Channel Messages

[2017-08-12 02:49:07] hostile : @hostile has joined the channel

[2017-08-12 02:49:07] hostile : set the channel description: DJI Goggles !

[2017-08-12 02:50:43] hostile : to change the startup sound: $ wget "<https://archive.org/download/FlavorFlavsBestYeahBoy128KbpsAudioOnly/Flavor%20Flav's%20best%20Yeah%20Boy%20128%20kbps%20(Audio%20Only).m4a>" $ ffmpeg -i Flavor\ Flav\'s\ best\ Yeah\ Boy\ 128\ kbps\ $Audio\ Only$.m4a audio.wav $ adb shell "mount -o remount,rw /system" $ adb push audio.wav /system/gui/boot_animation $ adb shell reboot

[2017-08-12 03:04:55] hostile : ffplay -f h264 video.264 works on Goggles video

[2017-08-12 03:05:34] hostile : /system/bin/dji_animation is the startup animation it opens /system/gui/boot_animation/video.264 and /system/gui/boot_animation/audio.wav

[2017-08-12 04:04:51] hostile : This is close to the .264 video

[2017-08-12 04:05:12] hostile : $ wget <https://media.giphy.com/media/GkPlZwH59uUi4/giphy.mp4> $ ffmpeg -i giphy.mp4 -vf 'format=yuv420p' -s 1920x1080 -r 60 -colorspace bt709 -color_primaries bt709 -color_trc bt709 video2.264

[2017-08-12 04:23:46] hostile : $ ffprobe video.264 (original startup video)

[2017-08-12 04:23:54] hostile : Stream #0:0: Video: h264 (High), yuv420p(tv, bt709, progressive), 1920x1080, 60 fps, 60 tbr, 1200k tbn, 120 tbc

[2017-08-12 04:24:05] hostile : $ ffprobe video2.264

[2017-08-12 04:24:24] hostile : Stream #0:0: Video: h264 (High), yuv420p(tv, bt709, progressive), 1920x1080 [SAR 9:16 DAR 1:1], 60 fps, 60 tbr, 1200k tbn, 120 tbc

[2017-08-12 04:24:39] hostile : need to drop the '[SAR 9:16 DAR 1:1]'

[2017-08-12 21:52:06] asdrubale : @asdrubale has joined the channel

[2017-08-13 22:52:17] golumn69 : @golumn69 has joined the channel

[2017-08-14 09:32:29] andreasnofear : @andreasnofear has joined the channel

[2017-08-14 21:31:09] paoloone : @paoloone has joined the channel

[2017-08-14 22:46:43] nocommie : @nocommie has joined the channel

[2017-08-15 06:43:38] kilrah : @kilrah has joined the channel

[2017-08-16 20:31:35] tylkologin : @tylkologin has joined the channel

[2017-08-18 01:42:36] hotelzululima : @hotelzululima has joined the channel

[2017-08-18 10:03:50] d51 : @d51 has joined the channel

[2017-08-23 06:50:55] dmanisgnarly : @dmanisgnarly has joined the channel

[2017-08-23 11:02:28] rulppa : @rulppa has joined the channel

[2017-08-24 08:27:37] goof : @goof has joined the channel

[2017-08-29 12:45:39] e.vahdat : @e.vahdat has joined the channel

[2017-09-05 22:41:13] weprul0r : @weprul0r has joined the channel

[2017-09-07 21:00:06] chrarne : @chrarne has joined the channel

[2017-09-08 16:42:11] pure3d : @pure3d has joined the channel

[2017-09-09 03:42:33] jezzab : @jezzab has joined the channel

[2017-09-11 11:49:02] skywalk3r : @skywalk3r has joined the channel

[2017-09-12 05:02:20] jezzab : Enabled Debug Menu, Plane Match, Motion Mode, Device Info via removal of remarked lines in tool_setting_glasses_menu.xml

[2017-09-12 05:02:26] jezzab : Still seeing what they can do

[2017-09-12 05:30:47] jezzab : Screenshot seems to reboot it. Clear screenshot doesnt, just says "Screen shot cleared"

[2017-09-12 05:37:53] jezzab : cpdebug shows rows of cp uav cp ground cp rc but all 0.00's

[2017-09-12 13:23:57] hostile : <https://dji-rev.slack.com/archives/C5ZR0QXUG/p1505213819000067>

[2017-09-12 22:33:45] jezzab : root@wm220_dz_ah0001_v5:/system/bin # gui_debugger gui_debugger Usage: gui_debugger [cmd_id] [argument](optional)

[2017-09-12 22:34:02] jezzab : root@wm220_dz_ah0001_v5:/system/bin # gui_debugger get_input_focus gui_debugger get_input_focus command id 7 connect successfully win:top_menu_win|wgt:top_menu|idx:1

[2017-09-12 22:37:02] jezzab : commands: print_menu_state show_window get_input_focus enable_imu_control disable_imu_control print_profile_state hide_window dump_window log_level_control

[2017-09-12 22:44:48] jezzab : Stereo DAC Gain = System Volume/HDMI Volume (Master)

[2017-09-12 23:26:25] versuher : @versuher has joined the channel

[2017-09-13 08:56:24] jezzab : The dji_glass runs over the top of the background image in the frame buffer. If you kill the process it will clear leaving the bg image and then auto restart

[2017-09-13 08:56:51] jezzab : most of the df_xxxxxxx direct frame buffer examples fail to run

[2017-09-13 13:16:48] hostile : thx for sharing @jezzab I need to get back into mine and play some more!

[2017-09-13 13:21:59] jezzab : Sorry if it's all over the place I'm kinda using it as a place to brain storm ATM.

[2017-09-13 13:22:33] jezzab : I have compiled ffmpeg but it keeps skipping ffplay which I really want

[2017-09-13 13:22:39] hostile : nope... SAME here... if you scroll up =]

[2017-09-13 13:22:44] hostile : I think in real time and share...

[2017-09-13 13:22:52] hostile : for anyone that happens to be watching

[2017-09-13 13:23:13] jezzab : Want to see if I can get it to play video with sound

[2017-09-13 13:24:38] jezzab : Then see if I can script it or edit the GUI to play files off the sdcard via ffplay

[2017-09-13 13:25:43] jezzab : Need to checked again because debug was showing a /dev/sda1 mounted for recording all x264 stream

[2017-09-13 13:26:16] jezzab : Using a otg cable and USB stick didn't show a lot in dmesg

[2017-09-13 13:28:03] hostile : I was wondering if we could hang a USB stick off it and make the Spark cable free =]

[2017-09-13 13:28:14] hostile : it would take some love with wpa_supplicant

[2017-09-13 13:28:23] hostile : and perhaps some iptables rules but I bet it could be done

[2017-09-13 13:29:00] jezzab : Yeah. I have noticed the wpa key in the goggles but it's probably just cross over crap from the rest

[2017-09-13 13:29:09] jezzab : PSK sorry

[2017-09-13 13:30:08] jezzab : Would be cool

[2017-09-13 13:43:59] jezzab : Sticky tape on the two speakers helps not piss the rest of the house of at 2am as well. Wish the volume applied to the startup

[2017-09-14 10:41:34] jezzab : given up on trying to get ffplay to compile. Its ignored because of SDL requirements and I cannot get it to pull them in for some reason.

[2017-09-14 13:05:00] hostile : Did you compile SDL?

[2017-09-14 13:06:00] hostile : You can also try hacking an existing android arm ffplay binary from somewhere compiled static.

[2017-09-14 13:06:37] hostile : Similar to <https://www.androidtcpdump.com/android-tcpdump/downloads>

[2017-09-14 13:10:37] jezzab : looked everywhere

[2017-09-14 13:10:42] jezzab : there is no ffplay for arm

[2017-09-14 13:11:00] jezzab : ffprobe, ffmpeg etc but no ffplay

[2017-09-14 13:13:12] hostile : <https://play.google.com/store/apps/details?id=com.AndroidA.OnlinePlayer&hl=en>

[2017-09-14 13:13:16] hostile : it is inside this apk?

[2017-09-14 13:13:56] hostile : <https://apkpure.com/ffmpeg-cli/org.magiclen.ffmpeg.cli>

[2017-09-14 13:14:12] jezzab : I played a bit with the whole apk thing. But its comes to that I think

[2017-09-14 13:14:42] hostile : <https://apkpure.com/ffmpeg-cmd/com.kkomanjell.ffmpeg.cmd>

[2017-09-14 13:14:46] hostile : extract em

[2017-09-14 13:14:48] hostile : jack the binary

[2017-09-14 13:14:49] hostile : fuck it

[2017-09-14 13:14:50] hostile : =]

[2017-09-14 13:14:51] jezzab : ffmpeg is fine. I can run that and ive compiled it

[2017-09-14 13:15:07] jezzab : and ffprobe, etc

[2017-09-14 13:15:13] hostile : ONE of those packages has to have ffplay =]

[2017-09-14 13:15:22] hostile : anPlayer is supposed to

[2017-09-14 13:15:27] jezzab : tell me about it. Its my arch nemesis for the moment

[2017-09-14 13:15:44] jezzab : driving me nuts and if I compile ffmpeg one more god damn time

[2017-09-14 13:15:57] hostile : did you hand compile sdl?

[2017-09-14 13:16:10] hostile : when you ./configure it should tell you it can or can't find those libraries

[2017-09-14 13:16:15] jezzab : it wont pull it in. Everything is all good.

[2017-09-14 13:16:41] jezzab : Even guys just compiling it on plain old x86/64 linux have dramas

[2017-09-14 13:17:20] jezzab : I just want a simple CLI media player that plays video and audio and I can point both at X device and Y device

[2017-09-14 13:17:28] jezzab : surely it cant be that hard lol.

[2017-09-14 13:18:09] jezzab : Ive stepped back from it until tomorrow arvo. Look at it again with fresh eyes but another app is on the cards

[2017-09-14 13:18:20] jezzab : I even tried to get a static bin of omxplayer

[2017-09-14 13:18:25] jezzab : no go

[2017-09-14 13:22:47] hostile : aye

[2017-09-14 13:33:24] jezzab : Might grab another rpi and install KitKat. Cross compiling is hell.

[2017-09-18 03:51:21] jezzab : Dammit. Chrome Cast doesnt display right on the Goggles. Cuts out

[2017-09-19 09:28:55] jezzab : Im pretty sure they are running DISKO library for the GUI. Called from dji_glasses

[2017-09-19 09:29:04] jezzab : <http://www.diskohq.com/>

[2017-09-19 10:18:05] jezzab : 100% they are. And using GStreamer lib for video but it doesn't look like they use xine

[2017-09-19 13:55:05] hostile : I really need to dust my goggles off and get back to them!

[2017-09-19 16:10:48] unusuario128 : Definitely, the NDK toolchain is evil.

[2017-09-19 16:34:30] unusuario128 : ...And SDL is the greater odd job that I've seen in years.

[2017-09-19 16:35:27] unusuario128 : `#define main SDL_main`

[2017-09-19 16:37:01] unusuario128 : The SDL_main problem with this trick is that scrambles the object files in a way that they don't have entry point unless they are linked against libSDL. SDL_mainly adding defines like this one is really weird.

[2017-09-19 23:08:25] jezzab : so close now. Had to add a lot of android native libs as well @unusuario128

[2017-09-19 23:09:54] jezzab : I set display with: export DISPLAY=:0.0

[2017-09-20 00:09:41] hostile : finally got it!?

[2017-09-20 00:09:58] hostile : congrats @jezzab!

[2017-09-20 00:10:03] jezzab : Nah :disappointed:

[2017-09-20 00:10:09] jezzab : something wrong

[2017-09-20 00:10:11] jezzab : killing me here

[2017-09-20 00:10:16] hostile : xhost + ?

[2017-09-20 00:10:22] hostile : strace it

[2017-09-20 00:10:37] jezzab : its this whole SDL bullshit

[2017-09-20 00:10:41] jezzab : I dont know why we need it

[2017-09-20 00:11:52] jezzab : Thinking maybe I can make a hack up program to use the Disko lib

[2017-09-20 00:12:42] hostile : strace it

[2017-09-20 00:12:44] hostile : it will tell us

[2017-09-20 00:13:40] jezzab : ive never actually used strace :confused:

[2017-09-20 00:13:48] hostile : great time to learn!

[2017-09-20 00:13:56] jezzab : not wrong lol

[2017-09-20 00:14:13] hostile : strace /path/to/binary arguments here

[2017-09-20 00:14:26] hostile : or strace -p <pid>

[2017-09-20 00:14:35] hostile : best modifier flags "-vvv" really verbose

[2017-09-20 00:14:46] hostile : -o <filename>

[2017-09-20 00:14:51] hostile : ^--- saves to file

[2017-09-20 00:15:18] hostile : and "-ff" for follow forks and forks of forks

[2017-09-20 00:15:33] hostile : "-i" shows instruction pointer (good for writing exploits)

[2017-09-20 00:15:51] hostile : example... "strace -ivffp <pid>"

[2017-09-20 00:16:06] hostile : you can likely pull strace binary from your mavic... and use on goggles

[2017-09-20 00:16:08] hostile : no compile needed

[2017-09-20 00:16:36] jezzab : yup, -vvv is verbose alright lol

[2017-09-20 00:17:53] hostile : -s2000 to make the strings longer

[2017-09-20 00:18:04] hostile : they get abreviated at like 115 chars or something short

[2017-09-20 00:19:39] hostile : if you -o the file... I am happy to eyeball lit

[2017-09-20 00:19:50] hostile : note if you are following forks... you get filename.pid

[2017-09-20 00:19:53] hostile : for every pid in the fork

[2017-09-20 00:20:12] hostile : so may have outputfile.123, outputfile.127, outputfile.133, etc

[2017-09-20 00:39:34] jezzab : ok got it

[2017-09-20 00:41:11] jezzab : shit

[2017-09-20 00:41:15] jezzab : i see one problem

[2017-09-20 00:41:31] jezzab : i have symlinked the lib files btw. there isnt enough room on /system/lib

[2017-09-20 00:46:25] hostile : LD_LIBRARY_PATH should sort that out

[2017-09-20 00:46:28] hostile : put em where ya want

[2017-09-20 00:46:40] hostile : looks like default search path is /lib, then /vendor/lib, then /system/lib

[2017-09-20 00:46:47] jezzab : yup

[2017-09-20 00:47:00] hostile : futex(0xb6ce4d84, FUTEX_WAKE_PRIVATE, 2147483647) = 0 mprotect(0xb6f9d000, 4096, PROT_READ|PROT_WRITE) = 0 mprotect(0xb6f9d000, 4096, PROT_READ) = 0 mprotect(0xb6f9d000, 4096, PROT_READ|PROT_WRITE) = 0 mprotect(0xb6f9d000, 4096, PROT_READ) = 0 mprotect(0xb6f9d000, 4096, PROT_READ|PROT_WRITE) = 0 mprotect(0xb6f9d000, 4096, PROT_READ) = 0 mprotect(0xb6f9d000, 4096, PROT_READ|PROT_WRITE) = 0 mprotect(0xb6f9d000, 4096, PROT_READ) = 0 mprotect(0xb6f9d000, 4096, PROT_READ|PROT_WRITE) = 0 mprotect(0xb6f9d000, 4096, PROT_READ) = 0

[2017-09-20 00:47:14] hostile : @martinbogo previously mentioned that the kernel may not allow permissions to that memory

[2017-09-20 00:47:51] jezzab : mmmmm

[2017-09-20 00:48:21] hostile : lots of memory access failures

[2017-09-20 00:48:22] hostile : mmap2(NULL, 4096, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_ANONYMOUS, -1, 0) = 0xb6f9d000 madvise(0xb6f9d000, 4096, 0xc /* MADV_??? */) = -1 EINVAL (Invalid argument) mprotect(0xb6f9d000, 4096, PROT_READ) = 0 mprotect(0xb4c70000, 4092, PROT_READ|PROT_WRITE) = 0 mprotect(0xb5034000, 4092, PROT_READ|PROT_WRITE) = 0 mprotect(0xb55de000, 4092, PROT_READ|PROT_WRITE) = 0 mprotect(0xb5777000, 4092, PROT_READ|PROT_WRITE) = 0

[2017-09-20 00:50:35] jezzab : probably why no one seems to use SDL on arm lol

[2017-09-20 00:51:43] hostile : strace the applications that DO play...

[2017-09-20 00:51:49] hostile : just as a quick comparison

[2017-09-20 00:51:55] jezzab : ill do dji_animation

[2017-09-20 00:52:13] hostile : feeling better about strace yet?

[2017-09-20 00:52:18] hostile : it is pretty simple to grok

[2017-09-20 00:52:42] hostile : do we know the video card in this beast btw?

[2017-09-20 00:52:55] jezzab : yeah im getting the gist now

[2017-09-20 00:53:26] jezzab : I havent posted the dmesg log in here i dont think

[2017-09-20 00:55:07] hostile : can you add "-s2000"

[2017-09-20 00:55:11] jezzab : ok

[2017-09-20 00:55:12] hostile : so the strings don't get truncated

[2017-09-20 00:55:26] hostile : vfork() = 32744

[2017-09-20 00:55:30] hostile : also -ff

[2017-09-20 00:59:28] hostile : uio/uio0/ usage

[2017-09-20 01:00:52] hostile : close(13) = 0 close(15) = 0 writev(5, [{"\4", 1}, {"Hantro_omx_OSAL\0", 16}, {"Close ion! alloc 0xb6f76b2c ion_client 7\n\0", 42}], 3) = 59 close(7) = 0

[2017-09-20 01:01:00] hostile : see nothing like that in the ffplay

[2017-09-20 01:01:10] hostile : writev(5, [{"\4", 1}, {"Hantro_omx_OSAL\0", 16}, {"open ion success!alloc 0xb6f76b2c ion_client 7\n\0", 48}], 3) = 65 open("/dev/hx170dec", O_RDWR) = 12 open("/dev/on2psm", O_RDWR) = 13 ioctl(13, 0x6b03, 0xd) = 0 ioctl(13, 0x6b0a, 0x1) = 0 open("/dev/on2map", O_RDWR|O_DSYNC) = 14 open("/dev/graphics/fb1", O_RDWR) = 15

[2017-09-20 01:01:11] jezzab : yeah coz its doing SDL crap

[2017-09-20 01:01:40] jezzab : wonder if I can find a hantro example

[2017-09-20 01:01:54] hostile : that shit is popular

[2017-09-20 01:02:50] hostile : <https://github.com/legotheboss/YouTube-files/wiki/(RPi)-Compile-FFmpeg-with-the-OpenMAX-H.264-GPU-acceleration>

[2017-09-20 01:02:53] jezzab : i wonder if that is what they have done though. used the example which plays video only and then spliced in an auto decoder

[2017-09-20 01:03:03] jezzab : hance why they demux the a/v

[2017-09-20 01:03:13] hostile : --enable-omx on your compile perhaps

[2017-09-20 01:06:19] hostile : stat64("/data/dji/dump", {st_mode=S_IFDIR|0700, st_size=4096, ...}) = 0 open("/data/dji/dump", O_RDONLY|O_DIRECTORY) = 5 brk(0xb6f77000) = 0xb6f77000 getdents64(5, /* 15 entries */, 4200) = 464 futex(0xb6ce7350, FUTEX_WAIT_PRIVATE, 2, NULL) = 0 futex(0xb6ce7350, FUTEX_WAKE_PRIVATE, 1) = 0 lstat64("/data/dji/dump/host_0xe00", {st_mode=S_IFREG|0600, st_size=223875, ...}) = 0 lstat64("/data/dji/dump/host_0x1c01", {st_mode=S_IFREG|0600, st_size=1016135, ...}) = 0 lstat64("/data/dji/dump/host_0x1c03", {st_mode=S_IFREG|0600, st_size=52895, ...}) = 0 lstat64("/data/dji/dump/host_0x1c07", {st_mode=S_IFREG|0600, st_size=67151, ...}) = 0 lstat64("/data/dji/dump/host_0x0", {st_mode=S_IFREG|0600, st_size=9863, ...}) = 0 lstat64("/data/dji/dump/host_0x1f00", {st_mode=S_IFREG|0600, st_size=31511, ...}) = 0 lstat64("/data/dji/dump/host_0xa04", {st_mode=S_IFREG|0600, st_size=45679, ...}) = 0 lstat64("/data/dji/dump/host_0x1c00", {st_mode=S_IFREG|0600, st_size=213909, ...}) = 0 lstat64("/data/dji/dump/host_0x200", {st_mode=S_IFREG|0600, st_size=156625, ...}) = 0 lstat64("/data/dji/dump/host_0x1c02", {st_mode=S_IFREG|0600, st_size=21210, ...}) = 0 lstat64("/data/dji/dump/host_0x1c04", {st_mode=S_IFREG|0600, st_size=21210, ...}) = 0 lstat64("/data/dji/dump/host_0x1e00", {st_mode=S_IFREG|0600, st_size=5465, ...}) = 0 lstat64("/data/dji/dump/host_0x1c06", {st_mode=S_IFREG|0600, st_size=785504, ...}) = 0

[2017-09-20 01:06:20] hostile : that is weird

[2017-09-20 01:06:26] jezzab : <https://github.com/linux4sam/gst1-hantro-g1>

[2017-09-20 01:06:28] hostile : can you mkdir one of those file paths?

[2017-09-20 01:06:37] hostile : I wonder if that shit is for debugging of dsuss shit

[2017-09-20 01:06:42] jezzab : give it a crack

[2017-09-20 01:06:51] hostile : dont' have my shit handy

[2017-09-20 01:07:15] jezzab : they already exist

[2017-09-20 01:07:27] hostile : anything in em?

[2017-09-20 01:07:45] jezzab : pulling them down

[2017-09-20 01:09:39] hostile : this is the Hantaro... <http://www.verisilicon.com/IPPortfolio_14_58_2_HantroG1.html>

[2017-09-20 01:10:55] jezzab : Nothing exciting in them

[2017-09-20 01:11:28] hostile : <https://news.ycombinator.com/item?id=7811327>

[2017-09-20 01:11:37] hostile : more mention of Rockchip and hantaro shit

[2017-09-20 01:11:55] jezzab : libOMX.hantro.H1.video.encoder.so

[2017-09-20 01:12:12] hostile : read(21, "", 4096) = 0 writev(5, [{"\4", 1}, {"Hantro_omx_OSAL\0", 16}, {"OSAL_AllocatorFreeMem line 482 bus_data 0xb49c3040 bus_address 0x1acfc040 size 3657664 \0", 88}], 3) = 105 munmap(0xb49c3000, 3657728) = 0 close(22) = 0 ioctl(18, IIOCNETAIF, 0xb5139a5c) = 0 writev(5, [{"\4", 1}, {"g1decoder_dwl\0", 14}, {"DWLFreeLinear\t size:3657664, busAddress:0x1acfc040, virAddr:0xb49c3040\n\0", 72}], 3) = 87 writev(5, [{"\4", 1}, {"Hantro_omx_OSAL\0", 16}, {"OSAL_AllocatorFreeMem line 482 bus_data 0xb4646040 bus_address 0x1b079040 size 3657664 \0", 88}], 3) = 105 munmap(0xb4646000, 3657728) = 0 close(23) = 0 ioctl(18, IIOCNETAIF, 0xb5139a5c) = 0 writev(5, [{"\4", 1}, {"g1decoder_dwl\0", 14}, {"DWLFreeLinear\t size:3657664, busAddress:0x1b079040, virAddr:0xb4646040\n\0", 72}], 3) = 87 writev(5, [{"\4", 1}, {"Hantro_omx_OSAL\0", 16}, {"OSAL_AllocatorFreeMem line 482 bus_data 0xb4644040 bus_address 0x1b3f6040 size 8128 \0", 85}], 3) = 102 munmap(0xb4644000, 8192) = 0 close(24) = 0 ioctl(18, IIOCNETAIF, 0xb5139a5c) = 0 writev(5, [{"\4", 1}, {"g1decoder_dwl\0", 14}, {"DWLFreeLinear\t size:8128, busAddress:0x1b3f6040, virAddr:0xb4644040\n\0", 69}], 3) = 84 close(19) = 0 close(16) = 0 writev(5, [{"\4", 1}, {"Hantro_omx_OSAL\0", 16}, {"Close ion! alloc 0xb6f737ac ion_client 18\n\0", 43}], 3) = 60 close(18) = 0 ioctl(17, 0x6b06, 0x1) = 0 ioctl(17, 0x6b09, 0x1) = 0 close(17) = 0 writev(5, [{"\4", 1}, {"Hantro_omx_OSAL\0", 16}, {"OSAL_AllocatorFreeMem line 482 bus_data 0xb4d40040 bus_address 0x1aa04040 size 3112896 \0", 88}], 3) = 105 munmap(0xb4d40000, 3112960) = 0 close(20) = 0 ioctl(7, IIOCNETAIF, 0xb5139b84) = 0

[2017-09-20 01:12:22] hostile : uses some ion client calls I think

[2017-09-20 01:12:43] hostile : yeah here is where it acquired it

[2017-09-20 01:12:43] hostile : open("/dev/ion", O_RDWR) = 18 writev(5, [{"\4", 1}, {"ion\0", 4}, {"open /dev/ion success fd 18!\n\0", 30}], 3) = 35 writev(5, [{"\4", 1}, {"Hantro_omx_OSAL\0", 16}, {"open ion success!alloc 0xb6f737ac ion_client 18\n\0", 49}], 3) = 66 open("/dev/on2map", O_RDWR|O_DSYNC) = 19 brk(0xb6f80000) = 0xb6f80000 ioctl(7, 0xc0144900, 0xb51399d4) = 0 ioctl(7, IIOCNETGCF, 0xb51399f0) = 0 mmap2(NULL, 3112960, PROT_READ|PROT_WRITE, MAP_SHARED, 20, 0) = 0xb4d40000 ioctl(7, IIOCNETDNM, 0xb51399d4) = 0 writev(5, [{"\4", 1}, {"Hantro_omx_OSAL\0", 16}, {"OSAL_AllocatorAllocMem2 line 303 **bus_data 0xb4d40040 **bus_address 0x1aa04040 *size 3112896 , share_fd=20\0", 106}], 3) = 123 open("/system/gui/boot_animation/video.264", O_RDONLY) = 21

[2017-09-20 01:14:02] hostile : audio gets hammered down controlC0

[2017-09-20 01:14:03] hostile : ioctl(26, 0x400c4150, 0xb4545bc4) = 0 writev(5, [{"\6", 1}, {"DUSS&5d[ bs_audio_task: 120]:\0", 40}, {"no more data\n\0", 14}], 3) = 55 open("/dev/snd/controlC0", O_RDWR) = 27 ioctl(27, USBDEVFS_RELEASEINTERFACE, 0xb4545958) = 0 brk(0xb6f89000) = 0xb6f89000

[2017-09-20 01:17:37] jezzab : ok so dji_animation is calling it directly and when you play demo videos with the GUI its using Gstreamer plug in Disko by the looks

[2017-09-20 01:17:39] hostile : try to build OMXPlayer perhaps?

[2017-09-20 01:17:40] hostile : <http://elinux.org/Omxplayer>

[2017-09-20 01:18:00] jezzab : I actually looked and downloaded that when i started but for the life of me cant get it to compile

[2017-09-20 01:18:08] jezzab : maybe @unusuario128 can

[2017-09-20 01:18:16] hostile : extracted a file from the .deb?

[2017-09-20 01:18:26] jezzab : tried that too lol

[2017-09-20 01:18:37] hostile : word!

[2017-09-20 01:20:45] jezzab : hmm i wonder if I can do a Gstreamer call with the gui debugger on there

[2017-09-20 01:26:11] jezzab : Hmm but running on an MP4:

[2017-09-20 01:26:32] hostile : paste the test_dec binary here

[2017-09-20 01:33:32] hostile : wife just got home so I am gonna be a bit quiet

[2017-09-20 01:33:48] jezzab : No worries

[2017-09-20 01:33:55] jezzab : Thanks for taking a look

[2017-09-20 01:34:59] hostile : duss_hal_device_open duss_hal_device_start duss_hal_vmem_alloc duss_hal_vmem_free duss_hal_device_stop

[2017-09-20 01:35:06] hostile : it has that fucking weird DJI hardware layer shit built in

[2017-09-20 01:35:12] jezzab : I KNOW!

[2017-09-20 01:35:21] jezzab : the DUSS stuff is EVERYWHERE

[2017-09-20 01:35:56] hostile : I bet ya anything there is some other kind of *USS

[2017-09-20 01:35:59] hostile : and they just renamed it

[2017-09-20 01:36:07] jezzab : haha could be

[2017-09-20 01:37:00] jezzab : wanna make em into a HTPC lol

[2017-09-20 01:37:01] jezzab : <https://sourceforge.net/projects/mms4l/>

[2017-09-20 01:37:03] hostile : dji/duml/testing/test_dec/main.c

[2017-09-20 01:57:39] jezzab : wtf the demo videos have sound?? they didnt even demux it they just added a wav file of the same track on the video LOL

[2017-09-20 02:04:53] jezzab : strace of playing a video in the GUI:

[2017-09-20 02:12:48] jezzab : system/lib/libduml_ffremux.so

[2017-09-20 02:23:09] jezzab : Whats the go with the XMP stuff

[2017-09-20 02:46:51] jezzab : @hostile libduml_ffremux contains portions of ffmpeg and it was compiled with '--enable-gpl'.

[2017-09-20 02:48:16] jezzab : I dont really know what that all effects regarding GPL etc. But anyway. Just noticed it in the lib. Ill upload it now incase ive made a mistake

[2017-09-20 08:49:21] jezzab : all the directfb examples on the googles actually DO run. It was because the examples they used had deps on .png files in /system/usr/directfb. You put them in and they run. So thats optimistic. mplayer looks like a good candidate

[2017-09-20 09:20:37] unusuario128 : My bad. This error happens because SDL_main is not calling the initialization routines.

[2017-09-20 09:21:03] unusuario128 : These routines have something to do with JNI and Android.

[2017-09-20 09:21:44] jezzab : I think your right its kinda interwoven

[2017-09-20 09:22:36] jezzab : Would it be possible to cross compile mplayer? <http://www.mplayerhq.hu/design7/info.html> seems like it can do just about everything

[2017-09-20 09:22:52] unusuario128 : Let me check.

[2017-09-20 09:32:42] jezzab : Every fb demo in the goggles plays 90deg CW. So the far right is the "top" of the display

[2017-09-20 09:33:36] jezzab : They have another on the df_ve_test. Can't find the source. It actually turned off the left eye. Freaked me out lol

[2017-09-20 09:34:57] jezzab : God I wish they were seperate screens. That would be aaaaaaaawsome. Stereo and a gyro

[2017-09-20 09:39:55] unusuario128 : What about `ls /dev/fb* /dev/graphics/*`?

[2017-09-20 09:44:06] jezzab : give me a sec but /dev/graphics has fb0 I think

[2017-09-20 09:44:39] jezzab : just had them on charge. they were running for 9 hours lol

[2017-09-20 09:47:14] jezzab : yup no /dev/fb0

[2017-09-20 09:47:17] jezzab : but graphics:

[2017-09-20 09:53:02] unusuario128 : Two framebuffers.

[2017-09-20 09:53:11] unusuario128 : Seems good...

[2017-09-20 09:54:13] unusuario128 : You can try to do a quick trick to test if they are independent.

[2017-09-20 09:54:18] jezzab : ok

[2017-09-20 09:54:41] unusuario128 : 1. Run `cat /dev/urandom > /dev/graphics/fb0`

[2017-09-20 09:54:42] unusuario128 : 2. Stop it `CTRL` + `C`

[2017-09-20 09:54:58] unusuario128 : 3. Run `cat /dev/urandom > /dev/graphics/fb1`

[2017-09-20 09:55:10] unusuario128 : 4. Stop it.

[2017-09-20 09:55:32] unusuario128 : You should see "noise" in the screens.

[2017-09-20 09:55:59] jezzab : ok so the first one fb0

[2017-09-20 09:56:03] unusuario128 : Look between the steps 2 and 3 to ensure that the screens are independient.

[2017-09-20 09:56:55] jezzab : they arnt :disappointed:

[2017-09-20 09:57:07] jezzab : fb0 seems to write it a lot 'lighter' like its at the back

[2017-09-20 09:57:17] jezzab : and then fb1 is darker 'front'?

[2017-09-20 09:57:27] jezzab : kinda like that

[2017-09-20 09:57:55] unusuario128 : Well, then wait for MPlayer to do more civilized tests.

[2017-09-20 09:58:17] unusuario128 : Maybe there are two screens along with a semitransparent mirror.

[2017-09-20 10:01:20] unusuario128 : Ok, so we have a beam splitter.

[2017-09-20 10:02:10] jezzab : I don't quite understand it. Glad you do lol

[2017-09-20 10:04:43] unusuario128 : <https://en.wikipedia.org/wiki/Beam_splitter>

[2017-09-20 10:05:15] jezzab : So it's a 2160 wide display?

[2017-09-20 10:05:39] jezzab : Or 1080 split out to each eye?

[2017-09-20 10:06:32] jezzab : Sorry 1920

[2017-09-20 10:06:45] jezzab : Or double that?

[2017-09-20 10:20:41] unusuario128 : It is a 1080p display that has in front an optical device that multiplexes the picture to both eyes.

[2017-09-20 10:21:04] unusuario128 : Unless they've done something special, you won't get stereo vision.

[2017-09-20 10:21:57] jezzab : That's ok. I wasn't expecting it

[2017-09-20 10:22:07] kilrah : it's supposed to be 2 displays

[2017-09-20 10:22:20] unusuario128 : Maybe they added some sort of polarizer in the middle of the mirror light path so they can alternate the output and create pseudo-stereo by persistance of vision.

[2017-09-20 10:22:56] kilrah : with each one being routed **mostly** to one eye but with leakage

[2017-09-20 10:23:20] jezzab : Thru the goggles then thru each eye

[2017-09-20 10:23:43] unusuario128 : Yes, I was reading the DJI product page and it seems to have two displays.

[2017-09-20 10:23:44] kilrah : @jezzab how did you look at your dark/bright observation above

[2017-09-20 10:24:08] kilrah : would match that if you were looking only at one side

[2017-09-20 10:24:16] jezzab : It was like one was semi transparent

[2017-09-20 10:24:43] jezzab : fb0

[2017-09-20 10:24:49] unusuario128 : Maybe semi-polarized?

[2017-09-20 10:24:59] kilrah : did you look at both eyes one after the other?

[2017-09-20 10:25:16] jezzab : I'll do, again right now

[2017-09-20 10:25:29] kilrah : if understood correctly one should look bright to an eye and dark to the other, and the other the opposite

[2017-09-20 10:25:49] unusuario128 : Seems plausible

[2017-09-20 10:26:05] kilrah : i really need to take my 2nd pair apart

[2017-09-20 10:26:07] kilrah : :smile:

[2017-09-20 10:26:29] jezzab : ok so writing fb0

[2017-09-20 10:26:38] jezzab : same both eyes checked with one eye

[2017-09-20 10:26:42] jezzab : if that makes sense

[2017-09-20 10:26:57] jezzab : and i can just seen the DJI background a little

[2017-09-20 10:27:22] unusuario128 : This would mean that it has two screens (or one) in the front panel, one vertically polarized and one horizontally polarized. Then, the eyepiece form each side has 90º phased polarization.

[2017-09-20 10:27:57] jezzab : fb1 overwrites all but im going to kill dji_glasses and do it again starting with fb1

[2017-09-20 10:28:35] jezzab : its two layers :slightly_smiling_face:

[2017-09-20 10:28:57] jezzab : fb1 is not transparent at all, cant see the DJI background but the icons are over the top

[2017-09-20 10:29:23] jezzab : fb0 you can see the DJI slightly but the icons are overwritten

[2017-09-20 10:30:09] kilrah : it's possible there's a polarizer that needs active control if the 2 screens are used independently

[2017-09-20 10:30:25] kilrah : (like shutter glasses)

[2017-09-20 10:30:47] kilrah : so might need to do something else somewhere else (which might not even be present in current FW)

[2017-09-20 10:32:25] kilrah : ah so maybe they actually use the 2 screens separately, fb0 gets a clean video feed and fb1 just the data overlay!

[2017-09-20 10:32:47] unusuario128 : MPlayer. Static build. FB help here: <http://www.mplayerhq.hu/DOCS/HTML/en/fbdev.html>

[2017-09-20 10:33:41] unusuario128 : Includes zlib and fbdev support. Just add water.

[2017-09-20 10:33:44] jezzab : so far so good. Runs.

[2017-09-20 10:33:48] jezzab : now to test

[2017-09-20 10:39:37] jezzab : hmm

[2017-09-20 10:39:48] jezzab : wont listen to my -fb

[2017-09-20 10:45:39] unusuario128 : Can you elaborate a bit?

[2017-09-20 10:46:57] jezzab : apparently playing but not showing. tested fb1 and fb0

[2017-09-20 10:47:37] unusuario128 : @jezzab: Do you have any program using the FB?

[2017-09-20 10:48:04] jezzab : there were directFB demos on the goggles

[2017-09-20 10:48:07] jezzab : they work

[2017-09-20 10:48:11] unusuario128 : Did you try to specify a different mode as stated on the aforementioned documentation?

[2017-09-20 10:48:26] unusuario128 : Do you have DIrectFB support, then?

[2017-09-20 10:48:46] jezzab : im bumbling my way through it mate

[2017-09-20 10:55:30] jezzab : oohh close now

[2017-09-20 10:56:09] jezzab : /data/mplayer -ao pcm -vo fbdev:/dev/graphics/fb0 FOF.mp4

[2017-09-20 10:56:25] jezzab : play sideways all all of the fb demos do and half screen

[2017-09-20 10:56:30] jezzab : and slow as f...

[2017-09-20 10:56:37] jezzab : because not accelerated

[2017-09-20 10:57:36] unusuario128 : Can you `find / -name *directfb.so`?

[2017-09-20 11:01:05] unusuario128 : Can you give me that file?

[2017-09-20 11:01:24] jezzab : sure

[2017-09-20 11:01:45] unusuario128 : Do you have also mesa and/or drm?

[2017-09-20 11:02:26] jezzab : want a list of /system/lib ?

[2017-09-20 11:02:52] unusuario128 : Yes

[2017-09-20 11:03:34] jezzab : ignore the symlinks. Thats because there was no space for the libs for your ffplay

[2017-09-20 11:06:14] kilrah : okay the fb1 and fb0 are a software thing for the overlay it seems, the screens are really one for each eye

[2017-09-20 11:06:22] kilrah : top one right eye, back one left eye

[2017-09-20 11:06:46] jezzab : ah k

[2017-09-20 11:23:14] kilrah : so in theory, 3D possible...

[2017-09-20 11:28:13] jezzab : Would be pretty cool

[2017-09-20 11:50:53] unusuario128 : Building over the directfb on your system is nearly impossible.

[2017-09-20 11:56:00] unusuario128 : Also, the project is not actively maintained, which is not great at all.

[2017-09-20 11:56:39] jezzab : Bugger

[2017-09-20 11:59:09] unusuario128 : Exclamation or noun? :thinking_face:

[2017-09-20 11:59:50] unusuario128 : Well, what kind of FB demo do you have?

[2017-09-20 12:03:41] jezzab : The ones on there goggles are:

[2017-09-20 12:09:32] jezzab : "Bugger is a word that is quite often thrown around in the Aussie language, and can be used in various contexts to cover a wide range of expression needs. The main “oh, bugger” phrase is often used to show a bit of disappointment."

[2017-09-20 12:09:49] jezzab : :wink:

[2017-09-20 12:12:47] jezzab : That a demo video playing with mplayer. Very very slowy

[2017-09-20 13:41:12] unusuario128 : @jezzab: I'll try to look again at directfb.

[2017-09-20 13:41:26] hostile : damn you guys are hustling

[2017-09-20 13:42:26] jezzab : @unusuario128 has been a great help. And nice to see @kilrah popping in too. Been a bit lonely in here

[2017-09-20 13:42:46] unusuario128 : @jezzab: Lunch time. See you later.

[2017-09-20 13:43:14] unusuario128 : @jezzab: This evening I'll be traveling, but I'll return home tonight.

[2017-09-20 13:43:30] unusuario128 : I'll try to build directfb support into the player.

[2017-09-20 13:44:09] unusuario128 : Also, you seem to have GLES support..

[2017-09-20 13:44:12] jezzab : Sounds great :) I'll tinker around a bit more. They do weird stuff to use the acceleration

[2017-09-20 13:44:16] jezzab : Yes

[2017-09-20 13:44:29] hostile : lol "Been a bit lonely in here" you telling me! I wish I wasn't burried busy! I coulda used the company last month. =]

[2017-09-20 13:44:44] jezzab : Hastro OMX for the accel.

[2017-09-20 13:45:20] hostile : Hantaro... <http://www.verisilicon.com/IPPortfolio_14_58_2_HantroG1.html>

[2017-09-20 13:45:42] jezzab : @hostile your all good. I didn't have them then or know about the channel. Then did exactly what you did, found the channel and realised you'd done it a month before lol re boot anim etc

[2017-09-20 13:46:52] jezzab : Oh and you get over 8 hours run time on the bench with USB in lol

[2017-09-20 13:47:08] hostile : truth!

[2017-09-20 13:47:14] hostile : I feel like the Goggles Battery NEVER dies

[2017-09-20 13:47:40] jezzab : They're awesome to work on. I thought the RC was good but have a new winner

[2017-09-20 13:48:18] jezzab : Oh and killing dji_glasses fixes most stuff from video errors testing it reloading changed GUI files

[2017-09-20 13:48:33] jezzab : Saves a reboot and all the noise

[2017-09-20 20:41:34] unusuario128 : Arrived now.

[2017-09-20 20:42:04] unusuario128 : I'll see if I can do something with the directfb stuff.

[2017-09-20 20:42:41] unusuario128 : Can anybody `uname -a` the goggles?

[2017-09-20 22:12:21] jezzab : `Linux localhost 3.10.62 #1 SMP PREEMPT Fri Apr 21 02:56:08 CST 2017 armv7l GNU/Linux`

[2017-09-20 22:56:26] jezzab : `/data/mplayer -fs -vf rotate=8,flip -vo fbdev:/dev/graphics/fb0 video.mp4` plays full screen, correct rotation

[2017-09-20 22:56:34] jezzab : ....at a blistering 2fps

[2017-09-20 22:57:28] jezzab : no audio but I havent lookin into that yet and it would be hard to play audio synced I guess

[2017-09-20 23:05:21] jezzab : trying to play a wav file only

[2017-09-20 23:52:01] hostile : you need MPlayer OMX? <https://gitorious.org/ac100/mplayer-omx.git/>

[2017-09-21 00:38:20] jezzab : Could do. I think it needs alsa as well. there is a tinyalsa.so on there. I have tried outputting pcm direct to /dev/snd/pcmC0D0p etc and its a no go BUT even when I use tinyplay to test a wav it doesnt work most of the time. It did once. Thinking some permissions maybe. I think the hard bit will be getting the hardware decoding to work, maybe

[2017-09-21 00:46:28] jezzab : tried -hwdec=auto too.

[2017-09-21 08:17:16] unusuario128 : If it supports OMX, maybe will be the way to go.

[2017-09-21 08:19:50] jezzab : So how do we get hardware acceleration?

[2017-09-21 08:19:53] jezzab : with OMX?

[2017-09-21 08:20:41] unusuario128 : I'll try first directfb, because OMX requires further investigation.

[2017-09-21 08:21:35] jezzab : So directfb can access the hantro libs?

[2017-09-21 08:22:08] unusuario128 : I don't know.

[2017-09-21 08:22:09] jezzab : See they are all over the place. They call their DUSS crap with a special lib that is basically ffmpeg

[2017-09-21 08:22:33] jezzab : there are some straces ^^^^^ when running dji_animation

[2017-09-21 08:22:42] unusuario128 : The hardware decoding will be accessible with OMX, if applicable.

[2017-09-21 08:22:44] jezzab : which plays a video and wav file

[2017-09-21 08:22:57] jezzab : every video they play they remux it

[2017-09-21 08:23:09] jezzab : even the demo videos are aaa.mp4 aaa.wav

[2017-09-21 08:23:35] unusuario128 : Let me check the strace logs. I have to climb into the transcript. :wink:

[2017-09-21 08:23:46] jezzab : libduml_ffremux.so

[2017-09-21 08:24:05] jezzab : sorry mate. there is some good info there amongst the rambling lol

[2017-09-21 08:24:28] jezzab : Your doing awesome for not even having the platform

[2017-09-21 08:24:36] jezzab : I really appreciate your help

[2017-09-21 08:25:33] jezzab : But in short. I dont want to play crap thats been demuxed. Want to just play a media file straight up :slightly_smiling_face:

[2017-09-21 08:26:08] jezzab : And possibly set it up so I can just throw in an SD card with videos and watch em :slightly_smiling_face:

[2017-09-21 08:26:42] unusuario128 : :slightly_smiling_face:

[2017-09-21 08:30:54] unusuario128 : Do you have OMX libraries? (autoanswered: yes)

[2017-09-21 08:32:12] jezzab : see that /system/lib list^^

[2017-09-21 08:32:18] jezzab : they were the hantro ones from memory

[2017-09-21 08:32:43] jezzab : can certainly grab them

[2017-09-21 08:34:29] unusuario128 : Can you attach them?

[2017-09-21 08:34:34] jezzab : sure

[2017-09-21 08:34:41] jezzab : one sec ill fire up the goggs

[2017-09-21 08:35:33] jezzab : so you want anything with **omx**?

[2017-09-21 08:35:38] unusuario128 : yes

[2017-09-21 08:36:15] jezzab : ok there is just one lol

[2017-09-21 08:36:17] jezzab : makes it easy

[2017-09-21 08:36:31] jezzab : `libon2_hantro_omx_core.so`

[2017-09-21 08:38:03] jezzab : ahh

[2017-09-21 08:38:05] jezzab : there is more

[2017-09-21 08:38:08] jezzab : one moment

[2017-09-21 08:51:54] jezzab : They have the stagefright libs which seem to be linked in with the whole OMX thing: ```-rw-r--r-- root root [9412 2008-08-01](tel:94122008-08-01) 12:00 libon2_stagefrighthw.so -rw-rw-rw- root root [79256 2015-02-09](tel:792562015-02-09) 12:00 libstagefright_foundation.so```

[2017-09-21 09:07:59] unusuario128 : Can you pull libdl.so from your goggles?

[2017-09-21 09:10:15] unusuario128 : libstagefright support was removed from ffmpeg because it was obscure and not useful :disappointed:

[2017-09-21 09:42:13] jezzab : There you go @unusuario128 thats the entire /system/lib dir

[2017-09-21 09:42:34] unusuario128 : @jezzab: Thanks!

[2017-09-21 09:42:56] jezzab : give me a yell when you get it and ill delete it. this joint runs out of space pretty quick

[2017-09-21 09:58:00] unusuario128 : Ok

[2017-09-21 09:59:38] unusuario128 : @jezzab: Yeeeeeeeell!

[2017-09-21 12:29:09] unusuario128 : Discarded OMXPlayer definitely.

[2017-09-21 12:32:29] jezzab : No good?

[2017-09-21 12:34:04] unusuario128 : No, 0xBAAD

[2017-09-21 12:34:19] jezzab : 0xDEAD

[2017-09-21 12:34:34] jezzab : Need a 0xCAFE

[2017-09-21 12:35:03] unusuario128 : I'll eat 0xF00D in half an hour :slightly_smiling_face:

[2017-09-21 12:35:15] jezzab : Hehe

[2017-09-21 12:35:20] unusuario128 : Seriously, OMXPlayer is a no-go

[2017-09-21 12:35:37] unusuario128 : It was tailored for RPi Broadcom video decoder

[2017-09-21 12:35:47] jezzab : Dammit

[2017-09-21 12:36:05] unusuario128 : I'll need to put both eyes over directfb

[2017-09-21 12:36:13] jezzab : Do you have any suggestions?

[2017-09-21 12:36:41] jezzab : Could mplayer be setup to use the hastro stuff easily?

[2017-09-21 12:36:49] jezzab : Or a real PITA?

[2017-09-21 12:38:46] unusuario128 : If I manage to support directfb it should go smooth.

[2017-09-21 12:38:58] jezzab : Oh cool

[2017-09-21 12:39:13] unusuario128 : If not, let's hope the help of some graphics guru here.

[2017-09-21 12:39:47] jezzab : Fingers and toes crossed :)

[2017-09-22 10:39:18] unusuario128 : Does the working internal video player have any option to receive streams?

[2017-09-22 10:39:46] jezzab : They use disko

[2017-09-22 10:40:30] jezzab : Yes

[2017-09-22 10:40:38] jezzab : And tie it together

[2017-09-22 10:42:03] jezzab : I can't work out why they just don't play it straight up as disko uses xine and GStreamer

[2017-09-22 10:42:17] jezzab : But I don't think they compiled with xine

[2017-09-22 10:42:42] jezzab : <http://www.diskohq.com/products/disko-framework/overview>

[2017-09-22 10:43:27] unusuario128 : But, do you have some command-line program that accepts two separate streams for audio and video?

[2017-09-22 10:43:52] jezzab : dji_animation

[2017-09-22 10:44:05] unusuario128 : And, does it accept stdin input?

[2017-09-22 10:44:39] jezzab : Just -a audo.wav and -v video.mp4

[2017-09-22 10:45:03] jezzab : But the MP4 has to be specific and no wrapper

[2017-09-22 10:45:07] jezzab : X264

[2017-09-22 10:45:28] jezzab : First posts in this channel from hostile show the ffprobe

[2017-09-22 10:45:30] unusuario128 : Does your shell support `<(command)` descriptor redirection?

[2017-09-22 10:45:50] jezzab : I don't know

[2017-09-22 10:46:01] unusuario128 : (i. e.) `cat <(echo hello) <(echo goodbye)`

[2017-09-22 10:46:08] jezzab : What would you like me to try?

[2017-09-22 10:46:20] unusuario128 : If you can...

[2017-09-22 10:48:34] jezzab : Just getting rid of the P4 crap off the bench and setting up the goggs

[2017-09-22 10:50:59] jezzab : ```root@wm220_dz_ah0001_v5:/ # cat <(echo hello) <(echo goodbye) cat <(echo hello) <(echo goodbye) /system/bin/sh: syntax error: '(' unexpected```

[2017-09-22 10:52:03] jezzab : ```1|root@wm220_dz_ah0001_v5:/ # dji_animation -h dji_animation -h usage: dji_animation [-h] help [-v] boot animation video file name [-a] boot animation audio file name```

[2017-09-22 11:14:16] unusuario128 : As this is for the bootanimation, it takes the audio from a separate file to comply with the Android "standard" (I don't to which extent) bootanimation packages.

[2017-09-22 11:17:00] jezzab : Oh ok

[2017-09-22 11:19:21] unusuario128 : It relies on the DUML operative system abstraction layer (hype-deserving name for a couple of function wrappers).

[2017-09-22 11:19:44] jezzab : Yup

[2017-09-22 11:20:00] jezzab : Basically ffmpeg by the looks

[2017-09-22 11:20:17] jezzab : And compiled with --gpl

[2017-09-22 11:20:57] jezzab : Ie libduml_ffremux.so

[2017-09-22 11:30:17] unusuario128 : What contains `/system/etc/dji.json`?

[2017-09-22 11:30:50] jezzab : It's in most devices. It's how they setup their hardware

[2017-09-22 11:30:56] jezzab : I'll upload

[2017-09-22 11:34:06] unusuario128 : `/dev/display0` looks promising. So it does `/dev/display1`

[2017-09-22 11:34:24] unusuario128 : Can you try to fill them with `/dev/urandom`?

[2017-09-22 11:35:56] unusuario128 : First of all, please check that they are not symlinks to `/dev/fb0` and `/dev/fb1`

[2017-09-22 11:40:45] jezzab : there is no /dev/display1 :confused:

[2017-09-22 11:41:48] unusuario128 : I don't see display0 neither. :confused:

[2017-09-22 11:47:42] unusuario128 : Can you `ls -la /dev/*`?

[2017-09-22 11:47:55] unusuario128 : ...or even `find /dev`

[2017-09-22 11:51:06] jezzab : ```root@wm220_dz_ah0001_v5:/dev # busybox find|grep "display" busybox find|grep "display"```

[2017-09-22 11:58:52] unusuario128 : ...because the program opens (or tries to open) that file device

[2017-09-22 12:04:04] jezzab : Hmm

[2017-09-22 12:06:21] jezzab : I did upload an strace of running ^^^ I'm pretty sure

[2017-09-22 12:06:34] jezzab : Can do again if needed

[2017-09-22 12:06:52] unusuario128 : No worries. Let's use the infamous transcript search.

[2017-09-22 12:07:01] jezzab : Black magic lol

[2017-09-22 12:13:31] unusuario128 : _Lagarto, lagarto_ Well, it is not actually called on the strace.

[2017-09-22 12:19:32] unusuario128 : Lunch time. See you later.

[2017-09-22 12:19:52] unusuario128 : Found some interesting `strace` strings.

[2017-09-22 12:20:55] unusuario128 : So, we have Hantro x170

[2017-09-22 12:21:02] unusuario128 : (or so I think)

[2017-09-22 12:25:38] jezzab : I agree. Searching it seems to be the Hantro x170

[2017-09-22 13:53:40] unusuario128 : Related: <https://forum.armbian.com/index.php?/topic/4806-the-vpu-driver/>

[2017-09-22 14:00:31] darksimpson : Moving here from hardware, I look at FPGA docs and need to say that it seems possible to synthesize configuration for FPGA that will split input LVDS steram say at 1920x1080 to two different output LVDS streams with half-height or half-width and rescale it back to 1920x1080 to output it to screens, using only one linebuffer that will fit in FPGA RAM. So we can technically able to get 3d working. But unfortunately I'm not FPGA guru :disappointed: And other heavy-lifring work must be done, like investigating how to update FPGA configuration programmatically, not disassembling goggles on bench and investigate all signal lines going to/from FPGA (may be need to desolder FPGA to test it all or/and use x-ray).

[2017-09-22 20:30:05] unusuario128 : @jezzab: Please try this `ffmpeg`.

[2017-09-22 21:04:13] unusuario128 : It was dynamically linked to your sysroot and between the components, so you will like to symlink/copy all the items of the folders to their correspondent locations.

[2017-09-22 21:09:56] unusuario128 : What I would like is to know if you can use the `ffmpeg` command-line tool to decode `h264` video using the `h264_omx` codec (allegedly hardware-accelerated by the Hantro x170).

[2017-09-22 21:32:08] unusuario128 : Also, would be great if you can benchmark the de/encoding time of the `h264_omx` codec compared to the normal `h264` codec.

[2017-09-22 22:48:18] jezzab : giving it a go now

[2017-09-22 23:18:30] jezzab : @unusuario128 seems its looking for `libOMX_Core.so` and `libOmxCore.so`

[2017-09-23 09:51:58] unusuario128 : @jezzab: I'll try to shim that library from the hantro core.

[2017-09-23 10:56:40] unusuario128 : Patching odd job. May fail.

[2017-09-23 10:57:23] unusuario128 : @jezzab: If it doesn't work as `libOMX_Core.so`, try renaming it to `libOmxCore.so`

[2017-09-23 10:58:25] unusuario128 : There aren't two OpenMAX implementations with the same exposed API. And it is supposed to be an standard...

[2017-09-23 10:58:37] unusuario128 : Let's hope the best.

[2017-09-23 11:13:13] jezzab : @unusuario128 I think its looking for both

[2017-09-23 11:13:32] jezzab : ```root@wm220_dz_ah0001_v5:/data/ffmpeg/bin # ./ffmpeg -i video.264 -framerate 30 -video_size 720x404 -vcodec h264_omx -maxrate 768k -bufsize 8080k -vf "format=yuv420p" -g 60 test.mp4 8080k -vf "format=yuv420p" -g 60 test.mp4 < ffmpeg version 3.3.git Copyright (c) 2000-2017 the FFmpeg developers built with gcc 4.9.x (GCC) 20150123 (prerelease) configuration: --enable-cross-compile --enable-shared --disable-static --arch=arm --target-os=linux --enable-parsers --enable-shared --enable-gpl --enable-version3 --enable-protocols --enable-nonfree --enable-pthreads --enable-pic --enable-hardcoded-tables --enable-omx --enable-decoder=mpegvideo --enable-decoder=mpeg1video --enable-decoder=mpeg2video --enable-decoder=mjpeg --enable-decoder=mjpegb --enable-decoder=mpeg4 --enable-decoder=h264 --enable-decoder=theora --enable-decoder=vc1 --enable-decoder=wmv3 --enable-decoder=vp6 --enable-decoder=vp6f --enable-decoder=vp8 --enable-decoder=opus --cross-prefix=arm-linux-androideabi- --sysroot=/home/casa/android-ndk-r15c/platforms/android-14/arch-arm --sysinclude=/home/casa/android-ndk-r15c/platforms/android-14/arch-arm/usr/include libavutil 55. 76.100 / 55. 76.100 libavcodec 57.106.101 / 57.106.101 libavformat 57. 82.101 / 57. 82.101 libavdevice 57. 8.101 / 57. 8.101 libavfilter 6.106.100 / 6.106.100 libswscale 4. 7.103 / 4. 7.103 libswresample 2. 8.100 / 2. 8.100 libpostproc 54. 6.100 / 54. 6.100 [0;35m[h264 @ 0x46290] [0m[0;33mStream #0: not enough frames to estimate rate; consider increasing probesize [0mInput #0, h264, from 'video.264': Duration: N/A, bitrate: N/A Stream #0:0: Video: h264 (High), yuv420p(tv, bt709, progressive), 1920x1080, 60 fps, 60 tbr, 1200k tbn, 120 tbc File 'test.mp4' already exists. Overwrite ? [y/N] y y Stream mapping: Stream #0:0 -> #0:0 (h264 (native) -> h264 (h264_omx)) Press [q] to stop, [?] for help [1;36m[h264_omx @ 0x47c60] [0m[0;33mlibOMX_Core.so not found [0m[1;36m[h264_omx @ 0x47c60] [0m[0;33mNot all functions found in libOmxCore.so [0m[1;31mError initializing output stream 0:0 -- Error while opening encoder for output stream #0:0 - maybe incorrect parameters such as bit_rate, rate, width or height [0mConversion failed!```

[2017-09-23 11:13:41] jezzab : ill try and copy it to the same as the other one

[2017-09-23 11:13:43] jezzab : double up

[2017-09-23 11:15:02] jezzab : ```1|root@wm220_dz_ah0001_v5:/data/ffmpeg/bin # cp /system/lib/libOmxCore.so /system/lib/libOMX_Core.so bOmxCore.so /system/lib/libOMX_Core.so < root@wm220_dz_ah0001_v5:/data/ffmpeg/bin # ./ffmpeg -i video.264 -framerate 30 -video_size 720x404 -vcodec h264_omx -maxrate 768k -bufsize 8080k -vf "format=yuv420p" -g 60 test.mp4 8080k -vf "format=yuv420p" -g 60 test.mp4 < ffmpeg version 3.3.git Copyright (c) 2000-2017 the FFmpeg developers built with gcc 4.9.x (GCC) 20150123 (prerelease) configuration: --enable-cross-compile --enable-shared --disable-static --arch=arm --target-os=linux --enable-parsers --enable-shared --enable-gpl --enable-version3 --enable-protocols --enable-nonfree --enable-pthreads --enable-pic --enable-hardcoded-tables --enable-omx --enable-decoder=mpegvideo --enable-decoder=mpeg1video --enable-decoder=mpeg2video --enable-decoder=mjpeg --enable-decoder=mjpegb --enable-decoder=mpeg4 --enable-decoder=h264 --enable-decoder=theora --enable-decoder=vc1 --enable-decoder=wmv3 --enable-decoder=vp6 --enable-decoder=vp6f --enable-decoder=vp8 --enable-decoder=opus --cross-prefix=arm-linux-androideabi- --sysroot=/home/casa/android-ndk-r15c/platforms/android-14/arch-arm --sysinclude=/home/casa/android-ndk-r15c/platforms/android-14/arch-arm/usr/include libavutil 55. 76.100 / 55. 76.100 libavcodec 57.106.101 / 57.106.101 libavformat 57. 82.101 / 57. 82.101 libavdevice 57. 8.101 / 57. 8.101 libavfilter 6.106.100 / 6.106.100 libswscale 4. 7.103 / 4. 7.103 libswresample 2. 8.100 / 2. 8.100 libpostproc 54. 6.100 / 54. 6.100 [0;35m[h264 @ 0x46290] [0m[0;33mStream #0: not enough frames to estimate rate; consider increasing probesize [0mInput #0, h264, from 'video.264': Duration: N/A, bitrate: N/A Stream #0:0: Video: h264 (High), yuv420p(tv, bt709, progressive), 1920x1080, 60 fps, 60 tbr, 1200k tbn, 120 tbc File 'test.mp4' already exists. Overwrite ? [y/N] y y Stream mapping: Stream #0:0 -> #0:0 (h264 (native) -> h264 (h264_omx)) Press [q] to stop, [?] for help [1;36m[h264_omx @ 0x47c60] [0m[0;33mNot all functions found in libOMX_Core.so [0m[1;36m[h264_omx @ 0x47c60] [0m[0;33mNot all functions found in libOmxCore.so [0m[1;31mError initializing output stream 0:0 -- Error while opening encoder for output stream #0:0 - maybe incorrect parameters such as bit_rate, rate, width or height [0mConversion failed!```

[2017-09-23 11:15:47] jezzab : excuse the color chars, i didnt remove them this time. My terminal cant parse them

[2017-09-23 11:19:59] unusuario128 : They don't affect the output.

[2017-09-23 11:20:19] unusuario128 : Let's see what function needs it apart of the Hantro ones...

[2017-09-23 11:30:48] unusuario128 : I'm building again ffmpeg with all the bells and whistles...

[2017-09-23 11:31:01] unusuario128 : ...and some neat addons :wink:

[2017-09-23 11:38:52] jezzab : Nice :)

[2017-09-23 11:41:46] unusuario128 : Just a question: why does your terminal spit the color codes without interpreting them? Is ADB eating the 0x1B scape sequence?

[2017-09-23 11:42:04] jezzab : Yup

[2017-09-23 11:42:24] jezzab : If I use cygwin and then run adb it's ok

[2017-09-23 11:42:32] jezzab : Just DOS crap

[2017-09-23 11:43:09] unusuario128 : Oooopsie... Windows user! Arghhhh!

[2017-09-23 11:43:11] unusuario128 : :wink:

[2017-09-23 11:43:48] jezzab : Haha yeah for this. My Linux server is on my ESXi machine. Which I use for most stuff

[2017-09-23 11:43:49] unusuario128 : Uploading ffmpeg...

[2017-09-23 11:44:59] unusuario128 : So, do you virtualize Linux with ESXi?

[2017-09-23 11:45:59] unusuario128 : Or do you use it as host?

[2017-09-23 11:46:34] unusuario128 : ^ - Try the above ffmpeg when running the "conflictive" command.

[2017-09-23 11:46:45] jezzab : Yup. My main Linux server VM that runs everything I do (plex, OpenHAB 2, http, samba and MYSQL) then another Linux test bed, and a couple of windows VMs

[2017-09-23 11:46:53] jezzab : No, a local rack

[2017-09-23 11:49:43] unusuario128 : I tend to be more self-contained and host myself all the things. I run GNU/Linux or macOS on all my machines and only have a virtualized Windows XP (yes, XP) for dynamical reverse engineering of Windows software.

[2017-09-23 11:54:16] jezzab : My rack server on the floor of my office here lol

[2017-09-23 11:55:44] jezzab : Will i have to copy over all these libs again?

[2017-09-23 11:56:03] jezzab : Or can i just do the main executables?

[2017-09-23 11:58:12] unusuario128 : You need to copy the libraries also.

[2017-09-23 11:58:33] unusuario128 : Or, at least, copy `libavcodec`, which should be enough

[2017-09-23 11:59:12] jezzab : ok

[2017-09-23 11:59:16] jezzab : libs are a pita

[2017-09-23 12:02:26] jezzab : just a few more options there lol

[2017-09-23 12:02:48] jezzab : but still the same ofcourse. looking for the omx stuff

[2017-09-23 12:05:32] unusuario128 : Ooopsie. I miss something.

[2017-09-23 12:06:43] unusuario128 : Did you copy all the libraries?

[2017-09-23 12:07:45] jezzab : no i didnt just libavcodec.so

[2017-09-23 12:08:44] jezzab : but then again its probably looking for libavcodec.so.57

[2017-09-23 12:08:48] unusuario128 : Sorry, my fail.

[2017-09-23 12:08:56] jezzab : there is symlinks and crap every where when its tar.gz

[2017-09-23 12:09:06] unusuario128 : Don't bother copyng more things.

[2017-09-23 12:09:18] jezzab : i cant just use the so files i had to run it and check the calling and then the symlink back to its etc last time

[2017-09-23 12:09:29] jezzab : ok

[2017-09-23 12:09:47] unusuario128 : Uploading.

[2017-09-23 12:10:12] unusuario128 : Now you should be fine copying only libavcodec.so

[2017-09-23 12:10:23] jezzab : ok cool

[2017-09-23 12:10:39] jezzab : so the new executables are just looking for the .so files not the [so.xxx](http://so.xxx) files like before?

[2017-09-23 12:11:33] unusuario128 : You should replace exactly `libavcodec.so.57.106.101`

[2017-09-23 12:11:44] unusuario128 : No less, no more.

[2017-09-23 12:14:38] jezzab : ill have to rename it to libavcodec.so.57

[2017-09-23 12:14:40] jezzab : on sec

[2017-09-23 12:14:42] unusuario128 : Nope

[2017-09-23 12:14:47] jezzab : thats what its lookin for

[2017-09-23 12:14:58] jezzab : ```root@wm220_dz_ah0001_v5:/data/ffmpeg/bin # ./ffmpeg -i video.264 -framerate 30 -video_size 720x404 -vcodec h264_omx -maxrate 768k -bufsize 8080k -vf "format=yuv420p" -g 60 test.mp4 8080k -vf "format=yuv420p" -g 60 test.mp4 < CANNOT LINK EXECUTABLE: could not load library "libavdevice.so.57" needed by "./ffmpeg"; caused by could not load library "libavfilter.so.6" needed by "libavdevice.so.57"; caused by could not load library "libavformat.so.57" needed by "libavfilter.so.6"; caused by could not load library "libavcodec.so.57" needed by "libavformat.so.57"; caused by library "libavcodec.so.57" not found```

[2017-09-23 12:15:56] unusuario128 : Both `libavcodec.so` and `libavcodec.so.57` should be symliks pointing to `libavcodec.so.57.106.101`

[2017-09-23 12:17:03] unusuario128 : So the `libavcodec.so.57.106.101` is the real library, which should be conveniently symlinked to its "friendly" name(s).

[2017-09-23 12:18:53] unusuario128 : Now it "works"

[2017-09-23 12:18:56] unusuario128 : :wink:

[2017-09-23 12:19:17] unusuario128 : Something went wrong into my shimmed libraries.

[2017-09-23 12:19:51] unusuario128 : However, it did not find anything "exotical"

[2017-09-23 12:20:09] unusuario128 : This means that probably I can get OMX to work.

[2017-09-23 12:20:23] jezzab : w00t!

[2017-09-23 12:20:29] jezzab : Well done

[2017-09-23 12:20:49] unusuario128 : However, let's wait. I don't really know if I'm right.

[2017-09-23 12:21:04] jezzab : Ok, i'll be quietly optimistic

[2017-09-23 13:03:26] unusuario128 : @jezzab Please check the new shim.

[2017-09-23 13:03:41] unusuario128 : This time it was made from "scratch"

[2017-09-23 13:16:09] unusuario128 : @jezzab: Theoretically, it should work to some extent.

[2017-09-23 13:29:13] jezzab : hmm same

[2017-09-23 13:29:24] jezzab : ahh coz its libOmxCore.so

[2017-09-23 13:29:42] jezzab : ill copy it over

[2017-09-23 13:30:39] jezzab : Shorter fail list

[2017-09-23 13:31:31] jezzab : so just to reiterate. libOMX_Core.so and libOmxCore.so are the same file

[2017-09-23 13:32:24] unusuario128 : Yes. You should delete one of them and use the other. I'll provide you `libOMX_Core.so`, so please delete `libOmxCore.so`

[2017-09-23 13:32:44] unusuario128 : I need to look over these functions. Please wait a bit.

[2017-09-23 13:33:13] jezzab : Ok

[2017-09-23 13:34:02] jezzab : the current ffmpeg build looks for both files.....

[2017-09-23 13:34:47] unusuario128 : Yes, but is redundancy. If it fails using the first, it tries to use the second.

[2017-09-23 13:34:55] jezzab : ok cool

[2017-09-23 13:35:02] jezzab : I understand now

[2017-09-23 13:38:20] unusuario128 : @jezzab: Try this now ^

[2017-09-23 13:51:24] jezzab : ```Stream mapping: Stream #0:0 -> #0:0 (h264 (native) -> h264 (h264_omx)) Press [q] to stop, [?] for help [1;36m[h264_omx @ 0x47c60] [0m[0;33munusuario128: Lack of OMX_GetComponentsOfRole function [0m[1;36m[h264_omx @ 0x47c60] [0m[0;33munusuario128: Lack of OMX_GetRolesOfComponent function [0m[1;36m[h264_omx @ 0x47c60] [0m[0;33mNot all functions found in libOMX_Core.so [0m[1;36m[h264_omx @ 0x47c60] [0m[0;33mlibOmxCore.so not found [0m[1;31mError initializing output stream 0:0 -- Error while opening encoder for output stream #0:0 - maybe incorrect parameters such as bit_rate, rate, width or height [0mConversion failed!```

[2017-09-23 13:53:33] unusuario128 : Did you use the latest file from 3:38 PM?

[2017-09-23 13:53:44] unusuario128 : It should have those symbols...

[2017-09-23 13:55:34] jezzab : I used the last file yiu sent. It's 11:55pm here lol

[2017-09-23 13:55:58] unusuario128 : Well, I've messed up with the timezone. I'm on CEST.

[2017-09-23 13:56:13] jezzab : All good :)

[2017-09-23 13:56:16] unusuario128 : md5=375c94f11dd8103a18bf6575893f10b

[2017-09-23 13:56:32] jezzab : For thr so file?

[2017-09-23 13:57:13] unusuario128 : Yes.

[2017-09-23 13:57:28] unusuario128 : For `libOMX_Core.so`

[2017-09-23 13:59:41] jezzab : my fault i ran from the wrong dir

[2017-09-23 13:59:49] jezzab : seg fault now

[2017-09-23 13:59:50] unusuario128 : `e375c94f11dd8103a18bf6575893f10b libOMX_Core.so`

[2017-09-23 13:59:59] jezzab : ```Stream mapping: Stream #0:0 -> #0:0 (h264 (native) -> h264 (h264_omx)) Press [q] to stop, [?] for help Segmentation fault```

[2017-09-23 14:00:19] unusuario128 : Well, something went wrong.

[2017-09-23 14:00:32] unusuario128 : I'll take a look again after lunch time.

[2017-09-23 14:01:19] jezzab : Ok mate. Ill probably be hitting the sack soon but if you upload I will test in the morning

[2017-09-23 14:02:18] jezzab : Thank you

[2017-09-23 14:02:23] codeforge : hi, is there an fcc mod for dji goggles?

[2017-09-23 14:02:59] jezzab : As per the AC, the FCC mod etc is for SENDING, and not for video

[2017-09-23 14:03:04] jezzab : on the RC

[2017-09-23 14:04:24] codeforge : i'm using spark with usb cable so i cannot connect goggles via ocusync like mavic.

[2017-09-23 14:04:42] codeforge : the dji go 4 within goggles i think will be modded

[2017-09-23 14:08:26] jezzab : Then the RC is doing the work

[2017-09-23 14:16:27] codeforge : ok thanks, so with spark what version do you think i need to install? thanks

[2017-09-23 14:19:03] jezzab : I'm not sure with Spark sorry

[2017-09-23 14:23:41] codeforge : ok ok thanks for reply

[2017-09-23 14:54:34] unusuario128 : @jezzab: Please try that.

[2017-09-23 14:55:54] unusuario128 : @jezzab: It may segfault yet.

[2017-09-23 14:57:59] jezzab : Segfault :disappointed:

[2017-09-23 14:58:12] jezzab : Ok mate im gonna head to bed.

[2017-09-23 14:58:42] unusuario128 : Ok. Good night!

[2017-09-23 14:59:05] jezzab : If you have anything else please upload it and ill give it a crack when I get up

[2017-09-23 14:59:13] unusuario128 : Ok.

[2017-09-23 15:21:09] unusuario128 : Information about the hardware video codec: The OMX API version seems to be OpenMAX IL 1.1.2 The IP core for video decoding seems to be Hantro x170

[2017-09-23 18:15:51] unusuario128 : `efdc771d5b80019cd1b20c0a28002591 libOMX_Core.so`

[2017-09-23 19:45:44] unusuario128 : @jezzab: Please try the file attached above

[2017-09-23 23:56:16] jezzab : ```Stream mapping: Stream #0:0 -> #0:0 (h264 (native) -> h264 (h264_omx)) Press [q] to stop, [?] for help [h264_omx @ 0x47c60] Using OMX.hantro.H1.video.encoder [h264_omx @ 0x47c60] err 80001019 (-2147479527) on line 498 Error initializing output stream 0:0 -- Error while opening encoder for output stream #0:0 - maybe incorrect parameters such as bit_rate, rate, width or height Conversion failed!```

[2017-09-23 23:56:29] jezzab : @unusuario128 ^^^

[2017-09-24 08:32:14] unusuario128 : @jezzab: Can you try the inverse operation (decoding)?

[2017-09-24 08:32:43] jezzab : Yeah. Not too sure how to do it but Ill give it a crack lol

[2017-09-24 08:35:15] jezzab : hmm im sorry but I dont know how to do it? do you have a command line suggestion?

[2017-09-24 08:36:05] jezzab : ...plus today consisted of a bbq and a substantial amount of beer

[2017-09-24 08:37:29] unusuario128 : Don't worry.

[2017-09-24 08:38:15] jezzab : More then happy to test if it will help mate. I've got everything setup

[2017-09-24 08:41:36] unusuario128 : `ASYNC: unsupported color format`: `0x80001019`

[2017-09-24 08:42:11] jezzab : ahh

[2017-09-24 08:42:12] unusuario128 : `API: unsupported state: %s\n`: `0x80001019`

[2017-09-24 08:42:20] jezzab : thats the "error line 498"

[2017-09-24 08:42:30] unusuario128 : There are various errors with the same code.

[2017-09-24 08:42:51] jezzab : Is it because of what im trying to do with the encode?

[2017-09-24 08:43:00] unusuario128 : May be.

[2017-09-24 08:43:08] jezzab : I just grabbed an example but it was encoding to a stream 'flv'

[2017-09-24 08:43:22] unusuario128 : `API: Only uncompressed video accepted for input`: 0x80001019

[2017-09-24 08:43:42] jezzab : ill just try another video for a sec

[2017-09-24 08:43:44] unusuario128 : `ASYNC: unsupported input format. No such video codec`: `0x80001019`

[2017-09-24 08:43:50] jezzab : the one im using is the boot animation

[2017-09-24 08:43:56] jezzab : ill try one of the demo videos on there

[2017-09-24 08:45:14] unusuario128 : Maybe you can find the exact error string on android `logcat` or `dmesg` or whatever it uses.

[2017-09-24 08:45:40] unusuario128 : Grepping for the strings that I've identified you can probably find the exact error.

[2017-09-24 08:46:09] jezzab : using a different video yields a different error:

[2017-09-24 08:47:04] unusuario128 : This is your fail.

[2017-09-24 08:47:23] unusuario128 : The line wrapping of your terminal deceived you.

[2017-09-24 08:47:35] jezzab : ahh

[2017-09-24 08:47:49] jezzab : im going to try a more simpler example from now on

[2017-09-24 08:48:55] unusuario128 : `test.mp448080k` replaced by `test.mp4 48080k`

[2017-09-24 08:49:10] jezzab : simple example shows the same error as previously today

[2017-09-24 08:49:24] jezzab : ```Input #0, mov,mp4,m4a,3gp,3g2,mj2, from 'snow.mp4': Metadata: major_brand : mp42 minor_version : 0 compatible_brands: mp42mp41 creation_time : 2017-03-28T04:06:12.000000Z Duration: 00:01:02.21, start: 0.000000, bitrate: 11125 kb/s Stream #0:0(eng): Video: h264 (Main) (avc1 / 0x31637661), yuv420p(tv, bt709), 1920x1080, 10810 kb/s, 29.97 fps, 29.97 tbr, 30k tbn, 59.94 tbc (default) Metadata: creation_time : 2017-03-28T04:06:12.000000Z handler_name : Alias Data Handler encoder : AVC Coding Stream #0:1(eng): Audio: aac (LC) (mp4a / 0x6134706D), 48000 Hz, stereo, fltp, 317 kb/s (default) Metadata: creation_time : 2017-03-28T04:06:12.000000Z handler_name : Alias Data Handler Stream mapping: Stream #0:0 -> #0:0 (h264 (native) -> h264 (h264_omx)) Stream #0:1 -> #0:1 (aac (native) -> ac3 (native)) Press [q] to stop, [?] for help [1;36m[h264_omx @ 0x493b0] [0mUsing OMX.hantro.H1.video.encoder [1;36m[h264_omx @ 0x493b0] [0m[1;31merr 80001019 (-2147479527) on line 498 [0m[1;31mError initializing output stream 0:0 -- Error while opening encoder for output stream #0:0 - maybe incorrect parameters such as bit_rate, rate, width or height [0mConversion failed!```

[2017-09-24 08:49:39] unusuario128 : You should run `logcat` and/or `dmesg` to see exactly what failed.

[2017-09-24 08:51:17] unusuario128 : I think that you could run `logcat | busybox grep h1_video_encoder`

[2017-09-24 08:51:42] jezzab : ok. dmesg wasnt useful

[2017-09-24 08:51:57] jezzab : ```root@wm220_dz_ah0001_v5:/data/ffmpeg/bin # logcat | busybox grep h1_video_encoder rep h1_video_encoder < E/h1_video_encoder(23550): API: Only uncompressed video accepted for input```

[2017-09-24 08:52:32] unusuario128 : Here you have the error :slightly_smiling_face:

[2017-09-24 08:53:23] unusuario128 : You'll want to "somehow" have a video without compression to test (maybe h264).

[2017-09-24 08:54:58] jezzab : how the hell would you have uncompressed H264??

[2017-09-24 08:55:20] unusuario128 : Good question.

[2017-09-24 08:55:36] jezzab : trying to find a simple raw avi

[2017-09-24 08:55:38] unusuario128 : ...which I couldn't answer without using St. Google.

[2017-09-24 08:55:56] jezzab : ```Input #0, avi, from 'drop.avi': Duration: 00:00:06.07, start: 0.000000, bitrate: 891 kb/s Stream #0:0: Video: indeo4 (IV41 / 0x31345649), yuv410p, 256x240, 887 kb/s, 30 fps, 30 tbr, 30 tbn, 30 tbc File 'output.avi' already exists. Overwrite ? [y/N] y y Stream mapping: Stream #0:0 -> #0:0 (indeo4 (native) -> h264 (h264_omx)) Press [q] to stop, [?] for help [1;36m[h264_omx @ 0x4bee0] [0mUsing OMX.hantro.H1.video.encoder [1;36m[h264_omx @ 0x4bee0] [0m[1;31merr 80001019 (-2147479527) on line 498 [0m[1;31mError initializing output stream 0:0 -- Error while opening encoder for output stream #0:0 - maybe incorrect parameters such as bit_rate, rate, width or height [0mConversion failed!```

[2017-09-24 08:56:16] unusuario128 : Do `logcat` now

[2017-09-24 08:58:04] jezzab : logcat | busybox grep h1_video_encoder shows nothing

[2017-09-24 08:58:58] unusuario128 : Oh! Wait... The H264 video **encoder** will accept only **raw** video (not compressed). What we need is the decoder.

[2017-09-24 08:59:14] jezzab : oh

[2017-09-24 08:59:53] jezzab : `./ffmpeg -i drop.avi -vcodec h264_omx output.avi` is my current command btw

[2017-09-24 08:59:55] unusuario128 : So, this means that what we should try is to get an H264 video and convert it to some raw uncompressed format with the OMX codec.

[2017-09-24 09:00:08] jezzab : riiiiight

[2017-09-24 09:00:40] jezzab : whats the -vcodec for 'raw'?

[2017-09-24 09:00:45] unusuario128 : And with `-vcodec h264_omx` you won't really want to get an avi file, but some sort of mp4...

[2017-09-24 09:01:04] unusuario128 : You should do something like this:

[2017-09-24 09:01:20] unusuario128 : `./ffmpeg -i drop.avi -vcodec h264_omx output.mp4`

[2017-09-24 09:01:42] unusuario128 : But, to test what we want to test, the command would be this:

[2017-09-24 09:01:57] unusuario128 : `./ffmpeg -i sample.mp4 -vcodec h264_omx output.avi`

[2017-09-24 09:02:24] jezzab : thats basically what i was doing before

[2017-09-24 09:02:27] jezzab : the last one

[2017-09-24 09:03:00] jezzab : results in the 80001019 error

[2017-09-24 09:03:04] jezzab : ill have a play

[2017-09-24 09:04:08] unusuario128 : When you get this error, run `logcat | busybox grep h1_video_encoder` to know what was the real issue.

[2017-09-24 09:04:50] unusuario128 : I'm not a ffmpeg user nor something like. In fact I only used it to see the information of video files.

[2017-09-24 09:04:54] unusuario128 : :disappointed:

[2017-09-24 09:05:48] jezzab : ```Stream mapping: Stream #0:0 -> #0:0 (h264 (native) -> h264 (h264_omx)) Stream #0:1 -> #0:1 (aac (native) -> ac3 (native)) Press [q] to stop, [?] for help [1;36m[h264_omx @ 0x493b0] [0mUsing OMX.hantro.H1.video.encoder [1;36m[h264_omx @ 0x493b0] [0m[1;31merr 80001019 (-2147479527) on line 498 [0m[1;31mError initializing output stream 0:0 -- Error while opening encoder for output stream #0:0 - maybe incorrect parameters such as bit_rate, rate, width or height [0mConversion failed! 1|root@wm220_dz_ah0001_v5:/data/ffmpeg/bin # logcat | busybox grep h1_video_encoder grep h1_video_encoder < E/h1_video_encoder(10750): API: Only uncompressed video accepted for input```

[2017-09-24 09:06:16] jezzab : same. my knowledge on it is limited. I have to goggle a lot. Most things ive used it with are GUI front end

[2017-09-24 09:06:17] unusuario128 : What is your command?

[2017-09-24 09:06:29] jezzab : `./ffmpeg -i snow.mp4 -vcodec h264_omx output.avi`

[2017-09-24 09:07:05] unusuario128 : `Stream #0:0 -> #0:0 (h264 (native) -> h264 (h264_omx))` means that you're using the codec for exactly the wrong purpose.

[2017-09-24 09:07:06] jezzab : but this reads to me as its trying to read in an MP4 and then use the h264_omx codec to save it as an avi

[2017-09-24 09:07:15] unusuario128 : This is the fail.

[2017-09-24 09:07:20] unusuario128 : Let me check the manpage.

[2017-09-24 09:07:33] jezzab : im gonna search for a raw uncompressed avi

[2017-09-24 09:09:11] unusuario128 : Can you try this? `ffmpeg -i snow.mp4 -c:v rawvideo -pix_fmt yuv420p out.yuv`

[2017-09-24 09:09:57] jezzab : ohhhh

[2017-09-24 09:09:59] jezzab : its decoding

[2017-09-24 09:10:22] jezzab : I had to use the bloody big video too didnt i

[2017-09-24 09:10:26] jezzab : ffs

[2017-09-24 09:10:39] jezzab : ```Stream mapping: Stream #0:0 -> #0:0 (h264 (native) -> rawvideo (native)) Press [q] to stop, [?] for help Output #0, rawvideo, to 'out.yuv': Metadata: major_brand : mp42 minor_version : 0 compatible_brands: mp42mp41 encoder : Lavf57.82.101 Stream #0:0(eng): Video: rawvideo (I420 / 0x30323449), yuv420p, 1920x1080, q=2-31, 745750 kb/s, 29.97 fps, 29.97 tbn, 29.97 tbc (default) Metadata: creation_time : 2017-03-28T04:06:12.000000Z handler_name : Alias Data Handler encoder : Lavc57.106.101 rawvideo frame= 90 fps=2.4 q=-0.0 size= 273152kB time=00:00:03.00 bitrate=745141.9kbits/s speed=0.0812x```

[2017-09-24 09:11:21] jezzab : hmm process died

[2017-09-24 09:11:34] jezzab : ill try the boot anim vid (video.264)

[2017-09-24 09:11:36] unusuario128 : You weren't using OMX at all...

[2017-09-24 09:11:53] unusuario128 : It used the native (CPU) h264 codec

[2017-09-24 09:12:16] jezzab : yup

[2017-09-24 09:12:27] unusuario128 : Try `./ffmpeg -vcodec h264_omx -i snow.mp4 output.avi`

[2017-09-24 09:12:32] jezzab : 2.4fps dec

[2017-09-24 09:13:24] unusuario128 : The `-vcodec` should be placed before the `-i` to set the codec of the input, if you place it after, it sets the output codec.

[2017-09-24 09:14:05] jezzab : Unknown decoder 'h264_omx'

[2017-09-24 09:14:14] jezzab : ah i see

[2017-09-24 09:15:54] unusuario128 : Have to leave for a while. Let me know if it works.

[2017-09-24 09:16:37] jezzab : ok

[2017-09-24 09:23:14] jezzab : `Unknown decoder 'h264_omx'`

[2017-09-24 09:28:37] jezzab : I converted an mp4 to raw yuv then tested using it as the input to make an mp4 with h264_omx:

[2017-09-24 09:28:41] jezzab : ```root@wm220_dz_ah0001_v5:/data/ffmpeg/bin # ./ffmpeg -s:v 1920x1080 -r 60 -i out.yuv -vcodec h264_omx output.mp4 <snip> [rawvideo @ 0x46200] [0m[0;33mEstimating duration from bitrate, this may be inaccurate Input #0, rawvideo, from 'out.yuv': Duration: 00:00:03.36, start: 0.000000, bitrate: 622884 kb/s Stream #0:0: Video: rawvideo (I420 / 0x30323449), yuv420p, 1920x1080, 622080 kb/s, 25 tbr, 25 tbn, 25 tbc Stream mapping: Stream #0:0 -> #0:0 (rawvideo (native) -> h264 (h264_omx)) Press [q] to stop, [?] for help [h264_omx @ 0x48a50] [0mUsing OMX.hantro.H1.video.encoder [h264_omx @ 0x48a50] [0m[1;31merr 80001019 (-2147479527) on line 498 Error initializing output stream 0:0 -- Error while opening encoder for output stream #0:0 - maybe incorrect parameters such as bit_rate, rate, width or height Conversion failed!```

[2017-09-24 09:29:57] jezzab : ```1|root@wm220_dz_ah0001_v5:/data/ffmpeg/bin # logcat | busybox grep h1_video_encoder grep h1_video_encoder```

[2017-09-24 09:34:00] unusuario128 : @jezzab: So, there is not OMX decoder?

[2017-09-24 09:34:22] jezzab : checking that right now with `./ffmpeg -decoders`

[2017-09-24 09:34:59] jezzab : Nope

[2017-09-24 09:40:12] unusuario128 : Well, what I've done until now is this: <https://github.com/crushedice2000/hantro-omx-shim>

[2017-09-24 09:40:31] unusuario128 : Now let me see how to implement the decoder if possible.

[2017-09-24 09:41:12] jezzab : But I would have thought that it should be able to encode the raw yuv file I tried to encode?

[2017-09-24 09:43:25] unusuario128 : Yes, it should be able to encode raw video, however it may be picky with the pixel, color or encoding format.

[2017-09-24 09:44:27] jezzab : Ok so your trying to use its own decoder so that it will be in a format that it accepts back in? Thats the plan for testing?

[2017-09-24 09:46:49] unusuario128 : Yes, but there isn't decoder... Or better yet, there should be one, but not exposed to ffmpeg.

[2017-09-24 09:47:06] jezzab : ok

[2017-09-24 10:06:53] unusuario128 : Just asking: is Dalvik running? Is there any chance of using JNI?

[2017-09-24 10:08:10] jezzab : How do I check?

[2017-09-24 10:08:37] jezzab : But from memory you can't install apk

[2017-09-24 10:12:26] unusuario128 : It shouldn't be needed. Only to have the ability if running Java code.

[2017-09-24 10:13:06] unusuario128 : Maybe it appears on `ps aux`?

[2017-09-24 10:13:22] unusuario128 : However I'm looking first for a fully native implementation of the decoder.

[2017-09-24 10:23:33] unusuario128 : Seems like it is not running. At least no zygote process.

[2017-09-24 10:28:25] unusuario128 : I'll try with rockchip's MPP, which seems the official low-lovel way to go.

[2017-09-24 10:53:15] unusuario128 : @jezzab: Can you play a bit with these files?

[2017-09-24 10:53:38] unusuario128 : You should be able to do "something" with they

[2017-09-24 10:58:22] unusuario128 : (in theory, the executables and libraries bundled there may allow you to play videos or at least try the hardware codec)

[2017-09-24 11:00:28] jezzab : ok ill give it a crack

[2017-09-24 11:07:04] jezzab : `CANNOT LINK EXECUTABLE: could not load library "librockchip_mpp.so" needed by "./mpi_enc_test"; caused by library "librockchip_mpp.so" not found`

[2017-09-24 11:07:26] jezzab : Missing a lib

[2017-09-24 11:09:00] jezzab : All require that lib

[2017-09-24 11:09:50] unusuario128 : Nope, is a link to libmpp.so

[2017-09-24 11:11:36] unusuario128 : You may also need to do something similar with the libvpu.

[2017-09-24 11:13:10] unusuario128 : `librockchip_vpu.so` -> `libvpu.so` `librockchip_mpp.so` -> `libmpp.so`

[2017-09-24 11:13:21] jezzab : Ahh ok

[2017-09-24 11:14:55] jezzab : will it look for them in ./lib or /system/lib ?

[2017-09-24 11:18:05] unusuario128 : In the library path. `/system/lib`

[2017-09-24 11:19:20] unusuario128 : Maybe is handier for you adding `./lib` to the `LD_LIBRARY_PATH` like this: `LD_LIBRARY_PATH="./lib:$LD_LIBRARY_PATH"`, however, I don't know if this will work.

[2017-09-24 11:20:27] jezzab : hmm is there any command line functions with these?

[2017-09-24 11:20:33] jezzab : they run but there is no output

[2017-09-24 11:20:38] jezzab : and no -h -? --help etc

[2017-09-24 11:20:53] jezzab : ```root@wm220_dz_ah0001_v5:/data/mpp # ./mpi_dec_test ./mpi_dec_test 1|root@wm220_dz_ah0001_v5:/data/mpp # ./mpi_dec_test --help ./mpi_dec_test --help 1|root@wm220_dz_ah0001_v5:/data/mpp # ./mpi_dec_test -? ./mpi_dec_test -? 255|root@wm220_dz_ah0001_v5:/data/mpp # ./mpi_dec_test -h ./mpi_dec_test -h 127|root@wm220_dz_ah0001_v5:/data/mpp # ./mpi_test ./mpi_test 255|root@wm220_dz_ah0001_v5:/data/mpp # ./mpi_test -h ./mpi_test -h```

[2017-09-24 11:21:14] unusuario128 : mpi_dec_test should have help. Let me check the code.

[2017-09-24 11:21:55] unusuario128 : ```static OptionInfo mpi_dec_cmd[] = { {"i", "input_file", "input bitstream file"}, {"o", "output_file", "output bitstream file, "}, {"w", "width", "the width of input bitstream"}, {"h", "height", "the height of input bitstream"}, {"t", "type", "input stream coding type"}, {"d", "debug", "debug flag"}, {"x", "timeout", "output timeout interval"}, };```

[2017-09-24 11:24:05] jezzab : ```root@wm220_dz_ah0001_v5:/data/mpp # ./mpi_dec_test -i video.264 -o test.out -w 1920 -h 1080 -d eo.264 -o test.out -w 1920 -h 1080 -d```

[2017-09-24 11:24:22] jezzab : unsure on 'type'

[2017-09-24 11:25:17] jezzab : ok without -d test.out is written 0 bytes

[2017-09-24 11:32:30] jezzab : hmm I cant get any to do anything bar a 0byte file :confused:

[2017-09-24 11:34:45] jezzab : if I use the -t with say -t h264 it doesnt save anything

[2017-09-24 11:37:37] jezzab : Nah there is something wrong

[2017-09-24 11:37:56] jezzab : Ive checked the source and if it cant find the input file name it should print an error. It doesnt

[2017-09-24 11:38:17] jezzab : ```if (cmd->have_input) { data.fp_input = fopen(cmd->file_input, "rb"); if (NULL == data.fp_input) { mpp_err("failed to open input file %s\n", cmd->file_input); goto MPP_TEST_OUT; }```

[2017-09-24 11:38:56] jezzab : ```255|root@wm220_dz_ah0001_v5:/data/mpp # ./mpi_dec_test -i foo.mp4 -o test.avi -t 264 -w 1920 -h 1080 oo.mp4 -o test.avi -t 264 -w 1920 -h 1080 < 255|root@wm220_dz_ah0001_v5:/data/mpp #```

[2017-09-24 11:39:21] jezzab : `foo.mp4` does not exist

[2017-09-24 11:48:22] unusuario128 : Let's add some debug...

[2017-09-24 11:49:42] unusuario128 : Or, better yet, I'll try to build the mpp into ffmpeg to try it as library

[2017-09-24 15:00:57] unusuario128 : About the test stuff, the type argument should be 7 for H264.

[2017-09-24 15:18:53] unusuario128 : Talking about serious things: I've built `ffmpeg` with support for the rockchip hardware codec with MPP.

[2017-09-24 15:19:19] unusuario128 : Be careful because the `lib` folder has some mandatory dependencies.

[2017-09-24 15:19:52] unusuario128 : You should be able to use it to **decode** any H264 video.

[2017-09-24 15:21:09] unusuario128 : Exactly, the decoder name is `h264_rkmpp`

[2017-09-24 15:22:05] unusuario128 : Theoretically, you should be able to run the following command (or anything similar) successfully:

[2017-09-24 15:22:10] unusuario128 : `./ffmpeg -vcodec h264_rkmpp -i snow.mp4 output.avi`

[2017-09-24 15:24:52] unusuario128 : If that behaves as expected, you may be able to output the video to any framebuffer referring to this example:

[2017-09-24 15:24:53] unusuario128 : <https://www.ffmpeg.org/ffmpeg-devices.html#Examples-13>

[2017-09-24 15:25:17] unusuario128 : You should choose the correct pixel format, though.

[2017-09-24 15:42:48] unusuario128 : Also, sorry for the intensive test requests, but when I reach a point where I can't test something, I need to rely on you. I hope that this time it works as it should.

[2017-09-24 20:27:00] unusuario128 : Tomorrow I'll be traveling, however I hope to have time to, at least, check slack.

[2017-09-24 20:28:16] unusuario128 : As my workstation will be fixed at home, I won't be able to do much, however, I'll try to leave open the ssh port and cross my fingers to "avoid" one of the frequent power outages.

[2017-09-24 20:37:35] kilrah : ugh :confused: UPS?

[2017-09-24 22:21:43] jezzab : No worries @unusuario128 I'll test today. And re: the intensive testing. I don't mind at all, this is beyond my skill set so am happy to do anything I can to help, which I could do more. And I appreciate all the help you have given and the time you have put in.

[2017-09-24 22:47:05] jezzab : ```root@wm220_dz_ah0001_v5:/data/ffmpeg/bin # ./ffmpeg ./ffmpeg CANNOT LINK EXECUTABLE: could not load library "libavdevice.so.57" needed by "./ffmpeg"; caused by could not load library "libavfilter.so.6" needed by "libavdevice.so.57"; caused by could not load library "libswscale.so.4" needed by "libavfilter.so.6"; caused by could not load library "libavutil.so.55" needed by "libswscale.so.4"; caused by could not load library "libdrm.so" needed by "libavutil.so.55"; caused by cannot locate symbol "log2" referenced by "libdrm.so"...```

[2017-09-24 22:49:12] jezzab : as suggested, doing this made things a lot easier with the libs too `export LD_LIBRARY_PATH="/data/ffmpeg/lib/:$LD_LIBRARY_PATH"`

[2017-09-25 04:22:56] unusuario128 : @jezzab: Did you finally get the hardware codec working?

[2017-09-25 04:25:54] jezzab : as above. It all seems to fail because of `caused by cannot locate symbol "log2" referenced by "libdrm.so`

[2017-09-25 04:26:18] jezzab : unless ive done something wrong this time

[2017-09-25 04:26:50] unusuario128 : Let me check. 6:28 AM here. :grinning:

[2017-09-25 04:27:01] jezzab : Hope you have a coffee!

[2017-09-25 04:27:31] jezzab : 2:28pm here. couple of hours and it will be beer o'clock

[2017-09-25 04:30:01] unusuario128 : Is it android KitKat?

[2017-09-25 04:30:36] jezzab : yup

[2017-09-25 04:30:37] jezzab : 4.4

[2017-09-25 04:30:59] jezzab : but chopped up a bit

[2017-09-25 04:31:38] jezzab : I only found one reference to that error googling

[2017-09-25 04:32:26] unusuario128 : Then, seems like something went wrong. You don't have the `log2` function.

[2017-09-25 04:33:44] unusuario128 : I'll attach you the complete `android-19` sysroot.

[2017-09-25 04:34:02] unusuario128 : However, you only need to use libm, which is attached as separate file.

[2017-09-25 04:34:38] jezzab : ok ill just get rid of this Mavic and put the goggles on.

[2017-09-25 04:35:31] unusuario128 : I have to take a quick breakfast and do a road trip of ~1.5h, until I'll arrive, even if ? will be far from the desk, I can probably help you.

[2017-09-25 04:35:37] unusuario128 : See you soon!

[2017-09-25 04:35:46] unusuario128 : (and good luck with that)

[2017-09-25 04:35:57] jezzab : No worries :slightly_smiling_face: fingers crossed

[2017-09-25 04:39:52] jezzab : same error :disappointed:

[2017-09-25 04:40:05] jezzab : `caused by cannot locate symbol "log2" referenced by "libdrm.so"`

[2017-09-25 04:45:39] jezzab : hmm something strange too. when i link the custom lib folder with libm.so in it, it kills `ls`

[2017-09-25 04:45:58] jezzab : ```1|root@wm220_dz_ah0001_v5:/data/ffmpeg/bin # ls ls CANNOT LINK EXECUTABLE: could not load library "libcutils.so" needed by "ls"; caused by could not load library "liblog.so" needed by "libcutils.so"; caused by could not load library "libm.so" needed by "liblog.so"; caused by empty/missing DT_HASH in "libm.so" (built with --hash-style=gnu?)```

[2017-09-25 04:46:18] jezzab : wtf

[2017-09-25 04:47:06] jezzab : ahh its part of the system already :confused:

[2017-09-25 04:49:19] jezzab : only reference i could find @unusuario128 was this `when I build libdrm b556ea1, I need to add "LDFLAGS=-lm", or I will get ../.libs/libdrm.so: undefined reference to 'log2' `

[2017-09-25 05:44:24] jezzab : hmm

[2017-09-25 06:17:53] jezzab : <https://vendor2.nginfotpdx.net/gitlab/upstream/drm/commit/51d72a619e8648544053c0c0f56a88eeef601cff>

[2017-09-25 08:00:36] unusuario128 : Did you try the libdrm.tar.gz attached?

[2017-09-25 08:02:38] unusuario128 : It contains libdrm linked against libm

[2017-09-25 08:04:33] unusuario128 : It should solve all your issues.

[2017-09-25 08:05:04] unusuario128 : I'm now using Slack for Android. :grinning: I'll check later you answer.

[2017-09-25 08:39:02] jezzab : but how do I use the `libdrm.a` ? I've always just seen them as left overs from compiling

[2017-09-25 08:44:42] jezzab : Thats whats thrown me lol I

[2017-09-25 08:45:18] jezzab : going to need some spoon feeding :confused:

[2017-09-25 11:02:34] unusuario128 : `.a` = static library

[2017-09-25 11:02:54] unusuario128 : `.so` = "dynamic" library (shared)

[2017-09-25 11:03:14] unusuario128 : Try the .so files attached.

[2017-09-25 11:03:51] unusuario128 : Sorry for the mistake, i thought that you would use `ar` and `gcc` if neeed, but forgot about the toolchain cross-compile stuff.

[2017-09-25 11:04:24] unusuario128 : Traveling again. In approximately 1h if you're awake yet I'll see if you encountered some problems more. I hope that not.

[2017-09-25 11:04:57] jezzab : all good

[2017-09-25 11:04:58] jezzab : trying now

[2017-09-25 11:05:21] unusuario128 : If you go to sleep before I reach my next destination, and you don't get more issues, plase benchmark the hardware codec.

[2017-09-25 11:05:29] unusuario128 : Bye! and good luck with that.

[2017-09-25 11:05:33] unusuario128 : :smile:

[2017-09-25 11:05:39] jezzab : ohhh

[2017-09-25 11:05:44] jezzab : runs!

[2017-09-25 11:05:57] jezzab : and i have color coz im using dropbear ssh server on it now lol

[2017-09-25 12:06:41] jezzab : The `h264_rkmpp` hardware decoder isn't working

[2017-09-25 12:07:12] jezzab : also trying to load one of the demo vids get "Aborted", even running ffprobe on the demo vids. The boot anim raw h264 is fine though

[2017-09-25 12:11:19] jezzab : ```69|root@wm220_dz_ah0001_v5:/data/ffmpeg/bin # ./ffmpeg -vcodec h264_rkmpp -i video.264 -f null - ffmpeg version N-87384-gee8067a Copyright (c) 2000-2017 the FFmpeg developers built with gcc 4.9.x (GCC) 20150123 (prerelease) configuration: --enable-cross-compile --enable-shared --disable-static --arch=arm --target-os=linux --enable-parsers --enable-shared --enable-gpl --enable-version3 --enable-protocols --enable-nonfree --enable-pthreads --enable-pic --enable-hardcoded-tables --enable-rkmpp --enable-libdrm --enable-decoder=mpegvideo --enable-decoder=mpeg1video --enable-decoder=mpeg2video --enable-decoder=mjpeg --enable-decoder=mjpegb --enable-decoder=mpeg4 --enable-decoder=h264 --enable-decoder=theora --enable-decoder=vc1 --enable-decoder=wmv3 --enable-decoder=vp6 --enable-decoder=vp6f --enable-decoder=vp8 --enable-decoder=opus --cross-prefix=arm-linux-androideabi- --sysroot=/home/casa/android-ndk-r15c/platforms/android-19/arch-arm --sysinclude=/home/casa/android-ndk-r15c/platforms/android-19/arch-arm/usr/include --prefix=./prefix --pkg-config=/home/casa/android-ndk-r15c/sources/mpp/build/android/arm/FFmpeg/pkgconfig.sh libavutil 55. 76.100 / 55. 76.100 libavcodec 57.106.101 / 57.106.101 libavformat 57. 82.101 / 57. 82.101 libavdevice 57. 8.101 / 57. 8.101 libavfilter 6.106.100 / 6.106.100 libswscale 4. 7.103 / 4. 7.103 libswresample 2. 8.100 / 2. 8.100 libpostproc 54. 6.100 / 54. 6.100 [h264 @ 0x441d0] Stream #0: not enough frames to estimate rate; consider increasing probesize Input #0, h264, from 'video.264': Duration: N/A, bitrate: N/A Stream #0:0: Video: h264 (High), yuv420p(tv, bt709, progressive), 1920x1080, 60 fps, 60 tbr, 1200k tbn, 120 tbc Stream mapping: Stream #0:0 -> #0:0 (h264 (h264_rkmpp) -> wrapped_avframe (native)) Press [q] to stop, [?] for help Error while decoding stream #0:0: Try again7014:32:22.77 bitrate=N/A speed=N/A Last message repeated 323 times Finishing stream 0:0 without any data written to it.2.77 bitrate=N/A speed=N/A Output #0, null, to 'pipe:': Metadata: encoder : Lavf57.82.101 Stream #0:0: Video: wrapped_avframe, yuv420p, 1920x1080, q=2-31, 200 kb/s, 60 fps, 60 tbn, 60 tbc Metadata: encoder : Lavc57.106.101 wrapped_avframe frame= 0 fps=0.0 q=0.0 Lsize=N/A time=00:00:00.00 bitrate=N/A speed= 0x```

[2017-09-25 12:12:07] jezzab : so thats a simple benchmark to null using the decoder but no encoder. It just sits there doing nothing. If I dont use the h264_rkmpp decoder it runs fine (around 28fps. CPU decoded)

[2017-09-25 12:12:42] jezzab : Heres the log from logcat while its running:

[2017-09-25 12:12:50] jezzab : ```E/mpp_buffer(18155): mpp_buffer_create failed to create buffer with size 2097152 E/mpp_ion (18155): ion_ioctl c0144900 failed with code -1: No such device E/mpp_ion (18155): allocator_ion_alloc ion_alloc failed ret -19 E/mpp_ion (18155): ion_ioctl c0144900 failed with code -1: No such device E/mpp_ion (18155): ion_ioctl c0144900 failed with code -1: No such device E/mpp_buffer(18155): mpp_buffer_create failed to create buffer with size 2097152 E/mpp_ion (18155): ion_ioctl c0144900 failed with code -1: No such device E/mpp_ion (18155): ion_ioctl c0144900 failed with code -1: No such device E/mpp_ion (18155): allocator_ion_alloc ion_alloc failed ret -19 E/mpp_ion (18155): ion_ioctl c0144900 failed with code -1: No such device E/mpp_ion (18155): allocator_ion_alloc ion_alloc failed ret -19 E/mpp_buffer(18155): mpp_buffer_create failed to create buffer with size 2097152 E/mpp_ion (18155): ion_ioctl c0144900 failed with code -1: No such device E/mpp_ion (18155): allocator_ion_alloc ion_alloc failed ret -19 E/mpp_ion (18155): ion_ioctl c0144900 failed with code -1: No such device```

[2017-09-25 12:13:26] jezzab : closer! a lot closer :slightly_smiling_face:

[2017-09-25 12:16:07] jezzab : also the kernel kills ffmpeg if it starts getting too memory hungry

[2017-09-25 22:36:40] jezzab : Any idea's @unusuario128? Seems like the ion memory management?

[2017-09-25 22:48:53] hostile : strace it

[2017-09-25 22:48:57] hostile : lets see what ion is bitchign abotu

[2017-09-25 22:49:55] jezzab : ill give it a go. Gotta charge em up a bit. Finally went flat. New record, 13 hours

[2017-09-25 23:04:01] jezzab : ```open("/dev/ion", O_RDWR) = 10 open("/proc/device-tree", O_RDONLY|O_DIRECTORY) = -1 ENOENT (No such file or directory) writev(4, [{"\4", 1}, {"mpp_ion\0", 8}, {"scan /proc/device-tree for vpu_service return -1\n\0", 50}], 3) = 59 open("/proc/device-tree", O_RDONLY|O_DIRECTORY) = -1 ENOENT (No such file or directory) writev(4, [{"\4", 1}, {"mpp_ion\0", 8}, {"scan /proc/device-tree for hevc_service return -1\n\0", 51}], 3) = 60 open("/proc/device-tree", O_RDONLY|O_DIRECTORY) = -1 ENOENT (No such file or directory) writev(4, [{"\4", 1}, {"mpp_ion\0", 8}, {"scan /proc/device-tree for rkvdec return -1\n\0", 45}], 3) = 54 open("/proc/device-tree", O_RDONLY|O_DIRECTORY) = -1 ENOENT (No such file or directory) writev(4, [{"\4", 1}, {"mpp_ion\0", 8}, {"scan /proc/device-tree for rkvenc return -1\n\0", 45}], 3) = 54 open("/proc/device-tree", O_RDONLY|O_DIRECTORY) = -1 ENOENT (No such file or directory) writev(4, [{"\4", 1}, {"mpp_ion\0", 8}, {"scan /proc/device-tree for vpu_combo return -1\n\0", 48}], 3) = 57 open("/sys/kernel/debug/ion/heaps", O_RDONLY|O_DIRECTORY) = -1 ENOENT (No such file or directory) writev(4, [{"\4", 1}, {"mpp_ion\0", 8}, {"scan /sys/kernel/debug/ion/heaps for vmalloc return -1\n\0", 56}], 3) = 65 open("/sys/kernel/debug/ion/heaps", O_RDONLY|O_DIRECTORY) = -1 ENOENT (No such file or directory) writev(4, [{"\4", 1}, {"mpp_ion\0", 8}, {"scan /sys/kernel/debug/ion/heaps for system-heap return -1\n\0", 60}], 3) = 69 writev(4, [{"\6", 1}, {"mpp_ion\0", 8}, {"can not find any hint from all possible devices\n\0", 49}], 3) = 58 writev(4, [{"\4", 1}, {"mpp_ion\0", 8}, {"using ion heap ION_HEAP_TYPE_CARVEOUT\n\0", 39}], 3) = 48 open("/dev/ion", O_RDWR) = 11```

[2017-09-26 00:53:57] hostile : looks like you missing some services it wants

[2017-09-26 00:54:21] hostile : <http://lkml.iu.edu/hypermail/linux/kernel/1608.3/01722.html>

[2017-09-26 03:51:47] jezzab : Im interested in this: `Assertion p_hal->vcodec_type & ((0x00000200) | (0x00000001) | (0x00000002)) failed at hal_h264d_init:187`

[2017-09-26 03:52:16] jezzab : I think its referring to `mpp_assert(p_hal->vcodec_type & (HAVE_RKVDEC | HAVE_VPU1 | HAVE_VPU2));`

[2017-09-26 03:54:38] hostile : <https://irclog.whitequark.org/linux-rockchip/2017-08-02>

[2017-09-26 03:54:54] hostile : “08:58 <Myy> Alright, turns out that -DHAVE_DRM=ON and -DRKPLATFORM=ON were needed ”

[2017-09-26 03:55:00] jezzab : You stalkin me

[2017-09-26 03:55:03] jezzab : I was reading that lol

[2017-09-26 03:55:15] hostile : similar google foo

[2017-09-26 03:55:27] hostile : I mean… I am tapped into the DJI remote tracking server LOL

[2017-09-26 03:55:32] hostile : I can see your every URL post

[2017-09-26 03:55:34] jezzab : BWAHAHA

[2017-09-26 03:55:34] hostile : LOLOLol

[2017-09-26 03:55:44] jezzab : Shit I was using WeChat!! How did you see me!

[2017-09-26 03:56:09] hostile : Official government request spoofed.

[2017-09-26 03:56:36] jezzab : I for one welcome our new Chinese overlords

[2017-09-26 03:56:41] jezzab : *looks around*

[2017-09-26 03:57:12] jezzab : This GPU stuff has become a mission. Its close but wow.

[2017-09-26 03:57:34] jezzab : But once its under control it does offer a lot of possibilities

[2017-09-26 03:58:09] jezzab : Be nice if they had wifi though

[2017-09-26 03:58:34] jezzab : I tried OTG with them. Not the same as Mavic, they dont allow OTG

[2017-09-26 08:27:32] unusuario128 : Hello, @jezzab

[2017-09-26 08:28:36] jezzab : Hey mate :slightly_smiling_face:

[2017-09-26 08:29:01] unusuario128 : I'm out of my main home and in Android mode. For more INRI, I have now 4.4 on the tablet and therefore I can't use Termux. However, I'll try to find something to ssh home.

[2017-09-26 08:29:18] unusuario128 : Let me check if I can do some now.

[2017-09-26 08:29:25] jezzab : Ok coool!

[2017-09-26 08:29:38] jezzab : Ive included some straces and things if they help

[2017-09-26 08:29:46] jezzab : its very very close

[2017-09-26 08:46:31] codeforge : is there a way to record video footage in goggles microSD when flying with spark?

[2017-09-26 08:55:35] unusuario128 : Tmate web won't accept keybord input on Android, and their ssh server seems to be picky about the key negociation system. Do you know of any ready-make apk of dbclient for android?

[2017-09-26 08:56:44] unusuario128 : @codeforge: Maybe reading from fb0... I don't know if there is an easy way.

[2017-09-26 09:14:31] jezzab : That sucks

[2017-09-26 09:14:53] jezzab : Looking for an apk but no luck yet

[2017-09-26 09:18:34] jezzab : <https://play.google.com/store/apps/details?id=com.sonelli.juicessh&hl=en>

[2017-09-26 09:18:43] jezzab : havent used it personally but comes recommended

[2017-09-28 20:48:55] unusuario128 : @jezzab: Sorry. I'm currently _very_ busy and yet away of my main home. To broaden my ~~steamy~~ STEAM education I'm learning serigraphy... Once I relocate completely to the new home and I adapt to the new life cycle (I hope that soon) I'll continue with this. Is awful to be in Android mode every day every time and without any fully usable computer.

[2017-09-28 20:49:50] hostile : steamy education sounds so much better! lol

[2017-09-28 20:49:54] hostile : ow ow

[2017-09-29 01:39:31] jezzab : All good mate. Steamy lol

[2017-09-29 08:09:09] unusuario128 : Oopsie. Meant the first and formal meaning. Lol.

[2017-09-29 08:09:40] unusuario128 : I didn't knew the second

[2017-09-29 08:10:45] unusuario128 : ...handicaps of not being an native speaker.

[2017-09-29 08:16:16] unusuario128 : My English is self-taught and learned in < 9 years of formal/technical usage via Internet. Mainly reading and, sporadically chatting and/or listening some videos.

[2017-09-29 08:18:30] unusuario128 : My father started to teach me the ABC of the English language when I was 7 and only a bit later I started to get interested in STEM, things, specifically computing science (what an awful name) and electronics.

[2017-09-29 08:20:03] unusuario128 : ~~blatantly off-topic~~ for the good stuff I'll need to bring the workstation to the new home. I hope that this weekend. So much kilometers and I don't have age to drive yet :-(

[2017-09-29 09:47:03] jezzab : Your English is very very good @unusuario128, you shouldn't worry :slightly_smiling_face:

[2017-09-29 13:40:13] hostile : @unusuario128 you speak just as well as many American’s I know that are unable to formally speak their own language with poise. =]

[2017-09-30 23:22:24] codeforge : can i use conditionedRace to root goggles?

[2017-10-01 01:32:52] hostile : @codeforge I’m not sure anyone has tried... your need to find the correct module, but yes I don’t see any reason why not!

[2017-10-05 09:51:41] coldflake : Hello guys!

[2017-10-05 09:55:46] jezzab : Welcome!

[2017-10-05 15:26:58] pathogen : :spock-hand:

[2017-10-08 07:59:20] unusuario128 : @jezzab Sorry for the late reply. Can you post the output of these commands? `ls /proc/device-tree` `ls /sys/firmware/devicetree/base` `ls /dev/ion` `/sys/kernel/debug/ion/*`

[2017-10-08 07:59:44] unusuario128 : I was a bit busy. _A bit_ to say something

[2017-10-08 08:00:21] jezzab : Can do. I'll fire them up. They have been moved back to the case. Give me a few mins

[2017-10-08 08:00:23] unusuario128 : Also, pleaase try the libraries above.

[2017-10-08 08:06:21] jezzab : ```root@wm220_dz_ah0001_v5:/ # ls /proc/device-tree /proc/device-tree: No such file or directory ```

[2017-10-08 08:06:37] jezzab : ```130|root@wm220_dz_ah0001_v5:/ # ls /sys/firmware/devicetree/base /sys/firmware/devicetree/base: No such file or directory```

[2017-10-08 08:07:03] jezzab : ```130|root@wm220_dz_ah0001_v5:/ # ls -la /dev/ion crw------- root root 10, 61 1980-01-01 00:00 ion```

[2017-10-08 08:08:07] jezzab : ```130|root@wm220_dz_ah0001_v5:/ # ls /sys/kernel/debug/ion /sys/kernel/debug/ion: No such file or directory```

[2017-10-08 08:08:29] unusuario128 : Maybe that kernel was built without DT support, or, at least, without support for exposing the device tree to the userspace via `sysfs`

[2017-10-08 08:08:31] jezzab : `/sys/kernel/debug` does exist but nothing in it

[2017-10-08 08:09:14] unusuario128 : Can you try to use the libraries above for the MPP stuff?

[2017-10-08 08:09:29] jezzab : ok

[2017-10-08 08:20:01] jezzab : No good with the replacement libs

[2017-10-08 08:20:05] jezzab : ```root@wm220_dz_ah0001_v5:/data/ffmpeg/bin # ./ffmpeg -vcodec h264_rkmpp -i video.264 -f null - ffmpeg version N-87384-gee8067a Copyright (c) 2000-2017 the FFmpeg developers built with gcc 4.9.x (GCC) 20150123 (prerelease) configuration: --enable-cross-compile --enable-shared --disable-static --arch=arm --target-os=linux --enable-parsers --enable-shared --enable-gpl --enable-version3 --enable-protocols --enable-nonfree --enable-pthreads --enable-pic --enable-hardcoded-tables --enable-rkmpp --enable-libdrm --enable-decoder=mpegvideo --enable-decoder=mpeg1video --enable-decoder=mpeg2video --enable-decoder=mjpeg --enable-decoder=mjpegb --enable-decoder=mpeg4 --enable-decoder=h264 --enable-decoder=theora --enable-decoder=vc1 --enable-decoder=wmv3 --enable-decoder=vp6 --enable-decoder=vp6f --enable-decoder=vp8 --enable-decoder=opus --cross-prefix=arm-linux-androideabi- --sysroot=/home/casa/android-ndk-r15c/platforms/android-19/arch-arm --sysinclude=/home/casa/android-ndk-r15c/platforms/android-19/arch-arm/usr/include --prefix=./prefix --pkg-config=/home/casa/android-ndk-r15c/sources/mpp/build/android/arm/FFmpeg/pkgconfig.sh libavutil 55. 76.100 / 55. 76.100 libavcodec 57.106.101 / 57.106.101 libavformat 57. 82.101 / 57. 82.101 libavdevice 57. 8.101 / 57. 8.101 libavfilter 6.106.100 / 6.106.100 libswscale 4. 7.103 / 4. 7.103 libswresample 2. 8.100 / 2. 8.100 libpostproc 54. 6.100 / 54. 6.100 [h264 @ 0x441d0] Stream #0: not enough frames to estimate rate; consider increasing probesize Input #0, h264, from 'video.264': Duration: N/A, bitrate: N/A Stream #0:0: Video: h264 (High), yuv420p(tv, bt709, progressive), 1920x1080, 60 fps, 60 tbr, 1200k tbn, 120 tbc [h264_rkmpp @ 0x56ef20] Failed to assign buffer group (code = -1) [h264_rkmpp @ 0x56ef20] Failed to initialize RKMPP decoder. Stream mapping: Stream #0:0 -> #0:0 (h264 (h264_rkmpp) -> wrapped_avframe (native)) Error while opening decoder for input stream #0:0 : Unknown error occurred ```

[2017-10-08 08:23:51] unusuario128 : This time it was built with `-DHAVE_DRM=ON` and `-DRKPLATFORM=ON`

[2017-10-08 08:33:04] jezzab : ok got an strace of it

[2017-10-08 11:37:47] unusuario128 : Seems like there are some device nodes missing.

[2017-10-08 11:39:06] unusuario128 : Maybe the kernel was built without support for exposing the hardware decoders in `/dev` or maybe we need some trick to ccreate them

[2017-10-08 11:39:16] unusuario128 : Can you `ls /dev`?

[2017-10-08 11:46:22] unusuario128 : ...Maybe our problem has something to do with this: <https://github.com/Miouyouyou/rockchip-vcodec/>

[2017-10-08 11:50:46] unusuario128 : What `uname -r`? Does it use 4.X or 3.X kernel?

[2017-10-08 12:51:37] jezzab : Sorry mate, about to hit the hay but:

[2017-10-08 12:51:47] jezzab : kernel is `3.10.62`

[2017-10-08 15:25:34] unusuario128 : Sleeping is a good habit. I also do that once each day and for no less than 9 hours.

[2017-10-08 15:26:31] unusuario128 : Seems like something went wrong with the vpu.

[2017-10-08 15:48:27] unusuario128 : The only related files that I can see on `/dev` are the Hantro x170 ones.

[2017-10-08 15:50:13] unusuario128 : And the Hantro x170 could maybe used via `gstreamer`, which may become a little nightmare if I decide to compile it for Android. However it may be the preferred way to go.

[2017-10-08 15:52:28] unusuario128 : I don't know to which extent will be useful to try it, however it looks promising. The only thing I have unclear yet is if the x170 GStreamer plugin that was originally written for the Atmel SoC will also fit on Rockchip, which seems to use some sort of obscure proccess for the video acceleration stuff.

[2017-10-08 15:56:51] unusuario128 : I have a really awful schedule at the lyceum (or high scool or whatever it is called generally) where I start at 13:35 the most of the days and then I spend the next 6+ hours there. This is good because you go to sleep approximately at 13:00 or so, then the morning time that I get free after doing the basic things can be happily invested here. :slightly_smiling_face:

[2017-10-08 15:57:12] unusuario128 : All the times are CEST

[2017-10-08 15:57:44] unusuario128 : See you on monday!

[2017-10-22 14:10:03] unusuario128 : @jezzab: I'm very sorry. Finally I hadn't time to do anything. Between the studies, the flu (pity that isn't a computer virus) and the Fauré Requiem op. 48, I'm nearly K.O. :disappointed:

[2017-10-22 14:25:27] hostile : yeah I have been busy as shit too!

[2017-10-22 21:47:23] jezzab : All good mate. I’ve been busy popping the hood on the DJI iOS app

[2017-10-26 06:49:26] pathogen : As we have seen recent success in custom firmware using mixed modules, is there any hope in applying such a technique, to the end of having the Mavic on a firmware higher than 0700, while being able to start up for a long range flight without the app, and fly truly goggles-only? Is there any hope said restriction is due to a specific module? I ask because I'm hoping in a future firmware the head tracking is made smooth enough to use for cinematic shots.

[2017-10-26 12:19:37] b_hargis : @pathogen I would assume using only the FC from 400, that all the goggles pieces would work correctly using mixed module since the functionality supporting them is coming from the latest firmware. Why don't you try it? (Or mail me your goggles and I will try it)

[2017-10-26 21:14:58] pathogen : Thanks! No need to try it just yet though as there is no confirmation of actual smooth head tracking.

[2017-10-26 21:15:30] jezzab : In the later goggles fw you can slow down the speed

[2017-10-26 21:19:03] pathogen : Yea I did try G-1000 MP-0400 but while there was indeed ability to dampen things, the tracking still 'clunks' to a stop at the last moment and 'lurches' from stand still. It will need to behave exactly like the RC wheel to be useful for cinematic dolly tracking shots.

[2017-10-26 21:19:57] jezzab : Im surpised it worked at all

[2017-10-26 21:20:15] jezzab : If I have the MP on 700 and goggles on 1000 it breaks the head tracking completely

[2017-10-26 21:21:00] jezzab : and 400 didnt support goggles?

[2017-10-26 21:21:31] pathogen : I was too! Actually the yaw did indeed break but the pitch did function, which is actually EXACTLY the best setup for that type of shot, sadly the jerking issue persists.

[2017-10-26 21:23:14] jezzab : It would be a good test to do a clean .1000 AC and .1000 GL

[2017-10-26 21:26:11] pathogen : I have refrained from pushing MP beyond 0700 because as I understand that will introduce the requirement to always start the GO app preflight, thus requiring a phone while in the field.

[2017-10-26 21:26:48] jezzab : I purely mean as a test case

[2017-10-26 21:30:56] pathogen : I did want to test 1000 without the FC modules but since I'm the biggest chicken around Im just waiting for someone else to do it. :rooster:

[2017-10-27 08:28:26] pathogen : Well I found my big boy pants and flashed AC RC and G to 0000, the AC firmware with deleted 305, 306, 905 modules. Flashes AC twice. Everything boots up and appears functional, however I am unable to do a dumldore backup of AC, it writes 2 module files and then stops reporting it has lost ftp connection, even though I can see the ftp is still available via explorer.

[2017-10-27 08:28:56] jezzab : What version of DD?

[2017-10-27 08:29:01] pathogen : 1.92

[2017-10-27 08:29:11] jezzab : Must be one hell of a time out

[2017-10-27 08:29:21] jezzab : DUMLdone wont time out for 2 minutes

[2017-10-27 08:29:44] pathogen : I don't recall the exact delay but that does sound about right

[2017-10-27 08:29:47] jezzab : And i know it works coz i tested backup on .0000 the day it came out

[2017-10-27 08:30:18] jezzab : The longest a file takes to do its whole encryption bullshit on the bird is around 60 secs. I just went overkill

[2017-10-27 08:30:30] jezzab : try and download the same file by hand with ftp

[2017-10-27 08:30:38] jezzab : why do you need a backup anyway?

[2017-10-27 08:31:34] pathogen : hmm well i think I was going to see if I could so some sort of manual verify on the modules. I will try that manual copy from ftp, good thinkin'

[2017-10-27 08:33:13] pathogen : Is there any possible reason you can imagine if including 905 (nfz) in the module deletion would affect it?

[2017-10-27 08:33:50] jezzab : Maybe try one of the other channels mate. I have no idea sorry

[2017-10-27 08:34:21] pathogen : roger that, thanks and cheers!

[2017-10-27 08:43:22] pathogen : I can't figure out the path for the firmware. I see some sigs in <ftp://192.168.42.2/upgrade/upgrade/> but that cant be right, its missing the old 305,306,905 ones in there. Even so I get some timeout errors while copying the larger sigs from there. The timeout is happening in about 15 seconds after trying to copy a file larger than 14MB

[2017-10-27 08:57:20] haloweenhamster : With 400 305 & 306 do you need to worry about 905?

[2017-10-27 08:57:48] pathogen : no idea i just didnt see a reason to keep it

[2017-10-27 08:58:33] pathogen : ok now im ftp via IE and I see a few sig.tmp files, that cant be good! one of them doesnt even have the same file size as its original

[2017-10-27 08:58:55] pathogen : IE isnt timing out on large file copy though

[2017-10-27 08:59:44] pathogen : btw it was from a 0700 rotted ac rather than a 400

[2017-10-27 09:10:18] pathogen : maybe i should see sigs in /dji/device/firmware/release/module? empty folder though

[2017-10-27 09:11:48] pathogen : lol im going to stop spamming this goggles channel now, thankyou and good day dont eat yellow snow, a bird in the hand is worth two in the bush

[2017-10-27 19:03:39] kilrah : I don't really understand why that's a problem. Skipping on having a phone as a config interface and backup display (the goggles DO occasionally crash/reboot) doesn't make much sense

[2017-10-28 03:17:44] pathogen : I dont mind if the goggles crash. Ill just shut the RC off to trigger an RTH. I hate the drain my phone puts on the RC, without it the RC battery barely drops. And so I did try the 0000 with 305/306 removed, and flew with goggles only out beyond 1KM no problem. So that's that!

[2017-11-09 23:44:41] yegoshin.uas : So has anyone tried goggles on higher firmware then .700 (to be able to adjust sensitivity) and AC with firmware that matches (with deleted modules of course) so that Goggles actually work correctly with head tracking modes and tilt works?

[2017-11-10 00:14:02] nocommie : I havent even powered on my goggles in ages. I have mixed modules on my mav. I can test it and see. Not sure if it the sens adj is there or not with app 413

[2017-11-10 01:05:04] yegoshin.uas : its on the goggles themselves in settings in firmware .800 and above I believe

[2017-11-10 01:07:16] jezzab : I cant see why mixed wouldnt work.

[2017-11-10 01:09:57] jezzab : Give it a whirl :slightly_smiling_face:

[2017-12-23 17:52:03] irischnpak : Just recevied my racing edition googles. Anything I can do to help while they are in a virgin state? FW dump, etc?

[2017-12-23 17:57:11] hostile : ping the guys in ~goggles_rooting

[2017-12-23 17:57:18] hostile : maybe ask @kilrah

[2017-12-24 07:50:46] kilrah : I don't think so... did provide a dump of mine already...

[2018-02-19 22:51:03] pawelsky : Given that now in the new firmware we have a video player for user content and a pano photo viewer with headtracker support, I wonder how long it will take before someone does the 360 video player (and then add 3D on top of that) :slightly_smiling_face:

[2018-04-13 12:13:06] nbm2010 : any news here? anybody working on DJI Googles RE ?

[2018-12-26 22:09:16] chipmangini : Any information on getting these working with NLD, or something???

[2018-12-27 15:43:13] kilrah : ??

[2018-12-27 21:32:22] chipmangini : Don't the Google Goggles run a version of DJI Go? I've never even seen a set, but an interested....

[2018-12-27 23:13:39] pawelsky : No, not even close

[2019-04-13 11:18:59] dronepilot : Dji googles RE range issue on Mavic 2, any one has a clue for a quik fix? Thanks

[2019-04-24 01:50:46] dronepilot : Anyone nows if this new M2 update firmware fixed the range issue on the dji goggles?

[2019-04-24 05:59:56] kilrah : Well it's one data point, but I flew once with the goggles on 400 and it's the first time I had no problem. So a solid **maybe**.

[2019-04-24 08:29:14] evgeniychuiko : And I have opposite experience with white goggles. For me issue is still there

[2019-04-24 08:30:20] evgeniychuiko : I guess they will announce it in revision info, if they solve that issue

[2019-04-24 09:54:41] dronepilot : @evgeniychuiko did you try the latest fw .400 on the M2? Still get the same range issue? Thanks

[2019-04-24 10:21:41] evgeniychuiko : We talking about 0400, aren’t?

[2019-04-24 13:20:39] dronepilot : Yes

[2019-04-28 00:50:15] digital1 : I have read on DJI Forum if your still having issues to send the Google’s in for replacement. It’s said DJI made a hardware change in Feb this year that fixes the problems but this is on the RE Goggles. Dji have been telling people to send them in now. Not sure how it can have a hardware issue when it was fine with Mavic Pro. anyone know anything on this ? What the hw problem may be ?

[2019-04-28 10:13:56] kilrah : link?

[2019-04-28 11:59:23] digital1 : ‪DJI acknowledges Goggles RE design flaw - FROM DJI FORUM <http://forum.dji.com/forum.php?mod=viewthread&tid=184488>

[2019-04-28 12:00:54] digital1 : Actually reading again and it’s not fixed it seems.

[2019-08-07 10:31:05] dronepilot : Hi is there anyway to fcc the goggles re?

[2019-08-08 13:39:29] evgeniychuiko : The googles are slave device. It will follow drone settings

[2019-08-08 15:59:08] evgeniychuiko : Are you sure it is able to show fcc? Any screenshots from US?

[2019-08-09 06:40:26] evgeniychuiko : As I said, googles are slave device, and have same destination code as remote controller. If your drone (master) has fcc settings for SDR, all slave devices will follow it

[2019-09-01 00:28:43] dronepilot : Anyone figured out how to get better range with dji goggles re with the mavic 2 standard rc?

[2020-07-08 05:57:52] almog.baku : hey, since the history is cleared… what’s the status today? any blogpost?

[2021-10-30 12:19:46] jj : madmaqx joined the channel.

[2021-10-31 00:21:53] will : will joined the channel.

[2021-11-02 07:05:29] newlc : newlc joined the channel.

[2021-11-09 07:38:34] cannon : cannonf0815 joined the channel.

[2021-11-12 10:40:44] cs2000 : cs2000 joined the channel.

[2021-11-12 10:40:51] cs2000 : dji-rev-bot added to the channel by cs2000.

[2021-11-12 10:40:53] cs2000 : @cs2000 left the channel.

[2021-12-30 12:51:12] reaktor : Hi all. Just putting it out there that I have the DJI FPV (on my 3rd actually) and the DJI v2 Goggles that come with the bundle. If anyone needs to test a mod I would give it a go. As long as the risk of bricking isnt too high.

[2021-12-30 12:57:12] reaktor : The way I see it the Goggles are the centre of everything. All software updates and activations for the quad & remote, etc occur through the Goggles. If you can root them you probably have a way into the system. The last firmware update locked things down somewhat but the FCC mod still works, as long as you dont plug the Goggles into the Fly app. I'm using the Beyond app on my Android phone which still works a treat on the latest firmware.

[2021-12-30 13:03:38] reaktor : Some people dont like the FPV but screaming along at 140km/h is nuts! I also have the Motion Controller. It is so intuitive anyone can fly like a maniac out of the box. A bit dangerous really :skull_and_crossbones:

[2021-12-30 13:13:21] reaktor : On the negative side, the fact that DJI needs you to sideload your app gives them too much control, especially in Android. I was grounded when my latest replacement FPV came. I had to connect the Goggles to the Fly app to activate the Quad. It then grounded all my hardware until I updates everything to latest firmware. This I dont like. It gives DJI the ability to force users to upload their bugfixes and fix any exploits that may have been found. The Android apps are no longer on the Google Play store so God knows what backdoors, etc we are installing.

[2021-12-30 13:13:47] reaktor : ./end rant

[2022-01-13 15:10:31] idzik : idzik joined the channel.

[2022-01-14 13:14:41] leo : pilipala008 joined the channel.

[2022-01-18 00:49:26] il1oo0 : il1oo0 joined the channel.

[2022-01-31 08:47:27] jackmax : jackson joined the channel.

[2022-03-20 20:24:32] hrisfpv : hrisfpv joined the channel.

[2022-03-25 23:34:59] joonas : joonas joined the channel.

[2022-04-01 10:59:51] kon : kon joined the channel.

[2022-04-06 17:08:37] ttdqj4wj3hgcy5zg85qi : ttdqj4wj3hgcy5zg85qi joined the channel.

[2022-04-18 09:25:18] enigma2 : enigma2 joined the channel.

[2022-04-18 09:25:49] enigma2 : @enigma2 left the channel.

[2022-05-23 20:49:25] jjbyrnes29 : jjbyrnes29 joined the channel.

[2022-05-26 13:53:07] item1979 : item1979 joined the channel.

[2022-06-04 19:58:01] mavic2reverser : mavic2reverser joined the channel.

[2022-06-06 19:15:56] socializednerd : socializednerd joined the channel.

[2022-09-15 20:54:35] d3vl_jack : d3vl_jack joined the channel.

[2022-09-26 01:23:06] jackmax : jackmax joined the channel.

[2022-10-16 11:05:45] jack117wb : jack117wb joined the channel.

[2022-10-16 13:40:46] cr : datum joined the channel.

[2022-10-24 07:49:10] c1dbr : c1dbr joined the channel.

[2022-11-12 06:21:41] pingspike : pingspike joined the channel.

[2023-03-02 00:10:56] ibndias : ibndias joined the channel.

[2023-03-24 08:20:48] retrocall : retrocall joined the channel.

[2023-04-08 06:09:09] gudvin : gudvin joined the channel.

[2023-04-08 06:09:33] gudvin : @gudvin left the channel.

[2023-05-29 01:39:28] revnmavic : revnmavic joined the channel.

[2023-06-07 12:33:24] revnmavic : @revnmavic left the channel.

[2023-06-11 01:14:00] milenovic : milenovic joined the channel.

[2023-07-21 23:20:58] tmbinc : tmbinc joined the channel.

[2023-08-01 13:50:00] crashing_bird : crashing_bird joined the channel.

[2023-09-13 10:14:38] molda : molda joined the channel.

[2023-10-15 10:22:02] sparkyws : sparkyws joined the channel.

[2023-10-15 20:50:35] dumldore_newbi : dumldore_newbi joined the channel.

[2023-10-23 09:15:26] jdan7387 : jdan7387 joined the channel.

[2023-10-24 17:34:50] caseygibson : caseygibson joined the channel.

[2023-12-08 04:10:30] seraph1573 : seraph1573 joined the channel.

[2024-01-15 14:54:00] shinoby : shinoby joined the channel.

[2024-01-16 14:36:44] zjm605186980 : zjm605186980 joined the channel.

[2024-01-18 15:43:40] basilius : basilius joined the channel.

[2024-02-06 14:06:08] xpk : xpk joined the channel.

[2024-02-06 19:58:54] ryantkasher : ryantkasher joined the channel.

[2024-02-10 03:22:48] accountfrompl : accountfrompl joined the channel.

[2024-02-12 20:45:19] lining-preps.0u : lining-preps.0u joined the channel.

[2024-03-03 23:16:58] gizmo : gizmo joined the channel.

[2024-03-24 22:29:45] deadman96385 : deadman96385 joined the channel.

[2024-04-11 13:10:14] halowarrior8 : halowarrior8 joined the channel.

[2024-05-10 08:20:41] dreamtree : dreamtree joined the channel.

[2024-05-21 07:01:01] wag-on : wag-on joined the channel.

[2024-07-01 17:29:41] mrsmith : mrsmith joined the channel.

[2024-07-17 03:02:10] propz : propz joined the channel.

[2024-07-26 15:53:07] ogini_ayotanom : ogini_ayotanom joined the channel.

[2024-08-16 13:55:26] yoyo : yoyo joined the channel.

[2024-08-16 13:59:04] yoyo : @yoyo left the channel.

[2024-11-26 16:58:58] ox3d : ox3d joined the channel.

DJI-Rev Mattermost Archive

Messages in goggles_rooting

DJI-Rev Mattermost Archive

Messages in goggles_rooting

User Info