Displaying an array of floats with OpenGL Interop

Accelerated Computing CUDA CUDA Programming and Performance
1

Hi everybody!

Sorry if this topic has been discussed but I did not find anything External Image

My problem is simple. I have an image stored in and array of float, I mean

float4 image

I want to display it using OpenGl interop, how can I achieve this? I tried with:

cudaGraphicsMapResources(1, &pboCUDA, 0);

size_t num_bytes; 

cudaGraphicsResourceGetMappedPointer((void**)&image, &num_bytes, pboCUDA);

kernel<<<g1,b1>>>(image, ...);

cudaGraphicsUnmapResources(1, &pboCUDA, 0);

and the init code

GLuint pbo;

struct cudaGraphicsResource* pboCUDA;

glGenBuffers(1, &pbo);

glBindBuffer(GL_ARRAY_BUFFER, pbo);

unsigned int size = width* height * 4 * sizeof(float);

glBufferData(GL_ARRAY_BUFFER, size, 0, GL_DYNAMIC_DRAW);

glBindBuffer(GL_ARRAY_BUFFER, 0);

cudaGraphicsGLRegisterBuffer(&pboCUDA, pbo, cudaGraphicsMapFlagsWriteDiscard);

drawing code

glClear(GL_COLOR_BUFFER_BIT | GL_DEPTH_BUFFER_BIT);

glBindBuffer(GL_ARRAY_BUFFER, pbo);

glVertexPointer(4, GL_FLOAT, 0, 0);

glEnableClientState(GL_VERTEX_ARRAY);

glDrawArrays(GL_POINTS, 0, width* height);

glDisableClientState(GL_VERTEX_ARRAY);

but I found an error trying to register de cudaGraphicsResource…do you know how I can simply display the image?

Because I found that OpenGLs primitives are not fast…

Thank you very much