bpf.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Andrii Nakryiko <andrii.nakryiko@gmail.com>
To: "Toke Høiland-Jørgensen" <toke@redhat.com>
Cc: Daniel Borkmann <daniel@iogearbox.net>,
	Alexei Starovoitov <ast@kernel.org>,
	Martin KaFai Lau <kafai@fb.com>, Song Liu <songliubraving@fb.com>,
	Yonghong Song <yhs@fb.com>,
	Jesper Dangaard Brouer <brouer@redhat.com>,
	David Miller <davem@davemloft.net>,
	Networking <netdev@vger.kernel.org>, bpf <bpf@vger.kernel.org>
Subject: Re: [PATCH RFC bpf-next 1/3] libbpf: Add new bpf_object__load2() using new-style opts
Date: Thu, 19 Dec 2019 15:50:57 -0800	[thread overview]
Message-ID: <CAEf4BzYKpstQk8JO_iOws93VpHEEs+J+z+ZO7cKRiKRNvN1zMg@mail.gmail.com> (raw)
In-Reply-To: <157676577159.957277.7471130922810004500.stgit@toke.dk>

On Thu, Dec 19, 2019 at 6:29 AM Toke Høiland-Jørgensen <toke@redhat.com> wrote:
>
> From: Toke Høiland-Jørgensen <toke@redhat.com>
>
> Since we introduced DECLARE_LIBBPF_OPTS and related macros for declaring
> function options, that is now the preferred way to extend APIs. Introduce a
> variant of the bpf_object__load() function that uses this function, and
> deprecate the _xattr variant. Since all the good function names were taken,
> the new function is unimaginatively called bpf_object__load2().
>
> Signed-off-by: Toke Høiland-Jørgensen <toke@redhat.com>
> ---

I've been thinking about options for load quite a bit lately, and I'm
leaning towards an opinion, that bpf_object__load() shouldn't take any
options, and all the various per-bpf_object options have to be
specified in bpf_object_open_opts and stored, if necessary for
load/attach phase. So I'd rather move target_btf_path and log_level to
open_opts instead.

For cases where we need to tune load behavior/options per individual
BPF program, bpf_object_load_opts don't work either way, and we'll
have to add some getters/setters for bpf_program objects, anyways. So
I think it's overall cleaner to specify everything per-bpf_object at
"creation time" (which is open), and keep load()/attach() option-less.

>  tools/lib/bpf/libbpf.c   |   31 ++++++++++++++++++-------------
>  tools/lib/bpf/libbpf.h   |   13 +++++++++++++
>  tools/lib/bpf/libbpf.map |    1 +
>  3 files changed, 32 insertions(+), 13 deletions(-)
>
> diff --git a/tools/lib/bpf/libbpf.c b/tools/lib/bpf/libbpf.c
> index febbaac3daf4..266b725e444b 100644
> --- a/tools/lib/bpf/libbpf.c
> +++ b/tools/lib/bpf/libbpf.c
> @@ -4844,15 +4844,12 @@ static int bpf_object__resolve_externs(struct bpf_object *obj,
>         return 0;
>  }
>
> -int bpf_object__load_xattr(struct bpf_object_load_attr *attr)
> +int bpf_object__load2(struct bpf_object *obj,
> +                     const struct bpf_object_load_opts *opts)
>  {
> -       struct bpf_object *obj;
>         int err, i;
>
> -       if (!attr)
> -               return -EINVAL;
> -       obj = attr->obj;
> -       if (!obj)
> +       if (!obj || !OPTS_VALID(opts, bpf_object_load_opts))
>                 return -EINVAL;
>
>         if (obj->loaded) {
> @@ -4867,8 +4864,10 @@ int bpf_object__load_xattr(struct bpf_object_load_attr *attr)
>         err = err ? : bpf_object__sanitize_and_load_btf(obj);
>         err = err ? : bpf_object__sanitize_maps(obj);
>         err = err ? : bpf_object__create_maps(obj);
> -       err = err ? : bpf_object__relocate(obj, attr->target_btf_path);
> -       err = err ? : bpf_object__load_progs(obj, attr->log_level);
> +       err = err ? : bpf_object__relocate(obj,
> +                                          OPTS_GET(opts, target_btf_path, NULL));
> +       err = err ? : bpf_object__load_progs(obj,
> +                                            OPTS_GET(opts, log_level, 0));
>         if (err)
>                 goto out;
>
> @@ -4884,13 +4883,19 @@ int bpf_object__load_xattr(struct bpf_object_load_attr *attr)
>         return err;
>  }
>
> -int bpf_object__load(struct bpf_object *obj)
> +int bpf_object__load_xattr(struct bpf_object_load_attr *attr)
>  {
> -       struct bpf_object_load_attr attr = {
> -               .obj = obj,
> -       };
> +       DECLARE_LIBBPF_OPTS(bpf_object_load_opts, opts,
> +           .log_level = attr->log_level,
> +           .target_btf_path = attr->target_btf_path,
> +       );
> +
> +       return bpf_object__load2(attr->obj, &opts);
> +}
>
> -       return bpf_object__load_xattr(&attr);
> +int bpf_object__load(struct bpf_object *obj)
> +{
> +       return bpf_object__load2(obj, NULL);
>  }
>
>  static int make_parent_dir(const char *path)
> diff --git a/tools/lib/bpf/libbpf.h b/tools/lib/bpf/libbpf.h
> index fe592ef48f1b..ce86277d7445 100644
> --- a/tools/lib/bpf/libbpf.h
> +++ b/tools/lib/bpf/libbpf.h
> @@ -132,8 +132,21 @@ struct bpf_object_load_attr {
>         const char *target_btf_path;
>  };
>
> +struct bpf_object_load_opts {
> +       /* size of this struct, for forward/backward compatiblity */
> +       size_t sz;
> +       /* log level on load */
> +       int log_level;
> +       /* BTF path (for CO-RE relocations) */
> +       const char *target_btf_path;
> +};
> +#define bpf_object_load_opts__last_field target_btf_path
> +
>  /* Load/unload object into/from kernel */
>  LIBBPF_API int bpf_object__load(struct bpf_object *obj);
> +LIBBPF_API int bpf_object__load2(struct bpf_object *obj,
> +                                const struct bpf_object_load_opts *opts);
> +/* deprecated, use bpf_object__load2() instead */
>  LIBBPF_API int bpf_object__load_xattr(struct bpf_object_load_attr *attr);
>  LIBBPF_API int bpf_object__unload(struct bpf_object *obj);
>
> diff --git a/tools/lib/bpf/libbpf.map b/tools/lib/bpf/libbpf.map
> index e3a471f38a71..d6cb860763d1 100644
> --- a/tools/lib/bpf/libbpf.map
> +++ b/tools/lib/bpf/libbpf.map
> @@ -217,6 +217,7 @@ LIBBPF_0.0.7 {
>                 bpf_object__attach_skeleton;
>                 bpf_object__destroy_skeleton;
>                 bpf_object__detach_skeleton;
> +               bpf_object__load2;
>                 bpf_object__load_skeleton;
>                 bpf_object__open_skeleton;
>                 bpf_program__attach;
>

  reply	other threads:[~2019-12-19 23:51 UTC|newest]

Thread overview: 13+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2019-12-19 14:29 [PATCH RFC bpf-next 0/3] libbpf: Add support for extern function calls Toke Høiland-Jørgensen
2019-12-19 14:29 ` [PATCH RFC bpf-next 1/3] libbpf: Add new bpf_object__load2() using new-style opts Toke Høiland-Jørgensen
2019-12-19 23:50   ` Andrii Nakryiko [this message]
2019-12-20 10:50     ` Toke Høiland-Jørgensen
2019-12-19 14:29 ` [PATCH RFC bpf-next 2/3] libbpf: Handle function externs and support static linking Toke Høiland-Jørgensen
2019-12-19 16:24   ` Yonghong Song
2019-12-19 16:59     ` Toke Høiland-Jørgensen
2019-12-20  0:02   ` Andrii Nakryiko
2019-12-20 10:47     ` Toke Høiland-Jørgensen
2019-12-20 17:28       ` Andrii Nakryiko
2019-12-19 14:29 ` [PATCH RFC bpf-next 3/3] selftests/bpf: Add selftest for XDP multiprogs Toke Høiland-Jørgensen
2019-12-20 20:30 ` [PATCH RFC bpf-next 0/3] libbpf: Add support for extern function calls Alexei Starovoitov
2019-12-21 16:24   ` Toke Høiland-Jørgensen

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=CAEf4BzYKpstQk8JO_iOws93VpHEEs+J+z+ZO7cKRiKRNvN1zMg@mail.gmail.com \
    --to=andrii.nakryiko@gmail.com \
    --cc=ast@kernel.org \
    --cc=bpf@vger.kernel.org \
    --cc=brouer@redhat.com \
    --cc=daniel@iogearbox.net \
    --cc=davem@davemloft.net \
    --cc=kafai@fb.com \
    --cc=netdev@vger.kernel.org \
    --cc=songliubraving@fb.com \
    --cc=toke@redhat.com \
    --cc=yhs@fb.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).