Merge branch 'bpf_metadata'

Stanislav Fomichev says:

====================
Currently, if a user wants to store arbitrary metadata for an eBPF
program, for example, the program build commit hash or version, they
could store it in a map, and conveniently libbpf uses .data section to
populate an internal map. However, if the program does not actually
reference the map, then the map would be de-refcounted and freed.

This patch set introduces a new syscall BPF_PROG_BIND_MAP to add a map
to a program's used_maps, even if the program instructions does not
reference the map.

libbpf is extended to always BPF_PROG_BIND_MAP .rodata section so the
metadata is kept in place.
bpftool is also extended to print metadata in the 'bpftool prog' list.

The variable is considered metadata if it starts with the
magic 'bpf_metadata_' prefix; everything after the prefix is the
metadata name.

An example use of this would be BPF C file declaring:

  volatile const char bpf_metadata_commit_hash[] SEC(".rodata") = "abcdef123456";

and bpftool would emit:

  $ bpftool prog
  [...]
        metadata:
                commit_hash = "abcdef123456"

v6 changes:
* libbpf: drop FEAT_GLOBAL_DATA from probe_prog_bind_map (Andrii Nakryiko)
* bpftool: combine find_metadata_map_id & find_metadata;
  drops extra bpf_map_get_fd_by_id and bpf_map_get_fd_by_id (Andrii Nakryiko)
* bpftool: use strncmp instead of strstr (Andrii Nakryiko)
* bpftool: memset(map_info) and extra empty line (Andrii Nakryiko)

v5 changes:
* selftest: verify that prog holds rodata (Andrii Nakryiko)
* selftest: use volatile for metadata (Andrii Nakryiko)
* bpftool: use sizeof in BPF_METADATA_PREFIX_LEN (Andrii Nakryiko)
* bpftool: new find_metadata that does map lookup (Andrii Nakryiko)
* libbpf: don't generalize probe_create_global_data (Andrii Nakryiko)
* libbpf: use OPTS_VALID in bpf_prog_bind_map (Andrii Nakryiko)
* libbpf: keep LIBBPF_0.2.0 sorted (Andrii Nakryiko)

v4 changes:
* Don't return EEXIST from syscall if already bound (Andrii Nakryiko)
* Removed --metadata argument (Andrii Nakryiko)
* Removed custom .metadata section (Alexei Starovoitov)
* Addressed Andrii's suggestions about btf helpers and vsi (Andrii Nakryiko)
* Moved bpf_prog_find_metadata into bpftool (Alexei Starovoitov)

v3 changes:
* API changes for bpf_prog_find_metadata (Toke Høiland-Jørgensen)

v2 changes:
* Made struct bpf_prog_bind_opts in libbpf so flags is optional.
* Deduped probe_kern_global_data and probe_prog_bind_map into a common
  helper.
* Added comment regarding why EEXIST is ignored in libbpf bind map.
* Froze all LIBBPF_MAP_METADATA internal maps.
* Moved bpf_prog_bind_map into new LIBBPF_0.1.1 in libbpf.map.
* Added p_err() calls on error cases in bpftool show_prog_metadata.
* Reverse christmas tree coding style in bpftool show_prog_metadata.
* Made bpftool gen skeleton recognize .metadata as an internal map and
  generate datasec definition in skeleton.
* Added C test using skeleton to see asset that the metadata is what we
  expect and rebinding causes EEXIST.

v1 changes:
* Fixed a few missing unlocks, and missing close while iterating map fds.
* Move mutex initialization to right after prog aux allocation, and mutex
  destroy to right after prog aux free.
* s/ADD_MAP/BIND_MAP/
* Use mutex only instead of RCU to protect the used_map array & count.

Cc: YiFei Zhu <zhuyifei1999@gmail.com>
====================

Acked-by: Andrii Nakryiko <andriin@fb.com>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>

3 files changed, 208 insertions, 0 deletions

diff --git a/tools/bpf/bpftool/json_writer.c b/tools/bpf/bpftool/json_writer.c
index 86501cd3c763..7fea83bedf48 100644
--- a/tools/bpf/bpftool/json_writer.c
+++ b/tools/bpf/bpftool/json_writer.c
@@ -119,6 +119,12 @@ void jsonw_pretty(json_writer_t *self, bool on)
 	self->pretty = on;
 }
 
+void jsonw_reset(json_writer_t *self)
+{
+	assert(self->depth == 0);
+	self->sep = '\0';
+}
+
 /* Basic blocks */
 static void jsonw_begin(json_writer_t *self, int c)
 {
diff --git a/tools/bpf/bpftool/json_writer.h b/tools/bpf/bpftool/json_writer.h
index 35cf1f00f96c..8ace65cdb92f 100644
--- a/tools/bpf/bpftool/json_writer.h
+++ b/tools/bpf/bpftool/json_writer.h
@@ -27,6 +27,9 @@ void jsonw_destroy(json_writer_t **self_p);
 /* Cause output to have pretty whitespace */
 void jsonw_pretty(json_writer_t *self, bool on);
 
+/* Reset separator to create new JSON */
+void jsonw_reset(json_writer_t *self);
+
 /* Add property name */
 void jsonw_name(json_writer_t *self, const char *name);
 
diff --git a/tools/bpf/bpftool/prog.c b/tools/bpf/bpftool/prog.c
index f7923414a052..d942c1e3372c 100644
--- a/tools/bpf/bpftool/prog.c
+++ b/tools/bpf/bpftool/prog.c
@@ -29,6 +29,9 @@
 #include "main.h"
 #include "xlated_dumper.h"
 
+#define BPF_METADATA_PREFIX "bpf_metadata_"
+#define BPF_METADATA_PREFIX_LEN (sizeof(BPF_METADATA_PREFIX) - 1)
+
 const char * const prog_type_name[] = {
 	[BPF_PROG_TYPE_UNSPEC]			= "unspec",
 	[BPF_PROG_TYPE_SOCKET_FILTER]		= "socket_filter",
@@ -151,6 +154,198 @@ static void show_prog_maps(int fd, __u32 num_maps)
 	}
 }
 
+static void *find_metadata(int prog_fd, struct bpf_map_info *map_info)
+{
+	struct bpf_prog_info prog_info;
+	__u32 prog_info_len;
+	__u32 map_info_len;
+	void *value = NULL;
+	__u32 *map_ids;
+	int nr_maps;
+	int key = 0;
+	int map_fd;
+	int ret;
+	__u32 i;
+
+	memset(&prog_info, 0, sizeof(prog_info));
+	prog_info_len = sizeof(prog_info);
+	ret = bpf_obj_get_info_by_fd(prog_fd, &prog_info, &prog_info_len);
+	if (ret)
+		return NULL;
+
+	if (!prog_info.nr_map_ids)
+		return NULL;
+
+	map_ids = calloc(prog_info.nr_map_ids, sizeof(__u32));
+	if (!map_ids)
+		return NULL;
+
+	nr_maps = prog_info.nr_map_ids;
+	memset(&prog_info, 0, sizeof(prog_info));
+	prog_info.nr_map_ids = nr_maps;
+	prog_info.map_ids = ptr_to_u64(map_ids);
+	prog_info_len = sizeof(prog_info);
+
+	ret = bpf_obj_get_info_by_fd(prog_fd, &prog_info, &prog_info_len);
+	if (ret)
+		goto free_map_ids;
+
+	for (i = 0; i < prog_info.nr_map_ids; i++) {
+		map_fd = bpf_map_get_fd_by_id(map_ids[i]);
+		if (map_fd < 0)
+			goto free_map_ids;
+
+		memset(map_info, 0, sizeof(*map_info));
+		map_info_len = sizeof(*map_info);
+		ret = bpf_obj_get_info_by_fd(map_fd, map_info, &map_info_len);
+		if (ret < 0) {
+			close(map_fd);
+			goto free_map_ids;
+		}
+
+		if (map_info->type != BPF_MAP_TYPE_ARRAY ||
+		    map_info->key_size != sizeof(int) ||
+		    map_info->max_entries != 1 ||
+		    !map_info->btf_value_type_id ||
+		    !strstr(map_info->name, ".rodata")) {
+			close(map_fd);
+			continue;
+		}
+
+		value = malloc(map_info->value_size);
+		if (!value) {
+			close(map_fd);
+			goto free_map_ids;
+		}
+
+		if (bpf_map_lookup_elem(map_fd, &key, value)) {
+			close(map_fd);
+			free(value);
+			value = NULL;
+			goto free_map_ids;
+		}
+
+		close(map_fd);
+		break;
+	}
+
+free_map_ids:
+	free(map_ids);
+	return value;
+}
+
+static bool has_metadata_prefix(const char *s)
+{
+	return strncmp(s, BPF_METADATA_PREFIX, BPF_METADATA_PREFIX_LEN) == 0;
+}
+
+static void show_prog_metadata(int fd, __u32 num_maps)
+{
+	const struct btf_type *t_datasec, *t_var;
+	struct bpf_map_info map_info;
+	struct btf_var_secinfo *vsi;
+	bool printed_header = false;
+	struct btf *btf = NULL;
+	unsigned int i, vlen;
+	void *value = NULL;
+	const char *name;
+	int err;
+
+	if (!num_maps)
+		return;
+
+	memset(&map_info, 0, sizeof(map_info));
+	value = find_metadata(fd, &map_info);
+	if (!value)
+		return;
+
+	err = btf__get_from_id(map_info.btf_id, &btf);
+	if (err || !btf)
+		goto out_free;
+
+	t_datasec = btf__type_by_id(btf, map_info.btf_value_type_id);
+	if (!btf_is_datasec(t_datasec))
+		goto out_free;
+
+	vlen = btf_vlen(t_datasec);
+	vsi = btf_var_secinfos(t_datasec);
+
+	/* We don't proceed to check the kinds of the elements of the DATASEC.
+	 * The verifier enforces them to be BTF_KIND_VAR.
+	 */
+
+	if (json_output) {
+		struct btf_dumper d = {
+			.btf = btf,
+			.jw = json_wtr,
+			.is_plain_text = false,
+		};
+
+		for (i = 0; i < vlen; i++, vsi++) {
+			t_var = btf__type_by_id(btf, vsi->type);
+			name = btf__name_by_offset(btf, t_var->name_off);
+
+			if (!has_metadata_prefix(name))
+				continue;
+
+			if (!printed_header) {
+				jsonw_name(json_wtr, "metadata");
+				jsonw_start_object(json_wtr);
+				printed_header = true;
+			}
+
+			jsonw_name(json_wtr, name + BPF_METADATA_PREFIX_LEN);
+			err = btf_dumper_type(&d, t_var->type, value + vsi->offset);
+			if (err) {
+				p_err("btf dump failed: %d", err);
+				break;
+			}
+		}
+		if (printed_header)
+			jsonw_end_object(json_wtr);
+	} else {
+		json_writer_t *btf_wtr = jsonw_new(stdout);
+		struct btf_dumper d = {
+			.btf = btf,
+			.jw = btf_wtr,
+			.is_plain_text = true,
+		};
+
+		if (!btf_wtr) {
+			p_err("jsonw alloc failed");
+			goto out_free;
+		}
+
+		for (i = 0; i < vlen; i++, vsi++) {
+			t_var = btf__type_by_id(btf, vsi->type);
+			name = btf__name_by_offset(btf, t_var->name_off);
+
+			if (!has_metadata_prefix(name))
+				continue;
+
+			if (!printed_header) {
+				printf("\tmetadata:");
+				printed_header = true;
+			}
+
+			printf("\n\t\t%s = ", name + BPF_METADATA_PREFIX_LEN);
+
+			jsonw_reset(btf_wtr);
+			err = btf_dumper_type(&d, t_var->type, value + vsi->offset);
+			if (err) {
+				p_err("btf dump failed: %d", err);
+				break;
+			}
+		}
+		if (printed_header)
+			jsonw_destroy(&btf_wtr);
+	}
+
+out_free:
+	btf__free(btf);
+	free(value);
+}
+
 static void print_prog_header_json(struct bpf_prog_info *info)
 {
 	jsonw_uint_field(json_wtr, "id", info->id);
@@ -228,6 +423,8 @@ static void print_prog_json(struct bpf_prog_info *info, int fd)
 
 	emit_obj_refs_json(&refs_table, info->id, json_wtr);
 
+	show_prog_metadata(fd, info->nr_map_ids);
+
 	jsonw_end_object(json_wtr);
 }
 
@@ -297,6 +494,8 @@ static void print_prog_plain(struct bpf_prog_info *info, int fd)
 	emit_obj_refs_plain(&refs_table, info->id, "\n\tpids ");
 
 	printf("\n");
+
+	show_prog_metadata(fd, info->nr_map_ids);
 }
 
 static int show_prog(int fd)