[ parted-Bugs-303417 ] Assertion (metadata_length > 0) at dos.c:2011 [...] failed.

[noreply]

Bugs item #303417, was opened at 2006-05-04 10:00
You can respond by visiting: 
http://alioth.debian.org/tracker/?func=detail&atid=410685&aid=303417&group_id=30287

Category: parted
Group: None
Status: Open
Resolution: None
Priority: 5
Submitted By: Nobody (None)
Assigned to: Andrew Clausen (clausen-guest)
Summary: Assertion (metadata_length > 0) at dos.c:2011 [...] failed.

Initial Comment:
I tried to resize an ext3 partition, using the latest development version on 
Debian Sarge. It gracefully refused to do, saying "Assertion (metadata_length > 
0) at dos.c:2011 [...] failed.". fsck afterwards didn't report errors.
Parted believes i found a bug. The FAQ at
http://www.gnu.org/software/parted/faq.html says it was fixed some time ago.
Leslie P. Polzer at
http://www.mail-archive.com/address@hidden/msg01772.html says this bug
was fixed in 1.7rc.

I have included the output you reqire at the bug-report page and a full copy of 
the console output when trying to resize.

Best Regards
--Jakob


Appendix
------------------------------------------------------------------

$ uname -a
Linux pz 2.6.8-2-k7 #1 Tue Aug 16 14:00:15 UTC 2005 i686 GNU/Linux

------------------------------------------------------------------

$ sudo ./parted /dev/hdc
GNU Parted 1.7.0rc5
Using /dev/hdc
Welcome to GNU Parted! Type 'help' to view a list of commands.
(parted) resize 5 58589118s 135717182s
[33] ext2.c:686 (ext2_determine_itoffset): start = 0, it = 4
[33] ext2.c:686 (ext2_determine_itoffset): start = 32768, it = 32772
[33] ext2.c:686 (ext2_determine_itoffset): start = 65536, it = 65540
Warning: A resize operation on this file system will use EXPERIMENTAL code
that
MAY CORRUPT it (although it hasn't done so yet).You should at least backup
your
data and run 'e2fsck -f' afterwards.
OK/Cancel? OK
Error: Unable to satisfy all constraints on the partition.


You found a bug in GNU Parted! Here's what you have to do:

Don't panic! The bug has most likely not affected any of your data.
Help us to fix this bug by doing the following:

Check whether the bug has already been fixed by checking
the last version of GNU Parted that you can find at:

        http://ftp.gnu.org/gnu/parted/

Please check this version prior to bug reporting.

If this has not been fixed yet or if you don't know how to check,
please visit the GNU Parted website:

        http://www.gnu.org/software/parted

for further information.

Your report should contain the version of this release (1.7.0rc5)
along with the error message below, the output of

        parted DEVICE unit co print unit s print

and additional information about your setup you consider important.

Assertion (metadata_length > 0) at dos.c:2011 in function
add_logical_part_metadata() failed.

Ignore/Cancel? C
(parted) q
Information: Don't forget to update /etc/fstab, if necessary.

------------------------------------------------------------------

$ sudo ./parted /dev/hdc unit co print unit s print

Disk /dev/hdc: 82,3GB
Sector size (logical/physical): 512B/512B
Partition Table: msdos

Number  Start   End     Size    Type      File system  Flags
 1      32,3kB  3997MB  3997MB  primary   ext3         boot, raid
 3      3997MB  4499MB  502MB   primary   linux-swap   raid
 4      4499MB  30,0GB  25,5GB  primary   ext3         raid
 2      30,0GB  82,3GB  52,3GB  extended
 5      30,0GB  44,4GB  14,4GB  logical   ext3
 6      69,5GB  82,3GB  12,9GB  logical   ext3


Disk /dev/hdc: 160836479s
Sector size (logical/physical): 512B/512B
Partition Table: msdos

Number  Start       End         Size        Type      File system  Flags
 1      63s         7807589s    7807527s    primary   ext3         boot,
raid
 3      7807590s    8787554s    979965s     primary   linux-swap   raid
 4      8787555s    58589054s   49801500s   primary   ext3         raid
 2      58589055s   160826714s  102237660s  extended
 5      58589118s   86799194s   28210077s   logical   ext3
 6      135717183s  160826714s  25109532s   logical   ext3

Information: Don't forget to update /etc/fstab, if necessary.

------------------------------------------------------------------

$ sudo fdisk -l u /dev/hdc

Disk /dev/hdc: 82.3 GB, 82348277760 bytes
255 heads, 63 sectors/track, 10011 cylinders
Units = cylinders of 16065 * 512 = 8225280 bytes

   Device Boot      Start         End      Blocks   Id  System
/dev/hdc1   *           1         486     3903763+  fd  Linux raid
autodetect
/dev/hdc2            3648       10011    51118830    5  Extended
/dev/hdc3             487         547      489982+  fd  Linux raid
autodetect
/dev/hdc4             548        3647    24900750   fd  Linux raid
autodetect
/dev/hdc5            3648        5403    14105038+  83  Linux
/dev/hdc6            8449       10011    12554766   83  Linux

Partition table entries are not in disk order

------------------------------------------------------------------

$ sudo fdisk -l -u /dev/hdc

Disk /dev/hdc: 82.3 GB, 82348277760 bytes
255 heads, 63 sectors/track, 10011 cylinders, total 160836480 sectors
Units = sectors of 1 * 512 = 512 bytes

   Device Boot      Start         End      Blocks   Id  System
/dev/hdc1   *          63     7807589     3903763+  fd  Linux raid
autodetect
/dev/hdc2        58589055   160826714    51118830    5  Extended
/dev/hdc3         7807590     8787554      489982+  fd  Linux raid
autodetect
/dev/hdc4         8787555    58589054    24900750   fd  Linux raid
autodetect
/dev/hdc5        58589118    86799194    14105038+  83  Linux
/dev/hdc6       135717183   160826714    12554766   83  Linux

Partition table entries are not in disk order


----------------------------------------------------------------------

>Comment By: Leslie P. Polzer (dejari-guest)
Date: 2006-05-25 23:08

Message:
Logged In: YES 
user_id=12119

Thanks a lot for the extensive information.

> 5) I'm starting to become afraid that there could be
> something wrong with my setup.
My guess is that it's the new experimental code that is at 
fault.  This new code is responsible for resizing ext2/3 file systems with a 
layout that originally was not supported, and I anticipated it to have quirks.
All we can do is debug it now.
Your setup should be perfectly well.

> As I stated in the original report, I'm running pure
> Debian Sarge (=stable), and this is where I compiled
> parted 1.7.0 from source.
> Is that supposed to work at all?
I can't think of any reason why it shouldn't work.  Parted does most things 
itself;  it hardly relies on other libraries for its core functionality.

----------------------------------------------------------------------

Comment By: Nobody (None)
Date: 2006-05-25 22:24

Message:
Logged In: NO 

A lot of news:

1) In the meantime, the partition layout on hdc has changed. I did a "poor 
man's" resize - shovel as much data as possible off the partitions, delete 
them, recreate a larger one.

It now looks like this:
------------------------------------------------------------------
$ sudo sfdisk -l -x -uS /dev/hdc

Disk /dev/hdc: 10011 cylinders, 255 heads, 63 sectors/track
Units = sectors of 512 bytes, counting from 0

   Device Boot    Start       End   #sectors  Id  System
/dev/hdc1   *        63   7807589    7807527  fd  Linux raid autodetect
/dev/hdc2      58589055 160826714  102237660   5  Extended
/dev/hdc3       7807590   8787554     979965  fd  Linux raid autodetect
/dev/hdc4       8787555  58589054   49801500  fd  Linux raid autodetect

/dev/hdc5      58589118 135717119   77128002  83  Linux
    -          58589055  58589054          0   0  Empty
    -          58589055  58589054          0   0  Empty
    -          58589055  58589054          0   0  Empty
------------------------------------------------------------------

2) So I want to resize /dev/hdc5 (still ext3) to use all availible space.

3) The problem has changed. Probably someone should update the bug title...

Trying to resize gives me a bunch of warnings, the "Assertion (metadata_length 
> 0) at dos.c:2011 [...] failed." thing is gone. I noticed that you have 
updated the FAQ
------------------------------------------------------------------
Warning: Block 1835521 shouldn't have been marked (0, 1)!
Warning: Block 1835522 shouldn't have been marked (0, 1)!
[...]
Warning: Block 9175556 shouldn't have been marked (0, 1)!
Warning: Block 9175557 shouldn't have been marked (0, 1)!
[...]
Assertion (block < EXT2_SUPER_BLOCKS_COUNT(fs->sb)) at ext2.h:226 in function
ext2_is_data_block() failed.

Ignore/Cancel? Terminated [<- That was me]
------------------------------------------------------------------
If I didn't kill parted here, I would have experienced this (and I did the 
first time):
http://alioth.debian.org/tracker/index.php?func=detail&aid=303476&group_id=30287&atid=410685


4) I tried to gdb it.
------------------------------------------------------------------
$ sudo gdb --args ./parted /dev/hdc
(gdb) break ext2_is_data_block
Function "ext2_is_data_block" not defined.
Make breakpoint pending on future shared library load? (y or [n]) y
Breakpoint 1 (ext2_is_data_block) pending.
(gdb) run
Starting program: /home/jakob/parted-1.7.0/installed/sbin/parted /dev/hdc
Breakpoint 2 at 0x40037d0e
Pending breakpoint "ext2_is_data_block" resolved
GNU Parted 1.7.0
Using /dev/hdc
[...Warnings; You have found a bug etc...]

Assertion (block < EXT2_SUPER_BLOCKS_COUNT(fs->sb)) at ext2.h:226 in function
ext2_is_data_block() failed.

Ignore/Cancel?
Program received signal SIGTERM, Terminated. [<- That was me]
0x401628de in read () from /lib/tls/libc.so.6
(gdb)
------------------------------------------------------------------
But it won't break at ext2_is_data_block for whatever reason.

5) I'm starting to become afraid that there could be something wrong with my 
setup.
As I stated in the original report, I'm running pure Debian Sarge (=stable), 
and this is where I compiled parted 1.7.0 from source. Is that supposed to work 
at all?

Several days ago, a debian-unstable package has been built for parted-1.7.0.
Looking at the dependencies listed there i noticed that I didn't have installed 
several of them.
For that reason, installing the unstable .deb fails, as it requires libraries 
that are in the unstable repository.
I'm not used to compiling stuff myself so there may things that I have 
overlooked.
In any case, parted-1.7.0 does seem to compile OK.
It came to my mind that shared libraries may be the problem, so here is the 
output of ldd:
------------------------------------------------------------------
~/parted-1.7.0/installed/sbin$ ldd parted
libparted-1.7.so.0 => /home/jakob/parted-1.7.0/installed/lib/libparted-1.7.so.0 
(0x40018000)
libuuid.so.1 => /lib/libuuid.so.1 (0x40092000)
libdl.so.2 => /lib/tls/libdl.so.2 (0x40095000)
libc.so.6 => /lib/tls/libc.so.6 (0x40098000)
/lib/ld-linux.so.2 => /lib/ld-linux.so.2 (0x40000000)
------------------------------------------------------------------

But there _is_ someone else experiencing the same problem:
http://lists.gnu.org/archive/html/bug-parted/2006-05/msg00016.html posted 
2006-05-09


Regards,
--Jakob

----------------------------------------------------------------------

Comment By: Leslie P. Polzer (dejari-guest)
Date: 2006-05-25 12:44

Message:
Logged In: YES 
user_id=12119

Here goes:

gdb parted --args /dev/hdc
(gdb) break add_logical_part_metadata
(gdb) run
[enter commands and wait until it break at the offending point.  then:]
(gdb) bt full
[and send us the output]
(gdb) quit

If it break more than once, please try to send a backtrace for each occurence.  
 But the most important one is of course that last one, before it stops working.

----------------------------------------------------------------------

Comment By: Nobody (None)
Date: 2006-05-25 12:09

Message:
Logged In: NO 

Providing root SSH access is not possible i'm afraid.

A GDB backtrace would be entirely possible, i'd even do with vague directions 
probably.

Regards,
--Jakob

----------------------------------------------------------------------

Comment By: Leslie P. Polzer (dejari-guest)
Date: 2006-05-24 20:20

Message:
Logged In: YES 
user_id=12119

Thanks for the info so far.

Is it possible for you to provide us SSH access?
Andrew Clausen said he cannot reproduce the bug.  If not, do you think you 
could provide us, given the exact instructions, with a GDB backtrace?

----------------------------------------------------------------------

Comment By: Nobody (None)
Date: 2006-05-21 15:23

Message:
Logged In: NO 

Of course.
--Jakob Unterwurzacher


$ sudo sfdisk -l -x -uS /dev/hdc

Disk /dev/hdc: 10011 cylinders, 255 heads, 63 sectors/track
Units = sectors of 512 bytes, counting from 0

   Device Boot    Start       End   #sectors  Id  System
/dev/hdc1   *        63   7807589    7807527  fd  Linux raid autodetect
/dev/hdc2      58589055 160826714  102237660   5  Extended
/dev/hdc3       7807590   8787554     979965  fd  Linux raid autodetect
/dev/hdc4       8787555  58589054   49801500  fd  Linux raid autodetect

/dev/hdc5      58589118 135717119   77128002  83  Linux
    -         135717120 160826714   25109595   5  Extended
    -          58589055  58589054          0   0  Empty
    -          58589055  58589054          0   0  Empty

/dev/hdc6     135717183 160826714   25109532  83  Linux
    -         135717120 135717119          0   0  Empty
    -         135717120 135717119          0   0  Empty
    -         135717120 135717119          0   0  Empty

----------------------------------------------------------------------

Comment By: Leslie P. Polzer (dejari-guest)
Date: 2006-05-20 15:18

Message:
Logged In: YES 
user_id=12119

Can we have the output of

  /sbin/sfdisk -l -x -uS /dev/hdc

please?

----------------------------------------------------------------------

Comment By: Nobody (None)
Date: 2006-05-17 19:33

Message:
Logged In: NO 

I updated parted to 1.7.0, problem persists.

Some additional info that may be useful:
parted-1.7.0 source package, configured --without-readline

Greetings
-- Jakob Unterwurzacher

----------------------------------------------------------------------

You can respond by visiting: 
http://alioth.debian.org/tracker/?func=detail&atid=410685&aid=303417&group_id=30287