[Mlir-commits] [mlir] [mlir][vector] Refactor vector-transfer-flatten.mlir (nfc) (3/n) (PR #95745)

[Andrzej Warzyński]

================
@@ -131,10 +131,42 @@ func.func @transfer_read_dims_mismatch_non_contiguous_non_zero_indices(
 
 // -----
 
-// The input memref has a dynamic trailing shape and hence is not flattened.
-// TODO: This case could be supported via memref.dim
+/// The leading dynamic shapes don't affect whether this example is flattenable
+/// or not. Indeed, those dynamic shapes are not candidates for flattening anyway.
----------------
banach-space wrote:

Sorry about that, it wasn't intentional.

For context, some people use different format for comments in tests (e.g. `///` instead of `//`) as apparently their editors handle that better (i.e. use different highlighting for comment and `RUN`/`CHECK` lines).

In any case, we should always prioritise consistency - I've just updated these from `///` to `//`.

https://github.com/llvm/llvm-project/pull/95745