Rockbox Project (M6)

Originally Posted by roelandb

so there is no risks to bricking the meizu m6 by doing anything wrong by this?

There are risk as any other attempt to change anything in a player.

It has to be a way to solve such errors, like a DFU program or anything that let's the users recover the player. Or simply good instructions on how to do the installation. This is up to Rockbox developers.

By personal experience if there is a way to recover my player nothing would stop me to try a new resource in my player...

great to hear about this progress! very good work guys!

yesterday i wanted to make a compile to try something by myself.
i've downloaded the VMplayer, the debian image and the current source. i got a few errors while compiling. the target seems to be little endian and some of the to linking objects seems to be big endian. i tried to change the option for the gcc from big to little endian (-mlittle-endian). but no luck.
can someone help me regarding this? maybe there is a patch for the arm missing (big endian patch)?
i dont have to much time, but i would like to try something to support the port to the meizu player.

cheers
jens

Originally Posted by Jensu

great to hear about this progress! very good work yesterday i wanted to make a compile to try something by myself.
i've downloaded the VMplayer, the debian image and the current source. i got a few errors while compiling. the target seems to be little endian and some of the to linking objects seems to be big endian. i tried to change the option for the gcc from big to little endian (-mlittle-endian). but no luck.
can someone help me regarding this? maybe there is a patch for the arm missing (big endian patch)?

Hi Jensu, I added big-endian support to the rockbox ARM cross-compiler when we started the port, but I guess the debian image hasn't been updated after that. You can build the cross-compiler by running tools/rockboxdev.sh from the rockbox source tree.

Good luck

thank you markun! the compiling of the bootloader was no problem after this.
the next problem i was running in was to compile the dfu tool. i got there a lot of errors (conficting/incomplete types undeclared definitions). do i need anything else what's not included in the rockbox source path?

Originally Posted by Jensu

thank you markun! the compiling of the bootloader was no problem after this.
the next problem i was running in was to compile the dfu tool. i got there a lot of errors (conficting/incomplete types undeclared definitions). do i need anything else what's not included in the rockbox source path?

This means that're step advanced... great!!!

Originally Posted by Jensu

thank you markun! the compiling of the bootloader was no problem after this.
the next problem i was running in was to compile the dfu tool. i got there a lot of errors (conficting/incomplete types undeclared definitions). do i need anything else what's not included in the rockbox source path?

Yes, I think you need to install libusb, but I'm not sure if it will work inside vmware.

Originally Posted by DaremoS

This means that're step advanced... great!!!

This is what will become the bootloader, it's not actually working yet

with the libusb installed i could make the dfu tool. thanks again.
but it seems that you are right with the vmplayer. the samsung usb device can be initialized successfully, but the first try with usb_control_msg() fails
i'm not sure if it could work this way. but i'm getting a valid usb handle...?!
what do you think markun?

cheers
jens
btw. i've a meizu M6 SL

We managed to read the touchpad and output the value by blinking the backlight (code is in SVN) Next will probably be the display.

Me and another rockbox dev also have M6SL's. Another dev has an M6SP and M3 I think.

Great! Keep up the good work!

this all sounds very promising, what you are doing now is making the drivers for the screen right? is there anything we can do to help on this?
could you make a video on youtube from the current progress, and post it here (when you find the time of course)

Originally Posted by markun

We managed to read the touchpad and output the value by blinking the backlight (code is in SVN) Next will probably be the display.

Me and another rockbox dev also have M6SL's. Another dev has an M6SP and M3 I think.

great! i will try it now! do you meen wpyh? he has m6 sp and m3.

I've have received my M6sl few days ago (after my m6sp have been stolen), I think I might help you on drivers developpement. I just need to find time to first finish another project, then I will focus on rockbox development. I'm so happy that rockbox will be available for the m6, when I bought it, the only bad point I saw is that I couldn't install rockbox on it. And as that this point will be removed, it will be the perfect player :D

++
LTourist

thank you for your offer, we can use all the help we can get to get the drivers to work, since it would be very nice to have gapless playback and lots of other improvements like no library updates.
Are the drivers different for the different types of players: sp/tp/sl ?

You need different Firmwares, so In guess the drivers are different, too.
But who knows, maybe they are similar.

ofcourse the drivers are not different. we need to make separate parts of prt gor sp/sl/tp from that moment

Most important drivers to write are display, NAND and USB I think. We already have drivers for our FM and audio chips from other ports. The NAND driver might be similar to one in one of the other ports being worked on and the USB driver will be the same as the one for the Cowon D2.

Hopefully we can report some progress on the display driver this week.

I found this, which seems to be a driver for a S6D0129 (look at the bottom of the page). Obviously, the page seems to be in chinese or japanese, but maybe the code can be usefull :D

lcd_driver_step_by (based_samsung_s6 ¿ª·¢¹¤¾ß£ºads1.2 Ö÷ÒªIC£ºMCU£ºlpc2210(NXP) LCDÇý¶¯¼°¿ØÖÆ s6d012( ) Embeded-SCM Develop ǶÈëʽ/µ¥Æ¬»ú±à³Ì - www.pudn.com

EDIT : found a complete driver :
http://ww1.microchip.com/downloads/e...ry%20v1.52.zip

In Microchip\Drivers\...
Maybe we can take some part of the code, cant we ?

good find, i think we can. the second driver seems to be full. we can use parts of it

Originally Posted by vcf

good find, i think we can. the second driver seems to be full. we can use parts of it

No we can't. There is such a thing as licensing

You could at least read the code and see how it's working and then make the driver easily, no ?