possible bug in "--pad" option for mkfs.jffs2?

possible bug in "--pad" option for mkfs.jffs2?

[Ryan B. Lynch]

I don't know if this is really a bug, but mkfs.jffs2 doesn't seem to be
working entirely as advertised for me.  I'm trying to use the "--pad"
option to create some relatively large JFFS2 images (~200-950 MB).  I
know that these are larger than is common, so I'm also on the lookout
for size-related limitations.

I run 'mkfs.jffs2' with the following command line:

 $ /usr/sbin/mkfs.jffs2 \
 >    -r /tmp/root \
 >    -o /tmp/image-20050815.160044.jffs2 \
 >    -e 0x10000 \
 >    -p 214171648

The image size is NOT padded to 214,171,648 bytes, though--the size
depends on the contents of the filesystem (I tried varying it a little
to see what happens).  I'm assuming that "-e" and "-p" both take values
in bytes... Maybe someone can correct me on that?

Can anyone correct me, or give me some troubleshooting direction?

Thanks for the help,
Ryan

Re: possible bug in "--pad" option for mkfs.jffs2?

[Ryan B. Lynch]

I did a little more testing with this, and the "--pad" option doesn't
work AT ALL on my installation.  Something is definitely wrong, here.

I compiled JFFS2 from source, using a CVS snapshot from 2005/08/14, with
the "jffs2-summary-20050811-v2.patch" applied.

Can anyone confirm this?

-Ryan


Ryan B. Lynch wrote:

>I don't know if this is really a bug, but mkfs.jffs2 doesn't seem to be
>working entirely as advertised for me.  I'm trying to use the "--pad"
>option to create some relatively large JFFS2 images (~200-950 MB).  I
>know that these are larger than is common, so I'm also on the lookout
>for size-related limitations.
>
>I run 'mkfs.jffs2' with the following command line:
>
> $ /usr/sbin/mkfs.jffs2 \
> >    -r /tmp/root \
> >    -o /tmp/image-20050815.160044.jffs2 \
> >    -e 0x10000 \
> >    -p 214171648
>
>The image size is NOT padded to 214,171,648 bytes, though--the size
>depends on the contents of the filesystem (I tried varying it a little
>to see what happens).  I'm assuming that "-e" and "-p" both take values
>in bytes... Maybe someone can correct me on that?
>
>Can anyone correct me, or give me some troubleshooting direction?
>
>Thanks for the help,
>Ryan
>
>______________________________________________________
>Linux MTD discussion mailing list
>http://lists.infradead.org/mailman/listinfo/linux-mtd/
>
>  
>

Re: possible bug in "--pad" option for mkfs.jffs2?

[Artem B. Bityuckiy]

Ryan B. Lynch wrote:
> I did a little more testing with this, and the "--pad" option doesn't
> work AT ALL on my installation.  Something is definitely wrong, here.
> 
> I compiled JFFS2 from source, using a CVS snapshot from 2005/08/14, with
> the "jffs2-summary-20050811-v2.patch" applied.
> 
> Can anyone confirm this?

Dunno, never used this option (what for?). Glance at utils/mkfs.jffs2.c. :-)

-- 
Best Regards,
Artem B. Bityuckiy,
St.-Petersburg, Russia.

Re: possible bug in "--pad" option for mkfs.jffs2?

[Ryan B. Lynch]

Hey,

Artem B. Bityuckiy wrote:

> Ryan B. Lynch wrote:
>
> Dunno, never used this option (what for?). Glance at
> utils/mkfs.jffs2.c. :-) \


The "--pad" option?  From the manpage:

"-p, --pad[=SIZE]
    Pad output to SIZE bytes with 0xFF.  If SIZE is not specified, the
output is padded to the end of the final erase block."

I'm looking at the C source right now, too... gonna see if I can figure
out WTF is going on.

-Ryan

Re: possible bug in "--pad" option for mkfs.jffs2?

[Ferenc Havasi]

Hi Ryan,

>I run 'mkfs.jffs2' with the following command line:
>
> $ /usr/sbin/mkfs.jffs2 \
> >    -r /tmp/root \
> >    -o /tmp/image-20050815.160044.jffs2 \
> >    -e 0x10000 \
> >    -p 214171648
>
>The image size is NOT padded to 214,171,648 bytes, though--the size
>  
>

Do not use space between -p and its parameter. It will work.

It is getopt bug (or feature) which happens using optional parameter.

If we whould change it to required parameter, than both "-p X" and "-pX"
would be OK, but in this case "-p X" means "-p" without parameter and X
- if I'm right.

Ferenc