Tom,

Also, you may want to check out 
http://wiki.davincidsp.com/index.php?title=HOWTO_Create_Filesystems_on_DaVinci 
in addition to the sites you listed below.

Sincerely,
Chase Maupin
Software Applications
Catalog DSP Products
e-mail: [EMAIL PROTECTED]<mailto:[EMAIL PROTECTED]>
phone: (281) 274-3285
________________________________
From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of Maupin, Chase
Sent: Tuesday, August 05, 2008 11:07 AM
To: Tom Briggs; [email protected]
Subject: RE: DM6446 + Linux + UBoot + NAND = long sad days

Tom,

My first thought is that your nand write command does not look correct.  If you 
do a help on the 'nand' command you should see that you need to do

nand write 0x80700000 <address in nand> <size>

I know that on the DM6467 that the address in NAND should not be entered with 
the 0x2000000 added to it.  I believe this is also true for the DM644x.  So 
your command should look something like
nand erase 0x60000 0x154000
nand write 0x80700000 0x60000 0x154000

Sincerely,
Chase Maupin
Software Applications
Catalog DSP Products
e-mail: [EMAIL PROTECTED]<mailto:[EMAIL PROTECTED]>
phone: (281) 274-3285
________________________________
From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of Tom Briggs
Sent: Tuesday, August 05, 2008 8:58 AM
To: [email protected]
Subject: DM6446 + Linux + UBoot + NAND = long sad days

Hello all, I am hoping someone here can point me in the right direction.  I am 
working with the DM6446 dvevm board and am trying to get Linux / JFFS2 to play 
nicely with UBoot and boot the system from the on-board NAND.

I have UBoot on the NAND chip (flashed with DVFlasher).  I tried the TI UBoot 
as well as the latest 1.3.4 from 'git'.

I'm following a tutorial which suggests the following commands to put the 
kernel into NAND [1], with a change to where the kernel goes:

tftp 0x80700000 uImage
nand erase 0x2060000 0x154000
nand write 0x2060000 0x154000

and, lo, I can boot Linux, which shows the default MTD partitions:

NAND device: Manufacturer ID: 0xec, Chip ID: 0x36 (Samsung NAND 64MiB 1,8V 8-bi)
Scanning device for bad blocks
Creating 4 MTD partitions on "nand_davinci.0":
0x00000000-0x00040000 : "bootloader"
0x00040000-0x00060000 : "params"
0x00060000-0x00460000 : "kernel"
0x00460000-0x04000000 : "filesystem"
nand_davinci nand_davinci.0: hardware revision: 2.1

If I read this correctly, I've put the kernel at 0x2060000, which *should* be 
at the beginning of the "kernel" partition, right?

Next, I erase and create the file system on the "filesystem" partition 
(mtd3/mtdblock3):

flash_eraseall -j /dev/mtd3
mount -t jffs2 /dev/mtdblock3 /mnt/nand
cd /mnt/nand
tar xvfz .../rootfs.tgz

and, after some time rolls by, lo, I have a jffs2 file system that appears to 
be good (and there was much rejoicing).

But, woe unto those who think they are done....

After a power cycle, UBoot reports that the kernel image is no longer valid.   
If I reflash the kernel, and reboot, Linux reports that the file system is no 
longer valid.

So...... Is it the case that the sizes reported by MTD linearlly map onto 
addresses relative to the 0x2000000 start address for the NAND?  i.e. is it 
true that /dev/mtd0 starts at 0x2000000, /dev/mtd1 at 0x2040000, /dev/mtd2 at 
0x2060000, /dev/mtd3 at 0x2460000 ?

Any help would be greatly appreciated!  (and thanks to those who may have 
written the cited tutorials, they really helped)


[1] 
http://linux.omap.com/pipermail/davinci-linux-open-source/2007-September/004150.html

[2] http://wiki.davincidsp.com/index.php?title=Filesystem_in_NOR_or_NAND
--------------------------------------------------------
Tom Briggs
[EMAIL PROTECTED]<mailto:[EMAIL PROTECTED]>
_______________________________________________
Davinci-linux-open-source mailing list
[email protected]
http://linux.davincidsp.com/mailman/listinfo/davinci-linux-open-source

Reply via email to