#asahi-dev · 2026-01-16 22:22
jannauwithout thunderbolt support though from reading the commit message closely#asahi-dev · 2026-01-16 22:21
jannausven: as far as I'm aware yes. you could ask in #aarch64-laptops. slightly related support for the retimer/mux is already upstream: https://lore.kernel.org/linux-arm-msm/20250206-x1e80100-ps8830-v6-2-60b1e49cfa8d@linaro.org/#asahi-dev · 2026-01-16 20:36
jannau`clpc=0` in bootargs makes things slow as well#asahi-dev · 2026-01-16 20:34
jannauit can be very slow with tracing#asahi-dev · 2026-01-16 20:34
jannau\o/!!!#asahi-dev · 2026-01-16 18:45
jannauAsking if the optical cable will need OS support. I would have assumed cd321x handles it transparently#asahi-dev · 2026-01-16 18:44
jannausven: sorry. I was trying to ask a question#asahi-dev · 2026-01-16 18:39
jannauand share free space that way#asahi-dev · 2026-01-16 18:38
jannauoptical cables require OS support beyond what cd321x will do anyway#asahi-dev · 2026-01-16 18:32
jannauowc has shorter and less expensive optical cables: https://www.owc.com/eu/solutions/usb4-cables €112 for 3 meters#asahi-dev · 2026-01-16 18:16
jannauyou had so far no need a 50m long cable? https://www.bhphotovideo.com/c/product/1577012-REG/optical_cables_by_corning_cor_aoc_ccu6jpn050m20_thunderbolt_3_usb_c_optical.html/overview for only $480#asahi-dev · 2026-01-16 18:09
jannauthat was already added in v6.9#asahi-dev · 2026-01-16 17:55
jannauthe gpu on ultra devices needs it as well downstream dt has a workaround#asahi-dev · 2026-01-16 17:51
jannauadp has 2 power-domains in the device tree but no code for it unless devm_drm_dev_alloc takes care of it#asahi-dev · 2026-01-16 17:34
jannausimpledrm / and simplefb as well. I think for some downstream drivers we modified pmgr dependencies to avoid using more than power-domain#asahi-dev · 2026-01-16 12:17
jannauyeah! pipewire 1.4.10 with "Fix a regression in restoring volumes on nodes." will also fix the webcam on devices using gst pipewiresrc#asahi-dev · 2026-01-15 20:44
jannauI thought I took care of forward compatibility :( but I guess I forgot something in the DT#asahi-dev · 2026-01-15 14:07
jannauI agree depends on how close to a page boundary the original ADT lands#asahi-dev · 2026-01-15 14:02
jannauthe commit changed the ADT data which is unexpected#asahi-dev · 2026-01-15 14:01
jannauthe size differs because adt.build() can't replicate the exact number of trailing zeros#asahi-dev · 2026-01-15 13:55
jannaumacos hotplugs fixed#asahi-dev · 2026-01-15 13:45
jannauIASInstallPhaseList I suppose they are using the same thing on intel macs#asahi-dev · 2026-01-15 13:40
jannauwhy is there a plist with "EFI Firmware Update" in the ADT?#asahi-dev · 2026-01-15 13:40
jannaunot sure if fetching parts of the ipsw from apple's cdn is a good idea#asahi-dev · 2026-01-15 13:31
jannaushould be fine to use a template ADT from the ipsw#asahi-dev · 2026-01-15 13:25
jannaumaybe time to add a -m m1n1.adt test to the CI#asahi-dev · 2026-01-15 13:24
jannauhttps://github.com/AsahiLinux/m1n1/pull/532 I can test macos hotplug irqs later today as well#asahi-dev · 2026-01-15 13:19
jannauthat fixes at least the m1n1.adt issue#asahi-dev · 2026-01-15 13:18
jannaucan you quickly test https://paste.debian.net/hidden/94ccc61e#asahi-dev · 2026-01-15 13:17
jannaujust 3 missing underscores#asahi-dev · 2026-01-15 13:16
jannauthat's an easy fix#asahi-dev · 2026-01-15 13:13
jannau`python3 -m m1n1.adt adt.bin adt_broken.bin` with edc2178e5977 look wild. it embeds "pmgr-u8id" in the output because it's a property of the python object#asahi-dev · 2026-01-15 11:40
jannauif it's an issue in adt parsing adt.py should notice it#asahi-dev · 2026-01-15 11:40
jannauthanks, I noticed that tipd interrupts aren't reaching macos anymore but hadn't a cha chance to debug#asahi-dev · 2026-01-15 10:18
jannaucreated upstream-pre-review#asahi-dev · 2026-01-15 10:12
jannauI wonder if it would make sense to create a faux merge request for reviewing. I can push a v6.19-rc1 based branch to the asahilinux linux repo#asahi-dev · 2026-01-15 10:05
jannauintegralpilot: I'll take a look. sorry I didn't get to it before#asahi-dev · 2026-01-15 07:39
jannaukitlith: looks like it but do you see more than panvk and imagination doing that?#asahi-dev · 2026-01-15 07:02
jannaukitlith: I'm not sure how the virtio gpu device presents itself. I could be required to check for is_virtio and return false in that case (in addition to panvk_can_present_on_device())#asahi-dev · 2026-01-15 06:53
jannaukitlith: both panvk_can_present_on_device and v3dv_wsi_can_present_on_device look valid. I'm not sure if renderonly will be initialized for hk#asahi-dev · 2026-01-15 06:46
jannaukitlith: does can_present_on_device() with `device->bustype == DRM_BUS_PLATFORM` alone fixes the issue why you started to look into this?#asahi-dev · 2026-01-15 06:44
jannaukitlith: looking at it more supports_scanout and can_present_on_device are not related. supports_scanout should probably remain false#asahi-dev · 2026-01-15 06:40
jannaukitlith: I don't know if it is not supposed to work. I wouldn't be surprised if the honeycomb's wsi is incomplete#asahi-dev · 2026-01-15 06:35
jannaukitlith: see the can_present in src/broadcom/vulkan/v3dv_wsi.c, the kms and render device fd are in struct renderonly in agx_device#asahi-dev · 2026-01-15 06:28
jannaukitlith: does can_present even matter with `wsi_device.supports_scanout = false` in hk_init_wsi()?#asahi-dev · 2026-01-14 19:52
jannausee also https://asahilinux.org/docs/sw/m1n1-hypervisor/#asahi-dev · 2026-01-14 19:52
jannaugatallah: the m3 support isn't yet in a state where support in the curl | sh installer would be useful#asahi-dev · 2026-01-14 17:27
jannauespecially since it would require vendoring it into the kernel. Producing code seems to be orders of magnitude faster than compiling it#asahi-dev · 2026-01-14 14:54
jannaurewriting the versions macro in syn doesn't seem to affect the compile time much (if at all). within 1% (build time of asahi.o, around 24.3s) and I'm not motivated for statistics#asahi-dev · 2026-01-13 20:10
jannaunot sure if we actually need that or are just cargo-culting what macos / the adt does