[Yaffs] Re: Using yaffs images to program NAND flash YAFFS partitions

[Charles Manning]

On Thursday 24 February 2005 01:17, Wookey wrote:
> +++ =C0=CC=C0=E5=BF=F8 [05-02-23 16:28 +0900]:
> > Hi,
> > I've learned before that when using a YAFFS image (made by mkyaffsima=
ge)
> > and dumping it on a NAND device, the first block of the YAFFS partiti=
on
> > should be not written, as the first block is reserved for YAFFS.
> > But in that case, what should I do if the first block of the partitio=
n is
> > to be marked as bad?
>
> If that is the case then you have to consider the whole device bad. My
> understanding is that the first block is supposed to be guaranteed work=
ing.

That is true for the first block of the **whole** NAND - not necessarily =
for=20
the first block of a **partition** if the partition is not at the beginni=
ng=20
of NAND. The manufacturers guarantee that the first block in NAND is good=
.

Regardless, YAFFS does not care.

>
> In fact it is only 'reserved' in the sense that we use 0 to mean someth=
ing
> else so block 0 is not used for real data. This means that it will prob=
ably
> work fine even if block 0 is duff.
>
> I think that's right, and hasn't changed for YAFFS2. (charles?).

THis has not been changed in YAFFS2.

-- Charles