#asahi-gpu · 2023-08-13 22:21
alyssafor vulkan? pick a failing test and fix it#asahi-gpu · 2023-08-13 22:21
handlerugyeah that#asahi-gpu · 2023-08-13 22:21
handlerugbut then, is there a ticket tracker of some sort#asahi-gpu · 2023-08-13 22:20
rosefromthedeadpraying for another copy paste-based good first issue :)#asahi-gpu · 2023-08-13 22:20
alyssathere is.. a nonzero amount of pester, yes.#asahi-gpu · 2023-08-13 22:20
handlerugas i said earlier i don't know anything about graphics programming, was just wondering if maybe someone keeps pestering you about "vulkan when vulkan when"#asahi-gpu · 2023-08-13 22:19
alyssahandlerug: are you interested? o:)#asahi-gpu · 2023-08-13 22:19
handlerugare you recruiting?#asahi-gpu · 2023-08-13 22:15
alyssaIf you would like to see Vulkan support on M1, help make it happen. https://gitlab.freedesktop.org/asahi/mesa/-/tree/agxv/main is the branch to contribute to. no, "testing" is not a contribution.#asahi-gpu · 2023-08-13 21:29
Marybtw I did some run yesterday and scripted something to grab all assert for every test crashes https://cloud.mary.zone/s/ayFBHsB43qLpHCM#asahi-gpu · 2023-08-13 21:17
alyssaIf you would like to see Vulkan support on M1, help make it happen. https://gitlab.freedesktop.org/asahi/mesa/-/tree/agxv/main is the branch to contribute to. no, "testing" is not a contribution.#asahi-gpu · 2023-08-12 15:42
alyssaI guess see iris_screen_create for example of how driParseConfigFiles and driQueryOption works#asahi-gpu · 2023-08-12 15:41
alyssaand handle that in the driver somehow, i've never done this uh#asahi-gpu · 2023-08-12 15:41
alyssamatch firefox + asahi#asahi-gpu · 2023-08-12 15:41
alyssano_16bit#asahi-gpu · 2023-08-12 15:40
alyssasrc/util/00-mesa-defaults.conf#asahi-gpu · 2023-08-12 15:40
alyssaumm#asahi-gpu · 2023-08-12 14:57
linaalyssa: How do we do that?#asahi-gpu · 2023-08-12 14:26
alyssaI'm fine with that#asahi-gpu · 2023-08-12 06:04
linaalyssa: I think we want to no16 shame firefox... they haven't replied to my bug report (with patch) about the Sheets bug yet, but there's also all the broken WebGL games/apps...#asahi-gpu · 2023-08-11 20:45
Maryyay#asahi-gpu · 2023-08-11 16:03
alyssathere is now NVK in the AGXV tree for easier copying :3#asahi-gpu · 2023-08-11 16:03
alyssaand rebased#asahi-gpu · 2023-08-11 15:59
alyssaMary: i'm rebasing agxv now fwiw#asahi-gpu · 2023-08-11 13:29
alyssadidn't realize that real apps were hitting that#asahi-gpu · 2023-08-11 13:27
alyssaregister allocation bug fix#asahi-gpu · 2023-08-11 13:27
alyssalina: 424ad888aeb ("agx: Don't reuse ssa_to_reg across blocks")#asahi-gpu · 2023-08-11 13:26
linaIt could also be randomly "fixed" though, given that it's a heisenbug..#asahi-gpu · 2023-08-11 13:26
alyssaWeird, usually I only fix fake bugs..#asahi-gpu · 2023-08-11 13:26
linaThat one's been broken in firefox for ages...#asahi-gpu · 2023-08-11 13:25
linaPlease...#asahi-gpu · 2023-08-11 13:25
alyssashould I bisect?#asahi-gpu · 2023-08-11 13:25
linaOh.#asahi-gpu · 2023-08-11 13:25
linaYeah, I didn't see anything obvious but I didn't trace every single line...#asahi-gpu · 2023-08-11 13:25
alyssalina: It's fixed on my next/ branch#asahi-gpu · 2023-08-11 13:25
alyssageneral compiler snafu not related to floating point crimes sounds likely then#asahi-gpu · 2023-08-11 13:25
alyssaalright#asahi-gpu · 2023-08-11 13:24
linaSo I think it has to do something with that#asahi-gpu · 2023-08-11 13:24
linaBut one side effect of that is the order of the basic blocks, and the sense of the condition variable, get flipped in the asm#asahi-gpu · 2023-08-11 13:24
linaSo one thing that fixes it is changing the condition in fetch_clip_area() to `if (index >= CLIP_TASK_EMPTY && index != 0xdead) {` (which should be a no-op as long as the index isn't that weird value)#asahi-gpu · 2023-08-11 13:23
alyssa(If this were VK and it used KHR_float_controls, we'd have a different conversation)#asahi-gpu · 2023-08-11 13:23
alyssaso if a GL shader is depending on IEEE compliant NaN behaviour, that's usually its own fault#asahi-gpu · 2023-08-11 13:23
alyssabut GL isn't expected to be IEEE compliant and we're allowed to optimize assuming that there are no NaNs and such#asahi-gpu · 2023-08-11 13:22
alyssain general I'd expect the hardware to be IEEE compliant#asahi-gpu · 2023-08-11 13:22
linaI mean what about cmpsel type stuff?#asahi-gpu · 2023-08-11 13:22
alyssaand, IDK, correctly?#asahi-gpu · 2023-08-11 13:22
alyssa`--sb --wait` was my friend#asahi-gpu · 2023-08-11 13:22
linaDo we know how NaN propagation works in general on AGX?#asahi-gpu · 2023-08-11 13:22
linaThe square is supposed to be a map thing, not glitched out#asahi-gpu · 2023-08-11 13:21
alyssayeah ok thanks