[issue1077] Kernel lockup with SMP on -HEAD

[Vladimir <sinknull@xxxxxxxxxxxxxxxxxxxxxxx>]

Vladimir <vladimir.n.solovyov@gmail.com> added the comment:

Good day.
As I have grabbed russiane39's machine at my placement, I have performed some
diagnostics.

First of all, HDD connected via SATA in IDE mode ('ad' device) in this box.

Second, I have compiled the latest kernel (cvsuped just some hours ago) with SMP
and still receive the same problem, but occasionally I get lock-up in the
completely different place (before lo0 initialization), sadly just once...

Third thing I should notice is that after lock-up and pressing Ctrl+Alt+Del I
could see some shutdown actions performed by init process:
"
lo0: flags=8049<UP,LOOPBACK,RUNNING,MULTICAST> mtu 16384
	inet 127.0.0.1 netmask 0xff000000 
inetd not running? (check /var/run/inetd.pid).
Shutting down daemon processes:.
cron not running? (check /var/run/cron.pid).
Shutting down local daemons:.
.
" (and continuation of the lock-up).

These are booting logs.
1. Odd one lock-up: http://pastebin.ca/1077610
2. Common lock-up (with 'ps' and 'trace'): http://pastebin.ca/1077611
3. Working non-SMP kernel: http://pastebin.ca/1077609

Now I am trying to get functions call backtrace using the kgdb through the
serial port, but I couldn't fight up the problem with connection speed (noticed
early by russiane39). I've found a possible solution in one manual which
recommended typing 'set baudrate 9600' but didn't specify where to type this
line. I've tried doing so in system shell (setting environment variable) and in
the gdb shell (it doesn't know such command at all) but with no results...

_____________________________________________________
DragonFly issue tracker <bugs@lists.dragonflybsd.org>
<https://bugs.dragonflybsd.org/issue1077>
_____________________________________________________