netdev
[Top] [All Lists]

RE: Fw: throttling kernel messages: KERNEL: assertion (flags &MSG_PEEK)

To: "KOVACS Krisztian" <hidden@xxxxxxxxxx>, "Olaf Kirch" <okir@xxxxxxx>
Subject: RE: Fw: throttling kernel messages: KERNEL: assertion (flags &MSG_PEEK) failed at net/ipv4/tcp.c (1282)
From: "Brandeburg, Jesse" <jesse.brandeburg@xxxxxxxxx>
Date: Wed, 13 Apr 2005 09:06:42 -0700
Cc: "Randy.Dunlap" <rddunlap@xxxxxxxx>, "netdev" <netdev@xxxxxxxxxxx>
Sender: netdev-bounce@xxxxxxxxxxx
Thread-index: AcU//cD71UWNH0kxSUOK78RwUNqC3wARLubA
Thread-topic: Fw: throttling kernel messages: KERNEL: assertion (flags &MSG_PEEK) failed at net/ipv4/tcp.c (1282)
>From: netdev-bounce@xxxxxxxxxxx [mailto:netdev-bounce@xxxxxxxxxxx] On
>Behalf Of KOVACS Krisztian
>2005-04-12, k keltezéssel 22.39-kor Olaf Kirch ezt írta:
>> >   I also hit these assertions today. In my case it was probably because
>> > of a slab corruption problem caused by the e100 driver. (I had
>> > CONFIG_DEBUG_SLAB enabled, so a lot of slab corruption messages
>appeared
>> > right before these assertion errors.)
>>
>> Yes, we've seen these symptoms in various bug reports in various 2.6
>> versions, and after some debugging narrowed it down to an e100 driver
>> issue. It went away with 2.6.11 final.
>
>  For me it definitely did not go away with 2.6.11. However, as soo as I
>applied e100-napi-fixes.patch the problem magically disappeared. So I
>suspect this NAPI state machine problem might have been the culprit.
>

This was the exact failure mode of this e100 bug.  It would exhibit with slab 
corruption, only when in NAPI mode.  Glad it fixed your issue as well.

Jesse


<Prev in Thread] Current Thread [Next in Thread>
  • RE: Fw: throttling kernel messages: KERNEL: assertion (flags &MSG_PEEK) failed at net/ipv4/tcp.c (1282), Brandeburg, Jesse <=