#asahi-dev · 2022-06-06 07:50
marcanwonder why that happens#asahi-dev · 2022-06-06 07:50
marcanhuh... I guess the fix should be conditional on aarch64 then :/#asahi-dev · 2022-06-05 17:36
marcanduty cycle is more often a fraction but if it says nanoseconds then it's relative to period#asahi-dev · 2022-06-05 17:36
marcancorrect, I would say#asahi-dev · 2022-06-05 07:15
marcansure :)#asahi-dev · 2022-06-05 06:54
marcanI'll probably make m1n1 init this eventually since it probably makes sense for it to deal with this and let Linux just read the data#asahi-dev · 2022-06-05 06:51
marcanlet me see if I can de-uglify the init and make sure I didn't miss anything#asahi-dev · 2022-06-05 06:50
marcan(including all the hardcoded stuff that wasn't properly broken out)#asahi-dev · 2022-06-05 06:49
marcanI assume you duplicated the init for the p cluster onto the second one?#asahi-dev · 2022-06-05 06:46
marcanit's possible the init is still wrong#asahi-dev · 2022-06-04 21:16
marcanyou can look at the old applesmc driver for stuff, fan/etc things should be the same I think?#asahi-dev · 2022-06-04 21:15
marcanvery little is from macos/fieldtest, they have few actual references in the software#asahi-dev · 2022-06-04 21:15
marcanguess work, apple schematic references, experimentation#asahi-dev · 2022-06-04 21:14
marcanhttps://hub.marcan.st/t/smc-key-info.txt just scattered notes#asahi-dev · 2022-06-04 21:13
marcanjeffmiw: I have a big TXT of SMC key stuff, let me throw it somewhere#asahi-dev · 2022-06-04 15:56
marcannothing specific to the vuart but those things touched the general exception entry stuff#asahi-dev · 2022-06-04 15:56
marcanand related changes#asahi-dev · 2022-06-04 15:56
marcan(prior)#asahi-dev · 2022-06-04 15:56
marcanjannau: if you have updated recently, the gdbstub was merged, which touched a bunch of stuff. there was also the spinlock thing and a fix for that.#asahi-dev · 2022-06-04 13:28
marcanwill look in more detail later#asahi-dev · 2022-06-04 13:28
marcanand there are significant differences between pcores and ecores so it seems plausible that the units differ#asahi-dev · 2022-06-04 13:28
marcanso they're reading those and computing transposed arrays of that info#asahi-dev · 2022-06-04 13:28
marcanchadmed: some of the init constants do seem different, also those regs are arrays of pstates (8 bits each) and they are preceded in the full HV log (which I did not paste verbatim) by reads of 8 regs that seem to contain pstate info for 8 pstates#asahi-dev · 2022-06-04 13:22
marcancluster globals yes#asahi-dev · 2022-06-04 13:18
marcanlet me look closer#asahi-dev · 2022-06-04 13:17
marcanI might have the init wrong#asahi-dev · 2022-06-04 13:17
marcanyeah#asahi-dev · 2022-06-04 13:11
marcanalso some counters aren't working yet#asahi-dev · 2022-06-04 13:07
marcannot sure if all those magic constants are the same for everyone or calibration stuff calculated off of the values read or what#asahi-dev · 2022-06-04 13:07
marcanso that means the dynamic power per clock is the same, and due to static power, higher clocks actually save (a tiny bit of) energy#asahi-dev · 2022-06-04 13:06
marcanwhich makes sense, since those boost states all use the same voltage AIUI#asahi-dev · 2022-06-04 13:06
marcaninterestingly the energy counters seem to be the same for all the boost states of the pcores, in fact it peaks a bit higher at 2988 MHz#asahi-dev · 2022-06-04 13:02
marcanalso the pcore/ecore units seem to be different for some reason#asahi-dev · 2022-06-04 13:02
marcannot sure what the units are either#asahi-dev · 2022-06-04 13:02
marcanthis is for t8103/mac mini#asahi-dev · 2022-06-04 13:02
marcannot sure how much of that init is required, but some of it seems to be to get good numbers for the energy counter, I think?#asahi-dev · 2022-06-04 13:01
marcanchadmed: https://mrcn.st/p/vIb0vP7u#asahi-dev · 2022-06-04 11:32
marcanlet me give it 30 mins or something#asahi-dev · 2022-06-04 11:32
marcanclpc node is under pmgr so yeah, must be raw#asahi-dev · 2022-06-04 11:30
marcanchadmed: want me to take a quick look at that see if I find it quickly?#asahi-dev · 2022-06-04 11:30
marcanso maybe those are raw#asahi-dev · 2022-06-04 11:30
marcanactually hold on, CLPC keeps complaining and I don't *think* that hits PMP#asahi-dev · 2022-06-04 11:30
marcanand measure in accumulated joules I believe?#asahi-dev · 2022-06-04 11:29
marcanthose should give you better data than SMC#asahi-dev · 2022-06-04 11:29
marcanAIUI the SoC has internal power meters, which are probably readable through PMP if not directly#asahi-dev · 2022-06-04 11:28
marcanmaybe there should be an official-ish list of "things for folks to pick up"#asahi-dev · 2022-06-04 11:27
marcanalso reminds me someone needs to write hwmon, it's on my endless pile but that's the kind of thing anyone can pick up given the smc scaffolding already in place#asahi-dev · 2022-06-04 11:27
marcanchadmed: unsurprising tbh :)#asahi-dev · 2022-06-04 11:27
marcanwe have some folks doing unofficial debian images but nothing official that I'm aware#asahi-dev · 2022-06-04 11:26
marcannorb: if you're still a debian project member, I'd love to see official work on that end for integrating with these machines :) (even if just discussion at this stage)