[PATCH] D121157: [AMDGPU] always use underlying object in the pointsToConstantMemory

[Stanislav Mekhanoshin via Phabricator via llvm-commits]

rampitec added a comment.

And even if it was not vectorized we cannot get the argument after the argument lowering anymore anyway:

  %arg.const.load = load i32 addrspace(4)*, i32 addrspace(4)* addrspace(4)* %arg.const.kernarg.offset.cast, align 4, !invariant.load !0

Apparently pointsToConstantMemory returns false for this.


CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D121157/new/

https://reviews.llvm.org/D121157