Re: fsck seg fault failure on vmware -i386?

[yancm%sdf.lonestar.org@localhost]

> On Mon, Jan 25, 2010 at 12:07:58AM +0000, Eduardo Horvath wrote:
>  > You shouldn't need to do anything with the kernel, just gdb running on
>  > fsck_ffs.  You probably need to recompile fsck_ffs with symbols.
>
> This.
>
> To do this, do 'make cleandir && make dependall DBG=-g' in
> src/sbin/fsck_ffs on a different machine (so as not to touch the
> broken volume) then run it.
>
> Running it in gdb from single-user mode may not be feasible, in which
> case the best approach may be to mount a ram disk, cd there, and let
> it dump core, then copy the core file out.
>
>  > OTOH this is probably a lost case.  While fsck_ffs should not be
>  > generating segmentation faults, even if it were fixed it probably
>  > wouldn't be able to repair your filesystem.
>
> That is probably true; one can easily get unfixable volumes from e.g.
> mounting with -o async.
>
> Oh, another important issue: was it running wapbl, softupdates, or
> vanilla ffs?

The virtual machine that crashed was pretty close to Generic...
so vanilla ffs.

Since it's the /usr partition, and this was my only --current VM,
I'm following this tack as Plan A:

1) fortunately, /usr/src is on a separate partition and is likely OK.
2) I seem to be able to mount /usr itself read-only.
   (I believe the damage is isolated in the /usr/obj subdirectory...
   one of those theory things...)
3) Create a new disk with a new (empty) /usr partition
4) mount the old /usr on a temporary mount point read-only and
   copy out all I can onto the new /usr, avoiding /usr/obj...
5) Fix up the fstab to use new partition scheme and...
6) hopefully boot back into multi-user.
7) Then try to debug the fsck dumps per directions provided...

Plan B is to build a new VM off of an install CD iso I created from
my "production" 5_Stable box. Then mount my --current /usr/src on
it and bring up to --current.

Plan A seems more pleasing aesthetically and possibly quicker.

I'll post a follow up when I have something to report...