drbd: Backport the "events2" command
Hello Andreas Gruenbacher,
The patch a29728463b25: "drbd: Backport the "events2" command" from Jul 31, 2014, leads to the following static checker warning:
drivers/block/drbd/drbd_nl.c:4934 get_initial_state() error: dereferencing freed memory 'skb'
drivers/block/drbd/drbd_nl.c
4880 static int get_initial_state(struct sk_buff *skb, struct netlink_callback *cb) 4881 {
4882 struct drbd_state_change *state_change = (struct drbd_state_change *)cb->args[0];
4883 unsigned int seq = cb->args[2];
4884 unsigned int n;
4885 enum drbd_notification_type flags = 0;
4886
4887 /* There is no need for taking notification_mutex here: it doesn't
4888 matter if the initial state events mix with later state chage
4889 events; we can always tell the events apart by the NOTIFY_EXISTS
4890 flag. */
4891
4892 cb->args[5]--;
4893 if (cb->args[5] == 1) {
4894 notify_initial_state_done(skb, seq);
^^^skb is freed on error inside notify_initial_state_done().
4895 goto out;
4896 }
4897 n = cb->args[4]++;
4898 if (cb->args[4] < cb->args[3])
4899 flags |= NOTIFY_CONTINUES;
4900 if (n < 1) {
4901 notify_resource_state_change(skb, seq, state_change->resource,
4902 NOTIFY_EXISTS | flags);
4903 goto next;
4904 }
4905 n--;
4906 if (n < state_change->n_connections) {
4907 notify_connection_state_change(skb, seq, &state_change->connections[n],
4908 NOTIFY_EXISTS | flags);
4909 goto next;
4910 }
4911 n -= state_change->n_connections;
4912 if (n < state_change->n_devices) {
4913 notify_device_state_change(skb, seq, &state_change->devices[n],
4914 NOTIFY_EXISTS | flags);
4915 goto next;
4916 }
4917 n -= state_change->n_devices;
4918 if (n < state_change->n_devices * state_change->n_connections) {
4919 notify_peer_device_state_change(skb, seq, &state_change->peer_devices[n],
4920 NOTIFY_EXISTS | flags);
4921 goto next;
4922 }
4923
4924 next:
4925 if (cb->args[4] == cb->args[3]) {
4926 struct drbd_state_change *next_state_change =
4927 list_entry(state_change->list.next,
4928 struct drbd_state_change, list);
4929 cb->args[0] = (long)next_state_change;
4930 cb->args[3] = notifications_for_state_change(next_state_change);
4931 cb->args[4] = 0;
4932 }
4933 out:
4934 return skb->len;
^^^^^^^^Dereference.
4935 }
regards,
dan carpenter
Lars Ellenberg
maintainerQuoted 1 line · Dan Carpenter
Hello Andreas Gruenbacher,
Andreas has since moved to more exiting challenges :)
Quoted 23 lines
The patch
a29728463b25: "drbd: Backport the "events2" command" from Jul 31, 2014, leads to the following static checker warning:drivers/block/drbd/drbd_nl.c:4934 get_initial_state() error: dereferencing freed memory 'skb'
drivers/block/drbd/drbd_nl.c
4880 static intget_initial_state(struct sk_buff *skb, struct netlink_callback *cb)4881 {4882 struct drbd_state_change *state_change = (struct drbd_state_change *)cb->args[0]; 4883 unsigned int seq = cb->args[2]; 4884 unsigned int n; 4885 enum drbd_notification_type flags = 0; 4886 4887 /* There is no need for taking notification_mutex here: it doesn't 4888 matter if the initial state events mix with later state chage 4889 events; we can always tell the events apart by the NOTIFY_EXISTS 4890 flag. */ 4891 4892 cb->args[5]--; 4893 if (cb->args[5] == 1) { 4894 notify_initial_state_done(skb, seq); ^^^
Quoted 1 line
skb is freed on error inside
notify_initial_state_done().
So notify_resource_state_change needs to become non void, and we need to change notify_initial_state_done(); goto out; to return notify_initial_state_done();
right?
Signature · 6 lines
: Lars Ellenberg : LINBIT | Keeping the Digital World Running : DRBD -- Heartbeat -- Corosync -- Pacemaker : R&D, Integration, Ops, Consulting, Support DRBD® and LINBIT® are registered trademarks of LINBIT