testers wanted Next major upgrade v5.60

[Igor]

On Which image can I test? Any

 

How? Go to armbian-config -> System -> Nightly and proceed. If this option is not there, defreeze upgrading

 

Restart and after your system is up, proceed to one of those areas:

 

- desktop install from armbian-config on top of CLI (any board)
- set first_run_network configuration, setting ip within /boot/armbian_first_run.txt (any board except Espressobin)
- kernel switching or upgrade (any board, especially Odroid XU4 NEXT) 
- setting fixed/dhcp network with armbian-config (any board except Espressobin)
- configure frame buffer within h3disp utility (any h3 board with a legacy kernel)
- DTS sound on DEV kernel (miqi, Tinkerboard)
- ROCK 64 default kernel
- check effects of changing everything to Network Manager withing firstrun/armhwinfo and other scripty
- check install from a blank image (https://dl.armbian.com/odroidc2/nightly/ and https://dl.armbian.com/orangepiplus2e/nightly/)
- set AP mode (any board)

All changes are reflected in a beta repository and are coming from the build script, branch "development". If you want to build this state, use LIB_TAG="development" while running the script. Make sure you are doing this on sine test install.

 

In case you find a bug:

- report it here

- push a fix to the build script, branch development

[TonyMac32]

Sound was added to Next for Tinker, and Dev for MiQi.  I will try to bring them to equality today.  

[jeanrhum]

hi,

 

Writing from: Linux orangepiplus2e 4.14.27-sunxi #105 SMP Thu Mar 15 22:24:12 CET 2018 armv7l armv7l armv7l GNU/Linux

I'm testing xubuntu server nightly image. I successfully installed desktop and I'm writing this post inside chromium. Navigation on several simple websites is smooth and I was able to play a video from youtube.

I configure wifi using armbian-config (default dhcp), then I defined a fixed IP.

I was able to play a video with mpv, but it only plays smoothly inside a small window (default size). Sound is working but I had to launch alsamixer to increase the volume (~40% by default) because I was not able to change this in pulse audio gui.

Cpu-freq adjust cpu frequency between 480MHz and 1.3GHz as expected

 

More details: http://ix.io/YdQ

 

Major features seems to work, great job!

 

I will try to take time to test another image on my opi one upgrade to nightly and/or on a pine64.

[ebin-dev]

I am testing Armbian 5.42 on an EspressoBin 1G (boots from SPI/sata):

 

It boots just fine:

Welcome to ARMBIAN 5.42.180317 nightly Debian GNU/Linux 9 (stretch) 4.14.27-mvebu64 

Linux version 4.14.27-mvebu64 (root@xeon) (gcc version 7.2.1 20171011 (Linaro GCC 7.2-2017.11)) #4 SMP PREEMPT Fri Mar 16 20:18:10 CET 2018

 

However:

After switching from stable branch (kernel 4.14.24) to nightly (kernel 4.14.27) I observe two kernel installations in /boot: 4.14.27 and 4.4.121 (however, only 4.14.27 is configured - I just deleted *4.4.* also from /lib/modules etc.). (This may be specific to my installation)

 

The only problem I observe is that with a mainline kernel on SSD (sata) reboot stops after shutdown is completed (with a manual reset it boots again):

 

[  OK  ] Reached target Shutdown.
[ 234.934734] reboot: Power down
ERROR:   a3700_systPANIC at PC : 0x0000000004023248

 

This is also reported in the forum to happen with a mainline kernel on SD.

 

[Christos]

@Igor

Do we need to do git clone with -b "development" from scratch or is it ok/suffice to use LIB_TAG="development" with the current compile.sh  from master ?

 

 

/PS: Tried to git clone -b development but every time compile.sh is started, it always checks out to master (Switched to branch 'master'), how can we compile from the development build branch?

christos@Ubuntu16042VM:~/armbian/build$ git checkout development
Already on 'development'
Your branch is up-to-date with 'origin/development'.
christos@Ubuntu16042VM:~/armbian/build$ ./compile.sh LIB_TAG=development
[ o.k. ] Using config file [ config-default.conf ]
[ warn ] This script requires root privileges, trying to use sudo 
[ o.k. ] Using config file [ config-default.conf ]
[ o.k. ] This script will try to update
Already up-to-date.
Switched to branch 'master'
Your branch is up-to-date with 'origin/master'.

 

[Igor]

23 hours ago, jeanrhum said:

 

I will try to take time to test another image on my opi one upgrade to nightly and/or on a pine64.

Thanks! Upgrades from some random old builds would be nice.

 

19 hours ago, ebin-dev said:

After switching from stable branch (kernel 4.14.24) to nightly (kernel 4.14.27) I observe two kernel installations

You must have them installed manually since from menu, it uninstalls the previous version and installs new one and while switching to beta, a normal package procedure goes on. But let's keep an eye if perhaps those post-install scripts are not in a perfect condition.

 

19 hours ago, ebin-dev said:

The only problem I observe is that with a mainline kernel on SSD (sata) reboot stops after shutdown is completed (with a manual reset it boots again):

For this, we can only write down as a known problem. I just did.

 

11 hours ago, Christos said:

it ok/suffice to use LIB_TAG="development"

It should. I did it this way.

[jeanrhum]

I successfully upgrade my opi one (debian 9 5.38 next) to nightly using armbian-config.

I have 2 pine64 boards, but I suspect one to have hardware issues. With this one, I did not achieve the upgrade when starting from latest 5.38 (debian or xenial server 3.4 kernel), the screen stay blue after some downloads and I can't do anything even after more than 10 minutes. With the other one (running a 5.31 ubuntu desktop I think) I update&upgrade to 5.38, after that when I want to switch to nightly, it downloads package and propose to reboot, but after reboot I'm still on kernel 3.14. Armbian-config title says 5.42 but kernel stay at 3.14 even if I try to swith to dev or next.

[Christos]

@Igor

Can you please have a look at this compile.sh output?

Trying to do a clean build based on the new development branch.

It is a fresh new git clone and use your "LIB_TAG=development" argument, yet upon start it reports that is still using 'master'..  is that ok?

Select then to build a OPiPC dev (latest kernel) xenial server image.

It also later on, reports that it will generate '5.41' deb and img whereas I guess it should say '5.42' as revision, is that ok?

And all generated debs are 5.41 revision.

christos@Ubuntu16042VM:~/armbian/build$ git checkout master
Already on 'master'
Your branch is up-to-date with 'origin/master'.
christos@Ubuntu16042VM:~/armbian/build$ ./compile.sh LIB_TAG=development
[ o.k. ] Using config file [ config-default.conf ]
[ warn ] This script requires root privileges, trying to use sudo 
[ o.k. ] Using config file [ config-default.conf ]
[ o.k. ] This script will try to update
Already up-to-date.
Already on 'master'
Your branch is up-to-date with 'origin/master'.
[ o.k. ] Command line: setting LIB_TAG to [ development ]
[ o.k. ] Preparing [ host ]
[ o.k. ] Build host OS release [ xenial ]
[ o.k. ] Syncing clock [ host ]

..

[ o.k. ] Building deb [ linux-u-boot-dev-orangepipc_5.41_armhf.deb ]

..

[ o.k. ] Done building [ /home/christos/armbian/build/output/images/Armbian_5.41_Orangepipc_Ubuntu_xenial_dev_4.16.0-rc5.img ]



christos@Ubuntu16042VM:~/armbian/build/output/debs$ ls -l
total 76632
-rw-r--r-- 1 root root    30752 Μάρ  18 15:19 armbian-config_5.41_all.deb
-rw-r--r-- 1 root root  2115308 Μάρ  18 15:19 armbian-firmware_5.41_all.deb
-rw-r--r-- 1 root root 76109880 Μάρ  18 15:22 armbian-firmware-full_5.41_all.deb
-rw-r--r-- 1 root root    11588 Μάρ  18 15:22 armbian-tools-xenial_5.41_armhf.deb
drwxrwsr-x 2 root sudo     4096 Μάρ  18 00:43 extra
-rw-r--r-- 1 root root   187638 Μάρ  18 15:38 linux-u-boot-dev-orangepipc_5.41_armhf.deb
drwxrwsr-x 2 root sudo     4096 Μάρ  18 15:38 xenial
christos@Ubuntu16042VM:~/armbian/build/output/debs$ 

 

Thus it looks the compile.sh even with the "LIB_TAG=development" eventually used the master branch and not the development.

 

 

Then I tried to build by first selecting manually the development branch and use your "LIB_TAG=development" argument, yet upon start it reports that is still using 'master'  !

Select then to build a OPiPC dev (latest kernel) xenial server image.

And again later on, reports that it will generate '5.41' deb and img whereas I guess it should say '5.42' as revision.

christos@Ubuntu16042VM:~/armbian/build$ git checkout development
Switched to branch 'development'
Your branch is up-to-date with 'origin/development'.
christos@Ubuntu16042VM:~/armbian/build$ ./compile.sh LIB_TAG=development
[ o.k. ] Using config file [ config-default.conf ]
[ warn ] This script requires root privileges, trying to use sudo 
[ o.k. ] Using config file [ config-default.conf ]
[ o.k. ] This script will try to update
Already up-to-date.
Switched to branch 'master'
Your branch is up-to-date with 'origin/master'.
[ o.k. ] Command line: setting LIB_TAG to [ development ]
[ o.k. ] Preparing [ host ]
[ o.k. ] Build host OS release [ xenial ]
[ o.k. ] Syncing clock [ host ]

..

[ o.k. ] Building deb [ linux-u-boot-dev-orangepipc_5.41_armhf.deb ]

..

 

 

So, tried both ways, either from master branch with "LIB_TAG=development" argument or from manually switching to development executing compile.sh with "LIB_TAG=development" argument and in both tries the compile.sh script reverted to master.

Am I making something wrong here?

 

 

[Igor]

56 minutes ago, Christos said:

Am I making something wrong here?

Don't alter it manually and it should work.

1. Git clone as usually

2. I added LIB_TAG to default config

... and it worked for me.

[Christos]

@Igor

 

I tried without manually specifiyng branch at first as I show above but it did not work with the shown log.

 

Where and how exactly should I use the LIB_TAG ?  not as an argument in compile.sh ?

[Igor]

Just now, Christos said:

Where and how exactly should I use the LIB_TAG ?  not as an argument in compile.sh ?

I added it to config-default.conf ... and it works for me. Try that.

[Christos]

Thanks @Igor  , that did it and now it can compile from the development branch.

Just one more thing though, can it use the 'dev' kernel? meaning the latest 4.16 and not the 4.14?

I see as options only the legacy and next, not dev.

[Igor]

Just now, Christos said:

I see as options only the legacy and next, not dev.

You need to be an expert to see this  Add EXPERT="yes" to the configuration.

[TonyMac32]

It would appear there's a regression in 4.14 concerning sound on rk3288 SoC drivers.  This has been seen on Tinkerboard Next and some reports came in on @Myy's github about the same thing.  I'll look into the fix, but that's been the holdup. 

[JMCC]

Would it be useful if we tested with a Debian Stretch image, or it is better to leave that for later and stick with Xenial for now?

[JMCC]

Testing with Xenial on Tinkerboard default kernel.

Had a wifi connection configured through Network-Manager GUI, and tried to disable it and creat a hotspot with armbian-config

 

1. First, I tried by deactivating the current wifi connection (Network -> Wifi -> Select connection and <Deactivate>).  Then, I went back to the menu and selected "Hotspot", the script threw a lot of dialog errors. The config files created also had those errors. This is how the head of the files looked like:

   /etc/NetworkManager/conf.d/10-ignore-interfaces.conf:
   
   [keyfile]
   unmanaged-devices=interface-name:
   Error: Expected at least 5 tokens for --menu, have 4.
   Use --help to list options.
   
   
   /etc/dnsmasq.conf:
   
   interface=
   Error: Expected at least 5 tokens for --menu, have 4.
   Use --help to list options.                             # Use interface
   Error: Expected at least 5 tokens for --menu, have 4.
   Use --help to list options.

   As you see, there are dialog error messages in the place where "wlan0" should be.

2. After fixing the mess, I tried again, but instead of disabling the connection in the "Wifi" module, I went to "Network -> Forget" at the bottom of the menu. Everything worked fine this time.

Proposed fix: What I would do, is just have the "Hotspot" submodule wipe all connections by invoking the "Forget" submodule  at the beginning. You can put a warning dialog (like "This module will clear all your saved WiFi connections and create a Hotspot. Proceeed? <OK> <Cancel>"). I think nobody will complain if you do that, and it will save you time and sweat.

[JMCC]

• Switching to nightly in a non-updated system (armbian OdroidXU4 5.37 without any upgrades) causes some packages to have missing dependencies and remain unconfigured, until you do manually apt -f install
• Installing desktop on top of CLI works fine in XU4.  Only that, after installation is finished, the user is presented with the desktop while the script is still running in another virtual terminal. No big deal, but can be fixed by:
  • Terminating the script after installation
  • Getting the virtual terminal id at the beginning with "SCRIPTVT=$(fgconsole)", and then returning there at the end with "chvt $SCRIPTVT"
• OpenGL-ES and Webgl also work, I'll post a script in other thread. 
• Wi-Fi hotspot with Odroid module 3 (rtl8192cu) works out of the box.

 

[Igor]

8 hours ago, JMCC said:

Switching to nightly in a non-updated system (armbian OdroidXU4 5.37 without any upgrades) causes some packages to have missing dependencies and remain unconfigured, until you do manually apt -f install

You mean image from the download section? Which one exactly?
 

8 hours ago, JMCC said:

OpenGL-ES and Webgl also work, I'll post a script in other thread. 

Ideally, we should pack those things in the build process. 

Thank you. I'll try to cope with those problems ASAP.

[JMCC]

5 hours ago, Igor said:

Which one exactly?

It was this one: https://dl.armbian.com/odroidxu4/Ubuntu_xenial_next.7z . But I haven't been able to reproduce the error today, so maybe it was a network or SD card failure.

 

5 hours ago, Igor said:

Ideally, we should pack those things in the build process. 

It shouldn't be too difficult to incorporate the script to the build process. For XU4, though, there is a small but annoying bug that we should fix or work around before. It is described here:

Besides that, the script should not be too hard to implement, and there are not too many extra packages to be included.

 

Maybe we could also consider to add something similar for RK3288 images. In that case, we don't even have any important bug to worry about.

[TonyMac32]

8 minutes ago, JMCC said:

Maybe we could also consider to add something similar for RK3288 images.

That would be great, I'm afraid my script-fu is not up to it.  :-/. This is very much a learning experience for me...

[jeanrhum]

Hi,

I succesfully boot last image for pine64: Linux pine64 4.14.29-sunxi64 #57 SMP Fri Mar 23 12:09:14 CET 2018 aarch64 GNU/Linux

I use debian stretch server image and after freezing kernel upgrade, I install desktop from armbian-config: http://ix.io/12BY

Desktop does not feel so smooth than on my opi+2e.

 

Before that, I upgraded without freezing kernel upgrades and than I was not able to boot again. I will try again and will look at uart, but I had no u-boot screen. After defreeze, these 4 packages can be upgraded:

  linux-dtb-next-sunxi64 linux-image-next-sunxi64
  linux-stretch-root-next-pine64 linux-u-boot-pine64-next
 

Edit: Here is the UART info:

 

U-Boot SPL 2017.11-armbian (Mar 24 2018 - 11:04:27)
DRAM: 0 MiB
### ERROR ### Please RESET the board ###

 

[suchende]

Usually I compile the kernel on my own because I need apparmor. The current 4.14.29 fails with this log. I tried it again with the Debian stretch next nightly. However this leads to the same:

Spoiler

œ

U-Boot 2017.05-armbian (Mar 24 2018 - 10:51:25 +0100) for ODROID-XU4

CPU:   Exynos5422 @ 800 MHz
Model: Odroid XU4 based on EXYNOS5422
Board: Odroid XU4 based on EXYNOS5422
Type:  xu4
DRAM:  2 GiB
MMC:   EXYNOS DWMMC: 0, EXYNOS DWMMC: 1
MMC Device 0 ( SD ): 14.6 GiB
Card did not respond to voltage select!
mmc_init: -95, time 11
*** Warning - bad CRC, using default environment

In:    serial
Out:   serial
Err:   serial
Net:   No ethernet found.
Press quickly 'Enter' twice to stop autoboot:  1  0 
** Unrecognized filesystem type **
** File not found /boot.ini **
12404 bytes read in 26 ms (465.8 KiB/s)
cfgload addr = 0x50000000, Loading boot.ini from ext4 0:1 /boot/boot.ini
cfgload: applying boot.ini...
cfgload: setenv initrd_high "0xffffffff"
cfgload: setenv fdt_high "0xffffffff"
cfgload: setenv macaddr "00:1e:06:61:7a:55"
cfgload: setenv rootdev "UUID=82df4dbc-e903-4279-b505-a39341a72d5f"
cfgload: setenv rootfstype "ext4"
cfgload: setenv console "both"
cfgload: setenv verbosity "1"
cfgload: if ext4load mmc 0:1 0x44000000 /boot/armbianEnv.txt || fatload mmc 0:1 0x44000000 armbianEnv.txt || ext4load mmc 0:1 0x44000000 armbianEnv.txt; then env import -t 0x44000000 ${filesize}; fi
** File not found /boot/armbianEnv.txt **
** Unrecognized filesystem type **
** File not found armbianEnv.txt **
cfgload: if test "${console}" = "display" || test "${console}" = "both"; then setenv consoleargs "console=tty1"; fi
cfgload: if test "${console}" = "serial" || test "${console}" = "both"; then setenv consoleargs "${consoleargs} console=ttySAC2,115200n8"; fi
cfgload: setenv bootrootfs "${consoleargs} consoleblank=0 loglevel=${verbosity} panic=10 root=${rootdev} rootfstype=${rootfstype} rootwait rw"
cfgload: setenv vout "hdmi"
cfgload: setenv cecenable "false" # false or true
cfgload: setenv governor "performance"
cfgload: setenv ddr_freq 825
cfgload: setenv HPD "true"
cfgload: setenv hdmi_tx_amp_lvl  "31"
cfgload: setenv hdmi_tx_lvl_ch0      "3"
cfgload: setenv hdmi_tx_lvl_ch1      "3"
cfgload: setenv hdmi_tx_lvl_ch2      "3"
cfgload: setenv hdmi_tx_emp_lvl      "6"
cfgload: setenv hdmi_clk_amp_lvl     "31"
cfgload: setenv hdmi_tx_res      "0"
cfgload: setenv hdmi_phy_control "hdmi_tx_amp_lvl=${hdmi_tx_amp_lvl} hdmi_tx_lvl_ch0=${hdmi_tx_lvl_ch0} hdmi_tx_lvl_ch1=${hdmi_tx_lvl_ch1} hdmi_tx_lvl_ch2=${hdmi_tx_lvl_ch2} hdmi_tx_emp_lvl=${hdmi_tx_emp_lvl} hdmi_clk_amp_lvl=${hdmi_clk_amp_lvl} hdmi_tx_res=${hdmi_tx_res} HPD=${HPD} vout=${vout}"
cfgload: ext4load mmc 0:1 0x40008000 /boot/zImage || fatload mmc 0:1 0x40008000 zImage || ext4load mmc 0:1 0x40008000 zImage
5057824 bytes read in 413 ms (11.7 MiB/s)
cfgload: ext4load mmc 0:1 0x42000000 /boot/uInitrd || fatload mmc 0:1 0x42000000 uInitrd || ext4load mmc 0:1 0x42000000 uInitrd
5743973 bytes read in 462 ms (11.9 MiB/s)
cfgload: if test "${board_name}" = "xu4"; then setenv fdtfile "exynos5422-odroidxu4.dtb"; fi
cfgload: if test "${board_name}" = "xu3"; then setenv fdtfile "exynos5422-odroidxu3.dtb"; fi
cfgload: if test "${board_name}" = "xu3l"; then setenv fdtfile "exynos5422-odroidxu3-lite.dtb"; fi
cfgload: if ext4load mmc 0:1 0x00000000 "/boot/.next" || fatload mmc 0:1 0x00000000 ".next"  || ext4load mmc 0:1 0x00000000 ".next"; then echo "Found mainline kernel configuration"; else setenv fdtfile "exynos5422-odroidxu3.dtb"; fi
0 bytes read in 18 ms (0 Bytes/s)
Found mainline kernel configuration
cfgload: ext4load mmc 0:1 0x44000000 /boot/dtb/${fdtfile} || fatload mmc 0:1 0x44000000 dtb/${fdtfile} || ext4load mmc 0:1 0x44000000 dtb/${fdtfile}
63776 bytes read in 67 ms (928.7 KiB/s)
cfgload: fdt addr 0x44000000
cfgload: if test "${cecenable}" = "false"; then fdt rm /cec@101B0000; fi
libfdt fdt_path_offset() returned FDT_ERR_NOTFOUND
cfgload: setenv bootargs "${bootrootfs} ${videoconfig} smsc95xx.macaddr=${macaddr} governor=${governor} ${hdmi_phy_control} usb-storage.quirks=${usbstoragequirks} ${extraargs}"
cfgload: dmc ${ddr_freq}
cfgload: bootz 0x40008000 0x42000000 0x44000000
Kernel image @ 0x40008000 [ 0x000000 - 0x4d2d20 ]
## Loading init Ramdisk from Legacy Image at 42000000 ...
   Image Name:   uInitrd
   Image Type:   ARM Linux RAMDisk Image (gzip compressed)
   Data Size:    5743909 Bytes = 5.5 MiB
   Load Address: 00000000
   Entry Point:  00000000
   Verifying Checksum ... OK
## Flattened Device Tree blob at 44000000
   Booting using the fdt blob at 0x44000000
   Using Device Tree in place at 44000000, end 4401291f

Starting kernel ...

 

Blue and green LEDs get off and that's it.

 

And a little bit offtopic: How do I fix a broken kernel...

[Igor]

4 minutes ago, suchende said:

And a little bit offtopic: How do I fix a broken kernel...

It is failing for me as well and we have to go back and check what was changed. Upstream, since we didn't do anything that would affect.

[suchende]

So far Kernel 4.14.20 worked for me. (I need however a way to restore it on to the sdcard)

[JMCC]

Upstream Mainline based kernel solves the file corruption on shutdown issue in XU4. But no sound.

[Igor]

35 minutes ago, JMCC said:

But no sound.

Sound was on the previous build but not on this one, with kernel 4.14.y ?

 

BTW. I nulled beta repository, recreating it, desktop install package naming is changed and few other things.

 

[JMCC]

Yes, there was sound in the previous 4.14.y (Saturday), but not now.

------------------------------------------------

 

Also, USB 3.0 ports don't work at all. They didn't work either in last Saturday's 4.14.y build, but I remember they worked the first time I tried the board on Thursday. I am pretty  100% sure that was 4.14.y too (EDIT: it was 4.14 for sure, because I used a wifi dongle that didn't work on 4.9).

------------------------------------------------

 

BTW, I was able to reproduce the failure I had when switching from stable to nightly in a freshly installed system. The problem is that I ran the script while Armbian was doing unattended upgrades in the background,  so when the script tried to install the packages, it couldn't do it. But anyway it went through and asked me to reboot, so it rebooted in the middle of an upgrade, and that's why there were broken packages left.

 

As a fix, I suggest checking for /var/lib/dpkg/lock before trying to install the packages.

[Igor]

My findings. USB3 is working all the time, while the sound is broken ... I guess I need to create a new patch. Attached 4.14.y with working sound from the previous week.

 

https://github.com/hardkernel/linux/issues/337

 

 

9 hours ago, JMCC said:

As a fix, I suggest checking for /var/lib/dpkg/lock before trying to install the packages.

Yes, this is probably needed in few places. Checking later.

linux-image-next-odroidxu4_5.42.180320_armhf.deb

linux-headers-next-odroidxu4_5.42.180320_armhf.deb

linux-dtb-next-odroidxu4_5.42.180320_armhf.deb

[Igor]

13 hours ago, JMCC said:

Also, USB 3.0 ports don't work at all. They didn't work either in last Saturday's 4.14.y build, but I remember they worked the first time I tried the board on Thursday. I am pretty  100% sure that was 4.14.y too (EDIT: it was 4.14 for sure, because I used a wifi dongle that didn't work on 4.9).

1

 

https://github.com/armbian/build/commit/fb390a0b52f9268907b23f2233c7b523a41b5f91

- USB3 works

- sound works

- it boots

Logs: http://ix.io/13l1

[jeanrhum]

I think it would be interesting to list the preferred boards that should be used for testing (maybe in a specific thread). I look at armbian website, documentation and forum, but this kind of information is not defined in a clear and unique place. In the first post of this thread, we can identify a few of them but that's all.

If we want to help in testing and for instance buy a new board for that purpose, it could guide us. It can also give everbody a clue about best supported boards and their probably long term support by armbian.

 

For instance:

- Officially supported boards for testing: orange pi plus 2E (H3), pine64+ (A64), tinker board (rk3288), odroid c2 (s905), odroid xu4 (Exynos5422)...

- WIP boards for testing: espresso bin (Armada 3700), rock64 (rk3328), NanoPi Neo Core2 (h5)...

 

In the previous list, I only chose 1 board for a given soc, but it is just an example and it is probably difficult to define this list with most people agreeing. For instance, nighlty builds are not generated for all boards, so it can be a criterion to select these boards. The WIP boards list can also be fed by the board bring up section when a board is validated for official support and software is mature enough. With this list, if a thread like this one is created before each new release, it think it will focus testing on these preferred boards.