nicolas17 has quit [Quit: Konversation terminated!]
kov has joined #asahi-dev
balrog has quit [Quit: Bye]
balrog has joined #asahi-dev
Ariadne has quit [Read error: Connection reset by peer]
Ariadne has joined #asahi-dev
<marcan>
we need the atomic mailbox stuff to make SMC shutdown/reboot ops work properly :/
<marcan>
have I mentioned I hate mailbox yet?
<sven>
I think I’m staring to understand that now
alexsv has joined #asahi-dev
alexsv has quit [Ping timeout: 480 seconds]
the_lanetly_052___ has joined #asahi-dev
boardwalk has quit [Quit: Ping timeout (120 seconds)]
boardwalk has joined #asahi-dev
the_lanetly_052__ has quit [Ping timeout: 480 seconds]
<marcan>
at this point I'd almost consider dropping that mailbox driver, moving the register pokes to something near rtkit, and just having it follow the existing DT binding... especially if it takes more than a day or so to make this mess work with the RTKit refactoring
<marcan>
that thing is holding us back and provides, as far as I can tell, absolutely zero actual useful benefit/code at this point
<marcan>
poor abstractions are a code smell
<marcan>
I'm okay with us spending time yak shaving stuff like sleep, but there is no way this much pain is justified for a couple mailbox registers with dedicated IRQs and hardware FIFOs we aren't even using
<sven>
marcan: yeah, I’m gonna give work queues one more try
<sven>
if that doesn’t work I can at least send out rtkit the way it is now and we can “fix” then mailbox crap once you upstream SMC
<sven>
*that
boardwalk has quit [Quit: Ping timeout (120 seconds)]
boardwalk has joined #asahi-dev
user982492 has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
jluthra has quit [Remote host closed the connection]
jluthra has joined #asahi-dev
gladiac has quit [Quit: k thx bye]
gladiac has joined #asahi-dev
<kettenis>
kevans91: yes that needs more work to support suspend/resume and runtime power management
r_rei[m] has joined #asahi-dev
the_lanetly_052___ has quit [Remote host closed the connection]
the_lanetly_052___ has joined #asahi-dev
millenialhacker has joined #asahi-dev
<millenialhacker>
Hi, regarding missing hw support, who are items that are already WIP by other members?
jakebot has joined #asahi-dev
<j`ey>
rqou_: did some experiments with bluetooth and prores and jpeg, thats all I know of
<Jamie[m]1>
i'm making reasonable progress on AVD re
<j`ey>
ah cool
<sven>
i have a wip branch for usb3 support and made very early progress with thunderbolt
<sven>
but right now i first want to focus on getting rtkit/nvme upstream
<Jamie[m]1>
millenialhacker: is there a specific feature you're interested in?
<j`ey>
oh and jannau is working on dcp
<kettenis>
I was going to say that you're not allowed to work on new stuff before nvme has been upstreamed ;)
<sven>
:-P
<millenialhacker>
Cool, has anyone tried to get full power management?
<chadmed_>
i dont think anyone's had a real look at the webcams, SEP or neural engine
<millenialhacker>
mmm, webcam sounds look a good start, I'm not a HW engineer, but want to do something for learning and fun
<j`ey>
sven: dont listen to kettenis! do whatever you want!.. just make sure what you want to do is upstreaming nvme!
<chadmed_>
last i heard about PM was that some stuff we need to work is being blocked upstream by some neccessary changes to PSCI not being there yet
mps_ has joined #asahi-dev
<_jannau_>
dcp has a wip driver and Lina and Alyssa are looking after agx
<j`ey>
and phire for starting the agx driver side
mps has quit [Ping timeout: 480 seconds]
<sven>
j`ey: :D
<phire>
Lina is doing a great job taking my very WIP agx branch and progressing fowards
<_jannau_>
ChaosPrincess started looking at the touchbar
<j`ey>
phire: just want to make sure you get some credit :-)
<phire>
:)
<millenialhacker>
What's preventing nvme to be upstreamed?
<j`ey>
nothing
<j`ey>
I mean, nothing will stop it being upstreamed, just v2 needs to be sent out
mps_ is now known as mps
<sven>
and likely also v3 and possibly v4 after that ;)
<maz>
the main thing is to keep the versions coming once a week. once you reach the point where you have a couple of consequtive versions without any significant comment, the maintainer is likely to pick it up.
<millenialhacker>
That's a good technique, at work I called digestible CRs
<millenialhacker>
There is always rejection to big chunk of codes no matter what
<maz>
the volume doesn't really change here. it is about posting regularly new updates to the series, making sure it is always fresh in the reviewer's head.
<maz>
I'm definitely likely to drop series if the author only posts once per cycle, and I'm sure I'm not the only one.
<millenialhacker>
A function of volume and frequency I would say
PhilippvK has quit [Quit: No Ping reply in 180 seconds.]
phiologe has joined #asahi-dev
millenialhacker has quit [Quit: Konversation terminated!]
millenialhacker has joined #asahi-dev
the_lanetly_052___ has quit [Remote host closed the connection]
the_lanetly_052___ has joined #asahi-dev
<Jamie[m]1>
do we know anything about how the fingerprint sensor works, out of interest?
<Jamie[m]1>
i did some RE/exploitation of a Windows Hello fingerprint sensor a couple of years ago, it established a TLS-PSK connection between the sensor's STM32 and the SGX applet
<chadmed_>
we know the "true" biometrics are stored in the SEP and that there is communication directly between the two
<Jamie[m]1>
(over SPI via the windows driver)
<Jamie[m]1>
ah, so the SEP has hardware connection directly to the sensor, not through the CPU?
<chadmed_>
i havent seen anything to suggest that we can get useful data out of it without SEP however happy to be proven wrong
<matthewayers[m]>
I wonder if the fingerprint sensor is hard-wired to a secure enclave on a hardware level
<Jamie[m]1>
^^
<chadmed_>
it cant be since the bluetooth apple keyboard has one on it
<Jamie[m]1>
yeah tbh I wouldn't want to build a driver that encourages people to store their unchangable credentials anywhere but a secure enclave
<chadmed_>
unless it is somehow exposed transparently over bluetooth as a separate serial device to the rest of the keyboard or something
kameks has joined #asahi-dev
<matthewayers[m]>
It seems like it could be? That’s a good point
<chadmed_>
but that seems very insecure and clunky
<matthewayers[m]>
It probably encrypts the data and sends it over as a string of text like typing on a keyboard
<matthewayers[m]>
And whatever prompt asking for it will listen for input only from the fingerprint sensor “key”
<chadmed_>
no it doesnt do that
<sven>
the fingerprint sensor is on a SPI bus
<Jamie[m]1>
yeah that would be suprisingly hacking matthewayers
<Jamie[m]1>
*hacky
<Jamie[m]1>
ah sweet sven
<sven>
but it's married to the SEP and the CPU just forwards communication between them afaik
<chadmed_>
the fingerprint sensor only directly talks to SEP
<chadmed_>
yeah what sven said :D
<sven>
no
<sven>
the fingerprint talks to the CPU
<sven>
and the CPU talks to SEP
<sven>
but it's an encrypted channel afaik
<chadmed_>
yeah that
<sven>
presumably the magic keyboard one just has the same protocol but over bluetooth
<Jamie[m]1>
sven: cool, same as the one i've dealt with previously then
<sven>
note that i'm not 100% sure on everything here, I just briefly looked SEP
<sven>
I know that it's fingerprint sensor <--SPI--> CPU <--> SEP sbio endpoint but i don't know much about the protocol itself
<Jamie[m]1>
i'll have a quick look in case it's the same protocol i've already REd lol
<Jamie[m]1>
sven: i don't see any mention of fingerprint/touch.?id in ioreg, am i grepping for the wrong stuff or is it not in there?
<sven>
euhm, it has a weird name
<Jamie[m]1>
ah AppleSandDollar
<Jamie[m]1>
lol
<Jamie[m]1>
or mesa
<_jannau_>
it's mesa in the ADT
<Jamie[m]1>
alright it's definitely not the goodix protocol
<Jamie[m]1>
which makes sense because that was broken in ways apple definitely wouldn't have accepted lol
<matthewayers[m]>
Maybe we’ll have to wait until Apple releases a repair manual for the MBP and (hopefully) talks about Touch ID replacement
Shiz has joined #asahi-dev
<chadmed_>
not really. we know the address it's on, we know it's an spi device and we know what apple call it so we can figure it out by tracing what macos does with it when its use is required
<chadmed_>
my concern is just that it's pretty useless without SEP so without knowing how SEP deals with it theres no real point
xet7 has joined #asahi-dev
<chadmed_>
it would be exceedingly dumb to have touch id support before SEP and then have users storing their biometrics pretty much in the clear in main storage
Shiz has quit [Ping timeout: 480 seconds]
thevar1able has joined #asahi-dev
robinp has quit [Remote host closed the connection]
<sven>
dunno if I want to keep that apple_rtkit_send_message api change, that's left from my first approach that I then discarded
bisko has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
nicolas17 has quit [Quit: Konversation terminated!]
user982492 has joined #asahi-dev
user982492 has quit []
bisko has joined #asahi-dev
bisko has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
bisko has joined #asahi-dev
possiblemeatball has quit [Quit: left]
user982492 has joined #asahi-dev
<rqou_>
chadmed_: "i dont think anyone's had a real look at the webcams, SEP or neural engine" <-- geohot has done some work on neural engine independent of the asahi project