[AArch64] Unfold adds when eliminating frame index with scalable offset

llvm/lib/Target/AArch64/AArch64InstrInfo.cpp

@@ -6263,6 +6263,11 @@ void llvm::emitFrameOffset(MachineBasicBlock &MBB,
   AArch64InstrInfo::decomposeStackOffsetForFrameOffsets(
       Offset, Bytes, NumPredicateVectors, NumDataVectors);
 
+  // Insert ADDSXri for scalable offset at the end.
+  bool NeedInsertADDS = SetNZCV && (NumPredicateVectors || NumDataVectors);
+  if (NeedInsertADDS)
+    SetNZCV = false;
+
   // First emit non-scalable frame offsets, or a simple 'mov'.
   if (Bytes || (!Offset && SrcReg != DestReg)) {
     assert((DestReg != AArch64::SP || Bytes % 8 == 0) &&
@@ -6282,8 +6287,6 @@ void llvm::emitFrameOffset(MachineBasicBlock &MBB,
     FrameReg = DestReg;
   }
 
-  assert(!(SetNZCV && (NumPredicateVectors || NumDataVectors)) &&
-         "SetNZCV not supported with SVE vectors");
   assert(!(NeedsWinCFI && NumPredicateVectors) &&
          "WinCFI can't allocate fractions of an SVE data vector");
 
@@ -6303,6 +6306,12 @@ void llvm::emitFrameOffset(MachineBasicBlock &MBB,
                        Flag, NeedsWinCFI, HasWinCFI, EmitCFAOffset, CFAOffset,
                        FrameReg);
   }
+
+  if (NeedInsertADDS)
+    BuildMI(MBB, MBBI, DL, TII->get(AArch64::ADDSXri), DestReg)
+        .addReg(DestReg)
+        .addImm(0)
+        .addImm(0);
 }
 
 MachineInstr *AArch64InstrInfo::foldMemoryOperandImpl(

llvm/test/CodeGen/AArch64/pr157252.mir

@@ -0,0 +1,103 @@
+# NOTE: Assertions have been autogenerated by utils/update_mir_test_checks.py UTC_ARGS: --version 5
+# RUN: llc -mtriple=aarch64-none-linux-gnu -run-pass prologepilog -o - %s | FileCheck %s
+--- |
+  define void @i(ptr %ad, ptr %0) #0 {
+  entry:
+    ret void
+  }
+  declare i32 @g(...)
+  attributes #0 = { "frame-pointer"="non-leaf" "target-features"="+sve" }
+...
+---
+name:            i
+alignment:       4
+exposesReturnsTwice: false
+legalized:       false
+regBankSelected: false
+selected:        false
+failedISel:      false
+tracksRegLiveness: true
+hasWinCFI:       false
+noPhis:          true
+isSSA:           false
+noVRegs:         true
+hasFakeUses:     false
+callsEHReturn:   false
+callsUnwindInit: false
+hasEHContTarget: false
+hasEHScopes:     false
+hasEHFunclets:   false
+isOutlined:      false
+debugInstrRef:   false
+failsVerification: false
+tracksDebugUserValues: true
+registers:       []
+liveins:         []
+frameInfo:
+  isFrameAddressTaken: false
+  isReturnAddressTaken: false
+  hasStackMap:     false
+  hasPatchPoint:   false
+  stackSize:       0
+  offsetAdjustment: 0
+  maxAlignment:    16
+  adjustsStack:    true
+  hasCalls:        true
+  stackProtector:  ''
+  functionContext: ''
+  maxCallFrameSize: 0
+  cvBytesOfCalleeSavedRegisters: 0
+  hasOpaqueSPAdjustment: false
+  hasVAStart:      false
+  hasMustTailInVarArgFunc: false
+  hasTailCall:     false
+  isCalleeSavedInfoValid: false
+  localFrameSize:  0
+fixedStack:      []
+stack:
+  - { id: 0, name: '', type: default, offset: 0, size: 4, alignment: 4,
+      stack-id: default, callee-saved-register: '', callee-saved-restored: true,
+      local-offset: -4, debug-info-variable: '', debug-info-expression: '',
+      debug-info-location: '' }
+  - { id: 3, name: '', type: spill-slot, offset: 0, size: 16, alignment: 16,
+      stack-id: scalable-vector, callee-saved-register: '', callee-saved-restored: true,
+      debug-info-variable: '', debug-info-expression: '', debug-info-location: '' }
+entry_values:    []
+callSites:       []
+debugValueSubstitutions: []
+constants:       []
+machineFunctionInfo: {}
+body:             |
+  bb.0:
+    ; CHECK-LABEL: name: i
+    ; CHECK: liveins: $d11, $lr, $x20, $x28
+    ; CHECK-NEXT: {{  $}}
+    ; CHECK-NEXT: early-clobber $sp = frame-setup STRDpre killed $d11, $sp, -48 :: (store (s64) into %stack.6)
+    ; CHECK-NEXT: frame-setup STPXi killed $fp, killed $lr, $sp, 2 :: (store (s64) into %stack.5), (store (s64) into %stack.4)
+    ; CHECK-NEXT: frame-setup STPXi killed $x28, killed $x20, $sp, 4 :: (store (s64) into %stack.3), (store (s64) into %stack.2)
+    ; CHECK-NEXT: $fp = frame-setup ADDXri $sp, 16, 0
+    ; CHECK-NEXT: $sp = frame-setup ADDVL_XXI $sp, -1, implicit $vg
+    ; CHECK-NEXT: frame-setup CFI_INSTRUCTION def_cfa $w29, 32
+    ; CHECK-NEXT: frame-setup CFI_INSTRUCTION offset $w20, -8
+    ; CHECK-NEXT: frame-setup CFI_INSTRUCTION offset $w28, -16
+    ; CHECK-NEXT: frame-setup CFI_INSTRUCTION offset $w30, -24
+    ; CHECK-NEXT: frame-setup CFI_INSTRUCTION offset $w29, -32
+    ; CHECK-NEXT: frame-setup CFI_INSTRUCTION offset $b11, -48
+    ; CHECK-NEXT: $w20 = MOVi32imm 1
+    ; CHECK-NEXT: INLINEASM &"", 1 /* sideeffect attdialect */, 12 /* clobber */, implicit-def dead early-clobber $q11
+    ; CHECK-NEXT: BL @g, csr_aarch64_aapcs, implicit-def dead $lr, implicit $sp, implicit-def $sp, implicit-def dead $w0
+    ; CHECK-NEXT: $x8 = ADDXri $sp, 12, 0
+    ; CHECK-NEXT: $x8 = ADDVL_XXI $x8, 1, implicit $vg
+    ; CHECK-NEXT: $x8 = ADDSXri $x8, 0, 0, implicit-def $nzcv
+    ; CHECK-NEXT: $sp = frame-destroy ADDVL_XXI $sp, 1, implicit $vg
+    ; CHECK-NEXT: $x28, $x20 = frame-destroy LDPXi $sp, 4 :: (load (s64) from %stack.3), (load (s64) from %stack.2)
+    ; CHECK-NEXT: $fp, $lr = frame-destroy LDPXi $sp, 2 :: (load (s64) from %stack.5), (load (s64) from %stack.4)
+    ; CHECK-NEXT: early-clobber $sp, $d11 = frame-destroy LDRDpost $sp, 48 :: (load (s64) from %stack.6)
+    ; CHECK-NEXT: RET_ReallyLR
+    $w20 = MOVi32imm 1
+    INLINEASM &"", 1 /* sideeffect attdialect */, 12 /* clobber */, implicit-def dead early-clobber $q11
+    BL @g, csr_aarch64_aapcs, implicit-def dead $lr, implicit $sp, implicit-def $sp, implicit-def dead $w0
+    $x8 = ADDSXri %stack.0, 0, 0, implicit-def $nzcv
+    RET_ReallyLR
+...
+