Getting "incorrect number of arguments to numeric-type constructor" error in a shader

I've been learning Unity shader for couple of days and decided to port this oil shader from Shadertoy.

I get the following error:

incorrect number of arguments to numeric-type constructor at line 69 (on glcore)

Here is the relevant part of the shader code:

#define T (_Time/3.+5.)

fixed4 frag(v2f i) : SV_Target
{
    float4 fragColor = 0;
    float2 fragCoord = i.vertex.xy;

    float4 k = fragColor;
    float2 p = fragCoord;

    float s = 1.;

    #define rot(p,a) float2 sc = sin(float2(a,a + 1.6));  p *= float2x2(sc.y,-sc.x,sc);

    #define A float3(0,1,157)
    #define B {float2 m = frac(p),l = dot(p - m,A.yz) + A.xz,r = lerp(frac(57.*sin(l++)),frac(57.*sin(l)),(m *= m*(3. - m - m)).x); k += lerp(r.x,r.y,m.y) / (s += s); p *= float4(1,1,1,-1);}

    p *= log(T) / R.y;              // scaling (slow zoom out)
    p.x += T;                       // translation
    rot(p, T / 22.); //ERROR HERE

    //......the rest of the code below
}

The rot(p, T / 22.); line is where the error is at. The funny thing is that when I remove or comment-out this line of code, the shader compiles and is working fine. I still want to know why that line can't compile.

What's causing that error and how can I fix it?

EDIT:

As pointed out by helium, float3x3 is a matrix.

Below is the original rot function:

#define rot(p,a) vec2 sc=sin(vec2(a,a+1.6)); p*=mat2(sc.y,-sc.x,sc);

And the ported version:

#define rot(p,a) float2 sc = sin(float2(a,a + 1.6));  p *= float2x2(sc.y,-sc.x,sc);

Notice how I replaced mat2 with float2x2 because that's the equivalent type in HLSL.

I do believe that the problem is from float2x2(sc.y,-sc.x,sc);. What is the appropriate way to do this and still use the 3 parameters as the original code did?

Solution

The built-in value _Time is a float4 so you have to pick one of the components depending on the time scale you want:

_Time float4 (t/20, t, t*2, t*3), use to animate things inside the shaders.

for example #define T (_Time.y/3.+5.).

Another thing is the * operator for float2x2 represents a component-wise product of 2 matrices. For vector-matrix/matrix-matrix product you should use the mul function.

So in the rot function you should have p =mul(p, float2x2(sc.y,-sc.x,sc));.

For more info on this you can go to: https://en.wikibooks.org/wiki/Cg_Programming/Vector_and_Matrix_Operations.