x86: vdso: use __pvclock_read_cycles
The new simplified __pvclock_read_cycles does the same computation as vread_pvclock, except that (because it takes the pvclock_vcpu_time_info pointer) it has to be moved inside the loop. Since the loop is expected to never roll, this makes no difference. Acked-by: NAndy Lutomirski <luto@kernel.org> Signed-off-by: NPaolo Bonzini <pbonzini@redhat.com>
Showing
想要评论请 注册 或 登录