[flang-commits] [flang] [mlir] [mlir][gpu] Add stream-based async mode to gpu.launch (PR #213031)

[Fabian Mora via flang-commits]

================
@@ -1005,6 +1025,17 @@ ParseResult LaunchOp::parse(OpAsmParser &parser, OperationState &result) {
   SmallVector<OpAsmParser::UnresolvedOperand, 16> regionArgs(
       LaunchOp::kNumConfigRegionAttributes);
 
+  // Parse optional asyncObject: < value : type >
+  OpAsmParser::UnresolvedOperand asyncObjectOperand;
+  Type asyncObjectType;
+  bool hasAsyncObject = false;
+  if (succeeded(parser.parseOptionalLess())) {
+    hasAsyncObject = true;
+    if (parser.parseOperand(asyncObjectOperand) || parser.parseColon() ||
+        parser.parseType(asyncObjectType) || parser.parseGreater())
+      return failure();
+  }
----------------
fabianmcg wrote:

Is this the same format `gpu.launch_func` uses? If it is, can we factor out common code?

https://github.com/llvm/llvm-project/pull/213031