bpf.h 74.5 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;
197 198 199 200 201 202 203 204 205 206 207
	/* 'Ownership' of program-containing map 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, JITed flag and xdp_has_frags flag.
	 */
	struct {
		spinlock_t lock;
		enum bpf_prog_type type;
		bool jited;
		bool xdp_has_frags;
	} owner;
208 209
};

210 211 212 213 214
static inline bool map_value_has_spin_lock(const struct bpf_map *map)
{
	return map->spin_lock_off >= 0;
}

215
static inline bool map_value_has_timer(const struct bpf_map *map)
216
{
217
	return map->timer_off >= 0;
218 219
}

220 221 222 223 224 225 226 227 228 229 230
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. */
231 232
static inline void copy_map_value(struct bpf_map *map, void *dst, void *src)
{
233 234
	u32 s_off = 0, s_sz = 0, t_off = 0, t_sz = 0;

235
	if (unlikely(map_value_has_spin_lock(map))) {
236 237 238 239 240 241
		s_off = map->spin_lock_off;
		s_sz = sizeof(struct bpf_spin_lock);
	} else if (unlikely(map_value_has_timer(map))) {
		t_off = map->timer_off;
		t_sz = sizeof(struct bpf_timer);
	}
242

243 244 245 246 247 248 249 250 251 252 253 254
	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);
255 256 257 258
	} else {
		memcpy(dst, src, map->value_size);
	}
}
259 260
void copy_map_value_locked(struct bpf_map *map, void *dst, void *src,
			   bool lock_src);
A
Alexei Starovoitov 已提交
261
void bpf_timer_cancel_and_free(void *timer);
262
int bpf_obj_name_cpy(char *dst, const char *src, unsigned int size);
263

264
struct bpf_offload_dev;
265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289
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);
}

290 291 292 293 294
static inline bool bpf_map_offload_neutral(const struct bpf_map *map)
{
	return map->map_type == BPF_MAP_TYPE_PERF_EVENT_ARRAY;
}

295 296
static inline bool bpf_map_support_seq_show(const struct bpf_map *map)
{
297 298
	return (map->btf_value_type_id || map->btf_vmlinux_value_type_id) &&
		map->ops->map_seq_show_elem;
299 300
}

301
int map_check_no_btf(const struct bpf_map *map,
302
		     const struct btf *btf,
303 304 305
		     const struct btf_type *key_type,
		     const struct btf_type *value_type);

306 307 308
bool bpf_map_meta_equal(const struct bpf_map *meta0,
			const struct bpf_map *meta1);

309 310
extern const struct bpf_map_ops bpf_map_offload_ops;

311 312 313 314 315 316 317 318 319 320 321 322 323 324
/* 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),

325 326 327
	/* 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
328 329
	MEM_RDONLY		= BIT(1 + BPF_BASE_TYPE_BITS),

330 331 332 333 334
	/* 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),

335 336 337 338
	/* MEM is in user address space. */
	MEM_USER		= BIT(3 + BPF_BASE_TYPE_BITS),

	__BPF_TYPE_LAST_FLAG	= MEM_USER,
339 340 341 342 343 344 345 346
};

/* 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))

347 348
/* function argument constraints */
enum bpf_arg_type {
349
	ARG_DONTCARE = 0,	/* unused argument in helper function */
350 351 352 353 354 355 356

	/* 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 */
357
	ARG_PTR_TO_UNINIT_MAP_VALUE,	/* pointer to valid memory used to store a map value */
358 359 360 361

	/* the following constraints used to prototype bpf_memcmp() and other
	 * functions that access data on eBPF program stack
	 */
362 363 364 365
	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.
366 367
				 */

368 369
	ARG_CONST_SIZE,		/* number of bytes accessed from memory */
	ARG_CONST_SIZE_OR_ZERO,	/* number of bytes accessed from memory or 0 */
370

371
	ARG_PTR_TO_CTX,		/* pointer to context */
372
	ARG_ANYTHING,		/* any (initialized) argument is ok */
373
	ARG_PTR_TO_SPIN_LOCK,	/* pointer to bpf_spin_lock */
374
	ARG_PTR_TO_SOCK_COMMON,	/* pointer to sock_common */
375 376
	ARG_PTR_TO_INT,		/* pointer to int */
	ARG_PTR_TO_LONG,	/* pointer to long */
377
	ARG_PTR_TO_SOCKET,	/* pointer to bpf_sock (fullsock) */
378
	ARG_PTR_TO_BTF_ID,	/* pointer to in-kernel struct */
379 380
	ARG_PTR_TO_ALLOC_MEM,	/* pointer to dynamically allocated memory */
	ARG_CONST_ALLOC_SIZE_OR_ZERO,	/* number of allocated bytes requested */
381
	ARG_PTR_TO_BTF_ID_SOCK_COMMON,	/* pointer to in-kernel sock_common or bpf-mirrored bpf_sock */
H
Hao Luo 已提交
382
	ARG_PTR_TO_PERCPU_BTF_ID,	/* pointer to in-kernel percpu type */
383
	ARG_PTR_TO_FUNC,	/* pointer to a bpf program function */
384
	ARG_PTR_TO_STACK,	/* pointer to stack */
385
	ARG_PTR_TO_CONST_STR,	/* pointer to a null terminated read-only string */
A
Alexei Starovoitov 已提交
386
	ARG_PTR_TO_TIMER,	/* pointer to bpf_timer */
387
	__BPF_ARG_TYPE_MAX,
388

389 390 391 392 393 394 395 396
	/* 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,

397 398 399 400
	/* 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,
401
};
402
static_assert(__BPF_ARG_TYPE_MAX <= BPF_BASE_TYPE_LIMIT);
403 404 405 406 407

/* type of values returned from helper functions */
enum bpf_return_type {
	RET_INTEGER,			/* function returns integer */
	RET_VOID,			/* function doesn't return anything */
408
	RET_PTR_TO_MAP_VALUE,		/* returns a pointer to map elem value */
409 410 411 412
	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
413
	RET_PTR_TO_MEM_OR_BTF_ID,	/* returns a pointer to a valid memory or a btf_id */
414
	RET_PTR_TO_BTF_ID,		/* returns a pointer to a btf_id */
415 416
	__BPF_RET_TYPE_MAX,

417 418 419 420 421
	/* 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,
422
	RET_PTR_TO_ALLOC_MEM_OR_NULL	= PTR_MAYBE_NULL | MEM_ALLOC | RET_PTR_TO_ALLOC_MEM,
423 424
	RET_PTR_TO_BTF_ID_OR_NULL	= PTR_MAYBE_NULL | RET_PTR_TO_BTF_ID,

425 426 427 428
	/* 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,
429
};
430
static_assert(__BPF_RET_TYPE_MAX <= BPF_BASE_TYPE_LIMIT);
431

432 433 434 435 436 437 438
/* 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;
439
	bool pkt_access;
440
	enum bpf_return_type ret_type;
441 442 443 444 445 446 447 448 449 450
	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];
	};
451 452 453 454 455 456 457 458 459 460
	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];
	};
461
	int *ret_btf_id; /* return value btf_id */
J
Jiri Olsa 已提交
462
	bool (*allowed)(const struct bpf_prog *prog);
463 464 465 466 467 468 469 470 471 472 473
};

/* 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
474 475
};

476
/* types of values stored in eBPF registers */
477 478 479 480 481 482 483 484 485
/* 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
 */
486 487
enum bpf_reg_type {
	NOT_INIT = 0,		 /* nothing was written into register */
488
	SCALAR_VALUE,		 /* reg doesn't contain a valid pointer */
489 490 491
	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 */
492
	PTR_TO_MAP_KEY,		 /* reg points to a map element key */
493
	PTR_TO_STACK,		 /* reg == frame_pointer + offset */
494
	PTR_TO_PACKET_META,	 /* skb->data - meta_len */
495
	PTR_TO_PACKET,		 /* reg points to skb->data */
496
	PTR_TO_PACKET_END,	 /* skb->data + headlen */
497
	PTR_TO_FLOW_KEYS,	 /* reg points to bpf_flow_keys */
498
	PTR_TO_SOCKET,		 /* reg points to struct bpf_sock */
499
	PTR_TO_SOCK_COMMON,	 /* reg points to sock_common */
500
	PTR_TO_TCP_SOCK,	 /* reg points to struct tcp_sock */
501
	PTR_TO_TP_BUFFER,	 /* reg points to a writable raw tp's buffer */
502
	PTR_TO_XDP_SOCK,	 /* reg points to struct xdp_sock */
503 504 505 506 507 508 509 510 511 512 513 514 515 516 517
	/* 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.
	 */
518
	PTR_TO_MEM,		 /* reg points to valid memory region */
H
Hao Luo 已提交
519
	PTR_TO_BUF,		 /* reg points to a read/write buffer */
H
Hao Luo 已提交
520
	PTR_TO_PERCPU_BTF_ID,	 /* reg points to a percpu kernel variable */
521
	PTR_TO_FUNC,		 /* reg points to a bpf program function */
522
	__BPF_REG_TYPE_MAX,
523

524 525 526 527 528 529 530
	/* 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,

531 532 533 534
	/* 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,
535
};
536
static_assert(__BPF_REG_TYPE_MAX <= BPF_BASE_TYPE_LIMIT);
537

538 539 540 541 542
/* The information passed from prog-specific *_is_valid_access
 * back to the verifier.
 */
struct bpf_insn_access_aux {
	enum bpf_reg_type reg_type;
543 544
	union {
		int ctx_field_size;
545 546 547 548
		struct {
			struct btf *btf;
			u32 btf_id;
		};
549 550
	};
	struct bpf_verifier_log *log; /* for verbose logs */
551 552
};

553 554 555 556 557 558
static inline void
bpf_ctx_record_field_size(struct bpf_insn_access_aux *aux, u32 size)
{
	aux->ctx_field_size = size;
}

559 560 561 562 563 564
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;
}

565 566 567 568 569
struct bpf_prog_ops {
	int (*test_run)(struct bpf_prog *prog, const union bpf_attr *kattr,
			union bpf_attr __user *uattr);
};

570 571
struct bpf_verifier_ops {
	/* return eBPF function prototype for verification */
572 573 574
	const struct bpf_func_proto *
	(*get_func_proto)(enum bpf_func_id func_id,
			  const struct bpf_prog *prog);
575 576 577 578

	/* return true if 'size' wide access at offset 'off' within bpf_context
	 * with 'type' (read or write) is allowed
	 */
579
	bool (*is_valid_access)(int off, int size, enum bpf_access_type type,
580
				const struct bpf_prog *prog,
581
				struct bpf_insn_access_aux *info);
582 583
	int (*gen_prologue)(struct bpf_insn *insn, bool direct_write,
			    const struct bpf_prog *prog);
584 585
	int (*gen_ld_abs)(const struct bpf_insn *orig,
			  struct bpf_insn *insn_buf);
586 587 588
	u32 (*convert_ctx_access)(enum bpf_access_type type,
				  const struct bpf_insn *src,
				  struct bpf_insn *dst,
589
				  struct bpf_prog *prog, u32 *target_size);
590
	int (*btf_struct_access)(struct bpf_verifier_log *log,
591
				 const struct btf *btf,
592 593
				 const struct btf_type *t, int off, int size,
				 enum bpf_access_type atype,
594
				 u32 *next_btf_id, enum bpf_type_flag *flag);
595 596
};

597
struct bpf_prog_offload_ops {
598
	/* verifier basic callbacks */
599 600
	int (*insn_hook)(struct bpf_verifier_env *env,
			 int insn_idx, int prev_insn_idx);
601
	int (*finalize)(struct bpf_verifier_env *env);
602 603 604 605 606
	/* 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 */
607 608
	int (*prepare)(struct bpf_prog *prog);
	int (*translate)(struct bpf_prog *prog);
609
	void (*destroy)(struct bpf_prog *prog);
610 611
};

612
struct bpf_prog_offload {
613 614
	struct bpf_prog		*prog;
	struct net_device	*netdev;
615
	struct bpf_offload_dev	*offdev;
616 617 618
	void			*dev_priv;
	struct list_head	offloads;
	bool			dev_state;
619
	bool			opt_failed;
620 621
	void			*jited_image;
	u32			jited_len;
622 623
};

624 625
enum bpf_cgroup_storage_type {
	BPF_CGROUP_STORAGE_SHARED,
626
	BPF_CGROUP_STORAGE_PERCPU,
627 628 629 630 631
	__BPF_CGROUP_STORAGE_MAX
};

#define MAX_BPF_CGROUP_STORAGE_TYPE __BPF_CGROUP_STORAGE_MAX

632 633 634 635 636
/* The longest tracepoint has 12 args.
 * See include/trace/bpf_probe.h
 */
#define MAX_BPF_FUNC_ARGS 12

637 638 639 640 641
/* 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
642 643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660
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)
661 662 663 664
/* 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)
665 666
/* Return the return value of fentry prog. Only used by bpf_struct_ops. */
#define BPF_TRAMP_F_RET_FENTRY_RET	BIT(4)
667

K
KP Singh 已提交
668 669 670
/* 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
 */
671
#define BPF_MAX_TRAMP_PROGS 38
K
KP Singh 已提交
672 673 674 675 676 677

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

A
Alexei Starovoitov 已提交
678 679 680 681 682 683 684 685 686 687 688 689 690 691 692 693 694 695 696 697
/* 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
698 699
struct bpf_tramp_image;
int arch_prepare_bpf_trampoline(struct bpf_tramp_image *tr, void *image, void *image_end,
700
				const struct btf_func_model *m, u32 flags,
K
KP Singh 已提交
701
				struct bpf_tramp_progs *tprogs,
A
Alexei Starovoitov 已提交
702 703
				void *orig_call);
/* these two functions are called from generated trampoline */
704
u64 notrace __bpf_prog_enter(struct bpf_prog *prog);
A
Alexei Starovoitov 已提交
705
void notrace __bpf_prog_exit(struct bpf_prog *prog, u64 start);
706
u64 notrace __bpf_prog_enter_sleepable(struct bpf_prog *prog);
707
void notrace __bpf_prog_exit_sleepable(struct bpf_prog *prog, u64 start);
A
Alexei Starovoitov 已提交
708 709
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
710

J
Jiri Olsa 已提交
711 712 713
struct bpf_ksym {
	unsigned long		 start;
	unsigned long		 end;
J
Jiri Olsa 已提交
714
	char			 name[KSYM_NAME_LEN];
715
	struct list_head	 lnode;
J
Jiri Olsa 已提交
716
	struct latch_tree_node	 tnode;
717
	bool			 prog;
J
Jiri Olsa 已提交
718 719
};

A
Alexei Starovoitov 已提交
720 721 722
enum bpf_tramp_prog_type {
	BPF_TRAMP_FENTRY,
	BPF_TRAMP_FEXIT,
K
KP Singh 已提交
723
	BPF_TRAMP_MODIFY_RETURN,
724 725
	BPF_TRAMP_MAX,
	BPF_TRAMP_REPLACE, /* more than MAX */
A
Alexei Starovoitov 已提交
726 727
};

A
Alexei Starovoitov 已提交
728 729 730 731 732 733 734 735 736 737 738 739
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
740 741 742 743 744 745 746 747 748 749
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;
750
		bool ftrace_managed;
A
Alexei Starovoitov 已提交
751
	} func;
752 753 754 755 756
	/* 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
757 758 759 760 761
	/* 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
762
	struct bpf_tramp_image *cur_image;
A
Alexei Starovoitov 已提交
763
	u64 selector;
764
	struct module *mod;
A
Alexei Starovoitov 已提交
765
};
B
Björn Töpel 已提交
766

767 768 769 770 771 772 773
struct bpf_attach_target_info {
	struct btf_func_model fmodel;
	long tgt_addr;
	const char *tgt_name;
	const struct btf_type *tgt_type;
};

774
#define BPF_DISPATCHER_MAX 48 /* Fits in 2048B */
B
Björn Töpel 已提交
775 776 777 778 779 780 781 782 783 784 785 786 787 788

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
789
	struct bpf_ksym ksym;
B
Björn Töpel 已提交
790 791
};

792
static __always_inline __nocfi unsigned int bpf_dispatcher_nop_func(
793 794 795 796 797 798 799
	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
800
#ifdef CONFIG_BPF_JIT
801 802
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);
803 804
struct bpf_trampoline *bpf_trampoline_get(u64 key,
					  struct bpf_attach_target_info *tgt_info);
A
Alexei Starovoitov 已提交
805
void bpf_trampoline_put(struct bpf_trampoline *tr);
806
int arch_prepare_bpf_dispatcher(void *image, s64 *funcs, int num_funcs);
J
Jiri Olsa 已提交
807 808 809 810 811 812 813 814 815 816 817
#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
818 819 820
}

#define DEFINE_BPF_DISPATCHER(name)					\
821
	noinline __nocfi unsigned int bpf_dispatcher_##name##_func(	\
B
Björn Töpel 已提交
822 823 824 825 826 827 828
		const void *ctx,					\
		const struct bpf_insn *insnsi,				\
		unsigned int (*bpf_func)(const void *,			\
					 const struct bpf_insn *))	\
	{								\
		return bpf_func(ctx, insnsi);				\
	}								\
829 830 831
	EXPORT_SYMBOL(bpf_dispatcher_##name##_func);			\
	struct bpf_dispatcher bpf_dispatcher_##name =			\
		BPF_DISPATCHER_INIT(bpf_dispatcher_##name);
B
Björn Töpel 已提交
832
#define DECLARE_BPF_DISPATCHER(name)					\
833
	unsigned int bpf_dispatcher_##name##_func(			\
B
Björn Töpel 已提交
834 835 836 837
		const void *ctx,					\
		const struct bpf_insn *insnsi,				\
		unsigned int (*bpf_func)(const void *,			\
					 const struct bpf_insn *));	\
838 839 840
	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
841 842
void bpf_dispatcher_change_prog(struct bpf_dispatcher *d, struct bpf_prog *from,
				struct bpf_prog *to);
J
Jiri Olsa 已提交
843
/* Called only from JIT-enabled code, so there's no need for stubs. */
J
Jiri Olsa 已提交
844
void *bpf_jit_alloc_exec_page(void);
J
Jiri Olsa 已提交
845 846
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
847 848
void bpf_ksym_add(struct bpf_ksym *ksym);
void bpf_ksym_del(struct bpf_ksym *ksym);
849 850
int bpf_jit_charge_modmem(u32 size);
void bpf_jit_uncharge_modmem(u32 size);
851
bool bpf_prog_has_trampoline(const struct bpf_prog *prog);
A
Alexei Starovoitov 已提交
852
#else
853 854
static inline int bpf_trampoline_link_prog(struct bpf_prog *prog,
					   struct bpf_trampoline *tr)
A
Alexei Starovoitov 已提交
855 856 857
{
	return -ENOTSUPP;
}
858 859
static inline int bpf_trampoline_unlink_prog(struct bpf_prog *prog,
					     struct bpf_trampoline *tr)
A
Alexei Starovoitov 已提交
860 861 862
{
	return -ENOTSUPP;
}
863 864 865 866 867
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
868
static inline void bpf_trampoline_put(struct bpf_trampoline *tr) {}
B
Björn Töpel 已提交
869 870
#define DEFINE_BPF_DISPATCHER(name)
#define DECLARE_BPF_DISPATCHER(name)
871
#define BPF_DISPATCHER_FUNC(name) bpf_dispatcher_nop_func
B
Björn Töpel 已提交
872 873 874 875
#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) {}
876 877 878 879
static inline bool is_bpf_image_address(unsigned long address)
{
	return false;
}
880 881 882 883
static inline bool bpf_prog_has_trampoline(const struct bpf_prog *prog)
{
	return false;
}
A
Alexei Starovoitov 已提交
884 885
#endif

886
struct bpf_func_info_aux {
887
	u16 linkage;
888 889 890
	bool unreliable;
};

891 892 893 894 895 896
enum bpf_jit_poke_reason {
	BPF_POKE_REASON_TAIL_CALL,
};

/* Descriptor of pokes pointing /into/ the JITed image. */
struct bpf_jit_poke_descriptor {
897
	void *tailcall_target;
898 899
	void *tailcall_bypass;
	void *bypass_addr;
900
	void *aux;
901 902 903 904 905 906
	union {
		struct {
			struct bpf_map *map;
			u32 key;
		} tail_call;
	};
907
	bool tailcall_target_stable;
908 909
	u8 adj_off;
	u16 reason;
910
	u32 insn_idx;
911 912
};

913 914 915 916
/* reg_type info for ctx arguments */
struct bpf_ctx_arg_aux {
	u32 offset;
	enum bpf_reg_type reg_type;
917
	u32 btf_id;
918 919
};

920 921 922 923 924
struct btf_mod_pair {
	struct btf *btf;
	struct module *module;
};

925 926
struct bpf_kfunc_desc_tab;

927
struct bpf_prog_aux {
928
	atomic64_t refcnt;
929
	u32 used_map_cnt;
930
	u32 used_btf_cnt;
931
	u32 max_ctx_offset;
932
	u32 max_pkt_offset;
933
	u32 max_tp_access;
934
	u32 stack_depth;
M
Martin KaFai Lau 已提交
935
	u32 id;
936 937
	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 */
938
	u32 attach_btf_id; /* in-kernel BTF type id to attach to */
939
	u32 ctx_arg_info_size;
940 941
	u32 max_rdonly_access;
	u32 max_rdwr_access;
942
	struct btf *attach_btf;
943
	const struct bpf_ctx_arg_aux *ctx_arg_info;
944 945 946
	struct mutex dst_mutex; /* protects dst_* pointers below, *after* prog becomes visible */
	struct bpf_prog *dst_prog;
	struct bpf_trampoline *dst_trampoline;
947 948
	enum bpf_prog_type saved_dst_prog_type;
	enum bpf_attach_type saved_dst_attach_type;
949
	bool verifier_zext; /* Zero extensions has been inserted by verifier. */
950
	bool offload_requested;
951
	bool attach_btf_trace; /* true if attaching to BTF-enabled raw tp */
952
	bool func_proto_unreliable;
953
	bool sleepable;
954
	bool tail_call_reachable;
955
	bool xdp_has_frags;
A
Alexei Starovoitov 已提交
956
	struct hlist_node tramp_hlist;
957 958 959 960
	/* 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;
961 962
	struct bpf_prog **func;
	void *jit_data; /* JIT specific data. arch dependent */
963
	struct bpf_jit_poke_descriptor *poke_tab;
964
	struct bpf_kfunc_desc_tab *kfunc_tab;
965
	struct bpf_kfunc_btf_tab *kfunc_btf_tab;
966
	u32 size_poke_tab;
J
Jiri Olsa 已提交
967
	struct bpf_ksym ksym;
968
	const struct bpf_prog_ops *ops;
969
	struct bpf_map **used_maps;
970
	struct mutex used_maps_mutex; /* mutex for used_maps and used_map_cnt */
971
	struct btf_mod_pair *used_btfs;
972
	struct bpf_prog *prog;
973
	struct user_struct *user;
974
	u64 load_time; /* ns since boottime */
975
	u32 verified_insns;
976
	struct bpf_map *cgroup_storage[MAX_BPF_CGROUP_STORAGE_TYPE];
977
	char name[BPF_OBJ_NAME_LEN];
978 979 980
#ifdef CONFIG_SECURITY
	void *security;
#endif
981
	struct bpf_prog_offload *offload;
Y
Yonghong Song 已提交
982
	struct btf *btf;
983
	struct bpf_func_info *func_info;
984
	struct bpf_func_info_aux *func_info_aux;
M
Martin KaFai Lau 已提交
985 986 987 988 989 990 991 992 993 994 995 996 997 998 999
	/* 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;
1000
	u32 func_info_cnt;
M
Martin KaFai Lau 已提交
1001 1002 1003 1004 1005 1006
	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;
1007 1008
	u32 num_exentries;
	struct exception_table_entry *extable;
1009 1010 1011 1012
	union {
		struct work_struct work;
		struct rcu_head	rcu;
	};
1013 1014
};

1015
struct bpf_array_aux {
1016 1017 1018 1019 1020
	/* 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;
1021 1022
};

1023 1024 1025 1026 1027 1028 1029 1030 1031 1032 1033 1034
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);
1035
	int (*detach)(struct bpf_link *link);
1036 1037 1038 1039 1040 1041 1042 1043 1044 1045 1046 1047 1048 1049
	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;
};

1050
struct bpf_struct_ops_value;
1051 1052 1053 1054 1055 1056 1057 1058
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);
1059 1060 1061 1062 1063
	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);
1064
	const struct btf_type *type;
1065
	const struct btf_type *value_type;
1066 1067 1068
	const char *name;
	struct btf_func_model func_models[BPF_STRUCT_OPS_MAX_NR_MEMBERS];
	u32 type_id;
1069
	u32 value_id;
1070 1071 1072
};

#if defined(CONFIG_BPF_JIT) && defined(CONFIG_BPF_SYSCALL)
1073
#define BPF_MODULE_OWNER ((void *)((0xeB9FUL << 2) + POISON_POINTER_DELTA))
1074
const struct bpf_struct_ops *bpf_struct_ops_find(u32 type_id);
1075
void bpf_struct_ops_init(struct btf *btf, struct bpf_verifier_log *log);
1076 1077 1078 1079
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);
1080 1081 1082 1083
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);
1084 1085 1086 1087 1088 1089 1090 1091 1092 1093 1094 1095 1096 1097
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);
}
1098 1099 1100 1101 1102 1103 1104 1105 1106 1107 1108 1109 1110 1111 1112 1113

#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
1114 1115 1116 1117 1118
#else
static inline const struct bpf_struct_ops *bpf_struct_ops_find(u32 type_id)
{
	return NULL;
}
1119 1120 1121 1122
static inline void bpf_struct_ops_init(struct btf *btf,
				       struct bpf_verifier_log *log)
{
}
1123 1124 1125 1126 1127 1128 1129 1130 1131 1132 1133 1134 1135 1136
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;
}
1137 1138
#endif

1139 1140 1141
struct bpf_array {
	struct bpf_map map;
	u32 elem_size;
1142
	u32 index_mask;
1143
	struct bpf_array_aux *aux;
1144 1145
	union {
		char value[0] __aligned(8);
1146
		void *ptrs[0] __aligned(8);
1147
		void __percpu *pptrs[0] __aligned(8);
1148 1149
	};
};
1150

1151
#define BPF_COMPLEXITY_LIMIT_INSNS      1000000 /* yes. 1M insns */
1152
#define MAX_TAIL_CALL_CNT 33
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 1180 1181 1182
#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);
}

1183 1184 1185 1186 1187 1188 1189
struct bpf_event_entry {
	struct perf_event *event;
	struct file *perf_file;
	struct file *map_file;
	struct rcu_head rcu;
};

1190 1191 1192 1193 1194 1195 1196 1197
static inline bool map_type_contains_progs(struct bpf_map *map)
{
	return map->map_type == BPF_MAP_TYPE_PROG_ARRAY ||
	       map->map_type == BPF_MAP_TYPE_DEVMAP ||
	       map->map_type == BPF_MAP_TYPE_CPUMAP;
}

bool bpf_prog_map_compatible(struct bpf_map *map, const struct bpf_prog *fp);
1198
int bpf_prog_calc_tag(struct bpf_prog *fp);
1199

1200
const struct bpf_func_proto *bpf_get_trace_printk_proto(void);
1201
const struct bpf_func_proto *bpf_get_trace_vprintk_proto(void);
1202 1203

typedef unsigned long (*bpf_ctx_copy_t)(void *dst, const void *src,
1204
					unsigned long off, unsigned long len);
1205 1206 1207 1208 1209
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);
1210 1211 1212

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);
1213

1214 1215 1216
/* an array of programs to be executed under rcu_lock.
 *
 * Typical usage:
1217
 * ret = BPF_PROG_RUN_ARRAY(&bpf_prog_array, ctx, bpf_prog_run);
1218 1219 1220 1221 1222 1223 1224 1225
 *
 * 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.
 */
1226 1227
struct bpf_prog_array_item {
	struct bpf_prog *prog;
1228 1229 1230 1231
	union {
		struct bpf_cgroup_storage *cgroup_storage[MAX_BPF_CGROUP_STORAGE_TYPE];
		u64 bpf_cookie;
	};
1232 1233
};

1234 1235
struct bpf_prog_array {
	struct rcu_head rcu;
1236
	struct bpf_prog_array_item items[];
1237 1238
};

1239 1240 1241 1242 1243 1244 1245 1246 1247 1248 1249 1250 1251
struct bpf_empty_prog_array {
	struct bpf_prog_array hdr;
	struct bpf_prog *null_prog;
};

/* to avoid allocating empty bpf_prog_array for cgroups that
 * don't have bpf program attached use one global 'bpf_empty_prog_array'
 * It will not be modified the caller of bpf_prog_array_alloc()
 * (since caller requested prog_cnt == 0)
 * that pointer should be 'freed' by bpf_prog_array_free()
 */
extern struct bpf_empty_prog_array bpf_empty_prog_array;

1252
struct bpf_prog_array *bpf_prog_array_alloc(u32 prog_cnt, gfp_t flags);
1253 1254
void bpf_prog_array_free(struct bpf_prog_array *progs);
int bpf_prog_array_length(struct bpf_prog_array *progs);
1255
bool bpf_prog_array_is_empty(struct bpf_prog_array *array);
1256
int bpf_prog_array_copy_to_user(struct bpf_prog_array *progs,
1257
				__u32 __user *prog_ids, u32 cnt);
1258

1259
void bpf_prog_array_delete_safe(struct bpf_prog_array *progs,
1260
				struct bpf_prog *old_prog);
1261 1262 1263
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);
1264
int bpf_prog_array_copy_info(struct bpf_prog_array *array,
1265 1266
			     u32 *prog_ids, u32 request_cnt,
			     u32 *prog_cnt);
1267
int bpf_prog_array_copy(struct bpf_prog_array *old_array,
1268 1269
			struct bpf_prog *exclude_prog,
			struct bpf_prog *include_prog,
1270
			u64 bpf_cookie,
1271 1272
			struct bpf_prog_array **new_array);

1273 1274 1275 1276
struct bpf_run_ctx {};

struct bpf_cg_run_ctx {
	struct bpf_run_ctx run_ctx;
1277
	const struct bpf_prog_array_item *prog_item;
1278
	int retval;
1279 1280
};

1281 1282 1283 1284 1285
struct bpf_trace_run_ctx {
	struct bpf_run_ctx run_ctx;
	u64 bpf_cookie;
};

1286 1287 1288 1289 1290 1291 1292 1293 1294 1295 1296 1297 1298 1299 1300 1301 1302 1303
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
}

1304 1305 1306 1307 1308
/* 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)

1309 1310
typedef u32 (*bpf_prog_run_fn)(const struct bpf_prog *prog, const void *ctx);

1311
static __always_inline int
1312 1313
BPF_PROG_RUN_ARRAY_CG_FLAGS(const struct bpf_prog_array __rcu *array_rcu,
			    const void *ctx, bpf_prog_run_fn run_prog,
1314
			    int retval, u32 *ret_flags)
1315 1316 1317 1318 1319 1320 1321 1322
{
	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 func_ret;

1323
	run_ctx.retval = retval;
1324 1325 1326 1327 1328 1329 1330 1331
	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);
1332
		if (!(func_ret & 1) && !IS_ERR_VALUE((long)run_ctx.retval))
1333
			run_ctx.retval = -EPERM;
1334 1335 1336 1337 1338 1339
		*(ret_flags) |= (func_ret >> 1);
		item++;
	}
	bpf_reset_run_ctx(old_run_ctx);
	rcu_read_unlock();
	migrate_enable();
1340
	return run_ctx.retval;
1341 1342
}

1343
static __always_inline int
1344
BPF_PROG_RUN_ARRAY_CG(const struct bpf_prog_array __rcu *array_rcu,
1345 1346
		      const void *ctx, bpf_prog_run_fn run_prog,
		      int retval)
1347 1348 1349 1350 1351 1352 1353
{
	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;

1354
	run_ctx.retval = retval;
1355 1356 1357 1358 1359 1360 1361
	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;
1362
		if (!run_prog(prog, ctx) && !IS_ERR_VALUE((long)run_ctx.retval))
1363
			run_ctx.retval = -EPERM;
1364 1365 1366 1367 1368
		item++;
	}
	bpf_reset_run_ctx(old_run_ctx);
	rcu_read_unlock();
	migrate_enable();
1369
	return run_ctx.retval;
1370 1371 1372 1373 1374 1375 1376 1377 1378
}

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;
1379 1380
	struct bpf_run_ctx *old_run_ctx;
	struct bpf_trace_run_ctx run_ctx;
1381 1382 1383 1384 1385 1386 1387
	u32 ret = 1;

	migrate_disable();
	rcu_read_lock();
	array = rcu_dereference(array_rcu);
	if (unlikely(!array))
		goto out;
1388
	old_run_ctx = bpf_set_run_ctx(&run_ctx.run_ctx);
1389 1390
	item = &array->items[0];
	while ((prog = READ_ONCE(item->prog))) {
1391
		run_ctx.bpf_cookie = item->bpf_cookie;
1392 1393 1394
		ret &= run_prog(prog, ctx);
		item++;
	}
1395
	bpf_reset_run_ctx(old_run_ctx);
1396 1397 1398 1399 1400
out:
	rcu_read_unlock();
	migrate_enable();
	return ret;
}
1401

1402 1403 1404 1405 1406 1407 1408 1409 1410 1411 1412 1413 1414 1415 1416 1417 1418 1419 1420 1421
/* 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
1422
 *   3: -err              skb should be dropped
1423 1424 1425
 */
#define BPF_PROG_CGROUP_INET_EGRESS_RUN_ARRAY(array, ctx, func)		\
	({						\
1426 1427 1428
		u32 _flags = 0;				\
		bool _cn;				\
		u32 _ret;				\
1429
		_ret = BPF_PROG_RUN_ARRAY_CG_FLAGS(array, ctx, func, 0, &_flags); \
1430
		_cn = _flags & BPF_RET_SET_CN;		\
1431 1432
		if (_ret && !IS_ERR_VALUE((long)_ret))	\
			_ret = -EFAULT;			\
1433
		if (!_ret)				\
1434 1435
			_ret = (_cn ? NET_XMIT_CN : NET_XMIT_SUCCESS);	\
		else					\
1436
			_ret = (_cn ? NET_XMIT_DROP : _ret);		\
1437 1438 1439
		_ret;					\
	})

1440
#ifdef CONFIG_BPF_SYSCALL
1441
DECLARE_PER_CPU(int, bpf_prog_active);
1442
extern struct mutex bpf_stats_enabled_mutex;
1443

1444 1445 1446 1447 1448 1449 1450 1451 1452
/*
 * 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();
1453
	this_cpu_inc(bpf_prog_active);
1454 1455 1456 1457
}

static inline void bpf_enable_instrumentation(void)
{
1458
	this_cpu_dec(bpf_prog_active);
1459 1460 1461
	migrate_enable();
}

1462 1463
extern const struct file_operations bpf_map_fops;
extern const struct file_operations bpf_prog_fops;
Y
Yonghong Song 已提交
1464
extern const struct file_operations bpf_iter_fops;
1465

A
Alexei Starovoitov 已提交
1466
#define BPF_PROG_TYPE(_id, _name, prog_ctx_type, kern_ctx_type) \
1467 1468
	extern const struct bpf_prog_ops _name ## _prog_ops; \
	extern const struct bpf_verifier_ops _name ## _verifier_ops;
1469 1470
#define BPF_MAP_TYPE(_id, _ops) \
	extern const struct bpf_map_ops _ops;
1471
#define BPF_LINK_TYPE(_id, _name)
1472 1473
#include <linux/bpf_types.h>
#undef BPF_PROG_TYPE
1474
#undef BPF_MAP_TYPE
1475
#undef BPF_LINK_TYPE
1476

1477
extern const struct bpf_prog_ops bpf_offload_prog_ops;
1478 1479 1480
extern const struct bpf_verifier_ops tc_cls_act_analyzer_ops;
extern const struct bpf_verifier_ops xdp_analyzer_ops;

1481
struct bpf_prog *bpf_prog_get(u32 ufd);
1482
struct bpf_prog *bpf_prog_get_type_dev(u32 ufd, enum bpf_prog_type type,
1483
				       bool attach_drv);
1484
void bpf_prog_add(struct bpf_prog *prog, int i);
1485
void bpf_prog_sub(struct bpf_prog *prog, int i);
1486
void bpf_prog_inc(struct bpf_prog *prog);
1487
struct bpf_prog * __must_check bpf_prog_inc_not_zero(struct bpf_prog *prog);
1488 1489
void bpf_prog_put(struct bpf_prog *prog);

1490
void bpf_prog_free_id(struct bpf_prog *prog, bool do_idr_lock);
1491
void bpf_map_free_id(struct bpf_map *map, bool do_idr_lock);
1492

1493
struct bpf_map *bpf_map_get(u32 ufd);
1494
struct bpf_map *bpf_map_get_with_uref(u32 ufd);
1495
struct bpf_map *__bpf_map_get(struct fd f);
1496 1497 1498
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);
1499
void bpf_map_put_with_uref(struct bpf_map *map);
1500
void bpf_map_put(struct bpf_map *map);
1501 1502
void *bpf_map_area_alloc(u64 size, int numa_node);
void *bpf_map_area_mmapable_alloc(u64 size, int numa_node);
1503
void bpf_map_area_free(void *base);
1504
bool bpf_map_write_active(const struct bpf_map *map);
1505
void bpf_map_init_from_attr(struct bpf_map *map, union bpf_attr *attr);
1506 1507
int  generic_map_lookup_batch(struct bpf_map *map,
			      const union bpf_attr *attr,
1508 1509 1510 1511 1512 1513
			      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,
1514
			      union bpf_attr __user *uattr);
Y
Yonghong Song 已提交
1515
struct bpf_map *bpf_map_get_curr_or_next(u32 *id);
A
Alexei Starovoitov 已提交
1516
struct bpf_prog *bpf_prog_get_curr_or_next(u32 *id);
1517

1518 1519 1520 1521 1522 1523 1524 1525 1526 1527 1528 1529 1530 1531 1532 1533 1534 1535 1536 1537 1538 1539 1540 1541 1542 1543 1544 1545
#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

1546 1547
extern int sysctl_unprivileged_bpf_disabled;

A
Alexei Starovoitov 已提交
1548 1549 1550 1551 1552
static inline bool bpf_allow_ptr_leaks(void)
{
	return perfmon_capable();
}

1553 1554 1555 1556 1557
static inline bool bpf_allow_uninit_stack(void)
{
	return perfmon_capable();
}

1558 1559 1560 1561 1562
static inline bool bpf_allow_ptr_to_map_access(void)
{
	return perfmon_capable();
}

A
Alexei Starovoitov 已提交
1563 1564 1565 1566 1567 1568 1569 1570 1571 1572
static inline bool bpf_bypass_spec_v1(void)
{
	return perfmon_capable();
}

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

1573
int bpf_map_new_fd(struct bpf_map *map, int flags);
1574 1575
int bpf_prog_new_fd(struct bpf_prog *prog);

1576
void bpf_link_init(struct bpf_link *link, enum bpf_link_type type,
A
Andrii Nakryiko 已提交
1577 1578 1579 1580
		   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);
1581 1582 1583
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);
1584
struct file *bpf_link_new_file(struct bpf_link *link, int *reserved_fd);
1585 1586
struct bpf_link *bpf_link_get_from_fd(u32 ufd);

1587
int bpf_obj_pin_user(u32 ufd, const char __user *pathname);
1588
int bpf_obj_get_user(const char __user *pathname, int flags);
1589

1590
#define BPF_ITER_FUNC_PREFIX "bpf_iter_"
1591
#define DEFINE_BPF_ITER_FUNC(target, args...)			\
1592 1593
	extern int bpf_iter_ ## target(args);			\
	int __init bpf_iter_ ## target(args) { return 0; }
1594

1595
struct bpf_iter_aux_info {
1596
	struct bpf_map *map;
1597 1598
};

1599 1600 1601 1602
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);
1603 1604 1605 1606
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);
1607 1608 1609
typedef const struct bpf_func_proto *
(*bpf_iter_get_func_proto_t)(enum bpf_func_id func_id,
			     const struct bpf_prog *prog);
1610

1611 1612 1613 1614
enum bpf_iter_feature {
	BPF_ITER_RESCHED	= BIT(0),
};

1615
#define BPF_ITER_CTX_ARG_MAX 2
1616 1617
struct bpf_iter_reg {
	const char *target;
1618 1619
	bpf_iter_attach_target_t attach_target;
	bpf_iter_detach_target_t detach_target;
1620 1621
	bpf_iter_show_fdinfo_t show_fdinfo;
	bpf_iter_fill_link_info_t fill_link_info;
1622
	bpf_iter_get_func_proto_t get_func_proto;
1623
	u32 ctx_arg_info_size;
1624
	u32 feature;
1625
	struct bpf_ctx_arg_aux ctx_arg_info[BPF_ITER_CTX_ARG_MAX];
1626
	const struct bpf_iter_seq_info *seq_info;
1627 1628
};

1629 1630 1631 1632 1633 1634
struct bpf_iter_meta {
	__bpf_md_ptr(struct seq_file *, seq);
	u64 session_id;
	u64 seq_num;
};

1635 1636 1637 1638 1639 1640 1641
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);
};

1642
int bpf_iter_reg_target(const struct bpf_iter_reg *reg_info);
1643
void bpf_iter_unreg_target(const struct bpf_iter_reg *reg_info);
1644
bool bpf_iter_prog_supported(struct bpf_prog *prog);
1645 1646
const struct bpf_func_proto *
bpf_iter_get_func_proto(enum bpf_func_id func_id, const struct bpf_prog *prog);
1647
int bpf_iter_link_attach(const union bpf_attr *attr, bpfptr_t uattr, struct bpf_prog *prog);
1648
int bpf_iter_new_fd(struct bpf_link *link);
Y
Yonghong Song 已提交
1649
bool bpf_link_is_iter(struct bpf_link *link);
1650 1651
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);
1652 1653 1654 1655
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);
1656

1657 1658 1659 1660
int map_set_for_each_callback_args(struct bpf_verifier_env *env,
				   struct bpf_func_state *caller,
				   struct bpf_func_state *callee);

1661 1662 1663 1664 1665 1666
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);
1667

1668
int bpf_stackmap_copy(struct bpf_map *map, void *key, void *value);
1669

1670 1671
int bpf_fd_array_map_update_elem(struct bpf_map *map, struct file *map_file,
				 void *key, void *value, u64 map_flags);
1672
int bpf_fd_array_map_lookup_elem(struct bpf_map *map, void *key, u32 *value);
M
Martin KaFai Lau 已提交
1673 1674
int bpf_fd_htab_map_update_elem(struct bpf_map *map, struct file *map_file,
				void *key, void *value, u64 map_flags);
1675
int bpf_fd_htab_map_lookup_elem(struct bpf_map *map, void *key, u32 *value);
1676

1677
int bpf_get_file_flag(int flags);
1678
int bpf_check_uarg_tail_zero(bpfptr_t uaddr, size_t expected_size,
1679
			     size_t actual_size);
1680

1681 1682 1683 1684 1685 1686 1687 1688 1689 1690 1691 1692 1693 1694 1695 1696
/* 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++;
}

1697
/* verify correctness of eBPF program */
1698
int bpf_check(struct bpf_prog **fp, union bpf_attr *attr, bpfptr_t uattr);
1699 1700

#ifndef CONFIG_BPF_JIT_ALWAYS_ON
1701
void bpf_patch_call_args(struct bpf_insn *insn, u32 stack_depth);
1702
#endif
1703

1704 1705
struct btf *bpf_get_btf_vmlinux(void);

1706
/* Map specifics */
1707
struct xdp_frame;
1708
struct sk_buff;
1709 1710
struct bpf_dtab_netdev;
struct bpf_cpu_map_entry;
1711

1712
void __dev_flush(void);
1713
int dev_xdp_enqueue(struct net_device *dev, struct xdp_frame *xdpf,
1714
		    struct net_device *dev_rx);
1715
int dev_map_enqueue(struct bpf_dtab_netdev *dst, struct xdp_frame *xdpf,
1716
		    struct net_device *dev_rx);
1717
int dev_map_enqueue_multi(struct xdp_frame *xdpf, struct net_device *dev_rx,
1718
			  struct bpf_map *map, bool exclude_ingress);
1719 1720
int dev_map_generic_redirect(struct bpf_dtab_netdev *dst, struct sk_buff *skb,
			     struct bpf_prog *xdp_prog);
1721 1722 1723
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);
1724

1725
void __cpu_map_flush(void);
1726
int cpu_map_enqueue(struct bpf_cpu_map_entry *rcpu, struct xdp_frame *xdpf,
1727
		    struct net_device *dev_rx);
1728 1729
int cpu_map_generic_redirect(struct bpf_cpu_map_entry *rcpu,
			     struct sk_buff *skb);
1730

1731 1732 1733 1734 1735 1736 1737
/* 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;
}

1738
struct bpf_prog *bpf_prog_get_type_path(const char *name, enum bpf_prog_type type);
1739
int array_map_alloc_check(union bpf_attr *attr);
1740

1741 1742 1743 1744
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);
1745 1746 1747
int bpf_prog_test_run_tracing(struct bpf_prog *prog,
			      const union bpf_attr *kattr,
			      union bpf_attr __user *uattr);
1748 1749 1750
int bpf_prog_test_run_flow_dissector(struct bpf_prog *prog,
				     const union bpf_attr *kattr,
				     union bpf_attr __user *uattr);
1751 1752 1753
int bpf_prog_test_run_raw_tp(struct bpf_prog *prog,
			     const union bpf_attr *kattr,
			     union bpf_attr __user *uattr);
1754 1755 1756
int bpf_prog_test_run_sk_lookup(struct bpf_prog *prog,
				const union bpf_attr *kattr,
				union bpf_attr __user *uattr);
1757 1758 1759
bool btf_ctx_access(int off, int size, enum bpf_access_type type,
		    const struct bpf_prog *prog,
		    struct bpf_insn_access_aux *info);
1760 1761 1762 1763 1764 1765 1766 1767 1768 1769 1770 1771 1772 1773 1774 1775 1776 1777 1778 1779 1780 1781 1782

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);
}

1783
int btf_struct_access(struct bpf_verifier_log *log, const struct btf *btf,
1784 1785
		      const struct btf_type *t, int off, int size,
		      enum bpf_access_type atype,
1786
		      u32 *next_btf_id, enum bpf_type_flag *flag);
1787
bool btf_struct_ids_match(struct bpf_verifier_log *log,
1788 1789
			  const struct btf *btf, u32 id, int off,
			  const struct btf *need_btf, u32 need_type_id);
1790

A
Alexei Starovoitov 已提交
1791 1792 1793 1794 1795 1796
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);

1797
struct bpf_reg_state;
1798 1799
int btf_check_subprog_arg_match(struct bpf_verifier_env *env, int subprog,
				struct bpf_reg_state *regs);
1800 1801 1802
int btf_check_kfunc_arg_match(struct bpf_verifier_env *env,
			      const struct btf *btf, u32 func_id,
			      struct bpf_reg_state *regs);
1803 1804
int btf_prepare_func_args(struct bpf_verifier_env *env, int subprog,
			  struct bpf_reg_state *reg);
1805
int btf_check_type_match(struct bpf_verifier_log *log, const struct bpf_prog *prog,
1806
			 struct btf *btf, const struct btf_type *t);
1807

1808
struct bpf_prog *bpf_prog_by_id(u32 id);
1809
struct bpf_link *bpf_link_by_id(u32 id);
1810

1811
const struct bpf_func_proto *bpf_base_func_proto(enum bpf_func_id func_id);
1812
void bpf_task_storage_free(struct task_struct *task);
1813 1814 1815 1816
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);
1817 1818 1819 1820 1821 1822 1823 1824
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);

1825
#else /* !CONFIG_BPF_SYSCALL */
1826 1827 1828 1829 1830
static inline struct bpf_prog *bpf_prog_get(u32 ufd)
{
	return ERR_PTR(-EOPNOTSUPP);
}

1831 1832
static inline struct bpf_prog *bpf_prog_get_type_dev(u32 ufd,
						     enum bpf_prog_type type,
1833
						     bool attach_drv)
1834 1835 1836 1837
{
	return ERR_PTR(-EOPNOTSUPP);
}

1838
static inline void bpf_prog_add(struct bpf_prog *prog, int i)
1839 1840
{
}
1841

1842 1843 1844 1845
static inline void bpf_prog_sub(struct bpf_prog *prog, int i)
{
}

1846 1847 1848
static inline void bpf_prog_put(struct bpf_prog *prog)
{
}
1849

1850
static inline void bpf_prog_inc(struct bpf_prog *prog)
1851 1852
{
}
1853

1854 1855 1856 1857 1858 1859
static inline struct bpf_prog *__must_check
bpf_prog_inc_not_zero(struct bpf_prog *prog)
{
	return ERR_PTR(-EOPNOTSUPP);
}

1860 1861 1862 1863 1864 1865 1866 1867 1868 1869 1870 1871 1872 1873 1874 1875 1876 1877 1878 1879 1880 1881 1882 1883 1884 1885 1886 1887 1888
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)
{
}

1889
static inline int bpf_obj_get_user(const char __user *pathname, int flags)
1890 1891 1892 1893
{
	return -EOPNOTSUPP;
}

1894
static inline void __dev_flush(void)
1895 1896
{
}
1897

1898
struct xdp_frame;
1899
struct bpf_dtab_netdev;
1900
struct bpf_cpu_map_entry;
1901

1902
static inline
1903
int dev_xdp_enqueue(struct net_device *dev, struct xdp_frame *xdpf,
1904 1905 1906 1907 1908
		    struct net_device *dev_rx)
{
	return 0;
}

1909
static inline
1910
int dev_map_enqueue(struct bpf_dtab_netdev *dst, struct xdp_frame *xdpf,
1911
		    struct net_device *dev_rx)
1912 1913 1914 1915
{
	return 0;
}

1916
static inline
1917
int dev_map_enqueue_multi(struct xdp_frame *xdpf, struct net_device *dev_rx,
1918 1919 1920 1921 1922
			  struct bpf_map *map, bool exclude_ingress)
{
	return 0;
}

1923 1924 1925 1926 1927 1928 1929 1930 1931
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;
}

1932 1933 1934 1935 1936 1937 1938 1939
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;
}

1940
static inline void __cpu_map_flush(void)
1941 1942 1943 1944
{
}

static inline int cpu_map_enqueue(struct bpf_cpu_map_entry *rcpu,
1945
				  struct xdp_frame *xdpf,
1946 1947 1948 1949
				  struct net_device *dev_rx)
{
	return 0;
}
1950

1951 1952 1953 1954 1955 1956
static inline int cpu_map_generic_redirect(struct bpf_cpu_map_entry *rcpu,
					   struct sk_buff *skb)
{
	return -EOPNOTSUPP;
}

1957 1958 1959 1960 1961
static inline struct bpf_prog *bpf_prog_get_type_path(const char *name,
				enum bpf_prog_type type)
{
	return ERR_PTR(-EOPNOTSUPP);
}
1962 1963 1964 1965 1966 1967 1968 1969 1970 1971 1972 1973 1974 1975 1976

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;
}

1977 1978 1979 1980 1981 1982 1983
static inline int bpf_prog_test_run_tracing(struct bpf_prog *prog,
					    const union bpf_attr *kattr,
					    union bpf_attr __user *uattr)
{
	return -ENOTSUPP;
}

1984 1985 1986 1987 1988 1989
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;
}
1990

1991 1992 1993 1994 1995 1996 1997
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;
}

1998 1999 2000
static inline void bpf_map_put(struct bpf_map *map)
{
}
2001 2002 2003 2004 2005

static inline struct bpf_prog *bpf_prog_by_id(u32 id)
{
	return ERR_PTR(-ENOTSUPP);
}
2006 2007 2008 2009 2010 2011

static inline const struct bpf_func_proto *
bpf_base_func_proto(enum bpf_func_id func_id)
{
	return NULL;
}
2012 2013 2014 2015

static inline void bpf_task_storage_free(struct task_struct *task)
{
}
2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 2026 2027

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;
}
2028
#endif /* CONFIG_BPF_SYSCALL */
2029

2030 2031 2032
void __bpf_free_used_btfs(struct bpf_prog_aux *aux,
			  struct btf_mod_pair *used_btfs, u32 len);

2033 2034 2035 2036 2037 2038
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);
}

2039 2040 2041
void __bpf_free_used_maps(struct bpf_prog_aux *aux,
			  struct bpf_map **used_maps, u32 len);

2042 2043
bool bpf_prog_get_ok(struct bpf_prog *, enum bpf_prog_type *, bool);

2044 2045
int bpf_prog_offload_compile(struct bpf_prog *prog);
void bpf_prog_offload_destroy(struct bpf_prog *prog);
2046 2047
int bpf_prog_offload_info_fill(struct bpf_prog_info *info,
			       struct bpf_prog *prog);
2048

2049 2050
int bpf_map_offload_info_fill(struct bpf_map_info *info, struct bpf_map *map);

2051 2052 2053 2054 2055 2056 2057
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);

2058
bool bpf_offload_prog_map_match(struct bpf_prog *prog, struct bpf_map *map);
2059

2060
struct bpf_offload_dev *
2061
bpf_offload_dev_create(const struct bpf_prog_offload_ops *ops, void *priv);
2062
void bpf_offload_dev_destroy(struct bpf_offload_dev *offdev);
2063
void *bpf_offload_dev_priv(struct bpf_offload_dev *offdev);
2064 2065 2066 2067
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);
2068
bool bpf_offload_dev_match(struct bpf_prog *prog, struct net_device *netdev);
2069

2070 2071 2072
#if defined(CONFIG_NET) && defined(CONFIG_BPF_SYSCALL)
int bpf_prog_offload_init(struct bpf_prog *prog, union bpf_attr *attr);

2073
static inline bool bpf_prog_is_dev_bound(const struct bpf_prog_aux *aux)
2074
{
2075
	return aux->offload_requested;
2076
}
2077 2078 2079 2080 2081 2082 2083 2084

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);
2085 2086 2087
int bpf_prog_test_run_syscall(struct bpf_prog *prog,
			      const union bpf_attr *kattr,
			      union bpf_attr __user *uattr);
2088 2089 2090 2091

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);
2092 2093 2094
int sock_map_bpf_prog_query(const union bpf_attr *attr,
			    union bpf_attr __user *uattr);

2095 2096
void sock_map_unhash(struct sock *sk);
void sock_map_close(struct sock *sk, long timeout);
2097 2098 2099 2100 2101 2102 2103 2104 2105 2106 2107
#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;
}
2108 2109 2110 2111 2112 2113 2114 2115 2116 2117 2118 2119 2120 2121

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)
{
}
2122 2123 2124 2125 2126 2127 2128

static inline int bpf_prog_test_run_syscall(struct bpf_prog *prog,
					    const union bpf_attr *kattr,
					    union bpf_attr __user *uattr)
{
	return -ENOTSUPP;
}
2129

2130
#ifdef CONFIG_BPF_SYSCALL
2131 2132
static inline int sock_map_get_from_fd(const union bpf_attr *attr,
				       struct bpf_prog *prog)
2133 2134 2135
{
	return -EINVAL;
}
2136 2137 2138 2139 2140 2141

static inline int sock_map_prog_detach(const union bpf_attr *attr,
				       enum bpf_prog_type ptype)
{
	return -EOPNOTSUPP;
}
2142 2143 2144 2145 2146 2147

static inline int sock_map_update_elem_sys(struct bpf_map *map, void *key, void *value,
					   u64 flags)
{
	return -EOPNOTSUPP;
}
2148 2149 2150 2151 2152 2153

static inline int sock_map_bpf_prog_query(const union bpf_attr *attr,
					  union bpf_attr __user *uattr)
{
	return -EINVAL;
}
2154 2155
#endif /* CONFIG_BPF_SYSCALL */
#endif /* CONFIG_NET && CONFIG_BPF_SYSCALL */
2156

2157 2158 2159 2160 2161 2162 2163 2164 2165 2166
#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)
{
}
2167

2168
#ifdef CONFIG_BPF_SYSCALL
2169 2170 2171 2172 2173 2174 2175 2176 2177 2178 2179 2180 2181 2182 2183
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) */

2184
/* verifier prototypes for helper functions called from eBPF programs */
2185 2186 2187
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
2188 2189 2190
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;
2191

2192
extern const struct bpf_func_proto bpf_get_prandom_u32_proto;
2193
extern const struct bpf_func_proto bpf_get_smp_processor_id_proto;
2194
extern const struct bpf_func_proto bpf_get_numa_node_id_proto;
2195
extern const struct bpf_func_proto bpf_tail_call_proto;
2196
extern const struct bpf_func_proto bpf_ktime_get_ns_proto;
2197
extern const struct bpf_func_proto bpf_ktime_get_boot_ns_proto;
2198 2199 2200
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;
2201
extern const struct bpf_func_proto bpf_get_stackid_proto;
Y
Yonghong Song 已提交
2202
extern const struct bpf_func_proto bpf_get_stack_proto;
2203
extern const struct bpf_func_proto bpf_get_task_stack_proto;
2204 2205
extern const struct bpf_func_proto bpf_get_stackid_proto_pe;
extern const struct bpf_func_proto bpf_get_stack_proto_pe;
2206
extern const struct bpf_func_proto bpf_sock_map_update_proto;
2207
extern const struct bpf_func_proto bpf_sock_hash_update_proto;
2208
extern const struct bpf_func_proto bpf_get_current_cgroup_id_proto;
2209
extern const struct bpf_func_proto bpf_get_current_ancestor_cgroup_id_proto;
2210 2211 2212 2213
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;
2214 2215
extern const struct bpf_func_proto bpf_spin_lock_proto;
extern const struct bpf_func_proto bpf_spin_unlock_proto;
2216
extern const struct bpf_func_proto bpf_get_local_storage_proto;
2217 2218
extern const struct bpf_func_proto bpf_strtol_proto;
extern const struct bpf_func_proto bpf_strtoul_proto;
2219
extern const struct bpf_func_proto bpf_tcp_sock_proto;
M
Martin KaFai Lau 已提交
2220
extern const struct bpf_func_proto bpf_jiffies64_proto;
2221
extern const struct bpf_func_proto bpf_get_ns_current_pid_tgid_proto;
2222
extern const struct bpf_func_proto bpf_event_output_data_proto;
2223 2224 2225 2226 2227
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;
2228
extern const struct bpf_func_proto bpf_skc_to_tcp6_sock_proto;
2229 2230 2231
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;
2232
extern const struct bpf_func_proto bpf_skc_to_udp6_sock_proto;
2233
extern const struct bpf_func_proto bpf_skc_to_unix_sock_proto;
2234
extern const struct bpf_func_proto bpf_copy_from_user_proto;
A
Alan Maguire 已提交
2235
extern const struct bpf_func_proto bpf_snprintf_btf_proto;
F
Florent Revest 已提交
2236
extern const struct bpf_func_proto bpf_snprintf_proto;
H
Hao Luo 已提交
2237
extern const struct bpf_func_proto bpf_per_cpu_ptr_proto;
H
Hao Luo 已提交
2238
extern const struct bpf_func_proto bpf_this_cpu_ptr_proto;
2239
extern const struct bpf_func_proto bpf_ktime_get_coarse_ns_proto;
2240
extern const struct bpf_func_proto bpf_sock_from_file_proto;
2241
extern const struct bpf_func_proto bpf_get_socket_ptr_cookie_proto;
2242 2243
extern const struct bpf_func_proto bpf_task_storage_get_proto;
extern const struct bpf_func_proto bpf_task_storage_delete_proto;
2244
extern const struct bpf_func_proto bpf_for_each_map_elem_proto;
2245
extern const struct bpf_func_proto bpf_btf_find_by_name_kind_proto;
2246 2247
extern const struct bpf_func_proto bpf_sk_setsockopt_proto;
extern const struct bpf_func_proto bpf_sk_getsockopt_proto;
2248
extern const struct bpf_func_proto bpf_kallsyms_lookup_name_proto;
S
Song Liu 已提交
2249
extern const struct bpf_func_proto bpf_find_vma_proto;
J
Joanne Koong 已提交
2250
extern const struct bpf_func_proto bpf_loop_proto;
H
Hou Tao 已提交
2251
extern const struct bpf_func_proto bpf_strncmp_proto;
2252
extern const struct bpf_func_proto bpf_copy_from_user_task_proto;
2253

2254 2255 2256
const struct bpf_func_proto *tracing_prog_func_proto(
  enum bpf_func_id func_id, const struct bpf_prog *prog);

2257 2258 2259
/* 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);
2260
u64 bpf_get_raw_cpu_id(u64 r1, u64 r2, u64 r3, u64 r4, u64 r5);
2261

2262
#if defined(CONFIG_NET)
2263 2264 2265
bool bpf_sock_common_is_valid_access(int off, int size,
				     enum bpf_access_type type,
				     struct bpf_insn_access_aux *info);
2266 2267 2268 2269 2270 2271 2272 2273
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
2274 2275 2276 2277 2278 2279
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;
}
2280 2281 2282 2283 2284 2285 2286 2287 2288 2289 2290 2291 2292 2293 2294 2295
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

2296
#ifdef CONFIG_INET
A
Alexei Starovoitov 已提交
2297 2298 2299 2300
struct sk_reuseport_kern {
	struct sk_buff *skb;
	struct sock *sk;
	struct sock *selected_sk;
2301
	struct sock *migrating_sk;
A
Alexei Starovoitov 已提交
2302 2303 2304 2305 2306
	void *data_end;
	u32 hash;
	u32 reuseport_id;
	bool bind_inany;
};
2307 2308 2309 2310 2311 2312 2313 2314
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);
2315 2316 2317 2318 2319 2320 2321 2322 2323

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);
2324 2325 2326 2327 2328 2329 2330 2331 2332 2333 2334 2335 2336 2337 2338 2339
#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;
}
2340 2341 2342 2343 2344 2345 2346 2347 2348 2349 2350 2351 2352 2353 2354
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;
}
2355 2356
#endif /* CONFIG_INET */

2357
enum bpf_text_poke_type {
2358 2359
	BPF_MOD_CALL,
	BPF_MOD_JUMP,
2360
};
2361

2362 2363 2364
int bpf_arch_text_poke(void *ip, enum bpf_text_poke_type t,
		       void *addr1, void *addr2);

S
Song Liu 已提交
2365 2366
void *bpf_arch_text_copy(void *dst, void *src, size_t len);

J
Jiri Olsa 已提交
2367
struct btf_id_set;
2368
bool btf_id_set_contains(const struct btf_id_set *set, u32 id);
J
Jiri Olsa 已提交
2369

2370 2371
#define MAX_BPRINTF_VARARGS		12

2372 2373 2374
int bpf_bprintf_prepare(char *fmt, u32 fmt_size, const u64 *raw_args,
			u32 **bin_buf, u32 num_args);
void bpf_bprintf_cleanup(void);
2375

2376
#endif /* _LINUX_BPF_H */