On 07/19/2011 05:14 PM, Joerg Roedel wrote:
On Tue, Jul 19, 2011 at 04:55:53PM +0300, Avi Kivity wrote: > On 07/19/2011 04:54 PM, Avi Kivity wrote: >> On 07/19/2011 04:30 PM, Joerg Roedel wrote: >>> Hmm, I planned to do the VMSTATE thing in a follow-on patch-set. The >>> plan is to read the VCPU tsc_freq at guest start time on !tsc-scale >>> hosts and migrate it over so that the destination host can set the >>> tsc-freq if it supports tsc-scaling. >> >> This can be done by a management tool if desired. >> > > Although, if we do this unconditionally (that is, also for tsc-scale > hosts) then we get stable tsc even without supplying a tsc frequency > argument... need to think about this. It has the advantage that it "just works", without the need to extend management tools and the like. And it makes migration more transparent to the guests.
Yes, exactly. The flip side is that automagic stuff is sometimes unexpected and leads to breakage. I'm not sure what the right thing is.
-- error compiling committee.c: too many arguments to function -- To unsubscribe from this list: send the line "unsubscribe kvm" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
