#asahi · 2022-11-24 15:56
marcanwe probably need to make the UCM config force set sane settings#asahi · 2022-11-22 15:47
marcanthanks :)#asahi · 2022-11-22 14:46
marcananything prior to the first official release is unsupported and you should probably reinstall or surgically upgrade manually#asahi · 2022-11-22 14:45
marcantobhe: not correctly, those are probably too old for the mechanism to work anyway#asahi · 2022-11-21 16:54
marcanthe underlying driver does not use either of those APIs#asahi · 2022-11-21 16:54
marcanGallium takes care of a lot of the common functionality and legacy GL features, but it's not "GL on top of GLES"#asahi · 2022-11-21 16:54
marcan(instead there's piglit, Alyssa only just got to trying it out)#asahi · 2022-11-21 16:53
marcanyou're probably confused because Lina and Alyssa keep tweeting about OpenGL ES compliance tests because official OpenGL compliance tests for older versions don't exist at all#asahi · 2022-11-21 16:53
marcanrowanG337: it's not an OpenGL ES driver, it's a Gallium driver#asahi · 2022-11-21 14:40
marcanbluetail: Gallium is like 6 years older than Metal so no#asahi · 2022-11-19 06:32
marcanyou can use two monitors over TB3 though (on M1 Pro/etc, and *possibly* with hacks and limitations on M1 but it's unclear), but not as MST, as separate tunneled DP streams#asahi · 2022-11-19 06:32
marcanchadmed: it doesn't sound like the firmware supports MST, and quite likely nor does the hardware#asahi · 2022-11-18 17:42
marcanforce discharge as in force the battery management into "discharging" state by disabling inflow entirely#asahi · 2022-11-18 17:41
marcanforce discharge means AC input is blocked, it's not going to literally burn power lol#asahi · 2022-11-09 10:55
marcanthere are font rendering settings in the KDE system settings#asahi · 2022-11-09 01:17
marcanLinux has better VFS than macOS in general AIUI#asahi · 2022-11-08 05:25
marcanDCP should give us the metadata we need to handle that properly, if userspace plays ball#asahi · 2022-11-08 05:25
marcanthat's entirely likely. I think Apple uses wider gamut framebuffers by default, and we have no way of expressing that in simpledrm so it'll just pretend it's sRGB.#asahi · 2022-11-01 17:46
marcanmacOS->asahi yes (bless), asahi->macOS not yet (needs nvram tooling)#asahi · 2022-11-01 17:46
marcannot right now#asahi · 2022-11-01 17:31
marcanno idea, but there's vfio_iommu_type1.allow_unsafe_interrupts=1#asahi · 2022-11-01 17:30
marcanvfio stuff is complicated, little chance it's going to magically work without someone spending time on it#asahi · 2022-11-01 17:30
marcanthat's at the very least one issue#asahi · 2022-11-01 17:29
marcansven: there is an IOMMU but there is no interrupt remapping, or at least we don't expose it I think (maybe APCI could do it?), so all the vfio stuff will fail by default#asahi · 2022-11-01 17:19
marcanyeah, it only looks for *an* ESP by looking at common mountpoints, so it's the vague equivalent of what I do in mount_boot_esp, but it's not mount_sys_esp (which is the more important one)#asahi · 2022-11-01 17:15
marcanWhyNotHugo: no, but it seems highly unlikely to work properly with our setup#asahi · 2022-11-01 14:32
marcanwhich apparently systemd doesn't provide? so you still need base#asahi · 2022-11-01 14:32
marcanbut also, these scripts expect a shell and stuff#asahi · 2022-11-01 14:31
marcanthe pre hook probably needs to run before systemd-udevd.service#asahi · 2022-11-01 14:30
marcanand /new_root is called /sysroot in the systemd world#asahi · 2022-11-01 14:29
marcansounds like the posthook stuff needs to run after initrd-fs.target is reached#asahi · 2022-11-01 14:26
marcanyes, and be ordered correctly (which is subtle)#asahi · 2022-11-01 14:25
marcanit literally replaces /init with systemd#asahi · 2022-11-01 14:25
marcanand makes it useless#asahi · 2022-11-01 14:25
marcanyes, but systemd overrides it#asahi · 2022-11-01 14:24
marcanbut right now that doesn't exist, so unless you want to write it... :)#asahi · 2022-11-01 14:24
marcansure, and we can provide a unit file to make it work with it#asahi · 2022-11-01 14:20
marcanWhyNotHugo: systemd in the initrd is unsupported, sorry, you're well on your own here#asahi · 2022-11-01 14:20
marcanit's right there in the docs#asahi · 2022-11-01 14:20
marcan"Other hooks with runtime components will need to be ported, and will not work as intended"#asahi · 2022-11-01 14:20
marcanyou are throwing away our hook (and the debug stuff, and everything)#asahi · 2022-11-01 14:19
marcanyeah that's not going to work#asahi · 2022-11-01 14:19
marcanthe systemd stuff *replaces* base?!#asahi · 2022-11-01 14:19
marcanoh *facepalm*#asahi · 2022-11-01 14:18
marcanthe hooks#asahi · 2022-11-01 14:18
marcanWhyNotHugo: what is your mkinitcpio config?#asahi · 2022-11-01 14:16
marcanis supposed to show you the initramfs stuff#asahi · 2022-11-01 14:16
marcanjournalctl -b 0 -t initramfs#asahi · 2022-11-01 14:11
marcanthat worries me more#asahi · 2022-11-01 14:11
marcanbut again, why are you not getting debug logs?