#asahi-dev · 2025-05-04 16:16
svenalso added some initial dt-bindings for dwc3 and atcphy#asahi-dev · 2025-05-04 15:42
svennice, looks like that svid = <0x..> hack is no longer required since Heikki cleaned all that up in 4aebc4f89f00#asahi-dev · 2025-05-04 11:37
svenhttps://github.com/AsahiLinux/linux/commits/sven/atcphy/ rebased on 6.15-rc1, first clean up pass, only compile tested so far#asahi-dev · 2025-05-04 10:52
svenso not very surprising the ports gets into a weird state eventually#asahi-dev · 2025-05-04 10:52
svenyeah, and resets are already a bit sketchy with the current dwc3/atcphy driver #asahi-dev · 2025-05-04 10:51
nickchan(right now it's the only machine I can sanely do that even)#asahi-dev · 2025-05-04 10:51
nickchanbut yeah I use it because the m1 pro is the fastest machine i can compile linux on and then boot it on an iphone#asahi-dev · 2025-05-04 10:51
nickchansven: right every checkm8 tool needs to do that#asahi-dev · 2025-05-04 10:48
svennot surprising then tbh, doesn't it trigger a reset in the middle of a transaction? #asahi-dev · 2025-05-04 10:47
svenyou mean running checkra1n on a asahi linux machine?#asahi-dev · 2025-05-04 10:40
nickchanokay it's not as bad as I made it out to be but it happens sometimes more than the background noise#asahi-dev · 2025-05-04 10:38
nickchananyways about the typec i will surely say that using checkra1n over and over is a great way to make the ports explode and dead until next reboot#asahi-dev · 2025-05-03 16:52
allmazzit worked, thank you#asahi-dev · 2025-05-03 16:50
chaos_princessprobably?#asahi-dev · 2025-05-03 16:49
allmazz2024-04-12?#asahi-dev · 2025-05-03 16:49
chaos_princesstry 1.86#asahi-dev · 2025-05-03 16:48
allmazzrustc 1.88.0-nightly (4824c2bb7 2025-05-02)#asahi-dev · 2025-05-03 16:48
chaos_princessrustc --version ?#asahi-dev · 2025-05-03 16:47
allmazzrust is available#asahi-dev · 2025-05-03 16:47
chaos_princesswhat does `make rustavailable` say#asahi-dev · 2025-05-03 16:44
allmazzhey, pls help me compile the kernel. i have latest rust nightly toolchain, but during compilation see "no method named sub_ptr found". as i see in code, required unstable feature is enabled. looks like i changed rustup config or something somewhere#asahi-dev · 2025-05-03 15:35
svensame difference#asahi-dev · 2025-05-03 15:15
nickchanthere are more than 1000 downstream commits pretty sure#asahi-dev · 2025-05-03 12:17
svenjannau: sounds reasonable from a quick look#asahi-dev · 2025-05-03 12:12
sveni'd just like to try because i really want to avoid another situation where we have >500 downstream commits for years which are never submitted#asahi-dev · 2025-05-03 12:11
chaos_princessi wouldn't trust _myself_ to be able to do that :P, but if you are confident, sure would be interesting to try it that way #asahi-dev · 2025-05-03 12:09
svenmy goal for next year (usual rule: if you ask i'll double the estimate) is to even bring up m3 or m4 upstream-first, i.e. once it boots locally just directly submit to ML#asahi-dev · 2025-05-03 12:09
jannauI think it would make sense to move from the oob hotplug to a dp-hpd drm_bridge. that would allow passing connection parameters to dcp#asahi-dev · 2025-05-03 12:08
chaos_princesssounds reasonable, and it is up to you at the end#asahi-dev · 2025-05-03 12:08
sventhe only truly critical stuff are the dt-bindings due to backwards/forwards compatible rules#asahi-dev · 2025-05-03 12:08
sveni'd test this for a short while downstream, then submit, then iterate until it's upstream and then just fix bugs there#asahi-dev · 2025-05-03 12:07
svenyeah, but i also don't wanna have this rolled out in our downstream tree for months+#asahi-dev · 2025-05-03 12:07
chaos_princessfair, my concern was more about the fact that users are great at finding random issues and it is faster to fix things when they are not upstream yet#asahi-dev · 2025-05-03 12:06
svenchaos_princess: more seriously though, i just want to have the stuff i work on in a shape where i could theoretically just hit b4 send#asahi-dev · 2025-05-03 12:05
sventbt + pcie side as well#asahi-dev · 2025-05-03 12:05
svennot just from the display side tbh#asahi-dev · 2025-05-03 12:04
jannaudt bindings are my biggest concern from display side#asahi-dev · 2025-05-03 12:04
sventhat's why i said "this year" :D#asahi-dev · 2025-05-03 12:03
chaos_princessdon't we want to roll it out to our lab rat^W^W users first?#asahi-dev · 2025-05-03 12:02
svenbut i'm going to be very happy if we get usb3/atcphy upstream this year#asahi-dev · 2025-05-03 12:00
sven*if i can get#asahi-dev · 2025-05-03 12:00
sveni'll clean that up a bit further to a series that can actually be submitted, then see if i can my old tbt code to work again and then we'll probably have to discuss dt bindings for the mess#asahi-dev · 2025-05-03 11:58
jannaubriefly boot tested on m1 mini#asahi-dev · 2025-05-03 11:58
sventhanks!#asahi-dev · 2025-05-03 11:58
jannausven: I've pushed atcphy/2025-experiments_v6.14 (cleaned up and rebased atcphy/2025-experiments, cleanup was dropping commits and their reverts from the old atcphy/dp series#asahi-dev · 2025-05-03 11:13
svenack. I’ll start by reading through the changes marcan did in early January #asahi-dev · 2025-05-03 10:33
jannausven: let me check if I have local changes (and which branch is the most recent)#asahi-dev · 2025-05-03 10:13
svenguess it's time to look into the dwc3/atcphy/tbt mess again 🫠#asahi-dev · 2025-05-03 10:13
svenhttps://lore.kernel.org/asahi/20250503-smc-6-15-v4-0-500b9b6546fc@svenpeter.dev/ smc, v4.#asahi-dev · 2025-05-02 13:46
svenif gpio wasn't already reviewed, etc. i would've only sent reboot even