[linux-audio-dev] deblocking snd_seq_event_input()

[Jens M Andreasen]

On Sun, 2005-05-08 at 12:50 +0200, Fons Adriaensen wrote: 
> On Sun, May 08, 2005 at 12:00:26PM +0200, Jens M Andreasen wrote:
> 
> > Can't you use:
> > 
> >   void pthread_cleanup_push(void (*routine) (void *), void *arg);
> >   void pthread_cleanup_pop(int execute);
> 
> Does the thread still exist as a thread when routine() is called, or more
> specific, is routine() called from the thread's own context ? This is not clear
> from the man page - routine() could be called from a different thread than
> the one that is being cancelled. 

The routine() is definately called from the threads own context. A very
common variant is routine() == free() and *arg is some malloced memory
that is only known to this thread.

> 
> But this is besides the point. In my original problem it is not the intention
> to terminate the thread - just to force the snd_seq_event_input() to return.
> The thread could well close the handle, open a new one and continue normal
> operation. The simplest solution is probably to send some fake event to the
> midi port.

Perhaps poll() with a shortish timeout? Wake up now and then and do some
housekeeping. Otherwise it looks like pthread_kill would or should have
been the one if not for implementation issues (although here I am
uncertain if it isn't so that the documentation of BUGS is outdated for
kernel 2.6?)

> 
-- 

   (
    )
  c[]  //  Jens M Andreasen