#
#asahi-re · 2021-01-08 16:55
marcanbputil does.... lots of things
#asahi-re · 2021-01-08 16:54
Shizi've been looking at how bputil and the Preboot partition interact
#asahi-re · 2021-01-08 16:54
Shizaw
#asahi-re · 2021-01-08 16:53
marcanI started writing a hardware wiki page at some point and apparently screwed up and closed the tab
#asahi-re · 2021-01-08 16:52
marcananyway, off to sleep :)
#asahi-re · 2021-01-08 16:52
marcanbut probably based on the same platform, assuming that's ace2
#asahi-re · 2021-01-08 16:52
marcanso it could also be an actual apple custom part
#asahi-re · 2021-01-08 16:51
marcancould just be a package difference but that feels odd
#asahi-re · 2021-01-08 16:51
marcanrwhitby: keep in mind the TPS65987 has a different package (which I noticed yesterday), so I'm not so sure any more
#asahi-re · 2021-01-08 16:51
marcanpuhitaku: that was already the plan, shhh, don't spoil it
#asahi-re · 2021-01-08 16:47
Shizi extracted the big sur dyld cache
#asahi-re · 2021-01-08 04:31
davidrysk[m]maximus64: okay yeah looking at the strings in the usb kexts they probably are dwc cores
#asahi-re · 2021-01-08 04:21
rwhitbyInteresting that MBP teardown shows CD3217 but Mini teardown shows CD3218
#asahi-re · 2021-01-08 04:13
rwhitbyhttps://www.ti.com/lit/ug/slvuan1a/slvuan1a.pdf
#asahi-re · 2021-01-08 04:12
rwhitbywe think it may be the same as the TPS65987, so I expect the same host interface will work to talk to it
#asahi-re · 2021-01-08 04:09
davidrysk[m]nvm I do :)
#asahi-re · 2021-01-08 04:09
davidrysk[m]I don't see a CD3217 on the MBP
#asahi-re · 2021-01-08 04:00
rwhitbyShould be able to talk the TI host interface protocol through that I2C bus.
#asahi-re · 2021-01-08 03:59
rwhitbyhttps://mrcn.st/p/YjUy8ROY - I see the CD3217 USB-PD controller in there as i2c0 with the expected I2C addresses of 0x38 and 0x3F for the two ports.
#asahi-re · 2021-01-07 21:56
roxfanmake it an easter egg
#asahi-re · 2021-01-07 21:35
puhitakujn__: it'll be so fun to show off with Bad Apple :O
#asahi-re · 2021-01-07 20:03
diz3yjn__: loading animation \o/
#asahi-re · 2021-01-07 19:59
jn__... or with Bad Apple frames :)
#asahi-re · 2021-01-07 19:41
puhitakuso we might be able to overwrite the white apple with the fresh asahi ringo logo ... that's cool, i hope the bootloader passes the fb as-is
#asahi-re · 2021-01-07 18:51
marcanbut the *apple* that shows up before that, with the "entering options" thing, *that* is the bootloader
#asahi-re · 2021-01-07 18:51
marcanthat "boot menu" is just an OSX app
#asahi-re · 2021-01-07 18:51
marcanactually that boot menu isn't that
#asahi-re · 2021-01-07 18:35
puhitakui've been forgotten that the boot menu showed up in your broadcast *slaps forehead* sry for misunderstanding because of not having the real hardware for now
#asahi-re · 2021-01-07 18:26
marcanof course actually implementing the display controller so you can do things like plug in external displays will be more fun
#asahi-re · 2021-01-07 18:25
marcanyou just write to memory
#asahi-re · 2021-01-07 18:25
marcanpuhitaku: the bootloader sets up the framebuffer for us
#asahi-re · 2021-01-07 18:24
roxfanbut I only see it called with NULL o.o
#asahi-re · 2021-01-07 18:21
roxfanthere's osfmk/console/video_console.c
#asahi-re · 2021-01-07 17:59
puhitakui'm a bit curious on how to find parts like DMAs in the vast memory map. from the source code of xnu?
#asahi-re · 2021-01-07 17:58
puhitakuthose steps sound natural while taking control of LCDs must be a journey as we have to make so many parts clear like DMAs and LCD interfaces before getting a fb... even we ignore the GPU stack entirely for the first step
#asahi-re · 2021-01-07 17:47
marcanbut the usb/serial thing is useful to get I/O
#asahi-re · 2021-01-07 17:47
marcanthere is, in fact that will be my first hello world, because poking a pixel on the framebuffer is slightly less error prone than sending stuff out of serial
#asahi-re · 2021-01-07 17:44
BlueriseI mean, if we can get an early glass framebuffer, that's enough for most people
#asahi-re · 2021-01-07 17:44
Blueriseaaand if there's a glass framebuffer, it's probably not really necesasry anyway
#asahi-re · 2021-01-07 17:38
marcanI just think I can avoid it being a blocker for anyone
#asahi-re · 2021-01-07 17:38
marcanbut I will be making an open hardware design for it anyway, it's a thing that's nice to have
#asahi-re · 2021-01-07 17:38
marcan(other than me)
#asahi-re · 2021-01-07 17:38
marcanso if everything goes according to plan, nobody will *really* need a serial cable to help out with the kernel bringup
#asahi-re · 2021-01-07 17:37
marcanpuhitaku: my first step will be a serial bootloader, and the second step will be adding usb-device support for that, and then I'll add earlyprintk support for that to linux (this will also be useful on the rPi, FWIW)
#asahi-re · 2021-01-07 16:12
puhitakunice details, also thanks for that, now i understood the overview
#asahi-re · 2021-01-07 15:44
FanfweBut yeah, essentially, being able to get UART on the USB-C is totally independent from what runs on the T2 and/or M1.
#asahi-re · 2021-01-07 15:40
Fanfweso you need a level translation circuit
#asahi-re · 2021-01-07 15:39
Fanfweonly thing is that logic level on those pins is not the usual 5V or 3.3V TTL
#asahi-re · 2021-01-07 15:38
FanfweApple has implemented a special vendor specific set of messages in the USB-PD protocol so you can request those two general purpose pins to be used as TX and RX lines of the machine's UART
#asahi-re · 2021-01-07 15:37
FanfweType-C has 2 "general purpose" pins that can carry whatever signal you like

plus ›