Re: [Xen-devel] [PATCH 04/11] x86/MM: Improve p2m type checks.

[Andrew Cooper <andrew.cooper3@xxxxxxxxxx>]

On 09/01/15 21:26, Ed White wrote:
> diff --git a/xen/include/asm-x86/p2m.h b/xen/include/asm-x86/p2m.h
> index 5f7fe71..8193901 100644
> --- a/xen/include/asm-x86/p2m.h
> +++ b/xen/include/asm-x86/p2m.h
> @@ -193,6 +193,9 @@ struct p2m_domain {
>       * threaded on in LRU order. */
>      struct list_head   np2m_list;
>  
> +    /* Does this p2m belong to the altp2m code? */
> +    bool_t alternate;
> +
>      /* Host p2m: Log-dirty ranges registered for the domain. */
>      struct rangeset   *logdirty_ranges;
>  
> @@ -290,7 +293,9 @@ struct p2m_domain *p2m_get_nestedp2m(struct vcpu *v, 
> uint64_t np2m_base);
>   */
>  struct p2m_domain *p2m_get_p2m(struct vcpu *v);
>  
> -#define p2m_is_nestedp2m(p2m)   ((p2m) != p2m_get_hostp2m((p2m->domain)))
> +#define p2m_is_hostp2m(p2m)   ((p2m) == p2m_get_hostp2m((p2m->domain)))
> +#define p2m_is_altp2m(p2m)    ((p2m)->alternate)
> +#define p2m_is_nestedp2m(p2m) (!p2m_is_altp2m(p2m) && !p2m_ishostp2m(p2m))

Might this be better expressed as a p2m type, currently of the set
{host, alt, nested} ?  p2m_is_nestedp2m() is starting to hide some
moderately complicated calculations.

~Andrew


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