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

[Pierre van Houtryve via Phabricator via llvm-commits]

Pierre-vh added inline comments.


================
Comment at: llvm/lib/Target/AMDGPU/AMDGPUInstructionSelector.cpp:1125
+
+  case CmpInst::FCMP_OEQ:
+    return Select(AMDGPU::V_CMP_EQ_F16_e64, AMDGPU::V_CMP_EQ_F16_t16_e64,
----------------
Note: I used the ops used in VOPCInstructions.td:1071 - not sure if it's correct or if we want different codegen for these intrinsics.


================
Comment at: llvm/lib/Target/AMDGPU/AMDGPUInstructionSelector.cpp:1297
+                     .addReg(Src1)
+                     .addImm(0); // omod
+  } else
----------------
Does omod need to be selected too? How?


Repository:
  rG LLVM Github Monorepo

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

https://reviews.llvm.org/D136592