From: Al Viro <viro@xxxxxxxxxxxxxxxxxx>
commit 451a2886b6bf90e2fb378f7c46c655450fb96e81 upstream.
unfortunately, allowing an arbitrary 16bit value means a possibility of
overflow in the calculation of total number of pages in bio_map_user_iov() -
we rely on there being no more than PAGE_SIZE members of sum in the
first loop there. If that sum wraps around, we end up allocating
too small array of pointers to pages and it's easy to overflow it in
the second loop.
X-Coverup: TINC (and there's no lumber cartel either)
Signed-off-by: Al Viro <viro@xxxxxxxxxxxxxxxxxx>
[bwh: s/MAX_UIOVEC/UIO_MAXIOV/. This was fixed upstream by commit
fdc81f45e9f5 ("sg_start_req(): use import_iovec()"), but we don't have
that function.]
Signed-off-by: Ben Hutchings <ben@xxxxxxxxxxxxxxx>
---
It looks like this bug was introduced in 2.6.28 by commit 10db10d144c0
("sg: convert the indirect IO path to use the block layer"), so the fix
is needed for all stable branches.
Ben.
drivers/scsi/sg.c | 3 +++
1 file changed, 3 insertions(+)
--- a/drivers/scsi/sg.c
+++ b/drivers/scsi/sg.c
@@ -1687,6 +1687,9 @@ static int sg_start_req(Sg_request *srp,
md->from_user = 0;
}
+ if (unlikely(iov_count > UIO_MAXIOV))
+ return -EINVAL;
+
if (iov_count) {
int len, size = sizeof(struct sg_iovec) * iov_count;
struct iovec *iov;
--
Ben Hutchings
One of the nice things about standards is that there are so many of them.
Attachment:
signature.asc
Description: This is a digitally signed message part