Question

I am pretty sure the answer is no, but I just wanted to make sure as I don’t want memory leaks.

I am using the following code

__constant__ void* VERTEX_NO_CONSTANT_PARAMETER;
HANDLE_ERROR( cudaMemcpyToSymbol( VERTEX_NO_CONSTANT_PARAMETER, &vertexNo, sizeof( int ) ) );
HANDLE_ERROR( cudaFree( VERTEX_NO_CONSTANT_PARAMETER ) );

It does not throw any errors at me which is putting doubt in my mind (I was hoping that cudaFree would error).

Thank you!

Kevin

Answer 1

No, you don’t. According to NVIDIA cuda library:

cudaFree (void *devPtr) Frees the memory space pointed to by devPtr,
which must have been returned by a previous call to cudaMalloc() or
cudaMallocPitch() […]