[EMAIL PROTECTED] said:
>  An other approach would be to create a block device driver for my
> paged 8Meg flash chip.

> Anyone have any thoughts as to which is a better approach?  The
> read-only characterisic of creating a block device driver isn't a
> problem.

It shouldn't need to be read-only. Creating a 'map' driver if your actual 
flash chip is already supported by the MTD code is fairly trivial.
See linux/drivers/mtd/octagon-5066.c for an example.

--
dwmw2


Reply via email to