Skip to contentSkip to conversation

pull request (net-next): ipsec-next 2013-02-14

PATCH New Steffen Klassert 2 participants 9 messages Feb 14, 2013

Subsystems (3)
lore
Share & tools
lore

Steffen Klassert

author maintainer
#0 lore raw

1) Remove a duplicated call to skb_orphan() in pf_key, from Cong Wang.

2) Prepare xfrm and pf_key for algorithms without pf_key support, from Jussi Kivilinna.

3) Fix an unbalanced lock in xfrm_output_one(), from Li RongQing.

4) Add an IPsec state resolution packet queue to handle packets that are send before the states are resolved.

5) xfrm4_policy_fini() is unused since 2.6.11, time to remove it. From Michal Kubecek.

6) The xfrm gc threshold was configurable just in the initial namespace, make it configurable in all namespaces. From Michal Kubecek.

7) We currently can not insert policies with mark and mask such that some flows would be matched from both policies. Allow this if the priorities of these policies are different, the one with the higher priority is used in this case.

Please pull or let me know if there are problems.

Thanks!

The following changes since commit cef401de7be8c4e155c6746bfccf721a4fa5fab9:

net: fix possible wrong checksum generation (2013-01-28 00:27:15 -0500)

are available in the git repository at:

git://git.kernel.org/pub/scm/linux/kernel/git/klassert/ipsec-next.git master

for you to fetch changes up to 7cb8a93968e395e40a72a50da0b6114e752304b4:

xfrm: Allow inserting policies with matching mark and different priorities (2013-02-11 14:07:01 +0100)

----------------------------------------------------------------
Cong Wang (1):
af_key: remove a duplicated skb_orphan()

Jussi Kivilinna (1):
pf_key/xfrm_algo: prepare pf_key and xfrm_algo for new algorithms without pfkey support

Li RongQing (1):
xfrm: fix a unbalanced lock

Michal Kubecek (2):
xfrm: remove unused xfrm4_policy_fini()
xfrm: make gc_thresh configurable in all namespaces

Steffen Klassert (2):
xfrm: Add a state resolution packet queue
xfrm: Allow inserting policies with matching mark and different priorities

 include/net/dst.h        |    1 +
 include/net/netns/ipv4.h |    1 +
 include/net/netns/ipv6.h |    1 +
 include/net/xfrm.h       |   12 ++-
 net/ipv4/xfrm4_policy.c  |   58 ++++++++---
 net/ipv6/xfrm6_policy.c  |   52 +++++++++-
 net/key/af_key.c         |   40 ++++++--
 net/xfrm/xfrm_algo.c     |   65 ++++++++++--
 net/xfrm/xfrm_output.c   |    2 +-
 net/xfrm/xfrm_policy.c   |  247 ++++++++++++++++++++++++++++++++++++++++++++--
 10 files changed, 438 insertions(+), 41 deletions(-)
author maintainer
#1 lore raw

PATCH 1/7 af_key: remove a duplicated skb_orphan()

From: Cong Wang <[email protected]>

skb_set_owner_r() will call skb_orphan(), I don't see any reason to call it twice.

Cc Steffen Klassert <[email protected]>
Cc Herbert Xu <[email protected]>
Cc "David S. Miller" <[email protected]>
---
 net/key/af_key.c |    1 -
 1 file changed, 1 deletion(-)
net/key/af_key.c +0 -1
index 5b426a6..6a6e684 100644
@@ -203,7 +203,6 @@ static int pfkey_broadcast_one(struct sk_buff *skb, struct sk_buff **skb2,
 	}
 	if (*skb2 != NULL) {
 		if (atomic_read(&sk->sk_rmem_alloc) <= sk->sk_rcvbuf) {
-			skb_orphan(*skb2);
 			skb_set_owner_r(*skb2, sk);
 			skb_queue_tail(&sk->sk_receive_queue, *skb2);
 			sk->sk_data_ready(sk, (*skb2)->len);
Signature · 1 line
1.7.9.5
author maintainer
#2 lore raw

PATCH 2/7 pf_key/xfrm_algo: prepare pf_key and xfrm_algo for new algorithms without pfkey support

In reply to Steffen Klassert

From: Jussi Kivilinna <[email protected]>

Mark existing algorithms as pfkey supported and make pfkey only use algorithms that have pfkey_supported set.

---
 include/net/xfrm.h   |    5 ++--
 net/key/af_key.c     |   39 +++++++++++++++++++++++-------
 net/xfrm/xfrm_algo.c |   65 +++++++++++++++++++++++++++++++++++++++++++++-----
 3 files changed, 93 insertions(+), 16 deletions(-)
include/net/xfrm.h +3 -2
index 421f764..814a1ba 100644
@@ -1320,6 +1320,7 @@ struct xfrm_algo_desc {
 	char *name;
 	char *compat;
 	u8 available:1;
+	u8 pfkey_supported:1;
 	union {
 		struct xfrm_algo_aead_info aead;
 		struct xfrm_algo_auth_info auth;
@@ -1561,8 +1562,8 @@ extern void xfrm_input_init(void);
 extern int xfrm_parse_spi(struct sk_buff *skb, u8 nexthdr, __be32 *spi, __be32 *seq);
 
 extern void xfrm_probe_algs(void);
-extern int xfrm_count_auth_supported(void);
-extern int xfrm_count_enc_supported(void);
+extern int xfrm_count_pfkey_auth_supported(void);
+extern int xfrm_count_pfkey_enc_supported(void);
 extern struct xfrm_algo_desc *xfrm_aalg_get_byidx(unsigned int idx);
 extern struct xfrm_algo_desc *xfrm_ealg_get_byidx(unsigned int idx);
 extern struct xfrm_algo_desc *xfrm_aalg_get_byid(int alg_id);
net/key/af_key.c +31 -8
index 6a6e684..cb75f9b 100644
@@ -815,18 +815,21 @@ static struct sk_buff *__pfkey_xfrm_state2msg(const struct xfrm_state *x,
 	sa->sadb_sa_auth = 0;
 	if (x->aalg) {
 		struct xfrm_algo_desc *a = xfrm_aalg_get_byname(x->aalg->alg_name, 0);
-		sa->sadb_sa_auth = a ? a->desc.sadb_alg_id : 0;
+		sa->sadb_sa_auth = (a && a->pfkey_supported) ?
+					a->desc.sadb_alg_id : 0;
 	}
 	sa->sadb_sa_encrypt = 0;
 	BUG_ON(x->ealg && x->calg);
 	if (x->ealg) {
 		struct xfrm_algo_desc *a = xfrm_ealg_get_byname(x->ealg->alg_name, 0);
-		sa->sadb_sa_encrypt = a ? a->desc.sadb_alg_id : 0;
+		sa->sadb_sa_encrypt = (a && a->pfkey_supported) ?
+					a->desc.sadb_alg_id : 0;
 	}
 	/* KAME compatible: sadb_sa_encrypt is overloaded with calg id */
 	if (x->calg) {
 		struct xfrm_algo_desc *a = xfrm_calg_get_byname(x->calg->alg_name, 0);
-		sa->sadb_sa_encrypt = a ? a->desc.sadb_alg_id : 0;
+		sa->sadb_sa_encrypt = (a && a->pfkey_supported) ?
+					a->desc.sadb_alg_id : 0;
 	}
 
 	sa->sadb_sa_flags = 0;
@@ -1137,7 +1140,7 @@ static struct xfrm_state * pfkey_msg2xfrm_state(struct net *net,
 	if (sa->sadb_sa_auth) {
 		int keysize = 0;
 		struct xfrm_algo_desc *a = xfrm_aalg_get_byid(sa->sadb_sa_auth);
-		if (!a) {
+		if (!a || !a->pfkey_supported) {
 			err = -ENOSYS;
 			goto out;
 		}
@@ -1159,7 +1162,7 @@ static struct xfrm_state * pfkey_msg2xfrm_state(struct net *net,
 	if (sa->sadb_sa_encrypt) {
 		if (hdr->sadb_msg_satype == SADB_X_SATYPE_IPCOMP) {
 			struct xfrm_algo_desc *a = xfrm_calg_get_byid(sa->sadb_sa_encrypt);
-			if (!a) {
+			if (!a || !a->pfkey_supported) {
 				err = -ENOSYS;
 				goto out;
 			}
@@ -1171,7 +1174,7 @@ static struct xfrm_state * pfkey_msg2xfrm_state(struct net *net,
 		} else {
 			int keysize = 0;
 			struct xfrm_algo_desc *a = xfrm_ealg_get_byid(sa->sadb_sa_encrypt);
-			if (!a) {
+			if (!a || !a->pfkey_supported) {
 				err = -ENOSYS;
 				goto out;
 			}
@@ -1577,13 +1580,13 @@ static struct sk_buff *compose_sadb_supported(const struct sadb_msg *orig,
 	struct sadb_msg *hdr;
 	int len, auth_len, enc_len, i;
 
-	auth_len = xfrm_count_auth_supported();
+	auth_len = xfrm_count_pfkey_auth_supported();
 	if (auth_len) {
 		auth_len *= sizeof(struct sadb_alg);
 		auth_len += sizeof(struct sadb_supported);
 	}
 
-	enc_len = xfrm_count_enc_supported();
+	enc_len = xfrm_count_pfkey_enc_supported();
 	if (enc_len) {
 		enc_len *= sizeof(struct sadb_alg);
 		enc_len += sizeof(struct sadb_supported);
@@ -1614,6 +1617,8 @@ static struct sk_buff *compose_sadb_supported(const struct sadb_msg *orig,
 			struct xfrm_algo_desc *aalg = xfrm_aalg_get_byidx(i);
 			if (!aalg)
 				break;
+			if (!aalg->pfkey_supported)
+				continue;
 			if (aalg->available)
 				*ap++ = aalg->desc;
 		}
@@ -1633,6 +1638,8 @@ static struct sk_buff *compose_sadb_supported(const struct sadb_msg *orig,
 			struct xfrm_algo_desc *ealg = xfrm_ealg_get_byidx(i);
 			if (!ealg)
 				break;
+			if (!ealg->pfkey_supported)
+				continue;
 			if (ealg->available)
 				*ap++ = ealg->desc;
 		}
@@ -2824,6 +2831,8 @@ static int count_ah_combs(const struct xfrm_tmpl *t)
 		const struct xfrm_algo_desc *aalg = xfrm_aalg_get_byidx(i);
 		if (!aalg)
 			break;
+		if (!aalg->pfkey_supported)
+			continue;
 		if (aalg_tmpl_set(t, aalg) && aalg->available)
 			sz += sizeof(struct sadb_comb);
 	}
@@ -2839,6 +2848,9 @@ static int count_esp_combs(const struct xfrm_tmpl *t)
 		if (!ealg)
 			break;
 
+		if (!ealg->pfkey_supported)
+			continue;
+
 		if (!(ealg_tmpl_set(t, ealg) && ealg->available))
 			continue;
 
@@ -2847,6 +2859,9 @@ static int count_esp_combs(const struct xfrm_tmpl *t)
 			if (!aalg)
 				break;
 
+			if (!aalg->pfkey_supported)
+				continue;
+
 			if (aalg_tmpl_set(t, aalg) && aalg->available)
 				sz += sizeof(struct sadb_comb);
 		}
@@ -2870,6 +2885,9 @@ static void dump_ah_combs(struct sk_buff *skb, const struct xfrm_tmpl *t)
 		if (!aalg)
 			break;
 
+		if (!aalg->pfkey_supported)
+			continue;
+
 		if (aalg_tmpl_set(t, aalg) && aalg->available) {
 			struct sadb_comb *c;
 			c = (struct sadb_comb*)skb_put(skb, sizeof(struct sadb_comb));
@@ -2902,6 +2920,9 @@ static void dump_esp_combs(struct sk_buff *skb, const struct xfrm_tmpl *t)
 		if (!ealg)
 			break;
 
+		if (!ealg->pfkey_supported)
+			continue;
+
 		if (!(ealg_tmpl_set(t, ealg) && ealg->available))
 			continue;
 
@@ -2910,6 +2931,8 @@ static void dump_esp_combs(struct sk_buff *skb, const struct xfrm_tmpl *t)
 			const struct xfrm_algo_desc *aalg = xfrm_aalg_get_byidx(k);
 			if (!aalg)
 				break;
+			if (!aalg->pfkey_supported)
+				continue;
 			if (!(aalg_tmpl_set(t, aalg) && aalg->available))
 				continue;
 			c = (struct sadb_comb*)skb_put(skb, sizeof(struct sadb_comb));
net/xfrm/xfrm_algo.c +59 -6

254 lines not on this page · raw message

Signature · 1 line
1.7.9.5
author maintainer
#3 lore raw

PATCH 3/7 xfrm: fix a unbalanced lock

In reply to Steffen Klassert

From: Li RongQing <[email protected]>

---
 net/xfrm/xfrm_output.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
net/xfrm/xfrm_output.c +1 -1
index 3670526..bcfda89 100644
@@ -64,7 +64,7 @@ static int xfrm_output_one(struct sk_buff *skb, int err)
 
 		if (unlikely(x->km.state != XFRM_STATE_VALID)) {
 			XFRM_INC_STATS(net, LINUX_MIB_XFRMOUTSTATEINVALID);
-			goto error_nolock;
+			goto error;
 		}
 
 		err = xfrm_state_check_expire(x);
Signature · 1 line
1.7.9.5
author maintainer
#4 lore raw

PATCH 4/7 xfrm: Add a state resolution packet queue

In reply to Steffen Klassert

As the default, we blackhole packets until the key manager resolves the states. This patch implements a packet queue where IPsec packets are queued until the states are resolved. We generate a dummy xfrm bundle, the output routine of the returned route enqueues the packet to a per policy queue and arms a timer that checks for state resolution when dst_output() is called. Once the states are resolved, the packets are sent out of the queue. If the states are not resolved after some time, the queue is flushed.

This patch keeps the defaut behaviour to blackhole packets as long as we have no states. To enable the packet queue the sysctl xfrm_larval_drop must be switched off.

---
 include/net/dst.h      |    1 +
 include/net/xfrm.h     |    7 ++
 net/xfrm/xfrm_policy.c |  229 +++++++++++++++++++++++++++++++++++++++++++++++-
 3 files changed, 233 insertions(+), 4 deletions(-)
include/net/dst.h +1 -0
index 9a78810..3da47e0 100644
@@ -61,6 +61,7 @@ struct dst_entry {
 #define DST_NOPEER		0x0040
 #define DST_FAKE_RTABLE		0x0080
 #define DST_XFRM_TUNNEL		0x0100
+#define DST_XFRM_QUEUE		0x0200
 
 	unsigned short		pending_confirm;
 
include/net/xfrm.h +7 -0
index 814a1ba..30f3e5b 100644
@@ -501,6 +501,12 @@ struct xfrm_policy_walk {
 	u32 seq;
 };
 
+struct xfrm_policy_queue {
+	struct sk_buff_head	hold_queue;
+	struct timer_list	hold_timer;
+	unsigned long		timeout;
+};
+
 struct xfrm_policy {
 #ifdef CONFIG_NET_NS
 	struct net		*xp_net;
@@ -522,6 +528,7 @@ struct xfrm_policy {
 	struct xfrm_lifetime_cfg lft;
 	struct xfrm_lifetime_cur curlft;
 	struct xfrm_policy_walk_entry walk;
+	struct xfrm_policy_queue polq;
 	u8			type;
 	u8			action;
 	u8			flags;
net/xfrm/xfrm_policy.c +225 -4

308 lines not on this page · raw message

Signature · 1 line
1.7.9.5
author maintainer
#5 lore raw

PATCH 5/7 xfrm: remove unused xfrm4_policy_fini()

In reply to Steffen Klassert

From: Michal Kubecek <[email protected]>

Function xfrm4_policy_fini() is unused since xfrm4_fini() was removed in 2.6.11.

---
 net/ipv4/xfrm4_policy.c |    9 ---------
 1 file changed, 9 deletions(-)
net/ipv4/xfrm4_policy.c +0 -9
index 3be0ac2..0e28383 100644
@@ -270,15 +270,6 @@ static void __init xfrm4_policy_init(void)
 	xfrm_policy_register_afinfo(&xfrm4_policy_afinfo);
 }
 
-static void __exit xfrm4_policy_fini(void)
-{
-#ifdef CONFIG_SYSCTL
-	if (sysctl_hdr)
-		unregister_net_sysctl_table(sysctl_hdr);
-#endif
-	xfrm_policy_unregister_afinfo(&xfrm4_policy_afinfo);
-}
-
 void __init xfrm4_init(void)
 {
 	dst_entries_init(&xfrm4_dst_ops);
Signature · 1 line
1.7.9.5
author maintainer
#6 lore raw

PATCH 6/7 xfrm: make gc_thresh configurable in all namespaces

In reply to Steffen Klassert

From: Michal Kubecek <[email protected]>

The xfrm gc threshold can be configured via xfrm{4,6}_gc_thresh sysctl but currently only in init_net, other namespaces always use the default value. This can substantially limit the number of IPsec tunnels that can be effectively used.

---
 include/net/netns/ipv4.h |    1 +
 include/net/netns/ipv6.h |    1 +
 net/ipv4/xfrm4_policy.c  |   49 ++++++++++++++++++++++++++++++++++++++++---
 net/ipv6/xfrm6_policy.c  |   52 +++++++++++++++++++++++++++++++++++++++++-----
 4 files changed, 95 insertions(+), 8 deletions(-)
include/net/netns/ipv4.h +1 -0
index 9b78862..2ba9de8 100644
@@ -22,6 +22,7 @@ struct netns_ipv4 {
 	struct ctl_table_header	*frags_hdr;
 	struct ctl_table_header	*ipv4_hdr;
 	struct ctl_table_header *route_hdr;
+	struct ctl_table_header *xfrm4_hdr;
 #endif
 	struct ipv4_devconf	*devconf_all;
 	struct ipv4_devconf	*devconf_dflt;
include/net/netns/ipv6.h +1 -0
index 214cb0a..1242f37 100644
@@ -16,6 +16,7 @@ struct netns_sysctl_ipv6 {
 	struct ctl_table_header *route_hdr;
 	struct ctl_table_header *icmp_hdr;
 	struct ctl_table_header *frags_hdr;
+	struct ctl_table_header *xfrm6_hdr;
 #endif
 	int bindv6only;
 	int flush_delay;
net/ipv4/xfrm4_policy.c +46 -3

61 lines not on this page · raw message

net/ipv6/xfrm6_policy.c +47 -5

70 lines not on this page · raw message

Signature · 1 line
1.7.9.5
author maintainer
#7 lore raw

PATCH 7/7 xfrm: Allow inserting policies with matching mark and different priorities

In reply to Steffen Klassert

We currently can not insert policies with mark and mask such that some flows would be matched from both policies. We make this possible when the priority of these policies are different. If both policies match a flow, the one with the higher priority is used.

---
 net/xfrm/xfrm_policy.c |   18 ++++++++++++++++--
 1 file changed, 16 insertions(+), 2 deletions(-)
net/xfrm/xfrm_policy.c +16 -2
index 456b11b..257dfb1 100644
@@ -607,6 +607,21 @@ static void xfrm_policy_requeue(struct xfrm_policy *old,
 	spin_unlock_bh(&pq->hold_queue.lock);
 }
 
+static bool xfrm_policy_mark_match(struct xfrm_policy *policy,
+				   struct xfrm_policy *pol)
+{
+	u32 mark = policy->mark.v & policy->mark.m;
+
+	if (policy->mark.v == pol->mark.v && policy->mark.m == pol->mark.m)
+		return true;
+
+	if ((mark & pol->mark.m) == pol->mark.v &&
+	    policy->priority == pol->priority)
+		return true;
+
+	return false;
+}
+
 int xfrm_policy_insert(int dir, struct xfrm_policy *policy, int excl)
 {
 	struct net *net = xp_net(policy);
@@ -614,7 +629,6 @@ int xfrm_policy_insert(int dir, struct xfrm_policy *policy, int excl)
 	struct xfrm_policy *delpol;
 	struct hlist_head *chain;
 	struct hlist_node *entry, *newpos;
-	u32 mark = policy->mark.v & policy->mark.m;
 
 	write_lock_bh(&xfrm_policy_lock);
 	chain = policy_hash_bysel(net, &policy->selector, policy->family, dir);
@@ -623,7 +637,7 @@ int xfrm_policy_insert(int dir, struct xfrm_policy *policy, int excl)
 	hlist_for_each_entry(pol, entry, chain, bydst) {
 		if (pol->type == policy->type &&
 		    !selector_cmp(&pol->selector, &policy->selector) &&
-		    (mark & pol->mark.m) == pol->mark.v &&
+		    xfrm_policy_mark_match(policy, pol) &&
 		    xfrm_sec_ctx_match(pol->security, policy->security) &&
 		    !WARN_ON(delpol)) {
 			if (excl) {
Signature · 1 line
1.7.9.5
maintainer
#8 lore raw
In reply to Steffen Klassert

From: Steffen Klassert <[email protected]>
Date: Thu, 14 Feb 2013 12:47:20 +0100

Quoted 17 lines

1) Remove a duplicated call to skb_orphan() in pf_key, from Cong Wang.

2) Prepare xfrm and pf_key for algorithms without pf_key support, from Jussi Kivilinna.

3) Fix an unbalanced lock in xfrm_output_one(), from Li RongQing.

4) Add an IPsec state resolution packet queue to handle packets that are send before the states are resolved.

5) xfrm4_policy_fini() is unused since 2.6.11, time to remove it. From Michal Kubecek.

6) The xfrm gc threshold was configurable just in the initial namespace, make it configurable in all namespaces. From Michal Kubecek.

Quoted 6 lines

7) We currently can not insert policies with mark and mask such that some flows would be matched from both policies. Allow this if the priorities of these policies are different, the one with the higher priority is used in this case.

Please pull or let me know if there are problems.

Pulled, thanks Steffen.