[PATCH] D72392: [MachineScheduler] Fix the TopDepth/BotHeightReduce latency heuristics

[Jay Foad via Phabricator via llvm-commits]

foad added a comment.

In D72392#1811691 <https://reviews.llvm.org/D72392#1811691>, @fhahn wrote:

> Also, do you have any performance/codesize numbers for impacted targets?


I did some performance testing of graphics shaders on the AMDGPU target. Out of ~350 shaders the average performance changed by less than 0.01%, only 5 shaders changed by more than 1%, and the biggest swings were -4%/+3%. So I think there is no significant overall trend.

I would welcome any help with performance testing on AArch64/PowerPC/X86.


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D72392/new/

https://reviews.llvm.org/D72392