#asahi · 2022-04-15 14:00
jannaujust crash dcp but not the actual display controller and you can turn the display off and on (not a serious suggestion)#asahi · 2022-04-15 08:32
jannauyou could also use qemu on the M1 to get a standard arm system#asahi · 2022-04-15 08:30
jannauthe inverse it not true#asahi · 2022-04-15 08:30
jannauusing packages compiled on a machine with 16kb page size on a system with 4kb page size should not be problem#asahi · 2022-04-15 08:28
jannaunote that not all arch linux arm packages work yet since it uses a page size of 16kb#asahi · 2022-04-15 08:27
jannauthe asahi linux reference distro is based on arch linux arm#asahi · 2022-04-15 08:26
jannauyes and it is well supported for that use case#asahi · 2022-04-14 17:09
jannauI care about it (my 4k display needs dp for 60 Hz, HDMI is only 30 Hz) but DCP needs work first#asahi · 2022-04-14 17:02
jannaudisplay port might not that hard but it depends on dcp#asahi · 2022-04-14 13:59
jannauwhat part of USB-C?#asahi · 2022-04-13 14:09
jannauno, chrome/chromium 102 is scheduled to be released end of may#asahi · 2022-04-13 10:54
jannauhttps://github.com/AsahiLinux/linux/releases/tag/asahi-20220310-5#asahi · 2022-04-13 10:54
jannaumps_: linux/asahi has tags used for the asahi linux package, appending those might be less confusing#asahi · 2022-04-13 10:49
jannauit's probably much closer to 5.18-rc1 than to 5.17. we will try to base on rcX tags going forward#asahi · 2022-04-12 19:14
jannauChainfire's arm64 16kb page size chromium patch is merged#asahi · 2022-04-11 18:43
jannaumore youtube mac studio storage module upgrade drama. nothing to interesting, swapping two 512GB modules works, using two 512GB modules (unknown vendors) in the same mac studio doesn't. not aware of the on package raw nand controllers#asahi · 2022-04-10 13:17
jannauI wouldn't be surprised if qemu-user has the same 4k/16k problem#asahi · 2022-04-10 09:37
jannauit's just not possible to use qemu+kvm to virtualize a Apple silicon machine capable of running macos or m1n1#asahi · 2022-04-10 09:36
jannauhectour[m]: no, kvm on asahi linux can be used to virtualize a standard aarch64 system#asahi · 2022-04-08 17:09
jannauLG_PAGE is only used in source files and internal headers in jemalloc#asahi · 2022-04-08 17:05
jannauI would hope that jemalloc's idea of the page size is strictly internal and does not bleed to its users#asahi · 2022-04-08 16:56
jannaunikwest[m]: if mariadb uses external jemalloc it should be fixed when https://github.com/archlinuxarm/PKGBUILDs/pull/1914 is merged#asahi · 2022-04-08 16:13
jannauthe rust change only works for packages with an embedded jemalloc copy#asahi · 2022-04-08 16:04
jannaujust disabling jemalloc will probably not be accepted, see how marcan built rust with a larger jemalloc page size in https://github.com/AsahiLinux/alarm-PKGBUILDs/commit/c2459a0ae6fc04b7fe98bb04f10795248eca949b#asahi · 2022-04-08 16:02
jannaunikwest[m]: there is https://github.com/AsahiLinux/PKGBUILDs but non-essential packages should not be added there but should be merged in https://github.com/archlinuxarm/PKGBUILDs#asahi · 2022-04-08 05:25
jannaugryfbane_[m]: is asahilinux-keyring still installed? I guess `pacman-key -l` doesn't list that key. modification time of /etc/pacman.d/gnupg/pubring.gpg is probably not helpful after --refresh-keys. `pacman-key --populate asahilinux` might fix the issue if the asahilinux-keyring is still installed. that doesn't help with finding the cause though#asahi · 2022-04-06 17:23
jannaucommand line diskutil https://github.com/AsahiLinux/docs/wiki/Partitioning-cheatsheet#asahi · 2022-04-06 17:17
jannauI would have hoped that 3rdparty distros look what asahi does and which exact software is used in the boot flow#asahi · 2022-04-06 17:14
jannauthe question is what u-boot does. m1n1 doesn't select anything. it has hard coded partuuid to chainload from and a hardcoded config to add the active ESP to the FDT. Only with the commit mentioned earlier u-boot reads the blessed ESP from the FDT and uses that as first boot partition#asahi · 2022-04-06 16:55
jannauthat doesn't have ESP selection and will boot from the first bootable partition on the nvme and then usb#asahi · 2022-04-06 16:41
jannauoh-teps[m]: u-boot works as expected, grub doesn't for some reason. is 725346d2-... listed in blkid in linux?#asahi · 2022-04-06 16:38
jannauas far as I know, it needs https://github.com/AsahiLinux/u-boot/commit/102777230b057efa00fe39c7e545e13ddead1177#asahi · 2022-04-06 16:38
jannauGlanzmann: which u-boot does the debian install use? the ESP selection works only with https://github.com/AsahiLinux/u-boot/tree/releng/installer-release#asahi · 2022-04-06 16:36
jannauoh-teps[m]: what is the last line u-boot prints? 'unable to read file ubootefi.var' is expected and not a problem#asahi · 2022-04-06 16:32
jannauoh-teps[m]: do you have any usb storage devices connected? what/how did you install?#asahi · 2022-04-06 14:05
jannauarnd: that's the asahi distro config. my config with most stuff built-in: https://www.jannau.net/asahi-t6002.config#asahi · 2022-04-06 11:30
jannaubecause apple used a hdmi 2.0 diplayport to hdmi conversion chip#asahi · 2022-04-05 21:49
jannauWhyNotHugo: rust is planned for the signature verification#asahi · 2022-04-03 14:26
jannauit did not, it moved from freenode to oftc for obvious reasons#asahi · 2022-04-03 13:01
jannauI assume for audio it's also a probe issue, i.e. the device is either there or not#asahi · 2022-04-01 18:30
jannaumps: still no contents? any helpful output? `git submodule sync && git submodule update -f --init`#asahi · 2022-04-01 18:25
jannaugit submodule update#asahi · 2022-03-30 17:57
jannauare you sure it's required on m1 pro/max? if so then it might be a limitation accel=hvf or qemu's implementation of it#asahi · 2022-03-30 17:54
jannauargh, it of course it works out of the box on m1 pro/max/ultra #asahi · 2022-03-30 17:50
jannauQEMU emulator version 6.2.0#asahi · 2022-03-30 17:50
jannauyes#asahi · 2022-03-30 17:50
jannau"qemu-system-aarch64 -machine virt,accel=kvm -cpu host -smp 16 -m 16384 ..." works for me#asahi · 2022-03-30 12:02
jannau`lsusb -v` should report whether the keyboard support it: "bInterfaceSubClass 1 Boot Interface Subclass"#asahi · 2022-03-30 12:00
jannauyes, it's most likely related to the boot protocol. that specifies a report format to avoid a full hid implementation in a bootloader/bios#asahi · 2022-03-29 19:22
jannaumulti core score remains mostly unchanged with CPU frequency scaling https://browser.geekbench.com/v5/cpu/13935568 looked like CPU cores were ~50% idle during the multi core test