[PATCH] D134230: AMDGPU: Fix expansion of 16-bit atomicrmw

[Matt Arsenault via Phabricator via llvm-commits]

arsenm added inline comments.


================
Comment at: llvm/test/Transforms/AtomicExpand/AMDGPU/expand-atomic-i16.ll:63
 ;
   %res = atomicrmw xchg i16 addrspace(1)* %ptr, i16 %value seq_cst, align 4
   ret i16 %res
----------------
rampitec wrote:
> Thanks! There seems to be a room for improvement in the expansion code for the aligned case, but it does not affect this patch.
I also realized this should be using llvm.ptrmask instead of this ptrtoint hacking


CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D134230/new/

https://reviews.llvm.org/D134230