LU-15440 lnet: lnet_peer_data_present() memory leak

If the ping buffer has nnis <= 1 then the ref on the ping buffer does
not get dropped. This causes a memory leak.

Lustre-change: https://review.whamcloud.com/46052
Lustre-commit: 56384a4fc39ff99c8abb3538f93d303f2be6ab45

Test-Parameters: trivial testlist=sanity-lnet
Signed-off-by: Chris Horn <chris.horn@hpe.com>
Change-Id: I5e3c651ffecbe4f8860afb86770cecef23ebe862
Reviewed-on: https://review.whamcloud.com/c/ex/lustre-release/+/51684
Tested-by: jenkins <devops@whamcloud.com>
Tested-by: Maloo <maloo@whamcloud.com>
Reviewed-by: Cyril Bordage <cbordage@whamcloud.com>
Reviewed-by: Andreas Dilger <adilger@whamcloud.com>

diff --git a/lnet/lnet/peer.c b/lnet/lnet/peer.c
index 45c74f6..fd899b0 100644 (file)
--- a/lnet/lnet/peer.c
+++ b/lnet/lnet/peer.c
@@ -3334,8 +3334,10 @@ __must_hold(&lp->lp_lock)
         * primary NID to the correct value here. Moreover, this peer
         * can show up with only the loopback NID in the ping buffer.
         */
-       if (pbuf->pb_info.pi_nnis <= 1)
+       if (pbuf->pb_info.pi_nnis <= 1) {
+               lnet_ping_buffer_decref(pbuf);
                goto out;
+       }
        nid = pbuf->pb_info.pi_ni[1].ns_nid;
        if (lp->lp_primary_nid == LNET_NID_LO_0) {
                rc = lnet_peer_set_primary_nid(lp, nid, flags);

diff --git a/lustre/tests/sanity-lnet.sh b/lustre/tests/sanity-lnet.sh
index 17f8f80..5dae873 100755 (executable)
--- a/lustre/tests/sanity-lnet.sh
+++ b/lustre/tests/sanity-lnet.sh
@@ -2120,26 +2120,58 @@ test_215() {
 }
 run_test 215 "Test lnetctl ping --source option"
 
-test_219() {
+test_216() {
+       local rc=0
+
        reinit_dlc || return $?
+
        add_net "tcp" "${INTERFACES[0]}" || return $?
        add_net "tcp1" "${INTERFACES[0]}" || return $?
 
-       local nid1=$(lctl list_nids | head -n 1)
-       local nid2=$(lctl list_nids | tail --lines 1)
+       local nids=( $($LCTL list_nids | xargs echo) )
 
-       do_lnetctl ping $nid1 ||
-               error "Ping failed $?"
-       do_lnetctl ping $nid2 ||
-               error "Ping failed $?"
+       do_lnetctl discover ${nids[0]} ||
+               error "Initial discovery failed"
 
-       do_lnetctl discover $nid2 ||
-               error "Discovery failed"
+       do_lnetctl ping --source ${nids[0]} ${nids[0]} ||
+               error "Initial ping failed $?"
 
-       $LNETCTL peer show --nid $nid1 | grep -q $nid2 ||
-               error "$nid2 is not listed under $nid1"
+       do_lnetctl ping --source ${nids[1]} ${nids[1]} ||
+               error "Initial ping failed $?"
+
+       local src dst
+       for src in ${nids[@]}; do
+               for dst in ${nids[@]}; do
+                       $LCTL net_drop_add -r 1 -s $src -d $dst -e network_timeout
+               done
+       done
+
+       do_lnetctl ping ${nids[0]} || rc=$?
+
+       $LCTL net_drop_del -a
+
+       [[ $rc -eq 0 ]] &&
+               error "expected ping to fail"
+
+       check_nid_in_recovq "-p" 0
+       check_nid_in_recovq "-l" 1
+
+       return 0
 }
-run_test 219 "Consolidate peer entries"
+run_test 216 "Failed send to peer NI owned by local host should not trigger peer NI recovery"
+
+test_217() {
+       reinit_dlc || return $?
+
+       [[ $($LNETCTL net show | grep -c nid) -ne 1 ]] &&
+               error "Unexpected number of NIs after initalizing DLC"
+
+       do_lnetctl discover 0@lo ||
+               error "Failed to discover 0@lo"
+
+       unload_modules
+}
+run_test 217 "Don't leak memory when discovering peer with nnis <= 1"
 
 test_218() {
        reinit_dlc || return $?
@@ -2194,6 +2226,27 @@ test_218() {
 }
 run_test 218 "Local recovery pings should exercise all available paths"
 
+test_219() {
+       reinit_dlc || return $?
+       add_net "tcp" "${INTERFACES[0]}" || return $?
+       add_net "tcp1" "${INTERFACES[0]}" || return $?
+
+       local nid1=$(lctl list_nids | head -n 1)
+       local nid2=$(lctl list_nids | tail --lines 1)
+
+       do_lnetctl ping $nid1 ||
+               error "Ping failed $?"
+       do_lnetctl ping $nid2 ||
+               error "Ping failed $?"
+
+       do_lnetctl discover $nid2 ||
+               error "Discovery failed"
+
+       $LNETCTL peer show --nid $nid1 | grep -q $nid2 ||
+               error "$nid2 is not listed under $nid1"
+}
+run_test 219 "Consolidate peer entries"
+
 test_230() {
        # LU-12815
        echo "Check valid values; Should succeed"