From: Andrii Nakryiko <andrii@kernel.org>
To: <bpf@vger.kernel.org>, <ast@kernel.org>, <daniel@iogearbox.net>,
<martin.lau@kernel.org>
Cc: <andrii@kernel.org>, <kernel-team@meta.com>
Subject: [PATCH bpf-next 0/6] BPF verifier rotating log
Date: Fri, 17 Mar 2023 15:03:45 -0700 [thread overview]
Message-ID: <20230317220351.2970665-1-andrii@kernel.org> (raw)
This patch set changes BPF verifier log behavior to behave as a rotating log,
by default. If user-supplied log buffer is big enough to contain entire
verifier log output, there is no effective difference. But where previously
user supplied too small log buffer and would get -ENOSPC error result and the
beginning part of the verifier log, now there will be no error and user will
get ending part of verifier log filling up user-supplied log buffer.
Which is, in absolute majority of cases, is exactly what's useful, relevant,
and what users want and need, as the ending of the verifier log is containing
details of verifier failure and relevant state that got us to that failure. So
this rotating mode is made a default, but for some niche advanced debugging
scenarios it's possible to request old behavior by specifying additional
BPF_LOG_FIXED (8) flag.
This patch set adjusts libbpf to allow specifying flags beyond 1 | 2 | 4. We
also add --log-size and --log-fixed options to veristat to be able to both
test this functionality manually, but also to be used in various debugging
scenarios. We also add selftests that tries many variants of log buffer size
to stress-test correctness of internal verifier log bookkeeping code.
v1->v2:
- fixed using logical end_pos, which could get out of buffer bounds, in
bpf_vlog_reset() (Alexei);
- added more strict testing of log buffer beyond allowed size, tested that
this caugh the above bug (before fixing it in kernel); also added
log_level 1 testing for program that fails to load (it stress tests
bpf_vlog_reset() some more);
- switched to div_u64_rem() for u64 modulo, reported by kernel test robot;
considered switching to u32 and handle overflows/underflows, but decided
against that because log->end_pos is used in some additional logic around
formatting verifier state next to each instruction and it was getting too
tricky to cleanly reason about underflows/overflows. Keeping it simple
with u64 (which are now always incresing and non-overflowing) and
div_u64_rem().
Andrii Nakryiko (6):
bpf: split off basic BPF verifier log into separate file
bpf: remove minimum size restrictions on verifier log buffer
bpf: switch BPF verifier log to be a rotating log by default
libbpf: don't enfore verifier log levels on libbpf side
selftests/bpf: add more veristat control over verifier log options
selftests/bpf: add fixed vs rotating verifier log tests
include/linux/bpf_verifier.h | 47 ++--
kernel/bpf/Makefile | 3 +-
kernel/bpf/log.c | 254 ++++++++++++++++++
kernel/bpf/verifier.c | 86 +-----
tools/lib/bpf/bpf.c | 2 -
.../selftests/bpf/prog_tests/log_fixup.c | 1 +
.../selftests/bpf/prog_tests/verifier_log.c | 156 +++++++++++
tools/testing/selftests/bpf/veristat.c | 42 ++-
8 files changed, 486 insertions(+), 105 deletions(-)
create mode 100644 kernel/bpf/log.c
create mode 100644 tools/testing/selftests/bpf/prog_tests/verifier_log.c
--
2.34.1
next reply other threads:[~2023-03-17 22:04 UTC|newest]
Thread overview: 15+ messages / expand[flat|nested] mbox.gz Atom feed top
2023-03-17 22:03 Andrii Nakryiko [this message]
2023-03-17 22:03 ` [PATCH bpf-next 1/6] bpf: split off basic BPF verifier log into separate file Andrii Nakryiko
2023-03-17 22:03 ` [PATCH bpf-next 2/6] bpf: remove minimum size restrictions on verifier log buffer Andrii Nakryiko
2023-03-17 22:03 ` [PATCH bpf-next 3/6] bpf: switch BPF verifier log to be a rotating log by default Andrii Nakryiko
2023-03-17 22:05 ` Andrii Nakryiko
2023-03-17 23:02 ` Daniel Borkmann
2023-03-17 23:12 ` Andrii Nakryiko
2023-03-20 16:10 ` Lorenz Bauer
2023-03-20 18:55 ` Andrii Nakryiko
2023-03-21 16:18 ` Lorenz Bauer
2023-03-22 21:27 ` Andrii Nakryiko
2023-03-17 22:03 ` [PATCH bpf-next 4/6] libbpf: don't enfore verifier log levels on libbpf side Andrii Nakryiko
2023-03-17 22:03 ` [PATCH bpf-next 5/6] selftests/bpf: add more veristat control over verifier log options Andrii Nakryiko
2023-03-17 22:03 ` [PATCH bpf-next 6/6] selftests/bpf: add fixed vs rotating verifier log tests Andrii Nakryiko
-- strict thread matches above, loose matches on Subject: below --
2023-03-16 18:30 [PATCH bpf-next 0/6] BPF verifier rotating log Andrii Nakryiko
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=20230317220351.2970665-1-andrii@kernel.org \
--to=andrii@kernel.org \
--cc=ast@kernel.org \
--cc=bpf@vger.kernel.org \
--cc=daniel@iogearbox.net \
--cc=kernel-team@meta.com \
--cc=martin.lau@kernel.org \
/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