Cowon D2

[Rapide_23]

I just flashed my D2 to the D2+ firmware with rockbox.

However, when I try to boot rockbox, all I get is "*PANIC* Max NAND write caches reached".

I'm using a 16GB D2. Did I not install something properly, or is rockbox still not quiet ready for the 16GB?

Any help would be greatly appreciated.

[shotofadds]

Maybe the SS_CTRL value should be increased. It seems to be used to determine if the stations signal level allow stereo decoding. With the actual value, I can hear some noise with some stereo stations.
In the OF, the SS_CTRL seems to be higher than rockbox.

I always found it quite annoying that the OF seems to need a particularly strong signal before it will enable stereo. The current behaviour in Rockbox seems to match what I'd expect from a portable device, so I think we should leave it how it is (and there's always the Force Mono option if you prefer it...)

Is it alowed to ask whether there is any development on the sd card driver (reading would be nice and write extreamely awsome!)?

You can ask but you might not necessarily get an answer! There was some work done on an SD driver a while ago, but I'm not sure of its current status.

First (neophyte) question: I know Rockbox bootloader is patched in the original firmware with mktccboot, does that mean Rockbox is like a "parasite" application ?  I mean does Rockbox send commands to the hardware through the original software or directly to the hardware ?

Rockbox is a complete firmware replacement, so it doesn't use any part of the original firmware. The only reason we patch the original firmware at all is to provide the dual-boot function. You could potentially flash just Rockbox, without the Cowon firmware, but that would be rather silly at this stage...

Second question: I remember having seen somewhere on this site something about database corruption being related to a bad bootloader patching or something...  Is that what I'm experiencing on the "original firmware side" (complete database corruption, like 'hpfmn' wrote) or this database corruption is really just from the new Cowon 2.59 firmware ?

It's extremely unlikely that Rockbox could affect anything in the Cowon firmware. I've never seen any reports of this or "bad bootloader patching"...

Ok, now some observations I made (could this help) on my Rockboxed (v. r20457-090322) 8GB D2:
1- when the backlight goes to sleep there is still a faint (so faint !) amount of light emitted when using a white background theme (like iCatcher)
2- I still got the two-bottom-lines-at-top problem, like once in 20 boots or so (random for sure)
3- I have a weird problem when getting the power button back from the locked position: most of the times the player goes to a paused state but gets back to the main menu and can not resume playing.  A less frequent symptom is the player shutting down after a couple of clicks.

1. Yeah, I see that too. It's because the LCD isn't actually being switched off, only the backlight. This is intentional for some Rockbox targets, as the screen can still be read with no backlight. I should fix that, as it's wasting power having the LCD switched on all the time.

2. It looks like there's still a timing problem in the LCD init code, but it seems to happen much less often now. I need to study the OF code again to see how they do it.

3. This is a flaw in the keymapping. To get out of hold mode you need to slide the hold switch and then press another button to see the screen. If you slide the switch all the way over, it acts as a POWER key press - this is mapped to STOP, which will stop playback and exit to the menu. I'm not sure how best to solve that.

Hi, I am looking for pre-built mktccboot (with associated bootloader.bin). I would try Rockbox on my D2 and haven't develop's environment.

First of all, Rockbox isn't really ready for general use (mainly just for development / testing at the moment) so we haven't made easy installation tools available yet. I did intend to provide some pre-built tools to aid people testing builds, but I wanted to wait until the NAND code was more reliable. It's not ready yet...

I just flashed my D2 to the D2+ firmware with rockbox.  However, when I try to boot rockbox, all I get is "*PANIC* Max NAND write caches reached".

What version of the bootloader did you use to patch the firmware (or did you download someone else's pre-patched build)? There was a fix a few weeks ago (r20168) which might help.

[Rapide_23]

What version of the bootloader did you use to patch the firmware (or did you download someone else's pre-patched build)? There was a fix a few weeks ago (r20168) which might help.

I'm using the one on iaudiophile: http://www.iaudiophile.net/forums/showthread.php?t=26360 (The patched 3.02 in the download link at the bottom of the first post).

[shotofadds]

That's the problem when other people distribute builds - I don't know exactly what version it is, or what other changes it might contain. So it's pretty much impossible to debug.

Are you able to use tcctool? If so could you try a debug build I made a few weeks ago for someone with a similar problem. See here.

[Rapide_23]

That's the problem when other people distribute builds - I don't know exactly what version it is, or what other changes it might contain. So it's pretty much impossible to debug.

Are you able to use tcctool? If so could you try a debug build I made a few weeks ago for someone with a similar problem. See here.

Thats the problem. I don't know how to build a bootloader  :'(

[Yeeze]

@Rapide:
Here: http://www.rockbox.org/twiki/bin/view/Main/DocsIndex#For_Developersis everything you need. It takes a little while to read and to setup everything, but then you can always build the newest version and you are independent.

@shotofadds:

It's because the LCD isn't actually being switched off, only the backlight. This is intentional for some Rockbox targets, as the screen can still be read with no backlight. I should fix that, as it's wasting power having the LCD switched on all the time.

As far as I can see, the OF also only switches the backlight off. The D2's screen is hard to read without backlight, but still possible under the right light conditions.
Do you really think it drains that much battery?

[shotofadds]

Do you really think it drains that much battery?

The LCD itself probably isn't that much - the backlight will be a much more significant battery drain. Still, on some other players (eg. iPod Video) a nice boost in battery lifetime has been gained by switching off the LCD when not in use. Although if we power off the LCD, it does take some time to power on again, so the "user experience" might not be ideal...

On a related note, I recently had Rockbox playing music constantly for over 24 hours, so our battery usage at the moment isn't exactly terrible. I'm not really too concerned about power consumption at the moment - I think the D2 port has rather bigger issues to worry about than that.

ps. I haven't mentioned it here yet, but the "user timer" has now been implemented, so things like the Metronome plugin and Doom should now work properly.

[luck3r]

The OF only turns off the LCD 30 seconds (or so) after the backlight was switched off, but stops updating the screen, as far as I can see.

Edit: I can confirm that Doom works now, although the sound quality is low and the MIDI music doesn't play yet. Radio is also working fine.

Something else: The delay until Rockbox turns off the device is too long. The device would automatically shutdown after a few seconds if the power switch was held for >1s AFAIK. Maybe this can be prevented by the firmware?

[markV]

Hi

Is it just me or is the current version of the bootlader broken ?

I bricked my D2 three times (on OF 2.59 and 3.57) trying to get the latest svn version...
It usually works well compiling and patching ...

Thanks
MarV

[andiator]

Well, the revision 20603 works.
Did you run the configure tool?
Running make you get the message that your version is old and you should run configure, is it that case?

[Rapide_23]

Andiator, I downloaded the patched 3.02 file and rockbox build from your website, but I still get the NAND panic after flashing and trying to boot into rockbox....

[markV]

Hi,

@Andiator,

svn  tells me "revision 20615" (today).

I've done:
make clean
configure --target=33 --type=B
make,
I've had no problem.

mktccboot bootloader d2N.bin d2new.bin

Thanks,
MarkV

[andiator]

Rapide_23,
do you have the new D2+ or the old (reflashed) one?
My 3.59 works fine, so I think it's the old NAND driver problem.
You can try to remove and copy files from/to D2 internal memory and then start rockbox. It helps sometimes.
Or you can backup your files and format the internal memory. Then copy back the .rockbox directory and try again.


@MarkV
so does

I've had no problem.

mean it worked? Anyway I'm going to try the new revision.

[Rapide_23]

Mine is the old D2 flashed to the newer 3.02 firmware.

[markV]

@andiator,

I'm sorry; I meant the "make" went well, but the bootloader bricked my D2.

Is there a way to check the mktccboot process ?  Is the "entry point" output tells something about the "patching" success ?

Thanks again,
MarkV