From patchwork Thu Aug 10 15:38:06 2017 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Ilya Maximets X-Patchwork-Id: 800242 Return-Path: X-Original-To: incoming@patchwork.ozlabs.org Delivered-To: patchwork-incoming@bilbo.ozlabs.org Authentication-Results: ozlabs.org; spf=pass (mailfrom) smtp.mailfrom=openvswitch.org (client-ip=140.211.169.12; helo=mail.linuxfoundation.org; envelope-from=ovs-dev-bounces@openvswitch.org; receiver=) Received: from mail.linuxfoundation.org (mail.linuxfoundation.org [140.211.169.12]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by ozlabs.org (Postfix) with ESMTPS id 3xSsnB31s2z9t2r for ; Fri, 11 Aug 2017 01:41:34 +1000 (AEST) Received: from mail.linux-foundation.org (localhost [127.0.0.1]) by mail.linuxfoundation.org (Postfix) with ESMTP id C13F4A7A; Thu, 10 Aug 2017 15:38:33 +0000 (UTC) X-Original-To: ovs-dev@openvswitch.org Delivered-To: ovs-dev@mail.linuxfoundation.org Received: from smtp1.linuxfoundation.org (smtp1.linux-foundation.org [172.17.192.35]) by mail.linuxfoundation.org (Postfix) with ESMTPS id 4975499F for ; Thu, 10 Aug 2017 15:38:32 +0000 (UTC) X-Greylist: domain auto-whitelisted by SQLgrey-1.7.6 Received: from mailout1.w1.samsung.com (mailout1.w1.samsung.com [210.118.77.11]) by smtp1.linuxfoundation.org (Postfix) with ESMTPS id ECA61350 for ; Thu, 10 Aug 2017 15:38:31 +0000 (UTC) Received: from eucas1p2.samsung.com (unknown [182.198.249.207]) by mailout1.w1.samsung.com (KnoxPortal) with ESMTP id 20170810153829euoutp01edc003ca0166a470096cec42ca935edd~ZhjRHNQSF0436604366euoutp01b; Thu, 10 Aug 2017 15:38:29 +0000 (GMT) Received: from eusmges4.samsung.com (unknown [203.254.199.244]) by eucas1p1.samsung.com (KnoxPortal) with ESMTP id 20170810153829eucas1p1215e211ae54c02a1da21261fa4a760c9~ZhjQcboJ11169411694eucas1p1M; Thu, 10 Aug 2017 15:38:29 +0000 (GMT) Received: from eucas1p1.samsung.com ( [182.198.249.206]) by eusmges4.samsung.com (EUCPMTA) with SMTP id 14.08.12944.4FD7C895; Thu, 10 Aug 2017 16:38:28 +0100 (BST) Received: from eusmgms2.samsung.com (unknown [182.198.249.180]) by eucas1p2.samsung.com (KnoxPortal) with ESMTP id 20170810153828eucas1p24d336949b681def57514e9c1eb3e6168~ZhjPi0pVg0083300833eucas1p22; Thu, 10 Aug 2017 15:38:28 +0000 (GMT) X-AuditID: cbfec7f4-f79ab6d000003290-e8-598c7df4cec2 Received: from eusync4.samsung.com ( [203.254.199.214]) by eusmgms2.samsung.com (EUCPMTA) with SMTP id 23.34.20118.4FD7C895; Thu, 10 Aug 2017 16:38:28 +0100 (BST) Received: from imaximets.rnd.samsung.ru ([106.109.129.180]) by eusync4.samsung.com (Oracle Communications Messaging Server 7.0.5.31.0 64bit (built May 5 2014)) with ESMTPA id <0OUH005CN7FL8NB0@eusync4.samsung.com>; Thu, 10 Aug 2017 16:38:28 +0100 (BST) From: Ilya Maximets To: ovs-dev@openvswitch.org, Bhanuprakash Bodireddy Date: Thu, 10 Aug 2017 18:38:06 +0300 Message-id: <1502379486-1568-5-git-send-email-i.maximets@samsung.com> X-Mailer: git-send-email 2.7.4 In-reply-to: <1502379486-1568-1-git-send-email-i.maximets@samsung.com> X-Brightmail-Tracker: H4sIAAAAAAAAA+NgFrrHIsWRmVeSWpSXmKPExsWy7djPc7pfansiDZ7vNLRY/YvTYuczZYtX kxsYLVr6ZzJbPH+xkNnizpWfbBbTPt9mt7jS/pPdYu2hD+wWcz89Z3Tg8li85yWTx7Ob/xk9 nl/rYfF4v+8qm0ffllWMHu/mv2ULYIvisklJzcksSy3St0vgynj55jtLQXN2xYLd+9gbGD96 djFyckgImEjcb+thhLDFJC7cW8/WxcjFISSwlFHi05vJrCAJIYHPjBLn2jNhGpZuaGeCKFrG KHFkymtmCKeZSWLBlsdgo9gEdCROrT4CZosIREg8+D6LHaSIWWAbk8SSx5PAEsICLhIbftxh A7FZBFQlfl5vBrN5BVwlruy4wQSxTk7i5rlOZhCbU8BNYuKeiSwggyQEutklFi05DVTEAeTI Smw6wAxR7yLxZ8VudghbWOLV8S1QtozE5cndUL3NjBINqy4xQjgTGCW+NC+H2mYvcermVTCb WYBPYtK26cwQC3glOtqEIEo8JGbsnMcGYTtKXN76gBXi/ZmMEj9WfWWawCizgJFhFaNIamlx bnpqsYlecWJucWleul5yfu4mRmCsn/53/MsOxsXHrA4xCnAwKvHwJoh2RwqxJpYVV+YeYpTg YFYS4e2o7IkU4k1JrKxKLcqPLyrNSS0+xCjNwaIkzmsb1RYpJJCeWJKanZpakFoEk2Xi4JRq YHQJ0D6b/3jCJrfoFXM5V3MtlnrxZ/tKwXvrplkfXVbyaf/Ro0mKAXf/cOWq+06caP0+JkXo 9nWD+T6Bc5uF3J8vvSG775B5hfEZ8Tkeub4Rxd6XbGMuWh/qdnzbo3eih7/qgahTo/CkSp3D b9XUq189qbbzvCBkr+F3bOqM23s2xPyvtfLtllBiKc5INNRiLipOBACXy4t78QIAAA== X-Brightmail-Tracker: H4sIAAAAAAAAA+NgFvrLLMWRmVeSWpSXmKPExsVy+t/xa7pfansiDf7NYLVY/YvTYuczZYtX kxsYLVr6ZzJbPH+xkNnizpWfbBbTPt9mt7jS/pPdYu2hD+wWcz89Z3Tg8li85yWTx7Ob/xk9 nl/rYfF4v+8qm0ffllWMHu/mv2ULYItys8lITUxJLVJIzUvOT8nMS7dVCg1x07VQUshLzE21 VYrQ9Q0JUlIoS8wpBfKMDNCAg3OAe7CSvl2CW8bLN99ZCpqzKxbs3sfewPjRs4uRk0NCwERi 6YZ2JghbTOLCvfVsXYxcHEICSxglrh87zA7htDJJHF3dyghSxSagI3Fq9REwW0QgQqJlznpG kCJmgW1MEo8vfGIBSQgLuEhs+HGHDcRmEVCV+Hm9GczmFXCVuLLjBtQ6OYmb5zqZQWxOATeJ iXsmgvUKAdWsPPeYaQIj7wJGhlWMIqmlxbnpucVGesWJucWleel6yfm5mxiBQb/t2M8tOxi7 3gUfYhTgYFTi4U0Q7Y4UYk0sK67MPcQowcGsJMLbUdkTKcSbklhZlVqUH19UmpNafIjRFOio icxSosn5wIjMK4k3NDE0tzQ0MrawMDcyUhLnVb/cFCkkkJ5YkpqdmlqQWgTTx8TBKdXAaJe5 ntmla1nEi8unD29Sddvt+bCo4cnJ6euuH0+rqzz30PXLrB32GbGRTvlfHO+zb/Ha28fYu7Jc ZnZ9MWP1Sm3Z5vtmabs/2p81nB+8qeRb9p561bsuIr5dV2Q+7ztboHJB3X7Jx3q+KyfkGhSP NW9OlNef/jR0waE966/EfDJjOJRrfuDvDiWW4oxEQy3mouJEAK2yX2CQAgAA X-MTR: 20000000000000000@CPGS X-CMS-MailID: 20170810153828eucas1p24d336949b681def57514e9c1eb3e6168 X-Msg-Generator: CA X-Sender-IP: 182.198.249.180 X-Local-Sender: =?UTF-8?B?SWx5YSBNYXhpbWV0cxtTUlItVmlydHVhbGl6YXRpb24gTGFi?= =?UTF-8?B?G+yCvOyEseyghOyekBtMZWFkaW5nIEVuZ2luZWVy?= X-Global-Sender: =?UTF-8?B?SWx5YSBNYXhpbWV0cxtTUlItVmlydHVhbGl6YXRpb24gTGFi?= =?UTF-8?B?G1NhbXN1bmcgRWxlY3Ryb25pY3MbTGVhZGluZyBFbmdpbmVlcg==?= X-Sender-Code: =?UTF-8?B?QzEwG0NJU0hRG0MxMEdEMDFHRDAxMDE1NA==?= CMS-TYPE: 201P X-CMS-RootMailID: 20170810153828eucas1p24d336949b681def57514e9c1eb3e6168 X-RootMTR: 20170810153828eucas1p24d336949b681def57514e9c1eb3e6168 References: <1502379486-1568-1-git-send-email-i.maximets@samsung.com> Cc: Heetae Ahn , Ilya Maximets Subject: [ovs-dev] [PATCH RFC v3 4/4] dpif-netdev: Time based output batching. X-BeenThere: ovs-dev@openvswitch.org X-Mailman-Version: 2.1.12 Precedence: list List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , MIME-Version: 1.0 Sender: ovs-dev-bounces@openvswitch.org Errors-To: ovs-dev-bounces@openvswitch.org This allows to collect packets from more than one RX burst and send them together with a configurable maximum latency. 'other_config:output-max-latency' can be used to configure time that a packet can wait in output batch for sending. Signed-off-by: Ilya Maximets --- Notes: * This is an RFC and should not be used for performance testing. * Millisecond granularity is used for now. Can be easily switched to use microseconds instead. lib/dpif-netdev.c | 121 ++++++++++++++++++++++++++++++++++++++++++--------- vswitchd/vswitch.xml | 15 +++++++ 2 files changed, 115 insertions(+), 21 deletions(-) diff --git a/lib/dpif-netdev.c b/lib/dpif-netdev.c index dcf55f3..0d78ae4 100644 --- a/lib/dpif-netdev.c +++ b/lib/dpif-netdev.c @@ -85,6 +85,9 @@ VLOG_DEFINE_THIS_MODULE(dpif_netdev); #define MAX_RECIRC_DEPTH 5 DEFINE_STATIC_PER_THREAD_DATA(uint32_t, recirc_depth, 0) +/* Use instant packet send by default. */ +#define DEFAULT_OUTPUT_MAX_LATENCY 0 + /* Configuration parameters. */ enum { MAX_FLOWS = 65536 }; /* Maximum number of flows in flow table. */ enum { MAX_METERS = 65536 }; /* Maximum number of meters. */ @@ -262,6 +265,9 @@ struct dp_netdev { struct hmap ports; struct seq *port_seq; /* Incremented whenever a port changes. */ + /* The time that a packet can wait in output batch for sending. */ + atomic_uint32_t output_max_latency; + /* Meters. */ struct ovs_mutex meter_locks[N_METER_LOCKS]; struct dp_meter *meters[MAX_METERS]; /* Meter bands. */ @@ -502,6 +508,7 @@ struct tx_port { int qid; long long last_used; struct hmap_node node; + long long output_time; struct dp_packet_batch output_pkts; }; @@ -574,6 +581,9 @@ struct dp_netdev_pmd_thread { * than 'cmap_count(dp->poll_threads)'. */ uint32_t static_tx_qid; + /* Number of filled output batches. */ + int n_output_batches; + struct ovs_mutex port_mutex; /* Mutex for 'poll_list' and 'tx_ports'. */ /* List of rx queues to poll. */ struct hmap poll_list OVS_GUARDED; @@ -669,9 +679,9 @@ static void dp_netdev_add_rxq_to_pmd(struct dp_netdev_pmd_thread *pmd, static void dp_netdev_del_rxq_from_pmd(struct dp_netdev_pmd_thread *pmd, struct rxq_poll *poll) OVS_REQUIRES(pmd->port_mutex); -static void +static int dp_netdev_pmd_flush_output_packets(struct dp_netdev_pmd_thread *pmd, - long long now); + long long now, bool force); static void reconfigure_datapath(struct dp_netdev *dp) OVS_REQUIRES(dp->port_mutex); static bool dp_netdev_pmd_try_ref(struct dp_netdev_pmd_thread *pmd); @@ -1193,6 +1203,7 @@ create_dp_netdev(const char *name, const struct dpif_class *class, conntrack_init(&dp->conntrack); atomic_init(&dp->emc_insert_min, DEFAULT_EM_FLOW_INSERT_MIN); + atomic_init(&dp->output_max_latency, DEFAULT_OUTPUT_MAX_LATENCY); cmap_init(&dp->poll_threads); @@ -2858,7 +2869,7 @@ dpif_netdev_execute(struct dpif *dpif, struct dpif_execute *execute) dp_packet_batch_init_packet(&pp, execute->packet); dp_netdev_execute_actions(pmd, &pp, false, execute->flow, execute->actions, execute->actions_len, now); - dp_netdev_pmd_flush_output_packets(pmd, now); + dp_netdev_pmd_flush_output_packets(pmd, now, true); if (pmd->core_id == NON_PMD_CORE_ID) { ovs_mutex_unlock(&dp->non_pmd_mutex); @@ -2907,6 +2918,16 @@ dpif_netdev_set_config(struct dpif *dpif, const struct smap *other_config) smap_get_ullong(other_config, "emc-insert-inv-prob", DEFAULT_EM_FLOW_INSERT_INV_PROB); uint32_t insert_min, cur_min; + uint32_t output_max_latency, cur_max_latency; + + output_max_latency = smap_get_int(other_config, "output-max-latency", + DEFAULT_OUTPUT_MAX_LATENCY); + atomic_read_relaxed(&dp->output_max_latency, &cur_max_latency); + if (output_max_latency != cur_max_latency) { + atomic_store_relaxed(&dp->output_max_latency, output_max_latency); + VLOG_INFO("Output maximum latency set to %"PRIu32" ms", + output_max_latency); + } if (!nullable_string_is_equal(dp->pmd_cmask, cmask)) { free(dp->pmd_cmask); @@ -3107,11 +3128,12 @@ cycles_count_intermediate(struct dp_netdev_pmd_thread *pmd, non_atomic_ullong_add(&pmd->cycles.n[type], interval); } -static void +static int dp_netdev_pmd_flush_output_on_port(struct dp_netdev_pmd_thread *pmd, struct tx_port *p, long long now) { int tx_qid; + int output_cnt; bool dynamic_txqs; dynamic_txqs = p->port->dynamic_txqs; @@ -3121,21 +3143,39 @@ dp_netdev_pmd_flush_output_on_port(struct dp_netdev_pmd_thread *pmd, tx_qid = pmd->static_tx_qid; } + output_cnt = dp_packet_batch_size(&p->output_pkts); netdev_send(p->port->netdev, tx_qid, &p->output_pkts, dynamic_txqs); dp_packet_batch_init(&p->output_pkts); + + if (output_cnt) { + ovs_assert(pmd->n_output_batches > 0); + pmd->n_output_batches--; + } + return output_cnt; } -static void +static int dp_netdev_pmd_flush_output_packets(struct dp_netdev_pmd_thread *pmd, - long long now) + long long now, bool force) { struct tx_port *p; + int output_cnt = 0; + + if (!pmd->n_output_batches) { + return 0; + } + + if (!now) { + now = time_msec(); + } HMAP_FOR_EACH (p, node, &pmd->send_port_cache) { - if (!dp_packet_batch_is_empty(&p->output_pkts)) { - dp_netdev_pmd_flush_output_on_port(pmd, p, now); + if (!dp_packet_batch_is_empty(&p->output_pkts) + && (force || p->output_time <= now)) { + output_cnt += dp_netdev_pmd_flush_output_on_port(pmd, p, now); } } + return output_cnt; } static int @@ -3145,7 +3185,7 @@ dp_netdev_process_rxq_port(struct dp_netdev_pmd_thread *pmd, { struct dp_packet_batch batch; int error; - int batch_cnt = 0; + int batch_cnt = 0, output_cnt = 0; dp_packet_batch_init(&batch); error = netdev_rxq_recv(rx, &batch); @@ -3156,7 +3196,7 @@ dp_netdev_process_rxq_port(struct dp_netdev_pmd_thread *pmd, batch_cnt = batch.count; dp_netdev_input(pmd, &batch, port_no, now); - dp_netdev_pmd_flush_output_packets(pmd, now); + output_cnt = dp_netdev_pmd_flush_output_packets(pmd, now, false); } else if (error != EAGAIN && error != EOPNOTSUPP) { static struct vlog_rate_limit rl = VLOG_RATE_LIMIT_INIT(1, 5); @@ -3164,7 +3204,7 @@ dp_netdev_process_rxq_port(struct dp_netdev_pmd_thread *pmd, netdev_rxq_get_name(rx), ovs_strerror(error)); } - return batch_cnt; + return batch_cnt + output_cnt; } static struct tx_port * @@ -3691,7 +3731,8 @@ dpif_netdev_run(struct dpif *dpif) struct dp_netdev *dp = get_dp_netdev(dpif); struct dp_netdev_pmd_thread *non_pmd; uint64_t new_tnl_seq; - int process_packets = 0; + int process_packets; + bool need_to_flush = true; ovs_mutex_lock(&dp->port_mutex); non_pmd = dp_netdev_get_pmd(dp, NON_PMD_CORE_ID); @@ -3707,12 +3748,25 @@ dpif_netdev_run(struct dpif *dpif) dp_netdev_process_rxq_port(non_pmd, port->rxqs[i].rx, port->port_no); - cycles_count_intermediate(non_pmd, process_packets ? - PMD_CYCLES_PROCESSING - : PMD_CYCLES_IDLE); + cycles_count_intermediate(non_pmd, process_packets + ? PMD_CYCLES_PROCESSING + : PMD_CYCLES_IDLE); + if (process_packets) { + need_to_flush = false; + } } } } + if (need_to_flush) { + /* We didn't receive anything in the process loop. + * Check if we need to send something. */ + process_packets = dp_netdev_pmd_flush_output_packets(non_pmd, + 0, false); + cycles_count_intermediate(non_pmd, process_packets + ? PMD_CYCLES_PROCESSING + : PMD_CYCLES_IDLE); + } + cycles_count_end(non_pmd, PMD_CYCLES_IDLE); dpif_netdev_xps_revalidate_pmd(non_pmd, time_msec(), false); ovs_mutex_unlock(&dp->non_pmd_mutex); @@ -3764,6 +3818,8 @@ pmd_free_cached_ports(struct dp_netdev_pmd_thread *pmd) { struct tx_port *tx_port_cached; + /* Flush all the queued packets. */ + dp_netdev_pmd_flush_output_packets(pmd, 0, true); /* Free all used tx queue ids. */ dpif_netdev_xps_revalidate_pmd(pmd, 0, true); @@ -3860,7 +3916,6 @@ pmd_thread_main(void *f_) bool exiting; int poll_cnt; int i; - int process_packets = 0; poll_list = NULL; @@ -3890,6 +3945,9 @@ reload: cycles_count_start(pmd); for (;;) { + int process_packets; + bool need_to_flush = true; + for (i = 0; i < poll_cnt; i++) { process_packets = dp_netdev_process_rxq_port(pmd, poll_list[i].rx, @@ -3897,6 +3955,19 @@ reload: cycles_count_intermediate(pmd, process_packets ? PMD_CYCLES_PROCESSING : PMD_CYCLES_IDLE); + if (process_packets) { + need_to_flush = false; + } + } + + if (need_to_flush) { + /* We didn't receive anything in the process loop. + * Check if we need to send something. */ + process_packets = dp_netdev_pmd_flush_output_packets(pmd, + 0, false); + cycles_count_intermediate(pmd, + process_packets ? PMD_CYCLES_PROCESSING + : PMD_CYCLES_IDLE); } if (lc++ > 1024) { @@ -4336,6 +4407,7 @@ dp_netdev_configure_pmd(struct dp_netdev_pmd_thread *pmd, struct dp_netdev *dp, pmd->core_id = core_id; pmd->numa_id = numa_id; pmd->need_reload = false; + pmd->n_output_batches = 0; ovs_refcount_init(&pmd->ref_cnt); latch_init(&pmd->exit_latch); @@ -4521,6 +4593,7 @@ dp_netdev_add_port_tx_to_pmd(struct dp_netdev_pmd_thread *pmd, tx->port = port; tx->qid = -1; + tx->output_time = 0LL; dp_packet_batch_init(&tx->output_pkts); hmap_insert(&pmd->tx_ports, &tx->node, hash_port_no(tx->port->port_no)); @@ -5197,14 +5270,20 @@ dp_execute_cb(void *aux_, struct dp_packet_batch *packets_, dp_netdev_pmd_flush_output_on_port(pmd, p, now); } #endif - - if (OVS_UNLIKELY(dp_packet_batch_size(&p->output_pkts) - + dp_packet_batch_size(packets_) > NETDEV_MAX_BURST)) { - /* Some packets was generated while input batch processing. - * Flush here to avoid overflow. */ + if (dp_packet_batch_size(&p->output_pkts) + + dp_packet_batch_size(packets_) > NETDEV_MAX_BURST) { + /* Flush here to avoid overflow. */ dp_netdev_pmd_flush_output_on_port(pmd, p, now); } + if (dp_packet_batch_is_empty(&p->output_pkts)) { + uint32_t cur_max_latency; + + atomic_read_relaxed(&dp->output_max_latency, &cur_max_latency); + p->output_time = now + cur_max_latency; + pmd->n_output_batches++; + } + DP_PACKET_BATCH_FOR_EACH (packet, packets_) { dp_packet_batch_add(&p->output_pkts, packet); } diff --git a/vswitchd/vswitch.xml b/vswitchd/vswitch.xml index 074535b..23930f0 100644 --- a/vswitchd/vswitch.xml +++ b/vswitchd/vswitch.xml @@ -344,6 +344,21 @@

+ +

+ Specifies the time in milliseconds that a packet can wait in output + batch for sending i.e. amount of time that packet can spend in an + intermediate output queue before sending to netdev. + This option can be used to configure balance between throughput + and latency. Lower values decreases latency while higher values + may be useful to achieve higher performance. +

+

+ Defaults to 0 i.e. instant packet sending (latency optimized). +

+ +