#asahi-dev · 2025-08-07 12:23
amarioguysnklp: saw your messages, wanted to clarify something about SEP, for the purposes of Asahi, we only really particularly care about AP<->SEP communication, because the SEP's firmware itself is basically a black box we have no control over apart from firmware version (it's signed by Apple with no means of signing our own)#asahi-dev · 2025-08-07 10:19
reptI see, it was more of a interrogation/challenge than anything, thank you for your answers :)#asahi-dev · 2025-08-07 10:14
chaos_princesshonestly, just do a full reinstall, it would be easier#asahi-dev · 2025-08-07 10:14
chaos_princess's partition uuid the same#asahi-dev · 2025-08-07 10:14
chaos_princessyou are making things harder for yourself. m1n1 finds the next step via matching on partition uuids, so you need to somehow reinstall it while keeping the ESP#asahi-dev · 2025-08-07 10:10
reptYes, so currently I have a (broken) linux installed, and I would like to erase & replace it with a new fresh one. I am guessing that it would be possible to remove the EFI & Linux Filesystem partitions from macos, and re-setup only thoses two?#asahi-dev · 2025-08-07 10:01
chaos_princessi _think_ it is possible, but the procedure would be complex enough to not be worth it in most cases#asahi-dev · 2025-08-07 09:59
chaos_princessso you still have the m1n1 installed?#asahi-dev · 2025-08-07 09:56
reptunderstand that it is only the 1st stage that requires going through the recovery os step? sorry for the dumb questions, sill very new to this ')#asahi-dev · 2025-08-07 09:56
reptThank you for your response! I still have the same macos, I followed the Uninstallation guide here https://asahilinux.org/docs/sw/deleting-an-install/#this-is-for-default-vanilla-asahi-installs-these-steps-do-not-apply-for-custom-partitioning-setups, and reinstalled with the curl | sh installer. I just wanted to know if it is possible to reinstall only the parts that comes after the m1n1 stage 1 (I #asahi-dev · 2025-08-07 09:45
chaos_princessif you reinstalled the instance of macos that was used for original installation, it would be better to reinstall m1n1 too#asahi-dev · 2025-08-07 09:33
reptHello there :) I am currently looking around the Asahi project, and I found myself reinstalling the whole thing a few times (deleting the partitions MacOS, and going through the RecoveryOS workflow for the install). I was wondering if it was possible to reinstall only the Linux part, and keep the m1n1 part that requires booting to RecoveryOS? Love the project! Thanks!#asahi-dev · 2025-08-05 05:45
nickchanso that settles it#asahi-dev · 2025-08-05 05:45
nickchanthe values read do look sensible but changing them doesn't really work#asahi-dev · 2025-08-05 05:43
nickchanchaos_princess: i did try the cpu frequency stuff myself on t6000 yesterday there's a lock of some sort I think#asahi-dev · 2025-08-04 18:09
nickchanerm I mean cpu performance counter event description here, uncore remains undocumented#asahi-dev · 2025-08-04 18:09
nickchanalso the event descriptions *are* in /usr/share/kpep at least on macOS 15.5 so that doesn't require the optimization guide#asahi-dev · 2025-08-04 18:03
nickchanerm I mean a grand total of two known events obviously there are more events#asahi-dev · 2025-08-04 18:01
nickchanevent 2 seems to be reference cycle and event 3 seems to the mach_absolute_time()#asahi-dev · 2025-08-04 18:01
nickchannow I understand how the Uncore performance counters are supposed to work, but only a grand total of two events#asahi-dev · 2025-08-04 13:54
chadmedfedoras being ddosed#asahi-dev · 2025-08-04 13:53
nickchan*sigh* packit failed on pr 479 and that appears to be a server issue#asahi-dev · 2025-08-04 13:53
jannauto understanding at most unprocessed parts in the log/journal after the reboot#asahi-dev · 2025-08-04 13:48
flokli(scnr)#asahi-dev · 2025-08-04 13:48
floklionly one nibble at a time?#asahi-dev · 2025-08-04 13:48
flokli> the bug doesn't eat (much) data#asahi-dev · 2025-08-04 13:41
jannauthere's unfortunately no power fail signal so the capacitance is wasted. sudden system reset / hang should trigger it as well#asahi-dev · 2025-08-04 13:37
chadmedive been very nasty to my ssds over the past few weeks on 6.15.x and theyre still here#asahi-dev · 2025-08-04 13:34
chaos_princessso, should not affect us since laptops have batteries and mac minis have a ton of capacitance in the psus?#asahi-dev · 2025-08-04 13:32
jannauto my understanding the bug doesn't eat (much) data and needs a power fail at the "right" time to trigger#asahi-dev · 2025-08-04 13:27
fl0_id¯\_(ツ)_/¯ #asahi-dev · 2025-08-04 13:22
chaos_princessand here i am thinking that surely "btrfs will eat your data" is an ancient meme at this point, right? :P#asahi-dev · 2025-08-04 13:17
fl0_idanother reminder why my arch nas is on an lts kernel...#asahi-dev · 2025-08-04 13:11
j`eyasahi isnt on 6.15 currently#asahi-dev · 2025-08-04 13:10
oliverbI assume that this is that bug? https://www.reddit.com/r/linux/comments/1mdxsev/btrfs_bug_bites_a_bunch_of_fedora_users/ or am I wrong? #asahi-dev · 2025-08-04 13:08
oliverbDo I read this correctly as "might break your system"? #asahi-dev · 2025-08-04 13:07
oliverb"- btrfs bug not resolved but fix sent upstream:" #asahi-dev · 2025-08-04 11:11
nickchanseems to be okay on t6000 as well#asahi-dev · 2025-08-04 11:11
nickchanupdated the two pre-M1 m1n1 PRs and added a third one to address the -march issue found yesterday#asahi-dev · 2025-08-04 08:53
leioMaybe a list of devices and notes who owns one could be useful in some wiki/docs, so can find people to cover needed devices#asahi-dev · 2025-08-04 08:34
chadmedi cant imagine that mattering _that_ much anyway but ik there were cts issues between the pro and max so who knows#asahi-dev · 2025-08-04 08:33
chadmedoh yeah right that is wasteful#asahi-dev · 2025-08-04 08:33
chaos_princessnot worth it for a one-off thing#asahi-dev · 2025-08-04 08:33
chaos_princesseh, that is kind of wasteful, it is nonsense like m1pro with 14 gpu cores, m1pro with 16gpu cores and so on#asahi-dev · 2025-08-04 08:30
chadmedwe have the money, we should probably ensure that between us we have enough devices to cover all possible test matrices #asahi-dev · 2025-08-04 08:30
jannauI doubt we can cover all GPU core variants ourself but I'm unsure if we need to#asahi-dev · 2025-08-04 08:23
chaos_princessi have too few devices to cover the test matrix on my own, so i would, but if we can do it ourselves - then no need#asahi-dev · 2025-08-04 08:20
jannauI'm undecided if the verification is something we should ship to users#asahi-dev · 2025-08-04 08:18
jannauno pressing need. I still need to verify that the m1n1 init is identical to the existing driver one#asahi-dev · 2025-08-04 08:14
chaos_princessshould i rebase gpu init 2.0 to it, or no need for now