panic in uma_startup for many-core amd64 system
Andriy Gapon
avg at freebsd.org
Tue Oct 19 06:55:34 UTC 2010
on 19/10/2010 00:01 Giovanni Trematerra said the following:
> On Mon, Oct 18, 2010 at 4:18 PM, Andriy Gapon <avg at freebsd.org> wrote:
>> Again, not sure if I follow you, I don't see relation between per-cpu caches and
>> dynamic slab size.
>
> Your patch seems just a work around about initial slab size where the
> keg is backed.
Well, setting aside my confusion with the terminology - yes, the patch is just
that, and precisely because I only tried to solve that particular problem.
> Having dynamic slab sizes would allow to have the keg backed on a larger slab
> without going OFFPAGE.
I agree in principle.
But without seeing code that implements that I can't guess if it would really be
more efficient or more maintainable, i.e. more useful in general.
Still a very good idea.
--
Andriy Gapon
More information about the freebsd-current
mailing list