All Activity

This week's work centers on release pipeline modernization, desktop and userland refinements, and board and kernel platform maintenance. On the release and CI side, the build matrix gained codename parameterisation with Ubuntu 26.04 "resolute" set as default, a dedicated Bianbu target, and exposed map overrides, while standard-support targets now include UEFI desktops and a plain cloud variant. The KDE fast-HDMI matrix was switched from kde-neon to kde-plasma, mesa-vpu was dropped from auto-attached extensions, and external CI now skips slots with a warning when upstream sources break. Supporting fixes route forky/loong64 base-files lookups to the main archive and add AI cover image generation to the blog workflow. Desktop and userland changes focus on the Bianbu environment, where PVR DRI was enabled, detection corrected, menu entries added, systemd suspend re-enabled on K1, and gnome-initial-setup purged post-install. Broader fixes pass --allow-downgrades on pinned package installs, align LAN/WAN labels across IPv4 and IPv6 rows in the MOTD, harden console-width handling against invalid COLUMNS values, and correct output to /etc/armbian-image-release. Platform support sees explicit ARCH=arm64 declarations on five inheriting boards, validate-board-config now following inheritance from ${SRC}/config/boards, and targeted fixes for imx8m binman hooks and rockchip family tweaks under forky (addgroup → groupadd). Kernel and DTS work restores 6.18.y on sm8550, syncs CAINIAO CNIoT-CORE DTS from 6.18 to 6.12, disables broken drm/xe patches under uefi-loong64-7.0, and improves the SMART AM40 and Retroid Pocket board definitions. AX210 firmware lands for mainline, and Seeed Studio reComputer images join the catalogue. #Armbian #EmbeddedLinux #ARM64 #Rockchip #Ubuntu #KDE #Mainline ChangesAdd AX210 firmware for mainline kernel. by @ginkage in armbian/firmware#126Add Seeed Studio reComputer board images and vendor logo. by @baorepo in armbian/armbian.github.io#297boards: declare ARCH=arm64 explicitly on five inheriting boards. by @igorpecovnik in armbian/build#9747ci/external: skip slot with warning when upstream source breaks. by @igorpecovnik in armbian/armbian.github.io#301ci/maintenance-unit-tests: list docker state on the runner. by @igorpecovnik in armbian/configng#903desktops/bianbu: enable PVR DRI, fix detection, add menu entries. by @igorpecovnik in armbian/configng#897desktops/bianbu: purge gnome-initial-setup after install. by @igorpecovnik in armbian/configng#902desktops/bianbu: re-enable systemd suspend on K1. by @igorpecovnik in armbian/configng#900desktops: pass --allow-downgrades on pinned package install. by @igorpecovnik in armbian/configng#899DTS: Sync CAINIAO CNIoT-CORE from 6.18 to 6.12. by @retro98boy in armbian/build#9736Fix workflow bugs and add AI cover image generation. by @igorpecovnik in armbian/armbian.github.io#298fix: rename imx8m bare hook so binman pkg_resources patch can run. by @igorpecovnik in armbian/build#9752fix: rockchip family_tweaks fails on forky — addgroup → groupadd. by @igorpecovnik in armbian/build#9753fix: route forky/loong64 base-files lookups to the main archive. by @igorpecovnik in armbian/armbian.github.io#299fixing output to /etc/armbian-image-release. by @dittodhole in armbian/build#9729Handle invalid data in COLUMNS env var when determining console width. by @scollinson in armbian/build#9631kernel: add linux-initramfs-tool as dep of linux-image package. by @amazingfate in armbian/build#9731motd/20-ip-info: align (LAN)/(WAN) labels across IPv4 + IPv6 rows. by @igorpecovnik in armbian/build#9728os-info: add Ubuntu 26.04 (resolute). by @igorpecovnik in armbian/imager#134patch: disable broken drm/xe patches under uefi-loong64-7.0. by @igorpecovnik in armbian/build#9745release-targets/standard-support.manual: add UEFI desktops + plain cloud. by @igorpecovnik in armbian/armbian.github.io#304release-targets: codename parameterisation, resolute default, Bianbu target, exposed.map overrides. by @igorpecovnik in armbian/armbian.github.io#295release-targets: drop mesa-vpu from auto-attached extensions. by @igorpecovnik in armbian/armbian.github.io#302release-targets: switch fast-HDMI KDE matrix from kde-neon to kde-plasma. by @igorpecovnik in armbian/armbian.github.io#303Retroidpocket: fix current and add edge as target. by @EvilOlaf in armbian/build#9724sm8550: Restore kernel 6.18.y for current and edge branch. by @kasimling in armbian/build#9737SMART AM40 improve. by @retro98boy in armbian/build#9735userpatches: skip jammy in items-from-inventory. by @igorpecovnik in armbian/os#460validate-board-config: follow source ${SRC}/config/boards inheritance. by @igorpecovnik in armbian/build#9748View the full article

You aren't telling us much, like what software you are using. I would expect you to need to give it something like smb://10.10.2.101/NAME_OF_SHARE/ or sftp://10.10.2.101:/path/to/folder You will also need to give it a username and password somehow. But first; have you checked you can reach it on the network?

On my Debian13 VM that I use to build images for my boards, I have started to notice new errors that I have not seen before after this part of the logs... [🐳|🌱] After updating and cleaning image apt list cache [ 30 files ] [🐳|🌱] Image installed package hash [ ✅ base-files = 1-trixie-1armbian1-Bb095-U13.8--deb13u4-R55fa ] [🐳|🌱] Image installed package hash [ ✅ linux-image-current-rockchip64 = 6.18.26-S1fe0-Dedd0-P1f35-Cc077-H8200-HK01ba-Vc222-Bed86-R448a ] [🐳|🌱] Image installed package hash [ ✅ armbian-firmware = 1-SAa38b-B96c8-R448a ] [🐳|🌱] Image installed package hash [ ✅ armbian-bsp-cli-rockpro64-current = 1-PC1e15-V9ab0-H3400-Bd2e9-Rf36a ] [🐳|🌱] Image installed package hash [ ✅ linux-u-boot-rockpro64-current = 2025.01-S6d41-P6e2c-Hbf76-V6f3c-Be4e9-R448a ] [🐳|🌱] Image installed package hash [ ✅ linux-dtb-current-rockchip64 = 6.18.26-S1fe0-Dedd0-P1f35-Cc077-H8200-HK01ba-Vc222-Bed86-R448a ] [🐳|🔨] removed '/armbian/.tmp/rootfs-no-uuidgen-yet-15898-49210/etc/resolv.conf' [🐳|🔨] removed '/armbian/.tmp/rootfs-no-uuidgen-yet-15898-49210/sbin/initctl' [🐳|🔨] removed '/armbian/.tmp/rootfs-no-uuidgen-yet-15898-49210/sbin/start-stop-daemon' [🐳|🔨] removed '/armbian/.tmp/rootfs-no-uuidgen-yet-15898-49210/usr/sbin/policy-rc.d' [🐳|🌱] Removing qemu-user-static binary from chroot [ qemu-aarch64-static during image ] [🐳|🔨] removed '/armbian/.tmp/rootfs-no-uuidgen-yet-15898-49210/usr/bin/qemu-aarch64-static' [🐳|🌱] Unmounting [ /armbian/.tmp/rootfs-no-uuidgen-yet-15898-49210 ] Once the unmounting of rootfs-nouuidgen-yet begins I notice the following errors spam in the logs, 1000+ lines, and finally the target unmounts successfully and the image finishes building. The images still boot fine. umount: /armbian/.tmp/rootfs-no-uuidgen-yet-15898-49210/dev: target is busy. [🐳|🔨] tmpfs /armbian/.tmp/rootfs-no-uuidgen-yet-15898-49210/dev tmpfs rw,nosuid,size=65536k,mode=755,inode64 0 0 umount: /armbian/.tmp/rootfs-no-uuidgen-yet-15898-49210/dev/pts: not mounted. umount: /armbian/.tmp/rootfs-no-uuidgen-yet-15898-49210/dev: target is busy. umount: /armbian/.tmp/rootfs-no-uuidgen-yet-15898-49210/proc: not mounted. umount: /armbian/.tmp/rootfs-no-uuidgen-yet-15898-49210/sys: not mounted umount: /armbian/.tmp/rootfs-no-uuidgen-yet-15898-49210/tmp: not mounted. umount: /armbian/.tmp/rootfs-no-uuidgen-yet-15898-49210/var/tmp: not mounted. umount: /armbian/.tmp/rootfs-no-uuidgen-yet-15898-49210/run/user/0: not mounted. Is there a way to adjust some kind of timeout for this to prevent the massive amount of "target is busy" outputs in the logs? Thank you for any help or insights!

Hi, I have a Samba server on a TV Box. I installed Samba, defined a location for the server (srv/Public), created and user and UFW (firewall) allowing Samba protocols. I also defined srv/Public in /etc/fstab (Windows uses NTFS). Pls, tell if you need additional info. Regards, Sergioclr

I don't own that box so I can only give you generic help. You really would need to hook up a usb-uart adapter to access the uboot debug output to see what is happening. The second thing I would say is that I have no idea what ophub builds do as they are not Armbian builds, ophub is a fork, they don't contribute to Armbian development nor do they participate in these forums. So per the Armbian amlogic TV box instructions, the boot environment must be a clean android boot environment. Installing or running any other distribution may cause Armbian to not boot. So I'd recommend either reinstalling a clean Android image to test from, or investigate any differences in boot scripts to see what might be going on. But again getting access to the new boot serial console would help a lot.

Hi Everyone! I'm trying to install it on a TV Box with an Allwinner IK316 SOC and a NANYA 1303 eMMC, but Armbian can't find the eMMC at all... has anyone found a solution for this NAND?

Hi all, I'm trying to boot Armbian from USB on a Foxconn BANDOTT BA101 (Amlogic S905D, 2GB RAM, 8GB eMMC — same SoC/RAM/storage as a Phicomm N1, originally shipped with Android TV 7.1.2). I'm running purely from USB, never touching eMMC. What works Armbian_24.5.0_amlogic_s905d_jammy_6.6.28_server_2024.04.25.img (ophub build) Flashed to USB with balenaEtcher Plugged into the box, powered on Boots directly with no modifications — no manual u-boot.ext, no dtb edits, no extlinux changes Network, IR, HDMI all good What doesn't work Armbian_community_26.2.0-trunk.821_Aml-s9xx-box_resolute_current_6.18.26.img (official community build) I tried the following on this image, in order: As-is flash to USB → stuck at the Android boot logo, never proceeds. Generated/copied u-boot.ext for S905D into the boot partition → still stuck at Android logo. Selected meson-gxl-s905d-phicomm-n1.dtb as the FDT (this is the dtb that the working 24.5.0 image uses). Edited /boot/extlinux/extlinux.conf to match the working 24.5.0 image exactly — same kernel/initrd path style, same APPEND line (root=, rootflags, console=, etc.), same FDT pointing at meson-gxl-s905d-phicomm-n1.dtb. After all four steps, the 26.2.0 image still won't boot — the box hangs at the Android logo and never hands off to U-Boot/kernel from USB. Question Given that the 24.5.0 ophub S905D server image works out of the box, and the 26.2.0 community Aml-s9xx-box image fails even after replicating the working boot chain (u-boot.ext + N1 dtb + matching extlinux.conf), what else differs between these two images that could prevent USB boot on a stock-Android S905D box? Specifically I'd like to understand: Does the community 26.2.0 / kernel 6.18 image use a different boot script signature (aml_autoscript / s905_autoscript) that the stock Android U-Boot on these boxes won't chain-load? Is there a known incompatibility between kernel 6.18 / the resolute-current package set and the S905D mainline u-boot used by the 24.5.0 build? Is the official "Aml-s9xx-box" community image even intended to boot on locked Android TV boxes from USB, or is it built only for boards that already run mainline u-boot from eMMC? Any pointers appreciated. Happy to attach a serial console log if someone can suggest where in the boot chain to look. Thanks! extlinux.conf uEnv.txt

Has anyone installed Waydroid on Armbian 26.2.4? Are there any manual for this?

Good morning. Is there any means of writing an Armbian image to Allwinner H313 emmc? Phoenixsuit does not allow Armbiam images to be written. I have tried to compile (Make) Livesuit, but I gave up due to several incompatibilities (kernel, libraries, etc.). Thanks in advance.

Hello everyone. I’m attempting to mount a Windows network share on Armbian, but something seems off—both smb://10.10.2.101 and sftp://10.10.2.101 show up in red. The same addresses work perfectly on Rocky Linux; any thoughts on what I might be overlooking?

You need to provide more informations. I wasn't able to repreoduce. I works here.

What you’re seeing is open-source maintenance in practice. It’s not centrally orchestrated. We only manage to do this for some key features while DVB support is present in some kernels, but it’s not guaranteed to be consistently enabled across all targets unless someone explicitly maintains this functionality. -> https://github.com/armbian/build/pulls

Looks like both current and edge on sunxi64 are missing dvb support. Is this intentional? sunxi kernels include it, as do most other families.

@Nick A Great! and Good job! I've tested both the server and desktop versions of v0.6.4 on the A7S, and the DP output works perfectly on both. I've attached a screenshot for your reference. I’m using a USB‑C to HDMI docking station to test it. When I repeatedly plug and unplug it, the display works fine. The dock also has a keyboard receiver plugged in, which works normally. However, if I then plug a USB flash drive into the dock, the display output stops working (probably due to power supply issues). If I unplug the USB drive and plug it back in, the display works normally again. Additionally, I tested the Mainline v0.1. It seems to have issues on the A7S: the blue LED flashes after boot, but neither the Ethernet nor WiFi connects, and there is no output on the DP port. It appears the system cannot be login (I haven't tested via UART). However, the A7Z handles dual screens fine over HDMI and DP.

$ grep -i error /var/log/armbian-hardware-monitor.log [ 2.990779] rockchip-pcie f8000000.pcie: probe with driver rockchip-pcie failed with error -110 Looking at this I see the following: https://patchwork.ozlabs.org/project/linux-pci/patch/20210421083115.30213-1-jinsiyu940203@163.com/ "In function rockchip_pcie_host_init_port(), it defines a timeout value of 500ms to wait for pcie training. However, it is not enough for samsung PM953 SSD drive and realtek RTL8111F network adapter, which leads to the following errors: [ 0.879663] rockchip-pcie f8000000.pcie: PCIe link training gen1 timeout! [ 0.880284] rockchip-pcie f8000000.pcie: deferred probe failed [ 0.880932] rockchip-pcie: probe of f8000000.pcie failed with error -110 The pcie spec only defines the min time of training, not the max one. So set a proper timeout value is important. Change the value to 1000ms will fix this bug." Is it possible for team to add patch such as ../patch/kernel/rockchip64-current/0002-rockchip-pcie-increase-timeout.patch which contains the fix specified in the patchwork.ozlabs.org link above? A humble thank you for any more help and consideration on this!

@well0nez I didn't even think about that. When I made the original post almost two years ago, AI was much less capable than today. You just gave me a few ideas for different projects. As I said, unfortunately I don't have the HY300 anymore, but I do have a Magcubic L018(same chip I think), though it's in use every day as a bedroom TV. I don't think my girlfriend would appreciate if I started to hack it 😉 Especially if core functions like projecting the things coming through the HDMI port to the wall. The L018 has worse software than the HY300, in my opinion, so I've connected it to a certified Android TV 14 Box.

Hello, I updated to Armbian 26.2.1 trixie installaed on emmc. If I poweroff the nas and power on again it does't work, but if I attach an usb cable to the console it works. I think the problem must be related to the u-boot part.

It worked seamlessly! Thanks!

I made some patches so that the unified driver / TIM-VX for the NPU works on the A7Z. I am also building a MLIR pipeline that emit TIM-VX code, so hopefully we can get more flexibility running ML models on the SBC soon. https://github.com/MaverickLong/Radxa-A733-NPU-Unified-Driver-Support-Package

@curse well yes i started this port because shift did a static analysis of the firmware with ai. but well its not usable (mostly) and there are obviously differences between the 2 projectors. i would somehow call it hacky, its nowhere near to be stable but for now its acceptable. i dont know if i will ever be able to complete it.

@well0nez Thank you. Interesting project. It seems to originally be based on a HY300 project. Unfortunately I don't have the Magcubic HY300 anymore, but both of the projects looks really promising, especially the HY310 project you linked that seems to be fully functional.

Hi everyone! I tried installing hardware decoding from this link - https://forum.armbian.com/topic/32449-repository-for-v4l2request-hardware-video-decoding-rockchip-allwinner/ , but it didn't work for me. My board version is 5.1. Has anyone been able to fix this?

I tried to enable the I2S / PCM5102 service with AI, but unfortunately, despite numerous attempts, I couldn't. But anyone can use a USB audio card. Maybe someone else will figure out how to enable it 🙂 Best regards

Support for DP Alt Mode is now functional for the A7S running kernel 6.6. https://github.com/NickAlilovic/build/releases/tag/Radxa-cubie-A7S-v0.6.4

have alook here: https://github.com/well0nez/allwinner-h713-linux its for an hy310 but its still a h731 ;)