lists.openwall.net   lists  /  announce  owl-users  owl-dev  john-users  john-dev  passwdqc-users  yescrypt  popa3d-users  /  oss-security  kernel-hardening  musl  sabotage  tlsify  passwords  /  crypt-dev  xvendor  /  Bugtraq  Full-Disclosure  linux-kernel  linux-netdev  linux-ext4  linux-hardening  linux-cve-announce  PHC 
Open Source and information security mailing list archives
 
Hash Suite: Windows password security audit tool. GUI, reports in PDF.
[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-Id: <4f9916058458b8d802ce47f5d19aba213e50b6bc.1657750543.git.jhpark1013@gmail.com>
Date:   Wed, 13 Jul 2022 15:37:19 -0700
From:   Jaehee Park <jhpark1013@...il.com>
To:     netdev@...r.kernel.org
Cc:     davem@...emloft.net, yoshfuji@...ux-ipv6.org, dsahern@...nel.org,
        dsahern@...il.com, edumazet@...gle.com, kuba@...nel.org,
        pabeni@...hat.com, shuah@...nel.org, linux-kernel@...r.kernel.org,
        aajith@...sta.com, roopa@...dia.com, roopa.prabhu@...il.com,
        aroulin@...dia.com, sbrivio@...hat.com, jhpark1013@...il.com
Subject: [PATCH v2 net-next 3/3] selftests: net: arp_ndisc_untracked_subnets: test for arp_accept and accept_untracked_na

ipv4 arp_accept has a new option '2' to create new neighbor entries
only if the src ip is in the same subnet as an address configured on
the interface that received the garp message. This selftest tests all
options in arp_accept.

ipv6 has a sysctl endpoint, accept_untracked_na, that defines the
behavior for accepting untracked neighbor advertisements. A new option
similar to that of arp_accept for learning only from the same subnet is
added to accept_untracked_na. This selftest tests this new feature.

Signed-off-by: Jaehee Park <jhpark1013@...il.com>
Suggested-by: Roopa Prabhu <roopa@...dia.com>
---
 tools/testing/selftests/net/Makefile          |   1 +
 .../net/arp_ndisc_untracked_subnets.sh        | 308 ++++++++++++++++++
 2 files changed, 309 insertions(+)
 create mode 100755 tools/testing/selftests/net/arp_ndisc_untracked_subnets.sh

diff --git a/tools/testing/selftests/net/Makefile b/tools/testing/selftests/net/Makefile
index ddad703ace34..9c2e9e303c37 100644
--- a/tools/testing/selftests/net/Makefile
+++ b/tools/testing/selftests/net/Makefile
@@ -38,6 +38,7 @@ TEST_PROGS += srv6_end_dt6_l3vpn_test.sh
 TEST_PROGS += vrf_strict_mode_test.sh
 TEST_PROGS += arp_ndisc_evict_nocarrier.sh
 TEST_PROGS += ndisc_unsolicited_na_test.sh
+TEST_PROGS += arp_ndisc_untracked_subnets.sh
 TEST_PROGS += stress_reuseport_listen.sh
 TEST_PROGS_EXTENDED := in_netns.sh setup_loopback.sh setup_veth.sh
 TEST_PROGS_EXTENDED += toeplitz_client.sh toeplitz.sh
diff --git a/tools/testing/selftests/net/arp_ndisc_untracked_subnets.sh b/tools/testing/selftests/net/arp_ndisc_untracked_subnets.sh
new file mode 100755
index 000000000000..689ecfacee10
--- /dev/null
+++ b/tools/testing/selftests/net/arp_ndisc_untracked_subnets.sh
@@ -0,0 +1,308 @@
+#!/bin/bash
+# SPDX-License-Identifier: GPL-2.0
+#
+# 2 namespaces: one host and one router. Use arping from the host to send a
+# garp to the router. Router accepts or ignores based on its arp_accept
+# or accept_untracked_na configuration.
+
+TESTS="arp ndisc"
+
+ROUTER_NS="ns-router"
+ROUTER_NS_V6="ns-router-v6"
+ROUTER_INTF="veth-router"
+ROUTER_ADDR="10.0.10.1"
+ROUTER_ADDR_V6="2001:db8:abcd:0012::1"
+
+HOST_NS="ns-host"
+HOST_NS_V6="ns-host-v6"
+HOST_INTF="veth-host"
+HOST_ADDR="10.0.10.2"
+HOST_ADDR_V6="2001:db8:abcd:0012::2"
+
+SUBNET_WIDTH=24
+PREFIX_WIDTH_V6=64
+
+cleanup() {
+        ip netns del ${HOST_NS}
+        ip netns del ${ROUTER_NS}
+}
+
+cleanup_v6() {
+        ip netns del ${HOST_NS_V6}
+        ip netns del ${ROUTER_NS_V6}
+}
+
+setup() {
+        set -e
+        local arp_accept=$1
+
+        # Set up two namespaces
+        ip netns add ${ROUTER_NS}
+        ip netns add ${HOST_NS}
+
+        # Set up interfaces veth0 and veth1, which are pairs in separate
+        # namespaces. veth0 is veth-router, veth1 is veth-host.
+        # first, set up the inteface's link to the namespace
+        # then, set the interface "up"
+        ip netns exec ${ROUTER_NS} ip link add name ${ROUTER_INTF} \
+                type veth peer name ${HOST_INTF}
+
+        ip netns exec ${ROUTER_NS} ip link set dev ${ROUTER_INTF} up
+        ip netns exec ${ROUTER_NS} ip link set dev ${HOST_INTF} netns ${HOST_NS}
+
+        ip netns exec ${HOST_NS} ip link set dev ${HOST_INTF} up
+        ip netns exec ${ROUTER_NS} ip addr add ${ROUTER_ADDR}/${SUBNET_WIDTH} \
+                dev ${ROUTER_INTF}
+
+        ip netns exec ${HOST_NS} ip addr add ${HOST_ADDR}/${SUBNET_WIDTH} \
+                dev ${HOST_INTF}
+        ip netns exec ${HOST_NS} ip route add default via ${HOST_ADDR} \
+                dev ${HOST_INTF}
+        ip netns exec ${ROUTER_NS} ip route add default via ${ROUTER_ADDR} \
+                dev ${ROUTER_INTF}
+
+        ROUTER_CONF=net.ipv4.conf.${ROUTER_INTF}
+        ip netns exec ${ROUTER_NS} sysctl -w \
+                ${ROUTER_CONF}.arp_accept=${arp_accept} >/dev/null 2>&1
+        set +e
+}
+
+setup_v6() {
+        set -e
+        local accept_untracked_na=$1
+
+        # Set up two namespaces
+        ip netns add ${ROUTER_NS_V6}
+        ip netns add ${HOST_NS_V6}
+
+        # Set up interfaces veth0 and veth1, which are pairs in separate
+        # namespaces. veth0 is veth-router, veth1 is veth-host.
+        # first, set up the inteface's link to the namespace
+        # then, set the interface "up"
+        ip -6 -netns ${ROUTER_NS_V6} link add name ${ROUTER_INTF} \
+                type veth peer name ${HOST_INTF}
+
+        ip -6 -netns ${ROUTER_NS_V6} link set dev ${ROUTER_INTF} up
+        ip -6 -netns ${ROUTER_NS_V6} link set dev ${HOST_INTF} netns \
+                ${HOST_NS_V6}
+
+        ip -6 -netns ${HOST_NS_V6} link set dev ${HOST_INTF} up
+        ip -6 -netns ${ROUTER_NS_V6} addr add \
+                ${ROUTER_ADDR_V6}/${PREFIX_WIDTH_V6} dev ${ROUTER_INTF} nodad
+
+        HOST_CONF=net.ipv6.conf.${HOST_INTF}
+        ip netns exec ${HOST_NS_V6} sysctl -qw ${HOST_CONF}.ndisc_notify=1
+        ip netns exec ${HOST_NS_V6} sysctl -qw ${HOST_CONF}.disable_ipv6=0
+        ip -6 -netns ${HOST_NS_V6} addr add ${HOST_ADDR_V6}/${PREFIX_WIDTH_V6} \
+                dev ${HOST_INTF}
+
+        ROUTER_CONF=net.ipv6.conf.${ROUTER_INTF}
+
+        ip netns exec ${ROUTER_NS_V6} sysctl -w \
+                ${ROUTER_CONF}.forwarding=1 >/dev/null 2>&1
+        ip netns exec ${ROUTER_NS_V6} sysctl -w \
+                ${ROUTER_CONF}.drop_unsolicited_na=0 >/dev/null 2>&1
+        ip netns exec ${ROUTER_NS_V6} sysctl -w \
+                ${ROUTER_CONF}.accept_untracked_na=${accept_untracked_na} \
+                >/dev/null 2>&1
+        set +e
+}
+
+verify_arp() {
+        local arp_accept=$1
+        local same_subnet=$2
+
+        neigh_show_output=$(ip netns exec ${ROUTER_NS} ip neigh get \
+                ${HOST_ADDR} dev ${ROUTER_INTF} 2>/dev/null)
+
+        if [ ${arp_accept} -eq 1 ]; then
+                # Neighbor entries expected
+                [[ ${neigh_show_output} ]]
+        elif [ ${arp_accept} -eq 2 ]; then
+                if [ ${same_subnet} -eq 1 ]; then
+                        # Neighbor entries expected
+                        [[ ${neigh_show_output} ]]
+                else
+                        [[ -z "${neigh_show_output}" ]]
+                fi
+        else
+                [[ -z "${neigh_show_output}" ]]
+        fi
+ }
+
+arp_test_gratuitous() {
+        set -e
+        local arp_accept=$1
+        local same_subnet=$2
+
+        if [ ${arp_accept} -eq 2 ]; then
+                test_msg=("test_arp: "
+                          "accept_arp=$1 "
+                          "same_subnet=$2")
+                if [ ${same_subnet} -eq 0 ]; then
+                        HOST_ADDR=10.0.11.3
+                else
+                        HOST_ADDR=10.0.10.3
+                fi
+        else
+                test_msg=("test_arp: "
+                          "accept_arp=$1")
+        fi
+        # Supply arp_accept option to set up which sets it in sysctl
+        setup ${arp_accept}
+        ip netns exec ${HOST_NS} arping -A -U ${HOST_ADDR} -c1 2>&1 >/dev/null
+
+        if verify_arp $1 $2; then
+                printf "    TEST: %-60s  [ OK ]\n" "${test_msg[*]}"
+        else
+                printf "    TEST: %-60s  [FAIL]\n" "${test_msg[*]}"
+        fi
+        cleanup
+        set +e
+}
+
+arp_test_gratuitous_combinations() {
+        arp_test_gratuitous 0
+        arp_test_gratuitous 1
+        arp_test_gratuitous 2 0 # Second entry indicates subnet or not
+        arp_test_gratuitous 2 1
+}
+
+cleanup_tcpdump() {
+        set -e
+        [[ ! -z  ${tcpdump_stdout} ]] && rm -f ${tcpdump_stdout}
+        [[ ! -z  ${tcpdump_stderr} ]] && rm -f ${tcpdump_stderr}
+        tcpdump_stdout=
+        tcpdump_stderr=
+        set +e
+}
+
+start_tcpdump() {
+        set -e
+        tcpdump_stdout=`mktemp`
+        tcpdump_stderr=`mktemp`
+        ip netns exec ${ROUTER_NS_V6} timeout 15s \
+                tcpdump --immediate-mode -tpni ${ROUTER_INTF} -c 1 \
+                "icmp6 && icmp6[0] == 136 && src ${HOST_ADDR_V6}" \
+                > ${tcpdump_stdout} 2> /dev/null
+        set +e
+}
+
+verify_ndisc() {
+        local accept_untracked_na=$1
+        local same_subnet=$2
+
+        neigh_show_output=$(ip -6 -netns ${ROUTER_NS_V6} neigh show \
+                to ${HOST_ADDR_V6} dev ${ROUTER_INTF} nud stale)
+
+        if [ ${accept_untracked_na} -eq 1 ]; then
+                # Neighbour entry expected to be present
+                [[ ${neigh_show_output} ]]
+        elif [ ${accept_untracked_na} -eq 2 ]; then
+                if [ ${same_subnet} -eq 1 ]; then
+                        [[ ${neigh_show_output} ]]
+                else
+                        [[ -z "${neigh_show_output}" ]]
+                fi
+        else
+                # Neighbour entry expected to be absent for all other cases
+                [[ -z "${neigh_show_output}" ]]
+        fi
+}
+
+ndisc_test_untracked_advertisements() {
+        set -e
+        test_msg=("test_ndisc: "
+                  "accept_untracked_na=$1")
+
+        local accept_untracked_na=$1
+        local same_subnet=$2
+        if [ ${accept_untracked_na} -eq 2 ]; then
+                test_msg=("test_ndisc: "
+                          "accept_untracked_na=$1 "
+                          "same_subnet=$2")
+                if [ ${same_subnet} -eq 0 ]; then
+                        # Not same subnet
+                        HOST_ADDR_V6=2000:db8:abcd:0013::4
+                else
+                        HOST_ADDR_V6=2001:db8:abcd:0012::3
+                fi
+        fi
+        setup_v6 $1 $2
+        start_tcpdump
+
+        if verify_ndisc $1 $2; then
+                printf "    TEST: %-60s  [ OK ]\n" "${test_msg[*]}"
+        else
+                printf "    TEST: %-60s  [FAIL]\n" "${test_msg[*]}"
+        fi
+
+        cleanup_tcpdump
+        cleanup_v6
+        set +e
+}
+
+ndisc_test_untracked_combinations() {
+        ndisc_test_untracked_advertisements 0
+        ndisc_test_untracked_advertisements 1
+        ndisc_test_untracked_advertisements 2 0
+        ndisc_test_untracked_advertisements 2 1
+}
+
+################################################################################
+# usage
+
+usage()
+{
+        cat <<EOF
+usage: ${0##*/} OPTS
+
+        -t <test>       Test(s) to run (default: all)
+                        (options: $TESTS)
+EOF
+}
+
+################################################################################
+# main
+
+while getopts ":t:h" opt; do
+        case $opt in
+                t) TESTS=$OPTARG;;
+                h) usage; exit 0;;
+                *) usage; exit 1;;
+        esac
+done
+
+if [ "$(id -u)" -ne 0 ];then
+	echo "SKIP: Need root privileges"
+	exit $ksft_skip;
+fi
+
+if [ ! -x "$(command -v ip)" ]; then
+	echo "SKIP: Could not run test without ip tool"
+	exit $ksft_skip
+fi
+
+if [ ! -x "$(command -v tcpdump)" ]; then
+	echo "SKIP: Could not run test without tcpdump tool"
+	exit $ksft_skip
+fi
+
+if [ ! -x "$(command -v arping)" ]; then
+	echo "SKIP: Could not run test without arping tool"
+	exit $ksft_skip
+fi
+
+# start clean
+cleanup &> /dev/null
+cleanup_v6 &> /dev/null
+
+for t in $TESTS
+do
+        case $t in
+        arp_test_gratuitous_combinations|arp) arp_test_gratuitous_combinations;;
+        ndisc_test_untracked_combinations|ndisc) \
+                ndisc_test_untracked_combinations;;
+        help) echo "Test names: $TESTS"; exit 0;;
+esac
+done
-- 
2.30.2

Powered by blists - more mailing lists

Powered by Openwall GNU/*/Linux Powered by OpenVZ