Re: [Xen-devel] [PATCH for 4.5] x86/viridian: Freeze time reference counter when domain is paused

["Jan Beulich" <JBeulich@xxxxxxxx>]

>>> On 10.10.14 at 11:07, <paul.durrant@xxxxxxxxxx> wrote:
> @@ -460,11 +495,12 @@ static int viridian_load_domain_ctxt(struct domain *d, 
> hvm_domain_context_t *h)
>  {
>      struct hvm_viridian_domain_context ctxt;
>  
> -    if ( hvm_load_entry(VIRIDIAN_DOMAIN, h, &ctxt) != 0 )
> +    if ( hvm_load_entry_zeroextend(VIRIDIAN_DOMAIN, h, &ctxt) != 0 )
>          return -EINVAL;
>  
> -    d->arch.hvm_domain.viridian.hypercall_gpa.raw = ctxt.hypercall_gpa;
> -    d->arch.hvm_domain.viridian.guest_os_id.raw   = ctxt.guest_os_id;
> +    d->arch.hvm_domain.viridian.time_ref_count.val = ctxt.time_ref_count;
> +    d->arch.hvm_domain.viridian.hypercall_gpa.raw  = ctxt.hypercall_gpa;
> +    d->arch.hvm_domain.viridian.guest_os_id.raw    = ctxt.guest_os_id;

Is loading zero here correct when migration comes from an older
hypervisor?

> @@ -706,6 +710,11 @@ void domain_shutdown(struct domain *d, u8 reason)
>          v->paused_for_shutdown = 1;
>      }
>  
> +#ifdef CONFIG_X86
> +    if ( has_viridian_time_ref_count(d) )
> +        viridian_time_ref_count_freeze(d);
> +#endif
> +

So in the description you say "but only if the domain is not shutting
down" - how does the above change fit that? Or is that comment
ambiguous whether it refers to both the freezing and thawing or only
the thawing?

Jan


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