From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from dpdk.org (dpdk.org [92.243.14.124]) by inbox.dpdk.org (Postfix) with ESMTP id 6698AA0521; Tue, 3 Nov 2020 19:02:04 +0100 (CET) Received: from [92.243.14.124] (localhost [127.0.0.1]) by dpdk.org (Postfix) with ESMTP id BD6272E8B; Tue, 3 Nov 2020 19:02:02 +0100 (CET) Received: from mga05.intel.com (mga05.intel.com [192.55.52.43]) by dpdk.org (Postfix) with ESMTP id 4005A2C7A for ; Tue, 3 Nov 2020 19:02:01 +0100 (CET) IronPort-SDR: HW3jyqgGFo3bVvNFGR2Av6Qo3lWKPzjV2GjXgXvOyTu6JdFDAu7UUKHGYGkalT1pa1+suKF4U4 rgp0LMu2ipKA== X-IronPort-AV: E=McAfee;i="6000,8403,9794"; a="253811952" X-IronPort-AV: E=Sophos;i="5.77,448,1596524400"; d="scan'208";a="253811952" X-Amp-Result: SKIPPED(no attachment in message) X-Amp-File-Uploaded: False Received: from orsmga005.jf.intel.com ([10.7.209.41]) by fmsmga105.fm.intel.com with ESMTP/TLS/ECDHE-RSA-AES256-GCM-SHA384; 03 Nov 2020 10:01:59 -0800 IronPort-SDR: h76zTRm/kiydIyzyMEzlkMx4dn3hwgIXh87nIOSsWKUEp5vZ9QpSONilNzvQ5/ug6ll7YAPSpk l2yIKdHC2ZLg== X-IronPort-AV: E=Sophos;i="5.77,448,1596524400"; d="scan'208";a="538581727" Received: from vmedvedk-mobl.ger.corp.intel.com (HELO [10.252.8.252]) ([10.252.8.252]) by orsmga005-auth.jf.intel.com with ESMTP/TLS/ECDHE-RSA-AES256-GCM-SHA384; 03 Nov 2020 10:01:57 -0800 To: Dharmik Thakkar , Honnappa Nagarahalli Cc: Bruce Richardson , "dev@dpdk.org" , nd References: <20201029153634.10647-1-dharmik.thakkar@arm.com> <20201102235203.6342-1-dharmik.thakkar@arm.com> <20201102235203.6342-5-dharmik.thakkar@arm.com> <9B5688D2-D668-4EE4-9ECF-150BDEF5A292@arm.com> <73640BD5-01D1-4DE3-B0E8-5A3404734DD3@arm.com> From: "Medvedkin, Vladimir" Message-ID: Date: Tue, 3 Nov 2020 18:01:55 +0000 User-Agent: Mozilla/5.0 (Windows NT 10.0; Win64; x64; rv:78.0) Gecko/20100101 Thunderbird/78.4.0 MIME-Version: 1.0 In-Reply-To: <73640BD5-01D1-4DE3-B0E8-5A3404734DD3@arm.com> Content-Type: text/plain; charset=utf-8; format=flowed Content-Language: en-US Content-Transfer-Encoding: 7bit Subject: Re: [dpdk-dev] [PATCH v2 4/4] test/lpm: avoid code duplication in rcu qsbr perf X-BeenThere: dev@dpdk.org X-Mailman-Version: 2.1.15 Precedence: list List-Id: DPDK patches and discussions List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: dev-bounces@dpdk.org Sender: "dev" Hi, On 03/11/2020 14:03, Dharmik Thakkar wrote: > > >> On Nov 2, 2020, at 11:32 PM, Honnappa Nagarahalli wrote: >> >> >> >>>>> >>>>> Avoid code duplication by combining single and multi threaded tests >>>>> >>>>> Signed-off-by: Dharmik Thakkar >>>>> Reviewed-by: Ruifeng Wang >>>>> --- >>>>> app/test/test_lpm_perf.c | 362 >>>>> ++++++++++----------------------------- >>>>> 1 file changed, 91 insertions(+), 271 deletions(-) >>>>> >>>>> diff --git a/app/test/test_lpm_perf.c b/app/test/test_lpm_perf.c >>>>> index >>>>> 224c92fa3d65..229c835c23f7 100644 >>>>> --- a/app/test/test_lpm_perf.c >>>>> +++ b/app/test/test_lpm_perf.c >>>>> @@ -67,6 +67,12 @@ enum { >>>>> IP_CLASS_C >>>>> }; >>>>> >>>>> +enum { >>>>> +SINGLE_WRITER = 1, >>>>> +MULTI_WRITER_1, >>>>> +MULTI_WRITER_2 >>>>> +}; >>>> Do we need this? Can we use the number of cores instead? >>>> >>> >>> There are 3 combinations of writes (adds/deletes): >>> 1. Write all the entries - in case of a single writer 2. Write half of the entries - >>> in case of multiple writers 3. Write remaining half of the entries - in case of >>> multiple writers >>> >>> So, I think this is required. >> IMO, this is not scalable. Essentially, we need 2 parameters to divide the routes among each writer thread. We need 2 parameters, 1) total number of writers 2) the core ID in the linear space. >> Creating a structure with these 2 and passing that to the writer thread would be better and scalable. > > Yes, agreed this is only applicable for 2 writers. Currently, the multi writer test is only limited to a maximum of 2 writers. > To support more number of writers, we need something like this (which I believe is in lines with your suggestion): > 1. Calculate what each writer will write: single_insert = TOTAL_WRITES / num_writers > 2. Pass core ID in linear space as an argument to the writer function: pos_core > 3. Calculate si and ei in the writer function: si = pos_core * single_insert; ei = si + single_insert > Agree to Honnappa suggestion, for me it looks good, better than previous implementation. > I can update the patch to enable more than 2 writers. > Do you also suggest we expand the scope of the test to test with more than 2 writers? > This will increase the time for which the test is running (which currently is significant even with 2 writers). > I don't see any reason to increase the number of writers more than 2. >> >>> >>>>> + >>>>> /* struct route_rule_count defines the total number of rules in >>>>> following a/b/c >>>>> * each item in a[]/b[]/c[] is the number of common IP address class >>>>> A/B/C, not >>>>> * including the ones for private local network. >>>>> @@ -430,11 +436,16 @@ test_lpm_rcu_qsbr_writer(void *arg) { >>> unsigned >>>>> int i, j, si, ei; uint64_t begin, total_cycles; -uint8_t core_id = >>>>> (uint8_t)((uintptr_t)arg); >>>>> +uint8_t writer_id = (uint8_t)((uintptr_t)arg); >>>>> uint32_t next_hop_add = 0xAA; >>>>> >>>>> -/* 2 writer threads are used */ >>>>> -if (core_id % 2 == 0) { >>>>> +/* Single writer (writer_id = 1) */ >>>>> +if (writer_id == SINGLE_WRITER) { >>>>> +si = 0; >>>>> +ei = NUM_LDEPTH_ROUTE_ENTRIES; >>>>> +} >>>>> +/* 2 Writers (writer_id = 2/3)*/ >>>>> +else if (writer_id == MULTI_WRITER_1) { >>>>> si = 0; >>>>> ei = NUM_LDEPTH_ROUTE_ENTRIES / 2; >>>>> } else { >>>>> @@ -447,29 +458,35 @@ test_lpm_rcu_qsbr_writer(void *arg) for (i = 0; >>>>> i < RCU_ITERATIONS; i++) { >>>>> /* Add all the entries */ >>>>> for (j = si; j < ei; j++) { >>>>> -pthread_mutex_lock(&lpm_mutex); >>>>> +if (writer_id != SINGLE_WRITER) >>>>> +pthread_mutex_lock(&lpm_mutex); >>>>> if (rte_lpm_add(lpm, large_ldepth_route_table[j].ip, >>>>> large_ldepth_route_table[j].depth, >>>>> next_hop_add) != 0) { >>>>> printf("Failed to add iteration %d, route# %d\n", i, j); >>>>> -pthread_mutex_unlock(&lpm_mutex); >>>>> +if (writer_id != SINGLE_WRITER) >>>>> + >>>>> pthread_mutex_unlock(&lpm_mutex); >>>>> return -1; >>>>> } >>>>> -pthread_mutex_unlock(&lpm_mutex); >>>>> +if (writer_id != SINGLE_WRITER) >>>>> +pthread_mutex_unlock(&lpm_mutex); >>>>> } >>>>> >>>>> /* Delete all the entries */ >>>>> for (j = si; j < ei; j++) { >>>>> -pthread_mutex_lock(&lpm_mutex); >>>>> +if (writer_id != SINGLE_WRITER) >>>>> +pthread_mutex_lock(&lpm_mutex); >>>>> if (rte_lpm_delete(lpm, >>>>> large_ldepth_route_table[j].ip, >>>>> large_ldepth_route_table[j].depth) != 0) { printf("Failed to delete >>>>> iteration %d, route# %d\n", i, j); -pthread_mutex_unlock(&lpm_mutex); >>>>> +if (writer_id != SINGLE_WRITER) >>>>> + >>>>> pthread_mutex_unlock(&lpm_mutex); >>>>> return -1; >>>>> } >>>>> -pthread_mutex_unlock(&lpm_mutex); >>>>> +if (writer_id != SINGLE_WRITER) >>>>> +pthread_mutex_unlock(&lpm_mutex); >>>>> } >>>>> } >>>>> >>>>> @@ -482,16 +499,17 @@ test_lpm_rcu_qsbr_writer(void *arg) >>>>> >>>>> /* >>>>> * Functional test: >>>>> - * 2 writers, rest are readers >>>>> + * 1/2 writers, rest are readers >>>>> */ >>>>> static int >>>>> -test_lpm_rcu_perf_multi_writer(void) >>>>> +test_lpm_rcu_perf_multi_writer(uint8_t use_rcu) >>>>> { >>>>> struct rte_lpm_config config; >>>>> size_t sz; >>>>> -unsigned int i; >>>>> +unsigned int i, j; >>>>> uint16_t core_id; >>>>> struct rte_lpm_rcu_config rcu_cfg = {0}; >>>>> +int (*reader_f)(void *arg) = NULL; >>>>> >>>>> if (rte_lcore_count() < 3) { >>>>> printf("Not enough cores for lpm_rcu_perf_autotest, expecting at >>>>> least 3\n"); @@ -504,273 +522,76 @@ >>>>> test_lpm_rcu_perf_multi_writer(void) >>>>> num_cores++; >>>>> } >>>>> >>>>> -printf("\nPerf test: 2 writers, %d readers, RCU integration >>>>> enabled\n", -num_cores - 2); >>>>> - >>>>> -/* Create LPM table */ >>>>> -config.max_rules = NUM_LDEPTH_ROUTE_ENTRIES; - >>> config.number_tbl8s = >>>>> NUM_LDEPTH_ROUTE_ENTRIES; -config.flags = 0; -lpm = >>>>> rte_lpm_create(__func__, SOCKET_ID_ANY, &config); >>>>> -TEST_LPM_ASSERT(lpm != NULL); >>>>> - >>>>> -/* Init RCU variable */ >>>>> -sz = rte_rcu_qsbr_get_memsize(num_cores); >>>>> -rv = (struct rte_rcu_qsbr *)rte_zmalloc("rcu0", sz, >>>>> -RTE_CACHE_LINE_SIZE); -rte_rcu_qsbr_init(rv, num_cores); >>>>> - >>>>> -rcu_cfg.v = rv; >>>>> -/* Assign the RCU variable to LPM */ -if (rte_lpm_rcu_qsbr_add(lpm, >>>>> &rcu_cfg) != 0) { -printf("RCU variable assignment failed\n"); -goto >>>>> error; -} >>>>> - >>>>> -writer_done = 0; >>>>> -__atomic_store_n(&gwrite_cycles, 0, __ATOMIC_RELAXED); >>>>> - >>>>> -__atomic_store_n(&thr_id, 0, __ATOMIC_SEQ_CST); >>>>> - >>>>> -/* Launch reader threads */ >>>>> -for (i = 2; i < num_cores; i++) >>>>> -rte_eal_remote_launch(test_lpm_rcu_qsbr_reader, NULL, >>>>> -enabled_core_ids[i]); >>>>> - >>>>> -/* Launch writer threads */ >>>>> -for (i = 0; i < 2; i++) >>>>> -rte_eal_remote_launch(test_lpm_rcu_qsbr_writer, >>>>> -(void *)(uintptr_t)i, >>>>> -enabled_core_ids[i]); >>>>> - >>>>> -/* Wait for writer threads */ >>>>> -for (i = 0; i < 2; i++) >>>>> -if (rte_eal_wait_lcore(enabled_core_ids[i]) < 0) -goto error; >>>>> - >>>>> -printf("Total LPM Adds: %d\n", TOTAL_WRITES); -printf("Total LPM >>>>> Deletes: %d\n", TOTAL_WRITES); -printf("Average LPM Add/Del: >>>>> %"PRIu64" cycles\n", -__atomic_load_n(&gwrite_cycles, >>>>> __ATOMIC_RELAXED) -/ TOTAL_WRITES); >>>>> - >>>>> -writer_done = 1; >>>>> -/* Wait until all readers have exited */ -for (i = 2; i < num_cores; >>>>> i++) -rte_eal_wait_lcore(enabled_core_ids[i]); >>>>> - >>>>> -rte_lpm_free(lpm); >>>>> -rte_free(rv); >>>>> -lpm = NULL; >>>>> -rv = NULL; >>>>> - >>>>> -/* Test without RCU integration */ >>>>> -printf("\nPerf test: 2 writers, %d readers, RCU integration >>>>> disabled\n", -num_cores - 2); >>>>> - >>>>> -/* Create LPM table */ >>>>> -config.max_rules = NUM_LDEPTH_ROUTE_ENTRIES; - >>> config.number_tbl8s = >>>>> NUM_LDEPTH_ROUTE_ENTRIES; -config.flags = 0; -lpm = >>>>> rte_lpm_create(__func__, SOCKET_ID_ANY, &config); >>>>> -TEST_LPM_ASSERT(lpm != NULL); >>>>> - >>>>> -writer_done = 0; >>>>> -__atomic_store_n(&gwrite_cycles, 0, __ATOMIC_RELAXED); >>>>> -__atomic_store_n(&thr_id, 0, __ATOMIC_SEQ_CST); >>>>> - >>>>> -/* Launch reader threads */ >>>>> -for (i = 2; i < num_cores; i++) >>>>> -rte_eal_remote_launch(test_lpm_reader, NULL, -enabled_core_ids[i]); >>>>> - >>>>> -/* Launch writer threads */ >>>>> -for (i = 0; i < 2; i++) >>>>> -rte_eal_remote_launch(test_lpm_rcu_qsbr_writer, >>>>> -(void *)(uintptr_t)i, >>>>> -enabled_core_ids[i]); >>>>> - >>>>> -/* Wait for writer threads */ >>>>> -for (i = 0; i < 2; i++) >>>>> -if (rte_eal_wait_lcore(enabled_core_ids[i]) < 0) -goto error; >>>>> - >>>>> -printf("Total LPM Adds: %d\n", TOTAL_WRITES); -printf("Total LPM >>>>> Deletes: %d\n", TOTAL_WRITES); -printf("Average LPM Add/Del: >>>>> %"PRIu64" cycles\n", -__atomic_load_n(&gwrite_cycles, >>>>> __ATOMIC_RELAXED) -/ TOTAL_WRITES); >>>>> - >>>>> -writer_done = 1; >>>>> -/* Wait until all readers have exited */ -for (i = 2; i < num_cores; >>>>> i++) -rte_eal_wait_lcore(enabled_core_ids[i]); >>>>> - >>>>> -rte_lpm_free(lpm); >>>>> - >>>>> -return 0; >>>>> - >>>>> -error: >>>>> -writer_done = 1; >>>>> -/* Wait until all readers have exited */ -rte_eal_mp_wait_lcore(); >>>>> - >>>>> -rte_lpm_free(lpm); >>>>> -rte_free(rv); >>>>> - >>>>> -return -1; >>>>> -} >>>>> - >>>>> -/* >>>>> - * Functional test: >>>>> - * Single writer, rest are readers >>>>> - */ >>>>> -static int >>>>> -test_lpm_rcu_perf(void) >>>>> -{ >>>>> -struct rte_lpm_config config; >>>>> -uint64_t begin, total_cycles; >>>>> -size_t sz; >>>>> -unsigned int i, j; >>>>> -uint16_t core_id; >>>>> -uint32_t next_hop_add = 0xAA; >>>>> -struct rte_lpm_rcu_config rcu_cfg = {0}; >>>>> - >>>>> -if (rte_lcore_count() < 2) { >>>>> -printf("Not enough cores for lpm_rcu_perf_autotest, expecting at >>>>> least 2\n"); -return TEST_SKIPPED; -} >>>>> - >>>>> -num_cores = 0; >>>>> -RTE_LCORE_FOREACH_WORKER(core_id) { >>>>> -enabled_core_ids[num_cores] = core_id; -num_cores++; -} >>>>> - >>>>> -printf("\nPerf test: 1 writer, %d readers, RCU integration >>>>> enabled\n", -num_cores); >>>>> - >>>>> -/* Create LPM table */ >>>>> -config.max_rules = NUM_LDEPTH_ROUTE_ENTRIES; - >>> config.number_tbl8s = >>>>> NUM_LDEPTH_ROUTE_ENTRIES; -config.flags = 0; -lpm = >>>>> rte_lpm_create(__func__, SOCKET_ID_ANY, &config); >>>>> -TEST_LPM_ASSERT(lpm != NULL); >>>>> - >>>>> -/* Init RCU variable */ >>>>> -sz = rte_rcu_qsbr_get_memsize(num_cores); >>>>> -rv = (struct rte_rcu_qsbr *)rte_zmalloc("rcu0", sz, >>>>> -RTE_CACHE_LINE_SIZE); -rte_rcu_qsbr_init(rv, num_cores); >>>>> - >>>>> -rcu_cfg.v = rv; >>>>> -/* Assign the RCU variable to LPM */ -if (rte_lpm_rcu_qsbr_add(lpm, >>>>> &rcu_cfg) != 0) { -printf("RCU variable assignment failed\n"); -goto >>>>> error; -} >>>>> - >>>>> -writer_done = 0; >>>>> -__atomic_store_n(&thr_id, 0, __ATOMIC_SEQ_CST); >>>>> - >>>>> -/* Launch reader threads */ >>>>> -for (i = 0; i < num_cores; i++) >>>>> -rte_eal_remote_launch(test_lpm_rcu_qsbr_reader, NULL, >>>>> -enabled_core_ids[i]); >>>>> - >>>>> -/* Measure add/delete. */ >>>>> -begin = rte_rdtsc_precise(); >>>>> -for (i = 0; i < RCU_ITERATIONS; i++) { >>>>> -/* Add all the entries */ >>>>> -for (j = 0; j < NUM_LDEPTH_ROUTE_ENTRIES; j++) -if (rte_lpm_add(lpm, >>>>> large_ldepth_route_table[j].ip, -large_ldepth_route_table[j].depth, >>>>> -next_hop_add) != 0) { >>>>> -printf("Failed to add iteration %d, route# %d\n", -i, j); >>>>> +for (j = 1; j < 3; j++) { >>>>> +if (use_rcu) >>>>> +printf("\nPerf test: %d writer(s), %d reader(s)," >>>>> + " RCU integration enabled\n", j, num_cores - j); else >>>>> +printf("\nPerf test: %d writer(s), %d reader(s)," >>>>> + " RCU integration disabled\n", j, num_cores - j); >>>>> + >>>>> +/* Create LPM table */ >>>>> +config.max_rules = NUM_LDEPTH_ROUTE_ENTRIES; >>> config.number_tbl8s = >>>>> +NUM_LDEPTH_ROUTE_ENTRIES; config.flags = 0; lpm = >>>>> +rte_lpm_create(__func__, SOCKET_ID_ANY, &config); >>>>> +TEST_LPM_ASSERT(lpm != NULL); >>>>> + >>>>> +/* Init RCU variable */ >>>>> +if (use_rcu) { >>>>> +sz = rte_rcu_qsbr_get_memsize(num_cores); >>>>> +rv = (struct rte_rcu_qsbr *)rte_zmalloc("rcu0", sz, >>>>> + >>>>> RTE_CACHE_LINE_SIZE); >>>>> +rte_rcu_qsbr_init(rv, num_cores); >>>>> + >>>>> +rcu_cfg.v = rv; >>>>> +/* Assign the RCU variable to LPM */ if (rte_lpm_rcu_qsbr_add(lpm, >>>>> +&rcu_cfg) != 0) { printf("RCU variable assignment failed\n"); >>>>> goto error; >>>>> } >>>>> >>>>> -/* Delete all the entries */ >>>>> -for (j = 0; j < NUM_LDEPTH_ROUTE_ENTRIES; j++) -if >>>>> (rte_lpm_delete(lpm, large_ldepth_route_table[j].ip, >>>>> -large_ldepth_route_table[j].depth) != 0) { -printf("Failed to delete >>>>> iteration %d, route# %d\n", -i, j); -goto error; -} -} -total_cycles >>>>> = rte_rdtsc_precise() - begin; >>>>> +reader_f = test_lpm_rcu_qsbr_reader; } else reader_f = >>>>> +test_lpm_reader; >>>>> >>>>> -printf("Total LPM Adds: %d\n", TOTAL_WRITES); -printf("Total LPM >>>>> Deletes: %d\n", TOTAL_WRITES); -printf("Average LPM Add/Del: %g >>>>> cycles\n", -(double)total_cycles / TOTAL_WRITES); >>>>> +writer_done = 0; >>>>> +__atomic_store_n(&gwrite_cycles, 0, __ATOMIC_RELAXED); >>>>> >>>>> -writer_done = 1; >>>>> -/* Wait until all readers have exited */ -for (i = 0; i < num_cores; >>>>> i++) -if (rte_eal_wait_lcore(enabled_core_ids[i]); >>>>> - >>>>> -rte_lpm_free(lpm); >>>>> -rte_free(rv); >>>>> -lpm = NULL; >>>>> -rv = NULL; >>>>> - >>>>> -/* Test without RCU integration */ >>>>> -printf("\nPerf test: 1 writer, %d readers, RCU integration >>>>> disabled\n", -num_cores); >>>>> - >>>>> -/* Create LPM table */ >>>>> -config.max_rules = NUM_LDEPTH_ROUTE_ENTRIES; - >>> config.number_tbl8s = >>>>> NUM_LDEPTH_ROUTE_ENTRIES; -config.flags = 0; -lpm = >>>>> rte_lpm_create(__func__, SOCKET_ID_ANY, &config); >>>>> -TEST_LPM_ASSERT(lpm != NULL); >>>>> +__atomic_store_n(&thr_id, 0, __ATOMIC_SEQ_CST); >>>>> >>>>> -writer_done = 0; >>>>> -__atomic_store_n(&thr_id, 0, __ATOMIC_SEQ_CST); >>>>> +/* Launch reader threads */ >>>>> +for (i = j; i < num_cores; i++) >>>>> +rte_eal_remote_launch(reader_f, NULL, >>>>> +enabled_core_ids[i]); >>>>> >>>>> -/* Launch reader threads */ >>>>> -for (i = 0; i < num_cores; i++) >>>>> -rte_eal_remote_launch(test_lpm_reader, NULL, >>>>> -enabled_core_ids[i]); >>>>> +/* Launch writer threads */ >>>>> +for (i = 0; i < j; i++) >>>>> +rte_eal_remote_launch(test_lpm_rcu_qsbr_writer, >>>>> +(void *)(uintptr_t)(i + j), >>>> This can be just 'j'? >>>> >>>>> +enabled_core_ids[i]); >>>>> >>>>> -/* Measure add/delete. */ >>>>> -begin = rte_rdtsc_precise(); >>>>> -for (i = 0; i < RCU_ITERATIONS; i++) { >>>>> -/* Add all the entries */ >>>>> -for (j = 0; j < NUM_LDEPTH_ROUTE_ENTRIES; j++) >>>>> -if (rte_lpm_add(lpm, large_ldepth_route_table[j].ip, >>>>> -large_ldepth_route_table[j].depth, >>>>> -next_hop_add) != 0) { >>>>> -printf("Failed to add iteration %d, route# >>>>> %d\n", >>>>> -i, j); >>>>> +/* Wait for writer threads */ >>>>> +for (i = 0; i < j; i++) >>>>> +if (rte_eal_wait_lcore(enabled_core_ids[i]) < 0) >>>>> goto error; >>>>> -} >>>>> >>>>> -/* Delete all the entries */ >>>>> -for (j = 0; j < NUM_LDEPTH_ROUTE_ENTRIES; j++) >>>>> -if (rte_lpm_delete(lpm, >>>>> large_ldepth_route_table[j].ip, >>>>> -large_ldepth_route_table[j].depth) != 0) { >>>>> -printf("Failed to delete iteration %d, route# >>>>> %d\n", >>>>> -i, j); >>>>> -goto error; >>>>> -} >>>>> +printf("Total LPM Adds: %d\n", TOTAL_WRITES); >>>>> +printf("Total LPM Deletes: %d\n", TOTAL_WRITES); >>>>> +printf("Average LPM Add/Del: %"PRIu64" cycles\n", >>>>> +__atomic_load_n(&gwrite_cycles, >>>>> __ATOMIC_RELAXED) >>>>> +/ TOTAL_WRITES); >>>>> + >>>>> +writer_done = 1; >>>>> +/* Wait until all readers have exited */ >>>>> +for (i = j; i < num_cores; i++) >>>>> +rte_eal_wait_lcore(enabled_core_ids[i]); >>>>> + >>>>> +rte_lpm_free(lpm); >>>>> +rte_free(rv); >>>>> +lpm = NULL; >>>>> +rv = NULL; >>>>> } >>>>> -total_cycles = rte_rdtsc_precise() - begin; >>>>> - >>>>> -printf("Total LPM Adds: %d\n", TOTAL_WRITES); >>>>> -printf("Total LPM Deletes: %d\n", TOTAL_WRITES); >>>>> -printf("Average LPM Add/Del: %g cycles\n", >>>>> -(double)total_cycles / TOTAL_WRITES); >>>>> - >>>>> -writer_done = 1; >>>>> -/* Wait until all readers have exited */ >>>>> -for (i = 0; i < num_cores; i++) >>>>> -rte_eal_wait_lcore(enabled_core_ids[i]); >>>>> - >>>>> -rte_lpm_free(lpm); >>>>> >>>>> return 0; >>>>> >>>>> @@ -946,9 +767,8 @@ test_lpm_perf(void) >>>>> rte_lpm_delete_all(lpm); >>>>> rte_lpm_free(lpm); >>>>> >>>>> -test_lpm_rcu_perf(); >>>>> - >>>>> -test_lpm_rcu_perf_multi_writer(); >>>>> +test_lpm_rcu_perf_multi_writer(0); >>>>> +test_lpm_rcu_perf_multi_writer(1); >>>>> >>>>> return 0; >>>>> } >>>>> -- >>>>> 2.17.1 > -- Regards, Vladimir