Seb_Lz

Thanks a lot @Juanjo ! Your dtb has indeed fixed eth0 interface on my Beelik GT1 (armbian bionic version 5.6.0-rc7-arm-64).

I have it installed on my Beelik GT1 following stricly those steps (and using qemuarm-64 for the superviser install): https://www.home-assistant.io/hassio/installation/#alternative-install-home-assistant-supervised-on-a-generic-linux-host Do you guys have a working ethernet connection with your GT1 (I tried khadas-vim2, q200, q201 and many other .dtb but I can only get wifi working, not eth0) ?

As I was still searching for a solution to get eth0 working on beelink gt1, I've come across this: https://github.com/150balbes/Build-Armbian/blob/7fa5c7b89c7444a2d503f064296e0dda219cee20/patch/kernel/meson64-dev/0361-arm64-dts-meson-gxm-add-beelink-gt1.patch My understanding is that the dev branch includes a patch that will generate an additional .dtb file specifically for beelink gt1. Therefore, I tried tried to compile the image from source code. Although this works fine with the "current" branch, compilation fails with the dev branch. I naively tried to add the gt1 patch as a userpatch for the 'current' branch, but the integration of the patch would fail too during the creation process of the image. Here is the command I used to build dev image: ./compile.sh BOARD=arm-64 BRANCH=dev RELEASE=bionic BUILD_MINIMAL=no BUILD_DESKTOP=no KERNEL_ONLY=no KERNEL_CONFIGURE=no COMPRESS_OUTPUTIMAGE=gz and here is the error (full logs in attachment): Displaying message: Creating packages Displaying message: ERROR in function compile_kernel compilation.sh:431 err Displaying message: Failed moving kernel DEBs err Displaying message: Process terminated info Any idea why the dev branch won't build ? log.7z

Small update on my research to make the image work with Beelink GT1: After trying a few things, I eventually decided to start again from scratch with Armbian_20.05.1_Arm-64_bionic_current_5.6.0-rc7_20200404.img.xz. After installing, same problem on clean install. I then tried to go back to 20200331 version, which was previously working. To my surprise, and despite using the exact same image (clean format using balenaEtcher), it was also not working anymore ! I thought about an hardware problem, but rebooting on android emmc system prove me wrong as ethernet was there working fine. I then flashed back the SD to 20200404 and used Q201 dtb file. The eth0 interface is recognized but there are still errors in kern.log and the interface is still faulty (can't get an IP from DHCP nor connect to it when set statically): root@hassio:~# dmesg -T --level=err,warn [Wed Apr 8 17:02:39 2020] KASLR disabled due to lack of seed [Wed Apr 8 17:02:41 2020] OCFS2 User DLM kernel interface loaded [Wed Apr 8 17:02:42 2020] meson8b-dwmac c9410000.ethernet: IRQ eth_wake_irq not found [Wed Apr 8 17:02:42 2020] meson8b-dwmac c9410000.ethernet: IRQ eth_lpi not found [Wed Apr 8 17:02:42 2020] meson8b-dwmac c9410000.ethernet: PTP uses main clock [Wed Apr 8 17:02:42 2020] dwc3 c9000000.dwc3: Failed to get clk 'ref': -2 [Wed Apr 8 17:02:43 2020] phy phy-d0078080.phy.3: unsupported PHY mode 5 [Wed Apr 8 17:02:43 2020] hub 2-0:1.0: config failed, hub doesn't have any ports! (err -19) [Wed Apr 8 17:02:43 2020] mmc2: queuing unknown CIS tuple 0x80 (2 bytes) [Wed Apr 8 17:02:43 2020] mmc2: queuing unknown CIS tuple 0x80 (3 bytes) [Wed Apr 8 17:02:43 2020] mmc2: queuing unknown CIS tuple 0x80 (3 bytes) [Wed Apr 8 17:02:43 2020] mmc2: queuing unknown CIS tuple 0x80 (7 bytes) [Wed Apr 8 17:02:47 2020] gx-sound-card sound: ASoC: no DMI vendor name! [Wed Apr 8 17:02:47 2020] panfrost d00c0000.gpu: d00c0000.gpu supply mali not found, using dummy regulator [Wed Apr 8 17:02:47 2020] panfrost d00c0000.gpu: GPU Fault 0x00000088 (UNKNOWN) at 0x000000fcffbfffc0 [Wed Apr 8 17:02:48 2020] meson_vdec: module is from the staging directory, the quality is unknown, you have been warned. [Wed Apr 8 17:02:48 2020] hci_uart_bcm serial0-0: serial0-0 supply vbat not found, using dummy regulator [Wed Apr 8 17:02:48 2020] hci_uart_bcm serial0-0: serial0-0 supply vddio not found, using dummy regulator [Wed Apr 8 17:02:48 2020] brcmfmac: brcmf_fw_alloc_request: using brcm/brcmfmac43455-sdio for chip BCM4345/6 [Wed Apr 8 17:02:48 2020] brcmfmac mmc2:0001:1: Direct firmware load for brcm/brcmfmac43455-sdio.amlogic,q201.txt failed with error -2 [Wed Apr 8 17:02:48 2020] bluetooth hci1: Direct firmware load for brcm/BCM4345C0.hcd failed with error -2 [Wed Apr 8 17:02:49 2020] meson8b-dwmac c9410000.ethernet eth0: PTP not supported by HW [Wed Apr 8 17:02:49 2020] ext4 filesystem being mounted at /var/log supports timestamps until 2038 (0x7fffffff) [Wed Apr 8 17:02:50 2020] brcmfmac: brcmf_fw_alloc_request: using brcm/brcmfmac43455-sdio for chip BCM4345/6 [Wed Apr 8 17:02:50 2020] brcmfmac: brcmf_c_preinit_dcmds: Firmware: BCM4345/6 wl0: Feb 27 2018 03:15:32 version 7.45.154 (r684107 CY) FWID 01-4fbe0b04 [Wed Apr 8 17:03:09 2020] process 'docker/tmp/qemu-check033460557/check' started with executable stack [Wed Apr 8 17:03:12 2020] overlayfs: NFS export requires "index=on", falling back to nfs_export=off. I've extracted the dtb file from the working Android system and from the 20200404 img and converted them to dts file. I tried to adapt some values in the ethernet@c9410000 {...} part to mimic what is in the android part, but this didn't help (in some cases, it created more errors). I'm really confused how as how to make progress on this, any kind of help would be very welcome (dts files in attachments). android-q201.dts meson-gxm-q201.dts

I have a similar issue, also with beelink GT1 2Gb ram (S912): with Armbian_20.05.0_Arm-64_bionic_current_5.6.0-rc6_20200331.img, eth0 and wlan0 eventually started to work using dtb q200 (was NOT the case with previous RC6 img, was previously running an old 4.20 kernel img) BUT today I had a small power failure and on reboot, I started to loose eth0 support again, with following error in kern.log: Apr 7 14:32:56 localhost kernel: [ 11.076849] meson8b-dwmac c9410000.ethernet eth0: no phy at addr -1 Apr 7 14:32:56 localhost kernel: [ 11.076859] meson8b-dwmac c9410000.ethernet eth0: stmmac_open: Cannot attach to PHY (error: -19) NB: directly after installing rc6_20200331, I did an apt update + apt upgrade and all went fine (I did not upgrade the kernel), so in my case the problem with eth0 does not seem to be related to upgrading some packages. Since this error started to appear just after a power failure, could this be linked to some corrupted file ? Any idea how I could restore them ?

Might also be worth mentionning that wifi is also not working (less of a problem in my case though). It seems so strange that it was working with an older kernel with vim2 dtb and not anymore with the new one. I'm wondering if I missed something, especially since this is quite a popular device and that I doubt being the only one trying to get the new armbian image working on it. (And I take this opportunity to thanks everyone, especially balbes150 ofc, involved in this project for the amaizing work done ).

Hello guys, I'm trying to get armbian 20.05 buster on a Beelink GT1 2GB, but unfortunately I cannot get ethernet to work. I had success with an older version (4.20.5-aml-s912 ubuntu armbian) with khadas-vim2 dtb, but the same dtb from 20.05 does not work (no eth0 interface). Using q201 dtb seems to make the eth0 interface available but it won't actually connect (dhcp failing, manual configuration also not working). Has anyone managed to get ethernet working on recent kernel with Beelink GT1 ?