Werner Almesberger wrote:
[EMAIL PROTECTED] wrote:
The code does't skip coorectly the bad block, because it use
start_addr instead
of i in the if code in main loop (so the condition is always true).
Then it must check the first and second page, byte 5 of 16 extrabyte.
Yes, I see that there are three changes in your patch:
- avoid false positives by only checking the beginning of a block
- check the second page as well
- do more wait_idle
The first two fix pretty obvious bugs. I'm curious about the third:
did you experience failures due to the lack of waiting here ?
No, The third is taken from u-boot. I don't see any failure at all.
Tomorrow I will add support for nand with large page size for a
different product. Maybe I will be able to post another patch.
I have no access to the neo hardware, but I have several s3c24xx
based boards, I would like to contribute to the project, for the
beats concerning u-boot and kernel; can you point me to something
that would be useful?
Regards Michael
- Werner
