On Wednesday 21 March 2007 23:34, Richard Genoud wrote:
> 2007/3/21, Vitaly Wool <vitalywool@gmail.com>:
> > On 3/21/07, Richard Genoud <richard.genoud@gmail.com> wrote:
> > > I trace the code, and it seems that it's a problem with getting the
> > > partitioning information.
> > > I'm going to search in this way.
> > >
> > > here's the trace I get :
> > > mount -tyaffs2 /dev/mtdblock0 /mnt/mtd0
> > >   into yaffs2_read_super_mtd
> > >     into get_sb_bdev
> > >       into open_bdev_excl
> > >         into do_open
> > > ->       get_gendisk : kobj_lookup returned NULL   <-
> > >         do_open : get_gendisk error
> > >       open_bdev_excl : blkdev_get error
> > >     open_bdev_excl error : IS_ERR(bdev) = -6
> > >   get_sb_bdev result=-6
> > >
> > > mount: Mounting /dev/mtdblock0 on /mnt/mtd0 failed: No such device or
> > > address
> >
> > Can you please post your .config here? First of all, I'd check that
> > MTD block device support is enabled at all.
> >
> > Vitaly
>
> Well, thanks for your help !
> I found it.
> It was, as you said, in the kernel :
> Caching block device access to MTD devices (MTD_BLOCK)
> I didn't enable it, and, as it is said in the help message :
> [quote]
> At the moment, it is also required for the Journalling Flash File
> System(s) to obtain a handle on the MTD device when it's mounted
> (although JFFS and JFFS2 don't actually use any of the functionality
> of the mtdblock device).
> [/quote]
> that's why i got jffs2 working and not yaffs...
> Anyway, thanks to you both !

Interesting... when YAFFS started off it used the same mechanisms as JFFS2 to 
locate the mtd device.

It looks like it is worth doing one of:
1) Change YAFFS to use mtd handles the same way.
- or -
2) Do something in Kconfig to prevent this sort of confusion.

Suggestions? Patches?

Thanx

-- CHarles