Mateusz Kielan
Mateusz Kielan
> The idea behind float2 being SSE is that we never convert back and forth between types. You've saved a register pack/unpack (or unaligned load, masked store), but now you're...
> This is really up to you, I wouldn't want it in all honesty but I wonder if you have hard requirements that hlsl++ doesn't meet and I can't fix....
> ```c++ > enum class Packing > { > Scalar, // align to component type > ConstantBuffer, // std140 in GLSL > StructuredBuffer // std430 in GLSL > }; >...
support for `scalarBlockLayout` seems to be ubiquitous across all Vulkan devices now, if you're targetting SPIR-V with DXC/HLSL use `-fvk-use-scalar-layout` however your `HLSLPP_SCALAR` does not mean that `alignof(float3)==sizeof(float)` and `sizeof(float3)==12`...
I have an implementation of this in my shoddy matrix class https://github.com/Devsh-Graphics-Programming/Nabla/blob/6dc8448ba5e5cef21b7eecc52038b389b8dc96a1/include/matrix3x4SIMD_impl.h#L318
> Hi @CUSkidmore it is always my plan to expand hlsl++ beyond its capabilities and make it as useful as possible. It isn't a priority right now as it's a...
>which as you said are multiplying by vectors and transposing I was refering to "turn row major into columns and do matrix-vector mult as FMA" as a particular strategy. My...
that would require DXC to spec its name mangling though...
but would be good if DXC gave you a warning if it encounters a function with same name as your entry point which is in a namespace
Replace `core::SRange` with `std::span` and other templated usages with `std::view` concept template
@pollend your wish was granted, now I use spans in a lot of places in `video` namespace on the `vulkan_1_3` branch