Re: [PATCH 2.6.21] cramfs: add cramfs Linear XIP

[Carsten Otte <cotte@de.ibm.com>]

Christoph Hellwig wrote:
> We can have a simpler variant as a start if we really want.  But we
> need to pass it through the mtd layer.  There is a reason we have this
> thing called devices drivers, and we don't want to add knowledge of
> ioremap to the filesystems and have users find out physical addresses
> of their flash to pass it as mount option.
I see your poing with regard to layering, and I agree to it.
In order to do ioremap/iounmap at least we need a counterpart 
put_xip_page thing to do iounmap in that path. Our dcss segments on 
390 did not raise that requirement: they had a permanent kernel mapping.
The quiz question to me is: why don't we establish a permanenet 
mapping of the entire thing from mount() to unmount(). That eliminates 
the need to do iomap/iounmap, eliminates the need to have 
put_xip_page, and eliminates to care about what layer would do this. 
Would work for cramfs, won't work for read+write flash filesystems.