Subject: Re: Help: Kernel panic
To: arix chen <arix@la29.com>
From: Chris Tribo <ctribo@dtcc.edu>
List: port-macppc
Date: 03/31/2005 16:20:35
Most of the code surrounding this issue was changed around the 2.0 
release, so it didn't make it in. I have -current kernels you can try 
and see if the issue still occurs:

http://user.dtcc.edu/~ctribo/macppc/netbsd-GENERIC.gz
MD5(netbsd-GENERIC.gz)= 7263fe515ba78aa6d3683c31b7e7518f
RIPEMD160(netbsd-GENERIC.gz)= 622dda1764ab6f053aec3fb9c6c891701b3df401

http://user.dtcc.edu/~ctribo/macppc/netbsd-GENERIC.MP.gz
MD5(netbsd-GENERIC.MP.gz)= e5f9fec0d31e0be09c7750c336dda670
RIPEMD160(netbsd-GENERIC.MP.gz)= 
4e291bd6240ed34881ed32a9c5855731d727e343


Hopefully this will be fixed for the upcoming 2.0.1 release, but I 
don't know the status of the pull up requests.

On Mar 31, 2005, at 4:52 AM, arix chen wrote:

> Thanks your reply. The following is the output of bt at the db> prompt 
> after the panic.
>
> 0x00494e50: at panic+0x19c
> 0x00494ee0: at __assert+0x28
> 0x00494ef0: at softintr__run+0xcc
> 0x00494f10: at do_pending_int+0x1d4
> 0x00494f50: at splx+0x40
> 0x00494f60: at ext_intr+0x174
> 0x00494fa0: at trapstart+0x880
> 0x0049ce30: at Idle+0x14
> 0x0049ce40: at mi_swich+0x198
> 0x0049ce80: at ltsleep+0x41c
> 0x0049cec0: at uvm_scheduler+0xb0
> 0x0049cef0: at main+0x6ec
> 0x0049cf30: at 0x100090
>
> Is it useful?
>
> p.s Sorry everybody, I just enable the Return Receipt in the last 
> mail.  Now  I've disabled it.
>
> Chris Tribo wrote:
>
>> On Tue, 29 Mar 2005, arix chen wrote:
>>
>>
>>> I have a PPC 8600 running NetBSd 2.0, but sometime the kernel will 
>>> panic and has the following message. It there anyone can help me to 
>>> find out the problem? Please tell me what information I should post.
>>>
>>> panic: kernel diagnostic assertion "si->si_refs > 0" failed: file
>>> "../../../../arch/powerpc/powerpc/softintr.c", line 116
>>>
>>
>> Your best bet at this point is to try a current kernel and see if it 
>> still
>> occurs, or wait for the next release; unless you feel like testing
>> bleeding edge patches to enable crash dumps and fix interrupt 
>> handling.
>>
>> A good start for what you have would be to type bt at the db> prompt 
>> after
>> the panic and send us that output
>>
>>
>
> -- 
> arix chen
> +852-93003194
> mailto:arix@la29.com
> -----------------------------------------------------
> Dell Latitude C840 & C/PORT II | P4-M 2.0 | 768MB RAM
> 60GB HD (WinXP SP2@Main) | 20GB HD(FreeBSD@Module Bay)
> SEC 15" UXGA LCD | 1600x1200 | Nvidia Geforce 4 440 Go 32MB.
> FreeBSD 5.3 | Xorg 6.8.1 | GNOME 2.8 | Nvidia native driver 1.0-6113