[PATCH] D92394: [amdgpu] Teach one more case for assumed global pointers.

[Michael Liao via Phabricator via llvm-commits]

hliao added inline comments.


================
Comment at: llvm/lib/Target/AMDGPU/AMDGPUTargetMachine.cpp:553
+  const Argument *Arg = dyn_cast_or_null<Argument>(getUnderlyingObject(Ptr));
+  if (Arg && Arg->onlyReadsMemory() && Arg->hasNoAliasAttr() &&
+      Arg->getParent()->getCallingConv() == CallingConv::AMDGPU_KERNEL)
----------------
arsenm wrote:
> Readonly is incorrect here, as it only indicates the function does not write to the memory and does not indicate the memory may be changed by another function/thread. I also don't think noalias is really the correct check for the owned object
That `readonly` atttribute is marked in function attribute deduction pass by checking every use of a pointer argument, including if it's passed into another sub-func. It's only marked as `readonly` when it's safe. It's safe to assume there's no write though a pointer argument marked with `readonly`.
`noalias` here is to ensure that there's no other path within the kernel function execution to modify that memory pointed by that pointer argument.
By checking both, we could ensure that memory is only prepared by the host.


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D92394/new/

https://reviews.llvm.org/D92394