00:01 <robinp> pipcet[m]: I'm guessing macOS turns off the backlight between brightness setting 1 and 0. My HP doesn't and its rather annoying.

00:26 <bloom> Dissecting the Apple M1 GPU, part III - https://rosenzweig.io/blog/asahi-gpu-part-3.html

00:26 <bloom> Get it while it's hot! (For those who are here but not #asahi-gpu. "Hot" being before one of you submits it to Hackernews 😋)

01:14 <marcan> \o/

04:53 <shenki> that's a great write up

05:47 <PthariensFlame[m> :tad

05:48 <PthariensFlame[m> * :tada:

05:48 <PthariensFlame[m> * 🎉

07:17 <pipcet[m]> robinp: if I'm reading the SMC return values correctly, turning off the backlight does save a lot of power, so that's good. What's not so good is it appears the Corellium code isn't sufficient to start charging the battery after the USB-C power cable is reconnected, although it does charge when booted with the power cable plugged in.

12:22 <bloom> thank you :)

12:49 <arnd> pipcet[m]: I'm fairly sure the Unix-like parts of MacOS use the normal 1970-based epoch, though there are still some traditional MacOS parts like HFS+ that use an epoch using unsigned integers starting in 1904, which means 32-bit time values are a little over 2 years off -- 1904 to 2040 instead of 1902 to 2038

12:50 <pipcet[m]> thanks, turns out it was nanoseconds since I got my mac rather than microseconds since the epoch. What's three orders of magnitude between friends? :-)

18:50 <pipcet[m]> I've documented a few of the SMC keys at https://github.com/AsahiLinux/docs/wiki/HW:SMC. I have a list of all the ones on my system here, obtained from the SMC itself, but I'm not sure whether it would be legal to post the entire list...

19:55 <kettenis> I don't see why that would be a problem

20:04 <kettenis> btw, gP0d is the key used to control power to the WiFi chip

20:06 <sven> i don't see any problem with posting such a list either. especially if it's easily obtained by asking the smc nicely

20:09 <pipcet[m]> doesn't give us a license to use it if Apple claim copyright on the names. I don't think they can, but I thought I'd ask to be safe

20:12 <kettenis> it is an interface

20:28 <pipcet[m]> any idea what gP09, gP0f, and gP14 are? They're also defined and reserved in the Corellium device tree, but they're not actually used, as far as I can tell.

20:30 <kettenis> gP09 is "function-reset" for the dp2hdmi-gpio

20:31 <kettenis> gP0f is "function-enable" for the stockholm node

20:32 <kettenis> gP14 is "function-redriver_enable" for the display-subsystem node

20:32 <pipcet[m]> ah, I can find gP0f, which...yes. No gP09 in my device tree, and no gP14 either, but I suppose they might exist on the Mini?

20:33 <kettenis> possibly; no hdmi on the laptops isn't it?

20:34 <pipcet[m]> I don't think so

20:35 <pipcet[m]> * I don't think I have HDMI, no