Emulator accesses are always done a page at a time, either by the emulator
itself (for fetches) or because we need to query the MMU for address
translations.  Speed up these accesses by using kvm_read_guest_page
and, in the case of fetches, by inlining kvm_read_guest_virt_helper and
dropping the loop around kvm_read_guest_page.

This final tweak saves 30-100 more clock cycles (4-10%), bringing the
count (as measured by kvm-unit-tests) down to 720-1100 clock cycles on
a Sandy Bridge Xeon host, compared to 2300-3200 before the whole series
and 925-1700 after the first two low-hanging fruit changes.
Signed-off-by: NPaolo Bonzini <pbonzini@redhat.com>