[maemo-developers] Failing to connect with flasher

From: Frantisek Dufka dufkaf at seznam.cz
Date: Wed Apr 11 10:17:28 EEST 2007
pancake wrote:
  > BTW If you are able to boot Maemo but you're not able to connect to NOLO
> you should probably think on reflashing nolo from the device (using
> mtdtools).

Did you try? Maybe this is not a good idea. I'm not sure bootloader 
(mtd0) partition is safe to write with mtdutils. So far in every n770 
kernel log I've seen on the net is something like this

[93032.218597] omap-hw-nand: OMAP NAND Controller rev. 1.1
[93032.218780] NAND device: Manufacturer ID: 0xec, Chip ID: 0xa1 
(Samsung NAND 128MiB 1,8V 8-bit)
[93032.218994] omap-hw-nand: using PSC values 2, 2, 3
[93032.219116] Scanning device for bad blocks
[93032.219238] Bad eraseblock 0 at 0x00000000
[93032.253875] 5 cmdlinepart partitions found on MTD device omap-nand
[93032.253997] Creating 5 MTD partitions on "omap-nand":
[93032.254119] 0x00000000-0x00020000 : "bootloader"

What makes me nervous is the "Bad eraseblock 0 at 0x00000000" line. This 
  may mean there is really a bad block but most likely it means the oob 
data has different format than with normal block so writing it with 
mtdutils would in theory repair/destroy the oob data. Also in reality it 
means this block cannot be written from linux kernel at all since it 
cannot write to bad blocks -> mtd utils will not write to first block. 
This may be good and may be not.

Also when I tried nanddump on /dev/mtd0 I got file with all FF and no 
errors. Maybe first bad block confuses it (-b does not help) maybe there 
is other problem  (dd if=/dev/mtdblock0 works fine however). This means 
I wouldn't trust nandwrite for writing correct data too.

Frantisek

More information about the maemo-developers mailing list