When I am debugging and looking at the trace-point output for a guest, there is no way for me to determine whether a given trace-point, say, {vmx|svm}_handle_exit() or {vmx|svm}_vcpu_run(), was executed as part of which L1/L2 guest. As VMs do such operations so many times in a short period of time, trace-point output looks overwhelming and is of no use when analyzed from this angle.
Does it makes sense to embed some sort of guest identification within the VMCS/VMCB, say, loaded_vmcs or vmcb, so that trace-points can spit out which guest/nested-guest executed the operation ?
-Krish
