* [PATCH bpf-next v2 0/5] libbpf 1.0: deprecate bpf_map__def() API
@ 2022-01-08 0:42 Christy Lee
2022-01-08 0:42 ` [PATCH bpf-next v2 1/5] samples/bpf: stop using " Christy Lee
` (5 more replies)
0 siblings, 6 replies; 7+ messages in thread
From: Christy Lee @ 2022-01-08 0:42 UTC (permalink / raw)
To: andrii, acme
Cc: christyc.y.lee, bpf, kernel-team, linux-perf-users, Christy Lee
bpf_map__def() is rarely used and non-extensible. bpf_map_def fields
can be accessed with appropriate map getters and setters instead.
Deprecate bpf_map__def() API and replace use cases with getters and
setters.
Changelog:
----------
v1 -> v2:
https://lore.kernel.org/all/20220105230057.853163-1-christylee@fb.com/
* Fixed commit messages to match commit titles
* Fixed indentation
* Removed bpf_map__def() usage that was missed in v1
Christy Lee (5):
samples/bpf: stop using bpf_map__def() API
bpftool: stop using bpf_map__def() API
perf: stop using bpf_map__def() API
selftests/bpf: stop using bpf_map__def() API
libbpf: deprecate bpf_map__def() API
samples/bpf/xdp_rxq_info_user.c | 10 +--
tools/bpf/bpftool/gen.c | 12 ++--
tools/bpf/bpftool/struct_ops.c | 4 +-
tools/lib/bpf/libbpf.h | 3 +-
tools/perf/util/bpf-loader.c | 64 ++++++++-----------
tools/perf/util/bpf_map.c | 28 ++++----
.../selftests/bpf/prog_tests/flow_dissector.c | 2 +-
.../selftests/bpf/prog_tests/global_data.c | 2 +-
.../bpf/prog_tests/global_data_init.c | 2 +-
.../selftests/bpf/prog_tests/sockmap_listen.c | 12 ++--
.../selftests/bpf/prog_tests/tailcalls.c | 36 +++++------
11 files changed, 81 insertions(+), 94 deletions(-)
--
2.30.2
^ permalink raw reply [flat|nested] 7+ messages in thread
* [PATCH bpf-next v2 1/5] samples/bpf: stop using bpf_map__def() API
2022-01-08 0:42 [PATCH bpf-next v2 0/5] libbpf 1.0: deprecate bpf_map__def() API Christy Lee
@ 2022-01-08 0:42 ` Christy Lee
2022-01-08 0:42 ` [PATCH bpf-next v2 2/5] bpftool: " Christy Lee
` (4 subsequent siblings)
5 siblings, 0 replies; 7+ messages in thread
From: Christy Lee @ 2022-01-08 0:42 UTC (permalink / raw)
To: andrii, acme
Cc: christyc.y.lee, bpf, kernel-team, linux-perf-users, Christy Lee
libbpf bpf_map__def() API is being deprecated, replace samples/bpf's
usage with the appropriate getters and setters.
Signed-off-by: Christy Lee <christylee@fb.com>
---
samples/bpf/xdp_rxq_info_user.c | 10 +++++-----
1 file changed, 5 insertions(+), 5 deletions(-)
diff --git a/samples/bpf/xdp_rxq_info_user.c b/samples/bpf/xdp_rxq_info_user.c
index 74a2926eba08..4033f345aa29 100644
--- a/samples/bpf/xdp_rxq_info_user.c
+++ b/samples/bpf/xdp_rxq_info_user.c
@@ -209,7 +209,7 @@ static struct datarec *alloc_record_per_cpu(void)
static struct record *alloc_record_per_rxq(void)
{
- unsigned int nr_rxqs = bpf_map__def(rx_queue_index_map)->max_entries;
+ unsigned int nr_rxqs = bpf_map__max_entries(rx_queue_index_map);
struct record *array;
array = calloc(nr_rxqs, sizeof(struct record));
@@ -222,7 +222,7 @@ static struct record *alloc_record_per_rxq(void)
static struct stats_record *alloc_stats_record(void)
{
- unsigned int nr_rxqs = bpf_map__def(rx_queue_index_map)->max_entries;
+ unsigned int nr_rxqs = bpf_map__max_entries(rx_queue_index_map);
struct stats_record *rec;
int i;
@@ -241,7 +241,7 @@ static struct stats_record *alloc_stats_record(void)
static void free_stats_record(struct stats_record *r)
{
- unsigned int nr_rxqs = bpf_map__def(rx_queue_index_map)->max_entries;
+ unsigned int nr_rxqs = bpf_map__max_entries(rx_queue_index_map);
int i;
for (i = 0; i < nr_rxqs; i++)
@@ -289,7 +289,7 @@ static void stats_collect(struct stats_record *rec)
map_collect_percpu(fd, 0, &rec->stats);
fd = bpf_map__fd(rx_queue_index_map);
- max_rxqs = bpf_map__def(rx_queue_index_map)->max_entries;
+ max_rxqs = bpf_map__max_entries(rx_queue_index_map);
for (i = 0; i < max_rxqs; i++)
map_collect_percpu(fd, i, &rec->rxq[i]);
}
@@ -335,7 +335,7 @@ static void stats_print(struct stats_record *stats_rec,
struct stats_record *stats_prev,
int action, __u32 cfg_opt)
{
- unsigned int nr_rxqs = bpf_map__def(rx_queue_index_map)->max_entries;
+ unsigned int nr_rxqs = bpf_map__max_entries(rx_queue_index_map);
unsigned int nr_cpus = bpf_num_possible_cpus();
double pps = 0, err = 0;
struct record *rec, *prev;
--
2.30.2
^ permalink raw reply related [flat|nested] 7+ messages in thread
* [PATCH bpf-next v2 2/5] bpftool: stop using bpf_map__def() API
2022-01-08 0:42 [PATCH bpf-next v2 0/5] libbpf 1.0: deprecate bpf_map__def() API Christy Lee
2022-01-08 0:42 ` [PATCH bpf-next v2 1/5] samples/bpf: stop using " Christy Lee
@ 2022-01-08 0:42 ` Christy Lee
2022-01-08 0:42 ` [PATCH bpf-next v2 3/5] perf: " Christy Lee
` (3 subsequent siblings)
5 siblings, 0 replies; 7+ messages in thread
From: Christy Lee @ 2022-01-08 0:42 UTC (permalink / raw)
To: andrii, acme
Cc: christyc.y.lee, bpf, kernel-team, linux-perf-users, Christy Lee
libbpf bpf_map__def() API is being deprecated, replace bpftool's
usage with the appropriate getters and setters
Signed-off-by: Christy Lee <christylee@fb.com>
---
tools/bpf/bpftool/gen.c | 12 ++++++------
tools/bpf/bpftool/struct_ops.c | 4 +---
2 files changed, 7 insertions(+), 9 deletions(-)
diff --git a/tools/bpf/bpftool/gen.c b/tools/bpf/bpftool/gen.c
index b4695df2ea3d..0b6f1553db0b 100644
--- a/tools/bpf/bpftool/gen.c
+++ b/tools/bpf/bpftool/gen.c
@@ -227,7 +227,7 @@ static int codegen_datasecs(struct bpf_object *obj, const char *obj_name)
/* only generate definitions for memory-mapped internal maps */
if (!bpf_map__is_internal(map))
continue;
- if (!(bpf_map__def(map)->map_flags & BPF_F_MMAPABLE))
+ if (!(bpf_map__map_flags(map) & BPF_F_MMAPABLE))
continue;
if (!get_map_ident(map, map_ident, sizeof(map_ident)))
@@ -468,7 +468,7 @@ static void codegen_destroy(struct bpf_object *obj, const char *obj_name)
if (!get_map_ident(map, ident, sizeof(ident)))
continue;
if (bpf_map__is_internal(map) &&
- (bpf_map__def(map)->map_flags & BPF_F_MMAPABLE))
+ (bpf_map__map_flags(map) & BPF_F_MMAPABLE))
printf("\tmunmap(skel->%1$s, %2$zd);\n",
ident, bpf_map_mmap_sz(map));
codegen("\
@@ -536,7 +536,7 @@ static int gen_trace(struct bpf_object *obj, const char *obj_name, const char *h
continue;
if (!bpf_map__is_internal(map) ||
- !(bpf_map__def(map)->map_flags & BPF_F_MMAPABLE))
+ !(bpf_map__map_flags(map) & BPF_F_MMAPABLE))
continue;
codegen("\
@@ -600,10 +600,10 @@ static int gen_trace(struct bpf_object *obj, const char *obj_name, const char *h
continue;
if (!bpf_map__is_internal(map) ||
- !(bpf_map__def(map)->map_flags & BPF_F_MMAPABLE))
+ !(bpf_map__map_flags(map) & BPF_F_MMAPABLE))
continue;
- if (bpf_map__def(map)->map_flags & BPF_F_RDONLY_PROG)
+ if (bpf_map__map_flags(map) & BPF_F_RDONLY_PROG)
mmap_flags = "PROT_READ";
else
mmap_flags = "PROT_READ | PROT_WRITE";
@@ -962,7 +962,7 @@ static int do_skeleton(int argc, char **argv)
i, bpf_map__name(map), i, ident);
/* memory-mapped internal maps */
if (bpf_map__is_internal(map) &&
- (bpf_map__def(map)->map_flags & BPF_F_MMAPABLE)) {
+ (bpf_map__map_flags(map) & BPF_F_MMAPABLE)) {
printf("\ts->maps[%zu].mmaped = (void **)&obj->%s;\n",
i, ident);
}
diff --git a/tools/bpf/bpftool/struct_ops.c b/tools/bpf/bpftool/struct_ops.c
index 2f693b082bdb..e08a6ff2866c 100644
--- a/tools/bpf/bpftool/struct_ops.c
+++ b/tools/bpf/bpftool/struct_ops.c
@@ -480,7 +480,6 @@ static int do_unregister(int argc, char **argv)
static int do_register(int argc, char **argv)
{
LIBBPF_OPTS(bpf_object_open_opts, open_opts);
- const struct bpf_map_def *def;
struct bpf_map_info info = {};
__u32 info_len = sizeof(info);
int nr_errs = 0, nr_maps = 0;
@@ -510,8 +509,7 @@ static int do_register(int argc, char **argv)
}
bpf_object__for_each_map(map, obj) {
- def = bpf_map__def(map);
- if (def->type != BPF_MAP_TYPE_STRUCT_OPS)
+ if (bpf_map__type(map) != BPF_MAP_TYPE_STRUCT_OPS)
continue;
link = bpf_map__attach_struct_ops(map);
--
2.30.2
^ permalink raw reply related [flat|nested] 7+ messages in thread
* [PATCH bpf-next v2 3/5] perf: stop using bpf_map__def() API
2022-01-08 0:42 [PATCH bpf-next v2 0/5] libbpf 1.0: deprecate bpf_map__def() API Christy Lee
2022-01-08 0:42 ` [PATCH bpf-next v2 1/5] samples/bpf: stop using " Christy Lee
2022-01-08 0:42 ` [PATCH bpf-next v2 2/5] bpftool: " Christy Lee
@ 2022-01-08 0:42 ` Christy Lee
2022-01-08 0:42 ` [PATCH bpf-next v2 4/5] selftests/bpf: " Christy Lee
` (2 subsequent siblings)
5 siblings, 0 replies; 7+ messages in thread
From: Christy Lee @ 2022-01-08 0:42 UTC (permalink / raw)
To: andrii, acme
Cc: christyc.y.lee, bpf, kernel-team, linux-perf-users, Christy Lee
libbpf bpf_map__def() API is being deprecated, replace perf's
usage with the appropriate getters and setters.
Signed-off-by: Christy Lee <christylee@fb.com>
---
tools/perf/util/bpf-loader.c | 64 +++++++++++++++---------------------
tools/perf/util/bpf_map.c | 28 ++++++++--------
2 files changed, 41 insertions(+), 51 deletions(-)
diff --git a/tools/perf/util/bpf-loader.c b/tools/perf/util/bpf-loader.c
index 528aeb0ab79d..4631cac3957f 100644
--- a/tools/perf/util/bpf-loader.c
+++ b/tools/perf/util/bpf-loader.c
@@ -1002,24 +1002,22 @@ __bpf_map__config_value(struct bpf_map *map,
{
struct bpf_map_op *op;
const char *map_name = bpf_map__name(map);
- const struct bpf_map_def *def = bpf_map__def(map);
- if (IS_ERR(def)) {
- pr_debug("Unable to get map definition from '%s'\n",
- map_name);
+ if (!map) {
+ pr_debug("Map '%s' is invalid\n", map_name);
return -BPF_LOADER_ERRNO__INTERNAL;
}
- if (def->type != BPF_MAP_TYPE_ARRAY) {
+ if (bpf_map__type(map) != BPF_MAP_TYPE_ARRAY) {
pr_debug("Map %s type is not BPF_MAP_TYPE_ARRAY\n",
map_name);
return -BPF_LOADER_ERRNO__OBJCONF_MAP_TYPE;
}
- if (def->key_size < sizeof(unsigned int)) {
+ if (bpf_map__key_size(map) < sizeof(unsigned int)) {
pr_debug("Map %s has incorrect key size\n", map_name);
return -BPF_LOADER_ERRNO__OBJCONF_MAP_KEYSIZE;
}
- switch (def->value_size) {
+ switch (bpf_map__value_size(map)) {
case 1:
case 2:
case 4:
@@ -1061,7 +1059,6 @@ __bpf_map__config_event(struct bpf_map *map,
struct parse_events_term *term,
struct evlist *evlist)
{
- const struct bpf_map_def *def;
struct bpf_map_op *op;
const char *map_name = bpf_map__name(map);
struct evsel *evsel = evlist__find_evsel_by_str(evlist, term->val.str);
@@ -1072,18 +1069,16 @@ __bpf_map__config_event(struct bpf_map *map,
return -BPF_LOADER_ERRNO__OBJCONF_MAP_NOEVT;
}
- def = bpf_map__def(map);
- if (IS_ERR(def)) {
- pr_debug("Unable to get map definition from '%s'\n",
- map_name);
- return PTR_ERR(def);
+ if (!map) {
+ pr_debug("Map '%s' is invalid\n", map_name);
+ return PTR_ERR(map);
}
/*
* No need to check key_size and value_size:
* kernel has already checked them.
*/
- if (def->type != BPF_MAP_TYPE_PERF_EVENT_ARRAY) {
+ if (bpf_map__type(map) != BPF_MAP_TYPE_PERF_EVENT_ARRAY) {
pr_debug("Map %s type is not BPF_MAP_TYPE_PERF_EVENT_ARRAY\n",
map_name);
return -BPF_LOADER_ERRNO__OBJCONF_MAP_TYPE;
@@ -1132,7 +1127,6 @@ config_map_indices_range_check(struct parse_events_term *term,
const char *map_name)
{
struct parse_events_array *array = &term->array;
- const struct bpf_map_def *def;
unsigned int i;
if (!array->nr_ranges)
@@ -1143,10 +1137,8 @@ config_map_indices_range_check(struct parse_events_term *term,
return -BPF_LOADER_ERRNO__INTERNAL;
}
- def = bpf_map__def(map);
- if (IS_ERR(def)) {
- pr_debug("ERROR: Unable to get map definition from '%s'\n",
- map_name);
+ if (!map) {
+ pr_debug("Map '%s' is invalid\n", map_name);
return -BPF_LOADER_ERRNO__INTERNAL;
}
@@ -1155,7 +1147,7 @@ config_map_indices_range_check(struct parse_events_term *term,
size_t length = array->ranges[i].length;
unsigned int idx = start + length - 1;
- if (idx >= def->max_entries) {
+ if (idx >= bpf_map__max_entries(map)) {
pr_debug("ERROR: index %d too large\n", idx);
return -BPF_LOADER_ERRNO__OBJCONF_MAP_IDX2BIG;
}
@@ -1248,21 +1240,21 @@ int bpf__config_obj(struct bpf_object *obj,
}
typedef int (*map_config_func_t)(const char *name, int map_fd,
- const struct bpf_map_def *pdef,
+ const struct bpf_map *map,
struct bpf_map_op *op,
void *pkey, void *arg);
static int
foreach_key_array_all(map_config_func_t func,
void *arg, const char *name,
- int map_fd, const struct bpf_map_def *pdef,
+ int map_fd, const struct bpf_map *map,
struct bpf_map_op *op)
{
unsigned int i;
int err;
- for (i = 0; i < pdef->max_entries; i++) {
- err = func(name, map_fd, pdef, op, &i, arg);
+ for (i = 0; i < bpf_map__max_entries(map); i++) {
+ err = func(name, map_fd, map, op, &i, arg);
if (err) {
pr_debug("ERROR: failed to insert value to %s[%u]\n",
name, i);
@@ -1275,7 +1267,7 @@ foreach_key_array_all(map_config_func_t func,
static int
foreach_key_array_ranges(map_config_func_t func, void *arg,
const char *name, int map_fd,
- const struct bpf_map_def *pdef,
+ const struct bpf_map *map,
struct bpf_map_op *op)
{
unsigned int i, j;
@@ -1288,7 +1280,7 @@ foreach_key_array_ranges(map_config_func_t func, void *arg,
for (j = 0; j < length; j++) {
unsigned int idx = start + j;
- err = func(name, map_fd, pdef, op, &idx, arg);
+ err = func(name, map_fd, map, op, &idx, arg);
if (err) {
pr_debug("ERROR: failed to insert value to %s[%u]\n",
name, idx);
@@ -1304,9 +1296,8 @@ bpf_map_config_foreach_key(struct bpf_map *map,
map_config_func_t func,
void *arg)
{
- int err, map_fd;
+ int err, map_fd, type;
struct bpf_map_op *op;
- const struct bpf_map_def *def;
const char *name = bpf_map__name(map);
struct bpf_map_priv *priv = bpf_map__priv(map);
@@ -1319,9 +1310,8 @@ bpf_map_config_foreach_key(struct bpf_map *map,
return 0;
}
- def = bpf_map__def(map);
- if (IS_ERR(def)) {
- pr_debug("ERROR: failed to get definition from map %s\n", name);
+ if (!map) {
+ pr_debug("Map '%s' is invalid\n", name);
return -BPF_LOADER_ERRNO__INTERNAL;
}
map_fd = bpf_map__fd(map);
@@ -1330,19 +1320,19 @@ bpf_map_config_foreach_key(struct bpf_map *map,
return map_fd;
}
+ type = bpf_map__type(map);
list_for_each_entry(op, &priv->ops_list, list) {
- switch (def->type) {
+ switch (type) {
case BPF_MAP_TYPE_ARRAY:
case BPF_MAP_TYPE_PERF_EVENT_ARRAY:
switch (op->key_type) {
case BPF_MAP_KEY_ALL:
err = foreach_key_array_all(func, arg, name,
- map_fd, def, op);
+ map_fd, map, op);
break;
case BPF_MAP_KEY_RANGES:
err = foreach_key_array_ranges(func, arg, name,
- map_fd, def,
- op);
+ map_fd, map, op);
break;
default:
pr_debug("ERROR: keytype for map '%s' invalid\n",
@@ -1451,7 +1441,7 @@ apply_config_evsel_for_key(const char *name, int map_fd, void *pkey,
static int
apply_obj_config_map_for_key(const char *name, int map_fd,
- const struct bpf_map_def *pdef,
+ const struct bpf_map *map,
struct bpf_map_op *op,
void *pkey, void *arg __maybe_unused)
{
@@ -1460,7 +1450,7 @@ apply_obj_config_map_for_key(const char *name, int map_fd,
switch (op->op_type) {
case BPF_MAP_OP_SET_VALUE:
err = apply_config_value_for_key(map_fd, pkey,
- pdef->value_size,
+ bpf_map__value_size(map),
op->v.value);
break;
case BPF_MAP_OP_SET_EVSEL:
diff --git a/tools/perf/util/bpf_map.c b/tools/perf/util/bpf_map.c
index eb853ca67cf4..c863ae0c5cb5 100644
--- a/tools/perf/util/bpf_map.c
+++ b/tools/perf/util/bpf_map.c
@@ -9,25 +9,25 @@
#include <stdlib.h>
#include <unistd.h>
-static bool bpf_map_def__is_per_cpu(const struct bpf_map_def *def)
+static bool bpf_map__is_per_cpu(enum bpf_map_type type)
{
- return def->type == BPF_MAP_TYPE_PERCPU_HASH ||
- def->type == BPF_MAP_TYPE_PERCPU_ARRAY ||
- def->type == BPF_MAP_TYPE_LRU_PERCPU_HASH ||
- def->type == BPF_MAP_TYPE_PERCPU_CGROUP_STORAGE;
+ return type == BPF_MAP_TYPE_PERCPU_HASH ||
+ type == BPF_MAP_TYPE_PERCPU_ARRAY ||
+ type == BPF_MAP_TYPE_LRU_PERCPU_HASH ||
+ type == BPF_MAP_TYPE_PERCPU_CGROUP_STORAGE;
}
-static void *bpf_map_def__alloc_value(const struct bpf_map_def *def)
+static void *bpf_map__alloc_value(const struct bpf_map *map)
{
- if (bpf_map_def__is_per_cpu(def))
- return malloc(round_up(def->value_size, 8) * sysconf(_SC_NPROCESSORS_CONF));
+ if (bpf_map__is_per_cpu(bpf_map__type(map)))
+ return malloc(round_up(bpf_map__value_size(map), 8) *
+ sysconf(_SC_NPROCESSORS_CONF));
- return malloc(def->value_size);
+ return malloc(bpf_map__value_size(map));
}
int bpf_map__fprintf(struct bpf_map *map, FILE *fp)
{
- const struct bpf_map_def *def = bpf_map__def(map);
void *prev_key = NULL, *key, *value;
int fd = bpf_map__fd(map), err;
int printed = 0;
@@ -35,15 +35,15 @@ int bpf_map__fprintf(struct bpf_map *map, FILE *fp)
if (fd < 0)
return fd;
- if (IS_ERR(def))
- return PTR_ERR(def);
+ if (!map)
+ return PTR_ERR(map);
err = -ENOMEM;
- key = malloc(def->key_size);
+ key = malloc(bpf_map__key_size(map));
if (key == NULL)
goto out;
- value = bpf_map_def__alloc_value(def);
+ value = bpf_map__alloc_value(map);
if (value == NULL)
goto out_free_key;
--
2.30.2
^ permalink raw reply related [flat|nested] 7+ messages in thread
* [PATCH bpf-next v2 4/5] selftests/bpf: stop using bpf_map__def() API
2022-01-08 0:42 [PATCH bpf-next v2 0/5] libbpf 1.0: deprecate bpf_map__def() API Christy Lee
` (2 preceding siblings ...)
2022-01-08 0:42 ` [PATCH bpf-next v2 3/5] perf: " Christy Lee
@ 2022-01-08 0:42 ` Christy Lee
2022-01-08 0:42 ` [PATCH bpf-next v2 5/5] libbpf: deprecate " Christy Lee
2022-01-10 2:10 ` [PATCH bpf-next v2 0/5] libbpf 1.0: " patchwork-bot+netdevbpf
5 siblings, 0 replies; 7+ messages in thread
From: Christy Lee @ 2022-01-08 0:42 UTC (permalink / raw)
To: andrii, acme
Cc: christyc.y.lee, bpf, kernel-team, linux-perf-users, Christy Lee
libbpf bpf_map__def() API is being deprecated, replace selftests/bpf's
usage with the appropriate getters and setters.
Signed-off-by: Christy Lee <christylee@fb.com>
---
.../selftests/bpf/prog_tests/flow_dissector.c | 2 +-
.../selftests/bpf/prog_tests/global_data.c | 2 +-
.../bpf/prog_tests/global_data_init.c | 2 +-
.../selftests/bpf/prog_tests/sockmap_listen.c | 12 +++----
.../selftests/bpf/prog_tests/tailcalls.c | 36 +++++++++----------
5 files changed, 26 insertions(+), 28 deletions(-)
diff --git a/tools/testing/selftests/bpf/prog_tests/flow_dissector.c b/tools/testing/selftests/bpf/prog_tests/flow_dissector.c
index ac54e3f91d42..dfafd62df50b 100644
--- a/tools/testing/selftests/bpf/prog_tests/flow_dissector.c
+++ b/tools/testing/selftests/bpf/prog_tests/flow_dissector.c
@@ -457,7 +457,7 @@ static int init_prog_array(struct bpf_object *obj, struct bpf_map *prog_array)
if (map_fd < 0)
return -1;
- for (i = 0; i < bpf_map__def(prog_array)->max_entries; i++) {
+ for (i = 0; i < bpf_map__max_entries(prog_array); i++) {
snprintf(prog_name, sizeof(prog_name), "flow_dissector_%d", i);
prog = bpf_object__find_program_by_name(obj, prog_name);
diff --git a/tools/testing/selftests/bpf/prog_tests/global_data.c b/tools/testing/selftests/bpf/prog_tests/global_data.c
index 9da131b32e13..917165e04427 100644
--- a/tools/testing/selftests/bpf/prog_tests/global_data.c
+++ b/tools/testing/selftests/bpf/prog_tests/global_data.c
@@ -121,7 +121,7 @@ static void test_global_data_rdonly(struct bpf_object *obj, __u32 duration)
if (CHECK_FAIL(map_fd < 0))
return;
- buff = malloc(bpf_map__def(map)->value_size);
+ buff = malloc(bpf_map__value_size(map));
if (buff)
err = bpf_map_update_elem(map_fd, &zero, buff, 0);
free(buff);
diff --git a/tools/testing/selftests/bpf/prog_tests/global_data_init.c b/tools/testing/selftests/bpf/prog_tests/global_data_init.c
index 1db86eab101b..57331c606964 100644
--- a/tools/testing/selftests/bpf/prog_tests/global_data_init.c
+++ b/tools/testing/selftests/bpf/prog_tests/global_data_init.c
@@ -20,7 +20,7 @@ void test_global_data_init(void)
if (CHECK_FAIL(!map || !bpf_map__is_internal(map)))
goto out;
- sz = bpf_map__def(map)->value_size;
+ sz = bpf_map__value_size(map);
newval = malloc(sz);
if (CHECK_FAIL(!newval))
goto out;
diff --git a/tools/testing/selftests/bpf/prog_tests/sockmap_listen.c b/tools/testing/selftests/bpf/prog_tests/sockmap_listen.c
index 7e21bfab6358..2cf0c7a3fe23 100644
--- a/tools/testing/selftests/bpf/prog_tests/sockmap_listen.c
+++ b/tools/testing/selftests/bpf/prog_tests/sockmap_listen.c
@@ -1413,14 +1413,12 @@ static void test_reuseport_mixed_groups(int family, int sotype, int sock_map,
static void test_ops_cleanup(const struct bpf_map *map)
{
- const struct bpf_map_def *def;
int err, mapfd;
u32 key;
- def = bpf_map__def(map);
mapfd = bpf_map__fd(map);
- for (key = 0; key < def->max_entries; key++) {
+ for (key = 0; key < bpf_map__max_entries(map); key++) {
err = bpf_map_delete_elem(mapfd, &key);
if (err && errno != EINVAL && errno != ENOENT)
FAIL_ERRNO("map_delete: expected EINVAL/ENOENT");
@@ -1443,13 +1441,13 @@ static const char *family_str(sa_family_t family)
static const char *map_type_str(const struct bpf_map *map)
{
- const struct bpf_map_def *def;
+ int type;
- def = bpf_map__def(map);
- if (IS_ERR(def))
+ if (!map)
return "invalid";
+ type = bpf_map__type(map);
- switch (def->type) {
+ switch (type) {
case BPF_MAP_TYPE_SOCKMAP:
return "sockmap";
case BPF_MAP_TYPE_SOCKHASH:
diff --git a/tools/testing/selftests/bpf/prog_tests/tailcalls.c b/tools/testing/selftests/bpf/prog_tests/tailcalls.c
index 5dc0f425bd11..796f231582f8 100644
--- a/tools/testing/selftests/bpf/prog_tests/tailcalls.c
+++ b/tools/testing/selftests/bpf/prog_tests/tailcalls.c
@@ -37,7 +37,7 @@ static void test_tailcall_1(void)
if (CHECK_FAIL(map_fd < 0))
goto out;
- for (i = 0; i < bpf_map__def(prog_array)->max_entries; i++) {
+ for (i = 0; i < bpf_map__max_entries(prog_array); i++) {
snprintf(prog_name, sizeof(prog_name), "classifier_%d", i);
prog = bpf_object__find_program_by_name(obj, prog_name);
@@ -53,7 +53,7 @@ static void test_tailcall_1(void)
goto out;
}
- for (i = 0; i < bpf_map__def(prog_array)->max_entries; i++) {
+ for (i = 0; i < bpf_map__max_entries(prog_array); i++) {
err = bpf_prog_test_run(main_fd, 1, buff, sizeof(buff), 0,
&duration, &retval, NULL);
CHECK(err || retval != i, "tailcall",
@@ -69,7 +69,7 @@ static void test_tailcall_1(void)
CHECK(err || retval != 3, "tailcall", "err %d errno %d retval %d\n",
err, errno, retval);
- for (i = 0; i < bpf_map__def(prog_array)->max_entries; i++) {
+ for (i = 0; i < bpf_map__max_entries(prog_array); i++) {
snprintf(prog_name, sizeof(prog_name), "classifier_%d", i);
prog = bpf_object__find_program_by_name(obj, prog_name);
@@ -90,8 +90,8 @@ static void test_tailcall_1(void)
CHECK(err || retval != 0, "tailcall", "err %d errno %d retval %d\n",
err, errno, retval);
- for (i = 0; i < bpf_map__def(prog_array)->max_entries; i++) {
- j = bpf_map__def(prog_array)->max_entries - 1 - i;
+ for (i = 0; i < bpf_map__max_entries(prog_array); i++) {
+ j = bpf_map__max_entries(prog_array) - 1 - i;
snprintf(prog_name, sizeof(prog_name), "classifier_%d", j);
prog = bpf_object__find_program_by_name(obj, prog_name);
@@ -107,8 +107,8 @@ static void test_tailcall_1(void)
goto out;
}
- for (i = 0; i < bpf_map__def(prog_array)->max_entries; i++) {
- j = bpf_map__def(prog_array)->max_entries - 1 - i;
+ for (i = 0; i < bpf_map__max_entries(prog_array); i++) {
+ j = bpf_map__max_entries(prog_array) - 1 - i;
err = bpf_prog_test_run(main_fd, 1, buff, sizeof(buff), 0,
&duration, &retval, NULL);
@@ -125,7 +125,7 @@ static void test_tailcall_1(void)
CHECK(err || retval != 3, "tailcall", "err %d errno %d retval %d\n",
err, errno, retval);
- for (i = 0; i < bpf_map__def(prog_array)->max_entries; i++) {
+ for (i = 0; i < bpf_map__max_entries(prog_array); i++) {
err = bpf_map_delete_elem(map_fd, &i);
if (CHECK_FAIL(err >= 0 || errno != ENOENT))
goto out;
@@ -175,7 +175,7 @@ static void test_tailcall_2(void)
if (CHECK_FAIL(map_fd < 0))
goto out;
- for (i = 0; i < bpf_map__def(prog_array)->max_entries; i++) {
+ for (i = 0; i < bpf_map__max_entries(prog_array); i++) {
snprintf(prog_name, sizeof(prog_name), "classifier_%d", i);
prog = bpf_object__find_program_by_name(obj, prog_name);
@@ -353,7 +353,7 @@ static void test_tailcall_4(void)
if (CHECK_FAIL(map_fd < 0))
return;
- for (i = 0; i < bpf_map__def(prog_array)->max_entries; i++) {
+ for (i = 0; i < bpf_map__max_entries(prog_array); i++) {
snprintf(prog_name, sizeof(prog_name), "classifier_%d", i);
prog = bpf_object__find_program_by_name(obj, prog_name);
@@ -369,7 +369,7 @@ static void test_tailcall_4(void)
goto out;
}
- for (i = 0; i < bpf_map__def(prog_array)->max_entries; i++) {
+ for (i = 0; i < bpf_map__max_entries(prog_array); i++) {
err = bpf_map_update_elem(data_fd, &zero, &i, BPF_ANY);
if (CHECK_FAIL(err))
goto out;
@@ -380,7 +380,7 @@ static void test_tailcall_4(void)
"err %d errno %d retval %d\n", err, errno, retval);
}
- for (i = 0; i < bpf_map__def(prog_array)->max_entries; i++) {
+ for (i = 0; i < bpf_map__max_entries(prog_array); i++) {
err = bpf_map_update_elem(data_fd, &zero, &i, BPF_ANY);
if (CHECK_FAIL(err))
goto out;
@@ -441,7 +441,7 @@ static void test_tailcall_5(void)
if (CHECK_FAIL(map_fd < 0))
return;
- for (i = 0; i < bpf_map__def(prog_array)->max_entries; i++) {
+ for (i = 0; i < bpf_map__max_entries(prog_array); i++) {
snprintf(prog_name, sizeof(prog_name), "classifier_%d", i);
prog = bpf_object__find_program_by_name(obj, prog_name);
@@ -457,7 +457,7 @@ static void test_tailcall_5(void)
goto out;
}
- for (i = 0; i < bpf_map__def(prog_array)->max_entries; i++) {
+ for (i = 0; i < bpf_map__max_entries(prog_array); i++) {
err = bpf_map_update_elem(data_fd, &zero, &key[i], BPF_ANY);
if (CHECK_FAIL(err))
goto out;
@@ -468,7 +468,7 @@ static void test_tailcall_5(void)
"err %d errno %d retval %d\n", err, errno, retval);
}
- for (i = 0; i < bpf_map__def(prog_array)->max_entries; i++) {
+ for (i = 0; i < bpf_map__max_entries(prog_array); i++) {
err = bpf_map_update_elem(data_fd, &zero, &key[i], BPF_ANY);
if (CHECK_FAIL(err))
goto out;
@@ -520,7 +520,7 @@ static void test_tailcall_bpf2bpf_1(void)
goto out;
/* nop -> jmp */
- for (i = 0; i < bpf_map__def(prog_array)->max_entries; i++) {
+ for (i = 0; i < bpf_map__max_entries(prog_array); i++) {
snprintf(prog_name, sizeof(prog_name), "classifier_%d", i);
prog = bpf_object__find_program_by_name(obj, prog_name);
@@ -681,7 +681,7 @@ static void test_tailcall_bpf2bpf_3(void)
if (CHECK_FAIL(map_fd < 0))
goto out;
- for (i = 0; i < bpf_map__def(prog_array)->max_entries; i++) {
+ for (i = 0; i < bpf_map__max_entries(prog_array); i++) {
snprintf(prog_name, sizeof(prog_name), "classifier_%d", i);
prog = bpf_object__find_program_by_name(obj, prog_name);
@@ -778,7 +778,7 @@ static void test_tailcall_bpf2bpf_4(bool noise)
if (CHECK_FAIL(map_fd < 0))
goto out;
- for (i = 0; i < bpf_map__def(prog_array)->max_entries; i++) {
+ for (i = 0; i < bpf_map__max_entries(prog_array); i++) {
snprintf(prog_name, sizeof(prog_name), "classifier_%d", i);
prog = bpf_object__find_program_by_name(obj, prog_name);
--
2.30.2
^ permalink raw reply related [flat|nested] 7+ messages in thread
* [PATCH bpf-next v2 5/5] libbpf: deprecate bpf_map__def() API
2022-01-08 0:42 [PATCH bpf-next v2 0/5] libbpf 1.0: deprecate bpf_map__def() API Christy Lee
` (3 preceding siblings ...)
2022-01-08 0:42 ` [PATCH bpf-next v2 4/5] selftests/bpf: " Christy Lee
@ 2022-01-08 0:42 ` Christy Lee
2022-01-10 2:10 ` [PATCH bpf-next v2 0/5] libbpf 1.0: " patchwork-bot+netdevbpf
5 siblings, 0 replies; 7+ messages in thread
From: Christy Lee @ 2022-01-08 0:42 UTC (permalink / raw)
To: andrii, acme
Cc: christyc.y.lee, bpf, kernel-team, linux-perf-users, Christy Lee
All fields accessed via bpf_map_def can now be accessed via
appropirate getters and setters. Mark bpf_map__def() API as deprecated.
Signed-off-by: Christy Lee <christylee@fb.com>
---
tools/lib/bpf/libbpf.h | 3 ++-
1 file changed, 2 insertions(+), 1 deletion(-)
diff --git a/tools/lib/bpf/libbpf.h b/tools/lib/bpf/libbpf.h
index 8b9bc5e90c2b..9728551501ae 100644
--- a/tools/lib/bpf/libbpf.h
+++ b/tools/lib/bpf/libbpf.h
@@ -706,7 +706,8 @@ bpf_object__prev_map(const struct bpf_object *obj, const struct bpf_map *map);
LIBBPF_API int bpf_map__fd(const struct bpf_map *map);
LIBBPF_API int bpf_map__reuse_fd(struct bpf_map *map, int fd);
/* get map definition */
-LIBBPF_API const struct bpf_map_def *bpf_map__def(const struct bpf_map *map);
+LIBBPF_API LIBBPF_DEPRECATED_SINCE(0, 8, "use appropriate getters or setters instead")
+const struct bpf_map_def *bpf_map__def(const struct bpf_map *map);
/* get map name */
LIBBPF_API const char *bpf_map__name(const struct bpf_map *map);
/* get/set map type */
--
2.30.2
^ permalink raw reply related [flat|nested] 7+ messages in thread
* Re: [PATCH bpf-next v2 0/5] libbpf 1.0: deprecate bpf_map__def() API
2022-01-08 0:42 [PATCH bpf-next v2 0/5] libbpf 1.0: deprecate bpf_map__def() API Christy Lee
` (4 preceding siblings ...)
2022-01-08 0:42 ` [PATCH bpf-next v2 5/5] libbpf: deprecate " Christy Lee
@ 2022-01-10 2:10 ` patchwork-bot+netdevbpf
5 siblings, 0 replies; 7+ messages in thread
From: patchwork-bot+netdevbpf @ 2022-01-10 2:10 UTC (permalink / raw)
To: Christy Lee
Cc: andrii, acme, christyc.y.lee, bpf, kernel-team, linux-perf-users
Hello:
This series was applied to bpf/bpf-next.git (master)
by Andrii Nakryiko <andrii@kernel.org>:
On Fri, 7 Jan 2022 16:42:13 -0800 you wrote:
> bpf_map__def() is rarely used and non-extensible. bpf_map_def fields
> can be accessed with appropriate map getters and setters instead.
> Deprecate bpf_map__def() API and replace use cases with getters and
> setters.
>
> Changelog:
>
> [...]
Here is the summary with links:
- [bpf-next,v2,1/5] samples/bpf: stop using bpf_map__def() API
https://git.kernel.org/bpf/bpf-next/c/fcaf77640892
- [bpf-next,v2,2/5] bpftool: stop using bpf_map__def() API
https://git.kernel.org/bpf/bpf-next/c/567df3b14606
- [bpf-next,v2,3/5] perf: stop using bpf_map__def() API
https://git.kernel.org/bpf/bpf-next/c/ad60db9aabb1
- [bpf-next,v2,4/5] selftests/bpf: stop using bpf_map__def() API
https://git.kernel.org/bpf/bpf-next/c/f8b287530aef
- [bpf-next,v2,5/5] libbpf: deprecate bpf_map__def() API
https://git.kernel.org/bpf/bpf-next/c/b95533311a0f
You are awesome, thank you!
--
Deet-doot-dot, I am a bot.
https://korg.docs.kernel.org/patchwork/pwbot.html
^ permalink raw reply [flat|nested] 7+ messages in thread
end of thread, other threads:[~2022-01-10 2:10 UTC | newest]
Thread overview: 7+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2022-01-08 0:42 [PATCH bpf-next v2 0/5] libbpf 1.0: deprecate bpf_map__def() API Christy Lee
2022-01-08 0:42 ` [PATCH bpf-next v2 1/5] samples/bpf: stop using " Christy Lee
2022-01-08 0:42 ` [PATCH bpf-next v2 2/5] bpftool: " Christy Lee
2022-01-08 0:42 ` [PATCH bpf-next v2 3/5] perf: " Christy Lee
2022-01-08 0:42 ` [PATCH bpf-next v2 4/5] selftests/bpf: " Christy Lee
2022-01-08 0:42 ` [PATCH bpf-next v2 5/5] libbpf: deprecate " Christy Lee
2022-01-10 2:10 ` [PATCH bpf-next v2 0/5] libbpf 1.0: " patchwork-bot+netdevbpf
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.