netdev - RE: Lots of "BUG eth1 code -5 qlen 0" messages in 2.6.24

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Date:	Tue, 29 Jan 2008 11:57:33 -0800
From:	"Waskiewicz Jr, Peter P" <peter.p.waskiewicz.jr@...el.com>
To:	"Erik Mouw" <mouw@...linux.org>, <netdev@...r.kernel.org>
Subject: RE: Lots of "BUG eth1 code -5 qlen 0" messages in 2.6.24

> I've just started to use 2.6.24 on my home firewall (before 
> it was running 2.6.24-rc2 for about 65 days) and I noticed a 
> couple of error messages I've never seen before:
> 
> Jan 29 07:50:54 gateway kernel: BUG eth1 code -5 qlen 0 Jan 
> 29 08:28:30 gateway kernel: BUG eth1 code -5 qlen 0 Jan 29 
> 08:57:30 gateway kernel: BUG eth1 code -5 qlen 0 Jan 29 
> 09:44:04 gateway kernel: BUG eth1 code -5 qlen 0 Jan 29 
> 10:01:35 gateway kernel: BUG eth1 code -5 qlen 0 Jan 29 
> 10:01:35 gateway last message repeated 2 times Jan 29 
> 10:16:48 gateway kernel: BUG eth1 code -5 qlen 0 Jan 29 
> 10:16:48 gateway last message repeated 2 times Jan 29 
> 10:45:48 gateway kernel: BUG eth1 code -5 qlen 0 Jan 29 
> 10:45:48 gateway last message repeated 2 times Jan 29 
> 11:10:01 gateway kernel: BUG eth1 code -5 qlen 0 Jan 29 
> 11:10:02 gateway last message repeated 9 times
> 
> The message seems to be coming from the qdisc_restart() in 
> net/sched/sch_generic.c which was changed with commit 
> 5f1a485d5905aa641f33009019b3699076666a4c .
> 
> The NIC is an IBM EtherJet cardbus card using the xircom_cb driver:

Are you using any specific qdisc, or just the default pfifo_fast?  Have
you done any specific tuning on your qdisc as well?  The default qlen
seems to have been changed.

Basically your queue is being overrun, and with the current checks in
the kernel in the stack, it's allowing the skb into the driver.  I've
known about this issue, and I'm hesitant to push a patch to re-add the
netif_queue_stopped() check into qdisc_restart().  I'd rather push a
one-time patch to the drivers that interacts with
netif_stop_subqueue(netdev, 0), so we can completely decouple the single
queue from the netdev.

I'd say you can somewhat ignore the messages for now.  But there is work
to be done here, and it's obvious I need to do this sooner than later.
Please let me know about the qdisc parameters though when you get a
chance.

Thanks,

-PJ Waskiewicz
<peter.p.waskiewicz.jr@...el.com>
--
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html