[PATCH net,v4,0/2] Fix a fib6 info notification bug

This patch set fixes a length calculation bug in fib6 info notification
and adds two testcases.

The test reult is listed as follows:

before the fix patch:

Fib6 info length calculation in route notify test
    TEST: ipv6 route add notify                      [FAIL]

Fib4 info length calculation in route notify test
    TEST: ipv4 route add notify                      [ OK ]

after the fix patch:

Fib6 info length calculation in route notify test
    TEST: ipv6 route add notify                      [ OK ]

Fib4 info length calculation in route notify test
    TEST: ipv4 route add notify                      [ OK ]

Lu Wei (2):
  ipv6: Add lwtunnel encap size of all siblings in nexthop calculation
  selftests: fib_tests: Add test cases for IPv4/IPv6 in route notify

 net/ipv6/route.c                         | 11 +--
 tools/testing/selftests/net/fib_tests.sh | 96 +++++++++++++++++++++++-
 2 files changed, 101 insertions(+), 6 deletions(-)

-- 
2.31.1

In function rt6_nlmsg_size(), the length of nexthop is calculated
by multipling the nexthop length of fib6_info and the number of
siblings. However if the fib6_info has no lwtunnel but the siblings
have lwtunnels, the nexthop length is less than it should be, and
it will trigger a warning in inet6_rt_notify() as follows:

WARNING: CPU: 0 PID: 6082 at net/ipv6/route.c:6180 inet6_rt_notify+0x120/0x130
......
Call Trace:
 <TASK>
 fib6_add_rt2node+0x685/0xa30
 fib6_add+0x96/0x1b0
 ip6_route_add+0x50/0xd0
 inet6_rtm_newroute+0x97/0xa0
 rtnetlink_rcv_msg+0x156/0x3d0
 netlink_rcv_skb+0x5a/0x110
 netlink_unicast+0x246/0x350
 netlink_sendmsg+0x250/0x4c0
 sock_sendmsg+0x66/0x70
 ___sys_sendmsg+0x7c/0xd0
 __sys_sendmsg+0x5d/0xb0
 do_syscall_64+0x3f/0x90
 entry_SYSCALL_64_after_hwframe+0x72/0xdc

This bug can be reproduced by script:

ip -6 addr add 2002::2/64 dev ens2
ip -6 route add 100::/64 via 2002::1 dev ens2 metric 100

for i in 10 20 30 40 50 60 70;
do
	ip link add link ens2 name ipv_$i type ipvlan
	ip -6 addr add 2002::$i/64 dev ipv_$i
	ifconfig ipv_$i up
done

for i in 10 20 30 40 50 60;
do
	ip -6 route append 100::/64 encap ip6 dst 2002::$i via 2002::1
dev ipv_$i metric 100
done

ip -6 route append 100::/64 via 2002::1 dev ipv_70 metric 100

This patch fixes it by adding nexthop_len of every siblings using
rt6_nh_nlmsg_size().

Fixes: beb1afac518d ("net: ipv6: Add support to dump multipath routes via RTA_MULTIPATH attribute")
Signed-off-by: Lu Wei <redacted>
---
 net/ipv6/route.c | 11 ++++++-----
 1 file changed, 6 insertions(+), 5 deletions(-)

diff --git a/net/ipv6/route.c b/net/ipv6/route.c
index c180c2ef17c5..0fdb03df2287 100644
--- a/net/ipv6/route.c
+++ b/net/ipv6/route.c

@@ -5533,16 +5533,17 @@ static size_t rt6_nlmsg_size(struct fib6_info *f6i)
 		nexthop_for_each_fib6_nh(f6i->nh, rt6_nh_nlmsg_size,
 					 &nexthop_len);
 	} else {
+		struct fib6_info *sibling, *next_sibling;
 		struct fib6_nh *nh = f6i->fib6_nh;
 
 		nexthop_len = 0;
 		if (f6i->fib6_nsiblings) {
-			nexthop_len = nla_total_size(0)	 /* RTA_MULTIPATH */
-				    + NLA_ALIGN(sizeof(struct rtnexthop))
-				    + nla_total_size(16) /* RTA_GATEWAY */
-				    + lwtunnel_get_encap_size(nh->fib_nh_lws);
+			rt6_nh_nlmsg_size(nh, &nexthop_len);
 
-			nexthop_len *= f6i->fib6_nsiblings;
+			list_for_each_entry_safe(sibling, next_sibling,
+						 &f6i->fib6_siblings, fib6_siblings) {
+				rt6_nh_nlmsg_size(sibling->fib6_nh, &nexthop_len);
+			}
 		}
 		nexthop_len += lwtunnel_get_encap_size(nh->fib_nh_lws);
 	}

-- 
2.31.1

Add tests to check whether the total fib info length is calculated
corretly in route notify process.

Signed-off-by: Lu Wei <redacted>
---
 tools/testing/selftests/net/fib_tests.sh | 96 +++++++++++++++++++++++-
 1 file changed, 95 insertions(+), 1 deletion(-)

diff --git a/tools/testing/selftests/net/fib_tests.sh b/tools/testing/selftests/net/fib_tests.sh
index 70ea8798b1f6..7da8ec838c63 100755
--- a/tools/testing/selftests/net/fib_tests.sh
+++ b/tools/testing/selftests/net/fib_tests.sh

@@ -9,7 +9,7 @@ ret=0
 ksft_skip=4
 
 # all tests in this script. Can be overridden with -t option
-TESTS="unregister down carrier nexthop suppress ipv6_rt ipv4_rt ipv6_addr_metric ipv4_addr_metric ipv6_route_metrics ipv4_route_metrics ipv4_route_v6_gw rp_filter ipv4_del_addr ipv4_mangle ipv6_mangle ipv4_bcast_neigh"
+TESTS="unregister down carrier nexthop suppress ipv6_notify ipv4_notify ipv6_rt ipv4_rt ipv6_addr_metric ipv4_addr_metric ipv6_route_metrics ipv4_route_metrics ipv4_route_v6_gw rp_filter ipv4_del_addr ipv4_mangle ipv6_mangle ipv4_bcast_neigh"
 
 VERBOSE=0
 PAUSE_ON_FAIL=no

@@ -655,6 +655,98 @@ fib_nexthop_test()
 	cleanup
 }
 
+fib6_notify_test()
+{
+	setup
+
+	echo
+	echo "Fib6 info length calculation in route notify test"
+	set -e
+
+	for i in 10 20 30 40 50 60 70;
+	do
+		$IP link add dummy_$i type dummy
+		$IP link set dev dummy_$i up
+		$IP -6 address add 2001:$i::1/64 dev dummy_$i
+	done
+
+	$NS_EXEC ip monitor route &> errors.txt &
+	sleep 2
+
+	$IP -6 route add 2001::/64 \
+                nexthop via 2001:10::2 dev dummy_10 \
+                nexthop encap ip6 dst 2002::20 via 2001:20::2 dev dummy_20 \
+                nexthop encap ip6 dst 2002::30 via 2001:30::2 dev dummy_30 \
+                nexthop encap ip6 dst 2002::40 via 2001:40::2 dev dummy_40 \
+                nexthop encap ip6 dst 2002::50 via 2001:50::2 dev dummy_50 \
+                nexthop encap ip6 dst 2002::60 via 2001:60::2 dev dummy_60 \
+                nexthop encap ip6 dst 2002::70 via 2001:70::2 dev dummy_70
+
+	set +e
+
+	err=`cat errors.txt |grep "Message too long"`
+	if [ -z "$err" ];then
+		ret=0
+	else
+		ret=1
+	fi
+
+	log_test $ret 0 "ipv6 route add notify"
+
+	{ kill %% && wait %%; } 2>/dev/null
+
+	#rm errors.txt
+
+	cleanup &> /dev/null
+}
+
+
+fib_notify_test()
+{
+	setup
+
+	echo
+	echo "Fib4 info length calculation in route notify test"
+
+	set -e
+
+	for i in 10 20 30 40 50 60 70;
+	do
+		$IP link add dummy_$i type dummy
+		$IP link set dev dummy_$i up
+		$IP address add 20.20.$i.2/24 dev dummy_$i
+	done
+
+	$NS_EXEC ip monitor route &> errors.txt &
+	sleep 2
+
+        $IP route add 10.0.0.0/24 \
+                nexthop via 20.20.10.1 dev dummy_10 \
+                nexthop encap ip dst 192.168.10.20 via 20.20.20.1 dev dummy_20 \
+                nexthop encap ip dst 192.168.10.30 via 20.20.30.1 dev dummy_30 \
+                nexthop encap ip dst 192.168.10.40 via 20.20.40.1 dev dummy_40 \
+                nexthop encap ip dst 192.168.10.50 via 20.20.50.1 dev dummy_50 \
+                nexthop encap ip dst 192.168.10.60 via 20.20.60.1 dev dummy_60 \
+                nexthop encap ip dst 192.168.10.70 via 20.20.70.1 dev dummy_70
+
+	set +e
+
+	err=`cat errors.txt |grep "Message too long"`
+	if [ -z "$err" ];then
+		ret=0
+	else
+		ret=1
+	fi
+
+	log_test $ret 0 "ipv4 route add notify"
+
+	{ kill %% && wait %%; } 2>/dev/null
+
+	rm  errors.txt
+
+	cleanup &> /dev/null
+}
+
 fib_suppress_test()
 {
 	echo

@@ -2111,6 +2203,8 @@ do
 	fib_carrier_test|carrier)	fib_carrier_test;;
 	fib_rp_filter_test|rp_filter)	fib_rp_filter_test;;
 	fib_nexthop_test|nexthop)	fib_nexthop_test;;
+	fib_notify_test|ipv4_notify)	fib_notify_test;;
+	fib6_notify_test|ipv6_notify)	fib6_notify_test;;
 	fib_suppress_test|suppress)	fib_suppress_test;;
 	ipv6_route_test|ipv6_rt)	ipv6_route_test;;
 	ipv4_route_test|ipv4_rt)	ipv4_route_test;;

-- 
2.31.1

On 2/22/23 1:36 AM, Lu Wei wrote:

In function rt6_nlmsg_size(), the length of nexthop is calculated
by multipling the nexthop length of fib6_info and the number of
siblings. However if the fib6_info has no lwtunnel but the siblings
have lwtunnels, the nexthop length is less than it should be, and
it will trigger a warning in inet6_rt_notify() as follows:

WARNING: CPU: 0 PID: 6082 at net/ipv6/route.c:6180 inet6_rt_notify+0x120/0x130
......
Call Trace:
 <TASK>
 fib6_add_rt2node+0x685/0xa30
 fib6_add+0x96/0x1b0
 ip6_route_add+0x50/0xd0
 inet6_rtm_newroute+0x97/0xa0
 rtnetlink_rcv_msg+0x156/0x3d0
 netlink_rcv_skb+0x5a/0x110
 netlink_unicast+0x246/0x350
 netlink_sendmsg+0x250/0x4c0
 sock_sendmsg+0x66/0x70
 ___sys_sendmsg+0x7c/0xd0
 __sys_sendmsg+0x5d/0xb0
 do_syscall_64+0x3f/0x90
 entry_SYSCALL_64_after_hwframe+0x72/0xdc

This bug can be reproduced by script:

ip -6 addr add 2002::2/64 dev ens2
ip -6 route add 100::/64 via 2002::1 dev ens2 metric 100

for i in 10 20 30 40 50 60 70;
do
	ip link add link ens2 name ipv_$i type ipvlan
	ip -6 addr add 2002::$i/64 dev ipv_$i
	ifconfig ipv_$i up
done

for i in 10 20 30 40 50 60;
do
	ip -6 route append 100::/64 encap ip6 dst 2002::$i via 2002::1
dev ipv_$i metric 100
done

ip -6 route append 100::/64 via 2002::1 dev ipv_70 metric 100

This patch fixes it by adding nexthop_len of every siblings using
rt6_nh_nlmsg_size().

Fixes: beb1afac518d ("net: ipv6: Add support to dump multipath routes via RTA_MULTIPATH attribute")
Signed-off-by: Lu Wei <redacted>
---
 net/ipv6/route.c | 11 ++++++-----
 1 file changed, 6 insertions(+), 5 deletions(-)

Reviewed-by: David Ahern <dsahern@kernel.org>

On 2/22/23 1:36 AM, Lu Wei wrote:

Add tests to check whether the total fib info length is calculated
corretly in route notify process.

Signed-off-by: Lu Wei <redacted>
---
 tools/testing/selftests/net/fib_tests.sh | 96 +++++++++++++++++++++++-
 1 file changed, 95 insertions(+), 1 deletion(-)

Reviewed-by: David Ahern <dsahern@kernel.org>

Hello:

This series was applied to netdev/net.git (master)
by Paolo Abeni [off-list ref]:

On Wed, 22 Feb 2023 16:36:27 +0800 you wrote:

This patch set fixes a length calculation bug in fib6 info notification
and adds two testcases.

The test reult is listed as follows:

before the fix patch:

[...]

Here is the summary with links:
  - [net,v4,1/2] ipv6: Add lwtunnel encap size of all siblings in nexthop calculation
    https://git.kernel.org/netdev/net/c/4cc59f386991
  - [net,v4,2/2] selftests: fib_tests: Add test cases for IPv4/IPv6 in route notify
    https://git.kernel.org/netdev/net/c/44bd0394fe10

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