[PATCH] D113779: [Clang] Add mfp16, mfp16fml and mdotprod flags for ARM target features.

Florian Hahn via Phabricator via cfe-commits cfe-commits at lists.llvm.org
Thu Dec 16 02:11:04 PST 2021


fhahn added a comment.

In D113779#3192009 <https://reviews.llvm.org/D113779#3192009>, @SjoerdMeijer wrote:

> Ok, fair enough, perhaps adding features is a valid use-case.
>
> I will refrain from commenting on "things are terribly broken".  I agree it is broken, but in a different way than suggested in previous comments. 
> If others also think this makes sense, then here a few follow up remarks from my side:
>
> - First of all, this (really) sets precedent for setting options in a different way. This needs documentation and release noting.
> - If we are going to do this, this should be the first patch in a series to fix this for all features. We can't just do a few of them.

Agreed.

> - There was a suggestion to allow adding features with -march=+feature. Was this dismissed in favour of how things works for x86 and be consistent with that?

This would be the easiest way to implement this, *but* it would require either to allow not specify an architecture version with `-march` (which seems a bit odd) or perhaps adding a `default` architecture version which just uses the default set. For our users, either would work, so I'd be happy to go with what seems most compelling to others.

But compatibility with X86 is IMO also valuable for people porting from X86->AArch64.

> It would be really good if we keep options consistent/compatible across the GCC and Clang toolchains. Any possibility to check with GCC community if they are open for this?

If anybody has contacts to GCC that would be very helpful. Unfortunately I don't think I will be able to drive this.


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D113779/new/

https://reviews.llvm.org/D113779



More information about the cfe-commits mailing list