Fwd: udf folders pointing to self and above

Fwd: udf folders pointing to self and above

[Pat LaVarre]

[-- Attachment #1: Type: text/plain, Size: 912 bytes --]

linux-scsi kindly suggests linux-fsdevel@vger.kernel.org folk may wish
to discuss:

-----Forwarded Message-----

From: Pat LaVarre <p.lavarre@ieee.org>
To: linux-scsi@vger.kernel.org
Subject: udf folders pointing to self and above
Date: 29 Sep 2003 13:37:42 -0600

To create a place to collect what you & Google teach me over time, I
will go ahead and mention:

Often I find if I create an initially zero + `mkudffs` disk of 1 GiB,
the Philips udf verifier udf_test complains as follows.

...
    309 read block
EFE  file type DIR    name: "lost+found"
EFE  28  icbtag error: parentICBLocation: (34,0),
- recommended: (0,0) for strategy type 4.
- Icbtag Parent ICB Location shall not point to itself
- or to a higher parent in the directory hierarchy.
- ECMA 3rd edition 4/8.10.1, 4/14.6.7, UDF 2.3.5.3.
- icbtag: ECMA 4/14.6.*, 4/A.5, UDF 2.3.5.*, 6.6
- name: "lost+found"
EFE  216 Embedded data, 40 bytes
...


[-- Attachment #2: zlxudf.udfct_1.0r2.txt.bz2 --]
[-- Type: application/x-bzip, Size: 3939 bytes --]

Re: Fwd: udf folders pointing to self and above

[Pat LaVarre]

Courtesy the linux-fsdevel "editable udf metadata" thread, specifically:
http://marc.theaimsgroup.com/?l=linux-fsdevel&m=106677828430258
we now have this same error printed in Linux.

In the example tty log below I override the --udfrev 0x0201 default of
mkudffs, and still I get much the same error.

Pat LaVarre

$ dd of=dd.bin bs=1M seek=15 count=1 if=/dev/zero
1+0 records in
1+0 records out
$
$ sudo losetup /dev/loop0 dd.bin
$
$ sudo mkudffs --udfrev 0x0201 /dev/loop0
start=0, blocks=16, type=RESERVED
...
start=8191, blocks=1, type=ANCHOR
$
$ sudo udf_test /dev/loop0 | grep -i -B 1 -A 7 'error:'
        EFE  file type DIR    name: "lost+found"
        EFE  28  icbtag error: parentICBLocation: (3,0),
-                       recommended: (0,0) for strategy type 4.
-                Icbtag Parent ICB Location shall not point to itself
-                or to a higher parent in the directory hierarchy.
-                ECMA 3rd edition 4/8.10.1, 4/14.6.7, UDF 2.3.5.3.
-                icbtag: ECMA 4/14.6.*, 4/A.5, UDF 2.3.5.*, 6.6
-                name: "lost+found"
        EFE  216 Embedded data, 40 bytes
$

Re: Fwd: udf folders pointing to self and above

[Pat LaVarre]

Kindly offline, in reply to this thread titled "udf folders pointing to
self and above", I received the hint:

        use cvs udftools

Anybody here already have a clue how I can click thru that link?

Apparently I'm testing the wrong code?

Pat LaVarre

-----Forwarded Message-----

To: kernelnewbies@nl.linux.org
Subject: use cvs udftools means what
...

Didn't kernelnewbies.org once contain a cvs tutorial?  Revisiting now,
all I find is an rtfm link? ...

I had reported that the udf mkfs fails the udf fsck.

By "the udf mkfs" I mean the udftools-1.0.0b2/mkudffs/mkudffs that
`make` creates from:

http://sourceforge.net/projects/linux-udf/
http://sourceforge.net/project/showfiles.php?group_id=295&release_id=74270
...

By "the udf fsck" I mean the udfct_1.0r2/src/udf_tester/udf_test that
`make` creates from:

http://www.extra.research.philips.com/udf/download/udfct_1_0r2.tar.gz

Re: Fwd: udf folders pointing to self and above

[Pat LaVarre]

>         use cvs udftools
> Anybody here already have a clue how I can click thru that link?
> Apparently I'm testing the wrong code?

kernelnewbies thread "Re: use cvs udftools means what" has an answer:

http://marc.theaimsgroup.com/?l=kernelnewbies&m=106686957020891

To get the mkudffs we want people to test, possibly all we have left to
discover is how to correct the spelling of the line:

export CVSROOT=:pserver:anonymous@www.diabetikware.org:/home/cvs

Pat LaVarre

P.S. Sorry the first post of this thread does not have two blanks past
the Re: in "Re:  Fwd: udf folders pointing to self and above".

Re: Fwd: udf folders pointing to self and above

[Pat LaVarre]

Now I am iterating over firewalls.

Failure looks like:

$ export
CVSROOT=:pserver:anonymous@cvs.sourceforge.net:/cvsroot/linux-udf
$ cvs login
Logging in to
:pserver:anonymous@cvs.sourceforge.net:2401/cvsroot/linux-udf
CVS password:
cvs [login aborted]: connect to cvs.sourceforge.net(66.35.250.207):2401
failed: Connection refused
$

Pat LaVarre

P.S.

I see no hits at:
http://groups.google.com/groups?q=CVSROOT+cvsroot+linux-udf

Instead I got to this guess by combining kernelnewbies "use cvs udftools
means what" with the web trail:

http://sourceforge.net/projects/linux-udf/

CVS
http://sourceforge.net/cvs/?group_id=295

Basic Introduction to CVS and SourceForge.net (SF.net) Project CVS Services
http://sourceforge.net/docman/display_doc.php?docid=14033&group_id=1

Re: Fwd: udf folders pointing to self and above

[Steve Jobs]

Versus the 2002 Feb 9 1.0.0b2 udftools of sourceforge, I yield 2643 
lines of diff -ur if I try:

% setenv CVSROOT 
:pserver:anonymous@cvs.sourceforge.net:/cvsroot/linux-udf
%
% cvs login
(Logging in to anonymous@cvs.sourceforge.net)
CVS password:
%
% cvs co udftools
cvs server: Updating udftools
U udftools/.cvsignore
...
U udftools/wrudf/wrudf.h
%
% cvs logout
(Logging out of anonymous@cvs.sourceforge.net)
%

Pat LaVarre

Re: Fwd: udf folders pointing to self and above

[Pat LaVarre]

Solved!  Indeed, solved by "use cvs udftools", precisely as hinted.

Thank you all.

Pat LaVarre

P.S. A tty log of the contrast follows ...

> > $ sudo udf_test /dev/loop0 | grep -i -B 1 -A 7 'error:'
> >         EFE  file type DIR    name: "lost+found"
> >         EFE  28  icbtag error: parentICBLocation: (3,0),
> > -                       recommended: (0,0) for strategy type 4.
> > -                Icbtag Parent ICB Location shall not point to itself
> > -                or to a higher parent in the directory hierarchy.
> > -                ECMA 3rd edition 4/8.10.1, 4/14.6.7, UDF 2.3.5.3.
> > -                icbtag: ECMA 4/14.6.*, 4/A.5, UDF 2.3.5.*, 6.6
> > -                name: "lost+found"
> >         EFE  216 Embedded data, 40 bytes
>
> Versus the 2002 Feb 9 1.0.0b2 udftools of
> sourceforge, I yield 2643 lines of diff -ur
> if I try:
>
> % setenv CVSROOT :pserver:anonymous@cvs.sourceforge.net:/cvsroot/linux-udf

$ find `which mkudffs` -printf '%l\n'
.../udftools-cvs/mkudffs/mkudffs
$
$ sudo losetup /dev/loop0 dd.bin
$
$ sudo mkudffs --udfrev=0x0102 /dev/loop0
$
$ sudo udf_test /dev/loop0 -udf 1.02 | egrep -i '(error|warning|note):'
$
$ sudo udf_test /dev/loop0 -udf 1.02 | egrep -i 'final udf rev'
  Final UDF Revision range: 1.02 only
$