Hardware Acceleration and Togl

I’m trying to use togl (an interface for openGL with Tcl/Tk) and was wondering about using it to get hardware accelerated rendering. I remember once upon a time creating a pbuffer on Windows using some wgl extensions and using glReadPixels and glDrawPixels to send and receive buffers to/from the card. Can anyone point me to a tutorial on how to set this up using the glx extensions (I’m now running on Linux)? Perhaps there’s another option - does double buffering work equally fast as rendering to a pbuffer (or is that just software rendering)? Togl seems pretty basic in its initialization so I’m guessing it does not do hardware accelerated rendering unless its tweaked to do so. We’re trying to do real-time rendering, ideally using constructive solid geometry. Any hints on how to proceed would be greatly appreciated.