[PATCH] D115881: [AMDGPU] Select no-return ds_* atomic ops in tblgen.

[Matt Arsenault via Phabricator via llvm-commits]

arsenm added inline comments.


================
Comment at: llvm/lib/Target/AMDGPU/DSInstructions.td:894-895
 
+// FIXME: (How) Should we write a GCNPat for DSAtomicNoRetPat? Currently noret
+// atomic patterns are derived from DSAtomicRetPat.
+
----------------
Is this fixme relevant since you handle this case


================
Comment at: llvm/lib/Target/AMDGPU/SIInstrInfo.td:294-297
+defm atomic_inc : binary_atomic_op_all_as<SIatomic_inc>;
+defm atomic_dec : binary_atomic_op_all_as<SIatomic_dec>;
+defm atomic_load_fmin : binary_atomic_op_all_as<SIatomic_fmin, 0>;
+defm atomic_load_fmax : binary_atomic_op_all_as<SIatomic_fmax, 0>;
----------------
Is this a separate cleanup?


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D115881/new/

https://reviews.llvm.org/D115881