[llvm-branch-commits] [llvm] [AMDGPU] Make amdgpu.noclobber respect acquire operations (PR #219991)
Jay Foad via llvm-branch-commits
llvm-branch-commits at lists.llvm.org
Tue Nov 3 06:46:43 PST 2026
jayfoad wrote:
Some review comments from Claude:
---
`fence syncscope("singlethread") acquire` can't synchronise with another thread, so it could be ignored.
Acquires marked `!mmra !{!"amdgcn-av", !"none"}` are not make-visible operations, so they could also be ignored.
A follow-up could ignore workgroup-or-narrower acquires when the kernel has no possibly-aliasing writes, no unknown calls, and no acquire wider than workgroup anywhere in the function.
- Why that is sound: any hb chain into a workgroup-scope acquire goes through threads of the same workgroup, which run the same function.
Test coverage:
- Add a fence case to `promote-kernel-arguments.ll`. The PR changes that pass's behaviour and no test shows it.
- Add a call to an external function, or a helper containing an acquire, to `noclobber-barrier.ll`. It works, but nothing pins it.
- The `noclobber-barrier.ll` header comment ("barrier or fence in between of loads is not considered a clobber") is now wrong for acquire fences.
`isReallyAClobber`'s comment says "Ignore non-acquire atomics not aliasing ... just like a fence". Fences are now *not* always ignored. Reword it, e.g. "Non-acquire atomics that don't alias are not clobbers. Acquire operations always are,
https://github.com/llvm/llvm-project/pull/219991
More information about the llvm-branch-commits
mailing list