WGSL compilation: usage of

[aleino-nv]

The following tests fail to produce WGSL due to usage of each corresponding function

• 'AllMemoryBarrier'
  • compute\texture-subscript
• 'AllMemoryBarrierWithGroupSync'
  • compute\half-rw-texture-convert2
• 'GetDimensions'
  • compute\texture-get-dimensions
• 'GroupMemoryBarrierWithGroupSync'
  • compute\groupshared
• 'InterlockedAdd'
  • compute\nonuniformres-atomic
• 'Load'
  • compute\constant-buffer-memory-packing
  • compute\dynamic-dispatch-13
  • compute\dynamic-dispatch-14
  • compute\dynamic-dispatch-15
  • compute\dynamic-dispatch-16
  • compute\dynamic-dispatch-17
  • compute\dynamic-dispatch-bindless-texture
  • compute\nonuniformres-array-of-textures
• 'NonUniformResourceIndex'
  • compute\nonuniformres-nested-rwstructuredbuf
  • compute\unbounded-array-of-array
  • compute\unbounded-array-of-array-syntax
• 'Sample'
  • compute\compile-time-loop
  • compute\discard-stmt
  • compute\texture-sampling
• 'SampleGrad'
  • compute\texture-sample-grad-offset-clamp
• 'SampleLevel'
  • compute\texture-simple
• 'SampleLevel'
  • compute\texture-simpler
• 'TraceRayInline'
  • compute\ray-tracing-inline
• 'all'
  • compute\texture-subscript-multisample
  • compute\vector-scalar-compare
• 'any'
  • compute\half-vector-compare
• 'asint'
  • compute\structured-buffer-of-matrices
• 'asuint'
  • compute\pack-any-value-16bit
• 'f32tof16_'
  • compute\half-opaque-convert
• 'max'
  • compute\ssa-reduce-bug
• 'mul'
  • compute\column-major
  • compute\default-major
  • compute\kernel-context-threading
  • compute\non-square-column-major
  • compute\non-square-row-major
  • compute\row-major
  • compute\static-const-matrix-array
• 'sin'
  • compute\transcendental
  • compute\transcendental-double

[aleino-nv]

I just resolved the asint issues. Took a while to figure out how to specify stuff for the capability system, but I hope the rest of the functions will go pretty smoothly now.

[aleino-nv]

This should be handled by #5053