04:49 <chadmed> i am quickly becoming of the opinion that the best way to handle kernel rust infrastructure for gentoo is to make it a free-for-all, at least in the mid term

04:51 <chadmed> rust itself is no issue, just add a use flag and depend on >=virtual/rust-1.62.0. bindgen-cli being a crate though makes it hard to automatically install and keep track of. the cargo eclass is not conducive

04:51 <chadmed> having to manually specify _all_ the crate dependencies of bindgen AND their specific versions is insane, and it doesnt even work

04:57 <sam_> i don't get what you mean

04:57 <sam_> it's pretty easy to do rust ebuilds, even if gross

05:00 <chadmed> i followed the example of cbindgen but it had a whinge, ill try again

05:01 <sam_> then ask me about it or #gentoo-rust

05:01 <sam_> it's not some sort of unsolvable problem

05:01 <sam_> do you want me to try?

06:59 <chadmed> nah all good i figured it out, something like this? https://github.com/chadmed/gentoo/tree/bindgen-wip/dev-util/rust-bindgen

11:11 <Ry_Darcy> Hallo all, running Debian testing on an M1 Mini. Is this the correct sequence to update the DTB's on my machine - cat /Transit/build/m1n1/m1n1.bin /Transit/linux-asahi.../arch/arm64/boot/dts/apple/*.dtb /Transit/u-boot-asahi/u-boot-nodtb.bin > /boot/efi/m1n1/boot.bin? with a subsequent echo display=3840x2160 >> /boot/efi/m1n1/boot.bin (I have a 4K Monitor).

11:47 <Glanzmann> Ry_Darcy: That, or you can try the new DCP driver from jannau and use this script to compile and install a kernel: https://pbot.rmdir.de/nTi2tcsdqE_M3v8zXNcAkw

11:48 <Glanzmann> In your case you can also drop the rust and LLVM than it would be: https://pbot.rmdir.de/9IU4QZfS_qMSvJDXePwrCA

11:51 <Ry_Darcy> Thanks Thomas. I have been having problems using the above sequence. I get an Asahi logo on reboot - nothing else - dead. I then have to reboot into 1TR and copy a known, working boot.bin to /boot/efi/m1n1/ to recover, which drops me back to 2K resolution also. Frustrating.

11:52 <j`ey> Ry_Darcy: you need to gzip uboot

11:52 <j`ey> I think that may help you

11:54 <Ry_Darcy> Tried the gzip varient also. It worked about two weeks ago. With my impatience and a bleeding edge Asahi kernel things have gone awry again. Maybe I have gone insane also. Not sure :)

11:54 <Ry_Darcy> variant

11:56 <Glanzmann> Ry_Darcy: Take the script that I posted, it works for Debian.

11:56 <Glanzmann> Put it in revision control (like git) and only modify it.

11:57 <Glanzmann> You can also go back easy to the old kernel by copying the m1n1 from macos and than select the right kernel from grub.

11:57 <Ry_Darcy> Will try https://pbot.rmdir.de/9IU4QZfS_qMSvJDXePwrCA this evening and report back. I first have to recover (1TR) things first before I try your magic.

11:57 <Glanzmann> j`ey: Good point with the gziped u-boot. In the old days it also worked without zipping but I think these days it is mandatory.

11:58 <j`ey> it's needed if you want to pass the display= arguments to m1n1

11:58 <Glanzmann> Ry_Darcy: As soon as markan releases the new kernel, I'll also update the Debian kernel.

11:58 <Glanzmann> j`ey: Ah, I see.

11:59 <Glanzmann> But the new kernel has dcp (asahi_drm) with that you don't need to pass the display parameters anymore.

11:59 <Glanzmann> Of course the graphics performance will be a little bit worse than with simpledrm.

11:59 <Ry_Darcy> Regarding using gzip... A lot of old/obsolete/misleading articles whether one should or not use gzip in this process.

11:59 <Glanzmann> Ry_Darcy: Always use it. Otherwise m1n1 can't tell the end of u-boot.

12:00 <Glanzmann> Ry_Darcy: I keep the m1-debian repository up to date.

12:00 <Glanzmann> So you can also look there how to do it properly.

12:01 <Ry_Darcy> These snippets of wisdom, heaven! I will report as stated earlier.

12:01 <Ry_Darcy> report back rather.

12:01 <Glanzmann> Thanks, if it does not work, I'll give you my build artefacts.

12:01 <Ry_Darcy> Grazie a tutti for the information.

12:05 <mps> Ry_Darcy: di niente ;)

12:13 <Ry_Darcy> :)

15:32 <Ry_Darcy> @Glanzmann, writing this after a reboot using your https://pbot.rmdir.de/9IU4QZfS_qMSvJDXePwrCA script. Thank you.

15:35 <Ry_Darcy> Having used XFCE for years, I changed to LXQT last year. After rebooting just now, LXQT is not working i.e. I see the blue Debian logo, mouse is working, but nothing else, in other words no (LXQT) desktop. Ditto for XFCE. I have the (dreadful) Plasma desktop working. This is where I am typing from at the moment.

15:37 <Ry_Darcy> Have I forgotten something? LXQT/XFCE were working yesterday. Is it a DCP setting somewhere that needs to be tweaked?

15:43 <Ry_Darcy> Some screenshots https://pastebin.com/E4CEMKBN

15:45 <Ry_Darcy> https://www.klgrth.io/paste/dhvse

15:49 <ashi> Is apple-drm supposed to work on a macbook air m2? for me there is no handover and if I do not configure simpledrm I get a black screen, also nothing in dmesg about apple-drm

15:52 <_jannau_> ashi: it is not supported

15:53 <ashi> _jannau_: thanks! good to know.

15:59 <Ry_Darcy> https://pastebin.com/EfxRBAUC

15:59 <Ry_Darcy> I am not sure what I else I should look for.

16:11 <sam_> chadmed: bang on

16:11 <sam_> chadmed: just fix LICENSE (the ()s are dodgy, I bet it's supposed to be || ( ... )?)

16:12 <sam_> (it's somehow misparsing the cargo metadata)

22:27 <Glanzmann> Ry_Darcy: You probably need to disable dcp (apple_drm) and use simpledrm again. Here is another script for you: https://tg.st/u/rydarcy.sh

22:27 <Glanzmann> I also write that to you via email.