Merge lp:~gandelman-a/ubuntu/precise/openvswitch/lp1044318 into lp:ubuntu/precise-proposed/openvswitch

[Adam Gandelman]

[lp:~gandelman-a/ubuntu/precise/openvswitch/lp1044318]

27. By Adam Gandelman on 2012-09-11

* Temporary PPA build to allow early RC1 testing while fix makes its way
  out to precise-updates.
     - debian/patches/lp1044318-Reset-upper-layer-protocol-info.patch: Cherry

28. By Adam Gandelman on 2012-09-12

Add .pc/lp1044318-Reset-upper-layer-protocol-info.patch.

[Stéphane Graber (stgraber)]

I'm seeing this in -proposed, marking it as merged.

=== modified file '.pc/applied-patches'
--- .pc/applied-patches	2012-07-26 11:23:06 +0000
+++ .pc/applied-patches	2012-09-12 17:06:26 +0000
@@ -1,2 +1,3 @@
 update_odputil_key_bytes.patch
 fix_ftbfs_big_endian.patch
+lp1044318-Reset-upper-layer-protocol-info.patch

=== added directory '.pc/lp1044318-Reset-upper-layer-protocol-info.patch'
=== added file '.pc/lp1044318-Reset-upper-layer-protocol-info.patch/.timestamp'
=== added directory '.pc/lp1044318-Reset-upper-layer-protocol-info.patch/datapath'
=== added file '.pc/lp1044318-Reset-upper-layer-protocol-info.patch/datapath/vport-internal_dev.c'
--- .pc/lp1044318-Reset-upper-layer-protocol-info.patch/datapath/vport-internal_dev.c	1970-01-01 00:00:00 +0000
+++ .pc/lp1044318-Reset-upper-layer-protocol-info.patch/datapath/vport-internal_dev.c	2012-09-12 17:06:26 +0000
@@ -0,0 +1,330 @@
+/*
+ * Copyright (c) 2007-2011 Nicira Networks.
+ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of version 2 of the GNU General Public
+ * License as published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope that it will be useful, but
+ * WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
+ * General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA
+ * 02110-1301, USA
+ */
+
+#include <linux/hardirq.h>
+#include <linux/if_vlan.h>
+#include <linux/kernel.h>
+#include <linux/netdevice.h>
+#include <linux/etherdevice.h>
+#include <linux/ethtool.h>
+#include <linux/skbuff.h>
+#include <linux/version.h>
+
+#include "checksum.h"
+#include "datapath.h"
+#include "vlan.h"
+#include "vport-generic.h"
+#include "vport-internal_dev.h"
+#include "vport-netdev.h"
+
+#if LINUX_VERSION_CODE >= KERNEL_VERSION(3,1,0)
+#define HAVE_NET_DEVICE_OPS
+#endif
+
+struct internal_dev {
+	struct vport *vport;
+#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,22)
+	struct net_device_stats stats;
+#endif
+};
+
+static struct internal_dev *internal_dev_priv(struct net_device *netdev)
+{
+	return netdev_priv(netdev);
+}
+
+/* This function is only called by the kernel network layer.*/
+#if LINUX_VERSION_CODE >= KERNEL_VERSION(2,6,36)
+static struct rtnl_link_stats64 *internal_dev_get_stats(struct net_device *netdev,
+							struct rtnl_link_stats64 *stats)
+{
+#else
+static struct net_device_stats *internal_dev_sys_stats(struct net_device *netdev)
+{
+#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,22)
+	struct net_device_stats *stats = &internal_dev_priv(netdev)->stats;
+#else
+	struct net_device_stats *stats = &netdev->stats;
+#endif
+#endif
+	struct vport *vport = ovs_internal_dev_get_vport(netdev);
+	struct ovs_vport_stats vport_stats;
+
+	ovs_vport_get_stats(vport, &vport_stats);
+
+	/* The tx and rx stats need to be swapped because the
+	 * switch and host OS have opposite perspectives. */
+	stats->rx_packets	= vport_stats.tx_packets;
+	stats->tx_packets	= vport_stats.rx_packets;
+	stats->rx_bytes		= vport_stats.tx_bytes;
+	stats->tx_bytes		= vport_stats.rx_bytes;
+	stats->rx_errors	= vport_stats.tx_errors;
+	stats->tx_errors	= vport_stats.rx_errors;
+	stats->rx_dropped	= vport_stats.tx_dropped;
+	stats->tx_dropped	= vport_stats.rx_dropped;
+
+	return stats;
+}
+
+static int internal_dev_mac_addr(struct net_device *dev, void *p)
+{
+	struct sockaddr *addr = p;
+
+	if (!is_valid_ether_addr(addr->sa_data))
+		return -EADDRNOTAVAIL;
+	memcpy(dev->dev_addr, addr->sa_data, dev->addr_len);
+	return 0;
+}
+
+/* Called with rcu_read_lock_bh. */
+static int internal_dev_xmit(struct sk_buff *skb, struct net_device *netdev)
+{
+	if (unlikely(compute_ip_summed(skb, true))) {
+		kfree_skb(skb);
+		return 0;
+	}
+
+	vlan_copy_skb_tci(skb);
+	OVS_CB(skb)->flow = NULL;
+
+	rcu_read_lock();
+	ovs_vport_receive(internal_dev_priv(netdev)->vport, skb);
+	rcu_read_unlock();
+	return 0;
+}
+
+static int internal_dev_open(struct net_device *netdev)
+{
+	netif_start_queue(netdev);
+	return 0;
+}
+
+static int internal_dev_stop(struct net_device *netdev)
+{
+	netif_stop_queue(netdev);
+	return 0;
+}
+
+static void internal_dev_getinfo(struct net_device *netdev,
+				 struct ethtool_drvinfo *info)
+{
+	strcpy(info->driver, "openvswitch");
+}
+
+static const struct ethtool_ops internal_dev_ethtool_ops = {
+	.get_drvinfo	= internal_dev_getinfo,
+	.get_link	= ethtool_op_get_link,
+#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,39)
+	.get_sg		= ethtool_op_get_sg,
+	.set_sg		= ethtool_op_set_sg,
+	.get_tx_csum	= ethtool_op_get_tx_csum,
+	.set_tx_csum	= ethtool_op_set_tx_hw_csum,
+	.get_tso	= ethtool_op_get_tso,
+	.set_tso	= ethtool_op_set_tso,
+#endif
+};
+
+static int internal_dev_change_mtu(struct net_device *netdev, int new_mtu)
+{
+	if (new_mtu < 68)
+		return -EINVAL;
+
+	netdev->mtu = new_mtu;
+	return 0;
+}
+
+static int internal_dev_do_ioctl(struct net_device *dev,
+				 struct ifreq *ifr, int cmd)
+{
+	if (ovs_dp_ioctl_hook)
+		return ovs_dp_ioctl_hook(dev, ifr, cmd);
+
+	return -EOPNOTSUPP;
+}
+
+static void internal_dev_destructor(struct net_device *dev)
+{
+	struct vport *vport = ovs_internal_dev_get_vport(dev);
+
+	ovs_vport_free(vport);
+	free_netdev(dev);
+}
+
+#ifdef HAVE_NET_DEVICE_OPS
+static const struct net_device_ops internal_dev_netdev_ops = {
+	.ndo_open = internal_dev_open,
+	.ndo_stop = internal_dev_stop,
+	.ndo_start_xmit = internal_dev_xmit,
+	.ndo_set_mac_address = internal_dev_mac_addr,
+	.ndo_do_ioctl = internal_dev_do_ioctl,
+	.ndo_change_mtu = internal_dev_change_mtu,
+#if LINUX_VERSION_CODE >= KERNEL_VERSION(2,6,36)
+	.ndo_get_stats64 = internal_dev_get_stats,
+#else
+	.ndo_get_stats = internal_dev_sys_stats,
+#endif
+};
+#endif
+
+static void do_setup(struct net_device *netdev)
+{
+	ether_setup(netdev);
+
+#ifdef HAVE_NET_DEVICE_OPS
+	netdev->netdev_ops = &internal_dev_netdev_ops;
+#else
+	netdev->do_ioctl = internal_dev_do_ioctl;
+	netdev->get_stats = internal_dev_sys_stats;
+	netdev->hard_start_xmit = internal_dev_xmit;
+	netdev->open = internal_dev_open;
+	netdev->stop = internal_dev_stop;
+	netdev->set_mac_address = internal_dev_mac_addr;
+	netdev->change_mtu = internal_dev_change_mtu;
+#endif
+
+	netdev->priv_flags &= ~IFF_TX_SKB_SHARING;
+	netdev->destructor = internal_dev_destructor;
+	SET_ETHTOOL_OPS(netdev, &internal_dev_ethtool_ops);
+	netdev->tx_queue_len = 0;
+
+	netdev->features = NETIF_F_LLTX | NETIF_F_SG | NETIF_F_FRAGLIST |
+				NETIF_F_HIGHDMA | NETIF_F_HW_CSUM | NETIF_F_TSO;
+
+#if LINUX_VERSION_CODE >= KERNEL_VERSION(2,6,27)
+	netdev->vlan_features = netdev->features;
+	netdev->features |= NETIF_F_HW_VLAN_TX;
+#endif
+
+#if LINUX_VERSION_CODE >= KERNEL_VERSION(2,6,39)
+	netdev->hw_features = netdev->features & ~NETIF_F_LLTX;
+#endif
+	random_ether_addr(netdev->dev_addr);
+}
+
+static struct vport *internal_dev_create(const struct vport_parms *parms)
+{
+	struct vport *vport;
+	struct netdev_vport *netdev_vport;
+	struct internal_dev *internal_dev;
+	int err;
+
+	vport = ovs_vport_alloc(sizeof(struct netdev_vport),
+				&ovs_internal_vport_ops, parms);
+	if (IS_ERR(vport)) {
+		err = PTR_ERR(vport);
+		goto error;
+	}
+
+	netdev_vport = netdev_vport_priv(vport);
+
+	netdev_vport->dev = alloc_netdev(sizeof(struct internal_dev),
+					 parms->name, do_setup);
+	if (!netdev_vport->dev) {
+		err = -ENOMEM;
+		goto error_free_vport;
+	}
+
+	internal_dev = internal_dev_priv(netdev_vport->dev);
+	internal_dev->vport = vport;
+
+	err = register_netdevice(netdev_vport->dev);
+	if (err)
+		goto error_free_netdev;
+
+	dev_set_promiscuity(netdev_vport->dev, 1);
+	netif_start_queue(netdev_vport->dev);
+
+	return vport;
+
+error_free_netdev:
+	free_netdev(netdev_vport->dev);
+error_free_vport:
+	ovs_vport_free(vport);
+error:
+	return ERR_PTR(err);
+}
+
+static void internal_dev_destroy(struct vport *vport)
+{
+	struct netdev_vport *netdev_vport = netdev_vport_priv(vport);
+
+	netif_stop_queue(netdev_vport->dev);
+	dev_set_promiscuity(netdev_vport->dev, -1);
+
+	/* unregister_netdevice() waits for an RCU grace period. */
+	unregister_netdevice(netdev_vport->dev);
+}
+
+static int internal_dev_recv(struct vport *vport, struct sk_buff *skb)
+{
+	struct net_device *netdev = netdev_vport_priv(vport)->dev;
+	int len;
+
+#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,37)
+	if (unlikely(vlan_deaccel_tag(skb)))
+		return 0;
+#endif
+
+	len = skb->len;
+	skb->dev = netdev;
+	skb->pkt_type = PACKET_HOST;
+	skb->protocol = eth_type_trans(skb, netdev);
+	forward_ip_summed(skb, false);
+
+	netif_rx(skb);
+
+#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,29)
+	netdev->last_rx = jiffies;
+#endif
+
+	return len;
+}
+
+const struct vport_ops ovs_internal_vport_ops = {
+	.type		= OVS_VPORT_TYPE_INTERNAL,
+	.flags		= VPORT_F_REQUIRED | VPORT_F_FLOW,
+	.create		= internal_dev_create,
+	.destroy	= internal_dev_destroy,
+	.set_addr	= ovs_netdev_set_addr,
+	.get_name	= ovs_netdev_get_name,
+	.get_addr	= ovs_netdev_get_addr,
+	.get_kobj	= ovs_netdev_get_kobj,
+	.get_dev_flags	= ovs_netdev_get_dev_flags,
+	.is_running	= ovs_netdev_is_running,
+	.get_operstate	= ovs_netdev_get_operstate,
+	.get_ifindex	= ovs_netdev_get_ifindex,
+	.get_mtu	= ovs_netdev_get_mtu,
+	.send		= internal_dev_recv,
+};
+
+int ovs_is_internal_dev(const struct net_device *netdev)
+{
+#ifdef HAVE_NET_DEVICE_OPS
+	return netdev->netdev_ops == &internal_dev_netdev_ops;
+#else
+	return netdev->open == internal_dev_open;
+#endif
+}
+
+struct vport *ovs_internal_dev_get_vport(struct net_device *netdev)
+{
+	if (!ovs_is_internal_dev(netdev))
+		return NULL;
+
+	return internal_dev_priv(netdev)->vport;
+}

=== modified file 'datapath/vport-internal_dev.c'
--- datapath/vport-internal_dev.c	2012-01-30 23:36:00 +0000
+++ datapath/vport-internal_dev.c	2012-09-12 17:06:26 +0000
@@ -25,6 +25,9 @@
 #include <linux/skbuff.h>
 #include <linux/version.h>
 
+#include <net/dst.h>
+#include <net/xfrm.h>
+
 #include "checksum.h"
 #include "datapath.h"
 #include "vlan.h"
@@ -281,6 +284,11 @@
 #endif
 
 	len = skb->len;
+
+	skb_dst_drop(skb);
+	nf_reset(skb);
+	secpath_reset(skb);
+
 	skb->dev = netdev;
 	skb->pkt_type = PACKET_HOST;
 	skb->protocol = eth_type_trans(skb, netdev);

=== modified file 'debian/changelog'
--- debian/changelog	2012-07-26 11:23:06 +0000
+++ debian/changelog	2012-09-12 17:06:26 +0000
@@ -1,3 +1,13 @@
+openvswitch (1.4.0-1ubuntu1.3~ppa) precise; urgency=low
+
+  * Temporary PPA build to allow early RC1 testing while fix makes its way
+    out to precise-updates.
+       - debian/patches/lp1044318-Reset-upper-layer-protocol-info.patch: Cherry
+         picked upstream patch to avoid critical issues with SNAT/DNAT when OVS
+         is chained with other Linux components. (LP: #1044318)
+
+ -- Adam Gandelman <[e-mail]>  Fri, 07 Sep 2012 15:38:56 -0700
+
 openvswitch (1.4.0-1ubuntu1.2) precise-proposed; urgency=low
 
   * debian/patches/fix_ftbfs_big_endian.patch: Fix FTBFS on PPC.

=== added file 'debian/patches/lp1044318-Reset-upper-layer-protocol-info.patch'
--- debian/patches/lp1044318-Reset-upper-layer-protocol-info.patch	1970-01-01 00:00:00 +0000
+++ debian/patches/lp1044318-Reset-upper-layer-protocol-info.patch	2012-09-12 17:06:26 +0000
@@ -0,0 +1,52 @@
+From: Adam Gandelman <[e-mail]>
+Author: Jesse Gross <[e-mail]>
+Date: Fri Sep  7 12:10:58 PDT 2012
+Bug-Ubuntu: https://bugs.launchpad.net/quantum/+bug/1044318
+X-Git-Url: http://openvswitch.org/cgi-bin/gitweb.cgi?p=openvswitch;a=commitdiff_plain;h=53e6421bc83918ac2d00ba5516f205fa7e394140
+Subject: datapath: Reset upper layer protocol info on internal devices.
+
+datapath: Reset upper layer protocol info on internal devices.
+
+It's possible that packets that are sent on internal devices (from
+the OVS perspective) have already traversed the local IP stack.
+After they go through the internal device, they will again travel
+through the IP stack which may get confused by the presence of
+existing information in the skb. The problem can be observed
+when switching between namespaces. This clears out that information
+to avoid problems but deliberately leaves other metadata alone.
+This is to provide maximum flexibility in chaining together OVS
+and other Linux components.
+
+Bug #10995
+
+Signed-off-by: Jesse Gross <[e-mail]>
+Acked-by: Ben Pfaff <[e-mail]>
+---
+
+
+Index: openvswitch/datapath/vport-internal_dev.c
+===================================================================
+--- openvswitch.orig/datapath/vport-internal_dev.c	2012-09-07 12:06:26.436795000 -0700
++++ openvswitch/datapath/vport-internal_dev.c	2012-09-07 12:09:09.223267544 -0700
+@@ -25,6 +25,9 @@
+ #include <linux/skbuff.h>
+ #include <linux/version.h>
+ 
++#include <net/dst.h>
++#include <net/xfrm.h>
++
+ #include "checksum.h"
+ #include "datapath.h"
+ #include "vlan.h"
+@@ -281,6 +284,11 @@
+ #endif
+ 
+ 	len = skb->len;
++
++	skb_dst_drop(skb);
++	nf_reset(skb);
++	secpath_reset(skb);
++
+ 	skb->dev = netdev;
+ 	skb->pkt_type = PACKET_HOST;
+ 	skb->protocol = eth_type_trans(skb, netdev);

=== modified file 'debian/patches/series'
--- debian/patches/series	2012-07-26 11:23:06 +0000
+++ debian/patches/series	2012-09-12 17:06:26 +0000
@@ -1,2 +1,3 @@
 update_odputil_key_bytes.patch
 fix_ftbfs_big_endian.patch
+lp1044318-Reset-upper-layer-protocol-info.patch