2026-01-29
01:39pb17 has quit [Ping timeout: 480 seconds]
01:51pb17 has joined #asahi-gpu
02:36alyssaolivial: the fact AGX can't read interleaved means we effectively can't use it for anything in Mesa.
02:40olivialthe way I'm handling this in panvk is to just not have a fast path for passthrough-gs-already-bound
02:50nightstrike_ has joined #asahi-gpu
03:00pb17 has quit [Ping timeout: 480 seconds]
03:56nightstrike_ is now known as nightstrike
04:01pb17 has joined #asahi-gpu
05:34JamesCalligeros[m]dcp supports actual agx framebuffers
05:34JamesCalligeros[m]we dont need to care about interchange unless its the only way to get framebuffers out of avd
05:35JamesCalligeros[m](note that the support for agx fbs is implied by firmware re and not m1n1 work. ive never actually seen macos pass an agx fb to dcp)
05:40pb17 has quit [Ping timeout: 480 seconds]
06:01JamesCalligeros[m]I got it to display them, but not properly. There are clearly flags or something we need to set to make it deal with the tiling and compression and addressing mode
06:02JamesCalligeros[m]I just dont know where those could be. I got partway through reversing iomfb_verify_surfaces in ghidra but ran out of holiday :p
06:12Z750 has quit [Quit: bye]
06:13Z750 has joined #asahi-gpu
06:13oliverCould you send me what you have? I was also doing that, but as soon as I enable compression and tiling, the dcp crashes with something like "isr0 queue not empty". I also looked into the firmware blob, and I think I understood the verify surfaces and verify agx surfaces methods. They do pass happily with compression set, but still crashes in the end..
06:16pb17 has joined #asahi-gpu
06:23tbodt has quit []
06:24tbodt has joined #asahi-gpu
06:35chadmed has joined #asahi-gpu
06:38ciara has quit [Remote host closed the connection]
06:38chadmedi can send you my ghidra project (#asahi-re if you wanna discuss this more) but in terms of making dcp show agx fbs i just added the format modifiers to the kernel driver and it showed them without crashing (with the above caveats that they looked wrong)
06:39ciara has joined #asahi-gpu
06:39chadmedmy ghidra project is probably less complete than yours, i spent maybe like 6 or so hours of actual work on it :P
06:42oliverhere are some pictures of rendering and scanning out interchange. i am using a little program to open the drm interface directly & render the triangle via gl, no wayland between.
06:42chadmedi got as far as verify_surfaces complaining that the surfaces havent been marked as having twiddled addresses
06:54lukeleppan has quit [Quit: The Lounge - https://thelounge.chat]
06:57lukeleppan has joined #asahi-gpu
07:29jannauoliver: have you looked at the interchange_compressed framebuffer in an image viewer? Does it look like the linear scanout of it?
07:34mattia013 has joined #asahi-gpu
07:48cr1901__ has joined #asahi-gpu
07:56cr1901_ has quit [Ping timeout: 480 seconds]
08:15pb17 has quit [Ping timeout: 480 seconds]
08:29jannauoliver: my point being that is that I expect a mostly solid color picture to compress well. As in just a few bytes per page sized tile. so your linear scanout doesn't look that unexpected to me
08:31jannauI expect the compression metadata to contain offset into the frame buffer to indicate where the tile data is stored. compare how Arm's compressed framebuffer format works
08:32ece314378925355451680698427415 has quit [Ping timeout: 480 seconds]
08:35jannauhttps://www.collabora.com/news-and-blog/blog/2020/04/08/adding-mainline-arm-frame-buffer-compression-support-for-rockchip/ has an overview how AFBC works
08:37[KongrooParadox] has joined #asahi-gpu
08:39KongrooParadox has quit [Ping timeout: 480 seconds]
09:03pb17 has joined #asahi-gpu
09:21olivialoof, putting together a CTS test for the xfb bug is kinda annoying cause it's only possible to actually trigger it with ESO
09:22olivialI was *gonna* test it against a version of panvk hacked to have the same bug, but panvk doesn't have ESO and I don't have hw to test against real hk :(
09:35KongrooParadox has joined #asahi-gpu
09:38[KongrooParadox] has quit [Ping timeout: 480 seconds]
10:15TheLink has quit [Quit: Bye!]
10:15TheLink has joined #asahi-gpu
10:17TheLink has quit []
10:18TheLink has joined #asahi-gpu
10:22[KongrooParadox] has joined #asahi-gpu
10:24KongrooParadox has quit [Ping timeout: 480 seconds]
10:44oliverjannau: I have looked at the images, yes. The interchange format compresses each 16x16 tile independently and writes one compressed tile per 1024 bytes. So I expect the picture to be mostly back and to have little dots of color all over the place.
11:30jsk has joined #asahi-gpu
11:40jannauoliver: I'm not sure I understand what you're exactly doing. are you trying to scanout a interchange_compressed surface created by a modified mesa? or a interchange surface rendered on macos and dumped?
12:13oliverThe first one. I've modified mesa and the kernel and build a small app to initialize the plane with interchange_compressed, render a triangle and scan that out. I looked at the raw compressed data (basically as linear would scan it out) to see how the compressed data looks like. I then compared what's in my rendered buffer with how the dumped buffer from macos looks like to get a feeling of how my data _should_ look like. Let me create
13:28alyssaoliver: I stand by what I said - effectively can't use interleaved compressed even for scanout in Mesa because AGX needs to be able to read things back.
13:33alyssa& the fact that iirc celebrity compositors like kwin *do* read back the scanout fb to implement screen grab
13:34alyssaand I won't be merging patches for any of this to Mesa unless it is very high quality, because this is subtle
13:35olivermeh. okay. any hints on why macos is using it then? they dont have any of those problems?
13:59oliverAfter reading gpu part 5 I don't understand how macos can use the interchange format. I can configure a metal application to do direct scanout in fullscreen (I think they call it direct-to-display). I recently did that and observed the dcp traces. The way I understood you this should not work when you have partial renders, as the gpu might need to read parts of it back. Do you have any insights in that?
15:30john-cabaj has joined #asahi-gpu
15:35john-cabaj has quit [Read error: No route to host]
15:35john-cabaj has joined #asahi-gpu
15:36john-cabaj has quit [Read error: No route to host]
15:36john-cabaj has joined #asahi-gpu
15:39john-cabaj1 has joined #asahi-gpu
15:39john-cabaj has quit [Read error: Connection reset by peer]
15:39john-cabaj1 is now known as john-cabaj
15:40john-cabaj has quit []
15:40john-cabaj has joined #asahi-gpu
15:40john-cabaj has quit []
15:40john-cabaj has joined #asahi-gpu
15:41john-cabaj has quit []
15:41john-cabaj has joined #asahi-gpu
15:44john-cabaj has quit []
15:45john-cabaj has joined #asahi-gpu
15:46john-cabaj has quit []
15:46john-cabaj has joined #asahi-gpu
15:59john-cabaj has quit [Quit: john-cabaj]
15:59john-cabaj has joined #asahi-gpu
16:06KongrooParadox has joined #asahi-gpu
16:07john-cabaj has quit [Ping timeout: 480 seconds]
16:08[KongrooParadox] has quit [Ping timeout: 480 seconds]
16:09john-cabaj has joined #asahi-gpu
16:17john-cabaj has quit [Ping timeout: 480 seconds]
16:54john-cabaj has joined #asahi-gpu
17:02john-cabaj has quit [Ping timeout: 480 seconds]
17:42john-cabaj has joined #asahi-gpu
18:58ipatch has quit [Read error: Connection reset by peer]
19:14rivendell has joined #asahi-gpu
19:19ipatch has joined #asahi-gpu
19:35john-cabaj1 has joined #asahi-gpu
19:38ipatch has quit [Quit: WeeChat 4.8.1]
19:39john-cabaj has quit [Ping timeout: 480 seconds]
19:39john-cabaj1 is now known as john-cabaj
19:47ipatch has joined #asahi-gpu
20:00john-cabaj1 has joined #asahi-gpu
20:05john-cabaj has quit [Ping timeout: 480 seconds]
20:05john-cabaj1 is now known as john-cabaj
20:19john-cabaj has quit [Ping timeout: 480 seconds]
20:21john-cabaj has joined #asahi-gpu
20:28rosalyntg has quit [Ping timeout: 480 seconds]
20:41rosalyntg has joined #asahi-gpu
20:43Tieppo[m] has quit [Ping timeout: 480 seconds]
20:43Bjarne[m] has quit [Ping timeout: 480 seconds]
20:43svenpeter[m] has quit [Ping timeout: 480 seconds]
20:43Moderation[m] has quit [Ping timeout: 480 seconds]
20:44JamesCalligeros[m] has quit [Ping timeout: 480 seconds]
20:44Felix[m]1 has quit [Ping timeout: 480 seconds]
20:44lacamar[m] has quit [Ping timeout: 480 seconds]
20:44Dcow[m] has quit [Ping timeout: 480 seconds]
21:33j`ey_ is now known as j`ey
22:16pb17 has quit [Ping timeout: 480 seconds]
22:33pb17 has joined #asahi-gpu
23:06john-cabaj has quit [Ping timeout: 480 seconds]