[Feature Request] Support all the GLSL unpack intrinsics in HLSL (including unpackUnorm4x8)

[jmoguill]

Is your feature request related to a problem? Please describe.
For example, GLSL provides unpackUnorm4x8.
HLSL's closest equivalent is unpack_u8u32 but this doesn't return normalized result, and returns as uint4 not float4.
Can HLSL support all the same unpack intrinsics as GLSL?
For example, if we're using HLSL with Vulkan, is there a performance tradeoff using HLSL vs GLSL (if we want to use unpackUnorm4x8)? Or can we accomplish same thing (with same performance) via HLSL?

Describe the solution you'd like
Support all the GLSL unpack intrinsics via HLSL

Describe alternatives you've considered
A clear and concise description of any alternative solutions or features you've considered.

Additional context
Add any other context or screenshots about the feature request here.

[s-perron]

The method we chose to expose individual SPIR-V instructions is to use inline spir-v. This works for all of the pack and unpack instructions. I have attached an example that show how to use all of the unpack instruction. You can write the definitions of the functions in a header file that you include in all of your shader that need to use them.

unpack_shader.hlsl.txt

[s-perron]

For example, if we're using HLSL with Vulkan, is there a performance tradeoff using HLSL vs GLSL (if we want to use unpackUnorm4x8)? Or can we accomplish same thing (with same performance) via HLSL?

I would expect the performance to be the same for both GLSL and HLSL.

[jmoguill]

Thanks @s-perron for the suggestions!