Igor

What do you mean by that? Images that are currently available are still from 25.2 ... new will be created soon. But you can anyway update them standard way: apt update + apt upgrade.

Also try to fix serial console, using different cable, different terminal application, ... doing this blindly ain't best practice.

We keep (most of) old images here: https://fi.mirror.armbian.de/archive/ https://fi.mirror.armbian.de/oldarchive/

This is what AI tells: Role of BL30 The BL30 component serves as the firmware for the Amlogic Secure Co-Processor (SCP). Its primary responsibilities include managing system-level operations such as Dynamic Voltage and Frequency Scaling (DVFS) and handling suspend/resume functionalities. These tasks are crucial for optimizing power consumption and ensuring efficient thermal management within the system. lists.denx.de In addition to BL30, there is a closely related component called BL301. BL301 acts as a board-specific firmware "plug-in" that provides customized parameters for DVFS and suspend/resume operations, tailored to the specific hardware configuration of the board. lists.denx.de Integration into the Boot Process The typical boot sequence in Amlogic devices follows this order: BL2 → BL30 → BL31 → BL33. Each stage is responsible for initializing specific aspects of the system, with BL30 focusing on power management and system control tasks. If any stage fails, the boot process may attempt to fall back to an alternative boot medium, although this behavior can vary depending on the specific SoC and its configuration. 7Ji’s Blog+1lists.denx.de+1 Understanding the role of BL30 is essential for developers and system integrators working with Amlogic platforms, especially when customizing firmware or troubleshooting boot-related issues. All Amlogic SoCs boards needs this. AFAIK here we don't have open source alternative. Only from the owner of this file after you sign certain papers https://www.amlogic.com/#Company/Contact/index.html This is complex operation and possible where / when there is a big enough interest and when more people join in this common interest and are willing to sacrifice months of their precious private time. https://github.com/crust-firmware/crust Here is an example of similar feature for some other SoC (allwinner). It works, but it's also not feature complete and probably never will. Many of those SoCs are already commercially obsolete, similar for the one you are asking.

Its 64bit, but support quality (tl;dr;) of Allwinner is behind Rockchip.

And I have ruled out that this is Armbian OS problem.

30-40% of GH runners we use are on arm64 platform. Compilation work in all combinations. Yes. This was lying around. Allwinner should work the same, but *edge kernels are experimental, adjust expectations. You have all tools you need, but nobody will be debugging some old experimental kernel just for you. Hint: when you generate image from sources, enable kernel headers install (i think parameter is INSTALL_HEADER=yes), so you have the one that are matching your kernel. In repository, they are probably different / not compatible. I checked if it works for latest stable (the only target that is worth spending time) kernel on supported hardware (that was around my desk). Unsupported hardware running unsupported kernel - it is expected that things will be failing.

This 100% works - we are even running daily automated tests on Bookworm, Jammy and Noble. https://github.com/armbian/os/actions/runs/15362508470/job/43232345405 Manual test on Rockchip64 (Bananapi M7) with 6.12.28-current, latest kernel from daily repository. System: https://paste.armbian.com/tiwisuhugi Here is build log:

Welcome to the club Perhaps you rather help us maintaining and fixing what is possible, so I would guide you away from things that are complete waste of time. Such as this. DKMS works on Armbian, but if this (shit) driver works, that is another question. In some cases it takes years before driver become usable ... while performance still sucks. https://docs.armbian.com/WifiPerformance/#xradio-xr819 AFAIK, there are no reliable driver for aic8800. Sorry for bringing bad news ...

Beta repository. - should be on servers where we push on directly and this error can't occur (unless something broke down) - is the same quality as other Linux distributions (there is no other distro that would maintain those custom devices) - currently assembly is broken as some patches are failing and thus (beta) repo is not getting any updates until its fixed https://github.com/armbian/os/actions/runs/15337468717/job/43160986093

Automation, not guys And they it can't be quicker then others elsewhere. Armbian adds a value on top of this. We have our EDGE kernels attached to latest mainline with additional stuff from Collabora and additional stuff from our own work. They are released daily in beta channels, which here doesn't have any fancy name (rawhide). https://github.com/armbian/build/blob/main/config/sources/families/include/rockchip64_common.inc#L36 https://github.com/armbian/build/blob/main/config/sources/mainline-kernel.conf.sh#L7-L14

I moved them to production already as I got another report they are working - so they are the same, but also community build images are mixed with those links ... this needs to be fixed eventually. Thank you for comprehensive testings. Here is image for 2w if anyone have it laying around: https://fi.mirror.armbian.de/incoming/igorpecovnik/orangepizero2w/archive/

Also need someone to test this: https://fi.mirror.armbian.de/incoming/igorpecovnik/orangepizero3/archive/

Yes. And we will keep fork maintained as long as possible. https://community.home-assistant.io/t/feedback-requested-deprecating-core-supervised-i386-armhf-armv7/880968/313

That was installed on my test machine which wasn't upgraded to latest (.115) yet.

Did you compare those things on the same hardware (architecture)?

We have vendor (v6.1.115) based kernel support here, where hw features usually works fine. I didn't do much of testings, only added board to the test farm. I never tried mainline here, perhaps someone else did. Logs: https://paste.armbian.com/uhofaquguz

Try running: ./compile.sh requirements And then try again.

Latest image: patrick@bananapif3:~$ sudo docker run hello-world [sudo] password for patrick: Unable to find image 'hello-world:latest' locally latest: Pulling from library/hello-world 2382d725750e: Pull complete Digest: sha256:dd01f97f252193ae3210da231b1dca0cffab4aadb3566692d6730bf93f123a48 Status: Downloaded newer image for hello-world:latest Hello from Docker! This message shows that your installation appears to be working correctly.

It would be nice to know what is the actual problem here, so we can at least add instructions. If XZ format is not supported, then probably it needs to be decompressed manually, perhaps renamed to ISO and then flash. Or there is some other trick here? P.S. Mirror fixed the problem on their side.

Strange. Rufus ... does it support compressed .xz ? What are exact steps which lead to this error? USB imager, as recommended, must work. I have manually tested some of those images and there were no problems.

Thank you for reporting. It's a problem with a mirror. Here is email I sent to admin: After this job finishes, mirror will be out of the pool, temporally. Here https://docs.armbian.com/Mirrors/#current-mirrors you can search for alternative download locations manually and try to use torrent - It is much faster.

Released so far: https://www.armbian.com/newsflash/armbian-25-5/ https://github.armbian.com/documentation/751/Release_Changelog/#v2551-2025-5-26 2025-05-31 14:04:06 nanopi-r5s/ 2025-05-31 13:20:18 nanopct4/ 2025-05-28 14:45:54 orangepizero2w/ ###################################### added may 31th 2025-05-28 17:50:12 orangepizero3/ 2025-05-28 13:31:42 sweet-potato/ (untested) ###################################### added may 29th 2025-05-28 06:44:05 bananapim4berry/ 2025-05-28 01:38:54 rock-5b/ 2025-05-27 23:26:24 khadas-vim4/ 2025-05-27 21:53:19 orangepi5/ 2025-05-26 23:30:28 rpi4b/ 2025-05-26 15:30:09 bananapim7/ 2025-05-26 12:44:43 orangepi5-plus/ 2025-05-26 06:54:00 khadas-vim1s/ #################################### added May 28th 2025-05-25 20:06:32 nanopct6/ 2025-05-25 18:58:55 khadas-edge2/ 2025-05-25 18:42:42 nanopi-r6s/ 2025-05-25 18:28:44 nanopi-m6/ 2025-05-25 18:13:14 rock-5t/ 2025-05-25 15:52:26 nanopct6-lts/ 2025-05-24 17:56:55 rock-5b-plus/ 2025-05-24 14:14:46 radxa-nio-12l/ #################################### added May 25th 2025-05-22 18:26:26 lepotato/ 2025-05-22 06:07:29 tritium-h3/ (untested) 2025-05-21 22:12:58 tritium-h5/ (untested) 2025-05-21 13:54:39 nanopi-r2s/ 2025-05-21 13:19:56 nanopi-r4s/ 2025-05-21 13:10:48 bigtreetech-cb2/ 2025-05-21 12:37:56 nanopi-r1/ 2025-05-21 08:24:34 odroidxu4/ 2025-05-20 21:02:21 bananapim2plus/ 2025-05-19 16:34:51 tinkerboard/ 2025-05-19 16:08:37 odroidc1/ 2025-05-19 13:41:51 cubox-i/ 2025-05-19 13:33:57 udoo/ 2025-05-18 18:14:19 rock-5c/ 2025-05-15 20:00:20 odroidm1/ 2025-05-15 19:45:23 odroidn2/ 2025-05-15 19:22:52 uefi-x86/ 2025-05-15 19:07:47 uefi-arm64/ 2025-05-15 17:31:47 odroidc4/ 2025-05-15 17:26:55 bigtreetech-cb1/ 2025-05-15 14:28:07 bananapicm4io/ 2025-05-15 14:10:36 rockpi-e/ 2025-05-15 13:15:45 helios4/ 2025-05-15 13:09:35 bananapim2pro/ 2025-05-15 12:47:30 clearfogpro/ 2025-05-15 12:14:00 nanopik2-s905/ 2025-05-15 11:21:35 mksklipad50/ 2025-05-15 10:05:49 odroidc2/ 2025-05-14 18:48:20 bananapif3/ I do this manually, every day few boards, those which I have around - basically I am manually updating my test farm. Many thanks to all that already helped in making this list longer. Release text is also WIP - even AI helps, its still quite a lot of work to convince (drunk / stupid) AI what do we want If you have time, please help - make image https://docs.armbian.com/Process_CI/#prepare-application-images-for-release-release-manager test it and provide feedback. In case you find something, hit red button here and make a short note. Thank you! code { font-family: Consolas,"courier new"; color: crimson; background-color: rgba(0, 0, 0, 0.2); padding: 2px; font-size: 105%; } code { font-family: Consolas,"courier new"; color: crimson; background-color: rgba(0, 0, 0, 0.2); padding: 2px; font-size: 105%; } code { font-family: Consolas,"courier new"; color: crimson; background-color: rgba(0, 0, 0, 0.2); padding: 2px; font-size: 105%; } Would you like to take on the role of maintainer for a specific board? https://docs.armbian.com/User-Guide_Board-Support-Rules/

It is diffucult to add certain advanced features to both / all base packages versions - which is the core difference between Debian and Ubuntu. And later more difficult and certainly less fun is keeping features operational. This is one of the main points behind providing both variants. Making assembly (Debian Bookworm, Ubuntu Jammy, Noble, ...) is easier, and cover several other aspects too, then adding and maintaining advanced features for both / all. Sometimes its nearly impossible (our resources are not endless and we have to remember we are in open source world - forks compete without contributions to commons) to add something to one, while its relatively easy adding it to another. At the end of the day, it is, like you said, more important that some feature works.

Network bridge is added within armbian-config Docker install ... Perhaps open a ticket https://github.com/armbian/configng so this is eventually done properly.