[net-next RFC 0/8] multiqueue API support for macvtap

[net-next RFC 0/8] multiqueue API support for macvtap

[Jason Wang]

Hi all:

This series implements a fully tuntap compatiable API which could be used by
userspace to manage multi macvtap queues. The main parts is to add TUNSETQUEUE
ioctl support for macvtap.

Patch 1 - 4 was some tuntap compatibility and misc cleanups.
Patch 5 removes the linear search in macvtap by reshuffling the macvtaps array
each time a queue is removed. After this, we could store both enabled and
disabled queues in the same array without introducing extra data structure.
Patch 6 let TUNSETIFF can create multiqueue device, nothing but some check were
added.
Patch 7 implement TUNSETQUEUE ioctl
Patch 8 reports IFF_MULTI_QUEUE to userspace to notify the userspace that the
multiqueue API is completed.

Flow caches implememtation were missed in this version, since I am doing rework
on the tuntap flow caches. Have some some stress test with both netperf and
pktgen.

Please review, thanks.

Jason Wang (8):
  macvlan: switch to use IS_ENABLED()
  macvtap: return -EBADFD when TUNGETIFF fails
  macvtap: introduce macvtap_get_vlan()
  macvlan: reduce the max number of taps to 8
  macvtap: eliminate linear search
  macvtap: allow TUNSETIFF to create multiqueue device
  macvtap: add TUNSETQUEUE ioctl
  macvtap: enable multiqueue flag

 drivers/net/macvtap.c      |  265 ++++++++++++++++++++++++++++++-------------
 include/linux/if_macvlan.h |   11 ++-
 2 files changed, 194 insertions(+), 82 deletions(-)

[net-next RFC 1/8] macvlan: switch to use IS_ENABLED()

[Jason Wang]

Signed-off-by: Jason Wang <jasowang@redhat.com>
---
 include/linux/if_macvlan.h |    2 +-
 1 files changed, 1 insertions(+), 1 deletions(-)

diff --git a/include/linux/if_macvlan.h b/include/linux/if_macvlan.h
index 84dde1d..e47ad46 100644
--- a/include/linux/if_macvlan.h
+++ b/include/linux/if_macvlan.h
@@ -8,7 +8,7 @@
 #include <net/netlink.h>
 #include <linux/u64_stats_sync.h>
 
-#if defined(CONFIG_MACVTAP) || defined(CONFIG_MACVTAP_MODULE)
+#if IS_ENABLED(CONFIG_MACVTAP)
 struct socket *macvtap_get_socket(struct file *);
 #else
 #include <linux/err.h>
-- 
1.7.1

[net-next RFC 2/8] macvtap: return -EBADFD when TUNGETIFF fails

[Jason Wang]

Tuntap return -EBADFD when TUNGETIFF fails, we should return the same value.

Signed-off-by: Jason Wang <jasowang@redhat.com>
---
 drivers/net/macvtap.c |    2 +-
 1 files changed, 1 insertions(+), 1 deletions(-)

diff --git a/drivers/net/macvtap.c b/drivers/net/macvtap.c
index 59e9605..ce1c72a 100644
--- a/drivers/net/macvtap.c
+++ b/drivers/net/macvtap.c
@@ -928,7 +928,7 @@ static long macvtap_ioctl(struct file *file, unsigned int cmd,
 		rcu_read_unlock_bh();
 
 		if (!vlan)
-			return -ENOLINK;
+			return -EBADFD;
 
 		ret = 0;
 		if (copy_to_user(&ifr->ifr_name, vlan->dev->name, IFNAMSIZ) ||
-- 
1.7.1

[net-next RFC 3/8] macvtap: introduce macvtap_get_vlan()

[Jason Wang]

Factor out the device holding logic to a macvtap_get_vlan(), this will be also
used by multiqueue API.

Signed-off-by: Jason Wang <jasowang@redhat.com>
---
 drivers/net/macvtap.c |   26 +++++++++++++++++++-------
 1 files changed, 19 insertions(+), 7 deletions(-)

diff --git a/drivers/net/macvtap.c b/drivers/net/macvtap.c
index ce1c72a..a36e49e 100644
--- a/drivers/net/macvtap.c
+++ b/drivers/net/macvtap.c
@@ -890,6 +890,23 @@ out:
 	return ret;
 }
 
+static struct macvlan_dev *macvtap_get_vlan(struct macvtap_queue *q)
+{
+	struct macvlan_dev *vlan;
+	rcu_read_lock_bh();
+	vlan = rcu_dereference_bh(q->vlan);
+	if (vlan)
+		dev_hold(vlan->dev);
+	rcu_read_unlock_bh();
+
+	return vlan;
+}
+
+static void macvtap_put_vlan(struct macvlan_dev *vlan)
+{
+	dev_put(vlan->dev);
+}
+
 /*
  * provide compatibility with generic tun/tap interface
  */
@@ -921,12 +938,7 @@ static long macvtap_ioctl(struct file *file, unsigned int cmd,
 		return ret;
 
 	case TUNGETIFF:
-		rcu_read_lock_bh();
-		vlan = rcu_dereference_bh(q->vlan);
-		if (vlan)
-			dev_hold(vlan->dev);
-		rcu_read_unlock_bh();
-
+		vlan = macvtap_get_vlan(q);
 		if (!vlan)
 			return -EBADFD;
 
@@ -934,7 +946,7 @@ static long macvtap_ioctl(struct file *file, unsigned int cmd,
 		if (copy_to_user(&ifr->ifr_name, vlan->dev->name, IFNAMSIZ) ||
 		    put_user(q->flags, &ifr->ifr_flags))
 			ret = -EFAULT;
-		dev_put(vlan->dev);
+		macvtap_put_vlan(vlan);
 		return ret;
 
 	case TUNGETFEATURES:
-- 
1.7.1

[net-next RFC 4/8] macvlan: reduce the max number of taps to 8

[Jason Wang]

To be same with tap.

Signed-off-by: Jason Wang <jasowang@redhat.com>
---
 include/linux/if_macvlan.h |    2 +-
 1 files changed, 1 insertions(+), 1 deletions(-)

diff --git a/include/linux/if_macvlan.h b/include/linux/if_macvlan.h
index e47ad46..32e943a 100644
--- a/include/linux/if_macvlan.h
+++ b/include/linux/if_macvlan.h
@@ -50,7 +50,7 @@ struct macvlan_pcpu_stats {
  * Maximum times a macvtap device can be opened. This can be used to
  * configure the number of receive queue, e.g. for multiqueue virtio.
  */
-#define MAX_MACVTAP_QUEUES	(NR_CPUS < 16 ? NR_CPUS : 16)
+#define MAX_MACVTAP_QUEUES	8
 
 #define MACVLAN_MC_FILTER_BITS	8
 #define MACVLAN_MC_FILTER_SZ	(1 << MACVLAN_MC_FILTER_BITS)
-- 
1.7.1

[net-next RFC 5/8] macvtap: eliminate linear search

[Jason Wang]

Linear search were used in both get_slot() and macvtap_get_queue(), this is
because:

- macvtap didn't reshuffle the array of taps when create or destroy a queue, so
  when adding a new queue, macvtap must do linear search to find a location for
  the new queue. This will also complicate the TUNSETQUEUE implementation for
  multiqueue API.
- the queue itself didn't track the queue index, so the we must do a linear
  search in the array to find the location of a existed queue.

The solution is straightforward: reshuffle the array and introduce a queue_index
to macvtap_queue.

Signed-off-by: Jason Wang <jasowang@redhat.com>
---
 drivers/net/macvtap.c |   63 +++++++++++++++---------------------------------
 1 files changed, 20 insertions(+), 43 deletions(-)

diff --git a/drivers/net/macvtap.c b/drivers/net/macvtap.c
index a36e49e..5fd341c 100644
--- a/drivers/net/macvtap.c
+++ b/drivers/net/macvtap.c
@@ -44,6 +44,7 @@ struct macvtap_queue {
 	struct macvlan_dev __rcu *vlan;
 	struct file *file;
 	unsigned int flags;
+	u16 queue_index;
 };
 
 static struct proto macvtap_proto = {
@@ -84,30 +85,10 @@ static const struct proto_ops macvtap_socket_ops;
  */
 static DEFINE_SPINLOCK(macvtap_lock);
 
-/*
- * get_slot: return a [unused/occupied] slot in vlan->taps[]:
- *	- if 'q' is NULL, return the first empty slot;
- *	- otherwise, return the slot this pointer occupies.
- */
-static int get_slot(struct macvlan_dev *vlan, struct macvtap_queue *q)
-{
-	int i;
-
-	for (i = 0; i < MAX_MACVTAP_QUEUES; i++) {
-		if (rcu_dereference_protected(vlan->taps[i],
-					      lockdep_is_held(&macvtap_lock)) == q)
-			return i;
-	}
-
-	/* Should never happen */
-	BUG_ON(1);
-}
-
 static int macvtap_set_queue(struct net_device *dev, struct file *file,
 				struct macvtap_queue *q)
 {
 	struct macvlan_dev *vlan = netdev_priv(dev);
-	int index;
 	int err = -EBUSY;
 
 	spin_lock(&macvtap_lock);
@@ -115,12 +96,12 @@ static int macvtap_set_queue(struct net_device *dev, struct file *file,
 		goto out;
 
 	err = 0;
-	index = get_slot(vlan, NULL);
 	rcu_assign_pointer(q->vlan, vlan);
-	rcu_assign_pointer(vlan->taps[index], q);
+	rcu_assign_pointer(vlan->taps[vlan->numvtaps], q);
 	sock_hold(&q->sk);
 
 	q->file = file;
+	q->queue_index = vlan->numvtaps;
 	file->private_data = q;
 
 	vlan->numvtaps++;
@@ -140,16 +121,23 @@ out:
  */
 static void macvtap_put_queue(struct macvtap_queue *q)
 {
+	struct macvtap_queue *nq;
 	struct macvlan_dev *vlan;
 
 	spin_lock(&macvtap_lock);
 	vlan = rcu_dereference_protected(q->vlan,
 					 lockdep_is_held(&macvtap_lock));
 	if (vlan) {
-		int index = get_slot(vlan, q);
+		int index = q->queue_index;
+		BUG_ON(index >= vlan->numvtaps);
 
-		RCU_INIT_POINTER(vlan->taps[index], NULL);
+		rcu_assign_pointer(vlan->taps[index],
+				   vlan->taps[vlan->numvtaps - 1]);
 		RCU_INIT_POINTER(q->vlan, NULL);
+		nq = rcu_dereference_protected(vlan->taps[index],
+					       lockdep_is_held(&macvtap_lock));
+		nq->queue_index = index;
+
 		sock_put(&q->sk);
 		--vlan->numvtaps;
 	}
@@ -182,8 +170,7 @@ static struct macvtap_queue *macvtap_get_queue(struct net_device *dev,
 	rxq = skb_get_rxhash(skb);
 	if (rxq) {
 		tap = rcu_dereference(vlan->taps[rxq % numvtaps]);
-		if (tap)
-			goto out;
+		goto out;
 	}
 
 	if (likely(skb_rx_queue_recorded(skb))) {
@@ -193,17 +180,10 @@ static struct macvtap_queue *macvtap_get_queue(struct net_device *dev,
 			rxq -= numvtaps;
 
 		tap = rcu_dereference(vlan->taps[rxq]);
-		if (tap)
-			goto out;
-	}
-
-	/* Everything failed - find first available queue */
-	for (rxq = 0; rxq < MAX_MACVTAP_QUEUES; rxq++) {
-		tap = rcu_dereference(vlan->taps[rxq]);
-		if (tap)
-			break;
+		goto out;
 	}
 
+	tap = rcu_dereference(vlan->taps[0]);
 out:
 	return tap;
 }
@@ -221,17 +201,14 @@ static void macvtap_del_queues(struct net_device *dev)
 
 	/* macvtap_put_queue can free some slots, so go through all slots */
 	spin_lock(&macvtap_lock);
-	for (i = 0; i < MAX_MACVTAP_QUEUES && vlan->numvtaps; i++) {
+	for (i = 0; i < vlan->numvtaps; i++) {
 		q = rcu_dereference_protected(vlan->taps[i],
 					      lockdep_is_held(&macvtap_lock));
-		if (q) {
-			qlist[j++] = q;
-			RCU_INIT_POINTER(vlan->taps[i], NULL);
-			RCU_INIT_POINTER(q->vlan, NULL);
-			vlan->numvtaps--;
-		}
+		BUG_ON(q == NULL);
+		qlist[j++] = q;
+		RCU_INIT_POINTER(vlan->taps[i], NULL);
+		RCU_INIT_POINTER(q->vlan, NULL);
 	}
-	BUG_ON(vlan->numvtaps != 0);
 	/* guarantee that any future macvtap_set_queue will fail */
 	vlan->numvtaps = MAX_MACVTAP_QUEUES;
 	spin_unlock(&macvtap_lock);
-- 
1.7.1

[net-next RFC 6/8] macvtap: allow TUNSETIFF to create multiqueue device

[Jason Wang]

Though the queue were in fact created by open(), we still need to add this check
to be compatible with tuntap which can let mgmt software use a single API to
manage queues. This patch only validates the device name and moves the TUNSETIFF
to a helper.

Signed-off-by: Jason Wang <jasowang@redhat.com>
---
 drivers/net/macvtap.c |   50 ++++++++++++++++++++++++++++++++++++++----------
 1 files changed, 39 insertions(+), 11 deletions(-)

diff --git a/drivers/net/macvtap.c b/drivers/net/macvtap.c
index 5fd341c..e3f9344 100644
--- a/drivers/net/macvtap.c
+++ b/drivers/net/macvtap.c
@@ -867,6 +867,7 @@ out:
 	return ret;
 }
 
+
 static struct macvlan_dev *macvtap_get_vlan(struct macvtap_queue *q)
 {
 	struct macvlan_dev *vlan;
@@ -884,6 +885,43 @@ static void macvtap_put_vlan(struct macvlan_dev *vlan)
 	dev_put(vlan->dev);
 }
 
+static int macvtap_set_iff(struct file *file, struct ifreq __user *ifr_u)
+{
+	struct macvtap_queue *q = file->private_data;
+	struct net *net = current->nsproxy->net_ns;
+	struct inode *inode = file_inode(file);
+	struct net_device *dev, *dev2;
+	struct ifreq ifr;
+
+	if (copy_from_user(&ifr, ifr_u, sizeof(struct ifreq)))
+		return -EFAULT;
+
+	if (ifr.ifr_flags & IFF_MULTI_QUEUE) {
+		dev = __dev_get_by_name(net, ifr.ifr_name);
+		if (!dev)
+			return -EINVAL;
+
+		dev2 = dev_get_by_macvtap_minor(iminor(inode));
+		if (!dev2)
+			return -EINVAL;
+
+		if (dev != dev2) {
+			dev_put(dev2);
+			return -EINVAL;
+		}
+
+		dev_put(dev2);
+	}
+
+	if ((ifr.ifr_flags & ~(IFF_VNET_HDR | IFF_MULTI_QUEUE)) !=
+	    (IFF_NO_PI | IFF_TAP))
+		return -EINVAL;
+	else
+		q->flags = ifr.ifr_flags;
+
+	return 0;
+}
+
 /*
  * provide compatibility with generic tun/tap interface
  */
@@ -902,17 +940,7 @@ static long macvtap_ioctl(struct file *file, unsigned int cmd,
 
 	switch (cmd) {
 	case TUNSETIFF:
-		/* ignore the name, just look at flags */
-		if (get_user(u, &ifr->ifr_flags))
-			return -EFAULT;
-
-		ret = 0;
-		if ((u & ~IFF_VNET_HDR) != (IFF_NO_PI | IFF_TAP))
-			ret = -EINVAL;
-		else
-			q->flags = u;
-
-		return ret;
+		return macvtap_set_iff(file, ifr);
 
 	case TUNGETIFF:
 		vlan = macvtap_get_vlan(q);
-- 
1.7.1

[net-next RFC 7/8] macvtap: add TUNSETQUEUE ioctl

[Jason Wang]

This patch adds TUNSETQUEUE ioctl to let userspace can temporarily disable or
enable a queue of macvtap. This is used to be compatible at API layer of tuntap
to simplify the userspace to manage the queues.

This is done by just introduce a pointer detached which was set when the queue
were detached. Only the queue with NULL set to this pointer can be used to
forward packets.

Signed-off-by: Jason Wang <jasowang@redhat.com>
---
 drivers/net/macvtap.c      |  147 +++++++++++++++++++++++++++++++++++--------
 include/linux/if_macvlan.h |    7 ++
 2 files changed, 126 insertions(+), 28 deletions(-)

diff --git a/drivers/net/macvtap.c b/drivers/net/macvtap.c
index e3f9344..06b10a5 100644
--- a/drivers/net/macvtap.c
+++ b/drivers/net/macvtap.c
@@ -85,67 +85,131 @@ static const struct proto_ops macvtap_socket_ops;
  */
 static DEFINE_SPINLOCK(macvtap_lock);
 
+static void macvtap_swap_slot(struct macvlan_dev *vlan, int a, int b)
+{
+	struct macvtap_queue *q1, *q2;
+
+	if (a == b)
+		return;
+
+	q1 = rcu_dereference_protected(vlan->taps[a],
+				       lockdep_is_held(&macvtap_lock));
+	q2 = rcu_dereference_protected(vlan->taps[b],
+				       lockdep_is_held(&macvtap_lock));
+
+	BUG_ON(q1 == NULL || q2 == NULL);
+
+	rcu_assign_pointer(vlan->taps[a], vlan->taps[b]);
+	rcu_assign_pointer(vlan->taps[b], q1);
+
+	q1->queue_index = b;
+	q2->queue_index = a;
+}
+
 static int macvtap_set_queue(struct net_device *dev, struct file *file,
-				struct macvtap_queue *q)
+			     struct macvtap_queue *q, bool enable)
 {
 	struct macvlan_dev *vlan = netdev_priv(dev);
 	int err = -EBUSY;
+	int total;
 
 	spin_lock(&macvtap_lock);
-	if (vlan->numvtaps == MAX_MACVTAP_QUEUES)
+
+	total = vlan->numvtaps + vlan->numdisabled;
+	if (!enable && total == MAX_MACVTAP_QUEUES)
+		goto out;
+
+	err = -EINVAL;
+	if (enable && q->queue_index < vlan->numvtaps)
 		goto out;
 
 	err = 0;
-	rcu_assign_pointer(q->vlan, vlan);
-	rcu_assign_pointer(vlan->taps[vlan->numvtaps], q);
-	sock_hold(&q->sk);
 
-	q->file = file;
-	q->queue_index = vlan->numvtaps;
-	file->private_data = q;
+	if (enable) {
+		BUG_ON(q->queue_index >= total);
+		macvtap_swap_slot(vlan, q->queue_index, vlan->numvtaps);
+		--vlan->numdisabled;
+	} else {
+		rcu_assign_pointer(q->vlan, vlan);
+		rcu_assign_pointer(vlan->taps[total], q);
+		sock_hold(&q->sk);
+
+		q->file = file;
+		q->queue_index = total;
+		file->private_data = q;
+		if (vlan->numdisabled)
+			macvtap_swap_slot(vlan, vlan->numvtaps, total);
+
+		/* Make sure the pointers were seen before numvtaps. */
+		smp_wmb();
+	}
 
 	vlan->numvtaps++;
-
 out:
 	spin_unlock(&macvtap_lock);
 	return err;
 }
 
 /*
- * The file owning the queue got closed, give up both
- * the reference that the files holds as well as the
- * one from the macvlan_dev if that still exists.
+ * The file owning the queue got closed or disabled.
+ *
+ * If closed give up both the reference that the files holds as well as the
+ * one from the macvlan_dev if that still exists. If disabled, reshuffle the
+ * array of taps.
  *
  * Using the spinlock makes sure that we don't get
  * to the queue again after destroying it.
  */
-static void macvtap_put_queue(struct macvtap_queue *q)
+static int macvtap_put_queue(struct macvtap_queue *q, bool clean)
 {
-	struct macvtap_queue *nq;
 	struct macvlan_dev *vlan;
+	int err = -EINVAL;
 
 	spin_lock(&macvtap_lock);
 	vlan = rcu_dereference_protected(q->vlan,
 					 lockdep_is_held(&macvtap_lock));
+
 	if (vlan) {
+		int total = vlan->numvtaps + vlan->numdisabled;
 		int index = q->queue_index;
-		BUG_ON(index >= vlan->numvtaps);
+		bool disabled = q->queue_index >= vlan->numvtaps;
 
-		rcu_assign_pointer(vlan->taps[index],
-				   vlan->taps[vlan->numvtaps - 1]);
-		RCU_INIT_POINTER(q->vlan, NULL);
-		nq = rcu_dereference_protected(vlan->taps[index],
-					       lockdep_is_held(&macvtap_lock));
-		nq->queue_index = index;
+		if (!clean) {
+			BUG_ON(q->queue_index >= total);
+			if (q->queue_index >= vlan->numvtaps)
+				goto out;
+		}
 
-		sock_put(&q->sk);
-		--vlan->numvtaps;
-	}
+		err = 0;
+		macvtap_swap_slot(vlan, index, total - 1);
+		if (!disabled && vlan->numdisabled)
+			macvtap_swap_slot(vlan, index, vlan->numvtaps - 1);
+
+		if (clean) {
+			RCU_INIT_POINTER(vlan->taps[total - 1], NULL);
+			RCU_INIT_POINTER(q->vlan, NULL);
+			sock_put(&q->sk);
+			if (disabled)
+				--vlan->numdisabled;
+			else
+				--vlan->numvtaps;
+		} else {
+			--vlan->numvtaps;
+			++vlan->numdisabled;
+		}
+
+	} else if (clean)
+		err = 0;
 
+out:
 	spin_unlock(&macvtap_lock);
 
 	synchronize_rcu();
-	sock_put(&q->sk);
+
+	if (clean)
+		sock_put(&q->sk);
+
+	return err;
 }
 
 /*
@@ -201,7 +265,7 @@ static void macvtap_del_queues(struct net_device *dev)
 
 	/* macvtap_put_queue can free some slots, so go through all slots */
 	spin_lock(&macvtap_lock);
-	for (i = 0; i < vlan->numvtaps; i++) {
+	for (i = 0; i < vlan->numvtaps + vlan->numdisabled; i++) {
 		q = rcu_dereference_protected(vlan->taps[i],
 					      lockdep_is_held(&macvtap_lock));
 		BUG_ON(q == NULL);
@@ -211,6 +275,7 @@ static void macvtap_del_queues(struct net_device *dev)
 	}
 	/* guarantee that any future macvtap_set_queue will fail */
 	vlan->numvtaps = MAX_MACVTAP_QUEUES;
+	vlan->numdisabled = 0;
 	spin_unlock(&macvtap_lock);
 
 	synchronize_rcu();
@@ -384,7 +449,7 @@ static int macvtap_open(struct inode *inode, struct file *file)
 	if ((dev->features & NETIF_F_HIGHDMA) && (dev->features & NETIF_F_SG))
 		sock_set_flag(&q->sk, SOCK_ZEROCOPY);
 
-	err = macvtap_set_queue(dev, file, q);
+	err = macvtap_set_queue(dev, file, q, false);
 	if (err)
 		sock_put(&q->sk);
 
@@ -398,7 +463,7 @@ out:
 static int macvtap_release(struct inode *inode, struct file *file)
 {
 	struct macvtap_queue *q = file->private_data;
-	macvtap_put_queue(q);
+	macvtap_put_queue(q, true);
 	return 0;
 }
 
@@ -922,6 +987,27 @@ static int macvtap_set_iff(struct file *file, struct ifreq __user *ifr_u)
 	return 0;
 }
 
+static int macvtap_ioctl_set_queue(struct file *file, unsigned int flags)
+{
+	struct macvtap_queue *q = file->private_data;
+	struct macvlan_dev *vlan;
+	int ret = -EINVAL;
+
+	vlan = macvtap_get_vlan(q);
+	if (!vlan)
+		goto done;
+
+	if (flags & IFF_ATTACH_QUEUE)
+		ret = macvtap_set_queue(vlan->dev, file, q, true);
+	else if (flags & IFF_DETACH_QUEUE)
+		ret = macvtap_put_queue(q, false);
+
+	macvtap_put_vlan(vlan);
+
+done:
+	return ret;
+}
+
 /*
  * provide compatibility with generic tun/tap interface
  */
@@ -954,6 +1040,11 @@ static long macvtap_ioctl(struct file *file, unsigned int cmd,
 		macvtap_put_vlan(vlan);
 		return ret;
 
+	case TUNSETQUEUE:
+		if (get_user(u, &ifr->ifr_flags))
+			return -EFAULT;
+		return macvtap_ioctl_set_queue(file, u);
+
 	case TUNGETFEATURES:
 		if (put_user(IFF_TAP | IFF_NO_PI | IFF_VNET_HDR, up))
 			return -EFAULT;
diff --git a/include/linux/if_macvlan.h b/include/linux/if_macvlan.h
index 32e943a..f134afb 100644
--- a/include/linux/if_macvlan.h
+++ b/include/linux/if_macvlan.h
@@ -69,8 +69,15 @@ struct macvlan_dev {
 	u16			flags;
 	int (*receive)(struct sk_buff *skb);
 	int (*forward)(struct net_device *dev, struct sk_buff *skb);
+	/* This array tracks all taps (include disabled ones) and will be
+	 * reshuffled to keep the following order:
+	 * [0, numvtaps) : enabled taps,
+	 * [numvtaps, numvtaps + numdisabled) : disabled taps,
+	 * [numvtaps + numdisabled, MAX_MACVTAP_QUEUES) : unused slots
+	 */
 	struct macvtap_queue	*taps[MAX_MACVTAP_QUEUES];
 	int			numvtaps;
+	int			numdisabled;
 	int			minor;
 };
 
-- 
1.7.1

[net-next RFC 8/8] macvtap: enable multiqueue flag

[Jason Wang]

Signed-off-by: Jason Wang <jasowang@redhat.com>
---
 drivers/net/macvtap.c |    7 ++-----
 1 files changed, 2 insertions(+), 5 deletions(-)

diff --git a/drivers/net/macvtap.c b/drivers/net/macvtap.c
index 06b10a5..6a4af74 100644
--- a/drivers/net/macvtap.c
+++ b/drivers/net/macvtap.c
@@ -31,10 +31,6 @@
  * macvtap_proto is used to allocate queues through the sock allocation
  * mechanism.
  *
- * TODO: multiqueue support is currently not implemented, even though
- * macvtap is basically prepared for that. We will need to add this
- * here as well as in virtio-net and qemu to get line rate on 10gbit
- * adapters from a guest.
  */
 struct macvtap_queue {
 	struct sock sk;
@@ -1046,7 +1042,8 @@ static long macvtap_ioctl(struct file *file, unsigned int cmd,
 		return macvtap_ioctl_set_queue(file, u);
 
 	case TUNGETFEATURES:
-		if (put_user(IFF_TAP | IFF_NO_PI | IFF_VNET_HDR, up))
+		if (put_user(IFF_TAP | IFF_NO_PI | IFF_VNET_HDR |
+			     IFF_MULTI_QUEUE, up))
 			return -EFAULT;
 		return 0;
 
-- 
1.7.1

Re: [net-next RFC 4/8] macvlan: reduce the max number of taps to 8

[Michael S. Tsirkin]

On Thu, May 23, 2013 at 11:12:29AM +0800, Jason Wang wrote:
> To be same with tap.
> 
> Signed-off-by: Jason Wang <jasowang@redhat.com>

Well for tap the very specific reason was that
there's an array of big queue structures,
so we need to limit it to make it fit in a page.
No such reason here right?

We need at least as much as tap to be compatible, so
let's just make it 16 unconditionally?

> ---
>  include/linux/if_macvlan.h |    2 +-
>  1 files changed, 1 insertions(+), 1 deletions(-)
> 
> diff --git a/include/linux/if_macvlan.h b/include/linux/if_macvlan.h
> index e47ad46..32e943a 100644
> --- a/include/linux/if_macvlan.h
> +++ b/include/linux/if_macvlan.h
> @@ -50,7 +50,7 @@ struct macvlan_pcpu_stats {
>   * Maximum times a macvtap device can be opened. This can be used to
>   * configure the number of receive queue, e.g. for multiqueue virtio.
>   */
> -#define MAX_MACVTAP_QUEUES	(NR_CPUS < 16 ? NR_CPUS : 16)
> +#define MAX_MACVTAP_QUEUES	8
>  
>  #define MACVLAN_MC_FILTER_BITS	8
>  #define MACVLAN_MC_FILTER_SZ	(1 << MACVLAN_MC_FILTER_BITS)
> -- 
> 1.7.1

Re: [net-next RFC 1/8] macvlan: switch to use IS_ENABLED()

[Michael S. Tsirkin]

On Thu, May 23, 2013 at 11:12:26AM +0800, Jason Wang wrote:
> Signed-off-by: Jason Wang <jasowang@redhat.com>

Acked-by: Michael S. Tsirkin <mst@redhat.com>

> ---
>  include/linux/if_macvlan.h |    2 +-
>  1 files changed, 1 insertions(+), 1 deletions(-)
> 
> diff --git a/include/linux/if_macvlan.h b/include/linux/if_macvlan.h
> index 84dde1d..e47ad46 100644
> --- a/include/linux/if_macvlan.h
> +++ b/include/linux/if_macvlan.h
> @@ -8,7 +8,7 @@
>  #include <net/netlink.h>
>  #include <linux/u64_stats_sync.h>
>  
> -#if defined(CONFIG_MACVTAP) || defined(CONFIG_MACVTAP_MODULE)
> +#if IS_ENABLED(CONFIG_MACVTAP)
>  struct socket *macvtap_get_socket(struct file *);
>  #else
>  #include <linux/err.h>
> -- 
> 1.7.1

Re: [net-next RFC 5/8] macvtap: eliminate linear search

[Michael S. Tsirkin]

On Thu, May 23, 2013 at 11:12:30AM +0800, Jason Wang wrote:
> Linear search were used in both get_slot() and macvtap_get_queue(), this is
> because:
> 
> - macvtap didn't reshuffle the array of taps when create or destroy a queue, so
>   when adding a new queue, macvtap must do linear search to find a location for
>   the new queue. This will also complicate the TUNSETQUEUE implementation for
>   multiqueue API.
> - the queue itself didn't track the queue index, so the we must do a linear
>   search in the array to find the location of a existed queue.
> 
> The solution is straightforward: reshuffle the array and introduce a queue_index
> to macvtap_queue.
> 
> Signed-off-by: Jason Wang <jasowang@redhat.com>
> ---
>  drivers/net/macvtap.c |   63 +++++++++++++++---------------------------------
>  1 files changed, 20 insertions(+), 43 deletions(-)
> 
> diff --git a/drivers/net/macvtap.c b/drivers/net/macvtap.c
> index a36e49e..5fd341c 100644
> --- a/drivers/net/macvtap.c
> +++ b/drivers/net/macvtap.c
> @@ -44,6 +44,7 @@ struct macvtap_queue {
>  	struct macvlan_dev __rcu *vlan;
>  	struct file *file;
>  	unsigned int flags;
> +	u16 queue_index;
>  };
>  
>  static struct proto macvtap_proto = {
> @@ -84,30 +85,10 @@ static const struct proto_ops macvtap_socket_ops;
>   */
>  static DEFINE_SPINLOCK(macvtap_lock);
>  
> -/*
> - * get_slot: return a [unused/occupied] slot in vlan->taps[]:
> - *	- if 'q' is NULL, return the first empty slot;
> - *	- otherwise, return the slot this pointer occupies.
> - */
> -static int get_slot(struct macvlan_dev *vlan, struct macvtap_queue *q)
> -{
> -	int i;
> -
> -	for (i = 0; i < MAX_MACVTAP_QUEUES; i++) {
> -		if (rcu_dereference_protected(vlan->taps[i],
> -					      lockdep_is_held(&macvtap_lock)) == q)
> -			return i;
> -	}
> -
> -	/* Should never happen */
> -	BUG_ON(1);
> -}
> -
>  static int macvtap_set_queue(struct net_device *dev, struct file *file,
>  				struct macvtap_queue *q)
>  {
>  	struct macvlan_dev *vlan = netdev_priv(dev);
> -	int index;
>  	int err = -EBUSY;
>  
>  	spin_lock(&macvtap_lock);
> @@ -115,12 +96,12 @@ static int macvtap_set_queue(struct net_device *dev, struct file *file,
>  		goto out;
>  
>  	err = 0;
> -	index = get_slot(vlan, NULL);
>  	rcu_assign_pointer(q->vlan, vlan);
> -	rcu_assign_pointer(vlan->taps[index], q);
> +	rcu_assign_pointer(vlan->taps[vlan->numvtaps], q);
>  	sock_hold(&q->sk);
>  
>  	q->file = file;
> +	q->queue_index = vlan->numvtaps;
>  	file->private_data = q;
>  
>  	vlan->numvtaps++;
> @@ -140,16 +121,23 @@ out:
>   */
>  static void macvtap_put_queue(struct macvtap_queue *q)
>  {
> +	struct macvtap_queue *nq;
>  	struct macvlan_dev *vlan;
>  
>  	spin_lock(&macvtap_lock);
>  	vlan = rcu_dereference_protected(q->vlan,
>  					 lockdep_is_held(&macvtap_lock));
>  	if (vlan) {
> -		int index = get_slot(vlan, q);
> +		int index = q->queue_index;
> +		BUG_ON(index >= vlan->numvtaps);
>  
> -		RCU_INIT_POINTER(vlan->taps[index], NULL);
> +		rcu_assign_pointer(vlan->taps[index],
> +				   vlan->taps[vlan->numvtaps - 1]);
>  		RCU_INIT_POINTER(q->vlan, NULL);
> +		nq = rcu_dereference_protected(vlan->taps[index],
> +					       lockdep_is_held(&macvtap_lock));
> +		nq->queue_index = index;
> +
>  		sock_put(&q->sk);

Hmm this looks like a bug BTW: sock_put should be after
no one uses this tap, so must be after synchronize_rcu.
No?

>  		--vlan->numvtaps;
>  	}

So:
		nq = rcu_dereference_protected(vlan->taps[vlan->numvtaps - 1],
					       lockdep_is_held(&macvtap_lock));
		rcu_assign_pointer(vlan->taps[index], nq);

same thing?


> @@ -182,8 +170,7 @@ static struct macvtap_queue *macvtap_get_queue(struct net_device *dev,
>  	rxq = skb_get_rxhash(skb);
>  	if (rxq) {
>  		tap = rcu_dereference(vlan->taps[rxq % numvtaps]);
> -		if (tap)
> -			goto out;
> +		goto out;
>  	}
>  
>  	if (likely(skb_rx_queue_recorded(skb))) {
> @@ -193,17 +180,10 @@ static struct macvtap_queue *macvtap_get_queue(struct net_device *dev,
>  			rxq -= numvtaps;
>  
>  		tap = rcu_dereference(vlan->taps[rxq]);
> -		if (tap)
> -			goto out;
> -	}
> -
> -	/* Everything failed - find first available queue */
> -	for (rxq = 0; rxq < MAX_MACVTAP_QUEUES; rxq++) {
> -		tap = rcu_dereference(vlan->taps[rxq]);
> -		if (tap)
> -			break;
> +		goto out;
>  	}
>  
> +	tap = rcu_dereference(vlan->taps[0]);
>  out:
>  	return tap;
>  }
> @@ -221,17 +201,14 @@ static void macvtap_del_queues(struct net_device *dev)
>  
>  	/* macvtap_put_queue can free some slots, so go through all slots */
>  	spin_lock(&macvtap_lock);
> -	for (i = 0; i < MAX_MACVTAP_QUEUES && vlan->numvtaps; i++) {
> +	for (i = 0; i < vlan->numvtaps; i++) {
>  		q = rcu_dereference_protected(vlan->taps[i],
>  					      lockdep_is_held(&macvtap_lock));
> -		if (q) {
> -			qlist[j++] = q;
> -			RCU_INIT_POINTER(vlan->taps[i], NULL);
> -			RCU_INIT_POINTER(q->vlan, NULL);
> -			vlan->numvtaps--;
> -		}
> +		BUG_ON(q == NULL);
> +		qlist[j++] = q;
> +		RCU_INIT_POINTER(vlan->taps[i], NULL);
> +		RCU_INIT_POINTER(q->vlan, NULL);
>  	}
> -	BUG_ON(vlan->numvtaps != 0);
>  	/* guarantee that any future macvtap_set_queue will fail */
>  	vlan->numvtaps = MAX_MACVTAP_QUEUES;
>  	spin_unlock(&macvtap_lock);
> -- 
> 1.7.1

Re: [net-next RFC 6/8] macvtap: allow TUNSETIFF to create multiqueue device

[Michael S. Tsirkin]

On Thu, May 23, 2013 at 11:12:31AM +0800, Jason Wang wrote:
> Though the queue were in fact created by open(), we still need to add this check
> to be compatible with tuntap which can let mgmt software use a single API to
> manage queues. This patch only validates the device name and moves the TUNSETIFF
> to a helper.
> 
> Signed-off-by: Jason Wang <jasowang@redhat.com>
> ---
>  drivers/net/macvtap.c |   50 ++++++++++++++++++++++++++++++++++++++----------
>  1 files changed, 39 insertions(+), 11 deletions(-)
> 
> diff --git a/drivers/net/macvtap.c b/drivers/net/macvtap.c
> index 5fd341c..e3f9344 100644
> --- a/drivers/net/macvtap.c
> +++ b/drivers/net/macvtap.c
> @@ -867,6 +867,7 @@ out:
>  	return ret;
>  }
>  
> +

Please don't add empty lines like this :).

>  static struct macvlan_dev *macvtap_get_vlan(struct macvtap_queue *q)
>  {
>  	struct macvlan_dev *vlan;
> @@ -884,6 +885,43 @@ static void macvtap_put_vlan(struct macvlan_dev *vlan)
>  	dev_put(vlan->dev);
>  }
>  
> +static int macvtap_set_iff(struct file *file, struct ifreq __user *ifr_u)
> +{
> +	struct macvtap_queue *q = file->private_data;
> +	struct net *net = current->nsproxy->net_ns;
> +	struct inode *inode = file_inode(file);
> +	struct net_device *dev, *dev2;
> +	struct ifreq ifr;
> +
> +	if (copy_from_user(&ifr, ifr_u, sizeof(struct ifreq)))
> +		return -EFAULT;
> +
> +	if (ifr.ifr_flags & IFF_MULTI_QUEUE) {

So why do we validate the name?
Pls add a comment.

> +		dev = __dev_get_by_name(net, ifr.ifr_name);
> +		if (!dev)
> +			return -EINVAL;
> +
> +		dev2 = dev_get_by_macvtap_minor(iminor(inode));
> +		if (!dev2)
> +			return -EINVAL;
> +
> +		if (dev != dev2) {
> +			dev_put(dev2);
> +			return -EINVAL;
> +		}
> +
> +		dev_put(dev2);
> +	}
> +
> +	if ((ifr.ifr_flags & ~(IFF_VNET_HDR | IFF_MULTI_QUEUE)) !=
> +	    (IFF_NO_PI | IFF_TAP))
> +		return -EINVAL;
> +	else
> +		q->flags = ifr.ifr_flags;
> +
> +	return 0;
> +}
> +
>  /*
>   * provide compatibility with generic tun/tap interface
>   */
> @@ -902,17 +940,7 @@ static long macvtap_ioctl(struct file *file, unsigned int cmd,
>  
>  	switch (cmd) {
>  	case TUNSETIFF:
> -		/* ignore the name, just look at flags */
> -		if (get_user(u, &ifr->ifr_flags))
> -			return -EFAULT;
> -
> -		ret = 0;
> -		if ((u & ~IFF_VNET_HDR) != (IFF_NO_PI | IFF_TAP))
> -			ret = -EINVAL;
> -		else
> -			q->flags = u;
> -
> -		return ret;
> +		return macvtap_set_iff(file, ifr);
>  
>  	case TUNGETIFF:
>  		vlan = macvtap_get_vlan(q);
> -- 
> 1.7.1

Re: [net-next RFC 7/8] macvtap: add TUNSETQUEUE ioctl

[Michael S. Tsirkin]

On Thu, May 23, 2013 at 11:12:32AM +0800, Jason Wang wrote:
> This patch adds TUNSETQUEUE ioctl to let userspace can temporarily disable or
> enable a queue of macvtap. This is used to be compatible at API layer of tuntap
> to simplify the userspace to manage the queues.
> 
> This is done by just introduce a pointer detached which was set when the queue
> were detached. Only the queue with NULL set to this pointer can be used to
> forward packets.
> 
> Signed-off-by: Jason Wang <jasowang@redhat.com>

Isn't this a bit too tricky?
Why keep disabled queues on the array at all?
Let's add a linked list of all queues for accounting.
Use array for enabled queues only, for data path.

> ---
>  drivers/net/macvtap.c      |  147 +++++++++++++++++++++++++++++++++++--------
>  include/linux/if_macvlan.h |    7 ++
>  2 files changed, 126 insertions(+), 28 deletions(-)
> 
> diff --git a/drivers/net/macvtap.c b/drivers/net/macvtap.c
> index e3f9344..06b10a5 100644
> --- a/drivers/net/macvtap.c
> +++ b/drivers/net/macvtap.c
> @@ -85,67 +85,131 @@ static const struct proto_ops macvtap_socket_ops;
>   */
>  static DEFINE_SPINLOCK(macvtap_lock);
>  
> +static void macvtap_swap_slot(struct macvlan_dev *vlan, int a, int b)
> +{
> +	struct macvtap_queue *q1, *q2;
> +
> +	if (a == b)
> +		return;
> +
> +	q1 = rcu_dereference_protected(vlan->taps[a],
> +				       lockdep_is_held(&macvtap_lock));
> +	q2 = rcu_dereference_protected(vlan->taps[b],
> +				       lockdep_is_held(&macvtap_lock));
> +
> +	BUG_ON(q1 == NULL || q2 == NULL);
> +
> +	rcu_assign_pointer(vlan->taps[a], vlan->taps[b]);
> +	rcu_assign_pointer(vlan->taps[b], q1);
> +
> +	q1->queue_index = b;
> +	q2->queue_index = a;
> +}
> +
>  static int macvtap_set_queue(struct net_device *dev, struct file *file,
> -				struct macvtap_queue *q)
> +			     struct macvtap_queue *q, bool enable)
>  {
>  	struct macvlan_dev *vlan = netdev_priv(dev);
>  	int err = -EBUSY;
> +	int total;
>  
>  	spin_lock(&macvtap_lock);
> -	if (vlan->numvtaps == MAX_MACVTAP_QUEUES)
> +
> +	total = vlan->numvtaps + vlan->numdisabled;
> +	if (!enable && total == MAX_MACVTAP_QUEUES)
> +		goto out;
> +
> +	err = -EINVAL;
> +	if (enable && q->queue_index < vlan->numvtaps)
>  		goto out;
>  
>  	err = 0;
> -	rcu_assign_pointer(q->vlan, vlan);
> -	rcu_assign_pointer(vlan->taps[vlan->numvtaps], q);
> -	sock_hold(&q->sk);
>  
> -	q->file = file;
> -	q->queue_index = vlan->numvtaps;
> -	file->private_data = q;
> +	if (enable) {
> +		BUG_ON(q->queue_index >= total);
> +		macvtap_swap_slot(vlan, q->queue_index, vlan->numvtaps);
> +		--vlan->numdisabled;
> +	} else {
> +		rcu_assign_pointer(q->vlan, vlan);
> +		rcu_assign_pointer(vlan->taps[total], q);
> +		sock_hold(&q->sk);
> +
> +		q->file = file;
> +		q->queue_index = total;
> +		file->private_data = q;
> +		if (vlan->numdisabled)
> +			macvtap_swap_slot(vlan, vlan->numvtaps, total);
> +
> +		/* Make sure the pointers were seen before numvtaps. */
> +		smp_wmb();
> +	}
>  
>  	vlan->numvtaps++;
> -
>  out:
>  	spin_unlock(&macvtap_lock);
>  	return err;
>  }
>  
>  /*
> - * The file owning the queue got closed, give up both
> - * the reference that the files holds as well as the
> - * one from the macvlan_dev if that still exists.
> + * The file owning the queue got closed or disabled.
> + *
> + * If closed give up both the reference that the files holds as well as the
> + * one from the macvlan_dev if that still exists. If disabled, reshuffle the
> + * array of taps.
>   *
>   * Using the spinlock makes sure that we don't get
>   * to the queue again after destroying it.
>   */
> -static void macvtap_put_queue(struct macvtap_queue *q)
> +static int macvtap_put_queue(struct macvtap_queue *q, bool clean)

s/clean/release/
and add a comment explaining what it is.

>  {
> -	struct macvtap_queue *nq;
>  	struct macvlan_dev *vlan;
> +	int err = -EINVAL;
>  
>  	spin_lock(&macvtap_lock);
>  	vlan = rcu_dereference_protected(q->vlan,
>  					 lockdep_is_held(&macvtap_lock));
> +
>  	if (vlan) {
> +		int total = vlan->numvtaps + vlan->numdisabled;
>  		int index = q->queue_index;
> -		BUG_ON(index >= vlan->numvtaps);
> +		bool disabled = q->queue_index >= vlan->numvtaps;
>  
> -		rcu_assign_pointer(vlan->taps[index],
> -				   vlan->taps[vlan->numvtaps - 1]);
> -		RCU_INIT_POINTER(q->vlan, NULL);
> -		nq = rcu_dereference_protected(vlan->taps[index],
> -					       lockdep_is_held(&macvtap_lock));
> -		nq->queue_index = index;
> +		if (!clean) {
> +			BUG_ON(q->queue_index >= total);
> +			if (q->queue_index >= vlan->numvtaps)
> +				goto out;
> +		}
>  
> -		sock_put(&q->sk);
> -		--vlan->numvtaps;
> -	}
> +		err = 0;
> +		macvtap_swap_slot(vlan, index, total - 1);
> +		if (!disabled && vlan->numdisabled)

What's the logic here?

> +			macvtap_swap_slot(vlan, index, vlan->numvtaps - 1);

So these swaps do a lot of RCU assign, presumably we need
synchronize_rcu as well?

> +
> +		if (clean) {
> +			RCU_INIT_POINTER(vlan->taps[total - 1], NULL);
> +			RCU_INIT_POINTER(q->vlan, NULL);
> +			sock_put(&q->sk);
> +			if (disabled)
> +				--vlan->numdisabled;
> +			else
> +				--vlan->numvtaps;
> +		} else {
> +			--vlan->numvtaps;
> +			++vlan->numdisabled;
> +		}
> +
> +	} else if (clean)
> +		err = 0;
>  
> +out:
>  	spin_unlock(&macvtap_lock);
>  
>  	synchronize_rcu();
> -	sock_put(&q->sk);
> +
> +	if (clean)
> +		sock_put(&q->sk);
> +
> +	return err;
>  }
>  
>  /*
> @@ -201,7 +265,7 @@ static void macvtap_del_queues(struct net_device *dev)
>  
>  	/* macvtap_put_queue can free some slots, so go through all slots */
>  	spin_lock(&macvtap_lock);
> -	for (i = 0; i < vlan->numvtaps; i++) {
> +	for (i = 0; i < vlan->numvtaps + vlan->numdisabled; i++) {
>  		q = rcu_dereference_protected(vlan->taps[i],
>  					      lockdep_is_held(&macvtap_lock));
>  		BUG_ON(q == NULL);
> @@ -211,6 +275,7 @@ static void macvtap_del_queues(struct net_device *dev)
>  	}
>  	/* guarantee that any future macvtap_set_queue will fail */
>  	vlan->numvtaps = MAX_MACVTAP_QUEUES;
> +	vlan->numdisabled = 0;
>  	spin_unlock(&macvtap_lock);
>  
>  	synchronize_rcu();
> @@ -384,7 +449,7 @@ static int macvtap_open(struct inode *inode, struct file *file)
>  	if ((dev->features & NETIF_F_HIGHDMA) && (dev->features & NETIF_F_SG))
>  		sock_set_flag(&q->sk, SOCK_ZEROCOPY);
>  
> -	err = macvtap_set_queue(dev, file, q);
> +	err = macvtap_set_queue(dev, file, q, false);
>  	if (err)
>  		sock_put(&q->sk);
>  
> @@ -398,7 +463,7 @@ out:
>  static int macvtap_release(struct inode *inode, struct file *file)
>  {
>  	struct macvtap_queue *q = file->private_data;
> -	macvtap_put_queue(q);
> +	macvtap_put_queue(q, true);
>  	return 0;
>  }
>  
> @@ -922,6 +987,27 @@ static int macvtap_set_iff(struct file *file, struct ifreq __user *ifr_u)
>  	return 0;
>  }
>  
> +static int macvtap_ioctl_set_queue(struct file *file, unsigned int flags)
> +{
> +	struct macvtap_queue *q = file->private_data;
> +	struct macvlan_dev *vlan;
> +	int ret = -EINVAL;
> +
> +	vlan = macvtap_get_vlan(q);
> +	if (!vlan)
> +		goto done;
> +
> +	if (flags & IFF_ATTACH_QUEUE)
> +		ret = macvtap_set_queue(vlan->dev, file, q, true);
> +	else if (flags & IFF_DETACH_QUEUE)
> +		ret = macvtap_put_queue(q, false);
> +
> +	macvtap_put_vlan(vlan);
> +
> +done:
> +	return ret;
> +}
> +
>  /*
>   * provide compatibility with generic tun/tap interface
>   */
> @@ -954,6 +1040,11 @@ static long macvtap_ioctl(struct file *file, unsigned int cmd,
>  		macvtap_put_vlan(vlan);
>  		return ret;
>  
> +	case TUNSETQUEUE:
> +		if (get_user(u, &ifr->ifr_flags))
> +			return -EFAULT;
> +		return macvtap_ioctl_set_queue(file, u);
> +
>  	case TUNGETFEATURES:
>  		if (put_user(IFF_TAP | IFF_NO_PI | IFF_VNET_HDR, up))
>  			return -EFAULT;
> diff --git a/include/linux/if_macvlan.h b/include/linux/if_macvlan.h
> index 32e943a..f134afb 100644
> --- a/include/linux/if_macvlan.h
> +++ b/include/linux/if_macvlan.h
> @@ -69,8 +69,15 @@ struct macvlan_dev {
>  	u16			flags;
>  	int (*receive)(struct sk_buff *skb);
>  	int (*forward)(struct net_device *dev, struct sk_buff *skb);
> +	/* This array tracks all taps (include disabled ones) and will be
> +	 * reshuffled to keep the following order:
> +	 * [0, numvtaps) : enabled taps,
> +	 * [numvtaps, numvtaps + numdisabled) : disabled taps,
> +	 * [numvtaps + numdisabled, MAX_MACVTAP_QUEUES) : unused slots
> +	 */
>  	struct macvtap_queue	*taps[MAX_MACVTAP_QUEUES];
>  	int			numvtaps;
> +	int			numdisabled;
>  	int			minor;
>  };
>  
> -- 
> 1.7.1

Re: [net-next RFC 0/8] multiqueue API support for macvtap

[Michael S. Tsirkin]

On Thu, May 23, 2013 at 11:12:25AM +0800, Jason Wang wrote:
> Hi all:
> 
> This series implements a fully tuntap compatiable API which could be used by
> userspace to manage multi macvtap queues. The main parts is to add TUNSETQUEUE
> ioctl support for macvtap.
> 
> Patch 1 - 4 was some tuntap compatibility and misc cleanups.
> Patch 5 removes the linear search in macvtap by reshuffling the macvtaps array
> each time a queue is removed. After this, we could store both enabled and
> disabled queues in the same array without introducing extra data structure.
> Patch 6 let TUNSETIFF can create multiqueue device, nothing but some check were
> added.
> Patch 7 implement TUNSETQUEUE ioctl
> Patch 8 reports IFF_MULTI_QUEUE to userspace to notify the userspace that the
> multiqueue API is completed.
> 
> Flow caches implememtation were missed in this version, since I am doing rework
> on the tuntap flow caches. Have some some stress test with both netperf and
> pktgen.
> 
> Please review, thanks.

Overall looks good. Sent some comments on specific patches.

Thanks!

> Jason Wang (8):
>   macvlan: switch to use IS_ENABLED()
>   macvtap: return -EBADFD when TUNGETIFF fails
>   macvtap: introduce macvtap_get_vlan()
>   macvlan: reduce the max number of taps to 8
>   macvtap: eliminate linear search
>   macvtap: allow TUNSETIFF to create multiqueue device
>   macvtap: add TUNSETQUEUE ioctl
>   macvtap: enable multiqueue flag
> 
>  drivers/net/macvtap.c      |  265 ++++++++++++++++++++++++++++++-------------
>  include/linux/if_macvlan.h |   11 ++-
>  2 files changed, 194 insertions(+), 82 deletions(-)

Re: [net-next RFC 2/8] macvtap: return -EBADFD when TUNGETIFF fails

[Michael S. Tsirkin]

On Thu, May 23, 2013 at 11:12:27AM +0800, Jason Wang wrote:
> Tuntap return -EBADFD when TUNGETIFF fails, we should return the same value.
> 
> Signed-off-by: Jason Wang <jasowang@redhat.com>

Can you add some more comments on why this matters?
Any userspace that cares will have to handle both, right?

> ---
>  drivers/net/macvtap.c |    2 +-
>  1 files changed, 1 insertions(+), 1 deletions(-)
> 
> diff --git a/drivers/net/macvtap.c b/drivers/net/macvtap.c
> index 59e9605..ce1c72a 100644
> --- a/drivers/net/macvtap.c
> +++ b/drivers/net/macvtap.c
> @@ -928,7 +928,7 @@ static long macvtap_ioctl(struct file *file, unsigned int cmd,
>  		rcu_read_unlock_bh();
>  
>  		if (!vlan)
> -			return -ENOLINK;
> +			return -EBADFD;
>  
>  		ret = 0;
>  		if (copy_to_user(&ifr->ifr_name, vlan->dev->name, IFNAMSIZ) ||
> -- 
> 1.7.1

Re: [net-next RFC 3/8] macvtap: introduce macvtap_get_vlan()

[Sergei Shtylyov]

Hello.

On 23-05-2013 7:12, Jason Wang wrote:

> Factor out the device holding logic to a macvtap_get_vlan(), this will be also
> used by multiqueue API.

> Signed-off-by: Jason Wang <jasowang@redhat.com>
> ---
>   drivers/net/macvtap.c |   26 +++++++++++++++++++-------
>   1 files changed, 19 insertions(+), 7 deletions(-)

> diff --git a/drivers/net/macvtap.c b/drivers/net/macvtap.c
> index ce1c72a..a36e49e 100644
> --- a/drivers/net/macvtap.c
> +++ b/drivers/net/macvtap.c
> @@ -890,6 +890,23 @@ out:
>   	return ret;
>   }
>
> +static struct macvlan_dev *macvtap_get_vlan(struct macvtap_queue *q)
> +{
> +	struct macvlan_dev *vlan;

    Empty line wouldn't hurt here, after declaration.

> +	rcu_read_lock_bh();
> +	vlan = rcu_dereference_bh(q->vlan);
> +	if (vlan)
> +		dev_hold(vlan->dev);
> +	rcu_read_unlock_bh();
> +
> +	return vlan;
> +}
[...]

WBR, Sergei

Re: [net-next RFC 4/8] macvlan: reduce the max number of taps to 8

[Jason Wang]

On 05/23/2013 02:37 PM, Michael S. Tsirkin wrote:
> On Thu, May 23, 2013 at 11:12:29AM +0800, Jason Wang wrote:
>> To be same with tap.
>>
>> Signed-off-by: Jason Wang <jasowang@redhat.com>
> Well for tap the very specific reason was that
> there's an array of big queue structures,
> so we need to limit it to make it fit in a page.
> No such reason here right?

Right, one of the reason is that the flow caches itself occupies too
much space, I'd move it out of tun_struct.
>
> We need at least as much as tap to be compatible, so
> let's just make it 16 unconditionally?

Ok, in fact we can make it even more.
>
>> ---
>>  include/linux/if_macvlan.h |    2 +-
>>  1 files changed, 1 insertions(+), 1 deletions(-)
>>
>> diff --git a/include/linux/if_macvlan.h b/include/linux/if_macvlan.h
>> index e47ad46..32e943a 100644
>> --- a/include/linux/if_macvlan.h
>> +++ b/include/linux/if_macvlan.h
>> @@ -50,7 +50,7 @@ struct macvlan_pcpu_stats {
>>   * Maximum times a macvtap device can be opened. This can be used to
>>   * configure the number of receive queue, e.g. for multiqueue virtio.
>>   */
>> -#define MAX_MACVTAP_QUEUES	(NR_CPUS < 16 ? NR_CPUS : 16)
>> +#define MAX_MACVTAP_QUEUES	8
>>  
>>  #define MACVLAN_MC_FILTER_BITS	8
>>  #define MACVLAN_MC_FILTER_SZ	(1 << MACVLAN_MC_FILTER_BITS)
>> -- 
>> 1.7.1

Re: [net-next RFC 5/8] macvtap: eliminate linear search

[Jason Wang]

On 05/23/2013 07:41 PM, Michael S. Tsirkin wrote:
> On Thu, May 23, 2013 at 11:12:30AM +0800, Jason Wang wrote:
>> Linear search were used in both get_slot() and macvtap_get_queue(), this is
>> because:
>>
>> - macvtap didn't reshuffle the array of taps when create or destroy a queue, so
>>   when adding a new queue, macvtap must do linear search to find a location for
>>   the new queue. This will also complicate the TUNSETQUEUE implementation for
>>   multiqueue API.
>> - the queue itself didn't track the queue index, so the we must do a linear
>>   search in the array to find the location of a existed queue.
>>
>> The solution is straightforward: reshuffle the array and introduce a queue_index
>> to macvtap_queue.
>>
>> Signed-off-by: Jason Wang <jasowang@redhat.com>
>> ---
>>  drivers/net/macvtap.c |   63 +++++++++++++++---------------------------------
>>  1 files changed, 20 insertions(+), 43 deletions(-)
>>
>> diff --git a/drivers/net/macvtap.c b/drivers/net/macvtap.c
>> index a36e49e..5fd341c 100644
>> --- a/drivers/net/macvtap.c
>> +++ b/drivers/net/macvtap.c
>> @@ -44,6 +44,7 @@ struct macvtap_queue {
>>  	struct macvlan_dev __rcu *vlan;
>>  	struct file *file;
>>  	unsigned int flags;
>> +	u16 queue_index;
>>  };
>>  
>>  static struct proto macvtap_proto = {
>> @@ -84,30 +85,10 @@ static const struct proto_ops macvtap_socket_ops;
>>   */
>>  static DEFINE_SPINLOCK(macvtap_lock);
>>  
>> -/*
>> - * get_slot: return a [unused/occupied] slot in vlan->taps[]:
>> - *	- if 'q' is NULL, return the first empty slot;
>> - *	- otherwise, return the slot this pointer occupies.
>> - */
>> -static int get_slot(struct macvlan_dev *vlan, struct macvtap_queue *q)
>> -{
>> -	int i;
>> -
>> -	for (i = 0; i < MAX_MACVTAP_QUEUES; i++) {
>> -		if (rcu_dereference_protected(vlan->taps[i],
>> -					      lockdep_is_held(&macvtap_lock)) == q)
>> -			return i;
>> -	}
>> -
>> -	/* Should never happen */
>> -	BUG_ON(1);
>> -}
>> -
>>  static int macvtap_set_queue(struct net_device *dev, struct file *file,
>>  				struct macvtap_queue *q)
>>  {
>>  	struct macvlan_dev *vlan = netdev_priv(dev);
>> -	int index;
>>  	int err = -EBUSY;
>>  
>>  	spin_lock(&macvtap_lock);
>> @@ -115,12 +96,12 @@ static int macvtap_set_queue(struct net_device *dev, struct file *file,
>>  		goto out;
>>  
>>  	err = 0;
>> -	index = get_slot(vlan, NULL);
>>  	rcu_assign_pointer(q->vlan, vlan);
>> -	rcu_assign_pointer(vlan->taps[index], q);
>> +	rcu_assign_pointer(vlan->taps[vlan->numvtaps], q);
>>  	sock_hold(&q->sk);
>>  
>>  	q->file = file;
>> +	q->queue_index = vlan->numvtaps;
>>  	file->private_data = q;
>>  
>>  	vlan->numvtaps++;
>> @@ -140,16 +121,23 @@ out:
>>   */
>>  static void macvtap_put_queue(struct macvtap_queue *q)
>>  {
>> +	struct macvtap_queue *nq;
>>  	struct macvlan_dev *vlan;
>>  
>>  	spin_lock(&macvtap_lock);
>>  	vlan = rcu_dereference_protected(q->vlan,
>>  					 lockdep_is_held(&macvtap_lock));
>>  	if (vlan) {
>> -		int index = get_slot(vlan, q);
>> +		int index = q->queue_index;
>> +		BUG_ON(index >= vlan->numvtaps);
>>  
>> -		RCU_INIT_POINTER(vlan->taps[index], NULL);
>> +		rcu_assign_pointer(vlan->taps[index],
>> +				   vlan->taps[vlan->numvtaps - 1]);
>>  		RCU_INIT_POINTER(q->vlan, NULL);
>> +		nq = rcu_dereference_protected(vlan->taps[index],
>> +					       lockdep_is_held(&macvtap_lock));
>> +		nq->queue_index = index;
>> +
>>  		sock_put(&q->sk);
> Hmm this looks like a bug BTW: sock_put should be after
> no one uses this tap, so must be after synchronize_rcu.
> No?

Two refcnts were held, one is the socket itself (and we can safely put
it here), another is held by device when doing macvtap_set_queue(), it
was released by another sock_put() after synchronize_rcu().
>
>>  		--vlan->numvtaps;
>>  	}
> So:
> 		nq = rcu_dereference_protected(vlan->taps[vlan->numvtaps - 1],
> 					       lockdep_is_held(&macvtap_lock));
> 		rcu_assign_pointer(vlan->taps[index], nq);
>
> same thing?

Same and cleaner, thanks.
>
>> @@ -182,8 +170,7 @@ static struct macvtap_queue *macvtap_get_queue(struct net_device *dev,
>>  	rxq = skb_get_rxhash(skb);
>>  	if (rxq) {
>>  		tap = rcu_dereference(vlan->taps[rxq % numvtaps]);
>> -		if (tap)
>> -			goto out;
>> +		goto out;
>>  	}
>>  
>>  	if (likely(skb_rx_queue_recorded(skb))) {
>> @@ -193,17 +180,10 @@ static struct macvtap_queue *macvtap_get_queue(struct net_device *dev,
>>  			rxq -= numvtaps;
>>  
>>  		tap = rcu_dereference(vlan->taps[rxq]);
>> -		if (tap)
>> -			goto out;
>> -	}
>> -
>> -	/* Everything failed - find first available queue */
>> -	for (rxq = 0; rxq < MAX_MACVTAP_QUEUES; rxq++) {
>> -		tap = rcu_dereference(vlan->taps[rxq]);
>> -		if (tap)
>> -			break;
>> +		goto out;
>>  	}
>>  
>> +	tap = rcu_dereference(vlan->taps[0]);
>>  out:
>>  	return tap;
>>  }
>> @@ -221,17 +201,14 @@ static void macvtap_del_queues(struct net_device *dev)
>>  
>>  	/* macvtap_put_queue can free some slots, so go through all slots */
>>  	spin_lock(&macvtap_lock);
>> -	for (i = 0; i < MAX_MACVTAP_QUEUES && vlan->numvtaps; i++) {
>> +	for (i = 0; i < vlan->numvtaps; i++) {
>>  		q = rcu_dereference_protected(vlan->taps[i],
>>  					      lockdep_is_held(&macvtap_lock));
>> -		if (q) {
>> -			qlist[j++] = q;
>> -			RCU_INIT_POINTER(vlan->taps[i], NULL);
>> -			RCU_INIT_POINTER(q->vlan, NULL);
>> -			vlan->numvtaps--;
>> -		}
>> +		BUG_ON(q == NULL);
>> +		qlist[j++] = q;
>> +		RCU_INIT_POINTER(vlan->taps[i], NULL);
>> +		RCU_INIT_POINTER(q->vlan, NULL);
>>  	}
>> -	BUG_ON(vlan->numvtaps != 0);
>>  	/* guarantee that any future macvtap_set_queue will fail */
>>  	vlan->numvtaps = MAX_MACVTAP_QUEUES;
>>  	spin_unlock(&macvtap_lock);
>> -- 
>> 1.7.1

Re: [net-next RFC 6/8] macvtap: allow TUNSETIFF to create multiqueue device

[Jason Wang]

On 05/23/2013 07:45 PM, Michael S. Tsirkin wrote:
> On Thu, May 23, 2013 at 11:12:31AM +0800, Jason Wang wrote:
>> Though the queue were in fact created by open(), we still need to add this check
>> to be compatible with tuntap which can let mgmt software use a single API to
>> manage queues. This patch only validates the device name and moves the TUNSETIFF
>> to a helper.
>>
>> Signed-off-by: Jason Wang <jasowang@redhat.com>
>> ---
>>  drivers/net/macvtap.c |   50 ++++++++++++++++++++++++++++++++++++++----------
>>  1 files changed, 39 insertions(+), 11 deletions(-)
>>
>> diff --git a/drivers/net/macvtap.c b/drivers/net/macvtap.c
>> index 5fd341c..e3f9344 100644
>> --- a/drivers/net/macvtap.c
>> +++ b/drivers/net/macvtap.c
>> @@ -867,6 +867,7 @@ out:
>>  	return ret;
>>  }
>>  
>> +
> Please don't add empty lines like this :).

Sure.
>
>>  static struct macvlan_dev *macvtap_get_vlan(struct macvtap_queue *q)
>>  {
>>  	struct macvlan_dev *vlan;
>> @@ -884,6 +885,43 @@ static void macvtap_put_vlan(struct macvlan_dev *vlan)
>>  	dev_put(vlan->dev);
>>  }
>>  
>> +static int macvtap_set_iff(struct file *file, struct ifreq __user *ifr_u)
>> +{
>> +	struct macvtap_queue *q = file->private_data;
>> +	struct net *net = current->nsproxy->net_ns;
>> +	struct inode *inode = file_inode(file);
>> +	struct net_device *dev, *dev2;
>> +	struct ifreq ifr;
>> +
>> +	if (copy_from_user(&ifr, ifr_u, sizeof(struct ifreq)))
>> +		return -EFAULT;
>> +
>> +	if (ifr.ifr_flags & IFF_MULTI_QUEUE) {
> So why do we validate the name?
> Pls add a comment.

Ok. The validation is to prevent the userspace create the queues on the
wrong device. Not sure whether or not this is needed for single queue case.
>
>> +		dev = __dev_get_by_name(net, ifr.ifr_name);
>> +		if (!dev)
>> +			return -EINVAL;
>> +
>> +		dev2 = dev_get_by_macvtap_minor(iminor(inode));
>> +		if (!dev2)
>> +			return -EINVAL;
>> +
>> +		if (dev != dev2) {
>> +			dev_put(dev2);
>> +			return -EINVAL;
>> +		}
>> +
>> +		dev_put(dev2);
>> +	}
>> +
>> +	if ((ifr.ifr_flags & ~(IFF_VNET_HDR | IFF_MULTI_QUEUE)) !=
>> +	    (IFF_NO_PI | IFF_TAP))
>> +		return -EINVAL;
>> +	else
>> +		q->flags = ifr.ifr_flags;
>> +
>> +	return 0;
>> +}
>> +
>>  /*
>>   * provide compatibility with generic tun/tap interface
>>   */
>> @@ -902,17 +940,7 @@ static long macvtap_ioctl(struct file *file, unsigned int cmd,
>>  
>>  	switch (cmd) {
>>  	case TUNSETIFF:
>> -		/* ignore the name, just look at flags */
>> -		if (get_user(u, &ifr->ifr_flags))
>> -			return -EFAULT;
>> -
>> -		ret = 0;
>> -		if ((u & ~IFF_VNET_HDR) != (IFF_NO_PI | IFF_TAP))
>> -			ret = -EINVAL;
>> -		else
>> -			q->flags = u;
>> -
>> -		return ret;
>> +		return macvtap_set_iff(file, ifr);
>>  
>>  	case TUNGETIFF:
>>  		vlan = macvtap_get_vlan(q);
>> -- 
>> 1.7.1

Re: [net-next RFC 7/8] macvtap: add TUNSETQUEUE ioctl

[Jason Wang]

On 05/23/2013 07:52 PM, Michael S. Tsirkin wrote:
> On Thu, May 23, 2013 at 11:12:32AM +0800, Jason Wang wrote:
>> This patch adds TUNSETQUEUE ioctl to let userspace can temporarily disable or
>> enable a queue of macvtap. This is used to be compatible at API layer of tuntap
>> to simplify the userspace to manage the queues.
>>
>> This is done by just introduce a pointer detached which was set when the queue
>> were detached. Only the queue with NULL set to this pointer can be used to
>> forward packets.
>>
>> Signed-off-by: Jason Wang <jasowang@redhat.com>
> Isn't this a bit too tricky?

Well, not tricky and it was just like what we did the reshuffling in
PATCH 5/8
> Why keep disabled queues on the array at all?

Both are ok, but this method need less data structures in macvlan_dev.
> Let's add a linked list of all queues for accounting.
> Use array for enabled queues only, for data path.

This would not be simpler than just do in place reshuffling.
>
>> ---
>>  drivers/net/macvtap.c      |  147 +++++++++++++++++++++++++++++++++++--------
>>  include/linux/if_macvlan.h |    7 ++
>>  2 files changed, 126 insertions(+), 28 deletions(-)
>>
>> diff --git a/drivers/net/macvtap.c b/drivers/net/macvtap.c
>> index e3f9344..06b10a5 100644
>> --- a/drivers/net/macvtap.c
>> +++ b/drivers/net/macvtap.c
>> @@ -85,67 +85,131 @@ static const struct proto_ops macvtap_socket_ops;
>>   */
>>  static DEFINE_SPINLOCK(macvtap_lock);
>>  
>> +static void macvtap_swap_slot(struct macvlan_dev *vlan, int a, int b)
>> +{
>> +	struct macvtap_queue *q1, *q2;
>> +
>> +	if (a == b)
>> +		return;
>> +
>> +	q1 = rcu_dereference_protected(vlan->taps[a],
>> +				       lockdep_is_held(&macvtap_lock));
>> +	q2 = rcu_dereference_protected(vlan->taps[b],
>> +				       lockdep_is_held(&macvtap_lock));
>> +
>> +	BUG_ON(q1 == NULL || q2 == NULL);
>> +
>> +	rcu_assign_pointer(vlan->taps[a], vlan->taps[b]);
>> +	rcu_assign_pointer(vlan->taps[b], q1);
>> +
>> +	q1->queue_index = b;
>> +	q2->queue_index = a;
>> +}
>> +
>>  static int macvtap_set_queue(struct net_device *dev, struct file *file,
>> -				struct macvtap_queue *q)
>> +			     struct macvtap_queue *q, bool enable)
>>  {
>>  	struct macvlan_dev *vlan = netdev_priv(dev);
>>  	int err = -EBUSY;
>> +	int total;
>>  
>>  	spin_lock(&macvtap_lock);
>> -	if (vlan->numvtaps == MAX_MACVTAP_QUEUES)
>> +
>> +	total = vlan->numvtaps + vlan->numdisabled;
>> +	if (!enable && total == MAX_MACVTAP_QUEUES)
>> +		goto out;
>> +
>> +	err = -EINVAL;
>> +	if (enable && q->queue_index < vlan->numvtaps)
>>  		goto out;
>>  
>>  	err = 0;
>> -	rcu_assign_pointer(q->vlan, vlan);
>> -	rcu_assign_pointer(vlan->taps[vlan->numvtaps], q);
>> -	sock_hold(&q->sk);
>>  
>> -	q->file = file;
>> -	q->queue_index = vlan->numvtaps;
>> -	file->private_data = q;
>> +	if (enable) {
>> +		BUG_ON(q->queue_index >= total);
>> +		macvtap_swap_slot(vlan, q->queue_index, vlan->numvtaps);
>> +		--vlan->numdisabled;
>> +	} else {
>> +		rcu_assign_pointer(q->vlan, vlan);
>> +		rcu_assign_pointer(vlan->taps[total], q);
>> +		sock_hold(&q->sk);
>> +
>> +		q->file = file;
>> +		q->queue_index = total;
>> +		file->private_data = q;
>> +		if (vlan->numdisabled)
>> +			macvtap_swap_slot(vlan, vlan->numvtaps, total);
>> +
>> +		/* Make sure the pointers were seen before numvtaps. */
>> +		smp_wmb();
>> +	}
>>  
>>  	vlan->numvtaps++;
>> -
>>  out:
>>  	spin_unlock(&macvtap_lock);
>>  	return err;
>>  }
>>  
>>  /*
>> - * The file owning the queue got closed, give up both
>> - * the reference that the files holds as well as the
>> - * one from the macvlan_dev if that still exists.
>> + * The file owning the queue got closed or disabled.
>> + *
>> + * If closed give up both the reference that the files holds as well as the
>> + * one from the macvlan_dev if that still exists. If disabled, reshuffle the
>> + * array of taps.
>>   *
>>   * Using the spinlock makes sure that we don't get
>>   * to the queue again after destroying it.
>>   */
>> -static void macvtap_put_queue(struct macvtap_queue *q)
>> +static int macvtap_put_queue(struct macvtap_queue *q, bool clean)
> s/clean/release/
> and add a comment explaining what it is.

Ok.
>>  {
>> -	struct macvtap_queue *nq;
>>  	struct macvlan_dev *vlan;
>> +	int err = -EINVAL;
>>  
>>  	spin_lock(&macvtap_lock);
>>  	vlan = rcu_dereference_protected(q->vlan,
>>  					 lockdep_is_held(&macvtap_lock));
>> +
>>  	if (vlan) {
>> +		int total = vlan->numvtaps + vlan->numdisabled;
>>  		int index = q->queue_index;
>> -		BUG_ON(index >= vlan->numvtaps);
>> +		bool disabled = q->queue_index >= vlan->numvtaps;
>>  
>> -		rcu_assign_pointer(vlan->taps[index],
>> -				   vlan->taps[vlan->numvtaps - 1]);
>> -		RCU_INIT_POINTER(q->vlan, NULL);
>> -		nq = rcu_dereference_protected(vlan->taps[index],
>> -					       lockdep_is_held(&macvtap_lock));
>> -		nq->queue_index = index;
>> +		if (!clean) {
>> +			BUG_ON(q->queue_index >= total);
>> +			if (q->queue_index >= vlan->numvtaps)
>> +				goto out;
>> +		}
>>  
>> -		sock_put(&q->sk);
>> -		--vlan->numvtaps;
>> -	}
>> +		err = 0;
>> +		macvtap_swap_slot(vlan, index, total - 1);
>> +		if (!disabled && vlan->numdisabled)
> What's the logic here?

If there's at least one disabled taps, the above swap will move a
disabled tap into enabled area. So we need to do swap it with the last
enabled tap to keep the right order.
>
>> +			macvtap_swap_slot(vlan, index, vlan->numvtaps - 1);
> So these swaps do a lot of RCU assign, presumably we need
> synchronize_rcu as well?

There is a synchronize_rcu() below, and we do the cleanup -sock_put()
after that. And since we purge the socket during its destruction, no
need to care about the little possibility that a packet were sent to the
disabled or removed queue.
>
>> +
>> +		if (clean) {
>> +			RCU_INIT_POINTER(vlan->taps[total - 1], NULL);
>> +			RCU_INIT_POINTER(q->vlan, NULL);
>> +			sock_put(&q->sk);
>> +			if (disabled)
>> +				--vlan->numdisabled;
>> +			else
>> +				--vlan->numvtaps;
>> +		} else {
>> +			--vlan->numvtaps;
>> +			++vlan->numdisabled;
>> +		}
>> +
>> +	} else if (clean)
>> +		err = 0;
>>  
>> +out:
>>  	spin_unlock(&macvtap_lock);
>>  
>>  	synchronize_rcu();
>> -	sock_put(&q->sk);
>> +
>> +	if (clean)
>> +		sock_put(&q->sk);
>> +
>> +	return err;
>>  }
>>  
>>  /*
>> @@ -201,7 +265,7 @@ static void macvtap_del_queues(struct net_device *dev)
>>  
>>  	/* macvtap_put_queue can free some slots, so go through all slots */
>>  	spin_lock(&macvtap_lock);
>> -	for (i = 0; i < vlan->numvtaps; i++) {
>> +	for (i = 0; i < vlan->numvtaps + vlan->numdisabled; i++) {
>>  		q = rcu_dereference_protected(vlan->taps[i],
>>  					      lockdep_is_held(&macvtap_lock));
>>  		BUG_ON(q == NULL);
>> @@ -211,6 +275,7 @@ static void macvtap_del_queues(struct net_device *dev)
>>  	}
>>  	/* guarantee that any future macvtap_set_queue will fail */
>>  	vlan->numvtaps = MAX_MACVTAP_QUEUES;
>> +	vlan->numdisabled = 0;
>>  	spin_unlock(&macvtap_lock);
>>  
>>  	synchronize_rcu();
>> @@ -384,7 +449,7 @@ static int macvtap_open(struct inode *inode, struct file *file)
>>  	if ((dev->features & NETIF_F_HIGHDMA) && (dev->features & NETIF_F_SG))
>>  		sock_set_flag(&q->sk, SOCK_ZEROCOPY);
>>  
>> -	err = macvtap_set_queue(dev, file, q);
>> +	err = macvtap_set_queue(dev, file, q, false);
>>  	if (err)
>>  		sock_put(&q->sk);
>>  
>> @@ -398,7 +463,7 @@ out:
>>  static int macvtap_release(struct inode *inode, struct file *file)
>>  {
>>  	struct macvtap_queue *q = file->private_data;
>> -	macvtap_put_queue(q);
>> +	macvtap_put_queue(q, true);
>>  	return 0;
>>  }
>>  
>> @@ -922,6 +987,27 @@ static int macvtap_set_iff(struct file *file, struct ifreq __user *ifr_u)
>>  	return 0;
>>  }
>>  
>> +static int macvtap_ioctl_set_queue(struct file *file, unsigned int flags)
>> +{
>> +	struct macvtap_queue *q = file->private_data;
>> +	struct macvlan_dev *vlan;
>> +	int ret = -EINVAL;
>> +
>> +	vlan = macvtap_get_vlan(q);
>> +	if (!vlan)
>> +		goto done;
>> +
>> +	if (flags & IFF_ATTACH_QUEUE)
>> +		ret = macvtap_set_queue(vlan->dev, file, q, true);
>> +	else if (flags & IFF_DETACH_QUEUE)
>> +		ret = macvtap_put_queue(q, false);
>> +
>> +	macvtap_put_vlan(vlan);
>> +
>> +done:
>> +	return ret;
>> +}
>> +
>>  /*
>>   * provide compatibility with generic tun/tap interface
>>   */
>> @@ -954,6 +1040,11 @@ static long macvtap_ioctl(struct file *file, unsigned int cmd,
>>  		macvtap_put_vlan(vlan);
>>  		return ret;
>>  
>> +	case TUNSETQUEUE:
>> +		if (get_user(u, &ifr->ifr_flags))
>> +			return -EFAULT;
>> +		return macvtap_ioctl_set_queue(file, u);
>> +
>>  	case TUNGETFEATURES:
>>  		if (put_user(IFF_TAP | IFF_NO_PI | IFF_VNET_HDR, up))
>>  			return -EFAULT;
>> diff --git a/include/linux/if_macvlan.h b/include/linux/if_macvlan.h
>> index 32e943a..f134afb 100644
>> --- a/include/linux/if_macvlan.h
>> +++ b/include/linux/if_macvlan.h
>> @@ -69,8 +69,15 @@ struct macvlan_dev {
>>  	u16			flags;
>>  	int (*receive)(struct sk_buff *skb);
>>  	int (*forward)(struct net_device *dev, struct sk_buff *skb);
>> +	/* This array tracks all taps (include disabled ones) and will be
>> +	 * reshuffled to keep the following order:
>> +	 * [0, numvtaps) : enabled taps,
>> +	 * [numvtaps, numvtaps + numdisabled) : disabled taps,
>> +	 * [numvtaps + numdisabled, MAX_MACVTAP_QUEUES) : unused slots
>> +	 */
>>  	struct macvtap_queue	*taps[MAX_MACVTAP_QUEUES];
>>  	int			numvtaps;
>> +	int			numdisabled;
>>  	int			minor;
>>  };
>>  
>> -- 
>> 1.7.1

Re: [net-next RFC 3/8] macvtap: introduce macvtap_get_vlan()

[Jason Wang]

On 05/23/2013 11:11 PM, Sergei Shtylyov wrote:
> Hello.
>
> On 23-05-2013 7:12, Jason Wang wrote:
>
>> Factor out the device holding logic to a macvtap_get_vlan(), this
>> will be also
>> used by multiqueue API.
>
>> Signed-off-by: Jason Wang <jasowang@redhat.com>
>> ---
>>   drivers/net/macvtap.c |   26 +++++++++++++++++++-------
>>   1 files changed, 19 insertions(+), 7 deletions(-)
>
>> diff --git a/drivers/net/macvtap.c b/drivers/net/macvtap.c
>> index ce1c72a..a36e49e 100644
>> --- a/drivers/net/macvtap.c
>> +++ b/drivers/net/macvtap.c
>> @@ -890,6 +890,23 @@ out:
>>       return ret;
>>   }
>>
>> +static struct macvlan_dev *macvtap_get_vlan(struct macvtap_queue *q)
>> +{
>> +    struct macvlan_dev *vlan;
>
>    Empty line wouldn't hurt here, after declaration.

Will add one line here. Thanks
>
>> +    rcu_read_lock_bh();
>> +    vlan = rcu_dereference_bh(q->vlan);
>> +    if (vlan)
>> +        dev_hold(vlan->dev);
>> +    rcu_read_unlock_bh();
>> +
>> +    return vlan;
>> +}
> [...]
>
> WBR, Sergei
>
> -- 
> To unsubscribe from this list: send the line "unsubscribe netdev" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: [net-next RFC 2/8] macvtap: return -EBADFD when TUNGETIFF fails

[Jason Wang]

On 05/23/2013 07:54 PM, Michael S. Tsirkin wrote:
> On Thu, May 23, 2013 at 11:12:27AM +0800, Jason Wang wrote:
>> Tuntap return -EBADFD when TUNGETIFF fails, we should return the same value.
>>
>> Signed-off-by: Jason Wang <jasowang@redhat.com>
> Can you add some more comments on why this matters?

Just to keep compatibility for tuntap.
> Any userspace that cares will have to handle both, right?

Ideally the userspace should expect the same behavior for both tap and
macvtap.

Maybe we should just keep this.
>
>> ---
>>  drivers/net/macvtap.c |    2 +-
>>  1 files changed, 1 insertions(+), 1 deletions(-)
>>
>> diff --git a/drivers/net/macvtap.c b/drivers/net/macvtap.c
>> index 59e9605..ce1c72a 100644
>> --- a/drivers/net/macvtap.c
>> +++ b/drivers/net/macvtap.c
>> @@ -928,7 +928,7 @@ static long macvtap_ioctl(struct file *file, unsigned int cmd,
>>  		rcu_read_unlock_bh();
>>  
>>  		if (!vlan)
>> -			return -ENOLINK;
>> +			return -EBADFD;
>>  
>>  		ret = 0;
>>  		if (copy_to_user(&ifr->ifr_name, vlan->dev->name, IFNAMSIZ) ||
>> -- 
>> 1.7.1