filter.h 23.5 KB
Newer Older
L
Linus Torvalds 已提交
1 2 3 4 5 6
/*
 * Linux Socket Filter Data Structures
 */
#ifndef __LINUX_FILTER_H__
#define __LINUX_FILTER_H__

7 8
#include <stdarg.h>

A
Arun Sharma 已提交
9
#include <linux/atomic.h>
10
#include <linux/compat.h>
11
#include <linux/skbuff.h>
12 13
#include <linux/linkage.h>
#include <linux/printk.h>
14
#include <linux/workqueue.h>
15
#include <linux/sched.h>
16
#include <linux/capability.h>
17
#include <linux/cryptohash.h>
18

19
#include <net/sch_generic.h>
20

21
#include <asm/cacheflush.h>
22 23

#include <uapi/linux/filter.h>
24
#include <uapi/linux/bpf.h>
25 26 27 28

struct sk_buff;
struct sock;
struct seccomp_data;
29
struct bpf_prog_aux;
30

31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46
/* ArgX, context and stack frame pointer register positions. Note,
 * Arg1, Arg2, Arg3, etc are used as argument mappings of function
 * calls in BPF_CALL instruction.
 */
#define BPF_REG_ARG1	BPF_REG_1
#define BPF_REG_ARG2	BPF_REG_2
#define BPF_REG_ARG3	BPF_REG_3
#define BPF_REG_ARG4	BPF_REG_4
#define BPF_REG_ARG5	BPF_REG_5
#define BPF_REG_CTX	BPF_REG_6
#define BPF_REG_FP	BPF_REG_10

/* Additional register mappings for converted user programs. */
#define BPF_REG_A	BPF_REG_0
#define BPF_REG_X	BPF_REG_7
#define BPF_REG_TMP	BPF_REG_8
47

48 49 50 51 52 53 54 55 56
/* Kernel hidden auxiliary/helper register for hardening step.
 * Only used by eBPF JITs. It's nothing more than a temporary
 * register that JITs use internally, only that here it's part
 * of eBPF instructions that have been rewritten for blinding
 * constants. See JIT pre-step in bpf_jit_blind_constants().
 */
#define BPF_REG_AX		MAX_BPF_REG
#define MAX_BPF_JIT_REG		(MAX_BPF_REG + 1)

57 58 59 60 61 62
/* As per nm, we expose JITed images as text (code) section for
 * kallsyms. That way, tools like perf can find it to match
 * addresses.
 */
#define BPF_SYM_ELF_TYPE	't'

63 64 65
/* BPF program can access up to 512 bytes of stack space. */
#define MAX_BPF_STACK	512

66 67
#define BPF_TAG_SIZE	8

68 69
/* Helper macros for filter block array initializers. */

70
/* ALU ops on registers, bpf_add|sub|...: dst_reg += src_reg */
71

72
#define BPF_ALU64_REG(OP, DST, SRC)				\
73
	((struct bpf_insn) {					\
74
		.code  = BPF_ALU64 | BPF_OP(OP) | BPF_X,	\
75 76
		.dst_reg = DST,					\
		.src_reg = SRC,					\
77 78 79
		.off   = 0,					\
		.imm   = 0 })

80
#define BPF_ALU32_REG(OP, DST, SRC)				\
81
	((struct bpf_insn) {					\
82
		.code  = BPF_ALU | BPF_OP(OP) | BPF_X,		\
83 84
		.dst_reg = DST,					\
		.src_reg = SRC,					\
85 86 87
		.off   = 0,					\
		.imm   = 0 })

88
/* ALU ops on immediates, bpf_add|sub|...: dst_reg += imm32 */
89

90
#define BPF_ALU64_IMM(OP, DST, IMM)				\
91
	((struct bpf_insn) {					\
92
		.code  = BPF_ALU64 | BPF_OP(OP) | BPF_K,	\
93 94
		.dst_reg = DST,					\
		.src_reg = 0,					\
95 96 97
		.off   = 0,					\
		.imm   = IMM })

98
#define BPF_ALU32_IMM(OP, DST, IMM)				\
99
	((struct bpf_insn) {					\
100
		.code  = BPF_ALU | BPF_OP(OP) | BPF_K,		\
101 102
		.dst_reg = DST,					\
		.src_reg = 0,					\
103 104 105 106 107
		.off   = 0,					\
		.imm   = IMM })

/* Endianess conversion, cpu_to_{l,b}e(), {l,b}e_to_cpu() */

108
#define BPF_ENDIAN(TYPE, DST, LEN)				\
109
	((struct bpf_insn) {					\
110
		.code  = BPF_ALU | BPF_END | BPF_SRC(TYPE),	\
111 112
		.dst_reg = DST,					\
		.src_reg = 0,					\
113 114 115
		.off   = 0,					\
		.imm   = LEN })

116
/* Short form of mov, dst_reg = src_reg */
117

118
#define BPF_MOV64_REG(DST, SRC)					\
119
	((struct bpf_insn) {					\
120
		.code  = BPF_ALU64 | BPF_MOV | BPF_X,		\
121 122
		.dst_reg = DST,					\
		.src_reg = SRC,					\
123 124 125
		.off   = 0,					\
		.imm   = 0 })

126
#define BPF_MOV32_REG(DST, SRC)					\
127
	((struct bpf_insn) {					\
128
		.code  = BPF_ALU | BPF_MOV | BPF_X,		\
129 130
		.dst_reg = DST,					\
		.src_reg = SRC,					\
131 132 133
		.off   = 0,					\
		.imm   = 0 })

134
/* Short form of mov, dst_reg = imm32 */
135

136
#define BPF_MOV64_IMM(DST, IMM)					\
137
	((struct bpf_insn) {					\
138
		.code  = BPF_ALU64 | BPF_MOV | BPF_K,		\
139 140
		.dst_reg = DST,					\
		.src_reg = 0,					\
141 142 143
		.off   = 0,					\
		.imm   = IMM })

144
#define BPF_MOV32_IMM(DST, IMM)					\
145
	((struct bpf_insn) {					\
146
		.code  = BPF_ALU | BPF_MOV | BPF_K,		\
147 148
		.dst_reg = DST,					\
		.src_reg = 0,					\
149 150 151
		.off   = 0,					\
		.imm   = IMM })

152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169
/* BPF_LD_IMM64 macro encodes single 'load 64-bit immediate' insn */
#define BPF_LD_IMM64(DST, IMM)					\
	BPF_LD_IMM64_RAW(DST, 0, IMM)

#define BPF_LD_IMM64_RAW(DST, SRC, IMM)				\
	((struct bpf_insn) {					\
		.code  = BPF_LD | BPF_DW | BPF_IMM,		\
		.dst_reg = DST,					\
		.src_reg = SRC,					\
		.off   = 0,					\
		.imm   = (__u32) (IMM) }),			\
	((struct bpf_insn) {					\
		.code  = 0, /* zero is reserved opcode */	\
		.dst_reg = 0,					\
		.src_reg = 0,					\
		.off   = 0,					\
		.imm   = ((__u64) (IMM)) >> 32 })

170 171 172 173
/* pseudo BPF_LD_IMM64 insn used to refer to process-local map_fd */
#define BPF_LD_MAP_FD(DST, MAP_FD)				\
	BPF_LD_IMM64_RAW(DST, BPF_PSEUDO_MAP_FD, MAP_FD)

174
/* Short form of mov based on type, BPF_X: dst_reg = src_reg, BPF_K: dst_reg = imm32 */
175

176
#define BPF_MOV64_RAW(TYPE, DST, SRC, IMM)			\
177
	((struct bpf_insn) {					\
178
		.code  = BPF_ALU64 | BPF_MOV | BPF_SRC(TYPE),	\
179 180
		.dst_reg = DST,					\
		.src_reg = SRC,					\
181 182 183
		.off   = 0,					\
		.imm   = IMM })

184
#define BPF_MOV32_RAW(TYPE, DST, SRC, IMM)			\
185
	((struct bpf_insn) {					\
186
		.code  = BPF_ALU | BPF_MOV | BPF_SRC(TYPE),	\
187 188
		.dst_reg = DST,					\
		.src_reg = SRC,					\
189 190 191
		.off   = 0,					\
		.imm   = IMM })

192
/* Direct packet access, R0 = *(uint *) (skb->data + imm32) */
193

194
#define BPF_LD_ABS(SIZE, IMM)					\
195
	((struct bpf_insn) {					\
196
		.code  = BPF_LD | BPF_SIZE(SIZE) | BPF_ABS,	\
197 198
		.dst_reg = 0,					\
		.src_reg = 0,					\
199
		.off   = 0,					\
200
		.imm   = IMM })
201

202
/* Indirect packet access, R0 = *(uint *) (skb->data + src_reg + imm32) */
203

204
#define BPF_LD_IND(SIZE, SRC, IMM)				\
205
	((struct bpf_insn) {					\
206
		.code  = BPF_LD | BPF_SIZE(SIZE) | BPF_IND,	\
207 208
		.dst_reg = 0,					\
		.src_reg = SRC,					\
209
		.off   = 0,					\
210
		.imm   = IMM })
211

212
/* Memory load, dst_reg = *(uint *) (src_reg + off16) */
213

214
#define BPF_LDX_MEM(SIZE, DST, SRC, OFF)			\
215
	((struct bpf_insn) {					\
216
		.code  = BPF_LDX | BPF_SIZE(SIZE) | BPF_MEM,	\
217 218
		.dst_reg = DST,					\
		.src_reg = SRC,					\
219 220 221
		.off   = OFF,					\
		.imm   = 0 })

222 223 224
/* Memory store, *(uint *) (dst_reg + off16) = src_reg */

#define BPF_STX_MEM(SIZE, DST, SRC, OFF)			\
225
	((struct bpf_insn) {					\
226
		.code  = BPF_STX | BPF_SIZE(SIZE) | BPF_MEM,	\
227 228
		.dst_reg = DST,					\
		.src_reg = SRC,					\
229 230 231
		.off   = OFF,					\
		.imm   = 0 })

232 233 234 235 236 237 238 239 240 241
/* Atomic memory add, *(uint *)(dst_reg + off16) += src_reg */

#define BPF_STX_XADD(SIZE, DST, SRC, OFF)			\
	((struct bpf_insn) {					\
		.code  = BPF_STX | BPF_SIZE(SIZE) | BPF_XADD,	\
		.dst_reg = DST,					\
		.src_reg = SRC,					\
		.off   = OFF,					\
		.imm   = 0 })

242 243 244
/* Memory store, *(uint *) (dst_reg + off16) = imm32 */

#define BPF_ST_MEM(SIZE, DST, OFF, IMM)				\
245
	((struct bpf_insn) {					\
246 247 248 249 250 251 252
		.code  = BPF_ST | BPF_SIZE(SIZE) | BPF_MEM,	\
		.dst_reg = DST,					\
		.src_reg = 0,					\
		.off   = OFF,					\
		.imm   = IMM })

/* Conditional jumps against registers, if (dst_reg 'op' src_reg) goto pc + off16 */
253

254
#define BPF_JMP_REG(OP, DST, SRC, OFF)				\
255
	((struct bpf_insn) {					\
256
		.code  = BPF_JMP | BPF_OP(OP) | BPF_X,		\
257 258
		.dst_reg = DST,					\
		.src_reg = SRC,					\
259 260 261
		.off   = OFF,					\
		.imm   = 0 })

262
/* Conditional jumps against immediates, if (dst_reg 'op' imm32) goto pc + off16 */
263

264
#define BPF_JMP_IMM(OP, DST, IMM, OFF)				\
265
	((struct bpf_insn) {					\
266
		.code  = BPF_JMP | BPF_OP(OP) | BPF_K,		\
267 268
		.dst_reg = DST,					\
		.src_reg = 0,					\
269 270 271 272 273 274
		.off   = OFF,					\
		.imm   = IMM })

/* Function call */

#define BPF_EMIT_CALL(FUNC)					\
275
	((struct bpf_insn) {					\
276
		.code  = BPF_JMP | BPF_CALL,			\
277 278
		.dst_reg = 0,					\
		.src_reg = 0,					\
279 280 281 282 283
		.off   = 0,					\
		.imm   = ((FUNC) - __bpf_call_base) })

/* Raw code statement block */

284
#define BPF_RAW_INSN(CODE, DST, SRC, OFF, IMM)			\
285
	((struct bpf_insn) {					\
286
		.code  = CODE,					\
287 288
		.dst_reg = DST,					\
		.src_reg = SRC,					\
289 290 291 292 293 294
		.off   = OFF,					\
		.imm   = IMM })

/* Program exit */

#define BPF_EXIT_INSN()						\
295
	((struct bpf_insn) {					\
296
		.code  = BPF_JMP | BPF_EXIT,			\
297 298
		.dst_reg = 0,					\
		.src_reg = 0,					\
299 300 301
		.off   = 0,					\
		.imm   = 0 })

302 303 304 305 306 307 308 309
/* Internal classic blocks for direct assignment */

#define __BPF_STMT(CODE, K)					\
	((struct sock_filter) BPF_STMT(CODE, K))

#define __BPF_JUMP(CODE, K, JT, JF)				\
	((struct sock_filter) BPF_JUMP(CODE, K, JT, JF))

310 311 312 313 314 315 316 317 318 319 320 321 322 323 324
#define bytes_to_bpf_size(bytes)				\
({								\
	int bpf_size = -EINVAL;					\
								\
	if (bytes == sizeof(u8))				\
		bpf_size = BPF_B;				\
	else if (bytes == sizeof(u16))				\
		bpf_size = BPF_H;				\
	else if (bytes == sizeof(u32))				\
		bpf_size = BPF_W;				\
	else if (bytes == sizeof(u64))				\
		bpf_size = BPF_DW;				\
								\
	bpf_size;						\
})
325

326 327 328 329 330 331 332 333 334 335 336 337 338 339
#define BPF_SIZEOF(type)					\
	({							\
		const int __size = bytes_to_bpf_size(sizeof(type)); \
		BUILD_BUG_ON(__size < 0);			\
		__size;						\
	})

#define BPF_FIELD_SIZEOF(type, field)				\
	({							\
		const int __size = bytes_to_bpf_size(FIELD_SIZEOF(type, field)); \
		BUILD_BUG_ON(__size < 0);			\
		__size;						\
	})

340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389
#define __BPF_MAP_0(m, v, ...) v
#define __BPF_MAP_1(m, v, t, a, ...) m(t, a)
#define __BPF_MAP_2(m, v, t, a, ...) m(t, a), __BPF_MAP_1(m, v, __VA_ARGS__)
#define __BPF_MAP_3(m, v, t, a, ...) m(t, a), __BPF_MAP_2(m, v, __VA_ARGS__)
#define __BPF_MAP_4(m, v, t, a, ...) m(t, a), __BPF_MAP_3(m, v, __VA_ARGS__)
#define __BPF_MAP_5(m, v, t, a, ...) m(t, a), __BPF_MAP_4(m, v, __VA_ARGS__)

#define __BPF_REG_0(...) __BPF_PAD(5)
#define __BPF_REG_1(...) __BPF_MAP(1, __VA_ARGS__), __BPF_PAD(4)
#define __BPF_REG_2(...) __BPF_MAP(2, __VA_ARGS__), __BPF_PAD(3)
#define __BPF_REG_3(...) __BPF_MAP(3, __VA_ARGS__), __BPF_PAD(2)
#define __BPF_REG_4(...) __BPF_MAP(4, __VA_ARGS__), __BPF_PAD(1)
#define __BPF_REG_5(...) __BPF_MAP(5, __VA_ARGS__)

#define __BPF_MAP(n, ...) __BPF_MAP_##n(__VA_ARGS__)
#define __BPF_REG(n, ...) __BPF_REG_##n(__VA_ARGS__)

#define __BPF_CAST(t, a)						       \
	(__force t)							       \
	(__force							       \
	 typeof(__builtin_choose_expr(sizeof(t) == sizeof(unsigned long),      \
				      (unsigned long)0, (t)0))) a
#define __BPF_V void
#define __BPF_N

#define __BPF_DECL_ARGS(t, a) t   a
#define __BPF_DECL_REGS(t, a) u64 a

#define __BPF_PAD(n)							       \
	__BPF_MAP(n, __BPF_DECL_ARGS, __BPF_N, u64, __ur_1, u64, __ur_2,       \
		  u64, __ur_3, u64, __ur_4, u64, __ur_5)

#define BPF_CALL_x(x, name, ...)					       \
	static __always_inline						       \
	u64 ____##name(__BPF_MAP(x, __BPF_DECL_ARGS, __BPF_V, __VA_ARGS__));   \
	u64 name(__BPF_REG(x, __BPF_DECL_REGS, __BPF_N, __VA_ARGS__));	       \
	u64 name(__BPF_REG(x, __BPF_DECL_REGS, __BPF_N, __VA_ARGS__))	       \
	{								       \
		return ____##name(__BPF_MAP(x,__BPF_CAST,__BPF_N,__VA_ARGS__));\
	}								       \
	static __always_inline						       \
	u64 ____##name(__BPF_MAP(x, __BPF_DECL_ARGS, __BPF_V, __VA_ARGS__))

#define BPF_CALL_0(name, ...)	BPF_CALL_x(0, name, __VA_ARGS__)
#define BPF_CALL_1(name, ...)	BPF_CALL_x(1, name, __VA_ARGS__)
#define BPF_CALL_2(name, ...)	BPF_CALL_x(2, name, __VA_ARGS__)
#define BPF_CALL_3(name, ...)	BPF_CALL_x(3, name, __VA_ARGS__)
#define BPF_CALL_4(name, ...)	BPF_CALL_x(4, name, __VA_ARGS__)
#define BPF_CALL_5(name, ...)	BPF_CALL_x(5, name, __VA_ARGS__)

390 391
#ifdef CONFIG_COMPAT
/* A struct sock_filter is architecture independent. */
392 393
struct compat_sock_fprog {
	u16		len;
394
	compat_uptr_t	filter;	/* struct sock_filter * */
395 396 397
};
#endif

398 399 400 401 402
struct sock_fprog_kern {
	u16			len;
	struct sock_filter	*filter;
};

403 404 405 406 407
struct bpf_binary_header {
	unsigned int pages;
	u8 image[];
};

408
struct bpf_prog {
409
	u16			pages;		/* Number of allocated pages */
410 411
	kmemcheck_bitfield_begin(meta);
	u16			jited:1,	/* Is our filter JIT'ed? */
412
				locked:1,	/* Program image locked? */
413
				gpl_compatible:1, /* Is filter GPL compatible? */
414
				cb_access:1,	/* Is control block accessed? */
415 416
				dst_needed:1,	/* Do we need dst entry? */
				xdp_adjust_head:1; /* Adjusting pkt head? */
417
	kmemcheck_bitfield_end(meta);
418
	enum bpf_prog_type	type;		/* Type of BPF program */
419
	u32			len;		/* Number of filter blocks */
420
	u8			tag[BPF_TAG_SIZE];
421
	struct bpf_prog_aux	*aux;		/* Auxiliary fields */
422
	struct sock_fprog_kern	*orig_prog;	/* Original BPF program */
423 424
	unsigned int		(*bpf_func)(const void *ctx,
					    const struct bpf_insn *insn);
425
	/* Instructions for interpreter */
426
	union {
427
		struct sock_filter	insns[0];
428
		struct bpf_insn		insnsi[0];
429
	};
430 431
};

432 433 434 435 436 437 438 439
struct sk_filter {
	atomic_t	refcnt;
	struct rcu_head	rcu;
	struct bpf_prog	*prog;
};

#define BPF_PROG_RUN(filter, ctx)  (*filter->bpf_func)(ctx, filter->insnsi)

440 441
#define BPF_SKB_CB_LEN QDISC_CB_PRIV_LEN

442 443 444 445 446
struct bpf_skb_data_end {
	struct qdisc_skb_cb qdisc_cb;
	void *data_end;
};

447 448 449
struct xdp_buff {
	void *data;
	void *data_end;
450
	void *data_hard_start;
451 452
};

453
/* compute the linear packet data range [data, data_end) which
454
 * will be accessed by cls_bpf, act_bpf and lwt programs
455 456 457 458 459 460 461 462 463
 */
static inline void bpf_compute_data_end(struct sk_buff *skb)
{
	struct bpf_skb_data_end *cb = (struct bpf_skb_data_end *)skb->cb;

	BUILD_BUG_ON(sizeof(*cb) > FIELD_SIZEOF(struct sk_buff, cb));
	cb->data_end = skb->data + skb_headlen(skb);
}

464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482
static inline u8 *bpf_skb_cb(struct sk_buff *skb)
{
	/* eBPF programs may read/write skb->cb[] area to transfer meta
	 * data between tail calls. Since this also needs to work with
	 * tc, that scratch memory is mapped to qdisc_skb_cb's data area.
	 *
	 * In some socket filter cases, the cb unfortunately needs to be
	 * saved/restored so that protocol specific skb->cb[] data won't
	 * be lost. In any case, due to unpriviledged eBPF programs
	 * attached to sockets, we need to clear the bpf_skb_cb() area
	 * to not leak previous contents to user space.
	 */
	BUILD_BUG_ON(FIELD_SIZEOF(struct __sk_buff, cb) != BPF_SKB_CB_LEN);
	BUILD_BUG_ON(FIELD_SIZEOF(struct __sk_buff, cb) !=
		     FIELD_SIZEOF(struct qdisc_skb_cb, data));

	return qdisc_skb_cb(skb)->data;
}

483 484 485
static inline u32 bpf_prog_run_save_cb(const struct bpf_prog *prog,
				       struct sk_buff *skb)
{
486 487
	u8 *cb_data = bpf_skb_cb(skb);
	u8 cb_saved[BPF_SKB_CB_LEN];
488 489 490
	u32 res;

	if (unlikely(prog->cb_access)) {
491 492
		memcpy(cb_saved, cb_data, sizeof(cb_saved));
		memset(cb_data, 0, sizeof(cb_saved));
493 494 495 496 497
	}

	res = BPF_PROG_RUN(prog, skb);

	if (unlikely(prog->cb_access))
498
		memcpy(cb_data, cb_saved, sizeof(cb_saved));
499 500 501 502 503 504 505

	return res;
}

static inline u32 bpf_prog_run_clear_cb(const struct bpf_prog *prog,
					struct sk_buff *skb)
{
506
	u8 *cb_data = bpf_skb_cb(skb);
507 508

	if (unlikely(prog->cb_access))
509 510
		memset(cb_data, 0, BPF_SKB_CB_LEN);

511 512 513
	return BPF_PROG_RUN(prog, skb);
}

514 515
static __always_inline u32 bpf_prog_run_xdp(const struct bpf_prog *prog,
					    struct xdp_buff *xdp)
516
{
517 518 519 520 521 522 523
	/* Caller needs to hold rcu_read_lock() (!), otherwise program
	 * can be released while still running, or map elements could be
	 * freed early while still having concurrent users. XDP fastpath
	 * already takes rcu_read_lock() when fetching the program, so
	 * it's not necessary here anymore.
	 */
	return BPF_PROG_RUN(prog, xdp);
524 525
}

526 527 528 529 530
static inline u32 bpf_prog_insn_size(const struct bpf_prog *prog)
{
	return prog->len * sizeof(struct bpf_insn);
}

531
static inline u32 bpf_prog_tag_scratch_size(const struct bpf_prog *prog)
532 533 534 535 536
{
	return round_up(bpf_prog_insn_size(prog) +
			sizeof(__be64) + 1, SHA_MESSAGE_BYTES);
}

537
static inline unsigned int bpf_prog_size(unsigned int proglen)
538
{
539 540
	return max(sizeof(struct bpf_prog),
		   offsetof(struct bpf_prog, insns[proglen]));
541 542
}

543 544 545 546 547 548 549 550 551 552
static inline bool bpf_prog_was_classic(const struct bpf_prog *prog)
{
	/* When classic BPF programs have been loaded and the arch
	 * does not have a classic BPF JIT (anymore), they have been
	 * converted via bpf_migrate_filter() to eBPF and thus always
	 * have an unspec program type.
	 */
	return prog->type == BPF_PROG_TYPE_UNSPEC;
}

553
#define bpf_classic_proglen(fprog) (fprog->len * sizeof(fprog->filter[0]))
554

555
#ifdef CONFIG_ARCH_HAS_SET_MEMORY
556 557
static inline void bpf_prog_lock_ro(struct bpf_prog *fp)
{
558 559
	fp->locked = 1;
	WARN_ON_ONCE(set_memory_ro((unsigned long)fp, fp->pages));
560 561 562 563
}

static inline void bpf_prog_unlock_ro(struct bpf_prog *fp)
{
564 565 566 567 568 569 570
	if (fp->locked) {
		WARN_ON_ONCE(set_memory_rw((unsigned long)fp, fp->pages));
		/* In case set_memory_rw() fails, we want to be the first
		 * to crash here instead of some random place later on.
		 */
		fp->locked = 0;
	}
571
}
572

573 574
static inline void bpf_jit_binary_lock_ro(struct bpf_binary_header *hdr)
{
575
	WARN_ON_ONCE(set_memory_ro((unsigned long)hdr, hdr->pages));
576 577
}

578 579
static inline void bpf_jit_binary_unlock_ro(struct bpf_binary_header *hdr)
{
580
	WARN_ON_ONCE(set_memory_rw((unsigned long)hdr, hdr->pages));
581
}
582 583 584 585 586 587 588 589
#else
static inline void bpf_prog_lock_ro(struct bpf_prog *fp)
{
}

static inline void bpf_prog_unlock_ro(struct bpf_prog *fp)
{
}
590

591 592 593 594
static inline void bpf_jit_binary_lock_ro(struct bpf_binary_header *hdr)
{
}

595 596 597
static inline void bpf_jit_binary_unlock_ro(struct bpf_binary_header *hdr)
{
}
598
#endif /* CONFIG_ARCH_HAS_SET_MEMORY */
599

600 601 602 603 604 605 606 607 608
static inline struct bpf_binary_header *
bpf_jit_binary_hdr(const struct bpf_prog *fp)
{
	unsigned long real_start = (unsigned long)fp->bpf_func;
	unsigned long addr = real_start & PAGE_MASK;

	return (void *)addr;
}

609 610 611 612 613
int sk_filter_trim_cap(struct sock *sk, struct sk_buff *skb, unsigned int cap);
static inline int sk_filter(struct sock *sk, struct sk_buff *skb)
{
	return sk_filter_trim_cap(sk, skb, 1);
}
614

615
struct bpf_prog *bpf_prog_select_runtime(struct bpf_prog *fp, int *err);
616
void bpf_prog_free(struct bpf_prog *fp);
617

618 619 620 621 622 623 624 625 626 627 628
struct bpf_prog *bpf_prog_alloc(unsigned int size, gfp_t gfp_extra_flags);
struct bpf_prog *bpf_prog_realloc(struct bpf_prog *fp_old, unsigned int size,
				  gfp_t gfp_extra_flags);
void __bpf_prog_free(struct bpf_prog *fp);

static inline void bpf_prog_unlock_free(struct bpf_prog *fp)
{
	bpf_prog_unlock_ro(fp);
	__bpf_prog_free(fp);
}

629 630 631
typedef int (*bpf_aux_classic_check_t)(struct sock_filter *filter,
				       unsigned int flen);

632
int bpf_prog_create(struct bpf_prog **pfp, struct sock_fprog_kern *fprog);
633
int bpf_prog_create_from_user(struct bpf_prog **pfp, struct sock_fprog *fprog,
634
			      bpf_aux_classic_check_t trans, bool save_orig);
635
void bpf_prog_destroy(struct bpf_prog *fp);
636

637
int sk_attach_filter(struct sock_fprog *fprog, struct sock *sk);
638
int sk_attach_bpf(u32 ufd, struct sock *sk);
639 640
int sk_reuseport_attach_filter(struct sock_fprog *fprog, struct sock *sk);
int sk_reuseport_attach_bpf(u32 ufd, struct sock *sk);
641 642 643 644
int sk_detach_filter(struct sock *sk);
int sk_get_filter(struct sock *sk, struct sock_filter __user *filter,
		  unsigned int len);

645
bool sk_filter_charge(struct sock *sk, struct sk_filter *fp);
646
void sk_filter_uncharge(struct sock *sk, struct sk_filter *fp);
647

648
u64 __bpf_call_base(u64 r1, u64 r2, u64 r3, u64 r4, u64 r5);
649 650

struct bpf_prog *bpf_int_jit_compile(struct bpf_prog *prog);
651
void bpf_jit_compile(struct bpf_prog *prog);
652
bool bpf_helper_changes_pkt_data(void *func);
653

654 655
struct bpf_prog *bpf_patch_insn_single(struct bpf_prog *prog, u32 off,
				       const struct bpf_insn *patch, u32 len);
656
void bpf_warn_invalid_xdp_action(u32 act);
657

658
#ifdef CONFIG_BPF_JIT
659
extern int bpf_jit_enable;
660
extern int bpf_jit_harden;
661
extern int bpf_jit_kallsyms;
662

663 664 665 666 667 668 669 670 671 672
typedef void (*bpf_jit_fill_hole_t)(void *area, unsigned int size);

struct bpf_binary_header *
bpf_jit_binary_alloc(unsigned int proglen, u8 **image_ptr,
		     unsigned int alignment,
		     bpf_jit_fill_hole_t bpf_fill_ill_insns);
void bpf_jit_binary_free(struct bpf_binary_header *hdr);

void bpf_jit_free(struct bpf_prog *fp);

673 674 675
struct bpf_prog *bpf_jit_blind_constants(struct bpf_prog *fp);
void bpf_jit_prog_release_other(struct bpf_prog *fp, struct bpf_prog *fp_other);

676 677 678
static inline void bpf_jit_dump(unsigned int flen, unsigned int proglen,
				u32 pass, void *image)
{
679 680 681
	pr_err("flen=%u proglen=%u pass=%u image=%pK from=%s pid=%d\n", flen,
	       proglen, pass, image, current->comm, task_pid_nr(current));

682 683 684 685
	if (image)
		print_hex_dump(KERN_ERR, "JIT code: ", DUMP_PREFIX_OFFSET,
			       16, 1, image, proglen, false);
}
686 687 688 689 690 691 692 693 694 695

static inline bool bpf_jit_is_ebpf(void)
{
# ifdef CONFIG_HAVE_EBPF_JIT
	return true;
# else
	return false;
# endif
}

696 697 698 699 700
static inline bool ebpf_jit_enabled(void)
{
	return bpf_jit_enable && bpf_jit_is_ebpf();
}

701 702 703 704 705
static inline bool bpf_prog_ebpf_jited(const struct bpf_prog *fp)
{
	return fp->jited && bpf_jit_is_ebpf();
}

706 707 708 709 710 711 712 713 714 715 716 717 718 719 720 721 722
static inline bool bpf_jit_blinding_enabled(void)
{
	/* These are the prerequisites, should someone ever have the
	 * idea to call blinding outside of them, we make sure to
	 * bail out.
	 */
	if (!bpf_jit_is_ebpf())
		return false;
	if (!bpf_jit_enable)
		return false;
	if (!bpf_jit_harden)
		return false;
	if (bpf_jit_harden == 1 && capable(CAP_SYS_ADMIN))
		return false;

	return true;
}
723 724 725 726 727 728 729 730 731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748 749 750 751 752 753 754 755 756 757 758 759 760

static inline bool bpf_jit_kallsyms_enabled(void)
{
	/* There are a couple of corner cases where kallsyms should
	 * not be enabled f.e. on hardening.
	 */
	if (bpf_jit_harden)
		return false;
	if (!bpf_jit_kallsyms)
		return false;
	if (bpf_jit_kallsyms == 1)
		return true;

	return false;
}

const char *__bpf_address_lookup(unsigned long addr, unsigned long *size,
				 unsigned long *off, char *sym);
bool is_bpf_text_address(unsigned long addr);
int bpf_get_kallsym(unsigned int symnum, unsigned long *value, char *type,
		    char *sym);

static inline const char *
bpf_address_lookup(unsigned long addr, unsigned long *size,
		   unsigned long *off, char **modname, char *sym)
{
	const char *ret = __bpf_address_lookup(addr, size, off, sym);

	if (ret && modname)
		*modname = NULL;
	return ret;
}

void bpf_prog_kallsyms_add(struct bpf_prog *fp);
void bpf_prog_kallsyms_del(struct bpf_prog *fp);

#else /* CONFIG_BPF_JIT */

761 762 763 764 765
static inline bool ebpf_jit_enabled(void)
{
	return false;
}

766 767 768 769 770
static inline bool bpf_prog_ebpf_jited(const struct bpf_prog *fp)
{
	return false;
}

771 772 773 774
static inline void bpf_jit_free(struct bpf_prog *fp)
{
	bpf_prog_unlock_free(fp);
}
775 776 777 778 779 780 781 782 783 784 785 786 787 788 789 790 791 792 793 794 795 796 797 798 799 800 801 802 803 804 805 806 807 808 809 810 811 812

static inline bool bpf_jit_kallsyms_enabled(void)
{
	return false;
}

static inline const char *
__bpf_address_lookup(unsigned long addr, unsigned long *size,
		     unsigned long *off, char *sym)
{
	return NULL;
}

static inline bool is_bpf_text_address(unsigned long addr)
{
	return false;
}

static inline int bpf_get_kallsym(unsigned int symnum, unsigned long *value,
				  char *type, char *sym)
{
	return -ERANGE;
}

static inline const char *
bpf_address_lookup(unsigned long addr, unsigned long *size,
		   unsigned long *off, char **modname, char *sym)
{
	return NULL;
}

static inline void bpf_prog_kallsyms_add(struct bpf_prog *fp)
{
}

static inline void bpf_prog_kallsyms_del(struct bpf_prog *fp)
{
}
813 814
#endif /* CONFIG_BPF_JIT */

815 816
#define BPF_ANC		BIT(15)

817 818 819 820 821 822 823 824 825 826 827 828 829 830 831 832 833 834 835
static inline bool bpf_needs_clear_a(const struct sock_filter *first)
{
	switch (first->code) {
	case BPF_RET | BPF_K:
	case BPF_LD | BPF_W | BPF_LEN:
		return false;

	case BPF_LD | BPF_W | BPF_ABS:
	case BPF_LD | BPF_H | BPF_ABS:
	case BPF_LD | BPF_B | BPF_ABS:
		if (first->k == SKF_AD_OFF + SKF_AD_ALU_XOR_X)
			return true;
		return false;

	default:
		return true;
	}
}

836 837 838 839 840 841 842 843 844 845 846 847 848 849 850 851 852 853 854 855 856 857 858 859 860 861
static inline u16 bpf_anc_helper(const struct sock_filter *ftest)
{
	BUG_ON(ftest->code & BPF_ANC);

	switch (ftest->code) {
	case BPF_LD | BPF_W | BPF_ABS:
	case BPF_LD | BPF_H | BPF_ABS:
	case BPF_LD | BPF_B | BPF_ABS:
#define BPF_ANCILLARY(CODE)	case SKF_AD_OFF + SKF_AD_##CODE:	\
				return BPF_ANC | SKF_AD_##CODE
		switch (ftest->k) {
		BPF_ANCILLARY(PROTOCOL);
		BPF_ANCILLARY(PKTTYPE);
		BPF_ANCILLARY(IFINDEX);
		BPF_ANCILLARY(NLATTR);
		BPF_ANCILLARY(NLATTR_NEST);
		BPF_ANCILLARY(MARK);
		BPF_ANCILLARY(QUEUE);
		BPF_ANCILLARY(HATYPE);
		BPF_ANCILLARY(RXHASH);
		BPF_ANCILLARY(CPU);
		BPF_ANCILLARY(ALU_XOR_X);
		BPF_ANCILLARY(VLAN_TAG);
		BPF_ANCILLARY(VLAN_TAG_PRESENT);
		BPF_ANCILLARY(PAY_OFFSET);
		BPF_ANCILLARY(RANDOM);
862
		BPF_ANCILLARY(VLAN_TPID);
863 864 865 866 867 868 869
		}
		/* Fallthrough. */
	default:
		return ftest->code;
	}
}

870 871 872 873 874 875 876 877 878 879 880 881
void *bpf_internal_load_pointer_neg_helper(const struct sk_buff *skb,
					   int k, unsigned int size);

static inline void *bpf_load_pointer(const struct sk_buff *skb, int k,
				     unsigned int size, void *buffer)
{
	if (k >= 0)
		return skb_header_pointer(skb, k, size, buffer);

	return bpf_internal_load_pointer_neg_helper(skb, k, size);
}

882 883
static inline int bpf_tell_extensions(void)
{
884
	return SKF_AD_MAX;
885 886
}

L
Linus Torvalds 已提交
887
#endif /* __LINUX_FILTER_H__ */