[ovs-dev] [PATCH ovn] Fix ovn-controller generated packets from getting dropped for reject ACL action.
numans at ovn.org
numans at ovn.org
Mon Jun 15 09:47:27 UTC 2020
From: Numan Siddique <numans at ovn.org>
TCP reset/ICMP packet generated by ovn-controller for the ACL reject action
gets dropped by ovs-vswithd with the below messages in ovs-vswitchd log
even though ovn-controller sets the in_port as OFPP_CONTROLLER.
----
ofproto_dpif_upcall(handler1)|INFO|received packet on unassociated datapath port 4294967295
ofproto_dpif_upcall(revalidator37)|WARN|Failed to acquire udpif_key corresponding to
unexpected flow (Invalid argument): ufid:0daac824-bda7-44d8-ad38-cdd9c5f0fc97
----
ovs-vswitchd drops the packet because the in_port is 0.
The below OF flow sets the in_port to 0 if 'MLF_ALLOW_LOOPBACK_BIT' is set in the REG0
in table 64.
priority=100,reg10=0x1/0x1,reg15=0x2,metadata=0x2 actions=push:NXM_OF_IN_PORT[],load:0->NXM_OF_IN_PORT[],resubmit(,65),pop:NXM_OF_IN_PORT[]
This patch fixes this issue by setting the in_port to OFPP_NONE so that ovs-vswitchd
doesn't drop the packet.
Reported-at: https://bugzilla.redhat.com/show_bug.cgi?id=1832176
Signed-off-by: Numan Siddique <numans at ovn.org>
---
controller/physical.c | 18 ++++++++++++------
tests/system-ovn.at | 32 ++++++++++++++++++++++++++++++++
2 files changed, 44 insertions(+), 6 deletions(-)
diff --git a/controller/physical.c b/controller/physical.c
index f06313b9d..d8e312426 100644
--- a/controller/physical.c
+++ b/controller/physical.c
@@ -765,12 +765,18 @@ put_local_common_flows(uint32_t dp_key, uint32_t port_key,
* - or if the destination is a nested container
* - or if "nested_container" flag is set and the destination is the
* parent port,
- * temporarily set the in_port to zero, resubmit to
+ * temporarily set the in_port to OFPP_NONE, resubmit to
* table 65 for logical-to-physical translation, then restore
* the port number.
*
* If 'parent_port_key' is set, then the 'port_key' represents a nested
- * container. */
+ * container.
+ *
+ * Note:We can set in_port to 0 too. But if recirculation happens
+ * later (eg. clone action to enter peer pipeline and a subsequent
+ * ct action), ovs-vswitchd will drop the packet if the frozen metadata
+ * in_port is 0.
+ * */
bool nested_container = parent_port_key ? true: false;
match_init_catchall(&match);
@@ -783,7 +789,7 @@ put_local_common_flows(uint32_t dp_key, uint32_t port_key,
}
put_stack(MFF_IN_PORT, ofpact_put_STACK_PUSH(ofpacts_p));
- put_load(0, MFF_IN_PORT, 0, 16, ofpacts_p);
+ put_load(ofp_to_u16(OFPP_NONE), MFF_IN_PORT, 0, 16, ofpacts_p);
put_resubmit(OFTABLE_LOG_TO_PHY, ofpacts_p);
put_stack(MFF_IN_PORT, ofpact_put_STACK_POP(ofpacts_p));
ofctrl_add_flow(flow_table, OFTABLE_SAVE_INPORT, 100, 0,
@@ -792,8 +798,8 @@ put_local_common_flows(uint32_t dp_key, uint32_t port_key,
if (nested_container) {
/* It's a nested container and when the packet from the nested
* container is to be sent to the parent port, "nested_container"
- * flag will be set. We need to temporarily set the in_port to zero
- * as mentioned in the comment above.
+ * flag will be set. We need to temporarily set the in_port to
+ * OFPP_NONE as mentioned in the comment above.
*
* If a parent port has multiple child ports, then this if condition
* will be hit multiple times, but we want to add only one flow.
@@ -814,7 +820,7 @@ put_local_common_flows(uint32_t dp_key, uint32_t port_key,
MLF_NESTED_CONTAINER, MLF_NESTED_CONTAINER);
put_stack(MFF_IN_PORT, ofpact_put_STACK_PUSH(ofpacts_p));
- put_load(0, MFF_IN_PORT, 0, 16, ofpacts_p);
+ put_load(ofp_to_u16(OFPP_NONE), MFF_IN_PORT, 0, 16, ofpacts_p);
put_resubmit(OFTABLE_LOG_TO_PHY, ofpacts_p);
put_stack(MFF_IN_PORT, ofpact_put_STACK_POP(ofpacts_p));
ofctrl_check_and_add_flow(flow_table, OFTABLE_SAVE_INPORT, 100, 0,
diff --git a/tests/system-ovn.at b/tests/system-ovn.at
index 9dfe6a4ad..52f05f07e 100644
--- a/tests/system-ovn.at
+++ b/tests/system-ovn.at
@@ -3927,6 +3927,24 @@ ovn-nbctl acl-add pg0 to-lport 1002 "outport == @pg0 && ip4 && ip4.src == 0.0.0.
ovn-nbctl --log acl-add pg0 to-lport 1004 "inport == @pg0 && ip && tcp && tcp.dst == 84" reject
ovn-nbctl --log acl-add pg0 to-lport 1004 "inport == @pg0 && ip && udp && udp.dst == 94" reject
+ovn-nbctl ls-add sw1
+ovn-nbctl lsp-add sw1 sw1-p1-rej
+ovn-nbctl lsp-set-addresses sw1-p1-rej "40:54:00:00:00:03 20.0.0.3"
+ovn-nbctl lsp-set-port-security sw1-p1-rej "40:54:00:00:00:03 20.0.0.3"
+
+ovn-nbctl lr-add lr0
+ovn-nbctl lrp-add lr0 lr0-sw0 00:00:00:00:ff:01 10.0.0.1/24
+ovn-nbctl lsp-add sw0 sw0-lr0
+ovn-nbctl lsp-set-type sw0-lr0 router
+ovn-nbctl lsp-set-addresses sw0-lr0 router
+ovn-nbctl lsp-set-options sw0-lr0 router-port=lr0-sw0
+
+ovn-nbctl lrp-add lr0 lr0-sw1 00:00:00:00:ff:02 20.0.0.1/24
+ovn-nbctl lsp-add sw1 sw1-lr0
+ovn-nbctl lsp-set-type sw1-lr0 router
+ovn-nbctl lsp-set-addresses sw1-lr0 router
+ovn-nbctl lsp-set-options sw1-lr0 router-port=lr0-sw1
+
OVN_POPULATE_ARP
ovn-nbctl --wait=hv sync
@@ -3941,6 +3959,10 @@ ADD_VETH(sw0-p2-rej, sw0-p2-rej, br-int, "10.0.0.4/24", "50:54:00:00:00:04", \
NS_CHECK_EXEC([sw0-p1-rej], [ip a a aef0::3/64 dev sw0-p1-rej], [0])
NS_CHECK_EXEC([sw0-p2-rej], [ip a a aef0::4/64 dev sw0-p2-rej], [0])
+ADD_NAMESPACES(sw1-p1-rej)
+ADD_VETH(sw1-p1-rej, sw1-p1-rej, br-int, "20.0.0.3/24", "40:54:00:00:00:03", \
+ "20.0.0.1")
+
sleep 1
# Capture packets in sw0-p1-rej.
@@ -3993,6 +4015,16 @@ OVS_WAIT_UNTIL([
test "${total}" = "2"
])
+ovn-nbctl acl-add sw1 from-lport 1004 "ip" allow-related
+ovn-nbctl acl-add sw1 to-lport 1004 "ip" allow-related
+ovn-nbctl --log acl-add pg0 to-lport 1004 "outport == @pg0 && ip && tcp && tcp.dst == 84" reject
+
+OVS_WAIT_UNTIL([
+ ip netns exec sw1-p1-rej nc 10.0.0.4 84 2> r
+ res=$(cat r)
+ test "$res" = "Ncat: Connection refused."
+])
+
# Now test for IPv4 UDP.
NS_CHECK_EXEC([sw0-p1-rej], [tcpdump -n -c 1 -i sw0-p1-rej udp port 90 > sw0-p1-rej-udp.pcap &], [0])
NS_CHECK_EXEC([sw0-p1-rej], [tcpdump -n -c 1 -i sw0-p1-rej icmp > sw0-p1-rej-icmp.pcap &], [0])
--
2.26.2
More information about the dev
mailing list