RE: [microblaze-uclinux] how to mount flash via jffs2 as dir /mnt ?

[Brettschneider Falk <fbrettschneider@xxxxxxxxxxxxxxxxx>]

Hi,

John Williams wrote:
> Brettschneider Falk wrote:
> > mtd0 is the flash, mtd2 is the root filesystem (romfs). 
> > mtd0 is the same as
> > mtd1 due doubled registration (likely a bug?) in 
> > mtd/maps/physmap.c line 145
> > and 181, but it actually doesn't matter to the system and 
> > can be ignored:
> >   /> cat /proc/mtd
> >   dev:    size   erasesize  name
> >   mtd0: 01000000 00010000 "Physically mapped flash"
> >   mtd1: 01000000 00010000 "Physically mapped flash"
> >   mtd2: 00058000 00001000 "RAM"
> > 
> 
> I don't think it will work if you try to get MTD to manage 
> only a small
> region of your flash.  It will want to control entire chip.
Yeah, that was the trap I felt into. With the physical mapping, I wanted to
tell MTD the flash is only from 12MB to 16MB. But that is not compatible
with all the auto-probe stuff.

> For this
> reason, I suggest you use a full MTD partition that covers the entire
> chip, but perhaps only create /dev nodes that correspond to the region
> you are interested in.
Yippie, the MTD command line parsing has helped!
  mtdparts=physmap:12m(reserved)ro,-(user) root=1f:02
It leads to mtd0 to mtd3 where mtd2 is my user partition.
Now I could 'eraseall /dev/mtd2' and 'mount -t jffs2 /dev/mtdblock2 /mnt'.
:-)

Now we're still fighting with the flash-connection to our FPGA memory
controller, some tweaking with the endianess and adressing but I can see
light at the end of the tunnel...
And I hope fast programming simply works for mtd/chips/cfi_cmdset_0002.c if
I cheekily turn it on again.

Thanks for your help
Cheers, F@lk
___________________________
microblaze-uclinux mailing list
microblaze-uclinux@xxxxxxxxxxxxxx
Project Home Page : http://www.itee.uq.edu.au/~jwilliams/mblaze-uclinux
Mailing List Archive : http://www.itee.uq.edu.au/~listarch/microblaze-uclinux/