#asahi-dev · 2026-02-08 21:21
svenotherwise everything should be incremental unless you’re doing something by weird #asahi-dev · 2026-02-08 21:21
svenccache makes the built time tolerable if you change header files a lot #asahi-dev · 2026-02-08 21:19
svenjust as built-in #asahi-dev · 2026-02-08 21:19
svenuh, what? it’s gonna compile the same stuff #asahi-dev · 2026-02-08 21:18
svenI don’t care much about wifi and other stuff that needs firmware though #asahi-dev · 2026-02-08 21:17
svenModule support doesn’t make the build much faster I think, just more convenient because you no longer need a initramfs#asahi-dev · 2026-02-08 21:15
svenalso no need to bother with initramfs then#asahi-dev · 2026-02-08 21:14
svenbuilding on my m1 then takes maybe 5 minutes at worst when I touch the wrong header file#asahi-dev · 2026-02-08 21:14
svenccache and disabling module support is usually what I got for #asahi-dev · 2026-02-08 18:36
sventhe m1 atcphy is already upstream 😅#asahi-dev · 2026-02-08 18:01
svenhuh#asahi-dev · 2026-02-08 17:52
svenand 99% of what’s required for Thunderbolt #asahi-dev · 2026-02-08 17:52
svenno, the upstream code already has everything required for DP altmode#asahi-dev · 2026-02-08 15:08
svennice work! hopefully without rewriting half of atcphy ;)#asahi-dev · 2026-02-07 14:51
svencdc_acm is probably still superior, this is mostly useful for early bringup and bare-metal linux I think#asahi-dev · 2026-02-07 14:50
svenhaven't tested speed yet, and i think it's only available on the dfu port#asahi-dev · 2026-02-07 14:31
svennot even a special cable, anything that does usb2 is enough#asahi-dev · 2026-02-07 14:31
svenshould also work on m3/m4 if we hook up the pd controller via spmi #asahi-dev · 2026-02-07 14:31
svenbest part: no extra hardware required, i could enter that special mode from m1n1 on the target just fine :)#asahi-dev · 2026-02-07 13:53
svendoesn't read any more data that should be there but still progress#asahi-dev · 2026-02-07 13:52
sven\o/#asahi-dev · 2026-02-07 13:52
svenHello, world from t8103 via dockchannel-uart!#asahi-dev · 2026-02-07 13:52
sven$ ./target/debug/asahi-kisd#asahi-dev · 2026-02-07 13:02
svenI suspect that register still exists then and is just read-only in raw boot mode there#asahi-dev · 2026-02-07 13:02
svenis 0 the correct value?#asahi-dev · 2026-02-07 09:21
svensomeone with a m4 should check if we can write to those or if they already have the correct value#asahi-dev · 2026-02-06 21:55
sven*thought#asahi-dev · 2026-02-06 21:54
svensaw it earlier today and was confused for a moment as well because i though i had already fixed it :/#asahi-dev · 2026-02-06 21:44
svenlooks like the usb curse got to you now@#asahi-dev · 2026-02-06 20:08
svenand if this follow the gpu rules you need an open-source userspace with a stable UABI before the kernel module will be accepted#asahi-dev · 2026-02-06 20:08
svenyeah, and that#asahi-dev · 2026-02-06 20:07
svenfrom what i remember the current driver has quite a bit of hacks, especially regarding the iommu handling#asahi-dev · 2026-02-06 20:04
svenit's simple really: if you want to get this into the tree work on getting it upstream. it'll end up in our tree soon after it hits -next#asahi-dev · 2026-02-06 19:23
svennoopwafel: looks like it changed from a quick glance#asahi-dev · 2026-02-06 19:00
svenif someone is interested in taking over, rebasing on the latest -wip and working on making it upstreaming ready that would be great though#asahi-dev · 2026-02-06 19:00
svenwe'd be adding unmaintained code to our downstream tree which is precisely what we've been working against in the past months#asahi-dev · 2026-02-06 18:58
sveni don't think anyone has worked on the neural engine in quite a while#asahi-dev · 2026-02-06 18:27
svenbut who knows with m3#asahi-dev · 2026-02-06 18:26
sventhat's usually not caused by tunables fwiw#asahi-dev · 2026-02-06 17:35
svenat the pipehandler lock? or while waiting for some other bit#asahi-dev · 2026-02-06 17:35
svenhow exactly does it fail?#asahi-dev · 2026-02-06 17:03
sveni suspect it's some kind of bug workaround anyway#asahi-dev · 2026-02-06 17:03
svensometimes usb3 still works even if that thing fails#asahi-dev · 2026-02-06 17:03
svenor maybe something else changed around there#asahi-dev · 2026-02-06 17:03
svenit won't work with incorrect tunables iirc#asahi-dev · 2026-02-06 16:06
svenplease keep this discussion to #asahi since it's not related to development#asahi-dev · 2026-02-06 16:04
svengetting debug logs is not the bottleneck there#asahi-dev · 2026-02-06 16:04
svensomeone with reverse engineering knowledge needs to work on that, yes