> >> Does bootbox work yet?
> > Yes, it works, but I'd like to gain a bit more confidence with
> > it. Only a few people participated in the beta test, iirc not
> > all all models were covered (Studio and Ondio SP untested).
> Pls remember that I tested on all my devices, among which are
> both an Ondio SP and a Studio 10...
Another case of bad iirc, please forgive me old senile.
> We could round up here. Afair, the bootloader determines the
> start of the second image from the start & length of the first
> one. Simply padding the first image a little with zero bytes to
> leave a bit more room should be enough. This would also have the
> advantage that we could use the same start address for the
> second image on all models.
The padding will have to be hand-adjusted for each model, to reach the same
destination. Plus bootbox is compressed, so we need a non-trivial padding.
> I'd say releasing it in parallel would be a good idea. The
> feature freeze phase should help weeding out bugs.
> Of course rockbox_flash should get the start address check
> added we already talked about. It will work fine as long as we
> don't change the linking convention, i.e. the startup code is
> located directly behind the vector table. Adding this check can
> happen before the freeze. This will prevent flashing a rombox
> image with the wrong start address.
I'll add the check, if I find a free minute.
> Release 2.5 rombox should then be built with the new rom_start
> address. Bootbox for the flash packages would be built from the
> same code base. Immediately after that the flash packages could
> be put together, tested & released.
5 GB Mailbox, 50 FreeSMS http://www.gmx.net/de/go/promail
+++ GMX - die erste Adresse für Mail, Message, More +++
Received on Thu Aug 11 12:55:14 2005