suser

So VY 73 and Grazie, de LZ4TU  taaa ti ti taaa
*Excuse me for the offtopic but could not resist, for those not knowing Morse code and are not HAMs each CW QSO finish with TU, on morse: _  .._ meaning Thank You!
Good team here: @jock, @ilmich, @fabiobassa and so much great contributors, ofcourse the leading @Igor , Thanks again, wish you all good luck!

i perfectly understand, I also hate when an available piece of hardware cannot be used, but really lost a lot of time in wifi drivers and, especially realtek drivers, need constant care because they are overbloated software written by simians.
When complied, any realtek driver is few megabytes large, while the average wifi driver is well below the 100kbytes.
 
The device id you see in rk322x-config 024c:8723 is clearly declares that your chip is rtl8723as (for some other references: https://forum.libreelec.tv/thread/27091-no-wifi-at-mxq-pro-4k-5g/)
 

@suser
your avatar is a smith cart about impedence, I suppose.. Italian HAM here

This make me think that you have the right approach to problem : WHY IS NOT WORKING !!!
And not the simple " ok it will work in othe manner" but " what hell is going now"
 
This is the spirit that animated @jock @ilmich and myself when we approached to rk322x world

Unfortunately lack of documentation = trial and errors and understand code and compilation !! And again unsuccess and again trial and errors. Even actually the evil nand driver is ONLY proprietary, even if jock and ilmich actively searching for months a workaround

Welcome in the club 😅
 

I have few USB dongles working, bough for 1.5EU each, it was not a matter of money, nor WiFi desperately needed. It's was a matter of "make things working" or at least "to know why it is not working". Thanks everybody here, you are such a talents !

@suser 8723 are pain in the a**
On other boards we have had even 8723cs but ONLY working with 8723bs driver.
And many other strange combinations 
My suggestion :
IF wifi strictly needed buy a few dollars external USB wifi dongle.
ask here which are the best supported

Hi all,
 
I've a box with rk3228b, nand, ddr2 etc etc and I'm using it as mediacenter with Libreelec 9.2 installed on sdcard.
The box recognize the nand, but I had never been able to install the operating system in it (tried with rkdeveloptool, dd and others but without succes).
Until today, when I've discovered this beautiful tool called multitool
I had to make a few small changes (basically removed offset and gpt creation in the nand flash code), but it works great.
So my question is, it's possible to add an option to write a generic image, in order to make this tool more useful!?
 
Last question, I think that the libreelec bootloader has some trouble when boot from NAND, but it's strange, because I've readed that the multitool one shoud be the same. It's true, or there are some changes that I can apply and retry?!
 
 
Thanks for your hard work.
Michele
 
 
 
 
 

Hello!
The problem is the missing driver for the rtl8723as. Actually, there is a driver floating around, but adapting and keeping it working on every kernel release require a lot of effort. I don't even have a board with such chip, so it is indeed very difficult, but it may become unworthy of the time spent because realtek drivers are very messy.
 
You could find the source code and compile it yourself, or perhaps desoldering the wifi module and change the chip with another one is easier!
It would be even easier to attach an USB wifi device.
 
BTW, in the forum thread there is a success story with rlt8723as driver, you may want to search with forum tools or google.

First I want to say BIG thanks to @jock for his hard work! I am runing Armbian_23.11.1_Rk3318-box_jammy_current_6.1.63_xfce_desktop.img on a T9 box with RK3318, even The Bluetooth is working! I want to say thanks to @paradigman for his post here https://forum.armbian.com/topic/26978-csc-armbian-for-rk3318rk3328-tv-box-boards/?do=findComment&comment=135469 , finally i succeeded to see something on my 7segment display and even if it is upside down i am very happy!
What i did is extacted kernel from Android, then extract DTB from it, than decompile to DTS and find that:
Than i correct the DTS from @paradigman like this:
Than installed the kernel headers from armbian-config and follow the steps from @paradigman
 
Now the question is: can i rotate the display and can i lit the WiFi and Network indicators and how? And also i will love the "Play" symbol to blink with heartbeat after booting by adding to "/etc/rc.local"  something like:
echo heartbeat > /sys/class/leds/working/trigger  

MyT9_dtc_out.dts

nmtui-edit Select interface, go to IPV6 and switch to "ignore". Perhaps you will need to restart.

@suser try to set another "display type" values in openvfd config: https://github.com/arthur-liberman/vfd-configurations

About OpenVFDService, had to study and understood that led cannot be triggerd without an userspace program with custom made functions.
https://github.com/arthur-liberman/linux_openvfd/blob/master/OpenVFDService.c
 
So decided to mod directly the OpenVFDService to include POWER, LAN and WIFI led.
My setup is:
POWER always on when the box is running LAN on when lan cable connected  
The OpenVFDService source is attached, and hereunder a short explanation of the mod. 
// display is controlled by just 7 words 0 -> lan, power, colon, wifi leds on/off 1 -> hour +1 2 -> hour 3 -> minute +1 4 -> minute 5 -> unused = 0 6 -> unused = 0 7 -> unused = 0 memset(wb,0x00, sz); // set all to 0 // byte 0 mapping wb[0] |= 0b01100000; // wifi wb[0] |= 0b00000100; // power wb[0] |= 0b00010000; // colon wb[0] |= 0b00001000; // lan // mod for time and colon wb[0] |= 0b00010000 & (data.colon_on << 4); // colon blink original function wb[1] = char_to_mask(data.time_date.hours/10); // hour+1 wb[2] = char_to_mask(data.time_date.hours%10); // hour wb[3] = char_to_mask(data.time_date.minutes/10); // minutes+1 wb[4] = char_to_mask(data.time_date.minutes%10); // minutes //write to display ret = write(openvfd_fd, wb, sz);  
LAN connected was pretty straightforward:
// check ethernet char buf_eth[1]; int fd = open("/sys/class/net/eth0/carrier", O_RDONLY); if (fd < 0) { perror("Open /sys/class/net/eth0/carrier failed.\n"); return; } else{ // read from fd read(fd, buf_eth, 1); close(fd); } // set display LAN led if ( buf_eth[0]=='1' ){ wb[0] |= 0b00001000; // lan }  
About Wifi led, my box do not use wifi, so decided to link this function the the IR remote receiver. Led on when a IR signal is received.
// create thread for LIRC dev listening in main ret = pthread_create(&lirc_id, NULL, lirc_thread_handler, &setup); //New thread - listen ot lirc device for IR signal void *lirc_thread_handler(void *arg) { lirc_loop(); pthread_exit(NULL); } // IR read loop and set flag void lirc_loop(){ char buf[8]; //printf("Start lirc wait signal:\n"); while(sync_data.isActive) { //printf("DEBUG new thread created!\n"); if ( !gotIRchar ){ // check lirc dev int fd = open("/dev/lirc0", O_RDONLY); if (fd < 0) { perror("Open /dev/lirc0 failed.\n"); return; } // read from fd read(fd, buf, sizeof(buf)); //printf("DEBUG read complete %i\r",sizeof(buf)); close(fd); if (buf[0] > 0){ //printf("Ir receive: %s\n",buf); gotIRchar = true; // set IR received FLAG } } } } // check flag and set display WIFI led if ( gotIRchar ){ wb[0] |= 0b01100000; // wifi gotIRchar = false; }  
OpenVFDService.c

It would be nice to have a proper kernel driver for such device. As far as I remember, it uses a i2c-like bus with some minor differences; I looked into some time ago, but had to give up due to not enough time, but a proper driver would wire the led chip hardware with the kernel led framework to exploit all the kernel triggers and goodies.

Updated to kernel 6.6.2-edge-rockchip64
 
Need to recompile openvfd driver, had to modify openvfd_drv.c.
Patch attached.
patch linux_openvfd/driver/openvfd_drv.c < openvfd_drv.c.patch  
openvfd_drv.c.patch

No way to have kodi acceleration working yet.
 
In the meanwhile I have been studying the front LCD, which was black [SOLVED]
 
1. Install OpenVFD
 
#install kernel headers (skip if already installed) sudo apt-get install linux-headers-`uname -r` # download the sources: git clone https://github.com/arthur-liberman/linux_openvfd # change the directory cd linux_openvfd/driver # edit the Makefile # change the line KERNELDIR = ../../../ # to this: KERNELDIR = /lib/modules/$(shell uname -r)/build # create a symlink to correct System.map in this KERNELDIR - in my case: # /lib/modules/6.1.7-rockchip64/build/System.map -> /boot/System.map-6.1.7-rockchip64 # compile the driver (in openvfd/driver): make -j4 # install module sudo make modules_install # load module modprobe openvfd  
2. Install OpenVFDService
cd linux_openvfd make -j4 OpenVFDService sudo cp OpenVFDService /usr/sbin/  
3. Update dts using armbian-config 
add at the end:
openvfd { compatible = "open,vfd"; dev_name = "openvfd"; status = "okay"; }; reboot and you will find new device-tree in /proc/device-tree/open-vfd
 
4. Create openvfd.service in /lib/systemd/system/openvfd.service 
 or download attached file
 
[Unit] Description=OpenVFD Service ConditionPathExists=/proc/device-tree/openvfd/ [Service] ExecStart=/bin/sh -c '[ `cat /proc/device-tree/openvfd/compatible` = "open,vfd" ] && /sbin/modprobe openvfd; /usr/sbin/OpenVFDService' ExecStop=/bin/kill -TERM $MAINPID ExecStopPost=-/usr/sbin/rmmod openvfd RemainAfterExit=yes [Install] WantedBy=basic.target  
create VFD config in /etc/modprobe.d/openvfd.conf
options openvfd vfd_gpio_clk="4,0x13,0" options openvfd vfd_gpio_dat="4,0x16,0" options openvfd vfd_gpio_stb="4,0x12,0" options openvfd vfd_chars="4,0,1,2,3" options openvfd vfd_dot_bits="0,1,3,2,4,5,6" #ID Display type #0x03 A display like on the A95X R2 (or the Abox A1 Max). <-- #.2 Reserved - must be 0. #.3 flags #.4 0 - > FD628 and compatible controllers options openvfd vfd_display_type="0x03,0x00,0x00,0x00"  
enable service and start (this will autoload the kernel module openvfd
sudo systemctl enable openvfd sudo systemctl start openvfd  
After this the system clock will successfully show up on the front LCD 
 
openvfd.service

RK3318 board (x88 pro 10)
Kernel: 6.6.2-edge-rockchip64
Ubuntu Jammy 22.04.3 LTS
 
Vidoe hardware acceleration working.
 
Clock and leds working using the OpenVFD service
 

 
See details here if interested:
 
 

You're welcome
 
This is actually not possible with the openvfd driver because it is not wired to the kernel led framework but is a totally custom module that just exposes the hardware to userspace.

Of course. I wrote a little script for it that shortens the process a bit. It is important to know that I am not working as a root user within armbian.
The necessary steps: 
   copy openvfd-fd6551.dts to /home/<username>    copy install_kernel_debs.sh to /home/<username>    copy OpenVFDService file to /usr/sbin,  AND set permission to 777 !    copy openvfd.service file to  /lib/systemd/system        sudo armbian-add-overlay openvfd-fd6551.dts    sudo /bin/bash install_kernel_debs.sh 
After reboot:    cd linux_openvfd/driver    edit MAKE -> KERNELDIR = /lib/modules/$(shell uname -r)/build    sudo make -j4    sudo make modules_install    sudo modprobe openvfd     sudo usermod -a -G i2c <username> Start the Service:
sudo systemctl enable openvfd sudo systemctl start openvfd

   
openvfd-fd6551.dts openvfd.service OpenVFDService install_kernel_debs.sh

@Jock
Here Are Some DTB Dumps From 2022 to 2023 (Android 9/10/11/12)
The File Names Are "CRC32".dtb
DTB Dumps (2022~2023) (CRC32.DTB)

Now RPI strongly loses in all respects to s9xxx and rk33xx TV boxes. The only place where the use of RPI is justified, systems that use a variety of external sensors and actuators, of which a large number are produced and there are ready-made systems for their use.
 
 
 
 LE did not use patches, but a ready-made closed binary module (driver), which was taken from the finished Android firmware. This binary module was built exclusively for one version of the 3.14 kernel and cannot work with any other versions of the kernel. The manufacturer of the chip does not give anyone the source code from this driver, so no one will spend their time disassembling the binary code and studying it to write a new driver for the kernel 4\5. If you want, you can do the job yourself.

I would propose to build most recent ./compile EXPERT="yes" # to see EOS targets - that should give you latest latest, but since its unmaintained ... I have no idea if it works. Probably it does.
 

This box came with at least two different onboard wifi chips, which means you have to play with a device tree blob configuration to make it work. Drivers are most likely present in the kernel for all variants.
 

If you don't have support for DVFS, CPU runs at IIRC 50 or 60% of full speed ... which is why temps will stay low.
 

It depends who's asking  TV boxes are without schematics and without certainty that one model is identical to the same model, which means this job is more like a hacking ... 

Partial solution to the problem!
I've upgraded the stock firmware+bootloader to another more recent (Android 9 with kernel 3.14 oficial) with Amlogic USB Flash tool, checking the option "Erase bootloader". After this, I've booted completely Android.
 
Next , I've booted Armbian, installed the WIFI, change the keyboard layout, timezone and hostname. And installed on eMMC.
 
Now, with Armbian 20.10 Focal 5.9 installed on eMMC, I've restarted hundreds of times and the "black-screen-non-booting" problem has disappeared.
 
So, if you don't have more official firmwares for your TV Box, try any unofficial ROM like ATV Experience. It seems that changing the bootloader and binary blobs can solve it.
 
My issue now is that sometimes on reboot, the screen doesn't get HDMI signal from TV Box and it doesn't boot. It boots OK if I unplug and plug again the power plug.
 
Anyone knows about this strange HDMI restarting issue?
 

@Werner
 
your point about not struggling through support issues is a good one and I will take a look at the units at your link. 
 
Thanks
 
P.S. The delayed response is because I'm new and am only allowed to make one post a day