public static void Execute(byte[] bitmap)
{
CudafyModule km = CudafyModule.TryDeserialize();
if (km == null || !km.TryVerifyChecksums())
{
km = CudafyTranslator.Cudafy(typeof(SphereOpenCL), typeof(ray_opencl_const));
km.TrySerialize();
}
GPGPU gpu = CudafyHost.GetDevice(CudafyModes.Target, CudafyModes.DeviceId);
gpu.LoadModule(km);
// capture the start time
gpu.StartTimer();
// allocate memory on the GPU for the bitmap (same size as ptr)
byte[] dev_bitmap = gpu.Allocate(bitmap);
// allocate memory for the Sphere dataset
//SphereOpenCL[] s = gpu.Allocate<SphereOpenCL>(SPHERES);
// allocate temp memory, initialize it, copy to constant memory on the GPU
SphereOpenCL[] temp_s = new SphereOpenCL[SPHERES];
for (int i = 0; i < SPHERES; i++)
{
temp_s[i].r = rnd(1.0f);
temp_s[i].g = rnd(1.0f);
temp_s[i].b = rnd(1.0f);
temp_s[i].x = rnd(1000.0f) - 500;
temp_s[i].y = rnd(1000.0f) - 500;
temp_s[i].z = rnd(1000.0f) - 500;
temp_s[i].radius = rnd(100.0f) + 20;
}
//gpu.CopyToDevice(temp_s, s);
gpu.CopyToConstantMemory(temp_s, spheres);
// generate a bitmap from our sphere data
dim3 grids = new dim3(ray_gui.DIM / 16, ray_gui.DIM / 16);
dim3 threads = new dim3(16, 16);
//gpu.Launch(grids, threads).kernel(s, dev_bitmap); // Dynamic
gpu.Launch(grids, threads, ((Action <GThread, byte[]>)thekernel), dev_bitmap); // Strongly typed
// copy our bitmap back from the GPU for display
gpu.CopyFromDevice(dev_bitmap, bitmap);
// get stop time, and display the timing results
float elapsedTime = gpu.StopTimer();
Console.WriteLine("Time to generate: {0} ms", elapsedTime);
gpu.FreeAll();
}
public static float hit(SphereOpenCL s, float ox1, float oy1, ref float n1)
{
float dx = ox1 - s.x;
float dy = oy1 - s.y;
if (dx * dx + dy * dy < s.radius * s.radius)
{
float dz = GMath.Sqrt(s.radius * s.radius - dx * dx - dy * dy);
n1 = dz / GMath.Sqrt(s.radius * s.radius);
return(dz + s.z);
}
return(-2e10f);
}
