Skip to content

[IR] Return correct memory effects for convergencectrl #129874

New issue

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

Sign up for GitHub

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

Already on GitHub? Sign in to your account

Merged
merged 4 commits into from
Mar 5, 2025
Merged

[IR] Return correct memory effects for convergencectrl #129874

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
4 commits
Select commit Hold shift + click to select a range
d28584f
[DCE] Add pre-commit tests. NFC.
dtcxzyw Mar 5, 2025
f30058b
[IR] Return correct memory effects for convergencectrl
dtcxzyw Mar 5, 2025
fce8210
[IRNormalizer] Fix test failure.
dtcxzyw Mar 5, 2025
74d812e
[InstCombine] Fix tests. NFC.
dtcxzyw Mar 5, 2025
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
8 changes: 5 additions & 3 deletions llvm/lib/IR/Instructions.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -609,15 +609,17 @@ bool CallBase::hasReadingOperandBundles() const {
// Implementation note: this is a conservative implementation of operand
// bundle semantics, where *any* non-assume operand bundle (other than
// ptrauth) forces a callsite to be at least readonly.
return hasOperandBundlesOtherThan(
{LLVMContext::OB_ptrauth, LLVMContext::OB_kcfi}) &&
return hasOperandBundlesOtherThan({LLVMContext::OB_ptrauth,
LLVMContext::OB_kcfi,
LLVMContext::OB_convergencectrl}) &&
getIntrinsicID() != Intrinsic::assume;
}

bool CallBase::hasClobberingOperandBundles() const {
return hasOperandBundlesOtherThan(
{LLVMContext::OB_deopt, LLVMContext::OB_funclet,
LLVMContext::OB_ptrauth, LLVMContext::OB_kcfi}) &&
LLVMContext::OB_ptrauth, LLVMContext::OB_kcfi,
LLVMContext::OB_convergencectrl}) &&
getIntrinsicID() != Intrinsic::assume;
}

Expand Down
12 changes: 12 additions & 0 deletions llvm/test/Transforms/DCE/op_bundles.ll
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -0,0 +1,12 @@
; NOTE: Assertions have been autogenerated by utils/update_test_checks.py UTC_ARGS: --version 5
; RUN: opt -S < %s -passes=dce | FileCheck %s

define void @dead_readfirstlane_convergencetoken(<2 x i32> %src) convergent {
; CHECK-LABEL: define void @dead_readfirstlane_convergencetoken(
; CHECK-SAME: <2 x i32> [[SRC:%.*]]) #[[ATTR0:[0-9]+]] {
; CHECK-NEXT: ret void
;
%t = tail call token @llvm.experimental.convergence.entry()
%vec = call <2 x i32> @llvm.amdgcn.readfirstlane.v2i32(<2 x i32> %src) [ "convergencectrl"(token %t) ]
ret void
}
8 changes: 4 additions & 4 deletions llvm/test/Transforms/IRNormalizer/regression-convergence-tokens.ll
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -7,14 +7,14 @@ define i32 @nested(i32 %src) #0 {
; CHECK-SAME: i32 [[A0:%.*]]) #[[ATTR0:[0-9]+]] {
; CHECK-NEXT: [[BB15160:.*:]]
; CHECK-NEXT: [[T1:%.*]] = call token @llvm.experimental.convergence.entry()
; CHECK-NEXT: %"vl15001llvm.experimental.convergence.anchor()" = call token @llvm.experimental.convergence.anchor()
; CHECK-NEXT: %"op68297llvm.amdgcn.readfirstlane.i32([[A0]], vl15001llvm.experimental.convergence.anchor())" = call i32 @llvm.amdgcn.readfirstlane.i32(i32 [[A0]]) [ "convergencectrl"(token %"vl15001llvm.experimental.convergence.anchor()") ]
; CHECK-NEXT: ret i32 undef
; CHECK-NEXT: %"vl77672llvm.experimental.convergence.anchor()" = call token @llvm.experimental.convergence.anchor()
; CHECK-NEXT: %"op68297(vl77672)" = call i32 @llvm.amdgcn.readfirstlane.i32(i32 [[A0]]) [ "convergencectrl"(token %"vl77672llvm.experimental.convergence.anchor()") ]
; CHECK-NEXT: ret i32 %"op68297(vl77672)"
;
%t1 = call token @llvm.experimental.convergence.entry()
%t2 = call token @llvm.experimental.convergence.anchor()
%r2 = call i32 @llvm.amdgcn.readfirstlane(i32 %src) [ "convergencectrl"(token %t2) ]
ret i32 undef
ret i32 %r2
Copy link
Member Author

Choose a reason for hiding this comment

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

IRNormalizer does not set name for dead values.

}

; Function Attrs: convergent nounwind readnone
Expand Down
1 change: 0 additions & 1 deletion llvm/test/Transforms/InstCombine/AMDGPU/simplify-demanded-vector-elts-lane-intrinsics.ll
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -349,7 +349,6 @@ define i32 @extract_elt0_v2i32_readfirstlane_convergencetoken(<2 x i32> %src) co
; CHECK-LABEL: define i32 @extract_elt0_v2i32_readfirstlane_convergencetoken(
; CHECK-SAME: <2 x i32> [[SRC:%.*]]) #[[ATTR1:[0-9]+]] {
; CHECK-NEXT: [[T:%.*]] = call token @llvm.experimental.convergence.entry()
; CHECK-NEXT: [[VEC:%.*]] = call <2 x i32> @llvm.amdgcn.readfirstlane.v2i32(<2 x i32> [[SRC]]) [ "convergencectrl"(token [[T]]) ]
; CHECK-NEXT: [[TMP1:%.*]] = extractelement <2 x i32> [[SRC]], i64 0
; CHECK-NEXT: [[ELT:%.*]] = call i32 @llvm.amdgcn.readfirstlane.i32(i32 [[TMP1]]) [ "convergencectrl"(token [[T]]) ]
; CHECK-NEXT: ret i32 [[ELT]]
Expand Down