Re: [Xen-devel] [v3][PATCH 03/16] xen/vtd: create RMRR mapping

["Tian, Kevin" <kevin.tian@xxxxxxxxx>]

> From: Chen, Tiejun
> Sent: Thursday, June 11, 2015 9:15 AM
> 
> RMRR reserved regions must be setup in the pfn space with an identity
> mapping to reported mfn. However existing code has problem to setup
> correct mapping when VT-d shares EPT page table, so lead to problem
> when assigning devices (e.g GPU) with RMRR reported. So instead, this
> patch aims to setup identity mapping in p2m layer, regardless of
> whether EPT is shared or not. And we still keep creating VT-d table.
> 
> Signed-off-by: Tiejun Chen <tiejun.chen@xxxxxxxxx>
> ---
>  xen/arch/x86/mm/p2m.c               | 10 ++++++++--
>  xen/drivers/passthrough/vtd/iommu.c |  3 +--
>  2 files changed, 9 insertions(+), 4 deletions(-)
> 
> diff --git a/xen/arch/x86/mm/p2m.c b/xen/arch/x86/mm/p2m.c
> index a6db236..c7198a5 100644
> --- a/xen/arch/x86/mm/p2m.c
> +++ b/xen/arch/x86/mm/p2m.c
> @@ -927,10 +927,16 @@ int set_identity_p2m_entry(struct domain *d, unsigned 
> long
> gfn,
>          }
> 
>          gfn_unlock(p2m, gfn, 0);
> -        return ret;
>      }
> +    else
> +        ret = 0;
> 
> -    return 0;
> +    if( ret == 0 )
> +    {
> +        ret = iommu_map_page(d, gfn, gfn, IOMMUF_readable|IOMMUF_writable);
> +    }
> +
> +    return ret;

p2m_set_entry will setup IOMMU pages already. You don't need
another explicit iommu map here. 

>  }
> 
>  /* Returns: 0 for success, -errno for failure */
> diff --git a/xen/drivers/passthrough/vtd/iommu.c
> b/xen/drivers/passthrough/vtd/iommu.c
> index 6a37624..31ce1af 100644
> --- a/xen/drivers/passthrough/vtd/iommu.c
> +++ b/xen/drivers/passthrough/vtd/iommu.c
> @@ -1856,8 +1856,7 @@ static int rmrr_identity_mapping(struct domain *d, 
> bool_t map,
> 
>      while ( base_pfn < end_pfn )
>      {
> -        int err = intel_iommu_map_page(d, base_pfn, base_pfn,
> -                                       IOMMUF_readable|IOMMUF_writable);
> +        int err = set_identity_p2m_entry(d, base_pfn, p2m_access_rw);
> 
>          if ( err )
>              return err;

Tim has another comment to replace earlier unmap with 
guest_physmap_remove_page() which will call iommu
unmap internally. Please include this change too.

Thanks
Kevin

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