So the Oculus Rift SDK accepts a framebuffer texture ID which is created with the glFramebufferTexture function.
Example:
[...]
GLuint l_FBOId;
glGenFramebuffers(1, &l_FBOId);
glBindFramebuffer(GL_FRAMEBUFFER, l_FBOId);
// The texture we're going to render to...
GLuint l_TextureId;
glGenTextures(1, &l_TextureId);
// "Bind" the newly created texture : all future texture functions will modify this texture...
glBindTexture(GL_TEXTURE_2D, l_TextureId);
// Give an empty image to OpenGL (the last "0")
glTexImage2D(GL_TEXTURE_2D, 0, GL_RGBA, l_TextureSize.w, l_TextureSize.h, 0, GL_RGBA, GL_UNSIGNED_BYTE, 0);
// Linear filtering...
glTexParameteri(GL_TEXTURE_2D, GL_TEXTURE_MAG_FILTER, GL_LINEAR);
glTexParameteri(GL_TEXTURE_2D, GL_TEXTURE_MIN_FILTER, GL_LINEAR);
// Create Depth Buffer...
GLuint l_DepthBufferId;
glGenRenderbuffers(1, &l_DepthBufferId);
glBindRenderbuffer(GL_RENDERBUFFER, l_DepthBufferId);
glRenderbufferStorage(GL_RENDERBUFFER, GL_DEPTH_COMPONENT, l_TextureSize.w, l_TextureSize.h);
glFramebufferRenderbuffer(GL_FRAMEBUFFER, GL_DEPTH_ATTACHMENT, GL_RENDERBUFFER, l_DepthBufferId);
[...]
/*
Oculus Rift API
*/
ovrGLTexture l_EyeTexture[2];
l_EyeTexture[0].OGL.Header.API = ovrRenderAPI_OpenGL;
l_EyeTexture[0].OGL.Header.TextureSize.w = l_TextureSize.w;
l_EyeTexture[0].OGL.Header.TextureSize.h = l_TextureSize.h;
l_EyeTexture[0].OGL.Header.RenderViewport = l_Eyes[0].RenderViewport;
l_EyeTexture[0].OGL.TexId = l_TextureId;
[...]
However, the glFramebufferTexture function is only available in OpenGL 3.2+.
Is it possible to create a workaround that will be functionally equivalent and acceptable to the Oculus Rift SDK?