<sven>
it probes and the BARs are correctly mapped
<sven>
and now it can also construct the correct firmware filename (minus that weird MUR vs. USI part I still have to figure out)
<j`ey>
which driver?
<sven>
uh, the one I wrote. there's unfortunately nothing that's similar
<j`ey>
oh
<sven>
R already did some nice reverse engineering, essentially it's HCI over what apple calls "Apple Converged IPC"
<j`ey>
that's a shame!
<sven>
yeah :(
the_lanetly_052 has joined #asahi-dev
<sven>
I kinda hoped it would just be similar enough to some other broadcom stuff but there's no pcie broadcom bluetooth driver in linux
<sven>
it sounds like apple was involved in the design though
the_lanetly_052 has quit [Ping timeout: 480 seconds]
<sven>
M=RASP m=6.11 V=m <-- that looks like part of the wifi otp, doesn't it?
<sven>
marcan: ^--
<j`ey>
RASP is part of the wifi fw name
<j`ey>
hm, part of the .txt: brcmfmac4378b1-pcie.apple,shikoku-RASP-m.txt
<sven>
okay, looks like that's the bluetooth OTP then. unfortunately it doesn't seem to contain that "USI" or "MUR" string that's part of the bt firmware name
gabuscus has joined #asahi-dev
c10l has quit [Quit: Bye o/]
c10l has joined #asahi-dev
<marcan>
sven: yes, that's the wifi otp format
<marcan>
sven: V=m = MUR, V=u = USI
<marcan>
m = murata
<marcan>
u = no idea but probably starts with USI
<marcan>
it's the module vendor
<sven>
ohhhh…. duh
<marcan>
we should just have the renamer do s/MUR/m/ s/USI/u/
<marcan>
keep with the format, avoid hardcoding nonsense in the driver
<sven>
it booted almost immediately but it hated me because I didn't setup irqs correctly
<sven>
at least that confirms that this abomination actually requires the msi doorbell addressed to be configured inside some register in the BAR though
<j`ey>
sven: did you see f3ce8a68d07d8001b8f90a908cd2e9d904e6df72 ? in m1n1