#asahi-dev · 2022-03-05 13:45
marcan#863: PRab = (flt , 0x85) 0.3517 <- that's 0.3W still going to the display somehow#asahi-dev · 2022-03-05 13:42
marcanlooking through SMC power counters now to see if I can spot some power hog#asahi-dev · 2022-03-05 13:42
marcananyway, turning stuff off I can get it down to 3.5W in m1n1, which is still kinda high but that's 21h battery runtime#asahi-dev · 2022-03-05 13:41
marcanalyssa: it has to work on all the machines, and have notch aversion#asahi-dev · 2022-03-05 13:40
marcanwhich I'm pretty sure DCP controls itself#asahi-dev · 2022-03-05 13:40
marcanthe GPIO is a gross hack we use for now#asahi-dev · 2022-03-05 13:40
marcanalyssa: no, everything DCP#asahi-dev · 2022-03-05 13:38
marcanalyssa: yes#asahi-dev · 2022-03-05 13:10
marcanI can tell you right now how much power the screen draws, actually#asahi-dev · 2022-03-05 13:10
marcanthese are two different things#asahi-dev · 2022-03-05 13:10
marcan>backlight#asahi-dev · 2022-03-05 13:10
marcan>screen#asahi-dev · 2022-03-05 13:07
marcanmaybe more since these are USB-C level measurements, there's going to be some loss and some power spent on trickle charging the battery#asahi-dev · 2022-03-05 13:07
marcananyway, with the deep sleep hack I would expect at least 24h of sleep battery life, which is enough to get you through randomly putting the laptop away for a day#asahi-dev · 2022-03-05 13:07
marcanit's purely an OS-level thing#asahi-dev · 2022-03-05 13:06
marcanDCP doesn't know about s2idle#asahi-dev · 2022-03-05 13:05
marcansince putting 3K onto the screen all the time can't be cheap, and in FB mode it has to keep refreshing from RAM, it can't know you didn't touch it#asahi-dev · 2022-03-05 13:05
marcanand then of course there's the screen, can't do much about that until the DCP driver is in place, but I bet that'll save a bunch of power too#asahi-dev · 2022-03-05 13:04
marcanletting the e-cluster go to sleep too saves us a bit more too#asahi-dev · 2022-03-05 13:03
marcanthat might make the case for doing an ugly hack in linux to support deep sleep for the time being, that's a significant win and should be trivial to just do as a throwaway patch#asahi-dev · 2022-03-05 13:02
marcanok, that's about 100mA = 2W gone there#asahi-dev · 2022-03-05 12:59
marcannow let me switch to light sleep#asahi-dev · 2022-03-05 12:59
marcanso smp_start_secondaries makes *no* difference to power usage, which is using deep sleep, so at least now we know that's pretty good#asahi-dev · 2022-03-05 12:58
marcanlet's see#asahi-dev · 2022-03-05 12:58
marcanso I'm getting about 5W power draw in suspend, which is not great. I wonder how much of that is the CPUs...#asahi-dev · 2022-03-05 12:47
marcanif there's nothing special to it other than them being applied in order#asahi-dev · 2022-03-05 12:47
marcansven: I would merge#asahi-dev · 2022-03-05 12:46
marcanbut we might just never want to do that, and leave full PCIe powergate for deep sleep#asahi-dev · 2022-03-05 12:46
marcanthough that would introduce a DT compat break if we ever remove it, for older kernels#asahi-dev · 2022-03-05 12:45
marcanwe could use *that* for pcie and I'd hope that doesn't lose PHY state#asahi-dev · 2022-03-05 12:45
marcanalso we have that power domain flag that says "never go below clkgate"#asahi-dev · 2022-03-05 12:41
marcanyeah#asahi-dev · 2022-03-05 12:41
marcanmaybe there's some idle state signaling that confuses things#asahi-dev · 2022-03-05 12:41
marcanbut who knows...#asahi-dev · 2022-03-05 12:41
marcanand is just straight through redriving#asahi-dev · 2022-03-05 12:41
marcanthough I want to say *that* doesn't have magic protocols#asahi-dev · 2022-03-05 12:41
marcanthe SS lanes do have redrivers#asahi-dev · 2022-03-05 12:41
marcanso yeah#asahi-dev · 2022-03-05 12:41
marcanwell... remember the usb2 nonsense...#asahi-dev · 2022-03-05 12:40
marcanso maybe it's time to bite that bullet#asahi-dev · 2022-03-05 12:40
marcanyeah :/#asahi-dev · 2022-03-05 12:39
marcanbut there's probably some way to just mark pcie as "don't do that"#asahi-dev · 2022-03-05 12:39
marcanbut I guess it makes sense#asahi-dev · 2022-03-05 12:39
marcanhonestly I'm surprised s2idle even tries to full sleep all devices like this#asahi-dev · 2022-03-05 12:38
marcanand then punt it to full system sleep which will happen via m1n1 anyway#asahi-dev · 2022-03-05 12:38
marcansven: if it does we can just stop powergating it for now, which we probably want to anyway for "normal" s2idle since people probably want wake-on-wlan?#asahi-dev · 2022-03-05 12:35
marcanand also figure out why brcmfmac outright blocks it#asahi-dev · 2022-03-05 12:34
marcanprobably apple-pcie just needs a huge pile of re-init ports en masse on resume?#asahi-dev · 2022-03-05 12:34
marcanwhich might be more painful, but maybe not *that* bad since all the scaffolding for suspend/resume should already be there?#asahi-dev · 2022-03-05 12:34
marcanyup