On Fri, Jun 16, 2017 at 10:25 +0200, Mike Belopuhov wrote: > Last time I've tried uebayashi's pvclock on Xen, it didn't > work for me. I didn't have time to investigate why but > probably because we need per-cpu readings. Which you do > for KVM. I'll test this on Xen as soon as I get to the > office. > [...] > > But this brings another point: where and how to perform > the pvclock initialization and attachment. In your diff > pvclock_xen_init comes a bit too early: none of the Xen > things are initialized at that point, shared info page > isn't allocated. >
As I've told jmatthew@ privately this doesn't work on Xen, but I've changed a few things and made it work: https://github.com/mbelop/src/commits/pvclock However, I'm observing a huge drift: 6 minutes in about an hour so it's not usable as it is. I think this is the same thing as I saw before. I'll ponder the per-CPU vcpu_info path and report back when I've got something. But in the meantime I hope that jmatthew@ will check in rdtsc and pvbus_init_cpu bits (we're lobbying for a s/_vcpu/_cpu/ change. :-)
