2026-01-01 07:53:44 <ncopa> https://wiki.gentoo.org/wiki/Milk-V_Pioneer_Box#Configuring_the_Linux_kernel there is a: Warning
2026-01-01 07:53:44 <ncopa> You must build the latest bootloader (zsbl, OpenSBI, and EDK2) to use an upstream kernel.
2026-01-01 18:15:23 <clandmeter> ncopa: I would exepect the bsp from their own CI to be workable. 
2026-01-01 18:15:41 <clandmeter> i am not sure what goes wrong, but this has happend before when i tried EDK2
2026-01-01 18:15:54 <clandmeter> i read some refs regarding memory
2026-01-02 12:03:08 <mps> mesa should be upgraded to 25.3, i tested it on riscv64, aarch64, x86_64 and loongarch64
2026-01-02 14:29:07 <achill> mps: then please look after https://gitlab.freedesktop.org/mesa/mesa/-/issues/14463
2026-01-02 14:29:13 <achill> thats the issue blocking the merge of mesa 25.3
2026-01-02 14:37:35 <mps> uhm, 32bit arm, didn't tried on it
2026-01-06 07:28:54 <mps> is pioneer with dev lxcs down again?
2026-01-06 07:31:13 <ikke> nld-bld-1, yes
2026-01-06 07:35:21 <mps> can someone boot it?
2026-01-06 07:41:44 <ikke> clandmeter needs to restart it 
2026-01-06 08:13:20 <clandmeter> should restart
2026-01-06 08:20:58 <mps> must restart ;)
2026-01-06 09:18:43 <mps> looks like host works now but my lxc isn't accessible
2026-01-06 09:21:14 <ikke> I suspect you can now
2026-01-06 09:22:53 <mps> i can, thanks
2026-01-06 09:25:59 <mps> will try to build and enable bpi-f3 in mainline u-boot
2026-01-06 09:32:33 <ikke> clandmeter: funny thing, now it's the other host that's unavailabl
2026-01-06 10:16:18 <ncopa> clandmeter: re failing to boot pioneer using edk2 https://github.com/sophgo/bootloader-riscv/issues/230 I think the problem is in grub
2026-01-06 11:30:30 <ncopa> we can solve it by either try fix grub, or create an UKI and boot the kernel/initramfs as an EFI executable
2026-01-06 11:30:51 <ncopa> former is better longter, latter is easier (does not require rebuild grub)
2026-01-06 11:31:42 <ncopa> I played with ukify a few days ago. https://gist.github.com/ncopa/00c6e791e3e56c6e0369e5a2aa2eb99f#create-the-image
2026-01-06 11:32:06 <achill> ohh that is nice
2026-01-06 11:32:17 <ncopa> there is even a trigger so the UKI can be auto generated on kernel updates
2026-01-06 11:32:56 <ncopa> I was thinking of maybe even ship a distro image as UKI
2026-01-06 11:34:37 <ncopa> ie. put the modloop and apks repo in initramfs and ship it as alpine-virt.efi
2026-01-06 12:05:11 <mps> nld-bld-1 is practically unusable for building anything bigger for me
2026-01-06 12:07:03 <clandmeter> ncopa: what about trying limine? 
2026-01-06 12:07:10 <clandmeter> or  what was the real name
2026-01-06 12:07:23 <clandmeter> its weird, didnt expect it was grub
2026-01-06 12:07:40 <clandmeter> saw no msg about it exec the bootloader
2026-01-06 12:09:43 <mps> I use limine an apple silicon and loongarch, didn't tried on riscv64
2026-01-06 12:10:25 <mps> and use it aarch64 qemu
2026-01-06 12:10:57 <mps> but it can't load compressed vmlinuz
2026-01-06 12:22:27 <ikke> unusable in what way?
2026-01-06 12:25:59 <mps> ikke: too often hangs build processes or fails in different ways
2026-01-06 12:42:56 <clandmeter> seems bld2 is hanging again
2026-01-06 12:43:05 <ikke> yes, I mentioned it
2026-01-06 12:43:12 <clandmeter> ill reboto it later and see if i can try limine to boot
2026-01-06 12:43:20 <clandmeter> or newer grub
2026-01-06 12:43:27 <clandmeter> not even sure which grub version it is
2026-01-06 12:43:50 <clandmeter> anyone tried our grub on rv64?
2026-01-06 12:53:54 <mps> clandmeter: I tried on VF2 1-2 years ago. it worked then
2026-01-06 12:54:23 <mps> and on qemu
2026-01-06 14:45:01 <clandmeter> rebooting bld1
2026-01-06 14:47:02 <clandmeter> \o/
2026-01-06 14:47:08 <clandmeter> ncopa: you were right, again of course
2026-01-06 14:47:16 <clandmeter> how did you figure that out? 
2026-01-06 14:47:23 <clandmeter> just by that error msg? 
2026-01-06 14:48:38 <clandmeter> Limine 10.6.1 (riscv64, UEFI)
2026-01-06 15:10:54 <ncopa> clandmeter: I pasted the entire log to chatgpt which claimed it to be grub. I asked how it knew it was bug and then I asked for references
2026-01-06 15:11:09 <clandmeter> lol
2026-01-06 15:11:28 <clandmeter> but not all of it is grub
2026-01-06 15:11:35 <clandmeter> only the last part i think
2026-01-06 15:11:59 <ncopa> yes. that is what chatgpt explained too
2026-01-06 15:13:59 <ncopa> so it boots with Limine now?
2026-01-06 15:15:07 <clandmeter> it loads the bootloader yes
2026-01-06 15:15:10 <clandmeter> need to config it now
2026-01-06 15:15:15 <clandmeter> see if it wants to load the kernel
2026-01-06 15:15:18 <ncopa> oh nice!
2026-01-06 15:15:44 <ncopa> the linux-lts kernel from edge should in theory work
2026-01-06 15:19:40 <ncopa> reminds me, I should followup on that limine thing for setup-disk
2026-01-06 15:20:07 <clandmeter> ok here she goes
2026-01-06 15:20:22 <clandmeter> lets see if it eats linux for dinner
2026-01-06 15:20:55 <ncopa> I joined the serial console. I hope i didnt kick you out
2026-01-06 15:21:15 <ncopa> i think i did... sorry
2026-01-06 15:22:17 <clandmeter> nope
2026-01-06 15:22:26 <clandmeter> weird, it cannot find the config
2026-01-06 15:25:50 <ncopa> it is on nvme? maybe it does not find nvme?
2026-01-06 15:26:07 <ncopa> limine probably needs the driver for the storage media
2026-01-06 15:30:13 <clandmeter> not sure, it does not mention anything that it needs something special for nvme
2026-01-06 15:58:46 <ncopa> but it found the kernel?
2026-01-06 15:59:13 <clandmeter> it seems it cannot find the disk
2026-01-06 15:59:19 <clandmeter> or im holding it wrong
2026-01-06 15:59:36 <clandmeter> not sure if i need to install anything else except the efi pkg
2026-01-06 16:00:02 <clandmeter> i put the config in esp
2026-01-06 16:00:16 <clandmeter> so boot() should be set to the esp
2026-01-06 16:00:24 <clandmeter> and boot(2) should be root drive
2026-01-06 16:00:33 <clandmeter> boot(1) is esp
2026-01-06 16:00:59 <clandmeter> also, if i dont put the config on the esp, it wont find it. 
2026-01-06 16:01:17 <clandmeter> but it does find the config if its next to it. 
2026-01-06 16:15:33 <mps> clandmeter: I put all on efi partitions for limine
2026-01-06 16:15:52 <clandmeter> yes i think thats the only way
2026-01-06 16:16:07 <mps> think it don't have drivers for other FS
2026-01-06 16:16:30 <clandmeter> oh hah, thats the problem yes
2026-01-06 16:16:32 <clandmeter> now i remember
2026-01-06 16:17:06 <clandmeter> thx for reminding me
2026-01-06 16:17:14 <clandmeter> i put boot on esp
2026-01-06 16:17:48 <mps> and i found it can boot only uncompressed or efi kernel
2026-01-06 16:19:10 <mps> type of kernel must be 'told' in limine.conf stanza
2026-01-06 16:20:02 <mps> initramfs can be compressed
2026-01-06 16:21:03 <mps> in 15 minutes i can post limine.conf for asahi alpine
2026-01-06 16:26:35 <mps> here is script to create alpine image for apple silicon installer https://tpaste.us/xDoQ
2026-01-06 16:27:08 <mps> there is part for bootloader - limine
2026-01-06 16:45:53 <clandmeter> ok, you are right, it does not take compressed kernel
2026-01-06 16:46:09 <clandmeter> is our kernel just gzipped? 
2026-01-06 16:47:36 <mps> i think so. once i simply gunziped kernel and it booted
2026-01-06 16:48:58 <clandmeter> 🤞
2026-01-06 16:50:03 <mps> you can check by `file` command
2026-01-06 16:51:16 <clandmeter> i did :)
2026-01-06 16:51:25 <clandmeter> ok it loaded the kernel and initramfs
2026-01-06 16:51:29 <clandmeter> but lost screen
2026-01-06 16:51:37 <clandmeter> linux: Loading kernel `boot():/vmlinuz-lts`...
2026-01-06 16:51:37 <clandmeter> linux: Loading module `boot():/initramfs-lts`...
2026-01-06 16:51:37 <clandmeter> PROGRESS CODE: V02010004 I0
2026-01-06 16:51:37 <clandmeter> PROGRESS CODE: V03101019 I0
2026-01-06 16:51:50 <mps> uh
2026-01-06 16:52:48 <mps> maybe it need few seconds
2026-01-06 16:53:11 <mps> 10-15 iirc in such cases
2026-01-06 16:53:26 <mps> seconds*
2026-01-06 16:54:03 <ncopa> do you have serial console?
2026-01-06 16:55:12 <mps> i can ping 172.16.0.30 from dmvpn
2026-01-06 16:55:49 <mps> nld-bld-1 is this?
2026-01-06 16:58:37 <clandmeter> i will update the cmdline, forgot about it
2026-01-06 17:00:38 <clandmeter> ncopa: i remember i did a in kernel console config
2026-01-06 17:00:44 <clandmeter> i guess you didnt do that for yours
2026-01-06 17:02:39 <mps> limine supports serial console
2026-01-06 17:03:14 <mps> and serial_baudrate
2026-01-06 17:04:02 <mps> also verbose could be useful
2026-01-06 17:04:04 <ikke> nld-bld-1 is 172.16.24.2
2026-01-06 17:04:11 <ikke> sorry, 30.20
2026-01-06 17:04:13 <ikke> sorry, 30.2
2026-01-06 17:04:26 <ikke> 172.16.0.30 is the router in front of them
2026-01-06 17:04:40 <mps> ah, thanks
2026-01-06 17:05:31 <mps> 172.16.30.2 also answer ping
2026-01-06 17:05:38 <clandmeter> its online now
2026-01-06 17:05:41 <clandmeter> but will reboot soon
2026-01-06 17:05:47 <clandmeter> mps: what serial support?
2026-01-06 17:05:55 <clandmeter> i mean i can see limine over serial just fine
2026-01-06 17:06:06 <clandmeter> but i need to see the kernel 
2026-01-06 17:06:11 <mps> this, i meant
2026-01-06 17:06:32 <mps> then do, please :)
2026-01-06 17:06:43 <clandmeter> i mean does it have any special serial cofnig?
2026-01-06 17:07:20 <mps> yes, these: serial and serial_baudrate
2026-01-06 17:07:29 <clandmeter> i dont see it in tthe docs
2026-01-06 17:07:43 <mps> serial=yes
2026-01-06 17:07:54 <clandmeter> in limine.conf?
2026-01-06 17:08:02 <mps> serial_baudrate default to 9600
2026-01-06 17:08:18 <mps> yes, limine.conf
2026-01-06 17:08:25 <clandmeter> show me where you find this? 
2026-01-06 17:08:49 <clandmeter> ah found it
2026-01-06 17:09:39 <mps> > clandmeter| show me where you find this?
2026-01-06 17:10:01 <mps> in local git cloned
2026-01-06 17:10:22 <clandmeter> baudrate is only for bios boot
2026-01-06 17:11:00 <mps> https://codeberg.org/Limine/Limine/src/branch/v10.x/CONFIG.md
2026-01-06 17:11:45 <clandmeter> ok last try
2026-01-06 17:11:49 <clandmeter> need to go home
2026-01-06 17:13:18 <clandmeter> nope serial stops
2026-01-06 17:13:28 <mps> huh
2026-01-06 17:13:29 <clandmeter> ncopa: im not sure what happends
2026-01-06 17:13:36 <clandmeter> linux: Loading kernel `boot():/vmlinuz-lts`...
2026-01-06 17:13:36 <clandmeter> PROGRESS CODE: V03101019 I0
2026-01-06 17:13:36 <clandmeter> PROGRESS CODE: V02010004 I0
2026-01-06 17:13:36 <clandmeter> linux: Loading module `boot():/initramfs-lts`...
2026-01-06 17:13:41 <clandmeter> it stops after this
2026-01-06 17:13:58 <clandmeter> this is your kernel
2026-01-06 17:14:19 <clandmeter> i remember when i build the kernel for pioneer I had to hard code the serial in the kernel
2026-01-06 17:14:39 <clandmeter> as else something breaks it, and im not sure what. 
2026-01-06 17:14:55 <clandmeter> and the GPU only runs when the modules are loaded
2026-01-06 17:16:24 <mps> clandmeter: try to put `cmdline` params, these are kernel cmdline
2026-01-06 17:17:21 <mps> f.e. cmdline: root=UUID=$rootuuid ro rootwait console=tty1 modules=sd-mod,usb-storage
2026-01-06 17:19:27 <clandmeter> ikke: im bringign it back online
2026-01-06 17:19:30 <clandmeter> stoppign now
2026-01-06 17:19:42 <clandmeter> hope it boots
2026-01-06 17:19:55 <clandmeter> it does
2026-01-06 17:20:23 <ikke> ok, all containers have an IP now
2026-01-06 17:20:44 <clandmeter> ncopa: maybe we can check tomrorow morning about the issue
2026-01-06 17:20:54 <clandmeter> something is up with your kernel
2026-01-06 17:26:12 <mps> maybe could test limine on armv7 chromebook, but have to find time for this
2026-01-06 17:28:14 <mps> (while impatiently waiting for spacemit musebook)
2026-01-07 06:40:05 <clandmeter> morning
2026-01-07 06:40:25 <clandmeter> ncopa: where is your kernel config? 
2026-01-07 07:09:58 <mps> i remember that with limine and linux-lts had to set 'protocol: efi' in limine.conf but then loading initramfs doesn't work
2026-01-07 07:11:08 <mps> because this i built kernels with needed FS drivers in kernel not as modules
2026-01-07 07:13:57 <mps> linux-lts for loongarch64 is efi or it called something like PE32 (forgot details) so in limine protocol must be efi and not linux
2026-01-07 07:14:38 <mps> at least in my test, would like to be proven wrong
2026-01-07 07:15:11 <mps> maybe i simply didn't understand something
2026-01-07 07:31:18 <clandmeter> i think our issue is not the bootloader, but we can try grub later. 
2026-01-07 07:35:23 <clandmeter> ncopa: maybe you can try to enable CONFIG_CMDLINE="console=ttyS0,115200"
2026-01-07 07:55:46 <ncopa> you find the config at /boot/config-6.18.1-1-lts 
2026-01-07 07:59:43 <mps> clandmeter: I just tell my issues with limine and kernel. ofc i don't know what is problem in pioneer and kernel 6.18 case
2026-01-07 08:35:21 <clandmeter> ncopa: can you build a new kernel with that option added? 
2026-01-07 08:35:33 <clandmeter> not sure which hw you build yours
2026-01-07 09:01:19 <ncopa> nld-bld-1 [~]# grep CMDLINE /boot/config-6.18.3-1-lts 
2026-01-07 09:01:19 <ncopa> CONFIG_CMDLINE="console=ttyS0,115200"
2026-01-07 09:01:22 <ncopa> kernel is installed
2026-01-07 09:01:44 <ncopa> as /boot/vmlinuz-lts
2026-01-07 09:02:27 <ncopa> can you run your serial console in tmux or something?
2026-01-07 09:07:20 <ncopa> the _CMDLINE_ options differs from sophgo though
2026-01-07 09:07:44 <ncopa> # grep CMDLINE /boot/config-6.1.80-0-sophgo  | tpaste
2026-01-07 09:07:44 <ncopa> https://tpaste.us/voQW
2026-01-07 09:07:57 <ncopa> # grep CMDLINE /boot/config-6.18.3-1-lts   | tpaste
2026-01-07 09:07:57 <ncopa> https://tpaste.us/QKy8
2026-01-07 09:08:35 <ncopa> sophgo kernel has CMDLINE_EXTEND, which meand that the CONFIG_CMDLINE="console=ttyS0,115200" is appended to the cmdline from bootloader
2026-01-07 09:09:08 <ncopa> while the lts kernel has CONFIG_CMDLINE_FALLBACK=y, which means that it will only use the CONFIG_CMDLINE="console=ttyS0,115200" if no cmdline is passed from boot loader
2026-01-07 09:12:04 <ncopa> oh the gentoo wiki about milk-v had lots of changes 30 dec
2026-01-07 09:12:16 <ncopa> it completely removed the vendored kernel'
2026-01-07 10:15:59 <ncopa> clandmeter: what serial console is connected? the USB-c on front?
2026-01-07 10:52:33 <clandmeter> yes front
2026-01-07 11:09:48 <clandmeter> ncopa: did you verify if all the config options in the gentoo wiki are set on ours?
2026-01-07 11:11:30 <clandmeter> there are some things incorrect on their wiki, or lets say missing. they are not aware of the BSP that sophgo is building on github ci. 
2026-01-07 11:12:21 <ncopa> yes, I have verified all config options from the gentoo wiki is there. chatgpt thinks it may be related the DTS, but im not so sure
2026-01-07 11:12:56 <ncopa> i dont think the CONFIG_CMDLINE= should make any difference, unless the bootloader fails to pass the boot options
2026-01-07 11:13:29 <ncopa> the current vmlinux-lts that is there now has the CONFIG_CMDLINE="console=ttyS0,115200"
2026-01-07 11:13:35 <ncopa> in case you want to test
2026-01-07 11:13:51 <ncopa> i am currently building a new kernel which enables the radeon driver
2026-01-07 11:14:07 <ncopa> but according the wiki we also need some non-upstreamed patches
2026-01-07 11:18:46 <clandmeter> CONFIG_DRM_RADEON
2026-01-07 11:18:49 <clandmeter> this is missing
2026-01-07 11:19:19 <clandmeter> CONFIG_FB_RADEON also
2026-01-07 11:20:27 <clandmeter> I lost bld1
2026-01-07 11:20:31 <ncopa> do you want test boot the kernel with CONFIG_CMDLINE="console=ttyS0,115200" before I replace the linux-lts?
2026-01-07 11:21:04 <clandmeter> is that an extended cmdline or it will replace? 
2026-01-07 11:21:30 <ncopa> as I understand it will only replace if cmdline is missing from bootloader
2026-01-07 11:22:26 <clandmeter> so if cmdline is set then it will append? 
2026-01-07 11:22:36 <clandmeter> or ignore?
2026-01-07 11:22:52 <ncopa> if cmdline is set from bootloader, it will ignore
2026-01-07 11:23:07 <clandmeter> CONFIG_RTC_DS1307 and CONFIG_RTC_DRV_EFI is also missing, not sure we need it. 
2026-01-07 11:24:41 <ncopa> according to wiki, probably not. I think its the RTC only, which means we will boot with the clock wrong
2026-01-07 11:26:06 <clandmeter> hmm you build the pci controller as module
2026-01-07 11:26:13 <clandmeter> is that wise ?
2026-01-07 11:27:04 <clandmeter> CONFIG_PCIE_SG2042_HOST=y
2026-01-07 11:27:04 <clandmeter> The PCIe controller on the board.
2026-01-07 11:27:45 <ncopa> it is a generic kernel. I assume that it is ok?
2026-01-07 11:28:32 <ncopa> if you give the current kernel a go, just to verify that the CONFIG_CMDLINE change
2026-01-07 11:28:43 <ncopa> then I can replace the kernel with one that has radeon enabled
2026-01-07 11:29:01 <clandmeter> i can check it, but we already set the cmdline
2026-01-07 11:29:06 <clandmeter> should i disable cmdline from bootloader? 
2026-01-07 11:29:10 <ncopa> nah
2026-01-07 11:29:20 <ncopa> the point is that the kernel gets the cmdline
2026-01-07 11:29:42 <ncopa> we coudl also try: earlycon=uart8250,mmio32,0x7040000000,115200 console=ttyS0,115200
2026-01-07 11:29:45 <clandmeter> CONFIG_SOPHGO_SG2042_PLL=m, CONFIG_CLK_SOPHGO_SG2042_CLKGEN=m, CONFIG_SOPHGO_SG2042_RPGATE=m  some are also not set
2026-01-07 11:29:57 <ncopa> oh?
2026-01-07 11:29:58 <ncopa> ok
2026-01-07 11:30:02 <clandmeter> ok lets just reboot it
2026-01-07 11:30:08 <ncopa> I will enable those as well
2026-01-07 11:30:18 <clandmeter> need to first change the sd
2026-01-07 11:30:58 <ncopa> do you have a tmux session so I also can join to watch the boot?
2026-01-07 11:35:06 <clandmeter> i tried
2026-01-07 11:35:13 <clandmeter> but screen in tmux doesnt work well
2026-01-07 11:35:17 <clandmeter> can you attach to screen?
2026-01-07 11:36:28 <clandmeter> ncopa: i enabled multiuser mode
2026-01-07 11:37:41 <ncopa> i think I am joined
2026-01-07 11:38:06 <clandmeter> its not that nice with 2ppl :)
2026-01-07 11:38:13 <clandmeter> ah its better now
2026-01-07 11:38:58 <clandmeter> but still same issue
2026-01-07 11:39:15 <ncopa> ok
2026-01-07 11:39:32 <ncopa> lets try kernel with radeon and maybe you get something on teh screen
2026-01-07 11:40:00 <ncopa> then we can try earlycon=uart8250,mmio32,0x7040000000,115200 console=ttyS0,115200
2026-01-07 11:40:58 <clandmeter> i just think if you do not build the pcie controller in the kernel, how would any device connected over pcie work on early boot?
2026-01-07 11:41:11 <clandmeter> or you need to add it to initramfs
2026-01-07 11:41:28 <clandmeter> the nvme is over pcie
2026-01-07 11:41:39 <ikke> tio is also an option to connect to serial devices, should work better in tmux
2026-01-07 11:42:18 <clandmeter> ok let me try that next
2026-01-07 11:42:43 <clandmeter> scrolling the readme, but looks nice
2026-01-07 11:55:11 <mps> ikke: thanks for mentioning tio, it looks better than screen for serial line works
2026-01-07 12:12:36 <ncopa> i dont think the serial console is over pcie. We do need it in the initramfs though
2026-01-07 12:13:49 <ncopa> i did read that the pioneer has MCU with UART. Maybe we can powercycle it over a second serial port?
2026-01-07 12:14:17 <ncopa> i use tio as well. it is nicer than the alternatives IMHO
2026-01-07 12:22:04 <clandmeter> yes, serial should not be over pcie, but disk is, so it wont be able to find root if not in initramfs.
2026-01-07 12:28:57 <ncopa> I would assume it is in the initramfs
2026-01-07 14:25:58 <clandmeter> why would you assume it? 
2026-01-07 14:26:09 <clandmeter> is it automatic?
2026-01-07 14:26:37 <clandmeter> ncopa: did you build a new one? 
2026-01-07 14:30:18 <mps> doesn't nvme feature pulls pci module in initramfs
2026-01-07 14:40:12 <clandmeter> i dont know, if so then its ok
2026-01-07 14:41:05 <mps> also idk but assume this
2026-01-07 14:45:41 <clandmeter> is it normal for tio to keep disconnecting? 
2026-01-07 14:46:26 <ikke> No
2026-01-07 14:47:39 <clandmeter> [14:45:22.028] Disconnected
2026-01-07 14:47:39 <clandmeter> [14:45:23.030] Connected to /dev/ttyUSB0
2026-01-07 14:47:39 <clandmeter> [14:45:25.631] Connected to /dev/ttyUSB0
2026-01-07 14:47:39 <clandmeter> [14:45:24.630] Disconnected
2026-01-07 14:48:51 <mps> this is nice feature. screen irritates me because disconnect on such cases
2026-01-07 14:49:30 <clandmeter> never had screen disconnecting
2026-01-07 14:49:37 <clandmeter> so i am not sure what this is 
2026-01-07 14:50:10 <mps> even when unplug usb ?
2026-01-07 14:52:11 <mps> on apple silicon usb driver is unstable and it too often disconnect interface
2026-01-07 16:16:18 <mps> clandmeter: success upgrade or ...?
2026-01-08 08:11:33 <clandmeter> mps: i think it was related to multiple screen sessions still active.
2026-01-08 08:11:46 <clandmeter> ncopa: i tried to boot the working kernel with the new bsp
2026-01-08 08:12:04 <clandmeter> it does boot and show itself, but crashes
2026-01-08 08:13:28 <clandmeter> looks like some memory alignment issue
2026-01-08 08:31:40 <mps> clandmeter: yes and no. screen exits or goes to background when serial interface disconnects or unplugged
2026-01-08 08:32:20 <mps> i have to kill screen and start it again in these cases
2026-01-08 09:54:33 <ncopa> the working kernel as 6.1.80-sophgo?
2026-01-08 11:09:39 <clandmeter> ncopa: yes
2026-01-08 11:09:46 <clandmeter> you can check tmux on bld2
2026-01-08 11:10:09 <clandmeter> I left it open for now, can reboot to live later.
2026-01-08 12:47:59 <ncopa> dont have time today sorry
2026-01-10 12:06:28 <mps> can't connect to my lxc on nld-bld-1, 172.16.30.101
2026-01-10 12:07:42 <ikke> Now you should
2026-01-10 12:08:29 <mps> yes, thank you
2026-01-10 12:09:04 <mps> netfilter issue or something else?
2026-01-10 12:10:17 <ikke> iptables, yes. Haven't had time to research it, but I need to explicitly allow forwarding traffic
2026-01-10 12:10:48 <mps> on host?
2026-01-10 12:23:14 <ikke> yes
2026-01-10 12:23:31 <ikke> Need to figure out what awall rule is missing
2026-01-10 12:24:23 <mps> hm, why not work 'net.ipv4.ip_forward = 1' in /etc/sysctl.d/01-local.conf
2026-01-10 12:24:48 <mps> iiuc this needs to be set
2026-01-10 12:24:50 <ikke> That's not the issue. The default policy that awall sets is block.
2026-01-10 12:24:53 <ikke> for FORWARD
2026-01-10 12:24:59 <mps> ah
2026-01-10 12:25:03 <ikke> So we need to make sure the correct traffic is allowed
2026-01-10 12:25:14 <ikke> right now I hotfix it by setting the policy to ACCEPT
2026-01-10 12:25:40 <mps> understand
2026-01-10 12:26:25 <mps> but I never used awall so have no idea what is solution
2026-01-11 16:32:16 <ech6f> hi, i have been trying to use the script here (https://arvanta.net/alpine/alpine-on-spacemit/) to generate in image that fully utalises my sd card, however when running te script i run into some errors. when the script reaches the "installing base packages" section apk throws errors, specifcally "ERROR: busybox-1.37.0-r31.trigger: script exited with error 127" and the same but with dbus-1.16.2-r1.trigger. after this it procedes no further.
2026-01-11 16:32:36 <ech6f> is there a way i can fix this? apologies if it is a stupid question i'm new to alpine
2026-01-11 16:51:19 <mps> ech6f: did you started qemu-binfmt, installed qemu-riscv64 and loaded loop kernel module
2026-01-11 16:53:48 <mps> i guess yes but ask just in case
2026-01-11 16:57:24 <mps> how to redirect stderr to file in chroot is question
2026-01-11 17:26:53 <ech6f> im fairly certain i have started qemu-binfmt, and i have installed qemu-riscv64, but i dont think i have loaded the loop kernel module
2026-01-11 17:28:39 <mps> also, what is /bin/sh? ash or something else
2026-01-11 17:29:35 <ech6f> yep, default 
2026-01-11 17:29:46 <mps> i.e. do you run it on alpine or some other distro
2026-01-11 17:29:58 <ech6f> im running it on an alpine install
2026-01-11 17:30:42 <mps> strange then, i run this script many times
2026-01-11 17:32:03 <mps> try with 'set -x' in script to see where it die
2026-01-11 17:33:08 <mps> also run with sudo or as root
2026-01-11 17:33:31 <ech6f> yep, running as root
2026-01-11 17:33:48 <ech6f> didnt throw an error when i changed to 'set -x'
2026-01-11 17:33:49 <mps> then all is correct
2026-01-12 00:01:21 <ech6f> thank you for the help, i will test out the image tomorrow :)