[PATCH] D90028: [AMDGPU] Add simplification/combines for llvm.amdgcn.fma.legacy

[Matt Arsenault via Phabricator via llvm-commits]

arsenm added inline comments.


================
Comment at: llvm/test/Transforms/InstCombine/AMDGPU/fma_legacy.ll:7
+; CHECK-LABEL: @test_zero(
+; CHECK-NEXT:    [[CALL:%.*]] = fadd float [[Z:%.*]], 0.000000e+00
+; CHECK-NEXT:    ret float [[CALL]]
----------------
Test preserving flags?


================
Comment at: llvm/test/Transforms/InstCombine/AMDGPU/fma_legacy.ll:48
+;
+  %call = call float @llvm.amdgcn.fma.legacy(float %x, float 99.5, float %z)
+  ret float %call
----------------
Should also test that flags are preserved


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D90028/new/

https://reviews.llvm.org/D90028