[K3]+[DXG] Un-demo information

[geekmaster]

Recent PM from megacoupe indicates serial now working; not fried.

[knc1]

Originally Posted by geekmaster

Recent PM from megacoupe indicates serial now working; not fried.

The only explanation for that is non-full disclosure of the facts behind the situation.
We where reading "filtered" information on this thread.

A common problem when trying to troubleshoot by "remote control".

[megacoupe]

Originally Posted by knc1

I suspect you skipped that part of the procedure (programming the levels to 1.8v with the special Windows driver.)
That is probably how someone else could use the board and you couldn't - they did not skip the "program it to correct voltage levels" step.

It's not a question of skipping the step intentionally; this is my first time attempting a project of this magnitude. I had no idea such a driver was available, and if I did, I wouldn't know how to set it to the correct voltage.

The only explanation for that is non-full disclosure of the facts behind the situation.
We where reading "filtered" information on this thread.

A common problem when trying to troubleshoot by "remote control".

I mentioned in the other thread that I'm now working with a different demo unit DXG. I'm still not sure if the first one got fried (stangri tells me he's used the 5V connector on two DXG's without issue), but since this unit seems to be communicating, I'll work on it first before returning to the first one.

[knc1]

We just don't have the information available to know the full what and why.

Since the units are programmable, we can not assume that yours came with the same eeprom programming of someone else's.

Notice in the picture that it is sitting on a plastic bag with the word "AmazonKindle" on it. Which might imply that someone re-programmed it from the default 5v mode.

I can not tell from where I sit if yours has been re-programmed (or the one pictured has been).

[dpavlin]

Thanks to this thread I was able today to re-flash two DXGs in demo mode to normal Kindle firmware, so I decided to give something back and share a few tips we found long the way.

Since I do own normal rooted DXG with firmware 3.2.1 I decided to make images for kernel and rootfs as described in https://www.mobileread.com/forums/sho...d.php?t=190149

Than my friend soldered three wires and we connected them to Bus Pirate for reliable serial connection using 115200N1 settings.

I used yifanlu's KindleTool from github to create root password because various other implementation in python and javascript didn't work for me (YMMV).

I also double checked passwords using framework account and cat /etc/passwd which I then feed into jack the ripper using fiona world list to be sure that it was correct.

I already copied images to Kindle (still in demo mode) using USB mount, and on first kindle I wasn't fast enough to enter into recovery so I decided to dd content while demo mode was running. Kernel (first file) went OK, but while I was coping root filesystem (second file) serial port decided to turn off power on serial port so I got corrupted root filesystem.

I slapped myself because if you examine /etc/init.d/power on Kindle you will notice that you can create file /mnt/us/DO_NOT_KILL_UART which will prevent it or just use

echo 0 > /sys/devices/platform/mxcintuart.0/uart_clk_state

after you log in.

After entering in recovery mode (after few tries) I used 3. Load MMC0 over USB storage (using same root password generated for login) and than on laptop used dd to dump root filesystem to first partition of exported USB disk. Problem solved :-)

One device was registered on demo account, and although it did connect to 3G it returned that account was disabled in web browser. After de-registration and login with existing Amazon account we where able to access web and Amazon web store from device (so serial numbers from demo devices don't seem blacklisted in any way -- thanks Amazon, we really appreciate it).

Thanks again to everyone for all hints and this discussed here.

[Begemot]

Just wanted to add that I too have been successful in flashing my demo Kindle DX(replacement from the infamous bangss, since the first one was broken).

I first used stangri's method in OP but I made the mistake of plugging in usb cable half way through the copying (was worried about battery running low) and got some extra buffer errors, which bricked it.

My last hope was yifanlu's passwordless kernel: http://yifan.lu/2011/04/02/recoverin...rupt-kindle-2/

Using a copy of good old friend HyperTerminal from XP copied to Windows 7 ( I am sure minicom on Linux would work too) I transferred the kernel image to memory (takes about 3 minutes on 115200 =~10kB/s for 1.8MB) and was able to boot and then mount MMCO as USB (option 3 in development menu).

After that it was a simple dd of a rootimg.rfs (which I extracted from a Kindle DX a .bin that is "floating" around the net using yifanlu's KindleTools that I compiled on Linux) from my Lubuntu virtual machine to the usb (i think output was to of=/dev/sdb1 but make sure with df or mount command first, you want the 400MB partition as the destination)

The process was quite slow at around 200kB/s thus taking about 40 minutes for full 400MB, but it worked!

I now have my Demo Kindle DX running 3.2.1.

One small tip, if you have to use u-boot a lot, you should change enviroment settings for boot to 5 seconds instead of 1 second and also you can set the reboot from inactivity to something large like 3600 seconds, so you do not worry about typing too slow.

[geekmaster]

Originally Posted by megacoupe

Here's a link to the adapter I bought: http://www.ebay.com/itm/180953407401

And here's a pic from user stranger_st of the very same adapter:



I don't know if you can tell, but there appear to be 3 LED's on the board: one that lights up when plugged into a USB port, and two others labeled TX and RX (which I've yet to see lit up).

As far as I can tell, two other users here (stranger_st and stangri) have managed to use this adapter without making any modifications to it. What was their secret?!?

I ordered three of these ONEFROG things when you first posted this. They FINALLY arrived. But they have had a name change. Although the boards look the same, the "ONEFROG" printed in white ink on your photo now says "USB to TTL", and the gold label with the two arrows that says "ONEFROG" is still the same gold and arrows, but now says "ONESANYA". What's a "sanya"?

[knc1]

Chinese for "one brick" ?

- - - - -

That is the device whose output levels are programmable (it has flash on-chip for configuration).

Test what levels it is working at, you may have to get the programming application to reset it to 1.8volt.

[qlob]

Sanya is a chinese city....

[stangri]

Weird, I've ordered mine about a month or so ago and it still said onefrog, which if you think about it doesn't make any more sense than onesanya.

[seawhite]

Onefrog is a store on taobao, sells converters etc., the store is based in xi'an
http://onefrog.taobao.com/

Maybe now they have a branch in SanYa.

You can try the module out with a handheld old gps, which outputs rs232 signal to pc

[geekmaster]

Originally Posted by seawhite

Onefrog is a store on taobao, sells converters etc., the store is based in xi'an
http://onefrog.taobao.com/

Maybe now they have a branch in SanYa.

You can try the module out with a handheld old gps, which outputs rs232 signal to pc

Yeah, but my old serial GPS units need to be outside to see any satellites, and still take at least five minutes to acquire a position.

The new ones are incredibly sensitive, even indoors, and get a position quickly (and they are tiny and low power). My oldest GPS is the size (and weight) of a friggin' brick.

You can also test serial devices by connecting the TxD and RxD wires, and checking for a character echo.

[seawhite]

Originally Posted by geekmaster

Yeah, but my old serial GPS units need to be outside to see any satellites, and still take at least five minutes to acquire a position.

The new ones are incredibly sensitive, even indoors, and get a position quickly (and they are tiny and low power). My oldest GPS is the size (and weight) of a friggin' brick.

You can also test serial devices by connecting the TxD and RxD wires, and checking for a character echo.

You are right, that's a simpler way to check it!

I bought a 3 in 1 converter from ebay and was worry about damaging kindle if I setup the jumper wrong
http://www.ebay.com/itm/3-in-1-USB-t...item232445d5dc

I thought about GPS solution because years ago I built a rs232 to usb (actually Axim interface) cable using max232, to just connect the old "brick" GPS to Dell Axim. Now I don't need to dig out the GPS anymore

[suspirit]

Hi, I'm new to DXG un-demo and after reading the thread carefully I have some questions before working on the un-demo.

From Stangri's post, it seems that running "dd" command (flashing rootfs.img) was in normal serial console. However in dpavlin and Begemot's posts they all flash the kernel and go to recovery mode and mount MMCO as USB, and then flash rootfs.img. I found the latter was discussed in http://yifan.lu/2011/04/02/recoverin...rupt-kindle-2/. I'm not sure whether flashing the kernel, entering recovery mode and mounting MMC0 as USB are necessary? My guess is it's only necessary when the original root file system is corrupted (bricked). Could somebody confirm?


[QUOTE=stangri;2210861]

In the serial console I logged in with the login root and the password from the KindleDiagnosticTool.exe and ran the following command:
dd if=/mnt/us/rootfs.img of=/dev/mmcblk0p1

At the end it started throwing errors like "Buffer I/O error on device mmcblk0p1, logical block nnnnnnnnnnnnn" -- as I'm guessing when it could no longer load next page/image. I've had trouble trying to reboot at this point thro the command line but after reboot with the hardware power slide I've got the working Kindle DXG with the 2.5.8 firmware.

[stangri]

Either method worked for me (yep, I've un-demoed more than one DXG ).