| lists.openwall.net | lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC | |
|
Open Source and information security mailing list archives
| ||
|
Message-ID: <CAO-hwJKwDWJ6ZKK=+BjrDhjfyG00VKFznJ+HO-0MV1AQ1U8D-Q@mail.gmail.com>
Date: Thu, 1 Dec 2022 15:22:49 +0100
From: Benjamin Tissoires <benjamin.tissoires@...hat.com>
To: Andrii Nakryiko <andrii.nakryiko@...il.com>
Cc: Greg KH <gregkh@...uxfoundation.org>,
Jiri Kosina <jikos@...nel.org>,
Alexei Starovoitov <ast@...nel.org>,
Daniel Borkmann <daniel@...earbox.net>,
Andrii Nakryiko <andrii@...nel.org>,
Dmitry Torokhov <dmitry.torokhov@...il.com>,
Tero Kristo <tero.kristo@...ux.intel.com>,
linux-kernel@...r.kernel.org, linux-input@...r.kernel.org,
netdev@...r.kernel.org, bpf@...r.kernel.org
Subject: Re: [RFC hid v1 01/10] bpftool: generate json output of skeletons
On Thu, Dec 1, 2022 at 12:06 AM Andrii Nakryiko
<andrii.nakryiko@...il.com> wrote:
>
> On Thu, Nov 24, 2022 at 7:16 AM Benjamin Tissoires
> <benjamin.tissoires@...hat.com> wrote:
> >
> > So we can then build light skeletons with loaders in any language.
> >
>
> It would be useful to include an example generated JSON. Other than
> that the overall idea makes sense. Kind of machine-friendly "BPF
> object schema" to allow automation.
>
Great :)
I'll add the generated example in v2 then.
However, I have problems figuring out how I can hit the
"codegen_datasecs()" path. I think that's the only code path I am not
duplicating from the do_skeleton() function, but I don't know what
kind of .bpf.c program will trigger that part.
Also, should I add some tests for it in tools/testing/selftests/bpf?
Cheers,
Benjamin
> > Signed-off-by: Benjamin Tissoires <benjamin.tissoires@...hat.com>
> > ---
> > tools/bpf/bpftool/gen.c | 95 +++++++++++++++++++++++++++++++++++++++++
> > 1 file changed, 95 insertions(+)
> >
> > diff --git a/tools/bpf/bpftool/gen.c b/tools/bpf/bpftool/gen.c
> > index cf8b4e525c88..818a5209b3ac 100644
> > --- a/tools/bpf/bpftool/gen.c
> > +++ b/tools/bpf/bpftool/gen.c
> > @@ -904,6 +904,96 @@ codegen_progs_skeleton(struct bpf_object *obj, size_t prog_cnt, bool populate_li
> > }
> > }
> >
> > +static int gen_json(struct bpf_object *obj, const char *obj_name, size_t file_sz, uint8_t *obj_data)
> > +{
> > + struct bpf_program *prog;
> > + struct bpf_map *map;
> > + char ident[256];
> > +
> > + jsonw_start_object(json_wtr); /* root object */
> > +
> > + jsonw_string_field(json_wtr, "name", obj_name);
> > +
> > + jsonw_bool_field(json_wtr, "use_loader", use_loader);
> > +
> > + /* print all maps */
> > + jsonw_name(json_wtr, "maps");
> > + jsonw_start_array(json_wtr);
> > + bpf_object__for_each_map(map, obj) {
> > + if (!get_map_ident(map, ident, sizeof(ident))) {
> > + p_err("ignoring unrecognized internal map '%s'...",
> > + bpf_map__name(map));
> > + continue;
> > + }
> > +
> > + jsonw_start_object(json_wtr); /* map object */
> > + jsonw_string_field(json_wtr, "ident", ident);
> > + jsonw_string_field(json_wtr, "name", bpf_map__name(map));
> > +
> > + /* print mmap data value */
> > + if (is_internal_mmapable_map(map, ident, sizeof(ident))) {
> > + const void *mmap_data = NULL;
> > + size_t mmap_size = 0;
> > +
> > + mmap_data = bpf_map__initial_value(map, &mmap_size);
> > +
> > + jsonw_uint_field(json_wtr, "size", mmap_size);
> > + jsonw_uint_field(json_wtr, "mmap_sz", bpf_map_mmap_sz(map));
> > + jsonw_name(json_wtr, "data");
> > + print_hex_data_json((uint8_t *)mmap_data, mmap_size);
> > +
> > + }
> > + jsonw_end_object(json_wtr); /* map object */
> > + }
> > + jsonw_end_array(json_wtr);
> > +
> > + /* print all progs */
> > + jsonw_name(json_wtr, "progs");
> > + jsonw_start_array(json_wtr);
> > + bpf_object__for_each_program(prog, obj) {
> > + jsonw_start_object(json_wtr); /* prog object */
> > + jsonw_string_field(json_wtr, "name", bpf_program__name(prog));
> > + jsonw_string_field(json_wtr, "sec", bpf_program__section_name(prog));
> > + jsonw_end_object(json_wtr); /* prog object */
> > + }
> > + jsonw_end_array(json_wtr);
> > +
> > + /* print object data */
> > + if (use_loader) {
> > + DECLARE_LIBBPF_OPTS(gen_loader_opts, opts);
> > + int err = 0;
> > +
> > + err = bpf_object__gen_loader(obj, &opts);
> > + if (err)
> > + return err;
> > +
> > + err = bpf_object__load(obj);
> > + if (err) {
> > + p_err("failed to load object file");
> > + return err;
> > + }
> > + /* If there was no error during load then gen_loader_opts
> > + * are populated with the loader program.
> > + */
> > +
> > + jsonw_uint_field(json_wtr, "data_sz", opts.data_sz);
> > + jsonw_name(json_wtr, "data");
> > + print_hex_data_json((uint8_t *)opts.data, opts.data_sz);
> > +
> > + jsonw_uint_field(json_wtr, "insns_sz", opts.insns_sz);
> > + jsonw_name(json_wtr, "insns");
> > + print_hex_data_json((uint8_t *)opts.insns, opts.insns_sz);
> > +
> > + } else {
> > + jsonw_name(json_wtr, "data");
> > + print_hex_data_json(obj_data, file_sz);
> > + }
> > +
> > + jsonw_end_object(json_wtr); /* root object */
> > +
> > + return 0;
> > +}
> > +
> > static int do_skeleton(int argc, char **argv)
> > {
> > char header_guard[MAX_OBJ_NAME_LEN + sizeof("__SKEL_H__")];
> > @@ -986,6 +1076,11 @@ static int do_skeleton(int argc, char **argv)
> > goto out;
> > }
> >
> > + if (json_output) {
> > + err = gen_json(obj, obj_name, file_sz, (uint8_t *)obj_data);
> > + goto out;
> > + }
> > +
> > bpf_object__for_each_map(map, obj) {
> > if (!get_map_ident(map, ident, sizeof(ident))) {
> > p_err("ignoring unrecognized internal map '%s'...",
> > --
> > 2.38.1
> >
>
Powered by blists - more mailing lists