=== Terminal output ===
# mount -o recovery /dev/md0 /mnt/root
mount: wrong fs type, bad option, bad superblock on /dev/md0,
       missing codepage or helper program, or other error
       In some cases useful info is found in syslog - try
       dmesg | tail  or so

# dmesg | tail
[58290.395538] device label root devid 1 transid 250160 /dev/md0
[58290.395891] btrfs: enabling auto recovery
[58290.395895] btrfs: disk space caching is enabled
[58290.396388] btrfs bad tree block start 0 139264
[58290.396401] btrfs: failed to read chunk root on md0
[58290.396660] btrfs: open_ctree failed

# btrfsck /dev/md0
Check tree block failed, want=139264, have=0
Check tree block failed, want=139264, have=0
Check tree block failed, want=139264, have=0
read block failed check_tree_block
Couldn't read chunk root
Segmentation fault

=======================

I'm having some problems with my btrfs partition.  It showed up around
the same time a disk got ejected from a RAID5 mdadm array, so the two
are probably related.  The array seems to be working properly now (or
thinks it is), having rebuilt itself and completed a full integrity
check  However, the array lists 1196 events and found 180156080
mismatches on the integrity check, so it may be serious issues with my
old disks from a Sun array.  Even if my data is completely ruined
(e.g. `dd if=/dev/urandom of=/dev/md0`), btrfsck should not segfault.
Running `btrfs-show-super -a /dev/md0` appears to work fine, except
"chunk_root_generation" is 12 lower than "generation" and
"dev_item.generation" is 0, but I'm not sure how significant that is.

I'm running 64-bit Linux Mint 14 KDE (Ubuntu 12.10 based), upgraded to
the 3.9-rc1 kernel and latest git pull of the official btrfs-progs as
of March 13, 2013.  It uses gcc 4.7.2 (Ubuntu/Linaro 4.7.2-2ubuntu1).
I have attached my backtrace (with commands I used to generate it).  I
could not find a debug version of libuuid to show where that is
crashing.

P.S.
Is there any hope for my data?  I have at least 1 small subvolume I'd
really like to recover.  Most of the rest is either backed up, copies
of media, or stuff I can download.  If it's all gone, I need to move
on with rebuilding, so I can get back to my normal work environment.

P.P.S.
Sorry if anyone is getting this twice.  The first message (with a core
dump included) did not get a reply, nor was it on the mailing list
archive.

Attachment: backtrace
Description: Binary data

Reply via email to