#
✕ nick:davidrysk[m]
#asahi-gpu · 2021-01-14 16:58
davidrysk[m]cool :)
#asahi-gpu · 2021-01-14 16:56
davidrysk[m](on the end of the person attempting, not on your end)
#asahi-gpu · 2021-01-14 16:56
davidrysk[m]bloom: that's fine, though I'm curious if there's feedback when someone attempts
#asahi-gpu · 2021-01-14 16:40
davidrysk[m]gross!
#asahi-gpu · 2021-01-14 14:45
davidrysk[m]okie :)
#asahi-gpu · 2021-01-14 14:14
davidrysk[m]bloom: can you intercept the result values with wrap?
#asahi-gpu · 2021-01-14 14:14
davidrysk[m]hthh_: you're saying that's intentional of Apple's API?
#asahi-gpu · 2021-01-13 18:36
davidrysk[m]for what bloom is doing? makefiles are more than sufficient :)
#asahi-gpu · 2021-01-13 18:36
davidrysk[m]cmake et al are useful for complex projects that pull in lots of dependencies and where you need to support Windows well
#asahi-gpu · 2021-01-13 18:32
davidrysk[m]my opinion? plain ol' makefiles are best for simple stuff
#asahi-gpu · 2021-01-13 18:32
davidrysk[m]cmake can use pkg-config -- https://cmake.org/cmake/help/latest/module/FindPkgConfig.html
#asahi-gpu · 2021-01-13 16:53
davidrysk[m]cmake greatly improved abstraction in version 3
#asahi-gpu · 2021-01-13 15:37
davidrysk[m]yeah that's the thing, cmake scripting is weird
#asahi-gpu · 2021-01-13 15:23
davidrysk[m]downside of cmake is a somewhat weird domain-specific language
#asahi-gpu · 2021-01-13 15:23
davidrysk[m]both meson and modern cmake are good... a lot of the criticisms of cmake from the meson side are due to older cmake syntax/design; modern cmake fixes most of that and is generally more flexible
#asahi-gpu · 2021-01-13 13:10
davidrysk[m]Or install using Macports or homebrew
#asahi-gpu · 2021-01-13 13:09
davidrysk[m](But here — there’s no need, we’re not doing anything crazy)
#asahi-gpu · 2021-01-13 13:09
davidrysk[m](I tend to be partial to cmake over meson myself)
#asahi-gpu · 2021-01-13 13:09
davidrysk[m]No cmake or meson
#asahi-gpu · 2021-01-13 13:08
davidrysk[m]Im suggesting plain make at this point
#asahi-gpu · 2021-01-13 13:08
davidrysk[m]I can write some example makefiles
#asahi-gpu · 2021-01-13 04:41
davidrysk[m]bloom: have you thought about adding Makefiles? :)
#asahi-gpu · 2021-01-13 04:41
davidrysk[m]Basically one would compile a simple shader, intercepting the system calls that push it to the GPU, and figure out the format, instruction set, etc
#asahi-gpu · 2021-01-13 04:39
davidrysk[m]take a look at bloom's article here: https://rosenzweig.io/blog/asahi-gpu-part-1.html
#asahi-gpu · 2021-01-13 04:36
davidrysk[m]People here are already doing blackbox RE of the GPU
#asahi-gpu · 2021-01-13 04:36
davidrysk[m]Yeah. That's more on topic for #asahi though.
#asahi-gpu · 2021-01-13 04:33
davidrysk[m]The documentation is there, but the tooling doesn't function yet
#asahi-gpu · 2021-01-13 04:33
davidrysk[m]The planned goal is to be able to dual boot, but Apple hasn't yet provided the tooling for signing a custom OS object.
#asahi-gpu · 2021-01-13 03:04
davidrysk[m]what do you normally see when AGX_SELECTOR_SET_API is used?
#asahi-gpu · 2021-01-13 02:33
davidrysk[m]I started digging at trying to figure out how it does error handling but got sidetracked :)
#asahi-gpu · 2021-01-13 02:31
davidrysk[m]bloom: are you still poking at IOConnectCallMethod?
#asahi-gpu · 2021-01-13 01:09
davidrysk[m]:(
#asahi-gpu · 2021-01-13 01:08
davidrysk[m]IOServiceCallMethod is what you're looking for?
#asahi-gpu · 2021-01-13 01:08
davidrysk[m]can you repeat the whole thing?
#asahi-gpu · 2021-01-13 01:00
davidrysk[m]IOServiceCallMethod?
#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 05:08
davidrysk[m]it's a GPU, it's entirely custom instructions
#asahi-gpu · 2021-01-12 05:05
davidrysk[m]the device probably maps better to VK
#asahi-gpu · 2021-01-11 17:33
davidrysk[m]bloom: "Does that require dual licensing the other files too?" APSL sections 1.5 and 4 seem to say no
#asahi-gpu · 2021-01-11 16:04
davidrysk[m]bloom: Is there any documentation on retrieving compiled and assembled object code of a shader?
#asahi-gpu · 2021-01-11 02:37
davidrysk[m]found this with a quick google :)
#asahi-gpu · 2021-01-11 02:37
davidrysk[m]https://developer.amd.com/wordpress/media/2012/10/gdc2008_ribble_maurice_TileBasedGpus.pdf
#asahi-gpu · 2021-01-11 00:34
davidrysk[m]I wish we knew more about Apple's (CPU) microarchitecture. From what I can tell, they're still using the scheduler model that was intended for Cyclone (A7) in LLVM, and that chip had much less instruction level parallelism available.
#asahi-gpu · 2021-01-11 00:32
davidrysk[m]This was making the rounds the other day, but these tests are too elementary to really provide any interesting info
#asahi-gpu · 2021-01-11 00:32
davidrysk[m]https://lemire.me/blog/2021/01/06/memory-access-on-the-apple-m1-processor/
#asahi-gpu · 2021-01-10 23:57
davidrysk[m](in their WWDC talks on Metal and their GPUs)
#asahi-gpu · 2021-01-10 23:57
davidrysk[m]makes sense — Apple has been very specific that they use tiling heavily
#asahi-gpu · 2021-01-10 23:57
davidrysk[m]:) nice!

plus ›