From: Ian Betts <ian.be...@intel.com> This commit adds an l3fwd derivative that allows multiple EAL threads to be run on a single physical core or multiple lightwieght threads to be run in an EAL thread.
Its purpose is to facilitate characterization of performance with different threading models. It depends on a simple cooperative scheduler included in this patchset. Signed-off-by: Ian Betts <ian.betts at intel.com> --- examples/performance-thread/l3fwd-thread/Makefile | 57 + examples/performance-thread/l3fwd-thread/main.c | 3355 +++++++++++++++++++++ 2 files changed, 3412 insertions(+) create mode 100644 examples/performance-thread/l3fwd-thread/Makefile create mode 100644 examples/performance-thread/l3fwd-thread/main.c diff --git a/examples/performance-thread/l3fwd-thread/Makefile b/examples/performance-thread/l3fwd-thread/Makefile new file mode 100644 index 0000000..d8fe5e6 --- /dev/null +++ b/examples/performance-thread/l3fwd-thread/Makefile @@ -0,0 +1,57 @@ +# BSD LICENSE +# +# Copyright(c) 2010-2015 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. + +ifeq ($(RTE_SDK),) +$(error "Please define RTE_SDK environment variable") +endif + +# Default target, can be overridden by command line or environment +RTE_TARGET ?= x86_64-native-linuxapp-gcc + +include $(RTE_SDK)/mk/rte.vars.mk + +# binary name +APP = l3fwd-thread + +# all source are stored in SRCS-y +SRCS-y := main.c + +include $(RTE_SDK)/examples/performance-thread/common/common.mk + +CFLAGS += -O3 -g $(USER_FLAGS) $(INCLUDES) $(WERROR_FLAGS) + +# workaround for a gcc bug with noreturn attribute +# http://gcc.gnu.org/bugzilla/show_bug.cgi?id=12603 +#ifeq ($(CONFIG_RTE_TOOLCHAIN_GCC),y) +CFLAGS_main.o += -Wno-return-type +#endif + +include $(RTE_SDK)/mk/rte.extapp.mk diff --git a/examples/performance-thread/l3fwd-thread/main.c b/examples/performance-thread/l3fwd-thread/main.c new file mode 100644 index 0000000..2708ec6 --- /dev/null +++ b/examples/performance-thread/l3fwd-thread/main.c @@ -0,0 +1,3355 @@ +/* + * BSD LICENSE + * + * Copyright(c) 2010-2015 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 <stdio.h> +#include <stdlib.h> +#include <stdint.h> +#include <inttypes.h> +#include <sys/types.h> +#include <string.h> +#include <sys/queue.h> +#include <stdarg.h> +#include <errno.h> +#include <getopt.h> + +#include <rte_common.h> +#include <rte_vect.h> +#include <rte_byteorder.h> +#include <rte_log.h> +#include <rte_memory.h> +#include <rte_memcpy.h> +#include <rte_memzone.h> +#include <rte_eal.h> +#include <rte_per_lcore.h> +#include <rte_launch.h> +#include <rte_atomic.h> +#include <rte_cycles.h> +#include <rte_prefetch.h> +#include <rte_lcore.h> +#include <rte_branch_prediction.h> +#include <rte_interrupts.h> +#include <rte_pci.h> +#include <rte_random.h> +#include <rte_debug.h> +#include <rte_ether.h> +#include <rte_ethdev.h> +#include <rte_ring.h> +#include <rte_mempool.h> +#include <rte_mbuf.h> +#include <rte_ip.h> +#include <rte_tcp.h> +#include <rte_udp.h> +#include <rte_string_fns.h> +#include <lthread_api.h> + +#include <cmdline_parse.h> +#include <cmdline_parse_etheraddr.h> + +#define APP_LOOKUP_EXACT_MATCH 0 +#define APP_LOOKUP_LPM 1 +#define DO_RFC_1812_CHECKS + +#ifndef APP_LOOKUP_METHOD +#define APP_LOOKUP_METHOD APP_LOOKUP_LPM +#endif + +/* + * When set to zero, simple forwarding path is enabled. + * When set to one, optimized forwarding path is enabled. + * Note that LPM optimization path uses SSE4.1 instructions. + */ +#if ((APP_LOOKUP_METHOD == APP_LOOKUP_LPM) && !defined(__SSE4_1__)) +#define ENABLE_MULTI_BUFFER_OPTIMIZE 0 +#else +#define ENABLE_MULTI_BUFFER_OPTIMIZE 1 +#endif + +#if (APP_LOOKUP_METHOD == APP_LOOKUP_EXACT_MATCH) +#include <rte_hash.h> +#elif (APP_LOOKUP_METHOD == APP_LOOKUP_LPM) +#include <rte_lpm.h> +#include <rte_lpm6.h> +#else +#error "APP_LOOKUP_METHOD set to incorrect value" +#endif + +#ifndef IPv6_BYTES +#define IPv6_BYTES_FMT "%02x%02x:%02x%02x:%02x%02x:%02x%02x:"\ + "%02x%02x:%02x%02x:%02x%02x:%02x%02x" +#define IPv6_BYTES(addr) \ + addr[0], addr[1], addr[2], addr[3], \ + addr[4], addr[5], addr[6], addr[7], \ + addr[8], addr[9], addr[10], addr[11],\ + addr[12], addr[13], addr[14], addr[15] +#endif + +#define RTE_LOGTYPE_L3FWD RTE_LOGTYPE_USER1 +#define MAX_JUMBO_PKT_LEN 9600 +#define IPV6_ADDR_LEN 16 +#define MEMPOOL_CACHE_SIZE 256 + +/* + * This expression is used to calculate the number of mbufs needed depending on + * user input, taking into account memory for rx and tx hardware rings, cache + * per lcore and mtable per port per lcore. + * RTE_MAX is used to ensure that NB_MBUF never goes below a minimum value + * of 8192 + */ + +#define NB_MBUF RTE_MAX( \ + (nb_ports * nb_rx_queue * RTE_TEST_RX_DESC_DEFAULT + \ + nb_ports * nb_lcores * MAX_PKT_BURST + \ + nb_ports * n_tx_queue * RTE_TEST_TX_DESC_DEFAULT + \ + nb_lcores * MEMPOOL_CACHE_SIZE), \ + (unsigned)8192) + +#define MAX_PKT_BURST 1024 +#define BURST_TX_DRAIN_NS 100000 /* TX drain every ~100000ns */ + +/* + * Try to avoid TX buffering if we have at least MAX_TX_BURST packets to send. + */ +#define MAX_TX_BURST (MAX_PKT_BURST / 2) + +#define BURST_SIZE 512 + +#define NB_SOCKETS 8 + +/* Configure how many packets ahead to prefetch, when reading packets */ +#define PREFETCH_OFFSET 3 + +/* Used to mark destination port as 'invalid'. */ +#define BAD_PORT ((uint16_t)-1) + +#define FWDSTEP 4 + +/* + * Configurable number of RX/TX ring descriptors + */ +#define RTE_TEST_RX_DESC_DEFAULT 128 +#define RTE_TEST_TX_DESC_DEFAULT 512 +static uint16_t nb_rxd = RTE_TEST_RX_DESC_DEFAULT; +static uint16_t nb_txd = RTE_TEST_TX_DESC_DEFAULT; + +/* ethernet addresses of ports */ +static uint64_t dest_eth_addr[RTE_MAX_ETHPORTS]; +static struct ether_addr ports_eth_addr[RTE_MAX_ETHPORTS]; + +static __m128i val_eth[RTE_MAX_ETHPORTS]; + +/* replace first 12B of the ethernet header. */ +#define MASK_ETH 0x3f + +/* mask of enabled ports */ +static uint32_t enabled_port_mask; +static int promiscuous_on; /**< Ports set in promiscuous mode off by default. */ +static int numa_on = 1; /**< NUMA is enabled by default. */ + +#if (APP_LOOKUP_METHOD == APP_LOOKUP_EXACT_MATCH) +static int ipv6; /**< ipv6 is false by default. */ +#endif + +struct mbuf_table { + uint16_t len; + struct rte_mbuf *m_table[MAX_PKT_BURST]; +}; + +struct thread_rx_queue { + uint8_t port_id; + uint8_t queue_id; +} __rte_cache_aligned; + +#define MAX_RX_QUEUE_PER_THREAD 16 +#define MAX_TX_PORT_PER_THREAD RTE_MAX_ETHPORTS +#define MAX_TX_QUEUE_PER_PORT RTE_MAX_ETHPORTS +#define MAX_RX_QUEUE_PER_PORT 128 + +#define MAX_THREAD_PARAMS 1024 + +/** + * Rx params + */ +struct thread_rx_params { + uint8_t port_id; + uint8_t queue_id; + uint8_t lcore_id; + uint8_t thread_id; +} __rte_cache_aligned; + +static struct thread_rx_params thread_rx_params_array[MAX_THREAD_PARAMS]; +static struct thread_rx_params thread_rx_params_array_default[] = { + {0, 0, 0, 0}, + {1, 0, 1, 1}, +}; + +static struct thread_rx_params *thread_rx_params = + thread_rx_params_array_default; +static uint16_t nb_thread_rx_params = RTE_DIM(thread_rx_params_array_default); + +/** + * Tx params + */ +struct thread_tx_params { + uint8_t port_id; /**< Default port */ + uint8_t lcore_id; /**< Initial lcore */ + uint8_t thread_id; /**< Tx thread index */ +} __rte_cache_aligned; + +static struct thread_tx_params thread_tx_params_array[MAX_THREAD_PARAMS]; +static struct thread_tx_params thread_tx_params_array_default[] = { + {0, 2, 0}, + {1, 3, 1}, + {0, 4, 2}, + {1, 5, 3}, + {0, 6, 4}, + {1, 7, 5}, +}; + +static struct thread_tx_params *thread_tx_params = + thread_tx_params_array_default; +static uint16_t nb_thread_tx_params = + RTE_DIM(thread_tx_params_array_default); + + +static struct rte_eth_conf port_conf = { + .rxmode = { + .mq_mode = ETH_MQ_RX_RSS, + .max_rx_pkt_len = ETHER_MAX_LEN, + .split_hdr_size = 0, + .header_split = 0, /**< Header Split disabled */ + .hw_ip_checksum = 1, /**< IP checksum offload enabled */ + .hw_vlan_filter = 0, /**< VLAN filtering disabled */ + .jumbo_frame = 0, /**< Jumbo Frame Support disabled */ + .hw_strip_crc = 0, /**< CRC stripped by hardware */ + }, + .rx_adv_conf = { + .rss_conf = { + .rss_key = NULL, + .rss_hf = ETH_RSS_IP, + }, + }, + .txmode = { + .mq_mode = ETH_MQ_TX_NONE, + }, +}; + +static struct rte_mempool *pktmbuf_pool[NB_SOCKETS]; + +#if (APP_LOOKUP_METHOD == APP_LOOKUP_EXACT_MATCH) + +#ifdef RTE_MACHINE_CPUFLAG_SSE4_2 +#include <rte_hash_crc.h> +#define DEFAULT_HASH_FUNC rte_hash_crc +#else +#include <rte_jhash.h> +#define DEFAULT_HASH_FUNC rte_jhash +#endif + +struct ipv4_5tuple { + uint32_t ip_dst; + uint32_t ip_src; + uint16_t port_dst; + uint16_t port_src; + uint8_t proto; +} __attribute__((__packed__)); + +union ipv4_5tuple_host { + struct { + uint8_t pad0; + uint8_t proto; + uint16_t pad1; + uint32_t ip_src; + uint32_t ip_dst; + uint16_t port_src; + uint16_t port_dst; + }; + __m128i xmm; +}; + +#define XMM_NUM_IN_IPV6_5TUPLE 3 + +struct ipv6_5tuple { + uint8_t ip_dst[IPV6_ADDR_LEN]; + uint8_t ip_src[IPV6_ADDR_LEN]; + uint16_t port_dst; + uint16_t port_src; + uint8_t proto; +} __attribute__((__packed__)); + +union ipv6_5tuple_host { + struct { + uint16_t pad0; + uint8_t proto; + uint8_t pad1; + uint8_t ip_src[IPV6_ADDR_LEN]; + uint8_t ip_dst[IPV6_ADDR_LEN]; + uint16_t port_src; + uint16_t port_dst; + uint64_t reserve; + }; + __m128i xmm[XMM_NUM_IN_IPV6_5TUPLE]; +}; + +struct ipv4_l3fwd_route { + struct ipv4_5tuple key; + uint8_t if_out; +}; + +struct ipv6_l3fwd_route { + struct ipv6_5tuple key; + uint8_t if_out; +}; + +static struct ipv4_l3fwd_route ipv4_l3fwd_route_array[] = { + {{IPv4(101, 0, 0, 0), IPv4(100, 10, 0, 1), 101, 11, IPPROTO_TCP}, 0}, + {{IPv4(201, 0, 0, 0), IPv4(200, 20, 0, 1), 102, 12, IPPROTO_TCP}, 1}, + {{IPv4(111, 0, 0, 0), IPv4(100, 30, 0, 1), 101, 11, IPPROTO_TCP}, 2}, + {{IPv4(211, 0, 0, 0), IPv4(200, 40, 0, 1), 102, 12, IPPROTO_TCP}, 3}, +}; + +static struct ipv6_l3fwd_route ipv6_l3fwd_route_array[] = { + {{ + {0xfe, 0x80, 0, 0, 0, 0, 0, 0, 0x02, 0x1e, 0x67, 0xff, 0xfe, 0, 0, 0}, + {0xfe, 0x80, 0, 0, 0, 0, 0, 0, 0x02, 0x1b, 0x21, 0xff, 0xfe, 0x91, 0x38, + 0x05}, + 101, 11, IPPROTO_TCP}, 0}, + + {{ + {0xfe, 0x90, 0, 0, 0, 0, 0, 0, 0x02, 0x1e, 0x67, 0xff, 0xfe, 0, 0, 0}, + {0xfe, 0x90, 0, 0, 0, 0, 0, 0, 0x02, 0x1b, 0x21, 0xff, 0xfe, 0x91, 0x38, + 0x05}, + 102, 12, IPPROTO_TCP}, 1}, + + {{ + {0xfe, 0xa0, 0, 0, 0, 0, 0, 0, 0x02, 0x1e, 0x67, 0xff, 0xfe, 0, 0, 0}, + {0xfe, 0xa0, 0, 0, 0, 0, 0, 0, 0x02, 0x1b, 0x21, 0xff, 0xfe, 0x91, 0x38, + 0x05}, + 101, 11, IPPROTO_TCP}, 2}, + + {{ + {0xfe, 0xb0, 0, 0, 0, 0, 0, 0, 0x02, 0x1e, 0x67, 0xff, 0xfe, 0, 0, 0}, + {0xfe, 0xb0, 0, 0, 0, 0, 0, 0, 0x02, 0x1b, 0x21, 0xff, 0xfe, 0x91, 0x38, + 0x05}, + 102, 12, IPPROTO_TCP}, 3}, +}; + +typedef struct rte_hash lookup_struct_t; +static lookup_struct_t *ipv4_l3fwd_lookup_struct[NB_SOCKETS]; +static lookup_struct_t *ipv6_l3fwd_lookup_struct[NB_SOCKETS]; + +#ifdef RTE_ARCH_X86_64 +/* default to 4 million hash entries (approx) */ +#define L3FWD_HASH_ENTRIES (1024*1024*4) +#else +/* 32-bit has less address-space for hugepage memory, limit to 1M entries */ +#define L3FWD_HASH_ENTRIES (1024*1024*1) +#endif +#define HASH_ENTRY_NUMBER_DEFAULT 4 + +static uint32_t hash_entry_number = HASH_ENTRY_NUMBER_DEFAULT; + +static inline uint32_t +ipv4_hash_crc(const void *data, __rte_unused uint32_t data_len, + uint32_t init_val) +{ + const union ipv4_5tuple_host *k; + uint32_t t; + const uint32_t *p; + + k = data; + t = k->proto; + p = (const uint32_t *)&k->port_src; + +#ifdef RTE_MACHINE_CPUFLAG_SSE4_2 + init_val = rte_hash_crc_4byte(t, init_val); + init_val = rte_hash_crc_4byte(k->ip_src, init_val); + init_val = rte_hash_crc_4byte(k->ip_dst, init_val); + init_val = rte_hash_crc_4byte(*p, init_val); +#else /* RTE_MACHINE_CPUFLAG_SSE4_2 */ + init_val = rte_jhash_1word(t, init_val); + init_val = rte_jhash_1word(k->ip_src, init_val); + init_val = rte_jhash_1word(k->ip_dst, init_val); + init_val = rte_jhash_1word(*p, init_val); +#endif /* RTE_MACHINE_CPUFLAG_SSE4_2 */ + return init_val; +} + +static inline uint32_t +ipv6_hash_crc(const void *data, __rte_unused uint32_t data_len, + uint32_t init_val) +{ + const union ipv6_5tuple_host *k; + uint32_t t; + const uint32_t *p; +#ifdef RTE_MACHINE_CPUFLAG_SSE4_2 + const uint32_t *ip_src0, *ip_src1, *ip_src2, *ip_src3; + const uint32_t *ip_dst0, *ip_dst1, *ip_dst2, *ip_dst3; +#endif /* RTE_MACHINE_CPUFLAG_SSE4_2 */ + + k = data; + t = k->proto; + p = (const uint32_t *)&k->port_src; + +#ifdef RTE_MACHINE_CPUFLAG_SSE4_2 + ip_src0 = (const uint32_t *) k->ip_src; + ip_src1 = (const uint32_t *)(k->ip_src+4); + ip_src2 = (const uint32_t *)(k->ip_src+8); + ip_src3 = (const uint32_t *)(k->ip_src+12); + ip_dst0 = (const uint32_t *) k->ip_dst; + ip_dst1 = (const uint32_t *)(k->ip_dst+4); + ip_dst2 = (const uint32_t *)(k->ip_dst+8); + ip_dst3 = (const uint32_t *)(k->ip_dst+12); + init_val = rte_hash_crc_4byte(t, init_val); + init_val = rte_hash_crc_4byte(*ip_src0, init_val); + init_val = rte_hash_crc_4byte(*ip_src1, init_val); + init_val = rte_hash_crc_4byte(*ip_src2, init_val); + init_val = rte_hash_crc_4byte(*ip_src3, init_val); + init_val = rte_hash_crc_4byte(*ip_dst0, init_val); + init_val = rte_hash_crc_4byte(*ip_dst1, init_val); + init_val = rte_hash_crc_4byte(*ip_dst2, init_val); + init_val = rte_hash_crc_4byte(*ip_dst3, init_val); + init_val = rte_hash_crc_4byte(*p, init_val); +#else /* RTE_MACHINE_CPUFLAG_SSE4_2 */ + init_val = rte_jhash_1word(t, init_val); + init_val = rte_jhash(k->ip_src, sizeof(uint8_t) * IPV6_ADDR_LEN, init_val); + init_val = rte_jhash(k->ip_dst, sizeof(uint8_t) * IPV6_ADDR_LEN, init_val); + init_val = rte_jhash_1word(*p, init_val); +#endif /* RTE_MACHINE_CPUFLAG_SSE4_2 */ + return init_val; +} + +#define IPV4_L3FWD_NUM_ROUTES RTE_DIM(ipv4_l3fwd_route_array) +#define IPV6_L3FWD_NUM_ROUTES RTE_DIM(ipv6_l3fwd_route_array) + +static uint8_t ipv4_l3fwd_out_if[L3FWD_HASH_ENTRIES] __rte_cache_aligned; +static uint8_t ipv6_l3fwd_out_if[L3FWD_HASH_ENTRIES] __rte_cache_aligned; + +#endif + +#if (APP_LOOKUP_METHOD == APP_LOOKUP_LPM) +struct ipv4_l3fwd_route { + uint32_t ip; + uint8_t depth; + uint8_t if_out; +}; + +struct ipv6_l3fwd_route { + uint8_t ip[16]; + uint8_t depth; + uint8_t if_out; +}; + +static struct ipv4_l3fwd_route ipv4_l3fwd_route_array[] = { + {IPv4(1, 1, 1, 0), 24, 0}, + {IPv4(2, 1, 1, 0), 24, 1}, + {IPv4(3, 1, 1, 0), 24, 2}, + {IPv4(4, 1, 1, 0), 24, 3}, + {IPv4(5, 1, 1, 0), 24, 4}, + {IPv4(6, 1, 1, 0), 24, 5}, + {IPv4(7, 1, 1, 0), 24, 6}, + {IPv4(8, 1, 1, 0), 24, 7}, +}; + +static struct ipv6_l3fwd_route ipv6_l3fwd_route_array[] = { + {{1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1}, 48, 0}, + {{2, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1}, 48, 1}, + {{3, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1}, 48, 2}, + {{4, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1}, 48, 3}, + {{5, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1}, 48, 4}, + {{6, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1}, 48, 5}, + {{7, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1}, 48, 6}, + {{8, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1}, 48, 7}, +}; + +#define IPV4_L3FWD_NUM_ROUTES RTE_DIM(ipv4_l3fwd_route_array) +#define IPV6_L3FWD_NUM_ROUTES RTE_DIM(ipv6_l3fwd_route_array) + +#define IPV4_L3FWD_LPM_MAX_RULES 1024 +#define IPV6_L3FWD_LPM_MAX_RULES 1024 +#define IPV6_L3FWD_LPM_NUMBER_TBL8S (1 << 16) + +typedef struct rte_lpm lookup_struct_t; +typedef struct rte_lpm6 lookup6_struct_t; +static lookup_struct_t *ipv4_l3fwd_lookup_struct[NB_SOCKETS]; +static lookup6_struct_t *ipv6_l3fwd_lookup_struct[NB_SOCKETS]; +#endif + +#define MAX_RX_THREAD 8 +#define MAX_TX_THREAD 8 +#define MAX_THREAD (MAX_RX_THREAD + MAX_TX_THREAD) + +/** + * Producers and consumers threads configuration + */ +static int lthreads_on = 1; /**< Use lthreads for processing*/ + +struct thread_rx_conf { + uint16_t lcore_id; /**< Initial lcore for rx thread */ + uint16_t thread_id; /**< Rx thread index */ + + uint16_t n_queue; /**< Number of queues to read by one thread */ + struct thread_rx_queue queue[MAX_RX_QUEUE_PER_THREAD]; + + uint16_t n_ring; /**< Number of output rings */ + struct rte_ring *ring[RTE_MAX_LCORE]; + struct lthread_cond *ready[RTE_MAX_LCORE]; + +} __rte_cache_aligned; + +struct thread_tx_conf { + uint16_t lcore_id; /**< Initial lcore for tx thread*/ + uint16_t thread_id; /**< Tx thread index */ + + uint16_t n_port_id; /**< Number of output ports */ + uint16_t port_id[MAX_TX_PORT_PER_THREAD]; /**< Default port_id */ + + uint16_t n_ring; /**< Number of input rings */ + struct rte_ring *ring[RTE_MAX_LCORE]; + struct lthread_cond **ready[RTE_MAX_LCORE]; + + uint16_t queue_id[RTE_MAX_ETHPORTS]; + struct mbuf_table mbuf[RTE_MAX_ETHPORTS]; + +} __rte_cache_aligned; + +struct thread_tx_ring_conf { + uint16_t ring_id; + struct thread_tx_conf *tx_conf; +} __rte_cache_aligned; + +/** + * lcore configuration + */ +struct lcore_conf { + void *data; + lookup_struct_t *ipv4_lookup_struct; +#if (APP_LOOKUP_METHOD == APP_LOOKUP_LPM) + lookup6_struct_t *ipv6_lookup_struct; +#else + lookup_struct_t *ipv6_lookup_struct; +#endif +} __rte_cache_aligned; + +static struct lcore_conf lcore_conf[RTE_MAX_LCORE]; +RTE_DEFINE_PER_LCORE(struct lcore_conf *, lcore_conf); + +uint16_t n_rx_thread; +struct thread_rx_conf rx_thread[MAX_RX_THREAD]; + +uint16_t n_tx_thread; +struct thread_tx_conf tx_thread[MAX_TX_THREAD]; + +/* Send burst of packets on an output interface */ +static inline int +send_burst(struct thread_tx_conf *tx_conf, uint16_t n, uint8_t port) +{ + struct rte_mbuf **m_table; + int ret; + uint16_t queueid; + + queueid = tx_conf->queue_id[port]; + m_table = (struct rte_mbuf **)tx_conf->mbuf[port].m_table; + + ret = rte_eth_tx_burst(port, queueid, m_table, n); + if (unlikely(ret < n)) + do { + rte_pktmbuf_free(m_table[ret]); + } while (++ret < n); + + return 0; +} + +/* Enqueue a single packet, and send burst if queue is filled */ +static inline int +send_single_packet(struct rte_mbuf *m, uint8_t port) +{ + uint16_t len; + struct thread_tx_conf *tx_conf; + + if (lthreads_on) + tx_conf = (struct thread_tx_conf *)lthread_get_data(); + else + tx_conf = (struct thread_tx_conf *)RTE_PER_LCORE(lcore_conf)->data; + + len = tx_conf->mbuf[port].len; + tx_conf->mbuf[port].m_table[len] = m; + len++; + + /* enough pkts to be sent */ + if (unlikely(len == MAX_PKT_BURST)) { + send_burst(tx_conf, MAX_PKT_BURST, port); + len = 0; + } + + tx_conf->mbuf[port].len = len; + return 0; +} + +static inline __attribute__((always_inline)) void +send_packetsx4(struct thread_tx_conf *tx_conf, uint8_t port, + struct rte_mbuf *m[], uint32_t num) +{ + uint32_t len, j, n; + + len = tx_conf->mbuf[port].len; + + /* + * If TX buffer for that queue is empty, and we have enough packets, + * then send them straightway. + */ + if (num >= MAX_TX_BURST && len == 0) { + n = rte_eth_tx_burst(port, tx_conf->queue_id[port], m, num); + if (unlikely(n < num)) { + do { + rte_pktmbuf_free(m[n]); + } while (++n < num); + } + return; + } + + /* + * Put packets into TX buffer for that queue. + */ + + n = len + num; + n = (n > MAX_PKT_BURST) ? MAX_PKT_BURST - len : num; + + j = 0; + switch (n % FWDSTEP) { + while (j < n) { + case 0: + tx_conf->mbuf[port].m_table[len + j] = m[j]; + j++; + case 3: + tx_conf->mbuf[port].m_table[len + j] = m[j]; + j++; + case 2: + tx_conf->mbuf[port].m_table[len + j] = m[j]; + j++; + case 1: + tx_conf->mbuf[port].m_table[len + j] = m[j]; + j++; + } + } + + len += n; + + /* enough pkts to be sent */ + if (unlikely(len == MAX_PKT_BURST)) { + + send_burst(tx_conf, MAX_PKT_BURST, port); + + /* copy rest of the packets into the TX buffer. */ + len = num - n; + j = 0; + switch (len % FWDSTEP) { + while (j < len) { + case 0: + tx_conf->mbuf[port].m_table[j] = m[n + j]; + j++; + case 3: + tx_conf->mbuf[port].m_table[j] = m[n + j]; + j++; + case 2: + tx_conf->mbuf[port].m_table[j] = m[n + j]; + j++; + case 1: + tx_conf->mbuf[port].m_table[j] = m[n + j]; + j++; + } + } + } + + tx_conf->mbuf[port].len = len; +} + +#ifdef DO_RFC_1812_CHECKS +static inline int +is_valid_ipv4_pkt(struct ipv4_hdr *pkt, uint32_t link_len) +{ + /* From http://www.rfc-editor.org/rfc/rfc1812.txt section 5.2.2 */ + /* + * 1. The packet length reported by the Link Layer must be large + * enough to hold the minimum length legal IP datagram (20 bytes). + */ + if (link_len < sizeof(struct ipv4_hdr)) + return -1; + + /* 2. The IP checksum must be correct. */ + /* this is checked in H/W */ + + /* + * 3. The IP version number must be 4. If the version number is not 4 + * then the packet may be another version of IP, such as IPng or + * ST-II. + */ + if (((pkt->version_ihl) >> 4) != 4) + return -3; + /* + * 4. The IP header length field must be large enough to hold the + * minimum length legal IP datagram (20 bytes = 5 words). + */ + if ((pkt->version_ihl & 0xf) < 5) + return -4; + + /* + * 5. The IP total length field must be large enough to hold the IP + * datagram header, whose length is specified in the IP header length + * field. + */ + if (rte_cpu_to_be_16(pkt->total_length) < sizeof(struct ipv4_hdr)) + return -5; + + return 0; +} +#endif + +#if (APP_LOOKUP_METHOD == APP_LOOKUP_EXACT_MATCH) + +static __m128i mask0; +static __m128i mask1; +static __m128i mask2; + +static inline uint8_t +get_ipv4_dst_port(void *ipv4_hdr, uint8_t portid, + lookup_struct_t *ipv4_l3fwd_lookup_struct) +{ + int ret = 0; + union ipv4_5tuple_host key; + + ipv4_hdr = (uint8_t *)ipv4_hdr + offsetof(struct ipv4_hdr, time_to_live); + __m128i data = _mm_loadu_si128((__m128i *)(ipv4_hdr)); + /* Get 5 tuple: dst port, src port, dst IP address, src IP address and + protocol */ + key.xmm = _mm_and_si128(data, mask0); + /* Find destination port */ + ret = rte_hash_lookup(ipv4_l3fwd_lookup_struct, (const void *)&key); + return (uint8_t)((ret < 0) ? portid : ipv4_l3fwd_out_if[ret]); +} + +static inline uint8_t +get_ipv6_dst_port(void *ipv6_hdr, uint8_t portid, + lookup_struct_t *ipv6_l3fwd_lookup_struct) +{ + int ret = 0; + union ipv6_5tuple_host key; + + ipv6_hdr = (uint8_t *)ipv6_hdr + offsetof(struct ipv6_hdr, payload_len); + __m128i data0 = _mm_loadu_si128((__m128i *)(ipv6_hdr)); + __m128i data1 = _mm_loadu_si128((__m128i *)(((uint8_t *)ipv6_hdr) + + sizeof(__m128i))); + __m128i data2 = _mm_loadu_si128((__m128i *)(((uint8_t *)ipv6_hdr) + + sizeof(__m128i) + sizeof(__m128i))); + /* Get part of 5 tuple: src IP address lower 96 bits and protocol */ + key.xmm[0] = _mm_and_si128(data0, mask1); + /* Get part of 5 tuple: dst IP address lower 96 bits and src IP address + higher 32 bits */ + key.xmm[1] = data1; + /* Get part of 5 tuple: dst port and src port and dst IP address higher + 32 bits */ + key.xmm[2] = _mm_and_si128(data2, mask2); + + /* Find destination port */ + ret = rte_hash_lookup(ipv6_l3fwd_lookup_struct, (const void *)&key); + return (uint8_t)((ret < 0) ? portid : ipv6_l3fwd_out_if[ret]); +} +#endif + +#if (APP_LOOKUP_METHOD == APP_LOOKUP_LPM) + +static inline uint8_t +get_ipv4_dst_port(void *ipv4_hdr, uint8_t portid, + lookup_struct_t *ipv4_l3fwd_lookup_struct) +{ + uint8_t next_hop; + + return (uint8_t) ((rte_lpm_lookup(ipv4_l3fwd_lookup_struct, + rte_be_to_cpu_32(((struct ipv4_hdr *)ipv4_hdr)->dst_addr), + &next_hop) == 0) ? next_hop : portid); +} + +static inline uint8_t +get_ipv6_dst_port(void *ipv6_hdr, uint8_t portid, + lookup6_struct_t *ipv6_l3fwd_lookup_struct) +{ + uint8_t next_hop; + + return (uint8_t) ((rte_lpm6_lookup(ipv6_l3fwd_lookup_struct, + ((struct ipv6_hdr *)ipv6_hdr)->dst_addr, &next_hop) == 0) ? + next_hop : portid); +} +#endif + +static inline void l3fwd_simple_forward(struct rte_mbuf *m, uint8_t portid) + __attribute__((unused)); + +#if ((APP_LOOKUP_METHOD == APP_LOOKUP_EXACT_MATCH) && \ + (ENABLE_MULTI_BUFFER_OPTIMIZE == 1)) + +#define MASK_ALL_PKTS 0xf +#define EXECLUDE_1ST_PKT 0xe +#define EXECLUDE_2ND_PKT 0xd +#define EXECLUDE_3RD_PKT 0xb +#define EXECLUDE_4TH_PKT 0x7 + +static inline void +simple_ipv4_fwd_4pkts(struct rte_mbuf *m[4], uint8_t portid, + struct thread_tx_conf *qconf) +{ + struct ether_hdr *eth_hdr[4]; + struct ipv4_hdr *ipv4_hdr[4]; + uint8_t dst_port[4]; + int32_t ret[4]; + union ipv4_5tuple_host key[4]; + __m128i data[4]; + + eth_hdr[0] = rte_pktmbuf_mtod(m[0], struct ether_hdr *); + eth_hdr[1] = rte_pktmbuf_mtod(m[1], struct ether_hdr *); + eth_hdr[2] = rte_pktmbuf_mtod(m[2], struct ether_hdr *); + eth_hdr[3] = rte_pktmbuf_mtod(m[3], struct ether_hdr *); + + /* Handle IPv4 headers.*/ + ipv4_hdr[0] = rte_pktmbuf_mtod_offset(m[0], struct ipv4_hdr *, + sizeof(struct ether_hdr)); + ipv4_hdr[1] = rte_pktmbuf_mtod_offset(m[1], struct ipv4_hdr *, + sizeof(struct ether_hdr)); + ipv4_hdr[2] = rte_pktmbuf_mtod_offset(m[2], struct ipv4_hdr *, + sizeof(struct ether_hdr)); + ipv4_hdr[3] = rte_pktmbuf_mtod_offset(m[3], struct ipv4_hdr *, + sizeof(struct ether_hdr)); + +#ifdef DO_RFC_1812_CHECKS + /* Check to make sure the packet is valid (RFC1812) */ + uint8_t valid_mask = MASK_ALL_PKTS; + + if (is_valid_ipv4_pkt(ipv4_hdr[0], m[0]->pkt_len) < 0) { + rte_pktmbuf_free(m[0]); + valid_mask &= EXECLUDE_1ST_PKT; + } + if (is_valid_ipv4_pkt(ipv4_hdr[1], m[1]->pkt_len) < 0) { + rte_pktmbuf_free(m[1]); + valid_mask &= EXECLUDE_2ND_PKT; + } + if (is_valid_ipv4_pkt(ipv4_hdr[2], m[2]->pkt_len) < 0) { + rte_pktmbuf_free(m[2]); + valid_mask &= EXECLUDE_3RD_PKT; + } + if (is_valid_ipv4_pkt(ipv4_hdr[3], m[3]->pkt_len) < 0) { + rte_pktmbuf_free(m[3]); + valid_mask &= EXECLUDE_4TH_PKT; + } + if (unlikely(valid_mask != MASK_ALL_PKTS)) { + if (valid_mask == 0) + return; + + uint8_t i = 0; + + for (i = 0; i < 4; i++) + if ((0x1 << i) & valid_mask) + l3fwd_simple_forward(m[i], portid, qconf); + + return; + } +#endif /* End of #ifdef DO_RFC_1812_CHECKS */ + + data[0] = _mm_loadu_si128(rte_pktmbuf_mtod_offset(m[0], __m128i *, + sizeof(struct ether_hdr) + offsetof(struct ipv4_hdr, + time_to_live))); + data[1] = _mm_loadu_si128(rte_pktmbuf_mtod_offset(m[1], __m128i *, + sizeof(struct ether_hdr) + offsetof(struct ipv4_hdr, + time_to_live))); + data[2] = _mm_loadu_si128(rte_pktmbuf_mtod_offset(m[2], __m128i *, + sizeof(struct ether_hdr) + offsetof(struct ipv4_hdr, + time_to_live))); + data[3] = _mm_loadu_si128(rte_pktmbuf_mtod_offset(m[3], __m128i *, + sizeof(struct ether_hdr) + offsetof(struct ipv4_hdr, + time_to_live))); + + key[0].xmm = _mm_and_si128(data[0], mask0); + key[1].xmm = _mm_and_si128(data[1], mask0); + key[2].xmm = _mm_and_si128(data[2], mask0); + key[3].xmm = _mm_and_si128(data[3], mask0); + + const void *key_array[4] = {&key[0], &key[1], &key[2], &key[3]}; + + rte_hash_lookup_multi(qconf->ipv4_lookup_struct, &key_array[0], 4, ret); + dst_port[0] = (uint8_t) ((ret[0] < 0) ? portid : ipv4_l3fwd_out_if[ret[0]]); + dst_port[1] = (uint8_t) ((ret[1] < 0) ? portid : ipv4_l3fwd_out_if[ret[1]]); + dst_port[2] = (uint8_t) ((ret[2] < 0) ? portid : ipv4_l3fwd_out_if[ret[2]]); + dst_port[3] = (uint8_t) ((ret[3] < 0) ? portid : ipv4_l3fwd_out_if[ret[3]]); + + if (dst_port[0] >= RTE_MAX_ETHPORTS || + (enabled_port_mask & 1 << dst_port[0]) == 0) + dst_port[0] = portid; + if (dst_port[1] >= RTE_MAX_ETHPORTS || + (enabled_port_mask & 1 << dst_port[1]) == 0) + dst_port[1] = portid; + if (dst_port[2] >= RTE_MAX_ETHPORTS || + (enabled_port_mask & 1 << dst_port[2]) == 0) + dst_port[2] = portid; + if (dst_port[3] >= RTE_MAX_ETHPORTS || + (enabled_port_mask & 1 << dst_port[3]) == 0) + dst_port[3] = portid; + +#ifdef DO_RFC_1812_CHECKS + /* Update time to live and header checksum */ + --(ipv4_hdr[0]->time_to_live); + --(ipv4_hdr[1]->time_to_live); + --(ipv4_hdr[2]->time_to_live); + --(ipv4_hdr[3]->time_to_live); + ++(ipv4_hdr[0]->hdr_checksum); + ++(ipv4_hdr[1]->hdr_checksum); + ++(ipv4_hdr[2]->hdr_checksum); + ++(ipv4_hdr[3]->hdr_checksum); +#endif + + /* dst addr */ + *(uint64_t *)ð_hdr[0]->d_addr = dest_eth_addr[dst_port[0]]; + *(uint64_t *)ð_hdr[1]->d_addr = dest_eth_addr[dst_port[1]]; + *(uint64_t *)ð_hdr[2]->d_addr = dest_eth_addr[dst_port[2]]; + *(uint64_t *)ð_hdr[3]->d_addr = dest_eth_addr[dst_port[3]]; + + /* src addr */ + ether_addr_copy(&ports_eth_addr[dst_port[0]], ð_hdr[0]->s_addr); + ether_addr_copy(&ports_eth_addr[dst_port[1]], ð_hdr[1]->s_addr); + ether_addr_copy(&ports_eth_addr[dst_port[2]], ð_hdr[2]->s_addr); + ether_addr_copy(&ports_eth_addr[dst_port[3]], ð_hdr[3]->s_addr); + + send_single_packet(m[0], (uint8_t)dst_port[0]); + send_single_packet(m[1], (uint8_t)dst_port[1]); + send_single_packet(m[2], (uint8_t)dst_port[2]); + send_single_packet(m[3], (uint8_t)dst_port[3]); + +} + +static inline void get_ipv6_5tuple(struct rte_mbuf *m0, __m128i mask0, + __m128i mask1, union ipv6_5tuple_host *key) +{ + __m128i tmpdata0 = _mm_loadu_si128(rte_pktmbuf_mtod_offset(m0, __m128i *, + sizeof(struct ether_hdr) + offsetof(struct ipv6_hdr, payload_len))); + __m128i tmpdata1 = _mm_loadu_si128(rte_pktmbuf_mtod_offset(m0, __m128i *, + sizeof(struct ether_hdr) + offsetof(struct ipv6_hdr, payload_len) + + sizeof(__m128i))); + __m128i tmpdata2 = _mm_loadu_si128(rte_pktmbuf_mtod_offset(m0, __m128i *, + sizeof(struct ether_hdr) + offsetof(struct ipv6_hdr, payload_len) + + sizeof(__m128i) + sizeof(__m128i))); + + key->xmm[0] = _mm_and_si128(tmpdata0, mask0); + key->xmm[1] = tmpdata1; + key->xmm[2] = _mm_and_si128(tmpdata2, mask1); +} + +static inline void +simple_ipv6_fwd_4pkts(struct rte_mbuf *m[4], uint8_t portid, + struct thread_tx_conf *qconf) +{ + struct ether_hdr *eth_hdr[4]; + uint8_t dst_port[4]; + int32_t ret[4]; + union ipv6_5tuple_host key[4]; + + __attribute__((unused)) struct ipv6_hdr *ipv6_hdr[4]; + + eth_hdr[0] = rte_pktmbuf_mtod(m[0], struct ether_hdr *); + eth_hdr[1] = rte_pktmbuf_mtod(m[1], struct ether_hdr *); + eth_hdr[2] = rte_pktmbuf_mtod(m[2], struct ether_hdr *); + eth_hdr[3] = rte_pktmbuf_mtod(m[3], struct ether_hdr *); + + /* Handle IPv6 headers.*/ + ipv6_hdr[0] = rte_pktmbuf_mtod_offset(m[0], struct ipv6_hdr *, + sizeof(struct ether_hdr)); + ipv6_hdr[1] = rte_pktmbuf_mtod_offset(m[1], struct ipv6_hdr *, + sizeof(struct ether_hdr)); + ipv6_hdr[2] = rte_pktmbuf_mtod_offset(m[2], struct ipv6_hdr *, + sizeof(struct ether_hdr)); + ipv6_hdr[3] = rte_pktmbuf_mtod_offset(m[3], struct ipv6_hdr *, + sizeof(struct ether_hdr)); + + get_ipv6_5tuple(m[0], mask1, mask2, &key[0]); + get_ipv6_5tuple(m[1], mask1, mask2, &key[1]); + get_ipv6_5tuple(m[2], mask1, mask2, &key[2]); + get_ipv6_5tuple(m[3], mask1, mask2, &key[3]); + + const void *key_array[4] = {&key[0], &key[1], &key[2], &key[3]}; + + rte_hash_lookup_multi(qconf->ipv6_lookup_struct, &key_array[0], 4, ret); + dst_port[0] = (uint8_t) ((ret[0] < 0) ? portid : ipv6_l3fwd_out_if[ret[0]]); + dst_port[1] = (uint8_t) ((ret[1] < 0) ? portid : ipv6_l3fwd_out_if[ret[1]]); + dst_port[2] = (uint8_t) ((ret[2] < 0) ? portid : ipv6_l3fwd_out_if[ret[2]]); + dst_port[3] = (uint8_t) ((ret[3] < 0) ? portid : ipv6_l3fwd_out_if[ret[3]]); + + if (dst_port[0] >= RTE_MAX_ETHPORTS || + (enabled_port_mask & 1 << dst_port[0]) == 0) + dst_port[0] = portid; + if (dst_port[1] >= RTE_MAX_ETHPORTS || + (enabled_port_mask & 1 << dst_port[1]) == 0) + dst_port[1] = portid; + if (dst_port[2] >= RTE_MAX_ETHPORTS || + (enabled_port_mask & 1 << dst_port[2]) == 0) + dst_port[2] = portid; + if (dst_port[3] >= RTE_MAX_ETHPORTS || + (enabled_port_mask & 1 << dst_port[3]) == 0) + dst_port[3] = portid; + + /* dst addr */ + *(uint64_t *)ð_hdr[0]->d_addr = dest_eth_addr[dst_port[0]]; + *(uint64_t *)ð_hdr[1]->d_addr = dest_eth_addr[dst_port[1]]; + *(uint64_t *)ð_hdr[2]->d_addr = dest_eth_addr[dst_port[2]]; + *(uint64_t *)ð_hdr[3]->d_addr = dest_eth_addr[dst_port[3]]; + + /* src addr */ + ether_addr_copy(&ports_eth_addr[dst_port[0]], ð_hdr[0]->s_addr); + ether_addr_copy(&ports_eth_addr[dst_port[1]], ð_hdr[1]->s_addr); + ether_addr_copy(&ports_eth_addr[dst_port[2]], ð_hdr[2]->s_addr); + ether_addr_copy(&ports_eth_addr[dst_port[3]], ð_hdr[3]->s_addr); + + send_single_packet(m[0], (uint8_t)dst_port[0]); + send_single_packet(m[1], (uint8_t)dst_port[1]); + send_single_packet(m[2], (uint8_t)dst_port[2]); + send_single_packet(m[3], (uint8_t)dst_port[3]); + +} +#endif /* APP_LOOKUP_METHOD */ + +static inline __attribute__((always_inline)) void +l3fwd_simple_forward(struct rte_mbuf *m, uint8_t portid) +{ + struct ether_hdr *eth_hdr; + struct ipv4_hdr *ipv4_hdr; + uint8_t dst_port; + + eth_hdr = rte_pktmbuf_mtod(m, struct ether_hdr *); + + if (RTE_ETH_IS_IPV4_HDR(m->packet_type)) { + /* Handle IPv4 headers.*/ + ipv4_hdr = rte_pktmbuf_mtod_offset(m, struct ipv4_hdr *, + sizeof(struct ether_hdr)); + +#ifdef DO_RFC_1812_CHECKS + /* Check to make sure the packet is valid (RFC1812) */ + if (is_valid_ipv4_pkt(ipv4_hdr, m->pkt_len) < 0) { + rte_pktmbuf_free(m); + return; + } +#endif + + dst_port = get_ipv4_dst_port(ipv4_hdr, portid, + RTE_PER_LCORE(lcore_conf)->ipv4_lookup_struct); + if (dst_port >= RTE_MAX_ETHPORTS || + (enabled_port_mask & 1 << dst_port) == 0) + dst_port = portid; + +#ifdef DO_RFC_1812_CHECKS + /* Update time to live and header checksum */ + --(ipv4_hdr->time_to_live); + ++(ipv4_hdr->hdr_checksum); +#endif + /* dst addr */ + *(uint64_t *)ð_hdr->d_addr = dest_eth_addr[dst_port]; + + /* src addr */ + ether_addr_copy(&ports_eth_addr[dst_port], ð_hdr->s_addr); + + send_single_packet(m, dst_port); + } else if (RTE_ETH_IS_IPV6_HDR(m->packet_type)) { + /* Handle IPv6 headers.*/ + struct ipv6_hdr *ipv6_hdr; + + ipv6_hdr = rte_pktmbuf_mtod_offset(m, struct ipv6_hdr *, + sizeof(struct ether_hdr)); + + dst_port = get_ipv6_dst_port(ipv6_hdr, portid, + RTE_PER_LCORE(lcore_conf)->ipv6_lookup_struct); + + if (dst_port >= RTE_MAX_ETHPORTS || + (enabled_port_mask & 1 << dst_port) == 0) + dst_port = portid; + + /* dst addr */ + *(uint64_t *)ð_hdr->d_addr = dest_eth_addr[dst_port]; + + /* src addr */ + ether_addr_copy(&ports_eth_addr[dst_port], ð_hdr->s_addr); + + send_single_packet(m, dst_port); + } else + /* Free the mbuf that contains non-IPV4/IPV6 packet */ + rte_pktmbuf_free(m); +} + +#ifdef DO_RFC_1812_CHECKS + +#define IPV4_MIN_VER_IHL 0x45 +#define IPV4_MAX_VER_IHL 0x4f +#define IPV4_MAX_VER_IHL_DIFF (IPV4_MAX_VER_IHL - IPV4_MIN_VER_IHL) + +/* Minimum value of IPV4 total length (20B) in network byte order. */ +#define IPV4_MIN_LEN_BE (sizeof(struct ipv4_hdr) << 8) + +/* + * From http://www.rfc-editor.org/rfc/rfc1812.txt section 5.2.2: + * - The IP version number must be 4. + * - The IP header length field must be large enough to hold the + * minimum length legal IP datagram (20 bytes = 5 words). + * - The IP total length field must be large enough to hold the IP + * datagram header, whose length is specified in the IP header length + * field. + * If we encounter invalid IPV4 packet, then set destination port for it + * to BAD_PORT value. + */ +static inline __attribute__((always_inline)) void +rfc1812_process(struct ipv4_hdr *ipv4_hdr, uint16_t *dp, uint32_t ptype) +{ + uint8_t ihl; + + if (RTE_ETH_IS_IPV4_HDR(ptype)) { + ihl = ipv4_hdr->version_ihl - IPV4_MIN_VER_IHL; + + ipv4_hdr->time_to_live--; + ipv4_hdr->hdr_checksum++; + + if (ihl > IPV4_MAX_VER_IHL_DIFF || + ((uint8_t)ipv4_hdr->total_length == 0 && + ipv4_hdr->total_length < IPV4_MIN_LEN_BE)) { + dp[0] = BAD_PORT; + } + } +} + +#else +#define rfc1812_process(mb, dp) do { } while (0) +#endif /* DO_RFC_1812_CHECKS */ + + +#if ((APP_LOOKUP_METHOD == APP_LOOKUP_LPM) && \ + (ENABLE_MULTI_BUFFER_OPTIMIZE == 1)) + +static inline __attribute__((always_inline)) uint16_t +get_dst_port(struct rte_mbuf *pkt, uint32_t dst_ipv4, uint8_t portid) +{ + uint8_t next_hop; + struct ipv6_hdr *ipv6_hdr; + struct ether_hdr *eth_hdr; + + if (RTE_ETH_IS_IPV4_HDR(pkt->packet_type)) { + if (rte_lpm_lookup(RTE_PER_LCORE(lcore_conf)->ipv4_lookup_struct, + dst_ipv4, &next_hop) != 0) + next_hop = portid; + } else if (RTE_ETH_IS_IPV6_HDR(pkt->packet_type)) { + eth_hdr = rte_pktmbuf_mtod(pkt, struct ether_hdr *); + ipv6_hdr = (struct ipv6_hdr *)(eth_hdr + 1); + if (rte_lpm6_lookup(RTE_PER_LCORE(lcore_conf)->ipv6_lookup_struct, + ipv6_hdr->dst_addr, &next_hop) != 0) + next_hop = portid; + } else { + next_hop = portid; + } + + return next_hop; +} + +static inline void +process_packet(struct rte_mbuf *pkt, uint16_t *dst_port, uint8_t portid) +{ + struct ether_hdr *eth_hdr; + struct ipv4_hdr *ipv4_hdr; + uint32_t dst_ipv4; + uint16_t dp; + __m128i te, ve; + + eth_hdr = rte_pktmbuf_mtod(pkt, struct ether_hdr *); + ipv4_hdr = (struct ipv4_hdr *)(eth_hdr + 1); + + dst_ipv4 = ipv4_hdr->dst_addr; + dst_ipv4 = rte_be_to_cpu_32(dst_ipv4); + dp = get_dst_port(pkt, dst_ipv4, portid); + + te = _mm_load_si128((__m128i *)eth_hdr); + ve = val_eth[dp]; + + dst_port[0] = dp; + rfc1812_process(ipv4_hdr, dst_port, pkt->packet_type); + + te = _mm_blend_epi16(te, ve, MASK_ETH); + _mm_store_si128((__m128i *)eth_hdr, te); +} + +/* + * Read packet_type and destination IPV4 addresses from 4 mbufs. + */ +static inline void +processx4_step1(struct rte_mbuf *pkt[FWDSTEP], __m128i *dip, + uint32_t *ipv4_flag) +{ + struct ipv4_hdr *ipv4_hdr; + struct ether_hdr *eth_hdr; + uint32_t x0, x1, x2, x3; + + eth_hdr = rte_pktmbuf_mtod(pkt[0], struct ether_hdr *); + ipv4_hdr = (struct ipv4_hdr *)(eth_hdr + 1); + x0 = ipv4_hdr->dst_addr; + ipv4_flag[0] = pkt[0]->packet_type & RTE_PTYPE_L3_IPV4; + + eth_hdr = rte_pktmbuf_mtod(pkt[1], struct ether_hdr *); + ipv4_hdr = (struct ipv4_hdr *)(eth_hdr + 1); + x1 = ipv4_hdr->dst_addr; + ipv4_flag[0] &= pkt[1]->packet_type; + + eth_hdr = rte_pktmbuf_mtod(pkt[2], struct ether_hdr *); + ipv4_hdr = (struct ipv4_hdr *)(eth_hdr + 1); + x2 = ipv4_hdr->dst_addr; + ipv4_flag[0] &= pkt[2]->packet_type; + + eth_hdr = rte_pktmbuf_mtod(pkt[3], struct ether_hdr *); + ipv4_hdr = (struct ipv4_hdr *)(eth_hdr + 1); + x3 = ipv4_hdr->dst_addr; + ipv4_flag[0] &= pkt[3]->packet_type; + + dip[0] = _mm_set_epi32(x3, x2, x1, x0); +} + +/* + * Lookup into LPM for destination port. + * If lookup fails, use incoming port (portid) as destination port. + */ +static inline void +processx4_step2(__m128i dip, uint32_t ipv4_flag, uint8_t portid, + struct rte_mbuf *pkt[FWDSTEP], uint16_t dprt[FWDSTEP]) +{ + rte_xmm_t dst; + const __m128i bswap_mask = _mm_set_epi8(12, 13, 14, 15, 8, 9, 10, 11, + 4, 5, 6, 7, 0, 1, 2, 3); + + /* Byte swap 4 IPV4 addresses. */ + dip = _mm_shuffle_epi8(dip, bswap_mask); + + /* if all 4 packets are IPV4. */ + if (likely(ipv4_flag)) { + rte_lpm_lookupx4(RTE_PER_LCORE(lcore_conf)->ipv4_lookup_struct, dip, + dprt, portid); + } else { + dst.x = dip; + dprt[0] = get_dst_port(pkt[0], dst.u32[0], portid); + dprt[1] = get_dst_port(pkt[1], dst.u32[1], portid); + dprt[2] = get_dst_port(pkt[2], dst.u32[2], portid); + dprt[3] = get_dst_port(pkt[3], dst.u32[3], portid); + } +} + +/* + * Update source and destination MAC addresses in the ethernet header. + * Perform RFC1812 checks and updates for IPV4 packets. + */ +static inline void +processx4_step3(struct rte_mbuf *pkt[FWDSTEP], uint16_t dst_port[FWDSTEP]) +{ + __m128i te[FWDSTEP]; + __m128i ve[FWDSTEP]; + __m128i *p[FWDSTEP]; + + p[0] = rte_pktmbuf_mtod(pkt[0], __m128i *); + p[1] = rte_pktmbuf_mtod(pkt[1], __m128i *); + p[2] = rte_pktmbuf_mtod(pkt[2], __m128i *); + p[3] = rte_pktmbuf_mtod(pkt[3], __m128i *); + + ve[0] = val_eth[dst_port[0]]; + te[0] = _mm_load_si128(p[0]); + + ve[1] = val_eth[dst_port[1]]; + te[1] = _mm_load_si128(p[1]); + + ve[2] = val_eth[dst_port[2]]; + te[2] = _mm_load_si128(p[2]); + + ve[3] = val_eth[dst_port[3]]; + te[3] = _mm_load_si128(p[3]); + + /* Update first 12 bytes, keep rest bytes intact. */ + te[0] = _mm_blend_epi16(te[0], ve[0], MASK_ETH); + te[1] = _mm_blend_epi16(te[1], ve[1], MASK_ETH); + te[2] = _mm_blend_epi16(te[2], ve[2], MASK_ETH); + te[3] = _mm_blend_epi16(te[3], ve[3], MASK_ETH); + + _mm_store_si128(p[0], te[0]); + _mm_store_si128(p[1], te[1]); + _mm_store_si128(p[2], te[2]); + _mm_store_si128(p[3], te[3]); + + rfc1812_process((struct ipv4_hdr *)((struct ether_hdr *)p[0] + 1), + &dst_port[0], pkt[0]->packet_type); + rfc1812_process((struct ipv4_hdr *)((struct ether_hdr *)p[1] + 1), + &dst_port[1], pkt[1]->packet_type); + rfc1812_process((struct ipv4_hdr *)((struct ether_hdr *)p[2] + 1), + &dst_port[2], pkt[2]->packet_type); + rfc1812_process((struct ipv4_hdr *)((struct ether_hdr *)p[3] + 1), + &dst_port[3], pkt[3]->packet_type); +} + +/* + * We group consecutive packets with the same destionation port into one burst. + * To avoid extra latency this is done together with some other packet + * processing, but after we made a final decision about packet's destination. + * To do this we maintain: + * pnum - array of number of consecutive packets with the same dest port for + * each packet in the input burst. + * lp - pointer to the last updated element in the pnum. + * dlp - dest port value lp corresponds to. + */ + +#define GRPSZ (1 << FWDSTEP) +#define GRPMSK (GRPSZ - 1) + +#define GROUP_PORT_STEP(dlp, dcp, lp, pn, idx) do { \ + if (likely((dlp) == (dcp)[(idx)])) { \ + (lp)[0]++; \ + } else { \ + (dlp) = (dcp)[idx]; \ + (lp) = (pn) + (idx); \ + (lp)[0] = 1; \ + } \ +} while (0) + +/* + * Group consecutive packets with the same destination port in bursts of 4. + * Suppose we have array of destionation ports: + * dst_port[] = {a, b, c, d,, e, ... } + * dp1 should contain: <a, b, c, d>, dp2: <b, c, d, e>. + * We doing 4 comparisions at once and the result is 4 bit mask. + * This mask is used as an index into prebuild array of pnum values. + */ +static inline uint16_t * +port_groupx4(uint16_t pn[FWDSTEP + 1], uint16_t *lp, __m128i dp1, __m128i dp2) +{ + static const struct { + uint64_t pnum; /* prebuild 4 values for pnum[]. */ + int32_t idx; /* index for new last updated elemnet. */ + uint16_t lpv; /* add value to the last updated element. */ + } gptbl[GRPSZ] = { + { + /* 0: a != b, b != c, c != d, d != e */ + .pnum = UINT64_C(0x0001000100010001), + .idx = 4, + .lpv = 0, + }, + { + /* 1: a == b, b != c, c != d, d != e */ + .pnum = UINT64_C(0x0001000100010002), + .idx = 4, + .lpv = 1, + }, + { + /* 2: a != b, b == c, c != d, d != e */ + .pnum = UINT64_C(0x0001000100020001), + .idx = 4, + .lpv = 0, + }, + { + /* 3: a == b, b == c, c != d, d != e */ + .pnum = UINT64_C(0x0001000100020003), + .idx = 4, + .lpv = 2, + }, + { + /* 4: a != b, b != c, c == d, d != e */ + .pnum = UINT64_C(0x0001000200010001), + .idx = 4, + .lpv = 0, + }, + { + /* 5: a == b, b != c, c == d, d != e */ + .pnum = UINT64_C(0x0001000200010002), + .idx = 4, + .lpv = 1, + }, + { + /* 6: a != b, b == c, c == d, d != e */ + .pnum = UINT64_C(0x0001000200030001), + .idx = 4, + .lpv = 0, + }, + { + /* 7: a == b, b == c, c == d, d != e */ + .pnum = UINT64_C(0x0001000200030004), + .idx = 4, + .lpv = 3, + }, + { + /* 8: a != b, b != c, c != d, d == e */ + .pnum = UINT64_C(0x0002000100010001), + .idx = 3, + .lpv = 0, + }, + { + /* 9: a == b, b != c, c != d, d == e */ + .pnum = UINT64_C(0x0002000100010002), + .idx = 3, + .lpv = 1, + }, + { + /* 0xa: a != b, b == c, c != d, d == e */ + .pnum = UINT64_C(0x0002000100020001), + .idx = 3, + .lpv = 0, + }, + { + /* 0xb: a == b, b == c, c != d, d == e */ + .pnum = UINT64_C(0x0002000100020003), + .idx = 3, + .lpv = 2, + }, + { + /* 0xc: a != b, b != c, c == d, d == e */ + .pnum = UINT64_C(0x0002000300010001), + .idx = 2, + .lpv = 0, + }, + { + /* 0xd: a == b, b != c, c == d, d == e */ + .pnum = UINT64_C(0x0002000300010002), + .idx = 2, + .lpv = 1, + }, + { + /* 0xe: a != b, b == c, c == d, d == e */ + .pnum = UINT64_C(0x0002000300040001), + .idx = 1, + .lpv = 0, + }, + { + /* 0xf: a == b, b == c, c == d, d == e */ + .pnum = UINT64_C(0x0002000300040005), + .idx = 0, + .lpv = 4, + }, + }; + + union { + uint16_t u16[FWDSTEP + 1]; + uint64_t u64; + } *pnum = (void *)pn; + + int32_t v; + + dp1 = _mm_cmpeq_epi16(dp1, dp2); + dp1 = _mm_unpacklo_epi16(dp1, dp1); + v = _mm_movemask_ps((__m128)dp1); + + /* update last port counter. */ + lp[0] += gptbl[v].lpv; + + /* if dest port value has changed. */ + if (v != GRPMSK) { + lp = pnum->u16 + gptbl[v].idx; + lp[0] = 1; + pnum->u64 = gptbl[v].pnum; + } + + return lp; +} + +#endif /* APP_LOOKUP_METHOD */ + +static inline void +process_burst(struct rte_mbuf *pkts_burst[MAX_PKT_BURST], int nb_rx, + struct thread_tx_conf *tx_conf, uint8_t port_id) +{ + int j; + +#if ((APP_LOOKUP_METHOD == APP_LOOKUP_LPM) && \ + (ENABLE_MULTI_BUFFER_OPTIMIZE == 1)) + int32_t k; + uint16_t dlp; + uint16_t *lp; + uint16_t dst_port[MAX_PKT_BURST]; + __m128i dip[MAX_PKT_BURST / FWDSTEP]; + uint32_t ipv4_flag[MAX_PKT_BURST / FWDSTEP]; + uint16_t pnum[MAX_PKT_BURST + 1]; +#endif + +#if (ENABLE_MULTI_BUFFER_OPTIMIZE == 1) +#if (APP_LOOKUP_METHOD == APP_LOOKUP_EXACT_MATCH) + { + /* + * Send nb_rx - nb_rx%4 packets + * in groups of 4. + */ + int32_t n = RTE_ALIGN_FLOOR(nb_rx, 4); + + for (j = 0; j < n ; j += 4) { + uint32_t pkt_type = + pkts_burst[j]->packet_type & + pkts_burst[j+1]->packet_type & + pkts_burst[j+2]->packet_type & + pkts_burst[j+3]->packet_type; + + if (pkt_type & RTE_PTYPE_L3_IPV4) { + simple_ipv4_fwd_4pkts( + &pkts_burst[j], port_id, tx_conf); + } else if (pkt_type & + RTE_PTYPE_L3_IPV6) { + simple_ipv6_fwd_4pkts(&pkts_burst[j], + port_id, tx_conf); + } else { + l3fwd_simple_forward(pkts_burst[j], + port_id, tx_conf); + l3fwd_simple_forward(pkts_burst[j+1], + port_id, tx_conf); + l3fwd_simple_forward(pkts_burst[j+2], + port_id, tx_conf); + l3fwd_simple_forward(pkts_burst[j+3], + port_id, tx_conf); + } + } + for (; j < nb_rx ; j++) { + l3fwd_simple_forward(pkts_burst[j], + port_id, tx_conf); + } + } +#elif (APP_LOOKUP_METHOD == APP_LOOKUP_LPM) + + k = RTE_ALIGN_FLOOR(nb_rx, FWDSTEP); + for (j = 0; j != k; j += FWDSTEP) { + processx4_step1(&pkts_burst[j], + &dip[j / FWDSTEP], + &ipv4_flag[j / FWDSTEP]); + } + + k = RTE_ALIGN_FLOOR(nb_rx, FWDSTEP); + for (j = 0; j != k; j += FWDSTEP) + processx4_step2(dip[j / FWDSTEP], ipv4_flag[j / FWDSTEP], + port_id, &pkts_burst[j], &dst_port[j]); + + /* + * Finish packet processing and group consecutive + * packets with the same destination port. + */ + k = RTE_ALIGN_FLOOR(nb_rx, FWDSTEP); + if (k != 0) { + __m128i dp1, dp2; + + lp = pnum; + lp[0] = 1; + + processx4_step3(pkts_burst, dst_port); + + /* dp1: <d[0], d[1], d[2], d[3], ... > */ + dp1 = _mm_loadu_si128((__m128i *)dst_port); + + for (j = FWDSTEP; j != k; j += FWDSTEP) { + processx4_step3(&pkts_burst[j], &dst_port[j]); + + /* + * dp2: + * <d[j-3], d[j-2], d[j-1], d[j], ... > + */ + dp2 = _mm_loadu_si128((__m128i *)&dst_port[j - + FWDSTEP + 1]); + lp = port_groupx4(&pnum[j - FWDSTEP], lp, dp1, dp2); + + /* + * dp1: + * <d[j], d[j+1], d[j+2], d[j+3], ... > + */ + dp1 = _mm_srli_si128(dp2, (FWDSTEP - 1) * + sizeof(dst_port[0])); + } + + /* + * dp2: <d[j-3], d[j-2], d[j-1], d[j-1], ... > + */ + dp2 = _mm_shufflelo_epi16(dp1, 0xf9); + lp = port_groupx4(&pnum[j - FWDSTEP], lp, dp1, dp2); + + /* + * remove values added by the last repeated + * dst port. + */ + lp[0]--; + dlp = dst_port[j - 1]; + } else { + /* set dlp and lp to the never used values. */ + dlp = BAD_PORT - 1; + lp = pnum + MAX_PKT_BURST; + } + + /* Process up to last 3 packets one by one. */ + switch (nb_rx % FWDSTEP) { + case 3: + process_packet(pkts_burst[j], dst_port + j, port_id); + GROUP_PORT_STEP(dlp, dst_port, lp, pnum, j); + j++; + case 2: + process_packet(pkts_burst[j], dst_port + j, port_id); + GROUP_PORT_STEP(dlp, dst_port, lp, pnum, j); + j++; + case 1: + process_packet(pkts_burst[j], dst_port + j, port_id); + GROUP_PORT_STEP(dlp, dst_port, lp, pnum, j); + j++; + } + + /* + * Send packets out, through destination port. + * Consecutive packets with the same destination port + * are already grouped together. + * If destination port for the packet equals BAD_PORT, + * then free the packet without sending it out. + */ + for (j = 0; j < nb_rx; j += k) { + + int32_t m; + uint16_t pn; + + pn = dst_port[j]; + k = pnum[j]; + + if (likely(pn != BAD_PORT)) { + send_packetsx4(tx_conf, pn, pkts_burst + j, k); + } else { + for (m = j; m != j + k; m++) + rte_pktmbuf_free(pkts_burst[m]); + } + } + +#endif /* APP_LOOKUP_METHOD */ +#else /* ENABLE_MULTI_BUFFER_OPTIMIZE == 0 */ + + /* Prefetch first packets */ + for (j = 0; j < PREFETCH_OFFSET && j < nb_rx; j++) + rte_prefetch0(rte_pktmbuf_mtod(pkts_burst[j], void *)); + + /* Prefetch and forward already prefetched packets */ + for (j = 0; j < (nb_rx - PREFETCH_OFFSET); j++) { + rte_prefetch0(rte_pktmbuf_mtod(pkts_burst[j + + PREFETCH_OFFSET], void *)); + l3fwd_simple_forward(pkts_burst[j], port_id, tx_conf); + } + + /* Forward remaining prefetched packets */ + for (; j < nb_rx; j++) + l3fwd_simple_forward(pkts_burst[j], port_id, tx_conf); + +#endif /* ENABLE_MULTI_BUFFER_OPTIMIZE */ +} + + +/* + * lthread processing loops + */ + + +/* + * Counter of spawned rx lthreads + */ +rte_atomic16_t rx_counter; + + +/* + * Null processing lthread loop + * + * This loop is used to start empty scheduler on lcore. + */ +static void +main_lthread_null_loop(__rte_unused void *args) +{ + int lcore_id = rte_lcore_id(); + + RTE_LOG(INFO, L3FWD, "Starting scheduler on lcore %d.\n", lcore_id); + lthread_exit(NULL); +} + +/* + * Rx processing loop + */ +static void +main_lthread_rx_loop(void *args) +{ + struct rte_mbuf *pkts_burst[MAX_PKT_BURST]; + int len[RTE_MAX_LCORE] = { 0 }; + + unsigned lcore_id; + uint16_t i, nb_rx; + + uint16_t worker_id; + + uint8_t port_id, queue_id; + struct thread_rx_conf *rx_conf; + int ret; + + rx_conf = (struct thread_rx_conf *)args; + + lthread_set_data((void *)rx_conf); + + /* + * Move this lthread to lcore + */ + lthread_set_affinity(rx_conf->lcore_id); + + lcore_id = rte_lcore_id(); + + for (i = 0; i < rx_conf->n_queue; i++) { + port_id = rx_conf->queue[i].port_id; + queue_id = rx_conf->queue[i].queue_id; + RTE_LOG(INFO, L3FWD, " -- lcoreid=%u portid=%hhu rxqueueid=%hhu\n", + lcore_id, port_id, queue_id); + } + + for (i = 0; i < rx_conf->n_ring; i++) + lthread_cond_init(NULL, &rx_conf->ready[i], NULL); + + rte_atomic16_inc(&rx_counter); + + RTE_LOG(INFO, L3FWD, "Entering Rx loop on lcore %u\n", lcore_id); + + worker_id = 0; + + while (1) { + + /* + * Read packet from RX queues + */ + for (i = 0; i < rx_conf->n_queue; ++i) { + lthread_yield(); + + port_id = rx_conf->queue[i].port_id; + queue_id = rx_conf->queue[i].queue_id; + nb_rx = rte_eth_rx_burst(port_id, queue_id, pkts_burst, + MAX_PKT_BURST); + + if (nb_rx == 0) + continue; + + worker_id = (worker_id + 1) % rx_conf->n_ring; + + ret = rte_ring_sp_enqueue_burst( + rx_conf->ring[worker_id], + (void **) pkts_burst, + nb_rx); + + len[worker_id] += nb_rx; + if (len[worker_id] >= BURST_SIZE) { + lthread_cond_signal(rx_conf->ready[worker_id]); + len[worker_id] = 0; + } + + if (unlikely(ret == -ENOBUFS)) { + uint32_t k; + + for (k = 0; k < nb_rx; k++) { + struct rte_mbuf *m = pkts_burst[k]; + + rte_pktmbuf_free(m); + } + } + } + } +} + +/* + * Tx processing loop (one per ring) + * + * This lthread waits for data from producer on condition variable "ready", and + * then process incomming packets. + */ +static void +main_lthread_tx_loop(void *args) +{ + struct rte_mbuf *pkts_burst[MAX_PKT_BURST]; + uint8_t port_id; + struct thread_tx_ring_conf *tx_ring_conf; + struct thread_tx_conf *tx_conf; + uint16_t n, nb_rx; + uint16_t ring_id; + + tx_ring_conf = (struct thread_tx_ring_conf *)args; + tx_conf = tx_ring_conf->tx_conf; + ring_id = tx_ring_conf->ring_id; + struct rte_ring *ring = tx_conf->ring[ring_id]; + struct lthread_cond *ready = *tx_conf->ready[ring_id]; + + lthread_set_data((void *)tx_conf); + + nb_rx = 0; + + while (1) { + /* + * Read packet from RX rings + */ + for (n = 0; n < tx_conf->n_port_id; n++) { + + /* + * Wait for signal only if needed + */ + if (nb_rx < BURST_SIZE) + lthread_cond_wait(ready, 0); + + nb_rx = rte_ring_sc_dequeue_burst(ring, (void **)pkts_burst, + BURST_SIZE); + + if (likely(nb_rx != 0)) { + port_id = tx_conf->port_id[n]; + process_burst(pkts_burst, nb_rx, tx_conf, port_id); + } + + lthread_yield(); + } + } +} + +/* + * Main tx-lthreads spawner lthread. + * + * This lthread is used to spawn one new lthread per ring from producers. + * + */ +static void +main_lthread_tx_main_loop(void *args) +{ + struct lthread *lt[MAX_TX_THREAD]; + struct thread_tx_ring_conf tx_ring_conf[RTE_MAX_LCORE]; + + unsigned lcore_id; + uint8_t portid; + struct thread_tx_conf *tx_conf; + uint16_t i; + + tx_conf = (struct thread_tx_conf *)args; + lthread_set_data((void *)tx_conf); + + /* + * Move this lthread to the selected lcore + */ + lthread_set_affinity(tx_conf->lcore_id); + + /* + * Spawn tx readers (one per input ring) + */ + for (i = 0; i < tx_conf->n_ring; ++i) { + tx_ring_conf[i].tx_conf = tx_conf; + tx_ring_conf[i].ring_id = i; + lthread_create(<[i], tx_conf->lcore_id, main_lthread_tx_loop, + (void *)&tx_ring_conf[i]); + } + + lcore_id = rte_lcore_id(); + RTE_LOG(INFO, L3FWD, "Entering Tx main loop on lcore %u\n", lcore_id); + + while (1) { + + lthread_sleep(BURST_TX_DRAIN_NS); + + /* + * TX burst queue drain + */ + for (portid = 0; portid < RTE_MAX_ETHPORTS; portid++) { + if (tx_conf->mbuf[portid].len == 0) + continue; + send_burst(tx_conf, tx_conf->mbuf[portid].len, portid); + tx_conf->mbuf[portid].len = 0; + } + + } +} + +/* + * Start scheduler with initial lthread on lcore + * + * This lthread loop spawns all rx and tx lthreads on master lcore + */ +static void +main_lthread_master_loop(__rte_unused void *arg) { + struct lthread *lt[MAX_THREAD]; + int i; + int n_thread = 0; + + /* + * Create producers (rx threads) on default lcore + */ + for (i = 0; i < n_rx_thread; i++) { + rx_thread[i].thread_id = i; + lthread_create(<[n_thread], -1, main_lthread_rx_loop, + (void *)&rx_thread[i]); + n_thread++; + } + + /* + * Wait for all producers. Until some producers can be started on the same + * scheduler as this lthread, yielding is required to let them to run and + * prevent deadlock here. + */ + while (rte_atomic16_read(&rx_counter) < n_rx_thread) + lthread_sleep(100000); + + /* + * Create consumers (tx threads) on default lcore_id + */ + for (i = 0; i < n_tx_thread; i++) { + tx_thread[i].thread_id = i; + lthread_create(<[n_thread], -1, main_lthread_tx_main_loop, + (void *)&tx_thread[i]); + n_thread++; + } + + /* + * Wait for all threads finished + */ + for (i = 0; i < n_thread; i++) + lthread_join(lt[i], NULL); + +} + + +/* + * Rx processing loop for pthreads + */ +static void +main_pthread_rx_loop(void *args) +{ + struct rte_mbuf *pkts_burst[MAX_PKT_BURST]; + + unsigned lcore_id; + uint16_t i, nb_rx; + + uint16_t worker_id = 0; + + uint8_t portid, queueid; + struct thread_rx_conf *rx_conf; + int ret; + + rx_conf = (struct thread_rx_conf *)args; + + lcore_id = rte_lcore_id(); + + for (i = 0; i < rx_conf->n_queue; i++) { + portid = rx_conf->queue[i].port_id; + queueid = rx_conf->queue[i].queue_id; + RTE_LOG(INFO, L3FWD, " -- lcoreid=%u portid=%hhu rxqueueid=%hhu\n", + lcore_id, portid, queueid); + } + + rte_atomic16_inc(&rx_counter); + + RTE_LOG(INFO, L3FWD, "Entering Rx loop on lcore %u\n", lcore_id); + + while (1) { + + /* + * Read packet from RX queues + */ + for (i = 0; i < rx_conf->n_queue; ++i) { + + portid = rx_conf->queue[i].port_id; + queueid = rx_conf->queue[i].queue_id; + nb_rx = rte_eth_rx_burst(portid, queueid, pkts_burst, + MAX_PKT_BURST); + + if (nb_rx == 0) { + sched_yield(); + continue; + } + + worker_id = (worker_id + 1) % rx_conf->n_ring; + + ret = rte_ring_sp_enqueue_burst( + rx_conf->ring[worker_id], + (void **) pkts_burst, + nb_rx); + + if (unlikely(ret == -ENOBUFS)) { + uint32_t k; + + for (k = 0; k < nb_rx; k++) { + struct rte_mbuf *m = pkts_burst[k]; + + rte_pktmbuf_free(m); + } + } + } + } +} + +/* + * Tx processing loop for pthreads + */ +static void +main_pthread_tx_loop(void *args) +{ + struct rte_mbuf *pkts_burst[MAX_PKT_BURST]; + uint8_t port_id; + struct thread_tx_conf *tx_conf; + uint16_t nb_rx; + uint16_t ring_id; + + uint64_t prev_tsc, diff_tsc, cur_tsc; + struct rte_ring *ring; + const uint64_t drain_tsc = (rte_get_tsc_hz() + US_PER_S - 1) / + US_PER_S * BURST_TX_DRAIN_NS / 1000; + + nb_rx = 0; + prev_tsc = 0; + tx_conf = (struct thread_tx_conf *)args; + + RTE_LOG(INFO, L3FWD, "Entering Tx loop on lcore %u\n", rte_lcore_id()); + + while (1) { + + cur_tsc = rte_rdtsc(); + + /* + * TX burst queue drain + */ + diff_tsc = cur_tsc - prev_tsc; + if (unlikely(diff_tsc > drain_tsc)) { + + for (port_id = 0; port_id < RTE_MAX_ETHPORTS; port_id++) { + if (tx_conf->mbuf[port_id].len == 0) + continue; + send_burst(tx_conf, tx_conf->mbuf[port_id].len, port_id); + tx_conf->mbuf[port_id].len = 0; + } + + prev_tsc = cur_tsc; + } + + /* + * Read packet from RX rings + */ + for (ring_id = 0; ring_id < tx_conf->n_ring; ring_id++) { + + ring = tx_conf->ring[ring_id]; + + nb_rx = rte_ring_sc_dequeue_burst(ring, (void **)pkts_burst, + BURST_SIZE); + + if (nb_rx == 0) { + sched_yield(); + continue; + } + + /** For p-threads we have one port per lcore */ + port_id = tx_conf->port_id[0]; + process_burst(pkts_burst, nb_rx, tx_conf, port_id); + + } + } +} + + +/* + * Initialize and configure application + */ + +static int +check_lcore_params(void) +{ + uint8_t queue, lcore; + uint16_t i; + int socketid; + + for (i = 0; i < nb_thread_rx_params; ++i) { + queue = thread_rx_params[i].queue_id; + if (queue >= MAX_RX_QUEUE_PER_PORT) { + printf("invalid queue number: %hhu\n", queue); + return -1; + } + lcore = thread_rx_params[i].lcore_id; + if (!rte_lcore_is_enabled(lcore)) { + printf("error: lcore %hhu is not enabled in lcore mask\n", lcore); + return -1; + } + socketid = rte_lcore_to_socket_id(lcore); + if ((socketid != 0) && (numa_on == 0)) + printf("warning: lcore %hhu is on socket %d with numa off\n", + lcore, socketid); + } + return 0; +} + +static int +check_port_config(const unsigned nb_ports) +{ + unsigned portid; + uint16_t i; + + for (i = 0; i < nb_thread_rx_params; ++i) { + portid = thread_rx_params[i].port_id; + if ((enabled_port_mask & (1 << portid)) == 0) { + printf("port %u is not enabled in port mask\n", portid); + return -1; + } + if (portid >= nb_ports) { + printf("port %u is not present on the board\n", portid); + return -1; + } + } + return 0; +} + +static uint8_t +get_port_n_rx_queues(const uint8_t port) +{ + int queue = -1; + uint16_t i; + + for (i = 0; i < nb_thread_rx_params; ++i) + if (thread_rx_params[i].port_id == port && + thread_rx_params[i].queue_id > queue) + queue = thread_rx_params[i].queue_id; + + return (uint8_t)(++queue); +} + +static int +init_lcore_rx_rings(void) +{ + unsigned rx_thread_id, tx_thread_id; + + struct thread_rx_conf *rx_conf; + struct thread_tx_conf *tx_conf; + + /* Initialize the rings for the RX side */ + for (rx_thread_id = 0; rx_thread_id < n_rx_thread; rx_thread_id++) { + unsigned socket_io; + + rx_conf = &rx_thread[rx_thread_id]; + + /* use socket from initial lcore for rx_thread */ + socket_io = rte_lcore_to_socket_id(rx_conf->lcore_id); + + for (tx_thread_id = 0; tx_thread_id < n_tx_thread; tx_thread_id++) { + char name[32]; + struct rte_ring *ring = NULL; + + tx_conf = &tx_thread[tx_thread_id]; + + printf("Creating ring to connect rx-lthread %u (socket %u) with " + "tx-lthread %u ...\n", rx_thread_id, socket_io, + tx_thread_id); + + snprintf(name, sizeof(name), "app_ring_rx_s%u_rx%u_tx%u", + socket_io, rx_thread_id, tx_thread_id); + + ring = rte_ring_create(name, 1024 * 1024 * 2, socket_io, + RING_F_SP_ENQ | RING_F_SC_DEQ); + + if (ring == NULL) { + rte_panic("Cannot create ring to connect rx-lthread %u with " + "tx-thread %u\n", rx_thread_id, tx_thread_id); + } + + rx_conf->ring[rx_conf->n_ring] = ring; + + tx_conf->ring[tx_conf->n_ring] = ring; + tx_conf->ready[tx_conf->n_ring] = &rx_conf->ready[rx_conf->n_ring]; + + rx_conf->n_ring++; + tx_conf->n_ring++; + } + } + return 0; +} + +static int +init_lcore_rx_queues(void) +{ + uint16_t i, nb_rx_queue; + uint8_t lcore_id, thread_id; + struct thread_rx_conf *rx_conf; + + n_rx_thread = 0; + for (i = 0; i < nb_thread_rx_params; ++i) { + lcore_id = thread_rx_params[i].lcore_id; + thread_id = thread_rx_params[i].thread_id; + + rx_conf = &rx_thread[thread_id]; + rx_conf->lcore_id = lcore_id; + + nb_rx_queue = rx_conf->n_queue; + if (nb_rx_queue >= MAX_RX_QUEUE_PER_THREAD) { + printf("error: too many queues (%u) for lcore: %u\n", + (unsigned)nb_rx_queue + 1, (unsigned)lcore_id); + return -1; + } + rx_conf->queue[nb_rx_queue].port_id = thread_rx_params[i].port_id; + rx_conf->queue[nb_rx_queue].queue_id = thread_rx_params[i].queue_id; + + rx_conf->n_queue++; + if (thread_rx_params[i].thread_id >= n_rx_thread) + n_rx_thread = thread_rx_params[i].thread_id + 1; + } + return 0; +} + +static int +init_lcore_tx_queues(void) +{ + uint16_t i; + uint8_t tid; + unsigned cid; + struct thread_tx_conf *tx_conf; + + n_tx_thread = 0; + for (i = 0; i < nb_thread_tx_params; ++i) { + cid = thread_tx_params[i].lcore_id; + tid = thread_tx_params[i].thread_id; + + tx_conf = &tx_thread[tid]; + tx_conf->lcore_id = cid; + tx_conf->port_id[tx_conf->n_port_id++] = thread_tx_params[i].port_id; + + if (thread_tx_params[i].thread_id >= n_tx_thread) + n_tx_thread = thread_tx_params[i].thread_id + 1; + + } + return 0; +} + +/* display usage */ +static void +print_usage(const char *prgname) +{ + printf("%s [EAL options] -- -p PORTMASK -P" + " [--rx (port,queue,lcore,thread)[,(port,queue,lcore,thread]]" + " [--tx (port,lcore,thread)[,(port,lcore,thread]]" + " [--enable-jumbo [--max-pkt-len PKTLEN]]\n" + " -p PORTMASK: hexadecimal bitmask of ports to configure\n" + " -P : enable promiscuous mode\n" + " --rx (port,queue,lcore,thread): rx queues configuration\n" + " --tx (port,lcore,thread): tx configuration\n" + " --eth-dest=X,MM:MM:MM:MM:MM:MM: optional, ethernet destination for port X\n" + " --no-numa: optional, disable numa awareness\n" + " --ipv6: optional, specify it if running ipv6 packets\n" + " --enable-jumbo: enable jumbo frame" + " which max packet len is PKTLEN in decimal (64-9600)\n" + " --hash-entry-num: specify the hash entry number in hexadecimal to be setup\n" + " --no-lthreads: optional, disable lthread model\n", + prgname); +} + +static int +parse_max_pkt_len(const char *pktlen) +{ + char *end = NULL; + unsigned long len; + + /* parse decimal string */ + len = strtoul(pktlen, &end, 10); + if ((pktlen[0] == '\0') || (end == NULL) || (*end != '\0')) + return -1; + + if (len == 0) + return -1; + + return len; +} + +static int +parse_portmask(const char *portmask) +{ + char *end = NULL; + unsigned long pm; + + /* parse hexadecimal string */ + pm = strtoul(portmask, &end, 16); + if ((portmask[0] == '\0') || (end == NULL) || (*end != '\0')) + return -1; + + if (pm == 0) + return -1; + + return pm; +} + +#if (APP_LOOKUP_METHOD == APP_LOOKUP_EXACT_MATCH) +static int +parse_hash_entry_number(const char *hash_entry_num) +{ + char *end = NULL; + unsigned long hash_en; + + /* parse hexadecimal string */ + hash_en = strtoul(hash_entry_num, &end, 16); + if ((hash_entry_num[0] == '\0') || (end == NULL) || (*end != '\0')) + return -1; + + if (hash_en == 0) + return -1; + + return hash_en; +} +#endif + +static int +parse_rx_config(const char *q_arg) +{ + char s[256]; + const char *p, *p0 = q_arg; + char *end; + enum fieldnames { + FLD_PORT = 0, + FLD_QUEUE, + FLD_LCORE, + FLD_THREAD, + _NUM_FLD + }; + unsigned long int_fld[_NUM_FLD]; + char *str_fld[_NUM_FLD]; + int i; + unsigned size; + + nb_thread_rx_params = 0; + + while ((p = strchr(p0, '(')) != NULL) { + ++p; + p0 = strchr(p, ')'); + if (p0 == NULL) + return -1; + + size = p0 - p; + if (size >= sizeof(s)) + return -1; + + snprintf(s, sizeof(s), "%.*s", size, p); + if (rte_strsplit(s, sizeof(s), str_fld, _NUM_FLD, ',') != _NUM_FLD) + return -1; + for (i = 0; i < _NUM_FLD; i++) { + errno = 0; + int_fld[i] = strtoul(str_fld[i], &end, 0); + if (errno != 0 || end == str_fld[i] || int_fld[i] > 255) + return -1; + } + if (nb_thread_rx_params >= MAX_THREAD_PARAMS) { + printf("exceeded max number of lcore params: %hu\n", + nb_thread_rx_params); + return -1; + } + thread_rx_params_array[nb_thread_rx_params].port_id = + (uint8_t)int_fld[FLD_PORT]; + thread_rx_params_array[nb_thread_rx_params].queue_id = + (uint8_t)int_fld[FLD_QUEUE]; + thread_rx_params_array[nb_thread_rx_params].lcore_id = + (uint8_t)int_fld[FLD_LCORE]; + thread_rx_params_array[nb_thread_rx_params].thread_id = + (uint8_t)int_fld[FLD_THREAD]; + + ++nb_thread_rx_params; + } + thread_rx_params = thread_rx_params_array; + return 0; +} + +static int +parse_tx_config(const char *q_arg) +{ + char s[256]; + const char *p, *p0 = q_arg; + char *end; + enum fieldnames { + FLD_PORT = 0, + FLD_LCORE, + FLD_THREAD, + _NUM_FLD + }; + unsigned long int_fld[_NUM_FLD]; + char *str_fld[_NUM_FLD]; + int i; + unsigned size; + + nb_thread_tx_params = 0; + + while ((p = strchr(p0, '(')) != NULL) { + ++p; + p0 = strchr(p, ')'); + if (p0 == NULL) + return -1; + + size = p0 - p; + if (size >= sizeof(s)) + return -1; + + snprintf(s, sizeof(s), "%.*s", size, p); + if (rte_strsplit(s, sizeof(s), str_fld, _NUM_FLD, ',') != _NUM_FLD) + return -1; + for (i = 0; i < _NUM_FLD; i++) { + errno = 0; + int_fld[i] = strtoul(str_fld[i], &end, 0); + if (errno != 0 || end == str_fld[i] || int_fld[i] > 255) + return -1; + } + if (nb_thread_rx_params >= MAX_THREAD_PARAMS) { + printf("exceeded max number of lcore params: %hu\n", + nb_thread_rx_params); + return -1; + } + thread_tx_params_array[nb_thread_tx_params].port_id = + (uint8_t)int_fld[FLD_PORT]; + thread_tx_params_array[nb_thread_tx_params].lcore_id = + (uint8_t)int_fld[FLD_LCORE]; + thread_tx_params_array[nb_thread_tx_params].thread_id = + (uint8_t)int_fld[FLD_THREAD]; + + ++nb_thread_tx_params; + } + thread_tx_params = thread_tx_params_array; + return 0; +} + +static void +parse_eth_dest(const char *optarg) +{ + uint8_t portid; + char *port_end; + uint8_t c, *dest, peer_addr[6]; + + errno = 0; + portid = strtoul(optarg, &port_end, 10); + if (errno != 0 || port_end == optarg || *port_end++ != ',') + rte_exit(EXIT_FAILURE, + "Invalid eth-dest: %s", optarg); + if (portid >= RTE_MAX_ETHPORTS) + rte_exit(EXIT_FAILURE, + "eth-dest: port %d >= RTE_MAX_ETHPORTS(%d)\n", + portid, RTE_MAX_ETHPORTS); + + if (cmdline_parse_etheraddr(NULL, port_end, + &peer_addr, sizeof(peer_addr)) < 0) + rte_exit(EXIT_FAILURE, + "Invalid ethernet address: %s\n", + port_end); + dest = (uint8_t *)&dest_eth_addr[portid]; + for (c = 0; c < 6; c++) + dest[c] = peer_addr[c]; + *(uint64_t *)(val_eth + portid) = dest_eth_addr[portid]; +} + +#define CMD_LINE_OPT_RX_CONFIG "rx" +#define CMD_LINE_OPT_TX_CONFIG "tx" +#define CMD_LINE_OPT_ETH_DEST "eth-dest" +#define CMD_LINE_OPT_NO_NUMA "no-numa" +#define CMD_LINE_OPT_IPV6 "ipv6" +#define CMD_LINE_OPT_ENABLE_JUMBO "enable-jumbo" +#define CMD_LINE_OPT_HASH_ENTRY_NUM "hash-entry-num" +#define CMD_LINE_OPT_NO_LTHREADS "no-lthreads" + +/* Parse the argument given in the command line of the application */ +static int +parse_args(int argc, char **argv) +{ + int opt, ret; + char **argvopt; + int option_index; + char *prgname = argv[0]; + static struct option lgopts[] = { + {CMD_LINE_OPT_RX_CONFIG, 1, 0, 0}, + {CMD_LINE_OPT_TX_CONFIG, 1, 0, 0}, + {CMD_LINE_OPT_ETH_DEST, 1, 0, 0}, + {CMD_LINE_OPT_NO_NUMA, 0, 0, 0}, + {CMD_LINE_OPT_IPV6, 0, 0, 0}, + {CMD_LINE_OPT_ENABLE_JUMBO, 0, 0, 0}, + {CMD_LINE_OPT_HASH_ENTRY_NUM, 1, 0, 0}, + {CMD_LINE_OPT_NO_LTHREADS, 0, 0, 0}, + {NULL, 0, 0, 0} + }; + + argvopt = argv; + + while ((opt = getopt_long(argc, argvopt, "p:P", + lgopts, &option_index)) != EOF) { + + switch (opt) { + /* portmask */ + case 'p': + enabled_port_mask = parse_portmask(optarg); + if (enabled_port_mask == 0) { + printf("invalid portmask\n"); + print_usage(prgname); + return -1; + } + break; + case 'P': + printf("Promiscuous mode selected\n"); + promiscuous_on = 1; + break; + + /* long options */ + case 0: + if (!strncmp(lgopts[option_index].name, CMD_LINE_OPT_RX_CONFIG, + sizeof(CMD_LINE_OPT_RX_CONFIG))) { + ret = parse_rx_config(optarg); + if (ret) { + printf("invalid rx config\n"); + print_usage(prgname); + return -1; + } + } + + if (!strncmp(lgopts[option_index].name, CMD_LINE_OPT_TX_CONFIG, + sizeof(CMD_LINE_OPT_TX_CONFIG))) { + ret = parse_tx_config(optarg); + if (ret) { + printf("invalid tx config\n"); + print_usage(prgname); + return -1; + } + } + + if (!strncmp(lgopts[option_index].name, CMD_LINE_OPT_ETH_DEST, + sizeof(CMD_LINE_OPT_ETH_DEST))) + parse_eth_dest(optarg); + + if (!strncmp(lgopts[option_index].name, CMD_LINE_OPT_NO_NUMA, + sizeof(CMD_LINE_OPT_NO_NUMA))) { + printf("numa is disabled\n"); + numa_on = 0; + } + +#if (APP_LOOKUP_METHOD == APP_LOOKUP_EXACT_MATCH) + if (!strncmp(lgopts[option_index].name, CMD_LINE_OPT_IPV6, + sizeof(CMD_LINE_OPT_IPV6))) { + printf("ipv6 is specified\n"); + ipv6 = 1; + } +#endif + + if (!strncmp(lgopts[option_index].name, CMD_LINE_OPT_NO_LTHREADS, + sizeof(CMD_LINE_OPT_NO_LTHREADS))) { + printf("l-threads model is disabled\n"); + lthreads_on = 0; + } + + + if (!strncmp(lgopts[option_index].name, CMD_LINE_OPT_ENABLE_JUMBO, + sizeof(CMD_LINE_OPT_ENABLE_JUMBO))) { + struct option lenopts = {"max-pkt-len", required_argument, 0, + 0}; + + printf("jumbo frame is enabled - disabling simple TX path\n"); + port_conf.rxmode.jumbo_frame = 1; + + /* if no max-pkt-len set, use the default value ETHER_MAX_LEN */ + if (0 == getopt_long(argc, argvopt, "", &lenopts, + &option_index)) { + ret = parse_max_pkt_len(optarg); + if ((ret < 64) || (ret > MAX_JUMBO_PKT_LEN)) { + printf("invalid packet length\n"); + print_usage(prgname); + return -1; + } + port_conf.rxmode.max_rx_pkt_len = ret; + } + printf("set jumbo frame max packet length to %u\n", + (unsigned int)port_conf.rxmode.max_rx_pkt_len); + } +#if (APP_LOOKUP_METHOD == APP_LOOKUP_EXACT_MATCH) + if (!strncmp(lgopts[option_index].name, CMD_LINE_OPT_HASH_ENTRY_NUM, + sizeof(CMD_LINE_OPT_HASH_ENTRY_NUM))) { + ret = parse_hash_entry_number(optarg); + if ((ret > 0) && (ret <= L3FWD_HASH_ENTRIES)) { + hash_entry_number = ret; + } else { + printf("invalid hash entry number\n"); + print_usage(prgname); + return -1; + } + } +#endif + break; + + default: + print_usage(prgname); + return -1; + } + } + + if (optind >= 0) + argv[optind-1] = prgname; + + ret = optind-1; + optind = 0; /* reset getopt lib */ + return ret; +} + +static void +print_ethaddr(const char *name, const struct ether_addr *eth_addr) +{ + char buf[ETHER_ADDR_FMT_SIZE]; + + ether_format_addr(buf, ETHER_ADDR_FMT_SIZE, eth_addr); + printf("%s%s", name, buf); +} + +#if (APP_LOOKUP_METHOD == APP_LOOKUP_EXACT_MATCH) + +static void convert_ipv4_5tuple(struct ipv4_5tuple *key1, + union ipv4_5tuple_host *key2) +{ + key2->ip_dst = rte_cpu_to_be_32(key1->ip_dst); + key2->ip_src = rte_cpu_to_be_32(key1->ip_src); + key2->port_dst = rte_cpu_to_be_16(key1->port_dst); + key2->port_src = rte_cpu_to_be_16(key1->port_src); + key2->proto = key1->proto; + key2->pad0 = 0; + key2->pad1 = 0; +} + +static void convert_ipv6_5tuple(struct ipv6_5tuple *key1, + union ipv6_5tuple_host *key2) +{ + uint32_t i; + + for (i = 0; i < 16; i++) { + key2->ip_dst[i] = key1->ip_dst[i]; + key2->ip_src[i] = key1->ip_src[i]; + } + key2->port_dst = rte_cpu_to_be_16(key1->port_dst); + key2->port_src = rte_cpu_to_be_16(key1->port_src); + key2->proto = key1->proto; + key2->pad0 = 0; + key2->pad1 = 0; + key2->reserve = 0; +} + +#define BYTE_VALUE_MAX 256 +#define ALL_32_BITS 0xffffffff +#define BIT_8_TO_15 0x0000ff00 + +static inline void +populate_ipv4_few_flow_into_table(const struct rte_hash *h) +{ + uint32_t i; + int32_t ret; + uint32_t array_len = RTE_DIM(ipv4_l3fwd_route_array); + + mask0 = _mm_set_epi32(ALL_32_BITS, ALL_32_BITS, ALL_32_BITS, BIT_8_TO_15); + for (i = 0; i < array_len; i++) { + struct ipv4_l3fwd_route entry; + union ipv4_5tuple_host newkey; + + entry = ipv4_l3fwd_route_array[i]; + convert_ipv4_5tuple(&entry.key, &newkey); + ret = rte_hash_add_key(h, (void *) &newkey); + if (ret < 0) { + rte_exit(EXIT_FAILURE, "Unable to add entry %" PRIu32 + " to the l3fwd hash.\n", i); + } + ipv4_l3fwd_out_if[ret] = entry.if_out; + } + printf("Hash: Adding 0x%" PRIx32 " keys\n", array_len); +} + +#define BIT_16_TO_23 0x00ff0000 +static inline void +populate_ipv6_few_flow_into_table(const struct rte_hash *h) +{ + uint32_t i; + int32_t ret; + uint32_t array_len = RTE_DIM(ipv6_l3fwd_route_array); + + mask1 = _mm_set_epi32(ALL_32_BITS, ALL_32_BITS, ALL_32_BITS, BIT_16_TO_23); + mask2 = _mm_set_epi32(0, 0, ALL_32_BITS, ALL_32_BITS); + + for (i = 0; i < array_len; i++) { + struct ipv6_l3fwd_route entry; + union ipv6_5tuple_host newkey; + + entry = ipv6_l3fwd_route_array[i]; + convert_ipv6_5tuple(&entry.key, &newkey); + + ret = rte_hash_add_key(h, (void *)&newkey); + if (ret < 0) + rte_exit(EXIT_FAILURE, "Unable to add entry %" PRIu32 + " to the l3fwd hash.\n", i); + + ipv6_l3fwd_out_if[ret] = entry.if_out; + } + printf("Hash: Adding 0x%" PRIx32 "keys\n", array_len); +} + +#define NUMBER_PORT_USED 4 +static inline void +populate_ipv4_many_flow_into_table(const struct rte_hash *h, + unsigned int nr_flow) +{ + unsigned i; + + mask0 = _mm_set_epi32(ALL_32_BITS, ALL_32_BITS, ALL_32_BITS, BIT_8_TO_15); + for (i = 0; i < nr_flow; i++) { + struct ipv4_l3fwd_route entry; + union ipv4_5tuple_host newkey; + uint8_t a = (uint8_t) ((i / NUMBER_PORT_USED) % BYTE_VALUE_MAX); + uint8_t b = (uint8_t) (((i / NUMBER_PORT_USED) / BYTE_VALUE_MAX) % + BYTE_VALUE_MAX); + uint8_t c = (uint8_t) ((i / NUMBER_PORT_USED) / (BYTE_VALUE_MAX * + BYTE_VALUE_MAX)); + + /* Create the ipv4 exact match flow */ + memset(&entry, 0, sizeof(entry)); + switch (i & (NUMBER_PORT_USED - 1)) { + case 0: + entry = ipv4_l3fwd_route_array[0]; + entry.key.ip_dst = IPv4(101, c, b, a); + break; + case 1: + entry = ipv4_l3fwd_route_array[1]; + entry.key.ip_dst = IPv4(201, c, b, a); + break; + case 2: + entry = ipv4_l3fwd_route_array[2]; + entry.key.ip_dst = IPv4(111, c, b, a); + break; + case 3: + entry = ipv4_l3fwd_route_array[3]; + entry.key.ip_dst = IPv4(211, c, b, a); + break; + }; + convert_ipv4_5tuple(&entry.key, &newkey); + + int32_t ret = rte_hash_add_key(h, (void *) &newkey); + + if (ret < 0) + rte_exit(EXIT_FAILURE, "Unable to add entry %u\n", i); + + ipv4_l3fwd_out_if[ret] = (uint8_t) entry.if_out; + + } + printf("Hash: Adding 0x%x keys\n", nr_flow); +} + +static inline void +populate_ipv6_many_flow_into_table(const struct rte_hash *h, + unsigned int nr_flow) +{ + unsigned i; + + mask1 = _mm_set_epi32(ALL_32_BITS, ALL_32_BITS, ALL_32_BITS, BIT_16_TO_23); + mask2 = _mm_set_epi32(0, 0, ALL_32_BITS, ALL_32_BITS); + + for (i = 0; i < nr_flow; i++) { + struct ipv6_l3fwd_route entry; + union ipv6_5tuple_host newkey; + uint8_t a = (uint8_t) ((i / NUMBER_PORT_USED) % BYTE_VALUE_MAX); + uint8_t b = (uint8_t) (((i / NUMBER_PORT_USED) / BYTE_VALUE_MAX) % + BYTE_VALUE_MAX); + uint8_t c = (uint8_t) ((i / NUMBER_PORT_USED) / (BYTE_VALUE_MAX * + BYTE_VALUE_MAX)); + + /* Create the ipv6 exact match flow */ + memset(&entry, 0, sizeof(entry)); + switch (i & (NUMBER_PORT_USED - 1)) { + case 0: + entry = ipv6_l3fwd_route_array[0]; + break; + case 1: + entry = ipv6_l3fwd_route_array[1]; + break; + case 2: + entry = ipv6_l3fwd_route_array[2]; + break; + case 3: + entry = ipv6_l3fwd_route_array[3]; + break; + }; + entry.key.ip_dst[13] = c; + entry.key.ip_dst[14] = b; + entry.key.ip_dst[15] = a; + convert_ipv6_5tuple(&entry.key, &newkey); + + int32_t ret = rte_hash_add_key(h, (void *) &newkey); + + if (ret < 0) + rte_exit(EXIT_FAILURE, "Unable to add entry %u\n", i); + + ipv6_l3fwd_out_if[ret] = (uint8_t) entry.if_out; + + } + printf("Hash: Adding 0x%x keys\n", nr_flow); +} + +static void +setup_hash(int socketid) +{ + struct rte_hash_parameters ipv4_l3fwd_hash_params = { + .name = NULL, + .entries = L3FWD_HASH_ENTRIES, + .key_len = sizeof(union ipv4_5tuple_host), + .hash_func = ipv4_hash_crc, + .hash_func_init_val = 0, + }; + + struct rte_hash_parameters ipv6_l3fwd_hash_params = { + .name = NULL, + .entries = L3FWD_HASH_ENTRIES, + .key_len = sizeof(union ipv6_5tuple_host), + .hash_func = ipv6_hash_crc, + .hash_func_init_val = 0, + }; + + char s[64]; + + /* create ipv4 hash */ + snprintf(s, sizeof(s), "ipv4_l3fwd_hash_%d", socketid); + ipv4_l3fwd_hash_params.name = s; + ipv4_l3fwd_hash_params.socket_id = socketid; + ipv4_l3fwd_lookup_struct[socketid] = + rte_hash_create(&ipv4_l3fwd_hash_params); + if (ipv4_l3fwd_lookup_struct[socketid] == NULL) + rte_exit(EXIT_FAILURE, "Unable to create the l3fwd hash on socket %d\n", + socketid); + + /* create ipv6 hash */ + snprintf(s, sizeof(s), "ipv6_l3fwd_hash_%d", socketid); + ipv6_l3fwd_hash_params.name = s; + ipv6_l3fwd_hash_params.socket_id = socketid; + ipv6_l3fwd_lookup_struct[socketid] = + rte_hash_create(&ipv6_l3fwd_hash_params); + if (ipv6_l3fwd_lookup_struct[socketid] == NULL) + rte_exit(EXIT_FAILURE, "Unable to create the l3fwd hash on socket %d\n", + socketid); + + if (hash_entry_number != HASH_ENTRY_NUMBER_DEFAULT) { + /* For testing hash matching with a large number of flows we + * generate millions of IP 5-tuples with an incremented dst + * address to initialize the hash table. */ + if (ipv6 == 0) { + /* populate the ipv4 hash */ + populate_ipv4_many_flow_into_table( + ipv4_l3fwd_lookup_struct[socketid], hash_entry_number); + } else { + /* populate the ipv6 hash */ + populate_ipv6_many_flow_into_table( + ipv6_l3fwd_lookup_struct[socketid], hash_entry_number); + } + } else { + /* Use data in ipv4/ipv6 l3fwd lookup table directly to initialize + the hash table */ + if (ipv6 == 0) { + /* populate the ipv4 hash */ + populate_ipv4_few_flow_into_table( + ipv4_l3fwd_lookup_struct[socketid]); + } else { + /* populate the ipv6 hash */ + populate_ipv6_few_flow_into_table( + ipv6_l3fwd_lookup_struct[socketid]); + } + } +} +#endif + +#if (APP_LOOKUP_METHOD == APP_LOOKUP_LPM) +static void +setup_lpm(int socketid) +{ + struct rte_lpm6_config config; + unsigned i; + int ret; + char s[64]; + + /* create the LPM table */ + snprintf(s, sizeof(s), "IPV4_L3FWD_LPM_%d", socketid); + ipv4_l3fwd_lookup_struct[socketid] = rte_lpm_create(s, socketid, + IPV4_L3FWD_LPM_MAX_RULES, 0); + if (ipv4_l3fwd_lookup_struct[socketid] == NULL) + rte_exit(EXIT_FAILURE, "Unable to create the l3fwd LPM table " + "on socket %d\n", socketid); + + /* populate the LPM table */ + for (i = 0; i < IPV4_L3FWD_NUM_ROUTES; i++) { + + /* skip unused ports */ + if ((1 << ipv4_l3fwd_route_array[i].if_out & + enabled_port_mask) == 0) + continue; + + ret = rte_lpm_add(ipv4_l3fwd_lookup_struct[socketid], + ipv4_l3fwd_route_array[i].ip, + ipv4_l3fwd_route_array[i].depth, + ipv4_l3fwd_route_array[i].if_out); + + if (ret < 0) { + rte_exit(EXIT_FAILURE, "Unable to add entry %u to the " + "l3fwd LPM table on socket %d\n", + i, socketid); + } + + printf("LPM: Adding route 0x%08x / %d (%d)\n", + (unsigned)ipv4_l3fwd_route_array[i].ip, + ipv4_l3fwd_route_array[i].depth, + ipv4_l3fwd_route_array[i].if_out); + } + + /* create the LPM6 table */ + snprintf(s, sizeof(s), "IPV6_L3FWD_LPM_%d", socketid); + + config.max_rules = IPV6_L3FWD_LPM_MAX_RULES; + config.number_tbl8s = IPV6_L3FWD_LPM_NUMBER_TBL8S; + config.flags = 0; + ipv6_l3fwd_lookup_struct[socketid] = rte_lpm6_create(s, socketid, + &config); + if (ipv6_l3fwd_lookup_struct[socketid] == NULL) + rte_exit(EXIT_FAILURE, "Unable to create the l3fwd LPM table" + " on socket %d\n", socketid); + + /* populate the LPM table */ + for (i = 0; i < IPV6_L3FWD_NUM_ROUTES; i++) { + + /* skip unused ports */ + if ((1 << ipv6_l3fwd_route_array[i].if_out & + enabled_port_mask) == 0) + continue; + + ret = rte_lpm6_add(ipv6_l3fwd_lookup_struct[socketid], + ipv6_l3fwd_route_array[i].ip, + ipv6_l3fwd_route_array[i].depth, + ipv6_l3fwd_route_array[i].if_out); + + if (ret < 0) { + rte_exit(EXIT_FAILURE, "Unable to add entry %u to the " + "l3fwd LPM table on socket %d\n", + i, socketid); + } + + printf("LPM: Adding route %s / %d (%d)\n", + "IPV6", + ipv6_l3fwd_route_array[i].depth, + ipv6_l3fwd_route_array[i].if_out); + } +} +#endif + +static int +init_mem(unsigned nb_mbuf) +{ + struct lcore_conf *qconf; + int socketid; + unsigned 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; + + if (numa_on) + socketid = rte_lcore_to_socket_id(lcore_id); + else + socketid = 0; + + if (socketid >= NB_SOCKETS) { + rte_exit(EXIT_FAILURE, "Socket %d of lcore %u is out of range %d\n", + socketid, lcore_id, NB_SOCKETS); + } + if (pktmbuf_pool[socketid] == NULL) { + snprintf(s, sizeof(s), "mbuf_pool_%d", socketid); + pktmbuf_pool[socketid] = + rte_pktmbuf_pool_create(s, nb_mbuf, + MEMPOOL_CACHE_SIZE, 0, + RTE_MBUF_DEFAULT_BUF_SIZE, socketid); + if (pktmbuf_pool[socketid] == NULL) + rte_exit(EXIT_FAILURE, + "Cannot init mbuf pool on socket %d\n", socketid); + else + printf("Allocated mbuf pool on socket %d\n", socketid); + +#if (APP_LOOKUP_METHOD == APP_LOOKUP_LPM) + setup_lpm(socketid); +#else + setup_hash(socketid); +#endif + } + qconf = &lcore_conf[lcore_id]; + qconf->ipv4_lookup_struct = + ipv4_l3fwd_lookup_struct[socketid]; + qconf->ipv6_lookup_struct = + ipv6_l3fwd_lookup_struct[socketid]; + } + return 0; +} + +/* Check the link status of all ports in up to 9s, and print them finally */ +static void +check_all_ports_link_status(uint8_t port_num, uint32_t port_mask) +{ +#define CHECK_INTERVAL 100 /* 100ms */ +#define MAX_CHECK_TIME 90 /* 9s (90 * 100ms) in total */ + uint8_t portid, count, all_ports_up, print_flag = 0; + struct rte_eth_link link; + + printf("\nChecking link status"); + fflush(stdout); + for (count = 0; count <= MAX_CHECK_TIME; count++) { + all_ports_up = 1; + for (portid = 0; portid < port_num; portid++) { + if ((port_mask & (1 << portid)) == 0) + continue; + memset(&link, 0, sizeof(link)); + rte_eth_link_get_nowait(portid, &link); + /* print link status if flag set */ + if (print_flag == 1) { + if (link.link_status) + printf("Port %d Link Up - speed %u " + "Mbps - %s\n", (uint8_t)portid, + (unsigned)link.link_speed, + (link.link_duplex == ETH_LINK_FULL_DUPLEX) ? + ("full-duplex") : ("half-duplex\n")); + else + printf("Port %d Link Down\n", + (uint8_t)portid); + continue; + } + /* clear all_ports_up flag if any link down */ + if (link.link_status == 0) { + all_ports_up = 0; + break; + } + } + /* after finally printing all link status, get out */ + if (print_flag == 1) + break; + + if (all_ports_up == 0) { + printf("."); + fflush(stdout); + rte_delay_ms(CHECK_INTERVAL); + } + + /* set the print_flag if all ports up or timeout */ + if (all_ports_up == 1 || count == (MAX_CHECK_TIME - 1)) { + print_flag = 1; + printf("done\n"); + } + } +} + +/* + * Start scheduler on lcore. + */ +static int +lthread_slave_spawner(__rte_unused void *arg) { + struct lthread *lt; + int lcore_id = rte_lcore_id(); + + RTE_PER_LCORE(lcore_conf) = &lcore_conf[lcore_id]; + lthread_create(<, -1, main_lthread_null_loop, NULL); + lthread_run(); + return 0; +} + +/* + * Start master scheduler with initial lthread spawning rx and tx lthreads + * (main_lthread_master). + */ +static int +lthread_master_spawner(__rte_unused void *arg) { + struct lthread *lt; + int lcore_id = rte_lcore_id(); + + RTE_PER_LCORE(lcore_conf) = &lcore_conf[lcore_id]; + lthread_create(<, -1, main_lthread_master_loop, NULL); + lthread_run(); + return 0; +} + +/* + * Start rx/tx threads on lcore. + */ +static int +pthread_spawner(__rte_unused void *arg) { + int lcore_id = rte_lcore_id(); + int i; + + RTE_PER_LCORE(lcore_conf) = &lcore_conf[lcore_id]; + + for (i = 0; i < n_rx_thread; i++) + if (rx_thread[i].lcore_id == lcore_id) { + RTE_PER_LCORE(lcore_conf)->data = (void *)&rx_thread[i]; + main_pthread_rx_loop((void *)&rx_thread[i]); + return 0; + } + + for (i = 0; i < n_tx_thread; i++) + if (tx_thread[i].lcore_id == lcore_id) { + RTE_PER_LCORE(lcore_conf)->data = (void *)&tx_thread[i]; + main_pthread_tx_loop((void *)&tx_thread[i]); + return 0; + } + + return 0; +} + +/* + * Dump current configuration + */ +static void +print_config(void) +{ + unsigned rx, tx; + unsigned i; + struct thread_rx_conf *rx_conf; + struct thread_tx_conf *tx_conf; + + printf("\n\n--- Configuration ---\n\n"); + + printf("rx lthreads: %d\n", n_rx_thread); + for (rx = 0; rx < n_rx_thread; rx++) { + rx_conf = &rx_thread[rx]; + printf(" %2d. lcore=%d, %d queue(s):\n", rx, rx_conf->lcore_id, + rx_conf->n_queue); + for (i = 0; i < rx_conf->n_queue; i++) + printf(" %2d. Rx-port %d, queue %d\n", i, + rx_conf->queue[i].port_id, rx_conf->queue[i].queue_id); + } + + printf("tx lthreads: %d\n", n_tx_thread); + for (tx = 0; tx < n_tx_thread; tx++) { + tx_conf = &tx_thread[tx]; + printf(" %2d. lcore %d, lthread %d\n", tx, tx_conf->lcore_id, + tx_conf->thread_id); + for (i = 0; i < tx_conf->n_port_id; i++) + printf(" %2d. Tx-port %d\n", i, tx_conf->port_id[i]); + } + +} + +int +main(int argc, char **argv) +{ + struct rte_eth_dev_info dev_info; + struct rte_eth_txconf *txconf; + int ret; + unsigned nb_ports; + uint16_t queueid; + unsigned lcore_id, lthread_id; + uint32_t n_tx_queue, nb_lcores; + uint8_t portid, nb_rx_queue, queue, socketid; + + /* init EAL */ + ret = rte_eal_init(argc, argv); + if (ret < 0) + rte_exit(EXIT_FAILURE, "Invalid EAL parameters\n"); + argc -= ret; + argv += ret; + + /* pre-init dst MACs for all ports to 02:00:00:00:00:xx */ + for (portid = 0; portid < RTE_MAX_ETHPORTS; portid++) { + dest_eth_addr[portid] = ETHER_LOCAL_ADMIN_ADDR + ((uint64_t)portid << + 40); + *(uint64_t *)(val_eth + portid) = dest_eth_addr[portid]; + } + + /* parse application arguments (after the EAL ones) */ + ret = parse_args(argc, argv); + if (ret < 0) + rte_exit(EXIT_FAILURE, "Invalid L3FWD parameters\n"); + + if (check_lcore_params() < 0) + rte_exit(EXIT_FAILURE, "check_lcore_params failed\n"); + + ret = init_lcore_rx_queues(); + if (ret < 0) + rte_exit(EXIT_FAILURE, "init_lcore_rx_queues failed\n"); + + ret = init_lcore_tx_queues(); + if (ret < 0) + rte_exit(EXIT_FAILURE, "init_lcore_tx_queues failed\n"); + + ret = init_lcore_rx_rings(); + if (ret < 0) + rte_exit(EXIT_FAILURE, "init_lcore_rx_rings failed\n"); + + nb_ports = rte_eth_dev_count(); + if (nb_ports > RTE_MAX_ETHPORTS) + nb_ports = RTE_MAX_ETHPORTS; + + if (check_port_config(nb_ports) < 0) + rte_exit(EXIT_FAILURE, "check_port_config failed\n"); + + nb_lcores = rte_lcore_count(); + + /* initialize all ports */ + for (portid = 0; portid < nb_ports; portid++) { + /* skip ports that are not enabled */ + if ((enabled_port_mask & (1 << portid)) == 0) { + printf("\nSkipping disabled port %d\n", portid); + continue; + } + + /* init port */ + printf("Initializing port %d ... ", portid); + fflush(stdout); + + nb_rx_queue = get_port_n_rx_queues(portid); + n_tx_queue = nb_lcores; + + if (n_tx_queue > MAX_TX_QUEUE_PER_PORT) + n_tx_queue = MAX_TX_QUEUE_PER_PORT; + + printf("Creating queues: nb_rxq=%d nb_txq=%u... ", nb_rx_queue, + (unsigned)n_tx_queue); + + ret = rte_eth_dev_configure(portid, nb_rx_queue, (uint16_t)n_tx_queue, + &port_conf); + if (ret < 0) + rte_exit(EXIT_FAILURE, "Cannot configure device: err=%d, port=%d\n", + ret, portid); + + rte_eth_macaddr_get(portid, &ports_eth_addr[portid]); + print_ethaddr(" Address:", &ports_eth_addr[portid]); + printf(", "); + print_ethaddr("Destination:", + (const struct ether_addr *)&dest_eth_addr[portid]); + printf(", "); + + /* + * prepare src MACs for each port. + */ + ether_addr_copy(&ports_eth_addr[portid], + (struct ether_addr *)(val_eth + portid) + 1); + + /* init memory */ + ret = init_mem(NB_MBUF); + if (ret < 0) + rte_exit(EXIT_FAILURE, "init_mem failed\n"); + + /* init one TX queue per couple (lcore,port) */ + queueid = 0; + for (lcore_id = 0; lcore_id < RTE_MAX_LCORE; lcore_id++) { + if (rte_lcore_is_enabled(lcore_id) == 0) + continue; + + if (numa_on) + socketid = (uint8_t)rte_lcore_to_socket_id(lcore_id); + else + socketid = 0; + + printf("txq=%u,%d,%d ", lcore_id, queueid, socketid); + fflush(stdout); + + rte_eth_dev_info_get(portid, &dev_info); + txconf = &dev_info.default_txconf; + if (port_conf.rxmode.jumbo_frame) + txconf->txq_flags = 0; + ret = rte_eth_tx_queue_setup(portid, queueid, nb_txd, + socketid, txconf); + if (ret < 0) + rte_exit(EXIT_FAILURE, "rte_eth_tx_queue_setup: err=%d, " + "port=%d\n", ret, portid); + + queueid++; + } + + /* + * Set other tx queue number for every thread + */ + for (lthread_id = 0; lthread_id < n_tx_thread; lthread_id++) + tx_thread[lthread_id].queue_id[portid] = lthread_id; + + + printf("\n"); + } + + for (lcore_id = 0; lcore_id < RTE_MAX_LCORE; lcore_id++) { + if (rte_lcore_is_enabled(lcore_id) == 0) + continue; + + /* init RX queues */ + for (lthread_id = 0; lthread_id < n_rx_thread; lthread_id++) + for (queue = 0; queue < rx_thread[lthread_id].n_queue; ++queue) { + + portid = rx_thread[lthread_id].queue[queue].port_id; + queueid = rx_thread[lthread_id].queue[queue].queue_id; + + if (numa_on) + socketid = (uint8_t)rte_lcore_to_socket_id(lcore_id); + else + socketid = 0; + + printf("rxq=%d,%d,%d ", portid, queueid, socketid); + fflush(stdout); + + ret = rte_eth_rx_queue_setup(portid, queueid, nb_rxd, socketid, + NULL, pktmbuf_pool[socketid]); + if (ret < 0) + rte_exit(EXIT_FAILURE, "rte_eth_rx_queue_setup: err=%d," + "port=%d\n", ret, portid); + } + } + + printf("\n"); + + /* start ports */ + for (portid = 0; portid < nb_ports; portid++) { + if ((enabled_port_mask & (1 << portid)) == 0) + continue; + + /* Start device */ + ret = rte_eth_dev_start(portid); + if (ret < 0) + rte_exit(EXIT_FAILURE, "rte_eth_dev_start: err=%d, port=%d\n", + ret, portid); + + /* + * If enabled, put device in promiscuous mode. + * This allows IO forwarding mode to forward packets + * to itself through 2 cross-connected ports of the + * target machine. + */ + if (promiscuous_on) + rte_eth_promiscuous_enable(portid); + } + + /* + * Dump rx/tx lthreads configuration + */ + print_config(); + + check_all_ports_link_status((uint8_t)nb_ports, enabled_port_mask); + + /* + * Set number of schedulers to wait for all schedulers + */ + if (lthreads_on) { + printf("Starting L-Threading Model\n"); + lthread_num_schedulers_set(nb_lcores); + rte_eal_mp_remote_launch(lthread_slave_spawner, NULL, SKIP_MASTER); + lthread_master_spawner(NULL); + } else { + printf("Starting P-Threading Model"); + rte_eal_mp_remote_launch(pthread_spawner, NULL, CALL_MASTER); + } + + RTE_LCORE_FOREACH_SLAVE(lcore_id) { + if (rte_eal_wait_lcore(lcore_id) < 0) + return -1; + } + + return 0; +} -- 1.9.3
