#asahi-dev · 2023-01-22 12:07
marcanthat's another reason to deprecate it#asahi-dev · 2023-01-22 12:07
marcanno#asahi-dev · 2023-01-22 12:06
marcanthe fan control keys match, that's the #1 thing that will share nicely#asahi-dev · 2023-01-22 11:48
marcanit doesn't have to be consistent other than a few very well defined and stable keys, it has to be useful.#asahi-dev · 2023-01-22 11:47
marcankettenis_: I'd just include both, flood users with info. repair people will love it, and it's cool. for stuff we don't have the info for, just roll with what we can guess.#asahi-dev · 2023-01-22 11:47
marcanplease do#asahi-dev · 2023-01-22 11:46
marcansven: works for me, we already have apple,dwc3 for this. might need a denylist in the generic dwc3 driver to make sure it doesn't probe on snps,dwc3 first.#asahi-dev · 2023-01-22 11:44
marcanit'd cause some divergence but that's fine#asahi-dev · 2023-01-22 11:44
marcanChaosPrincess: I don't think intel machines have anywhere near as many sensors as these; we could just have some defaults hardcoded in the driver and probe for existing keys like kettenis_ does for those#asahi-dev · 2023-01-22 11:43
marcan(of course we don't have that data for everything, but we do for some machines, and that's nice)#asahi-dev · 2023-01-22 11:43
marcanwhich varies from board to board#asahi-dev · 2023-01-22 11:43
marcanI'd rather name those after the power rail they drive or what that power rail is for#asahi-dev · 2023-01-22 11:42
marcane.g. MPMU buck #1 might be two totally different things on different machines, and naming it "MPMU buck #1" erases all information useful for end users#asahi-dev · 2023-01-22 11:42
marcankettenis_: some keys have a different meaning on different systems because they are abstractly attached to power rails#asahi-dev · 2023-01-22 11:31
marcanwe push back against those silly rules/guidelines :p#asahi-dev · 2023-01-22 11:31
marcanchadmed_: yeah the "don't build the hwmon_channel_info on the fly" thing falls under "lol subsystem maintainers think they can dictate how hardware works"#asahi-dev · 2023-01-22 08:53
marcanso we probably want to have it in the DT as separate includes much like the platforms themselves are handled#asahi-dev · 2023-01-22 08:52
marcanplus there are going to be combinatronics involved, e.g. you have certain SoC-tied things (like CLVR voltage rails on Pro/Max) and some machine-tied things (e.g. j314 vs j316 differences)#asahi-dev · 2023-01-22 08:50
marcanand could even gain new entries with firmware upgrades#asahi-dev · 2023-01-22 08:50
marcanit is tied to the platform and SMC firmware#asahi-dev · 2023-01-22 08:49
marcansince it is fundamentally not tied to the SoC at all#asahi-dev · 2023-01-22 08:49
marcanso I was thinking we want all those key/name mappings in the DT#asahi-dev · 2023-01-22 08:48
marcanI don't get it, that function doesn't do any pointer math at all...#asahi-dev · 2023-01-22 08:47
marcanare you actually giving it a pointer to the int to write to? :p#asahi-dev · 2023-01-22 08:46
marcana... null deref? for a *float conversion*?#asahi-dev · 2023-01-22 04:09
marcaneiln: nice!#asahi-dev · 2023-01-21 03:22
marcanbut that doesn't really buy you much#asahi-dev · 2023-01-21 03:21
marcanmaybe it's just "deliver as IRQ directly" (like the GIC management IRQ, bypassing AIC)#asahi-dev · 2023-01-21 03:21
marcanit doesn't make much sense either, it'd require a back channel from the CPUs to AIC which is implausible#asahi-dev · 2023-01-21 03:21
marcanI don't think I ever saw IRQ mode working#asahi-dev · 2023-01-18 05:55
marcanpovik: ahh, quantization noise?#asahi-dev · 2023-01-17 17:03
marcankinda surprising it's that jagged though, unless the input wasn't a sine wave?#asahi-dev · 2023-01-17 17:03
marcanpovik: nice! :D#asahi-dev · 2023-01-17 16:09
marcanlol ok#asahi-dev · 2023-01-17 16:09
marcanthough I still need to figure out rpm for the DARTs properly#asahi-dev · 2023-01-17 16:09
marcanbut that should work ish with the rpm stuff I think?#asahi-dev · 2023-01-17 16:09
marcansure, as long as we can still sanely power sequence it to off for power saving#asahi-dev · 2023-01-17 16:09
marcanoh lol#asahi-dev · 2023-01-17 16:08
marcancan you bring that up independent of any thunderbolt/hotplug/etc stuff?#asahi-dev · 2023-01-17 16:07
marcanyou mean have the DART depend on a clock supplied by PCIe?#asahi-dev · 2023-01-17 16:07
marcanhow would that work?#asahi-dev · 2023-01-17 16:06
marcaneither that or maybe trigger on IOMMU domain attach? not sure if that would work#asahi-dev · 2023-01-17 16:06
marcanalso, for the DART, vague idea I had was to play with runtime-pm somehow? like have the DART driver defer doing anything and start up in rpm sleep state, only to be woken up when the PCIe dependency needs it. or something something.#asahi-dev · 2023-01-17 16:05
marcanheh#asahi-dev · 2023-01-17 16:05
marcanthat intel chip should be what's in the dongle bridging to PCIe#asahi-dev · 2023-01-17 16:04
marcanno that tracks#asahi-dev · 2023-01-17 16:04
marcan< sven> pretty sure the second thing is a lie :D#asahi-dev · 2023-01-17 03:30
marcanbut not SPI (and then the driver would need a codepath to trigger the wakeup anyway)#asahi-dev · 2023-01-17 03:29
marcanIRQ behavior on s2idle is opt-in, right now we opt in all mailboxes under the assumption that mostly everything but SMC will be shut down anyway#asahi-dev · 2023-01-17 03:28
marcanyeah, wake on keyboard is not implemented