Skip to contentSkip to conversation

3.6.11.2-stable review

PATCH In review Steven Rostedt 6 participants 189 messages Apr 11, 2013 – Oct 23, 2013 Page 2 of 2

Subsystems (77)
lore
Share & tools
lore
author maintainer
#170 lore raw

PATCH 154/171 unix: fix a race condition in unix_release()

In reply to Steven Rostedt

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

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

From: Paul Moore <[email protected]>

[ Upstream commit ded34e0fe8fe8c2d595bfa30626654e4b87621e0 ]

As reported by Jan, and others over the past few years, there is a race condition caused by unix_release setting the sock->sk pointer to NULL before properly marking the socket as dead/orphaned. This can cause a problem with the LSM hook security_unix_may_send() if there is another socket attempting to write to this partially released socket in between when sock->sk is set to NULL and it is marked as dead/orphaned. This patch fixes this by only setting sock->sk to NULL after the socket has been marked as dead; I also take the opportunity to make unix_release_sock() a void function as it only ever returned 0/success.

Dave, I think this one should go on the -stable pile.

Special thanks to Jan for coming up with a reproducer for this problem.

---
 net/unix/af_unix.c |    7 +++----
 1 file changed, 3 insertions(+), 4 deletions(-)
net/unix/af_unix.c +3 -4
index c5ee4ff..1b3abc8 100644
@@ -384,7 +384,7 @@ static void unix_sock_destructor(struct sock *sk)
 #endif
 }
 
-static int unix_release_sock(struct sock *sk, int embrion)
+static void unix_release_sock(struct sock *sk, int embrion)
 {
 	struct unix_sock *u = unix_sk(sk);
 	struct path path;
@@ -453,8 +453,6 @@ static int unix_release_sock(struct sock *sk, int embrion)
 
 	if (unix_tot_inflight)
 		unix_gc();		/* Garbage collect fds */
-
-	return 0;
 }
 
 static void init_peercred(struct sock *sk)
@@ -704,9 +702,10 @@ static int unix_release(struct socket *sock)
 	if (!sk)
 		return 0;
 
+	unix_release_sock(sk, 0);
 	sock->sk = NULL;
 
-	return unix_release_sock(sk, 0);
+	return 0;
 }
 
 static int unix_autobind(struct socket *sock)
Signature · 1 line
1.7.10.4
author maintainer
#171 lore raw

PATCH 155/171 bonding: remove already created master sysfs link on failure

In reply to Steven Rostedt

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

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

From: Veaceslav Falico <[email protected]>

[ Upstream commit 9fe16b78ee17579cb4f333534cf7043e94c67024 ]

If slave sysfs symlink failes to be created - we end up without removing the master sysfs symlink. Remove it in case of failure.

---
 drivers/net/bonding/bond_sysfs.c |    5 +++++
 1 file changed, 5 insertions(+)
drivers/net/bonding/bond_sysfs.c +5 -0
index 21609f9..ecd47a3 100644
@@ -183,6 +183,11 @@ int bond_create_slave_symlinks(struct net_device *master,
 	sprintf(linkname, "slave_%s", slave->name);
 	ret = sysfs_create_link(&(master->dev.kobj), &(slave->dev.kobj),
 				linkname);
+
+	/* free the master link created earlier in case of error */
+	if (ret)
+		sysfs_remove_link(&(slave->dev.kobj), "master");
+
 	return ret;
 
 }
Signature · 1 line
1.7.10.4
author maintainer
#172 lore raw

PATCH 156/171 bonding: fix miimon and arp_interval delayed work race conditions

In reply to Steven Rostedt

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

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

From: "[email protected]" <[email protected]>

[ Upstream commit fbb0c41b814d497c656fc7be9e35456f139cb2fb ]

First I would give three observations which will be used later. Observation 1: if (delayed_work_pending(wq)) cancel_delayed_work(wq)
This usage is wrong because the pending bit is cleared just before the work's fn is executed and if the function re-arms itself we might end up with the work still running. It's safe to call cancel_delayed_work_sync() even if the work is not queued at all.
Observation 2: Use of INIT_DELAYED_WORK()
Work needs to be initialized only once prior to (de/en)queueing.
Observation 3: IFF_UP is set only after ndo_open is called

Related race conditions:
1. Race between bonding_store_miimon() and bonding_store_arp_interval() Because of Obs.1 we can end up having both works enqueued.
2. Multiple races with INIT_DELAYED_WORK()
Since the works are not protected by anything between INIT_DELAYED_WORK() and calls to (en/de)queue it is possible for races between the following functions:
(races are also possible between the calls to INIT_DELAYED_WORK()
and workqueue code)
bonding_store_miimon() - bonding_store_arp_interval(), bond_close(),
bond_open(), enqueued functions
bonding_store_arp_interval() - bonding_store_miimon(), bond_close(),
bond_open(), enqueued functions
3. By Obs.1 we need to change bond_cancel_all()

Bugs 1 and 2 are fixed by moving all work initializations in bond_open which by Obs. 2 and Obs. 3 and the fact that we make sure that all works are cancelled in bond_close(), is guaranteed not to have any work enqueued.
Also RTNL lock is now acquired in bonding_store_miimon/arp_interval so they can't race with bond_close and bond_open. The opposing work is cancelled only if the IFF_UP flag is set and it is cancelled unconditionally. The opposing work is already cancelled if the interface is down so no need to cancel it again. This way we don't need new synchronizations for the bonding workqueue. These bugs (and fixes) are tied together and belong in the same patch.
Note: I have left 1 line intentionally over 80 characters (84) because I
didn't like how it looks broken down. If you'd prefer it otherwise, then simply break it.

v2: Make description text < 75 columns

---
 drivers/net/bonding/bond_main.c  |   88 +++++++++++---------------------------
 drivers/net/bonding/bond_sysfs.c |   34 +++++----------
 2 files changed, 36 insertions(+), 86 deletions(-)
drivers/net/bonding/bond_main.c +26 -62
index 704d0ad..a9a07ce 100644
@@ -3411,6 +3411,28 @@ static int bond_xmit_hash_policy_l2(struct sk_buff *skb, int count)
 
 /*-------------------------- Device entry points ----------------------------*/
 
+static void bond_work_init_all(struct bonding *bond)
+{
+	INIT_DELAYED_WORK(&bond->mcast_work,
+			  bond_resend_igmp_join_requests_delayed);
+	INIT_DELAYED_WORK(&bond->alb_work, bond_alb_monitor);
+	INIT_DELAYED_WORK(&bond->mii_work, bond_mii_monitor);
+	if (bond->params.mode == BOND_MODE_ACTIVEBACKUP)
+		INIT_DELAYED_WORK(&bond->arp_work, bond_activebackup_arp_mon);
+	else
+		INIT_DELAYED_WORK(&bond->arp_work, bond_loadbalance_arp_mon);
+	INIT_DELAYED_WORK(&bond->ad_work, bond_3ad_state_machine_handler);
+}
+
+static void bond_work_cancel_all(struct bonding *bond)
+{
+	cancel_delayed_work_sync(&bond->mii_work);
+	cancel_delayed_work_sync(&bond->arp_work);
+	cancel_delayed_work_sync(&bond->alb_work);
+	cancel_delayed_work_sync(&bond->ad_work);
+	cancel_delayed_work_sync(&bond->mcast_work);
+}
+
 static int bond_open(struct net_device *bond_dev)
 {
 	struct bonding *bond = netdev_priv(bond_dev);
@@ -3433,41 +3455,27 @@ static int bond_open(struct net_device *bond_dev)
 	}
 	read_unlock(&bond->lock);
 
-	INIT_DELAYED_WORK(&bond->mcast_work, bond_resend_igmp_join_requests_delayed);
+	bond_work_init_all(bond);
 
 	if (bond_is_lb(bond)) {
 		/* bond_alb_initialize must be called before the timer
 		 * is started.
 		 */
-		if (bond_alb_initialize(bond, (bond->params.mode == BOND_MODE_ALB))) {
-			/* something went wrong - fail the open operation */
+		if (bond_alb_initialize(bond, (bond->params.mode == BOND_MODE_ALB)))
 			return -ENOMEM;
-		}
-
-		INIT_DELAYED_WORK(&bond->alb_work, bond_alb_monitor);
 		queue_delayed_work(bond->wq, &bond->alb_work, 0);
 	}
 
-	if (bond->params.miimon) {  /* link check interval, in milliseconds. */
-		INIT_DELAYED_WORK(&bond->mii_work, bond_mii_monitor);
+	if (bond->params.miimon)  /* link check interval, in milliseconds. */
 		queue_delayed_work(bond->wq, &bond->mii_work, 0);
-	}
 
 	if (bond->params.arp_interval) {  /* arp interval, in milliseconds. */
-		if (bond->params.mode == BOND_MODE_ACTIVEBACKUP)
-			INIT_DELAYED_WORK(&bond->arp_work,
-					  bond_activebackup_arp_mon);
-		else
-			INIT_DELAYED_WORK(&bond->arp_work,
-					  bond_loadbalance_arp_mon);
-
 		queue_delayed_work(bond->wq, &bond->arp_work, 0);
 		if (bond->params.arp_validate)
 			bond->recv_probe = bond_arp_rcv;
 	}
 
 	if (bond->params.mode == BOND_MODE_8023AD) {
-		INIT_DELAYED_WORK(&bond->ad_work, bond_3ad_state_machine_handler);
 		queue_delayed_work(bond->wq, &bond->ad_work, 0);
 		/* register to receive LACPDUs */
 		bond->recv_probe = bond_3ad_lacpdu_recv;
@@ -3482,34 +3490,10 @@ static int bond_close(struct net_device *bond_dev)
 	struct bonding *bond = netdev_priv(bond_dev);
 
 	write_lock_bh(&bond->lock);
-
 	bond->send_peer_notif = 0;
-
 	write_unlock_bh(&bond->lock);
 
-	if (bond->params.miimon) {  /* link check interval, in milliseconds. */
-		cancel_delayed_work_sync(&bond->mii_work);
-	}
-
-	if (bond->params.arp_interval) {  /* arp interval, in milliseconds. */
-		cancel_delayed_work_sync(&bond->arp_work);
-	}
-
-	switch (bond->params.mode) {
-	case BOND_MODE_8023AD:
-		cancel_delayed_work_sync(&bond->ad_work);
-		break;
-	case BOND_MODE_TLB:
-	case BOND_MODE_ALB:
-		cancel_delayed_work_sync(&bond->alb_work);
-		break;
-	default:
-		break;
-	}
-
-	if (delayed_work_pending(&bond->mcast_work))
-		cancel_delayed_work_sync(&bond->mcast_work);
-
+	bond_work_cancel_all(bond);
 	if (bond_is_lb(bond)) {
 		/* Must be called only after all
 		 * slaves have been released
@@ -4388,26 +4372,6 @@ static void bond_setup(struct net_device *bond_dev)
 	bond_dev->features |= bond_dev->hw_features;
 }
 
-static void bond_work_cancel_all(struct bonding *bond)
-{
-	if (bond->params.miimon && delayed_work_pending(&bond->mii_work))
-		cancel_delayed_work_sync(&bond->mii_work);
-
-	if (bond->params.arp_interval && delayed_work_pending(&bond->arp_work))
-		cancel_delayed_work_sync(&bond->arp_work);
-
-	if (bond->params.mode == BOND_MODE_ALB &&
-	    delayed_work_pending(&bond->alb_work))
-		cancel_delayed_work_sync(&bond->alb_work);
-
-	if (bond->params.mode == BOND_MODE_8023AD &&
-	    delayed_work_pending(&bond->ad_work))
-		cancel_delayed_work_sync(&bond->ad_work);
-
-	if (delayed_work_pending(&bond->mcast_work))
-		cancel_delayed_work_sync(&bond->mcast_work);
-}
-
 /*
 * Destroy a bonding device.
 * Must be under rtnl_lock when this function is called.
drivers/net/bonding/bond_sysfs.c +10 -24
index ecd47a3..13e37cc 100644
@@ -518,6 +518,8 @@ static ssize_t bonding_store_arp_interval(struct device *d,
 	int new_value, ret = count;
 	struct bonding *bond = to_bond(d);
 
+	if (!rtnl_trylock())
+		return restart_syscall();
 	if (sscanf(buf, "%d", &new_value) != 1) {
 		pr_err("%s: no arp_interval value specified.\n",
 		       bond->dev->name);
@@ -544,10 +546,6 @@ static ssize_t bonding_store_arp_interval(struct device *d,
 		pr_info("%s: ARP monitoring cannot be used with MII monitoring. %s Disabling MII monitoring.\n",
 			bond->dev->name, bond->dev->name);
 		bond->params.miimon = 0;
-		if (delayed_work_pending(&bond->mii_work)) {
-			cancel_delayed_work(&bond->mii_work);
-			flush_workqueue(bond->wq);
-		}
 	}
 	if (!bond->params.arp_targets[0]) {
 		pr_info("%s: ARP monitoring has been set up, but no ARP targets have been specified.\n",
@@ -559,19 +557,12 @@ static ssize_t bonding_store_arp_interval(struct device *d,
 		 * timer will get fired off when the open function
 		 * is called.
 		 */
-		if (!delayed_work_pending(&bond->arp_work)) {
-			if (bond->params.mode == BOND_MODE_ACTIVEBACKUP)
-				INIT_DELAYED_WORK(&bond->arp_work,
-						  bond_activebackup_arp_mon);
-			else
-				INIT_DELAYED_WORK(&bond->arp_work,
-						  bond_loadbalance_arp_mon);
-
-			queue_delayed_work(bond->wq, &bond->arp_work, 0);
-		}
+		cancel_delayed_work_sync(&bond->mii_work);
+		queue_delayed_work(bond->wq, &bond->arp_work, 0);
 	}
 
 out:
+	rtnl_unlock();
 	return ret;
 }
 static DEVICE_ATTR(arp_interval, S_IRUGO | S_IWUSR,
@@ -967,6 +958,8 @@ static ssize_t bonding_store_miimon(struct device *d,
 	int new_value, ret = count;
 	struct bonding *bond = to_bond(d);
 
+	if (!rtnl_trylock())
+		return restart_syscall();
 	if (sscanf(buf, "%d", &new_value) != 1) {
 		pr_err("%s: no miimon value specified.\n",
 		       bond->dev->name);
@@ -998,10 +991,6 @@ static ssize_t bonding_store_miimon(struct device *d,
 				bond->params.arp_validate =
 					BOND_ARP_VALIDATE_NONE;
 			}
-			if (delayed_work_pending(&bond->arp_work)) {
-				cancel_delayed_work(&bond->arp_work);
-				flush_workqueue(bond->wq);
-			}
 		}
 
 		if (bond->dev->flags & IFF_UP) {
@@ -1010,15 +999,12 @@ static ssize_t bonding_store_miimon(struct device *d,
 			 * timer will get fired off when the open function
 			 * is called.
 			 */
-			if (!delayed_work_pending(&bond->mii_work)) {
-				INIT_DELAYED_WORK(&bond->mii_work,
-						  bond_mii_monitor);
-				queue_delayed_work(bond->wq,
-						   &bond->mii_work, 0);
-			}
+			cancel_delayed_work_sync(&bond->arp_work);
+			queue_delayed_work(bond->wq, &bond->mii_work, 0);
 		}
 	}
 out:
+	rtnl_unlock();
 	return ret;
 }
 static DEVICE_ATTR(miimon, S_IRUGO | S_IWUSR,
Signature · 1 line
1.7.10.4
author maintainer
#173 lore raw

PATCH 157/171 bonding: fix disabling of arp_interval and miimon

In reply to Steven Rostedt

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

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

From: "[email protected]" <[email protected]>

[ Upstream commit 1bc7db16782c2a581fb4d53ca853631050f31611 ]

Currently if either arp_interval or miimon is disabled, they both get disabled, and upon disabling they get executed once more which is not the proper behaviour. Also when doing a no-op and disabling an already disabled one, the other again gets disabled.
Also fix the error messages with the proper valid ranges, and a small typo fix in the up delay error message (outputting "down delay", instead of "up delay").

---
 drivers/net/bonding/bond_sysfs.c |   92 ++++++++++++++++++++------------------
 1 file changed, 48 insertions(+), 44 deletions(-)
drivers/net/bonding/bond_sysfs.c +48 -44
index 13e37cc..53eeac2 100644
@@ -527,7 +527,7 @@ static ssize_t bonding_store_arp_interval(struct device *d,
 		goto out;
 	}
 	if (new_value < 0) {
-		pr_err("%s: Invalid arp_interval value %d not in range 1-%d; rejected.\n",
+		pr_err("%s: Invalid arp_interval value %d not in range 0-%d; rejected.\n",
 		       bond->dev->name, new_value, INT_MAX);
 		ret = -EINVAL;
 		goto out;
@@ -542,14 +542,15 @@ static ssize_t bonding_store_arp_interval(struct device *d,
 	pr_info("%s: Setting ARP monitoring interval to %d.\n",
 		bond->dev->name, new_value);
 	bond->params.arp_interval = new_value;
-	if (bond->params.miimon) {
-		pr_info("%s: ARP monitoring cannot be used with MII monitoring. %s Disabling MII monitoring.\n",
-			bond->dev->name, bond->dev->name);
-		bond->params.miimon = 0;
-	}
-	if (!bond->params.arp_targets[0]) {
-		pr_info("%s: ARP monitoring has been set up, but no ARP targets have been specified.\n",
-			bond->dev->name);
+	if (new_value) {
+		if (bond->params.miimon) {
+			pr_info("%s: ARP monitoring cannot be used with MII monitoring. %s Disabling MII monitoring.\n",
+				bond->dev->name, bond->dev->name);
+			bond->params.miimon = 0;
+		}
+		if (!bond->params.arp_targets[0])
+			pr_info("%s: ARP monitoring has been set up, but no ARP targets have been specified.\n",
+				bond->dev->name);
 	}
 	if (bond->dev->flags & IFF_UP) {
 		/* If the interface is up, we may need to fire off
@@ -557,10 +558,13 @@ static ssize_t bonding_store_arp_interval(struct device *d,
 		 * timer will get fired off when the open function
 		 * is called.
 		 */
-		cancel_delayed_work_sync(&bond->mii_work);
-		queue_delayed_work(bond->wq, &bond->arp_work, 0);
+		if (!new_value) {
+			cancel_delayed_work_sync(&bond->arp_work);
+		} else {
+			cancel_delayed_work_sync(&bond->mii_work);
+			queue_delayed_work(bond->wq, &bond->arp_work, 0);
+		}
 	}
-
 out:
 	rtnl_unlock();
 	return ret;
@@ -702,7 +706,7 @@ static ssize_t bonding_store_downdelay(struct device *d,
 	}
 	if (new_value < 0) {
 		pr_err("%s: Invalid down delay value %d not in range %d-%d; rejected.\n",
-		       bond->dev->name, new_value, 1, INT_MAX);
+		       bond->dev->name, new_value, 0, INT_MAX);
 		ret = -EINVAL;
 		goto out;
 	} else {
@@ -757,8 +761,8 @@ static ssize_t bonding_store_updelay(struct device *d,
 		goto out;
 	}
 	if (new_value < 0) {
-		pr_err("%s: Invalid down delay value %d not in range %d-%d; rejected.\n",
-		       bond->dev->name, new_value, 1, INT_MAX);
+		pr_err("%s: Invalid up delay value %d not in range %d-%d; rejected.\n",
+		       bond->dev->name, new_value, 0, INT_MAX);
 		ret = -EINVAL;
 		goto out;
 	} else {
@@ -968,37 +972,37 @@ static ssize_t bonding_store_miimon(struct device *d,
 	}
 	if (new_value < 0) {
 		pr_err("%s: Invalid miimon value %d not in range %d-%d; rejected.\n",
-		       bond->dev->name, new_value, 1, INT_MAX);
+		       bond->dev->name, new_value, 0, INT_MAX);
 		ret = -EINVAL;
 		goto out;
-	} else {
-		pr_info("%s: Setting MII monitoring interval to %d.\n",
-			bond->dev->name, new_value);
-		bond->params.miimon = new_value;
-		if (bond->params.updelay)
-			pr_info("%s: Note: Updating updelay (to %d) since it is a multiple of the miimon value.\n",
-				bond->dev->name,
-				bond->params.updelay * bond->params.miimon);
-		if (bond->params.downdelay)
-			pr_info("%s: Note: Updating downdelay (to %d) since it is a multiple of the miimon value.\n",
-				bond->dev->name,
-				bond->params.downdelay * bond->params.miimon);
-		if (bond->params.arp_interval) {
-			pr_info("%s: MII monitoring cannot be used with ARP monitoring. Disabling ARP monitoring...\n",
-				bond->dev->name);
-			bond->params.arp_interval = 0;
-			if (bond->params.arp_validate) {
-				bond->params.arp_validate =
-					BOND_ARP_VALIDATE_NONE;
-			}
-		}
-
-		if (bond->dev->flags & IFF_UP) {
-			/* If the interface is up, we may need to fire off
-			 * the MII timer. If the interface is down, the
-			 * timer will get fired off when the open function
-			 * is called.
-			 */
+	}
+	pr_info("%s: Setting MII monitoring interval to %d.\n",
+		bond->dev->name, new_value);
+	bond->params.miimon = new_value;
+	if (bond->params.updelay)
+		pr_info("%s: Note: Updating updelay (to %d) since it is a multiple of the miimon value.\n",
+			bond->dev->name,
+			bond->params.updelay * bond->params.miimon);
+	if (bond->params.downdelay)
+		pr_info("%s: Note: Updating downdelay (to %d) since it is a multiple of the miimon value.\n",
+			bond->dev->name,
+			bond->params.downdelay * bond->params.miimon);
+	if (new_value && bond->params.arp_interval) {
+		pr_info("%s: MII monitoring cannot be used with ARP monitoring. Disabling ARP monitoring...\n",
+			bond->dev->name);
+		bond->params.arp_interval = 0;
+		if (bond->params.arp_validate)
+			bond->params.arp_validate = BOND_ARP_VALIDATE_NONE;
+	}
+	if (bond->dev->flags & IFF_UP) {
+		/* If the interface is up, we may need to fire off
+		 * the MII timer. If the interface is down, the
+		 * timer will get fired off when the open function
+		 * is called.
+		 */
+		if (!new_value) {
+			cancel_delayed_work_sync(&bond->mii_work);
+		} else {
 			cancel_delayed_work_sync(&bond->arp_work);
 			queue_delayed_work(bond->wq, &bond->mii_work, 0);
 		}
Signature · 1 line
1.7.10.4
author maintainer
#174 lore raw

PATCH 158/171 drivers: net: ethernet: davinci_emac: use netif_wake_queue() while restarting tx queue

In reply to Steven Rostedt

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

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

From: Mugunthan V N <[email protected]>

[ Upstream commit 7e51cde276ca820d526c6c21cf8147df595a36bf ]

To restart tx queue use netif_wake_queue() intead of netif_start_queue() so that net schedule will restart transmission immediately which will increase network performance while doing huge data transfers.

---
 drivers/net/ethernet/ti/davinci_emac.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/net/ethernet/ti/davinci_emac.c +1 -1
index fce89a0..93c040d 100644
@@ -1055,7 +1055,7 @@ static void emac_tx_handler(void *token, int len, int status)
 	atomic_dec(&priv->cur_tx);
 
 	if (unlikely(netif_queue_stopped(ndev)))
-		netif_start_queue(ndev);
+		netif_wake_queue(ndev);
 	ndev->stats.tx_packets++;
 	ndev->stats.tx_bytes += len;
 	dev_kfree_skb_any(skb);
Signature · 1 line
1.7.10.4
author maintainer
#175 lore raw

PATCH 159/171 drivers: net: ethernet: cpsw: use netif_wake_queue() while restarting tx queue

In reply to Steven Rostedt

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

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

From: Mugunthan V N <[email protected]>

[ Upstream commit b56d6b3fca6d1214dbc9c5655f26e5d4ec04afc8 ]

To restart tx queue use netif_wake_queue() intead of netif_start_queue() so that net schedule will restart transmission immediately which will increase network performance while doing huge data transfers.

---
 drivers/net/ethernet/ti/cpsw.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/net/ethernet/ti/cpsw.c +1 -1
index 1e5d85b..0ab0a10 100644
@@ -250,7 +250,7 @@ void cpsw_tx_handler(void *token, int len, int status)
 	struct cpsw_priv	*priv = netdev_priv(ndev);
 
 	if (unlikely(netif_queue_stopped(ndev)))
-		netif_start_queue(ndev);
+		netif_wake_queue(ndev);
 	priv->stats.tx_packets++;
 	priv->stats.tx_bytes += len;
 	dev_kfree_skb_any(skb);
Signature · 1 line
1.7.10.4
author maintainer
#176 lore raw

PATCH 160/171 aoe: reserve enough headroom on skbs

In reply to Steven Rostedt

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

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

From: Eric Dumazet <[email protected]>

[ Upstream commit 91c5746425aed8f7188a351f1224a26aa232e4b3 ]

Some network drivers use a non default hard_header_len

Transmitted skb should take into account dev->hard_header_len, or risk crashes or expensive reallocations.

In the case of aoe, lets reserve MAX_HEADER bytes.

David reported a crash in defxx driver, solved by this patch.

---
 drivers/block/aoe/aoecmd.c |    3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)
drivers/block/aoe/aoecmd.c +2 -1
index 887f68f..db30542 100644
@@ -30,8 +30,9 @@ new_skb(ulong len)
 {
 	struct sk_buff *skb;
 
-	skb = alloc_skb(len, GFP_ATOMIC);
+	skb = alloc_skb(len + MAX_HEADER, GFP_ATOMIC);
 	if (skb) {
+		skb_reserve(skb, MAX_HEADER);
 		skb_reset_mac_header(skb);
 		skb_reset_network_header(skb);
 		skb->protocol = __constant_htons(ETH_P_AOE);
Signature · 1 line
1.7.10.4
author maintainer
#177 lore raw

PATCH 161/171 atl1e: drop pci-msi support because of packet corruption

In reply to Steven Rostedt

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

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

From: Hannes Frederic Sowa <[email protected]>

[ Upstream commit 188ab1b105c96656f6bcfb49d0d8bb1b1936b632 ]

Usage of pci-msi results in corrupted dma packet transfers to the host.

---
 drivers/net/ethernet/atheros/atl1e/atl1e.h      |    1 -
 drivers/net/ethernet/atheros/atl1e/atl1e_main.c |   19 ++-----------------
 2 files changed, 2 insertions(+), 18 deletions(-)
drivers/net/ethernet/atheros/atl1e/atl1e.h +0 -1
index 829b5ad..edfdf6b 100644
@@ -438,7 +438,6 @@ struct atl1e_adapter {
 	struct atl1e_hw        hw;
 	struct atl1e_hw_stats  hw_stats;
 
-	bool have_msi;
 	u32 wol;
 	u16 link_speed;
 	u16 link_duplex;
drivers/net/ethernet/atheros/atl1e/atl1e_main.c +2 -17
index a98acc8..b986503 100644
@@ -1851,34 +1851,19 @@ static void atl1e_free_irq(struct atl1e_adapter *adapter)
 	struct net_device *netdev = adapter->netdev;
 
 	free_irq(adapter->pdev->irq, netdev);
-
-	if (adapter->have_msi)
-		pci_disable_msi(adapter->pdev);
 }
 
 static int atl1e_request_irq(struct atl1e_adapter *adapter)
 {
 	struct pci_dev    *pdev   = adapter->pdev;
 	struct net_device *netdev = adapter->netdev;
-	int flags = 0;
 	int err = 0;
 
-	adapter->have_msi = true;
-	err = pci_enable_msi(pdev);
-	if (err) {
-		netdev_dbg(netdev,
-			   "Unable to allocate MSI interrupt Error: %d\n", err);
-		adapter->have_msi = false;
-	}
-
-	if (!adapter->have_msi)
-		flags |= IRQF_SHARED;
-	err = request_irq(pdev->irq, atl1e_intr, flags, netdev->name, netdev);
+	err = request_irq(pdev->irq, atl1e_intr, IRQF_SHARED, netdev->name,
+			  netdev);
 	if (err) {
 		netdev_dbg(adapter->netdev,
 			   "Unable to allocate interrupt Error: %d\n", err);
-		if (adapter->have_msi)
-			pci_disable_msi(pdev);
 		return err;
 	}
 	netdev_dbg(netdev, "atl1e_request_irq OK\n");
Signature · 1 line
1.7.10.4
author maintainer
#178 lore raw

PATCH 162/171 DM9000B: driver initialization upgrade

In reply to Steven Rostedt

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

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

From: Joseph CHANG <[email protected]>

[ Upstream commit 6741f40d198c6a5feb23653a1efd4ca47f93d83d ]

Fix bug for DM9000 revision B which contain a DSP PHY

DM9000B use DSP PHY instead previouse DM9000 revisions' analog PHY, So need extra change in initialization, For
explicity PHY Reset and PHY init parameter, and
first DM9000_NCR reset need NCR_MAC_LBK bit by dm9000_probe().

Following DM9000_NCR reset cause by dm9000_open() clear the NCR_MAC_LBK bit.

Without this fix, Power-up FIFO pointers error happen around 2% rate among Davicom's customers' boards. With this fix, All above cases can be solved.

---
 drivers/net/ethernet/davicom/dm9000.c |  214 +++++++++++++++++----------------
 drivers/net/ethernet/davicom/dm9000.h |   11 +-
 2 files changed, 120 insertions(+), 105 deletions(-)
drivers/net/ethernet/davicom/dm9000.c +110 -104
index 36499d5..a9628b6 100644
@@ -257,6 +257,107 @@ static void dm9000_dumpblk_32bit(void __iomem *reg, int count)
 		tmp = readl(reg);
 }
 
+/*
+ * Sleep, either by using msleep() or if we are suspending, then
+ * use mdelay() to sleep.
+ */
+static void dm9000_msleep(board_info_t *db, unsigned int ms)
+{
+	if (db->in_suspend)
+		mdelay(ms);
+	else
+		msleep(ms);
+}
+
+/* Read a word from phyxcer */
+static int
+dm9000_phy_read(struct net_device *dev, int phy_reg_unused, int reg)
+{
+	board_info_t *db = netdev_priv(dev);
+	unsigned long flags;
+	unsigned int reg_save;
+	int ret;
+
+	mutex_lock(&db->addr_lock);
+
+	spin_lock_irqsave(&db->lock, flags);
+
+	/* Save previous register address */
+	reg_save = readb(db->io_addr);
+
+	/* Fill the phyxcer register into REG_0C */
+	iow(db, DM9000_EPAR, DM9000_PHY | reg);
+
+	/* Issue phyxcer read command */
+	iow(db, DM9000_EPCR, EPCR_ERPRR | EPCR_EPOS);
+
+	writeb(reg_save, db->io_addr);
+	spin_unlock_irqrestore(&db->lock, flags);
+
+	dm9000_msleep(db, 1);		/* Wait read complete */
+
+	spin_lock_irqsave(&db->lock, flags);
+	reg_save = readb(db->io_addr);
+
+	iow(db, DM9000_EPCR, 0x0);	/* Clear phyxcer read command */
+
+	/* The read data keeps on REG_0D & REG_0E */
+	ret = (ior(db, DM9000_EPDRH) << 8) | ior(db, DM9000_EPDRL);
+
+	/* restore the previous address */
+	writeb(reg_save, db->io_addr);
+	spin_unlock_irqrestore(&db->lock, flags);
+
+	mutex_unlock(&db->addr_lock);
+
+	dm9000_dbg(db, 5, "phy_read[%02x] -> %04x\n", reg, ret);
+	return ret;
+}
+
+/* Write a word to phyxcer */
+static void
+dm9000_phy_write(struct net_device *dev,
+		 int phyaddr_unused, int reg, int value)
+{
+	board_info_t *db = netdev_priv(dev);
+	unsigned long flags;
+	unsigned long reg_save;
+
+	dm9000_dbg(db, 5, "phy_write[%02x] = %04x\n", reg, value);
+	mutex_lock(&db->addr_lock);
+
+	spin_lock_irqsave(&db->lock, flags);
+
+	/* Save previous register address */
+	reg_save = readb(db->io_addr);
+
+	/* Fill the phyxcer register into REG_0C */
+	iow(db, DM9000_EPAR, DM9000_PHY | reg);
+
+	/* Fill the written data into REG_0D & REG_0E */
+	iow(db, DM9000_EPDRL, value);
+	iow(db, DM9000_EPDRH, value >> 8);
+
+	/* Issue phyxcer write command */
+	iow(db, DM9000_EPCR, EPCR_EPOS | EPCR_ERPRW);
+
+	writeb(reg_save, db->io_addr);
+	spin_unlock_irqrestore(&db->lock, flags);
+
+	dm9000_msleep(db, 1);		/* Wait write complete */
+
+	spin_lock_irqsave(&db->lock, flags);
+	reg_save = readb(db->io_addr);
+
+	iow(db, DM9000_EPCR, 0x0);	/* Clear phyxcer write command */
+
+	/* restore the previous address */
+	writeb(reg_save, db->io_addr);
+
+	spin_unlock_irqrestore(&db->lock, flags);
+	mutex_unlock(&db->addr_lock);
+}
+
 /* dm9000_set_io
  *
  * select the specified set of io routines to use with the
@@ -794,6 +895,9 @@ dm9000_init_dm9000(struct net_device *dev)
 
 	iow(db, DM9000_GPCR, GPCR_GEP_CNTL);	/* Let GPIO0 output */
 
+	dm9000_phy_write(dev, 0, MII_BMCR, BMCR_RESET); /* PHY RESET */
+	dm9000_phy_write(dev, 0, MII_DM_DSPCR, DSPCR_INIT_PARAM); /* Init */
+
 	ncr = (db->flags & DM9000_PLATF_EXT_PHY) ? NCR_EXT_PHY : 0;
 
 	/* if wol is needed, then always set NCR_WAKEEN otherwise we end
@@ -1200,109 +1304,6 @@ dm9000_open(struct net_device *dev)
 	return 0;
 }
 
-/*
- * Sleep, either by using msleep() or if we are suspending, then
- * use mdelay() to sleep.
- */
-static void dm9000_msleep(board_info_t *db, unsigned int ms)
-{
-	if (db->in_suspend)
-		mdelay(ms);
-	else
-		msleep(ms);
-}
-
-/*
- *   Read a word from phyxcer
- */
-static int
-dm9000_phy_read(struct net_device *dev, int phy_reg_unused, int reg)
-{
-	board_info_t *db = netdev_priv(dev);
-	unsigned long flags;
-	unsigned int reg_save;
-	int ret;
-
-	mutex_lock(&db->addr_lock);
-
-	spin_lock_irqsave(&db->lock,flags);
-
-	/* Save previous register address */
-	reg_save = readb(db->io_addr);
-
-	/* Fill the phyxcer register into REG_0C */
-	iow(db, DM9000_EPAR, DM9000_PHY | reg);
-
-	iow(db, DM9000_EPCR, EPCR_ERPRR | EPCR_EPOS);	/* Issue phyxcer read command */
-
-	writeb(reg_save, db->io_addr);
-	spin_unlock_irqrestore(&db->lock,flags);
-
-	dm9000_msleep(db, 1);		/* Wait read complete */
-
-	spin_lock_irqsave(&db->lock,flags);
-	reg_save = readb(db->io_addr);
-
-	iow(db, DM9000_EPCR, 0x0);	/* Clear phyxcer read command */
-
-	/* The read data keeps on REG_0D & REG_0E */
-	ret = (ior(db, DM9000_EPDRH) << 8) | ior(db, DM9000_EPDRL);
-
-	/* restore the previous address */
-	writeb(reg_save, db->io_addr);
-	spin_unlock_irqrestore(&db->lock,flags);
-
-	mutex_unlock(&db->addr_lock);
-
-	dm9000_dbg(db, 5, "phy_read[%02x] -> %04x\n", reg, ret);
-	return ret;
-}
-
-/*
- *   Write a word to phyxcer
- */
-static void
-dm9000_phy_write(struct net_device *dev,
-		 int phyaddr_unused, int reg, int value)
-{
-	board_info_t *db = netdev_priv(dev);
-	unsigned long flags;
-	unsigned long reg_save;
-
-	dm9000_dbg(db, 5, "phy_write[%02x] = %04x\n", reg, value);
-	mutex_lock(&db->addr_lock);
-
-	spin_lock_irqsave(&db->lock,flags);
-
-	/* Save previous register address */
-	reg_save = readb(db->io_addr);
-
-	/* Fill the phyxcer register into REG_0C */
-	iow(db, DM9000_EPAR, DM9000_PHY | reg);
-
-	/* Fill the written data into REG_0D & REG_0E */
-	iow(db, DM9000_EPDRL, value);
-	iow(db, DM9000_EPDRH, value >> 8);
-
-	iow(db, DM9000_EPCR, EPCR_EPOS | EPCR_ERPRW);	/* Issue phyxcer write command */
-
-	writeb(reg_save, db->io_addr);
-	spin_unlock_irqrestore(&db->lock, flags);
-
-	dm9000_msleep(db, 1);		/* Wait write complete */
-
-	spin_lock_irqsave(&db->lock,flags);
-	reg_save = readb(db->io_addr);
-
-	iow(db, DM9000_EPCR, 0x0);	/* Clear phyxcer write command */
-
-	/* restore the previous address */
-	writeb(reg_save, db->io_addr);
-
-	spin_unlock_irqrestore(&db->lock, flags);
-	mutex_unlock(&db->addr_lock);
-}
-
 static void
 dm9000_shutdown(struct net_device *dev)
 {
@@ -1501,7 +1502,12 @@ dm9000_probe(struct platform_device *pdev)
 	db->flags |= DM9000_PLATF_SIMPLE_PHY;
 #endif
 
-	dm9000_reset(db);
+	/* Fixing bug on dm9000_probe, takeover dm9000_reset(db),
+	 * Need 'NCR_MAC_LBK' bit to indeed stable our DM9000 fifo
+	 * while probe stage.
+	 */
+
+	iow(db, DM9000_NCR, NCR_MAC_LBK | NCR_RST);
 
 	/* try multiple times, DM9000 sometimes gets the read wrong */
 	for (i = 0; i < 8; i++) {
drivers/net/ethernet/davicom/dm9000.h +10 -1
index 55688bd..9ce058a 100644
@@ -69,7 +69,9 @@
 #define NCR_WAKEEN          (1<<6)
 #define NCR_FCOL            (1<<4)
 #define NCR_FDX             (1<<3)
-#define NCR_LBK             (3<<1)
+
+#define NCR_RESERVED        (3<<1)
+#define NCR_MAC_LBK         (1<<1)
 #define NCR_RST	            (1<<0)
 
 #define NSR_SPEED           (1<<7)
@@ -167,5 +169,12 @@
 #define ISR_LNKCHNG		(1<<5)
 #define ISR_UNDERRUN		(1<<4)
 
+/* Davicom MII registers.
+ */
+
+#define MII_DM_DSPCR		0x1b    /* DSP Control Register */
+
+#define DSPCR_INIT_PARAM	0xE100	/* DSP init parameter */
+
 #endif /* _DM9000X_H_ */
 
Signature · 1 line
1.7.10.4
author maintainer
#179 lore raw

PATCH 163/171 ipv6: fix bad free of addrconf_init_net

In reply to Steven Rostedt

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

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

From: Hong Zhiguo <[email protected]>

[ Upstream commit a79ca223e029aa4f09abb337accf1812c900a800 ]

---
 net/ipv6/addrconf.c |   26 ++++++++++----------------
 1 file changed, 10 insertions(+), 16 deletions(-)
net/ipv6/addrconf.c +10 -16
index fd82a30..ec627b5 100644
@@ -4653,26 +4653,20 @@ static void addrconf_sysctl_unregister(struct inet6_dev *idev)
 
 static int __net_init addrconf_init_net(struct net *net)
 {
-	int err;
+	int err = -ENOMEM;
 	struct ipv6_devconf *all, *dflt;
 
-	err = -ENOMEM;
-	all = &ipv6_devconf;
-	dflt = &ipv6_devconf_dflt;
+	all = kmemdup(&ipv6_devconf, sizeof(ipv6_devconf), GFP_KERNEL);
+	if (all == NULL)
+		goto err_alloc_all;
 
-	if (!net_eq(net, &init_net)) {
-		all = kmemdup(all, sizeof(ipv6_devconf), GFP_KERNEL);
-		if (all == NULL)
-			goto err_alloc_all;
+	dflt = kmemdup(&ipv6_devconf_dflt, sizeof(ipv6_devconf_dflt), GFP_KERNEL);
+	if (dflt == NULL)
+		goto err_alloc_dflt;
 
-		dflt = kmemdup(dflt, sizeof(ipv6_devconf_dflt), GFP_KERNEL);
-		if (dflt == NULL)
-			goto err_alloc_dflt;
-	} else {
-		/* these will be inherited by all namespaces */
-		dflt->autoconf = ipv6_defaults.autoconf;
-		dflt->disable_ipv6 = ipv6_defaults.disable_ipv6;
-	}
+	/* these will be inherited by all namespaces */
+	dflt->autoconf = ipv6_defaults.autoconf;
+	dflt->disable_ipv6 = ipv6_defaults.disable_ipv6;
 
 	net->ipv6.devconf_all = all;
 	net->ipv6.devconf_dflt = dflt;
Signature · 1 line
1.7.10.4
#180 lore raw

Re: [ 163/171 ] ipv6: fix bad free of addrconf_init_net

On 04/12/2013 04:27 AM, Steven Rostedt wrote:
On linux with stable kernel v3.4.39, when this patch (ipv6: fix bad free of addrconf_init_net) is applied. Then I run the following commands:

1. the default value of ipv6 forwarding is 0

2. we change it to 1 firstly via:

# vi /etc/sysctl.conf
Uncomment: net.ipv6.conf.all.forwarding=1
#sysctl �p
Then we can verify via
#sysctl �a | grep forwarding

3, then we create a new namespace via:

# ip netns add fib_100
# ip netns exec fib_100 sysctl -a | grep forwarding

We�ll see in this new name space, the value of ipv6 forwarding will be set to �1� in v3.4.38, but in v3.4.39, the value is still be �0�

without this patch (ipv6: fix bad free of addrconf_init_net), the value of ipv6 forwarding will be set to �1�.

Would you like to tell me which one (v4.3.39 and v4.3.38) is correct?

Best Regards!
Zhu Yanjun

author maintainer
#181 lore raw

PATCH 164/171 ipv6: dont accept multicast traffic with scope 0

In reply to Steven Rostedt

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

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

From: Hannes Frederic Sowa <[email protected]>

[ Upstream commit 20314092c1b41894d8c181bf9aa6f022be2416aa ]

v2:
a) moved before multicast source address check
b) changed comment to netdev style

---
 net/ipv6/ip6_input.c |    9 +++++++++
 1 file changed, 9 insertions(+)
net/ipv6/ip6_input.c +9 -0
index b196852..db70b88 100644
@@ -118,6 +118,15 @@ int ipv6_rcv(struct sk_buff *skb, struct net_device *dev, struct packet_type *pt
 	    ipv6_addr_loopback(&hdr->daddr))
 		goto err;
 
+	/* RFC4291 2.7
+	 * Nodes must not originate a packet to a multicast address whose scope
+	 * field contains the reserved value 0; if such a packet is received, it
+	 * must be silently dropped.
+	 */
+	if (ipv6_addr_is_multicast(&hdr->daddr) &&
+	    IPV6_ADDR_MC_SCOPE(&hdr->daddr) == 0)
+		goto err;
+
 	/*
 	 * RFC4291 2.7
 	 * Multicast addresses must not be used as source addresses in IPv6
Signature · 1 line
1.7.10.4
author maintainer
#182 lore raw

PATCH 165/171 ipv6: dont accept node local multicast traffic from the wire

In reply to Steven Rostedt

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

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

From: Hannes Frederic Sowa <[email protected]>

[ Upstream commit 1c4a154e5253687c51123956dfcee9e9dfa8542d ]

Erik Hugne's errata proposal (Errata ID: 3480) to RFC4291 has been
verified: http://www.rfc-editor.org/errata_search.php?eid=3480

We have to check for pkt_type and loopback flag because either the packets are allowed to travel over the loopback interface (in which case pkt_type is PACKET_HOST and IFF_LOOPBACK flag is set) or they travel over a non-loopback interface back to us (in which case PACKET_TYPE is PACKET_LOOPBACK and IFF_LOOPBACK flag is not set).

---
 net/ipv6/ip6_input.c |   12 ++++++++++++
 1 file changed, 12 insertions(+)
net/ipv6/ip6_input.c +12 -0
index db70b88..dee9964 100644
@@ -118,6 +118,18 @@ int ipv6_rcv(struct sk_buff *skb, struct net_device *dev, struct packet_type *pt
 	    ipv6_addr_loopback(&hdr->daddr))
 		goto err;
 
+	/* RFC4291 Errata ID: 3480
+	 * Interface-Local scope spans only a single interface on a
+	 * node and is useful only for loopback transmission of
+	 * multicast.  Packets with interface-local scope received
+	 * from another node must be discarded.
+	 */
+	if (!(skb->pkt_type == PACKET_LOOPBACK ||
+	      dev->flags & IFF_LOOPBACK) &&
+	    ipv6_addr_is_multicast(&hdr->daddr) &&
+	    IPV6_ADDR_MC_SCOPE(&hdr->daddr) == 1)
+		goto err;
+
 	/* RFC4291 2.7
 	 * Nodes must not originate a packet to a multicast address whose scope
 	 * field contains the reserved value 0; if such a packet is received, it
Signature · 1 line
1.7.10.4
author maintainer
#183 lore raw

PATCH 166/171 ks8851: Fix interpretation of rxlen field.

In reply to Steven Rostedt

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

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

From: "[email protected]" <[email protected]>

[ Upstream commit 14bc435ea54cb888409efb54fc6b76c13ef530e9 ]

According to the Datasheet (page 52):
15-12 Reserved
11-0 RXBC Receive Byte Count
This field indicates the present received frame byte size.

The code has a bug:

                 rxh = ks8851_rdreg32(ks, KS_RXFHSR);
                 rxstat = rxh & 0xffff;
                 rxlen = rxh >> 16; // BUG!!! 0xFFF mask should be applied
---
 drivers/net/ethernet/micrel/ks8851.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/net/ethernet/micrel/ks8851.c +1 -1
index 1540ebe..a4cd9a0 100644
@@ -547,7 +547,7 @@ static void ks8851_rx_pkts(struct ks8851_net *ks)
 	for (; rxfc != 0; rxfc--) {
 		rxh = ks8851_rdreg32(ks, KS_RXFHSR);
 		rxstat = rxh & 0xffff;
-		rxlen = rxh >> 16;
+		rxlen = (rxh >> 16) & 0xfff;
 
 		netif_dbg(ks, rx_status, ks->netdev,
 			  "rx: stat 0x%04x, len 0x%04x\n", rxstat, rxlen);
Signature · 1 line
1.7.10.4
author maintainer
#184 lore raw

PATCH 167/171 net: add a synchronize_net() in netdev_rx_handler_unregister()

In reply to Steven Rostedt

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

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

From: Eric Dumazet <[email protected]>

[ Upstream commit 00cfec37484761a44a3b6f4675a54caa618210ae ]

commit 35d48903e97819 (bonding: fix rx_handler locking) added a race in bonding driver, reported by Steven Rostedt who did a very good diagnosis :

<quoting Steven>

I'm currently debugging a crash in an old 3.0-rt kernel that one of our customers is seeing. The bug happens with a stress test that loads and unloads the bonding module in a loop (I don't know all the details as I'm not the one that is directly interacting with the customer). But the bug looks to be something that may still be present and possibly present in mainline too. It will just be much harder to trigger it in mainline.

In -rt, interrupts are threads, and can schedule in and out just like any other thread. Note, mainline now supports interrupt threads so this may be easily reproducible in mainline as well. I don't have the ability to tell the customer to try mainline or other kernels, so my hands are somewhat tied to what I can do.

But according to a core dump, I tracked down that the eth irq thread crashed in bond_handle_frame() here:

        slave = bond_slave_get_rcu(skb->dev);
        bond = slave->bond; <--- BUG

the slave returned was NULL and accessing slave->bond caused a NULL pointer dereference.

Looking at the code that unregisters the handler:

void netdev_rx_handler_unregister(struct net_device *dev) {

        ASSERT_RTNL();
        RCU_INIT_POINTER(dev->rx_handler, NULL);
        RCU_INIT_POINTER(dev->rx_handler_data, NULL);
}

Which is basically:

        dev->rx_handler = NULL;
        dev->rx_handler_data = NULL;

And looking at __netif_receive_skb() we have:

        rx_handler = rcu_dereference(skb->dev->rx_handler);
        if (rx_handler) {
                if (pt_prev) {
                        ret = deliver_skb(skb, pt_prev, orig_dev);
                        pt_prev = NULL;
                }
                switch (rx_handler(&skb)) {

My question to all of you is, what stops this interrupt from happening while the bonding module is unloading? What happens if the interrupt triggers and we have this:

        CPU0                    CPU1
        ----                    ----

rx_handler = skb->dev->rx_handler

                        netdev_rx_handler_unregister() {
                           dev->rx_handler = NULL;
                           dev->rx_handler_data = NULL;
  rx_handler()
   bond_handle_frame() {
    slave = skb->dev->rx_handler;
    bond = slave->bond; <-- NULL pointer dereference!!!

What protection am I missing in the bond release handler that would prevent the above from happening?

</quoting Steven>

We can fix bug this in two ways. First is adding a test in bond_handle_frame() and others to check if rx_handler_data is NULL.

A second way is adding a synchronize_net() in netdev_rx_handler_unregister() to make sure that a rcu protected reader has the guarantee to see a non NULL rx_handler_data.

The second way is better as it avoids an extra test in fast path.

---
 net/core/dev.c |    6 ++++++
 1 file changed, 6 insertions(+)
net/core/dev.c +6 -0
index d283020..9509394 100644
@@ -3153,6 +3153,7 @@ int netdev_rx_handler_register(struct net_device *dev,
 	if (dev->rx_handler)
 		return -EBUSY;
 
+	/* Note: rx_handler_data must be set before rx_handler */
 	rcu_assign_pointer(dev->rx_handler_data, rx_handler_data);
 	rcu_assign_pointer(dev->rx_handler, rx_handler);
 
@@ -3173,6 +3174,11 @@ void netdev_rx_handler_unregister(struct net_device *dev)
 
 	ASSERT_RTNL();
 	RCU_INIT_POINTER(dev->rx_handler, NULL);
+	/* a reader seeing a non NULL rx_handler in a rcu_read_lock()
+	 * section has a guarantee to see a non NULL rx_handler_data
+	 * as well.
+	 */
+	synchronize_net();
 	RCU_INIT_POINTER(dev->rx_handler_data, NULL);
 }
 EXPORT_SYMBOL_GPL(netdev_rx_handler_unregister);
Signature · 1 line
1.7.10.4
author maintainer
#185 lore raw

PATCH 168/171 net: fq_codel: Fix off-by-one error

In reply to Steven Rostedt

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

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

From: Vijay Subramanian <[email protected]>

[ Upstream commit cd68ddd4c29ab523440299f24ff2417fe7a0dca6 ]

Currently, we hold a max of sch->limit -1 number of packets instead of sch->limit packets. Fix this off-by-one error.

---
 net/sched/sch_fq_codel.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
net/sched/sch_fq_codel.c +1 -1
index 4e606fc..5578628 100644
@@ -195,7 +195,7 @@ static int fq_codel_enqueue(struct sk_buff *skb, struct Qdisc *sch)
 		flow->deficit = q->quantum;
 		flow->dropped = 0;
 	}
-	if (++sch->q.qlen < sch->limit)
+	if (++sch->q.qlen <= sch->limit)
 		return NET_XMIT_SUCCESS;
 
 	q->drop_overlimit++;
Signature · 1 line
1.7.10.4
author maintainer
#186 lore raw

PATCH 169/171 pch_gbe: fix ip_summed checksum reporting on rx

In reply to Steven Rostedt

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

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

From: Veaceslav Falico <[email protected]>

[ Upstream commit 76a0e68129d7d24eb995a6871ab47081bbfa0acc ]

skb->ip_summed should be CHECKSUM_UNNECESSARY when the driver reports that checksums were correct and CHECKSUM_NONE in any other case. They're currently placed vice versa, which breaks the forwarding scenario. Fix it by placing them as described above.

---
 drivers/net/ethernet/oki-semi/pch_gbe/pch_gbe_main.c |    4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)
drivers/net/ethernet/oki-semi/pch_gbe/pch_gbe_main.c +2 -2
index feb85d5..bfc2d64 100644
@@ -1777,9 +1777,9 @@ pch_gbe_clean_rx(struct pch_gbe_adapter *adapter,
 
 			skb->protocol = eth_type_trans(skb, netdev);
 			if (tcp_ip_status & PCH_GBE_RXD_ACC_STAT_TCPIPOK)
-				skb->ip_summed = CHECKSUM_NONE;
-			else
 				skb->ip_summed = CHECKSUM_UNNECESSARY;
+			else
+				skb->ip_summed = CHECKSUM_NONE;
 
 			napi_gro_receive(&adapter->napi, skb);
 			(*work_done)++;
Signature · 1 line
1.7.10.4
author maintainer
#187 lore raw

PATCH 170/171 smsc75xx: fix jumbo frame support

In reply to Steven Rostedt

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

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

From: Steve Glendinning <[email protected]>

[ Upstream commit 4c51e53689569398d656e631c17308d9b8e84650 ]

This patch enables RX of jumbo frames for LAN7500.

Previously the driver would transmit jumbo frames succesfully but would drop received jumbo frames (incrementing the interface errors count).

With this patch applied the device can succesfully receive jumbo frames up to MTU 9000 (9014 bytes on the wire including ethernet header).

---
 drivers/net/usb/smsc75xx.c |   12 ++++++++----
 1 file changed, 8 insertions(+), 4 deletions(-)
drivers/net/usb/smsc75xx.c +8 -4
index 376143e..0d9a895 100644
@@ -724,8 +724,12 @@ static int smsc75xx_set_rx_max_frame_length(struct usbnet *dev, int size)
 static int smsc75xx_change_mtu(struct net_device *netdev, int new_mtu)
 {
 	struct usbnet *dev = netdev_priv(netdev);
+	int ret;
+
+	if (new_mtu > MAX_SINGLE_PACKET_SIZE)
+		return -EINVAL;
 
-	int ret = smsc75xx_set_rx_max_frame_length(dev, new_mtu);
+	ret = smsc75xx_set_rx_max_frame_length(dev, new_mtu + ETH_HLEN);
 	check_warn_return(ret, "Failed to set mac rx frame length");
 
 	return usbnet_change_mtu(netdev, new_mtu);
@@ -983,7 +987,7 @@ static int smsc75xx_reset(struct usbnet *dev)
 
 	netif_dbg(dev, ifup, dev->net, "FCT_TX_CTL set to 0x%08x", buf);
 
-	ret = smsc75xx_set_rx_max_frame_length(dev, 1514);
+	ret = smsc75xx_set_rx_max_frame_length(dev, dev->net->mtu + ETH_HLEN);
 	check_warn_return(ret, "Failed to set max rx frame length");
 
 	ret = smsc75xx_read_reg(dev, MAC_RX, &buf);
@@ -1127,8 +1131,8 @@ static int smsc75xx_rx_fixup(struct usbnet *dev, struct sk_buff *skb)
 			else if (rx_cmd_a & (RX_CMD_A_LONG | RX_CMD_A_RUNT))
 				dev->net->stats.rx_frame_errors++;
 		} else {
-			/* ETH_FRAME_LEN + 4(CRC) + 2(COE) + 4(Vlan) */
-			if (unlikely(size > (ETH_FRAME_LEN + 12))) {
+			/* MAX_SINGLE_PACKET_SIZE + 4(CRC) + 2(COE) + 4(Vlan) */
+			if (unlikely(size > (MAX_SINGLE_PACKET_SIZE + ETH_HLEN + 12))) {
 				netif_dbg(dev, rx_err, dev->net,
 					"size err rx_cmd_a=0x%08x", rx_cmd_a);
 				return 0;
Signature · 1 line
1.7.10.4
author maintainer
#188 lore raw

PATCH 171/171 bonding: get netdev_rx_handler_unregister out of locks

In reply to Steven Rostedt

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

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

From: Veaceslav Falico <[email protected]>

[ Upstream commit fcd99434fb5c137274d2e15dd2a6a7455f0f29ff ]

Now that netdev_rx_handler_unregister contains synchronize_net(), we need to call it outside of bond->lock, cause it might sleep. Also, remove the already unneded synchronize_net().

---
 drivers/net/bonding/bond_main.c |    3 +--
 1 file changed, 1 insertion(+), 2 deletions(-)
drivers/net/bonding/bond_main.c +1 -2
index a9a07ce..83ae683 100644
@@ -1961,12 +1961,11 @@ int bond_release(struct net_device *bond_dev, struct net_device *slave_dev)
 		return -EINVAL;
 	}
 
+	write_unlock_bh(&bond->lock);
 	/* unregister rx_handler early so bond_handle_frame wouldn't be called
 	 * for this slave anymore.
 	 */
 	netdev_rx_handler_unregister(slave_dev);
-	write_unlock_bh(&bond->lock);
-	synchronize_net();
 	write_lock_bh(&bond->lock);
 
 	if (!bond->params.fail_over_mac) {
Signature · 1 line
1.7.10.4