On Thu, Oct 14, 1999 at 09:38:33PM +0200, jkunz@unixag-kl.fh-kl.de wrote:
> Hi.
> 
> I made a update from 1.4 to 1.4.1 last week. (I comiled the 1.4.1 kernel
> and after a reboot I compiled the userland) Since this I get the
> folowing messages in syslog:
> Oct 14 00:31:46 MissSophie /netbsd: WARNING: mclpool limit reached; increase NMB
> CLUSTERS
> 
> Oct 14 04:20:00 MissSophie /netbsd: Data modified on freelist: word 0 of object 
> 0xf04c7c60 size 4 previous type ifaddr (0xf05f4270 != 0xdeadbeef)
> 
> And I had a kernel panic:
> Oct 13 04:08:40 MissSophie savecore: reboot after panic: pfil_list_remove
> This is somthing new to me. I'm using NetBSD for nearly two years and I
> never had a kernel panic. (...on my i386, sparc, hp300, ... boxes. I had
> several on my VAXstation, but that is an other story.)
> 
> What does this mean? How to solve this?

The problem is that the kernel ran out of resources for network buffers
(the panic may be related but in any cases it's, wrong :)
The problem may be that your network driver pre-allocates too much resources.
A workaround has been commited but is not in 1.4.1. You may want to upgrade
your kernel with the sources from the release branch (available either from
sup or anoncvs, I'm pretty sure all the infos are on the www site).

--
Manuel Bouyer, LIP6, Universite Paris VI.           Manuel.Bouyer@lip6.fr
--