x86/hvm: be more strict with XENMAPSPACE_gmfn source types
PATCH 1/1 x86/hvm: be more strict with XENMAPSPACE_gmfn source types
XENMAPSPACE_gmfn{_range} allows moving gfn around the guest p2m: the mfn behind the source gfn is zapped from the origin and mapped at the requested destination gfn. The destination p2m entries are always created with type p2m_ram_rw.
With the current checking done in xenmem_add_to_physmap_one() it's possible to use XENMAPSPACE_gmfn{_range} to change the type of a p2m entry. The source gfn is only checked to be not shared, and that the underlying page is owned by the domain.
Make the source checks more strict, by checking that the source gfn is of type p2m_ram_rw. That prevents the operation from inadvertently changing the type as part of the move.
---
The change was discussed internally by the security team and deemed not a security issue.
--- xen/arch/x86/mm/p2m.c | 10 +++++++--- 1 file changed, 7 insertions(+), 3 deletions(-)
xen/arch/x86/mm/p2m.c +7 -3
@@ -2007,11 +2007,15 @@ int xenmem_add_to_physmap_one( { gmfn = idx; mfn =get_gfn_unshare (d, gmfn, &p2mt); -/* If the page is still shared, exit early */ -if (p2m_is_shared (p2mt) ) +/* +* The entry at the destination gfn will be created as type p2m_ram_rw. +* Only allow moving source gfns with p2m_ram_rw type to avoid +* unexpected p2m type changes as a result of the operation. +*/ +if ( p2mt != p2m_ram_rw ) {put_gfn (d, gmfn); -return -ENOMEM ; +return -EACCES ; } page =get_page_from_mfn (mfn, d);if (unlikely (!page) )
Signature · 1 line
2.51.0
Quoted 4 lines · Roger Pau Monne
xen/arch/x86/mm/p2m.c +0
-0@@ -2007,11 +2007,15 @@ int xenmem_add_to_physmap_one( {
xen/arch/x86/mm/p2m.c int xenmem_add_to_physmap_one( · line 2015
Quoted 10 lines
xen/arch/x86/mm/p2m.c +6
-2gmfn = idx; mfn =get_gfn_unshare (d, gmfn, &p2mt); -/* If the page is still shared, exit early */ -if (p2m_is_shared (p2mt) ) +/* +* The entry at the destination gfn will be created as type p2m_ram_rw. +* Only allow moving source gfns with p2m_ram_rw type to avoid +* unexpected p2m type changes as a result of the operation. +*/ +if ( p2mt != p2m_ram_rw )
As asked before - what about p2m_log_dirty? Imo that needs permitting here as well. Making it become p2m_ram_rw is "natural", as long as the (new) GFN is suitably marked dirty (which p2m_add_page() looks to be doing).
xen/arch/x86/mm/p2m.c int xenmem_add_to_physmap_one( · line 2018
Quoted 4 lines
xen/arch/x86/mm/p2m.c +1
-1{put_gfn (d, gmfn); -return -ENOMEM ; +return -EACCES ;
Since we tried to unshare, imo ENOMEM should continue to be returned for p2m_is_shared().
With both changes:
Jan
Quoted 5 lines · Roger Pau Monne
XENMAPSPACE_gmfn{_range} allows moving gfn around the guest p2m: the mfn behind the source gfn is zapped from the origin and mapped at the requested destination gfn. The destination p2m entries are always created with type
p2m_ram_rw.
Quoted 8 lines
With the current checking done in
xenmem_add_to_physmap_one()it's possible to use XENMAPSPACE_gmfn{_range} to change the type of a p2m entry. The source gfn is only checked to be not shared, and that the underlying page is owned by the domain.Make the source checks more strict, by checking that the source gfn is of type
p2m_ram_rw. That prevents the operation from inadvertently changing the type as part of the move.
This is missing:
The hypercall was missing any p2m type checks since introduction. It's possible the get_page() seemed enough, but it was dangerous to not account for new incompatible p2m types being added down the road.
Thanks, Roger.