[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

RE: [Xen-devel] problem with the pro/1000 driver for intel gigabitethernet card



I have seen this (intermittent) problem on two 4P boxes and a laptop, all using e1000 driver.

What kind of network input triggers the end of the hang? I have seen this hang on a busy network with lots of broadcast traffic.  Did not see it with a network restart that was separate from domain shutdown.

Bug filed: http://bugzilla.xensource.com/cgi-bin/bugzilla/show_bug.cgi?id=35

Rich

-----Original Message-----
From: xen-devel-bounces@xxxxxxxxxxxxxxxxxxx on behalf of Kip Macy
Sent: Wed 5/11/2005 12:06 PM
To: Jon Mason
Cc: peter bier; xen-devel@xxxxxxxxxxxxxxxxxxx
Subject: Re: [Xen-devel] problem with the pro/1000 driver for intel gigabitethernet card

We're not seeing it on 2.0.5 either. The first thing I would do would
be to diff the driver between 2.6.10 and 2.6.11.

                      -Kip

On 5/11/05, Jon Mason <jdmason@xxxxxxxxxx> wrote:
> I have been seeing the exact same problem for a while (on xen-unstable).  I
> think I didn't see this problem on Xen 2.0.5 (but I'm not 100% sure).
>
> I can open a bug in bugzilla.xensource.com if no one else is willing.
>
> Thanks,
> Jon
>
> On Wednesday 11 May 2005 12:41 pm, Kip Macy wrote:
> > I can confirm that I've seen the exact same behaviour.
> >
> >                     -Kip
> >
> > On 5/11/05, peter bier <peter_bier@xxxxxx> wrote:
> > > I encountered a problem with dom0 when rebooting and running an intel
> > > gigabit card using the pro/1000 driver. When the system tries to shut
> > > down I get the following output of an strace command :
> > >
> > >    .........
> > > 17:10:49.063616 socket(PF_INET, SOCK_DGRAM, IPPROTO_IP) = 3 <0.000029>
> > > 17:10:49.063689 ioctl(3, 0x8913, 0xbffffc80) = 0 <0.000018>
> > > 17:10:49.063751 ioctl(3, 0x8914, 0xbffffc80) = 0 <535.114722>
> > > 17:19:44.178567 close(3)                = 0 <0.000026>
> > > 17:19:44.178679 exit_group(0)           = ?
> > >
> > > observe the more than 500 seconds it takes for the second ioctl to
> > > complete !! I have found out that any input on the console keyboard or
> > > network interface immediately causes the system call to return.
> > >
> > > Is there a problem with interrupt delivery ?
> > >
> > > Peter Bier
> > >
> > > _______________________________________________
> > > Xen-devel mailing list
> > > Xen-devel@xxxxxxxxxxxxxxxxxxx
> > > http://lists.xensource.com/xen-devel
> >
> > _______________________________________________
> > Xen-devel mailing list
> > Xen-devel@xxxxxxxxxxxxxxxxxxx
> > http://lists.xensource.com/xen-devel
>
> _______________________________________________
> Xen-devel mailing list
> Xen-devel@xxxxxxxxxxxxxxxxxxx
> http://lists.xensource.com/xen-devel
>

_______________________________________________
Xen-devel mailing list
Xen-devel@xxxxxxxxxxxxxxxxxxx
http://lists.xensource.com/xen-devel

_______________________________________________
Xen-devel mailing list
Xen-devel@xxxxxxxxxxxxxxxxxxx
http://lists.xensource.com/xen-devel

 


Rackspace

Lists.xenproject.org is hosted with RackSpace, monitoring our
servers 24x7x365 and backed by RackSpace's Fanatical Support®.