As recently specified by commit 0ea09cbf8350 ("docs: netdev: add a note on selftest posting") in net-next, the selftest is therefore shipped in this series. However, this selftest does not really test this series. It needs this series to avoid crashing the kernel. What it really tests, thanks to kmemleak, is what was fixed by the following commits: - commit c71a192976de ("net: ipv6: fix dst refleaks in rpl, seg6 and ioam6 lwtunnels") - commit 92191dd10730 ("net: ipv6: fix dst ref loops in rpl, seg6 and ioam6 lwtunnels") - commit c64a0727f9b1 ("net: ipv6: fix dst ref loop on input in seg6 lwt") - commit 13e55fbaec17 ("net: ipv6: fix dst ref loop on input in rpl lwt") - commit 0e7633d7b95b ("net: ipv6: fix dst ref loop in ila lwtunnel") - commit 5da15a9c11c1 ("net: ipv6: fix missing dst ref drop in ila lwtunnel")
Signed-off-by: Justin Iurman justin.iurman@uliege.be --- Cc: Shuah Khan shuah@kernel.org Cc: linux-kselftest@vger.kernel.org --- tools/testing/selftests/net/Makefile | 1 + tools/testing/selftests/net/config | 2 + .../selftests/net/lwt_dst_cache_ref_loop.sh | 246 ++++++++++++++++++ 3 files changed, 249 insertions(+) create mode 100755 tools/testing/selftests/net/lwt_dst_cache_ref_loop.sh
diff --git a/tools/testing/selftests/net/Makefile b/tools/testing/selftests/net/Makefile index 5916f3b81c39..843ab747645d 100644 --- a/tools/testing/selftests/net/Makefile +++ b/tools/testing/selftests/net/Makefile @@ -101,6 +101,7 @@ TEST_PROGS += vlan_bridge_binding.sh TEST_PROGS += bpf_offload.py TEST_PROGS += ipv6_route_update_soft_lockup.sh TEST_PROGS += busy_poll_test.sh +TEST_PROGS += lwt_dst_cache_ref_loop.sh
# YNL files, must be before "include ..lib.mk" YNL_GEN_FILES := busy_poller netlink-dumps diff --git a/tools/testing/selftests/net/config b/tools/testing/selftests/net/config index 5b9baf708950..61e5116987f3 100644 --- a/tools/testing/selftests/net/config +++ b/tools/testing/selftests/net/config @@ -107,3 +107,5 @@ CONFIG_XFRM_INTERFACE=m CONFIG_XFRM_USER=m CONFIG_IP_NF_MATCH_RPFILTER=m CONFIG_IP6_NF_MATCH_RPFILTER=m +CONFIG_IPV6_ILA=m +CONFIG_IPV6_RPL_LWTUNNEL=y diff --git a/tools/testing/selftests/net/lwt_dst_cache_ref_loop.sh b/tools/testing/selftests/net/lwt_dst_cache_ref_loop.sh new file mode 100755 index 000000000000..881eb399798f --- /dev/null +++ b/tools/testing/selftests/net/lwt_dst_cache_ref_loop.sh @@ -0,0 +1,246 @@ +#!/bin/bash +# SPDX-License-Identifier: GPL-2.0+ +# +# Author: Justin Iurman justin.iurman@uliege.be +# +# WARNING +# ------- +# This is just a dummy script that triggers encap cases with possible dst cache +# reference loops in affected lwt users (see list below). Some cases are +# pathological configurations for simplicity, others are valid. Overall, we +# don't want this issue to happen, no matter what. In order to catch any +# reference loops, kmemleak MUST be used. The results alone are always blindly +# successful, don't rely on them. Note that the following tests may crash the +# kernel if the fix to prevent lwtunnel_{input|output|xmit}() reentry loops is +# not present. +# +# Affected lwt users so far (please update accordingly if needed): +# - ila_lwt (output only) +# - ioam6_iptunnel (output only) +# - rpl_iptunnel (both input and output) +# - seg6_iptunnel (both input and output) + +source lib.sh + +check_compatibility() +{ + setup_ns tmp_node &>/dev/null + if [ $? != 0 ]; then + echo "SKIP: Cannot create netns." + exit $ksft_skip + fi + + ip link add name veth0 netns $tmp_node type veth \ + peer name veth1 netns $tmp_node &>/dev/null + local ret=$? + + ip -netns $tmp_node link set veth0 up &>/dev/null + ret=$((ret + $?)) + + ip -netns $tmp_node link set veth1 up &>/dev/null + ret=$((ret + $?)) + + if [ $ret != 0 ]; then + echo "SKIP: Cannot configure links." + cleanup_ns $tmp_node + exit $ksft_skip + fi + + lsmod 2>/dev/null | grep -q "ila" + ila_lsmod=$? + [ $ila_lsmod != 0 ] && modprobe ila &>/dev/null + + ip -netns $tmp_node route add 2001:db8:1::/64 \ + encap ila 1:2:3:4 csum-mode no-action ident-type luid \ + hook-type output \ + dev veth0 &>/dev/null + + ip -netns $tmp_node route add 2001:db8:2::/64 \ + encap ioam6 trace prealloc type 0x800000 ns 0 size 4 \ + dev veth0 &>/dev/null + + ip -netns $tmp_node route add 2001:db8:3::/64 \ + encap rpl segs 2001:db8:3::1 dev veth0 &>/dev/null + + ip -netns $tmp_node route add 2001:db8:4::/64 \ + encap seg6 mode inline segs 2001:db8:4::1 dev veth0 &>/dev/null + + ip -netns $tmp_node -6 route 2>/dev/null | grep -q "encap ila" + skip_ila=$? + + ip -netns $tmp_node -6 route 2>/dev/null | grep -q "encap ioam6" + skip_ioam6=$? + + ip -netns $tmp_node -6 route 2>/dev/null | grep -q "encap rpl" + skip_rpl=$? + + ip -netns $tmp_node -6 route 2>/dev/null | grep -q "encap seg6" + skip_seg6=$? + + cleanup_ns $tmp_node +} + +setup() +{ + setup_ns alpha beta gamma &>/dev/null + + ip link add name veth-alpha netns $alpha type veth \ + peer name veth-betaL netns $beta &>/dev/null + + ip link add name veth-betaR netns $beta type veth \ + peer name veth-gamma netns $gamma &>/dev/null + + ip -netns $alpha link set veth-alpha name veth0 &>/dev/null + ip -netns $beta link set veth-betaL name veth0 &>/dev/null + ip -netns $beta link set veth-betaR name veth1 &>/dev/null + ip -netns $gamma link set veth-gamma name veth0 &>/dev/null + + ip -netns $alpha addr add 2001:db8:1::2/64 dev veth0 &>/dev/null + ip -netns $alpha link set veth0 up &>/dev/null + ip -netns $alpha link set lo up &>/dev/null + ip -netns $alpha route add 2001:db8:2::/64 \ + via 2001:db8:1::1 dev veth0 &>/dev/null + + ip -netns $beta addr add 2001:db8:1::1/64 dev veth0 &>/dev/null + ip -netns $beta addr add 2001:db8:2::1/64 dev veth1 &>/dev/null + ip -netns $beta link set veth0 up &>/dev/null + ip -netns $beta link set veth1 up &>/dev/null + ip -netns $beta link set lo up &>/dev/null + ip -netns $beta route del 2001:db8:2::/64 + ip -netns $beta route add 2001:db8:2::/64 dev veth1 + ip netns exec $beta \ + sysctl -wq net.ipv6.conf.all.forwarding=1 &>/dev/null + + ip -netns $gamma addr add 2001:db8:2::2/64 dev veth0 &>/dev/null + ip -netns $gamma link set veth0 up &>/dev/null + ip -netns $gamma link set lo up &>/dev/null + ip -netns $gamma route add 2001:db8:1::/64 \ + via 2001:db8:2::1 dev veth0 &>/dev/null + + sleep 1 + + ip netns exec $alpha ping6 -c 5 -W 1 2001:db8:2::2 &>/dev/null + if [ $? != 0 ]; then + echo "SKIP: Setup failed." + exit $ksft_skip + fi + + sleep 1 +} + +cleanup() +{ + cleanup_ns $alpha $beta $gamma + [ $ila_lsmod != 0 ] && modprobe -r ila &>/dev/null +} + +run_ila() +{ + if [ $skip_ila != 0 ]; then + echo "SKIP: ila (output)" + return + fi + + ip -netns $beta route del 2001:db8:2::/64 + ip -netns $beta route add 2001:db8:2:0:0:0:0:2/128 \ + encap ila 2001:db8:2:0 csum-mode no-action ident-type luid \ + hook-type output \ + dev veth1 &>/dev/null + sleep 1 + + echo "TEST: ila (output)" + ip netns exec $beta ping6 -c 2 -W 1 2001:db8:2::2 &>/dev/null + sleep 1 + + ip -netns $beta route del 2001:db8:2:0:0:0:0:2/128 + ip -netns $beta route add 2001:db8:2::/64 dev veth1 + sleep 1 +} + +run_ioam6() +{ + if [ $skip_ioam6 != 0 ]; then + echo "SKIP: ioam6 (output)" + return + fi + + ip -netns $beta route change 2001:db8:2::/64 \ + encap ioam6 trace prealloc type 0x800000 ns 1 size 4 \ + dev veth1 &>/dev/null + sleep 1 + + echo "TEST: ioam6 (output)" + ip netns exec $beta ping6 -c 2 -W 1 2001:db8:2::2 &>/dev/null + sleep 1 +} + +run_rpl() +{ + if [ $skip_rpl != 0 ]; then + echo "SKIP: rpl (input)" + echo "SKIP: rpl (output)" + return + fi + + ip -netns $beta route change 2001:db8:2::/64 \ + encap rpl segs 2001:db8:2::2 \ + dev veth1 &>/dev/null + sleep 1 + + echo "TEST: rpl (input)" + ip netns exec $alpha ping6 -c 2 -W 1 2001:db8:2::2 &>/dev/null + sleep 1 + + echo "TEST: rpl (output)" + ip netns exec $beta ping6 -c 2 -W 1 2001:db8:2::2 &>/dev/null + sleep 1 +} + +run_seg6() +{ + if [ $skip_seg6 != 0 ]; then + echo "SKIP: seg6 (input)" + echo "SKIP: seg6 (output)" + return + fi + + ip -netns $beta route change 2001:db8:2::/64 \ + encap seg6 mode inline segs 2001:db8:2::2 \ + dev veth1 &>/dev/null + sleep 1 + + echo "TEST: seg6 (input)" + ip netns exec $alpha ping6 -c 2 -W 1 2001:db8:2::2 &>/dev/null + sleep 1 + + echo "TEST: seg6 (output)" + ip netns exec $beta ping6 -c 2 -W 1 2001:db8:2::2 &>/dev/null + sleep 1 +} + +run() +{ + run_ila + run_ioam6 + run_rpl + run_seg6 +} + +if [ "$(id -u)" -ne 0 ]; then + echo "SKIP: Need root privileges." + exit $ksft_skip +fi + +if [ ! -x "$(command -v ip)" ]; then + echo "SKIP: Could not run test without ip tool." + exit $ksft_skip +fi + +check_compatibility + +trap cleanup EXIT + +setup +run + +exit $ksft_pass
On 3/14/25 1:00 PM, Justin Iurman wrote:
As recently specified by commit 0ea09cbf8350 ("docs: netdev: add a note on selftest posting") in net-next, the selftest is therefore shipped in this series. However, this selftest does not really test this series. It needs this series to avoid crashing the kernel. What it really tests, thanks to kmemleak,
As a net-next follow-up you could force a kmemleak scan and check the result after each test case to really output a pass/fail message.
Also, still for net-next, please investigate if dropping or reducing the many sleep below could be possible (it's not clear to me why they are needed).
I'll take is as-is to avoid blocking the fixes for trivial matters.
Thanks,
Paolo
On 3/20/25 11:25, Paolo Abeni wrote:
On 3/14/25 1:00 PM, Justin Iurman wrote:
As recently specified by commit 0ea09cbf8350 ("docs: netdev: add a note on selftest posting") in net-next, the selftest is therefore shipped in this series. However, this selftest does not really test this series. It needs this series to avoid crashing the kernel. What it really tests, thanks to kmemleak,
As a net-next follow-up you could force a kmemleak scan and check the result after each test case to really output a pass/fail message.
Sure. Could you point me to an example doing that? Not sure how I'd force it within the selftest.
Also, still for net-next, please investigate if dropping or reducing the many sleep below could be possible (it's not clear to me why they are needed).
Of course, will do.
I'll take is as-is to avoid blocking the fixes for trivial matters.
Thanks, Justin
Thanks,
Paolo
Hi Justin,
On 14/03/2025 13:00, Justin Iurman wrote:
As recently specified by commit 0ea09cbf8350 ("docs: netdev: add a note on selftest posting") in net-next, the selftest is therefore shipped in this series. However, this selftest does not really test this series. It needs this series to avoid crashing the kernel. What it really tests, thanks to kmemleak, is what was fixed by the following commits:
- commit c71a192976de ("net: ipv6: fix dst refleaks in rpl, seg6 and
ioam6 lwtunnels")
- commit 92191dd10730 ("net: ipv6: fix dst ref loops in rpl, seg6 and
ioam6 lwtunnels")
- commit c64a0727f9b1 ("net: ipv6: fix dst ref loop on input in seg6
lwt")
- commit 13e55fbaec17 ("net: ipv6: fix dst ref loop on input in rpl
lwt")
- commit 0e7633d7b95b ("net: ipv6: fix dst ref loop in ila lwtunnel")
- commit 5da15a9c11c1 ("net: ipv6: fix missing dst ref drop in ila
lwtunnel")
Signed-off-by: Justin Iurman justin.iurman@uliege.be
Cc: Shuah Khan shuah@kernel.org Cc: linux-kselftest@vger.kernel.org
tools/testing/selftests/net/Makefile | 1 + tools/testing/selftests/net/config | 2 + .../selftests/net/lwt_dst_cache_ref_loop.sh | 246 ++++++++++++++++++ 3 files changed, 249 insertions(+) create mode 100755 tools/testing/selftests/net/lwt_dst_cache_ref_loop.sh
diff --git a/tools/testing/selftests/net/Makefile b/tools/testing/selftests/net/Makefile index 5916f3b81c39..843ab747645d 100644 --- a/tools/testing/selftests/net/Makefile +++ b/tools/testing/selftests/net/Makefile @@ -101,6 +101,7 @@ TEST_PROGS += vlan_bridge_binding.sh TEST_PROGS += bpf_offload.py TEST_PROGS += ipv6_route_update_soft_lockup.sh TEST_PROGS += busy_poll_test.sh +TEST_PROGS += lwt_dst_cache_ref_loop.sh # YNL files, must be before "include ..lib.mk" YNL_GEN_FILES := busy_poller netlink-dumps diff --git a/tools/testing/selftests/net/config b/tools/testing/selftests/net/config index 5b9baf708950..61e5116987f3 100644 --- a/tools/testing/selftests/net/config +++ b/tools/testing/selftests/net/config @@ -107,3 +107,5 @@ CONFIG_XFRM_INTERFACE=m CONFIG_XFRM_USER=m CONFIG_IP_NF_MATCH_RPFILTER=m CONFIG_IP6_NF_MATCH_RPFILTER=m +CONFIG_IPV6_ILA=m +CONFIG_IPV6_RPL_LWTUNNEL=y
FYI, we got a small conflict when merging 'net' in 'net-next' in the MPTCP tree due to this patch applied in 'net':
3ed61b8938c6 ("selftests: net: test for lwtunnel dst ref loops")
and these ones from 'net-next':
03544faad761 ("selftest: net: add proc_net_pktgen") 85cb3711acb8 ("selftests: net: Add test cases for link and peer netns")
----- Generic Message ----- The best is to avoid conflicts between 'net' and 'net-next' trees but if they cannot be avoided when preparing patches, a note about how to fix them is much appreciated.
The conflict has been resolved on our side[1] and the resolution we suggest is attached to this email. Please report any issues linked to this conflict resolution as it might be used by others. If you worked on the mentioned patches, don't hesitate to ACK this conflict resolution. ---------------------------
Regarding these conflicts, that's because the new test and configs have been added at the end. It is usually recommended to avoid that for long list, e.g. inserting new entries at the right place if the list is sorted by alphabetical order (not the case here), grouping by theme, or keeping the same order as in the .config, etc. For the resolutions here, I simply kept the new tests and configs from both sides at the same place, adding the new ones from net at the end.
Rerere cache is available in [2].
Cheers, Matt
[1] https://github.com/multipath-tcp/mptcp_net-next/commit/89220b9164c7 [2] https://github.com/multipath-tcp/mptcp-upstream-rr-cache/commit/d240
Cheers, Matt
linux-kselftest-mirror@lists.linaro.org
-
Justin Iurman -
Matthieu Baerts -
Paolo Abeni