[PATCH] D136592: [AMDGPU][GISel] Select llvm.amdgcn.fcmp intrinsics

[Matt Arsenault via Phabricator via llvm-commits]

arsenm added inline comments.


================
Comment at: llvm/lib/Target/AMDGPU/AMDGPUInstructionSelector.cpp:1289
+        STI.isWave32() ? AMDGPU::S_MOV_B32 : AMDGPU::S_MOV_B64;
+    BuildMI(*BB, &I, DL, TII.get(Opcode), Dst).addImm(Value);
+    I.eraseFromParent();
----------------
This has different behavior with exec. This doesn't and exec. For some reason, we do have V_CMP_T_* and V_CMP_F_* you can directly select to, without having to write special case code for these


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D136592/new/

https://reviews.llvm.org/D136592