14:04 <apritzel> diego71: can you try this DT? https://gist.github.com/apritzel/60fd7379b9560a15b8b352986b69d39f

14:07 <apritzel> diego71: it's this kernel tree: https://github.com/apritzel/linux/commits/a733-rfc/

14:07 <apritzel> I would hope that the Radxa U-Boot can load mainline kernels ...

15:51 <MoeIcenowy> paulk: I got the ddk 1.11 public headers from yuzuki

15:52 <paulk> MoeIcenowy: neat! can you share them or should I ask directly?

15:52 <paulk> they're all MIT

15:53 <MoeIcenowy> should I mail you ?

15:53 <MoeIcenowy> the sys-base mail address

15:54 <paulk> yes perfect, thanks!

15:57 <paulk> not sure I mentioned it here but I'm working to integrate powervr rogue blobs for A133 with the GBM/Wayland WSEGL plugin from Renesas

15:58 <paulk> which should allow running any wayland compositor and any software using direct kms with EGL

15:58 <paulk> but I hear there's not that much interest in A133 anyway :)

15:59 <MoeIcenowy> well at least a few people should be interested

16:01 <paulk> yes people are making products with it, but less interest from the community

16:02 <paulk> with display and gpu working, the next step will be cedrus

16:02 <paulk> and I'd really like to work on the camera side too

16:06 <MoeIcenowy> paulk: got the mail?

16:06 <paulk> MoeIcenowy: yes all good, many many thanks!

16:08 <apritzel> well, the lack of HDMI is a bit of a bummer, at least for me

16:11 <paulk> yes but I guess it reduces the price tag, which makes it interesting for all things using a single lcd

16:12 <paulk> though lvds is limited to 1366x768

16:13 <paulk> (in single mode, dual goes up to 1920x1200)

16:13 <paulk> mhh maybe that's always the case for lvds

17:59 <diego71> apritzel: I've tried but I've no output on serial, so I don't know if it's working or not

18:08 <diego71> you can find here the log messages on serial:

18:08 <diego71> https://eoval.org/MoT3sEJv

18:10 <diego71> I don't know if it is relevant: with kernel 6.17-rc1 with you patches it consume 0.11A, with kernel 5.15 from radxa consume 0.55A

19:01 <apritzel> diego71: ah, many thanks for trying. I gracefully forgot about the completely bonkers command line that those BSP kernels use

19:02 <apritzel> so can you change that? You would just need: "console=ttyS0,115200n8 earlycon"

19:04 <diego71> ok, I'll try

19:07 <diego71> ok, now it writes something

19:09 <diego71> https://eoval.org/gCuMRzgM

19:18 <diego71> complete log: https://eoval.org/gCuMRzgM

19:20 <apritzel> that's the same short one?

19:22 <apritzel> diego71: so what happens at the end? Do you get a panic because of a missing rootfs? Do you have some basic/recovery initrd, which drops you to a shell?

19:29 <diego71> complete log is from a new boot, saving the serial on file

19:30 <diego71> and yes it panics because don't find the block device

19:30 <diego71> but I see no shell

19:32 <diego71> no more output after: Dropping to a shell!

19:38 <diego71> sorry, I see now, that I send you the wrong link. https://eoval.org/Lx7hz4rz#screenlog.0

19:40 <apritzel> ah yeah, see it now, thanks

19:41 <apritzel> but that's still a filtered log, right? Just showing any problematic lines, not the normal ones.

19:45 <diego71> yes, I can try with earlycon

19:46 <diego71> or better, I see a loglevel=4 in the kernel cmdline

19:46 <diego71> I'll increase it

19:50 <apritzel> did you change the kernel's default loglevel? That should print just the normal output, with a lot more messages

19:54 <diego71> i'm using u-boot from raxda, with its default cmdline

19:55 <apritzel> can't you completely replace/overwrite the command line? You literally should just need the two parameters I posted

19:58 <apritzel> I just booted a similarly stripped DTB on the Cubie A5E, and that booted to an initrd prompt, even with those "Couldn't get bank PF regulator" messages (which are mostly temporary)

19:58 <diego71> yes, i'll do it. BTW loglevel=6 give the same messages

20:00 <apritzel> defconfig default is 7, I think

20:00 <apritzel> as grown kernel developers, I think we can cope with the unfiltered view ;-)

20:02 <diego71> it looks better

20:03 <diego71> https://eoval.org/z2PN96Dq#screenlog.0

20:04 <apritzel> ah yeah, much better, thanks!

20:05 <diego71> I'm going to sleep now, if you need more test I can do them tomorrow

20:08 <apritzel> many thanks, that's already very helpful!

20:20 <spawacz> When plugging my trusty D1 based sbc to the power, it often does not boot, imediatelly prompting from u-boot for reset. I added a script in u-boot that it flashes the led when it reaches grub, so if I don't see it I power cycle the board. Now, is there any reason for it ask for reset just immediatelly after powering on?

20:20 <spawacz> It does it on this very first stage of u-boot, just miliseconds after powering on

20:21 <spawacz> the thing of course is that I am not plugged into UART all the time