#
✕ nick:kettenis
#asahi-dev · 2021-12-16 09:30
kettenisalso, I'm seeing a "read" packet for "device" 0 with "type" 0xe010 and I'm wondering if that provides some sort of identification of the device
#asahi-dev · 2021-12-16 09:24
kettenisjannau: the j314 touchpad coordinates seem to roughly match the j293 values I'm seeing
#asahi-dev · 2021-12-15 16:22
kettenismarcan: heh, I can probably write something up for the next report...
#asahi-dev · 2021-12-15 12:59
kettenisalso, where did you get the device IDs from? Did you just invent one?
#asahi-dev · 2021-12-15 12:57
kettenisbut I'm wondering whether German (QWERTZ) or French (AZERTY) work correctly without userland configuration
#asahi-dev · 2021-12-15 12:56
kettenismy Dutch ISO keyboard is to close to the Internation English ISO variant to make a distinction
#asahi-dev · 2021-12-15 12:54
kettenisand then there are language variants of ANSI and ISO
#asahi-dev · 2021-12-15 12:54
kettenisthere are basically three variants: ANSI, ISO and Japanese
#asahi-dev · 2021-12-15 12:53
kettenisapple keyboard layouts are a bit weird
#asahi-dev · 2021-12-15 12:52
kettenisthe other question is whether we need to add somephysical parameters for the touchpad in the device tree
#asahi-dev · 2021-12-15 12:50
kettenisjannau: no, the keyboard layout/language thing is one reason
#asahi-dev · 2021-12-15 12:21
kettenis_jannau_: what keyboard layout are you using?
#asahi-dev · 2021-12-15 11:35
kettenisso we're going with apple,asc-mailbox-v4 for the generic compatible for the mailboxes?
#asahi-dev · 2021-12-14 19:23
kettenisis *not* very explicit I mean
#asahi-dev · 2021-12-14 19:22
kettenisyes, but the device tee structure in linux is very explicit
#asahi-dev · 2021-12-14 19:21
kettenisbut for pci devices the match function checks vendor and device ID
#asahi-dev · 2021-12-14 19:21
kettenison arm64 this uses the device tree and the match functions basically check compatible strings
#asahi-dev · 2021-12-14 19:20
kettenisdrivers have a match function to see whether they should attach or not
#asahi-dev · 2021-12-14 19:20
kettenisthe idea is that drivers form a hirarchy that matches the bus hierarchy
#asahi-dev · 2021-12-14 19:19
kettenisdesigned by Chris Torek IIRC
#asahi-dev · 2021-12-14 19:19
kettenisalyssa: it's magic ;)
#asahi-dev · 2021-12-14 19:14
kettenisI think it is easier to read, but I may be a bit biassed ;)
#asahi-dev · 2021-12-14 19:13
ketteniss/write/wrote/
#asahi-dev · 2021-12-14 19:11
kettenisThat is the OpenBSD RTC driver that I write after studying the corellium code
#asahi-dev · 2021-12-14 19:11
kettenishttps://github.com/openbsd/src/blob/master/sys/arch/arm64/dev/aplpmu.c
#asahi-dev · 2021-12-14 19:08
kettenisI don't disagree ;) (we ripped out bluetooth support in OpenBSD years ago since the implementation wasn't up to our standards)
#asahi-dev · 2021-12-14 19:06
kettenisalyssa: yes, I have seen no code for the bluetooth part of the wifi device
#asahi-dev · 2021-12-14 14:55
kettenisand given how they operate, that also explains why the chip isn't publically available
#asahi-dev · 2021-12-14 14:54
kettenismakes sense; apple boasted about supporting high-impedance headphones on the m1 max/pro laptops in their marketing
#asahi-dev · 2021-12-14 14:30
kettenisso it's not just a simple quirk where they basically undone the banking?
#asahi-dev · 2021-12-14 13:58
kettenisthere are many chips that do it like that
#asahi-dev · 2021-12-14 13:57
kettenisyou mean register addresses?
#asahi-dev · 2021-12-14 11:16
kettenisgoes back to answering e-mails about VLBI
#asahi-dev · 2021-12-14 11:15
kettenissven: ah yes, $DAYJOB
#asahi-dev · 2021-12-14 11:13
kettenisand getting usable nvme out there will actually allow other people to look into it
#asahi-dev · 2021-12-14 11:12
kettenisI'd say the write cache buffer flushing slowness is something folks could live with for a while
#asahi-dev · 2021-12-14 11:11
kettenisheh
#asahi-dev · 2021-12-14 11:10
kettenissven: don't let perfect be the enemy of good enough
#asahi-dev · 2021-12-14 10:41
kettenis"raw" mode just extends the packet with touch information (which isn't really HID-compliant)
#asahi-dev · 2021-12-14 10:40
kettenisand it continues to be sent, even in "raw" mode
#asahi-dev · 2021-12-14 10:40
kettenisbtw, you're right, the "mouse" HID report is at the start of the packet
#asahi-dev · 2021-12-14 10:33
kettenis_jannau_: same here ;)
#asahi-dev · 2021-12-14 10:29
kettenisOpenBSD only flushes the disk cache before shutdown/powerdown basically
#asahi-dev · 2021-12-14 10:28
kettenisflushing the cache is defenitely expected to be "slower"
#asahi-dev · 2021-12-14 10:24
kettenissven: do those sync commands result in an NVMe flush command being issued?
#asahi-dev · 2021-12-14 10:13
kettenisbut the float values don't seem to provide more resolution
#asahi-dev · 2021-12-14 10:13
kettenissome values are available both as an integer and as a float
#asahi-dev · 2021-12-14 10:11
kettenisfunny thing is that most values are native-endian, but the remaining capacity seems to be reverse endian
#asahi-dev · 2021-12-14 10:10
kettenisI did a bit more exploring of the battery-related SMC keys yesterday and added stuff to the wiki page
#asahi-dev · 2021-12-14 10:02
kettenismarcan: maybe mention that basic M1 support (USB type-C only) has been upstreamed in U-Boot now as well and will be part of U-Boot 2022.01?

plus ›