[Nut-upsuser] tripplite_usb/nut-2.2.1: Driver not connected.

[Charles Lepple]

On Thu, May 29, 2008 at 1:00 PM, Forest Bond <forest at alittletooquiet.net> wrote:
> Hi,
>
> On Wed, May 28, 2008 at 10:01:58PM -0400, Charles Lepple wrote:
>> On Wed, May 28, 2008 at 4:55 PM, Forest Bond <forest at alittletooquiet.net> wrote:
>> > May 28 15:43:34 logicserver tripplite_usb[27408]: Error reading B value: Device detached? (error 0: error sending control message: Operation not permitted)
>> > May 28 15:43:34 logicserver upsd[27410]: Data for UPS [tripplite] is stale - check driver
>> > May 28 15:43:34 logicserver tripplite_usb[27408]: Reconnect attempt #1
>> > May 28 15:43:34 logicserver upsd[27410]: UPS [tripplite] disconnected - check driver
>> > May 28 15:43:34 logicserver upsd[27410]: Can't connect to UPS [tripplite] (tripplite_usb-tripplite): Connection refused
>> > May 28 15:43:34 logicserver kernel: [1257714.534999] tripplite_usb[27408]: segfault at 0000001d eip b7e664ac esp bfd478a8 error 4
>>
>> Given that this is Ubuntu Hardy, Is there a file with a name like
>> /var/crash/*tripplite_usb.txt?
>>
>> If so, it might be best to send it to me off-list (I don't know how
>> big those files typically are).
>
> No such file existed.  I realized that apport wasn't installed, so I installed
> that and triggered the crash again.  I couldn't get apport to generate any such
> crash report.

OK. It was a shot in the dark - I haven't had a chance to work with
apport, but in theory it looks very powerful.

> Next, I tried to generate a plain old core file with:
>
> # ulimit -c 50000
> # /lib/nut/tripplite_usb -a tripplite -DDD
>
> I still get the segfault, but I can't seem to generate a core file.  Would there
> be something funny about this issue that would prevent me from getting a core
> dump?  I don't know much about these things, to be honest, but I'd love to help
> debug.

Not sure why it won't generate a core file, although it may have
something to do with switching userids. Can the 'nut' system user
write to the directory where you started the driver? Did it put the
file in one of the NUT system directories? (/var/lib/nut,
/var/run/nut, etc.)

Arjen pointed out what is the most likely cause of the segfault (our
reconnection code has changed a bit since the driver was first
written), but the problem is that I won't have time to work on that
for a week or so.

-- 
- Charles Lepple