[mythtv-users] DEBUGGING INFO: Kernel message out of nowhere

R. G. Newbury newbury at mandamus.org
Sun Aug 19 05:22:37 UTC 2007

Lan Barnes wrote:
> On Sat, August 18, 2007 12:03 pm, Andy Burns wrote:
>> On 18/08/2007 19:46, Lan Barnes wrote:
>>> I would also encourage all us users to NOT respond so it doesn't turn
>>> into
>>> a useless noise thread.
>> As far as I can see, it's already noise, because it's a crash caught in
>> the kernel (and a tainted one at that) rather than myth, further looks
>> like it was bash within one of your SSH sessions that caused it.
> A rather harsh judgement, considering any crash report shows some effort
> to be of help, but I'll shine it on as either unintended or a personality
> thing.
> Curious that bash/ssh would choose to do that on a myth machine having
> never done that previously in years of use on any of my other FC/RH boxes.
> But obviously I passed this on because I'm not qualified to assess it.
> BTW, I was too sanguine in reporting that all was well. Myth FE is frozen
> on xena, and (as you might expect from your analysis of the crash), I can
> ping but I can't ssh in to do a clean shutdown.
> I'll leave it up a little longer in case anyone suggests a worthwhile
> experiment in this condition and then do a cold boot.
> Hmm ... and maybe I'll try to ssh from a different box (unlikely to work)
> and to run a remote FE (might tell me if the BE is still OK).
Earlier this afternoon, I noticed that xorg.conf has 2 options 
"AllowDeactivateGrabs" and "AllowCloseDownGrabs" (see 'man xorg.conf').

If I read these correctly, if enabled it might allow you to grab back 
the keyboard and/or mouse which have been frozen by the locked program 
(mythfrontend in this case). I haven't had a chance to check this yet, 
but when mythfrontend locks up, it *always* freezes the keyboard too....

Just a thought for a future case..


More information about the mythtv-users mailing list