Looping SIGALARMs

Hi,

I've not posted to any MySQL list before so if this is the wrong place to
post this, please redirect me.

I'm running mysqld 5.022 on NetBSD 3.0_STABLE on i386 and I've it got
starting up on boot up from /etc/rc.local.

After firing up mysqld I've got a 10 second sleep and then I've got a few
..sql scripts that run and appear to complete successfully. The problem is
that mysqld ends up running at 100% processor usage after bootup. I removed
my 3 scripts from rc.local and after startup everything appears to be
normal. I've also tried running mysqld and the scripts manually as root
after bootup and everything runs normally. So it appears that the problem is
limited to when mysqld and the scripts and run from rc.local at startup.

The same scripts were running fine on an i386 platform running mysqld 4.1.9
on NetBSD 2.0_STABLE.

I recompiled mysqld 5.022 --with-debug and ran a debug trace. Below is a
snippet of the trace:

T@2 : | >hash_free

T@2 : | | enter: hash: 0x852e76cd

T@2 : | | >my_free

T@2 : | | | my: ptr: 0x8534400

T@2 : | |
T@2 : |
T@2 : | >hash_init

T@2 : | | enter: hash: 0x852e76c size: 128

T@2 : | | >init_dynamic_array

T@2 : | | | >my_malloc

T@2 : | | | | my: size: 1024 my_flags: 16

T@2 : | | | | exit: ptr: 0x8534400

T@2 : | | |
T@2 : | |
T@2 : |
T@2 : | >process_alarm

T@2 : | | info: sig: 14 active alarms: 0

T@2 : |
T@2 : | >process_alarm

T@2 : | | info: sig: 14 active alarms: 0

T@2 : |
T@2 : | >process_alarm

T@2 : | | info: sig: 14 active alarms: 0

T@2 : |
T@2 : | >process_alarm

T@2 : | | info: sig: 14 active alarms: 0

T@2 : |
T@2 : | >process_alarm

T@2 : | | info: sig: 14 active alarms: 0

T@2 : |
..

..

..

Basically the process_alarm gets called indefinitely and it appears to be in
a really tight loop, note that at this point all scripts have completed and
there are no clients interacting with mysqld.

I've tried increasing the delay between starting up mysqld and my scripts
but it makes no difference. I've tried running only 1 of the 3 scripts after
starting mysqld but it also makes no difference. My guess would be that any
mysql client access in the boot-up rc.local script results in the above
problem.

Has anyone encountered this before?

Please CC me in any response as I'm not registered on this list.



Thanks in advance,

Brad


--
MySQL Bugs Mailing List
For list archives: http://lists.mysql.com/bugs
To unsubscribe: http://lists.mysql.com/bugs?unsub=gcdmb-bugs@m.gmane.org