On Jan 31, 2008 3:10 AM, Mike Battersby <[EMAIL PROTECTED]> wrote:
> Derrick Brashear wrote:
> > which thread?
>
> The process forked at afsd.c:2001.
>
> > code = fork();
> > if (code == 0) {
> > /* Child */
> > SET_RX_RTPRI(); /* max advised for non-interrupts */
> > call_syscall(AFSOP_RXLISTENER_DAEMON, preallocs,
> enable_peer_stats,
> > enable_process_stats);
> > exit(1);
> > }
>
> The afs_osi_Invisible call sets SSYS on this forked child which exits.
> Bad idea.
>
All the forked children exit, eventually, but...
>
> > where are we calling thread_create at all?
>
> Where did I say you were?
>
Actually, we are in rxk_Listener... (rx_kcommon.c) which is why this is
exiting immediately. That's the piece of information I wanted.
In any case, the patch will deal with this.
