[Mlir-commits] [mlir] [mlir] Added `Convergent` trait that matches LLVM's semantics (PR #152358)

[Nikolay Panchenko]

================
@@ -1054,7 +1058,8 @@ def NVVM_CpAsyncWaitGroupOp : NVVM_Op<"cp.async.wait.group">,
   let assemblyFormat = "$n attr-dict";
 }
 
-def NVVM_CpAsyncMBarrierArriveOp : NVVM_Op<"cp.async.mbarrier.arrive"> {
+def NVVM_CpAsyncMBarrierArriveOp : NVVM_Op<"cp.async.mbarrier.arrive",
+  [Convergent]> {
----------------
npanchen wrote:

@AlexMaclean @Prince781 that will be great to revisit nvvm's intrinsic properties. I know PTX doc definitely describes `nvvm_barrier_cta_sync_aligned_all` as convergent, but I couldn't find such statement for others

https://github.com/llvm/llvm-project/pull/152358