Date: Sat, 3 Jun 2006 20:14:55 -0400 (EDT) From: Brian Tao <taob@risc.org> To: FREEBSD-STABLE <freebsd-stable@freebsd.org> Subject: 6.1 kernel unable to find /dev ? Message-ID: <20060603195754.Q15261-100000@as2.dm.egate.net>
next in thread | raw e-mail | index | archive | help
I had a very stable 6.1-R amd64 server (once I swapped out some
bad RAM, that is) that needed a couple more hard drives installed.
There were some problems with the upgrade (device renumbering woes,
basically... topic of another thread), and it had to be rolled back.
Upon rolling back, the previously-good kernel would no longer
complete the boot after the device probe. I saw two types of panics
on the serial console:
| Trying to mount root from ufs:/dev/ad4s1a
| Lookup of /dev for devfs, error: 20
| exec /sbin/init: error 20
| exec /sbin/oinit: error 20
| exec /sbin/init.bak: error 20
| exec /rescue/init: error 20
| exec /stand/sysinstall: error 20
| init: not found in path
| /sbin/init:/sbin/oinit:/sbin/init.bak:/rescue/init:/stand/sysinstall
| panic: no init
| Uptime: 8s
| Cannot dump. No dump device defined.
| Automatic reboot in 15 seconds - press a key on the console to abort
| --> Press a key on the console to reboot,
| --> or switch off the system now.
... and:
| Trying to mount root from ufs:/dev/ad4s1a
| pid 47 (sh), uid 0: exited on signal 11
| TPTE at 0xffff8000040028e0 IS ZERO @ VA 80051c000
| panic: bad pte
| Uptime: 8s
| Cannot dump. No dump device defined.
| Automatic reboot in 15 seconds - press a key on the console to abort
| --> Press a key on the console to reboot,
| --> or switch off the system now.
The first one is suggesting that /dev does not exist (or is not a
directory)... I'm thinking this means that devfs is somehow
unavailable, but I did not think it is even possible to disable devfs
via the kernel config file these days.
The second one leaves me clueless... I have not been able to find
any useful information on that panic during boot. Granted, I've only
see the "bad pte" panic twice... all other reboot attempts result in
the first type of problem.
Fortunately, I did happen to keep an old 6.0-RELEASE-p6 kernel
around (Apr 15 2006 build). That kernel boots fine, using the same
filesystem as newer kernels on that drive. I am up-to-date with the
RELENG_6_1 tag. Should I perhaps to a make installkernel installworld
before rebooting? The installed binaries on the server are from an
early 6.1-RELEASE (which *was* successfully booted by this server). I
am running into a few minor but surmountable problems because of the
older kernel version, but I obviously would like to get my world and
kernel back in sync ASAP.
Any comments on this?
--
Brian Tao (BT300, taob@risc.org)
"Though this be madness, yet there is method in't"
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20060603195754.Q15261-100000>