[PATCH] D106891: [AMDGPU] [Remarks] Emit optimization remarks for FP atomics

[Matt Arsenault via Phabricator via llvm-commits]

arsenm added a comment.

I think getting too specific as to why in the messages may confuse people. Especially "might not update memory" is not super helpful. It sounds like the instruction is just entirely unreliable



================
Comment at: llvm/lib/Target/AMDGPU/SIISelLowering.cpp:12212
+            "A floating-point atomic instruction will generate an unsafe"
+            " hardware instruction which may not follow denorm mode");
+      }
----------------
Should not abbreviate denormal


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D106891/new/

https://reviews.llvm.org/D106891