Mateusz Kielan comments

Results 448 comments of


                                            Mateusz Kielan

Compute pathtracing PR

btw did you have to make any changes to the GLSL or was it a simple copy & paste from the previous example?

script: gen_profiles_file.py generating profiles doesn't work as expected

I think I know what the problem might be, the properties which are `false` are simply not being outputted by vulkan.gpuinfo.org, so the JSONs are invalid? Edit: this seems to...

shader cache polishing

Did you test the hash cache works nicely on example 23 (arithmetic test) and 62 CAD ?

shader cache polishing

@Erfan-Ahmadi you can merge, this will probably help you a lot with the cache size

Parallel GLSL Queue implementation

Hehehe, this is UB without Forward progress guarantees https://github.com/KhronosGroup/Vulkan-Docs/issues/2233 https://raphlinus.github.io/gpu/2020/04/30/prefix-sum.html

[SPIR-V] Why is the `vk::BufferPointer` itself getting loaded with `Align 4` ?

@danbrown-amd

Proposal: Re-think WindowManager API

you shouldn't be able to set the display mode dynamically on a window, should be done at creation time. Also on android, WebGPU, etc. you won't be able to create...

Proposal: Re-think WindowManager API

I'll let @Erfan-Ahmadi deal wtih this, just remember that windows need to keep the window manager alive. And on Win32 and other fun legacy platforms, every window method needs to...

[Feature Request] What's the way to set `OpMemoryModel` ?

@s-perron I see that in https://github.com/microsoft/DirectXShaderCompiler/commit/06da989c4ab615cf7ea4bd1c5c1de00d0f7339e6 you added `-fspv_use_vulkan_memory_model` but its not a recognised option anymore? https://godbolt.org/z/6rdheznWP

[Feature Request] What's the way to set `OpMemoryModel` ?

btw I've ran into limitations of not being able to codegen VMM directly before (being gatekept by the UpdateMemorModelPass - having the intial codegen happen in GLSL450 model) https://github.com/microsoft/DirectXShaderCompiler/issues/7533