#asahi-dev · 2022-03-07 14:32
marcanwould be really cool to have the speakers working to some extent on all models#asahi-dev · 2022-03-07 14:32
marcanpovik: if you think it's good enough I could sneak it into the first release, plus the m1n1 change#asahi-dev · 2022-03-07 14:14
marcanfair enough#asahi-dev · 2022-03-07 14:08
marcanare the required settings represented somewhere in the ADT, or is it hardcoded?#asahi-dev · 2022-03-07 14:07
marcanpovik: m1n1 should probably do that if it's a one-time thing#asahi-dev · 2022-03-07 14:06
marcannice!!#asahi-dev · 2022-03-07 13:30
marcanlet me try mine#asahi-dev · 2022-03-07 13:30
marcanwhich is probably required per spec?#asahi-dev · 2022-03-07 13:29
marcanisn't the USB thing a billboard device?#asahi-dev · 2022-03-06 06:39
marcanbut if it's Power Management that might be one of the magical tickets to lower sleep power#asahi-dev · 2022-03-06 06:39
marcanI never found out if PMP is Power Management or Power Measurement#asahi-dev · 2022-03-06 06:39
marcanand then we can see if we can get power stats from it which would be cool#asahi-dev · 2022-03-06 06:39
marcanmight be a good time to write a PMP driver too, if only to supply these notifications - it's another trivial driver now that RTKit is in#asahi-dev · 2022-03-06 06:38
marcanmaybe it's okay if we just ignore this and assume the fe -> cpu -> (dcp runs and does whatever) -> !cpu -> !fe ordering makes it all work out even if both touch the same register#asahi-dev · 2022-03-06 06:37
marcanthe interesting thing though is that nominally cpu depends on fe#asahi-dev · 2022-03-06 06:37
marcanone test would be to read the PS register after unplug, and see if it has changed#asahi-dev · 2022-03-06 06:37
marcanand that might also mean we shouldn't be handling _fe *at all* if DCP is going to be doing that for us#asahi-dev · 2022-03-06 06:36
marcanthen I bet it's that#asahi-dev · 2022-03-06 06:36
marcanahh, yeah, _fe heh#asahi-dev · 2022-03-06 06:36
marcanmaybe DCP *itself* tries to talk to PMP via one of the backchannels#asahi-dev · 2022-03-06 06:36
marcanbut wait, does disp0 (not cpu) have that?#asahi-dev · 2022-03-06 06:35
marcanthat's unexpected since the notification is something done by the AP normally#asahi-dev · 2022-03-06 06:35
marcanwe should make that an explicit dep then#asahi-dev · 2022-03-06 06:35
marcannot running or anything?#asahi-dev · 2022-03-06 06:35
marcanbut it just needs to be powered on?#asahi-dev · 2022-03-06 06:35
marcanoh, interesting, huh#asahi-dev · 2022-03-06 06:34
marcanIIRC I labeled that based on what power domains get reported to PMP when it's up#asahi-dev · 2022-03-06 06:34
marcanhm, what do you mean?#asahi-dev · 2022-03-06 06:16
marcansorry...#asahi-dev · 2022-03-06 06:15
marcanjannau: argh, yeah, how did I miss that#asahi-dev · 2022-03-06 06:15
marcankov: you mean the defconfig in the kernel tree? yeah, that probably needs updating#asahi-dev · 2022-03-05 17:32
marcananyway, I need to get some sleep, and I think I'll disconnect for tomorrow... it's apartment cleanup day ;)#asahi-dev · 2022-03-05 17:32
marcanthe map is by index though, isn't it? IIRC#asahi-dev · 2022-03-05 17:25
marcando we even need to touch disp0?#asahi-dev · 2022-03-05 17:15
marcanwhich is configured in the DAPF, but we don't inspect that#asahi-dev · 2022-03-05 17:15
marcanthe offset that is added/ORed to DART IOVAs to yield RTKit DVAs#asahi-dev · 2022-03-05 17:06
marcanjannau: at the very least the DVA offset is different, if that isn't a DT prop yet#asahi-dev · 2022-03-05 14:34
marcanor maybe the RAM isn't configured to go into lower power states right now#asahi-dev · 2022-03-05 14:34
marcanmaybe PMP manages some of that, who knows#asahi-dev · 2022-03-05 14:33
marcananyway, dunno, will have to cross check this with osx and see if there's some magic that makes it use less power#asahi-dev · 2022-03-05 14:32
marcannope#asahi-dev · 2022-03-05 14:24
marcanwell, could be pcie? let me nuke that#asahi-dev · 2022-03-05 14:24
marcanthere's about 1W going to the SoC CLVRs, which also doesn't sound right for mostly being idle#asahi-dev · 2022-03-05 14:17
marcanthen again, the way I brute force powered things down via PMGR and the PMU GPIOs may not be correct :p#asahi-dev · 2022-03-05 14:17
marcanthat's 22h runtime#asahi-dev · 2022-03-05 14:16
marcanok, so the lowest I can get things is 3.3W, but there's like a whole 1W unaccounted for in the SPMU for some reason, so I don't know what that's about (assuming I guessed the SMC keys right, of course)#asahi-dev · 2022-03-05 13:47
marcanit was DISPEXT still on#asahi-dev · 2022-03-05 13:47
marcanaha!#asahi-dev · 2022-03-05 13:47
marcan#863: PRab = (flt , 0x85) 0.0#asahi-dev · 2022-03-05 13:45
marcanactually, not even, to disp0