-----BEGIN PGP SIGNED MESSAGE-----
Helo all,
On 18-Mar-02 Oleg Drokin tried to scribble about:
> Hello!
>
> On Fri, Mar 15, 2002 at 07:19:19PM +0100, Tels wrote:
>
>> > Ok. you need to use losetup, then (man losetup)
>> It worked (sort of ;)
> Great!
>> I found that losetup let's me do a reiserfsck, and that it will no
>> longer report "ignore --fix-fixable". However, it did *not* fix the
>> corruptions (log appended). They were present even after running it
>> multiply times with - -o and -x (or --fix-fixable).
> --fix-fixable won't fix all the problems.
> Some problems can only be fixed with rebuild tree.
The funny thing is that reiserfsck told me:
"Found 6 corruptions that can be fixed with --fix-fixable".
but it couldn't fix them - it needed rebuild :)
>> In the end I settled for a --rebuild-tree and it worked great! I could
>> delete the offending directory, and the only lost things (seem) to be
>> some unimportant konqueror cache files ;)
>
> Great news again ;)
>
>> (--rebuild-tree told me for about a dozend blocks that it could not read
>> them. Does this mean these blocks are bad blocks on my HD? I hope not.)
> No, not likely.
> If you'd post exact error messages, that would certainly help.
> ("bad_stat_data: 24338 is shared by at least two files" is not a message
> of block that cannot be read.)
Sorry, that was the message that came before I did the rebuild-tree. I
didn't capture the rebuild messages - the were essentially "can not read
block xxxxxx" where xxxxxx was some number (it counted down from
3,000,000something to 0.
>> losetup should be mentioned in the reiserfsck man page. Should I prepare
>> a
>> patch (where to get the newest version so that I can patch this?)?
>
> Actually it is already mentioned in mount(8) man page, and if you use
> loop devices, you should be aware of that.
I do think putting it into the reiserfsck is better.
man mount mentions this on my system:
One further possible type is a mount via the loop device.
For example, the command
mount /tmp/fdimage /mnt -t msdos -o loop=/dev/loop3,blocksize=1024
will set up the loop device /dev/loop3 to correspond to
the file /tmp/fdimage, and then mount this device on /mnt.
This type of mount knows about three options, namely loop,
offset and encryption, that are really options to los
etup(8). If no explicit loop device is mentioned (but
just an option `-o loop' is given), then mount will try to
find some unused loop device and use that. If you are not
so unwise as to make /etc/mtab a symbolic link to
/proc/mounts then any loop device allocated by mount will
be freed by umount. You can also free a loop device by
hand, using `losetup -d', see losetup(8).
But that doesn't help the "unexperienced" - since you can't mount the
device, you *must* use losetup. And this isn't what you would get from the
man page of mount. Also the man page of losetup isn't very helpfull. It
explaines how losetup works, but it doesn't give you an idea *why* losetup
is different than mount and that this is the trick to get reiserfsck to
work.
And, reading the man page of reiserfsck is the first thing one would do ;)
One more hint can't hurt. I always hate it when I hunt around manpages, and
then overlook the only one, halfsentence mention in brief passing by what
is the solution, but looks totally unreleated ;)
> Though I think it worth including into our FAQ.
Great! (I looked at the FAQ, so I would have found it ;)
Thanx in advance,
Tels
- --
"Why do you go so slowly? Do you think this is some kind of game?"
PGP key available on http://bloodgate.com/tels.asc or via email.
perl -MDev::Bollocks -e'print Dev::Bollocks->rand(),"\n"'
confidentially facilitate web-enabled markets
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.0.6 (GNU/Linux)
iQEVAwUBPJYncncLPEOTuEwVAQGjzAf+OcPuU6Yi0DSIZ6mwAoYvxIqu9M8AgvjD
Ht1TyYuv13XQhh7yHsaZ0xorkYUhurzgWssd27XrDer9Y7vThh1+1OwN8aFgMJuH
5FSbc7g6B+tixzxKFD3alGFxC9b8mCjndUWg3vlyp9q5r2A1oiD6GUIEHQsRCtxI
Pi5Ncs53WZzVp9/5yt6xqUv3oyrcvlXc+aUnBmpR2AAUYY6n5lOgiht0q6Y3nTFl
gCy/Uj/ucI2r+ad84OZwgnA0mYwUVrlxvdfM5OSjt6Sr45uvZIydgX8Fra0OgXsp
GSs9ARkRIY5WPnTq25+DMNbn5RcUDOBkDoeP5pyw9tLaT8zAkFAIaA==
=TB1r
-----END PGP SIGNATURE-----
