Subject: Re: Interesting current MP lockup
To: John Klos <john@ziaspace.com>
From: Chuck Silvers <chuq@chuq.com>
List: port-macppc
Date: 01/02/2005 16:04:45
hi,

I just checked in a fix for this, update powerpc/trap.c.
apparently pthreads never worked on MP powerpc.

-Chuck


On Sun, Dec 19, 2004 at 02:26:34PM -0800, John Klos wrote:
> Hi,
> 
> I just discovered that I can reliably crash a current MP kernel from the 
> 13th tree by trying to launch BIND 9.3.0 on this dual 200 MHz 604e system:
> 
> Dec 19 14:24:32 bunny named[383]: none:0: open: /etc/rndc.key: file not 
> found
> Dec 19 14:24:32 bunny named[383]: none:0: open: /etc/rndc.key: file not 
> found
> Dec 19 14:24:32 bunny named[383]: couldn't open pid file '/varpanic: 
> kernel diagnostic assertion "curcpu()->ci_data.cpu_biglock_count == 0" 
> failed: file "/usr/current/src/sys/kern/kern_lock.c", line 1480
> Stopped in pid 383.3 (named) at netbsd:cpu_Debugger+0x10:       lwz 
> r0, r1,
> 0
> x14
> db{1}> tr
> 0xd5933c60: at panic+0x19c
> 0xd5933cf0: at __assert+0x28
> 0xd5933d00: at _kernel_lock_acquire_count+0x114
> 0xd5933d30: at mi_switch+0x174
> 0xd5933d70: at ltsleep+0x560
> 0xd5933db0: at lwp_wait1+0xf0
> 0xd5933de0: at exit_lwps+0x1c8
> 0xd5933e20: at exit1+0x118
> 0xd5933eb0: at sys_exit+0x38
> 0xd5933ed0: at syscall_plain+0xc8
> 0xd5933f40: user SC trap #1 by 0x41cd1c3c: srr1=0xd032
>             r1=0x449ffcd0 cr=0x22042042 xer=0 ctr=0x41cd1c34