freeze and crashes

[Alexander Sack]

Hi Alain:

Without more details its hard to understand where your hang or freeze
is coming from.  What I DO suggest is that you build a debug kernel
and/or minimally enable the kernel debugger (DDB).  Then when the box
is frozen you can get into the debugger (CTRL-ALT-ESC), type "bt", and
post it on the list to give everybody a chance to understand what is
hanging ("show msgbuf" as well actually).

Here is some good documentation on how to build kernels (and debug
ones) if you don't already know:

http://www.freebsd.org/doc/en_US.ISO8859-1/books/handbook/kernelconfig-building.html

Generally speaking I would enable in my KERNCONF file (GENERIC, etc.
under src/sys/<arch>/conf/)

# Turn off debugging symbols
makeoptions	DEBUG=-g

# Debugger support
options		KDB
options		DDB
options		KDB_TRACE

If you want to go further, for example you believe this is a locking
issue then you can go whole hog and do:

# Heavy duty debugger options
options	DEBUG_LOCKS
options	INVARIANTS
options	INVARIANT_SUPPORT
options	WITNESS
options	WITNESS_SKIPSPIN

Be warned that the kernel.debug KO (kernel object) will run a lot
slower than a stock non-debug kernel (the extra metadata the kernel is
keeping track off etc.) so there are instances where running a debug
kernels could make a timing bug disappear or appear more/less
frequently.  In any event, if the problem is easily reproducible, I
whole-heartily recommend you minimally enable KDB/DDB support so you
can dump things on the appropriate list AS WELL AS build a complete
debug kernel that may catch the issue (you will see console messages
including lock order reversals and other nastiness).

Here is some good documentation about kernel debugging:

http://www.freebsd.org/doc/en_US.ISO8859-1/books/developers-handbook/kerneldebug.html

Let us know how things go,

-aps

PS Make sure that new disk is good!
On Fri, Apr 25, 2008 at 9:01 AM, Alain G. Fabry <alainfabry at belgacom.net> wrote:
> Have a issue here. I've been running FreeBSD 6.3 release for about 6 months now without problem. Then last week I purchase another external HD and installed release 7.0.
>  Everything went great for about 3 weeks. Then suddenly my 7.0 started to crash (kde freezing, sudden reboots, etc). The first time happened when I was performing a portsnap on a FreeBSD Qemu session. It came up to the point where this 7.0 installation wasn't workable anymore.
>
>  Since I need to work to earn money as the rest of us, I fell back to my 6.3 release external HD -> copied over my data files and now I'm starting to get the same symthoms.
>
>  Input/output errors, reboot when performing a simple 'man something'. -> could this be a virus or something?
>
>  I'm not sure how or what I can do to find out what could be the cause of this problem?
>
>  Is there anywhere I can find out more about debugging such strange behavior. (everything works fine and then suddenly it freezes permanently or just reboots on its own)
>
>
>  Thanks,
>
>  PS: I'm about to reinstall the 7.0 release and hope that that will work normal again.
>
>  Alain
>  _______________________________________________
>  freebsd-questions at freebsd.org mailing list
>  http://lists.freebsd.org/mailman/listinfo/freebsd-questions
>  To unsubscribe, send any mail to "freebsd-questions-unsubscribe at freebsd.org"
>