Current-Users archive

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index][Old Index]

Re: NetBSD 3.x, IPv6 and IPF

On Mon, 18 Aug 2008 14:30:00 +0200, Dominik Bia³y 
> On Mon, 18 Aug 2008 11:18:31 +0200, Dominik Bia³y 
> <>
> wrote:
>> On Sun, 17 Aug 2008 17:47:38 +0000 (UTC), Michael van Elst 
>> <>
>> wrote:
>>> (Dominik =?iso-8859-1?Q?Bia=B3y?=) 
>>> writes:
>>>>>>fr_checkicmp6matchingstate(c0911980,0,28,c0911980,8) at 
>>>>> Sounds like an old bug, which was 'fixed' by an ipf upgrade on the
>>>>> way to netbsd-4. I doubt that this will be backported to netbsd-3.
>>>>I have also hit this bug. Is there any workaround? I see the bug is in
>>>>fragmented ipv6 packets handling (am I right?). Would not using
>>>>``keep frags'' be sufficient?
>>> I don't think so. One step of the fix was sys/dist/ipf/netinet/fil.c 1.38
>>> but since multiple ipf upgrades happened since netbsd-3 I doubt that
>>> this will solve your problem.
>> I've upgraded to NetBSD 4.0_STABLE from 20080813 and got another panic
>> today. Alas, no backtrace. Trying gdb netbsd.8 netbsd.8.core, but gdb
>> says netbsd.8.core is not a corefile. How do I get the backtrace?
> Never mind my last question. I found the answer, but sadly gdb only
> gives this:
[ snip ]
I got the panic once again. This time I had ddb.onpanic=1. I got
backtrace from ddb -- it was almost the same backtrace which was
reported by Martti Kuparinen at the beginning of this thread. This
is on NetBSD 4.0_STABLE, so the problem doesn't touch only netbsd-3.
I think it's really worth fixing.

Sorry not posting backtrace, but dmesg on the core file showed nothing
and gdb shows nothing useful, too.

I hope you guys will fix it, so we have working NetBSD + ipf + ipv6
again ;-)

Thank you
        Dominik Bialy

Serwer darmowych kont shell:
tel. +48 698 182 161; email:

Home | Main Index | Thread Index | Old Index