#asahi-gpu · 2023-02-15 15:32
linaOne little thing: sometimes you get GPU faults when killing things. Since everything is async now and BOs are tracked by userspace, if userspace dies in-progress GPU operations lose their BOs. The driver itself keeps all the critical structures alive until the jobs complete/fail, so nothing breaks, you just get GPU fault logs.#asahi-gpu · 2023-02-15 15:26
linaPushed all the explicit sync stuff to gpu/explicit-sync (kernel) and agx/explicit-sync (mesa), seems to work well ^^#asahi-gpu · 2023-02-15 15:18
lina800x600 surfaceless glmark score: 4496#asahi-gpu · 2023-02-15 13:33
linaI did fix some more threading bugs ^^#asahi-gpu · 2023-02-15 13:33
linaalyssa: With that commented out, we pass all the EGL tests except robust_gl_3* context creation stuff which sounds not applicable... ^^#asahi-gpu · 2023-02-15 12:52
alyssa(unlikely to be today)#asahi-gpu · 2023-02-15 12:52
alyssalina: K, will look when I get a chance#asahi-gpu · 2023-02-15 12:20
danielsthe latter is good anyway if you don't want your running session to get trashed#asahi-gpu · 2023-02-15 12:19
danielstry either skipping that subset of tests, or EGL_PLATFORM=surfaceless#asahi-gpu · 2023-02-15 12:18
danielsah yes#asahi-gpu · 2023-02-15 12:11
linadaniels: x11/egl (I think that's what the dEQP build picked by default?)#asahi-gpu · 2023-02-15 11:58
danielslina: which winsys?#asahi-gpu · 2023-02-15 10:07
linadEQP-EGL.functional.swap_buffers_with_damage.resize_after_swap.buffer_age_render#asahi-gpu · 2023-02-15 10:07
linaalyssa: This is triggering on the EGL testsuite: Assertion `maxx > minx && maxy > miny' failed#asahi-gpu · 2023-02-15 03:26
alyssalina: Any chance I could get some pre-stream code review? *sweats*#asahi-gpu · 2023-02-14 21:59
alyssasee https://developer.apple.com/documentation/metal/mtlrenderpipelinedescriptor?language=objc for the pseudo code, of course the actual expression is implementation-defined#asahi-gpu · 2023-02-14 21:58
alyssahttps://backtick.town/~bloom/alpha-to-coverage.txt#asahi-gpu · 2023-02-14 21:57
alyssaIt just doesn't make sense what's happening#asahi-gpu · 2023-02-14 21:57
alyssadottedmag: No, I don't have any reason to think it's wrong#asahi-gpu · 2023-02-14 21:56
dottedmagAbility of Apple engineers to make bugs?#asahi-gpu · 2023-02-14 21:47
alyssathis, this doesn't make sense to me why they would do that, which makes me think i'm missing something#asahi-gpu · 2023-02-14 21:47
alyssaso end up with a float in the range [0, 1132462080.0) i guess#asahi-gpu · 2023-02-14 21:46
alyssamantissa is free (save for denormals which have been flushed), so we end up encoding [0, 135 << 23)#asahi-gpu · 2023-02-14 21:45
alyssaand then adding back 0 we get an exponent range of [0, 134]#asahi-gpu · 2023-02-14 21:45
alyssasign bit is clear#asahi-gpu · 2023-02-14 21:45
alyssaencoded as [1, 134]#asahi-gpu · 2023-02-14 21:44
alyssaRange (0.0, 255.0] corresponds to an exponent of [-126, 7] #asahi-gpu · 2023-02-14 21:42
alyssaThen they... reinterpret the float as a 32-bit integer and convert it to a float?#asahi-gpu · 2023-02-14 21:41
alyssaSo now they have a 32-bit float in the range of [0.0, 255.0]#asahi-gpu · 2023-02-14 21:41
alyssaThen, they multiply it with 255.0#asahi-gpu · 2023-02-14 21:41
alyssaFirst, they calculate the alpha and clamp it to [0.0, 1.0]#asahi-gpu · 2023-02-14 21:41
alyssathis is for alpha to coverage with msaa 4x#asahi-gpu · 2023-02-14 21:40
alyssafor anyone following along at home#asahi-gpu · 2023-02-14 21:40
alyssabe91c8280000 is the spicy convert#asahi-gpu · 2023-02-14 21:38
alyssawhat kind of witchcraft is this#asahi-gpu · 2023-02-14 21:38
alyssathis doesn't make any sense, why are they converting a float from uint to float#asahi-gpu · 2023-02-14 21:35
alyssaer no it's a funny encoding of convert#asahi-gpu · 2023-02-14 21:32
alyssalikelier than you would think#asahi-gpu · 2023-02-14 21:32
alyssaan unknown ALU instruction? on M1?#asahi-gpu · 2023-02-14 06:11
i509vcbI am interested in trying to see if I *could* help with agxv, but I guess I am a bit unfamiliar with the hardware and I imagine some stuff is probably blocking larger open contribution on that#asahi-gpu · 2023-02-14 06:10
i509vcbhey ella-0, I recall hearing you have gotten vkcube to "work" on agxv. vkcube obviously requires WSI and agxv is probably some time away from working wsi so I imagine that showcase was probably from some fork of vkcube. I was wondering if you had the vkcube fork somewhere (if it even exists) to experiment with?#asahi-gpu · 2023-02-14 04:48
TellowKrinkleActually maybe not#asahi-gpu · 2023-02-14 04:47
TellowKrinkle It usually sets up a valid pair before each of those calls#asahi-gpu · 2023-02-14 04:47
TellowKrinkleI think the bindless register references are register pairs and we're not decompiling correctly#asahi-gpu · 2023-02-14 04:11
alyssamaybe I just need to try to wire this stuff up and see what breaks.#asahi-gpu · 2023-02-14 04:09
alyssaare these addresses 32-bits or not?!#asahi-gpu · 2023-02-14 04:08
alyssaI still feel like I'm missing something#asahi-gpu · 2023-02-14 03:52
alyssanext question is whether there's proper bindless image load i missed#asahi-gpu · 2023-02-14 03:51
alyssaok, so there's full bindless image_write, so what? well..#asahi-gpu · 2023-02-14 03:49
alyssai just mised the second add due to the scheduling