X-Git-Url: http://git.droids-corp.org/?a=blobdiff_plain;f=app%2Ftest-eventdev%2Ftest_perf_common.c;h=cc100650c21e4a9e814b179564c3a41849b215e3;hb=2d084f69aa268634fa7ccefa52c980c1e17e17d0;hp=01f782820962ec496ef2c532f3304a7394ff3512;hpb=35b2d13fd6fdcbd191f2a30d74648faeb1186c65;p=dpdk.git diff --git a/app/test-eventdev/test_perf_common.c b/app/test-eventdev/test_perf_common.c index 01f7828209..cc100650c2 100644 --- a/app/test-eventdev/test_perf_common.c +++ b/app/test-eventdev/test_perf_common.c @@ -2,6 +2,8 @@ * Copyright(c) 2017 Cavium, Inc */ +#include + #include "test_perf_common.h" int @@ -28,6 +30,7 @@ perf_test_result(struct evt_test *test, struct evt_options *opt) static inline int perf_producer(void *arg) { + int i; struct prod_data *p = arg; struct test_perf *t = p->t; struct evt_options *opt = t->opt; @@ -38,7 +41,7 @@ perf_producer(void *arg) const uint32_t nb_flows = t->nb_flows; uint32_t flow_counter = 0; uint64_t count = 0; - struct perf_elt *m; + struct perf_elt *m[BURST_SIZE + 1] = {NULL}; struct rte_event ev; if (opt->verbose_level > 1) @@ -54,19 +57,21 @@ perf_producer(void *arg) ev.sub_event_type = 0; /* stage 0 */ while (count < nb_pkts && t->done == false) { - if (rte_mempool_get(pool, (void **)&m) < 0) + if (rte_mempool_get_bulk(pool, (void **)m, BURST_SIZE) < 0) continue; - - ev.flow_id = flow_counter++ % nb_flows; - ev.event_ptr = m; - m->timestamp = rte_get_timer_cycles(); - while (rte_event_enqueue_burst(dev_id, port, &ev, 1) != 1) { - if (t->done) - break; - rte_pause(); - m->timestamp = rte_get_timer_cycles(); + for (i = 0; i < BURST_SIZE; i++) { + ev.flow_id = flow_counter++ % nb_flows; + ev.event_ptr = m[i]; + m[i]->timestamp = rte_get_timer_cycles(); + while (rte_event_enqueue_burst(dev_id, + port, &ev, 1) != 1) { + if (t->done) + break; + rte_pause(); + m[i]->timestamp = rte_get_timer_cycles(); + } } - count++; + count += BURST_SIZE; } return 0; @@ -75,6 +80,7 @@ perf_producer(void *arg) static inline int perf_event_timer_producer(void *arg) { + int i; struct prod_data *p = arg; struct test_perf *t = p->t; struct evt_options *opt = t->opt; @@ -85,17 +91,19 @@ perf_event_timer_producer(void *arg) const uint32_t nb_flows = t->nb_flows; const uint64_t nb_timers = opt->nb_timers; struct rte_mempool *pool = t->pool; - struct perf_elt *m; + struct perf_elt *m[BURST_SIZE + 1] = {NULL}; struct rte_event_timer_adapter **adptr = t->timer_adptr; struct rte_event_timer tim; uint64_t timeout_ticks = opt->expiry_nsec / opt->timer_tick_nsec; memset(&tim, 0, sizeof(struct rte_event_timer)); - timeout_ticks = opt->optm_timer_tick_nsec ? - (timeout_ticks * opt->timer_tick_nsec) - / opt->optm_timer_tick_nsec : timeout_ticks; + timeout_ticks = + opt->optm_timer_tick_nsec + ? ceil((double)(timeout_ticks * opt->timer_tick_nsec) / + opt->optm_timer_tick_nsec) + : timeout_ticks; timeout_ticks += timeout_ticks ? 0 : 1; - tim.ev.event_type = RTE_EVENT_TYPE_TIMER; + tim.ev.event_type = RTE_EVENT_TYPE_TIMER; tim.ev.op = RTE_EVENT_OP_NEW; tim.ev.sched_type = t->opt->sched_type_list[0]; tim.ev.queue_id = p->queue_id; @@ -107,29 +115,31 @@ perf_event_timer_producer(void *arg) printf("%s(): lcore %d\n", __func__, rte_lcore_id()); while (count < nb_timers && t->done == false) { - if (rte_mempool_get(pool, (void **)&m) < 0) + if (rte_mempool_get_bulk(pool, (void **)m, BURST_SIZE) < 0) continue; - - m->tim = tim; - m->tim.ev.flow_id = flow_counter++ % nb_flows; - m->tim.ev.event_ptr = m; - m->timestamp = rte_get_timer_cycles(); - while (rte_event_timer_arm_burst( - adptr[flow_counter % nb_timer_adptrs], - (struct rte_event_timer **)&m, 1) != 1) { - if (t->done) - break; - rte_pause(); - m->timestamp = rte_get_timer_cycles(); + for (i = 0; i < BURST_SIZE; i++) { + rte_prefetch0(m[i + 1]); + m[i]->tim = tim; + m[i]->tim.ev.flow_id = flow_counter++ % nb_flows; + m[i]->tim.ev.event_ptr = m[i]; + m[i]->timestamp = rte_get_timer_cycles(); + while (rte_event_timer_arm_burst( + adptr[flow_counter % nb_timer_adptrs], + (struct rte_event_timer **)&m[i], 1) != 1) { + if (t->done) + break; + m[i]->timestamp = rte_get_timer_cycles(); + } + arm_latency += rte_get_timer_cycles() - m[i]->timestamp; } - arm_latency += rte_get_timer_cycles() - m->timestamp; - count++; + count += BURST_SIZE; } fflush(stdout); rte_delay_ms(1000); printf("%s(): lcore %d Average event timer arm latency = %.3f us\n", - __func__, rte_lcore_id(), (float)(arm_latency / count) / - (rte_get_timer_hz() / 1000000)); + __func__, rte_lcore_id(), + count ? (float)(arm_latency / count) / + (rte_get_timer_hz() / 1000000) : 0); return 0; } @@ -153,11 +163,13 @@ perf_event_timer_producer_burst(void *arg) uint64_t timeout_ticks = opt->expiry_nsec / opt->timer_tick_nsec; memset(&tim, 0, sizeof(struct rte_event_timer)); - timeout_ticks = opt->optm_timer_tick_nsec ? - (timeout_ticks * opt->timer_tick_nsec) - / opt->optm_timer_tick_nsec : timeout_ticks; + timeout_ticks = + opt->optm_timer_tick_nsec + ? ceil((double)(timeout_ticks * opt->timer_tick_nsec) / + opt->optm_timer_tick_nsec) + : timeout_ticks; timeout_ticks += timeout_ticks ? 0 : 1; - tim.ev.event_type = RTE_EVENT_TYPE_TIMER; + tim.ev.event_type = RTE_EVENT_TYPE_TIMER; tim.ev.op = RTE_EVENT_OP_NEW; tim.ev.sched_type = t->opt->sched_type_list[0]; tim.ev.queue_id = p->queue_id; @@ -189,8 +201,9 @@ perf_event_timer_producer_burst(void *arg) fflush(stdout); rte_delay_ms(1000); printf("%s(): lcore %d Average event timer arm latency = %.3f us\n", - __func__, rte_lcore_id(), (float)(arm_latency / count) / - (rte_get_timer_hz() / 1000000)); + __func__, rte_lcore_id(), + count ? (float)(arm_latency / count) / + (rte_get_timer_hz() / 1000000) : 0); return 0; } @@ -217,7 +230,6 @@ processed_pkts(struct test_perf *t) uint8_t i; uint64_t total = 0; - rte_smp_rmb(); for (i = 0; i < t->nb_workers; i++) total += t->worker[i].processed_pkts; @@ -230,7 +242,6 @@ total_latency(struct test_perf *t) uint8_t i; uint64_t total = 0; - rte_smp_rmb(); for (i = 0; i < t->nb_workers; i++) total += t->worker[i].latency; @@ -247,7 +258,7 @@ perf_launch_lcores(struct evt_test *test, struct evt_options *opt, int port_idx = 0; /* launch workers */ - RTE_LCORE_FOREACH_SLAVE(lcore_id) { + RTE_LCORE_FOREACH_WORKER(lcore_id) { if (!(opt->wlcores[lcore_id])) continue; @@ -261,7 +272,7 @@ perf_launch_lcores(struct evt_test *test, struct evt_options *opt, } /* launch producers */ - RTE_LCORE_FOREACH_SLAVE(lcore_id) { + RTE_LCORE_FOREACH_WORKER(lcore_id) { if (!(opt->plcores[lcore_id])) continue; @@ -320,7 +331,6 @@ perf_launch_lcores(struct evt_test *test, struct evt_options *opt, opt->prod_type == EVT_PROD_TYPE_EVENT_TIMER_ADPTR) { t->done = true; - rte_smp_wmb(); break; } } @@ -334,7 +344,6 @@ perf_launch_lcores(struct evt_test *test, struct evt_options *opt, rte_event_dev_dump(opt->dev_id, stdout); evt_err("No schedules for seconds, deadlock"); t->done = true; - rte_smp_wmb(); break; } dead_lock_remaining = remaining; @@ -434,7 +443,7 @@ perf_event_timer_adapter_setup(struct test_perf *t) if (!(adapter_info.caps & RTE_EVENT_TIMER_ADAPTER_CAP_INTERNAL_PORT)) { - uint32_t service_id; + uint32_t service_id = -1U; rte_event_timer_adapter_service_id_get(wl, &service_id); @@ -534,8 +543,8 @@ perf_opt_check(struct evt_options *opt, uint64_t nb_queues) { unsigned int lcores; - /* N producer + N worker + 1 master when producer cores are used - * Else N worker + 1 master when Rx adapter is used + /* N producer + N worker + main when producer cores are used + * Else N worker + main when Rx adapter is used */ lcores = opt->prod_type == EVT_PROD_TYPE_SYNT ? 3 : 2; @@ -545,8 +554,8 @@ perf_opt_check(struct evt_options *opt, uint64_t nb_queues) } /* Validate worker lcores */ - if (evt_lcores_has_overlap(opt->wlcores, rte_get_master_lcore())) { - evt_err("worker lcores overlaps with master lcore"); + if (evt_lcores_has_overlap(opt->wlcores, rte_get_main_lcore())) { + evt_err("worker lcores overlaps with main lcore"); return -1; } if (evt_lcores_has_overlap_multi(opt->wlcores, opt->plcores)) { @@ -562,11 +571,12 @@ perf_opt_check(struct evt_options *opt, uint64_t nb_queues) return -1; } - if (opt->prod_type == EVT_PROD_TYPE_SYNT) { + if (opt->prod_type == EVT_PROD_TYPE_SYNT || + opt->prod_type == EVT_PROD_TYPE_EVENT_TIMER_ADPTR) { /* Validate producer lcores */ if (evt_lcores_has_overlap(opt->plcores, - rte_get_master_lcore())) { - evt_err("producer lcores overlaps with master lcore"); + rte_get_main_lcore())) { + evt_err("producer lcores overlaps with main lcore"); return -1; } if (evt_has_disabled_lcore(opt->plcores)) { @@ -654,6 +664,7 @@ int perf_ethdev_setup(struct evt_test *test, struct evt_options *opt) { uint16_t i; + int ret; struct test_perf *t = evt_test_priv(test); struct rte_eth_conf port_conf = { .rxmode = { @@ -682,7 +693,12 @@ perf_ethdev_setup(struct evt_test *test, struct evt_options *opt) struct rte_eth_dev_info dev_info; struct rte_eth_conf local_port_conf = port_conf; - rte_eth_dev_info_get(i, &dev_info); + ret = rte_eth_dev_info_get(i, &dev_info); + if (ret != 0) { + evt_err("Error during getting device (port %u) info: %s\n", + i, strerror(-ret)); + return ret; + } local_port_conf.rx_adv_conf.rss_conf.rss_hf &= dev_info.flow_type_rss_offloads; @@ -714,7 +730,12 @@ perf_ethdev_setup(struct evt_test *test, struct evt_options *opt) return -EINVAL; } - rte_eth_promiscuous_enable(i); + ret = rte_eth_promiscuous_enable(i); + if (ret != 0) { + evt_err("Failed to enable promiscuous mode for eth port [%d]: %s", + i, rte_strerror(-ret)); + return ret; + } } return 0;