Merge branch 'ipv6-honor-oif-when-choosing-nexthop-for-locally-generated-traffic'

Ido Schimmel says:

====================
ipv6: Honor oif when choosing nexthop for locally generated traffic

Patch #1 is a preparation patch following the comment from Sashiko on
v2. See details in the commit message.

Patch #2 aligns IPv6 with IPv4 and changes IPv6 route lookup to prefer a
nexthop whose nexthop device matches the specified oif.

Patch #3 adds a selftest.
====================

Link: https://patch.msgid.link/20260611154605.992528-1-idosch@nvidia.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

2 files changed, 263 insertions, 10 deletions

diff --git a/net/ipv6/route.c b/net/ipv6/route.c
index 8259c7527aa4..c14b078b0249 100644
--- a/net/ipv6/route.c
+++ b/net/ipv6/route.c
@@ -819,9 +819,11 @@ static int rt6_nh_find_match(struct fib6_nh *nh, void *_arg)
 {
 	struct fib6_nh_frl_arg *arg = _arg;
 
-	arg->nh = nh;
-	return find_match(nh, arg->flags, arg->oif, arg->strict,
-			  arg->mpri, arg->do_rr);
+	if (find_match(nh, arg->flags, arg->oif, arg->strict, arg->mpri,
+		       arg->do_rr))
+		arg->nh = nh;
+
+	return 0;
 }
 
 static void __find_rr_leaf(struct fib6_info *f6i_start,
@@ -861,11 +863,10 @@ static void __find_rr_leaf(struct fib6_info *f6i_start,
 				res->nh = nexthop_fib6_nh(f6i->nh);
 				return;
 			}
-			if (nexthop_for_each_fib6_nh(f6i->nh, rt6_nh_find_match,
-						     &arg)) {
-				matched = true;
-				nh = arg.nh;
-			}
+			nexthop_for_each_fib6_nh(f6i->nh, rt6_nh_find_match,
+						 &arg);
+			matched = !!arg.nh;
+			nh = arg.nh;
 		} else {
 			nh = f6i->fib6_nh;
 			if (find_match(nh, f6i->fib6_flags, oif, strict,
@@ -2275,6 +2276,7 @@ struct rt6_info *ip6_pol_route(struct net *net, struct fib6_table *table,
 {
 	struct fib6_result res = {};
 	struct rt6_info *rt = NULL;
+	bool have_oif_match;
 	int strict = 0;
 
 	WARN_ON_ONCE((flags & RT6_LOOKUP_F_DST_NOREF) &&
@@ -2291,7 +2293,9 @@ struct rt6_info *ip6_pol_route(struct net *net, struct fib6_table *table,
 	if (res.f6i == net->ipv6.fib6_null_entry)
 		goto out;
 
-	fib6_select_path(net, &res, fl6, oif, false, skb, strict);
+	have_oif_match = fl6->flowi6_iif == LOOPBACK_IFINDEX &&
+			 oif == res.nh->fib_nh_dev->ifindex;
+	fib6_select_path(net, &res, fl6, oif, have_oif_match, skb, strict);
 
 	/*Search through exception table */
 	rt = rt6_find_cached_rt(&res, &fl6->daddr, &fl6->saddr);
diff --git a/tools/testing/selftests/net/fib_tests.sh b/tools/testing/selftests/net/fib_tests.sh
index 8f10de0eb985..b338bfb196a2 100755
--- a/tools/testing/selftests/net/fib_tests.sh
+++ b/tools/testing/selftests/net/fib_tests.sh
@@ -12,7 +12,9 @@ TESTS="unregister down carrier nexthop suppress ipv6_notify ipv4_notify \
        ipv4_route_metrics ipv4_route_v6_gw rp_filter ipv4_del_addr \
        ipv6_del_addr ipv4_mangle ipv6_mangle ipv4_bcast_neigh fib6_gc_test \
        ipv4_mpath_list ipv6_mpath_list ipv4_mpath_balance ipv6_mpath_balance \
-       ipv4_mpath_balance_preferred fib6_ra_to_static fib6_temp_addr_renewal"
+       ipv4_mpath_balance_preferred ipv4_mpath_oif ipv4_mpath_oif_nh \
+       ipv4_mpath_oif_vrf ipv6_mpath_oif ipv6_mpath_oif_nh ipv6_mpath_oif_vrf \
+       fib6_ra_to_static fib6_temp_addr_renewal"
 
 VERBOSE=0
 PAUSE_ON_FAIL=no
@@ -2971,6 +2973,247 @@ ipv6_mpath_balance_test()
 	forwarding_cleanup
 }
 
+ipv4_mpath_oif_test_common()
+{
+	local get_param=$1; shift
+	local expected_oif=$1; shift
+	local test_name=$1; shift
+	local tmp_file
+
+	tmp_file=$(mktemp)
+
+	for i in {1..100}; do
+		$IP route get 203.0.113.${i} $get_param >> "$tmp_file"
+	done
+
+	[[ $(grep "$expected_oif" "$tmp_file" | wc -l) -eq 100 ]]
+	log_test $? 0 "$test_name"
+
+	rm "$tmp_file"
+}
+
+ipv4_mpath_oif_test()
+{
+	echo
+	echo "IPv4 multipath oif test"
+
+	setup
+
+	set -e
+	$IP link add dummy1 up type dummy
+	$IP address add 192.0.2.1/28 dev dummy1
+	$IP address add 192.0.2.17/32 dev lo
+
+	$IP route add 203.0.113.0/24 \
+		nexthop via 198.51.100.2 dev dummy0 \
+		nexthop via 192.0.2.2 dev dummy1
+	set +e
+
+	ipv4_mpath_oif_test_common "oif dummy0" "dummy0" \
+		"IPv4 multipath via first nexthop"
+
+	ipv4_mpath_oif_test_common "oif dummy1" "dummy1" \
+		"IPv4 multipath via second nexthop"
+
+	ipv4_mpath_oif_test_common "oif dummy0 from 192.0.2.17" "dummy0" \
+		"IPv4 multipath via first nexthop with source address"
+
+	ipv4_mpath_oif_test_common "oif dummy1 from 192.0.2.17" "dummy1" \
+		"IPv4 multipath via second nexthop with source address"
+
+	cleanup
+}
+
+ipv4_mpath_oif_nh_test()
+{
+	echo
+	echo "IPv4 multipath oif with nexthop object test"
+
+	setup
+
+	set -e
+	$IP link add dummy1 up type dummy
+	$IP address add 192.0.2.1/28 dev dummy1
+	$IP address add 192.0.2.17/32 dev lo
+
+	$IP nexthop add id 1 via 198.51.100.2 dev dummy0
+	$IP nexthop add id 2 via 192.0.2.2 dev dummy1
+	$IP nexthop add id 3 group 1/2
+	$IP route add 203.0.113.0/24 nhid 3
+	set +e
+
+	ipv4_mpath_oif_test_common "oif dummy0" "dummy0" \
+		"IPv4 multipath via first nexthop"
+
+	ipv4_mpath_oif_test_common "oif dummy1" "dummy1" \
+		"IPv4 multipath via second nexthop"
+
+	ipv4_mpath_oif_test_common "oif dummy0 from 192.0.2.17" "dummy0" \
+		"IPv4 multipath via first nexthop with source address"
+
+	ipv4_mpath_oif_test_common "oif dummy1 from 192.0.2.17" "dummy1" \
+		"IPv4 multipath via second nexthop with source address"
+
+	cleanup
+}
+
+ipv4_mpath_oif_vrf_test()
+{
+	echo
+	echo "IPv4 multipath oif with VRF test"
+
+	setup
+
+	set -e
+	$IP -4 rule add pref 32765 table local
+	$IP -4 rule del pref 0
+	$IP link add name vrf-123 up type vrf table 123
+	$IP link set dev dummy0 master vrf-123
+	$IP link add dummy1 up master vrf-123 type dummy
+	$IP address add 192.0.2.1/28 dev dummy1
+	$IP address add 192.0.2.17/32 dev vrf-123
+
+	$IP route add 203.0.113.0/24 vrf vrf-123 \
+		nexthop via 198.51.100.2 dev dummy0 \
+		nexthop via 192.0.2.2 dev dummy1
+	set +e
+
+	ipv4_mpath_oif_test_common "oif dummy0" "dummy0" \
+		"IPv4 multipath via first nexthop"
+
+	ipv4_mpath_oif_test_common "oif dummy1" "dummy1" \
+		"IPv4 multipath via second nexthop"
+
+	ipv4_mpath_oif_test_common "oif dummy0 from 192.0.2.17" "dummy0" \
+		"IPv4 multipath via first nexthop with source address"
+
+	ipv4_mpath_oif_test_common "oif dummy1 from 192.0.2.17" "dummy1" \
+		"IPv4 multipath via second nexthop with source address"
+
+	cleanup
+}
+
+ipv6_mpath_oif_test_common()
+{
+	local get_param=$1; shift
+	local expected_oif=$1; shift
+	local test_name=$1; shift
+	local tmp_file
+
+	tmp_file=$(mktemp)
+
+	for i in {1..100}; do
+		$IP route get 2001:db8:10::${i} $get_param >> "$tmp_file"
+	done
+
+	[[ $(grep "$expected_oif" "$tmp_file" | wc -l) -eq 100 ]]
+	log_test $? 0 "$test_name"
+
+	rm "$tmp_file"
+}
+
+ipv6_mpath_oif_test()
+{
+	echo
+	echo "IPv6 multipath oif test"
+
+	setup
+
+	set -e
+	$IP link add dummy1 up type dummy
+	$IP address add 2001:db8:2::1/64 dev dummy1
+	$IP address add 2001:db8:100::1/128 dev lo
+
+	$IP route add 2001:db8:10::/64 \
+		nexthop via 2001:db8:1::2 dev dummy0 \
+		nexthop via 2001:db8:2::2 dev dummy1
+	set +e
+
+	ipv6_mpath_oif_test_common "oif dummy0" "dummy0" \
+		"IPv6 multipath via first nexthop"
+
+	ipv6_mpath_oif_test_common "oif dummy1" "dummy1" \
+		"IPv6 multipath via second nexthop"
+
+	ipv6_mpath_oif_test_common "oif dummy0 from 2001:db8:100::1" "dummy0" \
+		"IPv6 multipath via first nexthop with source address"
+
+	ipv6_mpath_oif_test_common "oif dummy1 from 2001:db8:100::1" "dummy1" \
+		"IPv6 multipath via second nexthop with source address"
+
+	cleanup
+}
+
+ipv6_mpath_oif_nh_test()
+{
+	echo
+	echo "IPv6 multipath oif with nexthop object test"
+
+	setup
+
+	set -e
+	$IP link add dummy1 up type dummy
+	$IP address add 2001:db8:2::1/64 dev dummy1
+	$IP address add 2001:db8:100::1/128 dev lo
+
+	$IP nexthop add id 1 via 2001:db8:1::2 dev dummy0
+	$IP nexthop add id 2 via 2001:db8:2::2 dev dummy1
+	$IP nexthop add id 3 group 1/2
+	$IP route add 2001:db8:10::/64 nhid 3
+	set +e
+
+	ipv6_mpath_oif_test_common "oif dummy0" "dummy0" \
+		"IPv6 multipath via first nexthop"
+
+	ipv6_mpath_oif_test_common "oif dummy1" "dummy1" \
+		"IPv6 multipath via second nexthop"
+
+	ipv6_mpath_oif_test_common "oif dummy0 from 2001:db8:100::1" "dummy0" \
+		"IPv6 multipath via first nexthop with source address"
+
+	ipv6_mpath_oif_test_common "oif dummy1 from 2001:db8:100::1" "dummy1" \
+		"IPv6 multipath via second nexthop with source address"
+
+	cleanup
+}
+
+ipv6_mpath_oif_vrf_test()
+{
+	echo
+	echo "IPv6 multipath oif with VRF test"
+
+	setup
+
+	set -e
+	$NS_EXEC sysctl -qw net.ipv6.conf.all.keep_addr_on_down=1
+	$IP -6 rule add pref 32765 table local
+	$IP -6 rule del pref 0
+	$IP link add name vrf-123 up type vrf table 123
+	$IP link set dev dummy0 master vrf-123
+	$IP link add dummy1 up master vrf-123 type dummy
+	$IP address add 2001:db8:2::1/64 dev dummy1
+	$IP address add 2001:db8:100::1/128 dev vrf-123
+
+	$IP route add 2001:db8:10::/64 vrf vrf-123 \
+		nexthop via 2001:db8:1::2 dev dummy0 \
+		nexthop via 2001:db8:2::2 dev dummy1
+	set +e
+
+	ipv6_mpath_oif_test_common "oif dummy0" "dummy0" \
+		"IPv6 multipath via first nexthop"
+
+	ipv6_mpath_oif_test_common "oif dummy1" "dummy1" \
+		"IPv6 multipath via second nexthop"
+
+	ipv6_mpath_oif_test_common "oif dummy0 from 2001:db8:100::1" "dummy0" \
+		"IPv6 multipath via first nexthop with source address"
+
+	ipv6_mpath_oif_test_common "oif dummy1 from 2001:db8:100::1" "dummy1" \
+		"IPv6 multipath via second nexthop with source address"
+
+	cleanup
+}
+
 ################################################################################
 # usage
 
@@ -3057,6 +3300,12 @@ do
 	ipv4_mpath_balance)		ipv4_mpath_balance_test;;
 	ipv6_mpath_balance)		ipv6_mpath_balance_test;;
 	ipv4_mpath_balance_preferred)	ipv4_mpath_balance_preferred_test;;
+	ipv4_mpath_oif)			ipv4_mpath_oif_test;;
+	ipv4_mpath_oif_nh)		ipv4_mpath_oif_nh_test;;
+	ipv4_mpath_oif_vrf)		ipv4_mpath_oif_vrf_test;;
+	ipv6_mpath_oif)			ipv6_mpath_oif_test;;
+	ipv6_mpath_oif_nh)		ipv6_mpath_oif_nh_test;;
+	ipv6_mpath_oif_vrf)		ipv6_mpath_oif_vrf_test;;
 	fib6_ra_to_static)		fib6_ra_to_static;;
 	fib6_temp_addr_renewal)		fib6_temp_addr_renewal;;