#asahi-dev · 2026-11-07 07:32
JamesCalligeros[m]Yeah it was busted on 7.2 as well, just havent been at my desk in a while :p#asahi-dev · 2026-11-07 07:22
JamesCalligeros[m]And i fixed it locally#asahi-dev · 2026-11-07 07:22
JamesCalligeros[m]Also no that commit did nothing by itself but it gave me a hint on what the problem was#asahi-dev · 2026-11-07 07:21
JamesCalligeros[m]jannau: ipv6 in general and mdns are both broken on ethernet only. Ran out of time to do more diagnosis but both work fine over wifi#asahi-dev · 2026-11-06 05:56
JamesCalligeros[m]It’s worth#asahi-dev · 2026-11-05 06:35
JamesCalligeros[m]Cosmic rays?#asahi-dev · 2026-11-03 00:02
JamesCalligeros[m]thanks! #asahi-dev · 2026-11-02 23:59
JamesCalligeros[m]Michael Reeves: would you mind sharing the FNV screenshots again? sorry.#asahi-dev · 2026-11-02 01:19
JamesCalligeros[m]Yay, i can send macaudio v4#asahi-dev · 2026-11-01 09:31
JamesCalligeros[m]Whatre we trying to do :p#asahi-dev · 2026-11-01 09:28
JamesCalligeros[m]Can we build it and just vendor it in the tarball#asahi-dev · 2026-11-01 06:37
JamesCalligeros[m]i believe something similar was reported with intel display engines so maybe it's a kwin bug?#asahi-dev · 2026-11-01 06:36
JamesCalligeros[m]i cant see anything we could possibly be doing wrong in the dcp driver though#asahi-dev · 2026-11-01 06:36
JamesCalligeros[m]unfortunately my hypothesis that the blending issues were a result of quantising the 8 bit alpha channel were incorrect#asahi-dev · 2026-11-01 06:34
JamesCalligeros[m]AhGR ended up Just Working so thats nice#asahi-dev · 2026-10-31 04:20
JamesCalligeros[m]but the bottommost layer should always be fully opaque aiui#asahi-dev · 2026-10-31 04:20
JamesCalligeros[m]could be related to the fact that we're trying to blend a surface with 8 bits of alpha against a surface with 2 bits #asahi-dev · 2026-10-31 04:19
JamesCalligeros[m]if you notice when stuff with alpha is on an overlay, the blending with the lower layers changes visibly#asahi-dev · 2026-10-31 04:18
JamesCalligeros[m]theres something sussy going on with blending in general that i havent been able to figure out#asahi-dev · 2026-10-31 03:52
JamesCalligeros[m]boy it sure would be nice to just be able to git push --force and fix this instantly for everyone involved and not worry about sending a v2 and cluttering up a mailing list....#asahi-dev · 2026-10-31 03:51
JamesCalligeros[m]oh and theres a better fix for the problem too! #asahi-dev · 2026-10-30 21:38
JamesCalligeros[m]\o/#asahi-dev · 2026-10-29 23:08
JamesCalligeros[m]there's corresponding strings in dcp fw too#asahi-dev · 2026-10-29 10:37
JamesCalligeros[m]Maybe when they make having an MBA illegal we will get company leadership that thinks like this#asahi-dev · 2026-10-29 06:23
JamesCalligeros[m]Wheres that screen geometry/occlusion protocol at these days#asahi-dev · 2026-10-29 06:22
JamesCalligeros[m]Genuinely wtf are they even doing#asahi-dev · 2026-10-29 06:22
JamesCalligeros[m]Ugh did anyone notice that the new mbp has rounded bottom corners too#asahi-dev · 2026-10-28 09:37
JamesCalligeros[m]i forget this sometimes too#asahi-dev · 2026-10-28 07:45
JamesCalligeros[m]Not#asahi-dev · 2026-10-28 07:45
JamesCalligeros[m]Why cant they just#asahi-dev · 2026-10-28 07:45
JamesCalligeros[m]Eyyyy#asahi-dev · 2026-10-28 06:38
JamesCalligeros[m]Its a proprietary isa with an architecture manual and standard extensions that are expected to be used. Apple are free to do whatever they want in their walled garden per their architecture licence, we however are not.#asahi-dev · 2026-10-28 06:12
JamesCalligeros[m]We will never enable AMX. It is a proprietary ISA extension that has no business in upstream Linux. Newer machines support SVE and SME, which are the replacements for AMX.#asahi-dev · 2026-10-27 21:41
JamesCalligeros[m]Is this still only failing on m3 max#asahi-dev · 2026-10-27 11:42
JamesCalligeros[m]fw crash?#asahi-dev · 2026-10-27 06:30
JamesCalligeros[m]Yeah#asahi-dev · 2026-10-27 06:28
JamesCalligeros[m]Or at the very least wilfully ignorant#asahi-dev · 2026-10-27 06:28
JamesCalligeros[m]That seems at least partially deliberate to me#asahi-dev · 2026-10-27 06:23
JamesCalligeros[m]As we are unfortunately all too aware of, sometimes the intention is to obfuscate and erase attribution in the first instance so yknow#asahi-dev · 2026-10-27 06:17
JamesCalligeros[m]Submodules exist for a reason#asahi-dev · 2026-10-27 06:17
JamesCalligeros[m]Yeah its silly#asahi-dev · 2026-10-27 06:16
JamesCalligeros[m]Hacking on m1n1?#asahi-dev · 2026-10-26 21:50
JamesCalligeros[m]Whered you get that m1n1#asahi-dev · 2026-10-26 10:30
JamesCalligeros[m](may be platform-specific)#asahi-dev · 2026-10-26 10:30
JamesCalligeros[m]not sure if the fix for it is gonna fly upstream though#asahi-dev · 2026-10-26 10:30
JamesCalligeros[m]its what i originally thought#asahi-dev · 2026-10-26 10:30
JamesCalligeros[m]lets go found the tas2770 issue#asahi-dev · 2026-10-25 22:33
JamesCalligeros[m]is there some power domain we're not switching on?#asahi-dev · 2026-10-25 22:32
JamesCalligeros[m]ok i feel a lot better if its just the max, but that is weird#asahi-dev · 2026-10-25 22:32
JamesCalligeros[m]sofus and a-ramses both have DT MRs that supposedly work