diff mbox

[dpdk-dev,v2] doc: flow classify guides

Message ID 1509707635-24405-2-git-send-email-bernard.iremonger@intel.com (mailing list archive)
State Accepted, archived
Headers show

Checks

Context Check Description
ci/checkpatch warning coding style issues
ci/Intel-compilation success Compilation OK

Commit Message

Bernard Iremonger Nov. 3, 2017, 11:13 a.m. UTC
The Flow Classify Library Programmers Guide documents
librte_flow_classify.

The Flow Classify Sample Application Guide documents the
flow_classify sample application which is used to
demonstate the use of the Flow Classify Library,
librte_flow_classify.

Updated MAINTAINERS file

Signed-off-by: Bernard Iremonger <bernard.iremonger@intel.com>
---
 MAINTAINERS                                 |   2 +
 doc/guides/prog_guide/flow_classify_lib.rst | 427 +++++++++++++++++++++
 doc/guides/prog_guide/index.rst             |   1 +
 doc/guides/sample_app_ug/flow_classify.rst  | 575 ++++++++++++++++++++++++++++
 doc/guides/sample_app_ug/index.rst          |   1 +
 5 files changed, 1006 insertions(+)
 create mode 100644 doc/guides/prog_guide/flow_classify_lib.rst
 create mode 100644 doc/guides/sample_app_ug/flow_classify.rst

Comments

Mcnamara, John Nov. 3, 2017, 3:34 p.m. UTC | #1
> -----Original Message-----
> From: Iremonger, Bernard
> Sent: Friday, November 3, 2017 11:14 AM
> To: dev@dpdk.org; Mcnamara, John <john.mcnamara@intel.com>
> Cc: Iremonger, Bernard <bernard.iremonger@intel.com>
> Subject: [PATCH v2] doc: flow classify guides
> 
> The Flow Classify Library Programmers Guide documents
> librte_flow_classify.
> 
> The Flow Classify Sample Application Guide documents the flow_classify
> sample application which is used to demonstate the use of the Flow
> Classify Library, librte_flow_classify.
> 
> Updated MAINTAINERS file
> 
> Signed-off-by: Bernard Iremonger <bernard.iremonger@intel.com>

Acked-by: John McNamara <john.mcnamara@intel.com>
Thomas Monjalon Nov. 7, 2017, 10 p.m. UTC | #2
> > The Flow Classify Library Programmers Guide documents
> > librte_flow_classify.
> > 
> > The Flow Classify Sample Application Guide documents the flow_classify
> > sample application which is used to demonstate the use of the Flow
> > Classify Library, librte_flow_classify.
> > 
> > Updated MAINTAINERS file
> > 
> > Signed-off-by: Bernard Iremonger <bernard.iremonger@intel.com>
> 
> Acked-by: John McNamara <john.mcnamara@intel.com>

Applied, thanks
diff mbox

Patch

diff --git a/MAINTAINERS b/MAINTAINERS
index 8ab08d2..7d30bda 100644
--- a/MAINTAINERS
+++ b/MAINTAINERS
@@ -725,6 +725,8 @@  M: Bernard Iremonger <bernard.iremonger@intel.com>
 F: lib/librte_flow_classify/
 F: test/test/test_flow_classify*
 F: examples/flow_classify/
+F: doc/guides/prog_guide/flow_classify_lib.rst
+F: doc/guides/sample_app_ug/flow_classify.rst
 
 Distributor
 M: Bruce Richardson <bruce.richardson@intel.com>
diff --git a/doc/guides/prog_guide/flow_classify_lib.rst b/doc/guides/prog_guide/flow_classify_lib.rst
new file mode 100644
index 0000000..820dc72
--- /dev/null
+++ b/doc/guides/prog_guide/flow_classify_lib.rst
@@ -0,0 +1,427 @@ 
+..  BSD LICENSE
+    Copyright(c) 2017 Intel Corporation. All rights reserved.
+    All rights reserved.
+
+    Redistribution and use in source and binary forms, with or without
+    modification, are permitted provided that the following conditions
+    are met:
+
+    * Redistributions of source code must retain the above copyright
+    notice, this list of conditions and the following disclaimer.
+    * Redistributions in binary form must reproduce the above copyright
+    notice, this list of conditions and the following disclaimer in
+    the documentation and/or other materials provided with the
+    distribution.
+    * Neither the name of Intel Corporation nor the names of its
+    contributors may be used to endorse or promote products derived
+    from this software without specific prior written permission.
+
+    THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
+    "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
+    LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
+    A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
+    OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
+    SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
+    LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
+    DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
+    THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
+    (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
+    OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+
+Flow Classification Library
+===========================
+
+DPDK provides a Flow Classification library that provides the ability
+to classify an input packet by matching it against a set of Flow rules.
+
+The initial implementation supports counting of IPv4 5-tuple packets which match
+a particular Flow rule only.
+
+Please refer to the
+:doc:`./rte_flow`
+for more information.
+
+The Flow Classification library uses the ``librte_table`` API for managing Flow
+rules and matching packets against the Flow rules.
+The library is table agnostic and can use the following tables:
+``Access Control List``, ``Hash`` and ``Longest Prefix Match(LPM)``.
+The ``Access Control List`` table is used in the initial implementation.
+
+Please refer to the
+:doc:`./packet_framework`
+for more information.on ``librte_table``.
+
+DPDK provides an Access Control List library that provides the ability to
+classify an input packet based on a set of classification rules.
+
+Please refer to the
+:doc:`./packet_classif_access_ctrl`
+library for more information on ``librte_acl``.
+
+There is also a Flow Classify sample application which demonstrates the use of
+the Flow Classification Library API's.
+
+Please refer to the
+:doc:`../sample_app_ug/flow_classify`
+for more information on the ``flow_classify`` sample application.
+
+Overview
+--------
+
+The library has the following API's
+
+.. code-block:: c
+
+    /**
+     * Flow classifier create
+     *
+     * @param params
+     *   Parameters for flow classifier creation
+     * @return
+     *   Handle to flow classifier instance on success or NULL otherwise
+     */
+    struct rte_flow_classifier *
+    rte_flow_classifier_create(struct rte_flow_classifier_params *params);
+
+    /**
+     * Flow classifier free
+     *
+     * @param cls
+     *   Handle to flow classifier instance
+     * @return
+     *   0 on success, error code otherwise
+     */
+    int
+    rte_flow_classifier_free(struct rte_flow_classifier *cls);
+
+    /**
+     * Flow classify table create
+     *
+     * @param cls
+     *   Handle to flow classifier instance
+     * @param params
+     *   Parameters for flow_classify table creation
+     * @param table_id
+     *   Table ID. Valid only within the scope of table IDs of the current
+     *   classifier. Only returned after a successful invocation.
+     * @return
+     *   0 on success, error code otherwise
+     */
+    int
+    rte_flow_classify_table_create(struct rte_flow_classifier *cls,
+           struct rte_flow_classify_table_params *params,
+           uint32_t *table_id);
+
+    /**
+     * Add a flow classify rule to the flow_classifier table.
+     *
+     * @param[in] cls
+     *   Flow classifier handle
+     * @param[in] table_id
+     *   id of table
+     * @param[in] attr
+     *   Flow rule attributes
+     * @param[in] pattern
+     *   Pattern specification (list terminated by the END pattern item).
+     * @param[in] actions
+     *   Associated actions (list terminated by the END pattern item).
+     * @param[out] error
+     *   Perform verbose error reporting if not NULL. Structure
+     *   initialised in case of error only.
+     * @return
+     *   A valid handle in case of success, NULL otherwise.
+     */
+    struct rte_flow_classify_rule *
+    rte_flow_classify_table_entry_add(struct rte_flow_classifier *cls,
+            uint32_t table_id,
+            const struct rte_flow_attr *attr,
+            const struct rte_flow_item pattern[],
+            const struct rte_flow_action actions[],
+            struct rte_flow_error *error);
+
+    /**
+     * Delete a flow classify rule from the flow_classifier table.
+     *
+     * @param[in] cls
+     *   Flow classifier handle
+     * @param[in] table_id
+     *   id of table
+     * @param[in] rule
+     *   Flow classify rule
+     * @return
+     *   0 on success, error code otherwise.
+     */
+    int
+    rte_flow_classify_table_entry_delete(struct rte_flow_classifier *cls,
+            uint32_t table_id,
+            struct rte_flow_classify_rule *rule);
+
+    /**
+     * Query flow classifier for given rule.
+     *
+     * @param[in] cls
+     *   Flow classifier handle
+     * @param[in] table_id
+     *   id of table
+     * @param[in] pkts
+     *   Pointer to packets to process
+     * @param[in] nb_pkts
+     *   Number of packets to process
+     * @param[in] rule
+     *   Flow classify rule
+     * @param[in] stats
+     *   Flow classify stats
+     *
+     * @return
+     *   0 on success, error code otherwise.
+     */
+    int
+    rte_flow_classifier_query(struct rte_flow_classifier *cls,
+            uint32_t table_id,
+            struct rte_mbuf **pkts,
+            const uint16_t nb_pkts,
+            struct rte_flow_classify_rule *rule,
+            struct rte_flow_classify_stats *stats);
+
+Classifier creation
+~~~~~~~~~~~~~~~~~~~
+
+The application creates the ``Classifier`` using the
+``rte_flow_classifier_create`` API.
+The ``rte_flow_classify_params`` structure must be initialised by the
+application before calling the API.
+
+.. code-block:: c
+
+    struct rte_flow_classifier_params {
+        /** flow classifier name */
+        const char *name;
+
+        /** CPU socket ID where memory for the flow classifier and its */
+        /** elements (tables) should be allocated */
+        int socket_id;
+
+        /** Table type */
+        enum rte_flow_classify_table_type type;
+    };
+
+The ``Classifier`` has the following internal structures:
+
+.. code-block:: c
+
+    struct rte_table {
+        /* Input parameters */
+        struct rte_table_ops ops;
+        uint32_t entry_size;
+        enum rte_flow_classify_table_type type;
+
+        /* Handle to the low-level table object */
+        void *h_table;
+    };
+
+    #define RTE_FLOW_CLASSIFIER_MAX_NAME_SZ 256
+
+    struct rte_flow_classifier {
+        /* Input parameters */
+        char name[RTE_FLOW_CLASSIFIER_MAX_NAME_SZ];
+        int socket_id;
+        enum rte_flow_classify_table_type type;
+
+        /* Internal tables */
+        struct rte_table tables[RTE_FLOW_CLASSIFY_TABLE_MAX];
+        uint32_t num_tables;
+        uint16_t nb_pkts;
+        struct rte_flow_classify_table_entry
+            *entries[RTE_PORT_IN_BURST_SIZE_MAX];
+    } __rte_cache_aligned;
+
+Adding a table to the Classifier
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+
+The application adds a table to the ``Classifier`` using the
+``rte_flow_classify_table_create`` API.
+The ``rte_flow_classify_table_params`` structure must be initialised by the
+application before calling the API.
+
+.. code-block:: c
+
+    struct rte_flow_classify_table_params {
+        /** Table operations (specific to each table type) */
+        struct rte_table_ops *ops;
+
+        /** Opaque param to be passed to the table create operation */
+        void *arg_create;
+
+        /** Memory size to be reserved per classifier object entry for */
+        /** storing meta data */
+        uint32_t table_metadata_size;
+     };
+
+To create an ACL table the ``rte_table_acl_params`` structure must be
+initialised and assigned to ``arg_create`` in the
+``rte_flow_classify_table_params`` structure.
+
+.. code-block:: c
+
+    struct rte_table_acl_params {
+        /** Name */
+        const char *name;
+
+        /** Maximum number of ACL rules in the table */
+        uint32_t n_rules;
+
+        /** Number of fields in the ACL rule specification */
+        uint32_t n_rule_fields;
+
+        /** Format specification of the fields of the ACL rule */
+        struct rte_acl_field_def field_format[RTE_ACL_MAX_FIELDS];
+    };
+
+The fields for the ACL rule must also be initialised by the application.
+
+An ACL table can be added to the ``Classifier`` for each ACL rule, for example
+another table could be added for the IPv6 5-tuple rule.
+
+Flow Parsing
+~~~~~~~~~~~~
+
+The library currently supports three IPv4 5-tuple flow patterns, for UDP, TCP
+and SCTP.
+
+.. code-block:: c
+
+    /* Pattern for IPv4 5-tuple UDP filter */
+    static enum rte_flow_item_type pattern_ntuple_1[] = {
+        RTE_FLOW_ITEM_TYPE_ETH,
+        RTE_FLOW_ITEM_TYPE_IPV4,
+        RTE_FLOW_ITEM_TYPE_UDP,
+        RTE_FLOW_ITEM_TYPE_END,
+    };
+
+    /* Pattern for IPv4 5-tuple TCP filter */
+    static enum rte_flow_item_type pattern_ntuple_2[] = {
+        RTE_FLOW_ITEM_TYPE_ETH,
+        RTE_FLOW_ITEM_TYPE_IPV4,
+        RTE_FLOW_ITEM_TYPE_TCP,
+        RTE_FLOW_ITEM_TYPE_END,
+    };
+
+    /* Pattern for IPv4 5-tuple SCTP filter */
+    static enum rte_flow_item_type pattern_ntuple_3[] = {
+        RTE_FLOW_ITEM_TYPE_ETH,
+        RTE_FLOW_ITEM_TYPE_IPV4,
+        RTE_FLOW_ITEM_TYPE_SCTP,
+        RTE_FLOW_ITEM_TYPE_END,
+    };
+
+The internal function ``flow_classify_parse_flow`` parses the
+IPv4 5-tuple pattern, attributes and actions and returns the 5-tuple data in the
+``rte_eth_ntuple_filter`` structure.
+
+.. code-block:: c
+
+    static int
+    flow_classify_parse_flow(
+                   const struct rte_flow_attr *attr,
+                   const struct rte_flow_item pattern[],
+                   const struct rte_flow_action actions[],
+                   struct rte_flow_error *error)
+
+Adding Flow Rules
+~~~~~~~~~~~~~~~~~
+
+The ``rte_flow_classify_table_entry_add`` API creates an
+``rte_flow_classify`` object which contains the flow_classify id and type, the
+action, a union of add and delete keys and a union of rules.
+It uses the ``flow_classify_parse_flow`` internal function for parsing the
+flow parameters.
+The 5-tuple ACL key data is obtained from the ``rte_eth_ntuple_filter``
+structure populated by the ``classify_parse_ntuple_filter`` function which
+parses the Flow rule.
+
+.. code-block:: c
+
+    struct acl_keys {
+        struct rte_table_acl_rule_add_params key_add; /* add key */
+        struct rte_table_acl_rule_delete_params	key_del; /* delete key */
+    };
+
+    struct classify_rules {
+        enum rte_flow_classify_rule_type type;
+        union {
+            struct rte_flow_classify_ipv4_5tuple ipv4_5tuple;
+        } u;
+    };
+
+    struct rte_flow_classify {
+        uint32_t id;  /* unique ID of classify object */
+        struct rte_flow_action action; /* action when match found */
+	struct classify_rules rules; /* union of rules */
+        union {
+            struct acl_keys key;
+        } u;
+        int key_found; /* rule key found in table */
+        void *entry; /* pointer to buffer to hold rule meta data */
+        void *entry_ptr; /* handle to the table entry for rule meta data */
+    };
+
+It then calls the ``table[table_id].ops.f_add`` API to add the rule to the ACL
+table.
+
+Deleting Flow Rules
+~~~~~~~~~~~~~~~~~~~
+
+The ``rte_flow_classify_table_entry_delete`` API calls the
+``table[table_id].ops.f_delete`` API to delete a rule from the ACL table.
+
+Packet Matching
+~~~~~~~~~~~~~~~
+
+The ``rte_flow_classifier_query`` API is used to find packets which match a
+given flow Flow rule in the table.
+This API calls the flow_classify_run internal function which calls the
+``table[table_id].ops.f_lookup`` API to see if any packets in a burst match any
+of the Flow rules in the table.
+The meta data for the highest priority rule matched for each packet is returned
+in the entries array in the ``rte_flow_classify`` object.
+The internal function ``action_apply`` implements the ``Count`` action which is
+used to return data which matches a particular Flow rule.
+
+The rte_flow_classifier_query API uses the following structures to return data
+to the application.
+
+.. code-block:: c
+
+    /** IPv4 5-tuple data */
+    struct rte_flow_classify_ipv4_5tuple {
+        uint32_t dst_ip;         /**< Destination IP address in big endian. */
+        uint32_t dst_ip_mask;    /**< Mask of destination IP address. */
+        uint32_t src_ip;         /**< Source IP address in big endian. */
+        uint32_t src_ip_mask;    /**< Mask of destination IP address. */
+        uint16_t dst_port;       /**< Destination port in big endian. */
+        uint16_t dst_port_mask;  /**< Mask of destination port. */
+        uint16_t src_port;       /**< Source Port in big endian. */
+        uint16_t src_port_mask;  /**< Mask of source port. */
+        uint8_t proto;           /**< L4 protocol. */
+        uint8_t proto_mask;      /**< Mask of L4 protocol. */
+    };
+
+    /**
+     * Flow stats
+     *
+     * For the count action, stats can be returned by the query API.
+     *
+     * Storage for stats is provided by the application.
+     *
+     *
+     */
+    struct rte_flow_classify_stats {
+        void *stats;
+    };
+
+    struct rte_flow_classify_5tuple_stats {
+        /** count of packets that match IPv4 5tuple pattern */
+        uint64_t counter1;
+        /** IPv4 5tuple data */
+        struct rte_flow_classify_ipv4_5tuple ipv4_5tuple;
+    };
diff --git a/doc/guides/prog_guide/index.rst b/doc/guides/prog_guide/index.rst
index 9759264..21d1c3d 100644
--- a/doc/guides/prog_guide/index.rst
+++ b/doc/guides/prog_guide/index.rst
@@ -72,6 +72,7 @@  Programmer's Guide
     vhost_lib
     metrics_lib
     port_hotplug_framework
+    flow_classify_lib
     source_org
     dev_kit_build_system
     dev_kit_root_make_help
diff --git a/doc/guides/sample_app_ug/flow_classify.rst b/doc/guides/sample_app_ug/flow_classify.rst
new file mode 100644
index 0000000..bc12b87
--- /dev/null
+++ b/doc/guides/sample_app_ug/flow_classify.rst
@@ -0,0 +1,575 @@ 
+..  BSD LICENSE
+    Copyright(c) 2017 Intel Corporation. All rights reserved.
+    All rights reserved.
+
+    Redistribution and use in source and binary forms, with or without
+    modification, are permitted provided that the following conditions
+    are met:
+
+    * Redistributions of source code must retain the above copyright
+    notice, this list of conditions and the following disclaimer.
+    * Redistributions in binary form must reproduce the above copyright
+    notice, this list of conditions and the following disclaimer in
+    the documentation and/or other materials provided with the
+    distribution.
+    * Neither the name of Intel Corporation nor the names of its
+    contributors may be used to endorse or promote products derived
+    from this software without specific prior written permission.
+
+    THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
+    "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
+    LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
+    A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
+    OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
+    SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
+    LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
+    DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
+    THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
+    (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
+    OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+
+Flow Classify Sample Application
+================================
+
+The Flow Classify sample application is based on the simple *skeleton* example
+of a forwarding application.
+
+It is intended as a demonstration of the basic components of a DPDK forwarding
+application which uses the Flow Classify library API's.
+
+Please refer to the
+:doc:`../prog_guide/flow_classify_lib`
+for more information.
+
+Compiling the Application
+-------------------------
+
+To compile the sample application see :doc:`compiling`.
+
+The application is located in the ``flow_classify`` sub-directory.
+
+Running the Application
+-----------------------
+
+To run the example in a ``linuxapp`` environment:
+
+.. code-block:: console
+
+    cd ~/dpdk/examples/flow_classify
+    ./build/flow_classify -c 4 -n 4 -- --rule_ipv4="../ipv4_rules_file.txt"
+
+Please refer to the *DPDK Getting Started Guide*, section
+:doc:`../linux_gsg/build_sample_apps`
+for general information on running applications and the Environment Abstraction
+Layer (EAL) options.
+
+
+Sample ipv4_rules_file.txt
+--------------------------
+
+.. code-block:: console
+
+    #file format:
+    #src_ip/masklen dst_ip/masklen src_port : mask dst_port : mask proto/mask priority
+    #
+    2.2.2.3/24 2.2.2.7/24 32 : 0xffff 33 : 0xffff 17/0xff 0
+    9.9.9.3/24 9.9.9.7/24 32 : 0xffff 33 : 0xffff 17/0xff 1
+    9.9.9.3/24 9.9.9.7/24 32 : 0xffff 33 : 0xffff 6/0xff 2
+    9.9.8.3/24 9.9.8.7/24 32 : 0xffff 33 : 0xffff 6/0xff 3
+    6.7.8.9/24 2.3.4.5/24 32 : 0x0000 33 : 0x0000 132/0xff 4
+
+Explanation
+-----------
+
+The following sections provide an explanation of the main components of the
+code.
+
+All DPDK library functions used in the sample code are prefixed with ``rte_``
+and are explained in detail in the *DPDK API Documentation*.
+
+ACL field definitions for the IPv4 5 tuple rule
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+
+The following field definitions are used when creating the ACL table during
+initialisation of the ``Flow Classify`` application..
+
+.. code-block:: c
+
+     enum {
+         PROTO_FIELD_IPV4,
+         SRC_FIELD_IPV4,
+         DST_FIELD_IPV4,
+         SRCP_FIELD_IPV4,
+         DSTP_FIELD_IPV4,
+         NUM_FIELDS_IPV4
+    };
+
+    enum {
+        PROTO_INPUT_IPV4,
+        SRC_INPUT_IPV4,
+        DST_INPUT_IPV4,
+        SRCP_DESTP_INPUT_IPV4
+    };
+
+    static struct rte_acl_field_def ipv4_defs[NUM_FIELDS_IPV4] = {
+        /* first input field - always one byte long. */
+        {
+            .type = RTE_ACL_FIELD_TYPE_BITMASK,
+            .size = sizeof(uint8_t),
+            .field_index = PROTO_FIELD_IPV4,
+            .input_index = PROTO_INPUT_IPV4,
+            .offset = sizeof(struct ether_hdr) +
+                offsetof(struct ipv4_hdr, next_proto_id),
+        },
+        /* next input field (IPv4 source address) - 4 consecutive bytes. */
+        {
+            /* rte_flow uses a bit mask for IPv4 addresses */
+            .type = RTE_ACL_FIELD_TYPE_BITMASK,
+            .size = sizeof(uint32_t),
+            .field_index = SRC_FIELD_IPV4,
+            .input_index = SRC_INPUT_IPV4,
+            .offset = sizeof(struct ether_hdr) +
+                offsetof(struct ipv4_hdr, src_addr),
+        },
+        /* next input field (IPv4 destination address) - 4 consecutive bytes. */
+        {
+            /* rte_flow uses a bit mask for IPv4 addresses */
+            .type = RTE_ACL_FIELD_TYPE_BITMASK,
+            .size = sizeof(uint32_t),
+            .field_index = DST_FIELD_IPV4,
+            .input_index = DST_INPUT_IPV4,
+            .offset = sizeof(struct ether_hdr) +
+                offsetof(struct ipv4_hdr, dst_addr),
+        },
+        /*
+         * Next 2 fields (src & dst ports) form 4 consecutive bytes.
+         * They share the same input index.
+         */
+	{
+            /* rte_flow uses a bit mask for protocol ports */
+            .type = RTE_ACL_FIELD_TYPE_BITMASK,
+            .size = sizeof(uint16_t),
+            .field_index = SRCP_FIELD_IPV4,
+            .input_index = SRCP_DESTP_INPUT_IPV4,
+            .offset = sizeof(struct ether_hdr) +
+                sizeof(struct ipv4_hdr) +
+                offsetof(struct tcp_hdr, src_port),
+        },
+        {
+             /* rte_flow uses a bit mask for protocol ports */
+             .type = RTE_ACL_FIELD_TYPE_BITMASK,
+             .size = sizeof(uint16_t),
+             .field_index = DSTP_FIELD_IPV4,
+             .input_index = SRCP_DESTP_INPUT_IPV4,
+             .offset = sizeof(struct ether_hdr) +
+                 sizeof(struct ipv4_hdr) +
+                 offsetof(struct tcp_hdr, dst_port),
+        },
+    };
+
+The Main Function
+~~~~~~~~~~~~~~~~~
+
+The ``main()`` function performs the initialization and calls the execution
+threads for each lcore.
+
+The first task is to initialize the Environment Abstraction Layer (EAL).
+The ``argc`` and ``argv`` arguments are provided to the ``rte_eal_init()``
+function. The value returned is the number of parsed arguments:
+
+.. code-block:: c
+
+    int ret = rte_eal_init(argc, argv);
+    if (ret < 0)
+        rte_exit(EXIT_FAILURE, "Error with EAL initialization\n");
+
+It then parses the flow_classify application arguments
+
+.. code-block:: c
+
+    ret = parse_args(argc, argv);
+    if (ret < 0)
+        rte_exit(EXIT_FAILURE, "Invalid flow_classify parameters\n");
+
+The ``main()`` function also allocates a mempool to hold the mbufs
+(Message Buffers) used by the application:
+
+.. code-block:: c
+
+    mbuf_pool = rte_mempool_create("MBUF_POOL",
+                                   NUM_MBUFS * nb_ports,
+                                   MBUF_SIZE,
+                                   MBUF_CACHE_SIZE,
+                                   sizeof(struct rte_pktmbuf_pool_private),
+                                   rte_pktmbuf_pool_init, NULL,
+                                   rte_pktmbuf_init, NULL,
+                                   rte_socket_id(),
+                                   0);
+
+mbufs are the packet buffer structure used by DPDK. They are explained in
+detail in the "Mbuf Library" section of the *DPDK Programmer's Guide*.
+
+The ``main()`` function also initializes all the ports using the user defined
+``port_init()`` function which is explained in the next section:
+
+.. code-block:: c
+
+    for (portid = 0; portid < nb_ports; portid++) {
+        if (port_init(portid, mbuf_pool) != 0) {
+            rte_exit(EXIT_FAILURE,
+                     "Cannot init port %" PRIu8 "\n", portid);
+        }
+    }
+
+The ``main()`` function creates the ``flow classifier object`` and adds an ``ACL
+table`` to the flow classifier.
+
+.. code-block:: c
+
+    struct flow_classifier {
+        struct rte_flow_classifier *cls;
+        uint32_t table_id[RTE_FLOW_CLASSIFY_TABLE_MAX];
+    };
+
+    struct flow_classifier_acl {
+        struct flow_classifier cls;
+    } __rte_cache_aligned;
+
+    /* Memory allocation */
+    size = RTE_CACHE_LINE_ROUNDUP(sizeof(struct flow_classifier_acl));
+    cls_app = rte_zmalloc(NULL, size, RTE_CACHE_LINE_SIZE);
+    if (cls_app == NULL)
+        rte_exit(EXIT_FAILURE, "Cannot allocate classifier memory\n");
+
+    cls_params.name = "flow_classifier";
+    cls_params.socket_id = socket_id;
+    cls_params.type = RTE_FLOW_CLASSIFY_TABLE_TYPE_ACL;
+
+    cls_app->cls = rte_flow_classifier_create(&cls_params);
+    if (cls_app->cls == NULL) {
+        rte_free(cls_app);
+        rte_exit(EXIT_FAILURE, "Cannot create classifier\n");
+    }
+
+    /* initialise ACL table params */
+    table_acl_params.name = "table_acl_ipv4_5tuple";
+    table_acl_params.n_rule_fields = RTE_DIM(ipv4_defs);
+    table_acl_params.n_rules = FLOW_CLASSIFY_MAX_RULE_NUM;
+    memcpy(table_acl_params.field_format, ipv4_defs, sizeof(ipv4_defs));
+
+    /* initialise table create params */
+    cls_table_params.ops = &rte_table_acl_ops,
+    cls_table_params.arg_create = &table_acl_params,
+    cls_table_params.table_metadata_size = 0;
+
+    ret = rte_flow_classify_table_create(cls_app->cls, &cls_table_params,
+                  &cls->table_id[0]);
+    if (ret) {
+        rte_flow_classifier_free(cls_app->cls);
+        rte_free(cls);
+        rte_exit(EXIT_FAILURE, "Failed to create classifier table\n");
+    }
+
+It then reads the ipv4_rules_file.txt file and initialises the parameters for
+the ``rte_flow_classify_table_entry_add`` API.
+This API adds a rule to the ACL table.
+
+.. code-block:: c
+
+    if (add_rules(parm_config.rule_ipv4_name)) {
+        rte_flow_classifier_free(cls_app->cls);
+        rte_free(cls_app);
+        rte_exit(EXIT_FAILURE, "Failed to add rules\n");
+    }
+
+Once the initialization is complete, the application is ready to launch a
+function on an lcore. In this example ``lcore_main()`` is called on a single
+lcore.
+
+.. code-block:: c
+
+    lcore_main(cls_app);
+
+The ``lcore_main()`` function is explained below.
+
+The Port Initialization  Function
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+
+The main functional part of the port initialization used in the Basic
+Forwarding application is shown below:
+
+.. code-block:: c
+
+    static inline int
+    port_init(uint8_t port, struct rte_mempool *mbuf_pool)
+    {
+        struct rte_eth_conf port_conf = port_conf_default;
+        const uint16_t rx_rings = 1, tx_rings = 1;
+        struct ether_addr addr;
+        int retval;
+        uint16_t q;
+
+        if (port >= rte_eth_dev_count())
+            return -1;
+
+        /* Configure the Ethernet device. */
+        retval = rte_eth_dev_configure(port, rx_rings, tx_rings, &port_conf);
+        if (retval != 0)
+            return retval;
+
+        /* Allocate and set up 1 RX queue per Ethernet port. */
+        for (q = 0; q < rx_rings; q++) {
+            retval = rte_eth_rx_queue_setup(port, q, RX_RING_SIZE,
+                    rte_eth_dev_socket_id(port), NULL, mbuf_pool);
+            if (retval < 0)
+                return retval;
+        }
+
+        /* Allocate and set up 1 TX queue per Ethernet port. */
+        for (q = 0; q < tx_rings; q++) {
+            retval = rte_eth_tx_queue_setup(port, q, TX_RING_SIZE,
+                    rte_eth_dev_socket_id(port), NULL);
+            if (retval < 0)
+                return retval;
+        }
+
+        /* Start the Ethernet port. */
+        retval = rte_eth_dev_start(port);
+        if (retval < 0)
+            return retval;
+
+        /* Display the port MAC address. */
+        rte_eth_macaddr_get(port, &addr);
+        printf("Port %u MAC: %02" PRIx8 " %02" PRIx8 " %02" PRIx8
+               " %02" PRIx8 " %02" PRIx8 " %02" PRIx8 "\n",
+               port,
+               addr.addr_bytes[0], addr.addr_bytes[1],
+               addr.addr_bytes[2], addr.addr_bytes[3],
+               addr.addr_bytes[4], addr.addr_bytes[5]);
+
+        /* Enable RX in promiscuous mode for the Ethernet device. */
+        rte_eth_promiscuous_enable(port);
+
+        return 0;
+    }
+
+The Ethernet ports are configured with default settings using the
+``rte_eth_dev_configure()`` function and the ``port_conf_default`` struct.
+
+.. code-block:: c
+
+    static const struct rte_eth_conf port_conf_default = {
+        .rxmode = { .max_rx_pkt_len = ETHER_MAX_LEN }
+    };
+
+For this example the ports are set up with 1 RX and 1 TX queue using the
+``rte_eth_rx_queue_setup()`` and ``rte_eth_tx_queue_setup()`` functions.
+
+The Ethernet port is then started:
+
+.. code-block:: c
+
+    retval  = rte_eth_dev_start(port);
+
+
+Finally the RX port is set in promiscuous mode:
+
+.. code-block:: c
+
+    rte_eth_promiscuous_enable(port);
+
+The Add Rules function
+~~~~~~~~~~~~~~~~~~~~~~
+
+The ``add_rules`` function reads the ``ipv4_rules_file.txt`` file and calls the
+``add_classify_rule`` function which calls the
+``rte_flow_classify_table_entry_add`` API.
+
+.. code-block:: c
+
+    static int
+    add_rules(const char *rule_path)
+    {
+        FILE *fh;
+        char buff[LINE_MAX];
+        unsigned int i = 0;
+        unsigned int total_num = 0;
+        struct rte_eth_ntuple_filter ntuple_filter;
+
+        fh = fopen(rule_path, "rb");
+        if (fh == NULL)
+            rte_exit(EXIT_FAILURE, "%s: Open %s failed\n", __func__,
+                     rule_path);
+
+        fseek(fh, 0, SEEK_SET);
+
+        i = 0;
+        while (fgets(buff, LINE_MAX, fh) != NULL) {
+            i++;
+
+            if (is_bypass_line(buff))
+                continue;
+
+            if (total_num >= FLOW_CLASSIFY_MAX_RULE_NUM - 1) {
+                printf("\nINFO: classify rule capacity %d reached\n",
+                       total_num);
+                break;
+            }
+
+            if (parse_ipv4_5tuple_rule(buff, &ntuple_filter) != 0)
+                rte_exit(EXIT_FAILURE,
+                         "%s Line %u: parse rules error\n",
+                         rule_path, i);
+
+            if (add_classify_rule(&ntuple_filter) != 0)
+                rte_exit(EXIT_FAILURE, "add rule error\n");
+
+            total_num++;
+	}
+
+	fclose(fh);
+	return 0;
+    }
+
+
+The Lcore Main function
+~~~~~~~~~~~~~~~~~~~~~~~
+
+As we saw above the ``main()`` function calls an application function on the
+available lcores.
+The ``lcore_main`` function calls the ``rte_flow_classifier_query`` API.
+For the Basic Forwarding application the ``lcore_main`` function looks like the
+following:
+
+.. code-block:: c
+
+    /* flow classify data */
+    static int num_classify_rules;
+    static struct rte_flow_classify_rule *rules[MAX_NUM_CLASSIFY];
+    static struct rte_flow_classify_ipv4_5tuple_stats ntuple_stats;
+    static struct rte_flow_classify_stats classify_stats = {
+            .stats = (void *)&ntuple_stats
+    };
+
+    static __attribute__((noreturn)) void
+    lcore_main(cls_app)
+    {
+        const uint8_t nb_ports = rte_eth_dev_count();
+        uint8_t port;
+
+        /*
+         * Check that the port is on the same NUMA node as the polling thread
+         * for best performance.
+         */
+        for (port = 0; port < nb_ports; port++)
+            if (rte_eth_dev_socket_id(port) > 0 &&
+                rte_eth_dev_socket_id(port) != (int)rte_socket_id()) {
+                printf("\n\n");
+                printf("WARNING: port %u is on remote NUMA node\n",
+                       port);
+                printf("to polling thread.\n");
+                printf("Performance will not be optimal.\n");
+
+                printf("\nCore %u forwarding packets. \n",
+                       rte_lcore_id());
+                printf("[Ctrl+C to quit]\n
+            }
+
+        /* Run until the application is quit or killed. */
+        for (;;) {
+            /*
+             * Receive packets on a port and forward them on the paired
+             * port. The mapping is 0 -> 1, 1 -> 0, 2 -> 3, 3 -> 2, etc.
+             */
+            for (port = 0; port < nb_ports; port++) {
+
+                /* Get burst of RX packets, from first port of pair. */
+                struct rte_mbuf *bufs[BURST_SIZE];
+                const uint16_t nb_rx = rte_eth_rx_burst(port, 0,
+                        bufs, BURST_SIZE);
+
+                if (unlikely(nb_rx == 0))
+                    continue;
+
+                for (i = 0; i < MAX_NUM_CLASSIFY; i++) {
+                    if (rules[i]) {
+                        ret = rte_flow_classifier_query(
+                            cls_app->cls,
+                            cls_app->table_id[0],
+                            bufs, nb_rx, rules[i],
+                            &classify_stats);
+                        if (ret)
+                            printf(
+                                "rule [%d] query failed ret [%d]\n\n",
+                                i, ret);
+                        else {
+                            printf(
+                                "rule[%d] count=%"PRIu64"\n",
+                                i, ntuple_stats.counter1);
+
+                            printf("proto = %d\n",
+                                ntuple_stats.ipv4_5tuple.proto);
+                        }
+                     }
+                 }
+
+                /* Send burst of TX packets, to second port of pair. */
+                const uint16_t nb_tx = rte_eth_tx_burst(port ^ 1, 0,
+                        bufs, nb_rx);
+
+                /* Free any unsent packets. */
+                if (unlikely(nb_tx < nb_rx)) {
+                    uint16_t buf;
+                    for (buf = nb_tx; buf < nb_rx; buf++)
+                        rte_pktmbuf_free(bufs[buf]);
+                }
+            }
+        }
+    }
+
+The main work of the application is done within the loop:
+
+.. code-block:: c
+
+        for (;;) {
+            for (port = 0; port < nb_ports; port++) {
+
+                /* Get burst of RX packets, from first port of pair. */
+                struct rte_mbuf *bufs[BURST_SIZE];
+                const uint16_t nb_rx = rte_eth_rx_burst(port, 0,
+                        bufs, BURST_SIZE);
+
+                if (unlikely(nb_rx == 0))
+                    continue;
+
+                /* Send burst of TX packets, to second port of pair. */
+                const uint16_t nb_tx = rte_eth_tx_burst(port ^ 1, 0,
+                        bufs, nb_rx);
+
+                /* Free any unsent packets. */
+                if (unlikely(nb_tx < nb_rx)) {
+                    uint16_t buf;
+                    for (buf = nb_tx; buf < nb_rx; buf++)
+                        rte_pktmbuf_free(bufs[buf]);
+                }
+            }
+        }
+
+Packets are received in bursts on the RX ports and transmitted in bursts on
+the TX ports. The ports are grouped in pairs with a simple mapping scheme
+using the an XOR on the port number::
+
+    0 -> 1
+    1 -> 0
+
+    2 -> 3
+    3 -> 2
+
+    etc.
+
+The ``rte_eth_tx_burst()`` function frees the memory buffers of packets that
+are transmitted. If packets fail to transmit, ``(nb_tx < nb_rx)``, then they
+must be freed explicitly using ``rte_pktmbuf_free()``.
+
+The forwarding loop can be interrupted and the application closed using
+``Ctrl-C``.
diff --git a/doc/guides/sample_app_ug/index.rst b/doc/guides/sample_app_ug/index.rst
index 1e18d88..b62046a 100644
--- a/doc/guides/sample_app_ug/index.rst
+++ b/doc/guides/sample_app_ug/index.rst
@@ -79,6 +79,7 @@  Sample Applications User Guides
     ptpclient
     performance_thread
     ipsec_secgw
+    flow_classify
 
 **Figures**