BPF List
 help / color / mirror / Atom feed
From: Eduard Zingerman <eddyz87@gmail.com>
To: bpf@vger.kernel.org, ast@kernel.org
Cc: andrii@kernel.org, daniel@iogearbox.net, martin.lau@linux.dev,
	kernel-team@fb.com, yonghong.song@linux.dev, memxor@gmail.com,
	awerner32@gmail.com, Eduard Zingerman <eddyz87@gmail.com>
Subject: [PATCH bpf 03/12] selftests/bpf: fix bpf_loop_bench for new callback verification scheme
Date: Thu, 16 Nov 2023 04:17:54 +0200	[thread overview]
Message-ID: <20231116021803.9982-4-eddyz87@gmail.com> (raw)
In-Reply-To: <20231116021803.9982-1-eddyz87@gmail.com>

The patch a few patches from this one changes logic for callbacks
handling. While previously callbacks were verified as a single
function call, new scheme takes into account that callbacks could be
executed unknown number of times.

This has dire implications for bpf_loop_bench:

    SEC("fentry/" SYS_PREFIX "sys_getpgid")
    int benchmark(void *ctx)
    {
            for (int i = 0; i < 1000; i++) {
                    bpf_loop(nr_loops, empty_callback, NULL, 0);
                    __sync_add_and_fetch(&hits, nr_loops);
            }
            return 0;
    }

W/o callbacks change for verifier it merely represents 1000 calls to
empty_callback(). However, with callbacks change things become
exponential:
- i=0: state exploring empty_callback is scheduled with i=0 (a);
- i=1: state exploring empty_callback is scheduled with i=1;
  ...
- i=999: state exploring empty_callback is scheduled with i=999;
- state (a) is popped from stack;
- i=1: state exploring empty_callback is scheduled with i=1;
  ...

Avoid this issue by rewriting outer loop as bpf_loop().
Unfortunately, this adds a function call to a loop at runtime, which
negatively affects performance:

            throughput               latency
   before:  149.919 ± 0.168 M ops/s, 6.670 ns/op
   after :  137.040 ± 0.187 M ops/s, 7.297 ns/op

Signed-off-by: Eduard Zingerman <eddyz87@gmail.com>
---
 tools/testing/selftests/bpf/progs/bpf_loop_bench.c | 13 ++++++++-----
 1 file changed, 8 insertions(+), 5 deletions(-)

diff --git a/tools/testing/selftests/bpf/progs/bpf_loop_bench.c b/tools/testing/selftests/bpf/progs/bpf_loop_bench.c
index 4ce76eb064c4..d461746fd3c1 100644
--- a/tools/testing/selftests/bpf/progs/bpf_loop_bench.c
+++ b/tools/testing/selftests/bpf/progs/bpf_loop_bench.c
@@ -15,13 +15,16 @@ static int empty_callback(__u32 index, void *data)
 	return 0;
 }
 
+static int outer_loop(__u32 index, void *data)
+{
+	bpf_loop(nr_loops, empty_callback, NULL, 0);
+	__sync_add_and_fetch(&hits, nr_loops);
+	return 0;
+}
+
 SEC("fentry/" SYS_PREFIX "sys_getpgid")
 int benchmark(void *ctx)
 {
-	for (int i = 0; i < 1000; i++) {
-		bpf_loop(nr_loops, empty_callback, NULL, 0);
-
-		__sync_add_and_fetch(&hits, nr_loops);
-	}
+	bpf_loop(1000, outer_loop, NULL, 0);
 	return 0;
 }
-- 
2.42.0


  parent reply	other threads:[~2023-11-16  2:18 UTC|newest]

Thread overview: 49+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2023-11-16  2:17 [PATCH bpf 00/12] verify callbacks as if they are called unknown number of times Eduard Zingerman
2023-11-16  2:17 ` [PATCH bpf 01/12] selftests/bpf: track tcp payload offset as scalar in xdp_synproxy Eduard Zingerman
2023-11-17 16:46   ` Andrii Nakryiko
2023-11-16  2:17 ` [PATCH bpf 02/12] selftests/bpf: track string payload offset as scalar in strobemeta Eduard Zingerman
2023-11-17 16:46   ` Andrii Nakryiko
2023-11-17 18:52     ` Eduard Zingerman
2023-11-16  2:17 ` Eduard Zingerman [this message]
2023-11-17 16:46   ` [PATCH bpf 03/12] selftests/bpf: fix bpf_loop_bench for new callback verification scheme Andrii Nakryiko
2023-11-17 18:52     ` Eduard Zingerman
2023-11-17 21:38   ` Alexei Starovoitov
2023-11-17 21:43     ` Eduard Zingerman
2023-11-17 21:47       ` Alexei Starovoitov
2023-11-17 21:50         ` Eduard Zingerman
2023-11-17 21:55           ` Alexei Starovoitov
2023-11-16  2:17 ` [PATCH bpf 04/12] bpf: extract __check_reg_arg() utility function Eduard Zingerman
2023-11-17 16:46   ` Andrii Nakryiko
2023-11-16  2:17 ` [PATCH bpf 05/12] bpf: extract setup_func_entry() " Eduard Zingerman
2023-11-17 16:46   ` Andrii Nakryiko
2023-11-17 18:52     ` Eduard Zingerman
2023-11-16  2:17 ` [PATCH bpf 06/12] bpf: verify callbacks as if they are called unknown number of times Eduard Zingerman
2023-11-17 16:46   ` Andrii Nakryiko
2023-11-17 18:52     ` Eduard Zingerman
2023-11-17 20:27       ` Andrii Nakryiko
2023-11-17 21:03         ` Eduard Zingerman
2023-11-16  2:17 ` [PATCH bpf 07/12] selftests/bpf: tests for iterating callbacks Eduard Zingerman
2023-11-17 16:46   ` Andrii Nakryiko
2023-11-16  2:17 ` [PATCH bpf 08/12] bpf: widening for callback iterators Eduard Zingerman
2023-11-17 16:46   ` Andrii Nakryiko
2023-11-16  2:18 ` [PATCH bpf 09/12] selftests/bpf: test widening for iterating callbacks Eduard Zingerman
2023-11-17 16:47   ` Andrii Nakryiko
2023-11-17 18:53     ` Eduard Zingerman
2023-11-17 20:28       ` Andrii Nakryiko
2023-11-16  2:18 ` [PATCH bpf 10/12] bpf: keep track of max number of bpf_loop callback iterations Eduard Zingerman
2023-11-16 14:08   ` Andrii Nakryiko
2023-11-16 14:13     ` Eduard Zingerman
2023-11-17 16:47   ` Andrii Nakryiko
2023-11-17 18:53     ` Eduard Zingerman
2023-11-17 20:30       ` Andrii Nakryiko
2023-11-16  2:18 ` [PATCH bpf 11/12] selftests/bpf: add __not_msg annotation for test_loader based tests Eduard Zingerman
2023-11-17 16:45   ` Andrii Nakryiko
2023-11-17 18:53     ` Eduard Zingerman
2023-11-17 20:31       ` Andrii Nakryiko
2023-11-17 21:10         ` Eduard Zingerman
2023-11-17 21:33           ` Alexei Starovoitov
2023-11-16  2:18 ` [PATCH bpf 12/12] selftests/bpf: check if max number of bpf_loop iterations is tracked Eduard Zingerman
2023-11-17 16:47   ` Andrii Nakryiko
2023-11-17 18:53     ` Eduard Zingerman
2023-11-17 20:32       ` Andrii Nakryiko
2023-11-17 21:18         ` Eduard Zingerman

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=20231116021803.9982-4-eddyz87@gmail.com \
    --to=eddyz87@gmail.com \
    --cc=andrii@kernel.org \
    --cc=ast@kernel.org \
    --cc=awerner32@gmail.com \
    --cc=bpf@vger.kernel.org \
    --cc=daniel@iogearbox.net \
    --cc=kernel-team@fb.com \
    --cc=martin.lau@linux.dev \
    --cc=memxor@gmail.com \
    --cc=yonghong.song@linux.dev \
    /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