Strange vkernel crash

Dave Hayes dave at jetcafe.org
Thu Apr 5 02:10:36 PDT 2007


Matthew Dillon <dillon at apollo.backplane.com> writes:
>     That was supposed to be appended to the standard /etc/ttys, not
>     replace it!

Yes, well I copied exactly what was in the man page. :) Perhaps the
man page should be changed to reflect this new information? 

>     However, when I do that on my test box, it still seems to work just
>     fine, so that doesn't appear to be the cause.

Both 1.8.1_Slip and HEAD exhibit the same crash now. I'm at a loss
here, mostly because I cannot "see" what's wrong.

Here's the script I use to generate the vkernel root image:

#!/bin/sh
#######
# mkvkernel - Make a vkernel filesystem image per man page
#
# 1) cd to your src directory
# 2) use "mkvkernel <identifier>"
# 3) You should be dropped into a directory with an already built kernel
#

ID=$1
if [ "x$ID" = "x" ]; then
        ID=`pwd | sed -e 's/\//_/g;'`
        ID="${ID}.$$"
fi

VKIMG=/var/vkernel/rootimg.$ID

echo "Making vkernel at $VKIMG"

rm -rf $VKIMG >/dev/null 2>&1
dd if=/dev/zero of=$VKIMG bs=1m count=2048
vnconfig -c -s labels vn0 $VKIMG
disklabel -r -w vn0 auto
disklabel -R vn0 /home/dave/dfltlabel
newfs /dev/vn0a
mount /dev/vn0a /mnt
trap "umount /mnt ; vnconfig -u vn0; echo 'cleaned up'" 2
make installworld DESTDIR=/mnt
cd etc
make distribution DESTDIR=/mnt
echo '/dev/vkd0a / ufs rw 1 1' >/mnt/etc/fstab
echo 'console "/usr/libexec/getty Pc" cons25 on secure' >>/mnt/etc/ttys
umount /mnt
vnconfig -u vn0
cd ../sys/config
config VKERNEL
cd ../compile/VKERNEL
make depend && make

echo "Vkernel make done ... to test:"
echo "   cd `pwd`"
echo "   ./kernel.debug -m 64m -r $VKIMG "
######### End

In /home/dave/dfltlabel we have:

# /dev/vn0c:
type: unknown
disk: amnesiac
label: 
flags:
bytes/sector: 512
sectors/track: 32
tracks/cylinder: 64
sectors/cylinder: 2048
cylinders: 2048
sectors/unit: 4194304
rpm: 3600
interleave: 1
trackskew: 0
cylinderskew: 0
headswitch: 0           # milliseconds
track-to-track seek: 0  # milliseconds
drivedata: 0 

16 partitions:
#        size   offset    fstype   [fsize bsize bps/cpg]
  a:  4194304        0    4.2BSD        0     0         # (Cyl.    0 - 2047)
  c:  4194304        0    unused        0     0         # (Cyl.    0 - 2047)

If I missed something, let me know. This could just as easily be my
naivete as a real bug. 
------
Dave Hayes - Consultant - Altadena CA, USA - dave at jetcafe.org 
>>> The opinions expressed above are entirely my own <<<

Nasrudin was sitting talking with a friend as dusk
fell. "Light a candle", the man said, "because it is dark
now. There is one just by your left side."  "How can I tell
my right from my left in the dark, you fool?" came the
reply.







More information about the Kernel mailing list