Re: Utility of PV on non-obsolete hardware

[tg%gmplib.org@localhost (Torbjörn Granlund)]

Greg Troxel <gdt%lexort.com@localhost> writes:

  Is this slowness of PV expected and understood?  Is it about the cost of page
  table update hypercalls?  Could it be a difference in using multiple
  cores vs not?

It is not understood by me :) but it likely has to do with inherent VM
overhead for PV.

Note that my Dom0 runs Gentoo GNU/Linux and that Linux PV guests run
slowly compared to the same guests in PVH or PVHVM mode.  So this is not
a NetBSD problem.  NetBSD PV is no slower than e.g. Debian PV guests.

My guests all use a single core (except the Dom0 "guest" and the
${host}gentoo64 guest, which have several).

-- 
Torbjörn
Please encrypt, key id 0xC8601622