CVS updates

[Xavier DEBREUIL]

Abraham vd Merwe a écrit :
> 
> Hi Erik!
> 
> > > I've added a standalone driver for the 28F160xx chips on LART. I've also
> > > removed the LART definitions in the sa1100 maps driver (it's wasn't used in
> > > anycase and it's obsolete now anyway). Same for the LART endian definitions
> > > in cfi_endian.h and the cfi chip driver Config file.
> >
> > Why? If I understood the MTD stuff properly it is layered with the chip
> > drivers at the lowest level, the maps driver on that, and the device
> > drivers on top of the maps driver. As far as I can see (and what my
> > plan was before you started) you only need a new chip driver for LART
> > and reuse the other parts. Because the LART is an sa1100 design we
> > should avoid special casing and just use the sa1100 maps driver.
> 
> The driver I wrote for LART is a standalone driver, so it doesn't need aby
> maps driver.
> 
> Originally I wanted to change the maps driver and just write a chip driver,
> but the sa1100 maps driver is dependant on CFI being configured. Since the
> 28F160xx chips isn't CFI compliant, it makes no sense to put the map
> definitions in there (It's only about 20 lines of code anyway: read/write
> routines and the partition definition), so I decided to just make a
> standalone driver.

Does it mean that I can not use do_map_probe("cfi", ...) with 28F160B3
chips ?

Xavier