Werner

There are multiple lines starting with overlay= Put all in one line like overlays=uart1 uart2 uart3 uart4 i2s i2c whatever Make sure the enabled overlays don't share same GPIO pins. Check your boards datasheet/manual.

I noticed it is applied somewhere between the existing patches. Have you tried to rename it to something like zzzz_yourpatchfile.patch so, since they are ordered alphabetically, it is applied at the very end, just like your chances have been made with all the previous patches already applied.

Maybe this is why:https://patchwork.kernel.org/project/linux-rockchip/patch/20241220163227.1501912-1-alicja.michalska@9elements.com/

moved If it is a 1:1 clone of opi pc check if i2s overlay is enabled. Try armbian-config or the manual way by setting in /boot/armbianEnv.txt Check /boot subfolders for available overlays.

There is no package simply called qemu: https://packages.debian.org/search?keywords=qemu&searchon=names&suite=bookworm&section=all

Debian has some weird behaviour when using "su" because $PATH lacks /sbin for whatever reason. /sbin/chroot instead of chroot might work

If somebody adds support for it, why not? https://www.cnx-software.com/2024/12/21/rockchip-rk3588-mainline-linux-support-current-status-and-future-work-for-2025/ tl;dr: even if rk abandons it seems like collabora will continue.

Isn't the hdmirx on rk3588 able to do 4k60? https://www.rock-chips.com/uploads/pdf/2022.8.26/192/RK3588 Brief Datasheet.pdf

Should be rk3588-hdmirx.dtbo

moved

hdmirx is disabled by default since the code is poorly implemented and causes load for no reason. If you still want to play with this enable it via overlay.

Best chances is getting yourself familiar with the build framework and try to port the necessary patch to 6.12.y What I'd do is probably using kernel-patch command along with ./compile.sh (like "./compile.sh BOARD=udoo BRANCH=current kernel-patch") and then manually adding the patch contents. Then copy over the newly created patch into patch/kernel/archive/imx6-6.12 and try to compile. If it works try booting and see if USB are back again. PR is welcome Did a look myself and it seems like it's not that simple. the dts looks good already. But the drivers/usb/misc/onboard_usb_hub.c file is gone entirely. Therefore I guess CONFIG_USB_ONBOARD_HUB=y doesn't make sense in the kernel configuration either. Ahaa, there we go. https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/drivers/usb/misc?h=v6.12.13&id=31e7f6c015d9eb35e77ae9868801c53ab0ff19ac So things need adaption to the new name

That's what you get with bleeding edge. Use vendor if you need stuff working. Here is great overview of what is working in mainline: https://gitlab.collabora.com/hardware-enablement/rockchip-3588/notes-for-rockchip-3588/-/blob/main/mainline-status.md

I guess the patch has not been carried over to 6.12 since the folder is empty: https://github.com/armbian/build/tree/main/patch/kernel/archive/imx6-6.12

Yes. Tools like etcher or USBimager can decompress images on-the-fly. dd cannot do that. Latter since the image may already contain partitions, depending on board. Not really. If it is big enough to fit the image it is good enough. However microSD cards with bigger size are usually a lot faster as well. I personally use and can recommend sandisk extreme (the golden ones) with either 32G or 64G capacity. Too large is also possible since the SoCs may have a limit. So if your SD card is 2TB or something...try something a little smaller it never occurred to me that I had to install any kind of dependency to use it. Strange. Maybe there is another option available: https://gitlab.com/bztsrc/usbimager

known issue. its being worked on. Retry in a few days.

Of course, dd works too. The major issue with using this though is the lack of verification. While rare either dd or the target device can cause issues which are not detected while writing but when trying to read from the written device. USBimager and others like etcher verify all written data to make sure there is no issue. So if you run into weird hard to explain troubles, this could be a reason. arm64 download: https://gitlab.com/bztsrc/usbimager/-/blob/binaries/usbimager_1.0.10-aarch64-linux-x11.zip

I think this is part of the ramlog service. To reduce wear on microsd cards stuff is logged into memory and periodically written to sd. Since you are using a nvme this can be disabled. Check /etc/default/armbian-... files.

Wireguard is shipped by default with almost all Armbian images. Simply do code { font-family: Consolas,"courier new"; color: crimson; background-color: rgba(0, 0, 0, 0.2); padding: 2px; font-size: 105%; } sudo modprobe wireguard All you need is wireguard-tools which can be installed with --no-install-recommends to avoid unnecessary bloat/dependencies.

Ah. I guess nobody ported the overlays from old bsp over yet. Well also nobody was asking for them yet either lol Anyway this may work with "modern" bsp too: /dts-v1/; /plugin/; / { fragment@0 { target = <&spi3>; __overlay__ { status = "okay"; spidev@0 { compatible = "rockchip,spidev"; status = "okay"; reg = <0>; spi-max-frequency = <50000000>; }; }; }; };

How do you connect a NVMe to this board? Never mind. Wrong prefix. Anyway you need boot loader present on either spi, emmc or sd. Direct boot from nvme isnt possible.

No clue, sorry. I never actually used i2c

Hm vendor seem lacking all rk356x overlays except rk3566-roc-pc-sata2.dtbo. Need to check current. Edit: current and edge only have a few more but also none spi related for rk3566. Odd...

rockchip bsp kernel has recently been bumped to 6.1.99. Maybe things run smoother using it. You can use the build framework to create a set of kernel package to install via dpkg. Or use armbian-config to switch to beta repo which should also have it.

Can you provide me with the output of code { font-family: Consolas,"courier new"; color: crimson; background-color: rgba(0, 0, 0, 0.2); padding: 2px; font-size: 105%; } ls /boot/dtb/rockchip/overlay ?