Running mysqld-4.0.15 on a -current of about 07/15/03, causes 
the machine to come to a grinding halt.  It's unpingable and 
doesn't respond at the console.  One can however activate
DDB and force a reboot using cpu_reset.

A 'ps' from within ddb yields the following:

db{1}> ps
 PID           PPID     PGRP        UID S   FLAGS LWPS          COMMAND    WAIT
>26741        23941    23941       1015 2  0x4502    4           mysqld       *
 23941        28892    23941          0 2  0x4002    1               sh    wait


So it looks like there's something about the native threadss that mysqld
is tickling... This happens repeatably immediately after starting mysqld.

The environment is a dual PII-300, NetBSD-current of mid-july, and running
SMP...

If anyone wants to look at this, I'm willing to do a certain amount of
debuggery but it's a production machine so long periods of downtime 
or frequent reboots need to be scheduled.


---
Beheading servers since 1999 -- The PC Weasel! http://www.realweasel.com