[PATCH] D99743: [AMDGPU] Use enum for flat variants. NFC

[Jay Foad via Phabricator via llvm-commits]

foad added a comment.

In D99743#2674198 <https://reviews.llvm.org/D99743#2674198>, @sebastian-ne wrote:

> Use SIInstrFlags instead of new enum.

Thanks for trying it out! I think it is good to avoid introducing another enum for the same thing. I'm not thrilled about having to pass around a "uint64_t" argument for it, or about the name "SIInstrFlags::IsFlatScratch" which sounds more like a predicate function (but that's not your fault). What do others think?



================
Comment at: llvm/lib/Target/AMDGPU/AMDGPUInstructionSelector.h:16
 
+#include "SIDefines.h"
 #include "llvm/CodeGen/GlobalISel/InstructionSelector.h"
----------------
I don't think you need this.


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D99743/new/

https://reviews.llvm.org/D99743