[PATCH] D75016: [X86][TwoAddressInstructionPass] Teach tryInstructionCommute to continue checking for commutable FMA operands in more cases.

[Simon Pilgrim via Phabricator via llvm-commits]

RKSimon added a comment.

There seems to be a few extra register moves with this change



================
Comment at: llvm/test/CodeGen/X86/recip-fastmath.ll:819
+; HASWELL-NEXT:    vfnmadd132ps {{.*#+}} ymm1 = -(ymm1 * ymm2) + ymm2
+; HASWELL-NEXT:    vmovaps %ymm1, %ymm0
 ; HASWELL-NEXT:    retq
----------------
Any chance that we can avoid this regression (and the other ones above)?


CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D75016/new/

https://reviews.llvm.org/D75016