rj-jesus wrote:

> The part that alters -mcpu=grace sounds good. I believe that can be committed 
> separately and isn't dependant on the rest. The part that alters Host.cpp is 
> related to -mcpu=native and could be better included of the other patch, so 
> that -mcpu=native works correctly.

Thanks, that should be done now. Depending on the PR that's merged first, I'll 
update the other to remove the corresponding FIXME.

> (I think how it would best work is to use hpcaps on linux and trust all the 
> feaures we have, but that is a larger change).

Agreed, I think that's how we should be doing it. Maybe some of the FMV code in 
compiler-rt could be used for it?

https://github.com/llvm/llvm-project/pull/166387
_______________________________________________
cfe-commits mailing list
[email protected]
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

Reply via email to