#
✕ nick:sven
#asahi-dev · 2024-02-02 16:25
svenfirst step would be to reproduce the issue. I believe some people reported that Bluetooth devices won’t reconnect after a suspend
#asahi-dev · 2024-02-02 16:24
sveneventually you’d have to send a patch to the mailing list but we can add the fix to our downstream kernel first
#asahi-dev · 2024-02-01 18:15
sventhere have been reports that devices disconnect and don’t connect again. I suspect it’s just a bug in the driver because I didn’t see any special Bluetooth messages before suspend in macos
#asahi-dev · 2024-01-22 11:12
svennow if only I could connect my external display to it ;p
#asahi-dev · 2024-01-22 11:11
svenI’m still very happy with my m1 air but I’ve always been a fan of smaller laptops
#asahi-dev · 2024-01-22 09:33
sven+and
#asahi-dev · 2024-01-22 09:33
svenbut maybe the AOP needs to wake the bt controller up every now then
#asahi-dev · 2024-01-22 09:32
svenhrm, i assumed the find my stuff was mostly implemented inside the bluetooth firmware
#asahi-dev · 2024-01-21 21:20
svenouch. I very much did not enjoy writing that thing originally
#asahi-dev · 2024-01-21 21:15
svenwait, there’s something else that uses the mess that is afk?
#asahi-dev · 2024-01-16 11:06
svenI'm using a similar setup with tmux, so I don't really need such a feature <-- same here
#asahi-dev · 2024-01-15 13:23
svenALS?
#asahi-dev · 2024-01-11 17:16
sventhe rest is just half-knowledge and there are better people to ask :)
#asahi-dev · 2024-01-11 17:16
svenand that’s also where most of what I actually know about brcmfmac stops
#asahi-dev · 2024-01-11 17:14
sventhe Bluetooth hardware works pretty much like this
#asahi-dev · 2024-01-11 17:14
svenfrom what little I’ve seen of bcrmfmac it also does something similar to this
#asahi-dev · 2024-01-11 17:13
svenso essentially you have one (or more) ringbuffers where one sides writes messages and increments the write pointer (which usually triggers an irq for the other side) and the other side reads messages and increments the read pointer
#asahi-dev · 2024-01-11 06:34
svenprobably better to talk to dberlin though, he seems to actually understand what’s going on in there
#asahi-dev · 2024-01-11 06:33
svenfrom what I know it’s essentially a bunch of message rings used to exchange information
#asahi-dev · 2024-01-06 10:05
svenworks for me but it's probably best to ask janneg, he's been taking care of dptxep.c (and dcp) lately
#asahi-dev · 2024-01-06 10:04
svenwe should just add those headers directly
#asahi-dev · 2024-01-06 10:02
svenyeah
#asahi-dev · 2024-01-06 10:02
svengood catch, I keep forgetting some of those headers that get included by accident
#asahi-dev · 2024-01-06 10:02
sven:D
#asahi-dev · 2024-01-04 19:16
svenbut I’ll take a closer look
#asahi-dev · 2024-01-04 19:16
svenapple uses a custom firmware update mechanism I think
#asahi-dev · 2024-01-02 15:14
svenplease move this somewhere else
#asahi-dev · 2024-01-02 14:59
svenmight be a good initial project as well, I think dberlin already worked out how it works and what changed
#asahi-dev · 2024-01-02 14:58
svenLinux kinda needs a working interrupt controller to boot :D
#asahi-dev · 2024-01-02 14:58
svenoh, right, so first step would be to add aic3 support
#asahi-dev · 2024-01-02 14:58
svenand then see what works and what doesn’t. usb won’t work since the tipd chip has a new bus (spmi instead of i2c)
#asahi-dev · 2024-01-02 14:57
svenfirst step would be to get that running, then the next step would be to create an initial device tree for m3
#asahi-dev · 2024-01-02 14:56
svenI think there are some pull requests for m1n1 for initial m3 support
#asahi-dev · 2024-01-02 14:55
svenah, I thought it was a pro but I probably misremembered
#asahi-dev · 2024-01-02 14:54
sven:D
#asahi-dev · 2024-01-02 14:54
svendberlin has done some initial m3 bringup work fwiw
#asahi-dev · 2024-01-02 14:54
svenI don’t think lack of hardware is the bottleneck for initial bringup right now anyway
#asahi-dev · 2024-01-02 14:51
svendunno how likely that is though
#asahi-dev · 2024-01-02 14:51
svenonce we have thunderbolt some dock might use a renesas xhci controller I guess
#asahi-dev · 2024-01-02 12:15
svenand the atcphy code I wrote during 37c3 and on that train broke usb3 again it seems *sigh*
#asahi-dev · 2024-01-02 10:02
svenyeah, that too
#asahi-dev · 2024-01-02 09:55
svenremote access is usually not that helpful unless something very specific has to be tested fwiw
#asahi-dev · 2023-12-29 14:33
svenThat’s not going to work without a screen, keyboard and mouse attached
#asahi-dev · 2023-12-29 14:27
svenit’s like a four hour train ride and I need to rewrite what feels like half of atcphy!
#asahi-dev · 2023-12-29 14:27
svenyup :D
#asahi-dev · 2023-12-29 14:26
svenI don’t even know what userspace I have but I can build rust in a local vm and transfer that binary over
#asahi-dev · 2023-12-29 14:25
svenPerfect, that’s just what I was looking for
#asahi-dev · 2023-12-29 14:21
svennice, let me try that
#asahi-dev · 2023-12-29 14:15
svenis there any way to disable that startup chime from Linux or m1n1? I’m on a train and having that annoying sound every few minutes might piss off the other passengers rather quickly :D
#asahi-dev · 2023-12-29 11:42
sveni think we can rid of that entire weird synchronization dance with dwc3

plus ›