[PATCH] D147216: [AMDGPU] Reduce repetition in SMEM instruction definitions

[Ivan Kosarev via Phabricator via llvm-commits]

kosarev added a comment.

More of that would be great.



================
Comment at: llvm/lib/Target/AMDGPU/SMInstructions.td:347
 let SubtargetPredicate = HasScalarFlatScratchInsts, Uses = [FLAT_SCR] in {
-defm S_SCRATCH_LOAD_DWORD    : SM_Pseudo_Loads <"s_scratch_load_dword",   SReg_64, SReg_32_XM0_XEXEC>;
-defm S_SCRATCH_LOAD_DWORDX2  : SM_Pseudo_Loads <"s_scratch_load_dwordx2", SReg_64, SReg_64_XEXEC>;
-defm S_SCRATCH_LOAD_DWORDX4  : SM_Pseudo_Loads <"s_scratch_load_dwordx4", SReg_64, SReg_128>;
+defm S_SCRATCH_LOAD_DWORD    : SM_Pseudo_Loads <  SReg_64, SReg_32_XM0_XEXEC>;
+defm S_SCRATCH_LOAD_DWORDX2  : SM_Pseudo_Loads <SReg_64, SReg_64_XEXEC>;
----------------
nit: These spaces after `<` here and below might need some more treatment.


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D147216/new/

https://reviews.llvm.org/D147216