[PATCH] D75392: [AMDGPU] Fix the gfx10 scheduling model for f32 conversions

[Jay Foad via Phabricator via llvm-commits]

foad marked 2 inline comments as done.
foad added inline comments.


================
Comment at: llvm/lib/Target/AMDGPU/VOP1Instructions.td:199
 let FPDPRounding = 1 in {
 defm V_CVT_F16_F32 : VOP1Inst <"v_cvt_f16_f32", VOP_F16_F32, fpround>;
 } // End FPDPRounding = 1
----------------
rampitec wrote:
> As far as I can tell this is quarter rate.
For anyone following at home, we got confirmation from an authority at AMD that V_CVT_F16_F32 is full rate on gfx10.


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D75392/new/

https://reviews.llvm.org/D75392