[PATCH] D59536: [AMDGPU] Add convergent attribute to WWM.

[Neil Henning via Phabricator via llvm-commits]

sheredom created this revision.
sheredom added reviewers: arsenm, nhaehnle, cwabbott.
sheredom added a project: AMDGPU.
Herald added subscribers: llvm-commits, t-tye, tpr, dstuttard, yaxunl, wdng, jvesely, kzhuravl.
Herald added a project: LLVM.

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


Repository:
  rL LLVM

https://reviews.llvm.org/D59536

Files:
  include/llvm/IR/IntrinsicsAMDGPU.td


Index: include/llvm/IR/IntrinsicsAMDGPU.td
===================================================================
--- include/llvm/IR/IntrinsicsAMDGPU.td
+++ include/llvm/IR/IntrinsicsAMDGPU.td
@@ -1361,7 +1361,7 @@
 // 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


-------------- next part --------------
A non-text attachment was scrubbed...
Name: D59536.191259.patch
Type: text/x-patch
Size: 561 bytes
Desc: not available
URL: <http://lists.llvm.org/pipermail/llvm-commits/attachments/20190319/876790c5/attachment.bin>