OpenGL texture into CUDA

Hi folks,

Sorry if this has been asked before.

I’m rendering to a texture in OpenGL (via a FBO) and I want to do some processing with the information in that texture using CUDA. What is the fastest way to make that texture accessible to CUDA?