On Mon, Nov 13, 2017 at 06:05:59PM +0800, Quan Xu wrote:ÂÂ Konrad, I can't follow 'b/c' here.. sorry.
From: Yang Zhang <yang.zhang.wz@xxxxxxxxx>Meaning an VMEXIT b/c it is an 'halt' operation ? And then going
Some latency-intensive workload have seen obviously performance
drop when running inside VM. The main reason is that the overhead
is amplified when running inside VM. The most cost I have seen is
inside idle path.
back in guest (VMRESUME) takes time. And hence your latency gets
all whacked b/c of this?
So if I understand - you want to use your _full_ timeslice (of the guest)ÂÂÂ as much as possible.
without ever (or as much as possible) to go in the hypervisor?
Which means in effect you don't care about power-saving or CPUfreq actually, we care about power-saving. The poll duration is self-tuning, otherwise it is almost as the same as
savings, you just want to eat the full CPU for snack?
 rescheduled of guest scheduler..This patch introduces a new mechanism to poll for a while beforeSchedule of what? The guest or the host?
entering idle state. If schedule is needed during poll, then we
don't need to goes through the heavy overhead path.