[PATCH net-next 00/11] Netfilter updates for net-next

[PATCH net-next 00/11] Netfilter updates for net-next

[Pablo Neira Ayuso]

Hi,

The following patchset contains Netfilter updates for net-next:

1) Replace unnecessary list_for_each_entry_continue() in nf_tables,
   from Jakob Koschel.

2) Add struct nf_conntrack_net_ecache to conntrack event cache and
   use it, from Florian Westphal.

3) Refactor ctnetlink_dump_list(), also from Florian.

4) Bump module reference counter on cttimeout object addition/removal,
   from Florian.

5) Consolidate nf_log MAC printer, from Phil Sutter.

6) Add basic logging support for unknown ethertype, from Phil Sutter.

7) Consolidate check for sysctl nf_log_all_netns toggle, also from Phil.

8) Replace hardcode value in nft_bitwise, from Jeremy Sowden.

9) Rename BASIC-like goto tags in nft_bitwise to more meaningful names,
   also from Jeremy.

10) nft_fib support for reverse path filtering with policy-based routing
    on iif. Extend selftests to cover for this new usecase, from Florian.

Please, pull these changes from:

  git://git.kernel.org/pub/scm/linux/kernel/git/netfilter/nf-next.git

Thanks.

----------------------------------------------------------------

The following changes since commit 2975dbdc3989cd66a4cb5a7c5510de2de8ee4d14:

  Merge tag 'net-5.18-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net (2022-03-31 11:23:31 -0700)

are available in the Git repository at:

  git://git.kernel.org/pub/scm/linux/kernel/git/netfilter/nf-next.git HEAD

for you to fetch changes up to 0c7b27616fbd64b3b86c59ad5441f82a1a0c4176:

  selftests: netfilter: add fib expression forward test case (2022-04-11 12:10:09 +0200)

----------------------------------------------------------------
Florian Westphal (4):
      netfilter: ecache: move to separate structure
      netfilter: conntrack: split inner loop of list dumping to own function
      netfilter: cttimeout: inc/dec module refcount per object, not per use refcount
      selftests: netfilter: add fib expression forward test case

Jakob Koschel (1):
      netfilter: nf_tables: replace unnecessary use of list_for_each_entry_continue()

Jeremy Sowden (2):
      netfilter: bitwise: replace hard-coded size with `sizeof` expression
      netfilter: bitwise: improve error goto labels

Pablo Neira Ayuso (1):
      netfilter: nft_fib: reverse path filter for policy-based routing on iif

Phil Sutter (3):
      netfilter: nf_log_syslog: Merge MAC header dumpers
      netfilter: nf_log_syslog: Don't ignore unknown protocols
      netfilter: nf_log_syslog: Consolidate entry checks

 include/net/netfilter/nf_conntrack.h         |   8 +-
 net/ipv4/netfilter/nft_fib_ipv4.c            |   4 +
 net/ipv6/netfilter/nft_fib_ipv6.c            |   4 +
 net/netfilter/nf_conntrack_ecache.c          |  19 ++--
 net/netfilter/nf_conntrack_netlink.c         |  68 +++++++++-----
 net/netfilter/nf_log_syslog.c                | 136 +++++++++++++--------------
 net/netfilter/nf_tables_api.c                |   6 +-
 net/netfilter/nfnetlink_cttimeout.c          |  14 +--
 net/netfilter/nft_bitwise.c                  |  13 +--
 net/netfilter/nft_fib.c                      |   4 +
 tools/testing/selftests/netfilter/nft_fib.sh |  50 ++++++++++
 11 files changed, 199 insertions(+), 127 deletions(-)

[PATCH net-next 01/11] netfilter: nf_tables: replace unnecessary use of list_for_each_entry_continue()

[Pablo Neira Ayuso]

From: Jakob Koschel <jakobkoschel@gmail.com>

Since there is no way for list_for_each_entry_continue() to start
interating in the middle of the list they can be replaced with a call
to list_for_each_entry().

In preparation to limit the scope of the list iterator to the list
traversal loop, the list iterator variable 'rule' should not be used
past the loop.

v1->v2:
- also replace first usage of list_for_each_entry_continue() (Florian
Westphal)

Signed-off-by: Jakob Koschel <jakobkoschel@gmail.com>
Reviewed-by: Florian Westphal <fw@strlen.de>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
---
 net/netfilter/nf_tables_api.c | 6 ++----
 1 file changed, 2 insertions(+), 4 deletions(-)

diff --git a/net/netfilter/nf_tables_api.c b/net/netfilter/nf_tables_api.c
index 5ddfdb2adaf1..060aa56e54d9 100644
--- a/net/netfilter/nf_tables_api.c
+++ b/net/netfilter/nf_tables_api.c
@@ -8367,10 +8367,8 @@ static int nf_tables_commit_chain_prepare(struct net *net, struct nft_chain *cha
 	if (chain->blob_next || !nft_is_active_next(net, chain))
 		return 0;
 
-	rule = list_entry(&chain->rules, struct nft_rule, list);
-
 	data_size = 0;
-	list_for_each_entry_continue(rule, &chain->rules, list) {
+	list_for_each_entry(rule, &chain->rules, list) {
 		if (nft_is_active_next(net, rule)) {
 			data_size += sizeof(*prule) + rule->dlen;
 			if (data_size > INT_MAX)
@@ -8387,7 +8385,7 @@ static int nf_tables_commit_chain_prepare(struct net *net, struct nft_chain *cha
 	data_boundary = data + data_size;
 	size = 0;
 
-	list_for_each_entry_continue(rule, &chain->rules, list) {
+	list_for_each_entry(rule, &chain->rules, list) {
 		if (!nft_is_active_next(net, rule))
 			continue;
 
-- 
2.30.2

Re: [PATCH net-next 01/11] netfilter: nf_tables: replace unnecessary use of list_for_each_entry_continue()

[patchwork-bot+netdevbpf]

Hello:

This series was applied to netdev/net-next.git (master)
by Pablo Neira Ayuso <pablo@netfilter.org>:

On Mon, 11 Apr 2022 12:27:34 +0200 you wrote:
> From: Jakob Koschel <jakobkoschel@gmail.com>
> 
> Since there is no way for list_for_each_entry_continue() to start
> interating in the middle of the list they can be replaced with a call
> to list_for_each_entry().
> 
> In preparation to limit the scope of the list iterator to the list
> traversal loop, the list iterator variable 'rule' should not be used
> past the loop.
> 
> [...]

Here is the summary with links:
  - [net-next,01/11] netfilter: nf_tables: replace unnecessary use of list_for_each_entry_continue()
    https://git.kernel.org/netdev/net-next/c/10377d42281e
  - [net-next,02/11] netfilter: ecache: move to separate structure
    https://git.kernel.org/netdev/net-next/c/9027ce0b071a
  - [net-next,03/11] netfilter: conntrack: split inner loop of list dumping to own function
    https://git.kernel.org/netdev/net-next/c/49001a2e83a8
  - [net-next,04/11] netfilter: cttimeout: inc/dec module refcount per object, not per use refcount
    https://git.kernel.org/netdev/net-next/c/523895e5b278
  - [net-next,05/11] netfilter: nf_log_syslog: Merge MAC header dumpers
    https://git.kernel.org/netdev/net-next/c/39ab798fc14d
  - [net-next,06/11] netfilter: nf_log_syslog: Don't ignore unknown protocols
    https://git.kernel.org/netdev/net-next/c/0c8783806f63
  - [net-next,07/11] netfilter: nf_log_syslog: Consolidate entry checks
    https://git.kernel.org/netdev/net-next/c/c3e348666713
  - [net-next,08/11] netfilter: bitwise: replace hard-coded size with `sizeof` expression
    https://git.kernel.org/netdev/net-next/c/c70b921fc1e8
  - [net-next,09/11] netfilter: bitwise: improve error goto labels
    https://git.kernel.org/netdev/net-next/c/00bd435208e5
  - [net-next,10/11] netfilter: nft_fib: reverse path filter for policy-based routing on iif
    https://git.kernel.org/netdev/net-next/c/be8be04e5ddb
  - [net-next,11/11] selftests: netfilter: add fib expression forward test case
    https://git.kernel.org/netdev/net-next/c/0c7b27616fbd

You are awesome, thank you!
-- 
Deet-doot-dot, I am a bot.
https://korg.docs.kernel.org/patchwork/pwbot.html

[PATCH net-next 02/11] netfilter: ecache: move to separate structure

[Pablo Neira Ayuso]

From: Florian Westphal <fw@strlen.de>

This makes it easier for a followup patch to only expose ecache
related parts of nf_conntrack_net structure.

Signed-off-by: Florian Westphal <fw@strlen.de>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
---
 include/net/netfilter/nf_conntrack.h |  8 ++++++--
 net/netfilter/nf_conntrack_ecache.c  | 19 ++++++++++---------
 2 files changed, 16 insertions(+), 11 deletions(-)

diff --git a/include/net/netfilter/nf_conntrack.h b/include/net/netfilter/nf_conntrack.h
index b08b70989d2c..69e6c6a218be 100644
--- a/include/net/netfilter/nf_conntrack.h
+++ b/include/net/netfilter/nf_conntrack.h
@@ -43,6 +43,11 @@ union nf_conntrack_expect_proto {
 	/* insert expect proto private data here */
 };
 
+struct nf_conntrack_net_ecache {
+	struct delayed_work dwork;
+	struct netns_ct *ct_net;
+};
+
 struct nf_conntrack_net {
 	/* only used when new connection is allocated: */
 	atomic_t count;
@@ -58,8 +63,7 @@ struct nf_conntrack_net {
 	struct ctl_table_header	*sysctl_header;
 #endif
 #ifdef CONFIG_NF_CONNTRACK_EVENTS
-	struct delayed_work ecache_dwork;
-	struct netns_ct *ct_net;
+	struct nf_conntrack_net_ecache ecache;
 #endif
 };
 
diff --git a/net/netfilter/nf_conntrack_ecache.c b/net/netfilter/nf_conntrack_ecache.c
index 07e65b4e92f8..0cb2da0a759a 100644
--- a/net/netfilter/nf_conntrack_ecache.c
+++ b/net/netfilter/nf_conntrack_ecache.c
@@ -96,8 +96,8 @@ static enum retry_state ecache_work_evict_list(struct ct_pcpu *pcpu)
 
 static void ecache_work(struct work_struct *work)
 {
-	struct nf_conntrack_net *cnet = container_of(work, struct nf_conntrack_net, ecache_dwork.work);
-	struct netns_ct *ctnet = cnet->ct_net;
+	struct nf_conntrack_net *cnet = container_of(work, struct nf_conntrack_net, ecache.dwork.work);
+	struct netns_ct *ctnet = cnet->ecache.ct_net;
 	int cpu, delay = -1;
 	struct ct_pcpu *pcpu;
 
@@ -127,7 +127,7 @@ static void ecache_work(struct work_struct *work)
 
 	ctnet->ecache_dwork_pending = delay > 0;
 	if (delay >= 0)
-		schedule_delayed_work(&cnet->ecache_dwork, delay);
+		schedule_delayed_work(&cnet->ecache.dwork, delay);
 }
 
 static int __nf_conntrack_eventmask_report(struct nf_conntrack_ecache *e,
@@ -293,12 +293,12 @@ void nf_conntrack_ecache_work(struct net *net, enum nf_ct_ecache_state state)
 	struct nf_conntrack_net *cnet = nf_ct_pernet(net);
 
 	if (state == NFCT_ECACHE_DESTROY_FAIL &&
-	    !delayed_work_pending(&cnet->ecache_dwork)) {
-		schedule_delayed_work(&cnet->ecache_dwork, HZ);
+	    !delayed_work_pending(&cnet->ecache.dwork)) {
+		schedule_delayed_work(&cnet->ecache.dwork, HZ);
 		net->ct.ecache_dwork_pending = true;
 	} else if (state == NFCT_ECACHE_DESTROY_SENT) {
 		net->ct.ecache_dwork_pending = false;
-		mod_delayed_work(system_wq, &cnet->ecache_dwork, 0);
+		mod_delayed_work(system_wq, &cnet->ecache.dwork, 0);
 	}
 }
 
@@ -310,8 +310,9 @@ void nf_conntrack_ecache_pernet_init(struct net *net)
 	struct nf_conntrack_net *cnet = nf_ct_pernet(net);
 
 	net->ct.sysctl_events = nf_ct_events;
-	cnet->ct_net = &net->ct;
-	INIT_DELAYED_WORK(&cnet->ecache_dwork, ecache_work);
+
+	cnet->ecache.ct_net = &net->ct;
+	INIT_DELAYED_WORK(&cnet->ecache.dwork, ecache_work);
 
 	BUILD_BUG_ON(__IPCT_MAX >= 16);	/* e->ctmask is u16 */
 }
@@ -320,5 +321,5 @@ void nf_conntrack_ecache_pernet_fini(struct net *net)
 {
 	struct nf_conntrack_net *cnet = nf_ct_pernet(net);
 
-	cancel_delayed_work_sync(&cnet->ecache_dwork);
+	cancel_delayed_work_sync(&cnet->ecache.dwork);
 }
-- 
2.30.2

[PATCH net-next 03/11] netfilter: conntrack: split inner loop of list dumping to own function

[Pablo Neira Ayuso]

From: Florian Westphal <fw@strlen.de>

This allows code re-use in the followup patch.
No functional changes intended.

Signed-off-by: Florian Westphal <fw@strlen.de>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
---
 net/netfilter/nf_conntrack_netlink.c | 68 ++++++++++++++++++----------
 1 file changed, 43 insertions(+), 25 deletions(-)

diff --git a/net/netfilter/nf_conntrack_netlink.c b/net/netfilter/nf_conntrack_netlink.c
index 1ea2ad732d57..924d766e6c53 100644
--- a/net/netfilter/nf_conntrack_netlink.c
+++ b/net/netfilter/nf_conntrack_netlink.c
@@ -1708,6 +1708,47 @@ static int ctnetlink_done_list(struct netlink_callback *cb)
 	return 0;
 }
 
+static int ctnetlink_dump_one_entry(struct sk_buff *skb,
+				    struct netlink_callback *cb,
+				    struct nf_conn *ct,
+				    bool dying)
+{
+	struct ctnetlink_list_dump_ctx *ctx = (void *)cb->ctx;
+	struct nfgenmsg *nfmsg = nlmsg_data(cb->nlh);
+	u8 l3proto = nfmsg->nfgen_family;
+	int res;
+
+	if (l3proto && nf_ct_l3num(ct) != l3proto)
+		return 0;
+
+	if (ctx->last) {
+		if (ct != ctx->last)
+			return 0;
+
+		ctx->last = NULL;
+	}
+
+	/* We can't dump extension info for the unconfirmed
+	 * list because unconfirmed conntracks can have
+	 * ct->ext reallocated (and thus freed).
+	 *
+	 * In the dying list case ct->ext can't be free'd
+	 * until after we drop pcpu->lock.
+	 */
+	res = ctnetlink_fill_info(skb, NETLINK_CB(cb->skb).portid,
+				  cb->nlh->nlmsg_seq,
+				  NFNL_MSG_TYPE(cb->nlh->nlmsg_type),
+				  ct, dying, 0);
+	if (res < 0) {
+		if (!refcount_inc_not_zero(&ct->ct_general.use))
+			return 0;
+
+		ctx->last = ct;
+	}
+
+	return res;
+}
+
 static int
 ctnetlink_dump_list(struct sk_buff *skb, struct netlink_callback *cb, bool dying)
 {
@@ -1715,12 +1756,9 @@ ctnetlink_dump_list(struct sk_buff *skb, struct netlink_callback *cb, bool dying
 	struct nf_conn *ct, *last;
 	struct nf_conntrack_tuple_hash *h;
 	struct hlist_nulls_node *n;
-	struct nfgenmsg *nfmsg = nlmsg_data(cb->nlh);
-	u_int8_t l3proto = nfmsg->nfgen_family;
-	int res;
-	int cpu;
 	struct hlist_nulls_head *list;
 	struct net *net = sock_net(skb->sk);
+	int res, cpu;
 
 	if (ctx->done)
 		return 0;
@@ -1739,30 +1777,10 @@ ctnetlink_dump_list(struct sk_buff *skb, struct netlink_callback *cb, bool dying
 restart:
 		hlist_nulls_for_each_entry(h, n, list, hnnode) {
 			ct = nf_ct_tuplehash_to_ctrack(h);
-			if (l3proto && nf_ct_l3num(ct) != l3proto)
-				continue;
-			if (ctx->last) {
-				if (ct != last)
-					continue;
-				ctx->last = NULL;
-			}
 
-			/* We can't dump extension info for the unconfirmed
-			 * list because unconfirmed conntracks can have
-			 * ct->ext reallocated (and thus freed).
-			 *
-			 * In the dying list case ct->ext can't be free'd
-			 * until after we drop pcpu->lock.
-			 */
-			res = ctnetlink_fill_info(skb, NETLINK_CB(cb->skb).portid,
-						  cb->nlh->nlmsg_seq,
-						  NFNL_MSG_TYPE(cb->nlh->nlmsg_type),
-						  ct, dying, 0);
+			res = ctnetlink_dump_one_entry(skb, cb, ct, dying);
 			if (res < 0) {
-				if (!refcount_inc_not_zero(&ct->ct_general.use))
-					continue;
 				ctx->cpu = cpu;
-				ctx->last = ct;
 				spin_unlock_bh(&pcpu->lock);
 				goto out;
 			}
-- 
2.30.2

[PATCH net-next 04/11] netfilter: cttimeout: inc/dec module refcount per object, not per use refcount

[Pablo Neira Ayuso]

From: Florian Westphal <fw@strlen.de>

There is no need to increment the module refcount again, its enough to
obtain one reference per object, i.e. take a reference on object
creation and put it on object destruction.

Signed-off-by: Florian Westphal <fw@strlen.de>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
---
 net/netfilter/nfnetlink_cttimeout.c | 14 +++++---------
 1 file changed, 5 insertions(+), 9 deletions(-)

diff --git a/net/netfilter/nfnetlink_cttimeout.c b/net/netfilter/nfnetlink_cttimeout.c
index b0d8888a539b..eea486f32971 100644
--- a/net/netfilter/nfnetlink_cttimeout.c
+++ b/net/netfilter/nfnetlink_cttimeout.c
@@ -158,6 +158,7 @@ static int cttimeout_new_timeout(struct sk_buff *skb,
 	timeout->timeout.l3num = l3num;
 	timeout->timeout.l4proto = l4proto;
 	refcount_set(&timeout->refcnt, 1);
+	__module_get(THIS_MODULE);
 	list_add_tail_rcu(&timeout->head, &pernet->nfct_timeout_list);
 
 	return 0;
@@ -506,13 +507,8 @@ static struct nf_ct_timeout *ctnl_timeout_find_get(struct net *net,
 		if (strncmp(timeout->name, name, CTNL_TIMEOUT_NAME_MAX) != 0)
 			continue;
 
-		if (!try_module_get(THIS_MODULE))
+		if (!refcount_inc_not_zero(&timeout->refcnt))
 			goto err;
-
-		if (!refcount_inc_not_zero(&timeout->refcnt)) {
-			module_put(THIS_MODULE);
-			goto err;
-		}
 		matching = timeout;
 		break;
 	}
@@ -525,10 +521,10 @@ static void ctnl_timeout_put(struct nf_ct_timeout *t)
 	struct ctnl_timeout *timeout =
 		container_of(t, struct ctnl_timeout, timeout);
 
-	if (refcount_dec_and_test(&timeout->refcnt))
+	if (refcount_dec_and_test(&timeout->refcnt)) {
 		kfree_rcu(timeout, rcu_head);
-
-	module_put(THIS_MODULE);
+		module_put(THIS_MODULE);
+	}
 }
 
 static const struct nfnl_callback cttimeout_cb[IPCTNL_MSG_TIMEOUT_MAX] = {
-- 
2.30.2

[PATCH net-next 05/11] netfilter: nf_log_syslog: Merge MAC header dumpers

[Pablo Neira Ayuso]

From: Phil Sutter <phil@nwl.cc>

The functions for IPv4 and IPv6 were almost identical apart from extra
SIT tunnel device handling in the latter.

Signed-off-by: Phil Sutter <phil@nwl.cc>
Acked-by: Florian Westphal <fw@strlen.de>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
---
 net/netfilter/nf_log_syslog.c | 91 ++++++++++-------------------------
 1 file changed, 25 insertions(+), 66 deletions(-)

diff --git a/net/netfilter/nf_log_syslog.c b/net/netfilter/nf_log_syslog.c
index 13234641cdb3..d1dcf36545d7 100644
--- a/net/netfilter/nf_log_syslog.c
+++ b/net/netfilter/nf_log_syslog.c
@@ -766,9 +766,9 @@ dump_ipv6_packet(struct net *net, struct nf_log_buf *m,
 		nf_log_buf_add(m, "MARK=0x%x ", skb->mark);
 }
 
-static void dump_ipv4_mac_header(struct nf_log_buf *m,
-				 const struct nf_loginfo *info,
-				 const struct sk_buff *skb)
+static void dump_mac_header(struct nf_log_buf *m,
+			    const struct nf_loginfo *info,
+			    const struct sk_buff *skb)
 {
 	struct net_device *dev = skb->dev;
 	unsigned int logflags = 0;
@@ -798,9 +798,26 @@ static void dump_ipv4_mac_header(struct nf_log_buf *m,
 		const unsigned char *p = skb_mac_header(skb);
 		unsigned int i;
 
-		nf_log_buf_add(m, "%02x", *p++);
-		for (i = 1; i < dev->hard_header_len; i++, p++)
-			nf_log_buf_add(m, ":%02x", *p);
+		if (dev->type == ARPHRD_SIT) {
+			p -= ETH_HLEN;
+
+			if (p < skb->head)
+				p = NULL;
+		}
+
+		if (p) {
+			nf_log_buf_add(m, "%02x", *p++);
+			for (i = 1; i < dev->hard_header_len; i++)
+				nf_log_buf_add(m, ":%02x", *p++);
+		}
+
+		if (dev->type == ARPHRD_SIT) {
+			const struct iphdr *iph =
+				(struct iphdr *)skb_mac_header(skb);
+
+			nf_log_buf_add(m, " TUNNEL=%pI4->%pI4", &iph->saddr,
+				       &iph->daddr);
+		}
 	}
 	nf_log_buf_add(m, " ");
 }
@@ -827,7 +844,7 @@ static void nf_log_ip_packet(struct net *net, u_int8_t pf,
 				  out, loginfo, prefix);
 
 	if (in)
-		dump_ipv4_mac_header(m, loginfo, skb);
+		dump_mac_header(m, loginfo, skb);
 
 	dump_ipv4_packet(net, m, loginfo, skb, 0);
 
@@ -841,64 +858,6 @@ static struct nf_logger nf_ip_logger __read_mostly = {
 	.me		= THIS_MODULE,
 };
 
-static void dump_ipv6_mac_header(struct nf_log_buf *m,
-				 const struct nf_loginfo *info,
-				 const struct sk_buff *skb)
-{
-	struct net_device *dev = skb->dev;
-	unsigned int logflags = 0;
-
-	if (info->type == NF_LOG_TYPE_LOG)
-		logflags = info->u.log.logflags;
-
-	if (!(logflags & NF_LOG_MACDECODE))
-		goto fallback;
-
-	switch (dev->type) {
-	case ARPHRD_ETHER:
-		nf_log_buf_add(m, "MACSRC=%pM MACDST=%pM ",
-			       eth_hdr(skb)->h_source, eth_hdr(skb)->h_dest);
-		nf_log_dump_vlan(m, skb);
-		nf_log_buf_add(m, "MACPROTO=%04x ",
-			       ntohs(eth_hdr(skb)->h_proto));
-		return;
-	default:
-		break;
-	}
-
-fallback:
-	nf_log_buf_add(m, "MAC=");
-	if (dev->hard_header_len &&
-	    skb->mac_header != skb->network_header) {
-		const unsigned char *p = skb_mac_header(skb);
-		unsigned int len = dev->hard_header_len;
-		unsigned int i;
-
-		if (dev->type == ARPHRD_SIT) {
-			p -= ETH_HLEN;
-
-			if (p < skb->head)
-				p = NULL;
-		}
-
-		if (p) {
-			nf_log_buf_add(m, "%02x", *p++);
-			for (i = 1; i < len; i++)
-				nf_log_buf_add(m, ":%02x", *p++);
-		}
-		nf_log_buf_add(m, " ");
-
-		if (dev->type == ARPHRD_SIT) {
-			const struct iphdr *iph =
-				(struct iphdr *)skb_mac_header(skb);
-			nf_log_buf_add(m, "TUNNEL=%pI4->%pI4 ", &iph->saddr,
-				       &iph->daddr);
-		}
-	} else {
-		nf_log_buf_add(m, " ");
-	}
-}
-
 static void nf_log_ip6_packet(struct net *net, u_int8_t pf,
 			      unsigned int hooknum, const struct sk_buff *skb,
 			      const struct net_device *in,
@@ -921,7 +880,7 @@ static void nf_log_ip6_packet(struct net *net, u_int8_t pf,
 				  loginfo, prefix);
 
 	if (in)
-		dump_ipv6_mac_header(m, loginfo, skb);
+		dump_mac_header(m, loginfo, skb);
 
 	dump_ipv6_packet(net, m, loginfo, skb, skb_network_offset(skb), 1);
 
-- 
2.30.2

[PATCH net-next 06/11] netfilter: nf_log_syslog: Don't ignore unknown protocols

[Pablo Neira Ayuso]

From: Phil Sutter <phil@nwl.cc>

With netdev and bridge nfprotos, loggers may see arbitrary ethernet
frames. Print at least basic info like interfaces and MAC header data.

Signed-off-by: Phil Sutter <phil@nwl.cc>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
---
 net/netfilter/nf_log_syslog.c | 31 +++++++++++++++++++++++++++++++
 1 file changed, 31 insertions(+)

diff --git a/net/netfilter/nf_log_syslog.c b/net/netfilter/nf_log_syslog.c
index d1dcf36545d7..a7ff6fdbafc9 100644
--- a/net/netfilter/nf_log_syslog.c
+++ b/net/netfilter/nf_log_syslog.c
@@ -894,6 +894,33 @@ static struct nf_logger nf_ip6_logger __read_mostly = {
 	.me		= THIS_MODULE,
 };
 
+static void nf_log_unknown_packet(struct net *net, u_int8_t pf,
+				  unsigned int hooknum,
+				  const struct sk_buff *skb,
+				  const struct net_device *in,
+				  const struct net_device *out,
+				  const struct nf_loginfo *loginfo,
+				  const char *prefix)
+{
+	struct nf_log_buf *m;
+
+	/* FIXME: Disabled from containers until syslog ns is supported */
+	if (!net_eq(net, &init_net) && !sysctl_nf_log_all_netns)
+		return;
+
+	m = nf_log_buf_open();
+
+	if (!loginfo)
+		loginfo = &default_loginfo;
+
+	nf_log_dump_packet_common(m, pf, hooknum, skb, in, out, loginfo,
+				  prefix);
+
+	dump_mac_header(m, loginfo, skb);
+
+	nf_log_buf_close(m);
+}
+
 static void nf_log_netdev_packet(struct net *net, u_int8_t pf,
 				 unsigned int hooknum,
 				 const struct sk_buff *skb,
@@ -913,6 +940,10 @@ static void nf_log_netdev_packet(struct net *net, u_int8_t pf,
 	case htons(ETH_P_RARP):
 		nf_log_arp_packet(net, pf, hooknum, skb, in, out, loginfo, prefix);
 		break;
+	default:
+		nf_log_unknown_packet(net, pf, hooknum, skb,
+				      in, out, loginfo, prefix);
+		break;
 	}
 }
 
-- 
2.30.2

[PATCH net-next 07/11] netfilter: nf_log_syslog: Consolidate entry checks

[Pablo Neira Ayuso]

From: Phil Sutter <phil@nwl.cc>

Every syslog logging callback has to perform the same check to cover for
rogue containers, introduce a helper for clarity. Drop the FIXME as
there is a viable solution since commit 2851940ffee31 ("netfilter: allow
logging from non-init namespaces").

Suggested-by: Florian Westphal <fw@strlen.de>
Signed-off-by: Phil Sutter <phil@nwl.cc>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
---
 net/netfilter/nf_log_syslog.c | 18 ++++++++++--------
 1 file changed, 10 insertions(+), 8 deletions(-)

diff --git a/net/netfilter/nf_log_syslog.c b/net/netfilter/nf_log_syslog.c
index a7ff6fdbafc9..77bcb10fc586 100644
--- a/net/netfilter/nf_log_syslog.c
+++ b/net/netfilter/nf_log_syslog.c
@@ -40,6 +40,12 @@ struct arppayload {
 	unsigned char ip_dst[4];
 };
 
+/* Guard against containers flooding syslog. */
+static bool nf_log_allowed(const struct net *net)
+{
+	return net_eq(net, &init_net) || sysctl_nf_log_all_netns;
+}
+
 static void nf_log_dump_vlan(struct nf_log_buf *m, const struct sk_buff *skb)
 {
 	u16 vid;
@@ -133,8 +139,7 @@ static void nf_log_arp_packet(struct net *net, u_int8_t pf,
 {
 	struct nf_log_buf *m;
 
-	/* FIXME: Disabled from containers until syslog ns is supported */
-	if (!net_eq(net, &init_net) && !sysctl_nf_log_all_netns)
+	if (!nf_log_allowed(net))
 		return;
 
 	m = nf_log_buf_open();
@@ -831,8 +836,7 @@ static void nf_log_ip_packet(struct net *net, u_int8_t pf,
 {
 	struct nf_log_buf *m;
 
-	/* FIXME: Disabled from containers until syslog ns is supported */
-	if (!net_eq(net, &init_net) && !sysctl_nf_log_all_netns)
+	if (!nf_log_allowed(net))
 		return;
 
 	m = nf_log_buf_open();
@@ -867,8 +871,7 @@ static void nf_log_ip6_packet(struct net *net, u_int8_t pf,
 {
 	struct nf_log_buf *m;
 
-	/* FIXME: Disabled from containers until syslog ns is supported */
-	if (!net_eq(net, &init_net) && !sysctl_nf_log_all_netns)
+	if (!nf_log_allowed(net))
 		return;
 
 	m = nf_log_buf_open();
@@ -904,8 +907,7 @@ static void nf_log_unknown_packet(struct net *net, u_int8_t pf,
 {
 	struct nf_log_buf *m;
 
-	/* FIXME: Disabled from containers until syslog ns is supported */
-	if (!net_eq(net, &init_net) && !sysctl_nf_log_all_netns)
+	if (!nf_log_allowed(net))
 		return;
 
 	m = nf_log_buf_open();
-- 
2.30.2

[PATCH net-next 08/11] netfilter: bitwise: replace hard-coded size with `sizeof` expression

[Pablo Neira Ayuso]

From: Jeremy Sowden <jeremy@azazel.net>

When calculating the length of an array, use the appropriate `sizeof`
expression for its type, rather than an integer literal.

Signed-off-by: Jeremy Sowden <jeremy@azazel.net>
Signed-off-by: Florian Westphal <fw@strlen.de>
---
 net/netfilter/nft_bitwise.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/net/netfilter/nft_bitwise.c b/net/netfilter/nft_bitwise.c
index 38caa66632b4..dc5759fac5b6 100644
--- a/net/netfilter/nft_bitwise.c
+++ b/net/netfilter/nft_bitwise.c
@@ -30,7 +30,7 @@ static void nft_bitwise_eval_bool(u32 *dst, const u32 *src,
 {
 	unsigned int i;
 
-	for (i = 0; i < DIV_ROUND_UP(priv->len, 4); i++)
+	for (i = 0; i < DIV_ROUND_UP(priv->len, sizeof(u32)); i++)
 		dst[i] = (src[i] & priv->mask.data[i]) ^ priv->xor.data[i];
 }
 
-- 
2.30.2

[PATCH net-next 09/11] netfilter: bitwise: improve error goto labels

[Pablo Neira Ayuso]

From: Jeremy Sowden <jeremy@azazel.net>

Replace two labels (`err1` and `err2`) with more informative ones.

Signed-off-by: Jeremy Sowden <jeremy@azazel.net>
Signed-off-by: Florian Westphal <fw@strlen.de>
---
 net/netfilter/nft_bitwise.c | 11 ++++++-----
 1 file changed, 6 insertions(+), 5 deletions(-)

diff --git a/net/netfilter/nft_bitwise.c b/net/netfilter/nft_bitwise.c
index dc5759fac5b6..d72143622f22 100644
--- a/net/netfilter/nft_bitwise.c
+++ b/net/netfilter/nft_bitwise.c
@@ -109,22 +109,23 @@ static int nft_bitwise_init_bool(struct nft_bitwise *priv,
 		return err;
 	if (mask.type != NFT_DATA_VALUE || mask.len != priv->len) {
 		err = -EINVAL;
-		goto err1;
+		goto err_mask_release;
 	}
 
 	err = nft_data_init(NULL, &priv->xor, sizeof(priv->xor), &xor,
 			    tb[NFTA_BITWISE_XOR]);
 	if (err < 0)
-		goto err1;
+		goto err_mask_release;
 	if (xor.type != NFT_DATA_VALUE || xor.len != priv->len) {
 		err = -EINVAL;
-		goto err2;
+		goto err_xor_release;
 	}
 
 	return 0;
-err2:
+
+err_xor_release:
 	nft_data_release(&priv->xor, xor.type);
-err1:
+err_mask_release:
 	nft_data_release(&priv->mask, mask.type);
 	return err;
 }
-- 
2.30.2

[PATCH net-next 10/11] netfilter: nft_fib: reverse path filter for policy-based routing on iif

[Pablo Neira Ayuso]

If policy-based routing using the iif selector is used, then the fib
expression fails to look up for the reverse path from the prerouting
hook because the input interface cannot be inferred. In order to support
this scenario, extend the fib expression to allow to use after the route
lookup, from the forward hook.

This patch also adds support for the input hook for usability reasons.
Since the prerouting hook cannot be used for the scenario described
above, users need two rules: one for the forward chain and another rule
for the input chain to check for the reverse path check for locally
targeted traffic.

Acked-by: Florian Westphal <fw@strlen.de>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
---
 net/ipv4/netfilter/nft_fib_ipv4.c | 4 ++++
 net/ipv6/netfilter/nft_fib_ipv6.c | 4 ++++
 net/netfilter/nft_fib.c           | 4 ++++
 3 files changed, 12 insertions(+)

diff --git a/net/ipv4/netfilter/nft_fib_ipv4.c b/net/ipv4/netfilter/nft_fib_ipv4.c
index 4151eb1262dd..b75cac69bd7e 100644
--- a/net/ipv4/netfilter/nft_fib_ipv4.c
+++ b/net/ipv4/netfilter/nft_fib_ipv4.c
@@ -112,6 +112,10 @@ void nft_fib4_eval(const struct nft_expr *expr, struct nft_regs *regs,
 		fl4.daddr = iph->daddr;
 		fl4.saddr = get_saddr(iph->saddr);
 	} else {
+		if (nft_hook(pkt) == NF_INET_FORWARD &&
+		    priv->flags & NFTA_FIB_F_IIF)
+			fl4.flowi4_iif = nft_out(pkt)->ifindex;
+
 		fl4.daddr = iph->saddr;
 		fl4.saddr = get_saddr(iph->daddr);
 	}
diff --git a/net/ipv6/netfilter/nft_fib_ipv6.c b/net/ipv6/netfilter/nft_fib_ipv6.c
index b3f163b40c2b..8970d0b4faeb 100644
--- a/net/ipv6/netfilter/nft_fib_ipv6.c
+++ b/net/ipv6/netfilter/nft_fib_ipv6.c
@@ -30,6 +30,10 @@ static int nft_fib6_flowi_init(struct flowi6 *fl6, const struct nft_fib *priv,
 		fl6->daddr = iph->daddr;
 		fl6->saddr = iph->saddr;
 	} else {
+		if (nft_hook(pkt) == NF_INET_FORWARD &&
+		    priv->flags & NFTA_FIB_F_IIF)
+			fl6->flowi6_iif = nft_out(pkt)->ifindex;
+
 		fl6->daddr = iph->saddr;
 		fl6->saddr = iph->daddr;
 	}
diff --git a/net/netfilter/nft_fib.c b/net/netfilter/nft_fib.c
index f198f2d9ef90..1f12d7ade606 100644
--- a/net/netfilter/nft_fib.c
+++ b/net/netfilter/nft_fib.c
@@ -35,6 +35,10 @@ int nft_fib_validate(const struct nft_ctx *ctx, const struct nft_expr *expr,
 	case NFT_FIB_RESULT_OIF:
 	case NFT_FIB_RESULT_OIFNAME:
 		hooks = (1 << NF_INET_PRE_ROUTING);
+		if (priv->flags & NFTA_FIB_F_IIF) {
+			hooks |= (1 << NF_INET_LOCAL_IN) |
+				 (1 << NF_INET_FORWARD);
+		}
 		break;
 	case NFT_FIB_RESULT_ADDRTYPE:
 		if (priv->flags & NFTA_FIB_F_IIF)
-- 
2.30.2

[PATCH net-next 11/11] selftests: netfilter: add fib expression forward test case

[Pablo Neira Ayuso]

From: Florian Westphal <fw@strlen.de>

Its now possible to use fib expression in the forward chain (where both
the input and output interfaces are known).

Add a simple test case for this.

Signed-off-by: Florian Westphal <fw@strlen.de>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
---
 tools/testing/selftests/netfilter/nft_fib.sh | 50 ++++++++++++++++++++
 1 file changed, 50 insertions(+)

diff --git a/tools/testing/selftests/netfilter/nft_fib.sh b/tools/testing/selftests/netfilter/nft_fib.sh
index 695a1958723f..fd76b69635a4 100755
--- a/tools/testing/selftests/netfilter/nft_fib.sh
+++ b/tools/testing/selftests/netfilter/nft_fib.sh
@@ -66,6 +66,20 @@ table inet filter {
 EOF
 }
 
+load_pbr_ruleset() {
+	local netns=$1
+
+ip netns exec ${netns} nft -f /dev/stdin <<EOF
+table inet filter {
+	chain forward {
+		type filter hook forward priority raw;
+		fib saddr . iif oif gt 0 accept
+		log drop
+	}
+}
+EOF
+}
+
 load_ruleset_count() {
 	local netns=$1
 
@@ -219,4 +233,40 @@ sleep 2
 ip netns exec ${ns1} ping -c 3 -q 1c3::c01d > /dev/null
 check_fib_counter 3 ${nsrouter} 1c3::c01d || exit 1
 
+# delete all rules
+ip netns exec ${ns1} nft flush ruleset
+ip netns exec ${ns2} nft flush ruleset
+ip netns exec ${nsrouter} nft flush ruleset
+
+ip -net ${ns1} addr add 10.0.1.99/24 dev eth0
+ip -net ${ns1} addr add dead:1::99/64 dev eth0
+
+ip -net ${ns1} addr del 10.0.2.99/24 dev eth0
+ip -net ${ns1} addr del dead:2::99/64 dev eth0
+
+ip -net ${nsrouter} addr del dead:2::1/64 dev veth0
+
+# ... pbr ruleset for the router, check iif+oif.
+load_pbr_ruleset ${nsrouter}
+if [ $? -ne 0 ] ; then
+	echo "SKIP: Could not load fib forward ruleset"
+	exit $ksft_skip
+fi
+
+ip -net ${nsrouter} rule add from all table 128
+ip -net ${nsrouter} rule add from all iif veth0 table 129
+ip -net ${nsrouter} route add table 128 to 10.0.1.0/24 dev veth0
+ip -net ${nsrouter} route add table 129 to 10.0.2.0/24 dev veth1
+
+# drop main ipv4 table
+ip -net ${nsrouter} -4 rule delete table main
+
+test_ping 10.0.2.99 dead:2::99
+if [ $? -ne 0 ] ; then
+	ip -net ${nsrouter} nft list ruleset
+	echo "FAIL: fib mismatch in pbr setup"
+	exit 1
+fi
+
+echo "PASS: fib expression forward check with policy based routing"
 exit 0
-- 
2.30.2