#
…

2022-12-21

00:00mort_ has joined #asahi-gpu
00:41swaggie has joined #asahi-gpu
00:46ChanServ changed the topic of #asahi-gpu to: Asahi Linux GPU development (no user support, NO binary reversing) | Keep things on topic | GitHub: https://alx.sh/g | Wiki: https://alx.sh/w | Logs: https://alx.sh/l/asahi-gpu
01:13Major_Biscuit has quit [Ping timeout: 480 seconds]
02:16pthariensflame has joined #asahi-gpu
02:16pthariensflame has quit []
02:49Dcow has joined #asahi-gpu
02:57Dcow has quit [Ping timeout: 480 seconds]
03:11zzywysm has quit [Quit: Textual IRC Client: www.textualapp.com]
03:15zzywysm has joined #asahi-gpu
03:26zzywysm has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
03:27zzywysm has joined #asahi-gpu
03:32zzywysm has quit []
03:32zzywysm has joined #asahi-gpu
05:00WindowPain_ is now known as WindowPain
05:42Hibyehello_ has quit [Ping timeout: 480 seconds]
05:53kode54 has quit [Quit: Ping timeout (120 seconds)]
05:53kode54 has joined #asahi-gpu
05:57SSJ_GZ has joined #asahi-gpu
06:07kenzie7 has quit []
06:08kenzie7 has joined #asahi-gpu
07:31Dcow has joined #asahi-gpu
07:35Major_Biscuit has joined #asahi-gpu
07:40Dcow has quit [Ping timeout: 480 seconds]
07:42MajorBiscuit has joined #asahi-gpu
07:44Major_Biscuit has quit [Ping timeout: 480 seconds]
08:17MajorBiscuit has quit [Ping timeout: 480 seconds]
08:19Dcow has joined #asahi-gpu
08:27Dcow has quit [Remote host closed the connection]
08:28Dcow has joined #asahi-gpu
08:51bluetail8 has quit [Ping timeout: 480 seconds]
08:58Dcow has quit [Remote host closed the connection]
09:12Dcow has joined #asahi-gpu
09:15linamkurz: Sorry about that, it's fixed now! ^^
09:19Dcow has quit [Remote host closed the connection]
09:24Dcow has joined #asahi-gpu
09:33Dcow has quit [Ping timeout: 480 seconds]
09:35mkurzlina: Thanks! ;)
09:35mkurz has quit [Quit: Leaving]
09:48Dcow has joined #asahi-gpu
09:48swaggie has quit [Remote host closed the connection]
09:50swaggie has joined #asahi-gpu
09:56Dcow has quit [Ping timeout: 480 seconds]
10:05swaggie has quit [Ping timeout: 480 seconds]
10:09jbowen_ has quit [Read error: Network is unreachable]
10:09jbowen_ has joined #asahi-gpu
10:14merry has quit [Quit: ZNC - https://znc.in]
10:38swaggie has joined #asahi-gpu
10:52swaggie has quit [Ping timeout: 480 seconds]
10:56cylm_ has joined #asahi-gpu
10:58iaguis has joined #asahi-gpu
11:01SSJ_GZ has quit [Remote host closed the connection]
11:01SSJ_GZ has joined #asahi-gpu
11:04swaggie has joined #asahi-gpu
11:20swaggie has quit [Ping timeout: 480 seconds]
11:34cylm_ has quit [Quit: WeeChat 3.6]
11:35cylm has joined #asahi-gpu
11:47Dcow has joined #asahi-gpu
11:58Dcow has quit [Ping timeout: 480 seconds]
12:20Dcow has joined #asahi-gpu
12:25swaggie has joined #asahi-gpu
12:28Dcow has quit [Ping timeout: 480 seconds]
12:34swaggie has quit [Ping timeout: 480 seconds]
12:42Dcow has joined #asahi-gpu
12:44jannaumarcan: can you tag mesa commits which are used in mesa-asahi-edge? old commits are garbage collected quickly after rebases
12:44marcanhm, guess I could push dated tags yeah
12:45marcan(don't have a better scheme)
12:46jannauyes, that's ok and matches the package version
12:50Dcow has quit [Ping timeout: 480 seconds]
13:01Dcow has joined #asahi-gpu
13:03swaggie has joined #asahi-gpu
13:11swaggie has quit [Ping timeout: 480 seconds]
13:13Dcow has quit [Ping timeout: 480 seconds]
13:37swaggie has joined #asahi-gpu
13:46swaggie has quit [Ping timeout: 480 seconds]
13:48Dcow has joined #asahi-gpu
13:57Dcow has quit [Ping timeout: 480 seconds]
14:23Dcow has joined #asahi-gpu
14:25swaggie has joined #asahi-gpu
14:31Dcow has quit [Ping timeout: 480 seconds]
14:33swaggie has quit [Ping timeout: 480 seconds]
14:57swaggie has joined #asahi-gpu
15:02Dcow has joined #asahi-gpu
15:06swaggie has quit [Ping timeout: 480 seconds]
15:06jbowen_ has quit []
15:06jbowen_ has joined #asahi-gpu
15:07jbowen_ has quit []
15:08jbowen has joined #asahi-gpu
15:09linajannau: I just found something interesting
15:09linaYou were wondering why 10-bit mode broke on the macbooks, right?
15:10linakCVPixelFormatType_30RGBLEPackedWideGamut = 'w30r', /* little-endian RGB101010, 2 MSB are zero, wide-gamut (384-895) */
15:10Dcow has quit [Ping timeout: 480 seconds]
15:10linaIt looks like that pixel format itself is inherently wide gamut, so what you need is another pixel format
15:13jannaulina: yes, I already noticed. dcp says the two other linear uncompressed 30-bit pixel formats are not supported
15:15linaMaybe you need to use the matrix stuff to correct it then?
15:15jannauwhich is strange since it is supported in the iboot ep
15:16linaYeah... maybe that endpoint automatically sets that conversion factor?
15:20jannaudcp starts with w30r with an identity matrix but macos probably uses wide gamut not that it matters much for the boot process
15:25marcanI have an idea
15:33LinuxM1 has joined #asahi-gpu
15:47LinuxM1 has quit [Quit: Leaving]
16:04marcanjannau: interesting. so it's not w30r, it's w18p. I dumped the surface descriptor from DCP memory.
16:05marcan2 planes, no idea what the second one is? it's 8bpp and has the bootup apple (and has had always had it with simpledrmfb and I don't see any apple-shaped artifacts so...)
16:05marcanhowever... using that and faking the second plane, KDE is still crunched.
16:06marcanso that's not it.
16:06marcanlet me see if I find any other unknowns around here...
16:07jannauI think the second plane is 8-bit alpha
16:10marcanif so the apple isn't quite opaque, which is interesting
16:10marcanI see 0xdf values
16:15jannauno. I was mixing it up with "b3a8" which is still wide range. pixel format descriptions dumped from macos: https://www.jannau.net/asahi/pixelformats_fmt.txt
16:16jannauno w18p but I guess the 'w' would indicate wide gamut as well
16:16marcanaand fixed it.
16:16marcanlet me clean this up.
16:18bcrumb has joined #asahi-gpu
16:18marcanwait no sec
16:19* marcanconfused himself
16:21marcannope, not that :(
16:35jannauany other interesting things set in the surface descriptor?
16:54marcannope... I think I have it copied exactly at this point
17:03jannaualso it is converted correctly on external displays as long as the color mode uses gamma-SDR as EOTF
17:05jannauthe only issue is that the single color mode on the mbp 14 is bogus since that seems to be only populated from edid
17:18marcanjannau: do you think it's possible iBoot issues a set_matrix? (that is a thing in the iboot interface too)
17:23iaguis has quit [Quit: leaving]
17:23jannauI would think
17:23jannauthat's unlikely
17:23jannaubut who knows
17:27systwi has quit [Ping timeout: 480 seconds]
17:27systwi has joined #asahi-gpu
17:32alyssanote to self: optimizing scissor_enable may improve perf slightly
17:32marcanjannau: so I tried the iboot interface from python and it works fine, the image looks normal (with w30r) after doing a swap
17:33alyssapass type has a huge impact on perf
17:33alyssanot that we can do much about that, but
17:33marcanohh. but.
17:34jannaumarcan: I guess you could try to set an identity matrix and see if that breaks it
17:36jannauI suspect it's the gamma SDR EOTF which makes it work
17:36Dcow has joined #asahi-gpu
17:44Dcow has quit [Ping timeout: 480 seconds]
18:03jannauany use for inverted or greyscale rendering? set_color_remap_mode does that
18:04jannauhuh, I completely missed splc_{get,set}_brightness
18:28bcrumb has quit [Quit: WeeChat 3.7.1]
18:53alyssabah, screw this, let's just play supertuxkart :-p
19:02marcanjannau: ok, so for w18p the extra plane is alpha, and the interesting thing is that we can't see it in the boot FB because there is some kind of optimization where if the plane covers the whole screen alpha is disabled
19:03marcanif I change the bounding box, the apple alpha mask shows up
19:03marcanbut also, the colorspace stuff is working properly in iboot
19:03marcan... so why isn't it working in the main dcp?
19:05Dcow has joined #asahi-gpu
19:06marcan... oh ffs.
19:07marcanjannau: it works on surface 1. it does not work on surface 0.
19:07marcansomehow.
19:07jannaudcp_iboot.py with 'IBootLayerInfo.colorspace = Colorimetry.DCIP3' looks similar
19:07marcanbut iBoot nominally uses surface 0, I *think*, though there is some weird allocation-ish stuff that might have me confused
19:14Dcow has quit [Ping timeout: 480 seconds]
19:19jannaurustc 1.66 is too new for rust4linux / asahi (yes I know 1.62 is the blessed version but 1.65 was working)
19:20marcanyes, we talked about that on the fedora channel
19:22jannauno surface[1] has still off colors for me with xfer_func = 13, colorspace = 2
19:26marcanjannau: colorspace=12
19:27marcanbut I don't understand why it works in iBoot with what I believe is surface 0 (unless this changed in a newer firmware, I'm looking at an old one)
19:27marcanI guess I should switch to 12.3 to double check...
19:27jannauI was testing the same with colorspace = 2
19:28marcanbut as far as I can tell in iboot it works with all planes, though I *think* I saw it glitch into bad colors at some point but I couldn't reproduce...
19:28marcanso maybe it's a firmware bug?
19:29marcanFWIW iBoot maps the EOTFs 1->13, 2->16 and the colorspaces 1->0, 2->12, 3->9
19:31marcanI guess it's possible we're missing some swap flags to make the colorspace properly effective? dunno, maybe I should see if I can find the iBoot swap struct in RAM too...
19:32* marcansleeps
19:32Dcow has joined #asahi-gpu
19:35alyssamarcan: nini
19:36jannaufix confirmed. what an annoying bug
19:37alyssalet's do uhhh
19:37alyssatransform feedback!
19:37alyssaeveryone's least favourite GL feature
19:40Dcow has quit [Ping timeout: 480 seconds]
19:46jannauwonderful, using surface[3] results in black screen on t6001
20:06Dcow has joined #asahi-gpu
20:14Dcow has quit [Ping timeout: 480 seconds]
20:15marcanjannau: I think only 0-2 are supposed to work?
20:22Gues__________________________ has joined #asahi-gpu
20:22Gues__________________________ is now known as evtyz
20:22dottedmag has quit [Quit: QUIT]
20:22evtyz has quit []
20:23Gues__________________________ has joined #asahi-gpu
20:23dottedmag has joined #asahi-gpu
20:25Gues__________________________ has quit []
20:33alyssaHmmm I wonder if I should be too clever for my own good
20:33alyssapanfrost's code here is dumb but I really don't want to copy it
20:36swaggie has joined #asahi-gpu
20:37alyssaoh I see why the panfrost code is silly
20:37alyssayeah let's not do that.
20:41Dcow has joined #asahi-gpu
20:50Dcow has quit [Ping timeout: 480 seconds]
20:55Dcow has joined #asahi-gpu
20:59Gues__________________________ has joined #asahi-gpu
21:00Gues__________________________ has quit []
21:08Dcow has quit [Remote host closed the connection]
21:09Dcow has joined #asahi-gpu
21:14c10l has quit [Quit: Bye o/]
21:15Dcow has quit [Remote host closed the connection]
21:15c10l has joined #asahi-gpu
21:15Dcow has joined #asahi-gpu
21:16c10l has quit []
21:16c10l has joined #asahi-gpu
21:24cylm_ has joined #asahi-gpu
21:25cylm has quit [Ping timeout: 480 seconds]
22:15mkurz has joined #asahi-gpu
22:32Dcow has quit [Remote host closed the connection]
22:48Dcow has joined #asahi-gpu
22:56Dcow has quit [Ping timeout: 480 seconds]
23:16Dcow has joined #asahi-gpu
23:18SSJ_GZ has quit [Ping timeout: 480 seconds]
23:25Dcow has quit [Ping timeout: 480 seconds]
23:43fossdd has quit [Ping timeout: 480 seconds]
23:51Dcow has joined #asahi-gpu
23:59Dcow has quit [Ping timeout: 480 seconds]
…