Documentation on the custom syntax for existing operations


when writing MLIR code, I sometimes find it hard to figure out the proper syntax of some operations. E.g. the semantics of the index_cast operation in the standard dialect are documented, but for its syntax there is neither an example or a description. Is there any documentation on the custom syntax for operations?

Kind regards

The documentation on custom syntax would live with the operation documentation, but looks like index_cast is missing an example or syntax explanation in its description. It also uses a manually-implemented custom assembly format that predates the declarative assembly format, which automatically adds the syntax to the generated documentation (e.g. Probably the best way to make sure these are documented would be to migrate as many as possible to the declarative format.

Ok, I understand. Thanks!