Commen=
t # 24
on bug 83436<=
/a>
from Dieter N=C3=BCtzel
(In reply to comment #=
23)
> (In reply to comment #22)
> > (In reply to comment #21<=
/a>)
> > > (In reply to comment=
#20)
> > > > (In reply to co=
mment #19)
> > > > > (In reply to comment #18)
> > > > > > Just tried oibaf's radeonsi_dri.so to copy h=
ere on Debian (just to be sure
> > > > > > :D) and it suffers from the same problem with=
-march=3Dgeneric ;)
> > > > >=20
> > > > > I mean -mtune=3Dgeneric of course :).
> > > > >=20
> > > > > Now to bisect where that SSE4.1 code start to com=
pile this slow (4
> > > > > minututes on Kabini just for that) :D Issue also n=
ot present on 64bit ;)
> > > >=20
> > > > I haven't seen the perf issue on my pure 64 bit setup, =
but have noticed that
> > > > the compile sits on sse4.1 for a while recently - just =
using one core, like
> > > > it's blocking the other threads of a make -j5 till it's=
done.
> > >=20
> > > Oh i bisected that ine, it takes 200X more time to compile =
on 32bit :D ,
> > > actually compile libmesa_gallium i think this is the commit:
> > >=20
> > > h=
ttp://cgit.freedesktop.org/mesa/mesa/commit/
> > > ?id=3Dd55f77b503ab7b59ecdd8f31c4f7dc498710e75b
> >=20
> > For this look, here:
> >=20
> > http://lists.freedesktop.org/archives/mesa-dev/2014-Augus=
t/065823.html
> >=20
> > Greetings and happy 'bisecting'...;-)
> >=20
> > Dieter
>=20
> So what is the solution for that after one month? Simply to use -O0 m=
aybe :)
Ping Jason Ekstrand, maybe 8-)