#asahi-dev · 2026-08-10 12:22
jannauit would be silly to use merge commits for the asahi-installer-data repo. the merges are expected to contain a single fast-forward mergeable commit#asahi-dev · 2026-08-10 12:20
jannausobek[m]: github does a rebase for the "Rebase and Merge" merge strategy resulting in PR source and destination branches not being identical#asahi-dev · 2026-08-10 12:18
jannaudoes anyone know if it is possible to access files via the latest release#asahi-dev · 2026-08-10 12:17
jannauprod is only expected to lag behind for testing#asahi-dev · 2026-08-10 12:17
jannauI'm annoyed by the lack of support for fast-forward merges on github. This is very annoying for the asahi-installer-data repo with its main and prod branches which are expected to be mostly identical#asahi-dev · 2026-08-10 12:13
jannauI'd first look for updated dxvk requirements. I think it would be unlikely that hk drops features unintentionally and I'm not aware of intentionally dropped vulkan extensions#asahi-dev · 2026-08-08 20:43
jannauhttps://lore.kernel.org/linux-usb/20260611084845.GX2990@black.igk.intel.com/#asahi-dev · 2026-08-08 20:42
jannautry the thunderbolt pull request for 7.2#asahi-dev · 2026-08-08 17:06
jannauignore the modprobe, I combine it with the result of `make modules-cpio-pkg`#asahi-dev · 2026-08-08 17:05
jannausven: I use https://gist.github.com/jannau/a680545d6a1dc9e083f64a6e671c17c8 to generate an initramfs from alpine's mini rootfs. not sure if that has enough to mount an overlayfs to install other packages#asahi-dev · 2026-08-08 14:48
jannauI don't think tested tbt on anything but m1 based on irc logs and I've found no branches with t600x dts thunderbolt. I might remember macos tracing on m1 max#asahi-dev · 2026-08-08 14:15
jannauexcept for the strange issue with samsung ssds#asahi-dev · 2026-08-08 14:14
jannauiirc the old working branch did work on m1 max#asahi-dev · 2026-08-08 12:36
jannauI meant to close it and wrote already the comment but never sent it#asahi-dev · 2026-08-08 11:46
jannauat least for that we can depend on 64BIT#asahi-dev · 2026-08-08 11:45
jannauchaos_princess: sigh, sorry. tell the clanker that ARCH_APPLE is little-endian only and COMPILE_TEST is not hardware#asahi-dev · 2026-08-08 11:43
jannauand released https://github.com/AsahiLinux/m1n1/releases/tag/v1.6.1#asahi-dev · 2026-08-08 11:38
jannaum1n1 v1.6.1 tagged#asahi-dev · 2026-08-08 11:35
jannauisn't that just the short form for frame data mismatch#asahi-dev · 2026-08-08 11:34
jannauafaik netflix uses film-grain. but it's not fully required. the standard is explicitly written to allow HW decoders + filmgrain generation via GPU shaders#asahi-dev · 2026-08-08 09:14
jannaumy usb3.2 gen2 ssd is faster than that#asahi-dev · 2026-08-08 08:07
jannauI think so but hidden behind the mtp coproc#asahi-dev · 2026-08-08 07:58
jannauin 7.2 there's usb4stream as well which looks useful as well#asahi-dev · 2026-08-08 06:45
jannaunot happy with the artifact names: https://github.com/AsahiLinux/m1n1/pull/654 is better #asahi-dev · 2026-08-08 06:44
jannaubroken as in support doesn't exists yet, or?#asahi-dev · 2026-08-08 06:39
jannaugithub releases feel more natural to me#asahi-dev · 2026-08-08 06:31
jannauI guess we want to create github releases from the tags to attach the CI artifacts#asahi-dev · 2026-08-08 06:24
jannauI'll tag m1n1 v1.6.1-rc1 to test the release CI#asahi-dev · 2026-08-08 05:58
jannaufluster's GStreamer-H.265-VL2SL mostly works (137/147) on fedora with asahi-7.1.6-1 and avd-fw 0.1#asahi-dev · 2026-08-08 05:44
jannauhow is hevc not working? not at all or broken decoded frames?#asahi-dev · 2026-08-08 05:25
jannauchadmed: do you mean t8122 (M3)? av1 is M3 and later. hevc was working on t8112 (M2) when I tested the last time#asahi-dev · 2026-08-08 04:20
jannausven: https://pastebin.com/j92uYEXy#asahi-dev · 2026-08-08 04:20
jannaunice, except for the ping latency#asahi-dev · 2026-08-07 17:39
jannauI know my boot cpu switch PR but I don't think I'll take it. there's always the next release#asahi-dev · 2026-08-07 17:39
jannauis there anything outstanding m3 related which would be nice to have?#asahi-dev · 2026-08-07 17:37
jannauI plan to tag m1n1 v1.6.1 soon (to have it in fedora rawhide before f45 is branched next tuesday)#asahi-dev · 2026-08-07 16:25
jannau26 minutes#asahi-dev · 2026-08-07 13:36
jannauthe main endpoint for external displays is dptx#asahi-dev · 2026-08-07 13:35
jannauor is that the endpoint we use for edid? Apple's naming is a little confusing. there are multiple similar named things#asahi-dev · 2026-08-07 13:33
jannauno, it's used for dp-audio#asahi-dev · 2026-08-07 10:31
jannauyes, looks as expected for enabled SLC#asahi-dev · 2026-08-07 10:12
jannauI'm asking because it works differently on t8122 and t6030: https://github.com/AsahiLinux/m1n1/pull/653#asahi-dev · 2026-08-07 10:11
jannauyuka or chaos_princess: can you test if we enable the system cache correctly on m3 max. I test with https://github.com/FedeParola/memory-latency. with slc the latency for mem size 16 - 48MB should be lower than for hundreds of MB#asahi-dev · 2026-08-07 09:11
jannauthanks#asahi-dev · 2026-08-06 16:09
jannausupporting a single display using a single DP tunnel on a specific port might mostly work. more displays / ports work as long as there are display controllers. switching between dp alt mode and tunneling on the same port should be easy to support#asahi-dev · 2026-08-06 15:19
jannauTested-by: me#asahi-dev · 2026-08-06 12:46
jannauthe patch fixes it here on system-fw 14.5#asahi-dev · 2026-08-06 12:44
jannaugoes away after reverting "nvme-apple: Drop the PRP null check chicken bit"#asahi-dev · 2026-08-06 12:21
jannaunvme-isa-0000 Composite: +40.9°C#asahi-dev · 2026-08-06 12:21
jannauI think without the patches there is at least a ssd temp sensor. also not seeing that error