[Mlir-commits] [llvm] [mlir] [XeGPU][Transform] Add XeGPU array length optimization pass (PR #194062)

[Charitha Saumya]

================
@@ -556,11 +556,20 @@ LogicalResult LoadNdOp::verify() {
     }
   }
 
+  // Handle array_length. Two result shape conventions are accepted:
+  //   * Legacy: leading array_length dimension prepended, e.g. descriptor
+  //     16x16 with array_length=2 -> [2, 16, 16].
----------------
charithaintc wrote:

why is this still accepted?

https://github.com/llvm/llvm-project/pull/194062