[Clang] Pass correct lane mask for match helpers #138693
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
Summary: We use the ballot to get the proper lane mask after we've masked off the threads already done. This isn't an issue on AMDGPU but could cause problems for post-Volta since it's saying that threads are active when they aren't.
|
@llvm/pr-subscribers-backend-x86 @llvm/pr-subscribers-clang Author: Joseph Huber (jhuber6) ChangesSummary: Full diff: https://github.com/llvm/llvm-project/pull/138693.diff 1 Files Affected:
diff --git a/clang/lib/Headers/gpuintrin.h b/clang/lib/Headers/gpuintrin.h
index d308cc959be84..7afc82413996b 100644
--- a/clang/lib/Headers/gpuintrin.h
+++ b/clang/lib/Headers/gpuintrin.h
@@ -264,9 +264,10 @@ __gpu_match_any_u32_impl(uint64_t __lane_mask, uint32_t __x) {
uint64_t __match_mask = 0;
bool __done = 0;
- while (__gpu_ballot(__lane_mask, !__done)) {
+ for (uint64_t __active_mask = __lane_mask; __active_mask;
+ __active_mask = __gpu_ballot(__lane_mask, !__done)) {
if (!__done) {
- uint32_t __first = __gpu_read_first_lane_u32(__lane_mask, __x);
+ uint32_t __first = __gpu_read_first_lane_u32(__active_mask, __x);
if (__first == __x) {
__match_mask = __gpu_lane_mask();
__done = 1;
@@ -283,9 +284,10 @@ __gpu_match_any_u64_impl(uint64_t __lane_mask, uint64_t __x) {
uint64_t __match_mask = 0;
bool __done = 0;
- while (__gpu_ballot(__lane_mask, !__done)) {
+ for (uint64_t __active_mask = __lane_mask; __active_mask;
+ __active_mask = __gpu_ballot(__lane_mask, !__done)) {
if (!__done) {
- uint64_t __first = __gpu_read_first_lane_u64(__lane_mask, __x);
+ uint64_t __first = __gpu_read_first_lane_u64(__active_mask, __x);
if (__first == __x) {
__match_mask = __gpu_lane_mask();
__done = 1;
|
Artem-B
approved these changes
May 7, 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.
Summary:
We use the ballot to get the proper lane mask after we've masked off the
threads already done. This isn't an issue on AMDGPU but could cause
problems for post-Volta since it's saying that threads are active when
they aren't.