[PATCH] D102334: [AMDGPU] Sort LDS globals based on thier size and alignment.

[Stanislav Mekhanoshin via Phabricator via llvm-commits]

rampitec added a comment.

As far as I understand all kernels in the module will use all the LDS globals irrespective if they actually need it? We may run out of LDS.



================
Comment at: llvm/lib/Target/AMDGPU/AMDGPUMachineFunction.cpp:51
+  // available.
+  auto Iter = std::find(AMDGPUTargetMachine::SortedLDSGlobals.begin(),
+                        AMDGPUTargetMachine::SortedLDSGlobals.end(), &GV);
----------------
llvm::find() and you do not need to include algorithm then.


================
Comment at: llvm/lib/Target/AMDGPU/AMDGPUTargetMachine.h:41
+  // soon as possible.
+  static std::vector<GlobalVariable *> SortedLDSGlobals;
 
----------------
I do not think you can keep per-module stuff in the TM.


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D102334/new/

https://reviews.llvm.org/D102334