From: Brian Vazquez <brianvv@google.com>
To: Brian Vazquez <brianvv.kernel@gmail.com>,
Alexei Starovoitov <ast@kernel.org>,
Daniel Borkmann <daniel@iogearbox.net>,
"David S . Miller" <davem@davemloft.net>
Cc: Stanislav Fomichev <sdf@google.com>,
Willem de Bruijn <willemb@google.com>,
Petar Penkov <ppenkov@google.com>,
linux-kernel@vger.kernel.org, netdev@vger.kernel.org,
bpf@vger.kernel.org, Brian Vazquez <brianvv@google.com>
Subject: [PATCH bpf-next 6/6] selftests/bpf: add test to measure performance of BPF_MAP_DUMP
Date: Wed, 24 Jul 2019 09:58:03 -0700 [thread overview]
Message-ID: <20190724165803.87470-7-brianvv@google.com> (raw)
In-Reply-To: <20190724165803.87470-1-brianvv@google.com>
This tests compares the amount of time that takes to read an entire
table of 100K elements on a bpf hashmap using both BPF_MAP_DUMP and
BPF_MAP_GET_NEXT_KEY + BPF_MAP_LOOKUP_ELEM.
Signed-off-by: Brian Vazquez <brianvv@google.com>
---
tools/testing/selftests/bpf/test_maps.c | 65 +++++++++++++++++++++++++
1 file changed, 65 insertions(+)
diff --git a/tools/testing/selftests/bpf/test_maps.c b/tools/testing/selftests/bpf/test_maps.c
index f7ab401399d40..c4593a8904ca6 100644
--- a/tools/testing/selftests/bpf/test_maps.c
+++ b/tools/testing/selftests/bpf/test_maps.c
@@ -18,6 +18,7 @@
#include <sys/socket.h>
#include <netinet/in.h>
#include <linux/bpf.h>
+#include <linux/time64.h>
#include <bpf/bpf.h>
#include <bpf/libbpf.h>
@@ -389,6 +390,69 @@ static void test_hashmap_dump(void)
close(fd);
}
+static void test_hashmap_dump_perf(void)
+{
+ int fd, i, max_entries = 100000;
+ uint64_t key, value, next_key;
+ bool next_key_valid = true;
+ void *buf;
+ u32 buf_len, entries;
+ int j = 0;
+ int clk_id = CLOCK_MONOTONIC;
+ struct timespec begin, end;
+ long long time_spent, dump_time_spent;
+ double res;
+ int tests[] = {1, 2, 230, 5000, 73000, 100000, 234567};
+ int test_len = ARRAY_SIZE(tests);
+ const int elem_size = sizeof(key) + sizeof(value);
+
+ fd = helper_fill_hashmap(max_entries);
+ // Alloc memory considering the largest buffer
+ buf = malloc(elem_size * tests[test_len-1]);
+ assert(buf != NULL);
+
+test:
+ entries = tests[j];
+ buf_len = elem_size*tests[j];
+ j++;
+ clock_gettime(clk_id, &begin);
+ errno = 0;
+ i = 0;
+ while (errno == 0) {
+ bpf_map_dump(fd, !i ? NULL : &key,
+ buf, &buf_len);
+ if (errno)
+ break;
+ if (!i)
+ key = *((uint64_t *)(buf + buf_len - elem_size));
+ i += buf_len / elem_size;
+ }
+ clock_gettime(clk_id, &end);
+ assert(i == max_entries);
+ dump_time_spent = NSEC_PER_SEC * (end.tv_sec - begin.tv_sec) +
+ end.tv_nsec - begin.tv_nsec;
+ next_key_valid = true;
+ clock_gettime(clk_id, &begin);
+ assert(bpf_map_get_next_key(fd, NULL, &key) == 0);
+ for (i = 0; next_key_valid; i++) {
+ next_key_valid = bpf_map_get_next_key(fd, &key, &next_key) == 0;
+ assert(bpf_map_lookup_elem(fd, &key, &value) == 0);
+ key = next_key;
+ }
+ clock_gettime(clk_id, &end);
+ time_spent = NSEC_PER_SEC * (end.tv_sec - begin.tv_sec) +
+ end.tv_nsec - begin.tv_nsec;
+ res = (1-((double)dump_time_spent/time_spent))*100;
+ printf("buf_len_%u:\t %llu entry-by-entry: %llu improvement %lf\n",
+ entries, dump_time_spent, time_spent, res);
+ assert(i == max_entries);
+
+ if (j < test_len)
+ goto test;
+ free(buf);
+ close(fd);
+}
+
static void test_hashmap_zero_seed(void)
{
int i, first, second, old_flags;
@@ -1758,6 +1822,7 @@ static void run_all_tests(void)
test_hashmap_walk(0, NULL);
test_hashmap_zero_seed();
test_hashmap_dump();
+ test_hashmap_dump_perf();
test_arraymap(0, NULL);
test_arraymap_percpu(0, NULL);
--
2.22.0.657.g960e92d24f-goog
next prev parent reply other threads:[~2019-07-24 16:58 UTC|newest]
Thread overview: 31+ messages / expand[flat|nested] mbox.gz Atom feed top
2019-07-24 16:57 [PATCH bpf-next 0/6] bpf: add BPF_MAP_DUMP command to dump more than one entry per call Brian Vazquez
2019-07-24 16:57 ` [PATCH bpf-next 1/6] bpf: add bpf_map_value_size and bp_map_copy_value helper functions Brian Vazquez
2019-07-24 20:53 ` Song Liu
2019-07-24 16:57 ` [PATCH bpf-next 2/6] bpf: add BPF_MAP_DUMP command to dump more than one entry per call Brian Vazquez
2019-07-24 19:54 ` Willem de Bruijn
2019-07-24 22:26 ` Brian Vazquez
2019-07-24 22:33 ` Willem de Bruijn
2019-07-24 21:40 ` Song Liu
2019-07-24 22:44 ` Brian Vazquez
2019-07-24 23:04 ` Song Liu
2019-07-25 23:25 ` Brian Vazquez
2019-07-25 23:54 ` Alexei Starovoitov
2019-07-26 1:02 ` Willem de Bruijn
2019-07-26 1:24 ` Brian Vazquez
2019-07-26 1:47 ` Alexei Starovoitov
2019-07-26 6:10 ` Yonghong Song
2019-07-26 23:36 ` Brian Vazquez
2019-07-27 0:02 ` Jakub Kicinski
2019-07-27 17:54 ` Yonghong Song
2019-07-24 16:58 ` [PATCH bpf-next 3/6] bpf: keep bpf.h in sync with tools/ Brian Vazquez
2019-07-24 21:41 ` Song Liu
2019-07-24 23:10 ` Andrii Nakryiko
2019-07-25 23:27 ` Brian Vazquez
2019-07-24 16:58 ` [PATCH bpf-next 4/6] libbpf: support BPF_MAP_DUMP command Brian Vazquez
2019-07-24 19:51 ` Willem de Bruijn
2019-07-24 16:58 ` [PATCH bpf-next 5/6] selftests/bpf: test BPF_MAP_DUMP command on a bpf hashmap Brian Vazquez
2019-07-24 21:58 ` Song Liu
2019-07-24 16:58 ` Brian Vazquez [this message]
2019-07-24 22:01 ` [PATCH bpf-next 6/6] selftests/bpf: add test to measure performance of BPF_MAP_DUMP Song Liu
2019-07-24 19:20 ` [PATCH bpf-next 0/6] bpf: add BPF_MAP_DUMP command to dump more than one entry per call Song Liu
2019-07-24 22:15 ` Brian Vazquez
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20190724165803.87470-7-brianvv@google.com \
--to=brianvv@google.com \
--cc=ast@kernel.org \
--cc=bpf@vger.kernel.org \
--cc=brianvv.kernel@gmail.com \
--cc=daniel@iogearbox.net \
--cc=davem@davemloft.net \
--cc=linux-kernel@vger.kernel.org \
--cc=netdev@vger.kernel.org \
--cc=ppenkov@google.com \
--cc=sdf@google.com \
--cc=willemb@google.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).