Re: [linuxdc-dev] vmublk

[M. R. B. <mr...@0x...>]

* Adrian McMenamin <ad...@mc...> on Wed, Sep 18, 2002:

> There is now a block device driver for the vmu in the cvs. This is very=
=20
> closely modelled on the existing mtdblock driver - though is, IMHO, bette=
r=20
> suited to a small footprint system like the DC (it's amaller code),=20
> particularly with its small erase size on its flash (it doesn't employ th=
e=20
> same cacheing mechanism).
>=20

How much less of a footprint does your driver offer than the standard
mtdblock?  Is it on the order of megabytes, hundreds of kilobytes, what?
Is there any other reason we should deviate from the standard code?  How's
the caching mechansim different?

M. R.