All Activity

I usually check here every couple of days: https://patchwork.kernel.org/project/linux-rockchip/list/ Yes, they will be included in edge kernel and in a year or two when the next LTS kernel hits they'll be included in current as well.

Thank you @Werner, that makes, sense and I'll follow your advice. Can you please tell me where do you check those updates? And should I assume that when those changes are completed in mainline, they'll make it to the Armbian distribution with the regular updates? Or will I need to change to a different one if noone updates this one? Thank you again. Kind regards

Which version or image is that ? From here ? https://joshua-riek.github.io/ubuntu-rockchip-download/boards/rock-5b-plus.html 5.10 kernel or 6.1 kernel? As already indicated, 'images' usually do have a too big scope for identifying the core issue for flaky boot or crashes etc. Also the Armbian '25.5.1' and '25.2.2' are just a text in some file. For my installations that have been always in-place upgraded, they are wrong anyway. Stating buster when I just did reboot into upgraded trixie for example (NanoPi-NEO). W.r.t. PCI-E, and other specific Rockchip HW, 5.10 (legacy) likely has best support. At least for my ROCK3A, I then have RPi camera V1 support (not tested but required overlays are missing in 6.1). Also w.r.t. the earlier mentioned PCI-E NVME+SATA issue, I got it working by merging the Radxa 5.10 ('latest') kernel + Radxa U-Boot '2017' with some Noble userspace from around December 2024. Now Bookworm by the way as I don't want netplan.io. In a running system (NVME working), you could run 'sudo lspci -vv' and look at Capabilities. When non-working NVME, so boot from SD-card and maybe SPI empty or disabled so only the U-Boot from the SD-card is used, it also might give hints. At least I see that SSD can do 8GT/s 4-lane, while RK3588s (NanoPi-R6C) then has to use 5GT/s 1-lane. It is 6.16.1-edge kernel with EDK2-UEFI v1.1 in eMMC. I don't understand all items, but at least some, w.r.t. power-save (ASPM) among those.

Hello In the past I installed armbian into emmc (a95xf3-air) with a USB key. All was ok. Now I would like to do a fresh install with the current release : I burned the image, make the changes (extlinux …) Boot with the "toothpick" method, ssh , the armbian release is the old one from the emmc, not the recent from the key. How is it possible to boot with the recent one with ssh (no display no keyboard) ? And after copy the new release to emmc. Thanks for help

@robertoj thank you for the answer. I will look at it. @Nick A Hi Nick, for your x98h box, is the wifi working after a reboot command?

I am observing the same thing. This can be easily fixed by setting the max frequency manually; https://www.incertum.net/post/2024/armbian-cm3588/ I checked `cpufreq-info | grep "available frequency steps"` And made this to set the correct speed for all the cores on the RK3588 NAS KIT # Little cluster (A55 cores: cpu0–cpu3) for c in 0 1 2 3; do cpufreq-set -c $c -g schedutil -u 1.80GHz done # Big cluster (A76 cores: cpu4–cpu7) for c in 4 5 6 7; do cpufreq-set -c $c -g schedutil -u 2.40GHz done Now running `watch -n1 "cat /sys/devices/system/cpu/cpu*/cpufreq/scaling_cur_freq"` together with `for i in $(seq 1 $(nproc)); do yes > /dev/null & done` shows the frequencies going up correctly, and down too when I `killall yes`

Not sure for what you need it but here: https://github.com/u-boot/u-boot tag specification and defined patch sources are here:

Get the microsd card mounted in your linux laptop Open /boot/armbianEnv.txt Add the line: extraargs=plymouth.enable=0 disablehooks=plymouth https://wiki.archlinux.org/title/Plymouth When you get booted in your armbian OS: sudo apt remove plymouth

I don't know anything about this patch, but check this: * Linux version you are using, vs Linux version corneliius has used * Try deleting the hunk in the patch that caused the problem you saw, and see if other hunks cause patch failures * If it is only one hunk causing the problem, look at the file that is affected by the patch drivers/auxdisplay/Makefile: do the lines mentioned in the hunk exist in the Makefile? In the expected line number?

Well apparently it is an issue, as stated above by @JFL. Also, after a RECALL, I'm the one who did the PR to change it from =y to =m. https://github.com/armbian/build/pull/8300/commits/bb71177580b34c1aeaeb601696d603b6dc49d326 As for udev and delay, you could edit the udev system file; /lib/udev/rules.d/50-udev-default.rules The line you want to edit is about 15 lines down. Change from rtc0 to rtc1.

Interesting. I wonder how the FriendlyElec stock kernel is assembled, because according to the log the stock system doesn’t use any Rockchip clock—only the HYM8563.

No issue, because it would behave the same as the default Rockchip system clock. Moreover, by default `fake-hwclock` is running, so after connecting a battery the user would just need to uninstall `fake-hwclock`. That shouldn’t be a problem.

I am trying building image natively on orangepiplus, but the scripts are not designed to complete it. System incompatibility, lack of packages for this distribution etc make it fail. How do you manage building it with this Armbian's official building tool ? I am very curious to discover the right approach as I need adding in-kernel modules, which you can't build as stand alone. At the same time, where can u-boot be downloaded from? I understand, that it was updated in 2024 : [🌿] {u-boot:1} Preparing u-boot targets packaging [ 2024.01 ] [🌿] {u-boot:1} Deploying u-boot binary target [ 2024.01 :: u-boot-sunxi-with-spl.bin ]

Hello again, Yes, I read the documentation, that's why I said my UART adapter doesn't work and I ordered a compatible one which is on its way. I managed to somehow solve my issue. Luckily I had another SSD lying around (Kingston) which works without any issue and is recognized every time and bootable. My original SSD (Crucial) has no hardware problems though, because I used it in an x86 Ubuntu machine and works without any problem whatsoever. So although I managed to solve my situation, the underlying problem is still there and may re-surface to someone else. I would like to re-state that the 25.5.1 armbian desktop does not even boot (edge or vendor) and the 25.2.2 boots but barely works on my board. Thanks for the time. PS. I ended up using JR ubuntu, because although non-supported anymore, it was the only one working. If you find my situation is worth further investigation and you would like to look into it, I can send you the serial logs once I get the proper adapter.

Same issue on Orange Pi 5, it used to start fast then hangs for ~ 1mn after some upgrades. systemd-analyze blame 1min 31.472s plymouth-quit-wait.service 25.068s apt-xapian-index.service 2.713s NetworkManager-wait-online.service 1.086s xrdp.service 1.074s NetworkManager.service 604ms snapd.seeded.service 538ms dev-nvme0n1p1.device 501ms snapd.service

On orangepi-5-plus without RTC battery setting `CONFIG_RTC_DRV_HYM8563=y` will cause the CPU Loadaverage to be always greater than 1. The rockchip64 kernel currently use `CONFIG_RTC_DRV_HYM8563=m` and blacklist rtc_hym8563 resolved the high CPU loadaverage.

Hi! Great to see Odroid C1 is being maintained again! Since I am using C1 for many years in my homelab I've given it a try. Unfortunately, after some time under heavier load I start to receive errors, system files becomes unreadable or straight kernel panic: The SD card is all right, tried different ones, checked cooling, applied fresh paste etc. Nothing. On the other hand, I've run this board for years with uptimes measured in months with rocksolid 5.11 build by @balbes150 from this thread: Since download links there are not available anymore, I share mine /boot/ here, maybe it will help somebody: https://mega.nz/file/nVdjFKSD#zGGyDAJ5YHEY3clyezbHUHMhsItgpzTyvW0Y9uKFVNQ I've went back to this kernel and it is stable again. Doing well with current stable trixie even. Could you use it to investigate stability issues? Or maybe @balbes150 could share his old patches with current maintainers? Thanks for your hard work, I appreciate it!

Has anyone been able to use the remote control that comes with the tv bos under Armbian? Does it reqire additional driver(s)? I want to run Kodi (or any better mediacenter) under Armbian.

For those without a battery connected wouldn't this be an issue?

This device is community-maintained. Armbian core team does not monitor its status nor gets involved into development here. However anyone from the community can step up and send fixes. I saw some stuff upstreamed from time to time for 5 ultra and max, hit somewhere between 6.16 and 6.18 I guess. When rockchip64 gets bumped further they will be available for these boards. Since mainline is still under development though sticking with 6.1.y vendor kernel isn't a bad option since most things work here.

Quite possible, actually expected as there is no support for current kernel: https://github.com/armbian/build/blob/main/config/boards/orangepi5pro.csc#L8 and build target is not maintained by core Armbian team https://docs.armbian.com/User-Guide_Board-Support-Rules/#community-maintained

We have been stuck on version 6.1 for almost half a year ever since the initial attempt to migrate to 6.12 caused some issues with HDMI if I remember correctly. Anyone knows if 6.12 or 6.14 are still in the works for this SBC? Thank you!

I have a different Orange PI board, but I have two partitiions of Armbian installed (25.8.1), one with the edge kernel, the other with the current (mainline) kernel. I checked my /boot/dtb/rockchip/ directories in each for the Orange PI 5 Pro, and there is a device tree for it with the edge kernel (rk3588s-orangepi-5-pro.dtb), but not with the current kernel. Perhaps a question would be: Is that devicetree (.dtb file) being utilized with the edge kernel? Might be worth looking into.

Thanks for the help. But the moment when the RTC is symlinked does matter. I mean, if the symlink changes too late, some services (for example, NTP synchronization) may be affected. I use specific software that starts as a service and immediately begins logging its activity. So if the system time changes significantly during boot, it causes problems for this software. I think this is a much better and more proper solution

sources are defined per board family, in your case sunxi:https://github.com/armbian/build/blob/main/config/sources/families/include/sunxi_common.inc Before attempting u-boot for a whole family I suggest to do some small scale tests at board level first by setting an override in the board config file. Example for a different board:https://github.com/armbian/build/blob/a7c19f1e35a65daf42f090ecc34ee1151ee6db23/config/boards/orangepi5-plus.conf#L31-L52