On Jul 26, 2006, at 1:41 PM, Andrew Doran wrote:

> Hi Chris,
>
> On Wed, Jul 26, 2006 at 12:18:29PM -0400, Chris Tribo wrote:
>
>> In my quest to reproduce the previous panic I found a new one
>
> This should no longer happen with revision 1.15 of kern/kern_auth.c
>
> http://mail-index.netbsd.org/source-changes/2006/07/26/0034.html

Cool, that seems to fix it. BTW, kern_auth.c looks like it's missing  
an RCSID.

This is still the old code, but is this related to the panic I  
reported, or something different:

panic: spinlock_switchcheck: CPU 0 has 1 spin locks
db{0}> reboot 0x0
sched_lock lock held

lock heldlock: 0xc08d26b4, currently at: /usr/src/sys/kern/ 
subr_pool.c:907 on CPU 0
lask locked: /usr/src/sys/kern/kern_synch.c:885
las unlocked: /usr/src/sys/kern/kern_synch.c:1202
_prop_dictionary_keysym32_pool(c3724fdc,0,0,c3724fdc,0) at 0xc3c3c800
Bad frame pointer: 0xc905f600
db{0}> reboot 0x0
uhci_idone: ii=0xc3c0096c is done!
uhci_idone: ii=0xc3c0096c is done!
uhci_idone: ii=0xc3c0096c is done!
(endless repetition)