[Qemu-devel] [PATCH] vfio-pci: Fix BAR size overflow

[Qemu-devel] [PATCH] vfio-pci: Fix BAR size overflow

[Alex Williamson]

We use an unsigned int when working with the PCI BAR size, which can
obviously overflow if the BAR is 4GB or larger.  This needs to change
to an unsigned long.  A similar issue is possible, though even more
unlikely, when mapping the region above an MSI-X table.  The start of
the table must be below 4GB, but the end, and therefore the start of
the next mapping region, could still land at 4GB.

Suggested-by: Nishank Trivedi <nishank.trivedi@netapp.com>
Signed-off-by: Alex Williamson <alex.williamson@redhat.com>
---

 hw/vfio/pci.c |    4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/hw/vfio/pci.c b/hw/vfio/pci.c
index b4e73d1..03790a8 100644
--- a/hw/vfio/pci.c
+++ b/hw/vfio/pci.c
@@ -2301,7 +2301,7 @@ static void vfio_unmap_bar(VFIOPCIDevice *vdev, int nr)
 static void vfio_map_bar(VFIOPCIDevice *vdev, int nr)
 {
     VFIOBAR *bar = &vdev->bars[nr];
-    unsigned size = bar->region.size;
+    unsigned long size = bar->region.size;
     char name[64];
     uint32_t pci_bar;
     uint8_t type;
@@ -2351,7 +2351,7 @@ static void vfio_map_bar(VFIOPCIDevice *vdev, int nr)
     }
 
     if (vdev->msix && vdev->msix->table_bar == nr) {
-        unsigned start;
+        unsigned long start;
 
         start = HOST_PAGE_ALIGN(vdev->msix->table_offset +
                                 (vdev->msix->entries * PCI_MSIX_ENTRY_SIZE));

Re: [Qemu-devel] [PATCH] vfio-pci: Fix BAR size overflow

[Don Slutz]

On 01/06/15 19:03, Alex Williamson wrote:
> We use an unsigned int when working with the PCI BAR size, which can
> obviously overflow if the BAR is 4GB or larger.  This needs to change
> to an unsigned long.  A similar issue is possible, though even more
> unlikely, when mapping the region above an MSI-X table.  The start of
> the table must be below 4GB, but the end, and therefore the start of
> the next mapping region, could still land at 4GB.
>
> Suggested-by: Nishank Trivedi <nishank.trivedi@netapp.com>
> Signed-off-by: Alex Williamson <alex.williamson@redhat.com>
> ---
>
>   hw/vfio/pci.c |    4 ++--
>   1 file changed, 2 insertions(+), 2 deletions(-)
>
> diff --git a/hw/vfio/pci.c b/hw/vfio/pci.c
> index b4e73d1..03790a8 100644
> --- a/hw/vfio/pci.c
> +++ b/hw/vfio/pci.c
> @@ -2301,7 +2301,7 @@ static void vfio_unmap_bar(VFIOPCIDevice *vdev, int nr)
>   static void vfio_map_bar(VFIOPCIDevice *vdev, int nr)
>   {
>       VFIOBAR *bar = &vdev->bars[nr];
> -    unsigned size = bar->region.size;
> +    unsigned long size = bar->region.size;

On a 32bit build, this does not fix the issue.
     -Don Slutz

>       char name[64];
>       uint32_t pci_bar;
>       uint8_t type;
> @@ -2351,7 +2351,7 @@ static void vfio_map_bar(VFIOPCIDevice *vdev, int nr)
>       }
>   
>       if (vdev->msix && vdev->msix->table_bar == nr) {
> -        unsigned start;
> +        unsigned long start;
>   
>           start = HOST_PAGE_ALIGN(vdev->msix->table_offset +
>                                   (vdev->msix->entries * PCI_MSIX_ENTRY_SIZE));
>
>

Re: [Qemu-devel] [PATCH] vfio-pci: Fix BAR size overflow

[Alex Williamson]

----- Original Message -----
> On 01/06/15 19:03, Alex Williamson wrote:
> > We use an unsigned int when working with the PCI BAR size, which can
> > obviously overflow if the BAR is 4GB or larger.  This needs to change
> > to an unsigned long.  A similar issue is possible, though even more
> > unlikely, when mapping the region above an MSI-X table.  The start of
> > the table must be below 4GB, but the end, and therefore the start of
> > the next mapping region, could still land at 4GB.
> >
> > Suggested-by: Nishank Trivedi <nishank.trivedi@netapp.com>
> > Signed-off-by: Alex Williamson <alex.williamson@redhat.com>
> > ---
> >
> >   hw/vfio/pci.c |    4 ++--
> >   1 file changed, 2 insertions(+), 2 deletions(-)
> >
> > diff --git a/hw/vfio/pci.c b/hw/vfio/pci.c
> > index b4e73d1..03790a8 100644
> > --- a/hw/vfio/pci.c
> > +++ b/hw/vfio/pci.c
> > @@ -2301,7 +2301,7 @@ static void vfio_unmap_bar(VFIOPCIDevice *vdev, int
> > nr)
> >   static void vfio_map_bar(VFIOPCIDevice *vdev, int nr)
> >   {
> >       VFIOBAR *bar = &vdev->bars[nr];
> > -    unsigned size = bar->region.size;
> > +    unsigned long size = bar->region.size;
> 
> On a 32bit build, this does not fix the issue.

Very true.  Thanks for the review,

Alex


> >       char name[64];
> >       uint32_t pci_bar;
> >       uint8_t type;
> > @@ -2351,7 +2351,7 @@ static void vfio_map_bar(VFIOPCIDevice *vdev, int nr)
> >       }
> >   
> >       if (vdev->msix && vdev->msix->table_bar == nr) {
> > -        unsigned start;
> > +        unsigned long start;
> >   
> >           start = HOST_PAGE_ALIGN(vdev->msix->table_offset +
> >                                   (vdev->msix->entries *
> >                                   PCI_MSIX_ENTRY_SIZE));
> >
> >
> 
>