mirror of
https://git.hardenedbsd.org/hardenedbsd/HardenedBSD.git
synced 2024-11-23 12:41:10 +01:00
effaab8861
Forwarded packets passed through PFIL_OUT, which made it difficult for firewalls to figure out if they were forwarding or producing packets. This in turn is an issue for pf for IPv6 fragment handling: it needs to call ip6_output() or ip6_forward() to handle the fragments. Figuring out which was difficult (and until now, incorrect). Having pfil distinguish the two removes an ugly piece of code from pf. Introduce a new variant of the netpfil callbacks with a flags variable, which has PFIL_FWD set for forwarded packets. This allows pf to reliably work out if a packet is forwarded. Reviewed by: ae, kevans Differential Revision: https://reviews.freebsd.org/D13715
132 lines
4.6 KiB
C
132 lines
4.6 KiB
C
/* $FreeBSD$ */
|
|
/* $NetBSD: pfil.h,v 1.22 2003/06/23 12:57:08 martin Exp $ */
|
|
|
|
/*-
|
|
* SPDX-License-Identifier: BSD-3-Clause
|
|
*
|
|
* Copyright (c) 1996 Matthew R. Green
|
|
* All rights reserved.
|
|
*
|
|
* Redistribution and use in source and binary forms, with or without
|
|
* modification, are permitted provided that the following conditions
|
|
* are met:
|
|
* 1. Redistributions of source code must retain the above copyright
|
|
* notice, this list of conditions and the following disclaimer.
|
|
* 2. Redistributions in binary form must reproduce the above copyright
|
|
* notice, this list of conditions and the following disclaimer in the
|
|
* documentation and/or other materials provided with the distribution.
|
|
* 3. The name of the author may not be used to endorse or promote products
|
|
* derived from this software without specific prior written permission.
|
|
*
|
|
* THIS SOFTWARE IS PROVIDED BY THE AUTHOR ``AS IS'' AND ANY EXPRESS OR
|
|
* IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES
|
|
* OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED.
|
|
* IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR ANY DIRECT, INDIRECT,
|
|
* INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING,
|
|
* BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES;
|
|
* LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED
|
|
* AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY,
|
|
* OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
|
|
* OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
|
|
* SUCH DAMAGE.
|
|
*/
|
|
|
|
#ifndef _NET_PFIL_H_
|
|
#define _NET_PFIL_H_
|
|
|
|
#include <sys/systm.h>
|
|
#include <sys/queue.h>
|
|
#include <sys/_lock.h>
|
|
#include <sys/_mutex.h>
|
|
#include <sys/lock.h>
|
|
#include <sys/rmlock.h>
|
|
#include <net/vnet.h>
|
|
|
|
struct mbuf;
|
|
struct ifnet;
|
|
struct inpcb;
|
|
|
|
typedef int (*pfil_func_t)(void *, struct mbuf **, struct ifnet *, int,
|
|
struct inpcb *);
|
|
typedef int (*pfil_func_flags_t)(void *, struct mbuf **, struct ifnet *,
|
|
int, int, struct inpcb *);
|
|
|
|
/*
|
|
* The packet filter hooks are designed for anything to call them to
|
|
* possibly intercept the packet. Multiple filter hooks are chained
|
|
* together and after each other in the specified order.
|
|
*/
|
|
struct packet_filter_hook {
|
|
TAILQ_ENTRY(packet_filter_hook) pfil_chain;
|
|
pfil_func_t pfil_func;
|
|
pfil_func_flags_t pfil_func_flags;
|
|
void *pfil_arg;
|
|
};
|
|
|
|
#define PFIL_IN 0x00000001
|
|
#define PFIL_OUT 0x00000002
|
|
#define PFIL_WAITOK 0x00000004
|
|
#define PFIL_FWD 0x00000008
|
|
#define PFIL_ALL (PFIL_IN|PFIL_OUT)
|
|
|
|
typedef TAILQ_HEAD(pfil_chain, packet_filter_hook) pfil_chain_t;
|
|
|
|
#define PFIL_TYPE_AF 1 /* key is AF_* type */
|
|
#define PFIL_TYPE_IFNET 2 /* key is ifnet pointer */
|
|
|
|
#define PFIL_FLAG_PRIVATE_LOCK 0x01 /* Personal lock instead of global */
|
|
|
|
/*
|
|
* A pfil head is created by each protocol or packet intercept point.
|
|
* For packet is then run through the hook chain for inspection.
|
|
*/
|
|
struct pfil_head {
|
|
pfil_chain_t ph_in;
|
|
pfil_chain_t ph_out;
|
|
int ph_type;
|
|
int ph_nhooks;
|
|
#if defined( __linux__ ) || defined( _WIN32 )
|
|
rwlock_t ph_mtx;
|
|
#else
|
|
struct rmlock *ph_plock; /* Pointer to the used lock */
|
|
struct rmlock ph_lock; /* Private lock storage */
|
|
int flags;
|
|
#endif
|
|
union {
|
|
u_long phu_val;
|
|
void *phu_ptr;
|
|
} ph_un;
|
|
#define ph_af ph_un.phu_val
|
|
#define ph_ifnet ph_un.phu_ptr
|
|
LIST_ENTRY(pfil_head) ph_list;
|
|
};
|
|
|
|
VNET_DECLARE(struct rmlock, pfil_lock);
|
|
#define V_pfil_lock VNET(pfil_lock)
|
|
|
|
/* Public functions for pfil hook management by packet filters. */
|
|
struct pfil_head *pfil_head_get(int, u_long);
|
|
int pfil_add_hook_flags(pfil_func_flags_t, void *, int, struct pfil_head *);
|
|
int pfil_add_hook(pfil_func_t, void *, int, struct pfil_head *);
|
|
int pfil_remove_hook_flags(pfil_func_flags_t, void *, int, struct pfil_head *);
|
|
int pfil_remove_hook(pfil_func_t, void *, int, struct pfil_head *);
|
|
#define PFIL_HOOKED(p) ((p)->ph_nhooks > 0)
|
|
|
|
/* Public functions to run the packet inspection by protocols. */
|
|
int pfil_run_hooks(struct pfil_head *, struct mbuf **, struct ifnet *, int,
|
|
int, struct inpcb *inp);
|
|
|
|
/* Public functions for pfil head management by protocols. */
|
|
int pfil_head_register(struct pfil_head *);
|
|
int pfil_head_unregister(struct pfil_head *);
|
|
|
|
/* Public pfil locking functions for self managed locks by packet filters. */
|
|
int pfil_try_rlock(struct pfil_head *, struct rm_priotracker *);
|
|
void pfil_rlock(struct pfil_head *, struct rm_priotracker *);
|
|
void pfil_runlock(struct pfil_head *, struct rm_priotracker *);
|
|
void pfil_wlock(struct pfil_head *);
|
|
void pfil_wunlock(struct pfil_head *);
|
|
int pfil_wowned(struct pfil_head *ph);
|
|
|
|
#endif /* _NET_PFIL_H_ */
|