According to blymn:
>
>According to Jukka Marin:
>>
>>Yep, that's too sad.  One would think there was a -debug switch or something,
>>but no.  I added some printf's to the code and saw that the serial port
>>thread was not running.  Also, ktrace output shows that gnokii never even
>>tries calling open(2) on the tty port.  Then, the first thread gets tired of
>>waiting for a flag set by the serial thread, prints the error message and
>>quits.  I have no idea why the threads do not work.
>>
>
>Very odd - I was using the unproven-pthreads package as my threading
>package, I am sure other things we have work with pthreads.  If I get
>time I shall see if I can dig into why that open is not happening.
>

OK, I did some digging on this on the weekend.  The results are that,
yes, the serial port is opened and the parameters set.  Stuff is
output on the line and then the thread waits for responses.  The
thread is waiting for a SIGIO to be delivered but it never gets one.
At the moment I cannot tell you if this is because the phone never
responds (unlikely, others have this working) or because the signal is
being blocked.

I know there is some voodoo surrounding pthreads and signals, I am
wondering if this is the cause of the failure.

-- 
===============================================================================
Brett Lymn, Computer Systems Administrator, BAE SYSTEMS
===============================================================================