[PATCH] D84995: [AMDGPU][CostModel] Add f16, f64 and contract cases to fused costs estimation.

[Stanislav Mekhanoshin via Phabricator via llvm-commits]

rampitec added inline comments.


================
Comment at: llvm/test/Analysis/CostModel/AMDGPU/fused_costs.ll:6
+; RUN: opt -cost-model -cost-kind=code-size -analyze -mtriple=amdgcn-unknown-amdhsa -mcpu=gfx900 -denormal-fp-math-f32=ieee -denormal-fp-math=ieee -fp-contract=on < %s | FileCheck -check-prefixes=SLOW,NOCONTRACT,ALL %s
+; RUN: opt -cost-model -cost-kind=code-size -analyze -mtriple=amdgcn-unknown-amdhsa -mcpu=gfx900 -denormal-fp-math-f32=ieee -denormal-fp-math=ieee -fp-contract=fast < %s | FileCheck -check-prefixes=FUSED32,FUSED16,CONTRACT,ALL %s
+
----------------
Add a run line with -mcpu=gfx1030 and disabled fp32 denorm support. It does not have f32 mad/mac and shall not be contracted as a result.


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D84995/new/

https://reviews.llvm.org/D84995