Matthias Urlichs ha scritto:
> Hi,
>
> Massimo Nuvoli:
>>> ast_mutex_lock(&ast_chan->lock);
>>> +
>>> + if (!ast_chan)
>>> + {
>>> + ast_log(LOG_WARNING, "Unexpexted ast_chan NULL\n");
>>> + return;
>>> + }
>>> The second hunk, testing ast_chan again after ast_mutex_lock(), does make
>>> no sense whatsoever. You have a few other places where you do this.
>> I found sometime after the mutex lock the ast_chan is NULL.
>> I think while stopped by the mutex the chan is closed so this is the
>> cause. Maybe i am wrong.. :-)
>>
> Umm, ast_chan is a local variable. It can't just transform itself to
> zero, unless something else is going very very wrong.
> First it is probabile that not every check is "right" (i added a lot of them in a seamlessy right position), second my idea is that while the program is waiting on the mutex the ast_chan struct is unallocated, and the mutex is released by a destroy, not by a unlock. :-) This week the "ISDN tester" come back to me and i do some test, i try to check. Bye.
signature.asc
Description: OpenPGP digital signature
_______________________________________________ Visdn-hackers mailing list [email protected] https://mailman.uli.it/mailman/listinfo/visdn-hackers
