From mboxrd@z Thu Jan 1 00:00:00 1970 From: Andrew Morton Subject: Re: 2.5.70-bk+ broken networking Date: Wed, 4 Jun 2003 18:56:52 -0700 Sender: linux-kernel-owner@vger.kernel.org Message-ID: <20030604185652.31958d1f.akpm@digeo.com> References: <20030604161437.2b4d3a79.shemminger@osdl.org> <3EDE7FEB.2C7FAEC7@digeo.com> Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Return-path: To: shemminger@osdl.org, jgarzik@pobox.com, davem@redhat.com, netdev@oss.sgi.com, linux-kernel@vger.kernel.org In-Reply-To: <3EDE7FEB.2C7FAEC7@digeo.com> List-Id: netdev.vger.kernel.org Andrew Morton wrote: > > Stephen Hemminger wrote: > > > > Test machine running 2.5.70-bk latest can't boot because eth2 won't > > come up. The same machine and configuration successfully brings up > > all the devices and runs on 2.5.70. > > kjournald is stuck waiting for IO to complete against some buffer > during transaction commit. > > I'd be suspecting block layer or device drivers. What device driver > is handling your /var/log? I take that back. Your sysrq-T woke up syslogd which did a synchronous write which poked kjournald. You happened to catch it in mid-commit. So that's all normal and sane. Something is up with netdevice initialisation. My eth0 (e100) is in a strange half-there state and won't come up. Reverting the post-2.5.70 e100 changes does not help. It's something which went into the tree today I think.