[Nut-upsuser] Kernel crash when using usbhid-ups driver.

[Paul Whittaker]

Hi Arnaud,

Great to hear this is an isolated case; hopefully it means we've found a
bug we can help solve.

On 27/11/12 22:50, Arnaud Quette wrote:
> please send the log here, in compressed form.
> are the photos taken with a still cam?
> please put these on your website (or any other public storage) and
> point the link
For each of the four crashes I've got syslog excerpts (which include
some messages from just before and just after the crash) and photos of
the screen.  I couldn't take the whole screen at once clearly, so I took
several photos that overlap.  The photos were pretty bad (low end
smartphone - sorry) but they should be legible; I've treated them to
improve clarity and image size.

Crashes 1 and 2 used the same (fairly basic) setup.  After crash 2 I
added the 'pollonly' driver flag, and after crash 3 I removed that and
was running the driver with -D instead.

Crash 1 - A.png <http://ubuntuone.com/0J9nFa80fvGSou2eytvm92> B.png
<http://ubuntuone.com/5xJS7R2NuKR7jlDsyHa60B> (syslog.1.gz attached)
Crash 2 - A.png <http://ubuntuone.com/3oDzexDvymLlEZp9n08yi5> B.png
<http://ubuntuone.com/4E75bm8hP1Dn26ZALc38P2> C.png
<http://ubuntuone.com/5R9lx8sxILvfTSLElZwB75> D.png
<http://ubuntuone.com/2Z9f2KhxgsuoQUZKCgyWUr> (syslog.2.gz attached)
Crash 3 - A.png <http://ubuntuone.com/3dLbezmhjk94EH1NWebId8> B.png
<http://ubuntuone.com/74ZOCo4ymmm4ONUsRgdeSL> (syslog.3.gz attached)
Crash 4 - A.png <http://ubuntuone.com/0BSntOkyOuB935R6oxie7v> B.png
<http://ubuntuone.com/735zksHQYCzeidhTjJM28P> C.png
<http://ubuntuone.com/1Bs3sn2rIRU5vkqXe8Pc2x> (syslog.4.gz attached)

There should be most kernel details you need in the screenshots.  This
is running the latest nut-server package from Ubuntu Precise 64-bit
(version 2.6.3-1ubuntu1.1 at present) on a Dell PowerEdge R210 II.

> a dump file would be preferable:
> https://help.ubuntu.com/12.04/serverguide/kernel-crash-dump.html
I'll see what I can do; this is our main development server, so I'll
probably be limited to weekends.

I'm not very familiar with what data crash dumps record, but do know
that I can only post one if I'm sure it doesn't contain company file
data, encryption keys, VM state, etc. (for obvious reasons).  Since I
know in advance I'll be taking a crash dump, are there any steps I can
take steps to avoid such data?  If not, I can still run tools on the
dump locally (out of hours) and send backtraces or other output you
might need.

Let me know if any other details, config files, etc., would help.

Cheers,

Paul.

PS I've replied only to the list; do you prefer that I reply to you and
CC: the list instead?

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.alioth.debian.org/pipermail/nut-upsuser/attachments/20121128/c3246939/attachment.html>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: syslog.1.gz
Type: application/x-gzip
Size: 919 bytes
Desc: not available
URL: <http://lists.alioth.debian.org/pipermail/nut-upsuser/attachments/20121128/c3246939/attachment.bin>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: syslog.2.gz
Type: application/x-gzip
Size: 1669 bytes
Desc: not available
URL: <http://lists.alioth.debian.org/pipermail/nut-upsuser/attachments/20121128/c3246939/attachment-0001.bin>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: syslog.3.gz
Type: application/x-gzip
Size: 2707 bytes
Desc: not available
URL: <http://lists.alioth.debian.org/pipermail/nut-upsuser/attachments/20121128/c3246939/attachment-0002.bin>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: syslog.4.gz
Type: application/x-gzip
Size: 769 bytes
Desc: not available
URL: <http://lists.alioth.debian.org/pipermail/nut-upsuser/attachments/20121128/c3246939/attachment-0003.bin>