struct VS_OUTPUT        {
    float4 Position             : SV_Position;
    float4 clip0                : SV_ClipDistance0;  // multiple semantic IDs, two vec4s (no extra packing)
    float4 clip1                : SV_ClipDistance1;  // ...
};

float4 main(VS_OUTPUT v) : SV_Target0
{
    return v.Position + v.clip0 + v.clip1;
}