Skip site navigation (1)Skip section navigation (2)
Date:      Sun, 10 Dec 2006 08:42:35 +0000 (GMT)
From:      Robert Watson <rwatson@FreeBSD.org>
To:        Maxim Konovalov <maxim@macomnet.ru>
Cc:        freebsd-current@FreeBSD.org, yal <yal@yal.hopto.org>
Subject:   Re: CURRENT freezes on Laitude D520
Message-ID:  <20061210083752.G9926@fledge.watson.org>
In-Reply-To: <20061210013735.D11309@mp2.macomnet.net>
References:  <52944.192.168.1.110.1165679313.squirrel@yal.hopto.org> <20061209195519.B60055@mp2.macomnet.net> <20061209204924.N9926@fledge.watson.org> <20061210013735.D11309@mp2.macomnet.net>

next in thread | previous in thread | raw e-mail | index | archive | help
On Sun, 10 Dec 2006, Maxim Konovalov wrote:

>> Do you have any information about the actual source of this problem?
>
> No, I don't.  The whole system is unstable.  E.g. ifconfig hangs at the 
> random time.  Wireless stops working, mouse is lagging.  Suddenly the whole 
> system just hangs.
>
> yal@ replied in the private email debug.mpsafenet="0" did help him too.
>
>> Would it motivate solving the problem if I were to remove the 
>> debug.mpsafenet tunable? :-)
>
> I didn't suggest to turn off mpsafenet forever and forget, I just wanted to 
> check my guess.  I would like to help to debug the problem but I need some 
> initial instructions to start.  There is a firewire console.  What do I need 
> to check?

Start with the information in my followup e-mail to Andrew:

- Configure WITNESS and see if you get any console output regarding lock order
   problems.

- Try setting net.isr.direct=0 and see if the problem goes away.

- Try removing options PREEMPTION and see if the problem goes away.

Don't try them all at once, rather, try them individually until you've managed 
to find some arrangement that makes the source of the problem more clear 
(i.e., a WITNESS warning involving, say, the network stack).

When suggesting debug.mpsafenet=0 as a debugging step, please make it clear 
that this is not a solution.  The hope is that in 7.0, all components of the 
network stack will be MPSAFE, so whether in the short or long term, it is 
going away, and workarounds masking the problem don't get us any closer to 
solving the problem.

Robert N M Watson
Computer Laboratory
University of Cambridge



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20061210083752.G9926>