Refcount table entries have a field to store the offset of the
refcount block. The rest of the bits of the entry are currently
reserved.

The offset is always taken from the entry using REFT_OFFSET_MASK to
ensure that we only use the bits that belong to that field.

While that mask is used every time we read from the refcount table, it
is never used when we write to it. Due to the other constraints of the
qcow2 format QEMU can never produce refcount block offsets that don't
fit in that field so any such offset when allocating a refcount block
would indicate a bug in QEMU.

Signed-off-by: Alberto Garcia <berto@igalia.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
---
v2: Add missing Signed-off-by: line
---
 block/qcow2-refcount.c | 3 +++
 1 file changed, 3 insertions(+)

diff --git a/block/qcow2-refcount.c b/block/qcow2-refcount.c
index 46082aeac1..31a2e1f845 100644
--- a/block/qcow2-refcount.c
+++ b/block/qcow2-refcount.c
@@ -367,6 +367,9 @@ static int alloc_refcount_block(BlockDriverState *bs,
         return new_block;
     }
 
+    /* The offset must fit in the offset field of the refcount table entry */
+    assert((new_block & REFT_OFFSET_MASK) == new_block);
+
     /* If we're allocating the block at offset 0 then something is wrong */
     if (new_block == 0) {
         qcow2_signal_corruption(bs, true, -1, -1, "Preventing invalid "
-- 
2.11.0

ping

I suppose this one can go in now.

On Wed 14 Nov 2018 03:58:57 PM CET, Alberto Garcia wrote:
> Refcount table entries have a field to store the offset of the
> refcount block. The rest of the bits of the entry are currently
> reserved.
>
> The offset is always taken from the entry using REFT_OFFSET_MASK to
> ensure that we only use the bits that belong to that field.
>
> While that mask is used every time we read from the refcount table, it
> is never used when we write to it. Due to the other constraints of the
> qcow2 format QEMU can never produce refcount block offsets that don't
> fit in that field so any such offset when allocating a refcount block
> would indicate a bug in QEMU.
>
> Signed-off-by: Alberto Garcia <berto@igalia.com>
> Reviewed-by: Eric Blake <eblake@redhat.com>
> ---
> v2: Add missing Signed-off-by: line
> ---
>  block/qcow2-refcount.c | 3 +++
>  1 file changed, 3 insertions(+)
>
> diff --git a/block/qcow2-refcount.c b/block/qcow2-refcount.c
> index 46082aeac1..31a2e1f845 100644
> --- a/block/qcow2-refcount.c
> +++ b/block/qcow2-refcount.c
> @@ -367,6 +367,9 @@ static int alloc_refcount_block(BlockDriverState *bs,
>          return new_block;
>      }
>  
> +    /* The offset must fit in the offset field of the refcount table entry */
> +    assert((new_block & REFT_OFFSET_MASK) == new_block);
> +
>      /* If we're allocating the block at offset 0 then something is wrong */
>      if (new_block == 0) {
>          qcow2_signal_corruption(bs, true, -1, -1, "Preventing invalid "
> -- 
> 2.11.0

Am 14.11.2018 um 15:58 hat Alberto Garcia geschrieben:
> Refcount table entries have a field to store the offset of the
> refcount block. The rest of the bits of the entry are currently
> reserved.
> 
> The offset is always taken from the entry using REFT_OFFSET_MASK to
> ensure that we only use the bits that belong to that field.
> 
> While that mask is used every time we read from the refcount table, it
> is never used when we write to it. Due to the other constraints of the
> qcow2 format QEMU can never produce refcount block offsets that don't
> fit in that field so any such offset when allocating a refcount block
> would indicate a bug in QEMU.
> 
> Signed-off-by: Alberto Garcia <berto@igalia.com>
> Reviewed-by: Eric Blake <eblake@redhat.com>

Thanks, applied to the block branch.

Kevin

ping 2

On Wed 14 Nov 2018 03:58:57 PM CET, Alberto Garcia wrote:
> Refcount table entries have a field to store the offset of the
> refcount block. The rest of the bits of the entry are currently
> reserved.
>
> The offset is always taken from the entry using REFT_OFFSET_MASK to
> ensure that we only use the bits that belong to that field.
>
> While that mask is used every time we read from the refcount table, it
> is never used when we write to it. Due to the other constraints of the
> qcow2 format QEMU can never produce refcount block offsets that don't
> fit in that field so any such offset when allocating a refcount block
> would indicate a bug in QEMU.
>
> Signed-off-by: Alberto Garcia <berto@igalia.com>
> Reviewed-by: Eric Blake <eblake@redhat.com>
> ---
> v2: Add missing Signed-off-by: line
> ---
>  block/qcow2-refcount.c | 3 +++
>  1 file changed, 3 insertions(+)
>
> diff --git a/block/qcow2-refcount.c b/block/qcow2-refcount.c
> index 46082aeac1..31a2e1f845 100644
> --- a/block/qcow2-refcount.c
> +++ b/block/qcow2-refcount.c
> @@ -367,6 +367,9 @@ static int alloc_refcount_block(BlockDriverState *bs,
>          return new_block;
>      }
>  
> +    /* The offset must fit in the offset field of the refcount table entry */
> +    assert((new_block & REFT_OFFSET_MASK) == new_block);
> +
>      /* If we're allocating the block at offset 0 then something is wrong */
>      if (new_block == 0) {
>          qcow2_signal_corruption(bs, true, -1, -1, "Preventing invalid "
> -- 
> 2.11.0