Error 'overlapping register semantics not yet implemented' in VertexShader

Question

I am trying to perform diffuse reflection in hlsl. Currently I am working on vertex shader. Unfortunatelly I get following error, when trying to compile with fxc.exe:

C:\Users\BBaczek\Projects\MyApp\VertexShader.vs.hlsl(2,10-25)
: error X4500: overlapping register semantics not yet implemented 'c1'
C:\Users\BBaczek\Projects\MyApp\VertexShader.vs.hlsl(2,10-25)
: error X4500: overlapping register semantics not yet implemented 'c2'
C:\Users\BBaczek\Projects\MyApp\VertexShader.vs.hlsl(2,10-25)
: error X4500: overlapping register semantics not yet implemented 'c3'

Vertex shader code:

    float4x4 WorldViewProj : register(c0);
float4x4 inv_world_matrix : register(c1);
float4 LightAmbient;
float4 LightPosition;


struct VertexData
{
    float4 Position : POSITION;
    float4 Normal : NORMAL;
    float3 UV : TEXCOORD;
};

struct VertexShaderOutput
{
    float4 Position : POSITION;
    float3 Color: COLOR;
};

VertexShaderOutput main(VertexData vertex)
{
    VertexShaderOutput output;


    vertex.Normal = normalize(vertex.Normal);

    float4 newColor = LightAmbient;

    vector obj_light = mul(LightPosition, inv_world_matrix);
    vector LightDir = normalize(obj_light - vertex.Position);
    float DiffuseAttn = max(0, dot(vertex.Normal, LightDir));
    vector light = { 0.8, 0.8, 0.8, 1 };
    newColor += light * DiffuseAttn;

    output.Position = mul(vertex.Position, WorldViewProj);
    output.Color = float3(newColor.r, newColor.g, newColor.b);
    return output;
}

And command I use to perform compilation:

fxc /T vs_2_0 /O3 /Zpr /Fo VertexShader.vs VertexShader.vs.hlsl

Why am I getting this error? What can I do to prevent this?

Answer 1

Found it out - I am not deleting this question, because someone might find it useful.

What you need to do is changing

float4x4 WorldViewProj : register(c0);
float4x4 inv_world_matrix : register(c1);

to

float4x4 WorldViewProj : register(c0);
float4x4 inv_world_matrix : register(c4);

I am not sure what is that ok, but I assume that float4x4 is going to take more space in that buffer (4x4 - so it takes 4 places). I think that explanation is a bit silly but it works.