vpci: add permission checks to map_range()
PATCH v2 1/1 vpci: add permission checks to map_range()
Just like it's done for the XEN_DOMCTL_memory_mapping hypercall, add the permissions checks to vPCI map_range(), which is used to map the BARs into the domain p2m.
Adding those checks requires that for x86 PVH hardware domain builder the permissions are set before initializing the IOMMU, or else attempts to initialize vPCI done as part of IOMMU device setup will fail due to missing permissions to create the BAR mappings.
While moving the call to dom0_setup_permissions() convert the panic() used for error handling to a printk, the caller will already panic if required.
---
Changes since v1:
- Fix printk calls.
- Expand commit message.
---
I'm unsure whether on ARM MMIO permissions are properly set for the hardware domain, but I don't have a system to test with.
--- xen/arch/x86/hvm/dom0_build.c | 21 ++++++++++++++------- xen/drivers/vpci/header.c | 20 ++++++++++++++++++++ 2 files changed, 34 insertions(+), 7 deletions(-)
xen/arch/x86/hvm/dom0_build.c +14 -7
@@ -715,13 +715,6 @@ static int __init pvh_setup_cpus(struct domain *d, paddr_t entry,return rc; } - rc =dom0_setup_permissions (d); -if ( rc ) - { -panic ("Unable to setup Dom0 permissions: %d\n" , rc); -return rc; - } -update_domain_wallclock_time (d); v->is_initialised =1 ;
@@ -1184,6 +1177,20 @@ int __init dom0_construct_pvh(struct domain *d, const module_t *image,printk (XENLOG_INFO "*** Building a PVH Dom%d ***\n" , d->domain_id); +if (is_hardware_domain (d) ) + { +/* +* Setup permissions early so that calls to add MMIO regions to the +* p2m as part of vPCI setup don't fail due to permission checks. +*/ + rc =dom0_setup_permissions (d); +if ( rc ) + { +printk ("%pd unable to setup permissions: %d\n" , d, rc); +return rc; + } + } +/* * NB: MMCFG initialization needs to be performed before iommu * initialization so the iommu code can fetch the MMCFG regions used by the
xen/drivers/vpci/header.c +20 -0
@@ -17,10 +17,13 @@* License along with this program; If not, see <http://www.gnu.org/licenses/>. */ +#include <xen/iocap.h> #include <xen/sched.h> #include <xen/softirq.h> #include <xen/vpci.h> +#include <xsm/xsm.h> +#include <asm/event.h> #include <asm/p2m.h>
@@ -43,6 +46,23 @@ static int cf_check map_range( {unsigned long size = e - s +1 ; +if ( !iomem_access_permitted (map->d, s, e) ) + { +printk (XENLOG_G_WARNING +"%pd denied access to MMIO range [%#lx, %#lx]\n" , + map->d, s, e); +return -EPERM ; + } + + rc =xsm_iomem_mapping (XSM_HOOK , map->d, s, e, map->map); +if ( rc ) + { +printk (XENLOG_G_WARNING +"%pd XSM denied access to MMIO range [%#lx, %#lx]: %d\n" , + map->d, s, e, rc); +return rc; + } +/* * ARM TODOs: * - On ARM whether the memory is prefetchable or not should be passed
Signature · 1 line
2.41.0
Quoted 9 lines · Roger Pau Monne
Just like it's done for the
XEN_DOMCTL_memory_mappinghypercall, add the permissions checks to vPCImap_range(), which is used to map the BARs into the domain p2m.Adding those checks requires that for x86 PVH hardware domain builder the permissions are set before initializing the IOMMU, or else attempts to initialize vPCI done as part of IOMMU device setup will fail due to missing permissions to create the BAR mappings.
Quoted 6 lines
While moving the call to
dom0_setup_permissions()convert thepanic()used for error handling to a printk, the caller will already panic if required.Fixes 9c244fdef7e7 ('vpci: add header handlers')Signed-off-by Roger Pau Monné <[email protected]>
Hi Roger,
Quoted 11 lines · Roger Pau Monne
Just like it's done for the
XEN_DOMCTL_memory_mappinghypercall, add the permissions checks to vPCImap_range(), which is used to map the BARs into the domain p2m.Adding those checks requires that for x86 PVH hardware domain builder the permissions are set before initializing the IOMMU, or else attempts to initialize vPCI done as part of IOMMU device setup will fail due to missing permissions to create the BAR mappings.
Quoted 6 lines
While moving the call to
dom0_setup_permissions()convert thepanic()used for error handling to a printk, the caller will already panic if required.Fixes 9c244fdef7e7 ('vpci: add header handlers')Signed-off-by Roger Pau Monné <[email protected]>
I tested the patch on ARM board with vPCI enabled everything works.
Regards,
Rahul
Quoted 9 lines · Roger Pau Monne
Just like it's done for the
XEN_DOMCTL_memory_mappinghypercall, add the permissions checks to vPCImap_range(), which is used to map the BARs into the domain p2m.Adding those checks requires that for x86 PVH hardware domain builder the permissions are set before initializing the IOMMU, or else attempts to initialize vPCI done as part of IOMMU device setup will fail due to missing permissions to create the BAR mappings.
Quoted 6 lines
While moving the call to
dom0_setup_permissions()convert thepanic()used for error handling to a printk, the caller will already panic if required.Fixes 9c244fdef7e7 ('vpci: add header handlers')Signed-off-by Roger Pau Monné <[email protected]>
I've committed this, but despite the Fixes: tag I'm not sure this wants backporting. Thoughts?
Jan
Quoted 14 lines · Jan Beulich
Quoted 13 lines · Roger Pau Monne
Just like it's done for the
XEN_DOMCTL_memory_mappinghypercall, add the permissions checks to vPCImap_range(), which is used to map the BARs into the domain p2m.Adding those checks requires that for x86 PVH hardware domain builder the permissions are set before initializing the IOMMU, or else attempts to initialize vPCI done as part of IOMMU device setup will fail due to missing permissions to create the BAR mappings.
While moving the call to
dom0_setup_permissions()convert thepanic()used for error handling to a printk, the caller will already panic if required.
Quoted 5 lines
Quoted 2 lines
Fixes 9c244fdef7e7 ('vpci: add header handlers')Signed-off-by Roger Pau Monné <[email protected]>I've committed this, but despite the Fixes: tag I'm not sure this wants backporting. Thoughts?
It was IMO an omission from that commit, however given vPCI so far is only used by dom0 (an in experimental mode) I don't see much reason to backport it.
Thanks, Roger.
Quoted 14 lines · Jan Beulich
Quoted 13 lines · Roger Pau Monne
Just like it's done for the
XEN_DOMCTL_memory_mappinghypercall, add the permissions checks to vPCImap_range(), which is used to map the BARs into the domain p2m.Adding those checks requires that for x86 PVH hardware domain builder the permissions are set before initializing the IOMMU, or else attempts to initialize vPCI done as part of IOMMU device setup will fail due to missing permissions to create the BAR mappings.
While moving the call to
dom0_setup_permissions()convert thepanic()used for error handling to a printk, the caller will already panic if required.
Quoted 7 lines
Quoted 2 lines
Fixes 9c244fdef7e7 ('vpci: add header handlers')Signed-off-by Roger Pau Monné <[email protected]>I've committed this, but despite the Fixes: tag I'm not sure this wants backporting. Thoughts?
Jan
From a cursory review thus far, since this introduced a new XSM hook site, shouldn't this have at least had an Rb by an XSM reviewer/maintainer? I would have replied sooner, but have been on holiday for last two weeks.
V/r,
Daniel P. Smith
Quoted 20 lines · Daniel P. Smith
Quoted 17 lines · Jan Beulich
Quoted 15 lines · Roger Pau Monne
Just like it's done for the
XEN_DOMCTL_memory_mappinghypercall, add the permissions checks to vPCImap_range(), which is used to map the BARs into the domain p2m.Adding those checks requires that for x86 PVH hardware domain builder the permissions are set before initializing the IOMMU, or else attempts to initialize vPCI done as part of IOMMU device setup will fail due to missing permissions to create the BAR mappings.
While moving the call to
dom0_setup_permissions()convert thepanic()used for error handling to a printk, the caller will already panic if required.Fixes 9c244fdef7e7 ('vpci: add header handlers')Signed-off-by Roger Pau Monné <[email protected]>
Quoted 6 lines
Quoted 2 lines
I've committed this, but despite the Fixes: tag I'm not sure this wants backporting. Thoughts?
From a cursory review thus far, since this introduced a new XSM hook
site, shouldn't this have at least had an Rb by an XSM reviewer/maintainer?
Probably, but already back then I said this model isn't going to work flawlessly.
Quoted 1 line
I would have replied sooner, but have been on holiday for last two weeks.
I guess there was no way for us to know without you sending a note to private@ (which, I will admit, you may not even have been aware of).
Jan