korg/linux
0
0
Fork 0
mirror of https://mirrors.bfsu.edu.cn/git/linux.git synced 2024-12-26 12:34:41 +08:00
Code Issues Packages Projects Releases Wiki Activity
ee31fff091
linux/tools/testing/selftests/bpf/bench.c

651 lines
18 KiB
C
Raw Normal View History

selftests/bpf: Add benchmark runner infrastructure While working on BPF ringbuf implementation, testing, and benchmarking, I've developed a pretty generic and modular benchmark runner, which seems to be generically useful, as I've already used it for one more purpose (testing fastest way to trigger BPF program, to minimize overhead of in-kernel code). This patch adds generic part of benchmark runner and sets up Makefile for extending it with more sets of benchmarks. Benchmarker itself operates by spinning up specified number of producer and consumer threads, setting up interval timer sending SIGALARM signal to application once a second. Every second, current snapshot with hits/drops counters are collected and stored in an array. Drops are useful for producer/consumer benchmarks in which producer might overwhelm consumers. Once test finishes after given amount of warm-up and testing seconds, mean and stddev are calculated (ignoring warm-up results) and is printed out to stdout. This setup seems to give consistent and accurate results. To validate behavior, I added two atomic counting tests: global and local. For global one, all the producer threads are atomically incrementing same counter as fast as possible. This, of course, leads to huge drop of performance once there is more than one producer thread due to CPUs fighting for the same memory location. Local counting, on the other hand, maintains one counter per each producer thread, incremented independently. Once per second, all counters are read and added together to form final "counting throughput" measurement. As expected, such setup demonstrates linear scalability with number of producers (as long as there are enough physical CPU cores, of course). See example output below. Also, this setup can nicely demonstrate disastrous effects of false sharing, if care is not taken to take those per-producer counters apart into independent cache lines. Demo output shows global counter first with 1 producer, then with 4. Both total and per-producer performance significantly drop. The last run is local counter with 4 producers, demonstrating near-perfect scalability. $ ./bench -a -w1 -d2 -p1 count-global Setting up benchmark 'count-global'... Benchmark 'count-global' started. Iter 0 ( 24.822us): hits 148.179M/s (148.179M/prod), drops 0.000M/s Iter 1 ( 37.939us): hits 149.308M/s (149.308M/prod), drops 0.000M/s Iter 2 (-10.774us): hits 150.717M/s (150.717M/prod), drops 0.000M/s Iter 3 ( 3.807us): hits 151.435M/s (151.435M/prod), drops 0.000M/s Summary: hits 150.488 ± 1.079M/s (150.488M/prod), drops 0.000 ± 0.000M/s $ ./bench -a -w1 -d2 -p4 count-global Setting up benchmark 'count-global'... Benchmark 'count-global' started. Iter 0 ( 60.659us): hits 53.910M/s ( 13.477M/prod), drops 0.000M/s Iter 1 (-17.658us): hits 53.722M/s ( 13.431M/prod), drops 0.000M/s Iter 2 ( 5.865us): hits 53.495M/s ( 13.374M/prod), drops 0.000M/s Iter 3 ( 0.104us): hits 53.606M/s ( 13.402M/prod), drops 0.000M/s Summary: hits 53.608 ± 0.113M/s ( 13.402M/prod), drops 0.000 ± 0.000M/s $ ./bench -a -w1 -d2 -p4 count-local Setting up benchmark 'count-local'... Benchmark 'count-local' started. Iter 0 ( 23.388us): hits 640.450M/s (160.113M/prod), drops 0.000M/s Iter 1 ( 2.291us): hits 605.661M/s (151.415M/prod), drops 0.000M/s Iter 2 ( -6.415us): hits 607.092M/s (151.773M/prod), drops 0.000M/s Iter 3 ( -1.361us): hits 601.796M/s (150.449M/prod), drops 0.000M/s Summary: hits 604.849 ± 2.739M/s (151.212M/prod), drops 0.000 ± 0.000M/s Benchmark runner supports setting thread affinity for producer and consumer threads. You can use -a flag for default CPU selection scheme, where first consumer gets CPU #0, next one gets CPU #1, and so on. Then producer threads pick up next CPU and increment one-by-one as well. But user can also specify a set of CPUs independently for producers and consumers with --prod-affinity 1,2-10,15 and --cons-affinity <set-of-cpus>. The latter allows to force producers and consumers to share same set of CPUs, if necessary. Signed-off-by: Andrii Nakryiko <andriin@fb.com> Signed-off-by: Alexei Starovoitov <ast@kernel.org> Acked-by: Yonghong Song <yhs@fb.com> Link: https://lore.kernel.org/bpf/20200512192445.2351848-3-andriin@fb.com
2020-05-13 03:24:43 +08:00
// SPDX-License-Identifier: GPL-2.0
/* Copyright (c) 2020 Facebook */
#define _GNU_SOURCE
#include <argp.h>
#include <linux/compiler.h>
#include <sys/time.h>
#include <sched.h>
#include <fcntl.h>
#include <pthread.h>
#include <sys/sysinfo.h>
#include <signal.h>
#include "bench.h"
#include "testing_helpers.h"
struct env env = {
.warmup_sec = 1,
.duration_sec = 5,
.affinity = false,
.consumer_cnt = 1,
.producer_cnt = 1,
};
static int libbpf_print_fn(enum libbpf_print_level level,
const char *format, va_list args)
{
if (level == LIBBPF_DEBUG && !env.verbose)
return 0;
return vfprintf(stderr, format, args);
}
selftests/bpf: Fix checkpatch error on empty function parameter Fix checkpatch error: "ERROR: Bad function definition - void foo() should probably be void foo(void)". Most replacements are done by the following command: sed -i 's#\([a-z]\)()$#\1(void)#g' testing/selftests/bpf/benchs/*.c Signed-off-by: Hou Tao <houtao1@huawei.com> Signed-off-by: Alexei Starovoitov <ast@kernel.org> Link: https://lore.kernel.org/bpf/20211210141652.877186-3-houtao1@huawei.com
2021-12-10 22:16:50 +08:00
void setup_libbpf(void)
selftests/bpf: Add benchmark runner infrastructure While working on BPF ringbuf implementation, testing, and benchmarking, I've developed a pretty generic and modular benchmark runner, which seems to be generically useful, as I've already used it for one more purpose (testing fastest way to trigger BPF program, to minimize overhead of in-kernel code). This patch adds generic part of benchmark runner and sets up Makefile for extending it with more sets of benchmarks. Benchmarker itself operates by spinning up specified number of producer and consumer threads, setting up interval timer sending SIGALARM signal to application once a second. Every second, current snapshot with hits/drops counters are collected and stored in an array. Drops are useful for producer/consumer benchmarks in which producer might overwhelm consumers. Once test finishes after given amount of warm-up and testing seconds, mean and stddev are calculated (ignoring warm-up results) and is printed out to stdout. This setup seems to give consistent and accurate results. To validate behavior, I added two atomic counting tests: global and local. For global one, all the producer threads are atomically incrementing same counter as fast as possible. This, of course, leads to huge drop of performance once there is more than one producer thread due to CPUs fighting for the same memory location. Local counting, on the other hand, maintains one counter per each producer thread, incremented independently. Once per second, all counters are read and added together to form final "counting throughput" measurement. As expected, such setup demonstrates linear scalability with number of producers (as long as there are enough physical CPU cores, of course). See example output below. Also, this setup can nicely demonstrate disastrous effects of false sharing, if care is not taken to take those per-producer counters apart into independent cache lines. Demo output shows global counter first with 1 producer, then with 4. Both total and per-producer performance significantly drop. The last run is local counter with 4 producers, demonstrating near-perfect scalability. $ ./bench -a -w1 -d2 -p1 count-global Setting up benchmark 'count-global'... Benchmark 'count-global' started. Iter 0 ( 24.822us): hits 148.179M/s (148.179M/prod), drops 0.000M/s Iter 1 ( 37.939us): hits 149.308M/s (149.308M/prod), drops 0.000M/s Iter 2 (-10.774us): hits 150.717M/s (150.717M/prod), drops 0.000M/s Iter 3 ( 3.807us): hits 151.435M/s (151.435M/prod), drops 0.000M/s Summary: hits 150.488 ± 1.079M/s (150.488M/prod), drops 0.000 ± 0.000M/s $ ./bench -a -w1 -d2 -p4 count-global Setting up benchmark 'count-global'... Benchmark 'count-global' started. Iter 0 ( 60.659us): hits 53.910M/s ( 13.477M/prod), drops 0.000M/s Iter 1 (-17.658us): hits 53.722M/s ( 13.431M/prod), drops 0.000M/s Iter 2 ( 5.865us): hits 53.495M/s ( 13.374M/prod), drops 0.000M/s Iter 3 ( 0.104us): hits 53.606M/s ( 13.402M/prod), drops 0.000M/s Summary: hits 53.608 ± 0.113M/s ( 13.402M/prod), drops 0.000 ± 0.000M/s $ ./bench -a -w1 -d2 -p4 count-local Setting up benchmark 'count-local'... Benchmark 'count-local' started. Iter 0 ( 23.388us): hits 640.450M/s (160.113M/prod), drops 0.000M/s Iter 1 ( 2.291us): hits 605.661M/s (151.415M/prod), drops 0.000M/s Iter 2 ( -6.415us): hits 607.092M/s (151.773M/prod), drops 0.000M/s Iter 3 ( -1.361us): hits 601.796M/s (150.449M/prod), drops 0.000M/s Summary: hits 604.849 ± 2.739M/s (151.212M/prod), drops 0.000 ± 0.000M/s Benchmark runner supports setting thread affinity for producer and consumer threads. You can use -a flag for default CPU selection scheme, where first consumer gets CPU #0, next one gets CPU #1, and so on. Then producer threads pick up next CPU and increment one-by-one as well. But user can also specify a set of CPUs independently for producers and consumers with --prod-affinity 1,2-10,15 and --cons-affinity <set-of-cpus>. The latter allows to force producers and consumers to share same set of CPUs, if necessary. Signed-off-by: Andrii Nakryiko <andriin@fb.com> Signed-off-by: Alexei Starovoitov <ast@kernel.org> Acked-by: Yonghong Song <yhs@fb.com> Link: https://lore.kernel.org/bpf/20200512192445.2351848-3-andriin@fb.com
2020-05-13 03:24:43 +08:00
{
selftests/bpf: Turn on libbpf 1.0 mode and fix all IS_ERR checks Turn ony libbpf 1.0 mode. Fix all the explicit IS_ERR checks that now will be broken because libbpf returns NULL on error (and sets errno). Fix ASSERT_OK_PTR and ASSERT_ERR_PTR to work for both old mode and new modes and use them throughout selftests. This is trivial to do by using libbpf_get_error() API that all libbpf users are supposed to use, instead of IS_ERR checks. A bunch of checks also did explicit -1 comparison for various fd-returning APIs. Such checks are replaced with >= 0 or < 0 cases. There were also few misuses of bpf_object__find_map_by_name() in test_maps. Those are fixed in this patch as well. Signed-off-by: Andrii Nakryiko <andrii@kernel.org> Signed-off-by: Alexei Starovoitov <ast@kernel.org> Acked-by: John Fastabend <john.fastabend@gmail.com> Acked-by: Toke Høiland-Jørgensen <toke@redhat.com> Link: https://lore.kernel.org/bpf/20210525035935.1461796-3-andrii@kernel.org
2021-05-25 11:59:32 +08:00
libbpf_set_strict_mode(LIBBPF_STRICT_ALL);
selftests/bpf: Add benchmark runner infrastructure While working on BPF ringbuf implementation, testing, and benchmarking, I've developed a pretty generic and modular benchmark runner, which seems to be generically useful, as I've already used it for one more purpose (testing fastest way to trigger BPF program, to minimize overhead of in-kernel code). This patch adds generic part of benchmark runner and sets up Makefile for extending it with more sets of benchmarks. Benchmarker itself operates by spinning up specified number of producer and consumer threads, setting up interval timer sending SIGALARM signal to application once a second. Every second, current snapshot with hits/drops counters are collected and stored in an array. Drops are useful for producer/consumer benchmarks in which producer might overwhelm consumers. Once test finishes after given amount of warm-up and testing seconds, mean and stddev are calculated (ignoring warm-up results) and is printed out to stdout. This setup seems to give consistent and accurate results. To validate behavior, I added two atomic counting tests: global and local. For global one, all the producer threads are atomically incrementing same counter as fast as possible. This, of course, leads to huge drop of performance once there is more than one producer thread due to CPUs fighting for the same memory location. Local counting, on the other hand, maintains one counter per each producer thread, incremented independently. Once per second, all counters are read and added together to form final "counting throughput" measurement. As expected, such setup demonstrates linear scalability with number of producers (as long as there are enough physical CPU cores, of course). See example output below. Also, this setup can nicely demonstrate disastrous effects of false sharing, if care is not taken to take those per-producer counters apart into independent cache lines. Demo output shows global counter first with 1 producer, then with 4. Both total and per-producer performance significantly drop. The last run is local counter with 4 producers, demonstrating near-perfect scalability. $ ./bench -a -w1 -d2 -p1 count-global Setting up benchmark 'count-global'... Benchmark 'count-global' started. Iter 0 ( 24.822us): hits 148.179M/s (148.179M/prod), drops 0.000M/s Iter 1 ( 37.939us): hits 149.308M/s (149.308M/prod), drops 0.000M/s Iter 2 (-10.774us): hits 150.717M/s (150.717M/prod), drops 0.000M/s Iter 3 ( 3.807us): hits 151.435M/s (151.435M/prod), drops 0.000M/s Summary: hits 150.488 ± 1.079M/s (150.488M/prod), drops 0.000 ± 0.000M/s $ ./bench -a -w1 -d2 -p4 count-global Setting up benchmark 'count-global'... Benchmark 'count-global' started. Iter 0 ( 60.659us): hits 53.910M/s ( 13.477M/prod), drops 0.000M/s Iter 1 (-17.658us): hits 53.722M/s ( 13.431M/prod), drops 0.000M/s Iter 2 ( 5.865us): hits 53.495M/s ( 13.374M/prod), drops 0.000M/s Iter 3 ( 0.104us): hits 53.606M/s ( 13.402M/prod), drops 0.000M/s Summary: hits 53.608 ± 0.113M/s ( 13.402M/prod), drops 0.000 ± 0.000M/s $ ./bench -a -w1 -d2 -p4 count-local Setting up benchmark 'count-local'... Benchmark 'count-local' started. Iter 0 ( 23.388us): hits 640.450M/s (160.113M/prod), drops 0.000M/s Iter 1 ( 2.291us): hits 605.661M/s (151.415M/prod), drops 0.000M/s Iter 2 ( -6.415us): hits 607.092M/s (151.773M/prod), drops 0.000M/s Iter 3 ( -1.361us): hits 601.796M/s (150.449M/prod), drops 0.000M/s Summary: hits 604.849 ± 2.739M/s (151.212M/prod), drops 0.000 ± 0.000M/s Benchmark runner supports setting thread affinity for producer and consumer threads. You can use -a flag for default CPU selection scheme, where first consumer gets CPU #0, next one gets CPU #1, and so on. Then producer threads pick up next CPU and increment one-by-one as well. But user can also specify a set of CPUs independently for producers and consumers with --prod-affinity 1,2-10,15 and --cons-affinity <set-of-cpus>. The latter allows to force producers and consumers to share same set of CPUs, if necessary. Signed-off-by: Andrii Nakryiko <andriin@fb.com> Signed-off-by: Alexei Starovoitov <ast@kernel.org> Acked-by: Yonghong Song <yhs@fb.com> Link: https://lore.kernel.org/bpf/20200512192445.2351848-3-andriin@fb.com
2020-05-13 03:24:43 +08:00
libbpf_set_print(libbpf_print_fn);
}
bpf/benchs: Add benchmark tests for bloom filter throughput + false positive This patch adds benchmark tests for the throughput (for lookups + updates) and the false positive rate of bloom filter lookups, as well as some minor refactoring of the bash script for running the benchmarks. These benchmarks show that as the number of hash functions increases, the throughput and the false positive rate of the bloom filter decreases. >From the benchmark data, the approximate average false-positive rates are roughly as follows: 1 hash function = ~30% 2 hash functions = ~15% 3 hash functions = ~5% 4 hash functions = ~2.5% 5 hash functions = ~1% 6 hash functions = ~0.5% 7 hash functions = ~0.35% 8 hash functions = ~0.15% 9 hash functions = ~0.1% 10 hash functions = ~0% For reference data, the benchmarks run on one thread on a machine with one numa node for 1 to 5 hash functions for 8-byte and 64-byte values are as follows: 1 hash function: 50k entries 8-byte value Lookups - 51.1 M/s operations Updates - 33.6 M/s operations False positive rate: 24.15% 64-byte value Lookups - 15.7 M/s operations Updates - 15.1 M/s operations False positive rate: 24.2% 100k entries 8-byte value Lookups - 51.0 M/s operations Updates - 33.4 M/s operations False positive rate: 24.04% 64-byte value Lookups - 15.6 M/s operations Updates - 14.6 M/s operations False positive rate: 24.06% 500k entries 8-byte value Lookups - 50.5 M/s operations Updates - 33.1 M/s operations False positive rate: 27.45% 64-byte value Lookups - 15.6 M/s operations Updates - 14.2 M/s operations False positive rate: 27.42% 1 mil entries 8-byte value Lookups - 49.7 M/s operations Updates - 32.9 M/s operations False positive rate: 27.45% 64-byte value Lookups - 15.4 M/s operations Updates - 13.7 M/s operations False positive rate: 27.58% 2.5 mil entries 8-byte value Lookups - 47.2 M/s operations Updates - 31.8 M/s operations False positive rate: 30.94% 64-byte value Lookups - 15.3 M/s operations Updates - 13.2 M/s operations False positive rate: 30.95% 5 mil entries 8-byte value Lookups - 41.1 M/s operations Updates - 28.1 M/s operations False positive rate: 31.01% 64-byte value Lookups - 13.3 M/s operations Updates - 11.4 M/s operations False positive rate: 30.98% 2 hash functions: 50k entries 8-byte value Lookups - 34.1 M/s operations Updates - 20.1 M/s operations False positive rate: 9.13% 64-byte value Lookups - 8.4 M/s operations Updates - 7.9 M/s operations False positive rate: 9.21% 100k entries 8-byte value Lookups - 33.7 M/s operations Updates - 18.9 M/s operations False positive rate: 9.13% 64-byte value Lookups - 8.4 M/s operations Updates - 7.7 M/s operations False positive rate: 9.19% 500k entries 8-byte value Lookups - 32.7 M/s operations Updates - 18.1 M/s operations False positive rate: 12.61% 64-byte value Lookups - 8.4 M/s operations Updates - 7.5 M/s operations False positive rate: 12.61% 1 mil entries 8-byte value Lookups - 30.6 M/s operations Updates - 18.9 M/s operations False positive rate: 12.54% 64-byte value Lookups - 8.0 M/s operations Updates - 7.0 M/s operations False positive rate: 12.52% 2.5 mil entries 8-byte value Lookups - 25.3 M/s operations Updates - 16.7 M/s operations False positive rate: 16.77% 64-byte value Lookups - 7.9 M/s operations Updates - 6.5 M/s operations False positive rate: 16.88% 5 mil entries 8-byte value Lookups - 20.8 M/s operations Updates - 14.7 M/s operations False positive rate: 16.78% 64-byte value Lookups - 7.0 M/s operations Updates - 6.0 M/s operations False positive rate: 16.78% 3 hash functions: 50k entries 8-byte value Lookups - 25.1 M/s operations Updates - 14.6 M/s operations False positive rate: 7.65% 64-byte value Lookups - 5.8 M/s operations Updates - 5.5 M/s operations False positive rate: 7.58% 100k entries 8-byte value Lookups - 24.7 M/s operations Updates - 14.1 M/s operations False positive rate: 7.71% 64-byte value Lookups - 5.8 M/s operations Updates - 5.3 M/s operations False positive rate: 7.62% 500k entries 8-byte value Lookups - 22.9 M/s operations Updates - 13.9 M/s operations False positive rate: 2.62% 64-byte value Lookups - 5.6 M/s operations Updates - 4.8 M/s operations False positive rate: 2.7% 1 mil entries 8-byte value Lookups - 19.8 M/s operations Updates - 12.6 M/s operations False positive rate: 2.60% 64-byte value Lookups - 5.3 M/s operations Updates - 4.4 M/s operations False positive rate: 2.69% 2.5 mil entries 8-byte value Lookups - 16.2 M/s operations Updates - 10.7 M/s operations False positive rate: 4.49% 64-byte value Lookups - 4.9 M/s operations Updates - 4.1 M/s operations False positive rate: 4.41% 5 mil entries 8-byte value Lookups - 18.8 M/s operations Updates - 9.2 M/s operations False positive rate: 4.45% 64-byte value Lookups - 5.2 M/s operations Updates - 3.9 M/s operations False positive rate: 4.54% 4 hash functions: 50k entries 8-byte value Lookups - 19.7 M/s operations Updates - 11.1 M/s operations False positive rate: 1.01% 64-byte value Lookups - 4.4 M/s operations Updates - 4.0 M/s operations False positive rate: 1.00% 100k entries 8-byte value Lookups - 19.5 M/s operations Updates - 10.9 M/s operations False positive rate: 1.00% 64-byte value Lookups - 4.3 M/s operations Updates - 3.9 M/s operations False positive rate: 0.97% 500k entries 8-byte value Lookups - 18.2 M/s operations Updates - 10.6 M/s operations False positive rate: 2.05% 64-byte value Lookups - 4.3 M/s operations Updates - 3.7 M/s operations False positive rate: 2.05% 1 mil entries 8-byte value Lookups - 15.5 M/s operations Updates - 9.6 M/s operations False positive rate: 1.99% 64-byte value Lookups - 4.0 M/s operations Updates - 3.4 M/s operations False positive rate: 1.99% 2.5 mil entries 8-byte value Lookups - 13.8 M/s operations Updates - 7.7 M/s operations False positive rate: 3.91% 64-byte value Lookups - 3.7 M/s operations Updates - 3.6 M/s operations False positive rate: 3.78% 5 mil entries 8-byte value Lookups - 13.0 M/s operations Updates - 6.9 M/s operations False positive rate: 3.93% 64-byte value Lookups - 3.5 M/s operations Updates - 3.7 M/s operations False positive rate: 3.39% 5 hash functions: 50k entries 8-byte value Lookups - 16.4 M/s operations Updates - 9.1 M/s operations False positive rate: 0.78% 64-byte value Lookups - 3.5 M/s operations Updates - 3.2 M/s operations False positive rate: 0.77% 100k entries 8-byte value Lookups - 16.3 M/s operations Updates - 9.0 M/s operations False positive rate: 0.79% 64-byte value Lookups - 3.5 M/s operations Updates - 3.2 M/s operations False positive rate: 0.78% 500k entries 8-byte value Lookups - 15.1 M/s operations Updates - 8.8 M/s operations False positive rate: 1.82% 64-byte value Lookups - 3.4 M/s operations Updates - 3.0 M/s operations False positive rate: 1.78% 1 mil entries 8-byte value Lookups - 13.2 M/s operations Updates - 7.8 M/s operations False positive rate: 1.81% 64-byte value Lookups - 3.2 M/s operations Updates - 2.8 M/s operations False positive rate: 1.80% 2.5 mil entries 8-byte value Lookups - 10.5 M/s operations Updates - 5.9 M/s operations False positive rate: 0.29% 64-byte value Lookups - 3.2 M/s operations Updates - 2.4 M/s operations False positive rate: 0.28% 5 mil entries 8-byte value Lookups - 9.6 M/s operations Updates - 5.7 M/s operations False positive rate: 0.30% 64-byte value Lookups - 3.2 M/s operations Updates - 2.7 M/s operations False positive rate: 0.30% Signed-off-by: Joanne Koong <joannekoong@fb.com> Signed-off-by: Alexei Starovoitov <ast@kernel.org> Acked-by: Andrii Nakryiko <andrii@kernel.org> Link: https://lore.kernel.org/bpf/20211027234504.30744-5-joannekoong@fb.com
2021-10-28 07:45:03 +08:00
void false_hits_report_progress(int iter, struct bench_res *res, long delta_ns)
{
long total = res->false_hits + res->hits + res->drops;
printf("Iter %3d (%7.3lfus): ",
iter, (delta_ns - 1000000000) / 1000.0);
printf("%ld false hits of %ld total operations. Percentage = %2.2f %%\n",
res->false_hits, total, ((float)res->false_hits / total) * 100);
}
void false_hits_report_final(struct bench_res res[], int res_cnt)
{
long total_hits = 0, total_drops = 0, total_false_hits = 0, total_ops = 0;
int i;
for (i = 0; i < res_cnt; i++) {
total_hits += res[i].hits;
total_false_hits += res[i].false_hits;
total_drops += res[i].drops;
}
total_ops = total_hits + total_false_hits + total_drops;
printf("Summary: %ld false hits of %ld total operations. ",
total_false_hits, total_ops);
printf("Percentage = %2.2f %%\n",
((float)total_false_hits / total_ops) * 100);
}
selftests/bpf: Add benchmark runner infrastructure While working on BPF ringbuf implementation, testing, and benchmarking, I've developed a pretty generic and modular benchmark runner, which seems to be generically useful, as I've already used it for one more purpose (testing fastest way to trigger BPF program, to minimize overhead of in-kernel code). This patch adds generic part of benchmark runner and sets up Makefile for extending it with more sets of benchmarks. Benchmarker itself operates by spinning up specified number of producer and consumer threads, setting up interval timer sending SIGALARM signal to application once a second. Every second, current snapshot with hits/drops counters are collected and stored in an array. Drops are useful for producer/consumer benchmarks in which producer might overwhelm consumers. Once test finishes after given amount of warm-up and testing seconds, mean and stddev are calculated (ignoring warm-up results) and is printed out to stdout. This setup seems to give consistent and accurate results. To validate behavior, I added two atomic counting tests: global and local. For global one, all the producer threads are atomically incrementing same counter as fast as possible. This, of course, leads to huge drop of performance once there is more than one producer thread due to CPUs fighting for the same memory location. Local counting, on the other hand, maintains one counter per each producer thread, incremented independently. Once per second, all counters are read and added together to form final "counting throughput" measurement. As expected, such setup demonstrates linear scalability with number of producers (as long as there are enough physical CPU cores, of course). See example output below. Also, this setup can nicely demonstrate disastrous effects of false sharing, if care is not taken to take those per-producer counters apart into independent cache lines. Demo output shows global counter first with 1 producer, then with 4. Both total and per-producer performance significantly drop. The last run is local counter with 4 producers, demonstrating near-perfect scalability. $ ./bench -a -w1 -d2 -p1 count-global Setting up benchmark 'count-global'... Benchmark 'count-global' started. Iter 0 ( 24.822us): hits 148.179M/s (148.179M/prod), drops 0.000M/s Iter 1 ( 37.939us): hits 149.308M/s (149.308M/prod), drops 0.000M/s Iter 2 (-10.774us): hits 150.717M/s (150.717M/prod), drops 0.000M/s Iter 3 ( 3.807us): hits 151.435M/s (151.435M/prod), drops 0.000M/s Summary: hits 150.488 ± 1.079M/s (150.488M/prod), drops 0.000 ± 0.000M/s $ ./bench -a -w1 -d2 -p4 count-global Setting up benchmark 'count-global'... Benchmark 'count-global' started. Iter 0 ( 60.659us): hits 53.910M/s ( 13.477M/prod), drops 0.000M/s Iter 1 (-17.658us): hits 53.722M/s ( 13.431M/prod), drops 0.000M/s Iter 2 ( 5.865us): hits 53.495M/s ( 13.374M/prod), drops 0.000M/s Iter 3 ( 0.104us): hits 53.606M/s ( 13.402M/prod), drops 0.000M/s Summary: hits 53.608 ± 0.113M/s ( 13.402M/prod), drops 0.000 ± 0.000M/s $ ./bench -a -w1 -d2 -p4 count-local Setting up benchmark 'count-local'... Benchmark 'count-local' started. Iter 0 ( 23.388us): hits 640.450M/s (160.113M/prod), drops 0.000M/s Iter 1 ( 2.291us): hits 605.661M/s (151.415M/prod), drops 0.000M/s Iter 2 ( -6.415us): hits 607.092M/s (151.773M/prod), drops 0.000M/s Iter 3 ( -1.361us): hits 601.796M/s (150.449M/prod), drops 0.000M/s Summary: hits 604.849 ± 2.739M/s (151.212M/prod), drops 0.000 ± 0.000M/s Benchmark runner supports setting thread affinity for producer and consumer threads. You can use -a flag for default CPU selection scheme, where first consumer gets CPU #0, next one gets CPU #1, and so on. Then producer threads pick up next CPU and increment one-by-one as well. But user can also specify a set of CPUs independently for producers and consumers with --prod-affinity 1,2-10,15 and --cons-affinity <set-of-cpus>. The latter allows to force producers and consumers to share same set of CPUs, if necessary. Signed-off-by: Andrii Nakryiko <andriin@fb.com> Signed-off-by: Alexei Starovoitov <ast@kernel.org> Acked-by: Yonghong Song <yhs@fb.com> Link: https://lore.kernel.org/bpf/20200512192445.2351848-3-andriin@fb.com
2020-05-13 03:24:43 +08:00
void hits_drops_report_progress(int iter, struct bench_res *res, long delta_ns)
{
double hits_per_sec, drops_per_sec;
double hits_per_prod;
hits_per_sec = res->hits / 1000000.0 / (delta_ns / 1000000000.0);
hits_per_prod = hits_per_sec / env.producer_cnt;
drops_per_sec = res->drops / 1000000.0 / (delta_ns / 1000000000.0);
printf("Iter %3d (%7.3lfus): ",
iter, (delta_ns - 1000000000) / 1000.0);
bpf/benchs: Add benchmarks for comparing hashmap lookups w/ vs. w/out bloom filter This patch adds benchmark tests for comparing the performance of hashmap lookups without the bloom filter vs. hashmap lookups with the bloom filter. Checking the bloom filter first for whether the element exists should overall enable a higher throughput for hashmap lookups, since if the element does not exist in the bloom filter, we can avoid a costly lookup in the hashmap. On average, using 5 hash functions in the bloom filter tended to perform the best across the widest range of different entry sizes. The benchmark results using 5 hash functions (running on 8 threads on a machine with one numa node, and taking the average of 3 runs) were roughly as follows: value_size = 4 bytes - 10k entries: 30% faster 50k entries: 40% faster 100k entries: 40% faster 500k entres: 70% faster 1 million entries: 90% faster 5 million entries: 140% faster value_size = 8 bytes - 10k entries: 30% faster 50k entries: 40% faster 100k entries: 50% faster 500k entres: 80% faster 1 million entries: 100% faster 5 million entries: 150% faster value_size = 16 bytes - 10k entries: 20% faster 50k entries: 30% faster 100k entries: 35% faster 500k entres: 65% faster 1 million entries: 85% faster 5 million entries: 110% faster value_size = 40 bytes - 10k entries: 5% faster 50k entries: 15% faster 100k entries: 20% faster 500k entres: 65% faster 1 million entries: 75% faster 5 million entries: 120% faster Signed-off-by: Joanne Koong <joannekoong@fb.com> Signed-off-by: Alexei Starovoitov <ast@kernel.org> Link: https://lore.kernel.org/bpf/20211027234504.30744-6-joannekoong@fb.com
2021-10-28 07:45:04 +08:00
printf("hits %8.3lfM/s (%7.3lfM/prod), drops %8.3lfM/s, total operations %8.3lfM/s\n",
hits_per_sec, hits_per_prod, drops_per_sec, hits_per_sec + drops_per_sec);
selftests/bpf: Add benchmark runner infrastructure While working on BPF ringbuf implementation, testing, and benchmarking, I've developed a pretty generic and modular benchmark runner, which seems to be generically useful, as I've already used it for one more purpose (testing fastest way to trigger BPF program, to minimize overhead of in-kernel code). This patch adds generic part of benchmark runner and sets up Makefile for extending it with more sets of benchmarks. Benchmarker itself operates by spinning up specified number of producer and consumer threads, setting up interval timer sending SIGALARM signal to application once a second. Every second, current snapshot with hits/drops counters are collected and stored in an array. Drops are useful for producer/consumer benchmarks in which producer might overwhelm consumers. Once test finishes after given amount of warm-up and testing seconds, mean and stddev are calculated (ignoring warm-up results) and is printed out to stdout. This setup seems to give consistent and accurate results. To validate behavior, I added two atomic counting tests: global and local. For global one, all the producer threads are atomically incrementing same counter as fast as possible. This, of course, leads to huge drop of performance once there is more than one producer thread due to CPUs fighting for the same memory location. Local counting, on the other hand, maintains one counter per each producer thread, incremented independently. Once per second, all counters are read and added together to form final "counting throughput" measurement. As expected, such setup demonstrates linear scalability with number of producers (as long as there are enough physical CPU cores, of course). See example output below. Also, this setup can nicely demonstrate disastrous effects of false sharing, if care is not taken to take those per-producer counters apart into independent cache lines. Demo output shows global counter first with 1 producer, then with 4. Both total and per-producer performance significantly drop. The last run is local counter with 4 producers, demonstrating near-perfect scalability. $ ./bench -a -w1 -d2 -p1 count-global Setting up benchmark 'count-global'... Benchmark 'count-global' started. Iter 0 ( 24.822us): hits 148.179M/s (148.179M/prod), drops 0.000M/s Iter 1 ( 37.939us): hits 149.308M/s (149.308M/prod), drops 0.000M/s Iter 2 (-10.774us): hits 150.717M/s (150.717M/prod), drops 0.000M/s Iter 3 ( 3.807us): hits 151.435M/s (151.435M/prod), drops 0.000M/s Summary: hits 150.488 ± 1.079M/s (150.488M/prod), drops 0.000 ± 0.000M/s $ ./bench -a -w1 -d2 -p4 count-global Setting up benchmark 'count-global'... Benchmark 'count-global' started. Iter 0 ( 60.659us): hits 53.910M/s ( 13.477M/prod), drops 0.000M/s Iter 1 (-17.658us): hits 53.722M/s ( 13.431M/prod), drops 0.000M/s Iter 2 ( 5.865us): hits 53.495M/s ( 13.374M/prod), drops 0.000M/s Iter 3 ( 0.104us): hits 53.606M/s ( 13.402M/prod), drops 0.000M/s Summary: hits 53.608 ± 0.113M/s ( 13.402M/prod), drops 0.000 ± 0.000M/s $ ./bench -a -w1 -d2 -p4 count-local Setting up benchmark 'count-local'... Benchmark 'count-local' started. Iter 0 ( 23.388us): hits 640.450M/s (160.113M/prod), drops 0.000M/s Iter 1 ( 2.291us): hits 605.661M/s (151.415M/prod), drops 0.000M/s Iter 2 ( -6.415us): hits 607.092M/s (151.773M/prod), drops 0.000M/s Iter 3 ( -1.361us): hits 601.796M/s (150.449M/prod), drops 0.000M/s Summary: hits 604.849 ± 2.739M/s (151.212M/prod), drops 0.000 ± 0.000M/s Benchmark runner supports setting thread affinity for producer and consumer threads. You can use -a flag for default CPU selection scheme, where first consumer gets CPU #0, next one gets CPU #1, and so on. Then producer threads pick up next CPU and increment one-by-one as well. But user can also specify a set of CPUs independently for producers and consumers with --prod-affinity 1,2-10,15 and --cons-affinity <set-of-cpus>. The latter allows to force producers and consumers to share same set of CPUs, if necessary. Signed-off-by: Andrii Nakryiko <andriin@fb.com> Signed-off-by: Alexei Starovoitov <ast@kernel.org> Acked-by: Yonghong Song <yhs@fb.com> Link: https://lore.kernel.org/bpf/20200512192445.2351848-3-andriin@fb.com
2020-05-13 03:24:43 +08:00
}
selftests/bpf: Add benchmark for local_storage RCU Tasks Trace usage This benchmark measures grace period latency and kthread cpu usage of RCU Tasks Trace when many processes are creating/deleting BPF local_storage. Intent here is to quantify improvement on these metrics after Paul's recent RCU Tasks patches [0]. Specifically, fork 15k tasks which call a bpf prog that creates/destroys task local_storage and sleep in a loop, resulting in many call_rcu_tasks_trace calls. To determine grace period latency, trace time elapsed between rcu_tasks_trace_pregp_step and rcu_tasks_trace_postgp; for cpu usage look at rcu_task_trace_kthread's stime in /proc/PID/stat. On my virtualized test environment (Skylake, 8 cpus) benchmark results demonstrate significant improvement: BEFORE Paul's patches: SUMMARY tasks_trace grace period latency avg 22298.551 us stddev 1302.165 us SUMMARY ticks per tasks_trace grace period avg 2.291 stddev 0.324 AFTER Paul's patches: SUMMARY tasks_trace grace period latency avg 16969.197 us stddev 2525.053 us SUMMARY ticks per tasks_trace grace period avg 1.146 stddev 0.178 Note that since these patches are not in bpf-next benchmarking was done by cherry-picking this patch onto rcu tree. [0] https://lore.kernel.org/rcu/20220620225402.GA3842369@paulmck-ThinkPad-P17-Gen-1/ Signed-off-by: Dave Marchevsky <davemarchevsky@fb.com> Signed-off-by: Daniel Borkmann <daniel@iogearbox.net> Acked-by: Paul E. McKenney <paulmck@kernel.org> Acked-by: Martin KaFai Lau <kafai@fb.com> Link: https://lore.kernel.org/bpf/20220705190018.3239050-1-davemarchevsky@fb.com
2022-07-06 03:00:18 +08:00
void
grace_period_latency_basic_stats(struct bench_res res[], int res_cnt, struct basic_stats *gp_stat)
{
int i;
memset(gp_stat, 0, sizeof(struct basic_stats));
for (i = 0; i < res_cnt; i++)
gp_stat->mean += res[i].gp_ns / 1000.0 / (double)res[i].gp_ct / (0.0 + res_cnt);
#define IT_MEAN_DIFF (res[i].gp_ns / 1000.0 / (double)res[i].gp_ct - gp_stat->mean)
if (res_cnt > 1) {
for (i = 0; i < res_cnt; i++)
gp_stat->stddev += (IT_MEAN_DIFF * IT_MEAN_DIFF) / (res_cnt - 1.0);
}
gp_stat->stddev = sqrt(gp_stat->stddev);
#undef IT_MEAN_DIFF
}
void
grace_period_ticks_basic_stats(struct bench_res res[], int res_cnt, struct basic_stats *gp_stat)
{
int i;
memset(gp_stat, 0, sizeof(struct basic_stats));
for (i = 0; i < res_cnt; i++)
gp_stat->mean += res[i].stime / (double)res[i].gp_ct / (0.0 + res_cnt);
#define IT_MEAN_DIFF (res[i].stime / (double)res[i].gp_ct - gp_stat->mean)
if (res_cnt > 1) {
for (i = 0; i < res_cnt; i++)
gp_stat->stddev += (IT_MEAN_DIFF * IT_MEAN_DIFF) / (res_cnt - 1.0);
}
gp_stat->stddev = sqrt(gp_stat->stddev);
#undef IT_MEAN_DIFF
}
selftests/bpf: Add benchmark runner infrastructure While working on BPF ringbuf implementation, testing, and benchmarking, I've developed a pretty generic and modular benchmark runner, which seems to be generically useful, as I've already used it for one more purpose (testing fastest way to trigger BPF program, to minimize overhead of in-kernel code). This patch adds generic part of benchmark runner and sets up Makefile for extending it with more sets of benchmarks. Benchmarker itself operates by spinning up specified number of producer and consumer threads, setting up interval timer sending SIGALARM signal to application once a second. Every second, current snapshot with hits/drops counters are collected and stored in an array. Drops are useful for producer/consumer benchmarks in which producer might overwhelm consumers. Once test finishes after given amount of warm-up and testing seconds, mean and stddev are calculated (ignoring warm-up results) and is printed out to stdout. This setup seems to give consistent and accurate results. To validate behavior, I added two atomic counting tests: global and local. For global one, all the producer threads are atomically incrementing same counter as fast as possible. This, of course, leads to huge drop of performance once there is more than one producer thread due to CPUs fighting for the same memory location. Local counting, on the other hand, maintains one counter per each producer thread, incremented independently. Once per second, all counters are read and added together to form final "counting throughput" measurement. As expected, such setup demonstrates linear scalability with number of producers (as long as there are enough physical CPU cores, of course). See example output below. Also, this setup can nicely demonstrate disastrous effects of false sharing, if care is not taken to take those per-producer counters apart into independent cache lines. Demo output shows global counter first with 1 producer, then with 4. Both total and per-producer performance significantly drop. The last run is local counter with 4 producers, demonstrating near-perfect scalability. $ ./bench -a -w1 -d2 -p1 count-global Setting up benchmark 'count-global'... Benchmark 'count-global' started. Iter 0 ( 24.822us): hits 148.179M/s (148.179M/prod), drops 0.000M/s Iter 1 ( 37.939us): hits 149.308M/s (149.308M/prod), drops 0.000M/s Iter 2 (-10.774us): hits 150.717M/s (150.717M/prod), drops 0.000M/s Iter 3 ( 3.807us): hits 151.435M/s (151.435M/prod), drops 0.000M/s Summary: hits 150.488 ± 1.079M/s (150.488M/prod), drops 0.000 ± 0.000M/s $ ./bench -a -w1 -d2 -p4 count-global Setting up benchmark 'count-global'... Benchmark 'count-global' started. Iter 0 ( 60.659us): hits 53.910M/s ( 13.477M/prod), drops 0.000M/s Iter 1 (-17.658us): hits 53.722M/s ( 13.431M/prod), drops 0.000M/s Iter 2 ( 5.865us): hits 53.495M/s ( 13.374M/prod), drops 0.000M/s Iter 3 ( 0.104us): hits 53.606M/s ( 13.402M/prod), drops 0.000M/s Summary: hits 53.608 ± 0.113M/s ( 13.402M/prod), drops 0.000 ± 0.000M/s $ ./bench -a -w1 -d2 -p4 count-local Setting up benchmark 'count-local'... Benchmark 'count-local' started. Iter 0 ( 23.388us): hits 640.450M/s (160.113M/prod), drops 0.000M/s Iter 1 ( 2.291us): hits 605.661M/s (151.415M/prod), drops 0.000M/s Iter 2 ( -6.415us): hits 607.092M/s (151.773M/prod), drops 0.000M/s Iter 3 ( -1.361us): hits 601.796M/s (150.449M/prod), drops 0.000M/s Summary: hits 604.849 ± 2.739M/s (151.212M/prod), drops 0.000 ± 0.000M/s Benchmark runner supports setting thread affinity for producer and consumer threads. You can use -a flag for default CPU selection scheme, where first consumer gets CPU #0, next one gets CPU #1, and so on. Then producer threads pick up next CPU and increment one-by-one as well. But user can also specify a set of CPUs independently for producers and consumers with --prod-affinity 1,2-10,15 and --cons-affinity <set-of-cpus>. The latter allows to force producers and consumers to share same set of CPUs, if necessary. Signed-off-by: Andrii Nakryiko <andriin@fb.com> Signed-off-by: Alexei Starovoitov <ast@kernel.org> Acked-by: Yonghong Song <yhs@fb.com> Link: https://lore.kernel.org/bpf/20200512192445.2351848-3-andriin@fb.com
2020-05-13 03:24:43 +08:00
void hits_drops_report_final(struct bench_res res[], int res_cnt)
{
int i;
bpf/benchs: Add benchmarks for comparing hashmap lookups w/ vs. w/out bloom filter This patch adds benchmark tests for comparing the performance of hashmap lookups without the bloom filter vs. hashmap lookups with the bloom filter. Checking the bloom filter first for whether the element exists should overall enable a higher throughput for hashmap lookups, since if the element does not exist in the bloom filter, we can avoid a costly lookup in the hashmap. On average, using 5 hash functions in the bloom filter tended to perform the best across the widest range of different entry sizes. The benchmark results using 5 hash functions (running on 8 threads on a machine with one numa node, and taking the average of 3 runs) were roughly as follows: value_size = 4 bytes - 10k entries: 30% faster 50k entries: 40% faster 100k entries: 40% faster 500k entres: 70% faster 1 million entries: 90% faster 5 million entries: 140% faster value_size = 8 bytes - 10k entries: 30% faster 50k entries: 40% faster 100k entries: 50% faster 500k entres: 80% faster 1 million entries: 100% faster 5 million entries: 150% faster value_size = 16 bytes - 10k entries: 20% faster 50k entries: 30% faster 100k entries: 35% faster 500k entres: 65% faster 1 million entries: 85% faster 5 million entries: 110% faster value_size = 40 bytes - 10k entries: 5% faster 50k entries: 15% faster 100k entries: 20% faster 500k entres: 65% faster 1 million entries: 75% faster 5 million entries: 120% faster Signed-off-by: Joanne Koong <joannekoong@fb.com> Signed-off-by: Alexei Starovoitov <ast@kernel.org> Link: https://lore.kernel.org/bpf/20211027234504.30744-6-joannekoong@fb.com
2021-10-28 07:45:04 +08:00
double hits_mean = 0.0, drops_mean = 0.0, total_ops_mean = 0.0;
double hits_stddev = 0.0, drops_stddev = 0.0, total_ops_stddev = 0.0;
double total_ops;
selftests/bpf: Add benchmark runner infrastructure While working on BPF ringbuf implementation, testing, and benchmarking, I've developed a pretty generic and modular benchmark runner, which seems to be generically useful, as I've already used it for one more purpose (testing fastest way to trigger BPF program, to minimize overhead of in-kernel code). This patch adds generic part of benchmark runner and sets up Makefile for extending it with more sets of benchmarks. Benchmarker itself operates by spinning up specified number of producer and consumer threads, setting up interval timer sending SIGALARM signal to application once a second. Every second, current snapshot with hits/drops counters are collected and stored in an array. Drops are useful for producer/consumer benchmarks in which producer might overwhelm consumers. Once test finishes after given amount of warm-up and testing seconds, mean and stddev are calculated (ignoring warm-up results) and is printed out to stdout. This setup seems to give consistent and accurate results. To validate behavior, I added two atomic counting tests: global and local. For global one, all the producer threads are atomically incrementing same counter as fast as possible. This, of course, leads to huge drop of performance once there is more than one producer thread due to CPUs fighting for the same memory location. Local counting, on the other hand, maintains one counter per each producer thread, incremented independently. Once per second, all counters are read and added together to form final "counting throughput" measurement. As expected, such setup demonstrates linear scalability with number of producers (as long as there are enough physical CPU cores, of course). See example output below. Also, this setup can nicely demonstrate disastrous effects of false sharing, if care is not taken to take those per-producer counters apart into independent cache lines. Demo output shows global counter first with 1 producer, then with 4. Both total and per-producer performance significantly drop. The last run is local counter with 4 producers, demonstrating near-perfect scalability. $ ./bench -a -w1 -d2 -p1 count-global Setting up benchmark 'count-global'... Benchmark 'count-global' started. Iter 0 ( 24.822us): hits 148.179M/s (148.179M/prod), drops 0.000M/s Iter 1 ( 37.939us): hits 149.308M/s (149.308M/prod), drops 0.000M/s Iter 2 (-10.774us): hits 150.717M/s (150.717M/prod), drops 0.000M/s Iter 3 ( 3.807us): hits 151.435M/s (151.435M/prod), drops 0.000M/s Summary: hits 150.488 ± 1.079M/s (150.488M/prod), drops 0.000 ± 0.000M/s $ ./bench -a -w1 -d2 -p4 count-global Setting up benchmark 'count-global'... Benchmark 'count-global' started. Iter 0 ( 60.659us): hits 53.910M/s ( 13.477M/prod), drops 0.000M/s Iter 1 (-17.658us): hits 53.722M/s ( 13.431M/prod), drops 0.000M/s Iter 2 ( 5.865us): hits 53.495M/s ( 13.374M/prod), drops 0.000M/s Iter 3 ( 0.104us): hits 53.606M/s ( 13.402M/prod), drops 0.000M/s Summary: hits 53.608 ± 0.113M/s ( 13.402M/prod), drops 0.000 ± 0.000M/s $ ./bench -a -w1 -d2 -p4 count-local Setting up benchmark 'count-local'... Benchmark 'count-local' started. Iter 0 ( 23.388us): hits 640.450M/s (160.113M/prod), drops 0.000M/s Iter 1 ( 2.291us): hits 605.661M/s (151.415M/prod), drops 0.000M/s Iter 2 ( -6.415us): hits 607.092M/s (151.773M/prod), drops 0.000M/s Iter 3 ( -1.361us): hits 601.796M/s (150.449M/prod), drops 0.000M/s Summary: hits 604.849 ± 2.739M/s (151.212M/prod), drops 0.000 ± 0.000M/s Benchmark runner supports setting thread affinity for producer and consumer threads. You can use -a flag for default CPU selection scheme, where first consumer gets CPU #0, next one gets CPU #1, and so on. Then producer threads pick up next CPU and increment one-by-one as well. But user can also specify a set of CPUs independently for producers and consumers with --prod-affinity 1,2-10,15 and --cons-affinity <set-of-cpus>. The latter allows to force producers and consumers to share same set of CPUs, if necessary. Signed-off-by: Andrii Nakryiko <andriin@fb.com> Signed-off-by: Alexei Starovoitov <ast@kernel.org> Acked-by: Yonghong Song <yhs@fb.com> Link: https://lore.kernel.org/bpf/20200512192445.2351848-3-andriin@fb.com
2020-05-13 03:24:43 +08:00
for (i = 0; i < res_cnt; i++) {
hits_mean += res[i].hits / 1000000.0 / (0.0 + res_cnt);
drops_mean += res[i].drops / 1000000.0 / (0.0 + res_cnt);
}
bpf/benchs: Add benchmarks for comparing hashmap lookups w/ vs. w/out bloom filter This patch adds benchmark tests for comparing the performance of hashmap lookups without the bloom filter vs. hashmap lookups with the bloom filter. Checking the bloom filter first for whether the element exists should overall enable a higher throughput for hashmap lookups, since if the element does not exist in the bloom filter, we can avoid a costly lookup in the hashmap. On average, using 5 hash functions in the bloom filter tended to perform the best across the widest range of different entry sizes. The benchmark results using 5 hash functions (running on 8 threads on a machine with one numa node, and taking the average of 3 runs) were roughly as follows: value_size = 4 bytes - 10k entries: 30% faster 50k entries: 40% faster 100k entries: 40% faster 500k entres: 70% faster 1 million entries: 90% faster 5 million entries: 140% faster value_size = 8 bytes - 10k entries: 30% faster 50k entries: 40% faster 100k entries: 50% faster 500k entres: 80% faster 1 million entries: 100% faster 5 million entries: 150% faster value_size = 16 bytes - 10k entries: 20% faster 50k entries: 30% faster 100k entries: 35% faster 500k entres: 65% faster 1 million entries: 85% faster 5 million entries: 110% faster value_size = 40 bytes - 10k entries: 5% faster 50k entries: 15% faster 100k entries: 20% faster 500k entres: 65% faster 1 million entries: 75% faster 5 million entries: 120% faster Signed-off-by: Joanne Koong <joannekoong@fb.com> Signed-off-by: Alexei Starovoitov <ast@kernel.org> Link: https://lore.kernel.org/bpf/20211027234504.30744-6-joannekoong@fb.com
2021-10-28 07:45:04 +08:00
total_ops_mean = hits_mean + drops_mean;
selftests/bpf: Add benchmark runner infrastructure While working on BPF ringbuf implementation, testing, and benchmarking, I've developed a pretty generic and modular benchmark runner, which seems to be generically useful, as I've already used it for one more purpose (testing fastest way to trigger BPF program, to minimize overhead of in-kernel code). This patch adds generic part of benchmark runner and sets up Makefile for extending it with more sets of benchmarks. Benchmarker itself operates by spinning up specified number of producer and consumer threads, setting up interval timer sending SIGALARM signal to application once a second. Every second, current snapshot with hits/drops counters are collected and stored in an array. Drops are useful for producer/consumer benchmarks in which producer might overwhelm consumers. Once test finishes after given amount of warm-up and testing seconds, mean and stddev are calculated (ignoring warm-up results) and is printed out to stdout. This setup seems to give consistent and accurate results. To validate behavior, I added two atomic counting tests: global and local. For global one, all the producer threads are atomically incrementing same counter as fast as possible. This, of course, leads to huge drop of performance once there is more than one producer thread due to CPUs fighting for the same memory location. Local counting, on the other hand, maintains one counter per each producer thread, incremented independently. Once per second, all counters are read and added together to form final "counting throughput" measurement. As expected, such setup demonstrates linear scalability with number of producers (as long as there are enough physical CPU cores, of course). See example output below. Also, this setup can nicely demonstrate disastrous effects of false sharing, if care is not taken to take those per-producer counters apart into independent cache lines. Demo output shows global counter first with 1 producer, then with 4. Both total and per-producer performance significantly drop. The last run is local counter with 4 producers, demonstrating near-perfect scalability. $ ./bench -a -w1 -d2 -p1 count-global Setting up benchmark 'count-global'... Benchmark 'count-global' started. Iter 0 ( 24.822us): hits 148.179M/s (148.179M/prod), drops 0.000M/s Iter 1 ( 37.939us): hits 149.308M/s (149.308M/prod), drops 0.000M/s Iter 2 (-10.774us): hits 150.717M/s (150.717M/prod), drops 0.000M/s Iter 3 ( 3.807us): hits 151.435M/s (151.435M/prod), drops 0.000M/s Summary: hits 150.488 ± 1.079M/s (150.488M/prod), drops 0.000 ± 0.000M/s $ ./bench -a -w1 -d2 -p4 count-global Setting up benchmark 'count-global'... Benchmark 'count-global' started. Iter 0 ( 60.659us): hits 53.910M/s ( 13.477M/prod), drops 0.000M/s Iter 1 (-17.658us): hits 53.722M/s ( 13.431M/prod), drops 0.000M/s Iter 2 ( 5.865us): hits 53.495M/s ( 13.374M/prod), drops 0.000M/s Iter 3 ( 0.104us): hits 53.606M/s ( 13.402M/prod), drops 0.000M/s Summary: hits 53.608 ± 0.113M/s ( 13.402M/prod), drops 0.000 ± 0.000M/s $ ./bench -a -w1 -d2 -p4 count-local Setting up benchmark 'count-local'... Benchmark 'count-local' started. Iter 0 ( 23.388us): hits 640.450M/s (160.113M/prod), drops 0.000M/s Iter 1 ( 2.291us): hits 605.661M/s (151.415M/prod), drops 0.000M/s Iter 2 ( -6.415us): hits 607.092M/s (151.773M/prod), drops 0.000M/s Iter 3 ( -1.361us): hits 601.796M/s (150.449M/prod), drops 0.000M/s Summary: hits 604.849 ± 2.739M/s (151.212M/prod), drops 0.000 ± 0.000M/s Benchmark runner supports setting thread affinity for producer and consumer threads. You can use -a flag for default CPU selection scheme, where first consumer gets CPU #0, next one gets CPU #1, and so on. Then producer threads pick up next CPU and increment one-by-one as well. But user can also specify a set of CPUs independently for producers and consumers with --prod-affinity 1,2-10,15 and --cons-affinity <set-of-cpus>. The latter allows to force producers and consumers to share same set of CPUs, if necessary. Signed-off-by: Andrii Nakryiko <andriin@fb.com> Signed-off-by: Alexei Starovoitov <ast@kernel.org> Acked-by: Yonghong Song <yhs@fb.com> Link: https://lore.kernel.org/bpf/20200512192445.2351848-3-andriin@fb.com
2020-05-13 03:24:43 +08:00
if (res_cnt > 1) {
for (i = 0; i < res_cnt; i++) {
hits_stddev += (hits_mean - res[i].hits / 1000000.0) *
(hits_mean - res[i].hits / 1000000.0) /
(res_cnt - 1.0);
drops_stddev += (drops_mean - res[i].drops / 1000000.0) *
(drops_mean - res[i].drops / 1000000.0) /
(res_cnt - 1.0);
bpf/benchs: Add benchmarks for comparing hashmap lookups w/ vs. w/out bloom filter This patch adds benchmark tests for comparing the performance of hashmap lookups without the bloom filter vs. hashmap lookups with the bloom filter. Checking the bloom filter first for whether the element exists should overall enable a higher throughput for hashmap lookups, since if the element does not exist in the bloom filter, we can avoid a costly lookup in the hashmap. On average, using 5 hash functions in the bloom filter tended to perform the best across the widest range of different entry sizes. The benchmark results using 5 hash functions (running on 8 threads on a machine with one numa node, and taking the average of 3 runs) were roughly as follows: value_size = 4 bytes - 10k entries: 30% faster 50k entries: 40% faster 100k entries: 40% faster 500k entres: 70% faster 1 million entries: 90% faster 5 million entries: 140% faster value_size = 8 bytes - 10k entries: 30% faster 50k entries: 40% faster 100k entries: 50% faster 500k entres: 80% faster 1 million entries: 100% faster 5 million entries: 150% faster value_size = 16 bytes - 10k entries: 20% faster 50k entries: 30% faster 100k entries: 35% faster 500k entres: 65% faster 1 million entries: 85% faster 5 million entries: 110% faster value_size = 40 bytes - 10k entries: 5% faster 50k entries: 15% faster 100k entries: 20% faster 500k entres: 65% faster 1 million entries: 75% faster 5 million entries: 120% faster Signed-off-by: Joanne Koong <joannekoong@fb.com> Signed-off-by: Alexei Starovoitov <ast@kernel.org> Link: https://lore.kernel.org/bpf/20211027234504.30744-6-joannekoong@fb.com
2021-10-28 07:45:04 +08:00
total_ops = res[i].hits + res[i].drops;
total_ops_stddev += (total_ops_mean - total_ops / 1000000.0) *
(total_ops_mean - total_ops / 1000000.0) /
(res_cnt - 1.0);
selftests/bpf: Add benchmark runner infrastructure While working on BPF ringbuf implementation, testing, and benchmarking, I've developed a pretty generic and modular benchmark runner, which seems to be generically useful, as I've already used it for one more purpose (testing fastest way to trigger BPF program, to minimize overhead of in-kernel code). This patch adds generic part of benchmark runner and sets up Makefile for extending it with more sets of benchmarks. Benchmarker itself operates by spinning up specified number of producer and consumer threads, setting up interval timer sending SIGALARM signal to application once a second. Every second, current snapshot with hits/drops counters are collected and stored in an array. Drops are useful for producer/consumer benchmarks in which producer might overwhelm consumers. Once test finishes after given amount of warm-up and testing seconds, mean and stddev are calculated (ignoring warm-up results) and is printed out to stdout. This setup seems to give consistent and accurate results. To validate behavior, I added two atomic counting tests: global and local. For global one, all the producer threads are atomically incrementing same counter as fast as possible. This, of course, leads to huge drop of performance once there is more than one producer thread due to CPUs fighting for the same memory location. Local counting, on the other hand, maintains one counter per each producer thread, incremented independently. Once per second, all counters are read and added together to form final "counting throughput" measurement. As expected, such setup demonstrates linear scalability with number of producers (as long as there are enough physical CPU cores, of course). See example output below. Also, this setup can nicely demonstrate disastrous effects of false sharing, if care is not taken to take those per-producer counters apart into independent cache lines. Demo output shows global counter first with 1 producer, then with 4. Both total and per-producer performance significantly drop. The last run is local counter with 4 producers, demonstrating near-perfect scalability. $ ./bench -a -w1 -d2 -p1 count-global Setting up benchmark 'count-global'... Benchmark 'count-global' started. Iter 0 ( 24.822us): hits 148.179M/s (148.179M/prod), drops 0.000M/s Iter 1 ( 37.939us): hits 149.308M/s (149.308M/prod), drops 0.000M/s Iter 2 (-10.774us): hits 150.717M/s (150.717M/prod), drops 0.000M/s Iter 3 ( 3.807us): hits 151.435M/s (151.435M/prod), drops 0.000M/s Summary: hits 150.488 ± 1.079M/s (150.488M/prod), drops 0.000 ± 0.000M/s $ ./bench -a -w1 -d2 -p4 count-global Setting up benchmark 'count-global'... Benchmark 'count-global' started. Iter 0 ( 60.659us): hits 53.910M/s ( 13.477M/prod), drops 0.000M/s Iter 1 (-17.658us): hits 53.722M/s ( 13.431M/prod), drops 0.000M/s Iter 2 ( 5.865us): hits 53.495M/s ( 13.374M/prod), drops 0.000M/s Iter 3 ( 0.104us): hits 53.606M/s ( 13.402M/prod), drops 0.000M/s Summary: hits 53.608 ± 0.113M/s ( 13.402M/prod), drops 0.000 ± 0.000M/s $ ./bench -a -w1 -d2 -p4 count-local Setting up benchmark 'count-local'... Benchmark 'count-local' started. Iter 0 ( 23.388us): hits 640.450M/s (160.113M/prod), drops 0.000M/s Iter 1 ( 2.291us): hits 605.661M/s (151.415M/prod), drops 0.000M/s Iter 2 ( -6.415us): hits 607.092M/s (151.773M/prod), drops 0.000M/s Iter 3 ( -1.361us): hits 601.796M/s (150.449M/prod), drops 0.000M/s Summary: hits 604.849 ± 2.739M/s (151.212M/prod), drops 0.000 ± 0.000M/s Benchmark runner supports setting thread affinity for producer and consumer threads. You can use -a flag for default CPU selection scheme, where first consumer gets CPU #0, next one gets CPU #1, and so on. Then producer threads pick up next CPU and increment one-by-one as well. But user can also specify a set of CPUs independently for producers and consumers with --prod-affinity 1,2-10,15 and --cons-affinity <set-of-cpus>. The latter allows to force producers and consumers to share same set of CPUs, if necessary. Signed-off-by: Andrii Nakryiko <andriin@fb.com> Signed-off-by: Alexei Starovoitov <ast@kernel.org> Acked-by: Yonghong Song <yhs@fb.com> Link: https://lore.kernel.org/bpf/20200512192445.2351848-3-andriin@fb.com
2020-05-13 03:24:43 +08:00
}
hits_stddev = sqrt(hits_stddev);
drops_stddev = sqrt(drops_stddev);
bpf/benchs: Add benchmarks for comparing hashmap lookups w/ vs. w/out bloom filter This patch adds benchmark tests for comparing the performance of hashmap lookups without the bloom filter vs. hashmap lookups with the bloom filter. Checking the bloom filter first for whether the element exists should overall enable a higher throughput for hashmap lookups, since if the element does not exist in the bloom filter, we can avoid a costly lookup in the hashmap. On average, using 5 hash functions in the bloom filter tended to perform the best across the widest range of different entry sizes. The benchmark results using 5 hash functions (running on 8 threads on a machine with one numa node, and taking the average of 3 runs) were roughly as follows: value_size = 4 bytes - 10k entries: 30% faster 50k entries: 40% faster 100k entries: 40% faster 500k entres: 70% faster 1 million entries: 90% faster 5 million entries: 140% faster value_size = 8 bytes - 10k entries: 30% faster 50k entries: 40% faster 100k entries: 50% faster 500k entres: 80% faster 1 million entries: 100% faster 5 million entries: 150% faster value_size = 16 bytes - 10k entries: 20% faster 50k entries: 30% faster 100k entries: 35% faster 500k entres: 65% faster 1 million entries: 85% faster 5 million entries: 110% faster value_size = 40 bytes - 10k entries: 5% faster 50k entries: 15% faster 100k entries: 20% faster 500k entres: 65% faster 1 million entries: 75% faster 5 million entries: 120% faster Signed-off-by: Joanne Koong <joannekoong@fb.com> Signed-off-by: Alexei Starovoitov <ast@kernel.org> Link: https://lore.kernel.org/bpf/20211027234504.30744-6-joannekoong@fb.com
2021-10-28 07:45:04 +08:00
total_ops_stddev = sqrt(total_ops_stddev);
selftests/bpf: Add benchmark runner infrastructure While working on BPF ringbuf implementation, testing, and benchmarking, I've developed a pretty generic and modular benchmark runner, which seems to be generically useful, as I've already used it for one more purpose (testing fastest way to trigger BPF program, to minimize overhead of in-kernel code). This patch adds generic part of benchmark runner and sets up Makefile for extending it with more sets of benchmarks. Benchmarker itself operates by spinning up specified number of producer and consumer threads, setting up interval timer sending SIGALARM signal to application once a second. Every second, current snapshot with hits/drops counters are collected and stored in an array. Drops are useful for producer/consumer benchmarks in which producer might overwhelm consumers. Once test finishes after given amount of warm-up and testing seconds, mean and stddev are calculated (ignoring warm-up results) and is printed out to stdout. This setup seems to give consistent and accurate results. To validate behavior, I added two atomic counting tests: global and local. For global one, all the producer threads are atomically incrementing same counter as fast as possible. This, of course, leads to huge drop of performance once there is more than one producer thread due to CPUs fighting for the same memory location. Local counting, on the other hand, maintains one counter per each producer thread, incremented independently. Once per second, all counters are read and added together to form final "counting throughput" measurement. As expected, such setup demonstrates linear scalability with number of producers (as long as there are enough physical CPU cores, of course). See example output below. Also, this setup can nicely demonstrate disastrous effects of false sharing, if care is not taken to take those per-producer counters apart into independent cache lines. Demo output shows global counter first with 1 producer, then with 4. Both total and per-producer performance significantly drop. The last run is local counter with 4 producers, demonstrating near-perfect scalability. $ ./bench -a -w1 -d2 -p1 count-global Setting up benchmark 'count-global'... Benchmark 'count-global' started. Iter 0 ( 24.822us): hits 148.179M/s (148.179M/prod), drops 0.000M/s Iter 1 ( 37.939us): hits 149.308M/s (149.308M/prod), drops 0.000M/s Iter 2 (-10.774us): hits 150.717M/s (150.717M/prod), drops 0.000M/s Iter 3 ( 3.807us): hits 151.435M/s (151.435M/prod), drops 0.000M/s Summary: hits 150.488 ± 1.079M/s (150.488M/prod), drops 0.000 ± 0.000M/s $ ./bench -a -w1 -d2 -p4 count-global Setting up benchmark 'count-global'... Benchmark 'count-global' started. Iter 0 ( 60.659us): hits 53.910M/s ( 13.477M/prod), drops 0.000M/s Iter 1 (-17.658us): hits 53.722M/s ( 13.431M/prod), drops 0.000M/s Iter 2 ( 5.865us): hits 53.495M/s ( 13.374M/prod), drops 0.000M/s Iter 3 ( 0.104us): hits 53.606M/s ( 13.402M/prod), drops 0.000M/s Summary: hits 53.608 ± 0.113M/s ( 13.402M/prod), drops 0.000 ± 0.000M/s $ ./bench -a -w1 -d2 -p4 count-local Setting up benchmark 'count-local'... Benchmark 'count-local' started. Iter 0 ( 23.388us): hits 640.450M/s (160.113M/prod), drops 0.000M/s Iter 1 ( 2.291us): hits 605.661M/s (151.415M/prod), drops 0.000M/s Iter 2 ( -6.415us): hits 607.092M/s (151.773M/prod), drops 0.000M/s Iter 3 ( -1.361us): hits 601.796M/s (150.449M/prod), drops 0.000M/s Summary: hits 604.849 ± 2.739M/s (151.212M/prod), drops 0.000 ± 0.000M/s Benchmark runner supports setting thread affinity for producer and consumer threads. You can use -a flag for default CPU selection scheme, where first consumer gets CPU #0, next one gets CPU #1, and so on. Then producer threads pick up next CPU and increment one-by-one as well. But user can also specify a set of CPUs independently for producers and consumers with --prod-affinity 1,2-10,15 and --cons-affinity <set-of-cpus>. The latter allows to force producers and consumers to share same set of CPUs, if necessary. Signed-off-by: Andrii Nakryiko <andriin@fb.com> Signed-off-by: Alexei Starovoitov <ast@kernel.org> Acked-by: Yonghong Song <yhs@fb.com> Link: https://lore.kernel.org/bpf/20200512192445.2351848-3-andriin@fb.com
2020-05-13 03:24:43 +08:00
}
printf("Summary: hits %8.3lf \u00B1 %5.3lfM/s (%7.3lfM/prod), ",
hits_mean, hits_stddev, hits_mean / env.producer_cnt);
bpf/benchs: Add benchmarks for comparing hashmap lookups w/ vs. w/out bloom filter This patch adds benchmark tests for comparing the performance of hashmap lookups without the bloom filter vs. hashmap lookups with the bloom filter. Checking the bloom filter first for whether the element exists should overall enable a higher throughput for hashmap lookups, since if the element does not exist in the bloom filter, we can avoid a costly lookup in the hashmap. On average, using 5 hash functions in the bloom filter tended to perform the best across the widest range of different entry sizes. The benchmark results using 5 hash functions (running on 8 threads on a machine with one numa node, and taking the average of 3 runs) were roughly as follows: value_size = 4 bytes - 10k entries: 30% faster 50k entries: 40% faster 100k entries: 40% faster 500k entres: 70% faster 1 million entries: 90% faster 5 million entries: 140% faster value_size = 8 bytes - 10k entries: 30% faster 50k entries: 40% faster 100k entries: 50% faster 500k entres: 80% faster 1 million entries: 100% faster 5 million entries: 150% faster value_size = 16 bytes - 10k entries: 20% faster 50k entries: 30% faster 100k entries: 35% faster 500k entres: 65% faster 1 million entries: 85% faster 5 million entries: 110% faster value_size = 40 bytes - 10k entries: 5% faster 50k entries: 15% faster 100k entries: 20% faster 500k entres: 65% faster 1 million entries: 75% faster 5 million entries: 120% faster Signed-off-by: Joanne Koong <joannekoong@fb.com> Signed-off-by: Alexei Starovoitov <ast@kernel.org> Link: https://lore.kernel.org/bpf/20211027234504.30744-6-joannekoong@fb.com
2021-10-28 07:45:04 +08:00
printf("drops %8.3lf \u00B1 %5.3lfM/s, ",
selftests/bpf: Add benchmark runner infrastructure While working on BPF ringbuf implementation, testing, and benchmarking, I've developed a pretty generic and modular benchmark runner, which seems to be generically useful, as I've already used it for one more purpose (testing fastest way to trigger BPF program, to minimize overhead of in-kernel code). This patch adds generic part of benchmark runner and sets up Makefile for extending it with more sets of benchmarks. Benchmarker itself operates by spinning up specified number of producer and consumer threads, setting up interval timer sending SIGALARM signal to application once a second. Every second, current snapshot with hits/drops counters are collected and stored in an array. Drops are useful for producer/consumer benchmarks in which producer might overwhelm consumers. Once test finishes after given amount of warm-up and testing seconds, mean and stddev are calculated (ignoring warm-up results) and is printed out to stdout. This setup seems to give consistent and accurate results. To validate behavior, I added two atomic counting tests: global and local. For global one, all the producer threads are atomically incrementing same counter as fast as possible. This, of course, leads to huge drop of performance once there is more than one producer thread due to CPUs fighting for the same memory location. Local counting, on the other hand, maintains one counter per each producer thread, incremented independently. Once per second, all counters are read and added together to form final "counting throughput" measurement. As expected, such setup demonstrates linear scalability with number of producers (as long as there are enough physical CPU cores, of course). See example output below. Also, this setup can nicely demonstrate disastrous effects of false sharing, if care is not taken to take those per-producer counters apart into independent cache lines. Demo output shows global counter first with 1 producer, then with 4. Both total and per-producer performance significantly drop. The last run is local counter with 4 producers, demonstrating near-perfect scalability. $ ./bench -a -w1 -d2 -p1 count-global Setting up benchmark 'count-global'... Benchmark 'count-global' started. Iter 0 ( 24.822us): hits 148.179M/s (148.179M/prod), drops 0.000M/s Iter 1 ( 37.939us): hits 149.308M/s (149.308M/prod), drops 0.000M/s Iter 2 (-10.774us): hits 150.717M/s (150.717M/prod), drops 0.000M/s Iter 3 ( 3.807us): hits 151.435M/s (151.435M/prod), drops 0.000M/s Summary: hits 150.488 ± 1.079M/s (150.488M/prod), drops 0.000 ± 0.000M/s $ ./bench -a -w1 -d2 -p4 count-global Setting up benchmark 'count-global'... Benchmark 'count-global' started. Iter 0 ( 60.659us): hits 53.910M/s ( 13.477M/prod), drops 0.000M/s Iter 1 (-17.658us): hits 53.722M/s ( 13.431M/prod), drops 0.000M/s Iter 2 ( 5.865us): hits 53.495M/s ( 13.374M/prod), drops 0.000M/s Iter 3 ( 0.104us): hits 53.606M/s ( 13.402M/prod), drops 0.000M/s Summary: hits 53.608 ± 0.113M/s ( 13.402M/prod), drops 0.000 ± 0.000M/s $ ./bench -a -w1 -d2 -p4 count-local Setting up benchmark 'count-local'... Benchmark 'count-local' started. Iter 0 ( 23.388us): hits 640.450M/s (160.113M/prod), drops 0.000M/s Iter 1 ( 2.291us): hits 605.661M/s (151.415M/prod), drops 0.000M/s Iter 2 ( -6.415us): hits 607.092M/s (151.773M/prod), drops 0.000M/s Iter 3 ( -1.361us): hits 601.796M/s (150.449M/prod), drops 0.000M/s Summary: hits 604.849 ± 2.739M/s (151.212M/prod), drops 0.000 ± 0.000M/s Benchmark runner supports setting thread affinity for producer and consumer threads. You can use -a flag for default CPU selection scheme, where first consumer gets CPU #0, next one gets CPU #1, and so on. Then producer threads pick up next CPU and increment one-by-one as well. But user can also specify a set of CPUs independently for producers and consumers with --prod-affinity 1,2-10,15 and --cons-affinity <set-of-cpus>. The latter allows to force producers and consumers to share same set of CPUs, if necessary. Signed-off-by: Andrii Nakryiko <andriin@fb.com> Signed-off-by: Alexei Starovoitov <ast@kernel.org> Acked-by: Yonghong Song <yhs@fb.com> Link: https://lore.kernel.org/bpf/20200512192445.2351848-3-andriin@fb.com
2020-05-13 03:24:43 +08:00
drops_mean, drops_stddev);
bpf/benchs: Add benchmarks for comparing hashmap lookups w/ vs. w/out bloom filter This patch adds benchmark tests for comparing the performance of hashmap lookups without the bloom filter vs. hashmap lookups with the bloom filter. Checking the bloom filter first for whether the element exists should overall enable a higher throughput for hashmap lookups, since if the element does not exist in the bloom filter, we can avoid a costly lookup in the hashmap. On average, using 5 hash functions in the bloom filter tended to perform the best across the widest range of different entry sizes. The benchmark results using 5 hash functions (running on 8 threads on a machine with one numa node, and taking the average of 3 runs) were roughly as follows: value_size = 4 bytes - 10k entries: 30% faster 50k entries: 40% faster 100k entries: 40% faster 500k entres: 70% faster 1 million entries: 90% faster 5 million entries: 140% faster value_size = 8 bytes - 10k entries: 30% faster 50k entries: 40% faster 100k entries: 50% faster 500k entres: 80% faster 1 million entries: 100% faster 5 million entries: 150% faster value_size = 16 bytes - 10k entries: 20% faster 50k entries: 30% faster 100k entries: 35% faster 500k entres: 65% faster 1 million entries: 85% faster 5 million entries: 110% faster value_size = 40 bytes - 10k entries: 5% faster 50k entries: 15% faster 100k entries: 20% faster 500k entres: 65% faster 1 million entries: 75% faster 5 million entries: 120% faster Signed-off-by: Joanne Koong <joannekoong@fb.com> Signed-off-by: Alexei Starovoitov <ast@kernel.org> Link: https://lore.kernel.org/bpf/20211027234504.30744-6-joannekoong@fb.com
2021-10-28 07:45:04 +08:00
printf("total operations %8.3lf \u00B1 %5.3lfM/s\n",
total_ops_mean, total_ops_stddev);
selftests/bpf: Add benchmark runner infrastructure While working on BPF ringbuf implementation, testing, and benchmarking, I've developed a pretty generic and modular benchmark runner, which seems to be generically useful, as I've already used it for one more purpose (testing fastest way to trigger BPF program, to minimize overhead of in-kernel code). This patch adds generic part of benchmark runner and sets up Makefile for extending it with more sets of benchmarks. Benchmarker itself operates by spinning up specified number of producer and consumer threads, setting up interval timer sending SIGALARM signal to application once a second. Every second, current snapshot with hits/drops counters are collected and stored in an array. Drops are useful for producer/consumer benchmarks in which producer might overwhelm consumers. Once test finishes after given amount of warm-up and testing seconds, mean and stddev are calculated (ignoring warm-up results) and is printed out to stdout. This setup seems to give consistent and accurate results. To validate behavior, I added two atomic counting tests: global and local. For global one, all the producer threads are atomically incrementing same counter as fast as possible. This, of course, leads to huge drop of performance once there is more than one producer thread due to CPUs fighting for the same memory location. Local counting, on the other hand, maintains one counter per each producer thread, incremented independently. Once per second, all counters are read and added together to form final "counting throughput" measurement. As expected, such setup demonstrates linear scalability with number of producers (as long as there are enough physical CPU cores, of course). See example output below. Also, this setup can nicely demonstrate disastrous effects of false sharing, if care is not taken to take those per-producer counters apart into independent cache lines. Demo output shows global counter first with 1 producer, then with 4. Both total and per-producer performance significantly drop. The last run is local counter with 4 producers, demonstrating near-perfect scalability. $ ./bench -a -w1 -d2 -p1 count-global Setting up benchmark 'count-global'... Benchmark 'count-global' started. Iter 0 ( 24.822us): hits 148.179M/s (148.179M/prod), drops 0.000M/s Iter 1 ( 37.939us): hits 149.308M/s (149.308M/prod), drops 0.000M/s Iter 2 (-10.774us): hits 150.717M/s (150.717M/prod), drops 0.000M/s Iter 3 ( 3.807us): hits 151.435M/s (151.435M/prod), drops 0.000M/s Summary: hits 150.488 ± 1.079M/s (150.488M/prod), drops 0.000 ± 0.000M/s $ ./bench -a -w1 -d2 -p4 count-global Setting up benchmark 'count-global'... Benchmark 'count-global' started. Iter 0 ( 60.659us): hits 53.910M/s ( 13.477M/prod), drops 0.000M/s Iter 1 (-17.658us): hits 53.722M/s ( 13.431M/prod), drops 0.000M/s Iter 2 ( 5.865us): hits 53.495M/s ( 13.374M/prod), drops 0.000M/s Iter 3 ( 0.104us): hits 53.606M/s ( 13.402M/prod), drops 0.000M/s Summary: hits 53.608 ± 0.113M/s ( 13.402M/prod), drops 0.000 ± 0.000M/s $ ./bench -a -w1 -d2 -p4 count-local Setting up benchmark 'count-local'... Benchmark 'count-local' started. Iter 0 ( 23.388us): hits 640.450M/s (160.113M/prod), drops 0.000M/s Iter 1 ( 2.291us): hits 605.661M/s (151.415M/prod), drops 0.000M/s Iter 2 ( -6.415us): hits 607.092M/s (151.773M/prod), drops 0.000M/s Iter 3 ( -1.361us): hits 601.796M/s (150.449M/prod), drops 0.000M/s Summary: hits 604.849 ± 2.739M/s (151.212M/prod), drops 0.000 ± 0.000M/s Benchmark runner supports setting thread affinity for producer and consumer threads. You can use -a flag for default CPU selection scheme, where first consumer gets CPU #0, next one gets CPU #1, and so on. Then producer threads pick up next CPU and increment one-by-one as well. But user can also specify a set of CPUs independently for producers and consumers with --prod-affinity 1,2-10,15 and --cons-affinity <set-of-cpus>. The latter allows to force producers and consumers to share same set of CPUs, if necessary. Signed-off-by: Andrii Nakryiko <andriin@fb.com> Signed-off-by: Alexei Starovoitov <ast@kernel.org> Acked-by: Yonghong Song <yhs@fb.com> Link: https://lore.kernel.org/bpf/20200512192445.2351848-3-andriin@fb.com
2020-05-13 03:24:43 +08:00
}
selftest/bpf/benchs: Add bpf_loop benchmark Add benchmark to measure the throughput and latency of the bpf_loop call. Testing this on my dev machine on 1 thread, the data is as follows: nr_loops: 10 bpf_loop - throughput: 198.519 ± 0.155 M ops/s, latency: 5.037 ns/op nr_loops: 100 bpf_loop - throughput: 247.448 ± 0.305 M ops/s, latency: 4.041 ns/op nr_loops: 500 bpf_loop - throughput: 260.839 ± 0.380 M ops/s, latency: 3.834 ns/op nr_loops: 1000 bpf_loop - throughput: 262.806 ± 0.629 M ops/s, latency: 3.805 ns/op nr_loops: 5000 bpf_loop - throughput: 264.211 ± 1.508 M ops/s, latency: 3.785 ns/op nr_loops: 10000 bpf_loop - throughput: 265.366 ± 3.054 M ops/s, latency: 3.768 ns/op nr_loops: 50000 bpf_loop - throughput: 235.986 ± 20.205 M ops/s, latency: 4.238 ns/op nr_loops: 100000 bpf_loop - throughput: 264.482 ± 0.279 M ops/s, latency: 3.781 ns/op nr_loops: 500000 bpf_loop - throughput: 309.773 ± 87.713 M ops/s, latency: 3.228 ns/op nr_loops: 1000000 bpf_loop - throughput: 262.818 ± 4.143 M ops/s, latency: 3.805 ns/op >From this data, we can see that the latency per loop decreases as the number of loops increases. On this particular machine, each loop had an overhead of about ~4 ns, and we were able to run ~250 million loops per second. Signed-off-by: Joanne Koong <joannekoong@fb.com> Signed-off-by: Alexei Starovoitov <ast@kernel.org> Acked-by: Andrii Nakryiko <andrii@kernel.org> Link: https://lore.kernel.org/bpf/20211130030622.4131246-5-joannekoong@fb.com
2021-11-30 11:06:22 +08:00
void ops_report_progress(int iter, struct bench_res *res, long delta_ns)
{
double hits_per_sec, hits_per_prod;
hits_per_sec = res->hits / 1000000.0 / (delta_ns / 1000000000.0);
hits_per_prod = hits_per_sec / env.producer_cnt;
printf("Iter %3d (%7.3lfus): ", iter, (delta_ns - 1000000000) / 1000.0);
printf("hits %8.3lfM/s (%7.3lfM/prod)\n", hits_per_sec, hits_per_prod);
}
void ops_report_final(struct bench_res res[], int res_cnt)
{
double hits_mean = 0.0, hits_stddev = 0.0;
int i;
for (i = 0; i < res_cnt; i++)
hits_mean += res[i].hits / 1000000.0 / (0.0 + res_cnt);
if (res_cnt > 1) {
for (i = 0; i < res_cnt; i++)
hits_stddev += (hits_mean - res[i].hits / 1000000.0) *
(hits_mean - res[i].hits / 1000000.0) /
(res_cnt - 1.0);
hits_stddev = sqrt(hits_stddev);
}
printf("Summary: throughput %8.3lf \u00B1 %5.3lf M ops/s (%7.3lfM ops/prod), ",
hits_mean, hits_stddev, hits_mean / env.producer_cnt);
printf("latency %8.3lf ns/op\n", 1000.0 / hits_mean * env.producer_cnt);
}
selftests/bpf: Add benchmark for local_storage get Add a benchmarks to demonstrate the performance cliff for local_storage get as the number of local_storage maps increases beyond current local_storage implementation's cache size. "sequential get" and "interleaved get" benchmarks are added, both of which do many bpf_task_storage_get calls on sets of task local_storage maps of various counts, while considering a single specific map to be 'important' and counting task_storage_gets to the important map separately in addition to normal 'hits' count of all gets. Goal here is to mimic scenario where a particular program using one map - the important one - is running on a system where many other local_storage maps exist and are accessed often. While "sequential get" benchmark does bpf_task_storage_get for map 0, 1, ..., {9, 99, 999} in order, "interleaved" benchmark interleaves 4 bpf_task_storage_gets for the important map for every 10 map gets. This is meant to highlight performance differences when important map is accessed far more frequently than non-important maps. A "hashmap control" benchmark is also included for easy comparison of standard bpf hashmap lookup vs local_storage get. The benchmark is similar to "sequential get", but creates and uses BPF_MAP_TYPE_HASH instead of local storage. Only one inner map is created - a hashmap meant to hold tid -> data mapping for all tasks. Size of the hashmap is hardcoded to my system's PID_MAX_LIMIT (4,194,304). The number of these keys which are actually fetched as part of the benchmark is configurable. Addition of this benchmark is inspired by conversation with Alexei in a previous patchset's thread [0], which highlighted the need for such a benchmark to motivate and validate improvements to local_storage implementation. My approach in that series focused on improving performance for explicitly-marked 'important' maps and was rejected with feedback to make more generally-applicable improvements while avoiding explicitly marking maps as important. Thus the benchmark reports both general and important-map-focused metrics, so effect of future work on both is clear. Regarding the benchmark results. On a powerful system (Skylake, 20 cores, 256gb ram): Hashmap Control =============== num keys: 10 hashmap (control) sequential get: hits throughput: 20.900 ± 0.334 M ops/s, hits latency: 47.847 ns/op, important_hits throughput: 20.900 ± 0.334 M ops/s num keys: 1000 hashmap (control) sequential get: hits throughput: 13.758 ± 0.219 M ops/s, hits latency: 72.683 ns/op, important_hits throughput: 13.758 ± 0.219 M ops/s num keys: 10000 hashmap (control) sequential get: hits throughput: 6.995 ± 0.034 M ops/s, hits latency: 142.959 ns/op, important_hits throughput: 6.995 ± 0.034 M ops/s num keys: 100000 hashmap (control) sequential get: hits throughput: 4.452 ± 0.371 M ops/s, hits latency: 224.635 ns/op, important_hits throughput: 4.452 ± 0.371 M ops/s num keys: 4194304 hashmap (control) sequential get: hits throughput: 3.043 ± 0.033 M ops/s, hits latency: 328.587 ns/op, important_hits throughput: 3.043 ± 0.033 M ops/s Local Storage ============= num_maps: 1 local_storage cache sequential get: hits throughput: 47.298 ± 0.180 M ops/s, hits latency: 21.142 ns/op, important_hits throughput: 47.298 ± 0.180 M ops/s local_storage cache interleaved get: hits throughput: 55.277 ± 0.888 M ops/s, hits latency: 18.091 ns/op, important_hits throughput: 55.277 ± 0.888 M ops/s num_maps: 10 local_storage cache sequential get: hits throughput: 40.240 ± 0.802 M ops/s, hits latency: 24.851 ns/op, important_hits throughput: 4.024 ± 0.080 M ops/s local_storage cache interleaved get: hits throughput: 48.701 ± 0.722 M ops/s, hits latency: 20.533 ns/op, important_hits throughput: 17.393 ± 0.258 M ops/s num_maps: 16 local_storage cache sequential get: hits throughput: 44.515 ± 0.708 M ops/s, hits latency: 22.464 ns/op, important_hits throughput: 2.782 ± 0.044 M ops/s local_storage cache interleaved get: hits throughput: 49.553 ± 2.260 M ops/s, hits latency: 20.181 ns/op, important_hits throughput: 15.767 ± 0.719 M ops/s num_maps: 17 local_storage cache sequential get: hits throughput: 38.778 ± 0.302 M ops/s, hits latency: 25.788 ns/op, important_hits throughput: 2.284 ± 0.018 M ops/s local_storage cache interleaved get: hits throughput: 43.848 ± 1.023 M ops/s, hits latency: 22.806 ns/op, important_hits throughput: 13.349 ± 0.311 M ops/s num_maps: 24 local_storage cache sequential get: hits throughput: 19.317 ± 0.568 M ops/s, hits latency: 51.769 ns/op, important_hits throughput: 0.806 ± 0.024 M ops/s local_storage cache interleaved get: hits throughput: 24.397 ± 0.272 M ops/s, hits latency: 40.989 ns/op, important_hits throughput: 6.863 ± 0.077 M ops/s num_maps: 32 local_storage cache sequential get: hits throughput: 13.333 ± 0.135 M ops/s, hits latency: 75.000 ns/op, important_hits throughput: 0.417 ± 0.004 M ops/s local_storage cache interleaved get: hits throughput: 16.898 ± 0.383 M ops/s, hits latency: 59.178 ns/op, important_hits throughput: 4.717 ± 0.107 M ops/s num_maps: 100 local_storage cache sequential get: hits throughput: 6.360 ± 0.107 M ops/s, hits latency: 157.233 ns/op, important_hits throughput: 0.064 ± 0.001 M ops/s local_storage cache interleaved get: hits throughput: 7.303 ± 0.362 M ops/s, hits latency: 136.930 ns/op, important_hits throughput: 1.907 ± 0.094 M ops/s num_maps: 1000 local_storage cache sequential get: hits throughput: 0.452 ± 0.010 M ops/s, hits latency: 2214.022 ns/op, important_hits throughput: 0.000 ± 0.000 M ops/s local_storage cache interleaved get: hits throughput: 0.542 ± 0.007 M ops/s, hits latency: 1843.341 ns/op, important_hits throughput: 0.136 ± 0.002 M ops/s Looking at the "sequential get" results, it's clear that as the number of task local_storage maps grows beyond the current cache size (16), there's a significant reduction in hits throughput. Note that current local_storage implementation assigns a cache_idx to maps as they are created. Since "sequential get" is creating maps 0..n in order and then doing bpf_task_storage_get calls in the same order, the benchmark is effectively ensuring that a map will not be in cache when the program tries to access it. For "interleaved get" results, important-map hits throughput is greatly increased as the important map is more likely to be in cache by virtue of being accessed far more frequently. Throughput still reduces as # maps increases, though. To get a sense of the overhead of the benchmark program, I commented out bpf_task_storage_get/bpf_map_lookup_elem in local_storage_bench.c and ran the benchmark on the same host as the 'real' run. Results: Hashmap Control =============== num keys: 10 hashmap (control) sequential get: hits throughput: 54.288 ± 0.655 M ops/s, hits latency: 18.420 ns/op, important_hits throughput: 54.288 ± 0.655 M ops/s num keys: 1000 hashmap (control) sequential get: hits throughput: 52.913 ± 0.519 M ops/s, hits latency: 18.899 ns/op, important_hits throughput: 52.913 ± 0.519 M ops/s num keys: 10000 hashmap (control) sequential get: hits throughput: 53.480 ± 1.235 M ops/s, hits latency: 18.699 ns/op, important_hits throughput: 53.480 ± 1.235 M ops/s num keys: 100000 hashmap (control) sequential get: hits throughput: 54.982 ± 1.902 M ops/s, hits latency: 18.188 ns/op, important_hits throughput: 54.982 ± 1.902 M ops/s num keys: 4194304 hashmap (control) sequential get: hits throughput: 50.858 ± 0.707 M ops/s, hits latency: 19.662 ns/op, important_hits throughput: 50.858 ± 0.707 M ops/s Local Storage ============= num_maps: 1 local_storage cache sequential get: hits throughput: 110.990 ± 4.828 M ops/s, hits latency: 9.010 ns/op, important_hits throughput: 110.990 ± 4.828 M ops/s local_storage cache interleaved get: hits throughput: 161.057 ± 4.090 M ops/s, hits latency: 6.209 ns/op, important_hits throughput: 161.057 ± 4.090 M ops/s num_maps: 10 local_storage cache sequential get: hits throughput: 112.930 ± 1.079 M ops/s, hits latency: 8.855 ns/op, important_hits throughput: 11.293 ± 0.108 M ops/s local_storage cache interleaved get: hits throughput: 115.841 ± 2.088 M ops/s, hits latency: 8.633 ns/op, important_hits throughput: 41.372 ± 0.746 M ops/s num_maps: 16 local_storage cache sequential get: hits throughput: 115.653 ± 0.416 M ops/s, hits latency: 8.647 ns/op, important_hits throughput: 7.228 ± 0.026 M ops/s local_storage cache interleaved get: hits throughput: 138.717 ± 1.649 M ops/s, hits latency: 7.209 ns/op, important_hits throughput: 44.137 ± 0.525 M ops/s num_maps: 17 local_storage cache sequential get: hits throughput: 112.020 ± 1.649 M ops/s, hits latency: 8.927 ns/op, important_hits throughput: 6.598 ± 0.097 M ops/s local_storage cache interleaved get: hits throughput: 128.089 ± 1.960 M ops/s, hits latency: 7.807 ns/op, important_hits throughput: 38.995 ± 0.597 M ops/s num_maps: 24 local_storage cache sequential get: hits throughput: 92.447 ± 5.170 M ops/s, hits latency: 10.817 ns/op, important_hits throughput: 3.855 ± 0.216 M ops/s local_storage cache interleaved get: hits throughput: 128.844 ± 2.808 M ops/s, hits latency: 7.761 ns/op, important_hits throughput: 36.245 ± 0.790 M ops/s num_maps: 32 local_storage cache sequential get: hits throughput: 102.042 ± 1.462 M ops/s, hits latency: 9.800 ns/op, important_hits throughput: 3.194 ± 0.046 M ops/s local_storage cache interleaved get: hits throughput: 126.577 ± 1.818 M ops/s, hits latency: 7.900 ns/op, important_hits throughput: 35.332 ± 0.507 M ops/s num_maps: 100 local_storage cache sequential get: hits throughput: 111.327 ± 1.401 M ops/s, hits latency: 8.983 ns/op, important_hits throughput: 1.113 ± 0.014 M ops/s local_storage cache interleaved get: hits throughput: 131.327 ± 1.339 M ops/s, hits latency: 7.615 ns/op, important_hits throughput: 34.302 ± 0.350 M ops/s num_maps: 1000 local_storage cache sequential get: hits throughput: 101.978 ± 0.563 M ops/s, hits latency: 9.806 ns/op, important_hits throughput: 0.102 ± 0.001 M ops/s local_storage cache interleaved get: hits throughput: 141.084 ± 1.098 M ops/s, hits latency: 7.088 ns/op, important_hits throughput: 35.430 ± 0.276 M ops/s Adjusting for overhead, latency numbers for "hashmap control" and "sequential get" are: hashmap_control_1k: ~53.8ns hashmap_control_10k: ~124.2ns hashmap_control_100k: ~206.5ns sequential_get_1: ~12.1ns sequential_get_10: ~16.0ns sequential_get_16: ~13.8ns sequential_get_17: ~16.8ns sequential_get_24: ~40.9ns sequential_get_32: ~65.2ns sequential_get_100: ~148.2ns sequential_get_1000: ~2204ns Clearly demonstrating a cliff. In the discussion for v1 of this patch, Alexei noted that local_storage was 2.5x faster than a large hashmap when initially implemented [1]. The benchmark results show that local_storage is 5-10x faster: a long-running BPF application putting some pid-specific info into a hashmap for each pid it sees will probably see on the order of 10-100k pids. Bench numbers for hashmaps of this size are ~10x slower than sequential_get_16, but as the number of local_storage maps grows far past local_storage cache size the performance advantage shrinks and eventually reverses. When running the benchmarks it may be necessary to bump 'open files' ulimit for a successful run. [0]: https://lore.kernel.org/all/20220420002143.1096548-1-davemarchevsky@fb.com [1]: https://lore.kernel.org/bpf/20220511173305.ftldpn23m4ski3d3@MBP-98dd607d3435.dhcp.thefacebook.com/ Signed-off-by: Dave Marchevsky <davemarchevsky@fb.com> Link: https://lore.kernel.org/r/20220620222554.270578-1-davemarchevsky@fb.com Signed-off-by: Alexei Starovoitov <ast@kernel.org>
2022-06-21 06:25:54 +08:00
void local_storage_report_progress(int iter, struct bench_res *res,
long delta_ns)
{
double important_hits_per_sec, hits_per_sec;
double delta_sec = delta_ns / 1000000000.0;
hits_per_sec = res->hits / 1000000.0 / delta_sec;
important_hits_per_sec = res->important_hits / 1000000.0 / delta_sec;
printf("Iter %3d (%7.3lfus): ", iter, (delta_ns - 1000000000) / 1000.0);
printf("hits %8.3lfM/s ", hits_per_sec);
printf("important_hits %8.3lfM/s\n", important_hits_per_sec);
}
void local_storage_report_final(struct bench_res res[], int res_cnt)
{
double important_hits_mean = 0.0, important_hits_stddev = 0.0;
double hits_mean = 0.0, hits_stddev = 0.0;
int i;
for (i = 0; i < res_cnt; i++) {
hits_mean += res[i].hits / 1000000.0 / (0.0 + res_cnt);
important_hits_mean += res[i].important_hits / 1000000.0 / (0.0 + res_cnt);
}
if (res_cnt > 1) {
for (i = 0; i < res_cnt; i++) {
hits_stddev += (hits_mean - res[i].hits / 1000000.0) *
(hits_mean - res[i].hits / 1000000.0) /
(res_cnt - 1.0);
important_hits_stddev +=
(important_hits_mean - res[i].important_hits / 1000000.0) *
(important_hits_mean - res[i].important_hits / 1000000.0) /
(res_cnt - 1.0);
}
hits_stddev = sqrt(hits_stddev);
important_hits_stddev = sqrt(important_hits_stddev);
}
printf("Summary: hits throughput %8.3lf \u00B1 %5.3lf M ops/s, ",
hits_mean, hits_stddev);
printf("hits latency %8.3lf ns/op, ", 1000.0 / hits_mean);
printf("important_hits throughput %8.3lf \u00B1 %5.3lf M ops/s\n",
important_hits_mean, important_hits_stddev);
}
selftests/bpf: Add benchmark runner infrastructure While working on BPF ringbuf implementation, testing, and benchmarking, I've developed a pretty generic and modular benchmark runner, which seems to be generically useful, as I've already used it for one more purpose (testing fastest way to trigger BPF program, to minimize overhead of in-kernel code). This patch adds generic part of benchmark runner and sets up Makefile for extending it with more sets of benchmarks. Benchmarker itself operates by spinning up specified number of producer and consumer threads, setting up interval timer sending SIGALARM signal to application once a second. Every second, current snapshot with hits/drops counters are collected and stored in an array. Drops are useful for producer/consumer benchmarks in which producer might overwhelm consumers. Once test finishes after given amount of warm-up and testing seconds, mean and stddev are calculated (ignoring warm-up results) and is printed out to stdout. This setup seems to give consistent and accurate results. To validate behavior, I added two atomic counting tests: global and local. For global one, all the producer threads are atomically incrementing same counter as fast as possible. This, of course, leads to huge drop of performance once there is more than one producer thread due to CPUs fighting for the same memory location. Local counting, on the other hand, maintains one counter per each producer thread, incremented independently. Once per second, all counters are read and added together to form final "counting throughput" measurement. As expected, such setup demonstrates linear scalability with number of producers (as long as there are enough physical CPU cores, of course). See example output below. Also, this setup can nicely demonstrate disastrous effects of false sharing, if care is not taken to take those per-producer counters apart into independent cache lines. Demo output shows global counter first with 1 producer, then with 4. Both total and per-producer performance significantly drop. The last run is local counter with 4 producers, demonstrating near-perfect scalability. $ ./bench -a -w1 -d2 -p1 count-global Setting up benchmark 'count-global'... Benchmark 'count-global' started. Iter 0 ( 24.822us): hits 148.179M/s (148.179M/prod), drops 0.000M/s Iter 1 ( 37.939us): hits 149.308M/s (149.308M/prod), drops 0.000M/s Iter 2 (-10.774us): hits 150.717M/s (150.717M/prod), drops 0.000M/s Iter 3 ( 3.807us): hits 151.435M/s (151.435M/prod), drops 0.000M/s Summary: hits 150.488 ± 1.079M/s (150.488M/prod), drops 0.000 ± 0.000M/s $ ./bench -a -w1 -d2 -p4 count-global Setting up benchmark 'count-global'... Benchmark 'count-global' started. Iter 0 ( 60.659us): hits 53.910M/s ( 13.477M/prod), drops 0.000M/s Iter 1 (-17.658us): hits 53.722M/s ( 13.431M/prod), drops 0.000M/s Iter 2 ( 5.865us): hits 53.495M/s ( 13.374M/prod), drops 0.000M/s Iter 3 ( 0.104us): hits 53.606M/s ( 13.402M/prod), drops 0.000M/s Summary: hits 53.608 ± 0.113M/s ( 13.402M/prod), drops 0.000 ± 0.000M/s $ ./bench -a -w1 -d2 -p4 count-local Setting up benchmark 'count-local'... Benchmark 'count-local' started. Iter 0 ( 23.388us): hits 640.450M/s (160.113M/prod), drops 0.000M/s Iter 1 ( 2.291us): hits 605.661M/s (151.415M/prod), drops 0.000M/s Iter 2 ( -6.415us): hits 607.092M/s (151.773M/prod), drops 0.000M/s Iter 3 ( -1.361us): hits 601.796M/s (150.449M/prod), drops 0.000M/s Summary: hits 604.849 ± 2.739M/s (151.212M/prod), drops 0.000 ± 0.000M/s Benchmark runner supports setting thread affinity for producer and consumer threads. You can use -a flag for default CPU selection scheme, where first consumer gets CPU #0, next one gets CPU #1, and so on. Then producer threads pick up next CPU and increment one-by-one as well. But user can also specify a set of CPUs independently for producers and consumers with --prod-affinity 1,2-10,15 and --cons-affinity <set-of-cpus>. The latter allows to force producers and consumers to share same set of CPUs, if necessary. Signed-off-by: Andrii Nakryiko <andriin@fb.com> Signed-off-by: Alexei Starovoitov <ast@kernel.org> Acked-by: Yonghong Song <yhs@fb.com> Link: https://lore.kernel.org/bpf/20200512192445.2351848-3-andriin@fb.com
2020-05-13 03:24:43 +08:00
const char *argp_program_version = "benchmark";
const char *argp_program_bug_address = "<bpf@vger.kernel.org>";
const char argp_program_doc[] =
"benchmark Generic benchmarking framework.\n"
"\n"
"This tool runs benchmarks.\n"
"\n"
"USAGE: benchmark <bench-name>\n"
"\n"
"EXAMPLES:\n"
" # run 'count-local' benchmark with 1 producer and 1 consumer\n"
" benchmark count-local\n"
" # run 'count-local' with 16 producer and 8 consumer thread, pinned to CPUs\n"
" benchmark -p16 -c8 -a count-local\n";
enum {
ARG_PROD_AFFINITY_SET = 1000,
ARG_CONS_AFFINITY_SET = 1001,
};
static const struct argp_option opts[] = {
{ "list", 'l', NULL, 0, "List available benchmarks"},
{ "duration", 'd', "SEC", 0, "Duration of benchmark, seconds"},
{ "warmup", 'w', "SEC", 0, "Warm-up period, seconds"},
{ "producers", 'p', "NUM", 0, "Number of producer threads"},
{ "consumers", 'c', "NUM", 0, "Number of consumer threads"},
{ "verbose", 'v', NULL, 0, "Verbose debug output"},
{ "affinity", 'a', NULL, 0, "Set consumer/producer thread affinity"},
{ "prod-affinity", ARG_PROD_AFFINITY_SET, "CPUSET", 0,
"Set of CPUs for producer threads; implies --affinity"},
{ "cons-affinity", ARG_CONS_AFFINITY_SET, "CPUSET", 0,
"Set of CPUs for consumer threads; implies --affinity"},
{},
};
bpf: Add BPF ringbuf and perf buffer benchmarks Extend bench framework with ability to have benchmark-provided child argument parser for custom benchmark-specific parameters. This makes bench generic code modular and independent from any specific benchmark. Also implement a set of benchmarks for new BPF ring buffer and existing perf buffer. 4 benchmarks were implemented: 2 variations for each of BPF ringbuf and perfbuf:, - rb-libbpf utilizes stock libbpf ring_buffer manager for reading data; - rb-custom implements custom ring buffer setup and reading code, to eliminate overheads inherent in generic libbpf code due to callback functions and the need to update consumer position after each consumed record, instead of batching updates (due to pessimistic assumption that user callback might take long time and thus could unnecessarily hold ring buffer space for too long); - pb-libbpf uses stock libbpf perf_buffer code with all the default settings, though uses higher-performance raw event callback to minimize unnecessary overhead; - pb-custom implements its own custom consumer code to minimize any possible overhead of generic libbpf implementation and indirect function calls. All of the test support default, no data notification skipped, mode, as well as sampled mode (with --rb-sampled flag), which allows to trigger epoll notification less frequently and reduce overhead. As will be shown, this mode is especially critical for perf buffer, which suffers from high overhead of wakeups in kernel. Otherwise, all benchamrks implement similar way to generate a batch of records by using fentry/sys_getpgid BPF program, which pushes a bunch of records in a tight loop and records number of successful and dropped samples. Each record is a small 8-byte integer, to minimize the effect of memory copying with bpf_perf_event_output() and bpf_ringbuf_output(). Benchmarks that have only one producer implement optional back-to-back mode, in which record production and consumption is alternating on the same CPU. This is the highest-throughput happy case, showing ultimate performance achievable with either BPF ringbuf or perfbuf. All the below scenarios are implemented in a script in benchs/run_bench_ringbufs.sh. Tests were performed on 28-core/56-thread Intel Xeon CPU E5-2680 v4 @ 2.40GHz CPU. Single-producer, parallel producer ================================== rb-libbpf 12.054 ± 0.320M/s (drops 0.000 ± 0.000M/s) rb-custom 8.158 ± 0.118M/s (drops 0.001 ± 0.003M/s) pb-libbpf 0.931 ± 0.007M/s (drops 0.000 ± 0.000M/s) pb-custom 0.965 ± 0.003M/s (drops 0.000 ± 0.000M/s) Single-producer, parallel producer, sampled notification ======================================================== rb-libbpf 11.563 ± 0.067M/s (drops 0.000 ± 0.000M/s) rb-custom 15.895 ± 0.076M/s (drops 0.000 ± 0.000M/s) pb-libbpf 9.889 ± 0.032M/s (drops 0.000 ± 0.000M/s) pb-custom 9.866 ± 0.028M/s (drops 0.000 ± 0.000M/s) Single producer on one CPU, consumer on another one, both running at full speed. Curiously, rb-libbpf has higher throughput than objectively faster (due to more lightweight consumer code path) rb-custom. It appears that faster consumer causes kernel to send notifications more frequently, because consumer appears to be caught up more frequently. Performance of perfbuf suffers from default "no sampling" policy and huge overhead that causes. In sampled mode, rb-custom is winning very significantly eliminating too frequent in-kernel wakeups, the gain appears to be more than 2x. Perf buffer achieves even more impressive wins, compared to stock perfbuf settings, with 10x improvements in throughput with 1:500 sampling rate. The trade-off is that with sampling, application might not get next X events until X+1st arrives, which is not always acceptable. With steady influx of events, though, this shouldn't be a problem. Overall, single-producer performance of ring buffers seems to be better no matter the sampled/non-sampled modes, but it especially beats ring buffer without sampling due to its adaptive notification approach. Single-producer, back-to-back mode ================================== rb-libbpf 15.507 ± 0.247M/s (drops 0.000 ± 0.000M/s) rb-libbpf-sampled 14.692 ± 0.195M/s (drops 0.000 ± 0.000M/s) rb-custom 21.449 ± 0.157M/s (drops 0.000 ± 0.000M/s) rb-custom-sampled 20.024 ± 0.386M/s (drops 0.000 ± 0.000M/s) pb-libbpf 1.601 ± 0.015M/s (drops 0.000 ± 0.000M/s) pb-libbpf-sampled 8.545 ± 0.064M/s (drops 0.000 ± 0.000M/s) pb-custom 1.607 ± 0.022M/s (drops 0.000 ± 0.000M/s) pb-custom-sampled 8.988 ± 0.144M/s (drops 0.000 ± 0.000M/s) Here we test a back-to-back mode, which is arguably best-case scenario both for BPF ringbuf and perfbuf, because there is no contention and for ringbuf also no excessive notification, because consumer appears to be behind after the first record. For ringbuf, custom consumer code clearly wins with 21.5 vs 16 million records per second exchanged between producer and consumer. Sampled mode actually hurts a bit due to slightly slower producer logic (it needs to fetch amount of data available to decide whether to skip or force notification). Perfbuf with wakeup sampling gets 5.5x throughput increase, compared to no-sampling version. There also doesn't seem to be noticeable overhead from generic libbpf handling code. Perfbuf back-to-back, effect of sample rate =========================================== pb-sampled-1 1.035 ± 0.012M/s (drops 0.000 ± 0.000M/s) pb-sampled-5 3.476 ± 0.087M/s (drops 0.000 ± 0.000M/s) pb-sampled-10 5.094 ± 0.136M/s (drops 0.000 ± 0.000M/s) pb-sampled-25 7.118 ± 0.153M/s (drops 0.000 ± 0.000M/s) pb-sampled-50 8.169 ± 0.156M/s (drops 0.000 ± 0.000M/s) pb-sampled-100 8.887 ± 0.136M/s (drops 0.000 ± 0.000M/s) pb-sampled-250 9.180 ± 0.209M/s (drops 0.000 ± 0.000M/s) pb-sampled-500 9.353 ± 0.281M/s (drops 0.000 ± 0.000M/s) pb-sampled-1000 9.411 ± 0.217M/s (drops 0.000 ± 0.000M/s) pb-sampled-2000 9.464 ± 0.167M/s (drops 0.000 ± 0.000M/s) pb-sampled-3000 9.575 ± 0.273M/s (drops 0.000 ± 0.000M/s) This benchmark shows the effect of event sampling for perfbuf. Back-to-back mode for highest throughput. Just doing every 5th record notification gives 3.5x speed up. 250-500 appears to be the point of diminishing return, with almost 9x speed up. Most benchmarks use 500 as the default sampling for pb-raw and pb-custom. Ringbuf back-to-back, effect of sample rate =========================================== rb-sampled-1 1.106 ± 0.010M/s (drops 0.000 ± 0.000M/s) rb-sampled-5 4.746 ± 0.149M/s (drops 0.000 ± 0.000M/s) rb-sampled-10 7.706 ± 0.164M/s (drops 0.000 ± 0.000M/s) rb-sampled-25 12.893 ± 0.273M/s (drops 0.000 ± 0.000M/s) rb-sampled-50 15.961 ± 0.361M/s (drops 0.000 ± 0.000M/s) rb-sampled-100 18.203 ± 0.445M/s (drops 0.000 ± 0.000M/s) rb-sampled-250 19.962 ± 0.786M/s (drops 0.000 ± 0.000M/s) rb-sampled-500 20.881 ± 0.551M/s (drops 0.000 ± 0.000M/s) rb-sampled-1000 21.317 ± 0.532M/s (drops 0.000 ± 0.000M/s) rb-sampled-2000 21.331 ± 0.535M/s (drops 0.000 ± 0.000M/s) rb-sampled-3000 21.688 ± 0.392M/s (drops 0.000 ± 0.000M/s) Similar benchmark for ring buffer also shows a great advantage (in terms of throughput) of skipping notifications. Skipping every 5th one gives 4x boost. Also similar to perfbuf case, 250-500 seems to be the point of diminishing returns, giving roughly 20x better results. Keep in mind, for this test, notifications are controlled manually with BPF_RB_NO_WAKEUP and BPF_RB_FORCE_WAKEUP. As can be seen from previous benchmarks, adaptive notifications based on consumer's positions provides same (or even slightly better due to simpler load generator on BPF side) benefits in favorable back-to-back scenario. Over zealous and fast consumer, which is almost always caught up, will make thoughput numbers smaller. That's the case when manual notification control might prove to be extremely beneficial. Ringbuf back-to-back, reserve+commit vs output ============================================== reserve 22.819 ± 0.503M/s (drops 0.000 ± 0.000M/s) output 18.906 ± 0.433M/s (drops 0.000 ± 0.000M/s) Ringbuf sampled, reserve+commit vs output ========================================= reserve-sampled 15.350 ± 0.132M/s (drops 0.000 ± 0.000M/s) output-sampled 14.195 ± 0.144M/s (drops 0.000 ± 0.000M/s) BPF ringbuf supports two sets of APIs with various usability and performance tradeoffs: bpf_ringbuf_reserve()+bpf_ringbuf_commit() vs bpf_ringbuf_output(). This benchmark clearly shows superiority of reserve+commit approach, despite using a small 8-byte record size. Single-producer, consumer/producer competing on the same CPU, low batch count ============================================================================= rb-libbpf 3.045 ± 0.020M/s (drops 3.536 ± 0.148M/s) rb-custom 3.055 ± 0.022M/s (drops 3.893 ± 0.066M/s) pb-libbpf 1.393 ± 0.024M/s (drops 0.000 ± 0.000M/s) pb-custom 1.407 ± 0.016M/s (drops 0.000 ± 0.000M/s) This benchmark shows one of the worst-case scenarios, in which producer and consumer do not coordinate *and* fight for the same CPU. No batch count and sampling settings were able to eliminate drops for ringbuffer, producer is just too fast for consumer to keep up. But ringbuf and perfbuf still able to pass through quite a lot of messages, which is more than enough for a lot of applications. Ringbuf, multi-producer contention ================================== rb-libbpf nr_prod 1 10.916 ± 0.399M/s (drops 0.000 ± 0.000M/s) rb-libbpf nr_prod 2 4.931 ± 0.030M/s (drops 0.000 ± 0.000M/s) rb-libbpf nr_prod 3 4.880 ± 0.006M/s (drops 0.000 ± 0.000M/s) rb-libbpf nr_prod 4 3.926 ± 0.004M/s (drops 0.000 ± 0.000M/s) rb-libbpf nr_prod 8 4.011 ± 0.004M/s (drops 0.000 ± 0.000M/s) rb-libbpf nr_prod 12 3.967 ± 0.016M/s (drops 0.000 ± 0.000M/s) rb-libbpf nr_prod 16 2.604 ± 0.030M/s (drops 0.001 ± 0.002M/s) rb-libbpf nr_prod 20 2.233 ± 0.003M/s (drops 0.000 ± 0.000M/s) rb-libbpf nr_prod 24 2.085 ± 0.015M/s (drops 0.000 ± 0.000M/s) rb-libbpf nr_prod 28 2.055 ± 0.004M/s (drops 0.000 ± 0.000M/s) rb-libbpf nr_prod 32 1.962 ± 0.004M/s (drops 0.000 ± 0.000M/s) rb-libbpf nr_prod 36 2.089 ± 0.005M/s (drops 0.000 ± 0.000M/s) rb-libbpf nr_prod 40 2.118 ± 0.006M/s (drops 0.000 ± 0.000M/s) rb-libbpf nr_prod 44 2.105 ± 0.004M/s (drops 0.000 ± 0.000M/s) rb-libbpf nr_prod 48 2.120 ± 0.058M/s (drops 0.000 ± 0.001M/s) rb-libbpf nr_prod 52 2.074 ± 0.024M/s (drops 0.007 ± 0.014M/s) Ringbuf uses a very short-duration spinlock during reservation phase, to check few invariants, increment producer count and set record header. This is the biggest point of contention for ringbuf implementation. This benchmark evaluates the effect of multiple competing writers on overall throughput of a single shared ringbuffer. Overall throughput drops almost 2x when going from single to two highly-contended producers, gradually dropping with additional competing producers. Performance drop stabilizes at around 20 producers and hovers around 2mln even with 50+ fighting producers, which is a 5x drop compared to non-contended case. Good kernel implementation in kernel helps maintain decent performance here. Note, that in the intended real-world scenarios, it's not expected to get even close to such a high levels of contention. But if contention will become a problem, there is always an option of sharding few ring buffers across a set of CPUs. Signed-off-by: Andrii Nakryiko <andriin@fb.com> Signed-off-by: Daniel Borkmann <daniel@iogearbox.net> Link: https://lore.kernel.org/bpf/20200529075424.3139988-5-andriin@fb.com Signed-off-by: Alexei Starovoitov <ast@kernel.org>
2020-05-29 15:54:23 +08:00
extern struct argp bench_ringbufs_argp;
bpf/benchs: Add benchmark tests for bloom filter throughput + false positive This patch adds benchmark tests for the throughput (for lookups + updates) and the false positive rate of bloom filter lookups, as well as some minor refactoring of the bash script for running the benchmarks. These benchmarks show that as the number of hash functions increases, the throughput and the false positive rate of the bloom filter decreases. >From the benchmark data, the approximate average false-positive rates are roughly as follows: 1 hash function = ~30% 2 hash functions = ~15% 3 hash functions = ~5% 4 hash functions = ~2.5% 5 hash functions = ~1% 6 hash functions = ~0.5% 7 hash functions = ~0.35% 8 hash functions = ~0.15% 9 hash functions = ~0.1% 10 hash functions = ~0% For reference data, the benchmarks run on one thread on a machine with one numa node for 1 to 5 hash functions for 8-byte and 64-byte values are as follows: 1 hash function: 50k entries 8-byte value Lookups - 51.1 M/s operations Updates - 33.6 M/s operations False positive rate: 24.15% 64-byte value Lookups - 15.7 M/s operations Updates - 15.1 M/s operations False positive rate: 24.2% 100k entries 8-byte value Lookups - 51.0 M/s operations Updates - 33.4 M/s operations False positive rate: 24.04% 64-byte value Lookups - 15.6 M/s operations Updates - 14.6 M/s operations False positive rate: 24.06% 500k entries 8-byte value Lookups - 50.5 M/s operations Updates - 33.1 M/s operations False positive rate: 27.45% 64-byte value Lookups - 15.6 M/s operations Updates - 14.2 M/s operations False positive rate: 27.42% 1 mil entries 8-byte value Lookups - 49.7 M/s operations Updates - 32.9 M/s operations False positive rate: 27.45% 64-byte value Lookups - 15.4 M/s operations Updates - 13.7 M/s operations False positive rate: 27.58% 2.5 mil entries 8-byte value Lookups - 47.2 M/s operations Updates - 31.8 M/s operations False positive rate: 30.94% 64-byte value Lookups - 15.3 M/s operations Updates - 13.2 M/s operations False positive rate: 30.95% 5 mil entries 8-byte value Lookups - 41.1 M/s operations Updates - 28.1 M/s operations False positive rate: 31.01% 64-byte value Lookups - 13.3 M/s operations Updates - 11.4 M/s operations False positive rate: 30.98% 2 hash functions: 50k entries 8-byte value Lookups - 34.1 M/s operations Updates - 20.1 M/s operations False positive rate: 9.13% 64-byte value Lookups - 8.4 M/s operations Updates - 7.9 M/s operations False positive rate: 9.21% 100k entries 8-byte value Lookups - 33.7 M/s operations Updates - 18.9 M/s operations False positive rate: 9.13% 64-byte value Lookups - 8.4 M/s operations Updates - 7.7 M/s operations False positive rate: 9.19% 500k entries 8-byte value Lookups - 32.7 M/s operations Updates - 18.1 M/s operations False positive rate: 12.61% 64-byte value Lookups - 8.4 M/s operations Updates - 7.5 M/s operations False positive rate: 12.61% 1 mil entries 8-byte value Lookups - 30.6 M/s operations Updates - 18.9 M/s operations False positive rate: 12.54% 64-byte value Lookups - 8.0 M/s operations Updates - 7.0 M/s operations False positive rate: 12.52% 2.5 mil entries 8-byte value Lookups - 25.3 M/s operations Updates - 16.7 M/s operations False positive rate: 16.77% 64-byte value Lookups - 7.9 M/s operations Updates - 6.5 M/s operations False positive rate: 16.88% 5 mil entries 8-byte value Lookups - 20.8 M/s operations Updates - 14.7 M/s operations False positive rate: 16.78% 64-byte value Lookups - 7.0 M/s operations Updates - 6.0 M/s operations False positive rate: 16.78% 3 hash functions: 50k entries 8-byte value Lookups - 25.1 M/s operations Updates - 14.6 M/s operations False positive rate: 7.65% 64-byte value Lookups - 5.8 M/s operations Updates - 5.5 M/s operations False positive rate: 7.58% 100k entries 8-byte value Lookups - 24.7 M/s operations Updates - 14.1 M/s operations False positive rate: 7.71% 64-byte value Lookups - 5.8 M/s operations Updates - 5.3 M/s operations False positive rate: 7.62% 500k entries 8-byte value Lookups - 22.9 M/s operations Updates - 13.9 M/s operations False positive rate: 2.62% 64-byte value Lookups - 5.6 M/s operations Updates - 4.8 M/s operations False positive rate: 2.7% 1 mil entries 8-byte value Lookups - 19.8 M/s operations Updates - 12.6 M/s operations False positive rate: 2.60% 64-byte value Lookups - 5.3 M/s operations Updates - 4.4 M/s operations False positive rate: 2.69% 2.5 mil entries 8-byte value Lookups - 16.2 M/s operations Updates - 10.7 M/s operations False positive rate: 4.49% 64-byte value Lookups - 4.9 M/s operations Updates - 4.1 M/s operations False positive rate: 4.41% 5 mil entries 8-byte value Lookups - 18.8 M/s operations Updates - 9.2 M/s operations False positive rate: 4.45% 64-byte value Lookups - 5.2 M/s operations Updates - 3.9 M/s operations False positive rate: 4.54% 4 hash functions: 50k entries 8-byte value Lookups - 19.7 M/s operations Updates - 11.1 M/s operations False positive rate: 1.01% 64-byte value Lookups - 4.4 M/s operations Updates - 4.0 M/s operations False positive rate: 1.00% 100k entries 8-byte value Lookups - 19.5 M/s operations Updates - 10.9 M/s operations False positive rate: 1.00% 64-byte value Lookups - 4.3 M/s operations Updates - 3.9 M/s operations False positive rate: 0.97% 500k entries 8-byte value Lookups - 18.2 M/s operations Updates - 10.6 M/s operations False positive rate: 2.05% 64-byte value Lookups - 4.3 M/s operations Updates - 3.7 M/s operations False positive rate: 2.05% 1 mil entries 8-byte value Lookups - 15.5 M/s operations Updates - 9.6 M/s operations False positive rate: 1.99% 64-byte value Lookups - 4.0 M/s operations Updates - 3.4 M/s operations False positive rate: 1.99% 2.5 mil entries 8-byte value Lookups - 13.8 M/s operations Updates - 7.7 M/s operations False positive rate: 3.91% 64-byte value Lookups - 3.7 M/s operations Updates - 3.6 M/s operations False positive rate: 3.78% 5 mil entries 8-byte value Lookups - 13.0 M/s operations Updates - 6.9 M/s operations False positive rate: 3.93% 64-byte value Lookups - 3.5 M/s operations Updates - 3.7 M/s operations False positive rate: 3.39% 5 hash functions: 50k entries 8-byte value Lookups - 16.4 M/s operations Updates - 9.1 M/s operations False positive rate: 0.78% 64-byte value Lookups - 3.5 M/s operations Updates - 3.2 M/s operations False positive rate: 0.77% 100k entries 8-byte value Lookups - 16.3 M/s operations Updates - 9.0 M/s operations False positive rate: 0.79% 64-byte value Lookups - 3.5 M/s operations Updates - 3.2 M/s operations False positive rate: 0.78% 500k entries 8-byte value Lookups - 15.1 M/s operations Updates - 8.8 M/s operations False positive rate: 1.82% 64-byte value Lookups - 3.4 M/s operations Updates - 3.0 M/s operations False positive rate: 1.78% 1 mil entries 8-byte value Lookups - 13.2 M/s operations Updates - 7.8 M/s operations False positive rate: 1.81% 64-byte value Lookups - 3.2 M/s operations Updates - 2.8 M/s operations False positive rate: 1.80% 2.5 mil entries 8-byte value Lookups - 10.5 M/s operations Updates - 5.9 M/s operations False positive rate: 0.29% 64-byte value Lookups - 3.2 M/s operations Updates - 2.4 M/s operations False positive rate: 0.28% 5 mil entries 8-byte value Lookups - 9.6 M/s operations Updates - 5.7 M/s operations False positive rate: 0.30% 64-byte value Lookups - 3.2 M/s operations Updates - 2.7 M/s operations False positive rate: 0.30% Signed-off-by: Joanne Koong <joannekoong@fb.com> Signed-off-by: Alexei Starovoitov <ast@kernel.org> Acked-by: Andrii Nakryiko <andrii@kernel.org> Link: https://lore.kernel.org/bpf/20211027234504.30744-5-joannekoong@fb.com
2021-10-28 07:45:03 +08:00
extern struct argp bench_bloom_map_argp;
selftest/bpf/benchs: Add bpf_loop benchmark Add benchmark to measure the throughput and latency of the bpf_loop call. Testing this on my dev machine on 1 thread, the data is as follows: nr_loops: 10 bpf_loop - throughput: 198.519 ± 0.155 M ops/s, latency: 5.037 ns/op nr_loops: 100 bpf_loop - throughput: 247.448 ± 0.305 M ops/s, latency: 4.041 ns/op nr_loops: 500 bpf_loop - throughput: 260.839 ± 0.380 M ops/s, latency: 3.834 ns/op nr_loops: 1000 bpf_loop - throughput: 262.806 ± 0.629 M ops/s, latency: 3.805 ns/op nr_loops: 5000 bpf_loop - throughput: 264.211 ± 1.508 M ops/s, latency: 3.785 ns/op nr_loops: 10000 bpf_loop - throughput: 265.366 ± 3.054 M ops/s, latency: 3.768 ns/op nr_loops: 50000 bpf_loop - throughput: 235.986 ± 20.205 M ops/s, latency: 4.238 ns/op nr_loops: 100000 bpf_loop - throughput: 264.482 ± 0.279 M ops/s, latency: 3.781 ns/op nr_loops: 500000 bpf_loop - throughput: 309.773 ± 87.713 M ops/s, latency: 3.228 ns/op nr_loops: 1000000 bpf_loop - throughput: 262.818 ± 4.143 M ops/s, latency: 3.805 ns/op >From this data, we can see that the latency per loop decreases as the number of loops increases. On this particular machine, each loop had an overhead of about ~4 ns, and we were able to run ~250 million loops per second. Signed-off-by: Joanne Koong <joannekoong@fb.com> Signed-off-by: Alexei Starovoitov <ast@kernel.org> Acked-by: Andrii Nakryiko <andrii@kernel.org> Link: https://lore.kernel.org/bpf/20211130030622.4131246-5-joannekoong@fb.com
2021-11-30 11:06:22 +08:00
extern struct argp bench_bpf_loop_argp;
selftests/bpf: Add benchmark for local_storage get Add a benchmarks to demonstrate the performance cliff for local_storage get as the number of local_storage maps increases beyond current local_storage implementation's cache size. "sequential get" and "interleaved get" benchmarks are added, both of which do many bpf_task_storage_get calls on sets of task local_storage maps of various counts, while considering a single specific map to be 'important' and counting task_storage_gets to the important map separately in addition to normal 'hits' count of all gets. Goal here is to mimic scenario where a particular program using one map - the important one - is running on a system where many other local_storage maps exist and are accessed often. While "sequential get" benchmark does bpf_task_storage_get for map 0, 1, ..., {9, 99, 999} in order, "interleaved" benchmark interleaves 4 bpf_task_storage_gets for the important map for every 10 map gets. This is meant to highlight performance differences when important map is accessed far more frequently than non-important maps. A "hashmap control" benchmark is also included for easy comparison of standard bpf hashmap lookup vs local_storage get. The benchmark is similar to "sequential get", but creates and uses BPF_MAP_TYPE_HASH instead of local storage. Only one inner map is created - a hashmap meant to hold tid -> data mapping for all tasks. Size of the hashmap is hardcoded to my system's PID_MAX_LIMIT (4,194,304). The number of these keys which are actually fetched as part of the benchmark is configurable. Addition of this benchmark is inspired by conversation with Alexei in a previous patchset's thread [0], which highlighted the need for such a benchmark to motivate and validate improvements to local_storage implementation. My approach in that series focused on improving performance for explicitly-marked 'important' maps and was rejected with feedback to make more generally-applicable improvements while avoiding explicitly marking maps as important. Thus the benchmark reports both general and important-map-focused metrics, so effect of future work on both is clear. Regarding the benchmark results. On a powerful system (Skylake, 20 cores, 256gb ram): Hashmap Control =============== num keys: 10 hashmap (control) sequential get: hits throughput: 20.900 ± 0.334 M ops/s, hits latency: 47.847 ns/op, important_hits throughput: 20.900 ± 0.334 M ops/s num keys: 1000 hashmap (control) sequential get: hits throughput: 13.758 ± 0.219 M ops/s, hits latency: 72.683 ns/op, important_hits throughput: 13.758 ± 0.219 M ops/s num keys: 10000 hashmap (control) sequential get: hits throughput: 6.995 ± 0.034 M ops/s, hits latency: 142.959 ns/op, important_hits throughput: 6.995 ± 0.034 M ops/s num keys: 100000 hashmap (control) sequential get: hits throughput: 4.452 ± 0.371 M ops/s, hits latency: 224.635 ns/op, important_hits throughput: 4.452 ± 0.371 M ops/s num keys: 4194304 hashmap (control) sequential get: hits throughput: 3.043 ± 0.033 M ops/s, hits latency: 328.587 ns/op, important_hits throughput: 3.043 ± 0.033 M ops/s Local Storage ============= num_maps: 1 local_storage cache sequential get: hits throughput: 47.298 ± 0.180 M ops/s, hits latency: 21.142 ns/op, important_hits throughput: 47.298 ± 0.180 M ops/s local_storage cache interleaved get: hits throughput: 55.277 ± 0.888 M ops/s, hits latency: 18.091 ns/op, important_hits throughput: 55.277 ± 0.888 M ops/s num_maps: 10 local_storage cache sequential get: hits throughput: 40.240 ± 0.802 M ops/s, hits latency: 24.851 ns/op, important_hits throughput: 4.024 ± 0.080 M ops/s local_storage cache interleaved get: hits throughput: 48.701 ± 0.722 M ops/s, hits latency: 20.533 ns/op, important_hits throughput: 17.393 ± 0.258 M ops/s num_maps: 16 local_storage cache sequential get: hits throughput: 44.515 ± 0.708 M ops/s, hits latency: 22.464 ns/op, important_hits throughput: 2.782 ± 0.044 M ops/s local_storage cache interleaved get: hits throughput: 49.553 ± 2.260 M ops/s, hits latency: 20.181 ns/op, important_hits throughput: 15.767 ± 0.719 M ops/s num_maps: 17 local_storage cache sequential get: hits throughput: 38.778 ± 0.302 M ops/s, hits latency: 25.788 ns/op, important_hits throughput: 2.284 ± 0.018 M ops/s local_storage cache interleaved get: hits throughput: 43.848 ± 1.023 M ops/s, hits latency: 22.806 ns/op, important_hits throughput: 13.349 ± 0.311 M ops/s num_maps: 24 local_storage cache sequential get: hits throughput: 19.317 ± 0.568 M ops/s, hits latency: 51.769 ns/op, important_hits throughput: 0.806 ± 0.024 M ops/s local_storage cache interleaved get: hits throughput: 24.397 ± 0.272 M ops/s, hits latency: 40.989 ns/op, important_hits throughput: 6.863 ± 0.077 M ops/s num_maps: 32 local_storage cache sequential get: hits throughput: 13.333 ± 0.135 M ops/s, hits latency: 75.000 ns/op, important_hits throughput: 0.417 ± 0.004 M ops/s local_storage cache interleaved get: hits throughput: 16.898 ± 0.383 M ops/s, hits latency: 59.178 ns/op, important_hits throughput: 4.717 ± 0.107 M ops/s num_maps: 100 local_storage cache sequential get: hits throughput: 6.360 ± 0.107 M ops/s, hits latency: 157.233 ns/op, important_hits throughput: 0.064 ± 0.001 M ops/s local_storage cache interleaved get: hits throughput: 7.303 ± 0.362 M ops/s, hits latency: 136.930 ns/op, important_hits throughput: 1.907 ± 0.094 M ops/s num_maps: 1000 local_storage cache sequential get: hits throughput: 0.452 ± 0.010 M ops/s, hits latency: 2214.022 ns/op, important_hits throughput: 0.000 ± 0.000 M ops/s local_storage cache interleaved get: hits throughput: 0.542 ± 0.007 M ops/s, hits latency: 1843.341 ns/op, important_hits throughput: 0.136 ± 0.002 M ops/s Looking at the "sequential get" results, it's clear that as the number of task local_storage maps grows beyond the current cache size (16), there's a significant reduction in hits throughput. Note that current local_storage implementation assigns a cache_idx to maps as they are created. Since "sequential get" is creating maps 0..n in order and then doing bpf_task_storage_get calls in the same order, the benchmark is effectively ensuring that a map will not be in cache when the program tries to access it. For "interleaved get" results, important-map hits throughput is greatly increased as the important map is more likely to be in cache by virtue of being accessed far more frequently. Throughput still reduces as # maps increases, though. To get a sense of the overhead of the benchmark program, I commented out bpf_task_storage_get/bpf_map_lookup_elem in local_storage_bench.c and ran the benchmark on the same host as the 'real' run. Results: Hashmap Control =============== num keys: 10 hashmap (control) sequential get: hits throughput: 54.288 ± 0.655 M ops/s, hits latency: 18.420 ns/op, important_hits throughput: 54.288 ± 0.655 M ops/s num keys: 1000 hashmap (control) sequential get: hits throughput: 52.913 ± 0.519 M ops/s, hits latency: 18.899 ns/op, important_hits throughput: 52.913 ± 0.519 M ops/s num keys: 10000 hashmap (control) sequential get: hits throughput: 53.480 ± 1.235 M ops/s, hits latency: 18.699 ns/op, important_hits throughput: 53.480 ± 1.235 M ops/s num keys: 100000 hashmap (control) sequential get: hits throughput: 54.982 ± 1.902 M ops/s, hits latency: 18.188 ns/op, important_hits throughput: 54.982 ± 1.902 M ops/s num keys: 4194304 hashmap (control) sequential get: hits throughput: 50.858 ± 0.707 M ops/s, hits latency: 19.662 ns/op, important_hits throughput: 50.858 ± 0.707 M ops/s Local Storage ============= num_maps: 1 local_storage cache sequential get: hits throughput: 110.990 ± 4.828 M ops/s, hits latency: 9.010 ns/op, important_hits throughput: 110.990 ± 4.828 M ops/s local_storage cache interleaved get: hits throughput: 161.057 ± 4.090 M ops/s, hits latency: 6.209 ns/op, important_hits throughput: 161.057 ± 4.090 M ops/s num_maps: 10 local_storage cache sequential get: hits throughput: 112.930 ± 1.079 M ops/s, hits latency: 8.855 ns/op, important_hits throughput: 11.293 ± 0.108 M ops/s local_storage cache interleaved get: hits throughput: 115.841 ± 2.088 M ops/s, hits latency: 8.633 ns/op, important_hits throughput: 41.372 ± 0.746 M ops/s num_maps: 16 local_storage cache sequential get: hits throughput: 115.653 ± 0.416 M ops/s, hits latency: 8.647 ns/op, important_hits throughput: 7.228 ± 0.026 M ops/s local_storage cache interleaved get: hits throughput: 138.717 ± 1.649 M ops/s, hits latency: 7.209 ns/op, important_hits throughput: 44.137 ± 0.525 M ops/s num_maps: 17 local_storage cache sequential get: hits throughput: 112.020 ± 1.649 M ops/s, hits latency: 8.927 ns/op, important_hits throughput: 6.598 ± 0.097 M ops/s local_storage cache interleaved get: hits throughput: 128.089 ± 1.960 M ops/s, hits latency: 7.807 ns/op, important_hits throughput: 38.995 ± 0.597 M ops/s num_maps: 24 local_storage cache sequential get: hits throughput: 92.447 ± 5.170 M ops/s, hits latency: 10.817 ns/op, important_hits throughput: 3.855 ± 0.216 M ops/s local_storage cache interleaved get: hits throughput: 128.844 ± 2.808 M ops/s, hits latency: 7.761 ns/op, important_hits throughput: 36.245 ± 0.790 M ops/s num_maps: 32 local_storage cache sequential get: hits throughput: 102.042 ± 1.462 M ops/s, hits latency: 9.800 ns/op, important_hits throughput: 3.194 ± 0.046 M ops/s local_storage cache interleaved get: hits throughput: 126.577 ± 1.818 M ops/s, hits latency: 7.900 ns/op, important_hits throughput: 35.332 ± 0.507 M ops/s num_maps: 100 local_storage cache sequential get: hits throughput: 111.327 ± 1.401 M ops/s, hits latency: 8.983 ns/op, important_hits throughput: 1.113 ± 0.014 M ops/s local_storage cache interleaved get: hits throughput: 131.327 ± 1.339 M ops/s, hits latency: 7.615 ns/op, important_hits throughput: 34.302 ± 0.350 M ops/s num_maps: 1000 local_storage cache sequential get: hits throughput: 101.978 ± 0.563 M ops/s, hits latency: 9.806 ns/op, important_hits throughput: 0.102 ± 0.001 M ops/s local_storage cache interleaved get: hits throughput: 141.084 ± 1.098 M ops/s, hits latency: 7.088 ns/op, important_hits throughput: 35.430 ± 0.276 M ops/s Adjusting for overhead, latency numbers for "hashmap control" and "sequential get" are: hashmap_control_1k: ~53.8ns hashmap_control_10k: ~124.2ns hashmap_control_100k: ~206.5ns sequential_get_1: ~12.1ns sequential_get_10: ~16.0ns sequential_get_16: ~13.8ns sequential_get_17: ~16.8ns sequential_get_24: ~40.9ns sequential_get_32: ~65.2ns sequential_get_100: ~148.2ns sequential_get_1000: ~2204ns Clearly demonstrating a cliff. In the discussion for v1 of this patch, Alexei noted that local_storage was 2.5x faster than a large hashmap when initially implemented [1]. The benchmark results show that local_storage is 5-10x faster: a long-running BPF application putting some pid-specific info into a hashmap for each pid it sees will probably see on the order of 10-100k pids. Bench numbers for hashmaps of this size are ~10x slower than sequential_get_16, but as the number of local_storage maps grows far past local_storage cache size the performance advantage shrinks and eventually reverses. When running the benchmarks it may be necessary to bump 'open files' ulimit for a successful run. [0]: https://lore.kernel.org/all/20220420002143.1096548-1-davemarchevsky@fb.com [1]: https://lore.kernel.org/bpf/20220511173305.ftldpn23m4ski3d3@MBP-98dd607d3435.dhcp.thefacebook.com/ Signed-off-by: Dave Marchevsky <davemarchevsky@fb.com> Link: https://lore.kernel.org/r/20220620222554.270578-1-davemarchevsky@fb.com Signed-off-by: Alexei Starovoitov <ast@kernel.org>
2022-06-21 06:25:54 +08:00
extern struct argp bench_local_storage_argp;
selftests/bpf: Add benchmark for local_storage RCU Tasks Trace usage This benchmark measures grace period latency and kthread cpu usage of RCU Tasks Trace when many processes are creating/deleting BPF local_storage. Intent here is to quantify improvement on these metrics after Paul's recent RCU Tasks patches [0]. Specifically, fork 15k tasks which call a bpf prog that creates/destroys task local_storage and sleep in a loop, resulting in many call_rcu_tasks_trace calls. To determine grace period latency, trace time elapsed between rcu_tasks_trace_pregp_step and rcu_tasks_trace_postgp; for cpu usage look at rcu_task_trace_kthread's stime in /proc/PID/stat. On my virtualized test environment (Skylake, 8 cpus) benchmark results demonstrate significant improvement: BEFORE Paul's patches: SUMMARY tasks_trace grace period latency avg 22298.551 us stddev 1302.165 us SUMMARY ticks per tasks_trace grace period avg 2.291 stddev 0.324 AFTER Paul's patches: SUMMARY tasks_trace grace period latency avg 16969.197 us stddev 2525.053 us SUMMARY ticks per tasks_trace grace period avg 1.146 stddev 0.178 Note that since these patches are not in bpf-next benchmarking was done by cherry-picking this patch onto rcu tree. [0] https://lore.kernel.org/rcu/20220620225402.GA3842369@paulmck-ThinkPad-P17-Gen-1/ Signed-off-by: Dave Marchevsky <davemarchevsky@fb.com> Signed-off-by: Daniel Borkmann <daniel@iogearbox.net> Acked-by: Paul E. McKenney <paulmck@kernel.org> Acked-by: Martin KaFai Lau <kafai@fb.com> Link: https://lore.kernel.org/bpf/20220705190018.3239050-1-davemarchevsky@fb.com
2022-07-06 03:00:18 +08:00
extern struct argp bench_local_storage_rcu_tasks_trace_argp;
selftests/bpf: Add benchmark for bpf_strncmp() helper Add benchmark to compare the performance between home-made strncmp() in bpf program and bpf_strncmp() helper. In summary, the performance win of bpf_strncmp() under x86-64 is greater than 18% when the compared string length is greater than 64, and is 179% when the length is 4095. Under arm64 the performance win is even bigger: 33% when the length is greater than 64 and 600% when the length is 4095. The following is the details: no-helper-X: use home-made strncmp() to compare X-sized string helper-Y: use bpf_strncmp() to compare Y-sized string Under x86-64: no-helper-1 3.504 ± 0.000M/s (drops 0.000 ± 0.000M/s) helper-1 3.347 ± 0.001M/s (drops 0.000 ± 0.000M/s) no-helper-8 3.357 ± 0.001M/s (drops 0.000 ± 0.000M/s) helper-8 3.307 ± 0.001M/s (drops 0.000 ± 0.000M/s) no-helper-32 3.064 ± 0.000M/s (drops 0.000 ± 0.000M/s) helper-32 3.253 ± 0.001M/s (drops 0.000 ± 0.000M/s) no-helper-64 2.563 ± 0.001M/s (drops 0.000 ± 0.000M/s) helper-64 3.040 ± 0.001M/s (drops 0.000 ± 0.000M/s) no-helper-128 1.975 ± 0.000M/s (drops 0.000 ± 0.000M/s) helper-128 2.641 ± 0.000M/s (drops 0.000 ± 0.000M/s) no-helper-512 0.759 ± 0.000M/s (drops 0.000 ± 0.000M/s) helper-512 1.574 ± 0.000M/s (drops 0.000 ± 0.000M/s) no-helper-2048 0.329 ± 0.000M/s (drops 0.000 ± 0.000M/s) helper-2048 0.602 ± 0.000M/s (drops 0.000 ± 0.000M/s) no-helper-4095 0.117 ± 0.000M/s (drops 0.000 ± 0.000M/s) helper-4095 0.327 ± 0.000M/s (drops 0.000 ± 0.000M/s) Under arm64: no-helper-1 2.806 ± 0.004M/s (drops 0.000 ± 0.000M/s) helper-1 2.819 ± 0.002M/s (drops 0.000 ± 0.000M/s) no-helper-8 2.797 ± 0.109M/s (drops 0.000 ± 0.000M/s) helper-8 2.786 ± 0.025M/s (drops 0.000 ± 0.000M/s) no-helper-32 2.399 ± 0.011M/s (drops 0.000 ± 0.000M/s) helper-32 2.703 ± 0.002M/s (drops 0.000 ± 0.000M/s) no-helper-64 2.020 ± 0.015M/s (drops 0.000 ± 0.000M/s) helper-64 2.702 ± 0.073M/s (drops 0.000 ± 0.000M/s) no-helper-128 1.604 ± 0.001M/s (drops 0.000 ± 0.000M/s) helper-128 2.516 ± 0.002M/s (drops 0.000 ± 0.000M/s) no-helper-512 0.699 ± 0.000M/s (drops 0.000 ± 0.000M/s) helper-512 2.106 ± 0.003M/s (drops 0.000 ± 0.000M/s) no-helper-2048 0.215 ± 0.000M/s (drops 0.000 ± 0.000M/s) helper-2048 1.223 ± 0.003M/s (drops 0.000 ± 0.000M/s) no-helper-4095 0.112 ± 0.000M/s (drops 0.000 ± 0.000M/s) helper-4095 0.796 ± 0.000M/s (drops 0.000 ± 0.000M/s) Signed-off-by: Hou Tao <houtao1@huawei.com> Signed-off-by: Alexei Starovoitov <ast@kernel.org> Link: https://lore.kernel.org/bpf/20211210141652.877186-4-houtao1@huawei.com
2021-12-10 22:16:51 +08:00
extern struct argp bench_strncmp_argp;
bpf: Add BPF ringbuf and perf buffer benchmarks Extend bench framework with ability to have benchmark-provided child argument parser for custom benchmark-specific parameters. This makes bench generic code modular and independent from any specific benchmark. Also implement a set of benchmarks for new BPF ring buffer and existing perf buffer. 4 benchmarks were implemented: 2 variations for each of BPF ringbuf and perfbuf:, - rb-libbpf utilizes stock libbpf ring_buffer manager for reading data; - rb-custom implements custom ring buffer setup and reading code, to eliminate overheads inherent in generic libbpf code due to callback functions and the need to update consumer position after each consumed record, instead of batching updates (due to pessimistic assumption that user callback might take long time and thus could unnecessarily hold ring buffer space for too long); - pb-libbpf uses stock libbpf perf_buffer code with all the default settings, though uses higher-performance raw event callback to minimize unnecessary overhead; - pb-custom implements its own custom consumer code to minimize any possible overhead of generic libbpf implementation and indirect function calls. All of the test support default, no data notification skipped, mode, as well as sampled mode (with --rb-sampled flag), which allows to trigger epoll notification less frequently and reduce overhead. As will be shown, this mode is especially critical for perf buffer, which suffers from high overhead of wakeups in kernel. Otherwise, all benchamrks implement similar way to generate a batch of records by using fentry/sys_getpgid BPF program, which pushes a bunch of records in a tight loop and records number of successful and dropped samples. Each record is a small 8-byte integer, to minimize the effect of memory copying with bpf_perf_event_output() and bpf_ringbuf_output(). Benchmarks that have only one producer implement optional back-to-back mode, in which record production and consumption is alternating on the same CPU. This is the highest-throughput happy case, showing ultimate performance achievable with either BPF ringbuf or perfbuf. All the below scenarios are implemented in a script in benchs/run_bench_ringbufs.sh. Tests were performed on 28-core/56-thread Intel Xeon CPU E5-2680 v4 @ 2.40GHz CPU. Single-producer, parallel producer ================================== rb-libbpf 12.054 ± 0.320M/s (drops 0.000 ± 0.000M/s) rb-custom 8.158 ± 0.118M/s (drops 0.001 ± 0.003M/s) pb-libbpf 0.931 ± 0.007M/s (drops 0.000 ± 0.000M/s) pb-custom 0.965 ± 0.003M/s (drops 0.000 ± 0.000M/s) Single-producer, parallel producer, sampled notification ======================================================== rb-libbpf 11.563 ± 0.067M/s (drops 0.000 ± 0.000M/s) rb-custom 15.895 ± 0.076M/s (drops 0.000 ± 0.000M/s) pb-libbpf 9.889 ± 0.032M/s (drops 0.000 ± 0.000M/s) pb-custom 9.866 ± 0.028M/s (drops 0.000 ± 0.000M/s) Single producer on one CPU, consumer on another one, both running at full speed. Curiously, rb-libbpf has higher throughput than objectively faster (due to more lightweight consumer code path) rb-custom. It appears that faster consumer causes kernel to send notifications more frequently, because consumer appears to be caught up more frequently. Performance of perfbuf suffers from default "no sampling" policy and huge overhead that causes. In sampled mode, rb-custom is winning very significantly eliminating too frequent in-kernel wakeups, the gain appears to be more than 2x. Perf buffer achieves even more impressive wins, compared to stock perfbuf settings, with 10x improvements in throughput with 1:500 sampling rate. The trade-off is that with sampling, application might not get next X events until X+1st arrives, which is not always acceptable. With steady influx of events, though, this shouldn't be a problem. Overall, single-producer performance of ring buffers seems to be better no matter the sampled/non-sampled modes, but it especially beats ring buffer without sampling due to its adaptive notification approach. Single-producer, back-to-back mode ================================== rb-libbpf 15.507 ± 0.247M/s (drops 0.000 ± 0.000M/s) rb-libbpf-sampled 14.692 ± 0.195M/s (drops 0.000 ± 0.000M/s) rb-custom 21.449 ± 0.157M/s (drops 0.000 ± 0.000M/s) rb-custom-sampled 20.024 ± 0.386M/s (drops 0.000 ± 0.000M/s) pb-libbpf 1.601 ± 0.015M/s (drops 0.000 ± 0.000M/s) pb-libbpf-sampled 8.545 ± 0.064M/s (drops 0.000 ± 0.000M/s) pb-custom 1.607 ± 0.022M/s (drops 0.000 ± 0.000M/s) pb-custom-sampled 8.988 ± 0.144M/s (drops 0.000 ± 0.000M/s) Here we test a back-to-back mode, which is arguably best-case scenario both for BPF ringbuf and perfbuf, because there is no contention and for ringbuf also no excessive notification, because consumer appears to be behind after the first record. For ringbuf, custom consumer code clearly wins with 21.5 vs 16 million records per second exchanged between producer and consumer. Sampled mode actually hurts a bit due to slightly slower producer logic (it needs to fetch amount of data available to decide whether to skip or force notification). Perfbuf with wakeup sampling gets 5.5x throughput increase, compared to no-sampling version. There also doesn't seem to be noticeable overhead from generic libbpf handling code. Perfbuf back-to-back, effect of sample rate =========================================== pb-sampled-1 1.035 ± 0.012M/s (drops 0.000 ± 0.000M/s) pb-sampled-5 3.476 ± 0.087M/s (drops 0.000 ± 0.000M/s) pb-sampled-10 5.094 ± 0.136M/s (drops 0.000 ± 0.000M/s) pb-sampled-25 7.118 ± 0.153M/s (drops 0.000 ± 0.000M/s) pb-sampled-50 8.169 ± 0.156M/s (drops 0.000 ± 0.000M/s) pb-sampled-100 8.887 ± 0.136M/s (drops 0.000 ± 0.000M/s) pb-sampled-250 9.180 ± 0.209M/s (drops 0.000 ± 0.000M/s) pb-sampled-500 9.353 ± 0.281M/s (drops 0.000 ± 0.000M/s) pb-sampled-1000 9.411 ± 0.217M/s (drops 0.000 ± 0.000M/s) pb-sampled-2000 9.464 ± 0.167M/s (drops 0.000 ± 0.000M/s) pb-sampled-3000 9.575 ± 0.273M/s (drops 0.000 ± 0.000M/s) This benchmark shows the effect of event sampling for perfbuf. Back-to-back mode for highest throughput. Just doing every 5th record notification gives 3.5x speed up. 250-500 appears to be the point of diminishing return, with almost 9x speed up. Most benchmarks use 500 as the default sampling for pb-raw and pb-custom. Ringbuf back-to-back, effect of sample rate =========================================== rb-sampled-1 1.106 ± 0.010M/s (drops 0.000 ± 0.000M/s) rb-sampled-5 4.746 ± 0.149M/s (drops 0.000 ± 0.000M/s) rb-sampled-10 7.706 ± 0.164M/s (drops 0.000 ± 0.000M/s) rb-sampled-25 12.893 ± 0.273M/s (drops 0.000 ± 0.000M/s) rb-sampled-50 15.961 ± 0.361M/s (drops 0.000 ± 0.000M/s) rb-sampled-100 18.203 ± 0.445M/s (drops 0.000 ± 0.000M/s) rb-sampled-250 19.962 ± 0.786M/s (drops 0.000 ± 0.000M/s) rb-sampled-500 20.881 ± 0.551M/s (drops 0.000 ± 0.000M/s) rb-sampled-1000 21.317 ± 0.532M/s (drops 0.000 ± 0.000M/s) rb-sampled-2000 21.331 ± 0.535M/s (drops 0.000 ± 0.000M/s) rb-sampled-3000 21.688 ± 0.392M/s (drops 0.000 ± 0.000M/s) Similar benchmark for ring buffer also shows a great advantage (in terms of throughput) of skipping notifications. Skipping every 5th one gives 4x boost. Also similar to perfbuf case, 250-500 seems to be the point of diminishing returns, giving roughly 20x better results. Keep in mind, for this test, notifications are controlled manually with BPF_RB_NO_WAKEUP and BPF_RB_FORCE_WAKEUP. As can be seen from previous benchmarks, adaptive notifications based on consumer's positions provides same (or even slightly better due to simpler load generator on BPF side) benefits in favorable back-to-back scenario. Over zealous and fast consumer, which is almost always caught up, will make thoughput numbers smaller. That's the case when manual notification control might prove to be extremely beneficial. Ringbuf back-to-back, reserve+commit vs output ============================================== reserve 22.819 ± 0.503M/s (drops 0.000 ± 0.000M/s) output 18.906 ± 0.433M/s (drops 0.000 ± 0.000M/s) Ringbuf sampled, reserve+commit vs output ========================================= reserve-sampled 15.350 ± 0.132M/s (drops 0.000 ± 0.000M/s) output-sampled 14.195 ± 0.144M/s (drops 0.000 ± 0.000M/s) BPF ringbuf supports two sets of APIs with various usability and performance tradeoffs: bpf_ringbuf_reserve()+bpf_ringbuf_commit() vs bpf_ringbuf_output(). This benchmark clearly shows superiority of reserve+commit approach, despite using a small 8-byte record size. Single-producer, consumer/producer competing on the same CPU, low batch count ============================================================================= rb-libbpf 3.045 ± 0.020M/s (drops 3.536 ± 0.148M/s) rb-custom 3.055 ± 0.022M/s (drops 3.893 ± 0.066M/s) pb-libbpf 1.393 ± 0.024M/s (drops 0.000 ± 0.000M/s) pb-custom 1.407 ± 0.016M/s (drops 0.000 ± 0.000M/s) This benchmark shows one of the worst-case scenarios, in which producer and consumer do not coordinate *and* fight for the same CPU. No batch count and sampling settings were able to eliminate drops for ringbuffer, producer is just too fast for consumer to keep up. But ringbuf and perfbuf still able to pass through quite a lot of messages, which is more than enough for a lot of applications. Ringbuf, multi-producer contention ================================== rb-libbpf nr_prod 1 10.916 ± 0.399M/s (drops 0.000 ± 0.000M/s) rb-libbpf nr_prod 2 4.931 ± 0.030M/s (drops 0.000 ± 0.000M/s) rb-libbpf nr_prod 3 4.880 ± 0.006M/s (drops 0.000 ± 0.000M/s) rb-libbpf nr_prod 4 3.926 ± 0.004M/s (drops 0.000 ± 0.000M/s) rb-libbpf nr_prod 8 4.011 ± 0.004M/s (drops 0.000 ± 0.000M/s) rb-libbpf nr_prod 12 3.967 ± 0.016M/s (drops 0.000 ± 0.000M/s) rb-libbpf nr_prod 16 2.604 ± 0.030M/s (drops 0.001 ± 0.002M/s) rb-libbpf nr_prod 20 2.233 ± 0.003M/s (drops 0.000 ± 0.000M/s) rb-libbpf nr_prod 24 2.085 ± 0.015M/s (drops 0.000 ± 0.000M/s) rb-libbpf nr_prod 28 2.055 ± 0.004M/s (drops 0.000 ± 0.000M/s) rb-libbpf nr_prod 32 1.962 ± 0.004M/s (drops 0.000 ± 0.000M/s) rb-libbpf nr_prod 36 2.089 ± 0.005M/s (drops 0.000 ± 0.000M/s) rb-libbpf nr_prod 40 2.118 ± 0.006M/s (drops 0.000 ± 0.000M/s) rb-libbpf nr_prod 44 2.105 ± 0.004M/s (drops 0.000 ± 0.000M/s) rb-libbpf nr_prod 48 2.120 ± 0.058M/s (drops 0.000 ± 0.001M/s) rb-libbpf nr_prod 52 2.074 ± 0.024M/s (drops 0.007 ± 0.014M/s) Ringbuf uses a very short-duration spinlock during reservation phase, to check few invariants, increment producer count and set record header. This is the biggest point of contention for ringbuf implementation. This benchmark evaluates the effect of multiple competing writers on overall throughput of a single shared ringbuffer. Overall throughput drops almost 2x when going from single to two highly-contended producers, gradually dropping with additional competing producers. Performance drop stabilizes at around 20 producers and hovers around 2mln even with 50+ fighting producers, which is a 5x drop compared to non-contended case. Good kernel implementation in kernel helps maintain decent performance here. Note, that in the intended real-world scenarios, it's not expected to get even close to such a high levels of contention. But if contention will become a problem, there is always an option of sharding few ring buffers across a set of CPUs. Signed-off-by: Andrii Nakryiko <andriin@fb.com> Signed-off-by: Daniel Borkmann <daniel@iogearbox.net> Link: https://lore.kernel.org/bpf/20200529075424.3139988-5-andriin@fb.com Signed-off-by: Alexei Starovoitov <ast@kernel.org>
2020-05-29 15:54:23 +08:00
static const struct argp_child bench_parsers[] = {
{ &bench_ringbufs_argp, 0, "Ring buffers benchmark", 0 },
bpf/benchs: Add benchmark tests for bloom filter throughput + false positive This patch adds benchmark tests for the throughput (for lookups + updates) and the false positive rate of bloom filter lookups, as well as some minor refactoring of the bash script for running the benchmarks. These benchmarks show that as the number of hash functions increases, the throughput and the false positive rate of the bloom filter decreases. >From the benchmark data, the approximate average false-positive rates are roughly as follows: 1 hash function = ~30% 2 hash functions = ~15% 3 hash functions = ~5% 4 hash functions = ~2.5% 5 hash functions = ~1% 6 hash functions = ~0.5% 7 hash functions = ~0.35% 8 hash functions = ~0.15% 9 hash functions = ~0.1% 10 hash functions = ~0% For reference data, the benchmarks run on one thread on a machine with one numa node for 1 to 5 hash functions for 8-byte and 64-byte values are as follows: 1 hash function: 50k entries 8-byte value Lookups - 51.1 M/s operations Updates - 33.6 M/s operations False positive rate: 24.15% 64-byte value Lookups - 15.7 M/s operations Updates - 15.1 M/s operations False positive rate: 24.2% 100k entries 8-byte value Lookups - 51.0 M/s operations Updates - 33.4 M/s operations False positive rate: 24.04% 64-byte value Lookups - 15.6 M/s operations Updates - 14.6 M/s operations False positive rate: 24.06% 500k entries 8-byte value Lookups - 50.5 M/s operations Updates - 33.1 M/s operations False positive rate: 27.45% 64-byte value Lookups - 15.6 M/s operations Updates - 14.2 M/s operations False positive rate: 27.42% 1 mil entries 8-byte value Lookups - 49.7 M/s operations Updates - 32.9 M/s operations False positive rate: 27.45% 64-byte value Lookups - 15.4 M/s operations Updates - 13.7 M/s operations False positive rate: 27.58% 2.5 mil entries 8-byte value Lookups - 47.2 M/s operations Updates - 31.8 M/s operations False positive rate: 30.94% 64-byte value Lookups - 15.3 M/s operations Updates - 13.2 M/s operations False positive rate: 30.95% 5 mil entries 8-byte value Lookups - 41.1 M/s operations Updates - 28.1 M/s operations False positive rate: 31.01% 64-byte value Lookups - 13.3 M/s operations Updates - 11.4 M/s operations False positive rate: 30.98% 2 hash functions: 50k entries 8-byte value Lookups - 34.1 M/s operations Updates - 20.1 M/s operations False positive rate: 9.13% 64-byte value Lookups - 8.4 M/s operations Updates - 7.9 M/s operations False positive rate: 9.21% 100k entries 8-byte value Lookups - 33.7 M/s operations Updates - 18.9 M/s operations False positive rate: 9.13% 64-byte value Lookups - 8.4 M/s operations Updates - 7.7 M/s operations False positive rate: 9.19% 500k entries 8-byte value Lookups - 32.7 M/s operations Updates - 18.1 M/s operations False positive rate: 12.61% 64-byte value Lookups - 8.4 M/s operations Updates - 7.5 M/s operations False positive rate: 12.61% 1 mil entries 8-byte value Lookups - 30.6 M/s operations Updates - 18.9 M/s operations False positive rate: 12.54% 64-byte value Lookups - 8.0 M/s operations Updates - 7.0 M/s operations False positive rate: 12.52% 2.5 mil entries 8-byte value Lookups - 25.3 M/s operations Updates - 16.7 M/s operations False positive rate: 16.77% 64-byte value Lookups - 7.9 M/s operations Updates - 6.5 M/s operations False positive rate: 16.88% 5 mil entries 8-byte value Lookups - 20.8 M/s operations Updates - 14.7 M/s operations False positive rate: 16.78% 64-byte value Lookups - 7.0 M/s operations Updates - 6.0 M/s operations False positive rate: 16.78% 3 hash functions: 50k entries 8-byte value Lookups - 25.1 M/s operations Updates - 14.6 M/s operations False positive rate: 7.65% 64-byte value Lookups - 5.8 M/s operations Updates - 5.5 M/s operations False positive rate: 7.58% 100k entries 8-byte value Lookups - 24.7 M/s operations Updates - 14.1 M/s operations False positive rate: 7.71% 64-byte value Lookups - 5.8 M/s operations Updates - 5.3 M/s operations False positive rate: 7.62% 500k entries 8-byte value Lookups - 22.9 M/s operations Updates - 13.9 M/s operations False positive rate: 2.62% 64-byte value Lookups - 5.6 M/s operations Updates - 4.8 M/s operations False positive rate: 2.7% 1 mil entries 8-byte value Lookups - 19.8 M/s operations Updates - 12.6 M/s operations False positive rate: 2.60% 64-byte value Lookups - 5.3 M/s operations Updates - 4.4 M/s operations False positive rate: 2.69% 2.5 mil entries 8-byte value Lookups - 16.2 M/s operations Updates - 10.7 M/s operations False positive rate: 4.49% 64-byte value Lookups - 4.9 M/s operations Updates - 4.1 M/s operations False positive rate: 4.41% 5 mil entries 8-byte value Lookups - 18.8 M/s operations Updates - 9.2 M/s operations False positive rate: 4.45% 64-byte value Lookups - 5.2 M/s operations Updates - 3.9 M/s operations False positive rate: 4.54% 4 hash functions: 50k entries 8-byte value Lookups - 19.7 M/s operations Updates - 11.1 M/s operations False positive rate: 1.01% 64-byte value Lookups - 4.4 M/s operations Updates - 4.0 M/s operations False positive rate: 1.00% 100k entries 8-byte value Lookups - 19.5 M/s operations Updates - 10.9 M/s operations False positive rate: 1.00% 64-byte value Lookups - 4.3 M/s operations Updates - 3.9 M/s operations False positive rate: 0.97% 500k entries 8-byte value Lookups - 18.2 M/s operations Updates - 10.6 M/s operations False positive rate: 2.05% 64-byte value Lookups - 4.3 M/s operations Updates - 3.7 M/s operations False positive rate: 2.05% 1 mil entries 8-byte value Lookups - 15.5 M/s operations Updates - 9.6 M/s operations False positive rate: 1.99% 64-byte value Lookups - 4.0 M/s operations Updates - 3.4 M/s operations False positive rate: 1.99% 2.5 mil entries 8-byte value Lookups - 13.8 M/s operations Updates - 7.7 M/s operations False positive rate: 3.91% 64-byte value Lookups - 3.7 M/s operations Updates - 3.6 M/s operations False positive rate: 3.78% 5 mil entries 8-byte value Lookups - 13.0 M/s operations Updates - 6.9 M/s operations False positive rate: 3.93% 64-byte value Lookups - 3.5 M/s operations Updates - 3.7 M/s operations False positive rate: 3.39% 5 hash functions: 50k entries 8-byte value Lookups - 16.4 M/s operations Updates - 9.1 M/s operations False positive rate: 0.78% 64-byte value Lookups - 3.5 M/s operations Updates - 3.2 M/s operations False positive rate: 0.77% 100k entries 8-byte value Lookups - 16.3 M/s operations Updates - 9.0 M/s operations False positive rate: 0.79% 64-byte value Lookups - 3.5 M/s operations Updates - 3.2 M/s operations False positive rate: 0.78% 500k entries 8-byte value Lookups - 15.1 M/s operations Updates - 8.8 M/s operations False positive rate: 1.82% 64-byte value Lookups - 3.4 M/s operations Updates - 3.0 M/s operations False positive rate: 1.78% 1 mil entries 8-byte value Lookups - 13.2 M/s operations Updates - 7.8 M/s operations False positive rate: 1.81% 64-byte value Lookups - 3.2 M/s operations Updates - 2.8 M/s operations False positive rate: 1.80% 2.5 mil entries 8-byte value Lookups - 10.5 M/s operations Updates - 5.9 M/s operations False positive rate: 0.29% 64-byte value Lookups - 3.2 M/s operations Updates - 2.4 M/s operations False positive rate: 0.28% 5 mil entries 8-byte value Lookups - 9.6 M/s operations Updates - 5.7 M/s operations False positive rate: 0.30% 64-byte value Lookups - 3.2 M/s operations Updates - 2.7 M/s operations False positive rate: 0.30% Signed-off-by: Joanne Koong <joannekoong@fb.com> Signed-off-by: Alexei Starovoitov <ast@kernel.org> Acked-by: Andrii Nakryiko <andrii@kernel.org> Link: https://lore.kernel.org/bpf/20211027234504.30744-5-joannekoong@fb.com
2021-10-28 07:45:03 +08:00
{ &bench_bloom_map_argp, 0, "Bloom filter map benchmark", 0 },
selftest/bpf/benchs: Add bpf_loop benchmark Add benchmark to measure the throughput and latency of the bpf_loop call. Testing this on my dev machine on 1 thread, the data is as follows: nr_loops: 10 bpf_loop - throughput: 198.519 ± 0.155 M ops/s, latency: 5.037 ns/op nr_loops: 100 bpf_loop - throughput: 247.448 ± 0.305 M ops/s, latency: 4.041 ns/op nr_loops: 500 bpf_loop - throughput: 260.839 ± 0.380 M ops/s, latency: 3.834 ns/op nr_loops: 1000 bpf_loop - throughput: 262.806 ± 0.629 M ops/s, latency: 3.805 ns/op nr_loops: 5000 bpf_loop - throughput: 264.211 ± 1.508 M ops/s, latency: 3.785 ns/op nr_loops: 10000 bpf_loop - throughput: 265.366 ± 3.054 M ops/s, latency: 3.768 ns/op nr_loops: 50000 bpf_loop - throughput: 235.986 ± 20.205 M ops/s, latency: 4.238 ns/op nr_loops: 100000 bpf_loop - throughput: 264.482 ± 0.279 M ops/s, latency: 3.781 ns/op nr_loops: 500000 bpf_loop - throughput: 309.773 ± 87.713 M ops/s, latency: 3.228 ns/op nr_loops: 1000000 bpf_loop - throughput: 262.818 ± 4.143 M ops/s, latency: 3.805 ns/op >From this data, we can see that the latency per loop decreases as the number of loops increases. On this particular machine, each loop had an overhead of about ~4 ns, and we were able to run ~250 million loops per second. Signed-off-by: Joanne Koong <joannekoong@fb.com> Signed-off-by: Alexei Starovoitov <ast@kernel.org> Acked-by: Andrii Nakryiko <andrii@kernel.org> Link: https://lore.kernel.org/bpf/20211130030622.4131246-5-joannekoong@fb.com
2021-11-30 11:06:22 +08:00
{ &bench_bpf_loop_argp, 0, "bpf_loop helper benchmark", 0 },
selftests/bpf: Add benchmark for local_storage get Add a benchmarks to demonstrate the performance cliff for local_storage get as the number of local_storage maps increases beyond current local_storage implementation's cache size. "sequential get" and "interleaved get" benchmarks are added, both of which do many bpf_task_storage_get calls on sets of task local_storage maps of various counts, while considering a single specific map to be 'important' and counting task_storage_gets to the important map separately in addition to normal 'hits' count of all gets. Goal here is to mimic scenario where a particular program using one map - the important one - is running on a system where many other local_storage maps exist and are accessed often. While "sequential get" benchmark does bpf_task_storage_get for map 0, 1, ..., {9, 99, 999} in order, "interleaved" benchmark interleaves 4 bpf_task_storage_gets for the important map for every 10 map gets. This is meant to highlight performance differences when important map is accessed far more frequently than non-important maps. A "hashmap control" benchmark is also included for easy comparison of standard bpf hashmap lookup vs local_storage get. The benchmark is similar to "sequential get", but creates and uses BPF_MAP_TYPE_HASH instead of local storage. Only one inner map is created - a hashmap meant to hold tid -> data mapping for all tasks. Size of the hashmap is hardcoded to my system's PID_MAX_LIMIT (4,194,304). The number of these keys which are actually fetched as part of the benchmark is configurable. Addition of this benchmark is inspired by conversation with Alexei in a previous patchset's thread [0], which highlighted the need for such a benchmark to motivate and validate improvements to local_storage implementation. My approach in that series focused on improving performance for explicitly-marked 'important' maps and was rejected with feedback to make more generally-applicable improvements while avoiding explicitly marking maps as important. Thus the benchmark reports both general and important-map-focused metrics, so effect of future work on both is clear. Regarding the benchmark results. On a powerful system (Skylake, 20 cores, 256gb ram): Hashmap Control =============== num keys: 10 hashmap (control) sequential get: hits throughput: 20.900 ± 0.334 M ops/s, hits latency: 47.847 ns/op, important_hits throughput: 20.900 ± 0.334 M ops/s num keys: 1000 hashmap (control) sequential get: hits throughput: 13.758 ± 0.219 M ops/s, hits latency: 72.683 ns/op, important_hits throughput: 13.758 ± 0.219 M ops/s num keys: 10000 hashmap (control) sequential get: hits throughput: 6.995 ± 0.034 M ops/s, hits latency: 142.959 ns/op, important_hits throughput: 6.995 ± 0.034 M ops/s num keys: 100000 hashmap (control) sequential get: hits throughput: 4.452 ± 0.371 M ops/s, hits latency: 224.635 ns/op, important_hits throughput: 4.452 ± 0.371 M ops/s num keys: 4194304 hashmap (control) sequential get: hits throughput: 3.043 ± 0.033 M ops/s, hits latency: 328.587 ns/op, important_hits throughput: 3.043 ± 0.033 M ops/s Local Storage ============= num_maps: 1 local_storage cache sequential get: hits throughput: 47.298 ± 0.180 M ops/s, hits latency: 21.142 ns/op, important_hits throughput: 47.298 ± 0.180 M ops/s local_storage cache interleaved get: hits throughput: 55.277 ± 0.888 M ops/s, hits latency: 18.091 ns/op, important_hits throughput: 55.277 ± 0.888 M ops/s num_maps: 10 local_storage cache sequential get: hits throughput: 40.240 ± 0.802 M ops/s, hits latency: 24.851 ns/op, important_hits throughput: 4.024 ± 0.080 M ops/s local_storage cache interleaved get: hits throughput: 48.701 ± 0.722 M ops/s, hits latency: 20.533 ns/op, important_hits throughput: 17.393 ± 0.258 M ops/s num_maps: 16 local_storage cache sequential get: hits throughput: 44.515 ± 0.708 M ops/s, hits latency: 22.464 ns/op, important_hits throughput: 2.782 ± 0.044 M ops/s local_storage cache interleaved get: hits throughput: 49.553 ± 2.260 M ops/s, hits latency: 20.181 ns/op, important_hits throughput: 15.767 ± 0.719 M ops/s num_maps: 17 local_storage cache sequential get: hits throughput: 38.778 ± 0.302 M ops/s, hits latency: 25.788 ns/op, important_hits throughput: 2.284 ± 0.018 M ops/s local_storage cache interleaved get: hits throughput: 43.848 ± 1.023 M ops/s, hits latency: 22.806 ns/op, important_hits throughput: 13.349 ± 0.311 M ops/s num_maps: 24 local_storage cache sequential get: hits throughput: 19.317 ± 0.568 M ops/s, hits latency: 51.769 ns/op, important_hits throughput: 0.806 ± 0.024 M ops/s local_storage cache interleaved get: hits throughput: 24.397 ± 0.272 M ops/s, hits latency: 40.989 ns/op, important_hits throughput: 6.863 ± 0.077 M ops/s num_maps: 32 local_storage cache sequential get: hits throughput: 13.333 ± 0.135 M ops/s, hits latency: 75.000 ns/op, important_hits throughput: 0.417 ± 0.004 M ops/s local_storage cache interleaved get: hits throughput: 16.898 ± 0.383 M ops/s, hits latency: 59.178 ns/op, important_hits throughput: 4.717 ± 0.107 M ops/s num_maps: 100 local_storage cache sequential get: hits throughput: 6.360 ± 0.107 M ops/s, hits latency: 157.233 ns/op, important_hits throughput: 0.064 ± 0.001 M ops/s local_storage cache interleaved get: hits throughput: 7.303 ± 0.362 M ops/s, hits latency: 136.930 ns/op, important_hits throughput: 1.907 ± 0.094 M ops/s num_maps: 1000 local_storage cache sequential get: hits throughput: 0.452 ± 0.010 M ops/s, hits latency: 2214.022 ns/op, important_hits throughput: 0.000 ± 0.000 M ops/s local_storage cache interleaved get: hits throughput: 0.542 ± 0.007 M ops/s, hits latency: 1843.341 ns/op, important_hits throughput: 0.136 ± 0.002 M ops/s Looking at the "sequential get" results, it's clear that as the number of task local_storage maps grows beyond the current cache size (16), there's a significant reduction in hits throughput. Note that current local_storage implementation assigns a cache_idx to maps as they are created. Since "sequential get" is creating maps 0..n in order and then doing bpf_task_storage_get calls in the same order, the benchmark is effectively ensuring that a map will not be in cache when the program tries to access it. For "interleaved get" results, important-map hits throughput is greatly increased as the important map is more likely to be in cache by virtue of being accessed far more frequently. Throughput still reduces as # maps increases, though. To get a sense of the overhead of the benchmark program, I commented out bpf_task_storage_get/bpf_map_lookup_elem in local_storage_bench.c and ran the benchmark on the same host as the 'real' run. Results: Hashmap Control =============== num keys: 10 hashmap (control) sequential get: hits throughput: 54.288 ± 0.655 M ops/s, hits latency: 18.420 ns/op, important_hits throughput: 54.288 ± 0.655 M ops/s num keys: 1000 hashmap (control) sequential get: hits throughput: 52.913 ± 0.519 M ops/s, hits latency: 18.899 ns/op, important_hits throughput: 52.913 ± 0.519 M ops/s num keys: 10000 hashmap (control) sequential get: hits throughput: 53.480 ± 1.235 M ops/s, hits latency: 18.699 ns/op, important_hits throughput: 53.480 ± 1.235 M ops/s num keys: 100000 hashmap (control) sequential get: hits throughput: 54.982 ± 1.902 M ops/s, hits latency: 18.188 ns/op, important_hits throughput: 54.982 ± 1.902 M ops/s num keys: 4194304 hashmap (control) sequential get: hits throughput: 50.858 ± 0.707 M ops/s, hits latency: 19.662 ns/op, important_hits throughput: 50.858 ± 0.707 M ops/s Local Storage ============= num_maps: 1 local_storage cache sequential get: hits throughput: 110.990 ± 4.828 M ops/s, hits latency: 9.010 ns/op, important_hits throughput: 110.990 ± 4.828 M ops/s local_storage cache interleaved get: hits throughput: 161.057 ± 4.090 M ops/s, hits latency: 6.209 ns/op, important_hits throughput: 161.057 ± 4.090 M ops/s num_maps: 10 local_storage cache sequential get: hits throughput: 112.930 ± 1.079 M ops/s, hits latency: 8.855 ns/op, important_hits throughput: 11.293 ± 0.108 M ops/s local_storage cache interleaved get: hits throughput: 115.841 ± 2.088 M ops/s, hits latency: 8.633 ns/op, important_hits throughput: 41.372 ± 0.746 M ops/s num_maps: 16 local_storage cache sequential get: hits throughput: 115.653 ± 0.416 M ops/s, hits latency: 8.647 ns/op, important_hits throughput: 7.228 ± 0.026 M ops/s local_storage cache interleaved get: hits throughput: 138.717 ± 1.649 M ops/s, hits latency: 7.209 ns/op, important_hits throughput: 44.137 ± 0.525 M ops/s num_maps: 17 local_storage cache sequential get: hits throughput: 112.020 ± 1.649 M ops/s, hits latency: 8.927 ns/op, important_hits throughput: 6.598 ± 0.097 M ops/s local_storage cache interleaved get: hits throughput: 128.089 ± 1.960 M ops/s, hits latency: 7.807 ns/op, important_hits throughput: 38.995 ± 0.597 M ops/s num_maps: 24 local_storage cache sequential get: hits throughput: 92.447 ± 5.170 M ops/s, hits latency: 10.817 ns/op, important_hits throughput: 3.855 ± 0.216 M ops/s local_storage cache interleaved get: hits throughput: 128.844 ± 2.808 M ops/s, hits latency: 7.761 ns/op, important_hits throughput: 36.245 ± 0.790 M ops/s num_maps: 32 local_storage cache sequential get: hits throughput: 102.042 ± 1.462 M ops/s, hits latency: 9.800 ns/op, important_hits throughput: 3.194 ± 0.046 M ops/s local_storage cache interleaved get: hits throughput: 126.577 ± 1.818 M ops/s, hits latency: 7.900 ns/op, important_hits throughput: 35.332 ± 0.507 M ops/s num_maps: 100 local_storage cache sequential get: hits throughput: 111.327 ± 1.401 M ops/s, hits latency: 8.983 ns/op, important_hits throughput: 1.113 ± 0.014 M ops/s local_storage cache interleaved get: hits throughput: 131.327 ± 1.339 M ops/s, hits latency: 7.615 ns/op, important_hits throughput: 34.302 ± 0.350 M ops/s num_maps: 1000 local_storage cache sequential get: hits throughput: 101.978 ± 0.563 M ops/s, hits latency: 9.806 ns/op, important_hits throughput: 0.102 ± 0.001 M ops/s local_storage cache interleaved get: hits throughput: 141.084 ± 1.098 M ops/s, hits latency: 7.088 ns/op, important_hits throughput: 35.430 ± 0.276 M ops/s Adjusting for overhead, latency numbers for "hashmap control" and "sequential get" are: hashmap_control_1k: ~53.8ns hashmap_control_10k: ~124.2ns hashmap_control_100k: ~206.5ns sequential_get_1: ~12.1ns sequential_get_10: ~16.0ns sequential_get_16: ~13.8ns sequential_get_17: ~16.8ns sequential_get_24: ~40.9ns sequential_get_32: ~65.2ns sequential_get_100: ~148.2ns sequential_get_1000: ~2204ns Clearly demonstrating a cliff. In the discussion for v1 of this patch, Alexei noted that local_storage was 2.5x faster than a large hashmap when initially implemented [1]. The benchmark results show that local_storage is 5-10x faster: a long-running BPF application putting some pid-specific info into a hashmap for each pid it sees will probably see on the order of 10-100k pids. Bench numbers for hashmaps of this size are ~10x slower than sequential_get_16, but as the number of local_storage maps grows far past local_storage cache size the performance advantage shrinks and eventually reverses. When running the benchmarks it may be necessary to bump 'open files' ulimit for a successful run. [0]: https://lore.kernel.org/all/20220420002143.1096548-1-davemarchevsky@fb.com [1]: https://lore.kernel.org/bpf/20220511173305.ftldpn23m4ski3d3@MBP-98dd607d3435.dhcp.thefacebook.com/ Signed-off-by: Dave Marchevsky <davemarchevsky@fb.com> Link: https://lore.kernel.org/r/20220620222554.270578-1-davemarchevsky@fb.com Signed-off-by: Alexei Starovoitov <ast@kernel.org>
2022-06-21 06:25:54 +08:00
{ &bench_local_storage_argp, 0, "local_storage benchmark", 0 },
selftests/bpf: Add benchmark for bpf_strncmp() helper Add benchmark to compare the performance between home-made strncmp() in bpf program and bpf_strncmp() helper. In summary, the performance win of bpf_strncmp() under x86-64 is greater than 18% when the compared string length is greater than 64, and is 179% when the length is 4095. Under arm64 the performance win is even bigger: 33% when the length is greater than 64 and 600% when the length is 4095. The following is the details: no-helper-X: use home-made strncmp() to compare X-sized string helper-Y: use bpf_strncmp() to compare Y-sized string Under x86-64: no-helper-1 3.504 ± 0.000M/s (drops 0.000 ± 0.000M/s) helper-1 3.347 ± 0.001M/s (drops 0.000 ± 0.000M/s) no-helper-8 3.357 ± 0.001M/s (drops 0.000 ± 0.000M/s) helper-8 3.307 ± 0.001M/s (drops 0.000 ± 0.000M/s) no-helper-32 3.064 ± 0.000M/s (drops 0.000 ± 0.000M/s) helper-32 3.253 ± 0.001M/s (drops 0.000 ± 0.000M/s) no-helper-64 2.563 ± 0.001M/s (drops 0.000 ± 0.000M/s) helper-64 3.040 ± 0.001M/s (drops 0.000 ± 0.000M/s) no-helper-128 1.975 ± 0.000M/s (drops 0.000 ± 0.000M/s) helper-128 2.641 ± 0.000M/s (drops 0.000 ± 0.000M/s) no-helper-512 0.759 ± 0.000M/s (drops 0.000 ± 0.000M/s) helper-512 1.574 ± 0.000M/s (drops 0.000 ± 0.000M/s) no-helper-2048 0.329 ± 0.000M/s (drops 0.000 ± 0.000M/s) helper-2048 0.602 ± 0.000M/s (drops 0.000 ± 0.000M/s) no-helper-4095 0.117 ± 0.000M/s (drops 0.000 ± 0.000M/s) helper-4095 0.327 ± 0.000M/s (drops 0.000 ± 0.000M/s) Under arm64: no-helper-1 2.806 ± 0.004M/s (drops 0.000 ± 0.000M/s) helper-1 2.819 ± 0.002M/s (drops 0.000 ± 0.000M/s) no-helper-8 2.797 ± 0.109M/s (drops 0.000 ± 0.000M/s) helper-8 2.786 ± 0.025M/s (drops 0.000 ± 0.000M/s) no-helper-32 2.399 ± 0.011M/s (drops 0.000 ± 0.000M/s) helper-32 2.703 ± 0.002M/s (drops 0.000 ± 0.000M/s) no-helper-64 2.020 ± 0.015M/s (drops 0.000 ± 0.000M/s) helper-64 2.702 ± 0.073M/s (drops 0.000 ± 0.000M/s) no-helper-128 1.604 ± 0.001M/s (drops 0.000 ± 0.000M/s) helper-128 2.516 ± 0.002M/s (drops 0.000 ± 0.000M/s) no-helper-512 0.699 ± 0.000M/s (drops 0.000 ± 0.000M/s) helper-512 2.106 ± 0.003M/s (drops 0.000 ± 0.000M/s) no-helper-2048 0.215 ± 0.000M/s (drops 0.000 ± 0.000M/s) helper-2048 1.223 ± 0.003M/s (drops 0.000 ± 0.000M/s) no-helper-4095 0.112 ± 0.000M/s (drops 0.000 ± 0.000M/s) helper-4095 0.796 ± 0.000M/s (drops 0.000 ± 0.000M/s) Signed-off-by: Hou Tao <houtao1@huawei.com> Signed-off-by: Alexei Starovoitov <ast@kernel.org> Link: https://lore.kernel.org/bpf/20211210141652.877186-4-houtao1@huawei.com
2021-12-10 22:16:51 +08:00
{ &bench_strncmp_argp, 0, "bpf_strncmp helper benchmark", 0 },
selftests/bpf: Add benchmark for local_storage RCU Tasks Trace usage This benchmark measures grace period latency and kthread cpu usage of RCU Tasks Trace when many processes are creating/deleting BPF local_storage. Intent here is to quantify improvement on these metrics after Paul's recent RCU Tasks patches [0]. Specifically, fork 15k tasks which call a bpf prog that creates/destroys task local_storage and sleep in a loop, resulting in many call_rcu_tasks_trace calls. To determine grace period latency, trace time elapsed between rcu_tasks_trace_pregp_step and rcu_tasks_trace_postgp; for cpu usage look at rcu_task_trace_kthread's stime in /proc/PID/stat. On my virtualized test environment (Skylake, 8 cpus) benchmark results demonstrate significant improvement: BEFORE Paul's patches: SUMMARY tasks_trace grace period latency avg 22298.551 us stddev 1302.165 us SUMMARY ticks per tasks_trace grace period avg 2.291 stddev 0.324 AFTER Paul's patches: SUMMARY tasks_trace grace period latency avg 16969.197 us stddev 2525.053 us SUMMARY ticks per tasks_trace grace period avg 1.146 stddev 0.178 Note that since these patches are not in bpf-next benchmarking was done by cherry-picking this patch onto rcu tree. [0] https://lore.kernel.org/rcu/20220620225402.GA3842369@paulmck-ThinkPad-P17-Gen-1/ Signed-off-by: Dave Marchevsky <davemarchevsky@fb.com> Signed-off-by: Daniel Borkmann <daniel@iogearbox.net> Acked-by: Paul E. McKenney <paulmck@kernel.org> Acked-by: Martin KaFai Lau <kafai@fb.com> Link: https://lore.kernel.org/bpf/20220705190018.3239050-1-davemarchevsky@fb.com
2022-07-06 03:00:18 +08:00
{ &bench_local_storage_rcu_tasks_trace_argp, 0,
"local_storage RCU Tasks Trace slowdown benchmark", 0 },
bpf: Add BPF ringbuf and perf buffer benchmarks Extend bench framework with ability to have benchmark-provided child argument parser for custom benchmark-specific parameters. This makes bench generic code modular and independent from any specific benchmark. Also implement a set of benchmarks for new BPF ring buffer and existing perf buffer. 4 benchmarks were implemented: 2 variations for each of BPF ringbuf and perfbuf:, - rb-libbpf utilizes stock libbpf ring_buffer manager for reading data; - rb-custom implements custom ring buffer setup and reading code, to eliminate overheads inherent in generic libbpf code due to callback functions and the need to update consumer position after each consumed record, instead of batching updates (due to pessimistic assumption that user callback might take long time and thus could unnecessarily hold ring buffer space for too long); - pb-libbpf uses stock libbpf perf_buffer code with all the default settings, though uses higher-performance raw event callback to minimize unnecessary overhead; - pb-custom implements its own custom consumer code to minimize any possible overhead of generic libbpf implementation and indirect function calls. All of the test support default, no data notification skipped, mode, as well as sampled mode (with --rb-sampled flag), which allows to trigger epoll notification less frequently and reduce overhead. As will be shown, this mode is especially critical for perf buffer, which suffers from high overhead of wakeups in kernel. Otherwise, all benchamrks implement similar way to generate a batch of records by using fentry/sys_getpgid BPF program, which pushes a bunch of records in a tight loop and records number of successful and dropped samples. Each record is a small 8-byte integer, to minimize the effect of memory copying with bpf_perf_event_output() and bpf_ringbuf_output(). Benchmarks that have only one producer implement optional back-to-back mode, in which record production and consumption is alternating on the same CPU. This is the highest-throughput happy case, showing ultimate performance achievable with either BPF ringbuf or perfbuf. All the below scenarios are implemented in a script in benchs/run_bench_ringbufs.sh. Tests were performed on 28-core/56-thread Intel Xeon CPU E5-2680 v4 @ 2.40GHz CPU. Single-producer, parallel producer ================================== rb-libbpf 12.054 ± 0.320M/s (drops 0.000 ± 0.000M/s) rb-custom 8.158 ± 0.118M/s (drops 0.001 ± 0.003M/s) pb-libbpf 0.931 ± 0.007M/s (drops 0.000 ± 0.000M/s) pb-custom 0.965 ± 0.003M/s (drops 0.000 ± 0.000M/s) Single-producer, parallel producer, sampled notification ======================================================== rb-libbpf 11.563 ± 0.067M/s (drops 0.000 ± 0.000M/s) rb-custom 15.895 ± 0.076M/s (drops 0.000 ± 0.000M/s) pb-libbpf 9.889 ± 0.032M/s (drops 0.000 ± 0.000M/s) pb-custom 9.866 ± 0.028M/s (drops 0.000 ± 0.000M/s) Single producer on one CPU, consumer on another one, both running at full speed. Curiously, rb-libbpf has higher throughput than objectively faster (due to more lightweight consumer code path) rb-custom. It appears that faster consumer causes kernel to send notifications more frequently, because consumer appears to be caught up more frequently. Performance of perfbuf suffers from default "no sampling" policy and huge overhead that causes. In sampled mode, rb-custom is winning very significantly eliminating too frequent in-kernel wakeups, the gain appears to be more than 2x. Perf buffer achieves even more impressive wins, compared to stock perfbuf settings, with 10x improvements in throughput with 1:500 sampling rate. The trade-off is that with sampling, application might not get next X events until X+1st arrives, which is not always acceptable. With steady influx of events, though, this shouldn't be a problem. Overall, single-producer performance of ring buffers seems to be better no matter the sampled/non-sampled modes, but it especially beats ring buffer without sampling due to its adaptive notification approach. Single-producer, back-to-back mode ================================== rb-libbpf 15.507 ± 0.247M/s (drops 0.000 ± 0.000M/s) rb-libbpf-sampled 14.692 ± 0.195M/s (drops 0.000 ± 0.000M/s) rb-custom 21.449 ± 0.157M/s (drops 0.000 ± 0.000M/s) rb-custom-sampled 20.024 ± 0.386M/s (drops 0.000 ± 0.000M/s) pb-libbpf 1.601 ± 0.015M/s (drops 0.000 ± 0.000M/s) pb-libbpf-sampled 8.545 ± 0.064M/s (drops 0.000 ± 0.000M/s) pb-custom 1.607 ± 0.022M/s (drops 0.000 ± 0.000M/s) pb-custom-sampled 8.988 ± 0.144M/s (drops 0.000 ± 0.000M/s) Here we test a back-to-back mode, which is arguably best-case scenario both for BPF ringbuf and perfbuf, because there is no contention and for ringbuf also no excessive notification, because consumer appears to be behind after the first record. For ringbuf, custom consumer code clearly wins with 21.5 vs 16 million records per second exchanged between producer and consumer. Sampled mode actually hurts a bit due to slightly slower producer logic (it needs to fetch amount of data available to decide whether to skip or force notification). Perfbuf with wakeup sampling gets 5.5x throughput increase, compared to no-sampling version. There also doesn't seem to be noticeable overhead from generic libbpf handling code. Perfbuf back-to-back, effect of sample rate =========================================== pb-sampled-1 1.035 ± 0.012M/s (drops 0.000 ± 0.000M/s) pb-sampled-5 3.476 ± 0.087M/s (drops 0.000 ± 0.000M/s) pb-sampled-10 5.094 ± 0.136M/s (drops 0.000 ± 0.000M/s) pb-sampled-25 7.118 ± 0.153M/s (drops 0.000 ± 0.000M/s) pb-sampled-50 8.169 ± 0.156M/s (drops 0.000 ± 0.000M/s) pb-sampled-100 8.887 ± 0.136M/s (drops 0.000 ± 0.000M/s) pb-sampled-250 9.180 ± 0.209M/s (drops 0.000 ± 0.000M/s) pb-sampled-500 9.353 ± 0.281M/s (drops 0.000 ± 0.000M/s) pb-sampled-1000 9.411 ± 0.217M/s (drops 0.000 ± 0.000M/s) pb-sampled-2000 9.464 ± 0.167M/s (drops 0.000 ± 0.000M/s) pb-sampled-3000 9.575 ± 0.273M/s (drops 0.000 ± 0.000M/s) This benchmark shows the effect of event sampling for perfbuf. Back-to-back mode for highest throughput. Just doing every 5th record notification gives 3.5x speed up. 250-500 appears to be the point of diminishing return, with almost 9x speed up. Most benchmarks use 500 as the default sampling for pb-raw and pb-custom. Ringbuf back-to-back, effect of sample rate =========================================== rb-sampled-1 1.106 ± 0.010M/s (drops 0.000 ± 0.000M/s) rb-sampled-5 4.746 ± 0.149M/s (drops 0.000 ± 0.000M/s) rb-sampled-10 7.706 ± 0.164M/s (drops 0.000 ± 0.000M/s) rb-sampled-25 12.893 ± 0.273M/s (drops 0.000 ± 0.000M/s) rb-sampled-50 15.961 ± 0.361M/s (drops 0.000 ± 0.000M/s) rb-sampled-100 18.203 ± 0.445M/s (drops 0.000 ± 0.000M/s) rb-sampled-250 19.962 ± 0.786M/s (drops 0.000 ± 0.000M/s) rb-sampled-500 20.881 ± 0.551M/s (drops 0.000 ± 0.000M/s) rb-sampled-1000 21.317 ± 0.532M/s (drops 0.000 ± 0.000M/s) rb-sampled-2000 21.331 ± 0.535M/s (drops 0.000 ± 0.000M/s) rb-sampled-3000 21.688 ± 0.392M/s (drops 0.000 ± 0.000M/s) Similar benchmark for ring buffer also shows a great advantage (in terms of throughput) of skipping notifications. Skipping every 5th one gives 4x boost. Also similar to perfbuf case, 250-500 seems to be the point of diminishing returns, giving roughly 20x better results. Keep in mind, for this test, notifications are controlled manually with BPF_RB_NO_WAKEUP and BPF_RB_FORCE_WAKEUP. As can be seen from previous benchmarks, adaptive notifications based on consumer's positions provides same (or even slightly better due to simpler load generator on BPF side) benefits in favorable back-to-back scenario. Over zealous and fast consumer, which is almost always caught up, will make thoughput numbers smaller. That's the case when manual notification control might prove to be extremely beneficial. Ringbuf back-to-back, reserve+commit vs output ============================================== reserve 22.819 ± 0.503M/s (drops 0.000 ± 0.000M/s) output 18.906 ± 0.433M/s (drops 0.000 ± 0.000M/s) Ringbuf sampled, reserve+commit vs output ========================================= reserve-sampled 15.350 ± 0.132M/s (drops 0.000 ± 0.000M/s) output-sampled 14.195 ± 0.144M/s (drops 0.000 ± 0.000M/s) BPF ringbuf supports two sets of APIs with various usability and performance tradeoffs: bpf_ringbuf_reserve()+bpf_ringbuf_commit() vs bpf_ringbuf_output(). This benchmark clearly shows superiority of reserve+commit approach, despite using a small 8-byte record size. Single-producer, consumer/producer competing on the same CPU, low batch count ============================================================================= rb-libbpf 3.045 ± 0.020M/s (drops 3.536 ± 0.148M/s) rb-custom 3.055 ± 0.022M/s (drops 3.893 ± 0.066M/s) pb-libbpf 1.393 ± 0.024M/s (drops 0.000 ± 0.000M/s) pb-custom 1.407 ± 0.016M/s (drops 0.000 ± 0.000M/s) This benchmark shows one of the worst-case scenarios, in which producer and consumer do not coordinate *and* fight for the same CPU. No batch count and sampling settings were able to eliminate drops for ringbuffer, producer is just too fast for consumer to keep up. But ringbuf and perfbuf still able to pass through quite a lot of messages, which is more than enough for a lot of applications. Ringbuf, multi-producer contention ================================== rb-libbpf nr_prod 1 10.916 ± 0.399M/s (drops 0.000 ± 0.000M/s) rb-libbpf nr_prod 2 4.931 ± 0.030M/s (drops 0.000 ± 0.000M/s) rb-libbpf nr_prod 3 4.880 ± 0.006M/s (drops 0.000 ± 0.000M/s) rb-libbpf nr_prod 4 3.926 ± 0.004M/s (drops 0.000 ± 0.000M/s) rb-libbpf nr_prod 8 4.011 ± 0.004M/s (drops 0.000 ± 0.000M/s) rb-libbpf nr_prod 12 3.967 ± 0.016M/s (drops 0.000 ± 0.000M/s) rb-libbpf nr_prod 16 2.604 ± 0.030M/s (drops 0.001 ± 0.002M/s) rb-libbpf nr_prod 20 2.233 ± 0.003M/s (drops 0.000 ± 0.000M/s) rb-libbpf nr_prod 24 2.085 ± 0.015M/s (drops 0.000 ± 0.000M/s) rb-libbpf nr_prod 28 2.055 ± 0.004M/s (drops 0.000 ± 0.000M/s) rb-libbpf nr_prod 32 1.962 ± 0.004M/s (drops 0.000 ± 0.000M/s) rb-libbpf nr_prod 36 2.089 ± 0.005M/s (drops 0.000 ± 0.000M/s) rb-libbpf nr_prod 40 2.118 ± 0.006M/s (drops 0.000 ± 0.000M/s) rb-libbpf nr_prod 44 2.105 ± 0.004M/s (drops 0.000 ± 0.000M/s) rb-libbpf nr_prod 48 2.120 ± 0.058M/s (drops 0.000 ± 0.001M/s) rb-libbpf nr_prod 52 2.074 ± 0.024M/s (drops 0.007 ± 0.014M/s) Ringbuf uses a very short-duration spinlock during reservation phase, to check few invariants, increment producer count and set record header. This is the biggest point of contention for ringbuf implementation. This benchmark evaluates the effect of multiple competing writers on overall throughput of a single shared ringbuffer. Overall throughput drops almost 2x when going from single to two highly-contended producers, gradually dropping with additional competing producers. Performance drop stabilizes at around 20 producers and hovers around 2mln even with 50+ fighting producers, which is a 5x drop compared to non-contended case. Good kernel implementation in kernel helps maintain decent performance here. Note, that in the intended real-world scenarios, it's not expected to get even close to such a high levels of contention. But if contention will become a problem, there is always an option of sharding few ring buffers across a set of CPUs. Signed-off-by: Andrii Nakryiko <andriin@fb.com> Signed-off-by: Daniel Borkmann <daniel@iogearbox.net> Link: https://lore.kernel.org/bpf/20200529075424.3139988-5-andriin@fb.com Signed-off-by: Alexei Starovoitov <ast@kernel.org>
2020-05-29 15:54:23 +08:00
{},
};
selftests/bpf: Add benchmark runner infrastructure While working on BPF ringbuf implementation, testing, and benchmarking, I've developed a pretty generic and modular benchmark runner, which seems to be generically useful, as I've already used it for one more purpose (testing fastest way to trigger BPF program, to minimize overhead of in-kernel code). This patch adds generic part of benchmark runner and sets up Makefile for extending it with more sets of benchmarks. Benchmarker itself operates by spinning up specified number of producer and consumer threads, setting up interval timer sending SIGALARM signal to application once a second. Every second, current snapshot with hits/drops counters are collected and stored in an array. Drops are useful for producer/consumer benchmarks in which producer might overwhelm consumers. Once test finishes after given amount of warm-up and testing seconds, mean and stddev are calculated (ignoring warm-up results) and is printed out to stdout. This setup seems to give consistent and accurate results. To validate behavior, I added two atomic counting tests: global and local. For global one, all the producer threads are atomically incrementing same counter as fast as possible. This, of course, leads to huge drop of performance once there is more than one producer thread due to CPUs fighting for the same memory location. Local counting, on the other hand, maintains one counter per each producer thread, incremented independently. Once per second, all counters are read and added together to form final "counting throughput" measurement. As expected, such setup demonstrates linear scalability with number of producers (as long as there are enough physical CPU cores, of course). See example output below. Also, this setup can nicely demonstrate disastrous effects of false sharing, if care is not taken to take those per-producer counters apart into independent cache lines. Demo output shows global counter first with 1 producer, then with 4. Both total and per-producer performance significantly drop. The last run is local counter with 4 producers, demonstrating near-perfect scalability. $ ./bench -a -w1 -d2 -p1 count-global Setting up benchmark 'count-global'... Benchmark 'count-global' started. Iter 0 ( 24.822us): hits 148.179M/s (148.179M/prod), drops 0.000M/s Iter 1 ( 37.939us): hits 149.308M/s (149.308M/prod), drops 0.000M/s Iter 2 (-10.774us): hits 150.717M/s (150.717M/prod), drops 0.000M/s Iter 3 ( 3.807us): hits 151.435M/s (151.435M/prod), drops 0.000M/s Summary: hits 150.488 ± 1.079M/s (150.488M/prod), drops 0.000 ± 0.000M/s $ ./bench -a -w1 -d2 -p4 count-global Setting up benchmark 'count-global'... Benchmark 'count-global' started. Iter 0 ( 60.659us): hits 53.910M/s ( 13.477M/prod), drops 0.000M/s Iter 1 (-17.658us): hits 53.722M/s ( 13.431M/prod), drops 0.000M/s Iter 2 ( 5.865us): hits 53.495M/s ( 13.374M/prod), drops 0.000M/s Iter 3 ( 0.104us): hits 53.606M/s ( 13.402M/prod), drops 0.000M/s Summary: hits 53.608 ± 0.113M/s ( 13.402M/prod), drops 0.000 ± 0.000M/s $ ./bench -a -w1 -d2 -p4 count-local Setting up benchmark 'count-local'... Benchmark 'count-local' started. Iter 0 ( 23.388us): hits 640.450M/s (160.113M/prod), drops 0.000M/s Iter 1 ( 2.291us): hits 605.661M/s (151.415M/prod), drops 0.000M/s Iter 2 ( -6.415us): hits 607.092M/s (151.773M/prod), drops 0.000M/s Iter 3 ( -1.361us): hits 601.796M/s (150.449M/prod), drops 0.000M/s Summary: hits 604.849 ± 2.739M/s (151.212M/prod), drops 0.000 ± 0.000M/s Benchmark runner supports setting thread affinity for producer and consumer threads. You can use -a flag for default CPU selection scheme, where first consumer gets CPU #0, next one gets CPU #1, and so on. Then producer threads pick up next CPU and increment one-by-one as well. But user can also specify a set of CPUs independently for producers and consumers with --prod-affinity 1,2-10,15 and --cons-affinity <set-of-cpus>. The latter allows to force producers and consumers to share same set of CPUs, if necessary. Signed-off-by: Andrii Nakryiko <andriin@fb.com> Signed-off-by: Alexei Starovoitov <ast@kernel.org> Acked-by: Yonghong Song <yhs@fb.com> Link: https://lore.kernel.org/bpf/20200512192445.2351848-3-andriin@fb.com
2020-05-13 03:24:43 +08:00
static error_t parse_arg(int key, char *arg, struct argp_state *state)
{
static int pos_args;
switch (key) {
case 'v':
env.verbose = true;
break;
case 'l':
env.list = true;
break;
case 'd':
env.duration_sec = strtol(arg, NULL, 10);
if (env.duration_sec <= 0) {
fprintf(stderr, "Invalid duration: %s\n", arg);
argp_usage(state);
}
break;
case 'w':
env.warmup_sec = strtol(arg, NULL, 10);
if (env.warmup_sec <= 0) {
fprintf(stderr, "Invalid warm-up duration: %s\n", arg);
argp_usage(state);
}
break;
case 'p':
env.producer_cnt = strtol(arg, NULL, 10);
if (env.producer_cnt <= 0) {
fprintf(stderr, "Invalid producer count: %s\n", arg);
argp_usage(state);
}
break;
case 'c':
env.consumer_cnt = strtol(arg, NULL, 10);
if (env.consumer_cnt <= 0) {
fprintf(stderr, "Invalid consumer count: %s\n", arg);
argp_usage(state);
}
break;
case 'a':
env.affinity = true;
break;
case ARG_PROD_AFFINITY_SET:
env.affinity = true;
if (parse_num_list(arg, &env.prod_cpus.cpus,
&env.prod_cpus.cpus_len)) {
fprintf(stderr, "Invalid format of CPU set for producers.");
argp_usage(state);
}
break;
case ARG_CONS_AFFINITY_SET:
env.affinity = true;
if (parse_num_list(arg, &env.cons_cpus.cpus,
&env.cons_cpus.cpus_len)) {
fprintf(stderr, "Invalid format of CPU set for consumers.");
argp_usage(state);
}
break;
case ARGP_KEY_ARG:
if (pos_args++) {
fprintf(stderr,
"Unrecognized positional argument: %s\n", arg);
argp_usage(state);
}
env.bench_name = strdup(arg);
break;
default:
return ARGP_ERR_UNKNOWN;
}
return 0;
}
static void parse_cmdline_args(int argc, char **argv)
{
static const struct argp argp = {
.options = opts,
.parser = parse_arg,
.doc = argp_program_doc,
bpf: Add BPF ringbuf and perf buffer benchmarks Extend bench framework with ability to have benchmark-provided child argument parser for custom benchmark-specific parameters. This makes bench generic code modular and independent from any specific benchmark. Also implement a set of benchmarks for new BPF ring buffer and existing perf buffer. 4 benchmarks were implemented: 2 variations for each of BPF ringbuf and perfbuf:, - rb-libbpf utilizes stock libbpf ring_buffer manager for reading data; - rb-custom implements custom ring buffer setup and reading code, to eliminate overheads inherent in generic libbpf code due to callback functions and the need to update consumer position after each consumed record, instead of batching updates (due to pessimistic assumption that user callback might take long time and thus could unnecessarily hold ring buffer space for too long); - pb-libbpf uses stock libbpf perf_buffer code with all the default settings, though uses higher-performance raw event callback to minimize unnecessary overhead; - pb-custom implements its own custom consumer code to minimize any possible overhead of generic libbpf implementation and indirect function calls. All of the test support default, no data notification skipped, mode, as well as sampled mode (with --rb-sampled flag), which allows to trigger epoll notification less frequently and reduce overhead. As will be shown, this mode is especially critical for perf buffer, which suffers from high overhead of wakeups in kernel. Otherwise, all benchamrks implement similar way to generate a batch of records by using fentry/sys_getpgid BPF program, which pushes a bunch of records in a tight loop and records number of successful and dropped samples. Each record is a small 8-byte integer, to minimize the effect of memory copying with bpf_perf_event_output() and bpf_ringbuf_output(). Benchmarks that have only one producer implement optional back-to-back mode, in which record production and consumption is alternating on the same CPU. This is the highest-throughput happy case, showing ultimate performance achievable with either BPF ringbuf or perfbuf. All the below scenarios are implemented in a script in benchs/run_bench_ringbufs.sh. Tests were performed on 28-core/56-thread Intel Xeon CPU E5-2680 v4 @ 2.40GHz CPU. Single-producer, parallel producer ================================== rb-libbpf 12.054 ± 0.320M/s (drops 0.000 ± 0.000M/s) rb-custom 8.158 ± 0.118M/s (drops 0.001 ± 0.003M/s) pb-libbpf 0.931 ± 0.007M/s (drops 0.000 ± 0.000M/s) pb-custom 0.965 ± 0.003M/s (drops 0.000 ± 0.000M/s) Single-producer, parallel producer, sampled notification ======================================================== rb-libbpf 11.563 ± 0.067M/s (drops 0.000 ± 0.000M/s) rb-custom 15.895 ± 0.076M/s (drops 0.000 ± 0.000M/s) pb-libbpf 9.889 ± 0.032M/s (drops 0.000 ± 0.000M/s) pb-custom 9.866 ± 0.028M/s (drops 0.000 ± 0.000M/s) Single producer on one CPU, consumer on another one, both running at full speed. Curiously, rb-libbpf has higher throughput than objectively faster (due to more lightweight consumer code path) rb-custom. It appears that faster consumer causes kernel to send notifications more frequently, because consumer appears to be caught up more frequently. Performance of perfbuf suffers from default "no sampling" policy and huge overhead that causes. In sampled mode, rb-custom is winning very significantly eliminating too frequent in-kernel wakeups, the gain appears to be more than 2x. Perf buffer achieves even more impressive wins, compared to stock perfbuf settings, with 10x improvements in throughput with 1:500 sampling rate. The trade-off is that with sampling, application might not get next X events until X+1st arrives, which is not always acceptable. With steady influx of events, though, this shouldn't be a problem. Overall, single-producer performance of ring buffers seems to be better no matter the sampled/non-sampled modes, but it especially beats ring buffer without sampling due to its adaptive notification approach. Single-producer, back-to-back mode ================================== rb-libbpf 15.507 ± 0.247M/s (drops 0.000 ± 0.000M/s) rb-libbpf-sampled 14.692 ± 0.195M/s (drops 0.000 ± 0.000M/s) rb-custom 21.449 ± 0.157M/s (drops 0.000 ± 0.000M/s) rb-custom-sampled 20.024 ± 0.386M/s (drops 0.000 ± 0.000M/s) pb-libbpf 1.601 ± 0.015M/s (drops 0.000 ± 0.000M/s) pb-libbpf-sampled 8.545 ± 0.064M/s (drops 0.000 ± 0.000M/s) pb-custom 1.607 ± 0.022M/s (drops 0.000 ± 0.000M/s) pb-custom-sampled 8.988 ± 0.144M/s (drops 0.000 ± 0.000M/s) Here we test a back-to-back mode, which is arguably best-case scenario both for BPF ringbuf and perfbuf, because there is no contention and for ringbuf also no excessive notification, because consumer appears to be behind after the first record. For ringbuf, custom consumer code clearly wins with 21.5 vs 16 million records per second exchanged between producer and consumer. Sampled mode actually hurts a bit due to slightly slower producer logic (it needs to fetch amount of data available to decide whether to skip or force notification). Perfbuf with wakeup sampling gets 5.5x throughput increase, compared to no-sampling version. There also doesn't seem to be noticeable overhead from generic libbpf handling code. Perfbuf back-to-back, effect of sample rate =========================================== pb-sampled-1 1.035 ± 0.012M/s (drops 0.000 ± 0.000M/s) pb-sampled-5 3.476 ± 0.087M/s (drops 0.000 ± 0.000M/s) pb-sampled-10 5.094 ± 0.136M/s (drops 0.000 ± 0.000M/s) pb-sampled-25 7.118 ± 0.153M/s (drops 0.000 ± 0.000M/s) pb-sampled-50 8.169 ± 0.156M/s (drops 0.000 ± 0.000M/s) pb-sampled-100 8.887 ± 0.136M/s (drops 0.000 ± 0.000M/s) pb-sampled-250 9.180 ± 0.209M/s (drops 0.000 ± 0.000M/s) pb-sampled-500 9.353 ± 0.281M/s (drops 0.000 ± 0.000M/s) pb-sampled-1000 9.411 ± 0.217M/s (drops 0.000 ± 0.000M/s) pb-sampled-2000 9.464 ± 0.167M/s (drops 0.000 ± 0.000M/s) pb-sampled-3000 9.575 ± 0.273M/s (drops 0.000 ± 0.000M/s) This benchmark shows the effect of event sampling for perfbuf. Back-to-back mode for highest throughput. Just doing every 5th record notification gives 3.5x speed up. 250-500 appears to be the point of diminishing return, with almost 9x speed up. Most benchmarks use 500 as the default sampling for pb-raw and pb-custom. Ringbuf back-to-back, effect of sample rate =========================================== rb-sampled-1 1.106 ± 0.010M/s (drops 0.000 ± 0.000M/s) rb-sampled-5 4.746 ± 0.149M/s (drops 0.000 ± 0.000M/s) rb-sampled-10 7.706 ± 0.164M/s (drops 0.000 ± 0.000M/s) rb-sampled-25 12.893 ± 0.273M/s (drops 0.000 ± 0.000M/s) rb-sampled-50 15.961 ± 0.361M/s (drops 0.000 ± 0.000M/s) rb-sampled-100 18.203 ± 0.445M/s (drops 0.000 ± 0.000M/s) rb-sampled-250 19.962 ± 0.786M/s (drops 0.000 ± 0.000M/s) rb-sampled-500 20.881 ± 0.551M/s (drops 0.000 ± 0.000M/s) rb-sampled-1000 21.317 ± 0.532M/s (drops 0.000 ± 0.000M/s) rb-sampled-2000 21.331 ± 0.535M/s (drops 0.000 ± 0.000M/s) rb-sampled-3000 21.688 ± 0.392M/s (drops 0.000 ± 0.000M/s) Similar benchmark for ring buffer also shows a great advantage (in terms of throughput) of skipping notifications. Skipping every 5th one gives 4x boost. Also similar to perfbuf case, 250-500 seems to be the point of diminishing returns, giving roughly 20x better results. Keep in mind, for this test, notifications are controlled manually with BPF_RB_NO_WAKEUP and BPF_RB_FORCE_WAKEUP. As can be seen from previous benchmarks, adaptive notifications based on consumer's positions provides same (or even slightly better due to simpler load generator on BPF side) benefits in favorable back-to-back scenario. Over zealous and fast consumer, which is almost always caught up, will make thoughput numbers smaller. That's the case when manual notification control might prove to be extremely beneficial. Ringbuf back-to-back, reserve+commit vs output ============================================== reserve 22.819 ± 0.503M/s (drops 0.000 ± 0.000M/s) output 18.906 ± 0.433M/s (drops 0.000 ± 0.000M/s) Ringbuf sampled, reserve+commit vs output ========================================= reserve-sampled 15.350 ± 0.132M/s (drops 0.000 ± 0.000M/s) output-sampled 14.195 ± 0.144M/s (drops 0.000 ± 0.000M/s) BPF ringbuf supports two sets of APIs with various usability and performance tradeoffs: bpf_ringbuf_reserve()+bpf_ringbuf_commit() vs bpf_ringbuf_output(). This benchmark clearly shows superiority of reserve+commit approach, despite using a small 8-byte record size. Single-producer, consumer/producer competing on the same CPU, low batch count ============================================================================= rb-libbpf 3.045 ± 0.020M/s (drops 3.536 ± 0.148M/s) rb-custom 3.055 ± 0.022M/s (drops 3.893 ± 0.066M/s) pb-libbpf 1.393 ± 0.024M/s (drops 0.000 ± 0.000M/s) pb-custom 1.407 ± 0.016M/s (drops 0.000 ± 0.000M/s) This benchmark shows one of the worst-case scenarios, in which producer and consumer do not coordinate *and* fight for the same CPU. No batch count and sampling settings were able to eliminate drops for ringbuffer, producer is just too fast for consumer to keep up. But ringbuf and perfbuf still able to pass through quite a lot of messages, which is more than enough for a lot of applications. Ringbuf, multi-producer contention ================================== rb-libbpf nr_prod 1 10.916 ± 0.399M/s (drops 0.000 ± 0.000M/s) rb-libbpf nr_prod 2 4.931 ± 0.030M/s (drops 0.000 ± 0.000M/s) rb-libbpf nr_prod 3 4.880 ± 0.006M/s (drops 0.000 ± 0.000M/s) rb-libbpf nr_prod 4 3.926 ± 0.004M/s (drops 0.000 ± 0.000M/s) rb-libbpf nr_prod 8 4.011 ± 0.004M/s (drops 0.000 ± 0.000M/s) rb-libbpf nr_prod 12 3.967 ± 0.016M/s (drops 0.000 ± 0.000M/s) rb-libbpf nr_prod 16 2.604 ± 0.030M/s (drops 0.001 ± 0.002M/s) rb-libbpf nr_prod 20 2.233 ± 0.003M/s (drops 0.000 ± 0.000M/s) rb-libbpf nr_prod 24 2.085 ± 0.015M/s (drops 0.000 ± 0.000M/s) rb-libbpf nr_prod 28 2.055 ± 0.004M/s (drops 0.000 ± 0.000M/s) rb-libbpf nr_prod 32 1.962 ± 0.004M/s (drops 0.000 ± 0.000M/s) rb-libbpf nr_prod 36 2.089 ± 0.005M/s (drops 0.000 ± 0.000M/s) rb-libbpf nr_prod 40 2.118 ± 0.006M/s (drops 0.000 ± 0.000M/s) rb-libbpf nr_prod 44 2.105 ± 0.004M/s (drops 0.000 ± 0.000M/s) rb-libbpf nr_prod 48 2.120 ± 0.058M/s (drops 0.000 ± 0.001M/s) rb-libbpf nr_prod 52 2.074 ± 0.024M/s (drops 0.007 ± 0.014M/s) Ringbuf uses a very short-duration spinlock during reservation phase, to check few invariants, increment producer count and set record header. This is the biggest point of contention for ringbuf implementation. This benchmark evaluates the effect of multiple competing writers on overall throughput of a single shared ringbuffer. Overall throughput drops almost 2x when going from single to two highly-contended producers, gradually dropping with additional competing producers. Performance drop stabilizes at around 20 producers and hovers around 2mln even with 50+ fighting producers, which is a 5x drop compared to non-contended case. Good kernel implementation in kernel helps maintain decent performance here. Note, that in the intended real-world scenarios, it's not expected to get even close to such a high levels of contention. But if contention will become a problem, there is always an option of sharding few ring buffers across a set of CPUs. Signed-off-by: Andrii Nakryiko <andriin@fb.com> Signed-off-by: Daniel Borkmann <daniel@iogearbox.net> Link: https://lore.kernel.org/bpf/20200529075424.3139988-5-andriin@fb.com Signed-off-by: Alexei Starovoitov <ast@kernel.org>
2020-05-29 15:54:23 +08:00
.children = bench_parsers,
selftests/bpf: Add benchmark runner infrastructure While working on BPF ringbuf implementation, testing, and benchmarking, I've developed a pretty generic and modular benchmark runner, which seems to be generically useful, as I've already used it for one more purpose (testing fastest way to trigger BPF program, to minimize overhead of in-kernel code). This patch adds generic part of benchmark runner and sets up Makefile for extending it with more sets of benchmarks. Benchmarker itself operates by spinning up specified number of producer and consumer threads, setting up interval timer sending SIGALARM signal to application once a second. Every second, current snapshot with hits/drops counters are collected and stored in an array. Drops are useful for producer/consumer benchmarks in which producer might overwhelm consumers. Once test finishes after given amount of warm-up and testing seconds, mean and stddev are calculated (ignoring warm-up results) and is printed out to stdout. This setup seems to give consistent and accurate results. To validate behavior, I added two atomic counting tests: global and local. For global one, all the producer threads are atomically incrementing same counter as fast as possible. This, of course, leads to huge drop of performance once there is more than one producer thread due to CPUs fighting for the same memory location. Local counting, on the other hand, maintains one counter per each producer thread, incremented independently. Once per second, all counters are read and added together to form final "counting throughput" measurement. As expected, such setup demonstrates linear scalability with number of producers (as long as there are enough physical CPU cores, of course). See example output below. Also, this setup can nicely demonstrate disastrous effects of false sharing, if care is not taken to take those per-producer counters apart into independent cache lines. Demo output shows global counter first with 1 producer, then with 4. Both total and per-producer performance significantly drop. The last run is local counter with 4 producers, demonstrating near-perfect scalability. $ ./bench -a -w1 -d2 -p1 count-global Setting up benchmark 'count-global'... Benchmark 'count-global' started. Iter 0 ( 24.822us): hits 148.179M/s (148.179M/prod), drops 0.000M/s Iter 1 ( 37.939us): hits 149.308M/s (149.308M/prod), drops 0.000M/s Iter 2 (-10.774us): hits 150.717M/s (150.717M/prod), drops 0.000M/s Iter 3 ( 3.807us): hits 151.435M/s (151.435M/prod), drops 0.000M/s Summary: hits 150.488 ± 1.079M/s (150.488M/prod), drops 0.000 ± 0.000M/s $ ./bench -a -w1 -d2 -p4 count-global Setting up benchmark 'count-global'... Benchmark 'count-global' started. Iter 0 ( 60.659us): hits 53.910M/s ( 13.477M/prod), drops 0.000M/s Iter 1 (-17.658us): hits 53.722M/s ( 13.431M/prod), drops 0.000M/s Iter 2 ( 5.865us): hits 53.495M/s ( 13.374M/prod), drops 0.000M/s Iter 3 ( 0.104us): hits 53.606M/s ( 13.402M/prod), drops 0.000M/s Summary: hits 53.608 ± 0.113M/s ( 13.402M/prod), drops 0.000 ± 0.000M/s $ ./bench -a -w1 -d2 -p4 count-local Setting up benchmark 'count-local'... Benchmark 'count-local' started. Iter 0 ( 23.388us): hits 640.450M/s (160.113M/prod), drops 0.000M/s Iter 1 ( 2.291us): hits 605.661M/s (151.415M/prod), drops 0.000M/s Iter 2 ( -6.415us): hits 607.092M/s (151.773M/prod), drops 0.000M/s Iter 3 ( -1.361us): hits 601.796M/s (150.449M/prod), drops 0.000M/s Summary: hits 604.849 ± 2.739M/s (151.212M/prod), drops 0.000 ± 0.000M/s Benchmark runner supports setting thread affinity for producer and consumer threads. You can use -a flag for default CPU selection scheme, where first consumer gets CPU #0, next one gets CPU #1, and so on. Then producer threads pick up next CPU and increment one-by-one as well. But user can also specify a set of CPUs independently for producers and consumers with --prod-affinity 1,2-10,15 and --cons-affinity <set-of-cpus>. The latter allows to force producers and consumers to share same set of CPUs, if necessary. Signed-off-by: Andrii Nakryiko <andriin@fb.com> Signed-off-by: Alexei Starovoitov <ast@kernel.org> Acked-by: Yonghong Song <yhs@fb.com> Link: https://lore.kernel.org/bpf/20200512192445.2351848-3-andriin@fb.com
2020-05-13 03:24:43 +08:00
};
if (argp_parse(&argp, argc, argv, 0, NULL, NULL))
exit(1);
if (!env.list && !env.bench_name) {
argp_help(&argp, stderr, ARGP_HELP_DOC, "bench");
exit(1);
}
}
static void collect_measurements(long delta_ns);
static __u64 last_time_ns;
static void sigalarm_handler(int signo)
{
long new_time_ns = get_time_ns();
long delta_ns = new_time_ns - last_time_ns;
collect_measurements(delta_ns);
last_time_ns = new_time_ns;
}
/* set up periodic 1-second timer */
static void setup_timer()
{
static struct sigaction sigalarm_action = {
.sa_handler = sigalarm_handler,
};
struct itimerval timer_settings = {};
int err;
last_time_ns = get_time_ns();
err = sigaction(SIGALRM, &sigalarm_action, NULL);
if (err < 0) {
selftest/bpf: Fix spelling mistake "SIGALARM" -> "SIGALRM" There is a spelling mistake in an error message, fix it. Signed-off-by: Colin Ian King <colin.king@canonical.com> Signed-off-by: Alexei Starovoitov <ast@kernel.org> Acked-by: Yonghong Song <yhs@fb.com> Link: https://lore.kernel.org/bpf/20200514121529.259668-1-colin.king@canonical.com
2020-05-14 20:15:29 +08:00
fprintf(stderr, "failed to install SIGALRM handler: %d\n", -errno);
selftests/bpf: Add benchmark runner infrastructure While working on BPF ringbuf implementation, testing, and benchmarking, I've developed a pretty generic and modular benchmark runner, which seems to be generically useful, as I've already used it for one more purpose (testing fastest way to trigger BPF program, to minimize overhead of in-kernel code). This patch adds generic part of benchmark runner and sets up Makefile for extending it with more sets of benchmarks. Benchmarker itself operates by spinning up specified number of producer and consumer threads, setting up interval timer sending SIGALARM signal to application once a second. Every second, current snapshot with hits/drops counters are collected and stored in an array. Drops are useful for producer/consumer benchmarks in which producer might overwhelm consumers. Once test finishes after given amount of warm-up and testing seconds, mean and stddev are calculated (ignoring warm-up results) and is printed out to stdout. This setup seems to give consistent and accurate results. To validate behavior, I added two atomic counting tests: global and local. For global one, all the producer threads are atomically incrementing same counter as fast as possible. This, of course, leads to huge drop of performance once there is more than one producer thread due to CPUs fighting for the same memory location. Local counting, on the other hand, maintains one counter per each producer thread, incremented independently. Once per second, all counters are read and added together to form final "counting throughput" measurement. As expected, such setup demonstrates linear scalability with number of producers (as long as there are enough physical CPU cores, of course). See example output below. Also, this setup can nicely demonstrate disastrous effects of false sharing, if care is not taken to take those per-producer counters apart into independent cache lines. Demo output shows global counter first with 1 producer, then with 4. Both total and per-producer performance significantly drop. The last run is local counter with 4 producers, demonstrating near-perfect scalability. $ ./bench -a -w1 -d2 -p1 count-global Setting up benchmark 'count-global'... Benchmark 'count-global' started. Iter 0 ( 24.822us): hits 148.179M/s (148.179M/prod), drops 0.000M/s Iter 1 ( 37.939us): hits 149.308M/s (149.308M/prod), drops 0.000M/s Iter 2 (-10.774us): hits 150.717M/s (150.717M/prod), drops 0.000M/s Iter 3 ( 3.807us): hits 151.435M/s (151.435M/prod), drops 0.000M/s Summary: hits 150.488 ± 1.079M/s (150.488M/prod), drops 0.000 ± 0.000M/s $ ./bench -a -w1 -d2 -p4 count-global Setting up benchmark 'count-global'... Benchmark 'count-global' started. Iter 0 ( 60.659us): hits 53.910M/s ( 13.477M/prod), drops 0.000M/s Iter 1 (-17.658us): hits 53.722M/s ( 13.431M/prod), drops 0.000M/s Iter 2 ( 5.865us): hits 53.495M/s ( 13.374M/prod), drops 0.000M/s Iter 3 ( 0.104us): hits 53.606M/s ( 13.402M/prod), drops 0.000M/s Summary: hits 53.608 ± 0.113M/s ( 13.402M/prod), drops 0.000 ± 0.000M/s $ ./bench -a -w1 -d2 -p4 count-local Setting up benchmark 'count-local'... Benchmark 'count-local' started. Iter 0 ( 23.388us): hits 640.450M/s (160.113M/prod), drops 0.000M/s Iter 1 ( 2.291us): hits 605.661M/s (151.415M/prod), drops 0.000M/s Iter 2 ( -6.415us): hits 607.092M/s (151.773M/prod), drops 0.000M/s Iter 3 ( -1.361us): hits 601.796M/s (150.449M/prod), drops 0.000M/s Summary: hits 604.849 ± 2.739M/s (151.212M/prod), drops 0.000 ± 0.000M/s Benchmark runner supports setting thread affinity for producer and consumer threads. You can use -a flag for default CPU selection scheme, where first consumer gets CPU #0, next one gets CPU #1, and so on. Then producer threads pick up next CPU and increment one-by-one as well. But user can also specify a set of CPUs independently for producers and consumers with --prod-affinity 1,2-10,15 and --cons-affinity <set-of-cpus>. The latter allows to force producers and consumers to share same set of CPUs, if necessary. Signed-off-by: Andrii Nakryiko <andriin@fb.com> Signed-off-by: Alexei Starovoitov <ast@kernel.org> Acked-by: Yonghong Song <yhs@fb.com> Link: https://lore.kernel.org/bpf/20200512192445.2351848-3-andriin@fb.com
2020-05-13 03:24:43 +08:00
exit(1);
}
timer_settings.it_interval.tv_sec = 1;
timer_settings.it_value.tv_sec = 1;
err = setitimer(ITIMER_REAL, &timer_settings, NULL);
if (err < 0) {
fprintf(stderr, "failed to arm interval timer: %d\n", -errno);
exit(1);
}
}
static void set_thread_affinity(pthread_t thread, int cpu)
{
cpu_set_t cpuset;
CPU_ZERO(&cpuset);
CPU_SET(cpu, &cpuset);
if (pthread_setaffinity_np(thread, sizeof(cpuset), &cpuset)) {
fprintf(stderr, "setting affinity to CPU #%d failed: %d\n",
cpu, errno);
exit(1);
}
}
static int next_cpu(struct cpu_set *cpu_set)
{
if (cpu_set->cpus) {
int i;
/* find next available CPU */
for (i = cpu_set->next_cpu; i < cpu_set->cpus_len; i++) {
if (cpu_set->cpus[i]) {
cpu_set->next_cpu = i + 1;
return i;
}
}
fprintf(stderr, "Not enough CPUs specified, need CPU #%d or higher.\n", i);
exit(1);
}
return cpu_set->next_cpu++;
}
static struct bench_state {
int res_cnt;
struct bench_res *results;
pthread_t *consumers;
pthread_t *producers;
} state;
const struct bench *bench = NULL;
extern const struct bench bench_count_global;
extern const struct bench bench_count_local;
selftest/bpf: Fmod_ret prog and implement test_overhead as part of bench Add fmod_ret BPF program to existing test_overhead selftest. Also re-implement user-space benchmarking part into benchmark runner to compare results. Results with ./bench are consistently somewhat lower than test_overhead's, but relative performance of various types of BPF programs stay consisten (e.g., kretprobe is noticeably slower). This slowdown seems to be coming from the fact that test_overhead is single-threaded, while benchmark always spins off at least one thread for producer. This has been confirmed by hacking multi-threaded test_overhead variant and also single-threaded bench variant. Resutls are below. run_bench_rename.sh script from benchs/ subdirectory was used to produce results for ./bench. Single-threaded implementations =============================== /* bench: single-threaded, atomics */ base : 4.622 ± 0.049M/s kprobe : 3.673 ± 0.052M/s kretprobe : 2.625 ± 0.052M/s rawtp : 4.369 ± 0.089M/s fentry : 4.201 ± 0.558M/s fexit : 4.309 ± 0.148M/s fmodret : 4.314 ± 0.203M/s /* selftest: single-threaded, no atomics */ task_rename base 4555K events per sec task_rename kprobe 3643K events per sec task_rename kretprobe 2506K events per sec task_rename raw_tp 4303K events per sec task_rename fentry 4307K events per sec task_rename fexit 4010K events per sec task_rename fmod_ret 3984K events per sec Multi-threaded implementations ============================== /* bench: multi-threaded w/ atomics */ base : 3.910 ± 0.023M/s kprobe : 3.048 ± 0.037M/s kretprobe : 2.300 ± 0.015M/s rawtp : 3.687 ± 0.034M/s fentry : 3.740 ± 0.087M/s fexit : 3.510 ± 0.009M/s fmodret : 3.485 ± 0.050M/s /* selftest: multi-threaded w/ atomics */ task_rename base 3872K events per sec task_rename kprobe 3068K events per sec task_rename kretprobe 2350K events per sec task_rename raw_tp 3731K events per sec task_rename fentry 3639K events per sec task_rename fexit 3558K events per sec task_rename fmod_ret 3511K events per sec /* selftest: multi-threaded, no atomics */ task_rename base 3945K events per sec task_rename kprobe 3298K events per sec task_rename kretprobe 2451K events per sec task_rename raw_tp 3718K events per sec task_rename fentry 3782K events per sec task_rename fexit 3543K events per sec task_rename fmod_ret 3526K events per sec Note that the fact that ./bench benchmark always uses atomic increments for counting, while test_overhead doesn't, doesn't influence test results all that much. Signed-off-by: Andrii Nakryiko <andriin@fb.com> Signed-off-by: Alexei Starovoitov <ast@kernel.org> Acked-by: John Fastabend <john.fastabend@gmail.com> Acked-by: Yonghong Song <yhs@fb.com> Link: https://lore.kernel.org/bpf/20200512192445.2351848-4-andriin@fb.com
2020-05-13 03:24:44 +08:00
extern const struct bench bench_rename_base;
extern const struct bench bench_rename_kprobe;
extern const struct bench bench_rename_kretprobe;
extern const struct bench bench_rename_rawtp;
extern const struct bench bench_rename_fentry;
extern const struct bench bench_rename_fexit;
selftest/bpf: Add BPF triggering benchmark It is sometimes desirable to be able to trigger BPF program from user-space with minimal overhead. sys_enter would seem to be a good candidate, yet in a lot of cases there will be a lot of noise from syscalls triggered by other processes on the system. So while searching for low-overhead alternative, I've stumbled upon getpgid() syscall, which seems to be specific enough to not suffer from accidental syscall by other apps. This set of benchmarks compares tp, raw_tp w/ filtering by syscall ID, kprobe, fentry and fmod_ret with returning error (so that syscall would not be executed), to determine the lowest-overhead way. Here are results on my machine (using benchs/run_bench_trigger.sh script): base : 9.200 ± 0.319M/s tp : 6.690 ± 0.125M/s rawtp : 8.571 ± 0.214M/s kprobe : 6.431 ± 0.048M/s fentry : 8.955 ± 0.241M/s fmodret : 8.903 ± 0.135M/s So it seems like fmodret doesn't give much benefit for such lightweight syscall. Raw tracepoint is pretty decent despite additional filtering logic, but it will be called for any other syscall in the system, which rules it out. Fentry, though, seems to be adding the least amoung of overhead and achieves 97.3% of performance of baseline no-BPF-attached syscall. Using getpgid() seems to be preferable to set_task_comm() approach from test_overhead, as it's about 2.35x faster in a baseline performance. Signed-off-by: Andrii Nakryiko <andriin@fb.com> Signed-off-by: Alexei Starovoitov <ast@kernel.org> Acked-by: John Fastabend <john.fastabend@gmail.com> Acked-by: Yonghong Song <yhs@fb.com> Link: https://lore.kernel.org/bpf/20200512192445.2351848-5-andriin@fb.com
2020-05-13 03:24:45 +08:00
extern const struct bench bench_trig_base;
extern const struct bench bench_trig_tp;
extern const struct bench bench_trig_rawtp;
extern const struct bench bench_trig_kprobe;
extern const struct bench bench_trig_fentry;
selftests/bpf: Add sleepable tests Modify few tests to sanity test sleepable bpf functionality. Running 'bench trig-fentry-sleep' vs 'bench trig-fentry' and 'perf report': sleepable with SRCU: 3.86% bench [k] __srcu_read_unlock 3.22% bench [k] __srcu_read_lock 0.92% bench [k] bpf_prog_740d4210cdcd99a3_bench_trigger_fentry_sleep 0.50% bench [k] bpf_trampoline_10297 0.26% bench [k] __bpf_prog_exit_sleepable 0.21% bench [k] __bpf_prog_enter_sleepable sleepable with RCU_TRACE: 0.79% bench [k] bpf_prog_740d4210cdcd99a3_bench_trigger_fentry_sleep 0.72% bench [k] bpf_trampoline_10381 0.31% bench [k] __bpf_prog_exit_sleepable 0.29% bench [k] __bpf_prog_enter_sleepable non-sleepable with RCU: 0.88% bench [k] bpf_prog_740d4210cdcd99a3_bench_trigger_fentry 0.84% bench [k] bpf_trampoline_10297 0.13% bench [k] __bpf_prog_enter 0.12% bench [k] __bpf_prog_exit Signed-off-by: Alexei Starovoitov <ast@kernel.org> Signed-off-by: Daniel Borkmann <daniel@iogearbox.net> Acked-by: KP Singh <kpsingh@google.com> Link: https://lore.kernel.org/bpf/20200827220114.69225-6-alexei.starovoitov@gmail.com
2020-08-28 06:01:14 +08:00
extern const struct bench bench_trig_fentry_sleep;
selftest/bpf: Add BPF triggering benchmark It is sometimes desirable to be able to trigger BPF program from user-space with minimal overhead. sys_enter would seem to be a good candidate, yet in a lot of cases there will be a lot of noise from syscalls triggered by other processes on the system. So while searching for low-overhead alternative, I've stumbled upon getpgid() syscall, which seems to be specific enough to not suffer from accidental syscall by other apps. This set of benchmarks compares tp, raw_tp w/ filtering by syscall ID, kprobe, fentry and fmod_ret with returning error (so that syscall would not be executed), to determine the lowest-overhead way. Here are results on my machine (using benchs/run_bench_trigger.sh script): base : 9.200 ± 0.319M/s tp : 6.690 ± 0.125M/s rawtp : 8.571 ± 0.214M/s kprobe : 6.431 ± 0.048M/s fentry : 8.955 ± 0.241M/s fmodret : 8.903 ± 0.135M/s So it seems like fmodret doesn't give much benefit for such lightweight syscall. Raw tracepoint is pretty decent despite additional filtering logic, but it will be called for any other syscall in the system, which rules it out. Fentry, though, seems to be adding the least amoung of overhead and achieves 97.3% of performance of baseline no-BPF-attached syscall. Using getpgid() seems to be preferable to set_task_comm() approach from test_overhead, as it's about 2.35x faster in a baseline performance. Signed-off-by: Andrii Nakryiko <andriin@fb.com> Signed-off-by: Alexei Starovoitov <ast@kernel.org> Acked-by: John Fastabend <john.fastabend@gmail.com> Acked-by: Yonghong Song <yhs@fb.com> Link: https://lore.kernel.org/bpf/20200512192445.2351848-5-andriin@fb.com
2020-05-13 03:24:45 +08:00
extern const struct bench bench_trig_fmodret;
selftests/bpf: Add uprobe triggering overhead benchmarks Add benchmark to measure overhead of uprobes and uretprobes. Also have a baseline (no uprobe attached) benchmark. On my dev machine, baseline benchmark can trigger 130M user_target() invocations. When uprobe is attached, this falls to just 700K. With uretprobe, we get down to 520K: $ sudo ./bench trig-uprobe-base -a Summary: hits 131.289 ± 2.872M/s # UPROBE $ sudo ./bench -a trig-uprobe-without-nop Summary: hits 0.729 ± 0.007M/s $ sudo ./bench -a trig-uprobe-with-nop Summary: hits 1.798 ± 0.017M/s # URETPROBE $ sudo ./bench -a trig-uretprobe-without-nop Summary: hits 0.508 ± 0.012M/s $ sudo ./bench -a trig-uretprobe-with-nop Summary: hits 0.883 ± 0.008M/s So there is almost 2.5x performance difference between probing nop vs non-nop instruction for entry uprobe. And 1.7x difference for uretprobe. This means that non-nop uprobe overhead is around 1.4 microseconds for uprobe and 2 microseconds for non-nop uretprobe. For nop variants, uprobe and uretprobe overhead is down to 0.556 and 1.13 microseconds, respectively. For comparison, just doing a very low-overhead syscall (with no BPF programs attached anywhere) gives: $ sudo ./bench trig-base -a Summary: hits 4.830 ± 0.036M/s So uprobes are about 2.67x slower than pure context switch. Signed-off-by: Andrii Nakryiko <andrii@kernel.org> Signed-off-by: Daniel Borkmann <daniel@iogearbox.net> Link: https://lore.kernel.org/bpf/20211116013041.4072571-1-andrii@kernel.org
2021-11-16 09:30:41 +08:00
extern const struct bench bench_trig_uprobe_base;
extern const struct bench bench_trig_uprobe_with_nop;
extern const struct bench bench_trig_uretprobe_with_nop;
extern const struct bench bench_trig_uprobe_without_nop;
extern const struct bench bench_trig_uretprobe_without_nop;
bpf: Add BPF ringbuf and perf buffer benchmarks Extend bench framework with ability to have benchmark-provided child argument parser for custom benchmark-specific parameters. This makes bench generic code modular and independent from any specific benchmark. Also implement a set of benchmarks for new BPF ring buffer and existing perf buffer. 4 benchmarks were implemented: 2 variations for each of BPF ringbuf and perfbuf:, - rb-libbpf utilizes stock libbpf ring_buffer manager for reading data; - rb-custom implements custom ring buffer setup and reading code, to eliminate overheads inherent in generic libbpf code due to callback functions and the need to update consumer position after each consumed record, instead of batching updates (due to pessimistic assumption that user callback might take long time and thus could unnecessarily hold ring buffer space for too long); - pb-libbpf uses stock libbpf perf_buffer code with all the default settings, though uses higher-performance raw event callback to minimize unnecessary overhead; - pb-custom implements its own custom consumer code to minimize any possible overhead of generic libbpf implementation and indirect function calls. All of the test support default, no data notification skipped, mode, as well as sampled mode (with --rb-sampled flag), which allows to trigger epoll notification less frequently and reduce overhead. As will be shown, this mode is especially critical for perf buffer, which suffers from high overhead of wakeups in kernel. Otherwise, all benchamrks implement similar way to generate a batch of records by using fentry/sys_getpgid BPF program, which pushes a bunch of records in a tight loop and records number of successful and dropped samples. Each record is a small 8-byte integer, to minimize the effect of memory copying with bpf_perf_event_output() and bpf_ringbuf_output(). Benchmarks that have only one producer implement optional back-to-back mode, in which record production and consumption is alternating on the same CPU. This is the highest-throughput happy case, showing ultimate performance achievable with either BPF ringbuf or perfbuf. All the below scenarios are implemented in a script in benchs/run_bench_ringbufs.sh. Tests were performed on 28-core/56-thread Intel Xeon CPU E5-2680 v4 @ 2.40GHz CPU. Single-producer, parallel producer ================================== rb-libbpf 12.054 ± 0.320M/s (drops 0.000 ± 0.000M/s) rb-custom 8.158 ± 0.118M/s (drops 0.001 ± 0.003M/s) pb-libbpf 0.931 ± 0.007M/s (drops 0.000 ± 0.000M/s) pb-custom 0.965 ± 0.003M/s (drops 0.000 ± 0.000M/s) Single-producer, parallel producer, sampled notification ======================================================== rb-libbpf 11.563 ± 0.067M/s (drops 0.000 ± 0.000M/s) rb-custom 15.895 ± 0.076M/s (drops 0.000 ± 0.000M/s) pb-libbpf 9.889 ± 0.032M/s (drops 0.000 ± 0.000M/s) pb-custom 9.866 ± 0.028M/s (drops 0.000 ± 0.000M/s) Single producer on one CPU, consumer on another one, both running at full speed. Curiously, rb-libbpf has higher throughput than objectively faster (due to more lightweight consumer code path) rb-custom. It appears that faster consumer causes kernel to send notifications more frequently, because consumer appears to be caught up more frequently. Performance of perfbuf suffers from default "no sampling" policy and huge overhead that causes. In sampled mode, rb-custom is winning very significantly eliminating too frequent in-kernel wakeups, the gain appears to be more than 2x. Perf buffer achieves even more impressive wins, compared to stock perfbuf settings, with 10x improvements in throughput with 1:500 sampling rate. The trade-off is that with sampling, application might not get next X events until X+1st arrives, which is not always acceptable. With steady influx of events, though, this shouldn't be a problem. Overall, single-producer performance of ring buffers seems to be better no matter the sampled/non-sampled modes, but it especially beats ring buffer without sampling due to its adaptive notification approach. Single-producer, back-to-back mode ================================== rb-libbpf 15.507 ± 0.247M/s (drops 0.000 ± 0.000M/s) rb-libbpf-sampled 14.692 ± 0.195M/s (drops 0.000 ± 0.000M/s) rb-custom 21.449 ± 0.157M/s (drops 0.000 ± 0.000M/s) rb-custom-sampled 20.024 ± 0.386M/s (drops 0.000 ± 0.000M/s) pb-libbpf 1.601 ± 0.015M/s (drops 0.000 ± 0.000M/s) pb-libbpf-sampled 8.545 ± 0.064M/s (drops 0.000 ± 0.000M/s) pb-custom 1.607 ± 0.022M/s (drops 0.000 ± 0.000M/s) pb-custom-sampled 8.988 ± 0.144M/s (drops 0.000 ± 0.000M/s) Here we test a back-to-back mode, which is arguably best-case scenario both for BPF ringbuf and perfbuf, because there is no contention and for ringbuf also no excessive notification, because consumer appears to be behind after the first record. For ringbuf, custom consumer code clearly wins with 21.5 vs 16 million records per second exchanged between producer and consumer. Sampled mode actually hurts a bit due to slightly slower producer logic (it needs to fetch amount of data available to decide whether to skip or force notification). Perfbuf with wakeup sampling gets 5.5x throughput increase, compared to no-sampling version. There also doesn't seem to be noticeable overhead from generic libbpf handling code. Perfbuf back-to-back, effect of sample rate =========================================== pb-sampled-1 1.035 ± 0.012M/s (drops 0.000 ± 0.000M/s) pb-sampled-5 3.476 ± 0.087M/s (drops 0.000 ± 0.000M/s) pb-sampled-10 5.094 ± 0.136M/s (drops 0.000 ± 0.000M/s) pb-sampled-25 7.118 ± 0.153M/s (drops 0.000 ± 0.000M/s) pb-sampled-50 8.169 ± 0.156M/s (drops 0.000 ± 0.000M/s) pb-sampled-100 8.887 ± 0.136M/s (drops 0.000 ± 0.000M/s) pb-sampled-250 9.180 ± 0.209M/s (drops 0.000 ± 0.000M/s) pb-sampled-500 9.353 ± 0.281M/s (drops 0.000 ± 0.000M/s) pb-sampled-1000 9.411 ± 0.217M/s (drops 0.000 ± 0.000M/s) pb-sampled-2000 9.464 ± 0.167M/s (drops 0.000 ± 0.000M/s) pb-sampled-3000 9.575 ± 0.273M/s (drops 0.000 ± 0.000M/s) This benchmark shows the effect of event sampling for perfbuf. Back-to-back mode for highest throughput. Just doing every 5th record notification gives 3.5x speed up. 250-500 appears to be the point of diminishing return, with almost 9x speed up. Most benchmarks use 500 as the default sampling for pb-raw and pb-custom. Ringbuf back-to-back, effect of sample rate =========================================== rb-sampled-1 1.106 ± 0.010M/s (drops 0.000 ± 0.000M/s) rb-sampled-5 4.746 ± 0.149M/s (drops 0.000 ± 0.000M/s) rb-sampled-10 7.706 ± 0.164M/s (drops 0.000 ± 0.000M/s) rb-sampled-25 12.893 ± 0.273M/s (drops 0.000 ± 0.000M/s) rb-sampled-50 15.961 ± 0.361M/s (drops 0.000 ± 0.000M/s) rb-sampled-100 18.203 ± 0.445M/s (drops 0.000 ± 0.000M/s) rb-sampled-250 19.962 ± 0.786M/s (drops 0.000 ± 0.000M/s) rb-sampled-500 20.881 ± 0.551M/s (drops 0.000 ± 0.000M/s) rb-sampled-1000 21.317 ± 0.532M/s (drops 0.000 ± 0.000M/s) rb-sampled-2000 21.331 ± 0.535M/s (drops 0.000 ± 0.000M/s) rb-sampled-3000 21.688 ± 0.392M/s (drops 0.000 ± 0.000M/s) Similar benchmark for ring buffer also shows a great advantage (in terms of throughput) of skipping notifications. Skipping every 5th one gives 4x boost. Also similar to perfbuf case, 250-500 seems to be the point of diminishing returns, giving roughly 20x better results. Keep in mind, for this test, notifications are controlled manually with BPF_RB_NO_WAKEUP and BPF_RB_FORCE_WAKEUP. As can be seen from previous benchmarks, adaptive notifications based on consumer's positions provides same (or even slightly better due to simpler load generator on BPF side) benefits in favorable back-to-back scenario. Over zealous and fast consumer, which is almost always caught up, will make thoughput numbers smaller. That's the case when manual notification control might prove to be extremely beneficial. Ringbuf back-to-back, reserve+commit vs output ============================================== reserve 22.819 ± 0.503M/s (drops 0.000 ± 0.000M/s) output 18.906 ± 0.433M/s (drops 0.000 ± 0.000M/s) Ringbuf sampled, reserve+commit vs output ========================================= reserve-sampled 15.350 ± 0.132M/s (drops 0.000 ± 0.000M/s) output-sampled 14.195 ± 0.144M/s (drops 0.000 ± 0.000M/s) BPF ringbuf supports two sets of APIs with various usability and performance tradeoffs: bpf_ringbuf_reserve()+bpf_ringbuf_commit() vs bpf_ringbuf_output(). This benchmark clearly shows superiority of reserve+commit approach, despite using a small 8-byte record size. Single-producer, consumer/producer competing on the same CPU, low batch count ============================================================================= rb-libbpf 3.045 ± 0.020M/s (drops 3.536 ± 0.148M/s) rb-custom 3.055 ± 0.022M/s (drops 3.893 ± 0.066M/s) pb-libbpf 1.393 ± 0.024M/s (drops 0.000 ± 0.000M/s) pb-custom 1.407 ± 0.016M/s (drops 0.000 ± 0.000M/s) This benchmark shows one of the worst-case scenarios, in which producer and consumer do not coordinate *and* fight for the same CPU. No batch count and sampling settings were able to eliminate drops for ringbuffer, producer is just too fast for consumer to keep up. But ringbuf and perfbuf still able to pass through quite a lot of messages, which is more than enough for a lot of applications. Ringbuf, multi-producer contention ================================== rb-libbpf nr_prod 1 10.916 ± 0.399M/s (drops 0.000 ± 0.000M/s) rb-libbpf nr_prod 2 4.931 ± 0.030M/s (drops 0.000 ± 0.000M/s) rb-libbpf nr_prod 3 4.880 ± 0.006M/s (drops 0.000 ± 0.000M/s) rb-libbpf nr_prod 4 3.926 ± 0.004M/s (drops 0.000 ± 0.000M/s) rb-libbpf nr_prod 8 4.011 ± 0.004M/s (drops 0.000 ± 0.000M/s) rb-libbpf nr_prod 12 3.967 ± 0.016M/s (drops 0.000 ± 0.000M/s) rb-libbpf nr_prod 16 2.604 ± 0.030M/s (drops 0.001 ± 0.002M/s) rb-libbpf nr_prod 20 2.233 ± 0.003M/s (drops 0.000 ± 0.000M/s) rb-libbpf nr_prod 24 2.085 ± 0.015M/s (drops 0.000 ± 0.000M/s) rb-libbpf nr_prod 28 2.055 ± 0.004M/s (drops 0.000 ± 0.000M/s) rb-libbpf nr_prod 32 1.962 ± 0.004M/s (drops 0.000 ± 0.000M/s) rb-libbpf nr_prod 36 2.089 ± 0.005M/s (drops 0.000 ± 0.000M/s) rb-libbpf nr_prod 40 2.118 ± 0.006M/s (drops 0.000 ± 0.000M/s) rb-libbpf nr_prod 44 2.105 ± 0.004M/s (drops 0.000 ± 0.000M/s) rb-libbpf nr_prod 48 2.120 ± 0.058M/s (drops 0.000 ± 0.001M/s) rb-libbpf nr_prod 52 2.074 ± 0.024M/s (drops 0.007 ± 0.014M/s) Ringbuf uses a very short-duration spinlock during reservation phase, to check few invariants, increment producer count and set record header. This is the biggest point of contention for ringbuf implementation. This benchmark evaluates the effect of multiple competing writers on overall throughput of a single shared ringbuffer. Overall throughput drops almost 2x when going from single to two highly-contended producers, gradually dropping with additional competing producers. Performance drop stabilizes at around 20 producers and hovers around 2mln even with 50+ fighting producers, which is a 5x drop compared to non-contended case. Good kernel implementation in kernel helps maintain decent performance here. Note, that in the intended real-world scenarios, it's not expected to get even close to such a high levels of contention. But if contention will become a problem, there is always an option of sharding few ring buffers across a set of CPUs. Signed-off-by: Andrii Nakryiko <andriin@fb.com> Signed-off-by: Daniel Borkmann <daniel@iogearbox.net> Link: https://lore.kernel.org/bpf/20200529075424.3139988-5-andriin@fb.com Signed-off-by: Alexei Starovoitov <ast@kernel.org>
2020-05-29 15:54:23 +08:00
extern const struct bench bench_rb_libbpf;
extern const struct bench bench_rb_custom;
extern const struct bench bench_pb_libbpf;
extern const struct bench bench_pb_custom;
bpf/benchs: Add benchmark tests for bloom filter throughput + false positive This patch adds benchmark tests for the throughput (for lookups + updates) and the false positive rate of bloom filter lookups, as well as some minor refactoring of the bash script for running the benchmarks. These benchmarks show that as the number of hash functions increases, the throughput and the false positive rate of the bloom filter decreases. >From the benchmark data, the approximate average false-positive rates are roughly as follows: 1 hash function = ~30% 2 hash functions = ~15% 3 hash functions = ~5% 4 hash functions = ~2.5% 5 hash functions = ~1% 6 hash functions = ~0.5% 7 hash functions = ~0.35% 8 hash functions = ~0.15% 9 hash functions = ~0.1% 10 hash functions = ~0% For reference data, the benchmarks run on one thread on a machine with one numa node for 1 to 5 hash functions for 8-byte and 64-byte values are as follows: 1 hash function: 50k entries 8-byte value Lookups - 51.1 M/s operations Updates - 33.6 M/s operations False positive rate: 24.15% 64-byte value Lookups - 15.7 M/s operations Updates - 15.1 M/s operations False positive rate: 24.2% 100k entries 8-byte value Lookups - 51.0 M/s operations Updates - 33.4 M/s operations False positive rate: 24.04% 64-byte value Lookups - 15.6 M/s operations Updates - 14.6 M/s operations False positive rate: 24.06% 500k entries 8-byte value Lookups - 50.5 M/s operations Updates - 33.1 M/s operations False positive rate: 27.45% 64-byte value Lookups - 15.6 M/s operations Updates - 14.2 M/s operations False positive rate: 27.42% 1 mil entries 8-byte value Lookups - 49.7 M/s operations Updates - 32.9 M/s operations False positive rate: 27.45% 64-byte value Lookups - 15.4 M/s operations Updates - 13.7 M/s operations False positive rate: 27.58% 2.5 mil entries 8-byte value Lookups - 47.2 M/s operations Updates - 31.8 M/s operations False positive rate: 30.94% 64-byte value Lookups - 15.3 M/s operations Updates - 13.2 M/s operations False positive rate: 30.95% 5 mil entries 8-byte value Lookups - 41.1 M/s operations Updates - 28.1 M/s operations False positive rate: 31.01% 64-byte value Lookups - 13.3 M/s operations Updates - 11.4 M/s operations False positive rate: 30.98% 2 hash functions: 50k entries 8-byte value Lookups - 34.1 M/s operations Updates - 20.1 M/s operations False positive rate: 9.13% 64-byte value Lookups - 8.4 M/s operations Updates - 7.9 M/s operations False positive rate: 9.21% 100k entries 8-byte value Lookups - 33.7 M/s operations Updates - 18.9 M/s operations False positive rate: 9.13% 64-byte value Lookups - 8.4 M/s operations Updates - 7.7 M/s operations False positive rate: 9.19% 500k entries 8-byte value Lookups - 32.7 M/s operations Updates - 18.1 M/s operations False positive rate: 12.61% 64-byte value Lookups - 8.4 M/s operations Updates - 7.5 M/s operations False positive rate: 12.61% 1 mil entries 8-byte value Lookups - 30.6 M/s operations Updates - 18.9 M/s operations False positive rate: 12.54% 64-byte value Lookups - 8.0 M/s operations Updates - 7.0 M/s operations False positive rate: 12.52% 2.5 mil entries 8-byte value Lookups - 25.3 M/s operations Updates - 16.7 M/s operations False positive rate: 16.77% 64-byte value Lookups - 7.9 M/s operations Updates - 6.5 M/s operations False positive rate: 16.88% 5 mil entries 8-byte value Lookups - 20.8 M/s operations Updates - 14.7 M/s operations False positive rate: 16.78% 64-byte value Lookups - 7.0 M/s operations Updates - 6.0 M/s operations False positive rate: 16.78% 3 hash functions: 50k entries 8-byte value Lookups - 25.1 M/s operations Updates - 14.6 M/s operations False positive rate: 7.65% 64-byte value Lookups - 5.8 M/s operations Updates - 5.5 M/s operations False positive rate: 7.58% 100k entries 8-byte value Lookups - 24.7 M/s operations Updates - 14.1 M/s operations False positive rate: 7.71% 64-byte value Lookups - 5.8 M/s operations Updates - 5.3 M/s operations False positive rate: 7.62% 500k entries 8-byte value Lookups - 22.9 M/s operations Updates - 13.9 M/s operations False positive rate: 2.62% 64-byte value Lookups - 5.6 M/s operations Updates - 4.8 M/s operations False positive rate: 2.7% 1 mil entries 8-byte value Lookups - 19.8 M/s operations Updates - 12.6 M/s operations False positive rate: 2.60% 64-byte value Lookups - 5.3 M/s operations Updates - 4.4 M/s operations False positive rate: 2.69% 2.5 mil entries 8-byte value Lookups - 16.2 M/s operations Updates - 10.7 M/s operations False positive rate: 4.49% 64-byte value Lookups - 4.9 M/s operations Updates - 4.1 M/s operations False positive rate: 4.41% 5 mil entries 8-byte value Lookups - 18.8 M/s operations Updates - 9.2 M/s operations False positive rate: 4.45% 64-byte value Lookups - 5.2 M/s operations Updates - 3.9 M/s operations False positive rate: 4.54% 4 hash functions: 50k entries 8-byte value Lookups - 19.7 M/s operations Updates - 11.1 M/s operations False positive rate: 1.01% 64-byte value Lookups - 4.4 M/s operations Updates - 4.0 M/s operations False positive rate: 1.00% 100k entries 8-byte value Lookups - 19.5 M/s operations Updates - 10.9 M/s operations False positive rate: 1.00% 64-byte value Lookups - 4.3 M/s operations Updates - 3.9 M/s operations False positive rate: 0.97% 500k entries 8-byte value Lookups - 18.2 M/s operations Updates - 10.6 M/s operations False positive rate: 2.05% 64-byte value Lookups - 4.3 M/s operations Updates - 3.7 M/s operations False positive rate: 2.05% 1 mil entries 8-byte value Lookups - 15.5 M/s operations Updates - 9.6 M/s operations False positive rate: 1.99% 64-byte value Lookups - 4.0 M/s operations Updates - 3.4 M/s operations False positive rate: 1.99% 2.5 mil entries 8-byte value Lookups - 13.8 M/s operations Updates - 7.7 M/s operations False positive rate: 3.91% 64-byte value Lookups - 3.7 M/s operations Updates - 3.6 M/s operations False positive rate: 3.78% 5 mil entries 8-byte value Lookups - 13.0 M/s operations Updates - 6.9 M/s operations False positive rate: 3.93% 64-byte value Lookups - 3.5 M/s operations Updates - 3.7 M/s operations False positive rate: 3.39% 5 hash functions: 50k entries 8-byte value Lookups - 16.4 M/s operations Updates - 9.1 M/s operations False positive rate: 0.78% 64-byte value Lookups - 3.5 M/s operations Updates - 3.2 M/s operations False positive rate: 0.77% 100k entries 8-byte value Lookups - 16.3 M/s operations Updates - 9.0 M/s operations False positive rate: 0.79% 64-byte value Lookups - 3.5 M/s operations Updates - 3.2 M/s operations False positive rate: 0.78% 500k entries 8-byte value Lookups - 15.1 M/s operations Updates - 8.8 M/s operations False positive rate: 1.82% 64-byte value Lookups - 3.4 M/s operations Updates - 3.0 M/s operations False positive rate: 1.78% 1 mil entries 8-byte value Lookups - 13.2 M/s operations Updates - 7.8 M/s operations False positive rate: 1.81% 64-byte value Lookups - 3.2 M/s operations Updates - 2.8 M/s operations False positive rate: 1.80% 2.5 mil entries 8-byte value Lookups - 10.5 M/s operations Updates - 5.9 M/s operations False positive rate: 0.29% 64-byte value Lookups - 3.2 M/s operations Updates - 2.4 M/s operations False positive rate: 0.28% 5 mil entries 8-byte value Lookups - 9.6 M/s operations Updates - 5.7 M/s operations False positive rate: 0.30% 64-byte value Lookups - 3.2 M/s operations Updates - 2.7 M/s operations False positive rate: 0.30% Signed-off-by: Joanne Koong <joannekoong@fb.com> Signed-off-by: Alexei Starovoitov <ast@kernel.org> Acked-by: Andrii Nakryiko <andrii@kernel.org> Link: https://lore.kernel.org/bpf/20211027234504.30744-5-joannekoong@fb.com
2021-10-28 07:45:03 +08:00
extern const struct bench bench_bloom_lookup;
extern const struct bench bench_bloom_update;
extern const struct bench bench_bloom_false_positive;
bpf/benchs: Add benchmarks for comparing hashmap lookups w/ vs. w/out bloom filter This patch adds benchmark tests for comparing the performance of hashmap lookups without the bloom filter vs. hashmap lookups with the bloom filter. Checking the bloom filter first for whether the element exists should overall enable a higher throughput for hashmap lookups, since if the element does not exist in the bloom filter, we can avoid a costly lookup in the hashmap. On average, using 5 hash functions in the bloom filter tended to perform the best across the widest range of different entry sizes. The benchmark results using 5 hash functions (running on 8 threads on a machine with one numa node, and taking the average of 3 runs) were roughly as follows: value_size = 4 bytes - 10k entries: 30% faster 50k entries: 40% faster 100k entries: 40% faster 500k entres: 70% faster 1 million entries: 90% faster 5 million entries: 140% faster value_size = 8 bytes - 10k entries: 30% faster 50k entries: 40% faster 100k entries: 50% faster 500k entres: 80% faster 1 million entries: 100% faster 5 million entries: 150% faster value_size = 16 bytes - 10k entries: 20% faster 50k entries: 30% faster 100k entries: 35% faster 500k entres: 65% faster 1 million entries: 85% faster 5 million entries: 110% faster value_size = 40 bytes - 10k entries: 5% faster 50k entries: 15% faster 100k entries: 20% faster 500k entres: 65% faster 1 million entries: 75% faster 5 million entries: 120% faster Signed-off-by: Joanne Koong <joannekoong@fb.com> Signed-off-by: Alexei Starovoitov <ast@kernel.org> Link: https://lore.kernel.org/bpf/20211027234504.30744-6-joannekoong@fb.com
2021-10-28 07:45:04 +08:00
extern const struct bench bench_hashmap_without_bloom;
extern const struct bench bench_hashmap_with_bloom;
selftest/bpf/benchs: Add bpf_loop benchmark Add benchmark to measure the throughput and latency of the bpf_loop call. Testing this on my dev machine on 1 thread, the data is as follows: nr_loops: 10 bpf_loop - throughput: 198.519 ± 0.155 M ops/s, latency: 5.037 ns/op nr_loops: 100 bpf_loop - throughput: 247.448 ± 0.305 M ops/s, latency: 4.041 ns/op nr_loops: 500 bpf_loop - throughput: 260.839 ± 0.380 M ops/s, latency: 3.834 ns/op nr_loops: 1000 bpf_loop - throughput: 262.806 ± 0.629 M ops/s, latency: 3.805 ns/op nr_loops: 5000 bpf_loop - throughput: 264.211 ± 1.508 M ops/s, latency: 3.785 ns/op nr_loops: 10000 bpf_loop - throughput: 265.366 ± 3.054 M ops/s, latency: 3.768 ns/op nr_loops: 50000 bpf_loop - throughput: 235.986 ± 20.205 M ops/s, latency: 4.238 ns/op nr_loops: 100000 bpf_loop - throughput: 264.482 ± 0.279 M ops/s, latency: 3.781 ns/op nr_loops: 500000 bpf_loop - throughput: 309.773 ± 87.713 M ops/s, latency: 3.228 ns/op nr_loops: 1000000 bpf_loop - throughput: 262.818 ± 4.143 M ops/s, latency: 3.805 ns/op >From this data, we can see that the latency per loop decreases as the number of loops increases. On this particular machine, each loop had an overhead of about ~4 ns, and we were able to run ~250 million loops per second. Signed-off-by: Joanne Koong <joannekoong@fb.com> Signed-off-by: Alexei Starovoitov <ast@kernel.org> Acked-by: Andrii Nakryiko <andrii@kernel.org> Link: https://lore.kernel.org/bpf/20211130030622.4131246-5-joannekoong@fb.com
2021-11-30 11:06:22 +08:00
extern const struct bench bench_bpf_loop;
selftests/bpf: Add benchmark for bpf_strncmp() helper Add benchmark to compare the performance between home-made strncmp() in bpf program and bpf_strncmp() helper. In summary, the performance win of bpf_strncmp() under x86-64 is greater than 18% when the compared string length is greater than 64, and is 179% when the length is 4095. Under arm64 the performance win is even bigger: 33% when the length is greater than 64 and 600% when the length is 4095. The following is the details: no-helper-X: use home-made strncmp() to compare X-sized string helper-Y: use bpf_strncmp() to compare Y-sized string Under x86-64: no-helper-1 3.504 ± 0.000M/s (drops 0.000 ± 0.000M/s) helper-1 3.347 ± 0.001M/s (drops 0.000 ± 0.000M/s) no-helper-8 3.357 ± 0.001M/s (drops 0.000 ± 0.000M/s) helper-8 3.307 ± 0.001M/s (drops 0.000 ± 0.000M/s) no-helper-32 3.064 ± 0.000M/s (drops 0.000 ± 0.000M/s) helper-32 3.253 ± 0.001M/s (drops 0.000 ± 0.000M/s) no-helper-64 2.563 ± 0.001M/s (drops 0.000 ± 0.000M/s) helper-64 3.040 ± 0.001M/s (drops 0.000 ± 0.000M/s) no-helper-128 1.975 ± 0.000M/s (drops 0.000 ± 0.000M/s) helper-128 2.641 ± 0.000M/s (drops 0.000 ± 0.000M/s) no-helper-512 0.759 ± 0.000M/s (drops 0.000 ± 0.000M/s) helper-512 1.574 ± 0.000M/s (drops 0.000 ± 0.000M/s) no-helper-2048 0.329 ± 0.000M/s (drops 0.000 ± 0.000M/s) helper-2048 0.602 ± 0.000M/s (drops 0.000 ± 0.000M/s) no-helper-4095 0.117 ± 0.000M/s (drops 0.000 ± 0.000M/s) helper-4095 0.327 ± 0.000M/s (drops 0.000 ± 0.000M/s) Under arm64: no-helper-1 2.806 ± 0.004M/s (drops 0.000 ± 0.000M/s) helper-1 2.819 ± 0.002M/s (drops 0.000 ± 0.000M/s) no-helper-8 2.797 ± 0.109M/s (drops 0.000 ± 0.000M/s) helper-8 2.786 ± 0.025M/s (drops 0.000 ± 0.000M/s) no-helper-32 2.399 ± 0.011M/s (drops 0.000 ± 0.000M/s) helper-32 2.703 ± 0.002M/s (drops 0.000 ± 0.000M/s) no-helper-64 2.020 ± 0.015M/s (drops 0.000 ± 0.000M/s) helper-64 2.702 ± 0.073M/s (drops 0.000 ± 0.000M/s) no-helper-128 1.604 ± 0.001M/s (drops 0.000 ± 0.000M/s) helper-128 2.516 ± 0.002M/s (drops 0.000 ± 0.000M/s) no-helper-512 0.699 ± 0.000M/s (drops 0.000 ± 0.000M/s) helper-512 2.106 ± 0.003M/s (drops 0.000 ± 0.000M/s) no-helper-2048 0.215 ± 0.000M/s (drops 0.000 ± 0.000M/s) helper-2048 1.223 ± 0.003M/s (drops 0.000 ± 0.000M/s) no-helper-4095 0.112 ± 0.000M/s (drops 0.000 ± 0.000M/s) helper-4095 0.796 ± 0.000M/s (drops 0.000 ± 0.000M/s) Signed-off-by: Hou Tao <houtao1@huawei.com> Signed-off-by: Alexei Starovoitov <ast@kernel.org> Link: https://lore.kernel.org/bpf/20211210141652.877186-4-houtao1@huawei.com
2021-12-10 22:16:51 +08:00
extern const struct bench bench_strncmp_no_helper;
extern const struct bench bench_strncmp_helper;
selftest/bpf/benchs: Add bpf_map benchmark Add benchmark for hash_map to reproduce the worst case that non-stop update when map's free is zero. Just like this: ./run_bench_bpf_hashmap_full_update.sh Setting up benchmark 'bpf-hashmap-ful-update'... Benchmark 'bpf-hashmap-ful-update' started. 1:hash_map_full_perf 555830 events per sec ... Signed-off-by: Feng Zhou <zhoufeng.zf@bytedance.com> Link: https://lore.kernel.org/r/20220610023308.93798-3-zhoufeng.zf@bytedance.com Signed-off-by: Alexei Starovoitov <ast@kernel.org>
2022-06-10 10:33:08 +08:00
extern const struct bench bench_bpf_hashmap_full_update;
selftests/bpf: Add benchmark for local_storage get Add a benchmarks to demonstrate the performance cliff for local_storage get as the number of local_storage maps increases beyond current local_storage implementation's cache size. "sequential get" and "interleaved get" benchmarks are added, both of which do many bpf_task_storage_get calls on sets of task local_storage maps of various counts, while considering a single specific map to be 'important' and counting task_storage_gets to the important map separately in addition to normal 'hits' count of all gets. Goal here is to mimic scenario where a particular program using one map - the important one - is running on a system where many other local_storage maps exist and are accessed often. While "sequential get" benchmark does bpf_task_storage_get for map 0, 1, ..., {9, 99, 999} in order, "interleaved" benchmark interleaves 4 bpf_task_storage_gets for the important map for every 10 map gets. This is meant to highlight performance differences when important map is accessed far more frequently than non-important maps. A "hashmap control" benchmark is also included for easy comparison of standard bpf hashmap lookup vs local_storage get. The benchmark is similar to "sequential get", but creates and uses BPF_MAP_TYPE_HASH instead of local storage. Only one inner map is created - a hashmap meant to hold tid -> data mapping for all tasks. Size of the hashmap is hardcoded to my system's PID_MAX_LIMIT (4,194,304). The number of these keys which are actually fetched as part of the benchmark is configurable. Addition of this benchmark is inspired by conversation with Alexei in a previous patchset's thread [0], which highlighted the need for such a benchmark to motivate and validate improvements to local_storage implementation. My approach in that series focused on improving performance for explicitly-marked 'important' maps and was rejected with feedback to make more generally-applicable improvements while avoiding explicitly marking maps as important. Thus the benchmark reports both general and important-map-focused metrics, so effect of future work on both is clear. Regarding the benchmark results. On a powerful system (Skylake, 20 cores, 256gb ram): Hashmap Control =============== num keys: 10 hashmap (control) sequential get: hits throughput: 20.900 ± 0.334 M ops/s, hits latency: 47.847 ns/op, important_hits throughput: 20.900 ± 0.334 M ops/s num keys: 1000 hashmap (control) sequential get: hits throughput: 13.758 ± 0.219 M ops/s, hits latency: 72.683 ns/op, important_hits throughput: 13.758 ± 0.219 M ops/s num keys: 10000 hashmap (control) sequential get: hits throughput: 6.995 ± 0.034 M ops/s, hits latency: 142.959 ns/op, important_hits throughput: 6.995 ± 0.034 M ops/s num keys: 100000 hashmap (control) sequential get: hits throughput: 4.452 ± 0.371 M ops/s, hits latency: 224.635 ns/op, important_hits throughput: 4.452 ± 0.371 M ops/s num keys: 4194304 hashmap (control) sequential get: hits throughput: 3.043 ± 0.033 M ops/s, hits latency: 328.587 ns/op, important_hits throughput: 3.043 ± 0.033 M ops/s Local Storage ============= num_maps: 1 local_storage cache sequential get: hits throughput: 47.298 ± 0.180 M ops/s, hits latency: 21.142 ns/op, important_hits throughput: 47.298 ± 0.180 M ops/s local_storage cache interleaved get: hits throughput: 55.277 ± 0.888 M ops/s, hits latency: 18.091 ns/op, important_hits throughput: 55.277 ± 0.888 M ops/s num_maps: 10 local_storage cache sequential get: hits throughput: 40.240 ± 0.802 M ops/s, hits latency: 24.851 ns/op, important_hits throughput: 4.024 ± 0.080 M ops/s local_storage cache interleaved get: hits throughput: 48.701 ± 0.722 M ops/s, hits latency: 20.533 ns/op, important_hits throughput: 17.393 ± 0.258 M ops/s num_maps: 16 local_storage cache sequential get: hits throughput: 44.515 ± 0.708 M ops/s, hits latency: 22.464 ns/op, important_hits throughput: 2.782 ± 0.044 M ops/s local_storage cache interleaved get: hits throughput: 49.553 ± 2.260 M ops/s, hits latency: 20.181 ns/op, important_hits throughput: 15.767 ± 0.719 M ops/s num_maps: 17 local_storage cache sequential get: hits throughput: 38.778 ± 0.302 M ops/s, hits latency: 25.788 ns/op, important_hits throughput: 2.284 ± 0.018 M ops/s local_storage cache interleaved get: hits throughput: 43.848 ± 1.023 M ops/s, hits latency: 22.806 ns/op, important_hits throughput: 13.349 ± 0.311 M ops/s num_maps: 24 local_storage cache sequential get: hits throughput: 19.317 ± 0.568 M ops/s, hits latency: 51.769 ns/op, important_hits throughput: 0.806 ± 0.024 M ops/s local_storage cache interleaved get: hits throughput: 24.397 ± 0.272 M ops/s, hits latency: 40.989 ns/op, important_hits throughput: 6.863 ± 0.077 M ops/s num_maps: 32 local_storage cache sequential get: hits throughput: 13.333 ± 0.135 M ops/s, hits latency: 75.000 ns/op, important_hits throughput: 0.417 ± 0.004 M ops/s local_storage cache interleaved get: hits throughput: 16.898 ± 0.383 M ops/s, hits latency: 59.178 ns/op, important_hits throughput: 4.717 ± 0.107 M ops/s num_maps: 100 local_storage cache sequential get: hits throughput: 6.360 ± 0.107 M ops/s, hits latency: 157.233 ns/op, important_hits throughput: 0.064 ± 0.001 M ops/s local_storage cache interleaved get: hits throughput: 7.303 ± 0.362 M ops/s, hits latency: 136.930 ns/op, important_hits throughput: 1.907 ± 0.094 M ops/s num_maps: 1000 local_storage cache sequential get: hits throughput: 0.452 ± 0.010 M ops/s, hits latency: 2214.022 ns/op, important_hits throughput: 0.000 ± 0.000 M ops/s local_storage cache interleaved get: hits throughput: 0.542 ± 0.007 M ops/s, hits latency: 1843.341 ns/op, important_hits throughput: 0.136 ± 0.002 M ops/s Looking at the "sequential get" results, it's clear that as the number of task local_storage maps grows beyond the current cache size (16), there's a significant reduction in hits throughput. Note that current local_storage implementation assigns a cache_idx to maps as they are created. Since "sequential get" is creating maps 0..n in order and then doing bpf_task_storage_get calls in the same order, the benchmark is effectively ensuring that a map will not be in cache when the program tries to access it. For "interleaved get" results, important-map hits throughput is greatly increased as the important map is more likely to be in cache by virtue of being accessed far more frequently. Throughput still reduces as # maps increases, though. To get a sense of the overhead of the benchmark program, I commented out bpf_task_storage_get/bpf_map_lookup_elem in local_storage_bench.c and ran the benchmark on the same host as the 'real' run. Results: Hashmap Control =============== num keys: 10 hashmap (control) sequential get: hits throughput: 54.288 ± 0.655 M ops/s, hits latency: 18.420 ns/op, important_hits throughput: 54.288 ± 0.655 M ops/s num keys: 1000 hashmap (control) sequential get: hits throughput: 52.913 ± 0.519 M ops/s, hits latency: 18.899 ns/op, important_hits throughput: 52.913 ± 0.519 M ops/s num keys: 10000 hashmap (control) sequential get: hits throughput: 53.480 ± 1.235 M ops/s, hits latency: 18.699 ns/op, important_hits throughput: 53.480 ± 1.235 M ops/s num keys: 100000 hashmap (control) sequential get: hits throughput: 54.982 ± 1.902 M ops/s, hits latency: 18.188 ns/op, important_hits throughput: 54.982 ± 1.902 M ops/s num keys: 4194304 hashmap (control) sequential get: hits throughput: 50.858 ± 0.707 M ops/s, hits latency: 19.662 ns/op, important_hits throughput: 50.858 ± 0.707 M ops/s Local Storage ============= num_maps: 1 local_storage cache sequential get: hits throughput: 110.990 ± 4.828 M ops/s, hits latency: 9.010 ns/op, important_hits throughput: 110.990 ± 4.828 M ops/s local_storage cache interleaved get: hits throughput: 161.057 ± 4.090 M ops/s, hits latency: 6.209 ns/op, important_hits throughput: 161.057 ± 4.090 M ops/s num_maps: 10 local_storage cache sequential get: hits throughput: 112.930 ± 1.079 M ops/s, hits latency: 8.855 ns/op, important_hits throughput: 11.293 ± 0.108 M ops/s local_storage cache interleaved get: hits throughput: 115.841 ± 2.088 M ops/s, hits latency: 8.633 ns/op, important_hits throughput: 41.372 ± 0.746 M ops/s num_maps: 16 local_storage cache sequential get: hits throughput: 115.653 ± 0.416 M ops/s, hits latency: 8.647 ns/op, important_hits throughput: 7.228 ± 0.026 M ops/s local_storage cache interleaved get: hits throughput: 138.717 ± 1.649 M ops/s, hits latency: 7.209 ns/op, important_hits throughput: 44.137 ± 0.525 M ops/s num_maps: 17 local_storage cache sequential get: hits throughput: 112.020 ± 1.649 M ops/s, hits latency: 8.927 ns/op, important_hits throughput: 6.598 ± 0.097 M ops/s local_storage cache interleaved get: hits throughput: 128.089 ± 1.960 M ops/s, hits latency: 7.807 ns/op, important_hits throughput: 38.995 ± 0.597 M ops/s num_maps: 24 local_storage cache sequential get: hits throughput: 92.447 ± 5.170 M ops/s, hits latency: 10.817 ns/op, important_hits throughput: 3.855 ± 0.216 M ops/s local_storage cache interleaved get: hits throughput: 128.844 ± 2.808 M ops/s, hits latency: 7.761 ns/op, important_hits throughput: 36.245 ± 0.790 M ops/s num_maps: 32 local_storage cache sequential get: hits throughput: 102.042 ± 1.462 M ops/s, hits latency: 9.800 ns/op, important_hits throughput: 3.194 ± 0.046 M ops/s local_storage cache interleaved get: hits throughput: 126.577 ± 1.818 M ops/s, hits latency: 7.900 ns/op, important_hits throughput: 35.332 ± 0.507 M ops/s num_maps: 100 local_storage cache sequential get: hits throughput: 111.327 ± 1.401 M ops/s, hits latency: 8.983 ns/op, important_hits throughput: 1.113 ± 0.014 M ops/s local_storage cache interleaved get: hits throughput: 131.327 ± 1.339 M ops/s, hits latency: 7.615 ns/op, important_hits throughput: 34.302 ± 0.350 M ops/s num_maps: 1000 local_storage cache sequential get: hits throughput: 101.978 ± 0.563 M ops/s, hits latency: 9.806 ns/op, important_hits throughput: 0.102 ± 0.001 M ops/s local_storage cache interleaved get: hits throughput: 141.084 ± 1.098 M ops/s, hits latency: 7.088 ns/op, important_hits throughput: 35.430 ± 0.276 M ops/s Adjusting for overhead, latency numbers for "hashmap control" and "sequential get" are: hashmap_control_1k: ~53.8ns hashmap_control_10k: ~124.2ns hashmap_control_100k: ~206.5ns sequential_get_1: ~12.1ns sequential_get_10: ~16.0ns sequential_get_16: ~13.8ns sequential_get_17: ~16.8ns sequential_get_24: ~40.9ns sequential_get_32: ~65.2ns sequential_get_100: ~148.2ns sequential_get_1000: ~2204ns Clearly demonstrating a cliff. In the discussion for v1 of this patch, Alexei noted that local_storage was 2.5x faster than a large hashmap when initially implemented [1]. The benchmark results show that local_storage is 5-10x faster: a long-running BPF application putting some pid-specific info into a hashmap for each pid it sees will probably see on the order of 10-100k pids. Bench numbers for hashmaps of this size are ~10x slower than sequential_get_16, but as the number of local_storage maps grows far past local_storage cache size the performance advantage shrinks and eventually reverses. When running the benchmarks it may be necessary to bump 'open files' ulimit for a successful run. [0]: https://lore.kernel.org/all/20220420002143.1096548-1-davemarchevsky@fb.com [1]: https://lore.kernel.org/bpf/20220511173305.ftldpn23m4ski3d3@MBP-98dd607d3435.dhcp.thefacebook.com/ Signed-off-by: Dave Marchevsky <davemarchevsky@fb.com> Link: https://lore.kernel.org/r/20220620222554.270578-1-davemarchevsky@fb.com Signed-off-by: Alexei Starovoitov <ast@kernel.org>
2022-06-21 06:25:54 +08:00
extern const struct bench bench_local_storage_cache_seq_get;
extern const struct bench bench_local_storage_cache_interleaved_get;
extern const struct bench bench_local_storage_cache_hashmap_control;
selftests/bpf: Add benchmark for local_storage RCU Tasks Trace usage This benchmark measures grace period latency and kthread cpu usage of RCU Tasks Trace when many processes are creating/deleting BPF local_storage. Intent here is to quantify improvement on these metrics after Paul's recent RCU Tasks patches [0]. Specifically, fork 15k tasks which call a bpf prog that creates/destroys task local_storage and sleep in a loop, resulting in many call_rcu_tasks_trace calls. To determine grace period latency, trace time elapsed between rcu_tasks_trace_pregp_step and rcu_tasks_trace_postgp; for cpu usage look at rcu_task_trace_kthread's stime in /proc/PID/stat. On my virtualized test environment (Skylake, 8 cpus) benchmark results demonstrate significant improvement: BEFORE Paul's patches: SUMMARY tasks_trace grace period latency avg 22298.551 us stddev 1302.165 us SUMMARY ticks per tasks_trace grace period avg 2.291 stddev 0.324 AFTER Paul's patches: SUMMARY tasks_trace grace period latency avg 16969.197 us stddev 2525.053 us SUMMARY ticks per tasks_trace grace period avg 1.146 stddev 0.178 Note that since these patches are not in bpf-next benchmarking was done by cherry-picking this patch onto rcu tree. [0] https://lore.kernel.org/rcu/20220620225402.GA3842369@paulmck-ThinkPad-P17-Gen-1/ Signed-off-by: Dave Marchevsky <davemarchevsky@fb.com> Signed-off-by: Daniel Borkmann <daniel@iogearbox.net> Acked-by: Paul E. McKenney <paulmck@kernel.org> Acked-by: Martin KaFai Lau <kafai@fb.com> Link: https://lore.kernel.org/bpf/20220705190018.3239050-1-davemarchevsky@fb.com
2022-07-06 03:00:18 +08:00
extern const struct bench bench_local_storage_tasks_trace;
selftests/bpf: Add benchmark runner infrastructure While working on BPF ringbuf implementation, testing, and benchmarking, I've developed a pretty generic and modular benchmark runner, which seems to be generically useful, as I've already used it for one more purpose (testing fastest way to trigger BPF program, to minimize overhead of in-kernel code). This patch adds generic part of benchmark runner and sets up Makefile for extending it with more sets of benchmarks. Benchmarker itself operates by spinning up specified number of producer and consumer threads, setting up interval timer sending SIGALARM signal to application once a second. Every second, current snapshot with hits/drops counters are collected and stored in an array. Drops are useful for producer/consumer benchmarks in which producer might overwhelm consumers. Once test finishes after given amount of warm-up and testing seconds, mean and stddev are calculated (ignoring warm-up results) and is printed out to stdout. This setup seems to give consistent and accurate results. To validate behavior, I added two atomic counting tests: global and local. For global one, all the producer threads are atomically incrementing same counter as fast as possible. This, of course, leads to huge drop of performance once there is more than one producer thread due to CPUs fighting for the same memory location. Local counting, on the other hand, maintains one counter per each producer thread, incremented independently. Once per second, all counters are read and added together to form final "counting throughput" measurement. As expected, such setup demonstrates linear scalability with number of producers (as long as there are enough physical CPU cores, of course). See example output below. Also, this setup can nicely demonstrate disastrous effects of false sharing, if care is not taken to take those per-producer counters apart into independent cache lines. Demo output shows global counter first with 1 producer, then with 4. Both total and per-producer performance significantly drop. The last run is local counter with 4 producers, demonstrating near-perfect scalability. $ ./bench -a -w1 -d2 -p1 count-global Setting up benchmark 'count-global'... Benchmark 'count-global' started. Iter 0 ( 24.822us): hits 148.179M/s (148.179M/prod), drops 0.000M/s Iter 1 ( 37.939us): hits 149.308M/s (149.308M/prod), drops 0.000M/s Iter 2 (-10.774us): hits 150.717M/s (150.717M/prod), drops 0.000M/s Iter 3 ( 3.807us): hits 151.435M/s (151.435M/prod), drops 0.000M/s Summary: hits 150.488 ± 1.079M/s (150.488M/prod), drops 0.000 ± 0.000M/s $ ./bench -a -w1 -d2 -p4 count-global Setting up benchmark 'count-global'... Benchmark 'count-global' started. Iter 0 ( 60.659us): hits 53.910M/s ( 13.477M/prod), drops 0.000M/s Iter 1 (-17.658us): hits 53.722M/s ( 13.431M/prod), drops 0.000M/s Iter 2 ( 5.865us): hits 53.495M/s ( 13.374M/prod), drops 0.000M/s Iter 3 ( 0.104us): hits 53.606M/s ( 13.402M/prod), drops 0.000M/s Summary: hits 53.608 ± 0.113M/s ( 13.402M/prod), drops 0.000 ± 0.000M/s $ ./bench -a -w1 -d2 -p4 count-local Setting up benchmark 'count-local'... Benchmark 'count-local' started. Iter 0 ( 23.388us): hits 640.450M/s (160.113M/prod), drops 0.000M/s Iter 1 ( 2.291us): hits 605.661M/s (151.415M/prod), drops 0.000M/s Iter 2 ( -6.415us): hits 607.092M/s (151.773M/prod), drops 0.000M/s Iter 3 ( -1.361us): hits 601.796M/s (150.449M/prod), drops 0.000M/s Summary: hits 604.849 ± 2.739M/s (151.212M/prod), drops 0.000 ± 0.000M/s Benchmark runner supports setting thread affinity for producer and consumer threads. You can use -a flag for default CPU selection scheme, where first consumer gets CPU #0, next one gets CPU #1, and so on. Then producer threads pick up next CPU and increment one-by-one as well. But user can also specify a set of CPUs independently for producers and consumers with --prod-affinity 1,2-10,15 and --cons-affinity <set-of-cpus>. The latter allows to force producers and consumers to share same set of CPUs, if necessary. Signed-off-by: Andrii Nakryiko <andriin@fb.com> Signed-off-by: Alexei Starovoitov <ast@kernel.org> Acked-by: Yonghong Song <yhs@fb.com> Link: https://lore.kernel.org/bpf/20200512192445.2351848-3-andriin@fb.com
2020-05-13 03:24:43 +08:00
static const struct bench *benchs[] = {
&bench_count_global,
&bench_count_local,
selftest/bpf: Fmod_ret prog and implement test_overhead as part of bench Add fmod_ret BPF program to existing test_overhead selftest. Also re-implement user-space benchmarking part into benchmark runner to compare results. Results with ./bench are consistently somewhat lower than test_overhead's, but relative performance of various types of BPF programs stay consisten (e.g., kretprobe is noticeably slower). This slowdown seems to be coming from the fact that test_overhead is single-threaded, while benchmark always spins off at least one thread for producer. This has been confirmed by hacking multi-threaded test_overhead variant and also single-threaded bench variant. Resutls are below. run_bench_rename.sh script from benchs/ subdirectory was used to produce results for ./bench. Single-threaded implementations =============================== /* bench: single-threaded, atomics */ base : 4.622 ± 0.049M/s kprobe : 3.673 ± 0.052M/s kretprobe : 2.625 ± 0.052M/s rawtp : 4.369 ± 0.089M/s fentry : 4.201 ± 0.558M/s fexit : 4.309 ± 0.148M/s fmodret : 4.314 ± 0.203M/s /* selftest: single-threaded, no atomics */ task_rename base 4555K events per sec task_rename kprobe 3643K events per sec task_rename kretprobe 2506K events per sec task_rename raw_tp 4303K events per sec task_rename fentry 4307K events per sec task_rename fexit 4010K events per sec task_rename fmod_ret 3984K events per sec Multi-threaded implementations ============================== /* bench: multi-threaded w/ atomics */ base : 3.910 ± 0.023M/s kprobe : 3.048 ± 0.037M/s kretprobe : 2.300 ± 0.015M/s rawtp : 3.687 ± 0.034M/s fentry : 3.740 ± 0.087M/s fexit : 3.510 ± 0.009M/s fmodret : 3.485 ± 0.050M/s /* selftest: multi-threaded w/ atomics */ task_rename base 3872K events per sec task_rename kprobe 3068K events per sec task_rename kretprobe 2350K events per sec task_rename raw_tp 3731K events per sec task_rename fentry 3639K events per sec task_rename fexit 3558K events per sec task_rename fmod_ret 3511K events per sec /* selftest: multi-threaded, no atomics */ task_rename base 3945K events per sec task_rename kprobe 3298K events per sec task_rename kretprobe 2451K events per sec task_rename raw_tp 3718K events per sec task_rename fentry 3782K events per sec task_rename fexit 3543K events per sec task_rename fmod_ret 3526K events per sec Note that the fact that ./bench benchmark always uses atomic increments for counting, while test_overhead doesn't, doesn't influence test results all that much. Signed-off-by: Andrii Nakryiko <andriin@fb.com> Signed-off-by: Alexei Starovoitov <ast@kernel.org> Acked-by: John Fastabend <john.fastabend@gmail.com> Acked-by: Yonghong Song <yhs@fb.com> Link: https://lore.kernel.org/bpf/20200512192445.2351848-4-andriin@fb.com
2020-05-13 03:24:44 +08:00
&bench_rename_base,
&bench_rename_kprobe,
&bench_rename_kretprobe,
&bench_rename_rawtp,
&bench_rename_fentry,
&bench_rename_fexit,
selftest/bpf: Add BPF triggering benchmark It is sometimes desirable to be able to trigger BPF program from user-space with minimal overhead. sys_enter would seem to be a good candidate, yet in a lot of cases there will be a lot of noise from syscalls triggered by other processes on the system. So while searching for low-overhead alternative, I've stumbled upon getpgid() syscall, which seems to be specific enough to not suffer from accidental syscall by other apps. This set of benchmarks compares tp, raw_tp w/ filtering by syscall ID, kprobe, fentry and fmod_ret with returning error (so that syscall would not be executed), to determine the lowest-overhead way. Here are results on my machine (using benchs/run_bench_trigger.sh script): base : 9.200 ± 0.319M/s tp : 6.690 ± 0.125M/s rawtp : 8.571 ± 0.214M/s kprobe : 6.431 ± 0.048M/s fentry : 8.955 ± 0.241M/s fmodret : 8.903 ± 0.135M/s So it seems like fmodret doesn't give much benefit for such lightweight syscall. Raw tracepoint is pretty decent despite additional filtering logic, but it will be called for any other syscall in the system, which rules it out. Fentry, though, seems to be adding the least amoung of overhead and achieves 97.3% of performance of baseline no-BPF-attached syscall. Using getpgid() seems to be preferable to set_task_comm() approach from test_overhead, as it's about 2.35x faster in a baseline performance. Signed-off-by: Andrii Nakryiko <andriin@fb.com> Signed-off-by: Alexei Starovoitov <ast@kernel.org> Acked-by: John Fastabend <john.fastabend@gmail.com> Acked-by: Yonghong Song <yhs@fb.com> Link: https://lore.kernel.org/bpf/20200512192445.2351848-5-andriin@fb.com
2020-05-13 03:24:45 +08:00
&bench_trig_base,
&bench_trig_tp,
&bench_trig_rawtp,
&bench_trig_kprobe,
&bench_trig_fentry,
selftests/bpf: Add sleepable tests Modify few tests to sanity test sleepable bpf functionality. Running 'bench trig-fentry-sleep' vs 'bench trig-fentry' and 'perf report': sleepable with SRCU: 3.86% bench [k] __srcu_read_unlock 3.22% bench [k] __srcu_read_lock 0.92% bench [k] bpf_prog_740d4210cdcd99a3_bench_trigger_fentry_sleep 0.50% bench [k] bpf_trampoline_10297 0.26% bench [k] __bpf_prog_exit_sleepable 0.21% bench [k] __bpf_prog_enter_sleepable sleepable with RCU_TRACE: 0.79% bench [k] bpf_prog_740d4210cdcd99a3_bench_trigger_fentry_sleep 0.72% bench [k] bpf_trampoline_10381 0.31% bench [k] __bpf_prog_exit_sleepable 0.29% bench [k] __bpf_prog_enter_sleepable non-sleepable with RCU: 0.88% bench [k] bpf_prog_740d4210cdcd99a3_bench_trigger_fentry 0.84% bench [k] bpf_trampoline_10297 0.13% bench [k] __bpf_prog_enter 0.12% bench [k] __bpf_prog_exit Signed-off-by: Alexei Starovoitov <ast@kernel.org> Signed-off-by: Daniel Borkmann <daniel@iogearbox.net> Acked-by: KP Singh <kpsingh@google.com> Link: https://lore.kernel.org/bpf/20200827220114.69225-6-alexei.starovoitov@gmail.com
2020-08-28 06:01:14 +08:00
&bench_trig_fentry_sleep,
selftest/bpf: Add BPF triggering benchmark It is sometimes desirable to be able to trigger BPF program from user-space with minimal overhead. sys_enter would seem to be a good candidate, yet in a lot of cases there will be a lot of noise from syscalls triggered by other processes on the system. So while searching for low-overhead alternative, I've stumbled upon getpgid() syscall, which seems to be specific enough to not suffer from accidental syscall by other apps. This set of benchmarks compares tp, raw_tp w/ filtering by syscall ID, kprobe, fentry and fmod_ret with returning error (so that syscall would not be executed), to determine the lowest-overhead way. Here are results on my machine (using benchs/run_bench_trigger.sh script): base : 9.200 ± 0.319M/s tp : 6.690 ± 0.125M/s rawtp : 8.571 ± 0.214M/s kprobe : 6.431 ± 0.048M/s fentry : 8.955 ± 0.241M/s fmodret : 8.903 ± 0.135M/s So it seems like fmodret doesn't give much benefit for such lightweight syscall. Raw tracepoint is pretty decent despite additional filtering logic, but it will be called for any other syscall in the system, which rules it out. Fentry, though, seems to be adding the least amoung of overhead and achieves 97.3% of performance of baseline no-BPF-attached syscall. Using getpgid() seems to be preferable to set_task_comm() approach from test_overhead, as it's about 2.35x faster in a baseline performance. Signed-off-by: Andrii Nakryiko <andriin@fb.com> Signed-off-by: Alexei Starovoitov <ast@kernel.org> Acked-by: John Fastabend <john.fastabend@gmail.com> Acked-by: Yonghong Song <yhs@fb.com> Link: https://lore.kernel.org/bpf/20200512192445.2351848-5-andriin@fb.com
2020-05-13 03:24:45 +08:00
&bench_trig_fmodret,
selftests/bpf: Add uprobe triggering overhead benchmarks Add benchmark to measure overhead of uprobes and uretprobes. Also have a baseline (no uprobe attached) benchmark. On my dev machine, baseline benchmark can trigger 130M user_target() invocations. When uprobe is attached, this falls to just 700K. With uretprobe, we get down to 520K: $ sudo ./bench trig-uprobe-base -a Summary: hits 131.289 ± 2.872M/s # UPROBE $ sudo ./bench -a trig-uprobe-without-nop Summary: hits 0.729 ± 0.007M/s $ sudo ./bench -a trig-uprobe-with-nop Summary: hits 1.798 ± 0.017M/s # URETPROBE $ sudo ./bench -a trig-uretprobe-without-nop Summary: hits 0.508 ± 0.012M/s $ sudo ./bench -a trig-uretprobe-with-nop Summary: hits 0.883 ± 0.008M/s So there is almost 2.5x performance difference between probing nop vs non-nop instruction for entry uprobe. And 1.7x difference for uretprobe. This means that non-nop uprobe overhead is around 1.4 microseconds for uprobe and 2 microseconds for non-nop uretprobe. For nop variants, uprobe and uretprobe overhead is down to 0.556 and 1.13 microseconds, respectively. For comparison, just doing a very low-overhead syscall (with no BPF programs attached anywhere) gives: $ sudo ./bench trig-base -a Summary: hits 4.830 ± 0.036M/s So uprobes are about 2.67x slower than pure context switch. Signed-off-by: Andrii Nakryiko <andrii@kernel.org> Signed-off-by: Daniel Borkmann <daniel@iogearbox.net> Link: https://lore.kernel.org/bpf/20211116013041.4072571-1-andrii@kernel.org
2021-11-16 09:30:41 +08:00
&bench_trig_uprobe_base,
&bench_trig_uprobe_with_nop,
&bench_trig_uretprobe_with_nop,
&bench_trig_uprobe_without_nop,
&bench_trig_uretprobe_without_nop,
bpf: Add BPF ringbuf and perf buffer benchmarks Extend bench framework with ability to have benchmark-provided child argument parser for custom benchmark-specific parameters. This makes bench generic code modular and independent from any specific benchmark. Also implement a set of benchmarks for new BPF ring buffer and existing perf buffer. 4 benchmarks were implemented: 2 variations for each of BPF ringbuf and perfbuf:, - rb-libbpf utilizes stock libbpf ring_buffer manager for reading data; - rb-custom implements custom ring buffer setup and reading code, to eliminate overheads inherent in generic libbpf code due to callback functions and the need to update consumer position after each consumed record, instead of batching updates (due to pessimistic assumption that user callback might take long time and thus could unnecessarily hold ring buffer space for too long); - pb-libbpf uses stock libbpf perf_buffer code with all the default settings, though uses higher-performance raw event callback to minimize unnecessary overhead; - pb-custom implements its own custom consumer code to minimize any possible overhead of generic libbpf implementation and indirect function calls. All of the test support default, no data notification skipped, mode, as well as sampled mode (with --rb-sampled flag), which allows to trigger epoll notification less frequently and reduce overhead. As will be shown, this mode is especially critical for perf buffer, which suffers from high overhead of wakeups in kernel. Otherwise, all benchamrks implement similar way to generate a batch of records by using fentry/sys_getpgid BPF program, which pushes a bunch of records in a tight loop and records number of successful and dropped samples. Each record is a small 8-byte integer, to minimize the effect of memory copying with bpf_perf_event_output() and bpf_ringbuf_output(). Benchmarks that have only one producer implement optional back-to-back mode, in which record production and consumption is alternating on the same CPU. This is the highest-throughput happy case, showing ultimate performance achievable with either BPF ringbuf or perfbuf. All the below scenarios are implemented in a script in benchs/run_bench_ringbufs.sh. Tests were performed on 28-core/56-thread Intel Xeon CPU E5-2680 v4 @ 2.40GHz CPU. Single-producer, parallel producer ================================== rb-libbpf 12.054 ± 0.320M/s (drops 0.000 ± 0.000M/s) rb-custom 8.158 ± 0.118M/s (drops 0.001 ± 0.003M/s) pb-libbpf 0.931 ± 0.007M/s (drops 0.000 ± 0.000M/s) pb-custom 0.965 ± 0.003M/s (drops 0.000 ± 0.000M/s) Single-producer, parallel producer, sampled notification ======================================================== rb-libbpf 11.563 ± 0.067M/s (drops 0.000 ± 0.000M/s) rb-custom 15.895 ± 0.076M/s (drops 0.000 ± 0.000M/s) pb-libbpf 9.889 ± 0.032M/s (drops 0.000 ± 0.000M/s) pb-custom 9.866 ± 0.028M/s (drops 0.000 ± 0.000M/s) Single producer on one CPU, consumer on another one, both running at full speed. Curiously, rb-libbpf has higher throughput than objectively faster (due to more lightweight consumer code path) rb-custom. It appears that faster consumer causes kernel to send notifications more frequently, because consumer appears to be caught up more frequently. Performance of perfbuf suffers from default "no sampling" policy and huge overhead that causes. In sampled mode, rb-custom is winning very significantly eliminating too frequent in-kernel wakeups, the gain appears to be more than 2x. Perf buffer achieves even more impressive wins, compared to stock perfbuf settings, with 10x improvements in throughput with 1:500 sampling rate. The trade-off is that with sampling, application might not get next X events until X+1st arrives, which is not always acceptable. With steady influx of events, though, this shouldn't be a problem. Overall, single-producer performance of ring buffers seems to be better no matter the sampled/non-sampled modes, but it especially beats ring buffer without sampling due to its adaptive notification approach. Single-producer, back-to-back mode ================================== rb-libbpf 15.507 ± 0.247M/s (drops 0.000 ± 0.000M/s) rb-libbpf-sampled 14.692 ± 0.195M/s (drops 0.000 ± 0.000M/s) rb-custom 21.449 ± 0.157M/s (drops 0.000 ± 0.000M/s) rb-custom-sampled 20.024 ± 0.386M/s (drops 0.000 ± 0.000M/s) pb-libbpf 1.601 ± 0.015M/s (drops 0.000 ± 0.000M/s) pb-libbpf-sampled 8.545 ± 0.064M/s (drops 0.000 ± 0.000M/s) pb-custom 1.607 ± 0.022M/s (drops 0.000 ± 0.000M/s) pb-custom-sampled 8.988 ± 0.144M/s (drops 0.000 ± 0.000M/s) Here we test a back-to-back mode, which is arguably best-case scenario both for BPF ringbuf and perfbuf, because there is no contention and for ringbuf also no excessive notification, because consumer appears to be behind after the first record. For ringbuf, custom consumer code clearly wins with 21.5 vs 16 million records per second exchanged between producer and consumer. Sampled mode actually hurts a bit due to slightly slower producer logic (it needs to fetch amount of data available to decide whether to skip or force notification). Perfbuf with wakeup sampling gets 5.5x throughput increase, compared to no-sampling version. There also doesn't seem to be noticeable overhead from generic libbpf handling code. Perfbuf back-to-back, effect of sample rate =========================================== pb-sampled-1 1.035 ± 0.012M/s (drops 0.000 ± 0.000M/s) pb-sampled-5 3.476 ± 0.087M/s (drops 0.000 ± 0.000M/s) pb-sampled-10 5.094 ± 0.136M/s (drops 0.000 ± 0.000M/s) pb-sampled-25 7.118 ± 0.153M/s (drops 0.000 ± 0.000M/s) pb-sampled-50 8.169 ± 0.156M/s (drops 0.000 ± 0.000M/s) pb-sampled-100 8.887 ± 0.136M/s (drops 0.000 ± 0.000M/s) pb-sampled-250 9.180 ± 0.209M/s (drops 0.000 ± 0.000M/s) pb-sampled-500 9.353 ± 0.281M/s (drops 0.000 ± 0.000M/s) pb-sampled-1000 9.411 ± 0.217M/s (drops 0.000 ± 0.000M/s) pb-sampled-2000 9.464 ± 0.167M/s (drops 0.000 ± 0.000M/s) pb-sampled-3000 9.575 ± 0.273M/s (drops 0.000 ± 0.000M/s) This benchmark shows the effect of event sampling for perfbuf. Back-to-back mode for highest throughput. Just doing every 5th record notification gives 3.5x speed up. 250-500 appears to be the point of diminishing return, with almost 9x speed up. Most benchmarks use 500 as the default sampling for pb-raw and pb-custom. Ringbuf back-to-back, effect of sample rate =========================================== rb-sampled-1 1.106 ± 0.010M/s (drops 0.000 ± 0.000M/s) rb-sampled-5 4.746 ± 0.149M/s (drops 0.000 ± 0.000M/s) rb-sampled-10 7.706 ± 0.164M/s (drops 0.000 ± 0.000M/s) rb-sampled-25 12.893 ± 0.273M/s (drops 0.000 ± 0.000M/s) rb-sampled-50 15.961 ± 0.361M/s (drops 0.000 ± 0.000M/s) rb-sampled-100 18.203 ± 0.445M/s (drops 0.000 ± 0.000M/s) rb-sampled-250 19.962 ± 0.786M/s (drops 0.000 ± 0.000M/s) rb-sampled-500 20.881 ± 0.551M/s (drops 0.000 ± 0.000M/s) rb-sampled-1000 21.317 ± 0.532M/s (drops 0.000 ± 0.000M/s) rb-sampled-2000 21.331 ± 0.535M/s (drops 0.000 ± 0.000M/s) rb-sampled-3000 21.688 ± 0.392M/s (drops 0.000 ± 0.000M/s) Similar benchmark for ring buffer also shows a great advantage (in terms of throughput) of skipping notifications. Skipping every 5th one gives 4x boost. Also similar to perfbuf case, 250-500 seems to be the point of diminishing returns, giving roughly 20x better results. Keep in mind, for this test, notifications are controlled manually with BPF_RB_NO_WAKEUP and BPF_RB_FORCE_WAKEUP. As can be seen from previous benchmarks, adaptive notifications based on consumer's positions provides same (or even slightly better due to simpler load generator on BPF side) benefits in favorable back-to-back scenario. Over zealous and fast consumer, which is almost always caught up, will make thoughput numbers smaller. That's the case when manual notification control might prove to be extremely beneficial. Ringbuf back-to-back, reserve+commit vs output ============================================== reserve 22.819 ± 0.503M/s (drops 0.000 ± 0.000M/s) output 18.906 ± 0.433M/s (drops 0.000 ± 0.000M/s) Ringbuf sampled, reserve+commit vs output ========================================= reserve-sampled 15.350 ± 0.132M/s (drops 0.000 ± 0.000M/s) output-sampled 14.195 ± 0.144M/s (drops 0.000 ± 0.000M/s) BPF ringbuf supports two sets of APIs with various usability and performance tradeoffs: bpf_ringbuf_reserve()+bpf_ringbuf_commit() vs bpf_ringbuf_output(). This benchmark clearly shows superiority of reserve+commit approach, despite using a small 8-byte record size. Single-producer, consumer/producer competing on the same CPU, low batch count ============================================================================= rb-libbpf 3.045 ± 0.020M/s (drops 3.536 ± 0.148M/s) rb-custom 3.055 ± 0.022M/s (drops 3.893 ± 0.066M/s) pb-libbpf 1.393 ± 0.024M/s (drops 0.000 ± 0.000M/s) pb-custom 1.407 ± 0.016M/s (drops 0.000 ± 0.000M/s) This benchmark shows one of the worst-case scenarios, in which producer and consumer do not coordinate *and* fight for the same CPU. No batch count and sampling settings were able to eliminate drops for ringbuffer, producer is just too fast for consumer to keep up. But ringbuf and perfbuf still able to pass through quite a lot of messages, which is more than enough for a lot of applications. Ringbuf, multi-producer contention ================================== rb-libbpf nr_prod 1 10.916 ± 0.399M/s (drops 0.000 ± 0.000M/s) rb-libbpf nr_prod 2 4.931 ± 0.030M/s (drops 0.000 ± 0.000M/s) rb-libbpf nr_prod 3 4.880 ± 0.006M/s (drops 0.000 ± 0.000M/s) rb-libbpf nr_prod 4 3.926 ± 0.004M/s (drops 0.000 ± 0.000M/s) rb-libbpf nr_prod 8 4.011 ± 0.004M/s (drops 0.000 ± 0.000M/s) rb-libbpf nr_prod 12 3.967 ± 0.016M/s (drops 0.000 ± 0.000M/s) rb-libbpf nr_prod 16 2.604 ± 0.030M/s (drops 0.001 ± 0.002M/s) rb-libbpf nr_prod 20 2.233 ± 0.003M/s (drops 0.000 ± 0.000M/s) rb-libbpf nr_prod 24 2.085 ± 0.015M/s (drops 0.000 ± 0.000M/s) rb-libbpf nr_prod 28 2.055 ± 0.004M/s (drops 0.000 ± 0.000M/s) rb-libbpf nr_prod 32 1.962 ± 0.004M/s (drops 0.000 ± 0.000M/s) rb-libbpf nr_prod 36 2.089 ± 0.005M/s (drops 0.000 ± 0.000M/s) rb-libbpf nr_prod 40 2.118 ± 0.006M/s (drops 0.000 ± 0.000M/s) rb-libbpf nr_prod 44 2.105 ± 0.004M/s (drops 0.000 ± 0.000M/s) rb-libbpf nr_prod 48 2.120 ± 0.058M/s (drops 0.000 ± 0.001M/s) rb-libbpf nr_prod 52 2.074 ± 0.024M/s (drops 0.007 ± 0.014M/s) Ringbuf uses a very short-duration spinlock during reservation phase, to check few invariants, increment producer count and set record header. This is the biggest point of contention for ringbuf implementation. This benchmark evaluates the effect of multiple competing writers on overall throughput of a single shared ringbuffer. Overall throughput drops almost 2x when going from single to two highly-contended producers, gradually dropping with additional competing producers. Performance drop stabilizes at around 20 producers and hovers around 2mln even with 50+ fighting producers, which is a 5x drop compared to non-contended case. Good kernel implementation in kernel helps maintain decent performance here. Note, that in the intended real-world scenarios, it's not expected to get even close to such a high levels of contention. But if contention will become a problem, there is always an option of sharding few ring buffers across a set of CPUs. Signed-off-by: Andrii Nakryiko <andriin@fb.com> Signed-off-by: Daniel Borkmann <daniel@iogearbox.net> Link: https://lore.kernel.org/bpf/20200529075424.3139988-5-andriin@fb.com Signed-off-by: Alexei Starovoitov <ast@kernel.org>
2020-05-29 15:54:23 +08:00
&bench_rb_libbpf,
&bench_rb_custom,
&bench_pb_libbpf,
&bench_pb_custom,
bpf/benchs: Add benchmark tests for bloom filter throughput + false positive This patch adds benchmark tests for the throughput (for lookups + updates) and the false positive rate of bloom filter lookups, as well as some minor refactoring of the bash script for running the benchmarks. These benchmarks show that as the number of hash functions increases, the throughput and the false positive rate of the bloom filter decreases. >From the benchmark data, the approximate average false-positive rates are roughly as follows: 1 hash function = ~30% 2 hash functions = ~15% 3 hash functions = ~5% 4 hash functions = ~2.5% 5 hash functions = ~1% 6 hash functions = ~0.5% 7 hash functions = ~0.35% 8 hash functions = ~0.15% 9 hash functions = ~0.1% 10 hash functions = ~0% For reference data, the benchmarks run on one thread on a machine with one numa node for 1 to 5 hash functions for 8-byte and 64-byte values are as follows: 1 hash function: 50k entries 8-byte value Lookups - 51.1 M/s operations Updates - 33.6 M/s operations False positive rate: 24.15% 64-byte value Lookups - 15.7 M/s operations Updates - 15.1 M/s operations False positive rate: 24.2% 100k entries 8-byte value Lookups - 51.0 M/s operations Updates - 33.4 M/s operations False positive rate: 24.04% 64-byte value Lookups - 15.6 M/s operations Updates - 14.6 M/s operations False positive rate: 24.06% 500k entries 8-byte value Lookups - 50.5 M/s operations Updates - 33.1 M/s operations False positive rate: 27.45% 64-byte value Lookups - 15.6 M/s operations Updates - 14.2 M/s operations False positive rate: 27.42% 1 mil entries 8-byte value Lookups - 49.7 M/s operations Updates - 32.9 M/s operations False positive rate: 27.45% 64-byte value Lookups - 15.4 M/s operations Updates - 13.7 M/s operations False positive rate: 27.58% 2.5 mil entries 8-byte value Lookups - 47.2 M/s operations Updates - 31.8 M/s operations False positive rate: 30.94% 64-byte value Lookups - 15.3 M/s operations Updates - 13.2 M/s operations False positive rate: 30.95% 5 mil entries 8-byte value Lookups - 41.1 M/s operations Updates - 28.1 M/s operations False positive rate: 31.01% 64-byte value Lookups - 13.3 M/s operations Updates - 11.4 M/s operations False positive rate: 30.98% 2 hash functions: 50k entries 8-byte value Lookups - 34.1 M/s operations Updates - 20.1 M/s operations False positive rate: 9.13% 64-byte value Lookups - 8.4 M/s operations Updates - 7.9 M/s operations False positive rate: 9.21% 100k entries 8-byte value Lookups - 33.7 M/s operations Updates - 18.9 M/s operations False positive rate: 9.13% 64-byte value Lookups - 8.4 M/s operations Updates - 7.7 M/s operations False positive rate: 9.19% 500k entries 8-byte value Lookups - 32.7 M/s operations Updates - 18.1 M/s operations False positive rate: 12.61% 64-byte value Lookups - 8.4 M/s operations Updates - 7.5 M/s operations False positive rate: 12.61% 1 mil entries 8-byte value Lookups - 30.6 M/s operations Updates - 18.9 M/s operations False positive rate: 12.54% 64-byte value Lookups - 8.0 M/s operations Updates - 7.0 M/s operations False positive rate: 12.52% 2.5 mil entries 8-byte value Lookups - 25.3 M/s operations Updates - 16.7 M/s operations False positive rate: 16.77% 64-byte value Lookups - 7.9 M/s operations Updates - 6.5 M/s operations False positive rate: 16.88% 5 mil entries 8-byte value Lookups - 20.8 M/s operations Updates - 14.7 M/s operations False positive rate: 16.78% 64-byte value Lookups - 7.0 M/s operations Updates - 6.0 M/s operations False positive rate: 16.78% 3 hash functions: 50k entries 8-byte value Lookups - 25.1 M/s operations Updates - 14.6 M/s operations False positive rate: 7.65% 64-byte value Lookups - 5.8 M/s operations Updates - 5.5 M/s operations False positive rate: 7.58% 100k entries 8-byte value Lookups - 24.7 M/s operations Updates - 14.1 M/s operations False positive rate: 7.71% 64-byte value Lookups - 5.8 M/s operations Updates - 5.3 M/s operations False positive rate: 7.62% 500k entries 8-byte value Lookups - 22.9 M/s operations Updates - 13.9 M/s operations False positive rate: 2.62% 64-byte value Lookups - 5.6 M/s operations Updates - 4.8 M/s operations False positive rate: 2.7% 1 mil entries 8-byte value Lookups - 19.8 M/s operations Updates - 12.6 M/s operations False positive rate: 2.60% 64-byte value Lookups - 5.3 M/s operations Updates - 4.4 M/s operations False positive rate: 2.69% 2.5 mil entries 8-byte value Lookups - 16.2 M/s operations Updates - 10.7 M/s operations False positive rate: 4.49% 64-byte value Lookups - 4.9 M/s operations Updates - 4.1 M/s operations False positive rate: 4.41% 5 mil entries 8-byte value Lookups - 18.8 M/s operations Updates - 9.2 M/s operations False positive rate: 4.45% 64-byte value Lookups - 5.2 M/s operations Updates - 3.9 M/s operations False positive rate: 4.54% 4 hash functions: 50k entries 8-byte value Lookups - 19.7 M/s operations Updates - 11.1 M/s operations False positive rate: 1.01% 64-byte value Lookups - 4.4 M/s operations Updates - 4.0 M/s operations False positive rate: 1.00% 100k entries 8-byte value Lookups - 19.5 M/s operations Updates - 10.9 M/s operations False positive rate: 1.00% 64-byte value Lookups - 4.3 M/s operations Updates - 3.9 M/s operations False positive rate: 0.97% 500k entries 8-byte value Lookups - 18.2 M/s operations Updates - 10.6 M/s operations False positive rate: 2.05% 64-byte value Lookups - 4.3 M/s operations Updates - 3.7 M/s operations False positive rate: 2.05% 1 mil entries 8-byte value Lookups - 15.5 M/s operations Updates - 9.6 M/s operations False positive rate: 1.99% 64-byte value Lookups - 4.0 M/s operations Updates - 3.4 M/s operations False positive rate: 1.99% 2.5 mil entries 8-byte value Lookups - 13.8 M/s operations Updates - 7.7 M/s operations False positive rate: 3.91% 64-byte value Lookups - 3.7 M/s operations Updates - 3.6 M/s operations False positive rate: 3.78% 5 mil entries 8-byte value Lookups - 13.0 M/s operations Updates - 6.9 M/s operations False positive rate: 3.93% 64-byte value Lookups - 3.5 M/s operations Updates - 3.7 M/s operations False positive rate: 3.39% 5 hash functions: 50k entries 8-byte value Lookups - 16.4 M/s operations Updates - 9.1 M/s operations False positive rate: 0.78% 64-byte value Lookups - 3.5 M/s operations Updates - 3.2 M/s operations False positive rate: 0.77% 100k entries 8-byte value Lookups - 16.3 M/s operations Updates - 9.0 M/s operations False positive rate: 0.79% 64-byte value Lookups - 3.5 M/s operations Updates - 3.2 M/s operations False positive rate: 0.78% 500k entries 8-byte value Lookups - 15.1 M/s operations Updates - 8.8 M/s operations False positive rate: 1.82% 64-byte value Lookups - 3.4 M/s operations Updates - 3.0 M/s operations False positive rate: 1.78% 1 mil entries 8-byte value Lookups - 13.2 M/s operations Updates - 7.8 M/s operations False positive rate: 1.81% 64-byte value Lookups - 3.2 M/s operations Updates - 2.8 M/s operations False positive rate: 1.80% 2.5 mil entries 8-byte value Lookups - 10.5 M/s operations Updates - 5.9 M/s operations False positive rate: 0.29% 64-byte value Lookups - 3.2 M/s operations Updates - 2.4 M/s operations False positive rate: 0.28% 5 mil entries 8-byte value Lookups - 9.6 M/s operations Updates - 5.7 M/s operations False positive rate: 0.30% 64-byte value Lookups - 3.2 M/s operations Updates - 2.7 M/s operations False positive rate: 0.30% Signed-off-by: Joanne Koong <joannekoong@fb.com> Signed-off-by: Alexei Starovoitov <ast@kernel.org> Acked-by: Andrii Nakryiko <andrii@kernel.org> Link: https://lore.kernel.org/bpf/20211027234504.30744-5-joannekoong@fb.com
2021-10-28 07:45:03 +08:00
&bench_bloom_lookup,
&bench_bloom_update,
&bench_bloom_false_positive,
bpf/benchs: Add benchmarks for comparing hashmap lookups w/ vs. w/out bloom filter This patch adds benchmark tests for comparing the performance of hashmap lookups without the bloom filter vs. hashmap lookups with the bloom filter. Checking the bloom filter first for whether the element exists should overall enable a higher throughput for hashmap lookups, since if the element does not exist in the bloom filter, we can avoid a costly lookup in the hashmap. On average, using 5 hash functions in the bloom filter tended to perform the best across the widest range of different entry sizes. The benchmark results using 5 hash functions (running on 8 threads on a machine with one numa node, and taking the average of 3 runs) were roughly as follows: value_size = 4 bytes - 10k entries: 30% faster 50k entries: 40% faster 100k entries: 40% faster 500k entres: 70% faster 1 million entries: 90% faster 5 million entries: 140% faster value_size = 8 bytes - 10k entries: 30% faster 50k entries: 40% faster 100k entries: 50% faster 500k entres: 80% faster 1 million entries: 100% faster 5 million entries: 150% faster value_size = 16 bytes - 10k entries: 20% faster 50k entries: 30% faster 100k entries: 35% faster 500k entres: 65% faster 1 million entries: 85% faster 5 million entries: 110% faster value_size = 40 bytes - 10k entries: 5% faster 50k entries: 15% faster 100k entries: 20% faster 500k entres: 65% faster 1 million entries: 75% faster 5 million entries: 120% faster Signed-off-by: Joanne Koong <joannekoong@fb.com> Signed-off-by: Alexei Starovoitov <ast@kernel.org> Link: https://lore.kernel.org/bpf/20211027234504.30744-6-joannekoong@fb.com
2021-10-28 07:45:04 +08:00
&bench_hashmap_without_bloom,
&bench_hashmap_with_bloom,
selftest/bpf/benchs: Add bpf_loop benchmark Add benchmark to measure the throughput and latency of the bpf_loop call. Testing this on my dev machine on 1 thread, the data is as follows: nr_loops: 10 bpf_loop - throughput: 198.519 ± 0.155 M ops/s, latency: 5.037 ns/op nr_loops: 100 bpf_loop - throughput: 247.448 ± 0.305 M ops/s, latency: 4.041 ns/op nr_loops: 500 bpf_loop - throughput: 260.839 ± 0.380 M ops/s, latency: 3.834 ns/op nr_loops: 1000 bpf_loop - throughput: 262.806 ± 0.629 M ops/s, latency: 3.805 ns/op nr_loops: 5000 bpf_loop - throughput: 264.211 ± 1.508 M ops/s, latency: 3.785 ns/op nr_loops: 10000 bpf_loop - throughput: 265.366 ± 3.054 M ops/s, latency: 3.768 ns/op nr_loops: 50000 bpf_loop - throughput: 235.986 ± 20.205 M ops/s, latency: 4.238 ns/op nr_loops: 100000 bpf_loop - throughput: 264.482 ± 0.279 M ops/s, latency: 3.781 ns/op nr_loops: 500000 bpf_loop - throughput: 309.773 ± 87.713 M ops/s, latency: 3.228 ns/op nr_loops: 1000000 bpf_loop - throughput: 262.818 ± 4.143 M ops/s, latency: 3.805 ns/op >From this data, we can see that the latency per loop decreases as the number of loops increases. On this particular machine, each loop had an overhead of about ~4 ns, and we were able to run ~250 million loops per second. Signed-off-by: Joanne Koong <joannekoong@fb.com> Signed-off-by: Alexei Starovoitov <ast@kernel.org> Acked-by: Andrii Nakryiko <andrii@kernel.org> Link: https://lore.kernel.org/bpf/20211130030622.4131246-5-joannekoong@fb.com
2021-11-30 11:06:22 +08:00
&bench_bpf_loop,
selftests/bpf: Add benchmark for bpf_strncmp() helper Add benchmark to compare the performance between home-made strncmp() in bpf program and bpf_strncmp() helper. In summary, the performance win of bpf_strncmp() under x86-64 is greater than 18% when the compared string length is greater than 64, and is 179% when the length is 4095. Under arm64 the performance win is even bigger: 33% when the length is greater than 64 and 600% when the length is 4095. The following is the details: no-helper-X: use home-made strncmp() to compare X-sized string helper-Y: use bpf_strncmp() to compare Y-sized string Under x86-64: no-helper-1 3.504 ± 0.000M/s (drops 0.000 ± 0.000M/s) helper-1 3.347 ± 0.001M/s (drops 0.000 ± 0.000M/s) no-helper-8 3.357 ± 0.001M/s (drops 0.000 ± 0.000M/s) helper-8 3.307 ± 0.001M/s (drops 0.000 ± 0.000M/s) no-helper-32 3.064 ± 0.000M/s (drops 0.000 ± 0.000M/s) helper-32 3.253 ± 0.001M/s (drops 0.000 ± 0.000M/s) no-helper-64 2.563 ± 0.001M/s (drops 0.000 ± 0.000M/s) helper-64 3.040 ± 0.001M/s (drops 0.000 ± 0.000M/s) no-helper-128 1.975 ± 0.000M/s (drops 0.000 ± 0.000M/s) helper-128 2.641 ± 0.000M/s (drops 0.000 ± 0.000M/s) no-helper-512 0.759 ± 0.000M/s (drops 0.000 ± 0.000M/s) helper-512 1.574 ± 0.000M/s (drops 0.000 ± 0.000M/s) no-helper-2048 0.329 ± 0.000M/s (drops 0.000 ± 0.000M/s) helper-2048 0.602 ± 0.000M/s (drops 0.000 ± 0.000M/s) no-helper-4095 0.117 ± 0.000M/s (drops 0.000 ± 0.000M/s) helper-4095 0.327 ± 0.000M/s (drops 0.000 ± 0.000M/s) Under arm64: no-helper-1 2.806 ± 0.004M/s (drops 0.000 ± 0.000M/s) helper-1 2.819 ± 0.002M/s (drops 0.000 ± 0.000M/s) no-helper-8 2.797 ± 0.109M/s (drops 0.000 ± 0.000M/s) helper-8 2.786 ± 0.025M/s (drops 0.000 ± 0.000M/s) no-helper-32 2.399 ± 0.011M/s (drops 0.000 ± 0.000M/s) helper-32 2.703 ± 0.002M/s (drops 0.000 ± 0.000M/s) no-helper-64 2.020 ± 0.015M/s (drops 0.000 ± 0.000M/s) helper-64 2.702 ± 0.073M/s (drops 0.000 ± 0.000M/s) no-helper-128 1.604 ± 0.001M/s (drops 0.000 ± 0.000M/s) helper-128 2.516 ± 0.002M/s (drops 0.000 ± 0.000M/s) no-helper-512 0.699 ± 0.000M/s (drops 0.000 ± 0.000M/s) helper-512 2.106 ± 0.003M/s (drops 0.000 ± 0.000M/s) no-helper-2048 0.215 ± 0.000M/s (drops 0.000 ± 0.000M/s) helper-2048 1.223 ± 0.003M/s (drops 0.000 ± 0.000M/s) no-helper-4095 0.112 ± 0.000M/s (drops 0.000 ± 0.000M/s) helper-4095 0.796 ± 0.000M/s (drops 0.000 ± 0.000M/s) Signed-off-by: Hou Tao <houtao1@huawei.com> Signed-off-by: Alexei Starovoitov <ast@kernel.org> Link: https://lore.kernel.org/bpf/20211210141652.877186-4-houtao1@huawei.com
2021-12-10 22:16:51 +08:00
&bench_strncmp_no_helper,
&bench_strncmp_helper,
selftest/bpf/benchs: Add bpf_map benchmark Add benchmark for hash_map to reproduce the worst case that non-stop update when map's free is zero. Just like this: ./run_bench_bpf_hashmap_full_update.sh Setting up benchmark 'bpf-hashmap-ful-update'... Benchmark 'bpf-hashmap-ful-update' started. 1:hash_map_full_perf 555830 events per sec ... Signed-off-by: Feng Zhou <zhoufeng.zf@bytedance.com> Link: https://lore.kernel.org/r/20220610023308.93798-3-zhoufeng.zf@bytedance.com Signed-off-by: Alexei Starovoitov <ast@kernel.org>
2022-06-10 10:33:08 +08:00
&bench_bpf_hashmap_full_update,
selftests/bpf: Add benchmark for local_storage get Add a benchmarks to demonstrate the performance cliff for local_storage get as the number of local_storage maps increases beyond current local_storage implementation's cache size. "sequential get" and "interleaved get" benchmarks are added, both of which do many bpf_task_storage_get calls on sets of task local_storage maps of various counts, while considering a single specific map to be 'important' and counting task_storage_gets to the important map separately in addition to normal 'hits' count of all gets. Goal here is to mimic scenario where a particular program using one map - the important one - is running on a system where many other local_storage maps exist and are accessed often. While "sequential get" benchmark does bpf_task_storage_get for map 0, 1, ..., {9, 99, 999} in order, "interleaved" benchmark interleaves 4 bpf_task_storage_gets for the important map for every 10 map gets. This is meant to highlight performance differences when important map is accessed far more frequently than non-important maps. A "hashmap control" benchmark is also included for easy comparison of standard bpf hashmap lookup vs local_storage get. The benchmark is similar to "sequential get", but creates and uses BPF_MAP_TYPE_HASH instead of local storage. Only one inner map is created - a hashmap meant to hold tid -> data mapping for all tasks. Size of the hashmap is hardcoded to my system's PID_MAX_LIMIT (4,194,304). The number of these keys which are actually fetched as part of the benchmark is configurable. Addition of this benchmark is inspired by conversation with Alexei in a previous patchset's thread [0], which highlighted the need for such a benchmark to motivate and validate improvements to local_storage implementation. My approach in that series focused on improving performance for explicitly-marked 'important' maps and was rejected with feedback to make more generally-applicable improvements while avoiding explicitly marking maps as important. Thus the benchmark reports both general and important-map-focused metrics, so effect of future work on both is clear. Regarding the benchmark results. On a powerful system (Skylake, 20 cores, 256gb ram): Hashmap Control =============== num keys: 10 hashmap (control) sequential get: hits throughput: 20.900 ± 0.334 M ops/s, hits latency: 47.847 ns/op, important_hits throughput: 20.900 ± 0.334 M ops/s num keys: 1000 hashmap (control) sequential get: hits throughput: 13.758 ± 0.219 M ops/s, hits latency: 72.683 ns/op, important_hits throughput: 13.758 ± 0.219 M ops/s num keys: 10000 hashmap (control) sequential get: hits throughput: 6.995 ± 0.034 M ops/s, hits latency: 142.959 ns/op, important_hits throughput: 6.995 ± 0.034 M ops/s num keys: 100000 hashmap (control) sequential get: hits throughput: 4.452 ± 0.371 M ops/s, hits latency: 224.635 ns/op, important_hits throughput: 4.452 ± 0.371 M ops/s num keys: 4194304 hashmap (control) sequential get: hits throughput: 3.043 ± 0.033 M ops/s, hits latency: 328.587 ns/op, important_hits throughput: 3.043 ± 0.033 M ops/s Local Storage ============= num_maps: 1 local_storage cache sequential get: hits throughput: 47.298 ± 0.180 M ops/s, hits latency: 21.142 ns/op, important_hits throughput: 47.298 ± 0.180 M ops/s local_storage cache interleaved get: hits throughput: 55.277 ± 0.888 M ops/s, hits latency: 18.091 ns/op, important_hits throughput: 55.277 ± 0.888 M ops/s num_maps: 10 local_storage cache sequential get: hits throughput: 40.240 ± 0.802 M ops/s, hits latency: 24.851 ns/op, important_hits throughput: 4.024 ± 0.080 M ops/s local_storage cache interleaved get: hits throughput: 48.701 ± 0.722 M ops/s, hits latency: 20.533 ns/op, important_hits throughput: 17.393 ± 0.258 M ops/s num_maps: 16 local_storage cache sequential get: hits throughput: 44.515 ± 0.708 M ops/s, hits latency: 22.464 ns/op, important_hits throughput: 2.782 ± 0.044 M ops/s local_storage cache interleaved get: hits throughput: 49.553 ± 2.260 M ops/s, hits latency: 20.181 ns/op, important_hits throughput: 15.767 ± 0.719 M ops/s num_maps: 17 local_storage cache sequential get: hits throughput: 38.778 ± 0.302 M ops/s, hits latency: 25.788 ns/op, important_hits throughput: 2.284 ± 0.018 M ops/s local_storage cache interleaved get: hits throughput: 43.848 ± 1.023 M ops/s, hits latency: 22.806 ns/op, important_hits throughput: 13.349 ± 0.311 M ops/s num_maps: 24 local_storage cache sequential get: hits throughput: 19.317 ± 0.568 M ops/s, hits latency: 51.769 ns/op, important_hits throughput: 0.806 ± 0.024 M ops/s local_storage cache interleaved get: hits throughput: 24.397 ± 0.272 M ops/s, hits latency: 40.989 ns/op, important_hits throughput: 6.863 ± 0.077 M ops/s num_maps: 32 local_storage cache sequential get: hits throughput: 13.333 ± 0.135 M ops/s, hits latency: 75.000 ns/op, important_hits throughput: 0.417 ± 0.004 M ops/s local_storage cache interleaved get: hits throughput: 16.898 ± 0.383 M ops/s, hits latency: 59.178 ns/op, important_hits throughput: 4.717 ± 0.107 M ops/s num_maps: 100 local_storage cache sequential get: hits throughput: 6.360 ± 0.107 M ops/s, hits latency: 157.233 ns/op, important_hits throughput: 0.064 ± 0.001 M ops/s local_storage cache interleaved get: hits throughput: 7.303 ± 0.362 M ops/s, hits latency: 136.930 ns/op, important_hits throughput: 1.907 ± 0.094 M ops/s num_maps: 1000 local_storage cache sequential get: hits throughput: 0.452 ± 0.010 M ops/s, hits latency: 2214.022 ns/op, important_hits throughput: 0.000 ± 0.000 M ops/s local_storage cache interleaved get: hits throughput: 0.542 ± 0.007 M ops/s, hits latency: 1843.341 ns/op, important_hits throughput: 0.136 ± 0.002 M ops/s Looking at the "sequential get" results, it's clear that as the number of task local_storage maps grows beyond the current cache size (16), there's a significant reduction in hits throughput. Note that current local_storage implementation assigns a cache_idx to maps as they are created. Since "sequential get" is creating maps 0..n in order and then doing bpf_task_storage_get calls in the same order, the benchmark is effectively ensuring that a map will not be in cache when the program tries to access it. For "interleaved get" results, important-map hits throughput is greatly increased as the important map is more likely to be in cache by virtue of being accessed far more frequently. Throughput still reduces as # maps increases, though. To get a sense of the overhead of the benchmark program, I commented out bpf_task_storage_get/bpf_map_lookup_elem in local_storage_bench.c and ran the benchmark on the same host as the 'real' run. Results: Hashmap Control =============== num keys: 10 hashmap (control) sequential get: hits throughput: 54.288 ± 0.655 M ops/s, hits latency: 18.420 ns/op, important_hits throughput: 54.288 ± 0.655 M ops/s num keys: 1000 hashmap (control) sequential get: hits throughput: 52.913 ± 0.519 M ops/s, hits latency: 18.899 ns/op, important_hits throughput: 52.913 ± 0.519 M ops/s num keys: 10000 hashmap (control) sequential get: hits throughput: 53.480 ± 1.235 M ops/s, hits latency: 18.699 ns/op, important_hits throughput: 53.480 ± 1.235 M ops/s num keys: 100000 hashmap (control) sequential get: hits throughput: 54.982 ± 1.902 M ops/s, hits latency: 18.188 ns/op, important_hits throughput: 54.982 ± 1.902 M ops/s num keys: 4194304 hashmap (control) sequential get: hits throughput: 50.858 ± 0.707 M ops/s, hits latency: 19.662 ns/op, important_hits throughput: 50.858 ± 0.707 M ops/s Local Storage ============= num_maps: 1 local_storage cache sequential get: hits throughput: 110.990 ± 4.828 M ops/s, hits latency: 9.010 ns/op, important_hits throughput: 110.990 ± 4.828 M ops/s local_storage cache interleaved get: hits throughput: 161.057 ± 4.090 M ops/s, hits latency: 6.209 ns/op, important_hits throughput: 161.057 ± 4.090 M ops/s num_maps: 10 local_storage cache sequential get: hits throughput: 112.930 ± 1.079 M ops/s, hits latency: 8.855 ns/op, important_hits throughput: 11.293 ± 0.108 M ops/s local_storage cache interleaved get: hits throughput: 115.841 ± 2.088 M ops/s, hits latency: 8.633 ns/op, important_hits throughput: 41.372 ± 0.746 M ops/s num_maps: 16 local_storage cache sequential get: hits throughput: 115.653 ± 0.416 M ops/s, hits latency: 8.647 ns/op, important_hits throughput: 7.228 ± 0.026 M ops/s local_storage cache interleaved get: hits throughput: 138.717 ± 1.649 M ops/s, hits latency: 7.209 ns/op, important_hits throughput: 44.137 ± 0.525 M ops/s num_maps: 17 local_storage cache sequential get: hits throughput: 112.020 ± 1.649 M ops/s, hits latency: 8.927 ns/op, important_hits throughput: 6.598 ± 0.097 M ops/s local_storage cache interleaved get: hits throughput: 128.089 ± 1.960 M ops/s, hits latency: 7.807 ns/op, important_hits throughput: 38.995 ± 0.597 M ops/s num_maps: 24 local_storage cache sequential get: hits throughput: 92.447 ± 5.170 M ops/s, hits latency: 10.817 ns/op, important_hits throughput: 3.855 ± 0.216 M ops/s local_storage cache interleaved get: hits throughput: 128.844 ± 2.808 M ops/s, hits latency: 7.761 ns/op, important_hits throughput: 36.245 ± 0.790 M ops/s num_maps: 32 local_storage cache sequential get: hits throughput: 102.042 ± 1.462 M ops/s, hits latency: 9.800 ns/op, important_hits throughput: 3.194 ± 0.046 M ops/s local_storage cache interleaved get: hits throughput: 126.577 ± 1.818 M ops/s, hits latency: 7.900 ns/op, important_hits throughput: 35.332 ± 0.507 M ops/s num_maps: 100 local_storage cache sequential get: hits throughput: 111.327 ± 1.401 M ops/s, hits latency: 8.983 ns/op, important_hits throughput: 1.113 ± 0.014 M ops/s local_storage cache interleaved get: hits throughput: 131.327 ± 1.339 M ops/s, hits latency: 7.615 ns/op, important_hits throughput: 34.302 ± 0.350 M ops/s num_maps: 1000 local_storage cache sequential get: hits throughput: 101.978 ± 0.563 M ops/s, hits latency: 9.806 ns/op, important_hits throughput: 0.102 ± 0.001 M ops/s local_storage cache interleaved get: hits throughput: 141.084 ± 1.098 M ops/s, hits latency: 7.088 ns/op, important_hits throughput: 35.430 ± 0.276 M ops/s Adjusting for overhead, latency numbers for "hashmap control" and "sequential get" are: hashmap_control_1k: ~53.8ns hashmap_control_10k: ~124.2ns hashmap_control_100k: ~206.5ns sequential_get_1: ~12.1ns sequential_get_10: ~16.0ns sequential_get_16: ~13.8ns sequential_get_17: ~16.8ns sequential_get_24: ~40.9ns sequential_get_32: ~65.2ns sequential_get_100: ~148.2ns sequential_get_1000: ~2204ns Clearly demonstrating a cliff. In the discussion for v1 of this patch, Alexei noted that local_storage was 2.5x faster than a large hashmap when initially implemented [1]. The benchmark results show that local_storage is 5-10x faster: a long-running BPF application putting some pid-specific info into a hashmap for each pid it sees will probably see on the order of 10-100k pids. Bench numbers for hashmaps of this size are ~10x slower than sequential_get_16, but as the number of local_storage maps grows far past local_storage cache size the performance advantage shrinks and eventually reverses. When running the benchmarks it may be necessary to bump 'open files' ulimit for a successful run. [0]: https://lore.kernel.org/all/20220420002143.1096548-1-davemarchevsky@fb.com [1]: https://lore.kernel.org/bpf/20220511173305.ftldpn23m4ski3d3@MBP-98dd607d3435.dhcp.thefacebook.com/ Signed-off-by: Dave Marchevsky <davemarchevsky@fb.com> Link: https://lore.kernel.org/r/20220620222554.270578-1-davemarchevsky@fb.com Signed-off-by: Alexei Starovoitov <ast@kernel.org>
2022-06-21 06:25:54 +08:00
&bench_local_storage_cache_seq_get,
&bench_local_storage_cache_interleaved_get,
&bench_local_storage_cache_hashmap_control,
selftests/bpf: Add benchmark for local_storage RCU Tasks Trace usage This benchmark measures grace period latency and kthread cpu usage of RCU Tasks Trace when many processes are creating/deleting BPF local_storage. Intent here is to quantify improvement on these metrics after Paul's recent RCU Tasks patches [0]. Specifically, fork 15k tasks which call a bpf prog that creates/destroys task local_storage and sleep in a loop, resulting in many call_rcu_tasks_trace calls. To determine grace period latency, trace time elapsed between rcu_tasks_trace_pregp_step and rcu_tasks_trace_postgp; for cpu usage look at rcu_task_trace_kthread's stime in /proc/PID/stat. On my virtualized test environment (Skylake, 8 cpus) benchmark results demonstrate significant improvement: BEFORE Paul's patches: SUMMARY tasks_trace grace period latency avg 22298.551 us stddev 1302.165 us SUMMARY ticks per tasks_trace grace period avg 2.291 stddev 0.324 AFTER Paul's patches: SUMMARY tasks_trace grace period latency avg 16969.197 us stddev 2525.053 us SUMMARY ticks per tasks_trace grace period avg 1.146 stddev 0.178 Note that since these patches are not in bpf-next benchmarking was done by cherry-picking this patch onto rcu tree. [0] https://lore.kernel.org/rcu/20220620225402.GA3842369@paulmck-ThinkPad-P17-Gen-1/ Signed-off-by: Dave Marchevsky <davemarchevsky@fb.com> Signed-off-by: Daniel Borkmann <daniel@iogearbox.net> Acked-by: Paul E. McKenney <paulmck@kernel.org> Acked-by: Martin KaFai Lau <kafai@fb.com> Link: https://lore.kernel.org/bpf/20220705190018.3239050-1-davemarchevsky@fb.com
2022-07-06 03:00:18 +08:00
&bench_local_storage_tasks_trace,
selftests/bpf: Add benchmark runner infrastructure While working on BPF ringbuf implementation, testing, and benchmarking, I've developed a pretty generic and modular benchmark runner, which seems to be generically useful, as I've already used it for one more purpose (testing fastest way to trigger BPF program, to minimize overhead of in-kernel code). This patch adds generic part of benchmark runner and sets up Makefile for extending it with more sets of benchmarks. Benchmarker itself operates by spinning up specified number of producer and consumer threads, setting up interval timer sending SIGALARM signal to application once a second. Every second, current snapshot with hits/drops counters are collected and stored in an array. Drops are useful for producer/consumer benchmarks in which producer might overwhelm consumers. Once test finishes after given amount of warm-up and testing seconds, mean and stddev are calculated (ignoring warm-up results) and is printed out to stdout. This setup seems to give consistent and accurate results. To validate behavior, I added two atomic counting tests: global and local. For global one, all the producer threads are atomically incrementing same counter as fast as possible. This, of course, leads to huge drop of performance once there is more than one producer thread due to CPUs fighting for the same memory location. Local counting, on the other hand, maintains one counter per each producer thread, incremented independently. Once per second, all counters are read and added together to form final "counting throughput" measurement. As expected, such setup demonstrates linear scalability with number of producers (as long as there are enough physical CPU cores, of course). See example output below. Also, this setup can nicely demonstrate disastrous effects of false sharing, if care is not taken to take those per-producer counters apart into independent cache lines. Demo output shows global counter first with 1 producer, then with 4. Both total and per-producer performance significantly drop. The last run is local counter with 4 producers, demonstrating near-perfect scalability. $ ./bench -a -w1 -d2 -p1 count-global Setting up benchmark 'count-global'... Benchmark 'count-global' started. Iter 0 ( 24.822us): hits 148.179M/s (148.179M/prod), drops 0.000M/s Iter 1 ( 37.939us): hits 149.308M/s (149.308M/prod), drops 0.000M/s Iter 2 (-10.774us): hits 150.717M/s (150.717M/prod), drops 0.000M/s Iter 3 ( 3.807us): hits 151.435M/s (151.435M/prod), drops 0.000M/s Summary: hits 150.488 ± 1.079M/s (150.488M/prod), drops 0.000 ± 0.000M/s $ ./bench -a -w1 -d2 -p4 count-global Setting up benchmark 'count-global'... Benchmark 'count-global' started. Iter 0 ( 60.659us): hits 53.910M/s ( 13.477M/prod), drops 0.000M/s Iter 1 (-17.658us): hits 53.722M/s ( 13.431M/prod), drops 0.000M/s Iter 2 ( 5.865us): hits 53.495M/s ( 13.374M/prod), drops 0.000M/s Iter 3 ( 0.104us): hits 53.606M/s ( 13.402M/prod), drops 0.000M/s Summary: hits 53.608 ± 0.113M/s ( 13.402M/prod), drops 0.000 ± 0.000M/s $ ./bench -a -w1 -d2 -p4 count-local Setting up benchmark 'count-local'... Benchmark 'count-local' started. Iter 0 ( 23.388us): hits 640.450M/s (160.113M/prod), drops 0.000M/s Iter 1 ( 2.291us): hits 605.661M/s (151.415M/prod), drops 0.000M/s Iter 2 ( -6.415us): hits 607.092M/s (151.773M/prod), drops 0.000M/s Iter 3 ( -1.361us): hits 601.796M/s (150.449M/prod), drops 0.000M/s Summary: hits 604.849 ± 2.739M/s (151.212M/prod), drops 0.000 ± 0.000M/s Benchmark runner supports setting thread affinity for producer and consumer threads. You can use -a flag for default CPU selection scheme, where first consumer gets CPU #0, next one gets CPU #1, and so on. Then producer threads pick up next CPU and increment one-by-one as well. But user can also specify a set of CPUs independently for producers and consumers with --prod-affinity 1,2-10,15 and --cons-affinity <set-of-cpus>. The latter allows to force producers and consumers to share same set of CPUs, if necessary. Signed-off-by: Andrii Nakryiko <andriin@fb.com> Signed-off-by: Alexei Starovoitov <ast@kernel.org> Acked-by: Yonghong Song <yhs@fb.com> Link: https://lore.kernel.org/bpf/20200512192445.2351848-3-andriin@fb.com
2020-05-13 03:24:43 +08:00
};
static void setup_benchmark()
{
int i, err;
if (!env.bench_name) {
fprintf(stderr, "benchmark name is not specified\n");
exit(1);
}
for (i = 0; i < ARRAY_SIZE(benchs); i++) {
if (strcmp(benchs[i]->name, env.bench_name) == 0) {
bench = benchs[i];
break;
}
}
if (!bench) {
fprintf(stderr, "benchmark '%s' not found\n", env.bench_name);
exit(1);
}
printf("Setting up benchmark '%s'...\n", bench->name);
state.producers = calloc(env.producer_cnt, sizeof(*state.producers));
state.consumers = calloc(env.consumer_cnt, sizeof(*state.consumers));
state.results = calloc(env.duration_sec + env.warmup_sec + 2,
sizeof(*state.results));
if (!state.producers || !state.consumers || !state.results)
exit(1);
if (bench->validate)
bench->validate();
if (bench->setup)
bench->setup();
for (i = 0; i < env.consumer_cnt; i++) {
err = pthread_create(&state.consumers[i], NULL,
bench->consumer_thread, (void *)(long)i);
if (err) {
fprintf(stderr, "failed to create consumer thread #%d: %d\n",
i, -errno);
exit(1);
}
if (env.affinity)
set_thread_affinity(state.consumers[i],
next_cpu(&env.cons_cpus));
}
/* unless explicit producer CPU list is specified, continue after
* last consumer CPU
*/
if (!env.prod_cpus.cpus)
env.prod_cpus.next_cpu = env.cons_cpus.next_cpu;
for (i = 0; i < env.producer_cnt; i++) {
err = pthread_create(&state.producers[i], NULL,
bench->producer_thread, (void *)(long)i);
if (err) {
fprintf(stderr, "failed to create producer thread #%d: %d\n",
i, -errno);
exit(1);
}
if (env.affinity)
set_thread_affinity(state.producers[i],
next_cpu(&env.prod_cpus));
}
printf("Benchmark '%s' started.\n", bench->name);
}
static pthread_mutex_t bench_done_mtx = PTHREAD_MUTEX_INITIALIZER;
static pthread_cond_t bench_done = PTHREAD_COND_INITIALIZER;
static void collect_measurements(long delta_ns) {
int iter = state.res_cnt++;
struct bench_res *res = &state.results[iter];
bench->measure(res);
if (bench->report_progress)
bench->report_progress(iter, res, delta_ns);
if (iter == env.duration_sec + env.warmup_sec) {
pthread_mutex_lock(&bench_done_mtx);
pthread_cond_signal(&bench_done);
pthread_mutex_unlock(&bench_done_mtx);
}
}
int main(int argc, char **argv)
{
parse_cmdline_args(argc, argv);
if (env.list) {
int i;
printf("Available benchmarks:\n");
for (i = 0; i < ARRAY_SIZE(benchs); i++) {
printf("- %s\n", benchs[i]->name);
}
return 0;
}
setup_benchmark();
setup_timer();
pthread_mutex_lock(&bench_done_mtx);
pthread_cond_wait(&bench_done, &bench_done_mtx);
pthread_mutex_unlock(&bench_done_mtx);
if (bench->report_final)
/* skip first sample */
bench->report_final(state.results + env.warmup_sec,
state.res_cnt - env.warmup_sec);
return 0;
}
Reference in New Issue Copy Permalink