nsls' build-ins are translated to late in the shader generation process where a build-in name is replace by the same name if the target shader has the same build-in name. So the generation process will be caught in an infinite loop. This could be handled differently, but leads to other problems.
So lets just rename all build-in names in the nsl shader so that there will be no ambiguity in later stages in the generation process.
All build-ins should be transformed their respective name + some special sign like step -> :step:. So the name :step: becomes the opcode of the function.
nsls' build-ins are translated to late in the shader generation process where a build-in name is replace by the same name if the target shader has the same build-in name. So the generation process will be caught in an infinite loop. This could be handled differently, but leads to other problems.
So lets just rename all build-in names in the nsl shader so that there will be no ambiguity in later stages in the generation process.
All build-ins should be transformed their respective name + some special sign like
step->:step:. So the name:step:becomes the opcode of the function.