[net-next-2.6.git] / net / core / filter.c

/*
 * Linux Socket Filter - Kernel level socket filtering
 *
 * Author:
 *     Jay Schulist <jschlst@samba.org>
 *
 * Based on the design of:
 *     - The Berkeley Packet Filter
 *
 * This program is free software; you can redistribute it and/or
 * modify it under the terms of the GNU General Public License
 * as published by the Free Software Foundation; either version
 * 2 of the License, or (at your option) any later version.
 *
 * Andi Kleen - Fix a few bad bugs and races.
 * Kris Katterjohn - Added many additional checks in sk_chk_filter()
 */

#include <linux/module.h>
#include <linux/types.h>
#include <linux/mm.h>
#include <linux/fcntl.h>
#include <linux/socket.h>
#include <linux/in.h>
#include <linux/inet.h>
#include <linux/netdevice.h>
#include <linux/if_packet.h>
#include <net/ip.h>
#include <net/protocol.h>
#include <linux/skbuff.h>
#include <net/sock.h>
#include <linux/errno.h>
#include <linux/timer.h>
#include <asm/system.h>
#include <asm/uaccess.h>
#include <asm/unaligned.h>
#include <linux/filter.h>

/* No hurry in this branch */
static void *__load_pointer(struct sk_buff *skb, int k)
{
	u8 *ptr = NULL;

	if (k >= SKF_NET_OFF)
		ptr = skb_network_header(skb) + k - SKF_NET_OFF;
	else if (k >= SKF_LL_OFF)
		ptr = skb_mac_header(skb) + k - SKF_LL_OFF;

	if (ptr >= skb->head && ptr < skb_tail_pointer(skb))
		return ptr;
	return NULL;
}

static inline void *load_pointer(struct sk_buff *skb, int k,
				 unsigned int size, void *buffer)
{
	if (k >= 0)
		return skb_header_pointer(skb, k, size, buffer);
	else {
		if (k >= SKF_AD_OFF)
			return NULL;
		return __load_pointer(skb, k);
	}
}

/**
 *	sk_run_filter - run a filter on a socket
 *	@skb: buffer to run the filter on
 *	@filter: filter to apply
 *	@flen: length of filter
 *
 * Decode and apply filter instructions to the skb->data.
 * Return length to keep, 0 for none. skb is the data we are
 * filtering, filter is the array of filter instructions, and
 * len is the number of filter blocks in the array.
 */
unsigned int sk_run_filter(struct sk_buff *skb, struct sock_filter *filter, int flen)
{
	struct sock_filter *fentry;	/* We walk down these */
	void *ptr;
	u32 A = 0;			/* Accumulator */
	u32 X = 0;			/* Index Register */
	u32 mem[BPF_MEMWORDS];		/* Scratch Memory Store */
	u32 tmp;
	int k;
	int pc;

	/*
	 * Process array of filter instructions.
	 */
	for (pc = 0; pc < flen; pc++) {
		fentry = &filter[pc];

		switch (fentry->code) {
		case BPF_ALU|BPF_ADD|BPF_X:
			A += X;
			continue;
		case BPF_ALU|BPF_ADD|BPF_K:
			A += fentry->k;
			continue;
		case BPF_ALU|BPF_SUB|BPF_X:
			A -= X;
			continue;
		case BPF_ALU|BPF_SUB|BPF_K:
			A -= fentry->k;
			continue;
		case BPF_ALU|BPF_MUL|BPF_X:
			A *= X;
			continue;
		case BPF_ALU|BPF_MUL|BPF_K:
			A *= fentry->k;
			continue;
		case BPF_ALU|BPF_DIV|BPF_X:
			if (X == 0)
				return 0;
			A /= X;
			continue;
		case BPF_ALU|BPF_DIV|BPF_K:
			A /= fentry->k;
			continue;
		case BPF_ALU|BPF_AND|BPF_X:
			A &= X;
			continue;
		case BPF_ALU|BPF_AND|BPF_K:
			A &= fentry->k;
			continue;
		case BPF_ALU|BPF_OR|BPF_X:
			A |= X;
			continue;
		case BPF_ALU|BPF_OR|BPF_K:
			A |= fentry->k;
			continue;
		case BPF_ALU|BPF_LSH|BPF_X:
			A <<= X;
			continue;
		case BPF_ALU|BPF_LSH|BPF_K:
			A <<= fentry->k;
			continue;
		case BPF_ALU|BPF_RSH|BPF_X:
			A >>= X;
			continue;
		case BPF_ALU|BPF_RSH|BPF_K:
			A >>= fentry->k;
			continue;
		case BPF_ALU|BPF_NEG:
			A = -A;
			continue;
		case BPF_JMP|BPF_JA:
			pc += fentry->k;
			continue;
		case BPF_JMP|BPF_JGT|BPF_K:
			pc += (A > fentry->k) ? fentry->jt : fentry->jf;
			continue;
		case BPF_JMP|BPF_JGE|BPF_K:
			pc += (A >= fentry->k) ? fentry->jt : fentry->jf;
			continue;
		case BPF_JMP|BPF_JEQ|BPF_K:
			pc += (A == fentry->k) ? fentry->jt : fentry->jf;
			continue;
		case BPF_JMP|BPF_JSET|BPF_K:
			pc += (A & fentry->k) ? fentry->jt : fentry->jf;
			continue;
		case BPF_JMP|BPF_JGT|BPF_X:
			pc += (A > X) ? fentry->jt : fentry->jf;
			continue;
		case BPF_JMP|BPF_JGE|BPF_X:
			pc += (A >= X) ? fentry->jt : fentry->jf;
			continue;
		case BPF_JMP|BPF_JEQ|BPF_X:
			pc += (A == X) ? fentry->jt : fentry->jf;
			continue;
		case BPF_JMP|BPF_JSET|BPF_X:
			pc += (A & X) ? fentry->jt : fentry->jf;
			continue;
		case BPF_LD|BPF_W|BPF_ABS:
			k = fentry->k;
load_w:
			ptr = load_pointer(skb, k, 4, &tmp);
			if (ptr != NULL) {
				A = ntohl(get_unaligned((__be32 *)ptr));
				continue;
			}
			break;
		case BPF_LD|BPF_H|BPF_ABS:
			k = fentry->k;
load_h:
			ptr = load_pointer(skb, k, 2, &tmp);
			if (ptr != NULL) {
				A = ntohs(get_unaligned((__be16 *)ptr));
				continue;
			}
			break;
		case BPF_LD|BPF_B|BPF_ABS:
			k = fentry->k;
load_b:
			ptr = load_pointer(skb, k, 1, &tmp);
			if (ptr != NULL) {
				A = *(u8 *)ptr;
				continue;
			}
			break;
		case BPF_LD|BPF_W|BPF_LEN:
			A = skb->len;
			continue;
		case BPF_LDX|BPF_W|BPF_LEN:
			X = skb->len;
			continue;
		case BPF_LD|BPF_W|BPF_IND:
			k = X + fentry->k;
			goto load_w;
		case BPF_LD|BPF_H|BPF_IND:
			k = X + fentry->k;
			goto load_h;
		case BPF_LD|BPF_B|BPF_IND:
			k = X + fentry->k;
			goto load_b;
		case BPF_LDX|BPF_B|BPF_MSH:
			ptr = load_pointer(skb, fentry->k, 1, &tmp);
			if (ptr != NULL) {
				X = (*(u8 *)ptr & 0xf) << 2;
				continue;
			}
			return 0;
		case BPF_LD|BPF_IMM:
			A = fentry->k;
			continue;
		case BPF_LDX|BPF_IMM:
			X = fentry->k;
			continue;
		case BPF_LD|BPF_MEM:
			A = mem[fentry->k];
			continue;
		case BPF_LDX|BPF_MEM:
			X = mem[fentry->k];
			continue;
		case BPF_MISC|BPF_TAX:
			X = A;
			continue;
		case BPF_MISC|BPF_TXA:
			A = X;
			continue;
		case BPF_RET|BPF_K:
			return fentry->k;
		case BPF_RET|BPF_A:
			return A;
		case BPF_ST:
			mem[fentry->k] = A;
			continue;
		case BPF_STX:
			mem[fentry->k] = X;
			continue;
		default:
			WARN_ON(1);
			return 0;
		}

		/*
		 * Handle ancillary data, which are impossible
		 * (or very difficult) to get parsing packet contents.
		 */
		switch (k-SKF_AD_OFF) {
		case SKF_AD_PROTOCOL:
			A = ntohs(skb->protocol);
			continue;
		case SKF_AD_PKTTYPE:
			A = skb->pkt_type;
			continue;
		case SKF_AD_IFINDEX:
			A = skb->dev->ifindex;
			continue;
		default:
			return 0;
		}
	}

	return 0;
}
EXPORT_SYMBOL(sk_run_filter);

/**
 *	sk_chk_filter - verify socket filter code
 *	@filter: filter to verify
 *	@flen: length of filter
 *
 * Check the user's filter code. If we let some ugly
 * filter code slip through kaboom! The filter must contain
 * no references or jumps that are out of range, no illegal
 * instructions, and must end with a RET instruction.
 *
 * All jumps are forward as they are not signed.
 *
 * Returns 0 if the rule set is legal or -EINVAL if not.
 */
int sk_chk_filter(struct sock_filter *filter, int flen)
{
	struct sock_filter *ftest;
	int pc;

	if (flen == 0 || flen > BPF_MAXINSNS)
		return -EINVAL;

	/* check the filter code now */
	for (pc = 0; pc < flen; pc++) {
		ftest = &filter[pc];

		/* Only allow valid instructions */
		switch (ftest->code) {
		case BPF_ALU|BPF_ADD|BPF_K:
		case BPF_ALU|BPF_ADD|BPF_X:
		case BPF_ALU|BPF_SUB|BPF_K:
		case BPF_ALU|BPF_SUB|BPF_X:
		case BPF_ALU|BPF_MUL|BPF_K:
		case BPF_ALU|BPF_MUL|BPF_X:
		case BPF_ALU|BPF_DIV|BPF_X:
		case BPF_ALU|BPF_AND|BPF_K:
		case BPF_ALU|BPF_AND|BPF_X:
		case BPF_ALU|BPF_OR|BPF_K:
		case BPF_ALU|BPF_OR|BPF_X:
		case BPF_ALU|BPF_LSH|BPF_K:
		case BPF_ALU|BPF_LSH|BPF_X:
		case BPF_ALU|BPF_RSH|BPF_K:
		case BPF_ALU|BPF_RSH|BPF_X:
		case BPF_ALU|BPF_NEG:
		case BPF_LD|BPF_W|BPF_ABS:
		case BPF_LD|BPF_H|BPF_ABS:
		case BPF_LD|BPF_B|BPF_ABS:
		case BPF_LD|BPF_W|BPF_LEN:
		case BPF_LD|BPF_W|BPF_IND:
		case BPF_LD|BPF_H|BPF_IND:
		case BPF_LD|BPF_B|BPF_IND:
		case BPF_LD|BPF_IMM:
		case BPF_LDX|BPF_W|BPF_LEN:
		case BPF_LDX|BPF_B|BPF_MSH:
		case BPF_LDX|BPF_IMM:
		case BPF_MISC|BPF_TAX:
		case BPF_MISC|BPF_TXA:
		case BPF_RET|BPF_K:
		case BPF_RET|BPF_A:
			break;

		/* Some instructions need special checks */

		case BPF_ALU|BPF_DIV|BPF_K:
			/* check for division by zero */
			if (ftest->k == 0)
				return -EINVAL;
			break;

		case BPF_LD|BPF_MEM:
		case BPF_LDX|BPF_MEM:
		case BPF_ST:
		case BPF_STX:
			/* check for invalid memory addresses */
			if (ftest->k >= BPF_MEMWORDS)
				return -EINVAL;
			break;

		case BPF_JMP|BPF_JA:
			/*
			 * Note, the large ftest->k might cause loops.
			 * Compare this with conditional jumps below,
			 * where offsets are limited. --ANK (981016)
			 */
			if (ftest->k >= (unsigned)(flen-pc-1))
				return -EINVAL;
			break;

		case BPF_JMP|BPF_JEQ|BPF_K:
		case BPF_JMP|BPF_JEQ|BPF_X:
		case BPF_JMP|BPF_JGE|BPF_K:
		case BPF_JMP|BPF_JGE|BPF_X:
		case BPF_JMP|BPF_JGT|BPF_K:
		case BPF_JMP|BPF_JGT|BPF_X:
		case BPF_JMP|BPF_JSET|BPF_K:
		case BPF_JMP|BPF_JSET|BPF_X:
			/* for conditionals both must be safe */
			if (pc + ftest->jt + 1 >= flen ||
			    pc + ftest->jf + 1 >= flen)
				return -EINVAL;
			break;

		default:
			return -EINVAL;
		}
	}

	return (BPF_CLASS(filter[flen - 1].code) == BPF_RET) ? 0 : -EINVAL;
}
EXPORT_SYMBOL(sk_chk_filter);

/**
 * 	sk_filter_rcu_release: Release a socket filter by rcu_head
 *	@rcu: rcu_head that contains the sk_filter to free
 */
static void sk_filter_rcu_release(struct rcu_head *rcu)
{
	struct sk_filter *fp = container_of(rcu, struct sk_filter, rcu);

	sk_filter_release(fp);
}

static void sk_filter_delayed_uncharge(struct sock *sk, struct sk_filter *fp)
{
	unsigned int size = sk_filter_len(fp);

	atomic_sub(size, &sk->sk_omem_alloc);
	call_rcu_bh(&fp->rcu, sk_filter_rcu_release);
}

/**
 *	sk_attach_filter - attach a socket filter
 *	@fprog: the filter program
 *	@sk: the socket to use
 *
 * Attach the user's filter code. We first run some sanity checks on
 * it to make sure it does not explode on us later. If an error
 * occurs or there is insufficient memory for the filter a negative
 * errno code is returned. On success the return is zero.
 */
int sk_attach_filter(struct sock_fprog *fprog, struct sock *sk)
{
	struct sk_filter *fp, *old_fp;
	unsigned int fsize = sizeof(struct sock_filter) * fprog->len;
	int err;

	/* Make sure new filter is there and in the right amounts. */
	if (fprog->filter == NULL)
		return -EINVAL;

	fp = sock_kmalloc(sk, fsize+sizeof(*fp), GFP_KERNEL);
	if (!fp)
		return -ENOMEM;
	if (copy_from_user(fp->insns, fprog->filter, fsize)) {
		sock_kfree_s(sk, fp, fsize+sizeof(*fp));
		return -EFAULT;
	}

	atomic_set(&fp->refcnt, 1);
	fp->len = fprog->len;

	err = sk_chk_filter(fp->insns, fp->len);
	if (err) {
		sk_filter_uncharge(sk, fp);
		return err;
	}

	rcu_read_lock_bh();
	old_fp = rcu_dereference(sk->sk_filter);
	rcu_assign_pointer(sk->sk_filter, fp);
	rcu_read_unlock_bh();

	if (old_fp)
		sk_filter_delayed_uncharge(sk, old_fp);
	return 0;
}

int sk_detach_filter(struct sock *sk)
{
	int ret = -ENOENT;
	struct sk_filter *filter;

	rcu_read_lock_bh();
	filter = rcu_dereference(sk->sk_filter);
	if (filter) {
		rcu_assign_pointer(sk->sk_filter, NULL);
		sk_filter_delayed_uncharge(sk, filter);
		ret = 0;
	}
	rcu_read_unlock_bh();
	return ret;
}
Commit	Line	Data
1da177e4 LT	1	/*
	2	* Linux Socket Filter - Kernel level socket filtering
	3	*
	4	* Author:
	5	* Jay Schulist <jschlst@samba.org>
	6	*
	7	* Based on the design of:
	8	* - The Berkeley Packet Filter
	9	*
	10	* This program is free software; you can redistribute it and/or
	11	* modify it under the terms of the GNU General Public License
	12	* as published by the Free Software Foundation; either version
	13	* 2 of the License, or (at your option) any later version.
	14	*
	15	* Andi Kleen - Fix a few bad bugs and races.
93699863	16	* Kris Katterjohn - Added many additional checks in sk_chk_filter()
1da177e4 LT	17	*/
	18
	19	#include <linux/module.h>
	20	#include <linux/types.h>
1da177e4 LT	21	#include <linux/mm.h>
	22	#include <linux/fcntl.h>
	23	#include <linux/socket.h>
	24	#include <linux/in.h>
	25	#include <linux/inet.h>
	26	#include <linux/netdevice.h>
	27	#include <linux/if_packet.h>
	28	#include <net/ip.h>
	29	#include <net/protocol.h>
	30	#include <linux/skbuff.h>
	31	#include <net/sock.h>
	32	#include <linux/errno.h>
	33	#include <linux/timer.h>
	34	#include <asm/system.h>
	35	#include <asm/uaccess.h>
40daafc8	36	#include <asm/unaligned.h>
1da177e4 LT	37	#include <linux/filter.h>
	38
	39	/* No hurry in this branch */
0b05b2a4	40	static void __load_pointer(struct sk_buff skb, int k)
1da177e4 LT	41	{
	42	u8 *ptr = NULL;
	43
	44	if (k >= SKF_NET_OFF)
d56f90a7	45	ptr = skb_network_header(skb) + k - SKF_NET_OFF;
1da177e4	46	else if (k >= SKF_LL_OFF)
98e399f8	47	ptr = skb_mac_header(skb) + k - SKF_LL_OFF;
1da177e4	48
27a884dc	49	if (ptr >= skb->head && ptr < skb_tail_pointer(skb))
1da177e4 LT	50	return ptr;
	51	return NULL;
	52	}
	53
0b05b2a4	54	static inline void load_pointer(struct sk_buff skb, int k,
4ec93edb	55	unsigned int size, void *buffer)
0b05b2a4 PM	56	{
	57	if (k >= 0)
	58	return skb_header_pointer(skb, k, size, buffer);
	59	else {
	60	if (k >= SKF_AD_OFF)
	61	return NULL;
	62	return __load_pointer(skb, k);
	63	}
	64	}
	65
1da177e4	66	/**
2966b66c	67	* sk_run_filter - run a filter on a socket
1da177e4 LT	68	* @skb: buffer to run the filter on
	69	* @filter: filter to apply
	70	* @flen: length of filter
	71	*
	72	* Decode and apply filter instructions to the skb->data.
	73	* Return length to keep, 0 for none. skb is the data we are
	74	* filtering, filter is the array of filter instructions, and
	75	* len is the number of filter blocks in the array.
	76	*/
4bad4dc9	77	unsigned int sk_run_filter(struct sk_buff skb, struct sock_filter filter, int flen)
1da177e4	78	{
1da177e4	79	struct sock_filter fentry; / We walk down these */
0b05b2a4	80	void *ptr;
2966b66c KK	81	u32 A = 0; /* Accumulator */
2966b66c KK	82	u32 X = 0; /* Index Register */
1da177e4	83	u32 mem[BPF_MEMWORDS]; /* Scratch Memory Store */
0b05b2a4	84	u32 tmp;
1da177e4 LT	85	int k;
	86	int pc;
	87
	88	/*
	89	* Process array of filter instructions.
	90	*/
	91	for (pc = 0; pc < flen; pc++) {
	92	fentry = &filter[pc];
4ec93edb	93
1da177e4 LT	94	switch (fentry->code) {
	95	case BPF_ALU\|BPF_ADD\|BPF_X:
	96	A += X;
	97	continue;
	98	case BPF_ALU\|BPF_ADD\|BPF_K:
	99	A += fentry->k;
	100	continue;
	101	case BPF_ALU\|BPF_SUB\|BPF_X:
	102	A -= X;
	103	continue;
	104	case BPF_ALU\|BPF_SUB\|BPF_K:
	105	A -= fentry->k;
	106	continue;
	107	case BPF_ALU\|BPF_MUL\|BPF_X:
	108	A *= X;
	109	continue;
	110	case BPF_ALU\|BPF_MUL\|BPF_K:
	111	A *= fentry->k;
	112	continue;
	113	case BPF_ALU\|BPF_DIV\|BPF_X:
	114	if (X == 0)
	115	return 0;
	116	A /= X;
	117	continue;
	118	case BPF_ALU\|BPF_DIV\|BPF_K:
1da177e4 LT	119	A /= fentry->k;
	120	continue;
	121	case BPF_ALU\|BPF_AND\|BPF_X:
	122	A &= X;
	123	continue;
	124	case BPF_ALU\|BPF_AND\|BPF_K:
	125	A &= fentry->k;
	126	continue;
	127	case BPF_ALU\|BPF_OR\|BPF_X:
	128	A \|= X;
	129	continue;
	130	case BPF_ALU\|BPF_OR\|BPF_K:
	131	A \|= fentry->k;
	132	continue;
	133	case BPF_ALU\|BPF_LSH\|BPF_X:
	134	A <<= X;
	135	continue;
	136	case BPF_ALU\|BPF_LSH\|BPF_K:
	137	A <<= fentry->k;
	138	continue;
	139	case BPF_ALU\|BPF_RSH\|BPF_X:
	140	A >>= X;
	141	continue;
	142	case BPF_ALU\|BPF_RSH\|BPF_K:
	143	A >>= fentry->k;
	144	continue;
	145	case BPF_ALU\|BPF_NEG:
	146	A = -A;
	147	continue;
	148	case BPF_JMP\|BPF_JA:
	149	pc += fentry->k;
	150	continue;
	151	case BPF_JMP\|BPF_JGT\|BPF_K:
	152	pc += (A > fentry->k) ? fentry->jt : fentry->jf;
	153	continue;
	154	case BPF_JMP\|BPF_JGE\|BPF_K:
	155	pc += (A >= fentry->k) ? fentry->jt : fentry->jf;
	156	continue;
	157	case BPF_JMP\|BPF_JEQ\|BPF_K:
	158	pc += (A == fentry->k) ? fentry->jt : fentry->jf;
	159	continue;
	160	case BPF_JMP\|BPF_JSET\|BPF_K:
	161	pc += (A & fentry->k) ? fentry->jt : fentry->jf;
	162	continue;
	163	case BPF_JMP\|BPF_JGT\|BPF_X:
	164	pc += (A > X) ? fentry->jt : fentry->jf;
	165	continue;
	166	case BPF_JMP\|BPF_JGE\|BPF_X:
	167	pc += (A >= X) ? fentry->jt : fentry->jf;
	168	continue;
	169	case BPF_JMP\|BPF_JEQ\|BPF_X:
	170	pc += (A == X) ? fentry->jt : fentry->jf;
	171	continue;
	172	case BPF_JMP\|BPF_JSET\|BPF_X:
	173	pc += (A & X) ? fentry->jt : fentry->jf;
	174	continue;
	175	case BPF_LD\|BPF_W\|BPF_ABS:
	176	k = fentry->k;
e35bedf3	177	load_w:
0b05b2a4 PM	178	ptr = load_pointer(skb, k, 4, &tmp);
0b05b2a4 PM	179	if (ptr != NULL) {
252e3346	180	A = ntohl(get_unaligned((__be32 *)ptr));
0b05b2a4	181	continue;
1da177e4	182	}
1198ad00	183	break;
1da177e4 LT	184	case BPF_LD\|BPF_H\|BPF_ABS:
1da177e4 LT	185	k = fentry->k;
e35bedf3	186	load_h:
0b05b2a4 PM	187	ptr = load_pointer(skb, k, 2, &tmp);
0b05b2a4 PM	188	if (ptr != NULL) {
252e3346	189	A = ntohs(get_unaligned((__be16 *)ptr));
0b05b2a4	190	continue;
1da177e4	191	}
1198ad00	192	break;
1da177e4 LT	193	case BPF_LD\|BPF_B\|BPF_ABS:
	194	k = fentry->k;
	195	load_b:
0b05b2a4 PM	196	ptr = load_pointer(skb, k, 1, &tmp);
	197	if (ptr != NULL) {
	198	A = (u8 )ptr;
	199	continue;
1da177e4	200	}
1198ad00	201	break;
1da177e4	202	case BPF_LD\|BPF_W\|BPF_LEN:
3154e540	203	A = skb->len;
1da177e4 LT	204	continue;
1da177e4 LT	205	case BPF_LDX\|BPF_W\|BPF_LEN:
3154e540	206	X = skb->len;
1da177e4 LT	207	continue;
	208	case BPF_LD\|BPF_W\|BPF_IND:
	209	k = X + fentry->k;
	210	goto load_w;
	211	case BPF_LD\|BPF_H\|BPF_IND:
	212	k = X + fentry->k;
	213	goto load_h;
	214	case BPF_LD\|BPF_B\|BPF_IND:
	215	k = X + fentry->k;
	216	goto load_b;
	217	case BPF_LDX\|BPF_B\|BPF_MSH:
0b05b2a4 PM	218	ptr = load_pointer(skb, fentry->k, 1, &tmp);
	219	if (ptr != NULL) {
	220	X = ((u8 )ptr & 0xf) << 2;
	221	continue;
	222	}
	223	return 0;
1da177e4 LT	224	case BPF_LD\|BPF_IMM:
	225	A = fentry->k;
	226	continue;
	227	case BPF_LDX\|BPF_IMM:
	228	X = fentry->k;
	229	continue;
	230	case BPF_LD\|BPF_MEM:
	231	A = mem[fentry->k];
	232	continue;
	233	case BPF_LDX\|BPF_MEM:
	234	X = mem[fentry->k];
	235	continue;
	236	case BPF_MISC\|BPF_TAX:
	237	X = A;
	238	continue;
	239	case BPF_MISC\|BPF_TXA:
	240	A = X;
	241	continue;
	242	case BPF_RET\|BPF_K:
4bad4dc9	243	return fentry->k;
1da177e4	244	case BPF_RET\|BPF_A:
4bad4dc9	245	return A;
1da177e4 LT	246	case BPF_ST:
	247	mem[fentry->k] = A;
	248	continue;
	249	case BPF_STX:
	250	mem[fentry->k] = X;
	251	continue;
	252	default:
93699863	253	WARN_ON(1);
1da177e4 LT	254	return 0;
	255	}
	256
	257	/*
	258	* Handle ancillary data, which are impossible
	259	* (or very difficult) to get parsing packet contents.
	260	*/
	261	switch (k-SKF_AD_OFF) {
	262	case SKF_AD_PROTOCOL:
252e3346	263	A = ntohs(skb->protocol);
1da177e4 LT	264	continue;
	265	case SKF_AD_PKTTYPE:
	266	A = skb->pkt_type;
	267	continue;
	268	case SKF_AD_IFINDEX:
	269	A = skb->dev->ifindex;
	270	continue;
	271	default:
	272	return 0;
	273	}
	274	}
	275
	276	return 0;
	277	}
b715631f	278	EXPORT_SYMBOL(sk_run_filter);
1da177e4 LT	279
	280	/**
	281	* sk_chk_filter - verify socket filter code
	282	* @filter: filter to verify
	283	* @flen: length of filter
	284	*
	285	* Check the user's filter code. If we let some ugly
	286	* filter code slip through kaboom! The filter must contain
93699863 KK	287	* no references or jumps that are out of range, no illegal
93699863 KK	288	* instructions, and must end with a RET instruction.
1da177e4	289	*
7b11f69f KK	290	* All jumps are forward as they are not signed.
	291	*
	292	* Returns 0 if the rule set is legal or -EINVAL if not.
1da177e4 LT	293	*/
	294	int sk_chk_filter(struct sock_filter *filter, int flen)
	295	{
	296	struct sock_filter *ftest;
	297	int pc;
	298
1b93ae64	299	if (flen == 0 \|\| flen > BPF_MAXINSNS)
1da177e4 LT	300	return -EINVAL;
	301
	302	/* check the filter code now */
	303	for (pc = 0; pc < flen; pc++) {
1da177e4	304	ftest = &filter[pc];
1da177e4	305
93699863 KK	306	/* Only allow valid instructions */
	307	switch (ftest->code) {
	308	case BPF_ALU\|BPF_ADD\|BPF_K:
	309	case BPF_ALU\|BPF_ADD\|BPF_X:
	310	case BPF_ALU\|BPF_SUB\|BPF_K:
	311	case BPF_ALU\|BPF_SUB\|BPF_X:
	312	case BPF_ALU\|BPF_MUL\|BPF_K:
	313	case BPF_ALU\|BPF_MUL\|BPF_X:
	314	case BPF_ALU\|BPF_DIV\|BPF_X:
	315	case BPF_ALU\|BPF_AND\|BPF_K:
	316	case BPF_ALU\|BPF_AND\|BPF_X:
	317	case BPF_ALU\|BPF_OR\|BPF_K:
	318	case BPF_ALU\|BPF_OR\|BPF_X:
	319	case BPF_ALU\|BPF_LSH\|BPF_K:
	320	case BPF_ALU\|BPF_LSH\|BPF_X:
	321	case BPF_ALU\|BPF_RSH\|BPF_K:
	322	case BPF_ALU\|BPF_RSH\|BPF_X:
	323	case BPF_ALU\|BPF_NEG:
	324	case BPF_LD\|BPF_W\|BPF_ABS:
	325	case BPF_LD\|BPF_H\|BPF_ABS:
	326	case BPF_LD\|BPF_B\|BPF_ABS:
	327	case BPF_LD\|BPF_W\|BPF_LEN:
	328	case BPF_LD\|BPF_W\|BPF_IND:
	329	case BPF_LD\|BPF_H\|BPF_IND:
	330	case BPF_LD\|BPF_B\|BPF_IND:
	331	case BPF_LD\|BPF_IMM:
	332	case BPF_LDX\|BPF_W\|BPF_LEN:
	333	case BPF_LDX\|BPF_B\|BPF_MSH:
	334	case BPF_LDX\|BPF_IMM:
	335	case BPF_MISC\|BPF_TAX:
	336	case BPF_MISC\|BPF_TXA:
	337	case BPF_RET\|BPF_K:
	338	case BPF_RET\|BPF_A:
	339	break;
	340
	341	/* Some instructions need special checks */
fb0d366b	342
93699863 KK	343	case BPF_ALU\|BPF_DIV\|BPF_K:
	344	/* check for division by zero */
	345	if (ftest->k == 0)
1da177e4	346	return -EINVAL;
93699863 KK	347	break;
	348
	349	case BPF_LD\|BPF_MEM:
	350	case BPF_LDX\|BPF_MEM:
	351	case BPF_ST:
	352	case BPF_STX:
	353	/* check for invalid memory addresses */
	354	if (ftest->k >= BPF_MEMWORDS)
	355	return -EINVAL;
	356	break;
	357
	358	case BPF_JMP\|BPF_JA:
	359	/*
	360	* Note, the large ftest->k might cause loops.
	361	* Compare this with conditional jumps below,
	362	* where offsets are limited. --ANK (981016)
	363	*/
	364	if (ftest->k >= (unsigned)(flen-pc-1))
	365	return -EINVAL;
	366	break;
	367
	368	case BPF_JMP\|BPF_JEQ\|BPF_K:
	369	case BPF_JMP\|BPF_JEQ\|BPF_X:
	370	case BPF_JMP\|BPF_JGE\|BPF_K:
	371	case BPF_JMP\|BPF_JGE\|BPF_X:
	372	case BPF_JMP\|BPF_JGT\|BPF_K:
	373	case BPF_JMP\|BPF_JGT\|BPF_X:
	374	case BPF_JMP\|BPF_JSET\|BPF_K:
	375	case BPF_JMP\|BPF_JSET\|BPF_X:
	376	/* for conditionals both must be safe */
e35bedf3	377	if (pc + ftest->jt + 1 >= flen \|\|
93699863 KK	378	pc + ftest->jf + 1 >= flen)
	379	return -EINVAL;
	380	break;
	381
	382	default:
	383	return -EINVAL;
1da177e4 LT	384	}
	385	}
	386
e35bedf3	387	return (BPF_CLASS(filter[flen - 1].code) == BPF_RET) ? 0 : -EINVAL;
1da177e4	388	}
b715631f	389	EXPORT_SYMBOL(sk_chk_filter);
1da177e4	390
47e958ea PE	391	/**
	392	* sk_filter_rcu_release: Release a socket filter by rcu_head
	393	* @rcu: rcu_head that contains the sk_filter to free
	394	*/
	395	static void sk_filter_rcu_release(struct rcu_head *rcu)
	396	{
	397	struct sk_filter *fp = container_of(rcu, struct sk_filter, rcu);
	398
	399	sk_filter_release(fp);
	400	}
	401
	402	static void sk_filter_delayed_uncharge(struct sock sk, struct sk_filter fp)
	403	{
	404	unsigned int size = sk_filter_len(fp);
	405
	406	atomic_sub(size, &sk->sk_omem_alloc);
	407	call_rcu_bh(&fp->rcu, sk_filter_rcu_release);
	408	}
	409
1da177e4 LT	410	/**
	411	* sk_attach_filter - attach a socket filter
	412	* @fprog: the filter program
	413	* @sk: the socket to use
	414	*
	415	* Attach the user's filter code. We first run some sanity checks on
	416	* it to make sure it does not explode on us later. If an error
	417	* occurs or there is insufficient memory for the filter a negative
	418	* errno code is returned. On success the return is zero.
	419	*/
	420	int sk_attach_filter(struct sock_fprog fprog, struct sock sk)
	421	{
d3904b73	422	struct sk_filter fp, old_fp;
1da177e4 LT	423	unsigned int fsize = sizeof(struct sock_filter) * fprog->len;
	424	int err;
	425
	426	/* Make sure new filter is there and in the right amounts. */
e35bedf3 KK	427	if (fprog->filter == NULL)
e35bedf3 KK	428	return -EINVAL;
1da177e4 LT	429
	430	fp = sock_kmalloc(sk, fsize+sizeof(*fp), GFP_KERNEL);
	431	if (!fp)
	432	return -ENOMEM;
	433	if (copy_from_user(fp->insns, fprog->filter, fsize)) {
4ec93edb	434	sock_kfree_s(sk, fp, fsize+sizeof(*fp));
1da177e4 LT	435	return -EFAULT;
	436	}
	437
	438	atomic_set(&fp->refcnt, 1);
	439	fp->len = fprog->len;
	440
	441	err = sk_chk_filter(fp->insns, fp->len);
d3904b73 PE	442	if (err) {
	443	sk_filter_uncharge(sk, fp);
	444	return err;
1da177e4 LT	445	}
1da177e4 LT	446
d3904b73 PE	447	rcu_read_lock_bh();
	448	old_fp = rcu_dereference(sk->sk_filter);
	449	rcu_assign_pointer(sk->sk_filter, fp);
	450	rcu_read_unlock_bh();
	451
9b013e05 OJ	452	if (old_fp)
9b013e05 OJ	453	sk_filter_delayed_uncharge(sk, old_fp);
d3904b73	454	return 0;
1da177e4 LT	455	}
1da177e4 LT	456
55b33325 PE	457	int sk_detach_filter(struct sock *sk)
	458	{
	459	int ret = -ENOENT;
	460	struct sk_filter *filter;
	461
	462	rcu_read_lock_bh();
	463	filter = rcu_dereference(sk->sk_filter);
	464	if (filter) {
	465	rcu_assign_pointer(sk->sk_filter, NULL);
47e958ea	466	sk_filter_delayed_uncharge(sk, filter);
55b33325 PE	467	ret = 0;
	468	}
	469	rcu_read_unlock_bh();
	470	return ret;
	471	}