Rockbox

This is the scan I made of my fuze for the simulator. Quite bad quality, so I decided to not use it for now.

   sansafuze.png (842 KiB)

You’ll need to modify also mkamsboot (add a dummy case for the Fuze since A3 is the only button we know and it is checked anyway)

I think lcd file can be empty for now since according to disassembly the controller isn’t the same than the e200v2.
Also you have put code for a button wheel (maybe from e200v1?) : it should be left empty since the fuze, like its little brother the clip, has no wheel ;)

For the picture, I mentioned it on irc: a good picture for the Clip would serve also for the Fuze (and the reverse is true).

How come you think the Fuze doesn’t have a wheel? It has a wheel, and it’s turning around nicely. I haven’t modified mkamsboot since I’m not as up to the progress as you, so it’d probably better if you just insert something before committing.

If you do so, implement something which I can use to find out the button light gpio pin :)

Sorry, I only know the fuze by pictures, and I supposed the similarity with the clip would extend to the buttons (aka no wheel).
Here is a patch to mkamsboot.

To find the button light pin (if it really is on a gpio), you can do something like that in the bootloader’s main():

GPIOA_DIR = 0xff; /* all pins output */
GPIOA_PIN(0) = (1«0);
GPIOA_PIN(1) = (1«1);

etc .. and try pins one by one

   mkamsboot-fuze.diff (3.6 KiB)

Well, trying pins one by one isn’t hard. I’ve could done that myself :) I wasn’t clear but I searched for a loop that tries every pin on A(B,C,D) every second or so :)

int delay;

GPIOA_DIR = 0xff;

GPIOA_PIN(0) = (1«0);
delay = 0×100000;
while(delay–);

GPIOA_PIN(1) = (1«1);
delay = 0×100000;
while(delay–);

..

I can’t think of something else, especially if you have no output at all, I think each pin one by one is the best solution (there is “only” 32 of them)

This patch fixes some of linuxstb’s work. Correcting firmware/sources and button-target.h.

It still adds a preliminary version of a simulator.

   fuze-fixes-sim.diff (22.4 KiB)

Working buttonlight, backlight and lcd.

The meaning of the GPIO D7 (supposedly buttonlight pin) needs to be investigated though. Backlight didn’t work without that pin. Possibly lcd didn’t too, but I can’t tell if the lcd is working without backlight.

   fuze-fixes-sim-lcd.diff (27.2 KiB)

update: _buttonlight_on() isn’t actually needed.

Now with correct colors and a loop which shows brightness level and buttonlight change is working. Also commented storage_init() out, since it’s broken.

   fuze-fixes-sim-lcd.diff (28.9 KiB)

Ok, scratch that patch. svn diff doesn’t recognize if I “svn mv’d” a file…