[PATCH] D153544: [AMDGPU] Use V_FMA_MIX* more often

[Matt Arsenault via Phabricator via llvm-commits]

arsenm added inline comments.


================
Comment at: llvm/test/CodeGen/AMDGPU/mad-mix-lo.ll:2246
+.entry:
+  %mul = fmul reassoc nnan nsz arcp contract afn float %a, %b
+  %trunc = fptrunc float %mul to half
----------------
Drop unnecessary flags


================
Comment at: llvm/test/CodeGen/AMDGPU/mad-mix-lo.ll:2293
+.entry:
+  %mul = fmul reassoc nnan nsz arcp contract afn float %a, %b
+  %trunc = fptrunc float %mul to half
----------------
Drop unnecessary flags


================
Comment at: llvm/test/CodeGen/AMDGPU/mad-mix-lo.ll:2297
+}
+
 declare half @llvm.minnum.f16(half, half) #1
----------------
Add tests with source modifier usage


CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D153544/new/

https://reviews.llvm.org/D153544