c0rnelius

I'll see if I can get those release imgs updated to something more recent and inquire about the breakage.

Damn. That's not good. Sorry to hear that.

Hmm. It appears ur still on 6.6.16-current-meson64 which does not include the fix.

This has already been fixed upstream and backported to 6.6.y LTS. https://lore.kernel.org/lkml/20240322164525.2617508-1-christianshewitt@gmail.com/ My suggestion is to update the kernel. If you don't wanna update you may just be able to blacklist the driver, which was what I did until it was fixed. It comes down to what is required by Klipper in your case.

I suggest looking over the overlays available; https://github.com/BPI-SINOVOIP/pi-linux/tree/pi-6.1-sunxi/arch/arm64/boot/dts/allwinner/overlay They may or maynot need to be edited as BSP VS MAINLINE usually isn't identical.

Using overlays on the PI whilst on Armbian should be no different than doing so on RaspiOS. https://www.raspberrypi.com/documentation/computers/config_txt.html#overlay_prefix If you don't know how to use them I suggest you use google or check the foundation docs.

Try disabling it with an overlay. /dts-v1/; /plugin/; / { fragment@0 { target = <&mbus>; __overlay__ { status = "disabled"; }; }; }; If that doesn't work, set the wrong compat string. /dts-v1/; /plugin/; / { fragment@0 { target = <&mbus>; __overlay__ { compatible = "allwinner,sun8i-h3-mbus"; }; }; }; Use the `armbian-add-overlay` to compile and add the overlay. NOTE: The overlay(s) may not work because of how the current DTS is setup, which is why I suggested blacklisting the driver. The driver in question is: sun8i-a33-mbus https://lore.kernel.org/linux-arm-kernel/a48923b7-12f9-808e-1171-49b826bd5f1c@samsung.com/T/#ma00a3e07248dc7fb2d300b7c9c409f69ffa64c14

Try adding a delay and see if that corrects the DRAM detection. https://github.com/armbian/build/blob/main/patch/u-boot/v2024.01/board_bananapim4zero/006-sunxi-restore-modified-memory.patch#L149

Kali is in a lot of ways a polished SID with all of their own custom bits added. It can be debootstrap'd just like Debian, Devuan and Ubuntu. I did find some services didn't work on boot; 'ssh avahi bluetooth ntp'. But there could be some option they have I'm not aware of to allow those services to start out of the box. Creating a custom service to start those does the trick though. I would think adding this as an option could be doable, but the question then becomes who wants to handle the support required to maintain it?

Try adding it this way. extraargs=swiotlb=0

Add this to your cmdline. swiotlb=0 https://github.com/raspberrypi/linux/commit/3e0065d1cb5bd58e35e0c7722ffe6303dfe0364f

The 6.6.y kernel for reasons I have yet to figure out reads the MEM wrong. 6.7 and up is not affected by this. Nor is 6.1 to 6.5. I would have thought after so many REVS this would have been resolved, but I have yet to see it come up as a topic of discussion. This in my experience is also not PLATFORM specific. It happens on all my boards regardless of SoC and PLATFORM.

I suspect it has to do with the devfreq driver for the MBUS/DRAM controller. https://lore.kernel.org/linux-arm-kernel/a48923b7-12f9-808e-1171-49b826bd5f1c@samsung.com/T/#ma00a3e07248dc7fb2d300b7c9c409f69ffa64c14 In my personal experience this introduces instability into the a64/h5. You could try blacklisting the driver and see if it solves the issue. I personally remove the entry from the DTS and create an overlay to add it back if for some reason I need it. The main issue you reported `Kernel panics with headless boot` in my mind points to it being the driver.

My guess would be this patch; https://github.com/armbian/build/blob/main/patch/u-boot/v2022.10/meson64-boot-usb-nvme-scsi-first.patch But if its a spin up HDD, it could be PREBOOT and it isn't able to spin up the drive in time and halting the boot process.

and a DTS was created?

firefox-esr --kiosk --private-window http://armbian.com I think chromium is better for this, but I suppose it depends on what you are trying to load. If using Xfce4 you could use Settings > Session and Startup. I prefer using Openbox my self, as editing a file `.config/openbox/autostart` is more straight forward and makes it easier to load only the services you want. cat .config/openbox/autostart # Daemon xfsettingsd & # Display xset s off & xset -dpms & xset s noblank & unclutter & # Background feh --bg-fill /home/username/.background.png & # Clean rm -fdr ~/.cache/chromium & rm -fdr ~/.config/chromium & # Kiosk sleep 5s chromium --noerrdialogs --incognito --kiosk --force-device-scale-factor=1 --disk-cache-dir=/tmp/null --enable-features=WebUIDarkMode --force-dark-mode http://yourwebsite.whatever &

If its completely bricked you should be able to short the unit and force it into USB mode. From there you can use either Linux or Windows to re-flash the vendor firmware. I would google the box you have and see if someone else has already done this and if nothing found, do a general search for ideas of where to look. "tvbox short emmc pins" Example: Tin foil seems to work best, but a small flat head will also do the trick.

In my experience, its not just Armbian. I "think" I saw a commit at one point concerning a fix for this over at https://github.com/libre-computer-project. But I can't remember if it was u-boot, linux or boot fip related. I'm leaning towards fip.

Not ideal, but: sudo halt

This works for me. Edit it to fit the commands you want to be able to run. sudo tee /etc/sudoers.d/010_$USER-nopasswd <<EOF ${USER} ${HOSTNAME} = (root) NOPASSWD: /sbin/shutdown ${USER} ${HOSTNAME} = (root) NOPASSWD: /sbin/poweroff ${USER} ${HOSTNAME} = (root) NOPASSWD: /sbin/reboot EOF

Not sure about doing it for just one command, but you can make it so you don't need to authenticate when using sudo. sudo mkdir -p /etc/sudoers.d/ sudo tee /etc/sudoers.d/010_$USER-nopasswd <<EOF $USER ALL=(ALL) NOPASSWD: ALL EOF

I'm using overlay meson-g12a-radxa-zero-i2c-ee-m1-gpiox-10-gpiox-11.dtbo sudo i2cdetect -y 1 0 1 2 3 4 5 6 7 8 9 a b c d e f 00: -- -- -- -- -- -- -- -- 10: -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- 20: -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- 30: -- -- -- -- -- -- -- -- -- -- -- -- 3c -- -- -- 40: -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- 50: -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- 60: -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- 70: -- -- -- -- -- -- -- -- Overlay meson-g12a-radxa-zero-i2c-ee-m3-gpioa-14-gpioa-15.dtbo works for me as well.

Or temporarily patch the driver until a better solution presents its self. diff --git a/drivers/net/wireless/uwe5622/unisocwifi/tx_msg.c b/drivers/net/wireless/uwe5622/unisocwifi/tx_msg.c index 40d51a7130d9..f1c356306c80 100644 --- a/drivers/net/wireless/uwe5622/unisocwifi/tx_msg.c +++ b/drivers/net/wireless/uwe5622/unisocwifi/tx_msg.c @@ -317,7 +317,15 @@ void sprdwl_dequeue_data_list(struct mbuf_t *head, int num) /* seam for tx_thread */ void tx_down(struct sprdwl_tx_msg *tx_msg) { - wait_for_completion(&tx_msg->tx_completed); + int ret; + while (1) { + ret = wait_for_completion_interruptible(&tx_msg->tx_completed); + if (ret == -ERESTARTSYS) { + continue; + } + + return; + } } void tx_up(struct sprdwl_tx_msg *tx_msg) -- 2.39.2 https://github.com/orangepi-xunlong/linux-orangepi/issues/57

sudo nano /etc/asound.conf pcm.!default { type hw card 2 } ctl.!default { type hw card 2 } save and reboot https://www.alsa-project.org/wiki/Asoundrc

Fun read though.