filter.h 12.6 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

16
#include <asm/cacheflush.h>
17 18

#include <uapi/linux/filter.h>
19
#include <uapi/linux/bpf.h>
20 21 22 23

struct sk_buff;
struct sock;
struct seccomp_data;
24
struct bpf_prog_aux;
25

26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41
/* 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
42 43 44 45

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

46 47
/* Helper macros for filter block array initializers. */

48
/* ALU ops on registers, bpf_add|sub|...: dst_reg += src_reg */
49

50
#define BPF_ALU64_REG(OP, DST, SRC)				\
51
	((struct bpf_insn) {					\
52
		.code  = BPF_ALU64 | BPF_OP(OP) | BPF_X,	\
53 54
		.dst_reg = DST,					\
		.src_reg = SRC,					\
55 56 57
		.off   = 0,					\
		.imm   = 0 })

58
#define BPF_ALU32_REG(OP, DST, SRC)				\
59
	((struct bpf_insn) {					\
60
		.code  = BPF_ALU | BPF_OP(OP) | BPF_X,		\
61 62
		.dst_reg = DST,					\
		.src_reg = SRC,					\
63 64 65
		.off   = 0,					\
		.imm   = 0 })

66
/* ALU ops on immediates, bpf_add|sub|...: dst_reg += imm32 */
67

68
#define BPF_ALU64_IMM(OP, DST, IMM)				\
69
	((struct bpf_insn) {					\
70
		.code  = BPF_ALU64 | BPF_OP(OP) | BPF_K,	\
71 72
		.dst_reg = DST,					\
		.src_reg = 0,					\
73 74 75
		.off   = 0,					\
		.imm   = IMM })

76
#define BPF_ALU32_IMM(OP, DST, IMM)				\
77
	((struct bpf_insn) {					\
78
		.code  = BPF_ALU | BPF_OP(OP) | BPF_K,		\
79 80
		.dst_reg = DST,					\
		.src_reg = 0,					\
81 82 83 84 85
		.off   = 0,					\
		.imm   = IMM })

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

86
#define BPF_ENDIAN(TYPE, DST, LEN)				\
87
	((struct bpf_insn) {					\
88
		.code  = BPF_ALU | BPF_END | BPF_SRC(TYPE),	\
89 90
		.dst_reg = DST,					\
		.src_reg = 0,					\
91 92 93
		.off   = 0,					\
		.imm   = LEN })

94
/* Short form of mov, dst_reg = src_reg */
95

96
#define BPF_MOV64_REG(DST, SRC)					\
97
	((struct bpf_insn) {					\
98
		.code  = BPF_ALU64 | BPF_MOV | BPF_X,		\
99 100
		.dst_reg = DST,					\
		.src_reg = SRC,					\
101 102 103
		.off   = 0,					\
		.imm   = 0 })

104
#define BPF_MOV32_REG(DST, SRC)					\
105
	((struct bpf_insn) {					\
106
		.code  = BPF_ALU | BPF_MOV | BPF_X,		\
107 108
		.dst_reg = DST,					\
		.src_reg = SRC,					\
109 110 111
		.off   = 0,					\
		.imm   = 0 })

112
/* Short form of mov, dst_reg = imm32 */
113

114
#define BPF_MOV64_IMM(DST, IMM)					\
115
	((struct bpf_insn) {					\
116
		.code  = BPF_ALU64 | BPF_MOV | BPF_K,		\
117 118
		.dst_reg = DST,					\
		.src_reg = 0,					\
119 120 121
		.off   = 0,					\
		.imm   = IMM })

122
#define BPF_MOV32_IMM(DST, IMM)					\
123
	((struct bpf_insn) {					\
124
		.code  = BPF_ALU | BPF_MOV | BPF_K,		\
125 126
		.dst_reg = DST,					\
		.src_reg = 0,					\
127 128 129
		.off   = 0,					\
		.imm   = IMM })

130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147
/* 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 })

148 149 150 151
/* 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)

152
/* Short form of mov based on type, BPF_X: dst_reg = src_reg, BPF_K: dst_reg = imm32 */
153

154
#define BPF_MOV64_RAW(TYPE, DST, SRC, IMM)			\
155
	((struct bpf_insn) {					\
156
		.code  = BPF_ALU64 | BPF_MOV | BPF_SRC(TYPE),	\
157 158
		.dst_reg = DST,					\
		.src_reg = SRC,					\
159 160 161
		.off   = 0,					\
		.imm   = IMM })

162
#define BPF_MOV32_RAW(TYPE, DST, SRC, IMM)			\
163
	((struct bpf_insn) {					\
164
		.code  = BPF_ALU | BPF_MOV | BPF_SRC(TYPE),	\
165 166
		.dst_reg = DST,					\
		.src_reg = SRC,					\
167 168 169
		.off   = 0,					\
		.imm   = IMM })

170
/* Direct packet access, R0 = *(uint *) (skb->data + imm32) */
171

172
#define BPF_LD_ABS(SIZE, IMM)					\
173
	((struct bpf_insn) {					\
174
		.code  = BPF_LD | BPF_SIZE(SIZE) | BPF_ABS,	\
175 176
		.dst_reg = 0,					\
		.src_reg = 0,					\
177
		.off   = 0,					\
178
		.imm   = IMM })
179

180
/* Indirect packet access, R0 = *(uint *) (skb->data + src_reg + imm32) */
181

182
#define BPF_LD_IND(SIZE, SRC, IMM)				\
183
	((struct bpf_insn) {					\
184
		.code  = BPF_LD | BPF_SIZE(SIZE) | BPF_IND,	\
185 186
		.dst_reg = 0,					\
		.src_reg = SRC,					\
187
		.off   = 0,					\
188
		.imm   = IMM })
189

190
/* Memory load, dst_reg = *(uint *) (src_reg + off16) */
191

192
#define BPF_LDX_MEM(SIZE, DST, SRC, OFF)			\
193
	((struct bpf_insn) {					\
194
		.code  = BPF_LDX | BPF_SIZE(SIZE) | BPF_MEM,	\
195 196
		.dst_reg = DST,					\
		.src_reg = SRC,					\
197 198 199
		.off   = OFF,					\
		.imm   = 0 })

200 201 202
/* Memory store, *(uint *) (dst_reg + off16) = src_reg */

#define BPF_STX_MEM(SIZE, DST, SRC, OFF)			\
203
	((struct bpf_insn) {					\
204
		.code  = BPF_STX | BPF_SIZE(SIZE) | BPF_MEM,	\
205 206
		.dst_reg = DST,					\
		.src_reg = SRC,					\
207 208 209
		.off   = OFF,					\
		.imm   = 0 })

210 211 212 213 214 215 216 217 218 219
/* 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 })

220 221 222
/* Memory store, *(uint *) (dst_reg + off16) = imm32 */

#define BPF_ST_MEM(SIZE, DST, OFF, IMM)				\
223
	((struct bpf_insn) {					\
224 225 226 227 228 229 230
		.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 */
231

232
#define BPF_JMP_REG(OP, DST, SRC, OFF)				\
233
	((struct bpf_insn) {					\
234
		.code  = BPF_JMP | BPF_OP(OP) | BPF_X,		\
235 236
		.dst_reg = DST,					\
		.src_reg = SRC,					\
237 238 239
		.off   = OFF,					\
		.imm   = 0 })

240
/* Conditional jumps against immediates, if (dst_reg 'op' imm32) goto pc + off16 */
241

242
#define BPF_JMP_IMM(OP, DST, IMM, OFF)				\
243
	((struct bpf_insn) {					\
244
		.code  = BPF_JMP | BPF_OP(OP) | BPF_K,		\
245 246
		.dst_reg = DST,					\
		.src_reg = 0,					\
247 248 249 250 251 252
		.off   = OFF,					\
		.imm   = IMM })

/* Function call */

#define BPF_EMIT_CALL(FUNC)					\
253
	((struct bpf_insn) {					\
254
		.code  = BPF_JMP | BPF_CALL,			\
255 256
		.dst_reg = 0,					\
		.src_reg = 0,					\
257 258 259 260 261
		.off   = 0,					\
		.imm   = ((FUNC) - __bpf_call_base) })

/* Raw code statement block */

262
#define BPF_RAW_INSN(CODE, DST, SRC, OFF, IMM)			\
263
	((struct bpf_insn) {					\
264
		.code  = CODE,					\
265 266
		.dst_reg = DST,					\
		.src_reg = SRC,					\
267 268 269 270 271 272
		.off   = OFF,					\
		.imm   = IMM })

/* Program exit */

#define BPF_EXIT_INSN()						\
273
	((struct bpf_insn) {					\
274
		.code  = BPF_JMP | BPF_EXIT,			\
275 276
		.dst_reg = 0,					\
		.src_reg = 0,					\
277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294
		.off   = 0,					\
		.imm   = 0 })

#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;						\
})
295

296
/* Macro to invoke filter function. */
297 298
#define SK_RUN_FILTER(filter, ctx) \
	(*filter->prog->bpf_func)(ctx, filter->prog->insnsi)
299 300 301

#ifdef CONFIG_COMPAT
/* A struct sock_filter is architecture independent. */
302 303
struct compat_sock_fprog {
	u16		len;
304
	compat_uptr_t	filter;	/* struct sock_filter * */
305 306 307
};
#endif

308 309 310 311 312
struct sock_fprog_kern {
	u16			len;
	struct sock_filter	*filter;
};

313 314 315 316 317
struct bpf_binary_header {
	unsigned int pages;
	u8 image[];
};

318
struct bpf_prog {
319 320
	u16			pages;		/* Number of allocated pages */
	bool			jited;		/* Is our filter JIT'ed? */
321
	bool			gpl_compatible;	/* Is our filter GPL compatible? */
322
	u32			len;		/* Number of filter blocks */
323
	enum bpf_prog_type	type;		/* Type of BPF program */
324
	struct bpf_prog_aux	*aux;		/* Auxiliary fields */
325
	struct sock_fprog_kern	*orig_prog;	/* Original BPF program */
326
	unsigned int		(*bpf_func)(const struct sk_buff *skb,
327
					    const struct bpf_insn *filter);
328
	/* Instructions for interpreter */
329
	union {
330
		struct sock_filter	insns[0];
331
		struct bpf_insn		insnsi[0];
332
	};
333 334
};

335 336 337 338 339 340 341 342 343
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)

static inline unsigned int bpf_prog_size(unsigned int proglen)
344
{
345 346
	return max(sizeof(struct bpf_prog),
		   offsetof(struct bpf_prog, insns[proglen]));
347 348
}

349
#define bpf_classic_proglen(fprog) (fprog->len * sizeof(fprog->filter[0]))
350

351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370
#ifdef CONFIG_DEBUG_SET_MODULE_RONX
static inline void bpf_prog_lock_ro(struct bpf_prog *fp)
{
	set_memory_ro((unsigned long)fp, fp->pages);
}

static inline void bpf_prog_unlock_ro(struct bpf_prog *fp)
{
	set_memory_rw((unsigned long)fp, fp->pages);
}
#else
static inline void bpf_prog_lock_ro(struct bpf_prog *fp)
{
}

static inline void bpf_prog_unlock_ro(struct bpf_prog *fp)
{
}
#endif /* CONFIG_DEBUG_SET_MODULE_RONX */

371
int sk_filter(struct sock *sk, struct sk_buff *skb);
372

373 374
void bpf_prog_select_runtime(struct bpf_prog *fp);
void bpf_prog_free(struct bpf_prog *fp);
375

376 377 378 379 380 381 382 383 384 385 386
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);
}

387 388 389
typedef int (*bpf_aux_classic_check_t)(struct sock_filter *filter,
				       unsigned int flen);

390
int bpf_prog_create(struct bpf_prog **pfp, struct sock_fprog_kern *fprog);
391 392
int bpf_prog_create_from_user(struct bpf_prog **pfp, struct sock_fprog *fprog,
			      bpf_aux_classic_check_t trans);
393
void bpf_prog_destroy(struct bpf_prog *fp);
394

395
int sk_attach_filter(struct sock_fprog *fprog, struct sock *sk);
396
int sk_attach_bpf(u32 ufd, struct sock *sk);
397 398 399 400
int sk_detach_filter(struct sock *sk);
int sk_get_filter(struct sock *sk, struct sock_filter __user *filter,
		  unsigned int len);

401
bool sk_filter_charge(struct sock *sk, struct sk_filter *fp);
402
void sk_filter_uncharge(struct sock *sk, struct sk_filter *fp);
403

404
u64 __bpf_call_base(u64 r1, u64 r2, u64 r3, u64 r4, u64 r5);
405
void bpf_int_jit_compile(struct bpf_prog *fp);
406

407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438
#ifdef CONFIG_BPF_JIT
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_compile(struct bpf_prog *fp);
void bpf_jit_free(struct bpf_prog *fp);

static inline void bpf_jit_dump(unsigned int flen, unsigned int proglen,
				u32 pass, void *image)
{
	pr_err("flen=%u proglen=%u pass=%u image=%pK\n",
	       flen, proglen, pass, image);
	if (image)
		print_hex_dump(KERN_ERR, "JIT code: ", DUMP_PREFIX_OFFSET,
			       16, 1, image, proglen, false);
}
#else
static inline void bpf_jit_compile(struct bpf_prog *fp)
{
}

static inline void bpf_jit_free(struct bpf_prog *fp)
{
	bpf_prog_unlock_free(fp);
}
#endif /* CONFIG_BPF_JIT */

439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466
#define BPF_ANC		BIT(15)

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);
467
		BPF_ANCILLARY(VLAN_TPID);
468 469 470 471 472 473 474
		}
		/* Fallthrough. */
	default:
		return ftest->code;
	}
}

475 476 477 478 479 480 481 482 483 484 485 486
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);
}

487 488
static inline int bpf_tell_extensions(void)
{
489
	return SKF_AD_MAX;
490 491
}

L
Linus Torvalds 已提交
492
#endif /* __LINUX_FILTER_H__ */