#asahi-dev · 2026-05-18 19:10
yukait should work on: m1, m2, m4, a18 pro#asahi-dev · 2026-05-18 19:10
yukaI added something to find the base#asahi-dev · 2026-05-18 10:41
yukaI decided 0x134000 was likely to be the offset#asahi-dev · 2026-05-18 10:40
yukathe addresses all seemed to end in 0x?340??#asahi-dev · 2026-05-18 10:39
yukaof course it's somewhat arbitrary what I decided to use as the base#asahi-dev · 2026-05-18 10:37
yukaand the base for whatever device you sent me the captures of would be 0x23d000000#asahi-dev · 2026-05-18 10:36
yukathe base for My M1 Pro Macbook would be 0x0292400000#asahi-dev · 2026-05-18 10:35
yukaTheoretically I think it should work on the M1 / M1 Pro as well (it will use the correct usb endpoints depending on the device version), but I couldn't test that yet (have no second linux mac to put my m1 pro into debugusb)#asahi-dev · 2026-05-18 10:33
yukaand symlinks /dev/m1n1 to it#asahi-dev · 2026-05-18 10:33
yukait allocates a pty#asahi-dev · 2026-05-18 10:33
yukaanyways, https://cyberchaos.dev/yuka/kisd/, given the right base, works on m4 mac mini (base=0x3c8000000) as well as macbook neo (base=0x348000000)#asahi-dev · 2026-05-18 10:32
yukais that the ok status code?#asahi-dev · 2026-05-18 10:32
yukaeverything = 0#asahi-dev · 2026-05-18 10:32
yukait ecohed something like LOCK = 0 DBMa = 0 DVEn = 0 [...]#asahi-dev · 2026-05-18 10:31
yukacouldn't get it to work#asahi-dev · 2026-05-18 09:37
yukathx#asahi-dev · 2026-05-18 09:35
yukaack#asahi-dev · 2026-05-18 09:34
yukaI guess this would be the same i2c things done by tuxvdmtool#asahi-dev · 2026-05-18 09:34
yukasven, you said you had m1n1 patches to make it put itself into debugusb?#asahi-dev · 2026-05-18 09:33
yukaHmm, I can't get the central scrutinizer to get my device into debugusb#asahi-dev · 2026-05-17 19:26
yuka(would be interesting if M3 is also bcdDevice = 4.00)#asahi-dev · 2026-05-17 19:25
yukabcdDevice = 1.20 vs bcdDevice = 4.00#asahi-dev · 2026-05-17 19:25
yukathis might be different between the device version#asahi-dev · 2026-05-17 19:25
yukaI'm seeing the same#asahi-dev · 2026-05-17 19:25
yukayes#asahi-dev · 2026-05-17 19:25
yukabut something with an offset#asahi-dev · 2026-05-17 19:25
yukathe addresses are not in there verbatim#asahi-dev · 2026-05-17 19:24
yuka~120 bytes#asahi-dev · 2026-05-17 19:24
yukano, the one before that#asahi-dev · 2026-05-17 19:22
yukabetween m4 and a18 pro I think I can get the debugusb addresses from the first response (portal 0 PAR [...])#asahi-dev · 2026-05-17 19:20
yukawrong chat...#asahi-dev · 2026-05-17 19:20
yukahttps://github.com/KiaraGrouwstra#asahi-dev · 2026-05-17 18:42
yukaAnd otherwise I will try to get it working on my other devices#asahi-dev · 2026-05-17 18:41
yukaI think that is a bug in the dockchannel uart code on the m1n1 side#asahi-dev · 2026-05-17 18:41
yukachainload does not work just like from mac os#asahi-dev · 2026-05-17 18:41
yukaAppreciated#asahi-dev · 2026-05-17 18:40
yukaI have read and write working on my neo now, to the point that proxyclient shell works#asahi-dev · 2026-05-17 18:39
yukasven: debugusb dumps for the m3 (/pro) would be appreciated#asahi-dev · 2026-05-17 09:58
yukaI guess I can check the dumps you sent me as an additional sample (what device was that?)#asahi-dev · 2026-05-17 09:58
yukaFor now I can only write something which works for the 3 devices I have (M1 Pro, M4, A18 Pro)#asahi-dev · 2026-05-17 09:57
yukasven: I have a hypothesis where to get the address, but I'd need more sample data to confirm it#asahi-dev · 2026-05-17 08:26
yukathe other is ep 3, portal 0x13#asahi-dev · 2026-05-17 08:25
yukaone is on ep 1, portal 0x01#asahi-dev · 2026-05-17 08:25
yukaprobably#asahi-dev · 2026-05-17 08:24
yukaso very possible I will get to the same conclusion as you#asahi-dev · 2026-05-17 08:24
yukafor writing to the serial, I'm still testing that on different devices#asahi-dev · 2026-05-17 08:24
yukathose successfully enable the serial. I took the hex from the wireshark dumps.#asahi-dev · 2026-05-17 08:23
yukahttps://cyberchaos.dev/yuka/kisd/-/blob/main/src/main.rs?ref_type=heads#L92#asahi-dev · 2026-05-17 08:22
yukaand still get output on endpoint 4#asahi-dev · 2026-05-17 08:22
yukaI reduced the number of commands to just 2