Subject: Problems with raidframe
To: None <tech-kern@netbsd.org>
From: M Graff <explorer@flame.org>
List: tech-kern
Date: 09/24/1999 16:36:59
I have a 25G raidframe array, and after upgrading to the latest code
on the netbsd-1-4 branch, I got this lock up:

(system unresponsive)

[BREAK]Stopped at      _Debugger+0x4:  leave
db> ps
 PID             PPID       PGRP        UID S   FLAGS          COMMAND    WAIT
 6682             227        227          0 3     0x4            inetd   inode
 6681             227        227          0 3     0x4            inetd   inode
 6680             227        227          0 3     0x4            inetd   inode
 6679             227        227          0 3     0x4            inetd   inode
 6678             227        227          0 3     0x4            inetd   inode
 6677             227        227          0 3     0x4            inetd   inode
 6676             227        227          0 3     0x4            inetd   inode
 6675             227        227          0 3     0x4            inetd   inode
 6674             227        227          0 3     0x4            inetd   inode
 6673             227        227          0 3     0x4            inetd   inode
 6672             227        227          0 3     0x4            inetd   inode
 6671             227        227          0 3     0x4            inetd   inode
 6670             227        227          0 3     0x4            inetd   inode
 6669             227        227          0 3     0x4            inetd   inode
 6668             227        227          0 3     0x4            inetd   inode
 6667             227        227          0 3     0x4            inetd   inode
 6666             227        227          0 3     0x4            inetd   inode
 6665             227        227          0 3     0x4            inetd   inode
 6664             227        227          0 3     0x4            inetd   inode
 6663             227        227          0 3     0x4            inetd   inode
 6662             227        227          0 3     0x4            inetd   inode
...

My guess is that something is holding a lock when it shouldn't, and
the bug is brought to the surface in the raidframe code.

My other machine, one without raidframe, tends to lock up in a simular
way, but everything is waiting on a block rather than an inode.

--Michael