#asahi-dev · 2026-05-21 07:35
yukaI remember it worked early on in the M1 days#asahi-dev · 2026-05-21 07:34
yukaDoes setting screen brightness always go via the dcp?#asahi-dev · 2026-05-21 03:47
yukaIt confirmed my previous guessing was quite close, but I 1. still don't know where to apply phy ip tunables (there seems to be no reg that fits and is writable) and 2. once I see the root bridge I don't know how to test if it is really working (no other devices visible yet, but they might still need to be enabled?)#asahi-dev · 2026-05-19 16:40
yukaor 700000, in the case of M4 Pro#asahi-dev · 2026-05-19 16:40
yukaPro/Max are regularly +400000#asahi-dev · 2026-05-19 16:39
yukaI'm starting to see a pattern with the kisd base addresses#asahi-dev · 2026-05-19 10:26
yukaThe + in ^_+? means 'afterwards' not 'at the same time#asahi-dev · 2026-05-19 10:25
yukaRight#asahi-dev · 2026-05-19 10:24
yukanope, but make sure it has power#asahi-dev · 2026-05-19 10:24
yukaHow do you do that on a keyboard?#asahi-dev · 2026-05-19 10:24
yukaright#asahi-dev · 2026-05-19 10:23
yukaflokli: I'm pretty sure your central scrutinizer is fine#asahi-dev · 2026-05-19 10:22
yukaAnd the serial itself is split out on ACM1#asahi-dev · 2026-05-19 10:21
yukaThen the help command is just a '?' on ACM0#asahi-dev · 2026-05-19 10:21
yukaIf you have problems sending the escape codes you can try https://cyberchaos.dev/yuka/cs-sw#asahi-dev · 2026-05-19 10:21
yukaThe call for help is ctrl+?#asahi-dev · 2026-05-19 09:53
yukaMaybe useful for debugging#asahi-dev · 2026-05-19 09:52
yukaBtw kisd also works on macOS if you kill deviceinterfaced before ;)#asahi-dev · 2026-05-19 08:10
yukaand not per-model#asahi-dev · 2026-05-19 08:10
yukaBut I'm assuming for now the base is per-chip#asahi-dev · 2026-05-19 08:10
yukabut not to debugusb :p#asahi-dev · 2026-05-19 08:10
yukaairpods reply to a nop#asahi-dev · 2026-05-19 08:09
yukathey all have the same base#asahi-dev · 2026-05-19 08:09
yukaI now got several samples (>=3) from devices with M1 chip, including one Macbook Air, one Macbook Pro and one iPad Air#asahi-dev · 2026-05-19 06:40
yukatrue#asahi-dev · 2026-05-19 06:40
yuka> Permission to asahilinux/kisd.git denied to yuyuyureka.#asahi-dev · 2026-05-19 06:37
yukayuyuyureka#asahi-dev · 2026-05-18 23:19
yukaCould be useful!#asahi-dev · 2026-05-18 23:17
yukayup#asahi-dev · 2026-05-18 22:54
yukaI'm open to moving it to a repo under github.com/AsahiLinux for better visibility#asahi-dev · 2026-05-18 22:54
yukaI have polished kisd a bit more (better logging, added a readme with known base values and usage)#asahi-dev · 2026-05-18 21:43
yukaserial is still a must-have for development (for example linux bringup)#asahi-dev · 2026-05-18 21:42
yukaWe did that two weeks ago on an M5 Max#asahi-dev · 2026-05-18 21:42
yukaEven USB should be working in m1n1 on M5#asahi-dev · 2026-05-18 21:24
yukaIt shows up as "leftover cature data"#asahi-dev · 2026-05-18 21:23
yukaHmm, wireshark does not properly show the payloads#asahi-dev · 2026-05-18 21:22
yukain the worst case they have a table somewhere in their driver#asahi-dev · 2026-05-18 21:22
yukaI'm tempted to test an old version of mac os as client with a relatively new chip#asahi-dev · 2026-05-18 21:20
yukaAs in you didn't find an obvious way to get the offset, or you saw no packets at all?#asahi-dev · 2026-05-18 21:19
yukaanything?#asahi-dev · 2026-05-18 21:19
yukaI have no idea how to get it from the other communication#asahi-dev · 2026-05-18 21:18
yukaI got this from a dump from flokli#asahi-dev · 2026-05-18 21:14
yukaon the m4 pro#asahi-dev · 2026-05-18 21:14
yukajannau: try --base 0x548700000#asahi-dev · 2026-05-18 20:17
yukabut if you know the base from a capture, you can specify it with --base#asahi-dev · 2026-05-18 20:17
yukaeither the base guessing code does not work for m3 and m4 pro (I don't have dumps from them available), or maybe there is just a mistake in the code#asahi-dev · 2026-05-18 20:16
yukathat means the base is wrong#asahi-dev · 2026-05-18 19:28
yukaI had an m2 which only wanted to enter debugusb with a reboot#asahi-dev · 2026-05-18 19:11
yukaI still need more test data to figure out what's going on there#asahi-dev · 2026-05-18 19:11
yukaIt probably wouldn't work for m1 pro