"Boris B.Samorodov" wrote: > > >Submitter-Id: current-users > >Originator: Boris B. Samorodov > >Organization: InPharmTech Co. > >Confidential: no > >Synopsis: Fatal Trap 12 in kernel module uipc_mbuf2.c > >Severity: serious > >Priority: medium > >Category: kern > >Class: sw-bug > >Release: FreeBSD 5.3-BETA2 i386 > >Environment: > System: FreeBSD gw.ipt.ru 5.3-BETA2 FreeBSD 5.3-BETA2 #6: Sat Sep 4 17:47:55 MSD 2004 bsam_at_gw.ipt.ru:/usr/obj/usr/src/sys/GW i386 > > >Description: > Fatal Trap 12, pointer: 0xc053a1f0. > > # addr2line -e /root/kernel/kernel.debug 0xc053a1f0 > ----- > ../../../kern/uipc_mbuf2.c:389 > ----- > # gdb6 -k /root/kernel/kernel.debug vmcore.1 > ----- > GNU gdb 20040803 [GDB v6.x for FreeBSD] > Copyright 2004 Free Software Foundation, Inc. > GDB is free software, covered by the GNU General Public License, and you are > welcome to change it and/or distribute copies of it under certain conditions. > Type "show copying" to see the conditions. > There is absolutely no warranty for GDB. Type "show warranty" for details. > This GDB was configured as "i386-portbld-freebsd5.3"...can not access 0xc070b588, invalid address (c070b588) > can not access 0xc070b588, invalid address (c070b588) > > panic messages: > --- > dmesg: kvm_read: invalid address (c075cc54) > can not access 0xc070b6c0, invalid address (c070b6c0) > can not access 0xc070b6c0, invalid address (c070b6c0) > --- > #0 0x00000000 in ?? () > 0x00000000 in ?? () > (kgdb) list *m_tag_locate+0x40 > 0xc053a1f0 is in m_tag_locate (../../../kern/uipc_mbuf2.c:389). > 384 if (t == NULL) > 385 p = SLIST_FIRST(&m->m_pkthdr.tags); > 386 else > 387 p = SLIST_NEXT(t, m_tag_link); > 388 while (p != NULL) { > 389 if (p->m_tag_cookie == cookie && p->m_tag_id == type) > 390 return p; > 391 p = SLIST_NEXT(p, m_tag_link); > 392 } > 393 return NULL; > (kgdb) q > ----- > >How-To-Repeat: > The host have to upstream interfaces to ISP, three internal > networks with two interfaces (two internal networks are > aliased in one inteface). The default route is to the first > provider. We use policy routing to route our third network > to second provider. $inet1 and $inet2 are aliases on one > interface. > > After adding ipfw rule: > # ipfw add fwd $provider2 ip from $inet3 to { not $inet1 or not $inet2 } > gateway paniced. > > I know, that in fact the rule has to be "not $inet1,$inet2". > But logically the first rule has the value "all". OS must not > trap. Does it panic 'reliablily' in the same place or is it more of a random occurence? I'm not long in the office tomorrow (my test machine is there) but I'll try to reproduce it within the next few days. -- AndreReceived on Sun Sep 05 2004 - 20:59:04 UTC
This archive was generated by hypermail 2.4.0 : Wed May 19 2021 - 11:38:10 UTC