[PATCH] D134354: [AMDGPU][GlobalISel] Support mad/fma_mix selection

[Matt Arsenault via Phabricator via llvm-commits]

arsenm added inline comments.


================
Comment at: llvm/lib/Target/AMDGPU/AMDGPUInstructionSelector.cpp:580
+
+  I.removeFromParent();
+  return true;
----------------
eraseFromParent


================
Comment at: llvm/lib/Target/AMDGPU/AMDGPUInstructionSelector.cpp:3732
 
-  if (AllowAbs && MI && MI->getOpcode() == AMDGPU::G_FABS) {
+  // TODO: Should be a combine instead
+  if (MI->getOpcode() == AMDGPU::G_FSUB) {
----------------
This should definitely be a combine and should not be in the selector. This should only directly interpret fabs and fneg


================
Comment at: llvm/lib/Target/AMDGPU/AMDGPURegBankCombiner.cpp:248
 
+    return false;
   if (!K0->Value.isExactlyValue(0.0) || !K1->Value.isExactlyValue(1.0))
----------------
Don't see what happened here?


================
Comment at: llvm/lib/Target/AMDGPU/AMDGPURegBankCombiner.cpp:260
     Reg = Val;
+    dbgs() << "  yes:\n";
     return true;
----------------
Leftover debugging


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D134354/new/

https://reviews.llvm.org/D134354