bpf.h 73.4 KB
Newer Older
1
/* SPDX-License-Identifier: GPL-2.0-only */
2 3 4 5 6 7
/* Copyright (c) 2011-2014 PLUMgrid, http://plumgrid.com
 */
#ifndef _LINUX_BPF_H
#define _LINUX_BPF_H 1

#include <uapi/linux/bpf.h>
8

9
#include <linux/workqueue.h>
10
#include <linux/file.h>
11
#include <linux/percpu.h>
Z
Zi Shen Lim 已提交
12
#include <linux/err.h>
13
#include <linux/rbtree_latch.h>
14
#include <linux/numa.h>
15
#include <linux/mm_types.h>
16
#include <linux/wait.h>
A
Alexei Starovoitov 已提交
17 18
#include <linux/refcount.h>
#include <linux/mutex.h>
19
#include <linux/module.h>
J
Jiri Olsa 已提交
20
#include <linux/kallsyms.h>
A
Alexei Starovoitov 已提交
21
#include <linux/capability.h>
22 23
#include <linux/sched/mm.h>
#include <linux/slab.h>
A
Alexei Starovoitov 已提交
24
#include <linux/percpu-refcount.h>
25
#include <linux/bpfptr.h>
26

27
struct bpf_verifier_env;
28
struct bpf_verifier_log;
29
struct perf_event;
30
struct bpf_prog;
31
struct bpf_prog_aux;
32
struct bpf_map;
33
struct sock;
34
struct seq_file;
35
struct btf;
36
struct btf_type;
37
struct exception_table_entry;
38
struct seq_operations;
39
struct bpf_iter_aux_info;
K
KP Singh 已提交
40 41
struct bpf_local_storage;
struct bpf_local_storage_map;
42
struct kobject;
43
struct mem_cgroup;
44
struct module;
45
struct bpf_func_state;
46

47 48
extern struct idr btf_idr;
extern spinlock_t btf_idr_lock;
49
extern struct kobject *btf_kobj;
50

51
typedef u64 (*bpf_callback_t)(u64, u64, u64, u64, u64);
52 53
typedef int (*bpf_iter_init_seq_priv_t)(void *private_data,
					struct bpf_iter_aux_info *aux);
54 55 56 57 58 59 60 61
typedef void (*bpf_iter_fini_seq_priv_t)(void *private_data);
struct bpf_iter_seq_info {
	const struct seq_operations *seq_ops;
	bpf_iter_init_seq_priv_t init_seq_private;
	bpf_iter_fini_seq_priv_t fini_seq_private;
	u32 seq_priv_size;
};

62
/* map is generic key/value storage optionally accessible by eBPF programs */
63 64
struct bpf_map_ops {
	/* funcs callable from userspace (via syscall) */
65
	int (*map_alloc_check)(union bpf_attr *attr);
66
	struct bpf_map *(*map_alloc)(union bpf_attr *attr);
67 68
	void (*map_release)(struct bpf_map *map, struct file *map_file);
	void (*map_free)(struct bpf_map *map);
69
	int (*map_get_next_key)(struct bpf_map *map, void *key, void *next_key);
70
	void (*map_release_uref)(struct bpf_map *map);
71
	void *(*map_lookup_elem_sys_only)(struct bpf_map *map, void *key);
72 73
	int (*map_lookup_batch)(struct bpf_map *map, const union bpf_attr *attr,
				union bpf_attr __user *uattr);
74 75
	int (*map_lookup_and_delete_elem)(struct bpf_map *map, void *key,
					  void *value, u64 flags);
76 77 78
	int (*map_lookup_and_delete_batch)(struct bpf_map *map,
					   const union bpf_attr *attr,
					   union bpf_attr __user *uattr);
79 80 81 82
	int (*map_update_batch)(struct bpf_map *map, const union bpf_attr *attr,
				union bpf_attr __user *uattr);
	int (*map_delete_batch)(struct bpf_map *map, const union bpf_attr *attr,
				union bpf_attr __user *uattr);
83 84 85

	/* funcs callable from userspace and from eBPF programs */
	void *(*map_lookup_elem)(struct bpf_map *map, void *key);
86
	int (*map_update_elem)(struct bpf_map *map, void *key, void *value, u64 flags);
87
	int (*map_delete_elem)(struct bpf_map *map, void *key);
M
Mauricio Vasquez B 已提交
88 89 90
	int (*map_push_elem)(struct bpf_map *map, void *value, u64 flags);
	int (*map_pop_elem)(struct bpf_map *map, void *value);
	int (*map_peek_elem)(struct bpf_map *map, void *value);
91 92

	/* funcs called by prog_array and perf_event_array map */
93 94 95
	void *(*map_fd_get_ptr)(struct bpf_map *map, struct file *map_file,
				int fd);
	void (*map_fd_put_ptr)(void *ptr);
96
	int (*map_gen_lookup)(struct bpf_map *map, struct bpf_insn *insn_buf);
97
	u32 (*map_fd_sys_lookup_elem)(void *ptr);
98 99
	void (*map_seq_show_elem)(struct bpf_map *map, void *key,
				  struct seq_file *m);
100
	int (*map_check_btf)(const struct bpf_map *map,
101
			     const struct btf *btf,
102 103
			     const struct btf_type *key_type,
			     const struct btf_type *value_type);
104

105 106 107 108 109 110
	/* Prog poke tracking helpers. */
	int (*map_poke_track)(struct bpf_map *map, struct bpf_prog_aux *aux);
	void (*map_poke_untrack)(struct bpf_map *map, struct bpf_prog_aux *aux);
	void (*map_poke_run)(struct bpf_map *map, u32 key, struct bpf_prog *old,
			     struct bpf_prog *new);

111 112 113 114 115
	/* Direct value access helpers. */
	int (*map_direct_value_addr)(const struct bpf_map *map,
				     u64 *imm, u32 off);
	int (*map_direct_value_meta)(const struct bpf_map *map,
				     u64 imm, u32 *off);
116
	int (*map_mmap)(struct bpf_map *map, struct vm_area_struct *vma);
117 118
	__poll_t (*map_poll)(struct bpf_map *map, struct file *filp,
			     struct poll_table_struct *pts);
119

K
KP Singh 已提交
120 121 122 123 124 125
	/* Functions called by bpf_local_storage maps */
	int (*map_local_storage_charge)(struct bpf_local_storage_map *smap,
					void *owner, u32 size);
	void (*map_local_storage_uncharge)(struct bpf_local_storage_map *smap,
					   void *owner, u32 size);
	struct bpf_local_storage __rcu ** (*map_owner_storage_ptr)(void *owner);
126

127 128 129
	/* Misc helpers.*/
	int (*map_redirect)(struct bpf_map *map, u32 ifindex, u64 flags);

130 131 132 133 134 135 136 137 138 139 140 141
	/* map_meta_equal must be implemented for maps that can be
	 * used as an inner map.  It is a runtime check to ensure
	 * an inner map can be inserted to an outer map.
	 *
	 * Some properties of the inner map has been used during the
	 * verification time.  When inserting an inner map at the runtime,
	 * map_meta_equal has to ensure the inserting map has the same
	 * properties that the verifier has used earlier.
	 */
	bool (*map_meta_equal)(const struct bpf_map *meta0,
			       const struct bpf_map *meta1);

142 143 144 145

	int (*map_set_for_each_callback_args)(struct bpf_verifier_env *env,
					      struct bpf_func_state *caller,
					      struct bpf_func_state *callee);
146 147
	int (*map_for_each_callback)(struct bpf_map *map,
				     bpf_callback_t callback_fn,
148 149
				     void *callback_ctx, u64 flags);

150 151 152
	/* BTF name and id of struct allocated by map_alloc */
	const char * const map_btf_name;
	int *map_btf_id;
153 154 155

	/* bpf_iter info used to open a seq_file */
	const struct bpf_iter_seq_info *iter_seq_info;
156 157 158
};

struct bpf_map {
159
	/* The first two cachelines with read-mostly members of which some
160 161 162 163 164 165 166
	 * are also accessed in fast-path (e.g. ops, max_entries).
	 */
	const struct bpf_map_ops *ops ____cacheline_aligned;
	struct bpf_map *inner_map_meta;
#ifdef CONFIG_SECURITY
	void *security;
#endif
167 168 169 170
	enum bpf_map_type map_type;
	u32 key_size;
	u32 value_size;
	u32 max_entries;
171
	u64 map_extra; /* any per-map-type extra fields */
172
	u32 map_flags;
173
	int spin_lock_off; /* >=0 valid offset, <0 error */
A
Alexei Starovoitov 已提交
174
	int timer_off; /* >=0 valid offset, <0 error */
M
Martin KaFai Lau 已提交
175
	u32 id;
176
	int numa_node;
177 178
	u32 btf_key_type_id;
	u32 btf_value_type_id;
179
	u32 btf_vmlinux_value_type_id;
180
	struct btf *btf;
181 182 183
#ifdef CONFIG_MEMCG_KMEM
	struct mem_cgroup *memcg;
#endif
184
	char name[BPF_OBJ_NAME_LEN];
A
Alexei Starovoitov 已提交
185
	bool bypass_spec_v1;
186
	bool frozen; /* write-once; write-protected by freeze_mutex */
187
	/* 14 bytes hole */
188

189
	/* The 3rd and 4th cacheline with misc members to avoid false sharing
190 191
	 * particularly with refcounting.
	 */
192 193
	atomic64_t refcnt ____cacheline_aligned;
	atomic64_t usercnt;
194
	struct work_struct work;
195
	struct mutex freeze_mutex;
196
	atomic64_t writecnt;
197 198
};

199 200 201 202 203
static inline bool map_value_has_spin_lock(const struct bpf_map *map)
{
	return map->spin_lock_off >= 0;
}

204
static inline bool map_value_has_timer(const struct bpf_map *map)
205
{
206
	return map->timer_off >= 0;
207 208
}

209 210 211 212 213 214 215 216 217 218 219
static inline void check_and_init_map_value(struct bpf_map *map, void *dst)
{
	if (unlikely(map_value_has_spin_lock(map)))
		*(struct bpf_spin_lock *)(dst + map->spin_lock_off) =
			(struct bpf_spin_lock){};
	if (unlikely(map_value_has_timer(map)))
		*(struct bpf_timer *)(dst + map->timer_off) =
			(struct bpf_timer){};
}

/* copy everything but bpf_spin_lock and bpf_timer. There could be one of each. */
220 221
static inline void copy_map_value(struct bpf_map *map, void *dst, void *src)
{
222 223
	u32 s_off = 0, s_sz = 0, t_off = 0, t_sz = 0;

224
	if (unlikely(map_value_has_spin_lock(map))) {
225 226
		s_off = map->spin_lock_off;
		s_sz = sizeof(struct bpf_spin_lock);
227 228
	}
	if (unlikely(map_value_has_timer(map))) {
229 230 231
		t_off = map->timer_off;
		t_sz = sizeof(struct bpf_timer);
	}
232

233 234 235 236 237 238 239 240 241 242 243 244
	if (unlikely(s_sz || t_sz)) {
		if (s_off < t_off || !s_sz) {
			swap(s_off, t_off);
			swap(s_sz, t_sz);
		}
		memcpy(dst, src, t_off);
		memcpy(dst + t_off + t_sz,
		       src + t_off + t_sz,
		       s_off - t_off - t_sz);
		memcpy(dst + s_off + s_sz,
		       src + s_off + s_sz,
		       map->value_size - s_off - s_sz);
245 246 247 248
	} else {
		memcpy(dst, src, map->value_size);
	}
}
249 250
void copy_map_value_locked(struct bpf_map *map, void *dst, void *src,
			   bool lock_src);
A
Alexei Starovoitov 已提交
251
void bpf_timer_cancel_and_free(void *timer);
252
int bpf_obj_name_cpy(char *dst, const char *src, unsigned int size);
253

254
struct bpf_offload_dev;
255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279
struct bpf_offloaded_map;

struct bpf_map_dev_ops {
	int (*map_get_next_key)(struct bpf_offloaded_map *map,
				void *key, void *next_key);
	int (*map_lookup_elem)(struct bpf_offloaded_map *map,
			       void *key, void *value);
	int (*map_update_elem)(struct bpf_offloaded_map *map,
			       void *key, void *value, u64 flags);
	int (*map_delete_elem)(struct bpf_offloaded_map *map, void *key);
};

struct bpf_offloaded_map {
	struct bpf_map map;
	struct net_device *netdev;
	const struct bpf_map_dev_ops *dev_ops;
	void *dev_priv;
	struct list_head offloads;
};

static inline struct bpf_offloaded_map *map_to_offmap(struct bpf_map *map)
{
	return container_of(map, struct bpf_offloaded_map, map);
}

280 281 282 283 284
static inline bool bpf_map_offload_neutral(const struct bpf_map *map)
{
	return map->map_type == BPF_MAP_TYPE_PERF_EVENT_ARRAY;
}

285 286
static inline bool bpf_map_support_seq_show(const struct bpf_map *map)
{
287 288
	return (map->btf_value_type_id || map->btf_vmlinux_value_type_id) &&
		map->ops->map_seq_show_elem;
289 290
}

291
int map_check_no_btf(const struct bpf_map *map,
292
		     const struct btf *btf,
293 294 295
		     const struct btf_type *key_type,
		     const struct btf_type *value_type);

296 297 298
bool bpf_map_meta_equal(const struct bpf_map *meta0,
			const struct bpf_map *meta1);

299 300
extern const struct bpf_map_ops bpf_map_offload_ops;

301 302 303 304 305 306 307 308 309 310 311 312 313 314
/* bpf_type_flag contains a set of flags that are applicable to the values of
 * arg_type, ret_type and reg_type. For example, a pointer value may be null,
 * or a memory is read-only. We classify types into two categories: base types
 * and extended types. Extended types are base types combined with a type flag.
 *
 * Currently there are no more than 32 base types in arg_type, ret_type and
 * reg_types.
 */
#define BPF_BASE_TYPE_BITS	8

enum bpf_type_flag {
	/* PTR may be NULL. */
	PTR_MAYBE_NULL		= BIT(0 + BPF_BASE_TYPE_BITS),

315 316 317
	/* MEM is read-only. When applied on bpf_arg, it indicates the arg is
	 * compatible with both mutable and immutable memory.
	 */
H
Hao Luo 已提交
318 319
	MEM_RDONLY		= BIT(1 + BPF_BASE_TYPE_BITS),

320 321 322 323 324 325
	/* MEM was "allocated" from a different helper, and cannot be mixed
	 * with regular non-MEM_ALLOC'ed MEM types.
	 */
	MEM_ALLOC		= BIT(2 + BPF_BASE_TYPE_BITS),

	__BPF_TYPE_LAST_FLAG	= MEM_ALLOC,
326 327 328 329 330 331 332 333
};

/* Max number of base types. */
#define BPF_BASE_TYPE_LIMIT	(1UL << BPF_BASE_TYPE_BITS)

/* Max number of all types. */
#define BPF_TYPE_LIMIT		(__BPF_TYPE_LAST_FLAG | (__BPF_TYPE_LAST_FLAG - 1))

334 335
/* function argument constraints */
enum bpf_arg_type {
336
	ARG_DONTCARE = 0,	/* unused argument in helper function */
337 338 339 340 341 342 343

	/* the following constraints used to prototype
	 * bpf_map_lookup/update/delete_elem() functions
	 */
	ARG_CONST_MAP_PTR,	/* const argument used as pointer to bpf_map */
	ARG_PTR_TO_MAP_KEY,	/* pointer to stack used as map key */
	ARG_PTR_TO_MAP_VALUE,	/* pointer to stack used as map value */
344
	ARG_PTR_TO_UNINIT_MAP_VALUE,	/* pointer to valid memory used to store a map value */
345 346 347 348

	/* the following constraints used to prototype bpf_memcmp() and other
	 * functions that access data on eBPF program stack
	 */
349 350 351 352
	ARG_PTR_TO_MEM,		/* pointer to valid memory (stack, packet, map value) */
	ARG_PTR_TO_UNINIT_MEM,	/* pointer to memory does not need to be initialized,
				 * helper function must fill all bytes or clear
				 * them in error case.
353 354
				 */

355 356
	ARG_CONST_SIZE,		/* number of bytes accessed from memory */
	ARG_CONST_SIZE_OR_ZERO,	/* number of bytes accessed from memory or 0 */
357

358
	ARG_PTR_TO_CTX,		/* pointer to context */
359
	ARG_ANYTHING,		/* any (initialized) argument is ok */
360
	ARG_PTR_TO_SPIN_LOCK,	/* pointer to bpf_spin_lock */
361
	ARG_PTR_TO_SOCK_COMMON,	/* pointer to sock_common */
362 363
	ARG_PTR_TO_INT,		/* pointer to int */
	ARG_PTR_TO_LONG,	/* pointer to long */
364
	ARG_PTR_TO_SOCKET,	/* pointer to bpf_sock (fullsock) */
365
	ARG_PTR_TO_BTF_ID,	/* pointer to in-kernel struct */
366 367
	ARG_PTR_TO_ALLOC_MEM,	/* pointer to dynamically allocated memory */
	ARG_CONST_ALLOC_SIZE_OR_ZERO,	/* number of allocated bytes requested */
368
	ARG_PTR_TO_BTF_ID_SOCK_COMMON,	/* pointer to in-kernel sock_common or bpf-mirrored bpf_sock */
H
Hao Luo 已提交
369
	ARG_PTR_TO_PERCPU_BTF_ID,	/* pointer to in-kernel percpu type */
370
	ARG_PTR_TO_FUNC,	/* pointer to a bpf program function */
371
	ARG_PTR_TO_STACK,	/* pointer to stack */
372
	ARG_PTR_TO_CONST_STR,	/* pointer to a null terminated read-only string */
A
Alexei Starovoitov 已提交
373
	ARG_PTR_TO_TIMER,	/* pointer to bpf_timer */
374
	__BPF_ARG_TYPE_MAX,
375

376 377 378 379 380 381 382 383
	/* Extended arg_types. */
	ARG_PTR_TO_MAP_VALUE_OR_NULL	= PTR_MAYBE_NULL | ARG_PTR_TO_MAP_VALUE,
	ARG_PTR_TO_MEM_OR_NULL		= PTR_MAYBE_NULL | ARG_PTR_TO_MEM,
	ARG_PTR_TO_CTX_OR_NULL		= PTR_MAYBE_NULL | ARG_PTR_TO_CTX,
	ARG_PTR_TO_SOCKET_OR_NULL	= PTR_MAYBE_NULL | ARG_PTR_TO_SOCKET,
	ARG_PTR_TO_ALLOC_MEM_OR_NULL	= PTR_MAYBE_NULL | ARG_PTR_TO_ALLOC_MEM,
	ARG_PTR_TO_STACK_OR_NULL	= PTR_MAYBE_NULL | ARG_PTR_TO_STACK,

384 385 386 387
	/* This must be the last entry. Its purpose is to ensure the enum is
	 * wide enough to hold the higher bits reserved for bpf_type_flag.
	 */
	__BPF_ARG_TYPE_LIMIT	= BPF_TYPE_LIMIT,
388
};
389
static_assert(__BPF_ARG_TYPE_MAX <= BPF_BASE_TYPE_LIMIT);
390 391 392 393 394

/* type of values returned from helper functions */
enum bpf_return_type {
	RET_INTEGER,			/* function returns integer */
	RET_VOID,			/* function doesn't return anything */
395
	RET_PTR_TO_MAP_VALUE,		/* returns a pointer to map elem value */
396 397 398 399
	RET_PTR_TO_SOCKET,		/* returns a pointer to a socket */
	RET_PTR_TO_TCP_SOCK,		/* returns a pointer to a tcp_sock */
	RET_PTR_TO_SOCK_COMMON,		/* returns a pointer to a sock_common */
	RET_PTR_TO_ALLOC_MEM,		/* returns a pointer to dynamically allocated memory */
H
Hao Luo 已提交
400
	RET_PTR_TO_MEM_OR_BTF_ID,	/* returns a pointer to a valid memory or a btf_id */
401
	RET_PTR_TO_BTF_ID,		/* returns a pointer to a btf_id */
402 403
	__BPF_RET_TYPE_MAX,

404 405 406 407 408
	/* Extended ret_types. */
	RET_PTR_TO_MAP_VALUE_OR_NULL	= PTR_MAYBE_NULL | RET_PTR_TO_MAP_VALUE,
	RET_PTR_TO_SOCKET_OR_NULL	= PTR_MAYBE_NULL | RET_PTR_TO_SOCKET,
	RET_PTR_TO_TCP_SOCK_OR_NULL	= PTR_MAYBE_NULL | RET_PTR_TO_TCP_SOCK,
	RET_PTR_TO_SOCK_COMMON_OR_NULL	= PTR_MAYBE_NULL | RET_PTR_TO_SOCK_COMMON,
409
	RET_PTR_TO_ALLOC_MEM_OR_NULL	= PTR_MAYBE_NULL | MEM_ALLOC | RET_PTR_TO_ALLOC_MEM,
410 411
	RET_PTR_TO_BTF_ID_OR_NULL	= PTR_MAYBE_NULL | RET_PTR_TO_BTF_ID,

412 413 414 415
	/* This must be the last entry. Its purpose is to ensure the enum is
	 * wide enough to hold the higher bits reserved for bpf_type_flag.
	 */
	__BPF_RET_TYPE_LIMIT	= BPF_TYPE_LIMIT,
416
};
417
static_assert(__BPF_RET_TYPE_MAX <= BPF_BASE_TYPE_LIMIT);
418

419 420 421 422 423 424 425
/* eBPF function prototype used by verifier to allow BPF_CALLs from eBPF programs
 * to in-kernel helper functions and for adjusting imm32 field in BPF_CALL
 * instructions after verifying
 */
struct bpf_func_proto {
	u64 (*func)(u64 r1, u64 r2, u64 r3, u64 r4, u64 r5);
	bool gpl_only;
426
	bool pkt_access;
427
	enum bpf_return_type ret_type;
428 429 430 431 432 433 434 435 436 437
	union {
		struct {
			enum bpf_arg_type arg1_type;
			enum bpf_arg_type arg2_type;
			enum bpf_arg_type arg3_type;
			enum bpf_arg_type arg4_type;
			enum bpf_arg_type arg5_type;
		};
		enum bpf_arg_type arg_type[5];
	};
438 439 440 441 442 443 444 445 446 447
	union {
		struct {
			u32 *arg1_btf_id;
			u32 *arg2_btf_id;
			u32 *arg3_btf_id;
			u32 *arg4_btf_id;
			u32 *arg5_btf_id;
		};
		u32 *arg_btf_id[5];
	};
448
	int *ret_btf_id; /* return value btf_id */
J
Jiri Olsa 已提交
449
	bool (*allowed)(const struct bpf_prog *prog);
450 451 452 453 454 455 456 457 458 459 460
};

/* bpf_context is intentionally undefined structure. Pointer to bpf_context is
 * the first argument to eBPF programs.
 * For socket filters: 'struct bpf_context *' == 'struct sk_buff *'
 */
struct bpf_context;

enum bpf_access_type {
	BPF_READ = 1,
	BPF_WRITE = 2
461 462
};

463
/* types of values stored in eBPF registers */
464 465 466 467 468 469 470 471 472
/* Pointer types represent:
 * pointer
 * pointer + imm
 * pointer + (u16) var
 * pointer + (u16) var + imm
 * if (range > 0) then [ptr, ptr + range - off) is safe to access
 * if (id > 0) means that some 'var' was added
 * if (off > 0) means that 'imm' was added
 */
473 474
enum bpf_reg_type {
	NOT_INIT = 0,		 /* nothing was written into register */
475
	SCALAR_VALUE,		 /* reg doesn't contain a valid pointer */
476 477 478
	PTR_TO_CTX,		 /* reg points to bpf_context */
	CONST_PTR_TO_MAP,	 /* reg points to struct bpf_map */
	PTR_TO_MAP_VALUE,	 /* reg points to map element value */
479
	PTR_TO_MAP_KEY,		 /* reg points to a map element key */
480
	PTR_TO_STACK,		 /* reg == frame_pointer + offset */
481
	PTR_TO_PACKET_META,	 /* skb->data - meta_len */
482
	PTR_TO_PACKET,		 /* reg points to skb->data */
483
	PTR_TO_PACKET_END,	 /* skb->data + headlen */
484
	PTR_TO_FLOW_KEYS,	 /* reg points to bpf_flow_keys */
485
	PTR_TO_SOCKET,		 /* reg points to struct bpf_sock */
486
	PTR_TO_SOCK_COMMON,	 /* reg points to sock_common */
487
	PTR_TO_TCP_SOCK,	 /* reg points to struct tcp_sock */
488
	PTR_TO_TP_BUFFER,	 /* reg points to a writable raw tp's buffer */
489
	PTR_TO_XDP_SOCK,	 /* reg points to struct xdp_sock */
490 491 492 493 494 495 496 497 498 499 500 501 502 503 504
	/* PTR_TO_BTF_ID points to a kernel struct that does not need
	 * to be null checked by the BPF program. This does not imply the
	 * pointer is _not_ null and in practice this can easily be a null
	 * pointer when reading pointer chains. The assumption is program
	 * context will handle null pointer dereference typically via fault
	 * handling. The verifier must keep this in mind and can make no
	 * assumptions about null or non-null when doing branch analysis.
	 * Further, when passed into helpers the helpers can not, without
	 * additional context, assume the value is non-null.
	 */
	PTR_TO_BTF_ID,
	/* PTR_TO_BTF_ID_OR_NULL points to a kernel struct that has not
	 * been checked for null. Used primarily to inform the verifier
	 * an explicit null check is required for this struct.
	 */
505
	PTR_TO_MEM,		 /* reg points to valid memory region */
H
Hao Luo 已提交
506
	PTR_TO_BUF,		 /* reg points to a read/write buffer */
H
Hao Luo 已提交
507
	PTR_TO_PERCPU_BTF_ID,	 /* reg points to a percpu kernel variable */
508
	PTR_TO_FUNC,		 /* reg points to a bpf program function */
509
	__BPF_REG_TYPE_MAX,
510

511 512 513 514 515 516 517
	/* Extended reg_types. */
	PTR_TO_MAP_VALUE_OR_NULL	= PTR_MAYBE_NULL | PTR_TO_MAP_VALUE,
	PTR_TO_SOCKET_OR_NULL		= PTR_MAYBE_NULL | PTR_TO_SOCKET,
	PTR_TO_SOCK_COMMON_OR_NULL	= PTR_MAYBE_NULL | PTR_TO_SOCK_COMMON,
	PTR_TO_TCP_SOCK_OR_NULL		= PTR_MAYBE_NULL | PTR_TO_TCP_SOCK,
	PTR_TO_BTF_ID_OR_NULL		= PTR_MAYBE_NULL | PTR_TO_BTF_ID,

518 519 520 521
	/* This must be the last entry. Its purpose is to ensure the enum is
	 * wide enough to hold the higher bits reserved for bpf_type_flag.
	 */
	__BPF_REG_TYPE_LIMIT	= BPF_TYPE_LIMIT,
522
};
523
static_assert(__BPF_REG_TYPE_MAX <= BPF_BASE_TYPE_LIMIT);
524

525 526 527 528 529
/* The information passed from prog-specific *_is_valid_access
 * back to the verifier.
 */
struct bpf_insn_access_aux {
	enum bpf_reg_type reg_type;
530 531
	union {
		int ctx_field_size;
532 533 534 535
		struct {
			struct btf *btf;
			u32 btf_id;
		};
536 537
	};
	struct bpf_verifier_log *log; /* for verbose logs */
538 539
};

540 541 542 543 544 545
static inline void
bpf_ctx_record_field_size(struct bpf_insn_access_aux *aux, u32 size)
{
	aux->ctx_field_size = size;
}

546 547 548 549 550 551
static inline bool bpf_pseudo_func(const struct bpf_insn *insn)
{
	return insn->code == (BPF_LD | BPF_IMM | BPF_DW) &&
	       insn->src_reg == BPF_PSEUDO_FUNC;
}

552 553 554 555 556
struct bpf_prog_ops {
	int (*test_run)(struct bpf_prog *prog, const union bpf_attr *kattr,
			union bpf_attr __user *uattr);
};

557 558
struct bpf_verifier_ops {
	/* return eBPF function prototype for verification */
559 560 561
	const struct bpf_func_proto *
	(*get_func_proto)(enum bpf_func_id func_id,
			  const struct bpf_prog *prog);
562 563 564 565

	/* return true if 'size' wide access at offset 'off' within bpf_context
	 * with 'type' (read or write) is allowed
	 */
566
	bool (*is_valid_access)(int off, int size, enum bpf_access_type type,
567
				const struct bpf_prog *prog,
568
				struct bpf_insn_access_aux *info);
569 570
	int (*gen_prologue)(struct bpf_insn *insn, bool direct_write,
			    const struct bpf_prog *prog);
571 572
	int (*gen_ld_abs)(const struct bpf_insn *orig,
			  struct bpf_insn *insn_buf);
573 574 575
	u32 (*convert_ctx_access)(enum bpf_access_type type,
				  const struct bpf_insn *src,
				  struct bpf_insn *dst,
576
				  struct bpf_prog *prog, u32 *target_size);
577
	int (*btf_struct_access)(struct bpf_verifier_log *log,
578
				 const struct btf *btf,
579 580 581
				 const struct btf_type *t, int off, int size,
				 enum bpf_access_type atype,
				 u32 *next_btf_id);
582
	bool (*check_kfunc_call)(u32 kfunc_btf_id, struct module *owner);
583 584
};

585
struct bpf_prog_offload_ops {
586
	/* verifier basic callbacks */
587 588
	int (*insn_hook)(struct bpf_verifier_env *env,
			 int insn_idx, int prev_insn_idx);
589
	int (*finalize)(struct bpf_verifier_env *env);
590 591 592 593 594
	/* verifier optimization callbacks (called after .finalize) */
	int (*replace_insn)(struct bpf_verifier_env *env, u32 off,
			    struct bpf_insn *insn);
	int (*remove_insns)(struct bpf_verifier_env *env, u32 off, u32 cnt);
	/* program management callbacks */
595 596
	int (*prepare)(struct bpf_prog *prog);
	int (*translate)(struct bpf_prog *prog);
597
	void (*destroy)(struct bpf_prog *prog);
598 599
};

600
struct bpf_prog_offload {
601 602
	struct bpf_prog		*prog;
	struct net_device	*netdev;
603
	struct bpf_offload_dev	*offdev;
604 605 606
	void			*dev_priv;
	struct list_head	offloads;
	bool			dev_state;
607
	bool			opt_failed;
608 609
	void			*jited_image;
	u32			jited_len;
610 611
};

612 613
enum bpf_cgroup_storage_type {
	BPF_CGROUP_STORAGE_SHARED,
614
	BPF_CGROUP_STORAGE_PERCPU,
615 616 617 618 619
	__BPF_CGROUP_STORAGE_MAX
};

#define MAX_BPF_CGROUP_STORAGE_TYPE __BPF_CGROUP_STORAGE_MAX

620 621 622 623 624
/* The longest tracepoint has 12 args.
 * See include/trace/bpf_probe.h
 */
#define MAX_BPF_FUNC_ARGS 12

625 626 627 628 629
/* The maximum number of arguments passed through registers
 * a single function may have.
 */
#define MAX_BPF_FUNC_REG_ARGS 5

A
Alexei Starovoitov 已提交
630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648
struct btf_func_model {
	u8 ret_size;
	u8 nr_args;
	u8 arg_size[MAX_BPF_FUNC_ARGS];
};

/* Restore arguments before returning from trampoline to let original function
 * continue executing. This flag is used for fentry progs when there are no
 * fexit progs.
 */
#define BPF_TRAMP_F_RESTORE_REGS	BIT(0)
/* Call original function after fentry progs, but before fexit progs.
 * Makes sense for fentry/fexit, normal calls and indirect calls.
 */
#define BPF_TRAMP_F_CALL_ORIG		BIT(1)
/* Skip current frame and return to parent.  Makes sense for fentry/fexit
 * programs only. Should not be used with normal calls and indirect calls.
 */
#define BPF_TRAMP_F_SKIP_FRAME		BIT(2)
649 650 651 652
/* Store IP address of the caller on the trampoline stack,
 * so it's available for trampoline's programs.
 */
#define BPF_TRAMP_F_IP_ARG		BIT(3)
653 654
/* Return the return value of fentry prog. Only used by bpf_struct_ops. */
#define BPF_TRAMP_F_RET_FENTRY_RET	BIT(4)
655

K
KP Singh 已提交
656 657 658
/* Each call __bpf_prog_enter + call bpf_func + call __bpf_prog_exit is ~50
 * bytes on x86.  Pick a number to fit into BPF_IMAGE_SIZE / 2
 */
659
#define BPF_MAX_TRAMP_PROGS 38
K
KP Singh 已提交
660 661 662 663 664 665

struct bpf_tramp_progs {
	struct bpf_prog *progs[BPF_MAX_TRAMP_PROGS];
	int nr_progs;
};

A
Alexei Starovoitov 已提交
666 667 668 669 670 671 672 673 674 675 676 677 678 679 680 681 682 683 684 685
/* Different use cases for BPF trampoline:
 * 1. replace nop at the function entry (kprobe equivalent)
 *    flags = BPF_TRAMP_F_RESTORE_REGS
 *    fentry = a set of programs to run before returning from trampoline
 *
 * 2. replace nop at the function entry (kprobe + kretprobe equivalent)
 *    flags = BPF_TRAMP_F_CALL_ORIG | BPF_TRAMP_F_SKIP_FRAME
 *    orig_call = fentry_ip + MCOUNT_INSN_SIZE
 *    fentry = a set of program to run before calling original function
 *    fexit = a set of program to run after original function
 *
 * 3. replace direct call instruction anywhere in the function body
 *    or assign a function pointer for indirect call (like tcp_congestion_ops->cong_avoid)
 *    With flags = 0
 *      fentry = a set of programs to run before returning from trampoline
 *    With flags = BPF_TRAMP_F_CALL_ORIG
 *      orig_call = original callback addr or direct function addr
 *      fentry = a set of program to run before calling original function
 *      fexit = a set of program to run after original function
 */
A
Alexei Starovoitov 已提交
686 687
struct bpf_tramp_image;
int arch_prepare_bpf_trampoline(struct bpf_tramp_image *tr, void *image, void *image_end,
688
				const struct btf_func_model *m, u32 flags,
K
KP Singh 已提交
689
				struct bpf_tramp_progs *tprogs,
A
Alexei Starovoitov 已提交
690 691
				void *orig_call);
/* these two functions are called from generated trampoline */
692
u64 notrace __bpf_prog_enter(struct bpf_prog *prog);
A
Alexei Starovoitov 已提交
693
void notrace __bpf_prog_exit(struct bpf_prog *prog, u64 start);
694
u64 notrace __bpf_prog_enter_sleepable(struct bpf_prog *prog);
695
void notrace __bpf_prog_exit_sleepable(struct bpf_prog *prog, u64 start);
A
Alexei Starovoitov 已提交
696 697
void notrace __bpf_tramp_enter(struct bpf_tramp_image *tr);
void notrace __bpf_tramp_exit(struct bpf_tramp_image *tr);
A
Alexei Starovoitov 已提交
698

J
Jiri Olsa 已提交
699 700 701
struct bpf_ksym {
	unsigned long		 start;
	unsigned long		 end;
J
Jiri Olsa 已提交
702
	char			 name[KSYM_NAME_LEN];
703
	struct list_head	 lnode;
J
Jiri Olsa 已提交
704
	struct latch_tree_node	 tnode;
705
	bool			 prog;
J
Jiri Olsa 已提交
706 707
};

A
Alexei Starovoitov 已提交
708 709 710
enum bpf_tramp_prog_type {
	BPF_TRAMP_FENTRY,
	BPF_TRAMP_FEXIT,
K
KP Singh 已提交
711
	BPF_TRAMP_MODIFY_RETURN,
712 713
	BPF_TRAMP_MAX,
	BPF_TRAMP_REPLACE, /* more than MAX */
A
Alexei Starovoitov 已提交
714 715
};

A
Alexei Starovoitov 已提交
716 717 718 719 720 721 722 723 724 725 726 727
struct bpf_tramp_image {
	void *image;
	struct bpf_ksym ksym;
	struct percpu_ref pcref;
	void *ip_after_call;
	void *ip_epilogue;
	union {
		struct rcu_head rcu;
		struct work_struct work;
	};
};

A
Alexei Starovoitov 已提交
728 729 730 731 732 733 734 735 736 737
struct bpf_trampoline {
	/* hlist for trampoline_table */
	struct hlist_node hlist;
	/* serializes access to fields of this trampoline */
	struct mutex mutex;
	refcount_t refcnt;
	u64 key;
	struct {
		struct btf_func_model model;
		void *addr;
738
		bool ftrace_managed;
A
Alexei Starovoitov 已提交
739
	} func;
740 741 742 743 744
	/* if !NULL this is BPF_PROG_TYPE_EXT program that extends another BPF
	 * program by replacing one of its functions. func.addr is the address
	 * of the function it replaced.
	 */
	struct bpf_prog *extension_prog;
A
Alexei Starovoitov 已提交
745 746 747 748 749
	/* list of BPF programs using this trampoline */
	struct hlist_head progs_hlist[BPF_TRAMP_MAX];
	/* Number of attached programs. A counter per kind. */
	int progs_cnt[BPF_TRAMP_MAX];
	/* Executable image of trampoline */
A
Alexei Starovoitov 已提交
750
	struct bpf_tramp_image *cur_image;
A
Alexei Starovoitov 已提交
751
	u64 selector;
752
	struct module *mod;
A
Alexei Starovoitov 已提交
753
};
B
Björn Töpel 已提交
754

755 756 757 758 759 760 761
struct bpf_attach_target_info {
	struct btf_func_model fmodel;
	long tgt_addr;
	const char *tgt_name;
	const struct btf_type *tgt_type;
};

762
#define BPF_DISPATCHER_MAX 48 /* Fits in 2048B */
B
Björn Töpel 已提交
763 764 765 766 767 768 769 770 771 772 773 774 775 776

struct bpf_dispatcher_prog {
	struct bpf_prog *prog;
	refcount_t users;
};

struct bpf_dispatcher {
	/* dispatcher mutex */
	struct mutex mutex;
	void *func;
	struct bpf_dispatcher_prog progs[BPF_DISPATCHER_MAX];
	int num_progs;
	void *image;
	u32 image_off;
J
Jiri Olsa 已提交
777
	struct bpf_ksym ksym;
B
Björn Töpel 已提交
778 779
};

780
static __always_inline __nocfi unsigned int bpf_dispatcher_nop_func(
781 782 783 784 785 786 787
	const void *ctx,
	const struct bpf_insn *insnsi,
	unsigned int (*bpf_func)(const void *,
				 const struct bpf_insn *))
{
	return bpf_func(ctx, insnsi);
}
A
Alexei Starovoitov 已提交
788
#ifdef CONFIG_BPF_JIT
789 790
int bpf_trampoline_link_prog(struct bpf_prog *prog, struct bpf_trampoline *tr);
int bpf_trampoline_unlink_prog(struct bpf_prog *prog, struct bpf_trampoline *tr);
791 792
struct bpf_trampoline *bpf_trampoline_get(u64 key,
					  struct bpf_attach_target_info *tgt_info);
A
Alexei Starovoitov 已提交
793
void bpf_trampoline_put(struct bpf_trampoline *tr);
794
int arch_prepare_bpf_dispatcher(void *image, s64 *funcs, int num_funcs);
J
Jiri Olsa 已提交
795 796 797 798 799 800 801 802 803 804 805
#define BPF_DISPATCHER_INIT(_name) {				\
	.mutex = __MUTEX_INITIALIZER(_name.mutex),		\
	.func = &_name##_func,					\
	.progs = {},						\
	.num_progs = 0,						\
	.image = NULL,						\
	.image_off = 0,						\
	.ksym = {						\
		.name  = #_name,				\
		.lnode = LIST_HEAD_INIT(_name.ksym.lnode),	\
	},							\
B
Björn Töpel 已提交
806 807 808
}

#define DEFINE_BPF_DISPATCHER(name)					\
809
	noinline __nocfi unsigned int bpf_dispatcher_##name##_func(	\
B
Björn Töpel 已提交
810 811 812 813 814 815 816
		const void *ctx,					\
		const struct bpf_insn *insnsi,				\
		unsigned int (*bpf_func)(const void *,			\
					 const struct bpf_insn *))	\
	{								\
		return bpf_func(ctx, insnsi);				\
	}								\
817 818 819
	EXPORT_SYMBOL(bpf_dispatcher_##name##_func);			\
	struct bpf_dispatcher bpf_dispatcher_##name =			\
		BPF_DISPATCHER_INIT(bpf_dispatcher_##name);
B
Björn Töpel 已提交
820
#define DECLARE_BPF_DISPATCHER(name)					\
821
	unsigned int bpf_dispatcher_##name##_func(			\
B
Björn Töpel 已提交
822 823 824 825
		const void *ctx,					\
		const struct bpf_insn *insnsi,				\
		unsigned int (*bpf_func)(const void *,			\
					 const struct bpf_insn *));	\
826 827 828
	extern struct bpf_dispatcher bpf_dispatcher_##name;
#define BPF_DISPATCHER_FUNC(name) bpf_dispatcher_##name##_func
#define BPF_DISPATCHER_PTR(name) (&bpf_dispatcher_##name)
B
Björn Töpel 已提交
829 830
void bpf_dispatcher_change_prog(struct bpf_dispatcher *d, struct bpf_prog *from,
				struct bpf_prog *to);
J
Jiri Olsa 已提交
831
/* Called only from JIT-enabled code, so there's no need for stubs. */
J
Jiri Olsa 已提交
832
void *bpf_jit_alloc_exec_page(void);
J
Jiri Olsa 已提交
833 834
void bpf_image_ksym_add(void *data, struct bpf_ksym *ksym);
void bpf_image_ksym_del(struct bpf_ksym *ksym);
J
Jiri Olsa 已提交
835 836
void bpf_ksym_add(struct bpf_ksym *ksym);
void bpf_ksym_del(struct bpf_ksym *ksym);
A
Alexei Starovoitov 已提交
837 838
int bpf_jit_charge_modmem(u32 pages);
void bpf_jit_uncharge_modmem(u32 pages);
839
bool bpf_prog_has_trampoline(const struct bpf_prog *prog);
A
Alexei Starovoitov 已提交
840
#else
841 842
static inline int bpf_trampoline_link_prog(struct bpf_prog *prog,
					   struct bpf_trampoline *tr)
A
Alexei Starovoitov 已提交
843 844 845
{
	return -ENOTSUPP;
}
846 847
static inline int bpf_trampoline_unlink_prog(struct bpf_prog *prog,
					     struct bpf_trampoline *tr)
A
Alexei Starovoitov 已提交
848 849 850
{
	return -ENOTSUPP;
}
851 852 853 854 855
static inline struct bpf_trampoline *bpf_trampoline_get(u64 key,
							struct bpf_attach_target_info *tgt_info)
{
	return ERR_PTR(-EOPNOTSUPP);
}
A
Alexei Starovoitov 已提交
856
static inline void bpf_trampoline_put(struct bpf_trampoline *tr) {}
B
Björn Töpel 已提交
857 858
#define DEFINE_BPF_DISPATCHER(name)
#define DECLARE_BPF_DISPATCHER(name)
859
#define BPF_DISPATCHER_FUNC(name) bpf_dispatcher_nop_func
B
Björn Töpel 已提交
860 861 862 863
#define BPF_DISPATCHER_PTR(name) NULL
static inline void bpf_dispatcher_change_prog(struct bpf_dispatcher *d,
					      struct bpf_prog *from,
					      struct bpf_prog *to) {}
864 865 866 867
static inline bool is_bpf_image_address(unsigned long address)
{
	return false;
}
868 869 870 871
static inline bool bpf_prog_has_trampoline(const struct bpf_prog *prog)
{
	return false;
}
A
Alexei Starovoitov 已提交
872 873
#endif

874
struct bpf_func_info_aux {
875
	u16 linkage;
876 877 878
	bool unreliable;
};

879 880 881 882 883 884
enum bpf_jit_poke_reason {
	BPF_POKE_REASON_TAIL_CALL,
};

/* Descriptor of pokes pointing /into/ the JITed image. */
struct bpf_jit_poke_descriptor {
885
	void *tailcall_target;
886 887
	void *tailcall_bypass;
	void *bypass_addr;
888
	void *aux;
889 890 891 892 893 894
	union {
		struct {
			struct bpf_map *map;
			u32 key;
		} tail_call;
	};
895
	bool tailcall_target_stable;
896 897
	u8 adj_off;
	u16 reason;
898
	u32 insn_idx;
899 900
};

901 902 903 904
/* reg_type info for ctx arguments */
struct bpf_ctx_arg_aux {
	u32 offset;
	enum bpf_reg_type reg_type;
905
	u32 btf_id;
906 907
};

908 909 910 911 912
struct btf_mod_pair {
	struct btf *btf;
	struct module *module;
};

913 914
struct bpf_kfunc_desc_tab;

915
struct bpf_prog_aux {
916
	atomic64_t refcnt;
917
	u32 used_map_cnt;
918
	u32 used_btf_cnt;
919
	u32 max_ctx_offset;
920
	u32 max_pkt_offset;
921
	u32 max_tp_access;
922
	u32 stack_depth;
M
Martin KaFai Lau 已提交
923
	u32 id;
924 925
	u32 func_cnt; /* used by non-func prog as the number of func progs */
	u32 func_idx; /* 0 for non-func prog, the index in func array for func prog */
926
	u32 attach_btf_id; /* in-kernel BTF type id to attach to */
927
	u32 ctx_arg_info_size;
928 929
	u32 max_rdonly_access;
	u32 max_rdwr_access;
930
	struct btf *attach_btf;
931
	const struct bpf_ctx_arg_aux *ctx_arg_info;
932 933 934
	struct mutex dst_mutex; /* protects dst_* pointers below, *after* prog becomes visible */
	struct bpf_prog *dst_prog;
	struct bpf_trampoline *dst_trampoline;
935 936
	enum bpf_prog_type saved_dst_prog_type;
	enum bpf_attach_type saved_dst_attach_type;
937
	bool verifier_zext; /* Zero extensions has been inserted by verifier. */
938
	bool offload_requested;
939
	bool attach_btf_trace; /* true if attaching to BTF-enabled raw tp */
940
	bool func_proto_unreliable;
941
	bool sleepable;
942
	bool tail_call_reachable;
A
Alexei Starovoitov 已提交
943
	struct hlist_node tramp_hlist;
944 945 946 947
	/* BTF_KIND_FUNC_PROTO for valid attach_btf_id */
	const struct btf_type *attach_func_proto;
	/* function name for valid attach_btf_id */
	const char *attach_func_name;
948 949
	struct bpf_prog **func;
	void *jit_data; /* JIT specific data. arch dependent */
950
	struct bpf_jit_poke_descriptor *poke_tab;
951
	struct bpf_kfunc_desc_tab *kfunc_tab;
952
	struct bpf_kfunc_btf_tab *kfunc_btf_tab;
953
	u32 size_poke_tab;
J
Jiri Olsa 已提交
954
	struct bpf_ksym ksym;
955
	const struct bpf_prog_ops *ops;
956
	struct bpf_map **used_maps;
957
	struct mutex used_maps_mutex; /* mutex for used_maps and used_map_cnt */
958
	struct btf_mod_pair *used_btfs;
959
	struct bpf_prog *prog;
960
	struct user_struct *user;
961
	u64 load_time; /* ns since boottime */
962
	u32 verified_insns;
963
	struct bpf_map *cgroup_storage[MAX_BPF_CGROUP_STORAGE_TYPE];
964
	char name[BPF_OBJ_NAME_LEN];
965 966 967
#ifdef CONFIG_SECURITY
	void *security;
#endif
968
	struct bpf_prog_offload *offload;
Y
Yonghong Song 已提交
969
	struct btf *btf;
970
	struct bpf_func_info *func_info;
971
	struct bpf_func_info_aux *func_info_aux;
M
Martin KaFai Lau 已提交
972 973 974 975 976 977 978 979 980 981 982 983 984 985 986
	/* bpf_line_info loaded from userspace.  linfo->insn_off
	 * has the xlated insn offset.
	 * Both the main and sub prog share the same linfo.
	 * The subprog can access its first linfo by
	 * using the linfo_idx.
	 */
	struct bpf_line_info *linfo;
	/* jited_linfo is the jited addr of the linfo.  It has a
	 * one to one mapping to linfo:
	 * jited_linfo[i] is the jited addr for the linfo[i]->insn_off.
	 * Both the main and sub prog share the same jited_linfo.
	 * The subprog can access its first jited_linfo by
	 * using the linfo_idx.
	 */
	void **jited_linfo;
987
	u32 func_info_cnt;
M
Martin KaFai Lau 已提交
988 989 990 991 992 993
	u32 nr_linfo;
	/* subprog can use linfo_idx to access its first linfo and
	 * jited_linfo.
	 * main prog always has linfo_idx == 0
	 */
	u32 linfo_idx;
994 995
	u32 num_exentries;
	struct exception_table_entry *extable;
996 997 998 999
	union {
		struct work_struct work;
		struct rcu_head	rcu;
	};
1000 1001
};

1002 1003 1004 1005 1006 1007
struct bpf_array_aux {
	/* 'Ownership' of prog array is claimed by the first program that
	 * is going to use this map or by the first program which FD is
	 * stored in the map to make sure that all callers and callees have
	 * the same prog type and JITed flag.
	 */
1008 1009 1010 1011 1012
	struct {
		spinlock_t lock;
		enum bpf_prog_type type;
		bool jited;
	} owner;
1013 1014 1015 1016 1017
	/* Programs with direct jumps into programs part of this array. */
	struct list_head poke_progs;
	struct bpf_map *map;
	struct mutex poke_mutex;
	struct work_struct work;
1018 1019
};

1020 1021 1022 1023 1024 1025 1026 1027 1028 1029 1030 1031
struct bpf_link {
	atomic64_t refcnt;
	u32 id;
	enum bpf_link_type type;
	const struct bpf_link_ops *ops;
	struct bpf_prog *prog;
	struct work_struct work;
};

struct bpf_link_ops {
	void (*release)(struct bpf_link *link);
	void (*dealloc)(struct bpf_link *link);
1032
	int (*detach)(struct bpf_link *link);
1033 1034 1035 1036 1037 1038 1039 1040 1041 1042 1043 1044 1045 1046
	int (*update_prog)(struct bpf_link *link, struct bpf_prog *new_prog,
			   struct bpf_prog *old_prog);
	void (*show_fdinfo)(const struct bpf_link *link, struct seq_file *seq);
	int (*fill_link_info)(const struct bpf_link *link,
			      struct bpf_link_info *info);
};

struct bpf_link_primer {
	struct bpf_link *link;
	struct file *file;
	int fd;
	u32 id;
};

1047
struct bpf_struct_ops_value;
1048 1049 1050 1051 1052 1053 1054 1055
struct btf_member;

#define BPF_STRUCT_OPS_MAX_NR_MEMBERS 64
struct bpf_struct_ops {
	const struct bpf_verifier_ops *verifier_ops;
	int (*init)(struct btf *btf);
	int (*check_member)(const struct btf_type *t,
			    const struct btf_member *member);
1056 1057 1058 1059 1060
	int (*init_member)(const struct btf_type *t,
			   const struct btf_member *member,
			   void *kdata, const void *udata);
	int (*reg)(void *kdata);
	void (*unreg)(void *kdata);
1061
	const struct btf_type *type;
1062
	const struct btf_type *value_type;
1063 1064 1065
	const char *name;
	struct btf_func_model func_models[BPF_STRUCT_OPS_MAX_NR_MEMBERS];
	u32 type_id;
1066
	u32 value_id;
1067 1068 1069
};

#if defined(CONFIG_BPF_JIT) && defined(CONFIG_BPF_SYSCALL)
1070
#define BPF_MODULE_OWNER ((void *)((0xeB9FUL << 2) + POISON_POINTER_DELTA))
1071
const struct bpf_struct_ops *bpf_struct_ops_find(u32 type_id);
1072
void bpf_struct_ops_init(struct btf *btf, struct bpf_verifier_log *log);
1073 1074 1075 1076
bool bpf_struct_ops_get(const void *kdata);
void bpf_struct_ops_put(const void *kdata);
int bpf_struct_ops_map_sys_lookup_elem(struct bpf_map *map, void *key,
				       void *value);
1077 1078 1079 1080
int bpf_struct_ops_prepare_trampoline(struct bpf_tramp_progs *tprogs,
				      struct bpf_prog *prog,
				      const struct btf_func_model *model,
				      void *image, void *image_end);
1081 1082 1083 1084 1085 1086 1087 1088 1089 1090 1091 1092 1093 1094
static inline bool bpf_try_module_get(const void *data, struct module *owner)
{
	if (owner == BPF_MODULE_OWNER)
		return bpf_struct_ops_get(data);
	else
		return try_module_get(owner);
}
static inline void bpf_module_put(const void *data, struct module *owner)
{
	if (owner == BPF_MODULE_OWNER)
		bpf_struct_ops_put(data);
	else
		module_put(owner);
}
1095 1096 1097 1098 1099 1100 1101 1102 1103 1104 1105 1106 1107 1108 1109 1110

#ifdef CONFIG_NET
/* Define it here to avoid the use of forward declaration */
struct bpf_dummy_ops_state {
	int val;
};

struct bpf_dummy_ops {
	int (*test_1)(struct bpf_dummy_ops_state *cb);
	int (*test_2)(struct bpf_dummy_ops_state *cb, int a1, unsigned short a2,
		      char a3, unsigned long a4);
};

int bpf_struct_ops_test_run(struct bpf_prog *prog, const union bpf_attr *kattr,
			    union bpf_attr __user *uattr);
#endif
1111 1112 1113 1114 1115
#else
static inline const struct bpf_struct_ops *bpf_struct_ops_find(u32 type_id)
{
	return NULL;
}
1116 1117 1118 1119
static inline void bpf_struct_ops_init(struct btf *btf,
				       struct bpf_verifier_log *log)
{
}
1120 1121 1122 1123 1124 1125 1126 1127 1128 1129 1130 1131 1132 1133
static inline bool bpf_try_module_get(const void *data, struct module *owner)
{
	return try_module_get(owner);
}
static inline void bpf_module_put(const void *data, struct module *owner)
{
	module_put(owner);
}
static inline int bpf_struct_ops_map_sys_lookup_elem(struct bpf_map *map,
						     void *key,
						     void *value)
{
	return -EINVAL;
}
1134 1135
#endif

1136 1137 1138
struct bpf_array {
	struct bpf_map map;
	u32 elem_size;
1139
	u32 index_mask;
1140
	struct bpf_array_aux *aux;
1141 1142
	union {
		char value[0] __aligned(8);
1143
		void *ptrs[0] __aligned(8);
1144
		void __percpu *pptrs[0] __aligned(8);
1145 1146
	};
};
1147

1148
#define BPF_COMPLEXITY_LIMIT_INSNS      1000000 /* yes. 1M insns */
1149
#define MAX_TAIL_CALL_CNT 33
1150

1151 1152 1153 1154 1155 1156 1157 1158 1159 1160 1161 1162 1163 1164 1165 1166 1167 1168 1169 1170 1171 1172 1173 1174 1175 1176 1177 1178 1179
#define BPF_F_ACCESS_MASK	(BPF_F_RDONLY |		\
				 BPF_F_RDONLY_PROG |	\
				 BPF_F_WRONLY |		\
				 BPF_F_WRONLY_PROG)

#define BPF_MAP_CAN_READ	BIT(0)
#define BPF_MAP_CAN_WRITE	BIT(1)

static inline u32 bpf_map_flags_to_cap(struct bpf_map *map)
{
	u32 access_flags = map->map_flags & (BPF_F_RDONLY_PROG | BPF_F_WRONLY_PROG);

	/* Combination of BPF_F_RDONLY_PROG | BPF_F_WRONLY_PROG is
	 * not possible.
	 */
	if (access_flags & BPF_F_RDONLY_PROG)
		return BPF_MAP_CAN_READ;
	else if (access_flags & BPF_F_WRONLY_PROG)
		return BPF_MAP_CAN_WRITE;
	else
		return BPF_MAP_CAN_READ | BPF_MAP_CAN_WRITE;
}

static inline bool bpf_map_flags_access_ok(u32 access_flags)
{
	return (access_flags & (BPF_F_RDONLY_PROG | BPF_F_WRONLY_PROG)) !=
	       (BPF_F_RDONLY_PROG | BPF_F_WRONLY_PROG);
}

1180 1181 1182 1183 1184 1185 1186
struct bpf_event_entry {
	struct perf_event *event;
	struct file *perf_file;
	struct file *map_file;
	struct rcu_head rcu;
};

1187
bool bpf_prog_array_compatible(struct bpf_array *array, const struct bpf_prog *fp);
1188
int bpf_prog_calc_tag(struct bpf_prog *fp);
1189

1190
const struct bpf_func_proto *bpf_get_trace_printk_proto(void);
1191
const struct bpf_func_proto *bpf_get_trace_vprintk_proto(void);
1192 1193

typedef unsigned long (*bpf_ctx_copy_t)(void *dst, const void *src,
1194
					unsigned long off, unsigned long len);
1195 1196 1197 1198 1199
typedef u32 (*bpf_convert_ctx_access_t)(enum bpf_access_type type,
					const struct bpf_insn *src,
					struct bpf_insn *dst,
					struct bpf_prog *prog,
					u32 *target_size);
1200 1201 1202

u64 bpf_event_output(struct bpf_map *map, u64 flags, void *meta, u64 meta_size,
		     void *ctx, u64 ctx_size, bpf_ctx_copy_t ctx_copy);
1203

1204 1205 1206
/* an array of programs to be executed under rcu_lock.
 *
 * Typical usage:
1207
 * ret = BPF_PROG_RUN_ARRAY(&bpf_prog_array, ctx, bpf_prog_run);
1208 1209 1210 1211 1212 1213 1214 1215
 *
 * the structure returned by bpf_prog_array_alloc() should be populated
 * with program pointers and the last pointer must be NULL.
 * The user has to keep refcnt on the program and make sure the program
 * is removed from the array before bpf_prog_put().
 * The 'struct bpf_prog_array *' should only be replaced with xchg()
 * since other cpus are walking the array of pointers in parallel.
 */
1216 1217
struct bpf_prog_array_item {
	struct bpf_prog *prog;
1218 1219 1220 1221
	union {
		struct bpf_cgroup_storage *cgroup_storage[MAX_BPF_CGROUP_STORAGE_TYPE];
		u64 bpf_cookie;
	};
1222 1223
};

1224 1225
struct bpf_prog_array {
	struct rcu_head rcu;
1226
	struct bpf_prog_array_item items[];
1227 1228
};

1229
struct bpf_prog_array *bpf_prog_array_alloc(u32 prog_cnt, gfp_t flags);
1230 1231
void bpf_prog_array_free(struct bpf_prog_array *progs);
int bpf_prog_array_length(struct bpf_prog_array *progs);
1232
bool bpf_prog_array_is_empty(struct bpf_prog_array *array);
1233
int bpf_prog_array_copy_to_user(struct bpf_prog_array *progs,
1234
				__u32 __user *prog_ids, u32 cnt);
1235

1236
void bpf_prog_array_delete_safe(struct bpf_prog_array *progs,
1237
				struct bpf_prog *old_prog);
1238 1239 1240
int bpf_prog_array_delete_safe_at(struct bpf_prog_array *array, int index);
int bpf_prog_array_update_at(struct bpf_prog_array *array, int index,
			     struct bpf_prog *prog);
1241
int bpf_prog_array_copy_info(struct bpf_prog_array *array,
1242 1243
			     u32 *prog_ids, u32 request_cnt,
			     u32 *prog_cnt);
1244
int bpf_prog_array_copy(struct bpf_prog_array *old_array,
1245 1246
			struct bpf_prog *exclude_prog,
			struct bpf_prog *include_prog,
1247
			u64 bpf_cookie,
1248 1249
			struct bpf_prog_array **new_array);

1250 1251 1252 1253
struct bpf_run_ctx {};

struct bpf_cg_run_ctx {
	struct bpf_run_ctx run_ctx;
1254
	const struct bpf_prog_array_item *prog_item;
1255 1256
};

1257 1258 1259 1260 1261
struct bpf_trace_run_ctx {
	struct bpf_run_ctx run_ctx;
	u64 bpf_cookie;
};

1262 1263 1264 1265 1266 1267 1268 1269 1270 1271 1272 1273 1274 1275 1276 1277 1278 1279
static inline struct bpf_run_ctx *bpf_set_run_ctx(struct bpf_run_ctx *new_ctx)
{
	struct bpf_run_ctx *old_ctx = NULL;

#ifdef CONFIG_BPF_SYSCALL
	old_ctx = current->bpf_ctx;
	current->bpf_ctx = new_ctx;
#endif
	return old_ctx;
}

static inline void bpf_reset_run_ctx(struct bpf_run_ctx *old_ctx)
{
#ifdef CONFIG_BPF_SYSCALL
	current->bpf_ctx = old_ctx;
#endif
}

1280 1281 1282 1283 1284
/* BPF program asks to bypass CAP_NET_BIND_SERVICE in bind. */
#define BPF_RET_BIND_NO_CAP_NET_BIND_SERVICE			(1 << 0)
/* BPF program asks to set CN on the packet. */
#define BPF_RET_SET_CN						(1 << 0)

1285 1286 1287 1288 1289 1290 1291 1292 1293 1294 1295 1296 1297 1298 1299 1300 1301 1302 1303 1304 1305 1306 1307 1308 1309 1310 1311 1312 1313 1314 1315 1316 1317 1318 1319 1320 1321 1322 1323 1324 1325 1326 1327 1328 1329 1330 1331 1332 1333 1334 1335 1336 1337 1338 1339 1340 1341 1342 1343 1344 1345 1346 1347 1348 1349 1350 1351
typedef u32 (*bpf_prog_run_fn)(const struct bpf_prog *prog, const void *ctx);

static __always_inline u32
BPF_PROG_RUN_ARRAY_CG_FLAGS(const struct bpf_prog_array __rcu *array_rcu,
			    const void *ctx, bpf_prog_run_fn run_prog,
			    u32 *ret_flags)
{
	const struct bpf_prog_array_item *item;
	const struct bpf_prog *prog;
	const struct bpf_prog_array *array;
	struct bpf_run_ctx *old_run_ctx;
	struct bpf_cg_run_ctx run_ctx;
	u32 ret = 1;
	u32 func_ret;

	migrate_disable();
	rcu_read_lock();
	array = rcu_dereference(array_rcu);
	item = &array->items[0];
	old_run_ctx = bpf_set_run_ctx(&run_ctx.run_ctx);
	while ((prog = READ_ONCE(item->prog))) {
		run_ctx.prog_item = item;
		func_ret = run_prog(prog, ctx);
		ret &= (func_ret & 1);
		*(ret_flags) |= (func_ret >> 1);
		item++;
	}
	bpf_reset_run_ctx(old_run_ctx);
	rcu_read_unlock();
	migrate_enable();
	return ret;
}

static __always_inline u32
BPF_PROG_RUN_ARRAY_CG(const struct bpf_prog_array __rcu *array_rcu,
		      const void *ctx, bpf_prog_run_fn run_prog)
{
	const struct bpf_prog_array_item *item;
	const struct bpf_prog *prog;
	const struct bpf_prog_array *array;
	struct bpf_run_ctx *old_run_ctx;
	struct bpf_cg_run_ctx run_ctx;
	u32 ret = 1;

	migrate_disable();
	rcu_read_lock();
	array = rcu_dereference(array_rcu);
	item = &array->items[0];
	old_run_ctx = bpf_set_run_ctx(&run_ctx.run_ctx);
	while ((prog = READ_ONCE(item->prog))) {
		run_ctx.prog_item = item;
		ret &= run_prog(prog, ctx);
		item++;
	}
	bpf_reset_run_ctx(old_run_ctx);
	rcu_read_unlock();
	migrate_enable();
	return ret;
}

static __always_inline u32
BPF_PROG_RUN_ARRAY(const struct bpf_prog_array __rcu *array_rcu,
		   const void *ctx, bpf_prog_run_fn run_prog)
{
	const struct bpf_prog_array_item *item;
	const struct bpf_prog *prog;
	const struct bpf_prog_array *array;
1352 1353
	struct bpf_run_ctx *old_run_ctx;
	struct bpf_trace_run_ctx run_ctx;
1354 1355 1356 1357 1358 1359 1360
	u32 ret = 1;

	migrate_disable();
	rcu_read_lock();
	array = rcu_dereference(array_rcu);
	if (unlikely(!array))
		goto out;
1361
	old_run_ctx = bpf_set_run_ctx(&run_ctx.run_ctx);
1362 1363
	item = &array->items[0];
	while ((prog = READ_ONCE(item->prog))) {
1364
		run_ctx.bpf_cookie = item->bpf_cookie;
1365 1366 1367
		ret &= run_prog(prog, ctx);
		item++;
	}
1368
	bpf_reset_run_ctx(old_run_ctx);
1369 1370 1371 1372 1373
out:
	rcu_read_unlock();
	migrate_enable();
	return ret;
}
1374

1375 1376 1377 1378 1379 1380 1381 1382 1383 1384 1385 1386 1387 1388 1389 1390 1391 1392 1393 1394 1395 1396 1397 1398
/* To be used by __cgroup_bpf_run_filter_skb for EGRESS BPF progs
 * so BPF programs can request cwr for TCP packets.
 *
 * Current cgroup skb programs can only return 0 or 1 (0 to drop the
 * packet. This macro changes the behavior so the low order bit
 * indicates whether the packet should be dropped (0) or not (1)
 * and the next bit is a congestion notification bit. This could be
 * used by TCP to call tcp_enter_cwr()
 *
 * Hence, new allowed return values of CGROUP EGRESS BPF programs are:
 *   0: drop packet
 *   1: keep packet
 *   2: drop packet and cn
 *   3: keep packet and cn
 *
 * This macro then converts it to one of the NET_XMIT or an error
 * code that is then interpreted as drop packet (and no cn):
 *   0: NET_XMIT_SUCCESS  skb should be transmitted
 *   1: NET_XMIT_DROP     skb should be dropped and cn
 *   2: NET_XMIT_CN       skb should be transmitted and cn
 *   3: -EPERM            skb should be dropped
 */
#define BPF_PROG_CGROUP_INET_EGRESS_RUN_ARRAY(array, ctx, func)		\
	({						\
1399 1400 1401
		u32 _flags = 0;				\
		bool _cn;				\
		u32 _ret;				\
1402
		_ret = BPF_PROG_RUN_ARRAY_CG_FLAGS(array, ctx, func, &_flags); \
1403
		_cn = _flags & BPF_RET_SET_CN;		\
1404 1405 1406 1407 1408 1409 1410
		if (_ret)				\
			_ret = (_cn ? NET_XMIT_CN : NET_XMIT_SUCCESS);	\
		else					\
			_ret = (_cn ? NET_XMIT_DROP : -EPERM);		\
		_ret;					\
	})

1411
#ifdef CONFIG_BPF_SYSCALL
1412
DECLARE_PER_CPU(int, bpf_prog_active);
1413
extern struct mutex bpf_stats_enabled_mutex;
1414

1415 1416 1417 1418 1419 1420 1421 1422 1423
/*
 * Block execution of BPF programs attached to instrumentation (perf,
 * kprobes, tracepoints) to prevent deadlocks on map operations as any of
 * these events can happen inside a region which holds a map bucket lock
 * and can deadlock on it.
 */
static inline void bpf_disable_instrumentation(void)
{
	migrate_disable();
1424
	this_cpu_inc(bpf_prog_active);
1425 1426 1427 1428
}

static inline void bpf_enable_instrumentation(void)
{
1429
	this_cpu_dec(bpf_prog_active);
1430 1431 1432
	migrate_enable();
}

1433 1434
extern const struct file_operations bpf_map_fops;
extern const struct file_operations bpf_prog_fops;
Y
Yonghong Song 已提交
1435
extern const struct file_operations bpf_iter_fops;
1436

A
Alexei Starovoitov 已提交
1437
#define BPF_PROG_TYPE(_id, _name, prog_ctx_type, kern_ctx_type) \
1438 1439
	extern const struct bpf_prog_ops _name ## _prog_ops; \
	extern const struct bpf_verifier_ops _name ## _verifier_ops;
1440 1441
#define BPF_MAP_TYPE(_id, _ops) \
	extern const struct bpf_map_ops _ops;
1442
#define BPF_LINK_TYPE(_id, _name)
1443 1444
#include <linux/bpf_types.h>
#undef BPF_PROG_TYPE
1445
#undef BPF_MAP_TYPE
1446
#undef BPF_LINK_TYPE
1447

1448
extern const struct bpf_prog_ops bpf_offload_prog_ops;
1449 1450 1451
extern const struct bpf_verifier_ops tc_cls_act_analyzer_ops;
extern const struct bpf_verifier_ops xdp_analyzer_ops;

1452
struct bpf_prog *bpf_prog_get(u32 ufd);
1453
struct bpf_prog *bpf_prog_get_type_dev(u32 ufd, enum bpf_prog_type type,
1454
				       bool attach_drv);
1455
void bpf_prog_add(struct bpf_prog *prog, int i);
1456
void bpf_prog_sub(struct bpf_prog *prog, int i);
1457
void bpf_prog_inc(struct bpf_prog *prog);
1458
struct bpf_prog * __must_check bpf_prog_inc_not_zero(struct bpf_prog *prog);
1459 1460
void bpf_prog_put(struct bpf_prog *prog);

1461
void bpf_prog_free_id(struct bpf_prog *prog, bool do_idr_lock);
1462
void bpf_map_free_id(struct bpf_map *map, bool do_idr_lock);
1463

1464
struct bpf_map *bpf_map_get(u32 ufd);
1465
struct bpf_map *bpf_map_get_with_uref(u32 ufd);
1466
struct bpf_map *__bpf_map_get(struct fd f);
1467 1468 1469
void bpf_map_inc(struct bpf_map *map);
void bpf_map_inc_with_uref(struct bpf_map *map);
struct bpf_map * __must_check bpf_map_inc_not_zero(struct bpf_map *map);
1470
void bpf_map_put_with_uref(struct bpf_map *map);
1471
void bpf_map_put(struct bpf_map *map);
1472 1473
void *bpf_map_area_alloc(u64 size, int numa_node);
void *bpf_map_area_mmapable_alloc(u64 size, int numa_node);
1474
void bpf_map_area_free(void *base);
1475
bool bpf_map_write_active(const struct bpf_map *map);
1476
void bpf_map_init_from_attr(struct bpf_map *map, union bpf_attr *attr);
1477 1478
int  generic_map_lookup_batch(struct bpf_map *map,
			      const union bpf_attr *attr,
1479 1480 1481 1482 1483 1484
			      union bpf_attr __user *uattr);
int  generic_map_update_batch(struct bpf_map *map,
			      const union bpf_attr *attr,
			      union bpf_attr __user *uattr);
int  generic_map_delete_batch(struct bpf_map *map,
			      const union bpf_attr *attr,
1485
			      union bpf_attr __user *uattr);
Y
Yonghong Song 已提交
1486
struct bpf_map *bpf_map_get_curr_or_next(u32 *id);
A
Alexei Starovoitov 已提交
1487
struct bpf_prog *bpf_prog_get_curr_or_next(u32 *id);
1488

1489 1490 1491 1492 1493 1494 1495 1496 1497 1498 1499 1500 1501 1502 1503 1504 1505 1506 1507 1508 1509 1510 1511 1512 1513 1514 1515 1516
#ifdef CONFIG_MEMCG_KMEM
void *bpf_map_kmalloc_node(const struct bpf_map *map, size_t size, gfp_t flags,
			   int node);
void *bpf_map_kzalloc(const struct bpf_map *map, size_t size, gfp_t flags);
void __percpu *bpf_map_alloc_percpu(const struct bpf_map *map, size_t size,
				    size_t align, gfp_t flags);
#else
static inline void *
bpf_map_kmalloc_node(const struct bpf_map *map, size_t size, gfp_t flags,
		     int node)
{
	return kmalloc_node(size, flags, node);
}

static inline void *
bpf_map_kzalloc(const struct bpf_map *map, size_t size, gfp_t flags)
{
	return kzalloc(size, flags);
}

static inline void __percpu *
bpf_map_alloc_percpu(const struct bpf_map *map, size_t size, size_t align,
		     gfp_t flags)
{
	return __alloc_percpu_gfp(size, align, flags);
}
#endif

1517 1518
extern int sysctl_unprivileged_bpf_disabled;

A
Alexei Starovoitov 已提交
1519 1520 1521 1522 1523
static inline bool bpf_allow_ptr_leaks(void)
{
	return perfmon_capable();
}

1524 1525 1526 1527 1528
static inline bool bpf_allow_uninit_stack(void)
{
	return perfmon_capable();
}

1529 1530 1531 1532 1533
static inline bool bpf_allow_ptr_to_map_access(void)
{
	return perfmon_capable();
}

A
Alexei Starovoitov 已提交
1534 1535 1536 1537 1538 1539 1540 1541 1542 1543
static inline bool bpf_bypass_spec_v1(void)
{
	return perfmon_capable();
}

static inline bool bpf_bypass_spec_v4(void)
{
	return perfmon_capable();
}

1544
int bpf_map_new_fd(struct bpf_map *map, int flags);
1545 1546
int bpf_prog_new_fd(struct bpf_prog *prog);

1547
void bpf_link_init(struct bpf_link *link, enum bpf_link_type type,
A
Andrii Nakryiko 已提交
1548 1549 1550 1551
		   const struct bpf_link_ops *ops, struct bpf_prog *prog);
int bpf_link_prime(struct bpf_link *link, struct bpf_link_primer *primer);
int bpf_link_settle(struct bpf_link_primer *primer);
void bpf_link_cleanup(struct bpf_link_primer *primer);
1552 1553 1554
void bpf_link_inc(struct bpf_link *link);
void bpf_link_put(struct bpf_link *link);
int bpf_link_new_fd(struct bpf_link *link);
1555
struct file *bpf_link_new_file(struct bpf_link *link, int *reserved_fd);
1556 1557
struct bpf_link *bpf_link_get_from_fd(u32 ufd);

1558
int bpf_obj_pin_user(u32 ufd, const char __user *pathname);
1559
int bpf_obj_get_user(const char __user *pathname, int flags);
1560

1561
#define BPF_ITER_FUNC_PREFIX "bpf_iter_"
1562
#define DEFINE_BPF_ITER_FUNC(target, args...)			\
1563 1564
	extern int bpf_iter_ ## target(args);			\
	int __init bpf_iter_ ## target(args) { return 0; }
1565

1566
struct bpf_iter_aux_info {
1567
	struct bpf_map *map;
1568 1569
};

1570 1571 1572 1573
typedef int (*bpf_iter_attach_target_t)(struct bpf_prog *prog,
					union bpf_iter_link_info *linfo,
					struct bpf_iter_aux_info *aux);
typedef void (*bpf_iter_detach_target_t)(struct bpf_iter_aux_info *aux);
1574 1575 1576 1577
typedef void (*bpf_iter_show_fdinfo_t) (const struct bpf_iter_aux_info *aux,
					struct seq_file *seq);
typedef int (*bpf_iter_fill_link_info_t)(const struct bpf_iter_aux_info *aux,
					 struct bpf_link_info *info);
1578 1579 1580
typedef const struct bpf_func_proto *
(*bpf_iter_get_func_proto_t)(enum bpf_func_id func_id,
			     const struct bpf_prog *prog);
1581

1582 1583 1584 1585
enum bpf_iter_feature {
	BPF_ITER_RESCHED	= BIT(0),
};

1586
#define BPF_ITER_CTX_ARG_MAX 2
1587 1588
struct bpf_iter_reg {
	const char *target;
1589 1590
	bpf_iter_attach_target_t attach_target;
	bpf_iter_detach_target_t detach_target;
1591 1592
	bpf_iter_show_fdinfo_t show_fdinfo;
	bpf_iter_fill_link_info_t fill_link_info;
1593
	bpf_iter_get_func_proto_t get_func_proto;
1594
	u32 ctx_arg_info_size;
1595
	u32 feature;
1596
	struct bpf_ctx_arg_aux ctx_arg_info[BPF_ITER_CTX_ARG_MAX];
1597
	const struct bpf_iter_seq_info *seq_info;
1598 1599
};

1600 1601 1602 1603 1604 1605
struct bpf_iter_meta {
	__bpf_md_ptr(struct seq_file *, seq);
	u64 session_id;
	u64 seq_num;
};

1606 1607 1608 1609 1610 1611 1612
struct bpf_iter__bpf_map_elem {
	__bpf_md_ptr(struct bpf_iter_meta *, meta);
	__bpf_md_ptr(struct bpf_map *, map);
	__bpf_md_ptr(void *, key);
	__bpf_md_ptr(void *, value);
};

1613
int bpf_iter_reg_target(const struct bpf_iter_reg *reg_info);
1614
void bpf_iter_unreg_target(const struct bpf_iter_reg *reg_info);
1615
bool bpf_iter_prog_supported(struct bpf_prog *prog);
1616 1617
const struct bpf_func_proto *
bpf_iter_get_func_proto(enum bpf_func_id func_id, const struct bpf_prog *prog);
1618
int bpf_iter_link_attach(const union bpf_attr *attr, bpfptr_t uattr, struct bpf_prog *prog);
1619
int bpf_iter_new_fd(struct bpf_link *link);
Y
Yonghong Song 已提交
1620
bool bpf_link_is_iter(struct bpf_link *link);
1621 1622
struct bpf_prog *bpf_iter_get_info(struct bpf_iter_meta *meta, bool in_stop);
int bpf_iter_run_prog(struct bpf_prog *prog, void *ctx);
1623 1624 1625 1626
void bpf_iter_map_show_fdinfo(const struct bpf_iter_aux_info *aux,
			      struct seq_file *seq);
int bpf_iter_map_fill_link_info(const struct bpf_iter_aux_info *aux,
				struct bpf_link_info *info);
1627

1628 1629 1630 1631
int map_set_for_each_callback_args(struct bpf_verifier_env *env,
				   struct bpf_func_state *caller,
				   struct bpf_func_state *callee);

1632 1633 1634 1635 1636 1637
int bpf_percpu_hash_copy(struct bpf_map *map, void *key, void *value);
int bpf_percpu_array_copy(struct bpf_map *map, void *key, void *value);
int bpf_percpu_hash_update(struct bpf_map *map, void *key, void *value,
			   u64 flags);
int bpf_percpu_array_update(struct bpf_map *map, void *key, void *value,
			    u64 flags);
1638

1639
int bpf_stackmap_copy(struct bpf_map *map, void *key, void *value);
1640

1641 1642
int bpf_fd_array_map_update_elem(struct bpf_map *map, struct file *map_file,
				 void *key, void *value, u64 map_flags);
1643
int bpf_fd_array_map_lookup_elem(struct bpf_map *map, void *key, u32 *value);
M
Martin KaFai Lau 已提交
1644 1645
int bpf_fd_htab_map_update_elem(struct bpf_map *map, struct file *map_file,
				void *key, void *value, u64 map_flags);
1646
int bpf_fd_htab_map_lookup_elem(struct bpf_map *map, void *key, u32 *value);
1647

1648
int bpf_get_file_flag(int flags);
1649
int bpf_check_uarg_tail_zero(bpfptr_t uaddr, size_t expected_size,
1650
			     size_t actual_size);
1651

1652 1653 1654 1655 1656 1657 1658 1659 1660 1661 1662 1663 1664 1665 1666 1667
/* memcpy that is used with 8-byte aligned pointers, power-of-8 size and
 * forced to use 'long' read/writes to try to atomically copy long counters.
 * Best-effort only.  No barriers here, since it _will_ race with concurrent
 * updates from BPF programs. Called from bpf syscall and mostly used with
 * size 8 or 16 bytes, so ask compiler to inline it.
 */
static inline void bpf_long_memcpy(void *dst, const void *src, u32 size)
{
	const long *lsrc = src;
	long *ldst = dst;

	size /= sizeof(long);
	while (size--)
		*ldst++ = *lsrc++;
}

1668
/* verify correctness of eBPF program */
1669
int bpf_check(struct bpf_prog **fp, union bpf_attr *attr, bpfptr_t uattr);
1670 1671

#ifndef CONFIG_BPF_JIT_ALWAYS_ON
1672
void bpf_patch_call_args(struct bpf_insn *insn, u32 stack_depth);
1673
#endif
1674

1675 1676
struct btf *bpf_get_btf_vmlinux(void);

1677
/* Map specifics */
1678
struct xdp_frame;
1679
struct sk_buff;
1680 1681
struct bpf_dtab_netdev;
struct bpf_cpu_map_entry;
1682

1683
void __dev_flush(void);
1684
int dev_xdp_enqueue(struct net_device *dev, struct xdp_frame *xdpf,
1685
		    struct net_device *dev_rx);
1686
int dev_map_enqueue(struct bpf_dtab_netdev *dst, struct xdp_frame *xdpf,
1687
		    struct net_device *dev_rx);
1688
int dev_map_enqueue_multi(struct xdp_frame *xdpf, struct net_device *dev_rx,
1689
			  struct bpf_map *map, bool exclude_ingress);
1690 1691
int dev_map_generic_redirect(struct bpf_dtab_netdev *dst, struct sk_buff *skb,
			     struct bpf_prog *xdp_prog);
1692 1693 1694
int dev_map_redirect_multi(struct net_device *dev, struct sk_buff *skb,
			   struct bpf_prog *xdp_prog, struct bpf_map *map,
			   bool exclude_ingress);
1695

1696
void __cpu_map_flush(void);
1697
int cpu_map_enqueue(struct bpf_cpu_map_entry *rcpu, struct xdp_frame *xdpf,
1698
		    struct net_device *dev_rx);
1699 1700
int cpu_map_generic_redirect(struct bpf_cpu_map_entry *rcpu,
			     struct sk_buff *skb);
1701

1702 1703 1704 1705 1706 1707 1708
/* Return map's numa specified by userspace */
static inline int bpf_map_attr_numa_node(const union bpf_attr *attr)
{
	return (attr->map_flags & BPF_F_NUMA_NODE) ?
		attr->numa_node : NUMA_NO_NODE;
}

1709
struct bpf_prog *bpf_prog_get_type_path(const char *name, enum bpf_prog_type type);
1710
int array_map_alloc_check(union bpf_attr *attr);
1711

1712 1713 1714 1715
int bpf_prog_test_run_xdp(struct bpf_prog *prog, const union bpf_attr *kattr,
			  union bpf_attr __user *uattr);
int bpf_prog_test_run_skb(struct bpf_prog *prog, const union bpf_attr *kattr,
			  union bpf_attr __user *uattr);
1716 1717 1718
int bpf_prog_test_run_tracing(struct bpf_prog *prog,
			      const union bpf_attr *kattr,
			      union bpf_attr __user *uattr);
1719 1720 1721
int bpf_prog_test_run_flow_dissector(struct bpf_prog *prog,
				     const union bpf_attr *kattr,
				     union bpf_attr __user *uattr);
1722 1723 1724
int bpf_prog_test_run_raw_tp(struct bpf_prog *prog,
			     const union bpf_attr *kattr,
			     union bpf_attr __user *uattr);
1725 1726 1727
int bpf_prog_test_run_sk_lookup(struct bpf_prog *prog,
				const union bpf_attr *kattr,
				union bpf_attr __user *uattr);
1728
bool bpf_prog_test_check_kfunc_call(u32 kfunc_id, struct module *owner);
1729 1730 1731
bool btf_ctx_access(int off, int size, enum bpf_access_type type,
		    const struct bpf_prog *prog,
		    struct bpf_insn_access_aux *info);
1732 1733 1734 1735 1736 1737 1738 1739 1740 1741 1742 1743 1744 1745 1746 1747 1748 1749 1750 1751 1752 1753 1754

static inline bool bpf_tracing_ctx_access(int off, int size,
					  enum bpf_access_type type)
{
	if (off < 0 || off >= sizeof(__u64) * MAX_BPF_FUNC_ARGS)
		return false;
	if (type != BPF_READ)
		return false;
	if (off % size != 0)
		return false;
	return true;
}

static inline bool bpf_tracing_btf_ctx_access(int off, int size,
					      enum bpf_access_type type,
					      const struct bpf_prog *prog,
					      struct bpf_insn_access_aux *info)
{
	if (!bpf_tracing_ctx_access(off, size, type))
		return false;
	return btf_ctx_access(off, size, type, prog, info);
}

1755
int btf_struct_access(struct bpf_verifier_log *log, const struct btf *btf,
1756 1757 1758
		      const struct btf_type *t, int off, int size,
		      enum bpf_access_type atype,
		      u32 *next_btf_id);
1759
bool btf_struct_ids_match(struct bpf_verifier_log *log,
1760 1761
			  const struct btf *btf, u32 id, int off,
			  const struct btf *need_btf, u32 need_type_id);
1762

A
Alexei Starovoitov 已提交
1763 1764 1765 1766 1767 1768
int btf_distill_func_proto(struct bpf_verifier_log *log,
			   struct btf *btf,
			   const struct btf_type *func_proto,
			   const char *func_name,
			   struct btf_func_model *m);

1769
struct bpf_reg_state;
1770 1771
int btf_check_subprog_arg_match(struct bpf_verifier_env *env, int subprog,
				struct bpf_reg_state *regs);
1772 1773 1774
int btf_check_kfunc_arg_match(struct bpf_verifier_env *env,
			      const struct btf *btf, u32 func_id,
			      struct bpf_reg_state *regs);
1775 1776
int btf_prepare_func_args(struct bpf_verifier_env *env, int subprog,
			  struct bpf_reg_state *reg);
1777
int btf_check_type_match(struct bpf_verifier_log *log, const struct bpf_prog *prog,
1778
			 struct btf *btf, const struct btf_type *t);
1779

1780
struct bpf_prog *bpf_prog_by_id(u32 id);
1781
struct bpf_link *bpf_link_by_id(u32 id);
1782

1783
const struct bpf_func_proto *bpf_base_func_proto(enum bpf_func_id func_id);
1784
void bpf_task_storage_free(struct task_struct *task);
1785 1786 1787 1788
bool bpf_prog_has_kfunc_call(const struct bpf_prog *prog);
const struct btf_func_model *
bpf_jit_find_kfunc_model(const struct bpf_prog *prog,
			 const struct bpf_insn *insn);
1789 1790 1791 1792 1793 1794 1795 1796
struct bpf_core_ctx {
	struct bpf_verifier_log *log;
	const struct btf *btf;
};

int bpf_core_apply(struct bpf_core_ctx *ctx, const struct bpf_core_relo *relo,
		   int relo_idx, void *insn);

1797
#else /* !CONFIG_BPF_SYSCALL */
1798 1799 1800 1801 1802
static inline struct bpf_prog *bpf_prog_get(u32 ufd)
{
	return ERR_PTR(-EOPNOTSUPP);
}

1803 1804
static inline struct bpf_prog *bpf_prog_get_type_dev(u32 ufd,
						     enum bpf_prog_type type,
1805
						     bool attach_drv)
1806 1807 1808 1809
{
	return ERR_PTR(-EOPNOTSUPP);
}

1810
static inline void bpf_prog_add(struct bpf_prog *prog, int i)
1811 1812
{
}
1813

1814 1815 1816 1817
static inline void bpf_prog_sub(struct bpf_prog *prog, int i)
{
}

1818 1819 1820
static inline void bpf_prog_put(struct bpf_prog *prog)
{
}
1821

1822
static inline void bpf_prog_inc(struct bpf_prog *prog)
1823 1824
{
}
1825

1826 1827 1828 1829 1830 1831
static inline struct bpf_prog *__must_check
bpf_prog_inc_not_zero(struct bpf_prog *prog)
{
	return ERR_PTR(-EOPNOTSUPP);
}

1832 1833 1834 1835 1836 1837 1838 1839 1840 1841 1842 1843 1844 1845 1846 1847 1848 1849 1850 1851 1852 1853 1854 1855 1856 1857 1858 1859 1860
static inline void bpf_link_init(struct bpf_link *link, enum bpf_link_type type,
				 const struct bpf_link_ops *ops,
				 struct bpf_prog *prog)
{
}

static inline int bpf_link_prime(struct bpf_link *link,
				 struct bpf_link_primer *primer)
{
	return -EOPNOTSUPP;
}

static inline int bpf_link_settle(struct bpf_link_primer *primer)
{
	return -EOPNOTSUPP;
}

static inline void bpf_link_cleanup(struct bpf_link_primer *primer)
{
}

static inline void bpf_link_inc(struct bpf_link *link)
{
}

static inline void bpf_link_put(struct bpf_link *link)
{
}

1861
static inline int bpf_obj_get_user(const char __user *pathname, int flags)
1862 1863 1864 1865
{
	return -EOPNOTSUPP;
}

1866 1867 1868 1869
static inline bool dev_map_can_have_prog(struct bpf_map *map)
{
	return false;
}
1870

1871
static inline void __dev_flush(void)
1872 1873
{
}
1874

1875
struct xdp_frame;
1876
struct bpf_dtab_netdev;
1877
struct bpf_cpu_map_entry;
1878

1879
static inline
1880
int dev_xdp_enqueue(struct net_device *dev, struct xdp_frame *xdpf,
1881 1882 1883 1884 1885
		    struct net_device *dev_rx)
{
	return 0;
}

1886
static inline
1887
int dev_map_enqueue(struct bpf_dtab_netdev *dst, struct xdp_frame *xdpf,
1888
		    struct net_device *dev_rx)
1889 1890 1891 1892
{
	return 0;
}

1893
static inline
1894
int dev_map_enqueue_multi(struct xdp_frame *xdpf, struct net_device *dev_rx,
1895 1896 1897 1898 1899
			  struct bpf_map *map, bool exclude_ingress)
{
	return 0;
}

1900 1901 1902 1903 1904 1905 1906 1907 1908
struct sk_buff;

static inline int dev_map_generic_redirect(struct bpf_dtab_netdev *dst,
					   struct sk_buff *skb,
					   struct bpf_prog *xdp_prog)
{
	return 0;
}

1909 1910 1911 1912 1913 1914 1915 1916
static inline
int dev_map_redirect_multi(struct net_device *dev, struct sk_buff *skb,
			   struct bpf_prog *xdp_prog, struct bpf_map *map,
			   bool exclude_ingress)
{
	return 0;
}

1917
static inline void __cpu_map_flush(void)
1918 1919 1920 1921
{
}

static inline int cpu_map_enqueue(struct bpf_cpu_map_entry *rcpu,
1922
				  struct xdp_frame *xdpf,
1923 1924 1925 1926
				  struct net_device *dev_rx)
{
	return 0;
}
1927

1928 1929 1930 1931 1932 1933
static inline int cpu_map_generic_redirect(struct bpf_cpu_map_entry *rcpu,
					   struct sk_buff *skb)
{
	return -EOPNOTSUPP;
}

1934 1935 1936 1937 1938
static inline bool cpu_map_prog_allowed(struct bpf_map *map)
{
	return false;
}

1939 1940 1941 1942 1943
static inline struct bpf_prog *bpf_prog_get_type_path(const char *name,
				enum bpf_prog_type type)
{
	return ERR_PTR(-EOPNOTSUPP);
}
1944 1945 1946 1947 1948 1949 1950 1951 1952 1953 1954 1955 1956 1957 1958

static inline int bpf_prog_test_run_xdp(struct bpf_prog *prog,
					const union bpf_attr *kattr,
					union bpf_attr __user *uattr)
{
	return -ENOTSUPP;
}

static inline int bpf_prog_test_run_skb(struct bpf_prog *prog,
					const union bpf_attr *kattr,
					union bpf_attr __user *uattr)
{
	return -ENOTSUPP;
}

1959 1960 1961 1962 1963 1964 1965
static inline int bpf_prog_test_run_tracing(struct bpf_prog *prog,
					    const union bpf_attr *kattr,
					    union bpf_attr __user *uattr)
{
	return -ENOTSUPP;
}

1966 1967 1968 1969 1970 1971
static inline int bpf_prog_test_run_flow_dissector(struct bpf_prog *prog,
						   const union bpf_attr *kattr,
						   union bpf_attr __user *uattr)
{
	return -ENOTSUPP;
}
1972

1973 1974 1975 1976 1977 1978 1979
static inline int bpf_prog_test_run_sk_lookup(struct bpf_prog *prog,
					      const union bpf_attr *kattr,
					      union bpf_attr __user *uattr)
{
	return -ENOTSUPP;
}

1980 1981
static inline bool bpf_prog_test_check_kfunc_call(u32 kfunc_id,
						  struct module *owner)
1982 1983 1984 1985
{
	return false;
}

1986 1987 1988
static inline void bpf_map_put(struct bpf_map *map)
{
}
1989 1990 1991 1992 1993

static inline struct bpf_prog *bpf_prog_by_id(u32 id)
{
	return ERR_PTR(-ENOTSUPP);
}
1994 1995 1996 1997 1998 1999

static inline const struct bpf_func_proto *
bpf_base_func_proto(enum bpf_func_id func_id)
{
	return NULL;
}
2000 2001 2002 2003

static inline void bpf_task_storage_free(struct task_struct *task)
{
}
2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015

static inline bool bpf_prog_has_kfunc_call(const struct bpf_prog *prog)
{
	return false;
}

static inline const struct btf_func_model *
bpf_jit_find_kfunc_model(const struct bpf_prog *prog,
			 const struct bpf_insn *insn)
{
	return NULL;
}
2016
#endif /* CONFIG_BPF_SYSCALL */
2017

2018 2019 2020
void __bpf_free_used_btfs(struct bpf_prog_aux *aux,
			  struct btf_mod_pair *used_btfs, u32 len);

2021 2022 2023 2024 2025 2026
static inline struct bpf_prog *bpf_prog_get_type(u32 ufd,
						 enum bpf_prog_type type)
{
	return bpf_prog_get_type_dev(ufd, type, false);
}

2027 2028 2029
void __bpf_free_used_maps(struct bpf_prog_aux *aux,
			  struct bpf_map **used_maps, u32 len);

2030 2031
bool bpf_prog_get_ok(struct bpf_prog *, enum bpf_prog_type *, bool);

2032 2033
int bpf_prog_offload_compile(struct bpf_prog *prog);
void bpf_prog_offload_destroy(struct bpf_prog *prog);
2034 2035
int bpf_prog_offload_info_fill(struct bpf_prog_info *info,
			       struct bpf_prog *prog);
2036

2037 2038
int bpf_map_offload_info_fill(struct bpf_map_info *info, struct bpf_map *map);

2039 2040 2041 2042 2043 2044 2045
int bpf_map_offload_lookup_elem(struct bpf_map *map, void *key, void *value);
int bpf_map_offload_update_elem(struct bpf_map *map,
				void *key, void *value, u64 flags);
int bpf_map_offload_delete_elem(struct bpf_map *map, void *key);
int bpf_map_offload_get_next_key(struct bpf_map *map,
				 void *key, void *next_key);

2046
bool bpf_offload_prog_map_match(struct bpf_prog *prog, struct bpf_map *map);
2047

2048
struct bpf_offload_dev *
2049
bpf_offload_dev_create(const struct bpf_prog_offload_ops *ops, void *priv);
2050
void bpf_offload_dev_destroy(struct bpf_offload_dev *offdev);
2051
void *bpf_offload_dev_priv(struct bpf_offload_dev *offdev);
2052 2053 2054 2055
int bpf_offload_dev_netdev_register(struct bpf_offload_dev *offdev,
				    struct net_device *netdev);
void bpf_offload_dev_netdev_unregister(struct bpf_offload_dev *offdev,
				       struct net_device *netdev);
2056
bool bpf_offload_dev_match(struct bpf_prog *prog, struct net_device *netdev);
2057

2058 2059 2060
#if defined(CONFIG_NET) && defined(CONFIG_BPF_SYSCALL)
int bpf_prog_offload_init(struct bpf_prog *prog, union bpf_attr *attr);

2061
static inline bool bpf_prog_is_dev_bound(const struct bpf_prog_aux *aux)
2062
{
2063
	return aux->offload_requested;
2064
}
2065 2066 2067 2068 2069 2070 2071 2072

static inline bool bpf_map_is_dev_bound(struct bpf_map *map)
{
	return unlikely(map->ops == &bpf_map_offload_ops);
}

struct bpf_map *bpf_map_offload_map_alloc(union bpf_attr *attr);
void bpf_map_offload_map_free(struct bpf_map *map);
2073 2074 2075
int bpf_prog_test_run_syscall(struct bpf_prog *prog,
			      const union bpf_attr *kattr,
			      union bpf_attr __user *uattr);
2076 2077 2078 2079 2080 2081

int sock_map_get_from_fd(const union bpf_attr *attr, struct bpf_prog *prog);
int sock_map_prog_detach(const union bpf_attr *attr, enum bpf_prog_type ptype);
int sock_map_update_elem_sys(struct bpf_map *map, void *key, void *value, u64 flags);
void sock_map_unhash(struct sock *sk);
void sock_map_close(struct sock *sk, long timeout);
2082 2083 2084 2085 2086 2087 2088 2089 2090 2091 2092
#else
static inline int bpf_prog_offload_init(struct bpf_prog *prog,
					union bpf_attr *attr)
{
	return -EOPNOTSUPP;
}

static inline bool bpf_prog_is_dev_bound(struct bpf_prog_aux *aux)
{
	return false;
}
2093 2094 2095 2096 2097 2098 2099 2100 2101 2102 2103 2104 2105 2106

static inline bool bpf_map_is_dev_bound(struct bpf_map *map)
{
	return false;
}

static inline struct bpf_map *bpf_map_offload_map_alloc(union bpf_attr *attr)
{
	return ERR_PTR(-EOPNOTSUPP);
}

static inline void bpf_map_offload_map_free(struct bpf_map *map)
{
}
2107 2108 2109 2110 2111 2112 2113

static inline int bpf_prog_test_run_syscall(struct bpf_prog *prog,
					    const union bpf_attr *kattr,
					    union bpf_attr __user *uattr)
{
	return -ENOTSUPP;
}
2114

2115
#ifdef CONFIG_BPF_SYSCALL
2116 2117
static inline int sock_map_get_from_fd(const union bpf_attr *attr,
				       struct bpf_prog *prog)
2118 2119 2120
{
	return -EINVAL;
}
2121 2122 2123 2124 2125 2126

static inline int sock_map_prog_detach(const union bpf_attr *attr,
				       enum bpf_prog_type ptype)
{
	return -EOPNOTSUPP;
}
2127 2128 2129 2130 2131 2132

static inline int sock_map_update_elem_sys(struct bpf_map *map, void *key, void *value,
					   u64 flags)
{
	return -EOPNOTSUPP;
}
2133 2134
#endif /* CONFIG_BPF_SYSCALL */
#endif /* CONFIG_NET && CONFIG_BPF_SYSCALL */
2135

2136 2137 2138 2139 2140 2141 2142 2143 2144 2145
#if defined(CONFIG_INET) && defined(CONFIG_BPF_SYSCALL)
void bpf_sk_reuseport_detach(struct sock *sk);
int bpf_fd_reuseport_array_lookup_elem(struct bpf_map *map, void *key,
				       void *value);
int bpf_fd_reuseport_array_update_elem(struct bpf_map *map, void *key,
				       void *value, u64 map_flags);
#else
static inline void bpf_sk_reuseport_detach(struct sock *sk)
{
}
2146

2147
#ifdef CONFIG_BPF_SYSCALL
2148 2149 2150 2151 2152 2153 2154 2155 2156 2157 2158 2159 2160 2161 2162
static inline int bpf_fd_reuseport_array_lookup_elem(struct bpf_map *map,
						     void *key, void *value)
{
	return -EOPNOTSUPP;
}

static inline int bpf_fd_reuseport_array_update_elem(struct bpf_map *map,
						     void *key, void *value,
						     u64 map_flags)
{
	return -EOPNOTSUPP;
}
#endif /* CONFIG_BPF_SYSCALL */
#endif /* defined(CONFIG_INET) && defined(CONFIG_BPF_SYSCALL) */

2163
/* verifier prototypes for helper functions called from eBPF programs */
2164 2165 2166
extern const struct bpf_func_proto bpf_map_lookup_elem_proto;
extern const struct bpf_func_proto bpf_map_update_elem_proto;
extern const struct bpf_func_proto bpf_map_delete_elem_proto;
M
Mauricio Vasquez B 已提交
2167 2168 2169
extern const struct bpf_func_proto bpf_map_push_elem_proto;
extern const struct bpf_func_proto bpf_map_pop_elem_proto;
extern const struct bpf_func_proto bpf_map_peek_elem_proto;
2170

2171
extern const struct bpf_func_proto bpf_get_prandom_u32_proto;
2172
extern const struct bpf_func_proto bpf_get_smp_processor_id_proto;
2173
extern const struct bpf_func_proto bpf_get_numa_node_id_proto;
2174
extern const struct bpf_func_proto bpf_tail_call_proto;
2175
extern const struct bpf_func_proto bpf_ktime_get_ns_proto;
2176
extern const struct bpf_func_proto bpf_ktime_get_boot_ns_proto;
2177 2178 2179
extern const struct bpf_func_proto bpf_get_current_pid_tgid_proto;
extern const struct bpf_func_proto bpf_get_current_uid_gid_proto;
extern const struct bpf_func_proto bpf_get_current_comm_proto;
2180
extern const struct bpf_func_proto bpf_get_stackid_proto;
Y
Yonghong Song 已提交
2181
extern const struct bpf_func_proto bpf_get_stack_proto;
2182
extern const struct bpf_func_proto bpf_get_task_stack_proto;
2183 2184
extern const struct bpf_func_proto bpf_get_stackid_proto_pe;
extern const struct bpf_func_proto bpf_get_stack_proto_pe;
2185
extern const struct bpf_func_proto bpf_sock_map_update_proto;
2186
extern const struct bpf_func_proto bpf_sock_hash_update_proto;
2187
extern const struct bpf_func_proto bpf_get_current_cgroup_id_proto;
2188
extern const struct bpf_func_proto bpf_get_current_ancestor_cgroup_id_proto;
2189 2190 2191 2192
extern const struct bpf_func_proto bpf_msg_redirect_hash_proto;
extern const struct bpf_func_proto bpf_msg_redirect_map_proto;
extern const struct bpf_func_proto bpf_sk_redirect_hash_proto;
extern const struct bpf_func_proto bpf_sk_redirect_map_proto;
2193 2194
extern const struct bpf_func_proto bpf_spin_lock_proto;
extern const struct bpf_func_proto bpf_spin_unlock_proto;
2195
extern const struct bpf_func_proto bpf_get_local_storage_proto;
2196 2197
extern const struct bpf_func_proto bpf_strtol_proto;
extern const struct bpf_func_proto bpf_strtoul_proto;
2198
extern const struct bpf_func_proto bpf_tcp_sock_proto;
M
Martin KaFai Lau 已提交
2199
extern const struct bpf_func_proto bpf_jiffies64_proto;
2200
extern const struct bpf_func_proto bpf_get_ns_current_pid_tgid_proto;
2201
extern const struct bpf_func_proto bpf_event_output_data_proto;
2202 2203 2204 2205 2206
extern const struct bpf_func_proto bpf_ringbuf_output_proto;
extern const struct bpf_func_proto bpf_ringbuf_reserve_proto;
extern const struct bpf_func_proto bpf_ringbuf_submit_proto;
extern const struct bpf_func_proto bpf_ringbuf_discard_proto;
extern const struct bpf_func_proto bpf_ringbuf_query_proto;
2207
extern const struct bpf_func_proto bpf_skc_to_tcp6_sock_proto;
2208 2209 2210
extern const struct bpf_func_proto bpf_skc_to_tcp_sock_proto;
extern const struct bpf_func_proto bpf_skc_to_tcp_timewait_sock_proto;
extern const struct bpf_func_proto bpf_skc_to_tcp_request_sock_proto;
2211
extern const struct bpf_func_proto bpf_skc_to_udp6_sock_proto;
2212
extern const struct bpf_func_proto bpf_skc_to_unix_sock_proto;
2213
extern const struct bpf_func_proto bpf_copy_from_user_proto;
A
Alan Maguire 已提交
2214
extern const struct bpf_func_proto bpf_snprintf_btf_proto;
F
Florent Revest 已提交
2215
extern const struct bpf_func_proto bpf_snprintf_proto;
H
Hao Luo 已提交
2216
extern const struct bpf_func_proto bpf_per_cpu_ptr_proto;
H
Hao Luo 已提交
2217
extern const struct bpf_func_proto bpf_this_cpu_ptr_proto;
2218
extern const struct bpf_func_proto bpf_ktime_get_coarse_ns_proto;
2219
extern const struct bpf_func_proto bpf_sock_from_file_proto;
2220
extern const struct bpf_func_proto bpf_get_socket_ptr_cookie_proto;
2221 2222
extern const struct bpf_func_proto bpf_task_storage_get_proto;
extern const struct bpf_func_proto bpf_task_storage_delete_proto;
2223
extern const struct bpf_func_proto bpf_for_each_map_elem_proto;
2224
extern const struct bpf_func_proto bpf_btf_find_by_name_kind_proto;
2225 2226
extern const struct bpf_func_proto bpf_sk_setsockopt_proto;
extern const struct bpf_func_proto bpf_sk_getsockopt_proto;
2227
extern const struct bpf_func_proto bpf_kallsyms_lookup_name_proto;
S
Song Liu 已提交
2228
extern const struct bpf_func_proto bpf_find_vma_proto;
J
Joanne Koong 已提交
2229
extern const struct bpf_func_proto bpf_loop_proto;
H
Hou Tao 已提交
2230
extern const struct bpf_func_proto bpf_strncmp_proto;
2231

2232 2233 2234
const struct bpf_func_proto *tracing_prog_func_proto(
  enum bpf_func_id func_id, const struct bpf_prog *prog);

2235 2236 2237
/* Shared helpers among cBPF and eBPF. */
void bpf_user_rnd_init_once(void);
u64 bpf_user_rnd_u32(u64 r1, u64 r2, u64 r3, u64 r4, u64 r5);
2238
u64 bpf_get_raw_cpu_id(u64 r1, u64 r2, u64 r3, u64 r4, u64 r5);
2239

2240
#if defined(CONFIG_NET)
2241 2242 2243
bool bpf_sock_common_is_valid_access(int off, int size,
				     enum bpf_access_type type,
				     struct bpf_insn_access_aux *info);
2244 2245 2246 2247 2248 2249 2250 2251
bool bpf_sock_is_valid_access(int off, int size, enum bpf_access_type type,
			      struct bpf_insn_access_aux *info);
u32 bpf_sock_convert_ctx_access(enum bpf_access_type type,
				const struct bpf_insn *si,
				struct bpf_insn *insn_buf,
				struct bpf_prog *prog,
				u32 *target_size);
#else
2252 2253 2254 2255 2256 2257
static inline bool bpf_sock_common_is_valid_access(int off, int size,
						   enum bpf_access_type type,
						   struct bpf_insn_access_aux *info)
{
	return false;
}
2258 2259 2260 2261 2262 2263 2264 2265 2266 2267 2268 2269 2270 2271 2272 2273
static inline bool bpf_sock_is_valid_access(int off, int size,
					    enum bpf_access_type type,
					    struct bpf_insn_access_aux *info)
{
	return false;
}
static inline u32 bpf_sock_convert_ctx_access(enum bpf_access_type type,
					      const struct bpf_insn *si,
					      struct bpf_insn *insn_buf,
					      struct bpf_prog *prog,
					      u32 *target_size)
{
	return 0;
}
#endif

2274
#ifdef CONFIG_INET
A
Alexei Starovoitov 已提交
2275 2276 2277 2278
struct sk_reuseport_kern {
	struct sk_buff *skb;
	struct sock *sk;
	struct sock *selected_sk;
2279
	struct sock *migrating_sk;
A
Alexei Starovoitov 已提交
2280 2281 2282 2283 2284
	void *data_end;
	u32 hash;
	u32 reuseport_id;
	bool bind_inany;
};
2285 2286 2287 2288 2289 2290 2291 2292
bool bpf_tcp_sock_is_valid_access(int off, int size, enum bpf_access_type type,
				  struct bpf_insn_access_aux *info);

u32 bpf_tcp_sock_convert_ctx_access(enum bpf_access_type type,
				    const struct bpf_insn *si,
				    struct bpf_insn *insn_buf,
				    struct bpf_prog *prog,
				    u32 *target_size);
2293 2294 2295 2296 2297 2298 2299 2300 2301

bool bpf_xdp_sock_is_valid_access(int off, int size, enum bpf_access_type type,
				  struct bpf_insn_access_aux *info);

u32 bpf_xdp_sock_convert_ctx_access(enum bpf_access_type type,
				    const struct bpf_insn *si,
				    struct bpf_insn *insn_buf,
				    struct bpf_prog *prog,
				    u32 *target_size);
2302 2303 2304 2305 2306 2307 2308 2309 2310 2311 2312 2313 2314 2315 2316 2317
#else
static inline bool bpf_tcp_sock_is_valid_access(int off, int size,
						enum bpf_access_type type,
						struct bpf_insn_access_aux *info)
{
	return false;
}

static inline u32 bpf_tcp_sock_convert_ctx_access(enum bpf_access_type type,
						  const struct bpf_insn *si,
						  struct bpf_insn *insn_buf,
						  struct bpf_prog *prog,
						  u32 *target_size)
{
	return 0;
}
2318 2319 2320 2321 2322 2323 2324 2325 2326 2327 2328 2329 2330 2331 2332
static inline bool bpf_xdp_sock_is_valid_access(int off, int size,
						enum bpf_access_type type,
						struct bpf_insn_access_aux *info)
{
	return false;
}

static inline u32 bpf_xdp_sock_convert_ctx_access(enum bpf_access_type type,
						  const struct bpf_insn *si,
						  struct bpf_insn *insn_buf,
						  struct bpf_prog *prog,
						  u32 *target_size)
{
	return 0;
}
2333 2334
#endif /* CONFIG_INET */

2335
enum bpf_text_poke_type {
2336 2337
	BPF_MOD_CALL,
	BPF_MOD_JUMP,
2338
};
2339

2340 2341 2342
int bpf_arch_text_poke(void *ip, enum bpf_text_poke_type t,
		       void *addr1, void *addr2);

J
Jiri Olsa 已提交
2343
struct btf_id_set;
2344
bool btf_id_set_contains(const struct btf_id_set *set, u32 id);
J
Jiri Olsa 已提交
2345

2346 2347
#define MAX_BPRINTF_VARARGS		12

2348 2349 2350
int bpf_bprintf_prepare(char *fmt, u32 fmt_size, const u64 *raw_args,
			u32 **bin_buf, u32 num_args);
void bpf_bprintf_cleanup(void);
2351

2352
#endif /* _LINUX_BPF_H */