[Arm-netbook] Real size of Mele A1000 NAND

[David Given]

On 11/05/12 10:06, David Given wrote:
[...]
> Well, the block device layer is very small and very simple; it ought to
> be relatively trivial to throw it away and replace it with a much
> simpler block device layer that exposes a single big device and then
> uses Linux's own partitioning.

Not as trivial as you might think --- the block driver layer is a fair
mass of undocumented spaghetti. Rather than rewrite it entirely I just
hacked their driver to assume a single NAND partition occupying the
entire space, called /dev/nand0, into which we can then put a
traditional partition table with fdisk. Strictly the block driver needs
to be rewritten entirely.

If I create an fdisk partition table with a single partition
(/dev/nand0p1), and then format it ext4, I get 3.4GB free, which is
reasonable.

NB: I don't know how trustworthy their wear levelling code is... and
there's a bug (in their stuff) which causes a deadlock if you build the
driver as a module and then rmmod it. (I think the helper thread is
failing to exit on demand.)

This can also be used to back up the entire Android image by dding off
/dev/nand0 directly.

-- 
┌─── ｄｇ＠ｃｏｗｌａｒｋ．ｃｏｍ ───── http://www.cowlark.com ─────
│
│ "Never attribute to malice what can be adequately explained by
│ stupidity." --- Nick Diamos (Hanlon's Razor)
-------------- next part --------------
A non-text attachment was scrubbed...
Name: nand-partition.patch
Type: text/x-patch
Size: 3592 bytes
Desc: not available
Url : http://lists.phcomp.co.uk/pipermail/arm-netbook/attachments/20120512/d4c32262/attachment.bin 
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 254 bytes
Desc: OpenPGP digital signature
Url : http://lists.phcomp.co.uk/pipermail/arm-netbook/attachments/20120512/d4c32262/attachment-0001.bin