[PATCH] D33139: AMDGPU/SI: Move the local memory usage related checking after calling convention checking in PromoteAlloca

[Matt Arsenault via Phabricator via llvm-commits]

arsenm added inline comments.


================
Comment at: lib/Target/AMDGPU/AMDGPUPromoteAlloca.cpp:691
     DEBUG(dbgs() << " alloca is not a candidate for vectorization.\n");
-    return;
+    return true;
   }
----------------
Why is this true? It failed


================
Comment at: test/CodeGen/AMDGPU/vector-alloca.ll:149-153
+; EG: MOV
+; EG: MOV
+; EG: MOV
+; EG: MOV
+; EG: MOVA_INT
----------------
cfang wrote:
> arsenm wrote:
> > Should use the GCN instruction checks
> what is GCN instruction check? Do you mean should use llc to compile to ISA and do checking?
You had EG checks before


https://reviews.llvm.org/D33139