Paravirtualization has a lot of promise, in particular in its ability to deliver performance by allowing the hypervisor to be aware of the idioms in the operating system. Since kernel changes are necessary, it is very easy to get into a situation where the paravirtualized kernel is incapable of executing on a native machine, or on another hypervisor. It is also quite easy to expose too many hypervisor implementation details in the name of performance, which can impede the general development of the kernel with many hypervisor specific subtleties.
VMI, or the Virtual Machine Interface, is a clearly defined extensible specification for OS communication with the hypervisor. VMI delivers great performance without requiring that kernel developers be aware of concepts that are only relevant to the hypervisor. As a result, it can keep pace with the fast releases of the Linux kernel and a new kernel version can be trivially paravirtualized. With VMI, a single Linux kernel binary can run on a native machine and on one or more hypervisors.
In this paper, we discuss a working patch to Linux 2.6.16, the latest version of Linux as of this writing. We present performance data on native to show the negligible cost of VMI and on the VMware hypervisor to show its overhead compared with native. We also share some future work directions.