On Tue, Apr 26, 2005 at 07:59:34AM +0300, Martti Kuparinen wrote:
> Manuel Bouyer wrote:
> 
> >Do you have the kernel core dump ? If so, you should have it in
> >dmesg -M netbsd.core
> 
> panic: m_copym: m == 0, len 1 [!COPYALL]
> Begin traceback...
> m_copym0(0,484,1,1,0) at netbsd:m_copym0+0x459
> m_copym(c1e70200,484,4da,1,0) at netbsd:m_copym+0x19
> tcp_build_datapkt(c24e8190,c1d7b5f4,b68,4da,28) at 
> netbsd:tcp_build_datapkt+0x1f
> 0
> tcp_output(c24e8190,c0936cc8,c1d7b5f4,c0936de0,c24e8190) at 
> netbsd:tcp_output+0x
> 1029
> tcp_sack_newack(c24e8190,cb53e844,c0936de0,cb53e858,0) at 
> netbsd:tcp_sack_newack
> +0xf6
> tcp_input(c1d67300,14,6,1a,c0936e50) at netbsd:tcp_input+0x17bc
> ip_input(c1d67300,0,0,246,0) at netbsd:ip_input+0x587
> ipintr(ba40010,30,a67a0010,10,c0933000) at netbsd:ipintr+0x76
> DDB lost frame for netbsd:Xsoftnet+0x41, trying 0xc0936e80
> Xsoftnet() at netbsd:Xsoftnet+0x41

This can probably be worked around by disabling sack.

-- 
Manuel Bouyer <bouyer@antioche.eu.org>
     NetBSD: 26 ans d'experience feront toujours la difference
--