nand driver bug?

[Thomas Gleixner]

On Thursday 15 April 2004 09:49, Dmitry Konyshev wrote:
> Thomas Gleixner wrote:
> >>    The driver happens to fall in BUG in nand_select_chip when writing
> >>the last page of a device. It's trying to select_chip in nand_write_ecc
> >>even in case there's nothing left to write.
> >
> >Yep, you're right. I noticed this already and fixed it, but did not yet
> > commit it, as I changed a lot of other things.
> >Should be fixed until tomorrow morning
>
> Any changes should be done in boards' drivers? Fresh CVS version oopses
> in nand_scan.

Yep. I'm writing a short summary at the moment. 
Please check the changes in autcpu12.c

We need one additional buffer. The size is oobsize * pages per block, e.g.
blocksize 16384
pagesize 512
= 32 pages per block
* 16 byte oob
= 512 byte buffer
The pointer must be set before calling nand_scan


-- 
Thomas
________________________________________________________________________
"Free software" is a matter of liberty, not price. To understand the concept,
you should think of "free" as in "free speech,'' not as in "free beer".
________________________________________________________________________
linutronix - competence in embedded & realtime linux
http://www.linutronix.de
mail: tglx at linutronix.de