Bad Blocks on NAND and JFFS2

Thomas Gleixner gleixner at
Wed Mar 20 03:43:43 EST 2002

On Dienstag, 19. März 2002 10:40 , Thomas Locher wrote:
> I tried to put the JFFS2 on our NAND flash, but now I have troubles with
> bad blocks on the flash device.
> The actual CVS ( code seems to support bad
> block handling, but the code
> doesn't work correctly.
> What is the status of bad block handling in JFFS2?
> I ask you this because I think you have developped most of the bad block
> handling.
Please keep such questions on the mailing list. Others may be interrested 
too. :)

Which chips do you use ? Which file did you use as template to hack your 
chipdriver ? Have you adjusted 
        /* 20 us command delay time */
        this->chip_delay = 20;
properly ?

The bad block handling is fully functional. It uses the bad block marking 
scheme of NAND chips (Samsung/Toshiba) and Smartmediacards, which is the same.

You must enable
Enable ECC correction algorithm'  (CONFIG_MTD_NAND_ECC)

There were some modifications to JFFS2 recently, so the CVS seems to be 
broken for JFFS2 and NAND at the moment. I will check this anyway today. Keep 
an eye on the maillist and on linux-mtd-cvs at (cvs announce 
Thomas Gleixner <gleixner at>
autronix automation

More information about the linux-mtd mailing list