Hello, Last week I was working on restoring vmspaces of virtual processes in order to be able to run them after vkernel is restored from a checkpoint. Currently, vmspaces are simply recreated using cpu_vmspace_alloc() function and the correct one is chosen by calling vmspace_ctl(). There are still some issues that have to be addressed but virtual processes continue executing after restoring the vkernel from a checkpoint.
I also did some work necessary to get vkernel interrupts working. When restoring a program from a checkpoint flags of every open file are properly reset (so that kqueue used by vkernels have O_ASYNC set). Also, vkernel checkpoint signal handler sets kqueue owner the vkernel new PID. Paweł
