#asahi-dev · 2024-03-29 05:04
chadmedoh for fucks sake they changed the matching behaviour from AND to OR #asahi-dev · 2024-03-29 04:26
chadmedso we're almost ready for the cutover#asahi-dev · 2024-03-29 04:26
chadmedcurrently main is (mostly) working with wireplumber 0.5.x#asahi-dev · 2024-03-29 04:24
chadmedso im basically just going to say no more 1.x backports after so-and-so date and time#asahi-dev · 2024-03-29 04:24
chadmedbut obviously i cant keep doing this forever and supporting new devices across both is a PITA#asahi-dev · 2024-03-29 04:24
chadmedcurrent plan for asahi-audio is for main to continue tracking upstream wireplumber developments and maintain a 0.4.x compat branch which will continue to be tagged out as 1.x releases#asahi-dev · 2024-03-29 04:23
chadmednot officially supporting wireplumber 0.4.x at all after a certain point#asahi-dev · 2024-03-29 04:16
tpw_rulesnot allowing them?#asahi-dev · 2024-03-29 04:16
tpw_ruleswhat exactly does that mean#asahi-dev · 2024-03-29 04:09
chadmeddoes anyone have any objections if i put a hard limit on asahi-audio 1.x backports?#asahi-dev · 2024-03-28 21:21
jannegcounter resolution is 24 MHz though#asahi-dev · 2024-03-28 20:28
jannegsetting BIT(1) specifically#asahi-dev · 2024-03-28 20:22
jannegmsr(s3_1_c15_c1_5, 0x3uL) switches CNTFRQ_EL0 to 24 MHz#asahi-dev · 2024-03-28 20:10
jannegM3 performance cores come out of reset with 1GHz as CNTFRQ_EL0. I guess I should check the counter resolution and then see which reg write changes that #asahi-dev · 2024-03-28 17:40
nicolas17lota: how did you even get pacman/makepkg in fedora lol#asahi-dev · 2024-03-28 17:18
lotaI read an answer from 1 year ago but now it uses fedora ...#asahi-dev · 2024-03-28 17:17
lotaargh pacman is for arch and not fedora I think#asahi-dev · 2024-03-28 17:16
lotaany idea how to fix that ?#asahi-dev · 2024-03-28 17:16
lota==> ERROR: 'pacman' returned a fatal error (255):#asahi-dev · 2024-03-28 17:16
lotacould not create database#asahi-dev · 2024-03-28 17:16
lota(root: /, dbpath: /var/lib/pacman/)#asahi-dev · 2024-03-28 17:16
lotaerror: failed to initialize alpm library:#asahi-dev · 2024-03-28 17:16
lota==> Making package: visual-studio-code-bin 1.87.2-1 (Thu 28 Mar 2024 05:15:46 PM UTC)#asahi-dev · 2024-03-28 17:16
lota==> Checking runtime dependencies...#asahi-dev · 2024-03-28 17:16
lota~/Dev/visual-studio-code-bin$ makepkg -si#asahi-dev · 2024-03-27 21:08
leioyeah, I meant before realizing 6.8 itself works for what you were testing, but you were just checking if it'd be easy to figure out if it is indeed broken; seeing #-alt now#asahi-dev · 2024-03-27 21:05
jannegno, just commits changing something in net/bluetooth/ maybe 20 commits#asahi-dev · 2024-03-27 21:00
leioyou reverted half of 6.8 without it being broken? :)#asahi-dev · 2024-03-27 19:02
jannegseems to be false alarm. I can't reproduce#asahi-dev · 2024-03-27 18:50
jannegbcm4378 (mac mini m1) works here#asahi-dev · 2024-03-27 18:29
jannegyes, seems to be easy enough to bisect. reverted half of the v6.7..v6.8 changes and it's working#asahi-dev · 2024-03-27 18:12
svenbt broke again? :/#asahi-dev · 2024-03-27 18:04
leiook, thanks#asahi-dev · 2024-03-27 18:04
jannegleio: https://github.com/jannau/linux/tree/beta-6.8/asahi-wip bt le pairing seems to be broken, wifi works#asahi-dev · 2024-03-27 17:41
leioI'd actually prefer not being stuck in 6.6 :)#asahi-dev · 2024-03-27 17:41
leiofor the HDMI stuff#asahi-dev · 2024-03-27 17:41
leiojanneg: which branch I should just grab? I don't mind 6.8 if only bluetooth and/or wifi is broken, don't use them#asahi-dev · 2024-03-26 21:20
jannegI've really tried with 300ms. and it might be the delay and not the ignorant debounce implementation#asahi-dev · 2024-03-26 21:17
leioI think 6K would not work well for DPMS, so I stopped using it and turned monitor off for the night instead; that one needed the HDMI replugging gymnastics#asahi-dev · 2024-03-26 21:17
PaulFertserInteresting, 500 ms is really plenty and I assume 300 worked less reliably so it might be not a simple debounce but also a meaningful delay that helps.#asahi-dev · 2024-03-26 21:17
leiothanks, I'll try to try these all weekend or so. I'm been lucky lately with just locking the screen and waking up now with the old 4K monitor, even without the power-domains fix#asahi-dev · 2024-03-26 21:16
PaulFertsertypo#asahi-dev · 2024-03-26 21:16
PaulFertseralso a type in deboune#asahi-dev · 2024-03-26 21:15
PaulFertserThe string in the dev_info says 300 not 500.#asahi-dev · 2024-03-26 21:13
jannegleio: this seems to improve M2 desktop HDMI reliability: https://github.com/jannau/linux/commit/5ef7d920caa52ee0331ec1c71f3a67857a6a3427 (either use the branch with rust 1.76+ or that commit on-top of asahi-6.6-16)#asahi-dev · 2024-03-26 12:54
chadmedso i think theres some give and take here#asahi-dev · 2024-03-26 12:54
chadmedinserting yourself as a crucial part of the linux desktop supply chain while still being officially unstable then deciding to completely and utterly break virtually every single consumer of your api between versions is not very neighbourly#asahi-dev · 2024-03-26 12:51
leiodoesn't sound very "upstream first" :D#asahi-dev · 2024-03-26 12:50
chadmedlets go that works