From: "Tim Warnock" <timoid@getonit.net.au>
To: <linux-kernel@vger.kernel.org>
Cc: "Tim Warnock" <timoid@getonit.net.au>
Subject: RE: FW: Kernel oops v2.4.31 in e1000 network card driver.
Date: Fri, 23 Dec 2005 10:06:19 +1000 [thread overview]
Message-ID: <C67FBCB411B4024382B11B96D68E49E4079693@server.local.GetOffice> (raw)
> -----Original Message-----
> From: Willy Tarreau [mailto:willy@w.ods.org]
> Sent: Friday, 23 December 2005 9:17 AM
> To: Tim Warnock
> Cc: linux-kernel@vger.kernel.org
> Subject: Re: FW: Kernel oops v2.4.31 in e1000 network card driver.
>
> Hello,
>
> On Thu, Dec 22, 2005 at 07:10:04PM +1000, Tim Warnock wrote:
> > Further information to this:
> >
> > Network card causing the problem is the intel quad port
> > gigabit ethernet pci card.
> > I have tested also on 2.4.27, 2.4.32 and the latest 2.6
> > series kernel.
> >
> > Under load (10-15kpps) the network driver crashes. Under
> > increased load (20-30kpps) the driver will actually cause
> > a full kernel panic and reboot the box.
>
> What type of system is it ? P3/P4/K7/K8 ? UP/SMP ? do you have a PCI-X
> bus in it ? have you checked /proc/interrupts for strange behaviours ?
IBM x306 p4 3.0ghz UP/HT (acpi=ht) 1gb ram
It's a 64bit pci bus, the quad card is 64 bit also.
Underlying host os is debian sarge.
I never thought to check /proc/interrupts. I get the network card back
from remote soon, I will put it in a bench system and see what happens.
What should I be looking for in /proc/interrupts?
> > After replacing the card with a single port intel gigabit
> ethernet pci card, the system has been rock stable.
> >
> > According to intel, the quad port nic is based around the "Intel(r)
> > 82546EB" controller, the single port nic is based around
> > the "Intel(r) 82545" controller.
> >
> > Are there any other known problems with Intel(r) 82546EB controller
> > support with the current e1000 driver?
>
> Not to my knowledge. I have several of them running on moderate volume
> (50 Mbps) on production up to 50-60 kpps, and they have never
> ever caused any trouble after 2.5 years. I even use others in
> stress-testing machines which throw up to 500 kpps per port without
> any problem either. BTW, the ones in the stress-testers are more
> recent, they are the ones with the "toundra" PCI bridge.
>
> Do you encounter the problem in only one system ? I begin to suspect a
> hardware failure somewhere (CPU, RAM) which is triggered by
> higher loads.
>
next reply other threads:[~2005-12-23 0:06 UTC|newest]
Thread overview: 5+ messages / expand[flat|nested] mbox.gz Atom feed top
2005-12-23 0:06 Tim Warnock [this message]
2005-12-23 5:23 ` FW: Kernel oops v2.4.31 in e1000 network card driver Willy Tarreau
-- strict thread matches above, loose matches on Subject: below --
2005-12-22 9:10 Tim Warnock
2005-12-22 23:16 ` Willy Tarreau
2005-12-23 0:52 ` Jeff Kirsher
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=C67FBCB411B4024382B11B96D68E49E4079693@server.local.GetOffice \
--to=timoid@getonit.net.au \
--cc=linux-kernel@vger.kernel.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).