Re: kernel BUG at fs/btrfs/inode.c:149!

[whirm@gmx.com]

On Friday 13 May 2011 20:52:22 Josef Bacik wrote:
> On 05/13/2011 01:19 PM, whirm@gmx.com wrote:
> > On Thursday 05 May 2011 20:57:17 Josef Bacik wrote:
> > [..]
> > 
> >> It doesn't look like that bit had my debugging output.  Thanks,
> >> 
> >> Josef
> > 
> > Looks like the last message I sent didn't make to the list. Here's the
> > link to
> 
> > the debug log:
> So unfortunately I don't know how we ended up with duplicate entries in
> the free space cache, but I can make it so we discard the cache if this
> happens.  Please try the patch I just sent to the list
> 
> [PATCH] Btrfs: check for duplicate entries in the free space cache
> 
> this will make your fs able to be mounted at the very least.  I'll try
> and figure out how this sort of thing happens.  If you manage to make it
> happen on purpose let me know how you did it so I can figure out what
> I'm doing wrong.  Thanks,

I was able to mount it readonly and copy all the content from it without 
problems (with vanilla linus' tree). The only out of ordinary thing I did 
before the error coming up, was to defrag the filesystem (find ~/ -xdev -type 
f -print0 | xargs -0 sudo ./btrfs filesystem defragment -c) with btrfs-tools 
from the latest git. After this I think I rebooted once the same day and 
everything was working ok. The day after in the morning I powered up the 
laptop and I wasn't able to mount the home volume. 
I'm on a 64 bit Debian SID with custom kernel. Using btrfs on home and root 
and ext4 in 3 or 4 other FS, everything is on lvm in luks.

The only thing I think of I can try is to recreate the filesystem, copy the 
same data on it, defrag and try to umount/remount several times... (although 
the data will not be fragmented so we may not hit this error) What do you say?


Thanks,

-- 

Elric Milon
-- 

Elric Milon