rampitec wrote:

> > I like this. I agree we probably do not need an option, this is just the 
> > extra code.
> 
> aka we don't want to test whether we fill the feature map correctly? It is a 
> separate map from the one defined in `AMDGPU.td`.

Probably not. I have to update these tests more often than I actually catch 
anything useful form them.

https://github.com/llvm/llvm-project/pull/176533
_______________________________________________
cfe-commits mailing list
[email protected]
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

Reply via email to