* [dpdk-dev] [PATCH 0/2] Few enhancements for l3fwd-acl @ 2020-09-07 16:39 Konstantin Ananyev 2020-09-07 16:39 ` [dpdk-dev] [PATCH 1/2] examples/l3fwd-acl: add source and destination MAC update Konstantin Ananyev ` (2 more replies) 0 siblings, 3 replies; 21+ messages in thread From: Konstantin Ananyev @ 2020-09-07 16:39 UTC (permalink / raw) To: dev; +Cc: Konstantin Ananyev Konstantin Ananyev (2): examples/l3fwd-acl: add source and destination MAC update examples/l3fwd-acl: allow explicitly select classify method doc/guides/rel_notes/release_20_11.rst | 10 ++ .../sample_app_ug/l3_forward_access_ctrl.rst | 11 +- examples/l3fwd-acl/main.c | 154 ++++++++++++++++-- 3 files changed, 156 insertions(+), 19 deletions(-) -- 2.17.1 ^ permalink raw reply [flat|nested] 21+ messages in thread
* [dpdk-dev] [PATCH 1/2] examples/l3fwd-acl: add source and destination MAC update 2020-09-07 16:39 [dpdk-dev] [PATCH 0/2] Few enhancements for l3fwd-acl Konstantin Ananyev @ 2020-09-07 16:39 ` Konstantin Ananyev 2020-09-07 16:39 ` [dpdk-dev] [PATCH 2/2] examples/l3fwd-acl: allow explicitly select classify method Konstantin Ananyev 2020-09-15 17:04 ` [dpdk-dev] [PATCH v2 0/3] Few enhancements for l3fwd-acl Konstantin Ananyev 2 siblings, 0 replies; 21+ messages in thread From: Konstantin Ananyev @ 2020-09-07 16:39 UTC (permalink / raw) To: dev; +Cc: Konstantin Ananyev Introduces two changes into l3fwd-acl behaviour to make it behave in the same way as l3fwd: - Add a command-line parameter to allow the user to specify the destination mac address for each ethernet port used. - While forwarding the packet update source and destination mac addresses. Signed-off-by: Konstantin Ananyev <konstantin.ananyev@intel.com> --- doc/guides/rel_notes/release_20_11.rst | 5 ++ .../sample_app_ug/l3_forward_access_ctrl.rst | 4 +- examples/l3fwd-acl/main.c | 51 +++++++++++++++++-- 3 files changed, 54 insertions(+), 6 deletions(-) diff --git a/doc/guides/rel_notes/release_20_11.rst b/doc/guides/rel_notes/release_20_11.rst index df227a177..e9e1fba1c 100644 --- a/doc/guides/rel_notes/release_20_11.rst +++ b/doc/guides/rel_notes/release_20_11.rst @@ -55,6 +55,11 @@ New Features Also, make sure to start the actual text at the margin. ======================================================= +* **Add new command-line parameter for l3wfd-acl sample application.** + + Added new optional parameter ``--eth-dest`` for the ``l3fwd-acl`` to allow + the user to specify the destination mac address for each ethernet port used. + Removed Items ------------- diff --git a/doc/guides/sample_app_ug/l3_forward_access_ctrl.rst b/doc/guides/sample_app_ug/l3_forward_access_ctrl.rst index a44fbcd52..91004356e 100644 --- a/doc/guides/sample_app_ug/l3_forward_access_ctrl.rst +++ b/doc/guides/sample_app_ug/l3_forward_access_ctrl.rst @@ -236,7 +236,7 @@ The application has a number of command line options: .. code-block:: console - ./build/l3fwd-acl [EAL options] -- -p PORTMASK [-P] --config(port,queue,lcore)[,(port,queue,lcore)] --rule_ipv4 FILENAME rule_ipv6 FILENAME [--scalar] [--enable-jumbo [--max-pkt-len PKTLEN]] [--no-numa] + ./build/l3fwd-acl [EAL options] -- -p PORTMASK [-P] --config(port,queue,lcore)[,(port,queue,lcore)] --rule_ipv4 FILENAME rule_ipv6 FILENAME [--scalar] [--enable-jumbo [--max-pkt-len PKTLEN]] [--no-numa] [--eth-dest=X,MM:MM:MM:MM:MM:MM] where, @@ -260,6 +260,8 @@ where, * --no-numa: optional, disables numa awareness +* --eth-dest=X,MM:MM:MM:MM:MM:MM: optional, ethernet destination for port X + For example, consider a dual processor socket platform with 8 physical cores, where cores 0-7 and 16-23 appear on socket 0, while cores 8-15 and 24-31 appear on socket 1. diff --git a/examples/l3fwd-acl/main.c b/examples/l3fwd-acl/main.c index 112ec3d2c..39175398e 100644 --- a/examples/l3fwd-acl/main.c +++ b/examples/l3fwd-acl/main.c @@ -39,6 +39,9 @@ #include <rte_string_fns.h> #include <rte_acl.h> +#include <cmdline_parse.h> +#include <cmdline_parse_etheraddr.h> + #if RTE_LOG_DP_LEVEL >= RTE_LOG_DEBUG #define L3FWDACL_DEBUG #endif @@ -81,9 +84,6 @@ 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 struct rte_ether_addr ports_eth_addr[RTE_MAX_ETHPORTS]; - /* mask of enabled ports */ static uint32_t enabled_port_mask; static int promiscuous_on; /**< Ports set in promiscuous mode off by default. */ @@ -143,6 +143,9 @@ static struct rte_eth_conf port_conf = { static struct rte_mempool *pktmbuf_pool[NB_SOCKETS]; +/* ethernet addresses of ports */ +static struct rte_ether_hdr port_l2hdr[RTE_MAX_ETHPORTS]; + /***********************start of ACL part******************************/ #ifdef DO_RFC_1812_CHECKS static inline int @@ -164,6 +167,7 @@ send_single_packet(struct rte_mbuf *m, uint16_t port); #define OPTION_RULE_IPV4 "rule_ipv4" #define OPTION_RULE_IPV6 "rule_ipv6" #define OPTION_SCALAR "scalar" +#define OPTION_ETH_DEST "eth-dest" #define ACL_DENY_SIGNATURE 0xf0000000 #define RTE_LOGTYPE_L3FWDACL RTE_LOGTYPE_USER3 #define acl_log(format, ...) RTE_LOG(ERR, L3FWDACL, format, ##__VA_ARGS__) @@ -1275,9 +1279,14 @@ send_single_packet(struct rte_mbuf *m, uint16_t port) { uint32_t lcore_id; struct lcore_conf *qconf; + struct rte_ether_hdr *eh; lcore_id = rte_lcore_id(); + /* update src and dst mac*/ + eh = rte_pktmbuf_mtod(m, struct rte_ether_hdr *); + memcpy(eh, &port_l2hdr[port], sizeof(eh->d_addr) + sizeof(eh->s_addr)); + qconf = &lcore_conf[lcore_id]; rte_eth_tx_buffer(port, qconf->tx_queue_id[port], qconf->tx_buffer[port], m); @@ -1627,6 +1636,26 @@ parse_config(const char *q_arg) return 0; } +static const char * +parse_eth_dest(const char *optarg) +{ + unsigned long portid; + char *port_end; + + errno = 0; + portid = strtoul(optarg, &port_end, 0); + if (errno != 0 || port_end == optarg || *port_end++ != ',') + return "Invalid format"; + else if (portid >= RTE_MAX_ETHPORTS) + return "port value exceeds RTE_MAX_ETHPORTS(" + RTE_STR(RTE_MAX_ETHPORTS) ")"; + + if (cmdline_parse_etheraddr(NULL, port_end, &port_l2hdr[portid].d_addr, + sizeof(port_l2hdr[portid].d_addr)) < 0) + return "Invalid ethernet address"; + return NULL; +} + /* Parse the argument given in the command line of the application */ static int parse_args(int argc, char **argv) @@ -1642,6 +1671,7 @@ parse_args(int argc, char **argv) {OPTION_RULE_IPV4, 1, 0, 0}, {OPTION_RULE_IPV6, 1, 0, 0}, {OPTION_SCALAR, 0, 0, 0}, + {OPTION_ETH_DEST, 1, 0, 0}, {NULL, 0, 0, 0} }; @@ -1737,6 +1767,16 @@ parse_args(int argc, char **argv) OPTION_SCALAR, sizeof(OPTION_SCALAR))) parm_config.scalar = 1; + if (!strncmp(lgopts[option_index].name, OPTION_ETH_DEST, + sizeof(OPTION_ETH_DEST))) { + const char *serr = parse_eth_dest(optarg); + if (serr != NULL) { + printf("invalid %s value:\"%s\": %s\n", + OPTION_ETH_DEST, optarg, serr); + print_usage(prgname); + return -1; + } + } break; @@ -1965,13 +2005,14 @@ main(int argc, char **argv) "rte_eth_dev_adjust_nb_rx_tx_desc: err=%d, port=%d\n", ret, portid); - ret = rte_eth_macaddr_get(portid, &ports_eth_addr[portid]); + ret = rte_eth_macaddr_get(portid, &port_l2hdr[portid].s_addr); if (ret < 0) rte_exit(EXIT_FAILURE, "rte_eth_macaddr_get: err=%d, port=%d\n", ret, portid); - print_ethaddr(" Address:", &ports_eth_addr[portid]); + print_ethaddr("Dst MAC:", &port_l2hdr[portid].d_addr); + print_ethaddr(", Src MAC:", &port_l2hdr[portid].s_addr); printf(", "); /* init memory */ -- 2.17.1 ^ permalink raw reply [flat|nested] 21+ messages in thread
* [dpdk-dev] [PATCH 2/2] examples/l3fwd-acl: allow explicitly select classify method 2020-09-07 16:39 [dpdk-dev] [PATCH 0/2] Few enhancements for l3fwd-acl Konstantin Ananyev 2020-09-07 16:39 ` [dpdk-dev] [PATCH 1/2] examples/l3fwd-acl: add source and destination MAC update Konstantin Ananyev @ 2020-09-07 16:39 ` Konstantin Ananyev 2020-09-15 17:04 ` [dpdk-dev] [PATCH v2 0/3] Few enhancements for l3fwd-acl Konstantin Ananyev 2 siblings, 0 replies; 21+ messages in thread From: Konstantin Ananyev @ 2020-09-07 16:39 UTC (permalink / raw) To: dev; +Cc: Konstantin Ananyev Replace '--scalar' command-line option with new one: --alg=<algname> to allow user explicitly select desired classify method. This is an optional parameter, if not specified default classify algorithm will be used. Signed-off-by: Konstantin Ananyev <konstantin.ananyev@intel.com> --- doc/guides/rel_notes/release_20_11.rst | 5 + .../sample_app_ug/l3_forward_access_ctrl.rst | 9 +- examples/l3fwd-acl/main.c | 103 ++++++++++++++++-- 3 files changed, 103 insertions(+), 14 deletions(-) diff --git a/doc/guides/rel_notes/release_20_11.rst b/doc/guides/rel_notes/release_20_11.rst index e9e1fba1c..01b1aae2e 100644 --- a/doc/guides/rel_notes/release_20_11.rst +++ b/doc/guides/rel_notes/release_20_11.rst @@ -60,6 +60,11 @@ New Features Added new optional parameter ``--eth-dest`` for the ``l3fwd-acl`` to allow the user to specify the destination mac address for each ethernet port used. +* **Allow explicitly select classify method for l3fwd-acl sample application.** + + Replace ``--scalar`` command-line option with ``--alg=<value>``, to + to allow user explicitly select desired classify method. + Removed Items ------------- diff --git a/doc/guides/sample_app_ug/l3_forward_access_ctrl.rst b/doc/guides/sample_app_ug/l3_forward_access_ctrl.rst index 91004356e..80dd8ba00 100644 --- a/doc/guides/sample_app_ug/l3_forward_access_ctrl.rst +++ b/doc/guides/sample_app_ug/l3_forward_access_ctrl.rst @@ -236,7 +236,7 @@ The application has a number of command line options: .. code-block:: console - ./build/l3fwd-acl [EAL options] -- -p PORTMASK [-P] --config(port,queue,lcore)[,(port,queue,lcore)] --rule_ipv4 FILENAME rule_ipv6 FILENAME [--scalar] [--enable-jumbo [--max-pkt-len PKTLEN]] [--no-numa] [--eth-dest=X,MM:MM:MM:MM:MM:MM] + ./build/l3fwd-acl [EAL options] -- -p PORTMASK [-P] --config(port,queue,lcore)[,(port,queue,lcore)] --rule_ipv4 FILENAME rule_ipv6 FILENAME [--alg=<val>] [--enable-jumbo [--max-pkt-len PKTLEN]] [--no-numa] [--eth-dest=X,MM:MM:MM:MM:MM:MM] where, @@ -252,7 +252,8 @@ where, * --rule_ipv6 FILENAME: Specifies the IPv6 ACL and route rules file -* --scalar: Use a scalar function to perform rule lookup +* --alg=<val>: optional, ACL classify method to use, one of: + ``scalar|sse|avx2|neon|altivec`` * --enable-jumbo: optional, enables jumbo frames @@ -270,7 +271,7 @@ To enable L3 forwarding between two ports, assuming that both ports are in the s .. code-block:: console - ./build/l3fwd-acl -l 1,2 -n 4 -- -p 0x3 --config="(0,0,1),(1,0,2)" --rule_ipv4="./rule_ipv4.db" -- rule_ipv6="./rule_ipv6.db" --scalar + ./build/l3fwd-acl -l 1,2 -n 4 -- -p 0x3 --config="(0,0,1),(1,0,2)" --rule_ipv4="./rule_ipv4.db" -- rule_ipv6="./rule_ipv6.db" --alg=scalar In this command: @@ -296,7 +297,7 @@ In this command: * The --rule_ipv6 option specifies the reading of IPv6 rules sets from the ./ rule_ipv6.db file. -* The --scalar option specifies the performing of rule lookup with a scalar function. +* The --alg=scalar option specifies the performing of rule lookup with a scalar function. Explanation ----------- diff --git a/examples/l3fwd-acl/main.c b/examples/l3fwd-acl/main.c index 39175398e..e0f7701cd 100644 --- a/examples/l3fwd-acl/main.c +++ b/examples/l3fwd-acl/main.c @@ -146,6 +146,32 @@ static struct rte_mempool *pktmbuf_pool[NB_SOCKETS]; /* ethernet addresses of ports */ static struct rte_ether_hdr port_l2hdr[RTE_MAX_ETHPORTS]; +static const struct { + const char *name; + enum rte_acl_classify_alg alg; +} acl_alg[] = { + { + .name = "scalar", + .alg = RTE_ACL_CLASSIFY_SCALAR, + }, + { + .name = "sse", + .alg = RTE_ACL_CLASSIFY_SSE, + }, + { + .name = "avx2", + .alg = RTE_ACL_CLASSIFY_AVX2, + }, + { + .name = "neon", + .alg = RTE_ACL_CLASSIFY_NEON, + }, + { + .name = "altivec", + .alg = RTE_ACL_CLASSIFY_ALTIVEC, + }, +}; + /***********************start of ACL part******************************/ #ifdef DO_RFC_1812_CHECKS static inline int @@ -166,7 +192,7 @@ send_single_packet(struct rte_mbuf *m, uint16_t port); #define OPTION_ENBJMO "enable-jumbo" #define OPTION_RULE_IPV4 "rule_ipv4" #define OPTION_RULE_IPV6 "rule_ipv6" -#define OPTION_SCALAR "scalar" +#define OPTION_ALG "alg" #define OPTION_ETH_DEST "eth-dest" #define ACL_DENY_SIGNATURE 0xf0000000 #define RTE_LOGTYPE_L3FWDACL RTE_LOGTYPE_USER3 @@ -441,7 +467,7 @@ static struct { static struct{ const char *rule_ipv4_name; const char *rule_ipv6_name; - int scalar; + enum rte_acl_classify_alg alg; } parm_config; const char cb_port_delim[] = ":"; @@ -1094,13 +1120,58 @@ add_rules(const char *rule_path, return 0; } +static int +usage_acl_alg(char *buf, size_t sz) +{ + uint32_t i, n, rc, tn; + + n = 0; + tn = 0; + for (i = 0; i < RTE_DIM(acl_alg); i++) { + rc = snprintf(buf + n, sz - n, + i == RTE_DIM(acl_alg) - 1 ? "%s" : "%s|", + acl_alg[i].name); + tn += rc; + if (rc < sz - n) + n += rc; + } + + return tn; +} + +static const char * +str_acl_alg(enum rte_acl_classify_alg alg) +{ + uint32_t i; + + for (i = 0; i != RTE_DIM(acl_alg); i++) { + if (alg == acl_alg[i].alg) + return acl_alg[i].name; + } + + return "default"; +} + +static enum rte_acl_classify_alg +parse_acl_alg(const char *alg) +{ + uint32_t i; + + for (i = 0; i != RTE_DIM(acl_alg); i++) { + if (strcmp(alg, acl_alg[i].name) == 0) + return acl_alg[i].alg; + } + + return RTE_ACL_CLASSIFY_DEFAULT; +} + static void dump_acl_config(void) { printf("ACL option are:\n"); printf(OPTION_RULE_IPV4": %s\n", parm_config.rule_ipv4_name); printf(OPTION_RULE_IPV6": %s\n", parm_config.rule_ipv6_name); - printf(OPTION_SCALAR": %d\n", parm_config.scalar); + printf(OPTION_ALG": %s\n", str_acl_alg(parm_config.alg)); } static int @@ -1141,8 +1212,8 @@ setup_acl(struct rte_acl_rule *route_base, if ((context = rte_acl_create(&acl_param)) == NULL) rte_exit(EXIT_FAILURE, "Failed to create ACL context\n"); - if (parm_config.scalar && rte_acl_set_ctx_classify(context, - RTE_ACL_CLASSIFY_SCALAR) != 0) + if (parm_config.alg != RTE_ACL_CLASSIFY_DEFAULT && + rte_acl_set_ctx_classify(context, parm_config.alg) != 0) rte_exit(EXIT_FAILURE, "Failed to setup classify method for ACL context\n"); @@ -1525,6 +1596,9 @@ init_lcore_rx_queues(void) static void print_usage(const char *prgname) { + char alg[PATH_MAX]; + + usage_acl_alg(alg, sizeof(alg)); printf("%s [EAL options] -- -p PORTMASK -P" "--"OPTION_RULE_IPV4"=FILE" "--"OPTION_RULE_IPV6"=FILE" @@ -1546,8 +1620,8 @@ print_usage(const char *prgname) "character '%c'.\n" " --"OPTION_RULE_IPV6"=FILE: specify the ipv6 rules " "entries file.\n" - " --"OPTION_SCALAR": Use scalar function to do lookup\n", - prgname, ACL_LEAD_CHAR, ROUTE_LEAD_CHAR); + " --"OPTION_ALG": ACL classify method to use, one of: %s\n", + prgname, ACL_LEAD_CHAR, ROUTE_LEAD_CHAR, alg); } static int @@ -1656,6 +1730,7 @@ parse_eth_dest(const char *optarg) return NULL; } + /* Parse the argument given in the command line of the application */ static int parse_args(int argc, char **argv) @@ -1670,7 +1745,7 @@ parse_args(int argc, char **argv) {OPTION_ENBJMO, 0, 0, 0}, {OPTION_RULE_IPV4, 1, 0, 0}, {OPTION_RULE_IPV6, 1, 0, 0}, - {OPTION_SCALAR, 0, 0, 0}, + {OPTION_ALG, 1, 0, 0}, {OPTION_ETH_DEST, 1, 0, 0}, {NULL, 0, 0, 0} }; @@ -1764,8 +1839,16 @@ parse_args(int argc, char **argv) } if (!strncmp(lgopts[option_index].name, - OPTION_SCALAR, sizeof(OPTION_SCALAR))) - parm_config.scalar = 1; + OPTION_ALG, sizeof(OPTION_ALG))) { + parm_config.alg = parse_acl_alg(optarg); + if (parm_config.alg == + RTE_ACL_CLASSIFY_DEFAULT) { + printf("unknown %s value:\"%s\"\n", + OPTION_ALG, optarg); + print_usage(prgname); + return -1; + } + } if (!strncmp(lgopts[option_index].name, OPTION_ETH_DEST, sizeof(OPTION_ETH_DEST))) { -- 2.17.1 ^ permalink raw reply [flat|nested] 21+ messages in thread
* [dpdk-dev] [PATCH v2 0/3] Few enhancements for l3fwd-acl 2020-09-07 16:39 [dpdk-dev] [PATCH 0/2] Few enhancements for l3fwd-acl Konstantin Ananyev 2020-09-07 16:39 ` [dpdk-dev] [PATCH 1/2] examples/l3fwd-acl: add source and destination MAC update Konstantin Ananyev 2020-09-07 16:39 ` [dpdk-dev] [PATCH 2/2] examples/l3fwd-acl: allow explicitly select classify method Konstantin Ananyev @ 2020-09-15 17:04 ` Konstantin Ananyev 2020-09-15 17:04 ` [dpdk-dev] [PATCH v2 1/3] examples/l3fwd-acl: add source and destination MAC update Konstantin Ananyev ` (3 more replies) 2 siblings, 4 replies; 21+ messages in thread From: Konstantin Ananyev @ 2020-09-15 17:04 UTC (permalink / raw) To: dev; +Cc: Konstantin Ananyev Improve l3fwd-acl usability. Allow to specify destination MAC and explicitly select classify algorithm. hese patch series depends on: https://patches.dpdk.org/project/dpdk/list/?series=12241 to be applied first. Konstantin Ananyev (3): examples/l3fwd-acl: add source and destination MAC update examples/l3fwd-acl: allow explicitly select classify method examples/l3fwd-acl: add support for AVX512 doc/guides/rel_notes/release_20_11.rst | 10 ++ .../sample_app_ug/l3_forward_access_ctrl.rst | 11 +- examples/l3fwd-acl/main.c | 158 ++++++++++++++++-- 3 files changed, 160 insertions(+), 19 deletions(-) -- 2.17.1 ^ permalink raw reply [flat|nested] 21+ messages in thread
* [dpdk-dev] [PATCH v2 1/3] examples/l3fwd-acl: add source and destination MAC update 2020-09-15 17:04 ` [dpdk-dev] [PATCH v2 0/3] Few enhancements for l3fwd-acl Konstantin Ananyev @ 2020-09-15 17:04 ` Konstantin Ananyev 2020-09-15 17:04 ` [dpdk-dev] [PATCH v2 2/3] examples/l3fwd-acl: allow explicitly select classify method Konstantin Ananyev ` (2 subsequent siblings) 3 siblings, 0 replies; 21+ messages in thread From: Konstantin Ananyev @ 2020-09-15 17:04 UTC (permalink / raw) To: dev; +Cc: Konstantin Ananyev Introduces two changes into l3fwd-acl behaviour to make it behave in the same way as l3fwd: - Add a command-line parameter to allow the user to specify the destination mac address for each ethernet port used. - While forwarding the packet update source and destination mac addresses. Signed-off-by: Konstantin Ananyev <konstantin.ananyev@intel.com> --- doc/guides/rel_notes/release_20_11.rst | 5 ++ .../sample_app_ug/l3_forward_access_ctrl.rst | 4 +- examples/l3fwd-acl/main.c | 51 +++++++++++++++++-- 3 files changed, 54 insertions(+), 6 deletions(-) diff --git a/doc/guides/rel_notes/release_20_11.rst b/doc/guides/rel_notes/release_20_11.rst index acdd12ef9..cfc275f7e 100644 --- a/doc/guides/rel_notes/release_20_11.rst +++ b/doc/guides/rel_notes/release_20_11.rst @@ -60,6 +60,11 @@ New Features Added new ``RTE_ACL_CLASSIFY_AVX512`` vector implementation, which can processup to 32 flows in parallel. Requires AVX512 support. +* **Add new command-line parameter for l3wfd-acl sample application.** + + Added new optional parameter ``--eth-dest`` for the ``l3fwd-acl`` to allow + the user to specify the destination mac address for each ethernet port used. + Removed Items ------------- diff --git a/doc/guides/sample_app_ug/l3_forward_access_ctrl.rst b/doc/guides/sample_app_ug/l3_forward_access_ctrl.rst index a44fbcd52..91004356e 100644 --- a/doc/guides/sample_app_ug/l3_forward_access_ctrl.rst +++ b/doc/guides/sample_app_ug/l3_forward_access_ctrl.rst @@ -236,7 +236,7 @@ The application has a number of command line options: .. code-block:: console - ./build/l3fwd-acl [EAL options] -- -p PORTMASK [-P] --config(port,queue,lcore)[,(port,queue,lcore)] --rule_ipv4 FILENAME rule_ipv6 FILENAME [--scalar] [--enable-jumbo [--max-pkt-len PKTLEN]] [--no-numa] + ./build/l3fwd-acl [EAL options] -- -p PORTMASK [-P] --config(port,queue,lcore)[,(port,queue,lcore)] --rule_ipv4 FILENAME rule_ipv6 FILENAME [--scalar] [--enable-jumbo [--max-pkt-len PKTLEN]] [--no-numa] [--eth-dest=X,MM:MM:MM:MM:MM:MM] where, @@ -260,6 +260,8 @@ where, * --no-numa: optional, disables numa awareness +* --eth-dest=X,MM:MM:MM:MM:MM:MM: optional, ethernet destination for port X + For example, consider a dual processor socket platform with 8 physical cores, where cores 0-7 and 16-23 appear on socket 0, while cores 8-15 and 24-31 appear on socket 1. diff --git a/examples/l3fwd-acl/main.c b/examples/l3fwd-acl/main.c index 112ec3d2c..39175398e 100644 --- a/examples/l3fwd-acl/main.c +++ b/examples/l3fwd-acl/main.c @@ -39,6 +39,9 @@ #include <rte_string_fns.h> #include <rte_acl.h> +#include <cmdline_parse.h> +#include <cmdline_parse_etheraddr.h> + #if RTE_LOG_DP_LEVEL >= RTE_LOG_DEBUG #define L3FWDACL_DEBUG #endif @@ -81,9 +84,6 @@ 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 struct rte_ether_addr ports_eth_addr[RTE_MAX_ETHPORTS]; - /* mask of enabled ports */ static uint32_t enabled_port_mask; static int promiscuous_on; /**< Ports set in promiscuous mode off by default. */ @@ -143,6 +143,9 @@ static struct rte_eth_conf port_conf = { static struct rte_mempool *pktmbuf_pool[NB_SOCKETS]; +/* ethernet addresses of ports */ +static struct rte_ether_hdr port_l2hdr[RTE_MAX_ETHPORTS]; + /***********************start of ACL part******************************/ #ifdef DO_RFC_1812_CHECKS static inline int @@ -164,6 +167,7 @@ send_single_packet(struct rte_mbuf *m, uint16_t port); #define OPTION_RULE_IPV4 "rule_ipv4" #define OPTION_RULE_IPV6 "rule_ipv6" #define OPTION_SCALAR "scalar" +#define OPTION_ETH_DEST "eth-dest" #define ACL_DENY_SIGNATURE 0xf0000000 #define RTE_LOGTYPE_L3FWDACL RTE_LOGTYPE_USER3 #define acl_log(format, ...) RTE_LOG(ERR, L3FWDACL, format, ##__VA_ARGS__) @@ -1275,9 +1279,14 @@ send_single_packet(struct rte_mbuf *m, uint16_t port) { uint32_t lcore_id; struct lcore_conf *qconf; + struct rte_ether_hdr *eh; lcore_id = rte_lcore_id(); + /* update src and dst mac*/ + eh = rte_pktmbuf_mtod(m, struct rte_ether_hdr *); + memcpy(eh, &port_l2hdr[port], sizeof(eh->d_addr) + sizeof(eh->s_addr)); + qconf = &lcore_conf[lcore_id]; rte_eth_tx_buffer(port, qconf->tx_queue_id[port], qconf->tx_buffer[port], m); @@ -1627,6 +1636,26 @@ parse_config(const char *q_arg) return 0; } +static const char * +parse_eth_dest(const char *optarg) +{ + unsigned long portid; + char *port_end; + + errno = 0; + portid = strtoul(optarg, &port_end, 0); + if (errno != 0 || port_end == optarg || *port_end++ != ',') + return "Invalid format"; + else if (portid >= RTE_MAX_ETHPORTS) + return "port value exceeds RTE_MAX_ETHPORTS(" + RTE_STR(RTE_MAX_ETHPORTS) ")"; + + if (cmdline_parse_etheraddr(NULL, port_end, &port_l2hdr[portid].d_addr, + sizeof(port_l2hdr[portid].d_addr)) < 0) + return "Invalid ethernet address"; + return NULL; +} + /* Parse the argument given in the command line of the application */ static int parse_args(int argc, char **argv) @@ -1642,6 +1671,7 @@ parse_args(int argc, char **argv) {OPTION_RULE_IPV4, 1, 0, 0}, {OPTION_RULE_IPV6, 1, 0, 0}, {OPTION_SCALAR, 0, 0, 0}, + {OPTION_ETH_DEST, 1, 0, 0}, {NULL, 0, 0, 0} }; @@ -1737,6 +1767,16 @@ parse_args(int argc, char **argv) OPTION_SCALAR, sizeof(OPTION_SCALAR))) parm_config.scalar = 1; + if (!strncmp(lgopts[option_index].name, OPTION_ETH_DEST, + sizeof(OPTION_ETH_DEST))) { + const char *serr = parse_eth_dest(optarg); + if (serr != NULL) { + printf("invalid %s value:\"%s\": %s\n", + OPTION_ETH_DEST, optarg, serr); + print_usage(prgname); + return -1; + } + } break; @@ -1965,13 +2005,14 @@ main(int argc, char **argv) "rte_eth_dev_adjust_nb_rx_tx_desc: err=%d, port=%d\n", ret, portid); - ret = rte_eth_macaddr_get(portid, &ports_eth_addr[portid]); + ret = rte_eth_macaddr_get(portid, &port_l2hdr[portid].s_addr); if (ret < 0) rte_exit(EXIT_FAILURE, "rte_eth_macaddr_get: err=%d, port=%d\n", ret, portid); - print_ethaddr(" Address:", &ports_eth_addr[portid]); + print_ethaddr("Dst MAC:", &port_l2hdr[portid].d_addr); + print_ethaddr(", Src MAC:", &port_l2hdr[portid].s_addr); printf(", "); /* init memory */ -- 2.17.1 ^ permalink raw reply [flat|nested] 21+ messages in thread
* [dpdk-dev] [PATCH v2 2/3] examples/l3fwd-acl: allow explicitly select classify method 2020-09-15 17:04 ` [dpdk-dev] [PATCH v2 0/3] Few enhancements for l3fwd-acl Konstantin Ananyev 2020-09-15 17:04 ` [dpdk-dev] [PATCH v2 1/3] examples/l3fwd-acl: add source and destination MAC update Konstantin Ananyev @ 2020-09-15 17:04 ` Konstantin Ananyev 2020-09-15 17:04 ` [dpdk-dev] [PATCH v2 3/3] examples/l3fwd-acl: add support for AVX512 Konstantin Ananyev 2020-10-06 17:16 ` [dpdk-dev] [PATCH v3 0/3] Few enhancements for l3fwd-acl Konstantin Ananyev 3 siblings, 0 replies; 21+ messages in thread From: Konstantin Ananyev @ 2020-09-15 17:04 UTC (permalink / raw) To: dev; +Cc: Konstantin Ananyev Replace '--scalar' command-line option with new one: --alg=<algname> to allow user explicitly select desired classify method. This is an optional parameter, if not specified default classify algorithm will be used. Signed-off-by: Konstantin Ananyev <konstantin.ananyev@intel.com> --- doc/guides/rel_notes/release_20_11.rst | 5 + .../sample_app_ug/l3_forward_access_ctrl.rst | 9 +- examples/l3fwd-acl/main.c | 103 ++++++++++++++++-- 3 files changed, 103 insertions(+), 14 deletions(-) diff --git a/doc/guides/rel_notes/release_20_11.rst b/doc/guides/rel_notes/release_20_11.rst index cfc275f7e..1bc41b56d 100644 --- a/doc/guides/rel_notes/release_20_11.rst +++ b/doc/guides/rel_notes/release_20_11.rst @@ -65,6 +65,11 @@ New Features Added new optional parameter ``--eth-dest`` for the ``l3fwd-acl`` to allow the user to specify the destination mac address for each ethernet port used. +* **Allow explicitly select classify method for l3fwd-acl sample application.** + + Replace ``--scalar`` command-line option with ``--alg=<value>``, to + to allow user explicitly select desired classify method. + Removed Items ------------- diff --git a/doc/guides/sample_app_ug/l3_forward_access_ctrl.rst b/doc/guides/sample_app_ug/l3_forward_access_ctrl.rst index 91004356e..80dd8ba00 100644 --- a/doc/guides/sample_app_ug/l3_forward_access_ctrl.rst +++ b/doc/guides/sample_app_ug/l3_forward_access_ctrl.rst @@ -236,7 +236,7 @@ The application has a number of command line options: .. code-block:: console - ./build/l3fwd-acl [EAL options] -- -p PORTMASK [-P] --config(port,queue,lcore)[,(port,queue,lcore)] --rule_ipv4 FILENAME rule_ipv6 FILENAME [--scalar] [--enable-jumbo [--max-pkt-len PKTLEN]] [--no-numa] [--eth-dest=X,MM:MM:MM:MM:MM:MM] + ./build/l3fwd-acl [EAL options] -- -p PORTMASK [-P] --config(port,queue,lcore)[,(port,queue,lcore)] --rule_ipv4 FILENAME rule_ipv6 FILENAME [--alg=<val>] [--enable-jumbo [--max-pkt-len PKTLEN]] [--no-numa] [--eth-dest=X,MM:MM:MM:MM:MM:MM] where, @@ -252,7 +252,8 @@ where, * --rule_ipv6 FILENAME: Specifies the IPv6 ACL and route rules file -* --scalar: Use a scalar function to perform rule lookup +* --alg=<val>: optional, ACL classify method to use, one of: + ``scalar|sse|avx2|neon|altivec`` * --enable-jumbo: optional, enables jumbo frames @@ -270,7 +271,7 @@ To enable L3 forwarding between two ports, assuming that both ports are in the s .. code-block:: console - ./build/l3fwd-acl -l 1,2 -n 4 -- -p 0x3 --config="(0,0,1),(1,0,2)" --rule_ipv4="./rule_ipv4.db" -- rule_ipv6="./rule_ipv6.db" --scalar + ./build/l3fwd-acl -l 1,2 -n 4 -- -p 0x3 --config="(0,0,1),(1,0,2)" --rule_ipv4="./rule_ipv4.db" -- rule_ipv6="./rule_ipv6.db" --alg=scalar In this command: @@ -296,7 +297,7 @@ In this command: * The --rule_ipv6 option specifies the reading of IPv6 rules sets from the ./ rule_ipv6.db file. -* The --scalar option specifies the performing of rule lookup with a scalar function. +* The --alg=scalar option specifies the performing of rule lookup with a scalar function. Explanation ----------- diff --git a/examples/l3fwd-acl/main.c b/examples/l3fwd-acl/main.c index 39175398e..e0f7701cd 100644 --- a/examples/l3fwd-acl/main.c +++ b/examples/l3fwd-acl/main.c @@ -146,6 +146,32 @@ static struct rte_mempool *pktmbuf_pool[NB_SOCKETS]; /* ethernet addresses of ports */ static struct rte_ether_hdr port_l2hdr[RTE_MAX_ETHPORTS]; +static const struct { + const char *name; + enum rte_acl_classify_alg alg; +} acl_alg[] = { + { + .name = "scalar", + .alg = RTE_ACL_CLASSIFY_SCALAR, + }, + { + .name = "sse", + .alg = RTE_ACL_CLASSIFY_SSE, + }, + { + .name = "avx2", + .alg = RTE_ACL_CLASSIFY_AVX2, + }, + { + .name = "neon", + .alg = RTE_ACL_CLASSIFY_NEON, + }, + { + .name = "altivec", + .alg = RTE_ACL_CLASSIFY_ALTIVEC, + }, +}; + /***********************start of ACL part******************************/ #ifdef DO_RFC_1812_CHECKS static inline int @@ -166,7 +192,7 @@ send_single_packet(struct rte_mbuf *m, uint16_t port); #define OPTION_ENBJMO "enable-jumbo" #define OPTION_RULE_IPV4 "rule_ipv4" #define OPTION_RULE_IPV6 "rule_ipv6" -#define OPTION_SCALAR "scalar" +#define OPTION_ALG "alg" #define OPTION_ETH_DEST "eth-dest" #define ACL_DENY_SIGNATURE 0xf0000000 #define RTE_LOGTYPE_L3FWDACL RTE_LOGTYPE_USER3 @@ -441,7 +467,7 @@ static struct { static struct{ const char *rule_ipv4_name; const char *rule_ipv6_name; - int scalar; + enum rte_acl_classify_alg alg; } parm_config; const char cb_port_delim[] = ":"; @@ -1094,13 +1120,58 @@ add_rules(const char *rule_path, return 0; } +static int +usage_acl_alg(char *buf, size_t sz) +{ + uint32_t i, n, rc, tn; + + n = 0; + tn = 0; + for (i = 0; i < RTE_DIM(acl_alg); i++) { + rc = snprintf(buf + n, sz - n, + i == RTE_DIM(acl_alg) - 1 ? "%s" : "%s|", + acl_alg[i].name); + tn += rc; + if (rc < sz - n) + n += rc; + } + + return tn; +} + +static const char * +str_acl_alg(enum rte_acl_classify_alg alg) +{ + uint32_t i; + + for (i = 0; i != RTE_DIM(acl_alg); i++) { + if (alg == acl_alg[i].alg) + return acl_alg[i].name; + } + + return "default"; +} + +static enum rte_acl_classify_alg +parse_acl_alg(const char *alg) +{ + uint32_t i; + + for (i = 0; i != RTE_DIM(acl_alg); i++) { + if (strcmp(alg, acl_alg[i].name) == 0) + return acl_alg[i].alg; + } + + return RTE_ACL_CLASSIFY_DEFAULT; +} + static void dump_acl_config(void) { printf("ACL option are:\n"); printf(OPTION_RULE_IPV4": %s\n", parm_config.rule_ipv4_name); printf(OPTION_RULE_IPV6": %s\n", parm_config.rule_ipv6_name); - printf(OPTION_SCALAR": %d\n", parm_config.scalar); + printf(OPTION_ALG": %s\n", str_acl_alg(parm_config.alg)); } static int @@ -1141,8 +1212,8 @@ setup_acl(struct rte_acl_rule *route_base, if ((context = rte_acl_create(&acl_param)) == NULL) rte_exit(EXIT_FAILURE, "Failed to create ACL context\n"); - if (parm_config.scalar && rte_acl_set_ctx_classify(context, - RTE_ACL_CLASSIFY_SCALAR) != 0) + if (parm_config.alg != RTE_ACL_CLASSIFY_DEFAULT && + rte_acl_set_ctx_classify(context, parm_config.alg) != 0) rte_exit(EXIT_FAILURE, "Failed to setup classify method for ACL context\n"); @@ -1525,6 +1596,9 @@ init_lcore_rx_queues(void) static void print_usage(const char *prgname) { + char alg[PATH_MAX]; + + usage_acl_alg(alg, sizeof(alg)); printf("%s [EAL options] -- -p PORTMASK -P" "--"OPTION_RULE_IPV4"=FILE" "--"OPTION_RULE_IPV6"=FILE" @@ -1546,8 +1620,8 @@ print_usage(const char *prgname) "character '%c'.\n" " --"OPTION_RULE_IPV6"=FILE: specify the ipv6 rules " "entries file.\n" - " --"OPTION_SCALAR": Use scalar function to do lookup\n", - prgname, ACL_LEAD_CHAR, ROUTE_LEAD_CHAR); + " --"OPTION_ALG": ACL classify method to use, one of: %s\n", + prgname, ACL_LEAD_CHAR, ROUTE_LEAD_CHAR, alg); } static int @@ -1656,6 +1730,7 @@ parse_eth_dest(const char *optarg) return NULL; } + /* Parse the argument given in the command line of the application */ static int parse_args(int argc, char **argv) @@ -1670,7 +1745,7 @@ parse_args(int argc, char **argv) {OPTION_ENBJMO, 0, 0, 0}, {OPTION_RULE_IPV4, 1, 0, 0}, {OPTION_RULE_IPV6, 1, 0, 0}, - {OPTION_SCALAR, 0, 0, 0}, + {OPTION_ALG, 1, 0, 0}, {OPTION_ETH_DEST, 1, 0, 0}, {NULL, 0, 0, 0} }; @@ -1764,8 +1839,16 @@ parse_args(int argc, char **argv) } if (!strncmp(lgopts[option_index].name, - OPTION_SCALAR, sizeof(OPTION_SCALAR))) - parm_config.scalar = 1; + OPTION_ALG, sizeof(OPTION_ALG))) { + parm_config.alg = parse_acl_alg(optarg); + if (parm_config.alg == + RTE_ACL_CLASSIFY_DEFAULT) { + printf("unknown %s value:\"%s\"\n", + OPTION_ALG, optarg); + print_usage(prgname); + return -1; + } + } if (!strncmp(lgopts[option_index].name, OPTION_ETH_DEST, sizeof(OPTION_ETH_DEST))) { -- 2.17.1 ^ permalink raw reply [flat|nested] 21+ messages in thread
* [dpdk-dev] [PATCH v2 3/3] examples/l3fwd-acl: add support for AVX512 2020-09-15 17:04 ` [dpdk-dev] [PATCH v2 0/3] Few enhancements for l3fwd-acl Konstantin Ananyev 2020-09-15 17:04 ` [dpdk-dev] [PATCH v2 1/3] examples/l3fwd-acl: add source and destination MAC update Konstantin Ananyev 2020-09-15 17:04 ` [dpdk-dev] [PATCH v2 2/3] examples/l3fwd-acl: allow explicitly select classify method Konstantin Ananyev @ 2020-09-15 17:04 ` Konstantin Ananyev 2020-10-06 17:16 ` [dpdk-dev] [PATCH v3 0/3] Few enhancements for l3fwd-acl Konstantin Ananyev 3 siblings, 0 replies; 21+ messages in thread From: Konstantin Ananyev @ 2020-09-15 17:04 UTC (permalink / raw) To: dev; +Cc: Konstantin Ananyev Add ability to select AVX512 classify algorithm. Signed-off-by: Konstantin Ananyev <konstantin.ananyev@intel.com> --- doc/guides/sample_app_ug/l3_forward_access_ctrl.rst | 2 +- examples/l3fwd-acl/main.c | 4 ++++ 2 files changed, 5 insertions(+), 1 deletion(-) diff --git a/doc/guides/sample_app_ug/l3_forward_access_ctrl.rst b/doc/guides/sample_app_ug/l3_forward_access_ctrl.rst index 80dd8ba00..6e4d17a2a 100644 --- a/doc/guides/sample_app_ug/l3_forward_access_ctrl.rst +++ b/doc/guides/sample_app_ug/l3_forward_access_ctrl.rst @@ -253,7 +253,7 @@ where, * --rule_ipv6 FILENAME: Specifies the IPv6 ACL and route rules file * --alg=<val>: optional, ACL classify method to use, one of: - ``scalar|sse|avx2|neon|altivec`` + ``scalar|sse|avx2|neon|altivec|avx512`` * --enable-jumbo: optional, enables jumbo frames diff --git a/examples/l3fwd-acl/main.c b/examples/l3fwd-acl/main.c index e0f7701cd..1a27d9d75 100644 --- a/examples/l3fwd-acl/main.c +++ b/examples/l3fwd-acl/main.c @@ -170,6 +170,10 @@ static const struct { .name = "altivec", .alg = RTE_ACL_CLASSIFY_ALTIVEC, }, + { + .name = "avx512", + .alg = RTE_ACL_CLASSIFY_AVX512, + }, }; /***********************start of ACL part******************************/ -- 2.17.1 ^ permalink raw reply [flat|nested] 21+ messages in thread
* [dpdk-dev] [PATCH v3 0/3] Few enhancements for l3fwd-acl 2020-09-15 17:04 ` [dpdk-dev] [PATCH v2 0/3] Few enhancements for l3fwd-acl Konstantin Ananyev ` (2 preceding siblings ...) 2020-09-15 17:04 ` [dpdk-dev] [PATCH v2 3/3] examples/l3fwd-acl: add support for AVX512 Konstantin Ananyev @ 2020-10-06 17:16 ` Konstantin Ananyev 2020-10-06 17:16 ` [dpdk-dev] [PATCH v3 1/3] examples/l3fwd-acl: add source and destination MAC update Konstantin Ananyev ` (3 more replies) 3 siblings, 4 replies; 21+ messages in thread From: Konstantin Ananyev @ 2020-10-06 17:16 UTC (permalink / raw) To: dev; +Cc: Konstantin Ananyev Improve l3fwd-acl usability. Allow to specify destination MAC and explicitly select classify algorithm. Depends-on: series-12721 ("acl: introduce AVX512 classify methods") Konstantin Ananyev (3): examples/l3fwd-acl: add source and destination MAC update examples/l3fwd-acl: allow explicitly select classify method examples/l3fwd-acl: add support for AVX512 doc/guides/rel_notes/release_20_11.rst | 10 ++ .../sample_app_ug/l3_forward_access_ctrl.rst | 11 +- examples/l3fwd-acl/main.c | 162 ++++++++++++++++-- 3 files changed, 164 insertions(+), 19 deletions(-) -- 2.17.1 ^ permalink raw reply [flat|nested] 21+ messages in thread
* [dpdk-dev] [PATCH v3 1/3] examples/l3fwd-acl: add source and destination MAC update 2020-10-06 17:16 ` [dpdk-dev] [PATCH v3 0/3] Few enhancements for l3fwd-acl Konstantin Ananyev @ 2020-10-06 17:16 ` Konstantin Ananyev 2020-10-15 11:58 ` David Marchand 2020-10-06 17:16 ` [dpdk-dev] [PATCH v3 2/3] examples/l3fwd-acl: allow explicitly select classify method Konstantin Ananyev ` (2 subsequent siblings) 3 siblings, 1 reply; 21+ messages in thread From: Konstantin Ananyev @ 2020-10-06 17:16 UTC (permalink / raw) To: dev; +Cc: Konstantin Ananyev Introduces two changes into l3fwd-acl behaviour to make it behave in the same way as l3fwd: - Add a command-line parameter to allow the user to specify the destination mac address for each ethernet port used. - While forwarding the packet update source and destination mac addresses. Signed-off-by: Konstantin Ananyev <konstantin.ananyev@intel.com> --- doc/guides/rel_notes/release_20_11.rst | 5 ++ .../sample_app_ug/l3_forward_access_ctrl.rst | 4 +- examples/l3fwd-acl/main.c | 51 +++++++++++++++++-- 3 files changed, 54 insertions(+), 6 deletions(-) diff --git a/doc/guides/rel_notes/release_20_11.rst b/doc/guides/rel_notes/release_20_11.rst index fc6c5de8d9..53343c780a 100644 --- a/doc/guides/rel_notes/release_20_11.rst +++ b/doc/guides/rel_notes/release_20_11.rst @@ -119,6 +119,11 @@ New Features * Added new ``RTE_ACL_CLASSIFY_AVX512X32`` vector implementation, which can process up to 32 flows in parallel. Requires AVX512 support. +* **Add new command-line parameter for l3wfd-acl sample application.** + + Added new optional parameter ``--eth-dest`` for the ``l3fwd-acl`` to allow + the user to specify the destination mac address for each ethernet port used. + Removed Items ------------- diff --git a/doc/guides/sample_app_ug/l3_forward_access_ctrl.rst b/doc/guides/sample_app_ug/l3_forward_access_ctrl.rst index a44fbcd52c..91004356ed 100644 --- a/doc/guides/sample_app_ug/l3_forward_access_ctrl.rst +++ b/doc/guides/sample_app_ug/l3_forward_access_ctrl.rst @@ -236,7 +236,7 @@ The application has a number of command line options: .. code-block:: console - ./build/l3fwd-acl [EAL options] -- -p PORTMASK [-P] --config(port,queue,lcore)[,(port,queue,lcore)] --rule_ipv4 FILENAME rule_ipv6 FILENAME [--scalar] [--enable-jumbo [--max-pkt-len PKTLEN]] [--no-numa] + ./build/l3fwd-acl [EAL options] -- -p PORTMASK [-P] --config(port,queue,lcore)[,(port,queue,lcore)] --rule_ipv4 FILENAME rule_ipv6 FILENAME [--scalar] [--enable-jumbo [--max-pkt-len PKTLEN]] [--no-numa] [--eth-dest=X,MM:MM:MM:MM:MM:MM] where, @@ -260,6 +260,8 @@ where, * --no-numa: optional, disables numa awareness +* --eth-dest=X,MM:MM:MM:MM:MM:MM: optional, ethernet destination for port X + For example, consider a dual processor socket platform with 8 physical cores, where cores 0-7 and 16-23 appear on socket 0, while cores 8-15 and 24-31 appear on socket 1. diff --git a/examples/l3fwd-acl/main.c b/examples/l3fwd-acl/main.c index 4386584667..a5b8b11ee1 100644 --- a/examples/l3fwd-acl/main.c +++ b/examples/l3fwd-acl/main.c @@ -39,6 +39,9 @@ #include <rte_string_fns.h> #include <rte_acl.h> +#include <cmdline_parse.h> +#include <cmdline_parse_etheraddr.h> + #if RTE_LOG_DP_LEVEL >= RTE_LOG_DEBUG #define L3FWDACL_DEBUG #endif @@ -81,9 +84,6 @@ 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 struct rte_ether_addr ports_eth_addr[RTE_MAX_ETHPORTS]; - /* mask of enabled ports */ static uint32_t enabled_port_mask; static int promiscuous_on; /**< Ports set in promiscuous mode off by default. */ @@ -143,6 +143,9 @@ static struct rte_eth_conf port_conf = { static struct rte_mempool *pktmbuf_pool[NB_SOCKETS]; +/* ethernet addresses of ports */ +static struct rte_ether_hdr port_l2hdr[RTE_MAX_ETHPORTS]; + /***********************start of ACL part******************************/ #ifdef DO_RFC_1812_CHECKS static inline int @@ -164,6 +167,7 @@ send_single_packet(struct rte_mbuf *m, uint16_t port); #define OPTION_RULE_IPV4 "rule_ipv4" #define OPTION_RULE_IPV6 "rule_ipv6" #define OPTION_SCALAR "scalar" +#define OPTION_ETH_DEST "eth-dest" #define ACL_DENY_SIGNATURE 0xf0000000 #define RTE_LOGTYPE_L3FWDACL RTE_LOGTYPE_USER3 #define acl_log(format, ...) RTE_LOG(ERR, L3FWDACL, format, ##__VA_ARGS__) @@ -1275,9 +1279,14 @@ send_single_packet(struct rte_mbuf *m, uint16_t port) { uint32_t lcore_id; struct lcore_conf *qconf; + struct rte_ether_hdr *eh; lcore_id = rte_lcore_id(); + /* update src and dst mac*/ + eh = rte_pktmbuf_mtod(m, struct rte_ether_hdr *); + memcpy(eh, &port_l2hdr[port], sizeof(eh->d_addr) + sizeof(eh->s_addr)); + qconf = &lcore_conf[lcore_id]; rte_eth_tx_buffer(port, qconf->tx_queue_id[port], qconf->tx_buffer[port], m); @@ -1627,6 +1636,26 @@ parse_config(const char *q_arg) return 0; } +static const char * +parse_eth_dest(const char *optarg) +{ + unsigned long portid; + char *port_end; + + errno = 0; + portid = strtoul(optarg, &port_end, 0); + if (errno != 0 || port_end == optarg || *port_end++ != ',') + return "Invalid format"; + else if (portid >= RTE_MAX_ETHPORTS) + return "port value exceeds RTE_MAX_ETHPORTS(" + RTE_STR(RTE_MAX_ETHPORTS) ")"; + + if (cmdline_parse_etheraddr(NULL, port_end, &port_l2hdr[portid].d_addr, + sizeof(port_l2hdr[portid].d_addr)) < 0) + return "Invalid ethernet address"; + return NULL; +} + /* Parse the argument given in the command line of the application */ static int parse_args(int argc, char **argv) @@ -1642,6 +1671,7 @@ parse_args(int argc, char **argv) {OPTION_RULE_IPV4, 1, 0, 0}, {OPTION_RULE_IPV6, 1, 0, 0}, {OPTION_SCALAR, 0, 0, 0}, + {OPTION_ETH_DEST, 1, 0, 0}, {NULL, 0, 0, 0} }; @@ -1737,6 +1767,16 @@ parse_args(int argc, char **argv) OPTION_SCALAR, sizeof(OPTION_SCALAR))) parm_config.scalar = 1; + if (!strncmp(lgopts[option_index].name, OPTION_ETH_DEST, + sizeof(OPTION_ETH_DEST))) { + const char *serr = parse_eth_dest(optarg); + if (serr != NULL) { + printf("invalid %s value:\"%s\": %s\n", + OPTION_ETH_DEST, optarg, serr); + print_usage(prgname); + return -1; + } + } break; @@ -1962,13 +2002,14 @@ main(int argc, char **argv) "rte_eth_dev_adjust_nb_rx_tx_desc: err=%d, port=%d\n", ret, portid); - ret = rte_eth_macaddr_get(portid, &ports_eth_addr[portid]); + ret = rte_eth_macaddr_get(portid, &port_l2hdr[portid].s_addr); if (ret < 0) rte_exit(EXIT_FAILURE, "rte_eth_macaddr_get: err=%d, port=%d\n", ret, portid); - print_ethaddr(" Address:", &ports_eth_addr[portid]); + print_ethaddr("Dst MAC:", &port_l2hdr[portid].d_addr); + print_ethaddr(", Src MAC:", &port_l2hdr[portid].s_addr); printf(", "); /* init memory */ -- 2.17.1 ^ permalink raw reply [flat|nested] 21+ messages in thread
* Re: [dpdk-dev] [PATCH v3 1/3] examples/l3fwd-acl: add source and destination MAC update 2020-10-06 17:16 ` [dpdk-dev] [PATCH v3 1/3] examples/l3fwd-acl: add source and destination MAC update Konstantin Ananyev @ 2020-10-15 11:58 ` David Marchand 2020-10-15 13:06 ` Ananyev, Konstantin 0 siblings, 1 reply; 21+ messages in thread From: David Marchand @ 2020-10-15 11:58 UTC (permalink / raw) To: Konstantin Ananyev; +Cc: dev On Tue, Oct 6, 2020 at 7:17 PM Konstantin Ananyev <konstantin.ananyev@intel.com> wrote: > > Introduces two changes into l3fwd-acl behaviour to make > it behave in the same way as l3fwd: > - Add a command-line parameter to allow the user to specify the > destination mac address for each ethernet port used. > - While forwarding the packet update source and destination mac > addresses. This new parameter is optional, but I see no default for l2 destination in the patch. How does it work when you won't set this option? Bonus question, what keeps us from merging this example with l3fwd? -- David Marchand ^ permalink raw reply [flat|nested] 21+ messages in thread
* Re: [dpdk-dev] [PATCH v3 1/3] examples/l3fwd-acl: add source and destination MAC update 2020-10-15 11:58 ` David Marchand @ 2020-10-15 13:06 ` Ananyev, Konstantin 2020-10-16 12:19 ` David Marchand 0 siblings, 1 reply; 21+ messages in thread From: Ananyev, Konstantin @ 2020-10-15 13:06 UTC (permalink / raw) To: David Marchand; +Cc: dev > On Tue, Oct 6, 2020 at 7:17 PM Konstantin Ananyev > <konstantin.ananyev@intel.com> wrote: > > > > Introduces two changes into l3fwd-acl behaviour to make > > it behave in the same way as l3fwd: > > - Add a command-line parameter to allow the user to specify the > > destination mac address for each ethernet port used. > > - While forwarding the packet update source and destination mac > > addresses. > > This new parameter is optional, but I see no default for l2 > destination in the patch. > How does it work when you won't set this option? Ah yes, forgot to init l2 dst array with some default values. Will fix and re-spin then. > > > Bonus question, what keeps us from merging this example with l3fwd? Nothing, except the effort required. ^ permalink raw reply [flat|nested] 21+ messages in thread
* Re: [dpdk-dev] [PATCH v3 1/3] examples/l3fwd-acl: add source and destination MAC update 2020-10-15 13:06 ` Ananyev, Konstantin @ 2020-10-16 12:19 ` David Marchand 2020-10-16 13:14 ` Ananyev, Konstantin 0 siblings, 1 reply; 21+ messages in thread From: David Marchand @ 2020-10-16 12:19 UTC (permalink / raw) To: Ananyev, Konstantin; +Cc: dev On Thu, Oct 15, 2020 at 3:06 PM Ananyev, Konstantin <konstantin.ananyev@intel.com> wrote: > > On Tue, Oct 6, 2020 at 7:17 PM Konstantin Ananyev > > <konstantin.ananyev@intel.com> wrote: > > > > > > Introduces two changes into l3fwd-acl behaviour to make > > > it behave in the same way as l3fwd: > > > - Add a command-line parameter to allow the user to specify the > > > destination mac address for each ethernet port used. > > > - While forwarding the packet update source and destination mac > > > addresses. > > > > This new parameter is optional, but I see no default for l2 > > destination in the patch. > > How does it work when you won't set this option? > > Ah yes, forgot to init l2 dst array with some default values. > Will fix and re-spin then. Ok, waiting for it, and squash patch 3 as suggested. Thanks. -- David Marchand ^ permalink raw reply [flat|nested] 21+ messages in thread
* Re: [dpdk-dev] [PATCH v3 1/3] examples/l3fwd-acl: add source and destination MAC update 2020-10-16 12:19 ` David Marchand @ 2020-10-16 13:14 ` Ananyev, Konstantin 0 siblings, 0 replies; 21+ messages in thread From: Ananyev, Konstantin @ 2020-10-16 13:14 UTC (permalink / raw) To: David Marchand; +Cc: dev > On Thu, Oct 15, 2020 at 3:06 PM Ananyev, Konstantin > <konstantin.ananyev@intel.com> wrote: > > > On Tue, Oct 6, 2020 at 7:17 PM Konstantin Ananyev > > > <konstantin.ananyev@intel.com> wrote: > > > > > > > > Introduces two changes into l3fwd-acl behaviour to make > > > > it behave in the same way as l3fwd: > > > > - Add a command-line parameter to allow the user to specify the > > > > destination mac address for each ethernet port used. > > > > - While forwarding the packet update source and destination mac > > > > addresses. > > > > > > This new parameter is optional, but I see no default for l2 > > > destination in the patch. > > > How does it work when you won't set this option? > > > > Ah yes, forgot to init l2 dst array with some default values. > > Will fix and re-spin then. > > Ok, waiting for it, and squash patch 3 as suggested. > Thanks. > Just sent. Thanks. ^ permalink raw reply [flat|nested] 21+ messages in thread
* [dpdk-dev] [PATCH v3 2/3] examples/l3fwd-acl: allow explicitly select classify method 2020-10-06 17:16 ` [dpdk-dev] [PATCH v3 0/3] Few enhancements for l3fwd-acl Konstantin Ananyev 2020-10-06 17:16 ` [dpdk-dev] [PATCH v3 1/3] examples/l3fwd-acl: add source and destination MAC update Konstantin Ananyev @ 2020-10-06 17:16 ` Konstantin Ananyev 2020-10-06 17:16 ` [dpdk-dev] [PATCH v3 3/3] examples/l3fwd-acl: add support for AVX512 Konstantin Ananyev 2020-10-16 13:12 ` [dpdk-dev] [PATCH v4 0/2] Few enhancements for l3fwd-acl Konstantin Ananyev 3 siblings, 0 replies; 21+ messages in thread From: Konstantin Ananyev @ 2020-10-06 17:16 UTC (permalink / raw) To: dev; +Cc: Konstantin Ananyev Replace '--scalar' command-line option with new one: --alg=<algname> to allow user explicitly select desired classify method. This is an optional parameter, if not specified default classify algorithm will be used. Signed-off-by: Konstantin Ananyev <konstantin.ananyev@intel.com> --- doc/guides/rel_notes/release_20_11.rst | 5 + .../sample_app_ug/l3_forward_access_ctrl.rst | 9 +- examples/l3fwd-acl/main.c | 103 ++++++++++++++++-- 3 files changed, 103 insertions(+), 14 deletions(-) diff --git a/doc/guides/rel_notes/release_20_11.rst b/doc/guides/rel_notes/release_20_11.rst index 53343c780a..d5dbbd4e34 100644 --- a/doc/guides/rel_notes/release_20_11.rst +++ b/doc/guides/rel_notes/release_20_11.rst @@ -124,6 +124,11 @@ New Features Added new optional parameter ``--eth-dest`` for the ``l3fwd-acl`` to allow the user to specify the destination mac address for each ethernet port used. +* **Allow explicitly select classify method for l3fwd-acl sample application.** + + Replace ``--scalar`` command-line option with ``--alg=<value>``, to + to allow user explicitly select desired classify method. + Removed Items ------------- diff --git a/doc/guides/sample_app_ug/l3_forward_access_ctrl.rst b/doc/guides/sample_app_ug/l3_forward_access_ctrl.rst index 91004356ed..80dd8ba008 100644 --- a/doc/guides/sample_app_ug/l3_forward_access_ctrl.rst +++ b/doc/guides/sample_app_ug/l3_forward_access_ctrl.rst @@ -236,7 +236,7 @@ The application has a number of command line options: .. code-block:: console - ./build/l3fwd-acl [EAL options] -- -p PORTMASK [-P] --config(port,queue,lcore)[,(port,queue,lcore)] --rule_ipv4 FILENAME rule_ipv6 FILENAME [--scalar] [--enable-jumbo [--max-pkt-len PKTLEN]] [--no-numa] [--eth-dest=X,MM:MM:MM:MM:MM:MM] + ./build/l3fwd-acl [EAL options] -- -p PORTMASK [-P] --config(port,queue,lcore)[,(port,queue,lcore)] --rule_ipv4 FILENAME rule_ipv6 FILENAME [--alg=<val>] [--enable-jumbo [--max-pkt-len PKTLEN]] [--no-numa] [--eth-dest=X,MM:MM:MM:MM:MM:MM] where, @@ -252,7 +252,8 @@ where, * --rule_ipv6 FILENAME: Specifies the IPv6 ACL and route rules file -* --scalar: Use a scalar function to perform rule lookup +* --alg=<val>: optional, ACL classify method to use, one of: + ``scalar|sse|avx2|neon|altivec`` * --enable-jumbo: optional, enables jumbo frames @@ -270,7 +271,7 @@ To enable L3 forwarding between two ports, assuming that both ports are in the s .. code-block:: console - ./build/l3fwd-acl -l 1,2 -n 4 -- -p 0x3 --config="(0,0,1),(1,0,2)" --rule_ipv4="./rule_ipv4.db" -- rule_ipv6="./rule_ipv6.db" --scalar + ./build/l3fwd-acl -l 1,2 -n 4 -- -p 0x3 --config="(0,0,1),(1,0,2)" --rule_ipv4="./rule_ipv4.db" -- rule_ipv6="./rule_ipv6.db" --alg=scalar In this command: @@ -296,7 +297,7 @@ In this command: * The --rule_ipv6 option specifies the reading of IPv6 rules sets from the ./ rule_ipv6.db file. -* The --scalar option specifies the performing of rule lookup with a scalar function. +* The --alg=scalar option specifies the performing of rule lookup with a scalar function. Explanation ----------- diff --git a/examples/l3fwd-acl/main.c b/examples/l3fwd-acl/main.c index a5b8b11ee1..c27d4025e8 100644 --- a/examples/l3fwd-acl/main.c +++ b/examples/l3fwd-acl/main.c @@ -146,6 +146,32 @@ static struct rte_mempool *pktmbuf_pool[NB_SOCKETS]; /* ethernet addresses of ports */ static struct rte_ether_hdr port_l2hdr[RTE_MAX_ETHPORTS]; +static const struct { + const char *name; + enum rte_acl_classify_alg alg; +} acl_alg[] = { + { + .name = "scalar", + .alg = RTE_ACL_CLASSIFY_SCALAR, + }, + { + .name = "sse", + .alg = RTE_ACL_CLASSIFY_SSE, + }, + { + .name = "avx2", + .alg = RTE_ACL_CLASSIFY_AVX2, + }, + { + .name = "neon", + .alg = RTE_ACL_CLASSIFY_NEON, + }, + { + .name = "altivec", + .alg = RTE_ACL_CLASSIFY_ALTIVEC, + }, +}; + /***********************start of ACL part******************************/ #ifdef DO_RFC_1812_CHECKS static inline int @@ -166,7 +192,7 @@ send_single_packet(struct rte_mbuf *m, uint16_t port); #define OPTION_ENBJMO "enable-jumbo" #define OPTION_RULE_IPV4 "rule_ipv4" #define OPTION_RULE_IPV6 "rule_ipv6" -#define OPTION_SCALAR "scalar" +#define OPTION_ALG "alg" #define OPTION_ETH_DEST "eth-dest" #define ACL_DENY_SIGNATURE 0xf0000000 #define RTE_LOGTYPE_L3FWDACL RTE_LOGTYPE_USER3 @@ -441,7 +467,7 @@ static struct { static struct{ const char *rule_ipv4_name; const char *rule_ipv6_name; - int scalar; + enum rte_acl_classify_alg alg; } parm_config; const char cb_port_delim[] = ":"; @@ -1094,13 +1120,58 @@ add_rules(const char *rule_path, return 0; } +static int +usage_acl_alg(char *buf, size_t sz) +{ + uint32_t i, n, rc, tn; + + n = 0; + tn = 0; + for (i = 0; i < RTE_DIM(acl_alg); i++) { + rc = snprintf(buf + n, sz - n, + i == RTE_DIM(acl_alg) - 1 ? "%s" : "%s|", + acl_alg[i].name); + tn += rc; + if (rc < sz - n) + n += rc; + } + + return tn; +} + +static const char * +str_acl_alg(enum rte_acl_classify_alg alg) +{ + uint32_t i; + + for (i = 0; i != RTE_DIM(acl_alg); i++) { + if (alg == acl_alg[i].alg) + return acl_alg[i].name; + } + + return "default"; +} + +static enum rte_acl_classify_alg +parse_acl_alg(const char *alg) +{ + uint32_t i; + + for (i = 0; i != RTE_DIM(acl_alg); i++) { + if (strcmp(alg, acl_alg[i].name) == 0) + return acl_alg[i].alg; + } + + return RTE_ACL_CLASSIFY_DEFAULT; +} + static void dump_acl_config(void) { printf("ACL option are:\n"); printf(OPTION_RULE_IPV4": %s\n", parm_config.rule_ipv4_name); printf(OPTION_RULE_IPV6": %s\n", parm_config.rule_ipv6_name); - printf(OPTION_SCALAR": %d\n", parm_config.scalar); + printf(OPTION_ALG": %s\n", str_acl_alg(parm_config.alg)); } static int @@ -1141,8 +1212,8 @@ setup_acl(struct rte_acl_rule *route_base, if ((context = rte_acl_create(&acl_param)) == NULL) rte_exit(EXIT_FAILURE, "Failed to create ACL context\n"); - if (parm_config.scalar && rte_acl_set_ctx_classify(context, - RTE_ACL_CLASSIFY_SCALAR) != 0) + if (parm_config.alg != RTE_ACL_CLASSIFY_DEFAULT && + rte_acl_set_ctx_classify(context, parm_config.alg) != 0) rte_exit(EXIT_FAILURE, "Failed to setup classify method for ACL context\n"); @@ -1525,6 +1596,9 @@ init_lcore_rx_queues(void) static void print_usage(const char *prgname) { + char alg[PATH_MAX]; + + usage_acl_alg(alg, sizeof(alg)); printf("%s [EAL options] -- -p PORTMASK -P" "--"OPTION_RULE_IPV4"=FILE" "--"OPTION_RULE_IPV6"=FILE" @@ -1546,8 +1620,8 @@ print_usage(const char *prgname) "character '%c'.\n" " --"OPTION_RULE_IPV6"=FILE: specify the ipv6 rules " "entries file.\n" - " --"OPTION_SCALAR": Use scalar function to do lookup\n", - prgname, ACL_LEAD_CHAR, ROUTE_LEAD_CHAR); + " --"OPTION_ALG": ACL classify method to use, one of: %s\n", + prgname, ACL_LEAD_CHAR, ROUTE_LEAD_CHAR, alg); } static int @@ -1656,6 +1730,7 @@ parse_eth_dest(const char *optarg) return NULL; } + /* Parse the argument given in the command line of the application */ static int parse_args(int argc, char **argv) @@ -1670,7 +1745,7 @@ parse_args(int argc, char **argv) {OPTION_ENBJMO, 0, 0, 0}, {OPTION_RULE_IPV4, 1, 0, 0}, {OPTION_RULE_IPV6, 1, 0, 0}, - {OPTION_SCALAR, 0, 0, 0}, + {OPTION_ALG, 1, 0, 0}, {OPTION_ETH_DEST, 1, 0, 0}, {NULL, 0, 0, 0} }; @@ -1764,8 +1839,16 @@ parse_args(int argc, char **argv) } if (!strncmp(lgopts[option_index].name, - OPTION_SCALAR, sizeof(OPTION_SCALAR))) - parm_config.scalar = 1; + OPTION_ALG, sizeof(OPTION_ALG))) { + parm_config.alg = parse_acl_alg(optarg); + if (parm_config.alg == + RTE_ACL_CLASSIFY_DEFAULT) { + printf("unknown %s value:\"%s\"\n", + OPTION_ALG, optarg); + print_usage(prgname); + return -1; + } + } if (!strncmp(lgopts[option_index].name, OPTION_ETH_DEST, sizeof(OPTION_ETH_DEST))) { -- 2.17.1 ^ permalink raw reply [flat|nested] 21+ messages in thread
* [dpdk-dev] [PATCH v3 3/3] examples/l3fwd-acl: add support for AVX512 2020-10-06 17:16 ` [dpdk-dev] [PATCH v3 0/3] Few enhancements for l3fwd-acl Konstantin Ananyev 2020-10-06 17:16 ` [dpdk-dev] [PATCH v3 1/3] examples/l3fwd-acl: add source and destination MAC update Konstantin Ananyev 2020-10-06 17:16 ` [dpdk-dev] [PATCH v3 2/3] examples/l3fwd-acl: allow explicitly select classify method Konstantin Ananyev @ 2020-10-06 17:16 ` Konstantin Ananyev 2020-10-15 12:11 ` David Marchand 2020-10-16 13:12 ` [dpdk-dev] [PATCH v4 0/2] Few enhancements for l3fwd-acl Konstantin Ananyev 3 siblings, 1 reply; 21+ messages in thread From: Konstantin Ananyev @ 2020-10-06 17:16 UTC (permalink / raw) To: dev; +Cc: Konstantin Ananyev Add ability to select AVX512 classify algorithm. Signed-off-by: Konstantin Ananyev <konstantin.ananyev@intel.com> --- Depends-on: series-12721 ("acl: introduce AVX512 classify methods") doc/guides/sample_app_ug/l3_forward_access_ctrl.rst | 2 +- examples/l3fwd-acl/main.c | 8 ++++++++ 2 files changed, 9 insertions(+), 1 deletion(-) diff --git a/doc/guides/sample_app_ug/l3_forward_access_ctrl.rst b/doc/guides/sample_app_ug/l3_forward_access_ctrl.rst index 80dd8ba008..6e4d17a2a8 100644 --- a/doc/guides/sample_app_ug/l3_forward_access_ctrl.rst +++ b/doc/guides/sample_app_ug/l3_forward_access_ctrl.rst @@ -253,7 +253,7 @@ where, * --rule_ipv6 FILENAME: Specifies the IPv6 ACL and route rules file * --alg=<val>: optional, ACL classify method to use, one of: - ``scalar|sse|avx2|neon|altivec`` + ``scalar|sse|avx2|neon|altivec|avx512x16|avx512x32`` * --enable-jumbo: optional, enables jumbo frames diff --git a/examples/l3fwd-acl/main.c b/examples/l3fwd-acl/main.c index c27d4025e8..cbb92abdb3 100644 --- a/examples/l3fwd-acl/main.c +++ b/examples/l3fwd-acl/main.c @@ -170,6 +170,14 @@ static const struct { .name = "altivec", .alg = RTE_ACL_CLASSIFY_ALTIVEC, }, + { + .name = "avx512x16", + .alg = RTE_ACL_CLASSIFY_AVX512X16, + }, + { + .name = "avx512x32", + .alg = RTE_ACL_CLASSIFY_AVX512X32, + }, }; /***********************start of ACL part******************************/ -- 2.17.1 ^ permalink raw reply [flat|nested] 21+ messages in thread
* Re: [dpdk-dev] [PATCH v3 3/3] examples/l3fwd-acl: add support for AVX512 2020-10-06 17:16 ` [dpdk-dev] [PATCH v3 3/3] examples/l3fwd-acl: add support for AVX512 Konstantin Ananyev @ 2020-10-15 12:11 ` David Marchand 2020-10-15 12:42 ` Ananyev, Konstantin 0 siblings, 1 reply; 21+ messages in thread From: David Marchand @ 2020-10-15 12:11 UTC (permalink / raw) To: Konstantin Ananyev; +Cc: dev On Tue, Oct 6, 2020 at 7:17 PM Konstantin Ananyev <konstantin.ananyev@intel.com> wrote: > > Add ability to select AVX512 classify algorithm. > > Signed-off-by: Konstantin Ananyev <konstantin.ananyev@intel.com> No need to separate this from patch 2, right? -- David Marchand ^ permalink raw reply [flat|nested] 21+ messages in thread
* Re: [dpdk-dev] [PATCH v3 3/3] examples/l3fwd-acl: add support for AVX512 2020-10-15 12:11 ` David Marchand @ 2020-10-15 12:42 ` Ananyev, Konstantin 0 siblings, 0 replies; 21+ messages in thread From: Ananyev, Konstantin @ 2020-10-15 12:42 UTC (permalink / raw) To: David Marchand; +Cc: dev > On Tue, Oct 6, 2020 at 7:17 PM Konstantin Ananyev > <konstantin.ananyev@intel.com> wrote: > > > > Add ability to select AVX512 classify algorithm. > > > > Signed-off-by: Konstantin Ananyev <konstantin.ananyev@intel.com> > > No need to separate this from patch 2, right? > Yes, could be squashed into one, I think. ^ permalink raw reply [flat|nested] 21+ messages in thread
* [dpdk-dev] [PATCH v4 0/2] Few enhancements for l3fwd-acl 2020-10-06 17:16 ` [dpdk-dev] [PATCH v3 0/3] Few enhancements for l3fwd-acl Konstantin Ananyev ` (2 preceding siblings ...) 2020-10-06 17:16 ` [dpdk-dev] [PATCH v3 3/3] examples/l3fwd-acl: add support for AVX512 Konstantin Ananyev @ 2020-10-16 13:12 ` Konstantin Ananyev 2020-10-16 13:12 ` [dpdk-dev] [PATCH v4 1/2] examples/l3fwd-acl: add source and destination MAC update Konstantin Ananyev ` (2 more replies) 3 siblings, 3 replies; 21+ messages in thread From: Konstantin Ananyev @ 2020-10-16 13:12 UTC (permalink / raw) To: dev; +Cc: Konstantin Ananyev v3 -> v4: Address David comments: - squash patches #2 and #3 into one - set default values for dest MACs Improve l3fwd-acl usability. Allow to specify destination MAC and explicitly select classify algorithm. Konstantin Ananyev (2): examples/l3fwd-acl: add source and destination MAC update examples/l3fwd-acl: allow explicitly select classify method doc/guides/rel_notes/release_20_11.rst | 10 + .../sample_app_ug/l3_forward_access_ctrl.rst | 11 +- examples/l3fwd-acl/main.c | 178 ++++++++++++++++-- 3 files changed, 180 insertions(+), 19 deletions(-) -- 2.17.1 ^ permalink raw reply [flat|nested] 21+ messages in thread
* [dpdk-dev] [PATCH v4 1/2] examples/l3fwd-acl: add source and destination MAC update 2020-10-16 13:12 ` [dpdk-dev] [PATCH v4 0/2] Few enhancements for l3fwd-acl Konstantin Ananyev @ 2020-10-16 13:12 ` Konstantin Ananyev 2020-10-16 13:12 ` [dpdk-dev] [PATCH v4 2/2] examples/l3fwd-acl: allow explicitly select classify method Konstantin Ananyev 2020-10-19 7:03 ` [dpdk-dev] [PATCH v4 0/2] Few enhancements for l3fwd-acl David Marchand 2 siblings, 0 replies; 21+ messages in thread From: Konstantin Ananyev @ 2020-10-16 13:12 UTC (permalink / raw) To: dev; +Cc: Konstantin Ananyev Introduces two changes into l3fwd-acl behaviour to make it behave in the same way as l3fwd: - Add a command-line parameter to allow the user to specify the destination mac address for each ethernet port used. - While forwarding the packet update source and destination mac addresses. Signed-off-by: Konstantin Ananyev <konstantin.ananyev@intel.com> --- doc/guides/rel_notes/release_20_11.rst | 5 ++ .../sample_app_ug/l3_forward_access_ctrl.rst | 4 +- examples/l3fwd-acl/main.c | 67 +++++++++++++++++-- 3 files changed, 70 insertions(+), 6 deletions(-) diff --git a/doc/guides/rel_notes/release_20_11.rst b/doc/guides/rel_notes/release_20_11.rst index cda5b2f5b2..3b23a9a206 100644 --- a/doc/guides/rel_notes/release_20_11.rst +++ b/doc/guides/rel_notes/release_20_11.rst @@ -276,6 +276,11 @@ New Features * Added scatter gather support. * Added NIST GCMVS complaint GMAC test method support. +* **Add new command-line parameter for l3wfd-acl sample application.** + + Added new optional parameter ``--eth-dest`` for the ``l3fwd-acl`` to allow + the user to specify the destination mac address for each ethernet port used. + Removed Items ------------- diff --git a/doc/guides/sample_app_ug/l3_forward_access_ctrl.rst b/doc/guides/sample_app_ug/l3_forward_access_ctrl.rst index a44fbcd52c..91004356ed 100644 --- a/doc/guides/sample_app_ug/l3_forward_access_ctrl.rst +++ b/doc/guides/sample_app_ug/l3_forward_access_ctrl.rst @@ -236,7 +236,7 @@ The application has a number of command line options: .. code-block:: console - ./build/l3fwd-acl [EAL options] -- -p PORTMASK [-P] --config(port,queue,lcore)[,(port,queue,lcore)] --rule_ipv4 FILENAME rule_ipv6 FILENAME [--scalar] [--enable-jumbo [--max-pkt-len PKTLEN]] [--no-numa] + ./build/l3fwd-acl [EAL options] -- -p PORTMASK [-P] --config(port,queue,lcore)[,(port,queue,lcore)] --rule_ipv4 FILENAME rule_ipv6 FILENAME [--scalar] [--enable-jumbo [--max-pkt-len PKTLEN]] [--no-numa] [--eth-dest=X,MM:MM:MM:MM:MM:MM] where, @@ -260,6 +260,8 @@ where, * --no-numa: optional, disables numa awareness +* --eth-dest=X,MM:MM:MM:MM:MM:MM: optional, ethernet destination for port X + For example, consider a dual processor socket platform with 8 physical cores, where cores 0-7 and 16-23 appear on socket 0, while cores 8-15 and 24-31 appear on socket 1. diff --git a/examples/l3fwd-acl/main.c b/examples/l3fwd-acl/main.c index 4386584667..16e9530b3b 100644 --- a/examples/l3fwd-acl/main.c +++ b/examples/l3fwd-acl/main.c @@ -39,6 +39,9 @@ #include <rte_string_fns.h> #include <rte_acl.h> +#include <cmdline_parse.h> +#include <cmdline_parse_etheraddr.h> + #if RTE_LOG_DP_LEVEL >= RTE_LOG_DEBUG #define L3FWDACL_DEBUG #endif @@ -81,9 +84,6 @@ 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 struct rte_ether_addr ports_eth_addr[RTE_MAX_ETHPORTS]; - /* mask of enabled ports */ static uint32_t enabled_port_mask; static int promiscuous_on; /**< Ports set in promiscuous mode off by default. */ @@ -143,6 +143,9 @@ static struct rte_eth_conf port_conf = { static struct rte_mempool *pktmbuf_pool[NB_SOCKETS]; +/* ethernet addresses of ports */ +static struct rte_ether_hdr port_l2hdr[RTE_MAX_ETHPORTS]; + /***********************start of ACL part******************************/ #ifdef DO_RFC_1812_CHECKS static inline int @@ -164,6 +167,7 @@ send_single_packet(struct rte_mbuf *m, uint16_t port); #define OPTION_RULE_IPV4 "rule_ipv4" #define OPTION_RULE_IPV6 "rule_ipv6" #define OPTION_SCALAR "scalar" +#define OPTION_ETH_DEST "eth-dest" #define ACL_DENY_SIGNATURE 0xf0000000 #define RTE_LOGTYPE_L3FWDACL RTE_LOGTYPE_USER3 #define acl_log(format, ...) RTE_LOG(ERR, L3FWDACL, format, ##__VA_ARGS__) @@ -1275,9 +1279,14 @@ send_single_packet(struct rte_mbuf *m, uint16_t port) { uint32_t lcore_id; struct lcore_conf *qconf; + struct rte_ether_hdr *eh; lcore_id = rte_lcore_id(); + /* update src and dst mac*/ + eh = rte_pktmbuf_mtod(m, struct rte_ether_hdr *); + memcpy(eh, &port_l2hdr[port], sizeof(eh->d_addr) + sizeof(eh->s_addr)); + qconf = &lcore_conf[lcore_id]; rte_eth_tx_buffer(port, qconf->tx_queue_id[port], qconf->tx_buffer[port], m); @@ -1627,6 +1636,26 @@ parse_config(const char *q_arg) return 0; } +static const char * +parse_eth_dest(const char *optarg) +{ + unsigned long portid; + char *port_end; + + errno = 0; + portid = strtoul(optarg, &port_end, 0); + if (errno != 0 || port_end == optarg || *port_end++ != ',') + return "Invalid format"; + else if (portid >= RTE_MAX_ETHPORTS) + return "port value exceeds RTE_MAX_ETHPORTS(" + RTE_STR(RTE_MAX_ETHPORTS) ")"; + + if (cmdline_parse_etheraddr(NULL, port_end, &port_l2hdr[portid].d_addr, + sizeof(port_l2hdr[portid].d_addr)) < 0) + return "Invalid ethernet address"; + return NULL; +} + /* Parse the argument given in the command line of the application */ static int parse_args(int argc, char **argv) @@ -1642,6 +1671,7 @@ parse_args(int argc, char **argv) {OPTION_RULE_IPV4, 1, 0, 0}, {OPTION_RULE_IPV6, 1, 0, 0}, {OPTION_SCALAR, 0, 0, 0}, + {OPTION_ETH_DEST, 1, 0, 0}, {NULL, 0, 0, 0} }; @@ -1737,6 +1767,16 @@ parse_args(int argc, char **argv) OPTION_SCALAR, sizeof(OPTION_SCALAR))) parm_config.scalar = 1; + if (!strncmp(lgopts[option_index].name, OPTION_ETH_DEST, + sizeof(OPTION_ETH_DEST))) { + const char *serr = parse_eth_dest(optarg); + if (serr != NULL) { + printf("invalid %s value:\"%s\": %s\n", + OPTION_ETH_DEST, optarg, serr); + print_usage(prgname); + return -1; + } + } break; @@ -1862,6 +1902,20 @@ check_all_ports_link_status(uint32_t port_mask) } } +/* + * build-up default vaues for dest MACs. + */ +static void +set_default_dest_mac(void) +{ + uint32_t i; + + for (i = 0; i != RTE_DIM(port_l2hdr); i++) { + port_l2hdr[i].d_addr.addr_bytes[0] = RTE_ETHER_LOCAL_ADMIN_ADDR; + port_l2hdr[i].d_addr.addr_bytes[5] = i; + } +} + int main(int argc, char **argv) { @@ -1883,6 +1937,8 @@ main(int argc, char **argv) argc -= ret; argv += ret; + set_default_dest_mac(); + /* parse application arguments (after the EAL ones) */ ret = parse_args(argc, argv); if (ret < 0) @@ -1962,13 +2018,14 @@ main(int argc, char **argv) "rte_eth_dev_adjust_nb_rx_tx_desc: err=%d, port=%d\n", ret, portid); - ret = rte_eth_macaddr_get(portid, &ports_eth_addr[portid]); + ret = rte_eth_macaddr_get(portid, &port_l2hdr[portid].s_addr); if (ret < 0) rte_exit(EXIT_FAILURE, "rte_eth_macaddr_get: err=%d, port=%d\n", ret, portid); - print_ethaddr(" Address:", &ports_eth_addr[portid]); + print_ethaddr("Dst MAC:", &port_l2hdr[portid].d_addr); + print_ethaddr(", Src MAC:", &port_l2hdr[portid].s_addr); printf(", "); /* init memory */ -- 2.17.1 ^ permalink raw reply [flat|nested] 21+ messages in thread
* [dpdk-dev] [PATCH v4 2/2] examples/l3fwd-acl: allow explicitly select classify method 2020-10-16 13:12 ` [dpdk-dev] [PATCH v4 0/2] Few enhancements for l3fwd-acl Konstantin Ananyev 2020-10-16 13:12 ` [dpdk-dev] [PATCH v4 1/2] examples/l3fwd-acl: add source and destination MAC update Konstantin Ananyev @ 2020-10-16 13:12 ` Konstantin Ananyev 2020-10-19 7:03 ` [dpdk-dev] [PATCH v4 0/2] Few enhancements for l3fwd-acl David Marchand 2 siblings, 0 replies; 21+ messages in thread From: Konstantin Ananyev @ 2020-10-16 13:12 UTC (permalink / raw) To: dev; +Cc: Konstantin Ananyev Replace '--scalar' command-line option with new one: --alg=<algname> to allow user explicitly select desired classify method. This is an optional parameter, if not specified default classify algorithm will be used. Signed-off-by: Konstantin Ananyev <konstantin.ananyev@intel.com> --- doc/guides/rel_notes/release_20_11.rst | 5 + .../sample_app_ug/l3_forward_access_ctrl.rst | 9 +- examples/l3fwd-acl/main.c | 111 ++++++++++++++++-- 3 files changed, 111 insertions(+), 14 deletions(-) diff --git a/doc/guides/rel_notes/release_20_11.rst b/doc/guides/rel_notes/release_20_11.rst index 3b23a9a206..ee5781458b 100644 --- a/doc/guides/rel_notes/release_20_11.rst +++ b/doc/guides/rel_notes/release_20_11.rst @@ -281,6 +281,11 @@ New Features Added new optional parameter ``--eth-dest`` for the ``l3fwd-acl`` to allow the user to specify the destination mac address for each ethernet port used. +* **Allow explicitly select classify method for l3fwd-acl sample application.** + + Replace ``--scalar`` command-line option with ``--alg=<value>``, to + to allow user explicitly select desired classify method. + Removed Items ------------- diff --git a/doc/guides/sample_app_ug/l3_forward_access_ctrl.rst b/doc/guides/sample_app_ug/l3_forward_access_ctrl.rst index 91004356ed..c2d4ca73ab 100644 --- a/doc/guides/sample_app_ug/l3_forward_access_ctrl.rst +++ b/doc/guides/sample_app_ug/l3_forward_access_ctrl.rst @@ -236,7 +236,7 @@ The application has a number of command line options: .. code-block:: console - ./build/l3fwd-acl [EAL options] -- -p PORTMASK [-P] --config(port,queue,lcore)[,(port,queue,lcore)] --rule_ipv4 FILENAME rule_ipv6 FILENAME [--scalar] [--enable-jumbo [--max-pkt-len PKTLEN]] [--no-numa] [--eth-dest=X,MM:MM:MM:MM:MM:MM] + ./build/l3fwd-acl [EAL options] -- -p PORTMASK [-P] --config(port,queue,lcore)[,(port,queue,lcore)] --rule_ipv4 FILENAME rule_ipv6 FILENAME [--alg=<val>] [--enable-jumbo [--max-pkt-len PKTLEN]] [--no-numa] [--eth-dest=X,MM:MM:MM:MM:MM:MM] where, @@ -252,7 +252,8 @@ where, * --rule_ipv6 FILENAME: Specifies the IPv6 ACL and route rules file -* --scalar: Use a scalar function to perform rule lookup +* --alg=<val>: optional, ACL classify method to use, one of: + ``scalar|sse|avx2|neon|altivec|avx512x16|avx512x32`` * --enable-jumbo: optional, enables jumbo frames @@ -270,7 +271,7 @@ To enable L3 forwarding between two ports, assuming that both ports are in the s .. code-block:: console - ./build/l3fwd-acl -l 1,2 -n 4 -- -p 0x3 --config="(0,0,1),(1,0,2)" --rule_ipv4="./rule_ipv4.db" -- rule_ipv6="./rule_ipv6.db" --scalar + ./build/l3fwd-acl -l 1,2 -n 4 -- -p 0x3 --config="(0,0,1),(1,0,2)" --rule_ipv4="./rule_ipv4.db" -- rule_ipv6="./rule_ipv6.db" --alg=scalar In this command: @@ -296,7 +297,7 @@ In this command: * The --rule_ipv6 option specifies the reading of IPv6 rules sets from the ./ rule_ipv6.db file. -* The --scalar option specifies the performing of rule lookup with a scalar function. +* The --alg=scalar option specifies the performing of rule lookup with a scalar function. Explanation ----------- diff --git a/examples/l3fwd-acl/main.c b/examples/l3fwd-acl/main.c index 16e9530b3b..4aa325072a 100644 --- a/examples/l3fwd-acl/main.c +++ b/examples/l3fwd-acl/main.c @@ -146,6 +146,40 @@ static struct rte_mempool *pktmbuf_pool[NB_SOCKETS]; /* ethernet addresses of ports */ static struct rte_ether_hdr port_l2hdr[RTE_MAX_ETHPORTS]; +static const struct { + const char *name; + enum rte_acl_classify_alg alg; +} acl_alg[] = { + { + .name = "scalar", + .alg = RTE_ACL_CLASSIFY_SCALAR, + }, + { + .name = "sse", + .alg = RTE_ACL_CLASSIFY_SSE, + }, + { + .name = "avx2", + .alg = RTE_ACL_CLASSIFY_AVX2, + }, + { + .name = "neon", + .alg = RTE_ACL_CLASSIFY_NEON, + }, + { + .name = "altivec", + .alg = RTE_ACL_CLASSIFY_ALTIVEC, + }, + { + .name = "avx512x16", + .alg = RTE_ACL_CLASSIFY_AVX512X16, + }, + { + .name = "avx512x32", + .alg = RTE_ACL_CLASSIFY_AVX512X32, + }, +}; + /***********************start of ACL part******************************/ #ifdef DO_RFC_1812_CHECKS static inline int @@ -166,7 +200,7 @@ send_single_packet(struct rte_mbuf *m, uint16_t port); #define OPTION_ENBJMO "enable-jumbo" #define OPTION_RULE_IPV4 "rule_ipv4" #define OPTION_RULE_IPV6 "rule_ipv6" -#define OPTION_SCALAR "scalar" +#define OPTION_ALG "alg" #define OPTION_ETH_DEST "eth-dest" #define ACL_DENY_SIGNATURE 0xf0000000 #define RTE_LOGTYPE_L3FWDACL RTE_LOGTYPE_USER3 @@ -441,7 +475,7 @@ static struct { static struct{ const char *rule_ipv4_name; const char *rule_ipv6_name; - int scalar; + enum rte_acl_classify_alg alg; } parm_config; const char cb_port_delim[] = ":"; @@ -1094,13 +1128,58 @@ add_rules(const char *rule_path, return 0; } +static int +usage_acl_alg(char *buf, size_t sz) +{ + uint32_t i, n, rc, tn; + + n = 0; + tn = 0; + for (i = 0; i < RTE_DIM(acl_alg); i++) { + rc = snprintf(buf + n, sz - n, + i == RTE_DIM(acl_alg) - 1 ? "%s" : "%s|", + acl_alg[i].name); + tn += rc; + if (rc < sz - n) + n += rc; + } + + return tn; +} + +static const char * +str_acl_alg(enum rte_acl_classify_alg alg) +{ + uint32_t i; + + for (i = 0; i != RTE_DIM(acl_alg); i++) { + if (alg == acl_alg[i].alg) + return acl_alg[i].name; + } + + return "default"; +} + +static enum rte_acl_classify_alg +parse_acl_alg(const char *alg) +{ + uint32_t i; + + for (i = 0; i != RTE_DIM(acl_alg); i++) { + if (strcmp(alg, acl_alg[i].name) == 0) + return acl_alg[i].alg; + } + + return RTE_ACL_CLASSIFY_DEFAULT; +} + static void dump_acl_config(void) { printf("ACL option are:\n"); printf(OPTION_RULE_IPV4": %s\n", parm_config.rule_ipv4_name); printf(OPTION_RULE_IPV6": %s\n", parm_config.rule_ipv6_name); - printf(OPTION_SCALAR": %d\n", parm_config.scalar); + printf(OPTION_ALG": %s\n", str_acl_alg(parm_config.alg)); } static int @@ -1141,8 +1220,8 @@ setup_acl(struct rte_acl_rule *route_base, if ((context = rte_acl_create(&acl_param)) == NULL) rte_exit(EXIT_FAILURE, "Failed to create ACL context\n"); - if (parm_config.scalar && rte_acl_set_ctx_classify(context, - RTE_ACL_CLASSIFY_SCALAR) != 0) + if (parm_config.alg != RTE_ACL_CLASSIFY_DEFAULT && + rte_acl_set_ctx_classify(context, parm_config.alg) != 0) rte_exit(EXIT_FAILURE, "Failed to setup classify method for ACL context\n"); @@ -1525,6 +1604,9 @@ init_lcore_rx_queues(void) static void print_usage(const char *prgname) { + char alg[PATH_MAX]; + + usage_acl_alg(alg, sizeof(alg)); printf("%s [EAL options] -- -p PORTMASK -P" "--"OPTION_RULE_IPV4"=FILE" "--"OPTION_RULE_IPV6"=FILE" @@ -1546,8 +1628,8 @@ print_usage(const char *prgname) "character '%c'.\n" " --"OPTION_RULE_IPV6"=FILE: specify the ipv6 rules " "entries file.\n" - " --"OPTION_SCALAR": Use scalar function to do lookup\n", - prgname, ACL_LEAD_CHAR, ROUTE_LEAD_CHAR); + " --"OPTION_ALG": ACL classify method to use, one of: %s\n", + prgname, ACL_LEAD_CHAR, ROUTE_LEAD_CHAR, alg); } static int @@ -1656,6 +1738,7 @@ parse_eth_dest(const char *optarg) return NULL; } + /* Parse the argument given in the command line of the application */ static int parse_args(int argc, char **argv) @@ -1670,7 +1753,7 @@ parse_args(int argc, char **argv) {OPTION_ENBJMO, 0, 0, 0}, {OPTION_RULE_IPV4, 1, 0, 0}, {OPTION_RULE_IPV6, 1, 0, 0}, - {OPTION_SCALAR, 0, 0, 0}, + {OPTION_ALG, 1, 0, 0}, {OPTION_ETH_DEST, 1, 0, 0}, {NULL, 0, 0, 0} }; @@ -1764,8 +1847,16 @@ parse_args(int argc, char **argv) } if (!strncmp(lgopts[option_index].name, - OPTION_SCALAR, sizeof(OPTION_SCALAR))) - parm_config.scalar = 1; + OPTION_ALG, sizeof(OPTION_ALG))) { + parm_config.alg = parse_acl_alg(optarg); + if (parm_config.alg == + RTE_ACL_CLASSIFY_DEFAULT) { + printf("unknown %s value:\"%s\"\n", + OPTION_ALG, optarg); + print_usage(prgname); + return -1; + } + } if (!strncmp(lgopts[option_index].name, OPTION_ETH_DEST, sizeof(OPTION_ETH_DEST))) { -- 2.17.1 ^ permalink raw reply [flat|nested] 21+ messages in thread
* Re: [dpdk-dev] [PATCH v4 0/2] Few enhancements for l3fwd-acl 2020-10-16 13:12 ` [dpdk-dev] [PATCH v4 0/2] Few enhancements for l3fwd-acl Konstantin Ananyev 2020-10-16 13:12 ` [dpdk-dev] [PATCH v4 1/2] examples/l3fwd-acl: add source and destination MAC update Konstantin Ananyev 2020-10-16 13:12 ` [dpdk-dev] [PATCH v4 2/2] examples/l3fwd-acl: allow explicitly select classify method Konstantin Ananyev @ 2020-10-19 7:03 ` David Marchand 2 siblings, 0 replies; 21+ messages in thread From: David Marchand @ 2020-10-19 7:03 UTC (permalink / raw) To: Konstantin Ananyev; +Cc: dev On Fri, Oct 16, 2020 at 3:13 PM Konstantin Ananyev <konstantin.ananyev@intel.com> wrote: > > v3 -> v4: > Address David comments: > - squash patches #2 and #3 into one > - set default values for dest MACs > > Improve l3fwd-acl usability. > Allow to specify destination MAC and explicitly select classify algorithm. > > Konstantin Ananyev (2): > examples/l3fwd-acl: add source and destination MAC update > examples/l3fwd-acl: allow explicitly select classify method > > doc/guides/rel_notes/release_20_11.rst | 10 + > .../sample_app_ug/l3_forward_access_ctrl.rst | 11 +- > examples/l3fwd-acl/main.c | 178 ++++++++++++++++-- > 3 files changed, 180 insertions(+), 19 deletions(-) Rewrote release notes under a single "l3fwd-acl" entry. Series applied, thanks. -- David Marchand ^ permalink raw reply [flat|nested] 21+ messages in thread
end of thread, other threads:[~2020-10-19 7:04 UTC | newest] Thread overview: 21+ messages (download: mbox.gz / follow: Atom feed) -- links below jump to the message on this page -- 2020-09-07 16:39 [dpdk-dev] [PATCH 0/2] Few enhancements for l3fwd-acl Konstantin Ananyev 2020-09-07 16:39 ` [dpdk-dev] [PATCH 1/2] examples/l3fwd-acl: add source and destination MAC update Konstantin Ananyev 2020-09-07 16:39 ` [dpdk-dev] [PATCH 2/2] examples/l3fwd-acl: allow explicitly select classify method Konstantin Ananyev 2020-09-15 17:04 ` [dpdk-dev] [PATCH v2 0/3] Few enhancements for l3fwd-acl Konstantin Ananyev 2020-09-15 17:04 ` [dpdk-dev] [PATCH v2 1/3] examples/l3fwd-acl: add source and destination MAC update Konstantin Ananyev 2020-09-15 17:04 ` [dpdk-dev] [PATCH v2 2/3] examples/l3fwd-acl: allow explicitly select classify method Konstantin Ananyev 2020-09-15 17:04 ` [dpdk-dev] [PATCH v2 3/3] examples/l3fwd-acl: add support for AVX512 Konstantin Ananyev 2020-10-06 17:16 ` [dpdk-dev] [PATCH v3 0/3] Few enhancements for l3fwd-acl Konstantin Ananyev 2020-10-06 17:16 ` [dpdk-dev] [PATCH v3 1/3] examples/l3fwd-acl: add source and destination MAC update Konstantin Ananyev 2020-10-15 11:58 ` David Marchand 2020-10-15 13:06 ` Ananyev, Konstantin 2020-10-16 12:19 ` David Marchand 2020-10-16 13:14 ` Ananyev, Konstantin 2020-10-06 17:16 ` [dpdk-dev] [PATCH v3 2/3] examples/l3fwd-acl: allow explicitly select classify method Konstantin Ananyev 2020-10-06 17:16 ` [dpdk-dev] [PATCH v3 3/3] examples/l3fwd-acl: add support for AVX512 Konstantin Ananyev 2020-10-15 12:11 ` David Marchand 2020-10-15 12:42 ` Ananyev, Konstantin 2020-10-16 13:12 ` [dpdk-dev] [PATCH v4 0/2] Few enhancements for l3fwd-acl Konstantin Ananyev 2020-10-16 13:12 ` [dpdk-dev] [PATCH v4 1/2] examples/l3fwd-acl: add source and destination MAC update Konstantin Ananyev 2020-10-16 13:12 ` [dpdk-dev] [PATCH v4 2/2] examples/l3fwd-acl: allow explicitly select classify method Konstantin Ananyev 2020-10-19 7:03 ` [dpdk-dev] [PATCH v4 0/2] Few enhancements for l3fwd-acl David Marchand
This is a public inbox, see mirroring instructions for how to clone and mirror all data and code used for this inbox; as well as URLs for NNTP newsgroup(s).