Skip to contentSkip to conversation

Move bpf_helpers and add BPF_CORE_READ macros

PATCH Applied Andrii Nakryiko 2 participants 9 messages Oct 8, 2019

Subsystems (6)
lore
Share & tools
lore
#0 lore raw

v5 cover Move bpf_helpers and add BPF_CORE_READ macros

This patch set makes bpf_helpers.h and bpf_endian.h a part of libbpf itself for consumption by user BPF programs, not just selftests. It also splits off tracing helpers into bpf_tracing.h, which also becomes part of libbpf. Some of the legacy stuff (BPF_ANNOTATE_KV_PAIR, load_{byte,half,word}, bpf_map_def with unsupported fields, etc, is extracted into selftests-only bpf_legacy.h. All the selftests and samples are switched to use libbpf's headers and selftests' ones are removed.

As part of this patch set we also add BPF_CORE_READ variadic macros, that are simplifying BPF CO-RE reads, especially the ones that have to follow few pointers. E.g., what in non-BPF world (and when using BCC) would be:

int x = s->a->b.c->d; /* s, a, and b.c are pointers */

today would have to be written using explicit bpf_probe_read() calls as:

  void *t;
  int x;
  bpf_probe_read(&t, sizeof(t), s->a);
  bpf_probe_read(&t, sizeof(t), ((struct b *)t)->b.c);
  bpf_probe_read(&x, sizeof(x), ((struct c *)t)->d);

This is super inconvenient and distracts from program logic a lot. Now, with added BPF_CORE_READ() macros, you can write the above as:

  int x = BPF_CORE_READ(s, a, b.c, d);

Up to 9 levels of pointer chasing are supported, which should be enough for any practical purpose, hopefully, without adding too much boilerplate macro definitions (though there is admittedly some, given how variadic and recursive C macro have to be implemented).

There is also BPF_CORE_READ_INTO() variant, which relies on caller to allocate space for result:

  int x;
  BPF_CORE_READ_INTO(&x, s, a, b.c, d);

Result of last bpf_probe_read() call in the chain of calls is the result of BPF_CORE_READ_INTO(). If any intermediate bpf_probe_read() aall fails, then all the subsequent ones will fail too, so this is sufficient to know whether overall "operation" succeeded or not. No short-circuiting of bpf_probe_read()s is done, though.

BPF_CORE_READ_STR_INTO() is added as well, which differs from BPF_CORE_READ_INTO() only in that last bpf_probe_read() call (to read final field after chasing pointers) is replaced with bpf_probe_read_str(). Result of bpf_probe_read_str() is returned as a result of BPF_CORE_READ_STR_INTO() macro itself, so that applications can track return code and/or length of read string.

Patch set outline:
- patch #1 undoes previously added GCC-specific bpf-helpers.h include;
- patch #2 splits off legacy stuff we don't want to carry over;
- patch #3 adjusts CO-RE reloc tests to avoid subsequent naming conflict with BPF_CORE_READ;
- patch #4 splits off bpf_tracing.h;
- patch #5 moves bpf_{helpers,endian,tracing}.h and bpf_helper_defs.h generation into libbpf and adjusts Makefiles to include libbpf for header search;
- patch #6 adds variadic BPF_CORE_READ() macro family, as described above;
- patch #7 adds tests to verify all possible levels of pointer nestedness for BPF_CORE_READ(), as well as correctness test for BPF_CORE_READ_STR_INTO().

v4->v5:
- move BPF_CORE_READ() stuff into bpf_core_read.h header (Alexei);

v3->v4:
- rebase on latest bpf-next master;
- bpf_helper_defs.h generation is moved into libbpf's Makefile;

v2->v3:
- small formatting fixes and macro () fixes (Song);

v1->v2:
- fix CO-RE reloc tests before bpf_helpers.h move (Song);
- split off legacy stuff we don't want to carry over (Daniel, Toke);
- split off bpf_tracing.h (Daniel);
- fix samples/bpf build (assuming other fixes are applied);
- switch remaining maps either to bpf_map_def_legacy or BTF-defined maps;

Andrii Nakryiko (7):
selftests/bpf: undo GCC-specific bpf_helpers.h changes selftests/bpf: samples/bpf: split off legacy stuff from bpf_helpers.h selftests/bpf: adjust CO-RE reloc tests for new bpf_core_read() macro selftests/bpf: split off tracing-only helpers into bpf_tracing.h
libbpf: move bpf_{helpers,helper_defs,endian,tracing}.h into libbpf
libbpf: add BPF_CORE_READ/BPF_CORE_READ_INTO helpers selftests/bpf: add BPF_CORE_READ and BPF_CORE_READ_STR_INTO macro
tests

 samples/bpf/Makefile                          |   2 +-
 samples/bpf/hbm_kern.h                        |  27 ++-
 samples/bpf/map_perf_test_kern.c              |  24 +--
 samples/bpf/offwaketime_kern.c                |   1 +
 samples/bpf/parse_ldabs.c                     |   1 +
 samples/bpf/sampleip_kern.c                   |   1 +
 samples/bpf/sockex1_kern.c                    |   1 +
 samples/bpf/sockex2_kern.c                    |   1 +
 samples/bpf/sockex3_kern.c                    |   1 +
 samples/bpf/spintest_kern.c                   |   1 +
 samples/bpf/tcbpf1_kern.c                     |   1 +
 samples/bpf/test_map_in_map_kern.c            |  16 +-
 samples/bpf/test_overhead_kprobe_kern.c       |   1 +
 samples/bpf/test_probe_write_user_kern.c      |   1 +
 samples/bpf/trace_event_kern.c                |   1 +
 samples/bpf/tracex1_kern.c                    |   1 +
 samples/bpf/tracex2_kern.c                    |   1 +
 samples/bpf/tracex3_kern.c                    |   1 +
 samples/bpf/tracex4_kern.c                    |   1 +
 samples/bpf/tracex5_kern.c                    |   1 +
 tools/lib/bpf/.gitignore                      |   1 +
 tools/lib/bpf/Makefile                        |  17 +-
 tools/lib/bpf/bpf_core_read.h                 | 167 ++++++++++++++++++
 .../selftests => lib}/bpf/bpf_endian.h        |   0
 tools/lib/bpf/bpf_helpers.h                   |  41 +++++
 .../bpf_helpers.h => lib/bpf/bpf_tracing.h}   |  94 +---------
 tools/testing/selftests/bpf/.gitignore        |   1 -
 tools/testing/selftests/bpf/Makefile          |  10 +-
 tools/testing/selftests/bpf/bpf_legacy.h      |  39 ++++
 .../selftests/bpf/prog_tests/core_reloc.c     |   8 +-
 .../selftests/bpf/progs/core_reloc_types.h    |   9 +
 tools/testing/selftests/bpf/progs/loop1.c     |   1 +
 tools/testing/selftests/bpf/progs/loop2.c     |   1 +
 tools/testing/selftests/bpf/progs/loop3.c     |   1 +
 .../testing/selftests/bpf/progs/sockopt_sk.c  |  13 +-
 tools/testing/selftests/bpf/progs/tcp_rtt.c   |  13 +-
 .../selftests/bpf/progs/test_btf_haskv.c      |   1 +
 .../selftests/bpf/progs/test_btf_newkv.c      |   1 +
 .../bpf/progs/test_core_reloc_arrays.c        |  11 +-
 .../bpf/progs/test_core_reloc_flavors.c       |   9 +-
 .../bpf/progs/test_core_reloc_ints.c          |  19 +-
 .../bpf/progs/test_core_reloc_kernel.c        |  61 ++++++-
 .../bpf/progs/test_core_reloc_misc.c          |   9 +-
 .../bpf/progs/test_core_reloc_mods.c          |  19 +-
 .../bpf/progs/test_core_reloc_nesting.c       |   7 +-
 .../bpf/progs/test_core_reloc_primitives.c    |  13 +-
 .../bpf/progs/test_core_reloc_ptr_as_arr.c    |   5 +-
 47 files changed, 470 insertions(+), 186 deletions(-)
 create mode 100644 tools/lib/bpf/bpf_core_read.h
 rename tools/{testing/selftests => lib}/bpf/bpf_endian.h (100%)
 create mode 100644 tools/lib/bpf/bpf_helpers.h
 rename tools/{testing/selftests/bpf/bpf_helpers.h => lib/bpf/bpf_tracing.h} (68%)
 create mode 100644 tools/testing/selftests/bpf/bpf_legacy.h
Signature · 1 line
2.17.1
#1 lore raw

PATCH v5 1/7 selftests/bpf: undo GCC-specific bpf_helpers.h changes

Having GCC provide its own bpf-helper.h is not the right approach and is going to be changed. Undo bpf_helpers.h change before moving bpf_helpers.h into libbpf.

---
 tools/testing/selftests/bpf/bpf_helpers.h | 8 --------
 1 file changed, 8 deletions(-)
tools/testing/selftests/bpf/bpf_helpers.h +0 -8
index 15152280db6f..ffd4d8c9a087 100644
@@ -15,8 +15,6 @@
 			 ##__VA_ARGS__);		\
 })
 
-#ifdef __clang__
-
 /* helper macro to place programs, maps, license in
  * different sections in elf_bpf file. Section names
  * are interpreted by elf_bpf loader
@@ -47,12 +45,6 @@ struct bpf_map_def {
 	unsigned int numa_node;
 };
 
-#else
-
-#include <bpf-helpers.h>
-
-#endif
-
 #define BPF_ANNOTATE_KV_PAIR(name, type_key, type_val)		\
 	struct ____btf_map_##name {				\
 		type_key key;					\
Signature · 1 line
2.17.1
#2 lore raw

PATCH v5 2/7 selftests/bpf: samples/bpf: split off legacy stuff from bpf_helpers.h

In reply to Andrii Nakryiko

Split off few legacy things from bpf_helpers.h into separate bpf_legacy.h file:
- load_{byte|half|word};
- remove extra inner_idx and numa_node fields from bpf_map_def and introduce bpf_map_def_legacy for use in samples;
- move BPF_ANNOTATE_KV_PAIR into bpf_legacy.h.

Adjust samples and selftests accordingly by either including bpf_legacy.h and using bpf_map_def_legacy, or switching to BTF-defined maps altogether.

---
 samples/bpf/hbm_kern.h                        | 27 ++++++-------
 samples/bpf/map_perf_test_kern.c              | 23 +++++------
 samples/bpf/parse_ldabs.c                     |  1 +
 samples/bpf/sockex1_kern.c                    |  1 +
 samples/bpf/sockex2_kern.c                    |  1 +
 samples/bpf/sockex3_kern.c                    |  1 +
 samples/bpf/tcbpf1_kern.c                     |  1 +
 samples/bpf/test_map_in_map_kern.c            | 15 +++----
 tools/testing/selftests/bpf/bpf_helpers.h     | 24 +-----------
 tools/testing/selftests/bpf/bpf_legacy.h      | 39 +++++++++++++++++++
 .../testing/selftests/bpf/progs/sockopt_sk.c  | 13 +++----
 tools/testing/selftests/bpf/progs/tcp_rtt.c   | 13 +++----
 .../selftests/bpf/progs/test_btf_haskv.c      |  1 +
 .../selftests/bpf/progs/test_btf_newkv.c      |  1 +
 14 files changed, 91 insertions(+), 70 deletions(-)
 create mode 100644 tools/testing/selftests/bpf/bpf_legacy.h
samples/bpf/hbm_kern.h +12 -15
index aa207a2eebbd..4edaf47876ca 100644
@@ -59,21 +59,18 @@
 #define BYTES_PER_NS(delta, rate) ((((u64)(delta)) * (rate)) >> 20)
 #define BYTES_TO_NS(bytes, rate) div64_u64(((u64)(bytes)) << 20, (u64)(rate))
 
-struct bpf_map_def SEC("maps") queue_state = {
-	.type = BPF_MAP_TYPE_CGROUP_STORAGE,
-	.key_size = sizeof(struct bpf_cgroup_storage_key),
-	.value_size = sizeof(struct hbm_vqueue),
-};
-BPF_ANNOTATE_KV_PAIR(queue_state, struct bpf_cgroup_storage_key,
-		     struct hbm_vqueue);
-
-struct bpf_map_def SEC("maps") queue_stats = {
-	.type = BPF_MAP_TYPE_ARRAY,
-	.key_size = sizeof(u32),
-	.value_size = sizeof(struct hbm_queue_stats),
-	.max_entries = 1,
-};
-BPF_ANNOTATE_KV_PAIR(queue_stats, int, struct hbm_queue_stats);
+struct {
+	__uint(type, BPF_MAP_TYPE_CGROUP_STORAGE);
+	__type(key, struct bpf_cgroup_storage_key);
+	__type(value, struct hbm_vqueue);
+} queue_state SEC(".maps");
+
+struct {
+	__uint(type, BPF_MAP_TYPE_ARRAY);
+	__uint(max_entries, 1);
+	__type(key, u32);
+	__type(value, struct hvm_queue_stats);
+} queue_stats SEC(".maps");
 
 struct hbm_pkt_info {
 	int	cwnd;
samples/bpf/map_perf_test_kern.c +12 -11
index 2b2ffb97018b..f47ee513cb7c 100644
@@ -9,25 +9,26 @@
 #include <linux/version.h>
 #include <uapi/linux/bpf.h>
 #include "bpf_helpers.h"
+#include "bpf_legacy.h"
 
 #define MAX_ENTRIES 1000
 #define MAX_NR_CPUS 1024
 
-struct bpf_map_def SEC("maps") hash_map = {
+struct bpf_map_def_legacy SEC("maps") hash_map = {
 	.type = BPF_MAP_TYPE_HASH,
 	.key_size = sizeof(u32),
 	.value_size = sizeof(long),
 	.max_entries = MAX_ENTRIES,
 };
 
-struct bpf_map_def SEC("maps") lru_hash_map = {
+struct bpf_map_def_legacy SEC("maps") lru_hash_map = {
 	.type = BPF_MAP_TYPE_LRU_HASH,
 	.key_size = sizeof(u32),
 	.value_size = sizeof(long),
 	.max_entries = 10000,
 };
 
-struct bpf_map_def SEC("maps") nocommon_lru_hash_map = {
+struct bpf_map_def_legacy SEC("maps") nocommon_lru_hash_map = {
 	.type = BPF_MAP_TYPE_LRU_HASH,
 	.key_size = sizeof(u32),
 	.value_size = sizeof(long),
@@ -35,7 +36,7 @@ struct bpf_map_def SEC("maps") nocommon_lru_hash_map = {
 	.map_flags = BPF_F_NO_COMMON_LRU,
 };
 
-struct bpf_map_def SEC("maps") inner_lru_hash_map = {
+struct bpf_map_def_legacy SEC("maps") inner_lru_hash_map = {
 	.type = BPF_MAP_TYPE_LRU_HASH,
 	.key_size = sizeof(u32),
 	.value_size = sizeof(long),
@@ -44,20 +45,20 @@ struct bpf_map_def SEC("maps") inner_lru_hash_map = {
 	.numa_node = 0,
 };
 
-struct bpf_map_def SEC("maps") array_of_lru_hashs = {
+struct bpf_map_def_legacy SEC("maps") array_of_lru_hashs = {
 	.type = BPF_MAP_TYPE_ARRAY_OF_MAPS,
 	.key_size = sizeof(u32),
 	.max_entries = MAX_NR_CPUS,
 };
 
-struct bpf_map_def SEC("maps") percpu_hash_map = {
+struct bpf_map_def_legacy SEC("maps") percpu_hash_map = {
 	.type = BPF_MAP_TYPE_PERCPU_HASH,
 	.key_size = sizeof(u32),
 	.value_size = sizeof(long),
 	.max_entries = MAX_ENTRIES,
 };
 
-struct bpf_map_def SEC("maps") hash_map_alloc = {
+struct bpf_map_def_legacy SEC("maps") hash_map_alloc = {
 	.type = BPF_MAP_TYPE_HASH,
 	.key_size = sizeof(u32),
 	.value_size = sizeof(long),
@@ -65,7 +66,7 @@ struct bpf_map_def SEC("maps") hash_map_alloc = {
 	.map_flags = BPF_F_NO_PREALLOC,
 };
 
-struct bpf_map_def SEC("maps") percpu_hash_map_alloc = {
+struct bpf_map_def_legacy SEC("maps") percpu_hash_map_alloc = {
 	.type = BPF_MAP_TYPE_PERCPU_HASH,
 	.key_size = sizeof(u32),
 	.value_size = sizeof(long),
@@ -73,7 +74,7 @@ struct bpf_map_def SEC("maps") percpu_hash_map_alloc = {
 	.map_flags = BPF_F_NO_PREALLOC,
 };
 
-struct bpf_map_def SEC("maps") lpm_trie_map_alloc = {
+struct bpf_map_def_legacy SEC("maps") lpm_trie_map_alloc = {
 	.type = BPF_MAP_TYPE_LPM_TRIE,
 	.key_size = 8,
 	.value_size = sizeof(long),
@@ -81,14 +82,14 @@ struct bpf_map_def SEC("maps") lpm_trie_map_alloc = {
 	.map_flags = BPF_F_NO_PREALLOC,
 };
 
-struct bpf_map_def SEC("maps") array_map = {
+struct bpf_map_def_legacy SEC("maps") array_map = {
 	.type = BPF_MAP_TYPE_ARRAY,
 	.key_size = sizeof(u32),
 	.value_size = sizeof(long),
 	.max_entries = MAX_ENTRIES,
 };
 
-struct bpf_map_def SEC("maps") lru_hash_lookup_map = {
+struct bpf_map_def_legacy SEC("maps") lru_hash_lookup_map = {
 	.type = BPF_MAP_TYPE_LRU_HASH,
 	.key_size = sizeof(u32),
 	.value_size = sizeof(long),
samples/bpf/parse_ldabs.c +1 -0
index 6db6b21fdc6d..ef5892377beb 100644
@@ -12,6 +12,7 @@
 #include <linux/udp.h>
 #include <uapi/linux/bpf.h>
 #include "bpf_helpers.h"
+#include "bpf_legacy.h"
 
 #define DEFAULT_PKTGEN_UDP_PORT	9
 #define IP_MF			0x2000
samples/bpf/sockex1_kern.c +1 -0
index ed18e9a4909c..f96943f443ab 100644
@@ -3,6 +3,7 @@
 #include <uapi/linux/if_packet.h>
 #include <uapi/linux/ip.h>
 #include "bpf_helpers.h"
+#include "bpf_legacy.h"
 
 struct bpf_map_def SEC("maps") my_map = {
 	.type = BPF_MAP_TYPE_ARRAY,
samples/bpf/sockex2_kern.c +1 -0
index f2f9dbc021b0..5566fa7d92fa 100644
@@ -1,5 +1,6 @@
 #include <uapi/linux/bpf.h>
 #include "bpf_helpers.h"
+#include "bpf_legacy.h"
 #include <uapi/linux/in.h>
 #include <uapi/linux/if.h>
 #include <uapi/linux/if_ether.h>
samples/bpf/sockex3_kern.c +1 -0
index c527b57d3ec8..151dd842ecc0 100644
@@ -6,6 +6,7 @@
  */
 #include <uapi/linux/bpf.h>
 #include "bpf_helpers.h"
+#include "bpf_legacy.h"
 #include <uapi/linux/in.h>
 #include <uapi/linux/if.h>
 #include <uapi/linux/if_ether.h>
samples/bpf/tcbpf1_kern.c +1 -0
index 274c884c87fe..ff43341bdfce 100644
@@ -8,6 +8,7 @@
 #include <uapi/linux/filter.h>
 #include <uapi/linux/pkt_cls.h>
 #include "bpf_helpers.h"
+#include "bpf_legacy.h"
 
 /* compiler workaround */
 #define _htonl __builtin_bswap32
samples/bpf/test_map_in_map_kern.c +8 -7
index 42c44d091dd1..8101bf3dc7f7 100644
@@ -11,11 +11,12 @@
 #include <uapi/linux/bpf.h>
 #include <uapi/linux/in6.h>
 #include "bpf_helpers.h"
+#include "bpf_legacy.h"
 
 #define MAX_NR_PORTS 65536
 
 /* map #0 */
-struct bpf_map_def SEC("maps") port_a = {
+struct bpf_map_def_legacy SEC("maps") port_a = {
 	.type = BPF_MAP_TYPE_ARRAY,
 	.key_size = sizeof(u32),
 	.value_size = sizeof(int),
@@ -23,7 +24,7 @@ struct bpf_map_def SEC("maps") port_a = {
 };
 
 /* map #1 */
-struct bpf_map_def SEC("maps") port_h = {
+struct bpf_map_def_legacy SEC("maps") port_h = {
 	.type = BPF_MAP_TYPE_HASH,
 	.key_size = sizeof(u32),
 	.value_size = sizeof(int),
@@ -31,7 +32,7 @@ struct bpf_map_def SEC("maps") port_h = {
 };
 
 /* map #2 */
-struct bpf_map_def SEC("maps") reg_result_h = {
+struct bpf_map_def_legacy SEC("maps") reg_result_h = {
 	.type = BPF_MAP_TYPE_HASH,
 	.key_size = sizeof(u32),
 	.value_size = sizeof(int),
@@ -39,7 +40,7 @@ struct bpf_map_def SEC("maps") reg_result_h = {
 };
 
 /* map #3 */
-struct bpf_map_def SEC("maps") inline_result_h = {
+struct bpf_map_def_legacy SEC("maps") inline_result_h = {
 	.type = BPF_MAP_TYPE_HASH,
 	.key_size = sizeof(u32),
 	.value_size = sizeof(int),
@@ -47,7 +48,7 @@ struct bpf_map_def SEC("maps") inline_result_h = {
 };
 
 /* map #4 */ /* Test case #0 */
-struct bpf_map_def SEC("maps") a_of_port_a = {
+struct bpf_map_def_legacy SEC("maps") a_of_port_a = {
 	.type = BPF_MAP_TYPE_ARRAY_OF_MAPS,
 	.key_size = sizeof(u32),
 	.inner_map_idx = 0, /* map_fd[0] is port_a */
@@ -55,7 +56,7 @@ struct bpf_map_def SEC("maps") a_of_port_a = {
 };
 
 /* map #5 */ /* Test case #1 */
-struct bpf_map_def SEC("maps") h_of_port_a = {
+struct bpf_map_def_legacy SEC("maps") h_of_port_a = {
 	.type = BPF_MAP_TYPE_HASH_OF_MAPS,
 	.key_size = sizeof(u32),
 	.inner_map_idx = 0, /* map_fd[0] is port_a */
@@ -63,7 +64,7 @@ struct bpf_map_def SEC("maps") h_of_port_a = {
 };
 
 /* map #6 */ /* Test case #2 */
-struct bpf_map_def SEC("maps") h_of_port_h = {
+struct bpf_map_def_legacy SEC("maps") h_of_port_h = {
 	.type = BPF_MAP_TYPE_HASH_OF_MAPS,
 	.key_size = sizeof(u32),
 	.inner_map_idx = 1, /* map_fd[1] is port_h */
tools/testing/selftests/bpf/bpf_helpers.h +1 -23
index ffd4d8c9a087..c7cfc27063d4 100644
@@ -21,19 +21,8 @@
  */
 #define SEC(NAME) __attribute__((section(NAME), used))
 
-/* llvm builtin functions that eBPF C program may use to
- * emit BPF_LD_ABS and BPF_LD_IND instructions
- */
-struct sk_buff;
-unsigned long long load_byte(void *skb,
-			     unsigned long long off) asm("llvm.bpf.load.byte");
-unsigned long long load_half(void *skb,
-			     unsigned long long off) asm("llvm.bpf.load.half");
-unsigned long long load_word(void *skb,
-			     unsigned long long off) asm("llvm.bpf.load.word");
-
 /* a helper structure used by eBPF C program
- * to describe map attributes to elf_bpf loader
+ * to describe BPF map attributes to libbpf loader
  */
 struct bpf_map_def {
 	unsigned int type;
@@ -41,19 +30,8 @@ struct bpf_map_def {
 	unsigned int value_size;
 	unsigned int max_entries;
 	unsigned int map_flags;
-	unsigned int inner_map_idx;
-	unsigned int numa_node;
 };
 
-#define BPF_ANNOTATE_KV_PAIR(name, type_key, type_val)		\
-	struct ____btf_map_##name {				\
-		type_key key;					\
-		type_val value;					\
-	};							\
-	struct ____btf_map_##name				\
-	__attribute__ ((section(".maps." #name), used))		\
-		____btf_map_##name = { }
-
 /* Scan the ARCH passed in from ARCH env variable (see Makefile) */
 #if defined(__TARGET_ARCH_x86)
 	#define bpf_target_x86
tools/testing/selftests/bpf/bpf_legacy.h new file +39 -0
new file mode 100644
index 000000000000..6f8988738bc1
@@ -0,0 +1,39 @@
+/* SPDX-License-Identifier: (LGPL-2.1 OR BSD-2-Clause) */
+#ifndef __BPF_LEGACY__
+#define __BPF_LEGACY__
+
+/*
+ * legacy bpf_map_def with extra fields supported only by bpf_load(), do not
+ * use outside of samples/bpf
+ */
+struct bpf_map_def_legacy {
+	unsigned int type;
+	unsigned int key_size;
+	unsigned int value_size;
+	unsigned int max_entries;
+	unsigned int map_flags;
+	unsigned int inner_map_idx;
+	unsigned int numa_node;
+};
+
+#define BPF_ANNOTATE_KV_PAIR(name, type_key, type_val)		\
+	struct ____btf_map_##name {				\
+		type_key key;					\
+		type_val value;					\
+	};							\
+	struct ____btf_map_##name				\
+	__attribute__ ((section(".maps." #name), used))		\
+		____btf_map_##name = { }
+
+/* llvm builtin functions that eBPF C program may use to
+ * emit BPF_LD_ABS and BPF_LD_IND instructions
+ */
+unsigned long long load_byte(void *skb,
+			     unsigned long long off) asm("llvm.bpf.load.byte");
+unsigned long long load_half(void *skb,
+			     unsigned long long off) asm("llvm.bpf.load.half");
+unsigned long long load_word(void *skb,
+			     unsigned long long off) asm("llvm.bpf.load.word");
+
+#endif
+
tools/testing/selftests/bpf/progs/sockopt_sk.c +6 -7
index 9a3d1c79e6fe..1bafbb944e37 100644
@@ -14,13 +14,12 @@ struct sockopt_sk {
 	__u8 val;
 };
 
-struct bpf_map_def SEC("maps") socket_storage_map = {
-	.type = BPF_MAP_TYPE_SK_STORAGE,
-	.key_size = sizeof(int),
-	.value_size = sizeof(struct sockopt_sk),
-	.map_flags = BPF_F_NO_PREALLOC,
-};
-BPF_ANNOTATE_KV_PAIR(socket_storage_map, int, struct sockopt_sk);
+struct {
+	__uint(type, BPF_MAP_TYPE_SK_STORAGE);
+	__uint(map_flags, BPF_F_NO_PREALLOC);
+	__type(key, int);
+	__type(value, struct sockopt_sk);
+} socket_storage_map SEC(".maps");
 
 SEC("cgroup/getsockopt")
 int _getsockopt(struct bpf_sockopt *ctx)
tools/testing/selftests/bpf/progs/tcp_rtt.c +6 -7
index 233bdcb1659e..2cf813a06b83 100644
@@ -13,13 +13,12 @@ struct tcp_rtt_storage {
 	__u32 icsk_retransmits;
 };
 
-struct bpf_map_def SEC("maps") socket_storage_map = {
-	.type = BPF_MAP_TYPE_SK_STORAGE,
-	.key_size = sizeof(int),
-	.value_size = sizeof(struct tcp_rtt_storage),
-	.map_flags = BPF_F_NO_PREALLOC,
-};
-BPF_ANNOTATE_KV_PAIR(socket_storage_map, int, struct tcp_rtt_storage);
+struct {
+	__uint(type, BPF_MAP_TYPE_SK_STORAGE);
+	__uint(map_flags, BPF_F_NO_PREALLOC);
+	__type(key, int);
+	__type(value, struct tcp_rtt_storage);
+} socket_storage_map SEC(".maps");
 
 SEC("sockops")
 int _sockops(struct bpf_sock_ops *ctx)
tools/testing/selftests/bpf/progs/test_btf_haskv.c +1 -0
index e5c79fe0ffdb..763c51447c19 100644
@@ -2,6 +2,7 @@
 /* Copyright (c) 2018 Facebook */
 #include <linux/bpf.h>
 #include "bpf_helpers.h"
+#include "bpf_legacy.h"
 
 int _version SEC("version") = 1;
 
tools/testing/selftests/bpf/progs/test_btf_newkv.c +1 -0
index 5ee3622ddebb..96f9e8451029 100644
@@ -2,6 +2,7 @@
 /* Copyright (c) 2018 Facebook */
 #include <linux/bpf.h>
 #include "bpf_helpers.h"
+#include "bpf_legacy.h"
 
 int _version SEC("version") = 1;
 
Signature · 1 line
2.17.1
#3 lore raw

PATCH v5 3/7 selftests/bpf: adjust CO-RE reloc tests for new bpf_core_read() macro

In reply to Andrii Nakryiko

To allow adding a variadic BPF_CORE_READ macro with slightly different syntax and semantics, define CORE_READ in CO-RE reloc tests, which is a thin wrapper around low-level bpf_core_read() macro, which in turn is just a wrapper around bpf_probe_read().

---
 tools/testing/selftests/bpf/bpf_helpers.h      |  8 ++++----
 .../bpf/progs/test_core_reloc_arrays.c         | 10 ++++++----
 .../bpf/progs/test_core_reloc_flavors.c        |  8 +++++---
 .../selftests/bpf/progs/test_core_reloc_ints.c | 18 ++++++++++--------
 .../bpf/progs/test_core_reloc_kernel.c         |  6 ++++--
 .../selftests/bpf/progs/test_core_reloc_misc.c |  8 +++++---
 .../selftests/bpf/progs/test_core_reloc_mods.c | 18 ++++++++++--------
 .../bpf/progs/test_core_reloc_nesting.c        |  6 ++++--
 .../bpf/progs/test_core_reloc_primitives.c     | 12 +++++++-----
 .../bpf/progs/test_core_reloc_ptr_as_arr.c     |  4 +++-
 10 files changed, 58 insertions(+), 40 deletions(-)
tools/testing/selftests/bpf/bpf_helpers.h +4 -4
index c7cfc27063d4..4e6f97142cd8 100644
@@ -223,7 +223,7 @@ struct pt_regs;
 #endif
 
 /*
- * BPF_CORE_READ abstracts away bpf_probe_read() call and captures offset
+ * bpf_core_read() abstracts away bpf_probe_read() call and captures offset
  * relocation for source address using __builtin_preserve_access_index()
  * built-in, provided by Clang.
  *
@@ -238,8 +238,8 @@ struct pt_regs;
  * actual field offset, based on target kernel BTF type that matches original
  * (local) BTF, used to record relocation.
  */
-#define BPF_CORE_READ(dst, src)						\
-	bpf_probe_read((dst), sizeof(*(src)),				\
-		       __builtin_preserve_access_index(src))
+#define bpf_core_read(dst, sz, src)					    \
+	bpf_probe_read(dst, sz,						    \
+		       (const void *)__builtin_preserve_access_index(src))
 
 #endif
tools/testing/selftests/bpf/progs/test_core_reloc_arrays.c +6 -4
index bf67f0fdf743..8e3f6e6a90e7 100644
@@ -31,6 +31,8 @@ struct core_reloc_arrays {
 	struct core_reloc_arrays_substruct d[1][2];
 };
 
+#define CORE_READ(dst, src) bpf_core_read(dst, sizeof(*(dst)), src)
+
 SEC("raw_tracepoint/sys_enter")
 int test_core_arrays(void *ctx)
 {
@@ -38,16 +40,16 @@ int test_core_arrays(void *ctx)
 	struct core_reloc_arrays_output *out = (void *)&data.out;
 
 	/* in->a[2] */
-	if (BPF_CORE_READ(&out->a2, &in->a[2]))
+	if (CORE_READ(&out->a2, &in->a[2]))
 		return 1;
 	/* in->b[1][2][3] */
-	if (BPF_CORE_READ(&out->b123, &in->b[1][2][3]))
+	if (CORE_READ(&out->b123, &in->b[1][2][3]))
 		return 1;
 	/* in->c[1].c */
-	if (BPF_CORE_READ(&out->c1c, &in->c[1].c))
+	if (CORE_READ(&out->c1c, &in->c[1].c))
 		return 1;
 	/* in->d[0][0].d */
-	if (BPF_CORE_READ(&out->d00d, &in->d[0][0].d))
+	if (CORE_READ(&out->d00d, &in->d[0][0].d))
 		return 1;
 
 	return 0;
tools/testing/selftests/bpf/progs/test_core_reloc_flavors.c +5 -3
index 9fda73e87972..613474a18b45 100644
@@ -39,6 +39,8 @@ struct core_reloc_flavors___weird {
 	};
 };
 
+#define CORE_READ(dst, src) bpf_core_read(dst, sizeof(*(dst)), src)
+
 SEC("raw_tracepoint/sys_enter")
 int test_core_flavors(void *ctx)
 {
@@ -48,13 +50,13 @@ int test_core_flavors(void *ctx)
 	struct core_reloc_flavors *out = (void *)&data.out;
 
 	/* read a using weird layout */
-	if (BPF_CORE_READ(&out->a, &in_weird->a))
+	if (CORE_READ(&out->a, &in_weird->a))
 		return 1;
 	/* read b using reversed layout */
-	if (BPF_CORE_READ(&out->b, &in_rev->b))
+	if (CORE_READ(&out->b, &in_rev->b))
 		return 1;
 	/* read c using original layout */
-	if (BPF_CORE_READ(&out->c, &in_orig->c))
+	if (CORE_READ(&out->c, &in_orig->c))
 		return 1;
 
 	return 0;
tools/testing/selftests/bpf/progs/test_core_reloc_ints.c +10 -8
index d99233c8008a..7a88a3975455 100644
@@ -23,20 +23,22 @@ struct core_reloc_ints {
 	int64_t		s64_field;
 };
 
+#define CORE_READ(dst, src) bpf_core_read(dst, sizeof(*(dst)), src)
+
 SEC("raw_tracepoint/sys_enter")
 int test_core_ints(void *ctx)
 {
 	struct core_reloc_ints *in = (void *)&data.in;
 	struct core_reloc_ints *out = (void *)&data.out;
 
-	if (BPF_CORE_READ(&out->u8_field, &in->u8_field) ||
-	    BPF_CORE_READ(&out->s8_field, &in->s8_field) ||
-	    BPF_CORE_READ(&out->u16_field, &in->u16_field) ||
-	    BPF_CORE_READ(&out->s16_field, &in->s16_field) ||
-	    BPF_CORE_READ(&out->u32_field, &in->u32_field) ||
-	    BPF_CORE_READ(&out->s32_field, &in->s32_field) ||
-	    BPF_CORE_READ(&out->u64_field, &in->u64_field) ||
-	    BPF_CORE_READ(&out->s64_field, &in->s64_field))
+	if (CORE_READ(&out->u8_field, &in->u8_field) ||
+	    CORE_READ(&out->s8_field, &in->s8_field) ||
+	    CORE_READ(&out->u16_field, &in->u16_field) ||
+	    CORE_READ(&out->s16_field, &in->s16_field) ||
+	    CORE_READ(&out->u32_field, &in->u32_field) ||
+	    CORE_READ(&out->s32_field, &in->s32_field) ||
+	    CORE_READ(&out->u64_field, &in->u64_field) ||
+	    CORE_READ(&out->s64_field, &in->s64_field))
 		return 1;
 
 	return 0;
tools/testing/selftests/bpf/progs/test_core_reloc_kernel.c +4 -2
index 37e02aa3f0c8..684a06cf41ea 100644
@@ -17,6 +17,8 @@ struct task_struct {
 	int tgid;
 };
 
+#define CORE_READ(dst, src) bpf_core_read(dst, sizeof(*(dst)), src)
+
 SEC("raw_tracepoint/sys_enter")
 int test_core_kernel(void *ctx)
 {
@@ -24,8 +26,8 @@ int test_core_kernel(void *ctx)
 	uint64_t pid_tgid = bpf_get_current_pid_tgid();
 	int pid, tgid;
 
-	if (BPF_CORE_READ(&pid, &task->pid) ||
-	    BPF_CORE_READ(&tgid, &task->tgid))
+	if (CORE_READ(&pid, &task->pid) ||
+	    CORE_READ(&tgid, &task->tgid))
 		return 1;
 
 	/* validate pid + tgid matches */
tools/testing/selftests/bpf/progs/test_core_reloc_misc.c +5 -3
index c59984bd3e23..10bdb2050552 100644
@@ -32,6 +32,8 @@ struct core_reloc_misc_extensible {
 	int b;
 };
 
+#define CORE_READ(dst, src) bpf_core_read(dst, sizeof(*(dst)), src)
+
 SEC("raw_tracepoint/sys_enter")
 int test_core_misc(void *ctx)
 {
@@ -41,15 +43,15 @@ int test_core_misc(void *ctx)
 	struct core_reloc_misc_output *out = (void *)&data.out;
 
 	/* record two different relocations with the same accessor string */
-	if (BPF_CORE_READ(&out->a, &in_a->a1) ||	/* accessor: 0:0 */
-	    BPF_CORE_READ(&out->b, &in_b->b1))		/* accessor: 0:0 */
+	if (CORE_READ(&out->a, &in_a->a1) ||		/* accessor: 0:0 */
+	    CORE_READ(&out->b, &in_b->b1))		/* accessor: 0:0 */
 		return 1;
 
 	/* Validate relocations capture array-only accesses for structs with
 	 * fixed header, but with potentially extendable tail. This will read
 	 * first 4 bytes of 2nd element of in_ext array of potentially
 	 * variably sized struct core_reloc_misc_extensible. */ 
-	if (BPF_CORE_READ(&out->c, &in_ext[2]))		/* accessor: 2 */
+	if (CORE_READ(&out->c, &in_ext[2]))		/* accessor: 2 */
 		return 1;
 
 	return 0;
tools/testing/selftests/bpf/progs/test_core_reloc_mods.c +10 -8
index f98b942c062b..e930e7e88c5c 100644
@@ -41,20 +41,22 @@ struct core_reloc_mods {
 	core_reloc_mods_substruct_t h;
 };
 
+#define CORE_READ(dst, src) bpf_core_read(dst, sizeof(*(dst)), src)
+
 SEC("raw_tracepoint/sys_enter")
 int test_core_mods(void *ctx)
 {
 	struct core_reloc_mods *in = (void *)&data.in;
 	struct core_reloc_mods_output *out = (void *)&data.out;
 
-	if (BPF_CORE_READ(&out->a, &in->a) ||
-	    BPF_CORE_READ(&out->b, &in->b) ||
-	    BPF_CORE_READ(&out->c, &in->c) ||
-	    BPF_CORE_READ(&out->d, &in->d) ||
-	    BPF_CORE_READ(&out->e, &in->e[2]) ||
-	    BPF_CORE_READ(&out->f, &in->f[1]) ||
-	    BPF_CORE_READ(&out->g, &in->g.x) ||
-	    BPF_CORE_READ(&out->h, &in->h.y))
+	if (CORE_READ(&out->a, &in->a) ||
+	    CORE_READ(&out->b, &in->b) ||
+	    CORE_READ(&out->c, &in->c) ||
+	    CORE_READ(&out->d, &in->d) ||
+	    CORE_READ(&out->e, &in->e[2]) ||
+	    CORE_READ(&out->f, &in->f[1]) ||
+	    CORE_READ(&out->g, &in->g.x) ||
+	    CORE_READ(&out->h, &in->h.y))
 		return 1;
 
 	return 0;
tools/testing/selftests/bpf/progs/test_core_reloc_nesting.c +4 -2
index 3ca30cec2b39..b63007958290 100644
@@ -30,15 +30,17 @@ struct core_reloc_nesting {
 	} b;
 };
 
+#define CORE_READ(dst, src) bpf_core_read(dst, sizeof(*(dst)), src)
+
 SEC("raw_tracepoint/sys_enter")
 int test_core_nesting(void *ctx)
 {
 	struct core_reloc_nesting *in = (void *)&data.in;
 	struct core_reloc_nesting *out = (void *)&data.out;
 
-	if (BPF_CORE_READ(&out->a.a.a, &in->a.a.a))
+	if (CORE_READ(&out->a.a.a, &in->a.a.a))
 		return 1;
-	if (BPF_CORE_READ(&out->b.b.b, &in->b.b.b))
+	if (CORE_READ(&out->b.b.b, &in->b.b.b))
 		return 1;
 
 	return 0;
tools/testing/selftests/bpf/progs/test_core_reloc_primitives.c +7 -5
index add52f23ab35..7654f59914bc 100644
@@ -25,17 +25,19 @@ struct core_reloc_primitives {
 	int (*f)(const char *);
 };
 
+#define CORE_READ(dst, src) bpf_core_read(dst, sizeof(*(dst)), src)
+
 SEC("raw_tracepoint/sys_enter")
 int test_core_primitives(void *ctx)
 {
 	struct core_reloc_primitives *in = (void *)&data.in;
 	struct core_reloc_primitives *out = (void *)&data.out;
 
-	if (BPF_CORE_READ(&out->a, &in->a) ||
-	    BPF_CORE_READ(&out->b, &in->b) ||
-	    BPF_CORE_READ(&out->c, &in->c) ||
-	    BPF_CORE_READ(&out->d, &in->d) ||
-	    BPF_CORE_READ(&out->f, &in->f))
+	if (CORE_READ(&out->a, &in->a) ||
+	    CORE_READ(&out->b, &in->b) ||
+	    CORE_READ(&out->c, &in->c) ||
+	    CORE_READ(&out->d, &in->d) ||
+	    CORE_READ(&out->f, &in->f))
 		return 1;
 
 	return 0;
tools/testing/selftests/bpf/progs/test_core_reloc_ptr_as_arr.c +3 -1
index 526b7ddc7ea1..709f7cba453f 100644
@@ -16,13 +16,15 @@ struct core_reloc_ptr_as_arr {
 	int a;
 };
 
+#define CORE_READ(dst, src) bpf_core_read(dst, sizeof(*(dst)), src)
+
 SEC("raw_tracepoint/sys_enter")
 int test_core_ptr_as_arr(void *ctx)
 {
 	struct core_reloc_ptr_as_arr *in = (void *)&data.in;
 	struct core_reloc_ptr_as_arr *out = (void *)&data.out;
 
-	if (BPF_CORE_READ(&out->a, &in[2].a))
+	if (CORE_READ(&out->a, &in[2].a))
 		return 1;
 
 	return 0;
Signature · 1 line
2.17.1
#4 lore raw

PATCH v5 4/7 selftests/bpf: split off tracing-only helpers into bpf_tracing.h

In reply to Andrii Nakryiko

Split-off PT_REGS-related helpers into bpf_tracing.h header. Adjust selftests and samples to include it where necessary.

---
 samples/bpf/map_perf_test_kern.c          |   1 +
 samples/bpf/offwaketime_kern.c            |   1 +
 samples/bpf/sampleip_kern.c               |   1 +
 samples/bpf/spintest_kern.c               |   1 +
 samples/bpf/test_map_in_map_kern.c        |   1 +
 samples/bpf/test_overhead_kprobe_kern.c   |   1 +
 samples/bpf/test_probe_write_user_kern.c  |   1 +
 samples/bpf/trace_event_kern.c            |   1 +
 samples/bpf/tracex1_kern.c                |   1 +
 samples/bpf/tracex2_kern.c                |   1 +
 samples/bpf/tracex3_kern.c                |   1 +
 samples/bpf/tracex4_kern.c                |   1 +
 samples/bpf/tracex5_kern.c                |   1 +
 tools/testing/selftests/bpf/bpf_helpers.h | 190 ---------------------
 tools/testing/selftests/bpf/bpf_tracing.h | 195 ++++++++++++++++++++++
 tools/testing/selftests/bpf/progs/loop1.c |   1 +
 tools/testing/selftests/bpf/progs/loop2.c |   1 +
 tools/testing/selftests/bpf/progs/loop3.c |   1 +
 18 files changed, 211 insertions(+), 190 deletions(-)
 create mode 100644 tools/testing/selftests/bpf/bpf_tracing.h
samples/bpf/map_perf_test_kern.c +1 -0
index f47ee513cb7c..5c11aefbc489 100644
@@ -10,6 +10,7 @@
 #include <uapi/linux/bpf.h>
 #include "bpf_helpers.h"
 #include "bpf_legacy.h"
+#include "bpf_tracing.h"
 
 #define MAX_ENTRIES 1000
 #define MAX_NR_CPUS 1024
samples/bpf/offwaketime_kern.c +1 -0
index e7d9a0a3d45b..9cb5207a692f 100644
@@ -6,6 +6,7 @@
  */
 #include <uapi/linux/bpf.h>
 #include "bpf_helpers.h"
+#include "bpf_tracing.h"
 #include <uapi/linux/ptrace.h>
 #include <uapi/linux/perf_event.h>
 #include <linux/version.h>
samples/bpf/sampleip_kern.c +1 -0
index ceabf31079cf..4a190893894f 100644
@@ -9,6 +9,7 @@
 #include <uapi/linux/bpf.h>
 #include <uapi/linux/bpf_perf_event.h>
 #include "bpf_helpers.h"
+#include "bpf_tracing.h"
 
 #define MAX_IPS		8192
 
samples/bpf/spintest_kern.c +1 -0
index ce0167d09cdc..6e9478aa2938 100644
@@ -10,6 +10,7 @@
 #include <uapi/linux/bpf.h>
 #include <uapi/linux/perf_event.h>
 #include "bpf_helpers.h"
+#include "bpf_tracing.h"
 
 struct bpf_map_def SEC("maps") my_map = {
 	.type = BPF_MAP_TYPE_HASH,
samples/bpf/test_map_in_map_kern.c +1 -0
index 8101bf3dc7f7..4f80cbe74c72 100644
@@ -12,6 +12,7 @@
 #include <uapi/linux/in6.h>
 #include "bpf_helpers.h"
 #include "bpf_legacy.h"
+#include "bpf_tracing.h"
 
 #define MAX_NR_PORTS 65536
 
samples/bpf/test_overhead_kprobe_kern.c +1 -0
index 468a66a92ef9..8d2518e68db9 100644
@@ -8,6 +8,7 @@
 #include <linux/ptrace.h>
 #include <uapi/linux/bpf.h>
 #include "bpf_helpers.h"
+#include "bpf_tracing.h"
 
 #define _(P) ({typeof(P) val = 0; bpf_probe_read(&val, sizeof(val), &P); val;})
 
samples/bpf/test_probe_write_user_kern.c +1 -0
index 3a677c807044..a543358218e6 100644
@@ -9,6 +9,7 @@
 #include <uapi/linux/bpf.h>
 #include <linux/version.h>
 #include "bpf_helpers.h"
+#include "bpf_tracing.h"
 
 struct bpf_map_def SEC("maps") dnat_map = {
 	.type = BPF_MAP_TYPE_HASH,
samples/bpf/trace_event_kern.c +1 -0
index 7068fbdde951..8dc18d233a27 100644
@@ -10,6 +10,7 @@
 #include <uapi/linux/bpf_perf_event.h>
 #include <uapi/linux/perf_event.h>
 #include "bpf_helpers.h"
+#include "bpf_tracing.h"
 
 struct key_t {
 	char comm[TASK_COMM_LEN];
samples/bpf/tracex1_kern.c +1 -0
index 107da148820f..1a15f6605129 100644
@@ -9,6 +9,7 @@
 #include <uapi/linux/bpf.h>
 #include <linux/version.h>
 #include "bpf_helpers.h"
+#include "bpf_tracing.h"
 
 #define _(P) ({typeof(P) val = 0; bpf_probe_read(&val, sizeof(val), &P); val;})
 
samples/bpf/tracex2_kern.c +1 -0
index 5e11c20ce5ec..d70b3ea79ea7 100644
@@ -9,6 +9,7 @@
 #include <linux/version.h>
 #include <uapi/linux/bpf.h>
 #include "bpf_helpers.h"
+#include "bpf_tracing.h"
 
 struct bpf_map_def SEC("maps") my_map = {
 	.type = BPF_MAP_TYPE_HASH,
samples/bpf/tracex3_kern.c +1 -0
index ea1d4c19c132..9af546bebfa9 100644
@@ -9,6 +9,7 @@
 #include <linux/version.h>
 #include <uapi/linux/bpf.h>
 #include "bpf_helpers.h"
+#include "bpf_tracing.h"
 
 struct bpf_map_def SEC("maps") my_map = {
 	.type = BPF_MAP_TYPE_HASH,
samples/bpf/tracex4_kern.c +1 -0
index 6dd8e384de96..2a02cbe9d9a1 100644
@@ -8,6 +8,7 @@
 #include <linux/version.h>
 #include <uapi/linux/bpf.h>
 #include "bpf_helpers.h"
+#include "bpf_tracing.h"
 
 struct pair {
 	u64 val;
samples/bpf/tracex5_kern.c +1 -0
index 35cb0eed3be5..b3557b21a8fe 100644
@@ -11,6 +11,7 @@
 #include <uapi/linux/unistd.h>
 #include "syscall_nrs.h"
 #include "bpf_helpers.h"
+#include "bpf_tracing.h"
 
 #define PROG(F) SEC("kprobe/"__stringify(F)) int bpf_func_##F
 
tools/testing/selftests/bpf/bpf_helpers.h +0 -190

196 lines not on this page · raw message

tools/testing/selftests/bpf/bpf_tracing.h new file +195 -0

195 lines not on this page · raw message

tools/testing/selftests/bpf/progs/loop1.c +1 -0
index 7cdb7f878310..40ac722a9da5 100644
@@ -7,6 +7,7 @@
 #include <stdbool.h>
 #include <linux/bpf.h>
 #include "bpf_helpers.h"
+#include "bpf_tracing.h"
 
 char _license[] SEC("license") = "GPL";
 
tools/testing/selftests/bpf/progs/loop2.c +1 -0
index 9b2f808a2863..bb80f29aa7f7 100644
@@ -7,6 +7,7 @@
 #include <stdbool.h>
 #include <linux/bpf.h>
 #include "bpf_helpers.h"
+#include "bpf_tracing.h"
 
 char _license[] SEC("license") = "GPL";
 
tools/testing/selftests/bpf/progs/loop3.c +1 -0
index d727657d51e2..2b9165a7afe1 100644
@@ -7,6 +7,7 @@
 #include <stdbool.h>
 #include <linux/bpf.h>
 #include "bpf_helpers.h"
+#include "bpf_tracing.h"
 
 char _license[] SEC("license") = "GPL";
 
Signature · 1 line
2.17.1
#5 lore raw

PATCH v5 5/7 libbpf: move bpf_{helpers,helper_defs,endian,tracing}.h into libbpf

In reply to Andrii Nakryiko

Move bpf_helpers.h, bpf_tracing.h, and bpf_endian.h into libbpf. Move bpf_helper_defs.h generation into libbpf's Makefile. Ensure all those headers are installed along the other libbpf headers. Also, adjust selftests and samples include path to include libbpf now.

---
 samples/bpf/Makefile                             |  2 +-
 tools/lib/bpf/.gitignore                         |  1 +
 tools/lib/bpf/Makefile                           | 16 ++++++++++++----
 .../{testing/selftests => lib}/bpf/bpf_endian.h  |  0
 .../{testing/selftests => lib}/bpf/bpf_helpers.h |  0
 .../{testing/selftests => lib}/bpf/bpf_tracing.h |  0
 tools/testing/selftests/bpf/.gitignore           |  1 -
 tools/testing/selftests/bpf/Makefile             | 10 +++-------
 8 files changed, 17 insertions(+), 13 deletions(-)
 rename tools/{testing/selftests => lib}/bpf/bpf_endian.h (100%)
 rename tools/{testing/selftests => lib}/bpf/bpf_helpers.h (100%)
 rename tools/{testing/selftests => lib}/bpf/bpf_tracing.h (100%)
samples/bpf/Makefile +1 -1
index 42b571cde177..ecb3535d91e3 100644
@@ -283,7 +283,7 @@ $(obj)/hbm_edt_kern.o: $(src)/hbm.h $(src)/hbm_kern.h
 $(obj)/%.o: $(src)/%.c
 	@echo "  CLANG-bpf " $@
 	$(Q)$(CLANG) $(NOSTDINC_FLAGS) $(LINUXINCLUDE) $(EXTRA_CFLAGS) -I$(obj) \
-		-I$(srctree)/tools/testing/selftests/bpf/ \
+		-I$(srctree)/tools/testing/selftests/bpf/ -I$(srctree)/tools/lib/bpf/ \
 		-D__KERNEL__ -D__BPF_TRACING__ -Wno-unused-value -Wno-pointer-sign \
 		-D__TARGET_ARCH_$(SRCARCH) -Wno-compare-distinct-pointer-types \
 		-Wno-gnu-variable-sized-type-not-at-end \
tools/lib/bpf/.gitignore +1 -0
index 12382b0c71c7..35bf013e368c 100644
@@ -6,3 +6,4 @@ libbpf.so.*
 TAGS
 tags
 cscope.*
+/bpf_helper_defs.h
tools/lib/bpf/Makefile +12 -4
index 10b77644a17c..974453564f01 100644
@@ -157,7 +157,7 @@ all: fixdep
 
 all_cmd: $(CMD_TARGETS) check
 
-$(BPF_IN): force elfdep bpfdep
+$(BPF_IN): force elfdep bpfdep bpf_helper_defs.h
 	@(test -f ../../include/uapi/linux/bpf.h -a -f ../../../include/uapi/linux/bpf.h && ( \
 	(diff -B ../../include/uapi/linux/bpf.h ../../../include/uapi/linux/bpf.h >/dev/null) || \
 	echo "Warning: Kernel ABI header at 'tools/include/uapi/linux/bpf.h' differs from latest version at 'include/uapi/linux/bpf.h'" >&2 )) || true
@@ -175,6 +175,10 @@ $(BPF_IN): force elfdep bpfdep
 	echo "Warning: Kernel ABI header at 'tools/include/uapi/linux/if_xdp.h' differs from latest version at 'include/uapi/linux/if_xdp.h'" >&2 )) || true
 	$(Q)$(MAKE) $(build)=libbpf
 
+bpf_helper_defs.h: $(srctree)/include/uapi/linux/bpf.h
+	$(Q)$(srctree)/scripts/bpf_helpers_doc.py --header 		\
+		--file $(srctree)/include/uapi/linux/bpf.h > bpf_helper_defs.h
+
 $(OUTPUT)libbpf.so: $(OUTPUT)libbpf.so.$(LIBBPF_VERSION)
 
 $(OUTPUT)libbpf.so.$(LIBBPF_VERSION): $(BPF_IN)
@@ -236,13 +240,17 @@ install_lib: all_cmd
 		$(call do_install_mkdir,$(libdir_SQ)); \
 		cp -fpR $(LIB_FILE) $(DESTDIR)$(libdir_SQ)
 
-install_headers:
+install_headers: bpf_helper_defs.h
 	$(call QUIET_INSTALL, headers) \
 		$(call do_install,bpf.h,$(prefix)/include/bpf,644); \
 		$(call do_install,libbpf.h,$(prefix)/include/bpf,644); \
 		$(call do_install,btf.h,$(prefix)/include/bpf,644); \
 		$(call do_install,libbpf_util.h,$(prefix)/include/bpf,644); \
-		$(call do_install,xsk.h,$(prefix)/include/bpf,644);
+		$(call do_install,xsk.h,$(prefix)/include/bpf,644); \
+		$(call do_install,bpf_helpers.h,$(prefix)/include/bpf,644); \
+		$(call do_install,bpf_helper_defs.h,$(prefix)/include/bpf,644); \
+		$(call do_install,bpf_tracing.h,$(prefix)/include/bpf,644); \
+		$(call do_install,bpf_endian.h,$(prefix)/include/bpf,644);
 
 install_pkgconfig: $(PC_FILE)
 	$(call QUIET_INSTALL, $(PC_FILE)) \
@@ -259,7 +267,7 @@ config-clean:
 clean:
 	$(call QUIET_CLEAN, libbpf) $(RM) $(TARGETS) $(CXX_TEST_TARGET) \
 		*.o *~ *.a *.so *.so.$(LIBBPF_MAJOR_VERSION) .*.d .*.cmd \
-		*.pc LIBBPF-CFLAGS
+		*.pc LIBBPF-CFLAGS bpf_helper_defs.h
 	$(call QUIET_CLEAN, core-gen) $(RM) $(OUTPUT)FEATURE-DUMP.libbpf
 
 
tools/lib/bpf/bpf_endian.h renamed from tools/testing/selftests/bpf/bpf_endian.h +0 -0
similarity index 100%
rename from tools/testing/selftests/bpf/bpf_endian.h
rename to tools/lib/bpf/bpf_endian.h
tools/lib/bpf/bpf_helpers.h renamed from tools/testing/selftests/bpf/bpf_helpers.h +0 -0
similarity index 100%
rename from tools/testing/selftests/bpf/bpf_helpers.h
rename to tools/lib/bpf/bpf_helpers.h
tools/lib/bpf/bpf_tracing.h renamed from tools/testing/selftests/bpf/bpf_tracing.h +0 -0
similarity index 100%
rename from tools/testing/selftests/bpf/bpf_tracing.h
rename to tools/lib/bpf/bpf_tracing.h
tools/testing/selftests/bpf/.gitignore +0 -1
index 50063f66539d..7470327edcfe 100644
@@ -39,4 +39,3 @@ libbpf.so.*
 test_hashmap
 test_btf_dump
 xdping
-/bpf_helper_defs.h
tools/testing/selftests/bpf/Makefile +3 -7
index 771a4e82128b..90944b7a8274 100644
@@ -90,10 +90,6 @@ include ../lib.mk
 TEST_CUSTOM_PROGS = $(OUTPUT)/urandom_read
 all: $(TEST_CUSTOM_PROGS)
 
-bpf_helper_defs.h: $(APIDIR)/linux/bpf.h
-	$(BPFDIR)/../../../scripts/bpf_helpers_doc.py --header 		\
-		--file $(APIDIR)/linux/bpf.h > bpf_helper_defs.h
-
 $(OUTPUT)/urandom_read: $(OUTPUT)/%: %.c
 	$(CC) -o $@ $< -Wl,--build-id
 
@@ -127,7 +123,7 @@ $(OUTPUT)/test_cgroup_attach: cgroup_helpers.c
 # force a rebuild of BPFOBJ when its dependencies are updated
 force:
 
-$(BPFOBJ): force bpf_helper_defs.h
+$(BPFOBJ): force
 	$(MAKE) -C $(BPFDIR) OUTPUT=$(OUTPUT)/
 
 PROBE := $(shell $(LLC) -march=bpf -mcpu=probe -filetype=null /dev/null 2>&1)
@@ -152,7 +148,7 @@ $(shell $(1) -v -E - </dev/null 2>&1 \
 endef
 CLANG_SYS_INCLUDES = $(call get_sys_includes,$(CLANG))
 BPF_CFLAGS = -I. -I./include/uapi -I../../../include/uapi \
-	     -I$(OUTPUT)/../usr/include -D__TARGET_ARCH_$(SRCARCH)
+	     -I$(BPFDIR) -I$(OUTPUT)/../usr/include -D__TARGET_ARCH_$(SRCARCH)
 
 CLANG_CFLAGS = $(CLANG_SYS_INCLUDES) \
 	       -Wno-compare-distinct-pointer-types
@@ -323,4 +319,4 @@ $(VERIFIER_TESTS_H): $(VERIFIER_TEST_FILES) | $(VERIFIER_TESTS_DIR)
 
 EXTRA_CLEAN := $(TEST_CUSTOM_PROGS) $(ALU32_BUILD_DIR) $(BPF_GCC_BUILD_DIR) \
 	$(VERIFIER_TESTS_H) $(PROG_TESTS_H) $(MAP_TESTS_H) \
-	feature bpf_helper_defs.h
+	feature
Signature · 1 line
2.17.1
#6 lore raw

PATCH v5 6/7 libbpf: add BPF_CORE_READ/BPF_CORE_READ_INTO helpers

In reply to Andrii Nakryiko

Add few macros simplifying BCC-like multi-level probe reads, while also emitting CO-RE relocations for each read.

---
 tools/lib/bpf/Makefile                        |   3 +-
 tools/lib/bpf/bpf_core_read.h                 | 167 ++++++++++++++++++
 tools/lib/bpf/bpf_helpers.h                   |  32 +---
 .../bpf/progs/test_core_reloc_arrays.c        |   1 +
 .../bpf/progs/test_core_reloc_flavors.c       |   1 +
 .../bpf/progs/test_core_reloc_ints.c          |   1 +
 .../bpf/progs/test_core_reloc_kernel.c        |   1 +
 .../bpf/progs/test_core_reloc_misc.c          |   1 +
 .../bpf/progs/test_core_reloc_mods.c          |   1 +
 .../bpf/progs/test_core_reloc_nesting.c       |   1 +
 .../bpf/progs/test_core_reloc_primitives.c    |   1 +
 .../bpf/progs/test_core_reloc_ptr_as_arr.c    |   1 +
 12 files changed, 187 insertions(+), 24 deletions(-)
 create mode 100644 tools/lib/bpf/bpf_core_read.h
tools/lib/bpf/Makefile +2 -1
index 974453564f01..1270955e4845 100644
@@ -250,7 +250,8 @@ install_headers: bpf_helper_defs.h
 		$(call do_install,bpf_helpers.h,$(prefix)/include/bpf,644); \
 		$(call do_install,bpf_helper_defs.h,$(prefix)/include/bpf,644); \
 		$(call do_install,bpf_tracing.h,$(prefix)/include/bpf,644); \
-		$(call do_install,bpf_endian.h,$(prefix)/include/bpf,644);
+		$(call do_install,bpf_endian.h,$(prefix)/include/bpf,644); \
+		$(call do_install,bpf_core_read.h,$(prefix)/include/bpf,644);
 
 install_pkgconfig: $(PC_FILE)
 	$(call QUIET_INSTALL, $(PC_FILE)) \
tools/lib/bpf/bpf_core_read.h new file +167 -0

167 lines not on this page · raw message

tools/lib/bpf/bpf_helpers.h +9 -23
index 6d059c0a7845..2203595f38c3 100644
@@ -7,7 +7,7 @@
 #define __uint(name, val) int (*name)[val]
 #define __type(name, val) typeof(val) *name
 
-/* helper macro to print out debug messages */
+/* Helper macro to print out debug messages */
 #define bpf_printk(fmt, ...)				\
 ({							\
 	char ____fmt[] = fmt;				\
@@ -15,13 +15,19 @@
 			 ##__VA_ARGS__);		\
 })
 
-/* helper macro to place programs, maps, license in
+/*
+ * Helper macro to place programs, maps, license in
  * different sections in elf_bpf file. Section names
  * are interpreted by elf_bpf loader
  */
 #define SEC(NAME) __attribute__((section(NAME), used))
 
-/* a helper structure used by eBPF C program
+#ifndef __always_inline
+#define __always_inline __attribute__((always_inline))
+#endif
+
+/*
+ * Helper structure used by eBPF C program
  * to describe BPF map attributes to libbpf loader
  */
 struct bpf_map_def {
@@ -32,24 +38,4 @@ struct bpf_map_def {
 	unsigned int map_flags;
 };
 
-/*
- * bpf_core_read() abstracts away bpf_probe_read() call and captures offset
- * relocation for source address using __builtin_preserve_access_index()
- * built-in, provided by Clang.
- *
- * __builtin_preserve_access_index() takes as an argument an expression of
- * taking an address of a field within struct/union. It makes compiler emit
- * a relocation, which records BTF type ID describing root struct/union and an
- * accessor string which describes exact embedded field that was used to take
- * an address. See detailed description of this relocation format and
- * semantics in comments to struct bpf_offset_reloc in libbpf_internal.h.
- *
- * This relocation allows libbpf to adjust BPF instruction to use correct
- * actual field offset, based on target kernel BTF type that matches original
- * (local) BTF, used to record relocation.
- */
-#define bpf_core_read(dst, sz, src)					    \
-	bpf_probe_read(dst, sz,						    \
-		       (const void *)__builtin_preserve_access_index(src))
-
 #endif
tools/testing/selftests/bpf/progs/test_core_reloc_arrays.c +1 -0
index 8e3f6e6a90e7..96b1f5f3b07a 100644
@@ -4,6 +4,7 @@
 #include <linux/bpf.h>
 #include <stdint.h>
 #include "bpf_helpers.h"
+#include "bpf_core_read.h"
 
 char _license[] SEC("license") = "GPL";
 
tools/testing/selftests/bpf/progs/test_core_reloc_flavors.c +1 -0
index 613474a18b45..71fd7cebc9d7 100644
@@ -4,6 +4,7 @@
 #include <linux/bpf.h>
 #include <stdint.h>
 #include "bpf_helpers.h"
+#include "bpf_core_read.h"
 
 char _license[] SEC("license") = "GPL";
 
tools/testing/selftests/bpf/progs/test_core_reloc_ints.c +1 -0
index 7a88a3975455..ad5c3f59c9c6 100644
@@ -4,6 +4,7 @@
 #include <linux/bpf.h>
 #include <stdint.h>
 #include "bpf_helpers.h"
+#include "bpf_core_read.h"
 
 char _license[] SEC("license") = "GPL";
 
tools/testing/selftests/bpf/progs/test_core_reloc_kernel.c +1 -0
index 684a06cf41ea..14ce688463de 100644
@@ -4,6 +4,7 @@
 #include <linux/bpf.h>
 #include <stdint.h>
 #include "bpf_helpers.h"
+#include "bpf_core_read.h"
 
 char _license[] SEC("license") = "GPL";
 
tools/testing/selftests/bpf/progs/test_core_reloc_misc.c +1 -0
index 10bdb2050552..1a36b0856653 100644
@@ -4,6 +4,7 @@
 #include <linux/bpf.h>
 #include <stdint.h>
 #include "bpf_helpers.h"
+#include "bpf_core_read.h"
 
 char _license[] SEC("license") = "GPL";
 
tools/testing/selftests/bpf/progs/test_core_reloc_mods.c +1 -0
index e930e7e88c5c..3199fafede2c 100644
@@ -4,6 +4,7 @@
 #include <linux/bpf.h>
 #include <stdint.h>
 #include "bpf_helpers.h"
+#include "bpf_core_read.h"
 
 char _license[] SEC("license") = "GPL";
 
tools/testing/selftests/bpf/progs/test_core_reloc_nesting.c +1 -0
index b63007958290..98238cb64fbd 100644
@@ -4,6 +4,7 @@
 #include <linux/bpf.h>
 #include <stdint.h>
 #include "bpf_helpers.h"
+#include "bpf_core_read.h"
 
 char _license[] SEC("license") = "GPL";
 
tools/testing/selftests/bpf/progs/test_core_reloc_primitives.c +1 -0
index 7654f59914bc..4f3ecb9127bb 100644
@@ -4,6 +4,7 @@
 #include <linux/bpf.h>
 #include <stdint.h>
 #include "bpf_helpers.h"
+#include "bpf_core_read.h"
 
 char _license[] SEC("license") = "GPL";
 
tools/testing/selftests/bpf/progs/test_core_reloc_ptr_as_arr.c +1 -0
index 709f7cba453f..27f602f00419 100644
@@ -4,6 +4,7 @@
 #include <linux/bpf.h>
 #include <stdint.h>
 #include "bpf_helpers.h"
+#include "bpf_core_read.h"
 
 char _license[] SEC("license") = "GPL";
 
Signature · 1 line
2.17.1
#7 lore raw

PATCH v5 7/7 selftests/bpf: add BPF_CORE_READ and BPF_CORE_READ_STR_INTO macro tests

In reply to Andrii Nakryiko

Validate BPF_CORE_READ correctness and handling of up to 9 levels of nestedness using cyclic task->(group_leader->)*->tgid chains.

Also add a test of maximum-dpeth BPF_CORE_READ_STR_INTO() macro.

---
 .../selftests/bpf/prog_tests/core_reloc.c     |  8 ++-
 .../selftests/bpf/progs/core_reloc_types.h    |  9 ++++
 .../bpf/progs/test_core_reloc_kernel.c        | 54 ++++++++++++++++++-
 3 files changed, 68 insertions(+), 3 deletions(-)
tools/testing/selftests/bpf/prog_tests/core_reloc.c +6 -2
index f3863f976a48..21a0dff66241 100644
@@ -193,8 +193,12 @@ static struct core_reloc_test_case test_cases[] = {
 		.btf_src_file = NULL, /* load from /lib/modules/$(uname -r) */
 		.input = "",
 		.input_len = 0,
-		.output = "\1", /* true */
-		.output_len = 1,
+		.output = STRUCT_TO_CHAR_PTR(core_reloc_kernel_output) {
+			.valid = { 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, },
+			.comm = "test_progs\0\0\0\0\0",
+			.comm_len = 11,
+		},
+		.output_len = sizeof(struct core_reloc_kernel_output),
 	},
 
 	/* validate BPF program can use multiple flavors to match against
tools/testing/selftests/bpf/progs/core_reloc_types.h +9 -0
index f686a8138d90..9a6bdeb4894c 100644
@@ -1,5 +1,14 @@
 #include <stdint.h>
 #include <stdbool.h>
+/*
+ * KERNEL
+ */
+
+struct core_reloc_kernel_output {
+	int valid[10];
+	char comm[16];
+	int comm_len;
+};
 
 /*
  * FLAVORS
tools/testing/selftests/bpf/progs/test_core_reloc_kernel.c +53 -1

76 lines not on this page · raw message

Signature · 1 line
2.17.1
maintainer Applied
#8 lore raw
In reply to Andrii Nakryiko
Quoted 7 lines · Andrii Nakryiko

This patch set makes bpf_helpers.h and bpf_endian.h a part of libbpf itself for consumption by user BPF programs, not just selftests. It also splits off tracing helpers into bpf_tracing.h, which also becomes part of libbpf. Some of the legacy stuff (BPF_ANNOTATE_KV_PAIR, load_{byte,half,word}, bpf_map_def with unsupported fields, etc, is extracted into selftests-only bpf_legacy.h. All the selftests and samples are switched to use libbpf's headers and selftests' ones are removed.

[...]

Applied, thanks!