ffmpeg & ULE
Eduardo Morras
nec556 at retena.com
Tue Oct 18 16:41:35 UTC 2011
At 17:45 18/10/2011, Ivan Klymenko wrote:
>Ð Tue, 18 Oct 2011 16:30:22 +0100
>Vincent Hoffman <vince at unsane.co.uk> пиÑеÑ:
>
> > On 18/10/2011 16:24, Ivan Klymenko wrote:
> > > Ð Tue, 18 Oct 2011 23:02:47 +0800
> > > Adrian Chadd <adrian at freebsd.org> пиÑеÑ:
> > >
> > >> What is FBFS?
> > >>
> > >>
> > >>
> > >> Adrian
> > > :)
> > > http://rudot.blog.com/
> > or
> > http://wiki.freebsd.org/RudolfTomori/rudotSoC2011
> >
> > The blog entry was interesting in that it gives some context to the 2
> > images you previously posted as does the wiki entry.
> >
> > Vince
>
>they are very different four images - made on different systems
>Forgive me for offtopic :(
I think it's on topic. Andriy Gapon requested any
facts that demostrates or shows examples when ULE
scheduler is not so good than others. The images
shows how the 3 schedulers performs with the same
workload when number of cores is changed. Perhaps
it's not a scientific and statistically valid
but, for me, it shows when is ULE under other schedulers.
Also, some people has asked and pointed about
using 4BSD and not ULE in 8.x because it works
better for some type of workload.
More information about the freebsd-performance
mailing list