[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Xen-devel] [V2 PATCH 1/8] PVH dom0: iommu related changes



Mukesh Rathor <mukesh.rathor@xxxxxxxxxx> wrote:
>- For now, iommu is required for PVH dom0. Check for that.
>- For pvh, we need to do mfn_to_gmfn before calling mapping function
>  intel_iommu_map_page/amd_iommu_map_page which expects a gfn.
>
>Signed-off-by: Mukesh Rathor <mukesh.rathor@xxxxxxxxxx>
>---
> xen/drivers/passthrough/iommu.c |   17 +++++++++++++++--
> 1 files changed, 15 insertions(+), 2 deletions(-)
>
>diff --git a/xen/drivers/passthrough/iommu.c
>b/xen/drivers/passthrough/iommu.c
>index 93ad122..dbbd034 100644
>--- a/xen/drivers/passthrough/iommu.c
>+++ b/xen/drivers/passthrough/iommu.c
>@@ -125,15 +125,27 @@ int iommu_domain_init(struct domain *d)
>     return hd->platform_ops->init(d);
> }
> 
>+static __init void check_dom0_pvh_reqs(struct domain *d)
>+{
>+    if ( !iommu_enabled )
>+        panic("Presently, iommu must be enabled for pvh dom0\n");
>+
>+    if ( iommu_passthrough )
>+        panic("For pvh dom0, dom0-passthrough must not be enabled\n");
>+}

That is not very "check' type. You panic!

Could you instead just disable PVH dom0 booting and boot it as a normal PV 
guest (or try)?
>+
> void __init iommu_dom0_init(struct domain *d)
> {
>     struct hvm_iommu *hd = domain_hvm_iommu(d);
> 
>+    if ( is_pvh_domain(d) )
>+        check_dom0_pvh_reqs(d);
>+
>     if ( !iommu_enabled )
>         return;
> 
>     register_keyhandler('o', &iommu_p2m_table);
>-    d->need_iommu = !!iommu_dom0_strict;
>+    d->need_iommu = is_pvh_domain(d) || !!iommu_dom0_strict;
>     if ( need_iommu(d) )
>     {
>         struct page_info *page;
>@@ -141,12 +153,13 @@ void __init iommu_dom0_init(struct domain *d)
>         page_list_for_each ( page, &d->page_list )
>         {
>             unsigned long mfn = page_to_mfn(page);
>+            unsigned long gfn = mfn_to_gmfn(d, mfn);
>             unsigned int mapping = IOMMUF_readable;
>             if ( ((page->u.inuse.type_info & PGT_count_mask) == 0) ||
>                  ((page->u.inuse.type_info & PGT_type_mask)
>                   == PGT_writable_page) )
>                 mapping |= IOMMUF_writable;
>-            hd->platform_ops->map_page(d, mfn, mfn, mapping);
>+            hd->platform_ops->map_page(d, gfn, mfn, mapping);
>             if ( !(i++ & 0xfffff) )
>                 process_pending_softirqs();
>         }



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


 


Rackspace

Lists.xenproject.org is hosted with RackSpace, monitoring our
servers 24x7x365 and backed by RackSpace's Fanatical Support®.