Skip site navigation (1)Skip section navigation (2)
Date:      Sat, 16 Sep 2006 12:29:43 -0500
From:      "Christian S.J. Peron" <csjp@FreeBSD.org>
To:        Christian Brueffer <brueffer@FreeBSD.org>
Cc:        Andre Oppermann <andre@freebsd.org>, current@freebsd.org
Subject:   Re: em(4) interface wedges
Message-ID:  <450C3487.6050104@FreeBSD.org>
In-Reply-To: <20060915191008.GE2020@haakonia.hitnet.RWTH-Aachen.DE>
References:  <20060915165351.GC2020@haakonia.hitnet.RWTH-Aachen.DE>	<450ADE55.7090509@samsco.org> <450AEC50.2060404@freebsd.org> <20060915191008.GE2020@haakonia.hitnet.RWTH-Aachen.DE>

next in thread | previous in thread | raw e-mail | index | archive | help
Are you using dhclient (or do you have any other BPF peers attached to 
the network interface?)

Christian Brueffer wrote:
> On Fri, Sep 15, 2006 at 08:09:20PM +0200, Andre Oppermann wrote:
>   
>> Scott Long wrote:
>>     
>>> Christian Brueffer wrote:
>>>
>>>       
>>>> Hi,
>>>>
>>>> with the latest em(4) code in current (sans the printf/device_printf
>>>> change), my interface wedges every few hours.  It can be resurrected
>>>> by bringing it down and up again.
>>>>
>>>> The device in question is a E1000_DEV_ID_82540EP_LP built into my
>>>> Thinkpad.
>>>>
>>>> Anyone else seeing this?
>>>>
>>>> - Christian
>>>>
>>>>         
>>> Yes.  It's also almost guaranteed to wedge the first time I do a remote
>>> CVS operation after boot.
>>>       
>> When the interface wedges does it show the OACTIVE flag in ifconfig output?
>> This is an important hint.
>>
>>     
>
> Yeah, OACTIVE is set.
>
>   
>> If you are running at 100Mbit there seems to be bug in the em(4) harware
>> with TSO.  Try disabling TSO on it with 'ifconfig em0 -tso'.
>>
>>     
>
> Correct, I'm using it at 100Mbit and disabling TSO works around the
> issue.
>
> - Christian
>
>   


-- 
Christian S.J. Peron
csjp@FreeBSD.ORG
FreeBSD Committer
FreeBSD Security Team




Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?450C3487.6050104>