#
✕ nick:sven
#asahi · 2022-07-13 07:15
svenwould also be good to understand if it's uninitialized garbage or additional bits that have some meaning
#asahi · 2022-07-13 07:11
svenit's part of the core bluetooth stack
#asahi · 2022-07-13 07:11
svennope :D
#asahi · 2022-07-13 07:10
svenand there might be other places where we'll have to mask the upper bits as well
#asahi · 2022-07-13 07:09
svenit needs to be quirk at least
#asahi · 2022-07-13 07:09
svenprobably not
#asahi · 2022-07-13 07:03
sventhat needs a proper fix which will go into the repo eventually
#asahi · 2022-07-12 20:00
sveni wonder why those upper bits are messed up
#asahi · 2022-07-12 19:59
svenoh dear
#asahi · 2022-07-12 19:59
svenhahahahahaha :D
#asahi · 2022-07-12 19:51
svenhttps://f.svpe.de/6a34e3ae6485af13a43a7ab02ceccfefe76a4bda830cb124a6ad2e213eb5409d_0001-lol.patch
#asahi · 2022-07-12 19:49
svenLOL, that actually makes it discover more devices here
#asahi · 2022-07-12 19:47
svenjannau: so... uh.... can you put a evt_type &= 0xff; into ext_evt_type_to_legacy after the very first if where it checks for legacy events in net/bluetooth/hci_event.c? :D :D
#asahi · 2022-07-12 19:44
svendid they just mess up and put garbage in the upper half?
#asahi · 2022-07-12 19:43
sven..... LE_LEGACY_ADV_IND is 0x13
#asahi · 2022-07-12 19:39
svenbut btmon still seems to be able to parse most of it?
#asahi · 2022-07-12 19:38
svenwhy :-(
#asahi · 2022-07-12 19:38
svenugh
#asahi · 2022-07-12 19:38
svenLegacy PDU Type: Reserved (0x2613)
#asahi · 2022-07-12 19:35
sveni hope Apple/Broadcom didn't decide to re-invent the advertising report format for some reason
#asahi · 2022-07-12 19:35
svencan you paste both reports somewhere?
#asahi · 2022-07-12 19:24
svenI don't have any BLE devices here so can't really debug that :/
#asahi · 2022-07-12 19:24
svenbut dunno
#asahi · 2022-07-12 19:24
svenhrm, or maybe these BLE events just have a weird format in this controller
#asahi · 2022-07-12 19:17
svenI’m leaning a bit towards userspace fail and not just because it’s not my fault then: I can see LE events in btmon
#asahi · 2022-07-12 19:16
svenyeah, this either LE userspace fail or some magic incantation is required to make the controller actually support LE
#asahi · 2022-07-12 18:31
svenbut my Bluetooth low level knowledge is almost non existing so dunno what that means
#asahi · 2022-07-12 18:30
sventhe weird thing is that it seems to show up in btmon
#asahi · 2022-07-12 18:11
svenit could still be a driver issue (like some missing magic "enable LE" command) but i'd first like to confirm it's no userland problem
#asahi · 2022-07-12 18:10
sveni'm afraid i don't know enough about bluetooth to help
#asahi · 2022-07-12 18:02
svenhttps://wiki.archlinux.org/title/Bluetooth#Device_does_not_show_up_in_scan maybe? dunno
#asahi · 2022-07-12 17:58
svenr.e. clover[m]'s issue: i looked at the btmon log and devices do show up there, they're all bluetooth LE afaict though. this sounds like a userspace problem to me, and i'm not very familiar with the bluetooth stack there
#asahi · 2022-07-12 16:36
sven(Note that the output will contain lots of Mac addresses and other info you may or may not want to share publicly)
#asahi · 2022-07-12 16:34
sven can anyone run btmon before power and scan on?
#asahi · 2022-07-12 16:34
sventhat’s weird
#asahi · 2022-07-12 07:57
svenoh, unless you pinned an older version
#asahi · 2022-07-12 07:57
svenupdating dts should've also updated m1n1
#asahi · 2022-07-12 07:54
svenwrong device tree and/or outdated m1n1 then
#asahi · 2022-07-12 07:51
svenENOENT should only happen when OTP reading fails or when no firmware is found and in both cases it'll complain
#asahi · 2022-07-12 07:51
sveni'd rather take a closer look for error messages in dmesg
#asahi · 2022-07-12 07:51
svenand it should complain anyway
#asahi · 2022-07-12 07:50
sven-2 is ENOENT
#asahi · 2022-07-12 07:50
svenyou probably don't
#asahi · 2022-07-12 07:49
svenhow does it fail?
#asahi · 2022-07-07 20:16
svenwhy are you trying to load the dtb from grub?
#asahi · 2022-07-07 19:37
svenI don’t think that stuff is even remotely ready yet
#asahi · 2022-07-05 13:43
svenYou’ll also need to add the node to the device trees, see #asahi-dev this morning
#asahi · 2022-07-05 13:42
svenyeah, that’s also not the correct logic then but might work if the current one doesn’t :D
#asahi · 2022-07-05 13:32
svenoh, or maybe try to just set that m2m on ss reset thing to false. I think I messed up the logic there
#asahi · 2022-07-05 13:18
svenmuch easier for someone with a pro/max to figure that out

plus ›