#asahi-dev · 2025-04-06 11:20
nickchanmischa85: i do not think there is a timer in aic on 64-bit apple soc#asahi-dev · 2025-04-06 11:15
mischa85which was my first approach but the aic timer crashes as it's probably not enabled via pmgr#asahi-dev · 2025-04-06 11:15
svenbut first you need to figure out what currently causes wfi to terminate#asahi-dev · 2025-04-06 11:14
svenyou can configure it to just trigger an interrupt which will wake you from wfi#asahi-dev · 2025-04-06 11:14
svenwe use that in the hv as well#asahi-dev · 2025-04-06 11:14
sveni'm talking about the hardware timer#asahi-dev · 2025-04-06 11:13
mischa85to actually run the timer#asahi-dev · 2025-04-06 11:13
svenwhy do you need another cpu?#asahi-dev · 2025-04-06 11:11
mischa85the timer idea would probably work to fire an interrupt but only with smp working#asahi-dev · 2025-04-06 11:10
mischa85i can only say, the output is consistent#asahi-dev · 2025-04-06 11:10
mischa85sven: yes, i realized. so this is why this experiment will need to be done on another cpu and doesn't work with only the boot cpu#asahi-dev · 2025-04-06 11:10
nickchansven: i think interrupts are not used#asahi-dev · 2025-04-06 11:06
svenor wait.. do we these days?#asahi-dev · 2025-04-06 11:05
svenI don’t think we use interrupts for usb #asahi-dev · 2025-04-06 11:00
chaos_princessisn't the event literally the read64 in the next line that causes proxyclient to write to serial?#asahi-dev · 2025-04-06 10:53
svenif there's the HW magic that decides between retention/non-retention it will probably only move to non-retention if nothing happened for a longer time#asahi-dev · 2025-04-06 10:51
sventhere's probably still some event firing in the background, maybe some timer#asahi-dev · 2025-04-06 10:46
svenmischa85: uh, no. the point is to make wfi last longer but right now the core seems to exit that state pretty quickly again#asahi-dev · 2025-04-06 04:46
chadmedi wish i could attend#asahi-dev · 2025-04-06 04:46
chadmedinteresting development in the wireplumber space: apparently someone else has found a use for the DSP chain stuff we built and george and arun want to hack on it at the pw hackfest#asahi-dev · 2025-04-05 22:25
chadmedurgh waking up to a kernel build bot error email is probably the worst way to start a sunday#asahi-dev · 2025-04-05 21:21
mischa85but to be honest, i just hardly know what i'm doing#asahi-dev · 2025-04-05 19:58
mischa85even tho the bytes seem a little weird and i could not get the syntax correctly#asahi-dev · 2025-04-05 19:58
mischa85it always returns what goes in via x1#asahi-dev · 2025-04-05 19:58
mischa85https://github.com/mischa85/m1n1/blob/test/proxyclient/m1n1/sventest.py#asahi-dev · 2025-04-05 19:58
mischa85sven: i wrote this, i hope i understood you correctly#asahi-dev · 2025-04-05 17:15
mischa85i can just use CNTVCT_EL0#asahi-dev · 2025-04-05 17:15
mischa85oh wait#asahi-dev · 2025-04-05 17:07
mischa85_i cannot access the timer? read32(0x302580000) gives sync exception...#asahi-dev · 2025-04-05 12:09
sventhat means that the e-cores likely don't spin up either#asahi-dev · 2025-04-05 12:02
mischa85_this is where they fail: https://github.com/AsahiLinux/m1n1/blob/main/src/smp.c#L167#asahi-dev · 2025-04-05 12:00
mischa85_no, i could try to spin up the E-cores but they fail to do so, trying to do so on the P-cores just straight up crashes. that's why i set MAX_CPUS to 5#asahi-dev · 2025-04-05 11:58
svenhave you figured out how to start the other cores yet?#asahi-dev · 2025-04-05 11:56
mischa85_on it!#asahi-dev · 2025-04-05 10:45
svenx1 should be fine, that reg is just allowed to be discarded across function calls #asahi-dev · 2025-04-05 10:45
svenah, no, nvm#asahi-dev · 2025-04-05 10:45
svenor looking at https://github.com/AsahiLinux/m1n1/blob/5291d5c1ccee788086972dbad21ba9202be30b8d/src/utils_asm.S#L155 you probably want to use x18 instead of x1#asahi-dev · 2025-04-05 10:43
svenif state is never retained it always uses deeper sleep states#asahi-dev · 2025-04-05 10:43
svenif state is sometimes retained then there's some HW heuristic that decides if wfi uses deeper sleep states#asahi-dev · 2025-04-05 10:43
svenif state is always retained than wfi won't ever enter deep sleep#asahi-dev · 2025-04-05 10:42
svenmischa85_: (let's move back from -re): run m1n1 and write some code that prepare a timer interrupt or something, save all regs, set x1 to some magic value, run wfi, check if x1 still has that magic value, restore all regs or something like that#asahi-dev · 2025-04-05 09:56
svenmischa85_: also, I think m4 doesn't seem to require any chicken bits (or they're already set up in iboot before we get control)#asahi-dev · 2025-04-05 09:56
chadmedlatest passt seems to be broken with muvm again -.-#asahi-dev · 2025-04-05 08:22
svensee #asahi-re, i think cyc_ovrd just doesn't exist anymore on M4#asahi-dev · 2025-04-05 07:04
nickchanbut since ps-regs is not present no one really knows how to mess with power domains right now#asahi-dev · 2025-04-05 07:04
nickchanall i mean is disabling those power domains could possibly save power#asahi-dev · 2025-04-05 07:03
nickchanmischa85: i never said that#asahi-dev · 2025-04-04 23:25
chadmedi am deeply concerned about the fact that the upstream-approved tdm idle mask stuff ended up not working but i'll worry about that when it's worth worrying about #asahi-dev · 2025-04-04 23:24
chadmedi need to submit v4 of the codec patches upstream today without the tdm idle mask stuff. since that was the last controversial bit hopefully those get merged without a fuss#asahi-dev · 2025-04-04 23:24
chaos_princessNot like there is a ton of changes happening to those files