[PATCH] D140242: [AMDGPU] Modify adjustInliningThreshold to also consider the cost of passing function arguments through the stack

[Matt Arsenault via Phabricator via llvm-commits]

arsenm accepted this revision.
arsenm added a comment.

You don't need to re-post the review for every minor formatting fix

The SGPR argument thing doesn't really matter now, it's not an optimization we perform for arguments



================
Comment at: llvm/lib/Target/AMDGPU/AMDGPUTargetTransformInfo.cpp:1206
+  ArgStackCost += const_cast<GCNTTIImpl *>(TTIImpl)->getMemoryOpCost(
+      Instruction::Load, Type::getInt32Ty(Callee->getContext()), MaybeAlign(4),
+      AMDGPUAS::PRIVATE_ADDRESS, TTI::TCK_SizeAndLatency);
----------------
Align(4), not MaybeAlign


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D140242/new/

https://reviews.llvm.org/D140242