On Wed, Mar 10, 2021 at 11:44 AM Paolo Bonzini pbonzini@redhat.com wrote:
On 10/03/21 18:31, Marc Zyngier wrote:
Maintaining VM-global counters would require an atomic instruction and would suffer lots of cacheline bouncing even on architectures that have relaxed atomic memory operations.
Which is why we have per-cpu counters already. Making use of them doesn't seem that outlandish.
But you wouldn't be able to guarantee consistency anyway, would you? You *could* copy N*M counters to userspace, but there's no guarantee that they are consistent, neither within a single vCPU nor within a single counter.
Speed/efficiency of retrieving statistics is important, but let's keep in mind that the baseline for comparison is hundreds of syscalls and filesystem lookups.
Having that baseline in the cover letter would be a good start, as well as an indication of the frequency this is used at.
Can't disagree, especially on the latter which I have no idea about.
Paolo
Marc, Paolo, thanks for the comments. I will add some more information in the cover letter.
Thanks, Jing