Merge git://git.kernel.org/pub/scm/linux/kernel/git/pablo/nf

Pablo Neira Ayuso says:

====================
Netfilter fixes for net

The following patchset contains Netfilter fixes for net:

1) Do not delete clash entries on reply, let them expire instead,
from Florian Westphal.

2) Do not report EAGAIN to nfnetlink, otherwise this enters a busy loop.
Update nfnetlink_unicast() to translate EAGAIN to ENOBUFS.

3) Remove repeated words in code comments, from Randy Dunlap.

4) Several patches for the flowtable selftests, from Fabian Frederick.
====================

Signed-off-by: David S. Miller <[e-mail]>

include/linux/netfilter/nfnetlink.h

···
 int nfnetlink_send(struct sk_buff *skb, struct net *net, u32 portid,
 		   unsigned int group, int echo, gfp_t flags);
 int nfnetlink_set_err(struct net *net, u32 portid, u32 group, int error);
-int nfnetlink_unicast(struct sk_buff *skb, struct net *net, u32 portid,
-		      int flags);
+int nfnetlink_unicast(struct sk_buff *skb, struct net *net, u32 portid);
 
 static inline u16 nfnl_msg_type(u8 subsys, u8 msg_type)
 {

net/ipv4/netfilter/nf_nat_pptp.c

···
  * nf_nat_pptp.c
  *
  * NAT support for PPTP (Point to Point Tunneling Protocol).
- * PPTP is a a protocol for creating virtual private networks.
+ * PPTP is a protocol for creating virtual private networks.
  * It is a specification defined by Microsoft and some vendors
  * working with Microsoft.  PPTP is built on top of a modified
  * version of the Internet Generic Routing Encapsulation Protocol.

net/netfilter/nf_conntrack_pptp.c

···
 // SPDX-License-Identifier: GPL-2.0-only
 /*
  * Connection tracking support for PPTP (Point to Point Tunneling Protocol).
- * PPTP is a a protocol for creating virtual private networks.
+ * PPTP is a protocol for creating virtual private networks.
  * It is a specification defined by Microsoft and some vendors
  * working with Microsoft.  PPTP is built on top of a modified
  * version of the Internet Generic Routing Encapsulation Protocol.

net/netfilter/nf_conntrack_proto_tcp.c

···
 		   && (old_state == TCP_CONNTRACK_SYN_RECV
 		       || old_state == TCP_CONNTRACK_ESTABLISHED)
 		   && new_state == TCP_CONNTRACK_ESTABLISHED) {
-		/* Set ASSURED if we see see valid ack in ESTABLISHED
+		/* Set ASSURED if we see valid ack in ESTABLISHED
 		   after SYN_RECV or a valid answer for a picked up
 		   connection. */
 		set_bit(IPS_ASSURED_BIT, &ct->status);

net/netfilter/nf_conntrack_proto_udp.c

···
 	return false;
 }
 
-static void nf_conntrack_udp_refresh_unreplied(struct nf_conn *ct,
-					       struct sk_buff *skb,
-					       enum ip_conntrack_info ctinfo,
-					       u32 extra_jiffies)
-{
-	if (unlikely(ctinfo == IP_CT_ESTABLISHED_REPLY &&
-		     ct->status & IPS_NAT_CLASH))
-		nf_ct_kill(ct);
-	else
-		nf_ct_refresh_acct(ct, ctinfo, skb, extra_jiffies);
-}
-
 /* Returns verdict for packet, and may modify conntracktype */
 int nf_conntrack_udp_packet(struct nf_conn *ct,
 			    struct sk_buff *skb,
···
 
 		nf_ct_refresh_acct(ct, ctinfo, skb, extra);
 
+		/* never set ASSURED for IPS_NAT_CLASH, they time out soon */
+		if (unlikely((ct->status & IPS_NAT_CLASH)))
+			return NF_ACCEPT;
+
 		/* Also, more likely to be important, and not a probe */
 		if (!test_and_set_bit(IPS_ASSURED_BIT, &ct->status))
 			nf_conntrack_event_cache(IPCT_ASSURED, ct);
 	} else {
-		nf_conntrack_udp_refresh_unreplied(ct, skb, ctinfo,
-						   timeouts[UDP_CT_UNREPLIED]);
+		nf_ct_refresh_acct(ct, ctinfo, skb, timeouts[UDP_CT_UNREPLIED]);
 	}
 	return NF_ACCEPT;
 }
···
 	if (test_bit(IPS_SEEN_REPLY_BIT, &ct->status)) {
 		nf_ct_refresh_acct(ct, ctinfo, skb,
 				   timeouts[UDP_CT_REPLIED]);
+
+		if (unlikely((ct->status & IPS_NAT_CLASH)))
+			return NF_ACCEPT;
+
 		/* Also, more likely to be important, and not a probe */
 		if (!test_and_set_bit(IPS_ASSURED_BIT, &ct->status))
 			nf_conntrack_event_cache(IPCT_ASSURED, ct);
 	} else {
-		nf_conntrack_udp_refresh_unreplied(ct, skb, ctinfo,
-						   timeouts[UDP_CT_UNREPLIED]);
+		nf_ct_refresh_acct(ct, ctinfo, skb, timeouts[UDP_CT_UNREPLIED]);
 	}
 	return NF_ACCEPT;
 }

net/netfilter/nf_tables_api.c

···
 					nlh->nlmsg_seq, NFT_MSG_NEWTABLE, 0,
 					family, table);
 	if (err < 0)
-		goto err;
+		goto err_fill_table_info;
 
-	return nlmsg_unicast(nlsk, skb2, NETLINK_CB(skb).portid);
+	return nfnetlink_unicast(skb2, net, NETLINK_CB(skb).portid);
 
-err:
+err_fill_table_info:
 	kfree_skb(skb2);
 	return err;
 }
···
 					nlh->nlmsg_seq, NFT_MSG_NEWCHAIN, 0,
 					family, table, chain);
 	if (err < 0)
-		goto err;
+		goto err_fill_chain_info;
 
-	return nlmsg_unicast(nlsk, skb2, NETLINK_CB(skb).portid);
+	return nfnetlink_unicast(skb2, net, NETLINK_CB(skb).portid);
 
-err:
+err_fill_chain_info:
 	kfree_skb(skb2);
 	return err;
 }
···
 				       nlh->nlmsg_seq, NFT_MSG_NEWRULE, 0,
 				       family, table, chain, rule, NULL);
 	if (err < 0)
-		goto err;
+		goto err_fill_rule_info;
 
-	return nlmsg_unicast(nlsk, skb2, NETLINK_CB(skb).portid);
+	return nfnetlink_unicast(skb2, net, NETLINK_CB(skb).portid);
 
-err:
+err_fill_rule_info:
 	kfree_skb(skb2);
 	return err;
 }
···
 
 	err = nf_tables_fill_set(skb2, &ctx, set, NFT_MSG_NEWSET, 0);
 	if (err < 0)
-		goto err;
+		goto err_fill_set_info;
 
-	return nlmsg_unicast(nlsk, skb2, NETLINK_CB(skb).portid);
+	return nfnetlink_unicast(skb2, net, NETLINK_CB(skb).portid);
 
-err:
+err_fill_set_info:
 	kfree_skb(skb2);
 	return err;
 }
···
 	err = -ENOMEM;
 	skb = nlmsg_new(NLMSG_GOODSIZE, GFP_ATOMIC);
 	if (skb == NULL)
-		goto err1;
+		return err;
 
 	err = nf_tables_fill_setelem_info(skb, ctx, ctx->seq, ctx->portid,
 					  NFT_MSG_NEWSETELEM, 0, set, &elem);
 	if (err < 0)
-		goto err2;
+		goto err_fill_setelem;
 
-	err = nfnetlink_unicast(skb, ctx->net, ctx->portid, MSG_DONTWAIT);
-	/* This avoids a loop in nfnetlink. */
-	if (err < 0)
-		goto err1;
+	return nfnetlink_unicast(skb, ctx->net, ctx->portid);
 
-	return 0;
-err2:
+err_fill_setelem:
 	kfree_skb(skb);
-err1:
-	/* this avoids a loop in nfnetlink. */
-	return err == -EAGAIN ? -ENOBUFS : err;
+	return err;
 }
 
 /* called with rcu_read_lock held */
···
 				      nlh->nlmsg_seq, NFT_MSG_NEWOBJ, 0,
 				      family, table, obj, reset);
 	if (err < 0)
-		goto err;
+		goto err_fill_obj_info;
 
-	return nlmsg_unicast(nlsk, skb2, NETLINK_CB(skb).portid);
-err:
+	return nfnetlink_unicast(skb2, net, NETLINK_CB(skb).portid);
+
+err_fill_obj_info:
 	kfree_skb(skb2);
 	return err;
 }
···
 					    NFT_MSG_NEWFLOWTABLE, 0, family,
 					    flowtable, &flowtable->hook_list);
 	if (err < 0)
-		goto err;
+		goto err_fill_flowtable_info;
 
-	return nlmsg_unicast(nlsk, skb2, NETLINK_CB(skb).portid);
-err:
+	return nfnetlink_unicast(skb2, net, NETLINK_CB(skb).portid);
+
+err_fill_flowtable_info:
 	kfree_skb(skb2);
 	return err;
 }
···
 	err = nf_tables_fill_gen_info(skb2, net, NETLINK_CB(skb).portid,
 				      nlh->nlmsg_seq);
 	if (err < 0)
-		goto err;
+		goto err_fill_gen_info;
 
-	return nlmsg_unicast(nlsk, skb2, NETLINK_CB(skb).portid);
-err:
+	return nfnetlink_unicast(skb2, net, NETLINK_CB(skb).portid);
+
+err_fill_gen_info:
 	kfree_skb(skb2);
 	return err;
 }

net/netfilter/nfnetlink.c

···
 }
 EXPORT_SYMBOL_GPL(nfnetlink_set_err);
 
-int nfnetlink_unicast(struct sk_buff *skb, struct net *net, u32 portid,
-		      int flags)
+int nfnetlink_unicast(struct sk_buff *skb, struct net *net, u32 portid)
 {
-	return netlink_unicast(net->nfnl, skb, portid, flags);
+	int err;
+
+	err = nlmsg_unicast(net->nfnl, skb, portid);
+	if (err == -EAGAIN)
+		err = -ENOBUFS;
+
+	return err;
 }
 EXPORT_SYMBOL_GPL(nfnetlink_unicast);
 

net/netfilter/nfnetlink_log.c

···
 			goto out;
 		}
 	}
-	nfnetlink_unicast(inst->skb, inst->net, inst->peer_portid,
-			  MSG_DONTWAIT);
+	nfnetlink_unicast(inst->skb, inst->net, inst->peer_portid);
 out:
 	inst->qlen = 0;
 	inst->skb = NULL;

net/netfilter/nfnetlink_queue.c

···
 	*packet_id_ptr = htonl(entry->id);
 
 	/* nfnetlink_unicast will either free the nskb or add it to a socket */
-	err = nfnetlink_unicast(nskb, net, queue->peer_portid, MSG_DONTWAIT);
+	err = nfnetlink_unicast(nskb, net, queue->peer_portid);
 	if (err < 0) {
 		if (queue->flags & NFQA_CFG_F_FAIL_OPEN) {
 			failopen = 1;

net/netfilter/nft_flow_offload.c

···
 	}
 
 	if (nf_ct_ext_exist(ct, NF_CT_EXT_HELPER) ||
-	    ct->status & IPS_SEQ_ADJUST)
+	    ct->status & (IPS_SEQ_ADJUST | IPS_NAT_CLASH))
 		goto out;
 
 	if (!nf_ct_is_confirmed(ct))

net/netfilter/xt_recent.c

···
 	struct recent_table *t;
 
 	/* recent_net_exit() is called before recent_mt_destroy(). Make sure
-	 * that the parent xt_recent proc entry is is empty before trying to
+	 * that the parent xt_recent proc entry is empty before trying to
 	 * remove it.
 	 */
 	spin_lock_bh(&recent_lock);

tools/testing/selftests/netfilter/nft_flowtable.sh

···
 # result in fragmentation and/or PMTU discovery.
 #
 # You can check with different Orgininator/Link/Responder MTU eg:
-# sh nft_flowtable.sh -o1000 -l500 -r100
+# nft_flowtable.sh -o8000 -l1500 -r2000
 #
 
 
···
 log_netns=$(sysctl -n net.netfilter.nf_log_all_netns)
 
 checktool (){
-	$1 > /dev/null 2>&1
-	if [ $? -ne 0 ];then
+	if ! $1 > /dev/null 2>&1; then
 		echo "SKIP: Could not $2"
 		exit $ksft_skip
 	fi
···
 lmtu=1500
 rmtu=2000
 
+usage(){
+	echo "nft_flowtable.sh [OPTIONS]"
+	echo
+	echo "MTU options"
+	echo "   -o originator"
+	echo "   -l link"
+	echo "   -r responder"
+	exit 1
+}
+
 while getopts "o:l:r:" o
 do
 	case $o in
 		o) omtu=$OPTARG;;
 		l) lmtu=$OPTARG;;
 		r) rmtu=$OPTARG;;
+		*) usage;;
 	esac
 done
 
-ip -net nsr1 link set veth0 mtu $omtu
+if ! ip -net nsr1 link set veth0 mtu $omtu; then
+	exit 1
+fi
+
 ip -net ns1 link set eth0 mtu $omtu
 
-ip -net nsr2 link set veth1 mtu $rmtu
+if ! ip -net nsr2 link set veth1 mtu $rmtu; then
+	exit 1
+fi
+
 ip -net ns2 link set eth0 mtu $rmtu
 
 # transfer-net between nsr1 and nsr2.
···
   ip -net ns$i route add default via 10.0.$i.1
   ip -net ns$i addr add dead:$i::99/64 dev eth0
   ip -net ns$i route add default via dead:$i::1
-  ip netns exec ns$i sysctl net.ipv4.tcp_no_metrics_save=1 > /dev/null
+  if ! ip netns exec ns$i sysctl net.ipv4.tcp_no_metrics_save=1 > /dev/null; then
+	echo "ERROR: Check Originator/Responder values (problem during address addition)"
+	exit 1
+  fi
 
   # don't set ip DF bit for first two tests
   ip netns exec ns$i sysctl net.ipv4.ip_no_pmtu_disc=1 > /dev/null
···
 fi
 
 # test basic connectivity
-ip netns exec ns1 ping -c 1 -q 10.0.2.99 > /dev/null
-if [ $? -ne 0 ];then
+if ! ip netns exec ns1 ping -c 1 -q 10.0.2.99 > /dev/null; then
   echo "ERROR: ns1 cannot reach ns2" 1>&2
   bash
   exit 1
 fi
 
-ip netns exec ns2 ping -c 1 -q 10.0.1.99 > /dev/null
-if [ $? -ne 0 ];then
+if ! ip netns exec ns2 ping -c 1 -q 10.0.1.99 > /dev/null; then
   echo "ERROR: ns2 cannot reach ns1" 1>&2
   exit 1
 fi
···
 make_file()
 {
 	name=$1
-	who=$2
 
 	SIZE=$((RANDOM % (1024 * 8)))
 	TSIZE=$((SIZE * 1024))
···
 	out=$2
 	what=$3
 
-	cmp "$in" "$out" > /dev/null 2>&1
-	if [ $? -ne 0 ] ;then
+	if ! cmp "$in" "$out" > /dev/null 2>&1; then
 		echo "FAIL: file mismatch for $what" 1>&2
 		ls -l "$in"
 		ls -l "$out"
···
 
 	wait
 
-	check_transfer "$ns1in" "$ns2out" "ns1 -> ns2"
-	if [ $? -ne 0 ];then
+	if ! check_transfer "$ns1in" "$ns2out" "ns1 -> ns2"; then
 		lret=1
 	fi
 
-	check_transfer "$ns2in" "$ns1out" "ns1 <- ns2"
-	if [ $? -ne 0 ];then
+	if ! check_transfer "$ns2in" "$ns1out" "ns1 <- ns2"; then
 		lret=1
 	fi
 
···
 	return $lret
 }
 
-make_file "$ns1in" "ns1"
-make_file "$ns2in" "ns2"
+make_file "$ns1in"
+make_file "$ns2in"
 
 # First test:
 # No PMTU discovery, nsr1 is expected to fragment packets from ns1 to ns2 as needed.
-test_tcp_forwarding ns1 ns2
-if [ $? -eq 0 ] ;then
+if test_tcp_forwarding ns1 ns2; then
 	echo "PASS: flow offloaded for ns1/ns2"
 else
 	echo "FAIL: flow offload for ns1/ns2:" 1>&2
···
 }
 EOF
 
-test_tcp_forwarding_nat ns1 ns2
-
-if [ $? -eq 0 ] ;then
+if test_tcp_forwarding_nat ns1 ns2; then
 	echo "PASS: flow offloaded for ns1/ns2 with NAT"
 else
 	echo "FAIL: flow offload for ns1/ns2 with NAT" 1>&2
···
 # Same as second test, but with PMTU discovery enabled.
 handle=$(ip netns exec nsr1 nft -a list table inet filter | grep something-to-grep-for | cut -d \# -f 2)
 
-ip netns exec nsr1 nft delete rule inet filter forward $handle
-if [ $? -ne 0 ] ;then
+if ! ip netns exec nsr1 nft delete rule inet filter forward $handle; then
 	echo "FAIL: Could not delete large-packet accept rule"
 	exit 1
 fi
···
 ip netns exec ns1 sysctl net.ipv4.ip_no_pmtu_disc=0 > /dev/null
 ip netns exec ns2 sysctl net.ipv4.ip_no_pmtu_disc=0 > /dev/null
 
-test_tcp_forwarding_nat ns1 ns2
-if [ $? -eq 0 ] ;then
+if test_tcp_forwarding_nat ns1 ns2; then
 	echo "PASS: flow offloaded for ns1/ns2 with NAT and pmtu discovery"
 else
 	echo "FAIL: flow offload for ns1/ns2 with NAT and pmtu discovery" 1>&2
···
 ip -net ns2 route add default via 10.0.2.1
 ip -net ns2 route add default via dead:2::1
 
-test_tcp_forwarding ns1 ns2
-if [ $? -eq 0 ] ;then
+if test_tcp_forwarding ns1 ns2; then
 	echo "PASS: ipsec tunnel mode for ns1/ns2"
 else
 	echo "FAIL: ipsec tunnel mode for ns1/ns2"