Skip site navigation (1)Skip section navigation (2)
Date:      Fri, 15 Oct 2004 20:39:34 -0600
From:      Scott Long <scottl@freebsd.org>
To:        David Schultz <das@freebsd.org>
Cc:        current@freebsd.org
Subject:   Re: Stable panic on shutdown: swapoff: failed to locate N swap blocks
Message-ID:  <417089E6.8030105@freebsd.org>
In-Reply-To: <20041016021131.GA72979@VARK.MIT.EDU>
References:  <20041015075641.GA6820@nagual.pp.ru> <20041015171144.GA69709@VARK.MIT.EDU> <20041015171700.GA74901@nagual.pp.ru> <20041016021131.GA72979@VARK.MIT.EDU>

next in thread | previous in thread | raw e-mail | index | archive | help
David Schultz wrote:
> On Fri, Oct 15, 2004, Andrey Chernov wrote:
> 
>>On Fri, Oct 15, 2004 at 01:11:45PM -0400, David Schultz wrote:
>>
>>>On Fri, Oct 15, 2004, Andrey Chernov wrote:
>>>
>>>>N of blocks can be different.
>>>>Can't provide full debugging trace yet, excepting this one:
>>>
>>>The traceback won't be very helpful in this case because this most
>>>likely indicates a swap leak that happened some time earlier.
>>>What *would* be helpful is more information about your swap
>>>devices, any steps you need to take while the system is running to
>>>reproduce this, the approximate size of N, etc.
>>
>># swapinfo
>>Device          1K-blocks     Used    Avail Capacity
>>/dev/da0s1b        615408    41408   574000     7%
>>
>>Nothing special to reproduce, it is always reproduced automatically at 
>>shutdown.
> 
> 
> You're clearly using some of that swap space for something, so it
> would be great if you could narrow it down a bit more.  For
> instance, does the problem only occur if you shut the system down
> after having run particular applications?  Are there ever any disk
> errors?
> 
> I'll poke around and see if I can find anything by inspection.
> phk's rotitilling of the swap subsystem introduced a number of new
> nits, but no serious bugs that I can see...
> 
> 
>>Kernels from Aug 8 and below never do that, I see this panic constantly 
>>only with most recent kernels (last 2 days). I not test kernels between 
>>Aug 8 and last 2 days, so can't say anything.
> 
> 
> The bug was probably introduced earlier, either by me, Alan, or
> phk, but the swapoff codepath wasn't routinely exercised until
> recently.

FWIW, I think that doing a swapoff in the shutdown path is just asking
for trouble.  Fixing whatever bug this is would of course be nice, but
the need for swapoff here is a hack and only opens up up to problems.

Scott



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?417089E6.8030105>