All Activity

hi. can you make only cli image? i will test it

Thanks @AurealNix - that's good intel. Only 20 of those patches of the 109 failed - I had a quick look through them and I can see you are looking to prioritise the GPU (all those GOOGLE_ flags to make youtube stream better etc) amongst other things. For my project I only needed to run headless so I could install your image and uninstall KDE if it didn't work out with my current prototype for this outdoor noise cancelling thing (I was attracted to the opi 4a because it had a bit more all round grunt) I figured I'd stick with armbian as it would have a better chance of the PR for changes to dtb etc being accepted into "stable".. I think I might go with your .dts and manually apply the changes to the other files and then build from there - used to have the stomach for this back in the day when I used to hack stuff before the internet, but I really need to get onto the noise cancelling project as my wife complains about the noise of the road when we have a BBQ outdoors and I reckon i could go some way to disturb that with some software. Your retro gaming image sounds top notch! I'm still to Master Halo 4 on the X-Box one given all these other distractions. The board is now in the bottom drawer for now as the DJI Mic Mini 2 just turned up today and the other board recognises it ok and its on to stage 2 of this prototype. Enjoy the football .. Cheers

you fancy a fun distraction rather than a frustrating one: I'm also putting together a retro-gaming image for this SoC (HDMI to the TV + a gamepad = a neat little emulation box, and it's the same chip as some of the new handhelds). Early days, but it's in the works — one more reason to keep the board out of the eBay pile. 🙂

Hi @bickns, thanks a lot for giving it a go! 🙏 The reason dropping them into userpatches/kernel/archive/sunxi-6.18 fails is that these aren't Armbian-style patches — they're a git format-patch series on top of plain vanilla 6.18.38 (base commit 1db74811b), meant to be applied to a clean tree in order. Armbian applies its own sunxi patch set first, so they conflict against that. I didn't use the Armbian build system at all — I built mainline from scratch: Clone/checkout vanilla 6.18.38, then git am /path/to/patches/0*.patch (all 106, in order). Copy opi4a_blindboot_defconfig into arch/arm64/configs/ and make opi4a_blindboot_defconfig. make ARCH=arm64 CROSS_COMPILE=aarch64-linux-gnu- -j$(nproc) Image dtbs modules Wrap it for the BSP U-Boot (this is the important bit — it only boots a legacy uImage): mkimage -A arm -O linux -T kernel -C none -a 0x41000000 -e 0x41000000 -d arch/arm64/boot/Image uImage Two things to know: A handful of the patches are BSP-U-Boot workarounds (static /memory, display FDT stubs, etc.) — only needed to boot under the proprietary BSP U-Boot. If you're on mainline U-Boot / a normal Armbian boot flow, you'd skip those. The README groups the patches by destination so you can see which is which. If you want to integrate into Armbian proper, the cleaner path is probably the board DTS + defconfig + the generic sunxi fixes as individual patches, rather than the whole display series in one go (that one's a big port of the minimyth2/Suess work and would need adapting to Armbian's tree). Happy to help however's useful — glad it's of interest!

Yes, I already forgot this and maybe more; Armbian uses this by default now AFAIK, so this is another issue like netplan; I normally do in-place dist-upgrades, so my older SBC's I bought when Buster was the current Debian version, so that is still somewhere in a tag in a image generation file or so, but os-release is 13 and kernel 6.18. But for 1 case, I did take a recent/new Armbian image (Trixie) and it did cost a lot of time as the default settings for that don't work properly in my home network, so after reading freedesktop.org docs and with extra setting in conf file, it worked. But openresolv and a fixed resolv.conf is a simpler base.

@AurealNix - great effort and thanks for putting me in touch with Claude 5, I was just using Gemini and it was struggling a fair bit!! In fact I had just put this board on ebay as it was causing too much distraction so your last two posts looked hopeful. Sorry my spanish isn't that great so I'll stick with English. I was hoping to just put these 106 patch files directly into the userpatches/kernel/archive/sunxi-6.18 folder and then run compile.sh with USER_PATCHES=yes, but that didn't work as patch files in that location seem to be tempremental and it couldn't find a lot of the targets. I tried removing the ones with errors which didn't work either, so went down the route of adding running compile.sh with kernel-patch parameter then copying your sun55i-t527-orangepi-4a.dts into ./cache/sources/linux-kernel-worktree/6.18__sunxi64__arm64/arch/arm64/boot/dts/allwinner/sun55i-t527-orangepi-4a.dts to get a new patch file, but this failed to compile likely because it doesn't have the changes you made to the .dtsi and other files . It would be good to know how you and Claude were building the kernel - I can see in your README that you are doing it directly into linux 6.18.38 tree, whereas I was working out of a clone of armbian in github. I would like to spend more time on this but need to get onto the real reason I bought the board (an outdoor noise cancelling project) so I'm going to use another board in the meantime. Cheers

Hello, I am experiencing the same issue: CPU1-3 are offline, and they can't be activated by echo 1 > /sys/devices/system/cpu/cpu1/online or similar commands, the system is responding with "invalid argument" error. I did not notice this nuisance until last week. I would like not having to reinstall the hole system, because it is running my Nextcloud and some other home services. Is there any way to install a kernel that uses all 4 CPU, without having to reinstall the hole system? Will it work, if I just copy /boot from one of the older images onto the sdcard in use?

Now I attacked the issue from the clients side and tried to connect from a "real" (Debian) system. Client Syslog output on connect looks quite comparable, but on the "failing" AP I seem to have a DNS issue: ping and traceroute do not work from the client in that case (and work when I connect to the "Bookworm AP"). From the AP computers I can alway resolve hostnames. So I will investigate this further (on Bookworm the resolveconf service is used, on Trixie systemd-resolved: so I assume an issue in the DNS setup of dnsmasq). I will not follow the "RADIUS" path anymore (for now), the brcmf_psm_watchdog_notify issue will also remain untouched (it is with me for a long time on the BPiPro now).

Perhaps here: https://issues.chromium.org/issues?q=status:open as xtradeb ppa is packaking service only. I don't think they do anything else that that.

I've seen this same issue with Discord, since March/April this year. Just going to https://discord.com/app (no login necessary) is enough to trigger it. From what I gather, this seems to be some sort of problem with the Chromium build Armbian uses (it appears to be the same build from the xtradeb ppa). This seems to happen on: 1. Raspberry Pi 4 on Armbian Ubuntu Noble 2. Some random Ubuntu VM ARM64 on Apple Silicon using Chromium from xtradeb/apps PPA 3. Dragon Q6A w/ Armbian Ubuntu Noble No workarounds seem to work. Turning off GPU acceleration, disabling sandboxing, etc. all make no difference. But it does not occur on: 1. Google Chrome 2. Chromium 145 or earlier ARM64 from Armbian or Xtradeb PPA, on all devices. Only 146 or later is affected for some reason 3. Linux Mint's Chromium x64 build 4. x64 Chromium snap package on Ubuntu I have not tested xtradeb PPA (or Armbian)'s x64 build of Chromium. Based on the above, I do not think this is an Armbian specific issue. Anyone have any ideas on where to report this?

Where can I find the image?

Update: the ready-to-flash Debian 13 + KDE Plasma image is now available: https://github.com/ut-slayer/orangepi-4a-mainline/releases/tag/v0.1 Boots straight to an accelerated Plasma desktop. Feedback welcome!

So you try routed AP instead of bridged. Trixie might handle 'iptables' in a different way, depends also what alternatives you have for that and/or if it is netfilter underneath. This kind of manual setup (routed, hostapd) I did before Bookworm only a few times, rest is always bridged and basically using only NM. It seems to me that something on IP layer is not accepted by your phone, so it disconnects. You need to dig more and deeper and also still take in to account what is said about that brcmf_psm_watchdog_notify issue. But it seems yest another issues.

Unnoficial Image: Download Link enabled CMA, increased CMA to 32mb and enabled rknn Board logs: https://paste.armbian.com/hoqejiqeno Build logs: https://paste.armbian.eu/cekehuhule Installing ffmpeg with encoders and decoders, armhf - armv7: ffmpeg-rockchip-armhf.zip On rv1126 armbian device: cd /media/usb/esp32-rv1126-ffmpeg-rockchip ./install-on-target.sh Verify: /opt/ffmpeg-rockchip/bin/ffmpeg -hide_banner -encoders | grep -Ei 'rkmpp|mpp' /opt/ffmpeg-rockchip/bin/ffmpeg -hide_banner -decoders | grep -Ei 'rkmpp|mpp' /opt/ffmpeg-rockchip/bin/ffmpeg -hide_banner -filters | grep -Ei 'rkrga|rga' Please don't trust pre-compiled files. compile yourself: Use this only if you really want hardware H.264 encode for the go2rtc bridge. A normal upstream FFmpeg build with --enable-rkmpp may expose Rockchip decoders but not the encoder set needed here. For encoder support, use an encoder-capable Rockchip FFmpeg fork such as nyanmisaka/ffmpeg-rockchip, which documents h264_rkmpp, hevc_rkmpp, and mjpeg_rkmpp under encoders. Build FFmpeg With RKMpp Encoders: sudo apt update sudo apt install -y \ git build-essential pkg-config cmake meson ninja-build \ yasm nasm libdrm-dev libudev-dev The MPP and RGA sources below are hosted on Gitee. If git clone prompts with Username for 'https://gitee.com':, sign in with a Gitee account. That prompt is expected for these URLs and is not an FFmpeg/RKMpp build error. Build and install Rockchip MPP runtime and headers: mkdir -p ~/src cd ~/src if [ -d rkmpp/.git ]; then git -C rkmpp pull --ff-only elif [ -e rkmpp ]; then echo "rkmpp exists but is not a git checkout; move it aside or choose a new directory" >&2 exit 1 else git clone -b jellyfin-mpp --depth=1 https://gitee.com/nyanmisaka/mpp.git rkmpp fi cd rkmpp cmake -S . -B build \ -DCMAKE_BUILD_TYPE=Release \ -DCMAKE_INSTALL_PREFIX=/usr/local \ -DBUILD_SHARED_LIBS=ON \ -DBUILD_TEST=OFF cmake --build build -j"$(nproc)" sudo cmake --install build sudo ldconfig Build and install RGA. The FFmpeg fork can use RGA for hardware color conversion/scaling; this is useful because the ESP32 camera enters as raw YUYV and encoders usually want a YUV420/NV12-style input. cd ~/src if [ -d rkrga/.git ]; then git -C rkrga pull --ff-only elif [ -e rkrga ]; then echo "rkrga exists but is not a git checkout; move it aside or choose a new directory" >&2 exit 1 else git clone -b jellyfin-rga --depth=1 https://gitee.com/nyanmisaka/rga.git rkrga fi if [ -d rkrga_build ]; then meson setup --reconfigure rkrga_build else meson setup rkrga rkrga_build \ --prefix=/usr/local \ --libdir=lib \ --buildtype=release \ --default-library=shared \ -Dcpp_args=-fpermissive \ -Dlibdrm=false \ -Dlibrga_demo=false fi meson compile -C rkrga_build sudo meson install -C rkrga_build sudo ldconfig Build FFmpeg to /opt/ffmpeg-rockchip so it does not overwrite the system FFmpeg: set -e cd ~/src if [ -d ffmpeg-rockchip/.git ]; then git -C ffmpeg-rockchip pull --ff-only elif [ -e ffmpeg-rockchip ]; then echo "ffmpeg-rockchip exists but is not a git checkout; move it aside or choose a new directory" >&2 exit 1 else git clone --depth=1 https://github.com/nyanmisaka/ffmpeg-rockchip.git ffmpeg-rockchip fi cd ffmpeg-rockchip make distclean >/dev/null 2>&1 || true # RV1126 is ARMv7-A. Disabling ARMv5TE avoids assembler failures in # libavcodec/arm/mlpdsp_armv5te.S on some armhf toolchains. PKG_CONFIG_PATH=/usr/local/lib/pkgconfig:/usr/local/lib/arm-linux-gnueabihf/pkgconfig:/usr/lib/arm-linux-gnueabihf/pkgconfig \ ./configure \ --prefix=/opt/ffmpeg-rockchip \ --enable-gpl \ --enable-version3 \ --enable-libdrm \ --enable-rkmpp \ --enable-rkrga \ --disable-armv5te \ --enable-shared \ --disable-static \ --disable-doc \ --disable-debug make -j"$(nproc)" sudo make install echo /opt/ffmpeg-rockchip/lib | sudo tee /etc/ld.so.conf.d/ffmpeg-rockchip.conf sudo ldconfig This only disables FFmpeg's CPU hand-written assembly. It does not disable the Rockchip MPP encoder or RGA support, and it is acceptable for this camera path because the input is only 320x240 @ 3 fps. Verify that FFmpeg was built with actual encoders: /opt/ffmpeg-rockchip/bin/ffmpeg -hide_banner -encoders | grep -Ei 'rkmpp|mpp' /opt/ffmpeg-rockchip/bin/ffmpeg -hide_banner -decoders | grep -Ei 'rkmpp|mpp' /opt/ffmpeg-rockchip/bin/ffmpeg -hide_banner -filters | grep -Ei 'rkrga|rga' Expected Result: /opt/ffmpeg-rockchip/bin/ffmpeg -hide_banner -encoders | grep -Ei 'rkmpp|mpp' /opt/ffmpeg-rockchip/bin/ffmpeg -hide_banner -decoders | grep -Ei 'rkmpp|mpp' /opt/ffmpeg-rockchip/bin/ffmpeg -hide_banner -filters | grep -Ei 'rkrga|rga' V..... h264_rkmpp Rockchip MPP (Media Process Platform) H264 encoder (codec h264) V..... hevc_rkmpp Rockchip MPP (Media Process Platform) HEVC encoder (codec hevc) V..... mjpeg_rkmpp Rockchip MPP (Media Process Platform) MJPEG encoder (codec mjpeg) V..... av1_rkmpp Rockchip MPP (Media Process Platform) AV1 decoder (codec av1) V..... h263_rkmpp Rockchip MPP (Media Process Platform) H263 decoder (codec h263) V..... h264_rkmpp Rockchip MPP (Media Process Platform) H264 decoder (codec h264) V..... hevc_rkmpp Rockchip MPP (Media Process Platform) HEVC decoder (codec hevc) V..... mjpeg_rkmpp Rockchip MPP (Media Process Platform) MJPEG decoder (codec mjpeg) V..... mpeg1_rkmpp Rockchip MPP (Media Process Platform) MPEG1VIDEO decoder (codec mpeg1video) V..... mpeg2_rkmpp Rockchip MPP (Media Process Platform) MPEG2VIDEO decoder (codec mpeg2video) V..... mpeg4_rkmpp Rockchip MPP (Media Process Platform) MPEG4 decoder (codec mpeg4) V..... vp8_rkmpp Rockchip MPP (Media Process Platform) VP8 decoder (codec vp8) V..... vp9_rkmpp Rockchip MPP (Media Process Platform) VP9 decoder (codec vp9) ... overlay_rkrga VV->V Rockchip RGA (2D Raster Graphic Acceleration) video compositor ... scale_rkrga V->V Rockchip RGA (2D Raster Graphic Acceleration) video resizer and format converter ... vpp_rkrga V->V Rockchip RGA (2D Raster Graphic Acceleration) video post-process (scale/crop/transpose)

Gemini info 😞 ..... In the official, pure Armbian for Banana Pi M2 Pro, enabling SPI is currently extremely difficult (and even impossible for the average user) If you absolutely need the SPI bus for your project: Operating system change: SPI works without any problems on official system images provided directly by the board manufacturer (BPI-M2 Pro Debian/Ubuntu with kernel 4.9 from Amlogic). The manufacturer uses an older but fully modified kernel (Kernel BSP), where all pins are rigidly and correctly assigned. Or... Staying with Armbian: If you must stick with Armbian, it's safer to abandon SPI attempts and—if possible—connect the target device to a free I2C bus or UART serial communication (which work flawlessly in Armbian on this board). But ultimately, the Armbian system developer will have to decide. (Maybe there is some method to enable SPI on the latest Armbian images/firmware?) 🙂 Check https://docs.banana-pi.org/en/BPI-M2_Pro/BananaPi_BPI-M2_Pro Armbian older firmware (bananapi manufacturer's firmware) (bookworm): https://drive.google.com/drive/folders/1DPkc7AHEZVS7PT_yzhNH1Dfuiz99mE0p

I'm facing the same issue and haven't found a way to enable USB 3.0. There doesn't seem to be an overlay like the one in Radxa OS for enabling it. If anyone knows how to get USB 3.0 working, I'd really appreciate the help.

Now I have tried to bring my non-working Trixie setup as near as possible to my working Bookworm setup (both BananaPi Pro boards) (and: no bridging stuff here!): -I replaced systemd-networkd with NetworkManager on Trixie -I restricted NM to handling of end0 (wlan0 as "unmanaged"), set up everything "the old way" as it is working on Bookworm (manual hostapd, dnsmasq, iptables) And now I try to connect from my mobile phone to both APs: Bookworm is okay as always, Trixie says "connect w/o internet" (although the very first time before rebooting it connected w/ Internet). In the extended dnsmasq log I see in both cases lots of connection names, but anyway the Trixie AP claims to have "no internet". The only visible difference is the log output for hostapd (both v2.10): in the failing case the following lines are printed for every connection, whereas in the working case these are not there: 2026-07-14T19:15:41.505822+02:00 bananapipro hostapd: wlan0: STA 22:50:7b:ac:e3:2e IEEE 802.11: associated 2026-07-14T19:15:41.541156+02:00 bananapipro hostapd: wlan0: STA 22:50:7b:ac:e3:2e RADIUS: starting accounting session 0F32AA180BFF0C2C 2026-07-14T19:15:41.542161+02:00 bananapipro hostapd: wlan0: STA 22:50:7b:ac:e3:2e WPA: pairwise key handshake completed (RSN) ....(many lines of log output from dnsmasq) 2026-07-14T19:15:58.724493+02:00 bananapipro hostapd: wlan0: STA 22:50:7b:ac:e3:2e IEEE 802.11: disassociated I have no clue what activates this RADIUS stuff and why there is this associated / disassociated pair (or why it is not there in the Bookworm case).

Hi, I can confirm that the issue is present in Armbian v26.5.1 with current kernel versions 6.18.x, and that it is resolved in the edge kernels 6.19.x and 7.1.x. Armbian v26.5.1 - current kernel 6.18.x - NOT OK (many retrans and ~368 Mb/sec): root@odroidtest:~# grep -a --null-data U-Boot /dev/mtd0ro U-Boot SPL 2026.01_armbian-2026.01-S127a-Pb401-H8652-Vab81-Bd0d2-R448a (May 20 2026 - 22:28:40 +0000) root@odroidtest:~# uname -a Linux odroidtest 6.18.35-current-rockchip64 #2 SMP PREEMPT Tue Jun 9 10:28:53 UTC 2026 aarch64 GNU/Linux root@odroidtest:~# iperf3 -c 192.168.0.1 Connecting to host 192.168.0.1, port 5201 [ 5] local 192.168.0.15 port 42940 connected to 192.168.0.1 port 5201 [ ID] Interval Transfer Bitrate Retr Cwnd [ 5] 0.00-1.00 sec 43.0 MBytes 360 Mbits/sec 199 26.9 KBytes [ 5] 1.00-2.00 sec 44.9 MBytes 376 Mbits/sec 172 26.9 KBytes [ 5] 2.00-3.00 sec 41.8 MBytes 350 Mbits/sec 188 39.6 KBytes [ 5] 3.00-4.00 sec 45.1 MBytes 379 Mbits/sec 178 33.9 KBytes [ 5] 4.00-5.00 sec 41.9 MBytes 351 Mbits/sec 185 17.0 KBytes [ 5] 5.00-6.00 sec 46.5 MBytes 390 Mbits/sec 168 35.4 KBytes [ 5] 6.00-7.00 sec 42.8 MBytes 359 Mbits/sec 170 24.0 KBytes [ 5] 7.00-8.00 sec 47.0 MBytes 394 Mbits/sec 164 31.1 KBytes [ 5] 8.00-9.00 sec 41.0 MBytes 344 Mbits/sec 178 18.4 KBytes [ 5] 9.00-10.00 sec 44.8 MBytes 375 Mbits/sec 167 15.6 KBytes - - - - - - - - - - - - - - - - - - - - - - - - - [ ID] Interval Transfer Bitrate Retr [ 5] 0.00-10.00 sec 439 MBytes 368 Mbits/sec 1769 sender [ 5] 0.00-10.00 sec 438 MBytes 368 Mbits/sec receiver iperf Done. Armbian v26.5.1 - edge kernel 6.19 - OK root@odroidtest:~# grep -a --null-data U-Boot /dev/mtd0ro U-Boot SPL 2026.01_armbian-2026.01-S127a-Pb401-H8652-Vab81-Bd0d2-R448a (May 20 2026 - 22:28:40 +0000) root@odroidtest:~# uname -a Linux odroidtest 6.19.0-edge-rockchip64 #1 SMP PREEMPT Sun Feb 8 21:03:27 UTC 2026 aarch64 GNU/Linux root@odroidtest:~# iperf3 -c 192.168.0.1 Connecting to host 192.168.0.1, port 5201 [ 5] local 192.168.0.15 port 35634 connected to 192.168.0.1 port 5201 [ ID] Interval Transfer Bitrate Retr Cwnd [ 5] 0.00-1.00 sec 113 MBytes 950 Mbits/sec 0 1.81 MBytes [ 5] 1.00-2.00 sec 111 MBytes 933 Mbits/sec 0 1.81 MBytes [ 5] 2.00-3.00 sec 112 MBytes 944 Mbits/sec 0 1.81 MBytes [ 5] 3.00-4.00 sec 113 MBytes 947 Mbits/sec 0 1.81 MBytes [ 5] 4.00-5.00 sec 112 MBytes 941 Mbits/sec 0 1.81 MBytes [ 5] 5.00-6.00 sec 111 MBytes 933 Mbits/sec 0 1.81 MBytes [ 5] 6.00-7.00 sec 113 MBytes 948 Mbits/sec 0 1.81 MBytes [ 5] 7.00-8.00 sec 111 MBytes 934 Mbits/sec 1 929 KBytes [ 5] 8.00-9.00 sec 113 MBytes 948 Mbits/sec 0 929 KBytes [ 5] 9.00-10.00 sec 113 MBytes 944 Mbits/sec 0 929 KBytes - - - - - - - - - - - - - - - - - - - - - - - - - [ ID] Interval Transfer Bitrate Retr [ 5] 0.00-10.00 sec 1.10 GBytes 942 Mbits/sec 1 sender [ 5] 0.00-10.00 sec 1.09 GBytes 939 Mbits/sec receiver iperf Done. Armbian v26.5.1 - edge kernel 7.1 - OK root@odroidtest:~# grep -a --null-data U-Boot /dev/mtd0ro U-Boot SPL 2026.01_armbian-2026.01-S127a-Pb401-H8652-Vab81-Bd0d2-R448a (May 20 2026 - 22:28:40 +0000) root@odroidtest:~# uname -a Linux odroidtest 7.1.0-edge-rockchip64 #1 SMP PREEMPT Sun Jun 14 14:58:38 UTC 2026 aarch64 GNU/Linux root@odroidtest:~# iperf3 -c 192.168.0.1 Connecting to host 192.168.0.1, port 5201 [ 5] local 192.168.0.15 port 37256 connected to 192.168.0.1 port 5201 [ ID] Interval Transfer Bitrate Retr Cwnd [ 5] 0.00-1.00 sec 114 MBytes 958 Mbits/sec 0 1.72 MBytes [ 5] 1.00-2.00 sec 111 MBytes 933 Mbits/sec 0 1.72 MBytes [ 5] 2.00-3.00 sec 113 MBytes 948 Mbits/sec 0 1.72 MBytes [ 5] 3.00-4.00 sec 112 MBytes 944 Mbits/sec 0 1.72 MBytes [ 5] 4.00-5.00 sec 112 MBytes 936 Mbits/sec 0 1.72 MBytes [ 5] 5.00-6.00 sec 113 MBytes 947 Mbits/sec 0 1.72 MBytes [ 5] 6.00-7.00 sec 112 MBytes 935 Mbits/sec 0 1.72 MBytes [ 5] 7.00-8.00 sec 113 MBytes 945 Mbits/sec 0 1.72 MBytes [ 5] 8.00-9.00 sec 112 MBytes 943 Mbits/sec 0 2.19 MBytes [ 5] 9.00-10.00 sec 112 MBytes 937 Mbits/sec 0 2.19 MBytes - - - - - - - - - - - - - - - - - - - - - - - - - [ ID] Interval Transfer Bitrate Retr [ 5] 0.00-10.00 sec 1.10 GBytes 943 Mbits/sec 0 sender [ 5] 0.00-10.00 sec 1.10 GBytes 941 Mbits/sec receiver iperf Done.

Hi — I've been following this thread, the 100 Mbit ethernet cap and the silent mainline boot hang especially. I ended up doing a full mainline bring-up of the 4A on 6.18.38 and got gigabit ethernet working properly (RX/TX), plus the boot hang sorted, so I wanted to share the whole patch series in case it helps here. Full disclosure: I'm not a kernel developer by trade — I got most of this working with a lot of help from AI coding agents, leaning hard on the BSP 5.15 as the source of truth. In my own testing everything seems stable, but I'd genuinely welcome anyone who actually knows this stuff to review the patches and tell me what I got wrong or did in a hacky way. No egos here — I'm sharing it in case it's useful and to get feedback, not claiming it's perfect. 🙂 What works (confirmed on hardware): HDMI KMS display (DE33/DE3.5 + TCON-TV + DW-HDMI 2.0 + Inno PHY) — 720p/1080p, clean VSU8 scaling GPU Mali-G57 (Panfrost) — accelerated (KDE Plasma Wayland) HDMI audio (i2s2 → dw-hdmi) — PCM out to the TV Ethernet GMAC1 (YT8531 PHY, RGMII) — full gigabit RX/TX WiFi AP6256 (BCM43456, brcmfmac, SDIO) — 2.4/5 GHz scan + assoc reboot / poweroff (sunxi_wdt / AXP717) AFBC scanout (v350 AFBD decoder) USB — the 4 rear ports (all USB 2.0; the board doesn't wire the SoC's USB3) — HID + mass storage, hotplug THS thermal sensors (5 zones: cpu_l / cpu_b / gpu / npu / ddr) + 110 °C critical trip CPU cpufreq/DVFS (little cluster up to 1.416 GHz, big cluster up to 1.8 GHz) + thermal throttling GPU devfreq/DVFS (Panfrost, 150–600 MHz) + thermal throttling HDMI HPD / hotplug (native, no force) Known not working / help wanted: eMMC — not tested (I don't have a module). I saw the mmc2: error -110 reports earlier in the thread; happy to dig into it with anyone who has eMMC. Hardware video decode (VPU) — no driver in 6.18 yet (software decode works). Suspend/hibernate — untested. 4GB variant — expected to work out of the box (the bootloader auto-detects the RAM size and fills in /memory at boot; verified over UART on the 2GB board). I only own the 2GB variant, so if you have the 4GB one, a free -h confirmation would be very welcome. NVMe SSD — untested, no drive here. Everything so far has only run from microSD. Freshness note: the cpufreq/thermal/GPU-devfreq block landed this very week and is still under soak testing on my board — everything else has more mileage. I'll report anything that shakes out in this thread. Repo / patches: https://github.com/ut-slayer/orangepi-4a-mainline It's a git format-patch series (~106 patches) on top of 6.18.38 vanilla, plus a defconfig and the board .dts. The README groups the patches by intended destination: generic sunxi fixes (pinctrl, watchdog vs PSCI, AXP717 poweroff, a RESET_GPIO/mmc-pwrseq bug affecting any sunxi board with SDIO WiFi), the HDMI KMS display series (largely ported/adapted from the minimyth2 / Justin Suess H728 work, with original authorship kept in the patch headers), the A523 CPU CCU + thermal + DVFS, board integration (dts + defconfig), and a few U-Boot BSP workarounds that go away with mainline U-Boot. Coming next: a ready-to-flash Debian 13 (Trixie) image with accelerated Panfrost and KDE Plasma on Wayland — I'll share it in this thread once it's had a bit more soak testing. Happy to open a PR against armbian/build (sun55iw3 family) with the board dts + defconfig + patches if there's interest. Feedback and testers welcome! Credits: minimyth2 / Justin Suess for the H728 display groundwork, and the linux-sunxi community. Cheers.

moved to tvboxes

These error messages don't look good. Have you tried setting up and working with that wifi device alone (i.e. without all that bridging stuff)? This seems to me like a problem with either firmware loading for this adapter, or an unusable configuration, like unusable tx/rx power or channel for the specified region (e.g. if there has been no region code set for that interface).

@jshkol, I saw that an Orange4Pro build was approved yesterday—I have one of those myself. What’s the status regarding the NPU? I actually wanted to start a build myself at one point, and it’s great to see you so dedicated to it. How would I go about building it using the new image? I have a homelab and would love to test it out and share some feedback with you.

Hi ! I was trying to use SPI communication with Banana Pi M2 Pro. I am currently using Armbian 26.5.1 and I realized that I cannot find spidev from /boot/dev/ . I tried loading all dtbo files from armbian-config but still I can't get SPI. Is there something that I haven't check yet?

It seemed like EVERYONE was talking about "looking into" getting Armbian going on this box. I did my best. Built from the Rock5b images. WiFi: Broadcom BCM43752 / AP6275P 802.11ax "WiFi 6" Panthor + Mesa GPU stack Open PanVK Vulkan + Panfrost GL GPU driver Gpu no longer pinned to 800MHz. Power toggles cycle it 600MHz, 800MHz, and 1000MHz. (the CPU also scales) Desktop compositing on the GPU A command is needed to set monitor resolution. Default is 1080p. Other Findings: CPU frequency was scanned under load: LITTLE cores (Cortex-A55): 1800 MHz BIG cores (Cortex-A76): 2208 MHz Those were the scaling_cur_freq readings. The vendor DTB caps the A76 at 2208 MHz: the physical max is 2.4 GHz but that OPP entry isn't in the table. Github repo: DOWNLOAD RELEASE FILES READ THE README.md! REFLASH STOCK IMAGE: UNPACKED FLASHING FILES FOR STOCK ANDROID README.md

Next week we'll be in Santa Fe, Argentina, for DebConf 2026! Catch our 2 talks covering the latest Apertis developments and toolkits in Debian. View the full article