public static void complexAbs(GThread thread, ComplexF[,] a, ComplexF[,] c)
{
int x = thread.blockIdx.x;
int y = 0;
float j = 1.0F;
float k = 2.0F;
ComplexF cf = new ComplexF(3 + j, 7 + k);
while (y < YSIZE)
{
c[x, y].x = ComplexF.Abs(a[x, y]);
c[x, y].y = cf.y;// 0.0F;
y++;
}
}
public static void Execute()
{
CudafyModule km = CudafyTranslator.Cudafy();
GPGPU gpu = CudafyHost.GetDevice(CudafyModes.Target);
gpu.LoadModule(km);
// 2D
Console.WriteLine("kernel");
ComplexF[,] host_A = new ComplexF[XSIZE, YSIZE];
ComplexF[,] host_B = new ComplexF[XSIZE, YSIZE];
ComplexF[,] host_C = new ComplexF[XSIZE, YSIZE];
int i = 0;
for (int x = 0; x < XSIZE; x++)
{
for (int y = 0; y < YSIZE; y++)
{
host_A[x, y] = new ComplexF(i, i);
host_B[x, y] = new ComplexF(2, 0);
i++;
}
}
ComplexF[,] dev_A = gpu.CopyToDevice(host_A);
ComplexF[,] dev_B = gpu.CopyToDevice(host_B);
ComplexF[,] dev_C = gpu.Allocate <ComplexF>(XSIZE, YSIZE);
Console.WriteLine("complexAdd");
gpu.Launch(XSIZE, 1, "complexAdd", dev_A, dev_B, dev_C);
gpu.CopyFromDevice(dev_C, host_C);
i = 0;
bool pass = true;
for (int x = 0; x < XSIZE; x++)
{
for (int y = 0; y < YSIZE && pass; y++)
{
ComplexF expected = ComplexF.Add(host_A[x, y], host_B[x, y]);
pass = host_C[x, y].x == expected.x && host_C[x, y].y == expected.y;
}
}
Console.WriteLine(pass ? "Pass" : "Fail");
Console.WriteLine("complexSub");
gpu.Launch(XSIZE, 1, "complexSub", dev_A, dev_B, dev_C);
gpu.CopyFromDevice(dev_C, host_C);
i = 0;
pass = true;
for (int x = 0; x < XSIZE; x++)
{
for (int y = 0; y < YSIZE && pass; y++)
{
ComplexF expected = ComplexF.Subtract(host_A[x, y], host_B[x, y]);
pass = host_C[x, y].x == expected.x && host_C[x, y].y == expected.y;
}
}
Console.WriteLine(pass ? "Pass" : "Fail");
Console.WriteLine("complexMpy");
gpu.Launch(XSIZE, 1, "complexMpy", dev_A, dev_B, dev_C);
gpu.CopyFromDevice(dev_C, host_C);
i = 0;
pass = true;
for (int x = 0; x < XSIZE; x++)
{
for (int y = 0; y < YSIZE && pass; y++)
{
ComplexF expected = ComplexF.Multiply(host_A[x, y], host_B[x, y]);
//Console.WriteLine("{0} {1} : {2} {3}", host_C[x, y].R, host_C[x, y].I, expected.R, expected.I);
pass = Verify(host_C[x, y], expected, 1e-14F);
i++;
}
}
Console.WriteLine(pass ? "Pass" : "Fail");
Console.WriteLine("complexDiv");
gpu.Launch(XSIZE, 1, "complexDiv", dev_A, dev_B, dev_C);
gpu.CopyFromDevice(dev_C, host_C);
i = 0;
pass = true;
for (int x = 0; x < XSIZE; x++)
{
for (int y = 0; y < YSIZE && pass; y++)
{
ComplexF expected = ComplexF.Divide(host_A[x, y], host_B[x, y]);
//Console.WriteLine("{0} {1} : {2} {3}", host_C[x, y].R, host_C[x, y].I, expected.R, expected.I);
if (i > 0)
{
pass = Verify(host_C[x, y], expected, 1e-13F);
}
i++;
}
}
Console.WriteLine(pass ? "Pass" : "Fail");
Console.WriteLine("complexAbs");
gpu.Launch(XSIZE, 1, "complexAbs", dev_A, dev_C);
gpu.CopyFromDevice(dev_C, host_C);
i = 0;
pass = true;
for (int x = 0; x < XSIZE; x++)
{
for (int y = 0; y < YSIZE && pass; y++)
{
float expected = ComplexF.Abs(host_A[x, y]);
pass = Verify(host_C[x, y].x, expected, 1e-2F);
//Console.WriteLine("{0} {1} : {2}", host_C[x, y].x, host_C[x, y].y, expected);
i++;
}
}
Console.WriteLine(pass ? "Pass" : "Fail");
gpu.FreeAll();
}