[Buildroot] Buildroot, NAND flash and missing mkfs.jffs2 option

[Buildroot] Buildroot, NAND flash and missing mkfs.jffs2 option

[julien.boibessot at free.fr]

Hi !

I have found that on my NAND flash, if I want to burn a JFFS2 rootfs image
generated with buildroot, I need to add "-n" option to JFFS2_OPTS variable in
target/jffs2/jffs2.mk Makefile.

Otherwise when mounting my rootfs, I have a lot of warning messages like this
one:
    CLEANMARKER node found at 0x00...... has totlen 0xc != normal 0x0

Was anybody aware of that special case ?
Should I send you a patch to support NAND flash ?

Regards,
Julien

[Buildroot] Buildroot, NAND flash and missing mkfs.jffs2 option

[Bernhard Fischer]

On Tue, Nov 28, 2006 at 07:09:44PM +0100, julien.boibessot at free.fr wrote:
>Hi !
>
>I have found that on my NAND flash, if I want to burn a JFFS2 rootfs image
>generated with buildroot, I need to add "-n" option to JFFS2_OPTS variable in
>target/jffs2/jffs2.mk Makefile.
>
>Otherwise when mounting my rootfs, I have a lot of warning messages like this
>one:
>    CLEANMARKER node found at 0x00...... has totlen 0xc != normal 0x0
>
>Was anybody aware of that special case ?

>Should I send you a patch to support NAND flash ?

I think there is at least one report at bugs.busybox.net about jffs2,
namely an endian fix and maybe another report. It would be great if you
could verify that the endian patch does break nothing for you, so that i
can apply it safely and check if there is already a report WRT jffs2 and
-n. If there is already such a report, then i ask you to be so kind and
to point me to it and also point out further problems you did encounter
with jffs2 (like the -n you mentioned above).

TIA,
Bernhard

[Buildroot] Buildroot, NAND flash and missing mkfs.jffs2 option

[julien.boibessot at free.fr]

Hi !

> I think there is at least one report at bugs.busybox.net about jffs2,
> namely an endian fix and maybe another report. It would be great if you
> could verify that the endian patch does break nothing for you, so that i
> can apply it safely and check if there is already a report WRT jffs2 and
> -n. If there is already such a report, then i ask you to be so kind and
> to point me to it and also point out further problems you did encounter
> with jffs2 (like the -n you mentioned above).
>

I didn't find such a report (search with jffs2) so I have open a new one:
http://busybox.net/bugs/view.php?id=1105

If you need a patch to correct the problem, let me know, I can generate one
against Buildroot version 20060908 (the snapshot I'm using).

Regards,
Julien

[Buildroot] Buildroot, NAND flash and missing mkfs.jffs2 option

[Bernhard Fischer]

On Wed, Nov 29, 2006 at 02:04:21PM +0100, julien.boibessot at free.fr wrote:
>Hi !
>
>> I think there is at least one report at bugs.busybox.net about jffs2,
>> namely an endian fix and maybe another report. It would be great if you
>> could verify that the endian patch does break nothing for you, so that i
>> can apply it safely and check if there is already a report WRT jffs2 and
>> -n. If there is already such a report, then i ask you to be so kind and
>> to point me to it and also point out further problems you did encounter
>> with jffs2 (like the -n you mentioned above).
>>
>
>I didn't find such a report (search with jffs2) so I have open a new one:
>http://busybox.net/bugs/view.php?id=1105

My bad, it was a report about cramfs.
>
>If you need a patch to correct the problem, let me know, I can generate one
>against Buildroot version 20060908 (the snapshot I'm using).

Yes, please send a patch.

TIA,
Bernhard

[Buildroot] Buildroot, NAND flash and missing mkfs.jffs2 option

[julien.boibessot at free.fr]

Quoting Bernhard Fischer <rep.nop@aon.at>:
> >
> >If you need a patch to correct the problem, let me know, I can generate one
> >against Buildroot version 20060908 (the snapshot I'm using).
>
> Yes, please send a patch.
>

I have attached the patch to the bug report. It has been generated against
Buildroot 20061111.

I hope it will be helpfull !

Julien