#asahi-dev · 2023-03-04 15:12
marcan*strings*#asahi-dev · 2023-03-04 15:12
marcan813984 ./AppleMobileDispH13G-DFR.kext/Contents/MacOS/AppleMobileDispH13G-DFR#asahi-dev · 2023-03-04 15:12
marcanmarcan@raider:~/asahi/vendor-bins/Extensions/12.4 -$ wc -c ./AppleMobileDispH13G-DFR.kext/Contents/MacOS/AppleMobileDispH13G-DFR #asahi-dev · 2023-03-04 15:12
marcan3544864 ./AppleMobileDispH13G.kext/Contents/MacOS/AppleMobileDispH13G#asahi-dev · 2023-03-04 15:12
marcanmarcan@raider:~/asahi/vendor-bins/Extensions/12.4 -$ wc -c ./AppleMobileDispH13G.kext/Contents/MacOS/AppleMobileDispH13G #asahi-dev · 2023-03-04 15:11
marcanso I don't think so#asahi-dev · 2023-03-04 15:11
marcanit is *much* larger than the touch bar DFR driver#asahi-dev · 2023-03-04 15:10
marcanthere is a driver for disp0 without DCP#asahi-dev · 2023-03-04 15:10
marcanH13G is just the soc codename#asahi-dev · 2023-03-04 15:10
marcanor is this like a much older display controller?#asahi-dev · 2023-03-04 15:10
marcanoh huh... well then... err... isn't that what DCP drives behind the scenes?#asahi-dev · 2023-03-04 15:09
marcanso there's a good chance most of the work will be the MIPI junk#asahi-dev · 2023-03-04 15:09
marcanI would expect the actual display controller portion to be fairly dumb, if they only intended it for the touch bar (single plane, no scaling, etc.), though who knows#asahi-dev · 2023-03-04 15:08
marcanalyssa: does that sound reasonable?#asahi-dev · 2023-03-04 15:07
marcanChaosPrincess: ^#asahi-dev · 2023-03-04 15:05
marcanbut also given MIPI specs are unobtainium, divining MIPI-isms from other drivers is probably a good idea too#asahi-dev · 2023-03-04 15:05
marcan(1/3 of that is DSI because lol MIPI)#asahi-dev · 2023-03-04 15:04
marcanwhole driver is ~4400 lines#asahi-dev · 2023-03-04 15:04
marcaneven smaller, seems to only do MIPI#asahi-dev · 2023-03-04 15:04
marcankmb? TIL that's a thing#asahi-dev · 2023-03-04 15:04
marcanoh actually#asahi-dev · 2023-03-04 15:03
marcanquick search suggests rcar-du for a reasonable driver with MIPI#asahi-dev · 2023-03-04 15:01
marcanif MIPI panel stuff is involved (which it likely is if this is a bare metal MIPI PHY display controller), you probably want to look at other MIPI drivers and panel stuff for common code, and then you should end up with two drivers (MIPI controller + panel)#asahi-dev · 2023-03-04 15:00
marcanand seems to actively be getting patches#asahi-dev · 2023-03-04 15:00
marcandriver from 2016, not *too* ancient#asahi-dev · 2023-03-04 14:59
marcanokay yeah dcp is a bad choice and radeon is worse#asahi-dev · 2023-03-04 14:59
marcanI was going to suggest one of the more recentish bare metal mipi/whatever drivers#asahi-dev · 2023-03-04 14:57
marcanyes.#asahi-dev · 2023-03-04 14:56
marcan:D#asahi-dev · 2023-03-04 13:38
marcanchadmed: internet says the chiplets are just cache and everything else (GPU+media+phys+etc) is still in a single chip?#asahi-dev · 2023-03-04 13:35
marcanlots of non-GPU stuff#asahi-dev · 2023-03-04 13:35
marcankettenis: bottom left on that navi10 shot is PCIe PHYs, all the wide top/bottom white blocks are RAM interface, lighter blocks along the mid left are probably display PHYs, etc#asahi-dev · 2023-03-04 13:32
marcanI measure 56% GPU, 44% everything else#asahi-dev · 2023-03-04 13:32
marcanhttps://www.flickr.com/photos/130561288@N04/49411586768/in/photostream/ navi10 shot, the bulk of the center is obviously GPU but look at all that wide junk around the edges#asahi-dev · 2023-03-04 13:29
marcan(memory/SLC being most of that, but even just display+media stuff is ~half the size of the GPU)#asahi-dev · 2023-03-04 13:28
marcanon M1 Pro, memory+cache+AVE+scaler+APR are significantly larger than the whole GPU by area#asahi-dev · 2023-03-04 13:27
marcankettenis: you underestimate how chonky media encode/decode stuff, general memory interface, etc can be#asahi-dev · 2023-03-04 13:26
marcan(which sure is based on drm)#asahi-dev · 2023-03-04 13:26
marcantechnically not drm, accel framework :p#asahi-dev · 2023-03-04 13:25
marcanjust like there are several intel, AMD, and ARM drivers#asahi-dev · 2023-03-04 13:24
marcanthey know nothing about each other#asahi-dev · 2023-03-04 13:24
marcanthere is zero shared code, zero direct interaction#asahi-dev · 2023-03-04 13:24
marcanbut also the drivers literally have zero in common with each other#asahi-dev · 2023-03-04 13:23
marcandisplaylink.#asahi-dev · 2023-03-04 13:23
marcanChaosPrincess: people are already doing it.#asahi-dev · 2023-03-04 13:22
marcankettenis: we already have 2 display controllers on these machines (missing mipi-dsi for the touch bar) even#asahi-dev · 2023-03-04 13:22
marcanI'm tired of repeating it over and over...#asahi-dev · 2023-03-04 13:19
marcanand panfrost isn't under "arm"#asahi-dev · 2023-03-04 13:18
marcanit's /drivers/gpu/drm/{i810,i915}, without an "intel" level. Same with radeon and amd (though amd is actually amdgpu but whatever).#asahi-dev · 2023-03-04 13:18
marcannominally the next level after drm is the driver/module name