#
…

2025-12-30

00:02hdbngr has quit [Ping timeout: 480 seconds]
00:12hdbngr has joined #asahi-dev
00:21hdbngr has quit [Ping timeout: 480 seconds]
00:25mixi has quit [Ping timeout: 480 seconds]
00:36mixi has joined #asahi-dev
01:03michaelrostom has quit [Ping timeout: 480 seconds]
01:06terr has joined #asahi-dev
01:08pb17 has quit [Ping timeout: 480 seconds]
01:10michaelrostom has joined #asahi-dev
01:17pb17 has joined #asahi-dev
01:50nav-28 has joined #asahi-dev
01:50nav-28 has quit [Remote host closed the connection]
01:51tobhe_ has joined #asahi-dev
01:54ralim has quit [Read error: Connection reset by peer]
01:54ralim has joined #asahi-dev
01:56tobhe has quit [Ping timeout: 480 seconds]
02:14terr has quit [Ping timeout: 480 seconds]
02:41michaelrostom has quit [Ping timeout: 480 seconds]
02:47pb17 has quit [Remote host closed the connection]
02:47pb17 has joined #asahi-dev
02:51michaelrostom has joined #asahi-dev
03:07hamr8 has quit [Read error: Connection reset by peer]
03:07hamr8 has joined #asahi-dev
03:13abhay has quit [Read error: Connection reset by peer]
03:14abhay has joined #asahi-dev
03:19michaelrostom has quit [Ping timeout: 480 seconds]
03:23tobhe has joined #asahi-dev
03:24rany has quit [Remote host closed the connection]
03:26abhay has quit [Ping timeout: 480 seconds]
03:28rany has joined #asahi-dev
03:30tobhe_ has quit [Ping timeout: 480 seconds]
03:35abhay has joined #asahi-dev
03:36hexdump01 has joined #asahi-dev
03:38hexdump0815 has quit [Ping timeout: 480 seconds]
04:18hdbngr has joined #asahi-dev
04:20marvin24 has quit [Ping timeout: 480 seconds]
04:26pb17 has quit [Ping timeout: 480 seconds]
04:27hdbngr has quit [Ping timeout: 480 seconds]
04:29eluks has quit [Remote host closed the connection]
04:30eluks has joined #asahi-dev
04:32pb17 has joined #asahi-dev
04:39nora has joined #asahi-dev
04:41nora_ has quit [Ping timeout: 480 seconds]
04:41abhay has quit [Read error: Connection reset by peer]
04:42abhay has joined #asahi-dev
04:48terr has joined #asahi-dev
04:58terr has quit [Quit: Konversation terminated!]
04:59terr has joined #asahi-dev
05:01hdbngr has joined #asahi-dev
05:09hdbngr has quit [Ping timeout: 480 seconds]
06:00hdbngr has joined #asahi-dev
06:08hdbngr has quit [Ping timeout: 480 seconds]
06:25terr has quit [Quit: Konversation terminated!]
06:56terr has joined #asahi-dev
07:02hdbngr has joined #asahi-dev
07:05abhay has quit [Ping timeout: 480 seconds]
07:10hdbngr has quit [Ping timeout: 480 seconds]
07:13abhay has joined #asahi-dev
07:14terr has quit [Remote host closed the connection]
07:20pb17 has quit [Ping timeout: 480 seconds]
07:25pb17 has joined #asahi-dev
07:29abhay has quit [Ping timeout: 480 seconds]
07:44abhay has joined #asahi-dev
07:52abhay has quit [Ping timeout: 480 seconds]
08:01abhay has joined #asahi-dev
08:03hdbngr has joined #asahi-dev
08:09abhay has quit [Remote host closed the connection]
08:11hdbngr has quit [Ping timeout: 480 seconds]
08:19jvd_ has joined #asahi-dev
08:21jvd_ has quit []
08:22aissen has quit [Ping timeout: 480 seconds]
08:39alxdrl has joined #asahi-dev
08:40abhay has joined #asahi-dev
08:44abhay has quit [Read error: Connection reset by peer]
08:44abhay has joined #asahi-dev
08:50pb17 has quit [Ping timeout: 480 seconds]
09:04hdbngr has joined #asahi-dev
09:08integralpilot has joined #asahi-dev
09:08integralpilotsven: good luck with your talk! it starts in about an hour right? i'm sure it will go amazing!
09:09sventhanks!
09:09svenyes, about 50 minutes left
09:12hdbngr has quit [Ping timeout: 480 seconds]
09:17integralpilot has quit [Ping timeout: 480 seconds]
09:20rany has quit [Ping timeout: 480 seconds]
09:23pb17 has joined #asahi-dev
09:27campagnol has joined #asahi-dev
09:29integralpilot has joined #asahi-dev
09:31campagnolsven: good luck and, above all, enjoy the moment :) ! How many slides did you end up writing ? :P
09:31nickchangood luck!
09:32nickchanhm some iPads actually have horizontal framebuffer
09:36noopwafelintegralpilot: hey for the binary patches you're just trying to filter out changes to the registers which you set in the chicken bits, right?
09:37DiscreteVic has joined #asahi-dev
09:37nickchannoopwafel: those are attempts to access _EL2 or _EL12 registers from _EL1
09:37nickchan(which doesn't work)
09:41integralpilotthere's no sound :(
09:41leioit's in 18 minutes, I assume they intentionally don't give early setup sound
09:42integralpilotoh, makes sense
09:42leioit was some informational stuff before; some sound coming through just now too
09:42leioit switched to showing the room so early just as I was posting the link
09:43noopwafelfirst talk of the day in here so they are still doing setup etc
09:43nickchanthere's sound now
09:44integralpilotwhy is it like ominous video game boss music
09:45integralpilotwell i guess it's more intended as like techno background music while it shows the schedule and stuff as they set up
09:45leiofirst time I see some conference actually do a clap test and such. Should we go to another channel for the stream discussions though? :)
09:47integralpilot#asahi-stream ?
09:47nickchani mean there's #asahi-stream which is basically unused
09:47jannauI think #asahi would be fine
09:47campagnol has quit [Quit: campagnol]
09:48jannausven's talk isn't going to last 8+ hours
09:48campagnol has joined #asahi-dev
09:48noopwafelnickchan: is that new for the M3 setup? I went down a rabbit hole yesterday of trying to try hacr_trap_bits.py and then, oh, el1_call is broken (I have patch) after wondering if there was a nicer way to trap them
09:49abhay has quit [Read error: Connection reset by peer]
09:49integralpilotthey seem to be registers which, when there is nothing in EL2, access it allowed from EL1. but when there is something (m1n1) in EL2, EL1 is not allowed to touch them
09:50abhay has joined #asahi-dev
09:50integralpilotif we don't set the chickens, XNU hangs at some sha512 function, so i assume they are crypto related in some way
09:50chadmed has quit [Quit: Konversation terminated!]
09:50chadmed has joined #asahi-dev
09:51rany has joined #asahi-dev
09:51integralpilotwhich i guess makes sense for them to be protected like this, probably if you are a hypervisor running a sandboxed VM, you don't want it messing with sensitive registers like that
09:51abhay has quit [Read error: Connection reset by peer]
09:51abhay has joined #asahi-dev
09:51integralpilotto my knowledge there's no way for us in EL2 to allow EL1 to access these, so we have to binary patch, but there might be another way
09:51nickchansounds like hacr/hcr trapping to me
09:52noopwafelmakes sense, thanks :) we were trying to use a different kernel and we have the offsets to nop but it seemed horrible :D
09:52integralpilotit's annoying that it doesn't like trap to us in EL2 when EL1 touches these, it sends an exception to El1 and doesn't notify us
09:53noopwafelif it is still 'open' then maybe I will play a bit later today and see if I can blindly find any trap bits
09:53nickchantidcp should cause those to be trapped
09:53nickchanBut that's probably too many traps
09:54chaos_princessintegralpilot: "nothing in el2"? macos runs in el2 under vhe, no?
09:54integralpiloti thought on normal macOS boot there's nothing in EL2, but maybe i'm wrong
09:54nickchanon normal macOS boot macOS runs in EL2
09:55nickchanso of course it can access _EL2 and _EL12 registers then
09:55chaos_princessno, the keywords to search for is "virtualization host extensions", its an arm thing that makes el2 look kinda like el1
09:55integralpilotoh interesting, thank you!
09:55nickchanand vhe adds some register redirection so the same encoding can be a different register in different EL
09:55chaos_princessapple chips always run using vhe, old non-vhe virtualization is not supported
09:56nickchanlike im pretty sure s3_1_c15_c1_5 gets the vhe treatment with s3_4_c15_c14_6 being the corresponding _EL12 register
09:59Shizchaos and drama
10:00noopwafelstream is going to start now
10:00nickchanand some of those registers has to do with what appears to be a full set of custom parallel registers to ARM's arch timer
10:00nickchanas found in NXU sources
10:00nickchanxnu sources
10:00Shizi'm sure we can make it work later today :)
10:02nickchanhere's like 5 registers that's a clone of arm registers but probably runs at 24 MHz instead of 1 GHz
10:05hdbngr has joined #asahi-dev
10:13hdbngr has quit [Ping timeout: 480 seconds]
10:40alxdrl has quit [Quit: My Mac has gone to sleep. ZZZzzz…]
10:44integralpilotjust finished RE of new RTKit oslog message format, the updated driver is working now
10:44integralpilotuploading PR soon just doing some testing
10:45integralpilotgiven RTKit is also in mainline, once merged into asahi, should i bring this oslog patch upstream as well?
10:46integralpilotthere's definetely still more RTKit fixing to do but i feel like smaller, self-contained patches are better than just once big "RTKit M3 makeover" patch but not sure about upstream dynamics or when we should send stuff there
10:46nickchanrtkit is in drivers/soc/apple
10:47chadmedanything thats already upstream please just submit upstream
10:47nickchanand sven is a maintainer of that
10:50nickchanintegralpilot: that's why patch series is a thing
10:50nickchanmutliple patches in one series
10:51Shizyea, makes for nicer reviewing :)
10:51Shizcan def recommend small self-contained patches
10:51integralpilotthanks! so is it better to wait untill all of rtkit is ready for M3 and submit a series, or submit oslog now? and i shouldn't make a PR to asahi?
10:52Shizif it's a small self-contained patch, I'd just send it upstream directly and cc sven (personally, but I'm not super familiar with the process here)
10:52Shizbut considering sven is the maintainer, you'd cc him anyway :)
10:52integralpilotthank you, makes sense!
10:52chadmedi dont think theres any point in submitting the oslog thing by itself
10:53chadmedi would submit it as part of a "make rtkit on m3+ work" series
10:55integralpilotperfect, thanks!
10:57nickchanintegralpilot: did apple bump protocol version?
10:57nickchanor still 12
10:58nickchanupstream supports 11-12
10:59chadmednow that the cat's out of the bag ive pushed a rebased dcp/scanout branch
10:59chadmedhdr is a no-go until the colour op stuff is merged so dont even bother with that (cc oliver)
11:00hdbngr has joined #asahi-dev
11:00chadmedtrying to grab the hdr metadata off the connector at the point we need it for dcp to properly deal with mixed hdr/sdr surfaces is flaky, and compositors cant yet deal with it anyway
11:00integralpilotstill says 12, but it's sending oslog messages that the previous driver didn't know how to handle and said were an invalid format
11:01chadmedthis is doubly compounded when using multiple displays (especially across hdmi and dp) as the colour spaces can be different and we cant deal with that gracefully
11:03chadmednot sure what we need to do for VRR. hopefully its just a matter of filling out the drm_mode with the right info from TimingElements, setting a VRR mode then letting iomfb do the needful
11:03chadmedmacos doesnt want to do vrr with any of my monitors for some reason even though TimingElements brings back a bunch of supported VRR modes
11:04chaos_princesschadmed: how are you trying to make it do vrr?
11:04chadmedon macos? plug in my monitor then look for the setting in system settings
11:05chaos_princessi mean the content
11:05chadmedi dont get that far because macos doesnt recognise the monitors as vrr capable
11:05chadmedso i cant even arm the setting
11:06hdbngr has quit [Remote host closed the connection]
11:06hdbngr has joined #asahi-dev
11:06leiomy monitor only advertises VRR if it's in "game mode", so I was confused there for a while with mine
11:06chadmedyeah i mightve spent like 4 hours yesterday trying to figure out why HDR wasnt coming up in plasma
11:06chaos_princesswhat about the internal screen?
11:07unsecretised has joined #asahi-dev
11:07integralpilot has quit [Remote host closed the connection]
11:08integralpilot has joined #asahi-dev
11:08chadmedive been avoiding using that as a base because i want to see what dcp does for a display that behaves properly first
11:08chadmedeliminating the lack of edid/timingelements as a factor to consider
11:09unsecretised has quit [Remote host closed the connection]
11:09chadmedthough on the topic i am less sure about those timestamps now than i was a week ago
11:10chadmedthere seems to be a lot of slack and no rhyme/reason to the values. some swaps dont even populate them at all
11:16chadmedmy plan at some point is to see what happens when we just fill in the mode properly and modeset to the vrr timings
11:16chadmedbut im back at work after this week :(
11:17hdbngr has quit [Ping timeout: 480 seconds]
11:19integralpilot has quit [Ping timeout: 480 seconds]
11:21hdbngr has joined #asahi-dev
11:52paps__ has joined #asahi-dev
11:53campagnol has quit [Remote host closed the connection]
12:04oliverchadmed: would you mind explaining that a little bit please. regarding hdr and color spaces. What do you mean with color op? The changes in your branch? Also why do er need to care about mixed surfaces right now if kwin and mutter transformer everything to bt2020/pq anyway right now? And running one screen in sdr and one in hdr also looked fine yesterday. Mostly just for my understanding
12:11chadmedoliver: kwin can currently only put the connector into RGB BT.2020 with PQ transfer function
12:12oliverYea, but that's fine, isn't it?
12:12chadmedproblem 1 is that we dont know that "DCP_XFER_FUNC_HDR" actually corresponds to PQ, HLG, or HDR10 which all use different curves
12:12chadmedproblem 2 is that DCP's mode selection logic is stupid, and for some displays will select YCC BT.2020 primaries
12:12chadmedtrying to output an RGB BT.2020 surface on to a limited range YCbCr BT.2020YCC display looks like crap
12:13chadmedproblem 3 is that without flaky gross hacks that sometimes cause locks and hangs, we cannot switch between the correct surface primaries and transfer functions when enabling/disabling HDR at runtime
12:15chadmedso yeah it looks fine when everything lines up. my monitor plugged in to displayport with full range 10-bit RGB, PQ transfer func and BT.2020 colour space of course looks fantastic when kwin sets the surface framebuffer to ARGB2101010 and uses the PQ transfer function with BT.2020 colour space
12:16chadmedwhen i plug the exact same monitor in via hdmi, it looks like total garbage because dcp thinks limited YCbCr is the "best" colour mode for that display
12:17chadmedso im not terribly keen on enabling hdr until we have some more control over colour modes and what dcp is doing to blend surfaces. at a minimum we need to figure out wiht more certainty what all of the surface colourspaces and transfer functions are
12:18chadmedi havent tried plugging in my dolby vision oled yet but thats another factor we have to consider since DV uses completely different colour spaces again lol
12:18oliverProblem 1: what's the best way to find this out? Measure?
12:19chadmedget some source video that explicitly uses each and then play it back in quicktime while tracing
12:19chadmedthats how i found BT601 and BT709
12:20chadmedapples multimedia handling is in general very very good when the media is spec-conformant
12:21chadmede.g. an apple tv will hw decode and direct scanout h.265 video tagged with 'hvc1' but not 'hevc' or 'hev1' which are for some reason common nonstandard tags with.... "alternatively sourced" media
12:21chadmedso as long as your mp4 file has all the proper tags and colourspace metadata, quicktime will tell the compositor to tell dcp to do what you want it to
12:22oliverThat is quite nice
12:23aissen has joined #asahi-dev
12:23chadmedffmpeg does the right thing so its not terribly difficult to make known-good source files to play withf
12:23olivercould we enable pq/2020 just for the internal display earlier if the hdmi thing doesn't work out that well
12:23chadmedyeah if we sort the fake edids out then i dont see why not
12:24chadmedi have a hard time believing 'r03w' actually "looks fine" though :P
12:24chadmedtry using 'r01l' for A/XARGB2101010 instead
12:24integralpilot has joined #asahi-dev
12:24oliverYea I did change that in my github pr already.
12:25oliverTogether with bt2020 it looks good now
12:25oliverIs there anything I can support you with?
12:26chadmedmore RE if youre happy to do that i suppose. if you really want a challenge, youll notice that the framebuffer formats used by macos are quite odd. that is because they are "apple interchange/interchange compressed" which we dont know much about yet
12:26chadmedAGX supposedly can write "interchange" format textures but cannot read them so we never bothered looking at them
12:27oliverYes I saw those when dumping the surface Mac sends to the dcp
12:28chadmedbut looking at IOMFB, there is some flag we need to set _somewhere_ to make DCP scan out interchange or AGX format framebuffers. i havent been able to find this flag at all and id really like for us to be able to scan out AGX format framebuffers if possible
12:29chadmedif thats too daunting then lets work on finding the iosurface transfer functions and colourspaces
12:29chadmedscanning out nonlinear fbs is just a nice to have power optimisation that i suspect might also help with our dismal swap timing
12:30oliverCan we already create such buffers from mesa?
12:30chadmedyep, try setting DRM_FORMAT_MOD_APPLE_GPU_TILED and DRM_FORMAT_MOD_APPLE_GPU_TILED_COMPRESSED as valid format motifiers in the dcp driver and see what happens
12:31oliverI'll see what I can do. For my $job also starts again next week.
12:37trello has joined #asahi-dev
12:38trello has quit []
12:41bshephar has joined #asahi-dev
12:41integralpilot has quit [Ping timeout: 480 seconds]
12:42bshephar has quit []
12:42abhay has quit [Read error: Connection reset by peer]
12:42abhay has joined #asahi-dev
12:46alxdrl has joined #asahi-dev
12:46DiscreteVic has quit [Ping timeout: 480 seconds]
12:55abhay has quit [Ping timeout: 480 seconds]
13:06abhay has joined #asahi-dev
13:18alxdrl has quit [Quit: My Mac has gone to sleep. ZZZzzz…]
13:36DiscreteVic has joined #asahi-dev
13:37DiscreteVic has quit []
13:50michaelrostom has joined #asahi-dev
13:59michaelrostom has quit [Quit: Konversation terminated!]
14:07chadmed_ has joined #asahi-dev
14:07chadmed has quit [Read error: Connection reset by peer]
14:49oliveris the experiments/dcp.py script supposed to work with fw 13.5?
14:52jannaumaybe. I certainly used it with 13.5 but I probably have a few minor changes to make it work git-stashed away
15:02hdbngr has quit [Ping timeout: 480 seconds]
15:04hdbngr has joined #asahi-dev
15:09sativa93 has joined #asahi-dev
15:09sativa93 has quit [Remote host closed the connection]
15:13hdbngr has quit [Ping timeout: 480 seconds]
15:16hdbngr has joined #asahi-dev
15:26abhay has quit [Read error: Connection reset by peer]
15:26abhay has joined #asahi-dev
15:34abhay has quit [Ping timeout: 480 seconds]
15:37hdbngr has quit [Read error: No route to host]
16:05abhay has joined #asahi-dev
16:13satmaran has joined #asahi-dev
16:46alxdrl has joined #asahi-dev
16:55abhay has quit [Ping timeout: 480 seconds]
16:56abhay has joined #asahi-dev
17:02nazmul has joined #asahi-dev
17:06abhay has quit [Read error: Connection reset by peer]
17:07abhay has joined #asahi-dev
17:08nazmul has quit [Quit: https://quassel-irc.org - Chat comfortably. Anywhere.]
17:12nazmul has joined #asahi-dev
17:17oliverI am on the way of making it work by comparing it with the apple_dcp kernel module
17:49nazmul has quit [Quit: Page closed]
17:59abhay has quit [Ping timeout: 480 seconds]
18:00abhay has joined #asahi-dev
18:24alxdrl has quit [Quit: My Mac has gone to sleep. ZZZzzz…]
18:27yukagot a compile error on the fairydust branch
18:27yukawith the same config as I was using before (APPLE_MAILBOX=m)
18:35LarstiQintroduced by 7cc11916ccd291ed901d3bd01526a2730f6d76a5 ?
19:39asah has joined #asahi-dev
19:42asah has quit [Quit: Leaving]
19:49alxdrl has joined #asahi-dev
19:53michaelrostom has joined #asahi-dev
19:56abhay has quit [Remote host closed the connection]
19:56abhay has joined #asahi-dev
20:16paps__ has quit [Quit: Connection closed for inactivity]
20:23fl0_idoh noes, integralpilot sent me a device tree to test, but somehow I didn't see it on irc, no DMs were shown
20:23fl0_idmaybe it was because switiching between macos and linux irc clients and it somehow messed with the bouncer..
20:34paps__ has joined #asahi-dev
20:50michaelrostom has quit [Quit: Konversation terminated!]
21:03jannauAPPLE_MAILBOX=m should have always been broken with APPLE_SEP=m|y but the dependency is not expressed in Kconfig. not sure if I lost that or if it was alwyas
21:03jannaualways missing
21:08michaelrostom has joined #asahi-dev
21:10DiscreteVic has joined #asahi-dev
21:12DiscreteVicHi, Where is the tracking of the dp alt progress? Just to follow the last updates and if it's possible, to contribute. Thanks in advance!
21:16abhay has quit [Read error: No route to host]
21:16abhay has joined #asahi-dev
21:36DiscreteVic has quit [Quit: Lost terminal]
21:36yukaAlso had to set APPLE_RTKIT=y to make it compile
21:38cow123 has quit [Ping timeout: 480 seconds]
21:57pb17 has quit [Ping timeout: 480 seconds]
22:02michaelrostom has quit [Quit: Konversation terminated!]
22:18pb17 has joined #asahi-dev
23:02cow123 has joined #asahi-dev
23:09alxdrl has quit [Quit: My Mac has gone to sleep. ZZZzzz…]
23:10integralpilot has joined #asahi-dev
23:25hdbngr has joined #asahi-dev
23:32michaelrostom has joined #asahi-dev
23:34michaelrostom has quit [Remote host closed the connection]
23:35michaelrostom has joined #asahi-dev
23:36chadmed_ has quit [Quit: Konversation terminated!]
23:41chadmed has joined #asahi-dev
23:48integralpilot has quit [Read error: Connection timed out]
23:59hdbngr has quit [Remote host closed the connection]
23:59hdbngr has joined #asahi-dev
…