[PATCH] D139078: [X86][FP16] Do not combine fminnum/fmaxnum for FP16 emulation

[Phoebe Wang via Phabricator via llvm-commits]

pengfei added inline comments.


================
Comment at: llvm/test/CodeGen/X86/pr59258.ll:170
+
+declare <8 x half> @llvm.maxnum.v16f16(<8 x half>, <8 x half>)
+declare <8 x half> @llvm.minnum.v16f16(<8 x half>, <8 x half>)
----------------
spatel wrote:
> I don't think it changes anything, but shouldn't these calls be `@llvm.*.v8f16`?
Good catch, thanks!


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D139078/new/

https://reviews.llvm.org/D139078