Skip to content

[AMDGPU][SDAG] Enable ISD::PTRADD for 64-bit AS by default #146076

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
merged 3 commits into from
Oct 2, 2025
Merged

[AMDGPU][SDAG] Enable ISD::PTRADD for 64-bit AS by default #146076

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
3 commits
Select commit Hold shift + click to select a range
5f40d5c
[AMDGPU][SDAG] Enable ISD::PTRADD for 64-bit AS by default
ritter-x2a Jun 27, 2025
d99cb05
Update checks for updated test loop-prefetch-data.ll
ritter-x2a Aug 6, 2025
5f1afbb
Update checks for new test no-folding-imm-to-inst-with-fi.ll
ritter-x2a Aug 7, 2025
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
10 changes: 1 addition & 9 deletions llvm/lib/Target/AMDGPU/SIISelLowering.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -64,14 +64,6 @@ static cl::opt<bool> UseDivergentRegisterIndexing(
cl::desc("Use indirect register addressing for divergent indexes"),
cl::init(false));

// TODO: This option should be removed once we switch to always using PTRADD in
// the SelectionDAG.
static cl::opt<bool> UseSelectionDAGPTRADD(
"amdgpu-use-sdag-ptradd", cl::Hidden,
cl::desc("Generate ISD::PTRADD nodes for 64-bit pointer arithmetic in the "
"SelectionDAG ISel"),
cl::init(false));

static bool denormalModeIsFlushAllF32(const MachineFunction &MF) {
const SIMachineFunctionInfo *Info = MF.getInfo<SIMachineFunctionInfo>();
return Info->getMode().FP32Denormals == DenormalMode::getPreserveSign();
Expand Down Expand Up @@ -11466,7 +11458,7 @@ static bool isNoUnsignedWrap(SDValue Addr) {

bool SITargetLowering::shouldPreservePtrArith(const Function &F,
EVT PtrVT) const {
return UseSelectionDAGPTRADD && PtrVT == MVT::i64;
return PtrVT == MVT::i64;
}

bool SITargetLowering::canTransformPtrArithOutOfBounds(const Function &F,
Expand Down
331 changes: 192 additions & 139 deletions llvm/test/CodeGen/AMDGPU/identical-subrange-spill-infloop.ll
View file
Open in desktop

Large diffs are not rendered by default.

14 changes: 7 additions & 7 deletions llvm/test/CodeGen/AMDGPU/infer-addrspace-flat-atomic.ll
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -11,8 +11,8 @@ define protected amdgpu_kernel void @InferNothing(i32 %a, ptr %b, double %c) {
; CHECK-NEXT: v_mov_b32_e32 v2, s2
; CHECK-NEXT: v_mov_b32_e32 v3, s3
; CHECK-NEXT: s_lshl_b64 s[2:3], s[6:7], 3
; CHECK-NEXT: s_add_u32 s0, s2, s0
; CHECK-NEXT: s_addc_u32 s1, s3, s1
; CHECK-NEXT: s_add_u32 s0, s0, s2
; CHECK-NEXT: s_addc_u32 s1, s1, s3
; CHECK-NEXT: v_mov_b32_e32 v1, s1
; CHECK-NEXT: v_add_co_u32_e64 v0, vcc, -8, s0
; CHECK-NEXT: v_addc_co_u32_e32 v1, vcc, -1, v1, vcc
Expand Down Expand Up @@ -69,13 +69,13 @@ define protected amdgpu_kernel void @InferMixed(i32 %a, ptr addrspace(1) %b, dou
; CHECK-NEXT: s_lshl_b64 s[2:3], s[6:7], 3
; CHECK-NEXT: s_add_u32 s0, s0, s2
; CHECK-NEXT: s_addc_u32 s1, s1, s3
; CHECK-NEXT: s_add_u32 s0, s0, -8
; CHECK-NEXT: s_addc_u32 s1, s1, -1
; CHECK-NEXT: flat_atomic_add_f64 v[0:1], v[2:3]
; CHECK-NEXT: s_waitcnt vmcnt(0) lgkmcnt(0)
; CHECK-NEXT: buffer_wbinvl1_vol
; CHECK-NEXT: v_mov_b32_e32 v1, s1
; CHECK-NEXT: v_add_co_u32_e64 v0, vcc, -7, s0
; CHECK-NEXT: v_addc_co_u32_e32 v1, vcc, -1, v1, vcc
; CHECK-NEXT: flat_atomic_add_f64 v[0:1], v[2:3]
; CHECK-NEXT: v_pk_mov_b32 v[0:1], s[0:1], s[0:1] op_sel:[0,1]
; CHECK-NEXT: flat_atomic_add_f64 v[0:1], v[2:3] offset:1
; CHECK-NEXT: s_waitcnt vmcnt(0) lgkmcnt(0)
; CHECK-NEXT: buffer_wbinvl1_vol
; CHECK-NEXT: s_endpgm
Expand Down Expand Up @@ -113,7 +113,7 @@ define protected amdgpu_kernel void @InferPHI(i32 %a, ptr addrspace(1) %b, doubl
; CHECK-NEXT: s_addc_u32 s1, s1, s5
; CHECK-NEXT: s_add_u32 s4, s0, -8
; CHECK-NEXT: s_addc_u32 s5, s1, -1
; CHECK-NEXT: s_cmp_eq_u64 s[0:1], 9
; CHECK-NEXT: s_cmp_eq_u64 s[4:5], 1
; CHECK-NEXT: s_cselect_b64 s[0:1], -1, 0
; CHECK-NEXT: v_cndmask_b32_e64 v0, 0, 1, s[0:1]
; CHECK-NEXT: v_cmp_ne_u32_e64 s[0:1], 1, v0
Expand Down
8 changes: 4 additions & 4 deletions llvm/test/CodeGen/AMDGPU/lds-frame-extern.ll
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -46,8 +46,8 @@ define void @use_extern_normal() #0 {
; CHECK-NEXT: s_ashr_i32 s5, s15, 31
; CHECK-NEXT: v_mov_b32_e32 v0, 0x4048f5c3
; CHECK-NEXT: s_lshl_b64 s[4:5], s[4:5], 2
; CHECK-NEXT: s_add_u32 s4, s4, s6
; CHECK-NEXT: s_addc_u32 s5, s5, s7
; CHECK-NEXT: s_add_u32 s4, s6, s4
; CHECK-NEXT: s_addc_u32 s5, s7, s5
; CHECK-NEXT: s_load_dword s4, s[4:5], 0x0
; CHECK-NEXT: s_waitcnt lgkmcnt(0)
; CHECK-NEXT: v_mov_b32_e32 v1, s4
Expand All @@ -70,8 +70,8 @@ define void @use_extern_overalign() #0 {
; CHECK-NEXT: s_ashr_i32 s5, s15, 31
; CHECK-NEXT: v_mov_b32_e32 v0, 0x42280000
; CHECK-NEXT: s_lshl_b64 s[4:5], s[4:5], 2
; CHECK-NEXT: s_add_u32 s4, s4, s6
; CHECK-NEXT: s_addc_u32 s5, s5, s7
; CHECK-NEXT: s_add_u32 s4, s6, s4
; CHECK-NEXT: s_addc_u32 s5, s7, s5
; CHECK-NEXT: s_load_dword s4, s[4:5], 0x0
; CHECK-NEXT: s_waitcnt lgkmcnt(0)
; CHECK-NEXT: v_mov_b32_e32 v1, s4
Expand Down
16 changes: 8 additions & 8 deletions llvm/test/CodeGen/AMDGPU/loop-prefetch-data.ll
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -400,9 +400,9 @@ define amdgpu_kernel void @copy_flat_divergent(ptr nocapture %d, ptr nocapture r
; GFX12-NEXT: s_delay_alu instid0(VALU_DEP_1) | instskip(SKIP_1) | instid1(VALU_DEP_1)
; GFX12-NEXT: v_lshlrev_b32_e32 v0, 4, v0
; GFX12-NEXT: s_wait_kmcnt 0x0
; GFX12-NEXT: v_add_co_u32 v2, s1, v0, s6
; GFX12-NEXT: v_add_co_u32 v2, s1, s6, v0
; GFX12-NEXT: s_delay_alu instid0(VALU_DEP_1) | instskip(SKIP_1) | instid1(VALU_DEP_3)
; GFX12-NEXT: v_add_co_ci_u32_e64 v3, null, 0, s7, s1
; GFX12-NEXT: v_add_co_ci_u32_e64 v3, null, s7, 0, s1
; GFX12-NEXT: v_add_co_u32 v0, s1, s4, v0
; GFX12-NEXT: v_add_co_u32 v2, vcc_lo, 0xb0, v2
; GFX12-NEXT: s_wait_alu 0xf1ff
Expand Down Expand Up @@ -438,9 +438,9 @@ define amdgpu_kernel void @copy_flat_divergent(ptr nocapture %d, ptr nocapture r
; GFX12-SPREFETCH-NEXT: s_delay_alu instid0(VALU_DEP_1) | instskip(SKIP_1) | instid1(VALU_DEP_1)
; GFX12-SPREFETCH-NEXT: v_lshlrev_b32_e32 v0, 4, v0
; GFX12-SPREFETCH-NEXT: s_wait_kmcnt 0x0
; GFX12-SPREFETCH-NEXT: v_add_co_u32 v2, s1, v0, s6
; GFX12-SPREFETCH-NEXT: v_add_co_u32 v2, s1, s6, v0
; GFX12-SPREFETCH-NEXT: s_delay_alu instid0(VALU_DEP_1) | instskip(SKIP_1) | instid1(VALU_DEP_3)
; GFX12-SPREFETCH-NEXT: v_add_co_ci_u32_e64 v3, null, 0, s7, s1
; GFX12-SPREFETCH-NEXT: v_add_co_ci_u32_e64 v3, null, s7, 0, s1
; GFX12-SPREFETCH-NEXT: v_add_co_u32 v0, s1, s4, v0
; GFX12-SPREFETCH-NEXT: v_add_co_u32 v2, vcc_lo, 0xb0, v2
; GFX12-SPREFETCH-NEXT: s_wait_alu 0xf1ff
Expand Down Expand Up @@ -531,9 +531,9 @@ define amdgpu_kernel void @copy_global_divergent(ptr addrspace(1) nocapture %d,
; GFX12-NEXT: s_delay_alu instid0(VALU_DEP_1) | instskip(SKIP_1) | instid1(VALU_DEP_1)
; GFX12-NEXT: v_lshlrev_b32_e32 v0, 4, v0
; GFX12-NEXT: s_wait_kmcnt 0x0
; GFX12-NEXT: v_add_co_u32 v2, s1, v0, s6
; GFX12-NEXT: v_add_co_u32 v2, s1, s6, v0
; GFX12-NEXT: s_delay_alu instid0(VALU_DEP_1) | instskip(SKIP_1) | instid1(VALU_DEP_3)
; GFX12-NEXT: v_add_co_ci_u32_e64 v3, null, 0, s7, s1
; GFX12-NEXT: v_add_co_ci_u32_e64 v3, null, s7, 0, s1
; GFX12-NEXT: v_add_co_u32 v0, s1, s4, v0
; GFX12-NEXT: v_add_co_u32 v2, vcc_lo, 0xb0, v2
; GFX12-NEXT: s_wait_alu 0xf1ff
Expand Down Expand Up @@ -569,9 +569,9 @@ define amdgpu_kernel void @copy_global_divergent(ptr addrspace(1) nocapture %d,
; GFX12-SPREFETCH-NEXT: s_delay_alu instid0(VALU_DEP_1) | instskip(SKIP_1) | instid1(VALU_DEP_1)
; GFX12-SPREFETCH-NEXT: v_lshlrev_b32_e32 v0, 4, v0
; GFX12-SPREFETCH-NEXT: s_wait_kmcnt 0x0
; GFX12-SPREFETCH-NEXT: v_add_co_u32 v2, s1, v0, s6
; GFX12-SPREFETCH-NEXT: v_add_co_u32 v2, s1, s6, v0
; GFX12-SPREFETCH-NEXT: s_delay_alu instid0(VALU_DEP_1) | instskip(SKIP_1) | instid1(VALU_DEP_3)
; GFX12-SPREFETCH-NEXT: v_add_co_ci_u32_e64 v3, null, 0, s7, s1
; GFX12-SPREFETCH-NEXT: v_add_co_ci_u32_e64 v3, null, s7, 0, s1
; GFX12-SPREFETCH-NEXT: v_add_co_u32 v0, s1, s4, v0
; GFX12-SPREFETCH-NEXT: v_add_co_u32 v2, vcc_lo, 0xb0, v2
; GFX12-SPREFETCH-NEXT: s_wait_alu 0xf1ff
Expand Down
4 changes: 2 additions & 2 deletions llvm/test/CodeGen/AMDGPU/lower-module-lds-via-hybrid.ll
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -84,8 +84,8 @@ define void @f2() {
; GCN-NEXT: s_add_u32 s6, s6, llvm.amdgcn.lds.offset.table@rel32@lo+4
; GCN-NEXT: s_addc_u32 s7, s7, llvm.amdgcn.lds.offset.table@rel32@hi+12
; GCN-NEXT: s_lshl_b64 s[4:5], s[4:5], 2
; GCN-NEXT: s_add_u32 s4, s4, s6
; GCN-NEXT: s_addc_u32 s5, s5, s7
; GCN-NEXT: s_add_u32 s4, s6, s4
; GCN-NEXT: s_addc_u32 s5, s7, s5
; GCN-NEXT: s_load_dword s4, s[4:5], 0x0
; GCN-NEXT: s_waitcnt lgkmcnt(0)
; GCN-NEXT: v_mov_b32_e32 v2, s4
Expand Down
16 changes: 8 additions & 8 deletions llvm/test/CodeGen/AMDGPU/lower-module-lds-via-table.ll
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -49,8 +49,8 @@ define void @f0() {
; GCN-NEXT: s_add_u32 s6, s6, llvm.amdgcn.lds.offset.table@rel32@lo+4
; GCN-NEXT: s_addc_u32 s7, s7, llvm.amdgcn.lds.offset.table@rel32@hi+12
; GCN-NEXT: s_lshl_b64 s[4:5], s[4:5], 4
; GCN-NEXT: s_add_u32 s4, s4, s6
; GCN-NEXT: s_addc_u32 s5, s5, s7
; GCN-NEXT: s_add_u32 s4, s6, s4
; GCN-NEXT: s_addc_u32 s5, s7, s5
; GCN-NEXT: s_load_dword s4, s[4:5], 0x0
; GCN-NEXT: s_waitcnt lgkmcnt(0)
; GCN-NEXT: v_mov_b32_e32 v0, s4
Expand Down Expand Up @@ -90,8 +90,8 @@ define void @f1() {
; GCN-NEXT: s_add_u32 s6, s6, llvm.amdgcn.lds.offset.table@rel32@lo+8
; GCN-NEXT: s_addc_u32 s7, s7, llvm.amdgcn.lds.offset.table@rel32@hi+16
; GCN-NEXT: s_lshl_b64 s[4:5], s[4:5], 4
; GCN-NEXT: s_add_u32 s4, s4, s6
; GCN-NEXT: s_addc_u32 s5, s5, s7
; GCN-NEXT: s_add_u32 s4, s6, s4
; GCN-NEXT: s_addc_u32 s5, s7, s5
; GCN-NEXT: s_load_dword s4, s[4:5], 0x0
; GCN-NEXT: s_waitcnt lgkmcnt(0)
; GCN-NEXT: v_mov_b32_e32 v0, s4
Expand Down Expand Up @@ -131,8 +131,8 @@ define void @f2() {
; GCN-NEXT: s_add_u32 s6, s6, llvm.amdgcn.lds.offset.table@rel32@lo+12
; GCN-NEXT: s_addc_u32 s7, s7, llvm.amdgcn.lds.offset.table@rel32@hi+20
; GCN-NEXT: s_lshl_b64 s[4:5], s[4:5], 4
; GCN-NEXT: s_add_u32 s4, s4, s6
; GCN-NEXT: s_addc_u32 s5, s5, s7
; GCN-NEXT: s_add_u32 s4, s6, s4
; GCN-NEXT: s_addc_u32 s5, s7, s5
; GCN-NEXT: s_load_dword s4, s[4:5], 0x0
; GCN-NEXT: s_waitcnt lgkmcnt(0)
; GCN-NEXT: v_mov_b32_e32 v2, s4
Expand Down Expand Up @@ -172,8 +172,8 @@ define void @f3() {
; GCN-NEXT: s_add_u32 s6, s6, llvm.amdgcn.lds.offset.table@rel32@lo+16
; GCN-NEXT: s_addc_u32 s7, s7, llvm.amdgcn.lds.offset.table@rel32@hi+24
; GCN-NEXT: s_lshl_b64 s[4:5], s[4:5], 4
; GCN-NEXT: s_add_u32 s4, s4, s6
; GCN-NEXT: s_addc_u32 s5, s5, s7
; GCN-NEXT: s_add_u32 s4, s6, s4
; GCN-NEXT: s_addc_u32 s5, s7, s5
; GCN-NEXT: s_load_dword s4, s[4:5], 0x0
; GCN-NEXT: s_waitcnt lgkmcnt(0)
; GCN-NEXT: v_mov_b32_e32 v0, s4
Expand Down
22 changes: 10 additions & 12 deletions llvm/test/CodeGen/AMDGPU/match-perm-extract-vector-elt-bug.ll
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -13,9 +13,9 @@ define amdgpu_kernel void @test(ptr addrspace(1) %src, ptr addrspace(1) %dst) {
; GFX9-NEXT: s_and_b32 s4, s4, 0xffff
; GFX9-NEXT: s_mul_i32 s14, s14, s4
; GFX9-NEXT: s_add_i32 s5, s5, s14
; GFX9-NEXT: v_add_u32_e32 v0, s5, v0
; GFX9-NEXT: v_ashrrev_i32_e32 v1, 31, v0
; GFX9-NEXT: v_lshlrev_b64 v[4:5], 4, v[0:1]
; GFX9-NEXT: v_add_u32_e32 v1, s5, v0
; GFX9-NEXT: v_mov_b32_e32 v0, 0
; GFX9-NEXT: v_ashrrev_i64 v[4:5], 28, v[0:1]
; GFX9-NEXT: v_mov_b32_e32 v1, s1
; GFX9-NEXT: v_add_co_u32_e32 v0, vcc, s0, v4
; GFX9-NEXT: v_addc_co_u32_e32 v1, vcc, v1, v5, vcc
Expand All @@ -37,12 +37,12 @@ define amdgpu_kernel void @test(ptr addrspace(1) %src, ptr addrspace(1) %dst) {
; GFX10-NEXT: s_load_dword s4, s[8:9], 0x1c
; GFX10-NEXT: s_load_dword s5, s[8:9], 0x38
; GFX10-NEXT: s_load_dwordx4 s[0:3], s[8:9], 0x0
; GFX10-NEXT: v_mov_b32_e32 v1, 0
; GFX10-NEXT: s_waitcnt lgkmcnt(0)
; GFX10-NEXT: s_and_b32 s4, s4, 0xffff
; GFX10-NEXT: s_mul_i32 s14, s14, s4
; GFX10-NEXT: v_add3_u32 v0, s5, s14, v0
; GFX10-NEXT: v_ashrrev_i32_e32 v1, 31, v0
; GFX10-NEXT: v_lshlrev_b64 v[4:5], 4, v[0:1]
; GFX10-NEXT: v_add3_u32 v2, s5, s14, v0
; GFX10-NEXT: v_ashrrev_i64 v[4:5], 28, v[1:2]
; GFX10-NEXT: v_add_co_u32 v0, vcc_lo, s0, v4
; GFX10-NEXT: v_add_co_ci_u32_e64 v1, null, s1, v5, vcc_lo
; GFX10-NEXT: v_add_co_u32 v4, vcc_lo, s2, v4
Expand All @@ -62,21 +62,19 @@ define amdgpu_kernel void @test(ptr addrspace(1) %src, ptr addrspace(1) %dst) {
; GFX11-NEXT: s_load_b32 s6, s[4:5], 0x1c
; GFX11-NEXT: s_load_b32 s7, s[4:5], 0x38
; GFX11-NEXT: s_load_b128 s[0:3], s[4:5], 0x0
; GFX11-NEXT: v_and_b32_e32 v0, 0x3ff, v0
; GFX11-NEXT: v_dual_mov_b32 v0, 0 :: v_dual_and_b32 v1, 0x3ff, v0
; GFX11-NEXT: s_waitcnt lgkmcnt(0)
; GFX11-NEXT: s_and_b32 s4, s6, 0xffff
; GFX11-NEXT: s_delay_alu instid0(SALU_CYCLE_1)
; GFX11-NEXT: s_mul_i32 s13, s13, s4
; GFX11-NEXT: s_delay_alu instid0(VALU_DEP_1) | instid1(SALU_CYCLE_1)
; GFX11-NEXT: v_add3_u32 v0, s7, s13, v0
; GFX11-NEXT: s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1)
; GFX11-NEXT: v_ashrrev_i32_e32 v1, 31, v0
; GFX11-NEXT: v_lshlrev_b64 v[4:5], 4, v[0:1]
; GFX11-NEXT: v_add3_u32 v1, s7, s13, v1
; GFX11-NEXT: s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1)
; GFX11-NEXT: v_ashrrev_i64 v[4:5], 28, v[0:1]
; GFX11-NEXT: v_add_co_u32 v0, vcc_lo, s0, v4
; GFX11-NEXT: s_delay_alu instid0(VALU_DEP_1) | instskip(SKIP_1) | instid1(VALU_DEP_1)
; GFX11-NEXT: v_add_co_ci_u32_e64 v1, null, s1, v5, vcc_lo
; GFX11-NEXT: v_add_co_u32 v4, vcc_lo, s2, v4
; GFX11-NEXT: s_delay_alu instid0(VALU_DEP_1)
; GFX11-NEXT: v_add_co_ci_u32_e64 v5, null, s3, v5, vcc_lo
; GFX11-NEXT: global_load_b128 v[0:3], v[0:1], off
; GFX11-NEXT: s_waitcnt vmcnt(0)
Expand Down
16 changes: 8 additions & 8 deletions llvm/test/CodeGen/AMDGPU/memmove-var-size.ll
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -388,8 +388,8 @@ define void @memmove_p0_p3(ptr addrspace(0) align 1 %dst, ptr addrspace(3) align
; CHECK-NEXT: s_and_saveexec_b32 s7, s4
; CHECK-NEXT: s_cbranch_execz .LBB2_13
; CHECK-NEXT: ; %bb.11: ; %memmove_bwd_residual_loop.preheader
; CHECK-NEXT: v_add_co_u32 v9, s4, v3, v0
; CHECK-NEXT: v_add_co_ci_u32_e64 v10, null, v4, v1, s4
; CHECK-NEXT: v_add_co_u32 v9, s4, v0, v3
; CHECK-NEXT: v_add_co_ci_u32_e64 v10, null, v1, v4, s4
; CHECK-NEXT: v_add3_u32 v4, v3, v2, -1
; CHECK-NEXT: v_add_co_u32 v9, s4, v9, -1
; CHECK-NEXT: v_add_co_ci_u32_e64 v10, null, -1, v10, s4
Expand Down Expand Up @@ -684,8 +684,8 @@ define void @memmove_p0_p5(ptr addrspace(0) align 1 %dst, ptr addrspace(5) align
; CHECK-NEXT: s_and_saveexec_b32 s7, s4
; CHECK-NEXT: s_cbranch_execz .LBB4_13
; CHECK-NEXT: ; %bb.11: ; %memmove_bwd_residual_loop.preheader
; CHECK-NEXT: v_add_co_u32 v9, s4, v3, v0
; CHECK-NEXT: v_add_co_ci_u32_e64 v10, null, v4, v1, s4
; CHECK-NEXT: v_add_co_u32 v9, s4, v0, v3
; CHECK-NEXT: v_add_co_ci_u32_e64 v10, null, v1, v4, s4
; CHECK-NEXT: v_add3_u32 v4, v3, v2, -1
; CHECK-NEXT: v_add_co_u32 v9, s4, v9, -1
; CHECK-NEXT: v_add_co_ci_u32_e64 v10, null, -1, v10, s4
Expand Down Expand Up @@ -1411,8 +1411,8 @@ define void @memmove_p3_p0(ptr addrspace(3) align 1 %dst, ptr addrspace(0) align
; CHECK-NEXT: s_and_saveexec_b32 s7, s4
; CHECK-NEXT: s_cbranch_execz .LBB10_13
; CHECK-NEXT: ; %bb.11: ; %memmove_bwd_residual_loop.preheader
; CHECK-NEXT: v_add_co_u32 v9, s4, v3, v1
; CHECK-NEXT: v_add_co_ci_u32_e64 v10, null, v4, v2, s4
; CHECK-NEXT: v_add_co_u32 v9, s4, v1, v3
; CHECK-NEXT: v_add_co_ci_u32_e64 v10, null, v2, v4, s4
; CHECK-NEXT: v_add3_u32 v4, v3, v0, -1
; CHECK-NEXT: v_add_co_u32 v9, s4, v9, -1
; CHECK-NEXT: v_add_co_ci_u32_e64 v10, null, -1, v10, s4
Expand Down Expand Up @@ -1889,8 +1889,8 @@ define void @memmove_p5_p0(ptr addrspace(5) align 1 %dst, ptr addrspace(0) align
; CHECK-NEXT: s_and_saveexec_b32 s7, s4
; CHECK-NEXT: s_cbranch_execz .LBB15_13
; CHECK-NEXT: ; %bb.11: ; %memmove_bwd_residual_loop.preheader
; CHECK-NEXT: v_add_co_u32 v9, s4, v3, v1
; CHECK-NEXT: v_add_co_ci_u32_e64 v10, null, v4, v2, s4
; CHECK-NEXT: v_add_co_u32 v9, s4, v1, v3
; CHECK-NEXT: v_add_co_ci_u32_e64 v10, null, v2, v4, s4
; CHECK-NEXT: v_add3_u32 v4, v3, v0, -1
; CHECK-NEXT: v_add_co_u32 v9, s4, v9, -1
; CHECK-NEXT: v_add_co_ci_u32_e64 v10, null, -1, v10, s4
Expand Down
Loading