Date: Wed, 6 Jun 2007 18:49:17 +0400 From: Gleb Smirnoff <glebius@FreeBSD.org> To: Ed Schouten <ed@fxq.nl> Cc: current@FreeBSD.org Subject: Re: ifconfig carp0 destroy = kernel panic Message-ID: <20070606144917.GL89017@glebius.int.ru> In-Reply-To: <20070606143528.GB89502@hoeg.nl> References: <465BE214.8040502@monkeybrains.net> <20070530105120.GT89017@FreeBSD.org> <20070603144732.GA45756@hoeg.nl> <20070606143528.GB89502@hoeg.nl>
next in thread | previous in thread | raw e-mail | index | archive | help
On Wed, Jun 06, 2007 at 04:35:28PM +0200, Ed Schouten wrote: E> * Ed Schouten <ed@fxq.nl> wrote: E> > Please take a look at this patch as well. It has been lying around in E> > GNATS for some time and it really makes me go insane a lot of times: E> > E> > [...] E> E> Looks like it just got committed. Thanks a lot! :) Btw, the problems do not end at this. On HEAD (didn't check RELENG_6), the following sequence panics: ifconfig em0 10.0.0.2/24 ifconfig carp0 create ifconfig carp0 10.0.0.3/24 vhid 1 ifconfig carp1 create ifconfig carp1 10.0.0.4/24 vhid 1 (ifconfig returns EINVAL. This EINVAL comes via in_control()->in_ifinit()->carp_ioctl()->carp_setrun(). Really it should be EEXIST. I will change this later.) After this the last ifconfig command we have somewhat garbaged IP stack - the interface address lists and ia_hash reference a freed memory. The panic comes after an ARP request or if you repeat the command again: ifconfig carp1 10.0.0.4/24 vhid 1 I am now trying to understand the panic. -- Totus tuus, Glebius. GLEBIUS-RIPN GLEB-RIPE
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20070606144917.GL89017>