One of my domUs is crashing.  The problem is in coda, not xen, which
is why I'm using an expendable domU.


I see from previous posts that this is a known problem.  Am I doing
something wrong, or is this a bug?  Is there an open PR?
I'm guessing there needs to be a "programmed IO" path to call the
hypervisor with the dump, and that isn't written.


_open+0x49: movl    0x10(%eax),%eax
db> sync
syncing disks... done
unmounting file systems...
unmounting /coda (CODA)...unmount of /coda failed with error 16

unmounting /kern (kernfs)...
unmounting /tmp (mfs:190)...
unmounting /usr (/dev/xbd0e)...
unmounting /var (/dev/xbd0f)...
unmounting / (/dev/xbd0a)... done
WARNING: some file systems would not unmount

dumping to dev 142,1 offset 9063
dump device bad

inexact gdt 1 ~ > df
Filesystem  1024-blocks      Used     Avail Capacity  Mounted on
/dev/xbd0a       127023     31282     89390    25%    /
/dev/xbd0f       127023      5806    114866     4%    /var
/dev/xbd0e      3613038   1745314   1687074    50%    /usr
mfs:190           63471         2     60296     0%    /tmp
kernfs                1         1         0   100%    /kern
CODA            4000000         0   4000000     0%    /coda
inexact gdt 2 ~ > pstat -s
Device      1024-blocks     Used    Avail Capacity  Priority
/dev/xbd0b       262144        0   262144     0%    0
inexact gdt 3 ~ >