Date: Thu, 14 Sep 2000 11:43:56 +0200 (SAST) From: Reinier Bezuidenhout <rbezuide@oskar.nanoteq.co.za> To: freebsd-current@freebsd.org Subject: FW: (KAME-snap 3318) Panic on current (12 Sept) Message-ID: <XFMail.000914114356.rbezuide@oskar.nanoteq.co.za>
next in thread | raw e-mail | index | archive | help
Hi ...
I'm running a current machine of 12 Sept although this problem
also occured on a current of a few days earlier ...
This only happens when using the IPv6 IPSec code during the day,
it is readily reproduceable.
If during the day I load the racoon daemon and load keys and
establish a IPSec tunnel connection everything works fine till
2:00 am when the daily script runs OR if I run the daily script
by hand ... I generated the following dump and backtrace ...
It seems to crash in a makedev routine using FOREACH list macro's.
The problem doesn't seem to be with the list or the makedev function
in kern_conf.c. It seems to me that something in the kernel
corrupts the static list dev_hash when using the IPSec code.
Summary - when ising IPSec ... machine panics during daily
script execution.
If you need more info from the dumps, just ask !
Reinier
----- gdb bt follows qith a few prints -----
(kgdb)
#5 0xc0283d53 in trap (frame={tf_fs = 16, tf_es = -65520, tf_ds = -904265712,
tf_edi = 196611, tf_esi = 197379, tf_ebp = -904254412,
tf_isp = -904254444, tf_ebx = 5, tf_edx = 2, tf_ecx = -1057628672,
tf_eax = 197374, tf_trapno = 12, tf_err = 0, tf_eip = -1072335336,
tf_cs = 8, tf_eflags = 66050, tf_esp = 197379, tf_ss = -905316928})
at ../../i386/i386/trap.c:436
436 (void) trap_pfault(&frame, FALSE, eva);
(kgdb)
#6 0xc0157618 in makedev (x=3, y=196611) at ../../kern/kern_conf.c:232
232 LIST_FOREACH(si, &dev_hash[hash], si_hash) {
(kgdb) li
227
228 if (x == umajor(NOUDEV) && y == uminor(NOUDEV))
229 Debugger("makedev of NOUDEV");
230 udev = (x << 8) | y;
231 hash = udev % DEVT_HASH;
232 LIST_FOREACH(si, &dev_hash[hash], si_hash) {
233 if (si->si_udev == udev)
234 return (si);
235 }
236 si = allocdev();
(kgdb) p hash
$1 = 5
(kgdb) p dev_hash[hash]
$2 = {lh_first = 0x2}
(kgdb) p dev_hash[1]
$3 = {lh_first = 0xc0fb7980}
(kgdb) p dev_hash[2]
$4 = {lh_first = 0xc0fb7900}
(kgdb) p dev_hash[3]
$5 = {lh_first = 0xc0fb9d80}
(kgdb) p dev_hash[4]
$6 = {lh_first = 0x200} <----- OOPS !!
(kgdb) p dev_hash[5]
$7 = {lh_first = 0x2} <----- OOPS - the one we use ...
(kgdb) p dev_hash[6]
$8 = {lh_first = 0xc0fb5980}
(kgdb)
---------- panic message ------
Fatal trap 12: page fault while in kernel mode
fault virtual address = 0x1e
fault code = supervisor read, page not present
instruction pointer = 0x8:0xc0157618
stack pointer = 0x10:0xca1a2c28
frame pointer = 0x10:0xca1a2c34
code segment = base 0x0, limit 0xfffff, type 0x1b
= DPL 0, pres 1, def32 1, gran 1
processor eflags = interrupt enabled, resume, IOPL = 0
current process = 1910 (find)
trap number = 12
panic: page fault
syncing disks... 31 31 8 8 3 3
done
Uptime: 11h37m7s
###################################################################
# #
# R.N. Bezuidenhout NetSeq Firewall #
# rbezuide@oskar.nanoteq.co.za http://www.nanoteq.co.za #
# #
###################################################################
----------------------------------
Date: 14-Sep-00
Time: 11:43:22
This message was sent by XFMail
----------------------------------
To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe freebsd-current" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?XFMail.000914114356.rbezuide>