[PATCH] D65325: [AMDGPU] Enable v4f16 and above for v_pk_fma instructions

[Matt Arsenault via Phabricator via llvm-commits]

arsenm added inline comments.


================
Comment at: test/CodeGen/AMDGPU/llvm.fma.f16.ll:357
+  %mul.val = fmul reassoc nnan nsz arcp contract <4 x half> %a.val, %b.val
+  %r.val = fadd reassoc nnan nsz arcp contract <4 x half> %mul.val, %c.val
+  store <4 x half> %r.val, <4 x half> addrspace(1)* %r
----------------
Should test the intrinsic rather than the contraction


Repository:
  rL LLVM

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D65325/new/

https://reviews.llvm.org/D65325