Subject: kernel panic related to racoon/setkey
To: None <>
From: Jeff <>
List: current-users
Date: 12/13/2003 15:54:32
Lately I have been running into a panic that I've been able to
reproduce on two separate 1.6ZG installations.  The panic seems to
relate to setkey and/or racoon.  The scenarion is the following:

setkey has already been run, followed by starting racoon.  At this
point if I restart setkey (/etc/rc.d/ipsec) and then restart racoon
(/etc/rc.d/racoon) I get the following...

#2  0xc037a4e5 in panic (fmt=0xc069f945 "amap_wipeout: corrupt amap")
    at ../../../../kern/subr_prf.c:218
#3  0xc03e5a57 in amap_wipeout (amap=0xcb5b5914)
    at ../../../../uvm/uvm_amap.c:668
#4  0xc03ee30e in uvm_unmap_detach (first_entry=0xcb5cb4d0, flags=0)
    at ../../../../uvm/uvm_map.c:491
#5  0xc03f01d9 in uvmspace_free (vm=0xcb2f4c08)
    at ../../../../uvm/uvm_map.c:3590
#6  0xc035bff9 in reaper (arg=0xcb2f74c8) at 

NetBSD netbsd 1.6ZG NetBSD 1.6ZG (LAB) #0: Mon Dec  8 13:09:53 EST
2003  jeff@netbsd:/q/nbsd/src/sys/arch/i386/compile/LAB i386

$NetBSD: uvm_amap.c,v 1.52 2003/02/01 06:23:54 thorpej Exp $
$NetBSD: uvm_map.c,v 1.151 2003/11/13 02:44:02 chs Exp $

It should be noted that this panic does not occur *every* time, but if
I go through this restart scenario above ten times, I can guarantee a

If this should be opened as a PR, or if there is any further debug
information that would help diagnose this, please let me know.

Thank you,