#asahi · 2023-12-30 22:22
j`eyhttps://github.com/AsahiLinux/linux/tree/speakers/enablement-READ-COMMIT-MESSAGE#asahi · 2023-12-30 22:22
j`eysee the branch speakers/enablement-READ-COMMIT-MESSAGE#asahi · 2023-12-30 22:22
j`eykaazoo: the kernel needs extra patches#asahi · 2023-12-30 22:21
kaazooj'ey: No Ubuntu Asahi, with updated packages for the speaker support. Does that ENOENT error mean that macaudio doesn't enable the speakers?#asahi · 2023-12-30 22:16
nicolas17but I have never reversed that kind of thing and I have no clue how brcmfmac works for normal connections >.>#asahi · 2023-12-30 22:15
j`eykaazoo: are you using fedora?#asahi · 2023-12-30 22:14
nicolas17sven: that's probably the only way to do it on Apple Silicon yes#asahi · 2023-12-30 22:14
sventhen you at least get support on Apple silicon machines #asahi · 2023-12-30 22:14
kaazooI get this error during startup: speakersafetyd[4286]: thread 'main' panicked at 'Could not lock elem Left Woofer 1 VSENSE Switch. alsa-lib error: Error("snd_ctl_elem_lock", ENOENT)', src/helpers.rs:172:13#asahi · 2023-12-30 22:14
kaazooHi, does anybody here have working speakersafetyd on j414?#asahi · 2023-12-30 22:14
svensomeone who enjoys pain could also reverse engineer the AWDL firmware protocol and add support to brmcfmac :p#asahi · 2023-12-30 22:13
nicolas17"The reason why Apple prefers EAWs might have to do with the time that is required to perform a channel switch operation in the Wi-Fi chip. We found that a channel switch operation takes at least 8 ms (≈ 8 TU) using dump chanswitch of the wl utility." (from the first paper reverse-engineering AWDL, https://arxiv.org/pdf/1808.03156.pdf#page=13)#asahi · 2023-12-30 22:09
nicolas17so does the chip Mac uses, single channel at a time... it just has to be able to switch channels "fast enough" :D#asahi · 2023-12-30 22:08
PaulFertsernicolas17: not sure if it's really possible for most cards if AWDL requires some channel hopping as most cards only support working on single channel.#asahi · 2023-12-30 22:01
nicolas17but yeah if you're on mac80211 you could also... implement awdl in mac80211 and let it coexist with the normal mode :D#asahi · 2023-12-30 21:57
nicolas17oh, interesting#asahi · 2023-12-30 21:57
PaulFertsernicolas17: on mac80211 monitor mode interface can always be created in addition to existing running VIFs, so it's not a general rule. Probably true for nexmon-patched brcmfmac devices, no idea.#asahi · 2023-12-30 21:54
nicolas17PaulFertser: yes, but if you put it in monitor mode then you can't connect to an access point#asahi · 2023-12-30 21:24
svenor if that thing even was a code signature #asahi · 2023-12-30 21:24
svendunno if the wifi firmware is the same#asahi · 2023-12-30 21:24
sventhe bluetooth firmware looks like it has some kind of code signature fwiw #asahi · 2023-12-30 21:24
PaulFertser(probably a good project for someone interested in RE to try to add support for macs to nexmon)#asahi · 2023-12-30 21:20
PaulFertserAnd fullmac devices can support monitor mode with some convincing https://github.com/seemoo-lab/nexmon#asahi · 2023-12-30 21:16
svenoh, right, totally forgot about monitor mode #asahi · 2023-12-30 21:14
PaulFertserThere's a fully userspace implementations for cards which support active monitor mode: https://github.com/seemoo-lab/owl#asahi · 2023-12-30 21:09
nicolas17since the kernel is parsing the regular data frames anyway#asahi · 2023-12-30 21:08
nicolas17which can probably be implemented in the kernel for softmac devices#asahi · 2023-12-30 21:08
nicolas17it involves vendor-defined Wi-Fi data and action frames#asahi · 2023-12-30 21:05
svenI don’t have any idea how AWDL works but that sounds true since I assume it’s some kind of peer to peer protocol #asahi · 2023-12-30 21:03
nicolas17it does, but that could be offloading for power saving or something, I'm saying on fullmac it *can't* be implemented without firmware help#asahi · 2023-12-30 21:02
svendoesn’t the firmware contain strings that refer to AWDL anyway?#asahi · 2023-12-30 21:02
j`eyconfig BRCMFMAC: tristate "Broadcom FullMAC WLAN driver"#asahi · 2023-12-30 21:01
nicolas17I guess that means AWDL *necessarily* has to be implemented in the firmware, at least partially#asahi · 2023-12-30 21:00
svenfullmac #asahi · 2023-12-30 21:00
j`eytheyre brcmfmac.. so full?#asahi · 2023-12-30 20:59
nicolas17are Wi-Fi chips on Macs fullmac or softmac?#asahi · 2023-12-30 19:36
incogninto1-1Hello#asahi · 2023-12-30 18:31
StonecutterThis is a very cool project. Look forward to what the future has in store. Great job by the team.#asahi · 2023-12-30 18:28
j`eyalso iirc that was a later model than the other m2 ones, which were all sent at the same time#asahi · 2023-12-30 18:25
j`eynot much use to just the DT upstream#asahi · 2023-12-30 18:24
j`eynicolas17: maybe not even sent#asahi · 2023-12-30 18:24
nicolas17although that brings questions... j`ey any idea why stuff like "macbook air 15-inch devicetree" is still not upstream? just waiting for a release?#asahi · 2023-12-30 18:23
StonecutterI see#asahi · 2023-12-30 18:22
nicolas17where the feature support list says "linux-asahi", it means the upstream Linux kernel does *not* support that feature yet, you need asahi's own kernel fork#asahi · 2023-12-30 18:21
j`eyat least they claim to support armv8!#asahi · 2023-12-30 18:21
nicolas17it's up to EndeavourOS maintainers to package asahi's kernel, mesa, asahi-audio...#asahi · 2023-12-30 18:20
StonecutterAt least you said 'for now'..... so theres hope :)#asahi · 2023-12-30 18:19
StonecutterOh really. I see#asahi · 2023-12-30 18:19
j`eyno you need to keep macOS for now#asahi · 2023-12-30 18:19
StonecutterIdeally what I am hoping for is to install EndeavourOS with ease and also delete macOS entirely. Not sure if thats possible yet