On 02/13/2018 06:27 PM, Dr. David Alan Gilbert wrote: > * Roman Kagan (rka...@virtuozzo.com) wrote: >> On Tue, Feb 13, 2018 at 03:05:03PM +0000, Dr. David Alan Gilbert wrote: >>> * Denis V. Lunev (d...@virtuozzo.com) wrote: >>>> On 02/13/2018 05:59 PM, Dr. David Alan Gilbert wrote: >>>>> * Daniel P. Berrangé (berra...@redhat.com) wrote: >>>>>> That doesn't seem practical unless you can instantaneously write out >>>>>> the entire guest RAM to disk without blocking, or can somehow snapshot >>>>>> the RAM so you can write out a consistent view of the original RAM, >>>>>> while the guest continues to dirty RAM pages. >>>>> People have suggested doing something like that with userfault write >>>>> mode; but the same would also be doable just by write protecting the >>>>> whole of RAM and then following the faults. >>>> nope, userfault fd does not help :( We have tried, the functionality is not >>>> enough. Better to have small extension to KVM to protect all memory >>>> and notify QEMU with accessed address. >>> Can you explain why? I thought the write-protect mode of userfaultfd was >>> supposed to be able to do that; cc'ing in Andrea >> IIRC it would if it worked; it just didn't when we tried. > Hmm that doesn't seem to be the ideal reason to create new KVM > functionality, especially since there were a bunch of people wanting the > userfaultfd-write mode for other uses. This does not seems to work in practice. We status with that is the same for more than a year and a half AFAIR. Thus there is a lot of sense to create something feasible as async snapshots are badly wanted.