IBV

This should mean that the rtc is there. Take a look at https://wiki.banana-pi.org/BPI_RTC_real_time_Module You could try: echo ds3231 0x68 > /sys/class/i2c-adapter/i2c-0/new_device Then a dmesg to see if there's something new about rtc.

Try to enable i2c1 also (just a stupid suggestion) in armbianEnv.txt and scan the bus with i2cdetect. Who knows, maybe there's a weird mapping of the bus interfaces.

I believe the rtc_ds1307 is compatible with the ds3231. You can check that with "modinfo rtc_ds1307" I don't know the hardware, can you check that you don't have pin conflicts? Do you need these overlays : "pps-gpio w1-gpio" ? If not, you might remove them from armbianEnv.txt, boot and check again.

There seems to be a RTC detected on boot [ 1.018029] sun6i-rtc 1f00000.rtc: registered as rtc0 [ 1.018081] sun6i-rtc 1f00000.rtc: setting system clock to 1970-01-01T00:00:04 UTC (4) Just to confirm, you could disconnect your RTC and boot again. Then check the boot log if the system says anything about rtc.

After boot, can you show the output of "timedatectl status" ?

This would suggest that the battery of the rtc is dead, but it is confusing because you say it works with RPI zero. Could you provide logs with sudo armbianmonitor -u

Hi, is the hardware clock set before power off ? What does "sudo hwclock" say ? Why did you modify the hwclock-set script ?

Hi, take a look at this similar issue: https://forums.raspberrypi.com/viewtopic.php?t=306463 Accroding to the RFC, the behaviour you see is correct, when Classless-Static-Route (121) is offered, the router option (3) must be ignored.

Hi, how, which software do you use? - did you try to use "alsamixer" to check the volumes ? Just to rule out the mic being muted - please provide logs with: https://docs.armbian.com/User-Guide_Fine-Tuning/#how-to-provide-boot-logs-for-inspection

Indeed, the patch was for 32bit chips, I did not look too careful. But it touches the clk-pll.c which is common. Not sure if it is worth to try to apply it to the rockchip64 series. I had a similar issue with tinker board (rk3288) where the mainline kernel did not work with some HDMI panels but the vendor kernel did. In both cases I had EDID warnings. The problems was caused by clock issues.

Hi, as per this forum thread: and this PR: https://github.com/armbian/build/pull/7479 The patch clk-rockchip-max-frac-divider.patch was deactivated from patch/kernel/archive/rockchip-X.X/series.conf. The reason is that it broke the HDMI display for some TFT panels. However the patch broke things on Asus Tinkerboard (RK3288), but might work well for RK3566. I suggest you re-activate the patch in series.conf, rebuild your current or edge image and test. Maybe it works for you. When building, check the build log and confirm the above patch is listed and applied. Cheers

Link has an . at the end, the correct one: https://paste.armbian.com/febewikocu

Separate /var partition was never an issue for me with upgrades (I am talking about Debian now, but Armbian is Debian based). The ram log setting also should stay as it is if the package handling it is properly done. You can however test if you want to be sure: install an older release, do the configs and then upgrade to the latest release.

It would be interesting to see the error.. Also, without being knowledgeable with Ansible, to me it looks like a command timeout (5 minutes?). https://docs.ansible.com/ansible/latest/collections/ansible/builtin/shell_module.html

First method: even if you manage to mount the partition, you will be able to edit the /etc/fstab file, but moving the data in /var partition will be more difficult. You would need also the NVME disk available to the Ubuntu machine, then you will have to mount the new NVME partition and move the data. Maybe is simpler to boot the SD card and follow the steps I described as the 2 method.

OK, so it did not auto mount. Can you try to mount it manually ? sudo mount /dev/sdb1 /mnt If it works, you should see the contents in /mnt

Are you sure you properly prepared the SD card? https://docs.armbian.com/User-Guide_Getting-Started/#preparing-sd-card I just did a test write of the Orange Pi 5 max image, using dd on a usb stick. Granted, I used the Debian Bookworm minimal image. https://dl.armbian.com/orangepi5-max/Bookworm_current_minimal When I inserted the usb stick in my Ubuntu PC, it was automatically mounted. Can you post the output of "dmesg" command after you insert the SD card?

Yes you could. I gave the UUID example out of habit, I do this on VM's where the disks are /dev/sdX and I want to make sure I get exactly the partition I want on boot in case the kernel/udev is messing the device ordering. In your case /dev/nmve0n1p1 should be fine for example. You are right, you will need to move anyway the old /var contents to the new partition. In case you are able to mount the SD card on a Linux PC/VM, you could do that too before first boot. But maybe better to follow the procedure I describe later. Probably the easiest would be to use a Linux VM (eg. VirtualBox). You can then use a usb SD card reader and pass it through to the VM. My distribution does it (Ubuntu). Yes. After flash the Armbian image, boot it. Allow it to reboot and expand the filesystem. Then: - disable ram log (by editing the /etc/default/armbian-ramlog and setting ENABLED=false) then reboot - make sure you have the root password (you can set one with "sudo passwd" if you don't have it) - reboot in rescue mode (use the root password when prompted) sudo init 1 - mount the NVME partition you want to use as /var in /mnt. FYI: This step is to move the existing /var data to the new location. You need rescue mode so no services run and write stuff in /var when you do the move. Examples below assume that the partition you want to use is /dev/nvme0n1p1. mount /dev/nvme0n1p1 /mnt - move the data mv /var/* /mnt/ - edit /etc/fstab and add a line with the /var mount. /dev/nvme0n1p1 /var ext4 defaults 0 2 If you want to use UUID's you can run "blkid" command and you will get the UUID's of all disks/partitions. You can then take the one of the nvme0n1p1 and use it with UUID=<your uuid> like I showed you in the first answer. - Reboot

Hi, theoretically if you knew the UUID of the partition on the NVME disk that you want to use as /var partition, you could do the following: - after flashing the SD card with the Armbian image, put it in a reader (or a usb card reader) connected to a Linux PC. - the / partition on the SD card should be auto mounted - you can go and then edit the etc/fstab file on the SD card and add a line like this: UUID=<your UUID goes here> /var ext4 defaults 0 2 put the UUID of the partition you want to use as /var in the above line. - boot your SD card I am not sure how this will work with the ram logging, you should try it yourself. If it does not work I can show you how to move the /var partition on the NVME disk after Armbian boots the first time (I have done this myself and it works). This involves - ram log disabling - reboot in rescue mode - identify the UUID of your NVME partition for new /var - move the contents of the /var/ to the new /var - modify the /etc/fstab to mount the new /var on boot.

For the dts, I would take the patch/kernel/archive/rockchip64-6.12/dt/rk3399-orangepi-4-lts.dts , modify it according to your needs and put it in the user patches directory (using the recommended way to replace a patch).

Indeed you are trying to patch a file that is created by an armbian patch. I suppose you can overwrite the armbian patch that creates the rk3399-orangepi-4-lts.dts with a patch that creates the dts with your modifications.

Hi, your patch fails because the file rk3399-orangepi-4-lts.dts does not exist in the mainline kernel(s) you tried to build. Maybe previously you appplied it against the vendor kernel?

To provide logs: https://docs.armbian.com/User-Guide_Fine-Tuning/#how-to-provide-boot-logs-for-inspection There's also "armbianmonitor -h" which shows the possible options.

Hi, take a look at this thread: The user is using picocom to check the traffic on ttyS2: picocom /dev/ttyS2 --baud 115200 --echo

Hi, to me it looks like there's no difference in the overlay you took from the sunnxi-DT-overlays repo and the one in the Armbian image you're running which comes from this patch: https://github.com/armbian/build/blob/fad71e196230690a2c8205ec977dcc5a653f9746/patch/kernel/archive/sunxi-6.12/patches.armbian/arm-dts-overlay-Add-Overlays-for-sunxi.patch#L4136 Normally according to the doc, i2c-0 is the bus connected to pins 3 an 5 (PA11, PA12), isn't it ? https://wiki.banana-pi.org/Banana_Pi_BPI-M2_ZERO So the behaviour looks correct when you enable the built-in i2c0 overlay. I cannot say why the numbering changes when you use your own overlay. Eventually you could follow the below link and provide 1 log after you boot the build-in overlay and 1 log after you boot with the custom one.