Re(2): panic: sorele

Doug White wrote:

> On Thu, 23 Sep 2004, Evren Yurtesen wrote:
> 
> 
>>Does anybody know what is panic: sorele? :)
> 
> 
> This panic is triggered if a socket's reference count is reduced below 0.
> This usually indicates corruption of the socket structure.  If you got a
> crashdump it would be interesting to inspect the structure and see if its
> filled with 0xdeadc0de, which indicates reuse of a freed socket structure.
> 
> Can you do anything to reproduce this?


Actually I realized something else.

I already had the dumpdir variable set in rc.conf

dumpdev="/dev/ad0s1b"

and from fstab, we see that its the swap

# Device                Mountpoint      FStype  Options         Dump 
Pass#
/dev/ad0s1b             none            swap    sw              0       0

But I didnt have /var/crash. Although I dont think this is the problem 
but there is something weird in messages file. The message that savecore 
couldnt find any dumps...

messages:Sep 23 15:27:31 proxy squid[539]: NETDB state saved; 680 
entries, 25 msec
messages:Sep 23 16:15:07 proxy syslogd: kernel boot file is 
/boot/kernel/kernel
messages:Sep 23 16:15:07 proxy kernel: panic: sorele
messages:Sep 23 16:15:07 proxy kernel: cpuid = 1
messages:Sep 23 16:15:07 proxy kernel: boot() called on cpu#1
messages:Sep 23 16:15:07 proxy kernel: Uptime: 5d19h48m4s
...
...
...
messages:Sep 23 16:15:07 proxy savecore: no dumps found

So I dont know why savecore says that there were no dumps. Am I missing 
something? I know fixed the missing dumpdir directive in rc.conf also. 
But anyway, lets see what will happen next time when the machine crash.

The funnier thing is that it seems like the panic message came when the 
machine was booting already, and it still continued booting, isnt this a 
bit weird? Shouldnt it be like at about 15:27 when the machine actually 
crashed?


Evren