#asahi-dev · 2024-04-26 06:17
chadmedmake LLVM=1 LLVM_IAS=1#asahi-dev · 2024-04-26 06:17
leiohow did I change kernel compile to clang? :D#asahi-dev · 2024-04-26 06:16
chadmedcould be but i dont think if (x) { int y; } was ever defined behaviour right#asahi-dev · 2024-04-26 06:14
leiothat feels false positive warning if triggered for new variable decls after a block start#asahi-dev · 2024-04-26 06:14
chadmedhold on let me clean up, commit and push #asahi-dev · 2024-04-26 06:14
chadmedthe warning is just "you cant have a declaration after a label" and is turned on by -Wc2x-extensions#asahi-dev · 2024-04-26 06:11
jannauI haven't compiled with clang#asahi-dev · 2024-04-26 06:10
leioand the warning?#asahi-dev · 2024-04-26 06:10
jannaugcc was erroring out about the not being able to reduce the case statements to integer constants after rebasing onto 6.8.6-3#asahi-dev · 2024-04-26 06:08
jannauhttps://github.com/jannau/linux/blob/macsmc-hwmon/drivers/hwmon/macsmc-hwmon.c#asahi-dev · 2024-04-26 06:07
leiowhere can I see the code itself as opposed to a diff?#asahi-dev · 2024-04-26 06:06
leio(those that do these early testing with newer versions and warning options)#asahi-dev · 2024-04-26 06:05
leioI have fixed upstream clang compilation warnings in the same way successfully and without any further problems by gentoo clang users#asahi-dev · 2024-04-26 06:04
jannauI thought so too, that's why I wrote the misguided patch, but I guess goto and switch labels are special#asahi-dev · 2024-04-26 05:57
leioand that's what I see in the linked patch to be done before and after?#asahi-dev · 2024-04-26 05:57
leioshould be fine if it's at start of a block and you can start a block after case label#asahi-dev · 2024-04-26 05:53
chadmedyeah not a big deal just unsure how upstream will feel about it when we get to that stage#asahi-dev · 2024-04-26 05:44
jannauah, that should be fine. a way out of it would be to move the parsing to functions#asahi-dev · 2024-04-26 05:31
chadmedthats only officially adopted for C23#asahi-dev · 2024-04-26 05:31
chadmednah its the fact that there is a declaration after a label (the label being the case statement) #asahi-dev · 2024-04-26 05:29
jannauchadmed: about '_SMC_KEY("flt ")'? try https://gist.github.com/jannau/0c0b5d7b95ce00cd7c3a4f989ebccdd8 instead#asahi-dev · 2024-04-26 02:41
chadmedif we're only needing to support 'flt ' and 'ioft' then its fine imo#asahi-dev · 2024-04-26 02:40
chadmedeasiest way to make it shut up would be to go back to declaring all the key types we support at the top of the function, as annoying as that is#asahi-dev · 2024-04-26 02:37
chadmedgcc doesnt complain but clang emits a warning#asahi-dev · 2024-04-26 02:37
chadmedjannau: macsmc_hwmon_read_key is not valid c11, so idk if we can keep it #asahi-dev · 2024-04-25 07:11
jannaudone. added a note on the current state#asahi-dev · 2024-04-25 06:52
jannauthe soc entries are probably not needed anymore now that I have it working on all. When I added them t602x was still not working#asahi-dev · 2024-04-25 06:15
chadmedoh you added to the devices and the socs. i thought the soc ones were duplicates #asahi-dev · 2024-04-25 06:12
jannauchadmed: I added that 2 weeks ago, do we have now duplicates?#asahi-dev · 2024-04-25 05:40
chadmedjannau: hdmi audio was already a line item on the device specific feature tables :) #asahi-dev · 2024-04-24 18:31
j`eydonaldh_: there is a bunch of PR's to the m1n1 repo, that are adding M3 support, check those out#asahi-dev · 2024-04-24 18:28
donaldh_Hi, I have an M3 Max laptop and would like to contribute to M3 support. Is it possible to set up an M3 machine with m1n1 and is there anything a new contributor to asahi can work on?#asahi-dev · 2024-04-24 16:26
cyrinux3(on asahi fedora) #asahi-dev · 2024-04-24 16:25
cyrinux3can't enter luks password #asahi-dev · 2024-04-24 16:25
cyrinux3no keyboard at start #asahi-dev · 2024-04-24 16:25
cyrinux3warning, after updating m2 air#asahi-dev · 2024-04-24 16:25
cyrinux3hi#asahi-dev · 2024-04-23 22:20
chadmed_jannau: thats the same person who kept bugging me about the audio "firmware" #asahi-dev · 2024-04-23 20:03
jannausigh, Marek. just a few days and there's already the first "complaint" about the env based USB device ignore list: https://github.com/AsahiLinux/m1n1/issues/388#asahi-dev · 2024-04-23 07:24
chadmedSND_USE_CASE_DEV_HDMI#asahi-dev · 2024-04-23 07:24
chadmedthere are SectionDevice identifiers for HDMI so perhaps we start with the low-hanging fruit and implement a UCM file for it then work from there#asahi-dev · 2024-04-23 07:21
jannauthe current state on driver side is certainly incomplete, I'm just not sure in which direction#asahi-dev · 2024-04-23 07:20
chadmedlooks like it's going to be mostly UCM looking at what HDA does but looking at the driver it does seem a bit incomplete. perhaps because we are not advertising IEC958 modes it's mucking with what userspace thinks we can do idk #asahi-dev · 2024-04-23 07:03
jannauand it's not ASoC driver#asahi-dev · 2024-04-23 07:00
jannauno, it's a separate snd_card in drivers/gpu/drm/apple/audio.c and it's povik's code not mine#asahi-dev · 2024-04-23 06:57
chadmedjannau: i believe it is a combination of the two. i assume you made snd_soc_macaudio changes for dcp audio?#asahi-dev · 2024-04-22 21:40
jannauchadmed: any idea what's necessary to have DP / HDMI audio not show up as "Analog output (built-in audio stereo)"? UCM definitions or snd_card in kernel changes#asahi-dev · 2024-04-21 18:48
LarstiQjannau: exciting!#asahi-dev · 2024-04-21 17:28
jannauI still need to look at hotplug. I would be be pleasantly surprised if it works#asahi-dev · 2024-04-21 17:27
jannautested on m1, m2, m2 pro mac mini and m1 max 14 inch macbook pro