[Bf-committers] cycles fails to build with BMI enabled

[Marco Ziebell]

Anything new on this?

I glimpsed at the commits a while ago but haven't found anything
related. But it is easily missed with so many commits
...


Am Mon, 23 Feb 2015 00:08:39 +0600
schrieb Sergey Sharybin <sergey.vfx at gmail.com>:

> Hi,
> 
> The patch suggested in the gentoo tracker is not totally perfect to be
> applied as-is since it'll likely break Cycles on platform which
> doesn't have intrinsics which the builder machine had. But it for
> sure gives clues what's going on.
> 
> I'll look into the issue and do a proper solution from our side in
> the next couple of days, so stay tuned :)
> 
> On Sun, Feb 22, 2015 at 2:26 PM, Marco Ziebell
> <ziebell_marco at posteo.de> wrote:
> 
> >
> > Could someone please have a look at this report [1], it was
> > abandoned in favour of an mail to this list.
> >
> > There also exists a report in gentoo-bugzilla with further comments
> > an a patch?!?
> >
> > Compiling with "-march=" set properly ("native") is not that
> > mysterious. The only problematic thing about "march=native" in this
> > case is it enables some instruction sets supported by the processor,
> > which causes the build system to fail!
> >
> >
> > Thanks in advanced and best regards
> >
> >
> > Marco
> >
> > [1] https://developer.blender.org/T42279
> > [2] https://bugs.gentoo.org/show_bug.cgi?id=523928
> > _______________________________________________
> > Bf-committers mailing list
> > Bf-committers at blender.org
> > http://lists.blender.org/mailman/listinfo/bf-committers
> >
> 
> 
>