RFC: New mechanism for hard register operands to inline asm

Hi All,

Thank you very much for your feedback and opinions on this proposal. Based on the comments, this is what I'm going to do next.

1. Cross post this RFC to the Clang Developer's mailing list to get some more feedback.
2. Come up with draft implementations of how the implementation of the constraint would look like and hash out any open questions.

As others have mentioned in this thread, it would be okay to implement this if both GCC and Clang implement it (to maintain consistency). GCC hasn't implemented it yet (the RFC has been posted already to signal the intent of implementing it). So parallelly, I believe it would be good to get a few discussions going on possible implementation details and/or approaches to implement it on the Clang side.

Best Regards,
Anirudh Prasad

-----Anirudh Prasad/Canada/IBM wrote: -----