[PATCH 3.11 195/198] aio: fix kernel memory disclosure in io_getevents() introduced in v3.10

[Luis Henriques]

3.11.10.13 -stable review patch.  If anyone has any objections, please let me know.

------------------

From: Benjamin LaHaise <bcrl at kvack.org>

commit edfbbf388f293d70bf4b7c0bc38774d05e6f711a upstream.

A kernel memory disclosure was introduced in aio_read_events_ring() in v3.10
by commit a31ad380bed817aa25f8830ad23e1a0480fef797.  The changes made to
aio_read_events_ring() failed to correctly limit the index into
ctx->ring_pages[], allowing an attacked to cause the subsequent kmap() of
an arbitrary page with a copy_to_user() to copy the contents into userspace.
This vulnerability has been assigned CVE-2014-0206.  Thanks to Mateusz and
Petr for disclosing this issue.

This patch applies to v3.12+.  A separate backport is needed for 3.10/3.11.

[jmoyer at redhat.com: backported to 3.10]
Signed-off-by: Benjamin LaHaise <bcrl at kvack.org>
Cc: Mateusz Guzik <mguzik at redhat.com>
Cc: Petr Matousek <pmatouse at redhat.com>
Cc: Kent Overstreet <kmo at daterainc.com>
Cc: Jeff Moyer <jmoyer at redhat.com>
Signed-off-by: Luis Henriques <luis.henriques at canonical.com>
---
 fs/aio.c | 2 ++
 1 file changed, 2 insertions(+)

diff --git a/fs/aio.c b/fs/aio.c
index 48f02745b876..618021906aa9 100644
--- a/fs/aio.c
+++ b/fs/aio.c
@@ -719,6 +719,8 @@ static long aio_read_events_ring(struct kioctx *ctx,
 	if (head == ctx->tail)
 		goto out;
 
+	head %= ctx->nr_events;
+
 	while (ret < nr) {
 		long avail;
 		struct io_event *ev;
-- 
1.9.1