Hi,

Thanks for the problem report.

On Sat, Feb 17, 2007 at 10:40:37AM +0100, Markus W Kilbinger wrote:

> Do(es) the recent newlock2 changes / merge show up in a (n interim?)
> different bahvior for userland porograms?

It shouldn't result in the behaviour that you describe, no.
 
> I'm running asterisk on a cobalt qube2 which works quite flawlessly
> (well, only core dumps at shutdown) with recent pre-newlock2 -current
> kernel and userland (especially libpthread*).
> 
> After updating the whole machine to the newlock2 base (kernel +
> userland) asterisk seems to startup fine, but does no longer accept
> (all) incoming phone calls!?
> 
> Just reverting to my most recent pre-newlock2 kernel and libpthread*
> (source of Feb 7th) restores the old, correct asterisk behavior.
> 
> It does not seem to depend on asterisk compilation environment (it
> compiles fine with newlock and pre-newlock2 userland), newlock2 and
> pre-newlock2 compiled asterisks both show the faulty behavior with
> newlock2 kernel and libpthread* and both work fine with the
> pre-newlock2 kernel and libpthread*.
> 
> So, does the newlock2 merge result in a different pthread
> functionality for userland programs?
> 
> Any comments/explanations appreciated,

If you could file a PR about this it would be ideal. A good first step for
diagnostic the problem would be to attach to the process responsible for
handling incoming calls with "ktrace -di -p $pid", and make available the
ktrace.out file that is produced (or relevant excerpts from kdump -R,
afterwards).

Cheers,
Andrew