Skip site navigation (1)Skip section navigation (2)
Date:      Wed, 27 Nov 2013 02:05:54 -0800
From:      Adrian Chadd <adrian@freebsd.org>
To:        Craig Rodrigues <rodrigc@freebsd.org>
Cc:        thierry.herbelot@free.fr, "freebsd-virtualization@freebsd.org" <freebsd-virtualization@freebsd.org>
Subject:   Re: VIMAGE: Freed UMA keg was not empty
Message-ID:  <CAJ-VmonPZosKJ7dySRhCMPb0USsmaH9JNzNOZz_T0Z=8NPxNhw@mail.gmail.com>
In-Reply-To: <CAG=rPVc4dYcXXjSy%2BTV_GhrxKqszQ8JF5B56tUPQN-QmuDv7NA@mail.gmail.com>
References:  <CAG=rPVc4dYcXXjSy%2BTV_GhrxKqszQ8JF5B56tUPQN-QmuDv7NA@mail.gmail.com>

next in thread | previous in thread | raw e-mail | index | archive | help
Well, the first step is figuring out which UMA zones are actually
problematic. Isn't it logging which zones aren't empty?


-a


On 26 November 2013 20:01, Craig Rodrigues <rodrigc@freebsd.org> wrote:
> Adrian,
>
> As part of looking at this bug in FreeNAS:
>
> "Panic on new jail with vnet"
> https://bugs.freenas.org/issues/3102
>
> I started looking at this thread:
>
> http://lists.freebsd.org/pipermail/freebsd-current/2010-November/021280.html
>
> Based on this clue from Thiery:
>
> http://lists.freebsd.org/pipermail/freebsd-current/2010-November/021291.html
>
> I fixed some VIMAGE related memory leaks related to routetbl.
> I also fixed some VIMAGE related kernel panics when
> I tried creating and deleting lots of jails:
>
> http://lists.freebsd.org/pipermail/svn-src-all/2013-November/077178.html
> http://lists.freebsd.org/pipermail/svn-src-all/2013-November/077192.html
> http://lists.freebsd.org/pipermail/svn-src-all/2013-November/077195.html
>
> I haven't eliminated all the "Freed UMA keg was not empty"
> error messages.
>
> Do you have any recommendations for how I can track down the last
> few UMA related memory leaks?  I am new to debugging UMA so
> would appreciate any pointers that you may have.
> --
> Craig



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?CAJ-VmonPZosKJ7dySRhCMPb0USsmaH9JNzNOZz_T0Z=8NPxNhw>