[PATCH] D102334: [AMDGPU] Sort LDS globals based on thier size and alignment.
Matt Arsenault via Phabricator via llvm-commits
llvm-commits at lists.llvm.org
Wed May 12 10:31:18 PDT 2021
arsenm added a comment.
Why do we need another pass if we always just produce llvm.amdgcn.module.lds such that codegen will always allocate one single global?
================
Comment at: llvm/lib/Target/AMDGPU/AMDGPUTargetMachine.h:41
+ // soon as possible.
+ static std::vector<GlobalVariable *> SortedLDSGlobals;
----------------
rampitec wrote:
> I do not think you can keep per-module stuff in the TM.
You absolutely cannot do this here
Repository:
rG LLVM Github Monorepo
CHANGES SINCE LAST ACTION
https://reviews.llvm.org/D102334/new/
https://reviews.llvm.org/D102334
More information about the llvm-commits
mailing list