[AMDGPU] Add convergent attribute to WWM.

Add the convergent attribute to the WWM intrinsic to stop it ever being
sunk out of cfg.

Differential Revision: https://reviews.llvm.org/D59536

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@356470 91177308-0d34-0410-b5e6-96231b3b80d8

diff --git a/include/llvm/IR/IntrinsicsAMDGPU.td b/include/llvm/IR/IntrinsicsAMDGPU.td
index 6fc17277bede3a5bf0549bb752daa84c62c92620..6c493571d8591ff9f7c950f3e90a282ca25a4843 100644 (file)
--- a/include/llvm/IR/IntrinsicsAMDGPU.td
+++ b/include/llvm/IR/IntrinsicsAMDGPU.td
@@ -1361,7 +1361,7 @@ def int_amdgcn_kill : Intrinsic<[], [llvm_i1_ty], []>;
 // enabled, with a few exceptions: - Phi nodes with require WWM return an
 // undefined value.
 def int_amdgcn_wwm : Intrinsic<[llvm_any_ty],
-  [LLVMMatchType<0>], [IntrNoMem, IntrSpeculatable]
+  [LLVMMatchType<0>], [IntrNoMem, IntrSpeculatable, IntrConvergent]
 >;
 
 // Given a value, copies it while setting all the inactive lanes to a given