[PATCH] D119696: [AMDGPU] Improve v_cmpx usage on GFX10.3.

Thomas Symalla via Phabricator via llvm-commits llvm-commits at lists.llvm.org
Mon Mar 21 01:29:20 PDT 2022


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


================
Comment at: llvm/lib/Target/AMDGPU/SIOptimizeExecMasking.cpp:653
+  // to reduce pipeline stalls.
+  if (ST.hasGFX10_3Insts()) {
+    DenseMap<MachineInstr *, MachineInstr *> SaveExecVCmpMapping;
----------------
critson wrote:
> Could this be more future proof?
> e.g.
> `if (AMDGPU::isGFX10Plus(ST) && !ST.hasVcmpxExecWARHazard())`
I am not sure if the conditions check would be correct, as I don't know the logic behind setting the value returned by `hasVcmpxExecWARHazard()`. We don't want to have this transformation enabled on either GFX10.1 or GFX10.2 - so for now I'm leaving it as it is - we can change that later perhaps.


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D119696/new/

https://reviews.llvm.org/D119696



More information about the llvm-commits mailing list