test: add flow classify unit tests
authorBernard Iremonger <bernard.iremonger@intel.com>
Tue, 24 Oct 2017 17:28:03 +0000 (18:28 +0100)
committerThomas Monjalon <thomas@monjalon.net>
Tue, 24 Oct 2017 20:26:54 +0000 (22:26 +0200)
Add flow_classify_autotest program.

Set up IPv4 ACL field definitions.
Create table_acl for use by librte_flow_classify API's.
Create an mbuf pool for use by rte_flow_classify_query.

For each of the librte_flow_classify API's:
test with invalid parameters
test with invalid patterns
test with invalid actions
test with valid parameters

Initialise ipv4 udp traffic for use by the udp test for
rte_flow_classifier_run.

Initialise ipv4 tcp traffic for use by the tcp test for
rte_flow_classifier_run.

Initialise ipv4 sctp traffic for use by the sctp test for
rte_flow_classifier_run.

Signed-off-by: Bernard Iremonger <bernard.iremonger@intel.com>
Acked-by: Jasvinder Singh <jasvinder.singh@intel.com>
test/test/Makefile
test/test/test_flow_classify.c [new file with mode: 0644]
test/test/test_flow_classify.h [new file with mode: 0644]

index dcbe363..c2dbe40 100644 (file)
@@ -107,6 +107,7 @@ SRCS-y += test_table_tables.c
 SRCS-y += test_table_ports.c
 SRCS-y += test_table_combined.c
 SRCS-$(CONFIG_RTE_LIBRTE_ACL) += test_table_acl.c
+SRCS-$(CONFIG_RTE_LIBRTE_ACL) += test_flow_classify.c
 endif
 
 SRCS-y += test_rwlock.c
diff --git a/test/test/test_flow_classify.c b/test/test/test_flow_classify.c
new file mode 100644 (file)
index 0000000..9f331cd
--- /dev/null
@@ -0,0 +1,672 @@
+/*-
+ *   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.
+ */
+
+#include <string.h>
+#include <errno.h>
+
+#include "test.h"
+
+#include <rte_string_fns.h>
+#include <rte_mbuf.h>
+#include <rte_byteorder.h>
+#include <rte_ip.h>
+#include <rte_acl.h>
+#include <rte_common.h>
+#include <rte_table_acl.h>
+#include <rte_flow.h>
+#include <rte_flow_classify.h>
+
+#include "packet_burst_generator.h"
+#include "test_flow_classify.h"
+
+
+#define FLOW_CLASSIFY_MAX_RULE_NUM 100
+struct flow_classifier *cls;
+
+struct flow_classifier {
+       struct rte_flow_classifier *cls;
+       uint32_t table_id[RTE_FLOW_CLASSIFY_TABLE_MAX];
+       uint32_t n_tables;
+};
+
+struct flow_classifier_acl {
+       struct flow_classifier cls;
+} __rte_cache_aligned;
+
+/*
+ * test functions by passing invalid or
+ * non-workable parameters.
+ */
+static int
+test_invalid_parameters(void)
+{
+       struct rte_flow_classify_rule *rule;
+       int ret;
+
+       rule = rte_flow_classify_table_entry_add(NULL, 1, NULL, NULL, NULL,
+                       NULL, NULL);
+       if (rule) {
+               printf("Line %i: flow_classifier_table_entry_add", __LINE__);
+               printf(" with NULL param should have failed!\n");
+               return -1;
+       }
+
+       ret = rte_flow_classify_table_entry_delete(NULL, 1, NULL);
+       if (!ret) {
+               printf("Line %i: rte_flow_classify_table_entry_delete",
+                       __LINE__);
+               printf(" with NULL param should have failed!\n");
+               return -1;
+       }
+
+       ret = rte_flow_classifier_query(NULL, 1, NULL, 0, NULL, NULL);
+       if (!ret) {
+               printf("Line %i: flow_classifier_query", __LINE__);
+               printf(" with NULL param should have failed!\n");
+               return -1;
+       }
+
+       rule = rte_flow_classify_table_entry_add(NULL, 1, NULL, NULL, NULL,
+               NULL, &error);
+       if (rule) {
+               printf("Line %i: flow_classify_table_entry_add ", __LINE__);
+               printf("with NULL param should have failed!\n");
+               return -1;
+       }
+
+       ret = rte_flow_classify_table_entry_delete(NULL, 1, NULL);
+       if (!ret) {
+               printf("Line %i: rte_flow_classify_table_entry_delete",
+                       __LINE__);
+               printf("with NULL param should have failed!\n");
+               return -1;
+       }
+
+       ret = rte_flow_classifier_query(NULL, 1, NULL, 0, NULL, NULL);
+       if (!ret) {
+               printf("Line %i: flow_classifier_query", __LINE__);
+               printf(" with NULL param should have failed!\n");
+               return -1;
+       }
+       return 0;
+}
+
+static int
+test_valid_parameters(void)
+{
+       struct rte_flow_classify_rule *rule;
+       int ret;
+       int key_found;
+
+       /*
+        * set up parameters for rte_flow_classify_table_entry_add and
+        * rte_flow_classify_table_entry_delete
+        */
+
+       attr.ingress = 1;
+       attr.priority = 1;
+       pattern[0] = eth_item;
+       pattern[1] = ipv4_udp_item_1;
+       pattern[2] = udp_item_1;
+       pattern[3] = end_item;
+       actions[0] = count_action;
+       actions[1] = end_action;
+
+       rule = rte_flow_classify_table_entry_add(cls->cls, 0, &key_found,
+                       &attr, pattern, actions, &error);
+       if (!rule) {
+               printf("Line %i: flow_classify_table_entry_add", __LINE__);
+               printf(" should not have failed!\n");
+               return -1;
+       }
+
+       ret = rte_flow_classify_table_entry_delete(cls->cls, 0, rule);
+       if (ret) {
+               printf("Line %i: rte_flow_classify_table_entry_delete",
+                       __LINE__);
+               printf(" should not have failed!\n");
+               return -1;
+       }
+       return 0;
+}
+
+static int
+test_invalid_patterns(void)
+{
+       struct rte_flow_classify_rule *rule;
+       int ret;
+       int key_found;
+
+       /*
+        * set up parameters for rte_flow_classify_table_entry_add and
+        * rte_flow_classify_table_entry_delete
+        */
+
+       attr.ingress = 1;
+       attr.priority = 1;
+       pattern[0] = eth_item_bad;
+       pattern[1] = ipv4_udp_item_1;
+       pattern[2] = udp_item_1;
+       pattern[3] = end_item;
+       actions[0] = count_action;
+       actions[1] = end_action;
+
+       pattern[0] = eth_item;
+       pattern[1] = ipv4_udp_item_bad;
+       rule = rte_flow_classify_table_entry_add(cls->cls, 0, &key_found,
+                       &attr, pattern, actions, &error);
+       if (rule) {
+               printf("Line %i: flow_classify_table_entry_add", __LINE__);
+               printf(" should have failed!\n");
+               return -1;
+       }
+
+       ret = rte_flow_classify_table_entry_delete(cls->cls, 0, rule);
+       if (!ret) {
+               printf("Line %i: rte_flow_classify_table_entry_delete",
+                       __LINE__);
+               printf(" should have failed!\n");
+               return -1;
+       }
+
+       pattern[1] = ipv4_udp_item_1;
+       pattern[2] = udp_item_bad;
+       pattern[3] = end_item_bad;
+       rule = rte_flow_classify_table_entry_add(cls->cls, 0, &key_found,
+                       &attr, pattern, actions, &error);
+       if (rule) {
+               printf("Line %i: flow_classify_table_entry_add", __LINE__);
+               printf(" should have failed!\n");
+               return -1;
+       }
+
+       ret = rte_flow_classify_table_entry_delete(cls->cls, 0, rule);
+       if (!ret) {
+               printf("Line %i: rte_flow_classify_table_entry_delete",
+                       __LINE__);
+               printf(" should have failed!\n");
+               return -1;
+       }
+       return 0;
+}
+
+static int
+test_invalid_actions(void)
+{
+       struct rte_flow_classify_rule *rule;
+       int ret;
+       int key_found;
+
+       /*
+        * set up parameters for rte_flow_classify_table_entry_add and
+        * rte_flow_classify_table_entry_delete
+        */
+
+       attr.ingress = 1;
+       attr.priority = 1;
+       pattern[0] = eth_item;
+       pattern[1] = ipv4_udp_item_1;
+       pattern[2] = udp_item_1;
+       pattern[3] = end_item;
+       actions[0] = count_action_bad;
+       actions[1] = end_action;
+
+       rule = rte_flow_classify_table_entry_add(cls->cls, 0, &key_found,
+                       &attr, pattern, actions, &error);
+       if (rule) {
+               printf("Line %i: flow_classify_table_entry_add", __LINE__);
+               printf(" should have failed!\n");
+               return -1;
+       }
+
+       ret = rte_flow_classify_table_entry_delete(cls->cls, 0, rule);
+       if (!ret) {
+               printf("Line %i: rte_flow_classify_table_entry_delete",
+                       __LINE__);
+               printf(" should have failed!\n");
+               return -1;
+       }
+
+       actions[0] = count_action;
+       actions[1] = end_action_bad;
+
+       rule = rte_flow_classify_table_entry_add(cls->cls, 0, &key_found,
+                       &attr, pattern, actions, &error);
+       if (rule) {
+               printf("Line %i: flow_classify_table_entry_add", __LINE__);
+               printf(" should have failed!\n");
+               return -1;
+       }
+
+       ret = rte_flow_classify_table_entry_delete(cls->cls, 0, rule);
+       if (!ret) {
+               printf("Line %i: rte_flow_classify_table_entry_delete",
+                       __LINE__);
+               printf("should have failed!\n");
+               return -1;
+       }
+       return 0;
+}
+
+static int
+init_ipv4_udp_traffic(struct rte_mempool *mp,
+            struct rte_mbuf **pkts_burst, uint32_t burst_size)
+{
+       struct ether_hdr pkt_eth_hdr;
+       struct ipv4_hdr pkt_ipv4_hdr;
+       struct udp_hdr pkt_udp_hdr;
+       uint32_t src_addr = IPV4_ADDR(2, 2, 2, 3);
+       uint32_t dst_addr = IPV4_ADDR(2, 2, 2, 7);
+       uint16_t src_port = 32;
+       uint16_t dst_port = 33;
+       uint16_t pktlen;
+
+       static uint8_t src_mac[] = { 0x00, 0xFF, 0xAA, 0xFF, 0xAA, 0xFF };
+       static uint8_t dst_mac[] = { 0x00, 0xAA, 0xFF, 0xAA, 0xFF, 0xAA };
+
+       printf("Set up IPv4 UDP traffic\n");
+       initialize_eth_header(&pkt_eth_hdr,
+               (struct ether_addr *)src_mac,
+               (struct ether_addr *)dst_mac, ETHER_TYPE_IPv4, 0, 0);
+       pktlen = (uint16_t)(sizeof(struct ether_hdr));
+       printf("ETH  pktlen %u\n", pktlen);
+
+       pktlen = initialize_ipv4_header(&pkt_ipv4_hdr, src_addr, dst_addr,
+                                       pktlen);
+       printf("ETH + IPv4 pktlen %u\n", pktlen);
+
+       pktlen = initialize_udp_header(&pkt_udp_hdr, src_port, dst_port,
+                                       pktlen);
+       printf("ETH + IPv4 + UDP pktlen %u\n\n", pktlen);
+
+       return generate_packet_burst(mp, pkts_burst, &pkt_eth_hdr,
+                                    0, &pkt_ipv4_hdr, 1,
+                                    &pkt_udp_hdr, burst_size,
+                                    PACKET_BURST_GEN_PKT_LEN, 1);
+}
+
+static int
+init_ipv4_tcp_traffic(struct rte_mempool *mp,
+            struct rte_mbuf **pkts_burst, uint32_t burst_size)
+{
+       struct ether_hdr pkt_eth_hdr;
+       struct ipv4_hdr pkt_ipv4_hdr;
+       struct tcp_hdr pkt_tcp_hdr;
+       uint32_t src_addr = IPV4_ADDR(1, 2, 3, 4);
+       uint32_t dst_addr = IPV4_ADDR(5, 6, 7, 8);
+       uint16_t src_port = 16;
+       uint16_t dst_port = 17;
+       uint16_t pktlen;
+
+       static uint8_t src_mac[] = { 0x00, 0xFF, 0xAA, 0xFF, 0xAA, 0xFF };
+       static uint8_t dst_mac[] = { 0x00, 0xAA, 0xFF, 0xAA, 0xFF, 0xAA };
+
+       printf("Set up IPv4 TCP traffic\n");
+       initialize_eth_header(&pkt_eth_hdr,
+               (struct ether_addr *)src_mac,
+               (struct ether_addr *)dst_mac, ETHER_TYPE_IPv4, 0, 0);
+       pktlen = (uint16_t)(sizeof(struct ether_hdr));
+       printf("ETH  pktlen %u\n", pktlen);
+
+       pktlen = initialize_ipv4_header_proto(&pkt_ipv4_hdr, src_addr,
+                                       dst_addr, pktlen, IPPROTO_TCP);
+       printf("ETH + IPv4 pktlen %u\n", pktlen);
+
+       pktlen = initialize_tcp_header(&pkt_tcp_hdr, src_port, dst_port,
+                                       pktlen);
+       printf("ETH + IPv4 + TCP pktlen %u\n\n", pktlen);
+
+       return generate_packet_burst_proto(mp, pkts_burst, &pkt_eth_hdr,
+                                       0, &pkt_ipv4_hdr, 1, IPPROTO_TCP,
+                                       &pkt_tcp_hdr, burst_size,
+                                       PACKET_BURST_GEN_PKT_LEN, 1);
+}
+
+static int
+init_ipv4_sctp_traffic(struct rte_mempool *mp,
+            struct rte_mbuf **pkts_burst, uint32_t burst_size)
+{
+       struct ether_hdr pkt_eth_hdr;
+       struct ipv4_hdr pkt_ipv4_hdr;
+       struct sctp_hdr pkt_sctp_hdr;
+       uint32_t src_addr = IPV4_ADDR(11, 12, 13, 14);
+       uint32_t dst_addr = IPV4_ADDR(15, 16, 17, 18);
+       uint16_t src_port = 10;
+       uint16_t dst_port = 11;
+       uint16_t pktlen;
+
+       static uint8_t src_mac[] = { 0x00, 0xFF, 0xAA, 0xFF, 0xAA, 0xFF };
+       static uint8_t dst_mac[] = { 0x00, 0xAA, 0xFF, 0xAA, 0xFF, 0xAA };
+
+       printf("Set up IPv4 SCTP traffic\n");
+       initialize_eth_header(&pkt_eth_hdr,
+               (struct ether_addr *)src_mac,
+               (struct ether_addr *)dst_mac, ETHER_TYPE_IPv4, 0, 0);
+       pktlen = (uint16_t)(sizeof(struct ether_hdr));
+       printf("ETH  pktlen %u\n", pktlen);
+
+       pktlen = initialize_ipv4_header_proto(&pkt_ipv4_hdr, src_addr,
+                                       dst_addr, pktlen, IPPROTO_SCTP);
+       printf("ETH + IPv4 pktlen %u\n", pktlen);
+
+       pktlen = initialize_sctp_header(&pkt_sctp_hdr, src_port, dst_port,
+                                       pktlen);
+       printf("ETH + IPv4 + SCTP pktlen %u\n\n", pktlen);
+
+       return generate_packet_burst_proto(mp, pkts_burst, &pkt_eth_hdr,
+                                       0, &pkt_ipv4_hdr, 1, IPPROTO_SCTP,
+                                       &pkt_sctp_hdr, burst_size,
+                                       PACKET_BURST_GEN_PKT_LEN, 1);
+}
+
+static int
+init_mbufpool(void)
+{
+       int socketid;
+       int ret = 0;
+       unsigned int lcore_id;
+       char s[64];
+
+       for (lcore_id = 0; lcore_id < RTE_MAX_LCORE; lcore_id++) {
+               if (rte_lcore_is_enabled(lcore_id) == 0)
+                       continue;
+
+               socketid = rte_lcore_to_socket_id(lcore_id);
+               if (socketid >= NB_SOCKETS) {
+                       printf(
+                               "Socket %d of lcore %u is out of range %d\n",
+                               socketid, lcore_id, NB_SOCKETS);
+                       ret = -1;
+                       break;
+               }
+               if (mbufpool[socketid] == NULL) {
+                       snprintf(s, sizeof(s), "mbuf_pool_%d", socketid);
+                       mbufpool[socketid] =
+                               rte_pktmbuf_pool_create(s, NB_MBUF,
+                                       MEMPOOL_CACHE_SIZE, 0, MBUF_SIZE,
+                                       socketid);
+                       if (mbufpool[socketid]) {
+                               printf("Allocated mbuf pool on socket %d\n",
+                                       socketid);
+                       } else {
+                               printf("Cannot init mbuf pool on socket %d\n",
+                                       socketid);
+                               ret = -ENOMEM;
+                               break;
+                       }
+               }
+       }
+       return ret;
+}
+
+static int
+test_query_udp(void)
+{
+       struct rte_flow_error error;
+       struct rte_flow_classify_rule *rule;
+       int ret;
+       int i;
+       int key_found;
+
+       ret = init_ipv4_udp_traffic(mbufpool[0], bufs, MAX_PKT_BURST);
+       if (ret != MAX_PKT_BURST) {
+               printf("Line %i: init_udp_ipv4_traffic has failed!\n",
+                               __LINE__);
+               return -1;
+       }
+
+       for (i = 0; i < MAX_PKT_BURST; i++)
+               bufs[i]->packet_type = RTE_PTYPE_L3_IPV4;
+
+       /*
+        * set up parameters for rte_flow_classify_table_entry_add and
+        * rte_flow_classify_table_entry_delete
+        */
+
+       attr.ingress = 1;
+       attr.priority = 1;
+       pattern[0] = eth_item;
+       pattern[1] = ipv4_udp_item_1;
+       pattern[2] = udp_item_1;
+       pattern[3] = end_item;
+       actions[0] = count_action;
+       actions[1] = end_action;
+
+       rule = rte_flow_classify_table_entry_add(cls->cls, 0, &key_found,
+                       &attr, pattern, actions, &error);
+       if (!rule) {
+               printf("Line %i: flow_classify_table_entry_add", __LINE__);
+               printf(" should not have failed!\n");
+               return -1;
+       }
+
+       ret = rte_flow_classifier_query(cls->cls, 0, bufs, MAX_PKT_BURST,
+                       rule, &udp_classify_stats);
+       if (ret) {
+               printf("Line %i: flow_classifier_query", __LINE__);
+               printf(" should not have failed!\n");
+               return -1;
+       }
+
+       ret = rte_flow_classify_table_entry_delete(cls->cls, 0, rule);
+       if (ret) {
+               printf("Line %i: rte_flow_classify_table_entry_delete",
+                       __LINE__);
+               printf(" should not have failed!\n");
+               return -1;
+       }
+       return 0;
+}
+
+static int
+test_query_tcp(void)
+{
+       struct rte_flow_classify_rule *rule;
+       int ret;
+       int i;
+       int key_found;
+
+       ret = init_ipv4_tcp_traffic(mbufpool[0], bufs, MAX_PKT_BURST);
+       if (ret != MAX_PKT_BURST) {
+               printf("Line %i: init_ipv4_tcp_traffic has failed!\n",
+                               __LINE__);
+               return -1;
+       }
+
+       for (i = 0; i < MAX_PKT_BURST; i++)
+               bufs[i]->packet_type = RTE_PTYPE_L3_IPV4;
+
+       /*
+        * set up parameters for rte_flow_classify_table_entry_add and
+        * rte_flow_classify_table_entry_delete
+        */
+
+       attr.ingress = 1;
+       attr.priority = 1;
+       pattern[0] = eth_item;
+       pattern[1] = ipv4_tcp_item_1;
+       pattern[2] = tcp_item_1;
+       pattern[3] = end_item;
+       actions[0] = count_action;
+       actions[1] = end_action;
+
+       rule = rte_flow_classify_table_entry_add(cls->cls, 0, &key_found,
+                       &attr, pattern, actions, &error);
+       if (!rule) {
+               printf("Line %i: flow_classify_table_entry_add", __LINE__);
+               printf(" should not have failed!\n");
+               return -1;
+       }
+
+       ret = rte_flow_classifier_query(cls->cls, 0, bufs, MAX_PKT_BURST,
+                       rule, &tcp_classify_stats);
+       if (ret) {
+               printf("Line %i: flow_classifier_query", __LINE__);
+               printf(" should not have failed!\n");
+               return -1;
+       }
+
+       ret = rte_flow_classify_table_entry_delete(cls->cls, 0, rule);
+       if (ret) {
+               printf("Line %i: rte_flow_classify_table_entry_delete",
+                       __LINE__);
+               printf(" should not have failed!\n");
+               return -1;
+       }
+       return 0;
+}
+
+static int
+test_query_sctp(void)
+{
+       struct rte_flow_classify_rule *rule;
+       int ret;
+       int i;
+       int key_found;
+
+       ret = init_ipv4_sctp_traffic(mbufpool[0], bufs, MAX_PKT_BURST);
+       if (ret != MAX_PKT_BURST) {
+               printf("Line %i: init_ipv4_tcp_traffic has failed!\n",
+                       __LINE__);
+               return -1;
+       }
+
+       for (i = 0; i < MAX_PKT_BURST; i++)
+               bufs[i]->packet_type = RTE_PTYPE_L3_IPV4;
+
+       /*
+        * set up parameters rte_flow_classify_table_entry_add and
+        * rte_flow_classify_table_entry_delete
+        */
+
+       attr.ingress = 1;
+       attr.priority = 1;
+       pattern[0] = eth_item;
+       pattern[1] = ipv4_sctp_item_1;
+       pattern[2] = sctp_item_1;
+       pattern[3] = end_item;
+       actions[0] = count_action;
+       actions[1] = end_action;
+
+       rule = rte_flow_classify_table_entry_add(cls->cls, 0, &key_found,
+                       &attr, pattern, actions, &error);
+       if (!rule) {
+               printf("Line %i: flow_classify_table_entry_add", __LINE__);
+               printf(" should not have failed!\n");
+               return -1;
+       }
+
+       ret = rte_flow_classifier_query(cls->cls, 0, bufs, MAX_PKT_BURST,
+                       rule, &sctp_classify_stats);
+       if (ret) {
+               printf("Line %i: flow_classifier_query", __LINE__);
+               printf(" should not have failed!\n");
+               return -1;
+       }
+
+       ret = rte_flow_classify_table_entry_delete(cls->cls, 0, rule);
+       if (ret) {
+               printf("Line %i: rte_flow_classify_table_entry_delete",
+                       __LINE__);
+               printf(" should not have failed!\n");
+               return -1;
+       }
+       return 0;
+}
+
+static int
+test_flow_classify(void)
+{
+       struct rte_table_acl_params table_acl_params;
+       struct rte_flow_classify_table_params cls_table_params;
+       struct rte_flow_classifier_params cls_params;
+       int socket_id;
+       int ret;
+       uint32_t size;
+
+       socket_id = rte_eth_dev_socket_id(0);
+
+       /* Memory allocation */
+       size = RTE_CACHE_LINE_ROUNDUP(sizeof(struct flow_classifier_acl));
+       cls = rte_zmalloc(NULL, size, RTE_CACHE_LINE_SIZE);
+
+       cls_params.name = "flow_classifier";
+       cls_params.socket_id = socket_id;
+       cls_params.type = RTE_FLOW_CLASSIFY_TABLE_TYPE_ACL;
+       cls->cls = rte_flow_classifier_create(&cls_params);
+
+       /* initialise ACL table params */
+       table_acl_params.n_rule_fields = RTE_DIM(ipv4_defs);
+       table_acl_params.name = "table_acl_ipv4_5tuple";
+       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,
+
+       ret = rte_flow_classify_table_create(cls->cls, &cls_table_params,
+                       &cls->table_id[0]);
+       if (ret) {
+               printf("Line %i: f_create has failed!\n", __LINE__);
+               rte_flow_classifier_free(cls->cls);
+               rte_free(cls);
+               return -1;
+       }
+       printf("Created table_acl for for IPv4 five tuple packets\n");
+
+       ret = init_mbufpool();
+       if (ret) {
+               printf("Line %i: init_mbufpool has failed!\n", __LINE__);
+               return -1;
+       }
+
+       if (test_invalid_parameters() < 0)
+               return -1;
+       if (test_valid_parameters() < 0)
+               return -1;
+       if (test_invalid_patterns() < 0)
+               return -1;
+       if (test_invalid_actions() < 0)
+               return -1;
+       if (test_query_udp() < 0)
+               return -1;
+       if (test_query_tcp() < 0)
+               return -1;
+       if (test_query_sctp() < 0)
+               return -1;
+
+       return 0;
+}
+
+REGISTER_TEST_COMMAND(flow_classify_autotest, test_flow_classify);
diff --git a/test/test/test_flow_classify.h b/test/test/test_flow_classify.h
new file mode 100644 (file)
index 0000000..39535cf
--- /dev/null
@@ -0,0 +1,234 @@
+/*-
+ *   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.
+ */
+
+#ifndef TEST_FLOW_CLASSIFY_H_
+#define TEST_FLOW_CLASSIFY_H_
+
+#define MAX_PKT_BURST      (32)
+#define NB_SOCKETS         (1)
+#define MEMPOOL_CACHE_SIZE (256)
+#define MBUF_SIZE          (512)
+#define NB_MBUF            (512)
+
+/* test UDP, TCP and SCTP packets */
+static struct rte_mempool *mbufpool[NB_SOCKETS];
+static struct rte_mbuf *bufs[MAX_PKT_BURST];
+
+/* ACL field definitions for IPv4 5 tuple rule */
+
+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),
+       },
+};
+
+/* parameters for rte_flow_classify_validate and rte_flow_classify_create */
+
+/* test UDP pattern:
+ * "eth / ipv4 src spec 2.2.2.3 src mask 255.255.255.00 dst spec 2.2.2.7
+ *  dst mask 255.255.255.00 / udp src is 32 dst is 33 / end"
+ */
+static struct rte_flow_item_ipv4 ipv4_udp_spec_1 = {
+       { 0, 0, 0, 0, 0, 0, IPPROTO_UDP, 0, IPv4(2, 2, 2, 3), IPv4(2, 2, 2, 7)}
+};
+static const struct rte_flow_item_ipv4 ipv4_mask_24 = {
+       .hdr = {
+               .next_proto_id = 0xff,
+               .src_addr = 0xffffff00,
+               .dst_addr = 0xffffff00,
+       },
+};
+static struct rte_flow_item_udp udp_spec_1 = {
+       { 32, 33, 0, 0 }
+};
+
+static struct rte_flow_item  eth_item = { RTE_FLOW_ITEM_TYPE_ETH,
+       0, 0, 0 };
+static struct rte_flow_item  eth_item_bad = { -1, 0, 0, 0 };
+
+static struct rte_flow_item  ipv4_udp_item_1 = { RTE_FLOW_ITEM_TYPE_IPV4,
+       &ipv4_udp_spec_1, 0, &ipv4_mask_24};
+static struct rte_flow_item  ipv4_udp_item_bad = { RTE_FLOW_ITEM_TYPE_IPV4,
+       NULL, 0, NULL};
+
+static struct rte_flow_item  udp_item_1 = { RTE_FLOW_ITEM_TYPE_UDP,
+       &udp_spec_1, 0, &rte_flow_item_udp_mask};
+static struct rte_flow_item  udp_item_bad = { RTE_FLOW_ITEM_TYPE_UDP,
+       NULL, 0, NULL};
+
+static struct rte_flow_item  end_item = { RTE_FLOW_ITEM_TYPE_END,
+       0, 0, 0 };
+static struct rte_flow_item  end_item_bad = { -1, 0, 0, 0 };
+
+/* test TCP pattern:
+ * "eth / ipv4 src spec 1.2.3.4 src mask 255.255.255.00 dst spec 5.6.7.8
+ *  dst mask 255.255.255.00 / tcp src is 16 dst is 17 / end"
+ */
+static struct rte_flow_item_ipv4 ipv4_tcp_spec_1 = {
+       { 0, 0, 0, 0, 0, 0, IPPROTO_TCP, 0, IPv4(1, 2, 3, 4), IPv4(5, 6, 7, 8)}
+};
+
+static struct rte_flow_item_tcp tcp_spec_1 = {
+       { 16, 17, 0, 0, 0, 0, 0, 0, 0}
+};
+
+static struct rte_flow_item  ipv4_tcp_item_1 = { RTE_FLOW_ITEM_TYPE_IPV4,
+       &ipv4_tcp_spec_1, 0, &ipv4_mask_24};
+
+static struct rte_flow_item  tcp_item_1 = { RTE_FLOW_ITEM_TYPE_TCP,
+       &tcp_spec_1, 0, &rte_flow_item_tcp_mask};
+
+/* test SCTP pattern:
+ * "eth / ipv4 src spec 1.2.3.4 src mask 255.255.255.00 dst spec 5.6.7.8
+ *  dst mask 255.255.255.00 / sctp src is 16 dst is 17/ end"
+ */
+static struct rte_flow_item_ipv4 ipv4_sctp_spec_1 = {
+       { 0, 0, 0, 0, 0, 0, IPPROTO_SCTP, 0, IPv4(11, 12, 13, 14),
+       IPv4(15, 16, 17, 18)}
+};
+
+static struct rte_flow_item_sctp sctp_spec_1 = {
+       { 10, 11, 0, 0}
+};
+
+static struct rte_flow_item  ipv4_sctp_item_1 = { RTE_FLOW_ITEM_TYPE_IPV4,
+       &ipv4_sctp_spec_1, 0, &ipv4_mask_24};
+
+static struct rte_flow_item  sctp_item_1 = { RTE_FLOW_ITEM_TYPE_SCTP,
+       &sctp_spec_1, 0, &rte_flow_item_sctp_mask};
+
+
+/* test actions:
+ * "actions count / end"
+ */
+static struct rte_flow_action count_action = { RTE_FLOW_ACTION_TYPE_COUNT, 0};
+static struct rte_flow_action count_action_bad = { -1, 0};
+
+static struct rte_flow_action end_action = { RTE_FLOW_ACTION_TYPE_END, 0};
+static struct rte_flow_action end_action_bad = { -1, 0};
+
+static struct rte_flow_action actions[2];
+
+/* test attributes */
+static struct rte_flow_attr attr;
+
+/* test error */
+static struct rte_flow_error error;
+
+/* test pattern */
+static struct rte_flow_item  pattern[4];
+
+/* flow classify data for UDP burst */
+static struct rte_flow_classify_ipv4_5tuple_stats udp_ntuple_stats;
+static struct rte_flow_classify_stats udp_classify_stats = {
+               .stats = (void *)&udp_ntuple_stats
+};
+
+/* flow classify data for TCP burst */
+static struct rte_flow_classify_ipv4_5tuple_stats tcp_ntuple_stats;
+static struct rte_flow_classify_stats tcp_classify_stats = {
+               .stats = (void *)&tcp_ntuple_stats
+};
+
+/* flow classify data for SCTP burst */
+static struct rte_flow_classify_ipv4_5tuple_stats sctp_ntuple_stats;
+static struct rte_flow_classify_stats sctp_classify_stats = {
+               .stats = (void *)&sctp_ntuple_stats
+};
+#endif /* TEST_FLOW_CLASSIFY_H_ */