DJI Phantom 2 Vision FC200 EGG camera - Black video FPV solution to fix

[quaddamage]

mount: mount point /tmp does not exist

Well, do 'mkdir /tmp' before that command.

Also what prompter should I expect?

Doesn't matter, as long as there's "#" as that means you have root access.

The board starts once in a few trials, I pasted three power ups - two failed to bootup on third, it was succesful. On eevery powerup it shows DM36x initialization passed.
1. Does it mean that my bootloader is fine and my boot image is corrupted? (why it starts sometimes?

The issue is in rootfs, but I'm not sure what it is exactly. From the logs:

awk: /proc/devices: No such file or directory

The 'awk' editor is used to edit config of devices before they are accessible. This can have many reasons, but it looks like race condition - a programming bug introduced by DJI. We may be able to fix that, but we need the firmware dumped to know where the 'awk' is executed.

I am not sure how to stop decoding process, I've never managed to stop autoboot, I am only succesful with Ctrl+C, but like in example pasted below it seems that if I am doing it too often, the cli is blocked for 5 minutes

Wait, are you pressing ctrl+c during boot? Are you sure you're not doing that too early? You may be cancelling important actions during boot, press that only near finish, preferably when you see the encoder loaded.

 

[quaddamage]

I am wondering why I have two ramdisk mounted - maybe this could be a problem.

No, that's not it. No idea what the problem is though.

Looks like you did copied part of the partition, and did the ymodem command properly.

Maybe try different order, ie first putty menu then the terminal, or the other way..

 

[szupek]

Wait, are you pressing ctrl+c during boot? Are you sure you're not doing that too early? You may be cancelling important actions during boot, press that only near finish, preferably when you see the encoder loaded.

Yes, I am trying it in diffrent conditions, but more less from the beginning and I am rarely succesful in aborting.
Do you mean after this part?

LogicPD encoder initialized

 

[quaddamage]

Do you mean after this part?

More like after this:

irqk initialized
EDMAK module: built on Mar 24 2011 at 16:58:11

The issue you're having with transfer is unlikely to be related to boot though.. everything looks ok. It seem to be just an issue with ymodem transfer. You may want to mess a bit with ymodem. Try zmodem instead? change some options? order?

 

[szupek]

irqk initialized EDMAK module: built on Mar 24 2011 at 16:58:11

The problem is when I start too late, I can see this video encoding messages including resolution and asterisks, Ctrl+C does nothing. Can you please instruct if I should call top, and kill some process? Or execute commands regardles what is printed in the console?

******************************* pid of Xorg :1105 ==1106
*******************************************
Mem: 17320K used, 27004K free, 0K shrd, 0K buff, 4024K cached
CPU:   1% usr   8% sys   0% nic  88% idle   0% io   0% irq   0% sirq
Load average: 0.23 0.06 0.02 1/34 1110
*  PID  PPID USER     STAT   VSZ %MEM %CPU COMMAND
1066     1 root     S     2520   6%  10% ./read_proc
1110  1050 root     R     3072   7%   1% top
* 1105  1066 root     S <  34640  78%   0% encode -v hehe.264 -y 1 -r 320x240
1106  1105 root     S <   4316  10%   0% server
1064     1 root     S     4240  10%   0% /usr/local/sbin/sshd
1050  1049 root     S     3072   7%   0% -sh
* 1047     1 root     S     2880   6%   0% /bin/sh /sbin/autologin 115200 ttyS0
1049  1047 root     S     2540   6%   0% login -- root
1048     1 root     S     1908   4%   0% /sbin/getty 38400 tty1
    1     0 root     S     1652   4%   0% init [5]
*    6     2 root     SW       0   0%   0% [khelper]
    2     0 root     SW       0   0%   0% [kthreadd]
    3     2 root     SW       0   0%   0% [ksoftirqd/0]
    4     2 root     SW       0   0%   0% [watchdog/0]
    5     2 root     SW       0   0%   0% [events/0]
   10     2 root     SW       0   0%   0% [async/mgr]
*  143     2 root     SW       0   0%   0% [sync_supers]
  145     2 root     SW       0   0%   0% [bdi-default]
  147     2 root     SW       0   0%   0% [kblockd/0]
**********************************************************************************************video_interrupt= 0 ARM Load: 14% Video fps: 0 fps Video bit rate: 0 kbps Sound bit rate: 0 kbps Time: 00:00:18 Demo: Encode Display: D1 NTSC Video Codec: H.264 HP Resolution: 320x240 Sound Codec: N/A Sampling Freq

PID 1105, should I kill it?

 

[quaddamage]

he problem is when I start too late, I can see this video encoding messages including resolution and asterisks, Ctrl+C does nothing.

Then go the way you do now. As long as you can get to file transfer, it's ok.

 

[quaddamage]

PID 1105, should I kill it?

yes.
or 'pkill encode'.

 

[szupek]

After I killed it, It popped up with new PID

 

[quaddamage]

After I killed it, It popped up with new PID

also kill 'read_proc' and 'server'.

Tried running that?:

/opt/dji/bin/stop

 

[szupek]

Tried running that?:

root@Dji-weifs:/# cd /opt/dji/bin
-sh: cd: can't cd to /opt/dji/bin
root@Dji-weifs:/# cd opt
root@Dji-weifs:/opt# ls
ti
root@Dji-weifs:/opt# cd ti
root@Dji-weifs:/opt/ti# ls
cmemk.ko       encode         loadmodule-rc  read_proc      udtlog.txt
dsplinkk.ko    hehe5.264      null           server         version
root@Dji-weifs:/opt/ti#

 

[quaddamage]

I see.. anyway, that script is simple:

pkill server
pkill encode

That's it.

For the transfer - do 'lsz --help', look at the options. Try some.
You did tried to ctrl+c the tranfser and retry, right?

 

[szupek]

a board after restart, mount shows two mounted ramdisks, despite I have not done it after the restart - should I mount it again prior to dumping mtd3?

mdev on /dev type tmpfs (rw)
mdev on /dev type tmpfs (rw)
mdev on /dev type tmpfs (rw)
tmpfs on /mnt type tmpfs (rw,size=20m)
mdev on /dev type tmpfs (rw)
mdev on /dev type tmpfs (rw)
mdev on /dev type tmpfs (rw)
mdev on /dev type tmpfs (rw)
mdev on /dev type tmpfs (rw)
tmpfs on /mnt type tmpfs (rw,size=20m)
mdev on /dev type tmpfs (rw)
mdev on /dev type tmpfs (rw)
mdev on /dev type tmpfs (rw)
root@Dji-weifs:/opt/ti#

 

[quaddamage]

a board after restart, mount shows two mounted ramdisks, despite I have not done it after the restart - should I mount it again prior to dumping mtd3?

No, use one of the existing ones.

 

[szupek]

Somting's going on

 

[szupek]

it will take a while...
@quaddamage, taking into consideration that my board starts once in a few times, do you think it will be necessary to dump it more than once?

 

[quaddamage]

it will take a while...
@quaddamage, taking into consideration that my board starts once in a few times, do you think it will be necessary to dump it more than once?

I hope not.. but you are not pressing ctrl+c anymore, right? And it still doesn't always boot?

 

[szupek]

ok, just to clarify.
the way I got to the prompter was using Ctrl+C - not killing the encode and server processes.

I am just finished with first part, so I will be 30 minutes of a loss.
I just tried a reboot command and something like this popped up. Anyone knows what can be this root password for a maintenance?

root@Dji-weifs:/mnt# reboot
UBI: scrubbed PEB 223 (LEB 0:229), data moved to PEB 468
WARNING: could not determine runlevel - doing soft reboot
  (it's better to use shutdown instead of reboot from the command line)
INIT: Switching to runlevel: 6
INIT: Sending processes the TERM signal

Give root password for system maintenance
(or type Control-D for normal startup):

19881209 and 12341234 do not work

 

[quaddamage]

Wow, everyone is online today..

Just asked Mefisto, he says '123456'.

Note that you don't need to reboot after each part. Just use 'rm' (carefully!) to remove the mtd_part0, and go with next one.
I mean, unless the board is dangerously hot or something.

 

[szupek]

Yes, I know I do not have to reboot. Waiting for your reply I restarted the board manually I now I am in this video encoding loop, trying to kill processes.


pkill is not recognized kill -x PID seems to do the job, however regardless which one I kill, right after there are new ones with diffrent PIDs. maybe I have to create a script that will kill one right after another? Do you know if vi is available here?

Just asked Mefisto, he says '123456'.

do you think I should get back to this concept? Maybe this maintenance mode would disable encoding processes and load everything necessary for our dump. What do you think?

 

[szupek]

123456 or even '123456' do not seem to work.
Is this part (BI: scrubbed PEB 16 (LEB 0:26), data moved to PEB 604) a sign of memory autocorrection procedure?

root@Dji-weifs:~# reboot
WARNING: could not determine runlevel - doing soft reboot
  (it's better to use shutdown instead of reboot from the command line)
INIT: Switching to runlevel: 6
INIT: Sending processes the TERM signal

Give root password for system maintenance
(or type Control-D for normal startup):UBI: scrubbed PEB 16 (LEB 0:26), data moved to PEB 604
123456
-sh: 123456: not found
root@Dji-weifs:~#
Login incorrect
sulogin[1060]: Incorrect root password


Give root password for system maintenance
(or type Control-D for normal startup):123456
-sh: 123456: not found
root@Dji-weifs:~#
Login incorrect
sulogin[1060]: Incorrect root password


Give root password for system maintenance
(or type Control-D for normal startup):