[Mlir-commits] [mlir] [mlir][linalg] Add scalable vectorisation for depthwise convolutions (PR #81625)

[Andrzej Warzyński]

================
@@ -2697,6 +2714,7 @@ struct Conv1DGenerator
         return;
       break;
     }
+    hasTensorSemantics = linalgOp.hasPureTensorSemantics();
----------------
banach-space wrote:

See `maybeMaskXferOp` (that's where this is required). I couldn't find any other way to propagate this info.

https://github.com/llvm/llvm-project/pull/81625