AMDGPU: llvm.amdgcn.inverse.ballot needs to be convergent #155725
Merged
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
It is only defined for uniform inputs (instruction selection inserts v_readfirstlane as necessary).
|
@llvm/pr-subscribers-llvm-ir @llvm/pr-subscribers-backend-amdgpu Author: Nicolai Hähnle (nhaehnle) ChangesIt is only defined for uniform inputs (instruction selection inserts v_readfirstlane as necessary). Full diff: https://github.com/llvm/llvm-project/pull/155725.diff 2 Files Affected:
diff --git a/llvm/include/llvm/IR/IntrinsicsAMDGPU.td b/llvm/include/llvm/IR/IntrinsicsAMDGPU.td
index abd83c7c4d4a7..a4d4adae580d0 100644
--- a/llvm/include/llvm/IR/IntrinsicsAMDGPU.td
+++ b/llvm/include/llvm/IR/IntrinsicsAMDGPU.td
@@ -2434,9 +2434,13 @@ def int_amdgcn_ballot :
Intrinsic<[llvm_anyint_ty], [llvm_i1_ty],
[IntrNoMem, IntrConvergent, IntrWillReturn, IntrNoCallback, IntrNoFree]>;
+// Inverse of ballot: return the bit corresponding to the current lane from the
+// given mask.
+//
+// This is only defined for dynamically uniform masks and therefore convergent.
def int_amdgcn_inverse_ballot :
Intrinsic<[llvm_i1_ty], [llvm_anyint_ty],
- [IntrNoMem, IntrWillReturn, IntrNoCallback, IntrNoFree]>;
+ [IntrNoMem, IntrConvergent, IntrWillReturn, IntrNoCallback, IntrNoFree]>;
// Lowers to S_BITREPLICATE_B64_B32.
// The argument must be uniform; otherwise, the result is undefined.
diff --git a/llvm/test/CodeGen/AMDGPU/llvm.amdgcn.inverse.ballot.i32.ll b/llvm/test/CodeGen/AMDGPU/llvm.amdgcn.inverse.ballot.i32.ll
index 835c92448a4c0..05b786bdb5c76 100644
--- a/llvm/test/CodeGen/AMDGPU/llvm.amdgcn.inverse.ballot.i32.ll
+++ b/llvm/test/CodeGen/AMDGPU/llvm.amdgcn.inverse.ballot.i32.ll
@@ -5,7 +5,7 @@
; RUN: not llc -mtriple=amdgcn -mcpu=gfx1100 -mattr=+wavefrontsize64 -global-isel=1 < %s 2>&1 | FileCheck -check-prefix=GISEL-ERR %s
; RUN: not --crash llc -mtriple=amdgcn -mcpu=gfx1100 -mattr=+wavefrontsize64 -global-isel=0 < %s 2>&1 | FileCheck -check-prefix=SDAG-ERR %s
-; GISEL-ERR: LLVM ERROR: cannot select: {{.*}} = G_INTRINSIC intrinsic(@llvm.amdgcn.inverse.ballot)
+; GISEL-ERR: LLVM ERROR: cannot select: {{.*}} = G_INTRINSIC_CONVERGENT intrinsic(@llvm.amdgcn.inverse.ballot)
; SDAG-ERR: LLVM ERROR: Cannot select: intrinsic %llvm.amdgcn.inverse.ballot
declare i1 @llvm.amdgcn.inverse.ballot(i32)
|
arsenm
approved these changes
Aug 28, 2025
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
It is only defined for uniform inputs (instruction selection inserts v_readfirstlane as necessary).