diff options
Diffstat (limited to 'tools/testing/selftests/netfilter')
-rw-r--r-- | tools/testing/selftests/netfilter/.gitignore | 2 | ||||
-rw-r--r-- | tools/testing/selftests/netfilter/Makefile | 13 | ||||
-rwxr-xr-x | tools/testing/selftests/netfilter/bridge_brouter.sh | 146 | ||||
-rw-r--r-- | tools/testing/selftests/netfilter/config | 8 | ||||
-rwxr-xr-x | tools/testing/selftests/netfilter/conntrack_icmp_related.sh | 315 | ||||
-rwxr-xr-x | tools/testing/selftests/netfilter/conntrack_vrf.sh | 241 | ||||
-rwxr-xr-x | tools/testing/selftests/netfilter/ipvs.sh | 228 | ||||
-rw-r--r-- | tools/testing/selftests/netfilter/nf-queue.c | 395 | ||||
-rwxr-xr-x | tools/testing/selftests/netfilter/nft_concat_range.sh | 1586 | ||||
-rwxr-xr-x | tools/testing/selftests/netfilter/nft_conntrack_helper.sh | 181 | ||||
-rwxr-xr-x | tools/testing/selftests/netfilter/nft_flowtable.sh | 420 | ||||
-rwxr-xr-x | tools/testing/selftests/netfilter/nft_meta.sh | 142 | ||||
-rwxr-xr-x | tools/testing/selftests/netfilter/nft_nat.sh | 914 | ||||
-rwxr-xr-x | tools/testing/selftests/netfilter/nft_queue.sh | 376 | ||||
-rwxr-xr-x | tools/testing/selftests/netfilter/nft_trans_stress.sh | 78 |
15 files changed, 5045 insertions, 0 deletions
diff --git a/tools/testing/selftests/netfilter/.gitignore b/tools/testing/selftests/netfilter/.gitignore new file mode 100644 index 000000000..8448f74ad --- /dev/null +++ b/tools/testing/selftests/netfilter/.gitignore @@ -0,0 +1,2 @@ +# SPDX-License-Identifier: GPL-2.0-only +nf-queue diff --git a/tools/testing/selftests/netfilter/Makefile b/tools/testing/selftests/netfilter/Makefile new file mode 100644 index 000000000..a56cfc4f2 --- /dev/null +++ b/tools/testing/selftests/netfilter/Makefile @@ -0,0 +1,13 @@ +# SPDX-License-Identifier: GPL-2.0 +# Makefile for netfilter selftests + +TEST_PROGS := nft_trans_stress.sh nft_nat.sh bridge_brouter.sh \ + conntrack_icmp_related.sh nft_flowtable.sh ipvs.sh \ + nft_concat_range.sh nft_conntrack_helper.sh \ + nft_queue.sh nft_meta.sh \ + conntrack_vrf.sh + +LDLIBS = -lmnl +TEST_GEN_FILES = nf-queue + +include ../lib.mk diff --git a/tools/testing/selftests/netfilter/bridge_brouter.sh b/tools/testing/selftests/netfilter/bridge_brouter.sh new file mode 100755 index 000000000..29f3955b9 --- /dev/null +++ b/tools/testing/selftests/netfilter/bridge_brouter.sh @@ -0,0 +1,146 @@ +#!/bin/bash +# +# This test is for bridge 'brouting', i.e. make some packets being routed +# rather than getting bridged even though they arrive on interface that is +# part of a bridge. + +# eth0 br0 eth0 +# setup is: ns1 <-> ns0 <-> ns2 + +# Kselftest framework requirement - SKIP code is 4. +ksft_skip=4 +ret=0 + +ebtables -V > /dev/null 2>&1 +if [ $? -ne 0 ];then + echo "SKIP: Could not run test without ebtables" + exit $ksft_skip +fi + +ip -Version > /dev/null 2>&1 +if [ $? -ne 0 ];then + echo "SKIP: Could not run test without ip tool" + exit $ksft_skip +fi + +ip netns add ns0 +ip netns add ns1 +ip netns add ns2 + +ip link add veth0 netns ns0 type veth peer name eth0 netns ns1 +if [ $? -ne 0 ]; then + echo "SKIP: Can't create veth device" + exit $ksft_skip +fi +ip link add veth1 netns ns0 type veth peer name eth0 netns ns2 + +ip -net ns0 link set lo up +ip -net ns0 link set veth0 up +ip -net ns0 link set veth1 up + +ip -net ns0 link add br0 type bridge +if [ $? -ne 0 ]; then + echo "SKIP: Can't create bridge br0" + exit $ksft_skip +fi + +ip -net ns0 link set veth0 master br0 +ip -net ns0 link set veth1 master br0 +ip -net ns0 link set br0 up +ip -net ns0 addr add 10.0.0.1/24 dev br0 + +# place both in same subnet, ns1 and ns2 connected via ns0:br0 +for i in 1 2; do + ip -net ns$i link set lo up + ip -net ns$i link set eth0 up + ip -net ns$i addr add 10.0.0.1$i/24 dev eth0 +done + +test_ebtables_broute() +{ + local cipt + + # redirect is needed so the dstmac is rewritten to the bridge itself, + # ip stack won't process OTHERHOST (foreign unicast mac) packets. + ip netns exec ns0 ebtables -t broute -A BROUTING -p ipv4 --ip-protocol icmp -j redirect --redirect-target=DROP + if [ $? -ne 0 ]; then + echo "SKIP: Could not add ebtables broute redirect rule" + return $ksft_skip + fi + + # ping netns1, expected to not work (ip forwarding is off) + ip netns exec ns1 ping -q -c 1 10.0.0.12 > /dev/null 2>&1 + if [ $? -eq 0 ]; then + echo "ERROR: ping works, should have failed" 1>&2 + return 1 + fi + + # enable forwarding on both interfaces. + # neither needs an ip address, but at least the bridge needs + # an ip address in same network segment as ns1 and ns2 (ns0 + # needs to be able to determine route for to-be-forwarded packet). + ip netns exec ns0 sysctl -q net.ipv4.conf.veth0.forwarding=1 + ip netns exec ns0 sysctl -q net.ipv4.conf.veth1.forwarding=1 + + sleep 1 + + ip netns exec ns1 ping -q -c 1 10.0.0.12 > /dev/null + if [ $? -ne 0 ]; then + echo "ERROR: ping did not work, but it should (broute+forward)" 1>&2 + return 1 + fi + + echo "PASS: ns1/ns2 connectivity with active broute rule" + ip netns exec ns0 ebtables -t broute -F + + # ping netns1, expected to work (frames are bridged) + ip netns exec ns1 ping -q -c 1 10.0.0.12 > /dev/null + if [ $? -ne 0 ]; then + echo "ERROR: ping did not work, but it should (bridged)" 1>&2 + return 1 + fi + + ip netns exec ns0 ebtables -t filter -A FORWARD -p ipv4 --ip-protocol icmp -j DROP + + # ping netns1, expected to not work (DROP in bridge forward) + ip netns exec ns1 ping -q -c 1 10.0.0.12 > /dev/null 2>&1 + if [ $? -eq 0 ]; then + echo "ERROR: ping works, should have failed (icmp forward drop)" 1>&2 + return 1 + fi + + # re-activate brouter + ip netns exec ns0 ebtables -t broute -A BROUTING -p ipv4 --ip-protocol icmp -j redirect --redirect-target=DROP + + ip netns exec ns2 ping -q -c 1 10.0.0.11 > /dev/null + if [ $? -ne 0 ]; then + echo "ERROR: ping did not work, but it should (broute+forward 2)" 1>&2 + return 1 + fi + + echo "PASS: ns1/ns2 connectivity with active broute rule and bridge forward drop" + return 0 +} + +# test basic connectivity +ip netns exec ns1 ping -c 1 -q 10.0.0.12 > /dev/null +if [ $? -ne 0 ]; then + echo "ERROR: Could not reach ns2 from ns1" 1>&2 + ret=1 +fi + +ip netns exec ns2 ping -c 1 -q 10.0.0.11 > /dev/null +if [ $? -ne 0 ]; then + echo "ERROR: Could not reach ns1 from ns2" 1>&2 + ret=1 +fi + +if [ $ret -eq 0 ];then + echo "PASS: netns connectivity: ns1 and ns2 can reach each other" +fi + +test_ebtables_broute +ret=$? +for i in 0 1 2; do ip netns del ns$i;done + +exit $ret diff --git a/tools/testing/selftests/netfilter/config b/tools/testing/selftests/netfilter/config new file mode 100644 index 000000000..4faf2ce02 --- /dev/null +++ b/tools/testing/selftests/netfilter/config @@ -0,0 +1,8 @@ +CONFIG_NET_NS=y +CONFIG_NF_TABLES_INET=y +CONFIG_NFT_QUEUE=m +CONFIG_NFT_NAT=m +CONFIG_NFT_REDIR=m +CONFIG_NFT_MASQ=m +CONFIG_NFT_FLOW_OFFLOAD=m +CONFIG_NF_CT_NETLINK=m diff --git a/tools/testing/selftests/netfilter/conntrack_icmp_related.sh b/tools/testing/selftests/netfilter/conntrack_icmp_related.sh new file mode 100755 index 000000000..76645aaf2 --- /dev/null +++ b/tools/testing/selftests/netfilter/conntrack_icmp_related.sh @@ -0,0 +1,315 @@ +#!/bin/bash +# +# check that ICMP df-needed/pkttoobig icmp are set are set as related +# state +# +# Setup is: +# +# nsclient1 -> nsrouter1 -> nsrouter2 -> nsclient2 +# MTU 1500, except for nsrouter2 <-> nsclient2 link (1280). +# ping nsclient2 from nsclient1, checking that conntrack did set RELATED +# 'fragmentation needed' icmp packet. +# +# In addition, nsrouter1 will perform IP masquerading, i.e. also +# check the icmp errors are propagated to the correct host as per +# nat of "established" icmp-echo "connection". + +# Kselftest framework requirement - SKIP code is 4. +ksft_skip=4 +ret=0 + +nft --version > /dev/null 2>&1 +if [ $? -ne 0 ];then + echo "SKIP: Could not run test without nft tool" + exit $ksft_skip +fi + +ip -Version > /dev/null 2>&1 +if [ $? -ne 0 ];then + echo "SKIP: Could not run test without ip tool" + exit $ksft_skip +fi + +cleanup() { + for i in 1 2;do ip netns del nsclient$i;done + for i in 1 2;do ip netns del nsrouter$i;done +} + +trap cleanup EXIT + +ipv4() { + echo -n 192.168.$1.2 +} + +ipv6 () { + echo -n dead:$1::2 +} + +check_counter() +{ + ns=$1 + name=$2 + expect=$3 + local lret=0 + + cnt=$(ip netns exec $ns nft list counter inet filter "$name" | grep -q "$expect") + if [ $? -ne 0 ]; then + echo "ERROR: counter $name in $ns has unexpected value (expected $expect)" 1>&2 + ip netns exec $ns nft list counter inet filter "$name" 1>&2 + lret=1 + fi + + return $lret +} + +check_unknown() +{ + expect="packets 0 bytes 0" + for n in nsclient1 nsclient2 nsrouter1 nsrouter2; do + check_counter $n "unknown" "$expect" + if [ $? -ne 0 ] ;then + return 1 + fi + done + + return 0 +} + +for n in nsclient1 nsclient2 nsrouter1 nsrouter2; do + ip netns add $n + ip -net $n link set lo up +done + +DEV=veth0 +ip link add $DEV netns nsclient1 type veth peer name eth1 netns nsrouter1 +DEV=veth0 +ip link add $DEV netns nsclient2 type veth peer name eth1 netns nsrouter2 + +DEV=veth0 +ip link add $DEV netns nsrouter1 type veth peer name eth2 netns nsrouter2 + +DEV=veth0 +for i in 1 2; do + ip -net nsclient$i link set $DEV up + ip -net nsclient$i addr add $(ipv4 $i)/24 dev $DEV + ip -net nsclient$i addr add $(ipv6 $i)/64 dev $DEV +done + +ip -net nsrouter1 link set eth1 up +ip -net nsrouter1 link set veth0 up + +ip -net nsrouter2 link set eth1 up +ip -net nsrouter2 link set eth2 up + +ip -net nsclient1 route add default via 192.168.1.1 +ip -net nsclient1 -6 route add default via dead:1::1 + +ip -net nsclient2 route add default via 192.168.2.1 +ip -net nsclient2 route add default via dead:2::1 + +i=3 +ip -net nsrouter1 addr add 192.168.1.1/24 dev eth1 +ip -net nsrouter1 addr add 192.168.3.1/24 dev veth0 +ip -net nsrouter1 addr add dead:1::1/64 dev eth1 +ip -net nsrouter1 addr add dead:3::1/64 dev veth0 +ip -net nsrouter1 route add default via 192.168.3.10 +ip -net nsrouter1 -6 route add default via dead:3::10 + +ip -net nsrouter2 addr add 192.168.2.1/24 dev eth1 +ip -net nsrouter2 addr add 192.168.3.10/24 dev eth2 +ip -net nsrouter2 addr add dead:2::1/64 dev eth1 +ip -net nsrouter2 addr add dead:3::10/64 dev eth2 +ip -net nsrouter2 route add default via 192.168.3.1 +ip -net nsrouter2 route add default via dead:3::1 + +sleep 2 +for i in 4 6; do + ip netns exec nsrouter1 sysctl -q net.ipv$i.conf.all.forwarding=1 + ip netns exec nsrouter2 sysctl -q net.ipv$i.conf.all.forwarding=1 +done + +for netns in nsrouter1 nsrouter2; do +ip netns exec $netns nft -f - <<EOF +table inet filter { + counter unknown { } + counter related { } + chain forward { + type filter hook forward priority 0; policy accept; + meta l4proto icmpv6 icmpv6 type "packet-too-big" ct state "related" counter name "related" accept + meta l4proto icmp icmp type "destination-unreachable" ct state "related" counter name "related" accept + meta l4proto { icmp, icmpv6 } ct state new,established accept + counter name "unknown" drop + } +} +EOF +done + +ip netns exec nsclient1 nft -f - <<EOF +table inet filter { + counter unknown { } + counter related { } + counter redir4 { } + counter redir6 { } + chain input { + type filter hook input priority 0; policy accept; + + icmp type "redirect" ct state "related" counter name "redir4" accept + icmpv6 type "nd-redirect" ct state "related" counter name "redir6" accept + + meta l4proto { icmp, icmpv6 } ct state established,untracked accept + meta l4proto { icmp, icmpv6 } ct state "related" counter name "related" accept + + counter name "unknown" drop + } +} +EOF + +ip netns exec nsclient2 nft -f - <<EOF +table inet filter { + counter unknown { } + counter new { } + counter established { } + + chain input { + type filter hook input priority 0; policy accept; + meta l4proto { icmp, icmpv6 } ct state established,untracked accept + + meta l4proto { icmp, icmpv6 } ct state "new" counter name "new" accept + meta l4proto { icmp, icmpv6 } ct state "established" counter name "established" accept + counter name "unknown" drop + } + chain output { + type filter hook output priority 0; policy accept; + meta l4proto { icmp, icmpv6 } ct state established,untracked accept + + meta l4proto { icmp, icmpv6 } ct state "new" counter name "new" + meta l4proto { icmp, icmpv6 } ct state "established" counter name "established" + counter name "unknown" drop + } +} +EOF + + +# make sure NAT core rewrites adress of icmp error if nat is used according to +# conntrack nat information (icmp error will be directed at nsrouter1 address, +# but it needs to be routed to nsclient1 address). +ip netns exec nsrouter1 nft -f - <<EOF +table ip nat { + chain postrouting { + type nat hook postrouting priority 0; policy accept; + ip protocol icmp oifname "veth0" counter masquerade + } +} +table ip6 nat { + chain postrouting { + type nat hook postrouting priority 0; policy accept; + ip6 nexthdr icmpv6 oifname "veth0" counter masquerade + } +} +EOF + +ip netns exec nsrouter2 ip link set eth1 mtu 1280 +ip netns exec nsclient2 ip link set veth0 mtu 1280 +sleep 1 + +ip netns exec nsclient1 ping -c 1 -s 1000 -q -M do 192.168.2.2 >/dev/null +if [ $? -ne 0 ]; then + echo "ERROR: netns ip routing/connectivity broken" 1>&2 + cleanup + exit 1 +fi +ip netns exec nsclient1 ping6 -q -c 1 -s 1000 dead:2::2 >/dev/null +if [ $? -ne 0 ]; then + echo "ERROR: netns ipv6 routing/connectivity broken" 1>&2 + cleanup + exit 1 +fi + +check_unknown +if [ $? -ne 0 ]; then + ret=1 +fi + +expect="packets 0 bytes 0" +for netns in nsrouter1 nsrouter2 nsclient1;do + check_counter "$netns" "related" "$expect" + if [ $? -ne 0 ]; then + ret=1 + fi +done + +expect="packets 2 bytes 2076" +check_counter nsclient2 "new" "$expect" +if [ $? -ne 0 ]; then + ret=1 +fi + +ip netns exec nsclient1 ping -q -c 1 -s 1300 -M do 192.168.2.2 > /dev/null +if [ $? -eq 0 ]; then + echo "ERROR: ping should have failed with PMTU too big error" 1>&2 + ret=1 +fi + +# nsrouter2 should have generated the icmp error, so +# related counter should be 0 (its in forward). +expect="packets 0 bytes 0" +check_counter "nsrouter2" "related" "$expect" +if [ $? -ne 0 ]; then + ret=1 +fi + +# but nsrouter1 should have seen it, same for nsclient1. +expect="packets 1 bytes 576" +for netns in nsrouter1 nsclient1;do + check_counter "$netns" "related" "$expect" + if [ $? -ne 0 ]; then + ret=1 + fi +done + +ip netns exec nsclient1 ping6 -c 1 -s 1300 dead:2::2 > /dev/null +if [ $? -eq 0 ]; then + echo "ERROR: ping6 should have failed with PMTU too big error" 1>&2 + ret=1 +fi + +expect="packets 2 bytes 1856" +for netns in nsrouter1 nsclient1;do + check_counter "$netns" "related" "$expect" + if [ $? -ne 0 ]; then + ret=1 + fi +done + +if [ $ret -eq 0 ];then + echo "PASS: icmp mtu error had RELATED state" +else + echo "ERROR: icmp error RELATED state test has failed" +fi + +# add 'bad' route, expect icmp REDIRECT to be generated +ip netns exec nsclient1 ip route add 192.168.1.42 via 192.168.1.1 +ip netns exec nsclient1 ip route add dead:1::42 via dead:1::1 + +ip netns exec "nsclient1" ping -q -c 2 192.168.1.42 > /dev/null + +expect="packets 1 bytes 112" +check_counter nsclient1 "redir4" "$expect" +if [ $? -ne 0 ];then + ret=1 +fi + +ip netns exec "nsclient1" ping -c 1 dead:1::42 > /dev/null +expect="packets 1 bytes 192" +check_counter nsclient1 "redir6" "$expect" +if [ $? -ne 0 ];then + ret=1 +fi + +if [ $ret -eq 0 ];then + echo "PASS: icmp redirects had RELATED state" +else + echo "ERROR: icmp redirect RELATED state test has failed" +fi + +exit $ret diff --git a/tools/testing/selftests/netfilter/conntrack_vrf.sh b/tools/testing/selftests/netfilter/conntrack_vrf.sh new file mode 100755 index 000000000..8b5ea9234 --- /dev/null +++ b/tools/testing/selftests/netfilter/conntrack_vrf.sh @@ -0,0 +1,241 @@ +#!/bin/sh + +# This script demonstrates interaction of conntrack and vrf. +# The vrf driver calls the netfilter hooks again, with oif/iif +# pointing at the VRF device. +# +# For ingress, this means first iteration has iifname of lower/real +# device. In this script, thats veth0. +# Second iteration is iifname set to vrf device, tvrf in this script. +# +# For egress, this is reversed: first iteration has the vrf device, +# second iteration is done with the lower/real/veth0 device. +# +# test_ct_zone_in demonstrates unexpected change of nftables +# behavior # caused by commit 09e856d54bda5f28 "vrf: Reset skb conntrack +# connection on VRF rcv" +# +# It was possible to assign conntrack zone to a packet (or mark it for +# `notracking`) in the prerouting chain before conntrack, based on real iif. +# +# After the change, the zone assignment is lost and the zone is assigned based +# on the VRF master interface (in case such a rule exists). +# assignment is lost. Instead, assignment based on the `iif` matching +# Thus it is impossible to distinguish packets based on the original +# interface. +# +# test_masquerade_vrf and test_masquerade_veth0 demonstrate the problem +# that was supposed to be fixed by the commit mentioned above to make sure +# that any fix to test case 1 won't break masquerade again. + +ksft_skip=4 + +IP0=172.30.30.1 +IP1=172.30.30.2 +PFXL=30 +ret=0 + +sfx=$(mktemp -u "XXXXXXXX") +ns0="ns0-$sfx" +ns1="ns1-$sfx" + +cleanup() +{ + ip netns pids $ns0 | xargs kill 2>/dev/null + ip netns pids $ns1 | xargs kill 2>/dev/null + + ip netns del $ns0 $ns1 +} + +nft --version > /dev/null 2>&1 +if [ $? -ne 0 ];then + echo "SKIP: Could not run test without nft tool" + exit $ksft_skip +fi + +ip -Version > /dev/null 2>&1 +if [ $? -ne 0 ];then + echo "SKIP: Could not run test without ip tool" + exit $ksft_skip +fi + +ip netns add "$ns0" +if [ $? -ne 0 ];then + echo "SKIP: Could not create net namespace $ns0" + exit $ksft_skip +fi +ip netns add "$ns1" + +trap cleanup EXIT + +ip netns exec $ns0 sysctl -q -w net.ipv4.conf.default.rp_filter=0 +ip netns exec $ns0 sysctl -q -w net.ipv4.conf.all.rp_filter=0 +ip netns exec $ns0 sysctl -q -w net.ipv4.conf.all.rp_filter=0 + +ip link add veth0 netns "$ns0" type veth peer name veth0 netns "$ns1" > /dev/null 2>&1 +if [ $? -ne 0 ];then + echo "SKIP: Could not add veth device" + exit $ksft_skip +fi + +ip -net $ns0 li add tvrf type vrf table 9876 +if [ $? -ne 0 ];then + echo "SKIP: Could not add vrf device" + exit $ksft_skip +fi + +ip -net $ns0 li set lo up + +ip -net $ns0 li set veth0 master tvrf +ip -net $ns0 li set tvrf up +ip -net $ns0 li set veth0 up +ip -net $ns1 li set veth0 up + +ip -net $ns0 addr add $IP0/$PFXL dev veth0 +ip -net $ns1 addr add $IP1/$PFXL dev veth0 + +ip netns exec $ns1 iperf3 -s > /dev/null 2>&1& +if [ $? -ne 0 ];then + echo "SKIP: Could not start iperf3" + exit $ksft_skip +fi + +# test vrf ingress handling. +# The incoming connection should be placed in conntrack zone 1, +# as decided by the first iteration of the ruleset. +test_ct_zone_in() +{ +ip netns exec $ns0 nft -f - <<EOF +table testct { + chain rawpre { + type filter hook prerouting priority raw; + + iif { veth0, tvrf } counter meta nftrace set 1 + iif veth0 counter ct zone set 1 counter return + iif tvrf counter ct zone set 2 counter return + ip protocol icmp counter + notrack counter + } + + chain rawout { + type filter hook output priority raw; + + oif veth0 counter ct zone set 1 counter return + oif tvrf counter ct zone set 2 counter return + notrack counter + } +} +EOF + ip netns exec $ns1 ping -W 1 -c 1 -I veth0 $IP0 > /dev/null + + # should be in zone 1, not zone 2 + count=$(ip netns exec $ns0 conntrack -L -s $IP1 -d $IP0 -p icmp --zone 1 2>/dev/null | wc -l) + if [ $count -eq 1 ]; then + echo "PASS: entry found in conntrack zone 1" + else + echo "FAIL: entry not found in conntrack zone 1" + count=$(ip netns exec $ns0 conntrack -L -s $IP1 -d $IP0 -p icmp --zone 2 2> /dev/null | wc -l) + if [ $count -eq 1 ]; then + echo "FAIL: entry found in zone 2 instead" + else + echo "FAIL: entry not in zone 1 or 2, dumping table" + ip netns exec $ns0 conntrack -L + ip netns exec $ns0 nft list ruleset + fi + fi +} + +# add masq rule that gets evaluated w. outif set to vrf device. +# This tests the first iteration of the packet through conntrack, +# oifname is the vrf device. +test_masquerade_vrf() +{ + local qdisc=$1 + + if [ "$qdisc" != "default" ]; then + tc -net $ns0 qdisc add dev tvrf root $qdisc + fi + + ip netns exec $ns0 conntrack -F 2>/dev/null + +ip netns exec $ns0 nft -f - <<EOF +flush ruleset +table ip nat { + chain rawout { + type filter hook output priority raw; + + oif tvrf ct state untracked counter + } + chain postrouting2 { + type filter hook postrouting priority mangle; + + oif tvrf ct state untracked counter + } + chain postrouting { + type nat hook postrouting priority 0; + # NB: masquerade should always be combined with 'oif(name) bla', + # lack of this is intentional here, we want to exercise double-snat. + ip saddr 172.30.30.0/30 counter masquerade random + } +} +EOF + ip netns exec $ns0 ip vrf exec tvrf iperf3 -t 1 -c $IP1 >/dev/null + if [ $? -ne 0 ]; then + echo "FAIL: iperf3 connect failure with masquerade + sport rewrite on vrf device" + ret=1 + return + fi + + # must also check that nat table was evaluated on second (lower device) iteration. + ip netns exec $ns0 nft list table ip nat |grep -q 'counter packets 2' && + ip netns exec $ns0 nft list table ip nat |grep -q 'untracked counter packets [1-9]' + if [ $? -eq 0 ]; then + echo "PASS: iperf3 connect with masquerade + sport rewrite on vrf device ($qdisc qdisc)" + else + echo "FAIL: vrf rules have unexpected counter value" + ret=1 + fi + + if [ "$qdisc" != "default" ]; then + tc -net $ns0 qdisc del dev tvrf root + fi +} + +# add masq rule that gets evaluated w. outif set to veth device. +# This tests the 2nd iteration of the packet through conntrack, +# oifname is the lower device (veth0 in this case). +test_masquerade_veth() +{ + ip netns exec $ns0 conntrack -F 2>/dev/null +ip netns exec $ns0 nft -f - <<EOF +flush ruleset +table ip nat { + chain postrouting { + type nat hook postrouting priority 0; + meta oif veth0 ip saddr 172.30.30.0/30 counter masquerade random + } +} +EOF + ip netns exec $ns0 ip vrf exec tvrf iperf3 -t 1 -c $IP1 > /dev/null + if [ $? -ne 0 ]; then + echo "FAIL: iperf3 connect failure with masquerade + sport rewrite on veth device" + ret=1 + return + fi + + # must also check that nat table was evaluated on second (lower device) iteration. + ip netns exec $ns0 nft list table ip nat |grep -q 'counter packets 2' + if [ $? -eq 0 ]; then + echo "PASS: iperf3 connect with masquerade + sport rewrite on veth device" + else + echo "FAIL: vrf masq rule has unexpected counter value" + ret=1 + fi +} + +test_ct_zone_in +test_masquerade_vrf "default" +test_masquerade_vrf "pfifo" +test_masquerade_veth + +exit $ret diff --git a/tools/testing/selftests/netfilter/ipvs.sh b/tools/testing/selftests/netfilter/ipvs.sh new file mode 100755 index 000000000..c3b8f90c4 --- /dev/null +++ b/tools/testing/selftests/netfilter/ipvs.sh @@ -0,0 +1,228 @@ +#!/bin/sh +# SPDX-License-Identifier: GPL-2.0 +# +# End-to-end ipvs test suite +# Topology: +#--------------------------------------------------------------+ +# | | +# ns0 | ns1 | +# ----------- | ----------- ----------- | +# | veth01 | --------- | veth10 | | veth12 | | +# ----------- peer ----------- ----------- | +# | | | | +# ----------- | | | +# | br0 | |----------------- peer |--------------| +# ----------- | | | +# | | | | +# ---------- peer ---------- ----------- | +# | veth02 | --------- | veth20 | | veth21 | | +# ---------- | ---------- ----------- | +# | ns2 | +# | | +#--------------------------------------------------------------+ +# +# We assume that all network driver are loaded +# + +# Kselftest framework requirement - SKIP code is 4. +ksft_skip=4 +ret=0 +GREEN='\033[0;92m' +RED='\033[0;31m' +NC='\033[0m' # No Color + +readonly port=8080 + +readonly vip_v4=207.175.44.110 +readonly cip_v4=10.0.0.2 +readonly gip_v4=10.0.0.1 +readonly dip_v4=172.16.0.1 +readonly rip_v4=172.16.0.2 +readonly sip_v4=10.0.0.3 + +readonly infile="$(mktemp)" +readonly outfile="$(mktemp)" +readonly datalen=32 + +sysipvsnet="/proc/sys/net/ipv4/vs/" +if [ ! -d $sysipvsnet ]; then + modprobe -q ip_vs + if [ $? -ne 0 ]; then + echo "skip: could not run test without ipvs module" + exit $ksft_skip + fi +fi + +ip -Version > /dev/null 2>&1 +if [ $? -ne 0 ]; then + echo "SKIP: Could not run test without ip tool" + exit $ksft_skip +fi + +ipvsadm -v > /dev/null 2>&1 +if [ $? -ne 0 ]; then + echo "SKIP: Could not run test without ipvsadm" + exit $ksft_skip +fi + +setup() { + ip netns add ns0 + ip netns add ns1 + ip netns add ns2 + + ip link add veth01 netns ns0 type veth peer name veth10 netns ns1 + ip link add veth02 netns ns0 type veth peer name veth20 netns ns2 + ip link add veth12 netns ns1 type veth peer name veth21 netns ns2 + + ip netns exec ns0 ip link set veth01 up + ip netns exec ns0 ip link set veth02 up + ip netns exec ns0 ip link add br0 type bridge + ip netns exec ns0 ip link set veth01 master br0 + ip netns exec ns0 ip link set veth02 master br0 + ip netns exec ns0 ip link set br0 up + ip netns exec ns0 ip addr add ${cip_v4}/24 dev br0 + + ip netns exec ns1 ip link set lo up + ip netns exec ns1 ip link set veth10 up + ip netns exec ns1 ip addr add ${gip_v4}/24 dev veth10 + ip netns exec ns1 ip link set veth12 up + ip netns exec ns1 ip addr add ${dip_v4}/24 dev veth12 + + ip netns exec ns2 ip link set lo up + ip netns exec ns2 ip link set veth21 up + ip netns exec ns2 ip addr add ${rip_v4}/24 dev veth21 + ip netns exec ns2 ip link set veth20 up + ip netns exec ns2 ip addr add ${sip_v4}/24 dev veth20 + + sleep 1 + + dd if=/dev/urandom of="${infile}" bs="${datalen}" count=1 status=none +} + +cleanup() { + for i in 0 1 2 + do + ip netns del ns$i > /dev/null 2>&1 + done + + if [ -f "${outfile}" ]; then + rm "${outfile}" + fi + if [ -f "${infile}" ]; then + rm "${infile}" + fi +} + +server_listen() { + ip netns exec ns2 nc -l -p 8080 > "${outfile}" & + server_pid=$! + sleep 0.2 +} + +client_connect() { + ip netns exec ns0 timeout 2 nc -w 1 ${vip_v4} ${port} < "${infile}" +} + +verify_data() { + wait "${server_pid}" + cmp "$infile" "$outfile" 2>/dev/null +} + +test_service() { + server_listen + client_connect + verify_data +} + + +test_dr() { + ip netns exec ns0 ip route add ${vip_v4} via ${gip_v4} dev br0 + + ip netns exec ns1 sysctl -qw net.ipv4.ip_forward=1 + ip netns exec ns1 ipvsadm -A -t ${vip_v4}:${port} -s rr + ip netns exec ns1 ipvsadm -a -t ${vip_v4}:${port} -r ${rip_v4}:${port} + ip netns exec ns1 ip addr add ${vip_v4}/32 dev lo:1 + + # avoid incorrect arp response + ip netns exec ns2 sysctl -qw net.ipv4.conf.all.arp_ignore=1 + ip netns exec ns2 sysctl -qw net.ipv4.conf.all.arp_announce=2 + # avoid reverse route lookup + ip netns exec ns2 sysctl -qw net.ipv4.conf.all.rp_filter=0 + ip netns exec ns2 sysctl -qw net.ipv4.conf.veth21.rp_filter=0 + ip netns exec ns2 ip addr add ${vip_v4}/32 dev lo:1 + + test_service +} + +test_nat() { + ip netns exec ns0 ip route add ${vip_v4} via ${gip_v4} dev br0 + + ip netns exec ns1 sysctl -qw net.ipv4.ip_forward=1 + ip netns exec ns1 ipvsadm -A -t ${vip_v4}:${port} -s rr + ip netns exec ns1 ipvsadm -a -m -t ${vip_v4}:${port} -r ${rip_v4}:${port} + ip netns exec ns1 ip addr add ${vip_v4}/32 dev lo:1 + + ip netns exec ns2 ip link del veth20 + ip netns exec ns2 ip route add default via ${dip_v4} dev veth21 + + test_service +} + +test_tun() { + ip netns exec ns0 ip route add ${vip_v4} via ${gip_v4} dev br0 + + ip netns exec ns1 modprobe ipip + ip netns exec ns1 ip link set tunl0 up + ip netns exec ns1 sysctl -qw net.ipv4.ip_forward=0 + ip netns exec ns1 sysctl -qw net.ipv4.conf.all.send_redirects=0 + ip netns exec ns1 sysctl -qw net.ipv4.conf.default.send_redirects=0 + ip netns exec ns1 ipvsadm -A -t ${vip_v4}:${port} -s rr + ip netns exec ns1 ipvsadm -a -i -t ${vip_v4}:${port} -r ${rip_v4}:${port} + ip netns exec ns1 ip addr add ${vip_v4}/32 dev lo:1 + + ip netns exec ns2 modprobe ipip + ip netns exec ns2 ip link set tunl0 up + ip netns exec ns2 sysctl -qw net.ipv4.conf.all.arp_ignore=1 + ip netns exec ns2 sysctl -qw net.ipv4.conf.all.arp_announce=2 + ip netns exec ns2 sysctl -qw net.ipv4.conf.all.rp_filter=0 + ip netns exec ns2 sysctl -qw net.ipv4.conf.tunl0.rp_filter=0 + ip netns exec ns2 sysctl -qw net.ipv4.conf.veth21.rp_filter=0 + ip netns exec ns2 ip addr add ${vip_v4}/32 dev lo:1 + + test_service +} + +run_tests() { + local errors= + + echo "Testing DR mode..." + cleanup + setup + test_dr + errors=$(( $errors + $? )) + + echo "Testing NAT mode..." + cleanup + setup + test_nat + errors=$(( $errors + $? )) + + echo "Testing Tunnel mode..." + cleanup + setup + test_tun + errors=$(( $errors + $? )) + + return $errors +} + +trap cleanup EXIT + +run_tests + +if [ $? -ne 0 ]; then + echo -e "$(basename $0): ${RED}FAIL${NC}" + exit 1 +fi +echo -e "$(basename $0): ${GREEN}PASS${NC}" +exit 0 diff --git a/tools/testing/selftests/netfilter/nf-queue.c b/tools/testing/selftests/netfilter/nf-queue.c new file mode 100644 index 000000000..9e56b9d47 --- /dev/null +++ b/tools/testing/selftests/netfilter/nf-queue.c @@ -0,0 +1,395 @@ +// SPDX-License-Identifier: GPL-2.0 + +#include <errno.h> +#include <stdbool.h> +#include <stdio.h> +#include <stdint.h> +#include <stdlib.h> +#include <unistd.h> +#include <string.h> +#include <time.h> +#include <arpa/inet.h> + +#include <libmnl/libmnl.h> +#include <linux/netfilter.h> +#include <linux/netfilter/nfnetlink.h> +#include <linux/netfilter/nfnetlink_queue.h> + +struct options { + bool count_packets; + bool gso_enabled; + int verbose; + unsigned int queue_num; + unsigned int timeout; + uint32_t verdict; + uint32_t delay_ms; +}; + +static unsigned int queue_stats[5]; +static struct options opts; + +static void help(const char *p) +{ + printf("Usage: %s [-c|-v [-vv] ] [-t timeout] [-q queue_num] [-Qdst_queue ] [ -d ms_delay ] [-G]\n", p); +} + +static int parse_attr_cb(const struct nlattr *attr, void *data) +{ + const struct nlattr **tb = data; + int type = mnl_attr_get_type(attr); + + /* skip unsupported attribute in user-space */ + if (mnl_attr_type_valid(attr, NFQA_MAX) < 0) + return MNL_CB_OK; + + switch (type) { + case NFQA_MARK: + case NFQA_IFINDEX_INDEV: + case NFQA_IFINDEX_OUTDEV: + case NFQA_IFINDEX_PHYSINDEV: + case NFQA_IFINDEX_PHYSOUTDEV: + if (mnl_attr_validate(attr, MNL_TYPE_U32) < 0) { + perror("mnl_attr_validate"); + return MNL_CB_ERROR; + } + break; + case NFQA_TIMESTAMP: + if (mnl_attr_validate2(attr, MNL_TYPE_UNSPEC, + sizeof(struct nfqnl_msg_packet_timestamp)) < 0) { + perror("mnl_attr_validate2"); + return MNL_CB_ERROR; + } + break; + case NFQA_HWADDR: + if (mnl_attr_validate2(attr, MNL_TYPE_UNSPEC, + sizeof(struct nfqnl_msg_packet_hw)) < 0) { + perror("mnl_attr_validate2"); + return MNL_CB_ERROR; + } + break; + case NFQA_PAYLOAD: + break; + } + tb[type] = attr; + return MNL_CB_OK; +} + +static int queue_cb(const struct nlmsghdr *nlh, void *data) +{ + struct nlattr *tb[NFQA_MAX+1] = { 0 }; + struct nfqnl_msg_packet_hdr *ph = NULL; + uint32_t id = 0; + + (void)data; + + mnl_attr_parse(nlh, sizeof(struct nfgenmsg), parse_attr_cb, tb); + if (tb[NFQA_PACKET_HDR]) { + ph = mnl_attr_get_payload(tb[NFQA_PACKET_HDR]); + id = ntohl(ph->packet_id); + + if (opts.verbose > 0) + printf("packet hook=%u, hwproto 0x%x", + ntohs(ph->hw_protocol), ph->hook); + + if (ph->hook >= 5) { + fprintf(stderr, "Unknown hook %d\n", ph->hook); + return MNL_CB_ERROR; + } + + if (opts.verbose > 0) { + uint32_t skbinfo = 0; + + if (tb[NFQA_SKB_INFO]) + skbinfo = ntohl(mnl_attr_get_u32(tb[NFQA_SKB_INFO])); + if (skbinfo & NFQA_SKB_CSUMNOTREADY) + printf(" csumnotready"); + if (skbinfo & NFQA_SKB_GSO) + printf(" gso"); + if (skbinfo & NFQA_SKB_CSUM_NOTVERIFIED) + printf(" csumnotverified"); + puts(""); + } + + if (opts.count_packets) + queue_stats[ph->hook]++; + } + + return MNL_CB_OK + id; +} + +static struct nlmsghdr * +nfq_build_cfg_request(char *buf, uint8_t command, int queue_num) +{ + struct nlmsghdr *nlh = mnl_nlmsg_put_header(buf); + struct nfqnl_msg_config_cmd cmd = { + .command = command, + .pf = htons(AF_INET), + }; + struct nfgenmsg *nfg; + + nlh->nlmsg_type = (NFNL_SUBSYS_QUEUE << 8) | NFQNL_MSG_CONFIG; + nlh->nlmsg_flags = NLM_F_REQUEST; + + nfg = mnl_nlmsg_put_extra_header(nlh, sizeof(*nfg)); + + nfg->nfgen_family = AF_UNSPEC; + nfg->version = NFNETLINK_V0; + nfg->res_id = htons(queue_num); + + mnl_attr_put(nlh, NFQA_CFG_CMD, sizeof(cmd), &cmd); + + return nlh; +} + +static struct nlmsghdr * +nfq_build_cfg_params(char *buf, uint8_t mode, int range, int queue_num) +{ + struct nlmsghdr *nlh = mnl_nlmsg_put_header(buf); + struct nfqnl_msg_config_params params = { + .copy_range = htonl(range), + .copy_mode = mode, + }; + struct nfgenmsg *nfg; + + nlh->nlmsg_type = (NFNL_SUBSYS_QUEUE << 8) | NFQNL_MSG_CONFIG; + nlh->nlmsg_flags = NLM_F_REQUEST; + + nfg = mnl_nlmsg_put_extra_header(nlh, sizeof(*nfg)); + nfg->nfgen_family = AF_UNSPEC; + nfg->version = NFNETLINK_V0; + nfg->res_id = htons(queue_num); + + mnl_attr_put(nlh, NFQA_CFG_PARAMS, sizeof(params), ¶ms); + + return nlh; +} + +static struct nlmsghdr * +nfq_build_verdict(char *buf, int id, int queue_num, uint32_t verd) +{ + struct nfqnl_msg_verdict_hdr vh = { + .verdict = htonl(verd), + .id = htonl(id), + }; + struct nlmsghdr *nlh; + struct nfgenmsg *nfg; + + nlh = mnl_nlmsg_put_header(buf); + nlh->nlmsg_type = (NFNL_SUBSYS_QUEUE << 8) | NFQNL_MSG_VERDICT; + nlh->nlmsg_flags = NLM_F_REQUEST; + nfg = mnl_nlmsg_put_extra_header(nlh, sizeof(*nfg)); + nfg->nfgen_family = AF_UNSPEC; + nfg->version = NFNETLINK_V0; + nfg->res_id = htons(queue_num); + + mnl_attr_put(nlh, NFQA_VERDICT_HDR, sizeof(vh), &vh); + + return nlh; +} + +static void print_stats(void) +{ + unsigned int last, total; + int i; + + total = 0; + last = queue_stats[0]; + + for (i = 0; i < 5; i++) { + printf("hook %d packets %08u\n", i, queue_stats[i]); + last = queue_stats[i]; + total += last; + } + + printf("%u packets total\n", total); +} + +struct mnl_socket *open_queue(void) +{ + char buf[MNL_SOCKET_BUFFER_SIZE]; + unsigned int queue_num; + struct mnl_socket *nl; + struct nlmsghdr *nlh; + struct timeval tv; + uint32_t flags; + + nl = mnl_socket_open(NETLINK_NETFILTER); + if (nl == NULL) { + perror("mnl_socket_open"); + exit(EXIT_FAILURE); + } + + if (mnl_socket_bind(nl, 0, MNL_SOCKET_AUTOPID) < 0) { + perror("mnl_socket_bind"); + exit(EXIT_FAILURE); + } + + queue_num = opts.queue_num; + nlh = nfq_build_cfg_request(buf, NFQNL_CFG_CMD_BIND, queue_num); + + if (mnl_socket_sendto(nl, nlh, nlh->nlmsg_len) < 0) { + perror("mnl_socket_sendto"); + exit(EXIT_FAILURE); + } + + nlh = nfq_build_cfg_params(buf, NFQNL_COPY_PACKET, 0xFFFF, queue_num); + + flags = opts.gso_enabled ? NFQA_CFG_F_GSO : 0; + flags |= NFQA_CFG_F_UID_GID; + mnl_attr_put_u32(nlh, NFQA_CFG_FLAGS, htonl(flags)); + mnl_attr_put_u32(nlh, NFQA_CFG_MASK, htonl(flags)); + + if (mnl_socket_sendto(nl, nlh, nlh->nlmsg_len) < 0) { + perror("mnl_socket_sendto"); + exit(EXIT_FAILURE); + } + + memset(&tv, 0, sizeof(tv)); + tv.tv_sec = opts.timeout; + if (opts.timeout && setsockopt(mnl_socket_get_fd(nl), + SOL_SOCKET, SO_RCVTIMEO, + &tv, sizeof(tv))) { + perror("setsockopt(SO_RCVTIMEO)"); + exit(EXIT_FAILURE); + } + + return nl; +} + +static void sleep_ms(uint32_t delay) +{ + struct timespec ts = { .tv_sec = delay / 1000 }; + + delay %= 1000; + + ts.tv_nsec = delay * 1000llu * 1000llu; + + nanosleep(&ts, NULL); +} + +static int mainloop(void) +{ + unsigned int buflen = 64 * 1024 + MNL_SOCKET_BUFFER_SIZE; + struct mnl_socket *nl; + struct nlmsghdr *nlh; + unsigned int portid; + char *buf; + int ret; + + buf = malloc(buflen); + if (!buf) { + perror("malloc"); + exit(EXIT_FAILURE); + } + + nl = open_queue(); + portid = mnl_socket_get_portid(nl); + + for (;;) { + uint32_t id; + + ret = mnl_socket_recvfrom(nl, buf, buflen); + if (ret == -1) { + if (errno == ENOBUFS || errno == EINTR) + continue; + + if (errno == EAGAIN) { + errno = 0; + ret = 0; + break; + } + + perror("mnl_socket_recvfrom"); + exit(EXIT_FAILURE); + } + + ret = mnl_cb_run(buf, ret, 0, portid, queue_cb, NULL); + if (ret < 0) { + perror("mnl_cb_run"); + exit(EXIT_FAILURE); + } + + id = ret - MNL_CB_OK; + if (opts.delay_ms) + sleep_ms(opts.delay_ms); + + nlh = nfq_build_verdict(buf, id, opts.queue_num, opts.verdict); + if (mnl_socket_sendto(nl, nlh, nlh->nlmsg_len) < 0) { + perror("mnl_socket_sendto"); + exit(EXIT_FAILURE); + } + } + + mnl_socket_close(nl); + + return ret; +} + +static void parse_opts(int argc, char **argv) +{ + int c; + + while ((c = getopt(argc, argv, "chvt:q:Q:d:G")) != -1) { + switch (c) { + case 'c': + opts.count_packets = true; + break; + case 'h': + help(argv[0]); + exit(0); + break; + case 'q': + opts.queue_num = atoi(optarg); + if (opts.queue_num > 0xffff) + opts.queue_num = 0; + break; + case 'Q': + opts.verdict = atoi(optarg); + if (opts.verdict > 0xffff) { + fprintf(stderr, "Expected destination queue number\n"); + exit(1); + } + + opts.verdict <<= 16; + opts.verdict |= NF_QUEUE; + break; + case 'd': + opts.delay_ms = atoi(optarg); + if (opts.delay_ms == 0) { + fprintf(stderr, "Expected nonzero delay (in milliseconds)\n"); + exit(1); + } + break; + case 't': + opts.timeout = atoi(optarg); + break; + case 'G': + opts.gso_enabled = false; + break; + case 'v': + opts.verbose++; + break; + } + } + + if (opts.verdict != NF_ACCEPT && (opts.verdict >> 16 == opts.queue_num)) { + fprintf(stderr, "Cannot use same destination and source queue\n"); + exit(1); + } +} + +int main(int argc, char *argv[]) +{ + int ret; + + opts.verdict = NF_ACCEPT; + opts.gso_enabled = true; + + parse_opts(argc, argv); + + ret = mainloop(); + if (opts.count_packets) + print_stats(); + + return ret; +} diff --git a/tools/testing/selftests/netfilter/nft_concat_range.sh b/tools/testing/selftests/netfilter/nft_concat_range.sh new file mode 100755 index 000000000..af3461cb5 --- /dev/null +++ b/tools/testing/selftests/netfilter/nft_concat_range.sh @@ -0,0 +1,1586 @@ +#!/bin/sh +# SPDX-License-Identifier: GPL-2.0 +# +# nft_concat_range.sh - Tests for sets with concatenation of ranged fields +# +# Copyright (c) 2019 Red Hat GmbH +# +# Author: Stefano Brivio <sbrivio@redhat.com> +# +# shellcheck disable=SC2154,SC2034,SC2016,SC2030,SC2031 +# ^ Configuration and templates sourced with eval, counters reused in subshells + +KSELFTEST_SKIP=4 + +# Available test groups: +# - reported_issues: check for issues that were reported in the past +# - correctness: check that packets match given entries, and only those +# - concurrency: attempt races between insertion, deletion and lookup +# - timeout: check that packets match entries until they expire +# - performance: estimate matching rate, compare with rbtree and hash baselines +TESTS="reported_issues correctness concurrency timeout" +[ "${quicktest}" != "1" ] && TESTS="${TESTS} performance" + +# Set types, defined by TYPE_ variables below +TYPES="net_port port_net net6_port port_proto net6_port_mac net6_port_mac_proto + net_port_net net_mac net_mac_icmp net6_mac_icmp net6_port_net6_port + net_port_mac_proto_net" + +# Reported bugs, also described by TYPE_ variables below +BUGS="flush_remove_add reload" + +# List of possible paths to pktgen script from kernel tree for performance tests +PKTGEN_SCRIPT_PATHS=" + ../../../../samples/pktgen/pktgen_bench_xmit_mode_netif_receive.sh + pktgen/pktgen_bench_xmit_mode_netif_receive.sh" + +# Definition of set types: +# display display text for test report +# type_spec nftables set type specifier +# chain_spec nftables type specifier for rules mapping to set +# dst call sequence of format_*() functions for destination fields +# src call sequence of format_*() functions for source fields +# start initial integer used to generate addresses and ports +# count count of entries to generate and match +# src_delta number summed to destination generator for source fields +# tools list of tools for correctness and timeout tests, any can be used +# proto L4 protocol of test packets +# +# race_repeat race attempts per thread, 0 disables concurrency test for type +# flood_tools list of tools for concurrency tests, any can be used +# flood_proto L4 protocol of test packets for concurrency tests +# flood_spec nftables type specifier for concurrency tests +# +# perf_duration duration of single pktgen injection test +# perf_spec nftables type specifier for performance tests +# perf_dst format_*() functions for destination fields in performance test +# perf_src format_*() functions for source fields in performance test +# perf_entries number of set entries for performance test +# perf_proto L3 protocol of test packets +TYPE_net_port=" +display net,port +type_spec ipv4_addr . inet_service +chain_spec ip daddr . udp dport +dst addr4 port +src +start 1 +count 5 +src_delta 2000 +tools sendip nc bash +proto udp + +race_repeat 3 +flood_tools iperf3 iperf netperf +flood_proto udp +flood_spec ip daddr . udp dport + +perf_duration 5 +perf_spec ip daddr . udp dport +perf_dst addr4 port +perf_src +perf_entries 1000 +perf_proto ipv4 +" + +TYPE_port_net=" +display port,net +type_spec inet_service . ipv4_addr +chain_spec udp dport . ip daddr +dst port addr4 +src +start 1 +count 5 +src_delta 2000 +tools sendip nc bash +proto udp + +race_repeat 3 +flood_tools iperf3 iperf netperf +flood_proto udp +flood_spec udp dport . ip daddr + +perf_duration 5 +perf_spec udp dport . ip daddr +perf_dst port addr4 +perf_src +perf_entries 100 +perf_proto ipv4 +" + +TYPE_net6_port=" +display net6,port +type_spec ipv6_addr . inet_service +chain_spec ip6 daddr . udp dport +dst addr6 port +src +start 10 +count 5 +src_delta 2000 +tools sendip nc bash +proto udp6 + +race_repeat 3 +flood_tools iperf3 iperf netperf +flood_proto tcp6 +flood_spec ip6 daddr . udp dport + +perf_duration 5 +perf_spec ip6 daddr . udp dport +perf_dst addr6 port +perf_src +perf_entries 1000 +perf_proto ipv6 +" + +TYPE_port_proto=" +display port,proto +type_spec inet_service . inet_proto +chain_spec udp dport . meta l4proto +dst port proto +src +start 1 +count 5 +src_delta 2000 +tools sendip nc bash +proto udp + +race_repeat 0 + +perf_duration 5 +perf_spec udp dport . meta l4proto +perf_dst port proto +perf_src +perf_entries 30000 +perf_proto ipv4 +" + +TYPE_net6_port_mac=" +display net6,port,mac +type_spec ipv6_addr . inet_service . ether_addr +chain_spec ip6 daddr . udp dport . ether saddr +dst addr6 port +src mac +start 10 +count 5 +src_delta 2000 +tools sendip nc bash +proto udp6 + +race_repeat 0 + +perf_duration 5 +perf_spec ip6 daddr . udp dport . ether daddr +perf_dst addr6 port mac +perf_src +perf_entries 10 +perf_proto ipv6 +" + +TYPE_net6_port_mac_proto=" +display net6,port,mac,proto +type_spec ipv6_addr . inet_service . ether_addr . inet_proto +chain_spec ip6 daddr . udp dport . ether saddr . meta l4proto +dst addr6 port +src mac proto +start 10 +count 5 +src_delta 2000 +tools sendip nc bash +proto udp6 + +race_repeat 0 + +perf_duration 5 +perf_spec ip6 daddr . udp dport . ether daddr . meta l4proto +perf_dst addr6 port mac proto +perf_src +perf_entries 1000 +perf_proto ipv6 +" + +TYPE_net_port_net=" +display net,port,net +type_spec ipv4_addr . inet_service . ipv4_addr +chain_spec ip daddr . udp dport . ip saddr +dst addr4 port +src addr4 +start 1 +count 5 +src_delta 2000 +tools sendip nc bash +proto udp + +race_repeat 3 +flood_tools iperf3 iperf netperf +flood_proto tcp +flood_spec ip daddr . udp dport . ip saddr + +perf_duration 0 +" + +TYPE_net6_port_net6_port=" +display net6,port,net6,port +type_spec ipv6_addr . inet_service . ipv6_addr . inet_service +chain_spec ip6 daddr . udp dport . ip6 saddr . udp sport +dst addr6 port +src addr6 port +start 10 +count 5 +src_delta 2000 +tools sendip nc +proto udp6 + +race_repeat 3 +flood_tools iperf3 iperf netperf +flood_proto tcp6 +flood_spec ip6 daddr . tcp dport . ip6 saddr . tcp sport + +perf_duration 0 +" + +TYPE_net_port_mac_proto_net=" +display net,port,mac,proto,net +type_spec ipv4_addr . inet_service . ether_addr . inet_proto . ipv4_addr +chain_spec ip daddr . udp dport . ether saddr . meta l4proto . ip saddr +dst addr4 port +src mac proto addr4 +start 1 +count 5 +src_delta 2000 +tools sendip nc bash +proto udp + +race_repeat 0 + +perf_duration 0 +" + +TYPE_net_mac=" +display net,mac +type_spec ipv4_addr . ether_addr +chain_spec ip daddr . ether saddr +dst addr4 +src mac +start 1 +count 5 +src_delta 2000 +tools sendip nc bash +proto udp + +race_repeat 0 + +perf_duration 5 +perf_spec ip daddr . ether daddr +perf_dst addr4 mac +perf_src +perf_entries 1000 +perf_proto ipv4 +" + +TYPE_net_mac_icmp=" +display net,mac - ICMP +type_spec ipv4_addr . ether_addr +chain_spec ip daddr . ether saddr +dst addr4 +src mac +start 1 +count 5 +src_delta 2000 +tools ping +proto icmp + +race_repeat 0 + +perf_duration 0 +" + +TYPE_net6_mac_icmp=" +display net6,mac - ICMPv6 +type_spec ipv6_addr . ether_addr +chain_spec ip6 daddr . ether saddr +dst addr6 +src mac +start 10 +count 50 +src_delta 2000 +tools ping +proto icmp6 + +race_repeat 0 + +perf_duration 0 +" + +TYPE_net_port_proto_net=" +display net,port,proto,net +type_spec ipv4_addr . inet_service . inet_proto . ipv4_addr +chain_spec ip daddr . udp dport . meta l4proto . ip saddr +dst addr4 port proto +src addr4 +start 1 +count 5 +src_delta 2000 +tools sendip nc +proto udp + +race_repeat 3 +flood_tools iperf3 iperf netperf +flood_proto tcp +flood_spec ip daddr . tcp dport . meta l4proto . ip saddr + +perf_duration 0 +" + +# Definition of tests for bugs reported in the past: +# display display text for test report +TYPE_flush_remove_add=" +display Add two elements, flush, re-add +" + +TYPE_reload=" +display net,mac with reload +type_spec ipv4_addr . ether_addr +chain_spec ip daddr . ether saddr +dst addr4 +src mac +start 1 +count 1 +src_delta 2000 +tools sendip nc bash +proto udp + +race_repeat 0 + +perf_duration 0 +" + +# Set template for all tests, types and rules are filled in depending on test +set_template=' +flush ruleset + +table inet filter { + counter test { + packets 0 bytes 0 + } + + set test { + type ${type_spec} + flags interval,timeout + } + + chain input { + type filter hook prerouting priority 0; policy accept; + ${chain_spec} @test counter name \"test\" + } +} + +table netdev perf { + counter test { + packets 0 bytes 0 + } + + counter match { + packets 0 bytes 0 + } + + set test { + type ${type_spec} + flags interval + } + + set norange { + type ${type_spec} + } + + set noconcat { + type ${type_spec%% *} + flags interval + } + + chain test { + type filter hook ingress device veth_a priority 0; + } +} +' + +err_buf= +info_buf= + +# Append string to error buffer +err() { + err_buf="${err_buf}${1} +" +} + +# Append string to information buffer +info() { + info_buf="${info_buf}${1} +" +} + +# Flush error buffer to stdout +err_flush() { + printf "%s" "${err_buf}" + err_buf= +} + +# Flush information buffer to stdout +info_flush() { + printf "%s" "${info_buf}" + info_buf= +} + +# Setup veth pair: this namespace receives traffic, B generates it +setup_veth() { + ip netns add B + ip link add veth_a type veth peer name veth_b || return 1 + + ip link set veth_a up + ip link set veth_b netns B + + ip -n B link set veth_b up + + ip addr add dev veth_a 10.0.0.1 + ip route add default dev veth_a + + ip -6 addr add fe80::1/64 dev veth_a nodad + ip -6 addr add 2001:db8::1/64 dev veth_a nodad + ip -6 route add default dev veth_a + + ip -n B route add default dev veth_b + + ip -6 -n B addr add fe80::2/64 dev veth_b nodad + ip -6 -n B addr add 2001:db8::2/64 dev veth_b nodad + ip -6 -n B route add default dev veth_b + + B() { + ip netns exec B "$@" >/dev/null 2>&1 + } + + sleep 2 +} + +# Fill in set template and initialise set +setup_set() { + eval "echo \"${set_template}\"" | nft -f - +} + +# Check that at least one of the needed tools is available +check_tools() { + [ -z "${tools}" ] && return 0 + + __tools= + for tool in ${tools}; do + if [ "${tool}" = "nc" ] && [ "${proto}" = "udp6" ] && \ + ! nc -u -w0 1.1.1.1 1 2>/dev/null; then + # Some GNU netcat builds might not support IPv6 + __tools="${__tools} netcat-openbsd" + continue + fi + __tools="${__tools} ${tool}" + + command -v "${tool}" >/dev/null && return 0 + done + err "need one of:${__tools}, skipping" && return 1 +} + +# Set up function to send ICMP packets +setup_send_icmp() { + send_icmp() { + B ping -c1 -W1 "${dst_addr4}" >/dev/null 2>&1 + } +} + +# Set up function to send ICMPv6 packets +setup_send_icmp6() { + if command -v ping6 >/dev/null; then + send_icmp6() { + ip -6 addr add "${dst_addr6}" dev veth_a nodad \ + 2>/dev/null + B ping6 -q -c1 -W1 "${dst_addr6}" + } + else + send_icmp6() { + ip -6 addr add "${dst_addr6}" dev veth_a nodad \ + 2>/dev/null + B ping -q -6 -c1 -W1 "${dst_addr6}" + } + fi +} + +# Set up function to send single UDP packets on IPv4 +setup_send_udp() { + if command -v sendip >/dev/null; then + send_udp() { + [ -n "${src_port}" ] && src_port="-us ${src_port}" + [ -n "${dst_port}" ] && dst_port="-ud ${dst_port}" + [ -n "${src_addr4}" ] && src_addr4="-is ${src_addr4}" + + # shellcheck disable=SC2086 # sendip needs split options + B sendip -p ipv4 -p udp ${src_addr4} ${src_port} \ + ${dst_port} "${dst_addr4}" + + src_port= + dst_port= + src_addr4= + } + elif command -v nc >/dev/null; then + if nc -u -w0 1.1.1.1 1 2>/dev/null; then + # OpenBSD netcat + nc_opt="-w0" + else + # GNU netcat + nc_opt="-q0" + fi + + send_udp() { + if [ -n "${src_addr4}" ]; then + B ip addr add "${src_addr4}" dev veth_b + __src_addr4="-s ${src_addr4}" + fi + ip addr add "${dst_addr4}" dev veth_a 2>/dev/null + [ -n "${src_port}" ] && src_port="-p ${src_port}" + + echo "" | B nc -u "${nc_opt}" "${__src_addr4}" \ + "${src_port}" "${dst_addr4}" "${dst_port}" + + src_addr4= + src_port= + } + elif [ -z "$(bash -c 'type -p')" ]; then + send_udp() { + ip addr add "${dst_addr4}" dev veth_a 2>/dev/null + if [ -n "${src_addr4}" ]; then + B ip addr add "${src_addr4}/16" dev veth_b + B ip route add default dev veth_b + fi + + B bash -c "echo > /dev/udp/${dst_addr4}/${dst_port}" + + if [ -n "${src_addr4}" ]; then + B ip addr del "${src_addr4}/16" dev veth_b + fi + src_addr4= + } + else + return 1 + fi +} + +# Set up function to send single UDP packets on IPv6 +setup_send_udp6() { + if command -v sendip >/dev/null; then + send_udp6() { + [ -n "${src_port}" ] && src_port="-us ${src_port}" + [ -n "${dst_port}" ] && dst_port="-ud ${dst_port}" + if [ -n "${src_addr6}" ]; then + src_addr6="-6s ${src_addr6}" + else + src_addr6="-6s 2001:db8::2" + fi + ip -6 addr add "${dst_addr6}" dev veth_a nodad \ + 2>/dev/null + + # shellcheck disable=SC2086 # this needs split options + B sendip -p ipv6 -p udp ${src_addr6} ${src_port} \ + ${dst_port} "${dst_addr6}" + + src_port= + dst_port= + src_addr6= + } + elif command -v nc >/dev/null && nc -u -w0 1.1.1.1 1 2>/dev/null; then + # GNU netcat might not work with IPv6, try next tool + send_udp6() { + ip -6 addr add "${dst_addr6}" dev veth_a nodad \ + 2>/dev/null + if [ -n "${src_addr6}" ]; then + B ip addr add "${src_addr6}" dev veth_b nodad + else + src_addr6="2001:db8::2" + fi + [ -n "${src_port}" ] && src_port="-p ${src_port}" + + # shellcheck disable=SC2086 # this needs split options + echo "" | B nc -u w0 "-s${src_addr6}" ${src_port} \ + ${dst_addr6} ${dst_port} + + src_addr6= + src_port= + } + elif [ -z "$(bash -c 'type -p')" ]; then + send_udp6() { + ip -6 addr add "${dst_addr6}" dev veth_a nodad \ + 2>/dev/null + B ip addr add "${src_addr6}" dev veth_b nodad + B bash -c "echo > /dev/udp/${dst_addr6}/${dst_port}" + ip -6 addr del "${dst_addr6}" dev veth_a 2>/dev/null + } + else + return 1 + fi +} + +# Set up function to send TCP traffic on IPv4 +setup_flood_tcp() { + if command -v iperf3 >/dev/null; then + flood_tcp() { + [ -n "${dst_port}" ] && dst_port="-p ${dst_port}" + if [ -n "${src_addr4}" ]; then + B ip addr add "${src_addr4}/16" dev veth_b + src_addr4="-B ${src_addr4}" + else + B ip addr add dev veth_b 10.0.0.2 + src_addr4="-B 10.0.0.2" + fi + if [ -n "${src_port}" ]; then + src_port="--cport ${src_port}" + fi + B ip route add default dev veth_b 2>/dev/null + ip addr add "${dst_addr4}" dev veth_a 2>/dev/null + + # shellcheck disable=SC2086 # this needs split options + iperf3 -s -DB "${dst_addr4}" ${dst_port} >/dev/null 2>&1 + sleep 2 + + # shellcheck disable=SC2086 # this needs split options + B iperf3 -c "${dst_addr4}" ${dst_port} ${src_port} \ + ${src_addr4} -l16 -t 1000 + + src_addr4= + src_port= + dst_port= + } + elif command -v iperf >/dev/null; then + flood_tcp() { + [ -n "${dst_port}" ] && dst_port="-p ${dst_port}" + if [ -n "${src_addr4}" ]; then + B ip addr add "${src_addr4}/16" dev veth_b + src_addr4="-B ${src_addr4}" + else + B ip addr add dev veth_b 10.0.0.2 2>/dev/null + src_addr4="-B 10.0.0.2" + fi + if [ -n "${src_port}" ]; then + src_addr4="${src_addr4}:${src_port}" + fi + B ip route add default dev veth_b + ip addr add "${dst_addr4}" dev veth_a 2>/dev/null + + # shellcheck disable=SC2086 # this needs split options + iperf -s -DB "${dst_addr4}" ${dst_port} >/dev/null 2>&1 + sleep 2 + + # shellcheck disable=SC2086 # this needs split options + B iperf -c "${dst_addr4}" ${dst_port} ${src_addr4} \ + -l20 -t 1000 + + src_addr4= + src_port= + dst_port= + } + elif command -v netperf >/dev/null; then + flood_tcp() { + [ -n "${dst_port}" ] && dst_port="-p ${dst_port}" + if [ -n "${src_addr4}" ]; then + B ip addr add "${src_addr4}/16" dev veth_b + else + B ip addr add dev veth_b 10.0.0.2 + src_addr4="10.0.0.2" + fi + if [ -n "${src_port}" ]; then + dst_port="${dst_port},${src_port}" + fi + B ip route add default dev veth_b + ip addr add "${dst_addr4}" dev veth_a 2>/dev/null + + # shellcheck disable=SC2086 # this needs split options + netserver -4 ${dst_port} -L "${dst_addr4}" \ + >/dev/null 2>&1 + sleep 2 + + # shellcheck disable=SC2086 # this needs split options + B netperf -4 -H "${dst_addr4}" ${dst_port} \ + -L "${src_addr4}" -l 1000 -t TCP_STREAM + + src_addr4= + src_port= + dst_port= + } + else + return 1 + fi +} + +# Set up function to send TCP traffic on IPv6 +setup_flood_tcp6() { + if command -v iperf3 >/dev/null; then + flood_tcp6() { + [ -n "${dst_port}" ] && dst_port="-p ${dst_port}" + if [ -n "${src_addr6}" ]; then + B ip addr add "${src_addr6}" dev veth_b nodad + src_addr6="-B ${src_addr6}" + else + src_addr6="-B 2001:db8::2" + fi + if [ -n "${src_port}" ]; then + src_port="--cport ${src_port}" + fi + B ip route add default dev veth_b + ip -6 addr add "${dst_addr6}" dev veth_a nodad \ + 2>/dev/null + + # shellcheck disable=SC2086 # this needs split options + iperf3 -s -DB "${dst_addr6}" ${dst_port} >/dev/null 2>&1 + sleep 2 + + # shellcheck disable=SC2086 # this needs split options + B iperf3 -c "${dst_addr6}" ${dst_port} \ + ${src_port} ${src_addr6} -l16 -t 1000 + + src_addr6= + src_port= + dst_port= + } + elif command -v iperf >/dev/null; then + flood_tcp6() { + [ -n "${dst_port}" ] && dst_port="-p ${dst_port}" + if [ -n "${src_addr6}" ]; then + B ip addr add "${src_addr6}" dev veth_b nodad + src_addr6="-B ${src_addr6}" + else + src_addr6="-B 2001:db8::2" + fi + if [ -n "${src_port}" ]; then + src_addr6="${src_addr6}:${src_port}" + fi + B ip route add default dev veth_b + ip -6 addr add "${dst_addr6}" dev veth_a nodad \ + 2>/dev/null + + # shellcheck disable=SC2086 # this needs split options + iperf -s -VDB "${dst_addr6}" ${dst_port} >/dev/null 2>&1 + sleep 2 + + # shellcheck disable=SC2086 # this needs split options + B iperf -c "${dst_addr6}" -V ${dst_port} \ + ${src_addr6} -l1 -t 1000 + + src_addr6= + src_port= + dst_port= + } + elif command -v netperf >/dev/null; then + flood_tcp6() { + [ -n "${dst_port}" ] && dst_port="-p ${dst_port}" + if [ -n "${src_addr6}" ]; then + B ip addr add "${src_addr6}" dev veth_b nodad + else + src_addr6="2001:db8::2" + fi + if [ -n "${src_port}" ]; then + dst_port="${dst_port},${src_port}" + fi + B ip route add default dev veth_b + ip -6 addr add "${dst_addr6}" dev veth_a nodad \ + 2>/dev/null + + # shellcheck disable=SC2086 # this needs split options + netserver -6 ${dst_port} -L "${dst_addr6}" \ + >/dev/null 2>&1 + sleep 2 + + # shellcheck disable=SC2086 # this needs split options + B netperf -6 -H "${dst_addr6}" ${dst_port} \ + -L "${src_addr6}" -l 1000 -t TCP_STREAM + + src_addr6= + src_port= + dst_port= + } + else + return 1 + fi +} + +# Set up function to send UDP traffic on IPv4 +setup_flood_udp() { + if command -v iperf3 >/dev/null; then + flood_udp() { + [ -n "${dst_port}" ] && dst_port="-p ${dst_port}" + if [ -n "${src_addr4}" ]; then + B ip addr add "${src_addr4}/16" dev veth_b + src_addr4="-B ${src_addr4}" + else + B ip addr add dev veth_b 10.0.0.2 2>/dev/null + src_addr4="-B 10.0.0.2" + fi + if [ -n "${src_port}" ]; then + src_port="--cport ${src_port}" + fi + B ip route add default dev veth_b + ip addr add "${dst_addr4}" dev veth_a 2>/dev/null + + # shellcheck disable=SC2086 # this needs split options + iperf3 -s -DB "${dst_addr4}" ${dst_port} + sleep 2 + + # shellcheck disable=SC2086 # this needs split options + B iperf3 -u -c "${dst_addr4}" -Z -b 100M -l16 -t1000 \ + ${dst_port} ${src_port} ${src_addr4} + + src_addr4= + src_port= + dst_port= + } + elif command -v iperf >/dev/null; then + flood_udp() { + [ -n "${dst_port}" ] && dst_port="-p ${dst_port}" + if [ -n "${src_addr4}" ]; then + B ip addr add "${src_addr4}/16" dev veth_b + src_addr4="-B ${src_addr4}" + else + B ip addr add dev veth_b 10.0.0.2 + src_addr4="-B 10.0.0.2" + fi + if [ -n "${src_port}" ]; then + src_addr4="${src_addr4}:${src_port}" + fi + B ip route add default dev veth_b + ip addr add "${dst_addr4}" dev veth_a 2>/dev/null + + # shellcheck disable=SC2086 # this needs split options + iperf -u -sDB "${dst_addr4}" ${dst_port} >/dev/null 2>&1 + sleep 2 + + # shellcheck disable=SC2086 # this needs split options + B iperf -u -c "${dst_addr4}" -b 100M -l1 -t1000 \ + ${dst_port} ${src_addr4} + + src_addr4= + src_port= + dst_port= + } + elif command -v netperf >/dev/null; then + flood_udp() { + [ -n "${dst_port}" ] && dst_port="-p ${dst_port}" + if [ -n "${src_addr4}" ]; then + B ip addr add "${src_addr4}/16" dev veth_b + else + B ip addr add dev veth_b 10.0.0.2 + src_addr4="10.0.0.2" + fi + if [ -n "${src_port}" ]; then + dst_port="${dst_port},${src_port}" + fi + B ip route add default dev veth_b + ip addr add "${dst_addr4}" dev veth_a 2>/dev/null + + # shellcheck disable=SC2086 # this needs split options + netserver -4 ${dst_port} -L "${dst_addr4}" \ + >/dev/null 2>&1 + sleep 2 + + # shellcheck disable=SC2086 # this needs split options + B netperf -4 -H "${dst_addr4}" ${dst_port} \ + -L "${src_addr4}" -l 1000 -t UDP_STREAM + + src_addr4= + src_port= + dst_port= + } + else + return 1 + fi +} + +# Find pktgen script and set up function to start pktgen injection +setup_perf() { + for pktgen_script_path in ${PKTGEN_SCRIPT_PATHS} __notfound; do + command -v "${pktgen_script_path}" >/dev/null && break + done + [ "${pktgen_script_path}" = "__notfound" ] && return 1 + + perf_ipv4() { + ${pktgen_script_path} -s80 \ + -i veth_a -d "${dst_addr4}" -p "${dst_port}" \ + -m "${dst_mac}" \ + -t $(($(nproc) / 5 + 1)) -b10000 -n0 2>/dev/null & + perf_pid=$! + } + perf_ipv6() { + IP6=6 ${pktgen_script_path} -s100 \ + -i veth_a -d "${dst_addr6}" -p "${dst_port}" \ + -m "${dst_mac}" \ + -t $(($(nproc) / 5 + 1)) -b10000 -n0 2>/dev/null & + perf_pid=$! + } +} + +# Clean up before each test +cleanup() { + nft reset counter inet filter test >/dev/null 2>&1 + nft flush ruleset >/dev/null 2>&1 + ip link del dummy0 2>/dev/null + ip route del default 2>/dev/null + ip -6 route del default 2>/dev/null + ip netns del B 2>/dev/null + ip link del veth_a 2>/dev/null + timeout= + killall iperf3 2>/dev/null + killall iperf 2>/dev/null + killall netperf 2>/dev/null + killall netserver 2>/dev/null + rm -f ${tmp} + sleep 2 +} + +# Entry point for setup functions +setup() { + if [ "$(id -u)" -ne 0 ]; then + echo " need to run as root" + exit ${KSELFTEST_SKIP} + fi + + cleanup + check_tools || return 1 + for arg do + if ! eval setup_"${arg}"; then + err " ${arg} not supported" + return 1 + fi + done +} + +# Format integer into IPv4 address, summing 10.0.0.5 (arbitrary) to it +format_addr4() { + a=$((${1} + 16777216 * 10 + 5)) + printf "%i.%i.%i.%i" \ + "$((a / 16777216))" "$((a % 16777216 / 65536))" \ + "$((a % 65536 / 256))" "$((a % 256))" +} + +# Format integer into IPv6 address, summing 2001:db8:: to it +format_addr6() { + printf "2001:db8::%04x:%04x" "$((${1} / 65536))" "$((${1} % 65536))" +} + +# Format integer into EUI-48 address, summing 00:01:00:00:00:00 to it +format_mac() { + printf "00:01:%02x:%02x:%02x:%02x" \ + "$((${1} / 16777216))" "$((${1} % 16777216 / 65536))" \ + "$((${1} % 65536 / 256))" "$((${1} % 256))" +} + +# Format integer into port, avoid 0 port +format_port() { + printf "%i" "$((${1} % 65534 + 1))" +} + +# Drop suffixed '6' from L4 protocol, if any +format_proto() { + printf "%s" "${proto}" | tr -d 6 +} + +# Format destination and source fields into nft concatenated type +format() { + __start= + __end= + __expr="{ " + + for f in ${dst}; do + [ "${__expr}" != "{ " ] && __expr="${__expr} . " + + __start="$(eval format_"${f}" "${start}")" + __end="$(eval format_"${f}" "${end}")" + + if [ "${f}" = "proto" ]; then + __expr="${__expr}${__start}" + else + __expr="${__expr}${__start}-${__end}" + fi + done + for f in ${src}; do + __expr="${__expr} . " + __start="$(eval format_"${f}" "${srcstart}")" + __end="$(eval format_"${f}" "${srcend}")" + + if [ "${f}" = "proto" ]; then + __expr="${__expr}${__start}" + else + __expr="${__expr}${__start}-${__end}" + fi + done + + if [ -n "${timeout}" ]; then + echo "${__expr} timeout ${timeout}s }" + else + echo "${__expr} }" + fi +} + +# Format destination and source fields into nft type, start element only +format_norange() { + __expr="{ " + + for f in ${dst}; do + [ "${__expr}" != "{ " ] && __expr="${__expr} . " + + __expr="${__expr}$(eval format_"${f}" "${start}")" + done + for f in ${src}; do + __expr="${__expr} . $(eval format_"${f}" "${start}")" + done + + echo "${__expr} }" +} + +# Format first destination field into nft type +format_noconcat() { + for f in ${dst}; do + __start="$(eval format_"${f}" "${start}")" + __end="$(eval format_"${f}" "${end}")" + + if [ "${f}" = "proto" ]; then + echo "{ ${__start} }" + else + echo "{ ${__start}-${__end} }" + fi + return + done +} + +# Add single entry to 'test' set in 'inet filter' table +add() { + if ! nft add element inet filter test "${1}"; then + err "Failed to add ${1} given ruleset:" + err "$(nft -a list ruleset)" + return 1 + fi +} + +# Format and output entries for sets in 'netdev perf' table +add_perf() { + if [ "${1}" = "test" ]; then + echo "add element netdev perf test $(format)" + elif [ "${1}" = "norange" ]; then + echo "add element netdev perf norange $(format_norange)" + elif [ "${1}" = "noconcat" ]; then + echo "add element netdev perf noconcat $(format_noconcat)" + fi +} + +# Add single entry to 'norange' set in 'netdev perf' table +add_perf_norange() { + if ! nft add element netdev perf norange "${1}"; then + err "Failed to add ${1} given ruleset:" + err "$(nft -a list ruleset)" + return 1 + fi +} + +# Add single entry to 'noconcat' set in 'netdev perf' table +add_perf_noconcat() { + if ! nft add element netdev perf noconcat "${1}"; then + err "Failed to add ${1} given ruleset:" + err "$(nft -a list ruleset)" + return 1 + fi +} + +# Delete single entry from set +del() { + if ! nft delete element inet filter test "${1}"; then + err "Failed to delete ${1} given ruleset:" + err "$(nft -a list ruleset)" + return 1 + fi +} + +# Return packet count from 'test' counter in 'inet filter' table +count_packets() { + found=0 + for token in $(nft list counter inet filter test); do + [ ${found} -eq 1 ] && echo "${token}" && return + [ "${token}" = "packets" ] && found=1 + done +} + +# Return packet count from 'test' counter in 'netdev perf' table +count_perf_packets() { + found=0 + for token in $(nft list counter netdev perf test); do + [ ${found} -eq 1 ] && echo "${token}" && return + [ "${token}" = "packets" ] && found=1 + done +} + +# Set MAC addresses, send traffic according to specifier +flood() { + ip link set veth_a address "$(format_mac "${1}")" + ip -n B link set veth_b address "$(format_mac "${2}")" + + for f in ${dst}; do + eval dst_"$f"=\$\(format_\$f "${1}"\) + done + for f in ${src}; do + eval src_"$f"=\$\(format_\$f "${2}"\) + done + eval flood_\$proto +} + +# Set MAC addresses, start pktgen injection +perf() { + dst_mac="$(format_mac "${1}")" + ip link set veth_a address "${dst_mac}" + + for f in ${dst}; do + eval dst_"$f"=\$\(format_\$f "${1}"\) + done + for f in ${src}; do + eval src_"$f"=\$\(format_\$f "${2}"\) + done + eval perf_\$perf_proto +} + +# Set MAC addresses, send single packet, check that it matches, reset counter +send_match() { + ip link set veth_a address "$(format_mac "${1}")" + ip -n B link set veth_b address "$(format_mac "${2}")" + + for f in ${dst}; do + eval dst_"$f"=\$\(format_\$f "${1}"\) + done + for f in ${src}; do + eval src_"$f"=\$\(format_\$f "${2}"\) + done + eval send_\$proto + if [ "$(count_packets)" != "1" ]; then + err "${proto} packet to:" + err " $(for f in ${dst}; do + eval format_\$f "${1}"; printf ' '; done)" + err "from:" + err " $(for f in ${src}; do + eval format_\$f "${2}"; printf ' '; done)" + err "should have matched ruleset:" + err "$(nft -a list ruleset)" + return 1 + fi + nft reset counter inet filter test >/dev/null +} + +# Set MAC addresses, send single packet, check that it doesn't match +send_nomatch() { + ip link set veth_a address "$(format_mac "${1}")" + ip -n B link set veth_b address "$(format_mac "${2}")" + + for f in ${dst}; do + eval dst_"$f"=\$\(format_\$f "${1}"\) + done + for f in ${src}; do + eval src_"$f"=\$\(format_\$f "${2}"\) + done + eval send_\$proto + if [ "$(count_packets)" != "0" ]; then + err "${proto} packet to:" + err " $(for f in ${dst}; do + eval format_\$f "${1}"; printf ' '; done)" + err "from:" + err " $(for f in ${src}; do + eval format_\$f "${2}"; printf ' '; done)" + err "should not have matched ruleset:" + err "$(nft -a list ruleset)" + return 1 + fi +} + +# Correctness test template: +# - add ranged element, check that packets match it +# - check that packets outside range don't match it +# - remove some elements, check that packets don't match anymore +test_correctness() { + setup veth send_"${proto}" set || return ${KSELFTEST_SKIP} + + range_size=1 + for i in $(seq "${start}" $((start + count))); do + end=$((start + range_size)) + + # Avoid negative or zero-sized port ranges + if [ $((end / 65534)) -gt $((start / 65534)) ]; then + start=${end} + end=$((end + 1)) + fi + srcstart=$((start + src_delta)) + srcend=$((end + src_delta)) + + add "$(format)" || return 1 + for j in $(seq ${start} $((range_size / 2 + 1)) ${end}); do + send_match "${j}" $((j + src_delta)) || return 1 + done + send_nomatch $((end + 1)) $((end + 1 + src_delta)) || return 1 + + # Delete elements now and then + if [ $((i % 3)) -eq 0 ]; then + del "$(format)" || return 1 + for j in $(seq ${start} \ + $((range_size / 2 + 1)) ${end}); do + send_nomatch "${j}" $((j + src_delta)) \ + || return 1 + done + fi + + range_size=$((range_size + 1)) + start=$((end + range_size)) + done +} + +# Concurrency test template: +# - add all the elements +# - start a thread for each physical thread that: +# - adds all the elements +# - flushes the set +# - adds all the elements +# - flushes the entire ruleset +# - adds the set back +# - adds all the elements +# - delete all the elements +test_concurrency() { + proto=${flood_proto} + tools=${flood_tools} + chain_spec=${flood_spec} + setup veth flood_"${proto}" set || return ${KSELFTEST_SKIP} + + range_size=1 + cstart=${start} + flood_pids= + for i in $(seq ${start} $((start + count))); do + end=$((start + range_size)) + srcstart=$((start + src_delta)) + srcend=$((end + src_delta)) + + add "$(format)" || return 1 + + flood "${i}" $((i + src_delta)) & flood_pids="${flood_pids} $!" + + range_size=$((range_size + 1)) + start=$((end + range_size)) + done + + sleep 10 + + pids= + for c in $(seq 1 "$(nproc)"); do ( + for r in $(seq 1 "${race_repeat}"); do + range_size=1 + + # $start needs to be local to this subshell + # shellcheck disable=SC2030 + start=${cstart} + for i in $(seq ${start} $((start + count))); do + end=$((start + range_size)) + srcstart=$((start + src_delta)) + srcend=$((end + src_delta)) + + add "$(format)" 2>/dev/null + + range_size=$((range_size + 1)) + start=$((end + range_size)) + done + + nft flush inet filter test 2>/dev/null + + range_size=1 + start=${cstart} + for i in $(seq ${start} $((start + count))); do + end=$((start + range_size)) + srcstart=$((start + src_delta)) + srcend=$((end + src_delta)) + + add "$(format)" 2>/dev/null + + range_size=$((range_size + 1)) + start=$((end + range_size)) + done + + nft flush ruleset + setup set 2>/dev/null + + range_size=1 + start=${cstart} + for i in $(seq ${start} $((start + count))); do + end=$((start + range_size)) + srcstart=$((start + src_delta)) + srcend=$((end + src_delta)) + + add "$(format)" 2>/dev/null + + range_size=$((range_size + 1)) + start=$((end + range_size)) + done + + range_size=1 + start=${cstart} + for i in $(seq ${start} $((start + count))); do + end=$((start + range_size)) + srcstart=$((start + src_delta)) + srcend=$((end + src_delta)) + + del "$(format)" 2>/dev/null + + range_size=$((range_size + 1)) + start=$((end + range_size)) + done + done + ) & pids="${pids} $!" + done + + # shellcheck disable=SC2046,SC2086 # word splitting wanted here + wait $(for pid in ${pids}; do echo ${pid}; done) + # shellcheck disable=SC2046,SC2086 + kill $(for pid in ${flood_pids}; do echo ${pid}; done) 2>/dev/null + # shellcheck disable=SC2046,SC2086 + wait $(for pid in ${flood_pids}; do echo ${pid}; done) 2>/dev/null + + return 0 +} + +# Timeout test template: +# - add all the elements with 3s timeout while checking that packets match +# - wait 3s after the last insertion, check that packets don't match any entry +test_timeout() { + setup veth send_"${proto}" set || return ${KSELFTEST_SKIP} + + timeout=3 + range_size=1 + for i in $(seq "${start}" $((start + count))); do + end=$((start + range_size)) + srcstart=$((start + src_delta)) + srcend=$((end + src_delta)) + + add "$(format)" || return 1 + + for j in $(seq ${start} $((range_size / 2 + 1)) ${end}); do + send_match "${j}" $((j + src_delta)) || return 1 + done + + range_size=$((range_size + 1)) + start=$((end + range_size)) + done + sleep 3 + for i in $(seq ${start} $((start + count))); do + end=$((start + range_size)) + srcstart=$((start + src_delta)) + srcend=$((end + src_delta)) + + for j in $(seq ${start} $((range_size / 2 + 1)) ${end}); do + send_nomatch "${j}" $((j + src_delta)) || return 1 + done + + range_size=$((range_size + 1)) + start=$((end + range_size)) + done +} + +# Performance test template: +# - add concatenated ranged entries +# - add non-ranged concatenated entries (for hash set matching rate baseline) +# - add ranged entries with first field only (for rbhash baseline) +# - start pktgen injection directly on device rx path of this namespace +# - measure drop only rate, hash and rbtree baselines, then matching rate +test_performance() { + chain_spec=${perf_spec} + dst="${perf_dst}" + src="${perf_src}" + setup veth perf set || return ${KSELFTEST_SKIP} + + first=${start} + range_size=1 + for set in test norange noconcat; do + start=${first} + for i in $(seq ${start} $((start + perf_entries))); do + end=$((start + range_size)) + srcstart=$((start + src_delta)) + srcend=$((end + src_delta)) + + if [ $((end / 65534)) -gt $((start / 65534)) ]; then + start=${end} + end=$((end + 1)) + elif [ ${start} -eq ${end} ]; then + end=$((start + 1)) + fi + + add_perf ${set} + + start=$((end + range_size)) + done > "${tmp}" + nft -f "${tmp}" + done + + perf $((end - 1)) ${srcstart} + + sleep 2 + + nft add rule netdev perf test counter name \"test\" drop + nft reset counter netdev perf test >/dev/null 2>&1 + sleep "${perf_duration}" + pps="$(printf %10s $(($(count_perf_packets) / perf_duration)))" + info " baseline (drop from netdev hook): ${pps}pps" + handle="$(nft -a list chain netdev perf test | grep counter)" + handle="${handle##* }" + nft delete rule netdev perf test handle "${handle}" + + nft add rule "netdev perf test ${chain_spec} @norange \ + counter name \"test\" drop" + nft reset counter netdev perf test >/dev/null 2>&1 + sleep "${perf_duration}" + pps="$(printf %10s $(($(count_perf_packets) / perf_duration)))" + info " baseline hash (non-ranged entries): ${pps}pps" + handle="$(nft -a list chain netdev perf test | grep counter)" + handle="${handle##* }" + nft delete rule netdev perf test handle "${handle}" + + nft add rule "netdev perf test ${chain_spec%%. *} @noconcat \ + counter name \"test\" drop" + nft reset counter netdev perf test >/dev/null 2>&1 + sleep "${perf_duration}" + pps="$(printf %10s $(($(count_perf_packets) / perf_duration)))" + info " baseline rbtree (match on first field only): ${pps}pps" + handle="$(nft -a list chain netdev perf test | grep counter)" + handle="${handle##* }" + nft delete rule netdev perf test handle "${handle}" + + nft add rule "netdev perf test ${chain_spec} @test \ + counter name \"test\" drop" + nft reset counter netdev perf test >/dev/null 2>&1 + sleep "${perf_duration}" + pps="$(printf %10s $(($(count_perf_packets) / perf_duration)))" + p5="$(printf %5s "${perf_entries}")" + info " set with ${p5} full, ranged entries: ${pps}pps" + kill "${perf_pid}" +} + +test_bug_flush_remove_add() { + set_cmd='{ set s { type ipv4_addr . inet_service; flags interval; }; }' + elem1='{ 10.0.0.1 . 22-25, 10.0.0.1 . 10-20 }' + elem2='{ 10.0.0.1 . 10-20, 10.0.0.1 . 22-25 }' + for i in `seq 1 100`; do + nft add table t ${set_cmd} || return ${KSELFTEST_SKIP} + nft add element t s ${elem1} 2>/dev/null || return 1 + nft flush set t s 2>/dev/null || return 1 + nft add element t s ${elem2} 2>/dev/null || return 1 + done + nft flush ruleset +} + +# - add ranged element, check that packets match it +# - reload the set, check packets still match +test_bug_reload() { + setup veth send_"${proto}" set || return ${KSELFTEST_SKIP} + rstart=${start} + + range_size=1 + for i in $(seq "${start}" $((start + count))); do + end=$((start + range_size)) + + # Avoid negative or zero-sized port ranges + if [ $((end / 65534)) -gt $((start / 65534)) ]; then + start=${end} + end=$((end + 1)) + fi + srcstart=$((start + src_delta)) + srcend=$((end + src_delta)) + + add "$(format)" || return 1 + range_size=$((range_size + 1)) + start=$((end + range_size)) + done + + # check kernel does allocate pcpu sctrach map + # for reload with no elemet add/delete + ( echo flush set inet filter test ; + nft list set inet filter test ) | nft -f - + + start=${rstart} + range_size=1 + + for i in $(seq "${start}" $((start + count))); do + end=$((start + range_size)) + + # Avoid negative or zero-sized port ranges + if [ $((end / 65534)) -gt $((start / 65534)) ]; then + start=${end} + end=$((end + 1)) + fi + srcstart=$((start + src_delta)) + srcend=$((end + src_delta)) + + for j in $(seq ${start} $((range_size / 2 + 1)) ${end}); do + send_match "${j}" $((j + src_delta)) || return 1 + done + + range_size=$((range_size + 1)) + start=$((end + range_size)) + done + + nft flush ruleset +} + +test_reported_issues() { + eval test_bug_"${subtest}" +} + +# Run everything in a separate network namespace +[ "${1}" != "run" ] && { unshare -n "${0}" run; exit $?; } +tmp="$(mktemp)" +trap cleanup EXIT + +# Entry point for test runs +passed=0 +for name in ${TESTS}; do + printf "TEST: %s\n" "$(echo ${name} | tr '_' ' ')" + if [ "${name}" = "reported_issues" ]; then + SUBTESTS="${BUGS}" + else + SUBTESTS="${TYPES}" + fi + + for subtest in ${SUBTESTS}; do + eval desc=\$TYPE_"${subtest}" + IFS=' +' + for __line in ${desc}; do + # shellcheck disable=SC2086 + eval ${__line%% *}=\"${__line##* }\"; + done + IFS=' +' + + if [ "${name}" = "concurrency" ] && \ + [ "${race_repeat}" = "0" ]; then + continue + fi + if [ "${name}" = "performance" ] && \ + [ "${perf_duration}" = "0" ]; then + continue + fi + + printf " %-60s " "${display}" + eval test_"${name}" + ret=$? + + if [ $ret -eq 0 ]; then + printf "[ OK ]\n" + info_flush + passed=$((passed + 1)) + elif [ $ret -eq 1 ]; then + printf "[FAIL]\n" + err_flush + exit 1 + elif [ $ret -eq ${KSELFTEST_SKIP} ]; then + printf "[SKIP]\n" + err_flush + fi + done +done + +[ ${passed} -eq 0 ] && exit ${KSELFTEST_SKIP} || exit 0 diff --git a/tools/testing/selftests/netfilter/nft_conntrack_helper.sh b/tools/testing/selftests/netfilter/nft_conntrack_helper.sh new file mode 100755 index 000000000..bf6b9626c --- /dev/null +++ b/tools/testing/selftests/netfilter/nft_conntrack_helper.sh @@ -0,0 +1,181 @@ +#!/bin/bash +# +# This tests connection tracking helper assignment: +# 1. can attach ftp helper to a connection from nft ruleset. +# 2. auto-assign still works. +# +# Kselftest framework requirement - SKIP code is 4. +ksft_skip=4 +ret=0 + +sfx=$(mktemp -u "XXXXXXXX") +ns1="ns1-$sfx" +ns2="ns2-$sfx" +testipv6=1 + +cleanup() +{ + ip netns del ${ns1} + ip netns del ${ns2} +} + +nft --version > /dev/null 2>&1 +if [ $? -ne 0 ];then + echo "SKIP: Could not run test without nft tool" + exit $ksft_skip +fi + +ip -Version > /dev/null 2>&1 +if [ $? -ne 0 ];then + echo "SKIP: Could not run test without ip tool" + exit $ksft_skip +fi + +conntrack -V > /dev/null 2>&1 +if [ $? -ne 0 ];then + echo "SKIP: Could not run test without conntrack tool" + exit $ksft_skip +fi + +which nc >/dev/null 2>&1 +if [ $? -ne 0 ];then + echo "SKIP: Could not run test without netcat tool" + exit $ksft_skip +fi + +trap cleanup EXIT + +ip netns add ${ns1} +ip netns add ${ns2} + +ip link add veth0 netns ${ns1} type veth peer name veth0 netns ${ns2} > /dev/null 2>&1 +if [ $? -ne 0 ];then + echo "SKIP: No virtual ethernet pair device support in kernel" + exit $ksft_skip +fi + +ip -net ${ns1} link set lo up +ip -net ${ns1} link set veth0 up + +ip -net ${ns2} link set lo up +ip -net ${ns2} link set veth0 up + +ip -net ${ns1} addr add 10.0.1.1/24 dev veth0 +ip -net ${ns1} addr add dead:1::1/64 dev veth0 + +ip -net ${ns2} addr add 10.0.1.2/24 dev veth0 +ip -net ${ns2} addr add dead:1::2/64 dev veth0 + +load_ruleset_family() { + local family=$1 + local ns=$2 + +ip netns exec ${ns} nft -f - <<EOF +table $family raw { + ct helper ftp { + type "ftp" protocol tcp + } + chain pre { + type filter hook prerouting priority 0; policy accept; + tcp dport 2121 ct helper set "ftp" + } + chain output { + type filter hook output priority 0; policy accept; + tcp dport 2121 ct helper set "ftp" + } +} +EOF + return $? +} + +check_for_helper() +{ + local netns=$1 + local message=$2 + local port=$3 + + if echo $message |grep -q 'ipv6';then + local family="ipv6" + else + local family="ipv4" + fi + + ip netns exec ${netns} conntrack -L -f $family -p tcp --dport $port 2> /dev/null |grep -q 'helper=ftp' + if [ $? -ne 0 ] ; then + echo "FAIL: ${netns} did not show attached helper $message" 1>&2 + ret=1 + fi + + echo "PASS: ${netns} connection on port $port has ftp helper attached" 1>&2 + return 0 +} + +test_helper() +{ + local port=$1 + local msg=$2 + + sleep 3 | ip netns exec ${ns2} nc -w 2 -l -p $port > /dev/null & + + sleep 1 | ip netns exec ${ns1} nc -w 2 10.0.1.2 $port > /dev/null & + sleep 1 + + check_for_helper "$ns1" "ip $msg" $port + check_for_helper "$ns2" "ip $msg" $port + + wait + + if [ $testipv6 -eq 0 ] ;then + return 0 + fi + + ip netns exec ${ns1} conntrack -F 2> /dev/null + ip netns exec ${ns2} conntrack -F 2> /dev/null + + sleep 3 | ip netns exec ${ns2} nc -w 2 -6 -l -p $port > /dev/null & + + sleep 1 | ip netns exec ${ns1} nc -w 2 -6 dead:1::2 $port > /dev/null & + sleep 1 + + check_for_helper "$ns1" "ipv6 $msg" $port + check_for_helper "$ns2" "ipv6 $msg" $port + + wait +} + +load_ruleset_family ip ${ns1} +if [ $? -ne 0 ];then + echo "FAIL: ${ns1} cannot load ip ruleset" 1>&2 + exit 1 +fi + +load_ruleset_family ip6 ${ns1} +if [ $? -ne 0 ];then + echo "SKIP: ${ns1} cannot load ip6 ruleset" 1>&2 + testipv6=0 +fi + +load_ruleset_family inet ${ns2} +if [ $? -ne 0 ];then + echo "SKIP: ${ns1} cannot load inet ruleset" 1>&2 + load_ruleset_family ip ${ns2} + if [ $? -ne 0 ];then + echo "FAIL: ${ns2} cannot load ip ruleset" 1>&2 + exit 1 + fi + + if [ $testipv6 -eq 1 ] ;then + load_ruleset_family ip6 ${ns2} + if [ $? -ne 0 ];then + echo "FAIL: ${ns2} cannot load ip6 ruleset" 1>&2 + exit 1 + fi + fi +fi + +test_helper 2121 "set via ruleset" +ip netns exec ${ns1} sysctl -q 'net.netfilter.nf_conntrack_helper=1' +ip netns exec ${ns2} sysctl -q 'net.netfilter.nf_conntrack_helper=1' +test_helper 21 "auto-assign" + +exit $ret diff --git a/tools/testing/selftests/netfilter/nft_flowtable.sh b/tools/testing/selftests/netfilter/nft_flowtable.sh new file mode 100755 index 000000000..aefe50e0e --- /dev/null +++ b/tools/testing/selftests/netfilter/nft_flowtable.sh @@ -0,0 +1,420 @@ +#!/bin/bash +# SPDX-License-Identifier: GPL-2.0 +# +# This tests basic flowtable functionality. +# Creates following default topology: +# +# Originator (MTU 9000) <-Router1-> MTU 1500 <-Router2-> Responder (MTU 2000) +# Router1 is the one doing flow offloading, Router2 has no special +# purpose other than having a link that is smaller than either Originator +# and responder, i.e. TCPMSS announced values are too large and will still +# result in fragmentation and/or PMTU discovery. +# +# You can check with different Orgininator/Link/Responder MTU eg: +# nft_flowtable.sh -o8000 -l1500 -r2000 +# + + +# Kselftest framework requirement - SKIP code is 4. +ksft_skip=4 +ret=0 + +ns1in="" +ns2in="" +ns1out="" +ns2out="" + +log_netns=$(sysctl -n net.netfilter.nf_log_all_netns) + +checktool (){ + if ! $1 > /dev/null 2>&1; then + echo "SKIP: Could not $2" + exit $ksft_skip + fi +} + +checktool "nft --version" "run test without nft tool" +checktool "ip -Version" "run test without ip tool" +checktool "which nc" "run test without nc (netcat)" +checktool "ip netns add nsr1" "create net namespace" + +ip netns add ns1 +ip netns add ns2 + +ip netns add nsr2 + +cleanup() { + for i in 1 2; do + ip netns del ns$i + ip netns del nsr$i + done + + rm -f "$ns1in" "$ns1out" + rm -f "$ns2in" "$ns2out" + + [ $log_netns -eq 0 ] && sysctl -q net.netfilter.nf_log_all_netns=$log_netns +} + +trap cleanup EXIT + +sysctl -q net.netfilter.nf_log_all_netns=1 + +ip link add veth0 netns nsr1 type veth peer name eth0 netns ns1 +ip link add veth1 netns nsr1 type veth peer name veth0 netns nsr2 + +ip link add veth1 netns nsr2 type veth peer name eth0 netns ns2 + +for dev in lo veth0 veth1; do + for i in 1 2; do + ip -net nsr$i link set $dev up + done +done + +ip -net nsr1 addr add 10.0.1.1/24 dev veth0 +ip -net nsr1 addr add dead:1::1/64 dev veth0 + +ip -net nsr2 addr add 10.0.2.1/24 dev veth1 +ip -net nsr2 addr add dead:2::1/64 dev veth1 + +# set different MTUs so we need to push packets coming from ns1 (large MTU) +# to ns2 (smaller MTU) to stack either to perform fragmentation (ip_no_pmtu_disc=1), +# or to do PTMU discovery (send ICMP error back to originator). +# ns2 is going via nsr2 with a smaller mtu, so that TCPMSS announced by both peers +# is NOT the lowest link mtu. + +omtu=9000 +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 + +if ! ip -net nsr1 link set veth0 mtu $omtu; then + exit 1 +fi + +ip -net ns1 link set eth0 mtu $omtu + +if ! ip -net nsr2 link set veth1 mtu $rmtu; then + exit 1 +fi + +ip -net ns2 link set eth0 mtu $rmtu + +# transfer-net between nsr1 and nsr2. +# these addresses are not used for connections. +ip -net nsr1 addr add 192.168.10.1/24 dev veth1 +ip -net nsr1 addr add fee1:2::1/64 dev veth1 + +ip -net nsr2 addr add 192.168.10.2/24 dev veth0 +ip -net nsr2 addr add fee1:2::2/64 dev veth0 + +for i in 1 2; do + ip netns exec nsr$i sysctl net.ipv4.conf.veth0.forwarding=1 > /dev/null + ip netns exec nsr$i sysctl net.ipv4.conf.veth1.forwarding=1 > /dev/null + + ip -net ns$i link set lo up + ip -net ns$i link set eth0 up + ip -net ns$i addr add 10.0.$i.99/24 dev eth0 + ip -net ns$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:$i::1 + if ! ip netns exec ns$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 +done + +ip -net nsr1 route add default via 192.168.10.2 +ip -net nsr2 route add default via 192.168.10.1 + +ip netns exec nsr1 nft -f - <<EOF +table inet filter { + flowtable f1 { + hook ingress priority 0 + devices = { veth0, veth1 } + } + + chain forward { + type filter hook forward priority 0; policy drop; + + # flow offloaded? Tag ct with mark 1, so we can detect when it fails. + meta oif "veth1" tcp dport 12345 flow offload @f1 counter + + # use packet size to trigger 'should be offloaded by now'. + # otherwise, if 'flow offload' expression never offloads, the + # test will pass. + tcp dport 12345 meta length gt 200 ct mark set 1 counter + + # this turns off flow offloading internally, so expect packets again + tcp flags fin,rst ct mark set 0 accept + + # this allows large packets from responder, we need this as long + # as PMTUd is off. + # This rule is deleted for the last test, when we expect PMTUd + # to kick in and ensure all packets meet mtu requirements. + meta length gt $lmtu accept comment something-to-grep-for + + # next line blocks connection w.o. working offload. + # we only do this for reverse dir, because we expect packets to + # enter slow path due to MTU mismatch of veth0 and veth1. + tcp sport 12345 ct mark 1 counter log prefix "mark failure " drop + + ct state established,related accept + + # for packets that we can't offload yet, i.e. SYN (any ct that is not confirmed) + meta length lt 200 oif "veth1" tcp dport 12345 counter accept + + meta nfproto ipv4 meta l4proto icmp accept + meta nfproto ipv6 meta l4proto icmpv6 accept + } +} +EOF + +if [ $? -ne 0 ]; then + echo "SKIP: Could not load nft ruleset" + exit $ksft_skip +fi + +# test basic connectivity +if ! ip netns exec ns1 ping -c 1 -q 10.0.2.99 > /dev/null; then + echo "ERROR: ns1 cannot reach ns2" 1>&2 + exit 1 +fi + +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 + +if [ $ret -eq 0 ];then + echo "PASS: netns routing/connectivity: ns1 can reach ns2" +fi + +ns1in=$(mktemp) +ns1out=$(mktemp) +ns2in=$(mktemp) +ns2out=$(mktemp) + +make_file() +{ + name=$1 + + SIZE=$((RANDOM % (1024 * 8))) + TSIZE=$((SIZE * 1024)) + + dd if=/dev/urandom of="$name" bs=1024 count=$SIZE 2> /dev/null + + SIZE=$((RANDOM % 1024)) + SIZE=$((SIZE + 128)) + TSIZE=$((TSIZE + SIZE)) + dd if=/dev/urandom conf=notrunc of="$name" bs=1 count=$SIZE 2> /dev/null +} + +check_transfer() +{ + in=$1 + out=$2 + what=$3 + + if ! cmp "$in" "$out" > /dev/null 2>&1; then + echo "FAIL: file mismatch for $what" 1>&2 + ls -l "$in" + ls -l "$out" + return 1 + fi + + return 0 +} + +test_tcp_forwarding_ip() +{ + local nsa=$1 + local nsb=$2 + local dstip=$3 + local dstport=$4 + local lret=0 + + ip netns exec $nsb nc -w 5 -l -p 12345 < "$ns2in" > "$ns2out" & + lpid=$! + + sleep 1 + ip netns exec $nsa nc -w 4 "$dstip" "$dstport" < "$ns1in" > "$ns1out" & + cpid=$! + + sleep 3 + + if ps -p $lpid > /dev/null;then + kill $lpid + fi + + if ps -p $cpid > /dev/null;then + kill $cpid + fi + + wait + + if ! check_transfer "$ns1in" "$ns2out" "ns1 -> ns2"; then + lret=1 + fi + + if ! check_transfer "$ns2in" "$ns1out" "ns1 <- ns2"; then + lret=1 + fi + + return $lret +} + +test_tcp_forwarding() +{ + test_tcp_forwarding_ip "$1" "$2" 10.0.2.99 12345 + + return $? +} + +test_tcp_forwarding_nat() +{ + local lret + + test_tcp_forwarding_ip "$1" "$2" 10.0.2.99 12345 + lret=$? + + if [ $lret -eq 0 ] ; then + test_tcp_forwarding_ip "$1" "$2" 10.6.6.6 1666 + lret=$? + fi + + return $lret +} + +make_file "$ns1in" +make_file "$ns2in" + +# First test: +# No PMTU discovery, nsr1 is expected to fragment packets from ns1 to ns2 as needed. +if test_tcp_forwarding ns1 ns2; then + echo "PASS: flow offloaded for ns1/ns2" +else + echo "FAIL: flow offload for ns1/ns2:" 1>&2 + ip netns exec nsr1 nft list ruleset + ret=1 +fi + +# delete default route, i.e. ns2 won't be able to reach ns1 and +# will depend on ns1 being masqueraded in nsr1. +# expect ns1 has nsr1 address. +ip -net ns2 route del default via 10.0.2.1 +ip -net ns2 route del default via dead:2::1 +ip -net ns2 route add 192.168.10.1 via 10.0.2.1 + +# Second test: +# Same, but with NAT enabled. +ip netns exec nsr1 nft -f - <<EOF +table ip nat { + chain prerouting { + type nat hook prerouting priority 0; policy accept; + meta iif "veth0" ip daddr 10.6.6.6 tcp dport 1666 counter dnat ip to 10.0.2.99:12345 + } + + chain postrouting { + type nat hook postrouting priority 0; policy accept; + meta oifname "veth1" counter masquerade + } +} +EOF + +if test_tcp_forwarding_nat ns1 ns2; then + echo "PASS: flow offloaded for ns1/ns2 with NAT" +else + echo "FAIL: flow offload for ns1/ns2 with NAT" 1>&2 + ip netns exec nsr1 nft list ruleset + ret=1 +fi + +# Third test: +# Same as second test, but with PMTU discovery enabled. +handle=$(ip netns exec nsr1 nft -a list table inet filter | grep something-to-grep-for | cut -d \# -f 2) + +if ! ip netns exec nsr1 nft delete rule inet filter forward $handle; then + echo "FAIL: Could not delete large-packet accept rule" + exit 1 +fi + +ip netns exec ns1 sysctl net.ipv4.ip_no_pmtu_disc=0 > /dev/null +ip netns exec ns2 sysctl net.ipv4.ip_no_pmtu_disc=0 > /dev/null + +if test_tcp_forwarding_nat ns1 ns2; then + echo "PASS: flow offloaded for ns1/ns2 with NAT and pmtu discovery" +else + echo "FAIL: flow offload for ns1/ns2 with NAT and pmtu discovery" 1>&2 + ip netns exec nsr1 nft list ruleset +fi + +KEY_SHA="0x"$(ps -xaf | sha1sum | cut -d " " -f 1) +KEY_AES="0x"$(ps -xaf | md5sum | cut -d " " -f 1) +SPI1=$RANDOM +SPI2=$RANDOM + +if [ $SPI1 -eq $SPI2 ]; then + SPI2=$((SPI2+1)) +fi + +do_esp() { + local ns=$1 + local me=$2 + local remote=$3 + local lnet=$4 + local rnet=$5 + local spi_out=$6 + local spi_in=$7 + + ip -net $ns xfrm state add src $remote dst $me proto esp spi $spi_in enc aes $KEY_AES auth sha1 $KEY_SHA mode tunnel sel src $rnet dst $lnet + ip -net $ns xfrm state add src $me dst $remote proto esp spi $spi_out enc aes $KEY_AES auth sha1 $KEY_SHA mode tunnel sel src $lnet dst $rnet + + # to encrypt packets as they go out (includes forwarded packets that need encapsulation) + ip -net $ns xfrm policy add src $lnet dst $rnet dir out tmpl src $me dst $remote proto esp mode tunnel priority 1 action allow + # to fwd decrypted packets after esp processing: + ip -net $ns xfrm policy add src $rnet dst $lnet dir fwd tmpl src $remote dst $me proto esp mode tunnel priority 1 action allow + +} + +do_esp nsr1 192.168.10.1 192.168.10.2 10.0.1.0/24 10.0.2.0/24 $SPI1 $SPI2 + +do_esp nsr2 192.168.10.2 192.168.10.1 10.0.2.0/24 10.0.1.0/24 $SPI2 $SPI1 + +ip netns exec nsr1 nft delete table ip nat + +# restore default routes +ip -net ns2 route del 192.168.10.1 via 10.0.2.1 +ip -net ns2 route add default via 10.0.2.1 +ip -net ns2 route add default via dead:2::1 + +if test_tcp_forwarding ns1 ns2; then + echo "PASS: ipsec tunnel mode for ns1/ns2" +else + echo "FAIL: ipsec tunnel mode for ns1/ns2" + ip netns exec nsr1 nft list ruleset 1>&2 + ip netns exec nsr1 cat /proc/net/xfrm_stat 1>&2 +fi + +exit $ret diff --git a/tools/testing/selftests/netfilter/nft_meta.sh b/tools/testing/selftests/netfilter/nft_meta.sh new file mode 100755 index 000000000..f33154c04 --- /dev/null +++ b/tools/testing/selftests/netfilter/nft_meta.sh @@ -0,0 +1,142 @@ +#!/bin/bash + +# check iif/iifname/oifgroup/iiftype match. + +# Kselftest framework requirement - SKIP code is 4. +ksft_skip=4 +sfx=$(mktemp -u "XXXXXXXX") +ns0="ns0-$sfx" + +if ! nft --version > /dev/null 2>&1; then + echo "SKIP: Could not run test without nft tool" + exit $ksft_skip +fi + +cleanup() +{ + ip netns del "$ns0" +} + +ip netns add "$ns0" +ip -net "$ns0" link set lo up +ip -net "$ns0" addr add 127.0.0.1 dev lo + +trap cleanup EXIT + +currentyear=$(date +%Y) +lastyear=$((currentyear-1)) +ip netns exec "$ns0" nft -f /dev/stdin <<EOF +table inet filter { + counter iifcount {} + counter iifnamecount {} + counter iifgroupcount {} + counter iiftypecount {} + counter infproto4count {} + counter il4protocounter {} + counter imarkcounter {} + counter icpu0counter {} + counter ilastyearcounter {} + counter icurrentyearcounter {} + + counter oifcount {} + counter oifnamecount {} + counter oifgroupcount {} + counter oiftypecount {} + counter onfproto4count {} + counter ol4protocounter {} + counter oskuidcounter {} + counter oskgidcounter {} + counter omarkcounter {} + + chain input { + type filter hook input priority 0; policy accept; + + meta iif lo counter name "iifcount" + meta iifname "lo" counter name "iifnamecount" + meta iifgroup "default" counter name "iifgroupcount" + meta iiftype "loopback" counter name "iiftypecount" + meta nfproto ipv4 counter name "infproto4count" + meta l4proto icmp counter name "il4protocounter" + meta mark 42 counter name "imarkcounter" + meta cpu 0 counter name "icpu0counter" + meta time "$lastyear-01-01" - "$lastyear-12-31" counter name ilastyearcounter + meta time "$currentyear-01-01" - "$currentyear-12-31" counter name icurrentyearcounter + } + + chain output { + type filter hook output priority 0; policy accept; + meta oif lo counter name "oifcount" counter + meta oifname "lo" counter name "oifnamecount" + meta oifgroup "default" counter name "oifgroupcount" + meta oiftype "loopback" counter name "oiftypecount" + meta nfproto ipv4 counter name "onfproto4count" + meta l4proto icmp counter name "ol4protocounter" + meta skuid 0 counter name "oskuidcounter" + meta skgid 0 counter name "oskgidcounter" + meta mark 42 counter name "omarkcounter" + } +} +EOF + +if [ $? -ne 0 ]; then + echo "SKIP: Could not add test ruleset" + exit $ksft_skip +fi + +ret=0 + +check_one_counter() +{ + local cname="$1" + local want="packets $2" + local verbose="$3" + + if ! ip netns exec "$ns0" nft list counter inet filter $cname | grep -q "$want"; then + echo "FAIL: $cname, want \"$want\", got" + ret=1 + ip netns exec "$ns0" nft list counter inet filter $cname + fi +} + +check_lo_counters() +{ + local want="$1" + local verbose="$2" + local counter + + for counter in iifcount iifnamecount iifgroupcount iiftypecount infproto4count \ + oifcount oifnamecount oifgroupcount oiftypecount onfproto4count \ + il4protocounter icurrentyearcounter ol4protocounter \ + ; do + check_one_counter "$counter" "$want" "$verbose" + done +} + +check_lo_counters "0" false +ip netns exec "$ns0" ping -q -c 1 127.0.0.1 -m 42 > /dev/null + +check_lo_counters "2" true + +check_one_counter oskuidcounter "1" true +check_one_counter oskgidcounter "1" true +check_one_counter imarkcounter "1" true +check_one_counter omarkcounter "1" true +check_one_counter ilastyearcounter "0" true + +if [ $ret -eq 0 ];then + echo "OK: nftables meta iif/oif counters at expected values" +else + exit $ret +fi + +#First CPU execution and counter +taskset -p 01 $$ > /dev/null +ip netns exec "$ns0" nft reset counters > /dev/null +ip netns exec "$ns0" ping -q -c 1 127.0.0.1 > /dev/null +check_one_counter icpu0counter "2" true + +if [ $ret -eq 0 ];then + echo "OK: nftables meta cpu counter at expected values" +fi + +exit $ret diff --git a/tools/testing/selftests/netfilter/nft_nat.sh b/tools/testing/selftests/netfilter/nft_nat.sh new file mode 100755 index 000000000..67697d8ea --- /dev/null +++ b/tools/testing/selftests/netfilter/nft_nat.sh @@ -0,0 +1,914 @@ +#!/bin/bash +# +# This test is for basic NAT functionality: snat, dnat, redirect, masquerade. +# + +# Kselftest framework requirement - SKIP code is 4. +ksft_skip=4 +ret=0 +test_inet_nat=true + +sfx=$(mktemp -u "XXXXXXXX") +ns0="ns0-$sfx" +ns1="ns1-$sfx" +ns2="ns2-$sfx" + +cleanup() +{ + for i in 0 1 2; do ip netns del ns$i-"$sfx";done +} + +nft --version > /dev/null 2>&1 +if [ $? -ne 0 ];then + echo "SKIP: Could not run test without nft tool" + exit $ksft_skip +fi + +ip -Version > /dev/null 2>&1 +if [ $? -ne 0 ];then + echo "SKIP: Could not run test without ip tool" + exit $ksft_skip +fi + +ip netns add "$ns0" +if [ $? -ne 0 ];then + echo "SKIP: Could not create net namespace $ns0" + exit $ksft_skip +fi + +trap cleanup EXIT + +ip netns add "$ns1" +if [ $? -ne 0 ];then + echo "SKIP: Could not create net namespace $ns1" + exit $ksft_skip +fi + +ip netns add "$ns2" +if [ $? -ne 0 ];then + echo "SKIP: Could not create net namespace $ns2" + exit $ksft_skip +fi + +ip link add veth0 netns "$ns0" type veth peer name eth0 netns "$ns1" > /dev/null 2>&1 +if [ $? -ne 0 ];then + echo "SKIP: No virtual ethernet pair device support in kernel" + exit $ksft_skip +fi +ip link add veth1 netns "$ns0" type veth peer name eth0 netns "$ns2" + +ip -net "$ns0" link set lo up +ip -net "$ns0" link set veth0 up +ip -net "$ns0" addr add 10.0.1.1/24 dev veth0 +ip -net "$ns0" addr add dead:1::1/64 dev veth0 + +ip -net "$ns0" link set veth1 up +ip -net "$ns0" addr add 10.0.2.1/24 dev veth1 +ip -net "$ns0" addr add dead:2::1/64 dev veth1 + +for i in 1 2; do + ip -net ns$i-$sfx link set lo up + ip -net ns$i-$sfx link set eth0 up + ip -net ns$i-$sfx addr add 10.0.$i.99/24 dev eth0 + ip -net ns$i-$sfx route add default via 10.0.$i.1 + ip -net ns$i-$sfx addr add dead:$i::99/64 dev eth0 + ip -net ns$i-$sfx route add default via dead:$i::1 +done + +bad_counter() +{ + local ns=$1 + local counter=$2 + local expect=$3 + local tag=$4 + + echo "ERROR: $counter counter in $ns has unexpected value (expected $expect) at $tag" 1>&2 + ip netns exec $ns nft list counter inet filter $counter 1>&2 +} + +check_counters() +{ + ns=$1 + local lret=0 + + cnt=$(ip netns exec $ns nft list counter inet filter ns0in | grep -q "packets 1 bytes 84") + if [ $? -ne 0 ]; then + bad_counter $ns ns0in "packets 1 bytes 84" "check_counters 1" + lret=1 + fi + cnt=$(ip netns exec $ns nft list counter inet filter ns0out | grep -q "packets 1 bytes 84") + if [ $? -ne 0 ]; then + bad_counter $ns ns0out "packets 1 bytes 84" "check_counters 2" + lret=1 + fi + + expect="packets 1 bytes 104" + cnt=$(ip netns exec $ns nft list counter inet filter ns0in6 | grep -q "$expect") + if [ $? -ne 0 ]; then + bad_counter $ns ns0in6 "$expect" "check_counters 3" + lret=1 + fi + cnt=$(ip netns exec $ns nft list counter inet filter ns0out6 | grep -q "$expect") + if [ $? -ne 0 ]; then + bad_counter $ns ns0out6 "$expect" "check_counters 4" + lret=1 + fi + + return $lret +} + +check_ns0_counters() +{ + local ns=$1 + local lret=0 + + cnt=$(ip netns exec "$ns0" nft list counter inet filter ns0in | grep -q "packets 0 bytes 0") + if [ $? -ne 0 ]; then + bad_counter "$ns0" ns0in "packets 0 bytes 0" "check_ns0_counters 1" + lret=1 + fi + + cnt=$(ip netns exec "$ns0" nft list counter inet filter ns0in6 | grep -q "packets 0 bytes 0") + if [ $? -ne 0 ]; then + bad_counter "$ns0" ns0in6 "packets 0 bytes 0" + lret=1 + fi + + cnt=$(ip netns exec "$ns0" nft list counter inet filter ns0out | grep -q "packets 0 bytes 0") + if [ $? -ne 0 ]; then + bad_counter "$ns0" ns0out "packets 0 bytes 0" "check_ns0_counters 2" + lret=1 + fi + cnt=$(ip netns exec "$ns0" nft list counter inet filter ns0out6 | grep -q "packets 0 bytes 0") + if [ $? -ne 0 ]; then + bad_counter "$ns0" ns0out6 "packets 0 bytes 0" "check_ns0_counters3 " + lret=1 + fi + + for dir in "in" "out" ; do + expect="packets 1 bytes 84" + cnt=$(ip netns exec "$ns0" nft list counter inet filter ${ns}${dir} | grep -q "$expect") + if [ $? -ne 0 ]; then + bad_counter "$ns0" $ns$dir "$expect" "check_ns0_counters 4" + lret=1 + fi + + expect="packets 1 bytes 104" + cnt=$(ip netns exec "$ns0" nft list counter inet filter ${ns}${dir}6 | grep -q "$expect") + if [ $? -ne 0 ]; then + bad_counter "$ns0" $ns$dir6 "$expect" "check_ns0_counters 5" + lret=1 + fi + done + + return $lret +} + +reset_counters() +{ + for i in 0 1 2;do + ip netns exec ns$i-$sfx nft reset counters inet > /dev/null + done +} + +test_local_dnat6() +{ + local family=$1 + local lret=0 + local IPF="" + + if [ $family = "inet" ];then + IPF="ip6" + fi + +ip netns exec "$ns0" nft -f /dev/stdin <<EOF +table $family nat { + chain output { + type nat hook output priority 0; policy accept; + ip6 daddr dead:1::99 dnat $IPF to dead:2::99 + } +} +EOF + if [ $? -ne 0 ]; then + echo "SKIP: Could not add add $family dnat hook" + return $ksft_skip + fi + + # ping netns1, expect rewrite to netns2 + ip netns exec "$ns0" ping -q -c 1 dead:1::99 > /dev/null + if [ $? -ne 0 ]; then + lret=1 + echo "ERROR: ping6 failed" + return $lret + fi + + expect="packets 0 bytes 0" + for dir in "in6" "out6" ; do + cnt=$(ip netns exec "$ns0" nft list counter inet filter ns1${dir} | grep -q "$expect") + if [ $? -ne 0 ]; then + bad_counter "$ns0" ns1$dir "$expect" "test_local_dnat6 1" + lret=1 + fi + done + + expect="packets 1 bytes 104" + for dir in "in6" "out6" ; do + cnt=$(ip netns exec "$ns0" nft list counter inet filter ns2${dir} | grep -q "$expect") + if [ $? -ne 0 ]; then + bad_counter "$ns0" ns2$dir "$expect" "test_local_dnat6 2" + lret=1 + fi + done + + # expect 0 count in ns1 + expect="packets 0 bytes 0" + for dir in "in6" "out6" ; do + cnt=$(ip netns exec "$ns1" nft list counter inet filter ns0${dir} | grep -q "$expect") + if [ $? -ne 0 ]; then + bad_counter "$ns1" ns0$dir "$expect" "test_local_dnat6 3" + lret=1 + fi + done + + # expect 1 packet in ns2 + expect="packets 1 bytes 104" + for dir in "in6" "out6" ; do + cnt=$(ip netns exec "$ns2" nft list counter inet filter ns0${dir} | grep -q "$expect") + if [ $? -ne 0 ]; then + bad_counter "$ns2" ns0$dir "$expect" "test_local_dnat6 4" + lret=1 + fi + done + + test $lret -eq 0 && echo "PASS: ipv6 ping to $ns1 was $family NATted to $ns2" + ip netns exec "$ns0" nft flush chain ip6 nat output + + return $lret +} + +test_local_dnat() +{ + local family=$1 + local lret=0 + local IPF="" + + if [ $family = "inet" ];then + IPF="ip" + fi + +ip netns exec "$ns0" nft -f /dev/stdin <<EOF 2>/dev/null +table $family nat { + chain output { + type nat hook output priority 0; policy accept; + ip daddr 10.0.1.99 dnat $IPF to 10.0.2.99 + } +} +EOF + if [ $? -ne 0 ]; then + if [ $family = "inet" ];then + echo "SKIP: inet nat tests" + test_inet_nat=false + return $ksft_skip + fi + echo "SKIP: Could not add add $family dnat hook" + return $ksft_skip + fi + + # ping netns1, expect rewrite to netns2 + ip netns exec "$ns0" ping -q -c 1 10.0.1.99 > /dev/null + if [ $? -ne 0 ]; then + lret=1 + echo "ERROR: ping failed" + return $lret + fi + + expect="packets 0 bytes 0" + for dir in "in" "out" ; do + cnt=$(ip netns exec "$ns0" nft list counter inet filter ns1${dir} | grep -q "$expect") + if [ $? -ne 0 ]; then + bad_counter "$ns0" ns1$dir "$expect" "test_local_dnat 1" + lret=1 + fi + done + + expect="packets 1 bytes 84" + for dir in "in" "out" ; do + cnt=$(ip netns exec "$ns0" nft list counter inet filter ns2${dir} | grep -q "$expect") + if [ $? -ne 0 ]; then + bad_counter "$ns0" ns2$dir "$expect" "test_local_dnat 2" + lret=1 + fi + done + + # expect 0 count in ns1 + expect="packets 0 bytes 0" + for dir in "in" "out" ; do + cnt=$(ip netns exec "$ns1" nft list counter inet filter ns0${dir} | grep -q "$expect") + if [ $? -ne 0 ]; then + bad_counter "$ns1" ns0$dir "$expect" "test_local_dnat 3" + lret=1 + fi + done + + # expect 1 packet in ns2 + expect="packets 1 bytes 84" + for dir in "in" "out" ; do + cnt=$(ip netns exec "$ns2" nft list counter inet filter ns0${dir} | grep -q "$expect") + if [ $? -ne 0 ]; then + bad_counter "$ns2" ns0$dir "$expect" "test_local_dnat 4" + lret=1 + fi + done + + test $lret -eq 0 && echo "PASS: ping to $ns1 was $family NATted to $ns2" + + ip netns exec "$ns0" nft flush chain $family nat output + + reset_counters + ip netns exec "$ns0" ping -q -c 1 10.0.1.99 > /dev/null + if [ $? -ne 0 ]; then + lret=1 + echo "ERROR: ping failed" + return $lret + fi + + expect="packets 1 bytes 84" + for dir in "in" "out" ; do + cnt=$(ip netns exec "$ns0" nft list counter inet filter ns1${dir} | grep -q "$expect") + if [ $? -ne 0 ]; then + bad_counter "$ns1" ns1$dir "$expect" "test_local_dnat 5" + lret=1 + fi + done + expect="packets 0 bytes 0" + for dir in "in" "out" ; do + cnt=$(ip netns exec "$ns0" nft list counter inet filter ns2${dir} | grep -q "$expect") + if [ $? -ne 0 ]; then + bad_counter "$ns0" ns2$dir "$expect" "test_local_dnat 6" + lret=1 + fi + done + + # expect 1 count in ns1 + expect="packets 1 bytes 84" + for dir in "in" "out" ; do + cnt=$(ip netns exec "$ns1" nft list counter inet filter ns0${dir} | grep -q "$expect") + if [ $? -ne 0 ]; then + bad_counter "$ns0" ns0$dir "$expect" "test_local_dnat 7" + lret=1 + fi + done + + # expect 0 packet in ns2 + expect="packets 0 bytes 0" + for dir in "in" "out" ; do + cnt=$(ip netns exec "$ns2" nft list counter inet filter ns0${dir} | grep -q "$expect") + if [ $? -ne 0 ]; then + bad_counter "$ns2" ns0$dir "$expect" "test_local_dnat 8" + lret=1 + fi + done + + test $lret -eq 0 && echo "PASS: ping to $ns1 OK after $family nat output chain flush" + + return $lret +} + +test_local_dnat_portonly() +{ + local family=$1 + local daddr=$2 + local lret=0 + local sr_s + local sr_r + +ip netns exec "$ns0" nft -f /dev/stdin <<EOF +table $family nat { + chain output { + type nat hook output priority 0; policy accept; + meta l4proto tcp dnat to :2000 + + } +} +EOF + if [ $? -ne 0 ]; then + if [ $family = "inet" ];then + echo "SKIP: inet port test" + test_inet_nat=false + return + fi + echo "SKIP: Could not add $family dnat hook" + return + fi + + echo SERVER-$family | ip netns exec "$ns1" timeout 5 socat -u STDIN TCP-LISTEN:2000 & + sc_s=$! + + sleep 1 + + result=$(ip netns exec "$ns0" timeout 1 socat TCP:$daddr:2000 STDOUT) + + if [ "$result" = "SERVER-inet" ];then + echo "PASS: inet port rewrite without l3 address" + else + echo "ERROR: inet port rewrite" + ret=1 + fi +} + +test_masquerade6() +{ + local family=$1 + local natflags=$2 + local lret=0 + + ip netns exec "$ns0" sysctl net.ipv6.conf.all.forwarding=1 > /dev/null + + ip netns exec "$ns2" ping -q -c 1 dead:1::99 > /dev/null # ping ns2->ns1 + if [ $? -ne 0 ] ; then + echo "ERROR: cannot ping $ns1 from $ns2 via ipv6" + return 1 + lret=1 + fi + + expect="packets 1 bytes 104" + for dir in "in6" "out6" ; do + cnt=$(ip netns exec "$ns1" nft list counter inet filter ns2${dir} | grep -q "$expect") + if [ $? -ne 0 ]; then + bad_counter "$ns1" ns2$dir "$expect" "test_masquerade6 1" + lret=1 + fi + + cnt=$(ip netns exec "$ns2" nft list counter inet filter ns1${dir} | grep -q "$expect") + if [ $? -ne 0 ]; then + bad_counter "$ns2" ns1$dir "$expect" "test_masquerade6 2" + lret=1 + fi + done + + reset_counters + +# add masquerading rule +ip netns exec "$ns0" nft -f /dev/stdin <<EOF +table $family nat { + chain postrouting { + type nat hook postrouting priority 0; policy accept; + meta oif veth0 masquerade $natflags + } +} +EOF + if [ $? -ne 0 ]; then + echo "SKIP: Could not add add $family masquerade hook" + return $ksft_skip + fi + + ip netns exec "$ns2" ping -q -c 1 dead:1::99 > /dev/null # ping ns2->ns1 + if [ $? -ne 0 ] ; then + echo "ERROR: cannot ping $ns1 from $ns2 with active $family masquerade $natflags" + lret=1 + fi + + # ns1 should have seen packets from ns0, due to masquerade + expect="packets 1 bytes 104" + for dir in "in6" "out6" ; do + cnt=$(ip netns exec "$ns1" nft list counter inet filter ns0${dir} | grep -q "$expect") + if [ $? -ne 0 ]; then + bad_counter "$ns1" ns0$dir "$expect" "test_masquerade6 3" + lret=1 + fi + + cnt=$(ip netns exec "$ns2" nft list counter inet filter ns1${dir} | grep -q "$expect") + if [ $? -ne 0 ]; then + bad_counter "$ns2" ns1$dir "$expect" "test_masquerade6 4" + lret=1 + fi + done + + # ns1 should not have seen packets from ns2, due to masquerade + expect="packets 0 bytes 0" + for dir in "in6" "out6" ; do + cnt=$(ip netns exec "$ns1" nft list counter inet filter ns2${dir} | grep -q "$expect") + if [ $? -ne 0 ]; then + bad_counter "$ns1" ns0$dir "$expect" "test_masquerade6 5" + lret=1 + fi + + cnt=$(ip netns exec "$ns0" nft list counter inet filter ns1${dir} | grep -q "$expect") + if [ $? -ne 0 ]; then + bad_counter "$ns0" ns1$dir "$expect" "test_masquerade6 6" + lret=1 + fi + done + + ip netns exec "$ns2" ping -q -c 1 dead:1::99 > /dev/null # ping ns2->ns1 + if [ $? -ne 0 ] ; then + echo "ERROR: cannot ping $ns1 from $ns2 with active ipv6 masquerade $natflags (attempt 2)" + lret=1 + fi + + ip netns exec "$ns0" nft flush chain $family nat postrouting + if [ $? -ne 0 ]; then + echo "ERROR: Could not flush $family nat postrouting" 1>&2 + lret=1 + fi + + test $lret -eq 0 && echo "PASS: $family IPv6 masquerade $natflags for $ns2" + + return $lret +} + +test_masquerade() +{ + local family=$1 + local natflags=$2 + local lret=0 + + ip netns exec "$ns0" sysctl net.ipv4.conf.veth0.forwarding=1 > /dev/null + ip netns exec "$ns0" sysctl net.ipv4.conf.veth1.forwarding=1 > /dev/null + + ip netns exec "$ns2" ping -q -c 1 10.0.1.99 > /dev/null # ping ns2->ns1 + if [ $? -ne 0 ] ; then + echo "ERROR: cannot ping $ns1 from "$ns2" $natflags" + lret=1 + fi + + expect="packets 1 bytes 84" + for dir in "in" "out" ; do + cnt=$(ip netns exec "$ns1" nft list counter inet filter ns2${dir} | grep -q "$expect") + if [ $? -ne 0 ]; then + bad_counter "$ns1" ns2$dir "$expect" "test_masquerade 1" + lret=1 + fi + + cnt=$(ip netns exec "$ns2" nft list counter inet filter ns1${dir} | grep -q "$expect") + if [ $? -ne 0 ]; then + bad_counter "$ns2" ns1$dir "$expect" "test_masquerade 2" + lret=1 + fi + done + + reset_counters + +# add masquerading rule +ip netns exec "$ns0" nft -f /dev/stdin <<EOF +table $family nat { + chain postrouting { + type nat hook postrouting priority 0; policy accept; + meta oif veth0 masquerade $natflags + } +} +EOF + if [ $? -ne 0 ]; then + echo "SKIP: Could not add add $family masquerade hook" + return $ksft_skip + fi + + ip netns exec "$ns2" ping -q -c 1 10.0.1.99 > /dev/null # ping ns2->ns1 + if [ $? -ne 0 ] ; then + echo "ERROR: cannot ping $ns1 from $ns2 with active $family masquerade $natflags" + lret=1 + fi + + # ns1 should have seen packets from ns0, due to masquerade + expect="packets 1 bytes 84" + for dir in "in" "out" ; do + cnt=$(ip netns exec "$ns1" nft list counter inet filter ns0${dir} | grep -q "$expect") + if [ $? -ne 0 ]; then + bad_counter "$ns1" ns0$dir "$expect" "test_masquerade 3" + lret=1 + fi + + cnt=$(ip netns exec "$ns2" nft list counter inet filter ns1${dir} | grep -q "$expect") + if [ $? -ne 0 ]; then + bad_counter "$ns2" ns1$dir "$expect" "test_masquerade 4" + lret=1 + fi + done + + # ns1 should not have seen packets from ns2, due to masquerade + expect="packets 0 bytes 0" + for dir in "in" "out" ; do + cnt=$(ip netns exec "$ns1" nft list counter inet filter ns2${dir} | grep -q "$expect") + if [ $? -ne 0 ]; then + bad_counter "$ns1" ns0$dir "$expect" "test_masquerade 5" + lret=1 + fi + + cnt=$(ip netns exec "$ns0" nft list counter inet filter ns1${dir} | grep -q "$expect") + if [ $? -ne 0 ]; then + bad_counter "$ns0" ns1$dir "$expect" "test_masquerade 6" + lret=1 + fi + done + + ip netns exec "$ns2" ping -q -c 1 10.0.1.99 > /dev/null # ping ns2->ns1 + if [ $? -ne 0 ] ; then + echo "ERROR: cannot ping $ns1 from $ns2 with active ip masquerade $natflags (attempt 2)" + lret=1 + fi + + ip netns exec "$ns0" nft flush chain $family nat postrouting + if [ $? -ne 0 ]; then + echo "ERROR: Could not flush $family nat postrouting" 1>&2 + lret=1 + fi + + test $lret -eq 0 && echo "PASS: $family IP masquerade $natflags for $ns2" + + return $lret +} + +test_redirect6() +{ + local family=$1 + local lret=0 + + ip netns exec "$ns0" sysctl net.ipv6.conf.all.forwarding=1 > /dev/null + + ip netns exec "$ns2" ping -q -c 1 dead:1::99 > /dev/null # ping ns2->ns1 + if [ $? -ne 0 ] ; then + echo "ERROR: cannnot ping $ns1 from $ns2 via ipv6" + lret=1 + fi + + expect="packets 1 bytes 104" + for dir in "in6" "out6" ; do + cnt=$(ip netns exec "$ns1" nft list counter inet filter ns2${dir} | grep -q "$expect") + if [ $? -ne 0 ]; then + bad_counter "$ns1" ns2$dir "$expect" "test_redirect6 1" + lret=1 + fi + + cnt=$(ip netns exec "$ns2" nft list counter inet filter ns1${dir} | grep -q "$expect") + if [ $? -ne 0 ]; then + bad_counter "$ns2" ns1$dir "$expect" "test_redirect6 2" + lret=1 + fi + done + + reset_counters + +# add redirect rule +ip netns exec "$ns0" nft -f /dev/stdin <<EOF +table $family nat { + chain prerouting { + type nat hook prerouting priority 0; policy accept; + meta iif veth1 meta l4proto icmpv6 ip6 saddr dead:2::99 ip6 daddr dead:1::99 redirect + } +} +EOF + if [ $? -ne 0 ]; then + echo "SKIP: Could not add add $family redirect hook" + return $ksft_skip + fi + + ip netns exec "$ns2" ping -q -c 1 dead:1::99 > /dev/null # ping ns2->ns1 + if [ $? -ne 0 ] ; then + echo "ERROR: cannot ping $ns1 from $ns2 via ipv6 with active $family redirect" + lret=1 + fi + + # ns1 should have seen no packets from ns2, due to redirection + expect="packets 0 bytes 0" + for dir in "in6" "out6" ; do + cnt=$(ip netns exec "$ns1" nft list counter inet filter ns2${dir} | grep -q "$expect") + if [ $? -ne 0 ]; then + bad_counter "$ns1" ns0$dir "$expect" "test_redirect6 3" + lret=1 + fi + done + + # ns0 should have seen packets from ns2, due to masquerade + expect="packets 1 bytes 104" + for dir in "in6" "out6" ; do + cnt=$(ip netns exec "$ns0" nft list counter inet filter ns2${dir} | grep -q "$expect") + if [ $? -ne 0 ]; then + bad_counter "$ns1" ns0$dir "$expect" "test_redirect6 4" + lret=1 + fi + done + + ip netns exec "$ns0" nft delete table $family nat + if [ $? -ne 0 ]; then + echo "ERROR: Could not delete $family nat table" 1>&2 + lret=1 + fi + + test $lret -eq 0 && echo "PASS: $family IPv6 redirection for $ns2" + + return $lret +} + +test_redirect() +{ + local family=$1 + local lret=0 + + ip netns exec "$ns0" sysctl net.ipv4.conf.veth0.forwarding=1 > /dev/null + ip netns exec "$ns0" sysctl net.ipv4.conf.veth1.forwarding=1 > /dev/null + + ip netns exec "$ns2" ping -q -c 1 10.0.1.99 > /dev/null # ping ns2->ns1 + if [ $? -ne 0 ] ; then + echo "ERROR: cannot ping $ns1 from $ns2" + lret=1 + fi + + expect="packets 1 bytes 84" + for dir in "in" "out" ; do + cnt=$(ip netns exec "$ns1" nft list counter inet filter ns2${dir} | grep -q "$expect") + if [ $? -ne 0 ]; then + bad_counter "$ns1" $ns2$dir "$expect" "test_redirect 1" + lret=1 + fi + + cnt=$(ip netns exec "$ns2" nft list counter inet filter ns1${dir} | grep -q "$expect") + if [ $? -ne 0 ]; then + bad_counter "$ns2" ns1$dir "$expect" "test_redirect 2" + lret=1 + fi + done + + reset_counters + +# add redirect rule +ip netns exec "$ns0" nft -f /dev/stdin <<EOF +table $family nat { + chain prerouting { + type nat hook prerouting priority 0; policy accept; + meta iif veth1 ip protocol icmp ip saddr 10.0.2.99 ip daddr 10.0.1.99 redirect + } +} +EOF + if [ $? -ne 0 ]; then + echo "SKIP: Could not add add $family redirect hook" + return $ksft_skip + fi + + ip netns exec "$ns2" ping -q -c 1 10.0.1.99 > /dev/null # ping ns2->ns1 + if [ $? -ne 0 ] ; then + echo "ERROR: cannot ping $ns1 from $ns2 with active $family ip redirect" + lret=1 + fi + + # ns1 should have seen no packets from ns2, due to redirection + expect="packets 0 bytes 0" + for dir in "in" "out" ; do + + cnt=$(ip netns exec "$ns1" nft list counter inet filter ns2${dir} | grep -q "$expect") + if [ $? -ne 0 ]; then + bad_counter "$ns1" ns0$dir "$expect" "test_redirect 3" + lret=1 + fi + done + + # ns0 should have seen packets from ns2, due to masquerade + expect="packets 1 bytes 84" + for dir in "in" "out" ; do + cnt=$(ip netns exec "$ns0" nft list counter inet filter ns2${dir} | grep -q "$expect") + if [ $? -ne 0 ]; then + bad_counter "$ns0" ns0$dir "$expect" "test_redirect 4" + lret=1 + fi + done + + ip netns exec "$ns0" nft delete table $family nat + if [ $? -ne 0 ]; then + echo "ERROR: Could not delete $family nat table" 1>&2 + lret=1 + fi + + test $lret -eq 0 && echo "PASS: $family IP redirection for $ns2" + + return $lret +} + + +# ip netns exec "$ns0" ping -c 1 -q 10.0.$i.99 +for i in 0 1 2; do +ip netns exec ns$i-$sfx nft -f /dev/stdin <<EOF +table inet filter { + counter ns0in {} + counter ns1in {} + counter ns2in {} + + counter ns0out {} + counter ns1out {} + counter ns2out {} + + counter ns0in6 {} + counter ns1in6 {} + counter ns2in6 {} + + counter ns0out6 {} + counter ns1out6 {} + counter ns2out6 {} + + map nsincounter { + type ipv4_addr : counter + elements = { 10.0.1.1 : "ns0in", + 10.0.2.1 : "ns0in", + 10.0.1.99 : "ns1in", + 10.0.2.99 : "ns2in" } + } + + map nsincounter6 { + type ipv6_addr : counter + elements = { dead:1::1 : "ns0in6", + dead:2::1 : "ns0in6", + dead:1::99 : "ns1in6", + dead:2::99 : "ns2in6" } + } + + map nsoutcounter { + type ipv4_addr : counter + elements = { 10.0.1.1 : "ns0out", + 10.0.2.1 : "ns0out", + 10.0.1.99: "ns1out", + 10.0.2.99: "ns2out" } + } + + map nsoutcounter6 { + type ipv6_addr : counter + elements = { dead:1::1 : "ns0out6", + dead:2::1 : "ns0out6", + dead:1::99 : "ns1out6", + dead:2::99 : "ns2out6" } + } + + chain input { + type filter hook input priority 0; policy accept; + counter name ip saddr map @nsincounter + icmpv6 type { "echo-request", "echo-reply" } counter name ip6 saddr map @nsincounter6 + } + chain output { + type filter hook output priority 0; policy accept; + counter name ip daddr map @nsoutcounter + icmpv6 type { "echo-request", "echo-reply" } counter name ip6 daddr map @nsoutcounter6 + } +} +EOF +done + +sleep 3 +# test basic connectivity +for i in 1 2; do + ip netns exec "$ns0" ping -c 1 -q 10.0.$i.99 > /dev/null + if [ $? -ne 0 ];then + echo "ERROR: Could not reach other namespace(s)" 1>&2 + ret=1 + fi + + ip netns exec "$ns0" ping -c 1 -q dead:$i::99 > /dev/null + if [ $? -ne 0 ];then + echo "ERROR: Could not reach other namespace(s) via ipv6" 1>&2 + ret=1 + fi + check_counters ns$i-$sfx + if [ $? -ne 0 ]; then + ret=1 + fi + + check_ns0_counters ns$i + if [ $? -ne 0 ]; then + ret=1 + fi + reset_counters +done + +if [ $ret -eq 0 ];then + echo "PASS: netns routing/connectivity: $ns0 can reach $ns1 and $ns2" +fi + +reset_counters +test_local_dnat ip +test_local_dnat6 ip6 + +reset_counters +test_local_dnat_portonly inet 10.0.1.99 + +reset_counters +$test_inet_nat && test_local_dnat inet +$test_inet_nat && test_local_dnat6 inet + +for flags in "" "fully-random"; do +reset_counters +test_masquerade ip $flags +test_masquerade6 ip6 $flags +reset_counters +$test_inet_nat && test_masquerade inet $flags +$test_inet_nat && test_masquerade6 inet $flags +done + +reset_counters +test_redirect ip +test_redirect6 ip6 +reset_counters +$test_inet_nat && test_redirect inet +$test_inet_nat && test_redirect6 inet + +if [ $ret -ne 0 ];then + echo -n "FAIL: " + nft --version +fi + +exit $ret diff --git a/tools/testing/selftests/netfilter/nft_queue.sh b/tools/testing/selftests/netfilter/nft_queue.sh new file mode 100755 index 000000000..3d202b90b --- /dev/null +++ b/tools/testing/selftests/netfilter/nft_queue.sh @@ -0,0 +1,376 @@ +#!/bin/bash +# +# This tests nf_queue: +# 1. can process packets from all hooks +# 2. support running nfqueue from more than one base chain +# +# Kselftest framework requirement - SKIP code is 4. +ksft_skip=4 +ret=0 + +sfx=$(mktemp -u "XXXXXXXX") +ns1="ns1-$sfx" +ns2="ns2-$sfx" +nsrouter="nsrouter-$sfx" +timeout=4 + +cleanup() +{ + ip netns del ${ns1} + ip netns del ${ns2} + ip netns del ${nsrouter} + rm -f "$TMPFILE0" + rm -f "$TMPFILE1" + rm -f "$TMPFILE2" "$TMPFILE3" +} + +nft --version > /dev/null 2>&1 +if [ $? -ne 0 ];then + echo "SKIP: Could not run test without nft tool" + exit $ksft_skip +fi + +ip -Version > /dev/null 2>&1 +if [ $? -ne 0 ];then + echo "SKIP: Could not run test without ip tool" + exit $ksft_skip +fi + +ip netns add ${nsrouter} +if [ $? -ne 0 ];then + echo "SKIP: Could not create net namespace" + exit $ksft_skip +fi + +TMPFILE0=$(mktemp) +TMPFILE1=$(mktemp) +TMPFILE2=$(mktemp) +TMPFILE3=$(mktemp) +trap cleanup EXIT + +ip netns add ${ns1} +ip netns add ${ns2} + +ip link add veth0 netns ${nsrouter} type veth peer name eth0 netns ${ns1} > /dev/null 2>&1 +if [ $? -ne 0 ];then + echo "SKIP: No virtual ethernet pair device support in kernel" + exit $ksft_skip +fi +ip link add veth1 netns ${nsrouter} type veth peer name eth0 netns ${ns2} + +ip -net ${nsrouter} link set lo up +ip -net ${nsrouter} link set veth0 up +ip -net ${nsrouter} addr add 10.0.1.1/24 dev veth0 +ip -net ${nsrouter} addr add dead:1::1/64 dev veth0 + +ip -net ${nsrouter} link set veth1 up +ip -net ${nsrouter} addr add 10.0.2.1/24 dev veth1 +ip -net ${nsrouter} addr add dead:2::1/64 dev veth1 + +ip -net ${ns1} link set lo up +ip -net ${ns1} link set eth0 up + +ip -net ${ns2} link set lo up +ip -net ${ns2} link set eth0 up + +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} route add default via 10.0.1.1 +ip -net ${ns1} route add default via dead:1::1 + +ip -net ${ns2} addr add 10.0.2.99/24 dev eth0 +ip -net ${ns2} addr add dead:2::99/64 dev eth0 +ip -net ${ns2} route add default via 10.0.2.1 +ip -net ${ns2} route add default via dead:2::1 + +load_ruleset() { + local name=$1 + local prio=$2 + +ip netns exec ${nsrouter} nft -f /dev/stdin <<EOF +table inet $name { + chain nfq { + ip protocol icmp queue bypass + icmpv6 type { "echo-request", "echo-reply" } queue num 1 bypass + } + chain pre { + type filter hook prerouting priority $prio; policy accept; + jump nfq + } + chain input { + type filter hook input priority $prio; policy accept; + jump nfq + } + chain forward { + type filter hook forward priority $prio; policy accept; + tcp dport 12345 queue num 2 + jump nfq + } + chain output { + type filter hook output priority $prio; policy accept; + tcp dport 12345 queue num 3 + jump nfq + } + chain post { + type filter hook postrouting priority $prio; policy accept; + jump nfq + } +} +EOF +} + +load_counter_ruleset() { + local prio=$1 + +ip netns exec ${nsrouter} nft -f /dev/stdin <<EOF +table inet countrules { + chain pre { + type filter hook prerouting priority $prio; policy accept; + counter + } + chain input { + type filter hook input priority $prio; policy accept; + counter + } + chain forward { + type filter hook forward priority $prio; policy accept; + counter + } + chain output { + type filter hook output priority $prio; policy accept; + counter + } + chain post { + type filter hook postrouting priority $prio; policy accept; + counter + } +} +EOF +} + +test_ping() { + ip netns exec ${ns1} ping -c 1 -q 10.0.2.99 > /dev/null + if [ $? -ne 0 ];then + return 1 + fi + + ip netns exec ${ns1} ping -c 1 -q dead:2::99 > /dev/null + if [ $? -ne 0 ];then + return 1 + fi + + return 0 +} + +test_ping_router() { + ip netns exec ${ns1} ping -c 1 -q 10.0.2.1 > /dev/null + if [ $? -ne 0 ];then + return 1 + fi + + ip netns exec ${ns1} ping -c 1 -q dead:2::1 > /dev/null + if [ $? -ne 0 ];then + return 1 + fi + + return 0 +} + +test_queue_blackhole() { + local proto=$1 + +ip netns exec ${nsrouter} nft -f /dev/stdin <<EOF +table $proto blackh { + chain forward { + type filter hook forward priority 0; policy accept; + queue num 600 + } +} +EOF + if [ $proto = "ip" ] ;then + ip netns exec ${ns1} ping -W 2 -c 1 -q 10.0.2.99 > /dev/null + lret=$? + elif [ $proto = "ip6" ]; then + ip netns exec ${ns1} ping -W 2 -c 1 -q dead:2::99 > /dev/null + lret=$? + else + lret=111 + fi + + # queue without bypass keyword should drop traffic if no listener exists. + if [ $lret -eq 0 ];then + echo "FAIL: $proto expected failure, got $lret" 1>&2 + exit 1 + fi + + ip netns exec ${nsrouter} nft delete table $proto blackh + if [ $? -ne 0 ] ;then + echo "FAIL: $proto: Could not delete blackh table" + exit 1 + fi + + echo "PASS: $proto: statement with no listener results in packet drop" +} + +test_queue() +{ + local expected=$1 + local last="" + + # spawn nf-queue listeners + ip netns exec ${nsrouter} ./nf-queue -c -q 0 -t $timeout > "$TMPFILE0" & + ip netns exec ${nsrouter} ./nf-queue -c -q 1 -t $timeout > "$TMPFILE1" & + sleep 1 + test_ping + ret=$? + if [ $ret -ne 0 ];then + echo "FAIL: netns routing/connectivity with active listener on queue $queue: $ret" 1>&2 + exit $ret + fi + + test_ping_router + ret=$? + if [ $ret -ne 0 ];then + echo "FAIL: netns router unreachable listener on queue $queue: $ret" 1>&2 + exit $ret + fi + + wait + ret=$? + + for file in $TMPFILE0 $TMPFILE1; do + last=$(tail -n1 "$file") + if [ x"$last" != x"$expected packets total" ]; then + echo "FAIL: Expected $expected packets total, but got $last" 1>&2 + cat "$file" 1>&2 + + ip netns exec ${nsrouter} nft list ruleset + exit 1 + fi + done + + echo "PASS: Expected and received $last" +} + +test_tcp_forward() +{ + ip netns exec ${nsrouter} ./nf-queue -q 2 -t $timeout & + local nfqpid=$! + + tmpfile=$(mktemp) || exit 1 + dd conv=sparse status=none if=/dev/zero bs=1M count=200 of=$tmpfile + ip netns exec ${ns2} nc -w 5 -l -p 12345 <"$tmpfile" >/dev/null & + local rpid=$! + + sleep 1 + ip netns exec ${ns1} nc -w 5 10.0.2.99 12345 <"$tmpfile" >/dev/null & + + rm -f "$tmpfile" + + wait $rpid + wait $lpid + [ $? -eq 0 ] && echo "PASS: tcp and nfqueue in forward chain" +} + +test_tcp_localhost() +{ + tmpfile=$(mktemp) || exit 1 + + dd conv=sparse status=none if=/dev/zero bs=1M count=200 of=$tmpfile + ip netns exec ${nsrouter} nc -w 5 -l -p 12345 <"$tmpfile" >/dev/null & + local rpid=$! + + ip netns exec ${nsrouter} ./nf-queue -q 3 -t $timeout & + local nfqpid=$! + + sleep 1 + ip netns exec ${nsrouter} nc -w 5 127.0.0.1 12345 <"$tmpfile" > /dev/null + rm -f "$tmpfile" + + wait $rpid + [ $? -eq 0 ] && echo "PASS: tcp via loopback" + wait 2>/dev/null +} + +test_tcp_localhost_requeue() +{ +ip netns exec ${nsrouter} nft -f /dev/stdin <<EOF +flush ruleset +table inet filter { + chain output { + type filter hook output priority 0; policy accept; + tcp dport 12345 limit rate 1/second burst 1 packets counter queue num 0 + } + chain post { + type filter hook postrouting priority 0; policy accept; + tcp dport 12345 limit rate 1/second burst 1 packets counter queue num 0 + } +} +EOF + tmpfile=$(mktemp) || exit 1 + dd conv=sparse status=none if=/dev/zero bs=1M count=200 of=$tmpfile + ip netns exec ${nsrouter} nc -w 5 -l -p 12345 <"$tmpfile" >/dev/null & + local rpid=$! + + ip netns exec ${nsrouter} ./nf-queue -c -q 1 -t $timeout > "$TMPFILE2" & + + # nfqueue 1 will be called via output hook. But this time, + # re-queue the packet to nfqueue program on queue 2. + ip netns exec ${nsrouter} ./nf-queue -G -d 150 -c -q 0 -Q 1 -t $timeout > "$TMPFILE3" & + + sleep 1 + ip netns exec ${nsrouter} nc -w 5 127.0.0.1 12345 <"$tmpfile" > /dev/null + rm -f "$tmpfile" + + wait + + if ! diff -u "$TMPFILE2" "$TMPFILE3" ; then + echo "FAIL: lost packets during requeue?!" 1>&2 + return + fi + + echo "PASS: tcp via loopback and re-queueing" +} + +ip netns exec ${nsrouter} sysctl net.ipv6.conf.all.forwarding=1 > /dev/null +ip netns exec ${nsrouter} sysctl net.ipv4.conf.veth0.forwarding=1 > /dev/null +ip netns exec ${nsrouter} sysctl net.ipv4.conf.veth1.forwarding=1 > /dev/null + +load_ruleset "filter" 0 + +sleep 3 + +test_ping +ret=$? +if [ $ret -eq 0 ];then + # queue bypass works (rules were skipped, no listener) + echo "PASS: ${ns1} can reach ${ns2}" +else + echo "FAIL: ${ns1} cannot reach ${ns2}: $ret" 1>&2 + exit $ret +fi + +test_queue_blackhole ip +test_queue_blackhole ip6 + +# dummy ruleset to add base chains between the +# queueing rules. We don't want the second reinject +# to re-execute the old hooks. +load_counter_ruleset 10 + +# we are hooking all: prerouting/input/forward/output/postrouting. +# we ping ${ns2} from ${ns1} via ${nsrouter} using ipv4 and ipv6, so: +# 1x icmp prerouting,forward,postrouting -> 3 queue events (6 incl. reply). +# 1x icmp prerouting,input,output postrouting -> 4 queue events incl. reply. +# so we expect that userspace program receives 10 packets. +test_queue 10 + +# same. We queue to a second program as well. +load_ruleset "filter2" 20 +test_queue 20 + +test_tcp_forward +test_tcp_localhost +test_tcp_localhost_requeue + +exit $ret diff --git a/tools/testing/selftests/netfilter/nft_trans_stress.sh b/tools/testing/selftests/netfilter/nft_trans_stress.sh new file mode 100755 index 000000000..f1affd12c --- /dev/null +++ b/tools/testing/selftests/netfilter/nft_trans_stress.sh @@ -0,0 +1,78 @@ +#!/bin/bash +# +# This test is for stress-testing the nf_tables config plane path vs. +# packet path processing: Make sure we never release rules that are +# still visible to other cpus. +# +# set -e + +# Kselftest framework requirement - SKIP code is 4. +ksft_skip=4 + +testns=testns1 +tables="foo bar baz quux" + +nft --version > /dev/null 2>&1 +if [ $? -ne 0 ];then + echo "SKIP: Could not run test without nft tool" + exit $ksft_skip +fi + +ip -Version > /dev/null 2>&1 +if [ $? -ne 0 ];then + echo "SKIP: Could not run test without ip tool" + exit $ksft_skip +fi + +tmp=$(mktemp) + +for table in $tables; do + echo add table inet "$table" >> "$tmp" + echo flush table inet "$table" >> "$tmp" + + echo "add chain inet $table INPUT { type filter hook input priority 0; }" >> "$tmp" + echo "add chain inet $table OUTPUT { type filter hook output priority 0; }" >> "$tmp" + for c in $(seq 1 400); do + chain=$(printf "chain%03u" "$c") + echo "add chain inet $table $chain" >> "$tmp" + done + + for c in $(seq 1 400); do + chain=$(printf "chain%03u" "$c") + for BASE in INPUT OUTPUT; do + echo "add rule inet $table $BASE counter jump $chain" >> "$tmp" + done + echo "add rule inet $table $chain counter return" >> "$tmp" + done +done + +ip netns add "$testns" +ip -netns "$testns" link set lo up + +lscpu | grep ^CPU\(s\): | ( read cpu cpunum ; +cpunum=$((cpunum-1)) +for i in $(seq 0 $cpunum);do + mask=$(printf 0x%x $((1<<$i))) + ip netns exec "$testns" taskset $mask ping -4 127.0.0.1 -fq > /dev/null & + ip netns exec "$testns" taskset $mask ping -6 ::1 -fq > /dev/null & +done) + +sleep 1 + +for i in $(seq 1 10) ; do ip netns exec "$testns" nft -f "$tmp" & done + +for table in $tables;do + randsleep=$((RANDOM%10)) + sleep $randsleep + ip netns exec "$testns" nft delete table inet $table 2>/dev/null +done + +randsleep=$((RANDOM%10)) +sleep $randsleep + +pkill -9 ping + +wait + +rm -f "$tmp" +ip netns del "$testns" |