This post can be read here, but there isn't too much of value in this particular post. Though Zack does state, "As it stands we have an excellent framework in Gallium3D but not a lot of open drivers for it. Ironically it's our new software driver, llvmpipe, or more precisely a mutation of it, which has the potential to fix some of our GPU issues in the future. With the continues generalization of GPUs my hope is that all we'll need is DRM code (memory management, kernel modesetting, command submission) and LLVM->GPU code generator. It's not exactly a trivial amount of code by any stretch of the imagination but smaller than what we'd need right now and once it would be done for one or two GPUs it would certainly become a lot simpler. Plus GPGPU will eventually make the latter part mandatory anyway. Having that would get us a working driver right away and after that we could play with texture sampling and vertex paths (which will likely stay as dedicated units for a while) as optimizations."
The possibility of using a mutation of LLVMpipe to submit commands to the GPU has been brought up somewhat recently in our forums particularly after delivering updated LLVMpipe benchmarks. Next week we will also have out more LLVMpipe results from an Intel Core i3 system as we look at its performance with and without Hyper Threading, at different clock frequencies, and how it compares to Intel's Arrandale graphics core found on these new processors.