Re: [Xen-devel] [PATCH] xen: Don't disable the REP emulation optimizations for regular IO

[Tamas K Lengyel <tamas@xxxxxxxxxxxxx>]

On Mar 10, 2016 11:09, "Razvan Cojocaru" <rcojocaru@xxxxxxxxxxxxxxx> wrote:
>
> Currently REP emulations optimizations remain disabled even if
> the emulation does not happen as a result of a vm_event reply
> requestion emulation (i.e. even for regular IO). This patch takes
> emulate_each_rep into account only if emulation has been requested
> by a vm_event-capable application, and is a noticeable speed
> optimization for monitored guests.
>
> Signed-off-by: Razvan Cojocaru <rcojocaru@xxxxxxxxxxxxxxx>

Reviewed-by: Tamas K Lengyel <tamas@xxxxxxxxxxxxx>

> ---
>  xen/arch/x86/hvm/emulate.c | 18 ++++++++++++------
>  1 file changed, 12 insertions(+), 6 deletions(-)
>
> diff --git a/xen/arch/x86/hvm/emulate.c b/xen/arch/x86/hvm/emulate.c
> index 082aa30..ddc8007 100644
> --- a/xen/arch/x86/hvm/emulate.c
> +++ b/xen/arch/x86/hvm/emulate.c
> @@ -498,6 +498,7 @@ static int hvmemul_virtual_to_linear(
>  {
>      struct segment_register *reg;
>      int okay;
> +    unsigned long max_reps = 4096;
>
>      if ( seg == x86_seg_none )
>      {
> @@ -506,16 +507,21 @@ static int hvmemul_virtual_to_linear(
>      }
>
>      /*
> +     * If introspection has been enabled for this domain, and we're emulating
> +     * becase a vm_reply asked us to (i.e. not doing regular IO) reps should
> +     * be at most 1, since optimization might otherwise cause a single
> +     * vm_event being triggered for repeated writes to a whole page.
> +     */
> +    if ( unlikely(current->domain->arch.mem_access_emulate_each_rep) &&
> +         current->arch.vm_event->emulate_flags != 0 )
> +       max_reps = 1;
> +
> +    /*
>       * Clip repetitions to avoid overflow when multiplying by @bytes_per_rep.
>       * The chosen maximum is very conservative but it's what we use in
>       * hvmemul_linear_to_phys() so there is no point in using a larger value.
> -     * If introspection has been enabled for this domain, *reps should be
> -     * at most 1, since optimization might otherwise cause a single vm_event
> -     * being triggered for repeated writes to a whole page.
>       */
> -    *reps = min_t(unsigned long, *reps,
> -                  unlikely(current->domain->arch.mem_access_emulate_each_rep)
> -                           ? 1 : 4096);
> +    *reps = min_t(unsigned long, *reps, max_reps);
>
>      reg = hvmemul_get_seg_reg(seg, hvmemul_ctxt);
>
> --
> 1.9.1
>
>
> _______________________________________________
> Xen-devel mailing list
> Xen-devel@xxxxxxxxxxxxx
> http://lists.xen.org/xen-devel

_______________________________________________
Xen-devel mailing list
Xen-devel@xxxxxxxxxxxxx
http://lists.xen.org/xen-devel