In the forwarding testcase, it opens a server and a client with the nc. The server receives the correct message from NC, it prints OK. The server prints FAIL if it receives the wrong message from the client.
But If the server can't receive any message, it will not close so the amt.sh waits forever. There are several reasons. 1. crash of smcrouted. 2. Send a message from the client to the server before the server is up.
To avoid this problem, the server waits only for 10 seconds. The client sends messages for 10 seconds. If the server is successfully closed, it kills the client.
Fixes: c08e8baea78e ("selftests: add amt interface selftest script") Signed-off-by: Taehee Yoo ap420073@gmail.com --- tools/testing/selftests/net/amt.sh | 63 +++++++++++++++++++----------- 1 file changed, 40 insertions(+), 23 deletions(-)
diff --git a/tools/testing/selftests/net/amt.sh b/tools/testing/selftests/net/amt.sh index 75528788cb95..16641d3dccce 100755 --- a/tools/testing/selftests/net/amt.sh +++ b/tools/testing/selftests/net/amt.sh @@ -77,6 +77,7 @@ readonly LISTENER=$(mktemp -u listener-XXXXXXXX) readonly GATEWAY=$(mktemp -u gateway-XXXXXXXX) readonly RELAY=$(mktemp -u relay-XXXXXXXX) readonly SOURCE=$(mktemp -u source-XXXXXXXX) +readonly RESULT=$(mktemp -p /tmp amt-XXXXXXXX) ERR=4 err=0
@@ -85,6 +86,10 @@ exit_cleanup() for ns in "$@"; do ip netns delete "${ns}" 2>/dev/null || true done + rm $RESULT + smcpid=$(< $SMCROUTEDIR/amt.pid) + kill $smcpid + rm -rf $SMCROUTEDIR
exit $ERR } @@ -167,7 +172,9 @@ setup_iptables()
setup_mcast_routing() { - ip netns exec "${RELAY}" smcrouted + SMCROUTEDIR="$(mktemp -d)" + + ip netns exec "${RELAY}" smcrouted -P $SMCROUTEDIR/amt.pid ip netns exec "${RELAY}" smcroutectl a relay_src \ 172.17.0.2 239.0.0.1 amtr ip netns exec "${RELAY}" smcroutectl a relay_src \ @@ -210,40 +217,52 @@ check_features()
test_ipv4_forward() { - RESULT4=$(ip netns exec "${LISTENER}" nc -w 1 -l -u 239.0.0.1 4000) + echo "" > $RESULT + bash -c "$(ip netns exec "${LISTENER}" \ + timeout 10s nc -w 1 -l -u 239.0.0.1 4000 > $RESULT)" + RESULT4=$(< $RESULT) if [ "$RESULT4" == "172.17.0.2" ]; then printf "TEST: %-60s [ OK ]\n" "IPv4 amt multicast forwarding" - exit 0 else printf "TEST: %-60s [FAIL]\n" "IPv4 amt multicast forwarding" - exit 1 fi + }
test_ipv6_forward() { - RESULT6=$(ip netns exec "${LISTENER}" nc -w 1 -l -u ff0e::5:6 6000) + echo "" > $RESULT + bash -c "$(ip netns exec "${LISTENER}" \ + timeout 10s nc -w 1 -l -u ff0e::5:6 6000 > $RESULT)" + RESULT6=$(< $RESULT) if [ "$RESULT6" == "2001:db8:3::2" ]; then printf "TEST: %-60s [ OK ]\n" "IPv6 amt multicast forwarding" - exit 0 else printf "TEST: %-60s [FAIL]\n" "IPv6 amt multicast forwarding" - exit 1 fi + }
send_mcast4() { sleep 2 - ip netns exec "${SOURCE}" bash -c \ - 'echo 172.17.0.2 | nc -w 1 -u 239.0.0.1 4000' & + for n in {0..10}; do + ip netns exec "${SOURCE}" bash -c \ + 'echo 172.17.0.2 | nc -w 1 -u 239.0.0.1 4000' + sleep 1 + done + }
send_mcast6() { sleep 2 - ip netns exec "${SOURCE}" bash -c \ - 'echo 2001:db8:3::2 | nc -w 1 -u ff0e::5:6 6000' & + for n in {0..10}; do + ip netns exec "${SOURCE}" bash -c \ + 'echo 2001:db8:3::2 | nc -w 1 -u ff0e::5:6 6000' + sleep 1 + done + }
check_features @@ -259,19 +278,17 @@ setup_iptables setup_mcast_routing test_remote_ip test_ipv4_forward & -pid=$! -send_mcast4 -wait $pid || err=$? -if [ $err -eq 1 ]; then - ERR=1 -fi +spid=$! +send_mcast4 & +cpid=$! +wait $spid +kill $cpid test_ipv6_forward & -pid=$! -send_mcast6 -wait $pid || err=$? -if [ $err -eq 1 ]; then - ERR=1 -fi +spid=$! +send_mcast6 & +cpid=$! +wait $spid +kill $cpid send_mcast_torture4 printf "TEST: %-60s [ OK ]\n" "IPv4 amt traffic forwarding torture" send_mcast_torture6
On Wed, May 08, 2024 at 04:06:43AM +0000, Taehee Yoo wrote:
In the forwarding testcase, it opens a server and a client with the nc. The server receives the correct message from NC, it prints OK. The server prints FAIL if it receives the wrong message from the client.
But If the server can't receive any message, it will not close so the amt.sh waits forever. There are several reasons.
- crash of smcrouted.
- Send a message from the client to the server before the server is up.
To avoid this problem, the server waits only for 10 seconds. The client sends messages for 10 seconds. If the server is successfully closed, it kills the client.
Fixes: c08e8baea78e ("selftests: add amt interface selftest script") Signed-off-by: Taehee Yoo ap420073@gmail.com
tools/testing/selftests/net/amt.sh | 63 +++++++++++++++++++----------- 1 file changed, 40 insertions(+), 23 deletions(-)
diff --git a/tools/testing/selftests/net/amt.sh b/tools/testing/selftests/net/amt.sh index 75528788cb95..16641d3dccce 100755 --- a/tools/testing/selftests/net/amt.sh +++ b/tools/testing/selftests/net/amt.sh @@ -77,6 +77,7 @@ readonly LISTENER=$(mktemp -u listener-XXXXXXXX) readonly GATEWAY=$(mktemp -u gateway-XXXXXXXX) readonly RELAY=$(mktemp -u relay-XXXXXXXX) readonly SOURCE=$(mktemp -u source-XXXXXXXX) +readonly RESULT=$(mktemp -p /tmp amt-XXXXXXXX) ERR=4 err=0 @@ -85,6 +86,10 @@ exit_cleanup() for ns in "$@"; do ip netns delete "${ns}" 2>/dev/null || true done
- rm $RESULT
- smcpid=$(< $SMCROUTEDIR/amt.pid)
- kill $smcpid
- rm -rf $SMCROUTEDIR
Hi Taehee Yoo,
I think this cleanup may be executed before SMCROUTEDIR exists.
For consistency with other temp files, perhaps perpahps it is best to move the creation of SMCROUTEDIR up to where RESULT is instantiated above.
And perhaps the pid handling can be made conditional on the existence of $SMCROUTEDIR/amt.pid
if [ -f "$SMCROUTEDIR/amt.pid" ]; then ... fi
exit $ERR } @@ -167,7 +172,9 @@ setup_iptables() setup_mcast_routing() {
- ip netns exec "${RELAY}" smcrouted
- SMCROUTEDIR="$(mktemp -d)"
- ip netns exec "${RELAY}" smcrouted -P $SMCROUTEDIR/amt.pid ip netns exec "${RELAY}" smcroutectl a relay_src \ 172.17.0.2 239.0.0.1 amtr ip netns exec "${RELAY}" smcroutectl a relay_src \
@@ -210,40 +217,52 @@ check_features() test_ipv4_forward() {
- RESULT4=$(ip netns exec "${LISTENER}" nc -w 1 -l -u 239.0.0.1 4000)
- echo "" > $RESULT
- bash -c "$(ip netns exec "${LISTENER}" \
timeout 10s nc -w 1 -l -u 239.0.0.1 4000 > $RESULT)"
Hi,
It's unclear to me what the purpose of the bash -c "$(...)" construction is here. Can the same be achieved using simply:
ip netns exec "${LISTENER}" \ timeout 10s nc -w 1 -l -u 239.0.0.1 4000 > $RESULT
Also, not strictly related to this patch, it seems a little odd here, and elsewhere, to call bash in a /bin/sh script.
- RESULT4=$(< $RESULT) if [ "$RESULT4" == "172.17.0.2" ]; then printf "TEST: %-60s [ OK ]\n" "IPv4 amt multicast forwarding"
exit 0exit 1
}
...
send_mcast4() { sleep 2
- ip netns exec "${SOURCE}" bash -c \
'echo 172.17.0.2 | nc -w 1 -u 239.0.0.1 4000' &
- for n in {0..10}; do
ip netns exec "${SOURCE}" bash -c \'echo 172.17.0.2 | nc -w 1 -u 239.0.0.1 4000'sleep 1- done
} send_mcast6() { sleep 2
- ip netns exec "${SOURCE}" bash -c \
'echo 2001:db8:3::2 | nc -w 1 -u ff0e::5:6 6000' &
- for n in {0..10}; do
ip netns exec "${SOURCE}" bash -c \'echo 2001:db8:3::2 | nc -w 1 -u ff0e::5:6 6000'sleep 1- done
} check_features
...
On Thu, May 9, 2024 at 5:38 PM Simon Horman horms@kernel.org wrote:
Hi Simon, Thanks a lot for the review!
On Wed, May 08, 2024 at 04:06:43AM +0000, Taehee Yoo wrote:
In the forwarding testcase, it opens a server and a client with the nc. The server receives the correct message from NC, it prints OK. The server prints FAIL if it receives the wrong message from the client.
But If the server can't receive any message, it will not close so the amt.sh waits forever. There are several reasons.
- crash of smcrouted.
- Send a message from the client to the server before the server is up.
To avoid this problem, the server waits only for 10 seconds. The client sends messages for 10 seconds. If the server is successfully closed, it kills the client.
Fixes: c08e8baea78e ("selftests: add amt interface selftest script") Signed-off-by: Taehee Yoo ap420073@gmail.com
tools/testing/selftests/net/amt.sh | 63 +++++++++++++++++++----------- 1 file changed, 40 insertions(+), 23 deletions(-)
diff --git a/tools/testing/selftests/net/amt.sh b/tools/testing/selftests/net/amt.sh index 75528788cb95..16641d3dccce 100755 --- a/tools/testing/selftests/net/amt.sh +++ b/tools/testing/selftests/net/amt.sh @@ -77,6 +77,7 @@ readonly LISTENER=$(mktemp -u listener-XXXXXXXX) readonly GATEWAY=$(mktemp -u gateway-XXXXXXXX) readonly RELAY=$(mktemp -u relay-XXXXXXXX) readonly SOURCE=$(mktemp -u source-XXXXXXXX) +readonly RESULT=$(mktemp -p /tmp amt-XXXXXXXX) ERR=4 err=0
@@ -85,6 +86,10 @@ exit_cleanup() for ns in "$@"; do ip netns delete "${ns}" 2>/dev/null || true done
rm $RESULTsmcpid=$(< $SMCROUTEDIR/amt.pid)kill $smcpidrm -rf $SMCROUTEDIRHi Taehee Yoo,
I think this cleanup may be executed before SMCROUTEDIR exists.
For consistency with other temp files, perhaps perpahps it is best to move the creation of SMCROUTEDIR up to where RESULT is instantiated above.
And perhaps the pid handling can be made conditional on the existence of $SMCROUTEDIR/amt.pid
if [ -f "$SMCROUTEDIR/amt.pid" ]; then ... fi
Thanks! I will check a pid file before kills smcrouted.
exit $ERR} @@ -167,7 +172,9 @@ setup_iptables()
setup_mcast_routing() {
ip netns exec "${RELAY}" smcrouted
SMCROUTEDIR="$(mktemp -d)"ip netns exec "${RELAY}" smcrouted -P $SMCROUTEDIR/amt.pid ip netns exec "${RELAY}" smcroutectl a relay_src \ 172.17.0.2 239.0.0.1 amtr ip netns exec "${RELAY}" smcroutectl a relay_src \@@ -210,40 +217,52 @@ check_features()
test_ipv4_forward() {
RESULT4=$(ip netns exec "${LISTENER}" nc -w 1 -l -u 239.0.0.1 4000)
echo "" > $RESULTbash -c "$(ip netns exec "${LISTENER}" \timeout 10s nc -w 1 -l -u 239.0.0.1 4000 > $RESULT)"Hi,
It's unclear to me what the purpose of the bash -c "$(...)" construction is here. Can the same be achieved using simply:
ip netns exec "${LISTENER}" \ timeout 10s nc -w 1 -l -u 239.0.0.1 4000 > $RESULT
The purpose of using bash -s was to avoid exiting main bash program by timeout expiration due to 'set -e' option. But Jakub avoided that problem by adding (|| true) in the recent patch.
Also, not strictly related to this patch, it seems a little odd here, and elsewhere, to call bash in a /bin/sh script.
Oh Thanks, Shebang should be bash, not sh. I will fix it.
RESULT4=$(< $RESULT) if [ "$RESULT4" == "172.17.0.2" ]; then printf "TEST: %-60s [ OK ]\n" "IPv4 amt multicast forwarding"
exit 0 else printf "TEST: %-60s [FAIL]\n" "IPv4 amt multicast forwarding"exit 1 fi
}
...
send_mcast4() { sleep 2
ip netns exec "${SOURCE}" bash -c \'echo 172.17.0.2 | nc -w 1 -u 239.0.0.1 4000' &
for n in {0..10}; doip netns exec "${SOURCE}" bash -c \'echo 172.17.0.2 | nc -w 1 -u 239.0.0.1 4000'sleep 1done}
send_mcast6() { sleep 2
ip netns exec "${SOURCE}" bash -c \'echo 2001:db8:3::2 | nc -w 1 -u ff0e::5:6 6000' &
for n in {0..10}; doip netns exec "${SOURCE}" bash -c \'echo 2001:db8:3::2 | nc -w 1 -u ff0e::5:6 6000'sleep 1done}
check_features
...
-- pw-bot: under-review
Thanks a lot! Taehee Yoo
On Wed, 2024-05-08 at 04:06 +0000, Taehee Yoo wrote:
@@ -210,40 +217,52 @@ check_features() test_ipv4_forward() {
- RESULT4=$(ip netns exec "${LISTENER}" nc -w 1 -l -u 239.0.0.1 4000)
- echo "" > $RESULT
- bash -c "$(ip netns exec "${LISTENER}" \
timeout 10s > $RESULT)"- RESULT4=$(< $RESULT)
if you instead do:
RESULT4=$(timeout 10s ip netns exec \ "${LISTENER}" nc -w 1 -l -u 239.0.0.1 4000)
You can avoid the additional tmp file (RESULT)
if [ "$RESULT4" == "172.17.0.2" ]; then printf "TEST: %-60s [ OK ]\n" "IPv4 amt multicast forwarding"
exit 0exit 1
}
[...]
@@ -259,19 +278,17 @@ setup_iptables setup_mcast_routing test_remote_ip test_ipv4_forward & -pid=$! -send_mcast4 -wait $pid || err=$? -if [ $err -eq 1 ]; then
- ERR=1
-fi +spid=$! +send_mcast4 & +cpid=$! +wait $spid
It looks like you don't capture anymore the return code from test_ipv4_forward, why?
That will foul the test runner infra to think that this test is always successful.
Paolo
On Thu, May 9, 2024 at 6:36 PM Paolo Abeni pabeni@redhat.com wrote:
Hi Paolo, Thank you for the review!
On Wed, 2024-05-08 at 04:06 +0000, Taehee Yoo wrote:
@@ -210,40 +217,52 @@ check_features()
test_ipv4_forward() {
- RESULT4=$(ip netns exec "${LISTENER}" nc -w 1 -l -u 239.0.0.1 4000)
- echo "" > $RESULT
- bash -c "$(ip netns exec "${LISTENER}" \
- timeout 10s > $RESULT)"
- RESULT4=$(< $RESULT)
if you instead do:
RESULT4=$(timeout 10s ip netns exec \ "${LISTENER}" nc -w 1 -l -u 239.0.0.1 4000)
You can avoid the additional tmp file (RESULT)
Thanks, In the recent patch from Jakub, a variable is used instead of a file.
if [ "$RESULT4" == "172.17.0.2" ]; then printf "TEST: %-60s [ OK ]\n" "IPv4 amt multicast forwarding"
- exit 0
else printf "TEST: %-60s [FAIL]\n" "IPv4 amt multicast forwarding"
- exit 1
fi
}
[...]
@@ -259,19 +278,17 @@ setup_iptables setup_mcast_routing test_remote_ip test_ipv4_forward & -pid=$! -send_mcast4 -wait $pid || err=$? -if [ $err -eq 1 ]; then
- ERR=1
-fi +spid=$! +send_mcast4 & +cpid=$! +wait $spid
It looks like you don't capture anymore the return code from test_ipv4_forward, why?
That will foul the test runner infra to think that this test is always successful.
You're right, Sorry, I didn't consider it. It should not be changed.
Paolo
Thanks a lot! Taehee Yoo
On Wed, 8 May 2024 04:06:43 +0000 Taehee Yoo wrote:
In the forwarding testcase, it opens a server and a client with the nc. The server receives the correct message from NC, it prints OK. The server prints FAIL if it receives the wrong message from the client.
But If the server can't receive any message, it will not close so the amt.sh waits forever. There are several reasons.
- crash of smcrouted.
- Send a message from the client to the server before the server is up.
To avoid this problem, the server waits only for 10 seconds. The client sends messages for 10 seconds. If the server is successfully closed, it kills the client.
Since this didn't fix the problem of smcroute crashing I had to take a closer look myself.
I filed https://github.com/troglobit/smcroute/issues/207 for smcroute
And sent: https://lore.kernel.org/all/20240509161919.3939966-1-kuba@kernel.org/ and https://lore.kernel.org/all/20240509161952.3940476-1-kuba@kernel.org/
Please don't use netcat in tests in the future. There are two incompatible implementations which always cause hard to repro issues.
On Fri, May 10, 2024 at 1:23 AM Jakub Kicinski kuba@kernel.org wrote:
Hi Jakub, Thanks a lot for the review and looking into the bug.
On Wed, 8 May 2024 04:06:43 +0000 Taehee Yoo wrote:
In the forwarding testcase, it opens a server and a client with the nc. The server receives the correct message from NC, it prints OK. The server prints FAIL if it receives the wrong message from the client.
But If the server can't receive any message, it will not close so the amt.sh waits forever. There are several reasons.
- crash of smcrouted.
- Send a message from the client to the server before the server is up.
To avoid this problem, the server waits only for 10 seconds. The client sends messages for 10 seconds. If the server is successfully closed, it kills the client.
Since this didn't fix the problem of smcroute crashing I had to take a closer look myself.
I filed https://github.com/troglobit/smcroute/issues/207 for smcroute
Thank you so much for looking into this bug! I will test it.
And sent: https://lore.kernel.org/all/20240509161919.3939966-1-kuba@kernel.org/ and https://lore.kernel.org/all/20240509161952.3940476-1-kuba@kernel.org/
Please don't use netcat in tests in the future. There are two incompatible implementations which always cause hard to repro issues.
Okay, I will not use netcat in the future.
Thanks a lot! Taehee Yoo
linux-kselftest-mirror@lists.linaro.org
-
Jakub Kicinski -
Paolo Abeni -
Simon Horman -
Taehee Yoo