#asahi-gpu · 2021-01-13 01:09
davidrysk[m]:(#asahi-gpu · 2021-01-13 01:09
bloomIOConnectCallMethod#asahi-gpu · 2021-01-13 01:09
bloomugh I'm tired#asahi-gpu · 2021-01-13 01:09
bloomer#asahi-gpu · 2021-01-13 01:09
bloomyes#asahi-gpu · 2021-01-13 01:08
davidrysk[m]IOServiceCallMethod is what you're looking for?#asahi-gpu · 2021-01-13 01:08
bloomyes?#asahi-gpu · 2021-01-13 01:08
davidrysk[m]can you repeat the whole thing?#asahi-gpu · 2021-01-13 01:01
bloom*Connect#asahi-gpu · 2021-01-13 01:00
davidrysk[m]IOServiceCallMethod?#asahi-gpu · 2021-01-13 00:59
bloomAnyone know what KERN_INVALID_ADDRESS means from IOServiceCallMethod?#asahi-gpu · 2021-01-13 00:02
bloomok, next up is figuring out some of these memory allocations#asahi-gpu · 2021-01-12 19:19
marcanbloom: I mean it doesn't mean it won't perform well :p#asahi-gpu · 2021-01-12 19:15
bloommarcan: bwah?#asahi-gpu · 2021-01-12 18:53
austriancoderdavidrysk[m]: they are part of Mesa but the different drivers are not sharing much code as for OpenGL. For OpenGL (ES/CL) there is a common interface called gallium#asahi-gpu · 2021-01-12 17:17
marcanthat's not what that means#asahi-gpu · 2021-01-12 16:18
neobrainthough the "bigbanging magic numbers" bit sounds like it wouldn't be a particularly nice or performant driver#asahi-gpu · 2021-01-12 16:17
neobrainI'm actually surprised to hear implementing a Gallium driver requires less understanding of hardware than a Vulkan one, but then again I've never done the former :)#asahi-gpu · 2021-01-12 16:07
davidrysk[m]> LunarG compared Vulkan to "Gallium3D 2.0." << heh#asahi-gpu · 2021-01-12 16:02
davidrysk[m]they seem to be part of mesa#asahi-gpu · 2021-01-12 16:01
davidrysk[m]bloom: now I'm curious -- just how *are* vulkan drivers on Linux implemented#asahi-gpu · 2021-01-12 15:46
modwizcodeIt's the state machine for mesa right?#asahi-gpu · 2021-01-12 15:32
brentr123[m]what is gallian? ive never heard of that#asahi-gpu · 2021-01-12 15:31
j`eydejavu#asahi-gpu · 2021-01-12 15:31
bloomaustriancoder: right, the problem with that is -- afaiu -- Vulkan requires a deep understanding of the hardware, whereas I'm at the point I could throw together a gallium driver passing most of deqp-gles2 by just bitbanging magic numbers#asahi-gpu · 2021-01-12 13:51
bloomespecially on arm#asahi-gpu · 2021-01-12 13:50
bloomaustriancoder: yeah, that's the other thing, there's no point in paying vulkan tax if the entire userspace is going through zink, and that's one more layer of abstraction to debug#asahi-gpu · 2021-01-12 13:41
austriancoderbloom: passing most of deqp-gles2 .. I think this at least needs one more week for you. but yes I am in the same boat.. i know gallium quite well but never touched vulkan. also in linux userspace is more gles then vulkan atm#asahi-gpu · 2021-01-12 13:33
bloom(and more to the point - I am intimately familiar with Gallium, I do not know Vulkan)#asahi-gpu · 2021-01-12 13:33
bloomaustriancoder: right, the problem with that is -- afaiu -- Vulkan requires a deep understanding of the hardware, whereas I'm at the point I could throw together a gallium driver passing most of deqp-gles2 by just bitbanging magic numbers#asahi-gpu · 2021-01-12 07:53
marcan(I figured out this behavior via experimentation, then many years later when the official docs leaked, it turns out they don't document it properly... my docs are better than the official ones :))#asahi-gpu · 2021-01-12 07:52
marcanso if you don't align it ends up *before* the buffer#asahi-gpu · 2021-01-12 07:52
marcanthat reminds me of the wii NAND ECC calculation hardware... it ostensibly writes the calculated ECC buffer after the ECC spare buffer, but it actually writes it to the address XOR 0x40#asahi-gpu · 2021-01-12 07:51
marcanbloom: that makes sense#asahi-gpu · 2021-01-12 07:05
austriancoderbloom: Vulkan first :)#asahi-gpu · 2021-01-12 05:41
artemistLooking at benchmarks, Zink performans way better than I would expect#asahi-gpu · 2021-01-12 05:13
ConeOfAttack[m]Sorry, I was looking at an earlier discussion.#asahi-gpu · 2021-01-12 05:11
ConeOfAttack[m]... I figured they'd reuse as much between CPU and GPU as possible. So I was referring to ARM instructions.#asahi-gpu · 2021-01-12 05:08
davidrysk[m]it's a GPU, it's entirely custom instructions#asahi-gpu · 2021-01-12 05:08
ConeOfAttack[m]Quick question, I'm out of my league here but don't we already know Apple is running some custom instructions?#asahi-gpu · 2021-01-12 05:05
davidrysk[m]the device probably maps better to VK#asahi-gpu · 2021-01-12 03:00
bloomGL-on-VK is a thing but meh#asahi-gpu · 2021-01-12 03:00
brentr123[m]Ah#asahi-gpu · 2021-01-12 03:00
brentr123[m]Get the best of both worlds#asahi-gpu · 2021-01-12 03:00
bloomWith 2x the work..#asahi-gpu · 2021-01-12 03:00
brentr123[m]Anyway to do both?#asahi-gpu · 2021-01-12 02:57
bloomI am still undecided about GL or Vulkan first.#asahi-gpu · 2021-01-12 02:57
bloom^^ that#asahi-gpu · 2021-01-12 02:51
icecream95brentr123[m]: AFAIK the plan is to write a native OpenGL and/or Vulkan driver with no Metal involved#asahi-gpu · 2021-01-12 02:48
brentr123[m]So is metal going to be the graphics driver or will it be based off of moltenvk/molten gl?