Cowon D2

[andiator]

That's strange, the rev. 20615 works well on my D2 (3.59). Did you try if my build (20603) works?
I'm using the debian virtual machine, downloaded somewhere here, for build.

mktccboot says:
2.59:
  original entry point: 0x2000006c
  new entry point: 0x203204fc
3.57:
  original: 0x000068
  new: 0x2037b658

Greets,
  Andreas

[markV]

Hi Andi,

Here both OF (2.59 & 3.57) give:
original firmware entry point: 0xe59f00ec
New entry point: 0x2001afd4

I'm on a complete Linux system (Mandriva 2008).
Maybe I should check my make system ? (whatever that means ?)

And here, the rockbox-info.txt output:
Target: cowond2
Target id: 34
Target define: -DCOWON_D2
Memory: 32
CPU: arm
Manufacturer: tcc780x
Version: r20615-090405
Binary: rockbox.d2
Binary size: 110540
Actual size: 110540
RAM usage: 570006100
gcc: arm-elf-gcc (GCC) 4.0.3
ld: GNU ar 2.16.1
Host gcc: gcc (GCC) 4.2.2 20071128 (prerelease) (4.2.2-3.1mdv2008.0)
Host system: Linux

My D2 seems (hard to say) to shut off rapidly in the firmware upgrade process...?

Thanks,
MarkV

[andiator]

Do you use the right d2n.bin? from the directory '1' ?

[markV]

Hi Andiator,

Strange was the key word in your preceding reply... 
When it's strange it is usually due to a fool's mistake !
Well I'm a fool: I inverted the arguments on the mktccboot line !    I'm sorry...

Thank you very much for your patience,
MarkV

[andiator]

No problem,
I'm glad it's working again. 

And you don't want to know what stupid bugs I do always hide im my own code   

[Rotareneg]

According to the wiki, the lack of a USB connection to dump the raw flash data is one thing that's holding up a proper flash driver. Has the regular audio output been considered? The player seems to have enough RAM and processor power to allow enough compression to reduce the amount of data sent to a manageable level (at least if you make sure that any unallocated space is cleared, either by formatting or adding big empty files.) You'd have to write software on the PC side to receive the data as well, but altogether it would seem to be a relatively simple process?

[gevaerts]

According to the wiki, the lack of a USB connection to dump the raw flash data is one thing that's holding up a proper flash driver. Has the regular audio output been considered? The player seems to have enough RAM and processor power to allow enough compression to reduce the amount of data sent to a manageable level (at least if you make sure that any unallocated space is cleared, either by formatting or adding big empty files.) You'd have to write software on the PC side to receive the data as well, but altogether it would seem to be a relatively simple process?

Does the wiki still say that? If so, it's outdated. Shotofadds has used USB to get at the data.

[Rotareneg]

Yep,

Investigate and implement a software USB stack (the ability to dump raw NAND data to a PC would help item 1 considerably).

[shotofadds]

I'll fix the wiki page - the D2 has had semi-working software USB for some time now, although it's disabled in the main builds. In fact, it's the main reason that reading from the internal flash is as reliable as it is today.

The wiki page could really do with a bit of general updating - at the moment it doesn't even have a status table of what works and what doesn't.

[grantmasterflash]

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.

There must be some boot issue with Rockbox on D2's that have been upgraded to D2+ firmware, cause I'm having this same  "*PANIC* Max NAND write caches reached" problem referred to by Rapide_23.  I also notice a few people have posted in the D2 Rockbox iAudiophile forum with this issue on D2->D2+ (for example, http://iaudiophile.net/forums/showpost.php?p=230625&postcount=6). 

I used to play around with Rockbox when I was on D2 3.57 with no boot problems (never experienced this PANIC message), but since flashing my D2 to D2+ 3.02, I can no longer boot Rockbox because I always see this PANIC message.

I've tried this with Andiator's 3.02 Rockbox bootloader available from his ftp site (2009.04.01 rev20603) and with a bootloader I compiled myself earlier today with the daily build (090425).

I'll be glad to help with debugging this issue on my device if anyone can tell me what is needed.  I'm not completely new to using/installing Rockbox, but I am new to debugging it.

[shotofadds]

I'm still trying to work out why upgrading to the D2+ firmware causes these problems for some people.

The D2+ firmware doesn't seem to use a new version of the Telechips NAND driver (both D2+ x.02 and D2 x.58 contain the identifer string "SIGBYAHONG_TNFTL_TCC78X_V000474", whereas this version number did change slightly in some older versions) - so there shouldn't be any difference as far as Rockbox is concerned.

The "obvious" thing might just be to try upping the MAX_WRITE_CACHES value in firmware\target\arm\ata-nand-telechips.c (eg. from 8 to 16), and then re-compiling the bootloader. I can't see why that should be necessary, though.

If that still doesn't work, I'll need to write some code to print out some more debug information from your D2. Without being able to reproduce the problem myself or be able to retrieve data directly from your D2, this could be tricky.

[grantmasterflash]

The "obvious" thing might just be to try upping the MAX_WRITE_CACHES value in firmware\target\arm\ata-nand-telechips.c (eg. from 8 to 16), and then re-compiling the bootloader. I can't see why that should be necessary, though.

If that still doesn't work, I'll need to write some code to print out some more debug information from your D2. Without being able to reproduce the problem myself or be able to retrieve data directly from your D2, this could be tricky.

I upped MAX_WRITE_CACHES in firmware\target\arm\ata-nand-telechips.c from 8 to 16 and recompiled the bootloader.  Unfortunately, I still get the PANIC message.

If there's anything I can do to provide some debug, I'll be glad to help.  Just let me know.

[grantmasterflash]

Regarding these "*PANIC* Max NAND write caches reached" errors during Rockbox boot I've been receiving since flashing my D2 to the D2+ firmware:

I flashed back to D2 3.59 firmware, then flashed to D2+ 3.02 firmware, then put the latest daily Rockbox bootloader and zip build on my device.  Rockbox is now booting without a problem and I haven't seen these messages at all.  I don't know if the reflashing fixed this or if something has changed in the most recent Rockbox builds.

[shotofadds]

Well, that's interesting, but unfortunately it gives me nothing to go on in order to fix the problem. If you find a way of making these errors happen repeatably, do let me know...

[3amsleep]

I (really) hate to be such a nuisance, but I been looking every day here and in the trunk for the last year for any significant updates... I would really like to know if im just losing my time cause it seems to me that shotofadds lost interest in the project and though many people have offered help with the nand driver I've never seen any response from him regarding that.

is there anybody actually developing MMC support?, are there any NAND improvements not jet released?, shotofadds if you're just over the whole D2 thing, please just tell us so we can figure a way to keep the port developing.