xfrm: interface: support collect metadata mode

This commit adds support for 'collect_md' mode on xfrm interfaces.

Each net can have one collect_md device, created by providing the
IFLA_XFRM_COLLECT_METADATA flag at creation. This device cannot be
altered and has no if_id or link device attributes.

On transmit to this device, the if_id is fetched from the attached dst
metadata on the skb. If exists, the link property is also fetched from
the metadata. The dst metadata type used is METADATA_XFRM which holds
these properties.

On the receive side, xfrmi_rcv_cb() populates a dst metadata for each
packet received and attaches it to the skb. The if_id used in this case is
fetched from the xfrm state, and the link is fetched from the incoming
device. This information can later be used by upper layers such as tc,
ebpf, and ip rules.

Because the skb is scrubed in xfrmi_rcv_cb(), the attachment of the dst
metadata is postponed until after scrubing. Similarly, xfrm_input() is
adapted to avoid dropping metadata dsts by only dropping 'valid'
(skb_valid_dst(skb) == true) dsts.

Policy matching on packets arriving from collect_md xfrmi devices is
done by using the xfrm state existing in the skb's sec_path.
The xfrm_if_cb.decode_cb() interface implemented by xfrmi_decode_session()
is changed to keep the details of the if_id extraction tucked away
in xfrm_interface.c.

Reviewed-by: Nicolas Dichtel <nicolas.dichtel@6wind.com>
Reviewed-by: Nikolay Aleksandrov <razor@blackwall.org>
Signed-off-by: Eyal Birger <eyal.birger@gmail.com>
Signed-off-by: Steffen Klassert <steffen.klassert@secunet.com>

1 files changed, 101 insertions, 20 deletions

diff --git a/net/xfrm/xfrm_interface.c b/net/xfrm/xfrm_interface.c
index 5113fa0fbcee..e9a355047468 100644
--- a/net/xfrm/xfrm_interface.c
+++ b/net/xfrm/xfrm_interface.c
@@ -41,6 +41,7 @@
 #include <net/addrconf.h>
 #include <net/xfrm.h>
 #include <net/net_namespace.h>
+#include <net/dst_metadata.h>
 #include <net/netns/generic.h>
 #include <linux/etherdevice.h>
 
@@ -56,6 +57,7 @@ static const struct net_device_ops xfrmi_netdev_ops;
 struct xfrmi_net {
 	/* lists for storing interfaces in use */
 	struct xfrm_if __rcu *xfrmi[XFRMI_HASH_SIZE];
+	struct xfrm_if __rcu *collect_md_xfrmi;
 };
 
 #define for_each_xfrmi_rcu(start, xi) \
@@ -77,17 +79,23 @@ static struct xfrm_if *xfrmi_lookup(struct net *net, struct xfrm_state *x)
 			return xi;
 	}
 
+	xi = rcu_dereference(xfrmn->collect_md_xfrmi);
+	if (xi && (xi->dev->flags & IFF_UP))
+		return xi;
+
 	return NULL;
 }
 
-static struct xfrm_if *xfrmi_decode_session(struct sk_buff *skb,
-					    unsigned short family)
+static bool xfrmi_decode_session(struct sk_buff *skb,
+				 unsigned short family,
+				 struct xfrm_if_decode_session_result *res)
 {
 	struct net_device *dev;
+	struct xfrm_if *xi;
 	int ifindex = 0;
 
 	if (!secpath_exists(skb) || !skb->dev)
-		return NULL;
+		return false;
 
 	switch (family) {
 	case AF_INET6:
@@ -107,11 +115,18 @@ static struct xfrm_if *xfrmi_decode_session(struct sk_buff *skb,
 	}
 
 	if (!dev || !(dev->flags & IFF_UP))
-		return NULL;
+		return false;
 	if (dev->netdev_ops != &xfrmi_netdev_ops)
-		return NULL;
+		return false;
 
-	return netdev_priv(dev);
+	xi = netdev_priv(dev);
+	res->net = xi->net;
+
+	if (xi->p.collect_md)
+		res->if_id = xfrm_input_state(skb)->if_id;
+	else
+		res->if_id = xi->p.if_id;
+	return true;
 }
 
 static void xfrmi_link(struct xfrmi_net *xfrmn, struct xfrm_if *xi)
@@ -157,7 +172,10 @@ static int xfrmi_create(struct net_device *dev)
 	if (err < 0)
 		goto out;
 
-	xfrmi_link(xfrmn, xi);
+	if (xi->p.collect_md)
+		rcu_assign_pointer(xfrmn->collect_md_xfrmi, xi);
+	else
+		xfrmi_link(xfrmn, xi);
 
 	return 0;
 
@@ -185,7 +203,10 @@ static void xfrmi_dev_uninit(struct net_device *dev)
 	struct xfrm_if *xi = netdev_priv(dev);
 	struct xfrmi_net *xfrmn = net_generic(xi->net, xfrmi_net_id);
 
-	xfrmi_unlink(xfrmn, xi);
+	if (xi->p.collect_md)
+		RCU_INIT_POINTER(xfrmn->collect_md_xfrmi, NULL);
+	else
+		xfrmi_unlink(xfrmn, xi);
 }
 
 static void xfrmi_scrub_packet(struct sk_buff *skb, bool xnet)
@@ -214,6 +235,7 @@ static int xfrmi_rcv_cb(struct sk_buff *skb, int err)
 	struct xfrm_state *x;
 	struct xfrm_if *xi;
 	bool xnet;
+	int link;
 
 	if (err && !secpath_exists(skb))
 		return 0;
@@ -224,6 +246,7 @@ static int xfrmi_rcv_cb(struct sk_buff *skb, int err)
 	if (!xi)
 		return 1;
 
+	link = skb->dev->ifindex;
 	dev = xi->dev;
 	skb->dev = dev;
 
@@ -254,6 +277,17 @@ static int xfrmi_rcv_cb(struct sk_buff *skb, int err)
 	}
 
 	xfrmi_scrub_packet(skb, xnet);
+	if (xi->p.collect_md) {
+		struct metadata_dst *md_dst;
+
+		md_dst = metadata_dst_alloc(0, METADATA_XFRM, GFP_ATOMIC);
+		if (!md_dst)
+			return -ENOMEM;
+
+		md_dst->u.xfrm_info.if_id = x->if_id;
+		md_dst->u.xfrm_info.link = link;
+		skb_dst_set(skb, (struct dst_entry *)md_dst);
+	}
 	dev_sw_netstats_rx_add(dev, skb->len);
 
 	return 0;
@@ -269,10 +303,23 @@ xfrmi_xmit2(struct sk_buff *skb, struct net_device *dev, struct flowi *fl)
 	struct net_device *tdev;
 	struct xfrm_state *x;
 	int err = -1;
+	u32 if_id;
 	int mtu;
 
+	if (xi->p.collect_md) {
+		struct xfrm_md_info *md_info = skb_xfrm_md_info(skb);
+
+		if (unlikely(!md_info))
+			return -EINVAL;
+
+		if_id = md_info->if_id;
+		fl->flowi_oif = md_info->link;
+	} else {
+		if_id = xi->p.if_id;
+	}
+
 	dst_hold(dst);
-	dst = xfrm_lookup_with_ifid(xi->net, dst, fl, NULL, 0, xi->p.if_id);
+	dst = xfrm_lookup_with_ifid(xi->net, dst, fl, NULL, 0, if_id);
 	if (IS_ERR(dst)) {
 		err = PTR_ERR(dst);
 		dst = NULL;
@@ -283,7 +330,7 @@ xfrmi_xmit2(struct sk_buff *skb, struct net_device *dev, struct flowi *fl)
 	if (!x)
 		goto tx_err_link_failure;
 
-	if (x->if_id != xi->p.if_id)
+	if (x->if_id != if_id)
 		goto tx_err_link_failure;
 
 	tdev = dst->dev;
@@ -633,6 +680,9 @@ static void xfrmi_netlink_parms(struct nlattr *data[],
 
 	if (data[IFLA_XFRM_IF_ID])
 		parms->if_id = nla_get_u32(data[IFLA_XFRM_IF_ID]);
+
+	if (data[IFLA_XFRM_COLLECT_METADATA])
+		parms->collect_md = true;
 }
 
 static int xfrmi_newlink(struct net *src_net, struct net_device *dev,
@@ -645,14 +695,27 @@ static int xfrmi_newlink(struct net *src_net, struct net_device *dev,
 	int err;
 
 	xfrmi_netlink_parms(data, &p);
-	if (!p.if_id) {
-		NL_SET_ERR_MSG(extack, "if_id must be non zero");
-		return -EINVAL;
-	}
+	if (p.collect_md) {
+		struct xfrmi_net *xfrmn = net_generic(net, xfrmi_net_id);
 
-	xi = xfrmi_locate(net, &p);
-	if (xi)
-		return -EEXIST;
+		if (p.link || p.if_id) {
+			NL_SET_ERR_MSG(extack, "link and if_id must be zero");
+			return -EINVAL;
+		}
+
+		if (rtnl_dereference(xfrmn->collect_md_xfrmi))
+			return -EEXIST;
+
+	} else {
+		if (!p.if_id) {
+			NL_SET_ERR_MSG(extack, "if_id must be non zero");
+			return -EINVAL;
+		}
+
+		xi = xfrmi_locate(net, &p);
+		if (xi)
+			return -EEXIST;
+	}
 
 	xi = netdev_priv(dev);
 	xi->p = p;
@@ -682,12 +745,22 @@ static int xfrmi_changelink(struct net_device *dev, struct nlattr *tb[],
 		return -EINVAL;
 	}
 
+	if (p.collect_md) {
+		NL_SET_ERR_MSG(extack, "collect_md can't be changed");
+		return -EINVAL;
+	}
+
 	xi = xfrmi_locate(net, &p);
 	if (!xi) {
 		xi = netdev_priv(dev);
 	} else {
 		if (xi->dev != dev)
 			return -EEXIST;
+		if (xi->p.collect_md) {
+			NL_SET_ERR_MSG(extack,
+				       "device can't be changed to collect_md");
+			return -EINVAL;
+		}
 	}
 
 	return xfrmi_update(xi, &p);
@@ -700,6 +773,8 @@ static size_t xfrmi_get_size(const struct net_device *dev)
 		nla_total_size(4) +
 		/* IFLA_XFRM_IF_ID */
 		nla_total_size(4) +
+		/* IFLA_XFRM_COLLECT_METADATA */
+		nla_total_size(0) +
 		0;
 }
 
@@ -709,7 +784,8 @@ static int xfrmi_fill_info(struct sk_buff *skb, const struct net_device *dev)
 	struct xfrm_if_parms *parm = &xi->p;
 
 	if (nla_put_u32(skb, IFLA_XFRM_LINK, parm->link) ||
-	    nla_put_u32(skb, IFLA_XFRM_IF_ID, parm->if_id))
+	    nla_put_u32(skb, IFLA_XFRM_IF_ID, parm->if_id) ||
+	    (xi->p.collect_md && nla_put_flag(skb, IFLA_XFRM_COLLECT_METADATA)))
 		goto nla_put_failure;
 	return 0;
 
@@ -725,8 +801,10 @@ static struct net *xfrmi_get_link_net(const struct net_device *dev)
 }
 
 static const struct nla_policy xfrmi_policy[IFLA_XFRM_MAX + 1] = {
-	[IFLA_XFRM_LINK]	= { .type = NLA_U32 },
-	[IFLA_XFRM_IF_ID]	= { .type = NLA_U32 },
+	[IFLA_XFRM_UNSPEC]		= { .strict_start_type = IFLA_XFRM_COLLECT_METADATA },
+	[IFLA_XFRM_LINK]		= { .type = NLA_U32 },
+	[IFLA_XFRM_IF_ID]		= { .type = NLA_U32 },
+	[IFLA_XFRM_COLLECT_METADATA]	= { .type = NLA_FLAG },
 };
 
 static struct rtnl_link_ops xfrmi_link_ops __read_mostly = {
@@ -762,6 +840,9 @@ static void __net_exit xfrmi_exit_batch_net(struct list_head *net_exit_list)
 			     xip = &xi->next)
 				unregister_netdevice_queue(xi->dev, &list);
 		}
+		xi = rtnl_dereference(xfrmn->collect_md_xfrmi);
+		if (xi)
+			unregister_netdevice_queue(xi->dev, &list);
 	}
 	unregister_netdevice_many(&list);
 	rtnl_unlock();