#
✕ nick:sven
#asahi-dev · 2022-07-06 20:32
sventhere's also only bluetooth firmware for "formosa"
#asahi-dev · 2022-07-06 20:32
svenyeah
#asahi-dev · 2022-07-06 20:31
svenor, well, can i call (?) that method from the bluetooth device?
#asahi-dev · 2022-07-06 20:30
svenyeah, but does that also apply to the Bluetooth device?
#asahi-dev · 2022-07-06 20:21
svenI think you can use acpixtract on that to get back to binary. I can also ask someone with a t2 Mac to use another tool to generate the dump otherwise
#asahi-dev · 2022-07-06 19:25
svenit doesn’t matter much since the t2 only comes with a single board type but I’d like to be sure we can’t just grab it from acpi without horrible hacks
#asahi-dev · 2022-07-06 19:24
svenmarcan: you understand acpi, don’t you? can you check https://github.com/Dunedan/mbp-2016-linux/blob/master/MacBookPro15%2C4/acpidump and confirm that there’s no acpi data for the Bluetooth function?
#asahi-dev · 2022-07-06 11:21
svenhuh, so apparently bluetooth addresses are meant to be stored in little endian (i.e. the opposite of mac addresses)
#asahi-dev · 2022-07-05 09:25
svenhuh, those are both acl and should work fine
#asahi-dev · 2022-07-05 09:16
svendo you know if this is in SCO (bidirectional, lower quality) or ACL mode?
#asahi-dev · 2022-07-05 09:09
svenhavent noticed them here
#asahi-dev · 2022-07-05 09:09
svenhmmm, I wonder what causes those drops
#asahi-dev · 2022-07-05 09:04
svenwith that it makes everything very choppy for me as well
#asahi-dev · 2022-07-05 09:04
svenstill try without DEBUG
#asahi-dev · 2022-07-05 09:02
svenit’ll spam a lot with that and slow everything down
#asahi-dev · 2022-07-05 09:02
sventhat’s probably because debug is on
#asahi-dev · 2022-07-05 09:02
svennice!
#asahi-dev · 2022-07-05 08:55
svenyou also should get more debug output, but dunno why that is broken for you
#asahi-dev · 2022-07-05 08:55
svenyes, but it would complain otherwise
#asahi-dev · 2022-07-05 08:54
svenok, then your m1n1 is probably too old and doesn’t forward the bt blobs yet
#asahi-dev · 2022-07-05 08:53
svenin the device tree
#asahi-dev · 2022-07-05 08:52
svenand does the bluetooth0 alias exist?
#asahi-dev · 2022-07-05 08:52
svendo you have a recent m1n1?
#asahi-dev · 2022-07-05 08:52
svenyes
#asahi-dev · 2022-07-05 08:45
svenyes
#asahi-dev · 2022-07-05 08:39
sventry Bluetoothcl first and if that also doesn’t work add #define DEBUG at the top
#asahi-dev · 2022-07-05 08:21
svenif you are running in macos the fw is in /usr/share/firmware/Bluetooth
#asahi-dev · 2022-07-05 08:20
svenpython -m firmware.bluetooth /path/to/firmware fw.tar
#asahi-dev · 2022-07-05 08:16
svenfair :D
#asahi-dev · 2022-07-05 08:15
sveni just need to be able to run the hv ;)
#asahi-dev · 2022-07-05 08:15
svenmy definition of useable is a bit different I think :D
#asahi-dev · 2022-07-05 08:11
svenI haven’t done that because I usually work off some rc instead of asahi because it makes my workflow a bit easier
#asahi-dev · 2022-07-05 08:10
svenjust look for the WiFi board type and c/p that
#asahi-dev · 2022-07-05 07:06
svenwrong DT entry should just result in the wrong firmware name and/or a wrong address
#asahi-dev · 2022-07-05 07:06
sventrue, sounds like I have the wrong OTP offset for the pro though
#asahi-dev · 2022-07-04 21:36
sventhe firmware for T2 is at least called .signed.bin
#asahi-dev · 2022-07-04 21:28
sventhere might be something fancier but that was the quickest way
#asahi-dev · 2022-07-04 21:28
svenhrm, or I think I actually patched the C code, don’t remember
#asahi-dev · 2022-07-04 21:27
sventooling is very fancy word for monkey patching the sync trace handler with if addr == whatever.. ;p
#asahi-dev · 2022-07-04 21:26
svenand I didn’t realize that it does that from user space yet
#asahi-dev · 2022-07-04 21:26
sveni wanted to figure out how it does firmware selection
#asahi-dev · 2022-07-04 21:26
svenpatched the OTP from the hv and suddenly noticed how it used the other blob
#asahi-dev · 2022-07-04 21:25
svenbar1 has four or so windows and bar2 has another one
#asahi-dev · 2022-07-04 21:25
svenand that strange pcie config space dance just configures the view we get into the chips address space
#asahi-dev · 2022-07-04 21:24
svenbf = beam forming which I guess was only introduced in later revisions
#asahi-dev · 2022-07-04 21:23
svendunno if you saw it, but one of the two calibration blobs is used depending on the chip stepping stored in their OTP
#asahi-dev · 2022-07-04 21:23
svendefinitely!
#asahi-dev · 2022-07-04 21:22
svenyeah, I almost didn’t look at what macOS does and just played with the hw until it worked
#asahi-dev · 2022-07-04 21:21
sveni never use that other doorbell
#asahi-dev · 2022-07-04 21:20
svenyup

plus ›