#asahi · 2021-12-22 20:27
marcanzimsneexh: it depends on both#asahi · 2021-12-22 12:23
marcanI disabled applesmc since upstream fails horribly on T2#asahi · 2021-12-22 12:23
marcanthe machines are just bad#asahi · 2021-12-22 12:23
marcanRedecorating[m]: the fans are automatically scaling properly, and it does this on macOS too#asahi · 2021-12-22 11:57
marcanamw: it should work on any version#asahi · 2021-12-22 11:55
marcan(holy shit these intel machines really are terribad at power consumption... even just idle with an external screen the fans are already annoyingly loud, since I guess that clocks up the GPU?)#asahi · 2021-12-22 11:55
marcanbuilding linux now...#asahi · 2021-12-22 11:54
marcanRedecorating[m]: I have a 4364b3 too on this T2 I'm borrowing, so let's see what I can find#asahi · 2021-12-22 03:47
marcanlet's wait until I try the T2 machine and see if I run into the same issue#asahi · 2021-12-22 03:47
marcanRedecorating[m]: weird, thanks for checking#asahi · 2021-12-21 19:05
marcanyou can run it as many times as you want#asahi · 2021-12-21 19:04
marcankettenis: the pcie devices script does a clean shutdown#asahi · 2021-12-21 19:01
marcanthanks!#asahi · 2021-12-21 19:00
marcanwould be nice if you can write that out somewhere, I assume it's not too complicated?#asahi · 2021-12-21 19:00
marcankettenis: awesome :)#asahi · 2021-12-21 19:00
marcanespecially once we have a couple examples#asahi · 2021-12-21 19:00
marcanso working out SMC features and adding subdrivers is good territory for less experienced folks to help out#asahi · 2021-12-21 18:59
marcanfwiw the API is very simple and there will definitely be some kind of debug interface#asahi · 2021-12-21 18:59
marcanI'm going to start with the critical ones (gpio, battery) and then we'll see about adding more features as we go along#asahi · 2021-12-21 18:59
marcanfan control another, temperatures another, etc#asahi · 2021-12-21 18:59
marcanbattery management would be another#asahi · 2021-12-21 18:59
marcangpio would be one, that's the wifi enable bit (among others)#asahi · 2021-12-21 18:58
marcanI intend to build it as subdrivers, so we can incrementally add features#asahi · 2021-12-21 18:58
marcanyes#asahi · 2021-12-21 18:58
marcanthat's coming up next after I spend tomorrow making this work on T2 too so I can submit it#asahi · 2021-12-21 18:58
marcanthat requires the SMC driver which does not exist yet#asahi · 2021-12-21 18:58
marcanyes#asahi · 2021-12-21 18:56
marcanso we can deal with that if/when it happens#asahi · 2021-12-21 18:56
marcaneither way brcmfmac already has several firmware version feature gate mechanisms and I just added yet another one#asahi · 2021-12-21 18:56
marcanpresumably firmware versions do change e.g. to fix security bugs, but I'd hope they don't introduce breaking changes#asahi · 2021-12-21 18:55
marcanor more than 1000 in the original apple tree#asahi · 2021-12-21 18:55
marcanthat's why it's like 170 files#asahi · 2021-12-21 18:55
marcan(even T2 ones)#asahi · 2021-12-21 18:55
marcanGlanzmann: the firmware is universal, it includes everything for every mac#asahi · 2021-12-21 18:54
marcannice :)#asahi · 2021-12-21 18:53
marcanI will be merging in updates and pushing to asahi later#asahi · 2021-12-21 18:53
marcanplease don't bother testing anything other than wifi on that one#asahi · 2021-12-21 18:53
marcanthe wifi branch is out of date WRT other patchsets like the touchpad#asahi · 2021-12-21 18:27
marcan(AFAIK)#asahi · 2021-12-21 18:27
marcanshould still work#asahi · 2021-12-21 18:27
marcanfirefox317: that's normal#asahi · 2021-12-21 18:27
marcanjmr2: sometimes it is set from the AP's country code too#asahi · 2021-12-21 18:18
marcanonce we confirm this isn't blatantly broken and I add T2 support though, I'll send an RFC to the maintainers anyway, see what they have to say#asahi · 2021-12-21 18:17
marcanwhether we're supposed to set that from somewhere or what#asahi · 2021-12-21 18:16
marcanI still need to take a look at country codes, not sure what the situation is there yet#asahi · 2021-12-21 18:16
marcanand yeah, some "general use / perf tests" would be good; I haven't really stressed it yet. it would at least validate that 1) performance is good with all the proper config files we use, and 2) it's reasonably stable, no weirdo random crash bugs#asahi · 2021-12-21 18:15
marcanghz, hopefully :p#asahi · 2021-12-21 18:14
marcanI moved the problem structure from stack to heap and I don't think there's anything overflowable on the stack left, so... I think that's probably you're still running the old one :)#asahi · 2021-12-21 18:11
marcanGlanzmann: thanks#asahi · 2021-12-21 18:11
marcan(assuming this isn't another bug)