Dataman Forum

Hi, we're trying to program a Micron MT29F2G08AADWP NAND with a Dataman 48+ Pro, with the following partition layout:

1. Uboot
2. kernel1
3. kernel2
4. jffs2 root filesystem1
5. jffs2 root filesystem2
6. jffs2 logging filesystem

The source comes from a BIN file, since we had trouble generating a correct HEX file.

I have two questions:

What is the correct procedure for performing such a flash, especially w.r.t. factory bad blocks, and to keep a consistent JFFS2 filesystem (via linux MTD).

And, is there a way to "clone" a NAND by reading a pre-programmed NAND, writing a different NAND, but adjusting for the per-device bad block tables?

mrybczyn

Hello mrybczyn,

I strongly recommend that you read some common materials about NAND Flash and also our application note for NAND Flash programming using our programmers which can be found HERE.

> > What is the correct procedure for performing such a flash, especially w.r.t. factory bad blocks, and to keep a consistent JFFS2 filesystem (via linux MTD).

Support for the Linux MTD compatible partitions is implemented into the software for the Dataman-48Pro2, 448Pro2 and 848Pro2 programmers only. Full details are covered in the above application note.

> > And, is there a way to "clone" a NAND by reading a pre-programmed NAND, writing a different NAND, but adjusting for the per-device bad block tables?

See my first comment. I can not imagine this being possible... . Given the nature of how NAND Flash works, this type of exercise is going to be extremely difficult (if not impossible).

If you have any further questions, please contact me.

Regards,

Neil Parker
Dataman Programmers

I have exactly the same issue, ie how to get the programmer to satisfy the Linux MTD requirements for 1-bit ECC in the spare blocks.

Could anyone share the solution with me.?

I am trying to get nanobootloader, uboot, Linux A, Linux B, and JFFS2 on one Nand flash partition.

Cloning would be good but I guess this is not possible.

Regards
David