#
✕ nick:_jannau_
#asahi · 2022-08-12 09:39
_jannau_I would not expect that you can notice a difference by using the a system for 30 minutes
#asahi · 2022-08-12 09:36
_jannau_reset the stats if you have to switch the governor
#asahi · 2022-08-12 09:35
_jannau_that might be a decision which needs a re-evaluation given apple silicon's efficiency. a more reactive governor and just limit the top frequency state is probably better
#asahi · 2022-08-12 09:29
_jannau_well as400[m] claimed that schedutil won't go into the 600MHz state and you claimed that conservative is better
#asahi · 2022-08-12 09:27
_jannau_so it's 89% of the time at 600MHz
#asahi · 2022-08-12 09:17
_jannau_schedutil might boost them if it detects interactive use so if you check the current state it might be never in the lowest state
#asahi · 2022-08-12 09:16
_jannau_for my it's mostly in the lowest state
#asahi · 2022-08-12 09:15
_jannau_mps: are you sure the performance cores never go into the lowest frequency state? Please check /sys/devices/system/cpu/cpu4/cpufreq/stats/time_in_state
#asahi · 2022-08-12 09:07
_jannau_chadmed was looking into providing information in dt to make schedutil's energy aware scheduling work
#asahi · 2022-08-08 16:35
_jannau_kit_ty_kate1: no although not running sway. Do you see in dmesg many SPI HID CRC check failed messages?
#asahi · 2022-08-08 16:34
_jannau_Arian[m]: you need a virtualized metal gpu to run macos with UI in a VM, so no and probably not any time soon
#asahi · 2022-08-05 12:52
_jannau_nvme dt nodes not yet upstrean
#asahi · 2022-08-05 12:08
_jannau_ah, I assumed using the dt from the built kerne;
#asahi · 2022-08-05 12:04
_jannau_usb (usb-c) works without dt entries?
#asahi · 2022-08-04 15:10
_jannau_NikolayNikolaev[m]: most likely configure page size as 16k
#asahi · 2022-08-03 08:43
_jannau_whynothugo: are you connected to a 2.4GHz wlan?
#asahi · 2022-08-03 08:32
_jannau_you have to get the pairing keys from mac os and update the linux ones, see https://wiki.archlinux.org/title/bluetooth#For_macOS
#asahi · 2022-08-03 07:42
_jannau_not sure if we want to or can support u-boot variable files on the esp. Our single use case is to boot "/EFI/BOOTAA64.EFI" from the matched ESP. anything else is not supported
#asahi · 2022-08-03 07:40
_jannau_I don't know yet if it will work
#asahi · 2022-08-03 07:36
_jannau_not for the case we want to support: boot "/EFI/BOOTAA64.EFI" from the matched ESP. the plan would be to build the variables for that on the fly
#asahi · 2022-08-03 07:27
_jannau_planned/ongoing work to change how distro-boot is implemented in u-boot. I want to look into implementing it using efibootmgr
#asahi · 2022-08-02 16:02
_jannau_pull request open since the end of march
#asahi · 2022-08-02 16:02
_jannau_the system jemalloc will not work since that uses the build page size of 4k as well
#asahi · 2022-08-02 15:59
_jannau_and I think the fd package use it's own jemalloc, so it has to be fixed as well
#asahi · 2022-08-02 15:58
_jannau_and I hoped for a moment alarm rebased and merged the rust fix which sits there for almost 2 months now
#asahi · 2022-08-02 15:54
_jannau_whynothugo: which distro? for asahilinux / arch linux arm there are afaik 2 issues whih remain unfixed
#asahi · 2022-08-02 15:49
_jannau_not sure if Leif is still available he was planning to go on vacation
#asahi · 2022-08-02 15:48
_jannau_maybe "-asahi" vs. "-rc7.asahi" is an issue
#asahi · 2022-08-02 15:37
_jannau_what's your concern? it's the based on tag names in https://github.com/AsahiLinux/linux/tags
#asahi · 2022-08-02 14:31
_jannau_yes, it should be cortex-x1 + a78, maybe with some custom qualcomm sprinkles
#asahi · 2022-08-02 14:29
_jannau_+we
#asahi · 2022-08-02 14:28
_jannau_why would limit the selection that way if qualcomm itself doesn't say it uses ARM Ltd. based cores
#asahi · 2022-08-01 10:52
_jannau_it also handles the keyboard
#asahi · 2022-07-28 07:34
_jannau_I checked with `git tag --contains $FIX`
#asahi · 2022-07-28 07:33
_jannau_and the commit applies cleanly against fedora 36 glvnd package and a rebuilt package fixes the issue
#asahi · 2022-07-28 07:32
_jannau_j`ey: if I'm not mistaken the fix is in no libglvnd release
#asahi · 2022-07-27 14:26
_jannau_janrinze: from earlier: "the entire ATC thing is too entangled with the thunderbolt mess and I'm not confident enough the DT can stay this way"
#asahi · 2022-07-27 13:25
_jannau_no, that PR is not compatible with with the way the t600x devicetree looks after adding m1 ultra support
#asahi · 2022-07-27 13:23
_jannau_it does not contain the necessary device tree changes for any t600x device
#asahi · 2022-07-27 13:20
_jannau_sigh, why didn't use more distinct marketing names for their tablet and laptop/desktop chips
#asahi · 2022-07-27 13:18
_jannau_janrinze: no, pushed that branch just last weekend, not really compatible with using it for 2 months, also that branch works only on m1 (non pro/max/ultra) devices
#asahi · 2022-07-27 13:07
_jannau_yes, those have to be redone on top of asahi
#asahi · 2022-07-27 13:04
_jannau_I should send out the refix for that
#asahi · 2022-07-27 13:04
_jannau_rquo has a PR for that https://github.com/AsahiLinux/linux/pull/22
#asahi · 2022-07-27 13:03
_jannau_but does not include changes for m1 pro/max/ultra
#asahi · 2022-07-27 13:03
_jannau_janrinze: https://github.com/jannau/linux/tree/atc-WIP-5.19 this is on top of asahi
#asahi · 2022-07-27 09:35
_jannau_mps: I did the benchmarks already I just need to look up the results
#asahi · 2022-07-27 09:35
_jannau_m1 max and m1 pro should not differ much wrt. cpu performance
#asahi · 2022-07-27 09:33
_jannau_mps: I can look for kernel compilation benchmarks comparing m1 max vs m2 tonight
#asahi · 2022-07-27 09:32
_jannau_mps: 8 high performnce 2 high efficiency cores

plus ›