[PATCH] D46085: AMDGPU/SI: Don't promote alloca to vector for atomic load/store

[Matt Arsenault via Phabricator via llvm-commits]

arsenm accepted this revision.
arsenm added a comment.
This revision is now accepted and ready to land.

LGTM with the check lines enhanced



================
Comment at: test/CodeGen/AMDGPU/vector-alloca-atomic.ll:29
+; OPT: store i32 2
+; OPT: load atomic i32
+define amdgpu_kernel void @vector_alloca_atomic_read(i32 addrspace(1)* %out, i64 %index) {
----------------
These check lines should include the addrspace of the pointers used


================
Comment at: test/CodeGen/AMDGPU/vector-alloca-atomic.ll:50
+; OPT: store atomic i32 1
+; OPT: store atomic i32 2
+; OPT: load i32
----------------
Ditto


https://reviews.llvm.org/D46085