SanDisk Sansa c200v2, m200v4, clipv1, clipv2, clip+, and fuzev2

[willwolfe]

Don't forget the major bug remaining : we can't access data past 1GB on the internal storage.

I think the emphasis at this time should be on the "internal" storage.

[LambdaCalculus]

Just a reminder for everyone that we're starting to veer off-topic on this thread. Let's get back to the business of porting, shall we?

[kugel.]

I can confirm the Fuze buttons do not work in the main Rockbox.

More hacking is needed to make it work. Enabling button_int() for the main build in button-fuze.c does NOT give anything but a highly flickering screen and unresponsive.

Funnily enough, the screen actually reacts on your button presses, e.g. pressing up will make it invert the color, pressing center will make it flicker in a slower fashion.

Anyway, getting the buttons to work (and fixing the lcd) is not trivial, at least nor for me. I played around with irq, without success, although interrupt issues seem logical to me, since lcd and buttons are accessing the same GPIO port (lcd via switched to DBOP).
So I tried disable_irq and enable_irq, which leads to a non booting Rockbox, giving a "Divide by Zero" error at the boot for a split second (it's so fast you can't note down the address). Using disable_irq_save() and restore_irq() didn't seem to do anything.

Apart from the buttons, the lcd ain't stable. Even without buttons enabled, the display flickers frequently.

Edit: The button driver seems to crash when booting in the main. I replaced the button read function with something that toggles the buttonlight everytime that function is called. It just blinks once!
Then I placed a splashf which displays an increasing counter (which is increased everytime the funtion is called). It only counted upto 3.

Edit2: Scratch the crashing button driver. It's not crashing, that was just an issue with xpd.
But I still can't get the buttons to work in the main binary (always keep in mind: They work just fine in the bootloader).

I tried to read them seperately. I can read the Power and Down button pretty well, but reading e.g. center or right doesn't work at all. The code is the same, just the corresponding pin replaced. And they behave differently.
I'm out of ideas very soon. I'm now trying to get the button via DBOP since GPIOB and C are kinda connected to it by a physical pin or something, but I'm probably out of hope here.

Sadly, the OF doesn't reveal much at all about the buttons (at least not in my disassembly of the firmware block). I can hardly find mentions of GPIOB and GPIOC. Does anyone have an idea if the button stuff is another than the firmware block?

It would be great if I could get help on this one. Also, keep in mind, that this problems afaik are also on the e200v2 (not Fuze-only).

[sko]

Hi, I have an e200v2 and I'm watching this thread and the SVN since some weeks. I am looking at the SVN and testing the changes nearly every day.

Also, keep in mind, that this problems afaik are also on the e200v2 (not Fuze-only)

Yes, thats right, the buttons are working in the bootloader (at least Buttons for choosing the OF or RB), but in the Rockbox menu they aren't working.

Second thing (it hasn't been reported in this thread until now): since revision 19234 there is only a blank white screen on startup, no rockbox logo, no rockbox menu, but booting OF with the <<-button works, so the bootloader must be working I think.

EDIT: The issue with the white screen seems to be fixed with revision 19276.

[funman]

Hello all!

I wanted to tell you that since today we have sound working !

The code isn't committed yet, because it required quite some changes and we still experience a lot of unexplained deadlocks/reboots/shutdown etc .. but sometimes we can hear our songs

See FS#9592 for work in progress

@johnnyz86 : like saratoga told you there is still a lot of work to do (he only forgot the FM radio).

Something you'll need to work on the Sansa AMS is the AS3525 datasheet.
Unfortunately it's not redistributable, but if you come on irc (and prove you are interested) you can ask Bagder for it (he is allowed to redistribute it to developers on an individual basis, at the condition to not redistribute / make it public of course).

[funman]

Hello, some time has elapsed without updates, so I'll give you the status on the SansaV2:

ROCKBOX DOES NOT WORK YET not even for casual use; so really there is no point in using it if you're not going to develop and make it better.

A lot of people have been asking for status recently, i hope it's made clear.
For future reference, you can track the status here

Recently we have been working on the FM radio, and it has been tested on M200v4 and Clip (radio in Fuze and e200v2 should 'just work' when we'll be able to test it)

Here is what is missing at this day:

• BUTTONS on e200v2 and Fuze (work in progress by kugel and daniel_at)
• STORAGE : we can access fully 1GB and 2GB models, but we can only access up to 1GB on >2GB devices
• RECORDING
• USB : could be pretty hard since we do not have any form of documentation, so long and tedious reverse engineering might be our only way
• BUFFERING : sound has only been tested on target with buttons (and which happen to only have 2MB of SDRAM) and random crashes appear, maybe due to buffering since applying FS#9332 helps a lot => needs to be investigated
• PLUGINS : Fuze and m200v4 miss keymaps in the plugins. Note if you are not a programmer but know how to build rockbox you can definitely help here by testing your work in a simulator !
• MANUAL : Rockbox proudly ships with a very decent manual, and the Sansa AMS shall make no exception, the manual should list what are the specific issues/capabilities on these models
• KEYMAP : The e200v2 uses the same keymap than e200v1, but other models will need reflexion on a good keymap to use. Note : this is a point where non programmers can help too
• ART : some plugins might need screen-size specific bitmaps and/or a new wps
• REVERSE ENGINEERING : The Clipv2, c200v2, and Fuzev2 (announced a short time ago on Sandisk Sansa forums) have currently no code in SVN because we need to figure at least one button for dual boot to work. If you are willing to modify a few lines of assembly code and risk definitely bricking your device you can step up here.
• RBUTIL : mkamsboot should be integrated into rbutilqt, and a howto for manual install written in the wiki

[Hillshum]

Manual is another thing non-coders can write

[battlemac]

when you are referring to storage are you referring to the built in storage or are you referring to the SD card storage?

[gabe565]

It is the internal storage with the no more than 1GB problem.

[Hillshum]

is that 1GB as in 10⁹ or 1024³?

[RockRabbit]

Hello, some time has elapsed without updates, so I'll give you the status on the SansaV2:

• KEYMAP : The e200v2 uses the same keymap than e200v1, but other models will need reflexion on a good keymap to use. Note : this is a point where non programmers can help too

I might be able to help, provided I am given clear instructions what to do. I have been able to compile Rockbox, and am running it on a c250 version 1 (its great by the way - though of course you all know that already). I have a version 2 I might be willing to risk, providing I don't have to guess my way thru any of it.

I am willing to bet the player on the fact that the keymap is more or less the same as the e200 (excepting the wheel?). Perhaps someone can supply code that specifies the most obvious button that will be the same (the menu button ?).

[funman]

@ hillshum : I don't know exactly since I don't own such a device, but I record the reported size was a bit less than 1000000000 (so less than 1GB AND 1GiB)

@RockRabbit : you must first revert utils/AMS/hacking to revision 18663 to get the old basic mkamsboot (which inserts a delay)
Then if you are not familiar with assembly, you should come on the IRC channel and see if someone (like me if i'm present) can give you the steps

[Llorean]

Alright, finally got my Clip. I'm looking at the keymap now, and I'll probably post two patches later (in the next few days) one "low-controversy" one where I try to fix the real problems, and a more controversial one where I try to make available the full functionality of Rockbox while still being what *I* consider usable (which may not be the same for everyone).

Alongside that, I'll try to look at the m200v4's keymap as well. I suspect though that it should just have the same keymap as the c200 in 95% of cases since the Recording button isn't often used, I think.

As a note, sound doesn't seem to be working again. At least, when I attempt to play something, it says at 0:00 on the Clip, and doesn't progress, this is with r19386

[funman]

Alright, finally got my Clip. I'm looking at the keymap now, and I'll probably post two patches later

Nice!

As a note, sound doesn't seem to be working again. At least, when I attempt to play something, it says at 0:00 on the Clip, and doesn't progress, this is with r19386

Playback is quite buggy on the Clip/m200v4; you should try to apply FS#9332 (flash buffering) which gives better results.
Without it rockbox crashes because of corrupted data (we suspect that buffering.c has some overflows since replacing this file gives better results; but we are not sure yet)

You can also try FS#9611 which is FM radio support (The FM output is directly mixed on the headphones output, so it doesn't use the pcm driver)

[smdas]

I just got a brand new AMS Sansa e260 in the mail today. I've done a bit cross-arch embedded before, and I'd like to see that RB boot screen come up on my Sansa...

I've got a small amount of RB experience as a user, in that I own a fourth-generation ipod that was given to me (which was broken, and I fixed). I've been using RB on that for a bit over two years, but that port is prone to frequent lock-up/crashes, which is super annoying. I was inclined to help there, but I really don't like the Apple hardware that much, so, I went and bought a Sansa, but it's the second-gen AMS-based hardware... so I figured that I aught to finally give back.

I have a bit of firmware experience bringing up audio hardware (bootloader work, initialization & driving of DACs/ADCs, loading DSPs but not coding them, etc), and working with various busses. Is there any work in those areas that I could help with? I don't have a micro sd/sdhc card, so probably not there. Perhaps I2C to the buttons? I dunno 

At any rate, my question is, has anyone had any further progress with rescuing bricked non-e200's, JTAG or otherwise? If I kill my Sansa then that will basically end my ability to help develop in a hurry. If there is a (reasonably) safe way to contribute to the firmware, then I'd love to do so.