Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:18 as permitted sender) client-ip=2620:137:e000::1:18;
MIME-Version: 1.0
References: <20220217131916.50615-1-jolsa@kernel.org>
In-Reply-To: <20220217131916.50615-1-jolsa@kernel.org>
From:   Andrii Nakryiko <andrii.nakryiko@gmail.com>
Date:   Thu, 17 Feb 2022 13:55:13 -0800
Message-ID: <CAEf4BzY+_3vjtN3dJjU4deVR131=Dz-9adYQ+mntVqgAOfh4RA@mail.gmail.com>
Subject: Re: [PATCHv2 0/3] perf/bpf: Replace deprecated code
To:     Jiri Olsa <jolsa@kernel.org>
Cc:     Arnaldo Carvalho de Melo <acme@kernel.org>,
        lkml <linux-kernel@vger.kernel.org>,
        Peter Zijlstra <a.p.zijlstra@chello.nl>,
        Ingo Molnar <mingo@kernel.org>,
        Mark Rutland <mark.rutland@arm.com>,
        Namhyung Kim <namhyung@kernel.org>,
        Alexander Shishkin <alexander.shishkin@linux.intel.com>,
        Ian Rogers <irogers@google.com>,
        "linux-perf-use." <linux-perf-users@vger.kernel.org>,
        bpf <bpf@vger.kernel.org>, Andrii Nakryiko <andrii@kernel.org>
Content-Type: text/plain; charset="UTF-8"
Precedence: bulk

On Thu, Feb 17, 2022 at 5:19 AM Jiri Olsa <jolsa@kernel.org> wrote:
>
> hi,
> the original patchset [1] removed the whole perf functionality
> with the hope nobody's using that. But it turned out there's
> actually bpf script using prologue functionality, so there
> might be users of this.
>
> This patchset gets rid of and adds workaround (and keeps the
> current functionality) for following deprecated libbpf
> functions/struct:
>
>   bpf_program__set_priv
>   bpf_program__priv
>   bpf_map__set_priv
>   bpf_map__priv
>   bpf_program__set_prep
>   bpf_program__nth_fd
>   struct bpf_prog_prep_result
>
> Basically it implements workarounds suggested by Andrii in [2].
>
> I tested with script from examples/bpf that are working for me:
>
>   examples/bpf/hello.c
>   examples/bpf/5sec.c
>
> The rest seem to fail for various reasons even without this
> change..  they seem unmaintained for some time now, but I might
> have wrong setup.
>
> Also available in here:
>   git://git.kernel.org/pub/scm/linux/kernel/git/jolsa/perf.git
>   perf/depre
>
> thanks,
> jirka
>
>
> [1] https://lore.kernel.org/linux-perf-users/YgoPxhE3OEEmZqla@krava/T/#t
> [2] https://lore.kernel.org/linux-perf-users/YgoPxhE3OEEmZqla@krava/T/#md3ccab9fe70a4583e94603b1a562e369bd67b17d
> ---
> Jiri Olsa (3):
>       perf tools: Remove bpf_program__set_priv/bpf_program__priv usage
>       perf tools: Remove bpf_map__set_priv/bpf_map__priv usage
>       perf tools: Rework prologue generation code
>

It's great that you are deprecating these, thanks a lot for that! I
suggest to also doing libbpf_set_strict_mode(LIBBPF_STRICT_ALL) to
check that libbpf 1.0 won't break anything. For example, you'll need
to use a custom SEC() handler to handle those quirky sections that
perf allows. This patch set has landed in bpf-next, so you should be
good to go.


>  tools/perf/util/bpf-loader.c | 267 +++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++-----------------
>  1 file changed, 230 insertions(+), 37 deletions(-)