From: Herbert Xu <herbert.xu@redhat.com>
Date: Wed, 6 May 2009 11:37:16 +0800
Subject: [net] tun: add packet accounting
Message-id: 20090506033716.GA27876@gondor.apana.org.au
O-Subject: Re: [RHEL5.4 PATCH] tun: Add packet accounting
Bugzilla: 495863
RH-Acked-by: Neil Horman <nhorman@redhat.com>
RH-Acked-by: Michael S. Tsirkin <mst@redhat.com>

Hi:

RHEL5.4 BZ #495863

We need to add packet accounting to the tun driver so that users
such as virtio-net gets congestion feedback which is necessary to
prevent packet loss for protocols lacking congestion conctrol (such
as UDP) when used in a guest.

This is a backport of the following upstream patches:

commit 33dccbb050bbe35b88ca8cf1228dcf3e4d4b3554
Author: Herbert Xu <herbert@gondor.apana.org.au>
Date:   Thu Feb 5 21:25:32 2009 -0800

    tun: Limit amount of queued packets per device

    Unlike a normal socket path, the tuntap device send path does
    not have any accounting.  This means that the user-space sender
    may be able to pin down arbitrary amounts of kernel memory by
    continuing to send data to an end-point that is congested.

    Even when this isn't an issue because of limited queueing at
    most end points, this can also be a problem because its only
    response to congestion is packet loss.  That is, when those
    local queues at the end-point fills up, the tuntap device will
    start wasting system time because it will continue to send
    data there which simply gets dropped straight away.

    Of course one could argue that everybody should do congestion
    control end-to-end, unfortunately there are people in this world
    still hooked on UDP, and they don't appear to be going away
    anywhere fast.  In fact, we've always helped them by performing
    accounting in our UDP code, the sole purpose of which is to
    provide congestion feedback other than through packet loss.

    This patch attempts to apply the same bandaid to the tuntap device.
    It creates a pseudo-socket object which is used to account our
    packets just as a normal socket does for UDP.  Of course things
    are a little complex because we're actually reinjecting traffic
    back into the stack rather than out of the stack.

    The stack complexities however should have been resolved by preceding
    patches.  So this one can simply start using skb_set_owner_w.

    For now the accounting is essentially disabled by default for
    backwards compatibility.  In particular, we set the cap to INT_MAX.
    This is so that existing applications don't get confused by the
    sudden arrival EAGAIN errors.

    In future we may wish (or be forced to) do this by default.

    Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
    Signed-off-by: David S. Miller <davem@davemloft.net>

commit 4cc7f68d65558f683c702d4fe3a5aac4c5227b97
Author: Herbert Xu <herbert@gondor.apana.org.au>
Date:   Wed Feb 4 16:55:54 2009 -0800

    net: Reexport sock_alloc_send_pskb

    The function sock_alloc_send_pskb is completely useless if not
    exported since most of the code in it won't be used as is.  In
    fact, this code has already been duplicated in the tun driver.

    Now that we need accounting in the tun driver, we can in fact
    use this function as is.  So this patch marks it for export again.

    Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
    Signed-off-by: David S. Miller <davem@davemloft.net>

commit 9a279bcbe347496799711155ed41a89bc40f79c5
Author: Herbert Xu <herbert@gondor.apana.org.au>
Date:   Wed Feb 4 16:55:27 2009 -0800

    net: Partially allow skb destructors to be used on receive path

    As it currently stands, skb destructors are forbidden on the
    receive path because the protocol end-points will overwrite
    any existing destructor with their own.

    This is the reason why we have to call skb_orphan in the loopback
    driver before we reinject the packet back into the stack, thus
    creating a period during which loopback traffic isn't charged
    to any socket.

    With virtualisation, we have a similar problem in that traffic
    is reinjected into the stack without being associated with any
    socket entity, thus providing no natural congestion push-back
    for those poor folks still stuck with UDP.

    Now had we been consistent in telling them that UDP simply has
    no congestion feedback, I could just fob them off.  Unfortunately,
    we appear to have gone to some length in catering for this on
    the standard UDP path, with skb/socket accounting so that has
    created a very unhealthy dependency.

    Alas habits are difficult to break out of, so we may just have
    to allow skb destructors on the receive path.

    It turns out that making skb destructors useable on the receive path
    isn't as easy as it seems.  For instance, simply adding skb_orphan
    to skb_set_owner_r isn't enough.  This is because we assume all
    over the IP stack that skb->sk is an IP socket if present.

    The new transparent proxy code goes one step further and assumes
    that skb->sk is the receiving socket if present.

    Now all of this can be dealt with by adding simple checks such
    as only treating skb->sk as an IP socket if skb->sk->sk_family
    matches.  However, it turns out that for bridging at least we
    don't need to do all of this work.

    This is of interest because most virtualisation setups use bridging
    so we don't actually go through the IP stack on the host (with
    the exception of our old nemesis the bridge netfilter, but that's
    easily taken care of).

    So this patch simply adds skb_orphan to the point just before we
    enter the IP stack, but after we've gone through the bridge on the
    receive path.  It also adds an skb_orphan to the one place in
    netfilter that touches skb->sk/skb->destructor, that is, tproxy.

    One word of caution, because of the internal code structure, anyone
    wishing to deploy this must use skb_set_owner_w as opposed to
    skb_set_owner_r since many functions that create a new skb from
    an existing one will invoke skb_set_owner_w on the new skb.

    Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
    Signed-off-by: David S. Miller <davem@davemloft.net>

diff --git a/drivers/net/tun.c b/drivers/net/tun.c
index 9a35086..3a4afe0 100644
--- a/drivers/net/tun.c
+++ b/drivers/net/tun.c
@@ -59,10 +59,16 @@
 #include <linux/if_tun.h>
 #include <linux/crc32.h>
 #include <linux/virtio_net.h>
+#include <net/sock.h>
 
 #include <asm/system.h>
 #include <asm/uaccess.h>
 
+struct tun_sock {
+	struct sock		sk;
+	struct tun_struct	*tun;
+};
+
 #ifdef TUN_DEBUG
 static int debug;
 #endif
@@ -72,6 +78,11 @@ static int debug;
 static LIST_HEAD(tun_dev_list);
 static struct ethtool_ops tun_ethtool_ops;
 
+static inline struct tun_sock *tun_sk(struct sock *sk)
+{
+	return container_of(sk, struct tun_sock, sk);
+}
+
 /* Net device open. */
 static int tun_net_open(struct net_device *dev)
 {
@@ -223,7 +234,8 @@ static void tun_net_init(struct net_device *dev)
 static unsigned int tun_chr_poll(struct file *file, poll_table * wait)
 {  
 	struct tun_struct *tun = file->private_data;
-	unsigned int mask = POLLOUT | POLLWRNORM;
+	struct sock *sk = tun->sk;
+	unsigned int mask = 0;
 
 	if (!tun)
 		return -EBADFD;
@@ -235,11 +247,44 @@ static unsigned int tun_chr_poll(struct file *file, poll_table * wait)
 	if (!skb_queue_empty(&tun->readq))
 		mask |= POLLIN | POLLRDNORM;
 
+	if (sock_writeable(sk) ||
+	    (!test_and_set_bit(SOCK_ASYNC_NOSPACE, &sk->sk_socket->flags) &&
+	     sock_writeable(sk)))
+		mask |= POLLOUT | POLLWRNORM;
+
 	return mask;
 }
 
+/* prepad is the amount to reserve at front.  len is length after that.
+ * linear is a hint as to how much to copy (usually headers). */
+static inline struct sk_buff *tun_alloc_skb(struct tun_struct *tun,
+					    size_t prepad, size_t len,
+					    size_t linear, int noblock)
+{
+	struct sock *sk = tun->sk;
+	struct sk_buff *skb;
+	int err;
+
+	/* Always linear for now. */
+	linear = len;
+
+	skb = sock_alloc_send_pskb(sk, prepad + linear, len - linear, noblock,
+				   &err);
+	if (!skb)
+		return ERR_PTR(err);
+
+	skb_reserve(skb, prepad);
+	skb_put(skb, linear);
+	skb->data_len = len - linear;
+	skb->len += len - linear;
+
+	return skb;
+}
+
 /* Get packet from user space buffer */
-static __inline__ ssize_t tun_get_user(struct tun_struct *tun, struct iovec *iv, size_t count)
+static __inline__ ssize_t tun_get_user(struct tun_struct *tun,
+				       struct iovec *iv, size_t count,
+				       int noblock)
 {
 	struct tun_pi pi = { 0, __constant_htons(ETH_P_IP) };
 	struct sk_buff *skb;
@@ -268,16 +313,13 @@ static __inline__ ssize_t tun_get_user(struct tun_struct *tun, struct iovec *iv,
 	if ((tun->flags & TUN_TYPE_MASK) == TUN_TAP_DEV)
 		align = NET_IP_ALIGN;
  
-	if (!(skb = alloc_skb(len + align, GFP_KERNEL))) {
-		tun->stats.rx_dropped++;
-		return -ENOMEM;
+	skb = tun_alloc_skb(tun, align, len, gso.hdr_len, noblock);
+	if (IS_ERR(skb)) {
+		if (PTR_ERR(skb) != -EAGAIN)
+			tun->stats.rx_dropped++;
+		return PTR_ERR(skb);
 	}
 
-	if (align)
-		skb_reserve(skb, align);
-
-	skb_put(skb, len);
-
 	if (gso.flags & VIRTIO_NET_HDR_F_NEEDS_CSUM) {
 		unsigned int csum = 0;
 
@@ -392,7 +434,8 @@ static ssize_t tun_chr_writev(struct file * file, const struct iovec *iv,
 
 	DBG(KERN_INFO "%s: tun_chr_write %ld\n", tun->dev->name, count);
 
-	return tun_get_user(tun, (struct iovec *) iv, iov_total(iv, count));
+	return tun_get_user(tun, (struct iovec *) iv, iov_total(iv, count),
+			    file->f_flags & O_NONBLOCK);
 }
 
 /* Write */
@@ -591,8 +634,37 @@ static struct tun_struct *tun_get_by_name(const char *name)
 	return NULL;
 }
 
+static void tun_sock_write_space(struct sock *sk)
+{
+	struct tun_struct *tun;
+
+	if (!sock_writeable(sk))
+		return;
+
+	if (sk->sk_sleep && waitqueue_active(sk->sk_sleep))
+		wake_up_interruptible_sync(sk->sk_sleep);
+
+	if (!test_and_clear_bit(SOCK_ASYNC_NOSPACE, &sk->sk_socket->flags))
+		return;
+
+	tun = container_of(sk, struct tun_sock, sk)->tun;
+	kill_fasync(&tun->fasync, SIGIO, POLL_OUT);
+}
+
+static void tun_sock_destruct(struct sock *sk)
+{
+	dev_put(container_of(sk, struct tun_sock, sk)->tun->dev);
+}
+
+static struct proto tun_proto = {
+	.name		= "tun",
+	.owner		= THIS_MODULE,
+	.obj_size	= sizeof(struct tun_sock),
+};
+
 static int tun_set_iff(struct file *file, struct ifreq *ifr)
 {
+	struct sock *sk;
 	struct tun_struct *tun;
 	struct net_device *dev;
 	int err;
@@ -648,17 +720,33 @@ static int tun_set_iff(struct file *file, struct ifreq *ifr)
 		get_random_bytes(tun->dev_addr + sizeof(u16), 4);
 		memset(tun->chr_filter, 0, sizeof tun->chr_filter);
 
+		err = -ENOMEM;
+		sk = sk_alloc(AF_UNSPEC, GFP_KERNEL, &tun_proto, 1);
+		if (!sk)
+			goto err_free_dev;
+
+		/* This ref count is for tun->sk. */
+		dev_hold(dev);
+		sock_init_data(&tun->socket, sk);
+		sk->sk_write_space = tun_sock_write_space;
+		sk->sk_destruct = tun_sock_destruct;
+		sk->sk_sndbuf = INT_MAX;
+		sk->sk_sleep = &tun->read_wait;
+
+		tun->sk = sk;
+		container_of(sk, struct tun_sock, sk)->tun = tun;
+
 		tun_net_init(dev);
 
 		if (strchr(dev->name, '%')) {
 			err = dev_alloc_name(dev, dev->name);
 			if (err < 0)
-				goto err_free_dev;
+				goto err_free_sk;
 		}
 
 		err = register_netdevice(tun->dev);
 		if (err < 0)
-			goto err_free_dev;
+			goto err_free_sk;
 	
 		list_add(&tun->list, &tun_dev_list);
 	}
@@ -682,6 +770,8 @@ static int tun_set_iff(struct file *file, struct ifreq *ifr)
 	strcpy(ifr->ifr_name, tun->dev->name);
 	return 0;
 
+ err_free_sk:
+	sock_put(sk);
  err_free_dev:
 	free_netdev(dev);
  failed:
@@ -764,6 +854,7 @@ static int tun_chr_ioctl(struct inode *inode, struct file *file,
 	struct tun_struct *tun = file->private_data;
 	void __user* argp = (void __user*)arg;
 	struct ifreq ifr;
+	int sndbuf;
 	int ret;
 
 	if (cmd == TUNSETIFF || _IOC_TYPE(cmd) == 0x89)
@@ -919,6 +1010,21 @@ static int tun_chr_ioctl(struct inode *inode, struct file *file,
 				(u8)ifr.ifr_hwaddr.sa_data[4], (u8)ifr.ifr_hwaddr.sa_data[5]);
 		return 0;
 
+	case TUNGETSNDBUF:
+		sndbuf = tun->sk->sk_sndbuf;
+		if (copy_to_user(argp, &sndbuf, sizeof(sndbuf)))
+			ret = -EFAULT;
+		break;
+
+	case TUNSETSNDBUF:
+		if (copy_from_user(&sndbuf, argp, sizeof(sndbuf))) {
+			ret = -EFAULT;
+			break;
+		}
+
+		tun->sk->sk_sndbuf = sndbuf;
+		break;
+
 	default:
 		return -EINVAL;
 	};
@@ -979,6 +1085,7 @@ static int tun_chr_close(struct inode *inode, struct file *file)
 
 	if (!(tun->flags & TUN_PERSIST)) {
 		list_del(&tun->list);
+		sock_put(tun->sk);
 		unregister_netdevice(tun->dev);
 	}
 
diff --git a/include/linux/if_tun.h b/include/linux/if_tun.h
index 6860229..25660dd 100644
--- a/include/linux/if_tun.h
+++ b/include/linux/if_tun.h
@@ -23,6 +23,8 @@
 
 #ifdef __KERNEL__
 
+#include <linux/net.h>
+
 #ifdef TUN_DEBUG
 #define DBG  if(tun->debug)printk
 #define DBG1 if(debug==2)printk
@@ -45,6 +47,9 @@ struct tun_struct {
 
 	struct fasync_struct    *fasync;
 
+	struct sock		*sk;
+	struct socket		socket;
+
 	unsigned long if_flags;
 	u8 dev_addr[ETH_ALEN];
 	u32 chr_filter[2];
@@ -82,6 +87,8 @@ struct tun_struct {
 #define TUNGETFEATURES _IOR('T', 207, unsigned int)
 #define TUNSETOFFLOAD _IOW('T', 208, unsigned int)
 #define TUNGETIFF      _IOR('T', 210, unsigned int)
+#define TUNGETSNDBUF  _IOR('T', 211, int)
+#define TUNSETSNDBUF  _IOW('T', 212, int)
 
 /* TUNSETIFF ifr flags */
 #define IFF_TUN		0x0001
diff --git a/include/net/sock.h b/include/net/sock.h
index c74d782..091750b 100644
--- a/include/net/sock.h
+++ b/include/net/sock.h
@@ -858,6 +858,11 @@ extern struct sk_buff 		*sock_alloc_send_skb(struct sock *sk,
 						     unsigned long size,
 						     int noblock,
 						     int *errcode);
+extern struct sk_buff 		*sock_alloc_send_pskb(struct sock *sk,
+						      unsigned long header_len,
+						      unsigned long data_len,
+						      int noblock,
+						      int *errcode);
 extern void *sock_kmalloc(struct sock *sk, int size,
 			  gfp_t priority);
 extern void sock_kfree_s(struct sock *sk, void *mem, int size);
diff --git a/net/core/dev.c b/net/core/dev.c
index dffc3e3..1e6b1cf 100644
--- a/net/core/dev.c
+++ b/net/core/dev.c
@@ -1907,6 +1907,8 @@ ncls:
 	if (handle_bridge(&skb, &pt_prev, &ret, orig_dev))
 		goto out;
 
+	skb_orphan(skb);
+
 	type = skb->protocol;
 	list_for_each_entry_rcu(ptype, &ptype_base[ntohs(type)&15], list) {
 		if (ptype->type == type &&
diff --git a/net/core/sock.c b/net/core/sock.c
index 4bb1732..418c6cd 100644
--- a/net/core/sock.c
+++ b/net/core/sock.c
@@ -1151,10 +1151,9 @@ static long sock_wait_for_wmem(struct sock * sk, long timeo)
  *	Generic send/receive buffer handlers
  */
 
-static struct sk_buff *sock_alloc_send_pskb(struct sock *sk,
-					    unsigned long header_len,
-					    unsigned long data_len,
-					    int noblock, int *errcode)
+struct sk_buff *sock_alloc_send_pskb(struct sock *sk, unsigned long header_len,
+				     unsigned long data_len, int noblock,
+				     int *errcode)
 {
 	struct sk_buff *skb;
 	gfp_t gfp_mask;
@@ -1234,8 +1233,9 @@ failure:
 	*errcode = err;
 	return NULL;
 }
+EXPORT_SYMBOL(sock_alloc_send_pskb);
 
-struct sk_buff *sock_alloc_send_skb(struct sock *sk, unsigned long size, 
+struct sk_buff *sock_alloc_send_skb(struct sock *sk, unsigned long size,
 				    int noblock, int *errcode)
 {
 	return sock_alloc_send_pskb(sk, size, 0, noblock, errcode);