[vlc-devel] [PATCH 05/25] cpu: force feature support runtime check

[Victorien Le Couviour]

On Tue, Apr 14, 2020 at 5:42 PM Rémi Denis-Courmont <remi at remlab.net> wrote:

> Le tiistaina 14. huhtikuuta 2020, 18.14.52 EEST Thomas Guillem a écrit :
> > On Tue, Apr 14, 2020, at 16:57, Alexandre Janniaux wrote:
> > > Hi,
> > >
> > > On Tue, Apr 14, 2020 at 05:45:18PM +0300, Rémi Denis-Courmont wrote:
> > > > Le tiistaina 14. huhtikuuta 2020, 17.35.42 EEST Alexandre Janniaux a
> écrit :
> > > > > Hi,
> > > > >
> > > > > Isn't vlc_CPU checked only in initialization steps to choose
> > > > > the correct variant?
> > > >
> > > > No.
> > >
> > > Can't we just remove the few case where it's not calling it
> > > during initialization? At the same time, it would remove
> > > uneeded atomic operation on hot paths.
> >
> > Indeed, we could fix the few case where vlc_CPU is checked often.
> >
> > In that case, I'm OK with this patch, that is simplifying it a lot.
>
> And I'm not OK with this complete nonsense patch.
>
> If this file is too complicated for you, nobody's asking you to touch it.
> Just
> don't remove other people's optimisations.
>

Can you explain how this removes any optimization?
It just force check for the feature to have been detected on the first call
to vlc_CPU.


>
> --
> Rémi Denis-Courmont
> Hagalund ny stad, Nylands republik
>
>
>
> _______________________________________________
> vlc-devel mailing list
> To unsubscribe or modify your subscription options:
> https://mailman.videolan.org/listinfo/vlc-devel
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.videolan.org/pipermail/vlc-devel/attachments/20200414/8963e3ef/attachment.html>