* [PATCH bpf-next 0/7] libbpf: deprecate some setter and getter APIs
@ 2022-01-21 0:41 Andrii Nakryiko
2022-01-21 0:41 ` [PATCH bpf-next 1/7] libbpf: hide and discourage inconsistently named getters Andrii Nakryiko
` (6 more replies)
0 siblings, 7 replies; 9+ messages in thread
From: Andrii Nakryiko @ 2022-01-21 0:41 UTC (permalink / raw)
To: bpf, ast, daniel; +Cc: andrii, kernel-team
Another batch of simple deprecations. One of the last few, hopefully, as we
are getting close to deprecating all the planned APIs/features. See
individual patches for details.
Andrii Nakryiko (7):
libbpf: hide and discourage inconsistently named getters
libbpf: deprecate bpf_map__resize()
libbpf: deprecate bpf_program__is_<type>() and
bpf_program__set_<type>() APIs
bpftool: use preferred setters/getters instead of deprecated ones
selftests/bpf: use preferred setter/getter APIs instead of deprecated
ones
samples/bpf: use preferred getters/setters instead of deprecated ones
perf: use generic bpf_program__set_type() to set BPF prog type
samples/bpf/map_perf_test_user.c | 2 +-
samples/bpf/xdp_redirect_cpu_user.c | 2 +-
samples/bpf/xdp_sample_user.c | 2 +-
samples/bpf/xdp_sample_user.h | 2 +-
tools/bpf/bpftool/gen.c | 2 +-
tools/bpf/bpftool/prog.c | 8 ++---
tools/lib/bpf/btf.h | 5 +--
tools/lib/bpf/libbpf.c | 21 ++++++------
tools/lib/bpf/libbpf.h | 32 +++++++++++++++++--
tools/lib/bpf/libbpf.map | 2 ++
tools/lib/bpf/libbpf_internal.h | 3 ++
tools/lib/bpf/libbpf_legacy.h | 17 ++++++++++
tools/perf/util/bpf-loader.c | 4 +--
.../selftests/bpf/benchs/bench_ringbufs.c | 2 +-
.../selftests/bpf/prog_tests/fexit_bpf2bpf.c | 2 +-
.../prog_tests/get_stackid_cannot_attach.c | 2 +-
.../bpf/prog_tests/stacktrace_build_id_nmi.c | 2 +-
17 files changed, 79 insertions(+), 31 deletions(-)
--
2.30.2
^ permalink raw reply [flat|nested] 9+ messages in thread
* [PATCH bpf-next 1/7] libbpf: hide and discourage inconsistently named getters
2022-01-21 0:41 [PATCH bpf-next 0/7] libbpf: deprecate some setter and getter APIs Andrii Nakryiko
@ 2022-01-21 0:41 ` Andrii Nakryiko
2022-01-21 0:41 ` [PATCH bpf-next 2/7] libbpf: deprecate bpf_map__resize() Andrii Nakryiko
` (5 subsequent siblings)
6 siblings, 0 replies; 9+ messages in thread
From: Andrii Nakryiko @ 2022-01-21 0:41 UTC (permalink / raw)
To: bpf, ast, daniel; +Cc: andrii, kernel-team
Move a bunch of "getters" into libbpf_legacy.h to keep them there in
libbpf 1.0. See [0] for discussion of "Discouraged APIs". These getters
don't add any maintenance burden and are simple alias, but they are
inconsistent in naming. So keep them in libbpf_legacy.h instead of
libbpf.h to "hide" them in favor of preferred getters ([1]). Also add two
missing getters: bpf_program__type() and bpf_program__expected_attach_type().
[0] https://github.com/libbpf/libbpf/wiki/Libbpf:-the-road-to-v1.0#handling-deprecation-of-apis-and-functionality
[1] Closes: https://github.com/libbpf/libbpf/issues/307
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
---
tools/lib/bpf/btf.h | 5 +----
tools/lib/bpf/libbpf.c | 21 ++++++++++++---------
tools/lib/bpf/libbpf.h | 5 ++---
tools/lib/bpf/libbpf.map | 2 ++
tools/lib/bpf/libbpf_internal.h | 3 +++
tools/lib/bpf/libbpf_legacy.h | 17 +++++++++++++++++
6 files changed, 37 insertions(+), 16 deletions(-)
diff --git a/tools/lib/bpf/btf.h b/tools/lib/bpf/btf.h
index 51862fdee850..96b44d55db6e 100644
--- a/tools/lib/bpf/btf.h
+++ b/tools/lib/bpf/btf.h
@@ -147,8 +147,6 @@ LIBBPF_API int btf__resolve_type(const struct btf *btf, __u32 type_id);
LIBBPF_API int btf__align_of(const struct btf *btf, __u32 id);
LIBBPF_API int btf__fd(const struct btf *btf);
LIBBPF_API void btf__set_fd(struct btf *btf, int fd);
-LIBBPF_DEPRECATED_SINCE(0, 7, "use btf__raw_data() instead")
-LIBBPF_API const void *btf__get_raw_data(const struct btf *btf, __u32 *size);
LIBBPF_API const void *btf__raw_data(const struct btf *btf, __u32 *size);
LIBBPF_API const char *btf__name_by_offset(const struct btf *btf, __u32 offset);
LIBBPF_API const char *btf__str_by_offset(const struct btf *btf, __u32 offset);
@@ -159,8 +157,7 @@ LIBBPF_API int btf__get_map_kv_tids(const struct btf *btf, const char *map_name,
LIBBPF_API struct btf_ext *btf_ext__new(const __u8 *data, __u32 size);
LIBBPF_API void btf_ext__free(struct btf_ext *btf_ext);
-LIBBPF_API const void *btf_ext__get_raw_data(const struct btf_ext *btf_ext,
- __u32 *size);
+LIBBPF_API const void *btf_ext__raw_data(const struct btf_ext *btf_ext, __u32 *size);
LIBBPF_API LIBBPF_DEPRECATED("btf_ext__reloc_func_info was never meant as a public API and has wrong assumptions embedded in it; it will be removed in the future libbpf versions")
int btf_ext__reloc_func_info(const struct btf *btf,
const struct btf_ext *btf_ext,
diff --git a/tools/lib/bpf/libbpf.c b/tools/lib/bpf/libbpf.c
index fdb3536afa7d..4f35fcaf7eea 100644
--- a/tools/lib/bpf/libbpf.c
+++ b/tools/lib/bpf/libbpf.c
@@ -7883,10 +7883,8 @@ int bpf_map__set_pin_path(struct bpf_map *map, const char *path)
return 0;
}
-const char *bpf_map__get_pin_path(const struct bpf_map *map)
-{
- return map->pin_path;
-}
+__alias(bpf_map__pin_path)
+const char *bpf_map__get_pin_path(const struct bpf_map *map);
const char *bpf_map__pin_path(const struct bpf_map *map)
{
@@ -8451,7 +8449,10 @@ static int bpf_program_nth_fd(const struct bpf_program *prog, int n)
return fd;
}
-enum bpf_prog_type bpf_program__get_type(const struct bpf_program *prog)
+__alias(bpf_program__type)
+enum bpf_prog_type bpf_program__get_type(const struct bpf_program *prog);
+
+enum bpf_prog_type bpf_program__type(const struct bpf_program *prog)
{
return prog->type;
}
@@ -8495,8 +8496,10 @@ BPF_PROG_TYPE_FNS(struct_ops, BPF_PROG_TYPE_STRUCT_OPS);
BPF_PROG_TYPE_FNS(extension, BPF_PROG_TYPE_EXT);
BPF_PROG_TYPE_FNS(sk_lookup, BPF_PROG_TYPE_SK_LOOKUP);
-enum bpf_attach_type
-bpf_program__get_expected_attach_type(const struct bpf_program *prog)
+__alias(bpf_program__expected_attach_type)
+enum bpf_attach_type bpf_program__get_expected_attach_type(const struct bpf_program *prog);
+
+enum bpf_attach_type bpf_program__expected_attach_type(const struct bpf_program *prog)
{
return prog->expected_attach_type;
}
@@ -9460,7 +9463,7 @@ static int bpf_prog_load_xattr2(const struct bpf_prog_load_attr *attr,
bpf_program__set_expected_attach_type(prog,
attach_type);
}
- if (bpf_program__get_type(prog) == BPF_PROG_TYPE_UNSPEC) {
+ if (bpf_program__type(prog) == BPF_PROG_TYPE_UNSPEC) {
/*
* we haven't guessed from section name and user
* didn't provide a fallback type, too bad...
@@ -10511,7 +10514,7 @@ bpf_program__attach_fd(const struct bpf_program *prog, int target_fd, int btf_id
return libbpf_err_ptr(-ENOMEM);
link->detach = &bpf_link__detach_fd;
- attach_type = bpf_program__get_expected_attach_type(prog);
+ attach_type = bpf_program__expected_attach_type(prog);
link_fd = bpf_link_create(prog_fd, target_fd, attach_type, &opts);
if (link_fd < 0) {
link_fd = -errno;
diff --git a/tools/lib/bpf/libbpf.h b/tools/lib/bpf/libbpf.h
index 9728551501ae..dbf37c0fa531 100644
--- a/tools/lib/bpf/libbpf.h
+++ b/tools/lib/bpf/libbpf.h
@@ -605,12 +605,12 @@ LIBBPF_API int bpf_program__set_struct_ops(struct bpf_program *prog);
LIBBPF_API int bpf_program__set_extension(struct bpf_program *prog);
LIBBPF_API int bpf_program__set_sk_lookup(struct bpf_program *prog);
-LIBBPF_API enum bpf_prog_type bpf_program__get_type(const struct bpf_program *prog);
+LIBBPF_API enum bpf_prog_type bpf_program__type(const struct bpf_program *prog);
LIBBPF_API void bpf_program__set_type(struct bpf_program *prog,
enum bpf_prog_type type);
LIBBPF_API enum bpf_attach_type
-bpf_program__get_expected_attach_type(const struct bpf_program *prog);
+bpf_program__expected_attach_type(const struct bpf_program *prog);
LIBBPF_API void
bpf_program__set_expected_attach_type(struct bpf_program *prog,
enum bpf_attach_type type);
@@ -758,7 +758,6 @@ LIBBPF_API bool bpf_map__is_offload_neutral(const struct bpf_map *map);
*/
LIBBPF_API bool bpf_map__is_internal(const struct bpf_map *map);
LIBBPF_API int bpf_map__set_pin_path(struct bpf_map *map, const char *path);
-LIBBPF_API const char *bpf_map__get_pin_path(const struct bpf_map *map);
LIBBPF_API const char *bpf_map__pin_path(const struct bpf_map *map);
LIBBPF_API bool bpf_map__is_pinned(const struct bpf_map *map);
LIBBPF_API int bpf_map__pin(struct bpf_map *map, const char *path);
diff --git a/tools/lib/bpf/libbpf.map b/tools/lib/bpf/libbpf.map
index 8262cfca2240..e33432c5097c 100644
--- a/tools/lib/bpf/libbpf.map
+++ b/tools/lib/bpf/libbpf.map
@@ -424,10 +424,12 @@ LIBBPF_0.6.0 {
LIBBPF_0.7.0 {
global:
bpf_btf_load;
+ bpf_program__expected_attach_type;
bpf_program__log_buf;
bpf_program__log_level;
bpf_program__set_log_buf;
bpf_program__set_log_level;
+ bpf_program__type;
libbpf_probe_bpf_helper;
libbpf_probe_bpf_map_type;
libbpf_probe_bpf_prog_type;
diff --git a/tools/lib/bpf/libbpf_internal.h b/tools/lib/bpf/libbpf_internal.h
index 1565679eb432..bc86b82e90d1 100644
--- a/tools/lib/bpf/libbpf_internal.h
+++ b/tools/lib/bpf/libbpf_internal.h
@@ -92,6 +92,9 @@
# define offsetofend(TYPE, FIELD) \
(offsetof(TYPE, FIELD) + sizeof(((TYPE *)0)->FIELD))
#endif
+#ifndef __alias
+#define __alias(symbol) __attribute__((alias(#symbol)))
+#endif
/* Check whether a string `str` has prefix `pfx`, regardless if `pfx` is
* a string literal known at compilation time or char * pointer known only at
diff --git a/tools/lib/bpf/libbpf_legacy.h b/tools/lib/bpf/libbpf_legacy.h
index 79131f761a27..116d54d2be7a 100644
--- a/tools/lib/bpf/libbpf_legacy.h
+++ b/tools/lib/bpf/libbpf_legacy.h
@@ -81,6 +81,23 @@ LIBBPF_API int libbpf_set_strict_mode(enum libbpf_strict_mode mode);
#define DECLARE_LIBBPF_OPTS LIBBPF_OPTS
+/* "Discouraged" APIs which don't follow consistent libbpf naming patterns.
+ * They are normally a trivial aliases or wrappers for proper APIs and are
+ * left to minimize unnecessary disruption for users of libbpf. But they
+ * shouldn't be used going forward.
+ */
+
+struct bpf_program;
+struct bpf_map;
+struct btf;
+struct btf_ext;
+
+LIBBPF_API enum bpf_prog_type bpf_program__get_type(const struct bpf_program *prog);
+LIBBPF_API enum bpf_attach_type bpf_program__get_expected_attach_type(const struct bpf_program *prog);
+LIBBPF_API const char *bpf_map__get_pin_path(const struct bpf_map *map);
+LIBBPF_API const void *btf__get_raw_data(const struct btf *btf, __u32 *size);
+LIBBPF_API const void *btf_ext__get_raw_data(const struct btf_ext *btf_ext, __u32 *size);
+
#ifdef __cplusplus
} /* extern "C" */
#endif
--
2.30.2
^ permalink raw reply related [flat|nested] 9+ messages in thread
* [PATCH bpf-next 2/7] libbpf: deprecate bpf_map__resize()
2022-01-21 0:41 [PATCH bpf-next 0/7] libbpf: deprecate some setter and getter APIs Andrii Nakryiko
2022-01-21 0:41 ` [PATCH bpf-next 1/7] libbpf: hide and discourage inconsistently named getters Andrii Nakryiko
@ 2022-01-21 0:41 ` Andrii Nakryiko
2022-01-21 0:43 ` Andrii Nakryiko
2022-01-21 0:41 ` [PATCH bpf-next 3/7] libbpf: deprecate bpf_program__is_<type>() and bpf_program__set_<type>() APIs Andrii Nakryiko
` (4 subsequent siblings)
6 siblings, 1 reply; 9+ messages in thread
From: Andrii Nakryiko @ 2022-01-21 0:41 UTC (permalink / raw)
To: bpf, ast, daniel; +Cc: andrii, kernel-team
Deprecated bpf_map__resize() in favor of bpf_map__set_max_entries()
setter. In addition to having a surprising name (users often don't
realize that they need to use bpf_map__resize()), the name also implies
some magic way of resizing BPF map after it is created, which is clearly
not the case.
Another minor annoyance is that bpf_map__resize() disallows 0 value for
max_entries, which in some cases is totally acceptable (e.g., like for
BPF perf buf case to let libbpf auto-create one buffer per each
available CPU core).
[0] https://github.com/libbpf/libbpf/issues/304
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
---
tools/lib/bpf/libbpf.h | 1 +
1 file changed, 1 insertion(+)
diff --git a/tools/lib/bpf/libbpf.h b/tools/lib/bpf/libbpf.h
index dbf37c0fa531..6f8e6b3cff84 100644
--- a/tools/lib/bpf/libbpf.h
+++ b/tools/lib/bpf/libbpf.h
@@ -716,6 +716,7 @@ LIBBPF_API int bpf_map__set_type(struct bpf_map *map, enum bpf_map_type type);
/* get/set map size (max_entries) */
LIBBPF_API __u32 bpf_map__max_entries(const struct bpf_map *map);
LIBBPF_API int bpf_map__set_max_entries(struct bpf_map *map, __u32 max_entries);
+LIBBPF_DEPRECATED_SINCE(0, 8, "use bpf_map__set_max_entries() instead")
LIBBPF_API int bpf_map__resize(struct bpf_map *map, __u32 max_entries);
/* get/set map flags */
LIBBPF_API __u32 bpf_map__map_flags(const struct bpf_map *map);
--
2.30.2
^ permalink raw reply related [flat|nested] 9+ messages in thread
* [PATCH bpf-next 3/7] libbpf: deprecate bpf_program__is_<type>() and bpf_program__set_<type>() APIs
2022-01-21 0:41 [PATCH bpf-next 0/7] libbpf: deprecate some setter and getter APIs Andrii Nakryiko
2022-01-21 0:41 ` [PATCH bpf-next 1/7] libbpf: hide and discourage inconsistently named getters Andrii Nakryiko
2022-01-21 0:41 ` [PATCH bpf-next 2/7] libbpf: deprecate bpf_map__resize() Andrii Nakryiko
@ 2022-01-21 0:41 ` Andrii Nakryiko
2022-01-21 0:41 ` [PATCH bpf-next 4/7] bpftool: use preferred setters/getters instead of deprecated ones Andrii Nakryiko
` (3 subsequent siblings)
6 siblings, 0 replies; 9+ messages in thread
From: Andrii Nakryiko @ 2022-01-21 0:41 UTC (permalink / raw)
To: bpf, ast, daniel; +Cc: andrii, kernel-team
Not sure why these APIs were added in the first place instead of
a completely generic (and not requiring constantly adding new APIs with
each new BPF program type) bpf_program__type() and
bpf_program__set_type() APIs. But as it is right now, there are 13 such
specialized is_type/set_type APIs, while latest kernel is already at 30+
BPF program types.
Instead of completing the set of APIs and keep chasing kernel's
bpf_prog_type enum, deprecate existing subset and recommend generic
bpf_program__type() and bpf_program__set_type() APIs.
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
---
tools/lib/bpf/libbpf.h | 26 ++++++++++++++++++++++++++
1 file changed, 26 insertions(+)
diff --git a/tools/lib/bpf/libbpf.h b/tools/lib/bpf/libbpf.h
index 6f8e6b3cff84..572b3b967c0d 100644
--- a/tools/lib/bpf/libbpf.h
+++ b/tools/lib/bpf/libbpf.h
@@ -591,18 +591,31 @@ LIBBPF_API int bpf_program__nth_fd(const struct bpf_program *prog, int n);
/*
* Adjust type of BPF program. Default is kprobe.
*/
+LIBBPF_DEPRECATED_SINCE(0, 8, "use bpf_program__set_type() instead")
LIBBPF_API int bpf_program__set_socket_filter(struct bpf_program *prog);
+LIBBPF_DEPRECATED_SINCE(0, 8, "use bpf_program__set_type() instead")
LIBBPF_API int bpf_program__set_tracepoint(struct bpf_program *prog);
+LIBBPF_DEPRECATED_SINCE(0, 8, "use bpf_program__set_type() instead")
LIBBPF_API int bpf_program__set_raw_tracepoint(struct bpf_program *prog);
+LIBBPF_DEPRECATED_SINCE(0, 8, "use bpf_program__set_type() instead")
LIBBPF_API int bpf_program__set_kprobe(struct bpf_program *prog);
+LIBBPF_DEPRECATED_SINCE(0, 8, "use bpf_program__set_type() instead")
LIBBPF_API int bpf_program__set_lsm(struct bpf_program *prog);
+LIBBPF_DEPRECATED_SINCE(0, 8, "use bpf_program__set_type() instead")
LIBBPF_API int bpf_program__set_sched_cls(struct bpf_program *prog);
+LIBBPF_DEPRECATED_SINCE(0, 8, "use bpf_program__set_type() instead")
LIBBPF_API int bpf_program__set_sched_act(struct bpf_program *prog);
+LIBBPF_DEPRECATED_SINCE(0, 8, "use bpf_program__set_type() instead")
LIBBPF_API int bpf_program__set_xdp(struct bpf_program *prog);
+LIBBPF_DEPRECATED_SINCE(0, 8, "use bpf_program__set_type() instead")
LIBBPF_API int bpf_program__set_perf_event(struct bpf_program *prog);
+LIBBPF_DEPRECATED_SINCE(0, 8, "use bpf_program__set_type() instead")
LIBBPF_API int bpf_program__set_tracing(struct bpf_program *prog);
+LIBBPF_DEPRECATED_SINCE(0, 8, "use bpf_program__set_type() instead")
LIBBPF_API int bpf_program__set_struct_ops(struct bpf_program *prog);
+LIBBPF_DEPRECATED_SINCE(0, 8, "use bpf_program__set_type() instead")
LIBBPF_API int bpf_program__set_extension(struct bpf_program *prog);
+LIBBPF_DEPRECATED_SINCE(0, 8, "use bpf_program__set_type() instead")
LIBBPF_API int bpf_program__set_sk_lookup(struct bpf_program *prog);
LIBBPF_API enum bpf_prog_type bpf_program__type(const struct bpf_program *prog);
@@ -631,18 +644,31 @@ LIBBPF_API int
bpf_program__set_attach_target(struct bpf_program *prog, int attach_prog_fd,
const char *attach_func_name);
+LIBBPF_DEPRECATED_SINCE(0, 8, "use bpf_program__type() instead")
LIBBPF_API bool bpf_program__is_socket_filter(const struct bpf_program *prog);
+LIBBPF_DEPRECATED_SINCE(0, 8, "use bpf_program__type() instead")
LIBBPF_API bool bpf_program__is_tracepoint(const struct bpf_program *prog);
+LIBBPF_DEPRECATED_SINCE(0, 8, "use bpf_program__type() instead")
LIBBPF_API bool bpf_program__is_raw_tracepoint(const struct bpf_program *prog);
+LIBBPF_DEPRECATED_SINCE(0, 8, "use bpf_program__type() instead")
LIBBPF_API bool bpf_program__is_kprobe(const struct bpf_program *prog);
+LIBBPF_DEPRECATED_SINCE(0, 8, "use bpf_program__type() instead")
LIBBPF_API bool bpf_program__is_lsm(const struct bpf_program *prog);
+LIBBPF_DEPRECATED_SINCE(0, 8, "use bpf_program__type() instead")
LIBBPF_API bool bpf_program__is_sched_cls(const struct bpf_program *prog);
+LIBBPF_DEPRECATED_SINCE(0, 8, "use bpf_program__type() instead")
LIBBPF_API bool bpf_program__is_sched_act(const struct bpf_program *prog);
+LIBBPF_DEPRECATED_SINCE(0, 8, "use bpf_program__type() instead")
LIBBPF_API bool bpf_program__is_xdp(const struct bpf_program *prog);
+LIBBPF_DEPRECATED_SINCE(0, 8, "use bpf_program__type() instead")
LIBBPF_API bool bpf_program__is_perf_event(const struct bpf_program *prog);
+LIBBPF_DEPRECATED_SINCE(0, 8, "use bpf_program__type() instead")
LIBBPF_API bool bpf_program__is_tracing(const struct bpf_program *prog);
+LIBBPF_DEPRECATED_SINCE(0, 8, "use bpf_program__type() instead")
LIBBPF_API bool bpf_program__is_struct_ops(const struct bpf_program *prog);
+LIBBPF_DEPRECATED_SINCE(0, 8, "use bpf_program__type() instead")
LIBBPF_API bool bpf_program__is_extension(const struct bpf_program *prog);
+LIBBPF_DEPRECATED_SINCE(0, 8, "use bpf_program__type() instead")
LIBBPF_API bool bpf_program__is_sk_lookup(const struct bpf_program *prog);
/*
--
2.30.2
^ permalink raw reply related [flat|nested] 9+ messages in thread
* [PATCH bpf-next 4/7] bpftool: use preferred setters/getters instead of deprecated ones
2022-01-21 0:41 [PATCH bpf-next 0/7] libbpf: deprecate some setter and getter APIs Andrii Nakryiko
` (2 preceding siblings ...)
2022-01-21 0:41 ` [PATCH bpf-next 3/7] libbpf: deprecate bpf_program__is_<type>() and bpf_program__set_<type>() APIs Andrii Nakryiko
@ 2022-01-21 0:41 ` Andrii Nakryiko
2022-01-21 0:41 ` [PATCH bpf-next 5/7] selftests/bpf: use preferred setter/getter APIs " Andrii Nakryiko
` (2 subsequent siblings)
6 siblings, 0 replies; 9+ messages in thread
From: Andrii Nakryiko @ 2022-01-21 0:41 UTC (permalink / raw)
To: bpf, ast, daniel; +Cc: andrii, kernel-team
Use bpf_program__type() instead of discouraged bpf_program__get_type().
Also switch to bpf_map__set_max_entries() instead of bpf_map__resize().
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
---
tools/bpf/bpftool/gen.c | 2 +-
tools/bpf/bpftool/prog.c | 8 ++++----
2 files changed, 5 insertions(+), 5 deletions(-)
diff --git a/tools/bpf/bpftool/gen.c b/tools/bpf/bpftool/gen.c
index 43e3f8700ecc..8f78c27d41f0 100644
--- a/tools/bpf/bpftool/gen.c
+++ b/tools/bpf/bpftool/gen.c
@@ -378,7 +378,7 @@ static void codegen_attach_detach(struct bpf_object *obj, const char *obj_name)
int prog_fd = skel->progs.%2$s.prog_fd; \n\
", obj_name, bpf_program__name(prog));
- switch (bpf_program__get_type(prog)) {
+ switch (bpf_program__type(prog)) {
case BPF_PROG_TYPE_RAW_TRACEPOINT:
tp_name = strchr(bpf_program__section_name(prog), '/') + 1;
printf("\tint fd = bpf_raw_tracepoint_open(\"%s\", prog_fd);\n", tp_name);
diff --git a/tools/bpf/bpftool/prog.c b/tools/bpf/bpftool/prog.c
index cf935c63e6f5..87593f98d2d1 100644
--- a/tools/bpf/bpftool/prog.c
+++ b/tools/bpf/bpftool/prog.c
@@ -2283,10 +2283,10 @@ static int do_profile(int argc, char **argv)
profile_obj->rodata->num_metric = num_metric;
/* adjust map sizes */
- bpf_map__resize(profile_obj->maps.events, num_metric * num_cpu);
- bpf_map__resize(profile_obj->maps.fentry_readings, num_metric);
- bpf_map__resize(profile_obj->maps.accum_readings, num_metric);
- bpf_map__resize(profile_obj->maps.counts, 1);
+ bpf_map__set_max_entries(profile_obj->maps.events, num_metric * num_cpu);
+ bpf_map__set_max_entries(profile_obj->maps.fentry_readings, num_metric);
+ bpf_map__set_max_entries(profile_obj->maps.accum_readings, num_metric);
+ bpf_map__set_max_entries(profile_obj->maps.counts, 1);
/* change target name */
profile_tgt_name = profile_target_name(profile_tgt_fd);
--
2.30.2
^ permalink raw reply related [flat|nested] 9+ messages in thread
* [PATCH bpf-next 5/7] selftests/bpf: use preferred setter/getter APIs instead of deprecated ones
2022-01-21 0:41 [PATCH bpf-next 0/7] libbpf: deprecate some setter and getter APIs Andrii Nakryiko
` (3 preceding siblings ...)
2022-01-21 0:41 ` [PATCH bpf-next 4/7] bpftool: use preferred setters/getters instead of deprecated ones Andrii Nakryiko
@ 2022-01-21 0:41 ` Andrii Nakryiko
2022-01-21 0:41 ` [PATCH bpf-next 6/7] samples/bpf: use preferred getters/setters " Andrii Nakryiko
2022-01-21 0:41 ` [PATCH bpf-next 7/7] perf: use generic bpf_program__set_type() to set BPF prog type Andrii Nakryiko
6 siblings, 0 replies; 9+ messages in thread
From: Andrii Nakryiko @ 2022-01-21 0:41 UTC (permalink / raw)
To: bpf, ast, daniel; +Cc: andrii, kernel-team
Switch to using preferred setters and getters instead of deprecated ones.
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
---
tools/testing/selftests/bpf/benchs/bench_ringbufs.c | 2 +-
tools/testing/selftests/bpf/prog_tests/fexit_bpf2bpf.c | 2 +-
.../selftests/bpf/prog_tests/get_stackid_cannot_attach.c | 2 +-
.../testing/selftests/bpf/prog_tests/stacktrace_build_id_nmi.c | 2 +-
4 files changed, 4 insertions(+), 4 deletions(-)
diff --git a/tools/testing/selftests/bpf/benchs/bench_ringbufs.c b/tools/testing/selftests/bpf/benchs/bench_ringbufs.c
index da8593b3494a..c2554f9695ff 100644
--- a/tools/testing/selftests/bpf/benchs/bench_ringbufs.c
+++ b/tools/testing/selftests/bpf/benchs/bench_ringbufs.c
@@ -151,7 +151,7 @@ static struct ringbuf_bench *ringbuf_setup_skeleton(void)
/* record data + header take 16 bytes */
skel->rodata->wakeup_data_size = args.sample_rate * 16;
- bpf_map__resize(skel->maps.ringbuf, args.ringbuf_sz);
+ bpf_map__set_max_entries(skel->maps.ringbuf, args.ringbuf_sz);
if (ringbuf_bench__load(skel)) {
fprintf(stderr, "failed to load skeleton\n");
diff --git a/tools/testing/selftests/bpf/prog_tests/fexit_bpf2bpf.c b/tools/testing/selftests/bpf/prog_tests/fexit_bpf2bpf.c
index c52f99f6a909..e83575e5480f 100644
--- a/tools/testing/selftests/bpf/prog_tests/fexit_bpf2bpf.c
+++ b/tools/testing/selftests/bpf/prog_tests/fexit_bpf2bpf.c
@@ -132,7 +132,7 @@ static void test_fexit_bpf2bpf_common(const char *obj_file,
&link_info, &info_len);
ASSERT_OK(err, "link_fd_get_info");
ASSERT_EQ(link_info.tracing.attach_type,
- bpf_program__get_expected_attach_type(prog[i]),
+ bpf_program__expected_attach_type(prog[i]),
"link_attach_type");
ASSERT_EQ(link_info.tracing.target_obj_id, tgt_prog_id, "link_tgt_obj_id");
ASSERT_EQ(link_info.tracing.target_btf_id, btf_id, "link_tgt_btf_id");
diff --git a/tools/testing/selftests/bpf/prog_tests/get_stackid_cannot_attach.c b/tools/testing/selftests/bpf/prog_tests/get_stackid_cannot_attach.c
index 8d5a6023a1bb..5308de1ed478 100644
--- a/tools/testing/selftests/bpf/prog_tests/get_stackid_cannot_attach.c
+++ b/tools/testing/selftests/bpf/prog_tests/get_stackid_cannot_attach.c
@@ -27,7 +27,7 @@ void test_get_stackid_cannot_attach(void)
return;
/* override program type */
- bpf_program__set_perf_event(skel->progs.oncpu);
+ bpf_program__set_type(skel->progs.oncpu, BPF_PROG_TYPE_PERF_EVENT);
err = test_stacktrace_build_id__load(skel);
if (CHECK(err, "skel_load", "skeleton load failed: %d\n", err))
diff --git a/tools/testing/selftests/bpf/prog_tests/stacktrace_build_id_nmi.c b/tools/testing/selftests/bpf/prog_tests/stacktrace_build_id_nmi.c
index 0a91d8d9954b..f45a1d7b0a28 100644
--- a/tools/testing/selftests/bpf/prog_tests/stacktrace_build_id_nmi.c
+++ b/tools/testing/selftests/bpf/prog_tests/stacktrace_build_id_nmi.c
@@ -42,7 +42,7 @@ void test_stacktrace_build_id_nmi(void)
return;
/* override program type */
- bpf_program__set_perf_event(skel->progs.oncpu);
+ bpf_program__set_type(skel->progs.oncpu, BPF_PROG_TYPE_PERF_EVENT);
err = test_stacktrace_build_id__load(skel);
if (CHECK(err, "skel_load", "skeleton load failed: %d\n", err))
--
2.30.2
^ permalink raw reply related [flat|nested] 9+ messages in thread
* [PATCH bpf-next 6/7] samples/bpf: use preferred getters/setters instead of deprecated ones
2022-01-21 0:41 [PATCH bpf-next 0/7] libbpf: deprecate some setter and getter APIs Andrii Nakryiko
` (4 preceding siblings ...)
2022-01-21 0:41 ` [PATCH bpf-next 5/7] selftests/bpf: use preferred setter/getter APIs " Andrii Nakryiko
@ 2022-01-21 0:41 ` Andrii Nakryiko
2022-01-21 0:41 ` [PATCH bpf-next 7/7] perf: use generic bpf_program__set_type() to set BPF prog type Andrii Nakryiko
6 siblings, 0 replies; 9+ messages in thread
From: Andrii Nakryiko @ 2022-01-21 0:41 UTC (permalink / raw)
To: bpf, ast, daniel; +Cc: andrii, kernel-team
Use preferred setter and getter APIs instead of deprecated ones.
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
---
samples/bpf/map_perf_test_user.c | 2 +-
samples/bpf/xdp_redirect_cpu_user.c | 2 +-
samples/bpf/xdp_sample_user.c | 2 +-
samples/bpf/xdp_sample_user.h | 2 +-
4 files changed, 4 insertions(+), 4 deletions(-)
diff --git a/samples/bpf/map_perf_test_user.c b/samples/bpf/map_perf_test_user.c
index 319fd31522f3..e69651a6902f 100644
--- a/samples/bpf/map_perf_test_user.c
+++ b/samples/bpf/map_perf_test_user.c
@@ -413,7 +413,7 @@ static void fixup_map(struct bpf_object *obj)
for (i = 0; i < NR_TESTS; i++) {
if (!strcmp(test_map_names[i], name) &&
(check_test_flags(i))) {
- bpf_map__resize(map, num_map_entries);
+ bpf_map__set_max_entries(map, num_map_entries);
continue;
}
}
diff --git a/samples/bpf/xdp_redirect_cpu_user.c b/samples/bpf/xdp_redirect_cpu_user.c
index a81704d3317b..5f74a70a9021 100644
--- a/samples/bpf/xdp_redirect_cpu_user.c
+++ b/samples/bpf/xdp_redirect_cpu_user.c
@@ -70,7 +70,7 @@ static void print_avail_progs(struct bpf_object *obj)
printf(" Programs to be used for -p/--progname:\n");
bpf_object__for_each_program(pos, obj) {
- if (bpf_program__is_xdp(pos)) {
+ if (bpf_program__type(pos) == BPF_PROG_TYPE_XDP) {
if (!strncmp(bpf_program__name(pos), "xdp_prognum",
sizeof("xdp_prognum") - 1))
printf(" %s\n", bpf_program__name(pos));
diff --git a/samples/bpf/xdp_sample_user.c b/samples/bpf/xdp_sample_user.c
index 8740838e7767..44598df4c84b 100644
--- a/samples/bpf/xdp_sample_user.c
+++ b/samples/bpf/xdp_sample_user.c
@@ -1218,7 +1218,7 @@ int sample_setup_maps(struct bpf_map **maps)
default:
return -EINVAL;
}
- if (bpf_map__resize(sample_map[i], sample_map_count[i]) < 0)
+ if (bpf_map__set_max_entries(sample_map[i], sample_map_count[i]) < 0)
return -errno;
}
sample_map[MAP_DEVMAP_XMIT_MULTI] = maps[MAP_DEVMAP_XMIT_MULTI];
diff --git a/samples/bpf/xdp_sample_user.h b/samples/bpf/xdp_sample_user.h
index 5f44b877ecf5..f45051679977 100644
--- a/samples/bpf/xdp_sample_user.h
+++ b/samples/bpf/xdp_sample_user.h
@@ -61,7 +61,7 @@ static inline char *safe_strncpy(char *dst, const char *src, size_t size)
#define __attach_tp(name) \
({ \
- if (!bpf_program__is_tracing(skel->progs.name)) \
+ if (bpf_program__type(skel->progs.name) != BPF_PROG_TYPE_TRACING)\
return -EINVAL; \
skel->links.name = bpf_program__attach(skel->progs.name); \
if (!skel->links.name) \
--
2.30.2
^ permalink raw reply related [flat|nested] 9+ messages in thread
* [PATCH bpf-next 7/7] perf: use generic bpf_program__set_type() to set BPF prog type
2022-01-21 0:41 [PATCH bpf-next 0/7] libbpf: deprecate some setter and getter APIs Andrii Nakryiko
` (5 preceding siblings ...)
2022-01-21 0:41 ` [PATCH bpf-next 6/7] samples/bpf: use preferred getters/setters " Andrii Nakryiko
@ 2022-01-21 0:41 ` Andrii Nakryiko
6 siblings, 0 replies; 9+ messages in thread
From: Andrii Nakryiko @ 2022-01-21 0:41 UTC (permalink / raw)
To: bpf, ast, daniel; +Cc: andrii, kernel-team
bpf_program__set_<type>() APIs are deprecated, use generic
bpf_program__set_type() APIs instead.
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
---
tools/perf/util/bpf-loader.c | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/tools/perf/util/bpf-loader.c b/tools/perf/util/bpf-loader.c
index 4631cac3957f..b54415f2495d 100644
--- a/tools/perf/util/bpf-loader.c
+++ b/tools/perf/util/bpf-loader.c
@@ -651,11 +651,11 @@ int bpf__probe(struct bpf_object *obj)
}
if (priv->is_tp) {
- bpf_program__set_tracepoint(prog);
+ bpf_program__set_type(prog, BPF_PROG_TYPE_TRACEPOINT);
continue;
}
- bpf_program__set_kprobe(prog);
+ bpf_program__set_type(prog, BPF_PROG_TYPE_KPROBE);
pev = &priv->pev;
err = convert_perf_probe_events(pev, 1);
--
2.30.2
^ permalink raw reply related [flat|nested] 9+ messages in thread
* Re: [PATCH bpf-next 2/7] libbpf: deprecate bpf_map__resize()
2022-01-21 0:41 ` [PATCH bpf-next 2/7] libbpf: deprecate bpf_map__resize() Andrii Nakryiko
@ 2022-01-21 0:43 ` Andrii Nakryiko
0 siblings, 0 replies; 9+ messages in thread
From: Andrii Nakryiko @ 2022-01-21 0:43 UTC (permalink / raw)
To: Andrii Nakryiko; +Cc: bpf, Alexei Starovoitov, Daniel Borkmann, Kernel Team
On Thu, Jan 20, 2022 at 4:41 PM Andrii Nakryiko <andrii@kernel.org> wrote:
>
> Deprecated bpf_map__resize() in favor of bpf_map__set_max_entries()
> setter. In addition to having a surprising name (users often don't
> realize that they need to use bpf_map__resize()), the name also implies
> some magic way of resizing BPF map after it is created, which is clearly
> not the case.
>
> Another minor annoyance is that bpf_map__resize() disallows 0 value for
> max_entries, which in some cases is totally acceptable (e.g., like for
> BPF perf buf case to let libbpf auto-create one buffer per each
> available CPU core).
>
> [0] https://github.com/libbpf/libbpf/issues/304
This is supposed to have Closes: there. I'll fix it in the next
revision after waiting for any comments.
>
> Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
> ---
> tools/lib/bpf/libbpf.h | 1 +
> 1 file changed, 1 insertion(+)
>
> diff --git a/tools/lib/bpf/libbpf.h b/tools/lib/bpf/libbpf.h
> index dbf37c0fa531..6f8e6b3cff84 100644
> --- a/tools/lib/bpf/libbpf.h
> +++ b/tools/lib/bpf/libbpf.h
> @@ -716,6 +716,7 @@ LIBBPF_API int bpf_map__set_type(struct bpf_map *map, enum bpf_map_type type);
> /* get/set map size (max_entries) */
> LIBBPF_API __u32 bpf_map__max_entries(const struct bpf_map *map);
> LIBBPF_API int bpf_map__set_max_entries(struct bpf_map *map, __u32 max_entries);
> +LIBBPF_DEPRECATED_SINCE(0, 8, "use bpf_map__set_max_entries() instead")
> LIBBPF_API int bpf_map__resize(struct bpf_map *map, __u32 max_entries);
> /* get/set map flags */
> LIBBPF_API __u32 bpf_map__map_flags(const struct bpf_map *map);
> --
> 2.30.2
>
^ permalink raw reply [flat|nested] 9+ messages in thread
end of thread, other threads:[~2022-01-21 0:44 UTC | newest]
Thread overview: 9+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2022-01-21 0:41 [PATCH bpf-next 0/7] libbpf: deprecate some setter and getter APIs Andrii Nakryiko
2022-01-21 0:41 ` [PATCH bpf-next 1/7] libbpf: hide and discourage inconsistently named getters Andrii Nakryiko
2022-01-21 0:41 ` [PATCH bpf-next 2/7] libbpf: deprecate bpf_map__resize() Andrii Nakryiko
2022-01-21 0:43 ` Andrii Nakryiko
2022-01-21 0:41 ` [PATCH bpf-next 3/7] libbpf: deprecate bpf_program__is_<type>() and bpf_program__set_<type>() APIs Andrii Nakryiko
2022-01-21 0:41 ` [PATCH bpf-next 4/7] bpftool: use preferred setters/getters instead of deprecated ones Andrii Nakryiko
2022-01-21 0:41 ` [PATCH bpf-next 5/7] selftests/bpf: use preferred setter/getter APIs " Andrii Nakryiko
2022-01-21 0:41 ` [PATCH bpf-next 6/7] samples/bpf: use preferred getters/setters " Andrii Nakryiko
2022-01-21 0:41 ` [PATCH bpf-next 7/7] perf: use generic bpf_program__set_type() to set BPF prog type Andrii Nakryiko
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).