[PATCH] D130579: AMDGPU: Remove manual selection for atomic fadd

[Jay Foad via Phabricator via llvm-commits]

foad added a reviewer: AMDGPU.
foad added a comment.

Overall I think this looks good. Can you precommit all the changes in test/ (consider it pre approved) and then rebase this patch?



================
Comment at: llvm/lib/Target/AMDGPU/AMDGPULegalizerInfo.cpp:1344
     Atomic.legalFor({{S32, GlobalPtr}});
+  if (ST.hasGFX940Insts())
+    Atomic.legalFor({{S32, FlatPtr}, {S64, FlatPtr}, {V2S16, FlatPtr}});
----------------
Apparently gfx90a has flat_atomic_add for f64 (but not f32 or v2f16). Are there any tests for that?


CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D130579/new/

https://reviews.llvm.org/D130579