Cowon D2

[ThibG]

Thanks. I'll take a look later and maybe see if there's a way to optimize that add/sub stuff. Out of interest, what fps do you get in test_fps (both at 48MHz and 192MHz)?

48Mhz:
  1/1: 19.2 fps
  1/4: 75.5 fps
192Mhz:
  1/1: 75.0 fps
  1/4: 291.0 fps

[andiator]

About the NAND, did anyone looked at TCC78xx_MultiMedia_Application_SDK_v044_release.zip linked in the post #96 on this thread?
Under project/FS there are a dma.c and tnftl.h (Telechips Nand Flash Translation Layer) found.

[shotofadds]

Even if it contained anything useful, there's no way we could use code from the SDK. Anything going into Rockbox must either be written from scratch or come from another GPL-compatible source.

In any case, it appears that the TNFTL stuff is supplied as a pre-compiled binary library, which wouldn't be any use anyway.

[El3ctroded]

There's no way to get data back from the D2 to the PC... When you're dealing with 4Gb+ of data, it's quite a painful experience.

That would be painful. But why is there no way to get data from the D2 to the PC? There's still the USB connection...

Another idea that I've been meaning to try is this: from the PC, write every sector on the UMS volume with a unique identifier based on its LBA address. Then, using the current Rockbox driver, read every sector to find which ones are read incorrectly. If we then search the NAND for the correct identifier, it might give us some clues about how the scheme really works.

Sounds like a good plan.

El3ctroded

[Dicky Dave]

A few days ago, I spent some time going over the code in the SDK to see if there were any clues in there. However, the most useful routines were supplied pre-compiled. objdump -D did a fairly good job of disassembling it but its pretty tough trying to work through it. My goal was to clarify the unknown fields that shotofadds identified so as to be understand the fs structure.

[sikon]

Hi Everybody! I have just installed the latest build of rockbox (r18451-080908) and i have a problem: the music not plays. I tried mp3 and flac formats. What happend?
PS: sorry for my English

[shotofadds]

This is a known problem - music does not start playing until the decoding buffer has been fully filled. This might take up to about 10 seconds, so just wait a little while.

If the music still doesn't play, it's probably because Rockbox cannot yet read reliably from the internal NAND filesystem. Try copying some more files to the D2 and then try again.

Both of these issues have been covered before in this thread, and are listed as Known Problems on the wiki page (but the wiki is down right now, so I'll let you off... ).

[FirleFanz]

because the touchscreendriver. I think what ever, it never can two buttons, at the same time. But do´nt feel sad, a idea to get control of that touchy thing. I have.  Maybe a controlpad on the miniusbport is an ability?!?

I have installed that Rockbox and it´s awesome - I can play PACMAN whoa, same time AC-DC.

Great work for all here!

And also - I think it must be a specification about the touchscreendriver!

• when you want select with LCD, there can´t be 3x3 Pad available
• The 3x3 pad is a must, change  SelectWithLCD is optional to implement or not to implement in the plugins
• a simple Layer that allows that a Plugin can work on different Touchscreens(not only the D2 has one)
• you can define areas(Shapes) in the layer and get an event, when touched(touch begin/end?)
• you get the x and y position of that area in percent (left/top=0% right/bottom=100%)

this allows that other touchscreens only nead a new Layer, and the Plugins are ok!

                                                               
sorry for my english, foreigner...  thanks for that chess!

[shotofadds]

If this thread has been quiet it just means there's been nothing significant to report for a while. Don't read anything more into it than that.

At the moment I'm concentrating on getting the filesystem driver working more reliably. I've said this 101 times already, but it's going to take some time to get that right. The buggy file access is the #1 most important problem right now, and is responsible for many of the bugs/weird behaviour seen in the D2 port (and is now also holding back the upcoming iAudio7 port).

For this reason bug reports aren't really very useful at the moment, unless you're certain they aren't related to unreliable file access (e.g. most problems during music/video playback are almost certainly filesystem-related). But you're certainly free to install it and try things out, if you're feeling brave!

[FirleFanz]

Rockbox 3.0 I´m right way to get 3.0 with SVN?

I have no idea what to do 

compiled with the Debian4VMwareImage, the following error ocurred. But also with the 2.x build it was the same! since yesterday 3.0 i tryed again.

Consoleoutput
.
.
.
LD rockbox.elf
.../arm-elf/bin/ld: ERROR: .Dir/shrink/by/me/libgcc.a( udivsi3.o) uses FPA instructions, whereas ..D/S/By/Me/rockbox.elf does not
.
goes on with a lot other Errors
.
.
.
.

[saratoga]

Rockbox 3.0 I´m right way to get 3.0 with SVN?

No if you do a standard checkout, you'll get the current SVN version, which is about 3 weeks newer then 3.0 as of today (3.0 is a fork from the end of August).  3.0 wasn't even released for the D2, so it doesn't really apply anyway.

compiled with the Debian4VMwareImage, the following error ocurred. But also with the 2.x build it was the same! since yesterday 3.0 i tryed again.

Consoleoutput
.
.
.
LD rockbox.elf
.../arm-elf/bin/ld: ERROR: .Dir/shrink/by/me/libgcc.a( udivsi3.o) uses FPA instructions, whereas ..D/S/By/Me/rockbox.elf does not
.
goes on with a lot other Errors
.

You don't have the proper gcc build for the D2's CPU, so compiling fails.  You should be able to fix it just by running the build script in the tools directory to build a new version of gcc.  Since the D2 is a new port, the vmware image probably hasn't been updated.  See the SimpleGuideToCompiling wiki page for info if you're not sure how to run the script.  Also, you should probably post a thread in Getting Started and Compiling forum if you have further trouble setting up your build tools rather then posting here.

[FirleFanz]

  - i will do so...

...ok i have Compiled the uptodate build and the Simulator now - awesome

my Steps
ubuntu installed
sudo apt-get install subversion
sudo apt-get install build-essential libc6-dev
sudo apt-get install patch
sudo apt-get install libsdl1.2-dev libsdl1.2debian

to conclude the LinuxSimpleGuideToCompiling form the docs index
http://www.rockbox.org/twiki/bin/view/Main/LinuxSimpleGuideToCompiling
and the SimpleGuideToCompiling
http://www.rockbox.org/twiki/bin/view/Main/SimpleGuideToCompiling
and the most problem was that ubuntu hides files in Nautilus! 

[FirleFanz]

about the D2 - Simulator

there is a picture in the sim ordner UI256.bmp



but when I start ruckboxui there is´nt the picture around the screen.
Or is it future release?

[karashata]

You have to run the background.bat file (for Windows simulators at least) to get the image to show up.  The regular rockboxui.exe doesn't automatically display the image.

Not sure about the linux simulators though...

[ikari_01]

Hi,

I am a little puzzled right now.
I've built rockbox for D2 from SVN as described at http://www.rockbox.org/twiki/bin/view/Main/CowonD2Info#Steps_to_Install
There were no errors along the way and tcctool does not complain about anything.
However, after uploading the bootloader, I get the message:

Code: [Select]

 "PANIC: unknown NAND: 0xEC 0xD7 0x55 0xB6 0x78" 

This seems illogical as nand_id.c contains an ID block that would match 0xEC 0xD7 0x55.

I've tried to force the detected NAND type using this:

Code: [Select]

return &all[0].info[3];

instead of the detected returncode.
Then, the bootloader seems to hang, it does not even boot the original firmware when the hold switch is off...

EDIT: disregard that, I accidentally mktccbooted the wrong binary. It boots now  (albeit with forced NAND detection.)