Re: Rename xmm-tn-p to float-register-tn-p

CI fails on Mac build. But all places are only in x86-64.

Is it really related to this one or something else?
________________________________
Från: arthur miller <[email protected]>
Skickat: den 20 augusti 2026 19:58
Till: Stas Boukarev <[email protected]>
Kopia: sbcl-devel <[email protected]>
Ämne: Re: [Sbcl-devel] Rename xmm-tn-p to float-register-tn-p

Ha! I just produced new one and was about to send it 🙂 👍

I also wrote a less verbose commit message in the new one,
Without unnecessary deatils, but it is ok; I'll rebase mine branch.

Thanks.

________________________________
Från: Stas Boukarev <[email protected]>
Skickat: den 20 augusti 2026 19:44
Till: arthur miller <[email protected]>
Kopia: sbcl-devel <[email protected]>
Ämne: Re: Rename xmm-tn-p to float-register-tn-p

I modified your patch to float-tn-p and applied it.

On Thu, Aug 20, 2026 at 8:37 PM Stas Boukarev <[email protected]> wrote:
>
> I think float-tn-p is a better name.
>
> > Another question: do you prefer patches, pull requests on GH, or just to
> > notify you when I have something? 🙂 I see you have applied
> > the GH branch, not the patch I send yesterday.
>
> That was just a diff, not a proper patch.
>
> On Thu, Aug 20, 2026 at 8:34 PM arthur miller <[email protected]> wrote:
> >
> > I would like to refactor an older function for Intel simd support.
> > Is that OK please? The reason:
> >
> > The name xmm-tn-p is a misnomer. The code checks whether a register is in
> > storage base for float registers which can be xmm, ymm, zmm, float, double,
> > complex float and complex double registers. While it is true xmm-tn-p asks if a
> > TN is in float registers SB, it also suggests we are differentiating between
> > ymm-tn-p, zmm-tn-p etc, which is not really the case here. It caused me a bug
> > and some investigation while working on mask registers support.
> >
> > Another question: do you prefer patches, pull requests on GH, or just to
> > notify you when I have something? 🙂 I see you have applied
> > the GH branch, not the patch I send yesterday.

_______________________________________________
Sbcl-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/sbcl-devel