Thanks Ilya. I will work on these as well.
-----Original Message-----
From: Ilya Maximets [mailto:i.maxim...@samsung.com]
Sent: Friday, January 18, 2019 5:33 PM
To: Anju Thomas <anju.tho...@ericsson.com>; d...@openvswitch.org
Cc: Keshav Gupta <keshugup...@gmail.com>; Stokes, Ian <ian.sto...@intel.com>;
Ben Pfaff <b...@ovn.org>
Subject: Re: [ovs-dev,v6] Improved Packet Drop Statistics in OVS
Hi.
Thanks for working on this.
Some general comments:
1. This patch consists of two separate features:
- Reporting the drop reason in flow dumps.
- Counters for different drop types.
I still think that there should be patch-set of two separate patches.
This will simplify the review and allow to apply them separately/speed
up accepting.
2. There are some issues with the patch formatting. The commit message is
indented like in the output of 'git show' command. So, it's double
shifted after applying in git. Please remove the additional indentation.
You probably should start using 'git format-patch'.
Second issue:
Applying: Improved Packet Drop Statistics in OVS
.git/rebase-apply/patch:415: new blank line at EOF.
+
warning: 1 line adds whitespace errors.
One more thing is that for better readability it's a common practice to
limit the width of lines in commit-message to 72 characters. For example,
you may strip not important fields of the dump-flows output. Like this:
recirc_id(0),in_port(5),<...>, actions:drop:recursion too deep
Other comments inline.
Best regards, Ilya Maximets.
On 17.01.2019 7:49, Anju Thomas wrote:
> Currently OVS maintains explicit packet drop/error counters only on port
> level. Packets that are dropped as part of normal OpenFlow processing are
> counted in flow stats of “drop” flows or as table misses in table stats.
> These can only be interpreted by controllers that know the semantics of
> the configured OpenFlow pipeline. Without that knowledge, it is impossible
> for an OVS user to obtain e.g. the total number of packets dropped due to
> OpenFlow rules.
>
> Furthermore, there are numerous other reasons for which packets can be
> dropped by OVS slow path that are not related to the OpenFlow pipeline.
> The generated datapath flow entries include a drop action to avoid further
> expensive upcalls to the slow path, but subsequent packets dropped by the
> datapath are not accounted anywhere.
>
> Finally, the datapath itself drops packets in certain error situations.
> Also, these drops are today not accounted for.
>
> This makes it difficult for OVS users to monitor packet drop in an OVS
> instance and to alert a management system in case of a unexpected increase
> of such drops. Also OVS trouble-shooters face difficulties in analysing
> packet drops.
>
> With this patch we implement following changes to address the issues
> mentioned above.
>
> 1. Account and categorize all the packet drops in OVS.
> 2. Account & classify “drop” action packet drops according to the drop
> reason.
> 3. Identify and account all the silent packet drop scenarios.
> 4. Display these drops in ovs-appctl coverage/show
> 5. Modified ovs-appctl dpcls/dump-flows and ovs-appctl dpif/dump-flows
> to print drop reason along with drop action
>
> A detailed presentation on this was presented at OvS conference 2017 and
> link for the corresponding presentation is available at:
>
> https://www.slideshare.net/LF_OpenvSwitch/lfovs17troubleshooting-the-data-plane-in-ovs-82280329
>
> Sample ovs-appctl dpcls/dump-flows & ovs-appctl dpif/dump-flows
> displaying drop reason along with drop action.
>
> The idea is to use the coverage infrastructure to maintain the drops
>
> $ ovs-appctl dpctl/dump-flows netdev at ovs-netdev
> flow-dump from pmd on cpu core: 0
>
> recirc_id(0),in_port(5),packet_type(ns=0,id=0),eth_type(0x0800),ipv4(frag=no),
> packets:12, bytes:1176, used:0.884s, actions:drop:recursion too deep
>
> $ ovs-appctl dpif/dump-flows br-int
>
> recirc_id(0),in_port(5),packet_type(ns=0,id=0),eth_type(0x0800),ipv4(frag=no),
> packets:25, bytes:2450, used:5.008s, actions:drop:recursion too deep
> recirc_id(0),in_port(5),packet_type(ns=0,id=0),eth_type(0x0806),
> packets:7, bytes:294, used:0.009s, actions:drop:recursion too deep
>
> In subsequent commits, we are planning to see if we can use this
> infrastructure to create a
> wrapper to clear and display counters as well.
>
> Co-authored-by: Rohith Basavaraja <rohith.basavar...@gmail.com>
> Co-authored-by: Keshav Gupta <keshugup...@gmail.com>
> Signed-off-by: Anju Thomas <anju.tho...@ericsson.com>
> Signed-off-by: Rohith Basavaraja <rohith.basavar...@gmail.com>
> Signed-off-by: Keshav Gupta <keshugup...@gmail.com>
> Signed-off-by: Anju Thomas <anju.tho...@ericsson.com>
> Signed-off-by: Rohith Basavaraja <rohith.basavar...@gmail.com>
> Signed-off-by: Keshav Gupta <keshugup...@gmail.com>
> ---
> datapath/linux/compat/include/linux/openvswitch.h | 51 ++++++
> lib/coverage.c | 6 +-
> lib/coverage.h | 3 +-
> lib/dpif-netdev.c | 105 +++++++++++-
> lib/dpif.c | 10 +-
> lib/dpif.h | 3 +
> lib/netdev-dpdk.c | 4 +
> lib/odp-execute.c | 50 ++++--
> lib/odp-execute.h | 10 +-
> lib/odp-util.c | 54 +++++-
> ofproto/ofproto-dpif-ipfix.c | 1 +
> ofproto/ofproto-dpif-sflow.c | 1 +
> ofproto/ofproto-dpif-upcall.c | 3 +-
> ofproto/ofproto-dpif-xlate.c | 64 +++++++
> ofproto/ofproto-dpif-xlate.h | 4 +
> ofproto/ofproto-dpif.c | 11 +-
> ofproto/ofproto-dpif.h | 8 +-
> tests/automake.mk | 3 +-
> tests/bundle.at | 2 +-
> tests/classifier.at | 10 +-
> tests/dpif-netdev.at | 17 +-
> tests/drop-stats.at | 197
> ++++++++++++++++++++++
> tests/ofproto-dpif.at | 118 ++++++-------
> tests/ovs-ofctl.at | 2 +-
> tests/packet-type-aware.at | 17 +-
> tests/testsuite.at | 1 +
> tests/tunnel-push-pop.at | 26 ++-
> tests/tunnel.at | 21 ++-
> 28 files changed, 695 insertions(+), 107 deletions(-)
> create mode 100644 tests/drop-stats.at
>
> diff --git a/datapath/linux/compat/include/linux/openvswitch.h
> b/datapath/linux/compat/include/linux/openvswitch.h
> index 9b087f1..b66b46f 100644
> --- a/datapath/linux/compat/include/linux/openvswitch.h
> +++ b/datapath/linux/compat/include/linux/openvswitch.h
> @@ -819,6 +819,56 @@ struct ovs_action_push_eth {
> };
>
> /**
> + * enum ovs_drop_reason - Reasons for installing %OVS_ACTION_ATTR_DROP.
> + * @OVS_DROP_REASON_OF_PIPELINE: Explicit drop action in the pipeline.
> + * @OVS_DROP_REASON_BRIDGE_NOT_FOUND: Xlation error generated due to
> + * unable to determine bridge.
> + * @OVS_DROP_REASON_RECURSION_TOO_DEEP: Xlation error generated due to
> + * recursion reached maximum depth.
> + * @OVS_DROP_REASON_TOO_MANY_RESUBMITS: Xlation error generated due to
> + * too many resubmits.
> + * @OVS_DROP_REASON_STACK_TOO_DEEP: Xlation error generated due to stack
> + * too deep.
> + * @OVS_DROP_REASON_NO_RECIRCULATION_CONTEXT: Xlation error generated
> + * due to no recirculation context.
> + * @OVS_DROP_REASON_RECIRCULATION_CONFLICT: Xlation error generated due to
> + * conflict in recirculation context.
> + * @OVS_DROP_REASON_TOO_MANY_MPLS_LABELS: Xlation error generated due to
> + * too many mpls labels.
> + * @OVS_DROP_REASON_INVALID_TUNNEL_METADATA: Xlation error generated due to
> + * invalid tunnel metadata.
> + * @OVS_DROP_REASON_UNSUPPORTED_PACKET_TYPE: Xlation error generated due to
> + * unsupported packet type.
> + * @OVS_DROP_REASON_CONGESTION: Xlation error generated due to ecn mismatch
> + * during tunnel decapsulation.
> + * @OVS_DROP_REASON_FORWARDING_DISABLED: Xlation error generated due to
> + * forwarding is disabled.
> + */
> +enum ovs_drop_reason {
> + OVS_DROP_REASON_OF_PIPELINE = 0,
> + OVS_DROP_REASON_BRIDGE_NOT_FOUND,
> + OVS_DROP_REASON_RECURSION_TOO_DEEP,
> + OVS_DROP_REASON_TOO_MANY_RESUBMITS,
> + OVS_DROP_REASON_STACK_TOO_DEEP,
> + OVS_DROP_REASON_NO_RECIRCULATION_CONTEXT,
> + OVS_DROP_REASON_RECIRCULATION_CONFLICT,
> + OVS_DROP_REASON_TOO_MANY_MPLS_LABELS,
> + OVS_DROP_REASON_INVALID_TUNNEL_METADATA,
> + OVS_DROP_REASON_UNSUPPORTED_PACKET_TYPE,
> + OVS_DROP_REASON_CONGESTION,
> + OVS_DROP_REASON_FORWARDING_DISABLED,
> + OVS_DROP_REASON_MAX,
> +};
> +
> +/*
> + * struct ovs_action_drop - %OVS_ACTION_ATTR_DROP action argument.
> + * @drop_reason: Reason for installing drop action.
> + */
> +struct ovs_action_drop {
> + enum ovs_drop_reason drop_reason;
> +};
> +
> +/**
> * enum ovs_nat_attr - Attributes for %OVS_CT_ATTR_NAT.
> *
> * @OVS_NAT_ATTR_SRC: Flag for Source NAT (mangle source address/port).
> @@ -938,6 +988,7 @@ enum ovs_action_attr {
> OVS_ACTION_ATTR_POP_NSH, /* No argument. */
> OVS_ACTION_ATTR_METER, /* u32 meter number. */
> OVS_ACTION_ATTR_CLONE, /* Nested OVS_CLONE_ATTR_*. */
> + OVS_ACTION_ATTR_DROP, /* explicit drop action. */
This is the linux kernel header. You should use tabs here.
>
> #ifndef __KERNEL__
> OVS_ACTION_ATTR_TUNNEL_PUSH, /* struct ovs_action_push_tnl*/
> diff --git a/lib/coverage.c b/lib/coverage.c
> index 6cef826..c9acc37 100644
> --- a/lib/coverage.c
> +++ b/lib/coverage.c
> @@ -32,9 +32,7 @@ VLOG_DEFINE_THIS_MODULE(coverage);
> static struct coverage_counter **coverage_counters = NULL;
> static size_t n_coverage_counters = 0;
> static size_t allocated_coverage_counters = 0;
> -
> -static struct ovs_mutex coverage_mutex = OVS_MUTEX_INITIALIZER;
This is leftover from the previous versions. No need to expose
'coverage_mutex' in current patch.
> -
> +struct ovs_mutex coverage_mutex = OVS_MUTEX_INITIALIZER;
> DEFINE_STATIC_PER_THREAD_DATA(long long int, coverage_clear_time, LLONG_MIN);
> static long long int coverage_run_time = LLONG_MIN;
>
> @@ -52,7 +50,7 @@ coverage_counter_register(struct coverage_counter* counter)
> if (n_coverage_counters >= allocated_coverage_counters) {
> coverage_counters = x2nrealloc(coverage_counters,
> &allocated_coverage_counters,
> - sizeof(struct coverage_counter*));
> + sizeof(struct coverage_counter *));
this change is also not needed.
> }
> coverage_counters[n_coverage_counters++] = counter;
> }
> diff --git a/lib/coverage.h b/lib/coverage.h
> index dea990e..45b1dbd 100644
> --- a/lib/coverage.h
> +++ b/lib/coverage.h
> @@ -54,7 +54,8 @@ struct coverage_counter {
> unsigned int hr[HR_AVG_LEN];
> };
>
> -void coverage_counter_register(struct coverage_counter*);
> +extern struct ovs_mutex coverage_mutex;
> +void coverage_counter_register(struct coverage_counter *);
This is leftover from the previous versions. No need to expose
'coverage_mutex' in current patch.
>
> /* Defines COUNTER. There must be exactly one such definition at file scope
> * within a program. */
> diff --git a/lib/dpif-netdev.c b/lib/dpif-netdev.c
> index be529b6..8749cba 100644
> --- a/lib/dpif-netdev.c
> +++ b/lib/dpif-netdev.c
> @@ -100,6 +100,30 @@ enum { MAX_METERS = 65536 }; /* Maximum number of
> meters. */
> enum { MAX_BANDS = 8 }; /* Maximum number of bands / meter. */
> enum { N_METER_LOCKS = 64 }; /* Maximum number of meters. */
>
> +
> +COVERAGE_DEFINE(drop_action_of_pipeline);
> +COVERAGE_DEFINE(drop_action_bridge_not_found);
> +COVERAGE_DEFINE(drop_action_recursion_too_deep);
> +COVERAGE_DEFINE(drop_action_too_many_resubmit);
> +COVERAGE_DEFINE(drop_action_stack_too_deep);
> +COVERAGE_DEFINE(drop_action_no_recirculation_context);
> +COVERAGE_DEFINE(drop_action_recirculation_conflict);
> +COVERAGE_DEFINE(drop_action_too_many_mpls_labels);
> +COVERAGE_DEFINE(drop_action_invalid_tunnel_metadata);
> +COVERAGE_DEFINE(drop_action_unsupported_packet_type);
> +COVERAGE_DEFINE(drop_action_congestion);
> +COVERAGE_DEFINE(drop_action_forwarding_disabled);
Above counters could be defined inside odp-execute.c. You already have
2 other counters there. In this case there will be no need to introduce
additional 'dp_update_drop_action_counter_cb' callback and pass it around.
All the required logic will also be moved to odp-execute.c.
> +COVERAGE_DEFINE(dp_meter_drop);
> +COVERAGE_DEFINE(dp_upcall_error_drop);
> +COVERAGE_DEFINE(dp_lock_error_drop);
> +COVERAGE_DEFINE(dp_userspace_action_error_drop);
> +COVERAGE_DEFINE(dp_tunnel_push_error_drop);
> +COVERAGE_DEFINE(dp_tunnel_pop_error_drop);
> +COVERAGE_DEFINE(dp_recirc_error_drop);
> +COVERAGE_DEFINE(dp_invalid_port_drop);
> +COVERAGE_DEFINE(dp_invalid_tnl_port_drop);
> +COVERAGE_DEFINE(rx_invalid_packet_drop);
What do you think about renaming above counters to 'datapath_drop_*' ?
IMHO, it'll be more visually recognizable.
Like 'datapath_drop_rx_invalid_packet' or 'datapath_drop_upcall_error'.
> +
> /* Protects against changes to 'dp_netdevs'. */
> static struct ovs_mutex dp_netdev_mutex = OVS_MUTEX_INITIALIZER;
>
> @@ -829,7 +853,6 @@ static inline bool
> pmd_perf_metrics_enabled(const struct dp_netdev_pmd_thread *pmd);
> static void queue_netdev_flow_del(struct dp_netdev_pmd_thread *pmd,
> struct dp_netdev_flow *flow);
> -
No need to remove an empty line.
> static void
> emc_cache_init(struct emc_cache *flow_cache)
> {
> @@ -1388,6 +1411,7 @@ dpif_netdev_init(void)
> return 0;
> }
>
> +
No need to add an empty line.
> static int
> dpif_netdev_enumerate(struct sset *all_dps,
> const struct dpif_class *dpif_class)
> @@ -5563,7 +5587,7 @@ dp_netdev_run_meter(struct dp_netdev *dp, struct
> dp_packet_batch *packets_,
> band = &meter->bands[exceeded_band[j]];
> band->packet_count += 1;
> band->byte_count += dp_packet_size(packet);
> -
Please, keep this empty line for readability.
> + COVERAGE_INC(dp_meter_drop);
> dp_packet_delete(packet);
> } else {
> /* Meter accepts packet. */
> @@ -6320,6 +6344,7 @@ dfc_processing(struct dp_netdev_pmd_thread *pmd,
>
> if (OVS_UNLIKELY(dp_packet_size(packet) < ETH_HEADER_LEN)) {
> dp_packet_delete(packet);
> + COVERAGE_INC(rx_invalid_packet_drop);
> continue;
> }
>
> @@ -6446,6 +6471,7 @@ handle_packet_upcall(struct dp_netdev_pmd_thread *pmd,
> put_actions);
> if (OVS_UNLIKELY(error && error != ENOSPC)) {
> dp_packet_delete(packet);
> + COVERAGE_INC(dp_upcall_error_drop);
> return error;
> }
>
> @@ -6577,6 +6603,7 @@ fast_path_processing(struct dp_netdev_pmd_thread *pmd,
> DP_PACKET_BATCH_FOR_EACH (i, packet, packets_) {
> if (OVS_UNLIKELY(!rules[i])) {
> dp_packet_delete(packet);
> + COVERAGE_INC(dp_lock_error_drop);
> upcall_fail_cnt++;
> }
> }
> @@ -6846,10 +6873,59 @@ dp_execute_userspace_action(struct
> dp_netdev_pmd_thread *pmd,
> actions->data, actions->size);
> } else if (should_steal) {
> dp_packet_delete(packet);
> + COVERAGE_INC(dp_userspace_action_error_drop);
> }
> }
>
> static void
> +dp_update_drop_action_counter_cb(enum ovs_drop_reason drop_reason,
> + int delta)
Indents are a bit off.
> + OVS_NO_THREAD_SAFETY_ANALYSIS
Is it necessary to disable thread safety analysis ?
> +{
> + switch (drop_reason) {
> + case OVS_DROP_REASON_OF_PIPELINE:
> + COVERAGE_ADD(drop_action_of_pipeline, delta);
> + break;
> + case OVS_DROP_REASON_BRIDGE_NOT_FOUND:
> + COVERAGE_ADD(drop_action_bridge_not_found, delta);
> + break;
> + case OVS_DROP_REASON_RECURSION_TOO_DEEP:
> + COVERAGE_ADD(drop_action_recursion_too_deep, delta);
> + break;
> + case OVS_DROP_REASON_TOO_MANY_RESUBMITS:
> + COVERAGE_ADD(drop_action_too_many_resubmit, delta);
> + break;
> + case OVS_DROP_REASON_STACK_TOO_DEEP:
> + COVERAGE_ADD(drop_action_stack_too_deep, delta);
> + break;
> + case OVS_DROP_REASON_NO_RECIRCULATION_CONTEXT:
> + COVERAGE_ADD(drop_action_no_recirculation_context, delta);
> + break;
> + case OVS_DROP_REASON_RECIRCULATION_CONFLICT:
> + COVERAGE_ADD(drop_action_recirculation_conflict, delta);
> + break;
> + case OVS_DROP_REASON_TOO_MANY_MPLS_LABELS:
> + COVERAGE_ADD(drop_action_too_many_mpls_labels, delta);
> + break;
> + case OVS_DROP_REASON_INVALID_TUNNEL_METADATA:
> + COVERAGE_ADD(drop_action_invalid_tunnel_metadata, delta);
> + break;
> + case OVS_DROP_REASON_UNSUPPORTED_PACKET_TYPE:
> + COVERAGE_ADD(drop_action_unsupported_packet_type, delta);
> + break;
> + case OVS_DROP_REASON_CONGESTION:
> + COVERAGE_ADD(drop_action_congestion, delta);
> + break;
> + case OVS_DROP_REASON_FORWARDING_DISABLED:
> + COVERAGE_ADD(drop_action_forwarding_disabled, delta);
> + break;
> + case OVS_DROP_REASON_MAX:
> + default:
> + VLOG_ERR("Invalid Drop reason type:%d",drop_reason);
We might want the rate limiting for above message to not have it for each
packet.
> + }
> +}
> +
> +static void
> dp_execute_cb(void *aux_, struct dp_packet_batch *packets_,
> const struct nlattr *a, bool should_steal)
> OVS_NO_THREAD_SAFETY_ANALYSIS
> @@ -6860,6 +6936,7 @@ dp_execute_cb(void *aux_, struct dp_packet_batch
> *packets_,
> struct dp_netdev *dp = pmd->dp;
> int type = nl_attr_type(a);
> struct tx_port *p;
> + uint32_t packet_count, packet_dropped;
>
> switch ((enum ovs_action_attr)type) {
> case OVS_ACTION_ATTR_OUTPUT:
> @@ -6901,6 +6978,8 @@ dp_execute_cb(void *aux_, struct dp_packet_batch
> *packets_,
> dp_packet_batch_add(&p->output_pkts, packet);
> }
> return;
> + } else {
> + COVERAGE_ADD(dp_invalid_port_drop, packets_->count);
> }
> break;
>
> @@ -6910,10 +6989,13 @@ dp_execute_cb(void *aux_, struct dp_packet_batch
> *packets_,
> * the ownership of these packets. Thus, we can avoid performing
> * the action, because the caller will not use the result anyway.
> * Just break to free the batch. */
> + COVERAGE_ADD(dp_tunnel_push_error_drop, packets_->count);
> break;
> }
> dp_packet_batch_apply_cutlen(packets_);
> - push_tnl_action(pmd, a, packets_);
> + if (push_tnl_action(pmd, a, packets_)) {
> + COVERAGE_ADD(dp_tunnel_push_error_drop, packets_->count);
> + }
> return;
>
> case OVS_ACTION_ATTR_TUNNEL_POP:
> @@ -6933,7 +7015,12 @@ dp_execute_cb(void *aux_, struct dp_packet_batch
> *packets_,
>
> dp_packet_batch_apply_cutlen(packets_);
>
> + packet_count = packets_->count;
> netdev_pop_header(p->port->netdev, packets_);
> + packet_dropped = packet_count - packets_->count;
> + if (packet_dropped) {
> + COVERAGE_ADD(dp_tunnel_pop_error_drop, packet_dropped);
> + }
> if (dp_packet_batch_is_empty(packets_)) {
> return;
> }
> @@ -6947,7 +7034,11 @@ dp_execute_cb(void *aux_, struct dp_packet_batch
> *packets_,
> dp_netdev_recirculate(pmd, packets_);
> (*depth)--;
> return;
> + } else {
> + COVERAGE_ADD(dp_invalid_tnl_port_drop, packets_->count);
> }
> + } else {
> + COVERAGE_ADD(dp_recirc_error_drop, packets_->count);
> }
> break;
>
> @@ -6991,6 +7082,8 @@ dp_execute_cb(void *aux_, struct dp_packet_batch
> *packets_,
> fat_rwlock_unlock(&dp->upcall_rwlock);
>
> return;
> + } else {
> + COVERAGE_ADD(dp_lock_error_drop, packets_->count);
> }
> break;
>
> @@ -7013,6 +7106,8 @@ dp_execute_cb(void *aux_, struct dp_packet_batch
> *packets_,
> (*depth)--;
>
> return;
> + } else {
> + COVERAGE_ADD(dp_recirc_error_drop, packets_->count);
> }
'Else' condition is not needed here as we're returning inside the 'if'.
This is actually applicable to most of other places in this function,
but in this one we already have a warning below without 'else' condition.
>
> VLOG_WARN("Packet dropped. Max recirculation depth exceeded.");
> @@ -7167,6 +7262,7 @@ dp_execute_cb(void *aux_, struct dp_packet_batch
> *packets_,
> case OVS_ACTION_ATTR_PUSH_NSH:
> case OVS_ACTION_ATTR_POP_NSH:
> case OVS_ACTION_ATTR_CT_CLEAR:
> + case OVS_ACTION_ATTR_DROP:
> case __OVS_ACTION_ATTR_MAX:
> OVS_NOT_REACHED();
> }
> @@ -7183,7 +7279,8 @@ dp_netdev_execute_actions(struct dp_netdev_pmd_thread
> *pmd,
> struct dp_netdev_execute_aux aux = { pmd, flow };
>
> odp_execute_actions(&aux, packets, should_steal, actions,
> - actions_len, dp_execute_cb);
> + actions_len, dp_execute_cb,
> + dp_update_drop_action_counter_cb);
As I already wrote, we'll not need to have a callback function if you'll
move the 'drop_action_*' counters to odp-execute.c.
> }
>
> struct dp_netdev_ct_dump {
> diff --git a/lib/dpif.c b/lib/dpif.c
> index e35f111..21f9f54 100644
> --- a/lib/dpif.c
> +++ b/lib/dpif.c
> @@ -1273,6 +1273,7 @@ dpif_execute_helper_cb(void *aux_, struct
> dp_packet_batch *packets_,
> case OVS_ACTION_ATTR_PUSH_NSH:
> case OVS_ACTION_ATTR_POP_NSH:
> case OVS_ACTION_ATTR_CT_CLEAR:
> + case OVS_ACTION_ATTR_DROP:
> case OVS_ACTION_ATTR_UNSPEC:
> case __OVS_ACTION_ATTR_MAX:
> OVS_NOT_REACHED();
> @@ -1295,7 +1296,7 @@ dpif_execute_with_help(struct dpif *dpif, struct
> dpif_execute *execute)
>
> dp_packet_batch_init_packet(&pb, execute->packet);
> odp_execute_actions(&aux, &pb, false, execute->actions,
> - execute->actions_len, dpif_execute_helper_cb);
> + execute->actions_len, dpif_execute_helper_cb, NULL);
> return aux.error;
> }
>
> @@ -1879,6 +1880,12 @@ dpif_supports_tnl_push_pop(const struct dpif *dpif)
> return dpif_is_netdev(dpif);
> }
>
> +bool
> +dpif_supports_explicit_drop_action(const struct dpif *dpif)
> +{
> + return dpif_is_netdev(dpif);
> +}
> +
> /* Meters */
> void
> dpif_meter_get_features(const struct dpif *dpif,
> @@ -1976,3 +1983,4 @@ dpif_meter_del(struct dpif *dpif, ofproto_meter_id
> meter_id,
> }
> return error;
> }
> +
> diff --git a/lib/dpif.h b/lib/dpif.h
> index 475d5a6..e799da8 100644
> --- a/lib/dpif.h
> +++ b/lib/dpif.h
> @@ -888,6 +888,9 @@ int dpif_get_pmds_for_port(const struct dpif * dpif,
> odp_port_t port_no,
>
> char *dpif_get_dp_version(const struct dpif *);
> bool dpif_supports_tnl_push_pop(const struct dpif *);
> +bool dpif_supports_explicit_drop_action(const struct dpif *);
> +int dpif_show_drop_stats_support(struct dpif *dpif, bool detail,
> + struct ds *reply);
>
> /* Log functions. */
> struct vlog_module;
> diff --git a/lib/netdev-dpdk.c b/lib/netdev-dpdk.c
> index 320422b..0c3bc9c 100644
> --- a/lib/netdev-dpdk.c
> +++ b/lib/netdev-dpdk.c
> @@ -2395,6 +2395,10 @@ netdev_dpdk_send__(struct netdev_dpdk *dev, int qid,
> bool concurrent_txq)
> {
> if (OVS_UNLIKELY(!(dev->flags & NETDEV_UP))) {
> + int batch_cnt = dp_packet_batch_size(batch);
> + rte_spinlock_lock(&dev->stats_lock);
> + dev->stats.tx_dropped += batch_cnt;
> + rte_spinlock_unlock(&dev->stats_lock);
This is the unrelated bug fix that should be sent as a separate patch.
In this case we'll be able to apply it and backport to previous versions.
Adding Ian to CC.
> dp_packet_delete_batch(batch, true);
> return;
> }
> diff --git a/lib/odp-execute.c b/lib/odp-execute.c
> index 3b6890e..c0db93f 100644
> --- a/lib/odp-execute.c
> +++ b/lib/odp-execute.c
> @@ -25,6 +25,7 @@
> #include <stdlib.h>
> #include <string.h>
>
> +#include "coverage.h"
> #include "dp-packet.h"
> #include "dpif.h"
> #include "netlink.h"
> @@ -37,6 +38,10 @@
> #include "csum.h"
> #include "conntrack.h"
>
> +COVERAGE_DEFINE(dp_sample_error_drop);
> +COVERAGE_DEFINE(dp_nsh_decap_error_drop);
> +
> +
> /* Masked copy of an ethernet address. 'src' is already properly masked. */
> static void
> ether_addr_copy_masked(struct eth_addr *dst, const struct eth_addr src,
> @@ -575,7 +580,9 @@ odp_execute_masked_set_action(struct dp_packet *packet,
> static void
> odp_execute_sample(void *dp, struct dp_packet *packet, bool steal,
> const struct nlattr *action,
> - odp_execute_cb dp_execute_action)
> + odp_execute_cb dp_execute_action,
> + odp_update_drop_action_counter_cb
> + dp_update_drop_action_counter)
> {
> const struct nlattr *subactions = NULL;
> const struct nlattr *a;
> @@ -589,6 +596,7 @@ odp_execute_sample(void *dp, struct dp_packet *packet,
> bool steal,
> case OVS_SAMPLE_ATTR_PROBABILITY:
> if (random_uint32() >= nl_attr_get_u32(a)) {
> if (steal) {
> + COVERAGE_ADD(dp_sample_error_drop, 1);
> dp_packet_delete(packet);
> }
> return;
> @@ -616,13 +624,16 @@ odp_execute_sample(void *dp, struct dp_packet *packet,
> bool steal,
> }
> dp_packet_batch_init_packet(&pb, packet);
> odp_execute_actions(dp, &pb, true, nl_attr_get(subactions),
> - nl_attr_get_size(subactions), dp_execute_action);
> + nl_attr_get_size(subactions), dp_execute_action,
> + dp_update_drop_action_counter);
> }
>
> static void
> odp_execute_clone(void *dp, struct dp_packet_batch *batch, bool steal,
> const struct nlattr *actions,
> - odp_execute_cb dp_execute_action)
> + odp_execute_cb dp_execute_action,
> + odp_update_drop_action_counter_cb
> + dp_update_drop_action_counter)
> {
> if (!steal) {
> /* The 'actions' may modify the packet, but the modification
> @@ -634,11 +645,12 @@ odp_execute_clone(void *dp, struct dp_packet_batch
> *batch, bool steal,
> dp_packet_batch_clone(&clone_pkt_batch, batch);
> dp_packet_batch_reset_cutlen(batch);
> odp_execute_actions(dp, &clone_pkt_batch, true, nl_attr_get(actions),
> - nl_attr_get_size(actions), dp_execute_action);
> - }
> - else {
> + nl_attr_get_size(actions), dp_execute_action,
> + dp_update_drop_action_counter);
> + } else {
> odp_execute_actions(dp, batch, true, nl_attr_get(actions),
> - nl_attr_get_size(actions), dp_execute_action);
> + nl_attr_get_size(actions), dp_execute_action,
> + dp_update_drop_action_counter);
> }
> }
>
> @@ -673,6 +685,7 @@ requires_datapath_assistance(const struct nlattr *a)
> case OVS_ACTION_ATTR_PUSH_NSH:
> case OVS_ACTION_ATTR_POP_NSH:
> case OVS_ACTION_ATTR_CT_CLEAR:
> + case OVS_ACTION_ATTR_DROP:
> return false;
>
> case OVS_ACTION_ATTR_UNSPEC:
> @@ -699,12 +712,15 @@ requires_datapath_assistance(const struct nlattr *a)
> void
> odp_execute_actions(void *dp, struct dp_packet_batch *batch, bool steal,
> const struct nlattr *actions, size_t actions_len,
> - odp_execute_cb dp_execute_action)
> + odp_execute_cb dp_execute_action,
> + odp_update_drop_action_counter_cb
> + dp_update_drop_action_counter)
> {
> struct dp_packet *packet;
> const struct nlattr *a;
> unsigned int left;
>
> +
No need to add an empty line.
> NL_ATTR_FOR_EACH_UNSAFE (a, left, actions, actions_len) {
> int type = nl_attr_type(a);
> bool last_action = (left <= NLA_ALIGN(a->nla_len));
> @@ -822,7 +838,8 @@ odp_execute_actions(void *dp, struct dp_packet_batch
> *batch, bool steal,
> case OVS_ACTION_ATTR_SAMPLE:
> DP_PACKET_BATCH_FOR_EACH (i, packet, batch) {
> odp_execute_sample(dp, packet, steal && last_action, a,
> - dp_execute_action);
> + dp_execute_action,
> + dp_update_drop_action_counter);
> }
>
> if (last_action) {
> @@ -845,7 +862,8 @@ odp_execute_actions(void *dp, struct dp_packet_batch
> *batch, bool steal,
>
> case OVS_ACTION_ATTR_CLONE:
> odp_execute_clone(dp, batch, steal && last_action, a,
> - dp_execute_action);
> + dp_execute_action,
> + dp_update_drop_action_counter);
> if (last_action) {
> /* We do not need to free the packets. odp_execute_clone()
> has
> * stolen them. */
> @@ -889,6 +907,7 @@ odp_execute_actions(void *dp, struct dp_packet_batch
> *batch, bool steal,
> if (pop_nsh(packet)) {
> dp_packet_batch_refill(batch, packet, i);
> } else {
> + COVERAGE_INC(dp_nsh_decap_error_drop);
> dp_packet_delete(packet);
> }
> }
> @@ -900,6 +919,17 @@ odp_execute_actions(void *dp, struct dp_packet_batch
> *batch, bool steal,
> }
> break;
>
> + case OVS_ACTION_ATTR_DROP: {
> + const struct ovs_action_drop *drop_action = nl_attr_get(a);
> + enum ovs_drop_reason drop_reason = drop_action->drop_reason;
> + if ((drop_reason < OVS_DROP_REASON_MAX) &&
> + dp_update_drop_action_counter) {
> + dp_update_drop_action_counter(drop_reason, batch->count);
> + }
> + dp_packet_delete_batch(batch, steal);
> + return;
> + }
> +
> case OVS_ACTION_ATTR_OUTPUT:
> case OVS_ACTION_ATTR_TUNNEL_PUSH:
> case OVS_ACTION_ATTR_TUNNEL_POP:
> diff --git a/lib/odp-execute.h b/lib/odp-execute.h
> index a3578a5..a84ecbb 100644
> --- a/lib/odp-execute.h
> +++ b/lib/odp-execute.h
> @@ -22,6 +22,8 @@
> #include <stddef.h>
> #include <stdint.h>
> #include "openvswitch/types.h"
> +#include "ovs-atomic.h"
> +#include "dpif.h"
>
> struct nlattr;
> struct dp_packet;
> @@ -31,6 +33,10 @@ struct dp_packet_batch;
> typedef void (*odp_execute_cb)(void *dp, struct dp_packet_batch *batch,
> const struct nlattr *action, bool
> should_steal);
>
> +typedef void (*odp_update_drop_action_counter_cb) (
> + enum ovs_drop_reason drop_reason,
> + int delta);
> +
> /* Actions that need to be executed in the context of a datapath are handed
> * to 'dp_execute_action', if non-NULL. Currently this is called only for
> * actions OVS_ACTION_ATTR_OUTPUT and OVS_ACTION_ATTR_USERSPACE so
> @@ -38,5 +44,7 @@ typedef void (*odp_execute_cb)(void *dp, struct
> dp_packet_batch *batch,
> void odp_execute_actions(void *dp, struct dp_packet_batch *batch,
> bool steal,
> const struct nlattr *actions, size_t actions_len,
> - odp_execute_cb dp_execute_action);
> + odp_execute_cb dp_execute_action,
> + odp_update_drop_action_counter_cb
> + dp_update_drop_action_counter_cb);
> #endif
> diff --git a/lib/odp-util.c b/lib/odp-util.c
> index 0491bed..9754fd5 100644
> --- a/lib/odp-util.c
> +++ b/lib/odp-util.c
> @@ -131,6 +131,7 @@ odp_action_len(uint16_t type)
> case OVS_ACTION_ATTR_CLONE: return ATTR_LEN_VARIABLE;
> case OVS_ACTION_ATTR_PUSH_NSH: return ATTR_LEN_VARIABLE;
> case OVS_ACTION_ATTR_POP_NSH: return 0;
> + case OVS_ACTION_ATTR_DROP: return sizeof(struct ovs_action_drop);
>
> case OVS_ACTION_ATTR_UNSPEC:
> case __OVS_ACTION_ATTR_MAX:
> @@ -345,6 +346,49 @@ format_nsh_key_mask(struct ds *ds, const struct
> ovs_key_nsh *key,
> }
> }
>
> +static const char *
> +dropreason_str(enum ovs_drop_reason reason)
> +{
> + switch (reason) {
> + case OVS_DROP_REASON_OF_PIPELINE:
> + return "pipeline-drop";
> + case OVS_DROP_REASON_BRIDGE_NOT_FOUND:
> + return "bridge not found";
> + case OVS_DROP_REASON_RECURSION_TOO_DEEP:
> + return "recursion too deep";
> + case OVS_DROP_REASON_TOO_MANY_RESUBMITS:
> + return "too many resubmits";
> + case OVS_DROP_REASON_STACK_TOO_DEEP:
> + return "stack too deep";
> + case OVS_DROP_REASON_NO_RECIRCULATION_CONTEXT:
> + return "no recirculation context";
> + case OVS_DROP_REASON_RECIRCULATION_CONFLICT:
> + return "recirculation conflict";
> + case OVS_DROP_REASON_TOO_MANY_MPLS_LABELS:
> + return "too many mpls labels";
> + case OVS_DROP_REASON_INVALID_TUNNEL_METADATA:
> + return "invalid tunnel metadata";
> + case OVS_DROP_REASON_UNSUPPORTED_PACKET_TYPE:
> + return "unsupported packet type";
> + case OVS_DROP_REASON_CONGESTION:
> + return "ecn mismatch at tunnel decapsulation";
> + case OVS_DROP_REASON_FORWARDING_DISABLED:
> + return "forwarding disabled (stp/rstp)";
> + case OVS_DROP_REASON_MAX:
> + default:
> + return "unknown reason";
> + }
> + return "unknown reason";
> +}
> +
> +static void
> +format_odp_drop_action(struct ds *ds,
> + const struct ovs_action_drop *drop_action)
> +{
> + ds_put_format(ds, "drop:%s",
> + dropreason_str(drop_action->drop_reason));
As I already wrote for the previous versions it'll be good to add
some visible markers around the drop reason. Spaces in flow dumps
are usually not used and it could add some complications for the
parsing scripts. You may use brackets like "drop(stack too deep)"
or replace the spaces with '-' or '_' like "drop:stack_too_deep".
> +}
> +
> static void
> format_odp_push_nsh_action(struct ds *ds,
> const struct nsh_hdr *nsh_hdr)
> @@ -1181,6 +1225,9 @@ format_odp_action(struct ds *ds, const struct nlattr *a,
> case OVS_ACTION_ATTR_POP_NSH:
> ds_put_cstr(ds, "pop_nsh()");
> break;
> + case OVS_ACTION_ATTR_DROP:
> + format_odp_drop_action(ds, nl_attr_get(a));
> + break;
> case OVS_ACTION_ATTR_UNSPEC:
> case __OVS_ACTION_ATTR_MAX:
> default:
> @@ -2427,8 +2474,13 @@ odp_actions_from_string(const char *s, const struct
> simap *port_names,
> struct ofpbuf *actions)
> {
> size_t old_size;
> + struct ovs_action_drop drop_action;
>
> - if (!strcasecmp(s, "drop")) {
> + if ((!strcasecmp(s, "drop") ||
> + !strcasecmp(s, "drop:pipeline-drop"))) {
Why you're adding parse only for this type of drop action ?
> + drop_action.drop_reason = OVS_DROP_REASON_OF_PIPELINE;
> + nl_msg_put_unspec(actions, OVS_ACTION_ATTR_DROP,
> + &drop_action, sizeof drop_action);
> return 0;
> }
>
> diff --git a/ofproto/ofproto-dpif-ipfix.c b/ofproto/ofproto-dpif-ipfix.c
> index 4029806..1d23a5a 100644
> --- a/ofproto/ofproto-dpif-ipfix.c
> +++ b/ofproto/ofproto-dpif-ipfix.c
> @@ -3015,6 +3015,7 @@ dpif_ipfix_read_actions(const struct flow *flow,
> case OVS_ACTION_ATTR_PUSH_NSH:
> case OVS_ACTION_ATTR_POP_NSH:
> case OVS_ACTION_ATTR_UNSPEC:
> + case OVS_ACTION_ATTR_DROP:
> case __OVS_ACTION_ATTR_MAX:
> default:
> break;
> diff --git a/ofproto/ofproto-dpif-sflow.c b/ofproto/ofproto-dpif-sflow.c
> index 7da3175..69ed7b8 100644
> --- a/ofproto/ofproto-dpif-sflow.c
> +++ b/ofproto/ofproto-dpif-sflow.c
> @@ -1222,6 +1222,7 @@ dpif_sflow_read_actions(const struct flow *flow,
> case OVS_ACTION_ATTR_PUSH_NSH:
> case OVS_ACTION_ATTR_POP_NSH:
> case OVS_ACTION_ATTR_UNSPEC:
> + case OVS_ACTION_ATTR_DROP:
> case __OVS_ACTION_ATTR_MAX:
> default:
> break;
> diff --git a/ofproto/ofproto-dpif-upcall.c b/ofproto/ofproto-dpif-upcall.c
> index dc30824..b1ce2ec 100644
> --- a/ofproto/ofproto-dpif-upcall.c
> +++ b/ofproto/ofproto-dpif-upcall.c
> @@ -1154,7 +1154,7 @@ upcall_receive(struct upcall *upcall, const struct
> dpif_backer *backer,
> return 0;
> }
>
> -static void
> +static enum xlate_error
> upcall_xlate(struct udpif *udpif, struct upcall *upcall,
> struct ofpbuf *odp_actions, struct flow_wildcards *wc)
> {
> @@ -1244,6 +1244,7 @@ upcall_xlate(struct udpif *udpif, struct upcall *upcall,
> if (upcall->type == MISS_UPCALL) {
> upcall->ukey = ukey_create_from_upcall(upcall, wc);
> }
> + return xerr;
The only user of this function is 'process_upcall' and it does not check
the result.
> }
>
> static void
> diff --git a/ofproto/ofproto-dpif-xlate.c b/ofproto/ofproto-dpif-xlate.c
> index 839fddd..32c4edf 100644
> --- a/ofproto/ofproto-dpif-xlate.c
> +++ b/ofproto/ofproto-dpif-xlate.c
> @@ -444,10 +444,46 @@ const char *xlate_strerror(enum xlate_error error)
> return "Invalid tunnel metadata";
> case XLATE_UNSUPPORTED_PACKET_TYPE:
> return "Unsupported packet type";
> + case XLATE_CONGESTION_DROP:
> + return "CONGESTION DROP";
> + case XLATE_FORWARDING_DISABLED:
> + return "Forwarding is disabled";
> +
> }
> return "Unknown error";
> }
>
> +enum ovs_drop_reason xlate_error_to_drop_reason(enum xlate_error error)
> +{
> + switch (error) {
> + case XLATE_OK:
> + return OVS_DROP_REASON_OF_PIPELINE;
> + case XLATE_BRIDGE_NOT_FOUND:
> + return OVS_DROP_REASON_BRIDGE_NOT_FOUND;
> + case XLATE_RECURSION_TOO_DEEP:
> + return OVS_DROP_REASON_RECURSION_TOO_DEEP;
> + case XLATE_TOO_MANY_RESUBMITS:
> + return OVS_DROP_REASON_TOO_MANY_RESUBMITS;
> + case XLATE_STACK_TOO_DEEP:
> + return OVS_DROP_REASON_STACK_TOO_DEEP;
> + case XLATE_NO_RECIRCULATION_CONTEXT:
> + return OVS_DROP_REASON_NO_RECIRCULATION_CONTEXT;
> + case XLATE_RECIRCULATION_CONFLICT:
> + return OVS_DROP_REASON_RECIRCULATION_CONFLICT;
> + case XLATE_TOO_MANY_MPLS_LABELS:
> + return OVS_DROP_REASON_TOO_MANY_MPLS_LABELS;
> + case XLATE_INVALID_TUNNEL_METADATA:
> + return OVS_DROP_REASON_INVALID_TUNNEL_METADATA;
> + case XLATE_UNSUPPORTED_PACKET_TYPE:
> + return OVS_DROP_REASON_UNSUPPORTED_PACKET_TYPE;
> + case XLATE_CONGESTION_DROP:
> + return OVS_DROP_REASON_CONGESTION;
> + case XLATE_FORWARDING_DISABLED:
> + return OVS_DROP_REASON_MAX;
> + }
> + return OVS_DROP_REASON_OF_PIPELINE;
> +}
> +
> static void xlate_action_set(struct xlate_ctx *ctx);
> static void xlate_commit_actions(struct xlate_ctx *ctx);
>
> @@ -5921,6 +5957,17 @@ put_ct_label(const struct flow *flow, struct ofpbuf
> *odp_actions,
> }
>
> static void
> +put_drop_action(struct ofpbuf *odp_actions, enum xlate_error error)
> +{
> + struct ovs_action_drop drop_action;
> +
> + drop_action.drop_reason = xlate_error_to_drop_reason(error);
> + nl_msg_put_unspec(odp_actions, OVS_ACTION_ATTR_DROP,
> + &drop_action, sizeof drop_action);
Indents are a bit off.
> +
> +}
> +
> +static void
> put_ct_helper(struct xlate_ctx *ctx,
> struct ofpbuf *odp_actions, struct ofpact_conntrack *ofc)
> {
> @@ -7383,6 +7430,10 @@ xlate_actions(struct xlate_in *xin, struct xlate_out
> *xout)
> }
> size_t sample_actions_len = ctx.odp_actions->size;
>
> + if (!tnl_process_ecn(flow)) {
> + ctx.error = XLATE_CONGESTION_DROP;
> + }
> +
> if (tnl_process_ecn(flow)
> && (!in_port || may_receive(in_port, &ctx))) {
> const struct ofpact *ofpacts;
> @@ -7415,6 +7466,7 @@ xlate_actions(struct xlate_in *xin, struct xlate_out
> *xout)
> ctx.odp_actions->size = sample_actions_len;
> ctx_cancel_freeze(&ctx);
> ofpbuf_clear(&ctx.action_set);
> + ctx.error = XLATE_FORWARDING_DISABLED;
> }
>
> if (!ctx.freezing) {
> @@ -7522,6 +7574,18 @@ exit:
> ofpbuf_clear(xin->odp_actions);
> }
> }
> +
> + /*
> + * If we are going to install "drop" action, check whether
> + * datapath supports explicit "drop"action. If datapath
> + * supports explicit "drop"action then install the "drop"
> + * action containing the drop reason.
> + */
> + if (xin->odp_actions && !xin->odp_actions->size &&
> + ovs_explicit_drop_action_supported(ctx.xbridge->ofproto)) {
> + put_drop_action(xin->odp_actions, ctx.error);
> + }
> +
> return ctx.error;
> }
>
> diff --git a/ofproto/ofproto-dpif-xlate.h b/ofproto/ofproto-dpif-xlate.h
> index 0a5a528..bb3f9ff 100644
> --- a/ofproto/ofproto-dpif-xlate.h
> +++ b/ofproto/ofproto-dpif-xlate.h
> @@ -216,12 +216,16 @@ enum xlate_error {
> XLATE_TOO_MANY_MPLS_LABELS,
> XLATE_INVALID_TUNNEL_METADATA,
> XLATE_UNSUPPORTED_PACKET_TYPE,
> + XLATE_CONGESTION_DROP,
> + XLATE_FORWARDING_DISABLED,
> };
>
> const char *xlate_strerror(enum xlate_error error);
>
> enum xlate_error xlate_actions(struct xlate_in *, struct xlate_out *);
>
> +enum ovs_drop_reason xlate_error_to_drop_reason(enum xlate_error error);
> +
> void xlate_in_init(struct xlate_in *, struct ofproto_dpif *, ovs_version_t,
> const struct flow *, ofp_port_t in_port, struct rule_dpif
> *,
> uint16_t tcp_flags, const struct dp_packet *packet,
> diff --git a/ofproto/ofproto-dpif.c b/ofproto/ofproto-dpif.c
> index 14fe6fc..609226a 100644
> --- a/ofproto/ofproto-dpif.c
> +++ b/ofproto/ofproto-dpif.c
> @@ -827,6 +827,12 @@ ovs_native_tunneling_is_on(struct ofproto_dpif *ofproto)
> && atomic_count_get(&ofproto->backer->tnl_count);
> }
>
> +bool
> +ovs_explicit_drop_action_supported(struct ofproto_dpif *ofproto)
> +{
> + return ofproto->backer->rt_support.explicit_drop_action;
> +}
> +
> /* Tests whether 'backer''s datapath supports recirculation. Only newer
> * datapaths support OVS_KEY_ATTR_RECIRC_ID in keys. We need to disable some
> * features on older datapaths that don't support this feature.
> @@ -1397,6 +1403,8 @@ check_support(struct dpif_backer *backer)
> backer->rt_support.ct_eventmask = check_ct_eventmask(backer);
> backer->rt_support.ct_clear = check_ct_clear(backer);
> backer->rt_support.max_hash_alg = check_max_dp_hash_alg(backer);
> + backer->rt_support.explicit_drop_action =
> + dpif_supports_explicit_drop_action(backer->dpif);
>
> /* Flow fields. */
> backer->rt_support.odp.ct_state = check_ct_state(backer);
> @@ -5776,6 +5784,7 @@ ofproto_unixctl_dpif_set_dp_features(struct
> unixctl_conn *conn,
> ds_destroy(&ds);
> }
>
> +
No need to add an empty line.
> static void
> ofproto_unixctl_init(void)
> {
> @@ -5809,7 +5818,7 @@ ofproto_unixctl_init(void)
> unixctl_command_register("dpif/set-dp-features", "bridge", 1, 3 ,
> ofproto_unixctl_dpif_set_dp_features, NULL);
> }
> -
> +
Don't drop the ^L character.
> static odp_port_t
> ofp_port_to_odp_port(const struct ofproto_dpif *ofproto, ofp_port_t ofp_port)
> {
> diff --git a/ofproto/ofproto-dpif.h b/ofproto/ofproto-dpif.h
> index 1a404c8..9162ba0 100644
> --- a/ofproto/ofproto-dpif.h
> +++ b/ofproto/ofproto-dpif.h
> @@ -106,6 +106,7 @@ struct rule_dpif *rule_dpif_lookup_from_table(struct
> ofproto_dpif *,
> bool honor_table_miss,
> struct xlate_cache *);
>
> +
No need to add an empty line.
> void rule_dpif_credit_stats(struct rule_dpif *,
> const struct dpif_flow_stats *);
>
> @@ -192,7 +193,10 @@ struct group_dpif *group_dpif_lookup(struct ofproto_dpif
> *,
> DPIF_SUPPORT_FIELD(bool, ct_clear, "Conntrack clear") \
> \
> /* Highest supported dp_hash algorithm. */ \
> - DPIF_SUPPORT_FIELD(size_t, max_hash_alg, "Max dp_hash algorithm")
> + DPIF_SUPPORT_FIELD(size_t, max_hash_alg, "Max dp_hash algorithm") \
> + \
> + /* True if the datapath supports explicit drop action. */ \
> + DPIF_SUPPORT_FIELD(bool, explicit_drop_action, "Explicit Drop action")
>
> /* Stores the various features which the corresponding backer supports. */
> struct dpif_backer_support {
> @@ -361,4 +365,6 @@ int ofproto_dpif_delete_internal_flow(struct ofproto_dpif
> *, struct match *,
>
> bool ovs_native_tunneling_is_on(struct ofproto_dpif *);
>
> +bool ovs_explicit_drop_action_supported(struct ofproto_dpif *);
> +
> #endif /* ofproto-dpif.h */
> diff --git a/tests/automake.mk b/tests/automake.mk
> index 92d56b2..22b4722 100644
> --- a/tests/automake.mk
> +++ b/tests/automake.mk
> @@ -108,7 +108,8 @@ TESTSUITE_AT = \
> tests/ovn-controller-vtep.at \
> tests/mcast-snooping.at \
> tests/packet-type-aware.at \
> - tests/nsh.at
> + tests/nsh.at \
> + tests/drop-stats.at
automake.mk uses tabs.
>
> EXTRA_DIST += $(FUZZ_REGRESSION_TESTS)
> FUZZ_REGRESSION_TESTS = \
> diff --git a/tests/bundle.at b/tests/bundle.at
> index 0a4eadc..33fe249 100644
> --- a/tests/bundle.at
> +++ b/tests/bundle.at
> @@ -241,7 +241,7 @@ AT_CHECK([tail -1 stdout], [0],
> AT_CHECK([ovs-ofctl mod-port br0 p2 down])
> AT_CHECK([ovs-appctl ofproto/trace br0
> 'in_port=LOCAL,dl_src=50:54:00:00:00:05,dl_dst=50:54:00:00:00:06'], [0],
> [stdout])
> AT_CHECK([tail -1 stdout], [0],
> - [Datapath actions: drop
> + [Datapath actions: drop:pipeline-drop
> ])
> AT_CHECK([ovs-ofctl mod-port br0 p1 up])
> AT_CHECK([ovs-appctl ofproto/trace br0
> 'in_port=LOCAL,dl_src=50:54:00:00:00:05,dl_dst=50:54:00:00:00:06'], [0],
> [stdout])
> diff --git a/tests/classifier.at b/tests/classifier.at
> index 86f872d..a7378a7 100644
> --- a/tests/classifier.at
> +++ b/tests/classifier.at
> @@ -50,12 +50,12 @@ Datapath actions: 1
> AT_CHECK([ovs-appctl ofproto/trace br0
> 'in_port=1,dl_src=50:54:00:00:00:05,dl_dst=50:54:00:00:00:07,dl_type=0x0800,nw_src=192.168.0.1,nw_dst=11.1.2.15,nw_proto=6,nw_tos=0,nw_ttl=128,tp_src=8,tp_dst=80'],
> [0], [stdout])
> AT_CHECK([tail -2 stdout], [0],
> [Megaflow: recirc_id=0,eth,ip,in_port=1,nw_dst=11.0.0.0/8,nw_frag=no
> -Datapath actions: drop
> +Datapath actions: drop:pipeline-drop
> ])
> AT_CHECK([ovs-appctl ofproto/trace br0
> 'in_port=1,dl_src=50:54:00:00:00:05,dl_dst=50:54:00:00:00:07,dl_type=0x0800,nw_src=192.168.0.1,nw_dst=10.1.2.15,nw_proto=6,nw_tos=0,nw_ttl=128,tp_src=8,tp_dst=80'],
> [0], [stdout])
> AT_CHECK([tail -2 stdout], [0],
> [Megaflow:
> recirc_id=0,eth,tcp,in_port=1,nw_dst=10.1.2.15,nw_frag=no,tp_dst=80
> -Datapath actions: drop
> +Datapath actions: drop:pipeline-drop
> ])
> AT_CHECK([ovs-appctl ofproto/trace br0
> 'in_port=1,dl_src=50:54:00:00:00:05,dl_dst=50:54:00:00:00:07,dl_type=0x0800,nw_src=192.168.0.1,nw_dst=10.1.2.15,nw_proto=6,nw_tos=0,nw_ttl=128,tp_src=8,tp_dst=79'],
> [0], [stdout])
> AT_CHECK([tail -2 stdout], [0],
> @@ -88,7 +88,7 @@ AT_CHECK([ovs-ofctl add-flows br0 flows.txt])
> AT_CHECK([ovs-appctl ofproto/trace br0
> 'in_port=1,dl_src=50:54:00:00:00:05,dl_dst=50:54:00:00:00:07,dl_type=0x0800,nw_src=192.168.0.1,nw_dst=192.168.0.2,nw_proto=6,nw_tos=0,nw_ttl=128,tp_src=8,tp_dst=80'],
> [0], [stdout])
> AT_CHECK([tail -2 stdout], [0],
> [Megaflow: recirc_id=0,eth,tcp,in_port=1,nw_dst=192.168.0.0/16,nw_frag=no
> -Datapath actions: drop
> +Datapath actions: drop:pipeline-drop
> ])
>
> AT_CHECK([ovs-vsctl set Flow_Table t0 prefixes=ipv6_label], [0])
> @@ -103,7 +103,7 @@ AT_CHECK([ovs-vsctl set Flow_Table t0 prefixes=nw_dst],
> [0])
> AT_CHECK([ovs-appctl ofproto/trace br0
> 'in_port=1,dl_src=50:54:00:00:00:05,dl_dst=50:54:00:00:00:07,dl_type=0x0800,nw_src=192.168.0.1,nw_dst=192.168.0.2,nw_proto=6,nw_tos=0,nw_ttl=128,tp_src=8,tp_dst=80'],
> [0], [stdout])
> AT_CHECK([tail -2 stdout], [0],
> [Megaflow: recirc_id=0,eth,tcp,in_port=1,nw_dst=192.168.0.0/16,nw_frag=no
> -Datapath actions: drop
> +Datapath actions: drop:pipeline-drop
> ])
> AT_CHECK([ovs-appctl ofproto/trace br0
> 'in_port=2,dl_src=50:54:00:00:00:05,dl_dst=50:54:00:00:00:07,dl_type=0x0800,nw_src=192.168.0.1,nw_dst=192.168.0.2,nw_proto=6,nw_tos=0,nw_ttl=128,tp_src=8,tp_dst=80'],
> [0], [stdout])
> AT_CHECK([tail -2 stdout], [0],
> @@ -113,7 +113,7 @@ Datapath actions: 1
> AT_CHECK([ovs-appctl ofproto/trace br0
> 'in_port=1,dl_src=50:54:00:00:00:05,dl_dst=50:54:00:00:00:07,dl_type=0x0800,nw_src=192.168.0.1,nw_dst=10.1.2.15,nw_proto=6,nw_tos=0,nw_ttl=128,tp_src=8,tp_dst=80'],
> [0], [stdout])
> AT_CHECK([tail -2 stdout], [0],
> [Megaflow:
> recirc_id=0,eth,tcp,in_port=1,nw_dst=10.1.2.15,nw_frag=no,tp_dst=80
> -Datapath actions: drop
> +Datapath actions: drop:pipeline-drop
> ])
> AT_CHECK([ovs-appctl ofproto/trace br0
> 'in_port=1,dl_src=50:54:00:00:00:05,dl_dst=50:54:00:00:00:07,dl_type=0x0800,nw_src=192.168.0.1,nw_dst=10.1.2.15,nw_proto=6,nw_tos=0,nw_ttl=128,tp_src=8,tp_dst=79'],
> [0], [stdout])
> AT_CHECK([tail -2 stdout], [0],
> diff --git a/tests/dpif-netdev.at b/tests/dpif-netdev.at
> index 6915d43..08349f6 100644
> --- a/tests/dpif-netdev.at
> +++ b/tests/dpif-netdev.at
> @@ -1,4 +1,4 @@
> -AT_BANNER([dpif-netdev])
> +T_BANNER([dpif-netdev])
Something wrong here.
>
> m4_divert_push([PREPARE_TESTS])
> [
> @@ -281,6 +281,7 @@ type=drop rate=1 burst_size=2
> ])
>
> ovs-appctl time/warp 5000
> +sleep 10
Do we really need to sleep for so long ?
> AT_CHECK([ovs-appctl netdev-dummy/receive p7
> 'in_port(7),packet_type(ns=0,id=0),eth(src=50:54:00:00:00:09,dst=50:54:00:00:00:0a),eth_type(0x0800),ipv4(src=10.0.0.2,dst=10.0.0.1,proto=1,tos=0,ttl=64,frag=no),icmp(type=8,code=0)'
> --len 60])
> AT_CHECK([ovs-appctl netdev-dummy/receive p7
> 'in_port(7),packet_type(ns=0,id=0),eth(src=50:54:00:00:00:09,dst=50:54:00:00:00:0a),eth_type(0x0800),ipv4(src=10.0.0.2,dst=10.0.0.1,proto=1,tos=0,ttl=64,frag=no),icmp(type=8,code=0)'
> --len 60])
> AT_CHECK([ovs-appctl netdev-dummy/receive p7
> 'in_port(7),packet_type(ns=0,id=0),eth(src=50:54:00:00:00:09,dst=50:54:00:00:00:0a),eth_type(0x0800),ipv4(src=10.0.0.2,dst=10.0.0.1,proto=1,tos=0,ttl=64,frag=no),icmp(type=8,code=0)'
> --len 60])
> @@ -291,7 +292,7 @@ AT_CHECK([ovs-appctl netdev-dummy/receive p8
> 'in_port(8),packet_type(ns=0,id=0),
> AT_CHECK([ovs-appctl netdev-dummy/receive p8
> 'in_port(8),packet_type(ns=0,id=0),eth(src=50:54:00:00:00:0b,dst=50:54:00:00:00:0c),eth_type(0x0800),ipv4(src=10.0.0.3,dst=10.0.0.4,proto=1,tos=0,ttl=64,frag=no),icmp(type=8,code=0)'
> --len 60])
> AT_CHECK([ovs-appctl netdev-dummy/receive p8
> 'in_port(8),packet_type(ns=0,id=0),eth(src=50:54:00:00:00:0b,dst=50:54:00:00:00:0c),eth_type(0x0800),ipv4(src=10.0.0.3,dst=10.0.0.4,proto=1,tos=0,ttl=64,frag=no),icmp(type=8,code=0)'
> --len 60])
> AT_CHECK([ovs-appctl netdev-dummy/receive p8
> 'in_port(8),packet_type(ns=0,id=0),eth(src=50:54:00:00:00:0b,dst=50:54:00:00:00:0c),eth_type(0x0800),ipv4(src=10.0.0.3,dst=10.0.0.4,proto=1,tos=0,ttl=64,frag=no),icmp(type=8,code=0)'
> --len 60])
> -sleep 1 # wait for forwarders process packets
> +# wait for forwarders process packets
Something wrong here too.
>
> # Meter 1 is measuring packets, allowing one packet per second with
> # bursts of one packet, so 4 out of 5 packets should hit the drop
> @@ -320,7 +321,8 @@ AT_CHECK([ovs-appctl netdev-dummy/receive p8
> 'in_port(8),packet_type(ns=0,id=0),
> AT_CHECK([ovs-appctl netdev-dummy/receive p8
> 'in_port(8),packet_type(ns=0,id=0),eth(src=50:54:00:00:00:0b,dst=50:54:00:00:00:0c),eth_type(0x0800),ipv4(src=10.0.0.3,dst=10.0.0.4,proto=1,tos=0,ttl=64,frag=no),icmp(type=8,code=0)'
> --len 60])
> AT_CHECK([ovs-appctl netdev-dummy/receive p8
> 'in_port(8),packet_type(ns=0,id=0),eth(src=50:54:00:00:00:0b,dst=50:54:00:00:00:0c),eth_type(0x0800),ipv4(src=10.0.0.3,dst=10.0.0.4,proto=1,tos=0,ttl=64,frag=no),icmp(type=8,code=0)'
> --len 60])
> AT_CHECK([ovs-appctl netdev-dummy/receive p8
> 'in_port(8),packet_type(ns=0,id=0),eth(src=50:54:00:00:00:0b,dst=50:54:00:00:00:0c),eth_type(0x0800),ipv4(src=10.0.0.3,dst=10.0.0.4,proto=1,tos=0,ttl=64,frag=no),icmp(type=8,code=0)'
> --len 60])
> -sleep 1 # wait for forwarders process packets
> +
> +sleep 10 # wait for forwarders process packets
>
> # Meter 1 is measuring packets, allowing one packet per second with
> # bursts of one packet, so all 5 of the new packets should hit the drop
> @@ -337,6 +339,15 @@ meter:2 flow_count:1 packet_in_count:10
> byte_in_count:600 duration:0.0s bands:
> 0: packet_count:5 byte_count:300
> ])
>
> +ovs-appctl time/warp 5000
> +sleep 10
> +
> +AT_CHECK([
> +ovs-appctl coverage/show | grep "dp_meter_drop" | cut -d':' -f2|sed 's/ //'
> +], [0], [dnl
> +14
> +])
> +
> AT_CHECK([cat ovs-vswitchd.log | filter_flow_install |
> strip_xout_keep_actions], [0], [dnl
>
> recirc_id(0),in_port(1),packet_type(ns=0,id=0),eth_type(0x0800),ipv4(frag=no),
> actions:meter(0),7
>
> recirc_id(0),in_port(2),packet_type(ns=0,id=0),eth_type(0x0800),ipv4(frag=no),
> actions:8
> diff --git a/tests/drop-stats.at b/tests/drop-stats.at
> new file mode 100644
> index 0000000..be8b8ae
> --- /dev/null
> +++ b/tests/drop-stats.at
> @@ -0,0 +1,197 @@
> +AT_BANNER([drop-stats])
> +
> +AT_SETUP([drop-stats - cli tests])
> +
> +OVS_VSWITCHD_START([dnl
> + set bridge br0 datapath_type=dummy \
> + protocols=OpenFlow10,OpenFlow13,OpenFlow14,OpenFlow15 -- \
> + add-port br0 p1 -- set Interface p1 type=dummy ofport_request=1])
> +
> +AT_DATA([flows.txt], [dnl
> +table=0,in_port=1,actions=drop
> +])
> +
> +AT_CHECK([
> + ovs-ofctl del-flows br0
> + ovs-ofctl -Oopenflow13 add-flows br0 flows.txt
> + ovs-ofctl -Oopenflow13 dump-flows br0 | ofctl_strip | sort | grep actions
> +], [0], [dnl
> + in_port=1 actions=drop
> +])
> +
> +AT_CHECK([
> + ovs-appctl netdev-dummy/receive p1
> 1e2ce92a669e3a6dd2099cab0800450000548a53400040011addc0a80a0ac0a80a1e08006f200a4d0001fc509a58000000002715020000000000101112131415161718191a1b1c1d1e1f202122232425262728292a2b2c2d2e2f3031323334353637
> + ovs-appctl netdev-dummy/receive p1
> 1e2ce92a669e3a6dd2099cab0800450000548a53400040011addc0a80a0ac0a80a1e08006f200a4d0001fc509a58000000002715020000000000101112131415161718191a1b1c1d1e1f202122232425262728292a2b2c2d2e2f3031323334353637
> + ovs-appctl netdev-dummy/receive p1
> 1e2ce92a669e3a6dd2099cab0800450000548a53400040011addc0a80a0ac0a80a1e08006f200a4d0001fc509a58000000002715020000000000101112131415161718191a1b1c1d1e1f202122232425262728292a2b2c2d2e2f3031323334353637
> +], [0], [ignore])
> +
> +AT_CHECK([ovs-appctl dpctl/dump-flows | sed
> 's/used:[[0-9]].[[0-9]]*s/used:0.0/' | sort], [0],
> +[flow-dump from non-dpdk interfaces:
> +recirc_id(0),in_port(1),packet_type(ns=0,id=0),eth_type(0x0800),ipv4(frag=no),
> packets:2, bytes:196, used:0.0, actions:drop:pipeline-drop
> +])
> +
> +sleep 1
> +
> +
> +AT_CHECK([
> +ovs-appctl coverage/show | grep "drop_action_of_pipeline" | cut -d':'
> -f2|sed 's/ //'
> +], [0], [dnl
> +3
> +])
> +
> +
> +OVS_VSWITCHD_STOP
> +AT_CLEANUP
> +
> +AT_SETUP([drop-stats - pipeline and recurssion drops])
> +
> +OVS_VSWITCHD_START([dnl
> + set bridge br0 datapath_type=dummy \
> + protocols=OpenFlow10,OpenFlow13,OpenFlow14,OpenFlow15 -- \
> + add-port br0 p1 -- set Interface p1 type=dummy ofport_request=1 -- \
> + add-port br0 p2 -- set Interface p2 type=dummy ofport_request=2])
> +
> +AT_DATA([flows.txt], [dnl
> +table=0,in_port=1,actions=drop
> +])
> +
> +AT_CHECK([
> + ovs-ofctl del-flows br0
> + ovs-ofctl -Oopenflow13 add-flows br0 flows.txt
> + ovs-ofctl -Oopenflow13 dump-flows br0 | ofctl_strip | sort | grep actions
> +], [0], [dnl
> + in_port=1 actions=drop
> +])
> +
> +AT_CHECK([
> + ovs-appctl netdev-dummy/receive p1
> 1e2ce92a669e3a6dd2099cab0800450000548a53400040011addc0a80a0ac0a80a1e08006f200a4d0001fc509a58000000002715020000000000101112131415161718191a1b1c1d1e1f202122232425262728292a2b2c2d2e2f3031323334353637
If possible, it's preferred to use odp format (i.e. declarative) for
packets definition. This allowes user to see what packet is sent without
parsing it by external tools.
Same for all other places.
> +], [0], [ignore])
> +
> +sleep 1
> +
> +
> +AT_CHECK([
> +ovs-appctl coverage/show | grep "drop_action_of_pipeline" | cut -d':'
> -f2|sed 's/ //'
> +], [0], [dnl
> +1
> +])
> +
> +
> +AT_DATA([flows.txt], [dnl
> +table=0, in_port=1, actions=goto_table:1
> +table=1, in_port=1, actions=goto_table:2
> +table=2, in_port=1, actions=resubmit(,1)
> +])
> +
> +AT_CHECK([
> + ovs-ofctl del-flows br0
> + ovs-ofctl -Oopenflow13 add-flows br0 flows.txt
> + ovs-ofctl -Oopenflow13 dump-flows br0 | ofctl_strip | sort | grep actions
> +], [0], [ignore])
> +
> +AT_CHECK([
> + ovs-appctl netdev-dummy/receive p1
> 1e2ce92a669e3a6dd2099cab0800450000548a53400040011addc0a80a0ac0a80a1e08006f200a4d0001fc509a58000000002715020000000000101112131415161718191a1b1c1d1e1f202122232425262728292a2b2c2d2e2f3031323334353637
> +], [0], [ignore])
> +
> +sleep 1
> +
> +AT_CHECK([
> +ovs-appctl coverage/show | grep "drop_action_recursion_too_deep" | cut -d':'
> -f2|sed 's/ //'
> +], [0], [dnl
> +1
> +])
> +
> +
> +OVS_VSWITCHD_STOP(["/|WARN|/d"])
> +AT_CLEANUP
> +
> +AT_SETUP([drop-stats - too many resubmit])
> +
> +OVS_VSWITCHD_START
> +add_of_ports br0 1
> +(for i in `seq 1 64`; do
> + j=`expr $i + 1`
> + echo "in_port=$i, actions=resubmit:$j, resubmit:$j, local"
> + done
> + echo "in_port=65, actions=local") > flows.txt
> +
> +AT_CHECK([
> + ovs-ofctl del-flows br0
> + ovs-ofctl -Oopenflow13 add-flows br0 flows.txt
> +], [0], [ignore])
> +
> +ovs-appctl netdev-dummy/receive p1
> 'in_port(1),eth(src=50:54:00:00:00:09,dst=50:54:00:00:00:0a),eth_type(0x1234)'
> +
> +sleep 1
> +
> +AT_CHECK([
> +ovs-appctl coverage/show | grep "drop_action_too_many_resubmit" | cut -d':'
> -f2|sed 's/ //'
> +], [0], [dnl
> +1
> +])
> +
> +OVS_VSWITCHD_STOP(["/|WARN|/d"])
> +AT_CLEANUP
> +
> +
> +AT_SETUP([drop-stats - stack too deep])
> +OVS_VSWITCHD_START
> +add_of_ports br0 1
> +(for i in `seq 1 12`; do
> + j=`expr $i + 1`
> + echo "in_port=$i, actions=resubmit:$j, resubmit:$j, local"
> + done
> + push="push:NXM_NX_REG0[[]]"
> + echo "in_port=13, actions=$push,$push,$push,$push,$push,$push,$push,$push")
> > flows
> + AT_CHECK([ovs-ofctl add-flows br0 flows])
> +
> +ovs-appctl netdev-dummy/receive p1
> 'in_port(1),eth(src=50:54:00:00:00:09,dst=50:54:00:00:00:0a),eth_type(0x1234)'
> +
> +sleep 1
> +
> +AT_CHECK([
> +ovs-appctl coverage/show | grep "drop_action_stack_too_deep" | cut -d':'
> -f2|sed 's/ //'
> +], [0], [dnl
> +1
> +])
> +
> +
> +OVS_VSWITCHD_STOP(["/resubmits yielded over 64 kB of stack/d"])
> +AT_CLEANUP
> +
> +AT_SETUP([drop-stats - too many mpls labels])
> +
> +OVS_VSWITCHD_START([dnl
> + set bridge br0 datapath_type=dummy \
> + protocols=OpenFlow10,OpenFlow13,OpenFlow14,OpenFlow15 -- \
> + add-port br0 p1 -- set Interface p1 type=dummy ofport_request=1 -- \
> + add-port br0 p2 -- set Interface p2 type=dummy ofport_request=2])
> +
> +AT_DATA([flows.txt], [dnl
> +table=0, in_port=1, actions=push_mpls:0x8847, resubmit:3
> +table=0, in_port=3, actions=push_mpls:0x8847, set_field:10->mpls_label,
> set_field:15->mpls_label, resubmit:4
> +table=0, in_port=4, actions=push_mpls:0x8847, set_field:11->mpls_label,
> resubmit:5
> +table=0, in_port=5, actions=push_mpls:0x8847, set_field:12->mpls_label,
> resubmit:6
> +table=0, in_port=6, actions=push_mpls:0x8847, set_field:13->mpls_label,
> output:2
> +])
> +
> +AT_CHECK([
> + ovs-ofctl del-flows br0
> + ovs-ofctl -Oopenflow13 add-flows br0 flows.txt
> +])
> +
> +AT_CHECK([
> + ovs-appctl netdev-dummy/receive p1
> 1e2ce92a669e3a6dd2099cab0800450000548a53400040011addc0a80a0ac0a80a1e08006f200a4d0001fc509a58000000002715020000000000101112131415161718191a1b1c1d1e1f202122232425262728292a2b2c2d2e2f3031323334353637
> +], [0], [ignore])
> +
> +sleep 1
> +
> +AT_CHECK([
> +ovs-appctl coverage/show | grep "drop_action_too_many_mpls_labels" | cut
> -d':' -f2|sed 's/ //'
> +], [0], [dnl
> +1
> +])
> +
> +
> +OVS_VSWITCHD_STOP(["/|WARN|/d"])
> +AT_CLEANUP
> diff --git a/tests/ofproto-dpif.at b/tests/ofproto-dpif.at
> index ded2ef0..298aa9a 100644
> --- a/tests/ofproto-dpif.at
> +++ b/tests/ofproto-dpif.at
> @@ -3496,51 +3496,51 @@ dnl Each of these specifies an in_port by number, a
> VLAN VID (or "none"),
> dnl a VLAN PCP (used if the VID isn't "none") and the expected set of
> datapath
> dnl actions.
> for tuple in \
> - "100 none 0 drop" \
> - "100 0 0 drop" \
> - "100 0 1 drop" \
> + "100 none 0 drop:pipeline-drop" \
> + "100 0 0 drop:pipeline-drop" \
> + "100 0 1 drop:pipeline-drop" \
> "100 10 0 1,5,6,7,8,pop_vlan,2,9" \
> "100 10 1 1,5,6,7,8,pop_vlan,2,9" \
> "100 11 0 5,7" \
> "100 11 1 5,7" \
> "100 12 0 1,5,6,pop_vlan,3,4,7,8,11,12" \
> "100 12 1 1,5,6,pop_vlan,4,7,11,push_vlan(vid=0,pcp=1),3,8,12" \
> - "1 none 0 drop" \
> - "1 0 0 drop" \
> - "1 0 1 drop" \
> + "1 none 0 drop:pipeline-drop" \
> + "1 0 0 drop:pipeline-drop" \
> + "1 0 1 drop:pipeline-drop" \
> "1 10 0 5,6,7,8,100,pop_vlan,2,9" \
> "1 10 1 5,6,7,8,100,pop_vlan,2,9" \
> - "1 11 0 drop" \
> - "1 11 1 drop" \
> + "1 11 0 drop:pipeline-drop" \
> + "1 11 1 drop:pipeline-drop" \
> "1 12 0 5,6,100,pop_vlan,3,4,7,8,11,12" \
> "1 12 1 5,6,100,pop_vlan,4,7,11,push_vlan(vid=0,pcp=1),3,8,12" \
> "2 none 0 9,push_vlan(vid=10,pcp=0),1,5,6,7,8,100" \
> "2 0 0 pop_vlan,9,push_vlan(vid=10,pcp=0),1,5,6,7,8,100" \
> "2 0 1 pop_vlan,9,push_vlan(vid=10,pcp=1),1,5,6,7,8,100" \
> - "2 10 0 drop" \
> - "2 10 1 drop" \
> - "2 11 0 drop" \
> - "2 11 1 drop" \
> - "2 12 0 drop" \
> - "2 12 1 drop" \
> + "2 10 0 drop:pipeline-drop" \
> + "2 10 1 drop:pipeline-drop" \
> + "2 11 0 drop:pipeline-drop" \
> + "2 11 1 drop:pipeline-drop" \
> + "2 12 0 drop:pipeline-drop" \
> + "2 12 1 drop:pipeline-drop" \
> "3 none 0 4,7,8,11,12,push_vlan(vid=12,pcp=0),1,5,6,100" \
> "3 0 0 pop_vlan,4,7,8,11,12,push_vlan(vid=12,pcp=0),1,5,6,100" \
> "3 0 1 8,12,pop_vlan,4,7,11,push_vlan(vid=12,pcp=1),1,5,6,100" \
> - "3 10 0 drop" \
> - "3 10 1 drop" \
> - "3 11 0 drop" \
> - "3 11 1 drop" \
> - "3 12 0 drop" \
> - "3 12 1 drop" \
> + "3 10 0 drop:pipeline-drop" \
> + "3 10 1 drop:pipeline-drop" \
> + "3 11 0 drop:pipeline-drop" \
> + "3 11 1 drop:pipeline-drop" \
> + "3 12 0 drop:pipeline-drop" \
> + "3 12 1 drop:pipeline-drop" \
> "4 none 0 3,7,8,11,12,push_vlan(vid=12,pcp=0),1,5,6,100" \
> "4 0 0 pop_vlan,3,7,8,11,12,push_vlan(vid=12,pcp=0),1,5,6,100" \
> "4 0 1 3,8,12,pop_vlan,7,11,push_vlan(vid=12,pcp=1),1,5,6,100" \
> - "4 10 0 drop" \
> - "4 10 1 drop" \
> - "4 11 0 drop" \
> - "4 11 1 drop" \
> - "4 12 0 drop" \
> - "4 12 1 drop" \
> + "4 10 0 drop:pipeline-drop" \
> + "4 10 1 drop:pipeline-drop" \
> + "4 11 0 drop:pipeline-drop" \
> + "4 11 1 drop:pipeline-drop" \
> + "4 12 0 drop:pipeline-drop" \
> + "4 12 1 drop:pipeline-drop" \
> "5 none 0 2,9,push_vlan(vid=10,pcp=0),1,6,7,8,100" \
> "5 0 0 pop_vlan,2,9,push_vlan(vid=10,pcp=0),1,6,7,8,100" \
> "5 0 1 pop_vlan,2,9,push_vlan(vid=10,pcp=1),1,6,7,8,100" \
> @@ -3555,8 +3555,8 @@ for tuple in \
> "6 0 1 pop_vlan,2,9,push_vlan(vid=10,pcp=1),1,5,7,8,100" \
> "6 10 0 1,5,7,8,100,pop_vlan,2,9" \
> "6 10 1 1,5,7,8,100,pop_vlan,2,9" \
> - "6 11 0 drop" \
> - "6 11 1 drop" \
> + "6 11 0 drop:pipeline-drop" \
> + "6 11 1 drop:pipeline-drop" \
> "6 12 0 1,5,100,pop_vlan,3,4,7,8,11,12" \
> "6 12 1 1,5,100,pop_vlan,4,7,11,push_vlan(vid=0,pcp=1),3,8,12" \
> "7 none 0 3,4,8,11,12,push_vlan(vid=12,pcp=0),1,5,6,100" \
> @@ -3573,16 +3573,16 @@ for tuple in \
> "8 0 1 3,12,pop_vlan,4,7,11,push_vlan(vid=12,pcp=1),1,5,6,100" \
> "8 10 0 1,5,6,7,100,pop_vlan,2,9" \
> "8 10 1 1,5,6,7,100,pop_vlan,2,9" \
> - "8 11 0 drop" \
> - "8 11 1 drop" \
> + "8 11 0 drop:pipeline-drop" \
> + "8 11 1 drop:pipeline-drop" \
> "8 12 0 1,5,6,100,pop_vlan,3,4,7,11,12" \
> "8 12 1 1,5,6,100,pop_vlan,4,7,11,push_vlan(vid=0,pcp=1),3,12" \
> "9 none 0 2,push_vlan(vid=10,pcp=0),1,5,6,7,8,100" \
> "9 10 0 10,push_vlan(vid=10,pcp=0),1,5,6,7,8,100" \
> "9 11 0 push_vlan(vid=10,pcp=0),1,5,6,7,8,100" \
> - "10 none 0 drop" \
> - "10 0 0 drop" \
> - "10 11 0 drop" \
> + "10 none 0 drop:pipeline-drop" \
> + "10 0 0 drop:pipeline-drop" \
> + "10 11 0 drop:pipeline-drop" \
> "10 12 0 9,push_vlan(vid=10,pcp=0),1,5,6,7,8,100" \
> "11 10 0 7,8,12,push_vlan(vid=12,pcp=0),1,5,6,100" \
> "11 10 1 7,8,12,push_vlan(vid=12,pcp=0),1,5,6,100"
> @@ -4322,11 +4322,11 @@ no_flow="$base_flow,frag=no),tcp(src=12345,dst=80)"
> first_flow="$base_flow,frag=first),tcp(src=12345,dst=80)"
> later_flow="$base_flow,frag=later)"
>
> - # mode no first later
> + # mode no first later
> for tuple in \
> - 'normal 1 5 6' \
> - 'drop 1 drop drop' \
> - 'nx-match 1 2 6'
> + 'normal 1 5 6' \
> + 'drop 1 drop:pipeline-drop drop:pipeline-drop' \
> + 'nx-match 1 2 6'
> do
> set $tuple
> mode=$1
> @@ -4404,8 +4404,8 @@ done
> AT_CHECK([ovs-appctl dpctl/dump-flows], [0], [dnl
> flow-dump from non-dpdk interfaces:
>
> recirc_id(0),in_port(90),packet_type(ns=0,id=0),eth_type(0x0800),ipv4(proto=6,frag=no),tcp(dst=80),
> packets:0, bytes:0, used:never, actions:set(tcp(dst=81)),1
> -recirc_id(0),in_port(90),packet_type(ns=0,id=0),eth_type(0x0800),ipv4(frag=first),
> packets:0, bytes:0, used:never, actions:drop
> -recirc_id(0),in_port(90),packet_type(ns=0,id=0),eth_type(0x0800),ipv4(frag=later),
> packets:0, bytes:0, used:never, actions:drop
> +recirc_id(0),in_port(90),packet_type(ns=0,id=0),eth_type(0x0800),ipv4(frag=first),
> packets:0, bytes:0, used:never, actions:drop:pipeline-drop
> +recirc_id(0),in_port(90),packet_type(ns=0,id=0),eth_type(0x0800),ipv4(frag=later),
> packets:0, bytes:0, used:never, actions:drop:pipeline-drop
> ])
>
> mode=nx-match
> @@ -5751,7 +5751,7 @@ bridge("br0")
>
> Final flow: <cleared>
> Megaflow: <cleared>
> -Datapath actions: drop
> +Datapath actions: drop:pipeline-drop
> ])
>
> dnl Now, try again without megaflows:
> @@ -5772,7 +5772,7 @@ bridge("br0")
>
> Final flow: <cleared>
> Megaflow: <cleared>
> -Datapath actions: drop
> +Datapath actions: drop:pipeline-drop
> ])
>
> OVS_VSWITCHD_STOP
> @@ -7066,7 +7066,7 @@ for i in `seq 1 3`; do
> done
> AT_CHECK([ovs-appctl dpctl/dump-flows | sed 's/.*\(packets:\)/\1/' | sed
> 's/used:[[0-9]].[[0-9]]*s/used:0.001s/'], [0], [dnl
> flow-dump from non-dpdk interfaces:
> -packets:2, bytes:68, used:0.001s, actions:drop
> +packets:2, bytes:68, used:0.001s, actions:drop:pipeline-drop
> ])
>
> OVS_VSWITCHD_STOP(["/sending to collector failed/d"])
> @@ -7158,7 +7158,7 @@ for i in `seq 1 3`; do
> done
> AT_CHECK([ovs-appctl dpctl/dump-flows | sed 's/.*\(packets:\)/\1/' | sed
> 's/used:[[0-9]].[[0-9]]*s/used:0.001s/'], [0], [dnl
> flow-dump from non-dpdk interfaces:
> -packets:2, bytes:68, used:0.001s, actions:drop
> +packets:2, bytes:68, used:0.001s, actions:drop:pipeline-drop
> ])
>
> OVS_VSWITCHD_STOP(["/sending to collector failed/d
> @@ -7832,21 +7832,21 @@ AT_CHECK([ovs-appctl netdev-dummy/receive p2
> 'in_port(2),eth(src=50:54:00:00:00:
> AT_CHECK([ovs-appctl netdev-dummy/receive p3
> 'in_port(3),eth(src=50:54:00:00:00:09,dst=50:54:00:00:00:0a),eth_type(0x0800),ipv4(src=10.0.0.2,dst=10.0.0.1,proto=1,tos=0,ttl=64,frag=no),icmp(type=8,code=0)'])
> ovs-appctl revalidator/wait
> AT_CHECK([ovs-appctl dpif/dump-flows br0 | strip_ufid | strip_used | sort],
> [0], [dnl
> -recirc_id(0),in_port(1),packet_type(ns=0,id=0),eth_type(0x0800),ipv4(frag=no),
> packets:0, bytes:0, used:never, actions:drop
> -recirc_id(0),in_port(2),packet_type(ns=0,id=0),eth_type(0x0800),ipv4(frag=no),
> packets:0, bytes:0, used:never, actions:drop
> +recirc_id(0),in_port(1),packet_type(ns=0,id=0),eth_type(0x0800),ipv4(frag=no),
> packets:0, bytes:0, used:never, actions:drop:pipeline-drop
> +recirc_id(0),in_port(2),packet_type(ns=0,id=0),eth_type(0x0800),ipv4(frag=no),
> packets:0, bytes:0, used:never, actions:drop:pipeline-drop
> ])
>
> AT_CHECK([ovs-appctl dpif/dump-flows br1 | strip_ufid | strip_used | sort],
> [0], [dnl
> -recirc_id(0),in_port(3),packet_type(ns=0,id=0),eth_type(0x0800),ipv4(frag=no),
> packets:0, bytes:0, used:never, actions:drop
> +recirc_id(0),in_port(3),packet_type(ns=0,id=0),eth_type(0x0800),ipv4(frag=no),
> packets:0, bytes:0, used:never, actions:drop:pipeline-drop
> ])
>
> AT_CHECK([ovs-appctl dpif/dump-flows -m br0 | strip_ufid | strip_used |
> sort], [0], [dnl
> -skb_priority(0/0),skb_mark(0/0),ct_state(0/0),ct_zone(0/0),ct_mark(0/0),ct_label(0/0),recirc_id(0),dp_hash(0/0),in_port(p1),packet_type(ns=0,id=0),eth(src=50:54:00:00:00:05/00:00:00:00:00:00,dst=50:54:00:00:00:07/00:00:00:00:00:00),eth_type(0x0800),ipv4(src=192.168.0.1/0.0.0.0,dst=192.168.0.2/0.0.0.0,proto=1/0,tos=0/0,ttl=64/0,frag=no),icmp(type=8/0,code=0/0),
> packets:0, bytes:0, used:never, actions:drop
> -skb_priority(0/0),skb_mark(0/0),ct_state(0/0),ct_zone(0/0),ct_mark(0/0),ct_label(0/0),recirc_id(0),dp_hash(0/0),in_port(p2),packet_type(ns=0,id=0),eth(src=50:54:00:00:00:07/00:00:00:00:00:00,dst=50:54:00:00:00:05/00:00:00:00:00:00),eth_type(0x0800),ipv4(src=192.168.0.2/0.0.0.0,dst=192.168.0.1/0.0.0.0,proto=1/0,tos=0/0,ttl=64/0,frag=no),icmp(type=0/0,code=0/0),
> packets:0, bytes:0, used:never, actions:drop
> +skb_priority(0/0),skb_mark(0/0),ct_state(0/0),ct_zone(0/0),ct_mark(0/0),ct_label(0/0),recirc_id(0),dp_hash(0/0),in_port(p1),packet_type(ns=0,id=0),eth(src=50:54:00:00:00:05/00:00:00:00:00:00,dst=50:54:00:00:00:07/00:00:00:00:00:00),eth_type(0x0800),ipv4(src=192.168.0.1/0.0.0.0,dst=192.168.0.2/0.0.0.0,proto=1/0,tos=0/0,ttl=64/0,frag=no),icmp(type=8/0,code=0/0),
> packets:0, bytes:0, used:never, actions:drop:pipeline-drop
> +skb_priority(0/0),skb_mark(0/0),ct_state(0/0),ct_zone(0/0),ct_mark(0/0),ct_label(0/0),recirc_id(0),dp_hash(0/0),in_port(p2),packet_type(ns=0,id=0),eth(src=50:54:00:00:00:07/00:00:00:00:00:00,dst=50:54:00:00:00:05/00:00:00:00:00:00),eth_type(0x0800),ipv4(src=192.168.0.2/0.0.0.0,dst=192.168.0.1/0.0.0.0,proto=1/0,tos=0/0,ttl=64/0,frag=no),icmp(type=0/0,code=0/0),
> packets:0, bytes:0, used:never, actions:drop:pipeline-drop
> ])
>
> AT_CHECK([ovs-appctl dpif/dump-flows -m br1 | strip_ufid | strip_used |
> sort], [0], [dnl
> -skb_priority(0/0),skb_mark(0/0),ct_state(0/0),ct_zone(0/0),ct_mark(0/0),ct_label(0/0),recirc_id(0),dp_hash(0/0),in_port(p3),packet_type(ns=0,id=0),eth(src=50:54:00:00:00:09/00:00:00:00:00:00,dst=50:54:00:00:00:0a/00:00:00:00:00:00),eth_type(0x0800),ipv4(src=10.0.0.2/0.0.0.0,dst=10.0.0.1/0.0.0.0,proto=1/0,tos=0/0,ttl=64/0,frag=no),icmp(type=8/0,code=0/0),
> packets:0, bytes:0, used:never, actions:drop
> +skb_priority(0/0),skb_mark(0/0),ct_state(0/0),ct_zone(0/0),ct_mark(0/0),ct_label(0/0),recirc_id(0),dp_hash(0/0),in_port(p3),packet_type(ns=0,id=0),eth(src=50:54:00:00:00:09/00:00:00:00:00:00,dst=50:54:00:00:00:0a/00:00:00:00:00:00),eth_type(0x0800),ipv4(src=10.0.0.2/0.0.0.0,dst=10.0.0.1/0.0.0.0,proto=1/0,tos=0/0,ttl=64/0,frag=no),icmp(type=8/0,code=0/0),
> packets:0, bytes:0, used:never, actions:drop:pipeline-drop
> ])
>
> OVS_VSWITCHD_STOP
> @@ -7869,7 +7869,7 @@ m4_define([OFPROTO_DPIF_GET_FLOW],
>
> UFID=`sed -n 's/\(ufid:[[-0-9a-fA-F]]*\).*/\1/p' stdout`
> AT_CHECK([ovs-appctl dpctl/get-flow $UFID], [0], [dnl
> -recirc_id(0),in_port(1),packet_type(ns=0,id=0),eth_type(0x0800),ipv4(frag=no),
> packets:0, bytes:0, used:never, actions:drop
> +recirc_id(0),in_port(1),packet_type(ns=0,id=0),eth_type(0x0800),ipv4(frag=no),
> packets:0, bytes:0, used:never, actions:drop:pipeline-drop
> ])
>
> OVS_VSWITCHD_STOP
> @@ -8617,11 +8617,11 @@ table=0 in_port=1,ip,nw_dst=10.0.0.3 actions=drop
> sleep 1
> AT_CHECK([strip_ufid < ovs-vswitchd.log | filter_flow_install |
> strip_used], [0], [dnl
>
> skb_priority(0),skb_mark(0),ct_state(-new-est-rel-rpl-inv-trk-snat-dnat),ct_zone(0),ct_mark(0),ct_label(0),recirc_id(0),dp_hash(0),in_port(1),packet_type(ns=0,id=0),eth(src=50:54:00:00:00:09,dst=50:54:00:00:00:0a),eth_type(0x0800),ipv4(src=10.0.0.2,dst=10.0.0.1,proto=1,tos=0,ttl=64,frag=no),icmp(type=8,code=0),
> actions:2
> -skb_priority(0),skb_mark(0),ct_state(-new-est-rel-rpl-inv-trk-snat-dnat),ct_zone(0),ct_mark(0),ct_label(0),recirc_id(0),dp_hash(0),in_port(1),packet_type(ns=0,id=0),eth(src=50:54:00:00:00:0b,dst=50:54:00:00:00:0c),eth_type(0x0800),ipv4(src=10.0.0.4,dst=10.0.0.3,proto=1,tos=0,ttl=64,frag=no),icmp(type=8,code=0),
> actions:drop
> +skb_priority(0),skb_mark(0),ct_state(-new-est-rel-rpl-inv-trk-snat-dnat),ct_zone(0),ct_mark(0),ct_label(0),recirc_id(0),dp_hash(0),in_port(1),packet_type(ns=0,id=0),eth(src=50:54:00:00:00:0b,dst=50:54:00:00:00:0c),eth_type(0x0800),ipv4(src=10.0.0.4,dst=10.0.0.3,proto=1,tos=0,ttl=64,frag=no),icmp(type=8,code=0),
> actions:drop:pipeline-drop
> ])
> AT_CHECK([strip_ufid < ovs-vswitchd.log | filter_flow_dump | grep
> 'packets:3'], [0], [dnl
>
> skb_priority(0),skb_mark(0),ct_state(0/0xff),ct_zone(0),ct_mark(0),ct_label(0),recirc_id(0),dp_hash(0),in_port(1),packet_type(ns=0,id=0),eth(src=50:54:00:00:00:09,dst=50:54:00:00:00:0a),eth_type(0x0800),ipv4(src=10.0.0.2,dst=10.0.0.1,proto=1,tos=0,ttl=64,frag=no),icmp(type=8,code=0),
> packets:3, bytes:318, used:0.0s, actions:2
> -skb_priority(0),skb_mark(0),ct_state(0/0xff),ct_zone(0),ct_mark(0),ct_label(0),recirc_id(0),dp_hash(0),in_port(1),packet_type(ns=0,id=0),eth(src=50:54:00:00:00:0b,dst=50:54:00:00:00:0c),eth_type(0x0800),ipv4(src=10.0.0.4,dst=10.0.0.3,proto=1,tos=0,ttl=64,frag=no),icmp(type=8,code=0),
> packets:3, bytes:318, used:0.0s, actions:drop
> +skb_priority(0),skb_mark(0),ct_state(0/0xff),ct_zone(0),ct_mark(0),ct_label(0),recirc_id(0),dp_hash(0),in_port(1),packet_type(ns=0,id=0),eth(src=50:54:00:00:00:0b,dst=50:54:00:00:00:0c),eth_type(0x0800),ipv4(src=10.0.0.4,dst=10.0.0.3,proto=1,tos=0,ttl=64,frag=no),icmp(type=8,code=0),
> packets:3, bytes:318, used:0.0s, actions:drop:pipeline-drop
> ])
> OVS_VSWITCHD_STOP
> AT_CLEANUP])
> @@ -9348,7 +9348,7 @@ for i in 1 2 3; do
> done
>
> AT_CHECK([ovs-appctl dpif/dump-flows br0 | strip_ufid | strip_used | sort],
> [0], [dnl
> -recirc_id(0),in_port(1),packet_type(ns=0,id=0),eth_type(0x1234), packets:5,
> bytes:70, used:0.0s, actions:drop
> +recirc_id(0),in_port(1),packet_type(ns=0,id=0),eth_type(0x1234), packets:5,
> bytes:70, used:0.0s, actions:drop:pipeline-drop
> ])
>
> # Add a flow that matches the non-presence of a vlan tag, and check
> @@ -9377,16 +9377,16 @@ done
>
> AT_CHECK([ovs-appctl dpif/dump-flows br0 | strip_ufid | strip_used | sort],
> [0], [dnl
> recirc_id(0),in_port(1),packet_type(ns=0,id=0),eth_type(0x1234), packets:8,
> bytes:112, used:0.0s, actions:100
> -recirc_id(0),in_port(1),packet_type(ns=0,id=0),eth_type(0x8100),vlan(vid=99,pcp=7/0x0),encap(eth_type(0x1234)),
> packets:2, bytes:36, used:0.0s, actions:drop
> +recirc_id(0),in_port(1),packet_type(ns=0,id=0),eth_type(0x8100),vlan(vid=99,pcp=7/0x0),encap(eth_type(0x1234)),
> packets:2, bytes:36, used:0.0s, actions:drop:pipeline-drop
> ])
>
> # Check that the new flow matches the CFI bit, while both vid and pcp
> # are wildcarded.
> AT_CHECK([grep '\(modify\)\|\(flow_add\)' ovs-vswitchd.log | strip_ufid ],
> [0], [dnl
> dpif_netdev|DBG|flow_add:
> recirc_id(0),in_port(1),packet_type(ns=0,id=0),eth_type(0x1234), actions:100
> -dpif|DBG|dummy@ovs-dummy: put[[modify]]
> skb_priority(0/0),skb_mark(0/0),ct_state(0/0),ct_zone(0/0),ct_mark(0/0),ct_label(0/0),recirc_id(0),dp_hash(0/0),in_port(1),packet_type(ns=0,id=0),eth(src=50:54:00:00:00:09/00:00:00:00:00:00,dst=50:54:00:00:00:0a/00:00:00:00:00:00),eth_type(0x1234)
> +dpif|DBG|dummy@ovs-dummy: put[[modify]]
> skb_priority(0/0),skb_mark(0/0),ct_state(0/0),ct_zone(0/0),ct_mark(0/0),ct_label(0/0),recirc_id(0),dp_hash(0/0),in_port(1),packet_type(ns=0,id=0),eth(src=50:54:00:00:00:09/00:00:00:00:00:00,dst=50:54:00:00:00:0a/00:00:00:00:00:00),eth_type(0x1234),
> actions:drop:pipeline-drop
> dpif|DBG|dummy@ovs-dummy: put[[modify]]
> skb_priority(0/0),skb_mark(0/0),ct_state(0/0),ct_zone(0/0),ct_mark(0/0),ct_label(0/0),recirc_id(0),dp_hash(0/0),in_port(1),packet_type(ns=0,id=0),eth(src=50:54:00:00:00:09/00:00:00:00:00:00,dst=50:54:00:00:00:0a/00:00:00:00:00:00),eth_type(0x1234),
> actions:100
> -dpif_netdev|DBG|flow_add:
> recirc_id(0),in_port(1),packet_type(ns=0,id=0),eth_type(0x8100),vlan(vid=99,pcp=7/0x0),encap(eth_type(0x1234)),
> actions:drop
> +dpif_netdev|DBG|flow_add:
> recirc_id(0),in_port(1),packet_type(ns=0,id=0),eth_type(0x8100),vlan(vid=99,pcp=7/0x0),encap(eth_type(0x1234)),
> actions:drop:pipeline-drop
> ])
> OVS_VSWITCHD_STOP
> AT_CLEANUP
> @@ -9711,7 +9711,7 @@ AT_CHECK([ovs-appctl netdev-dummy/receive p2
> 'in_port(2),eth(src=50:54:00:00:00:
>
>
> AT_CHECK([cat ovs-vswitchd.log | strip_ufid | filter_flow_install], [0], [dnl
> -ct_state(+new-est+trk),recirc_id(0x1),in_port(2),packet_type(ns=0,id=0),eth_type(0x0800),ipv4(frag=no),
> actions:drop
> +ct_state(+new-est+trk),recirc_id(0x1),in_port(2),packet_type(ns=0,id=0),eth_type(0x0800),ipv4(frag=no),
> actions:drop:pipeline-drop
>
> ct_state(-new+est+trk),recirc_id(0x1),in_port(2),packet_type(ns=0,id=0),eth_type(0x0800),ipv4(proto=17,frag=no),
> actions:1
>
> recirc_id(0),in_port(1),packet_type(ns=0,id=0),eth_type(0x0800),ipv4(proto=17,frag=no),
> actions:ct(commit),2
>
> recirc_id(0),in_port(2),packet_type(ns=0,id=0),eth_type(0x0800),ipv4(proto=17,frag=no),
> actions:ct,recirc(0x1)
> @@ -10398,7 +10398,7 @@ AT_CHECK([ovs-ofctl add-flows br0 flows.txt])
>
> AT_CHECK([ovs-appctl ofproto/trace br0 'in_port=2,udp'], [0], [stdout])
> AT_CHECK([tail -1 stdout], [0],
> - [Datapath actions: drop
> + [Datapath actions: drop:pipeline-drop
> ])
>
> AT_CHECK([ovs-appctl ofproto/trace br0 'in_port=1,udp'], [0], [stdout])
> @@ -10490,7 +10490,7 @@ AT_CHECK([ovs-appctl netdev-dummy/receive p1
> 'recirc_id(0),in_port(1),eth_type(0
> ovs-appctl time/warp 5000
>
> AT_CHECK([strip_ufid < ovs-vswitchd.log | filter_flow_install | strip_used],
> [0], [dnl
> -recirc_id(0),in_port(1),packet_type(ns=0,id=0),eth_type(0x0800),ipv4(proto=17,frag=later),
> actions:drop
> +recirc_id(0),in_port(1),packet_type(ns=0,id=0),eth_type(0x0800),ipv4(proto=17,frag=later),
> actions:drop:pipeline-drop
> ])
>
> dnl Change the flow table. This will trigger revalidation of all the flows.
> diff --git a/tests/ovs-ofctl.at b/tests/ovs-ofctl.at
> index 6ae3470..cd19d41 100644
> --- a/tests/ovs-ofctl.at
> +++ b/tests/ovs-ofctl.at
> @@ -2999,7 +2999,7 @@ AT_CHECK([tail -1 stdout], [0],
> dnl Inbound web traffic with SYN bit without ACK or RST bits
> AT_CHECK([ovs-appctl ofproto/trace ovs-dummy
> 'in_port(2),eth(src=50:54:00:00:00:07,dst=50:54:00:00:00:05),eth_type(0x0800),ipv4(src=192.168.0.2,dst=192.168.0.1,proto=6,tos=0,ttl=64,frag=no),tcp(src=80,dst=8),tcp_flags(0xfeb)'],
> [0], [stdout])
> AT_CHECK([tail -1 stdout], [0],
> - [Datapath actions: drop
> + [Datapath actions: drop:pipeline-drop
> ])
>
> OVS_VSWITCHD_STOP
> diff --git a/tests/packet-type-aware.at b/tests/packet-type-aware.at
> index bfb47b4..7c1b551 100644
> --- a/tests/packet-type-aware.at
> +++ b/tests/packet-type-aware.at
> @@ -505,7 +505,7 @@ AT_CHECK([
> ovs-appctl dpctl/dump-flows --names dummy@ovs-dummy | strip_used | grep
> -v ipv6 | sort
> ], [0], [flow-dump from non-dpdk interfaces:
>
> recirc_id(0),in_port(n3),packet_type(ns=0,id=0),eth_type(0x0800),ipv4(dst=192.168.10.20,tos=0/0x3,frag=no),
> packets:1, bytes:98, used:0.0s,
> actions:pop_eth,clone(tnl_push(tnl_port(gre_sys),header(size=38,type=3,eth(dst=aa:55:00:00:00:02,src=aa:55:00:00:00:03,dl_type=0x0800),ipv4(src=30.0.0.3,dst=30.0.0.2,proto=47,tos=0,ttl=64,frag=0x4000),gre((flags=0x0,proto=0x800))),out_port(br-p3)),set(ipv4(src=20.0.0.3,dst=20.0.0.2)),tnl_pop(gre_sys))
> -tunnel(src=20.0.0.3,dst=20.0.0.2,flags(-df-csum)),recirc_id(0),in_port(gre_sys),packet_type(ns=1,id=0x800),ipv4(dst=192.168.10.20,frag=no),
> packets:1, bytes:84, used:0.0s, actions:drop
> +tunnel(src=20.0.0.3,dst=20.0.0.2,flags(-df-csum)),recirc_id(0),in_port(gre_sys),packet_type(ns=1,id=0x800),ipv4(dst=192.168.10.20,frag=no),
> packets:1, bytes:84, used:0.0s, actions:drop:pipeline-drop
> ])
>
> OVS_VSWITCHD_STOP(["/The Open vSwitch kernel module is probably not
> loaded/d"])
> @@ -565,7 +565,14 @@ ovs-appctl time/warp 1000
> AT_CHECK([
> ovs-appctl dpctl/dump-flows | strip_used | grep -v ipv6 | sort
> ], [0], [flow-dump from non-dpdk interfaces:
> -recirc_id(0),in_port(1),packet_type(ns=0,id=0),eth_type(0x0800),ipv4(frag=no),
> packets:1, bytes:98, used:0.0s, actions:drop
> +recirc_id(0),in_port(1),packet_type(ns=0,id=0),eth_type(0x0800),ipv4(frag=no),
> packets:1, bytes:98, used:0.0s, actions:drop:unsupported packet type
> +])
> +
> +
> +AT_CHECK([
> +ovs-appctl coverage/show | grep "drop_action_unsupported_packet_type" | cut
> -d':' -f2|sed 's/ //'
> +], [0], [dnl
> +2
> ])
>
> # Encap(ethernet) on Ethernet frame -> should be droped
> @@ -587,7 +594,7 @@ ovs-appctl time/warp 1000
> AT_CHECK([
> ovs-appctl dpctl/dump-flows | strip_used | grep -v ipv6 | sort
> ], [0], [flow-dump from non-dpdk interfaces:
> -recirc_id(0),in_port(1),packet_type(ns=0,id=0),eth_type(0x0800),ipv4(frag=no),
> packets:1, bytes:98, used:0.0s, actions:drop
> +recirc_id(0),in_port(1),packet_type(ns=0,id=0),eth_type(0x0800),ipv4(frag=no),
> packets:1, bytes:98, used:0.0s, actions:drop:pipeline-drop
> ])
>
> # Encap(ethernet) on VLAN tagged Ethernet frame -> should be droped
> @@ -609,7 +616,7 @@ ovs-appctl time/warp 1000
> AT_CHECK([
> ovs-appctl dpctl/dump-flows | strip_used | grep -v ipv6 | sort
> ], [0], [flow-dump from non-dpdk interfaces:
> -recirc_id(0),in_port(1),packet_type(ns=0,id=0),eth_type(0x0800),ipv4(frag=no),
> packets:1, bytes:98, used:0.0s, actions:drop
> +recirc_id(0),in_port(1),packet_type(ns=0,id=0),eth_type(0x0800),ipv4(frag=no),
> packets:1, bytes:98, used:0.0s, actions:drop:unsupported packet type
> ])
>
> OVS_VSWITCHD_STOP
> @@ -770,7 +777,7 @@ ovs-appctl time/warp 1000
> AT_CHECK([
> ovs-appctl dpctl/dump-flows --names dummy@ovs-dummy | strip_used | grep
> -v ipv6 | sort
> ], [0], [flow-dump from non-dpdk interfaces:
> -recirc_id(0),in_port(n0),packet_type(ns=0,id=0),eth_type(0x0800),ipv4(frag=no),
> packets:1, bytes:98, used:0.0s, actions:drop
> +recirc_id(0),in_port(n0),packet_type(ns=0,id=0),eth_type(0x0800),ipv4(frag=no),
> packets:1, bytes:98, used:0.0s, actions:drop:pipeline-drop
> ])
>
> AT_CHECK([
> diff --git a/tests/testsuite.at b/tests/testsuite.at
> index b840dbf..922ba48 100644
> --- a/tests/testsuite.at
> +++ b/tests/testsuite.at
> @@ -82,3 +82,4 @@ m4_include([tests/ovn-controller-vtep.at])
> m4_include([tests/mcast-snooping.at])
> m4_include([tests/packet-type-aware.at])
> m4_include([tests/nsh.at])
> +m4_include([tests/drop-stats.at])
> diff --git a/tests/tunnel-push-pop.at b/tests/tunnel-push-pop.at
> index f717243..385c114 100644
> --- a/tests/tunnel-push-pop.at
> +++ b/tests/tunnel-push-pop.at
> @@ -447,6 +447,27 @@ AT_CHECK([ovs-ofctl dump-ports int-br | grep 'port 7'],
> [0], [dnl
> port 7: rx pkts=3, bytes=252, drop=?, errs=?, frame=?, over=?, crc=?
> ])
>
> +AT_CHECK([ovs-appctl netdev-dummy/receive p0
> 'aa55aa550000001b213cab6408004500007079464000402fba600101025c0101025820000800000001c845000054ba200000400184861e0000011e00000200004227e75400030af3195500000000f265010000000000101112131415161718191a1b1c1d1e1f202122232425262728292a2b2c2d2e2f3031323334353637'])
> +
> +ovs-appctl time/warp 1200
> +
> +
> +AT_CHECK([
> +ovs-appctl coverage/show | grep "dp_tunnel_pop_error_drop" | cut -d':'
> -f2|sed 's/ //'
> +], [0], [dnl
> +1
> +])
> +
> +
> +AT_CHECK([ovs-appctl netdev-dummy/receive p0
> 'aa55aa550000001b213cab6408004503007079464000402fba600101025c0101025820000800000001c845000054ba200000400184861e0000011e00000200004227e75400030af3195500000000f265010000000000101112131415161718191a1b1c1d1e1f202122232425262728292a2b2c2d2e2f3031323334353637'])
> +ovs-appctl time/warp 1200
> +
> +AT_CHECK([
> +ovs-appctl coverage/show | grep "drop_action_congestion" | cut -d':' -f2|sed
> 's/ //'
> +], [0], [dnl
> +1
> +])
> +
> dnl Check GREL3 only accepts non-fragmented packets?
> AT_CHECK([ovs-appctl netdev-dummy/receive p0
> 'aa55aa550000001b213cab6408004500007e79464000402fba550101025c0101025820000800000001c8fe71d883724fbeb6f4e1494a080045000054ba200000400184861e0000011e00000200004227e75400030af3195500000000f265010000000000101112131415161718191a1b1c1d1e1f202122232425262728292a2b2c2d2e2f3031323334353637'])
>
> @@ -455,7 +476,7 @@ ovs-appctl time/warp 1000
>
> AT_CHECK([ovs-ofctl dump-ports int-br | grep 'port [[37]]' | sort], [0],
> [dnl
> port 3: rx pkts=3, bytes=294, drop=?, errs=?, frame=?, over=?, crc=?
> - port 7: rx pkts=4, bytes=350, drop=?, errs=?, frame=?, over=?, crc=?
> + port 7: rx pkts=5, bytes=434, drop=?, errs=?, frame=?, over=?, crc=?
> ])
>
> dnl Check decapsulation of Geneve packet with options
> @@ -510,7 +531,8 @@ AT_CHECK([ovs-appctl tnl/ports/show |sort], [0], [dnl
> Listening ports:
> ])
>
> -OVS_VSWITCHD_STOP
> +OVS_VSWITCHD_STOP(["/dropping tunnel packet marked ECN CE but is not ECN
> capable/d
> +/ip packet has invalid checksum/d"])
> AT_CLEANUP
>
> AT_SETUP([tunnel_push_pop - packet_out])
> diff --git a/tests/tunnel.at b/tests/tunnel.at
> index 55fb1d3..3646c06 100644
> --- a/tests/tunnel.at
> +++ b/tests/tunnel.at
> @@ -102,10 +102,12 @@ Datapath actions: set(ipv4(tos=0x3/0x3)),2
>
> dnl Tunnel CE and encapsulated packet Non-ECT
> AT_CHECK([ovs-appctl ofproto/trace ovs-dummy
> 'tunnel(src=1.1.1.1,dst=2.2.2.2,tos=0x3,ttl=64,flags()),in_port(1),eth(src=50:54:00:00:00:05,dst=50:54:00:00:00:07),eth_type(0x0800),ipv4(src=192.168.0.1,dst=192.168.0.2,proto=6,tos=0,ttl=64,frag=no),tcp(src=8,dst=9)'],
> [0], [stdout])
> -AT_CHECK([tail -2 stdout], [0],
> +AT_CHECK([tail -3 stdout], [0],
> [Megaflow:
> recirc_id=0,eth,ip,tun_id=0,tun_src=1.1.1.1,tun_dst=2.2.2.2,tun_tos=3,tun_flags=-df-csum-key,in_port=1,nw_ecn=0,nw_frag=no
> -Datapath actions: drop
> +Datapath actions: drop:ecn mismatch at tunnel decapsulation
> +Translation failed (CONGESTION DROP), packet is dropped.
> ])
> +
> OVS_VSWITCHD_STOP(["/dropping tunnel packet marked ECN CE but is not ECN
> capable/d"])
> AT_CLEANUP
>
> @@ -193,6 +195,15 @@ AT_CHECK([ovs-appctl ofproto/trace ovs-dummy
> 'in_port(2),eth(src=50:54:00:00:00:
> AT_CHECK([tail -1 stdout], [0],
> [Datapath actions:
> set(tunnel(tun_id=0x5,src=2.2.2.2,dst=1.1.1.1,ttl=64,flags(df|key))),set(skb_mark(0x2)),1
> ])
> +
> +AT_CHECK([ovs-appctl netdev-dummy/receive p2
> 'aa55aa550001f8bc124434b6080045000054ba20000040018486010103580101037001004227e75400030af3195500000000f265010000000000101112131415161718191a1b1c1d1e1f202122232425262728292a2b2c2d2e2f3031323334353637'])
> +sleep 2
> +
> +AT_CHECK([
> +ovs-appctl coverage/show | grep "dp_invalid_port_drop" | cut -d':' -f2|sed
> 's/ //'
> +], [0], [dnl
> +1
> +])
> OVS_VSWITCHD_STOP
> AT_CLEANUP
>
> @@ -364,7 +375,7 @@ Datapath actions: 4,3,5
>
> AT_CHECK([ovs-appctl ofproto/trace ovs-dummy
> 'tunnel(tun_id=0x0,src=1.1.1.1,dst=2.2.2.2,ttl=64,flags(key)),in_port(1),eth(src=50:54:00:00:00:05,dst=50:54:00:00:00:07),eth_type(0x0800),ipv4(src=192.168.0.1,dst=192.168.0.2,proto=6,tos=0,ttl=64,frag=no),tcp(src=8,dst=9)'],
> [0], [stdout])
> AT_CHECK([tail -1 stdout], [0], [dnl
> -Datapath actions: drop
> +Datapath actions: drop:pipeline-drop
> ])
>
> OVS_VSWITCHD_STOP
> @@ -571,7 +582,7 @@ dnl receive packet from ERSPAN port with wrong v1 metadata
> AT_CHECK([ovs-appctl ofproto/trace ovs-dummy
> 'recirc_id(0),tunnel(tun_id=0x1,src=1.1.1.1,dst=2.2.2.2,ttl=64,erspan(ver=1,idx=0xabcd),flags(df|key)),in_port(3),skb_mark(0),eth(src=50:54:00:00:00:05,dst=50:54:00:00:00:07),eth_type(0x0800),ipv4(frag=no)'],
> [0], [stdout])
> AT_CHECK([tail -2 stdout], [0],
> [Megaflow:
> recirc_id=0,eth,ip,tun_id=0x1,tun_src=1.1.1.1,tun_dst=2.2.2.2,tun_tos=0,tun_erspan_ver=1,tun_erspan_idx=0xabcd,tun_flags=+df-csum+key,in_port=3,nw_frag=no
> -Datapath actions: drop
> +Datapath actions: drop:pipeline-drop
> ])
>
> dnl receive packet from ERSPAN port with v2 metadata
> @@ -585,7 +596,7 @@ dnl receive packet from ERSPAN port with wrong v2 metadata
> AT_CHECK([ovs-appctl ofproto/trace ovs-dummy
> 'recirc_id(0),tunnel(tun_id=0x2,src=1.1.1.2,dst=2.2.2.2,ttl=64,erspan(ver=2,dir=0,hwid=0x17),flags(df|key)),in_port(3),skb_mark(0),eth(src=50:54:00:00:00:05,dst=50:54:00:00:00:07),eth_type(0x0800),ipv4(frag=no)'],
> [0], [stdout])
> AT_CHECK([tail -2 stdout], [0],
> [Megaflow:
> recirc_id=0,eth,ip,tun_id=0x2,tun_src=1.1.1.2,tun_dst=2.2.2.2,tun_tos=0,tun_erspan_ver=2,tun_erspan_dir=0,tun_erspan_hwid=0x1,tun_flags=+df-csum+key,in_port=4,nw_frag=no
> -Datapath actions: drop
> +Datapath actions: drop:pipeline-drop
> ])
>
> dnl test wildcard mask: recevie all v2 regardless of its metadata
>
_______________________________________________
dev mailing list
d...@openvswitch.org
https://mail.openvswitch.org/mailman/listinfo/ovs-dev
