[Mlir-commits] [mlir] [sparse] allow unpack op to return 0-ranked tensor type. (PR #66269)

[Aart Bik]

================
@@ -214,6 +214,8 @@ struct SparseTensorCodegenPass
     target.addLegalOp<GetStorageSpecifierOp>();
     target.addLegalOp<SetStorageSpecifierOp>();
     target.addLegalOp<StorageSpecifierInitOp>();
+    // tensor::FromElementsOp might be yield after lowering unpack.
----------------
aartbik wrote:

At first glance, this looks like commented out code ;-)
So please change to

// Note that `tensor::FromElementsOp` might be yield after lowering unpack.

https://github.com/llvm/llvm-project/pull/66269