    lui a2, 8
 ; ZVFHMIN-NEXT:    csrr a0, vlenb
-; ZVFHMIN-NEXT:    vsetvli zero, a1, e16, m8, ta, ma
 ; ZVFHMIN-NEXT:    vxor.vx v8, v8, a2, v0.t
 ; ZVFHMIN-NEXT:    slli a2, a0, 1
 ; ZVFHMIN-NEXT:    mv a3, a1
@@ -9277,10 +9277,10 @@ define <vscale x 32 x half> @vfmsub_vf_nxv32f16_unmasked(<vscale x 32 x half> %v
 ; ZVFHMIN-NEXT:    add a1, a1, a2
 ; ZVFHMIN-NEXT:    sub sp, sp, a1
 ; ZVFHMIN-NEXT:    .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x28, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 40 * vlenb
+; ZVFHMIN-NEXT:    vsetvli a1, zero, e8, m4, ta, ma
 ; ZVFHMIN-NEXT:    vmv8r.v v24, v16
 ; ZVFHMIN-NEXT:    fmv.x.h a2, fa0
 ; ZVFHMIN-NEXT:    lui a1, 8
-; ZVFHMIN-NEXT:    vsetvli a3, zero, e8, m4, ta, ma
 ; ZVFHMIN-NEXT:    vmset.m v7
 ; ZVFHMIN-NEXT:    csrr a3, vlenb
 ; ZVFHMIN-NEXT:    csrr a4, vlenb
@@ -10001,11 +10001,11 @@ define <vscale x 32 x half> @vfnmadd_vv_nxv32f16_unmasked_commuted(<vscale x 32
 ; ZVFHMIN-NEXT:    add a2, a2, a3
 ; ZVFHMIN-NEXT:    sub sp, sp, a2
 ; ZVFHMIN-NEXT:    .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x29, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 41 * vlenb
+; ZVFHMIN-NEXT:    vsetvli a2, zero, e8, m4, ta, ma
 ; ZVFHMIN-NEXT:    vmv8r.v v0, v16
 ; ZVFHMIN-NEXT:    vmv8r.v v16, v8
 ; ZVFHMIN-NEXT:    vl8re16.v v24, (a0)
 ; ZVFHMIN-NEXT:    lui a2, 8
-; ZVFHMIN-NEXT:    vsetvli a0, zero, e8, m4, ta, ma
 ; ZVFHMIN-NEXT:    vmset.m v8
 ; ZVFHMIN-NEXT:    csrr a3, vlenb
 ; ZVFHMIN-NEXT:    vsetvli zero, a1, e16, m8, ta, ma
@@ -11154,11 +11154,11 @@ define <vscale x 32 x half> @vfnmadd_vf_nxv32f16_neg_splat_unmasked_commute(<vsc
 ; ZVFHMIN-NEXT:    add a1, a1, a2
 ; ZVFHMIN-NEXT:    sub sp, sp, a1
 ; ZVFHMIN-NEXT:    .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x29, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 41 * vlenb
+; ZVFHMIN-NEXT:    vsetvli a1, zero, e16, m8, ta, ma
 ; ZVFHMIN-NEXT:    vmv8r.v v0, v16
 ; ZVFHMIN-NEXT:    vmv8r.v v16, v8
 ; ZVFHMIN-NEXT:    fmv.x.h a1, fa0
 ; ZVFHMIN-NEXT:    lui a2, 8
-; ZVFHMIN-NEXT:    vsetvli a3, zero, e16, m8, ta, ma
 ; ZVFHMIN-NEXT:    vmset.m v8
 ; ZVFHMIN-NEXT:    csrr a3, vlenb
 ; ZVFHMIN-NEXT:    vmv.v.x v24, a1
@@ -11766,11 +11766,11 @@ define <vscale x 32 x half> @vfnmsub_vv_nxv32f16_unmasked_commuted(<vscale x 32
 ; ZVFHMIN-NEXT:    add a2, a2, a3
 ; ZVFHMIN-NEXT:    sub sp, sp, a2
 ; ZVFHMIN-NEXT:    .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x29, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 41 * vlenb
+; ZVFHMIN-NEXT:    vsetvli a2, zero, e8, m4, ta, ma
 ; ZVFHMIN-NEXT:    vmv8r.v v0, v16
 ; ZVFHMIN-NEXT:    vmv8r.v v16, v8
 ; ZVFHMIN-NEXT:    vl8re16.v v24, (a0)
 ; ZVFHMIN-NEXT:    lui a2, 8
-; ZVFHMIN-NEXT:    vsetvli a0, zero, e8, m4, ta, ma
 ; ZVFHMIN-NEXT:    vmset.m v8
 ; ZVFHMIN-NEXT:    csrr a3, vlenb
 ; ZVFHMIN-NEXT:    vsetvli zero, a1, e16, m8, ta, ma
@@ -11924,11 +11924,11 @@ define <vscale x 32 x half> @vfnmsub_vf_nxv32f16(<vscale x 32 x half> %va, half
 ; ZVFHMIN-NEXT:    add a1, a1, a2
 ; ZVFHMIN-NEXT:    sub sp, sp, a1
 ; ZVFHMIN-NEXT:    .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x28, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 40 * vlenb
+; ZVFHMIN-NEXT:    vsetvli a1, zero, e16, m8, ta, ma
 ; ZVFHMIN-NEXT:    vmv8r.v v24, v16
 ; ZVFHMIN-NEXT:    fmv.x.h a2, fa0
 ; ZVFHMIN-NEXT:    lui a3, 8
 ; ZVFHMIN-NEXT:    csrr a1, vlenb
-; ZVFHMIN-NEXT:    vsetvli a4, zero, e16, m8, ta, ma
 ; ZVFHMIN-NEXT:    vmv.v.x v16, a2
 ; ZVFHMIN-NEXT:    csrr a2, vlenb
 ; ZVFHMIN-NEXT:    slli a2, a2, 5
@@ -12075,11 +12075,11 @@ define <vscale x 32 x half> @vfnmsub_vf_nxv32f16_commute(<vscale x 32 x half> %v
 ; ZVFHMIN-NEXT:    slli a1, a1, 5
 ; ZVFHMIN-NEXT:    sub sp, sp, a1
 ; ZVFHMIN-NEXT:    .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x20, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 32 * vlenb
+; ZVFHMIN-NEXT:    vsetvli a1, zero, e16, m8, ta, ma
 ; ZVFHMIN-NEXT:    vmv8r.v v24, v16
 ; ZVFHMIN-NEXT:    fmv.x.h a2, fa0
 ; ZVFHMIN-NEXT:    lui a3, 8
 ; ZVFHMIN-NEXT:    csrr a1, vlenb
-; ZVFHMIN-NEXT:    vsetvli a4, zero, e16, m8, ta, ma
 ; ZVFHMIN-NEXT:    vmv.v.x v16, a2
 ; ZVFHMIN-NEXT:    csrr a2, vlenb
 ; ZVFHMIN-NEXT:    slli a2, a2, 3
diff --git a/llvm/test/CodeGen/RISCV/rvv/vfmadd-constrained-sdnode.ll b/llvm/test/CodeGen/RISCV/rvv/vfmadd-constrained-sdnode.ll
index 1d471ab2404b1..a4f3a7d3a09a8 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vfmadd-constrained-sdnode.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vfmadd-constrained-sdnode.ll
@@ -227,6 +227,7 @@ define <vscale x 32 x bfloat> @vfmadd_vv_nxv32bf16(<vscale x 32 x bfloat> %va, <
 ; CHECK-NEXT:    slli a1, a1, 5
 ; CHECK-NEXT:    sub sp, sp, a1
 ; CHECK-NEXT:    .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x20, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 32 * vlenb
+; CHECK-NEXT:    vsetvli a1, zero, e16, m4, ta, ma
 ; CHECK-NEXT:    vmv8r.v v0, v16
 ; CHECK-NEXT:    addi a1, sp, 16
 ; CHECK-NEXT:    vs8r.v v16, (a1) # Unknown-size Folded Spill
@@ -237,7 +238,6 @@ define <vscale x 32 x bfloat> @vfmadd_vv_nxv32bf16(<vscale x 32 x bfloat> %va, <
 ; CHECK-NEXT:    add a0, sp, a0
 ; CHECK-NEXT:    addi a0, a0, 16
 ; CHECK-NEXT:    vs8r.v v8, (a0) # Unknown-size Folded Spill
-; CHECK-NEXT:    vsetvli a0, zero, e16, m4, ta, ma
 ; CHECK-NEXT:    vfwcvtbf16.f.f.v v24, v16
 ; CHECK-NEXT:    csrr a0, vlenb
 ; CHECK-NEXT:    li a1, 24
@@ -314,6 +314,7 @@ define <vscale x 32 x bfloat> @vfmadd_vf_nxv32bf16(<vscale x 32 x bfloat> %va, <
 ; CHECK-NEXT:    mul a0, a0, a1
 ; CHECK-NEXT:    sub sp, sp, a0
 ; CHECK-NEXT:    .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x18, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 24 * vlenb
+; CHECK-NEXT:    vsetvli a0, zero, e16, m4, ta, ma
 ; CHECK-NEXT:    vmv8r.v v24, v16
 ; CHECK-NEXT:    csrr a0, vlenb
 ; CHECK-NEXT:    slli a0, a0, 3
@@ -321,7 +322,6 @@ define <vscale x 32 x bfloat> @vfmadd_vf_nxv32bf16(<vscale x 32 x bfloat> %va, <
 ; CHECK-NEXT:    addi a0, a0, 16
 ; CHECK-NEXT:    vs8r.v v16, (a0) # Unknown-size Folded Spill
 ; CHECK-NEXT:    fmv.x.h a0, fa0
-; CHECK-NEXT:    vsetvli a1, zero, e16, m4, ta, ma
 ; CHECK-NEXT:    vfwcvtbf16.f.f.v v16, v8
 ; CHECK-NEXT:    addi a1, sp, 16
 ; CHECK-NEXT:    vs8r.v v16, (a1) # Unknown-size Folded Spill
@@ -664,6 +664,7 @@ define <vscale x 32 x half> @vfmadd_vv_nxv32f16(<vscale x 32 x half> %va, <vscal
 ; ZVFHMIN-NEXT:    slli a1, a1, 5
 ; ZVFHMIN-NEXT:    sub sp, sp, a1
 ; ZVFHMIN-NEXT:    .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x20, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 32 * vlenb
+; ZVFHMIN-NEXT:    vsetvli a1, zero, e16, m4, ta, ma
 ; ZVFHMIN-NEXT:    vmv8r.v v0, v16
 ; ZVFHMIN-NEXT:    addi a1, sp, 16
 ; ZVFHMIN-NEXT:    vs8r.v v16, (a1) # Unknown-size Folded Spill
@@ -674,7 +675,6 @@ define <vscale x 32 x half> @vfmadd_vv_nxv32f16(<vscale x 32 x half> %va, <vscal
 ; ZVFHMIN-NEXT:    add a0, sp, a0
 ; ZVFHMIN-NEXT:    addi a0, a0, 16
 ; ZVFHMIN-NEXT:    vs8r.v v8, (a0) # Unknown-size Folded Spill
-; ZVFHMIN-NEXT:    vsetvli a0, zero, e16, m4, ta, ma
 ; ZVFHMIN-NEXT:    vfwcvt.f.f.v v24, v16
 ; ZVFHMIN-NEXT:    csrr a0, vlenb
 ; ZVFHMIN-NEXT:    li a1, 24
@@ -757,6 +757,7 @@ define <vscale x 32 x half> @vfmadd_vf_nxv32f16(<vscale x 32 x half> %va, <vscal
 ; ZVFHMIN-NEXT:    mul a0, a0, a1
 ; ZVFHMIN-NEXT:    sub sp, sp, a0
 ; ZVFHMIN-NEXT:    .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x18, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 24 * vlenb
+; ZVFHMIN-NEXT:    vsetvli a0, zero, e16, m4, ta, ma
 ; ZVFHMIN-NEXT:    vmv8r.v v24, v16
 ; ZVFHMIN-NEXT:    csrr a0, vlenb
 ; ZVFHMIN-NEXT:    slli a0, a0, 3
@@ -764,7 +765,6 @@ define <vscale x 32 x half> @vfmadd_vf_nxv32f16(<vscale x 32 x half> %va, <vscal
 ; ZVFHMIN-NEXT:    addi a0, a0, 16
 ; ZVFHMIN-NEXT:    vs8r.v v16, (a0) # Unknown-size Folded Spill
 ; ZVFHMIN-NEXT:    fmv.x.h a0, fa0
-; ZVFHMIN-NEXT:    vsetvli a1, zero, e16, m4, ta, ma
 ; ZVFHMIN-NEXT:    vfwcvt.f.f.v v16, v8
 ; ZVFHMIN-NEXT:    addi a1, sp, 16
 ; ZVFHMIN-NEXT:    vs8r.v v16, (a1) # Unknown-size Folded Spill
diff --git a/llvm/test/CodeGen/RISCV/rvv/vfmadd-sdnode.ll b/llvm/test/CodeGen/RISCV/rvv/vfmadd-sdnode.ll
index 88fd81a5a2f7b..109023032e6ab 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vfmadd-sdnode.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vfmadd-sdnode.ll
@@ -226,6 +226,7 @@ define <vscale x 32 x bfloat> @vfmadd_vv_nxv32bf16(<vscale x 32 x bfloat> %va, <
 ; ZVFH-NEXT:    slli a1, a1, 5
 ; ZVFH-NEXT:    sub sp, sp, a1
 ; ZVFH-NEXT:    .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x20, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 32 * vlenb
+; ZVFH-NEXT:    vsetvli a1, zero, e16, m4, ta, ma
 ; ZVFH-NEXT:    vmv8r.v v0, v16
 ; ZVFH-NEXT:    addi a1, sp, 16
 ; ZVFH-NEXT:    vs8r.v v16, (a1) # Unknown-size Folded Spill
@@ -236,7 +237,6 @@ define <vscale x 32 x bfloat> @vfmadd_vv_nxv32bf16(<vscale x 32 x bfloat> %va, <
 ; ZVFH-NEXT:    add a0, sp, a0
 ; ZVFH-NEXT:    addi a0, a0, 16
 ; ZVFH-NEXT:    vs8r.v v8, (a0) # Unknown-size Folded Spill
-; ZVFH-NEXT:    vsetvli a0, zero, e16, m4, ta, ma
 ; ZVFH-NEXT:    vfwcvtbf16.f.f.v v24, v16
 ; ZVFH-NEXT:    csrr a0, vlenb
 ; ZVFH-NEXT:    slli a0, a0, 3
@@ -316,6 +316,7 @@ define <vscale x 32 x bfloat> @vfmadd_vv_nxv32bf16(<vscale x 32 x bfloat> %va, <
 ; ZVFHMIN-NEXT:    slli a1, a1, 5
 ; ZVFHMIN-NEXT:    sub sp, sp, a1
 ; ZVFHMIN-NEXT:    .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x20, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 32 * vlenb
+; ZVFHMIN-NEXT:    vsetvli a1, zero, e16, m4, ta, ma
 ; ZVFHMIN-NEXT:    vmv8r.v v0, v16
 ; ZVFHMIN-NEXT:    addi a1, sp, 16
 ; ZVFHMIN-NEXT:    vs8r.v v16, (a1) # Unknown-size Folded Spill
@@ -326,7 +327,6 @@ define <vscale x 32 x bfloat> @vfmadd_vv_nxv32bf16(<vscale x 32 x bfloat> %va, <
 ; ZVFHMIN-NEXT:    add a0, sp, a0
 ; ZVFHMIN-NEXT:    addi a0, a0, 16
 ; ZVFHMIN-NEXT:    vs8r.v v8, (a0) # Unknown-size Folded Spill
-; ZVFHMIN-NEXT:    vsetvli a0, zero, e16, m4, ta, ma
 ; ZVFHMIN-NEXT:    vfwcvtbf16.f.f.v v24, v16
 ; ZVFHMIN-NEXT:    csrr a0, vlenb
 ; ZVFHMIN-NEXT:    li a1, 24
@@ -402,12 +402,12 @@ define <vscale x 32 x bfloat> @vfmadd_vf_nxv32bf16(<vscale x 32 x bfloat> %va, <
 ; ZVFH-NEXT:    slli a0, a0, 5
 ; ZVFH-NEXT:    sub sp, sp, a0
 ; ZVFH-NEXT:    .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x20, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 32 * vlenb
+; ZVFH-NEXT:    vsetvli a0, zero, e16, m4, ta, ma
 ; ZVFH-NEXT:    vmv8r.v v0, v16
 ; ZVFH-NEXT:    addi a0, sp, 16
 ; ZVFH-NEXT:    vs8r.v v16, (a0) # Unknown-size Folded Spill
 ; ZVFH-NEXT:    vmv8r.v v16, v8
 ; ZVFH-NEXT:    fmv.x.h a0, fa0
-; ZVFH-NEXT:    vsetvli a1, zero, e16, m4, ta, ma
 ; ZVFH-NEXT:    vfwcvtbf16.f.f.v v24, v16
 ; ZVFH-NEXT:    csrr a1, vlenb
 ; ZVFH-NEXT:    slli a1, a1, 4
@@ -498,12 +498,12 @@ define <vscale x 32 x bfloat> @vfmadd_vf_nxv32bf16(<vscale x 32 x bfloat> %va, <
 ; ZVFHMIN-NEXT:    slli a0, a0, 5
 ; ZVFHMIN-NEXT:    sub sp, sp, a0
 ; ZVFHMIN-NEXT:    .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x20, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 32 * vlenb
+; ZVFHMIN-NEXT:    vsetvli a0, zero, e16, m4, ta, ma
 ; ZVFHMIN-NEXT:    vmv8r.v v0, v16
 ; ZVFHMIN-NEXT:    addi a0, sp, 16
 ; ZVFHMIN-NEXT:    vs8r.v v16, (a0) # Unknown-size Folded Spill
 ; ZVFHMIN-NEXT:    vmv8r.v v16, v8
 ; ZVFHMIN-NEXT:    fmv.x.h a0, fa0
-; ZVFHMIN-NEXT:    vsetvli a1, zero, e16, m4, ta, ma
 ; ZVFHMIN-NEXT:    vfwcvtbf16.f.f.v v24, v16
 ; ZVFHMIN-NEXT:    csrr a1, vlenb
 ; ZVFHMIN-NEXT:    slli a1, a1, 4
@@ -875,6 +875,7 @@ define <vscale x 32 x half> @vfmadd_vv_nxv32f16(<vscale x 32 x half> %va, <vscal
 ; ZVFHMIN-NEXT:    slli a1, a1, 5
 ; ZVFHMIN-NEXT:    sub sp, sp, a1
 ; ZVFHMIN-NEXT:    .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x20, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 32 * vlenb
+; ZVFHMIN-NEXT:    vsetvli a1, zero, e16, m4, ta, ma
 ; ZVFHMIN-NEXT:    vmv8r.v v0, v16
 ; ZVFHMIN-NEXT:    addi a1, sp, 16
 ; ZVFHMIN-NEXT:    vs8r.v v16, (a1) # Unknown-size Folded Spill
@@ -885,7 +886,6 @@ define <vscale x 32 x half> @vfmadd_vv_nxv32f16(<vscale x 32 x half> %va, <vscal
 ; ZVFHMIN-NEXT:    add a0, sp, a0
 ; ZVFHMIN-NEXT:    addi a0, a0, 16
 ; ZVFHMIN-NEXT:    vs8r.v v8, (a0) # Unknown-size Folded Spill
-; ZVFHMIN-NEXT:    vsetvli a0, zero, e16, m4, ta, ma
 ; ZVFHMIN-NEXT:    vfwcvt.f.f.v v24, v16
 ; ZVFHMIN-NEXT:    csrr a0, vlenb
 ; ZVFHMIN-NEXT:    li a1, 24
@@ -967,12 +967,12 @@ define <vscale x 32 x half> @vfmadd_vf_nxv32f16(<vscale x 32 x half> %va, <vscal
 ; ZVFHMIN-NEXT:    slli a0, a0, 5
 ; ZVFHMIN-NEXT:    sub sp, sp, a0
 ; ZVFHMIN-NEXT:    .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x20, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 32 * vlenb
+; ZVFHMIN-NEXT:    vsetvli a0, zero, e16, m4, ta, ma
 ; ZVFHMIN-NEXT:    vmv8r.v v0, v16
 ; ZVFHMIN-NEXT:    addi a0, sp, 16
 ; ZVFHMIN-NEXT:    vs8r.v v16, (a0) # Unknown-size Folded Spill
 ; ZVFHMIN-NEXT:    vmv8r.v v16, v8
 ; ZVFHMIN-NEXT:    fmv.x.h a0, fa0
-; ZVFHMIN-NEXT:    vsetvli a1, zero, e16, m4, ta, ma
 ; ZVFHMIN-NEXT:    vfwcvt.f.f.v v24, v16
 ; ZVFHMIN-NEXT:    csrr a1, vlenb
 ; ZVFHMIN-NEXT:    slli a1, a1, 4
diff --git a/llvm/test/CodeGen/RISCV/rvv/vfmax-vp.ll b/llvm/test/CodeGen/RISCV/rvv/vfmax-vp.ll
index dafcf8a1410d3..02d6229e99248 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vfmax-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vfmax-vp.ll
@@ -183,11 +183,11 @@ define <vscale x 32 x bfloat> @vfmax_vv_nxv32bf16(<vscale x 32 x bfloat> %va, <v
 ; CHECK-NEXT:    slli a1, a1, 3
 ; CHECK-NEXT:    sub sp, sp, a1
 ; CHECK-NEXT:    .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x08, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 8 * vlenb
+; CHECK-NEXT:    vsetvli a1, zero, e16, m4, ta, ma
 ; CHECK-NEXT:    vmv1r.v v7, v0
 ; CHECK-NEXT:    csrr a2, vlenb
 ; CHECK-NEXT:    addi a1, sp, 16
 ; CHECK-NEXT:    vs8r.v v16, (a1) # Unknown-size Folded Spill
-; CHECK-NEXT:    vsetvli a1, zero, e16, m4, ta, ma
 ; CHECK-NEXT:    vfwcvtbf16.f.f.v v24, v20
 ; CHECK-NEXT:    slli a1, a2, 1
 ; CHECK-NEXT:    srli a2, a2, 2
@@ -516,11 +516,11 @@ define <vscale x 32 x half> @vfmax_vv_nxv32f16(<vscale x 32 x half> %va, <vscale
 ; ZVFHMIN-NEXT:    slli a1, a1, 3
 ; ZVFHMIN-NEXT:    sub sp, sp, a1
 ; ZVFHMIN-NEXT:    .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x08, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 8 * vlenb
+; ZVFHMIN-NEXT:    vsetvli a1, zero, e16, m4, ta, ma
 ; ZVFHMIN-NEXT:    vmv1r.v v7, v0
 ; ZVFHMIN-NEXT:    csrr a2, vlenb
 ; ZVFHMIN-NEXT:    addi a1, sp, 16
 ; ZVFHMIN-NEXT:    vs8r.v v16, (a1) # Unknown-size Folded Spill
-; ZVFHMIN-NEXT:    vsetvli a1, zero, e16, m4, ta, ma
 ; ZVFHMIN-NEXT:    vfwcvt.f.f.v v24, v20
 ; ZVFHMIN-NEXT:    slli a1, a2, 1
 ; ZVFHMIN-NEXT:    srli a2, a2, 2
diff --git a/llvm/test/CodeGen/RISCV/rvv/vfmin-vp.ll b/llvm/test/CodeGen/RISCV/rvv/vfmin-vp.ll
index b3df6572f7936..f7f80299785d4 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vfmin-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vfmin-vp.ll
@@ -183,11 +183,11 @@ define <vscale x 32 x bfloat> @vfmin_vv_nxv32bf16(<vscale x 32 x bfloat> %va, <v
 ; CHECK-NEXT:    slli a1, a1, 3
 ; CHECK-NEXT:    sub sp, sp, a1
 ; CHECK-NEXT:    .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x08, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 8 * vlenb
+; CHECK-NEXT:    vsetvli a1, zero, e16, m4, ta, ma
 ; CHECK-NEXT:    vmv1r.v v7, v0
 ; CHECK-NEXT:    csrr a2, vlenb
 ; CHECK-NEXT:    addi a1, sp, 16
 ; CHECK-NEXT:    vs8r.v v16, (a1) # Unknown-size Folded Spill
-; CHECK-NEXT:    vsetvli a1, zero, e16, m4, ta, ma
 ; CHECK-NEXT:    vfwcvtbf16.f.f.v v24, v20
 ; CHECK-NEXT:    slli a1, a2, 1
 ; CHECK-NEXT:    srli a2, a2, 2
@@ -516,11 +516,11 @@ define <vscale x 32 x half> @vfmin_vv_nxv32f16(<vscale x 32 x half> %va, <vscale
 ; ZVFHMIN-NEXT:    slli a1, a1, 3
 ; ZVFHMIN-NEXT:    sub sp, sp, a1
 ; ZVFHMIN-NEXT:    .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x08, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 8 * vlenb
+; ZVFHMIN-NEXT:    vsetvli a1, zero, e16, m4, ta, ma
 ; ZVFHMIN-NEXT:    vmv1r.v v7, v0
 ; ZVFHMIN-NEXT:    csrr a2, vlenb
 ; ZVFHMIN-NEXT:    addi a1, sp, 16
 ; ZVFHMIN-NEXT:    vs8r.v v16, (a1) # Unknown-size Folded Spill
-; ZVFHMIN-NEXT:    vsetvli a1, zero, e16, m4, ta, ma
 ; ZVFHMIN-NEXT:    vfwcvt.f.f.v v24, v20
 ; ZVFHMIN-NEXT:    slli a1, a2, 1
 ; ZVFHMIN-NEXT:    srli a2, a2, 2
diff --git a/llvm/test/CodeGen/RISCV/rvv/vfmul-vp.ll b/llvm/test/CodeGen/RISCV/rvv/vfmul-vp.ll
index f4a236df4c9e4..7e5523044a010 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vfmul-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vfmul-vp.ll
@@ -495,11 +495,11 @@ define <vscale x 32 x half> @vfmul_vv_nxv32f16(<vscale x 32 x half> %va, <vscale
 ; ZVFHMIN-NEXT:    slli a1, a1, 3
 ; ZVFHMIN-NEXT:    sub sp, sp, a1
 ; ZVFHMIN-NEXT:    .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x08, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 8 * vlenb
+; ZVFHMIN-NEXT:    vsetvli a1, zero, e16, m4, ta, ma
 ; ZVFHMIN-NEXT:    vmv1r.v v7, v0
 ; ZVFHMIN-NEXT:    csrr a2, vlenb
 ; ZVFHMIN-NEXT:    addi a1, sp, 16
 ; ZVFHMIN-NEXT:    vs8r.v v16, (a1) # Unknown-size Folded Spill
-; ZVFHMIN-NEXT:    vsetvli a1, zero, e16, m4, ta, ma
 ; ZVFHMIN-NEXT:    vfwcvt.f.f.v v24, v20
 ; ZVFHMIN-NEXT:    slli a1, a2, 1
 ; ZVFHMIN-NEXT:    srli a2, a2, 2
@@ -614,12 +614,12 @@ define <vscale x 32 x half> @vfmul_vf_nxv32f16(<vscale x 32 x half> %va, half %b
 ; ZVFHMIN-NEXT:    add a1, a2, a1
 ; ZVFHMIN-NEXT:    sub sp, sp, a1
 ; ZVFHMIN-NEXT:    .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x11, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 17 * vlenb
+; ZVFHMIN-NEXT:    vsetvli a1, zero, e16, m4, ta, ma
 ; ZVFHMIN-NEXT:    vmv8r.v v16, v8
 ; ZVFHMIN-NEXT:    fmv.x.h a1, fa0
 ; ZVFHMIN-NEXT:    csrr a2, vlenb
 ; ZVFHMIN-NEXT:    addi a3, sp, 16
 ; ZVFHMIN-NEXT:    vs8r.v v8, (a3) # Unknown-size Folded Spill
-; ZVFHMIN-NEXT:    vsetvli a3, zero, e16, m4, ta, ma
 ; ZVFHMIN-NEXT:    vfwcvt.f.f.v v8, v20
 ; ZVFHMIN-NEXT:    vsetvli a3, zero, e16, m8, ta, ma
 ; ZVFHMIN-NEXT:    vmv.v.x v16, a1
@@ -706,10 +706,10 @@ define <vscale x 32 x half> @vfmul_vf_nxv32f16_unmasked(<vscale x 32 x half> %va
 ; ZVFHMIN-NEXT:    slli a1, a1, 4
 ; ZVFHMIN-NEXT:    sub sp, sp, a1
 ; ZVFHMIN-NEXT:    .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x10, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 16 * vlenb
+; ZVFHMIN-NEXT:    vsetvli a1, zero, e8, m4, ta, ma
 ; ZVFHMIN-NEXT:    vmv8r.v v16, v8
 ; ZVFHMIN-NEXT:    fmv.x.h a1, fa0
 ; ZVFHMIN-NEXT:    csrr a2, vlenb
-; ZVFHMIN-NEXT:    vsetvli a3, zero, e8, m4, ta, ma
 ; ZVFHMIN-NEXT:    vmset.m v7
 ; ZVFHMIN-NEXT:    addi a3, sp, 16
 ; ZVFHMIN-NEXT:    vs8r.v v8, (a3) # Unknown-size Folded Spill
diff --git a/llvm/test/CodeGen/RISCV/rvv/vfmuladd-vp.ll b/llvm/test/CodeGen/RISCV/rvv/vfmuladd-vp.ll
index d1702268f829f..901f3cd63fa9e 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vfmuladd-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vfmuladd-vp.ll
@@ -1112,6 +1112,7 @@ define <vscale x 16 x double> @vfma_vv_nxv16f64(<vscale x 16 x double> %va, <vsc
 ; CHECK-NEXT:    mul a1, a1, a3
 ; CHECK-NEXT:    sub sp, sp, a1
 ; CHECK-NEXT:    .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x30, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 48 * vlenb
+; CHECK-NEXT:    vsetvli a1, zero, e8, mf4, ta, ma
 ; CHECK-NEXT:    vmv1r.v v7, v0
 ; CHECK-NEXT:    csrr a1, vlenb
 ; CHECK-NEXT:    li a3, 24
@@ -1154,7 +1155,6 @@ define <vscale x 16 x double> @vfma_vv_nxv16f64(<vscale x 16 x double> %va, <vsc
 ; CHECK-NEXT:    vl8re64.v v8, (a0)
 ; CHECK-NEXT:    addi a0, sp, 16
 ; CHECK-NEXT:    vs8r.v v8, (a0) # Unknown-size Folded Spill
-; CHECK-NEXT:    vsetvli a0, zero, e8, mf4, ta, ma
 ; CHECK-NEXT:    vslidedown.vx v0, v0, a3
 ; CHECK-NEXT:    and a0, a7, a6
 ; CHECK-NEXT:    csrr a2, vlenb
diff --git a/llvm/test/CodeGen/RISCV/rvv/vfneg-vp.ll b/llvm/test/CodeGen/RISCV/rvv/vfneg-vp.ll
index 343098e87649e..bbab056f0ff46 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vfneg-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vfneg-vp.ll
@@ -450,11 +450,11 @@ declare <vscale x 16 x double> @llvm.vp.fneg.nxv16f64(<vscale x 16 x double>, <v
 define <vscale x 16 x double> @vfneg_vv_nxv16f64(<vscale x 16 x double> %va, <vscale x 16 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vfneg_vv_nxv16f64:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli a1, zero, e8, mf4, ta, ma
 ; CHECK-NEXT:    vmv1r.v v24, v0
 ; CHECK-NEXT:    csrr a1, vlenb
 ; CHECK-NEXT:    srli a2, a1, 3
 ; CHECK-NEXT:    sub a3, a0, a1
-; CHECK-NEXT:    vsetvli a4, zero, e8, mf4, ta, ma
 ; CHECK-NEXT:    vslidedown.vx v0, v0, a2
 ; CHECK-NEXT:    sltu a2, a0, a3
 ; CHECK-NEXT:    addi a2, a2, -1
diff --git a/llvm/test/CodeGen/RISCV/rvv/vfnmadd-constrained-sdnode.ll b/llvm/test/CodeGen/RISCV/rvv/vfnmadd-constrained-sdnode.ll
index 3705e73fda492..b8ec285b5c34e 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vfnmadd-constrained-sdnode.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vfnmadd-constrained-sdnode.ll
@@ -329,6 +329,7 @@ define <vscale x 32 x half> @vfnmsub_vv_nxv32f16(<vscale x 32 x half> %va, <vsca
 ; ZVFHMIN-NEXT:    slli a1, a1, 5
 ; ZVFHMIN-NEXT:    sub sp, sp, a1
 ; ZVFHMIN-NEXT:    .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x20, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 32 * vlenb
+; ZVFHMIN-NEXT:    vsetvli a1, zero, e16, m4, ta, ma
 ; ZVFHMIN-NEXT:    vmv8r.v v0, v16
 ; ZVFHMIN-NEXT:    csrr a1, vlenb
 ; ZVFHMIN-NEXT:    li a2, 24
@@ -338,7 +339,6 @@ define <vscale x 32 x half> @vfnmsub_vv_nxv32f16(<vscale x 32 x half> %va, <vsca
 ; ZVFHMIN-NEXT:    vs8r.v v8, (a1) # Unknown-size Folded Spill
 ; ZVFHMIN-NEXT:    vl8re16.v v16, (a0)
 ; ZVFHMIN-NEXT:    lui a0, 8
-; ZVFHMIN-NEXT:    vsetvli a1, zero, e16, m4, ta, ma
 ; ZVFHMIN-NEXT:    vfwcvt.f.f.v v24, v8
 ; ZVFHMIN-NEXT:    vsetvli a1, zero, e16, m8, ta, ma
 ; ZVFHMIN-NEXT:    vxor.vx v8, v0, a0
diff --git a/llvm/test/CodeGen/RISCV/rvv/vfnmsub-constrained-sdnode.ll b/llvm/test/CodeGen/RISCV/rvv/vfnmsub-constrained-sdnode.ll
index 80edf0e3a4d81..742083f6f684b 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vfnmsub-constrained-sdnode.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vfnmsub-constrained-sdnode.ll
@@ -310,6 +310,7 @@ define <vscale x 32 x half> @vfnmsub_vv_nxv32f16(<vscale x 32 x half> %va, <vsca
 ; ZVFHMIN-NEXT:    mul a1, a1, a2
 ; ZVFHMIN-NEXT:    sub sp, sp, a1
 ; ZVFHMIN-NEXT:    .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x18, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 24 * vlenb
+; ZVFHMIN-NEXT:    vsetvli a1, zero, e16, m4, ta, ma
 ; ZVFHMIN-NEXT:    vmv8r.v v0, v16
 ; ZVFHMIN-NEXT:    csrr a1, vlenb
 ; ZVFHMIN-NEXT:    slli a1, a1, 4
@@ -323,7 +324,6 @@ define <vscale x 32 x half> @vfnmsub_vv_nxv32f16(<vscale x 32 x half> %va, <vsca
 ; ZVFHMIN-NEXT:    vs8r.v v8, (a1) # Unknown-size Folded Spill
 ; ZVFHMIN-NEXT:    vl8re16.v v24, (a0)
 ; ZVFHMIN-NEXT:    lui a0, 8
-; ZVFHMIN-NEXT:    vsetvli a1, zero, e16, m4, ta, ma
 ; ZVFHMIN-NEXT:    vfwcvt.f.f.v v16, v0
 ; ZVFHMIN-NEXT:    vsetvli a1, zero, e16, m8, ta, ma
 ; ZVFHMIN-NEXT:    vxor.vx v0, v24, a0
@@ -389,6 +389,7 @@ define <vscale x 32 x half> @vfnmsub_vf_nxv32f16(<vscale x 32 x half> %va, <vsca
 ; ZVFHMIN-NEXT:    add a0, sp, a0
 ; ZVFHMIN-NEXT:    addi a0, a0, 16
 ; ZVFHMIN-NEXT:    vs8r.v v16, (a0) # Unknown-size Folded Spill
+; ZVFHMIN-NEXT:    vsetvli a0, zero, e16, m8, ta, ma
 ; ZVFHMIN-NEXT:    vmv8r.v v0, v8
 ; ZVFHMIN-NEXT:    csrr a0, vlenb
 ; ZVFHMIN-NEXT:    slli a0, a0, 4
@@ -396,7 +397,6 @@ define <vscale x 32 x half> @vfnmsub_vf_nxv32f16(<vscale x 32 x half> %va, <vsca
 ; ZVFHMIN-NEXT:    addi a0, a0, 16
 ; ZVFHMIN-NEXT:    vs8r.v v8, (a0) # Unknown-size Folded Spill
 ; ZVFHMIN-NEXT:    fmv.x.h a0, fa0
-; ZVFHMIN-NEXT:    vsetvli a1, zero, e16, m8, ta, ma
 ; ZVFHMIN-NEXT:    vmv.v.x v24, a0
 ; ZVFHMIN-NEXT:    lui a0, 8
 ; ZVFHMIN-NEXT:    vsetvli a1, zero, e16, m4, ta, ma
diff --git a/llvm/test/CodeGen/RISCV/rvv/vfpext-vp.ll b/llvm/test/CodeGen/RISCV/rvv/vfpext-vp.ll
index 341fe678183b6..6debc47460683 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vfpext-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vfpext-vp.ll
@@ -96,11 +96,11 @@ declare <vscale x 32 x float> @llvm.vp.fpext.nxv32f32.nxv32f16(<vscale x 32 x ha
 define <vscale x 32 x float> @vfpext_nxv32f16_nxv32f32(<vscale x 32 x half> %a, <vscale x 32 x i1> %m, i32 zeroext %vl) {
 ; CHECK-LABEL: vfpext_nxv32f16_nxv32f32:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli a1, zero, e8, mf2, ta, ma
 ; CHECK-NEXT:    vmv1r.v v24, v0
 ; CHECK-NEXT:    csrr a1, vlenb
 ; CHECK-NEXT:    srli a2, a1, 2
 ; CHECK-NEXT:    slli a1, a1, 1
-; CHECK-NEXT:    vsetvli a3, zero, e8, mf2, ta, ma
 ; CHECK-NEXT:    vslidedown.vx v0, v0, a2
 ; CHECK-NEXT:    sub a2, a0, a1
 ; CHECK-NEXT:    sltu a3, a0, a2
diff --git a/llvm/test/CodeGen/RISCV/rvv/vfptosi-vp.ll b/llvm/test/CodeGen/RISCV/rvv/vfptosi-vp.ll
index cf195c7c0935e..d990c74c67d5a 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vfptosi-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vfptosi-vp.ll
@@ -508,11 +508,11 @@ declare <vscale x 32 x i16> @llvm.vp.fptosi.nxv32i16.nxv32f32(<vscale x 32 x flo
 define <vscale x 32 x i16> @vfptosi_nxv32i16_nxv32f32(<vscale x 32 x float> %va, <vscale x 32 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vfptosi_nxv32i16_nxv32f32:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli a1, zero, e8, mf2, ta, ma
 ; CHECK-NEXT:    vmv1r.v v24, v0
 ; CHECK-NEXT:    csrr a1, vlenb
 ; CHECK-NEXT:    srli a2, a1, 2
 ; CHECK-NEXT:    slli a1, a1, 1
-; CHECK-NEXT:    vsetvli a3, zero, e8, mf2, ta, ma
 ; CHECK-NEXT:    vslidedown.vx v0, v0, a2
 ; CHECK-NEXT:    sub a2, a0, a1
 ; CHECK-NEXT:    sltu a3, a0, a2
@@ -538,11 +538,11 @@ declare <vscale x 32 x i32> @llvm.vp.fptosi.nxv32i32.nxv32f32(<vscale x 32 x flo
 define <vscale x 32 x i32> @vfptosi_nxv32i32_nxv32f32(<vscale x 32 x float> %va, <vscale x 32 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vfptosi_nxv32i32_nxv32f32:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli a1, zero, e8, mf2, ta, ma
 ; CHECK-NEXT:    vmv1r.v v24, v0
 ; CHECK-NEXT:    csrr a1, vlenb
 ; CHECK-NEXT:    srli a2, a1, 2
 ; CHECK-NEXT:    slli a1, a1, 1
-; CHECK-NEXT:    vsetvli a3, zero, e8, mf2, ta, ma
 ; CHECK-NEXT:    vslidedown.vx v0, v0, a2
 ; CHECK-NEXT:    sub a2, a0, a1
 ; CHECK-NEXT:    sltu a3, a0, a2
diff --git a/llvm/test/CodeGen/RISCV/rvv/vfptoui-vp.ll b/llvm/test/CodeGen/RISCV/rvv/vfptoui-vp.ll
index 952d28604b86c..3b24a648d97f5 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vfptoui-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vfptoui-vp.ll
@@ -508,11 +508,11 @@ declare <vscale x 32 x i16> @llvm.vp.fptoui.nxv32i16.nxv32f32(<vscale x 32 x flo
 define <vscale x 32 x i16> @vfptoui_nxv32i16_nxv32f32(<vscale x 32 x float> %va, <vscale x 32 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vfptoui_nxv32i16_nxv32f32:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli a1, zero, e8, mf2, ta, ma
 ; CHECK-NEXT:    vmv1r.v v24, v0
 ; CHECK-NEXT:    csrr a1, vlenb
 ; CHECK-NEXT:    srli a2, a1, 2
 ; CHECK-NEXT:    slli a1, a1, 1
-; CHECK-NEXT:    vsetvli a3, zero, e8, mf2, ta, ma
 ; CHECK-NEXT:    vslidedown.vx v0, v0, a2
 ; CHECK-NEXT:    sub a2, a0, a1
 ; CHECK-NEXT:    sltu a3, a0, a2
@@ -538,11 +538,11 @@ declare <vscale x 32 x i32> @llvm.vp.fptoui.nxv32i32.nxv32f32(<vscale x 32 x flo
 define <vscale x 32 x i32> @vfptoui_nxv32i32_nxv32f32(<vscale x 32 x float> %va, <vscale x 32 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vfptoui_nxv32i32_nxv32f32:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli a1, zero, e8, mf2, ta, ma
 ; CHECK-NEXT:    vmv1r.v v24, v0
 ; CHECK-NEXT:    csrr a1, vlenb
 ; CHECK-NEXT:    srli a2, a1, 2
 ; CHECK-NEXT:    slli a1, a1, 1
-; CHECK-NEXT:    vsetvli a3, zero, e8, mf2, ta, ma
 ; CHECK-NEXT:    vslidedown.vx v0, v0, a2
 ; CHECK-NEXT:    sub a2, a0, a1
 ; CHECK-NEXT:    sltu a3, a0, a2
diff --git a/llvm/test/CodeGen/RISCV/rvv/vfptrunc-vp.ll b/llvm/test/CodeGen/RISCV/rvv/vfptrunc-vp.ll
index 874813f057595..63156e1399293 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vfptrunc-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vfptrunc-vp.ll
@@ -102,13 +102,13 @@ define <vscale x 16 x float> @vfptrunc_nxv16f32_nxv16f64(<vscale x 16 x double>
 ; CHECK-NEXT:    slli a1, a1, 3
 ; CHECK-NEXT:    sub sp, sp, a1
 ; CHECK-NEXT:    .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x08, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 8 * vlenb
+; CHECK-NEXT:    vsetvli a1, zero, e8, mf4, ta, ma
 ; CHECK-NEXT:    vmv1r.v v7, v0
 ; CHECK-NEXT:    addi a1, sp, 16
 ; CHECK-NEXT:    vs8r.v v16, (a1) # Unknown-size Folded Spill
 ; CHECK-NEXT:    csrr a1, vlenb
 ; CHECK-NEXT:    srli a2, a1, 3
 ; CHECK-NEXT:    sub a3, a0, a1
-; CHECK-NEXT:    vsetvli a4, zero, e8, mf4, ta, ma
 ; CHECK-NEXT:    vslidedown.vx v0, v0, a2
 ; CHECK-NEXT:    sltu a2, a0, a3
 ; CHECK-NEXT:    addi a2, a2, -1
@@ -147,6 +147,7 @@ define <vscale x 32 x float> @vfptrunc_nxv32f32_nxv32f64(<vscale x 32 x double>
 ; CHECK-NEXT:    slli a1, a1, 4
 ; CHECK-NEXT:    sub sp, sp, a1
 ; CHECK-NEXT:    .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x10, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 16 * vlenb
+; CHECK-NEXT:    vsetvli a1, zero, e8, mf2, ta, ma
 ; CHECK-NEXT:    vmv1r.v v7, v0
 ; CHECK-NEXT:    addi a1, sp, 16
 ; CHECK-NEXT:    vs8r.v v16, (a1) # Unknown-size Folded Spill
@@ -160,7 +161,6 @@ define <vscale x 32 x float> @vfptrunc_nxv32f32_nxv32f64(<vscale x 32 x double>
 ; CHECK-NEXT:    srli a5, a1, 2
 ; CHECK-NEXT:    slli a6, a1, 3
 ; CHECK-NEXT:    slli a4, a1, 1
-; CHECK-NEXT:    vsetvli a7, zero, e8, mf2, ta, ma
 ; CHECK-NEXT:    vslidedown.vx v16, v0, a5
 ; CHECK-NEXT:    add a6, a0, a6
 ; CHECK-NEXT:    sub a5, a2, a4
diff --git a/llvm/test/CodeGen/RISCV/rvv/vfsqrt-vp.ll b/llvm/test/CodeGen/RISCV/rvv/vfsqrt-vp.ll
index 8edcf23988c7f..8e57be1e0697c 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vfsqrt-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vfsqrt-vp.ll
@@ -167,12 +167,12 @@ declare <vscale x 32 x bfloat> @llvm.vp.sqrt.nxv32bf16(<vscale x 32 x bfloat>, <
 define <vscale x 32 x bfloat> @vfsqrt_vv_nxv32bf16(<vscale x 32 x bfloat> %va, <vscale x 32 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vfsqrt_vv_nxv32bf16:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli a1, zero, e8, mf2, ta, ma
 ; CHECK-NEXT:    vmv1r.v v16, v0
 ; CHECK-NEXT:    csrr a2, vlenb
 ; CHECK-NEXT:    slli a1, a2, 1
 ; CHECK-NEXT:    srli a2, a2, 2
 ; CHECK-NEXT:    sub a3, a0, a1
-; CHECK-NEXT:    vsetvli a4, zero, e8, mf2, ta, ma
 ; CHECK-NEXT:    vslidedown.vx v0, v0, a2
 ; CHECK-NEXT:    sltu a2, a0, a3
 ; CHECK-NEXT:    addi a2, a2, -1
@@ -452,12 +452,12 @@ define <vscale x 32 x half> @vfsqrt_vv_nxv32f16(<vscale x 32 x half> %va, <vscal
 ;
 ; ZVFHMIN-LABEL: vfsqrt_vv_nxv32f16:
 ; ZVFHMIN:       # %bb.0:
+; ZVFHMIN-NEXT:    vsetvli a1, zero, e8, mf2, ta, ma
 ; ZVFHMIN-NEXT:    vmv1r.v v16, v0
 ; ZVFHMIN-NEXT:    csrr a2, vlenb
 ; ZVFHMIN-NEXT:    slli a1, a2, 1
 ; ZVFHMIN-NEXT:    srli a2, a2, 2
 ; ZVFHMIN-NEXT:    sub a3, a0, a1
-; ZVFHMIN-NEXT:    vsetvli a4, zero, e8, mf2, ta, ma
 ; ZVFHMIN-NEXT:    vslidedown.vx v0, v0, a2
 ; ZVFHMIN-NEXT:    sltu a2, a0, a3
 ; ZVFHMIN-NEXT:    addi a2, a2, -1
@@ -749,11 +749,11 @@ declare <vscale x 16 x double> @llvm.vp.sqrt.nxv16f64(<vscale x 16 x double>, <v
 define <vscale x 16 x double> @vfsqrt_vv_nxv16f64(<vscale x 16 x double> %va, <vscale x 16 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vfsqrt_vv_nxv16f64:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli a1, zero, e8, mf4, ta, ma
 ; CHECK-NEXT:    vmv1r.v v24, v0
 ; CHECK-NEXT:    csrr a1, vlenb
 ; CHECK-NEXT:    srli a2, a1, 3
 ; CHECK-NEXT:    sub a3, a0, a1
-; CHECK-NEXT:    vsetvli a4, zero, e8, mf4, ta, ma
 ; CHECK-NEXT:    vslidedown.vx v0, v0, a2
 ; CHECK-NEXT:    sltu a2, a0, a3
 ; CHECK-NEXT:    addi a2, a2, -1
diff --git a/llvm/test/CodeGen/RISCV/rvv/vfsub-vp.ll b/llvm/test/CodeGen/RISCV/rvv/vfsub-vp.ll
index 25a80e66c4a52..d034f65479a15 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vfsub-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vfsub-vp.ll
@@ -373,11 +373,11 @@ define <vscale x 32 x bfloat> @vfsub_vv_nxv32bf16(<vscale x 32 x bfloat> %va, <v
 ; CHECK-NEXT:    slli a1, a1, 3
 ; CHECK-NEXT:    sub sp, sp, a1
 ; CHECK-NEXT:    .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x08, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 8 * vlenb
+; CHECK-NEXT:    vsetvli a1, zero, e16, m4, ta, ma
 ; CHECK-NEXT:    vmv1r.v v7, v0
 ; CHECK-NEXT:    csrr a2, vlenb
 ; CHECK-NEXT:    addi a1, sp, 16
 ; CHECK-NEXT:    vs8r.v v16, (a1) # Unknown-size Folded Spill
-; CHECK-NEXT:    vsetvli a1, zero, e16, m4, ta, ma
 ; CHECK-NEXT:    vfwcvtbf16.f.f.v v24, v20
 ; CHECK-NEXT:    slli a1, a2, 1
 ; CHECK-NEXT:    srli a2, a2, 2
@@ -480,12 +480,12 @@ define <vscale x 32 x bfloat> @vfsub_vf_nxv32bf16(<vscale x 32 x bfloat> %va, bf
 ; CHECK-NEXT:    add a1, a2, a1
 ; CHECK-NEXT:    sub sp, sp, a1
 ; CHECK-NEXT:    .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x11, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 17 * vlenb
+; CHECK-NEXT:    vsetvli a1, zero, e16, m4, ta, ma
 ; CHECK-NEXT:    vmv8r.v v16, v8
 ; CHECK-NEXT:    fmv.x.h a1, fa0
 ; CHECK-NEXT:    csrr a2, vlenb
 ; CHECK-NEXT:    addi a3, sp, 16
 ; CHECK-NEXT:    vs8r.v v8, (a3) # Unknown-size Folded Spill
-; CHECK-NEXT:    vsetvli a3, zero, e16, m4, ta, ma
 ; CHECK-NEXT:    vfwcvtbf16.f.f.v v8, v20
 ; CHECK-NEXT:    vsetvli a3, zero, e16, m8, ta, ma
 ; CHECK-NEXT:    vmv.v.x v16, a1
@@ -566,10 +566,10 @@ define <vscale x 32 x bfloat> @vfsub_vf_nxv32bf16_unmasked(<vscale x 32 x bfloat
 ; CHECK-NEXT:    slli a1, a1, 4
 ; CHECK-NEXT:    sub sp, sp, a1
 ; CHECK-NEXT:    .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x10, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 16 * vlenb
+; CHECK-NEXT:    vsetvli a1, zero, e8, m4, ta, ma
 ; CHECK-NEXT:    vmv8r.v v16, v8
 ; CHECK-NEXT:    fmv.x.h a1, fa0
 ; CHECK-NEXT:    csrr a2, vlenb
-; CHECK-NEXT:    vsetvli a3, zero, e8, m4, ta, ma
 ; CHECK-NEXT:    vmset.m v7
 ; CHECK-NEXT:    addi a3, sp, 16
 ; CHECK-NEXT:    vs8r.v v8, (a3) # Unknown-size Folded Spill
@@ -1117,11 +1117,11 @@ define <vscale x 32 x half> @vfsub_vv_nxv32f16(<vscale x 32 x half> %va, <vscale
 ; ZVFHMIN-NEXT:    slli a1, a1, 3
 ; ZVFHMIN-NEXT:    sub sp, sp, a1
 ; ZVFHMIN-NEXT:    .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x08, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 8 * vlenb
+; ZVFHMIN-NEXT:    vsetvli a1, zero, e16, m4, ta, ma
 ; ZVFHMIN-NEXT:    vmv1r.v v7, v0
 ; ZVFHMIN-NEXT:    csrr a2, vlenb
 ; ZVFHMIN-NEXT:    addi a1, sp, 16
 ; ZVFHMIN-NEXT:    vs8r.v v16, (a1) # Unknown-size Folded Spill
-; ZVFHMIN-NEXT:    vsetvli a1, zero, e16, m4, ta, ma
 ; ZVFHMIN-NEXT:    vfwcvt.f.f.v v24, v20
 ; ZVFHMIN-NEXT:    slli a1, a2, 1
 ; ZVFHMIN-NEXT:    srli a2, a2, 2
@@ -1236,12 +1236,12 @@ define <vscale x 32 x half> @vfsub_vf_nxv32f16(<vscale x 32 x half> %va, half %b
 ; ZVFHMIN-NEXT:    add a1, a2, a1
 ; ZVFHMIN-NEXT:    sub sp, sp, a1
 ; ZVFHMIN-NEXT:    .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x11, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 17 * vlenb
+; ZVFHMIN-NEXT:    vsetvli a1, zero, e16, m4, ta, ma
 ; ZVFHMIN-NEXT:    vmv8r.v v16, v8
 ; ZVFHMIN-NEXT:    fmv.x.h a1, fa0
 ; ZVFHMIN-NEXT:    csrr a2, vlenb
 ; ZVFHMIN-NEXT:    addi a3, sp, 16
 ; ZVFHMIN-NEXT:    vs8r.v v8, (a3) # Unknown-size Folded Spill
-; ZVFHMIN-NEXT:    vsetvli a3, zero, e16, m4, ta, ma
 ; ZVFHMIN-NEXT:    vfwcvt.f.f.v v8, v20
 ; ZVFHMIN-NEXT:    vsetvli a3, zero, e16, m8, ta, ma
 ; ZVFHMIN-NEXT:    vmv.v.x v16, a1
@@ -1328,10 +1328,10 @@ define <vscale x 32 x half> @vfsub_vf_nxv32f16_unmasked(<vscale x 32 x half> %va
 ; ZVFHMIN-NEXT:    slli a1, a1, 4
 ; ZVFHMIN-NEXT:    sub sp, sp, a1
 ; ZVFHMIN-NEXT:    .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x10, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 16 * vlenb
+; ZVFHMIN-NEXT:    vsetvli a1, zero, e8, m4, ta, ma
 ; ZVFHMIN-NEXT:    vmv8r.v v16, v8
 ; ZVFHMIN-NEXT:    fmv.x.h a1, fa0
 ; ZVFHMIN-NEXT:    csrr a2, vlenb
-; ZVFHMIN-NEXT:    vsetvli a3, zero, e8, m4, ta, ma
 ; ZVFHMIN-NEXT:    vmset.m v7
 ; ZVFHMIN-NEXT:    addi a3, sp, 16
 ; ZVFHMIN-NEXT:    vs8r.v v8, (a3) # Unknown-size Folded Spill
diff --git a/llvm/test/CodeGen/RISCV/rvv/vl-opt.ll b/llvm/test/CodeGen/RISCV/rvv/vl-opt.ll
index 1a1472fcfc66f..ff4f3e24ec17b 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vl-opt.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vl-opt.ll
@@ -111,8 +111,8 @@ define <vscale x 4 x i32> @different_vl_with_ta(<vscale x 4 x i32> %a, <vscale x
 define <vscale x 4 x i32> @different_vl_with_tu(<vscale x 4 x i32> %passthru, <vscale x 4 x i32> %a, <vscale x 4 x i32> %b, iXLen %vl1, iXLen %vl2) {
 ; CHECK-LABEL: different_vl_with_tu:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv2r.v v14, v10
 ; CHECK-NEXT:    vsetvli zero, a0, e32, m2, tu, ma
+; CHECK-NEXT:    vmv2r.v v14, v10
 ; CHECK-NEXT:    vadd.vv v14, v10, v12
 ; CHECK-NEXT:    vsetvli zero, a1, e32, m2, tu, ma
 ; CHECK-NEXT:    vadd.vv v8, v14, v10
@@ -126,8 +126,8 @@ define <vscale x 4 x i32> @different_vl_with_tu(<vscale x 4 x i32> %passthru, <v
 define <vscale x 4 x i32> @different_imm_vl_with_tu(<vscale x 4 x i32> %passthru, <vscale x 4 x i32> %a, <vscale x 4 x i32> %b, iXLen %vl1, iXLen %vl2) {
 ; CHECK-LABEL: different_imm_vl_with_tu:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv2r.v v14, v10
 ; CHECK-NEXT:    vsetivli zero, 5, e32, m2, tu, ma
+; CHECK-NEXT:    vmv2r.v v14, v10
 ; CHECK-NEXT:    vadd.vv v14, v10, v12
 ; CHECK-NEXT:    vsetivli zero, 4, e32, m2, tu, ma
 ; CHECK-NEXT:    vadd.vv v8, v14, v10
diff --git a/llvm/test/CodeGen/RISCV/rvv/vlsegff-rv32-dead.ll b/llvm/test/CodeGen/RISCV/rvv/vlsegff-rv32-dead.ll
index 1516d656663b6..d8bff08ea5513 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vlsegff-rv32-dead.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vlsegff-rv32-dead.ll
@@ -51,9 +51,9 @@ entry:
 define <vscale x 16 x i16> @test_vlseg2ff_mask_dead_vl(target("riscv.vector.tuple", <vscale x 32 x i8>, 2) %val, ptr %base, i32 %vl, <vscale x 16 x i1> %mask) {
 ; CHECK-LABEL: test_vlseg2ff_mask_dead_vl:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, m4, ta, mu
 ; CHECK-NEXT:    vmv4r.v v4, v8
 ; CHECK-NEXT:    vmv4r.v v8, v12
-; CHECK-NEXT:    vsetvli zero, a1, e16, m4, ta, mu
 ; CHECK-NEXT:    vlseg2e16ff.v v4, (a0), v0.t
 ; CHECK-NEXT:    ret
 entry:
diff --git a/llvm/test/CodeGen/RISCV/rvv/vlsegff-rv32.ll b/llvm/test/CodeGen/RISCV/rvv/vlsegff-rv32.ll
index b89097b8ff974..03aee1881503a 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vlsegff-rv32.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vlsegff-rv32.ll
@@ -25,9 +25,9 @@ entry:
 define <vscale x 1 x i8> @test_vlseg2ff_mask_nxv1i8_triscv.vector.tuple_nxv1i8_2t(target("riscv.vector.tuple", <vscale x 1 x i8>, 2) %val, ptr %base, i32 %vl, <vscale x 1 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg2ff_mask_nxv1i8_triscv.vector.tuple_nxv1i8_2t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e8, mf8, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
-; CHECK-NEXT:    vsetvli zero, a1, e8, mf8, ta, mu
 ; CHECK-NEXT:    vlseg2e8ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
@@ -64,9 +64,9 @@ entry:
 define <vscale x 2 x i8> @test_vlseg2ff_mask_nxv2i8_triscv.vector.tuple_nxv2i8_2t(target("riscv.vector.tuple", <vscale x 2 x i8>, 2) %val, ptr %base, i32 %vl, <vscale x 2 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg2ff_mask_nxv2i8_triscv.vector.tuple_nxv2i8_2t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e8, mf4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
-; CHECK-NEXT:    vsetvli zero, a1, e8, mf4, ta, mu
 ; CHECK-NEXT:    vlseg2e8ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
@@ -103,9 +103,9 @@ entry:
 define <vscale x 4 x i8> @test_vlseg2ff_mask_nxv4i8_triscv.vector.tuple_nxv4i8_2t(target("riscv.vector.tuple", <vscale x 4 x i8>, 2) %val, ptr %base, i32 %vl, <vscale x 4 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg2ff_mask_nxv4i8_triscv.vector.tuple_nxv4i8_2t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e8, mf2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
-; CHECK-NEXT:    vsetvli zero, a1, e8, mf2, ta, mu
 ; CHECK-NEXT:    vlseg2e8ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
@@ -142,9 +142,9 @@ entry:
 define <vscale x 8 x i8> @test_vlseg2ff_mask_nxv8i8_triscv.vector.tuple_nxv8i8_2t(target("riscv.vector.tuple", <vscale x 8 x i8>, 2) %val, ptr %base, i32 %vl, <vscale x 8 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg2ff_mask_nxv8i8_triscv.vector.tuple_nxv8i8_2t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e8, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
-; CHECK-NEXT:    vsetvli zero, a1, e8, m1, ta, mu
 ; CHECK-NEXT:    vlseg2e8ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
@@ -181,9 +181,9 @@ entry:
 define <vscale x 16 x i8> @test_vlseg2ff_mask_nxv16i8_triscv.vector.tuple_nxv16i8_2t(target("riscv.vector.tuple", <vscale x 16 x i8>, 2) %val, ptr %base, i32 %vl, <vscale x 16 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg2ff_mask_nxv16i8_triscv.vector.tuple_nxv16i8_2t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e8, m2, ta, mu
 ; CHECK-NEXT:    vmv2r.v v6, v8
 ; CHECK-NEXT:    vmv2r.v v8, v10
-; CHECK-NEXT:    vsetvli zero, a1, e8, m2, ta, mu
 ; CHECK-NEXT:    vlseg2e8ff.v v6, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
@@ -220,9 +220,9 @@ entry:
 define <vscale x 32 x i8> @test_vlseg2ff_mask_nxv32i8_triscv.vector.tuple_nxv32i8_2t(target("riscv.vector.tuple", <vscale x 32 x i8>, 2) %val, ptr %base, i32 %vl, <vscale x 32 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg2ff_mask_nxv32i8_triscv.vector.tuple_nxv32i8_2t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e8, m4, ta, mu
 ; CHECK-NEXT:    vmv4r.v v4, v8
 ; CHECK-NEXT:    vmv4r.v v8, v12
-; CHECK-NEXT:    vsetvli zero, a1, e8, m4, ta, mu
 ; CHECK-NEXT:    vlseg2e8ff.v v4, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
@@ -259,10 +259,10 @@ entry:
 define <vscale x 1 x i8> @test_vlseg3ff_mask_nxv1i8_triscv.vector.tuple_nxv1i8_3t(target("riscv.vector.tuple", <vscale x 1 x i8>, 3) %val, ptr %base, i32 %vl, <vscale x 1 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg3ff_mask_nxv1i8_triscv.vector.tuple_nxv1i8_3t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e8, mf8, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
-; CHECK-NEXT:    vsetvli zero, a1, e8, mf8, ta, mu
 ; CHECK-NEXT:    vlseg3e8ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
@@ -299,10 +299,10 @@ entry:
 define <vscale x 2 x i8> @test_vlseg3ff_mask_nxv2i8_triscv.vector.tuple_nxv2i8_3t(target("riscv.vector.tuple", <vscale x 2 x i8>, 3) %val, ptr %base, i32 %vl, <vscale x 2 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg3ff_mask_nxv2i8_triscv.vector.tuple_nxv2i8_3t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e8, mf4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
-; CHECK-NEXT:    vsetvli zero, a1, e8, mf4, ta, mu
 ; CHECK-NEXT:    vlseg3e8ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
@@ -339,10 +339,10 @@ entry:
 define <vscale x 4 x i8> @test_vlseg3ff_mask_nxv4i8_triscv.vector.tuple_nxv4i8_3t(target("riscv.vector.tuple", <vscale x 4 x i8>, 3) %val, ptr %base, i32 %vl, <vscale x 4 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg3ff_mask_nxv4i8_triscv.vector.tuple_nxv4i8_3t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e8, mf2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
-; CHECK-NEXT:    vsetvli zero, a1, e8, mf2, ta, mu
 ; CHECK-NEXT:    vlseg3e8ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
@@ -379,10 +379,10 @@ entry:
 define <vscale x 8 x i8> @test_vlseg3ff_mask_nxv8i8_triscv.vector.tuple_nxv8i8_3t(target("riscv.vector.tuple", <vscale x 8 x i8>, 3) %val, ptr %base, i32 %vl, <vscale x 8 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg3ff_mask_nxv8i8_triscv.vector.tuple_nxv8i8_3t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e8, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
-; CHECK-NEXT:    vsetvli zero, a1, e8, m1, ta, mu
 ; CHECK-NEXT:    vlseg3e8ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
@@ -419,10 +419,10 @@ entry:
 define <vscale x 16 x i8> @test_vlseg3ff_mask_nxv16i8_triscv.vector.tuple_nxv16i8_3t(target("riscv.vector.tuple", <vscale x 16 x i8>, 3) %val, ptr %base, i32 %vl, <vscale x 16 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg3ff_mask_nxv16i8_triscv.vector.tuple_nxv16i8_3t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e8, m2, ta, mu
 ; CHECK-NEXT:    vmv2r.v v6, v8
 ; CHECK-NEXT:    vmv2r.v v8, v10
 ; CHECK-NEXT:    vmv2r.v v10, v12
-; CHECK-NEXT:    vsetvli zero, a1, e8, m2, ta, mu
 ; CHECK-NEXT:    vlseg3e8ff.v v6, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
@@ -459,11 +459,11 @@ entry:
 define <vscale x 1 x i8> @test_vlseg4ff_mask_nxv1i8_triscv.vector.tuple_nxv1i8_4t(target("riscv.vector.tuple", <vscale x 1 x i8>, 4) %val, ptr %base, i32 %vl, <vscale x 1 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg4ff_mask_nxv1i8_triscv.vector.tuple_nxv1i8_4t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e8, mf8, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
 ; CHECK-NEXT:    vmv1r.v v10, v11
-; CHECK-NEXT:    vsetvli zero, a1, e8, mf8, ta, mu
 ; CHECK-NEXT:    vlseg4e8ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
@@ -500,11 +500,11 @@ entry:
 define <vscale x 2 x i8> @test_vlseg4ff_mask_nxv2i8_triscv.vector.tuple_nxv2i8_4t(target("riscv.vector.tuple", <vscale x 2 x i8>, 4) %val, ptr %base, i32 %vl, <vscale x 2 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg4ff_mask_nxv2i8_triscv.vector.tuple_nxv2i8_4t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e8, mf4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
 ; CHECK-NEXT:    vmv1r.v v10, v11
-; CHECK-NEXT:    vsetvli zero, a1, e8, mf4, ta, mu
 ; CHECK-NEXT:    vlseg4e8ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
@@ -541,11 +541,11 @@ entry:
 define <vscale x 4 x i8> @test_vlseg4ff_mask_nxv4i8_triscv.vector.tuple_nxv4i8_4t(target("riscv.vector.tuple", <vscale x 4 x i8>, 4) %val, ptr %base, i32 %vl, <vscale x 4 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg4ff_mask_nxv4i8_triscv.vector.tuple_nxv4i8_4t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e8, mf2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
 ; CHECK-NEXT:    vmv1r.v v10, v11
-; CHECK-NEXT:    vsetvli zero, a1, e8, mf2, ta, mu
 ; CHECK-NEXT:    vlseg4e8ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
@@ -582,11 +582,11 @@ entry:
 define <vscale x 8 x i8> @test_vlseg4ff_mask_nxv8i8_triscv.vector.tuple_nxv8i8_4t(target("riscv.vector.tuple", <vscale x 8 x i8>, 4) %val, ptr %base, i32 %vl, <vscale x 8 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg4ff_mask_nxv8i8_triscv.vector.tuple_nxv8i8_4t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e8, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
 ; CHECK-NEXT:    vmv1r.v v10, v11
-; CHECK-NEXT:    vsetvli zero, a1, e8, m1, ta, mu
 ; CHECK-NEXT:    vlseg4e8ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
@@ -623,11 +623,11 @@ entry:
 define <vscale x 16 x i8> @test_vlseg4ff_mask_nxv16i8_triscv.vector.tuple_nxv16i8_4t(target("riscv.vector.tuple", <vscale x 16 x i8>, 4) %val, ptr %base, i32 %vl, <vscale x 16 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg4ff_mask_nxv16i8_triscv.vector.tuple_nxv16i8_4t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e8, m2, ta, mu
 ; CHECK-NEXT:    vmv2r.v v6, v8
 ; CHECK-NEXT:    vmv2r.v v8, v10
 ; CHECK-NEXT:    vmv2r.v v10, v12
 ; CHECK-NEXT:    vmv2r.v v12, v14
-; CHECK-NEXT:    vsetvli zero, a1, e8, m2, ta, mu
 ; CHECK-NEXT:    vlseg4e8ff.v v6, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
@@ -664,12 +664,12 @@ entry:
 define <vscale x 1 x i8> @test_vlseg5ff_mask_nxv1i8_triscv.vector.tuple_nxv1i8_5t(target("riscv.vector.tuple", <vscale x 1 x i8>, 5) %val, ptr %base, i32 %vl, <vscale x 1 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg5ff_mask_nxv1i8_triscv.vector.tuple_nxv1i8_5t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e8, mf8, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
 ; CHECK-NEXT:    vmv1r.v v10, v11
 ; CHECK-NEXT:    vmv1r.v v11, v12
-; CHECK-NEXT:    vsetvli zero, a1, e8, mf8, ta, mu
 ; CHECK-NEXT:    vlseg5e8ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
@@ -706,12 +706,12 @@ entry:
 define <vscale x 2 x i8> @test_vlseg5ff_mask_nxv2i8_triscv.vector.tuple_nxv2i8_5t(target("riscv.vector.tuple", <vscale x 2 x i8>, 5) %val, ptr %base, i32 %vl, <vscale x 2 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg5ff_mask_nxv2i8_triscv.vector.tuple_nxv2i8_5t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e8, mf4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
 ; CHECK-NEXT:    vmv1r.v v10, v11
 ; CHECK-NEXT:    vmv1r.v v11, v12
-; CHECK-NEXT:    vsetvli zero, a1, e8, mf4, ta, mu
 ; CHECK-NEXT:    vlseg5e8ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
@@ -748,12 +748,12 @@ entry:
 define <vscale x 4 x i8> @test_vlseg5ff_mask_nxv4i8_triscv.vector.tuple_nxv4i8_5t(target("riscv.vector.tuple", <vscale x 4 x i8>, 5) %val, ptr %base, i32 %vl, <vscale x 4 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg5ff_mask_nxv4i8_triscv.vector.tuple_nxv4i8_5t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e8, mf2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
 ; CHECK-NEXT:    vmv1r.v v10, v11
 ; CHECK-NEXT:    vmv1r.v v11, v12
-; CHECK-NEXT:    vsetvli zero, a1, e8, mf2, ta, mu
 ; CHECK-NEXT:    vlseg5e8ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
@@ -790,12 +790,12 @@ entry:
 define <vscale x 8 x i8> @test_vlseg5ff_mask_nxv8i8_triscv.vector.tuple_nxv8i8_5t(target("riscv.vector.tuple", <vscale x 8 x i8>, 5) %val, ptr %base, i32 %vl, <vscale x 8 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg5ff_mask_nxv8i8_triscv.vector.tuple_nxv8i8_5t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e8, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
 ; CHECK-NEXT:    vmv1r.v v10, v11
 ; CHECK-NEXT:    vmv1r.v v11, v12
-; CHECK-NEXT:    vsetvli zero, a1, e8, m1, ta, mu
 ; CHECK-NEXT:    vlseg5e8ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
@@ -832,13 +832,13 @@ entry:
 define <vscale x 1 x i8> @test_vlseg6ff_mask_nxv1i8_triscv.vector.tuple_nxv1i8_6t(target("riscv.vector.tuple", <vscale x 1 x i8>, 6) %val, ptr %base, i32 %vl, <vscale x 1 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg6ff_mask_nxv1i8_triscv.vector.tuple_nxv1i8_6t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e8, mf8, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
 ; CHECK-NEXT:    vmv1r.v v10, v11
 ; CHECK-NEXT:    vmv1r.v v11, v12
 ; CHECK-NEXT:    vmv1r.v v12, v13
-; CHECK-NEXT:    vsetvli zero, a1, e8, mf8, ta, mu
 ; CHECK-NEXT:    vlseg6e8ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
@@ -875,13 +875,13 @@ entry:
 define <vscale x 2 x i8> @test_vlseg6ff_mask_nxv2i8_triscv.vector.tuple_nxv2i8_6t(target("riscv.vector.tuple", <vscale x 2 x i8>, 6) %val, ptr %base, i32 %vl, <vscale x 2 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg6ff_mask_nxv2i8_triscv.vector.tuple_nxv2i8_6t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e8, mf4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
 ; CHECK-NEXT:    vmv1r.v v10, v11
 ; CHECK-NEXT:    vmv1r.v v11, v12
 ; CHECK-NEXT:    vmv1r.v v12, v13
-; CHECK-NEXT:    vsetvli zero, a1, e8, mf4, ta, mu
 ; CHECK-NEXT:    vlseg6e8ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
@@ -918,13 +918,13 @@ entry:
 define <vscale x 4 x i8> @test_vlseg6ff_mask_nxv4i8_triscv.vector.tuple_nxv4i8_6t(target("riscv.vector.tuple", <vscale x 4 x i8>, 6) %val, ptr %base, i32 %vl, <vscale x 4 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg6ff_mask_nxv4i8_triscv.vector.tuple_nxv4i8_6t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e8, mf2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
 ; CHECK-NEXT:    vmv1r.v v10, v11
 ; CHECK-NEXT:    vmv1r.v v11, v12
 ; CHECK-NEXT:    vmv1r.v v12, v13
-; CHECK-NEXT:    vsetvli zero, a1, e8, mf2, ta, mu
 ; CHECK-NEXT:    vlseg6e8ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
@@ -961,13 +961,13 @@ entry:
 define <vscale x 8 x i8> @test_vlseg6ff_mask_nxv8i8_triscv.vector.tuple_nxv8i8_6t(target("riscv.vector.tuple", <vscale x 8 x i8>, 6) %val, ptr %base, i32 %vl, <vscale x 8 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg6ff_mask_nxv8i8_triscv.vector.tuple_nxv8i8_6t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e8, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
 ; CHECK-NEXT:    vmv1r.v v10, v11
 ; CHECK-NEXT:    vmv1r.v v11, v12
 ; CHECK-NEXT:    vmv1r.v v12, v13
-; CHECK-NEXT:    vsetvli zero, a1, e8, m1, ta, mu
 ; CHECK-NEXT:    vlseg6e8ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
@@ -1004,6 +1004,7 @@ entry:
 define <vscale x 1 x i8> @test_vlseg7ff_mask_nxv1i8_triscv.vector.tuple_nxv1i8_7t(target("riscv.vector.tuple", <vscale x 1 x i8>, 7) %val, ptr %base, i32 %vl, <vscale x 1 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg7ff_mask_nxv1i8_triscv.vector.tuple_nxv1i8_7t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e8, mf8, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
@@ -1011,7 +1012,6 @@ define <vscale x 1 x i8> @test_vlseg7ff_mask_nxv1i8_triscv.vector.tuple_nxv1i8_7
 ; CHECK-NEXT:    vmv1r.v v11, v12
 ; CHECK-NEXT:    vmv1r.v v12, v13
 ; CHECK-NEXT:    vmv1r.v v13, v14
-; CHECK-NEXT:    vsetvli zero, a1, e8, mf8, ta, mu
 ; CHECK-NEXT:    vlseg7e8ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
@@ -1048,6 +1048,7 @@ entry:
 define <vscale x 2 x i8> @test_vlseg7ff_mask_nxv2i8_triscv.vector.tuple_nxv2i8_7t(target("riscv.vector.tuple", <vscale x 2 x i8>, 7) %val, ptr %base, i32 %vl, <vscale x 2 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg7ff_mask_nxv2i8_triscv.vector.tuple_nxv2i8_7t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e8, mf4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
@@ -1055,7 +1056,6 @@ define <vscale x 2 x i8> @test_vlseg7ff_mask_nxv2i8_triscv.vector.tuple_nxv2i8_7
 ; CHECK-NEXT:    vmv1r.v v11, v12
 ; CHECK-NEXT:    vmv1r.v v12, v13
 ; CHECK-NEXT:    vmv1r.v v13, v14
-; CHECK-NEXT:    vsetvli zero, a1, e8, mf4, ta, mu
 ; CHECK-NEXT:    vlseg7e8ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
@@ -1092,6 +1092,7 @@ entry:
 define <vscale x 4 x i8> @test_vlseg7ff_mask_nxv4i8_triscv.vector.tuple_nxv4i8_7t(target("riscv.vector.tuple", <vscale x 4 x i8>, 7) %val, ptr %base, i32 %vl, <vscale x 4 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg7ff_mask_nxv4i8_triscv.vector.tuple_nxv4i8_7t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e8, mf2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
@@ -1099,7 +1100,6 @@ define <vscale x 4 x i8> @test_vlseg7ff_mask_nxv4i8_triscv.vector.tuple_nxv4i8_7
 ; CHECK-NEXT:    vmv1r.v v11, v12
 ; CHECK-NEXT:    vmv1r.v v12, v13
 ; CHECK-NEXT:    vmv1r.v v13, v14
-; CHECK-NEXT:    vsetvli zero, a1, e8, mf2, ta, mu
 ; CHECK-NEXT:    vlseg7e8ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
@@ -1136,6 +1136,7 @@ entry:
 define <vscale x 8 x i8> @test_vlseg7ff_mask_nxv8i8_triscv.vector.tuple_nxv8i8_7t(target("riscv.vector.tuple", <vscale x 8 x i8>, 7) %val, ptr %base, i32 %vl, <vscale x 8 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg7ff_mask_nxv8i8_triscv.vector.tuple_nxv8i8_7t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e8, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
@@ -1143,7 +1144,6 @@ define <vscale x 8 x i8> @test_vlseg7ff_mask_nxv8i8_triscv.vector.tuple_nxv8i8_7
 ; CHECK-NEXT:    vmv1r.v v11, v12
 ; CHECK-NEXT:    vmv1r.v v12, v13
 ; CHECK-NEXT:    vmv1r.v v13, v14
-; CHECK-NEXT:    vsetvli zero, a1, e8, m1, ta, mu
 ; CHECK-NEXT:    vlseg7e8ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
@@ -1180,6 +1180,7 @@ entry:
 define <vscale x 1 x i8> @test_vlseg8ff_mask_nxv1i8_triscv.vector.tuple_nxv1i8_8t(target("riscv.vector.tuple", <vscale x 1 x i8>, 8) %val, ptr %base, i32 %vl, <vscale x 1 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg8ff_mask_nxv1i8_triscv.vector.tuple_nxv1i8_8t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e8, mf8, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
@@ -1188,7 +1189,6 @@ define <vscale x 1 x i8> @test_vlseg8ff_mask_nxv1i8_triscv.vector.tuple_nxv1i8_8
 ; CHECK-NEXT:    vmv1r.v v12, v13
 ; CHECK-NEXT:    vmv1r.v v13, v14
 ; CHECK-NEXT:    vmv1r.v v14, v15
-; CHECK-NEXT:    vsetvli zero, a1, e8, mf8, ta, mu
 ; CHECK-NEXT:    vlseg8e8ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
@@ -1225,6 +1225,7 @@ entry:
 define <vscale x 2 x i8> @test_vlseg8ff_mask_nxv2i8_triscv.vector.tuple_nxv2i8_8t(target("riscv.vector.tuple", <vscale x 2 x i8>, 8) %val, ptr %base, i32 %vl, <vscale x 2 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg8ff_mask_nxv2i8_triscv.vector.tuple_nxv2i8_8t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e8, mf4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
@@ -1233,7 +1234,6 @@ define <vscale x 2 x i8> @test_vlseg8ff_mask_nxv2i8_triscv.vector.tuple_nxv2i8_8
 ; CHECK-NEXT:    vmv1r.v v12, v13
 ; CHECK-NEXT:    vmv1r.v v13, v14
 ; CHECK-NEXT:    vmv1r.v v14, v15
-; CHECK-NEXT:    vsetvli zero, a1, e8, mf4, ta, mu
 ; CHECK-NEXT:    vlseg8e8ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
@@ -1270,6 +1270,7 @@ entry:
 define <vscale x 4 x i8> @test_vlseg8ff_mask_nxv4i8_triscv.vector.tuple_nxv4i8_8t(target("riscv.vector.tuple", <vscale x 4 x i8>, 8) %val, ptr %base, i32 %vl, <vscale x 4 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg8ff_mask_nxv4i8_triscv.vector.tuple_nxv4i8_8t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e8, mf2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
@@ -1278,7 +1279,6 @@ define <vscale x 4 x i8> @test_vlseg8ff_mask_nxv4i8_triscv.vector.tuple_nxv4i8_8
 ; CHECK-NEXT:    vmv1r.v v12, v13
 ; CHECK-NEXT:    vmv1r.v v13, v14
 ; CHECK-NEXT:    vmv1r.v v14, v15
-; CHECK-NEXT:    vsetvli zero, a1, e8, mf2, ta, mu
 ; CHECK-NEXT:    vlseg8e8ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
@@ -1315,6 +1315,7 @@ entry:
 define <vscale x 8 x i8> @test_vlseg8ff_mask_nxv8i8_triscv.vector.tuple_nxv8i8_8t(target("riscv.vector.tuple", <vscale x 8 x i8>, 8) %val, ptr %base, i32 %vl, <vscale x 8 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg8ff_mask_nxv8i8_triscv.vector.tuple_nxv8i8_8t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e8, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
@@ -1323,7 +1324,6 @@ define <vscale x 8 x i8> @test_vlseg8ff_mask_nxv8i8_triscv.vector.tuple_nxv8i8_8
 ; CHECK-NEXT:    vmv1r.v v12, v13
 ; CHECK-NEXT:    vmv1r.v v13, v14
 ; CHECK-NEXT:    vmv1r.v v14, v15
-; CHECK-NEXT:    vsetvli zero, a1, e8, m1, ta, mu
 ; CHECK-NEXT:    vlseg8e8ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
@@ -1359,9 +1359,9 @@ entry:
 define <vscale x 1 x i16> @test_vlseg2ff_mask_nxv1i16_triscv.vector.tuple_nxv2i8_2t(target("riscv.vector.tuple", <vscale x 2 x i8>, 2) %val, ptr %base, i32 %vl, <vscale x 1 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg2ff_mask_nxv1i16_triscv.vector.tuple_nxv2i8_2t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, mf4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
-; CHECK-NEXT:    vsetvli zero, a1, e16, mf4, ta, mu
 ; CHECK-NEXT:    vlseg2e16ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
@@ -1397,9 +1397,9 @@ entry:
 define <vscale x 2 x i16> @test_vlseg2ff_mask_nxv2i16_triscv.vector.tuple_nxv4i8_2t(target("riscv.vector.tuple", <vscale x 4 x i8>, 2) %val, ptr %base, i32 %vl, <vscale x 2 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg2ff_mask_nxv2i16_triscv.vector.tuple_nxv4i8_2t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, mf2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
-; CHECK-NEXT:    vsetvli zero, a1, e16, mf2, ta, mu
 ; CHECK-NEXT:    vlseg2e16ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
@@ -1435,9 +1435,9 @@ entry:
 define <vscale x 4 x i16> @test_vlseg2ff_mask_nxv4i16_triscv.vector.tuple_nxv8i8_2t(target("riscv.vector.tuple", <vscale x 8 x i8>, 2) %val, ptr %base, i32 %vl, <vscale x 4 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg2ff_mask_nxv4i16_triscv.vector.tuple_nxv8i8_2t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
-; CHECK-NEXT:    vsetvli zero, a1, e16, m1, ta, mu
 ; CHECK-NEXT:    vlseg2e16ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
@@ -1473,9 +1473,9 @@ entry:
 define <vscale x 8 x i16> @test_vlseg2ff_mask_nxv8i16_triscv.vector.tuple_nxv16i8_2t(target("riscv.vector.tuple", <vscale x 16 x i8>, 2) %val, ptr %base, i32 %vl, <vscale x 8 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg2ff_mask_nxv8i16_triscv.vector.tuple_nxv16i8_2t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, m2, ta, mu
 ; CHECK-NEXT:    vmv2r.v v6, v8
 ; CHECK-NEXT:    vmv2r.v v8, v10
-; CHECK-NEXT:    vsetvli zero, a1, e16, m2, ta, mu
 ; CHECK-NEXT:    vlseg2e16ff.v v6, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
@@ -1511,9 +1511,9 @@ entry:
 define <vscale x 16 x i16> @test_vlseg2ff_mask_nxv16i16_triscv.vector.tuple_nxv32i8_2t(target("riscv.vector.tuple", <vscale x 32 x i8>, 2) %val, ptr %base, i32 %vl, <vscale x 16 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg2ff_mask_nxv16i16_triscv.vector.tuple_nxv32i8_2t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, m4, ta, mu
 ; CHECK-NEXT:    vmv4r.v v4, v8
 ; CHECK-NEXT:    vmv4r.v v8, v12
-; CHECK-NEXT:    vsetvli zero, a1, e16, m4, ta, mu
 ; CHECK-NEXT:    vlseg2e16ff.v v4, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
@@ -1549,10 +1549,10 @@ entry:
 define <vscale x 1 x i16> @test_vlseg3ff_mask_nxv1i16_triscv.vector.tuple_nxv2i8_3t(target("riscv.vector.tuple", <vscale x 2 x i8>, 3) %val, ptr %base, i32 %vl, <vscale x 1 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg3ff_mask_nxv1i16_triscv.vector.tuple_nxv2i8_3t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, mf4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
-; CHECK-NEXT:    vsetvli zero, a1, e16, mf4, ta, mu
 ; CHECK-NEXT:    vlseg3e16ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
@@ -1588,10 +1588,10 @@ entry:
 define <vscale x 2 x i16> @test_vlseg3ff_mask_nxv2i16_triscv.vector.tuple_nxv4i8_3t(target("riscv.vector.tuple", <vscale x 4 x i8>, 3) %val, ptr %base, i32 %vl, <vscale x 2 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg3ff_mask_nxv2i16_triscv.vector.tuple_nxv4i8_3t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, mf2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
-; CHECK-NEXT:    vsetvli zero, a1, e16, mf2, ta, mu
 ; CHECK-NEXT:    vlseg3e16ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
@@ -1627,10 +1627,10 @@ entry:
 define <vscale x 4 x i16> @test_vlseg3ff_mask_nxv4i16_triscv.vector.tuple_nxv8i8_3t(target("riscv.vector.tuple", <vscale x 8 x i8>, 3) %val, ptr %base, i32 %vl, <vscale x 4 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg3ff_mask_nxv4i16_triscv.vector.tuple_nxv8i8_3t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
-; CHECK-NEXT:    vsetvli zero, a1, e16, m1, ta, mu
 ; CHECK-NEXT:    vlseg3e16ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
@@ -1666,10 +1666,10 @@ entry:
 define <vscale x 8 x i16> @test_vlseg3ff_mask_nxv8i16_triscv.vector.tuple_nxv16i8_3t(target("riscv.vector.tuple", <vscale x 16 x i8>, 3) %val, ptr %base, i32 %vl, <vscale x 8 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg3ff_mask_nxv8i16_triscv.vector.tuple_nxv16i8_3t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, m2, ta, mu
 ; CHECK-NEXT:    vmv2r.v v6, v8
 ; CHECK-NEXT:    vmv2r.v v8, v10
 ; CHECK-NEXT:    vmv2r.v v10, v12
-; CHECK-NEXT:    vsetvli zero, a1, e16, m2, ta, mu
 ; CHECK-NEXT:    vlseg3e16ff.v v6, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
@@ -1705,11 +1705,11 @@ entry:
 define <vscale x 1 x i16> @test_vlseg4ff_mask_nxv1i16_triscv.vector.tuple_nxv2i8_4t(target("riscv.vector.tuple", <vscale x 2 x i8>, 4) %val, ptr %base, i32 %vl, <vscale x 1 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg4ff_mask_nxv1i16_triscv.vector.tuple_nxv2i8_4t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, mf4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
 ; CHECK-NEXT:    vmv1r.v v10, v11
-; CHECK-NEXT:    vsetvli zero, a1, e16, mf4, ta, mu
 ; CHECK-NEXT:    vlseg4e16ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
@@ -1745,11 +1745,11 @@ entry:
 define <vscale x 2 x i16> @test_vlseg4ff_mask_nxv2i16_triscv.vector.tuple_nxv4i8_4t(target("riscv.vector.tuple", <vscale x 4 x i8>, 4) %val, ptr %base, i32 %vl, <vscale x 2 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg4ff_mask_nxv2i16_triscv.vector.tuple_nxv4i8_4t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, mf2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
 ; CHECK-NEXT:    vmv1r.v v10, v11
-; CHECK-NEXT:    vsetvli zero, a1, e16, mf2, ta, mu
 ; CHECK-NEXT:    vlseg4e16ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
@@ -1785,11 +1785,11 @@ entry:
 define <vscale x 4 x i16> @test_vlseg4ff_mask_nxv4i16_triscv.vector.tuple_nxv8i8_4t(target("riscv.vector.tuple", <vscale x 8 x i8>, 4) %val, ptr %base, i32 %vl, <vscale x 4 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg4ff_mask_nxv4i16_triscv.vector.tuple_nxv8i8_4t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
 ; CHECK-NEXT:    vmv1r.v v10, v11
-; CHECK-NEXT:    vsetvli zero, a1, e16, m1, ta, mu
 ; CHECK-NEXT:    vlseg4e16ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
@@ -1825,11 +1825,11 @@ entry:
 define <vscale x 8 x i16> @test_vlseg4ff_mask_nxv8i16_triscv.vector.tuple_nxv16i8_4t(target("riscv.vector.tuple", <vscale x 16 x i8>, 4) %val, ptr %base, i32 %vl, <vscale x 8 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg4ff_mask_nxv8i16_triscv.vector.tuple_nxv16i8_4t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, m2, ta, mu
 ; CHECK-NEXT:    vmv2r.v v6, v8
 ; CHECK-NEXT:    vmv2r.v v8, v10
 ; CHECK-NEXT:    vmv2r.v v10, v12
 ; CHECK-NEXT:    vmv2r.v v12, v14
-; CHECK-NEXT:    vsetvli zero, a1, e16, m2, ta, mu
 ; CHECK-NEXT:    vlseg4e16ff.v v6, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
@@ -1865,12 +1865,12 @@ entry:
 define <vscale x 1 x i16> @test_vlseg5ff_mask_nxv1i16_triscv.vector.tuple_nxv2i8_5t(target("riscv.vector.tuple", <vscale x 2 x i8>, 5) %val, ptr %base, i32 %vl, <vscale x 1 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg5ff_mask_nxv1i16_triscv.vector.tuple_nxv2i8_5t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, mf4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
 ; CHECK-NEXT:    vmv1r.v v10, v11
 ; CHECK-NEXT:    vmv1r.v v11, v12
-; CHECK-NEXT:    vsetvli zero, a1, e16, mf4, ta, mu
 ; CHECK-NEXT:    vlseg5e16ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
@@ -1906,12 +1906,12 @@ entry:
 define <vscale x 2 x i16> @test_vlseg5ff_mask_nxv2i16_triscv.vector.tuple_nxv4i8_5t(target("riscv.vector.tuple", <vscale x 4 x i8>, 5) %val, ptr %base, i32 %vl, <vscale x 2 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg5ff_mask_nxv2i16_triscv.vector.tuple_nxv4i8_5t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, mf2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
 ; CHECK-NEXT:    vmv1r.v v10, v11
 ; CHECK-NEXT:    vmv1r.v v11, v12
-; CHECK-NEXT:    vsetvli zero, a1, e16, mf2, ta, mu
 ; CHECK-NEXT:    vlseg5e16ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
@@ -1947,12 +1947,12 @@ entry:
 define <vscale x 4 x i16> @test_vlseg5ff_mask_nxv4i16_triscv.vector.tuple_nxv8i8_5t(target("riscv.vector.tuple", <vscale x 8 x i8>, 5) %val, ptr %base, i32 %vl, <vscale x 4 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg5ff_mask_nxv4i16_triscv.vector.tuple_nxv8i8_5t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
 ; CHECK-NEXT:    vmv1r.v v10, v11
 ; CHECK-NEXT:    vmv1r.v v11, v12
-; CHECK-NEXT:    vsetvli zero, a1, e16, m1, ta, mu
 ; CHECK-NEXT:    vlseg5e16ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
@@ -1988,13 +1988,13 @@ entry:
 define <vscale x 1 x i16> @test_vlseg6ff_mask_nxv1i16_triscv.vector.tuple_nxv2i8_6t(target("riscv.vector.tuple", <vscale x 2 x i8>, 6) %val, ptr %base, i32 %vl, <vscale x 1 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg6ff_mask_nxv1i16_triscv.vector.tuple_nxv2i8_6t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, mf4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
 ; CHECK-NEXT:    vmv1r.v v10, v11
 ; CHECK-NEXT:    vmv1r.v v11, v12
 ; CHECK-NEXT:    vmv1r.v v12, v13
-; CHECK-NEXT:    vsetvli zero, a1, e16, mf4, ta, mu
 ; CHECK-NEXT:    vlseg6e16ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
@@ -2030,13 +2030,13 @@ entry:
 define <vscale x 2 x i16> @test_vlseg6ff_mask_nxv2i16_triscv.vector.tuple_nxv4i8_6t(target("riscv.vector.tuple", <vscale x 4 x i8>, 6) %val, ptr %base, i32 %vl, <vscale x 2 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg6ff_mask_nxv2i16_triscv.vector.tuple_nxv4i8_6t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, mf2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
 ; CHECK-NEXT:    vmv1r.v v10, v11
 ; CHECK-NEXT:    vmv1r.v v11, v12
 ; CHECK-NEXT:    vmv1r.v v12, v13
-; CHECK-NEXT:    vsetvli zero, a1, e16, mf2, ta, mu
 ; CHECK-NEXT:    vlseg6e16ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
@@ -2072,13 +2072,13 @@ entry:
 define <vscale x 4 x i16> @test_vlseg6ff_mask_nxv4i16_triscv.vector.tuple_nxv8i8_6t(target("riscv.vector.tuple", <vscale x 8 x i8>, 6) %val, ptr %base, i32 %vl, <vscale x 4 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg6ff_mask_nxv4i16_triscv.vector.tuple_nxv8i8_6t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
 ; CHECK-NEXT:    vmv1r.v v10, v11
 ; CHECK-NEXT:    vmv1r.v v11, v12
 ; CHECK-NEXT:    vmv1r.v v12, v13
-; CHECK-NEXT:    vsetvli zero, a1, e16, m1, ta, mu
 ; CHECK-NEXT:    vlseg6e16ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
@@ -2114,6 +2114,7 @@ entry:
 define <vscale x 1 x i16> @test_vlseg7ff_mask_nxv1i16_triscv.vector.tuple_nxv2i8_7t(target("riscv.vector.tuple", <vscale x 2 x i8>, 7) %val, ptr %base, i32 %vl, <vscale x 1 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg7ff_mask_nxv1i16_triscv.vector.tuple_nxv2i8_7t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, mf4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
@@ -2121,7 +2122,6 @@ define <vscale x 1 x i16> @test_vlseg7ff_mask_nxv1i16_triscv.vector.tuple_nxv2i8
 ; CHECK-NEXT:    vmv1r.v v11, v12
 ; CHECK-NEXT:    vmv1r.v v12, v13
 ; CHECK-NEXT:    vmv1r.v v13, v14
-; CHECK-NEXT:    vsetvli zero, a1, e16, mf4, ta, mu
 ; CHECK-NEXT:    vlseg7e16ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
@@ -2157,6 +2157,7 @@ entry:
 define <vscale x 2 x i16> @test_vlseg7ff_mask_nxv2i16_triscv.vector.tuple_nxv4i8_7t(target("riscv.vector.tuple", <vscale x 4 x i8>, 7) %val, ptr %base, i32 %vl, <vscale x 2 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg7ff_mask_nxv2i16_triscv.vector.tuple_nxv4i8_7t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, mf2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
@@ -2164,7 +2165,6 @@ define <vscale x 2 x i16> @test_vlseg7ff_mask_nxv2i16_triscv.vector.tuple_nxv4i8
 ; CHECK-NEXT:    vmv1r.v v11, v12
 ; CHECK-NEXT:    vmv1r.v v12, v13
 ; CHECK-NEXT:    vmv1r.v v13, v14
-; CHECK-NEXT:    vsetvli zero, a1, e16, mf2, ta, mu
 ; CHECK-NEXT:    vlseg7e16ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
@@ -2200,6 +2200,7 @@ entry:
 define <vscale x 4 x i16> @test_vlseg7ff_mask_nxv4i16_triscv.vector.tuple_nxv8i8_7t(target("riscv.vector.tuple", <vscale x 8 x i8>, 7) %val, ptr %base, i32 %vl, <vscale x 4 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg7ff_mask_nxv4i16_triscv.vector.tuple_nxv8i8_7t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
@@ -2207,7 +2208,6 @@ define <vscale x 4 x i16> @test_vlseg7ff_mask_nxv4i16_triscv.vector.tuple_nxv8i8
 ; CHECK-NEXT:    vmv1r.v v11, v12
 ; CHECK-NEXT:    vmv1r.v v12, v13
 ; CHECK-NEXT:    vmv1r.v v13, v14
-; CHECK-NEXT:    vsetvli zero, a1, e16, m1, ta, mu
 ; CHECK-NEXT:    vlseg7e16ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
@@ -2243,6 +2243,7 @@ entry:
 define <vscale x 1 x i16> @test_vlseg8ff_mask_nxv1i16_triscv.vector.tuple_nxv2i8_8t(target("riscv.vector.tuple", <vscale x 2 x i8>, 8) %val, ptr %base, i32 %vl, <vscale x 1 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg8ff_mask_nxv1i16_triscv.vector.tuple_nxv2i8_8t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, mf4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
@@ -2251,7 +2252,6 @@ define <vscale x 1 x i16> @test_vlseg8ff_mask_nxv1i16_triscv.vector.tuple_nxv2i8
 ; CHECK-NEXT:    vmv1r.v v12, v13
 ; CHECK-NEXT:    vmv1r.v v13, v14
 ; CHECK-NEXT:    vmv1r.v v14, v15
-; CHECK-NEXT:    vsetvli zero, a1, e16, mf4, ta, mu
 ; CHECK-NEXT:    vlseg8e16ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
@@ -2287,6 +2287,7 @@ entry:
 define <vscale x 2 x i16> @test_vlseg8ff_mask_nxv2i16_triscv.vector.tuple_nxv4i8_8t(target("riscv.vector.tuple", <vscale x 4 x i8>, 8) %val, ptr %base, i32 %vl, <vscale x 2 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg8ff_mask_nxv2i16_triscv.vector.tuple_nxv4i8_8t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, mf2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
@@ -2295,7 +2296,6 @@ define <vscale x 2 x i16> @test_vlseg8ff_mask_nxv2i16_triscv.vector.tuple_nxv4i8
 ; CHECK-NEXT:    vmv1r.v v12, v13
 ; CHECK-NEXT:    vmv1r.v v13, v14
 ; CHECK-NEXT:    vmv1r.v v14, v15
-; CHECK-NEXT:    vsetvli zero, a1, e16, mf2, ta, mu
 ; CHECK-NEXT:    vlseg8e16ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
@@ -2331,6 +2331,7 @@ entry:
 define <vscale x 4 x i16> @test_vlseg8ff_mask_nxv4i16_triscv.vector.tuple_nxv8i8_8t(target("riscv.vector.tuple", <vscale x 8 x i8>, 8) %val, ptr %base, i32 %vl, <vscale x 4 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg8ff_mask_nxv4i16_triscv.vector.tuple_nxv8i8_8t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
@@ -2339,7 +2340,6 @@ define <vscale x 4 x i16> @test_vlseg8ff_mask_nxv4i16_triscv.vector.tuple_nxv8i8
 ; CHECK-NEXT:    vmv1r.v v12, v13
 ; CHECK-NEXT:    vmv1r.v v13, v14
 ; CHECK-NEXT:    vmv1r.v v14, v15
-; CHECK-NEXT:    vsetvli zero, a1, e16, m1, ta, mu
 ; CHECK-NEXT:    vlseg8e16ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
@@ -2375,9 +2375,9 @@ entry:
 define <vscale x 1 x i32> @test_vlseg2ff_mask_nxv1i32_triscv.vector.tuple_nxv4i8_2t(target("riscv.vector.tuple", <vscale x 4 x i8>, 2) %val, ptr %base, i32 %vl, <vscale x 1 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg2ff_mask_nxv1i32_triscv.vector.tuple_nxv4i8_2t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e32, mf2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
-; CHECK-NEXT:    vsetvli zero, a1, e32, mf2, ta, mu
 ; CHECK-NEXT:    vlseg2e32ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
@@ -2413,9 +2413,9 @@ entry:
 define <vscale x 2 x i32> @test_vlseg2ff_mask_nxv2i32_triscv.vector.tuple_nxv8i8_2t(target("riscv.vector.tuple", <vscale x 8 x i8>, 2) %val, ptr %base, i32 %vl, <vscale x 2 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg2ff_mask_nxv2i32_triscv.vector.tuple_nxv8i8_2t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e32, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
-; CHECK-NEXT:    vsetvli zero, a1, e32, m1, ta, mu
 ; CHECK-NEXT:    vlseg2e32ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
@@ -2451,9 +2451,9 @@ entry:
 define <vscale x 4 x i32> @test_vlseg2ff_mask_nxv4i32_triscv.vector.tuple_nxv16i8_2t(target("riscv.vector.tuple", <vscale x 16 x i8>, 2) %val, ptr %base, i32 %vl, <vscale x 4 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg2ff_mask_nxv4i32_triscv.vector.tuple_nxv16i8_2t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e32, m2, ta, mu
 ; CHECK-NEXT:    vmv2r.v v6, v8
 ; CHECK-NEXT:    vmv2r.v v8, v10
-; CHECK-NEXT:    vsetvli zero, a1, e32, m2, ta, mu
 ; CHECK-NEXT:    vlseg2e32ff.v v6, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
@@ -2489,9 +2489,9 @@ entry:
 define <vscale x 8 x i32> @test_vlseg2ff_mask_nxv8i32_triscv.vector.tuple_nxv32i8_2t(target("riscv.vector.tuple", <vscale x 32 x i8>, 2) %val, ptr %base, i32 %vl, <vscale x 8 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg2ff_mask_nxv8i32_triscv.vector.tuple_nxv32i8_2t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e32, m4, ta, mu
 ; CHECK-NEXT:    vmv4r.v v4, v8
 ; CHECK-NEXT:    vmv4r.v v8, v12
-; CHECK-NEXT:    vsetvli zero, a1, e32, m4, ta, mu
 ; CHECK-NEXT:    vlseg2e32ff.v v4, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
@@ -2527,10 +2527,10 @@ entry:
 define <vscale x 1 x i32> @test_vlseg3ff_mask_nxv1i32_triscv.vector.tuple_nxv4i8_3t(target("riscv.vector.tuple", <vscale x 4 x i8>, 3) %val, ptr %base, i32 %vl, <vscale x 1 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg3ff_mask_nxv1i32_triscv.vector.tuple_nxv4i8_3t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e32, mf2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
-; CHECK-NEXT:    vsetvli zero, a1, e32, mf2, ta, mu
 ; CHECK-NEXT:    vlseg3e32ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
@@ -2566,10 +2566,10 @@ entry:
 define <vscale x 2 x i32> @test_vlseg3ff_mask_nxv2i32_triscv.vector.tuple_nxv8i8_3t(target("riscv.vector.tuple", <vscale x 8 x i8>, 3) %val, ptr %base, i32 %vl, <vscale x 2 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg3ff_mask_nxv2i32_triscv.vector.tuple_nxv8i8_3t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e32, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
-; CHECK-NEXT:    vsetvli zero, a1, e32, m1, ta, mu
 ; CHECK-NEXT:    vlseg3e32ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
@@ -2605,10 +2605,10 @@ entry:
 define <vscale x 4 x i32> @test_vlseg3ff_mask_nxv4i32_triscv.vector.tuple_nxv16i8_3t(target("riscv.vector.tuple", <vscale x 16 x i8>, 3) %val, ptr %base, i32 %vl, <vscale x 4 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg3ff_mask_nxv4i32_triscv.vector.tuple_nxv16i8_3t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e32, m2, ta, mu
 ; CHECK-NEXT:    vmv2r.v v6, v8
 ; CHECK-NEXT:    vmv2r.v v8, v10
 ; CHECK-NEXT:    vmv2r.v v10, v12
-; CHECK-NEXT:    vsetvli zero, a1, e32, m2, ta, mu
 ; CHECK-NEXT:    vlseg3e32ff.v v6, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
@@ -2644,11 +2644,11 @@ entry:
 define <vscale x 1 x i32> @test_vlseg4ff_mask_nxv1i32_triscv.vector.tuple_nxv4i8_4t(target("riscv.vector.tuple", <vscale x 4 x i8>, 4) %val, ptr %base, i32 %vl, <vscale x 1 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg4ff_mask_nxv1i32_triscv.vector.tuple_nxv4i8_4t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e32, mf2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
 ; CHECK-NEXT:    vmv1r.v v10, v11
-; CHECK-NEXT:    vsetvli zero, a1, e32, mf2, ta, mu
 ; CHECK-NEXT:    vlseg4e32ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
@@ -2684,11 +2684,11 @@ entry:
 define <vscale x 2 x i32> @test_vlseg4ff_mask_nxv2i32_triscv.vector.tuple_nxv8i8_4t(target("riscv.vector.tuple", <vscale x 8 x i8>, 4) %val, ptr %base, i32 %vl, <vscale x 2 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg4ff_mask_nxv2i32_triscv.vector.tuple_nxv8i8_4t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e32, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
 ; CHECK-NEXT:    vmv1r.v v10, v11
-; CHECK-NEXT:    vsetvli zero, a1, e32, m1, ta, mu
 ; CHECK-NEXT:    vlseg4e32ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
@@ -2724,11 +2724,11 @@ entry:
 define <vscale x 4 x i32> @test_vlseg4ff_mask_nxv4i32_triscv.vector.tuple_nxv16i8_4t(target("riscv.vector.tuple", <vscale x 16 x i8>, 4) %val, ptr %base, i32 %vl, <vscale x 4 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg4ff_mask_nxv4i32_triscv.vector.tuple_nxv16i8_4t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e32, m2, ta, mu
 ; CHECK-NEXT:    vmv2r.v v6, v8
 ; CHECK-NEXT:    vmv2r.v v8, v10
 ; CHECK-NEXT:    vmv2r.v v10, v12
 ; CHECK-NEXT:    vmv2r.v v12, v14
-; CHECK-NEXT:    vsetvli zero, a1, e32, m2, ta, mu
 ; CHECK-NEXT:    vlseg4e32ff.v v6, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
@@ -2764,12 +2764,12 @@ entry:
 define <vscale x 1 x i32> @test_vlseg5ff_mask_nxv1i32_triscv.vector.tuple_nxv4i8_5t(target("riscv.vector.tuple", <vscale x 4 x i8>, 5) %val, ptr %base, i32 %vl, <vscale x 1 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg5ff_mask_nxv1i32_triscv.vector.tuple_nxv4i8_5t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e32, mf2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
 ; CHECK-NEXT:    vmv1r.v v10, v11
 ; CHECK-NEXT:    vmv1r.v v11, v12
-; CHECK-NEXT:    vsetvli zero, a1, e32, mf2, ta, mu
 ; CHECK-NEXT:    vlseg5e32ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
@@ -2805,12 +2805,12 @@ entry:
 define <vscale x 2 x i32> @test_vlseg5ff_mask_nxv2i32_triscv.vector.tuple_nxv8i8_5t(target("riscv.vector.tuple", <vscale x 8 x i8>, 5) %val, ptr %base, i32 %vl, <vscale x 2 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg5ff_mask_nxv2i32_triscv.vector.tuple_nxv8i8_5t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e32, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
 ; CHECK-NEXT:    vmv1r.v v10, v11
 ; CHECK-NEXT:    vmv1r.v v11, v12
-; CHECK-NEXT:    vsetvli zero, a1, e32, m1, ta, mu
 ; CHECK-NEXT:    vlseg5e32ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
@@ -2846,13 +2846,13 @@ entry:
 define <vscale x 1 x i32> @test_vlseg6ff_mask_nxv1i32_triscv.vector.tuple_nxv4i8_6t(target("riscv.vector.tuple", <vscale x 4 x i8>, 6) %val, ptr %base, i32 %vl, <vscale x 1 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg6ff_mask_nxv1i32_triscv.vector.tuple_nxv4i8_6t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e32, mf2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
 ; CHECK-NEXT:    vmv1r.v v10, v11
 ; CHECK-NEXT:    vmv1r.v v11, v12
 ; CHECK-NEXT:    vmv1r.v v12, v13
-; CHECK-NEXT:    vsetvli zero, a1, e32, mf2, ta, mu
 ; CHECK-NEXT:    vlseg6e32ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
@@ -2888,13 +2888,13 @@ entry:
 define <vscale x 2 x i32> @test_vlseg6ff_mask_nxv2i32_triscv.vector.tuple_nxv8i8_6t(target("riscv.vector.tuple", <vscale x 8 x i8>, 6) %val, ptr %base, i32 %vl, <vscale x 2 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg6ff_mask_nxv2i32_triscv.vector.tuple_nxv8i8_6t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e32, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
 ; CHECK-NEXT:    vmv1r.v v10, v11
 ; CHECK-NEXT:    vmv1r.v v11, v12
 ; CHECK-NEXT:    vmv1r.v v12, v13
-; CHECK-NEXT:    vsetvli zero, a1, e32, m1, ta, mu
 ; CHECK-NEXT:    vlseg6e32ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
@@ -2930,6 +2930,7 @@ entry:
 define <vscale x 1 x i32> @test_vlseg7ff_mask_nxv1i32_triscv.vector.tuple_nxv4i8_7t(target("riscv.vector.tuple", <vscale x 4 x i8>, 7) %val, ptr %base, i32 %vl, <vscale x 1 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg7ff_mask_nxv1i32_triscv.vector.tuple_nxv4i8_7t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e32, mf2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
@@ -2937,7 +2938,6 @@ define <vscale x 1 x i32> @test_vlseg7ff_mask_nxv1i32_triscv.vector.tuple_nxv4i8
 ; CHECK-NEXT:    vmv1r.v v11, v12
 ; CHECK-NEXT:    vmv1r.v v12, v13
 ; CHECK-NEXT:    vmv1r.v v13, v14
-; CHECK-NEXT:    vsetvli zero, a1, e32, mf2, ta, mu
 ; CHECK-NEXT:    vlseg7e32ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
@@ -2973,6 +2973,7 @@ entry:
 define <vscale x 2 x i32> @test_vlseg7ff_mask_nxv2i32_triscv.vector.tuple_nxv8i8_7t(target("riscv.vector.tuple", <vscale x 8 x i8>, 7) %val, ptr %base, i32 %vl, <vscale x 2 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg7ff_mask_nxv2i32_triscv.vector.tuple_nxv8i8_7t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e32, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
@@ -2980,7 +2981,6 @@ define <vscale x 2 x i32> @test_vlseg7ff_mask_nxv2i32_triscv.vector.tuple_nxv8i8
 ; CHECK-NEXT:    vmv1r.v v11, v12
 ; CHECK-NEXT:    vmv1r.v v12, v13
 ; CHECK-NEXT:    vmv1r.v v13, v14
-; CHECK-NEXT:    vsetvli zero, a1, e32, m1, ta, mu
 ; CHECK-NEXT:    vlseg7e32ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
@@ -3016,6 +3016,7 @@ entry:
 define <vscale x 1 x i32> @test_vlseg8ff_mask_nxv1i32_triscv.vector.tuple_nxv4i8_8t(target("riscv.vector.tuple", <vscale x 4 x i8>, 8) %val, ptr %base, i32 %vl, <vscale x 1 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg8ff_mask_nxv1i32_triscv.vector.tuple_nxv4i8_8t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e32, mf2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
@@ -3024,7 +3025,6 @@ define <vscale x 1 x i32> @test_vlseg8ff_mask_nxv1i32_triscv.vector.tuple_nxv4i8
 ; CHECK-NEXT:    vmv1r.v v12, v13
 ; CHECK-NEXT:    vmv1r.v v13, v14
 ; CHECK-NEXT:    vmv1r.v v14, v15
-; CHECK-NEXT:    vsetvli zero, a1, e32, mf2, ta, mu
 ; CHECK-NEXT:    vlseg8e32ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
@@ -3060,6 +3060,7 @@ entry:
 define <vscale x 2 x i32> @test_vlseg8ff_mask_nxv2i32_triscv.vector.tuple_nxv8i8_8t(target("riscv.vector.tuple", <vscale x 8 x i8>, 8) %val, ptr %base, i32 %vl, <vscale x 2 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg8ff_mask_nxv2i32_triscv.vector.tuple_nxv8i8_8t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e32, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
@@ -3068,7 +3069,6 @@ define <vscale x 2 x i32> @test_vlseg8ff_mask_nxv2i32_triscv.vector.tuple_nxv8i8
 ; CHECK-NEXT:    vmv1r.v v12, v13
 ; CHECK-NEXT:    vmv1r.v v13, v14
 ; CHECK-NEXT:    vmv1r.v v14, v15
-; CHECK-NEXT:    vsetvli zero, a1, e32, m1, ta, mu
 ; CHECK-NEXT:    vlseg8e32ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
@@ -3104,9 +3104,9 @@ entry:
 define <vscale x 1 x i64> @test_vlseg2ff_mask_nxv1i64_triscv.vector.tuple_nxv8i8_2t(target("riscv.vector.tuple", <vscale x 8 x i8>, 2) %val, ptr %base, i32 %vl, <vscale x 1 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg2ff_mask_nxv1i64_triscv.vector.tuple_nxv8i8_2t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e64, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
-; CHECK-NEXT:    vsetvli zero, a1, e64, m1, ta, mu
 ; CHECK-NEXT:    vlseg2e64ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
@@ -3142,9 +3142,9 @@ entry:
 define <vscale x 2 x i64> @test_vlseg2ff_mask_nxv2i64_triscv.vector.tuple_nxv16i8_2t(target("riscv.vector.tuple", <vscale x 16 x i8>, 2) %val, ptr %base, i32 %vl, <vscale x 2 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg2ff_mask_nxv2i64_triscv.vector.tuple_nxv16i8_2t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e64, m2, ta, mu
 ; CHECK-NEXT:    vmv2r.v v6, v8
 ; CHECK-NEXT:    vmv2r.v v8, v10
-; CHECK-NEXT:    vsetvli zero, a1, e64, m2, ta, mu
 ; CHECK-NEXT:    vlseg2e64ff.v v6, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
@@ -3180,9 +3180,9 @@ entry:
 define <vscale x 4 x i64> @test_vlseg2ff_mask_nxv4i64_triscv.vector.tuple_nxv32i8_2t(target("riscv.vector.tuple", <vscale x 32 x i8>, 2) %val, ptr %base, i32 %vl, <vscale x 4 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg2ff_mask_nxv4i64_triscv.vector.tuple_nxv32i8_2t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e64, m4, ta, mu
 ; CHECK-NEXT:    vmv4r.v v4, v8
 ; CHECK-NEXT:    vmv4r.v v8, v12
-; CHECK-NEXT:    vsetvli zero, a1, e64, m4, ta, mu
 ; CHECK-NEXT:    vlseg2e64ff.v v4, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
@@ -3218,10 +3218,10 @@ entry:
 define <vscale x 1 x i64> @test_vlseg3ff_mask_nxv1i64_triscv.vector.tuple_nxv8i8_3t(target("riscv.vector.tuple", <vscale x 8 x i8>, 3) %val, ptr %base, i32 %vl, <vscale x 1 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg3ff_mask_nxv1i64_triscv.vector.tuple_nxv8i8_3t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e64, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
-; CHECK-NEXT:    vsetvli zero, a1, e64, m1, ta, mu
 ; CHECK-NEXT:    vlseg3e64ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
@@ -3257,10 +3257,10 @@ entry:
 define <vscale x 2 x i64> @test_vlseg3ff_mask_nxv2i64_triscv.vector.tuple_nxv16i8_3t(target("riscv.vector.tuple", <vscale x 16 x i8>, 3) %val, ptr %base, i32 %vl, <vscale x 2 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg3ff_mask_nxv2i64_triscv.vector.tuple_nxv16i8_3t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e64, m2, ta, mu
 ; CHECK-NEXT:    vmv2r.v v6, v8
 ; CHECK-NEXT:    vmv2r.v v8, v10
 ; CHECK-NEXT:    vmv2r.v v10, v12
-; CHECK-NEXT:    vsetvli zero, a1, e64, m2, ta, mu
 ; CHECK-NEXT:    vlseg3e64ff.v v6, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
@@ -3296,11 +3296,11 @@ entry:
 define <vscale x 1 x i64> @test_vlseg4ff_mask_nxv1i64_triscv.vector.tuple_nxv8i8_4t(target("riscv.vector.tuple", <vscale x 8 x i8>, 4) %val, ptr %base, i32 %vl, <vscale x 1 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg4ff_mask_nxv1i64_triscv.vector.tuple_nxv8i8_4t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e64, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
 ; CHECK-NEXT:    vmv1r.v v10, v11
-; CHECK-NEXT:    vsetvli zero, a1, e64, m1, ta, mu
 ; CHECK-NEXT:    vlseg4e64ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
@@ -3336,11 +3336,11 @@ entry:
 define <vscale x 2 x i64> @test_vlseg4ff_mask_nxv2i64_triscv.vector.tuple_nxv16i8_4t(target("riscv.vector.tuple", <vscale x 16 x i8>, 4) %val, ptr %base, i32 %vl, <vscale x 2 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg4ff_mask_nxv2i64_triscv.vector.tuple_nxv16i8_4t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e64, m2, ta, mu
 ; CHECK-NEXT:    vmv2r.v v6, v8
 ; CHECK-NEXT:    vmv2r.v v8, v10
 ; CHECK-NEXT:    vmv2r.v v10, v12
 ; CHECK-NEXT:    vmv2r.v v12, v14
-; CHECK-NEXT:    vsetvli zero, a1, e64, m2, ta, mu
 ; CHECK-NEXT:    vlseg4e64ff.v v6, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
@@ -3376,12 +3376,12 @@ entry:
 define <vscale x 1 x i64> @test_vlseg5ff_mask_nxv1i64_triscv.vector.tuple_nxv8i8_5t(target("riscv.vector.tuple", <vscale x 8 x i8>, 5) %val, ptr %base, i32 %vl, <vscale x 1 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg5ff_mask_nxv1i64_triscv.vector.tuple_nxv8i8_5t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e64, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
 ; CHECK-NEXT:    vmv1r.v v10, v11
 ; CHECK-NEXT:    vmv1r.v v11, v12
-; CHECK-NEXT:    vsetvli zero, a1, e64, m1, ta, mu
 ; CHECK-NEXT:    vlseg5e64ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
@@ -3417,13 +3417,13 @@ entry:
 define <vscale x 1 x i64> @test_vlseg6ff_mask_nxv1i64_triscv.vector.tuple_nxv8i8_6t(target("riscv.vector.tuple", <vscale x 8 x i8>, 6) %val, ptr %base, i32 %vl, <vscale x 1 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg6ff_mask_nxv1i64_triscv.vector.tuple_nxv8i8_6t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e64, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
 ; CHECK-NEXT:    vmv1r.v v10, v11
 ; CHECK-NEXT:    vmv1r.v v11, v12
 ; CHECK-NEXT:    vmv1r.v v12, v13
-; CHECK-NEXT:    vsetvli zero, a1, e64, m1, ta, mu
 ; CHECK-NEXT:    vlseg6e64ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
@@ -3459,6 +3459,7 @@ entry:
 define <vscale x 1 x i64> @test_vlseg7ff_mask_nxv1i64_triscv.vector.tuple_nxv8i8_7t(target("riscv.vector.tuple", <vscale x 8 x i8>, 7) %val, ptr %base, i32 %vl, <vscale x 1 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg7ff_mask_nxv1i64_triscv.vector.tuple_nxv8i8_7t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e64, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
@@ -3466,7 +3467,6 @@ define <vscale x 1 x i64> @test_vlseg7ff_mask_nxv1i64_triscv.vector.tuple_nxv8i8
 ; CHECK-NEXT:    vmv1r.v v11, v12
 ; CHECK-NEXT:    vmv1r.v v12, v13
 ; CHECK-NEXT:    vmv1r.v v13, v14
-; CHECK-NEXT:    vsetvli zero, a1, e64, m1, ta, mu
 ; CHECK-NEXT:    vlseg7e64ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
@@ -3502,6 +3502,7 @@ entry:
 define <vscale x 1 x i64> @test_vlseg8ff_mask_nxv1i64_triscv.vector.tuple_nxv8i8_8t(target("riscv.vector.tuple", <vscale x 8 x i8>, 8) %val, ptr %base, i32 %vl, <vscale x 1 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg8ff_mask_nxv1i64_triscv.vector.tuple_nxv8i8_8t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e64, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
@@ -3510,7 +3511,6 @@ define <vscale x 1 x i64> @test_vlseg8ff_mask_nxv1i64_triscv.vector.tuple_nxv8i8
 ; CHECK-NEXT:    vmv1r.v v12, v13
 ; CHECK-NEXT:    vmv1r.v v13, v14
 ; CHECK-NEXT:    vmv1r.v v14, v15
-; CHECK-NEXT:    vsetvli zero, a1, e64, m1, ta, mu
 ; CHECK-NEXT:    vlseg8e64ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
@@ -3545,9 +3545,9 @@ entry:
 define <vscale x 1 x half> @test_vlseg2ff_mask_nxv1f16_triscv.vector.tuple_nxv2i8_2t(target("riscv.vector.tuple", <vscale x 2 x i8>, 2) %val, ptr %base, i32 %vl, <vscale x 1 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg2ff_mask_nxv1f16_triscv.vector.tuple_nxv2i8_2t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, mf4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
-; CHECK-NEXT:    vsetvli zero, a1, e16, mf4, ta, mu
 ; CHECK-NEXT:    vlseg2e16ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
@@ -3582,9 +3582,9 @@ entry:
 define <vscale x 2 x half> @test_vlseg2ff_mask_nxv2f16_triscv.vector.tuple_nxv4i8_2t(target("riscv.vector.tuple", <vscale x 4 x i8>, 2) %val, ptr %base, i32 %vl, <vscale x 2 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg2ff_mask_nxv2f16_triscv.vector.tuple_nxv4i8_2t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, mf2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
-; CHECK-NEXT:    vsetvli zero, a1, e16, mf2, ta, mu
 ; CHECK-NEXT:    vlseg2e16ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
@@ -3619,9 +3619,9 @@ entry:
 define <vscale x 4 x half> @test_vlseg2ff_mask_nxv4f16_triscv.vector.tuple_nxv8i8_2t(target("riscv.vector.tuple", <vscale x 8 x i8>, 2) %val, ptr %base, i32 %vl, <vscale x 4 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg2ff_mask_nxv4f16_triscv.vector.tuple_nxv8i8_2t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
-; CHECK-NEXT:    vsetvli zero, a1, e16, m1, ta, mu
 ; CHECK-NEXT:    vlseg2e16ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
@@ -3656,9 +3656,9 @@ entry:
 define <vscale x 8 x half> @test_vlseg2ff_mask_nxv8f16_triscv.vector.tuple_nxv16i8_2t(target("riscv.vector.tuple", <vscale x 16 x i8>, 2) %val, ptr %base, i32 %vl, <vscale x 8 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg2ff_mask_nxv8f16_triscv.vector.tuple_nxv16i8_2t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, m2, ta, mu
 ; CHECK-NEXT:    vmv2r.v v6, v8
 ; CHECK-NEXT:    vmv2r.v v8, v10
-; CHECK-NEXT:    vsetvli zero, a1, e16, m2, ta, mu
 ; CHECK-NEXT:    vlseg2e16ff.v v6, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
@@ -3693,9 +3693,9 @@ entry:
 define <vscale x 16 x half> @test_vlseg2ff_mask_nxv16f16_triscv.vector.tuple_nxv32i8_2t(target("riscv.vector.tuple", <vscale x 32 x i8>, 2) %val, ptr %base, i32 %vl, <vscale x 16 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg2ff_mask_nxv16f16_triscv.vector.tuple_nxv32i8_2t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, m4, ta, mu
 ; CHECK-NEXT:    vmv4r.v v4, v8
 ; CHECK-NEXT:    vmv4r.v v8, v12
-; CHECK-NEXT:    vsetvli zero, a1, e16, m4, ta, mu
 ; CHECK-NEXT:    vlseg2e16ff.v v4, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
@@ -3730,10 +3730,10 @@ entry:
 define <vscale x 1 x half> @test_vlseg3ff_mask_nxv1f16_triscv.vector.tuple_nxv2i8_3t(target("riscv.vector.tuple", <vscale x 2 x i8>, 3) %val, ptr %base, i32 %vl, <vscale x 1 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg3ff_mask_nxv1f16_triscv.vector.tuple_nxv2i8_3t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, mf4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
-; CHECK-NEXT:    vsetvli zero, a1, e16, mf4, ta, mu
 ; CHECK-NEXT:    vlseg3e16ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
@@ -3768,10 +3768,10 @@ entry:
 define <vscale x 2 x half> @test_vlseg3ff_mask_nxv2f16_triscv.vector.tuple_nxv4i8_3t(target("riscv.vector.tuple", <vscale x 4 x i8>, 3) %val, ptr %base, i32 %vl, <vscale x 2 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg3ff_mask_nxv2f16_triscv.vector.tuple_nxv4i8_3t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, mf2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
-; CHECK-NEXT:    vsetvli zero, a1, e16, mf2, ta, mu
 ; CHECK-NEXT:    vlseg3e16ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
@@ -3806,10 +3806,10 @@ entry:
 define <vscale x 4 x half> @test_vlseg3ff_mask_nxv4f16_triscv.vector.tuple_nxv8i8_3t(target("riscv.vector.tuple", <vscale x 8 x i8>, 3) %val, ptr %base, i32 %vl, <vscale x 4 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg3ff_mask_nxv4f16_triscv.vector.tuple_nxv8i8_3t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
-; CHECK-NEXT:    vsetvli zero, a1, e16, m1, ta, mu
 ; CHECK-NEXT:    vlseg3e16ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
@@ -3844,10 +3844,10 @@ entry:
 define <vscale x 8 x half> @test_vlseg3ff_mask_nxv8f16_triscv.vector.tuple_nxv16i8_3t(target("riscv.vector.tuple", <vscale x 16 x i8>, 3) %val, ptr %base, i32 %vl, <vscale x 8 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg3ff_mask_nxv8f16_triscv.vector.tuple_nxv16i8_3t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, m2, ta, mu
 ; CHECK-NEXT:    vmv2r.v v6, v8
 ; CHECK-NEXT:    vmv2r.v v8, v10
 ; CHECK-NEXT:    vmv2r.v v10, v12
-; CHECK-NEXT:    vsetvli zero, a1, e16, m2, ta, mu
 ; CHECK-NEXT:    vlseg3e16ff.v v6, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
@@ -3882,11 +3882,11 @@ entry:
 define <vscale x 1 x half> @test_vlseg4ff_mask_nxv1f16_triscv.vector.tuple_nxv2i8_4t(target("riscv.vector.tuple", <vscale x 2 x i8>, 4) %val, ptr %base, i32 %vl, <vscale x 1 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg4ff_mask_nxv1f16_triscv.vector.tuple_nxv2i8_4t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, mf4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
 ; CHECK-NEXT:    vmv1r.v v10, v11
-; CHECK-NEXT:    vsetvli zero, a1, e16, mf4, ta, mu
 ; CHECK-NEXT:    vlseg4e16ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
@@ -3921,11 +3921,11 @@ entry:
 define <vscale x 2 x half> @test_vlseg4ff_mask_nxv2f16_triscv.vector.tuple_nxv4i8_4t(target("riscv.vector.tuple", <vscale x 4 x i8>, 4) %val, ptr %base, i32 %vl, <vscale x 2 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg4ff_mask_nxv2f16_triscv.vector.tuple_nxv4i8_4t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, mf2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
 ; CHECK-NEXT:    vmv1r.v v10, v11
-; CHECK-NEXT:    vsetvli zero, a1, e16, mf2, ta, mu
 ; CHECK-NEXT:    vlseg4e16ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
@@ -3960,11 +3960,11 @@ entry:
 define <vscale x 4 x half> @test_vlseg4ff_mask_nxv4f16_triscv.vector.tuple_nxv8i8_4t(target("riscv.vector.tuple", <vscale x 8 x i8>, 4) %val, ptr %base, i32 %vl, <vscale x 4 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg4ff_mask_nxv4f16_triscv.vector.tuple_nxv8i8_4t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
 ; CHECK-NEXT:    vmv1r.v v10, v11
-; CHECK-NEXT:    vsetvli zero, a1, e16, m1, ta, mu
 ; CHECK-NEXT:    vlseg4e16ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
@@ -3999,11 +3999,11 @@ entry:
 define <vscale x 8 x half> @test_vlseg4ff_mask_nxv8f16_triscv.vector.tuple_nxv16i8_4t(target("riscv.vector.tuple", <vscale x 16 x i8>, 4) %val, ptr %base, i32 %vl, <vscale x 8 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg4ff_mask_nxv8f16_triscv.vector.tuple_nxv16i8_4t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, m2, ta, mu
 ; CHECK-NEXT:    vmv2r.v v6, v8
 ; CHECK-NEXT:    vmv2r.v v8, v10
 ; CHECK-NEXT:    vmv2r.v v10, v12
 ; CHECK-NEXT:    vmv2r.v v12, v14
-; CHECK-NEXT:    vsetvli zero, a1, e16, m2, ta, mu
 ; CHECK-NEXT:    vlseg4e16ff.v v6, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
@@ -4038,12 +4038,12 @@ entry:
 define <vscale x 1 x half> @test_vlseg5ff_mask_nxv1f16_triscv.vector.tuple_nxv2i8_5t(target("riscv.vector.tuple", <vscale x 2 x i8>, 5) %val, ptr %base, i32 %vl, <vscale x 1 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg5ff_mask_nxv1f16_triscv.vector.tuple_nxv2i8_5t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, mf4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
 ; CHECK-NEXT:    vmv1r.v v10, v11
 ; CHECK-NEXT:    vmv1r.v v11, v12
-; CHECK-NEXT:    vsetvli zero, a1, e16, mf4, ta, mu
 ; CHECK-NEXT:    vlseg5e16ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
@@ -4078,12 +4078,12 @@ entry:
 define <vscale x 2 x half> @test_vlseg5ff_mask_nxv2f16_triscv.vector.tuple_nxv4i8_5t(target("riscv.vector.tuple", <vscale x 4 x i8>, 5) %val, ptr %base, i32 %vl, <vscale x 2 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg5ff_mask_nxv2f16_triscv.vector.tuple_nxv4i8_5t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, mf2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
 ; CHECK-NEXT:    vmv1r.v v10, v11
 ; CHECK-NEXT:    vmv1r.v v11, v12
-; CHECK-NEXT:    vsetvli zero, a1, e16, mf2, ta, mu
 ; CHECK-NEXT:    vlseg5e16ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
@@ -4118,12 +4118,12 @@ entry:
 define <vscale x 4 x half> @test_vlseg5ff_mask_nxv4f16_triscv.vector.tuple_nxv8i8_5t(target("riscv.vector.tuple", <vscale x 8 x i8>, 5) %val, ptr %base, i32 %vl, <vscale x 4 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg5ff_mask_nxv4f16_triscv.vector.tuple_nxv8i8_5t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
 ; CHECK-NEXT:    vmv1r.v v10, v11
 ; CHECK-NEXT:    vmv1r.v v11, v12
-; CHECK-NEXT:    vsetvli zero, a1, e16, m1, ta, mu
 ; CHECK-NEXT:    vlseg5e16ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
@@ -4158,13 +4158,13 @@ entry:
 define <vscale x 1 x half> @test_vlseg6ff_mask_nxv1f16_triscv.vector.tuple_nxv2i8_6t(target("riscv.vector.tuple", <vscale x 2 x i8>, 6) %val, ptr %base, i32 %vl, <vscale x 1 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg6ff_mask_nxv1f16_triscv.vector.tuple_nxv2i8_6t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, mf4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
 ; CHECK-NEXT:    vmv1r.v v10, v11
 ; CHECK-NEXT:    vmv1r.v v11, v12
 ; CHECK-NEXT:    vmv1r.v v12, v13
-; CHECK-NEXT:    vsetvli zero, a1, e16, mf4, ta, mu
 ; CHECK-NEXT:    vlseg6e16ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
@@ -4199,13 +4199,13 @@ entry:
 define <vscale x 2 x half> @test_vlseg6ff_mask_nxv2f16_triscv.vector.tuple_nxv4i8_6t(target("riscv.vector.tuple", <vscale x 4 x i8>, 6) %val, ptr %base, i32 %vl, <vscale x 2 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg6ff_mask_nxv2f16_triscv.vector.tuple_nxv4i8_6t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, mf2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
 ; CHECK-NEXT:    vmv1r.v v10, v11
 ; CHECK-NEXT:    vmv1r.v v11, v12
 ; CHECK-NEXT:    vmv1r.v v12, v13
-; CHECK-NEXT:    vsetvli zero, a1, e16, mf2, ta, mu
 ; CHECK-NEXT:    vlseg6e16ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
@@ -4240,13 +4240,13 @@ entry:
 define <vscale x 4 x half> @test_vlseg6ff_mask_nxv4f16_triscv.vector.tuple_nxv8i8_6t(target("riscv.vector.tuple", <vscale x 8 x i8>, 6) %val, ptr %base, i32 %vl, <vscale x 4 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg6ff_mask_nxv4f16_triscv.vector.tuple_nxv8i8_6t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
 ; CHECK-NEXT:    vmv1r.v v10, v11
 ; CHECK-NEXT:    vmv1r.v v11, v12
 ; CHECK-NEXT:    vmv1r.v v12, v13
-; CHECK-NEXT:    vsetvli zero, a1, e16, m1, ta, mu
 ; CHECK-NEXT:    vlseg6e16ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
@@ -4281,6 +4281,7 @@ entry:
 define <vscale x 1 x half> @test_vlseg7ff_mask_nxv1f16_triscv.vector.tuple_nxv2i8_7t(target("riscv.vector.tuple", <vscale x 2 x i8>, 7) %val, ptr %base, i32 %vl, <vscale x 1 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg7ff_mask_nxv1f16_triscv.vector.tuple_nxv2i8_7t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, mf4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
@@ -4288,7 +4289,6 @@ define <vscale x 1 x half> @test_vlseg7ff_mask_nxv1f16_triscv.vector.tuple_nxv2i
 ; CHECK-NEXT:    vmv1r.v v11, v12
 ; CHECK-NEXT:    vmv1r.v v12, v13
 ; CHECK-NEXT:    vmv1r.v v13, v14
-; CHECK-NEXT:    vsetvli zero, a1, e16, mf4, ta, mu
 ; CHECK-NEXT:    vlseg7e16ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
@@ -4323,6 +4323,7 @@ entry:
 define <vscale x 2 x half> @test_vlseg7ff_mask_nxv2f16_triscv.vector.tuple_nxv4i8_7t(target("riscv.vector.tuple", <vscale x 4 x i8>, 7) %val, ptr %base, i32 %vl, <vscale x 2 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg7ff_mask_nxv2f16_triscv.vector.tuple_nxv4i8_7t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, mf2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
@@ -4330,7 +4331,6 @@ define <vscale x 2 x half> @test_vlseg7ff_mask_nxv2f16_triscv.vector.tuple_nxv4i
 ; CHECK-NEXT:    vmv1r.v v11, v12
 ; CHECK-NEXT:    vmv1r.v v12, v13
 ; CHECK-NEXT:    vmv1r.v v13, v14
-; CHECK-NEXT:    vsetvli zero, a1, e16, mf2, ta, mu
 ; CHECK-NEXT:    vlseg7e16ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
@@ -4365,6 +4365,7 @@ entry:
 define <vscale x 4 x half> @test_vlseg7ff_mask_nxv4f16_triscv.vector.tuple_nxv8i8_7t(target("riscv.vector.tuple", <vscale x 8 x i8>, 7) %val, ptr %base, i32 %vl, <vscale x 4 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg7ff_mask_nxv4f16_triscv.vector.tuple_nxv8i8_7t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
@@ -4372,7 +4373,6 @@ define <vscale x 4 x half> @test_vlseg7ff_mask_nxv4f16_triscv.vector.tuple_nxv8i
 ; CHECK-NEXT:    vmv1r.v v11, v12
 ; CHECK-NEXT:    vmv1r.v v12, v13
 ; CHECK-NEXT:    vmv1r.v v13, v14
-; CHECK-NEXT:    vsetvli zero, a1, e16, m1, ta, mu
 ; CHECK-NEXT:    vlseg7e16ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
@@ -4407,6 +4407,7 @@ entry:
 define <vscale x 1 x half> @test_vlseg8ff_mask_nxv1f16_triscv.vector.tuple_nxv2i8_8t(target("riscv.vector.tuple", <vscale x 2 x i8>, 8) %val, ptr %base, i32 %vl, <vscale x 1 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg8ff_mask_nxv1f16_triscv.vector.tuple_nxv2i8_8t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, mf4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
@@ -4415,7 +4416,6 @@ define <vscale x 1 x half> @test_vlseg8ff_mask_nxv1f16_triscv.vector.tuple_nxv2i
 ; CHECK-NEXT:    vmv1r.v v12, v13
 ; CHECK-NEXT:    vmv1r.v v13, v14
 ; CHECK-NEXT:    vmv1r.v v14, v15
-; CHECK-NEXT:    vsetvli zero, a1, e16, mf4, ta, mu
 ; CHECK-NEXT:    vlseg8e16ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
@@ -4450,6 +4450,7 @@ entry:
 define <vscale x 2 x half> @test_vlseg8ff_mask_nxv2f16_triscv.vector.tuple_nxv4i8_8t(target("riscv.vector.tuple", <vscale x 4 x i8>, 8) %val, ptr %base, i32 %vl, <vscale x 2 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg8ff_mask_nxv2f16_triscv.vector.tuple_nxv4i8_8t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, mf2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
@@ -4458,7 +4459,6 @@ define <vscale x 2 x half> @test_vlseg8ff_mask_nxv2f16_triscv.vector.tuple_nxv4i
 ; CHECK-NEXT:    vmv1r.v v12, v13
 ; CHECK-NEXT:    vmv1r.v v13, v14
 ; CHECK-NEXT:    vmv1r.v v14, v15
-; CHECK-NEXT:    vsetvli zero, a1, e16, mf2, ta, mu
 ; CHECK-NEXT:    vlseg8e16ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
@@ -4493,6 +4493,7 @@ entry:
 define <vscale x 4 x half> @test_vlseg8ff_mask_nxv4f16_triscv.vector.tuple_nxv8i8_8t(target("riscv.vector.tuple", <vscale x 8 x i8>, 8) %val, ptr %base, i32 %vl, <vscale x 4 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg8ff_mask_nxv4f16_triscv.vector.tuple_nxv8i8_8t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
@@ -4501,7 +4502,6 @@ define <vscale x 4 x half> @test_vlseg8ff_mask_nxv4f16_triscv.vector.tuple_nxv8i
 ; CHECK-NEXT:    vmv1r.v v12, v13
 ; CHECK-NEXT:    vmv1r.v v13, v14
 ; CHECK-NEXT:    vmv1r.v v14, v15
-; CHECK-NEXT:    vsetvli zero, a1, e16, m1, ta, mu
 ; CHECK-NEXT:    vlseg8e16ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
@@ -4536,9 +4536,9 @@ entry:
 define <vscale x 1 x float> @test_vlseg2ff_mask_nxv1f32_triscv.vector.tuple_nxv4i8_2t(target("riscv.vector.tuple", <vscale x 4 x i8>, 2) %val, ptr %base, i32 %vl, <vscale x 1 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg2ff_mask_nxv1f32_triscv.vector.tuple_nxv4i8_2t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e32, mf2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
-; CHECK-NEXT:    vsetvli zero, a1, e32, mf2, ta, mu
 ; CHECK-NEXT:    vlseg2e32ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
@@ -4573,9 +4573,9 @@ entry:
 define <vscale x 2 x float> @test_vlseg2ff_mask_nxv2f32_triscv.vector.tuple_nxv8i8_2t(target("riscv.vector.tuple", <vscale x 8 x i8>, 2) %val, ptr %base, i32 %vl, <vscale x 2 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg2ff_mask_nxv2f32_triscv.vector.tuple_nxv8i8_2t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e32, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
-; CHECK-NEXT:    vsetvli zero, a1, e32, m1, ta, mu
 ; CHECK-NEXT:    vlseg2e32ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
@@ -4610,9 +4610,9 @@ entry:
 define <vscale x 4 x float> @test_vlseg2ff_mask_nxv4f32_triscv.vector.tuple_nxv16i8_2t(target("riscv.vector.tuple", <vscale x 16 x i8>, 2) %val, ptr %base, i32 %vl, <vscale x 4 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg2ff_mask_nxv4f32_triscv.vector.tuple_nxv16i8_2t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e32, m2, ta, mu
 ; CHECK-NEXT:    vmv2r.v v6, v8
 ; CHECK-NEXT:    vmv2r.v v8, v10
-; CHECK-NEXT:    vsetvli zero, a1, e32, m2, ta, mu
 ; CHECK-NEXT:    vlseg2e32ff.v v6, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
@@ -4647,9 +4647,9 @@ entry:
 define <vscale x 8 x float> @test_vlseg2ff_mask_nxv8f32_triscv.vector.tuple_nxv32i8_2t(target("riscv.vector.tuple", <vscale x 32 x i8>, 2) %val, ptr %base, i32 %vl, <vscale x 8 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg2ff_mask_nxv8f32_triscv.vector.tuple_nxv32i8_2t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e32, m4, ta, mu
 ; CHECK-NEXT:    vmv4r.v v4, v8
 ; CHECK-NEXT:    vmv4r.v v8, v12
-; CHECK-NEXT:    vsetvli zero, a1, e32, m4, ta, mu
 ; CHECK-NEXT:    vlseg2e32ff.v v4, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
@@ -4684,10 +4684,10 @@ entry:
 define <vscale x 1 x float> @test_vlseg3ff_mask_nxv1f32_triscv.vector.tuple_nxv4i8_3t(target("riscv.vector.tuple", <vscale x 4 x i8>, 3) %val, ptr %base, i32 %vl, <vscale x 1 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg3ff_mask_nxv1f32_triscv.vector.tuple_nxv4i8_3t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e32, mf2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
-; CHECK-NEXT:    vsetvli zero, a1, e32, mf2, ta, mu
 ; CHECK-NEXT:    vlseg3e32ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
@@ -4722,10 +4722,10 @@ entry:
 define <vscale x 2 x float> @test_vlseg3ff_mask_nxv2f32_triscv.vector.tuple_nxv8i8_3t(target("riscv.vector.tuple", <vscale x 8 x i8>, 3) %val, ptr %base, i32 %vl, <vscale x 2 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg3ff_mask_nxv2f32_triscv.vector.tuple_nxv8i8_3t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e32, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
-; CHECK-NEXT:    vsetvli zero, a1, e32, m1, ta, mu
 ; CHECK-NEXT:    vlseg3e32ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
@@ -4760,10 +4760,10 @@ entry:
 define <vscale x 4 x float> @test_vlseg3ff_mask_nxv4f32_triscv.vector.tuple_nxv16i8_3t(target("riscv.vector.tuple", <vscale x 16 x i8>, 3) %val, ptr %base, i32 %vl, <vscale x 4 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg3ff_mask_nxv4f32_triscv.vector.tuple_nxv16i8_3t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e32, m2, ta, mu
 ; CHECK-NEXT:    vmv2r.v v6, v8
 ; CHECK-NEXT:    vmv2r.v v8, v10
 ; CHECK-NEXT:    vmv2r.v v10, v12
-; CHECK-NEXT:    vsetvli zero, a1, e32, m2, ta, mu
 ; CHECK-NEXT:    vlseg3e32ff.v v6, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
@@ -4798,11 +4798,11 @@ entry:
 define <vscale x 1 x float> @test_vlseg4ff_mask_nxv1f32_triscv.vector.tuple_nxv4i8_4t(target("riscv.vector.tuple", <vscale x 4 x i8>, 4) %val, ptr %base, i32 %vl, <vscale x 1 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg4ff_mask_nxv1f32_triscv.vector.tuple_nxv4i8_4t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e32, mf2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
 ; CHECK-NEXT:    vmv1r.v v10, v11
-; CHECK-NEXT:    vsetvli zero, a1, e32, mf2, ta, mu
 ; CHECK-NEXT:    vlseg4e32ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
@@ -4837,11 +4837,11 @@ entry:
 define <vscale x 2 x float> @test_vlseg4ff_mask_nxv2f32_triscv.vector.tuple_nxv8i8_4t(target("riscv.vector.tuple", <vscale x 8 x i8>, 4) %val, ptr %base, i32 %vl, <vscale x 2 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg4ff_mask_nxv2f32_triscv.vector.tuple_nxv8i8_4t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e32, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
 ; CHECK-NEXT:    vmv1r.v v10, v11
-; CHECK-NEXT:    vsetvli zero, a1, e32, m1, ta, mu
 ; CHECK-NEXT:    vlseg4e32ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
@@ -4876,11 +4876,11 @@ entry:
 define <vscale x 4 x float> @test_vlseg4ff_mask_nxv4f32_triscv.vector.tuple_nxv16i8_4t(target("riscv.vector.tuple", <vscale x 16 x i8>, 4) %val, ptr %base, i32 %vl, <vscale x 4 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg4ff_mask_nxv4f32_triscv.vector.tuple_nxv16i8_4t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e32, m2, ta, mu
 ; CHECK-NEXT:    vmv2r.v v6, v8
 ; CHECK-NEXT:    vmv2r.v v8, v10
 ; CHECK-NEXT:    vmv2r.v v10, v12
 ; CHECK-NEXT:    vmv2r.v v12, v14
-; CHECK-NEXT:    vsetvli zero, a1, e32, m2, ta, mu
 ; CHECK-NEXT:    vlseg4e32ff.v v6, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
@@ -4915,12 +4915,12 @@ entry:
 define <vscale x 1 x float> @test_vlseg5ff_mask_nxv1f32_triscv.vector.tuple_nxv4i8_5t(target("riscv.vector.tuple", <vscale x 4 x i8>, 5) %val, ptr %base, i32 %vl, <vscale x 1 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg5ff_mask_nxv1f32_triscv.vector.tuple_nxv4i8_5t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e32, mf2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
 ; CHECK-NEXT:    vmv1r.v v10, v11
 ; CHECK-NEXT:    vmv1r.v v11, v12
-; CHECK-NEXT:    vsetvli zero, a1, e32, mf2, ta, mu
 ; CHECK-NEXT:    vlseg5e32ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
@@ -4955,12 +4955,12 @@ entry:
 define <vscale x 2 x float> @test_vlseg5ff_mask_nxv2f32_triscv.vector.tuple_nxv8i8_5t(target("riscv.vector.tuple", <vscale x 8 x i8>, 5) %val, ptr %base, i32 %vl, <vscale x 2 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg5ff_mask_nxv2f32_triscv.vector.tuple_nxv8i8_5t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e32, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
 ; CHECK-NEXT:    vmv1r.v v10, v11
 ; CHECK-NEXT:    vmv1r.v v11, v12
-; CHECK-NEXT:    vsetvli zero, a1, e32, m1, ta, mu
 ; CHECK-NEXT:    vlseg5e32ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
@@ -4995,13 +4995,13 @@ entry:
 define <vscale x 1 x float> @test_vlseg6ff_mask_nxv1f32_triscv.vector.tuple_nxv4i8_6t(target("riscv.vector.tuple", <vscale x 4 x i8>, 6) %val, ptr %base, i32 %vl, <vscale x 1 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg6ff_mask_nxv1f32_triscv.vector.tuple_nxv4i8_6t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e32, mf2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
 ; CHECK-NEXT:    vmv1r.v v10, v11
 ; CHECK-NEXT:    vmv1r.v v11, v12
 ; CHECK-NEXT:    vmv1r.v v12, v13
-; CHECK-NEXT:    vsetvli zero, a1, e32, mf2, ta, mu
 ; CHECK-NEXT:    vlseg6e32ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
@@ -5036,13 +5036,13 @@ entry:
 define <vscale x 2 x float> @test_vlseg6ff_mask_nxv2f32_triscv.vector.tuple_nxv8i8_6t(target("riscv.vector.tuple", <vscale x 8 x i8>, 6) %val, ptr %base, i32 %vl, <vscale x 2 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg6ff_mask_nxv2f32_triscv.vector.tuple_nxv8i8_6t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e32, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
 ; CHECK-NEXT:    vmv1r.v v10, v11
 ; CHECK-NEXT:    vmv1r.v v11, v12
 ; CHECK-NEXT:    vmv1r.v v12, v13
-; CHECK-NEXT:    vsetvli zero, a1, e32, m1, ta, mu
 ; CHECK-NEXT:    vlseg6e32ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
@@ -5077,6 +5077,7 @@ entry:
 define <vscale x 1 x float> @test_vlseg7ff_mask_nxv1f32_triscv.vector.tuple_nxv4i8_7t(target("riscv.vector.tuple", <vscale x 4 x i8>, 7) %val, ptr %base, i32 %vl, <vscale x 1 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg7ff_mask_nxv1f32_triscv.vector.tuple_nxv4i8_7t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e32, mf2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
@@ -5084,7 +5085,6 @@ define <vscale x 1 x float> @test_vlseg7ff_mask_nxv1f32_triscv.vector.tuple_nxv4
 ; CHECK-NEXT:    vmv1r.v v11, v12
 ; CHECK-NEXT:    vmv1r.v v12, v13
 ; CHECK-NEXT:    vmv1r.v v13, v14
-; CHECK-NEXT:    vsetvli zero, a1, e32, mf2, ta, mu
 ; CHECK-NEXT:    vlseg7e32ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
@@ -5119,6 +5119,7 @@ entry:
 define <vscale x 2 x float> @test_vlseg7ff_mask_nxv2f32_triscv.vector.tuple_nxv8i8_7t(target("riscv.vector.tuple", <vscale x 8 x i8>, 7) %val, ptr %base, i32 %vl, <vscale x 2 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg7ff_mask_nxv2f32_triscv.vector.tuple_nxv8i8_7t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e32, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
@@ -5126,7 +5127,6 @@ define <vscale x 2 x float> @test_vlseg7ff_mask_nxv2f32_triscv.vector.tuple_nxv8
 ; CHECK-NEXT:    vmv1r.v v11, v12
 ; CHECK-NEXT:    vmv1r.v v12, v13
 ; CHECK-NEXT:    vmv1r.v v13, v14
-; CHECK-NEXT:    vsetvli zero, a1, e32, m1, ta, mu
 ; CHECK-NEXT:    vlseg7e32ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
@@ -5161,6 +5161,7 @@ entry:
 define <vscale x 1 x float> @test_vlseg8ff_mask_nxv1f32_triscv.vector.tuple_nxv4i8_8t(target("riscv.vector.tuple", <vscale x 4 x i8>, 8) %val, ptr %base, i32 %vl, <vscale x 1 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg8ff_mask_nxv1f32_triscv.vector.tuple_nxv4i8_8t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e32, mf2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
@@ -5169,7 +5170,6 @@ define <vscale x 1 x float> @test_vlseg8ff_mask_nxv1f32_triscv.vector.tuple_nxv4
 ; CHECK-NEXT:    vmv1r.v v12, v13
 ; CHECK-NEXT:    vmv1r.v v13, v14
 ; CHECK-NEXT:    vmv1r.v v14, v15
-; CHECK-NEXT:    vsetvli zero, a1, e32, mf2, ta, mu
 ; CHECK-NEXT:    vlseg8e32ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
@@ -5204,6 +5204,7 @@ entry:
 define <vscale x 2 x float> @test_vlseg8ff_mask_nxv2f32_triscv.vector.tuple_nxv8i8_8t(target("riscv.vector.tuple", <vscale x 8 x i8>, 8) %val, ptr %base, i32 %vl, <vscale x 2 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg8ff_mask_nxv2f32_triscv.vector.tuple_nxv8i8_8t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e32, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
@@ -5212,7 +5213,6 @@ define <vscale x 2 x float> @test_vlseg8ff_mask_nxv2f32_triscv.vector.tuple_nxv8
 ; CHECK-NEXT:    vmv1r.v v12, v13
 ; CHECK-NEXT:    vmv1r.v v13, v14
 ; CHECK-NEXT:    vmv1r.v v14, v15
-; CHECK-NEXT:    vsetvli zero, a1, e32, m1, ta, mu
 ; CHECK-NEXT:    vlseg8e32ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
@@ -5247,9 +5247,9 @@ entry:
 define <vscale x 1 x double> @test_vlseg2ff_mask_nxv1f64_triscv.vector.tuple_nxv8i8_2t(target("riscv.vector.tuple", <vscale x 8 x i8>, 2) %val, ptr %base, i32 %vl, <vscale x 1 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg2ff_mask_nxv1f64_triscv.vector.tuple_nxv8i8_2t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e64, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
-; CHECK-NEXT:    vsetvli zero, a1, e64, m1, ta, mu
 ; CHECK-NEXT:    vlseg2e64ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
@@ -5284,9 +5284,9 @@ entry:
 define <vscale x 2 x double> @test_vlseg2ff_mask_nxv2f64_triscv.vector.tuple_nxv16i8_2t(target("riscv.vector.tuple", <vscale x 16 x i8>, 2) %val, ptr %base, i32 %vl, <vscale x 2 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg2ff_mask_nxv2f64_triscv.vector.tuple_nxv16i8_2t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e64, m2, ta, mu
 ; CHECK-NEXT:    vmv2r.v v6, v8
 ; CHECK-NEXT:    vmv2r.v v8, v10
-; CHECK-NEXT:    vsetvli zero, a1, e64, m2, ta, mu
 ; CHECK-NEXT:    vlseg2e64ff.v v6, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
@@ -5321,9 +5321,9 @@ entry:
 define <vscale x 4 x double> @test_vlseg2ff_mask_nxv4f64_triscv.vector.tuple_nxv32i8_2t(target("riscv.vector.tuple", <vscale x 32 x i8>, 2) %val, ptr %base, i32 %vl, <vscale x 4 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg2ff_mask_nxv4f64_triscv.vector.tuple_nxv32i8_2t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e64, m4, ta, mu
 ; CHECK-NEXT:    vmv4r.v v4, v8
 ; CHECK-NEXT:    vmv4r.v v8, v12
-; CHECK-NEXT:    vsetvli zero, a1, e64, m4, ta, mu
 ; CHECK-NEXT:    vlseg2e64ff.v v4, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
@@ -5358,10 +5358,10 @@ entry:
 define <vscale x 1 x double> @test_vlseg3ff_mask_nxv1f64_triscv.vector.tuple_nxv8i8_3t(target("riscv.vector.tuple", <vscale x 8 x i8>, 3) %val, ptr %base, i32 %vl, <vscale x 1 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg3ff_mask_nxv1f64_triscv.vector.tuple_nxv8i8_3t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e64, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
-; CHECK-NEXT:    vsetvli zero, a1, e64, m1, ta, mu
 ; CHECK-NEXT:    vlseg3e64ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
@@ -5396,10 +5396,10 @@ entry:
 define <vscale x 2 x double> @test_vlseg3ff_mask_nxv2f64_triscv.vector.tuple_nxv16i8_3t(target("riscv.vector.tuple", <vscale x 16 x i8>, 3) %val, ptr %base, i32 %vl, <vscale x 2 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg3ff_mask_nxv2f64_triscv.vector.tuple_nxv16i8_3t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e64, m2, ta, mu
 ; CHECK-NEXT:    vmv2r.v v6, v8
 ; CHECK-NEXT:    vmv2r.v v8, v10
 ; CHECK-NEXT:    vmv2r.v v10, v12
-; CHECK-NEXT:    vsetvli zero, a1, e64, m2, ta, mu
 ; CHECK-NEXT:    vlseg3e64ff.v v6, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
@@ -5434,11 +5434,11 @@ entry:
 define <vscale x 1 x double> @test_vlseg4ff_mask_nxv1f64_triscv.vector.tuple_nxv8i8_4t(target("riscv.vector.tuple", <vscale x 8 x i8>, 4) %val, ptr %base, i32 %vl, <vscale x 1 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg4ff_mask_nxv1f64_triscv.vector.tuple_nxv8i8_4t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e64, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
 ; CHECK-NEXT:    vmv1r.v v10, v11
-; CHECK-NEXT:    vsetvli zero, a1, e64, m1, ta, mu
 ; CHECK-NEXT:    vlseg4e64ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
@@ -5473,11 +5473,11 @@ entry:
 define <vscale x 2 x double> @test_vlseg4ff_mask_nxv2f64_triscv.vector.tuple_nxv16i8_4t(target("riscv.vector.tuple", <vscale x 16 x i8>, 4) %val, ptr %base, i32 %vl, <vscale x 2 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg4ff_mask_nxv2f64_triscv.vector.tuple_nxv16i8_4t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e64, m2, ta, mu
 ; CHECK-NEXT:    vmv2r.v v6, v8
 ; CHECK-NEXT:    vmv2r.v v8, v10
 ; CHECK-NEXT:    vmv2r.v v10, v12
 ; CHECK-NEXT:    vmv2r.v v12, v14
-; CHECK-NEXT:    vsetvli zero, a1, e64, m2, ta, mu
 ; CHECK-NEXT:    vlseg4e64ff.v v6, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
@@ -5512,12 +5512,12 @@ entry:
 define <vscale x 1 x double> @test_vlseg5ff_mask_nxv1f64_triscv.vector.tuple_nxv8i8_5t(target("riscv.vector.tuple", <vscale x 8 x i8>, 5) %val, ptr %base, i32 %vl, <vscale x 1 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg5ff_mask_nxv1f64_triscv.vector.tuple_nxv8i8_5t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e64, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
 ; CHECK-NEXT:    vmv1r.v v10, v11
 ; CHECK-NEXT:    vmv1r.v v11, v12
-; CHECK-NEXT:    vsetvli zero, a1, e64, m1, ta, mu
 ; CHECK-NEXT:    vlseg5e64ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
@@ -5552,13 +5552,13 @@ entry:
 define <vscale x 1 x double> @test_vlseg6ff_mask_nxv1f64_triscv.vector.tuple_nxv8i8_6t(target("riscv.vector.tuple", <vscale x 8 x i8>, 6) %val, ptr %base, i32 %vl, <vscale x 1 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg6ff_mask_nxv1f64_triscv.vector.tuple_nxv8i8_6t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e64, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
 ; CHECK-NEXT:    vmv1r.v v10, v11
 ; CHECK-NEXT:    vmv1r.v v11, v12
 ; CHECK-NEXT:    vmv1r.v v12, v13
-; CHECK-NEXT:    vsetvli zero, a1, e64, m1, ta, mu
 ; CHECK-NEXT:    vlseg6e64ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
@@ -5593,6 +5593,7 @@ entry:
 define <vscale x 1 x double> @test_vlseg7ff_mask_nxv1f64_triscv.vector.tuple_nxv8i8_7t(target("riscv.vector.tuple", <vscale x 8 x i8>, 7) %val, ptr %base, i32 %vl, <vscale x 1 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg7ff_mask_nxv1f64_triscv.vector.tuple_nxv8i8_7t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e64, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
@@ -5600,7 +5601,6 @@ define <vscale x 1 x double> @test_vlseg7ff_mask_nxv1f64_triscv.vector.tuple_nxv
 ; CHECK-NEXT:    vmv1r.v v11, v12
 ; CHECK-NEXT:    vmv1r.v v12, v13
 ; CHECK-NEXT:    vmv1r.v v13, v14
-; CHECK-NEXT:    vsetvli zero, a1, e64, m1, ta, mu
 ; CHECK-NEXT:    vlseg7e64ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
@@ -5635,6 +5635,7 @@ entry:
 define <vscale x 1 x double> @test_vlseg8ff_mask_nxv1f64_triscv.vector.tuple_nxv8i8_8t(target("riscv.vector.tuple", <vscale x 8 x i8>, 8) %val, ptr %base, i32 %vl, <vscale x 1 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg8ff_mask_nxv1f64_triscv.vector.tuple_nxv8i8_8t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e64, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
@@ -5643,7 +5644,6 @@ define <vscale x 1 x double> @test_vlseg8ff_mask_nxv1f64_triscv.vector.tuple_nxv
 ; CHECK-NEXT:    vmv1r.v v12, v13
 ; CHECK-NEXT:    vmv1r.v v13, v14
 ; CHECK-NEXT:    vmv1r.v v14, v15
-; CHECK-NEXT:    vsetvli zero, a1, e64, m1, ta, mu
 ; CHECK-NEXT:    vlseg8e64ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
@@ -5678,9 +5678,9 @@ entry:
 define <vscale x 1 x bfloat> @test_vlseg2ff_mask_nxv1bf16_triscv.vector.tuple_nxv2i8_2t(target("riscv.vector.tuple", <vscale x 2 x i8>, 2) %val, ptr %base, i32 %vl, <vscale x 1 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg2ff_mask_nxv1bf16_triscv.vector.tuple_nxv2i8_2t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, mf4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
-; CHECK-NEXT:    vsetvli zero, a1, e16, mf4, ta, mu
 ; CHECK-NEXT:    vlseg2e16ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
@@ -5715,9 +5715,9 @@ entry:
 define <vscale x 2 x bfloat> @test_vlseg2ff_mask_nxv2bf16_triscv.vector.tuple_nxv4i8_2t(target("riscv.vector.tuple", <vscale x 4 x i8>, 2) %val, ptr %base, i32 %vl, <vscale x 2 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg2ff_mask_nxv2bf16_triscv.vector.tuple_nxv4i8_2t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, mf2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
-; CHECK-NEXT:    vsetvli zero, a1, e16, mf2, ta, mu
 ; CHECK-NEXT:    vlseg2e16ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
@@ -5752,9 +5752,9 @@ entry:
 define <vscale x 4 x bfloat> @test_vlseg2ff_mask_nxv4bf16_triscv.vector.tuple_nxv8i8_2t(target("riscv.vector.tuple", <vscale x 8 x i8>, 2) %val, ptr %base, i32 %vl, <vscale x 4 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg2ff_mask_nxv4bf16_triscv.vector.tuple_nxv8i8_2t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
-; CHECK-NEXT:    vsetvli zero, a1, e16, m1, ta, mu
 ; CHECK-NEXT:    vlseg2e16ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
@@ -5789,9 +5789,9 @@ entry:
 define <vscale x 8 x bfloat> @test_vlseg2ff_mask_nxv8bf16_triscv.vector.tuple_nxv16i8_2t(target("riscv.vector.tuple", <vscale x 16 x i8>, 2) %val, ptr %base, i32 %vl, <vscale x 8 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg2ff_mask_nxv8bf16_triscv.vector.tuple_nxv16i8_2t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, m2, ta, mu
 ; CHECK-NEXT:    vmv2r.v v6, v8
 ; CHECK-NEXT:    vmv2r.v v8, v10
-; CHECK-NEXT:    vsetvli zero, a1, e16, m2, ta, mu
 ; CHECK-NEXT:    vlseg2e16ff.v v6, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
@@ -5826,9 +5826,9 @@ entry:
 define <vscale x 16 x bfloat> @test_vlseg2ff_mask_nxv16bf16_triscv.vector.tuple_nxv32i8_2t(target("riscv.vector.tuple", <vscale x 32 x i8>, 2) %val, ptr %base, i32 %vl, <vscale x 16 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg2ff_mask_nxv16bf16_triscv.vector.tuple_nxv32i8_2t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, m4, ta, mu
 ; CHECK-NEXT:    vmv4r.v v4, v8
 ; CHECK-NEXT:    vmv4r.v v8, v12
-; CHECK-NEXT:    vsetvli zero, a1, e16, m4, ta, mu
 ; CHECK-NEXT:    vlseg2e16ff.v v4, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
@@ -5863,10 +5863,10 @@ entry:
 define <vscale x 1 x bfloat> @test_vlseg3ff_mask_nxv1bf16_triscv.vector.tuple_nxv2i8_3t(target("riscv.vector.tuple", <vscale x 2 x i8>, 3) %val, ptr %base, i32 %vl, <vscale x 1 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg3ff_mask_nxv1bf16_triscv.vector.tuple_nxv2i8_3t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, mf4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
-; CHECK-NEXT:    vsetvli zero, a1, e16, mf4, ta, mu
 ; CHECK-NEXT:    vlseg3e16ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
@@ -5901,10 +5901,10 @@ entry:
 define <vscale x 2 x bfloat> @test_vlseg3ff_mask_nxv2bf16_triscv.vector.tuple_nxv4i8_3t(target("riscv.vector.tuple", <vscale x 4 x i8>, 3) %val, ptr %base, i32 %vl, <vscale x 2 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg3ff_mask_nxv2bf16_triscv.vector.tuple_nxv4i8_3t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, mf2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
-; CHECK-NEXT:    vsetvli zero, a1, e16, mf2, ta, mu
 ; CHECK-NEXT:    vlseg3e16ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
@@ -5939,10 +5939,10 @@ entry:
 define <vscale x 4 x bfloat> @test_vlseg3ff_mask_nxv4bf16_triscv.vector.tuple_nxv8i8_3t(target("riscv.vector.tuple", <vscale x 8 x i8>, 3) %val, ptr %base, i32 %vl, <vscale x 4 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg3ff_mask_nxv4bf16_triscv.vector.tuple_nxv8i8_3t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
-; CHECK-NEXT:    vsetvli zero, a1, e16, m1, ta, mu
 ; CHECK-NEXT:    vlseg3e16ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
@@ -5977,10 +5977,10 @@ entry:
 define <vscale x 8 x bfloat> @test_vlseg3ff_mask_nxv8bf16_triscv.vector.tuple_nxv16i8_3t(target("riscv.vector.tuple", <vscale x 16 x i8>, 3) %val, ptr %base, i32 %vl, <vscale x 8 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg3ff_mask_nxv8bf16_triscv.vector.tuple_nxv16i8_3t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, m2, ta, mu
 ; CHECK-NEXT:    vmv2r.v v6, v8
 ; CHECK-NEXT:    vmv2r.v v8, v10
 ; CHECK-NEXT:    vmv2r.v v10, v12
-; CHECK-NEXT:    vsetvli zero, a1, e16, m2, ta, mu
 ; CHECK-NEXT:    vlseg3e16ff.v v6, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
@@ -6015,11 +6015,11 @@ entry:
 define <vscale x 1 x bfloat> @test_vlseg4ff_mask_nxv1bf16_triscv.vector.tuple_nxv2i8_4t(target("riscv.vector.tuple", <vscale x 2 x i8>, 4) %val, ptr %base, i32 %vl, <vscale x 1 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg4ff_mask_nxv1bf16_triscv.vector.tuple_nxv2i8_4t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, mf4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
 ; CHECK-NEXT:    vmv1r.v v10, v11
-; CHECK-NEXT:    vsetvli zero, a1, e16, mf4, ta, mu
 ; CHECK-NEXT:    vlseg4e16ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
@@ -6054,11 +6054,11 @@ entry:
 define <vscale x 2 x bfloat> @test_vlseg4ff_mask_nxv2bf16_triscv.vector.tuple_nxv4i8_4t(target("riscv.vector.tuple", <vscale x 4 x i8>, 4) %val, ptr %base, i32 %vl, <vscale x 2 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg4ff_mask_nxv2bf16_triscv.vector.tuple_nxv4i8_4t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, mf2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
 ; CHECK-NEXT:    vmv1r.v v10, v11
-; CHECK-NEXT:    vsetvli zero, a1, e16, mf2, ta, mu
 ; CHECK-NEXT:    vlseg4e16ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
@@ -6093,11 +6093,11 @@ entry:
 define <vscale x 4 x bfloat> @test_vlseg4ff_mask_nxv4bf16_triscv.vector.tuple_nxv8i8_4t(target("riscv.vector.tuple", <vscale x 8 x i8>, 4) %val, ptr %base, i32 %vl, <vscale x 4 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg4ff_mask_nxv4bf16_triscv.vector.tuple_nxv8i8_4t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
 ; CHECK-NEXT:    vmv1r.v v10, v11
-; CHECK-NEXT:    vsetvli zero, a1, e16, m1, ta, mu
 ; CHECK-NEXT:    vlseg4e16ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
@@ -6132,11 +6132,11 @@ entry:
 define <vscale x 8 x bfloat> @test_vlseg4ff_mask_nxv8bf16_triscv.vector.tuple_nxv16i8_4t(target("riscv.vector.tuple", <vscale x 16 x i8>, 4) %val, ptr %base, i32 %vl, <vscale x 8 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg4ff_mask_nxv8bf16_triscv.vector.tuple_nxv16i8_4t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, m2, ta, mu
 ; CHECK-NEXT:    vmv2r.v v6, v8
 ; CHECK-NEXT:    vmv2r.v v8, v10
 ; CHECK-NEXT:    vmv2r.v v10, v12
 ; CHECK-NEXT:    vmv2r.v v12, v14
-; CHECK-NEXT:    vsetvli zero, a1, e16, m2, ta, mu
 ; CHECK-NEXT:    vlseg4e16ff.v v6, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
@@ -6171,12 +6171,12 @@ entry:
 define <vscale x 1 x bfloat> @test_vlseg5ff_mask_nxv1bf16_triscv.vector.tuple_nxv2i8_5t(target("riscv.vector.tuple", <vscale x 2 x i8>, 5) %val, ptr %base, i32 %vl, <vscale x 1 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg5ff_mask_nxv1bf16_triscv.vector.tuple_nxv2i8_5t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, mf4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
 ; CHECK-NEXT:    vmv1r.v v10, v11
 ; CHECK-NEXT:    vmv1r.v v11, v12
-; CHECK-NEXT:    vsetvli zero, a1, e16, mf4, ta, mu
 ; CHECK-NEXT:    vlseg5e16ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
@@ -6211,12 +6211,12 @@ entry:
 define <vscale x 2 x bfloat> @test_vlseg5ff_mask_nxv2bf16_triscv.vector.tuple_nxv4i8_5t(target("riscv.vector.tuple", <vscale x 4 x i8>, 5) %val, ptr %base, i32 %vl, <vscale x 2 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg5ff_mask_nxv2bf16_triscv.vector.tuple_nxv4i8_5t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, mf2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
 ; CHECK-NEXT:    vmv1r.v v10, v11
 ; CHECK-NEXT:    vmv1r.v v11, v12
-; CHECK-NEXT:    vsetvli zero, a1, e16, mf2, ta, mu
 ; CHECK-NEXT:    vlseg5e16ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
@@ -6251,12 +6251,12 @@ entry:
 define <vscale x 4 x bfloat> @test_vlseg5ff_mask_nxv4bf16_triscv.vector.tuple_nxv8i8_5t(target("riscv.vector.tuple", <vscale x 8 x i8>, 5) %val, ptr %base, i32 %vl, <vscale x 4 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg5ff_mask_nxv4bf16_triscv.vector.tuple_nxv8i8_5t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
 ; CHECK-NEXT:    vmv1r.v v10, v11
 ; CHECK-NEXT:    vmv1r.v v11, v12
-; CHECK-NEXT:    vsetvli zero, a1, e16, m1, ta, mu
 ; CHECK-NEXT:    vlseg5e16ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
@@ -6291,13 +6291,13 @@ entry:
 define <vscale x 1 x bfloat> @test_vlseg6ff_mask_nxv1bf16_triscv.vector.tuple_nxv2i8_6t(target("riscv.vector.tuple", <vscale x 2 x i8>, 6) %val, ptr %base, i32 %vl, <vscale x 1 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg6ff_mask_nxv1bf16_triscv.vector.tuple_nxv2i8_6t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, mf4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
 ; CHECK-NEXT:    vmv1r.v v10, v11
 ; CHECK-NEXT:    vmv1r.v v11, v12
 ; CHECK-NEXT:    vmv1r.v v12, v13
-; CHECK-NEXT:    vsetvli zero, a1, e16, mf4, ta, mu
 ; CHECK-NEXT:    vlseg6e16ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
@@ -6332,13 +6332,13 @@ entry:
 define <vscale x 2 x bfloat> @test_vlseg6ff_mask_nxv2bf16_triscv.vector.tuple_nxv4i8_6t(target("riscv.vector.tuple", <vscale x 4 x i8>, 6) %val, ptr %base, i32 %vl, <vscale x 2 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg6ff_mask_nxv2bf16_triscv.vector.tuple_nxv4i8_6t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, mf2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
 ; CHECK-NEXT:    vmv1r.v v10, v11
 ; CHECK-NEXT:    vmv1r.v v11, v12
 ; CHECK-NEXT:    vmv1r.v v12, v13
-; CHECK-NEXT:    vsetvli zero, a1, e16, mf2, ta, mu
 ; CHECK-NEXT:    vlseg6e16ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
@@ -6373,13 +6373,13 @@ entry:
 define <vscale x 4 x bfloat> @test_vlseg6ff_mask_nxv4bf16_triscv.vector.tuple_nxv8i8_6t(target("riscv.vector.tuple", <vscale x 8 x i8>, 6) %val, ptr %base, i32 %vl, <vscale x 4 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg6ff_mask_nxv4bf16_triscv.vector.tuple_nxv8i8_6t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
 ; CHECK-NEXT:    vmv1r.v v10, v11
 ; CHECK-NEXT:    vmv1r.v v11, v12
 ; CHECK-NEXT:    vmv1r.v v12, v13
-; CHECK-NEXT:    vsetvli zero, a1, e16, m1, ta, mu
 ; CHECK-NEXT:    vlseg6e16ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
@@ -6414,6 +6414,7 @@ entry:
 define <vscale x 1 x bfloat> @test_vlseg7ff_mask_nxv1bf16_triscv.vector.tuple_nxv2i8_7t(target("riscv.vector.tuple", <vscale x 2 x i8>, 7) %val, ptr %base, i32 %vl, <vscale x 1 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg7ff_mask_nxv1bf16_triscv.vector.tuple_nxv2i8_7t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, mf4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
@@ -6421,7 +6422,6 @@ define <vscale x 1 x bfloat> @test_vlseg7ff_mask_nxv1bf16_triscv.vector.tuple_nx
 ; CHECK-NEXT:    vmv1r.v v11, v12
 ; CHECK-NEXT:    vmv1r.v v12, v13
 ; CHECK-NEXT:    vmv1r.v v13, v14
-; CHECK-NEXT:    vsetvli zero, a1, e16, mf4, ta, mu
 ; CHECK-NEXT:    vlseg7e16ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
@@ -6456,6 +6456,7 @@ entry:
 define <vscale x 2 x bfloat> @test_vlseg7ff_mask_nxv2bf16_triscv.vector.tuple_nxv4i8_7t(target("riscv.vector.tuple", <vscale x 4 x i8>, 7) %val, ptr %base, i32 %vl, <vscale x 2 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg7ff_mask_nxv2bf16_triscv.vector.tuple_nxv4i8_7t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, mf2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
@@ -6463,7 +6464,6 @@ define <vscale x 2 x bfloat> @test_vlseg7ff_mask_nxv2bf16_triscv.vector.tuple_nx
 ; CHECK-NEXT:    vmv1r.v v11, v12
 ; CHECK-NEXT:    vmv1r.v v12, v13
 ; CHECK-NEXT:    vmv1r.v v13, v14
-; CHECK-NEXT:    vsetvli zero, a1, e16, mf2, ta, mu
 ; CHECK-NEXT:    vlseg7e16ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
@@ -6498,6 +6498,7 @@ entry:
 define <vscale x 4 x bfloat> @test_vlseg7ff_mask_nxv4bf16_triscv.vector.tuple_nxv8i8_7t(target("riscv.vector.tuple", <vscale x 8 x i8>, 7) %val, ptr %base, i32 %vl, <vscale x 4 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg7ff_mask_nxv4bf16_triscv.vector.tuple_nxv8i8_7t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
@@ -6505,7 +6506,6 @@ define <vscale x 4 x bfloat> @test_vlseg7ff_mask_nxv4bf16_triscv.vector.tuple_nx
 ; CHECK-NEXT:    vmv1r.v v11, v12
 ; CHECK-NEXT:    vmv1r.v v12, v13
 ; CHECK-NEXT:    vmv1r.v v13, v14
-; CHECK-NEXT:    vsetvli zero, a1, e16, m1, ta, mu
 ; CHECK-NEXT:    vlseg7e16ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
@@ -6540,6 +6540,7 @@ entry:
 define <vscale x 1 x bfloat> @test_vlseg8ff_mask_nxv1bf16_triscv.vector.tuple_nxv2i8_8t(target("riscv.vector.tuple", <vscale x 2 x i8>, 8) %val, ptr %base, i32 %vl, <vscale x 1 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg8ff_mask_nxv1bf16_triscv.vector.tuple_nxv2i8_8t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, mf4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
@@ -6548,7 +6549,6 @@ define <vscale x 1 x bfloat> @test_vlseg8ff_mask_nxv1bf16_triscv.vector.tuple_nx
 ; CHECK-NEXT:    vmv1r.v v12, v13
 ; CHECK-NEXT:    vmv1r.v v13, v14
 ; CHECK-NEXT:    vmv1r.v v14, v15
-; CHECK-NEXT:    vsetvli zero, a1, e16, mf4, ta, mu
 ; CHECK-NEXT:    vlseg8e16ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
@@ -6583,6 +6583,7 @@ entry:
 define <vscale x 2 x bfloat> @test_vlseg8ff_mask_nxv2bf16_triscv.vector.tuple_nxv4i8_8t(target("riscv.vector.tuple", <vscale x 4 x i8>, 8) %val, ptr %base, i32 %vl, <vscale x 2 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg8ff_mask_nxv2bf16_triscv.vector.tuple_nxv4i8_8t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, mf2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
@@ -6591,7 +6592,6 @@ define <vscale x 2 x bfloat> @test_vlseg8ff_mask_nxv2bf16_triscv.vector.tuple_nx
 ; CHECK-NEXT:    vmv1r.v v12, v13
 ; CHECK-NEXT:    vmv1r.v v13, v14
 ; CHECK-NEXT:    vmv1r.v v14, v15
-; CHECK-NEXT:    vsetvli zero, a1, e16, mf2, ta, mu
 ; CHECK-NEXT:    vlseg8e16ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
@@ -6626,6 +6626,7 @@ entry:
 define <vscale x 4 x bfloat> @test_vlseg8ff_mask_nxv4bf16_triscv.vector.tuple_nxv8i8_8t(target("riscv.vector.tuple", <vscale x 8 x i8>, 8) %val, ptr %base, i32 %vl, <vscale x 4 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg8ff_mask_nxv4bf16_triscv.vector.tuple_nxv8i8_8t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
@@ -6634,7 +6635,6 @@ define <vscale x 4 x bfloat> @test_vlseg8ff_mask_nxv4bf16_triscv.vector.tuple_nx
 ; CHECK-NEXT:    vmv1r.v v12, v13
 ; CHECK-NEXT:    vmv1r.v v13, v14
 ; CHECK-NEXT:    vmv1r.v v14, v15
-; CHECK-NEXT:    vsetvli zero, a1, e16, m1, ta, mu
 ; CHECK-NEXT:    vlseg8e16ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sw a0, 0(a2)
diff --git a/llvm/test/CodeGen/RISCV/rvv/vlsegff-rv64-dead.ll b/llvm/test/CodeGen/RISCV/rvv/vlsegff-rv64-dead.ll
index 3dc0db90b6d85..05a5be295cc71 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vlsegff-rv64-dead.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vlsegff-rv64-dead.ll
@@ -51,9 +51,9 @@ entry:
 define <vscale x 16 x i16> @test_vlseg2ff_mask_dead_vl(target("riscv.vector.tuple", <vscale x 32 x i8>, 2) %val, ptr %base, i64 %vl, <vscale x 16 x i1> %mask) {
 ; CHECK-LABEL: test_vlseg2ff_mask_dead_vl:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, m4, ta, mu
 ; CHECK-NEXT:    vmv4r.v v4, v8
 ; CHECK-NEXT:    vmv4r.v v8, v12
-; CHECK-NEXT:    vsetvli zero, a1, e16, m4, ta, mu
 ; CHECK-NEXT:    vlseg2e16ff.v v4, (a0), v0.t
 ; CHECK-NEXT:    ret
 entry:
diff --git a/llvm/test/CodeGen/RISCV/rvv/vlsegff-rv64.ll b/llvm/test/CodeGen/RISCV/rvv/vlsegff-rv64.ll
index 68acb3beb0686..85008ff97143f 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vlsegff-rv64.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vlsegff-rv64.ll
@@ -25,9 +25,9 @@ entry:
 define <vscale x 1 x i8> @test_vlseg2ff_mask_nxv1i8_triscv.vector.tuple_nxv1i8_2t(target("riscv.vector.tuple", <vscale x 1 x i8>, 2) %val, ptr %base, i64 %vl, <vscale x 1 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg2ff_mask_nxv1i8_triscv.vector.tuple_nxv1i8_2t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e8, mf8, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
-; CHECK-NEXT:    vsetvli zero, a1, e8, mf8, ta, mu
 ; CHECK-NEXT:    vlseg2e8ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
@@ -64,9 +64,9 @@ entry:
 define <vscale x 2 x i8> @test_vlseg2ff_mask_nxv2i8_triscv.vector.tuple_nxv2i8_2t(target("riscv.vector.tuple", <vscale x 2 x i8>, 2) %val, ptr %base, i64 %vl, <vscale x 2 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg2ff_mask_nxv2i8_triscv.vector.tuple_nxv2i8_2t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e8, mf4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
-; CHECK-NEXT:    vsetvli zero, a1, e8, mf4, ta, mu
 ; CHECK-NEXT:    vlseg2e8ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
@@ -103,9 +103,9 @@ entry:
 define <vscale x 4 x i8> @test_vlseg2ff_mask_nxv4i8_triscv.vector.tuple_nxv4i8_2t(target("riscv.vector.tuple", <vscale x 4 x i8>, 2) %val, ptr %base, i64 %vl, <vscale x 4 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg2ff_mask_nxv4i8_triscv.vector.tuple_nxv4i8_2t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e8, mf2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
-; CHECK-NEXT:    vsetvli zero, a1, e8, mf2, ta, mu
 ; CHECK-NEXT:    vlseg2e8ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
@@ -142,9 +142,9 @@ entry:
 define <vscale x 8 x i8> @test_vlseg2ff_mask_nxv8i8_triscv.vector.tuple_nxv8i8_2t(target("riscv.vector.tuple", <vscale x 8 x i8>, 2) %val, ptr %base, i64 %vl, <vscale x 8 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg2ff_mask_nxv8i8_triscv.vector.tuple_nxv8i8_2t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e8, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
-; CHECK-NEXT:    vsetvli zero, a1, e8, m1, ta, mu
 ; CHECK-NEXT:    vlseg2e8ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
@@ -181,9 +181,9 @@ entry:
 define <vscale x 16 x i8> @test_vlseg2ff_mask_nxv16i8_triscv.vector.tuple_nxv16i8_2t(target("riscv.vector.tuple", <vscale x 16 x i8>, 2) %val, ptr %base, i64 %vl, <vscale x 16 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg2ff_mask_nxv16i8_triscv.vector.tuple_nxv16i8_2t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e8, m2, ta, mu
 ; CHECK-NEXT:    vmv2r.v v6, v8
 ; CHECK-NEXT:    vmv2r.v v8, v10
-; CHECK-NEXT:    vsetvli zero, a1, e8, m2, ta, mu
 ; CHECK-NEXT:    vlseg2e8ff.v v6, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
@@ -220,9 +220,9 @@ entry:
 define <vscale x 32 x i8> @test_vlseg2ff_mask_nxv32i8_triscv.vector.tuple_nxv32i8_2t(target("riscv.vector.tuple", <vscale x 32 x i8>, 2) %val, ptr %base, i64 %vl, <vscale x 32 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg2ff_mask_nxv32i8_triscv.vector.tuple_nxv32i8_2t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e8, m4, ta, mu
 ; CHECK-NEXT:    vmv4r.v v4, v8
 ; CHECK-NEXT:    vmv4r.v v8, v12
-; CHECK-NEXT:    vsetvli zero, a1, e8, m4, ta, mu
 ; CHECK-NEXT:    vlseg2e8ff.v v4, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
@@ -259,10 +259,10 @@ entry:
 define <vscale x 1 x i8> @test_vlseg3ff_mask_nxv1i8_triscv.vector.tuple_nxv1i8_3t(target("riscv.vector.tuple", <vscale x 1 x i8>, 3) %val, ptr %base, i64 %vl, <vscale x 1 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg3ff_mask_nxv1i8_triscv.vector.tuple_nxv1i8_3t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e8, mf8, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
-; CHECK-NEXT:    vsetvli zero, a1, e8, mf8, ta, mu
 ; CHECK-NEXT:    vlseg3e8ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
@@ -299,10 +299,10 @@ entry:
 define <vscale x 2 x i8> @test_vlseg3ff_mask_nxv2i8_triscv.vector.tuple_nxv2i8_3t(target("riscv.vector.tuple", <vscale x 2 x i8>, 3) %val, ptr %base, i64 %vl, <vscale x 2 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg3ff_mask_nxv2i8_triscv.vector.tuple_nxv2i8_3t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e8, mf4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
-; CHECK-NEXT:    vsetvli zero, a1, e8, mf4, ta, mu
 ; CHECK-NEXT:    vlseg3e8ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
@@ -339,10 +339,10 @@ entry:
 define <vscale x 4 x i8> @test_vlseg3ff_mask_nxv4i8_triscv.vector.tuple_nxv4i8_3t(target("riscv.vector.tuple", <vscale x 4 x i8>, 3) %val, ptr %base, i64 %vl, <vscale x 4 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg3ff_mask_nxv4i8_triscv.vector.tuple_nxv4i8_3t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e8, mf2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
-; CHECK-NEXT:    vsetvli zero, a1, e8, mf2, ta, mu
 ; CHECK-NEXT:    vlseg3e8ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
@@ -379,10 +379,10 @@ entry:
 define <vscale x 8 x i8> @test_vlseg3ff_mask_nxv8i8_triscv.vector.tuple_nxv8i8_3t(target("riscv.vector.tuple", <vscale x 8 x i8>, 3) %val, ptr %base, i64 %vl, <vscale x 8 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg3ff_mask_nxv8i8_triscv.vector.tuple_nxv8i8_3t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e8, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
-; CHECK-NEXT:    vsetvli zero, a1, e8, m1, ta, mu
 ; CHECK-NEXT:    vlseg3e8ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
@@ -419,10 +419,10 @@ entry:
 define <vscale x 16 x i8> @test_vlseg3ff_mask_nxv16i8_triscv.vector.tuple_nxv16i8_3t(target("riscv.vector.tuple", <vscale x 16 x i8>, 3) %val, ptr %base, i64 %vl, <vscale x 16 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg3ff_mask_nxv16i8_triscv.vector.tuple_nxv16i8_3t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e8, m2, ta, mu
 ; CHECK-NEXT:    vmv2r.v v6, v8
 ; CHECK-NEXT:    vmv2r.v v8, v10
 ; CHECK-NEXT:    vmv2r.v v10, v12
-; CHECK-NEXT:    vsetvli zero, a1, e8, m2, ta, mu
 ; CHECK-NEXT:    vlseg3e8ff.v v6, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
@@ -459,11 +459,11 @@ entry:
 define <vscale x 1 x i8> @test_vlseg4ff_mask_nxv1i8_triscv.vector.tuple_nxv1i8_4t(target("riscv.vector.tuple", <vscale x 1 x i8>, 4) %val, ptr %base, i64 %vl, <vscale x 1 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg4ff_mask_nxv1i8_triscv.vector.tuple_nxv1i8_4t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e8, mf8, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
 ; CHECK-NEXT:    vmv1r.v v10, v11
-; CHECK-NEXT:    vsetvli zero, a1, e8, mf8, ta, mu
 ; CHECK-NEXT:    vlseg4e8ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
@@ -500,11 +500,11 @@ entry:
 define <vscale x 2 x i8> @test_vlseg4ff_mask_nxv2i8_triscv.vector.tuple_nxv2i8_4t(target("riscv.vector.tuple", <vscale x 2 x i8>, 4) %val, ptr %base, i64 %vl, <vscale x 2 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg4ff_mask_nxv2i8_triscv.vector.tuple_nxv2i8_4t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e8, mf4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
 ; CHECK-NEXT:    vmv1r.v v10, v11
-; CHECK-NEXT:    vsetvli zero, a1, e8, mf4, ta, mu
 ; CHECK-NEXT:    vlseg4e8ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
@@ -541,11 +541,11 @@ entry:
 define <vscale x 4 x i8> @test_vlseg4ff_mask_nxv4i8_triscv.vector.tuple_nxv4i8_4t(target("riscv.vector.tuple", <vscale x 4 x i8>, 4) %val, ptr %base, i64 %vl, <vscale x 4 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg4ff_mask_nxv4i8_triscv.vector.tuple_nxv4i8_4t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e8, mf2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
 ; CHECK-NEXT:    vmv1r.v v10, v11
-; CHECK-NEXT:    vsetvli zero, a1, e8, mf2, ta, mu
 ; CHECK-NEXT:    vlseg4e8ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
@@ -582,11 +582,11 @@ entry:
 define <vscale x 8 x i8> @test_vlseg4ff_mask_nxv8i8_triscv.vector.tuple_nxv8i8_4t(target("riscv.vector.tuple", <vscale x 8 x i8>, 4) %val, ptr %base, i64 %vl, <vscale x 8 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg4ff_mask_nxv8i8_triscv.vector.tuple_nxv8i8_4t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e8, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
 ; CHECK-NEXT:    vmv1r.v v10, v11
-; CHECK-NEXT:    vsetvli zero, a1, e8, m1, ta, mu
 ; CHECK-NEXT:    vlseg4e8ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
@@ -623,11 +623,11 @@ entry:
 define <vscale x 16 x i8> @test_vlseg4ff_mask_nxv16i8_triscv.vector.tuple_nxv16i8_4t(target("riscv.vector.tuple", <vscale x 16 x i8>, 4) %val, ptr %base, i64 %vl, <vscale x 16 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg4ff_mask_nxv16i8_triscv.vector.tuple_nxv16i8_4t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e8, m2, ta, mu
 ; CHECK-NEXT:    vmv2r.v v6, v8
 ; CHECK-NEXT:    vmv2r.v v8, v10
 ; CHECK-NEXT:    vmv2r.v v10, v12
 ; CHECK-NEXT:    vmv2r.v v12, v14
-; CHECK-NEXT:    vsetvli zero, a1, e8, m2, ta, mu
 ; CHECK-NEXT:    vlseg4e8ff.v v6, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
@@ -664,12 +664,12 @@ entry:
 define <vscale x 1 x i8> @test_vlseg5ff_mask_nxv1i8_triscv.vector.tuple_nxv1i8_5t(target("riscv.vector.tuple", <vscale x 1 x i8>, 5) %val, ptr %base, i64 %vl, <vscale x 1 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg5ff_mask_nxv1i8_triscv.vector.tuple_nxv1i8_5t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e8, mf8, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
 ; CHECK-NEXT:    vmv1r.v v10, v11
 ; CHECK-NEXT:    vmv1r.v v11, v12
-; CHECK-NEXT:    vsetvli zero, a1, e8, mf8, ta, mu
 ; CHECK-NEXT:    vlseg5e8ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
@@ -706,12 +706,12 @@ entry:
 define <vscale x 2 x i8> @test_vlseg5ff_mask_nxv2i8_triscv.vector.tuple_nxv2i8_5t(target("riscv.vector.tuple", <vscale x 2 x i8>, 5) %val, ptr %base, i64 %vl, <vscale x 2 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg5ff_mask_nxv2i8_triscv.vector.tuple_nxv2i8_5t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e8, mf4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
 ; CHECK-NEXT:    vmv1r.v v10, v11
 ; CHECK-NEXT:    vmv1r.v v11, v12
-; CHECK-NEXT:    vsetvli zero, a1, e8, mf4, ta, mu
 ; CHECK-NEXT:    vlseg5e8ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
@@ -748,12 +748,12 @@ entry:
 define <vscale x 4 x i8> @test_vlseg5ff_mask_nxv4i8_triscv.vector.tuple_nxv4i8_5t(target("riscv.vector.tuple", <vscale x 4 x i8>, 5) %val, ptr %base, i64 %vl, <vscale x 4 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg5ff_mask_nxv4i8_triscv.vector.tuple_nxv4i8_5t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e8, mf2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
 ; CHECK-NEXT:    vmv1r.v v10, v11
 ; CHECK-NEXT:    vmv1r.v v11, v12
-; CHECK-NEXT:    vsetvli zero, a1, e8, mf2, ta, mu
 ; CHECK-NEXT:    vlseg5e8ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
@@ -790,12 +790,12 @@ entry:
 define <vscale x 8 x i8> @test_vlseg5ff_mask_nxv8i8_triscv.vector.tuple_nxv8i8_5t(target("riscv.vector.tuple", <vscale x 8 x i8>, 5) %val, ptr %base, i64 %vl, <vscale x 8 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg5ff_mask_nxv8i8_triscv.vector.tuple_nxv8i8_5t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e8, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
 ; CHECK-NEXT:    vmv1r.v v10, v11
 ; CHECK-NEXT:    vmv1r.v v11, v12
-; CHECK-NEXT:    vsetvli zero, a1, e8, m1, ta, mu
 ; CHECK-NEXT:    vlseg5e8ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
@@ -832,13 +832,13 @@ entry:
 define <vscale x 1 x i8> @test_vlseg6ff_mask_nxv1i8_triscv.vector.tuple_nxv1i8_6t(target("riscv.vector.tuple", <vscale x 1 x i8>, 6) %val, ptr %base, i64 %vl, <vscale x 1 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg6ff_mask_nxv1i8_triscv.vector.tuple_nxv1i8_6t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e8, mf8, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
 ; CHECK-NEXT:    vmv1r.v v10, v11
 ; CHECK-NEXT:    vmv1r.v v11, v12
 ; CHECK-NEXT:    vmv1r.v v12, v13
-; CHECK-NEXT:    vsetvli zero, a1, e8, mf8, ta, mu
 ; CHECK-NEXT:    vlseg6e8ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
@@ -875,13 +875,13 @@ entry:
 define <vscale x 2 x i8> @test_vlseg6ff_mask_nxv2i8_triscv.vector.tuple_nxv2i8_6t(target("riscv.vector.tuple", <vscale x 2 x i8>, 6) %val, ptr %base, i64 %vl, <vscale x 2 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg6ff_mask_nxv2i8_triscv.vector.tuple_nxv2i8_6t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e8, mf4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
 ; CHECK-NEXT:    vmv1r.v v10, v11
 ; CHECK-NEXT:    vmv1r.v v11, v12
 ; CHECK-NEXT:    vmv1r.v v12, v13
-; CHECK-NEXT:    vsetvli zero, a1, e8, mf4, ta, mu
 ; CHECK-NEXT:    vlseg6e8ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
@@ -918,13 +918,13 @@ entry:
 define <vscale x 4 x i8> @test_vlseg6ff_mask_nxv4i8_triscv.vector.tuple_nxv4i8_6t(target("riscv.vector.tuple", <vscale x 4 x i8>, 6) %val, ptr %base, i64 %vl, <vscale x 4 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg6ff_mask_nxv4i8_triscv.vector.tuple_nxv4i8_6t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e8, mf2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
 ; CHECK-NEXT:    vmv1r.v v10, v11
 ; CHECK-NEXT:    vmv1r.v v11, v12
 ; CHECK-NEXT:    vmv1r.v v12, v13
-; CHECK-NEXT:    vsetvli zero, a1, e8, mf2, ta, mu
 ; CHECK-NEXT:    vlseg6e8ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
@@ -961,13 +961,13 @@ entry:
 define <vscale x 8 x i8> @test_vlseg6ff_mask_nxv8i8_triscv.vector.tuple_nxv8i8_6t(target("riscv.vector.tuple", <vscale x 8 x i8>, 6) %val, ptr %base, i64 %vl, <vscale x 8 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg6ff_mask_nxv8i8_triscv.vector.tuple_nxv8i8_6t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e8, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
 ; CHECK-NEXT:    vmv1r.v v10, v11
 ; CHECK-NEXT:    vmv1r.v v11, v12
 ; CHECK-NEXT:    vmv1r.v v12, v13
-; CHECK-NEXT:    vsetvli zero, a1, e8, m1, ta, mu
 ; CHECK-NEXT:    vlseg6e8ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
@@ -1004,6 +1004,7 @@ entry:
 define <vscale x 1 x i8> @test_vlseg7ff_mask_nxv1i8_triscv.vector.tuple_nxv1i8_7t(target("riscv.vector.tuple", <vscale x 1 x i8>, 7) %val, ptr %base, i64 %vl, <vscale x 1 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg7ff_mask_nxv1i8_triscv.vector.tuple_nxv1i8_7t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e8, mf8, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
@@ -1011,7 +1012,6 @@ define <vscale x 1 x i8> @test_vlseg7ff_mask_nxv1i8_triscv.vector.tuple_nxv1i8_7
 ; CHECK-NEXT:    vmv1r.v v11, v12
 ; CHECK-NEXT:    vmv1r.v v12, v13
 ; CHECK-NEXT:    vmv1r.v v13, v14
-; CHECK-NEXT:    vsetvli zero, a1, e8, mf8, ta, mu
 ; CHECK-NEXT:    vlseg7e8ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
@@ -1048,6 +1048,7 @@ entry:
 define <vscale x 2 x i8> @test_vlseg7ff_mask_nxv2i8_triscv.vector.tuple_nxv2i8_7t(target("riscv.vector.tuple", <vscale x 2 x i8>, 7) %val, ptr %base, i64 %vl, <vscale x 2 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg7ff_mask_nxv2i8_triscv.vector.tuple_nxv2i8_7t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e8, mf4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
@@ -1055,7 +1056,6 @@ define <vscale x 2 x i8> @test_vlseg7ff_mask_nxv2i8_triscv.vector.tuple_nxv2i8_7
 ; CHECK-NEXT:    vmv1r.v v11, v12
 ; CHECK-NEXT:    vmv1r.v v12, v13
 ; CHECK-NEXT:    vmv1r.v v13, v14
-; CHECK-NEXT:    vsetvli zero, a1, e8, mf4, ta, mu
 ; CHECK-NEXT:    vlseg7e8ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
@@ -1092,6 +1092,7 @@ entry:
 define <vscale x 4 x i8> @test_vlseg7ff_mask_nxv4i8_triscv.vector.tuple_nxv4i8_7t(target("riscv.vector.tuple", <vscale x 4 x i8>, 7) %val, ptr %base, i64 %vl, <vscale x 4 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg7ff_mask_nxv4i8_triscv.vector.tuple_nxv4i8_7t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e8, mf2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
@@ -1099,7 +1100,6 @@ define <vscale x 4 x i8> @test_vlseg7ff_mask_nxv4i8_triscv.vector.tuple_nxv4i8_7
 ; CHECK-NEXT:    vmv1r.v v11, v12
 ; CHECK-NEXT:    vmv1r.v v12, v13
 ; CHECK-NEXT:    vmv1r.v v13, v14
-; CHECK-NEXT:    vsetvli zero, a1, e8, mf2, ta, mu
 ; CHECK-NEXT:    vlseg7e8ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
@@ -1136,6 +1136,7 @@ entry:
 define <vscale x 8 x i8> @test_vlseg7ff_mask_nxv8i8_triscv.vector.tuple_nxv8i8_7t(target("riscv.vector.tuple", <vscale x 8 x i8>, 7) %val, ptr %base, i64 %vl, <vscale x 8 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg7ff_mask_nxv8i8_triscv.vector.tuple_nxv8i8_7t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e8, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
@@ -1143,7 +1144,6 @@ define <vscale x 8 x i8> @test_vlseg7ff_mask_nxv8i8_triscv.vector.tuple_nxv8i8_7
 ; CHECK-NEXT:    vmv1r.v v11, v12
 ; CHECK-NEXT:    vmv1r.v v12, v13
 ; CHECK-NEXT:    vmv1r.v v13, v14
-; CHECK-NEXT:    vsetvli zero, a1, e8, m1, ta, mu
 ; CHECK-NEXT:    vlseg7e8ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
@@ -1180,6 +1180,7 @@ entry:
 define <vscale x 1 x i8> @test_vlseg8ff_mask_nxv1i8_triscv.vector.tuple_nxv1i8_8t(target("riscv.vector.tuple", <vscale x 1 x i8>, 8) %val, ptr %base, i64 %vl, <vscale x 1 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg8ff_mask_nxv1i8_triscv.vector.tuple_nxv1i8_8t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e8, mf8, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
@@ -1188,7 +1189,6 @@ define <vscale x 1 x i8> @test_vlseg8ff_mask_nxv1i8_triscv.vector.tuple_nxv1i8_8
 ; CHECK-NEXT:    vmv1r.v v12, v13
 ; CHECK-NEXT:    vmv1r.v v13, v14
 ; CHECK-NEXT:    vmv1r.v v14, v15
-; CHECK-NEXT:    vsetvli zero, a1, e8, mf8, ta, mu
 ; CHECK-NEXT:    vlseg8e8ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
@@ -1225,6 +1225,7 @@ entry:
 define <vscale x 2 x i8> @test_vlseg8ff_mask_nxv2i8_triscv.vector.tuple_nxv2i8_8t(target("riscv.vector.tuple", <vscale x 2 x i8>, 8) %val, ptr %base, i64 %vl, <vscale x 2 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg8ff_mask_nxv2i8_triscv.vector.tuple_nxv2i8_8t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e8, mf4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
@@ -1233,7 +1234,6 @@ define <vscale x 2 x i8> @test_vlseg8ff_mask_nxv2i8_triscv.vector.tuple_nxv2i8_8
 ; CHECK-NEXT:    vmv1r.v v12, v13
 ; CHECK-NEXT:    vmv1r.v v13, v14
 ; CHECK-NEXT:    vmv1r.v v14, v15
-; CHECK-NEXT:    vsetvli zero, a1, e8, mf4, ta, mu
 ; CHECK-NEXT:    vlseg8e8ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
@@ -1270,6 +1270,7 @@ entry:
 define <vscale x 4 x i8> @test_vlseg8ff_mask_nxv4i8_triscv.vector.tuple_nxv4i8_8t(target("riscv.vector.tuple", <vscale x 4 x i8>, 8) %val, ptr %base, i64 %vl, <vscale x 4 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg8ff_mask_nxv4i8_triscv.vector.tuple_nxv4i8_8t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e8, mf2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
@@ -1278,7 +1279,6 @@ define <vscale x 4 x i8> @test_vlseg8ff_mask_nxv4i8_triscv.vector.tuple_nxv4i8_8
 ; CHECK-NEXT:    vmv1r.v v12, v13
 ; CHECK-NEXT:    vmv1r.v v13, v14
 ; CHECK-NEXT:    vmv1r.v v14, v15
-; CHECK-NEXT:    vsetvli zero, a1, e8, mf2, ta, mu
 ; CHECK-NEXT:    vlseg8e8ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
@@ -1315,6 +1315,7 @@ entry:
 define <vscale x 8 x i8> @test_vlseg8ff_mask_nxv8i8_triscv.vector.tuple_nxv8i8_8t(target("riscv.vector.tuple", <vscale x 8 x i8>, 8) %val, ptr %base, i64 %vl, <vscale x 8 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg8ff_mask_nxv8i8_triscv.vector.tuple_nxv8i8_8t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e8, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
@@ -1323,7 +1324,6 @@ define <vscale x 8 x i8> @test_vlseg8ff_mask_nxv8i8_triscv.vector.tuple_nxv8i8_8
 ; CHECK-NEXT:    vmv1r.v v12, v13
 ; CHECK-NEXT:    vmv1r.v v13, v14
 ; CHECK-NEXT:    vmv1r.v v14, v15
-; CHECK-NEXT:    vsetvli zero, a1, e8, m1, ta, mu
 ; CHECK-NEXT:    vlseg8e8ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
@@ -1359,9 +1359,9 @@ entry:
 define <vscale x 1 x i16> @test_vlseg2ff_mask_nxv1i16_triscv.vector.tuple_nxv2i8_2t(target("riscv.vector.tuple", <vscale x 2 x i8>, 2) %val, ptr %base, i64 %vl, <vscale x 1 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg2ff_mask_nxv1i16_triscv.vector.tuple_nxv2i8_2t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, mf4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
-; CHECK-NEXT:    vsetvli zero, a1, e16, mf4, ta, mu
 ; CHECK-NEXT:    vlseg2e16ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
@@ -1397,9 +1397,9 @@ entry:
 define <vscale x 2 x i16> @test_vlseg2ff_mask_nxv2i16_triscv.vector.tuple_nxv4i8_2t(target("riscv.vector.tuple", <vscale x 4 x i8>, 2) %val, ptr %base, i64 %vl, <vscale x 2 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg2ff_mask_nxv2i16_triscv.vector.tuple_nxv4i8_2t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, mf2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
-; CHECK-NEXT:    vsetvli zero, a1, e16, mf2, ta, mu
 ; CHECK-NEXT:    vlseg2e16ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
@@ -1435,9 +1435,9 @@ entry:
 define <vscale x 4 x i16> @test_vlseg2ff_mask_nxv4i16_triscv.vector.tuple_nxv8i8_2t(target("riscv.vector.tuple", <vscale x 8 x i8>, 2) %val, ptr %base, i64 %vl, <vscale x 4 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg2ff_mask_nxv4i16_triscv.vector.tuple_nxv8i8_2t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
-; CHECK-NEXT:    vsetvli zero, a1, e16, m1, ta, mu
 ; CHECK-NEXT:    vlseg2e16ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
@@ -1473,9 +1473,9 @@ entry:
 define <vscale x 8 x i16> @test_vlseg2ff_mask_nxv8i16_triscv.vector.tuple_nxv16i8_2t(target("riscv.vector.tuple", <vscale x 16 x i8>, 2) %val, ptr %base, i64 %vl, <vscale x 8 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg2ff_mask_nxv8i16_triscv.vector.tuple_nxv16i8_2t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, m2, ta, mu
 ; CHECK-NEXT:    vmv2r.v v6, v8
 ; CHECK-NEXT:    vmv2r.v v8, v10
-; CHECK-NEXT:    vsetvli zero, a1, e16, m2, ta, mu
 ; CHECK-NEXT:    vlseg2e16ff.v v6, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
@@ -1511,9 +1511,9 @@ entry:
 define <vscale x 16 x i16> @test_vlseg2ff_mask_nxv16i16_triscv.vector.tuple_nxv32i8_2t(target("riscv.vector.tuple", <vscale x 32 x i8>, 2) %val, ptr %base, i64 %vl, <vscale x 16 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg2ff_mask_nxv16i16_triscv.vector.tuple_nxv32i8_2t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, m4, ta, mu
 ; CHECK-NEXT:    vmv4r.v v4, v8
 ; CHECK-NEXT:    vmv4r.v v8, v12
-; CHECK-NEXT:    vsetvli zero, a1, e16, m4, ta, mu
 ; CHECK-NEXT:    vlseg2e16ff.v v4, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
@@ -1549,10 +1549,10 @@ entry:
 define <vscale x 1 x i16> @test_vlseg3ff_mask_nxv1i16_triscv.vector.tuple_nxv2i8_3t(target("riscv.vector.tuple", <vscale x 2 x i8>, 3) %val, ptr %base, i64 %vl, <vscale x 1 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg3ff_mask_nxv1i16_triscv.vector.tuple_nxv2i8_3t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, mf4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
-; CHECK-NEXT:    vsetvli zero, a1, e16, mf4, ta, mu
 ; CHECK-NEXT:    vlseg3e16ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
@@ -1588,10 +1588,10 @@ entry:
 define <vscale x 2 x i16> @test_vlseg3ff_mask_nxv2i16_triscv.vector.tuple_nxv4i8_3t(target("riscv.vector.tuple", <vscale x 4 x i8>, 3) %val, ptr %base, i64 %vl, <vscale x 2 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg3ff_mask_nxv2i16_triscv.vector.tuple_nxv4i8_3t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, mf2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
-; CHECK-NEXT:    vsetvli zero, a1, e16, mf2, ta, mu
 ; CHECK-NEXT:    vlseg3e16ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
@@ -1627,10 +1627,10 @@ entry:
 define <vscale x 4 x i16> @test_vlseg3ff_mask_nxv4i16_triscv.vector.tuple_nxv8i8_3t(target("riscv.vector.tuple", <vscale x 8 x i8>, 3) %val, ptr %base, i64 %vl, <vscale x 4 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg3ff_mask_nxv4i16_triscv.vector.tuple_nxv8i8_3t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
-; CHECK-NEXT:    vsetvli zero, a1, e16, m1, ta, mu
 ; CHECK-NEXT:    vlseg3e16ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
@@ -1666,10 +1666,10 @@ entry:
 define <vscale x 8 x i16> @test_vlseg3ff_mask_nxv8i16_triscv.vector.tuple_nxv16i8_3t(target("riscv.vector.tuple", <vscale x 16 x i8>, 3) %val, ptr %base, i64 %vl, <vscale x 8 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg3ff_mask_nxv8i16_triscv.vector.tuple_nxv16i8_3t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, m2, ta, mu
 ; CHECK-NEXT:    vmv2r.v v6, v8
 ; CHECK-NEXT:    vmv2r.v v8, v10
 ; CHECK-NEXT:    vmv2r.v v10, v12
-; CHECK-NEXT:    vsetvli zero, a1, e16, m2, ta, mu
 ; CHECK-NEXT:    vlseg3e16ff.v v6, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
@@ -1705,11 +1705,11 @@ entry:
 define <vscale x 1 x i16> @test_vlseg4ff_mask_nxv1i16_triscv.vector.tuple_nxv2i8_4t(target("riscv.vector.tuple", <vscale x 2 x i8>, 4) %val, ptr %base, i64 %vl, <vscale x 1 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg4ff_mask_nxv1i16_triscv.vector.tuple_nxv2i8_4t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, mf4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
 ; CHECK-NEXT:    vmv1r.v v10, v11
-; CHECK-NEXT:    vsetvli zero, a1, e16, mf4, ta, mu
 ; CHECK-NEXT:    vlseg4e16ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
@@ -1745,11 +1745,11 @@ entry:
 define <vscale x 2 x i16> @test_vlseg4ff_mask_nxv2i16_triscv.vector.tuple_nxv4i8_4t(target("riscv.vector.tuple", <vscale x 4 x i8>, 4) %val, ptr %base, i64 %vl, <vscale x 2 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg4ff_mask_nxv2i16_triscv.vector.tuple_nxv4i8_4t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, mf2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
 ; CHECK-NEXT:    vmv1r.v v10, v11
-; CHECK-NEXT:    vsetvli zero, a1, e16, mf2, ta, mu
 ; CHECK-NEXT:    vlseg4e16ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
@@ -1785,11 +1785,11 @@ entry:
 define <vscale x 4 x i16> @test_vlseg4ff_mask_nxv4i16_triscv.vector.tuple_nxv8i8_4t(target("riscv.vector.tuple", <vscale x 8 x i8>, 4) %val, ptr %base, i64 %vl, <vscale x 4 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg4ff_mask_nxv4i16_triscv.vector.tuple_nxv8i8_4t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
 ; CHECK-NEXT:    vmv1r.v v10, v11
-; CHECK-NEXT:    vsetvli zero, a1, e16, m1, ta, mu
 ; CHECK-NEXT:    vlseg4e16ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
@@ -1825,11 +1825,11 @@ entry:
 define <vscale x 8 x i16> @test_vlseg4ff_mask_nxv8i16_triscv.vector.tuple_nxv16i8_4t(target("riscv.vector.tuple", <vscale x 16 x i8>, 4) %val, ptr %base, i64 %vl, <vscale x 8 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg4ff_mask_nxv8i16_triscv.vector.tuple_nxv16i8_4t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, m2, ta, mu
 ; CHECK-NEXT:    vmv2r.v v6, v8
 ; CHECK-NEXT:    vmv2r.v v8, v10
 ; CHECK-NEXT:    vmv2r.v v10, v12
 ; CHECK-NEXT:    vmv2r.v v12, v14
-; CHECK-NEXT:    vsetvli zero, a1, e16, m2, ta, mu
 ; CHECK-NEXT:    vlseg4e16ff.v v6, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
@@ -1865,12 +1865,12 @@ entry:
 define <vscale x 1 x i16> @test_vlseg5ff_mask_nxv1i16_triscv.vector.tuple_nxv2i8_5t(target("riscv.vector.tuple", <vscale x 2 x i8>, 5) %val, ptr %base, i64 %vl, <vscale x 1 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg5ff_mask_nxv1i16_triscv.vector.tuple_nxv2i8_5t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, mf4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
 ; CHECK-NEXT:    vmv1r.v v10, v11
 ; CHECK-NEXT:    vmv1r.v v11, v12
-; CHECK-NEXT:    vsetvli zero, a1, e16, mf4, ta, mu
 ; CHECK-NEXT:    vlseg5e16ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
@@ -1906,12 +1906,12 @@ entry:
 define <vscale x 2 x i16> @test_vlseg5ff_mask_nxv2i16_triscv.vector.tuple_nxv4i8_5t(target("riscv.vector.tuple", <vscale x 4 x i8>, 5) %val, ptr %base, i64 %vl, <vscale x 2 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg5ff_mask_nxv2i16_triscv.vector.tuple_nxv4i8_5t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, mf2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
 ; CHECK-NEXT:    vmv1r.v v10, v11
 ; CHECK-NEXT:    vmv1r.v v11, v12
-; CHECK-NEXT:    vsetvli zero, a1, e16, mf2, ta, mu
 ; CHECK-NEXT:    vlseg5e16ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
@@ -1947,12 +1947,12 @@ entry:
 define <vscale x 4 x i16> @test_vlseg5ff_mask_nxv4i16_triscv.vector.tuple_nxv8i8_5t(target("riscv.vector.tuple", <vscale x 8 x i8>, 5) %val, ptr %base, i64 %vl, <vscale x 4 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg5ff_mask_nxv4i16_triscv.vector.tuple_nxv8i8_5t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
 ; CHECK-NEXT:    vmv1r.v v10, v11
 ; CHECK-NEXT:    vmv1r.v v11, v12
-; CHECK-NEXT:    vsetvli zero, a1, e16, m1, ta, mu
 ; CHECK-NEXT:    vlseg5e16ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
@@ -1988,13 +1988,13 @@ entry:
 define <vscale x 1 x i16> @test_vlseg6ff_mask_nxv1i16_triscv.vector.tuple_nxv2i8_6t(target("riscv.vector.tuple", <vscale x 2 x i8>, 6) %val, ptr %base, i64 %vl, <vscale x 1 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg6ff_mask_nxv1i16_triscv.vector.tuple_nxv2i8_6t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, mf4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
 ; CHECK-NEXT:    vmv1r.v v10, v11
 ; CHECK-NEXT:    vmv1r.v v11, v12
 ; CHECK-NEXT:    vmv1r.v v12, v13
-; CHECK-NEXT:    vsetvli zero, a1, e16, mf4, ta, mu
 ; CHECK-NEXT:    vlseg6e16ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
@@ -2030,13 +2030,13 @@ entry:
 define <vscale x 2 x i16> @test_vlseg6ff_mask_nxv2i16_triscv.vector.tuple_nxv4i8_6t(target("riscv.vector.tuple", <vscale x 4 x i8>, 6) %val, ptr %base, i64 %vl, <vscale x 2 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg6ff_mask_nxv2i16_triscv.vector.tuple_nxv4i8_6t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, mf2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
 ; CHECK-NEXT:    vmv1r.v v10, v11
 ; CHECK-NEXT:    vmv1r.v v11, v12
 ; CHECK-NEXT:    vmv1r.v v12, v13
-; CHECK-NEXT:    vsetvli zero, a1, e16, mf2, ta, mu
 ; CHECK-NEXT:    vlseg6e16ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
@@ -2072,13 +2072,13 @@ entry:
 define <vscale x 4 x i16> @test_vlseg6ff_mask_nxv4i16_triscv.vector.tuple_nxv8i8_6t(target("riscv.vector.tuple", <vscale x 8 x i8>, 6) %val, ptr %base, i64 %vl, <vscale x 4 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg6ff_mask_nxv4i16_triscv.vector.tuple_nxv8i8_6t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
 ; CHECK-NEXT:    vmv1r.v v10, v11
 ; CHECK-NEXT:    vmv1r.v v11, v12
 ; CHECK-NEXT:    vmv1r.v v12, v13
-; CHECK-NEXT:    vsetvli zero, a1, e16, m1, ta, mu
 ; CHECK-NEXT:    vlseg6e16ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
@@ -2114,6 +2114,7 @@ entry:
 define <vscale x 1 x i16> @test_vlseg7ff_mask_nxv1i16_triscv.vector.tuple_nxv2i8_7t(target("riscv.vector.tuple", <vscale x 2 x i8>, 7) %val, ptr %base, i64 %vl, <vscale x 1 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg7ff_mask_nxv1i16_triscv.vector.tuple_nxv2i8_7t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, mf4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
@@ -2121,7 +2122,6 @@ define <vscale x 1 x i16> @test_vlseg7ff_mask_nxv1i16_triscv.vector.tuple_nxv2i8
 ; CHECK-NEXT:    vmv1r.v v11, v12
 ; CHECK-NEXT:    vmv1r.v v12, v13
 ; CHECK-NEXT:    vmv1r.v v13, v14
-; CHECK-NEXT:    vsetvli zero, a1, e16, mf4, ta, mu
 ; CHECK-NEXT:    vlseg7e16ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
@@ -2157,6 +2157,7 @@ entry:
 define <vscale x 2 x i16> @test_vlseg7ff_mask_nxv2i16_triscv.vector.tuple_nxv4i8_7t(target("riscv.vector.tuple", <vscale x 4 x i8>, 7) %val, ptr %base, i64 %vl, <vscale x 2 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg7ff_mask_nxv2i16_triscv.vector.tuple_nxv4i8_7t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, mf2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
@@ -2164,7 +2165,6 @@ define <vscale x 2 x i16> @test_vlseg7ff_mask_nxv2i16_triscv.vector.tuple_nxv4i8
 ; CHECK-NEXT:    vmv1r.v v11, v12
 ; CHECK-NEXT:    vmv1r.v v12, v13
 ; CHECK-NEXT:    vmv1r.v v13, v14
-; CHECK-NEXT:    vsetvli zero, a1, e16, mf2, ta, mu
 ; CHECK-NEXT:    vlseg7e16ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
@@ -2200,6 +2200,7 @@ entry:
 define <vscale x 4 x i16> @test_vlseg7ff_mask_nxv4i16_triscv.vector.tuple_nxv8i8_7t(target("riscv.vector.tuple", <vscale x 8 x i8>, 7) %val, ptr %base, i64 %vl, <vscale x 4 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg7ff_mask_nxv4i16_triscv.vector.tuple_nxv8i8_7t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
@@ -2207,7 +2208,6 @@ define <vscale x 4 x i16> @test_vlseg7ff_mask_nxv4i16_triscv.vector.tuple_nxv8i8
 ; CHECK-NEXT:    vmv1r.v v11, v12
 ; CHECK-NEXT:    vmv1r.v v12, v13
 ; CHECK-NEXT:    vmv1r.v v13, v14
-; CHECK-NEXT:    vsetvli zero, a1, e16, m1, ta, mu
 ; CHECK-NEXT:    vlseg7e16ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
@@ -2243,6 +2243,7 @@ entry:
 define <vscale x 1 x i16> @test_vlseg8ff_mask_nxv1i16_triscv.vector.tuple_nxv2i8_8t(target("riscv.vector.tuple", <vscale x 2 x i8>, 8) %val, ptr %base, i64 %vl, <vscale x 1 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg8ff_mask_nxv1i16_triscv.vector.tuple_nxv2i8_8t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, mf4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
@@ -2251,7 +2252,6 @@ define <vscale x 1 x i16> @test_vlseg8ff_mask_nxv1i16_triscv.vector.tuple_nxv2i8
 ; CHECK-NEXT:    vmv1r.v v12, v13
 ; CHECK-NEXT:    vmv1r.v v13, v14
 ; CHECK-NEXT:    vmv1r.v v14, v15
-; CHECK-NEXT:    vsetvli zero, a1, e16, mf4, ta, mu
 ; CHECK-NEXT:    vlseg8e16ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
@@ -2287,6 +2287,7 @@ entry:
 define <vscale x 2 x i16> @test_vlseg8ff_mask_nxv2i16_triscv.vector.tuple_nxv4i8_8t(target("riscv.vector.tuple", <vscale x 4 x i8>, 8) %val, ptr %base, i64 %vl, <vscale x 2 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg8ff_mask_nxv2i16_triscv.vector.tuple_nxv4i8_8t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, mf2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
@@ -2295,7 +2296,6 @@ define <vscale x 2 x i16> @test_vlseg8ff_mask_nxv2i16_triscv.vector.tuple_nxv4i8
 ; CHECK-NEXT:    vmv1r.v v12, v13
 ; CHECK-NEXT:    vmv1r.v v13, v14
 ; CHECK-NEXT:    vmv1r.v v14, v15
-; CHECK-NEXT:    vsetvli zero, a1, e16, mf2, ta, mu
 ; CHECK-NEXT:    vlseg8e16ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
@@ -2331,6 +2331,7 @@ entry:
 define <vscale x 4 x i16> @test_vlseg8ff_mask_nxv4i16_triscv.vector.tuple_nxv8i8_8t(target("riscv.vector.tuple", <vscale x 8 x i8>, 8) %val, ptr %base, i64 %vl, <vscale x 4 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg8ff_mask_nxv4i16_triscv.vector.tuple_nxv8i8_8t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
@@ -2339,7 +2340,6 @@ define <vscale x 4 x i16> @test_vlseg8ff_mask_nxv4i16_triscv.vector.tuple_nxv8i8
 ; CHECK-NEXT:    vmv1r.v v12, v13
 ; CHECK-NEXT:    vmv1r.v v13, v14
 ; CHECK-NEXT:    vmv1r.v v14, v15
-; CHECK-NEXT:    vsetvli zero, a1, e16, m1, ta, mu
 ; CHECK-NEXT:    vlseg8e16ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
@@ -2375,9 +2375,9 @@ entry:
 define <vscale x 1 x i32> @test_vlseg2ff_mask_nxv1i32_triscv.vector.tuple_nxv4i8_2t(target("riscv.vector.tuple", <vscale x 4 x i8>, 2) %val, ptr %base, i64 %vl, <vscale x 1 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg2ff_mask_nxv1i32_triscv.vector.tuple_nxv4i8_2t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e32, mf2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
-; CHECK-NEXT:    vsetvli zero, a1, e32, mf2, ta, mu
 ; CHECK-NEXT:    vlseg2e32ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
@@ -2413,9 +2413,9 @@ entry:
 define <vscale x 2 x i32> @test_vlseg2ff_mask_nxv2i32_triscv.vector.tuple_nxv8i8_2t(target("riscv.vector.tuple", <vscale x 8 x i8>, 2) %val, ptr %base, i64 %vl, <vscale x 2 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg2ff_mask_nxv2i32_triscv.vector.tuple_nxv8i8_2t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e32, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
-; CHECK-NEXT:    vsetvli zero, a1, e32, m1, ta, mu
 ; CHECK-NEXT:    vlseg2e32ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
@@ -2451,9 +2451,9 @@ entry:
 define <vscale x 4 x i32> @test_vlseg2ff_mask_nxv4i32_triscv.vector.tuple_nxv16i8_2t(target("riscv.vector.tuple", <vscale x 16 x i8>, 2) %val, ptr %base, i64 %vl, <vscale x 4 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg2ff_mask_nxv4i32_triscv.vector.tuple_nxv16i8_2t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e32, m2, ta, mu
 ; CHECK-NEXT:    vmv2r.v v6, v8
 ; CHECK-NEXT:    vmv2r.v v8, v10
-; CHECK-NEXT:    vsetvli zero, a1, e32, m2, ta, mu
 ; CHECK-NEXT:    vlseg2e32ff.v v6, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
@@ -2489,9 +2489,9 @@ entry:
 define <vscale x 8 x i32> @test_vlseg2ff_mask_nxv8i32_triscv.vector.tuple_nxv32i8_2t(target("riscv.vector.tuple", <vscale x 32 x i8>, 2) %val, ptr %base, i64 %vl, <vscale x 8 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg2ff_mask_nxv8i32_triscv.vector.tuple_nxv32i8_2t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e32, m4, ta, mu
 ; CHECK-NEXT:    vmv4r.v v4, v8
 ; CHECK-NEXT:    vmv4r.v v8, v12
-; CHECK-NEXT:    vsetvli zero, a1, e32, m4, ta, mu
 ; CHECK-NEXT:    vlseg2e32ff.v v4, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
@@ -2527,10 +2527,10 @@ entry:
 define <vscale x 1 x i32> @test_vlseg3ff_mask_nxv1i32_triscv.vector.tuple_nxv4i8_3t(target("riscv.vector.tuple", <vscale x 4 x i8>, 3) %val, ptr %base, i64 %vl, <vscale x 1 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg3ff_mask_nxv1i32_triscv.vector.tuple_nxv4i8_3t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e32, mf2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
-; CHECK-NEXT:    vsetvli zero, a1, e32, mf2, ta, mu
 ; CHECK-NEXT:    vlseg3e32ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
@@ -2566,10 +2566,10 @@ entry:
 define <vscale x 2 x i32> @test_vlseg3ff_mask_nxv2i32_triscv.vector.tuple_nxv8i8_3t(target("riscv.vector.tuple", <vscale x 8 x i8>, 3) %val, ptr %base, i64 %vl, <vscale x 2 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg3ff_mask_nxv2i32_triscv.vector.tuple_nxv8i8_3t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e32, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
-; CHECK-NEXT:    vsetvli zero, a1, e32, m1, ta, mu
 ; CHECK-NEXT:    vlseg3e32ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
@@ -2605,10 +2605,10 @@ entry:
 define <vscale x 4 x i32> @test_vlseg3ff_mask_nxv4i32_triscv.vector.tuple_nxv16i8_3t(target("riscv.vector.tuple", <vscale x 16 x i8>, 3) %val, ptr %base, i64 %vl, <vscale x 4 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg3ff_mask_nxv4i32_triscv.vector.tuple_nxv16i8_3t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e32, m2, ta, mu
 ; CHECK-NEXT:    vmv2r.v v6, v8
 ; CHECK-NEXT:    vmv2r.v v8, v10
 ; CHECK-NEXT:    vmv2r.v v10, v12
-; CHECK-NEXT:    vsetvli zero, a1, e32, m2, ta, mu
 ; CHECK-NEXT:    vlseg3e32ff.v v6, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
@@ -2644,11 +2644,11 @@ entry:
 define <vscale x 1 x i32> @test_vlseg4ff_mask_nxv1i32_triscv.vector.tuple_nxv4i8_4t(target("riscv.vector.tuple", <vscale x 4 x i8>, 4) %val, ptr %base, i64 %vl, <vscale x 1 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg4ff_mask_nxv1i32_triscv.vector.tuple_nxv4i8_4t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e32, mf2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
 ; CHECK-NEXT:    vmv1r.v v10, v11
-; CHECK-NEXT:    vsetvli zero, a1, e32, mf2, ta, mu
 ; CHECK-NEXT:    vlseg4e32ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
@@ -2684,11 +2684,11 @@ entry:
 define <vscale x 2 x i32> @test_vlseg4ff_mask_nxv2i32_triscv.vector.tuple_nxv8i8_4t(target("riscv.vector.tuple", <vscale x 8 x i8>, 4) %val, ptr %base, i64 %vl, <vscale x 2 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg4ff_mask_nxv2i32_triscv.vector.tuple_nxv8i8_4t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e32, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
 ; CHECK-NEXT:    vmv1r.v v10, v11
-; CHECK-NEXT:    vsetvli zero, a1, e32, m1, ta, mu
 ; CHECK-NEXT:    vlseg4e32ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
@@ -2724,11 +2724,11 @@ entry:
 define <vscale x 4 x i32> @test_vlseg4ff_mask_nxv4i32_triscv.vector.tuple_nxv16i8_4t(target("riscv.vector.tuple", <vscale x 16 x i8>, 4) %val, ptr %base, i64 %vl, <vscale x 4 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg4ff_mask_nxv4i32_triscv.vector.tuple_nxv16i8_4t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e32, m2, ta, mu
 ; CHECK-NEXT:    vmv2r.v v6, v8
 ; CHECK-NEXT:    vmv2r.v v8, v10
 ; CHECK-NEXT:    vmv2r.v v10, v12
 ; CHECK-NEXT:    vmv2r.v v12, v14
-; CHECK-NEXT:    vsetvli zero, a1, e32, m2, ta, mu
 ; CHECK-NEXT:    vlseg4e32ff.v v6, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
@@ -2764,12 +2764,12 @@ entry:
 define <vscale x 1 x i32> @test_vlseg5ff_mask_nxv1i32_triscv.vector.tuple_nxv4i8_5t(target("riscv.vector.tuple", <vscale x 4 x i8>, 5) %val, ptr %base, i64 %vl, <vscale x 1 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg5ff_mask_nxv1i32_triscv.vector.tuple_nxv4i8_5t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e32, mf2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
 ; CHECK-NEXT:    vmv1r.v v10, v11
 ; CHECK-NEXT:    vmv1r.v v11, v12
-; CHECK-NEXT:    vsetvli zero, a1, e32, mf2, ta, mu
 ; CHECK-NEXT:    vlseg5e32ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
@@ -2805,12 +2805,12 @@ entry:
 define <vscale x 2 x i32> @test_vlseg5ff_mask_nxv2i32_triscv.vector.tuple_nxv8i8_5t(target("riscv.vector.tuple", <vscale x 8 x i8>, 5) %val, ptr %base, i64 %vl, <vscale x 2 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg5ff_mask_nxv2i32_triscv.vector.tuple_nxv8i8_5t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e32, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
 ; CHECK-NEXT:    vmv1r.v v10, v11
 ; CHECK-NEXT:    vmv1r.v v11, v12
-; CHECK-NEXT:    vsetvli zero, a1, e32, m1, ta, mu
 ; CHECK-NEXT:    vlseg5e32ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
@@ -2846,13 +2846,13 @@ entry:
 define <vscale x 1 x i32> @test_vlseg6ff_mask_nxv1i32_triscv.vector.tuple_nxv4i8_6t(target("riscv.vector.tuple", <vscale x 4 x i8>, 6) %val, ptr %base, i64 %vl, <vscale x 1 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg6ff_mask_nxv1i32_triscv.vector.tuple_nxv4i8_6t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e32, mf2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
 ; CHECK-NEXT:    vmv1r.v v10, v11
 ; CHECK-NEXT:    vmv1r.v v11, v12
 ; CHECK-NEXT:    vmv1r.v v12, v13
-; CHECK-NEXT:    vsetvli zero, a1, e32, mf2, ta, mu
 ; CHECK-NEXT:    vlseg6e32ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
@@ -2888,13 +2888,13 @@ entry:
 define <vscale x 2 x i32> @test_vlseg6ff_mask_nxv2i32_triscv.vector.tuple_nxv8i8_6t(target("riscv.vector.tuple", <vscale x 8 x i8>, 6) %val, ptr %base, i64 %vl, <vscale x 2 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg6ff_mask_nxv2i32_triscv.vector.tuple_nxv8i8_6t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e32, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
 ; CHECK-NEXT:    vmv1r.v v10, v11
 ; CHECK-NEXT:    vmv1r.v v11, v12
 ; CHECK-NEXT:    vmv1r.v v12, v13
-; CHECK-NEXT:    vsetvli zero, a1, e32, m1, ta, mu
 ; CHECK-NEXT:    vlseg6e32ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
@@ -2930,6 +2930,7 @@ entry:
 define <vscale x 1 x i32> @test_vlseg7ff_mask_nxv1i32_triscv.vector.tuple_nxv4i8_7t(target("riscv.vector.tuple", <vscale x 4 x i8>, 7) %val, ptr %base, i64 %vl, <vscale x 1 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg7ff_mask_nxv1i32_triscv.vector.tuple_nxv4i8_7t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e32, mf2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
@@ -2937,7 +2938,6 @@ define <vscale x 1 x i32> @test_vlseg7ff_mask_nxv1i32_triscv.vector.tuple_nxv4i8
 ; CHECK-NEXT:    vmv1r.v v11, v12
 ; CHECK-NEXT:    vmv1r.v v12, v13
 ; CHECK-NEXT:    vmv1r.v v13, v14
-; CHECK-NEXT:    vsetvli zero, a1, e32, mf2, ta, mu
 ; CHECK-NEXT:    vlseg7e32ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
@@ -2973,6 +2973,7 @@ entry:
 define <vscale x 2 x i32> @test_vlseg7ff_mask_nxv2i32_triscv.vector.tuple_nxv8i8_7t(target("riscv.vector.tuple", <vscale x 8 x i8>, 7) %val, ptr %base, i64 %vl, <vscale x 2 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg7ff_mask_nxv2i32_triscv.vector.tuple_nxv8i8_7t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e32, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
@@ -2980,7 +2981,6 @@ define <vscale x 2 x i32> @test_vlseg7ff_mask_nxv2i32_triscv.vector.tuple_nxv8i8
 ; CHECK-NEXT:    vmv1r.v v11, v12
 ; CHECK-NEXT:    vmv1r.v v12, v13
 ; CHECK-NEXT:    vmv1r.v v13, v14
-; CHECK-NEXT:    vsetvli zero, a1, e32, m1, ta, mu
 ; CHECK-NEXT:    vlseg7e32ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
@@ -3016,6 +3016,7 @@ entry:
 define <vscale x 1 x i32> @test_vlseg8ff_mask_nxv1i32_triscv.vector.tuple_nxv4i8_8t(target("riscv.vector.tuple", <vscale x 4 x i8>, 8) %val, ptr %base, i64 %vl, <vscale x 1 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg8ff_mask_nxv1i32_triscv.vector.tuple_nxv4i8_8t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e32, mf2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
@@ -3024,7 +3025,6 @@ define <vscale x 1 x i32> @test_vlseg8ff_mask_nxv1i32_triscv.vector.tuple_nxv4i8
 ; CHECK-NEXT:    vmv1r.v v12, v13
 ; CHECK-NEXT:    vmv1r.v v13, v14
 ; CHECK-NEXT:    vmv1r.v v14, v15
-; CHECK-NEXT:    vsetvli zero, a1, e32, mf2, ta, mu
 ; CHECK-NEXT:    vlseg8e32ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
@@ -3060,6 +3060,7 @@ entry:
 define <vscale x 2 x i32> @test_vlseg8ff_mask_nxv2i32_triscv.vector.tuple_nxv8i8_8t(target("riscv.vector.tuple", <vscale x 8 x i8>, 8) %val, ptr %base, i64 %vl, <vscale x 2 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg8ff_mask_nxv2i32_triscv.vector.tuple_nxv8i8_8t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e32, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
@@ -3068,7 +3069,6 @@ define <vscale x 2 x i32> @test_vlseg8ff_mask_nxv2i32_triscv.vector.tuple_nxv8i8
 ; CHECK-NEXT:    vmv1r.v v12, v13
 ; CHECK-NEXT:    vmv1r.v v13, v14
 ; CHECK-NEXT:    vmv1r.v v14, v15
-; CHECK-NEXT:    vsetvli zero, a1, e32, m1, ta, mu
 ; CHECK-NEXT:    vlseg8e32ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
@@ -3104,9 +3104,9 @@ entry:
 define <vscale x 1 x i64> @test_vlseg2ff_mask_nxv1i64_triscv.vector.tuple_nxv8i8_2t(target("riscv.vector.tuple", <vscale x 8 x i8>, 2) %val, ptr %base, i64 %vl, <vscale x 1 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg2ff_mask_nxv1i64_triscv.vector.tuple_nxv8i8_2t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e64, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
-; CHECK-NEXT:    vsetvli zero, a1, e64, m1, ta, mu
 ; CHECK-NEXT:    vlseg2e64ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
@@ -3142,9 +3142,9 @@ entry:
 define <vscale x 2 x i64> @test_vlseg2ff_mask_nxv2i64_triscv.vector.tuple_nxv16i8_2t(target("riscv.vector.tuple", <vscale x 16 x i8>, 2) %val, ptr %base, i64 %vl, <vscale x 2 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg2ff_mask_nxv2i64_triscv.vector.tuple_nxv16i8_2t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e64, m2, ta, mu
 ; CHECK-NEXT:    vmv2r.v v6, v8
 ; CHECK-NEXT:    vmv2r.v v8, v10
-; CHECK-NEXT:    vsetvli zero, a1, e64, m2, ta, mu
 ; CHECK-NEXT:    vlseg2e64ff.v v6, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
@@ -3180,9 +3180,9 @@ entry:
 define <vscale x 4 x i64> @test_vlseg2ff_mask_nxv4i64_triscv.vector.tuple_nxv32i8_2t(target("riscv.vector.tuple", <vscale x 32 x i8>, 2) %val, ptr %base, i64 %vl, <vscale x 4 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg2ff_mask_nxv4i64_triscv.vector.tuple_nxv32i8_2t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e64, m4, ta, mu
 ; CHECK-NEXT:    vmv4r.v v4, v8
 ; CHECK-NEXT:    vmv4r.v v8, v12
-; CHECK-NEXT:    vsetvli zero, a1, e64, m4, ta, mu
 ; CHECK-NEXT:    vlseg2e64ff.v v4, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
@@ -3218,10 +3218,10 @@ entry:
 define <vscale x 1 x i64> @test_vlseg3ff_mask_nxv1i64_triscv.vector.tuple_nxv8i8_3t(target("riscv.vector.tuple", <vscale x 8 x i8>, 3) %val, ptr %base, i64 %vl, <vscale x 1 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg3ff_mask_nxv1i64_triscv.vector.tuple_nxv8i8_3t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e64, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
-; CHECK-NEXT:    vsetvli zero, a1, e64, m1, ta, mu
 ; CHECK-NEXT:    vlseg3e64ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
@@ -3257,10 +3257,10 @@ entry:
 define <vscale x 2 x i64> @test_vlseg3ff_mask_nxv2i64_triscv.vector.tuple_nxv16i8_3t(target("riscv.vector.tuple", <vscale x 16 x i8>, 3) %val, ptr %base, i64 %vl, <vscale x 2 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg3ff_mask_nxv2i64_triscv.vector.tuple_nxv16i8_3t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e64, m2, ta, mu
 ; CHECK-NEXT:    vmv2r.v v6, v8
 ; CHECK-NEXT:    vmv2r.v v8, v10
 ; CHECK-NEXT:    vmv2r.v v10, v12
-; CHECK-NEXT:    vsetvli zero, a1, e64, m2, ta, mu
 ; CHECK-NEXT:    vlseg3e64ff.v v6, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
@@ -3296,11 +3296,11 @@ entry:
 define <vscale x 1 x i64> @test_vlseg4ff_mask_nxv1i64_triscv.vector.tuple_nxv8i8_4t(target("riscv.vector.tuple", <vscale x 8 x i8>, 4) %val, ptr %base, i64 %vl, <vscale x 1 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg4ff_mask_nxv1i64_triscv.vector.tuple_nxv8i8_4t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e64, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
 ; CHECK-NEXT:    vmv1r.v v10, v11
-; CHECK-NEXT:    vsetvli zero, a1, e64, m1, ta, mu
 ; CHECK-NEXT:    vlseg4e64ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
@@ -3336,11 +3336,11 @@ entry:
 define <vscale x 2 x i64> @test_vlseg4ff_mask_nxv2i64_triscv.vector.tuple_nxv16i8_4t(target("riscv.vector.tuple", <vscale x 16 x i8>, 4) %val, ptr %base, i64 %vl, <vscale x 2 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg4ff_mask_nxv2i64_triscv.vector.tuple_nxv16i8_4t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e64, m2, ta, mu
 ; CHECK-NEXT:    vmv2r.v v6, v8
 ; CHECK-NEXT:    vmv2r.v v8, v10
 ; CHECK-NEXT:    vmv2r.v v10, v12
 ; CHECK-NEXT:    vmv2r.v v12, v14
-; CHECK-NEXT:    vsetvli zero, a1, e64, m2, ta, mu
 ; CHECK-NEXT:    vlseg4e64ff.v v6, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
@@ -3376,12 +3376,12 @@ entry:
 define <vscale x 1 x i64> @test_vlseg5ff_mask_nxv1i64_triscv.vector.tuple_nxv8i8_5t(target("riscv.vector.tuple", <vscale x 8 x i8>, 5) %val, ptr %base, i64 %vl, <vscale x 1 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg5ff_mask_nxv1i64_triscv.vector.tuple_nxv8i8_5t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e64, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
 ; CHECK-NEXT:    vmv1r.v v10, v11
 ; CHECK-NEXT:    vmv1r.v v11, v12
-; CHECK-NEXT:    vsetvli zero, a1, e64, m1, ta, mu
 ; CHECK-NEXT:    vlseg5e64ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
@@ -3417,13 +3417,13 @@ entry:
 define <vscale x 1 x i64> @test_vlseg6ff_mask_nxv1i64_triscv.vector.tuple_nxv8i8_6t(target("riscv.vector.tuple", <vscale x 8 x i8>, 6) %val, ptr %base, i64 %vl, <vscale x 1 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg6ff_mask_nxv1i64_triscv.vector.tuple_nxv8i8_6t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e64, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
 ; CHECK-NEXT:    vmv1r.v v10, v11
 ; CHECK-NEXT:    vmv1r.v v11, v12
 ; CHECK-NEXT:    vmv1r.v v12, v13
-; CHECK-NEXT:    vsetvli zero, a1, e64, m1, ta, mu
 ; CHECK-NEXT:    vlseg6e64ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
@@ -3459,6 +3459,7 @@ entry:
 define <vscale x 1 x i64> @test_vlseg7ff_mask_nxv1i64_triscv.vector.tuple_nxv8i8_7t(target("riscv.vector.tuple", <vscale x 8 x i8>, 7) %val, ptr %base, i64 %vl, <vscale x 1 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg7ff_mask_nxv1i64_triscv.vector.tuple_nxv8i8_7t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e64, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
@@ -3466,7 +3467,6 @@ define <vscale x 1 x i64> @test_vlseg7ff_mask_nxv1i64_triscv.vector.tuple_nxv8i8
 ; CHECK-NEXT:    vmv1r.v v11, v12
 ; CHECK-NEXT:    vmv1r.v v12, v13
 ; CHECK-NEXT:    vmv1r.v v13, v14
-; CHECK-NEXT:    vsetvli zero, a1, e64, m1, ta, mu
 ; CHECK-NEXT:    vlseg7e64ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
@@ -3502,6 +3502,7 @@ entry:
 define <vscale x 1 x i64> @test_vlseg8ff_mask_nxv1i64_triscv.vector.tuple_nxv8i8_8t(target("riscv.vector.tuple", <vscale x 8 x i8>, 8) %val, ptr %base, i64 %vl, <vscale x 1 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg8ff_mask_nxv1i64_triscv.vector.tuple_nxv8i8_8t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e64, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
@@ -3510,7 +3511,6 @@ define <vscale x 1 x i64> @test_vlseg8ff_mask_nxv1i64_triscv.vector.tuple_nxv8i8
 ; CHECK-NEXT:    vmv1r.v v12, v13
 ; CHECK-NEXT:    vmv1r.v v13, v14
 ; CHECK-NEXT:    vmv1r.v v14, v15
-; CHECK-NEXT:    vsetvli zero, a1, e64, m1, ta, mu
 ; CHECK-NEXT:    vlseg8e64ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
@@ -3545,9 +3545,9 @@ entry:
 define <vscale x 1 x half> @test_vlseg2ff_mask_nxv1f16_triscv.vector.tuple_nxv2i8_2t(target("riscv.vector.tuple", <vscale x 2 x i8>, 2) %val, ptr %base, i64 %vl, <vscale x 1 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg2ff_mask_nxv1f16_triscv.vector.tuple_nxv2i8_2t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, mf4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
-; CHECK-NEXT:    vsetvli zero, a1, e16, mf4, ta, mu
 ; CHECK-NEXT:    vlseg2e16ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
@@ -3582,9 +3582,9 @@ entry:
 define <vscale x 2 x half> @test_vlseg2ff_mask_nxv2f16_triscv.vector.tuple_nxv4i8_2t(target("riscv.vector.tuple", <vscale x 4 x i8>, 2) %val, ptr %base, i64 %vl, <vscale x 2 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg2ff_mask_nxv2f16_triscv.vector.tuple_nxv4i8_2t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, mf2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
-; CHECK-NEXT:    vsetvli zero, a1, e16, mf2, ta, mu
 ; CHECK-NEXT:    vlseg2e16ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
@@ -3619,9 +3619,9 @@ entry:
 define <vscale x 4 x half> @test_vlseg2ff_mask_nxv4f16_triscv.vector.tuple_nxv8i8_2t(target("riscv.vector.tuple", <vscale x 8 x i8>, 2) %val, ptr %base, i64 %vl, <vscale x 4 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg2ff_mask_nxv4f16_triscv.vector.tuple_nxv8i8_2t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
-; CHECK-NEXT:    vsetvli zero, a1, e16, m1, ta, mu
 ; CHECK-NEXT:    vlseg2e16ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
@@ -3656,9 +3656,9 @@ entry:
 define <vscale x 8 x half> @test_vlseg2ff_mask_nxv8f16_triscv.vector.tuple_nxv16i8_2t(target("riscv.vector.tuple", <vscale x 16 x i8>, 2) %val, ptr %base, i64 %vl, <vscale x 8 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg2ff_mask_nxv8f16_triscv.vector.tuple_nxv16i8_2t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, m2, ta, mu
 ; CHECK-NEXT:    vmv2r.v v6, v8
 ; CHECK-NEXT:    vmv2r.v v8, v10
-; CHECK-NEXT:    vsetvli zero, a1, e16, m2, ta, mu
 ; CHECK-NEXT:    vlseg2e16ff.v v6, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
@@ -3693,9 +3693,9 @@ entry:
 define <vscale x 16 x half> @test_vlseg2ff_mask_nxv16f16_triscv.vector.tuple_nxv32i8_2t(target("riscv.vector.tuple", <vscale x 32 x i8>, 2) %val, ptr %base, i64 %vl, <vscale x 16 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg2ff_mask_nxv16f16_triscv.vector.tuple_nxv32i8_2t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, m4, ta, mu
 ; CHECK-NEXT:    vmv4r.v v4, v8
 ; CHECK-NEXT:    vmv4r.v v8, v12
-; CHECK-NEXT:    vsetvli zero, a1, e16, m4, ta, mu
 ; CHECK-NEXT:    vlseg2e16ff.v v4, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
@@ -3730,10 +3730,10 @@ entry:
 define <vscale x 1 x half> @test_vlseg3ff_mask_nxv1f16_triscv.vector.tuple_nxv2i8_3t(target("riscv.vector.tuple", <vscale x 2 x i8>, 3) %val, ptr %base, i64 %vl, <vscale x 1 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg3ff_mask_nxv1f16_triscv.vector.tuple_nxv2i8_3t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, mf4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
-; CHECK-NEXT:    vsetvli zero, a1, e16, mf4, ta, mu
 ; CHECK-NEXT:    vlseg3e16ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
@@ -3768,10 +3768,10 @@ entry:
 define <vscale x 2 x half> @test_vlseg3ff_mask_nxv2f16_triscv.vector.tuple_nxv4i8_3t(target("riscv.vector.tuple", <vscale x 4 x i8>, 3) %val, ptr %base, i64 %vl, <vscale x 2 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg3ff_mask_nxv2f16_triscv.vector.tuple_nxv4i8_3t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, mf2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
-; CHECK-NEXT:    vsetvli zero, a1, e16, mf2, ta, mu
 ; CHECK-NEXT:    vlseg3e16ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
@@ -3806,10 +3806,10 @@ entry:
 define <vscale x 4 x half> @test_vlseg3ff_mask_nxv4f16_triscv.vector.tuple_nxv8i8_3t(target("riscv.vector.tuple", <vscale x 8 x i8>, 3) %val, ptr %base, i64 %vl, <vscale x 4 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg3ff_mask_nxv4f16_triscv.vector.tuple_nxv8i8_3t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
-; CHECK-NEXT:    vsetvli zero, a1, e16, m1, ta, mu
 ; CHECK-NEXT:    vlseg3e16ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
@@ -3844,10 +3844,10 @@ entry:
 define <vscale x 8 x half> @test_vlseg3ff_mask_nxv8f16_triscv.vector.tuple_nxv16i8_3t(target("riscv.vector.tuple", <vscale x 16 x i8>, 3) %val, ptr %base, i64 %vl, <vscale x 8 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg3ff_mask_nxv8f16_triscv.vector.tuple_nxv16i8_3t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, m2, ta, mu
 ; CHECK-NEXT:    vmv2r.v v6, v8
 ; CHECK-NEXT:    vmv2r.v v8, v10
 ; CHECK-NEXT:    vmv2r.v v10, v12
-; CHECK-NEXT:    vsetvli zero, a1, e16, m2, ta, mu
 ; CHECK-NEXT:    vlseg3e16ff.v v6, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
@@ -3882,11 +3882,11 @@ entry:
 define <vscale x 1 x half> @test_vlseg4ff_mask_nxv1f16_triscv.vector.tuple_nxv2i8_4t(target("riscv.vector.tuple", <vscale x 2 x i8>, 4) %val, ptr %base, i64 %vl, <vscale x 1 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg4ff_mask_nxv1f16_triscv.vector.tuple_nxv2i8_4t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, mf4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
 ; CHECK-NEXT:    vmv1r.v v10, v11
-; CHECK-NEXT:    vsetvli zero, a1, e16, mf4, ta, mu
 ; CHECK-NEXT:    vlseg4e16ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
@@ -3921,11 +3921,11 @@ entry:
 define <vscale x 2 x half> @test_vlseg4ff_mask_nxv2f16_triscv.vector.tuple_nxv4i8_4t(target("riscv.vector.tuple", <vscale x 4 x i8>, 4) %val, ptr %base, i64 %vl, <vscale x 2 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg4ff_mask_nxv2f16_triscv.vector.tuple_nxv4i8_4t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, mf2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
 ; CHECK-NEXT:    vmv1r.v v10, v11
-; CHECK-NEXT:    vsetvli zero, a1, e16, mf2, ta, mu
 ; CHECK-NEXT:    vlseg4e16ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
@@ -3960,11 +3960,11 @@ entry:
 define <vscale x 4 x half> @test_vlseg4ff_mask_nxv4f16_triscv.vector.tuple_nxv8i8_4t(target("riscv.vector.tuple", <vscale x 8 x i8>, 4) %val, ptr %base, i64 %vl, <vscale x 4 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg4ff_mask_nxv4f16_triscv.vector.tuple_nxv8i8_4t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
 ; CHECK-NEXT:    vmv1r.v v10, v11
-; CHECK-NEXT:    vsetvli zero, a1, e16, m1, ta, mu
 ; CHECK-NEXT:    vlseg4e16ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
@@ -3999,11 +3999,11 @@ entry:
 define <vscale x 8 x half> @test_vlseg4ff_mask_nxv8f16_triscv.vector.tuple_nxv16i8_4t(target("riscv.vector.tuple", <vscale x 16 x i8>, 4) %val, ptr %base, i64 %vl, <vscale x 8 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg4ff_mask_nxv8f16_triscv.vector.tuple_nxv16i8_4t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, m2, ta, mu
 ; CHECK-NEXT:    vmv2r.v v6, v8
 ; CHECK-NEXT:    vmv2r.v v8, v10
 ; CHECK-NEXT:    vmv2r.v v10, v12
 ; CHECK-NEXT:    vmv2r.v v12, v14
-; CHECK-NEXT:    vsetvli zero, a1, e16, m2, ta, mu
 ; CHECK-NEXT:    vlseg4e16ff.v v6, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
@@ -4038,12 +4038,12 @@ entry:
 define <vscale x 1 x half> @test_vlseg5ff_mask_nxv1f16_triscv.vector.tuple_nxv2i8_5t(target("riscv.vector.tuple", <vscale x 2 x i8>, 5) %val, ptr %base, i64 %vl, <vscale x 1 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg5ff_mask_nxv1f16_triscv.vector.tuple_nxv2i8_5t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, mf4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
 ; CHECK-NEXT:    vmv1r.v v10, v11
 ; CHECK-NEXT:    vmv1r.v v11, v12
-; CHECK-NEXT:    vsetvli zero, a1, e16, mf4, ta, mu
 ; CHECK-NEXT:    vlseg5e16ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
@@ -4078,12 +4078,12 @@ entry:
 define <vscale x 2 x half> @test_vlseg5ff_mask_nxv2f16_triscv.vector.tuple_nxv4i8_5t(target("riscv.vector.tuple", <vscale x 4 x i8>, 5) %val, ptr %base, i64 %vl, <vscale x 2 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg5ff_mask_nxv2f16_triscv.vector.tuple_nxv4i8_5t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, mf2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
 ; CHECK-NEXT:    vmv1r.v v10, v11
 ; CHECK-NEXT:    vmv1r.v v11, v12
-; CHECK-NEXT:    vsetvli zero, a1, e16, mf2, ta, mu
 ; CHECK-NEXT:    vlseg5e16ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
@@ -4118,12 +4118,12 @@ entry:
 define <vscale x 4 x half> @test_vlseg5ff_mask_nxv4f16_triscv.vector.tuple_nxv8i8_5t(target("riscv.vector.tuple", <vscale x 8 x i8>, 5) %val, ptr %base, i64 %vl, <vscale x 4 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg5ff_mask_nxv4f16_triscv.vector.tuple_nxv8i8_5t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
 ; CHECK-NEXT:    vmv1r.v v10, v11
 ; CHECK-NEXT:    vmv1r.v v11, v12
-; CHECK-NEXT:    vsetvli zero, a1, e16, m1, ta, mu
 ; CHECK-NEXT:    vlseg5e16ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
@@ -4158,13 +4158,13 @@ entry:
 define <vscale x 1 x half> @test_vlseg6ff_mask_nxv1f16_triscv.vector.tuple_nxv2i8_6t(target("riscv.vector.tuple", <vscale x 2 x i8>, 6) %val, ptr %base, i64 %vl, <vscale x 1 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg6ff_mask_nxv1f16_triscv.vector.tuple_nxv2i8_6t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, mf4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
 ; CHECK-NEXT:    vmv1r.v v10, v11
 ; CHECK-NEXT:    vmv1r.v v11, v12
 ; CHECK-NEXT:    vmv1r.v v12, v13
-; CHECK-NEXT:    vsetvli zero, a1, e16, mf4, ta, mu
 ; CHECK-NEXT:    vlseg6e16ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
@@ -4199,13 +4199,13 @@ entry:
 define <vscale x 2 x half> @test_vlseg6ff_mask_nxv2f16_triscv.vector.tuple_nxv4i8_6t(target("riscv.vector.tuple", <vscale x 4 x i8>, 6) %val, ptr %base, i64 %vl, <vscale x 2 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg6ff_mask_nxv2f16_triscv.vector.tuple_nxv4i8_6t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, mf2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
 ; CHECK-NEXT:    vmv1r.v v10, v11
 ; CHECK-NEXT:    vmv1r.v v11, v12
 ; CHECK-NEXT:    vmv1r.v v12, v13
-; CHECK-NEXT:    vsetvli zero, a1, e16, mf2, ta, mu
 ; CHECK-NEXT:    vlseg6e16ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
@@ -4240,13 +4240,13 @@ entry:
 define <vscale x 4 x half> @test_vlseg6ff_mask_nxv4f16_triscv.vector.tuple_nxv8i8_6t(target("riscv.vector.tuple", <vscale x 8 x i8>, 6) %val, ptr %base, i64 %vl, <vscale x 4 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg6ff_mask_nxv4f16_triscv.vector.tuple_nxv8i8_6t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
 ; CHECK-NEXT:    vmv1r.v v10, v11
 ; CHECK-NEXT:    vmv1r.v v11, v12
 ; CHECK-NEXT:    vmv1r.v v12, v13
-; CHECK-NEXT:    vsetvli zero, a1, e16, m1, ta, mu
 ; CHECK-NEXT:    vlseg6e16ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
@@ -4281,6 +4281,7 @@ entry:
 define <vscale x 1 x half> @test_vlseg7ff_mask_nxv1f16_triscv.vector.tuple_nxv2i8_7t(target("riscv.vector.tuple", <vscale x 2 x i8>, 7) %val, ptr %base, i64 %vl, <vscale x 1 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg7ff_mask_nxv1f16_triscv.vector.tuple_nxv2i8_7t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, mf4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
@@ -4288,7 +4289,6 @@ define <vscale x 1 x half> @test_vlseg7ff_mask_nxv1f16_triscv.vector.tuple_nxv2i
 ; CHECK-NEXT:    vmv1r.v v11, v12
 ; CHECK-NEXT:    vmv1r.v v12, v13
 ; CHECK-NEXT:    vmv1r.v v13, v14
-; CHECK-NEXT:    vsetvli zero, a1, e16, mf4, ta, mu
 ; CHECK-NEXT:    vlseg7e16ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
@@ -4323,6 +4323,7 @@ entry:
 define <vscale x 2 x half> @test_vlseg7ff_mask_nxv2f16_triscv.vector.tuple_nxv4i8_7t(target("riscv.vector.tuple", <vscale x 4 x i8>, 7) %val, ptr %base, i64 %vl, <vscale x 2 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg7ff_mask_nxv2f16_triscv.vector.tuple_nxv4i8_7t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, mf2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
@@ -4330,7 +4331,6 @@ define <vscale x 2 x half> @test_vlseg7ff_mask_nxv2f16_triscv.vector.tuple_nxv4i
 ; CHECK-NEXT:    vmv1r.v v11, v12
 ; CHECK-NEXT:    vmv1r.v v12, v13
 ; CHECK-NEXT:    vmv1r.v v13, v14
-; CHECK-NEXT:    vsetvli zero, a1, e16, mf2, ta, mu
 ; CHECK-NEXT:    vlseg7e16ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
@@ -4365,6 +4365,7 @@ entry:
 define <vscale x 4 x half> @test_vlseg7ff_mask_nxv4f16_triscv.vector.tuple_nxv8i8_7t(target("riscv.vector.tuple", <vscale x 8 x i8>, 7) %val, ptr %base, i64 %vl, <vscale x 4 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg7ff_mask_nxv4f16_triscv.vector.tuple_nxv8i8_7t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
@@ -4372,7 +4373,6 @@ define <vscale x 4 x half> @test_vlseg7ff_mask_nxv4f16_triscv.vector.tuple_nxv8i
 ; CHECK-NEXT:    vmv1r.v v11, v12
 ; CHECK-NEXT:    vmv1r.v v12, v13
 ; CHECK-NEXT:    vmv1r.v v13, v14
-; CHECK-NEXT:    vsetvli zero, a1, e16, m1, ta, mu
 ; CHECK-NEXT:    vlseg7e16ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
@@ -4407,6 +4407,7 @@ entry:
 define <vscale x 1 x half> @test_vlseg8ff_mask_nxv1f16_triscv.vector.tuple_nxv2i8_8t(target("riscv.vector.tuple", <vscale x 2 x i8>, 8) %val, ptr %base, i64 %vl, <vscale x 1 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg8ff_mask_nxv1f16_triscv.vector.tuple_nxv2i8_8t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, mf4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
@@ -4415,7 +4416,6 @@ define <vscale x 1 x half> @test_vlseg8ff_mask_nxv1f16_triscv.vector.tuple_nxv2i
 ; CHECK-NEXT:    vmv1r.v v12, v13
 ; CHECK-NEXT:    vmv1r.v v13, v14
 ; CHECK-NEXT:    vmv1r.v v14, v15
-; CHECK-NEXT:    vsetvli zero, a1, e16, mf4, ta, mu
 ; CHECK-NEXT:    vlseg8e16ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
@@ -4450,6 +4450,7 @@ entry:
 define <vscale x 2 x half> @test_vlseg8ff_mask_nxv2f16_triscv.vector.tuple_nxv4i8_8t(target("riscv.vector.tuple", <vscale x 4 x i8>, 8) %val, ptr %base, i64 %vl, <vscale x 2 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg8ff_mask_nxv2f16_triscv.vector.tuple_nxv4i8_8t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, mf2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
@@ -4458,7 +4459,6 @@ define <vscale x 2 x half> @test_vlseg8ff_mask_nxv2f16_triscv.vector.tuple_nxv4i
 ; CHECK-NEXT:    vmv1r.v v12, v13
 ; CHECK-NEXT:    vmv1r.v v13, v14
 ; CHECK-NEXT:    vmv1r.v v14, v15
-; CHECK-NEXT:    vsetvli zero, a1, e16, mf2, ta, mu
 ; CHECK-NEXT:    vlseg8e16ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
@@ -4493,6 +4493,7 @@ entry:
 define <vscale x 4 x half> @test_vlseg8ff_mask_nxv4f16_triscv.vector.tuple_nxv8i8_8t(target("riscv.vector.tuple", <vscale x 8 x i8>, 8) %val, ptr %base, i64 %vl, <vscale x 4 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg8ff_mask_nxv4f16_triscv.vector.tuple_nxv8i8_8t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
@@ -4501,7 +4502,6 @@ define <vscale x 4 x half> @test_vlseg8ff_mask_nxv4f16_triscv.vector.tuple_nxv8i
 ; CHECK-NEXT:    vmv1r.v v12, v13
 ; CHECK-NEXT:    vmv1r.v v13, v14
 ; CHECK-NEXT:    vmv1r.v v14, v15
-; CHECK-NEXT:    vsetvli zero, a1, e16, m1, ta, mu
 ; CHECK-NEXT:    vlseg8e16ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
@@ -4536,9 +4536,9 @@ entry:
 define <vscale x 1 x float> @test_vlseg2ff_mask_nxv1f32_triscv.vector.tuple_nxv4i8_2t(target("riscv.vector.tuple", <vscale x 4 x i8>, 2) %val, ptr %base, i64 %vl, <vscale x 1 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg2ff_mask_nxv1f32_triscv.vector.tuple_nxv4i8_2t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e32, mf2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
-; CHECK-NEXT:    vsetvli zero, a1, e32, mf2, ta, mu
 ; CHECK-NEXT:    vlseg2e32ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
@@ -4573,9 +4573,9 @@ entry:
 define <vscale x 2 x float> @test_vlseg2ff_mask_nxv2f32_triscv.vector.tuple_nxv8i8_2t(target("riscv.vector.tuple", <vscale x 8 x i8>, 2) %val, ptr %base, i64 %vl, <vscale x 2 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg2ff_mask_nxv2f32_triscv.vector.tuple_nxv8i8_2t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e32, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
-; CHECK-NEXT:    vsetvli zero, a1, e32, m1, ta, mu
 ; CHECK-NEXT:    vlseg2e32ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
@@ -4610,9 +4610,9 @@ entry:
 define <vscale x 4 x float> @test_vlseg2ff_mask_nxv4f32_triscv.vector.tuple_nxv16i8_2t(target("riscv.vector.tuple", <vscale x 16 x i8>, 2) %val, ptr %base, i64 %vl, <vscale x 4 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg2ff_mask_nxv4f32_triscv.vector.tuple_nxv16i8_2t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e32, m2, ta, mu
 ; CHECK-NEXT:    vmv2r.v v6, v8
 ; CHECK-NEXT:    vmv2r.v v8, v10
-; CHECK-NEXT:    vsetvli zero, a1, e32, m2, ta, mu
 ; CHECK-NEXT:    vlseg2e32ff.v v6, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
@@ -4647,9 +4647,9 @@ entry:
 define <vscale x 8 x float> @test_vlseg2ff_mask_nxv8f32_triscv.vector.tuple_nxv32i8_2t(target("riscv.vector.tuple", <vscale x 32 x i8>, 2) %val, ptr %base, i64 %vl, <vscale x 8 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg2ff_mask_nxv8f32_triscv.vector.tuple_nxv32i8_2t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e32, m4, ta, mu
 ; CHECK-NEXT:    vmv4r.v v4, v8
 ; CHECK-NEXT:    vmv4r.v v8, v12
-; CHECK-NEXT:    vsetvli zero, a1, e32, m4, ta, mu
 ; CHECK-NEXT:    vlseg2e32ff.v v4, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
@@ -4684,10 +4684,10 @@ entry:
 define <vscale x 1 x float> @test_vlseg3ff_mask_nxv1f32_triscv.vector.tuple_nxv4i8_3t(target("riscv.vector.tuple", <vscale x 4 x i8>, 3) %val, ptr %base, i64 %vl, <vscale x 1 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg3ff_mask_nxv1f32_triscv.vector.tuple_nxv4i8_3t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e32, mf2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
-; CHECK-NEXT:    vsetvli zero, a1, e32, mf2, ta, mu
 ; CHECK-NEXT:    vlseg3e32ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
@@ -4722,10 +4722,10 @@ entry:
 define <vscale x 2 x float> @test_vlseg3ff_mask_nxv2f32_triscv.vector.tuple_nxv8i8_3t(target("riscv.vector.tuple", <vscale x 8 x i8>, 3) %val, ptr %base, i64 %vl, <vscale x 2 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg3ff_mask_nxv2f32_triscv.vector.tuple_nxv8i8_3t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e32, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
-; CHECK-NEXT:    vsetvli zero, a1, e32, m1, ta, mu
 ; CHECK-NEXT:    vlseg3e32ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
@@ -4760,10 +4760,10 @@ entry:
 define <vscale x 4 x float> @test_vlseg3ff_mask_nxv4f32_triscv.vector.tuple_nxv16i8_3t(target("riscv.vector.tuple", <vscale x 16 x i8>, 3) %val, ptr %base, i64 %vl, <vscale x 4 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg3ff_mask_nxv4f32_triscv.vector.tuple_nxv16i8_3t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e32, m2, ta, mu
 ; CHECK-NEXT:    vmv2r.v v6, v8
 ; CHECK-NEXT:    vmv2r.v v8, v10
 ; CHECK-NEXT:    vmv2r.v v10, v12
-; CHECK-NEXT:    vsetvli zero, a1, e32, m2, ta, mu
 ; CHECK-NEXT:    vlseg3e32ff.v v6, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
@@ -4798,11 +4798,11 @@ entry:
 define <vscale x 1 x float> @test_vlseg4ff_mask_nxv1f32_triscv.vector.tuple_nxv4i8_4t(target("riscv.vector.tuple", <vscale x 4 x i8>, 4) %val, ptr %base, i64 %vl, <vscale x 1 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg4ff_mask_nxv1f32_triscv.vector.tuple_nxv4i8_4t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e32, mf2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
 ; CHECK-NEXT:    vmv1r.v v10, v11
-; CHECK-NEXT:    vsetvli zero, a1, e32, mf2, ta, mu
 ; CHECK-NEXT:    vlseg4e32ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
@@ -4837,11 +4837,11 @@ entry:
 define <vscale x 2 x float> @test_vlseg4ff_mask_nxv2f32_triscv.vector.tuple_nxv8i8_4t(target("riscv.vector.tuple", <vscale x 8 x i8>, 4) %val, ptr %base, i64 %vl, <vscale x 2 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg4ff_mask_nxv2f32_triscv.vector.tuple_nxv8i8_4t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e32, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
 ; CHECK-NEXT:    vmv1r.v v10, v11
-; CHECK-NEXT:    vsetvli zero, a1, e32, m1, ta, mu
 ; CHECK-NEXT:    vlseg4e32ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
@@ -4876,11 +4876,11 @@ entry:
 define <vscale x 4 x float> @test_vlseg4ff_mask_nxv4f32_triscv.vector.tuple_nxv16i8_4t(target("riscv.vector.tuple", <vscale x 16 x i8>, 4) %val, ptr %base, i64 %vl, <vscale x 4 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg4ff_mask_nxv4f32_triscv.vector.tuple_nxv16i8_4t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e32, m2, ta, mu
 ; CHECK-NEXT:    vmv2r.v v6, v8
 ; CHECK-NEXT:    vmv2r.v v8, v10
 ; CHECK-NEXT:    vmv2r.v v10, v12
 ; CHECK-NEXT:    vmv2r.v v12, v14
-; CHECK-NEXT:    vsetvli zero, a1, e32, m2, ta, mu
 ; CHECK-NEXT:    vlseg4e32ff.v v6, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
@@ -4915,12 +4915,12 @@ entry:
 define <vscale x 1 x float> @test_vlseg5ff_mask_nxv1f32_triscv.vector.tuple_nxv4i8_5t(target("riscv.vector.tuple", <vscale x 4 x i8>, 5) %val, ptr %base, i64 %vl, <vscale x 1 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg5ff_mask_nxv1f32_triscv.vector.tuple_nxv4i8_5t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e32, mf2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
 ; CHECK-NEXT:    vmv1r.v v10, v11
 ; CHECK-NEXT:    vmv1r.v v11, v12
-; CHECK-NEXT:    vsetvli zero, a1, e32, mf2, ta, mu
 ; CHECK-NEXT:    vlseg5e32ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
@@ -4955,12 +4955,12 @@ entry:
 define <vscale x 2 x float> @test_vlseg5ff_mask_nxv2f32_triscv.vector.tuple_nxv8i8_5t(target("riscv.vector.tuple", <vscale x 8 x i8>, 5) %val, ptr %base, i64 %vl, <vscale x 2 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg5ff_mask_nxv2f32_triscv.vector.tuple_nxv8i8_5t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e32, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
 ; CHECK-NEXT:    vmv1r.v v10, v11
 ; CHECK-NEXT:    vmv1r.v v11, v12
-; CHECK-NEXT:    vsetvli zero, a1, e32, m1, ta, mu
 ; CHECK-NEXT:    vlseg5e32ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
@@ -4995,13 +4995,13 @@ entry:
 define <vscale x 1 x float> @test_vlseg6ff_mask_nxv1f32_triscv.vector.tuple_nxv4i8_6t(target("riscv.vector.tuple", <vscale x 4 x i8>, 6) %val, ptr %base, i64 %vl, <vscale x 1 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg6ff_mask_nxv1f32_triscv.vector.tuple_nxv4i8_6t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e32, mf2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
 ; CHECK-NEXT:    vmv1r.v v10, v11
 ; CHECK-NEXT:    vmv1r.v v11, v12
 ; CHECK-NEXT:    vmv1r.v v12, v13
-; CHECK-NEXT:    vsetvli zero, a1, e32, mf2, ta, mu
 ; CHECK-NEXT:    vlseg6e32ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
@@ -5036,13 +5036,13 @@ entry:
 define <vscale x 2 x float> @test_vlseg6ff_mask_nxv2f32_triscv.vector.tuple_nxv8i8_6t(target("riscv.vector.tuple", <vscale x 8 x i8>, 6) %val, ptr %base, i64 %vl, <vscale x 2 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg6ff_mask_nxv2f32_triscv.vector.tuple_nxv8i8_6t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e32, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
 ; CHECK-NEXT:    vmv1r.v v10, v11
 ; CHECK-NEXT:    vmv1r.v v11, v12
 ; CHECK-NEXT:    vmv1r.v v12, v13
-; CHECK-NEXT:    vsetvli zero, a1, e32, m1, ta, mu
 ; CHECK-NEXT:    vlseg6e32ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
@@ -5077,6 +5077,7 @@ entry:
 define <vscale x 1 x float> @test_vlseg7ff_mask_nxv1f32_triscv.vector.tuple_nxv4i8_7t(target("riscv.vector.tuple", <vscale x 4 x i8>, 7) %val, ptr %base, i64 %vl, <vscale x 1 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg7ff_mask_nxv1f32_triscv.vector.tuple_nxv4i8_7t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e32, mf2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
@@ -5084,7 +5085,6 @@ define <vscale x 1 x float> @test_vlseg7ff_mask_nxv1f32_triscv.vector.tuple_nxv4
 ; CHECK-NEXT:    vmv1r.v v11, v12
 ; CHECK-NEXT:    vmv1r.v v12, v13
 ; CHECK-NEXT:    vmv1r.v v13, v14
-; CHECK-NEXT:    vsetvli zero, a1, e32, mf2, ta, mu
 ; CHECK-NEXT:    vlseg7e32ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
@@ -5119,6 +5119,7 @@ entry:
 define <vscale x 2 x float> @test_vlseg7ff_mask_nxv2f32_triscv.vector.tuple_nxv8i8_7t(target("riscv.vector.tuple", <vscale x 8 x i8>, 7) %val, ptr %base, i64 %vl, <vscale x 2 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg7ff_mask_nxv2f32_triscv.vector.tuple_nxv8i8_7t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e32, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
@@ -5126,7 +5127,6 @@ define <vscale x 2 x float> @test_vlseg7ff_mask_nxv2f32_triscv.vector.tuple_nxv8
 ; CHECK-NEXT:    vmv1r.v v11, v12
 ; CHECK-NEXT:    vmv1r.v v12, v13
 ; CHECK-NEXT:    vmv1r.v v13, v14
-; CHECK-NEXT:    vsetvli zero, a1, e32, m1, ta, mu
 ; CHECK-NEXT:    vlseg7e32ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
@@ -5161,6 +5161,7 @@ entry:
 define <vscale x 1 x float> @test_vlseg8ff_mask_nxv1f32_triscv.vector.tuple_nxv4i8_8t(target("riscv.vector.tuple", <vscale x 4 x i8>, 8) %val, ptr %base, i64 %vl, <vscale x 1 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg8ff_mask_nxv1f32_triscv.vector.tuple_nxv4i8_8t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e32, mf2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
@@ -5169,7 +5170,6 @@ define <vscale x 1 x float> @test_vlseg8ff_mask_nxv1f32_triscv.vector.tuple_nxv4
 ; CHECK-NEXT:    vmv1r.v v12, v13
 ; CHECK-NEXT:    vmv1r.v v13, v14
 ; CHECK-NEXT:    vmv1r.v v14, v15
-; CHECK-NEXT:    vsetvli zero, a1, e32, mf2, ta, mu
 ; CHECK-NEXT:    vlseg8e32ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
@@ -5204,6 +5204,7 @@ entry:
 define <vscale x 2 x float> @test_vlseg8ff_mask_nxv2f32_triscv.vector.tuple_nxv8i8_8t(target("riscv.vector.tuple", <vscale x 8 x i8>, 8) %val, ptr %base, i64 %vl, <vscale x 2 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg8ff_mask_nxv2f32_triscv.vector.tuple_nxv8i8_8t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e32, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
@@ -5212,7 +5213,6 @@ define <vscale x 2 x float> @test_vlseg8ff_mask_nxv2f32_triscv.vector.tuple_nxv8
 ; CHECK-NEXT:    vmv1r.v v12, v13
 ; CHECK-NEXT:    vmv1r.v v13, v14
 ; CHECK-NEXT:    vmv1r.v v14, v15
-; CHECK-NEXT:    vsetvli zero, a1, e32, m1, ta, mu
 ; CHECK-NEXT:    vlseg8e32ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
@@ -5247,9 +5247,9 @@ entry:
 define <vscale x 1 x double> @test_vlseg2ff_mask_nxv1f64_triscv.vector.tuple_nxv8i8_2t(target("riscv.vector.tuple", <vscale x 8 x i8>, 2) %val, ptr %base, i64 %vl, <vscale x 1 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg2ff_mask_nxv1f64_triscv.vector.tuple_nxv8i8_2t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e64, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
-; CHECK-NEXT:    vsetvli zero, a1, e64, m1, ta, mu
 ; CHECK-NEXT:    vlseg2e64ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
@@ -5284,9 +5284,9 @@ entry:
 define <vscale x 2 x double> @test_vlseg2ff_mask_nxv2f64_triscv.vector.tuple_nxv16i8_2t(target("riscv.vector.tuple", <vscale x 16 x i8>, 2) %val, ptr %base, i64 %vl, <vscale x 2 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg2ff_mask_nxv2f64_triscv.vector.tuple_nxv16i8_2t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e64, m2, ta, mu
 ; CHECK-NEXT:    vmv2r.v v6, v8
 ; CHECK-NEXT:    vmv2r.v v8, v10
-; CHECK-NEXT:    vsetvli zero, a1, e64, m2, ta, mu
 ; CHECK-NEXT:    vlseg2e64ff.v v6, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
@@ -5321,9 +5321,9 @@ entry:
 define <vscale x 4 x double> @test_vlseg2ff_mask_nxv4f64_triscv.vector.tuple_nxv32i8_2t(target("riscv.vector.tuple", <vscale x 32 x i8>, 2) %val, ptr %base, i64 %vl, <vscale x 4 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg2ff_mask_nxv4f64_triscv.vector.tuple_nxv32i8_2t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e64, m4, ta, mu
 ; CHECK-NEXT:    vmv4r.v v4, v8
 ; CHECK-NEXT:    vmv4r.v v8, v12
-; CHECK-NEXT:    vsetvli zero, a1, e64, m4, ta, mu
 ; CHECK-NEXT:    vlseg2e64ff.v v4, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
@@ -5358,10 +5358,10 @@ entry:
 define <vscale x 1 x double> @test_vlseg3ff_mask_nxv1f64_triscv.vector.tuple_nxv8i8_3t(target("riscv.vector.tuple", <vscale x 8 x i8>, 3) %val, ptr %base, i64 %vl, <vscale x 1 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg3ff_mask_nxv1f64_triscv.vector.tuple_nxv8i8_3t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e64, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
-; CHECK-NEXT:    vsetvli zero, a1, e64, m1, ta, mu
 ; CHECK-NEXT:    vlseg3e64ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
@@ -5396,10 +5396,10 @@ entry:
 define <vscale x 2 x double> @test_vlseg3ff_mask_nxv2f64_triscv.vector.tuple_nxv16i8_3t(target("riscv.vector.tuple", <vscale x 16 x i8>, 3) %val, ptr %base, i64 %vl, <vscale x 2 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg3ff_mask_nxv2f64_triscv.vector.tuple_nxv16i8_3t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e64, m2, ta, mu
 ; CHECK-NEXT:    vmv2r.v v6, v8
 ; CHECK-NEXT:    vmv2r.v v8, v10
 ; CHECK-NEXT:    vmv2r.v v10, v12
-; CHECK-NEXT:    vsetvli zero, a1, e64, m2, ta, mu
 ; CHECK-NEXT:    vlseg3e64ff.v v6, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
@@ -5434,11 +5434,11 @@ entry:
 define <vscale x 1 x double> @test_vlseg4ff_mask_nxv1f64_triscv.vector.tuple_nxv8i8_4t(target("riscv.vector.tuple", <vscale x 8 x i8>, 4) %val, ptr %base, i64 %vl, <vscale x 1 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg4ff_mask_nxv1f64_triscv.vector.tuple_nxv8i8_4t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e64, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
 ; CHECK-NEXT:    vmv1r.v v10, v11
-; CHECK-NEXT:    vsetvli zero, a1, e64, m1, ta, mu
 ; CHECK-NEXT:    vlseg4e64ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
@@ -5473,11 +5473,11 @@ entry:
 define <vscale x 2 x double> @test_vlseg4ff_mask_nxv2f64_triscv.vector.tuple_nxv16i8_4t(target("riscv.vector.tuple", <vscale x 16 x i8>, 4) %val, ptr %base, i64 %vl, <vscale x 2 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg4ff_mask_nxv2f64_triscv.vector.tuple_nxv16i8_4t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e64, m2, ta, mu
 ; CHECK-NEXT:    vmv2r.v v6, v8
 ; CHECK-NEXT:    vmv2r.v v8, v10
 ; CHECK-NEXT:    vmv2r.v v10, v12
 ; CHECK-NEXT:    vmv2r.v v12, v14
-; CHECK-NEXT:    vsetvli zero, a1, e64, m2, ta, mu
 ; CHECK-NEXT:    vlseg4e64ff.v v6, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
@@ -5512,12 +5512,12 @@ entry:
 define <vscale x 1 x double> @test_vlseg5ff_mask_nxv1f64_triscv.vector.tuple_nxv8i8_5t(target("riscv.vector.tuple", <vscale x 8 x i8>, 5) %val, ptr %base, i64 %vl, <vscale x 1 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg5ff_mask_nxv1f64_triscv.vector.tuple_nxv8i8_5t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e64, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
 ; CHECK-NEXT:    vmv1r.v v10, v11
 ; CHECK-NEXT:    vmv1r.v v11, v12
-; CHECK-NEXT:    vsetvli zero, a1, e64, m1, ta, mu
 ; CHECK-NEXT:    vlseg5e64ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
@@ -5552,13 +5552,13 @@ entry:
 define <vscale x 1 x double> @test_vlseg6ff_mask_nxv1f64_triscv.vector.tuple_nxv8i8_6t(target("riscv.vector.tuple", <vscale x 8 x i8>, 6) %val, ptr %base, i64 %vl, <vscale x 1 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg6ff_mask_nxv1f64_triscv.vector.tuple_nxv8i8_6t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e64, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
 ; CHECK-NEXT:    vmv1r.v v10, v11
 ; CHECK-NEXT:    vmv1r.v v11, v12
 ; CHECK-NEXT:    vmv1r.v v12, v13
-; CHECK-NEXT:    vsetvli zero, a1, e64, m1, ta, mu
 ; CHECK-NEXT:    vlseg6e64ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
@@ -5593,6 +5593,7 @@ entry:
 define <vscale x 1 x double> @test_vlseg7ff_mask_nxv1f64_triscv.vector.tuple_nxv8i8_7t(target("riscv.vector.tuple", <vscale x 8 x i8>, 7) %val, ptr %base, i64 %vl, <vscale x 1 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg7ff_mask_nxv1f64_triscv.vector.tuple_nxv8i8_7t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e64, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
@@ -5600,7 +5601,6 @@ define <vscale x 1 x double> @test_vlseg7ff_mask_nxv1f64_triscv.vector.tuple_nxv
 ; CHECK-NEXT:    vmv1r.v v11, v12
 ; CHECK-NEXT:    vmv1r.v v12, v13
 ; CHECK-NEXT:    vmv1r.v v13, v14
-; CHECK-NEXT:    vsetvli zero, a1, e64, m1, ta, mu
 ; CHECK-NEXT:    vlseg7e64ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
@@ -5635,6 +5635,7 @@ entry:
 define <vscale x 1 x double> @test_vlseg8ff_mask_nxv1f64_triscv.vector.tuple_nxv8i8_8t(target("riscv.vector.tuple", <vscale x 8 x i8>, 8) %val, ptr %base, i64 %vl, <vscale x 1 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg8ff_mask_nxv1f64_triscv.vector.tuple_nxv8i8_8t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e64, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
@@ -5643,7 +5644,6 @@ define <vscale x 1 x double> @test_vlseg8ff_mask_nxv1f64_triscv.vector.tuple_nxv
 ; CHECK-NEXT:    vmv1r.v v12, v13
 ; CHECK-NEXT:    vmv1r.v v13, v14
 ; CHECK-NEXT:    vmv1r.v v14, v15
-; CHECK-NEXT:    vsetvli zero, a1, e64, m1, ta, mu
 ; CHECK-NEXT:    vlseg8e64ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
@@ -5678,9 +5678,9 @@ entry:
 define <vscale x 1 x bfloat> @test_vlseg2ff_mask_nxv1bf16_triscv.vector.tuple_nxv2i8_2t(target("riscv.vector.tuple", <vscale x 2 x i8>, 2) %val, ptr %base, i64 %vl, <vscale x 1 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg2ff_mask_nxv1bf16_triscv.vector.tuple_nxv2i8_2t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, mf4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
-; CHECK-NEXT:    vsetvli zero, a1, e16, mf4, ta, mu
 ; CHECK-NEXT:    vlseg2e16ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
@@ -5715,9 +5715,9 @@ entry:
 define <vscale x 2 x bfloat> @test_vlseg2ff_mask_nxv2bf16_triscv.vector.tuple_nxv4i8_2t(target("riscv.vector.tuple", <vscale x 4 x i8>, 2) %val, ptr %base, i64 %vl, <vscale x 2 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg2ff_mask_nxv2bf16_triscv.vector.tuple_nxv4i8_2t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, mf2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
-; CHECK-NEXT:    vsetvli zero, a1, e16, mf2, ta, mu
 ; CHECK-NEXT:    vlseg2e16ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
@@ -5752,9 +5752,9 @@ entry:
 define <vscale x 4 x bfloat> @test_vlseg2ff_mask_nxv4bf16_triscv.vector.tuple_nxv8i8_2t(target("riscv.vector.tuple", <vscale x 8 x i8>, 2) %val, ptr %base, i64 %vl, <vscale x 4 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg2ff_mask_nxv4bf16_triscv.vector.tuple_nxv8i8_2t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
-; CHECK-NEXT:    vsetvli zero, a1, e16, m1, ta, mu
 ; CHECK-NEXT:    vlseg2e16ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
@@ -5789,9 +5789,9 @@ entry:
 define <vscale x 8 x bfloat> @test_vlseg2ff_mask_nxv8bf16_triscv.vector.tuple_nxv16i8_2t(target("riscv.vector.tuple", <vscale x 16 x i8>, 2) %val, ptr %base, i64 %vl, <vscale x 8 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg2ff_mask_nxv8bf16_triscv.vector.tuple_nxv16i8_2t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, m2, ta, mu
 ; CHECK-NEXT:    vmv2r.v v6, v8
 ; CHECK-NEXT:    vmv2r.v v8, v10
-; CHECK-NEXT:    vsetvli zero, a1, e16, m2, ta, mu
 ; CHECK-NEXT:    vlseg2e16ff.v v6, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
@@ -5826,9 +5826,9 @@ entry:
 define <vscale x 16 x bfloat> @test_vlseg2ff_mask_nxv16bf16_triscv.vector.tuple_nxv32i8_2t(target("riscv.vector.tuple", <vscale x 32 x i8>, 2) %val, ptr %base, i64 %vl, <vscale x 16 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg2ff_mask_nxv16bf16_triscv.vector.tuple_nxv32i8_2t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, m4, ta, mu
 ; CHECK-NEXT:    vmv4r.v v4, v8
 ; CHECK-NEXT:    vmv4r.v v8, v12
-; CHECK-NEXT:    vsetvli zero, a1, e16, m4, ta, mu
 ; CHECK-NEXT:    vlseg2e16ff.v v4, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
@@ -5863,10 +5863,10 @@ entry:
 define <vscale x 1 x bfloat> @test_vlseg3ff_mask_nxv1bf16_triscv.vector.tuple_nxv2i8_3t(target("riscv.vector.tuple", <vscale x 2 x i8>, 3) %val, ptr %base, i64 %vl, <vscale x 1 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg3ff_mask_nxv1bf16_triscv.vector.tuple_nxv2i8_3t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, mf4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
-; CHECK-NEXT:    vsetvli zero, a1, e16, mf4, ta, mu
 ; CHECK-NEXT:    vlseg3e16ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
@@ -5901,10 +5901,10 @@ entry:
 define <vscale x 2 x bfloat> @test_vlseg3ff_mask_nxv2bf16_triscv.vector.tuple_nxv4i8_3t(target("riscv.vector.tuple", <vscale x 4 x i8>, 3) %val, ptr %base, i64 %vl, <vscale x 2 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg3ff_mask_nxv2bf16_triscv.vector.tuple_nxv4i8_3t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, mf2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
-; CHECK-NEXT:    vsetvli zero, a1, e16, mf2, ta, mu
 ; CHECK-NEXT:    vlseg3e16ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
@@ -5939,10 +5939,10 @@ entry:
 define <vscale x 4 x bfloat> @test_vlseg3ff_mask_nxv4bf16_triscv.vector.tuple_nxv8i8_3t(target("riscv.vector.tuple", <vscale x 8 x i8>, 3) %val, ptr %base, i64 %vl, <vscale x 4 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg3ff_mask_nxv4bf16_triscv.vector.tuple_nxv8i8_3t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
-; CHECK-NEXT:    vsetvli zero, a1, e16, m1, ta, mu
 ; CHECK-NEXT:    vlseg3e16ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
@@ -5977,10 +5977,10 @@ entry:
 define <vscale x 8 x bfloat> @test_vlseg3ff_mask_nxv8bf16_triscv.vector.tuple_nxv16i8_3t(target("riscv.vector.tuple", <vscale x 16 x i8>, 3) %val, ptr %base, i64 %vl, <vscale x 8 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg3ff_mask_nxv8bf16_triscv.vector.tuple_nxv16i8_3t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, m2, ta, mu
 ; CHECK-NEXT:    vmv2r.v v6, v8
 ; CHECK-NEXT:    vmv2r.v v8, v10
 ; CHECK-NEXT:    vmv2r.v v10, v12
-; CHECK-NEXT:    vsetvli zero, a1, e16, m2, ta, mu
 ; CHECK-NEXT:    vlseg3e16ff.v v6, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
@@ -6015,11 +6015,11 @@ entry:
 define <vscale x 1 x bfloat> @test_vlseg4ff_mask_nxv1bf16_triscv.vector.tuple_nxv2i8_4t(target("riscv.vector.tuple", <vscale x 2 x i8>, 4) %val, ptr %base, i64 %vl, <vscale x 1 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg4ff_mask_nxv1bf16_triscv.vector.tuple_nxv2i8_4t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, mf4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
 ; CHECK-NEXT:    vmv1r.v v10, v11
-; CHECK-NEXT:    vsetvli zero, a1, e16, mf4, ta, mu
 ; CHECK-NEXT:    vlseg4e16ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
@@ -6054,11 +6054,11 @@ entry:
 define <vscale x 2 x bfloat> @test_vlseg4ff_mask_nxv2bf16_triscv.vector.tuple_nxv4i8_4t(target("riscv.vector.tuple", <vscale x 4 x i8>, 4) %val, ptr %base, i64 %vl, <vscale x 2 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg4ff_mask_nxv2bf16_triscv.vector.tuple_nxv4i8_4t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, mf2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
 ; CHECK-NEXT:    vmv1r.v v10, v11
-; CHECK-NEXT:    vsetvli zero, a1, e16, mf2, ta, mu
 ; CHECK-NEXT:    vlseg4e16ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
@@ -6093,11 +6093,11 @@ entry:
 define <vscale x 4 x bfloat> @test_vlseg4ff_mask_nxv4bf16_triscv.vector.tuple_nxv8i8_4t(target("riscv.vector.tuple", <vscale x 8 x i8>, 4) %val, ptr %base, i64 %vl, <vscale x 4 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg4ff_mask_nxv4bf16_triscv.vector.tuple_nxv8i8_4t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
 ; CHECK-NEXT:    vmv1r.v v10, v11
-; CHECK-NEXT:    vsetvli zero, a1, e16, m1, ta, mu
 ; CHECK-NEXT:    vlseg4e16ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
@@ -6132,11 +6132,11 @@ entry:
 define <vscale x 8 x bfloat> @test_vlseg4ff_mask_nxv8bf16_triscv.vector.tuple_nxv16i8_4t(target("riscv.vector.tuple", <vscale x 16 x i8>, 4) %val, ptr %base, i64 %vl, <vscale x 8 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg4ff_mask_nxv8bf16_triscv.vector.tuple_nxv16i8_4t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, m2, ta, mu
 ; CHECK-NEXT:    vmv2r.v v6, v8
 ; CHECK-NEXT:    vmv2r.v v8, v10
 ; CHECK-NEXT:    vmv2r.v v10, v12
 ; CHECK-NEXT:    vmv2r.v v12, v14
-; CHECK-NEXT:    vsetvli zero, a1, e16, m2, ta, mu
 ; CHECK-NEXT:    vlseg4e16ff.v v6, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
@@ -6171,12 +6171,12 @@ entry:
 define <vscale x 1 x bfloat> @test_vlseg5ff_mask_nxv1bf16_triscv.vector.tuple_nxv2i8_5t(target("riscv.vector.tuple", <vscale x 2 x i8>, 5) %val, ptr %base, i64 %vl, <vscale x 1 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg5ff_mask_nxv1bf16_triscv.vector.tuple_nxv2i8_5t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, mf4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
 ; CHECK-NEXT:    vmv1r.v v10, v11
 ; CHECK-NEXT:    vmv1r.v v11, v12
-; CHECK-NEXT:    vsetvli zero, a1, e16, mf4, ta, mu
 ; CHECK-NEXT:    vlseg5e16ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
@@ -6211,12 +6211,12 @@ entry:
 define <vscale x 2 x bfloat> @test_vlseg5ff_mask_nxv2bf16_triscv.vector.tuple_nxv4i8_5t(target("riscv.vector.tuple", <vscale x 4 x i8>, 5) %val, ptr %base, i64 %vl, <vscale x 2 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg5ff_mask_nxv2bf16_triscv.vector.tuple_nxv4i8_5t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, mf2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
 ; CHECK-NEXT:    vmv1r.v v10, v11
 ; CHECK-NEXT:    vmv1r.v v11, v12
-; CHECK-NEXT:    vsetvli zero, a1, e16, mf2, ta, mu
 ; CHECK-NEXT:    vlseg5e16ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
@@ -6251,12 +6251,12 @@ entry:
 define <vscale x 4 x bfloat> @test_vlseg5ff_mask_nxv4bf16_triscv.vector.tuple_nxv8i8_5t(target("riscv.vector.tuple", <vscale x 8 x i8>, 5) %val, ptr %base, i64 %vl, <vscale x 4 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg5ff_mask_nxv4bf16_triscv.vector.tuple_nxv8i8_5t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
 ; CHECK-NEXT:    vmv1r.v v10, v11
 ; CHECK-NEXT:    vmv1r.v v11, v12
-; CHECK-NEXT:    vsetvli zero, a1, e16, m1, ta, mu
 ; CHECK-NEXT:    vlseg5e16ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
@@ -6291,13 +6291,13 @@ entry:
 define <vscale x 1 x bfloat> @test_vlseg6ff_mask_nxv1bf16_triscv.vector.tuple_nxv2i8_6t(target("riscv.vector.tuple", <vscale x 2 x i8>, 6) %val, ptr %base, i64 %vl, <vscale x 1 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg6ff_mask_nxv1bf16_triscv.vector.tuple_nxv2i8_6t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, mf4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
 ; CHECK-NEXT:    vmv1r.v v10, v11
 ; CHECK-NEXT:    vmv1r.v v11, v12
 ; CHECK-NEXT:    vmv1r.v v12, v13
-; CHECK-NEXT:    vsetvli zero, a1, e16, mf4, ta, mu
 ; CHECK-NEXT:    vlseg6e16ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
@@ -6332,13 +6332,13 @@ entry:
 define <vscale x 2 x bfloat> @test_vlseg6ff_mask_nxv2bf16_triscv.vector.tuple_nxv4i8_6t(target("riscv.vector.tuple", <vscale x 4 x i8>, 6) %val, ptr %base, i64 %vl, <vscale x 2 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg6ff_mask_nxv2bf16_triscv.vector.tuple_nxv4i8_6t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, mf2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
 ; CHECK-NEXT:    vmv1r.v v10, v11
 ; CHECK-NEXT:    vmv1r.v v11, v12
 ; CHECK-NEXT:    vmv1r.v v12, v13
-; CHECK-NEXT:    vsetvli zero, a1, e16, mf2, ta, mu
 ; CHECK-NEXT:    vlseg6e16ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
@@ -6373,13 +6373,13 @@ entry:
 define <vscale x 4 x bfloat> @test_vlseg6ff_mask_nxv4bf16_triscv.vector.tuple_nxv8i8_6t(target("riscv.vector.tuple", <vscale x 8 x i8>, 6) %val, ptr %base, i64 %vl, <vscale x 4 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg6ff_mask_nxv4bf16_triscv.vector.tuple_nxv8i8_6t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
 ; CHECK-NEXT:    vmv1r.v v10, v11
 ; CHECK-NEXT:    vmv1r.v v11, v12
 ; CHECK-NEXT:    vmv1r.v v12, v13
-; CHECK-NEXT:    vsetvli zero, a1, e16, m1, ta, mu
 ; CHECK-NEXT:    vlseg6e16ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
@@ -6414,6 +6414,7 @@ entry:
 define <vscale x 1 x bfloat> @test_vlseg7ff_mask_nxv1bf16_triscv.vector.tuple_nxv2i8_7t(target("riscv.vector.tuple", <vscale x 2 x i8>, 7) %val, ptr %base, i64 %vl, <vscale x 1 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg7ff_mask_nxv1bf16_triscv.vector.tuple_nxv2i8_7t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, mf4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
@@ -6421,7 +6422,6 @@ define <vscale x 1 x bfloat> @test_vlseg7ff_mask_nxv1bf16_triscv.vector.tuple_nx
 ; CHECK-NEXT:    vmv1r.v v11, v12
 ; CHECK-NEXT:    vmv1r.v v12, v13
 ; CHECK-NEXT:    vmv1r.v v13, v14
-; CHECK-NEXT:    vsetvli zero, a1, e16, mf4, ta, mu
 ; CHECK-NEXT:    vlseg7e16ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
@@ -6456,6 +6456,7 @@ entry:
 define <vscale x 2 x bfloat> @test_vlseg7ff_mask_nxv2bf16_triscv.vector.tuple_nxv4i8_7t(target("riscv.vector.tuple", <vscale x 4 x i8>, 7) %val, ptr %base, i64 %vl, <vscale x 2 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg7ff_mask_nxv2bf16_triscv.vector.tuple_nxv4i8_7t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, mf2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
@@ -6463,7 +6464,6 @@ define <vscale x 2 x bfloat> @test_vlseg7ff_mask_nxv2bf16_triscv.vector.tuple_nx
 ; CHECK-NEXT:    vmv1r.v v11, v12
 ; CHECK-NEXT:    vmv1r.v v12, v13
 ; CHECK-NEXT:    vmv1r.v v13, v14
-; CHECK-NEXT:    vsetvli zero, a1, e16, mf2, ta, mu
 ; CHECK-NEXT:    vlseg7e16ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
@@ -6498,6 +6498,7 @@ entry:
 define <vscale x 4 x bfloat> @test_vlseg7ff_mask_nxv4bf16_triscv.vector.tuple_nxv8i8_7t(target("riscv.vector.tuple", <vscale x 8 x i8>, 7) %val, ptr %base, i64 %vl, <vscale x 4 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg7ff_mask_nxv4bf16_triscv.vector.tuple_nxv8i8_7t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
@@ -6505,7 +6506,6 @@ define <vscale x 4 x bfloat> @test_vlseg7ff_mask_nxv4bf16_triscv.vector.tuple_nx
 ; CHECK-NEXT:    vmv1r.v v11, v12
 ; CHECK-NEXT:    vmv1r.v v12, v13
 ; CHECK-NEXT:    vmv1r.v v13, v14
-; CHECK-NEXT:    vsetvli zero, a1, e16, m1, ta, mu
 ; CHECK-NEXT:    vlseg7e16ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
@@ -6540,6 +6540,7 @@ entry:
 define <vscale x 1 x bfloat> @test_vlseg8ff_mask_nxv1bf16_triscv.vector.tuple_nxv2i8_8t(target("riscv.vector.tuple", <vscale x 2 x i8>, 8) %val, ptr %base, i64 %vl, <vscale x 1 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg8ff_mask_nxv1bf16_triscv.vector.tuple_nxv2i8_8t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, mf4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
@@ -6548,7 +6549,6 @@ define <vscale x 1 x bfloat> @test_vlseg8ff_mask_nxv1bf16_triscv.vector.tuple_nx
 ; CHECK-NEXT:    vmv1r.v v12, v13
 ; CHECK-NEXT:    vmv1r.v v13, v14
 ; CHECK-NEXT:    vmv1r.v v14, v15
-; CHECK-NEXT:    vsetvli zero, a1, e16, mf4, ta, mu
 ; CHECK-NEXT:    vlseg8e16ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
@@ -6583,6 +6583,7 @@ entry:
 define <vscale x 2 x bfloat> @test_vlseg8ff_mask_nxv2bf16_triscv.vector.tuple_nxv4i8_8t(target("riscv.vector.tuple", <vscale x 4 x i8>, 8) %val, ptr %base, i64 %vl, <vscale x 2 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg8ff_mask_nxv2bf16_triscv.vector.tuple_nxv4i8_8t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, mf2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
@@ -6591,7 +6592,6 @@ define <vscale x 2 x bfloat> @test_vlseg8ff_mask_nxv2bf16_triscv.vector.tuple_nx
 ; CHECK-NEXT:    vmv1r.v v12, v13
 ; CHECK-NEXT:    vmv1r.v v13, v14
 ; CHECK-NEXT:    vmv1r.v v14, v15
-; CHECK-NEXT:    vsetvli zero, a1, e16, mf2, ta, mu
 ; CHECK-NEXT:    vlseg8e16ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
@@ -6626,6 +6626,7 @@ entry:
 define <vscale x 4 x bfloat> @test_vlseg8ff_mask_nxv4bf16_triscv.vector.tuple_nxv8i8_8t(target("riscv.vector.tuple", <vscale x 8 x i8>, 8) %val, ptr %base, i64 %vl, <vscale x 4 x i1> %mask, ptr %outvl) {
 ; CHECK-LABEL: test_vlseg8ff_mask_nxv4bf16_triscv.vector.tuple_nxv8i8_8t:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v7, v8
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    vmv1r.v v9, v10
@@ -6634,7 +6635,6 @@ define <vscale x 4 x bfloat> @test_vlseg8ff_mask_nxv4bf16_triscv.vector.tuple_nx
 ; CHECK-NEXT:    vmv1r.v v12, v13
 ; CHECK-NEXT:    vmv1r.v v13, v14
 ; CHECK-NEXT:    vmv1r.v v14, v15
-; CHECK-NEXT:    vsetvli zero, a1, e16, m1, ta, mu
 ; CHECK-NEXT:    vlseg8e16ff.v v7, (a0), v0.t
 ; CHECK-NEXT:    csrr a0, vl
 ; CHECK-NEXT:    sd a0, 0(a2)
diff --git a/llvm/test/CodeGen/RISCV/rvv/vmax-vp.ll b/llvm/test/CodeGen/RISCV/rvv/vmax-vp.ll
index 0b553d3cd6fdf..b839cd595f3be 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vmax-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vmax-vp.ll
@@ -412,8 +412,8 @@ declare <vscale x 128 x i8> @llvm.vp.smax.nxv128i8(<vscale x 128 x i8>, <vscale
 define <vscale x 128 x i8> @vmax_vx_nxv128i8(<vscale x 128 x i8> %va, i8 %b, <vscale x 128 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vmax_vx_nxv128i8:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v24, v0
 ; CHECK-NEXT:    vsetvli a3, zero, e8, m8, ta, ma
+; CHECK-NEXT:    vmv1r.v v24, v0
 ; CHECK-NEXT:    vlm.v v0, (a1)
 ; CHECK-NEXT:    csrr a1, vlenb
 ; CHECK-NEXT:    slli a1, a1, 3
@@ -974,11 +974,11 @@ declare <vscale x 32 x i32> @llvm.vp.smax.nxv32i32(<vscale x 32 x i32>, <vscale
 define <vscale x 32 x i32> @vmax_vx_nxv32i32(<vscale x 32 x i32> %va, i32 %b, <vscale x 32 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vmax_vx_nxv32i32:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli a2, zero, e8, mf2, ta, ma
 ; CHECK-NEXT:    vmv1r.v v24, v0
 ; CHECK-NEXT:    csrr a2, vlenb
 ; CHECK-NEXT:    srli a3, a2, 2
 ; CHECK-NEXT:    slli a2, a2, 1
-; CHECK-NEXT:    vsetvli a4, zero, e8, mf2, ta, ma
 ; CHECK-NEXT:    vslidedown.vx v0, v0, a3
 ; CHECK-NEXT:    sub a3, a1, a2
 ; CHECK-NEXT:    sltu a4, a1, a3
@@ -1034,11 +1034,11 @@ declare i32 @llvm.vscale.i32()
 define <vscale x 32 x i32> @vmax_vx_nxv32i32_evl_nx8(<vscale x 32 x i32> %va, i32 %b, <vscale x 32 x i1> %m) {
 ; CHECK-LABEL: vmax_vx_nxv32i32_evl_nx8:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli a1, zero, e8, mf2, ta, ma
 ; CHECK-NEXT:    vmv1r.v v24, v0
 ; CHECK-NEXT:    csrr a1, vlenb
 ; CHECK-NEXT:    srli a3, a1, 2
 ; CHECK-NEXT:    slli a2, a1, 1
-; CHECK-NEXT:    vsetvli a4, zero, e8, mf2, ta, ma
 ; CHECK-NEXT:    vslidedown.vx v0, v0, a3
 ; CHECK-NEXT:    sub a3, a1, a2
 ; CHECK-NEXT:    sltu a4, a1, a3
diff --git a/llvm/test/CodeGen/RISCV/rvv/vmaxu-vp.ll b/llvm/test/CodeGen/RISCV/rvv/vmaxu-vp.ll
index f6be882f74206..99e0dfaf90a2d 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vmaxu-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vmaxu-vp.ll
@@ -410,8 +410,8 @@ declare <vscale x 128 x i8> @llvm.vp.umax.nxv128i8(<vscale x 128 x i8>, <vscale
 define <vscale x 128 x i8> @vmaxu_vx_nxv128i8(<vscale x 128 x i8> %va, i8 %b, <vscale x 128 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vmaxu_vx_nxv128i8:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v24, v0
 ; CHECK-NEXT:    vsetvli a3, zero, e8, m8, ta, ma
+; CHECK-NEXT:    vmv1r.v v24, v0
 ; CHECK-NEXT:    vlm.v v0, (a1)
 ; CHECK-NEXT:    csrr a1, vlenb
 ; CHECK-NEXT:    slli a1, a1, 3
@@ -972,11 +972,11 @@ declare <vscale x 32 x i32> @llvm.vp.umax.nxv32i32(<vscale x 32 x i32>, <vscale
 define <vscale x 32 x i32> @vmaxu_vx_nxv32i32(<vscale x 32 x i32> %va, i32 %b, <vscale x 32 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vmaxu_vx_nxv32i32:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli a2, zero, e8, mf2, ta, ma
 ; CHECK-NEXT:    vmv1r.v v24, v0
 ; CHECK-NEXT:    csrr a2, vlenb
 ; CHECK-NEXT:    srli a3, a2, 2
 ; CHECK-NEXT:    slli a2, a2, 1
-; CHECK-NEXT:    vsetvli a4, zero, e8, mf2, ta, ma
 ; CHECK-NEXT:    vslidedown.vx v0, v0, a3
 ; CHECK-NEXT:    sub a3, a1, a2
 ; CHECK-NEXT:    sltu a4, a1, a3
@@ -1032,11 +1032,11 @@ declare i32 @llvm.vscale.i32()
 define <vscale x 32 x i32> @vmaxu_vx_nxv32i32_evl_nx8(<vscale x 32 x i32> %va, i32 %b, <vscale x 32 x i1> %m) {
 ; CHECK-LABEL: vmaxu_vx_nxv32i32_evl_nx8:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli a1, zero, e8, mf2, ta, ma
 ; CHECK-NEXT:    vmv1r.v v24, v0
 ; CHECK-NEXT:    csrr a1, vlenb
 ; CHECK-NEXT:    srli a3, a1, 2
 ; CHECK-NEXT:    slli a2, a1, 1
-; CHECK-NEXT:    vsetvli a4, zero, e8, mf2, ta, ma
 ; CHECK-NEXT:    vslidedown.vx v0, v0, a3
 ; CHECK-NEXT:    sub a3, a1, a2
 ; CHECK-NEXT:    sltu a4, a1, a3
diff --git a/llvm/test/CodeGen/RISCV/rvv/vmfeq.ll b/llvm/test/CodeGen/RISCV/rvv/vmfeq.ll
index 3ebfc68ddee4b..babf8de57b7ea 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vmfeq.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vmfeq.ll
@@ -34,8 +34,8 @@ declare <vscale x 1 x i1> @llvm.riscv.vmfeq.mask.nxv1f16(
 define <vscale x 1 x i1> @intrinsic_vmfeq_mask_vv_nxv1f16_nxv1f16(<vscale x 1 x i1> %0, <vscale x 1 x half> %1, <vscale x 1 x half> %2, <vscale x 1 x half> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmfeq_mask_vv_nxv1f16_nxv1f16:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e16, mf4, ta, mu
+; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmfeq.vv v0, v8, v9
 ; CHECK-NEXT:    vmfeq.vv v11, v9, v10, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v11
@@ -85,8 +85,8 @@ declare <vscale x 2 x i1> @llvm.riscv.vmfeq.mask.nxv2f16(
 define <vscale x 2 x i1> @intrinsic_vmfeq_mask_vv_nxv2f16_nxv2f16(<vscale x 2 x i1> %0, <vscale x 2 x half> %1, <vscale x 2 x half> %2, <vscale x 2 x half> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmfeq_mask_vv_nxv2f16_nxv2f16:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e16, mf2, ta, mu
+; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmfeq.vv v0, v8, v9
 ; CHECK-NEXT:    vmfeq.vv v11, v9, v10, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v11
@@ -136,8 +136,8 @@ declare <vscale x 4 x i1> @llvm.riscv.vmfeq.mask.nxv4f16(
 define <vscale x 4 x i1> @intrinsic_vmfeq_mask_vv_nxv4f16_nxv4f16(<vscale x 4 x i1> %0, <vscale x 4 x half> %1, <vscale x 4 x half> %2, <vscale x 4 x half> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmfeq_mask_vv_nxv4f16_nxv4f16:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e16, m1, ta, mu
+; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmfeq.vv v0, v8, v9
 ; CHECK-NEXT:    vmfeq.vv v11, v9, v10, v0.t
 ; CHECK-NEXT:    vmv.v.v v0, v11
@@ -187,8 +187,8 @@ declare <vscale x 8 x i1> @llvm.riscv.vmfeq.mask.nxv8f16(
 define <vscale x 8 x i1> @intrinsic_vmfeq_mask_vv_nxv8f16_nxv8f16(<vscale x 8 x i1> %0, <vscale x 8 x half> %1, <vscale x 8 x half> %2, <vscale x 8 x half> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmfeq_mask_vv_nxv8f16_nxv8f16:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v14, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e16, m2, ta, mu
+; CHECK-NEXT:    vmv1r.v v14, v0
 ; CHECK-NEXT:    vmfeq.vv v0, v8, v10
 ; CHECK-NEXT:    vmfeq.vv v14, v10, v12, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v14
@@ -238,8 +238,8 @@ declare <vscale x 16 x i1> @llvm.riscv.vmfeq.mask.nxv16f16(
 define <vscale x 16 x i1> @intrinsic_vmfeq_mask_vv_nxv16f16_nxv16f16(<vscale x 16 x i1> %0, <vscale x 16 x half> %1, <vscale x 16 x half> %2, <vscale x 16 x half> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmfeq_mask_vv_nxv16f16_nxv16f16:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v20, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e16, m4, ta, mu
+; CHECK-NEXT:    vmv1r.v v20, v0
 ; CHECK-NEXT:    vmfeq.vv v0, v8, v12
 ; CHECK-NEXT:    vmfeq.vv v20, v12, v16, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v20
@@ -289,8 +289,8 @@ declare <vscale x 1 x i1> @llvm.riscv.vmfeq.mask.nxv1f32(
 define <vscale x 1 x i1> @intrinsic_vmfeq_mask_vv_nxv1f32_nxv1f32(<vscale x 1 x i1> %0, <vscale x 1 x float> %1, <vscale x 1 x float> %2, <vscale x 1 x float> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmfeq_mask_vv_nxv1f32_nxv1f32:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e32, mf2, ta, mu
+; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmfeq.vv v0, v8, v9
 ; CHECK-NEXT:    vmfeq.vv v11, v9, v10, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v11
@@ -340,8 +340,8 @@ declare <vscale x 2 x i1> @llvm.riscv.vmfeq.mask.nxv2f32(
 define <vscale x 2 x i1> @intrinsic_vmfeq_mask_vv_nxv2f32_nxv2f32(<vscale x 2 x i1> %0, <vscale x 2 x float> %1, <vscale x 2 x float> %2, <vscale x 2 x float> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmfeq_mask_vv_nxv2f32_nxv2f32:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e32, m1, ta, mu
+; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmfeq.vv v0, v8, v9
 ; CHECK-NEXT:    vmfeq.vv v11, v9, v10, v0.t
 ; CHECK-NEXT:    vmv.v.v v0, v11
@@ -391,8 +391,8 @@ declare <vscale x 4 x i1> @llvm.riscv.vmfeq.mask.nxv4f32(
 define <vscale x 4 x i1> @intrinsic_vmfeq_mask_vv_nxv4f32_nxv4f32(<vscale x 4 x i1> %0, <vscale x 4 x float> %1, <vscale x 4 x float> %2, <vscale x 4 x float> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmfeq_mask_vv_nxv4f32_nxv4f32:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v14, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e32, m2, ta, mu
+; CHECK-NEXT:    vmv1r.v v14, v0
 ; CHECK-NEXT:    vmfeq.vv v0, v8, v10
 ; CHECK-NEXT:    vmfeq.vv v14, v10, v12, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v14
@@ -442,8 +442,8 @@ declare <vscale x 8 x i1> @llvm.riscv.vmfeq.mask.nxv8f32(
 define <vscale x 8 x i1> @intrinsic_vmfeq_mask_vv_nxv8f32_nxv8f32(<vscale x 8 x i1> %0, <vscale x 8 x float> %1, <vscale x 8 x float> %2, <vscale x 8 x float> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmfeq_mask_vv_nxv8f32_nxv8f32:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v20, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e32, m4, ta, mu
+; CHECK-NEXT:    vmv1r.v v20, v0
 ; CHECK-NEXT:    vmfeq.vv v0, v8, v12
 ; CHECK-NEXT:    vmfeq.vv v20, v12, v16, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v20
@@ -493,8 +493,8 @@ declare <vscale x 1 x i1> @llvm.riscv.vmfeq.mask.nxv1f64(
 define <vscale x 1 x i1> @intrinsic_vmfeq_mask_vv_nxv1f64_nxv1f64(<vscale x 1 x i1> %0, <vscale x 1 x double> %1, <vscale x 1 x double> %2, <vscale x 1 x double> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmfeq_mask_vv_nxv1f64_nxv1f64:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e64, m1, ta, mu
+; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmfeq.vv v0, v8, v9
 ; CHECK-NEXT:    vmfeq.vv v11, v9, v10, v0.t
 ; CHECK-NEXT:    vmv.v.v v0, v11
@@ -544,8 +544,8 @@ declare <vscale x 2 x i1> @llvm.riscv.vmfeq.mask.nxv2f64(
 define <vscale x 2 x i1> @intrinsic_vmfeq_mask_vv_nxv2f64_nxv2f64(<vscale x 2 x i1> %0, <vscale x 2 x double> %1, <vscale x 2 x double> %2, <vscale x 2 x double> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmfeq_mask_vv_nxv2f64_nxv2f64:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v14, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e64, m2, ta, mu
+; CHECK-NEXT:    vmv1r.v v14, v0
 ; CHECK-NEXT:    vmfeq.vv v0, v8, v10
 ; CHECK-NEXT:    vmfeq.vv v14, v10, v12, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v14
@@ -595,8 +595,8 @@ declare <vscale x 4 x i1> @llvm.riscv.vmfeq.mask.nxv4f64(
 define <vscale x 4 x i1> @intrinsic_vmfeq_mask_vv_nxv4f64_nxv4f64(<vscale x 4 x i1> %0, <vscale x 4 x double> %1, <vscale x 4 x double> %2, <vscale x 4 x double> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmfeq_mask_vv_nxv4f64_nxv4f64:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v20, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e64, m4, ta, mu
+; CHECK-NEXT:    vmv1r.v v20, v0
 ; CHECK-NEXT:    vmfeq.vv v0, v8, v12
 ; CHECK-NEXT:    vmfeq.vv v20, v12, v16, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v20
@@ -646,9 +646,9 @@ declare <vscale x 1 x i1> @llvm.riscv.vmfeq.mask.nxv1f16.f16(
 define <vscale x 1 x i1> @intrinsic_vmfeq_mask_vf_nxv1f16_f16(<vscale x 1 x i1> %0, <vscale x 1 x half> %1, half %2, <vscale x 1 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmfeq_mask_vf_nxv1f16_f16:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e16, mf4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a0, e16, mf4, ta, mu
 ; CHECK-NEXT:    vmfeq.vf v10, v8, fa0, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v10
 ; CHECK-NEXT:    ret
@@ -693,9 +693,9 @@ declare <vscale x 2 x i1> @llvm.riscv.vmfeq.mask.nxv2f16.f16(
 define <vscale x 2 x i1> @intrinsic_vmfeq_mask_vf_nxv2f16_f16(<vscale x 2 x i1> %0, <vscale x 2 x half> %1, half %2, <vscale x 2 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmfeq_mask_vf_nxv2f16_f16:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e16, mf2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a0, e16, mf2, ta, mu
 ; CHECK-NEXT:    vmfeq.vf v10, v8, fa0, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v10
 ; CHECK-NEXT:    ret
@@ -740,9 +740,9 @@ declare <vscale x 4 x i1> @llvm.riscv.vmfeq.mask.nxv4f16.f16(
 define <vscale x 4 x i1> @intrinsic_vmfeq_mask_vf_nxv4f16_f16(<vscale x 4 x i1> %0, <vscale x 4 x half> %1, half %2, <vscale x 4 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmfeq_mask_vf_nxv4f16_f16:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e16, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a0, e16, m1, ta, mu
 ; CHECK-NEXT:    vmfeq.vf v10, v8, fa0, v0.t
 ; CHECK-NEXT:    vmv.v.v v0, v10
 ; CHECK-NEXT:    ret
@@ -787,9 +787,9 @@ declare <vscale x 8 x i1> @llvm.riscv.vmfeq.mask.nxv8f16.f16(
 define <vscale x 8 x i1> @intrinsic_vmfeq_mask_vf_nxv8f16_f16(<vscale x 8 x i1> %0, <vscale x 8 x half> %1, half %2, <vscale x 8 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmfeq_mask_vf_nxv8f16_f16:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e16, m2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmv1r.v v0, v10
-; CHECK-NEXT:    vsetvli zero, a0, e16, m2, ta, mu
 ; CHECK-NEXT:    vmfeq.vf v11, v8, fa0, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v11
 ; CHECK-NEXT:    ret
@@ -834,9 +834,9 @@ declare <vscale x 16 x i1> @llvm.riscv.vmfeq.mask.nxv16f16.f16(
 define <vscale x 16 x i1> @intrinsic_vmfeq_mask_vf_nxv16f16_f16(<vscale x 16 x i1> %0, <vscale x 16 x half> %1, half %2, <vscale x 16 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmfeq_mask_vf_nxv16f16_f16:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e16, m4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v13, v0
 ; CHECK-NEXT:    vmv1r.v v0, v12
-; CHECK-NEXT:    vsetvli zero, a0, e16, m4, ta, mu
 ; CHECK-NEXT:    vmfeq.vf v13, v8, fa0, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v13
 ; CHECK-NEXT:    ret
@@ -881,9 +881,9 @@ declare <vscale x 1 x i1> @llvm.riscv.vmfeq.mask.nxv1f32.f32(
 define <vscale x 1 x i1> @intrinsic_vmfeq_mask_vf_nxv1f32_f32(<vscale x 1 x i1> %0, <vscale x 1 x float> %1, float %2, <vscale x 1 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmfeq_mask_vf_nxv1f32_f32:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e32, mf2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a0, e32, mf2, ta, mu
 ; CHECK-NEXT:    vmfeq.vf v10, v8, fa0, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v10
 ; CHECK-NEXT:    ret
@@ -928,9 +928,9 @@ declare <vscale x 2 x i1> @llvm.riscv.vmfeq.mask.nxv2f32.f32(
 define <vscale x 2 x i1> @intrinsic_vmfeq_mask_vf_nxv2f32_f32(<vscale x 2 x i1> %0, <vscale x 2 x float> %1, float %2, <vscale x 2 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmfeq_mask_vf_nxv2f32_f32:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e32, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a0, e32, m1, ta, mu
 ; CHECK-NEXT:    vmfeq.vf v10, v8, fa0, v0.t
 ; CHECK-NEXT:    vmv.v.v v0, v10
 ; CHECK-NEXT:    ret
@@ -975,9 +975,9 @@ declare <vscale x 4 x i1> @llvm.riscv.vmfeq.mask.nxv4f32.f32(
 define <vscale x 4 x i1> @intrinsic_vmfeq_mask_vf_nxv4f32_f32(<vscale x 4 x i1> %0, <vscale x 4 x float> %1, float %2, <vscale x 4 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmfeq_mask_vf_nxv4f32_f32:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e32, m2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmv1r.v v0, v10
-; CHECK-NEXT:    vsetvli zero, a0, e32, m2, ta, mu
 ; CHECK-NEXT:    vmfeq.vf v11, v8, fa0, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v11
 ; CHECK-NEXT:    ret
@@ -1022,9 +1022,9 @@ declare <vscale x 8 x i1> @llvm.riscv.vmfeq.mask.nxv8f32.f32(
 define <vscale x 8 x i1> @intrinsic_vmfeq_mask_vf_nxv8f32_f32(<vscale x 8 x i1> %0, <vscale x 8 x float> %1, float %2, <vscale x 8 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmfeq_mask_vf_nxv8f32_f32:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e32, m4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v13, v0
 ; CHECK-NEXT:    vmv1r.v v0, v12
-; CHECK-NEXT:    vsetvli zero, a0, e32, m4, ta, mu
 ; CHECK-NEXT:    vmfeq.vf v13, v8, fa0, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v13
 ; CHECK-NEXT:    ret
@@ -1069,9 +1069,9 @@ declare <vscale x 1 x i1> @llvm.riscv.vmfeq.mask.nxv1f64.f64(
 define <vscale x 1 x i1> @intrinsic_vmfeq_mask_vf_nxv1f64_f64(<vscale x 1 x i1> %0, <vscale x 1 x double> %1, double %2, <vscale x 1 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmfeq_mask_vf_nxv1f64_f64:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e64, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a0, e64, m1, ta, mu
 ; CHECK-NEXT:    vmfeq.vf v10, v8, fa0, v0.t
 ; CHECK-NEXT:    vmv.v.v v0, v10
 ; CHECK-NEXT:    ret
@@ -1116,9 +1116,9 @@ declare <vscale x 2 x i1> @llvm.riscv.vmfeq.mask.nxv2f64.f64(
 define <vscale x 2 x i1> @intrinsic_vmfeq_mask_vf_nxv2f64_f64(<vscale x 2 x i1> %0, <vscale x 2 x double> %1, double %2, <vscale x 2 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmfeq_mask_vf_nxv2f64_f64:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e64, m2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmv1r.v v0, v10
-; CHECK-NEXT:    vsetvli zero, a0, e64, m2, ta, mu
 ; CHECK-NEXT:    vmfeq.vf v11, v8, fa0, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v11
 ; CHECK-NEXT:    ret
@@ -1163,9 +1163,9 @@ declare <vscale x 4 x i1> @llvm.riscv.vmfeq.mask.nxv4f64.f64(
 define <vscale x 4 x i1> @intrinsic_vmfeq_mask_vf_nxv4f64_f64(<vscale x 4 x i1> %0, <vscale x 4 x double> %1, double %2, <vscale x 4 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmfeq_mask_vf_nxv4f64_f64:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e64, m4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v13, v0
 ; CHECK-NEXT:    vmv1r.v v0, v12
-; CHECK-NEXT:    vsetvli zero, a0, e64, m4, ta, mu
 ; CHECK-NEXT:    vmfeq.vf v13, v8, fa0, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v13
 ; CHECK-NEXT:    ret
diff --git a/llvm/test/CodeGen/RISCV/rvv/vmfge.ll b/llvm/test/CodeGen/RISCV/rvv/vmfge.ll
index e041e5874a8dc..4a9dd2f7d769d 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vmfge.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vmfge.ll
@@ -34,8 +34,8 @@ declare <vscale x 1 x i1> @llvm.riscv.vmfge.mask.nxv1f16(
 define <vscale x 1 x i1> @intrinsic_vmfge_mask_vv_nxv1f16_nxv1f16(<vscale x 1 x i1> %0, <vscale x 1 x half> %1, <vscale x 1 x half> %2, <vscale x 1 x half> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmfge_mask_vv_nxv1f16_nxv1f16:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e16, mf4, ta, mu
+; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmfle.vv v0, v9, v8
 ; CHECK-NEXT:    vmfle.vv v11, v10, v9, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v11
@@ -85,8 +85,8 @@ declare <vscale x 2 x i1> @llvm.riscv.vmfge.mask.nxv2f16(
 define <vscale x 2 x i1> @intrinsic_vmfge_mask_vv_nxv2f16_nxv2f16(<vscale x 2 x i1> %0, <vscale x 2 x half> %1, <vscale x 2 x half> %2, <vscale x 2 x half> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmfge_mask_vv_nxv2f16_nxv2f16:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e16, mf2, ta, mu
+; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmfle.vv v0, v9, v8
 ; CHECK-NEXT:    vmfle.vv v11, v10, v9, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v11
@@ -136,8 +136,8 @@ declare <vscale x 4 x i1> @llvm.riscv.vmfge.mask.nxv4f16(
 define <vscale x 4 x i1> @intrinsic_vmfge_mask_vv_nxv4f16_nxv4f16(<vscale x 4 x i1> %0, <vscale x 4 x half> %1, <vscale x 4 x half> %2, <vscale x 4 x half> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmfge_mask_vv_nxv4f16_nxv4f16:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e16, m1, ta, mu
+; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmfle.vv v0, v9, v8
 ; CHECK-NEXT:    vmfle.vv v11, v10, v9, v0.t
 ; CHECK-NEXT:    vmv.v.v v0, v11
@@ -187,8 +187,8 @@ declare <vscale x 8 x i1> @llvm.riscv.vmfge.mask.nxv8f16(
 define <vscale x 8 x i1> @intrinsic_vmfge_mask_vv_nxv8f16_nxv8f16(<vscale x 8 x i1> %0, <vscale x 8 x half> %1, <vscale x 8 x half> %2, <vscale x 8 x half> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmfge_mask_vv_nxv8f16_nxv8f16:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v14, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e16, m2, ta, mu
+; CHECK-NEXT:    vmv1r.v v14, v0
 ; CHECK-NEXT:    vmfle.vv v0, v10, v8
 ; CHECK-NEXT:    vmfle.vv v14, v12, v10, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v14
@@ -238,8 +238,8 @@ declare <vscale x 16 x i1> @llvm.riscv.vmfge.mask.nxv16f16(
 define <vscale x 16 x i1> @intrinsic_vmfge_mask_vv_nxv16f16_nxv16f16(<vscale x 16 x i1> %0, <vscale x 16 x half> %1, <vscale x 16 x half> %2, <vscale x 16 x half> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmfge_mask_vv_nxv16f16_nxv16f16:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v20, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e16, m4, ta, mu
+; CHECK-NEXT:    vmv1r.v v20, v0
 ; CHECK-NEXT:    vmfle.vv v0, v12, v8
 ; CHECK-NEXT:    vmfle.vv v20, v16, v12, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v20
@@ -289,8 +289,8 @@ declare <vscale x 1 x i1> @llvm.riscv.vmfge.mask.nxv1f32(
 define <vscale x 1 x i1> @intrinsic_vmfge_mask_vv_nxv1f32_nxv1f32(<vscale x 1 x i1> %0, <vscale x 1 x float> %1, <vscale x 1 x float> %2, <vscale x 1 x float> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmfge_mask_vv_nxv1f32_nxv1f32:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e32, mf2, ta, mu
+; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmfle.vv v0, v9, v8
 ; CHECK-NEXT:    vmfle.vv v11, v10, v9, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v11
@@ -340,8 +340,8 @@ declare <vscale x 2 x i1> @llvm.riscv.vmfge.mask.nxv2f32(
 define <vscale x 2 x i1> @intrinsic_vmfge_mask_vv_nxv2f32_nxv2f32(<vscale x 2 x i1> %0, <vscale x 2 x float> %1, <vscale x 2 x float> %2, <vscale x 2 x float> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmfge_mask_vv_nxv2f32_nxv2f32:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e32, m1, ta, mu
+; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmfle.vv v0, v9, v8
 ; CHECK-NEXT:    vmfle.vv v11, v10, v9, v0.t
 ; CHECK-NEXT:    vmv.v.v v0, v11
@@ -391,8 +391,8 @@ declare <vscale x 4 x i1> @llvm.riscv.vmfge.mask.nxv4f32(
 define <vscale x 4 x i1> @intrinsic_vmfge_mask_vv_nxv4f32_nxv4f32(<vscale x 4 x i1> %0, <vscale x 4 x float> %1, <vscale x 4 x float> %2, <vscale x 4 x float> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmfge_mask_vv_nxv4f32_nxv4f32:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v14, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e32, m2, ta, mu
+; CHECK-NEXT:    vmv1r.v v14, v0
 ; CHECK-NEXT:    vmfle.vv v0, v10, v8
 ; CHECK-NEXT:    vmfle.vv v14, v12, v10, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v14
@@ -442,8 +442,8 @@ declare <vscale x 8 x i1> @llvm.riscv.vmfge.mask.nxv8f32(
 define <vscale x 8 x i1> @intrinsic_vmfge_mask_vv_nxv8f32_nxv8f32(<vscale x 8 x i1> %0, <vscale x 8 x float> %1, <vscale x 8 x float> %2, <vscale x 8 x float> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmfge_mask_vv_nxv8f32_nxv8f32:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v20, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e32, m4, ta, mu
+; CHECK-NEXT:    vmv1r.v v20, v0
 ; CHECK-NEXT:    vmfle.vv v0, v12, v8
 ; CHECK-NEXT:    vmfle.vv v20, v16, v12, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v20
@@ -493,8 +493,8 @@ declare <vscale x 1 x i1> @llvm.riscv.vmfge.mask.nxv1f64(
 define <vscale x 1 x i1> @intrinsic_vmfge_mask_vv_nxv1f64_nxv1f64(<vscale x 1 x i1> %0, <vscale x 1 x double> %1, <vscale x 1 x double> %2, <vscale x 1 x double> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmfge_mask_vv_nxv1f64_nxv1f64:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e64, m1, ta, mu
+; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmfle.vv v0, v9, v8
 ; CHECK-NEXT:    vmfle.vv v11, v10, v9, v0.t
 ; CHECK-NEXT:    vmv.v.v v0, v11
@@ -544,8 +544,8 @@ declare <vscale x 2 x i1> @llvm.riscv.vmfge.mask.nxv2f64(
 define <vscale x 2 x i1> @intrinsic_vmfge_mask_vv_nxv2f64_nxv2f64(<vscale x 2 x i1> %0, <vscale x 2 x double> %1, <vscale x 2 x double> %2, <vscale x 2 x double> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmfge_mask_vv_nxv2f64_nxv2f64:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v14, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e64, m2, ta, mu
+; CHECK-NEXT:    vmv1r.v v14, v0
 ; CHECK-NEXT:    vmfle.vv v0, v10, v8
 ; CHECK-NEXT:    vmfle.vv v14, v12, v10, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v14
@@ -595,8 +595,8 @@ declare <vscale x 4 x i1> @llvm.riscv.vmfge.mask.nxv4f64(
 define <vscale x 4 x i1> @intrinsic_vmfge_mask_vv_nxv4f64_nxv4f64(<vscale x 4 x i1> %0, <vscale x 4 x double> %1, <vscale x 4 x double> %2, <vscale x 4 x double> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmfge_mask_vv_nxv4f64_nxv4f64:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v20, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e64, m4, ta, mu
+; CHECK-NEXT:    vmv1r.v v20, v0
 ; CHECK-NEXT:    vmfle.vv v0, v12, v8
 ; CHECK-NEXT:    vmfle.vv v20, v16, v12, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v20
@@ -646,9 +646,9 @@ declare <vscale x 1 x i1> @llvm.riscv.vmfge.mask.nxv1f16.f16(
 define <vscale x 1 x i1> @intrinsic_vmfge_mask_vf_nxv1f16_f16(<vscale x 1 x i1> %0, <vscale x 1 x half> %1, half %2, <vscale x 1 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmfge_mask_vf_nxv1f16_f16:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e16, mf4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a0, e16, mf4, ta, mu
 ; CHECK-NEXT:    vmfge.vf v10, v8, fa0, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v10
 ; CHECK-NEXT:    ret
@@ -693,9 +693,9 @@ declare <vscale x 2 x i1> @llvm.riscv.vmfge.mask.nxv2f16.f16(
 define <vscale x 2 x i1> @intrinsic_vmfge_mask_vf_nxv2f16_f16(<vscale x 2 x i1> %0, <vscale x 2 x half> %1, half %2, <vscale x 2 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmfge_mask_vf_nxv2f16_f16:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e16, mf2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a0, e16, mf2, ta, mu
 ; CHECK-NEXT:    vmfge.vf v10, v8, fa0, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v10
 ; CHECK-NEXT:    ret
@@ -740,9 +740,9 @@ declare <vscale x 4 x i1> @llvm.riscv.vmfge.mask.nxv4f16.f16(
 define <vscale x 4 x i1> @intrinsic_vmfge_mask_vf_nxv4f16_f16(<vscale x 4 x i1> %0, <vscale x 4 x half> %1, half %2, <vscale x 4 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmfge_mask_vf_nxv4f16_f16:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e16, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a0, e16, m1, ta, mu
 ; CHECK-NEXT:    vmfge.vf v10, v8, fa0, v0.t
 ; CHECK-NEXT:    vmv.v.v v0, v10
 ; CHECK-NEXT:    ret
@@ -787,9 +787,9 @@ declare <vscale x 8 x i1> @llvm.riscv.vmfge.mask.nxv8f16.f16(
 define <vscale x 8 x i1> @intrinsic_vmfge_mask_vf_nxv8f16_f16(<vscale x 8 x i1> %0, <vscale x 8 x half> %1, half %2, <vscale x 8 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmfge_mask_vf_nxv8f16_f16:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e16, m2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmv1r.v v0, v10
-; CHECK-NEXT:    vsetvli zero, a0, e16, m2, ta, mu
 ; CHECK-NEXT:    vmfge.vf v11, v8, fa0, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v11
 ; CHECK-NEXT:    ret
@@ -834,9 +834,9 @@ declare <vscale x 16 x i1> @llvm.riscv.vmfge.mask.nxv16f16.f16(
 define <vscale x 16 x i1> @intrinsic_vmfge_mask_vf_nxv16f16_f16(<vscale x 16 x i1> %0, <vscale x 16 x half> %1, half %2, <vscale x 16 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmfge_mask_vf_nxv16f16_f16:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e16, m4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v13, v0
 ; CHECK-NEXT:    vmv1r.v v0, v12
-; CHECK-NEXT:    vsetvli zero, a0, e16, m4, ta, mu
 ; CHECK-NEXT:    vmfge.vf v13, v8, fa0, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v13
 ; CHECK-NEXT:    ret
@@ -881,9 +881,9 @@ declare <vscale x 1 x i1> @llvm.riscv.vmfge.mask.nxv1f32.f32(
 define <vscale x 1 x i1> @intrinsic_vmfge_mask_vf_nxv1f32_f32(<vscale x 1 x i1> %0, <vscale x 1 x float> %1, float %2, <vscale x 1 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmfge_mask_vf_nxv1f32_f32:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e32, mf2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a0, e32, mf2, ta, mu
 ; CHECK-NEXT:    vmfge.vf v10, v8, fa0, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v10
 ; CHECK-NEXT:    ret
@@ -928,9 +928,9 @@ declare <vscale x 2 x i1> @llvm.riscv.vmfge.mask.nxv2f32.f32(
 define <vscale x 2 x i1> @intrinsic_vmfge_mask_vf_nxv2f32_f32(<vscale x 2 x i1> %0, <vscale x 2 x float> %1, float %2, <vscale x 2 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmfge_mask_vf_nxv2f32_f32:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e32, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a0, e32, m1, ta, mu
 ; CHECK-NEXT:    vmfge.vf v10, v8, fa0, v0.t
 ; CHECK-NEXT:    vmv.v.v v0, v10
 ; CHECK-NEXT:    ret
@@ -975,9 +975,9 @@ declare <vscale x 4 x i1> @llvm.riscv.vmfge.mask.nxv4f32.f32(
 define <vscale x 4 x i1> @intrinsic_vmfge_mask_vf_nxv4f32_f32(<vscale x 4 x i1> %0, <vscale x 4 x float> %1, float %2, <vscale x 4 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmfge_mask_vf_nxv4f32_f32:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e32, m2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmv1r.v v0, v10
-; CHECK-NEXT:    vsetvli zero, a0, e32, m2, ta, mu
 ; CHECK-NEXT:    vmfge.vf v11, v8, fa0, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v11
 ; CHECK-NEXT:    ret
@@ -1022,9 +1022,9 @@ declare <vscale x 8 x i1> @llvm.riscv.vmfge.mask.nxv8f32.f32(
 define <vscale x 8 x i1> @intrinsic_vmfge_mask_vf_nxv8f32_f32(<vscale x 8 x i1> %0, <vscale x 8 x float> %1, float %2, <vscale x 8 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmfge_mask_vf_nxv8f32_f32:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e32, m4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v13, v0
 ; CHECK-NEXT:    vmv1r.v v0, v12
-; CHECK-NEXT:    vsetvli zero, a0, e32, m4, ta, mu
 ; CHECK-NEXT:    vmfge.vf v13, v8, fa0, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v13
 ; CHECK-NEXT:    ret
@@ -1069,9 +1069,9 @@ declare <vscale x 1 x i1> @llvm.riscv.vmfge.mask.nxv1f64.f64(
 define <vscale x 1 x i1> @intrinsic_vmfge_mask_vf_nxv1f64_f64(<vscale x 1 x i1> %0, <vscale x 1 x double> %1, double %2, <vscale x 1 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmfge_mask_vf_nxv1f64_f64:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e64, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a0, e64, m1, ta, mu
 ; CHECK-NEXT:    vmfge.vf v10, v8, fa0, v0.t
 ; CHECK-NEXT:    vmv.v.v v0, v10
 ; CHECK-NEXT:    ret
@@ -1116,9 +1116,9 @@ declare <vscale x 2 x i1> @llvm.riscv.vmfge.mask.nxv2f64.f64(
 define <vscale x 2 x i1> @intrinsic_vmfge_mask_vf_nxv2f64_f64(<vscale x 2 x i1> %0, <vscale x 2 x double> %1, double %2, <vscale x 2 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmfge_mask_vf_nxv2f64_f64:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e64, m2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmv1r.v v0, v10
-; CHECK-NEXT:    vsetvli zero, a0, e64, m2, ta, mu
 ; CHECK-NEXT:    vmfge.vf v11, v8, fa0, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v11
 ; CHECK-NEXT:    ret
@@ -1163,9 +1163,9 @@ declare <vscale x 4 x i1> @llvm.riscv.vmfge.mask.nxv4f64.f64(
 define <vscale x 4 x i1> @intrinsic_vmfge_mask_vf_nxv4f64_f64(<vscale x 4 x i1> %0, <vscale x 4 x double> %1, double %2, <vscale x 4 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmfge_mask_vf_nxv4f64_f64:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e64, m4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v13, v0
 ; CHECK-NEXT:    vmv1r.v v0, v12
-; CHECK-NEXT:    vsetvli zero, a0, e64, m4, ta, mu
 ; CHECK-NEXT:    vmfge.vf v13, v8, fa0, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v13
 ; CHECK-NEXT:    ret
diff --git a/llvm/test/CodeGen/RISCV/rvv/vmfgt.ll b/llvm/test/CodeGen/RISCV/rvv/vmfgt.ll
index 0faaf4ebf255d..c9c5e84937cec 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vmfgt.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vmfgt.ll
@@ -34,8 +34,8 @@ declare <vscale x 1 x i1> @llvm.riscv.vmfgt.mask.nxv1f16(
 define <vscale x 1 x i1> @intrinsic_vmfgt_mask_vv_nxv1f16_nxv1f16(<vscale x 1 x i1> %0, <vscale x 1 x half> %1, <vscale x 1 x half> %2, <vscale x 1 x half> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmfgt_mask_vv_nxv1f16_nxv1f16:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e16, mf4, ta, mu
+; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmflt.vv v0, v9, v8
 ; CHECK-NEXT:    vmflt.vv v11, v10, v9, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v11
@@ -85,8 +85,8 @@ declare <vscale x 2 x i1> @llvm.riscv.vmfgt.mask.nxv2f16(
 define <vscale x 2 x i1> @intrinsic_vmfgt_mask_vv_nxv2f16_nxv2f16(<vscale x 2 x i1> %0, <vscale x 2 x half> %1, <vscale x 2 x half> %2, <vscale x 2 x half> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmfgt_mask_vv_nxv2f16_nxv2f16:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e16, mf2, ta, mu
+; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmflt.vv v0, v9, v8
 ; CHECK-NEXT:    vmflt.vv v11, v10, v9, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v11
@@ -136,8 +136,8 @@ declare <vscale x 4 x i1> @llvm.riscv.vmfgt.mask.nxv4f16(
 define <vscale x 4 x i1> @intrinsic_vmfgt_mask_vv_nxv4f16_nxv4f16(<vscale x 4 x i1> %0, <vscale x 4 x half> %1, <vscale x 4 x half> %2, <vscale x 4 x half> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmfgt_mask_vv_nxv4f16_nxv4f16:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e16, m1, ta, mu
+; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmflt.vv v0, v9, v8
 ; CHECK-NEXT:    vmflt.vv v11, v10, v9, v0.t
 ; CHECK-NEXT:    vmv.v.v v0, v11
@@ -187,8 +187,8 @@ declare <vscale x 8 x i1> @llvm.riscv.vmfgt.mask.nxv8f16(
 define <vscale x 8 x i1> @intrinsic_vmfgt_mask_vv_nxv8f16_nxv8f16(<vscale x 8 x i1> %0, <vscale x 8 x half> %1, <vscale x 8 x half> %2, <vscale x 8 x half> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmfgt_mask_vv_nxv8f16_nxv8f16:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v14, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e16, m2, ta, mu
+; CHECK-NEXT:    vmv1r.v v14, v0
 ; CHECK-NEXT:    vmflt.vv v0, v10, v8
 ; CHECK-NEXT:    vmflt.vv v14, v12, v10, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v14
@@ -238,8 +238,8 @@ declare <vscale x 16 x i1> @llvm.riscv.vmfgt.mask.nxv16f16(
 define <vscale x 16 x i1> @intrinsic_vmfgt_mask_vv_nxv16f16_nxv16f16(<vscale x 16 x i1> %0, <vscale x 16 x half> %1, <vscale x 16 x half> %2, <vscale x 16 x half> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmfgt_mask_vv_nxv16f16_nxv16f16:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v20, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e16, m4, ta, mu
+; CHECK-NEXT:    vmv1r.v v20, v0
 ; CHECK-NEXT:    vmflt.vv v0, v12, v8
 ; CHECK-NEXT:    vmflt.vv v20, v16, v12, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v20
@@ -289,8 +289,8 @@ declare <vscale x 1 x i1> @llvm.riscv.vmfgt.mask.nxv1f32(
 define <vscale x 1 x i1> @intrinsic_vmfgt_mask_vv_nxv1f32_nxv1f32(<vscale x 1 x i1> %0, <vscale x 1 x float> %1, <vscale x 1 x float> %2, <vscale x 1 x float> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmfgt_mask_vv_nxv1f32_nxv1f32:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e32, mf2, ta, mu
+; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmflt.vv v0, v9, v8
 ; CHECK-NEXT:    vmflt.vv v11, v10, v9, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v11
@@ -340,8 +340,8 @@ declare <vscale x 2 x i1> @llvm.riscv.vmfgt.mask.nxv2f32(
 define <vscale x 2 x i1> @intrinsic_vmfgt_mask_vv_nxv2f32_nxv2f32(<vscale x 2 x i1> %0, <vscale x 2 x float> %1, <vscale x 2 x float> %2, <vscale x 2 x float> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmfgt_mask_vv_nxv2f32_nxv2f32:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e32, m1, ta, mu
+; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmflt.vv v0, v9, v8
 ; CHECK-NEXT:    vmflt.vv v11, v10, v9, v0.t
 ; CHECK-NEXT:    vmv.v.v v0, v11
@@ -391,8 +391,8 @@ declare <vscale x 4 x i1> @llvm.riscv.vmfgt.mask.nxv4f32(
 define <vscale x 4 x i1> @intrinsic_vmfgt_mask_vv_nxv4f32_nxv4f32(<vscale x 4 x i1> %0, <vscale x 4 x float> %1, <vscale x 4 x float> %2, <vscale x 4 x float> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmfgt_mask_vv_nxv4f32_nxv4f32:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v14, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e32, m2, ta, mu
+; CHECK-NEXT:    vmv1r.v v14, v0
 ; CHECK-NEXT:    vmflt.vv v0, v10, v8
 ; CHECK-NEXT:    vmflt.vv v14, v12, v10, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v14
@@ -442,8 +442,8 @@ declare <vscale x 8 x i1> @llvm.riscv.vmfgt.mask.nxv8f32(
 define <vscale x 8 x i1> @intrinsic_vmfgt_mask_vv_nxv8f32_nxv8f32(<vscale x 8 x i1> %0, <vscale x 8 x float> %1, <vscale x 8 x float> %2, <vscale x 8 x float> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmfgt_mask_vv_nxv8f32_nxv8f32:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v20, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e32, m4, ta, mu
+; CHECK-NEXT:    vmv1r.v v20, v0
 ; CHECK-NEXT:    vmflt.vv v0, v12, v8
 ; CHECK-NEXT:    vmflt.vv v20, v16, v12, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v20
@@ -493,8 +493,8 @@ declare <vscale x 1 x i1> @llvm.riscv.vmfgt.mask.nxv1f64(
 define <vscale x 1 x i1> @intrinsic_vmfgt_mask_vv_nxv1f64_nxv1f64(<vscale x 1 x i1> %0, <vscale x 1 x double> %1, <vscale x 1 x double> %2, <vscale x 1 x double> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmfgt_mask_vv_nxv1f64_nxv1f64:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e64, m1, ta, mu
+; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmflt.vv v0, v9, v8
 ; CHECK-NEXT:    vmflt.vv v11, v10, v9, v0.t
 ; CHECK-NEXT:    vmv.v.v v0, v11
@@ -544,8 +544,8 @@ declare <vscale x 2 x i1> @llvm.riscv.vmfgt.mask.nxv2f64(
 define <vscale x 2 x i1> @intrinsic_vmfgt_mask_vv_nxv2f64_nxv2f64(<vscale x 2 x i1> %0, <vscale x 2 x double> %1, <vscale x 2 x double> %2, <vscale x 2 x double> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmfgt_mask_vv_nxv2f64_nxv2f64:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v14, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e64, m2, ta, mu
+; CHECK-NEXT:    vmv1r.v v14, v0
 ; CHECK-NEXT:    vmflt.vv v0, v10, v8
 ; CHECK-NEXT:    vmflt.vv v14, v12, v10, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v14
@@ -595,8 +595,8 @@ declare <vscale x 4 x i1> @llvm.riscv.vmfgt.mask.nxv4f64(
 define <vscale x 4 x i1> @intrinsic_vmfgt_mask_vv_nxv4f64_nxv4f64(<vscale x 4 x i1> %0, <vscale x 4 x double> %1, <vscale x 4 x double> %2, <vscale x 4 x double> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmfgt_mask_vv_nxv4f64_nxv4f64:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v20, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e64, m4, ta, mu
+; CHECK-NEXT:    vmv1r.v v20, v0
 ; CHECK-NEXT:    vmflt.vv v0, v12, v8
 ; CHECK-NEXT:    vmflt.vv v20, v16, v12, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v20
@@ -646,9 +646,9 @@ declare <vscale x 1 x i1> @llvm.riscv.vmfgt.mask.nxv1f16.f16(
 define <vscale x 1 x i1> @intrinsic_vmfgt_mask_vf_nxv1f16_f16(<vscale x 1 x i1> %0, <vscale x 1 x half> %1, half %2, <vscale x 1 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmfgt_mask_vf_nxv1f16_f16:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e16, mf4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a0, e16, mf4, ta, mu
 ; CHECK-NEXT:    vmfgt.vf v10, v8, fa0, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v10
 ; CHECK-NEXT:    ret
@@ -693,9 +693,9 @@ declare <vscale x 2 x i1> @llvm.riscv.vmfgt.mask.nxv2f16.f16(
 define <vscale x 2 x i1> @intrinsic_vmfgt_mask_vf_nxv2f16_f16(<vscale x 2 x i1> %0, <vscale x 2 x half> %1, half %2, <vscale x 2 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmfgt_mask_vf_nxv2f16_f16:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e16, mf2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a0, e16, mf2, ta, mu
 ; CHECK-NEXT:    vmfgt.vf v10, v8, fa0, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v10
 ; CHECK-NEXT:    ret
@@ -740,9 +740,9 @@ declare <vscale x 4 x i1> @llvm.riscv.vmfgt.mask.nxv4f16.f16(
 define <vscale x 4 x i1> @intrinsic_vmfgt_mask_vf_nxv4f16_f16(<vscale x 4 x i1> %0, <vscale x 4 x half> %1, half %2, <vscale x 4 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmfgt_mask_vf_nxv4f16_f16:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e16, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a0, e16, m1, ta, mu
 ; CHECK-NEXT:    vmfgt.vf v10, v8, fa0, v0.t
 ; CHECK-NEXT:    vmv.v.v v0, v10
 ; CHECK-NEXT:    ret
@@ -787,9 +787,9 @@ declare <vscale x 8 x i1> @llvm.riscv.vmfgt.mask.nxv8f16.f16(
 define <vscale x 8 x i1> @intrinsic_vmfgt_mask_vf_nxv8f16_f16(<vscale x 8 x i1> %0, <vscale x 8 x half> %1, half %2, <vscale x 8 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmfgt_mask_vf_nxv8f16_f16:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e16, m2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmv1r.v v0, v10
-; CHECK-NEXT:    vsetvli zero, a0, e16, m2, ta, mu
 ; CHECK-NEXT:    vmfgt.vf v11, v8, fa0, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v11
 ; CHECK-NEXT:    ret
@@ -834,9 +834,9 @@ declare <vscale x 16 x i1> @llvm.riscv.vmfgt.mask.nxv16f16.f16(
 define <vscale x 16 x i1> @intrinsic_vmfgt_mask_vf_nxv16f16_f16(<vscale x 16 x i1> %0, <vscale x 16 x half> %1, half %2, <vscale x 16 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmfgt_mask_vf_nxv16f16_f16:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e16, m4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v13, v0
 ; CHECK-NEXT:    vmv1r.v v0, v12
-; CHECK-NEXT:    vsetvli zero, a0, e16, m4, ta, mu
 ; CHECK-NEXT:    vmfgt.vf v13, v8, fa0, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v13
 ; CHECK-NEXT:    ret
@@ -881,9 +881,9 @@ declare <vscale x 1 x i1> @llvm.riscv.vmfgt.mask.nxv1f32.f32(
 define <vscale x 1 x i1> @intrinsic_vmfgt_mask_vf_nxv1f32_f32(<vscale x 1 x i1> %0, <vscale x 1 x float> %1, float %2, <vscale x 1 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmfgt_mask_vf_nxv1f32_f32:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e32, mf2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a0, e32, mf2, ta, mu
 ; CHECK-NEXT:    vmfgt.vf v10, v8, fa0, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v10
 ; CHECK-NEXT:    ret
@@ -928,9 +928,9 @@ declare <vscale x 2 x i1> @llvm.riscv.vmfgt.mask.nxv2f32.f32(
 define <vscale x 2 x i1> @intrinsic_vmfgt_mask_vf_nxv2f32_f32(<vscale x 2 x i1> %0, <vscale x 2 x float> %1, float %2, <vscale x 2 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmfgt_mask_vf_nxv2f32_f32:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e32, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a0, e32, m1, ta, mu
 ; CHECK-NEXT:    vmfgt.vf v10, v8, fa0, v0.t
 ; CHECK-NEXT:    vmv.v.v v0, v10
 ; CHECK-NEXT:    ret
@@ -975,9 +975,9 @@ declare <vscale x 4 x i1> @llvm.riscv.vmfgt.mask.nxv4f32.f32(
 define <vscale x 4 x i1> @intrinsic_vmfgt_mask_vf_nxv4f32_f32(<vscale x 4 x i1> %0, <vscale x 4 x float> %1, float %2, <vscale x 4 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmfgt_mask_vf_nxv4f32_f32:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e32, m2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmv1r.v v0, v10
-; CHECK-NEXT:    vsetvli zero, a0, e32, m2, ta, mu
 ; CHECK-NEXT:    vmfgt.vf v11, v8, fa0, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v11
 ; CHECK-NEXT:    ret
@@ -1022,9 +1022,9 @@ declare <vscale x 8 x i1> @llvm.riscv.vmfgt.mask.nxv8f32.f32(
 define <vscale x 8 x i1> @intrinsic_vmfgt_mask_vf_nxv8f32_f32(<vscale x 8 x i1> %0, <vscale x 8 x float> %1, float %2, <vscale x 8 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmfgt_mask_vf_nxv8f32_f32:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e32, m4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v13, v0
 ; CHECK-NEXT:    vmv1r.v v0, v12
-; CHECK-NEXT:    vsetvli zero, a0, e32, m4, ta, mu
 ; CHECK-NEXT:    vmfgt.vf v13, v8, fa0, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v13
 ; CHECK-NEXT:    ret
@@ -1069,9 +1069,9 @@ declare <vscale x 1 x i1> @llvm.riscv.vmfgt.mask.nxv1f64.f64(
 define <vscale x 1 x i1> @intrinsic_vmfgt_mask_vf_nxv1f64_f64(<vscale x 1 x i1> %0, <vscale x 1 x double> %1, double %2, <vscale x 1 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmfgt_mask_vf_nxv1f64_f64:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e64, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a0, e64, m1, ta, mu
 ; CHECK-NEXT:    vmfgt.vf v10, v8, fa0, v0.t
 ; CHECK-NEXT:    vmv.v.v v0, v10
 ; CHECK-NEXT:    ret
@@ -1116,9 +1116,9 @@ declare <vscale x 2 x i1> @llvm.riscv.vmfgt.mask.nxv2f64.f64(
 define <vscale x 2 x i1> @intrinsic_vmfgt_mask_vf_nxv2f64_f64(<vscale x 2 x i1> %0, <vscale x 2 x double> %1, double %2, <vscale x 2 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmfgt_mask_vf_nxv2f64_f64:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e64, m2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmv1r.v v0, v10
-; CHECK-NEXT:    vsetvli zero, a0, e64, m2, ta, mu
 ; CHECK-NEXT:    vmfgt.vf v11, v8, fa0, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v11
 ; CHECK-NEXT:    ret
@@ -1163,9 +1163,9 @@ declare <vscale x 4 x i1> @llvm.riscv.vmfgt.mask.nxv4f64.f64(
 define <vscale x 4 x i1> @intrinsic_vmfgt_mask_vf_nxv4f64_f64(<vscale x 4 x i1> %0, <vscale x 4 x double> %1, double %2, <vscale x 4 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmfgt_mask_vf_nxv4f64_f64:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e64, m4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v13, v0
 ; CHECK-NEXT:    vmv1r.v v0, v12
-; CHECK-NEXT:    vsetvli zero, a0, e64, m4, ta, mu
 ; CHECK-NEXT:    vmfgt.vf v13, v8, fa0, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v13
 ; CHECK-NEXT:    ret
diff --git a/llvm/test/CodeGen/RISCV/rvv/vmfle.ll b/llvm/test/CodeGen/RISCV/rvv/vmfle.ll
index ef5de6bc3481f..77d8dda258961 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vmfle.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vmfle.ll
@@ -34,8 +34,8 @@ declare <vscale x 1 x i1> @llvm.riscv.vmfle.mask.nxv1f16(
 define <vscale x 1 x i1> @intrinsic_vmfle_mask_vv_nxv1f16_nxv1f16(<vscale x 1 x i1> %0, <vscale x 1 x half> %1, <vscale x 1 x half> %2, <vscale x 1 x half> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmfle_mask_vv_nxv1f16_nxv1f16:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e16, mf4, ta, mu
+; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmfle.vv v0, v8, v9
 ; CHECK-NEXT:    vmfle.vv v11, v9, v10, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v11
@@ -85,8 +85,8 @@ declare <vscale x 2 x i1> @llvm.riscv.vmfle.mask.nxv2f16(
 define <vscale x 2 x i1> @intrinsic_vmfle_mask_vv_nxv2f16_nxv2f16(<vscale x 2 x i1> %0, <vscale x 2 x half> %1, <vscale x 2 x half> %2, <vscale x 2 x half> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmfle_mask_vv_nxv2f16_nxv2f16:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e16, mf2, ta, mu
+; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmfle.vv v0, v8, v9
 ; CHECK-NEXT:    vmfle.vv v11, v9, v10, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v11
@@ -136,8 +136,8 @@ declare <vscale x 4 x i1> @llvm.riscv.vmfle.mask.nxv4f16(
 define <vscale x 4 x i1> @intrinsic_vmfle_mask_vv_nxv4f16_nxv4f16(<vscale x 4 x i1> %0, <vscale x 4 x half> %1, <vscale x 4 x half> %2, <vscale x 4 x half> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmfle_mask_vv_nxv4f16_nxv4f16:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e16, m1, ta, mu
+; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmfle.vv v0, v8, v9
 ; CHECK-NEXT:    vmfle.vv v11, v9, v10, v0.t
 ; CHECK-NEXT:    vmv.v.v v0, v11
@@ -187,8 +187,8 @@ declare <vscale x 8 x i1> @llvm.riscv.vmfle.mask.nxv8f16(
 define <vscale x 8 x i1> @intrinsic_vmfle_mask_vv_nxv8f16_nxv8f16(<vscale x 8 x i1> %0, <vscale x 8 x half> %1, <vscale x 8 x half> %2, <vscale x 8 x half> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmfle_mask_vv_nxv8f16_nxv8f16:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v14, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e16, m2, ta, mu
+; CHECK-NEXT:    vmv1r.v v14, v0
 ; CHECK-NEXT:    vmfle.vv v0, v8, v10
 ; CHECK-NEXT:    vmfle.vv v14, v10, v12, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v14
@@ -238,8 +238,8 @@ declare <vscale x 16 x i1> @llvm.riscv.vmfle.mask.nxv16f16(
 define <vscale x 16 x i1> @intrinsic_vmfle_mask_vv_nxv16f16_nxv16f16(<vscale x 16 x i1> %0, <vscale x 16 x half> %1, <vscale x 16 x half> %2, <vscale x 16 x half> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmfle_mask_vv_nxv16f16_nxv16f16:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v20, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e16, m4, ta, mu
+; CHECK-NEXT:    vmv1r.v v20, v0
 ; CHECK-NEXT:    vmfle.vv v0, v8, v12
 ; CHECK-NEXT:    vmfle.vv v20, v12, v16, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v20
@@ -289,8 +289,8 @@ declare <vscale x 1 x i1> @llvm.riscv.vmfle.mask.nxv1f32(
 define <vscale x 1 x i1> @intrinsic_vmfle_mask_vv_nxv1f32_nxv1f32(<vscale x 1 x i1> %0, <vscale x 1 x float> %1, <vscale x 1 x float> %2, <vscale x 1 x float> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmfle_mask_vv_nxv1f32_nxv1f32:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e32, mf2, ta, mu
+; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmfle.vv v0, v8, v9
 ; CHECK-NEXT:    vmfle.vv v11, v9, v10, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v11
@@ -340,8 +340,8 @@ declare <vscale x 2 x i1> @llvm.riscv.vmfle.mask.nxv2f32(
 define <vscale x 2 x i1> @intrinsic_vmfle_mask_vv_nxv2f32_nxv2f32(<vscale x 2 x i1> %0, <vscale x 2 x float> %1, <vscale x 2 x float> %2, <vscale x 2 x float> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmfle_mask_vv_nxv2f32_nxv2f32:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e32, m1, ta, mu
+; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmfle.vv v0, v8, v9
 ; CHECK-NEXT:    vmfle.vv v11, v9, v10, v0.t
 ; CHECK-NEXT:    vmv.v.v v0, v11
@@ -391,8 +391,8 @@ declare <vscale x 4 x i1> @llvm.riscv.vmfle.mask.nxv4f32(
 define <vscale x 4 x i1> @intrinsic_vmfle_mask_vv_nxv4f32_nxv4f32(<vscale x 4 x i1> %0, <vscale x 4 x float> %1, <vscale x 4 x float> %2, <vscale x 4 x float> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmfle_mask_vv_nxv4f32_nxv4f32:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v14, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e32, m2, ta, mu
+; CHECK-NEXT:    vmv1r.v v14, v0
 ; CHECK-NEXT:    vmfle.vv v0, v8, v10
 ; CHECK-NEXT:    vmfle.vv v14, v10, v12, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v14
@@ -442,8 +442,8 @@ declare <vscale x 8 x i1> @llvm.riscv.vmfle.mask.nxv8f32(
 define <vscale x 8 x i1> @intrinsic_vmfle_mask_vv_nxv8f32_nxv8f32(<vscale x 8 x i1> %0, <vscale x 8 x float> %1, <vscale x 8 x float> %2, <vscale x 8 x float> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmfle_mask_vv_nxv8f32_nxv8f32:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v20, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e32, m4, ta, mu
+; CHECK-NEXT:    vmv1r.v v20, v0
 ; CHECK-NEXT:    vmfle.vv v0, v8, v12
 ; CHECK-NEXT:    vmfle.vv v20, v12, v16, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v20
@@ -493,8 +493,8 @@ declare <vscale x 1 x i1> @llvm.riscv.vmfle.mask.nxv1f64(
 define <vscale x 1 x i1> @intrinsic_vmfle_mask_vv_nxv1f64_nxv1f64(<vscale x 1 x i1> %0, <vscale x 1 x double> %1, <vscale x 1 x double> %2, <vscale x 1 x double> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmfle_mask_vv_nxv1f64_nxv1f64:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e64, m1, ta, mu
+; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmfle.vv v0, v8, v9
 ; CHECK-NEXT:    vmfle.vv v11, v9, v10, v0.t
 ; CHECK-NEXT:    vmv.v.v v0, v11
@@ -544,8 +544,8 @@ declare <vscale x 2 x i1> @llvm.riscv.vmfle.mask.nxv2f64(
 define <vscale x 2 x i1> @intrinsic_vmfle_mask_vv_nxv2f64_nxv2f64(<vscale x 2 x i1> %0, <vscale x 2 x double> %1, <vscale x 2 x double> %2, <vscale x 2 x double> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmfle_mask_vv_nxv2f64_nxv2f64:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v14, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e64, m2, ta, mu
+; CHECK-NEXT:    vmv1r.v v14, v0
 ; CHECK-NEXT:    vmfle.vv v0, v8, v10
 ; CHECK-NEXT:    vmfle.vv v14, v10, v12, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v14
@@ -595,8 +595,8 @@ declare <vscale x 4 x i1> @llvm.riscv.vmfle.mask.nxv4f64(
 define <vscale x 4 x i1> @intrinsic_vmfle_mask_vv_nxv4f64_nxv4f64(<vscale x 4 x i1> %0, <vscale x 4 x double> %1, <vscale x 4 x double> %2, <vscale x 4 x double> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmfle_mask_vv_nxv4f64_nxv4f64:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v20, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e64, m4, ta, mu
+; CHECK-NEXT:    vmv1r.v v20, v0
 ; CHECK-NEXT:    vmfle.vv v0, v8, v12
 ; CHECK-NEXT:    vmfle.vv v20, v12, v16, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v20
@@ -646,9 +646,9 @@ declare <vscale x 1 x i1> @llvm.riscv.vmfle.mask.nxv1f16.f16(
 define <vscale x 1 x i1> @intrinsic_vmfle_mask_vf_nxv1f16_f16(<vscale x 1 x i1> %0, <vscale x 1 x half> %1, half %2, <vscale x 1 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmfle_mask_vf_nxv1f16_f16:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e16, mf4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a0, e16, mf4, ta, mu
 ; CHECK-NEXT:    vmfle.vf v10, v8, fa0, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v10
 ; CHECK-NEXT:    ret
@@ -693,9 +693,9 @@ declare <vscale x 2 x i1> @llvm.riscv.vmfle.mask.nxv2f16.f16(
 define <vscale x 2 x i1> @intrinsic_vmfle_mask_vf_nxv2f16_f16(<vscale x 2 x i1> %0, <vscale x 2 x half> %1, half %2, <vscale x 2 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmfle_mask_vf_nxv2f16_f16:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e16, mf2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a0, e16, mf2, ta, mu
 ; CHECK-NEXT:    vmfle.vf v10, v8, fa0, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v10
 ; CHECK-NEXT:    ret
@@ -740,9 +740,9 @@ declare <vscale x 4 x i1> @llvm.riscv.vmfle.mask.nxv4f16.f16(
 define <vscale x 4 x i1> @intrinsic_vmfle_mask_vf_nxv4f16_f16(<vscale x 4 x i1> %0, <vscale x 4 x half> %1, half %2, <vscale x 4 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmfle_mask_vf_nxv4f16_f16:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e16, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a0, e16, m1, ta, mu
 ; CHECK-NEXT:    vmfle.vf v10, v8, fa0, v0.t
 ; CHECK-NEXT:    vmv.v.v v0, v10
 ; CHECK-NEXT:    ret
@@ -787,9 +787,9 @@ declare <vscale x 8 x i1> @llvm.riscv.vmfle.mask.nxv8f16.f16(
 define <vscale x 8 x i1> @intrinsic_vmfle_mask_vf_nxv8f16_f16(<vscale x 8 x i1> %0, <vscale x 8 x half> %1, half %2, <vscale x 8 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmfle_mask_vf_nxv8f16_f16:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e16, m2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmv1r.v v0, v10
-; CHECK-NEXT:    vsetvli zero, a0, e16, m2, ta, mu
 ; CHECK-NEXT:    vmfle.vf v11, v8, fa0, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v11
 ; CHECK-NEXT:    ret
@@ -834,9 +834,9 @@ declare <vscale x 16 x i1> @llvm.riscv.vmfle.mask.nxv16f16.f16(
 define <vscale x 16 x i1> @intrinsic_vmfle_mask_vf_nxv16f16_f16(<vscale x 16 x i1> %0, <vscale x 16 x half> %1, half %2, <vscale x 16 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmfle_mask_vf_nxv16f16_f16:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e16, m4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v13, v0
 ; CHECK-NEXT:    vmv1r.v v0, v12
-; CHECK-NEXT:    vsetvli zero, a0, e16, m4, ta, mu
 ; CHECK-NEXT:    vmfle.vf v13, v8, fa0, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v13
 ; CHECK-NEXT:    ret
@@ -881,9 +881,9 @@ declare <vscale x 1 x i1> @llvm.riscv.vmfle.mask.nxv1f32.f32(
 define <vscale x 1 x i1> @intrinsic_vmfle_mask_vf_nxv1f32_f32(<vscale x 1 x i1> %0, <vscale x 1 x float> %1, float %2, <vscale x 1 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmfle_mask_vf_nxv1f32_f32:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e32, mf2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a0, e32, mf2, ta, mu
 ; CHECK-NEXT:    vmfle.vf v10, v8, fa0, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v10
 ; CHECK-NEXT:    ret
@@ -928,9 +928,9 @@ declare <vscale x 2 x i1> @llvm.riscv.vmfle.mask.nxv2f32.f32(
 define <vscale x 2 x i1> @intrinsic_vmfle_mask_vf_nxv2f32_f32(<vscale x 2 x i1> %0, <vscale x 2 x float> %1, float %2, <vscale x 2 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmfle_mask_vf_nxv2f32_f32:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e32, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a0, e32, m1, ta, mu
 ; CHECK-NEXT:    vmfle.vf v10, v8, fa0, v0.t
 ; CHECK-NEXT:    vmv.v.v v0, v10
 ; CHECK-NEXT:    ret
@@ -975,9 +975,9 @@ declare <vscale x 4 x i1> @llvm.riscv.vmfle.mask.nxv4f32.f32(
 define <vscale x 4 x i1> @intrinsic_vmfle_mask_vf_nxv4f32_f32(<vscale x 4 x i1> %0, <vscale x 4 x float> %1, float %2, <vscale x 4 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmfle_mask_vf_nxv4f32_f32:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e32, m2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmv1r.v v0, v10
-; CHECK-NEXT:    vsetvli zero, a0, e32, m2, ta, mu
 ; CHECK-NEXT:    vmfle.vf v11, v8, fa0, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v11
 ; CHECK-NEXT:    ret
@@ -1022,9 +1022,9 @@ declare <vscale x 8 x i1> @llvm.riscv.vmfle.mask.nxv8f32.f32(
 define <vscale x 8 x i1> @intrinsic_vmfle_mask_vf_nxv8f32_f32(<vscale x 8 x i1> %0, <vscale x 8 x float> %1, float %2, <vscale x 8 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmfle_mask_vf_nxv8f32_f32:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e32, m4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v13, v0
 ; CHECK-NEXT:    vmv1r.v v0, v12
-; CHECK-NEXT:    vsetvli zero, a0, e32, m4, ta, mu
 ; CHECK-NEXT:    vmfle.vf v13, v8, fa0, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v13
 ; CHECK-NEXT:    ret
@@ -1069,9 +1069,9 @@ declare <vscale x 1 x i1> @llvm.riscv.vmfle.mask.nxv1f64.f64(
 define <vscale x 1 x i1> @intrinsic_vmfle_mask_vf_nxv1f64_f64(<vscale x 1 x i1> %0, <vscale x 1 x double> %1, double %2, <vscale x 1 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmfle_mask_vf_nxv1f64_f64:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e64, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a0, e64, m1, ta, mu
 ; CHECK-NEXT:    vmfle.vf v10, v8, fa0, v0.t
 ; CHECK-NEXT:    vmv.v.v v0, v10
 ; CHECK-NEXT:    ret
@@ -1116,9 +1116,9 @@ declare <vscale x 2 x i1> @llvm.riscv.vmfle.mask.nxv2f64.f64(
 define <vscale x 2 x i1> @intrinsic_vmfle_mask_vf_nxv2f64_f64(<vscale x 2 x i1> %0, <vscale x 2 x double> %1, double %2, <vscale x 2 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmfle_mask_vf_nxv2f64_f64:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e64, m2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmv1r.v v0, v10
-; CHECK-NEXT:    vsetvli zero, a0, e64, m2, ta, mu
 ; CHECK-NEXT:    vmfle.vf v11, v8, fa0, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v11
 ; CHECK-NEXT:    ret
@@ -1163,9 +1163,9 @@ declare <vscale x 4 x i1> @llvm.riscv.vmfle.mask.nxv4f64.f64(
 define <vscale x 4 x i1> @intrinsic_vmfle_mask_vf_nxv4f64_f64(<vscale x 4 x i1> %0, <vscale x 4 x double> %1, double %2, <vscale x 4 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmfle_mask_vf_nxv4f64_f64:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e64, m4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v13, v0
 ; CHECK-NEXT:    vmv1r.v v0, v12
-; CHECK-NEXT:    vsetvli zero, a0, e64, m4, ta, mu
 ; CHECK-NEXT:    vmfle.vf v13, v8, fa0, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v13
 ; CHECK-NEXT:    ret
diff --git a/llvm/test/CodeGen/RISCV/rvv/vmflt.ll b/llvm/test/CodeGen/RISCV/rvv/vmflt.ll
index 0b7740d5e0045..0fdae8abe8f6b 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vmflt.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vmflt.ll
@@ -34,8 +34,8 @@ declare <vscale x 1 x i1> @llvm.riscv.vmflt.mask.nxv1f16(
 define <vscale x 1 x i1> @intrinsic_vmflt_mask_vv_nxv1f16_nxv1f16(<vscale x 1 x i1> %0, <vscale x 1 x half> %1, <vscale x 1 x half> %2, <vscale x 1 x half> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmflt_mask_vv_nxv1f16_nxv1f16:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e16, mf4, ta, mu
+; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmflt.vv v0, v8, v9
 ; CHECK-NEXT:    vmflt.vv v11, v9, v10, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v11
@@ -85,8 +85,8 @@ declare <vscale x 2 x i1> @llvm.riscv.vmflt.mask.nxv2f16(
 define <vscale x 2 x i1> @intrinsic_vmflt_mask_vv_nxv2f16_nxv2f16(<vscale x 2 x i1> %0, <vscale x 2 x half> %1, <vscale x 2 x half> %2, <vscale x 2 x half> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmflt_mask_vv_nxv2f16_nxv2f16:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e16, mf2, ta, mu
+; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmflt.vv v0, v8, v9
 ; CHECK-NEXT:    vmflt.vv v11, v9, v10, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v11
@@ -136,8 +136,8 @@ declare <vscale x 4 x i1> @llvm.riscv.vmflt.mask.nxv4f16(
 define <vscale x 4 x i1> @intrinsic_vmflt_mask_vv_nxv4f16_nxv4f16(<vscale x 4 x i1> %0, <vscale x 4 x half> %1, <vscale x 4 x half> %2, <vscale x 4 x half> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmflt_mask_vv_nxv4f16_nxv4f16:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e16, m1, ta, mu
+; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmflt.vv v0, v8, v9
 ; CHECK-NEXT:    vmflt.vv v11, v9, v10, v0.t
 ; CHECK-NEXT:    vmv.v.v v0, v11
@@ -187,8 +187,8 @@ declare <vscale x 8 x i1> @llvm.riscv.vmflt.mask.nxv8f16(
 define <vscale x 8 x i1> @intrinsic_vmflt_mask_vv_nxv8f16_nxv8f16(<vscale x 8 x i1> %0, <vscale x 8 x half> %1, <vscale x 8 x half> %2, <vscale x 8 x half> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmflt_mask_vv_nxv8f16_nxv8f16:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v14, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e16, m2, ta, mu
+; CHECK-NEXT:    vmv1r.v v14, v0
 ; CHECK-NEXT:    vmflt.vv v0, v8, v10
 ; CHECK-NEXT:    vmflt.vv v14, v10, v12, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v14
@@ -238,8 +238,8 @@ declare <vscale x 16 x i1> @llvm.riscv.vmflt.mask.nxv16f16(
 define <vscale x 16 x i1> @intrinsic_vmflt_mask_vv_nxv16f16_nxv16f16(<vscale x 16 x i1> %0, <vscale x 16 x half> %1, <vscale x 16 x half> %2, <vscale x 16 x half> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmflt_mask_vv_nxv16f16_nxv16f16:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v20, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e16, m4, ta, mu
+; CHECK-NEXT:    vmv1r.v v20, v0
 ; CHECK-NEXT:    vmflt.vv v0, v8, v12
 ; CHECK-NEXT:    vmflt.vv v20, v12, v16, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v20
@@ -289,8 +289,8 @@ declare <vscale x 1 x i1> @llvm.riscv.vmflt.mask.nxv1f32(
 define <vscale x 1 x i1> @intrinsic_vmflt_mask_vv_nxv1f32_nxv1f32(<vscale x 1 x i1> %0, <vscale x 1 x float> %1, <vscale x 1 x float> %2, <vscale x 1 x float> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmflt_mask_vv_nxv1f32_nxv1f32:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e32, mf2, ta, mu
+; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmflt.vv v0, v8, v9
 ; CHECK-NEXT:    vmflt.vv v11, v9, v10, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v11
@@ -340,8 +340,8 @@ declare <vscale x 2 x i1> @llvm.riscv.vmflt.mask.nxv2f32(
 define <vscale x 2 x i1> @intrinsic_vmflt_mask_vv_nxv2f32_nxv2f32(<vscale x 2 x i1> %0, <vscale x 2 x float> %1, <vscale x 2 x float> %2, <vscale x 2 x float> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmflt_mask_vv_nxv2f32_nxv2f32:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e32, m1, ta, mu
+; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmflt.vv v0, v8, v9
 ; CHECK-NEXT:    vmflt.vv v11, v9, v10, v0.t
 ; CHECK-NEXT:    vmv.v.v v0, v11
@@ -391,8 +391,8 @@ declare <vscale x 4 x i1> @llvm.riscv.vmflt.mask.nxv4f32(
 define <vscale x 4 x i1> @intrinsic_vmflt_mask_vv_nxv4f32_nxv4f32(<vscale x 4 x i1> %0, <vscale x 4 x float> %1, <vscale x 4 x float> %2, <vscale x 4 x float> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmflt_mask_vv_nxv4f32_nxv4f32:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v14, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e32, m2, ta, mu
+; CHECK-NEXT:    vmv1r.v v14, v0
 ; CHECK-NEXT:    vmflt.vv v0, v8, v10
 ; CHECK-NEXT:    vmflt.vv v14, v10, v12, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v14
@@ -442,8 +442,8 @@ declare <vscale x 8 x i1> @llvm.riscv.vmflt.mask.nxv8f32(
 define <vscale x 8 x i1> @intrinsic_vmflt_mask_vv_nxv8f32_nxv8f32(<vscale x 8 x i1> %0, <vscale x 8 x float> %1, <vscale x 8 x float> %2, <vscale x 8 x float> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmflt_mask_vv_nxv8f32_nxv8f32:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v20, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e32, m4, ta, mu
+; CHECK-NEXT:    vmv1r.v v20, v0
 ; CHECK-NEXT:    vmflt.vv v0, v8, v12
 ; CHECK-NEXT:    vmflt.vv v20, v12, v16, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v20
@@ -493,8 +493,8 @@ declare <vscale x 1 x i1> @llvm.riscv.vmflt.mask.nxv1f64(
 define <vscale x 1 x i1> @intrinsic_vmflt_mask_vv_nxv1f64_nxv1f64(<vscale x 1 x i1> %0, <vscale x 1 x double> %1, <vscale x 1 x double> %2, <vscale x 1 x double> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmflt_mask_vv_nxv1f64_nxv1f64:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e64, m1, ta, mu
+; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmflt.vv v0, v8, v9
 ; CHECK-NEXT:    vmflt.vv v11, v9, v10, v0.t
 ; CHECK-NEXT:    vmv.v.v v0, v11
@@ -544,8 +544,8 @@ declare <vscale x 2 x i1> @llvm.riscv.vmflt.mask.nxv2f64(
 define <vscale x 2 x i1> @intrinsic_vmflt_mask_vv_nxv2f64_nxv2f64(<vscale x 2 x i1> %0, <vscale x 2 x double> %1, <vscale x 2 x double> %2, <vscale x 2 x double> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmflt_mask_vv_nxv2f64_nxv2f64:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v14, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e64, m2, ta, mu
+; CHECK-NEXT:    vmv1r.v v14, v0
 ; CHECK-NEXT:    vmflt.vv v0, v8, v10
 ; CHECK-NEXT:    vmflt.vv v14, v10, v12, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v14
@@ -595,8 +595,8 @@ declare <vscale x 4 x i1> @llvm.riscv.vmflt.mask.nxv4f64(
 define <vscale x 4 x i1> @intrinsic_vmflt_mask_vv_nxv4f64_nxv4f64(<vscale x 4 x i1> %0, <vscale x 4 x double> %1, <vscale x 4 x double> %2, <vscale x 4 x double> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmflt_mask_vv_nxv4f64_nxv4f64:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v20, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e64, m4, ta, mu
+; CHECK-NEXT:    vmv1r.v v20, v0
 ; CHECK-NEXT:    vmflt.vv v0, v8, v12
 ; CHECK-NEXT:    vmflt.vv v20, v12, v16, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v20
@@ -646,9 +646,9 @@ declare <vscale x 1 x i1> @llvm.riscv.vmflt.mask.nxv1f16.f16(
 define <vscale x 1 x i1> @intrinsic_vmflt_mask_vf_nxv1f16_f16(<vscale x 1 x i1> %0, <vscale x 1 x half> %1, half %2, <vscale x 1 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmflt_mask_vf_nxv1f16_f16:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e16, mf4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a0, e16, mf4, ta, mu
 ; CHECK-NEXT:    vmflt.vf v10, v8, fa0, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v10
 ; CHECK-NEXT:    ret
@@ -693,9 +693,9 @@ declare <vscale x 2 x i1> @llvm.riscv.vmflt.mask.nxv2f16.f16(
 define <vscale x 2 x i1> @intrinsic_vmflt_mask_vf_nxv2f16_f16(<vscale x 2 x i1> %0, <vscale x 2 x half> %1, half %2, <vscale x 2 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmflt_mask_vf_nxv2f16_f16:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e16, mf2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a0, e16, mf2, ta, mu
 ; CHECK-NEXT:    vmflt.vf v10, v8, fa0, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v10
 ; CHECK-NEXT:    ret
@@ -740,9 +740,9 @@ declare <vscale x 4 x i1> @llvm.riscv.vmflt.mask.nxv4f16.f16(
 define <vscale x 4 x i1> @intrinsic_vmflt_mask_vf_nxv4f16_f16(<vscale x 4 x i1> %0, <vscale x 4 x half> %1, half %2, <vscale x 4 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmflt_mask_vf_nxv4f16_f16:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e16, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a0, e16, m1, ta, mu
 ; CHECK-NEXT:    vmflt.vf v10, v8, fa0, v0.t
 ; CHECK-NEXT:    vmv.v.v v0, v10
 ; CHECK-NEXT:    ret
@@ -787,9 +787,9 @@ declare <vscale x 8 x i1> @llvm.riscv.vmflt.mask.nxv8f16.f16(
 define <vscale x 8 x i1> @intrinsic_vmflt_mask_vf_nxv8f16_f16(<vscale x 8 x i1> %0, <vscale x 8 x half> %1, half %2, <vscale x 8 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmflt_mask_vf_nxv8f16_f16:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e16, m2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmv1r.v v0, v10
-; CHECK-NEXT:    vsetvli zero, a0, e16, m2, ta, mu
 ; CHECK-NEXT:    vmflt.vf v11, v8, fa0, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v11
 ; CHECK-NEXT:    ret
@@ -834,9 +834,9 @@ declare <vscale x 16 x i1> @llvm.riscv.vmflt.mask.nxv16f16.f16(
 define <vscale x 16 x i1> @intrinsic_vmflt_mask_vf_nxv16f16_f16(<vscale x 16 x i1> %0, <vscale x 16 x half> %1, half %2, <vscale x 16 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmflt_mask_vf_nxv16f16_f16:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e16, m4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v13, v0
 ; CHECK-NEXT:    vmv1r.v v0, v12
-; CHECK-NEXT:    vsetvli zero, a0, e16, m4, ta, mu
 ; CHECK-NEXT:    vmflt.vf v13, v8, fa0, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v13
 ; CHECK-NEXT:    ret
@@ -881,9 +881,9 @@ declare <vscale x 1 x i1> @llvm.riscv.vmflt.mask.nxv1f32.f32(
 define <vscale x 1 x i1> @intrinsic_vmflt_mask_vf_nxv1f32_f32(<vscale x 1 x i1> %0, <vscale x 1 x float> %1, float %2, <vscale x 1 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmflt_mask_vf_nxv1f32_f32:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e32, mf2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a0, e32, mf2, ta, mu
 ; CHECK-NEXT:    vmflt.vf v10, v8, fa0, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v10
 ; CHECK-NEXT:    ret
@@ -928,9 +928,9 @@ declare <vscale x 2 x i1> @llvm.riscv.vmflt.mask.nxv2f32.f32(
 define <vscale x 2 x i1> @intrinsic_vmflt_mask_vf_nxv2f32_f32(<vscale x 2 x i1> %0, <vscale x 2 x float> %1, float %2, <vscale x 2 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmflt_mask_vf_nxv2f32_f32:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e32, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a0, e32, m1, ta, mu
 ; CHECK-NEXT:    vmflt.vf v10, v8, fa0, v0.t
 ; CHECK-NEXT:    vmv.v.v v0, v10
 ; CHECK-NEXT:    ret
@@ -975,9 +975,9 @@ declare <vscale x 4 x i1> @llvm.riscv.vmflt.mask.nxv4f32.f32(
 define <vscale x 4 x i1> @intrinsic_vmflt_mask_vf_nxv4f32_f32(<vscale x 4 x i1> %0, <vscale x 4 x float> %1, float %2, <vscale x 4 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmflt_mask_vf_nxv4f32_f32:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e32, m2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmv1r.v v0, v10
-; CHECK-NEXT:    vsetvli zero, a0, e32, m2, ta, mu
 ; CHECK-NEXT:    vmflt.vf v11, v8, fa0, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v11
 ; CHECK-NEXT:    ret
@@ -1022,9 +1022,9 @@ declare <vscale x 8 x i1> @llvm.riscv.vmflt.mask.nxv8f32.f32(
 define <vscale x 8 x i1> @intrinsic_vmflt_mask_vf_nxv8f32_f32(<vscale x 8 x i1> %0, <vscale x 8 x float> %1, float %2, <vscale x 8 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmflt_mask_vf_nxv8f32_f32:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e32, m4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v13, v0
 ; CHECK-NEXT:    vmv1r.v v0, v12
-; CHECK-NEXT:    vsetvli zero, a0, e32, m4, ta, mu
 ; CHECK-NEXT:    vmflt.vf v13, v8, fa0, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v13
 ; CHECK-NEXT:    ret
@@ -1069,9 +1069,9 @@ declare <vscale x 1 x i1> @llvm.riscv.vmflt.mask.nxv1f64.f64(
 define <vscale x 1 x i1> @intrinsic_vmflt_mask_vf_nxv1f64_f64(<vscale x 1 x i1> %0, <vscale x 1 x double> %1, double %2, <vscale x 1 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmflt_mask_vf_nxv1f64_f64:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e64, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a0, e64, m1, ta, mu
 ; CHECK-NEXT:    vmflt.vf v10, v8, fa0, v0.t
 ; CHECK-NEXT:    vmv.v.v v0, v10
 ; CHECK-NEXT:    ret
@@ -1116,9 +1116,9 @@ declare <vscale x 2 x i1> @llvm.riscv.vmflt.mask.nxv2f64.f64(
 define <vscale x 2 x i1> @intrinsic_vmflt_mask_vf_nxv2f64_f64(<vscale x 2 x i1> %0, <vscale x 2 x double> %1, double %2, <vscale x 2 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmflt_mask_vf_nxv2f64_f64:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e64, m2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmv1r.v v0, v10
-; CHECK-NEXT:    vsetvli zero, a0, e64, m2, ta, mu
 ; CHECK-NEXT:    vmflt.vf v11, v8, fa0, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v11
 ; CHECK-NEXT:    ret
@@ -1163,9 +1163,9 @@ declare <vscale x 4 x i1> @llvm.riscv.vmflt.mask.nxv4f64.f64(
 define <vscale x 4 x i1> @intrinsic_vmflt_mask_vf_nxv4f64_f64(<vscale x 4 x i1> %0, <vscale x 4 x double> %1, double %2, <vscale x 4 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmflt_mask_vf_nxv4f64_f64:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e64, m4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v13, v0
 ; CHECK-NEXT:    vmv1r.v v0, v12
-; CHECK-NEXT:    vsetvli zero, a0, e64, m4, ta, mu
 ; CHECK-NEXT:    vmflt.vf v13, v8, fa0, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v13
 ; CHECK-NEXT:    ret
diff --git a/llvm/test/CodeGen/RISCV/rvv/vmfne.ll b/llvm/test/CodeGen/RISCV/rvv/vmfne.ll
index 65a04e504a973..1d0227f793728 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vmfne.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vmfne.ll
@@ -34,8 +34,8 @@ declare <vscale x 1 x i1> @llvm.riscv.vmfne.mask.nxv1f16(
 define <vscale x 1 x i1> @intrinsic_vmfne_mask_vv_nxv1f16_nxv1f16(<vscale x 1 x i1> %0, <vscale x 1 x half> %1, <vscale x 1 x half> %2, <vscale x 1 x half> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmfne_mask_vv_nxv1f16_nxv1f16:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e16, mf4, ta, mu
+; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmfne.vv v0, v8, v9
 ; CHECK-NEXT:    vmfne.vv v11, v9, v10, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v11
@@ -85,8 +85,8 @@ declare <vscale x 2 x i1> @llvm.riscv.vmfne.mask.nxv2f16(
 define <vscale x 2 x i1> @intrinsic_vmfne_mask_vv_nxv2f16_nxv2f16(<vscale x 2 x i1> %0, <vscale x 2 x half> %1, <vscale x 2 x half> %2, <vscale x 2 x half> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmfne_mask_vv_nxv2f16_nxv2f16:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e16, mf2, ta, mu
+; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmfne.vv v0, v8, v9
 ; CHECK-NEXT:    vmfne.vv v11, v9, v10, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v11
@@ -136,8 +136,8 @@ declare <vscale x 4 x i1> @llvm.riscv.vmfne.mask.nxv4f16(
 define <vscale x 4 x i1> @intrinsic_vmfne_mask_vv_nxv4f16_nxv4f16(<vscale x 4 x i1> %0, <vscale x 4 x half> %1, <vscale x 4 x half> %2, <vscale x 4 x half> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmfne_mask_vv_nxv4f16_nxv4f16:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e16, m1, ta, mu
+; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmfne.vv v0, v8, v9
 ; CHECK-NEXT:    vmfne.vv v11, v9, v10, v0.t
 ; CHECK-NEXT:    vmv.v.v v0, v11
@@ -187,8 +187,8 @@ declare <vscale x 8 x i1> @llvm.riscv.vmfne.mask.nxv8f16(
 define <vscale x 8 x i1> @intrinsic_vmfne_mask_vv_nxv8f16_nxv8f16(<vscale x 8 x i1> %0, <vscale x 8 x half> %1, <vscale x 8 x half> %2, <vscale x 8 x half> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmfne_mask_vv_nxv8f16_nxv8f16:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v14, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e16, m2, ta, mu
+; CHECK-NEXT:    vmv1r.v v14, v0
 ; CHECK-NEXT:    vmfne.vv v0, v8, v10
 ; CHECK-NEXT:    vmfne.vv v14, v10, v12, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v14
@@ -238,8 +238,8 @@ declare <vscale x 16 x i1> @llvm.riscv.vmfne.mask.nxv16f16(
 define <vscale x 16 x i1> @intrinsic_vmfne_mask_vv_nxv16f16_nxv16f16(<vscale x 16 x i1> %0, <vscale x 16 x half> %1, <vscale x 16 x half> %2, <vscale x 16 x half> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmfne_mask_vv_nxv16f16_nxv16f16:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v20, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e16, m4, ta, mu
+; CHECK-NEXT:    vmv1r.v v20, v0
 ; CHECK-NEXT:    vmfne.vv v0, v8, v12
 ; CHECK-NEXT:    vmfne.vv v20, v12, v16, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v20
@@ -289,8 +289,8 @@ declare <vscale x 1 x i1> @llvm.riscv.vmfne.mask.nxv1f32(
 define <vscale x 1 x i1> @intrinsic_vmfne_mask_vv_nxv1f32_nxv1f32(<vscale x 1 x i1> %0, <vscale x 1 x float> %1, <vscale x 1 x float> %2, <vscale x 1 x float> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmfne_mask_vv_nxv1f32_nxv1f32:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e32, mf2, ta, mu
+; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmfne.vv v0, v8, v9
 ; CHECK-NEXT:    vmfne.vv v11, v9, v10, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v11
@@ -340,8 +340,8 @@ declare <vscale x 2 x i1> @llvm.riscv.vmfne.mask.nxv2f32(
 define <vscale x 2 x i1> @intrinsic_vmfne_mask_vv_nxv2f32_nxv2f32(<vscale x 2 x i1> %0, <vscale x 2 x float> %1, <vscale x 2 x float> %2, <vscale x 2 x float> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmfne_mask_vv_nxv2f32_nxv2f32:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e32, m1, ta, mu
+; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmfne.vv v0, v8, v9
 ; CHECK-NEXT:    vmfne.vv v11, v9, v10, v0.t
 ; CHECK-NEXT:    vmv.v.v v0, v11
@@ -391,8 +391,8 @@ declare <vscale x 4 x i1> @llvm.riscv.vmfne.mask.nxv4f32(
 define <vscale x 4 x i1> @intrinsic_vmfne_mask_vv_nxv4f32_nxv4f32(<vscale x 4 x i1> %0, <vscale x 4 x float> %1, <vscale x 4 x float> %2, <vscale x 4 x float> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmfne_mask_vv_nxv4f32_nxv4f32:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v14, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e32, m2, ta, mu
+; CHECK-NEXT:    vmv1r.v v14, v0
 ; CHECK-NEXT:    vmfne.vv v0, v8, v10
 ; CHECK-NEXT:    vmfne.vv v14, v10, v12, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v14
@@ -442,8 +442,8 @@ declare <vscale x 8 x i1> @llvm.riscv.vmfne.mask.nxv8f32(
 define <vscale x 8 x i1> @intrinsic_vmfne_mask_vv_nxv8f32_nxv8f32(<vscale x 8 x i1> %0, <vscale x 8 x float> %1, <vscale x 8 x float> %2, <vscale x 8 x float> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmfne_mask_vv_nxv8f32_nxv8f32:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v20, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e32, m4, ta, mu
+; CHECK-NEXT:    vmv1r.v v20, v0
 ; CHECK-NEXT:    vmfne.vv v0, v8, v12
 ; CHECK-NEXT:    vmfne.vv v20, v12, v16, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v20
@@ -493,8 +493,8 @@ declare <vscale x 1 x i1> @llvm.riscv.vmfne.mask.nxv1f64(
 define <vscale x 1 x i1> @intrinsic_vmfne_mask_vv_nxv1f64_nxv1f64(<vscale x 1 x i1> %0, <vscale x 1 x double> %1, <vscale x 1 x double> %2, <vscale x 1 x double> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmfne_mask_vv_nxv1f64_nxv1f64:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e64, m1, ta, mu
+; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmfne.vv v0, v8, v9
 ; CHECK-NEXT:    vmfne.vv v11, v9, v10, v0.t
 ; CHECK-NEXT:    vmv.v.v v0, v11
@@ -544,8 +544,8 @@ declare <vscale x 2 x i1> @llvm.riscv.vmfne.mask.nxv2f64(
 define <vscale x 2 x i1> @intrinsic_vmfne_mask_vv_nxv2f64_nxv2f64(<vscale x 2 x i1> %0, <vscale x 2 x double> %1, <vscale x 2 x double> %2, <vscale x 2 x double> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmfne_mask_vv_nxv2f64_nxv2f64:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v14, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e64, m2, ta, mu
+; CHECK-NEXT:    vmv1r.v v14, v0
 ; CHECK-NEXT:    vmfne.vv v0, v8, v10
 ; CHECK-NEXT:    vmfne.vv v14, v10, v12, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v14
@@ -595,8 +595,8 @@ declare <vscale x 4 x i1> @llvm.riscv.vmfne.mask.nxv4f64(
 define <vscale x 4 x i1> @intrinsic_vmfne_mask_vv_nxv4f64_nxv4f64(<vscale x 4 x i1> %0, <vscale x 4 x double> %1, <vscale x 4 x double> %2, <vscale x 4 x double> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmfne_mask_vv_nxv4f64_nxv4f64:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v20, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e64, m4, ta, mu
+; CHECK-NEXT:    vmv1r.v v20, v0
 ; CHECK-NEXT:    vmfne.vv v0, v8, v12
 ; CHECK-NEXT:    vmfne.vv v20, v12, v16, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v20
@@ -646,9 +646,9 @@ declare <vscale x 1 x i1> @llvm.riscv.vmfne.mask.nxv1f16.f16(
 define <vscale x 1 x i1> @intrinsic_vmfne_mask_vf_nxv1f16_f16(<vscale x 1 x i1> %0, <vscale x 1 x half> %1, half %2, <vscale x 1 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmfne_mask_vf_nxv1f16_f16:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e16, mf4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a0, e16, mf4, ta, mu
 ; CHECK-NEXT:    vmfne.vf v10, v8, fa0, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v10
 ; CHECK-NEXT:    ret
@@ -693,9 +693,9 @@ declare <vscale x 2 x i1> @llvm.riscv.vmfne.mask.nxv2f16.f16(
 define <vscale x 2 x i1> @intrinsic_vmfne_mask_vf_nxv2f16_f16(<vscale x 2 x i1> %0, <vscale x 2 x half> %1, half %2, <vscale x 2 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmfne_mask_vf_nxv2f16_f16:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e16, mf2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a0, e16, mf2, ta, mu
 ; CHECK-NEXT:    vmfne.vf v10, v8, fa0, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v10
 ; CHECK-NEXT:    ret
@@ -740,9 +740,9 @@ declare <vscale x 4 x i1> @llvm.riscv.vmfne.mask.nxv4f16.f16(
 define <vscale x 4 x i1> @intrinsic_vmfne_mask_vf_nxv4f16_f16(<vscale x 4 x i1> %0, <vscale x 4 x half> %1, half %2, <vscale x 4 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmfne_mask_vf_nxv4f16_f16:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e16, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a0, e16, m1, ta, mu
 ; CHECK-NEXT:    vmfne.vf v10, v8, fa0, v0.t
 ; CHECK-NEXT:    vmv.v.v v0, v10
 ; CHECK-NEXT:    ret
@@ -787,9 +787,9 @@ declare <vscale x 8 x i1> @llvm.riscv.vmfne.mask.nxv8f16.f16(
 define <vscale x 8 x i1> @intrinsic_vmfne_mask_vf_nxv8f16_f16(<vscale x 8 x i1> %0, <vscale x 8 x half> %1, half %2, <vscale x 8 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmfne_mask_vf_nxv8f16_f16:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e16, m2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmv1r.v v0, v10
-; CHECK-NEXT:    vsetvli zero, a0, e16, m2, ta, mu
 ; CHECK-NEXT:    vmfne.vf v11, v8, fa0, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v11
 ; CHECK-NEXT:    ret
@@ -834,9 +834,9 @@ declare <vscale x 16 x i1> @llvm.riscv.vmfne.mask.nxv16f16.f16(
 define <vscale x 16 x i1> @intrinsic_vmfne_mask_vf_nxv16f16_f16(<vscale x 16 x i1> %0, <vscale x 16 x half> %1, half %2, <vscale x 16 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmfne_mask_vf_nxv16f16_f16:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e16, m4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v13, v0
 ; CHECK-NEXT:    vmv1r.v v0, v12
-; CHECK-NEXT:    vsetvli zero, a0, e16, m4, ta, mu
 ; CHECK-NEXT:    vmfne.vf v13, v8, fa0, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v13
 ; CHECK-NEXT:    ret
@@ -881,9 +881,9 @@ declare <vscale x 1 x i1> @llvm.riscv.vmfne.mask.nxv1f32.f32(
 define <vscale x 1 x i1> @intrinsic_vmfne_mask_vf_nxv1f32_f32(<vscale x 1 x i1> %0, <vscale x 1 x float> %1, float %2, <vscale x 1 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmfne_mask_vf_nxv1f32_f32:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e32, mf2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a0, e32, mf2, ta, mu
 ; CHECK-NEXT:    vmfne.vf v10, v8, fa0, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v10
 ; CHECK-NEXT:    ret
@@ -928,9 +928,9 @@ declare <vscale x 2 x i1> @llvm.riscv.vmfne.mask.nxv2f32.f32(
 define <vscale x 2 x i1> @intrinsic_vmfne_mask_vf_nxv2f32_f32(<vscale x 2 x i1> %0, <vscale x 2 x float> %1, float %2, <vscale x 2 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmfne_mask_vf_nxv2f32_f32:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e32, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a0, e32, m1, ta, mu
 ; CHECK-NEXT:    vmfne.vf v10, v8, fa0, v0.t
 ; CHECK-NEXT:    vmv.v.v v0, v10
 ; CHECK-NEXT:    ret
@@ -975,9 +975,9 @@ declare <vscale x 4 x i1> @llvm.riscv.vmfne.mask.nxv4f32.f32(
 define <vscale x 4 x i1> @intrinsic_vmfne_mask_vf_nxv4f32_f32(<vscale x 4 x i1> %0, <vscale x 4 x float> %1, float %2, <vscale x 4 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmfne_mask_vf_nxv4f32_f32:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e32, m2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmv1r.v v0, v10
-; CHECK-NEXT:    vsetvli zero, a0, e32, m2, ta, mu
 ; CHECK-NEXT:    vmfne.vf v11, v8, fa0, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v11
 ; CHECK-NEXT:    ret
@@ -1022,9 +1022,9 @@ declare <vscale x 8 x i1> @llvm.riscv.vmfne.mask.nxv8f32.f32(
 define <vscale x 8 x i1> @intrinsic_vmfne_mask_vf_nxv8f32_f32(<vscale x 8 x i1> %0, <vscale x 8 x float> %1, float %2, <vscale x 8 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmfne_mask_vf_nxv8f32_f32:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e32, m4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v13, v0
 ; CHECK-NEXT:    vmv1r.v v0, v12
-; CHECK-NEXT:    vsetvli zero, a0, e32, m4, ta, mu
 ; CHECK-NEXT:    vmfne.vf v13, v8, fa0, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v13
 ; CHECK-NEXT:    ret
@@ -1069,9 +1069,9 @@ declare <vscale x 1 x i1> @llvm.riscv.vmfne.mask.nxv1f64.f64(
 define <vscale x 1 x i1> @intrinsic_vmfne_mask_vf_nxv1f64_f64(<vscale x 1 x i1> %0, <vscale x 1 x double> %1, double %2, <vscale x 1 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmfne_mask_vf_nxv1f64_f64:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e64, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a0, e64, m1, ta, mu
 ; CHECK-NEXT:    vmfne.vf v10, v8, fa0, v0.t
 ; CHECK-NEXT:    vmv.v.v v0, v10
 ; CHECK-NEXT:    ret
@@ -1116,9 +1116,9 @@ declare <vscale x 2 x i1> @llvm.riscv.vmfne.mask.nxv2f64.f64(
 define <vscale x 2 x i1> @intrinsic_vmfne_mask_vf_nxv2f64_f64(<vscale x 2 x i1> %0, <vscale x 2 x double> %1, double %2, <vscale x 2 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmfne_mask_vf_nxv2f64_f64:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e64, m2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmv1r.v v0, v10
-; CHECK-NEXT:    vsetvli zero, a0, e64, m2, ta, mu
 ; CHECK-NEXT:    vmfne.vf v11, v8, fa0, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v11
 ; CHECK-NEXT:    ret
@@ -1163,9 +1163,9 @@ declare <vscale x 4 x i1> @llvm.riscv.vmfne.mask.nxv4f64.f64(
 define <vscale x 4 x i1> @intrinsic_vmfne_mask_vf_nxv4f64_f64(<vscale x 4 x i1> %0, <vscale x 4 x double> %1, double %2, <vscale x 4 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmfne_mask_vf_nxv4f64_f64:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e64, m4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v13, v0
 ; CHECK-NEXT:    vmv1r.v v0, v12
-; CHECK-NEXT:    vsetvli zero, a0, e64, m4, ta, mu
 ; CHECK-NEXT:    vmfne.vf v13, v8, fa0, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v13
 ; CHECK-NEXT:    ret
diff --git a/llvm/test/CodeGen/RISCV/rvv/vmin-vp.ll b/llvm/test/CodeGen/RISCV/rvv/vmin-vp.ll
index 8690014cc2c9d..3441934fb1550 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vmin-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vmin-vp.ll
@@ -412,8 +412,8 @@ declare <vscale x 128 x i8> @llvm.vp.smin.nxv128i8(<vscale x 128 x i8>, <vscale
 define <vscale x 128 x i8> @vmin_vx_nxv128i8(<vscale x 128 x i8> %va, i8 %b, <vscale x 128 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vmin_vx_nxv128i8:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v24, v0
 ; CHECK-NEXT:    vsetvli a3, zero, e8, m8, ta, ma
+; CHECK-NEXT:    vmv1r.v v24, v0
 ; CHECK-NEXT:    vlm.v v0, (a1)
 ; CHECK-NEXT:    csrr a1, vlenb
 ; CHECK-NEXT:    slli a1, a1, 3
@@ -974,11 +974,11 @@ declare <vscale x 32 x i32> @llvm.vp.smin.nxv32i32(<vscale x 32 x i32>, <vscale
 define <vscale x 32 x i32> @vmin_vx_nxv32i32(<vscale x 32 x i32> %va, i32 %b, <vscale x 32 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vmin_vx_nxv32i32:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli a2, zero, e8, mf2, ta, ma
 ; CHECK-NEXT:    vmv1r.v v24, v0
 ; CHECK-NEXT:    csrr a2, vlenb
 ; CHECK-NEXT:    srli a3, a2, 2
 ; CHECK-NEXT:    slli a2, a2, 1
-; CHECK-NEXT:    vsetvli a4, zero, e8, mf2, ta, ma
 ; CHECK-NEXT:    vslidedown.vx v0, v0, a3
 ; CHECK-NEXT:    sub a3, a1, a2
 ; CHECK-NEXT:    sltu a4, a1, a3
@@ -1034,11 +1034,11 @@ declare i32 @llvm.vscale.i32()
 define <vscale x 32 x i32> @vmin_vx_nxv32i32_evl_nx8(<vscale x 32 x i32> %va, i32 %b, <vscale x 32 x i1> %m) {
 ; CHECK-LABEL: vmin_vx_nxv32i32_evl_nx8:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli a1, zero, e8, mf2, ta, ma
 ; CHECK-NEXT:    vmv1r.v v24, v0
 ; CHECK-NEXT:    csrr a1, vlenb
 ; CHECK-NEXT:    srli a3, a1, 2
 ; CHECK-NEXT:    slli a2, a1, 1
-; CHECK-NEXT:    vsetvli a4, zero, e8, mf2, ta, ma
 ; CHECK-NEXT:    vslidedown.vx v0, v0, a3
 ; CHECK-NEXT:    sub a3, a1, a2
 ; CHECK-NEXT:    sltu a4, a1, a3
diff --git a/llvm/test/CodeGen/RISCV/rvv/vminu-vp.ll b/llvm/test/CodeGen/RISCV/rvv/vminu-vp.ll
index 414807829d563..bbf4c886bfe9f 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vminu-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vminu-vp.ll
@@ -410,8 +410,8 @@ declare <vscale x 128 x i8> @llvm.vp.umin.nxv128i8(<vscale x 128 x i8>, <vscale
 define <vscale x 128 x i8> @vminu_vx_nxv128i8(<vscale x 128 x i8> %va, i8 %b, <vscale x 128 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vminu_vx_nxv128i8:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v24, v0
 ; CHECK-NEXT:    vsetvli a3, zero, e8, m8, ta, ma
+; CHECK-NEXT:    vmv1r.v v24, v0
 ; CHECK-NEXT:    vlm.v v0, (a1)
 ; CHECK-NEXT:    csrr a1, vlenb
 ; CHECK-NEXT:    slli a1, a1, 3
@@ -972,11 +972,11 @@ declare <vscale x 32 x i32> @llvm.vp.umin.nxv32i32(<vscale x 32 x i32>, <vscale
 define <vscale x 32 x i32> @vminu_vx_nxv32i32(<vscale x 32 x i32> %va, i32 %b, <vscale x 32 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vminu_vx_nxv32i32:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli a2, zero, e8, mf2, ta, ma
 ; CHECK-NEXT:    vmv1r.v v24, v0
 ; CHECK-NEXT:    csrr a2, vlenb
 ; CHECK-NEXT:    srli a3, a2, 2
 ; CHECK-NEXT:    slli a2, a2, 1
-; CHECK-NEXT:    vsetvli a4, zero, e8, mf2, ta, ma
 ; CHECK-NEXT:    vslidedown.vx v0, v0, a3
 ; CHECK-NEXT:    sub a3, a1, a2
 ; CHECK-NEXT:    sltu a4, a1, a3
@@ -1032,11 +1032,11 @@ declare i32 @llvm.vscale.i32()
 define <vscale x 32 x i32> @vminu_vx_nxv32i32_evl_nx8(<vscale x 32 x i32> %va, i32 %b, <vscale x 32 x i1> %m) {
 ; CHECK-LABEL: vminu_vx_nxv32i32_evl_nx8:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli a1, zero, e8, mf2, ta, ma
 ; CHECK-NEXT:    vmv1r.v v24, v0
 ; CHECK-NEXT:    csrr a1, vlenb
 ; CHECK-NEXT:    srli a3, a1, 2
 ; CHECK-NEXT:    slli a2, a1, 1
-; CHECK-NEXT:    vsetvli a4, zero, e8, mf2, ta, ma
 ; CHECK-NEXT:    vslidedown.vx v0, v0, a3
 ; CHECK-NEXT:    sub a3, a1, a2
 ; CHECK-NEXT:    sltu a4, a1, a3
diff --git a/llvm/test/CodeGen/RISCV/rvv/vmsbf.ll b/llvm/test/CodeGen/RISCV/rvv/vmsbf.ll
index d1f344d52763d..80e74faa8cd91 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vmsbf.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vmsbf.ll
@@ -31,9 +31,9 @@ declare <vscale x 1 x i1> @llvm.riscv.vmsbf.mask.nxv1i1(
 define <vscale x 1 x i1> @intrinsic_vmsbf_mask_m_nxv1i1_nxv1i1(<vscale x 1 x i1> %0, <vscale x 1 x i1> %1, <vscale x 1 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsbf_mask_m_nxv1i1_nxv1i1:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e8, mf8, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a0, e8, mf8, ta, mu
 ; CHECK-NEXT:    vmsbf.m v10, v8, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v10
 ; CHECK-NEXT:    ret
@@ -73,9 +73,9 @@ declare <vscale x 2 x i1> @llvm.riscv.vmsbf.mask.nxv2i1(
 define <vscale x 2 x i1> @intrinsic_vmsbf_mask_m_nxv2i1_nxv2i1(<vscale x 2 x i1> %0, <vscale x 2 x i1> %1, <vscale x 2 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsbf_mask_m_nxv2i1_nxv2i1:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e8, mf4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a0, e8, mf4, ta, mu
 ; CHECK-NEXT:    vmsbf.m v10, v8, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v10
 ; CHECK-NEXT:    ret
@@ -115,9 +115,9 @@ declare <vscale x 4 x i1> @llvm.riscv.vmsbf.mask.nxv4i1(
 define <vscale x 4 x i1> @intrinsic_vmsbf_mask_m_nxv4i1_nxv4i1(<vscale x 4 x i1> %0, <vscale x 4 x i1> %1, <vscale x 4 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsbf_mask_m_nxv4i1_nxv4i1:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e8, mf2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a0, e8, mf2, ta, mu
 ; CHECK-NEXT:    vmsbf.m v10, v8, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v10
 ; CHECK-NEXT:    ret
@@ -157,9 +157,9 @@ declare <vscale x 8 x i1> @llvm.riscv.vmsbf.mask.nxv8i1(
 define <vscale x 8 x i1> @intrinsic_vmsbf_mask_m_nxv8i1_nxv8i1(<vscale x 8 x i1> %0, <vscale x 8 x i1> %1, <vscale x 8 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsbf_mask_m_nxv8i1_nxv8i1:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e8, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a0, e8, m1, ta, mu
 ; CHECK-NEXT:    vmsbf.m v10, v8, v0.t
 ; CHECK-NEXT:    vmv.v.v v0, v10
 ; CHECK-NEXT:    ret
@@ -199,9 +199,9 @@ declare <vscale x 16 x i1> @llvm.riscv.vmsbf.mask.nxv16i1(
 define <vscale x 16 x i1> @intrinsic_vmsbf_mask_m_nxv16i1_nxv16i1(<vscale x 16 x i1> %0, <vscale x 16 x i1> %1, <vscale x 16 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsbf_mask_m_nxv16i1_nxv16i1:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e8, m2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a0, e8, m2, ta, mu
 ; CHECK-NEXT:    vmsbf.m v10, v8, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v10
 ; CHECK-NEXT:    ret
@@ -241,9 +241,9 @@ declare <vscale x 32 x i1> @llvm.riscv.vmsbf.mask.nxv32i1(
 define <vscale x 32 x i1> @intrinsic_vmsbf_mask_m_nxv32i1_nxv32i1(<vscale x 32 x i1> %0, <vscale x 32 x i1> %1, <vscale x 32 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsbf_mask_m_nxv32i1_nxv32i1:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e8, m4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a0, e8, m4, ta, mu
 ; CHECK-NEXT:    vmsbf.m v10, v8, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v10
 ; CHECK-NEXT:    ret
@@ -283,9 +283,9 @@ declare <vscale x 64 x i1> @llvm.riscv.vmsbf.mask.nxv64i1(
 define <vscale x 64 x i1> @intrinsic_vmsbf_mask_m_nxv64i1_nxv64i1(<vscale x 64 x i1> %0, <vscale x 64 x i1> %1, <vscale x 64 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsbf_mask_m_nxv64i1_nxv64i1:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e8, m8, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a0, e8, m8, ta, mu
 ; CHECK-NEXT:    vmsbf.m v10, v8, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v10
 ; CHECK-NEXT:    ret
diff --git a/llvm/test/CodeGen/RISCV/rvv/vmseq.ll b/llvm/test/CodeGen/RISCV/rvv/vmseq.ll
index 1fd2383c40d18..6407f39a65e8b 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vmseq.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vmseq.ll
@@ -34,8 +34,8 @@ declare <vscale x 1 x i1> @llvm.riscv.vmseq.mask.nxv1i8(
 define <vscale x 1 x i1> @intrinsic_vmseq_mask_vv_nxv1i8_nxv1i8(<vscale x 1 x i1> %0, <vscale x 1 x i8> %1, <vscale x 1 x i8> %2, <vscale x 1 x i8> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmseq_mask_vv_nxv1i8_nxv1i8:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e8, mf8, ta, mu
+; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmseq.vv v0, v8, v9
 ; CHECK-NEXT:    vmseq.vv v11, v9, v10, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v11
@@ -85,8 +85,8 @@ declare <vscale x 2 x i1> @llvm.riscv.vmseq.mask.nxv2i8(
 define <vscale x 2 x i1> @intrinsic_vmseq_mask_vv_nxv2i8_nxv2i8(<vscale x 2 x i1> %0, <vscale x 2 x i8> %1, <vscale x 2 x i8> %2, <vscale x 2 x i8> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmseq_mask_vv_nxv2i8_nxv2i8:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e8, mf4, ta, mu
+; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmseq.vv v0, v8, v9
 ; CHECK-NEXT:    vmseq.vv v11, v9, v10, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v11
@@ -136,8 +136,8 @@ declare <vscale x 4 x i1> @llvm.riscv.vmseq.mask.nxv4i8(
 define <vscale x 4 x i1> @intrinsic_vmseq_mask_vv_nxv4i8_nxv4i8(<vscale x 4 x i1> %0, <vscale x 4 x i8> %1, <vscale x 4 x i8> %2, <vscale x 4 x i8> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmseq_mask_vv_nxv4i8_nxv4i8:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e8, mf2, ta, mu
+; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmseq.vv v0, v8, v9
 ; CHECK-NEXT:    vmseq.vv v11, v9, v10, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v11
@@ -187,8 +187,8 @@ declare <vscale x 8 x i1> @llvm.riscv.vmseq.mask.nxv8i8(
 define <vscale x 8 x i1> @intrinsic_vmseq_mask_vv_nxv8i8_nxv8i8(<vscale x 8 x i1> %0, <vscale x 8 x i8> %1, <vscale x 8 x i8> %2, <vscale x 8 x i8> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmseq_mask_vv_nxv8i8_nxv8i8:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e8, m1, ta, mu
+; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmseq.vv v0, v8, v9
 ; CHECK-NEXT:    vmseq.vv v11, v9, v10, v0.t
 ; CHECK-NEXT:    vmv.v.v v0, v11
@@ -238,8 +238,8 @@ declare <vscale x 16 x i1> @llvm.riscv.vmseq.mask.nxv16i8(
 define <vscale x 16 x i1> @intrinsic_vmseq_mask_vv_nxv16i8_nxv16i8(<vscale x 16 x i1> %0, <vscale x 16 x i8> %1, <vscale x 16 x i8> %2, <vscale x 16 x i8> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmseq_mask_vv_nxv16i8_nxv16i8:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v14, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e8, m2, ta, mu
+; CHECK-NEXT:    vmv1r.v v14, v0
 ; CHECK-NEXT:    vmseq.vv v0, v8, v10
 ; CHECK-NEXT:    vmseq.vv v14, v10, v12, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v14
@@ -289,8 +289,8 @@ declare <vscale x 32 x i1> @llvm.riscv.vmseq.mask.nxv32i8(
 define <vscale x 32 x i1> @intrinsic_vmseq_mask_vv_nxv32i8_nxv32i8(<vscale x 32 x i1> %0, <vscale x 32 x i8> %1, <vscale x 32 x i8> %2, <vscale x 32 x i8> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmseq_mask_vv_nxv32i8_nxv32i8:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v20, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e8, m4, ta, mu
+; CHECK-NEXT:    vmv1r.v v20, v0
 ; CHECK-NEXT:    vmseq.vv v0, v8, v12
 ; CHECK-NEXT:    vmseq.vv v20, v12, v16, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v20
@@ -340,8 +340,8 @@ declare <vscale x 1 x i1> @llvm.riscv.vmseq.mask.nxv1i16(
 define <vscale x 1 x i1> @intrinsic_vmseq_mask_vv_nxv1i16_nxv1i16(<vscale x 1 x i1> %0, <vscale x 1 x i16> %1, <vscale x 1 x i16> %2, <vscale x 1 x i16> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmseq_mask_vv_nxv1i16_nxv1i16:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e16, mf4, ta, mu
+; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmseq.vv v0, v8, v9
 ; CHECK-NEXT:    vmseq.vv v11, v9, v10, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v11
@@ -391,8 +391,8 @@ declare <vscale x 2 x i1> @llvm.riscv.vmseq.mask.nxv2i16(
 define <vscale x 2 x i1> @intrinsic_vmseq_mask_vv_nxv2i16_nxv2i16(<vscale x 2 x i1> %0, <vscale x 2 x i16> %1, <vscale x 2 x i16> %2, <vscale x 2 x i16> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmseq_mask_vv_nxv2i16_nxv2i16:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e16, mf2, ta, mu
+; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmseq.vv v0, v8, v9
 ; CHECK-NEXT:    vmseq.vv v11, v9, v10, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v11
@@ -442,8 +442,8 @@ declare <vscale x 4 x i1> @llvm.riscv.vmseq.mask.nxv4i16(
 define <vscale x 4 x i1> @intrinsic_vmseq_mask_vv_nxv4i16_nxv4i16(<vscale x 4 x i1> %0, <vscale x 4 x i16> %1, <vscale x 4 x i16> %2, <vscale x 4 x i16> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmseq_mask_vv_nxv4i16_nxv4i16:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e16, m1, ta, mu
+; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmseq.vv v0, v8, v9
 ; CHECK-NEXT:    vmseq.vv v11, v9, v10, v0.t
 ; CHECK-NEXT:    vmv.v.v v0, v11
@@ -493,8 +493,8 @@ declare <vscale x 8 x i1> @llvm.riscv.vmseq.mask.nxv8i16(
 define <vscale x 8 x i1> @intrinsic_vmseq_mask_vv_nxv8i16_nxv8i16(<vscale x 8 x i1> %0, <vscale x 8 x i16> %1, <vscale x 8 x i16> %2, <vscale x 8 x i16> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmseq_mask_vv_nxv8i16_nxv8i16:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v14, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e16, m2, ta, mu
+; CHECK-NEXT:    vmv1r.v v14, v0
 ; CHECK-NEXT:    vmseq.vv v0, v8, v10
 ; CHECK-NEXT:    vmseq.vv v14, v10, v12, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v14
@@ -544,8 +544,8 @@ declare <vscale x 16 x i1> @llvm.riscv.vmseq.mask.nxv16i16(
 define <vscale x 16 x i1> @intrinsic_vmseq_mask_vv_nxv16i16_nxv16i16(<vscale x 16 x i1> %0, <vscale x 16 x i16> %1, <vscale x 16 x i16> %2, <vscale x 16 x i16> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmseq_mask_vv_nxv16i16_nxv16i16:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v20, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e16, m4, ta, mu
+; CHECK-NEXT:    vmv1r.v v20, v0
 ; CHECK-NEXT:    vmseq.vv v0, v8, v12
 ; CHECK-NEXT:    vmseq.vv v20, v12, v16, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v20
@@ -595,8 +595,8 @@ declare <vscale x 1 x i1> @llvm.riscv.vmseq.mask.nxv1i32(
 define <vscale x 1 x i1> @intrinsic_vmseq_mask_vv_nxv1i32_nxv1i32(<vscale x 1 x i1> %0, <vscale x 1 x i32> %1, <vscale x 1 x i32> %2, <vscale x 1 x i32> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmseq_mask_vv_nxv1i32_nxv1i32:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e32, mf2, ta, mu
+; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmseq.vv v0, v8, v9
 ; CHECK-NEXT:    vmseq.vv v11, v9, v10, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v11
@@ -646,8 +646,8 @@ declare <vscale x 2 x i1> @llvm.riscv.vmseq.mask.nxv2i32(
 define <vscale x 2 x i1> @intrinsic_vmseq_mask_vv_nxv2i32_nxv2i32(<vscale x 2 x i1> %0, <vscale x 2 x i32> %1, <vscale x 2 x i32> %2, <vscale x 2 x i32> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmseq_mask_vv_nxv2i32_nxv2i32:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e32, m1, ta, mu
+; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmseq.vv v0, v8, v9
 ; CHECK-NEXT:    vmseq.vv v11, v9, v10, v0.t
 ; CHECK-NEXT:    vmv.v.v v0, v11
@@ -697,8 +697,8 @@ declare <vscale x 4 x i1> @llvm.riscv.vmseq.mask.nxv4i32(
 define <vscale x 4 x i1> @intrinsic_vmseq_mask_vv_nxv4i32_nxv4i32(<vscale x 4 x i1> %0, <vscale x 4 x i32> %1, <vscale x 4 x i32> %2, <vscale x 4 x i32> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmseq_mask_vv_nxv4i32_nxv4i32:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v14, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e32, m2, ta, mu
+; CHECK-NEXT:    vmv1r.v v14, v0
 ; CHECK-NEXT:    vmseq.vv v0, v8, v10
 ; CHECK-NEXT:    vmseq.vv v14, v10, v12, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v14
@@ -748,8 +748,8 @@ declare <vscale x 8 x i1> @llvm.riscv.vmseq.mask.nxv8i32(
 define <vscale x 8 x i1> @intrinsic_vmseq_mask_vv_nxv8i32_nxv8i32(<vscale x 8 x i1> %0, <vscale x 8 x i32> %1, <vscale x 8 x i32> %2, <vscale x 8 x i32> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmseq_mask_vv_nxv8i32_nxv8i32:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v20, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e32, m4, ta, mu
+; CHECK-NEXT:    vmv1r.v v20, v0
 ; CHECK-NEXT:    vmseq.vv v0, v8, v12
 ; CHECK-NEXT:    vmseq.vv v20, v12, v16, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v20
@@ -799,8 +799,8 @@ declare <vscale x 1 x i1> @llvm.riscv.vmseq.mask.nxv1i64(
 define <vscale x 1 x i1> @intrinsic_vmseq_mask_vv_nxv1i64_nxv1i64(<vscale x 1 x i1> %0, <vscale x 1 x i64> %1, <vscale x 1 x i64> %2, <vscale x 1 x i64> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmseq_mask_vv_nxv1i64_nxv1i64:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e64, m1, ta, mu
+; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmseq.vv v0, v8, v9
 ; CHECK-NEXT:    vmseq.vv v11, v9, v10, v0.t
 ; CHECK-NEXT:    vmv.v.v v0, v11
@@ -850,8 +850,8 @@ declare <vscale x 2 x i1> @llvm.riscv.vmseq.mask.nxv2i64(
 define <vscale x 2 x i1> @intrinsic_vmseq_mask_vv_nxv2i64_nxv2i64(<vscale x 2 x i1> %0, <vscale x 2 x i64> %1, <vscale x 2 x i64> %2, <vscale x 2 x i64> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmseq_mask_vv_nxv2i64_nxv2i64:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v14, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e64, m2, ta, mu
+; CHECK-NEXT:    vmv1r.v v14, v0
 ; CHECK-NEXT:    vmseq.vv v0, v8, v10
 ; CHECK-NEXT:    vmseq.vv v14, v10, v12, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v14
@@ -901,8 +901,8 @@ declare <vscale x 4 x i1> @llvm.riscv.vmseq.mask.nxv4i64(
 define <vscale x 4 x i1> @intrinsic_vmseq_mask_vv_nxv4i64_nxv4i64(<vscale x 4 x i1> %0, <vscale x 4 x i64> %1, <vscale x 4 x i64> %2, <vscale x 4 x i64> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmseq_mask_vv_nxv4i64_nxv4i64:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v20, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e64, m4, ta, mu
+; CHECK-NEXT:    vmv1r.v v20, v0
 ; CHECK-NEXT:    vmseq.vv v0, v8, v12
 ; CHECK-NEXT:    vmseq.vv v20, v12, v16, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v20
@@ -952,9 +952,9 @@ declare <vscale x 1 x i1> @llvm.riscv.vmseq.mask.nxv1i8.i8(
 define <vscale x 1 x i1> @intrinsic_vmseq_mask_vx_nxv1i8_i8(<vscale x 1 x i1> %0, <vscale x 1 x i8> %1, i8 %2, <vscale x 1 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmseq_mask_vx_nxv1i8_i8:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e8, mf8, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a1, e8, mf8, ta, mu
 ; CHECK-NEXT:    vmseq.vx v10, v8, a0, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v10
 ; CHECK-NEXT:    ret
@@ -999,9 +999,9 @@ declare <vscale x 2 x i1> @llvm.riscv.vmseq.mask.nxv2i8.i8(
 define <vscale x 2 x i1> @intrinsic_vmseq_mask_vx_nxv2i8_i8(<vscale x 2 x i1> %0, <vscale x 2 x i8> %1, i8 %2, <vscale x 2 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmseq_mask_vx_nxv2i8_i8:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e8, mf4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a1, e8, mf4, ta, mu
 ; CHECK-NEXT:    vmseq.vx v10, v8, a0, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v10
 ; CHECK-NEXT:    ret
@@ -1046,9 +1046,9 @@ declare <vscale x 4 x i1> @llvm.riscv.vmseq.mask.nxv4i8.i8(
 define <vscale x 4 x i1> @intrinsic_vmseq_mask_vx_nxv4i8_i8(<vscale x 4 x i1> %0, <vscale x 4 x i8> %1, i8 %2, <vscale x 4 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmseq_mask_vx_nxv4i8_i8:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e8, mf2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a1, e8, mf2, ta, mu
 ; CHECK-NEXT:    vmseq.vx v10, v8, a0, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v10
 ; CHECK-NEXT:    ret
@@ -1093,9 +1093,9 @@ declare <vscale x 8 x i1> @llvm.riscv.vmseq.mask.nxv8i8.i8(
 define <vscale x 8 x i1> @intrinsic_vmseq_mask_vx_nxv8i8_i8(<vscale x 8 x i1> %0, <vscale x 8 x i8> %1, i8 %2, <vscale x 8 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmseq_mask_vx_nxv8i8_i8:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e8, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a1, e8, m1, ta, mu
 ; CHECK-NEXT:    vmseq.vx v10, v8, a0, v0.t
 ; CHECK-NEXT:    vmv.v.v v0, v10
 ; CHECK-NEXT:    ret
@@ -1140,9 +1140,9 @@ declare <vscale x 16 x i1> @llvm.riscv.vmseq.mask.nxv16i8.i8(
 define <vscale x 16 x i1> @intrinsic_vmseq_mask_vx_nxv16i8_i8(<vscale x 16 x i1> %0, <vscale x 16 x i8> %1, i8 %2, <vscale x 16 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmseq_mask_vx_nxv16i8_i8:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e8, m2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmv1r.v v0, v10
-; CHECK-NEXT:    vsetvli zero, a1, e8, m2, ta, mu
 ; CHECK-NEXT:    vmseq.vx v11, v8, a0, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v11
 ; CHECK-NEXT:    ret
@@ -1187,9 +1187,9 @@ declare <vscale x 32 x i1> @llvm.riscv.vmseq.mask.nxv32i8.i8(
 define <vscale x 32 x i1> @intrinsic_vmseq_mask_vx_nxv32i8_i8(<vscale x 32 x i1> %0, <vscale x 32 x i8> %1, i8 %2, <vscale x 32 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmseq_mask_vx_nxv32i8_i8:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e8, m4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v13, v0
 ; CHECK-NEXT:    vmv1r.v v0, v12
-; CHECK-NEXT:    vsetvli zero, a1, e8, m4, ta, mu
 ; CHECK-NEXT:    vmseq.vx v13, v8, a0, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v13
 ; CHECK-NEXT:    ret
@@ -1234,9 +1234,9 @@ declare <vscale x 1 x i1> @llvm.riscv.vmseq.mask.nxv1i16.i16(
 define <vscale x 1 x i1> @intrinsic_vmseq_mask_vx_nxv1i16_i16(<vscale x 1 x i1> %0, <vscale x 1 x i16> %1, i16 %2, <vscale x 1 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmseq_mask_vx_nxv1i16_i16:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, mf4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a1, e16, mf4, ta, mu
 ; CHECK-NEXT:    vmseq.vx v10, v8, a0, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v10
 ; CHECK-NEXT:    ret
@@ -1281,9 +1281,9 @@ declare <vscale x 2 x i1> @llvm.riscv.vmseq.mask.nxv2i16.i16(
 define <vscale x 2 x i1> @intrinsic_vmseq_mask_vx_nxv2i16_i16(<vscale x 2 x i1> %0, <vscale x 2 x i16> %1, i16 %2, <vscale x 2 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmseq_mask_vx_nxv2i16_i16:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, mf2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a1, e16, mf2, ta, mu
 ; CHECK-NEXT:    vmseq.vx v10, v8, a0, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v10
 ; CHECK-NEXT:    ret
@@ -1328,9 +1328,9 @@ declare <vscale x 4 x i1> @llvm.riscv.vmseq.mask.nxv4i16.i16(
 define <vscale x 4 x i1> @intrinsic_vmseq_mask_vx_nxv4i16_i16(<vscale x 4 x i1> %0, <vscale x 4 x i16> %1, i16 %2, <vscale x 4 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmseq_mask_vx_nxv4i16_i16:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a1, e16, m1, ta, mu
 ; CHECK-NEXT:    vmseq.vx v10, v8, a0, v0.t
 ; CHECK-NEXT:    vmv.v.v v0, v10
 ; CHECK-NEXT:    ret
@@ -1375,9 +1375,9 @@ declare <vscale x 8 x i1> @llvm.riscv.vmseq.mask.nxv8i16.i16(
 define <vscale x 8 x i1> @intrinsic_vmseq_mask_vx_nxv8i16_i16(<vscale x 8 x i1> %0, <vscale x 8 x i16> %1, i16 %2, <vscale x 8 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmseq_mask_vx_nxv8i16_i16:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, m2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmv1r.v v0, v10
-; CHECK-NEXT:    vsetvli zero, a1, e16, m2, ta, mu
 ; CHECK-NEXT:    vmseq.vx v11, v8, a0, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v11
 ; CHECK-NEXT:    ret
@@ -1422,9 +1422,9 @@ declare <vscale x 16 x i1> @llvm.riscv.vmseq.mask.nxv16i16.i16(
 define <vscale x 16 x i1> @intrinsic_vmseq_mask_vx_nxv16i16_i16(<vscale x 16 x i1> %0, <vscale x 16 x i16> %1, i16 %2, <vscale x 16 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmseq_mask_vx_nxv16i16_i16:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, m4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v13, v0
 ; CHECK-NEXT:    vmv1r.v v0, v12
-; CHECK-NEXT:    vsetvli zero, a1, e16, m4, ta, mu
 ; CHECK-NEXT:    vmseq.vx v13, v8, a0, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v13
 ; CHECK-NEXT:    ret
@@ -1469,9 +1469,9 @@ declare <vscale x 1 x i1> @llvm.riscv.vmseq.mask.nxv1i32.i32(
 define <vscale x 1 x i1> @intrinsic_vmseq_mask_vx_nxv1i32_i32(<vscale x 1 x i1> %0, <vscale x 1 x i32> %1, i32 %2, <vscale x 1 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmseq_mask_vx_nxv1i32_i32:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e32, mf2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a1, e32, mf2, ta, mu
 ; CHECK-NEXT:    vmseq.vx v10, v8, a0, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v10
 ; CHECK-NEXT:    ret
@@ -1516,9 +1516,9 @@ declare <vscale x 2 x i1> @llvm.riscv.vmseq.mask.nxv2i32.i32(
 define <vscale x 2 x i1> @intrinsic_vmseq_mask_vx_nxv2i32_i32(<vscale x 2 x i1> %0, <vscale x 2 x i32> %1, i32 %2, <vscale x 2 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmseq_mask_vx_nxv2i32_i32:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e32, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a1, e32, m1, ta, mu
 ; CHECK-NEXT:    vmseq.vx v10, v8, a0, v0.t
 ; CHECK-NEXT:    vmv.v.v v0, v10
 ; CHECK-NEXT:    ret
@@ -1563,9 +1563,9 @@ declare <vscale x 4 x i1> @llvm.riscv.vmseq.mask.nxv4i32.i32(
 define <vscale x 4 x i1> @intrinsic_vmseq_mask_vx_nxv4i32_i32(<vscale x 4 x i1> %0, <vscale x 4 x i32> %1, i32 %2, <vscale x 4 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmseq_mask_vx_nxv4i32_i32:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e32, m2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmv1r.v v0, v10
-; CHECK-NEXT:    vsetvli zero, a1, e32, m2, ta, mu
 ; CHECK-NEXT:    vmseq.vx v11, v8, a0, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v11
 ; CHECK-NEXT:    ret
@@ -1610,9 +1610,9 @@ declare <vscale x 8 x i1> @llvm.riscv.vmseq.mask.nxv8i32.i32(
 define <vscale x 8 x i1> @intrinsic_vmseq_mask_vx_nxv8i32_i32(<vscale x 8 x i1> %0, <vscale x 8 x i32> %1, i32 %2, <vscale x 8 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmseq_mask_vx_nxv8i32_i32:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e32, m4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v13, v0
 ; CHECK-NEXT:    vmv1r.v v0, v12
-; CHECK-NEXT:    vsetvli zero, a1, e32, m4, ta, mu
 ; CHECK-NEXT:    vmseq.vx v13, v8, a0, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v13
 ; CHECK-NEXT:    ret
@@ -1684,9 +1684,9 @@ define <vscale x 1 x i1> @intrinsic_vmseq_mask_vx_nxv1i64_i64(<vscale x 1 x i1>
 ;
 ; RV64-LABEL: intrinsic_vmseq_mask_vx_nxv1i64_i64:
 ; RV64:       # %bb.0: # %entry
+; RV64-NEXT:    vsetvli zero, a1, e64, m1, ta, mu
 ; RV64-NEXT:    vmv1r.v v10, v0
 ; RV64-NEXT:    vmv1r.v v0, v9
-; RV64-NEXT:    vsetvli zero, a1, e64, m1, ta, mu
 ; RV64-NEXT:    vmseq.vx v10, v8, a0, v0.t
 ; RV64-NEXT:    vmv.v.v v0, v10
 ; RV64-NEXT:    ret
@@ -1758,9 +1758,9 @@ define <vscale x 2 x i1> @intrinsic_vmseq_mask_vx_nxv2i64_i64(<vscale x 2 x i1>
 ;
 ; RV64-LABEL: intrinsic_vmseq_mask_vx_nxv2i64_i64:
 ; RV64:       # %bb.0: # %entry
+; RV64-NEXT:    vsetvli zero, a1, e64, m2, ta, mu
 ; RV64-NEXT:    vmv1r.v v11, v0
 ; RV64-NEXT:    vmv1r.v v0, v10
-; RV64-NEXT:    vsetvli zero, a1, e64, m2, ta, mu
 ; RV64-NEXT:    vmseq.vx v11, v8, a0, v0.t
 ; RV64-NEXT:    vmv1r.v v0, v11
 ; RV64-NEXT:    ret
@@ -1832,9 +1832,9 @@ define <vscale x 4 x i1> @intrinsic_vmseq_mask_vx_nxv4i64_i64(<vscale x 4 x i1>
 ;
 ; RV64-LABEL: intrinsic_vmseq_mask_vx_nxv4i64_i64:
 ; RV64:       # %bb.0: # %entry
+; RV64-NEXT:    vsetvli zero, a1, e64, m4, ta, mu
 ; RV64-NEXT:    vmv1r.v v13, v0
 ; RV64-NEXT:    vmv1r.v v0, v12
-; RV64-NEXT:    vsetvli zero, a1, e64, m4, ta, mu
 ; RV64-NEXT:    vmseq.vx v13, v8, a0, v0.t
 ; RV64-NEXT:    vmv1r.v v0, v13
 ; RV64-NEXT:    ret
@@ -1867,9 +1867,9 @@ entry:
 define <vscale x 1 x i1> @intrinsic_vmseq_mask_vi_nxv1i8_i8(<vscale x 1 x i1> %0, <vscale x 1 x i8> %1, <vscale x 1 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmseq_mask_vi_nxv1i8_i8:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e8, mf8, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a0, e8, mf8, ta, mu
 ; CHECK-NEXT:    vmseq.vi v10, v8, 9, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v10
 ; CHECK-NEXT:    ret
@@ -1902,9 +1902,9 @@ entry:
 define <vscale x 2 x i1> @intrinsic_vmseq_mask_vi_nxv2i8_i8(<vscale x 2 x i1> %0, <vscale x 2 x i8> %1, <vscale x 2 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmseq_mask_vi_nxv2i8_i8:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e8, mf4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a0, e8, mf4, ta, mu
 ; CHECK-NEXT:    vmseq.vi v10, v8, 9, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v10
 ; CHECK-NEXT:    ret
@@ -1937,9 +1937,9 @@ entry:
 define <vscale x 4 x i1> @intrinsic_vmseq_mask_vi_nxv4i8_i8(<vscale x 4 x i1> %0, <vscale x 4 x i8> %1, <vscale x 4 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmseq_mask_vi_nxv4i8_i8:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e8, mf2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a0, e8, mf2, ta, mu
 ; CHECK-NEXT:    vmseq.vi v10, v8, 9, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v10
 ; CHECK-NEXT:    ret
@@ -1972,9 +1972,9 @@ entry:
 define <vscale x 8 x i1> @intrinsic_vmseq_mask_vi_nxv8i8_i8(<vscale x 8 x i1> %0, <vscale x 8 x i8> %1, <vscale x 8 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmseq_mask_vi_nxv8i8_i8:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e8, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a0, e8, m1, ta, mu
 ; CHECK-NEXT:    vmseq.vi v10, v8, 9, v0.t
 ; CHECK-NEXT:    vmv.v.v v0, v10
 ; CHECK-NEXT:    ret
@@ -2007,9 +2007,9 @@ entry:
 define <vscale x 16 x i1> @intrinsic_vmseq_mask_vi_nxv16i8_i8(<vscale x 16 x i1> %0, <vscale x 16 x i8> %1, <vscale x 16 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmseq_mask_vi_nxv16i8_i8:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e8, m2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmv1r.v v0, v10
-; CHECK-NEXT:    vsetvli zero, a0, e8, m2, ta, mu
 ; CHECK-NEXT:    vmseq.vi v11, v8, 9, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v11
 ; CHECK-NEXT:    ret
@@ -2042,9 +2042,9 @@ entry:
 define <vscale x 32 x i1> @intrinsic_vmseq_mask_vi_nxv32i8_i8(<vscale x 32 x i1> %0, <vscale x 32 x i8> %1, <vscale x 32 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmseq_mask_vi_nxv32i8_i8:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e8, m4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v13, v0
 ; CHECK-NEXT:    vmv1r.v v0, v12
-; CHECK-NEXT:    vsetvli zero, a0, e8, m4, ta, mu
 ; CHECK-NEXT:    vmseq.vi v13, v8, 9, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v13
 ; CHECK-NEXT:    ret
@@ -2077,9 +2077,9 @@ entry:
 define <vscale x 1 x i1> @intrinsic_vmseq_mask_vi_nxv1i16_i16(<vscale x 1 x i1> %0, <vscale x 1 x i16> %1, <vscale x 1 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmseq_mask_vi_nxv1i16_i16:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e16, mf4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a0, e16, mf4, ta, mu
 ; CHECK-NEXT:    vmseq.vi v10, v8, 9, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v10
 ; CHECK-NEXT:    ret
@@ -2112,9 +2112,9 @@ entry:
 define <vscale x 2 x i1> @intrinsic_vmseq_mask_vi_nxv2i16_i16(<vscale x 2 x i1> %0, <vscale x 2 x i16> %1, <vscale x 2 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmseq_mask_vi_nxv2i16_i16:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e16, mf2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a0, e16, mf2, ta, mu
 ; CHECK-NEXT:    vmseq.vi v10, v8, 9, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v10
 ; CHECK-NEXT:    ret
@@ -2147,9 +2147,9 @@ entry:
 define <vscale x 4 x i1> @intrinsic_vmseq_mask_vi_nxv4i16_i16(<vscale x 4 x i1> %0, <vscale x 4 x i16> %1, <vscale x 4 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmseq_mask_vi_nxv4i16_i16:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e16, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a0, e16, m1, ta, mu
 ; CHECK-NEXT:    vmseq.vi v10, v8, 9, v0.t
 ; CHECK-NEXT:    vmv.v.v v0, v10
 ; CHECK-NEXT:    ret
@@ -2182,9 +2182,9 @@ entry:
 define <vscale x 8 x i1> @intrinsic_vmseq_mask_vi_nxv8i16_i16(<vscale x 8 x i1> %0, <vscale x 8 x i16> %1, <vscale x 8 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmseq_mask_vi_nxv8i16_i16:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e16, m2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmv1r.v v0, v10
-; CHECK-NEXT:    vsetvli zero, a0, e16, m2, ta, mu
 ; CHECK-NEXT:    vmseq.vi v11, v8, 9, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v11
 ; CHECK-NEXT:    ret
@@ -2217,9 +2217,9 @@ entry:
 define <vscale x 16 x i1> @intrinsic_vmseq_mask_vi_nxv16i16_i16(<vscale x 16 x i1> %0, <vscale x 16 x i16> %1, <vscale x 16 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmseq_mask_vi_nxv16i16_i16:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e16, m4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v13, v0
 ; CHECK-NEXT:    vmv1r.v v0, v12
-; CHECK-NEXT:    vsetvli zero, a0, e16, m4, ta, mu
 ; CHECK-NEXT:    vmseq.vi v13, v8, 9, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v13
 ; CHECK-NEXT:    ret
@@ -2252,9 +2252,9 @@ entry:
 define <vscale x 1 x i1> @intrinsic_vmseq_mask_vi_nxv1i32_i32(<vscale x 1 x i1> %0, <vscale x 1 x i32> %1, <vscale x 1 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmseq_mask_vi_nxv1i32_i32:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e32, mf2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a0, e32, mf2, ta, mu
 ; CHECK-NEXT:    vmseq.vi v10, v8, 9, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v10
 ; CHECK-NEXT:    ret
@@ -2287,9 +2287,9 @@ entry:
 define <vscale x 2 x i1> @intrinsic_vmseq_mask_vi_nxv2i32_i32(<vscale x 2 x i1> %0, <vscale x 2 x i32> %1, <vscale x 2 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmseq_mask_vi_nxv2i32_i32:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e32, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a0, e32, m1, ta, mu
 ; CHECK-NEXT:    vmseq.vi v10, v8, 9, v0.t
 ; CHECK-NEXT:    vmv.v.v v0, v10
 ; CHECK-NEXT:    ret
@@ -2322,9 +2322,9 @@ entry:
 define <vscale x 4 x i1> @intrinsic_vmseq_mask_vi_nxv4i32_i32(<vscale x 4 x i1> %0, <vscale x 4 x i32> %1, <vscale x 4 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmseq_mask_vi_nxv4i32_i32:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e32, m2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmv1r.v v0, v10
-; CHECK-NEXT:    vsetvli zero, a0, e32, m2, ta, mu
 ; CHECK-NEXT:    vmseq.vi v11, v8, 9, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v11
 ; CHECK-NEXT:    ret
@@ -2357,9 +2357,9 @@ entry:
 define <vscale x 8 x i1> @intrinsic_vmseq_mask_vi_nxv8i32_i32(<vscale x 8 x i1> %0, <vscale x 8 x i32> %1, <vscale x 8 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmseq_mask_vi_nxv8i32_i32:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e32, m4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v13, v0
 ; CHECK-NEXT:    vmv1r.v v0, v12
-; CHECK-NEXT:    vsetvli zero, a0, e32, m4, ta, mu
 ; CHECK-NEXT:    vmseq.vi v13, v8, 9, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v13
 ; CHECK-NEXT:    ret
@@ -2392,9 +2392,9 @@ entry:
 define <vscale x 1 x i1> @intrinsic_vmseq_mask_vi_nxv1i64_i64(<vscale x 1 x i1> %0, <vscale x 1 x i64> %1, <vscale x 1 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmseq_mask_vi_nxv1i64_i64:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e64, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a0, e64, m1, ta, mu
 ; CHECK-NEXT:    vmseq.vi v10, v8, 9, v0.t
 ; CHECK-NEXT:    vmv.v.v v0, v10
 ; CHECK-NEXT:    ret
@@ -2427,9 +2427,9 @@ entry:
 define <vscale x 2 x i1> @intrinsic_vmseq_mask_vi_nxv2i64_i64(<vscale x 2 x i1> %0, <vscale x 2 x i64> %1, <vscale x 2 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmseq_mask_vi_nxv2i64_i64:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e64, m2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmv1r.v v0, v10
-; CHECK-NEXT:    vsetvli zero, a0, e64, m2, ta, mu
 ; CHECK-NEXT:    vmseq.vi v11, v8, 9, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v11
 ; CHECK-NEXT:    ret
@@ -2462,9 +2462,9 @@ entry:
 define <vscale x 4 x i1> @intrinsic_vmseq_mask_vi_nxv4i64_i64(<vscale x 4 x i1> %0, <vscale x 4 x i64> %1, <vscale x 4 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmseq_mask_vi_nxv4i64_i64:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e64, m4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v13, v0
 ; CHECK-NEXT:    vmv1r.v v0, v12
-; CHECK-NEXT:    vsetvli zero, a0, e64, m4, ta, mu
 ; CHECK-NEXT:    vmseq.vi v13, v8, 9, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v13
 ; CHECK-NEXT:    ret
diff --git a/llvm/test/CodeGen/RISCV/rvv/vmsge.ll b/llvm/test/CodeGen/RISCV/rvv/vmsge.ll
index 2dc133d169f0a..45e3840f7e673 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vmsge.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vmsge.ll
@@ -34,8 +34,8 @@ declare <vscale x 1 x i1> @llvm.riscv.vmsge.mask.nxv1i8(
 define <vscale x 1 x i1> @intrinsic_vmsge_mask_vv_nxv1i8_nxv1i8(<vscale x 1 x i1> %0, <vscale x 1 x i8> %1, <vscale x 1 x i8> %2, <vscale x 1 x i8> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsge_mask_vv_nxv1i8_nxv1i8:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e8, mf8, ta, mu
+; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmsle.vv v0, v9, v8
 ; CHECK-NEXT:    vmsle.vv v11, v10, v9, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v11
@@ -85,8 +85,8 @@ declare <vscale x 2 x i1> @llvm.riscv.vmsge.mask.nxv2i8(
 define <vscale x 2 x i1> @intrinsic_vmsge_mask_vv_nxv2i8_nxv2i8(<vscale x 2 x i1> %0, <vscale x 2 x i8> %1, <vscale x 2 x i8> %2, <vscale x 2 x i8> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsge_mask_vv_nxv2i8_nxv2i8:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e8, mf4, ta, mu
+; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmsle.vv v0, v9, v8
 ; CHECK-NEXT:    vmsle.vv v11, v10, v9, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v11
@@ -136,8 +136,8 @@ declare <vscale x 4 x i1> @llvm.riscv.vmsge.mask.nxv4i8(
 define <vscale x 4 x i1> @intrinsic_vmsge_mask_vv_nxv4i8_nxv4i8(<vscale x 4 x i1> %0, <vscale x 4 x i8> %1, <vscale x 4 x i8> %2, <vscale x 4 x i8> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsge_mask_vv_nxv4i8_nxv4i8:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e8, mf2, ta, mu
+; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmsle.vv v0, v9, v8
 ; CHECK-NEXT:    vmsle.vv v11, v10, v9, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v11
@@ -187,8 +187,8 @@ declare <vscale x 8 x i1> @llvm.riscv.vmsge.mask.nxv8i8(
 define <vscale x 8 x i1> @intrinsic_vmsge_mask_vv_nxv8i8_nxv8i8(<vscale x 8 x i1> %0, <vscale x 8 x i8> %1, <vscale x 8 x i8> %2, <vscale x 8 x i8> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsge_mask_vv_nxv8i8_nxv8i8:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e8, m1, ta, mu
+; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmsle.vv v0, v9, v8
 ; CHECK-NEXT:    vmsle.vv v11, v10, v9, v0.t
 ; CHECK-NEXT:    vmv.v.v v0, v11
@@ -238,8 +238,8 @@ declare <vscale x 16 x i1> @llvm.riscv.vmsge.mask.nxv16i8(
 define <vscale x 16 x i1> @intrinsic_vmsge_mask_vv_nxv16i8_nxv16i8(<vscale x 16 x i1> %0, <vscale x 16 x i8> %1, <vscale x 16 x i8> %2, <vscale x 16 x i8> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsge_mask_vv_nxv16i8_nxv16i8:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v14, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e8, m2, ta, mu
+; CHECK-NEXT:    vmv1r.v v14, v0
 ; CHECK-NEXT:    vmsle.vv v0, v10, v8
 ; CHECK-NEXT:    vmsle.vv v14, v12, v10, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v14
@@ -289,8 +289,8 @@ declare <vscale x 32 x i1> @llvm.riscv.vmsge.mask.nxv32i8(
 define <vscale x 32 x i1> @intrinsic_vmsge_mask_vv_nxv32i8_nxv32i8(<vscale x 32 x i1> %0, <vscale x 32 x i8> %1, <vscale x 32 x i8> %2, <vscale x 32 x i8> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsge_mask_vv_nxv32i8_nxv32i8:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v20, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e8, m4, ta, mu
+; CHECK-NEXT:    vmv1r.v v20, v0
 ; CHECK-NEXT:    vmsle.vv v0, v12, v8
 ; CHECK-NEXT:    vmsle.vv v20, v16, v12, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v20
@@ -340,8 +340,8 @@ declare <vscale x 1 x i1> @llvm.riscv.vmsge.mask.nxv1i16(
 define <vscale x 1 x i1> @intrinsic_vmsge_mask_vv_nxv1i16_nxv1i16(<vscale x 1 x i1> %0, <vscale x 1 x i16> %1, <vscale x 1 x i16> %2, <vscale x 1 x i16> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsge_mask_vv_nxv1i16_nxv1i16:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e16, mf4, ta, mu
+; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmsle.vv v0, v9, v8
 ; CHECK-NEXT:    vmsle.vv v11, v10, v9, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v11
@@ -391,8 +391,8 @@ declare <vscale x 2 x i1> @llvm.riscv.vmsge.mask.nxv2i16(
 define <vscale x 2 x i1> @intrinsic_vmsge_mask_vv_nxv2i16_nxv2i16(<vscale x 2 x i1> %0, <vscale x 2 x i16> %1, <vscale x 2 x i16> %2, <vscale x 2 x i16> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsge_mask_vv_nxv2i16_nxv2i16:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e16, mf2, ta, mu
+; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmsle.vv v0, v9, v8
 ; CHECK-NEXT:    vmsle.vv v11, v10, v9, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v11
@@ -442,8 +442,8 @@ declare <vscale x 4 x i1> @llvm.riscv.vmsge.mask.nxv4i16(
 define <vscale x 4 x i1> @intrinsic_vmsge_mask_vv_nxv4i16_nxv4i16(<vscale x 4 x i1> %0, <vscale x 4 x i16> %1, <vscale x 4 x i16> %2, <vscale x 4 x i16> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsge_mask_vv_nxv4i16_nxv4i16:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e16, m1, ta, mu
+; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmsle.vv v0, v9, v8
 ; CHECK-NEXT:    vmsle.vv v11, v10, v9, v0.t
 ; CHECK-NEXT:    vmv.v.v v0, v11
@@ -493,8 +493,8 @@ declare <vscale x 8 x i1> @llvm.riscv.vmsge.mask.nxv8i16(
 define <vscale x 8 x i1> @intrinsic_vmsge_mask_vv_nxv8i16_nxv8i16(<vscale x 8 x i1> %0, <vscale x 8 x i16> %1, <vscale x 8 x i16> %2, <vscale x 8 x i16> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsge_mask_vv_nxv8i16_nxv8i16:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v14, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e16, m2, ta, mu
+; CHECK-NEXT:    vmv1r.v v14, v0
 ; CHECK-NEXT:    vmsle.vv v0, v10, v8
 ; CHECK-NEXT:    vmsle.vv v14, v12, v10, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v14
@@ -544,8 +544,8 @@ declare <vscale x 16 x i1> @llvm.riscv.vmsge.mask.nxv16i16(
 define <vscale x 16 x i1> @intrinsic_vmsge_mask_vv_nxv16i16_nxv16i16(<vscale x 16 x i1> %0, <vscale x 16 x i16> %1, <vscale x 16 x i16> %2, <vscale x 16 x i16> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsge_mask_vv_nxv16i16_nxv16i16:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v20, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e16, m4, ta, mu
+; CHECK-NEXT:    vmv1r.v v20, v0
 ; CHECK-NEXT:    vmsle.vv v0, v12, v8
 ; CHECK-NEXT:    vmsle.vv v20, v16, v12, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v20
@@ -595,8 +595,8 @@ declare <vscale x 1 x i1> @llvm.riscv.vmsge.mask.nxv1i32(
 define <vscale x 1 x i1> @intrinsic_vmsge_mask_vv_nxv1i32_nxv1i32(<vscale x 1 x i1> %0, <vscale x 1 x i32> %1, <vscale x 1 x i32> %2, <vscale x 1 x i32> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsge_mask_vv_nxv1i32_nxv1i32:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e32, mf2, ta, mu
+; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmsle.vv v0, v9, v8
 ; CHECK-NEXT:    vmsle.vv v11, v10, v9, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v11
@@ -646,8 +646,8 @@ declare <vscale x 2 x i1> @llvm.riscv.vmsge.mask.nxv2i32(
 define <vscale x 2 x i1> @intrinsic_vmsge_mask_vv_nxv2i32_nxv2i32(<vscale x 2 x i1> %0, <vscale x 2 x i32> %1, <vscale x 2 x i32> %2, <vscale x 2 x i32> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsge_mask_vv_nxv2i32_nxv2i32:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e32, m1, ta, mu
+; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmsle.vv v0, v9, v8
 ; CHECK-NEXT:    vmsle.vv v11, v10, v9, v0.t
 ; CHECK-NEXT:    vmv.v.v v0, v11
@@ -697,8 +697,8 @@ declare <vscale x 4 x i1> @llvm.riscv.vmsge.mask.nxv4i32(
 define <vscale x 4 x i1> @intrinsic_vmsge_mask_vv_nxv4i32_nxv4i32(<vscale x 4 x i1> %0, <vscale x 4 x i32> %1, <vscale x 4 x i32> %2, <vscale x 4 x i32> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsge_mask_vv_nxv4i32_nxv4i32:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v14, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e32, m2, ta, mu
+; CHECK-NEXT:    vmv1r.v v14, v0
 ; CHECK-NEXT:    vmsle.vv v0, v10, v8
 ; CHECK-NEXT:    vmsle.vv v14, v12, v10, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v14
@@ -748,8 +748,8 @@ declare <vscale x 8 x i1> @llvm.riscv.vmsge.mask.nxv8i32(
 define <vscale x 8 x i1> @intrinsic_vmsge_mask_vv_nxv8i32_nxv8i32(<vscale x 8 x i1> %0, <vscale x 8 x i32> %1, <vscale x 8 x i32> %2, <vscale x 8 x i32> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsge_mask_vv_nxv8i32_nxv8i32:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v20, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e32, m4, ta, mu
+; CHECK-NEXT:    vmv1r.v v20, v0
 ; CHECK-NEXT:    vmsle.vv v0, v12, v8
 ; CHECK-NEXT:    vmsle.vv v20, v16, v12, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v20
@@ -799,8 +799,8 @@ declare <vscale x 1 x i1> @llvm.riscv.vmsge.mask.nxv1i64(
 define <vscale x 1 x i1> @intrinsic_vmsge_mask_vv_nxv1i64_nxv1i64(<vscale x 1 x i1> %0, <vscale x 1 x i64> %1, <vscale x 1 x i64> %2, <vscale x 1 x i64> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsge_mask_vv_nxv1i64_nxv1i64:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e64, m1, ta, mu
+; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmsle.vv v0, v9, v8
 ; CHECK-NEXT:    vmsle.vv v11, v10, v9, v0.t
 ; CHECK-NEXT:    vmv.v.v v0, v11
@@ -850,8 +850,8 @@ declare <vscale x 2 x i1> @llvm.riscv.vmsge.mask.nxv2i64(
 define <vscale x 2 x i1> @intrinsic_vmsge_mask_vv_nxv2i64_nxv2i64(<vscale x 2 x i1> %0, <vscale x 2 x i64> %1, <vscale x 2 x i64> %2, <vscale x 2 x i64> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsge_mask_vv_nxv2i64_nxv2i64:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v14, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e64, m2, ta, mu
+; CHECK-NEXT:    vmv1r.v v14, v0
 ; CHECK-NEXT:    vmsle.vv v0, v10, v8
 ; CHECK-NEXT:    vmsle.vv v14, v12, v10, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v14
@@ -901,8 +901,8 @@ declare <vscale x 4 x i1> @llvm.riscv.vmsge.mask.nxv4i64(
 define <vscale x 4 x i1> @intrinsic_vmsge_mask_vv_nxv4i64_nxv4i64(<vscale x 4 x i1> %0, <vscale x 4 x i64> %1, <vscale x 4 x i64> %2, <vscale x 4 x i64> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsge_mask_vv_nxv4i64_nxv4i64:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v20, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e64, m4, ta, mu
+; CHECK-NEXT:    vmv1r.v v20, v0
 ; CHECK-NEXT:    vmsle.vv v0, v12, v8
 ; CHECK-NEXT:    vmsle.vv v20, v16, v12, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v20
@@ -953,9 +953,9 @@ declare <vscale x 1 x i1> @llvm.riscv.vmsge.mask.nxv1i8.i8(
 define <vscale x 1 x i1> @intrinsic_vmsge_mask_vx_nxv1i8_i8(<vscale x 1 x i1> %0, <vscale x 1 x i8> %1, i8 %2, <vscale x 1 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsge_mask_vx_nxv1i8_i8:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e8, mf8, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a1, e8, mf8, ta, mu
 ; CHECK-NEXT:    vmslt.vx v10, v8, a0, v0.t
 ; CHECK-NEXT:    vmxor.mm v0, v10, v9
 ; CHECK-NEXT:    ret
@@ -1001,9 +1001,9 @@ declare <vscale x 2 x i1> @llvm.riscv.vmsge.mask.nxv2i8.i8(
 define <vscale x 2 x i1> @intrinsic_vmsge_mask_vx_nxv2i8_i8(<vscale x 2 x i1> %0, <vscale x 2 x i8> %1, i8 %2, <vscale x 2 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsge_mask_vx_nxv2i8_i8:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e8, mf4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a1, e8, mf4, ta, mu
 ; CHECK-NEXT:    vmslt.vx v10, v8, a0, v0.t
 ; CHECK-NEXT:    vmxor.mm v0, v10, v9
 ; CHECK-NEXT:    ret
@@ -1049,9 +1049,9 @@ declare <vscale x 4 x i1> @llvm.riscv.vmsge.mask.nxv4i8.i8(
 define <vscale x 4 x i1> @intrinsic_vmsge_mask_vx_nxv4i8_i8(<vscale x 4 x i1> %0, <vscale x 4 x i8> %1, i8 %2, <vscale x 4 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsge_mask_vx_nxv4i8_i8:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e8, mf2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a1, e8, mf2, ta, mu
 ; CHECK-NEXT:    vmslt.vx v10, v8, a0, v0.t
 ; CHECK-NEXT:    vmxor.mm v0, v10, v9
 ; CHECK-NEXT:    ret
@@ -1069,11 +1069,11 @@ entry:
 define <vscale x 4 x i1> @intrinsic_vmsge_mask_vx_nxv4i8_i8_1(<vscale x 4 x i1> %0, <vscale x 4 x i8> %1, <vscale x 4 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsge_mask_vx_nxv4i8_i8_1:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e8, mf2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
-; CHECK-NEXT:    li a1, 99
+; CHECK-NEXT:    li a0, 99
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a0, e8, mf2, ta, mu
-; CHECK-NEXT:    vmsgt.vx v10, v8, a1, v0.t
+; CHECK-NEXT:    vmsgt.vx v10, v8, a0, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v10
 ; CHECK-NEXT:    ret
 entry:
@@ -1136,9 +1136,9 @@ declare <vscale x 8 x i1> @llvm.riscv.vmsge.mask.nxv8i8.i8(
 define <vscale x 8 x i1> @intrinsic_vmsge_mask_vx_nxv8i8_i8(<vscale x 8 x i1> %0, <vscale x 8 x i8> %1, i8 %2, <vscale x 8 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsge_mask_vx_nxv8i8_i8:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e8, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a1, e8, m1, ta, mu
 ; CHECK-NEXT:    vmslt.vx v10, v8, a0, v0.t
 ; CHECK-NEXT:    vmxor.mm v0, v10, v9
 ; CHECK-NEXT:    ret
@@ -1184,9 +1184,9 @@ declare <vscale x 16 x i1> @llvm.riscv.vmsge.mask.nxv16i8.i8(
 define <vscale x 16 x i1> @intrinsic_vmsge_mask_vx_nxv16i8_i8(<vscale x 16 x i1> %0, <vscale x 16 x i8> %1, i8 %2, <vscale x 16 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsge_mask_vx_nxv16i8_i8:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e8, m2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmv1r.v v0, v10
-; CHECK-NEXT:    vsetvli zero, a1, e8, m2, ta, mu
 ; CHECK-NEXT:    vmslt.vx v11, v8, a0, v0.t
 ; CHECK-NEXT:    vmxor.mm v0, v11, v10
 ; CHECK-NEXT:    ret
@@ -1232,9 +1232,9 @@ declare <vscale x 32 x i1> @llvm.riscv.vmsge.mask.nxv32i8.i8(
 define <vscale x 32 x i1> @intrinsic_vmsge_mask_vx_nxv32i8_i8(<vscale x 32 x i1> %0, <vscale x 32 x i8> %1, i8 %2, <vscale x 32 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsge_mask_vx_nxv32i8_i8:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e8, m4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v13, v0
 ; CHECK-NEXT:    vmv1r.v v0, v12
-; CHECK-NEXT:    vsetvli zero, a1, e8, m4, ta, mu
 ; CHECK-NEXT:    vmslt.vx v13, v8, a0, v0.t
 ; CHECK-NEXT:    vmxor.mm v0, v13, v12
 ; CHECK-NEXT:    ret
@@ -1280,9 +1280,9 @@ declare <vscale x 1 x i1> @llvm.riscv.vmsge.mask.nxv1i16.i16(
 define <vscale x 1 x i1> @intrinsic_vmsge_mask_vx_nxv1i16_i16(<vscale x 1 x i1> %0, <vscale x 1 x i16> %1, i16 %2, <vscale x 1 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsge_mask_vx_nxv1i16_i16:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, mf4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a1, e16, mf4, ta, mu
 ; CHECK-NEXT:    vmslt.vx v10, v8, a0, v0.t
 ; CHECK-NEXT:    vmxor.mm v0, v10, v9
 ; CHECK-NEXT:    ret
@@ -1328,9 +1328,9 @@ declare <vscale x 2 x i1> @llvm.riscv.vmsge.mask.nxv2i16.i16(
 define <vscale x 2 x i1> @intrinsic_vmsge_mask_vx_nxv2i16_i16(<vscale x 2 x i1> %0, <vscale x 2 x i16> %1, i16 %2, <vscale x 2 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsge_mask_vx_nxv2i16_i16:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, mf2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a1, e16, mf2, ta, mu
 ; CHECK-NEXT:    vmslt.vx v10, v8, a0, v0.t
 ; CHECK-NEXT:    vmxor.mm v0, v10, v9
 ; CHECK-NEXT:    ret
@@ -1376,9 +1376,9 @@ declare <vscale x 4 x i1> @llvm.riscv.vmsge.mask.nxv4i16.i16(
 define <vscale x 4 x i1> @intrinsic_vmsge_mask_vx_nxv4i16_i16(<vscale x 4 x i1> %0, <vscale x 4 x i16> %1, i16 %2, <vscale x 4 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsge_mask_vx_nxv4i16_i16:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a1, e16, m1, ta, mu
 ; CHECK-NEXT:    vmslt.vx v10, v8, a0, v0.t
 ; CHECK-NEXT:    vmxor.mm v0, v10, v9
 ; CHECK-NEXT:    ret
@@ -1424,9 +1424,9 @@ declare <vscale x 8 x i1> @llvm.riscv.vmsge.mask.nxv8i16.i16(
 define <vscale x 8 x i1> @intrinsic_vmsge_mask_vx_nxv8i16_i16(<vscale x 8 x i1> %0, <vscale x 8 x i16> %1, i16 %2, <vscale x 8 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsge_mask_vx_nxv8i16_i16:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, m2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmv1r.v v0, v10
-; CHECK-NEXT:    vsetvli zero, a1, e16, m2, ta, mu
 ; CHECK-NEXT:    vmslt.vx v11, v8, a0, v0.t
 ; CHECK-NEXT:    vmxor.mm v0, v11, v10
 ; CHECK-NEXT:    ret
@@ -1472,9 +1472,9 @@ declare <vscale x 16 x i1> @llvm.riscv.vmsge.mask.nxv16i16.i16(
 define <vscale x 16 x i1> @intrinsic_vmsge_mask_vx_nxv16i16_i16(<vscale x 16 x i1> %0, <vscale x 16 x i16> %1, i16 %2, <vscale x 16 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsge_mask_vx_nxv16i16_i16:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, m4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v13, v0
 ; CHECK-NEXT:    vmv1r.v v0, v12
-; CHECK-NEXT:    vsetvli zero, a1, e16, m4, ta, mu
 ; CHECK-NEXT:    vmslt.vx v13, v8, a0, v0.t
 ; CHECK-NEXT:    vmxor.mm v0, v13, v12
 ; CHECK-NEXT:    ret
@@ -1520,9 +1520,9 @@ declare <vscale x 1 x i1> @llvm.riscv.vmsge.mask.nxv1i32.i32(
 define <vscale x 1 x i1> @intrinsic_vmsge_mask_vx_nxv1i32_i32(<vscale x 1 x i1> %0, <vscale x 1 x i32> %1, i32 %2, <vscale x 1 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsge_mask_vx_nxv1i32_i32:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e32, mf2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a1, e32, mf2, ta, mu
 ; CHECK-NEXT:    vmslt.vx v10, v8, a0, v0.t
 ; CHECK-NEXT:    vmxor.mm v0, v10, v9
 ; CHECK-NEXT:    ret
@@ -1568,9 +1568,9 @@ declare <vscale x 2 x i1> @llvm.riscv.vmsge.mask.nxv2i32.i32(
 define <vscale x 2 x i1> @intrinsic_vmsge_mask_vx_nxv2i32_i32(<vscale x 2 x i1> %0, <vscale x 2 x i32> %1, i32 %2, <vscale x 2 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsge_mask_vx_nxv2i32_i32:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e32, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a1, e32, m1, ta, mu
 ; CHECK-NEXT:    vmslt.vx v10, v8, a0, v0.t
 ; CHECK-NEXT:    vmxor.mm v0, v10, v9
 ; CHECK-NEXT:    ret
@@ -1616,9 +1616,9 @@ declare <vscale x 4 x i1> @llvm.riscv.vmsge.mask.nxv4i32.i32(
 define <vscale x 4 x i1> @intrinsic_vmsge_mask_vx_nxv4i32_i32(<vscale x 4 x i1> %0, <vscale x 4 x i32> %1, i32 %2, <vscale x 4 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsge_mask_vx_nxv4i32_i32:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e32, m2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmv1r.v v0, v10
-; CHECK-NEXT:    vsetvli zero, a1, e32, m2, ta, mu
 ; CHECK-NEXT:    vmslt.vx v11, v8, a0, v0.t
 ; CHECK-NEXT:    vmxor.mm v0, v11, v10
 ; CHECK-NEXT:    ret
@@ -1664,9 +1664,9 @@ declare <vscale x 8 x i1> @llvm.riscv.vmsge.mask.nxv8i32.i32(
 define <vscale x 8 x i1> @intrinsic_vmsge_mask_vx_nxv8i32_i32(<vscale x 8 x i1> %0, <vscale x 8 x i32> %1, i32 %2, <vscale x 8 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsge_mask_vx_nxv8i32_i32:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e32, m4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v13, v0
 ; CHECK-NEXT:    vmv1r.v v0, v12
-; CHECK-NEXT:    vsetvli zero, a1, e32, m4, ta, mu
 ; CHECK-NEXT:    vmslt.vx v13, v8, a0, v0.t
 ; CHECK-NEXT:    vmxor.mm v0, v13, v12
 ; CHECK-NEXT:    ret
@@ -1739,9 +1739,9 @@ define <vscale x 1 x i1> @intrinsic_vmsge_mask_vx_nxv1i64_i64(<vscale x 1 x i1>
 ;
 ; RV64-LABEL: intrinsic_vmsge_mask_vx_nxv1i64_i64:
 ; RV64:       # %bb.0: # %entry
+; RV64-NEXT:    vsetvli zero, a1, e64, m1, ta, mu
 ; RV64-NEXT:    vmv1r.v v10, v0
 ; RV64-NEXT:    vmv1r.v v0, v9
-; RV64-NEXT:    vsetvli zero, a1, e64, m1, ta, mu
 ; RV64-NEXT:    vmslt.vx v10, v8, a0, v0.t
 ; RV64-NEXT:    vmxor.mm v0, v10, v9
 ; RV64-NEXT:    ret
@@ -1814,9 +1814,9 @@ define <vscale x 2 x i1> @intrinsic_vmsge_mask_vx_nxv2i64_i64(<vscale x 2 x i1>
 ;
 ; RV64-LABEL: intrinsic_vmsge_mask_vx_nxv2i64_i64:
 ; RV64:       # %bb.0: # %entry
+; RV64-NEXT:    vsetvli zero, a1, e64, m2, ta, mu
 ; RV64-NEXT:    vmv1r.v v11, v0
 ; RV64-NEXT:    vmv1r.v v0, v10
-; RV64-NEXT:    vsetvli zero, a1, e64, m2, ta, mu
 ; RV64-NEXT:    vmslt.vx v11, v8, a0, v0.t
 ; RV64-NEXT:    vmxor.mm v0, v11, v10
 ; RV64-NEXT:    ret
@@ -1889,9 +1889,9 @@ define <vscale x 4 x i1> @intrinsic_vmsge_mask_vx_nxv4i64_i64(<vscale x 4 x i1>
 ;
 ; RV64-LABEL: intrinsic_vmsge_mask_vx_nxv4i64_i64:
 ; RV64:       # %bb.0: # %entry
+; RV64-NEXT:    vsetvli zero, a1, e64, m4, ta, mu
 ; RV64-NEXT:    vmv1r.v v13, v0
 ; RV64-NEXT:    vmv1r.v v0, v12
-; RV64-NEXT:    vsetvli zero, a1, e64, m4, ta, mu
 ; RV64-NEXT:    vmslt.vx v13, v8, a0, v0.t
 ; RV64-NEXT:    vmxor.mm v0, v13, v12
 ; RV64-NEXT:    ret
@@ -1924,9 +1924,9 @@ entry:
 define <vscale x 1 x i1> @intrinsic_vmsge_mask_vi_nxv1i8_i8(<vscale x 1 x i1> %0, <vscale x 1 x i8> %1, <vscale x 1 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsge_mask_vi_nxv1i8_i8:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e8, mf8, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a0, e8, mf8, ta, mu
 ; CHECK-NEXT:    vmsgt.vi v10, v8, -15, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v10
 ; CHECK-NEXT:    ret
@@ -1959,9 +1959,9 @@ entry:
 define <vscale x 2 x i1> @intrinsic_vmsge_mask_vi_nxv2i8_i8(<vscale x 2 x i1> %0, <vscale x 2 x i8> %1, <vscale x 2 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsge_mask_vi_nxv2i8_i8:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e8, mf4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a0, e8, mf4, ta, mu
 ; CHECK-NEXT:    vmsgt.vi v10, v8, -13, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v10
 ; CHECK-NEXT:    ret
@@ -2025,9 +2025,9 @@ entry:
 define <vscale x 4 x i1> @intrinsic_vmsge_mask_vi_nxv4i8_i8(<vscale x 4 x i1> %0, <vscale x 4 x i8> %1, <vscale x 4 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsge_mask_vi_nxv4i8_i8:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e8, mf2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a0, e8, mf2, ta, mu
 ; CHECK-NEXT:    vmsgt.vi v10, v8, -11, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v10
 ; CHECK-NEXT:    ret
@@ -2060,9 +2060,9 @@ entry:
 define <vscale x 8 x i1> @intrinsic_vmsge_mask_vi_nxv8i8_i8(<vscale x 8 x i1> %0, <vscale x 8 x i8> %1, <vscale x 8 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsge_mask_vi_nxv8i8_i8:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e8, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a0, e8, m1, ta, mu
 ; CHECK-NEXT:    vmsgt.vi v10, v8, -9, v0.t
 ; CHECK-NEXT:    vmv.v.v v0, v10
 ; CHECK-NEXT:    ret
@@ -2095,9 +2095,9 @@ entry:
 define <vscale x 16 x i1> @intrinsic_vmsge_mask_vi_nxv16i8_i8(<vscale x 16 x i1> %0, <vscale x 16 x i8> %1, <vscale x 16 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsge_mask_vi_nxv16i8_i8:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e8, m2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmv1r.v v0, v10
-; CHECK-NEXT:    vsetvli zero, a0, e8, m2, ta, mu
 ; CHECK-NEXT:    vmsgt.vi v11, v8, -7, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v11
 ; CHECK-NEXT:    ret
@@ -2130,9 +2130,9 @@ entry:
 define <vscale x 32 x i1> @intrinsic_vmsge_mask_vi_nxv32i8_i8(<vscale x 32 x i1> %0, <vscale x 32 x i8> %1, <vscale x 32 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsge_mask_vi_nxv32i8_i8:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e8, m4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v13, v0
 ; CHECK-NEXT:    vmv1r.v v0, v12
-; CHECK-NEXT:    vsetvli zero, a0, e8, m4, ta, mu
 ; CHECK-NEXT:    vmsgt.vi v13, v8, -5, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v13
 ; CHECK-NEXT:    ret
@@ -2165,9 +2165,9 @@ entry:
 define <vscale x 1 x i1> @intrinsic_vmsge_mask_vi_nxv1i16_i16(<vscale x 1 x i1> %0, <vscale x 1 x i16> %1, <vscale x 1 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsge_mask_vi_nxv1i16_i16:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e16, mf4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a0, e16, mf4, ta, mu
 ; CHECK-NEXT:    vmsgt.vi v10, v8, -3, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v10
 ; CHECK-NEXT:    ret
@@ -2200,9 +2200,9 @@ entry:
 define <vscale x 2 x i1> @intrinsic_vmsge_mask_vi_nxv2i16_i16(<vscale x 2 x i1> %0, <vscale x 2 x i16> %1, <vscale x 2 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsge_mask_vi_nxv2i16_i16:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e16, mf2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a0, e16, mf2, ta, mu
 ; CHECK-NEXT:    vmsgt.vi v10, v8, -1, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v10
 ; CHECK-NEXT:    ret
@@ -2235,9 +2235,9 @@ entry:
 define <vscale x 4 x i1> @intrinsic_vmsge_mask_vi_nxv4i16_i16(<vscale x 4 x i1> %0, <vscale x 4 x i16> %1, <vscale x 4 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsge_mask_vi_nxv4i16_i16:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e16, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a0, e16, m1, ta, mu
 ; CHECK-NEXT:    vmsgt.vi v10, v8, 0, v0.t
 ; CHECK-NEXT:    vmv.v.v v0, v10
 ; CHECK-NEXT:    ret
@@ -2270,9 +2270,9 @@ entry:
 define <vscale x 8 x i1> @intrinsic_vmsge_mask_vi_nxv8i16_i16(<vscale x 8 x i1> %0, <vscale x 8 x i16> %1, <vscale x 8 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsge_mask_vi_nxv8i16_i16:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e16, m2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmv1r.v v0, v10
-; CHECK-NEXT:    vsetvli zero, a0, e16, m2, ta, mu
 ; CHECK-NEXT:    vmsgt.vi v11, v8, 2, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v11
 ; CHECK-NEXT:    ret
@@ -2305,9 +2305,9 @@ entry:
 define <vscale x 16 x i1> @intrinsic_vmsge_mask_vi_nxv16i16_i16(<vscale x 16 x i1> %0, <vscale x 16 x i16> %1, <vscale x 16 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsge_mask_vi_nxv16i16_i16:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e16, m4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v13, v0
 ; CHECK-NEXT:    vmv1r.v v0, v12
-; CHECK-NEXT:    vsetvli zero, a0, e16, m4, ta, mu
 ; CHECK-NEXT:    vmsgt.vi v13, v8, 4, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v13
 ; CHECK-NEXT:    ret
@@ -2340,9 +2340,9 @@ entry:
 define <vscale x 1 x i1> @intrinsic_vmsge_mask_vi_nxv1i32_i32(<vscale x 1 x i1> %0, <vscale x 1 x i32> %1, <vscale x 1 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsge_mask_vi_nxv1i32_i32:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e32, mf2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a0, e32, mf2, ta, mu
 ; CHECK-NEXT:    vmsgt.vi v10, v8, 6, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v10
 ; CHECK-NEXT:    ret
@@ -2375,9 +2375,9 @@ entry:
 define <vscale x 2 x i1> @intrinsic_vmsge_mask_vi_nxv2i32_i32(<vscale x 2 x i1> %0, <vscale x 2 x i32> %1, <vscale x 2 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsge_mask_vi_nxv2i32_i32:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e32, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a0, e32, m1, ta, mu
 ; CHECK-NEXT:    vmsgt.vi v10, v8, 8, v0.t
 ; CHECK-NEXT:    vmv.v.v v0, v10
 ; CHECK-NEXT:    ret
@@ -2410,9 +2410,9 @@ entry:
 define <vscale x 4 x i1> @intrinsic_vmsge_mask_vi_nxv4i32_i32(<vscale x 4 x i1> %0, <vscale x 4 x i32> %1, <vscale x 4 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsge_mask_vi_nxv4i32_i32:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e32, m2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmv1r.v v0, v10
-; CHECK-NEXT:    vsetvli zero, a0, e32, m2, ta, mu
 ; CHECK-NEXT:    vmsgt.vi v11, v8, 10, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v11
 ; CHECK-NEXT:    ret
@@ -2445,9 +2445,9 @@ entry:
 define <vscale x 8 x i1> @intrinsic_vmsge_mask_vi_nxv8i32_i32(<vscale x 8 x i1> %0, <vscale x 8 x i32> %1, <vscale x 8 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsge_mask_vi_nxv8i32_i32:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e32, m4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v13, v0
 ; CHECK-NEXT:    vmv1r.v v0, v12
-; CHECK-NEXT:    vsetvli zero, a0, e32, m4, ta, mu
 ; CHECK-NEXT:    vmsgt.vi v13, v8, 12, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v13
 ; CHECK-NEXT:    ret
@@ -2480,9 +2480,9 @@ entry:
 define <vscale x 1 x i1> @intrinsic_vmsge_mask_vi_nxv1i64_i64(<vscale x 1 x i1> %0, <vscale x 1 x i64> %1, <vscale x 1 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsge_mask_vi_nxv1i64_i64:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e64, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a0, e64, m1, ta, mu
 ; CHECK-NEXT:    vmsgt.vi v10, v8, 8, v0.t
 ; CHECK-NEXT:    vmv.v.v v0, v10
 ; CHECK-NEXT:    ret
@@ -2515,9 +2515,9 @@ entry:
 define <vscale x 2 x i1> @intrinsic_vmsge_mask_vi_nxv2i64_i64(<vscale x 2 x i1> %0, <vscale x 2 x i64> %1, <vscale x 2 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsge_mask_vi_nxv2i64_i64:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e64, m2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmv1r.v v0, v10
-; CHECK-NEXT:    vsetvli zero, a0, e64, m2, ta, mu
 ; CHECK-NEXT:    vmsgt.vi v11, v8, 8, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v11
 ; CHECK-NEXT:    ret
@@ -2550,9 +2550,9 @@ entry:
 define <vscale x 4 x i1> @intrinsic_vmsge_mask_vi_nxv4i64_i64(<vscale x 4 x i1> %0, <vscale x 4 x i64> %1, <vscale x 4 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsge_mask_vi_nxv4i64_i64:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e64, m4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v13, v0
 ; CHECK-NEXT:    vmv1r.v v0, v12
-; CHECK-NEXT:    vsetvli zero, a0, e64, m4, ta, mu
 ; CHECK-NEXT:    vmsgt.vi v13, v8, 8, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v13
 ; CHECK-NEXT:    ret
diff --git a/llvm/test/CodeGen/RISCV/rvv/vmsgeu.ll b/llvm/test/CodeGen/RISCV/rvv/vmsgeu.ll
index 69a3835cd4d67..e42be4faafefc 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vmsgeu.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vmsgeu.ll
@@ -34,8 +34,8 @@ declare <vscale x 1 x i1> @llvm.riscv.vmsgeu.mask.nxv1i8(
 define <vscale x 1 x i1> @intrinsic_vmsgeu_mask_vv_nxv1i8_nxv1i8(<vscale x 1 x i1> %0, <vscale x 1 x i8> %1, <vscale x 1 x i8> %2, <vscale x 1 x i8> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsgeu_mask_vv_nxv1i8_nxv1i8:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e8, mf8, ta, mu
+; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmsleu.vv v0, v9, v8
 ; CHECK-NEXT:    vmsleu.vv v11, v10, v9, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v11
@@ -85,8 +85,8 @@ declare <vscale x 2 x i1> @llvm.riscv.vmsgeu.mask.nxv2i8(
 define <vscale x 2 x i1> @intrinsic_vmsgeu_mask_vv_nxv2i8_nxv2i8(<vscale x 2 x i1> %0, <vscale x 2 x i8> %1, <vscale x 2 x i8> %2, <vscale x 2 x i8> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsgeu_mask_vv_nxv2i8_nxv2i8:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e8, mf4, ta, mu
+; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmsleu.vv v0, v9, v8
 ; CHECK-NEXT:    vmsleu.vv v11, v10, v9, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v11
@@ -136,8 +136,8 @@ declare <vscale x 4 x i1> @llvm.riscv.vmsgeu.mask.nxv4i8(
 define <vscale x 4 x i1> @intrinsic_vmsgeu_mask_vv_nxv4i8_nxv4i8(<vscale x 4 x i1> %0, <vscale x 4 x i8> %1, <vscale x 4 x i8> %2, <vscale x 4 x i8> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsgeu_mask_vv_nxv4i8_nxv4i8:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e8, mf2, ta, mu
+; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmsleu.vv v0, v9, v8
 ; CHECK-NEXT:    vmsleu.vv v11, v10, v9, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v11
@@ -187,8 +187,8 @@ declare <vscale x 8 x i1> @llvm.riscv.vmsgeu.mask.nxv8i8(
 define <vscale x 8 x i1> @intrinsic_vmsgeu_mask_vv_nxv8i8_nxv8i8(<vscale x 8 x i1> %0, <vscale x 8 x i8> %1, <vscale x 8 x i8> %2, <vscale x 8 x i8> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsgeu_mask_vv_nxv8i8_nxv8i8:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e8, m1, ta, mu
+; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmsleu.vv v0, v9, v8
 ; CHECK-NEXT:    vmsleu.vv v11, v10, v9, v0.t
 ; CHECK-NEXT:    vmv.v.v v0, v11
@@ -238,8 +238,8 @@ declare <vscale x 16 x i1> @llvm.riscv.vmsgeu.mask.nxv16i8(
 define <vscale x 16 x i1> @intrinsic_vmsgeu_mask_vv_nxv16i8_nxv16i8(<vscale x 16 x i1> %0, <vscale x 16 x i8> %1, <vscale x 16 x i8> %2, <vscale x 16 x i8> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsgeu_mask_vv_nxv16i8_nxv16i8:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v14, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e8, m2, ta, mu
+; CHECK-NEXT:    vmv1r.v v14, v0
 ; CHECK-NEXT:    vmsleu.vv v0, v10, v8
 ; CHECK-NEXT:    vmsleu.vv v14, v12, v10, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v14
@@ -289,8 +289,8 @@ declare <vscale x 32 x i1> @llvm.riscv.vmsgeu.mask.nxv32i8(
 define <vscale x 32 x i1> @intrinsic_vmsgeu_mask_vv_nxv32i8_nxv32i8(<vscale x 32 x i1> %0, <vscale x 32 x i8> %1, <vscale x 32 x i8> %2, <vscale x 32 x i8> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsgeu_mask_vv_nxv32i8_nxv32i8:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v20, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e8, m4, ta, mu
+; CHECK-NEXT:    vmv1r.v v20, v0
 ; CHECK-NEXT:    vmsleu.vv v0, v12, v8
 ; CHECK-NEXT:    vmsleu.vv v20, v16, v12, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v20
@@ -340,8 +340,8 @@ declare <vscale x 1 x i1> @llvm.riscv.vmsgeu.mask.nxv1i16(
 define <vscale x 1 x i1> @intrinsic_vmsgeu_mask_vv_nxv1i16_nxv1i16(<vscale x 1 x i1> %0, <vscale x 1 x i16> %1, <vscale x 1 x i16> %2, <vscale x 1 x i16> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsgeu_mask_vv_nxv1i16_nxv1i16:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e16, mf4, ta, mu
+; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmsleu.vv v0, v9, v8
 ; CHECK-NEXT:    vmsleu.vv v11, v10, v9, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v11
@@ -391,8 +391,8 @@ declare <vscale x 2 x i1> @llvm.riscv.vmsgeu.mask.nxv2i16(
 define <vscale x 2 x i1> @intrinsic_vmsgeu_mask_vv_nxv2i16_nxv2i16(<vscale x 2 x i1> %0, <vscale x 2 x i16> %1, <vscale x 2 x i16> %2, <vscale x 2 x i16> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsgeu_mask_vv_nxv2i16_nxv2i16:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e16, mf2, ta, mu
+; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmsleu.vv v0, v9, v8
 ; CHECK-NEXT:    vmsleu.vv v11, v10, v9, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v11
@@ -442,8 +442,8 @@ declare <vscale x 4 x i1> @llvm.riscv.vmsgeu.mask.nxv4i16(
 define <vscale x 4 x i1> @intrinsic_vmsgeu_mask_vv_nxv4i16_nxv4i16(<vscale x 4 x i1> %0, <vscale x 4 x i16> %1, <vscale x 4 x i16> %2, <vscale x 4 x i16> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsgeu_mask_vv_nxv4i16_nxv4i16:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e16, m1, ta, mu
+; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmsleu.vv v0, v9, v8
 ; CHECK-NEXT:    vmsleu.vv v11, v10, v9, v0.t
 ; CHECK-NEXT:    vmv.v.v v0, v11
@@ -493,8 +493,8 @@ declare <vscale x 8 x i1> @llvm.riscv.vmsgeu.mask.nxv8i16(
 define <vscale x 8 x i1> @intrinsic_vmsgeu_mask_vv_nxv8i16_nxv8i16(<vscale x 8 x i1> %0, <vscale x 8 x i16> %1, <vscale x 8 x i16> %2, <vscale x 8 x i16> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsgeu_mask_vv_nxv8i16_nxv8i16:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v14, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e16, m2, ta, mu
+; CHECK-NEXT:    vmv1r.v v14, v0
 ; CHECK-NEXT:    vmsleu.vv v0, v10, v8
 ; CHECK-NEXT:    vmsleu.vv v14, v12, v10, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v14
@@ -544,8 +544,8 @@ declare <vscale x 16 x i1> @llvm.riscv.vmsgeu.mask.nxv16i16(
 define <vscale x 16 x i1> @intrinsic_vmsgeu_mask_vv_nxv16i16_nxv16i16(<vscale x 16 x i1> %0, <vscale x 16 x i16> %1, <vscale x 16 x i16> %2, <vscale x 16 x i16> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsgeu_mask_vv_nxv16i16_nxv16i16:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v20, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e16, m4, ta, mu
+; CHECK-NEXT:    vmv1r.v v20, v0
 ; CHECK-NEXT:    vmsleu.vv v0, v12, v8
 ; CHECK-NEXT:    vmsleu.vv v20, v16, v12, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v20
@@ -595,8 +595,8 @@ declare <vscale x 1 x i1> @llvm.riscv.vmsgeu.mask.nxv1i32(
 define <vscale x 1 x i1> @intrinsic_vmsgeu_mask_vv_nxv1i32_nxv1i32(<vscale x 1 x i1> %0, <vscale x 1 x i32> %1, <vscale x 1 x i32> %2, <vscale x 1 x i32> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsgeu_mask_vv_nxv1i32_nxv1i32:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e32, mf2, ta, mu
+; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmsleu.vv v0, v9, v8
 ; CHECK-NEXT:    vmsleu.vv v11, v10, v9, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v11
@@ -646,8 +646,8 @@ declare <vscale x 2 x i1> @llvm.riscv.vmsgeu.mask.nxv2i32(
 define <vscale x 2 x i1> @intrinsic_vmsgeu_mask_vv_nxv2i32_nxv2i32(<vscale x 2 x i1> %0, <vscale x 2 x i32> %1, <vscale x 2 x i32> %2, <vscale x 2 x i32> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsgeu_mask_vv_nxv2i32_nxv2i32:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e32, m1, ta, mu
+; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmsleu.vv v0, v9, v8
 ; CHECK-NEXT:    vmsleu.vv v11, v10, v9, v0.t
 ; CHECK-NEXT:    vmv.v.v v0, v11
@@ -697,8 +697,8 @@ declare <vscale x 4 x i1> @llvm.riscv.vmsgeu.mask.nxv4i32(
 define <vscale x 4 x i1> @intrinsic_vmsgeu_mask_vv_nxv4i32_nxv4i32(<vscale x 4 x i1> %0, <vscale x 4 x i32> %1, <vscale x 4 x i32> %2, <vscale x 4 x i32> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsgeu_mask_vv_nxv4i32_nxv4i32:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v14, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e32, m2, ta, mu
+; CHECK-NEXT:    vmv1r.v v14, v0
 ; CHECK-NEXT:    vmsleu.vv v0, v10, v8
 ; CHECK-NEXT:    vmsleu.vv v14, v12, v10, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v14
@@ -748,8 +748,8 @@ declare <vscale x 8 x i1> @llvm.riscv.vmsgeu.mask.nxv8i32(
 define <vscale x 8 x i1> @intrinsic_vmsgeu_mask_vv_nxv8i32_nxv8i32(<vscale x 8 x i1> %0, <vscale x 8 x i32> %1, <vscale x 8 x i32> %2, <vscale x 8 x i32> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsgeu_mask_vv_nxv8i32_nxv8i32:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v20, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e32, m4, ta, mu
+; CHECK-NEXT:    vmv1r.v v20, v0
 ; CHECK-NEXT:    vmsleu.vv v0, v12, v8
 ; CHECK-NEXT:    vmsleu.vv v20, v16, v12, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v20
@@ -799,8 +799,8 @@ declare <vscale x 1 x i1> @llvm.riscv.vmsgeu.mask.nxv1i64(
 define <vscale x 1 x i1> @intrinsic_vmsgeu_mask_vv_nxv1i64_nxv1i64(<vscale x 1 x i1> %0, <vscale x 1 x i64> %1, <vscale x 1 x i64> %2, <vscale x 1 x i64> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsgeu_mask_vv_nxv1i64_nxv1i64:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e64, m1, ta, mu
+; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmsleu.vv v0, v9, v8
 ; CHECK-NEXT:    vmsleu.vv v11, v10, v9, v0.t
 ; CHECK-NEXT:    vmv.v.v v0, v11
@@ -850,8 +850,8 @@ declare <vscale x 2 x i1> @llvm.riscv.vmsgeu.mask.nxv2i64(
 define <vscale x 2 x i1> @intrinsic_vmsgeu_mask_vv_nxv2i64_nxv2i64(<vscale x 2 x i1> %0, <vscale x 2 x i64> %1, <vscale x 2 x i64> %2, <vscale x 2 x i64> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsgeu_mask_vv_nxv2i64_nxv2i64:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v14, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e64, m2, ta, mu
+; CHECK-NEXT:    vmv1r.v v14, v0
 ; CHECK-NEXT:    vmsleu.vv v0, v10, v8
 ; CHECK-NEXT:    vmsleu.vv v14, v12, v10, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v14
@@ -901,8 +901,8 @@ declare <vscale x 4 x i1> @llvm.riscv.vmsgeu.mask.nxv4i64(
 define <vscale x 4 x i1> @intrinsic_vmsgeu_mask_vv_nxv4i64_nxv4i64(<vscale x 4 x i1> %0, <vscale x 4 x i64> %1, <vscale x 4 x i64> %2, <vscale x 4 x i64> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsgeu_mask_vv_nxv4i64_nxv4i64:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v20, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e64, m4, ta, mu
+; CHECK-NEXT:    vmv1r.v v20, v0
 ; CHECK-NEXT:    vmsleu.vv v0, v12, v8
 ; CHECK-NEXT:    vmsleu.vv v20, v16, v12, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v20
@@ -953,9 +953,9 @@ declare <vscale x 1 x i1> @llvm.riscv.vmsgeu.mask.nxv1i8.i8(
 define <vscale x 1 x i1> @intrinsic_vmsgeu_mask_vx_nxv1i8_i8(<vscale x 1 x i1> %0, <vscale x 1 x i8> %1, i8 %2, <vscale x 1 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsgeu_mask_vx_nxv1i8_i8:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e8, mf8, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a1, e8, mf8, ta, mu
 ; CHECK-NEXT:    vmsltu.vx v10, v8, a0, v0.t
 ; CHECK-NEXT:    vmxor.mm v0, v10, v9
 ; CHECK-NEXT:    ret
@@ -1001,9 +1001,9 @@ declare <vscale x 2 x i1> @llvm.riscv.vmsgeu.mask.nxv2i8.i8(
 define <vscale x 2 x i1> @intrinsic_vmsgeu_mask_vx_nxv2i8_i8(<vscale x 2 x i1> %0, <vscale x 2 x i8> %1, i8 %2, <vscale x 2 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsgeu_mask_vx_nxv2i8_i8:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e8, mf4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a1, e8, mf4, ta, mu
 ; CHECK-NEXT:    vmsltu.vx v10, v8, a0, v0.t
 ; CHECK-NEXT:    vmxor.mm v0, v10, v9
 ; CHECK-NEXT:    ret
@@ -1049,9 +1049,9 @@ declare <vscale x 4 x i1> @llvm.riscv.vmsgeu.mask.nxv4i8.i8(
 define <vscale x 4 x i1> @intrinsic_vmsgeu_mask_vx_nxv4i8_i8(<vscale x 4 x i1> %0, <vscale x 4 x i8> %1, i8 %2, <vscale x 4 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsgeu_mask_vx_nxv4i8_i8:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e8, mf2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a1, e8, mf2, ta, mu
 ; CHECK-NEXT:    vmsltu.vx v10, v8, a0, v0.t
 ; CHECK-NEXT:    vmxor.mm v0, v10, v9
 ; CHECK-NEXT:    ret
@@ -1097,9 +1097,9 @@ declare <vscale x 8 x i1> @llvm.riscv.vmsgeu.mask.nxv8i8.i8(
 define <vscale x 8 x i1> @intrinsic_vmsgeu_mask_vx_nxv8i8_i8(<vscale x 8 x i1> %0, <vscale x 8 x i8> %1, i8 %2, <vscale x 8 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsgeu_mask_vx_nxv8i8_i8:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e8, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a1, e8, m1, ta, mu
 ; CHECK-NEXT:    vmsltu.vx v10, v8, a0, v0.t
 ; CHECK-NEXT:    vmxor.mm v0, v10, v9
 ; CHECK-NEXT:    ret
@@ -1145,9 +1145,9 @@ declare <vscale x 16 x i1> @llvm.riscv.vmsgeu.mask.nxv16i8.i8(
 define <vscale x 16 x i1> @intrinsic_vmsgeu_mask_vx_nxv16i8_i8(<vscale x 16 x i1> %0, <vscale x 16 x i8> %1, i8 %2, <vscale x 16 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsgeu_mask_vx_nxv16i8_i8:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e8, m2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmv1r.v v0, v10
-; CHECK-NEXT:    vsetvli zero, a1, e8, m2, ta, mu
 ; CHECK-NEXT:    vmsltu.vx v11, v8, a0, v0.t
 ; CHECK-NEXT:    vmxor.mm v0, v11, v10
 ; CHECK-NEXT:    ret
@@ -1193,9 +1193,9 @@ declare <vscale x 32 x i1> @llvm.riscv.vmsgeu.mask.nxv32i8.i8(
 define <vscale x 32 x i1> @intrinsic_vmsgeu_mask_vx_nxv32i8_i8(<vscale x 32 x i1> %0, <vscale x 32 x i8> %1, i8 %2, <vscale x 32 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsgeu_mask_vx_nxv32i8_i8:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e8, m4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v13, v0
 ; CHECK-NEXT:    vmv1r.v v0, v12
-; CHECK-NEXT:    vsetvli zero, a1, e8, m4, ta, mu
 ; CHECK-NEXT:    vmsltu.vx v13, v8, a0, v0.t
 ; CHECK-NEXT:    vmxor.mm v0, v13, v12
 ; CHECK-NEXT:    ret
@@ -1241,9 +1241,9 @@ declare <vscale x 1 x i1> @llvm.riscv.vmsgeu.mask.nxv1i16.i16(
 define <vscale x 1 x i1> @intrinsic_vmsgeu_mask_vx_nxv1i16_i16(<vscale x 1 x i1> %0, <vscale x 1 x i16> %1, i16 %2, <vscale x 1 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsgeu_mask_vx_nxv1i16_i16:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, mf4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a1, e16, mf4, ta, mu
 ; CHECK-NEXT:    vmsltu.vx v10, v8, a0, v0.t
 ; CHECK-NEXT:    vmxor.mm v0, v10, v9
 ; CHECK-NEXT:    ret
@@ -1289,9 +1289,9 @@ declare <vscale x 2 x i1> @llvm.riscv.vmsgeu.mask.nxv2i16.i16(
 define <vscale x 2 x i1> @intrinsic_vmsgeu_mask_vx_nxv2i16_i16(<vscale x 2 x i1> %0, <vscale x 2 x i16> %1, i16 %2, <vscale x 2 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsgeu_mask_vx_nxv2i16_i16:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, mf2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a1, e16, mf2, ta, mu
 ; CHECK-NEXT:    vmsltu.vx v10, v8, a0, v0.t
 ; CHECK-NEXT:    vmxor.mm v0, v10, v9
 ; CHECK-NEXT:    ret
@@ -1337,9 +1337,9 @@ declare <vscale x 4 x i1> @llvm.riscv.vmsgeu.mask.nxv4i16.i16(
 define <vscale x 4 x i1> @intrinsic_vmsgeu_mask_vx_nxv4i16_i16(<vscale x 4 x i1> %0, <vscale x 4 x i16> %1, i16 %2, <vscale x 4 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsgeu_mask_vx_nxv4i16_i16:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a1, e16, m1, ta, mu
 ; CHECK-NEXT:    vmsltu.vx v10, v8, a0, v0.t
 ; CHECK-NEXT:    vmxor.mm v0, v10, v9
 ; CHECK-NEXT:    ret
@@ -1385,9 +1385,9 @@ declare <vscale x 8 x i1> @llvm.riscv.vmsgeu.mask.nxv8i16.i16(
 define <vscale x 8 x i1> @intrinsic_vmsgeu_mask_vx_nxv8i16_i16(<vscale x 8 x i1> %0, <vscale x 8 x i16> %1, i16 %2, <vscale x 8 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsgeu_mask_vx_nxv8i16_i16:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, m2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmv1r.v v0, v10
-; CHECK-NEXT:    vsetvli zero, a1, e16, m2, ta, mu
 ; CHECK-NEXT:    vmsltu.vx v11, v8, a0, v0.t
 ; CHECK-NEXT:    vmxor.mm v0, v11, v10
 ; CHECK-NEXT:    ret
@@ -1433,9 +1433,9 @@ declare <vscale x 16 x i1> @llvm.riscv.vmsgeu.mask.nxv16i16.i16(
 define <vscale x 16 x i1> @intrinsic_vmsgeu_mask_vx_nxv16i16_i16(<vscale x 16 x i1> %0, <vscale x 16 x i16> %1, i16 %2, <vscale x 16 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsgeu_mask_vx_nxv16i16_i16:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, m4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v13, v0
 ; CHECK-NEXT:    vmv1r.v v0, v12
-; CHECK-NEXT:    vsetvli zero, a1, e16, m4, ta, mu
 ; CHECK-NEXT:    vmsltu.vx v13, v8, a0, v0.t
 ; CHECK-NEXT:    vmxor.mm v0, v13, v12
 ; CHECK-NEXT:    ret
@@ -1481,9 +1481,9 @@ declare <vscale x 1 x i1> @llvm.riscv.vmsgeu.mask.nxv1i32.i32(
 define <vscale x 1 x i1> @intrinsic_vmsgeu_mask_vx_nxv1i32_i32(<vscale x 1 x i1> %0, <vscale x 1 x i32> %1, i32 %2, <vscale x 1 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsgeu_mask_vx_nxv1i32_i32:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e32, mf2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a1, e32, mf2, ta, mu
 ; CHECK-NEXT:    vmsltu.vx v10, v8, a0, v0.t
 ; CHECK-NEXT:    vmxor.mm v0, v10, v9
 ; CHECK-NEXT:    ret
@@ -1529,9 +1529,9 @@ declare <vscale x 2 x i1> @llvm.riscv.vmsgeu.mask.nxv2i32.i32(
 define <vscale x 2 x i1> @intrinsic_vmsgeu_mask_vx_nxv2i32_i32(<vscale x 2 x i1> %0, <vscale x 2 x i32> %1, i32 %2, <vscale x 2 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsgeu_mask_vx_nxv2i32_i32:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e32, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a1, e32, m1, ta, mu
 ; CHECK-NEXT:    vmsltu.vx v10, v8, a0, v0.t
 ; CHECK-NEXT:    vmxor.mm v0, v10, v9
 ; CHECK-NEXT:    ret
@@ -1577,9 +1577,9 @@ declare <vscale x 4 x i1> @llvm.riscv.vmsgeu.mask.nxv4i32.i32(
 define <vscale x 4 x i1> @intrinsic_vmsgeu_mask_vx_nxv4i32_i32(<vscale x 4 x i1> %0, <vscale x 4 x i32> %1, i32 %2, <vscale x 4 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsgeu_mask_vx_nxv4i32_i32:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e32, m2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmv1r.v v0, v10
-; CHECK-NEXT:    vsetvli zero, a1, e32, m2, ta, mu
 ; CHECK-NEXT:    vmsltu.vx v11, v8, a0, v0.t
 ; CHECK-NEXT:    vmxor.mm v0, v11, v10
 ; CHECK-NEXT:    ret
@@ -1625,9 +1625,9 @@ declare <vscale x 8 x i1> @llvm.riscv.vmsgeu.mask.nxv8i32.i32(
 define <vscale x 8 x i1> @intrinsic_vmsgeu_mask_vx_nxv8i32_i32(<vscale x 8 x i1> %0, <vscale x 8 x i32> %1, i32 %2, <vscale x 8 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsgeu_mask_vx_nxv8i32_i32:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e32, m4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v13, v0
 ; CHECK-NEXT:    vmv1r.v v0, v12
-; CHECK-NEXT:    vsetvli zero, a1, e32, m4, ta, mu
 ; CHECK-NEXT:    vmsltu.vx v13, v8, a0, v0.t
 ; CHECK-NEXT:    vmxor.mm v0, v13, v12
 ; CHECK-NEXT:    ret
@@ -1700,9 +1700,9 @@ define <vscale x 1 x i1> @intrinsic_vmsgeu_mask_vx_nxv1i64_i64(<vscale x 1 x i1>
 ;
 ; RV64-LABEL: intrinsic_vmsgeu_mask_vx_nxv1i64_i64:
 ; RV64:       # %bb.0: # %entry
+; RV64-NEXT:    vsetvli zero, a1, e64, m1, ta, mu
 ; RV64-NEXT:    vmv1r.v v10, v0
 ; RV64-NEXT:    vmv1r.v v0, v9
-; RV64-NEXT:    vsetvli zero, a1, e64, m1, ta, mu
 ; RV64-NEXT:    vmsltu.vx v10, v8, a0, v0.t
 ; RV64-NEXT:    vmxor.mm v0, v10, v9
 ; RV64-NEXT:    ret
@@ -1775,9 +1775,9 @@ define <vscale x 2 x i1> @intrinsic_vmsgeu_mask_vx_nxv2i64_i64(<vscale x 2 x i1>
 ;
 ; RV64-LABEL: intrinsic_vmsgeu_mask_vx_nxv2i64_i64:
 ; RV64:       # %bb.0: # %entry
+; RV64-NEXT:    vsetvli zero, a1, e64, m2, ta, mu
 ; RV64-NEXT:    vmv1r.v v11, v0
 ; RV64-NEXT:    vmv1r.v v0, v10
-; RV64-NEXT:    vsetvli zero, a1, e64, m2, ta, mu
 ; RV64-NEXT:    vmsltu.vx v11, v8, a0, v0.t
 ; RV64-NEXT:    vmxor.mm v0, v11, v10
 ; RV64-NEXT:    ret
@@ -1850,9 +1850,9 @@ define <vscale x 4 x i1> @intrinsic_vmsgeu_mask_vx_nxv4i64_i64(<vscale x 4 x i1>
 ;
 ; RV64-LABEL: intrinsic_vmsgeu_mask_vx_nxv4i64_i64:
 ; RV64:       # %bb.0: # %entry
+; RV64-NEXT:    vsetvli zero, a1, e64, m4, ta, mu
 ; RV64-NEXT:    vmv1r.v v13, v0
 ; RV64-NEXT:    vmv1r.v v0, v12
-; RV64-NEXT:    vsetvli zero, a1, e64, m4, ta, mu
 ; RV64-NEXT:    vmsltu.vx v13, v8, a0, v0.t
 ; RV64-NEXT:    vmxor.mm v0, v13, v12
 ; RV64-NEXT:    ret
@@ -1885,9 +1885,9 @@ entry:
 define <vscale x 1 x i1> @intrinsic_vmsgeu_mask_vi_nxv1i8_i8(<vscale x 1 x i1> %0, <vscale x 1 x i8> %1, <vscale x 1 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsgeu_mask_vi_nxv1i8_i8:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e8, mf8, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a0, e8, mf8, ta, mu
 ; CHECK-NEXT:    vmsgtu.vi v10, v8, -15, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v10
 ; CHECK-NEXT:    ret
@@ -1920,9 +1920,9 @@ entry:
 define <vscale x 2 x i1> @intrinsic_vmsgeu_mask_vi_nxv2i8_i8(<vscale x 2 x i1> %0, <vscale x 2 x i8> %1, <vscale x 2 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsgeu_mask_vi_nxv2i8_i8:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e8, mf4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a0, e8, mf4, ta, mu
 ; CHECK-NEXT:    vmsgtu.vi v10, v8, -13, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v10
 ; CHECK-NEXT:    ret
@@ -1955,9 +1955,9 @@ entry:
 define <vscale x 4 x i1> @intrinsic_vmsgeu_mask_vi_nxv4i8_i8(<vscale x 4 x i1> %0, <vscale x 4 x i8> %1, <vscale x 4 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsgeu_mask_vi_nxv4i8_i8:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e8, mf2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a0, e8, mf2, ta, mu
 ; CHECK-NEXT:    vmsgtu.vi v10, v8, -11, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v10
 ; CHECK-NEXT:    ret
@@ -1975,11 +1975,11 @@ entry:
 define <vscale x 4 x i1> @intrinsic_vmsgeu_mask_vi_nxv4i8_i8_1(<vscale x 4 x i1> %0, <vscale x 4 x i8> %1, <vscale x 4 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsgeu_mask_vi_nxv4i8_i8_1:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e8, mf2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
-; CHECK-NEXT:    li a1, 99
+; CHECK-NEXT:    li a0, 99
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a0, e8, mf2, ta, mu
-; CHECK-NEXT:    vmsgtu.vx v10, v8, a1, v0.t
+; CHECK-NEXT:    vmsgtu.vx v10, v8, a0, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v10
 ; CHECK-NEXT:    ret
 entry:
@@ -2011,9 +2011,9 @@ entry:
 define <vscale x 8 x i1> @intrinsic_vmsgeu_mask_vi_nxv8i8_i8(<vscale x 8 x i1> %0, <vscale x 8 x i8> %1, <vscale x 8 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsgeu_mask_vi_nxv8i8_i8:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e8, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a0, e8, m1, ta, mu
 ; CHECK-NEXT:    vmsgtu.vi v10, v8, -9, v0.t
 ; CHECK-NEXT:    vmv.v.v v0, v10
 ; CHECK-NEXT:    ret
@@ -2046,9 +2046,9 @@ entry:
 define <vscale x 16 x i1> @intrinsic_vmsgeu_mask_vi_nxv16i8_i8(<vscale x 16 x i1> %0, <vscale x 16 x i8> %1, <vscale x 16 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsgeu_mask_vi_nxv16i8_i8:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e8, m2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmv1r.v v0, v10
-; CHECK-NEXT:    vsetvli zero, a0, e8, m2, ta, mu
 ; CHECK-NEXT:    vmsgtu.vi v11, v8, -7, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v11
 ; CHECK-NEXT:    ret
@@ -2081,9 +2081,9 @@ entry:
 define <vscale x 32 x i1> @intrinsic_vmsgeu_mask_vi_nxv32i8_i8(<vscale x 32 x i1> %0, <vscale x 32 x i8> %1, <vscale x 32 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsgeu_mask_vi_nxv32i8_i8:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e8, m4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v13, v0
 ; CHECK-NEXT:    vmv1r.v v0, v12
-; CHECK-NEXT:    vsetvli zero, a0, e8, m4, ta, mu
 ; CHECK-NEXT:    vmsgtu.vi v13, v8, -5, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v13
 ; CHECK-NEXT:    ret
@@ -2116,9 +2116,9 @@ entry:
 define <vscale x 1 x i1> @intrinsic_vmsgeu_mask_vi_nxv1i16_i16(<vscale x 1 x i1> %0, <vscale x 1 x i16> %1, <vscale x 1 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsgeu_mask_vi_nxv1i16_i16:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e16, mf4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a0, e16, mf4, ta, mu
 ; CHECK-NEXT:    vmsgtu.vi v10, v8, -3, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v10
 ; CHECK-NEXT:    ret
@@ -2214,9 +2214,9 @@ entry:
 define <vscale x 4 x i1> @intrinsic_vmsgeu_mask_vi_nxv4i16_i16(<vscale x 4 x i1> %0, <vscale x 4 x i16> %1, <vscale x 4 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsgeu_mask_vi_nxv4i16_i16:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e16, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a0, e16, m1, ta, mu
 ; CHECK-NEXT:    vmsgtu.vi v10, v8, 0, v0.t
 ; CHECK-NEXT:    vmv.v.v v0, v10
 ; CHECK-NEXT:    ret
@@ -2249,9 +2249,9 @@ entry:
 define <vscale x 8 x i1> @intrinsic_vmsgeu_mask_vi_nxv8i16_i16(<vscale x 8 x i1> %0, <vscale x 8 x i16> %1, <vscale x 8 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsgeu_mask_vi_nxv8i16_i16:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e16, m2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmv1r.v v0, v10
-; CHECK-NEXT:    vsetvli zero, a0, e16, m2, ta, mu
 ; CHECK-NEXT:    vmsgtu.vi v11, v8, 2, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v11
 ; CHECK-NEXT:    ret
@@ -2284,9 +2284,9 @@ entry:
 define <vscale x 16 x i1> @intrinsic_vmsgeu_mask_vi_nxv16i16_i16(<vscale x 16 x i1> %0, <vscale x 16 x i16> %1, <vscale x 16 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsgeu_mask_vi_nxv16i16_i16:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e16, m4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v13, v0
 ; CHECK-NEXT:    vmv1r.v v0, v12
-; CHECK-NEXT:    vsetvli zero, a0, e16, m4, ta, mu
 ; CHECK-NEXT:    vmsgtu.vi v13, v8, 4, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v13
 ; CHECK-NEXT:    ret
@@ -2319,9 +2319,9 @@ entry:
 define <vscale x 1 x i1> @intrinsic_vmsgeu_mask_vi_nxv1i32_i32(<vscale x 1 x i1> %0, <vscale x 1 x i32> %1, <vscale x 1 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsgeu_mask_vi_nxv1i32_i32:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e32, mf2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a0, e32, mf2, ta, mu
 ; CHECK-NEXT:    vmsgtu.vi v10, v8, 6, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v10
 ; CHECK-NEXT:    ret
@@ -2354,9 +2354,9 @@ entry:
 define <vscale x 2 x i1> @intrinsic_vmsgeu_mask_vi_nxv2i32_i32(<vscale x 2 x i1> %0, <vscale x 2 x i32> %1, <vscale x 2 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsgeu_mask_vi_nxv2i32_i32:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e32, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a0, e32, m1, ta, mu
 ; CHECK-NEXT:    vmsgtu.vi v10, v8, 8, v0.t
 ; CHECK-NEXT:    vmv.v.v v0, v10
 ; CHECK-NEXT:    ret
@@ -2389,9 +2389,9 @@ entry:
 define <vscale x 4 x i1> @intrinsic_vmsgeu_mask_vi_nxv4i32_i32(<vscale x 4 x i1> %0, <vscale x 4 x i32> %1, <vscale x 4 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsgeu_mask_vi_nxv4i32_i32:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e32, m2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmv1r.v v0, v10
-; CHECK-NEXT:    vsetvli zero, a0, e32, m2, ta, mu
 ; CHECK-NEXT:    vmsgtu.vi v11, v8, 10, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v11
 ; CHECK-NEXT:    ret
@@ -2424,9 +2424,9 @@ entry:
 define <vscale x 8 x i1> @intrinsic_vmsgeu_mask_vi_nxv8i32_i32(<vscale x 8 x i1> %0, <vscale x 8 x i32> %1, <vscale x 8 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsgeu_mask_vi_nxv8i32_i32:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e32, m4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v13, v0
 ; CHECK-NEXT:    vmv1r.v v0, v12
-; CHECK-NEXT:    vsetvli zero, a0, e32, m4, ta, mu
 ; CHECK-NEXT:    vmsgtu.vi v13, v8, 12, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v13
 ; CHECK-NEXT:    ret
@@ -2459,9 +2459,9 @@ entry:
 define <vscale x 1 x i1> @intrinsic_vmsgeu_mask_vi_nxv1i64_i64(<vscale x 1 x i1> %0, <vscale x 1 x i64> %1, <vscale x 1 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsgeu_mask_vi_nxv1i64_i64:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e64, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a0, e64, m1, ta, mu
 ; CHECK-NEXT:    vmsgtu.vi v10, v8, 14, v0.t
 ; CHECK-NEXT:    vmv.v.v v0, v10
 ; CHECK-NEXT:    ret
@@ -2494,9 +2494,9 @@ entry:
 define <vscale x 2 x i1> @intrinsic_vmsgeu_mask_vi_nxv2i64_i64(<vscale x 2 x i1> %0, <vscale x 2 x i64> %1, <vscale x 2 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsgeu_mask_vi_nxv2i64_i64:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e64, m2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmv1r.v v0, v10
-; CHECK-NEXT:    vsetvli zero, a0, e64, m2, ta, mu
 ; CHECK-NEXT:    vmsgtu.vi v11, v8, -16, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v11
 ; CHECK-NEXT:    ret
@@ -2529,9 +2529,9 @@ entry:
 define <vscale x 4 x i1> @intrinsic_vmsgeu_mask_vi_nxv4i64_i64(<vscale x 4 x i1> %0, <vscale x 4 x i64> %1, <vscale x 4 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsgeu_mask_vi_nxv4i64_i64:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e64, m4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v13, v0
 ; CHECK-NEXT:    vmv1r.v v0, v12
-; CHECK-NEXT:    vsetvli zero, a0, e64, m4, ta, mu
 ; CHECK-NEXT:    vmsgtu.vi v13, v8, -14, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v13
 ; CHECK-NEXT:    ret
diff --git a/llvm/test/CodeGen/RISCV/rvv/vmsgt.ll b/llvm/test/CodeGen/RISCV/rvv/vmsgt.ll
index d7dee2e1bc580..62ac44bfdf38c 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vmsgt.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vmsgt.ll
@@ -34,8 +34,8 @@ declare <vscale x 1 x i1> @llvm.riscv.vmsgt.mask.nxv1i8(
 define <vscale x 1 x i1> @intrinsic_vmsgt_mask_vv_nxv1i8_nxv1i8(<vscale x 1 x i1> %0, <vscale x 1 x i8> %1, <vscale x 1 x i8> %2, <vscale x 1 x i8> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsgt_mask_vv_nxv1i8_nxv1i8:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e8, mf8, ta, mu
+; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmslt.vv v0, v9, v8
 ; CHECK-NEXT:    vmslt.vv v11, v10, v9, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v11
@@ -85,8 +85,8 @@ declare <vscale x 2 x i1> @llvm.riscv.vmsgt.mask.nxv2i8(
 define <vscale x 2 x i1> @intrinsic_vmsgt_mask_vv_nxv2i8_nxv2i8(<vscale x 2 x i1> %0, <vscale x 2 x i8> %1, <vscale x 2 x i8> %2, <vscale x 2 x i8> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsgt_mask_vv_nxv2i8_nxv2i8:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e8, mf4, ta, mu
+; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmslt.vv v0, v9, v8
 ; CHECK-NEXT:    vmslt.vv v11, v10, v9, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v11
@@ -136,8 +136,8 @@ declare <vscale x 4 x i1> @llvm.riscv.vmsgt.mask.nxv4i8(
 define <vscale x 4 x i1> @intrinsic_vmsgt_mask_vv_nxv4i8_nxv4i8(<vscale x 4 x i1> %0, <vscale x 4 x i8> %1, <vscale x 4 x i8> %2, <vscale x 4 x i8> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsgt_mask_vv_nxv4i8_nxv4i8:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e8, mf2, ta, mu
+; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmslt.vv v0, v9, v8
 ; CHECK-NEXT:    vmslt.vv v11, v10, v9, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v11
@@ -187,8 +187,8 @@ declare <vscale x 8 x i1> @llvm.riscv.vmsgt.mask.nxv8i8(
 define <vscale x 8 x i1> @intrinsic_vmsgt_mask_vv_nxv8i8_nxv8i8(<vscale x 8 x i1> %0, <vscale x 8 x i8> %1, <vscale x 8 x i8> %2, <vscale x 8 x i8> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsgt_mask_vv_nxv8i8_nxv8i8:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e8, m1, ta, mu
+; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmslt.vv v0, v9, v8
 ; CHECK-NEXT:    vmslt.vv v11, v10, v9, v0.t
 ; CHECK-NEXT:    vmv.v.v v0, v11
@@ -238,8 +238,8 @@ declare <vscale x 16 x i1> @llvm.riscv.vmsgt.mask.nxv16i8(
 define <vscale x 16 x i1> @intrinsic_vmsgt_mask_vv_nxv16i8_nxv16i8(<vscale x 16 x i1> %0, <vscale x 16 x i8> %1, <vscale x 16 x i8> %2, <vscale x 16 x i8> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsgt_mask_vv_nxv16i8_nxv16i8:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v14, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e8, m2, ta, mu
+; CHECK-NEXT:    vmv1r.v v14, v0
 ; CHECK-NEXT:    vmslt.vv v0, v10, v8
 ; CHECK-NEXT:    vmslt.vv v14, v12, v10, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v14
@@ -289,8 +289,8 @@ declare <vscale x 32 x i1> @llvm.riscv.vmsgt.mask.nxv32i8(
 define <vscale x 32 x i1> @intrinsic_vmsgt_mask_vv_nxv32i8_nxv32i8(<vscale x 32 x i1> %0, <vscale x 32 x i8> %1, <vscale x 32 x i8> %2, <vscale x 32 x i8> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsgt_mask_vv_nxv32i8_nxv32i8:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v20, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e8, m4, ta, mu
+; CHECK-NEXT:    vmv1r.v v20, v0
 ; CHECK-NEXT:    vmslt.vv v0, v12, v8
 ; CHECK-NEXT:    vmslt.vv v20, v16, v12, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v20
@@ -340,8 +340,8 @@ declare <vscale x 1 x i1> @llvm.riscv.vmsgt.mask.nxv1i16(
 define <vscale x 1 x i1> @intrinsic_vmsgt_mask_vv_nxv1i16_nxv1i16(<vscale x 1 x i1> %0, <vscale x 1 x i16> %1, <vscale x 1 x i16> %2, <vscale x 1 x i16> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsgt_mask_vv_nxv1i16_nxv1i16:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e16, mf4, ta, mu
+; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmslt.vv v0, v9, v8
 ; CHECK-NEXT:    vmslt.vv v11, v10, v9, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v11
@@ -391,8 +391,8 @@ declare <vscale x 2 x i1> @llvm.riscv.vmsgt.mask.nxv2i16(
 define <vscale x 2 x i1> @intrinsic_vmsgt_mask_vv_nxv2i16_nxv2i16(<vscale x 2 x i1> %0, <vscale x 2 x i16> %1, <vscale x 2 x i16> %2, <vscale x 2 x i16> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsgt_mask_vv_nxv2i16_nxv2i16:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e16, mf2, ta, mu
+; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmslt.vv v0, v9, v8
 ; CHECK-NEXT:    vmslt.vv v11, v10, v9, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v11
@@ -442,8 +442,8 @@ declare <vscale x 4 x i1> @llvm.riscv.vmsgt.mask.nxv4i16(
 define <vscale x 4 x i1> @intrinsic_vmsgt_mask_vv_nxv4i16_nxv4i16(<vscale x 4 x i1> %0, <vscale x 4 x i16> %1, <vscale x 4 x i16> %2, <vscale x 4 x i16> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsgt_mask_vv_nxv4i16_nxv4i16:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e16, m1, ta, mu
+; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmslt.vv v0, v9, v8
 ; CHECK-NEXT:    vmslt.vv v11, v10, v9, v0.t
 ; CHECK-NEXT:    vmv.v.v v0, v11
@@ -493,8 +493,8 @@ declare <vscale x 8 x i1> @llvm.riscv.vmsgt.mask.nxv8i16(
 define <vscale x 8 x i1> @intrinsic_vmsgt_mask_vv_nxv8i16_nxv8i16(<vscale x 8 x i1> %0, <vscale x 8 x i16> %1, <vscale x 8 x i16> %2, <vscale x 8 x i16> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsgt_mask_vv_nxv8i16_nxv8i16:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v14, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e16, m2, ta, mu
+; CHECK-NEXT:    vmv1r.v v14, v0
 ; CHECK-NEXT:    vmslt.vv v0, v10, v8
 ; CHECK-NEXT:    vmslt.vv v14, v12, v10, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v14
@@ -544,8 +544,8 @@ declare <vscale x 16 x i1> @llvm.riscv.vmsgt.mask.nxv16i16(
 define <vscale x 16 x i1> @intrinsic_vmsgt_mask_vv_nxv16i16_nxv16i16(<vscale x 16 x i1> %0, <vscale x 16 x i16> %1, <vscale x 16 x i16> %2, <vscale x 16 x i16> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsgt_mask_vv_nxv16i16_nxv16i16:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v20, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e16, m4, ta, mu
+; CHECK-NEXT:    vmv1r.v v20, v0
 ; CHECK-NEXT:    vmslt.vv v0, v12, v8
 ; CHECK-NEXT:    vmslt.vv v20, v16, v12, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v20
@@ -595,8 +595,8 @@ declare <vscale x 1 x i1> @llvm.riscv.vmsgt.mask.nxv1i32(
 define <vscale x 1 x i1> @intrinsic_vmsgt_mask_vv_nxv1i32_nxv1i32(<vscale x 1 x i1> %0, <vscale x 1 x i32> %1, <vscale x 1 x i32> %2, <vscale x 1 x i32> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsgt_mask_vv_nxv1i32_nxv1i32:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e32, mf2, ta, mu
+; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmslt.vv v0, v9, v8
 ; CHECK-NEXT:    vmslt.vv v11, v10, v9, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v11
@@ -646,8 +646,8 @@ declare <vscale x 2 x i1> @llvm.riscv.vmsgt.mask.nxv2i32(
 define <vscale x 2 x i1> @intrinsic_vmsgt_mask_vv_nxv2i32_nxv2i32(<vscale x 2 x i1> %0, <vscale x 2 x i32> %1, <vscale x 2 x i32> %2, <vscale x 2 x i32> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsgt_mask_vv_nxv2i32_nxv2i32:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e32, m1, ta, mu
+; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmslt.vv v0, v9, v8
 ; CHECK-NEXT:    vmslt.vv v11, v10, v9, v0.t
 ; CHECK-NEXT:    vmv.v.v v0, v11
@@ -697,8 +697,8 @@ declare <vscale x 4 x i1> @llvm.riscv.vmsgt.mask.nxv4i32(
 define <vscale x 4 x i1> @intrinsic_vmsgt_mask_vv_nxv4i32_nxv4i32(<vscale x 4 x i1> %0, <vscale x 4 x i32> %1, <vscale x 4 x i32> %2, <vscale x 4 x i32> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsgt_mask_vv_nxv4i32_nxv4i32:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v14, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e32, m2, ta, mu
+; CHECK-NEXT:    vmv1r.v v14, v0
 ; CHECK-NEXT:    vmslt.vv v0, v10, v8
 ; CHECK-NEXT:    vmslt.vv v14, v12, v10, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v14
@@ -748,8 +748,8 @@ declare <vscale x 8 x i1> @llvm.riscv.vmsgt.mask.nxv8i32(
 define <vscale x 8 x i1> @intrinsic_vmsgt_mask_vv_nxv8i32_nxv8i32(<vscale x 8 x i1> %0, <vscale x 8 x i32> %1, <vscale x 8 x i32> %2, <vscale x 8 x i32> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsgt_mask_vv_nxv8i32_nxv8i32:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v20, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e32, m4, ta, mu
+; CHECK-NEXT:    vmv1r.v v20, v0
 ; CHECK-NEXT:    vmslt.vv v0, v12, v8
 ; CHECK-NEXT:    vmslt.vv v20, v16, v12, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v20
@@ -799,8 +799,8 @@ declare <vscale x 1 x i1> @llvm.riscv.vmsgt.mask.nxv1i64(
 define <vscale x 1 x i1> @intrinsic_vmsgt_mask_vv_nxv1i64_nxv1i64(<vscale x 1 x i1> %0, <vscale x 1 x i64> %1, <vscale x 1 x i64> %2, <vscale x 1 x i64> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsgt_mask_vv_nxv1i64_nxv1i64:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e64, m1, ta, mu
+; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmslt.vv v0, v9, v8
 ; CHECK-NEXT:    vmslt.vv v11, v10, v9, v0.t
 ; CHECK-NEXT:    vmv.v.v v0, v11
@@ -850,8 +850,8 @@ declare <vscale x 2 x i1> @llvm.riscv.vmsgt.mask.nxv2i64(
 define <vscale x 2 x i1> @intrinsic_vmsgt_mask_vv_nxv2i64_nxv2i64(<vscale x 2 x i1> %0, <vscale x 2 x i64> %1, <vscale x 2 x i64> %2, <vscale x 2 x i64> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsgt_mask_vv_nxv2i64_nxv2i64:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v14, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e64, m2, ta, mu
+; CHECK-NEXT:    vmv1r.v v14, v0
 ; CHECK-NEXT:    vmslt.vv v0, v10, v8
 ; CHECK-NEXT:    vmslt.vv v14, v12, v10, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v14
@@ -901,8 +901,8 @@ declare <vscale x 4 x i1> @llvm.riscv.vmsgt.mask.nxv4i64(
 define <vscale x 4 x i1> @intrinsic_vmsgt_mask_vv_nxv4i64_nxv4i64(<vscale x 4 x i1> %0, <vscale x 4 x i64> %1, <vscale x 4 x i64> %2, <vscale x 4 x i64> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsgt_mask_vv_nxv4i64_nxv4i64:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v20, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e64, m4, ta, mu
+; CHECK-NEXT:    vmv1r.v v20, v0
 ; CHECK-NEXT:    vmslt.vv v0, v12, v8
 ; CHECK-NEXT:    vmslt.vv v20, v16, v12, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v20
@@ -952,9 +952,9 @@ declare <vscale x 1 x i1> @llvm.riscv.vmsgt.mask.nxv1i8.i8(
 define <vscale x 1 x i1> @intrinsic_vmsgt_mask_vx_nxv1i8_i8(<vscale x 1 x i1> %0, <vscale x 1 x i8> %1, i8 %2, <vscale x 1 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsgt_mask_vx_nxv1i8_i8:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e8, mf8, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a1, e8, mf8, ta, mu
 ; CHECK-NEXT:    vmsgt.vx v10, v8, a0, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v10
 ; CHECK-NEXT:    ret
@@ -999,9 +999,9 @@ declare <vscale x 2 x i1> @llvm.riscv.vmsgt.mask.nxv2i8.i8(
 define <vscale x 2 x i1> @intrinsic_vmsgt_mask_vx_nxv2i8_i8(<vscale x 2 x i1> %0, <vscale x 2 x i8> %1, i8 %2, <vscale x 2 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsgt_mask_vx_nxv2i8_i8:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e8, mf4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a1, e8, mf4, ta, mu
 ; CHECK-NEXT:    vmsgt.vx v10, v8, a0, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v10
 ; CHECK-NEXT:    ret
@@ -1046,9 +1046,9 @@ declare <vscale x 4 x i1> @llvm.riscv.vmsgt.mask.nxv4i8.i8(
 define <vscale x 4 x i1> @intrinsic_vmsgt_mask_vx_nxv4i8_i8(<vscale x 4 x i1> %0, <vscale x 4 x i8> %1, i8 %2, <vscale x 4 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsgt_mask_vx_nxv4i8_i8:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e8, mf2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a1, e8, mf2, ta, mu
 ; CHECK-NEXT:    vmsgt.vx v10, v8, a0, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v10
 ; CHECK-NEXT:    ret
@@ -1093,9 +1093,9 @@ declare <vscale x 8 x i1> @llvm.riscv.vmsgt.mask.nxv8i8.i8(
 define <vscale x 8 x i1> @intrinsic_vmsgt_mask_vx_nxv8i8_i8(<vscale x 8 x i1> %0, <vscale x 8 x i8> %1, i8 %2, <vscale x 8 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsgt_mask_vx_nxv8i8_i8:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e8, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a1, e8, m1, ta, mu
 ; CHECK-NEXT:    vmsgt.vx v10, v8, a0, v0.t
 ; CHECK-NEXT:    vmv.v.v v0, v10
 ; CHECK-NEXT:    ret
@@ -1140,9 +1140,9 @@ declare <vscale x 16 x i1> @llvm.riscv.vmsgt.mask.nxv16i8.i8(
 define <vscale x 16 x i1> @intrinsic_vmsgt_mask_vx_nxv16i8_i8(<vscale x 16 x i1> %0, <vscale x 16 x i8> %1, i8 %2, <vscale x 16 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsgt_mask_vx_nxv16i8_i8:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e8, m2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmv1r.v v0, v10
-; CHECK-NEXT:    vsetvli zero, a1, e8, m2, ta, mu
 ; CHECK-NEXT:    vmsgt.vx v11, v8, a0, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v11
 ; CHECK-NEXT:    ret
@@ -1187,9 +1187,9 @@ declare <vscale x 32 x i1> @llvm.riscv.vmsgt.mask.nxv32i8.i8(
 define <vscale x 32 x i1> @intrinsic_vmsgt_mask_vx_nxv32i8_i8(<vscale x 32 x i1> %0, <vscale x 32 x i8> %1, i8 %2, <vscale x 32 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsgt_mask_vx_nxv32i8_i8:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e8, m4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v13, v0
 ; CHECK-NEXT:    vmv1r.v v0, v12
-; CHECK-NEXT:    vsetvli zero, a1, e8, m4, ta, mu
 ; CHECK-NEXT:    vmsgt.vx v13, v8, a0, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v13
 ; CHECK-NEXT:    ret
@@ -1234,9 +1234,9 @@ declare <vscale x 1 x i1> @llvm.riscv.vmsgt.mask.nxv1i16.i16(
 define <vscale x 1 x i1> @intrinsic_vmsgt_mask_vx_nxv1i16_i16(<vscale x 1 x i1> %0, <vscale x 1 x i16> %1, i16 %2, <vscale x 1 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsgt_mask_vx_nxv1i16_i16:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, mf4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a1, e16, mf4, ta, mu
 ; CHECK-NEXT:    vmsgt.vx v10, v8, a0, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v10
 ; CHECK-NEXT:    ret
@@ -1281,9 +1281,9 @@ declare <vscale x 2 x i1> @llvm.riscv.vmsgt.mask.nxv2i16.i16(
 define <vscale x 2 x i1> @intrinsic_vmsgt_mask_vx_nxv2i16_i16(<vscale x 2 x i1> %0, <vscale x 2 x i16> %1, i16 %2, <vscale x 2 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsgt_mask_vx_nxv2i16_i16:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, mf2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a1, e16, mf2, ta, mu
 ; CHECK-NEXT:    vmsgt.vx v10, v8, a0, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v10
 ; CHECK-NEXT:    ret
@@ -1328,9 +1328,9 @@ declare <vscale x 4 x i1> @llvm.riscv.vmsgt.mask.nxv4i16.i16(
 define <vscale x 4 x i1> @intrinsic_vmsgt_mask_vx_nxv4i16_i16(<vscale x 4 x i1> %0, <vscale x 4 x i16> %1, i16 %2, <vscale x 4 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsgt_mask_vx_nxv4i16_i16:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a1, e16, m1, ta, mu
 ; CHECK-NEXT:    vmsgt.vx v10, v8, a0, v0.t
 ; CHECK-NEXT:    vmv.v.v v0, v10
 ; CHECK-NEXT:    ret
@@ -1375,9 +1375,9 @@ declare <vscale x 8 x i1> @llvm.riscv.vmsgt.mask.nxv8i16.i16(
 define <vscale x 8 x i1> @intrinsic_vmsgt_mask_vx_nxv8i16_i16(<vscale x 8 x i1> %0, <vscale x 8 x i16> %1, i16 %2, <vscale x 8 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsgt_mask_vx_nxv8i16_i16:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, m2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmv1r.v v0, v10
-; CHECK-NEXT:    vsetvli zero, a1, e16, m2, ta, mu
 ; CHECK-NEXT:    vmsgt.vx v11, v8, a0, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v11
 ; CHECK-NEXT:    ret
@@ -1422,9 +1422,9 @@ declare <vscale x 16 x i1> @llvm.riscv.vmsgt.mask.nxv16i16.i16(
 define <vscale x 16 x i1> @intrinsic_vmsgt_mask_vx_nxv16i16_i16(<vscale x 16 x i1> %0, <vscale x 16 x i16> %1, i16 %2, <vscale x 16 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsgt_mask_vx_nxv16i16_i16:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, m4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v13, v0
 ; CHECK-NEXT:    vmv1r.v v0, v12
-; CHECK-NEXT:    vsetvli zero, a1, e16, m4, ta, mu
 ; CHECK-NEXT:    vmsgt.vx v13, v8, a0, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v13
 ; CHECK-NEXT:    ret
@@ -1469,9 +1469,9 @@ declare <vscale x 1 x i1> @llvm.riscv.vmsgt.mask.nxv1i32.i32(
 define <vscale x 1 x i1> @intrinsic_vmsgt_mask_vx_nxv1i32_i32(<vscale x 1 x i1> %0, <vscale x 1 x i32> %1, i32 %2, <vscale x 1 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsgt_mask_vx_nxv1i32_i32:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e32, mf2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a1, e32, mf2, ta, mu
 ; CHECK-NEXT:    vmsgt.vx v10, v8, a0, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v10
 ; CHECK-NEXT:    ret
@@ -1516,9 +1516,9 @@ declare <vscale x 2 x i1> @llvm.riscv.vmsgt.mask.nxv2i32.i32(
 define <vscale x 2 x i1> @intrinsic_vmsgt_mask_vx_nxv2i32_i32(<vscale x 2 x i1> %0, <vscale x 2 x i32> %1, i32 %2, <vscale x 2 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsgt_mask_vx_nxv2i32_i32:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e32, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a1, e32, m1, ta, mu
 ; CHECK-NEXT:    vmsgt.vx v10, v8, a0, v0.t
 ; CHECK-NEXT:    vmv.v.v v0, v10
 ; CHECK-NEXT:    ret
@@ -1563,9 +1563,9 @@ declare <vscale x 4 x i1> @llvm.riscv.vmsgt.mask.nxv4i32.i32(
 define <vscale x 4 x i1> @intrinsic_vmsgt_mask_vx_nxv4i32_i32(<vscale x 4 x i1> %0, <vscale x 4 x i32> %1, i32 %2, <vscale x 4 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsgt_mask_vx_nxv4i32_i32:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e32, m2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmv1r.v v0, v10
-; CHECK-NEXT:    vsetvli zero, a1, e32, m2, ta, mu
 ; CHECK-NEXT:    vmsgt.vx v11, v8, a0, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v11
 ; CHECK-NEXT:    ret
@@ -1610,9 +1610,9 @@ declare <vscale x 8 x i1> @llvm.riscv.vmsgt.mask.nxv8i32.i32(
 define <vscale x 8 x i1> @intrinsic_vmsgt_mask_vx_nxv8i32_i32(<vscale x 8 x i1> %0, <vscale x 8 x i32> %1, i32 %2, <vscale x 8 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsgt_mask_vx_nxv8i32_i32:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e32, m4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v13, v0
 ; CHECK-NEXT:    vmv1r.v v0, v12
-; CHECK-NEXT:    vsetvli zero, a1, e32, m4, ta, mu
 ; CHECK-NEXT:    vmsgt.vx v13, v8, a0, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v13
 ; CHECK-NEXT:    ret
@@ -1684,9 +1684,9 @@ define <vscale x 1 x i1> @intrinsic_vmsgt_mask_vx_nxv1i64_i64(<vscale x 1 x i1>
 ;
 ; RV64-LABEL: intrinsic_vmsgt_mask_vx_nxv1i64_i64:
 ; RV64:       # %bb.0: # %entry
+; RV64-NEXT:    vsetvli zero, a1, e64, m1, ta, mu
 ; RV64-NEXT:    vmv1r.v v10, v0
 ; RV64-NEXT:    vmv1r.v v0, v9
-; RV64-NEXT:    vsetvli zero, a1, e64, m1, ta, mu
 ; RV64-NEXT:    vmsgt.vx v10, v8, a0, v0.t
 ; RV64-NEXT:    vmv.v.v v0, v10
 ; RV64-NEXT:    ret
@@ -1758,9 +1758,9 @@ define <vscale x 2 x i1> @intrinsic_vmsgt_mask_vx_nxv2i64_i64(<vscale x 2 x i1>
 ;
 ; RV64-LABEL: intrinsic_vmsgt_mask_vx_nxv2i64_i64:
 ; RV64:       # %bb.0: # %entry
+; RV64-NEXT:    vsetvli zero, a1, e64, m2, ta, mu
 ; RV64-NEXT:    vmv1r.v v11, v0
 ; RV64-NEXT:    vmv1r.v v0, v10
-; RV64-NEXT:    vsetvli zero, a1, e64, m2, ta, mu
 ; RV64-NEXT:    vmsgt.vx v11, v8, a0, v0.t
 ; RV64-NEXT:    vmv1r.v v0, v11
 ; RV64-NEXT:    ret
@@ -1832,9 +1832,9 @@ define <vscale x 4 x i1> @intrinsic_vmsgt_mask_vx_nxv4i64_i64(<vscale x 4 x i1>
 ;
 ; RV64-LABEL: intrinsic_vmsgt_mask_vx_nxv4i64_i64:
 ; RV64:       # %bb.0: # %entry
+; RV64-NEXT:    vsetvli zero, a1, e64, m4, ta, mu
 ; RV64-NEXT:    vmv1r.v v13, v0
 ; RV64-NEXT:    vmv1r.v v0, v12
-; RV64-NEXT:    vsetvli zero, a1, e64, m4, ta, mu
 ; RV64-NEXT:    vmsgt.vx v13, v8, a0, v0.t
 ; RV64-NEXT:    vmv1r.v v0, v13
 ; RV64-NEXT:    ret
@@ -1867,9 +1867,9 @@ entry:
 define <vscale x 1 x i1> @intrinsic_vmsgt_mask_vi_nxv1i8_i8(<vscale x 1 x i1> %0, <vscale x 1 x i8> %1, <vscale x 1 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsgt_mask_vi_nxv1i8_i8:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e8, mf8, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a0, e8, mf8, ta, mu
 ; CHECK-NEXT:    vmsgt.vi v10, v8, 9, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v10
 ; CHECK-NEXT:    ret
@@ -1902,9 +1902,9 @@ entry:
 define <vscale x 2 x i1> @intrinsic_vmsgt_mask_vi_nxv2i8_i8(<vscale x 2 x i1> %0, <vscale x 2 x i8> %1, <vscale x 2 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsgt_mask_vi_nxv2i8_i8:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e8, mf4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a0, e8, mf4, ta, mu
 ; CHECK-NEXT:    vmsgt.vi v10, v8, 9, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v10
 ; CHECK-NEXT:    ret
@@ -1937,9 +1937,9 @@ entry:
 define <vscale x 4 x i1> @intrinsic_vmsgt_mask_vi_nxv4i8_i8(<vscale x 4 x i1> %0, <vscale x 4 x i8> %1, <vscale x 4 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsgt_mask_vi_nxv4i8_i8:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e8, mf2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a0, e8, mf2, ta, mu
 ; CHECK-NEXT:    vmsgt.vi v10, v8, 9, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v10
 ; CHECK-NEXT:    ret
@@ -1972,9 +1972,9 @@ entry:
 define <vscale x 8 x i1> @intrinsic_vmsgt_mask_vi_nxv8i8_i8(<vscale x 8 x i1> %0, <vscale x 8 x i8> %1, <vscale x 8 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsgt_mask_vi_nxv8i8_i8:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e8, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a0, e8, m1, ta, mu
 ; CHECK-NEXT:    vmsgt.vi v10, v8, 9, v0.t
 ; CHECK-NEXT:    vmv.v.v v0, v10
 ; CHECK-NEXT:    ret
@@ -2007,9 +2007,9 @@ entry:
 define <vscale x 16 x i1> @intrinsic_vmsgt_mask_vi_nxv16i8_i8(<vscale x 16 x i1> %0, <vscale x 16 x i8> %1, <vscale x 16 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsgt_mask_vi_nxv16i8_i8:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e8, m2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmv1r.v v0, v10
-; CHECK-NEXT:    vsetvli zero, a0, e8, m2, ta, mu
 ; CHECK-NEXT:    vmsgt.vi v11, v8, 9, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v11
 ; CHECK-NEXT:    ret
@@ -2042,9 +2042,9 @@ entry:
 define <vscale x 32 x i1> @intrinsic_vmsgt_mask_vi_nxv32i8_i8(<vscale x 32 x i1> %0, <vscale x 32 x i8> %1, <vscale x 32 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsgt_mask_vi_nxv32i8_i8:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e8, m4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v13, v0
 ; CHECK-NEXT:    vmv1r.v v0, v12
-; CHECK-NEXT:    vsetvli zero, a0, e8, m4, ta, mu
 ; CHECK-NEXT:    vmsgt.vi v13, v8, 9, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v13
 ; CHECK-NEXT:    ret
@@ -2077,9 +2077,9 @@ entry:
 define <vscale x 1 x i1> @intrinsic_vmsgt_mask_vi_nxv1i16_i16(<vscale x 1 x i1> %0, <vscale x 1 x i16> %1, <vscale x 1 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsgt_mask_vi_nxv1i16_i16:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e16, mf4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a0, e16, mf4, ta, mu
 ; CHECK-NEXT:    vmsgt.vi v10, v8, 9, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v10
 ; CHECK-NEXT:    ret
@@ -2112,9 +2112,9 @@ entry:
 define <vscale x 2 x i1> @intrinsic_vmsgt_mask_vi_nxv2i16_i16(<vscale x 2 x i1> %0, <vscale x 2 x i16> %1, <vscale x 2 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsgt_mask_vi_nxv2i16_i16:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e16, mf2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a0, e16, mf2, ta, mu
 ; CHECK-NEXT:    vmsgt.vi v10, v8, 9, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v10
 ; CHECK-NEXT:    ret
@@ -2147,9 +2147,9 @@ entry:
 define <vscale x 4 x i1> @intrinsic_vmsgt_mask_vi_nxv4i16_i16(<vscale x 4 x i1> %0, <vscale x 4 x i16> %1, <vscale x 4 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsgt_mask_vi_nxv4i16_i16:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e16, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a0, e16, m1, ta, mu
 ; CHECK-NEXT:    vmsgt.vi v10, v8, 9, v0.t
 ; CHECK-NEXT:    vmv.v.v v0, v10
 ; CHECK-NEXT:    ret
@@ -2182,9 +2182,9 @@ entry:
 define <vscale x 8 x i1> @intrinsic_vmsgt_mask_vi_nxv8i16_i16(<vscale x 8 x i1> %0, <vscale x 8 x i16> %1, <vscale x 8 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsgt_mask_vi_nxv8i16_i16:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e16, m2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmv1r.v v0, v10
-; CHECK-NEXT:    vsetvli zero, a0, e16, m2, ta, mu
 ; CHECK-NEXT:    vmsgt.vi v11, v8, 9, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v11
 ; CHECK-NEXT:    ret
@@ -2217,9 +2217,9 @@ entry:
 define <vscale x 16 x i1> @intrinsic_vmsgt_mask_vi_nxv16i16_i16(<vscale x 16 x i1> %0, <vscale x 16 x i16> %1, <vscale x 16 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsgt_mask_vi_nxv16i16_i16:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e16, m4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v13, v0
 ; CHECK-NEXT:    vmv1r.v v0, v12
-; CHECK-NEXT:    vsetvli zero, a0, e16, m4, ta, mu
 ; CHECK-NEXT:    vmsgt.vi v13, v8, 9, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v13
 ; CHECK-NEXT:    ret
@@ -2252,9 +2252,9 @@ entry:
 define <vscale x 1 x i1> @intrinsic_vmsgt_mask_vi_nxv1i32_i32(<vscale x 1 x i1> %0, <vscale x 1 x i32> %1, <vscale x 1 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsgt_mask_vi_nxv1i32_i32:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e32, mf2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a0, e32, mf2, ta, mu
 ; CHECK-NEXT:    vmsgt.vi v10, v8, 9, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v10
 ; CHECK-NEXT:    ret
@@ -2287,9 +2287,9 @@ entry:
 define <vscale x 2 x i1> @intrinsic_vmsgt_mask_vi_nxv2i32_i32(<vscale x 2 x i1> %0, <vscale x 2 x i32> %1, <vscale x 2 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsgt_mask_vi_nxv2i32_i32:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e32, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a0, e32, m1, ta, mu
 ; CHECK-NEXT:    vmsgt.vi v10, v8, 9, v0.t
 ; CHECK-NEXT:    vmv.v.v v0, v10
 ; CHECK-NEXT:    ret
@@ -2322,9 +2322,9 @@ entry:
 define <vscale x 4 x i1> @intrinsic_vmsgt_mask_vi_nxv4i32_i32(<vscale x 4 x i1> %0, <vscale x 4 x i32> %1, <vscale x 4 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsgt_mask_vi_nxv4i32_i32:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e32, m2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmv1r.v v0, v10
-; CHECK-NEXT:    vsetvli zero, a0, e32, m2, ta, mu
 ; CHECK-NEXT:    vmsgt.vi v11, v8, 9, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v11
 ; CHECK-NEXT:    ret
@@ -2357,9 +2357,9 @@ entry:
 define <vscale x 8 x i1> @intrinsic_vmsgt_mask_vi_nxv8i32_i32(<vscale x 8 x i1> %0, <vscale x 8 x i32> %1, <vscale x 8 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsgt_mask_vi_nxv8i32_i32:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e32, m4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v13, v0
 ; CHECK-NEXT:    vmv1r.v v0, v12
-; CHECK-NEXT:    vsetvli zero, a0, e32, m4, ta, mu
 ; CHECK-NEXT:    vmsgt.vi v13, v8, 9, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v13
 ; CHECK-NEXT:    ret
@@ -2392,9 +2392,9 @@ entry:
 define <vscale x 1 x i1> @intrinsic_vmsgt_mask_vi_nxv1i64_i64(<vscale x 1 x i1> %0, <vscale x 1 x i64> %1, <vscale x 1 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsgt_mask_vi_nxv1i64_i64:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e64, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a0, e64, m1, ta, mu
 ; CHECK-NEXT:    vmsgt.vi v10, v8, 9, v0.t
 ; CHECK-NEXT:    vmv.v.v v0, v10
 ; CHECK-NEXT:    ret
@@ -2427,9 +2427,9 @@ entry:
 define <vscale x 2 x i1> @intrinsic_vmsgt_mask_vi_nxv2i64_i64(<vscale x 2 x i1> %0, <vscale x 2 x i64> %1, <vscale x 2 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsgt_mask_vi_nxv2i64_i64:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e64, m2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmv1r.v v0, v10
-; CHECK-NEXT:    vsetvli zero, a0, e64, m2, ta, mu
 ; CHECK-NEXT:    vmsgt.vi v11, v8, 9, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v11
 ; CHECK-NEXT:    ret
@@ -2462,9 +2462,9 @@ entry:
 define <vscale x 4 x i1> @intrinsic_vmsgt_mask_vi_nxv4i64_i64(<vscale x 4 x i1> %0, <vscale x 4 x i64> %1, <vscale x 4 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsgt_mask_vi_nxv4i64_i64:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e64, m4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v13, v0
 ; CHECK-NEXT:    vmv1r.v v0, v12
-; CHECK-NEXT:    vsetvli zero, a0, e64, m4, ta, mu
 ; CHECK-NEXT:    vmsgt.vi v13, v8, 9, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v13
 ; CHECK-NEXT:    ret
diff --git a/llvm/test/CodeGen/RISCV/rvv/vmsgtu.ll b/llvm/test/CodeGen/RISCV/rvv/vmsgtu.ll
index fe9d522f6b401..d57b9cd5bae53 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vmsgtu.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vmsgtu.ll
@@ -34,8 +34,8 @@ declare <vscale x 1 x i1> @llvm.riscv.vmsgtu.mask.nxv1i8(
 define <vscale x 1 x i1> @intrinsic_vmsgtu_mask_vv_nxv1i8_nxv1i8(<vscale x 1 x i1> %0, <vscale x 1 x i8> %1, <vscale x 1 x i8> %2, <vscale x 1 x i8> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsgtu_mask_vv_nxv1i8_nxv1i8:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e8, mf8, ta, mu
+; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmsltu.vv v0, v9, v8
 ; CHECK-NEXT:    vmsltu.vv v11, v10, v9, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v11
@@ -85,8 +85,8 @@ declare <vscale x 2 x i1> @llvm.riscv.vmsgtu.mask.nxv2i8(
 define <vscale x 2 x i1> @intrinsic_vmsgtu_mask_vv_nxv2i8_nxv2i8(<vscale x 2 x i1> %0, <vscale x 2 x i8> %1, <vscale x 2 x i8> %2, <vscale x 2 x i8> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsgtu_mask_vv_nxv2i8_nxv2i8:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e8, mf4, ta, mu
+; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmsltu.vv v0, v9, v8
 ; CHECK-NEXT:    vmsltu.vv v11, v10, v9, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v11
@@ -136,8 +136,8 @@ declare <vscale x 4 x i1> @llvm.riscv.vmsgtu.mask.nxv4i8(
 define <vscale x 4 x i1> @intrinsic_vmsgtu_mask_vv_nxv4i8_nxv4i8(<vscale x 4 x i1> %0, <vscale x 4 x i8> %1, <vscale x 4 x i8> %2, <vscale x 4 x i8> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsgtu_mask_vv_nxv4i8_nxv4i8:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e8, mf2, ta, mu
+; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmsltu.vv v0, v9, v8
 ; CHECK-NEXT:    vmsltu.vv v11, v10, v9, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v11
@@ -187,8 +187,8 @@ declare <vscale x 8 x i1> @llvm.riscv.vmsgtu.mask.nxv8i8(
 define <vscale x 8 x i1> @intrinsic_vmsgtu_mask_vv_nxv8i8_nxv8i8(<vscale x 8 x i1> %0, <vscale x 8 x i8> %1, <vscale x 8 x i8> %2, <vscale x 8 x i8> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsgtu_mask_vv_nxv8i8_nxv8i8:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e8, m1, ta, mu
+; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmsltu.vv v0, v9, v8
 ; CHECK-NEXT:    vmsltu.vv v11, v10, v9, v0.t
 ; CHECK-NEXT:    vmv.v.v v0, v11
@@ -238,8 +238,8 @@ declare <vscale x 16 x i1> @llvm.riscv.vmsgtu.mask.nxv16i8(
 define <vscale x 16 x i1> @intrinsic_vmsgtu_mask_vv_nxv16i8_nxv16i8(<vscale x 16 x i1> %0, <vscale x 16 x i8> %1, <vscale x 16 x i8> %2, <vscale x 16 x i8> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsgtu_mask_vv_nxv16i8_nxv16i8:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v14, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e8, m2, ta, mu
+; CHECK-NEXT:    vmv1r.v v14, v0
 ; CHECK-NEXT:    vmsltu.vv v0, v10, v8
 ; CHECK-NEXT:    vmsltu.vv v14, v12, v10, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v14
@@ -289,8 +289,8 @@ declare <vscale x 32 x i1> @llvm.riscv.vmsgtu.mask.nxv32i8(
 define <vscale x 32 x i1> @intrinsic_vmsgtu_mask_vv_nxv32i8_nxv32i8(<vscale x 32 x i1> %0, <vscale x 32 x i8> %1, <vscale x 32 x i8> %2, <vscale x 32 x i8> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsgtu_mask_vv_nxv32i8_nxv32i8:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v20, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e8, m4, ta, mu
+; CHECK-NEXT:    vmv1r.v v20, v0
 ; CHECK-NEXT:    vmsltu.vv v0, v12, v8
 ; CHECK-NEXT:    vmsltu.vv v20, v16, v12, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v20
@@ -340,8 +340,8 @@ declare <vscale x 1 x i1> @llvm.riscv.vmsgtu.mask.nxv1i16(
 define <vscale x 1 x i1> @intrinsic_vmsgtu_mask_vv_nxv1i16_nxv1i16(<vscale x 1 x i1> %0, <vscale x 1 x i16> %1, <vscale x 1 x i16> %2, <vscale x 1 x i16> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsgtu_mask_vv_nxv1i16_nxv1i16:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e16, mf4, ta, mu
+; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmsltu.vv v0, v9, v8
 ; CHECK-NEXT:    vmsltu.vv v11, v10, v9, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v11
@@ -391,8 +391,8 @@ declare <vscale x 2 x i1> @llvm.riscv.vmsgtu.mask.nxv2i16(
 define <vscale x 2 x i1> @intrinsic_vmsgtu_mask_vv_nxv2i16_nxv2i16(<vscale x 2 x i1> %0, <vscale x 2 x i16> %1, <vscale x 2 x i16> %2, <vscale x 2 x i16> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsgtu_mask_vv_nxv2i16_nxv2i16:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e16, mf2, ta, mu
+; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmsltu.vv v0, v9, v8
 ; CHECK-NEXT:    vmsltu.vv v11, v10, v9, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v11
@@ -442,8 +442,8 @@ declare <vscale x 4 x i1> @llvm.riscv.vmsgtu.mask.nxv4i16(
 define <vscale x 4 x i1> @intrinsic_vmsgtu_mask_vv_nxv4i16_nxv4i16(<vscale x 4 x i1> %0, <vscale x 4 x i16> %1, <vscale x 4 x i16> %2, <vscale x 4 x i16> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsgtu_mask_vv_nxv4i16_nxv4i16:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e16, m1, ta, mu
+; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmsltu.vv v0, v9, v8
 ; CHECK-NEXT:    vmsltu.vv v11, v10, v9, v0.t
 ; CHECK-NEXT:    vmv.v.v v0, v11
@@ -493,8 +493,8 @@ declare <vscale x 8 x i1> @llvm.riscv.vmsgtu.mask.nxv8i16(
 define <vscale x 8 x i1> @intrinsic_vmsgtu_mask_vv_nxv8i16_nxv8i16(<vscale x 8 x i1> %0, <vscale x 8 x i16> %1, <vscale x 8 x i16> %2, <vscale x 8 x i16> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsgtu_mask_vv_nxv8i16_nxv8i16:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v14, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e16, m2, ta, mu
+; CHECK-NEXT:    vmv1r.v v14, v0
 ; CHECK-NEXT:    vmsltu.vv v0, v10, v8
 ; CHECK-NEXT:    vmsltu.vv v14, v12, v10, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v14
@@ -544,8 +544,8 @@ declare <vscale x 16 x i1> @llvm.riscv.vmsgtu.mask.nxv16i16(
 define <vscale x 16 x i1> @intrinsic_vmsgtu_mask_vv_nxv16i16_nxv16i16(<vscale x 16 x i1> %0, <vscale x 16 x i16> %1, <vscale x 16 x i16> %2, <vscale x 16 x i16> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsgtu_mask_vv_nxv16i16_nxv16i16:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v20, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e16, m4, ta, mu
+; CHECK-NEXT:    vmv1r.v v20, v0
 ; CHECK-NEXT:    vmsltu.vv v0, v12, v8
 ; CHECK-NEXT:    vmsltu.vv v20, v16, v12, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v20
@@ -595,8 +595,8 @@ declare <vscale x 1 x i1> @llvm.riscv.vmsgtu.mask.nxv1i32(
 define <vscale x 1 x i1> @intrinsic_vmsgtu_mask_vv_nxv1i32_nxv1i32(<vscale x 1 x i1> %0, <vscale x 1 x i32> %1, <vscale x 1 x i32> %2, <vscale x 1 x i32> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsgtu_mask_vv_nxv1i32_nxv1i32:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e32, mf2, ta, mu
+; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmsltu.vv v0, v9, v8
 ; CHECK-NEXT:    vmsltu.vv v11, v10, v9, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v11
@@ -646,8 +646,8 @@ declare <vscale x 2 x i1> @llvm.riscv.vmsgtu.mask.nxv2i32(
 define <vscale x 2 x i1> @intrinsic_vmsgtu_mask_vv_nxv2i32_nxv2i32(<vscale x 2 x i1> %0, <vscale x 2 x i32> %1, <vscale x 2 x i32> %2, <vscale x 2 x i32> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsgtu_mask_vv_nxv2i32_nxv2i32:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e32, m1, ta, mu
+; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmsltu.vv v0, v9, v8
 ; CHECK-NEXT:    vmsltu.vv v11, v10, v9, v0.t
 ; CHECK-NEXT:    vmv.v.v v0, v11
@@ -697,8 +697,8 @@ declare <vscale x 4 x i1> @llvm.riscv.vmsgtu.mask.nxv4i32(
 define <vscale x 4 x i1> @intrinsic_vmsgtu_mask_vv_nxv4i32_nxv4i32(<vscale x 4 x i1> %0, <vscale x 4 x i32> %1, <vscale x 4 x i32> %2, <vscale x 4 x i32> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsgtu_mask_vv_nxv4i32_nxv4i32:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v14, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e32, m2, ta, mu
+; CHECK-NEXT:    vmv1r.v v14, v0
 ; CHECK-NEXT:    vmsltu.vv v0, v10, v8
 ; CHECK-NEXT:    vmsltu.vv v14, v12, v10, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v14
@@ -748,8 +748,8 @@ declare <vscale x 8 x i1> @llvm.riscv.vmsgtu.mask.nxv8i32(
 define <vscale x 8 x i1> @intrinsic_vmsgtu_mask_vv_nxv8i32_nxv8i32(<vscale x 8 x i1> %0, <vscale x 8 x i32> %1, <vscale x 8 x i32> %2, <vscale x 8 x i32> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsgtu_mask_vv_nxv8i32_nxv8i32:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v20, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e32, m4, ta, mu
+; CHECK-NEXT:    vmv1r.v v20, v0
 ; CHECK-NEXT:    vmsltu.vv v0, v12, v8
 ; CHECK-NEXT:    vmsltu.vv v20, v16, v12, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v20
@@ -799,8 +799,8 @@ declare <vscale x 1 x i1> @llvm.riscv.vmsgtu.mask.nxv1i64(
 define <vscale x 1 x i1> @intrinsic_vmsgtu_mask_vv_nxv1i64_nxv1i64(<vscale x 1 x i1> %0, <vscale x 1 x i64> %1, <vscale x 1 x i64> %2, <vscale x 1 x i64> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsgtu_mask_vv_nxv1i64_nxv1i64:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e64, m1, ta, mu
+; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmsltu.vv v0, v9, v8
 ; CHECK-NEXT:    vmsltu.vv v11, v10, v9, v0.t
 ; CHECK-NEXT:    vmv.v.v v0, v11
@@ -850,8 +850,8 @@ declare <vscale x 2 x i1> @llvm.riscv.vmsgtu.mask.nxv2i64(
 define <vscale x 2 x i1> @intrinsic_vmsgtu_mask_vv_nxv2i64_nxv2i64(<vscale x 2 x i1> %0, <vscale x 2 x i64> %1, <vscale x 2 x i64> %2, <vscale x 2 x i64> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsgtu_mask_vv_nxv2i64_nxv2i64:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v14, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e64, m2, ta, mu
+; CHECK-NEXT:    vmv1r.v v14, v0
 ; CHECK-NEXT:    vmsltu.vv v0, v10, v8
 ; CHECK-NEXT:    vmsltu.vv v14, v12, v10, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v14
@@ -901,8 +901,8 @@ declare <vscale x 4 x i1> @llvm.riscv.vmsgtu.mask.nxv4i64(
 define <vscale x 4 x i1> @intrinsic_vmsgtu_mask_vv_nxv4i64_nxv4i64(<vscale x 4 x i1> %0, <vscale x 4 x i64> %1, <vscale x 4 x i64> %2, <vscale x 4 x i64> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsgtu_mask_vv_nxv4i64_nxv4i64:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v20, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e64, m4, ta, mu
+; CHECK-NEXT:    vmv1r.v v20, v0
 ; CHECK-NEXT:    vmsltu.vv v0, v12, v8
 ; CHECK-NEXT:    vmsltu.vv v20, v16, v12, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v20
@@ -952,9 +952,9 @@ declare <vscale x 1 x i1> @llvm.riscv.vmsgtu.mask.nxv1i8.i8(
 define <vscale x 1 x i1> @intrinsic_vmsgtu_mask_vx_nxv1i8_i8(<vscale x 1 x i1> %0, <vscale x 1 x i8> %1, i8 %2, <vscale x 1 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsgtu_mask_vx_nxv1i8_i8:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e8, mf8, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a1, e8, mf8, ta, mu
 ; CHECK-NEXT:    vmsgtu.vx v10, v8, a0, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v10
 ; CHECK-NEXT:    ret
@@ -999,9 +999,9 @@ declare <vscale x 2 x i1> @llvm.riscv.vmsgtu.mask.nxv2i8.i8(
 define <vscale x 2 x i1> @intrinsic_vmsgtu_mask_vx_nxv2i8_i8(<vscale x 2 x i1> %0, <vscale x 2 x i8> %1, i8 %2, <vscale x 2 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsgtu_mask_vx_nxv2i8_i8:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e8, mf4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a1, e8, mf4, ta, mu
 ; CHECK-NEXT:    vmsgtu.vx v10, v8, a0, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v10
 ; CHECK-NEXT:    ret
@@ -1046,9 +1046,9 @@ declare <vscale x 4 x i1> @llvm.riscv.vmsgtu.mask.nxv4i8.i8(
 define <vscale x 4 x i1> @intrinsic_vmsgtu_mask_vx_nxv4i8_i8(<vscale x 4 x i1> %0, <vscale x 4 x i8> %1, i8 %2, <vscale x 4 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsgtu_mask_vx_nxv4i8_i8:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e8, mf2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a1, e8, mf2, ta, mu
 ; CHECK-NEXT:    vmsgtu.vx v10, v8, a0, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v10
 ; CHECK-NEXT:    ret
@@ -1093,9 +1093,9 @@ declare <vscale x 8 x i1> @llvm.riscv.vmsgtu.mask.nxv8i8.i8(
 define <vscale x 8 x i1> @intrinsic_vmsgtu_mask_vx_nxv8i8_i8(<vscale x 8 x i1> %0, <vscale x 8 x i8> %1, i8 %2, <vscale x 8 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsgtu_mask_vx_nxv8i8_i8:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e8, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a1, e8, m1, ta, mu
 ; CHECK-NEXT:    vmsgtu.vx v10, v8, a0, v0.t
 ; CHECK-NEXT:    vmv.v.v v0, v10
 ; CHECK-NEXT:    ret
@@ -1140,9 +1140,9 @@ declare <vscale x 16 x i1> @llvm.riscv.vmsgtu.mask.nxv16i8.i8(
 define <vscale x 16 x i1> @intrinsic_vmsgtu_mask_vx_nxv16i8_i8(<vscale x 16 x i1> %0, <vscale x 16 x i8> %1, i8 %2, <vscale x 16 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsgtu_mask_vx_nxv16i8_i8:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e8, m2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmv1r.v v0, v10
-; CHECK-NEXT:    vsetvli zero, a1, e8, m2, ta, mu
 ; CHECK-NEXT:    vmsgtu.vx v11, v8, a0, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v11
 ; CHECK-NEXT:    ret
@@ -1187,9 +1187,9 @@ declare <vscale x 32 x i1> @llvm.riscv.vmsgtu.mask.nxv32i8.i8(
 define <vscale x 32 x i1> @intrinsic_vmsgtu_mask_vx_nxv32i8_i8(<vscale x 32 x i1> %0, <vscale x 32 x i8> %1, i8 %2, <vscale x 32 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsgtu_mask_vx_nxv32i8_i8:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e8, m4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v13, v0
 ; CHECK-NEXT:    vmv1r.v v0, v12
-; CHECK-NEXT:    vsetvli zero, a1, e8, m4, ta, mu
 ; CHECK-NEXT:    vmsgtu.vx v13, v8, a0, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v13
 ; CHECK-NEXT:    ret
@@ -1234,9 +1234,9 @@ declare <vscale x 1 x i1> @llvm.riscv.vmsgtu.mask.nxv1i16.i16(
 define <vscale x 1 x i1> @intrinsic_vmsgtu_mask_vx_nxv1i16_i16(<vscale x 1 x i1> %0, <vscale x 1 x i16> %1, i16 %2, <vscale x 1 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsgtu_mask_vx_nxv1i16_i16:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, mf4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a1, e16, mf4, ta, mu
 ; CHECK-NEXT:    vmsgtu.vx v10, v8, a0, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v10
 ; CHECK-NEXT:    ret
@@ -1281,9 +1281,9 @@ declare <vscale x 2 x i1> @llvm.riscv.vmsgtu.mask.nxv2i16.i16(
 define <vscale x 2 x i1> @intrinsic_vmsgtu_mask_vx_nxv2i16_i16(<vscale x 2 x i1> %0, <vscale x 2 x i16> %1, i16 %2, <vscale x 2 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsgtu_mask_vx_nxv2i16_i16:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, mf2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a1, e16, mf2, ta, mu
 ; CHECK-NEXT:    vmsgtu.vx v10, v8, a0, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v10
 ; CHECK-NEXT:    ret
@@ -1328,9 +1328,9 @@ declare <vscale x 4 x i1> @llvm.riscv.vmsgtu.mask.nxv4i16.i16(
 define <vscale x 4 x i1> @intrinsic_vmsgtu_mask_vx_nxv4i16_i16(<vscale x 4 x i1> %0, <vscale x 4 x i16> %1, i16 %2, <vscale x 4 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsgtu_mask_vx_nxv4i16_i16:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a1, e16, m1, ta, mu
 ; CHECK-NEXT:    vmsgtu.vx v10, v8, a0, v0.t
 ; CHECK-NEXT:    vmv.v.v v0, v10
 ; CHECK-NEXT:    ret
@@ -1375,9 +1375,9 @@ declare <vscale x 8 x i1> @llvm.riscv.vmsgtu.mask.nxv8i16.i16(
 define <vscale x 8 x i1> @intrinsic_vmsgtu_mask_vx_nxv8i16_i16(<vscale x 8 x i1> %0, <vscale x 8 x i16> %1, i16 %2, <vscale x 8 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsgtu_mask_vx_nxv8i16_i16:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, m2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmv1r.v v0, v10
-; CHECK-NEXT:    vsetvli zero, a1, e16, m2, ta, mu
 ; CHECK-NEXT:    vmsgtu.vx v11, v8, a0, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v11
 ; CHECK-NEXT:    ret
@@ -1422,9 +1422,9 @@ declare <vscale x 16 x i1> @llvm.riscv.vmsgtu.mask.nxv16i16.i16(
 define <vscale x 16 x i1> @intrinsic_vmsgtu_mask_vx_nxv16i16_i16(<vscale x 16 x i1> %0, <vscale x 16 x i16> %1, i16 %2, <vscale x 16 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsgtu_mask_vx_nxv16i16_i16:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, m4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v13, v0
 ; CHECK-NEXT:    vmv1r.v v0, v12
-; CHECK-NEXT:    vsetvli zero, a1, e16, m4, ta, mu
 ; CHECK-NEXT:    vmsgtu.vx v13, v8, a0, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v13
 ; CHECK-NEXT:    ret
@@ -1469,9 +1469,9 @@ declare <vscale x 1 x i1> @llvm.riscv.vmsgtu.mask.nxv1i32.i32(
 define <vscale x 1 x i1> @intrinsic_vmsgtu_mask_vx_nxv1i32_i32(<vscale x 1 x i1> %0, <vscale x 1 x i32> %1, i32 %2, <vscale x 1 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsgtu_mask_vx_nxv1i32_i32:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e32, mf2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a1, e32, mf2, ta, mu
 ; CHECK-NEXT:    vmsgtu.vx v10, v8, a0, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v10
 ; CHECK-NEXT:    ret
@@ -1516,9 +1516,9 @@ declare <vscale x 2 x i1> @llvm.riscv.vmsgtu.mask.nxv2i32.i32(
 define <vscale x 2 x i1> @intrinsic_vmsgtu_mask_vx_nxv2i32_i32(<vscale x 2 x i1> %0, <vscale x 2 x i32> %1, i32 %2, <vscale x 2 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsgtu_mask_vx_nxv2i32_i32:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e32, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a1, e32, m1, ta, mu
 ; CHECK-NEXT:    vmsgtu.vx v10, v8, a0, v0.t
 ; CHECK-NEXT:    vmv.v.v v0, v10
 ; CHECK-NEXT:    ret
@@ -1563,9 +1563,9 @@ declare <vscale x 4 x i1> @llvm.riscv.vmsgtu.mask.nxv4i32.i32(
 define <vscale x 4 x i1> @intrinsic_vmsgtu_mask_vx_nxv4i32_i32(<vscale x 4 x i1> %0, <vscale x 4 x i32> %1, i32 %2, <vscale x 4 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsgtu_mask_vx_nxv4i32_i32:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e32, m2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmv1r.v v0, v10
-; CHECK-NEXT:    vsetvli zero, a1, e32, m2, ta, mu
 ; CHECK-NEXT:    vmsgtu.vx v11, v8, a0, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v11
 ; CHECK-NEXT:    ret
@@ -1610,9 +1610,9 @@ declare <vscale x 8 x i1> @llvm.riscv.vmsgtu.mask.nxv8i32.i32(
 define <vscale x 8 x i1> @intrinsic_vmsgtu_mask_vx_nxv8i32_i32(<vscale x 8 x i1> %0, <vscale x 8 x i32> %1, i32 %2, <vscale x 8 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsgtu_mask_vx_nxv8i32_i32:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e32, m4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v13, v0
 ; CHECK-NEXT:    vmv1r.v v0, v12
-; CHECK-NEXT:    vsetvli zero, a1, e32, m4, ta, mu
 ; CHECK-NEXT:    vmsgtu.vx v13, v8, a0, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v13
 ; CHECK-NEXT:    ret
@@ -1684,9 +1684,9 @@ define <vscale x 1 x i1> @intrinsic_vmsgtu_mask_vx_nxv1i64_i64(<vscale x 1 x i1>
 ;
 ; RV64-LABEL: intrinsic_vmsgtu_mask_vx_nxv1i64_i64:
 ; RV64:       # %bb.0: # %entry
+; RV64-NEXT:    vsetvli zero, a1, e64, m1, ta, mu
 ; RV64-NEXT:    vmv1r.v v10, v0
 ; RV64-NEXT:    vmv1r.v v0, v9
-; RV64-NEXT:    vsetvli zero, a1, e64, m1, ta, mu
 ; RV64-NEXT:    vmsgtu.vx v10, v8, a0, v0.t
 ; RV64-NEXT:    vmv.v.v v0, v10
 ; RV64-NEXT:    ret
@@ -1758,9 +1758,9 @@ define <vscale x 2 x i1> @intrinsic_vmsgtu_mask_vx_nxv2i64_i64(<vscale x 2 x i1>
 ;
 ; RV64-LABEL: intrinsic_vmsgtu_mask_vx_nxv2i64_i64:
 ; RV64:       # %bb.0: # %entry
+; RV64-NEXT:    vsetvli zero, a1, e64, m2, ta, mu
 ; RV64-NEXT:    vmv1r.v v11, v0
 ; RV64-NEXT:    vmv1r.v v0, v10
-; RV64-NEXT:    vsetvli zero, a1, e64, m2, ta, mu
 ; RV64-NEXT:    vmsgtu.vx v11, v8, a0, v0.t
 ; RV64-NEXT:    vmv1r.v v0, v11
 ; RV64-NEXT:    ret
@@ -1832,9 +1832,9 @@ define <vscale x 4 x i1> @intrinsic_vmsgtu_mask_vx_nxv4i64_i64(<vscale x 4 x i1>
 ;
 ; RV64-LABEL: intrinsic_vmsgtu_mask_vx_nxv4i64_i64:
 ; RV64:       # %bb.0: # %entry
+; RV64-NEXT:    vsetvli zero, a1, e64, m4, ta, mu
 ; RV64-NEXT:    vmv1r.v v13, v0
 ; RV64-NEXT:    vmv1r.v v0, v12
-; RV64-NEXT:    vsetvli zero, a1, e64, m4, ta, mu
 ; RV64-NEXT:    vmsgtu.vx v13, v8, a0, v0.t
 ; RV64-NEXT:    vmv1r.v v0, v13
 ; RV64-NEXT:    ret
@@ -1867,9 +1867,9 @@ entry:
 define <vscale x 1 x i1> @intrinsic_vmsgtu_mask_vi_nxv1i8_i8(<vscale x 1 x i1> %0, <vscale x 1 x i8> %1, <vscale x 1 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsgtu_mask_vi_nxv1i8_i8:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e8, mf8, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a0, e8, mf8, ta, mu
 ; CHECK-NEXT:    vmsgtu.vi v10, v8, 9, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v10
 ; CHECK-NEXT:    ret
@@ -1902,9 +1902,9 @@ entry:
 define <vscale x 2 x i1> @intrinsic_vmsgtu_mask_vi_nxv2i8_i8(<vscale x 2 x i1> %0, <vscale x 2 x i8> %1, <vscale x 2 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsgtu_mask_vi_nxv2i8_i8:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e8, mf4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a0, e8, mf4, ta, mu
 ; CHECK-NEXT:    vmsgtu.vi v10, v8, 9, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v10
 ; CHECK-NEXT:    ret
@@ -1937,9 +1937,9 @@ entry:
 define <vscale x 4 x i1> @intrinsic_vmsgtu_mask_vi_nxv4i8_i8(<vscale x 4 x i1> %0, <vscale x 4 x i8> %1, <vscale x 4 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsgtu_mask_vi_nxv4i8_i8:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e8, mf2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a0, e8, mf2, ta, mu
 ; CHECK-NEXT:    vmsgtu.vi v10, v8, 9, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v10
 ; CHECK-NEXT:    ret
@@ -1972,9 +1972,9 @@ entry:
 define <vscale x 8 x i1> @intrinsic_vmsgtu_mask_vi_nxv8i8_i8(<vscale x 8 x i1> %0, <vscale x 8 x i8> %1, <vscale x 8 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsgtu_mask_vi_nxv8i8_i8:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e8, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a0, e8, m1, ta, mu
 ; CHECK-NEXT:    vmsgtu.vi v10, v8, 9, v0.t
 ; CHECK-NEXT:    vmv.v.v v0, v10
 ; CHECK-NEXT:    ret
@@ -2007,9 +2007,9 @@ entry:
 define <vscale x 16 x i1> @intrinsic_vmsgtu_mask_vi_nxv16i8_i8(<vscale x 16 x i1> %0, <vscale x 16 x i8> %1, <vscale x 16 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsgtu_mask_vi_nxv16i8_i8:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e8, m2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmv1r.v v0, v10
-; CHECK-NEXT:    vsetvli zero, a0, e8, m2, ta, mu
 ; CHECK-NEXT:    vmsgtu.vi v11, v8, 9, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v11
 ; CHECK-NEXT:    ret
@@ -2042,9 +2042,9 @@ entry:
 define <vscale x 32 x i1> @intrinsic_vmsgtu_mask_vi_nxv32i8_i8(<vscale x 32 x i1> %0, <vscale x 32 x i8> %1, <vscale x 32 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsgtu_mask_vi_nxv32i8_i8:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e8, m4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v13, v0
 ; CHECK-NEXT:    vmv1r.v v0, v12
-; CHECK-NEXT:    vsetvli zero, a0, e8, m4, ta, mu
 ; CHECK-NEXT:    vmsgtu.vi v13, v8, 9, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v13
 ; CHECK-NEXT:    ret
@@ -2077,9 +2077,9 @@ entry:
 define <vscale x 1 x i1> @intrinsic_vmsgtu_mask_vi_nxv1i16_i16(<vscale x 1 x i1> %0, <vscale x 1 x i16> %1, <vscale x 1 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsgtu_mask_vi_nxv1i16_i16:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e16, mf4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a0, e16, mf4, ta, mu
 ; CHECK-NEXT:    vmsgtu.vi v10, v8, 9, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v10
 ; CHECK-NEXT:    ret
@@ -2112,9 +2112,9 @@ entry:
 define <vscale x 2 x i1> @intrinsic_vmsgtu_mask_vi_nxv2i16_i16(<vscale x 2 x i1> %0, <vscale x 2 x i16> %1, <vscale x 2 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsgtu_mask_vi_nxv2i16_i16:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e16, mf2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a0, e16, mf2, ta, mu
 ; CHECK-NEXT:    vmsgtu.vi v10, v8, 9, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v10
 ; CHECK-NEXT:    ret
@@ -2147,9 +2147,9 @@ entry:
 define <vscale x 4 x i1> @intrinsic_vmsgtu_mask_vi_nxv4i16_i16(<vscale x 4 x i1> %0, <vscale x 4 x i16> %1, <vscale x 4 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsgtu_mask_vi_nxv4i16_i16:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e16, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a0, e16, m1, ta, mu
 ; CHECK-NEXT:    vmsgtu.vi v10, v8, 9, v0.t
 ; CHECK-NEXT:    vmv.v.v v0, v10
 ; CHECK-NEXT:    ret
@@ -2182,9 +2182,9 @@ entry:
 define <vscale x 8 x i1> @intrinsic_vmsgtu_mask_vi_nxv8i16_i16(<vscale x 8 x i1> %0, <vscale x 8 x i16> %1, <vscale x 8 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsgtu_mask_vi_nxv8i16_i16:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e16, m2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmv1r.v v0, v10
-; CHECK-NEXT:    vsetvli zero, a0, e16, m2, ta, mu
 ; CHECK-NEXT:    vmsgtu.vi v11, v8, 9, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v11
 ; CHECK-NEXT:    ret
@@ -2217,9 +2217,9 @@ entry:
 define <vscale x 16 x i1> @intrinsic_vmsgtu_mask_vi_nxv16i16_i16(<vscale x 16 x i1> %0, <vscale x 16 x i16> %1, <vscale x 16 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsgtu_mask_vi_nxv16i16_i16:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e16, m4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v13, v0
 ; CHECK-NEXT:    vmv1r.v v0, v12
-; CHECK-NEXT:    vsetvli zero, a0, e16, m4, ta, mu
 ; CHECK-NEXT:    vmsgtu.vi v13, v8, 9, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v13
 ; CHECK-NEXT:    ret
@@ -2252,9 +2252,9 @@ entry:
 define <vscale x 1 x i1> @intrinsic_vmsgtu_mask_vi_nxv1i32_i32(<vscale x 1 x i1> %0, <vscale x 1 x i32> %1, <vscale x 1 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsgtu_mask_vi_nxv1i32_i32:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e32, mf2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a0, e32, mf2, ta, mu
 ; CHECK-NEXT:    vmsgtu.vi v10, v8, 9, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v10
 ; CHECK-NEXT:    ret
@@ -2287,9 +2287,9 @@ entry:
 define <vscale x 2 x i1> @intrinsic_vmsgtu_mask_vi_nxv2i32_i32(<vscale x 2 x i1> %0, <vscale x 2 x i32> %1, <vscale x 2 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsgtu_mask_vi_nxv2i32_i32:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e32, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a0, e32, m1, ta, mu
 ; CHECK-NEXT:    vmsgtu.vi v10, v8, 9, v0.t
 ; CHECK-NEXT:    vmv.v.v v0, v10
 ; CHECK-NEXT:    ret
@@ -2322,9 +2322,9 @@ entry:
 define <vscale x 4 x i1> @intrinsic_vmsgtu_mask_vi_nxv4i32_i32(<vscale x 4 x i1> %0, <vscale x 4 x i32> %1, <vscale x 4 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsgtu_mask_vi_nxv4i32_i32:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e32, m2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmv1r.v v0, v10
-; CHECK-NEXT:    vsetvli zero, a0, e32, m2, ta, mu
 ; CHECK-NEXT:    vmsgtu.vi v11, v8, 9, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v11
 ; CHECK-NEXT:    ret
@@ -2357,9 +2357,9 @@ entry:
 define <vscale x 8 x i1> @intrinsic_vmsgtu_mask_vi_nxv8i32_i32(<vscale x 8 x i1> %0, <vscale x 8 x i32> %1, <vscale x 8 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsgtu_mask_vi_nxv8i32_i32:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e32, m4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v13, v0
 ; CHECK-NEXT:    vmv1r.v v0, v12
-; CHECK-NEXT:    vsetvli zero, a0, e32, m4, ta, mu
 ; CHECK-NEXT:    vmsgtu.vi v13, v8, 9, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v13
 ; CHECK-NEXT:    ret
@@ -2392,9 +2392,9 @@ entry:
 define <vscale x 1 x i1> @intrinsic_vmsgtu_mask_vi_nxv1i64_i64(<vscale x 1 x i1> %0, <vscale x 1 x i64> %1, <vscale x 1 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsgtu_mask_vi_nxv1i64_i64:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e64, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a0, e64, m1, ta, mu
 ; CHECK-NEXT:    vmsgtu.vi v10, v8, 9, v0.t
 ; CHECK-NEXT:    vmv.v.v v0, v10
 ; CHECK-NEXT:    ret
@@ -2427,9 +2427,9 @@ entry:
 define <vscale x 2 x i1> @intrinsic_vmsgtu_mask_vi_nxv2i64_i64(<vscale x 2 x i1> %0, <vscale x 2 x i64> %1, <vscale x 2 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsgtu_mask_vi_nxv2i64_i64:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e64, m2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmv1r.v v0, v10
-; CHECK-NEXT:    vsetvli zero, a0, e64, m2, ta, mu
 ; CHECK-NEXT:    vmsgtu.vi v11, v8, 9, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v11
 ; CHECK-NEXT:    ret
@@ -2462,9 +2462,9 @@ entry:
 define <vscale x 4 x i1> @intrinsic_vmsgtu_mask_vi_nxv4i64_i64(<vscale x 4 x i1> %0, <vscale x 4 x i64> %1, <vscale x 4 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsgtu_mask_vi_nxv4i64_i64:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e64, m4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v13, v0
 ; CHECK-NEXT:    vmv1r.v v0, v12
-; CHECK-NEXT:    vsetvli zero, a0, e64, m4, ta, mu
 ; CHECK-NEXT:    vmsgtu.vi v13, v8, 9, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v13
 ; CHECK-NEXT:    ret
diff --git a/llvm/test/CodeGen/RISCV/rvv/vmsif.ll b/llvm/test/CodeGen/RISCV/rvv/vmsif.ll
index 1dc52eb55455b..9c70dcab1efde 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vmsif.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vmsif.ll
@@ -31,9 +31,9 @@ declare <vscale x 1 x i1> @llvm.riscv.vmsif.mask.nxv1i1(
 define <vscale x 1 x i1> @intrinsic_vmsif_mask_m_nxv1i1_nxv1i1(<vscale x 1 x i1> %0, <vscale x 1 x i1> %1, <vscale x 1 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsif_mask_m_nxv1i1_nxv1i1:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e8, mf8, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a0, e8, mf8, ta, mu
 ; CHECK-NEXT:    vmsif.m v10, v8, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v10
 ; CHECK-NEXT:    ret
@@ -73,9 +73,9 @@ declare <vscale x 2 x i1> @llvm.riscv.vmsif.mask.nxv2i1(
 define <vscale x 2 x i1> @intrinsic_vmsif_mask_m_nxv2i1_nxv2i1(<vscale x 2 x i1> %0, <vscale x 2 x i1> %1, <vscale x 2 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsif_mask_m_nxv2i1_nxv2i1:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e8, mf4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a0, e8, mf4, ta, mu
 ; CHECK-NEXT:    vmsif.m v10, v8, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v10
 ; CHECK-NEXT:    ret
@@ -115,9 +115,9 @@ declare <vscale x 4 x i1> @llvm.riscv.vmsif.mask.nxv4i1(
 define <vscale x 4 x i1> @intrinsic_vmsif_mask_m_nxv4i1_nxv4i1(<vscale x 4 x i1> %0, <vscale x 4 x i1> %1, <vscale x 4 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsif_mask_m_nxv4i1_nxv4i1:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e8, mf2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a0, e8, mf2, ta, mu
 ; CHECK-NEXT:    vmsif.m v10, v8, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v10
 ; CHECK-NEXT:    ret
@@ -157,9 +157,9 @@ declare <vscale x 8 x i1> @llvm.riscv.vmsif.mask.nxv8i1(
 define <vscale x 8 x i1> @intrinsic_vmsif_mask_m_nxv8i1_nxv8i1(<vscale x 8 x i1> %0, <vscale x 8 x i1> %1, <vscale x 8 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsif_mask_m_nxv8i1_nxv8i1:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e8, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a0, e8, m1, ta, mu
 ; CHECK-NEXT:    vmsif.m v10, v8, v0.t
 ; CHECK-NEXT:    vmv.v.v v0, v10
 ; CHECK-NEXT:    ret
@@ -199,9 +199,9 @@ declare <vscale x 16 x i1> @llvm.riscv.vmsif.mask.nxv16i1(
 define <vscale x 16 x i1> @intrinsic_vmsif_mask_m_nxv16i1_nxv16i1(<vscale x 16 x i1> %0, <vscale x 16 x i1> %1, <vscale x 16 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsif_mask_m_nxv16i1_nxv16i1:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e8, m2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a0, e8, m2, ta, mu
 ; CHECK-NEXT:    vmsif.m v10, v8, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v10
 ; CHECK-NEXT:    ret
@@ -241,9 +241,9 @@ declare <vscale x 32 x i1> @llvm.riscv.vmsif.mask.nxv32i1(
 define <vscale x 32 x i1> @intrinsic_vmsif_mask_m_nxv32i1_nxv32i1(<vscale x 32 x i1> %0, <vscale x 32 x i1> %1, <vscale x 32 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsif_mask_m_nxv32i1_nxv32i1:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e8, m4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a0, e8, m4, ta, mu
 ; CHECK-NEXT:    vmsif.m v10, v8, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v10
 ; CHECK-NEXT:    ret
@@ -283,9 +283,9 @@ declare <vscale x 64 x i1> @llvm.riscv.vmsif.mask.nxv64i1(
 define <vscale x 64 x i1> @intrinsic_vmsif_mask_m_nxv64i1_nxv64i1(<vscale x 64 x i1> %0, <vscale x 64 x i1> %1, <vscale x 64 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsif_mask_m_nxv64i1_nxv64i1:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e8, m8, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a0, e8, m8, ta, mu
 ; CHECK-NEXT:    vmsif.m v10, v8, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v10
 ; CHECK-NEXT:    ret
diff --git a/llvm/test/CodeGen/RISCV/rvv/vmsle.ll b/llvm/test/CodeGen/RISCV/rvv/vmsle.ll
index bc98b31957b25..9653dfd2518d8 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vmsle.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vmsle.ll
@@ -34,8 +34,8 @@ declare <vscale x 1 x i1> @llvm.riscv.vmsle.mask.nxv1i8(
 define <vscale x 1 x i1> @intrinsic_vmsle_mask_vv_nxv1i8_nxv1i8(<vscale x 1 x i1> %0, <vscale x 1 x i8> %1, <vscale x 1 x i8> %2, <vscale x 1 x i8> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsle_mask_vv_nxv1i8_nxv1i8:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e8, mf8, ta, mu
+; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmsle.vv v0, v8, v9
 ; CHECK-NEXT:    vmsle.vv v11, v9, v10, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v11
@@ -85,8 +85,8 @@ declare <vscale x 2 x i1> @llvm.riscv.vmsle.mask.nxv2i8(
 define <vscale x 2 x i1> @intrinsic_vmsle_mask_vv_nxv2i8_nxv2i8(<vscale x 2 x i1> %0, <vscale x 2 x i8> %1, <vscale x 2 x i8> %2, <vscale x 2 x i8> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsle_mask_vv_nxv2i8_nxv2i8:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e8, mf4, ta, mu
+; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmsle.vv v0, v8, v9
 ; CHECK-NEXT:    vmsle.vv v11, v9, v10, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v11
@@ -136,8 +136,8 @@ declare <vscale x 4 x i1> @llvm.riscv.vmsle.mask.nxv4i8(
 define <vscale x 4 x i1> @intrinsic_vmsle_mask_vv_nxv4i8_nxv4i8(<vscale x 4 x i1> %0, <vscale x 4 x i8> %1, <vscale x 4 x i8> %2, <vscale x 4 x i8> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsle_mask_vv_nxv4i8_nxv4i8:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e8, mf2, ta, mu
+; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmsle.vv v0, v8, v9
 ; CHECK-NEXT:    vmsle.vv v11, v9, v10, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v11
@@ -187,8 +187,8 @@ declare <vscale x 8 x i1> @llvm.riscv.vmsle.mask.nxv8i8(
 define <vscale x 8 x i1> @intrinsic_vmsle_mask_vv_nxv8i8_nxv8i8(<vscale x 8 x i1> %0, <vscale x 8 x i8> %1, <vscale x 8 x i8> %2, <vscale x 8 x i8> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsle_mask_vv_nxv8i8_nxv8i8:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e8, m1, ta, mu
+; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmsle.vv v0, v8, v9
 ; CHECK-NEXT:    vmsle.vv v11, v9, v10, v0.t
 ; CHECK-NEXT:    vmv.v.v v0, v11
@@ -238,8 +238,8 @@ declare <vscale x 16 x i1> @llvm.riscv.vmsle.mask.nxv16i8(
 define <vscale x 16 x i1> @intrinsic_vmsle_mask_vv_nxv16i8_nxv16i8(<vscale x 16 x i1> %0, <vscale x 16 x i8> %1, <vscale x 16 x i8> %2, <vscale x 16 x i8> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsle_mask_vv_nxv16i8_nxv16i8:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v14, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e8, m2, ta, mu
+; CHECK-NEXT:    vmv1r.v v14, v0
 ; CHECK-NEXT:    vmsle.vv v0, v8, v10
 ; CHECK-NEXT:    vmsle.vv v14, v10, v12, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v14
@@ -289,8 +289,8 @@ declare <vscale x 32 x i1> @llvm.riscv.vmsle.mask.nxv32i8(
 define <vscale x 32 x i1> @intrinsic_vmsle_mask_vv_nxv32i8_nxv32i8(<vscale x 32 x i1> %0, <vscale x 32 x i8> %1, <vscale x 32 x i8> %2, <vscale x 32 x i8> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsle_mask_vv_nxv32i8_nxv32i8:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v20, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e8, m4, ta, mu
+; CHECK-NEXT:    vmv1r.v v20, v0
 ; CHECK-NEXT:    vmsle.vv v0, v8, v12
 ; CHECK-NEXT:    vmsle.vv v20, v12, v16, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v20
@@ -340,8 +340,8 @@ declare <vscale x 1 x i1> @llvm.riscv.vmsle.mask.nxv1i16(
 define <vscale x 1 x i1> @intrinsic_vmsle_mask_vv_nxv1i16_nxv1i16(<vscale x 1 x i1> %0, <vscale x 1 x i16> %1, <vscale x 1 x i16> %2, <vscale x 1 x i16> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsle_mask_vv_nxv1i16_nxv1i16:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e16, mf4, ta, mu
+; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmsle.vv v0, v8, v9
 ; CHECK-NEXT:    vmsle.vv v11, v9, v10, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v11
@@ -391,8 +391,8 @@ declare <vscale x 2 x i1> @llvm.riscv.vmsle.mask.nxv2i16(
 define <vscale x 2 x i1> @intrinsic_vmsle_mask_vv_nxv2i16_nxv2i16(<vscale x 2 x i1> %0, <vscale x 2 x i16> %1, <vscale x 2 x i16> %2, <vscale x 2 x i16> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsle_mask_vv_nxv2i16_nxv2i16:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e16, mf2, ta, mu
+; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmsle.vv v0, v8, v9
 ; CHECK-NEXT:    vmsle.vv v11, v9, v10, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v11
@@ -442,8 +442,8 @@ declare <vscale x 4 x i1> @llvm.riscv.vmsle.mask.nxv4i16(
 define <vscale x 4 x i1> @intrinsic_vmsle_mask_vv_nxv4i16_nxv4i16(<vscale x 4 x i1> %0, <vscale x 4 x i16> %1, <vscale x 4 x i16> %2, <vscale x 4 x i16> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsle_mask_vv_nxv4i16_nxv4i16:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e16, m1, ta, mu
+; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmsle.vv v0, v8, v9
 ; CHECK-NEXT:    vmsle.vv v11, v9, v10, v0.t
 ; CHECK-NEXT:    vmv.v.v v0, v11
@@ -493,8 +493,8 @@ declare <vscale x 8 x i1> @llvm.riscv.vmsle.mask.nxv8i16(
 define <vscale x 8 x i1> @intrinsic_vmsle_mask_vv_nxv8i16_nxv8i16(<vscale x 8 x i1> %0, <vscale x 8 x i16> %1, <vscale x 8 x i16> %2, <vscale x 8 x i16> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsle_mask_vv_nxv8i16_nxv8i16:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v14, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e16, m2, ta, mu
+; CHECK-NEXT:    vmv1r.v v14, v0
 ; CHECK-NEXT:    vmsle.vv v0, v8, v10
 ; CHECK-NEXT:    vmsle.vv v14, v10, v12, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v14
@@ -544,8 +544,8 @@ declare <vscale x 16 x i1> @llvm.riscv.vmsle.mask.nxv16i16(
 define <vscale x 16 x i1> @intrinsic_vmsle_mask_vv_nxv16i16_nxv16i16(<vscale x 16 x i1> %0, <vscale x 16 x i16> %1, <vscale x 16 x i16> %2, <vscale x 16 x i16> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsle_mask_vv_nxv16i16_nxv16i16:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v20, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e16, m4, ta, mu
+; CHECK-NEXT:    vmv1r.v v20, v0
 ; CHECK-NEXT:    vmsle.vv v0, v8, v12
 ; CHECK-NEXT:    vmsle.vv v20, v12, v16, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v20
@@ -595,8 +595,8 @@ declare <vscale x 1 x i1> @llvm.riscv.vmsle.mask.nxv1i32(
 define <vscale x 1 x i1> @intrinsic_vmsle_mask_vv_nxv1i32_nxv1i32(<vscale x 1 x i1> %0, <vscale x 1 x i32> %1, <vscale x 1 x i32> %2, <vscale x 1 x i32> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsle_mask_vv_nxv1i32_nxv1i32:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e32, mf2, ta, mu
+; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmsle.vv v0, v8, v9
 ; CHECK-NEXT:    vmsle.vv v11, v9, v10, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v11
@@ -646,8 +646,8 @@ declare <vscale x 2 x i1> @llvm.riscv.vmsle.mask.nxv2i32(
 define <vscale x 2 x i1> @intrinsic_vmsle_mask_vv_nxv2i32_nxv2i32(<vscale x 2 x i1> %0, <vscale x 2 x i32> %1, <vscale x 2 x i32> %2, <vscale x 2 x i32> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsle_mask_vv_nxv2i32_nxv2i32:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e32, m1, ta, mu
+; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmsle.vv v0, v8, v9
 ; CHECK-NEXT:    vmsle.vv v11, v9, v10, v0.t
 ; CHECK-NEXT:    vmv.v.v v0, v11
@@ -697,8 +697,8 @@ declare <vscale x 4 x i1> @llvm.riscv.vmsle.mask.nxv4i32(
 define <vscale x 4 x i1> @intrinsic_vmsle_mask_vv_nxv4i32_nxv4i32(<vscale x 4 x i1> %0, <vscale x 4 x i32> %1, <vscale x 4 x i32> %2, <vscale x 4 x i32> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsle_mask_vv_nxv4i32_nxv4i32:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v14, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e32, m2, ta, mu
+; CHECK-NEXT:    vmv1r.v v14, v0
 ; CHECK-NEXT:    vmsle.vv v0, v8, v10
 ; CHECK-NEXT:    vmsle.vv v14, v10, v12, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v14
@@ -748,8 +748,8 @@ declare <vscale x 8 x i1> @llvm.riscv.vmsle.mask.nxv8i32(
 define <vscale x 8 x i1> @intrinsic_vmsle_mask_vv_nxv8i32_nxv8i32(<vscale x 8 x i1> %0, <vscale x 8 x i32> %1, <vscale x 8 x i32> %2, <vscale x 8 x i32> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsle_mask_vv_nxv8i32_nxv8i32:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v20, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e32, m4, ta, mu
+; CHECK-NEXT:    vmv1r.v v20, v0
 ; CHECK-NEXT:    vmsle.vv v0, v8, v12
 ; CHECK-NEXT:    vmsle.vv v20, v12, v16, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v20
@@ -799,8 +799,8 @@ declare <vscale x 1 x i1> @llvm.riscv.vmsle.mask.nxv1i64(
 define <vscale x 1 x i1> @intrinsic_vmsle_mask_vv_nxv1i64_nxv1i64(<vscale x 1 x i1> %0, <vscale x 1 x i64> %1, <vscale x 1 x i64> %2, <vscale x 1 x i64> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsle_mask_vv_nxv1i64_nxv1i64:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e64, m1, ta, mu
+; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmsle.vv v0, v8, v9
 ; CHECK-NEXT:    vmsle.vv v11, v9, v10, v0.t
 ; CHECK-NEXT:    vmv.v.v v0, v11
@@ -850,8 +850,8 @@ declare <vscale x 2 x i1> @llvm.riscv.vmsle.mask.nxv2i64(
 define <vscale x 2 x i1> @intrinsic_vmsle_mask_vv_nxv2i64_nxv2i64(<vscale x 2 x i1> %0, <vscale x 2 x i64> %1, <vscale x 2 x i64> %2, <vscale x 2 x i64> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsle_mask_vv_nxv2i64_nxv2i64:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v14, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e64, m2, ta, mu
+; CHECK-NEXT:    vmv1r.v v14, v0
 ; CHECK-NEXT:    vmsle.vv v0, v8, v10
 ; CHECK-NEXT:    vmsle.vv v14, v10, v12, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v14
@@ -901,8 +901,8 @@ declare <vscale x 4 x i1> @llvm.riscv.vmsle.mask.nxv4i64(
 define <vscale x 4 x i1> @intrinsic_vmsle_mask_vv_nxv4i64_nxv4i64(<vscale x 4 x i1> %0, <vscale x 4 x i64> %1, <vscale x 4 x i64> %2, <vscale x 4 x i64> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsle_mask_vv_nxv4i64_nxv4i64:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v20, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e64, m4, ta, mu
+; CHECK-NEXT:    vmv1r.v v20, v0
 ; CHECK-NEXT:    vmsle.vv v0, v8, v12
 ; CHECK-NEXT:    vmsle.vv v20, v12, v16, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v20
@@ -952,9 +952,9 @@ declare <vscale x 1 x i1> @llvm.riscv.vmsle.mask.nxv1i8.i8(
 define <vscale x 1 x i1> @intrinsic_vmsle_mask_vx_nxv1i8_i8(<vscale x 1 x i1> %0, <vscale x 1 x i8> %1, i8 %2, <vscale x 1 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsle_mask_vx_nxv1i8_i8:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e8, mf8, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a1, e8, mf8, ta, mu
 ; CHECK-NEXT:    vmsle.vx v10, v8, a0, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v10
 ; CHECK-NEXT:    ret
@@ -999,9 +999,9 @@ declare <vscale x 2 x i1> @llvm.riscv.vmsle.mask.nxv2i8.i8(
 define <vscale x 2 x i1> @intrinsic_vmsle_mask_vx_nxv2i8_i8(<vscale x 2 x i1> %0, <vscale x 2 x i8> %1, i8 %2, <vscale x 2 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsle_mask_vx_nxv2i8_i8:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e8, mf4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a1, e8, mf4, ta, mu
 ; CHECK-NEXT:    vmsle.vx v10, v8, a0, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v10
 ; CHECK-NEXT:    ret
@@ -1046,9 +1046,9 @@ declare <vscale x 4 x i1> @llvm.riscv.vmsle.mask.nxv4i8.i8(
 define <vscale x 4 x i1> @intrinsic_vmsle_mask_vx_nxv4i8_i8(<vscale x 4 x i1> %0, <vscale x 4 x i8> %1, i8 %2, <vscale x 4 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsle_mask_vx_nxv4i8_i8:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e8, mf2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a1, e8, mf2, ta, mu
 ; CHECK-NEXT:    vmsle.vx v10, v8, a0, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v10
 ; CHECK-NEXT:    ret
@@ -1093,9 +1093,9 @@ declare <vscale x 8 x i1> @llvm.riscv.vmsle.mask.nxv8i8.i8(
 define <vscale x 8 x i1> @intrinsic_vmsle_mask_vx_nxv8i8_i8(<vscale x 8 x i1> %0, <vscale x 8 x i8> %1, i8 %2, <vscale x 8 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsle_mask_vx_nxv8i8_i8:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e8, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a1, e8, m1, ta, mu
 ; CHECK-NEXT:    vmsle.vx v10, v8, a0, v0.t
 ; CHECK-NEXT:    vmv.v.v v0, v10
 ; CHECK-NEXT:    ret
@@ -1140,9 +1140,9 @@ declare <vscale x 16 x i1> @llvm.riscv.vmsle.mask.nxv16i8.i8(
 define <vscale x 16 x i1> @intrinsic_vmsle_mask_vx_nxv16i8_i8(<vscale x 16 x i1> %0, <vscale x 16 x i8> %1, i8 %2, <vscale x 16 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsle_mask_vx_nxv16i8_i8:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e8, m2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmv1r.v v0, v10
-; CHECK-NEXT:    vsetvli zero, a1, e8, m2, ta, mu
 ; CHECK-NEXT:    vmsle.vx v11, v8, a0, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v11
 ; CHECK-NEXT:    ret
@@ -1187,9 +1187,9 @@ declare <vscale x 32 x i1> @llvm.riscv.vmsle.mask.nxv32i8.i8(
 define <vscale x 32 x i1> @intrinsic_vmsle_mask_vx_nxv32i8_i8(<vscale x 32 x i1> %0, <vscale x 32 x i8> %1, i8 %2, <vscale x 32 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsle_mask_vx_nxv32i8_i8:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e8, m4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v13, v0
 ; CHECK-NEXT:    vmv1r.v v0, v12
-; CHECK-NEXT:    vsetvli zero, a1, e8, m4, ta, mu
 ; CHECK-NEXT:    vmsle.vx v13, v8, a0, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v13
 ; CHECK-NEXT:    ret
@@ -1234,9 +1234,9 @@ declare <vscale x 1 x i1> @llvm.riscv.vmsle.mask.nxv1i16.i16(
 define <vscale x 1 x i1> @intrinsic_vmsle_mask_vx_nxv1i16_i16(<vscale x 1 x i1> %0, <vscale x 1 x i16> %1, i16 %2, <vscale x 1 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsle_mask_vx_nxv1i16_i16:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, mf4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a1, e16, mf4, ta, mu
 ; CHECK-NEXT:    vmsle.vx v10, v8, a0, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v10
 ; CHECK-NEXT:    ret
@@ -1281,9 +1281,9 @@ declare <vscale x 2 x i1> @llvm.riscv.vmsle.mask.nxv2i16.i16(
 define <vscale x 2 x i1> @intrinsic_vmsle_mask_vx_nxv2i16_i16(<vscale x 2 x i1> %0, <vscale x 2 x i16> %1, i16 %2, <vscale x 2 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsle_mask_vx_nxv2i16_i16:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, mf2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a1, e16, mf2, ta, mu
 ; CHECK-NEXT:    vmsle.vx v10, v8, a0, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v10
 ; CHECK-NEXT:    ret
@@ -1328,9 +1328,9 @@ declare <vscale x 4 x i1> @llvm.riscv.vmsle.mask.nxv4i16.i16(
 define <vscale x 4 x i1> @intrinsic_vmsle_mask_vx_nxv4i16_i16(<vscale x 4 x i1> %0, <vscale x 4 x i16> %1, i16 %2, <vscale x 4 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsle_mask_vx_nxv4i16_i16:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a1, e16, m1, ta, mu
 ; CHECK-NEXT:    vmsle.vx v10, v8, a0, v0.t
 ; CHECK-NEXT:    vmv.v.v v0, v10
 ; CHECK-NEXT:    ret
@@ -1375,9 +1375,9 @@ declare <vscale x 8 x i1> @llvm.riscv.vmsle.mask.nxv8i16.i16(
 define <vscale x 8 x i1> @intrinsic_vmsle_mask_vx_nxv8i16_i16(<vscale x 8 x i1> %0, <vscale x 8 x i16> %1, i16 %2, <vscale x 8 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsle_mask_vx_nxv8i16_i16:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, m2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmv1r.v v0, v10
-; CHECK-NEXT:    vsetvli zero, a1, e16, m2, ta, mu
 ; CHECK-NEXT:    vmsle.vx v11, v8, a0, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v11
 ; CHECK-NEXT:    ret
@@ -1422,9 +1422,9 @@ declare <vscale x 16 x i1> @llvm.riscv.vmsle.mask.nxv16i16.i16(
 define <vscale x 16 x i1> @intrinsic_vmsle_mask_vx_nxv16i16_i16(<vscale x 16 x i1> %0, <vscale x 16 x i16> %1, i16 %2, <vscale x 16 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsle_mask_vx_nxv16i16_i16:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, m4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v13, v0
 ; CHECK-NEXT:    vmv1r.v v0, v12
-; CHECK-NEXT:    vsetvli zero, a1, e16, m4, ta, mu
 ; CHECK-NEXT:    vmsle.vx v13, v8, a0, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v13
 ; CHECK-NEXT:    ret
@@ -1469,9 +1469,9 @@ declare <vscale x 1 x i1> @llvm.riscv.vmsle.mask.nxv1i32.i32(
 define <vscale x 1 x i1> @intrinsic_vmsle_mask_vx_nxv1i32_i32(<vscale x 1 x i1> %0, <vscale x 1 x i32> %1, i32 %2, <vscale x 1 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsle_mask_vx_nxv1i32_i32:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e32, mf2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a1, e32, mf2, ta, mu
 ; CHECK-NEXT:    vmsle.vx v10, v8, a0, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v10
 ; CHECK-NEXT:    ret
@@ -1516,9 +1516,9 @@ declare <vscale x 2 x i1> @llvm.riscv.vmsle.mask.nxv2i32.i32(
 define <vscale x 2 x i1> @intrinsic_vmsle_mask_vx_nxv2i32_i32(<vscale x 2 x i1> %0, <vscale x 2 x i32> %1, i32 %2, <vscale x 2 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsle_mask_vx_nxv2i32_i32:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e32, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a1, e32, m1, ta, mu
 ; CHECK-NEXT:    vmsle.vx v10, v8, a0, v0.t
 ; CHECK-NEXT:    vmv.v.v v0, v10
 ; CHECK-NEXT:    ret
@@ -1563,9 +1563,9 @@ declare <vscale x 4 x i1> @llvm.riscv.vmsle.mask.nxv4i32.i32(
 define <vscale x 4 x i1> @intrinsic_vmsle_mask_vx_nxv4i32_i32(<vscale x 4 x i1> %0, <vscale x 4 x i32> %1, i32 %2, <vscale x 4 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsle_mask_vx_nxv4i32_i32:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e32, m2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmv1r.v v0, v10
-; CHECK-NEXT:    vsetvli zero, a1, e32, m2, ta, mu
 ; CHECK-NEXT:    vmsle.vx v11, v8, a0, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v11
 ; CHECK-NEXT:    ret
@@ -1610,9 +1610,9 @@ declare <vscale x 8 x i1> @llvm.riscv.vmsle.mask.nxv8i32.i32(
 define <vscale x 8 x i1> @intrinsic_vmsle_mask_vx_nxv8i32_i32(<vscale x 8 x i1> %0, <vscale x 8 x i32> %1, i32 %2, <vscale x 8 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsle_mask_vx_nxv8i32_i32:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e32, m4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v13, v0
 ; CHECK-NEXT:    vmv1r.v v0, v12
-; CHECK-NEXT:    vsetvli zero, a1, e32, m4, ta, mu
 ; CHECK-NEXT:    vmsle.vx v13, v8, a0, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v13
 ; CHECK-NEXT:    ret
@@ -1684,9 +1684,9 @@ define <vscale x 1 x i1> @intrinsic_vmsle_mask_vx_nxv1i64_i64(<vscale x 1 x i1>
 ;
 ; RV64-LABEL: intrinsic_vmsle_mask_vx_nxv1i64_i64:
 ; RV64:       # %bb.0: # %entry
+; RV64-NEXT:    vsetvli zero, a1, e64, m1, ta, mu
 ; RV64-NEXT:    vmv1r.v v10, v0
 ; RV64-NEXT:    vmv1r.v v0, v9
-; RV64-NEXT:    vsetvli zero, a1, e64, m1, ta, mu
 ; RV64-NEXT:    vmsle.vx v10, v8, a0, v0.t
 ; RV64-NEXT:    vmv.v.v v0, v10
 ; RV64-NEXT:    ret
@@ -1758,9 +1758,9 @@ define <vscale x 2 x i1> @intrinsic_vmsle_mask_vx_nxv2i64_i64(<vscale x 2 x i1>
 ;
 ; RV64-LABEL: intrinsic_vmsle_mask_vx_nxv2i64_i64:
 ; RV64:       # %bb.0: # %entry
+; RV64-NEXT:    vsetvli zero, a1, e64, m2, ta, mu
 ; RV64-NEXT:    vmv1r.v v11, v0
 ; RV64-NEXT:    vmv1r.v v0, v10
-; RV64-NEXT:    vsetvli zero, a1, e64, m2, ta, mu
 ; RV64-NEXT:    vmsle.vx v11, v8, a0, v0.t
 ; RV64-NEXT:    vmv1r.v v0, v11
 ; RV64-NEXT:    ret
@@ -1832,9 +1832,9 @@ define <vscale x 4 x i1> @intrinsic_vmsle_mask_vx_nxv4i64_i64(<vscale x 4 x i1>
 ;
 ; RV64-LABEL: intrinsic_vmsle_mask_vx_nxv4i64_i64:
 ; RV64:       # %bb.0: # %entry
+; RV64-NEXT:    vsetvli zero, a1, e64, m4, ta, mu
 ; RV64-NEXT:    vmv1r.v v13, v0
 ; RV64-NEXT:    vmv1r.v v0, v12
-; RV64-NEXT:    vsetvli zero, a1, e64, m4, ta, mu
 ; RV64-NEXT:    vmsle.vx v13, v8, a0, v0.t
 ; RV64-NEXT:    vmv1r.v v0, v13
 ; RV64-NEXT:    ret
@@ -1867,9 +1867,9 @@ entry:
 define <vscale x 1 x i1> @intrinsic_vmsle_mask_vi_nxv1i8_i8(<vscale x 1 x i1> %0, <vscale x 1 x i8> %1, <vscale x 1 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsle_mask_vi_nxv1i8_i8:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e8, mf8, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a0, e8, mf8, ta, mu
 ; CHECK-NEXT:    vmsle.vi v10, v8, 9, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v10
 ; CHECK-NEXT:    ret
@@ -1902,9 +1902,9 @@ entry:
 define <vscale x 2 x i1> @intrinsic_vmsle_mask_vi_nxv2i8_i8(<vscale x 2 x i1> %0, <vscale x 2 x i8> %1, <vscale x 2 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsle_mask_vi_nxv2i8_i8:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e8, mf4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a0, e8, mf4, ta, mu
 ; CHECK-NEXT:    vmsle.vi v10, v8, 9, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v10
 ; CHECK-NEXT:    ret
@@ -1937,9 +1937,9 @@ entry:
 define <vscale x 4 x i1> @intrinsic_vmsle_mask_vi_nxv4i8_i8(<vscale x 4 x i1> %0, <vscale x 4 x i8> %1, <vscale x 4 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsle_mask_vi_nxv4i8_i8:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e8, mf2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a0, e8, mf2, ta, mu
 ; CHECK-NEXT:    vmsle.vi v10, v8, 9, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v10
 ; CHECK-NEXT:    ret
@@ -1972,9 +1972,9 @@ entry:
 define <vscale x 8 x i1> @intrinsic_vmsle_mask_vi_nxv8i8_i8(<vscale x 8 x i1> %0, <vscale x 8 x i8> %1, <vscale x 8 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsle_mask_vi_nxv8i8_i8:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e8, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a0, e8, m1, ta, mu
 ; CHECK-NEXT:    vmsle.vi v10, v8, 9, v0.t
 ; CHECK-NEXT:    vmv.v.v v0, v10
 ; CHECK-NEXT:    ret
@@ -2007,9 +2007,9 @@ entry:
 define <vscale x 16 x i1> @intrinsic_vmsle_mask_vi_nxv16i8_i8(<vscale x 16 x i1> %0, <vscale x 16 x i8> %1, <vscale x 16 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsle_mask_vi_nxv16i8_i8:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e8, m2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmv1r.v v0, v10
-; CHECK-NEXT:    vsetvli zero, a0, e8, m2, ta, mu
 ; CHECK-NEXT:    vmsle.vi v11, v8, 9, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v11
 ; CHECK-NEXT:    ret
@@ -2042,9 +2042,9 @@ entry:
 define <vscale x 32 x i1> @intrinsic_vmsle_mask_vi_nxv32i8_i8(<vscale x 32 x i1> %0, <vscale x 32 x i8> %1, <vscale x 32 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsle_mask_vi_nxv32i8_i8:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e8, m4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v13, v0
 ; CHECK-NEXT:    vmv1r.v v0, v12
-; CHECK-NEXT:    vsetvli zero, a0, e8, m4, ta, mu
 ; CHECK-NEXT:    vmsle.vi v13, v8, 9, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v13
 ; CHECK-NEXT:    ret
@@ -2077,9 +2077,9 @@ entry:
 define <vscale x 1 x i1> @intrinsic_vmsle_mask_vi_nxv1i16_i16(<vscale x 1 x i1> %0, <vscale x 1 x i16> %1, <vscale x 1 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsle_mask_vi_nxv1i16_i16:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e16, mf4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a0, e16, mf4, ta, mu
 ; CHECK-NEXT:    vmsle.vi v10, v8, 9, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v10
 ; CHECK-NEXT:    ret
@@ -2112,9 +2112,9 @@ entry:
 define <vscale x 2 x i1> @intrinsic_vmsle_mask_vi_nxv2i16_i16(<vscale x 2 x i1> %0, <vscale x 2 x i16> %1, <vscale x 2 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsle_mask_vi_nxv2i16_i16:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e16, mf2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a0, e16, mf2, ta, mu
 ; CHECK-NEXT:    vmsle.vi v10, v8, 9, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v10
 ; CHECK-NEXT:    ret
@@ -2147,9 +2147,9 @@ entry:
 define <vscale x 4 x i1> @intrinsic_vmsle_mask_vi_nxv4i16_i16(<vscale x 4 x i1> %0, <vscale x 4 x i16> %1, <vscale x 4 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsle_mask_vi_nxv4i16_i16:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e16, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a0, e16, m1, ta, mu
 ; CHECK-NEXT:    vmsle.vi v10, v8, 9, v0.t
 ; CHECK-NEXT:    vmv.v.v v0, v10
 ; CHECK-NEXT:    ret
@@ -2182,9 +2182,9 @@ entry:
 define <vscale x 8 x i1> @intrinsic_vmsle_mask_vi_nxv8i16_i16(<vscale x 8 x i1> %0, <vscale x 8 x i16> %1, <vscale x 8 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsle_mask_vi_nxv8i16_i16:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e16, m2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmv1r.v v0, v10
-; CHECK-NEXT:    vsetvli zero, a0, e16, m2, ta, mu
 ; CHECK-NEXT:    vmsle.vi v11, v8, 9, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v11
 ; CHECK-NEXT:    ret
@@ -2217,9 +2217,9 @@ entry:
 define <vscale x 16 x i1> @intrinsic_vmsle_mask_vi_nxv16i16_i16(<vscale x 16 x i1> %0, <vscale x 16 x i16> %1, <vscale x 16 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsle_mask_vi_nxv16i16_i16:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e16, m4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v13, v0
 ; CHECK-NEXT:    vmv1r.v v0, v12
-; CHECK-NEXT:    vsetvli zero, a0, e16, m4, ta, mu
 ; CHECK-NEXT:    vmsle.vi v13, v8, 9, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v13
 ; CHECK-NEXT:    ret
@@ -2252,9 +2252,9 @@ entry:
 define <vscale x 1 x i1> @intrinsic_vmsle_mask_vi_nxv1i32_i32(<vscale x 1 x i1> %0, <vscale x 1 x i32> %1, <vscale x 1 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsle_mask_vi_nxv1i32_i32:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e32, mf2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a0, e32, mf2, ta, mu
 ; CHECK-NEXT:    vmsle.vi v10, v8, 9, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v10
 ; CHECK-NEXT:    ret
@@ -2287,9 +2287,9 @@ entry:
 define <vscale x 2 x i1> @intrinsic_vmsle_mask_vi_nxv2i32_i32(<vscale x 2 x i1> %0, <vscale x 2 x i32> %1, <vscale x 2 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsle_mask_vi_nxv2i32_i32:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e32, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a0, e32, m1, ta, mu
 ; CHECK-NEXT:    vmsle.vi v10, v8, 9, v0.t
 ; CHECK-NEXT:    vmv.v.v v0, v10
 ; CHECK-NEXT:    ret
@@ -2322,9 +2322,9 @@ entry:
 define <vscale x 4 x i1> @intrinsic_vmsle_mask_vi_nxv4i32_i32(<vscale x 4 x i1> %0, <vscale x 4 x i32> %1, <vscale x 4 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsle_mask_vi_nxv4i32_i32:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e32, m2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmv1r.v v0, v10
-; CHECK-NEXT:    vsetvli zero, a0, e32, m2, ta, mu
 ; CHECK-NEXT:    vmsle.vi v11, v8, 9, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v11
 ; CHECK-NEXT:    ret
@@ -2357,9 +2357,9 @@ entry:
 define <vscale x 8 x i1> @intrinsic_vmsle_mask_vi_nxv8i32_i32(<vscale x 8 x i1> %0, <vscale x 8 x i32> %1, <vscale x 8 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsle_mask_vi_nxv8i32_i32:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e32, m4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v13, v0
 ; CHECK-NEXT:    vmv1r.v v0, v12
-; CHECK-NEXT:    vsetvli zero, a0, e32, m4, ta, mu
 ; CHECK-NEXT:    vmsle.vi v13, v8, 9, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v13
 ; CHECK-NEXT:    ret
@@ -2392,9 +2392,9 @@ entry:
 define <vscale x 1 x i1> @intrinsic_vmsle_mask_vi_nxv1i64_i64(<vscale x 1 x i1> %0, <vscale x 1 x i64> %1, <vscale x 1 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsle_mask_vi_nxv1i64_i64:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e64, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a0, e64, m1, ta, mu
 ; CHECK-NEXT:    vmsle.vi v10, v8, 9, v0.t
 ; CHECK-NEXT:    vmv.v.v v0, v10
 ; CHECK-NEXT:    ret
@@ -2427,9 +2427,9 @@ entry:
 define <vscale x 2 x i1> @intrinsic_vmsle_mask_vi_nxv2i64_i64(<vscale x 2 x i1> %0, <vscale x 2 x i64> %1, <vscale x 2 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsle_mask_vi_nxv2i64_i64:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e64, m2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmv1r.v v0, v10
-; CHECK-NEXT:    vsetvli zero, a0, e64, m2, ta, mu
 ; CHECK-NEXT:    vmsle.vi v11, v8, 9, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v11
 ; CHECK-NEXT:    ret
@@ -2462,9 +2462,9 @@ entry:
 define <vscale x 4 x i1> @intrinsic_vmsle_mask_vi_nxv4i64_i64(<vscale x 4 x i1> %0, <vscale x 4 x i64> %1, <vscale x 4 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsle_mask_vi_nxv4i64_i64:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e64, m4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v13, v0
 ; CHECK-NEXT:    vmv1r.v v0, v12
-; CHECK-NEXT:    vsetvli zero, a0, e64, m4, ta, mu
 ; CHECK-NEXT:    vmsle.vi v13, v8, 9, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v13
 ; CHECK-NEXT:    ret
diff --git a/llvm/test/CodeGen/RISCV/rvv/vmsleu.ll b/llvm/test/CodeGen/RISCV/rvv/vmsleu.ll
index 731989cfe15d9..25ecfa65c7c48 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vmsleu.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vmsleu.ll
@@ -34,8 +34,8 @@ declare <vscale x 1 x i1> @llvm.riscv.vmsleu.mask.nxv1i8(
 define <vscale x 1 x i1> @intrinsic_vmsleu_mask_vv_nxv1i8_nxv1i8(<vscale x 1 x i1> %0, <vscale x 1 x i8> %1, <vscale x 1 x i8> %2, <vscale x 1 x i8> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsleu_mask_vv_nxv1i8_nxv1i8:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e8, mf8, ta, mu
+; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmsleu.vv v0, v8, v9
 ; CHECK-NEXT:    vmsleu.vv v11, v9, v10, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v11
@@ -85,8 +85,8 @@ declare <vscale x 2 x i1> @llvm.riscv.vmsleu.mask.nxv2i8(
 define <vscale x 2 x i1> @intrinsic_vmsleu_mask_vv_nxv2i8_nxv2i8(<vscale x 2 x i1> %0, <vscale x 2 x i8> %1, <vscale x 2 x i8> %2, <vscale x 2 x i8> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsleu_mask_vv_nxv2i8_nxv2i8:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e8, mf4, ta, mu
+; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmsleu.vv v0, v8, v9
 ; CHECK-NEXT:    vmsleu.vv v11, v9, v10, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v11
@@ -136,8 +136,8 @@ declare <vscale x 4 x i1> @llvm.riscv.vmsleu.mask.nxv4i8(
 define <vscale x 4 x i1> @intrinsic_vmsleu_mask_vv_nxv4i8_nxv4i8(<vscale x 4 x i1> %0, <vscale x 4 x i8> %1, <vscale x 4 x i8> %2, <vscale x 4 x i8> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsleu_mask_vv_nxv4i8_nxv4i8:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e8, mf2, ta, mu
+; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmsleu.vv v0, v8, v9
 ; CHECK-NEXT:    vmsleu.vv v11, v9, v10, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v11
@@ -187,8 +187,8 @@ declare <vscale x 8 x i1> @llvm.riscv.vmsleu.mask.nxv8i8(
 define <vscale x 8 x i1> @intrinsic_vmsleu_mask_vv_nxv8i8_nxv8i8(<vscale x 8 x i1> %0, <vscale x 8 x i8> %1, <vscale x 8 x i8> %2, <vscale x 8 x i8> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsleu_mask_vv_nxv8i8_nxv8i8:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e8, m1, ta, mu
+; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmsleu.vv v0, v8, v9
 ; CHECK-NEXT:    vmsleu.vv v11, v9, v10, v0.t
 ; CHECK-NEXT:    vmv.v.v v0, v11
@@ -238,8 +238,8 @@ declare <vscale x 16 x i1> @llvm.riscv.vmsleu.mask.nxv16i8(
 define <vscale x 16 x i1> @intrinsic_vmsleu_mask_vv_nxv16i8_nxv16i8(<vscale x 16 x i1> %0, <vscale x 16 x i8> %1, <vscale x 16 x i8> %2, <vscale x 16 x i8> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsleu_mask_vv_nxv16i8_nxv16i8:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v14, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e8, m2, ta, mu
+; CHECK-NEXT:    vmv1r.v v14, v0
 ; CHECK-NEXT:    vmsleu.vv v0, v8, v10
 ; CHECK-NEXT:    vmsleu.vv v14, v10, v12, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v14
@@ -289,8 +289,8 @@ declare <vscale x 32 x i1> @llvm.riscv.vmsleu.mask.nxv32i8(
 define <vscale x 32 x i1> @intrinsic_vmsleu_mask_vv_nxv32i8_nxv32i8(<vscale x 32 x i1> %0, <vscale x 32 x i8> %1, <vscale x 32 x i8> %2, <vscale x 32 x i8> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsleu_mask_vv_nxv32i8_nxv32i8:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v20, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e8, m4, ta, mu
+; CHECK-NEXT:    vmv1r.v v20, v0
 ; CHECK-NEXT:    vmsleu.vv v0, v8, v12
 ; CHECK-NEXT:    vmsleu.vv v20, v12, v16, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v20
@@ -340,8 +340,8 @@ declare <vscale x 1 x i1> @llvm.riscv.vmsleu.mask.nxv1i16(
 define <vscale x 1 x i1> @intrinsic_vmsleu_mask_vv_nxv1i16_nxv1i16(<vscale x 1 x i1> %0, <vscale x 1 x i16> %1, <vscale x 1 x i16> %2, <vscale x 1 x i16> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsleu_mask_vv_nxv1i16_nxv1i16:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e16, mf4, ta, mu
+; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmsleu.vv v0, v8, v9
 ; CHECK-NEXT:    vmsleu.vv v11, v9, v10, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v11
@@ -391,8 +391,8 @@ declare <vscale x 2 x i1> @llvm.riscv.vmsleu.mask.nxv2i16(
 define <vscale x 2 x i1> @intrinsic_vmsleu_mask_vv_nxv2i16_nxv2i16(<vscale x 2 x i1> %0, <vscale x 2 x i16> %1, <vscale x 2 x i16> %2, <vscale x 2 x i16> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsleu_mask_vv_nxv2i16_nxv2i16:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e16, mf2, ta, mu
+; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmsleu.vv v0, v8, v9
 ; CHECK-NEXT:    vmsleu.vv v11, v9, v10, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v11
@@ -442,8 +442,8 @@ declare <vscale x 4 x i1> @llvm.riscv.vmsleu.mask.nxv4i16(
 define <vscale x 4 x i1> @intrinsic_vmsleu_mask_vv_nxv4i16_nxv4i16(<vscale x 4 x i1> %0, <vscale x 4 x i16> %1, <vscale x 4 x i16> %2, <vscale x 4 x i16> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsleu_mask_vv_nxv4i16_nxv4i16:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e16, m1, ta, mu
+; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmsleu.vv v0, v8, v9
 ; CHECK-NEXT:    vmsleu.vv v11, v9, v10, v0.t
 ; CHECK-NEXT:    vmv.v.v v0, v11
@@ -493,8 +493,8 @@ declare <vscale x 8 x i1> @llvm.riscv.vmsleu.mask.nxv8i16(
 define <vscale x 8 x i1> @intrinsic_vmsleu_mask_vv_nxv8i16_nxv8i16(<vscale x 8 x i1> %0, <vscale x 8 x i16> %1, <vscale x 8 x i16> %2, <vscale x 8 x i16> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsleu_mask_vv_nxv8i16_nxv8i16:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v14, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e16, m2, ta, mu
+; CHECK-NEXT:    vmv1r.v v14, v0
 ; CHECK-NEXT:    vmsleu.vv v0, v8, v10
 ; CHECK-NEXT:    vmsleu.vv v14, v10, v12, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v14
@@ -544,8 +544,8 @@ declare <vscale x 16 x i1> @llvm.riscv.vmsleu.mask.nxv16i16(
 define <vscale x 16 x i1> @intrinsic_vmsleu_mask_vv_nxv16i16_nxv16i16(<vscale x 16 x i1> %0, <vscale x 16 x i16> %1, <vscale x 16 x i16> %2, <vscale x 16 x i16> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsleu_mask_vv_nxv16i16_nxv16i16:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v20, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e16, m4, ta, mu
+; CHECK-NEXT:    vmv1r.v v20, v0
 ; CHECK-NEXT:    vmsleu.vv v0, v8, v12
 ; CHECK-NEXT:    vmsleu.vv v20, v12, v16, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v20
@@ -595,8 +595,8 @@ declare <vscale x 1 x i1> @llvm.riscv.vmsleu.mask.nxv1i32(
 define <vscale x 1 x i1> @intrinsic_vmsleu_mask_vv_nxv1i32_nxv1i32(<vscale x 1 x i1> %0, <vscale x 1 x i32> %1, <vscale x 1 x i32> %2, <vscale x 1 x i32> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsleu_mask_vv_nxv1i32_nxv1i32:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e32, mf2, ta, mu
+; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmsleu.vv v0, v8, v9
 ; CHECK-NEXT:    vmsleu.vv v11, v9, v10, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v11
@@ -646,8 +646,8 @@ declare <vscale x 2 x i1> @llvm.riscv.vmsleu.mask.nxv2i32(
 define <vscale x 2 x i1> @intrinsic_vmsleu_mask_vv_nxv2i32_nxv2i32(<vscale x 2 x i1> %0, <vscale x 2 x i32> %1, <vscale x 2 x i32> %2, <vscale x 2 x i32> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsleu_mask_vv_nxv2i32_nxv2i32:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e32, m1, ta, mu
+; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmsleu.vv v0, v8, v9
 ; CHECK-NEXT:    vmsleu.vv v11, v9, v10, v0.t
 ; CHECK-NEXT:    vmv.v.v v0, v11
@@ -697,8 +697,8 @@ declare <vscale x 4 x i1> @llvm.riscv.vmsleu.mask.nxv4i32(
 define <vscale x 4 x i1> @intrinsic_vmsleu_mask_vv_nxv4i32_nxv4i32(<vscale x 4 x i1> %0, <vscale x 4 x i32> %1, <vscale x 4 x i32> %2, <vscale x 4 x i32> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsleu_mask_vv_nxv4i32_nxv4i32:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v14, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e32, m2, ta, mu
+; CHECK-NEXT:    vmv1r.v v14, v0
 ; CHECK-NEXT:    vmsleu.vv v0, v8, v10
 ; CHECK-NEXT:    vmsleu.vv v14, v10, v12, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v14
@@ -748,8 +748,8 @@ declare <vscale x 8 x i1> @llvm.riscv.vmsleu.mask.nxv8i32(
 define <vscale x 8 x i1> @intrinsic_vmsleu_mask_vv_nxv8i32_nxv8i32(<vscale x 8 x i1> %0, <vscale x 8 x i32> %1, <vscale x 8 x i32> %2, <vscale x 8 x i32> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsleu_mask_vv_nxv8i32_nxv8i32:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v20, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e32, m4, ta, mu
+; CHECK-NEXT:    vmv1r.v v20, v0
 ; CHECK-NEXT:    vmsleu.vv v0, v8, v12
 ; CHECK-NEXT:    vmsleu.vv v20, v12, v16, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v20
@@ -799,8 +799,8 @@ declare <vscale x 1 x i1> @llvm.riscv.vmsleu.mask.nxv1i64(
 define <vscale x 1 x i1> @intrinsic_vmsleu_mask_vv_nxv1i64_nxv1i64(<vscale x 1 x i1> %0, <vscale x 1 x i64> %1, <vscale x 1 x i64> %2, <vscale x 1 x i64> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsleu_mask_vv_nxv1i64_nxv1i64:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e64, m1, ta, mu
+; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmsleu.vv v0, v8, v9
 ; CHECK-NEXT:    vmsleu.vv v11, v9, v10, v0.t
 ; CHECK-NEXT:    vmv.v.v v0, v11
@@ -850,8 +850,8 @@ declare <vscale x 2 x i1> @llvm.riscv.vmsleu.mask.nxv2i64(
 define <vscale x 2 x i1> @intrinsic_vmsleu_mask_vv_nxv2i64_nxv2i64(<vscale x 2 x i1> %0, <vscale x 2 x i64> %1, <vscale x 2 x i64> %2, <vscale x 2 x i64> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsleu_mask_vv_nxv2i64_nxv2i64:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v14, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e64, m2, ta, mu
+; CHECK-NEXT:    vmv1r.v v14, v0
 ; CHECK-NEXT:    vmsleu.vv v0, v8, v10
 ; CHECK-NEXT:    vmsleu.vv v14, v10, v12, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v14
@@ -901,8 +901,8 @@ declare <vscale x 4 x i1> @llvm.riscv.vmsleu.mask.nxv4i64(
 define <vscale x 4 x i1> @intrinsic_vmsleu_mask_vv_nxv4i64_nxv4i64(<vscale x 4 x i1> %0, <vscale x 4 x i64> %1, <vscale x 4 x i64> %2, <vscale x 4 x i64> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsleu_mask_vv_nxv4i64_nxv4i64:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v20, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e64, m4, ta, mu
+; CHECK-NEXT:    vmv1r.v v20, v0
 ; CHECK-NEXT:    vmsleu.vv v0, v8, v12
 ; CHECK-NEXT:    vmsleu.vv v20, v12, v16, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v20
@@ -952,9 +952,9 @@ declare <vscale x 1 x i1> @llvm.riscv.vmsleu.mask.nxv1i8.i8(
 define <vscale x 1 x i1> @intrinsic_vmsleu_mask_vx_nxv1i8_i8(<vscale x 1 x i1> %0, <vscale x 1 x i8> %1, i8 %2, <vscale x 1 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsleu_mask_vx_nxv1i8_i8:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e8, mf8, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a1, e8, mf8, ta, mu
 ; CHECK-NEXT:    vmsleu.vx v10, v8, a0, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v10
 ; CHECK-NEXT:    ret
@@ -999,9 +999,9 @@ declare <vscale x 2 x i1> @llvm.riscv.vmsleu.mask.nxv2i8.i8(
 define <vscale x 2 x i1> @intrinsic_vmsleu_mask_vx_nxv2i8_i8(<vscale x 2 x i1> %0, <vscale x 2 x i8> %1, i8 %2, <vscale x 2 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsleu_mask_vx_nxv2i8_i8:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e8, mf4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a1, e8, mf4, ta, mu
 ; CHECK-NEXT:    vmsleu.vx v10, v8, a0, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v10
 ; CHECK-NEXT:    ret
@@ -1046,9 +1046,9 @@ declare <vscale x 4 x i1> @llvm.riscv.vmsleu.mask.nxv4i8.i8(
 define <vscale x 4 x i1> @intrinsic_vmsleu_mask_vx_nxv4i8_i8(<vscale x 4 x i1> %0, <vscale x 4 x i8> %1, i8 %2, <vscale x 4 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsleu_mask_vx_nxv4i8_i8:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e8, mf2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a1, e8, mf2, ta, mu
 ; CHECK-NEXT:    vmsleu.vx v10, v8, a0, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v10
 ; CHECK-NEXT:    ret
@@ -1093,9 +1093,9 @@ declare <vscale x 8 x i1> @llvm.riscv.vmsleu.mask.nxv8i8.i8(
 define <vscale x 8 x i1> @intrinsic_vmsleu_mask_vx_nxv8i8_i8(<vscale x 8 x i1> %0, <vscale x 8 x i8> %1, i8 %2, <vscale x 8 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsleu_mask_vx_nxv8i8_i8:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e8, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a1, e8, m1, ta, mu
 ; CHECK-NEXT:    vmsleu.vx v10, v8, a0, v0.t
 ; CHECK-NEXT:    vmv.v.v v0, v10
 ; CHECK-NEXT:    ret
@@ -1140,9 +1140,9 @@ declare <vscale x 16 x i1> @llvm.riscv.vmsleu.mask.nxv16i8.i8(
 define <vscale x 16 x i1> @intrinsic_vmsleu_mask_vx_nxv16i8_i8(<vscale x 16 x i1> %0, <vscale x 16 x i8> %1, i8 %2, <vscale x 16 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsleu_mask_vx_nxv16i8_i8:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e8, m2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmv1r.v v0, v10
-; CHECK-NEXT:    vsetvli zero, a1, e8, m2, ta, mu
 ; CHECK-NEXT:    vmsleu.vx v11, v8, a0, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v11
 ; CHECK-NEXT:    ret
@@ -1187,9 +1187,9 @@ declare <vscale x 32 x i1> @llvm.riscv.vmsleu.mask.nxv32i8.i8(
 define <vscale x 32 x i1> @intrinsic_vmsleu_mask_vx_nxv32i8_i8(<vscale x 32 x i1> %0, <vscale x 32 x i8> %1, i8 %2, <vscale x 32 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsleu_mask_vx_nxv32i8_i8:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e8, m4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v13, v0
 ; CHECK-NEXT:    vmv1r.v v0, v12
-; CHECK-NEXT:    vsetvli zero, a1, e8, m4, ta, mu
 ; CHECK-NEXT:    vmsleu.vx v13, v8, a0, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v13
 ; CHECK-NEXT:    ret
@@ -1234,9 +1234,9 @@ declare <vscale x 1 x i1> @llvm.riscv.vmsleu.mask.nxv1i16.i16(
 define <vscale x 1 x i1> @intrinsic_vmsleu_mask_vx_nxv1i16_i16(<vscale x 1 x i1> %0, <vscale x 1 x i16> %1, i16 %2, <vscale x 1 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsleu_mask_vx_nxv1i16_i16:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, mf4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a1, e16, mf4, ta, mu
 ; CHECK-NEXT:    vmsleu.vx v10, v8, a0, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v10
 ; CHECK-NEXT:    ret
@@ -1281,9 +1281,9 @@ declare <vscale x 2 x i1> @llvm.riscv.vmsleu.mask.nxv2i16.i16(
 define <vscale x 2 x i1> @intrinsic_vmsleu_mask_vx_nxv2i16_i16(<vscale x 2 x i1> %0, <vscale x 2 x i16> %1, i16 %2, <vscale x 2 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsleu_mask_vx_nxv2i16_i16:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, mf2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a1, e16, mf2, ta, mu
 ; CHECK-NEXT:    vmsleu.vx v10, v8, a0, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v10
 ; CHECK-NEXT:    ret
@@ -1328,9 +1328,9 @@ declare <vscale x 4 x i1> @llvm.riscv.vmsleu.mask.nxv4i16.i16(
 define <vscale x 4 x i1> @intrinsic_vmsleu_mask_vx_nxv4i16_i16(<vscale x 4 x i1> %0, <vscale x 4 x i16> %1, i16 %2, <vscale x 4 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsleu_mask_vx_nxv4i16_i16:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a1, e16, m1, ta, mu
 ; CHECK-NEXT:    vmsleu.vx v10, v8, a0, v0.t
 ; CHECK-NEXT:    vmv.v.v v0, v10
 ; CHECK-NEXT:    ret
@@ -1375,9 +1375,9 @@ declare <vscale x 8 x i1> @llvm.riscv.vmsleu.mask.nxv8i16.i16(
 define <vscale x 8 x i1> @intrinsic_vmsleu_mask_vx_nxv8i16_i16(<vscale x 8 x i1> %0, <vscale x 8 x i16> %1, i16 %2, <vscale x 8 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsleu_mask_vx_nxv8i16_i16:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, m2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmv1r.v v0, v10
-; CHECK-NEXT:    vsetvli zero, a1, e16, m2, ta, mu
 ; CHECK-NEXT:    vmsleu.vx v11, v8, a0, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v11
 ; CHECK-NEXT:    ret
@@ -1422,9 +1422,9 @@ declare <vscale x 16 x i1> @llvm.riscv.vmsleu.mask.nxv16i16.i16(
 define <vscale x 16 x i1> @intrinsic_vmsleu_mask_vx_nxv16i16_i16(<vscale x 16 x i1> %0, <vscale x 16 x i16> %1, i16 %2, <vscale x 16 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsleu_mask_vx_nxv16i16_i16:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, m4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v13, v0
 ; CHECK-NEXT:    vmv1r.v v0, v12
-; CHECK-NEXT:    vsetvli zero, a1, e16, m4, ta, mu
 ; CHECK-NEXT:    vmsleu.vx v13, v8, a0, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v13
 ; CHECK-NEXT:    ret
@@ -1469,9 +1469,9 @@ declare <vscale x 1 x i1> @llvm.riscv.vmsleu.mask.nxv1i32.i32(
 define <vscale x 1 x i1> @intrinsic_vmsleu_mask_vx_nxv1i32_i32(<vscale x 1 x i1> %0, <vscale x 1 x i32> %1, i32 %2, <vscale x 1 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsleu_mask_vx_nxv1i32_i32:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e32, mf2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a1, e32, mf2, ta, mu
 ; CHECK-NEXT:    vmsleu.vx v10, v8, a0, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v10
 ; CHECK-NEXT:    ret
@@ -1516,9 +1516,9 @@ declare <vscale x 2 x i1> @llvm.riscv.vmsleu.mask.nxv2i32.i32(
 define <vscale x 2 x i1> @intrinsic_vmsleu_mask_vx_nxv2i32_i32(<vscale x 2 x i1> %0, <vscale x 2 x i32> %1, i32 %2, <vscale x 2 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsleu_mask_vx_nxv2i32_i32:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e32, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a1, e32, m1, ta, mu
 ; CHECK-NEXT:    vmsleu.vx v10, v8, a0, v0.t
 ; CHECK-NEXT:    vmv.v.v v0, v10
 ; CHECK-NEXT:    ret
@@ -1563,9 +1563,9 @@ declare <vscale x 4 x i1> @llvm.riscv.vmsleu.mask.nxv4i32.i32(
 define <vscale x 4 x i1> @intrinsic_vmsleu_mask_vx_nxv4i32_i32(<vscale x 4 x i1> %0, <vscale x 4 x i32> %1, i32 %2, <vscale x 4 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsleu_mask_vx_nxv4i32_i32:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e32, m2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmv1r.v v0, v10
-; CHECK-NEXT:    vsetvli zero, a1, e32, m2, ta, mu
 ; CHECK-NEXT:    vmsleu.vx v11, v8, a0, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v11
 ; CHECK-NEXT:    ret
@@ -1610,9 +1610,9 @@ declare <vscale x 8 x i1> @llvm.riscv.vmsleu.mask.nxv8i32.i32(
 define <vscale x 8 x i1> @intrinsic_vmsleu_mask_vx_nxv8i32_i32(<vscale x 8 x i1> %0, <vscale x 8 x i32> %1, i32 %2, <vscale x 8 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsleu_mask_vx_nxv8i32_i32:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e32, m4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v13, v0
 ; CHECK-NEXT:    vmv1r.v v0, v12
-; CHECK-NEXT:    vsetvli zero, a1, e32, m4, ta, mu
 ; CHECK-NEXT:    vmsleu.vx v13, v8, a0, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v13
 ; CHECK-NEXT:    ret
@@ -1684,9 +1684,9 @@ define <vscale x 1 x i1> @intrinsic_vmsleu_mask_vx_nxv1i64_i64(<vscale x 1 x i1>
 ;
 ; RV64-LABEL: intrinsic_vmsleu_mask_vx_nxv1i64_i64:
 ; RV64:       # %bb.0: # %entry
+; RV64-NEXT:    vsetvli zero, a1, e64, m1, ta, mu
 ; RV64-NEXT:    vmv1r.v v10, v0
 ; RV64-NEXT:    vmv1r.v v0, v9
-; RV64-NEXT:    vsetvli zero, a1, e64, m1, ta, mu
 ; RV64-NEXT:    vmsleu.vx v10, v8, a0, v0.t
 ; RV64-NEXT:    vmv.v.v v0, v10
 ; RV64-NEXT:    ret
@@ -1758,9 +1758,9 @@ define <vscale x 2 x i1> @intrinsic_vmsleu_mask_vx_nxv2i64_i64(<vscale x 2 x i1>
 ;
 ; RV64-LABEL: intrinsic_vmsleu_mask_vx_nxv2i64_i64:
 ; RV64:       # %bb.0: # %entry
+; RV64-NEXT:    vsetvli zero, a1, e64, m2, ta, mu
 ; RV64-NEXT:    vmv1r.v v11, v0
 ; RV64-NEXT:    vmv1r.v v0, v10
-; RV64-NEXT:    vsetvli zero, a1, e64, m2, ta, mu
 ; RV64-NEXT:    vmsleu.vx v11, v8, a0, v0.t
 ; RV64-NEXT:    vmv1r.v v0, v11
 ; RV64-NEXT:    ret
@@ -1832,9 +1832,9 @@ define <vscale x 4 x i1> @intrinsic_vmsleu_mask_vx_nxv4i64_i64(<vscale x 4 x i1>
 ;
 ; RV64-LABEL: intrinsic_vmsleu_mask_vx_nxv4i64_i64:
 ; RV64:       # %bb.0: # %entry
+; RV64-NEXT:    vsetvli zero, a1, e64, m4, ta, mu
 ; RV64-NEXT:    vmv1r.v v13, v0
 ; RV64-NEXT:    vmv1r.v v0, v12
-; RV64-NEXT:    vsetvli zero, a1, e64, m4, ta, mu
 ; RV64-NEXT:    vmsleu.vx v13, v8, a0, v0.t
 ; RV64-NEXT:    vmv1r.v v0, v13
 ; RV64-NEXT:    ret
@@ -1867,9 +1867,9 @@ entry:
 define <vscale x 1 x i1> @intrinsic_vmsleu_mask_vi_nxv1i8_i8(<vscale x 1 x i1> %0, <vscale x 1 x i8> %1, <vscale x 1 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsleu_mask_vi_nxv1i8_i8:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e8, mf8, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a0, e8, mf8, ta, mu
 ; CHECK-NEXT:    vmsleu.vi v10, v8, 9, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v10
 ; CHECK-NEXT:    ret
@@ -1902,9 +1902,9 @@ entry:
 define <vscale x 2 x i1> @intrinsic_vmsleu_mask_vi_nxv2i8_i8(<vscale x 2 x i1> %0, <vscale x 2 x i8> %1, <vscale x 2 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsleu_mask_vi_nxv2i8_i8:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e8, mf4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a0, e8, mf4, ta, mu
 ; CHECK-NEXT:    vmsleu.vi v10, v8, 9, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v10
 ; CHECK-NEXT:    ret
@@ -1937,9 +1937,9 @@ entry:
 define <vscale x 4 x i1> @intrinsic_vmsleu_mask_vi_nxv4i8_i8(<vscale x 4 x i1> %0, <vscale x 4 x i8> %1, <vscale x 4 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsleu_mask_vi_nxv4i8_i8:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e8, mf2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a0, e8, mf2, ta, mu
 ; CHECK-NEXT:    vmsleu.vi v10, v8, 9, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v10
 ; CHECK-NEXT:    ret
@@ -1972,9 +1972,9 @@ entry:
 define <vscale x 8 x i1> @intrinsic_vmsleu_mask_vi_nxv8i8_i8(<vscale x 8 x i1> %0, <vscale x 8 x i8> %1, <vscale x 8 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsleu_mask_vi_nxv8i8_i8:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e8, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a0, e8, m1, ta, mu
 ; CHECK-NEXT:    vmsleu.vi v10, v8, 9, v0.t
 ; CHECK-NEXT:    vmv.v.v v0, v10
 ; CHECK-NEXT:    ret
@@ -2007,9 +2007,9 @@ entry:
 define <vscale x 16 x i1> @intrinsic_vmsleu_mask_vi_nxv16i8_i8(<vscale x 16 x i1> %0, <vscale x 16 x i8> %1, <vscale x 16 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsleu_mask_vi_nxv16i8_i8:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e8, m2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmv1r.v v0, v10
-; CHECK-NEXT:    vsetvli zero, a0, e8, m2, ta, mu
 ; CHECK-NEXT:    vmsleu.vi v11, v8, 9, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v11
 ; CHECK-NEXT:    ret
@@ -2042,9 +2042,9 @@ entry:
 define <vscale x 32 x i1> @intrinsic_vmsleu_mask_vi_nxv32i8_i8(<vscale x 32 x i1> %0, <vscale x 32 x i8> %1, <vscale x 32 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsleu_mask_vi_nxv32i8_i8:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e8, m4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v13, v0
 ; CHECK-NEXT:    vmv1r.v v0, v12
-; CHECK-NEXT:    vsetvli zero, a0, e8, m4, ta, mu
 ; CHECK-NEXT:    vmsleu.vi v13, v8, 9, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v13
 ; CHECK-NEXT:    ret
@@ -2077,9 +2077,9 @@ entry:
 define <vscale x 1 x i1> @intrinsic_vmsleu_mask_vi_nxv1i16_i16(<vscale x 1 x i1> %0, <vscale x 1 x i16> %1, <vscale x 1 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsleu_mask_vi_nxv1i16_i16:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e16, mf4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a0, e16, mf4, ta, mu
 ; CHECK-NEXT:    vmsleu.vi v10, v8, 9, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v10
 ; CHECK-NEXT:    ret
@@ -2112,9 +2112,9 @@ entry:
 define <vscale x 2 x i1> @intrinsic_vmsleu_mask_vi_nxv2i16_i16(<vscale x 2 x i1> %0, <vscale x 2 x i16> %1, <vscale x 2 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsleu_mask_vi_nxv2i16_i16:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e16, mf2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a0, e16, mf2, ta, mu
 ; CHECK-NEXT:    vmsleu.vi v10, v8, 9, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v10
 ; CHECK-NEXT:    ret
@@ -2147,9 +2147,9 @@ entry:
 define <vscale x 4 x i1> @intrinsic_vmsleu_mask_vi_nxv4i16_i16(<vscale x 4 x i1> %0, <vscale x 4 x i16> %1, <vscale x 4 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsleu_mask_vi_nxv4i16_i16:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e16, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a0, e16, m1, ta, mu
 ; CHECK-NEXT:    vmsleu.vi v10, v8, 9, v0.t
 ; CHECK-NEXT:    vmv.v.v v0, v10
 ; CHECK-NEXT:    ret
@@ -2182,9 +2182,9 @@ entry:
 define <vscale x 8 x i1> @intrinsic_vmsleu_mask_vi_nxv8i16_i16(<vscale x 8 x i1> %0, <vscale x 8 x i16> %1, <vscale x 8 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsleu_mask_vi_nxv8i16_i16:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e16, m2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmv1r.v v0, v10
-; CHECK-NEXT:    vsetvli zero, a0, e16, m2, ta, mu
 ; CHECK-NEXT:    vmsleu.vi v11, v8, 9, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v11
 ; CHECK-NEXT:    ret
@@ -2217,9 +2217,9 @@ entry:
 define <vscale x 16 x i1> @intrinsic_vmsleu_mask_vi_nxv16i16_i16(<vscale x 16 x i1> %0, <vscale x 16 x i16> %1, <vscale x 16 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsleu_mask_vi_nxv16i16_i16:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e16, m4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v13, v0
 ; CHECK-NEXT:    vmv1r.v v0, v12
-; CHECK-NEXT:    vsetvli zero, a0, e16, m4, ta, mu
 ; CHECK-NEXT:    vmsleu.vi v13, v8, 9, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v13
 ; CHECK-NEXT:    ret
@@ -2252,9 +2252,9 @@ entry:
 define <vscale x 1 x i1> @intrinsic_vmsleu_mask_vi_nxv1i32_i32(<vscale x 1 x i1> %0, <vscale x 1 x i32> %1, <vscale x 1 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsleu_mask_vi_nxv1i32_i32:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e32, mf2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a0, e32, mf2, ta, mu
 ; CHECK-NEXT:    vmsleu.vi v10, v8, 9, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v10
 ; CHECK-NEXT:    ret
@@ -2287,9 +2287,9 @@ entry:
 define <vscale x 2 x i1> @intrinsic_vmsleu_mask_vi_nxv2i32_i32(<vscale x 2 x i1> %0, <vscale x 2 x i32> %1, <vscale x 2 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsleu_mask_vi_nxv2i32_i32:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e32, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a0, e32, m1, ta, mu
 ; CHECK-NEXT:    vmsleu.vi v10, v8, 9, v0.t
 ; CHECK-NEXT:    vmv.v.v v0, v10
 ; CHECK-NEXT:    ret
@@ -2322,9 +2322,9 @@ entry:
 define <vscale x 4 x i1> @intrinsic_vmsleu_mask_vi_nxv4i32_i32(<vscale x 4 x i1> %0, <vscale x 4 x i32> %1, <vscale x 4 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsleu_mask_vi_nxv4i32_i32:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e32, m2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmv1r.v v0, v10
-; CHECK-NEXT:    vsetvli zero, a0, e32, m2, ta, mu
 ; CHECK-NEXT:    vmsleu.vi v11, v8, 9, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v11
 ; CHECK-NEXT:    ret
@@ -2357,9 +2357,9 @@ entry:
 define <vscale x 8 x i1> @intrinsic_vmsleu_mask_vi_nxv8i32_i32(<vscale x 8 x i1> %0, <vscale x 8 x i32> %1, <vscale x 8 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsleu_mask_vi_nxv8i32_i32:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e32, m4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v13, v0
 ; CHECK-NEXT:    vmv1r.v v0, v12
-; CHECK-NEXT:    vsetvli zero, a0, e32, m4, ta, mu
 ; CHECK-NEXT:    vmsleu.vi v13, v8, 9, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v13
 ; CHECK-NEXT:    ret
@@ -2392,9 +2392,9 @@ entry:
 define <vscale x 1 x i1> @intrinsic_vmsleu_mask_vi_nxv1i64_i64(<vscale x 1 x i1> %0, <vscale x 1 x i64> %1, <vscale x 1 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsleu_mask_vi_nxv1i64_i64:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e64, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a0, e64, m1, ta, mu
 ; CHECK-NEXT:    vmsleu.vi v10, v8, 9, v0.t
 ; CHECK-NEXT:    vmv.v.v v0, v10
 ; CHECK-NEXT:    ret
@@ -2427,9 +2427,9 @@ entry:
 define <vscale x 2 x i1> @intrinsic_vmsleu_mask_vi_nxv2i64_i64(<vscale x 2 x i1> %0, <vscale x 2 x i64> %1, <vscale x 2 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsleu_mask_vi_nxv2i64_i64:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e64, m2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmv1r.v v0, v10
-; CHECK-NEXT:    vsetvli zero, a0, e64, m2, ta, mu
 ; CHECK-NEXT:    vmsleu.vi v11, v8, 9, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v11
 ; CHECK-NEXT:    ret
@@ -2462,9 +2462,9 @@ entry:
 define <vscale x 4 x i1> @intrinsic_vmsleu_mask_vi_nxv4i64_i64(<vscale x 4 x i1> %0, <vscale x 4 x i64> %1, <vscale x 4 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsleu_mask_vi_nxv4i64_i64:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e64, m4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v13, v0
 ; CHECK-NEXT:    vmv1r.v v0, v12
-; CHECK-NEXT:    vsetvli zero, a0, e64, m4, ta, mu
 ; CHECK-NEXT:    vmsleu.vi v13, v8, 9, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v13
 ; CHECK-NEXT:    ret
diff --git a/llvm/test/CodeGen/RISCV/rvv/vmslt.ll b/llvm/test/CodeGen/RISCV/rvv/vmslt.ll
index 407f85b4f5996..c17495e3b2119 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vmslt.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vmslt.ll
@@ -34,8 +34,8 @@ declare <vscale x 1 x i1> @llvm.riscv.vmslt.mask.nxv1i8(
 define <vscale x 1 x i1> @intrinsic_vmslt_mask_vv_nxv1i8_nxv1i8(<vscale x 1 x i1> %0, <vscale x 1 x i8> %1, <vscale x 1 x i8> %2, <vscale x 1 x i8> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmslt_mask_vv_nxv1i8_nxv1i8:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e8, mf8, ta, mu
+; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmslt.vv v0, v8, v9
 ; CHECK-NEXT:    vmslt.vv v11, v9, v10, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v11
@@ -85,8 +85,8 @@ declare <vscale x 2 x i1> @llvm.riscv.vmslt.mask.nxv2i8(
 define <vscale x 2 x i1> @intrinsic_vmslt_mask_vv_nxv2i8_nxv2i8(<vscale x 2 x i1> %0, <vscale x 2 x i8> %1, <vscale x 2 x i8> %2, <vscale x 2 x i8> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmslt_mask_vv_nxv2i8_nxv2i8:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e8, mf4, ta, mu
+; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmslt.vv v0, v8, v9
 ; CHECK-NEXT:    vmslt.vv v11, v9, v10, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v11
@@ -136,8 +136,8 @@ declare <vscale x 4 x i1> @llvm.riscv.vmslt.mask.nxv4i8(
 define <vscale x 4 x i1> @intrinsic_vmslt_mask_vv_nxv4i8_nxv4i8(<vscale x 4 x i1> %0, <vscale x 4 x i8> %1, <vscale x 4 x i8> %2, <vscale x 4 x i8> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmslt_mask_vv_nxv4i8_nxv4i8:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e8, mf2, ta, mu
+; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmslt.vv v0, v8, v9
 ; CHECK-NEXT:    vmslt.vv v11, v9, v10, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v11
@@ -187,8 +187,8 @@ declare <vscale x 8 x i1> @llvm.riscv.vmslt.mask.nxv8i8(
 define <vscale x 8 x i1> @intrinsic_vmslt_mask_vv_nxv8i8_nxv8i8(<vscale x 8 x i1> %0, <vscale x 8 x i8> %1, <vscale x 8 x i8> %2, <vscale x 8 x i8> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmslt_mask_vv_nxv8i8_nxv8i8:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e8, m1, ta, mu
+; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmslt.vv v0, v8, v9
 ; CHECK-NEXT:    vmslt.vv v11, v9, v10, v0.t
 ; CHECK-NEXT:    vmv.v.v v0, v11
@@ -238,8 +238,8 @@ declare <vscale x 16 x i1> @llvm.riscv.vmslt.mask.nxv16i8(
 define <vscale x 16 x i1> @intrinsic_vmslt_mask_vv_nxv16i8_nxv16i8(<vscale x 16 x i1> %0, <vscale x 16 x i8> %1, <vscale x 16 x i8> %2, <vscale x 16 x i8> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmslt_mask_vv_nxv16i8_nxv16i8:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v14, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e8, m2, ta, mu
+; CHECK-NEXT:    vmv1r.v v14, v0
 ; CHECK-NEXT:    vmslt.vv v0, v8, v10
 ; CHECK-NEXT:    vmslt.vv v14, v10, v12, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v14
@@ -289,8 +289,8 @@ declare <vscale x 32 x i1> @llvm.riscv.vmslt.mask.nxv32i8(
 define <vscale x 32 x i1> @intrinsic_vmslt_mask_vv_nxv32i8_nxv32i8(<vscale x 32 x i1> %0, <vscale x 32 x i8> %1, <vscale x 32 x i8> %2, <vscale x 32 x i8> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmslt_mask_vv_nxv32i8_nxv32i8:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v20, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e8, m4, ta, mu
+; CHECK-NEXT:    vmv1r.v v20, v0
 ; CHECK-NEXT:    vmslt.vv v0, v8, v12
 ; CHECK-NEXT:    vmslt.vv v20, v12, v16, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v20
@@ -340,8 +340,8 @@ declare <vscale x 1 x i1> @llvm.riscv.vmslt.mask.nxv1i16(
 define <vscale x 1 x i1> @intrinsic_vmslt_mask_vv_nxv1i16_nxv1i16(<vscale x 1 x i1> %0, <vscale x 1 x i16> %1, <vscale x 1 x i16> %2, <vscale x 1 x i16> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmslt_mask_vv_nxv1i16_nxv1i16:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e16, mf4, ta, mu
+; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmslt.vv v0, v8, v9
 ; CHECK-NEXT:    vmslt.vv v11, v9, v10, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v11
@@ -391,8 +391,8 @@ declare <vscale x 2 x i1> @llvm.riscv.vmslt.mask.nxv2i16(
 define <vscale x 2 x i1> @intrinsic_vmslt_mask_vv_nxv2i16_nxv2i16(<vscale x 2 x i1> %0, <vscale x 2 x i16> %1, <vscale x 2 x i16> %2, <vscale x 2 x i16> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmslt_mask_vv_nxv2i16_nxv2i16:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e16, mf2, ta, mu
+; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmslt.vv v0, v8, v9
 ; CHECK-NEXT:    vmslt.vv v11, v9, v10, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v11
@@ -442,8 +442,8 @@ declare <vscale x 4 x i1> @llvm.riscv.vmslt.mask.nxv4i16(
 define <vscale x 4 x i1> @intrinsic_vmslt_mask_vv_nxv4i16_nxv4i16(<vscale x 4 x i1> %0, <vscale x 4 x i16> %1, <vscale x 4 x i16> %2, <vscale x 4 x i16> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmslt_mask_vv_nxv4i16_nxv4i16:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e16, m1, ta, mu
+; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmslt.vv v0, v8, v9
 ; CHECK-NEXT:    vmslt.vv v11, v9, v10, v0.t
 ; CHECK-NEXT:    vmv.v.v v0, v11
@@ -493,8 +493,8 @@ declare <vscale x 8 x i1> @llvm.riscv.vmslt.mask.nxv8i16(
 define <vscale x 8 x i1> @intrinsic_vmslt_mask_vv_nxv8i16_nxv8i16(<vscale x 8 x i1> %0, <vscale x 8 x i16> %1, <vscale x 8 x i16> %2, <vscale x 8 x i16> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmslt_mask_vv_nxv8i16_nxv8i16:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v14, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e16, m2, ta, mu
+; CHECK-NEXT:    vmv1r.v v14, v0
 ; CHECK-NEXT:    vmslt.vv v0, v8, v10
 ; CHECK-NEXT:    vmslt.vv v14, v10, v12, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v14
@@ -544,8 +544,8 @@ declare <vscale x 16 x i1> @llvm.riscv.vmslt.mask.nxv16i16(
 define <vscale x 16 x i1> @intrinsic_vmslt_mask_vv_nxv16i16_nxv16i16(<vscale x 16 x i1> %0, <vscale x 16 x i16> %1, <vscale x 16 x i16> %2, <vscale x 16 x i16> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmslt_mask_vv_nxv16i16_nxv16i16:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v20, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e16, m4, ta, mu
+; CHECK-NEXT:    vmv1r.v v20, v0
 ; CHECK-NEXT:    vmslt.vv v0, v8, v12
 ; CHECK-NEXT:    vmslt.vv v20, v12, v16, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v20
@@ -595,8 +595,8 @@ declare <vscale x 1 x i1> @llvm.riscv.vmslt.mask.nxv1i32(
 define <vscale x 1 x i1> @intrinsic_vmslt_mask_vv_nxv1i32_nxv1i32(<vscale x 1 x i1> %0, <vscale x 1 x i32> %1, <vscale x 1 x i32> %2, <vscale x 1 x i32> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmslt_mask_vv_nxv1i32_nxv1i32:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e32, mf2, ta, mu
+; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmslt.vv v0, v8, v9
 ; CHECK-NEXT:    vmslt.vv v11, v9, v10, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v11
@@ -646,8 +646,8 @@ declare <vscale x 2 x i1> @llvm.riscv.vmslt.mask.nxv2i32(
 define <vscale x 2 x i1> @intrinsic_vmslt_mask_vv_nxv2i32_nxv2i32(<vscale x 2 x i1> %0, <vscale x 2 x i32> %1, <vscale x 2 x i32> %2, <vscale x 2 x i32> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmslt_mask_vv_nxv2i32_nxv2i32:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e32, m1, ta, mu
+; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmslt.vv v0, v8, v9
 ; CHECK-NEXT:    vmslt.vv v11, v9, v10, v0.t
 ; CHECK-NEXT:    vmv.v.v v0, v11
@@ -697,8 +697,8 @@ declare <vscale x 4 x i1> @llvm.riscv.vmslt.mask.nxv4i32(
 define <vscale x 4 x i1> @intrinsic_vmslt_mask_vv_nxv4i32_nxv4i32(<vscale x 4 x i1> %0, <vscale x 4 x i32> %1, <vscale x 4 x i32> %2, <vscale x 4 x i32> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmslt_mask_vv_nxv4i32_nxv4i32:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v14, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e32, m2, ta, mu
+; CHECK-NEXT:    vmv1r.v v14, v0
 ; CHECK-NEXT:    vmslt.vv v0, v8, v10
 ; CHECK-NEXT:    vmslt.vv v14, v10, v12, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v14
@@ -748,8 +748,8 @@ declare <vscale x 8 x i1> @llvm.riscv.vmslt.mask.nxv8i32(
 define <vscale x 8 x i1> @intrinsic_vmslt_mask_vv_nxv8i32_nxv8i32(<vscale x 8 x i1> %0, <vscale x 8 x i32> %1, <vscale x 8 x i32> %2, <vscale x 8 x i32> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmslt_mask_vv_nxv8i32_nxv8i32:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v20, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e32, m4, ta, mu
+; CHECK-NEXT:    vmv1r.v v20, v0
 ; CHECK-NEXT:    vmslt.vv v0, v8, v12
 ; CHECK-NEXT:    vmslt.vv v20, v12, v16, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v20
@@ -799,8 +799,8 @@ declare <vscale x 1 x i1> @llvm.riscv.vmslt.mask.nxv1i64(
 define <vscale x 1 x i1> @intrinsic_vmslt_mask_vv_nxv1i64_nxv1i64(<vscale x 1 x i1> %0, <vscale x 1 x i64> %1, <vscale x 1 x i64> %2, <vscale x 1 x i64> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmslt_mask_vv_nxv1i64_nxv1i64:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e64, m1, ta, mu
+; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmslt.vv v0, v8, v9
 ; CHECK-NEXT:    vmslt.vv v11, v9, v10, v0.t
 ; CHECK-NEXT:    vmv.v.v v0, v11
@@ -850,8 +850,8 @@ declare <vscale x 2 x i1> @llvm.riscv.vmslt.mask.nxv2i64(
 define <vscale x 2 x i1> @intrinsic_vmslt_mask_vv_nxv2i64_nxv2i64(<vscale x 2 x i1> %0, <vscale x 2 x i64> %1, <vscale x 2 x i64> %2, <vscale x 2 x i64> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmslt_mask_vv_nxv2i64_nxv2i64:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v14, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e64, m2, ta, mu
+; CHECK-NEXT:    vmv1r.v v14, v0
 ; CHECK-NEXT:    vmslt.vv v0, v8, v10
 ; CHECK-NEXT:    vmslt.vv v14, v10, v12, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v14
@@ -901,8 +901,8 @@ declare <vscale x 4 x i1> @llvm.riscv.vmslt.mask.nxv4i64(
 define <vscale x 4 x i1> @intrinsic_vmslt_mask_vv_nxv4i64_nxv4i64(<vscale x 4 x i1> %0, <vscale x 4 x i64> %1, <vscale x 4 x i64> %2, <vscale x 4 x i64> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmslt_mask_vv_nxv4i64_nxv4i64:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v20, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e64, m4, ta, mu
+; CHECK-NEXT:    vmv1r.v v20, v0
 ; CHECK-NEXT:    vmslt.vv v0, v8, v12
 ; CHECK-NEXT:    vmslt.vv v20, v12, v16, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v20
@@ -952,9 +952,9 @@ declare <vscale x 1 x i1> @llvm.riscv.vmslt.mask.nxv1i8.i8(
 define <vscale x 1 x i1> @intrinsic_vmslt_mask_vx_nxv1i8_i8(<vscale x 1 x i1> %0, <vscale x 1 x i8> %1, i8 %2, <vscale x 1 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmslt_mask_vx_nxv1i8_i8:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e8, mf8, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a1, e8, mf8, ta, mu
 ; CHECK-NEXT:    vmslt.vx v10, v8, a0, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v10
 ; CHECK-NEXT:    ret
@@ -999,9 +999,9 @@ declare <vscale x 2 x i1> @llvm.riscv.vmslt.mask.nxv2i8.i8(
 define <vscale x 2 x i1> @intrinsic_vmslt_mask_vx_nxv2i8_i8(<vscale x 2 x i1> %0, <vscale x 2 x i8> %1, i8 %2, <vscale x 2 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmslt_mask_vx_nxv2i8_i8:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e8, mf4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a1, e8, mf4, ta, mu
 ; CHECK-NEXT:    vmslt.vx v10, v8, a0, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v10
 ; CHECK-NEXT:    ret
@@ -1046,9 +1046,9 @@ declare <vscale x 4 x i1> @llvm.riscv.vmslt.mask.nxv4i8.i8(
 define <vscale x 4 x i1> @intrinsic_vmslt_mask_vx_nxv4i8_i8(<vscale x 4 x i1> %0, <vscale x 4 x i8> %1, i8 %2, <vscale x 4 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmslt_mask_vx_nxv4i8_i8:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e8, mf2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a1, e8, mf2, ta, mu
 ; CHECK-NEXT:    vmslt.vx v10, v8, a0, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v10
 ; CHECK-NEXT:    ret
@@ -1093,9 +1093,9 @@ declare <vscale x 8 x i1> @llvm.riscv.vmslt.mask.nxv8i8.i8(
 define <vscale x 8 x i1> @intrinsic_vmslt_mask_vx_nxv8i8_i8(<vscale x 8 x i1> %0, <vscale x 8 x i8> %1, i8 %2, <vscale x 8 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmslt_mask_vx_nxv8i8_i8:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e8, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a1, e8, m1, ta, mu
 ; CHECK-NEXT:    vmslt.vx v10, v8, a0, v0.t
 ; CHECK-NEXT:    vmv.v.v v0, v10
 ; CHECK-NEXT:    ret
@@ -1140,9 +1140,9 @@ declare <vscale x 16 x i1> @llvm.riscv.vmslt.mask.nxv16i8.i8(
 define <vscale x 16 x i1> @intrinsic_vmslt_mask_vx_nxv16i8_i8(<vscale x 16 x i1> %0, <vscale x 16 x i8> %1, i8 %2, <vscale x 16 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmslt_mask_vx_nxv16i8_i8:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e8, m2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmv1r.v v0, v10
-; CHECK-NEXT:    vsetvli zero, a1, e8, m2, ta, mu
 ; CHECK-NEXT:    vmslt.vx v11, v8, a0, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v11
 ; CHECK-NEXT:    ret
@@ -1187,9 +1187,9 @@ declare <vscale x 32 x i1> @llvm.riscv.vmslt.mask.nxv32i8.i8(
 define <vscale x 32 x i1> @intrinsic_vmslt_mask_vx_nxv32i8_i8(<vscale x 32 x i1> %0, <vscale x 32 x i8> %1, i8 %2, <vscale x 32 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmslt_mask_vx_nxv32i8_i8:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e8, m4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v13, v0
 ; CHECK-NEXT:    vmv1r.v v0, v12
-; CHECK-NEXT:    vsetvli zero, a1, e8, m4, ta, mu
 ; CHECK-NEXT:    vmslt.vx v13, v8, a0, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v13
 ; CHECK-NEXT:    ret
@@ -1234,9 +1234,9 @@ declare <vscale x 1 x i1> @llvm.riscv.vmslt.mask.nxv1i16.i16(
 define <vscale x 1 x i1> @intrinsic_vmslt_mask_vx_nxv1i16_i16(<vscale x 1 x i1> %0, <vscale x 1 x i16> %1, i16 %2, <vscale x 1 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmslt_mask_vx_nxv1i16_i16:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, mf4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a1, e16, mf4, ta, mu
 ; CHECK-NEXT:    vmslt.vx v10, v8, a0, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v10
 ; CHECK-NEXT:    ret
@@ -1281,9 +1281,9 @@ declare <vscale x 2 x i1> @llvm.riscv.vmslt.mask.nxv2i16.i16(
 define <vscale x 2 x i1> @intrinsic_vmslt_mask_vx_nxv2i16_i16(<vscale x 2 x i1> %0, <vscale x 2 x i16> %1, i16 %2, <vscale x 2 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmslt_mask_vx_nxv2i16_i16:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, mf2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a1, e16, mf2, ta, mu
 ; CHECK-NEXT:    vmslt.vx v10, v8, a0, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v10
 ; CHECK-NEXT:    ret
@@ -1328,9 +1328,9 @@ declare <vscale x 4 x i1> @llvm.riscv.vmslt.mask.nxv4i16.i16(
 define <vscale x 4 x i1> @intrinsic_vmslt_mask_vx_nxv4i16_i16(<vscale x 4 x i1> %0, <vscale x 4 x i16> %1, i16 %2, <vscale x 4 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmslt_mask_vx_nxv4i16_i16:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a1, e16, m1, ta, mu
 ; CHECK-NEXT:    vmslt.vx v10, v8, a0, v0.t
 ; CHECK-NEXT:    vmv.v.v v0, v10
 ; CHECK-NEXT:    ret
@@ -1375,9 +1375,9 @@ declare <vscale x 8 x i1> @llvm.riscv.vmslt.mask.nxv8i16.i16(
 define <vscale x 8 x i1> @intrinsic_vmslt_mask_vx_nxv8i16_i16(<vscale x 8 x i1> %0, <vscale x 8 x i16> %1, i16 %2, <vscale x 8 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmslt_mask_vx_nxv8i16_i16:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, m2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmv1r.v v0, v10
-; CHECK-NEXT:    vsetvli zero, a1, e16, m2, ta, mu
 ; CHECK-NEXT:    vmslt.vx v11, v8, a0, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v11
 ; CHECK-NEXT:    ret
@@ -1422,9 +1422,9 @@ declare <vscale x 16 x i1> @llvm.riscv.vmslt.mask.nxv16i16.i16(
 define <vscale x 16 x i1> @intrinsic_vmslt_mask_vx_nxv16i16_i16(<vscale x 16 x i1> %0, <vscale x 16 x i16> %1, i16 %2, <vscale x 16 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmslt_mask_vx_nxv16i16_i16:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, m4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v13, v0
 ; CHECK-NEXT:    vmv1r.v v0, v12
-; CHECK-NEXT:    vsetvli zero, a1, e16, m4, ta, mu
 ; CHECK-NEXT:    vmslt.vx v13, v8, a0, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v13
 ; CHECK-NEXT:    ret
@@ -1469,9 +1469,9 @@ declare <vscale x 1 x i1> @llvm.riscv.vmslt.mask.nxv1i32.i32(
 define <vscale x 1 x i1> @intrinsic_vmslt_mask_vx_nxv1i32_i32(<vscale x 1 x i1> %0, <vscale x 1 x i32> %1, i32 %2, <vscale x 1 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmslt_mask_vx_nxv1i32_i32:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e32, mf2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a1, e32, mf2, ta, mu
 ; CHECK-NEXT:    vmslt.vx v10, v8, a0, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v10
 ; CHECK-NEXT:    ret
@@ -1516,9 +1516,9 @@ declare <vscale x 2 x i1> @llvm.riscv.vmslt.mask.nxv2i32.i32(
 define <vscale x 2 x i1> @intrinsic_vmslt_mask_vx_nxv2i32_i32(<vscale x 2 x i1> %0, <vscale x 2 x i32> %1, i32 %2, <vscale x 2 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmslt_mask_vx_nxv2i32_i32:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e32, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a1, e32, m1, ta, mu
 ; CHECK-NEXT:    vmslt.vx v10, v8, a0, v0.t
 ; CHECK-NEXT:    vmv.v.v v0, v10
 ; CHECK-NEXT:    ret
@@ -1563,9 +1563,9 @@ declare <vscale x 4 x i1> @llvm.riscv.vmslt.mask.nxv4i32.i32(
 define <vscale x 4 x i1> @intrinsic_vmslt_mask_vx_nxv4i32_i32(<vscale x 4 x i1> %0, <vscale x 4 x i32> %1, i32 %2, <vscale x 4 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmslt_mask_vx_nxv4i32_i32:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e32, m2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmv1r.v v0, v10
-; CHECK-NEXT:    vsetvli zero, a1, e32, m2, ta, mu
 ; CHECK-NEXT:    vmslt.vx v11, v8, a0, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v11
 ; CHECK-NEXT:    ret
@@ -1610,9 +1610,9 @@ declare <vscale x 8 x i1> @llvm.riscv.vmslt.mask.nxv8i32.i32(
 define <vscale x 8 x i1> @intrinsic_vmslt_mask_vx_nxv8i32_i32(<vscale x 8 x i1> %0, <vscale x 8 x i32> %1, i32 %2, <vscale x 8 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmslt_mask_vx_nxv8i32_i32:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e32, m4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v13, v0
 ; CHECK-NEXT:    vmv1r.v v0, v12
-; CHECK-NEXT:    vsetvli zero, a1, e32, m4, ta, mu
 ; CHECK-NEXT:    vmslt.vx v13, v8, a0, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v13
 ; CHECK-NEXT:    ret
@@ -1684,9 +1684,9 @@ define <vscale x 1 x i1> @intrinsic_vmslt_mask_vx_nxv1i64_i64(<vscale x 1 x i1>
 ;
 ; RV64-LABEL: intrinsic_vmslt_mask_vx_nxv1i64_i64:
 ; RV64:       # %bb.0: # %entry
+; RV64-NEXT:    vsetvli zero, a1, e64, m1, ta, mu
 ; RV64-NEXT:    vmv1r.v v10, v0
 ; RV64-NEXT:    vmv1r.v v0, v9
-; RV64-NEXT:    vsetvli zero, a1, e64, m1, ta, mu
 ; RV64-NEXT:    vmslt.vx v10, v8, a0, v0.t
 ; RV64-NEXT:    vmv.v.v v0, v10
 ; RV64-NEXT:    ret
@@ -1758,9 +1758,9 @@ define <vscale x 2 x i1> @intrinsic_vmslt_mask_vx_nxv2i64_i64(<vscale x 2 x i1>
 ;
 ; RV64-LABEL: intrinsic_vmslt_mask_vx_nxv2i64_i64:
 ; RV64:       # %bb.0: # %entry
+; RV64-NEXT:    vsetvli zero, a1, e64, m2, ta, mu
 ; RV64-NEXT:    vmv1r.v v11, v0
 ; RV64-NEXT:    vmv1r.v v0, v10
-; RV64-NEXT:    vsetvli zero, a1, e64, m2, ta, mu
 ; RV64-NEXT:    vmslt.vx v11, v8, a0, v0.t
 ; RV64-NEXT:    vmv1r.v v0, v11
 ; RV64-NEXT:    ret
@@ -1832,9 +1832,9 @@ define <vscale x 4 x i1> @intrinsic_vmslt_mask_vx_nxv4i64_i64(<vscale x 4 x i1>
 ;
 ; RV64-LABEL: intrinsic_vmslt_mask_vx_nxv4i64_i64:
 ; RV64:       # %bb.0: # %entry
+; RV64-NEXT:    vsetvli zero, a1, e64, m4, ta, mu
 ; RV64-NEXT:    vmv1r.v v13, v0
 ; RV64-NEXT:    vmv1r.v v0, v12
-; RV64-NEXT:    vsetvli zero, a1, e64, m4, ta, mu
 ; RV64-NEXT:    vmslt.vx v13, v8, a0, v0.t
 ; RV64-NEXT:    vmv1r.v v0, v13
 ; RV64-NEXT:    ret
@@ -1867,9 +1867,9 @@ entry:
 define <vscale x 1 x i1> @intrinsic_vmslt_mask_vi_nxv1i8_i8(<vscale x 1 x i1> %0, <vscale x 1 x i8> %1, <vscale x 1 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmslt_mask_vi_nxv1i8_i8:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e8, mf8, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a0, e8, mf8, ta, mu
 ; CHECK-NEXT:    vmsle.vi v10, v8, -15, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v10
 ; CHECK-NEXT:    ret
@@ -1902,9 +1902,9 @@ entry:
 define <vscale x 2 x i1> @intrinsic_vmslt_mask_vi_nxv2i8_i8(<vscale x 2 x i1> %0, <vscale x 2 x i8> %1, <vscale x 2 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmslt_mask_vi_nxv2i8_i8:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e8, mf4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a0, e8, mf4, ta, mu
 ; CHECK-NEXT:    vmsle.vi v10, v8, -13, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v10
 ; CHECK-NEXT:    ret
@@ -1937,9 +1937,9 @@ entry:
 define <vscale x 4 x i1> @intrinsic_vmslt_mask_vi_nxv4i8_i8(<vscale x 4 x i1> %0, <vscale x 4 x i8> %1, <vscale x 4 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmslt_mask_vi_nxv4i8_i8:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e8, mf2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a0, e8, mf2, ta, mu
 ; CHECK-NEXT:    vmsle.vi v10, v8, -11, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v10
 ; CHECK-NEXT:    ret
@@ -1972,9 +1972,9 @@ entry:
 define <vscale x 8 x i1> @intrinsic_vmslt_mask_vi_nxv8i8_i8(<vscale x 8 x i1> %0, <vscale x 8 x i8> %1, <vscale x 8 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmslt_mask_vi_nxv8i8_i8:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e8, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a0, e8, m1, ta, mu
 ; CHECK-NEXT:    vmsle.vi v10, v8, -9, v0.t
 ; CHECK-NEXT:    vmv.v.v v0, v10
 ; CHECK-NEXT:    ret
@@ -2007,9 +2007,9 @@ entry:
 define <vscale x 16 x i1> @intrinsic_vmslt_mask_vi_nxv16i8_i8(<vscale x 16 x i1> %0, <vscale x 16 x i8> %1, <vscale x 16 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmslt_mask_vi_nxv16i8_i8:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e8, m2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmv1r.v v0, v10
-; CHECK-NEXT:    vsetvli zero, a0, e8, m2, ta, mu
 ; CHECK-NEXT:    vmsle.vi v11, v8, -7, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v11
 ; CHECK-NEXT:    ret
@@ -2042,9 +2042,9 @@ entry:
 define <vscale x 32 x i1> @intrinsic_vmslt_mask_vi_nxv32i8_i8(<vscale x 32 x i1> %0, <vscale x 32 x i8> %1, <vscale x 32 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmslt_mask_vi_nxv32i8_i8:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e8, m4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v13, v0
 ; CHECK-NEXT:    vmv1r.v v0, v12
-; CHECK-NEXT:    vsetvli zero, a0, e8, m4, ta, mu
 ; CHECK-NEXT:    vmsle.vi v13, v8, -5, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v13
 ; CHECK-NEXT:    ret
@@ -2077,9 +2077,9 @@ entry:
 define <vscale x 1 x i1> @intrinsic_vmslt_mask_vi_nxv1i16_i16(<vscale x 1 x i1> %0, <vscale x 1 x i16> %1, <vscale x 1 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmslt_mask_vi_nxv1i16_i16:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e16, mf4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a0, e16, mf4, ta, mu
 ; CHECK-NEXT:    vmsle.vi v10, v8, -3, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v10
 ; CHECK-NEXT:    ret
@@ -2112,9 +2112,9 @@ entry:
 define <vscale x 2 x i1> @intrinsic_vmslt_mask_vi_nxv2i16_i16(<vscale x 2 x i1> %0, <vscale x 2 x i16> %1, <vscale x 2 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmslt_mask_vi_nxv2i16_i16:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e16, mf2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a0, e16, mf2, ta, mu
 ; CHECK-NEXT:    vmslt.vx v10, v8, zero, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v10
 ; CHECK-NEXT:    ret
@@ -2147,9 +2147,9 @@ entry:
 define <vscale x 4 x i1> @intrinsic_vmslt_mask_vi_nxv4i16_i16(<vscale x 4 x i1> %0, <vscale x 4 x i16> %1, <vscale x 4 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmslt_mask_vi_nxv4i16_i16:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e16, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a0, e16, m1, ta, mu
 ; CHECK-NEXT:    vmsle.vi v10, v8, 0, v0.t
 ; CHECK-NEXT:    vmv.v.v v0, v10
 ; CHECK-NEXT:    ret
@@ -2182,9 +2182,9 @@ entry:
 define <vscale x 8 x i1> @intrinsic_vmslt_mask_vi_nxv8i16_i16(<vscale x 8 x i1> %0, <vscale x 8 x i16> %1, <vscale x 8 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmslt_mask_vi_nxv8i16_i16:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e16, m2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmv1r.v v0, v10
-; CHECK-NEXT:    vsetvli zero, a0, e16, m2, ta, mu
 ; CHECK-NEXT:    vmsle.vi v11, v8, 2, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v11
 ; CHECK-NEXT:    ret
@@ -2217,9 +2217,9 @@ entry:
 define <vscale x 16 x i1> @intrinsic_vmslt_mask_vi_nxv16i16_i16(<vscale x 16 x i1> %0, <vscale x 16 x i16> %1, <vscale x 16 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmslt_mask_vi_nxv16i16_i16:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e16, m4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v13, v0
 ; CHECK-NEXT:    vmv1r.v v0, v12
-; CHECK-NEXT:    vsetvli zero, a0, e16, m4, ta, mu
 ; CHECK-NEXT:    vmsle.vi v13, v8, 4, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v13
 ; CHECK-NEXT:    ret
@@ -2252,9 +2252,9 @@ entry:
 define <vscale x 1 x i1> @intrinsic_vmslt_mask_vi_nxv1i32_i32(<vscale x 1 x i1> %0, <vscale x 1 x i32> %1, <vscale x 1 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmslt_mask_vi_nxv1i32_i32:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e32, mf2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a0, e32, mf2, ta, mu
 ; CHECK-NEXT:    vmsle.vi v10, v8, 6, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v10
 ; CHECK-NEXT:    ret
@@ -2287,9 +2287,9 @@ entry:
 define <vscale x 2 x i1> @intrinsic_vmslt_mask_vi_nxv2i32_i32(<vscale x 2 x i1> %0, <vscale x 2 x i32> %1, <vscale x 2 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmslt_mask_vi_nxv2i32_i32:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e32, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a0, e32, m1, ta, mu
 ; CHECK-NEXT:    vmsle.vi v10, v8, 8, v0.t
 ; CHECK-NEXT:    vmv.v.v v0, v10
 ; CHECK-NEXT:    ret
@@ -2322,9 +2322,9 @@ entry:
 define <vscale x 4 x i1> @intrinsic_vmslt_mask_vi_nxv4i32_i32(<vscale x 4 x i1> %0, <vscale x 4 x i32> %1, <vscale x 4 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmslt_mask_vi_nxv4i32_i32:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e32, m2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmv1r.v v0, v10
-; CHECK-NEXT:    vsetvli zero, a0, e32, m2, ta, mu
 ; CHECK-NEXT:    vmsle.vi v11, v8, 10, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v11
 ; CHECK-NEXT:    ret
@@ -2357,9 +2357,9 @@ entry:
 define <vscale x 8 x i1> @intrinsic_vmslt_mask_vi_nxv8i32_i32(<vscale x 8 x i1> %0, <vscale x 8 x i32> %1, <vscale x 8 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmslt_mask_vi_nxv8i32_i32:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e32, m4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v13, v0
 ; CHECK-NEXT:    vmv1r.v v0, v12
-; CHECK-NEXT:    vsetvli zero, a0, e32, m4, ta, mu
 ; CHECK-NEXT:    vmsle.vi v13, v8, 12, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v13
 ; CHECK-NEXT:    ret
@@ -2392,9 +2392,9 @@ entry:
 define <vscale x 1 x i1> @intrinsic_vmslt_mask_vi_nxv1i64_i64(<vscale x 1 x i1> %0, <vscale x 1 x i64> %1, <vscale x 1 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmslt_mask_vi_nxv1i64_i64:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e64, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a0, e64, m1, ta, mu
 ; CHECK-NEXT:    vmsle.vi v10, v8, 8, v0.t
 ; CHECK-NEXT:    vmv.v.v v0, v10
 ; CHECK-NEXT:    ret
@@ -2427,9 +2427,9 @@ entry:
 define <vscale x 2 x i1> @intrinsic_vmslt_mask_vi_nxv2i64_i64(<vscale x 2 x i1> %0, <vscale x 2 x i64> %1, <vscale x 2 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmslt_mask_vi_nxv2i64_i64:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e64, m2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmv1r.v v0, v10
-; CHECK-NEXT:    vsetvli zero, a0, e64, m2, ta, mu
 ; CHECK-NEXT:    vmsle.vi v11, v8, 8, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v11
 ; CHECK-NEXT:    ret
@@ -2462,9 +2462,9 @@ entry:
 define <vscale x 4 x i1> @intrinsic_vmslt_mask_vi_nxv4i64_i64(<vscale x 4 x i1> %0, <vscale x 4 x i64> %1, <vscale x 4 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmslt_mask_vi_nxv4i64_i64:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e64, m4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v13, v0
 ; CHECK-NEXT:    vmv1r.v v0, v12
-; CHECK-NEXT:    vsetvli zero, a0, e64, m4, ta, mu
 ; CHECK-NEXT:    vmsle.vi v13, v8, 8, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v13
 ; CHECK-NEXT:    ret
diff --git a/llvm/test/CodeGen/RISCV/rvv/vmsltu.ll b/llvm/test/CodeGen/RISCV/rvv/vmsltu.ll
index e051b332018fd..a37a02848365d 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vmsltu.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vmsltu.ll
@@ -34,8 +34,8 @@ declare <vscale x 1 x i1> @llvm.riscv.vmsltu.mask.nxv1i8(
 define <vscale x 1 x i1> @intrinsic_vmsltu_mask_vv_nxv1i8_nxv1i8(<vscale x 1 x i1> %0, <vscale x 1 x i8> %1, <vscale x 1 x i8> %2, <vscale x 1 x i8> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsltu_mask_vv_nxv1i8_nxv1i8:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e8, mf8, ta, mu
+; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmsltu.vv v0, v8, v9
 ; CHECK-NEXT:    vmsltu.vv v11, v9, v10, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v11
@@ -85,8 +85,8 @@ declare <vscale x 2 x i1> @llvm.riscv.vmsltu.mask.nxv2i8(
 define <vscale x 2 x i1> @intrinsic_vmsltu_mask_vv_nxv2i8_nxv2i8(<vscale x 2 x i1> %0, <vscale x 2 x i8> %1, <vscale x 2 x i8> %2, <vscale x 2 x i8> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsltu_mask_vv_nxv2i8_nxv2i8:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e8, mf4, ta, mu
+; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmsltu.vv v0, v8, v9
 ; CHECK-NEXT:    vmsltu.vv v11, v9, v10, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v11
@@ -136,8 +136,8 @@ declare <vscale x 4 x i1> @llvm.riscv.vmsltu.mask.nxv4i8(
 define <vscale x 4 x i1> @intrinsic_vmsltu_mask_vv_nxv4i8_nxv4i8(<vscale x 4 x i1> %0, <vscale x 4 x i8> %1, <vscale x 4 x i8> %2, <vscale x 4 x i8> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsltu_mask_vv_nxv4i8_nxv4i8:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e8, mf2, ta, mu
+; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmsltu.vv v0, v8, v9
 ; CHECK-NEXT:    vmsltu.vv v11, v9, v10, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v11
@@ -187,8 +187,8 @@ declare <vscale x 8 x i1> @llvm.riscv.vmsltu.mask.nxv8i8(
 define <vscale x 8 x i1> @intrinsic_vmsltu_mask_vv_nxv8i8_nxv8i8(<vscale x 8 x i1> %0, <vscale x 8 x i8> %1, <vscale x 8 x i8> %2, <vscale x 8 x i8> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsltu_mask_vv_nxv8i8_nxv8i8:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e8, m1, ta, mu
+; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmsltu.vv v0, v8, v9
 ; CHECK-NEXT:    vmsltu.vv v11, v9, v10, v0.t
 ; CHECK-NEXT:    vmv.v.v v0, v11
@@ -238,8 +238,8 @@ declare <vscale x 16 x i1> @llvm.riscv.vmsltu.mask.nxv16i8(
 define <vscale x 16 x i1> @intrinsic_vmsltu_mask_vv_nxv16i8_nxv16i8(<vscale x 16 x i1> %0, <vscale x 16 x i8> %1, <vscale x 16 x i8> %2, <vscale x 16 x i8> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsltu_mask_vv_nxv16i8_nxv16i8:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v14, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e8, m2, ta, mu
+; CHECK-NEXT:    vmv1r.v v14, v0
 ; CHECK-NEXT:    vmsltu.vv v0, v8, v10
 ; CHECK-NEXT:    vmsltu.vv v14, v10, v12, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v14
@@ -289,8 +289,8 @@ declare <vscale x 32 x i1> @llvm.riscv.vmsltu.mask.nxv32i8(
 define <vscale x 32 x i1> @intrinsic_vmsltu_mask_vv_nxv32i8_nxv32i8(<vscale x 32 x i1> %0, <vscale x 32 x i8> %1, <vscale x 32 x i8> %2, <vscale x 32 x i8> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsltu_mask_vv_nxv32i8_nxv32i8:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v20, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e8, m4, ta, mu
+; CHECK-NEXT:    vmv1r.v v20, v0
 ; CHECK-NEXT:    vmsltu.vv v0, v8, v12
 ; CHECK-NEXT:    vmsltu.vv v20, v12, v16, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v20
@@ -340,8 +340,8 @@ declare <vscale x 1 x i1> @llvm.riscv.vmsltu.mask.nxv1i16(
 define <vscale x 1 x i1> @intrinsic_vmsltu_mask_vv_nxv1i16_nxv1i16(<vscale x 1 x i1> %0, <vscale x 1 x i16> %1, <vscale x 1 x i16> %2, <vscale x 1 x i16> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsltu_mask_vv_nxv1i16_nxv1i16:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e16, mf4, ta, mu
+; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmsltu.vv v0, v8, v9
 ; CHECK-NEXT:    vmsltu.vv v11, v9, v10, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v11
@@ -391,8 +391,8 @@ declare <vscale x 2 x i1> @llvm.riscv.vmsltu.mask.nxv2i16(
 define <vscale x 2 x i1> @intrinsic_vmsltu_mask_vv_nxv2i16_nxv2i16(<vscale x 2 x i1> %0, <vscale x 2 x i16> %1, <vscale x 2 x i16> %2, <vscale x 2 x i16> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsltu_mask_vv_nxv2i16_nxv2i16:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e16, mf2, ta, mu
+; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmsltu.vv v0, v8, v9
 ; CHECK-NEXT:    vmsltu.vv v11, v9, v10, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v11
@@ -442,8 +442,8 @@ declare <vscale x 4 x i1> @llvm.riscv.vmsltu.mask.nxv4i16(
 define <vscale x 4 x i1> @intrinsic_vmsltu_mask_vv_nxv4i16_nxv4i16(<vscale x 4 x i1> %0, <vscale x 4 x i16> %1, <vscale x 4 x i16> %2, <vscale x 4 x i16> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsltu_mask_vv_nxv4i16_nxv4i16:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e16, m1, ta, mu
+; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmsltu.vv v0, v8, v9
 ; CHECK-NEXT:    vmsltu.vv v11, v9, v10, v0.t
 ; CHECK-NEXT:    vmv.v.v v0, v11
@@ -493,8 +493,8 @@ declare <vscale x 8 x i1> @llvm.riscv.vmsltu.mask.nxv8i16(
 define <vscale x 8 x i1> @intrinsic_vmsltu_mask_vv_nxv8i16_nxv8i16(<vscale x 8 x i1> %0, <vscale x 8 x i16> %1, <vscale x 8 x i16> %2, <vscale x 8 x i16> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsltu_mask_vv_nxv8i16_nxv8i16:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v14, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e16, m2, ta, mu
+; CHECK-NEXT:    vmv1r.v v14, v0
 ; CHECK-NEXT:    vmsltu.vv v0, v8, v10
 ; CHECK-NEXT:    vmsltu.vv v14, v10, v12, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v14
@@ -544,8 +544,8 @@ declare <vscale x 16 x i1> @llvm.riscv.vmsltu.mask.nxv16i16(
 define <vscale x 16 x i1> @intrinsic_vmsltu_mask_vv_nxv16i16_nxv16i16(<vscale x 16 x i1> %0, <vscale x 16 x i16> %1, <vscale x 16 x i16> %2, <vscale x 16 x i16> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsltu_mask_vv_nxv16i16_nxv16i16:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v20, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e16, m4, ta, mu
+; CHECK-NEXT:    vmv1r.v v20, v0
 ; CHECK-NEXT:    vmsltu.vv v0, v8, v12
 ; CHECK-NEXT:    vmsltu.vv v20, v12, v16, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v20
@@ -595,8 +595,8 @@ declare <vscale x 1 x i1> @llvm.riscv.vmsltu.mask.nxv1i32(
 define <vscale x 1 x i1> @intrinsic_vmsltu_mask_vv_nxv1i32_nxv1i32(<vscale x 1 x i1> %0, <vscale x 1 x i32> %1, <vscale x 1 x i32> %2, <vscale x 1 x i32> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsltu_mask_vv_nxv1i32_nxv1i32:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e32, mf2, ta, mu
+; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmsltu.vv v0, v8, v9
 ; CHECK-NEXT:    vmsltu.vv v11, v9, v10, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v11
@@ -646,8 +646,8 @@ declare <vscale x 2 x i1> @llvm.riscv.vmsltu.mask.nxv2i32(
 define <vscale x 2 x i1> @intrinsic_vmsltu_mask_vv_nxv2i32_nxv2i32(<vscale x 2 x i1> %0, <vscale x 2 x i32> %1, <vscale x 2 x i32> %2, <vscale x 2 x i32> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsltu_mask_vv_nxv2i32_nxv2i32:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e32, m1, ta, mu
+; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmsltu.vv v0, v8, v9
 ; CHECK-NEXT:    vmsltu.vv v11, v9, v10, v0.t
 ; CHECK-NEXT:    vmv.v.v v0, v11
@@ -697,8 +697,8 @@ declare <vscale x 4 x i1> @llvm.riscv.vmsltu.mask.nxv4i32(
 define <vscale x 4 x i1> @intrinsic_vmsltu_mask_vv_nxv4i32_nxv4i32(<vscale x 4 x i1> %0, <vscale x 4 x i32> %1, <vscale x 4 x i32> %2, <vscale x 4 x i32> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsltu_mask_vv_nxv4i32_nxv4i32:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v14, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e32, m2, ta, mu
+; CHECK-NEXT:    vmv1r.v v14, v0
 ; CHECK-NEXT:    vmsltu.vv v0, v8, v10
 ; CHECK-NEXT:    vmsltu.vv v14, v10, v12, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v14
@@ -748,8 +748,8 @@ declare <vscale x 8 x i1> @llvm.riscv.vmsltu.mask.nxv8i32(
 define <vscale x 8 x i1> @intrinsic_vmsltu_mask_vv_nxv8i32_nxv8i32(<vscale x 8 x i1> %0, <vscale x 8 x i32> %1, <vscale x 8 x i32> %2, <vscale x 8 x i32> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsltu_mask_vv_nxv8i32_nxv8i32:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v20, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e32, m4, ta, mu
+; CHECK-NEXT:    vmv1r.v v20, v0
 ; CHECK-NEXT:    vmsltu.vv v0, v8, v12
 ; CHECK-NEXT:    vmsltu.vv v20, v12, v16, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v20
@@ -799,8 +799,8 @@ declare <vscale x 1 x i1> @llvm.riscv.vmsltu.mask.nxv1i64(
 define <vscale x 1 x i1> @intrinsic_vmsltu_mask_vv_nxv1i64_nxv1i64(<vscale x 1 x i1> %0, <vscale x 1 x i64> %1, <vscale x 1 x i64> %2, <vscale x 1 x i64> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsltu_mask_vv_nxv1i64_nxv1i64:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e64, m1, ta, mu
+; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmsltu.vv v0, v8, v9
 ; CHECK-NEXT:    vmsltu.vv v11, v9, v10, v0.t
 ; CHECK-NEXT:    vmv.v.v v0, v11
@@ -850,8 +850,8 @@ declare <vscale x 2 x i1> @llvm.riscv.vmsltu.mask.nxv2i64(
 define <vscale x 2 x i1> @intrinsic_vmsltu_mask_vv_nxv2i64_nxv2i64(<vscale x 2 x i1> %0, <vscale x 2 x i64> %1, <vscale x 2 x i64> %2, <vscale x 2 x i64> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsltu_mask_vv_nxv2i64_nxv2i64:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v14, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e64, m2, ta, mu
+; CHECK-NEXT:    vmv1r.v v14, v0
 ; CHECK-NEXT:    vmsltu.vv v0, v8, v10
 ; CHECK-NEXT:    vmsltu.vv v14, v10, v12, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v14
@@ -901,8 +901,8 @@ declare <vscale x 4 x i1> @llvm.riscv.vmsltu.mask.nxv4i64(
 define <vscale x 4 x i1> @intrinsic_vmsltu_mask_vv_nxv4i64_nxv4i64(<vscale x 4 x i1> %0, <vscale x 4 x i64> %1, <vscale x 4 x i64> %2, <vscale x 4 x i64> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsltu_mask_vv_nxv4i64_nxv4i64:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v20, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e64, m4, ta, mu
+; CHECK-NEXT:    vmv1r.v v20, v0
 ; CHECK-NEXT:    vmsltu.vv v0, v8, v12
 ; CHECK-NEXT:    vmsltu.vv v20, v12, v16, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v20
@@ -952,9 +952,9 @@ declare <vscale x 1 x i1> @llvm.riscv.vmsltu.mask.nxv1i8.i8(
 define <vscale x 1 x i1> @intrinsic_vmsltu_mask_vx_nxv1i8_i8(<vscale x 1 x i1> %0, <vscale x 1 x i8> %1, i8 %2, <vscale x 1 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsltu_mask_vx_nxv1i8_i8:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e8, mf8, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a1, e8, mf8, ta, mu
 ; CHECK-NEXT:    vmsltu.vx v10, v8, a0, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v10
 ; CHECK-NEXT:    ret
@@ -999,9 +999,9 @@ declare <vscale x 2 x i1> @llvm.riscv.vmsltu.mask.nxv2i8.i8(
 define <vscale x 2 x i1> @intrinsic_vmsltu_mask_vx_nxv2i8_i8(<vscale x 2 x i1> %0, <vscale x 2 x i8> %1, i8 %2, <vscale x 2 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsltu_mask_vx_nxv2i8_i8:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e8, mf4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a1, e8, mf4, ta, mu
 ; CHECK-NEXT:    vmsltu.vx v10, v8, a0, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v10
 ; CHECK-NEXT:    ret
@@ -1046,9 +1046,9 @@ declare <vscale x 4 x i1> @llvm.riscv.vmsltu.mask.nxv4i8.i8(
 define <vscale x 4 x i1> @intrinsic_vmsltu_mask_vx_nxv4i8_i8(<vscale x 4 x i1> %0, <vscale x 4 x i8> %1, i8 %2, <vscale x 4 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsltu_mask_vx_nxv4i8_i8:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e8, mf2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a1, e8, mf2, ta, mu
 ; CHECK-NEXT:    vmsltu.vx v10, v8, a0, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v10
 ; CHECK-NEXT:    ret
@@ -1093,9 +1093,9 @@ declare <vscale x 8 x i1> @llvm.riscv.vmsltu.mask.nxv8i8.i8(
 define <vscale x 8 x i1> @intrinsic_vmsltu_mask_vx_nxv8i8_i8(<vscale x 8 x i1> %0, <vscale x 8 x i8> %1, i8 %2, <vscale x 8 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsltu_mask_vx_nxv8i8_i8:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e8, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a1, e8, m1, ta, mu
 ; CHECK-NEXT:    vmsltu.vx v10, v8, a0, v0.t
 ; CHECK-NEXT:    vmv.v.v v0, v10
 ; CHECK-NEXT:    ret
@@ -1140,9 +1140,9 @@ declare <vscale x 16 x i1> @llvm.riscv.vmsltu.mask.nxv16i8.i8(
 define <vscale x 16 x i1> @intrinsic_vmsltu_mask_vx_nxv16i8_i8(<vscale x 16 x i1> %0, <vscale x 16 x i8> %1, i8 %2, <vscale x 16 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsltu_mask_vx_nxv16i8_i8:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e8, m2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmv1r.v v0, v10
-; CHECK-NEXT:    vsetvli zero, a1, e8, m2, ta, mu
 ; CHECK-NEXT:    vmsltu.vx v11, v8, a0, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v11
 ; CHECK-NEXT:    ret
@@ -1187,9 +1187,9 @@ declare <vscale x 32 x i1> @llvm.riscv.vmsltu.mask.nxv32i8.i8(
 define <vscale x 32 x i1> @intrinsic_vmsltu_mask_vx_nxv32i8_i8(<vscale x 32 x i1> %0, <vscale x 32 x i8> %1, i8 %2, <vscale x 32 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsltu_mask_vx_nxv32i8_i8:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e8, m4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v13, v0
 ; CHECK-NEXT:    vmv1r.v v0, v12
-; CHECK-NEXT:    vsetvli zero, a1, e8, m4, ta, mu
 ; CHECK-NEXT:    vmsltu.vx v13, v8, a0, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v13
 ; CHECK-NEXT:    ret
@@ -1234,9 +1234,9 @@ declare <vscale x 1 x i1> @llvm.riscv.vmsltu.mask.nxv1i16.i16(
 define <vscale x 1 x i1> @intrinsic_vmsltu_mask_vx_nxv1i16_i16(<vscale x 1 x i1> %0, <vscale x 1 x i16> %1, i16 %2, <vscale x 1 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsltu_mask_vx_nxv1i16_i16:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, mf4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a1, e16, mf4, ta, mu
 ; CHECK-NEXT:    vmsltu.vx v10, v8, a0, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v10
 ; CHECK-NEXT:    ret
@@ -1281,9 +1281,9 @@ declare <vscale x 2 x i1> @llvm.riscv.vmsltu.mask.nxv2i16.i16(
 define <vscale x 2 x i1> @intrinsic_vmsltu_mask_vx_nxv2i16_i16(<vscale x 2 x i1> %0, <vscale x 2 x i16> %1, i16 %2, <vscale x 2 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsltu_mask_vx_nxv2i16_i16:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, mf2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a1, e16, mf2, ta, mu
 ; CHECK-NEXT:    vmsltu.vx v10, v8, a0, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v10
 ; CHECK-NEXT:    ret
@@ -1328,9 +1328,9 @@ declare <vscale x 4 x i1> @llvm.riscv.vmsltu.mask.nxv4i16.i16(
 define <vscale x 4 x i1> @intrinsic_vmsltu_mask_vx_nxv4i16_i16(<vscale x 4 x i1> %0, <vscale x 4 x i16> %1, i16 %2, <vscale x 4 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsltu_mask_vx_nxv4i16_i16:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a1, e16, m1, ta, mu
 ; CHECK-NEXT:    vmsltu.vx v10, v8, a0, v0.t
 ; CHECK-NEXT:    vmv.v.v v0, v10
 ; CHECK-NEXT:    ret
@@ -1375,9 +1375,9 @@ declare <vscale x 8 x i1> @llvm.riscv.vmsltu.mask.nxv8i16.i16(
 define <vscale x 8 x i1> @intrinsic_vmsltu_mask_vx_nxv8i16_i16(<vscale x 8 x i1> %0, <vscale x 8 x i16> %1, i16 %2, <vscale x 8 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsltu_mask_vx_nxv8i16_i16:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, m2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmv1r.v v0, v10
-; CHECK-NEXT:    vsetvli zero, a1, e16, m2, ta, mu
 ; CHECK-NEXT:    vmsltu.vx v11, v8, a0, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v11
 ; CHECK-NEXT:    ret
@@ -1422,9 +1422,9 @@ declare <vscale x 16 x i1> @llvm.riscv.vmsltu.mask.nxv16i16.i16(
 define <vscale x 16 x i1> @intrinsic_vmsltu_mask_vx_nxv16i16_i16(<vscale x 16 x i1> %0, <vscale x 16 x i16> %1, i16 %2, <vscale x 16 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsltu_mask_vx_nxv16i16_i16:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, m4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v13, v0
 ; CHECK-NEXT:    vmv1r.v v0, v12
-; CHECK-NEXT:    vsetvli zero, a1, e16, m4, ta, mu
 ; CHECK-NEXT:    vmsltu.vx v13, v8, a0, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v13
 ; CHECK-NEXT:    ret
@@ -1469,9 +1469,9 @@ declare <vscale x 1 x i1> @llvm.riscv.vmsltu.mask.nxv1i32.i32(
 define <vscale x 1 x i1> @intrinsic_vmsltu_mask_vx_nxv1i32_i32(<vscale x 1 x i1> %0, <vscale x 1 x i32> %1, i32 %2, <vscale x 1 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsltu_mask_vx_nxv1i32_i32:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e32, mf2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a1, e32, mf2, ta, mu
 ; CHECK-NEXT:    vmsltu.vx v10, v8, a0, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v10
 ; CHECK-NEXT:    ret
@@ -1516,9 +1516,9 @@ declare <vscale x 2 x i1> @llvm.riscv.vmsltu.mask.nxv2i32.i32(
 define <vscale x 2 x i1> @intrinsic_vmsltu_mask_vx_nxv2i32_i32(<vscale x 2 x i1> %0, <vscale x 2 x i32> %1, i32 %2, <vscale x 2 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsltu_mask_vx_nxv2i32_i32:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e32, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a1, e32, m1, ta, mu
 ; CHECK-NEXT:    vmsltu.vx v10, v8, a0, v0.t
 ; CHECK-NEXT:    vmv.v.v v0, v10
 ; CHECK-NEXT:    ret
@@ -1563,9 +1563,9 @@ declare <vscale x 4 x i1> @llvm.riscv.vmsltu.mask.nxv4i32.i32(
 define <vscale x 4 x i1> @intrinsic_vmsltu_mask_vx_nxv4i32_i32(<vscale x 4 x i1> %0, <vscale x 4 x i32> %1, i32 %2, <vscale x 4 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsltu_mask_vx_nxv4i32_i32:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e32, m2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmv1r.v v0, v10
-; CHECK-NEXT:    vsetvli zero, a1, e32, m2, ta, mu
 ; CHECK-NEXT:    vmsltu.vx v11, v8, a0, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v11
 ; CHECK-NEXT:    ret
@@ -1610,9 +1610,9 @@ declare <vscale x 8 x i1> @llvm.riscv.vmsltu.mask.nxv8i32.i32(
 define <vscale x 8 x i1> @intrinsic_vmsltu_mask_vx_nxv8i32_i32(<vscale x 8 x i1> %0, <vscale x 8 x i32> %1, i32 %2, <vscale x 8 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsltu_mask_vx_nxv8i32_i32:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e32, m4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v13, v0
 ; CHECK-NEXT:    vmv1r.v v0, v12
-; CHECK-NEXT:    vsetvli zero, a1, e32, m4, ta, mu
 ; CHECK-NEXT:    vmsltu.vx v13, v8, a0, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v13
 ; CHECK-NEXT:    ret
@@ -1684,9 +1684,9 @@ define <vscale x 1 x i1> @intrinsic_vmsltu_mask_vx_nxv1i64_i64(<vscale x 1 x i1>
 ;
 ; RV64-LABEL: intrinsic_vmsltu_mask_vx_nxv1i64_i64:
 ; RV64:       # %bb.0: # %entry
+; RV64-NEXT:    vsetvli zero, a1, e64, m1, ta, mu
 ; RV64-NEXT:    vmv1r.v v10, v0
 ; RV64-NEXT:    vmv1r.v v0, v9
-; RV64-NEXT:    vsetvli zero, a1, e64, m1, ta, mu
 ; RV64-NEXT:    vmsltu.vx v10, v8, a0, v0.t
 ; RV64-NEXT:    vmv.v.v v0, v10
 ; RV64-NEXT:    ret
@@ -1758,9 +1758,9 @@ define <vscale x 2 x i1> @intrinsic_vmsltu_mask_vx_nxv2i64_i64(<vscale x 2 x i1>
 ;
 ; RV64-LABEL: intrinsic_vmsltu_mask_vx_nxv2i64_i64:
 ; RV64:       # %bb.0: # %entry
+; RV64-NEXT:    vsetvli zero, a1, e64, m2, ta, mu
 ; RV64-NEXT:    vmv1r.v v11, v0
 ; RV64-NEXT:    vmv1r.v v0, v10
-; RV64-NEXT:    vsetvli zero, a1, e64, m2, ta, mu
 ; RV64-NEXT:    vmsltu.vx v11, v8, a0, v0.t
 ; RV64-NEXT:    vmv1r.v v0, v11
 ; RV64-NEXT:    ret
@@ -1832,9 +1832,9 @@ define <vscale x 4 x i1> @intrinsic_vmsltu_mask_vx_nxv4i64_i64(<vscale x 4 x i1>
 ;
 ; RV64-LABEL: intrinsic_vmsltu_mask_vx_nxv4i64_i64:
 ; RV64:       # %bb.0: # %entry
+; RV64-NEXT:    vsetvli zero, a1, e64, m4, ta, mu
 ; RV64-NEXT:    vmv1r.v v13, v0
 ; RV64-NEXT:    vmv1r.v v0, v12
-; RV64-NEXT:    vsetvli zero, a1, e64, m4, ta, mu
 ; RV64-NEXT:    vmsltu.vx v13, v8, a0, v0.t
 ; RV64-NEXT:    vmv1r.v v0, v13
 ; RV64-NEXT:    ret
@@ -1867,9 +1867,9 @@ entry:
 define <vscale x 1 x i1> @intrinsic_vmsltu_mask_vi_nxv1i8_i8(<vscale x 1 x i1> %0, <vscale x 1 x i8> %1, <vscale x 1 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsltu_mask_vi_nxv1i8_i8:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e8, mf8, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a0, e8, mf8, ta, mu
 ; CHECK-NEXT:    vmsleu.vi v10, v8, -15, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v10
 ; CHECK-NEXT:    ret
@@ -1902,9 +1902,9 @@ entry:
 define <vscale x 2 x i1> @intrinsic_vmsltu_mask_vi_nxv2i8_i8(<vscale x 2 x i1> %0, <vscale x 2 x i8> %1, <vscale x 2 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsltu_mask_vi_nxv2i8_i8:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e8, mf4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a0, e8, mf4, ta, mu
 ; CHECK-NEXT:    vmsleu.vi v10, v8, -13, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v10
 ; CHECK-NEXT:    ret
@@ -1937,9 +1937,9 @@ entry:
 define <vscale x 4 x i1> @intrinsic_vmsltu_mask_vi_nxv4i8_i8(<vscale x 4 x i1> %0, <vscale x 4 x i8> %1, <vscale x 4 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsltu_mask_vi_nxv4i8_i8:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e8, mf2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a0, e8, mf2, ta, mu
 ; CHECK-NEXT:    vmsleu.vi v10, v8, -11, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v10
 ; CHECK-NEXT:    ret
@@ -1972,9 +1972,9 @@ entry:
 define <vscale x 8 x i1> @intrinsic_vmsltu_mask_vi_nxv8i8_i8(<vscale x 8 x i1> %0, <vscale x 8 x i8> %1, <vscale x 8 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsltu_mask_vi_nxv8i8_i8:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e8, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a0, e8, m1, ta, mu
 ; CHECK-NEXT:    vmsleu.vi v10, v8, -9, v0.t
 ; CHECK-NEXT:    vmv.v.v v0, v10
 ; CHECK-NEXT:    ret
@@ -2007,9 +2007,9 @@ entry:
 define <vscale x 16 x i1> @intrinsic_vmsltu_mask_vi_nxv16i8_i8(<vscale x 16 x i1> %0, <vscale x 16 x i8> %1, <vscale x 16 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsltu_mask_vi_nxv16i8_i8:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e8, m2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmv1r.v v0, v10
-; CHECK-NEXT:    vsetvli zero, a0, e8, m2, ta, mu
 ; CHECK-NEXT:    vmsleu.vi v11, v8, -7, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v11
 ; CHECK-NEXT:    ret
@@ -2042,9 +2042,9 @@ entry:
 define <vscale x 32 x i1> @intrinsic_vmsltu_mask_vi_nxv32i8_i8(<vscale x 32 x i1> %0, <vscale x 32 x i8> %1, <vscale x 32 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsltu_mask_vi_nxv32i8_i8:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e8, m4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v13, v0
 ; CHECK-NEXT:    vmv1r.v v0, v12
-; CHECK-NEXT:    vsetvli zero, a0, e8, m4, ta, mu
 ; CHECK-NEXT:    vmsleu.vi v13, v8, -5, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v13
 ; CHECK-NEXT:    ret
@@ -2077,9 +2077,9 @@ entry:
 define <vscale x 1 x i1> @intrinsic_vmsltu_mask_vi_nxv1i16_i16(<vscale x 1 x i1> %0, <vscale x 1 x i16> %1, <vscale x 1 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsltu_mask_vi_nxv1i16_i16:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e16, mf4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a0, e16, mf4, ta, mu
 ; CHECK-NEXT:    vmsleu.vi v10, v8, -3, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v10
 ; CHECK-NEXT:    ret
@@ -2112,9 +2112,9 @@ entry:
 define <vscale x 2 x i1> @intrinsic_vmsltu_mask_vi_nxv2i16_i16(<vscale x 2 x i1> %0, <vscale x 2 x i16> %1, <vscale x 2 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsltu_mask_vi_nxv2i16_i16:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e16, mf2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a0, e16, mf2, ta, mu
 ; CHECK-NEXT:    vmsltu.vx v10, v8, zero, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v10
 ; CHECK-NEXT:    ret
@@ -2147,9 +2147,9 @@ entry:
 define <vscale x 4 x i1> @intrinsic_vmsltu_mask_vi_nxv4i16_i16(<vscale x 4 x i1> %0, <vscale x 4 x i16> %1, <vscale x 4 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsltu_mask_vi_nxv4i16_i16:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e16, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a0, e16, m1, ta, mu
 ; CHECK-NEXT:    vmsleu.vi v10, v8, 0, v0.t
 ; CHECK-NEXT:    vmv.v.v v0, v10
 ; CHECK-NEXT:    ret
@@ -2182,9 +2182,9 @@ entry:
 define <vscale x 8 x i1> @intrinsic_vmsltu_mask_vi_nxv8i16_i16(<vscale x 8 x i1> %0, <vscale x 8 x i16> %1, <vscale x 8 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsltu_mask_vi_nxv8i16_i16:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e16, m2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmv1r.v v0, v10
-; CHECK-NEXT:    vsetvli zero, a0, e16, m2, ta, mu
 ; CHECK-NEXT:    vmsleu.vi v11, v8, 2, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v11
 ; CHECK-NEXT:    ret
@@ -2217,9 +2217,9 @@ entry:
 define <vscale x 16 x i1> @intrinsic_vmsltu_mask_vi_nxv16i16_i16(<vscale x 16 x i1> %0, <vscale x 16 x i16> %1, <vscale x 16 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsltu_mask_vi_nxv16i16_i16:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e16, m4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v13, v0
 ; CHECK-NEXT:    vmv1r.v v0, v12
-; CHECK-NEXT:    vsetvli zero, a0, e16, m4, ta, mu
 ; CHECK-NEXT:    vmsleu.vi v13, v8, 4, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v13
 ; CHECK-NEXT:    ret
@@ -2252,9 +2252,9 @@ entry:
 define <vscale x 1 x i1> @intrinsic_vmsltu_mask_vi_nxv1i32_i32(<vscale x 1 x i1> %0, <vscale x 1 x i32> %1, <vscale x 1 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsltu_mask_vi_nxv1i32_i32:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e32, mf2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a0, e32, mf2, ta, mu
 ; CHECK-NEXT:    vmsleu.vi v10, v8, 6, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v10
 ; CHECK-NEXT:    ret
@@ -2287,9 +2287,9 @@ entry:
 define <vscale x 2 x i1> @intrinsic_vmsltu_mask_vi_nxv2i32_i32(<vscale x 2 x i1> %0, <vscale x 2 x i32> %1, <vscale x 2 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsltu_mask_vi_nxv2i32_i32:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e32, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a0, e32, m1, ta, mu
 ; CHECK-NEXT:    vmsleu.vi v10, v8, 8, v0.t
 ; CHECK-NEXT:    vmv.v.v v0, v10
 ; CHECK-NEXT:    ret
@@ -2322,9 +2322,9 @@ entry:
 define <vscale x 4 x i1> @intrinsic_vmsltu_mask_vi_nxv4i32_i32(<vscale x 4 x i1> %0, <vscale x 4 x i32> %1, <vscale x 4 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsltu_mask_vi_nxv4i32_i32:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e32, m2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmv1r.v v0, v10
-; CHECK-NEXT:    vsetvli zero, a0, e32, m2, ta, mu
 ; CHECK-NEXT:    vmsleu.vi v11, v8, 10, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v11
 ; CHECK-NEXT:    ret
@@ -2357,9 +2357,9 @@ entry:
 define <vscale x 8 x i1> @intrinsic_vmsltu_mask_vi_nxv8i32_i32(<vscale x 8 x i1> %0, <vscale x 8 x i32> %1, <vscale x 8 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsltu_mask_vi_nxv8i32_i32:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e32, m4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v13, v0
 ; CHECK-NEXT:    vmv1r.v v0, v12
-; CHECK-NEXT:    vsetvli zero, a0, e32, m4, ta, mu
 ; CHECK-NEXT:    vmsleu.vi v13, v8, 12, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v13
 ; CHECK-NEXT:    ret
@@ -2392,9 +2392,9 @@ entry:
 define <vscale x 1 x i1> @intrinsic_vmsltu_mask_vi_nxv1i64_i64(<vscale x 1 x i1> %0, <vscale x 1 x i64> %1, <vscale x 1 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsltu_mask_vi_nxv1i64_i64:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e64, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a0, e64, m1, ta, mu
 ; CHECK-NEXT:    vmsleu.vi v10, v8, 14, v0.t
 ; CHECK-NEXT:    vmv.v.v v0, v10
 ; CHECK-NEXT:    ret
@@ -2427,9 +2427,9 @@ entry:
 define <vscale x 2 x i1> @intrinsic_vmsltu_mask_vi_nxv2i64_i64(<vscale x 2 x i1> %0, <vscale x 2 x i64> %1, <vscale x 2 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsltu_mask_vi_nxv2i64_i64:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e64, m2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmv1r.v v0, v10
-; CHECK-NEXT:    vsetvli zero, a0, e64, m2, ta, mu
 ; CHECK-NEXT:    vmsleu.vi v11, v8, -16, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v11
 ; CHECK-NEXT:    ret
@@ -2462,9 +2462,9 @@ entry:
 define <vscale x 4 x i1> @intrinsic_vmsltu_mask_vi_nxv4i64_i64(<vscale x 4 x i1> %0, <vscale x 4 x i64> %1, <vscale x 4 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsltu_mask_vi_nxv4i64_i64:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e64, m4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v13, v0
 ; CHECK-NEXT:    vmv1r.v v0, v12
-; CHECK-NEXT:    vsetvli zero, a0, e64, m4, ta, mu
 ; CHECK-NEXT:    vmsleu.vi v13, v8, -14, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v13
 ; CHECK-NEXT:    ret
diff --git a/llvm/test/CodeGen/RISCV/rvv/vmsne.ll b/llvm/test/CodeGen/RISCV/rvv/vmsne.ll
index 1e21b847ed20d..ed41a18dcc8d3 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vmsne.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vmsne.ll
@@ -34,8 +34,8 @@ declare <vscale x 1 x i1> @llvm.riscv.vmsne.mask.nxv1i8(
 define <vscale x 1 x i1> @intrinsic_vmsne_mask_vv_nxv1i8_nxv1i8(<vscale x 1 x i1> %0, <vscale x 1 x i8> %1, <vscale x 1 x i8> %2, <vscale x 1 x i8> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsne_mask_vv_nxv1i8_nxv1i8:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e8, mf8, ta, mu
+; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmsne.vv v0, v8, v9
 ; CHECK-NEXT:    vmsne.vv v11, v9, v10, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v11
@@ -85,8 +85,8 @@ declare <vscale x 2 x i1> @llvm.riscv.vmsne.mask.nxv2i8(
 define <vscale x 2 x i1> @intrinsic_vmsne_mask_vv_nxv2i8_nxv2i8(<vscale x 2 x i1> %0, <vscale x 2 x i8> %1, <vscale x 2 x i8> %2, <vscale x 2 x i8> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsne_mask_vv_nxv2i8_nxv2i8:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e8, mf4, ta, mu
+; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmsne.vv v0, v8, v9
 ; CHECK-NEXT:    vmsne.vv v11, v9, v10, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v11
@@ -136,8 +136,8 @@ declare <vscale x 4 x i1> @llvm.riscv.vmsne.mask.nxv4i8(
 define <vscale x 4 x i1> @intrinsic_vmsne_mask_vv_nxv4i8_nxv4i8(<vscale x 4 x i1> %0, <vscale x 4 x i8> %1, <vscale x 4 x i8> %2, <vscale x 4 x i8> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsne_mask_vv_nxv4i8_nxv4i8:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e8, mf2, ta, mu
+; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmsne.vv v0, v8, v9
 ; CHECK-NEXT:    vmsne.vv v11, v9, v10, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v11
@@ -187,8 +187,8 @@ declare <vscale x 8 x i1> @llvm.riscv.vmsne.mask.nxv8i8(
 define <vscale x 8 x i1> @intrinsic_vmsne_mask_vv_nxv8i8_nxv8i8(<vscale x 8 x i1> %0, <vscale x 8 x i8> %1, <vscale x 8 x i8> %2, <vscale x 8 x i8> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsne_mask_vv_nxv8i8_nxv8i8:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e8, m1, ta, mu
+; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmsne.vv v0, v8, v9
 ; CHECK-NEXT:    vmsne.vv v11, v9, v10, v0.t
 ; CHECK-NEXT:    vmv.v.v v0, v11
@@ -238,8 +238,8 @@ declare <vscale x 16 x i1> @llvm.riscv.vmsne.mask.nxv16i8(
 define <vscale x 16 x i1> @intrinsic_vmsne_mask_vv_nxv16i8_nxv16i8(<vscale x 16 x i1> %0, <vscale x 16 x i8> %1, <vscale x 16 x i8> %2, <vscale x 16 x i8> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsne_mask_vv_nxv16i8_nxv16i8:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v14, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e8, m2, ta, mu
+; CHECK-NEXT:    vmv1r.v v14, v0
 ; CHECK-NEXT:    vmsne.vv v0, v8, v10
 ; CHECK-NEXT:    vmsne.vv v14, v10, v12, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v14
@@ -289,8 +289,8 @@ declare <vscale x 32 x i1> @llvm.riscv.vmsne.mask.nxv32i8(
 define <vscale x 32 x i1> @intrinsic_vmsne_mask_vv_nxv32i8_nxv32i8(<vscale x 32 x i1> %0, <vscale x 32 x i8> %1, <vscale x 32 x i8> %2, <vscale x 32 x i8> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsne_mask_vv_nxv32i8_nxv32i8:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v20, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e8, m4, ta, mu
+; CHECK-NEXT:    vmv1r.v v20, v0
 ; CHECK-NEXT:    vmsne.vv v0, v8, v12
 ; CHECK-NEXT:    vmsne.vv v20, v12, v16, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v20
@@ -340,8 +340,8 @@ declare <vscale x 1 x i1> @llvm.riscv.vmsne.mask.nxv1i16(
 define <vscale x 1 x i1> @intrinsic_vmsne_mask_vv_nxv1i16_nxv1i16(<vscale x 1 x i1> %0, <vscale x 1 x i16> %1, <vscale x 1 x i16> %2, <vscale x 1 x i16> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsne_mask_vv_nxv1i16_nxv1i16:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e16, mf4, ta, mu
+; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmsne.vv v0, v8, v9
 ; CHECK-NEXT:    vmsne.vv v11, v9, v10, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v11
@@ -391,8 +391,8 @@ declare <vscale x 2 x i1> @llvm.riscv.vmsne.mask.nxv2i16(
 define <vscale x 2 x i1> @intrinsic_vmsne_mask_vv_nxv2i16_nxv2i16(<vscale x 2 x i1> %0, <vscale x 2 x i16> %1, <vscale x 2 x i16> %2, <vscale x 2 x i16> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsne_mask_vv_nxv2i16_nxv2i16:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e16, mf2, ta, mu
+; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmsne.vv v0, v8, v9
 ; CHECK-NEXT:    vmsne.vv v11, v9, v10, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v11
@@ -442,8 +442,8 @@ declare <vscale x 4 x i1> @llvm.riscv.vmsne.mask.nxv4i16(
 define <vscale x 4 x i1> @intrinsic_vmsne_mask_vv_nxv4i16_nxv4i16(<vscale x 4 x i1> %0, <vscale x 4 x i16> %1, <vscale x 4 x i16> %2, <vscale x 4 x i16> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsne_mask_vv_nxv4i16_nxv4i16:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e16, m1, ta, mu
+; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmsne.vv v0, v8, v9
 ; CHECK-NEXT:    vmsne.vv v11, v9, v10, v0.t
 ; CHECK-NEXT:    vmv.v.v v0, v11
@@ -493,8 +493,8 @@ declare <vscale x 8 x i1> @llvm.riscv.vmsne.mask.nxv8i16(
 define <vscale x 8 x i1> @intrinsic_vmsne_mask_vv_nxv8i16_nxv8i16(<vscale x 8 x i1> %0, <vscale x 8 x i16> %1, <vscale x 8 x i16> %2, <vscale x 8 x i16> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsne_mask_vv_nxv8i16_nxv8i16:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v14, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e16, m2, ta, mu
+; CHECK-NEXT:    vmv1r.v v14, v0
 ; CHECK-NEXT:    vmsne.vv v0, v8, v10
 ; CHECK-NEXT:    vmsne.vv v14, v10, v12, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v14
@@ -544,8 +544,8 @@ declare <vscale x 16 x i1> @llvm.riscv.vmsne.mask.nxv16i16(
 define <vscale x 16 x i1> @intrinsic_vmsne_mask_vv_nxv16i16_nxv16i16(<vscale x 16 x i1> %0, <vscale x 16 x i16> %1, <vscale x 16 x i16> %2, <vscale x 16 x i16> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsne_mask_vv_nxv16i16_nxv16i16:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v20, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e16, m4, ta, mu
+; CHECK-NEXT:    vmv1r.v v20, v0
 ; CHECK-NEXT:    vmsne.vv v0, v8, v12
 ; CHECK-NEXT:    vmsne.vv v20, v12, v16, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v20
@@ -595,8 +595,8 @@ declare <vscale x 1 x i1> @llvm.riscv.vmsne.mask.nxv1i32(
 define <vscale x 1 x i1> @intrinsic_vmsne_mask_vv_nxv1i32_nxv1i32(<vscale x 1 x i1> %0, <vscale x 1 x i32> %1, <vscale x 1 x i32> %2, <vscale x 1 x i32> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsne_mask_vv_nxv1i32_nxv1i32:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e32, mf2, ta, mu
+; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmsne.vv v0, v8, v9
 ; CHECK-NEXT:    vmsne.vv v11, v9, v10, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v11
@@ -646,8 +646,8 @@ declare <vscale x 2 x i1> @llvm.riscv.vmsne.mask.nxv2i32(
 define <vscale x 2 x i1> @intrinsic_vmsne_mask_vv_nxv2i32_nxv2i32(<vscale x 2 x i1> %0, <vscale x 2 x i32> %1, <vscale x 2 x i32> %2, <vscale x 2 x i32> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsne_mask_vv_nxv2i32_nxv2i32:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e32, m1, ta, mu
+; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmsne.vv v0, v8, v9
 ; CHECK-NEXT:    vmsne.vv v11, v9, v10, v0.t
 ; CHECK-NEXT:    vmv.v.v v0, v11
@@ -697,8 +697,8 @@ declare <vscale x 4 x i1> @llvm.riscv.vmsne.mask.nxv4i32(
 define <vscale x 4 x i1> @intrinsic_vmsne_mask_vv_nxv4i32_nxv4i32(<vscale x 4 x i1> %0, <vscale x 4 x i32> %1, <vscale x 4 x i32> %2, <vscale x 4 x i32> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsne_mask_vv_nxv4i32_nxv4i32:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v14, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e32, m2, ta, mu
+; CHECK-NEXT:    vmv1r.v v14, v0
 ; CHECK-NEXT:    vmsne.vv v0, v8, v10
 ; CHECK-NEXT:    vmsne.vv v14, v10, v12, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v14
@@ -748,8 +748,8 @@ declare <vscale x 8 x i1> @llvm.riscv.vmsne.mask.nxv8i32(
 define <vscale x 8 x i1> @intrinsic_vmsne_mask_vv_nxv8i32_nxv8i32(<vscale x 8 x i1> %0, <vscale x 8 x i32> %1, <vscale x 8 x i32> %2, <vscale x 8 x i32> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsne_mask_vv_nxv8i32_nxv8i32:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v20, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e32, m4, ta, mu
+; CHECK-NEXT:    vmv1r.v v20, v0
 ; CHECK-NEXT:    vmsne.vv v0, v8, v12
 ; CHECK-NEXT:    vmsne.vv v20, v12, v16, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v20
@@ -799,8 +799,8 @@ declare <vscale x 1 x i1> @llvm.riscv.vmsne.mask.nxv1i64(
 define <vscale x 1 x i1> @intrinsic_vmsne_mask_vv_nxv1i64_nxv1i64(<vscale x 1 x i1> %0, <vscale x 1 x i64> %1, <vscale x 1 x i64> %2, <vscale x 1 x i64> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsne_mask_vv_nxv1i64_nxv1i64:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e64, m1, ta, mu
+; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmsne.vv v0, v8, v9
 ; CHECK-NEXT:    vmsne.vv v11, v9, v10, v0.t
 ; CHECK-NEXT:    vmv.v.v v0, v11
@@ -850,8 +850,8 @@ declare <vscale x 2 x i1> @llvm.riscv.vmsne.mask.nxv2i64(
 define <vscale x 2 x i1> @intrinsic_vmsne_mask_vv_nxv2i64_nxv2i64(<vscale x 2 x i1> %0, <vscale x 2 x i64> %1, <vscale x 2 x i64> %2, <vscale x 2 x i64> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsne_mask_vv_nxv2i64_nxv2i64:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v14, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e64, m2, ta, mu
+; CHECK-NEXT:    vmv1r.v v14, v0
 ; CHECK-NEXT:    vmsne.vv v0, v8, v10
 ; CHECK-NEXT:    vmsne.vv v14, v10, v12, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v14
@@ -901,8 +901,8 @@ declare <vscale x 4 x i1> @llvm.riscv.vmsne.mask.nxv4i64(
 define <vscale x 4 x i1> @intrinsic_vmsne_mask_vv_nxv4i64_nxv4i64(<vscale x 4 x i1> %0, <vscale x 4 x i64> %1, <vscale x 4 x i64> %2, <vscale x 4 x i64> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsne_mask_vv_nxv4i64_nxv4i64:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v20, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e64, m4, ta, mu
+; CHECK-NEXT:    vmv1r.v v20, v0
 ; CHECK-NEXT:    vmsne.vv v0, v8, v12
 ; CHECK-NEXT:    vmsne.vv v20, v12, v16, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v20
@@ -952,9 +952,9 @@ declare <vscale x 1 x i1> @llvm.riscv.vmsne.mask.nxv1i8.i8(
 define <vscale x 1 x i1> @intrinsic_vmsne_mask_vx_nxv1i8_i8(<vscale x 1 x i1> %0, <vscale x 1 x i8> %1, i8 %2, <vscale x 1 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsne_mask_vx_nxv1i8_i8:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e8, mf8, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a1, e8, mf8, ta, mu
 ; CHECK-NEXT:    vmsne.vx v10, v8, a0, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v10
 ; CHECK-NEXT:    ret
@@ -999,9 +999,9 @@ declare <vscale x 2 x i1> @llvm.riscv.vmsne.mask.nxv2i8.i8(
 define <vscale x 2 x i1> @intrinsic_vmsne_mask_vx_nxv2i8_i8(<vscale x 2 x i1> %0, <vscale x 2 x i8> %1, i8 %2, <vscale x 2 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsne_mask_vx_nxv2i8_i8:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e8, mf4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a1, e8, mf4, ta, mu
 ; CHECK-NEXT:    vmsne.vx v10, v8, a0, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v10
 ; CHECK-NEXT:    ret
@@ -1046,9 +1046,9 @@ declare <vscale x 4 x i1> @llvm.riscv.vmsne.mask.nxv4i8.i8(
 define <vscale x 4 x i1> @intrinsic_vmsne_mask_vx_nxv4i8_i8(<vscale x 4 x i1> %0, <vscale x 4 x i8> %1, i8 %2, <vscale x 4 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsne_mask_vx_nxv4i8_i8:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e8, mf2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a1, e8, mf2, ta, mu
 ; CHECK-NEXT:    vmsne.vx v10, v8, a0, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v10
 ; CHECK-NEXT:    ret
@@ -1093,9 +1093,9 @@ declare <vscale x 8 x i1> @llvm.riscv.vmsne.mask.nxv8i8.i8(
 define <vscale x 8 x i1> @intrinsic_vmsne_mask_vx_nxv8i8_i8(<vscale x 8 x i1> %0, <vscale x 8 x i8> %1, i8 %2, <vscale x 8 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsne_mask_vx_nxv8i8_i8:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e8, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a1, e8, m1, ta, mu
 ; CHECK-NEXT:    vmsne.vx v10, v8, a0, v0.t
 ; CHECK-NEXT:    vmv.v.v v0, v10
 ; CHECK-NEXT:    ret
@@ -1140,9 +1140,9 @@ declare <vscale x 16 x i1> @llvm.riscv.vmsne.mask.nxv16i8.i8(
 define <vscale x 16 x i1> @intrinsic_vmsne_mask_vx_nxv16i8_i8(<vscale x 16 x i1> %0, <vscale x 16 x i8> %1, i8 %2, <vscale x 16 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsne_mask_vx_nxv16i8_i8:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e8, m2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmv1r.v v0, v10
-; CHECK-NEXT:    vsetvli zero, a1, e8, m2, ta, mu
 ; CHECK-NEXT:    vmsne.vx v11, v8, a0, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v11
 ; CHECK-NEXT:    ret
@@ -1187,9 +1187,9 @@ declare <vscale x 32 x i1> @llvm.riscv.vmsne.mask.nxv32i8.i8(
 define <vscale x 32 x i1> @intrinsic_vmsne_mask_vx_nxv32i8_i8(<vscale x 32 x i1> %0, <vscale x 32 x i8> %1, i8 %2, <vscale x 32 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsne_mask_vx_nxv32i8_i8:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e8, m4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v13, v0
 ; CHECK-NEXT:    vmv1r.v v0, v12
-; CHECK-NEXT:    vsetvli zero, a1, e8, m4, ta, mu
 ; CHECK-NEXT:    vmsne.vx v13, v8, a0, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v13
 ; CHECK-NEXT:    ret
@@ -1234,9 +1234,9 @@ declare <vscale x 1 x i1> @llvm.riscv.vmsne.mask.nxv1i16.i16(
 define <vscale x 1 x i1> @intrinsic_vmsne_mask_vx_nxv1i16_i16(<vscale x 1 x i1> %0, <vscale x 1 x i16> %1, i16 %2, <vscale x 1 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsne_mask_vx_nxv1i16_i16:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, mf4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a1, e16, mf4, ta, mu
 ; CHECK-NEXT:    vmsne.vx v10, v8, a0, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v10
 ; CHECK-NEXT:    ret
@@ -1281,9 +1281,9 @@ declare <vscale x 2 x i1> @llvm.riscv.vmsne.mask.nxv2i16.i16(
 define <vscale x 2 x i1> @intrinsic_vmsne_mask_vx_nxv2i16_i16(<vscale x 2 x i1> %0, <vscale x 2 x i16> %1, i16 %2, <vscale x 2 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsne_mask_vx_nxv2i16_i16:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, mf2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a1, e16, mf2, ta, mu
 ; CHECK-NEXT:    vmsne.vx v10, v8, a0, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v10
 ; CHECK-NEXT:    ret
@@ -1328,9 +1328,9 @@ declare <vscale x 4 x i1> @llvm.riscv.vmsne.mask.nxv4i16.i16(
 define <vscale x 4 x i1> @intrinsic_vmsne_mask_vx_nxv4i16_i16(<vscale x 4 x i1> %0, <vscale x 4 x i16> %1, i16 %2, <vscale x 4 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsne_mask_vx_nxv4i16_i16:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a1, e16, m1, ta, mu
 ; CHECK-NEXT:    vmsne.vx v10, v8, a0, v0.t
 ; CHECK-NEXT:    vmv.v.v v0, v10
 ; CHECK-NEXT:    ret
@@ -1375,9 +1375,9 @@ declare <vscale x 8 x i1> @llvm.riscv.vmsne.mask.nxv8i16.i16(
 define <vscale x 8 x i1> @intrinsic_vmsne_mask_vx_nxv8i16_i16(<vscale x 8 x i1> %0, <vscale x 8 x i16> %1, i16 %2, <vscale x 8 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsne_mask_vx_nxv8i16_i16:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, m2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmv1r.v v0, v10
-; CHECK-NEXT:    vsetvli zero, a1, e16, m2, ta, mu
 ; CHECK-NEXT:    vmsne.vx v11, v8, a0, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v11
 ; CHECK-NEXT:    ret
@@ -1422,9 +1422,9 @@ declare <vscale x 16 x i1> @llvm.riscv.vmsne.mask.nxv16i16.i16(
 define <vscale x 16 x i1> @intrinsic_vmsne_mask_vx_nxv16i16_i16(<vscale x 16 x i1> %0, <vscale x 16 x i16> %1, i16 %2, <vscale x 16 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsne_mask_vx_nxv16i16_i16:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e16, m4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v13, v0
 ; CHECK-NEXT:    vmv1r.v v0, v12
-; CHECK-NEXT:    vsetvli zero, a1, e16, m4, ta, mu
 ; CHECK-NEXT:    vmsne.vx v13, v8, a0, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v13
 ; CHECK-NEXT:    ret
@@ -1469,9 +1469,9 @@ declare <vscale x 1 x i1> @llvm.riscv.vmsne.mask.nxv1i32.i32(
 define <vscale x 1 x i1> @intrinsic_vmsne_mask_vx_nxv1i32_i32(<vscale x 1 x i1> %0, <vscale x 1 x i32> %1, i32 %2, <vscale x 1 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsne_mask_vx_nxv1i32_i32:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e32, mf2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a1, e32, mf2, ta, mu
 ; CHECK-NEXT:    vmsne.vx v10, v8, a0, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v10
 ; CHECK-NEXT:    ret
@@ -1516,9 +1516,9 @@ declare <vscale x 2 x i1> @llvm.riscv.vmsne.mask.nxv2i32.i32(
 define <vscale x 2 x i1> @intrinsic_vmsne_mask_vx_nxv2i32_i32(<vscale x 2 x i1> %0, <vscale x 2 x i32> %1, i32 %2, <vscale x 2 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsne_mask_vx_nxv2i32_i32:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e32, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a1, e32, m1, ta, mu
 ; CHECK-NEXT:    vmsne.vx v10, v8, a0, v0.t
 ; CHECK-NEXT:    vmv.v.v v0, v10
 ; CHECK-NEXT:    ret
@@ -1563,9 +1563,9 @@ declare <vscale x 4 x i1> @llvm.riscv.vmsne.mask.nxv4i32.i32(
 define <vscale x 4 x i1> @intrinsic_vmsne_mask_vx_nxv4i32_i32(<vscale x 4 x i1> %0, <vscale x 4 x i32> %1, i32 %2, <vscale x 4 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsne_mask_vx_nxv4i32_i32:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e32, m2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmv1r.v v0, v10
-; CHECK-NEXT:    vsetvli zero, a1, e32, m2, ta, mu
 ; CHECK-NEXT:    vmsne.vx v11, v8, a0, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v11
 ; CHECK-NEXT:    ret
@@ -1610,9 +1610,9 @@ declare <vscale x 8 x i1> @llvm.riscv.vmsne.mask.nxv8i32.i32(
 define <vscale x 8 x i1> @intrinsic_vmsne_mask_vx_nxv8i32_i32(<vscale x 8 x i1> %0, <vscale x 8 x i32> %1, i32 %2, <vscale x 8 x i1> %3, iXLen %4) nounwind {
 ; CHECK-LABEL: intrinsic_vmsne_mask_vx_nxv8i32_i32:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a1, e32, m4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v13, v0
 ; CHECK-NEXT:    vmv1r.v v0, v12
-; CHECK-NEXT:    vsetvli zero, a1, e32, m4, ta, mu
 ; CHECK-NEXT:    vmsne.vx v13, v8, a0, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v13
 ; CHECK-NEXT:    ret
@@ -1684,9 +1684,9 @@ define <vscale x 1 x i1> @intrinsic_vmsne_mask_vx_nxv1i64_i64(<vscale x 1 x i1>
 ;
 ; RV64-LABEL: intrinsic_vmsne_mask_vx_nxv1i64_i64:
 ; RV64:       # %bb.0: # %entry
+; RV64-NEXT:    vsetvli zero, a1, e64, m1, ta, mu
 ; RV64-NEXT:    vmv1r.v v10, v0
 ; RV64-NEXT:    vmv1r.v v0, v9
-; RV64-NEXT:    vsetvli zero, a1, e64, m1, ta, mu
 ; RV64-NEXT:    vmsne.vx v10, v8, a0, v0.t
 ; RV64-NEXT:    vmv.v.v v0, v10
 ; RV64-NEXT:    ret
@@ -1758,9 +1758,9 @@ define <vscale x 2 x i1> @intrinsic_vmsne_mask_vx_nxv2i64_i64(<vscale x 2 x i1>
 ;
 ; RV64-LABEL: intrinsic_vmsne_mask_vx_nxv2i64_i64:
 ; RV64:       # %bb.0: # %entry
+; RV64-NEXT:    vsetvli zero, a1, e64, m2, ta, mu
 ; RV64-NEXT:    vmv1r.v v11, v0
 ; RV64-NEXT:    vmv1r.v v0, v10
-; RV64-NEXT:    vsetvli zero, a1, e64, m2, ta, mu
 ; RV64-NEXT:    vmsne.vx v11, v8, a0, v0.t
 ; RV64-NEXT:    vmv1r.v v0, v11
 ; RV64-NEXT:    ret
@@ -1832,9 +1832,9 @@ define <vscale x 4 x i1> @intrinsic_vmsne_mask_vx_nxv4i64_i64(<vscale x 4 x i1>
 ;
 ; RV64-LABEL: intrinsic_vmsne_mask_vx_nxv4i64_i64:
 ; RV64:       # %bb.0: # %entry
+; RV64-NEXT:    vsetvli zero, a1, e64, m4, ta, mu
 ; RV64-NEXT:    vmv1r.v v13, v0
 ; RV64-NEXT:    vmv1r.v v0, v12
-; RV64-NEXT:    vsetvli zero, a1, e64, m4, ta, mu
 ; RV64-NEXT:    vmsne.vx v13, v8, a0, v0.t
 ; RV64-NEXT:    vmv1r.v v0, v13
 ; RV64-NEXT:    ret
@@ -1867,9 +1867,9 @@ entry:
 define <vscale x 1 x i1> @intrinsic_vmsne_mask_vi_nxv1i8_i8(<vscale x 1 x i1> %0, <vscale x 1 x i8> %1, <vscale x 1 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsne_mask_vi_nxv1i8_i8:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e8, mf8, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a0, e8, mf8, ta, mu
 ; CHECK-NEXT:    vmsne.vi v10, v8, 9, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v10
 ; CHECK-NEXT:    ret
@@ -1902,9 +1902,9 @@ entry:
 define <vscale x 2 x i1> @intrinsic_vmsne_mask_vi_nxv2i8_i8(<vscale x 2 x i1> %0, <vscale x 2 x i8> %1, <vscale x 2 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsne_mask_vi_nxv2i8_i8:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e8, mf4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a0, e8, mf4, ta, mu
 ; CHECK-NEXT:    vmsne.vi v10, v8, 9, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v10
 ; CHECK-NEXT:    ret
@@ -1937,9 +1937,9 @@ entry:
 define <vscale x 4 x i1> @intrinsic_vmsne_mask_vi_nxv4i8_i8(<vscale x 4 x i1> %0, <vscale x 4 x i8> %1, <vscale x 4 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsne_mask_vi_nxv4i8_i8:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e8, mf2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a0, e8, mf2, ta, mu
 ; CHECK-NEXT:    vmsne.vi v10, v8, 9, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v10
 ; CHECK-NEXT:    ret
@@ -1972,9 +1972,9 @@ entry:
 define <vscale x 8 x i1> @intrinsic_vmsne_mask_vi_nxv8i8_i8(<vscale x 8 x i1> %0, <vscale x 8 x i8> %1, <vscale x 8 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsne_mask_vi_nxv8i8_i8:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e8, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a0, e8, m1, ta, mu
 ; CHECK-NEXT:    vmsne.vi v10, v8, 9, v0.t
 ; CHECK-NEXT:    vmv.v.v v0, v10
 ; CHECK-NEXT:    ret
@@ -2007,9 +2007,9 @@ entry:
 define <vscale x 16 x i1> @intrinsic_vmsne_mask_vi_nxv16i8_i8(<vscale x 16 x i1> %0, <vscale x 16 x i8> %1, <vscale x 16 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsne_mask_vi_nxv16i8_i8:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e8, m2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmv1r.v v0, v10
-; CHECK-NEXT:    vsetvli zero, a0, e8, m2, ta, mu
 ; CHECK-NEXT:    vmsne.vi v11, v8, 9, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v11
 ; CHECK-NEXT:    ret
@@ -2042,9 +2042,9 @@ entry:
 define <vscale x 32 x i1> @intrinsic_vmsne_mask_vi_nxv32i8_i8(<vscale x 32 x i1> %0, <vscale x 32 x i8> %1, <vscale x 32 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsne_mask_vi_nxv32i8_i8:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e8, m4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v13, v0
 ; CHECK-NEXT:    vmv1r.v v0, v12
-; CHECK-NEXT:    vsetvli zero, a0, e8, m4, ta, mu
 ; CHECK-NEXT:    vmsne.vi v13, v8, 9, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v13
 ; CHECK-NEXT:    ret
@@ -2077,9 +2077,9 @@ entry:
 define <vscale x 1 x i1> @intrinsic_vmsne_mask_vi_nxv1i16_i16(<vscale x 1 x i1> %0, <vscale x 1 x i16> %1, <vscale x 1 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsne_mask_vi_nxv1i16_i16:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e16, mf4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a0, e16, mf4, ta, mu
 ; CHECK-NEXT:    vmsne.vi v10, v8, 9, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v10
 ; CHECK-NEXT:    ret
@@ -2112,9 +2112,9 @@ entry:
 define <vscale x 2 x i1> @intrinsic_vmsne_mask_vi_nxv2i16_i16(<vscale x 2 x i1> %0, <vscale x 2 x i16> %1, <vscale x 2 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsne_mask_vi_nxv2i16_i16:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e16, mf2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a0, e16, mf2, ta, mu
 ; CHECK-NEXT:    vmsne.vi v10, v8, 9, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v10
 ; CHECK-NEXT:    ret
@@ -2147,9 +2147,9 @@ entry:
 define <vscale x 4 x i1> @intrinsic_vmsne_mask_vi_nxv4i16_i16(<vscale x 4 x i1> %0, <vscale x 4 x i16> %1, <vscale x 4 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsne_mask_vi_nxv4i16_i16:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e16, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a0, e16, m1, ta, mu
 ; CHECK-NEXT:    vmsne.vi v10, v8, 9, v0.t
 ; CHECK-NEXT:    vmv.v.v v0, v10
 ; CHECK-NEXT:    ret
@@ -2182,9 +2182,9 @@ entry:
 define <vscale x 8 x i1> @intrinsic_vmsne_mask_vi_nxv8i16_i16(<vscale x 8 x i1> %0, <vscale x 8 x i16> %1, <vscale x 8 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsne_mask_vi_nxv8i16_i16:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e16, m2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmv1r.v v0, v10
-; CHECK-NEXT:    vsetvli zero, a0, e16, m2, ta, mu
 ; CHECK-NEXT:    vmsne.vi v11, v8, 9, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v11
 ; CHECK-NEXT:    ret
@@ -2217,9 +2217,9 @@ entry:
 define <vscale x 16 x i1> @intrinsic_vmsne_mask_vi_nxv16i16_i16(<vscale x 16 x i1> %0, <vscale x 16 x i16> %1, <vscale x 16 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsne_mask_vi_nxv16i16_i16:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e16, m4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v13, v0
 ; CHECK-NEXT:    vmv1r.v v0, v12
-; CHECK-NEXT:    vsetvli zero, a0, e16, m4, ta, mu
 ; CHECK-NEXT:    vmsne.vi v13, v8, 9, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v13
 ; CHECK-NEXT:    ret
@@ -2252,9 +2252,9 @@ entry:
 define <vscale x 1 x i1> @intrinsic_vmsne_mask_vi_nxv1i32_i32(<vscale x 1 x i1> %0, <vscale x 1 x i32> %1, <vscale x 1 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsne_mask_vi_nxv1i32_i32:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e32, mf2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a0, e32, mf2, ta, mu
 ; CHECK-NEXT:    vmsne.vi v10, v8, 9, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v10
 ; CHECK-NEXT:    ret
@@ -2287,9 +2287,9 @@ entry:
 define <vscale x 2 x i1> @intrinsic_vmsne_mask_vi_nxv2i32_i32(<vscale x 2 x i1> %0, <vscale x 2 x i32> %1, <vscale x 2 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsne_mask_vi_nxv2i32_i32:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e32, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a0, e32, m1, ta, mu
 ; CHECK-NEXT:    vmsne.vi v10, v8, 9, v0.t
 ; CHECK-NEXT:    vmv.v.v v0, v10
 ; CHECK-NEXT:    ret
@@ -2322,9 +2322,9 @@ entry:
 define <vscale x 4 x i1> @intrinsic_vmsne_mask_vi_nxv4i32_i32(<vscale x 4 x i1> %0, <vscale x 4 x i32> %1, <vscale x 4 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsne_mask_vi_nxv4i32_i32:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e32, m2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmv1r.v v0, v10
-; CHECK-NEXT:    vsetvli zero, a0, e32, m2, ta, mu
 ; CHECK-NEXT:    vmsne.vi v11, v8, 9, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v11
 ; CHECK-NEXT:    ret
@@ -2357,9 +2357,9 @@ entry:
 define <vscale x 8 x i1> @intrinsic_vmsne_mask_vi_nxv8i32_i32(<vscale x 8 x i1> %0, <vscale x 8 x i32> %1, <vscale x 8 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsne_mask_vi_nxv8i32_i32:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e32, m4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v13, v0
 ; CHECK-NEXT:    vmv1r.v v0, v12
-; CHECK-NEXT:    vsetvli zero, a0, e32, m4, ta, mu
 ; CHECK-NEXT:    vmsne.vi v13, v8, 9, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v13
 ; CHECK-NEXT:    ret
@@ -2392,9 +2392,9 @@ entry:
 define <vscale x 1 x i1> @intrinsic_vmsne_mask_vi_nxv1i64_i64(<vscale x 1 x i1> %0, <vscale x 1 x i64> %1, <vscale x 1 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsne_mask_vi_nxv1i64_i64:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e64, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a0, e64, m1, ta, mu
 ; CHECK-NEXT:    vmsne.vi v10, v8, 9, v0.t
 ; CHECK-NEXT:    vmv.v.v v0, v10
 ; CHECK-NEXT:    ret
@@ -2427,9 +2427,9 @@ entry:
 define <vscale x 2 x i1> @intrinsic_vmsne_mask_vi_nxv2i64_i64(<vscale x 2 x i1> %0, <vscale x 2 x i64> %1, <vscale x 2 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsne_mask_vi_nxv2i64_i64:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e64, m2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    vmv1r.v v0, v10
-; CHECK-NEXT:    vsetvli zero, a0, e64, m2, ta, mu
 ; CHECK-NEXT:    vmsne.vi v11, v8, 9, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v11
 ; CHECK-NEXT:    ret
@@ -2462,9 +2462,9 @@ entry:
 define <vscale x 4 x i1> @intrinsic_vmsne_mask_vi_nxv4i64_i64(<vscale x 4 x i1> %0, <vscale x 4 x i64> %1, <vscale x 4 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsne_mask_vi_nxv4i64_i64:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e64, m4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v13, v0
 ; CHECK-NEXT:    vmv1r.v v0, v12
-; CHECK-NEXT:    vsetvli zero, a0, e64, m4, ta, mu
 ; CHECK-NEXT:    vmsne.vi v13, v8, 9, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v13
 ; CHECK-NEXT:    ret
diff --git a/llvm/test/CodeGen/RISCV/rvv/vmsof.ll b/llvm/test/CodeGen/RISCV/rvv/vmsof.ll
index b0a28e6e455b0..4b818a2b1e58f 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vmsof.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vmsof.ll
@@ -31,9 +31,9 @@ declare <vscale x 1 x i1> @llvm.riscv.vmsof.mask.nxv1i1(
 define <vscale x 1 x i1> @intrinsic_vmsof_mask_m_nxv1i1_nxv1i1(<vscale x 1 x i1> %0, <vscale x 1 x i1> %1, <vscale x 1 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsof_mask_m_nxv1i1_nxv1i1:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e8, mf8, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a0, e8, mf8, ta, mu
 ; CHECK-NEXT:    vmsof.m v10, v8, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v10
 ; CHECK-NEXT:    ret
@@ -73,9 +73,9 @@ declare <vscale x 2 x i1> @llvm.riscv.vmsof.mask.nxv2i1(
 define <vscale x 2 x i1> @intrinsic_vmsof_mask_m_nxv2i1_nxv2i1(<vscale x 2 x i1> %0, <vscale x 2 x i1> %1, <vscale x 2 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsof_mask_m_nxv2i1_nxv2i1:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e8, mf4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a0, e8, mf4, ta, mu
 ; CHECK-NEXT:    vmsof.m v10, v8, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v10
 ; CHECK-NEXT:    ret
@@ -115,9 +115,9 @@ declare <vscale x 4 x i1> @llvm.riscv.vmsof.mask.nxv4i1(
 define <vscale x 4 x i1> @intrinsic_vmsof_mask_m_nxv4i1_nxv4i1(<vscale x 4 x i1> %0, <vscale x 4 x i1> %1, <vscale x 4 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsof_mask_m_nxv4i1_nxv4i1:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e8, mf2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a0, e8, mf2, ta, mu
 ; CHECK-NEXT:    vmsof.m v10, v8, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v10
 ; CHECK-NEXT:    ret
@@ -157,9 +157,9 @@ declare <vscale x 8 x i1> @llvm.riscv.vmsof.mask.nxv8i1(
 define <vscale x 8 x i1> @intrinsic_vmsof_mask_m_nxv8i1_nxv8i1(<vscale x 8 x i1> %0, <vscale x 8 x i1> %1, <vscale x 8 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsof_mask_m_nxv8i1_nxv8i1:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e8, m1, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a0, e8, m1, ta, mu
 ; CHECK-NEXT:    vmsof.m v10, v8, v0.t
 ; CHECK-NEXT:    vmv.v.v v0, v10
 ; CHECK-NEXT:    ret
@@ -199,9 +199,9 @@ declare <vscale x 16 x i1> @llvm.riscv.vmsof.mask.nxv16i1(
 define <vscale x 16 x i1> @intrinsic_vmsof_mask_m_nxv16i1_nxv16i1(<vscale x 16 x i1> %0, <vscale x 16 x i1> %1, <vscale x 16 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsof_mask_m_nxv16i1_nxv16i1:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e8, m2, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a0, e8, m2, ta, mu
 ; CHECK-NEXT:    vmsof.m v10, v8, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v10
 ; CHECK-NEXT:    ret
@@ -241,9 +241,9 @@ declare <vscale x 32 x i1> @llvm.riscv.vmsof.mask.nxv32i1(
 define <vscale x 32 x i1> @intrinsic_vmsof_mask_m_nxv32i1_nxv32i1(<vscale x 32 x i1> %0, <vscale x 32 x i1> %1, <vscale x 32 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsof_mask_m_nxv32i1_nxv32i1:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e8, m4, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a0, e8, m4, ta, mu
 ; CHECK-NEXT:    vmsof.m v10, v8, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v10
 ; CHECK-NEXT:    ret
@@ -283,9 +283,9 @@ declare <vscale x 64 x i1> @llvm.riscv.vmsof.mask.nxv64i1(
 define <vscale x 64 x i1> @intrinsic_vmsof_mask_m_nxv64i1_nxv64i1(<vscale x 64 x i1> %0, <vscale x 64 x i1> %1, <vscale x 64 x i1> %2, iXLen %3) nounwind {
 ; CHECK-LABEL: intrinsic_vmsof_mask_m_nxv64i1_nxv64i1:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli zero, a0, e8, m8, ta, mu
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v9
-; CHECK-NEXT:    vsetvli zero, a0, e8, m8, ta, mu
 ; CHECK-NEXT:    vmsof.m v10, v8, v0.t
 ; CHECK-NEXT:    vmv1r.v v0, v10
 ; CHECK-NEXT:    ret
diff --git a/llvm/test/CodeGen/RISCV/rvv/vmv.v.v-peephole.ll b/llvm/test/CodeGen/RISCV/rvv/vmv.v.v-peephole.ll
index 7f248a39b54fa..6345b90db23b8 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vmv.v.v-peephole.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vmv.v.v-peephole.ll
@@ -49,8 +49,8 @@ define <vscale x 4 x i32> @vadd_same_passthru(<vscale x 4 x i32> %passthru, <vsc
 define <vscale x 4 x i32> @unfoldable_diff_avl_unknown(<vscale x 4 x i32> %passthru, <vscale x 4 x i32> %a, <vscale x 4 x i32> %b, iXLen %vl1, iXLen %vl2) {
 ; CHECK-LABEL: unfoldable_diff_avl_unknown:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv2r.v v14, v8
 ; CHECK-NEXT:    vsetvli zero, a0, e32, m2, tu, ma
+; CHECK-NEXT:    vmv2r.v v14, v8
 ; CHECK-NEXT:    vadd.vv v14, v10, v12
 ; CHECK-NEXT:    vsetvli zero, a1, e32, m2, tu, ma
 ; CHECK-NEXT:    vmv.v.v v8, v14
diff --git a/llvm/test/CodeGen/RISCV/rvv/vp-cttz-elts.ll b/llvm/test/CodeGen/RISCV/rvv/vp-cttz-elts.ll
index f7ca65801dc87..b316f5f878816 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vp-cttz-elts.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vp-cttz-elts.ll
@@ -5,9 +5,9 @@
 define iXLen @bool_vec(<vscale x 2 x i1> %src, <vscale x 2 x i1> %m, i32 %evl) {
 ; RV32-LABEL: bool_vec:
 ; RV32:       # %bb.0:
+; RV32-NEXT:    vsetvli zero, a0, e8, mf4, ta, ma
 ; RV32-NEXT:    vmv1r.v v9, v0
 ; RV32-NEXT:    vmv1r.v v0, v8
-; RV32-NEXT:    vsetvli zero, a0, e8, mf4, ta, ma
 ; RV32-NEXT:    vfirst.m a1, v9, v0.t
 ; RV32-NEXT:    bltz a1, .LBB0_2
 ; RV32-NEXT:  # %bb.1:
@@ -17,6 +17,7 @@ define iXLen @bool_vec(<vscale x 2 x i1> %src, <vscale x 2 x i1> %m, i32 %evl) {
 ;
 ; RV64-LABEL: bool_vec:
 ; RV64:       # %bb.0:
+; RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; RV64-NEXT:    vmv1r.v v9, v0
 ; RV64-NEXT:    slli a0, a0, 32
 ; RV64-NEXT:    srli a0, a0, 32
@@ -35,14 +36,15 @@ define iXLen @bool_vec(<vscale x 2 x i1> %src, <vscale x 2 x i1> %m, i32 %evl) {
 define iXLen @bool_vec_zero_poison(<vscale x 2 x i1> %src, <vscale x 2 x i1> %m, i32 %evl) {
 ; RV32-LABEL: bool_vec_zero_poison:
 ; RV32:       # %bb.0:
+; RV32-NEXT:    vsetvli zero, a0, e8, mf4, ta, ma
 ; RV32-NEXT:    vmv1r.v v9, v0
 ; RV32-NEXT:    vmv1r.v v0, v8
-; RV32-NEXT:    vsetvli zero, a0, e8, mf4, ta, ma
 ; RV32-NEXT:    vfirst.m a0, v9, v0.t
 ; RV32-NEXT:    ret
 ;
 ; RV64-LABEL: bool_vec_zero_poison:
 ; RV64:       # %bb.0:
+; RV64-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; RV64-NEXT:    vmv1r.v v9, v0
 ; RV64-NEXT:    slli a0, a0, 32
 ; RV64-NEXT:    srli a0, a0, 32
diff --git a/llvm/test/CodeGen/RISCV/rvv/vp-select.ll b/llvm/test/CodeGen/RISCV/rvv/vp-select.ll
index c8a048971a803..652baf6692341 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vp-select.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vp-select.ll
@@ -12,6 +12,7 @@ define <vscale x 1 x i64> @all_ones(<vscale x 1 x i64> %true, <vscale x 1 x i64>
 define <vscale x 1 x i64> @all_zeroes(<vscale x 1 x i64> %true, <vscale x 1 x i64> %false, i32 %evl) {
 ; CHECK-LABEL: all_zeroes:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    ret
   %v = call <vscale x 1 x i64> @llvm.vp.select.nxv1i64(<vscale x 1 x i1> splat (i1 false), <vscale x 1 x i64> %true, <vscale x 1 x i64> %false, i32 %evl)
diff --git a/llvm/test/CodeGen/RISCV/rvv/vp-splice-mask-fixed-vectors.ll b/llvm/test/CodeGen/RISCV/rvv/vp-splice-mask-fixed-vectors.ll
index 2a137099bcb0f..745cec4e7c4f6 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vp-splice-mask-fixed-vectors.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vp-splice-mask-fixed-vectors.ll
@@ -10,9 +10,9 @@ declare <16 x i1> @llvm.experimental.vp.splice.v16i1(<16 x i1>, <16 x i1>, i32,
 define <2 x i1> @test_vp_splice_v2i1(<2 x i1> %va, <2 x i1> %vb, i32 zeroext %evla, i32 zeroext %evlb) {
 ; CHECK-LABEL: test_vp_splice_v2i1:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a1, e8, mf8, ta, ma
 ; CHECK-NEXT:    vmv1r.v v9, v0
 ; CHECK-NEXT:    vmv1r.v v0, v8
-; CHECK-NEXT:    vsetvli zero, a1, e8, mf8, ta, ma
 ; CHECK-NEXT:    vmv.v.i v8, 0
 ; CHECK-NEXT:    vmerge.vim v8, v8, 1, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e8, mf8, ta, ma
@@ -34,9 +34,9 @@ define <2 x i1> @test_vp_splice_v2i1(<2 x i1> %va, <2 x i1> %vb, i32 zeroext %ev
 define <2 x i1> @test_vp_splice_v2i1_negative_offset(<2 x i1> %va, <2 x i1> %vb, i32 zeroext %evla, i32 zeroext %evlb) {
 ; CHECK-LABEL: test_vp_splice_v2i1_negative_offset:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a1, e8, mf8, ta, ma
 ; CHECK-NEXT:    vmv1r.v v9, v0
 ; CHECK-NEXT:    vmv1r.v v0, v8
-; CHECK-NEXT:    vsetvli zero, a1, e8, mf8, ta, ma
 ; CHECK-NEXT:    vmv.v.i v8, 0
 ; CHECK-NEXT:    vmerge.vim v8, v8, 1, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e8, mf8, ta, ma
@@ -58,9 +58,9 @@ define <2 x i1> @test_vp_splice_v2i1_negative_offset(<2 x i1> %va, <2 x i1> %vb,
 define <2 x i1> @test_vp_splice_v2i1_masked(<2 x i1> %va, <2 x i1> %vb, <2 x i1> %mask, i32 zeroext %evla, i32 zeroext %evlb) {
 ; CHECK-LABEL: test_vp_splice_v2i1_masked:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a1, e8, mf8, ta, ma
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v8
-; CHECK-NEXT:    vsetvli zero, a1, e8, mf8, ta, ma
 ; CHECK-NEXT:    vmv.v.i v8, 0
 ; CHECK-NEXT:    vmerge.vim v8, v8, 1, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e8, mf8, ta, ma
@@ -83,9 +83,9 @@ define <2 x i1> @test_vp_splice_v2i1_masked(<2 x i1> %va, <2 x i1> %vb, <2 x i1>
 define <4 x i1> @test_vp_splice_v4i1(<4 x i1> %va, <4 x i1> %vb, i32 zeroext %evla, i32 zeroext %evlb) {
 ; CHECK-LABEL: test_vp_splice_v4i1:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a1, e8, mf4, ta, ma
 ; CHECK-NEXT:    vmv1r.v v9, v0
 ; CHECK-NEXT:    vmv1r.v v0, v8
-; CHECK-NEXT:    vsetvli zero, a1, e8, mf4, ta, ma
 ; CHECK-NEXT:    vmv.v.i v8, 0
 ; CHECK-NEXT:    vmerge.vim v8, v8, 1, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e8, mf4, ta, ma
@@ -107,9 +107,9 @@ define <4 x i1> @test_vp_splice_v4i1(<4 x i1> %va, <4 x i1> %vb, i32 zeroext %ev
 define <4 x i1> @test_vp_splice_v4i1_negative_offset(<4 x i1> %va, <4 x i1> %vb, i32 zeroext %evla, i32 zeroext %evlb) {
 ; CHECK-LABEL: test_vp_splice_v4i1_negative_offset:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a1, e8, mf4, ta, ma
 ; CHECK-NEXT:    vmv1r.v v9, v0
 ; CHECK-NEXT:    vmv1r.v v0, v8
-; CHECK-NEXT:    vsetvli zero, a1, e8, mf4, ta, ma
 ; CHECK-NEXT:    vmv.v.i v8, 0
 ; CHECK-NEXT:    vmerge.vim v8, v8, 1, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e8, mf4, ta, ma
@@ -131,9 +131,9 @@ define <4 x i1> @test_vp_splice_v4i1_negative_offset(<4 x i1> %va, <4 x i1> %vb,
 define <4 x i1> @test_vp_splice_v4i1_masked(<4 x i1> %va, <4 x i1> %vb, <4 x i1> %mask, i32 zeroext %evla, i32 zeroext %evlb) {
 ; CHECK-LABEL: test_vp_splice_v4i1_masked:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a1, e8, mf4, ta, ma
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v8
-; CHECK-NEXT:    vsetvli zero, a1, e8, mf4, ta, ma
 ; CHECK-NEXT:    vmv.v.i v8, 0
 ; CHECK-NEXT:    vmerge.vim v8, v8, 1, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e8, mf4, ta, ma
@@ -156,9 +156,9 @@ define <4 x i1> @test_vp_splice_v4i1_masked(<4 x i1> %va, <4 x i1> %vb, <4 x i1>
 define <8 x i1> @test_vp_splice_v8i1(<8 x i1> %va, <8 x i1> %vb, i32 zeroext %evla, i32 zeroext %evlb) {
 ; CHECK-LABEL: test_vp_splice_v8i1:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a1, e8, mf2, ta, ma
 ; CHECK-NEXT:    vmv1r.v v9, v0
 ; CHECK-NEXT:    vmv1r.v v0, v8
-; CHECK-NEXT:    vsetvli zero, a1, e8, mf2, ta, ma
 ; CHECK-NEXT:    vmv.v.i v8, 0
 ; CHECK-NEXT:    vmerge.vim v8, v8, 1, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e8, mf2, ta, ma
@@ -180,9 +180,9 @@ define <8 x i1> @test_vp_splice_v8i1(<8 x i1> %va, <8 x i1> %vb, i32 zeroext %ev
 define <8 x i1> @test_vp_splice_v8i1_negative_offset(<8 x i1> %va, <8 x i1> %vb, i32 zeroext %evla, i32 zeroext %evlb) {
 ; CHECK-LABEL: test_vp_splice_v8i1_negative_offset:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a1, e8, mf2, ta, ma
 ; CHECK-NEXT:    vmv1r.v v9, v0
 ; CHECK-NEXT:    vmv1r.v v0, v8
-; CHECK-NEXT:    vsetvli zero, a1, e8, mf2, ta, ma
 ; CHECK-NEXT:    vmv.v.i v8, 0
 ; CHECK-NEXT:    vmerge.vim v8, v8, 1, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e8, mf2, ta, ma
@@ -204,9 +204,9 @@ define <8 x i1> @test_vp_splice_v8i1_negative_offset(<8 x i1> %va, <8 x i1> %vb,
 define <8 x i1> @test_vp_splice_v8i1_masked(<8 x i1> %va, <8 x i1> %vb, <8 x i1> %mask, i32 zeroext %evla, i32 zeroext %evlb) {
 ; CHECK-LABEL: test_vp_splice_v8i1_masked:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a1, e8, mf2, ta, ma
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v8
-; CHECK-NEXT:    vsetvli zero, a1, e8, mf2, ta, ma
 ; CHECK-NEXT:    vmv.v.i v8, 0
 ; CHECK-NEXT:    vmerge.vim v8, v8, 1, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e8, mf2, ta, ma
@@ -229,9 +229,9 @@ define <8 x i1> @test_vp_splice_v8i1_masked(<8 x i1> %va, <8 x i1> %vb, <8 x i1>
 define <16 x i1> @test_vp_splice_v16i1(<16 x i1> %va, <16 x i1> %vb, i32 zeroext %evla, i32 zeroext %evlb) {
 ; CHECK-LABEL: test_vp_splice_v16i1:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a1, e8, m1, ta, ma
 ; CHECK-NEXT:    vmv1r.v v9, v0
 ; CHECK-NEXT:    vmv1r.v v0, v8
-; CHECK-NEXT:    vsetvli zero, a1, e8, m1, ta, ma
 ; CHECK-NEXT:    vmv.v.i v8, 0
 ; CHECK-NEXT:    vmerge.vim v8, v8, 1, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e8, m1, ta, ma
@@ -253,9 +253,9 @@ define <16 x i1> @test_vp_splice_v16i1(<16 x i1> %va, <16 x i1> %vb, i32 zeroext
 define <16 x i1> @test_vp_splice_v16i1_negative_offset(<16 x i1> %va, <16 x i1> %vb, i32 zeroext %evla, i32 zeroext %evlb) {
 ; CHECK-LABEL: test_vp_splice_v16i1_negative_offset:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a1, e8, m1, ta, ma
 ; CHECK-NEXT:    vmv1r.v v9, v0
 ; CHECK-NEXT:    vmv1r.v v0, v8
-; CHECK-NEXT:    vsetvli zero, a1, e8, m1, ta, ma
 ; CHECK-NEXT:    vmv.v.i v8, 0
 ; CHECK-NEXT:    vmerge.vim v8, v8, 1, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e8, m1, ta, ma
@@ -277,9 +277,9 @@ define <16 x i1> @test_vp_splice_v16i1_negative_offset(<16 x i1> %va, <16 x i1>
 define <16 x i1> @test_vp_splice_v16i1_masked(<16 x i1> %va, <16 x i1> %vb, <16 x i1> %mask, i32 zeroext %evla, i32 zeroext %evlb) {
 ; CHECK-LABEL: test_vp_splice_v16i1_masked:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a1, e8, m1, ta, ma
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v8
-; CHECK-NEXT:    vsetvli zero, a1, e8, m1, ta, ma
 ; CHECK-NEXT:    vmv.v.i v8, 0
 ; CHECK-NEXT:    vmerge.vim v8, v8, 1, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e8, m1, ta, ma
diff --git a/llvm/test/CodeGen/RISCV/rvv/vp-splice-mask-vectors.ll b/llvm/test/CodeGen/RISCV/rvv/vp-splice-mask-vectors.ll
index fc446d0a3a88a..3b0b183537468 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vp-splice-mask-vectors.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vp-splice-mask-vectors.ll
@@ -13,9 +13,9 @@ declare <vscale x 64 x i1> @llvm.experimental.vp.splice.nxv64i1(<vscale x 64 x i
 define <vscale x 1 x i1> @test_vp_splice_nxv1i1(<vscale x 1 x i1> %va, <vscale x 1 x i1> %vb, i32 zeroext %evla, i32 zeroext %evlb) {
 ; CHECK-LABEL: test_vp_splice_nxv1i1:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a1, e8, mf8, ta, ma
 ; CHECK-NEXT:    vmv1r.v v9, v0
 ; CHECK-NEXT:    vmv1r.v v0, v8
-; CHECK-NEXT:    vsetvli zero, a1, e8, mf8, ta, ma
 ; CHECK-NEXT:    vmv.v.i v8, 0
 ; CHECK-NEXT:    vmerge.vim v8, v8, 1, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e8, mf8, ta, ma
@@ -37,9 +37,9 @@ define <vscale x 1 x i1> @test_vp_splice_nxv1i1(<vscale x 1 x i1> %va, <vscale x
 define <vscale x 1 x i1> @test_vp_splice_nxv1i1_negative_offset(<vscale x 1 x i1> %va, <vscale x 1 x i1> %vb, i32 zeroext %evla, i32 zeroext %evlb) {
 ; CHECK-LABEL: test_vp_splice_nxv1i1_negative_offset:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a1, e8, mf8, ta, ma
 ; CHECK-NEXT:    vmv1r.v v9, v0
 ; CHECK-NEXT:    vmv1r.v v0, v8
-; CHECK-NEXT:    vsetvli zero, a1, e8, mf8, ta, ma
 ; CHECK-NEXT:    vmv.v.i v8, 0
 ; CHECK-NEXT:    vmerge.vim v8, v8, 1, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e8, mf8, ta, ma
@@ -61,9 +61,9 @@ define <vscale x 1 x i1> @test_vp_splice_nxv1i1_negative_offset(<vscale x 1 x i1
 define <vscale x 1 x i1> @test_vp_splice_nxv1i1_masked(<vscale x 1 x i1> %va, <vscale x 1 x i1> %vb, <vscale x 1 x i1> %mask, i32 zeroext %evla, i32 zeroext %evlb) {
 ; CHECK-LABEL: test_vp_splice_nxv1i1_masked:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a1, e8, mf8, ta, ma
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v8
-; CHECK-NEXT:    vsetvli zero, a1, e8, mf8, ta, ma
 ; CHECK-NEXT:    vmv.v.i v8, 0
 ; CHECK-NEXT:    vmerge.vim v8, v8, 1, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e8, mf8, ta, ma
@@ -86,9 +86,9 @@ define <vscale x 1 x i1> @test_vp_splice_nxv1i1_masked(<vscale x 1 x i1> %va, <v
 define <vscale x 2 x i1> @test_vp_splice_nxv2i1(<vscale x 2 x i1> %va, <vscale x 2 x i1> %vb, i32 zeroext %evla, i32 zeroext %evlb) {
 ; CHECK-LABEL: test_vp_splice_nxv2i1:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a1, e8, mf4, ta, ma
 ; CHECK-NEXT:    vmv1r.v v9, v0
 ; CHECK-NEXT:    vmv1r.v v0, v8
-; CHECK-NEXT:    vsetvli zero, a1, e8, mf4, ta, ma
 ; CHECK-NEXT:    vmv.v.i v8, 0
 ; CHECK-NEXT:    vmerge.vim v8, v8, 1, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e8, mf4, ta, ma
@@ -110,9 +110,9 @@ define <vscale x 2 x i1> @test_vp_splice_nxv2i1(<vscale x 2 x i1> %va, <vscale x
 define <vscale x 2 x i1> @test_vp_splice_nxv2i1_negative_offset(<vscale x 2 x i1> %va, <vscale x 2 x i1> %vb, i32 zeroext %evla, i32 zeroext %evlb) {
 ; CHECK-LABEL: test_vp_splice_nxv2i1_negative_offset:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a1, e8, mf4, ta, ma
 ; CHECK-NEXT:    vmv1r.v v9, v0
 ; CHECK-NEXT:    vmv1r.v v0, v8
-; CHECK-NEXT:    vsetvli zero, a1, e8, mf4, ta, ma
 ; CHECK-NEXT:    vmv.v.i v8, 0
 ; CHECK-NEXT:    vmerge.vim v8, v8, 1, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e8, mf4, ta, ma
@@ -134,9 +134,9 @@ define <vscale x 2 x i1> @test_vp_splice_nxv2i1_negative_offset(<vscale x 2 x i1
 define <vscale x 2 x i1> @test_vp_splice_nxv2i1_masked(<vscale x 2 x i1> %va, <vscale x 2 x i1> %vb, <vscale x 2 x i1> %mask, i32 zeroext %evla, i32 zeroext %evlb) {
 ; CHECK-LABEL: test_vp_splice_nxv2i1_masked:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a1, e8, mf4, ta, ma
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v8
-; CHECK-NEXT:    vsetvli zero, a1, e8, mf4, ta, ma
 ; CHECK-NEXT:    vmv.v.i v8, 0
 ; CHECK-NEXT:    vmerge.vim v8, v8, 1, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e8, mf4, ta, ma
@@ -159,9 +159,9 @@ define <vscale x 2 x i1> @test_vp_splice_nxv2i1_masked(<vscale x 2 x i1> %va, <v
 define <vscale x 4 x i1> @test_vp_splice_nxv4i1(<vscale x 4 x i1> %va, <vscale x 4 x i1> %vb, i32 zeroext %evla, i32 zeroext %evlb) {
 ; CHECK-LABEL: test_vp_splice_nxv4i1:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a1, e8, mf2, ta, ma
 ; CHECK-NEXT:    vmv1r.v v9, v0
 ; CHECK-NEXT:    vmv1r.v v0, v8
-; CHECK-NEXT:    vsetvli zero, a1, e8, mf2, ta, ma
 ; CHECK-NEXT:    vmv.v.i v8, 0
 ; CHECK-NEXT:    vmerge.vim v8, v8, 1, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e8, mf2, ta, ma
@@ -183,9 +183,9 @@ define <vscale x 4 x i1> @test_vp_splice_nxv4i1(<vscale x 4 x i1> %va, <vscale x
 define <vscale x 4 x i1> @test_vp_splice_nxv4i1_negative_offset(<vscale x 4 x i1> %va, <vscale x 4 x i1> %vb, i32 zeroext %evla, i32 zeroext %evlb) {
 ; CHECK-LABEL: test_vp_splice_nxv4i1_negative_offset:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a1, e8, mf2, ta, ma
 ; CHECK-NEXT:    vmv1r.v v9, v0
 ; CHECK-NEXT:    vmv1r.v v0, v8
-; CHECK-NEXT:    vsetvli zero, a1, e8, mf2, ta, ma
 ; CHECK-NEXT:    vmv.v.i v8, 0
 ; CHECK-NEXT:    vmerge.vim v8, v8, 1, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e8, mf2, ta, ma
@@ -207,9 +207,9 @@ define <vscale x 4 x i1> @test_vp_splice_nxv4i1_negative_offset(<vscale x 4 x i1
 define <vscale x 4 x i1> @test_vp_splice_nxv4i1_masked(<vscale x 4 x i1> %va, <vscale x 4 x i1> %vb, <vscale x 4 x i1> %mask, i32 zeroext %evla, i32 zeroext %evlb) {
 ; CHECK-LABEL: test_vp_splice_nxv4i1_masked:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a1, e8, mf2, ta, ma
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v8
-; CHECK-NEXT:    vsetvli zero, a1, e8, mf2, ta, ma
 ; CHECK-NEXT:    vmv.v.i v8, 0
 ; CHECK-NEXT:    vmerge.vim v8, v8, 1, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e8, mf2, ta, ma
@@ -232,9 +232,9 @@ define <vscale x 4 x i1> @test_vp_splice_nxv4i1_masked(<vscale x 4 x i1> %va, <v
 define <vscale x 8 x i1> @test_vp_splice_nxv8i1(<vscale x 8 x i1> %va, <vscale x 8 x i1> %vb, i32 zeroext %evla, i32 zeroext %evlb) {
 ; CHECK-LABEL: test_vp_splice_nxv8i1:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a1, e8, m1, ta, ma
 ; CHECK-NEXT:    vmv1r.v v9, v0
 ; CHECK-NEXT:    vmv1r.v v0, v8
-; CHECK-NEXT:    vsetvli zero, a1, e8, m1, ta, ma
 ; CHECK-NEXT:    vmv.v.i v8, 0
 ; CHECK-NEXT:    vmerge.vim v8, v8, 1, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e8, m1, ta, ma
@@ -256,9 +256,9 @@ define <vscale x 8 x i1> @test_vp_splice_nxv8i1(<vscale x 8 x i1> %va, <vscale x
 define <vscale x 8 x i1> @test_vp_splice_nxv8i1_negative_offset(<vscale x 8 x i1> %va, <vscale x 8 x i1> %vb, i32 zeroext %evla, i32 zeroext %evlb) {
 ; CHECK-LABEL: test_vp_splice_nxv8i1_negative_offset:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a1, e8, m1, ta, ma
 ; CHECK-NEXT:    vmv1r.v v9, v0
 ; CHECK-NEXT:    vmv1r.v v0, v8
-; CHECK-NEXT:    vsetvli zero, a1, e8, m1, ta, ma
 ; CHECK-NEXT:    vmv.v.i v8, 0
 ; CHECK-NEXT:    vmerge.vim v8, v8, 1, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e8, m1, ta, ma
@@ -280,9 +280,9 @@ define <vscale x 8 x i1> @test_vp_splice_nxv8i1_negative_offset(<vscale x 8 x i1
 define <vscale x 8 x i1> @test_vp_splice_nxv8i1_masked(<vscale x 8 x i1> %va, <vscale x 8 x i1> %vb, <vscale x 8 x i1> %mask, i32 zeroext %evla, i32 zeroext %evlb) {
 ; CHECK-LABEL: test_vp_splice_nxv8i1_masked:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a1, e8, m1, ta, ma
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v8
-; CHECK-NEXT:    vsetvli zero, a1, e8, m1, ta, ma
 ; CHECK-NEXT:    vmv.v.i v8, 0
 ; CHECK-NEXT:    vmerge.vim v8, v8, 1, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e8, m1, ta, ma
@@ -305,9 +305,9 @@ define <vscale x 8 x i1> @test_vp_splice_nxv8i1_masked(<vscale x 8 x i1> %va, <v
 define <vscale x 16 x i1> @test_vp_splice_nxv16i1(<vscale x 16 x i1> %va, <vscale x 16 x i1> %vb, i32 zeroext %evla, i32 zeroext %evlb) {
 ; CHECK-LABEL: test_vp_splice_nxv16i1:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a1, e8, m2, ta, ma
 ; CHECK-NEXT:    vmv1r.v v9, v0
 ; CHECK-NEXT:    vmv1r.v v0, v8
-; CHECK-NEXT:    vsetvli zero, a1, e8, m2, ta, ma
 ; CHECK-NEXT:    vmv.v.i v10, 0
 ; CHECK-NEXT:    vmerge.vim v10, v10, 1, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e8, m2, ta, ma
@@ -329,9 +329,9 @@ define <vscale x 16 x i1> @test_vp_splice_nxv16i1(<vscale x 16 x i1> %va, <vscal
 define <vscale x 16 x i1> @test_vp_splice_nxv16i1_negative_offset(<vscale x 16 x i1> %va, <vscale x 16 x i1> %vb, i32 zeroext %evla, i32 zeroext %evlb) {
 ; CHECK-LABEL: test_vp_splice_nxv16i1_negative_offset:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a1, e8, m2, ta, ma
 ; CHECK-NEXT:    vmv1r.v v9, v0
 ; CHECK-NEXT:    vmv1r.v v0, v8
-; CHECK-NEXT:    vsetvli zero, a1, e8, m2, ta, ma
 ; CHECK-NEXT:    vmv.v.i v10, 0
 ; CHECK-NEXT:    vmerge.vim v10, v10, 1, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e8, m2, ta, ma
@@ -353,9 +353,9 @@ define <vscale x 16 x i1> @test_vp_splice_nxv16i1_negative_offset(<vscale x 16 x
 define <vscale x 16 x i1> @test_vp_splice_nxv16i1_masked(<vscale x 16 x i1> %va, <vscale x 16 x i1> %vb, <vscale x 16 x i1> %mask, i32 zeroext %evla, i32 zeroext %evlb) {
 ; CHECK-LABEL: test_vp_splice_nxv16i1_masked:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a1, e8, m2, ta, ma
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v8
-; CHECK-NEXT:    vsetvli zero, a1, e8, m2, ta, ma
 ; CHECK-NEXT:    vmv.v.i v12, 0
 ; CHECK-NEXT:    vmerge.vim v12, v12, 1, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e8, m2, ta, ma
@@ -379,9 +379,9 @@ define <vscale x 16 x i1> @test_vp_splice_nxv16i1_masked(<vscale x 16 x i1> %va,
 define <vscale x 32 x i1> @test_vp_splice_nxv32i1(<vscale x 32 x i1> %va, <vscale x 32 x i1> %vb, i32 zeroext %evla, i32 zeroext %evlb) {
 ; CHECK-LABEL: test_vp_splice_nxv32i1:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a1, e8, m4, ta, ma
 ; CHECK-NEXT:    vmv1r.v v9, v0
 ; CHECK-NEXT:    vmv1r.v v0, v8
-; CHECK-NEXT:    vsetvli zero, a1, e8, m4, ta, ma
 ; CHECK-NEXT:    vmv.v.i v12, 0
 ; CHECK-NEXT:    vmerge.vim v12, v12, 1, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e8, m4, ta, ma
@@ -403,9 +403,9 @@ define <vscale x 32 x i1> @test_vp_splice_nxv32i1(<vscale x 32 x i1> %va, <vscal
 define <vscale x 32 x i1> @test_vp_splice_nxv32i1_negative_offset(<vscale x 32 x i1> %va, <vscale x 32 x i1> %vb, i32 zeroext %evla, i32 zeroext %evlb) {
 ; CHECK-LABEL: test_vp_splice_nxv32i1_negative_offset:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a1, e8, m4, ta, ma
 ; CHECK-NEXT:    vmv1r.v v9, v0
 ; CHECK-NEXT:    vmv1r.v v0, v8
-; CHECK-NEXT:    vsetvli zero, a1, e8, m4, ta, ma
 ; CHECK-NEXT:    vmv.v.i v12, 0
 ; CHECK-NEXT:    vmerge.vim v12, v12, 1, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e8, m4, ta, ma
@@ -427,9 +427,9 @@ define <vscale x 32 x i1> @test_vp_splice_nxv32i1_negative_offset(<vscale x 32 x
 define <vscale x 32 x i1> @test_vp_splice_nxv32i1_masked(<vscale x 32 x i1> %va, <vscale x 32 x i1> %vb, <vscale x 32 x i1> %mask, i32 zeroext %evla, i32 zeroext %evlb) {
 ; CHECK-LABEL: test_vp_splice_nxv32i1_masked:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a1, e8, m4, ta, ma
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v8
-; CHECK-NEXT:    vsetvli zero, a1, e8, m4, ta, ma
 ; CHECK-NEXT:    vmv.v.i v12, 0
 ; CHECK-NEXT:    vmerge.vim v12, v12, 1, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e8, m4, ta, ma
@@ -453,9 +453,9 @@ define <vscale x 32 x i1> @test_vp_splice_nxv32i1_masked(<vscale x 32 x i1> %va,
 define <vscale x 64 x i1> @test_vp_splice_nxv64i1(<vscale x 64 x i1> %va, <vscale x 64 x i1> %vb, i32 zeroext %evla, i32 zeroext %evlb) {
 ; CHECK-LABEL: test_vp_splice_nxv64i1:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a1, e8, m8, ta, ma
 ; CHECK-NEXT:    vmv1r.v v9, v0
 ; CHECK-NEXT:    vmv1r.v v0, v8
-; CHECK-NEXT:    vsetvli zero, a1, e8, m8, ta, ma
 ; CHECK-NEXT:    vmv.v.i v16, 0
 ; CHECK-NEXT:    vmerge.vim v16, v16, 1, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e8, m8, ta, ma
@@ -477,9 +477,9 @@ define <vscale x 64 x i1> @test_vp_splice_nxv64i1(<vscale x 64 x i1> %va, <vscal
 define <vscale x 64 x i1> @test_vp_splice_nxv64i1_negative_offset(<vscale x 64 x i1> %va, <vscale x 64 x i1> %vb, i32 zeroext %evla, i32 zeroext %evlb) {
 ; CHECK-LABEL: test_vp_splice_nxv64i1_negative_offset:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a1, e8, m8, ta, ma
 ; CHECK-NEXT:    vmv1r.v v9, v0
 ; CHECK-NEXT:    vmv1r.v v0, v8
-; CHECK-NEXT:    vsetvli zero, a1, e8, m8, ta, ma
 ; CHECK-NEXT:    vmv.v.i v16, 0
 ; CHECK-NEXT:    vmerge.vim v16, v16, 1, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e8, m8, ta, ma
@@ -501,9 +501,9 @@ define <vscale x 64 x i1> @test_vp_splice_nxv64i1_negative_offset(<vscale x 64 x
 define <vscale x 64 x i1> @test_vp_splice_nxv64i1_masked(<vscale x 64 x i1> %va, <vscale x 64 x i1> %vb, <vscale x 64 x i1> %mask, i32 zeroext %evla, i32 zeroext %evlb) {
 ; CHECK-LABEL: test_vp_splice_nxv64i1_masked:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a1, e8, m8, ta, ma
 ; CHECK-NEXT:    vmv1r.v v10, v0
 ; CHECK-NEXT:    vmv1r.v v0, v8
-; CHECK-NEXT:    vsetvli zero, a1, e8, m8, ta, ma
 ; CHECK-NEXT:    vmv.v.i v16, 0
 ; CHECK-NEXT:    vmerge.vim v16, v16, 1, v0
 ; CHECK-NEXT:    vsetvli zero, a0, e8, m8, ta, ma
diff --git a/llvm/test/CodeGen/RISCV/rvv/vpgather-sdnode.ll b/llvm/test/CodeGen/RISCV/rvv/vpgather-sdnode.ll
index 3e423c8ec9903..ca52ce6e2c4a1 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vpgather-sdnode.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vpgather-sdnode.ll
@@ -258,12 +258,12 @@ declare <vscale x 32 x i8> @llvm.vp.gather.nxv32i8.nxv32p0(<vscale x 32 x ptr>,
 define <vscale x 32 x i8> @vpgather_baseidx_nxv32i8(ptr %base, <vscale x 32 x i8> %idxs, <vscale x 32 x i1> %m, i32 zeroext %evl) {
 ; RV32-LABEL: vpgather_baseidx_nxv32i8:
 ; RV32:       # %bb.0:
+; RV32-NEXT:    vsetvli a2, zero, e8, mf2, ta, ma
 ; RV32-NEXT:    vmv1r.v v12, v0
 ; RV32-NEXT:    csrr a3, vlenb
 ; RV32-NEXT:    slli a2, a3, 1
 ; RV32-NEXT:    srli a3, a3, 2
 ; RV32-NEXT:    sub a4, a1, a2
-; RV32-NEXT:    vsetvli a5, zero, e8, mf2, ta, ma
 ; RV32-NEXT:    vslidedown.vx v0, v0, a3
 ; RV32-NEXT:    sltu a3, a1, a4
 ; RV32-NEXT:    addi a3, a3, -1
@@ -285,12 +285,12 @@ define <vscale x 32 x i8> @vpgather_baseidx_nxv32i8(ptr %base, <vscale x 32 x i8
 ;
 ; RV64-LABEL: vpgather_baseidx_nxv32i8:
 ; RV64:       # %bb.0:
+; RV64-NEXT:    vsetvli a2, zero, e8, mf2, ta, ma
 ; RV64-NEXT:    vmv1r.v v12, v0
 ; RV64-NEXT:    csrr a2, vlenb
 ; RV64-NEXT:    slli a3, a2, 1
 ; RV64-NEXT:    srli a4, a2, 2
 ; RV64-NEXT:    sub a5, a1, a3
-; RV64-NEXT:    vsetvli a6, zero, e8, mf2, ta, ma
 ; RV64-NEXT:    vslidedown.vx v13, v0, a4
 ; RV64-NEXT:    sltu a4, a1, a5
 ; RV64-NEXT:    addi a4, a4, -1
@@ -2457,11 +2457,11 @@ declare <vscale x 16 x double> @llvm.vp.gather.nxv16f64.nxv16p0(<vscale x 16 x p
 define <vscale x 16 x double> @vpgather_nxv16f64(<vscale x 16 x ptr> %ptrs, <vscale x 16 x i1> %m, i32 zeroext %evl) {
 ; RV32-LABEL: vpgather_nxv16f64:
 ; RV32:       # %bb.0:
+; RV32-NEXT:    vsetvli a1, zero, e8, mf4, ta, ma
 ; RV32-NEXT:    vmv1r.v v24, v0
 ; RV32-NEXT:    csrr a1, vlenb
 ; RV32-NEXT:    sub a2, a0, a1
 ; RV32-NEXT:    srli a3, a1, 3
-; RV32-NEXT:    vsetvli a4, zero, e8, mf4, ta, ma
 ; RV32-NEXT:    vslidedown.vx v0, v0, a3
 ; RV32-NEXT:    sltu a3, a0, a2
 ; RV32-NEXT:    addi a3, a3, -1
@@ -2480,11 +2480,11 @@ define <vscale x 16 x double> @vpgather_nxv16f64(<vscale x 16 x ptr> %ptrs, <vsc
 ;
 ; RV64-LABEL: vpgather_nxv16f64:
 ; RV64:       # %bb.0:
+; RV64-NEXT:    vsetvli a1, zero, e8, mf4, ta, ma
 ; RV64-NEXT:    vmv1r.v v24, v0
 ; RV64-NEXT:    csrr a1, vlenb
 ; RV64-NEXT:    sub a2, a0, a1
 ; RV64-NEXT:    srli a3, a1, 3
-; RV64-NEXT:    vsetvli a4, zero, e8, mf4, ta, ma
 ; RV64-NEXT:    vslidedown.vx v0, v0, a3
 ; RV64-NEXT:    sltu a3, a0, a2
 ; RV64-NEXT:    addi a3, a3, -1
@@ -2506,8 +2506,8 @@ define <vscale x 16 x double> @vpgather_nxv16f64(<vscale x 16 x ptr> %ptrs, <vsc
 define <vscale x 16 x double> @vpgather_baseidx_nxv16i16_nxv16f64(ptr %base, <vscale x 16 x i16> %idxs, <vscale x 16 x i1> %m, i32 zeroext %evl) {
 ; RV32-LABEL: vpgather_baseidx_nxv16i16_nxv16f64:
 ; RV32:       # %bb.0:
-; RV32-NEXT:    vmv1r.v v12, v0
 ; RV32-NEXT:    vsetvli a2, zero, e32, m8, ta, ma
+; RV32-NEXT:    vmv1r.v v12, v0
 ; RV32-NEXT:    vsext.vf2 v16, v8
 ; RV32-NEXT:    csrr a2, vlenb
 ; RV32-NEXT:    vsll.vi v24, v16, 3
@@ -2531,8 +2531,8 @@ define <vscale x 16 x double> @vpgather_baseidx_nxv16i16_nxv16f64(ptr %base, <vs
 ;
 ; RV64-LABEL: vpgather_baseidx_nxv16i16_nxv16f64:
 ; RV64:       # %bb.0:
-; RV64-NEXT:    vmv1r.v v12, v0
 ; RV64-NEXT:    vsetvli a2, zero, e64, m8, ta, ma
+; RV64-NEXT:    vmv1r.v v12, v0
 ; RV64-NEXT:    vsext.vf4 v16, v10
 ; RV64-NEXT:    csrr a2, vlenb
 ; RV64-NEXT:    vsll.vi v16, v16, 3
@@ -2564,8 +2564,8 @@ define <vscale x 16 x double> @vpgather_baseidx_nxv16i16_nxv16f64(ptr %base, <vs
 define <vscale x 16 x double> @vpgather_baseidx_sext_nxv16i16_nxv16f64(ptr %base, <vscale x 16 x i16> %idxs, <vscale x 16 x i1> %m, i32 zeroext %evl) {
 ; RV32-LABEL: vpgather_baseidx_sext_nxv16i16_nxv16f64:
 ; RV32:       # %bb.0:
-; RV32-NEXT:    vmv1r.v v12, v0
 ; RV32-NEXT:    vsetvli a2, zero, e32, m8, ta, ma
+; RV32-NEXT:    vmv1r.v v12, v0
 ; RV32-NEXT:    vsext.vf2 v16, v8
 ; RV32-NEXT:    csrr a2, vlenb
 ; RV32-NEXT:    vsll.vi v24, v16, 3
@@ -2589,8 +2589,8 @@ define <vscale x 16 x double> @vpgather_baseidx_sext_nxv16i16_nxv16f64(ptr %base
 ;
 ; RV64-LABEL: vpgather_baseidx_sext_nxv16i16_nxv16f64:
 ; RV64:       # %bb.0:
-; RV64-NEXT:    vmv1r.v v12, v0
 ; RV64-NEXT:    vsetvli a2, zero, e64, m8, ta, ma
+; RV64-NEXT:    vmv1r.v v12, v0
 ; RV64-NEXT:    vsext.vf4 v16, v10
 ; RV64-NEXT:    csrr a2, vlenb
 ; RV64-NEXT:    vsll.vi v16, v16, 3
@@ -2623,8 +2623,8 @@ define <vscale x 16 x double> @vpgather_baseidx_sext_nxv16i16_nxv16f64(ptr %base
 define <vscale x 16 x double> @vpgather_baseidx_zext_nxv16i16_nxv16f64(ptr %base, <vscale x 16 x i16> %idxs, <vscale x 16 x i1> %m, i32 zeroext %evl) {
 ; RV32-LABEL: vpgather_baseidx_zext_nxv16i16_nxv16f64:
 ; RV32:       # %bb.0:
-; RV32-NEXT:    vmv1r.v v12, v0
 ; RV32-NEXT:    vsetvli a2, zero, e32, m8, ta, ma
+; RV32-NEXT:    vmv1r.v v12, v0
 ; RV32-NEXT:    vzext.vf2 v16, v8
 ; RV32-NEXT:    csrr a2, vlenb
 ; RV32-NEXT:    vsll.vi v24, v16, 3
@@ -2648,8 +2648,8 @@ define <vscale x 16 x double> @vpgather_baseidx_zext_nxv16i16_nxv16f64(ptr %base
 ;
 ; RV64-LABEL: vpgather_baseidx_zext_nxv16i16_nxv16f64:
 ; RV64:       # %bb.0:
-; RV64-NEXT:    vmv1r.v v12, v0
 ; RV64-NEXT:    vsetvli a2, zero, e32, m8, ta, ma
+; RV64-NEXT:    vmv1r.v v12, v0
 ; RV64-NEXT:    vzext.vf2 v16, v8
 ; RV64-NEXT:    csrr a2, vlenb
 ; RV64-NEXT:    vsll.vi v24, v16, 3
diff --git a/llvm/test/CodeGen/RISCV/rvv/vpload.ll b/llvm/test/CodeGen/RISCV/rvv/vpload.ll
index bd7ea6c19d0b3..0844180e49612 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vpload.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vpload.ll
@@ -522,12 +522,12 @@ declare <vscale x 16 x double> @llvm.vp.load.nxv16f64.p0(ptr, <vscale x 16 x i1>
 define <vscale x 16 x double> @vpload_nxv16f64(ptr %ptr, <vscale x 16 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vpload_nxv16f64:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli a2, zero, e8, mf4, ta, ma
 ; CHECK-NEXT:    vmv1r.v v8, v0
 ; CHECK-NEXT:    csrr a2, vlenb
 ; CHECK-NEXT:    sub a3, a1, a2
 ; CHECK-NEXT:    slli a4, a2, 3
 ; CHECK-NEXT:    srli a5, a2, 3
-; CHECK-NEXT:    vsetvli a6, zero, e8, mf4, ta, ma
 ; CHECK-NEXT:    vslidedown.vx v0, v0, a5
 ; CHECK-NEXT:    sltu a5, a1, a3
 ; CHECK-NEXT:    addi a5, a5, -1
@@ -561,6 +561,7 @@ declare <vscale x 16 x double> @llvm.vector.extract.nxv16f64(<vscale x 17 x doub
 define <vscale x 16 x double> @vpload_nxv17f64(ptr %ptr, ptr %out, <vscale x 17 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vpload_nxv17f64:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-NEXT:    vmv1r.v v8, v0
 ; CHECK-NEXT:    csrr a3, vlenb
 ; CHECK-NEXT:    slli a5, a3, 1
diff --git a/llvm/test/CodeGen/RISCV/rvv/vpmerge-sdnode.ll b/llvm/test/CodeGen/RISCV/rvv/vpmerge-sdnode.ll
index f029d0b1b01bc..88a8ebcc90054 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vpmerge-sdnode.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vpmerge-sdnode.ll
@@ -361,12 +361,12 @@ define <vscale x 128 x i8> @vpmerge_vv_nxv128i8(<vscale x 128 x i8> %va, <vscale
 ; CHECK-NEXT:    slli a1, a1, 3
 ; CHECK-NEXT:    sub sp, sp, a1
 ; CHECK-NEXT:    .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x08, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 8 * vlenb
+; CHECK-NEXT:    vsetvli a1, zero, e8, m8, ta, ma
 ; CHECK-NEXT:    vmv1r.v v7, v0
 ; CHECK-NEXT:    vmv8r.v v24, v16
 ; CHECK-NEXT:    addi a1, sp, 16
 ; CHECK-NEXT:    vs8r.v v8, (a1) # Unknown-size Folded Spill
 ; CHECK-NEXT:    csrr a1, vlenb
-; CHECK-NEXT:    vsetvli a4, zero, e8, m8, ta, ma
 ; CHECK-NEXT:    vlm.v v0, (a2)
 ; CHECK-NEXT:    slli a1, a1, 3
 ; CHECK-NEXT:    add a2, a0, a1
@@ -401,8 +401,8 @@ define <vscale x 128 x i8> @vpmerge_vv_nxv128i8(<vscale x 128 x i8> %va, <vscale
 define <vscale x 128 x i8> @vpmerge_vx_nxv128i8(i8 %a, <vscale x 128 x i8> %vb, <vscale x 128 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vpmerge_vx_nxv128i8:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v24, v0
 ; CHECK-NEXT:    vsetvli a3, zero, e8, m8, ta, ma
+; CHECK-NEXT:    vmv1r.v v24, v0
 ; CHECK-NEXT:    vlm.v v0, (a1)
 ; CHECK-NEXT:    csrr a1, vlenb
 ; CHECK-NEXT:    slli a1, a1, 3
@@ -429,8 +429,8 @@ define <vscale x 128 x i8> @vpmerge_vx_nxv128i8(i8 %a, <vscale x 128 x i8> %vb,
 define <vscale x 128 x i8> @vpmerge_vi_nxv128i8(<vscale x 128 x i8> %vb, <vscale x 128 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vpmerge_vi_nxv128i8:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v24, v0
 ; CHECK-NEXT:    vsetvli a2, zero, e8, m8, ta, ma
+; CHECK-NEXT:    vmv1r.v v24, v0
 ; CHECK-NEXT:    vlm.v v0, (a0)
 ; CHECK-NEXT:    csrr a0, vlenb
 ; CHECK-NEXT:    slli a0, a0, 3
diff --git a/llvm/test/CodeGen/RISCV/rvv/vpstore.ll b/llvm/test/CodeGen/RISCV/rvv/vpstore.ll
index 8978dc268d4e5..7e7da529bf3d7 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vpstore.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vpstore.ll
@@ -468,6 +468,7 @@ define void @vpstore_nxv17f64(<vscale x 17 x double> %val, ptr %ptr, <vscale x 1
 ; CHECK-NEXT:    slli a3, a3, 3
 ; CHECK-NEXT:    sub sp, sp, a3
 ; CHECK-NEXT:    .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x08, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 8 * vlenb
+; CHECK-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-NEXT:    vmv1r.v v24, v0
 ; CHECK-NEXT:    addi a3, sp, 16
 ; CHECK-NEXT:    vs8r.v v16, (a3) # Unknown-size Folded Spill
diff --git a/llvm/test/CodeGen/RISCV/rvv/vreductions-mask-vp.ll b/llvm/test/CodeGen/RISCV/rvv/vreductions-mask-vp.ll
index 39666bb6119a0..18d20f66987b2 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vreductions-mask-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vreductions-mask-vp.ll
@@ -23,9 +23,9 @@ declare i1 @llvm.vp.reduce.or.nxv1i1(i1, <vscale x 1 x i1>, <vscale x 1 x i1>, i
 define zeroext i1 @vpreduce_or_nxv1i1(i1 zeroext %s, <vscale x 1 x i1> %v, <vscale x 1 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vpreduce_or_nxv1i1:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a1, e8, mf8, ta, ma
 ; CHECK-NEXT:    vmv1r.v v9, v0
 ; CHECK-NEXT:    vmv1r.v v0, v8
-; CHECK-NEXT:    vsetvli zero, a1, e8, mf8, ta, ma
 ; CHECK-NEXT:    vcpop.m a1, v9, v0.t
 ; CHECK-NEXT:    snez a1, a1
 ; CHECK-NEXT:    or a0, a1, a0
@@ -39,9 +39,9 @@ declare i1 @llvm.vp.reduce.xor.nxv1i1(i1, <vscale x 1 x i1>, <vscale x 1 x i1>,
 define zeroext i1 @vpreduce_xor_nxv1i1(i1 zeroext %s, <vscale x 1 x i1> %v, <vscale x 1 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vpreduce_xor_nxv1i1:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a1, e8, mf8, ta, ma
 ; CHECK-NEXT:    vmv1r.v v9, v0
 ; CHECK-NEXT:    vmv1r.v v0, v8
-; CHECK-NEXT:    vsetvli zero, a1, e8, mf8, ta, ma
 ; CHECK-NEXT:    vcpop.m a1, v9, v0.t
 ; CHECK-NEXT:    andi a1, a1, 1
 ; CHECK-NEXT:    xor a0, a1, a0
@@ -71,9 +71,9 @@ declare i1 @llvm.vp.reduce.or.nxv2i1(i1, <vscale x 2 x i1>, <vscale x 2 x i1>, i
 define zeroext i1 @vpreduce_or_nxv2i1(i1 zeroext %s, <vscale x 2 x i1> %v, <vscale x 2 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vpreduce_or_nxv2i1:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a1, e8, mf4, ta, ma
 ; CHECK-NEXT:    vmv1r.v v9, v0
 ; CHECK-NEXT:    vmv1r.v v0, v8
-; CHECK-NEXT:    vsetvli zero, a1, e8, mf4, ta, ma
 ; CHECK-NEXT:    vcpop.m a1, v9, v0.t
 ; CHECK-NEXT:    snez a1, a1
 ; CHECK-NEXT:    or a0, a1, a0
@@ -87,9 +87,9 @@ declare i1 @llvm.vp.reduce.xor.nxv2i1(i1, <vscale x 2 x i1>, <vscale x 2 x i1>,
 define zeroext i1 @vpreduce_xor_nxv2i1(i1 zeroext %s, <vscale x 2 x i1> %v, <vscale x 2 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vpreduce_xor_nxv2i1:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a1, e8, mf4, ta, ma
 ; CHECK-NEXT:    vmv1r.v v9, v0
 ; CHECK-NEXT:    vmv1r.v v0, v8
-; CHECK-NEXT:    vsetvli zero, a1, e8, mf4, ta, ma
 ; CHECK-NEXT:    vcpop.m a1, v9, v0.t
 ; CHECK-NEXT:    andi a1, a1, 1
 ; CHECK-NEXT:    xor a0, a1, a0
@@ -119,9 +119,9 @@ declare i1 @llvm.vp.reduce.or.nxv4i1(i1, <vscale x 4 x i1>, <vscale x 4 x i1>, i
 define zeroext i1 @vpreduce_or_nxv4i1(i1 zeroext %s, <vscale x 4 x i1> %v, <vscale x 4 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vpreduce_or_nxv4i1:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a1, e8, mf2, ta, ma
 ; CHECK-NEXT:    vmv1r.v v9, v0
 ; CHECK-NEXT:    vmv1r.v v0, v8
-; CHECK-NEXT:    vsetvli zero, a1, e8, mf2, ta, ma
 ; CHECK-NEXT:    vcpop.m a1, v9, v0.t
 ; CHECK-NEXT:    snez a1, a1
 ; CHECK-NEXT:    or a0, a1, a0
@@ -135,9 +135,9 @@ declare i1 @llvm.vp.reduce.xor.nxv4i1(i1, <vscale x 4 x i1>, <vscale x 4 x i1>,
 define zeroext i1 @vpreduce_xor_nxv4i1(i1 zeroext %s, <vscale x 4 x i1> %v, <vscale x 4 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vpreduce_xor_nxv4i1:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a1, e8, mf2, ta, ma
 ; CHECK-NEXT:    vmv1r.v v9, v0
 ; CHECK-NEXT:    vmv1r.v v0, v8
-; CHECK-NEXT:    vsetvli zero, a1, e8, mf2, ta, ma
 ; CHECK-NEXT:    vcpop.m a1, v9, v0.t
 ; CHECK-NEXT:    andi a1, a1, 1
 ; CHECK-NEXT:    xor a0, a1, a0
@@ -167,9 +167,9 @@ declare i1 @llvm.vp.reduce.or.nxv8i1(i1, <vscale x 8 x i1>, <vscale x 8 x i1>, i
 define zeroext i1 @vpreduce_or_nxv8i1(i1 zeroext %s, <vscale x 8 x i1> %v, <vscale x 8 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vpreduce_or_nxv8i1:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a1, e8, m1, ta, ma
 ; CHECK-NEXT:    vmv1r.v v9, v0
 ; CHECK-NEXT:    vmv1r.v v0, v8
-; CHECK-NEXT:    vsetvli zero, a1, e8, m1, ta, ma
 ; CHECK-NEXT:    vcpop.m a1, v9, v0.t
 ; CHECK-NEXT:    snez a1, a1
 ; CHECK-NEXT:    or a0, a1, a0
@@ -183,9 +183,9 @@ declare i1 @llvm.vp.reduce.xor.nxv8i1(i1, <vscale x 8 x i1>, <vscale x 8 x i1>,
 define zeroext i1 @vpreduce_xor_nxv8i1(i1 zeroext %s, <vscale x 8 x i1> %v, <vscale x 8 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vpreduce_xor_nxv8i1:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a1, e8, m1, ta, ma
 ; CHECK-NEXT:    vmv1r.v v9, v0
 ; CHECK-NEXT:    vmv1r.v v0, v8
-; CHECK-NEXT:    vsetvli zero, a1, e8, m1, ta, ma
 ; CHECK-NEXT:    vcpop.m a1, v9, v0.t
 ; CHECK-NEXT:    andi a1, a1, 1
 ; CHECK-NEXT:    xor a0, a1, a0
@@ -215,9 +215,9 @@ declare i1 @llvm.vp.reduce.or.nxv16i1(i1, <vscale x 16 x i1>, <vscale x 16 x i1>
 define zeroext i1 @vpreduce_or_nxv16i1(i1 zeroext %s, <vscale x 16 x i1> %v, <vscale x 16 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vpreduce_or_nxv16i1:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a1, e8, m2, ta, ma
 ; CHECK-NEXT:    vmv1r.v v9, v0
 ; CHECK-NEXT:    vmv1r.v v0, v8
-; CHECK-NEXT:    vsetvli zero, a1, e8, m2, ta, ma
 ; CHECK-NEXT:    vcpop.m a1, v9, v0.t
 ; CHECK-NEXT:    snez a1, a1
 ; CHECK-NEXT:    or a0, a1, a0
@@ -231,9 +231,9 @@ declare i1 @llvm.vp.reduce.xor.nxv16i1(i1, <vscale x 16 x i1>, <vscale x 16 x i1
 define zeroext i1 @vpreduce_xor_nxv16i1(i1 zeroext %s, <vscale x 16 x i1> %v, <vscale x 16 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vpreduce_xor_nxv16i1:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a1, e8, m2, ta, ma
 ; CHECK-NEXT:    vmv1r.v v9, v0
 ; CHECK-NEXT:    vmv1r.v v0, v8
-; CHECK-NEXT:    vsetvli zero, a1, e8, m2, ta, ma
 ; CHECK-NEXT:    vcpop.m a1, v9, v0.t
 ; CHECK-NEXT:    andi a1, a1, 1
 ; CHECK-NEXT:    xor a0, a1, a0
@@ -263,9 +263,9 @@ declare i1 @llvm.vp.reduce.or.nxv32i1(i1, <vscale x 32 x i1>, <vscale x 32 x i1>
 define zeroext i1 @vpreduce_or_nxv32i1(i1 zeroext %s, <vscale x 32 x i1> %v, <vscale x 32 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vpreduce_or_nxv32i1:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a1, e8, m4, ta, ma
 ; CHECK-NEXT:    vmv1r.v v9, v0
 ; CHECK-NEXT:    vmv1r.v v0, v8
-; CHECK-NEXT:    vsetvli zero, a1, e8, m4, ta, ma
 ; CHECK-NEXT:    vcpop.m a1, v9, v0.t
 ; CHECK-NEXT:    snez a1, a1
 ; CHECK-NEXT:    or a0, a1, a0
@@ -279,9 +279,9 @@ declare i1 @llvm.vp.reduce.xor.nxv32i1(i1, <vscale x 32 x i1>, <vscale x 32 x i1
 define zeroext i1 @vpreduce_xor_nxv32i1(i1 zeroext %s, <vscale x 32 x i1> %v, <vscale x 32 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vpreduce_xor_nxv32i1:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a1, e8, m4, ta, ma
 ; CHECK-NEXT:    vmv1r.v v9, v0
 ; CHECK-NEXT:    vmv1r.v v0, v8
-; CHECK-NEXT:    vsetvli zero, a1, e8, m4, ta, ma
 ; CHECK-NEXT:    vcpop.m a1, v9, v0.t
 ; CHECK-NEXT:    andi a1, a1, 1
 ; CHECK-NEXT:    xor a0, a1, a0
@@ -295,9 +295,9 @@ declare i1 @llvm.vp.reduce.or.nxv40i1(i1, <vscale x 40 x i1>, <vscale x 40 x i1>
 define zeroext i1 @vpreduce_or_nxv40i1(i1 zeroext %s, <vscale x 40 x i1> %v, <vscale x 40 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vpreduce_or_nxv40i1:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a1, e8, m8, ta, ma
 ; CHECK-NEXT:    vmv1r.v v9, v0
 ; CHECK-NEXT:    vmv1r.v v0, v8
-; CHECK-NEXT:    vsetvli zero, a1, e8, m8, ta, ma
 ; CHECK-NEXT:    vcpop.m a1, v9, v0.t
 ; CHECK-NEXT:    snez a1, a1
 ; CHECK-NEXT:    or a0, a1, a0
@@ -327,9 +327,9 @@ declare i1 @llvm.vp.reduce.or.nxv64i1(i1, <vscale x 64 x i1>, <vscale x 64 x i1>
 define zeroext i1 @vpreduce_or_nxv64i1(i1 zeroext %s, <vscale x 64 x i1> %v, <vscale x 64 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vpreduce_or_nxv64i1:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a1, e8, m8, ta, ma
 ; CHECK-NEXT:    vmv1r.v v9, v0
 ; CHECK-NEXT:    vmv1r.v v0, v8
-; CHECK-NEXT:    vsetvli zero, a1, e8, m8, ta, ma
 ; CHECK-NEXT:    vcpop.m a1, v9, v0.t
 ; CHECK-NEXT:    snez a1, a1
 ; CHECK-NEXT:    or a0, a1, a0
@@ -343,9 +343,9 @@ declare i1 @llvm.vp.reduce.xor.nxv64i1(i1, <vscale x 64 x i1>, <vscale x 64 x i1
 define zeroext i1 @vpreduce_xor_nxv64i1(i1 zeroext %s, <vscale x 64 x i1> %v, <vscale x 64 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vpreduce_xor_nxv64i1:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a1, e8, m8, ta, ma
 ; CHECK-NEXT:    vmv1r.v v9, v0
 ; CHECK-NEXT:    vmv1r.v v0, v8
-; CHECK-NEXT:    vsetvli zero, a1, e8, m8, ta, ma
 ; CHECK-NEXT:    vcpop.m a1, v9, v0.t
 ; CHECK-NEXT:    andi a1, a1, 1
 ; CHECK-NEXT:    xor a0, a1, a0
@@ -359,6 +359,7 @@ declare i1 @llvm.vp.reduce.or.nxv128i1(i1, <vscale x 128 x i1>, <vscale x 128 x
 define zeroext i1 @vpreduce_or_nxv128i1(i1 zeroext %s, <vscale x 128 x i1> %v, <vscale x 128 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vpreduce_or_nxv128i1:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-NEXT:    vmv1r.v v11, v0
 ; CHECK-NEXT:    csrr a2, vlenb
 ; CHECK-NEXT:    slli a2, a2, 3
@@ -390,9 +391,9 @@ declare i1 @llvm.vp.reduce.add.nxv1i1(i1, <vscale x 1 x i1>, <vscale x 1 x i1>,
 define zeroext i1 @vpreduce_add_nxv1i1(i1 zeroext %s, <vscale x 1 x i1> %v, <vscale x 1 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vpreduce_add_nxv1i1:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a1, e8, mf8, ta, ma
 ; CHECK-NEXT:    vmv1r.v v9, v0
 ; CHECK-NEXT:    vmv1r.v v0, v8
-; CHECK-NEXT:    vsetvli zero, a1, e8, mf8, ta, ma
 ; CHECK-NEXT:    vcpop.m a1, v9, v0.t
 ; CHECK-NEXT:    andi a1, a1, 1
 ; CHECK-NEXT:    xor a0, a1, a0
@@ -406,9 +407,9 @@ declare i1 @llvm.vp.reduce.add.nxv2i1(i1, <vscale x 2 x i1>, <vscale x 2 x i1>,
 define zeroext i1 @vpreduce_add_nxv2i1(i1 zeroext %s, <vscale x 2 x i1> %v, <vscale x 2 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vpreduce_add_nxv2i1:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a1, e8, mf4, ta, ma
 ; CHECK-NEXT:    vmv1r.v v9, v0
 ; CHECK-NEXT:    vmv1r.v v0, v8
-; CHECK-NEXT:    vsetvli zero, a1, e8, mf4, ta, ma
 ; CHECK-NEXT:    vcpop.m a1, v9, v0.t
 ; CHECK-NEXT:    andi a1, a1, 1
 ; CHECK-NEXT:    xor a0, a1, a0
@@ -422,9 +423,9 @@ declare i1 @llvm.vp.reduce.add.nxv4i1(i1, <vscale x 4 x i1>, <vscale x 4 x i1>,
 define zeroext i1 @vpreduce_add_nxv4i1(i1 zeroext %s, <vscale x 4 x i1> %v, <vscale x 4 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vpreduce_add_nxv4i1:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a1, e8, mf2, ta, ma
 ; CHECK-NEXT:    vmv1r.v v9, v0
 ; CHECK-NEXT:    vmv1r.v v0, v8
-; CHECK-NEXT:    vsetvli zero, a1, e8, mf2, ta, ma
 ; CHECK-NEXT:    vcpop.m a1, v9, v0.t
 ; CHECK-NEXT:    andi a1, a1, 1
 ; CHECK-NEXT:    xor a0, a1, a0
@@ -438,9 +439,9 @@ declare i1 @llvm.vp.reduce.add.nxv8i1(i1, <vscale x 8 x i1>, <vscale x 8 x i1>,
 define zeroext i1 @vpreduce_add_nxv8i1(i1 zeroext %s, <vscale x 8 x i1> %v, <vscale x 8 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vpreduce_add_nxv8i1:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a1, e8, m1, ta, ma
 ; CHECK-NEXT:    vmv1r.v v9, v0
 ; CHECK-NEXT:    vmv1r.v v0, v8
-; CHECK-NEXT:    vsetvli zero, a1, e8, m1, ta, ma
 ; CHECK-NEXT:    vcpop.m a1, v9, v0.t
 ; CHECK-NEXT:    andi a1, a1, 1
 ; CHECK-NEXT:    xor a0, a1, a0
@@ -454,9 +455,9 @@ declare i1 @llvm.vp.reduce.add.nxv16i1(i1, <vscale x 16 x i1>, <vscale x 16 x i1
 define zeroext i1 @vpreduce_add_nxv16i1(i1 zeroext %s, <vscale x 16 x i1> %v, <vscale x 16 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vpreduce_add_nxv16i1:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a1, e8, m2, ta, ma
 ; CHECK-NEXT:    vmv1r.v v9, v0
 ; CHECK-NEXT:    vmv1r.v v0, v8
-; CHECK-NEXT:    vsetvli zero, a1, e8, m2, ta, ma
 ; CHECK-NEXT:    vcpop.m a1, v9, v0.t
 ; CHECK-NEXT:    andi a1, a1, 1
 ; CHECK-NEXT:    xor a0, a1, a0
@@ -470,9 +471,9 @@ declare i1 @llvm.vp.reduce.add.nxv32i1(i1, <vscale x 32 x i1>, <vscale x 32 x i1
 define zeroext i1 @vpreduce_add_nxv32i1(i1 zeroext %s, <vscale x 32 x i1> %v, <vscale x 32 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vpreduce_add_nxv32i1:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a1, e8, m4, ta, ma
 ; CHECK-NEXT:    vmv1r.v v9, v0
 ; CHECK-NEXT:    vmv1r.v v0, v8
-; CHECK-NEXT:    vsetvli zero, a1, e8, m4, ta, ma
 ; CHECK-NEXT:    vcpop.m a1, v9, v0.t
 ; CHECK-NEXT:    andi a1, a1, 1
 ; CHECK-NEXT:    xor a0, a1, a0
@@ -486,9 +487,9 @@ declare i1 @llvm.vp.reduce.add.nxv64i1(i1, <vscale x 64 x i1>, <vscale x 64 x i1
 define zeroext i1 @vpreduce_add_nxv64i1(i1 zeroext %s, <vscale x 64 x i1> %v, <vscale x 64 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vpreduce_add_nxv64i1:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a1, e8, m8, ta, ma
 ; CHECK-NEXT:    vmv1r.v v9, v0
 ; CHECK-NEXT:    vmv1r.v v0, v8
-; CHECK-NEXT:    vsetvli zero, a1, e8, m8, ta, ma
 ; CHECK-NEXT:    vcpop.m a1, v9, v0.t
 ; CHECK-NEXT:    andi a1, a1, 1
 ; CHECK-NEXT:    xor a0, a1, a0
@@ -615,9 +616,9 @@ declare i1 @llvm.vp.reduce.smin.nxv1i1(i1, <vscale x 1 x i1>, <vscale x 1 x i1>,
 define zeroext i1 @vpreduce_smin_nxv1i1(i1 zeroext %s, <vscale x 1 x i1> %v, <vscale x 1 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vpreduce_smin_nxv1i1:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a1, e8, mf8, ta, ma
 ; CHECK-NEXT:    vmv1r.v v9, v0
 ; CHECK-NEXT:    vmv1r.v v0, v8
-; CHECK-NEXT:    vsetvli zero, a1, e8, mf8, ta, ma
 ; CHECK-NEXT:    vcpop.m a1, v9, v0.t
 ; CHECK-NEXT:    snez a1, a1
 ; CHECK-NEXT:    or a0, a1, a0
@@ -631,9 +632,9 @@ declare i1 @llvm.vp.reduce.smin.nxv2i1(i1, <vscale x 2 x i1>, <vscale x 2 x i1>,
 define zeroext i1 @vpreduce_smin_nxv2i1(i1 zeroext %s, <vscale x 2 x i1> %v, <vscale x 2 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vpreduce_smin_nxv2i1:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a1, e8, mf4, ta, ma
 ; CHECK-NEXT:    vmv1r.v v9, v0
 ; CHECK-NEXT:    vmv1r.v v0, v8
-; CHECK-NEXT:    vsetvli zero, a1, e8, mf4, ta, ma
 ; CHECK-NEXT:    vcpop.m a1, v9, v0.t
 ; CHECK-NEXT:    snez a1, a1
 ; CHECK-NEXT:    or a0, a1, a0
@@ -647,9 +648,9 @@ declare i1 @llvm.vp.reduce.smin.nxv4i1(i1, <vscale x 4 x i1>, <vscale x 4 x i1>,
 define zeroext i1 @vpreduce_smin_nxv4i1(i1 zeroext %s, <vscale x 4 x i1> %v, <vscale x 4 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vpreduce_smin_nxv4i1:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a1, e8, mf2, ta, ma
 ; CHECK-NEXT:    vmv1r.v v9, v0
 ; CHECK-NEXT:    vmv1r.v v0, v8
-; CHECK-NEXT:    vsetvli zero, a1, e8, mf2, ta, ma
 ; CHECK-NEXT:    vcpop.m a1, v9, v0.t
 ; CHECK-NEXT:    snez a1, a1
 ; CHECK-NEXT:    or a0, a1, a0
@@ -663,9 +664,9 @@ declare i1 @llvm.vp.reduce.smin.nxv8i1(i1, <vscale x 8 x i1>, <vscale x 8 x i1>,
 define zeroext i1 @vpreduce_smin_nxv8i1(i1 zeroext %s, <vscale x 8 x i1> %v, <vscale x 8 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vpreduce_smin_nxv8i1:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a1, e8, m1, ta, ma
 ; CHECK-NEXT:    vmv1r.v v9, v0
 ; CHECK-NEXT:    vmv1r.v v0, v8
-; CHECK-NEXT:    vsetvli zero, a1, e8, m1, ta, ma
 ; CHECK-NEXT:    vcpop.m a1, v9, v0.t
 ; CHECK-NEXT:    snez a1, a1
 ; CHECK-NEXT:    or a0, a1, a0
@@ -679,9 +680,9 @@ declare i1 @llvm.vp.reduce.smin.nxv16i1(i1, <vscale x 16 x i1>, <vscale x 16 x i
 define zeroext i1 @vpreduce_smin_nxv16i1(i1 zeroext %s, <vscale x 16 x i1> %v, <vscale x 16 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vpreduce_smin_nxv16i1:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a1, e8, m2, ta, ma
 ; CHECK-NEXT:    vmv1r.v v9, v0
 ; CHECK-NEXT:    vmv1r.v v0, v8
-; CHECK-NEXT:    vsetvli zero, a1, e8, m2, ta, ma
 ; CHECK-NEXT:    vcpop.m a1, v9, v0.t
 ; CHECK-NEXT:    snez a1, a1
 ; CHECK-NEXT:    or a0, a1, a0
@@ -695,9 +696,9 @@ declare i1 @llvm.vp.reduce.smin.nxv32i1(i1, <vscale x 32 x i1>, <vscale x 32 x i
 define zeroext i1 @vpreduce_smin_nxv32i1(i1 zeroext %s, <vscale x 32 x i1> %v, <vscale x 32 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vpreduce_smin_nxv32i1:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a1, e8, m4, ta, ma
 ; CHECK-NEXT:    vmv1r.v v9, v0
 ; CHECK-NEXT:    vmv1r.v v0, v8
-; CHECK-NEXT:    vsetvli zero, a1, e8, m4, ta, ma
 ; CHECK-NEXT:    vcpop.m a1, v9, v0.t
 ; CHECK-NEXT:    snez a1, a1
 ; CHECK-NEXT:    or a0, a1, a0
@@ -711,9 +712,9 @@ declare i1 @llvm.vp.reduce.smin.nxv64i1(i1, <vscale x 64 x i1>, <vscale x 64 x i
 define zeroext i1 @vpreduce_smin_nxv64i1(i1 zeroext %s, <vscale x 64 x i1> %v, <vscale x 64 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vpreduce_smin_nxv64i1:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a1, e8, m8, ta, ma
 ; CHECK-NEXT:    vmv1r.v v9, v0
 ; CHECK-NEXT:    vmv1r.v v0, v8
-; CHECK-NEXT:    vsetvli zero, a1, e8, m8, ta, ma
 ; CHECK-NEXT:    vcpop.m a1, v9, v0.t
 ; CHECK-NEXT:    snez a1, a1
 ; CHECK-NEXT:    or a0, a1, a0
@@ -727,9 +728,9 @@ declare i1 @llvm.vp.reduce.umax.nxv1i1(i1, <vscale x 1 x i1>, <vscale x 1 x i1>,
 define zeroext i1 @vpreduce_umax_nxv1i1(i1 zeroext %s, <vscale x 1 x i1> %v, <vscale x 1 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vpreduce_umax_nxv1i1:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a1, e8, mf8, ta, ma
 ; CHECK-NEXT:    vmv1r.v v9, v0
 ; CHECK-NEXT:    vmv1r.v v0, v8
-; CHECK-NEXT:    vsetvli zero, a1, e8, mf8, ta, ma
 ; CHECK-NEXT:    vcpop.m a1, v9, v0.t
 ; CHECK-NEXT:    snez a1, a1
 ; CHECK-NEXT:    or a0, a1, a0
@@ -743,9 +744,9 @@ declare i1 @llvm.vp.reduce.umax.nxv2i1(i1, <vscale x 2 x i1>, <vscale x 2 x i1>,
 define zeroext i1 @vpreduce_umax_nxv2i1(i1 zeroext %s, <vscale x 2 x i1> %v, <vscale x 2 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vpreduce_umax_nxv2i1:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a1, e8, mf4, ta, ma
 ; CHECK-NEXT:    vmv1r.v v9, v0
 ; CHECK-NEXT:    vmv1r.v v0, v8
-; CHECK-NEXT:    vsetvli zero, a1, e8, mf4, ta, ma
 ; CHECK-NEXT:    vcpop.m a1, v9, v0.t
 ; CHECK-NEXT:    snez a1, a1
 ; CHECK-NEXT:    or a0, a1, a0
@@ -759,9 +760,9 @@ declare i1 @llvm.vp.reduce.umax.nxv4i1(i1, <vscale x 4 x i1>, <vscale x 4 x i1>,
 define zeroext i1 @vpreduce_umax_nxv4i1(i1 zeroext %s, <vscale x 4 x i1> %v, <vscale x 4 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vpreduce_umax_nxv4i1:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a1, e8, mf2, ta, ma
 ; CHECK-NEXT:    vmv1r.v v9, v0
 ; CHECK-NEXT:    vmv1r.v v0, v8
-; CHECK-NEXT:    vsetvli zero, a1, e8, mf2, ta, ma
 ; CHECK-NEXT:    vcpop.m a1, v9, v0.t
 ; CHECK-NEXT:    snez a1, a1
 ; CHECK-NEXT:    or a0, a1, a0
@@ -775,9 +776,9 @@ declare i1 @llvm.vp.reduce.umax.nxv8i1(i1, <vscale x 8 x i1>, <vscale x 8 x i1>,
 define zeroext i1 @vpreduce_umax_nxv8i1(i1 zeroext %s, <vscale x 8 x i1> %v, <vscale x 8 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vpreduce_umax_nxv8i1:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a1, e8, m1, ta, ma
 ; CHECK-NEXT:    vmv1r.v v9, v0
 ; CHECK-NEXT:    vmv1r.v v0, v8
-; CHECK-NEXT:    vsetvli zero, a1, e8, m1, ta, ma
 ; CHECK-NEXT:    vcpop.m a1, v9, v0.t
 ; CHECK-NEXT:    snez a1, a1
 ; CHECK-NEXT:    or a0, a1, a0
@@ -791,9 +792,9 @@ declare i1 @llvm.vp.reduce.umax.nxv16i1(i1, <vscale x 16 x i1>, <vscale x 16 x i
 define zeroext i1 @vpreduce_umax_nxv16i1(i1 zeroext %s, <vscale x 16 x i1> %v, <vscale x 16 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vpreduce_umax_nxv16i1:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a1, e8, m2, ta, ma
 ; CHECK-NEXT:    vmv1r.v v9, v0
 ; CHECK-NEXT:    vmv1r.v v0, v8
-; CHECK-NEXT:    vsetvli zero, a1, e8, m2, ta, ma
 ; CHECK-NEXT:    vcpop.m a1, v9, v0.t
 ; CHECK-NEXT:    snez a1, a1
 ; CHECK-NEXT:    or a0, a1, a0
@@ -807,9 +808,9 @@ declare i1 @llvm.vp.reduce.umax.nxv32i1(i1, <vscale x 32 x i1>, <vscale x 32 x i
 define zeroext i1 @vpreduce_umax_nxv32i1(i1 zeroext %s, <vscale x 32 x i1> %v, <vscale x 32 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vpreduce_umax_nxv32i1:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a1, e8, m4, ta, ma
 ; CHECK-NEXT:    vmv1r.v v9, v0
 ; CHECK-NEXT:    vmv1r.v v0, v8
-; CHECK-NEXT:    vsetvli zero, a1, e8, m4, ta, ma
 ; CHECK-NEXT:    vcpop.m a1, v9, v0.t
 ; CHECK-NEXT:    snez a1, a1
 ; CHECK-NEXT:    or a0, a1, a0
@@ -823,9 +824,9 @@ declare i1 @llvm.vp.reduce.umax.nxv64i1(i1, <vscale x 64 x i1>, <vscale x 64 x i
 define zeroext i1 @vpreduce_umax_nxv64i1(i1 zeroext %s, <vscale x 64 x i1> %v, <vscale x 64 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vpreduce_umax_nxv64i1:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli zero, a1, e8, m8, ta, ma
 ; CHECK-NEXT:    vmv1r.v v9, v0
 ; CHECK-NEXT:    vmv1r.v v0, v8
-; CHECK-NEXT:    vsetvli zero, a1, e8, m8, ta, ma
 ; CHECK-NEXT:    vcpop.m a1, v9, v0.t
 ; CHECK-NEXT:    snez a1, a1
 ; CHECK-NEXT:    or a0, a1, a0
diff --git a/llvm/test/CodeGen/RISCV/rvv/vrgatherei16-subreg-liveness.ll b/llvm/test/CodeGen/RISCV/rvv/vrgatherei16-subreg-liveness.ll
index 1779fc12095e8..7b460f2c058f8 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vrgatherei16-subreg-liveness.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vrgatherei16-subreg-liveness.ll
@@ -22,8 +22,8 @@ define internal void @foo(<vscale x 1 x i16> %v15, <vscale x 1 x i16> %0, <vscal
 ; NOSUBREG-NEXT:  .LBB0_1: # %loopIR3.i.i
 ; NOSUBREG-NEXT:    # =>This Inner Loop Header: Depth=1
 ; NOSUBREG-NEXT:    vl1r.v v9, (zero)
-; NOSUBREG-NEXT:    vmv1r.v v13, v12
 ; NOSUBREG-NEXT:    vsetivli zero, 4, e8, m1, tu, ma
+; NOSUBREG-NEXT:    vmv1r.v v13, v12
 ; NOSUBREG-NEXT:    vrgatherei16.vv v13, v9, v10
 ; NOSUBREG-NEXT:    vsetvli a0, zero, e8, m1, ta, ma
 ; NOSUBREG-NEXT:    vand.vv v9, v8, v13
@@ -42,8 +42,8 @@ define internal void @foo(<vscale x 1 x i16> %v15, <vscale x 1 x i16> %0, <vscal
 ; SUBREG-NEXT:  .LBB0_1: # %loopIR3.i.i
 ; SUBREG-NEXT:    # =>This Inner Loop Header: Depth=1
 ; SUBREG-NEXT:    vl1r.v v9, (zero)
-; SUBREG-NEXT:    vmv1r.v v13, v12
 ; SUBREG-NEXT:    vsetivli zero, 4, e8, m1, tu, ma
+; SUBREG-NEXT:    vmv1r.v v13, v12
 ; SUBREG-NEXT:    vrgatherei16.vv v13, v9, v10
 ; SUBREG-NEXT:    vsetvli a0, zero, e8, m1, ta, ma
 ; SUBREG-NEXT:    vand.vv v9, v8, v13
diff --git a/llvm/test/CodeGen/RISCV/rvv/vsadd-vp.ll b/llvm/test/CodeGen/RISCV/rvv/vsadd-vp.ll
index 12c439346e356..3421c6af334bc 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vsadd-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vsadd-vp.ll
@@ -572,8 +572,8 @@ declare <vscale x 128 x i8> @llvm.vp.sadd.sat.nxv128i8(<vscale x 128 x i8>, <vsc
 define <vscale x 128 x i8> @vsadd_vi_nxv128i8(<vscale x 128 x i8> %va, <vscale x 128 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vsadd_vi_nxv128i8:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v24, v0
 ; CHECK-NEXT:    vsetvli a2, zero, e8, m8, ta, ma
+; CHECK-NEXT:    vmv1r.v v24, v0
 ; CHECK-NEXT:    vlm.v v0, (a0)
 ; CHECK-NEXT:    csrr a0, vlenb
 ; CHECK-NEXT:    slli a0, a0, 3
@@ -1350,11 +1350,11 @@ declare <vscale x 32 x i32> @llvm.vp.sadd.sat.nxv32i32(<vscale x 32 x i32>, <vsc
 define <vscale x 32 x i32> @vsadd_vi_nxv32i32(<vscale x 32 x i32> %va, <vscale x 32 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vsadd_vi_nxv32i32:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli a1, zero, e8, mf2, ta, ma
 ; CHECK-NEXT:    vmv1r.v v24, v0
 ; CHECK-NEXT:    csrr a1, vlenb
 ; CHECK-NEXT:    srli a2, a1, 2
 ; CHECK-NEXT:    slli a1, a1, 1
-; CHECK-NEXT:    vsetvli a3, zero, e8, mf2, ta, ma
 ; CHECK-NEXT:    vslidedown.vx v0, v0, a2
 ; CHECK-NEXT:    sub a2, a0, a1
 ; CHECK-NEXT:    sltu a3, a0, a2
diff --git a/llvm/test/CodeGen/RISCV/rvv/vsaddu-vp.ll b/llvm/test/CodeGen/RISCV/rvv/vsaddu-vp.ll
index d962f703abfd2..180e0799044e8 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vsaddu-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vsaddu-vp.ll
@@ -571,8 +571,8 @@ declare <vscale x 128 x i8> @llvm.vp.uadd.sat.nxv128i8(<vscale x 128 x i8>, <vsc
 define <vscale x 128 x i8> @vsaddu_vi_nxv128i8(<vscale x 128 x i8> %va, <vscale x 128 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vsaddu_vi_nxv128i8:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v24, v0
 ; CHECK-NEXT:    vsetvli a2, zero, e8, m8, ta, ma
+; CHECK-NEXT:    vmv1r.v v24, v0
 ; CHECK-NEXT:    vlm.v v0, (a0)
 ; CHECK-NEXT:    csrr a0, vlenb
 ; CHECK-NEXT:    slli a0, a0, 3
@@ -1349,11 +1349,11 @@ declare <vscale x 32 x i32> @llvm.vp.uadd.sat.nxv32i32(<vscale x 32 x i32>, <vsc
 define <vscale x 32 x i32> @vsaddu_vi_nxv32i32(<vscale x 32 x i32> %va, <vscale x 32 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vsaddu_vi_nxv32i32:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli a1, zero, e8, mf2, ta, ma
 ; CHECK-NEXT:    vmv1r.v v24, v0
 ; CHECK-NEXT:    csrr a1, vlenb
 ; CHECK-NEXT:    srli a2, a1, 2
 ; CHECK-NEXT:    slli a1, a1, 1
-; CHECK-NEXT:    vsetvli a3, zero, e8, mf2, ta, ma
 ; CHECK-NEXT:    vslidedown.vx v0, v0, a2
 ; CHECK-NEXT:    sub a2, a0, a1
 ; CHECK-NEXT:    sltu a3, a0, a2
diff --git a/llvm/test/CodeGen/RISCV/rvv/vselect-bf16.ll b/llvm/test/CodeGen/RISCV/rvv/vselect-bf16.ll
index a63d14e8b6c04..b793531e34069 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vselect-bf16.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vselect-bf16.ll
@@ -126,6 +126,7 @@ define <vscale x 8 x bfloat> @vmerge_truelhs_nxv8bf16_0(<vscale x 8 x bfloat> %v
 define <vscale x 8 x bfloat> @vmerge_falselhs_nxv8bf16_0(<vscale x 8 x bfloat> %va, <vscale x 8 x bfloat> %vb) {
 ; CHECK-LABEL: vmerge_falselhs_nxv8bf16_0:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-NEXT:    vmv2r.v v8, v10
 ; CHECK-NEXT:    ret
   %vc = select <vscale x 8 x i1> zeroinitializer, <vscale x 8 x bfloat> %va, <vscale x 8 x bfloat> %vb
diff --git a/llvm/test/CodeGen/RISCV/rvv/vselect-fp.ll b/llvm/test/CodeGen/RISCV/rvv/vselect-fp.ll
index 1fc33dc73a27d..be2fc6955294d 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vselect-fp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vselect-fp.ll
@@ -175,6 +175,7 @@ define <vscale x 8 x half> @vmerge_truelhs_nxv8f16_0(<vscale x 8 x half> %va, <v
 define <vscale x 8 x half> @vmerge_falselhs_nxv8f16_0(<vscale x 8 x half> %va, <vscale x 8 x half> %vb) {
 ; CHECK-LABEL: vmerge_falselhs_nxv8f16_0:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-NEXT:    vmv2r.v v8, v10
 ; CHECK-NEXT:    ret
   %vc = select <vscale x 8 x i1> zeroinitializer, <vscale x 8 x half> %va, <vscale x 8 x half> %vb
diff --git a/llvm/test/CodeGen/RISCV/rvv/vselect-int.ll b/llvm/test/CodeGen/RISCV/rvv/vselect-int.ll
index 9cafa28eb429f..4ec9e344e6278 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vselect-int.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vselect-int.ll
@@ -803,6 +803,7 @@ define <vscale x 8 x i64> @vmerge_truelhs_nxv8i64_0(<vscale x 8 x i64> %va, <vsc
 define <vscale x 8 x i64> @vmerge_falselhs_nxv8i64_0(<vscale x 8 x i64> %va, <vscale x 8 x i64> %vb) {
 ; CHECK-LABEL: vmerge_falselhs_nxv8i64_0:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-NEXT:    vmv8r.v v8, v16
 ; CHECK-NEXT:    ret
   %vc = select <vscale x 8 x i1> zeroinitializer, <vscale x 8 x i64> %va, <vscale x 8 x i64> %vb
diff --git a/llvm/test/CodeGen/RISCV/rvv/vselect-vp.ll b/llvm/test/CodeGen/RISCV/rvv/vselect-vp.ll
index bb51f0592dc17..d4ebe27420d7b 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vselect-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vselect-vp.ll
@@ -362,6 +362,7 @@ define <vscale x 32 x i32> @select_nxv32i32(<vscale x 32 x i1> %a, <vscale x 32
 ; CHECK-NEXT:    add a1, sp, a1
 ; CHECK-NEXT:    addi a1, a1, 16
 ; CHECK-NEXT:    vs8r.v v8, (a1) # Unknown-size Folded Spill
+; CHECK-NEXT:    vsetvli a1, zero, e8, mf2, ta, ma
 ; CHECK-NEXT:    vmv1r.v v24, v0
 ; CHECK-NEXT:    csrr a3, vlenb
 ; CHECK-NEXT:    slli a4, a3, 3
@@ -375,7 +376,6 @@ define <vscale x 32 x i32> @select_nxv32i32(<vscale x 32 x i1> %a, <vscale x 32
 ; CHECK-NEXT:    vl8re32.v v0, (a0)
 ; CHECK-NEXT:    addi a0, sp, 16
 ; CHECK-NEXT:    vs8r.v v0, (a0) # Unknown-size Folded Spill
-; CHECK-NEXT:    vsetvli a0, zero, e8, mf2, ta, ma
 ; CHECK-NEXT:    vslidedown.vx v0, v24, a3
 ; CHECK-NEXT:    and a4, a4, a5
 ; CHECK-NEXT:    vsetvli zero, a4, e32, m8, ta, ma
@@ -421,6 +421,7 @@ define <vscale x 32 x i32> @select_evl_nxv32i32(<vscale x 32 x i1> %a, <vscale x
 ; CHECK-NEXT:    add a1, sp, a1
 ; CHECK-NEXT:    addi a1, a1, 16
 ; CHECK-NEXT:    vs8r.v v8, (a1) # Unknown-size Folded Spill
+; CHECK-NEXT:    vsetvli a1, zero, e8, mf2, ta, ma
 ; CHECK-NEXT:    vmv1r.v v24, v0
 ; CHECK-NEXT:    csrr a1, vlenb
 ; CHECK-NEXT:    slli a3, a1, 3
@@ -434,7 +435,6 @@ define <vscale x 32 x i32> @select_evl_nxv32i32(<vscale x 32 x i1> %a, <vscale x
 ; CHECK-NEXT:    vl8re32.v v0, (a0)
 ; CHECK-NEXT:    addi a0, sp, 16
 ; CHECK-NEXT:    vs8r.v v0, (a0) # Unknown-size Folded Spill
-; CHECK-NEXT:    vsetvli a0, zero, e8, mf2, ta, ma
 ; CHECK-NEXT:    vslidedown.vx v0, v24, a4
 ; CHECK-NEXT:    and a3, a3, a5
 ; CHECK-NEXT:    vsetvli zero, a3, e32, m8, ta, ma
@@ -710,6 +710,7 @@ define <vscale x 16 x double> @select_nxv16f64(<vscale x 16 x i1> %a, <vscale x
 ; CHECK-NEXT:    add a1, sp, a1
 ; CHECK-NEXT:    addi a1, a1, 16
 ; CHECK-NEXT:    vs8r.v v8, (a1) # Unknown-size Folded Spill
+; CHECK-NEXT:    vsetvli a1, zero, e8, mf4, ta, ma
 ; CHECK-NEXT:    vmv1r.v v24, v0
 ; CHECK-NEXT:    csrr a1, vlenb
 ; CHECK-NEXT:    slli a3, a1, 3
@@ -722,7 +723,6 @@ define <vscale x 16 x double> @select_nxv16f64(<vscale x 16 x i1> %a, <vscale x
 ; CHECK-NEXT:    vl8re64.v v0, (a0)
 ; CHECK-NEXT:    addi a0, sp, 16
 ; CHECK-NEXT:    vs8r.v v0, (a0) # Unknown-size Folded Spill
-; CHECK-NEXT:    vsetvli a0, zero, e8, mf4, ta, ma
 ; CHECK-NEXT:    vslidedown.vx v0, v24, a3
 ; CHECK-NEXT:    and a4, a5, a4
 ; CHECK-NEXT:    vsetvli zero, a4, e64, m8, ta, ma
@@ -835,6 +835,7 @@ define <vscale x 2 x i1> @select_cond_x_cond(<vscale x 2 x i1> %x, <vscale x 2 x
 define <vscale x 2 x i1> @select_undef_T_F(<vscale x 2 x i1> %x, <vscale x 2 x i1> %y, i32 zeroext %evl) {
 ; CHECK-LABEL: select_undef_T_F:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-NEXT:    vmv1r.v v0, v8
 ; CHECK-NEXT:    ret
   %a = call <vscale x 2 x i1> @llvm.vp.select.nxv2i1(<vscale x 2 x i1> poison, <vscale x 2 x i1> %x, <vscale x 2 x i1> %y, i32 %evl)
@@ -852,6 +853,7 @@ define <vscale x 2 x i1> @select_undef_undef_F(<vscale x 2 x i1> %x, i32 zeroext
 define <vscale x 2 x i1> @select_unknown_undef_F(<vscale x 2 x i1> %x, <vscale x 2 x i1> %y, i32 zeroext %evl) {
 ; CHECK-LABEL: select_unknown_undef_F:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-NEXT:    vmv1r.v v0, v8
 ; CHECK-NEXT:    ret
   %a = call <vscale x 2 x i1> @llvm.vp.select.nxv2i1(<vscale x 2 x i1> %x, <vscale x 2 x i1> undef, <vscale x 2 x i1> %y, i32 %evl)
@@ -861,6 +863,7 @@ define <vscale x 2 x i1> @select_unknown_undef_F(<vscale x 2 x i1> %x, <vscale x
 define <vscale x 2 x i1> @select_unknown_T_undef(<vscale x 2 x i1> %x, <vscale x 2 x i1> %y, i32 zeroext %evl) {
 ; CHECK-LABEL: select_unknown_T_undef:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-NEXT:    vmv1r.v v0, v8
 ; CHECK-NEXT:    ret
   %a = call <vscale x 2 x i1> @llvm.vp.select.nxv2i1(<vscale x 2 x i1> %x, <vscale x 2 x i1> %y, <vscale x 2 x i1> poison, i32 %evl)
@@ -870,6 +873,7 @@ define <vscale x 2 x i1> @select_unknown_T_undef(<vscale x 2 x i1> %x, <vscale x
 define <vscale x 2 x i1> @select_false_T_F(<vscale x 2 x i1> %x, <vscale x 2 x i1> %y, <vscale x 2 x i1> %z, i32 zeroext %evl) {
 ; CHECK-LABEL: select_false_T_F:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-NEXT:    vmv1r.v v0, v9
 ; CHECK-NEXT:    ret
   %a = call <vscale x 2 x i1> @llvm.vp.select.nxv2i1(<vscale x 2 x i1> zeroinitializer, <vscale x 2 x i1> %y, <vscale x 2 x i1> %z, i32 %evl)
@@ -879,6 +883,7 @@ define <vscale x 2 x i1> @select_false_T_F(<vscale x 2 x i1> %x, <vscale x 2 x i
 define <vscale x 2 x i1> @select_unknown_T_T(<vscale x 2 x i1> %x, <vscale x 2 x i1> %y, i32 zeroext %evl) {
 ; CHECK-LABEL: select_unknown_T_T:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-NEXT:    vmv1r.v v0, v8
 ; CHECK-NEXT:    ret
   %a = call <vscale x 2 x i1> @llvm.vp.select.nxv2i1(<vscale x 2 x i1> %x, <vscale x 2 x i1> %y, <vscale x 2 x i1> %y, i32 %evl)
diff --git a/llvm/test/CodeGen/RISCV/rvv/vsetvli-insert-O0.ll b/llvm/test/CodeGen/RISCV/rvv/vsetvli-insert-O0.ll
index 33acfb7dceb94..d8620d0658261 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vsetvli-insert-O0.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vsetvli-insert-O0.ll
@@ -18,11 +18,11 @@ declare <vscale x 1 x i64> @llvm.riscv.vle.mask.nxv1i64(
 define <2 x double> @fixed_length(<2 x double> %a, <2 x double> %b) nounwind {
 ; CHECK-LABEL: fixed_length:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetivli zero, 2, e64, m1, ta, ma
 ; CHECK-NEXT:    vmv1r.v v10, v9
 ; CHECK-NEXT:    # kill: def $v11 killed $v10
 ; CHECK-NEXT:    # kill: def $v9 killed $v8
 ; CHECK-NEXT:    # implicit-def: $v9
-; CHECK-NEXT:    vsetivli zero, 2, e64, m1, ta, ma
 ; CHECK-NEXT:    vfadd.vv v9, v8, v10
 ; CHECK-NEXT:    # implicit-def: $v8
 ; CHECK-NEXT:    vfadd.vv v8, v9, v10
@@ -36,9 +36,9 @@ entry:
 define <vscale x 1 x double> @scalable(<vscale x 1 x double> %a, <vscale x 1 x double> %b) nounwind {
 ; CHECK-LABEL: scalable:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetvli a0, zero, e64, m1, ta, ma
 ; CHECK-NEXT:    vmv1r.v v10, v9
 ; CHECK-NEXT:    # implicit-def: $v9
-; CHECK-NEXT:    vsetvli a0, zero, e64, m1, ta, ma
 ; CHECK-NEXT:    vfadd.vv v9, v8, v10
 ; CHECK-NEXT:    # implicit-def: $v8
 ; CHECK-NEXT:    vfadd.vv v8, v9, v10
@@ -53,8 +53,8 @@ entry:
 define <vscale x 1 x double> @intrinsic_same_vlmax(<vscale x 1 x double> %a, <vscale x 1 x double> %b) nounwind {
 ; CHECK-LABEL: intrinsic_same_vlmax:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v10, v9
 ; CHECK-NEXT:    vsetvli a0, zero, e32, mf2, ta, ma
+; CHECK-NEXT:    vmv1r.v v10, v9
 ; CHECK-NEXT:    # implicit-def: $v9
 ; CHECK-NEXT:    vsetvli zero, a0, e64, m1, tu, ma
 ; CHECK-NEXT:    vfadd.vv v9, v8, v10
@@ -81,8 +81,8 @@ entry:
 define <vscale x 1 x double> @intrinsic_same_avl_imm(<vscale x 1 x double> %a, <vscale x 1 x double> %b) nounwind {
 ; CHECK-LABEL: intrinsic_same_avl_imm:
 ; CHECK:       # %bb.0: # %entry
-; CHECK-NEXT:    vmv1r.v v10, v9
 ; CHECK-NEXT:    vsetivli a0, 2, e32, mf2, ta, ma
+; CHECK-NEXT:    vmv1r.v v10, v9
 ; CHECK-NEXT:    # implicit-def: $v9
 ; CHECK-NEXT:    vsetvli zero, a0, e64, m1, tu, ma
 ; CHECK-NEXT:    vfadd.vv v9, v8, v10
@@ -108,6 +108,7 @@ entry:
 define <vscale x 1 x double> @intrinsic_same_avl_reg(i64 %avl, <vscale x 1 x double> %a, <vscale x 1 x double> %b) nounwind {
 ; CHECK-LABEL: intrinsic_same_avl_reg:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-NEXT:    vmv1r.v v10, v9
 ; CHECK-NEXT:    vsetvli a0, a0, e32, mf2, ta, ma
 ; CHECK-NEXT:    # implicit-def: $v9
@@ -135,6 +136,7 @@ entry:
 define <vscale x 1 x double> @intrinsic_diff_avl_reg(i64 %avl, i64 %avl2, <vscale x 1 x double> %a, <vscale x 1 x double> %b) nounwind {
 ; CHECK-LABEL: intrinsic_diff_avl_reg:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
 ; CHECK-NEXT:    vmv1r.v v10, v9
 ; CHECK-NEXT:    vsetvli a0, a0, e32, mf2, ta, ma
 ; CHECK-NEXT:    # implicit-def: $v9
diff --git a/llvm/test/CodeGen/RISCV/rvv/vsetvli-insert-crossbb.mir b/llvm/test/CodeGen/RISCV/rvv/vsetvli-insert-crossbb.mir
index bf93f5cc1f6f2..55cefbbea81b2 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vsetvli-insert-crossbb.mir
+++ b/llvm/test/CodeGen/RISCV/rvv/vsetvli-insert-crossbb.mir
@@ -835,8 +835,8 @@ body:             |
   ; CHECK-NEXT:   PseudoBR %bb.3
   ; CHECK-NEXT: {{  $}}
   ; CHECK-NEXT: bb.3:
+  ; CHECK-NEXT:   dead [[PseudoVSETVLIX0_1:%[0-9]+]]:gpr = PseudoVSETVLIX0 $x0, 197 /* e8, mf8, ta, ma */, implicit-def $vl, implicit-def $vtype
   ; CHECK-NEXT:   $v0 = COPY %mask
-  ; CHECK-NEXT:   dead $x0 = PseudoVSETVLIX0 killed $x0, 197 /* e8, mf8, ta, ma */, implicit-def $vl, implicit-def $vtype, implicit $vl
   ; CHECK-NEXT:   PseudoVSOXEI64_V_M1_MF8_MASK [[COPY]], %b, %idxs, $v0, -1, 3 /* e8 */, implicit $vl, implicit $vtype
   ; CHECK-NEXT:   PseudoRET
   bb.0:
diff --git a/llvm/test/CodeGen/RISCV/rvv/vsetvli-insert.ll b/llvm/test/CodeGen/RISCV/rvv/vsetvli-insert.ll
index b0cb6bc6125dd..8b48dc43eca29 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vsetvli-insert.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vsetvli-insert.ll
@@ -377,8 +377,8 @@ entry:
 define <vscale x 1 x double> @test19(<vscale x 1 x double> %a, double %b) nounwind {
 ; CHECK-LABEL: test19:
 ; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    vsetivli zero, 1, e64, m8, tu, ma
 ; CHECK-NEXT:    vmv1r.v v9, v8
-; CHECK-NEXT:    vsetivli zero, 2, e64, m1, tu, ma
 ; CHECK-NEXT:    vfmv.s.f v9, fa0
 ; CHECK-NEXT:    vsetvli a0, zero, e64, m1, ta, ma
 ; CHECK-NEXT:    vfadd.vv v8, v9, v8
diff --git a/llvm/test/CodeGen/RISCV/rvv/vsext-vp.ll b/llvm/test/CodeGen/RISCV/rvv/vsext-vp.ll
index d3b905ef897b1..3c91131fe4d12 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vsext-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vsext-vp.ll
@@ -151,11 +151,11 @@ declare <vscale x 32 x i32> @llvm.vp.sext.nxv32i32.nxv32i8(<vscale x 32 x i8>, <
 define <vscale x 32 x i32> @vsext_nxv32i8_nxv32i32(<vscale x 32 x i8> %a, <vscale x 32 x i1> %m, i32 zeroext %vl) {
 ; CHECK-LABEL: vsext_nxv32i8_nxv32i32:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli a1, zero, e8, mf2, ta, ma
 ; CHECK-NEXT:    vmv1r.v v12, v0
 ; CHECK-NEXT:    csrr a1, vlenb
 ; CHECK-NEXT:    srli a2, a1, 2
 ; CHECK-NEXT:    slli a1, a1, 1
-; CHECK-NEXT:    vsetvli a3, zero, e8, mf2, ta, ma
 ; CHECK-NEXT:    vslidedown.vx v0, v0, a2
 ; CHECK-NEXT:    sub a2, a0, a1
 ; CHECK-NEXT:    sltu a3, a0, a2
diff --git a/llvm/test/CodeGen/RISCV/rvv/vsitofp-vp.ll b/llvm/test/CodeGen/RISCV/rvv/vsitofp-vp.ll
index 581cc666b6cbd..44d3ee96f5e61 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vsitofp-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vsitofp-vp.ll
@@ -508,11 +508,11 @@ declare <vscale x 32 x half> @llvm.vp.sitofp.nxv32f16.nxv32i32(<vscale x 32 x i3
 define <vscale x 32 x half> @vsitofp_nxv32f16_nxv32i32(<vscale x 32 x i32> %va, <vscale x 32 x i1> %m, i32 zeroext %evl) {
 ; ZVFH-LABEL: vsitofp_nxv32f16_nxv32i32:
 ; ZVFH:       # %bb.0:
+; ZVFH-NEXT:    vsetvli a1, zero, e8, mf2, ta, ma
 ; ZVFH-NEXT:    vmv1r.v v24, v0
 ; ZVFH-NEXT:    csrr a1, vlenb
 ; ZVFH-NEXT:    srli a2, a1, 2
 ; ZVFH-NEXT:    slli a1, a1, 1
-; ZVFH-NEXT:    vsetvli a3, zero, e8, mf2, ta, ma
 ; ZVFH-NEXT:    vslidedown.vx v0, v0, a2
 ; ZVFH-NEXT:    sub a2, a0, a1
 ; ZVFH-NEXT:    sltu a3, a0, a2
@@ -532,11 +532,11 @@ define <vscale x 32 x half> @vsitofp_nxv32f16_nxv32i32(<vscale x 32 x i32> %va,
 ;
 ; ZVFHMIN-LABEL: vsitofp_nxv32f16_nxv32i32:
 ; ZVFHMIN:       # %bb.0:
+; ZVFHMIN-NEXT:    vsetvli a1, zero, e8, mf2, ta, ma
 ; ZVFHMIN-NEXT:    vmv1r.v v7, v0
 ; ZVFHMIN-NEXT:    csrr a1, vlenb
 ; ZVFHMIN-NEXT:    srli a2, a1, 2
 ; ZVFHMIN-NEXT:    slli a1, a1, 1
-; ZVFHMIN-NEXT:    vsetvli a3, zero, e8, mf2, ta, ma
 ; ZVFHMIN-NEXT:    vslidedown.vx v0, v0, a2
 ; ZVFHMIN-NEXT:    sub a2, a0, a1
 ; ZVFHMIN-NEXT:    sltu a3, a0, a2
@@ -566,11 +566,11 @@ declare <vscale x 32 x float> @llvm.vp.sitofp.nxv32f32.nxv32i32(<vscale x 32 x i
 define <vscale x 32 x float> @vsitofp_nxv32f32_nxv32i32(<vscale x 32 x i32> %va, <vscale x 32 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vsitofp_nxv32f32_nxv32i32:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli a1, zero, e8, mf2, ta, ma
 ; CHECK-NEXT:    vmv1r.v v24, v0
 ; CHECK-NEXT:    csrr a1, vlenb
 ; CHECK-NEXT:    srli a2, a1, 2
 ; CHECK-NEXT:    slli a1, a1, 1
-; CHECK-NEXT:    vsetvli a3, zero, e8, mf2, ta, ma
 ; CHECK-NEXT:    vslidedown.vx v0, v0, a2
 ; CHECK-NEXT:    sub a2, a0, a1
 ; CHECK-NEXT:    sltu a3, a0, a2
diff --git a/llvm/test/CodeGen/RISCV/rvv/vssub-vp.ll b/llvm/test/CodeGen/RISCV/rvv/vssub-vp.ll
index f9c24eeec31c5..7ee6ea9e19df0 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vssub-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vssub-vp.ll
@@ -590,8 +590,8 @@ declare <vscale x 128 x i8> @llvm.vp.ssub.sat.nxv128i8(<vscale x 128 x i8>, <vsc
 define <vscale x 128 x i8> @vssub_vi_nxv128i8(<vscale x 128 x i8> %va, <vscale x 128 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vssub_vi_nxv128i8:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v24, v0
 ; CHECK-NEXT:    vsetvli a2, zero, e8, m8, ta, ma
+; CHECK-NEXT:    vmv1r.v v24, v0
 ; CHECK-NEXT:    vlm.v v0, (a0)
 ; CHECK-NEXT:    csrr a0, vlenb
 ; CHECK-NEXT:    slli a0, a0, 3
@@ -1392,11 +1392,11 @@ declare <vscale x 32 x i32> @llvm.vp.ssub.sat.nxv32i32(<vscale x 32 x i32>, <vsc
 define <vscale x 32 x i32> @vssub_vi_nxv32i32(<vscale x 32 x i32> %va, <vscale x 32 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vssub_vi_nxv32i32:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli a1, zero, e8, mf2, ta, ma
 ; CHECK-NEXT:    vmv1r.v v24, v0
 ; CHECK-NEXT:    csrr a1, vlenb
 ; CHECK-NEXT:    srli a2, a1, 2
 ; CHECK-NEXT:    slli a1, a1, 1
-; CHECK-NEXT:    vsetvli a3, zero, e8, mf2, ta, ma
 ; CHECK-NEXT:    vslidedown.vx v0, v0, a2
 ; CHECK-NEXT:    sub a2, a0, a1
 ; CHECK-NEXT:    sltu a3, a0, a2
diff --git a/llvm/test/CodeGen/RISCV/rvv/vssubu-vp.ll b/llvm/test/CodeGen/RISCV/rvv/vssubu-vp.ll
index 04a1b522a8a33..7674a457ca961 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vssubu-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vssubu-vp.ll
@@ -588,8 +588,8 @@ declare <vscale x 128 x i8> @llvm.vp.usub.sat.nxv128i8(<vscale x 128 x i8>, <vsc
 define <vscale x 128 x i8> @vssubu_vi_nxv128i8(<vscale x 128 x i8> %va, <vscale x 128 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vssubu_vi_nxv128i8:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vmv1r.v v24, v0
 ; CHECK-NEXT:    vsetvli a2, zero, e8, m8, ta, ma
+; CHECK-NEXT:    vmv1r.v v24, v0
 ; CHECK-NEXT:    vlm.v v0, (a0)
 ; CHECK-NEXT:    csrr a0, vlenb
 ; CHECK-NEXT:    slli a0, a0, 3
@@ -1390,11 +1390,11 @@ declare <vscale x 32 x i32> @llvm.vp.usub.sat.nxv32i32(<vscale x 32 x i32>, <vsc
 define <vscale x 32 x i32> @vssubu_vi_nxv32i32(<vscale x 32 x i32> %va, <vscale x 32 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vssubu_vi_nxv32i32:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli a1, zero, e8, mf2, ta, ma
 ; CHECK-NEXT:    vmv1r.v v24, v0
 ; CHECK-NEXT:    csrr a1, vlenb
 ; CHECK-NEXT:    srli a2, a1, 2
 ; CHECK-NEXT:    slli a1, a1, 1
-; CHECK-NEXT:    vsetvli a3, zero, e8, mf2, ta, ma
 ; CHECK-NEXT:    vslidedown.vx v0, v0, a2
 ; CHECK-NEXT:    sub a2, a0, a1
 ; CHECK-NEXT:    sltu a3, a0, a2
diff --git a/llvm/test/CodeGen/RISCV/rvv/vtrunc-vp.ll b/llvm/test/CodeGen/RISCV/rvv/vtrunc-vp.ll
index e62b7a0039638..fd5bf4ebcede8 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vtrunc-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vtrunc-vp.ll
@@ -157,11 +157,11 @@ declare <vscale x 15 x i16> @llvm.vp.trunc.nxv15i16.nxv15i64(<vscale x 15 x i64>
 define <vscale x 15 x i16> @vtrunc_nxv15i16_nxv15i64(<vscale x 15 x i64> %a, <vscale x 15 x i1> %m, i32 zeroext %vl) {
 ; CHECK-LABEL: vtrunc_nxv15i16_nxv15i64:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli a1, zero, e8, mf4, ta, ma
 ; CHECK-NEXT:    vmv1r.v v24, v0
 ; CHECK-NEXT:    csrr a1, vlenb
 ; CHECK-NEXT:    srli a2, a1, 3
 ; CHECK-NEXT:    sub a3, a0, a1
-; CHECK-NEXT:    vsetvli a4, zero, e8, mf4, ta, ma
 ; CHECK-NEXT:    vslidedown.vx v0, v0, a2
 ; CHECK-NEXT:    sltu a2, a0, a3
 ; CHECK-NEXT:    addi a2, a2, -1
@@ -214,11 +214,11 @@ declare <vscale x 32 x i7> @llvm.vp.trunc.nxv32i7.nxv32i32(<vscale x 32 x i32>,
 define <vscale x 32 x i7> @vtrunc_nxv32i7_nxv32i32(<vscale x 32 x i32> %a, <vscale x 32 x i1> %m, i32 zeroext %vl) {
 ; CHECK-LABEL: vtrunc_nxv32i7_nxv32i32:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli a1, zero, e8, mf2, ta, ma
 ; CHECK-NEXT:    vmv1r.v v24, v0
 ; CHECK-NEXT:    csrr a1, vlenb
 ; CHECK-NEXT:    srli a2, a1, 2
 ; CHECK-NEXT:    slli a1, a1, 1
-; CHECK-NEXT:    vsetvli a3, zero, e8, mf2, ta, ma
 ; CHECK-NEXT:    vslidedown.vx v0, v0, a2
 ; CHECK-NEXT:    sub a2, a0, a1
 ; CHECK-NEXT:    sltu a3, a0, a2
@@ -248,11 +248,11 @@ declare <vscale x 32 x i8> @llvm.vp.trunc.nxv32i8.nxv32i32(<vscale x 32 x i32>,
 define <vscale x 32 x i8> @vtrunc_nxv32i8_nxv32i32(<vscale x 32 x i32> %a, <vscale x 32 x i1> %m, i32 zeroext %vl) {
 ; CHECK-LABEL: vtrunc_nxv32i8_nxv32i32:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli a1, zero, e8, mf2, ta, ma
 ; CHECK-NEXT:    vmv1r.v v24, v0
 ; CHECK-NEXT:    csrr a1, vlenb
 ; CHECK-NEXT:    srli a2, a1, 2
 ; CHECK-NEXT:    slli a1, a1, 1
-; CHECK-NEXT:    vsetvli a3, zero, e8, mf2, ta, ma
 ; CHECK-NEXT:    vslidedown.vx v0, v0, a2
 ; CHECK-NEXT:    sub a2, a0, a1
 ; CHECK-NEXT:    sltu a3, a0, a2
@@ -288,6 +288,7 @@ define <vscale x 32 x i32> @vtrunc_nxv32i64_nxv32i32(<vscale x 32 x i64> %a, <vs
 ; CHECK-NEXT:    slli a1, a1, 4
 ; CHECK-NEXT:    sub sp, sp, a1
 ; CHECK-NEXT:    .cfi_escape 0x0f, 0x0d, 0x72, 0x00, 0x11, 0x10, 0x22, 0x11, 0x10, 0x92, 0xa2, 0x38, 0x00, 0x1e, 0x22 # sp + 16 + 16 * vlenb
+; CHECK-NEXT:    vsetvli a1, zero, e8, mf2, ta, ma
 ; CHECK-NEXT:    vmv1r.v v7, v0
 ; CHECK-NEXT:    addi a1, sp, 16
 ; CHECK-NEXT:    vs8r.v v16, (a1) # Unknown-size Folded Spill
@@ -301,7 +302,6 @@ define <vscale x 32 x i32> @vtrunc_nxv32i64_nxv32i32(<vscale x 32 x i64> %a, <vs
 ; CHECK-NEXT:    srli a5, a1, 2
 ; CHECK-NEXT:    slli a6, a1, 3
 ; CHECK-NEXT:    slli a4, a1, 1
-; CHECK-NEXT:    vsetvli a7, zero, e8, mf2, ta, ma
 ; CHECK-NEXT:    vslidedown.vx v16, v0, a5
 ; CHECK-NEXT:    add a6, a0, a6
 ; CHECK-NEXT:    sub a5, a2, a4
diff --git a/llvm/test/CodeGen/RISCV/rvv/vuitofp-vp.ll b/llvm/test/CodeGen/RISCV/rvv/vuitofp-vp.ll
index 2c5a01279d5d3..f17cdd9c72bfc 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vuitofp-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vuitofp-vp.ll
@@ -500,11 +500,11 @@ declare <vscale x 32 x half> @llvm.vp.uitofp.nxv32f16.nxv32i32(<vscale x 32 x i3
 define <vscale x 32 x half> @vuitofp_nxv32f16_nxv32i32(<vscale x 32 x i32> %va, <vscale x 32 x i1> %m, i32 zeroext %evl) {
 ; ZVFH-LABEL: vuitofp_nxv32f16_nxv32i32:
 ; ZVFH:       # %bb.0:
+; ZVFH-NEXT:    vsetvli a1, zero, e8, mf2, ta, ma
 ; ZVFH-NEXT:    vmv1r.v v24, v0
 ; ZVFH-NEXT:    csrr a1, vlenb
 ; ZVFH-NEXT:    srli a2, a1, 2
 ; ZVFH-NEXT:    slli a1, a1, 1
-; ZVFH-NEXT:    vsetvli a3, zero, e8, mf2, ta, ma
 ; ZVFH-NEXT:    vslidedown.vx v0, v0, a2
 ; ZVFH-NEXT:    sub a2, a0, a1
 ; ZVFH-NEXT:    sltu a3, a0, a2
@@ -524,11 +524,11 @@ define <vscale x 32 x half> @vuitofp_nxv32f16_nxv32i32(<vscale x 32 x i32> %va,
 ;
 ; ZVFHMIN-LABEL: vuitofp_nxv32f16_nxv32i32:
 ; ZVFHMIN:       # %bb.0:
+; ZVFHMIN-NEXT:    vsetvli a1, zero, e8, mf2, ta, ma
 ; ZVFHMIN-NEXT:    vmv1r.v v7, v0
 ; ZVFHMIN-NEXT:    csrr a1, vlenb
 ; ZVFHMIN-NEXT:    srli a2, a1, 2
 ; ZVFHMIN-NEXT:    slli a1, a1, 1
-; ZVFHMIN-NEXT:    vsetvli a3, zero, e8, mf2, ta, ma
 ; ZVFHMIN-NEXT:    vslidedown.vx v0, v0, a2
 ; ZVFHMIN-NEXT:    sub a2, a0, a1
 ; ZVFHMIN-NEXT:    sltu a3, a0, a2
@@ -558,11 +558,11 @@ declare <vscale x 32 x float> @llvm.vp.uitofp.nxv32f32.nxv32i32(<vscale x 32 x i
 define <vscale x 32 x float> @vuitofp_nxv32f32_nxv32i32(<vscale x 32 x i32> %va, <vscale x 32 x i1> %m, i32 zeroext %evl) {
 ; CHECK-LABEL: vuitofp_nxv32f32_nxv32i32:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli a1, zero, e8, mf2, ta, ma
 ; CHECK-NEXT:    vmv1r.v v24, v0
 ; CHECK-NEXT:    csrr a1, vlenb
 ; CHECK-NEXT:    srli a2, a1, 2
 ; CHECK-NEXT:    slli a1, a1, 1
-; CHECK-NEXT:    vsetvli a3, zero, e8, mf2, ta, ma
 ; CHECK-NEXT:    vslidedown.vx v0, v0, a2
 ; CHECK-NEXT:    sub a2, a0, a1
 ; CHECK-NEXT:    sltu a3, a0, a2
diff --git a/llvm/test/CodeGen/RISCV/rvv/vzext-vp.ll b/llvm/test/CodeGen/RISCV/rvv/vzext-vp.ll
index 10e655c844540..934d7eb43ac2a 100644
--- a/llvm/test/CodeGen/RISCV/rvv/vzext-vp.ll
+++ b/llvm/test/CodeGen/RISCV/rvv/vzext-vp.ll
@@ -151,11 +151,11 @@ declare <vscale x 32 x i32> @llvm.vp.zext.nxv32i32.nxv32i8(<vscale x 32 x i8>, <
 define <vscale x 32 x i32> @vzext_nxv32i8_nxv32i32(<vscale x 32 x i8> %a, <vscale x 32 x i1> %m, i32 zeroext %vl) {
 ; CHECK-LABEL: vzext_nxv32i8_nxv32i32:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli a1, zero, e8, mf2, ta, ma
 ; CHECK-NEXT:    vmv1r.v v12, v0
 ; CHECK-NEXT:    csrr a1, vlenb
 ; CHECK-NEXT:    srli a2, a1, 2
 ; CHECK-NEXT:    slli a1, a1, 1
-; CHECK-NEXT:    vsetvli a3, zero, e8, mf2, ta, ma
 ; CHECK-NEXT:    vslidedown.vx v0, v0, a2
 ; CHECK-NEXT:    sub a2, a0, a1
 ; CHECK-NEXT:    sltu a3, a0, a2