#asahi-dev · 2025-09-09 17:49
chaos_princessoh, apple thingy is on?#asahi-dev · 2025-09-09 17:49
jannauno more broadcom wifi/bt: Apple N1 instead#asahi-dev · 2025-09-09 17:42
jannauhuh, I thought the square sensors in macbooks were based on iphone front camera designs#asahi-dev · 2025-09-08 16:51
svencause I don’t see a difference anymore and I don’t see the 5s delay on Mac OS #asahi-dev · 2025-09-08 16:50
svenyeah, probably even the xhci part of dwc3#asahi-dev · 2025-09-08 16:49
chaos_princessyou mean it is a difference between how the oses program the dwc3 core, not the phy?#asahi-dev · 2025-09-08 16:44
svenwith quite some hacks I can get the MMIO trace to essentially match macOS and yet I still see it #asahi-dev · 2025-09-08 16:43
sventhe 5s delay looks more and more like an xhci quirk to me #asahi-dev · 2025-09-08 11:00
nickchanin v1 so not the same case#asahi-dev · 2025-09-08 11:00
nickchanah alright then was looking at how FEAT_PMUv3 in kvm series is handled but it turns out that series did have some discussion about how things to be taken #asahi-dev · 2025-09-08 10:48
maznickchan: no, that's not something I can do. Will is the maintainer for this, and I cannot bypass him that way.#asahi-dev · 2025-09-08 10:41
nickchan(currently on v8 resend)#asahi-dev · 2025-09-08 10:41
nickchanthere's some addressed review comments on v7's resend https://lore.kernel.org/asahi/20250616-apple-cpmu-v7-0-df2778a44d5c@gmail.com/#asahi-dev · 2025-09-08 10:41
nickchanmaz: this might be stretching it but do you think you can take a7-a11's cpu pmu (patch 1-12) through kvmarm? it is your driver and you actually have apple silicon hardware after all. if one squint really hard then patch 2 mentions EL2 so it's not /completely/ off-topic: https://lore.kernel.org/asahi/20250908-apple-cpmu-v8-0-d83b4544dc14@gmail.com/#asahi-dev · 2025-09-07 20:02
kettenism2 pro (t6020) mini here#asahi-dev · 2025-09-07 19:47
jannauu-boot's read_poll_timeout() documentation is wrong, it says "@op: accessor function (takes @addr as its only argument)" while it takes trailing var args#asahi-dev · 2025-09-07 19:20
jannaukettenis: do you have a m2 (t8112) mac mini? the downstream u-boot xhci firmware loading either does not work or the xhci is broken for other reasons.#asahi-dev · 2025-09-07 09:15
jannauhttps://github.com/jannau/linux/commits/apple-soc/drivers-6.18/#asahi-dev · 2025-09-07 09:12
jannauI've rebase and fixed the conflicts for nvme#asahi-dev · 2025-09-07 09:11
jannausven: ok to answer to the t6020 series that we will pick acked changes through asahi/apple-soc/drivers-6.18?#asahi-dev · 2025-09-06 19:55
mjr23 Hello, I’m learning by writing my own stage-1 loader. Quick question about the boot chain: do I need to worry about RVBAR for stage-1, or is my entry point entirely defined by what I enroll with kmutil? From looking at m1n1:start.S, it seems only the Boot ROM deals with RVBAR.#asahi-dev · 2025-09-06 15:46
svenhttps://lore.kernel.org/all/20250906-atcphy-6-17-v2-0-52c348623ef6@kernel.org/#asahi-dev · 2025-09-06 13:49
svenplanning to submit this later as v2: https://git.kernel.org/pub/scm/linux/kernel/git/sven/linux.git/log/?h=wip/atcphy#asahi-dev · 2025-09-06 13:24
svenjust to confirm: if i run make dt_binding_check DT_SCHEMA_FILES=apple,dwc3.yaml and nothing shows up the schema is valid, isn't it?#asahi-dev · 2025-09-06 12:39
jannau+1#asahi-dev · 2025-09-06 12:29
svenI guess I’ll ignore that for now #asahi-dev · 2025-09-06 12:29
sven*hacked #asahi-dev · 2025-09-06 12:29
svenxnu seems to bring up dwc3 earlier and I Hacker that into Linux and still have the 5s delay :/#asahi-dev · 2025-09-06 12:05
svenyeah, that's what i think as well#asahi-dev · 2025-09-06 12:05
jannausven: I certainly used dp-altmode usb3 hub for testing back as well. we probably just didn't realize the 5 second delay is device dependent#asahi-dev · 2025-09-06 11:49
jannauI've tagged asahi-6.16.5-1 and pushed to github. rebase and bits/000-devicetree cleanup, another 50+ commits dropped (mostly by squasing but some duplicates or no longer needed changes)#asahi-dev · 2025-09-06 11:36
jannauI just saw uas errors on shutdown with the downstream kernel so those aren't new / a regression#asahi-dev · 2025-09-06 11:22
jannauhmm, I might misremember state then#asahi-dev · 2025-09-06 11:20
sveni also see the 5s delay on that experiements branch i think#asahi-dev · 2025-09-06 11:08
jannaudownstream kernel has the same behavior btw. device connected to an usb3 dp-altmode switch are detected without delay#asahi-dev · 2025-09-06 10:51
svendoesn't fix the 5s delay though#asahi-dev · 2025-09-06 10:51
sven+ reg = <0x3 0x82280000 0x0 0xcd00>, <0x3 0x8228cd00 0x0 0x3200>;#asahi-dev · 2025-09-06 10:51
sven- reg = <0x3 0x82280000 0x0 0xce00>, <0x3 0x8228ce00 0x0 0x3200>;#asahi-dev · 2025-09-06 10:47
svenack#asahi-dev · 2025-09-06 10:46
jannausven: pushed DT changes for ps_atc_usb_aon to https://github.com/jannau/linux/commits/wip/atcphy-20250906/#asahi-dev · 2025-09-06 10:23
svenyeah, same :)#asahi-dev · 2025-09-06 10:23
jannautunables don't trip "this looks a little weird" anymore#asahi-dev · 2025-09-06 10:20
svenokay, i'll try that one after lunch#asahi-dev · 2025-09-06 10:20
jannausven: no idea about a linux trace without delay. I think the atcphy/2025-experiments* branches don't have the delay. atcphy/2025-experiments_v6.14 would be the most recent#asahi-dev · 2025-09-06 10:17
svenah nvm, i already added it#asahi-dev · 2025-09-06 10:17
svenalso, should i directly add the t6020 compatible?#asahi-dev · 2025-09-06 10:14
svenor alternatively, do you still have a trace_atc.py from a linux tree where we didn't have that 5s delay?#asahi-dev · 2025-09-06 10:09
svenhttps://git.kernel.org/pub/scm/linux/kernel/git/sven/linux.git/commit/?h=wip/atcphy&id=1cf8afff18c8224b150fa458cb5722b079da9b56 i prefer the tunables like this#asahi-dev · 2025-09-06 10:07
svenjannau: do you remember which tree didn't have the 5s delay?