All of lore.kernel.org
 help / color / mirror / Atom feed
* Undeletable directory in reiser4
@ 2007-07-10 17:03 Marti Raudsepp
  2007-07-11 12:31 ` Edward Shishkin
  0 siblings, 1 reply; 6+ messages in thread
From: Marti Raudsepp @ 2007-07-10 17:03 UTC (permalink / raw)
  To: reiserfs-devel

Hi,

I seem to have found a tripped on a bug in reiser4: there is a
directory on my reiser4 partition that cannot be deleted since it is
supposedly "not empty", however, rm -rf cannot find any more files to
delete.

# rm -rf /var/tmp/portage
rm: cannot remove directory `/var/tmp/portage/sys-devel/gcc-4.1.2/work/
gcc-4.1.2/libjava/classpath/resource/gnu/java/locale': Directory not empty

I mounted the partition as read-only to get some more information out
of it with the debugfs.reiser4 tool. Note that the partition's vartmp/
directory is bind-mounted to /var/tmp.

Is there anything else I can do to help debug this? Will
CONFIG_REISER4_DEBUG help?

# debugfs.reiser4 /dev/primary/reiser4 --print-file /vartmp/portage/sys-devel/
gcc-4.1.2/work/gcc-4.1.2/libjava/classpath/resource/gnu/java/locale

exts:           3
mask:           0x7
plugin:         sdext_lw
offset:         2
len:            14
mode:           drwxr-xr-x
nlink:          2
size:           3
plugin:         sdext_unix
offset:         16
len:            28
uid:            250
gid:            250
atime:          Wed Jun 27 12:24:27 2007
mtime:          Wed Jun 27 13:16:01 2007
ctime:          Tue Jul 10 03:10:03 2007
rdev:           182
bytes:          182
plugin:         sdext_lt
offset:         44
len:            12
atime:          0
mtime:          0
ctime:          0


NR(2)  NAME     OFFSET HASH                              SDKEY
  0 .            54 0000000000000000:0000000000000000 041aac1:0041aad
  1 ..           78 0000000000000000:0000000000000000 041aab1:0041aac

# debugfs.reiser4 /dev/primary/reiser4 --print-file /vartmp/portage/sys-devel/
gcc-4.1.2/work/gcc-4.1.2/libjava/classpath/resource/gnu/java

exts:           3
mask:           0x7
plugin:         sdext_lw
offset:         2
len:            14
mode:           drwxr-xr-x
nlink:          3
size:           3
plugin:         sdext_unix
offset:         16
len:            28
uid:            250
gid:            250
atime:          Wed Jun 27 12:24:27 2007
mtime:          Wed Jun 27 13:16:01 2007
ctime:          Tue Jul 10 03:10:03 2007
rdev:           150
bytes:          150
plugin:         sdext_lt
offset:         44
len:            12
atime:          0
mtime:          0
ctime:          0


NR(3)  NAME     OFFSET HASH                              SDKEY
  0 .            80 0000000000000000:0000000000000000 041aab1:0041aac
  1 ..          104 0000000000000000:0000000000000000 041a981:0041aab
  2 locale      128 0000000000000000:0000000000000000 041aac1:0041aad


Hope this helps.

Regards,
Marti Raudsepp

^ permalink raw reply	[flat|nested] 6+ messages in thread

* Re: Undeletable directory in reiser4
  2007-07-10 17:03 Undeletable directory in reiser4 Marti Raudsepp
@ 2007-07-11 12:31 ` Edward Shishkin
  2007-07-11 15:35   ` Marti Raudsepp
  0 siblings, 1 reply; 6+ messages in thread
From: Edward Shishkin @ 2007-07-11 12:31 UTC (permalink / raw)
  To: Marti Raudsepp; +Cc: reiserfs-devel

Marti Raudsepp wrote:

> Hi,


Hi,
any ideas about kernel version that generated such
inconsistency? It seems this is an old bug..

Please, check your partition by fsck.reiser4.
To check root partition, boot in single mode and remount it as
readonly.

Thanks,
Edward.

>
> I seem to have found a tripped on a bug in reiser4: there is a
> directory on my reiser4 partition that cannot be deleted since it is
> supposedly "not empty", however, rm -rf cannot find any more files to
> delete.
>
> # rm -rf /var/tmp/portage
> rm: cannot remove directory `/var/tmp/portage/sys-devel/gcc-4.1.2/work/
> gcc-4.1.2/libjava/classpath/resource/gnu/java/locale': Directory not 
> empty
>
> I mounted the partition as read-only to get some more information out
> of it with the debugfs.reiser4 tool. Note that the partition's vartmp/
> directory is bind-mounted to /var/tmp.
>
> Is there anything else I can do to help debug this? Will
> CONFIG_REISER4_DEBUG help?
>
> # debugfs.reiser4 /dev/primary/reiser4 --print-file 
> /vartmp/portage/sys-devel/
> gcc-4.1.2/work/gcc-4.1.2/libjava/classpath/resource/gnu/java/locale
>
> exts:           3
> mask:           0x7
> plugin:         sdext_lw
> offset:         2
> len:            14
> mode:           drwxr-xr-x
> nlink:          2
> size:           3
> plugin:         sdext_unix
> offset:         16
> len:            28
> uid:            250
> gid:            250
> atime:          Wed Jun 27 12:24:27 2007
> mtime:          Wed Jun 27 13:16:01 2007
> ctime:          Tue Jul 10 03:10:03 2007
> rdev:           182
> bytes:          182
> plugin:         sdext_lt
> offset:         44
> len:            12
> atime:          0
> mtime:          0
> ctime:          0
>
>
> NR(2)  NAME     OFFSET HASH                              SDKEY
>  0 .            54 0000000000000000:0000000000000000 041aac1:0041aad
>  1 ..           78 0000000000000000:0000000000000000 041aab1:0041aac
>
> # debugfs.reiser4 /dev/primary/reiser4 --print-file 
> /vartmp/portage/sys-devel/
> gcc-4.1.2/work/gcc-4.1.2/libjava/classpath/resource/gnu/java
>
> exts:           3
> mask:           0x7
> plugin:         sdext_lw
> offset:         2
> len:            14
> mode:           drwxr-xr-x
> nlink:          3
> size:           3
> plugin:         sdext_unix
> offset:         16
> len:            28
> uid:            250
> gid:            250
> atime:          Wed Jun 27 12:24:27 2007
> mtime:          Wed Jun 27 13:16:01 2007
> ctime:          Tue Jul 10 03:10:03 2007
> rdev:           150
> bytes:          150
> plugin:         sdext_lt
> offset:         44
> len:            12
> atime:          0
> mtime:          0
> ctime:          0
>
>
> NR(3)  NAME     OFFSET HASH                              SDKEY
>  0 .            80 0000000000000000:0000000000000000 041aab1:0041aac
>  1 ..          104 0000000000000000:0000000000000000 041a981:0041aab
>  2 locale      128 0000000000000000:0000000000000000 041aac1:0041aad
>
>
> Hope this helps.
>
> Regards,
> Marti Raudsepp
> -
> To unsubscribe from this list: send the line "unsubscribe 
> reiserfs-devel" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html
>
>


^ permalink raw reply	[flat|nested] 6+ messages in thread

* Re: Undeletable directory in reiser4
  2007-07-11 12:31 ` Edward Shishkin
@ 2007-07-11 15:35   ` Marti Raudsepp
  2007-07-11 16:22     ` Edward Shishkin
  0 siblings, 1 reply; 6+ messages in thread
From: Marti Raudsepp @ 2007-07-11 15:35 UTC (permalink / raw)
  To: Edward Shishkin; +Cc: reiserfs-devel

Hi,

Forgot to Cc: the mailing list (I'm new to mailing lists), and also
forgot to mention that I'm on AMD64. The reiser4 file system lives on
a contiguous LVM2 volume.

Regards,
Marti Raudsepp

---------- Forwarded message ----------
From: Marti Raudsepp <marti@juffo.org>
Date: Jul 11, 2007 3:54 PM
Subject: Re: Undeletable directory in reiser4
To: Edward Shishkin <edward@namesys.com>

On 7/11/07, Edward Shishkin <edward@namesys.com> wrote:
> any ideas about kernel version that generated such
> inconsistency? It seems this is an old bug..

Oops, sorry, I knew I forgot something important in my rushed bug report.
I'm currently running kernel 2.6.21-gentoo-r3 with the appropriate
reiser4-for-2.6.21 patch. At the time the directory was created, I was
running 2.6.21-ck2.

> Please, check your partition by fsck.reiser4.

Here's what I got:
CHECKING THE SEMANTIC TREE
FSCK: obj40_repair.c: 223: obj40_stat_unix_check: Node (515), item (19),
[41aac:6c6f63616c6500:41aad] (stat40): wrong bytes (182), Should be (100).
FSCK: obj40_repair.c: 350: obj40_stat_lw_check: Node (515), item (19),
[41aac:6c6f63616c6500:41aad] (stat40): wrong size (3), Should be (2).
        Found 86670 objects (some could be encountered more then once).
        Time interval: Wed Jul 11 15:41:47 2007 - Wed Jul 11 15:42:00 2007
FSCK: repair.c: 550: repair_sem_fini: On-disk used block bitmap and
really used block bitmap
differ.
***** fsck.reiser4 finished at Wed Jul 11 15:42:00 2007
Closing fs...done

1 fatal corruptions were detected in FileSystem. Run with --build-fs
option to fix them.

I'll leave the file system in the broken state for now.

Regards,
Marti Raudsepp

^ permalink raw reply	[flat|nested] 6+ messages in thread

* Re: Undeletable directory in reiser4
  2007-07-11 15:35   ` Marti Raudsepp
@ 2007-07-11 16:22     ` Edward Shishkin
  2007-07-11 17:02       ` Marti Raudsepp
  0 siblings, 1 reply; 6+ messages in thread
From: Edward Shishkin @ 2007-07-11 16:22 UTC (permalink / raw)
  To: Marti Raudsepp; +Cc: reiserfs-devel

Marti Raudsepp wrote:

> Hi,
>
> Forgot to Cc: the mailing list (I'm new to mailing lists), and also
> forgot to mention that I'm on AMD64. The reiser4 file system lives on
> a contiguous LVM2 volume.
>
> Regards,
> Marti Raudsepp
>
> ---------- Forwarded message ----------
> From: Marti Raudsepp <marti@juffo.org>
> Date: Jul 11, 2007 3:54 PM
> Subject: Re: Undeletable directory in reiser4
> To: Edward Shishkin <edward@namesys.com>
>
> On 7/11/07, Edward Shishkin <edward@namesys.com> wrote:
>
>> any ideas about kernel version that generated such
>> inconsistency? It seems this is an old bug..
>
>
> Oops, sorry, I knew I forgot something important in my rushed bug report.
> I'm currently running kernel 2.6.21-gentoo-r3 with the appropriate
> reiser4-for-2.6.21 patch. At the time the directory was created, I was
> running 2.6.21-ck2.


Would you please pack metadata by
debugfs.reiser4 -P /dev/xxx | gzip > meta.gz
and let me download the file meta.gz

Thanks,
Edward.
 

>
>> Please, check your partition by fsck.reiser4.
>
>
> Here's what I got:
> CHECKING THE SEMANTIC TREE
> FSCK: obj40_repair.c: 223: obj40_stat_unix_check: Node (515), item (19),
> [41aac:6c6f63616c6500:41aad] (stat40): wrong bytes (182), Should be 
> (100).
> FSCK: obj40_repair.c: 350: obj40_stat_lw_check: Node (515), item (19),
> [41aac:6c6f63616c6500:41aad] (stat40): wrong size (3), Should be (2).
>        Found 86670 objects (some could be encountered more then once).
>        Time interval: Wed Jul 11 15:41:47 2007 - Wed Jul 11 15:42:00 2007
> FSCK: repair.c: 550: repair_sem_fini: On-disk used block bitmap and
> really used block bitmap
> differ.
> ***** fsck.reiser4 finished at Wed Jul 11 15:42:00 2007
> Closing fs...done
>
> 1 fatal corruptions were detected in FileSystem. Run with --build-fs
> option to fix them.
>
> I'll leave the file system in the broken state for now.
>
> Regards,
> Marti Raudsepp
> -
> To unsubscribe from this list: send the line "unsubscribe 
> reiserfs-devel" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html
>
>


^ permalink raw reply	[flat|nested] 6+ messages in thread

* Re: Undeletable directory in reiser4
  2007-07-11 16:22     ` Edward Shishkin
@ 2007-07-11 17:02       ` Marti Raudsepp
  2007-07-12  8:50         ` Christopher Covington
  0 siblings, 1 reply; 6+ messages in thread
From: Marti Raudsepp @ 2007-07-11 17:02 UTC (permalink / raw)
  To: Edward Shishkin; +Cc: reiserfs-devel

On 7/11/07, Edward Shishkin <edward@namesys.com> wrote:
> Would you please pack metadata by
> debugfs.reiser4 -P /dev/xxx | gzip > meta.gz
> and let me download the file meta.gz

Sure: http://juffo.org/stuff/reiser4_undeletable_dir_metadata.gz

The erroneous directory is located under /vartmp/portage_bug/...
(renamed from its original "portage")

Regards,
Marti Raudsepp

^ permalink raw reply	[flat|nested] 6+ messages in thread

* Re: Undeletable directory in reiser4
  2007-07-11 17:02       ` Marti Raudsepp
@ 2007-07-12  8:50         ` Christopher Covington
  0 siblings, 0 replies; 6+ messages in thread
From: Christopher Covington @ 2007-07-12  8:50 UTC (permalink / raw)
  To: reiserfs-devel

Although I don't have any useful information to report because this
happened to me a while ago and I was too lazy to report it or maybe
thought it was my own fault, I had the same problem (indeed with some
work directory of a java package on gentoo). Rebuilding the filesystem
may have fixed it,--'m relatively certain I'm rid of the annoyance now
and wouldn't have taken any steps more advanced than running fsck.

Regards,
Chris

^ permalink raw reply	[flat|nested] 6+ messages in thread

end of thread, other threads:[~2007-07-12  8:50 UTC | newest]

Thread overview: 6+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2007-07-10 17:03 Undeletable directory in reiser4 Marti Raudsepp
2007-07-11 12:31 ` Edward Shishkin
2007-07-11 15:35   ` Marti Raudsepp
2007-07-11 16:22     ` Edward Shishkin
2007-07-11 17:02       ` Marti Raudsepp
2007-07-12  8:50         ` Christopher Covington

This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.