Re: 1.2.11 + Linux 2.4.4 grief

[James Macnicol <James.Macnicol@faceng.anu.edu.au>]

Thanks to those of you who replied.  I'd like to make the following
extra observations/comments :

*	A number of you have suggested that I look again at the documents on
arena.sci.univr.it.  As far as I can tell this site has been down for some
time (doesn't work now, didn't work last week either).

*	The SIGSTOP behaviour I mentioned isn't the problem: strace send the
process to be traced one of these.  Sorry for not reading the manpage.

*       My latest theory is that this may be a problem with pthreads.  I
can't reproduce this just at the moment but yesterday I tried loading
the "hung" slapd in gdb (rather than strace) and it was stuck at
pthread_join().  When it has died I've noticed also that there is a slapd
process/thread (I assume thread) marked as <defunct> in the ps listing.
Maybe a thread has died and the parent is still waiting for it (forever in
this case ?).  Has anyone seen this before?


Thanks.


James