Haha - yes, I did eventually `saveenv'... just being super-cautious, and not sure what was going to work, given all the Kernel Panics. Also, in case some other poor guy comes by later and looks for some examples... If i boot the stock kernel/OS from eMMC, it loads (most of the time) and runs fine. I've had a few kernel panics, but usually it loads. If i boot fresh to the eMMC, it also seems to do better. There are some statements around the web that indicate uBoot might be 'hiding' errors prior to the booti, which causes people (like me) to tear their hair out looking in the wrong place -- i.e., register data is already bad before control is even transferred to the kernel. I did pull down armbian build and build out the latest kernel - 6.18, to see if that might make a difference, but no luck. If anything it caused more issues. I WAS able to get all the way booted up into Armbian one time -- adding the pcie power config switches (I thought), and it was stable for a while -- until i tried to `apt update', at which point it locked up. Haven't been able to get all the way up again. My current guesses to look into are: 1. make sure i'm not, as you said, overwriting myself during booti 2. continue to look for kernel config / versions that have better luck with SATA + mvebu64/power config. I'm guessing since I can boot the pre-installed OS that my hardware is actually OK. If you could provide your relevant ext4load (or similar) that you're using for SATA load, i'd appreciate it. I do get these bad rx status issues, and the networking isn't really coming up on the stock. -- Base install: [ 11.352864] mvneta d0030000.ethernet eth0: bad rx status 0cc10000 (crc error), size=75 [ 13.352762] mvneta d0030000.ethernet eth0: bad rx status 0cc10000 (crc error), size=75 root@ccpe10c1a5:~# cat /proc/cmdline console=ttyMV0,115200 earlycon=ar3700_uart,0xd0012000 root=/dev/mmcblk0p1 rw rootwait net.ifnames=0 biosdevname=0 root@ccpe10c1a5:~# uname -a Linux ccpe10c1a5 4.19.62-00036-gbc6a6e31fe72 #1 SMP PREEMPT Tue Jun 16 15:39:07 CST 2020 aarch64 aarch64 aarch64 GNU/Linux root@ccpe10c1a5:~# free -m total used free shared buff/cache available Mem: 993 59 793 2 140 867 Swap: 0 0 0 root@ccpe10c1a5:/media/disk0# dmesg | grep -i panic root@ccpe10c1a5:/media/disk0# w 18:17:28 up 19 min, 1 user, load average: 0.05, 0.03, 0.04

Ambian-imager requires a glibc 3 versions newer that this bookworm amd64 has, won't run at all. restricting dd to bs=512 gets only 2something megabytes a second, glacially slow. The more usual bs=4096 is 50x faster, niether of which reports any write errors. And its booted and configured. time about a month off. I run ntpsec on this machine so I'll follow the /etc/chrony/README and assign a source address by number. but ip a shows only an active ipv6 address while ip r returns nothing. hostname is bananapi-m5 which is unk on the local net. IOW its booted but network is only ipv6 & no route of either flavor is assigned. Is this exactly how it was spun to be? I've brought the card back to my reader, and will edit the /etc/hostname to match its ID for 192,168,71,2 in the /etc/hosts file I'll install. And reboot it. but the card won't mount despite the fact that it boots. from this bash xfc4 terminal: root@coyote:/etc# mount -t ext4 /dev/sdc1 /mnt/sdc1 mount: /mnt/sdc1: wrong fs type, bad option, bad superblock on /dev/loop0, missing codepage or helper program, or other error. dmesg(1) may have more information after failed mount system call. dmesg however has no additional info. only this: [1618874.568139] sd 37:0:0:0: [sdc] 249737216 512-byte logical blocks: (128 GB/119 GiB) [1618874.568944] sdc: detected capacity change from 0 to 249737216 [1618874.569933] sdc: sdc1 [1619402.494452] loop0: detected capacity change from 0 to 12943360 [1619402.494871] EXT4-fs (loop0): VFS: Can't find ext4 filesystem [1619458.923461] loop0: detected capacity change from 0 to 12943360 [1619458.923910] EXT4-fs (loop0): VFS: Can't find ext4 filesystem as I made two attempts to mount it. it loads and boots normally when installed in the bpi-m5. So other than the extra server line I just put in /etc/chrony/sources.d/ which doesn't work because no ipv4 net is even available AND the nearest ipv6 enabled anything is a minimum of 125 miles from me, its about as useful as the teats on a bore hog. Neither is the thought of editing on that tiny screen, the 40 or so lines of /etc/hosts that make up my /etc/hosts file. Doubly frustrating is the fact that the mouse speed is measurable fraction of C speed and totally disappears when inside the xfce4 terminal borders. But installed in the bpi-m5 it boots normally. back in my reader: root@coyote:/etc# e2fsck /dev/sdc e2fsck 1.47.0 (5-Feb-2023) ext2fs_open2: Bad magic number in super-block e2fsck: Superblock invalid, trying backup blocks... e2fsck: Bad magic number in super-block while trying to open /dev/sdc The superblock could not be read or does not describe a valid ext2/ext3/ext4 filesystem. If the device is valid and it really contains an ext2/ext3/ext4 filesystem (and not swap or ufs or something else), then the superblock is corrupt, and you might try running e2fsck with an alternate superblock: e2fsck -b 8193 <device> or e2fsck -b 32768 <device> Found a dos partition table in /dev/sdc ???????????? wtf??????????? And I get the same result for /dev/sdc1 obviously something is screwed. So I unpackedca new 200mb writeable sandisk u-sd & wrote the new image to it. fdisk then says its a dos partition table. Can a "dos" table be used at all on a 128gb card? I've heard not but don't know where the failure is. ipa shows only ipv6, ip r shows nothing. The only thing that could be used for net config is armbian.config which I think is NetworkManager . but that's by now known not to work. So whats left????? IOW what the hell is being used to configure network NOW with a 6.18 kernel ???? And how do I use the new tools if any, to set it up for a network address on 192.168.71.2, an ipv4 address THAT WORKS FOR EVERYTHING??????? Or do I go back to the soon to expire Armbian_25.5.1_Bananapim5_noble_current_6.12.28_xfce_desktop.img which I am using 2 more times here in front of klipper driven 3d printers. And which if asked to apt update, now want around 430 new files to do the update but failing in the make a new initrd stage. Start by putting a gpt partition table in the xz with a respin. And stop assuming that the whole planet has ipv6 connectivity, central WV is not the only ipv4 island on this planet so blessed.

Hello! How to download nightly build with 6.18 kernel? I've got 404 error when trying to download via link.

ok. I have 5 cubieboard 1. Not possible memory card for every one.

Nobody knows, nobody tried. Give it a shot. There is only one way to find out.