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

[funman]

This needs to be done by the e200 testers.

I remember fragilematter has made some tests but don't remember which exactly

As for overwriting the OF, its not visible through usb interface so no problem, we just need to find its exact size on the nand to skip it when reading.

[atomikpunk]

Yes right, but would you mind updating (and reuploading) an updated version of your mkamsboot patch including this change? I think you hardcoded the 0x100 multiplier if I remember correctly...

And right for the filesystem, that could do the trick (though I still have my reserve).

I'll try the (new) patched mkamsboot once it is done on an e200 firmware and check its disassembly in case I see something wrong with it. Maybe we can coordinate with the e200 testers so we can talk "live" about it, maybe this weekend?

[funman]

Here is a diff to svn checkout.
It uses 0xff * 0x200 as the "rom block" size (use normal test.S with it)

E200 testers, anyone ?

EDIT: this week end I'm available, I'm on European time (UTC+2) but I can stay awake late; let's meet on #rockbox ?
But next week end I live the country where I live in and I'll have to look for a place to live and a job so I will hold on rockbox hacking :'(

EDIT2: I remove the patch since it doesn't work

[fragilematter]

Okay, just so that anyone is advised, the mkamsboot 0xff * 0x200 rom block size firmware bricks sansa e200s... try it at your own risk

[atomikpunk]

Bha damn... Thanks fragile_matter. I'll check the patch, create a patched firmware tonight and see what it looks like. If the firmware looks clean, well I'll give up god damn :'( (obviously on that firmware block size field topic, I won't give up that easily on the project itself )

But I really wonder how come there is a field for that (firmware block size) in the header but they wouldn't use it

[funman]

Maybe they use this one to flash the ROM (maybe you can write in ROM only by 512 bytes (0x200) sectors a a time ?

Also note that there is reference to past firmware, below 128k offsets, to initalise the stack.

The offset might be replicated somewhere else to handle library blocks loading ..
I'll look at finding the firmware size at other places in the file and report.

EDIT:
I removed the previous patch because it was broken.
Following patch uses 0x100 blocks (maximum), replaces in total 7 occurences of firmware size (instead of 2 previously), and update the firmware block checksum (what the previous patch didn't do).
I also added a check which verifies that the blocks we lose are padded with 0xff or 0xdeadbeef
Note: it seems that all the firmwares end with the same padding, except that some firmwares (like clip v29) are filled with random data and not recognizable padding.
XXX: patch confirmed to NOT work on E200 by fragilematter

[tobi-lu]

Hi!

Haven't been here for a while (shitty exams ), nice to see your great progress!

Well, I also read the wiki-page again and those missing graphs annoyed me once more. So I took the code and fed it to dot manually. You can see the results in the attachment. Perhaps someone can put these images into the wiki instead of the code?


cu
Tobi

[fragilematter]

Heya tobi-lu.
funman, atomikpunk and I will probably do a testing session on #rockbox in about an hour or two. Maybe you could join us

[daniel_at]

Hey all,

I've found some button-mappings for E200-Players. Thankfully there are some direct GPIO-wired buttons, so that it is easy to check for a bootloader. Especially if we have not much space for the code.

Up/Down/Left/Right/Center are connected to GPIOC and can therefore easily be checked.
http://www.rockbox.org/twiki/bin/view/Main/SansaV2HardwareMappings

To all the Firmwarehackers:
If you have some usefull functions/snippets which may be interesting for other hackers to play with, pls add them here: http://www.rockbox.org/twiki/bin/view/Main/SansaV2FWHacks

ciao
Daniel

[funman]

Daniel I added the known buttons for the Clip

I use a patched mkamsboot so the format of my assembly file(s)  are a bit different, do you want it to look at useful snippets ?
I think posting code on the wiki is a bit disorienting, we better should open a Git repository on repo.cz to share our code.

Any volunteer admin for the repository ?

[atomikpunk]

Well why not ask for svn access to the rockbox SVN repository?

[saratoga]

Usually people get SVN access by having a few patches accepted, at least that is the tradition.  The powers that be might make an exception for some people in this thread, but if theres going to be more then 1 or 2 people using it, running your own repo might be considerably easier since you could add as many people as you like to it.

[kugel.]

I agree, a repo where everyone here would have commit access is better, at least until we produce a major breakthrough.

Another note: I'll gladly try to find some buttons on my Fuze and update the SansaV2HardwareMappings page accordingly. I'd just need some short introduction on how this works. I'll be in IRC tomorrow, until evening, so it'd be nice if we can meet, for better support.

[daniel_at]

Okay - an external Repo sounds like an good idea. First i thought i set up one on my own server - but found something looking quite okay:

http://gitorious.org/projects/rockbox_sansa_v2

I have created the initial project. To join you have just to register a user on that site and than i can add you (i hope) - But be aware: I am a git-noob (but thankfull that i now have a reason to get known something different than SVN)

@kugel
We just have found a reliable way of unbricking a E200-device, so the testcript for finding buttons have to be carefully adopted, so that the OF can easily (and reliable) loaded if something went wrong....

Okay - sofar
Daniel

PS: fixed typo, thx Hillshum

[Hillshum]

Well, it seems I will not have time to keep the wiki up to date, but I will be able to find time to code/learn to code.

*puts nose in big book*