Skip to content

[CGP] Eliminate noop bitcasts #146961

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

Jump to bottom
Open
wants to merge 1 commit into
base: main
Choose a base branch
from
Open

[CGP] Eliminate noop bitcasts #146961

wants to merge 1 commit into from
+260 −316

Conversation

preames
Copy link
Collaborator

@preames preames commented Jul 3, 2025

The change itself is very straight forward, it's mostly just updating tests.

Somewhat surprisingly, this does cause some minor codegen differences in a few tests.

@preames
[CGP] Eliminate noop bitcasts
902914a 
The change itself is very straight forward, it's mostly just updating tests.

Somewhat surprisingly, this does cause some minor codegen differences in
a few tests.
@preames preames requested review from nikic, RKSimon, e-kud and topperc July 3, 2025 21:09
@llvmbot
Copy link
Member

llvmbot commented Jul 3, 2025 •
edited
Loading

@llvm/pr-subscribers-backend-x86
@llvm/pr-subscribers-backend-aarch64

@llvm/pr-subscribers-llvm-globalisel

Author: Philip Reames (preames)

Changes

The change itself is very straight forward, it's mostly just updating tests.

Somewhat surprisingly, this does cause some minor codegen differences in a few tests.

Patch is 89.99 KiB, truncated to 20.00 KiB below, full version: https://github.com/llvm/llvm-project/pull/146961.diff

24 Files Affected:

  • (modified) llvm/lib/CodeGen/CodeGenPrepare.cpp (+7)
  • (modified) llvm/test/CodeGen/AArch64/memcpy-scoped-aa.ll (+10-10)
  • (modified) llvm/test/CodeGen/AArch64/merge-scoped-aa-store.ll (+2-2)
  • (modified) llvm/test/CodeGen/AMDGPU/GlobalISel/irtranslator-call-return-values.ll (+2-2)
  • (modified) llvm/test/CodeGen/AMDGPU/GlobalISel/irtranslator-call.ll (+2-2)
  • (modified) llvm/test/CodeGen/AMDGPU/GlobalISel/irtranslator-indirect-call.ll (+1-1)
  • (modified) llvm/test/CodeGen/AMDGPU/GlobalISel/irtranslator-sibling-call.ll (+1-1)
  • (modified) llvm/test/CodeGen/AMDGPU/branch-folding-implicit-def-subreg.ll (+2-2)
  • (modified) llvm/test/CodeGen/AMDGPU/partial-regcopy-and-spill-missed-at-regalloc.ll (+4-4)
  • (modified) llvm/test/CodeGen/AMDGPU/spill-vector-superclass.ll (+1-1)
  • (modified) llvm/test/CodeGen/SystemZ/isel-debug.ll (+1-1)
  • (modified) llvm/test/CodeGen/Thumb2/active_lane_mask.ll (+2-3)
  • (modified) llvm/test/CodeGen/X86/GlobalISel/x86_64-irtranslator-struct-return.ll (+27-27)
  • (modified) llvm/test/CodeGen/X86/memcpy-scoped-aa.ll (+18-18)
  • (modified) llvm/test/CodeGen/X86/merge-store-partially-alias-loads.ll (+1-1)
  • (modified) llvm/test/CodeGen/X86/pr44140.ll (+6-6)
  • (modified) llvm/test/CodeGen/X86/sse-intrinsics-fast-isel.ll (+120-160)
  • (modified) llvm/test/CodeGen/X86/statepoint-cmp-sunk-past-statepoint.ll (+4-4)
  • (modified) llvm/test/DebugInfo/Generic/assignment-tracking/codegenprepare/sunk-addr.ll (+1-2)
  • (modified) llvm/test/DebugInfo/X86/bitcast-di.ll (+2-33)
  • (modified) llvm/test/Transforms/CodeGenPrepare/X86/masked-gather-struct-gep.ll (+2-3)
  • (modified) llvm/test/Transforms/CodeGenPrepare/X86/sink-addr-recreate.ll (+4-7)
  • (modified) llvm/test/Transforms/CodeGenPrepare/X86/sink-addrmode-base.ll (+39-24)
  • (modified) llvm/test/Transforms/CodeGenPrepare/X86/sink-addrmode-two-phi.ll (+1-2)
diff --git a/llvm/lib/CodeGen/CodeGenPrepare.cpp b/llvm/lib/CodeGen/CodeGenPrepare.cpp
index 9bbb89e37865d..bb397c1c35bf8 100644
--- a/llvm/lib/CodeGen/CodeGenPrepare.cpp
+++ b/llvm/lib/CodeGen/CodeGenPrepare.cpp
@@ -8720,6 +8720,13 @@ bool CodeGenPrepare::optimizeInst(Instruction *I, ModifyDT &ModifiedDT) {
     if (isa<Constant>(CI->getOperand(0)))
       return AnyChange;
 
+    // Remove noop bitcasts
+    if (isa<BitCastInst>(I) && I->getType() == I->getOperand(0)->getType()) {
+      replaceAllUsesWith(I, I->getOperand(0), FreshBBs, IsHugeFunc);
+      I->eraseFromParent();
+      return true;
+    }
+
     if (OptimizeNoopCopyExpression(CI, *TLI, *DL))
       return true;
 
diff --git a/llvm/test/CodeGen/AArch64/memcpy-scoped-aa.ll b/llvm/test/CodeGen/AArch64/memcpy-scoped-aa.ll
index 7dde168024278..f77d300adb2c7 100644
--- a/llvm/test/CodeGen/AArch64/memcpy-scoped-aa.ll
+++ b/llvm/test/CodeGen/AArch64/memcpy-scoped-aa.ll
@@ -9,8 +9,8 @@
 ; MIR-DAG: ![[SET1:[0-9]+]] = !{![[SCOPE1]]}
 
 ; MIR-LABEL: name: test_memcpy
-; MIR:      %2:fpr128 = LDRQui %0, 1 :: (load (s128) from %ir.p1, align 4, !alias.scope ![[SET0]], !noalias ![[SET1]])
-; MIR-NEXT: STRQui killed %2, %0, 0 :: (store (s128) into %ir.p0, align 4, !alias.scope ![[SET0]], !noalias ![[SET1]])
+; MIR:      %2:fpr128 = LDRQui %0, 1 :: (load (s128) from %ir.add.ptr, align 4, !alias.scope ![[SET0]], !noalias ![[SET1]])
+; MIR-NEXT: STRQui killed %2, %0, 0 :: (store (s128) into %ir.p, align 4, !alias.scope ![[SET0]], !noalias ![[SET1]])
 define i32 @test_memcpy(ptr nocapture %p, ptr nocapture readonly %q) {
 ; CHECK-LABEL: test_memcpy:
 ; CHECK:       // %bb.0:
@@ -32,8 +32,8 @@ define i32 @test_memcpy(ptr nocapture %p, ptr nocapture readonly %q) {
 }
 
 ; MIR-LABEL: name: test_memcpy_inline
-; MIR:      %2:fpr128 = LDRQui %0, 1 :: (load (s128) from %ir.p1, align 4, !alias.scope ![[SET0]], !noalias ![[SET1]])
-; MIR-NEXT: STRQui killed %2, %0, 0 :: (store (s128) into %ir.p0, align 4, !alias.scope ![[SET0]], !noalias ![[SET1]])
+; MIR:      %2:fpr128 = LDRQui %0, 1 :: (load (s128) from %ir.add.ptr, align 4, !alias.scope ![[SET0]], !noalias ![[SET1]])
+; MIR-NEXT: STRQui killed %2, %0, 0 :: (store (s128) into %ir.p, align 4, !alias.scope ![[SET0]], !noalias ![[SET1]])
 define i32 @test_memcpy_inline(ptr nocapture %p, ptr nocapture readonly %q) {
 ; CHECK-LABEL: test_memcpy_inline:
 ; CHECK:       // %bb.0:
@@ -55,8 +55,8 @@ define i32 @test_memcpy_inline(ptr nocapture %p, ptr nocapture readonly %q) {
 }
 
 ; MIR-LABEL: name: test_memmove
-; MIR:      %2:fpr128 = LDRQui %0, 1 :: (load (s128) from %ir.p1, align 4, !alias.scope ![[SET0]], !noalias ![[SET1]])
-; MIR-NEXT: STRQui killed %2, %0, 0 :: (store (s128) into %ir.p0, align 4, !alias.scope ![[SET0]], !noalias ![[SET1]])
+; MIR:      %2:fpr128 = LDRQui %0, 1 :: (load (s128) from %ir.add.ptr, align 4, !alias.scope ![[SET0]], !noalias ![[SET1]])
+; MIR-NEXT: STRQui killed %2, %0, 0 :: (store (s128) into %ir.p, align 4, !alias.scope ![[SET0]], !noalias ![[SET1]])
 define i32 @test_memmove(ptr nocapture %p, ptr nocapture readonly %q) {
 ; CHECK-LABEL: test_memmove:
 ; CHECK:       // %bb.0:
@@ -79,8 +79,8 @@ define i32 @test_memmove(ptr nocapture %p, ptr nocapture readonly %q) {
 
 ; MIR-LABEL: name: test_memset
 ; MIR:      %2:gpr64 = MOVi64imm -6148914691236517206
-; MIR-NEXT: STRXui %2, %0, 1 :: (store (s64) into %ir.p0 + 8, align 4, !alias.scope ![[SET0]], !noalias ![[SET1]])
-; MIR-NEXT: STRXui %2, %0, 0 :: (store (s64) into %ir.p0, align 4, !alias.scope ![[SET0]], !noalias ![[SET1]])
+; MIR-NEXT: STRXui %2, %0, 1 :: (store (s64) into %ir.p + 8, align 4, !alias.scope ![[SET0]], !noalias ![[SET1]])
+; MIR-NEXT: STRXui %2, %0, 0 :: (store (s64) into %ir.p, align 4, !alias.scope ![[SET0]], !noalias ![[SET1]])
 define i32 @test_memset(ptr nocapture %p, ptr nocapture readonly %q) {
 ; CHECK-LABEL: test_memset:
 ; CHECK:       // %bb.0:
@@ -100,8 +100,8 @@ define i32 @test_memset(ptr nocapture %p, ptr nocapture readonly %q) {
 }
 
 ; MIR-LABEL: name: test_mempcpy
-; MIR:      %2:fpr128 = LDRQui %0, 1 :: (load (s128) from %ir.p1, align 1, !alias.scope ![[SET0]], !noalias ![[SET1]])
-; MIR-NEXT: STRQui killed %2, %0, 0 :: (store (s128) into %ir.p0, align 1, !alias.scope ![[SET0]], !noalias ![[SET1]])
+; MIR:      %2:fpr128 = LDRQui %0, 1 :: (load (s128) from %ir.add.ptr, align 1, !alias.scope ![[SET0]], !noalias ![[SET1]])
+; MIR-NEXT: STRQui killed %2, %0, 0 :: (store (s128) into %ir.p, align 1, !alias.scope ![[SET0]], !noalias ![[SET1]])
 define i32 @test_mempcpy(ptr nocapture %p, ptr nocapture readonly %q) {
 ; CHECK-LABEL: test_mempcpy:
 ; CHECK:       // %bb.0:
diff --git a/llvm/test/CodeGen/AArch64/merge-scoped-aa-store.ll b/llvm/test/CodeGen/AArch64/merge-scoped-aa-store.ll
index 6b2a8c5aaf8b8..42fda4b4309af 100644
--- a/llvm/test/CodeGen/AArch64/merge-scoped-aa-store.ll
+++ b/llvm/test/CodeGen/AArch64/merge-scoped-aa-store.ll
@@ -21,7 +21,7 @@ define void @blam0(ptr %g0, ptr %g1) {
   ; MIR-NEXT:   [[COPY:%[0-9]+]]:gpr64common = COPY $x1
   ; MIR-NEXT:   [[COPY1:%[0-9]+]]:gpr64common = COPY $x0
   ; MIR-NEXT:   [[LDRDui:%[0-9]+]]:fpr64 = LDRDui [[COPY1]], 0 :: (load (s64) from %ir.g0, align 4, !alias.scope !0, !noalias !3)
-  ; MIR-NEXT:   STRDui killed [[LDRDui]], [[COPY]], 0 :: (store (s64) into %ir.tmp41, align 4, !alias.scope !3, !noalias !0)
+  ; MIR-NEXT:   STRDui killed [[LDRDui]], [[COPY]], 0 :: (store (s64) into %ir.g1, align 4, !alias.scope !3, !noalias !0)
   ; MIR-NEXT:   RET_ReallyLR
   %tmp4 = getelementptr inbounds <3 x float>, ptr %g1, i64 0, i64 0
   %tmp5 = load <3 x float>, ptr %g0, align 4, !alias.scope !0, !noalias !1
@@ -43,7 +43,7 @@ define void @blam1(ptr %g0, ptr %g1) {
   ; MIR-NEXT:   [[COPY:%[0-9]+]]:gpr64common = COPY $x1
   ; MIR-NEXT:   [[COPY1:%[0-9]+]]:gpr64common = COPY $x0
   ; MIR-NEXT:   [[LDRDui:%[0-9]+]]:fpr64 = LDRDui [[COPY1]], 0 :: (load (s64) from %ir.g0, align 4, !alias.scope !0, !noalias !3)
-  ; MIR-NEXT:   STRDui killed [[LDRDui]], [[COPY]], 0 :: (store (s64) into %ir.tmp41, align 4, !alias.scope !9, !noalias !10)
+  ; MIR-NEXT:   STRDui killed [[LDRDui]], [[COPY]], 0 :: (store (s64) into %ir.g1, align 4, !alias.scope !9, !noalias !10)
   ; MIR-NEXT:   RET_ReallyLR
   %tmp4 = getelementptr inbounds <3 x float>, ptr %g1, i64 0, i64 0
   %tmp5 = load <3 x float>, ptr %g0, align 4, !alias.scope !0, !noalias !1
diff --git a/llvm/test/CodeGen/AMDGPU/GlobalISel/irtranslator-call-return-values.ll b/llvm/test/CodeGen/AMDGPU/GlobalISel/irtranslator-call-return-values.ll
index 96ee15f2eb78b..c7977c339880d 100644
--- a/llvm/test/CodeGen/AMDGPU/GlobalISel/irtranslator-call-return-values.ll
+++ b/llvm/test/CodeGen/AMDGPU/GlobalISel/irtranslator-call-return-values.ll
@@ -84,7 +84,7 @@ define amdgpu_kernel void @test_call_external_i32_func_i32_imm(ptr addrspace(1)
   ; GCN-NEXT:   [[COPY9:%[0-9]+]]:_(p4) = COPY $sgpr8_sgpr9
   ; GCN-NEXT:   [[C:%[0-9]+]]:_(s32) = G_CONSTANT i32 42
   ; GCN-NEXT:   [[INT:%[0-9]+]]:_(p4) = G_INTRINSIC intrinsic(@llvm.amdgcn.kernarg.segment.ptr)
-  ; GCN-NEXT:   [[LOAD:%[0-9]+]]:_(p1) = G_LOAD [[INT]](p4) :: (dereferenceable invariant load (p1) from %ir.out.kernarg.offset1, align 16, addrspace 4)
+  ; GCN-NEXT:   [[LOAD:%[0-9]+]]:_(p1) = G_LOAD [[INT]](p4) :: (dereferenceable invariant load (p1) from %ir.test_call_external_i32_func_i32_imm.kernarg.segment, align 16, addrspace 4)
   ; GCN-NEXT:   ADJCALLSTACKUP 0, 0, implicit-def $scc
   ; GCN-NEXT:   [[GV:%[0-9]+]]:_(p0) = G_GLOBAL_VALUE @external_i32_func_i32
   ; GCN-NEXT:   [[COPY10:%[0-9]+]]:_(p4) = COPY [[COPY8]]
@@ -2957,7 +2957,7 @@ define amdgpu_kernel void @test_call_external_v33i32_func_v33i32_i32(ptr addrspa
   ; GCN-NEXT:   [[COPY9:%[0-9]+]]:_(p4) = COPY $sgpr8_sgpr9
   ; GCN-NEXT:   [[DEF:%[0-9]+]]:_(p1) = G_IMPLICIT_DEF
   ; GCN-NEXT:   [[INT:%[0-9]+]]:_(p4) = G_INTRINSIC intrinsic(@llvm.amdgcn.kernarg.segment.ptr)
-  ; GCN-NEXT:   [[LOAD:%[0-9]+]]:_(p1) = G_LOAD [[INT]](p4) :: (dereferenceable invariant load (p1) from %ir.p.kernarg.offset1, align 16, addrspace 4)
+  ; GCN-NEXT:   [[LOAD:%[0-9]+]]:_(p1) = G_LOAD [[INT]](p4) :: (dereferenceable invariant load (p1) from %ir.test_call_external_v33i32_func_v33i32_i32.kernarg.segment, align 16, addrspace 4)
   ; GCN-NEXT:   [[C:%[0-9]+]]:_(s64) = G_CONSTANT i64 8
   ; GCN-NEXT:   %18:_(p4) = nuw nusw G_PTR_ADD [[INT]], [[C]](s64)
   ; GCN-NEXT:   [[LOAD1:%[0-9]+]]:_(s32) = G_LOAD %18(p4) :: (dereferenceable invariant load (s32) from %ir.idx.kernarg.offset, align 8, addrspace 4)
diff --git a/llvm/test/CodeGen/AMDGPU/GlobalISel/irtranslator-call.ll b/llvm/test/CodeGen/AMDGPU/GlobalISel/irtranslator-call.ll
index 92106d7e1d60d..862093ebf9416 100644
--- a/llvm/test/CodeGen/AMDGPU/GlobalISel/irtranslator-call.ll
+++ b/llvm/test/CodeGen/AMDGPU/GlobalISel/irtranslator-call.ll
@@ -1350,7 +1350,7 @@ define amdgpu_kernel void @test_call_external_void_func_p0_imm(ptr %arg) #0 {
   ; CHECK-NEXT:   [[COPY8:%[0-9]+]]:sgpr_64 = COPY $sgpr4_sgpr5
   ; CHECK-NEXT:   [[COPY9:%[0-9]+]]:_(p4) = COPY $sgpr8_sgpr9
   ; CHECK-NEXT:   [[INT:%[0-9]+]]:_(p4) = G_INTRINSIC intrinsic(@llvm.amdgcn.kernarg.segment.ptr)
-  ; CHECK-NEXT:   [[LOAD:%[0-9]+]]:_(p0) = G_LOAD [[INT]](p4) :: (dereferenceable invariant load (p0) from %ir.arg.kernarg.offset1, align 16, addrspace 4)
+  ; CHECK-NEXT:   [[LOAD:%[0-9]+]]:_(p0) = G_LOAD [[INT]](p4) :: (dereferenceable invariant load (p0) from %ir.test_call_external_void_func_p0_imm.kernarg.segment, align 16, addrspace 4)
   ; CHECK-NEXT:   ADJCALLSTACKUP 0, 0, implicit-def $scc
   ; CHECK-NEXT:   [[GV:%[0-9]+]]:_(p0) = G_GLOBAL_VALUE @external_void_func_p0
   ; CHECK-NEXT:   [[COPY10:%[0-9]+]]:_(p4) = COPY [[COPY8]]
@@ -4654,7 +4654,7 @@ define amdgpu_kernel void @stack_passed_arg_alignment_v32i32_f64(<32 x i32> %val
   ; CHECK-NEXT:   [[COPY8:%[0-9]+]]:sgpr_64 = COPY $sgpr4_sgpr5
   ; CHECK-NEXT:   [[COPY9:%[0-9]+]]:_(p4) = COPY $sgpr8_sgpr9
   ; CHECK-NEXT:   [[INT:%[0-9]+]]:_(p4) = G_INTRINSIC intrinsic(@llvm.amdgcn.kernarg.segment.ptr)
-  ; CHECK-NEXT:   [[LOAD:%[0-9]+]]:_(<32 x s32>) = G_LOAD [[INT]](p4) :: (dereferenceable invariant load (<32 x s32>) from %ir.val.kernarg.offset1, align 16, addrspace 4)
+  ; CHECK-NEXT:   [[LOAD:%[0-9]+]]:_(<32 x s32>) = G_LOAD [[INT]](p4) :: (dereferenceable invariant load (<32 x s32>) from %ir.stack_passed_arg_alignment_v32i32_f64.kernarg.segment, align 16, addrspace 4)
   ; CHECK-NEXT:   [[C:%[0-9]+]]:_(s64) = G_CONSTANT i64 128
   ; CHECK-NEXT:   %18:_(p4) = nuw nusw G_PTR_ADD [[INT]], [[C]](s64)
   ; CHECK-NEXT:   [[LOAD1:%[0-9]+]]:_(s64) = G_LOAD %18(p4) :: (dereferenceable invariant load (s64) from %ir.tmp.kernarg.offset, align 16, addrspace 4)
diff --git a/llvm/test/CodeGen/AMDGPU/GlobalISel/irtranslator-indirect-call.ll b/llvm/test/CodeGen/AMDGPU/GlobalISel/irtranslator-indirect-call.ll
index ac0d5ee78666e..bdf89f785ef99 100644
--- a/llvm/test/CodeGen/AMDGPU/GlobalISel/irtranslator-indirect-call.ll
+++ b/llvm/test/CodeGen/AMDGPU/GlobalISel/irtranslator-indirect-call.ll
@@ -17,7 +17,7 @@ define amdgpu_kernel void @test_indirect_call_sgpr_ptr(ptr %fptr) {
   ; CHECK-NEXT:   [[COPY8:%[0-9]+]]:sgpr_64 = COPY $sgpr4_sgpr5
   ; CHECK-NEXT:   [[COPY9:%[0-9]+]]:_(p4) = COPY $sgpr8_sgpr9
   ; CHECK-NEXT:   [[INT:%[0-9]+]]:_(p4) = G_INTRINSIC intrinsic(@llvm.amdgcn.kernarg.segment.ptr)
-  ; CHECK-NEXT:   [[LOAD:%[0-9]+]]:_(p0) = G_LOAD [[INT]](p4) :: (dereferenceable invariant load (p0) from %ir.fptr.kernarg.offset1, align 16, addrspace 4)
+  ; CHECK-NEXT:   [[LOAD:%[0-9]+]]:_(p0) = G_LOAD [[INT]](p4) :: (dereferenceable invariant load (p0) from %ir.test_indirect_call_sgpr_ptr.kernarg.segment, align 16, addrspace 4)
   ; CHECK-NEXT:   ADJCALLSTACKUP 0, 0, implicit-def $scc
   ; CHECK-NEXT:   [[COPY10:%[0-9]+]]:_(p4) = COPY [[COPY8]]
   ; CHECK-NEXT:   [[COPY11:%[0-9]+]]:_(p4) = COPY [[COPY7]]
diff --git a/llvm/test/CodeGen/AMDGPU/GlobalISel/irtranslator-sibling-call.ll b/llvm/test/CodeGen/AMDGPU/GlobalISel/irtranslator-sibling-call.ll
index ca580d8f29c84..a3f5d927d4b80 100644
--- a/llvm/test/CodeGen/AMDGPU/GlobalISel/irtranslator-sibling-call.ll
+++ b/llvm/test/CodeGen/AMDGPU/GlobalISel/irtranslator-sibling-call.ll
@@ -140,7 +140,7 @@ define amdgpu_kernel void @kernel_call_i32_fastcc_i32_i32_unused_result(i32 %a,
   ; GCN-NEXT:   [[C:%[0-9]+]]:_(s32) = G_CONSTANT i32 0
   ; GCN-NEXT:   [[C1:%[0-9]+]]:_(s32) = G_CONSTANT i32 1
   ; GCN-NEXT:   [[INT:%[0-9]+]]:_(p4) = G_INTRINSIC intrinsic(@llvm.amdgcn.kernarg.segment.ptr)
-  ; GCN-NEXT:   [[LOAD:%[0-9]+]]:_(<2 x s32>) = G_LOAD [[INT]](p4) :: (dereferenceable invariant load (<2 x s32>) from %ir.a.kernarg.offset1, align 16, addrspace 4)
+  ; GCN-NEXT:   [[LOAD:%[0-9]+]]:_(<2 x s32>) = G_LOAD [[INT]](p4) :: (dereferenceable invariant load (<2 x s32>) from %ir.kernel_call_i32_fastcc_i32_i32_unused_result.kernarg.segment, align 16, addrspace 4)
   ; GCN-NEXT:   [[EVEC:%[0-9]+]]:_(s32) = G_EXTRACT_VECTOR_ELT [[LOAD]](<2 x s32>), [[C]](s32)
   ; GCN-NEXT:   [[EVEC1:%[0-9]+]]:_(s32) = G_EXTRACT_VECTOR_ELT [[LOAD]](<2 x s32>), [[C1]](s32)
   ; GCN-NEXT:   ADJCALLSTACKUP 0, 0, implicit-def $scc
diff --git a/llvm/test/CodeGen/AMDGPU/branch-folding-implicit-def-subreg.ll b/llvm/test/CodeGen/AMDGPU/branch-folding-implicit-def-subreg.ll
index ae90cfb631e8d..412e3569c0d05 100644
--- a/llvm/test/CodeGen/AMDGPU/branch-folding-implicit-def-subreg.ll
+++ b/llvm/test/CodeGen/AMDGPU/branch-folding-implicit-def-subreg.ll
@@ -16,8 +16,8 @@ define amdgpu_kernel void @f1(ptr addrspace(1) %arg, ptr addrspace(1) %arg1, i64
   ; GFX90A-NEXT:   renamable $sgpr17 = S_LOAD_DWORD_IMM renamable $sgpr8_sgpr9, 24, 0 :: (dereferenceable invariant load (s32) from %ir.arg4.kernarg.offset.align.down, align 8, addrspace 4)
   ; GFX90A-NEXT:   early-clobber renamable $sgpr20_sgpr21_sgpr22_sgpr23 = S_LOAD_DWORDX4_IMM_ec renamable $sgpr8_sgpr9, 24, 0 :: (dereferenceable invariant load (s128) from %ir.arg6.kernarg.offset.align.down, align 8, addrspace 4)
   ; GFX90A-NEXT:   renamable $sgpr33 = S_LOAD_DWORD_IMM renamable $sgpr8_sgpr9, 40, 0 :: (dereferenceable invariant load (s32) from %ir.arg6.kernarg.offset.align.down + 16, align 8, addrspace 4)
-  ; GFX90A-NEXT:   renamable $sgpr24_sgpr25_sgpr26_sgpr27 = S_LOAD_DWORDX4_IMM renamable $sgpr8_sgpr9, 0, 0 :: (dereferenceable invariant load (s128) from %ir.arg.kernarg.offset1, addrspace 4)
-  ; GFX90A-NEXT:   renamable $sgpr46_sgpr47 = S_LOAD_DWORDX2_IMM renamable $sgpr8_sgpr9, 16, 0 :: (dereferenceable invariant load (s64) from %ir.arg.kernarg.offset1 + 16, align 16, addrspace 4)
+  ; GFX90A-NEXT:   renamable $sgpr24_sgpr25_sgpr26_sgpr27 = S_LOAD_DWORDX4_IMM renamable $sgpr8_sgpr9, 0, 0 :: (dereferenceable invariant load (s128) from %ir.f1.kernarg.segment, addrspace 4)
+  ; GFX90A-NEXT:   renamable $sgpr46_sgpr47 = S_LOAD_DWORDX2_IMM renamable $sgpr8_sgpr9, 16, 0 :: (dereferenceable invariant load (s64) from %ir.f1.kernarg.segment + 16, align 16, addrspace 4)
   ; GFX90A-NEXT:   S_BITCMP1_B32 renamable $sgpr17, 0, implicit-def $scc
   ; GFX90A-NEXT:   renamable $sgpr12_sgpr13 = S_CSELECT_B64 -1, 0, implicit killed $scc
   ; GFX90A-NEXT:   renamable $sgpr34_sgpr35 = S_MOV_B64 -1
diff --git a/llvm/test/CodeGen/AMDGPU/partial-regcopy-and-spill-missed-at-regalloc.ll b/llvm/test/CodeGen/AMDGPU/partial-regcopy-and-spill-missed-at-regalloc.ll
index 663fd98b46bf7..03d3724c232fa 100644
--- a/llvm/test/CodeGen/AMDGPU/partial-regcopy-and-spill-missed-at-regalloc.ll
+++ b/llvm/test/CodeGen/AMDGPU/partial-regcopy-and-spill-missed-at-regalloc.ll
@@ -15,7 +15,7 @@ define amdgpu_kernel void @partial_copy(<4 x i32> %arg) #0 {
   ; REGALLOC-GFX908-NEXT:   INLINEASM &"; def $0", 1 /* sideeffect attdialect */, 6094858 /* regdef:VReg_128 */, def %7
   ; REGALLOC-GFX908-NEXT:   INLINEASM &"; def $0", 1 /* sideeffect attdialect */, 3538954 /* regdef:VReg_64 */, def %8
   ; REGALLOC-GFX908-NEXT:   GLOBAL_STORE_DWORDX4 undef %15:vreg_64, %7, 0, 0, implicit $exec :: (volatile store (s128) into `ptr addrspace(1) poison`, addrspace 1)
-  ; REGALLOC-GFX908-NEXT:   renamable $sgpr0_sgpr1_sgpr2_sgpr3 = S_LOAD_DWORDX4_IMM killed renamable $sgpr4_sgpr5, 0, 0 :: (dereferenceable invariant load (s128) from %ir.arg.kernarg.offset1, addrspace 4)
+  ; REGALLOC-GFX908-NEXT:   renamable $sgpr0_sgpr1_sgpr2_sgpr3 = S_LOAD_DWORDX4_IMM killed renamable $sgpr4_sgpr5, 0, 0 :: (dereferenceable invariant load (s128) from %ir.partial_copy.kernarg.segment, addrspace 4)
   ; REGALLOC-GFX908-NEXT:   [[COPY:%[0-9]+]]:areg_128 = COPY killed renamable $sgpr0_sgpr1_sgpr2_sgpr3
   ; REGALLOC-GFX908-NEXT:   [[V_MOV_B32_e32_:%[0-9]+]]:vgpr_32 = V_MOV_B32_e32 1, implicit $exec
   ; REGALLOC-GFX908-NEXT:   [[V_MOV_B32_e32_1:%[0-9]+]]:vgpr_32 = V_MOV_B32_e32 2, implicit $exec
@@ -40,7 +40,7 @@ define amdgpu_kernel void @partial_copy(<4 x i32> %arg) #0 {
   ; PEI-GFX908-NEXT:   $agpr4 = V_ACCVGPR_WRITE_B32_e64 killed $vgpr1, implicit $exec, implicit killed $vgpr0_vgpr1
   ; PEI-GFX908-NEXT:   renamable $vgpr0_vgpr1_vgpr2_vgpr3 = COPY killed renamable $agpr0_agpr1_agpr2_agpr3, implicit $exec
   ; PEI-GFX908-NEXT:   GLOBAL_STORE_DWORDX4 undef renamable $vgpr0_vgpr1, killed renamable $vgpr0_vgpr1_vgpr2_vgpr3, 0, 0, implicit $exec :: (volatile store (s128) into `ptr addrspace(1) poison`, addrspace 1)
-  ; PEI-GFX908-NEXT:   renamable $sgpr0_sgpr1_sgpr2_sgpr3 = S_LOAD_DWORDX4_IMM killed renamable $sgpr4_sgpr5, 0, 0 :: (dereferenceable invariant load (s128) from %ir.arg.kernarg.offset1, addrspace 4)
+  ; PEI-GFX908-NEXT:   renamable $sgpr0_sgpr1_sgpr2_sgpr3 = S_LOAD_DWORDX4_IMM killed renamable $sgpr4_sgpr5, 0, 0 :: (dereferenceable invariant load (s128) from %ir.partial_copy.kernarg.segment, addrspace 4)
   ; PEI-GFX908-NEXT:   renamable $agpr0_agpr1_agpr2_agpr3 = COPY killed renamable $sgpr0_sgpr1_sgpr2_sgpr3, implicit $exec
   ; PEI-GFX908-NEXT:   renamable $vgpr0 = V_MOV_B32_e32 1, implicit $exec
   ; PEI-GFX908-NEXT:   renamable $vgpr1 = V_MOV_B32_e32 2, implicit $exec
@@ -60,7 +60,7 @@ define amdgpu_kernel void @partial_copy(<4 x i32> %arg) #0 {
   ; REGALLOC-GFX90A-NEXT:   INLINEASM &"; def $0", 1 /* sideeffect attdialect */, 6422538 /* regdef:VReg_128_Align2 */, def %7
   ; REGALLOC-GFX90A-NEXT:   INLINEASM &"; def $0", 1 /* sideeffect attdialect */, 3866634 /* regdef:VReg_64_Align2 */, def %8
   ; REGALLOC-GFX90A-NEXT:   GLOBAL_STORE_DWORDX4 undef %15:vreg_64_align2, %7, 0, 0, implicit $exec :: (volatile store (s128) into `ptr addrspace(1) poison`, addrspace 1)
-  ; REGALLOC-GFX90A-NEXT:   renamable $sgpr0_sgpr1_sgpr2_sgpr3 = S_LOAD_DWORDX4_IMM killed renamable $sgpr4_sgpr5, 0, 0 :: (dereferenceable invariant load (s128) from %ir.arg.kernarg.offset1, addrspace 4)
+  ; REGALLOC-GFX90A-NEXT:   renamable $sgpr0_sgpr1_sgpr2_sgpr3 = S_LOAD_DWORDX4_IMM killed renamable $sgpr4_sgpr5, 0, 0 :: (dereferenceable invariant load (s128) from %ir.partial_copy.kernarg.segment, addrspace 4)
   ; REGALLOC-GFX90A-NEXT:   [[COPY:%[0-9]+]]:areg_128_align2 = COPY killed renamable $sgpr0_sgpr1_sgpr2_sgpr3
   ; REGALLOC-GFX90A-NEXT:   [[V_MOV_B32_e32_:%[0-9]+]]:vgpr_32 = V_MOV_B32_e32 1, implicit $exec
   ; REGALLOC-GFX90A-NEXT:   [[V_MOV_B32_e32_1:%[0-9]+]]:vgpr_32 = V_MOV_B32_e32 2, implicit $exec
@@ -83,7 +83,7 @@ define amdgpu_kernel void @partial_copy(<4 x i32> %arg) #0 {
   ; PEI-GFX90A-NEXT:   BUFFER_STORE_DWORD_OFFSET killed $vgpr0, $sgpr12_sgpr13_sgpr14_sgpr15, 0, 0, 0, 0, implicit $exec, implicit-def $vgpr0_vgpr1, implicit $vgpr0_vgpr1 :: (store (s32) into %stack.0, addrspace 5)
   ; PEI-GFX90A-NEXT:   $agpr4 = V_ACCVGPR_WRITE_B32_e64 killed $vgpr1, implicit $exec, implicit killed $vgpr0_vgpr1
   ; PEI-GFX90A-NEXT:   GLOBAL_STORE_DWORDX4 undef renamable $vgpr0_vgpr1, killed renamable $agpr0_agpr1_agpr2_agpr3, 0, 0, implicit $exec :: (volatile store (s128) into `ptr addrspace(1) poison`, addrspace 1)
-  ; PEI-GFX90A-NEXT:   renamable $sgpr0_sgpr1_sgpr2_sgpr3 = S_LOAD_DWORDX4_IMM killed renamable $sgpr4_sgpr5, 0, 0 :: (dereferenceable invariant load (s128) from %ir.arg.kernarg.offset1, addrspace 4)
+  ; PEI-GFX90A-NEXT:   renamable $sgpr0_sgpr1_sgpr2_sgpr3 = S_LOAD_DWORDX4_IMM killed renamable $sgpr4_sgpr5, 0, 0 :: (dereferenceable invariant load (s128) from %ir.partial_copy.kernarg.segment, addrspace 4)
   ; PEI-GFX90A-NEXT:   renamable $agpr0_agpr1_agpr2_agpr3 = COPY killed renamable $sgpr0_sgpr1_sgpr2_sgpr3, implicit $exec
   ; PEI-GFX90A-NEXT:   renamable $vgpr0 = V_MOV_B32_e32 1, implicit $exec
   ; PEI-GFX90A-NEXT:   renamable $vgpr1 = V_MOV_B32_e32 2, implicit $exec
diff ...
[truncated]

@llvmbot
Copy link
Member

llvmbot commented Jul 3, 2025

@llvm/pr-subscribers-debuginfo

Author: Philip Reames (preames)

Changes

The change itself is very straight forward, it's mostly just updating tests.

Somewhat surprisingly, this does cause some minor codegen differences in a few tests.

Patch is 89.99 KiB, truncated to 20.00 KiB below, full version: https://github.com/llvm/llvm-project/pull/146961.diff

24 Files Affected:

  • (modified) llvm/lib/CodeGen/CodeGenPrepare.cpp (+7)
  • (modified) llvm/test/CodeGen/AArch64/memcpy-scoped-aa.ll (+10-10)
  • (modified) llvm/test/CodeGen/AArch64/merge-scoped-aa-store.ll (+2-2)
  • (modified) llvm/test/CodeGen/AMDGPU/GlobalISel/irtranslator-call-return-values.ll (+2-2)
  • (modified) llvm/test/CodeGen/AMDGPU/GlobalISel/irtranslator-call.ll (+2-2)
  • (modified) llvm/test/CodeGen/AMDGPU/GlobalISel/irtranslator-indirect-call.ll (+1-1)
  • (modified) llvm/test/CodeGen/AMDGPU/GlobalISel/irtranslator-sibling-call.ll (+1-1)
  • (modified) llvm/test/CodeGen/AMDGPU/branch-folding-implicit-def-subreg.ll (+2-2)
  • (modified) llvm/test/CodeGen/AMDGPU/partial-regcopy-and-spill-missed-at-regalloc.ll (+4-4)
  • (modified) llvm/test/CodeGen/AMDGPU/spill-vector-superclass.ll (+1-1)
  • (modified) llvm/test/CodeGen/SystemZ/isel-debug.ll (+1-1)
  • (modified) llvm/test/CodeGen/Thumb2/active_lane_mask.ll (+2-3)
  • (modified) llvm/test/CodeGen/X86/GlobalISel/x86_64-irtranslator-struct-return.ll (+27-27)
  • (modified) llvm/test/CodeGen/X86/memcpy-scoped-aa.ll (+18-18)
  • (modified) llvm/test/CodeGen/X86/merge-store-partially-alias-loads.ll (+1-1)
  • (modified) llvm/test/CodeGen/X86/pr44140.ll (+6-6)
  • (modified) llvm/test/CodeGen/X86/sse-intrinsics-fast-isel.ll (+120-160)
  • (modified) llvm/test/CodeGen/X86/statepoint-cmp-sunk-past-statepoint.ll (+4-4)
  • (modified) llvm/test/DebugInfo/Generic/assignment-tracking/codegenprepare/sunk-addr.ll (+1-2)
  • (modified) llvm/test/DebugInfo/X86/bitcast-di.ll (+2-33)
  • (modified) llvm/test/Transforms/CodeGenPrepare/X86/masked-gather-struct-gep.ll (+2-3)
  • (modified) llvm/test/Transforms/CodeGenPrepare/X86/sink-addr-recreate.ll (+4-7)
  • (modified) llvm/test/Transforms/CodeGenPrepare/X86/sink-addrmode-base.ll (+39-24)
  • (modified) llvm/test/Transforms/CodeGenPrepare/X86/sink-addrmode-two-phi.ll (+1-2)
diff --git a/llvm/lib/CodeGen/CodeGenPrepare.cpp b/llvm/lib/CodeGen/CodeGenPrepare.cpp
index 9bbb89e37865d..bb397c1c35bf8 100644
--- a/llvm/lib/CodeGen/CodeGenPrepare.cpp
+++ b/llvm/lib/CodeGen/CodeGenPrepare.cpp
@@ -8720,6 +8720,13 @@ bool CodeGenPrepare::optimizeInst(Instruction *I, ModifyDT &ModifiedDT) {
     if (isa<Constant>(CI->getOperand(0)))
       return AnyChange;
 
+    // Remove noop bitcasts
+    if (isa<BitCastInst>(I) && I->getType() == I->getOperand(0)->getType()) {
+      replaceAllUsesWith(I, I->getOperand(0), FreshBBs, IsHugeFunc);
+      I->eraseFromParent();
+      return true;
+    }
+
     if (OptimizeNoopCopyExpression(CI, *TLI, *DL))
       return true;
 
diff --git a/llvm/test/CodeGen/AArch64/memcpy-scoped-aa.ll b/llvm/test/CodeGen/AArch64/memcpy-scoped-aa.ll
index 7dde168024278..f77d300adb2c7 100644
--- a/llvm/test/CodeGen/AArch64/memcpy-scoped-aa.ll
+++ b/llvm/test/CodeGen/AArch64/memcpy-scoped-aa.ll
@@ -9,8 +9,8 @@
 ; MIR-DAG: ![[SET1:[0-9]+]] = !{![[SCOPE1]]}
 
 ; MIR-LABEL: name: test_memcpy
-; MIR:      %2:fpr128 = LDRQui %0, 1 :: (load (s128) from %ir.p1, align 4, !alias.scope ![[SET0]], !noalias ![[SET1]])
-; MIR-NEXT: STRQui killed %2, %0, 0 :: (store (s128) into %ir.p0, align 4, !alias.scope ![[SET0]], !noalias ![[SET1]])
+; MIR:      %2:fpr128 = LDRQui %0, 1 :: (load (s128) from %ir.add.ptr, align 4, !alias.scope ![[SET0]], !noalias ![[SET1]])
+; MIR-NEXT: STRQui killed %2, %0, 0 :: (store (s128) into %ir.p, align 4, !alias.scope ![[SET0]], !noalias ![[SET1]])
 define i32 @test_memcpy(ptr nocapture %p, ptr nocapture readonly %q) {
 ; CHECK-LABEL: test_memcpy:
 ; CHECK:       // %bb.0:
@@ -32,8 +32,8 @@ define i32 @test_memcpy(ptr nocapture %p, ptr nocapture readonly %q) {
 }
 
 ; MIR-LABEL: name: test_memcpy_inline
-; MIR:      %2:fpr128 = LDRQui %0, 1 :: (load (s128) from %ir.p1, align 4, !alias.scope ![[SET0]], !noalias ![[SET1]])
-; MIR-NEXT: STRQui killed %2, %0, 0 :: (store (s128) into %ir.p0, align 4, !alias.scope ![[SET0]], !noalias ![[SET1]])
+; MIR:      %2:fpr128 = LDRQui %0, 1 :: (load (s128) from %ir.add.ptr, align 4, !alias.scope ![[SET0]], !noalias ![[SET1]])
+; MIR-NEXT: STRQui killed %2, %0, 0 :: (store (s128) into %ir.p, align 4, !alias.scope ![[SET0]], !noalias ![[SET1]])
 define i32 @test_memcpy_inline(ptr nocapture %p, ptr nocapture readonly %q) {
 ; CHECK-LABEL: test_memcpy_inline:
 ; CHECK:       // %bb.0:
@@ -55,8 +55,8 @@ define i32 @test_memcpy_inline(ptr nocapture %p, ptr nocapture readonly %q) {
 }
 
 ; MIR-LABEL: name: test_memmove
-; MIR:      %2:fpr128 = LDRQui %0, 1 :: (load (s128) from %ir.p1, align 4, !alias.scope ![[SET0]], !noalias ![[SET1]])
-; MIR-NEXT: STRQui killed %2, %0, 0 :: (store (s128) into %ir.p0, align 4, !alias.scope ![[SET0]], !noalias ![[SET1]])
+; MIR:      %2:fpr128 = LDRQui %0, 1 :: (load (s128) from %ir.add.ptr, align 4, !alias.scope ![[SET0]], !noalias ![[SET1]])
+; MIR-NEXT: STRQui killed %2, %0, 0 :: (store (s128) into %ir.p, align 4, !alias.scope ![[SET0]], !noalias ![[SET1]])
 define i32 @test_memmove(ptr nocapture %p, ptr nocapture readonly %q) {
 ; CHECK-LABEL: test_memmove:
 ; CHECK:       // %bb.0:
@@ -79,8 +79,8 @@ define i32 @test_memmove(ptr nocapture %p, ptr nocapture readonly %q) {
 
 ; MIR-LABEL: name: test_memset
 ; MIR:      %2:gpr64 = MOVi64imm -6148914691236517206
-; MIR-NEXT: STRXui %2, %0, 1 :: (store (s64) into %ir.p0 + 8, align 4, !alias.scope ![[SET0]], !noalias ![[SET1]])
-; MIR-NEXT: STRXui %2, %0, 0 :: (store (s64) into %ir.p0, align 4, !alias.scope ![[SET0]], !noalias ![[SET1]])
+; MIR-NEXT: STRXui %2, %0, 1 :: (store (s64) into %ir.p + 8, align 4, !alias.scope ![[SET0]], !noalias ![[SET1]])
+; MIR-NEXT: STRXui %2, %0, 0 :: (store (s64) into %ir.p, align 4, !alias.scope ![[SET0]], !noalias ![[SET1]])
 define i32 @test_memset(ptr nocapture %p, ptr nocapture readonly %q) {
 ; CHECK-LABEL: test_memset:
 ; CHECK:       // %bb.0:
@@ -100,8 +100,8 @@ define i32 @test_memset(ptr nocapture %p, ptr nocapture readonly %q) {
 }
 
 ; MIR-LABEL: name: test_mempcpy
-; MIR:      %2:fpr128 = LDRQui %0, 1 :: (load (s128) from %ir.p1, align 1, !alias.scope ![[SET0]], !noalias ![[SET1]])
-; MIR-NEXT: STRQui killed %2, %0, 0 :: (store (s128) into %ir.p0, align 1, !alias.scope ![[SET0]], !noalias ![[SET1]])
+; MIR:      %2:fpr128 = LDRQui %0, 1 :: (load (s128) from %ir.add.ptr, align 1, !alias.scope ![[SET0]], !noalias ![[SET1]])
+; MIR-NEXT: STRQui killed %2, %0, 0 :: (store (s128) into %ir.p, align 1, !alias.scope ![[SET0]], !noalias ![[SET1]])
 define i32 @test_mempcpy(ptr nocapture %p, ptr nocapture readonly %q) {
 ; CHECK-LABEL: test_mempcpy:
 ; CHECK:       // %bb.0:
diff --git a/llvm/test/CodeGen/AArch64/merge-scoped-aa-store.ll b/llvm/test/CodeGen/AArch64/merge-scoped-aa-store.ll
index 6b2a8c5aaf8b8..42fda4b4309af 100644
--- a/llvm/test/CodeGen/AArch64/merge-scoped-aa-store.ll
+++ b/llvm/test/CodeGen/AArch64/merge-scoped-aa-store.ll
@@ -21,7 +21,7 @@ define void @blam0(ptr %g0, ptr %g1) {
   ; MIR-NEXT:   [[COPY:%[0-9]+]]:gpr64common = COPY $x1
   ; MIR-NEXT:   [[COPY1:%[0-9]+]]:gpr64common = COPY $x0
   ; MIR-NEXT:   [[LDRDui:%[0-9]+]]:fpr64 = LDRDui [[COPY1]], 0 :: (load (s64) from %ir.g0, align 4, !alias.scope !0, !noalias !3)
-  ; MIR-NEXT:   STRDui killed [[LDRDui]], [[COPY]], 0 :: (store (s64) into %ir.tmp41, align 4, !alias.scope !3, !noalias !0)
+  ; MIR-NEXT:   STRDui killed [[LDRDui]], [[COPY]], 0 :: (store (s64) into %ir.g1, align 4, !alias.scope !3, !noalias !0)
   ; MIR-NEXT:   RET_ReallyLR
   %tmp4 = getelementptr inbounds <3 x float>, ptr %g1, i64 0, i64 0
   %tmp5 = load <3 x float>, ptr %g0, align 4, !alias.scope !0, !noalias !1
@@ -43,7 +43,7 @@ define void @blam1(ptr %g0, ptr %g1) {
   ; MIR-NEXT:   [[COPY:%[0-9]+]]:gpr64common = COPY $x1
   ; MIR-NEXT:   [[COPY1:%[0-9]+]]:gpr64common = COPY $x0
   ; MIR-NEXT:   [[LDRDui:%[0-9]+]]:fpr64 = LDRDui [[COPY1]], 0 :: (load (s64) from %ir.g0, align 4, !alias.scope !0, !noalias !3)
-  ; MIR-NEXT:   STRDui killed [[LDRDui]], [[COPY]], 0 :: (store (s64) into %ir.tmp41, align 4, !alias.scope !9, !noalias !10)
+  ; MIR-NEXT:   STRDui killed [[LDRDui]], [[COPY]], 0 :: (store (s64) into %ir.g1, align 4, !alias.scope !9, !noalias !10)
   ; MIR-NEXT:   RET_ReallyLR
   %tmp4 = getelementptr inbounds <3 x float>, ptr %g1, i64 0, i64 0
   %tmp5 = load <3 x float>, ptr %g0, align 4, !alias.scope !0, !noalias !1
diff --git a/llvm/test/CodeGen/AMDGPU/GlobalISel/irtranslator-call-return-values.ll b/llvm/test/CodeGen/AMDGPU/GlobalISel/irtranslator-call-return-values.ll
index 96ee15f2eb78b..c7977c339880d 100644
--- a/llvm/test/CodeGen/AMDGPU/GlobalISel/irtranslator-call-return-values.ll
+++ b/llvm/test/CodeGen/AMDGPU/GlobalISel/irtranslator-call-return-values.ll
@@ -84,7 +84,7 @@ define amdgpu_kernel void @test_call_external_i32_func_i32_imm(ptr addrspace(1)
   ; GCN-NEXT:   [[COPY9:%[0-9]+]]:_(p4) = COPY $sgpr8_sgpr9
   ; GCN-NEXT:   [[C:%[0-9]+]]:_(s32) = G_CONSTANT i32 42
   ; GCN-NEXT:   [[INT:%[0-9]+]]:_(p4) = G_INTRINSIC intrinsic(@llvm.amdgcn.kernarg.segment.ptr)
-  ; GCN-NEXT:   [[LOAD:%[0-9]+]]:_(p1) = G_LOAD [[INT]](p4) :: (dereferenceable invariant load (p1) from %ir.out.kernarg.offset1, align 16, addrspace 4)
+  ; GCN-NEXT:   [[LOAD:%[0-9]+]]:_(p1) = G_LOAD [[INT]](p4) :: (dereferenceable invariant load (p1) from %ir.test_call_external_i32_func_i32_imm.kernarg.segment, align 16, addrspace 4)
   ; GCN-NEXT:   ADJCALLSTACKUP 0, 0, implicit-def $scc
   ; GCN-NEXT:   [[GV:%[0-9]+]]:_(p0) = G_GLOBAL_VALUE @external_i32_func_i32
   ; GCN-NEXT:   [[COPY10:%[0-9]+]]:_(p4) = COPY [[COPY8]]
@@ -2957,7 +2957,7 @@ define amdgpu_kernel void @test_call_external_v33i32_func_v33i32_i32(ptr addrspa
   ; GCN-NEXT:   [[COPY9:%[0-9]+]]:_(p4) = COPY $sgpr8_sgpr9
   ; GCN-NEXT:   [[DEF:%[0-9]+]]:_(p1) = G_IMPLICIT_DEF
   ; GCN-NEXT:   [[INT:%[0-9]+]]:_(p4) = G_INTRINSIC intrinsic(@llvm.amdgcn.kernarg.segment.ptr)
-  ; GCN-NEXT:   [[LOAD:%[0-9]+]]:_(p1) = G_LOAD [[INT]](p4) :: (dereferenceable invariant load (p1) from %ir.p.kernarg.offset1, align 16, addrspace 4)
+  ; GCN-NEXT:   [[LOAD:%[0-9]+]]:_(p1) = G_LOAD [[INT]](p4) :: (dereferenceable invariant load (p1) from %ir.test_call_external_v33i32_func_v33i32_i32.kernarg.segment, align 16, addrspace 4)
   ; GCN-NEXT:   [[C:%[0-9]+]]:_(s64) = G_CONSTANT i64 8
   ; GCN-NEXT:   %18:_(p4) = nuw nusw G_PTR_ADD [[INT]], [[C]](s64)
   ; GCN-NEXT:   [[LOAD1:%[0-9]+]]:_(s32) = G_LOAD %18(p4) :: (dereferenceable invariant load (s32) from %ir.idx.kernarg.offset, align 8, addrspace 4)
diff --git a/llvm/test/CodeGen/AMDGPU/GlobalISel/irtranslator-call.ll b/llvm/test/CodeGen/AMDGPU/GlobalISel/irtranslator-call.ll
index 92106d7e1d60d..862093ebf9416 100644
--- a/llvm/test/CodeGen/AMDGPU/GlobalISel/irtranslator-call.ll
+++ b/llvm/test/CodeGen/AMDGPU/GlobalISel/irtranslator-call.ll
@@ -1350,7 +1350,7 @@ define amdgpu_kernel void @test_call_external_void_func_p0_imm(ptr %arg) #0 {
   ; CHECK-NEXT:   [[COPY8:%[0-9]+]]:sgpr_64 = COPY $sgpr4_sgpr5
   ; CHECK-NEXT:   [[COPY9:%[0-9]+]]:_(p4) = COPY $sgpr8_sgpr9
   ; CHECK-NEXT:   [[INT:%[0-9]+]]:_(p4) = G_INTRINSIC intrinsic(@llvm.amdgcn.kernarg.segment.ptr)
-  ; CHECK-NEXT:   [[LOAD:%[0-9]+]]:_(p0) = G_LOAD [[INT]](p4) :: (dereferenceable invariant load (p0) from %ir.arg.kernarg.offset1, align 16, addrspace 4)
+  ; CHECK-NEXT:   [[LOAD:%[0-9]+]]:_(p0) = G_LOAD [[INT]](p4) :: (dereferenceable invariant load (p0) from %ir.test_call_external_void_func_p0_imm.kernarg.segment, align 16, addrspace 4)
   ; CHECK-NEXT:   ADJCALLSTACKUP 0, 0, implicit-def $scc
   ; CHECK-NEXT:   [[GV:%[0-9]+]]:_(p0) = G_GLOBAL_VALUE @external_void_func_p0
   ; CHECK-NEXT:   [[COPY10:%[0-9]+]]:_(p4) = COPY [[COPY8]]
@@ -4654,7 +4654,7 @@ define amdgpu_kernel void @stack_passed_arg_alignment_v32i32_f64(<32 x i32> %val
   ; CHECK-NEXT:   [[COPY8:%[0-9]+]]:sgpr_64 = COPY $sgpr4_sgpr5
   ; CHECK-NEXT:   [[COPY9:%[0-9]+]]:_(p4) = COPY $sgpr8_sgpr9
   ; CHECK-NEXT:   [[INT:%[0-9]+]]:_(p4) = G_INTRINSIC intrinsic(@llvm.amdgcn.kernarg.segment.ptr)
-  ; CHECK-NEXT:   [[LOAD:%[0-9]+]]:_(<32 x s32>) = G_LOAD [[INT]](p4) :: (dereferenceable invariant load (<32 x s32>) from %ir.val.kernarg.offset1, align 16, addrspace 4)
+  ; CHECK-NEXT:   [[LOAD:%[0-9]+]]:_(<32 x s32>) = G_LOAD [[INT]](p4) :: (dereferenceable invariant load (<32 x s32>) from %ir.stack_passed_arg_alignment_v32i32_f64.kernarg.segment, align 16, addrspace 4)
   ; CHECK-NEXT:   [[C:%[0-9]+]]:_(s64) = G_CONSTANT i64 128
   ; CHECK-NEXT:   %18:_(p4) = nuw nusw G_PTR_ADD [[INT]], [[C]](s64)
   ; CHECK-NEXT:   [[LOAD1:%[0-9]+]]:_(s64) = G_LOAD %18(p4) :: (dereferenceable invariant load (s64) from %ir.tmp.kernarg.offset, align 16, addrspace 4)
diff --git a/llvm/test/CodeGen/AMDGPU/GlobalISel/irtranslator-indirect-call.ll b/llvm/test/CodeGen/AMDGPU/GlobalISel/irtranslator-indirect-call.ll
index ac0d5ee78666e..bdf89f785ef99 100644
--- a/llvm/test/CodeGen/AMDGPU/GlobalISel/irtranslator-indirect-call.ll
+++ b/llvm/test/CodeGen/AMDGPU/GlobalISel/irtranslator-indirect-call.ll
@@ -17,7 +17,7 @@ define amdgpu_kernel void @test_indirect_call_sgpr_ptr(ptr %fptr) {
   ; CHECK-NEXT:   [[COPY8:%[0-9]+]]:sgpr_64 = COPY $sgpr4_sgpr5
   ; CHECK-NEXT:   [[COPY9:%[0-9]+]]:_(p4) = COPY $sgpr8_sgpr9
   ; CHECK-NEXT:   [[INT:%[0-9]+]]:_(p4) = G_INTRINSIC intrinsic(@llvm.amdgcn.kernarg.segment.ptr)
-  ; CHECK-NEXT:   [[LOAD:%[0-9]+]]:_(p0) = G_LOAD [[INT]](p4) :: (dereferenceable invariant load (p0) from %ir.fptr.kernarg.offset1, align 16, addrspace 4)
+  ; CHECK-NEXT:   [[LOAD:%[0-9]+]]:_(p0) = G_LOAD [[INT]](p4) :: (dereferenceable invariant load (p0) from %ir.test_indirect_call_sgpr_ptr.kernarg.segment, align 16, addrspace 4)
   ; CHECK-NEXT:   ADJCALLSTACKUP 0, 0, implicit-def $scc
   ; CHECK-NEXT:   [[COPY10:%[0-9]+]]:_(p4) = COPY [[COPY8]]
   ; CHECK-NEXT:   [[COPY11:%[0-9]+]]:_(p4) = COPY [[COPY7]]
diff --git a/llvm/test/CodeGen/AMDGPU/GlobalISel/irtranslator-sibling-call.ll b/llvm/test/CodeGen/AMDGPU/GlobalISel/irtranslator-sibling-call.ll
index ca580d8f29c84..a3f5d927d4b80 100644
--- a/llvm/test/CodeGen/AMDGPU/GlobalISel/irtranslator-sibling-call.ll
+++ b/llvm/test/CodeGen/AMDGPU/GlobalISel/irtranslator-sibling-call.ll
@@ -140,7 +140,7 @@ define amdgpu_kernel void @kernel_call_i32_fastcc_i32_i32_unused_result(i32 %a,
   ; GCN-NEXT:   [[C:%[0-9]+]]:_(s32) = G_CONSTANT i32 0
   ; GCN-NEXT:   [[C1:%[0-9]+]]:_(s32) = G_CONSTANT i32 1
   ; GCN-NEXT:   [[INT:%[0-9]+]]:_(p4) = G_INTRINSIC intrinsic(@llvm.amdgcn.kernarg.segment.ptr)
-  ; GCN-NEXT:   [[LOAD:%[0-9]+]]:_(<2 x s32>) = G_LOAD [[INT]](p4) :: (dereferenceable invariant load (<2 x s32>) from %ir.a.kernarg.offset1, align 16, addrspace 4)
+  ; GCN-NEXT:   [[LOAD:%[0-9]+]]:_(<2 x s32>) = G_LOAD [[INT]](p4) :: (dereferenceable invariant load (<2 x s32>) from %ir.kernel_call_i32_fastcc_i32_i32_unused_result.kernarg.segment, align 16, addrspace 4)
   ; GCN-NEXT:   [[EVEC:%[0-9]+]]:_(s32) = G_EXTRACT_VECTOR_ELT [[LOAD]](<2 x s32>), [[C]](s32)
   ; GCN-NEXT:   [[EVEC1:%[0-9]+]]:_(s32) = G_EXTRACT_VECTOR_ELT [[LOAD]](<2 x s32>), [[C1]](s32)
   ; GCN-NEXT:   ADJCALLSTACKUP 0, 0, implicit-def $scc
diff --git a/llvm/test/CodeGen/AMDGPU/branch-folding-implicit-def-subreg.ll b/llvm/test/CodeGen/AMDGPU/branch-folding-implicit-def-subreg.ll
index ae90cfb631e8d..412e3569c0d05 100644
--- a/llvm/test/CodeGen/AMDGPU/branch-folding-implicit-def-subreg.ll
+++ b/llvm/test/CodeGen/AMDGPU/branch-folding-implicit-def-subreg.ll
@@ -16,8 +16,8 @@ define amdgpu_kernel void @f1(ptr addrspace(1) %arg, ptr addrspace(1) %arg1, i64
   ; GFX90A-NEXT:   renamable $sgpr17 = S_LOAD_DWORD_IMM renamable $sgpr8_sgpr9, 24, 0 :: (dereferenceable invariant load (s32) from %ir.arg4.kernarg.offset.align.down, align 8, addrspace 4)
   ; GFX90A-NEXT:   early-clobber renamable $sgpr20_sgpr21_sgpr22_sgpr23 = S_LOAD_DWORDX4_IMM_ec renamable $sgpr8_sgpr9, 24, 0 :: (dereferenceable invariant load (s128) from %ir.arg6.kernarg.offset.align.down, align 8, addrspace 4)
   ; GFX90A-NEXT:   renamable $sgpr33 = S_LOAD_DWORD_IMM renamable $sgpr8_sgpr9, 40, 0 :: (dereferenceable invariant load (s32) from %ir.arg6.kernarg.offset.align.down + 16, align 8, addrspace 4)
-  ; GFX90A-NEXT:   renamable $sgpr24_sgpr25_sgpr26_sgpr27 = S_LOAD_DWORDX4_IMM renamable $sgpr8_sgpr9, 0, 0 :: (dereferenceable invariant load (s128) from %ir.arg.kernarg.offset1, addrspace 4)
-  ; GFX90A-NEXT:   renamable $sgpr46_sgpr47 = S_LOAD_DWORDX2_IMM renamable $sgpr8_sgpr9, 16, 0 :: (dereferenceable invariant load (s64) from %ir.arg.kernarg.offset1 + 16, align 16, addrspace 4)
+  ; GFX90A-NEXT:   renamable $sgpr24_sgpr25_sgpr26_sgpr27 = S_LOAD_DWORDX4_IMM renamable $sgpr8_sgpr9, 0, 0 :: (dereferenceable invariant load (s128) from %ir.f1.kernarg.segment, addrspace 4)
+  ; GFX90A-NEXT:   renamable $sgpr46_sgpr47 = S_LOAD_DWORDX2_IMM renamable $sgpr8_sgpr9, 16, 0 :: (dereferenceable invariant load (s64) from %ir.f1.kernarg.segment + 16, align 16, addrspace 4)
   ; GFX90A-NEXT:   S_BITCMP1_B32 renamable $sgpr17, 0, implicit-def $scc
   ; GFX90A-NEXT:   renamable $sgpr12_sgpr13 = S_CSELECT_B64 -1, 0, implicit killed $scc
   ; GFX90A-NEXT:   renamable $sgpr34_sgpr35 = S_MOV_B64 -1
diff --git a/llvm/test/CodeGen/AMDGPU/partial-regcopy-and-spill-missed-at-regalloc.ll b/llvm/test/CodeGen/AMDGPU/partial-regcopy-and-spill-missed-at-regalloc.ll
index 663fd98b46bf7..03d3724c232fa 100644
--- a/llvm/test/CodeGen/AMDGPU/partial-regcopy-and-spill-missed-at-regalloc.ll
+++ b/llvm/test/CodeGen/AMDGPU/partial-regcopy-and-spill-missed-at-regalloc.ll
@@ -15,7 +15,7 @@ define amdgpu_kernel void @partial_copy(<4 x i32> %arg) #0 {
   ; REGALLOC-GFX908-NEXT:   INLINEASM &"; def $0", 1 /* sideeffect attdialect */, 6094858 /* regdef:VReg_128 */, def %7
   ; REGALLOC-GFX908-NEXT:   INLINEASM &"; def $0", 1 /* sideeffect attdialect */, 3538954 /* regdef:VReg_64 */, def %8
   ; REGALLOC-GFX908-NEXT:   GLOBAL_STORE_DWORDX4 undef %15:vreg_64, %7, 0, 0, implicit $exec :: (volatile store (s128) into `ptr addrspace(1) poison`, addrspace 1)
-  ; REGALLOC-GFX908-NEXT:   renamable $sgpr0_sgpr1_sgpr2_sgpr3 = S_LOAD_DWORDX4_IMM killed renamable $sgpr4_sgpr5, 0, 0 :: (dereferenceable invariant load (s128) from %ir.arg.kernarg.offset1, addrspace 4)
+  ; REGALLOC-GFX908-NEXT:   renamable $sgpr0_sgpr1_sgpr2_sgpr3 = S_LOAD_DWORDX4_IMM killed renamable $sgpr4_sgpr5, 0, 0 :: (dereferenceable invariant load (s128) from %ir.partial_copy.kernarg.segment, addrspace 4)
   ; REGALLOC-GFX908-NEXT:   [[COPY:%[0-9]+]]:areg_128 = COPY killed renamable $sgpr0_sgpr1_sgpr2_sgpr3
   ; REGALLOC-GFX908-NEXT:   [[V_MOV_B32_e32_:%[0-9]+]]:vgpr_32 = V_MOV_B32_e32 1, implicit $exec
   ; REGALLOC-GFX908-NEXT:   [[V_MOV_B32_e32_1:%[0-9]+]]:vgpr_32 = V_MOV_B32_e32 2, implicit $exec
@@ -40,7 +40,7 @@ define amdgpu_kernel void @partial_copy(<4 x i32> %arg) #0 {
   ; PEI-GFX908-NEXT:   $agpr4 = V_ACCVGPR_WRITE_B32_e64 killed $vgpr1, implicit $exec, implicit killed $vgpr0_vgpr1
   ; PEI-GFX908-NEXT:   renamable $vgpr0_vgpr1_vgpr2_vgpr3 = COPY killed renamable $agpr0_agpr1_agpr2_agpr3, implicit $exec
   ; PEI-GFX908-NEXT:   GLOBAL_STORE_DWORDX4 undef renamable $vgpr0_vgpr1, killed renamable $vgpr0_vgpr1_vgpr2_vgpr3, 0, 0, implicit $exec :: (volatile store (s128) into `ptr addrspace(1) poison`, addrspace 1)
-  ; PEI-GFX908-NEXT:   renamable $sgpr0_sgpr1_sgpr2_sgpr3 = S_LOAD_DWORDX4_IMM killed renamable $sgpr4_sgpr5, 0, 0 :: (dereferenceable invariant load (s128) from %ir.arg.kernarg.offset1, addrspace 4)
+  ; PEI-GFX908-NEXT:   renamable $sgpr0_sgpr1_sgpr2_sgpr3 = S_LOAD_DWORDX4_IMM killed renamable $sgpr4_sgpr5, 0, 0 :: (dereferenceable invariant load (s128) from %ir.partial_copy.kernarg.segment, addrspace 4)
   ; PEI-GFX908-NEXT:   renamable $agpr0_agpr1_agpr2_agpr3 = COPY killed renamable $sgpr0_sgpr1_sgpr2_sgpr3, implicit $exec
   ; PEI-GFX908-NEXT:   renamable $vgpr0 = V_MOV_B32_e32 1, implicit $exec
   ; PEI-GFX908-NEXT:   renamable $vgpr1 = V_MOV_B32_e32 2, implicit $exec
@@ -60,7 +60,7 @@ define amdgpu_kernel void @partial_copy(<4 x i32> %arg) #0 {
   ; REGALLOC-GFX90A-NEXT:   INLINEASM &"; def $0", 1 /* sideeffect attdialect */, 6422538 /* regdef:VReg_128_Align2 */, def %7
   ; REGALLOC-GFX90A-NEXT:   INLINEASM &"; def $0", 1 /* sideeffect attdialect */, 3866634 /* regdef:VReg_64_Align2 */, def %8
   ; REGALLOC-GFX90A-NEXT:   GLOBAL_STORE_DWORDX4 undef %15:vreg_64_align2, %7, 0, 0, implicit $exec :: (volatile store (s128) into `ptr addrspace(1) poison`, addrspace 1)
-  ; REGALLOC-GFX90A-NEXT:   renamable $sgpr0_sgpr1_sgpr2_sgpr3 = S_LOAD_DWORDX4_IMM killed renamable $sgpr4_sgpr5, 0, 0 :: (dereferenceable invariant load (s128) from %ir.arg.kernarg.offset1, addrspace 4)
+  ; REGALLOC-GFX90A-NEXT:   renamable $sgpr0_sgpr1_sgpr2_sgpr3 = S_LOAD_DWORDX4_IMM killed renamable $sgpr4_sgpr5, 0, 0 :: (dereferenceable invariant load (s128) from %ir.partial_copy.kernarg.segment, addrspace 4)
   ; REGALLOC-GFX90A-NEXT:   [[COPY:%[0-9]+]]:areg_128_align2 = COPY killed renamable $sgpr0_sgpr1_sgpr2_sgpr3
   ; REGALLOC-GFX90A-NEXT:   [[V_MOV_B32_e32_:%[0-9]+]]:vgpr_32 = V_MOV_B32_e32 1, implicit $exec
   ; REGALLOC-GFX90A-NEXT:   [[V_MOV_B32_e32_1:%[0-9]+]]:vgpr_32 = V_MOV_B32_e32 2, implicit $exec
@@ -83,7 +83,7 @@ define amdgpu_kernel void @partial_copy(<4 x i32> %arg) #0 {
   ; PEI-GFX90A-NEXT:   BUFFER_STORE_DWORD_OFFSET killed $vgpr0, $sgpr12_sgpr13_sgpr14_sgpr15, 0, 0, 0, 0, implicit $exec, implicit-def $vgpr0_vgpr1, implicit $vgpr0_vgpr1 :: (store (s32) into %stack.0, addrspace 5)
   ; PEI-GFX90A-NEXT:   $agpr4 = V_ACCVGPR_WRITE_B32_e64 killed $vgpr1, implicit $exec, implicit killed $vgpr0_vgpr1
   ; PEI-GFX90A-NEXT:   GLOBAL_STORE_DWORDX4 undef renamable $vgpr0_vgpr1, killed renamable $agpr0_agpr1_agpr2_agpr3, 0, 0, implicit $exec :: (volatile store (s128) into `ptr addrspace(1) poison`, addrspace 1)
-  ; PEI-GFX90A-NEXT:   renamable $sgpr0_sgpr1_sgpr2_sgpr3 = S_LOAD_DWORDX4_IMM killed renamable $sgpr4_sgpr5, 0, 0 :: (dereferenceable invariant load (s128) from %ir.arg.kernarg.offset1, addrspace 4)
+  ; PEI-GFX90A-NEXT:   renamable $sgpr0_sgpr1_sgpr2_sgpr3 = S_LOAD_DWORDX4_IMM killed renamable $sgpr4_sgpr5, 0, 0 :: (dereferenceable invariant load (s128) from %ir.partial_copy.kernarg.segment, addrspace 4)
   ; PEI-GFX90A-NEXT:   renamable $agpr0_agpr1_agpr2_agpr3 = COPY killed renamable $sgpr0_sgpr1_sgpr2_sgpr3, implicit $exec
   ; PEI-GFX90A-NEXT:   renamable $vgpr0 = V_MOV_B32_e32 1, implicit $exec
   ; PEI-GFX90A-NEXT:   renamable $vgpr1 = V_MOV_B32_e32 2, implicit $exec
diff ...
[truncated]

@llvmbot
Copy link
Member

llvmbot commented Jul 3, 2025

@llvm/pr-subscribers-backend-systemz

Author: Philip Reames (preames)

Changes

The change itself is very straight forward, it's mostly just updating tests.

Somewhat surprisingly, this does cause some minor codegen differences in a few tests.

Patch is 89.99 KiB, truncated to 20.00 KiB below, full version: https://github.com/llvm/llvm-project/pull/146961.diff

24 Files Affected:

  • (modified) llvm/lib/CodeGen/CodeGenPrepare.cpp (+7)
  • (modified) llvm/test/CodeGen/AArch64/memcpy-scoped-aa.ll (+10-10)
  • (modified) llvm/test/CodeGen/AArch64/merge-scoped-aa-store.ll (+2-2)
  • (modified) llvm/test/CodeGen/AMDGPU/GlobalISel/irtranslator-call-return-values.ll (+2-2)
  • (modified) llvm/test/CodeGen/AMDGPU/GlobalISel/irtranslator-call.ll (+2-2)
  • (modified) llvm/test/CodeGen/AMDGPU/GlobalISel/irtranslator-indirect-call.ll (+1-1)
  • (modified) llvm/test/CodeGen/AMDGPU/GlobalISel/irtranslator-sibling-call.ll (+1-1)
  • (modified) llvm/test/CodeGen/AMDGPU/branch-folding-implicit-def-subreg.ll (+2-2)
  • (modified) llvm/test/CodeGen/AMDGPU/partial-regcopy-and-spill-missed-at-regalloc.ll (+4-4)
  • (modified) llvm/test/CodeGen/AMDGPU/spill-vector-superclass.ll (+1-1)
  • (modified) llvm/test/CodeGen/SystemZ/isel-debug.ll (+1-1)
  • (modified) llvm/test/CodeGen/Thumb2/active_lane_mask.ll (+2-3)
  • (modified) llvm/test/CodeGen/X86/GlobalISel/x86_64-irtranslator-struct-return.ll (+27-27)
  • (modified) llvm/test/CodeGen/X86/memcpy-scoped-aa.ll (+18-18)
  • (modified) llvm/test/CodeGen/X86/merge-store-partially-alias-loads.ll (+1-1)
  • (modified) llvm/test/CodeGen/X86/pr44140.ll (+6-6)
  • (modified) llvm/test/CodeGen/X86/sse-intrinsics-fast-isel.ll (+120-160)
  • (modified) llvm/test/CodeGen/X86/statepoint-cmp-sunk-past-statepoint.ll (+4-4)
  • (modified) llvm/test/DebugInfo/Generic/assignment-tracking/codegenprepare/sunk-addr.ll (+1-2)
  • (modified) llvm/test/DebugInfo/X86/bitcast-di.ll (+2-33)
  • (modified) llvm/test/Transforms/CodeGenPrepare/X86/masked-gather-struct-gep.ll (+2-3)
  • (modified) llvm/test/Transforms/CodeGenPrepare/X86/sink-addr-recreate.ll (+4-7)
  • (modified) llvm/test/Transforms/CodeGenPrepare/X86/sink-addrmode-base.ll (+39-24)
  • (modified) llvm/test/Transforms/CodeGenPrepare/X86/sink-addrmode-two-phi.ll (+1-2)
diff --git a/llvm/lib/CodeGen/CodeGenPrepare.cpp b/llvm/lib/CodeGen/CodeGenPrepare.cpp
index 9bbb89e37865d..bb397c1c35bf8 100644
--- a/llvm/lib/CodeGen/CodeGenPrepare.cpp
+++ b/llvm/lib/CodeGen/CodeGenPrepare.cpp
@@ -8720,6 +8720,13 @@ bool CodeGenPrepare::optimizeInst(Instruction *I, ModifyDT &ModifiedDT) {
     if (isa<Constant>(CI->getOperand(0)))
       return AnyChange;
 
+    // Remove noop bitcasts
+    if (isa<BitCastInst>(I) && I->getType() == I->getOperand(0)->getType()) {
+      replaceAllUsesWith(I, I->getOperand(0), FreshBBs, IsHugeFunc);
+      I->eraseFromParent();
+      return true;
+    }
+
     if (OptimizeNoopCopyExpression(CI, *TLI, *DL))
       return true;
 
diff --git a/llvm/test/CodeGen/AArch64/memcpy-scoped-aa.ll b/llvm/test/CodeGen/AArch64/memcpy-scoped-aa.ll
index 7dde168024278..f77d300adb2c7 100644
--- a/llvm/test/CodeGen/AArch64/memcpy-scoped-aa.ll
+++ b/llvm/test/CodeGen/AArch64/memcpy-scoped-aa.ll
@@ -9,8 +9,8 @@
 ; MIR-DAG: ![[SET1:[0-9]+]] = !{![[SCOPE1]]}
 
 ; MIR-LABEL: name: test_memcpy
-; MIR:      %2:fpr128 = LDRQui %0, 1 :: (load (s128) from %ir.p1, align 4, !alias.scope ![[SET0]], !noalias ![[SET1]])
-; MIR-NEXT: STRQui killed %2, %0, 0 :: (store (s128) into %ir.p0, align 4, !alias.scope ![[SET0]], !noalias ![[SET1]])
+; MIR:      %2:fpr128 = LDRQui %0, 1 :: (load (s128) from %ir.add.ptr, align 4, !alias.scope ![[SET0]], !noalias ![[SET1]])
+; MIR-NEXT: STRQui killed %2, %0, 0 :: (store (s128) into %ir.p, align 4, !alias.scope ![[SET0]], !noalias ![[SET1]])
 define i32 @test_memcpy(ptr nocapture %p, ptr nocapture readonly %q) {
 ; CHECK-LABEL: test_memcpy:
 ; CHECK:       // %bb.0:
@@ -32,8 +32,8 @@ define i32 @test_memcpy(ptr nocapture %p, ptr nocapture readonly %q) {
 }
 
 ; MIR-LABEL: name: test_memcpy_inline
-; MIR:      %2:fpr128 = LDRQui %0, 1 :: (load (s128) from %ir.p1, align 4, !alias.scope ![[SET0]], !noalias ![[SET1]])
-; MIR-NEXT: STRQui killed %2, %0, 0 :: (store (s128) into %ir.p0, align 4, !alias.scope ![[SET0]], !noalias ![[SET1]])
+; MIR:      %2:fpr128 = LDRQui %0, 1 :: (load (s128) from %ir.add.ptr, align 4, !alias.scope ![[SET0]], !noalias ![[SET1]])
+; MIR-NEXT: STRQui killed %2, %0, 0 :: (store (s128) into %ir.p, align 4, !alias.scope ![[SET0]], !noalias ![[SET1]])
 define i32 @test_memcpy_inline(ptr nocapture %p, ptr nocapture readonly %q) {
 ; CHECK-LABEL: test_memcpy_inline:
 ; CHECK:       // %bb.0:
@@ -55,8 +55,8 @@ define i32 @test_memcpy_inline(ptr nocapture %p, ptr nocapture readonly %q) {
 }
 
 ; MIR-LABEL: name: test_memmove
-; MIR:      %2:fpr128 = LDRQui %0, 1 :: (load (s128) from %ir.p1, align 4, !alias.scope ![[SET0]], !noalias ![[SET1]])
-; MIR-NEXT: STRQui killed %2, %0, 0 :: (store (s128) into %ir.p0, align 4, !alias.scope ![[SET0]], !noalias ![[SET1]])
+; MIR:      %2:fpr128 = LDRQui %0, 1 :: (load (s128) from %ir.add.ptr, align 4, !alias.scope ![[SET0]], !noalias ![[SET1]])
+; MIR-NEXT: STRQui killed %2, %0, 0 :: (store (s128) into %ir.p, align 4, !alias.scope ![[SET0]], !noalias ![[SET1]])
 define i32 @test_memmove(ptr nocapture %p, ptr nocapture readonly %q) {
 ; CHECK-LABEL: test_memmove:
 ; CHECK:       // %bb.0:
@@ -79,8 +79,8 @@ define i32 @test_memmove(ptr nocapture %p, ptr nocapture readonly %q) {
 
 ; MIR-LABEL: name: test_memset
 ; MIR:      %2:gpr64 = MOVi64imm -6148914691236517206
-; MIR-NEXT: STRXui %2, %0, 1 :: (store (s64) into %ir.p0 + 8, align 4, !alias.scope ![[SET0]], !noalias ![[SET1]])
-; MIR-NEXT: STRXui %2, %0, 0 :: (store (s64) into %ir.p0, align 4, !alias.scope ![[SET0]], !noalias ![[SET1]])
+; MIR-NEXT: STRXui %2, %0, 1 :: (store (s64) into %ir.p + 8, align 4, !alias.scope ![[SET0]], !noalias ![[SET1]])
+; MIR-NEXT: STRXui %2, %0, 0 :: (store (s64) into %ir.p, align 4, !alias.scope ![[SET0]], !noalias ![[SET1]])
 define i32 @test_memset(ptr nocapture %p, ptr nocapture readonly %q) {
 ; CHECK-LABEL: test_memset:
 ; CHECK:       // %bb.0:
@@ -100,8 +100,8 @@ define i32 @test_memset(ptr nocapture %p, ptr nocapture readonly %q) {
 }
 
 ; MIR-LABEL: name: test_mempcpy
-; MIR:      %2:fpr128 = LDRQui %0, 1 :: (load (s128) from %ir.p1, align 1, !alias.scope ![[SET0]], !noalias ![[SET1]])
-; MIR-NEXT: STRQui killed %2, %0, 0 :: (store (s128) into %ir.p0, align 1, !alias.scope ![[SET0]], !noalias ![[SET1]])
+; MIR:      %2:fpr128 = LDRQui %0, 1 :: (load (s128) from %ir.add.ptr, align 1, !alias.scope ![[SET0]], !noalias ![[SET1]])
+; MIR-NEXT: STRQui killed %2, %0, 0 :: (store (s128) into %ir.p, align 1, !alias.scope ![[SET0]], !noalias ![[SET1]])
 define i32 @test_mempcpy(ptr nocapture %p, ptr nocapture readonly %q) {
 ; CHECK-LABEL: test_mempcpy:
 ; CHECK:       // %bb.0:
diff --git a/llvm/test/CodeGen/AArch64/merge-scoped-aa-store.ll b/llvm/test/CodeGen/AArch64/merge-scoped-aa-store.ll
index 6b2a8c5aaf8b8..42fda4b4309af 100644
--- a/llvm/test/CodeGen/AArch64/merge-scoped-aa-store.ll
+++ b/llvm/test/CodeGen/AArch64/merge-scoped-aa-store.ll
@@ -21,7 +21,7 @@ define void @blam0(ptr %g0, ptr %g1) {
   ; MIR-NEXT:   [[COPY:%[0-9]+]]:gpr64common = COPY $x1
   ; MIR-NEXT:   [[COPY1:%[0-9]+]]:gpr64common = COPY $x0
   ; MIR-NEXT:   [[LDRDui:%[0-9]+]]:fpr64 = LDRDui [[COPY1]], 0 :: (load (s64) from %ir.g0, align 4, !alias.scope !0, !noalias !3)
-  ; MIR-NEXT:   STRDui killed [[LDRDui]], [[COPY]], 0 :: (store (s64) into %ir.tmp41, align 4, !alias.scope !3, !noalias !0)
+  ; MIR-NEXT:   STRDui killed [[LDRDui]], [[COPY]], 0 :: (store (s64) into %ir.g1, align 4, !alias.scope !3, !noalias !0)
   ; MIR-NEXT:   RET_ReallyLR
   %tmp4 = getelementptr inbounds <3 x float>, ptr %g1, i64 0, i64 0
   %tmp5 = load <3 x float>, ptr %g0, align 4, !alias.scope !0, !noalias !1
@@ -43,7 +43,7 @@ define void @blam1(ptr %g0, ptr %g1) {
   ; MIR-NEXT:   [[COPY:%[0-9]+]]:gpr64common = COPY $x1
   ; MIR-NEXT:   [[COPY1:%[0-9]+]]:gpr64common = COPY $x0
   ; MIR-NEXT:   [[LDRDui:%[0-9]+]]:fpr64 = LDRDui [[COPY1]], 0 :: (load (s64) from %ir.g0, align 4, !alias.scope !0, !noalias !3)
-  ; MIR-NEXT:   STRDui killed [[LDRDui]], [[COPY]], 0 :: (store (s64) into %ir.tmp41, align 4, !alias.scope !9, !noalias !10)
+  ; MIR-NEXT:   STRDui killed [[LDRDui]], [[COPY]], 0 :: (store (s64) into %ir.g1, align 4, !alias.scope !9, !noalias !10)
   ; MIR-NEXT:   RET_ReallyLR
   %tmp4 = getelementptr inbounds <3 x float>, ptr %g1, i64 0, i64 0
   %tmp5 = load <3 x float>, ptr %g0, align 4, !alias.scope !0, !noalias !1
diff --git a/llvm/test/CodeGen/AMDGPU/GlobalISel/irtranslator-call-return-values.ll b/llvm/test/CodeGen/AMDGPU/GlobalISel/irtranslator-call-return-values.ll
index 96ee15f2eb78b..c7977c339880d 100644
--- a/llvm/test/CodeGen/AMDGPU/GlobalISel/irtranslator-call-return-values.ll
+++ b/llvm/test/CodeGen/AMDGPU/GlobalISel/irtranslator-call-return-values.ll
@@ -84,7 +84,7 @@ define amdgpu_kernel void @test_call_external_i32_func_i32_imm(ptr addrspace(1)
   ; GCN-NEXT:   [[COPY9:%[0-9]+]]:_(p4) = COPY $sgpr8_sgpr9
   ; GCN-NEXT:   [[C:%[0-9]+]]:_(s32) = G_CONSTANT i32 42
   ; GCN-NEXT:   [[INT:%[0-9]+]]:_(p4) = G_INTRINSIC intrinsic(@llvm.amdgcn.kernarg.segment.ptr)
-  ; GCN-NEXT:   [[LOAD:%[0-9]+]]:_(p1) = G_LOAD [[INT]](p4) :: (dereferenceable invariant load (p1) from %ir.out.kernarg.offset1, align 16, addrspace 4)
+  ; GCN-NEXT:   [[LOAD:%[0-9]+]]:_(p1) = G_LOAD [[INT]](p4) :: (dereferenceable invariant load (p1) from %ir.test_call_external_i32_func_i32_imm.kernarg.segment, align 16, addrspace 4)
   ; GCN-NEXT:   ADJCALLSTACKUP 0, 0, implicit-def $scc
   ; GCN-NEXT:   [[GV:%[0-9]+]]:_(p0) = G_GLOBAL_VALUE @external_i32_func_i32
   ; GCN-NEXT:   [[COPY10:%[0-9]+]]:_(p4) = COPY [[COPY8]]
@@ -2957,7 +2957,7 @@ define amdgpu_kernel void @test_call_external_v33i32_func_v33i32_i32(ptr addrspa
   ; GCN-NEXT:   [[COPY9:%[0-9]+]]:_(p4) = COPY $sgpr8_sgpr9
   ; GCN-NEXT:   [[DEF:%[0-9]+]]:_(p1) = G_IMPLICIT_DEF
   ; GCN-NEXT:   [[INT:%[0-9]+]]:_(p4) = G_INTRINSIC intrinsic(@llvm.amdgcn.kernarg.segment.ptr)
-  ; GCN-NEXT:   [[LOAD:%[0-9]+]]:_(p1) = G_LOAD [[INT]](p4) :: (dereferenceable invariant load (p1) from %ir.p.kernarg.offset1, align 16, addrspace 4)
+  ; GCN-NEXT:   [[LOAD:%[0-9]+]]:_(p1) = G_LOAD [[INT]](p4) :: (dereferenceable invariant load (p1) from %ir.test_call_external_v33i32_func_v33i32_i32.kernarg.segment, align 16, addrspace 4)
   ; GCN-NEXT:   [[C:%[0-9]+]]:_(s64) = G_CONSTANT i64 8
   ; GCN-NEXT:   %18:_(p4) = nuw nusw G_PTR_ADD [[INT]], [[C]](s64)
   ; GCN-NEXT:   [[LOAD1:%[0-9]+]]:_(s32) = G_LOAD %18(p4) :: (dereferenceable invariant load (s32) from %ir.idx.kernarg.offset, align 8, addrspace 4)
diff --git a/llvm/test/CodeGen/AMDGPU/GlobalISel/irtranslator-call.ll b/llvm/test/CodeGen/AMDGPU/GlobalISel/irtranslator-call.ll
index 92106d7e1d60d..862093ebf9416 100644
--- a/llvm/test/CodeGen/AMDGPU/GlobalISel/irtranslator-call.ll
+++ b/llvm/test/CodeGen/AMDGPU/GlobalISel/irtranslator-call.ll
@@ -1350,7 +1350,7 @@ define amdgpu_kernel void @test_call_external_void_func_p0_imm(ptr %arg) #0 {
   ; CHECK-NEXT:   [[COPY8:%[0-9]+]]:sgpr_64 = COPY $sgpr4_sgpr5
   ; CHECK-NEXT:   [[COPY9:%[0-9]+]]:_(p4) = COPY $sgpr8_sgpr9
   ; CHECK-NEXT:   [[INT:%[0-9]+]]:_(p4) = G_INTRINSIC intrinsic(@llvm.amdgcn.kernarg.segment.ptr)
-  ; CHECK-NEXT:   [[LOAD:%[0-9]+]]:_(p0) = G_LOAD [[INT]](p4) :: (dereferenceable invariant load (p0) from %ir.arg.kernarg.offset1, align 16, addrspace 4)
+  ; CHECK-NEXT:   [[LOAD:%[0-9]+]]:_(p0) = G_LOAD [[INT]](p4) :: (dereferenceable invariant load (p0) from %ir.test_call_external_void_func_p0_imm.kernarg.segment, align 16, addrspace 4)
   ; CHECK-NEXT:   ADJCALLSTACKUP 0, 0, implicit-def $scc
   ; CHECK-NEXT:   [[GV:%[0-9]+]]:_(p0) = G_GLOBAL_VALUE @external_void_func_p0
   ; CHECK-NEXT:   [[COPY10:%[0-9]+]]:_(p4) = COPY [[COPY8]]
@@ -4654,7 +4654,7 @@ define amdgpu_kernel void @stack_passed_arg_alignment_v32i32_f64(<32 x i32> %val
   ; CHECK-NEXT:   [[COPY8:%[0-9]+]]:sgpr_64 = COPY $sgpr4_sgpr5
   ; CHECK-NEXT:   [[COPY9:%[0-9]+]]:_(p4) = COPY $sgpr8_sgpr9
   ; CHECK-NEXT:   [[INT:%[0-9]+]]:_(p4) = G_INTRINSIC intrinsic(@llvm.amdgcn.kernarg.segment.ptr)
-  ; CHECK-NEXT:   [[LOAD:%[0-9]+]]:_(<32 x s32>) = G_LOAD [[INT]](p4) :: (dereferenceable invariant load (<32 x s32>) from %ir.val.kernarg.offset1, align 16, addrspace 4)
+  ; CHECK-NEXT:   [[LOAD:%[0-9]+]]:_(<32 x s32>) = G_LOAD [[INT]](p4) :: (dereferenceable invariant load (<32 x s32>) from %ir.stack_passed_arg_alignment_v32i32_f64.kernarg.segment, align 16, addrspace 4)
   ; CHECK-NEXT:   [[C:%[0-9]+]]:_(s64) = G_CONSTANT i64 128
   ; CHECK-NEXT:   %18:_(p4) = nuw nusw G_PTR_ADD [[INT]], [[C]](s64)
   ; CHECK-NEXT:   [[LOAD1:%[0-9]+]]:_(s64) = G_LOAD %18(p4) :: (dereferenceable invariant load (s64) from %ir.tmp.kernarg.offset, align 16, addrspace 4)
diff --git a/llvm/test/CodeGen/AMDGPU/GlobalISel/irtranslator-indirect-call.ll b/llvm/test/CodeGen/AMDGPU/GlobalISel/irtranslator-indirect-call.ll
index ac0d5ee78666e..bdf89f785ef99 100644
--- a/llvm/test/CodeGen/AMDGPU/GlobalISel/irtranslator-indirect-call.ll
+++ b/llvm/test/CodeGen/AMDGPU/GlobalISel/irtranslator-indirect-call.ll
@@ -17,7 +17,7 @@ define amdgpu_kernel void @test_indirect_call_sgpr_ptr(ptr %fptr) {
   ; CHECK-NEXT:   [[COPY8:%[0-9]+]]:sgpr_64 = COPY $sgpr4_sgpr5
   ; CHECK-NEXT:   [[COPY9:%[0-9]+]]:_(p4) = COPY $sgpr8_sgpr9
   ; CHECK-NEXT:   [[INT:%[0-9]+]]:_(p4) = G_INTRINSIC intrinsic(@llvm.amdgcn.kernarg.segment.ptr)
-  ; CHECK-NEXT:   [[LOAD:%[0-9]+]]:_(p0) = G_LOAD [[INT]](p4) :: (dereferenceable invariant load (p0) from %ir.fptr.kernarg.offset1, align 16, addrspace 4)
+  ; CHECK-NEXT:   [[LOAD:%[0-9]+]]:_(p0) = G_LOAD [[INT]](p4) :: (dereferenceable invariant load (p0) from %ir.test_indirect_call_sgpr_ptr.kernarg.segment, align 16, addrspace 4)
   ; CHECK-NEXT:   ADJCALLSTACKUP 0, 0, implicit-def $scc
   ; CHECK-NEXT:   [[COPY10:%[0-9]+]]:_(p4) = COPY [[COPY8]]
   ; CHECK-NEXT:   [[COPY11:%[0-9]+]]:_(p4) = COPY [[COPY7]]
diff --git a/llvm/test/CodeGen/AMDGPU/GlobalISel/irtranslator-sibling-call.ll b/llvm/test/CodeGen/AMDGPU/GlobalISel/irtranslator-sibling-call.ll
index ca580d8f29c84..a3f5d927d4b80 100644
--- a/llvm/test/CodeGen/AMDGPU/GlobalISel/irtranslator-sibling-call.ll
+++ b/llvm/test/CodeGen/AMDGPU/GlobalISel/irtranslator-sibling-call.ll
@@ -140,7 +140,7 @@ define amdgpu_kernel void @kernel_call_i32_fastcc_i32_i32_unused_result(i32 %a,
   ; GCN-NEXT:   [[C:%[0-9]+]]:_(s32) = G_CONSTANT i32 0
   ; GCN-NEXT:   [[C1:%[0-9]+]]:_(s32) = G_CONSTANT i32 1
   ; GCN-NEXT:   [[INT:%[0-9]+]]:_(p4) = G_INTRINSIC intrinsic(@llvm.amdgcn.kernarg.segment.ptr)
-  ; GCN-NEXT:   [[LOAD:%[0-9]+]]:_(<2 x s32>) = G_LOAD [[INT]](p4) :: (dereferenceable invariant load (<2 x s32>) from %ir.a.kernarg.offset1, align 16, addrspace 4)
+  ; GCN-NEXT:   [[LOAD:%[0-9]+]]:_(<2 x s32>) = G_LOAD [[INT]](p4) :: (dereferenceable invariant load (<2 x s32>) from %ir.kernel_call_i32_fastcc_i32_i32_unused_result.kernarg.segment, align 16, addrspace 4)
   ; GCN-NEXT:   [[EVEC:%[0-9]+]]:_(s32) = G_EXTRACT_VECTOR_ELT [[LOAD]](<2 x s32>), [[C]](s32)
   ; GCN-NEXT:   [[EVEC1:%[0-9]+]]:_(s32) = G_EXTRACT_VECTOR_ELT [[LOAD]](<2 x s32>), [[C1]](s32)
   ; GCN-NEXT:   ADJCALLSTACKUP 0, 0, implicit-def $scc
diff --git a/llvm/test/CodeGen/AMDGPU/branch-folding-implicit-def-subreg.ll b/llvm/test/CodeGen/AMDGPU/branch-folding-implicit-def-subreg.ll
index ae90cfb631e8d..412e3569c0d05 100644
--- a/llvm/test/CodeGen/AMDGPU/branch-folding-implicit-def-subreg.ll
+++ b/llvm/test/CodeGen/AMDGPU/branch-folding-implicit-def-subreg.ll
@@ -16,8 +16,8 @@ define amdgpu_kernel void @f1(ptr addrspace(1) %arg, ptr addrspace(1) %arg1, i64
   ; GFX90A-NEXT:   renamable $sgpr17 = S_LOAD_DWORD_IMM renamable $sgpr8_sgpr9, 24, 0 :: (dereferenceable invariant load (s32) from %ir.arg4.kernarg.offset.align.down, align 8, addrspace 4)
   ; GFX90A-NEXT:   early-clobber renamable $sgpr20_sgpr21_sgpr22_sgpr23 = S_LOAD_DWORDX4_IMM_ec renamable $sgpr8_sgpr9, 24, 0 :: (dereferenceable invariant load (s128) from %ir.arg6.kernarg.offset.align.down, align 8, addrspace 4)
   ; GFX90A-NEXT:   renamable $sgpr33 = S_LOAD_DWORD_IMM renamable $sgpr8_sgpr9, 40, 0 :: (dereferenceable invariant load (s32) from %ir.arg6.kernarg.offset.align.down + 16, align 8, addrspace 4)
-  ; GFX90A-NEXT:   renamable $sgpr24_sgpr25_sgpr26_sgpr27 = S_LOAD_DWORDX4_IMM renamable $sgpr8_sgpr9, 0, 0 :: (dereferenceable invariant load (s128) from %ir.arg.kernarg.offset1, addrspace 4)
-  ; GFX90A-NEXT:   renamable $sgpr46_sgpr47 = S_LOAD_DWORDX2_IMM renamable $sgpr8_sgpr9, 16, 0 :: (dereferenceable invariant load (s64) from %ir.arg.kernarg.offset1 + 16, align 16, addrspace 4)
+  ; GFX90A-NEXT:   renamable $sgpr24_sgpr25_sgpr26_sgpr27 = S_LOAD_DWORDX4_IMM renamable $sgpr8_sgpr9, 0, 0 :: (dereferenceable invariant load (s128) from %ir.f1.kernarg.segment, addrspace 4)
+  ; GFX90A-NEXT:   renamable $sgpr46_sgpr47 = S_LOAD_DWORDX2_IMM renamable $sgpr8_sgpr9, 16, 0 :: (dereferenceable invariant load (s64) from %ir.f1.kernarg.segment + 16, align 16, addrspace 4)
   ; GFX90A-NEXT:   S_BITCMP1_B32 renamable $sgpr17, 0, implicit-def $scc
   ; GFX90A-NEXT:   renamable $sgpr12_sgpr13 = S_CSELECT_B64 -1, 0, implicit killed $scc
   ; GFX90A-NEXT:   renamable $sgpr34_sgpr35 = S_MOV_B64 -1
diff --git a/llvm/test/CodeGen/AMDGPU/partial-regcopy-and-spill-missed-at-regalloc.ll b/llvm/test/CodeGen/AMDGPU/partial-regcopy-and-spill-missed-at-regalloc.ll
index 663fd98b46bf7..03d3724c232fa 100644
--- a/llvm/test/CodeGen/AMDGPU/partial-regcopy-and-spill-missed-at-regalloc.ll
+++ b/llvm/test/CodeGen/AMDGPU/partial-regcopy-and-spill-missed-at-regalloc.ll
@@ -15,7 +15,7 @@ define amdgpu_kernel void @partial_copy(<4 x i32> %arg) #0 {
   ; REGALLOC-GFX908-NEXT:   INLINEASM &"; def $0", 1 /* sideeffect attdialect */, 6094858 /* regdef:VReg_128 */, def %7
   ; REGALLOC-GFX908-NEXT:   INLINEASM &"; def $0", 1 /* sideeffect attdialect */, 3538954 /* regdef:VReg_64 */, def %8
   ; REGALLOC-GFX908-NEXT:   GLOBAL_STORE_DWORDX4 undef %15:vreg_64, %7, 0, 0, implicit $exec :: (volatile store (s128) into `ptr addrspace(1) poison`, addrspace 1)
-  ; REGALLOC-GFX908-NEXT:   renamable $sgpr0_sgpr1_sgpr2_sgpr3 = S_LOAD_DWORDX4_IMM killed renamable $sgpr4_sgpr5, 0, 0 :: (dereferenceable invariant load (s128) from %ir.arg.kernarg.offset1, addrspace 4)
+  ; REGALLOC-GFX908-NEXT:   renamable $sgpr0_sgpr1_sgpr2_sgpr3 = S_LOAD_DWORDX4_IMM killed renamable $sgpr4_sgpr5, 0, 0 :: (dereferenceable invariant load (s128) from %ir.partial_copy.kernarg.segment, addrspace 4)
   ; REGALLOC-GFX908-NEXT:   [[COPY:%[0-9]+]]:areg_128 = COPY killed renamable $sgpr0_sgpr1_sgpr2_sgpr3
   ; REGALLOC-GFX908-NEXT:   [[V_MOV_B32_e32_:%[0-9]+]]:vgpr_32 = V_MOV_B32_e32 1, implicit $exec
   ; REGALLOC-GFX908-NEXT:   [[V_MOV_B32_e32_1:%[0-9]+]]:vgpr_32 = V_MOV_B32_e32 2, implicit $exec
@@ -40,7 +40,7 @@ define amdgpu_kernel void @partial_copy(<4 x i32> %arg) #0 {
   ; PEI-GFX908-NEXT:   $agpr4 = V_ACCVGPR_WRITE_B32_e64 killed $vgpr1, implicit $exec, implicit killed $vgpr0_vgpr1
   ; PEI-GFX908-NEXT:   renamable $vgpr0_vgpr1_vgpr2_vgpr3 = COPY killed renamable $agpr0_agpr1_agpr2_agpr3, implicit $exec
   ; PEI-GFX908-NEXT:   GLOBAL_STORE_DWORDX4 undef renamable $vgpr0_vgpr1, killed renamable $vgpr0_vgpr1_vgpr2_vgpr3, 0, 0, implicit $exec :: (volatile store (s128) into `ptr addrspace(1) poison`, addrspace 1)
-  ; PEI-GFX908-NEXT:   renamable $sgpr0_sgpr1_sgpr2_sgpr3 = S_LOAD_DWORDX4_IMM killed renamable $sgpr4_sgpr5, 0, 0 :: (dereferenceable invariant load (s128) from %ir.arg.kernarg.offset1, addrspace 4)
+  ; PEI-GFX908-NEXT:   renamable $sgpr0_sgpr1_sgpr2_sgpr3 = S_LOAD_DWORDX4_IMM killed renamable $sgpr4_sgpr5, 0, 0 :: (dereferenceable invariant load (s128) from %ir.partial_copy.kernarg.segment, addrspace 4)
   ; PEI-GFX908-NEXT:   renamable $agpr0_agpr1_agpr2_agpr3 = COPY killed renamable $sgpr0_sgpr1_sgpr2_sgpr3, implicit $exec
   ; PEI-GFX908-NEXT:   renamable $vgpr0 = V_MOV_B32_e32 1, implicit $exec
   ; PEI-GFX908-NEXT:   renamable $vgpr1 = V_MOV_B32_e32 2, implicit $exec
@@ -60,7 +60,7 @@ define amdgpu_kernel void @partial_copy(<4 x i32> %arg) #0 {
   ; REGALLOC-GFX90A-NEXT:   INLINEASM &"; def $0", 1 /* sideeffect attdialect */, 6422538 /* regdef:VReg_128_Align2 */, def %7
   ; REGALLOC-GFX90A-NEXT:   INLINEASM &"; def $0", 1 /* sideeffect attdialect */, 3866634 /* regdef:VReg_64_Align2 */, def %8
   ; REGALLOC-GFX90A-NEXT:   GLOBAL_STORE_DWORDX4 undef %15:vreg_64_align2, %7, 0, 0, implicit $exec :: (volatile store (s128) into `ptr addrspace(1) poison`, addrspace 1)
-  ; REGALLOC-GFX90A-NEXT:   renamable $sgpr0_sgpr1_sgpr2_sgpr3 = S_LOAD_DWORDX4_IMM killed renamable $sgpr4_sgpr5, 0, 0 :: (dereferenceable invariant load (s128) from %ir.arg.kernarg.offset1, addrspace 4)
+  ; REGALLOC-GFX90A-NEXT:   renamable $sgpr0_sgpr1_sgpr2_sgpr3 = S_LOAD_DWORDX4_IMM killed renamable $sgpr4_sgpr5, 0, 0 :: (dereferenceable invariant load (s128) from %ir.partial_copy.kernarg.segment, addrspace 4)
   ; REGALLOC-GFX90A-NEXT:   [[COPY:%[0-9]+]]:areg_128_align2 = COPY killed renamable $sgpr0_sgpr1_sgpr2_sgpr3
   ; REGALLOC-GFX90A-NEXT:   [[V_MOV_B32_e32_:%[0-9]+]]:vgpr_32 = V_MOV_B32_e32 1, implicit $exec
   ; REGALLOC-GFX90A-NEXT:   [[V_MOV_B32_e32_1:%[0-9]+]]:vgpr_32 = V_MOV_B32_e32 2, implicit $exec
@@ -83,7 +83,7 @@ define amdgpu_kernel void @partial_copy(<4 x i32> %arg) #0 {
   ; PEI-GFX90A-NEXT:   BUFFER_STORE_DWORD_OFFSET killed $vgpr0, $sgpr12_sgpr13_sgpr14_sgpr15, 0, 0, 0, 0, implicit $exec, implicit-def $vgpr0_vgpr1, implicit $vgpr0_vgpr1 :: (store (s32) into %stack.0, addrspace 5)
   ; PEI-GFX90A-NEXT:   $agpr4 = V_ACCVGPR_WRITE_B32_e64 killed $vgpr1, implicit $exec, implicit killed $vgpr0_vgpr1
   ; PEI-GFX90A-NEXT:   GLOBAL_STORE_DWORDX4 undef renamable $vgpr0_vgpr1, killed renamable $agpr0_agpr1_agpr2_agpr3, 0, 0, implicit $exec :: (volatile store (s128) into `ptr addrspace(1) poison`, addrspace 1)
-  ; PEI-GFX90A-NEXT:   renamable $sgpr0_sgpr1_sgpr2_sgpr3 = S_LOAD_DWORDX4_IMM killed renamable $sgpr4_sgpr5, 0, 0 :: (dereferenceable invariant load (s128) from %ir.arg.kernarg.offset1, addrspace 4)
+  ; PEI-GFX90A-NEXT:   renamable $sgpr0_sgpr1_sgpr2_sgpr3 = S_LOAD_DWORDX4_IMM killed renamable $sgpr4_sgpr5, 0, 0 :: (dereferenceable invariant load (s128) from %ir.partial_copy.kernarg.segment, addrspace 4)
   ; PEI-GFX90A-NEXT:   renamable $agpr0_agpr1_agpr2_agpr3 = COPY killed renamable $sgpr0_sgpr1_sgpr2_sgpr3, implicit $exec
   ; PEI-GFX90A-NEXT:   renamable $vgpr0 = V_MOV_B32_e32 1, implicit $exec
   ; PEI-GFX90A-NEXT:   renamable $vgpr1 = V_MOV_B32_e32 2, implicit $exec
diff ...
[truncated]

@llvmbot
Copy link
Member

llvmbot commented Jul 3, 2025

@llvm/pr-subscribers-llvm-transforms

Author: Philip Reames (preames)

Changes

The change itself is very straight forward, it's mostly just updating tests.

Somewhat surprisingly, this does cause some minor codegen differences in a few tests.

Patch is 89.99 KiB, truncated to 20.00 KiB below, full version: https://github.com/llvm/llvm-project/pull/146961.diff

24 Files Affected:

  • (modified) llvm/lib/CodeGen/CodeGenPrepare.cpp (+7)
  • (modified) llvm/test/CodeGen/AArch64/memcpy-scoped-aa.ll (+10-10)
  • (modified) llvm/test/CodeGen/AArch64/merge-scoped-aa-store.ll (+2-2)
  • (modified) llvm/test/CodeGen/AMDGPU/GlobalISel/irtranslator-call-return-values.ll (+2-2)
  • (modified) llvm/test/CodeGen/AMDGPU/GlobalISel/irtranslator-call.ll (+2-2)
  • (modified) llvm/test/CodeGen/AMDGPU/GlobalISel/irtranslator-indirect-call.ll (+1-1)
  • (modified) llvm/test/CodeGen/AMDGPU/GlobalISel/irtranslator-sibling-call.ll (+1-1)
  • (modified) llvm/test/CodeGen/AMDGPU/branch-folding-implicit-def-subreg.ll (+2-2)
  • (modified) llvm/test/CodeGen/AMDGPU/partial-regcopy-and-spill-missed-at-regalloc.ll (+4-4)
  • (modified) llvm/test/CodeGen/AMDGPU/spill-vector-superclass.ll (+1-1)
  • (modified) llvm/test/CodeGen/SystemZ/isel-debug.ll (+1-1)
  • (modified) llvm/test/CodeGen/Thumb2/active_lane_mask.ll (+2-3)
  • (modified) llvm/test/CodeGen/X86/GlobalISel/x86_64-irtranslator-struct-return.ll (+27-27)
  • (modified) llvm/test/CodeGen/X86/memcpy-scoped-aa.ll (+18-18)
  • (modified) llvm/test/CodeGen/X86/merge-store-partially-alias-loads.ll (+1-1)
  • (modified) llvm/test/CodeGen/X86/pr44140.ll (+6-6)
  • (modified) llvm/test/CodeGen/X86/sse-intrinsics-fast-isel.ll (+120-160)
  • (modified) llvm/test/CodeGen/X86/statepoint-cmp-sunk-past-statepoint.ll (+4-4)
  • (modified) llvm/test/DebugInfo/Generic/assignment-tracking/codegenprepare/sunk-addr.ll (+1-2)
  • (modified) llvm/test/DebugInfo/X86/bitcast-di.ll (+2-33)
  • (modified) llvm/test/Transforms/CodeGenPrepare/X86/masked-gather-struct-gep.ll (+2-3)
  • (modified) llvm/test/Transforms/CodeGenPrepare/X86/sink-addr-recreate.ll (+4-7)
  • (modified) llvm/test/Transforms/CodeGenPrepare/X86/sink-addrmode-base.ll (+39-24)
  • (modified) llvm/test/Transforms/CodeGenPrepare/X86/sink-addrmode-two-phi.ll (+1-2)
diff --git a/llvm/lib/CodeGen/CodeGenPrepare.cpp b/llvm/lib/CodeGen/CodeGenPrepare.cpp
index 9bbb89e37865d..bb397c1c35bf8 100644
--- a/llvm/lib/CodeGen/CodeGenPrepare.cpp
+++ b/llvm/lib/CodeGen/CodeGenPrepare.cpp
@@ -8720,6 +8720,13 @@ bool CodeGenPrepare::optimizeInst(Instruction *I, ModifyDT &ModifiedDT) {
     if (isa<Constant>(CI->getOperand(0)))
       return AnyChange;
 
+    // Remove noop bitcasts
+    if (isa<BitCastInst>(I) && I->getType() == I->getOperand(0)->getType()) {
+      replaceAllUsesWith(I, I->getOperand(0), FreshBBs, IsHugeFunc);
+      I->eraseFromParent();
+      return true;
+    }
+
     if (OptimizeNoopCopyExpression(CI, *TLI, *DL))
       return true;
 
diff --git a/llvm/test/CodeGen/AArch64/memcpy-scoped-aa.ll b/llvm/test/CodeGen/AArch64/memcpy-scoped-aa.ll
index 7dde168024278..f77d300adb2c7 100644
--- a/llvm/test/CodeGen/AArch64/memcpy-scoped-aa.ll
+++ b/llvm/test/CodeGen/AArch64/memcpy-scoped-aa.ll
@@ -9,8 +9,8 @@
 ; MIR-DAG: ![[SET1:[0-9]+]] = !{![[SCOPE1]]}
 
 ; MIR-LABEL: name: test_memcpy
-; MIR:      %2:fpr128 = LDRQui %0, 1 :: (load (s128) from %ir.p1, align 4, !alias.scope ![[SET0]], !noalias ![[SET1]])
-; MIR-NEXT: STRQui killed %2, %0, 0 :: (store (s128) into %ir.p0, align 4, !alias.scope ![[SET0]], !noalias ![[SET1]])
+; MIR:      %2:fpr128 = LDRQui %0, 1 :: (load (s128) from %ir.add.ptr, align 4, !alias.scope ![[SET0]], !noalias ![[SET1]])
+; MIR-NEXT: STRQui killed %2, %0, 0 :: (store (s128) into %ir.p, align 4, !alias.scope ![[SET0]], !noalias ![[SET1]])
 define i32 @test_memcpy(ptr nocapture %p, ptr nocapture readonly %q) {
 ; CHECK-LABEL: test_memcpy:
 ; CHECK:       // %bb.0:
@@ -32,8 +32,8 @@ define i32 @test_memcpy(ptr nocapture %p, ptr nocapture readonly %q) {
 }
 
 ; MIR-LABEL: name: test_memcpy_inline
-; MIR:      %2:fpr128 = LDRQui %0, 1 :: (load (s128) from %ir.p1, align 4, !alias.scope ![[SET0]], !noalias ![[SET1]])
-; MIR-NEXT: STRQui killed %2, %0, 0 :: (store (s128) into %ir.p0, align 4, !alias.scope ![[SET0]], !noalias ![[SET1]])
+; MIR:      %2:fpr128 = LDRQui %0, 1 :: (load (s128) from %ir.add.ptr, align 4, !alias.scope ![[SET0]], !noalias ![[SET1]])
+; MIR-NEXT: STRQui killed %2, %0, 0 :: (store (s128) into %ir.p, align 4, !alias.scope ![[SET0]], !noalias ![[SET1]])
 define i32 @test_memcpy_inline(ptr nocapture %p, ptr nocapture readonly %q) {
 ; CHECK-LABEL: test_memcpy_inline:
 ; CHECK:       // %bb.0:
@@ -55,8 +55,8 @@ define i32 @test_memcpy_inline(ptr nocapture %p, ptr nocapture readonly %q) {
 }
 
 ; MIR-LABEL: name: test_memmove
-; MIR:      %2:fpr128 = LDRQui %0, 1 :: (load (s128) from %ir.p1, align 4, !alias.scope ![[SET0]], !noalias ![[SET1]])
-; MIR-NEXT: STRQui killed %2, %0, 0 :: (store (s128) into %ir.p0, align 4, !alias.scope ![[SET0]], !noalias ![[SET1]])
+; MIR:      %2:fpr128 = LDRQui %0, 1 :: (load (s128) from %ir.add.ptr, align 4, !alias.scope ![[SET0]], !noalias ![[SET1]])
+; MIR-NEXT: STRQui killed %2, %0, 0 :: (store (s128) into %ir.p, align 4, !alias.scope ![[SET0]], !noalias ![[SET1]])
 define i32 @test_memmove(ptr nocapture %p, ptr nocapture readonly %q) {
 ; CHECK-LABEL: test_memmove:
 ; CHECK:       // %bb.0:
@@ -79,8 +79,8 @@ define i32 @test_memmove(ptr nocapture %p, ptr nocapture readonly %q) {
 
 ; MIR-LABEL: name: test_memset
 ; MIR:      %2:gpr64 = MOVi64imm -6148914691236517206
-; MIR-NEXT: STRXui %2, %0, 1 :: (store (s64) into %ir.p0 + 8, align 4, !alias.scope ![[SET0]], !noalias ![[SET1]])
-; MIR-NEXT: STRXui %2, %0, 0 :: (store (s64) into %ir.p0, align 4, !alias.scope ![[SET0]], !noalias ![[SET1]])
+; MIR-NEXT: STRXui %2, %0, 1 :: (store (s64) into %ir.p + 8, align 4, !alias.scope ![[SET0]], !noalias ![[SET1]])
+; MIR-NEXT: STRXui %2, %0, 0 :: (store (s64) into %ir.p, align 4, !alias.scope ![[SET0]], !noalias ![[SET1]])
 define i32 @test_memset(ptr nocapture %p, ptr nocapture readonly %q) {
 ; CHECK-LABEL: test_memset:
 ; CHECK:       // %bb.0:
@@ -100,8 +100,8 @@ define i32 @test_memset(ptr nocapture %p, ptr nocapture readonly %q) {
 }
 
 ; MIR-LABEL: name: test_mempcpy
-; MIR:      %2:fpr128 = LDRQui %0, 1 :: (load (s128) from %ir.p1, align 1, !alias.scope ![[SET0]], !noalias ![[SET1]])
-; MIR-NEXT: STRQui killed %2, %0, 0 :: (store (s128) into %ir.p0, align 1, !alias.scope ![[SET0]], !noalias ![[SET1]])
+; MIR:      %2:fpr128 = LDRQui %0, 1 :: (load (s128) from %ir.add.ptr, align 1, !alias.scope ![[SET0]], !noalias ![[SET1]])
+; MIR-NEXT: STRQui killed %2, %0, 0 :: (store (s128) into %ir.p, align 1, !alias.scope ![[SET0]], !noalias ![[SET1]])
 define i32 @test_mempcpy(ptr nocapture %p, ptr nocapture readonly %q) {
 ; CHECK-LABEL: test_mempcpy:
 ; CHECK:       // %bb.0:
diff --git a/llvm/test/CodeGen/AArch64/merge-scoped-aa-store.ll b/llvm/test/CodeGen/AArch64/merge-scoped-aa-store.ll
index 6b2a8c5aaf8b8..42fda4b4309af 100644
--- a/llvm/test/CodeGen/AArch64/merge-scoped-aa-store.ll
+++ b/llvm/test/CodeGen/AArch64/merge-scoped-aa-store.ll
@@ -21,7 +21,7 @@ define void @blam0(ptr %g0, ptr %g1) {
   ; MIR-NEXT:   [[COPY:%[0-9]+]]:gpr64common = COPY $x1
   ; MIR-NEXT:   [[COPY1:%[0-9]+]]:gpr64common = COPY $x0
   ; MIR-NEXT:   [[LDRDui:%[0-9]+]]:fpr64 = LDRDui [[COPY1]], 0 :: (load (s64) from %ir.g0, align 4, !alias.scope !0, !noalias !3)
-  ; MIR-NEXT:   STRDui killed [[LDRDui]], [[COPY]], 0 :: (store (s64) into %ir.tmp41, align 4, !alias.scope !3, !noalias !0)
+  ; MIR-NEXT:   STRDui killed [[LDRDui]], [[COPY]], 0 :: (store (s64) into %ir.g1, align 4, !alias.scope !3, !noalias !0)
   ; MIR-NEXT:   RET_ReallyLR
   %tmp4 = getelementptr inbounds <3 x float>, ptr %g1, i64 0, i64 0
   %tmp5 = load <3 x float>, ptr %g0, align 4, !alias.scope !0, !noalias !1
@@ -43,7 +43,7 @@ define void @blam1(ptr %g0, ptr %g1) {
   ; MIR-NEXT:   [[COPY:%[0-9]+]]:gpr64common = COPY $x1
   ; MIR-NEXT:   [[COPY1:%[0-9]+]]:gpr64common = COPY $x0
   ; MIR-NEXT:   [[LDRDui:%[0-9]+]]:fpr64 = LDRDui [[COPY1]], 0 :: (load (s64) from %ir.g0, align 4, !alias.scope !0, !noalias !3)
-  ; MIR-NEXT:   STRDui killed [[LDRDui]], [[COPY]], 0 :: (store (s64) into %ir.tmp41, align 4, !alias.scope !9, !noalias !10)
+  ; MIR-NEXT:   STRDui killed [[LDRDui]], [[COPY]], 0 :: (store (s64) into %ir.g1, align 4, !alias.scope !9, !noalias !10)
   ; MIR-NEXT:   RET_ReallyLR
   %tmp4 = getelementptr inbounds <3 x float>, ptr %g1, i64 0, i64 0
   %tmp5 = load <3 x float>, ptr %g0, align 4, !alias.scope !0, !noalias !1
diff --git a/llvm/test/CodeGen/AMDGPU/GlobalISel/irtranslator-call-return-values.ll b/llvm/test/CodeGen/AMDGPU/GlobalISel/irtranslator-call-return-values.ll
index 96ee15f2eb78b..c7977c339880d 100644
--- a/llvm/test/CodeGen/AMDGPU/GlobalISel/irtranslator-call-return-values.ll
+++ b/llvm/test/CodeGen/AMDGPU/GlobalISel/irtranslator-call-return-values.ll
@@ -84,7 +84,7 @@ define amdgpu_kernel void @test_call_external_i32_func_i32_imm(ptr addrspace(1)
   ; GCN-NEXT:   [[COPY9:%[0-9]+]]:_(p4) = COPY $sgpr8_sgpr9
   ; GCN-NEXT:   [[C:%[0-9]+]]:_(s32) = G_CONSTANT i32 42
   ; GCN-NEXT:   [[INT:%[0-9]+]]:_(p4) = G_INTRINSIC intrinsic(@llvm.amdgcn.kernarg.segment.ptr)
-  ; GCN-NEXT:   [[LOAD:%[0-9]+]]:_(p1) = G_LOAD [[INT]](p4) :: (dereferenceable invariant load (p1) from %ir.out.kernarg.offset1, align 16, addrspace 4)
+  ; GCN-NEXT:   [[LOAD:%[0-9]+]]:_(p1) = G_LOAD [[INT]](p4) :: (dereferenceable invariant load (p1) from %ir.test_call_external_i32_func_i32_imm.kernarg.segment, align 16, addrspace 4)
   ; GCN-NEXT:   ADJCALLSTACKUP 0, 0, implicit-def $scc
   ; GCN-NEXT:   [[GV:%[0-9]+]]:_(p0) = G_GLOBAL_VALUE @external_i32_func_i32
   ; GCN-NEXT:   [[COPY10:%[0-9]+]]:_(p4) = COPY [[COPY8]]
@@ -2957,7 +2957,7 @@ define amdgpu_kernel void @test_call_external_v33i32_func_v33i32_i32(ptr addrspa
   ; GCN-NEXT:   [[COPY9:%[0-9]+]]:_(p4) = COPY $sgpr8_sgpr9
   ; GCN-NEXT:   [[DEF:%[0-9]+]]:_(p1) = G_IMPLICIT_DEF
   ; GCN-NEXT:   [[INT:%[0-9]+]]:_(p4) = G_INTRINSIC intrinsic(@llvm.amdgcn.kernarg.segment.ptr)
-  ; GCN-NEXT:   [[LOAD:%[0-9]+]]:_(p1) = G_LOAD [[INT]](p4) :: (dereferenceable invariant load (p1) from %ir.p.kernarg.offset1, align 16, addrspace 4)
+  ; GCN-NEXT:   [[LOAD:%[0-9]+]]:_(p1) = G_LOAD [[INT]](p4) :: (dereferenceable invariant load (p1) from %ir.test_call_external_v33i32_func_v33i32_i32.kernarg.segment, align 16, addrspace 4)
   ; GCN-NEXT:   [[C:%[0-9]+]]:_(s64) = G_CONSTANT i64 8
   ; GCN-NEXT:   %18:_(p4) = nuw nusw G_PTR_ADD [[INT]], [[C]](s64)
   ; GCN-NEXT:   [[LOAD1:%[0-9]+]]:_(s32) = G_LOAD %18(p4) :: (dereferenceable invariant load (s32) from %ir.idx.kernarg.offset, align 8, addrspace 4)
diff --git a/llvm/test/CodeGen/AMDGPU/GlobalISel/irtranslator-call.ll b/llvm/test/CodeGen/AMDGPU/GlobalISel/irtranslator-call.ll
index 92106d7e1d60d..862093ebf9416 100644
--- a/llvm/test/CodeGen/AMDGPU/GlobalISel/irtranslator-call.ll
+++ b/llvm/test/CodeGen/AMDGPU/GlobalISel/irtranslator-call.ll
@@ -1350,7 +1350,7 @@ define amdgpu_kernel void @test_call_external_void_func_p0_imm(ptr %arg) #0 {
   ; CHECK-NEXT:   [[COPY8:%[0-9]+]]:sgpr_64 = COPY $sgpr4_sgpr5
   ; CHECK-NEXT:   [[COPY9:%[0-9]+]]:_(p4) = COPY $sgpr8_sgpr9
   ; CHECK-NEXT:   [[INT:%[0-9]+]]:_(p4) = G_INTRINSIC intrinsic(@llvm.amdgcn.kernarg.segment.ptr)
-  ; CHECK-NEXT:   [[LOAD:%[0-9]+]]:_(p0) = G_LOAD [[INT]](p4) :: (dereferenceable invariant load (p0) from %ir.arg.kernarg.offset1, align 16, addrspace 4)
+  ; CHECK-NEXT:   [[LOAD:%[0-9]+]]:_(p0) = G_LOAD [[INT]](p4) :: (dereferenceable invariant load (p0) from %ir.test_call_external_void_func_p0_imm.kernarg.segment, align 16, addrspace 4)
   ; CHECK-NEXT:   ADJCALLSTACKUP 0, 0, implicit-def $scc
   ; CHECK-NEXT:   [[GV:%[0-9]+]]:_(p0) = G_GLOBAL_VALUE @external_void_func_p0
   ; CHECK-NEXT:   [[COPY10:%[0-9]+]]:_(p4) = COPY [[COPY8]]
@@ -4654,7 +4654,7 @@ define amdgpu_kernel void @stack_passed_arg_alignment_v32i32_f64(<32 x i32> %val
   ; CHECK-NEXT:   [[COPY8:%[0-9]+]]:sgpr_64 = COPY $sgpr4_sgpr5
   ; CHECK-NEXT:   [[COPY9:%[0-9]+]]:_(p4) = COPY $sgpr8_sgpr9
   ; CHECK-NEXT:   [[INT:%[0-9]+]]:_(p4) = G_INTRINSIC intrinsic(@llvm.amdgcn.kernarg.segment.ptr)
-  ; CHECK-NEXT:   [[LOAD:%[0-9]+]]:_(<32 x s32>) = G_LOAD [[INT]](p4) :: (dereferenceable invariant load (<32 x s32>) from %ir.val.kernarg.offset1, align 16, addrspace 4)
+  ; CHECK-NEXT:   [[LOAD:%[0-9]+]]:_(<32 x s32>) = G_LOAD [[INT]](p4) :: (dereferenceable invariant load (<32 x s32>) from %ir.stack_passed_arg_alignment_v32i32_f64.kernarg.segment, align 16, addrspace 4)
   ; CHECK-NEXT:   [[C:%[0-9]+]]:_(s64) = G_CONSTANT i64 128
   ; CHECK-NEXT:   %18:_(p4) = nuw nusw G_PTR_ADD [[INT]], [[C]](s64)
   ; CHECK-NEXT:   [[LOAD1:%[0-9]+]]:_(s64) = G_LOAD %18(p4) :: (dereferenceable invariant load (s64) from %ir.tmp.kernarg.offset, align 16, addrspace 4)
diff --git a/llvm/test/CodeGen/AMDGPU/GlobalISel/irtranslator-indirect-call.ll b/llvm/test/CodeGen/AMDGPU/GlobalISel/irtranslator-indirect-call.ll
index ac0d5ee78666e..bdf89f785ef99 100644
--- a/llvm/test/CodeGen/AMDGPU/GlobalISel/irtranslator-indirect-call.ll
+++ b/llvm/test/CodeGen/AMDGPU/GlobalISel/irtranslator-indirect-call.ll
@@ -17,7 +17,7 @@ define amdgpu_kernel void @test_indirect_call_sgpr_ptr(ptr %fptr) {
   ; CHECK-NEXT:   [[COPY8:%[0-9]+]]:sgpr_64 = COPY $sgpr4_sgpr5
   ; CHECK-NEXT:   [[COPY9:%[0-9]+]]:_(p4) = COPY $sgpr8_sgpr9
   ; CHECK-NEXT:   [[INT:%[0-9]+]]:_(p4) = G_INTRINSIC intrinsic(@llvm.amdgcn.kernarg.segment.ptr)
-  ; CHECK-NEXT:   [[LOAD:%[0-9]+]]:_(p0) = G_LOAD [[INT]](p4) :: (dereferenceable invariant load (p0) from %ir.fptr.kernarg.offset1, align 16, addrspace 4)
+  ; CHECK-NEXT:   [[LOAD:%[0-9]+]]:_(p0) = G_LOAD [[INT]](p4) :: (dereferenceable invariant load (p0) from %ir.test_indirect_call_sgpr_ptr.kernarg.segment, align 16, addrspace 4)
   ; CHECK-NEXT:   ADJCALLSTACKUP 0, 0, implicit-def $scc
   ; CHECK-NEXT:   [[COPY10:%[0-9]+]]:_(p4) = COPY [[COPY8]]
   ; CHECK-NEXT:   [[COPY11:%[0-9]+]]:_(p4) = COPY [[COPY7]]
diff --git a/llvm/test/CodeGen/AMDGPU/GlobalISel/irtranslator-sibling-call.ll b/llvm/test/CodeGen/AMDGPU/GlobalISel/irtranslator-sibling-call.ll
index ca580d8f29c84..a3f5d927d4b80 100644
--- a/llvm/test/CodeGen/AMDGPU/GlobalISel/irtranslator-sibling-call.ll
+++ b/llvm/test/CodeGen/AMDGPU/GlobalISel/irtranslator-sibling-call.ll
@@ -140,7 +140,7 @@ define amdgpu_kernel void @kernel_call_i32_fastcc_i32_i32_unused_result(i32 %a,
   ; GCN-NEXT:   [[C:%[0-9]+]]:_(s32) = G_CONSTANT i32 0
   ; GCN-NEXT:   [[C1:%[0-9]+]]:_(s32) = G_CONSTANT i32 1
   ; GCN-NEXT:   [[INT:%[0-9]+]]:_(p4) = G_INTRINSIC intrinsic(@llvm.amdgcn.kernarg.segment.ptr)
-  ; GCN-NEXT:   [[LOAD:%[0-9]+]]:_(<2 x s32>) = G_LOAD [[INT]](p4) :: (dereferenceable invariant load (<2 x s32>) from %ir.a.kernarg.offset1, align 16, addrspace 4)
+  ; GCN-NEXT:   [[LOAD:%[0-9]+]]:_(<2 x s32>) = G_LOAD [[INT]](p4) :: (dereferenceable invariant load (<2 x s32>) from %ir.kernel_call_i32_fastcc_i32_i32_unused_result.kernarg.segment, align 16, addrspace 4)
   ; GCN-NEXT:   [[EVEC:%[0-9]+]]:_(s32) = G_EXTRACT_VECTOR_ELT [[LOAD]](<2 x s32>), [[C]](s32)
   ; GCN-NEXT:   [[EVEC1:%[0-9]+]]:_(s32) = G_EXTRACT_VECTOR_ELT [[LOAD]](<2 x s32>), [[C1]](s32)
   ; GCN-NEXT:   ADJCALLSTACKUP 0, 0, implicit-def $scc
diff --git a/llvm/test/CodeGen/AMDGPU/branch-folding-implicit-def-subreg.ll b/llvm/test/CodeGen/AMDGPU/branch-folding-implicit-def-subreg.ll
index ae90cfb631e8d..412e3569c0d05 100644
--- a/llvm/test/CodeGen/AMDGPU/branch-folding-implicit-def-subreg.ll
+++ b/llvm/test/CodeGen/AMDGPU/branch-folding-implicit-def-subreg.ll
@@ -16,8 +16,8 @@ define amdgpu_kernel void @f1(ptr addrspace(1) %arg, ptr addrspace(1) %arg1, i64
   ; GFX90A-NEXT:   renamable $sgpr17 = S_LOAD_DWORD_IMM renamable $sgpr8_sgpr9, 24, 0 :: (dereferenceable invariant load (s32) from %ir.arg4.kernarg.offset.align.down, align 8, addrspace 4)
   ; GFX90A-NEXT:   early-clobber renamable $sgpr20_sgpr21_sgpr22_sgpr23 = S_LOAD_DWORDX4_IMM_ec renamable $sgpr8_sgpr9, 24, 0 :: (dereferenceable invariant load (s128) from %ir.arg6.kernarg.offset.align.down, align 8, addrspace 4)
   ; GFX90A-NEXT:   renamable $sgpr33 = S_LOAD_DWORD_IMM renamable $sgpr8_sgpr9, 40, 0 :: (dereferenceable invariant load (s32) from %ir.arg6.kernarg.offset.align.down + 16, align 8, addrspace 4)
-  ; GFX90A-NEXT:   renamable $sgpr24_sgpr25_sgpr26_sgpr27 = S_LOAD_DWORDX4_IMM renamable $sgpr8_sgpr9, 0, 0 :: (dereferenceable invariant load (s128) from %ir.arg.kernarg.offset1, addrspace 4)
-  ; GFX90A-NEXT:   renamable $sgpr46_sgpr47 = S_LOAD_DWORDX2_IMM renamable $sgpr8_sgpr9, 16, 0 :: (dereferenceable invariant load (s64) from %ir.arg.kernarg.offset1 + 16, align 16, addrspace 4)
+  ; GFX90A-NEXT:   renamable $sgpr24_sgpr25_sgpr26_sgpr27 = S_LOAD_DWORDX4_IMM renamable $sgpr8_sgpr9, 0, 0 :: (dereferenceable invariant load (s128) from %ir.f1.kernarg.segment, addrspace 4)
+  ; GFX90A-NEXT:   renamable $sgpr46_sgpr47 = S_LOAD_DWORDX2_IMM renamable $sgpr8_sgpr9, 16, 0 :: (dereferenceable invariant load (s64) from %ir.f1.kernarg.segment + 16, align 16, addrspace 4)
   ; GFX90A-NEXT:   S_BITCMP1_B32 renamable $sgpr17, 0, implicit-def $scc
   ; GFX90A-NEXT:   renamable $sgpr12_sgpr13 = S_CSELECT_B64 -1, 0, implicit killed $scc
   ; GFX90A-NEXT:   renamable $sgpr34_sgpr35 = S_MOV_B64 -1
diff --git a/llvm/test/CodeGen/AMDGPU/partial-regcopy-and-spill-missed-at-regalloc.ll b/llvm/test/CodeGen/AMDGPU/partial-regcopy-and-spill-missed-at-regalloc.ll
index 663fd98b46bf7..03d3724c232fa 100644
--- a/llvm/test/CodeGen/AMDGPU/partial-regcopy-and-spill-missed-at-regalloc.ll
+++ b/llvm/test/CodeGen/AMDGPU/partial-regcopy-and-spill-missed-at-regalloc.ll
@@ -15,7 +15,7 @@ define amdgpu_kernel void @partial_copy(<4 x i32> %arg) #0 {
   ; REGALLOC-GFX908-NEXT:   INLINEASM &"; def $0", 1 /* sideeffect attdialect */, 6094858 /* regdef:VReg_128 */, def %7
   ; REGALLOC-GFX908-NEXT:   INLINEASM &"; def $0", 1 /* sideeffect attdialect */, 3538954 /* regdef:VReg_64 */, def %8
   ; REGALLOC-GFX908-NEXT:   GLOBAL_STORE_DWORDX4 undef %15:vreg_64, %7, 0, 0, implicit $exec :: (volatile store (s128) into `ptr addrspace(1) poison`, addrspace 1)
-  ; REGALLOC-GFX908-NEXT:   renamable $sgpr0_sgpr1_sgpr2_sgpr3 = S_LOAD_DWORDX4_IMM killed renamable $sgpr4_sgpr5, 0, 0 :: (dereferenceable invariant load (s128) from %ir.arg.kernarg.offset1, addrspace 4)
+  ; REGALLOC-GFX908-NEXT:   renamable $sgpr0_sgpr1_sgpr2_sgpr3 = S_LOAD_DWORDX4_IMM killed renamable $sgpr4_sgpr5, 0, 0 :: (dereferenceable invariant load (s128) from %ir.partial_copy.kernarg.segment, addrspace 4)
   ; REGALLOC-GFX908-NEXT:   [[COPY:%[0-9]+]]:areg_128 = COPY killed renamable $sgpr0_sgpr1_sgpr2_sgpr3
   ; REGALLOC-GFX908-NEXT:   [[V_MOV_B32_e32_:%[0-9]+]]:vgpr_32 = V_MOV_B32_e32 1, implicit $exec
   ; REGALLOC-GFX908-NEXT:   [[V_MOV_B32_e32_1:%[0-9]+]]:vgpr_32 = V_MOV_B32_e32 2, implicit $exec
@@ -40,7 +40,7 @@ define amdgpu_kernel void @partial_copy(<4 x i32> %arg) #0 {
   ; PEI-GFX908-NEXT:   $agpr4 = V_ACCVGPR_WRITE_B32_e64 killed $vgpr1, implicit $exec, implicit killed $vgpr0_vgpr1
   ; PEI-GFX908-NEXT:   renamable $vgpr0_vgpr1_vgpr2_vgpr3 = COPY killed renamable $agpr0_agpr1_agpr2_agpr3, implicit $exec
   ; PEI-GFX908-NEXT:   GLOBAL_STORE_DWORDX4 undef renamable $vgpr0_vgpr1, killed renamable $vgpr0_vgpr1_vgpr2_vgpr3, 0, 0, implicit $exec :: (volatile store (s128) into `ptr addrspace(1) poison`, addrspace 1)
-  ; PEI-GFX908-NEXT:   renamable $sgpr0_sgpr1_sgpr2_sgpr3 = S_LOAD_DWORDX4_IMM killed renamable $sgpr4_sgpr5, 0, 0 :: (dereferenceable invariant load (s128) from %ir.arg.kernarg.offset1, addrspace 4)
+  ; PEI-GFX908-NEXT:   renamable $sgpr0_sgpr1_sgpr2_sgpr3 = S_LOAD_DWORDX4_IMM killed renamable $sgpr4_sgpr5, 0, 0 :: (dereferenceable invariant load (s128) from %ir.partial_copy.kernarg.segment, addrspace 4)
   ; PEI-GFX908-NEXT:   renamable $agpr0_agpr1_agpr2_agpr3 = COPY killed renamable $sgpr0_sgpr1_sgpr2_sgpr3, implicit $exec
   ; PEI-GFX908-NEXT:   renamable $vgpr0 = V_MOV_B32_e32 1, implicit $exec
   ; PEI-GFX908-NEXT:   renamable $vgpr1 = V_MOV_B32_e32 2, implicit $exec
@@ -60,7 +60,7 @@ define amdgpu_kernel void @partial_copy(<4 x i32> %arg) #0 {
   ; REGALLOC-GFX90A-NEXT:   INLINEASM &"; def $0", 1 /* sideeffect attdialect */, 6422538 /* regdef:VReg_128_Align2 */, def %7
   ; REGALLOC-GFX90A-NEXT:   INLINEASM &"; def $0", 1 /* sideeffect attdialect */, 3866634 /* regdef:VReg_64_Align2 */, def %8
   ; REGALLOC-GFX90A-NEXT:   GLOBAL_STORE_DWORDX4 undef %15:vreg_64_align2, %7, 0, 0, implicit $exec :: (volatile store (s128) into `ptr addrspace(1) poison`, addrspace 1)
-  ; REGALLOC-GFX90A-NEXT:   renamable $sgpr0_sgpr1_sgpr2_sgpr3 = S_LOAD_DWORDX4_IMM killed renamable $sgpr4_sgpr5, 0, 0 :: (dereferenceable invariant load (s128) from %ir.arg.kernarg.offset1, addrspace 4)
+  ; REGALLOC-GFX90A-NEXT:   renamable $sgpr0_sgpr1_sgpr2_sgpr3 = S_LOAD_DWORDX4_IMM killed renamable $sgpr4_sgpr5, 0, 0 :: (dereferenceable invariant load (s128) from %ir.partial_copy.kernarg.segment, addrspace 4)
   ; REGALLOC-GFX90A-NEXT:   [[COPY:%[0-9]+]]:areg_128_align2 = COPY killed renamable $sgpr0_sgpr1_sgpr2_sgpr3
   ; REGALLOC-GFX90A-NEXT:   [[V_MOV_B32_e32_:%[0-9]+]]:vgpr_32 = V_MOV_B32_e32 1, implicit $exec
   ; REGALLOC-GFX90A-NEXT:   [[V_MOV_B32_e32_1:%[0-9]+]]:vgpr_32 = V_MOV_B32_e32 2, implicit $exec
@@ -83,7 +83,7 @@ define amdgpu_kernel void @partial_copy(<4 x i32> %arg) #0 {
   ; PEI-GFX90A-NEXT:   BUFFER_STORE_DWORD_OFFSET killed $vgpr0, $sgpr12_sgpr13_sgpr14_sgpr15, 0, 0, 0, 0, implicit $exec, implicit-def $vgpr0_vgpr1, implicit $vgpr0_vgpr1 :: (store (s32) into %stack.0, addrspace 5)
   ; PEI-GFX90A-NEXT:   $agpr4 = V_ACCVGPR_WRITE_B32_e64 killed $vgpr1, implicit $exec, implicit killed $vgpr0_vgpr1
   ; PEI-GFX90A-NEXT:   GLOBAL_STORE_DWORDX4 undef renamable $vgpr0_vgpr1, killed renamable $agpr0_agpr1_agpr2_agpr3, 0, 0, implicit $exec :: (volatile store (s128) into `ptr addrspace(1) poison`, addrspace 1)
-  ; PEI-GFX90A-NEXT:   renamable $sgpr0_sgpr1_sgpr2_sgpr3 = S_LOAD_DWORDX4_IMM killed renamable $sgpr4_sgpr5, 0, 0 :: (dereferenceable invariant load (s128) from %ir.arg.kernarg.offset1, addrspace 4)
+  ; PEI-GFX90A-NEXT:   renamable $sgpr0_sgpr1_sgpr2_sgpr3 = S_LOAD_DWORDX4_IMM killed renamable $sgpr4_sgpr5, 0, 0 :: (dereferenceable invariant load (s128) from %ir.partial_copy.kernarg.segment, addrspace 4)
   ; PEI-GFX90A-NEXT:   renamable $agpr0_agpr1_agpr2_agpr3 = COPY killed renamable $sgpr0_sgpr1_sgpr2_sgpr3, implicit $exec
   ; PEI-GFX90A-NEXT:   renamable $vgpr0 = V_MOV_B32_e32 1, implicit $exec
   ; PEI-GFX90A-NEXT:   renamable $vgpr1 = V_MOV_B32_e32 2, implicit $exec
diff ...
[truncated]

@preames preames mentioned this pull request Jul 3, 2025
Open
nikic
nikic reviewed Jul 3, 2025
View reviewed changes
Copy link
Contributor

@nikic nikic left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Aren't no-op bitcasts the way how ConstantHoisting conveys opaque values to the backend?

@preames
Copy link
Collaborator Author

preames commented Jul 3, 2025

Aren't no-op bitcasts the way how ConstantHoisting conveys opaque values to the backend?

I weakly believe that the opaque constant case is covered by the check just above:

if (isa<Constant>(CI->getOperand(0)))
      return AnyChange;

I say weakly here because I'm not familiar with constant hoisting beyond a cursory level, and it might be I'm wrong here. I did a quick search through the code, but if there's something more subtle going on - maybe around the constant expr stuff? - I might have missed it.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Reviewers

@nikic nikic nikic left review comments

@RKSimon RKSimon Awaiting requested review from RKSimon

@e-kud e-kud Awaiting requested review from e-kud

@topperc topperc Awaiting requested review from topperc

Assignees
No one assigned
Labels
backend:AArch64 backend:AMDGPU backend:SystemZ backend:X86 debuginfo llvm:codegen llvm:globalisel llvm:transforms
Projects
None yet
Milestone
No milestone
Development

Successfully merging this pull request may close these issues.
3 participants
@preames @llvmbot @nikic