Passive OS fingerprint xtables match.
Subsystems (3)
Hi.
Passive OS fingerprinting netfilter module allows to passively detect remote OS and perform various netfilter actions based on that knowledge. This module compares some data (WS, MSS, options and it's order, ttl, df and others) from packets with SYN bit set with dynamically loaded OS fingerprints.
Fingerprint matching rules can be downloaded from OpenBSD source tree and loaded via netlink connector into the kernel via special util found in archive. It will also listen for events about matching packets.
Archive also contains library file (also attached), which was shipped with iptables extensions some time ago (at least when ipt_osf existed in patch-o-matic).
This release implements suggestions found during the code review like codying style, structure split and tighter packing, bool and %pi4 usage and similar changes.
Fingerprints can be downloaded from
http://www.openbsd.org/cgi-bin/cvsweb/src/etc/pf.os
Example usage:
# modrpobe xt_osf # ./ucon_osf -f ./pf.os ^C Daemon will listen for incoming match events -d switch removes fingerprints # iptables -I INPUT -j ACCEPT -p tcp -m osf --genre Linux --log 0 --ttl 2 --connector
You will find something like this in the syslog:ipt_osf: Windows [2000:SP3:Windows XP Pro SP1, 2000 SP3]: 11.22.33.55:4024 -> 11.22.33.44:139
Please consider for inclusion.
Thank you.
Passive OS fingerprint homepage (archives, examples):
http://www.ioremap.net/projects/osf
include/linux/connector.h +2 -0
@@ -39,6 +39,8 @@#define CN_IDX_V86D 0x4#define CN_VAL_V86D_UVESAFB 0x1#define CN_IDX_BB 0x5/* BlackBoard, from the TSP GPL sampling framework */ +#define CN_IDX_OSF 0x6/* Passive OS fingerprint iptables module */ +#define CN_VAL_OSF 0x0#define CN_NETLINK_USERS 6
include/linux/netfilter/xt_osf.h new file +110 -0
110 lines not on this page · raw message
net/netfilter/Kconfig +13 -0
@@ -853,6 +853,19 @@ config NETFILTER_XT_MATCH_U32 Details and examples are in the kernel module source. +config NETFILTER_XT_MATCH_OSF + tristate 'Passive OS fingerprint match support' + depends on NETFILTER_ADVANCED + help + Passive OS fingerprint matching module. Allows to passively match + remote operation system analyzing incoming TCP packets with SYN + bit set. + + You should download and install rule loading software from + http://www.ioremap.net/projects/osf + + To compile it as a module, choose M here. If unsure, say N. + endif # NETFILTER_XTABLES endmenu
net/netfilter/Makefile +1 -0
@@ -89,6 +89,7 @@ obj-$(CONFIG_NETFILTER_XT_MATCH_STRING) += xt_string.o obj-$(CONFIG_NETFILTER_XT_MATCH_TCPMSS) += xt_tcpmss.o obj-$(CONFIG_NETFILTER_XT_MATCH_TIME) += xt_time.o obj-$(CONFIG_NETFILTER_XT_MATCH_U32) += xt_u32.o +obj-$(CONFIG_NETFILTER_XT_MATCH_OSF) += xt_osf.o # IPVS obj-$(CONFIG_IP_VS) += ipvs/
net/netfilter/xt_osf.c new file +475 -0
475 lines not on this page · raw message
Signature · 1 line
Evgeniy Polyakov
/*
* Copyright (c) 2003+ Evgeniy Polyakov <[email protected]>
*
*
* 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.
*
* This program is distributed in the hope that it will be useful,
* but WITHOUT ANY WARRANTY; without even the implied warranty of
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
* GNU General Public License for more details.
*
* You should have received a copy of the GNU General Public License
* along with this program; if not, write to the Free Software
* Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA */
/*
* iptables interface for OS fingerprint matching module. */
#include <stdio.h>
#include <netdb.h>
#include <string.h>
#include <stdlib.h>
#include <getopt.h>
#include <ctype.h>
#include <linux/types.h>
#include <xtables.h>
#include <netinet/ip.h>
#include <netinet/tcp.h>
#include <linux/netfilter/xt_osf.h>
static void osf_help(void)
{
printf("OS fingerprint match options:\n"
"--genre [!] string Match a OS genre by passive fingerprinting.\n"
"--ttl Use some TTL check extensions to determine OS:\n"
" 0 true ip and fingerprint TTL comparison. Works for LAN.\n"
" 1 check if ip TTL is less than fingerprint one. Works for global addresses.\n"
" 2 do not compare TTL at all. Allows to detect NMAP, but can produce false results.\n"
"--log level Log determined genres into dmesg even if they do not match desired one:\n"
" 0 log all matched or unknown signatures.\n"
" 1 log only first one.\n"
" 2 log all known matched signatures.\n"
"--connector Log through kernel connector.\n\n"
);
}
static const struct option osf_opts[] = {
{ .name = "genre", .has_arg = true, .val = '1' },
{ .name = "ttl", .has_arg = true, .val = '2' },
{ .name = "log", .has_arg = true, .val = '3' },
{ .name = "connector", .has_arg = false, .val = '5' },
{ .name = NULL }
};static void osf_init(struct xt_entry_match *m)
{
}
static void osf_parse_string(const unsigned char *s, struct ipt_osf_info *info)
{
if (strlen(s) < MAXGENRELEN)
strcpy(info->genre, s);
else
exit_error(PARAMETER_PROBLEM, "Genre string too long `%s' [%d], max=%d",
s, strlen(s), MAXGENRELEN);
}static int osf_parse(int c, char **argv, int invert, unsigned int *flags,
const void *entry,
struct xt_entry_match **match)
{
struct ipt_osf_info *info = (struct ipt_osf_info *)(*match)->data;
switch(c)
{
case '1': /* --genre */
if (*flags & IPT_OSF_GENRE)
exit_error(PARAMETER_PROBLEM, "Can't specify multiple genre parameter");
check_inverse(optarg, &invert, &optind, 0);
osf_parse_string(argv[optind-1], info);
if (invert)
info->flags |= IPT_OSF_INVERT;
info->len=strlen(info->genre);
*flags |= IPT_OSF_GENRE;
break;
case '2': /* --ttl */
if (*flags & IPT_OSF_TTL)
exit_error(PARAMETER_PROBLEM, "Can't specify multiple ttl parameter");
*flags |= IPT_OSF_TTL;
info->flags |= IPT_OSF_TTL;
info->ttl = atoi(argv[optind-1]);
break;
case '3': /* --log */
if (*flags & IPT_OSF_LOG)
exit_error(PARAMETER_PROBLEM, "Can't specify multiple log parameter");
*flags |= IPT_OSF_LOG;
info->loglevel = atoi(argv[optind-1]);
info->flags |= IPT_OSF_LOG;
break;
case '5': /* --connector */
if (*flags & IPT_OSF_CONNECTOR)
exit_error(PARAMETER_PROBLEM, "Can't specify multiple connector parameter");
*flags |= IPT_OSF_CONNECTOR;
info->flags |= IPT_OSF_CONNECTOR;
break;
default:
return 0;
}
return 1;
}static void osf_final_check(unsigned int flags)
{
if (!flags)
exit_error(PARAMETER_PROBLEM, "OS fingerprint match: You must specify `--genre'");
}static void osf_print(const void *ip, const struct xt_entry_match *match, int numeric)
{
const struct ipt_osf_info *info = (const struct ipt_osf_info*) match->data;
printf("OS fingerprint match %s%s ", (info->flags & IPT_OSF_INVERT) ? "! " : "", info->genre);
}static void osf_save(const void *ip, const struct xt_entry_match *match)
{
const struct ipt_osf_info *info = (const struct ipt_osf_info*) match->data;
printf("--genre %s%s ", (info->flags & IPT_OSF_INVERT) ? "! ": "", info->genre);
}static struct xtables_match osf_match = {
.name = "osf",
.version = XTABLES_VERSION,
.size = XT_ALIGN(sizeof(struct ipt_osf_info)),
.userspacesize = XT_ALIGN(sizeof(struct ipt_osf_info)),
.help = osf_help,
.init = osf_init,
.parse = osf_parse,
.print = osf_print,
.final_check = osf_final_check,
.save = osf_save,
.extra_opts = osf_opts
};void _init(void)
{
xtables_register_match(&osf_match);
}Quoted 10 lines · Evgeniy Polyakov
Passive OS fingerprinting netfilter module allows to passively detect remote OS and perform various netfilter actions based on that knowledge. This module compares some data (WS, MSS, options and it's order, ttl, df and others) from packets with SYN bit set with dynamically loaded OS fingerprints.
Fingerprint matching rules can be downloaded from OpenBSD source tree and loaded via netlink connector into the kernel via special util found in archive. It will also listen for events about matching packets.
Quoted 7 lines
Archive also contains library file (also attached), which was shipped with iptables extensions some time ago (at least when
ipt_osfexisted in patch-o-matic).This release implements suggestions found during the code review like codying style, structure split and tighter packing, bool and %pi4 usage and similar changes.
Not the latest version, it misses the following fix from the parallel tree.
Fixed TCP header copy to the userspace when given option is enabled.
net/netfilter/xt_osf.c +4 -2
@@ -83,7 +83,7 @@ static void ipt_osf_send_connector(struct ipt_osf_user_finger *f,struct ipt_osf_message *msg = &per_cpu (ipt_osf_mbuf,smp_processor_id ());struct ipt_osf_nlmsg *data = &msg->nlmsg;struct iphdr *iph =ip_hdr (skb); -struct tcphdr *tcph =tcp_hdr (skb); +struct tcphdr *tcp;memcpy (&msg->cmsg.id, &cn_osf_id,sizeof (struct cn_msg )); msg->cmsg.seq = osf_seq++;
@@ -92,7 +92,9 @@ static void ipt_osf_send_connector(struct ipt_osf_user_finger *f,memcpy (&data->f, f,sizeof (struct ipt_osf_user_finger ));memcpy (&data->ip, iph,sizeof (struct iphdr )); -memcpy (&data->tcp, tcph,sizeof (struct tcphdr )); + tcp =skb_header_pointer (skb,ip_hdrlen (skb),sizeof (struct tcphdr ), &data->tcp); +if (tcp) +memcpy (&data->tcp, tcp,sizeof (struct tcphdr ));cn_netlink_send (&msg->cmsg,CN_IDX_OSF ,GFP_ATOMIC ); }
Signature · 1 line
Evgeniy Polyakov
Quoted 2 lines · Evgeniy Polyakov
[
xt_osf]
This starts to get its shape. I'm in favor of it.
include/linux/netfilter/xt_osf.h line 3
Quoted 6 lines
include/linux/netfilter/xt_osf.h new file +3
-0@@ -0,0 +1,110 @@ +/* +* xt_osf.h +*
Common-sense policy wants that filenames like these not be part of the file.
Quoted 28 lines
include/linux/netfilter/xt_osf.h new file +28
-0+#ifndef _IPT_OSF_H +#define _IPT_OSF_H + +#define MAXGENRELEN 32 +#define MAXDETLEN 64 + +#define IPT_OSF_GENRE (1<<0) +#define IPT_OSF_TTL (1<<1) +#define IPT_OSF_LOG (1<<2) +#define IPT_OSF_UNUSED (1<<3) +#define IPT_OSF_CONNECTOR (1<<4) +#define IPT_OSF_INVERT (1<<5) + +#define IPT_OSF_LOGLEVEL_ALL 0 +#define IPT_OSF_LOGLEVEL_FIRST 1 +#define IPT_OSF_LOGLEVEL_ALL_KNOWN 2 + +#define IPT_OSF_TTL_TRUE 0/* True ip and fingerprint TTL comparison */ +#define IPT_OSF_TTL_LESS 1/* Check if ip TTL is less than fingerprint one */ +#define IPT_OSF_TTL_NOCHECK 2/* Do not compare ip and fingerprint TTL at all */ + +struct ipt_osf_info { +char genre[MAXGENRELEN ]; +__u32 len; +__u32 flags; +__u32 loglevel; +__u32 ttl; +};
Quoted 1 line
[more struct
ipt_osf...]
Since the module's name is xt_osf now, it would make sense to follow this in the struct names too.
net/netfilter/Kconfig config NETFILTER_XT_MATCH_U32 · line 857
Quoted 10 lines
net/netfilter/Kconfig +2
-0@@ -853,6 +853,19 @@ config NETFILTER_XT_MATCH_U32 Details and examples are in the kernel module source. +config NETFILTER_XT_MATCH_OSF + tristate 'Passive OS fingerprint match support'
Just a minor thing, could you add "osf" somewhere in the short-text so that people instantly know the name of the module (without having to consult the help text), like the other entries. For example
tristate '"osf" Passive OS fingerprint match'
net/netfilter/Kconfig config NETFILTER_XT_MATCH_U32 · line 862
Quoted 5 lines
net/netfilter/Kconfig +5
-0+ depends on NETFILTER_ADVANCED + help + Passive OS fingerprint matching module. Allows to passively match + remote operation system analyzing incoming TCP packets with SYN + bit set.
A bit of rewording I suggest.
This option selects the Passive OS Fingerprinting match module that allows to passively match the
remote operating system by analyzing incoming TCP SYN packets.
net/netfilter/Kconfig config NETFILTER_XT_MATCH_U32 · line 865
Quoted 3 lines
net/netfilter/Kconfig +3
-0+ + You should download and install rule loading software from + http://www.ioremap.net/projects/osf
should -> can
Rules and loading software can be downloaded from
http://ioremap.net/projects/osf/
Quoted 9 lines
net/netfilter/Makefile +0
-0@@ -89,6 +89,7 @@ obj-$(CONFIG_NETFILTER_XT_MATCH_STRING) += xt_string.oobj-$(CONFIG_NETFILTER_XT_MATCH_TCPMSS) += xt_tcpmss.o obj-$(CONFIG_NETFILTER_XT_MATCH_TIME) += xt_time.o obj-$(CONFIG_NETFILTER_XT_MATCH_U32) += xt_u32.o +obj-$(CONFIG_NETFILTER_XT_MATCH_OSF) += xt_osf.o
Please have the list sorted alphabetically. (This keeps merge conflicts down because everybody does not try to append at the end.)
net/netfilter/xt_osf.c line 86
Quoted 13 lines
net/netfilter/xt_osf.c new file +8
-0+ +static void ipt_osf_send_connector (struct ipt_osf_user_finger *f, +const struct sk_buff *skb) +{ +struct ipt_osf_message *msg = &per_cpu (ipt_osf_mbuf,smp_processor_id ()); +struct ipt_osf_nlmsg *data = &msg->nlmsg; +struct iphdr *iph =ip_hdr (skb); +struct tcphdr *tcph =tcp_hdr (skb);
iph and tcph are only used in memcpy and could be directly substituted:
memcpy(&data->ip, ip_hdr(skb), ...) memcpy(&data->tcp, tcp_hdr(skb), ...)
net/netfilter/xt_osf.c line 94
Quoted 7 lines
net/netfilter/xt_osf.c new file +7
-0+memcpy (&msg->cmsg.id, &cn_osf_id,sizeof (struct cn_msg )); + msg->cmsg.seq = osf_seq++; + msg->cmsg.ack =0 ; + msg->cmsg.len =sizeof (struct ipt_osf_nlmsg ); + +memcpy (&data->f, f,sizeof (struct ipt_osf_user_finger )); +memcpy (&data->ip, iph,sizeof (struct iphdr ));
xt_osf does not look at IP options, is that so?
net/netfilter/xt_osf.c line 95
Quoted 1 line
net/netfilter/xt_osf.c new file +1
-0+memcpy (&data->tcp, tcph,sizeof (struct tcphdr ));
xt_osf does not look at TCP options, is not it?
Note that you cannot directly use tcp_hdr(skb) as the skb->transport_header has not yet been initialized (it still points at skb->network_header) because the packet has not yet been seen by the next handler. This affects PREROUTING, and INPUT chains (and BROUTING, for ebtables people, but irrelevant here).
The packet may also be fragmented or non-linear.
You actually do the right thing in ipt_osf_match_packet.
Quoted 1 line
+static inline int ipt_osf_ttl(const struct
sk_buff*skb, structipt_osf_info*info,
const struct ipt_osf_info *
net/netfilter/xt_osf.c line 103
Quoted 3 lines
net/netfilter/xt_osf.c new file +3
-0+ unsigned char f_ttl) +{ +struct iphdr *ip =ip_hdr (skb);
const struct iphdr *iph
net/netfilter/xt_osf.c line 107
Quoted 3 lines
net/netfilter/xt_osf.c new file +3
-0+if (info->flags &IPT_OSF_TTL ) { +if (info->ttl ==IPT_OSF_TTL_TRUE ) +return (ip->ttl == f_ttl);
redundant pair of ()
Quoted 12 lines
net/netfilter/xt_osf.c new file +12
-0+if (info->ttl ==IPT_OSF_TTL_NOCHECK ) +return 1 ; +else if (ip->ttl <= f_ttl) +return 1 ; +else { +struct in_device *in_dev =in_dev_get (skb->dev); +int ret =0 ; + +for_ifa (in_dev) { +if (inet_ifa_match (ip->saddr, ifa)) { + ret = (ip->ttl == f_ttl); +break ;
net/netfilter/xt_osf.c line 129
Quoted 10 lines
net/netfilter/xt_osf.c new file +10
-0+ } + } +endfor_ifa (in_dev); + +in_dev_put (in_dev); +return ret; + } + } + +return (ip->ttl == f_ttl);
redundant pair of ()
net/netfilter/xt_osf.c line 135
Quoted 6 lines
net/netfilter/xt_osf.c new file +6
-0+} + +static bool ipt_osf_match_packet (const struct sk_buff *skb, +const struct xt_match_param *p) +{ +struct ipt_osf_info *info = (struct ipt_osf_info *)p->matchinfo;
no cast needed. const advised.
net/netfilter/xt_osf.c line 137
Quoted 2 lines
net/netfilter/xt_osf.c new file +2
-0+struct iphdr *ip; +struct tcphdr _tcph, *tcp;
const
net/netfilter/xt_osf.c line 142
Quoted 5 lines
net/netfilter/xt_osf.c new file +5
-0+int fmatch =FMATCH_WRONG , fcount =0 ; +unsigned int optsize =0 , check_WSS =0 ; +u16 window, totlen, mss =0 ; +bool df; +unsigned char *optp =NULL , *_optp =NULL ;
const
net/netfilter/xt_osf.c line 146
Quoted 5 lines
net/netfilter/xt_osf.c new file +4
-0+unsigned char opts[MAX_IPOPTLEN ]; +struct ipt_osf_finger *kf; +struct ipt_osf_user_finger *f; +struct ipt_osf_finger_storage *st;
three const, as far as possible
net/netfilter/xt_osf.c line 153
Quoted 7 lines
net/netfilter/xt_osf.c new file +7
-0+ +if (!info) +return false; + + ip =ip_hdr (skb); +if (!ip) +return false;
ip_hdr is always returning non-NULL anyway.
net/netfilter/xt_osf.c line 225
Quoted 5 lines
net/netfilter/xt_osf.c new file +5
-0+ +for (optnum =0 ; optnum < f->opt_num; ++optnum) { +if (f->opt[optnum].kind == (*optp)) { +__u32 len = f->opt[optnum].length; +__u8 *optend = optp + len;
const
Quoted 5 lines
[...]
net/netfilter/xt_osf.c new file +4
-0+ } + +if (fcount) + fmatch =FMATCH_OK ;
net/netfilter/xt_osf.c line 349
Quoted 10 lines
net/netfilter/xt_osf.c new file +10
-0+ +return fmatch ==FMATCH_OK ; +} + +static struct xt_match ipt_osf_match = { + .name ="osf" , + .revision =0 , + .family =NFPROTO_IPV4 , + .proto =IPPROTO_TCP , + .hooks = (1 <<NF_INET_LOCAL_IN ) | (1 <<NF_INET_PRE_ROUTING ),
Would allow FORWARD work?
net/netfilter/xt_osf.c line 354
Quoted 5 lines
net/netfilter/xt_osf.c new file +5
-0+ .match = ipt_osf_match_packet, + .matchsize = sizeof(struct ipt_osf_info), + .me = THIS_MODULE, +}; +
-Jan
Hi Jan.
Thanks for the review, I will incorporate the changes and respin it soon.
Signature · 1 line
Evgeniy Polyakov
Re: Passive OS fingerprint xtables match (iptables)
In reply to Evgeniy PolyakovQuoted 5 lines
static void
osf_init(struct xt_entry_match *m)
{
}
You can remove this function, as it does no initialization. Implicit NULL in struct xtables_match is handled.
Quoted 6 lines
case '2': /* --ttl */ if (*flags & IPT_OSF_TTL) exit_error(PARAMETER_PROBLEM, "Can't specify multiple ttl parameter"); *flags |= IPT_OSF_TTL; info->flags |= IPT_OSF_TTL; info->ttl = atoi(argv[optind-1]);
This would allow specifying --ttl 12345.
Use of xtables_strtoui (v1.4.3-rc1+git) for bounds checking, and use of optarg seem beneficial:
unsigned int num; if (!xtables_strtoui(optarg, NULL, &num, 0, UINT8_MAX)) exit_error(PARAMETER_PROBLEM, "*shrug*"); info->ttl = num;
Quoted 5 lines
case '3': /* --log */ if (*flags & IPT_OSF_LOG) exit_error(PARAMETER_PROBLEM, "Can't specify multiple log parameter"); *flags |= IPT_OSF_LOG; info->loglevel = atoi(argv[optind-1]);
similarly.
Quoted 12 lines
info->flags |= IPT_OSF_LOG; break; [...] return 1; }static void
osf_final_check(unsigned int flags){ if (!flags) exit_error(PARAMETER_PROBLEM, "OS fingerprint match: You must specify `--genre'"); }
Quoted 13 lines
static struct xtables_match osf_match = {
.name = "osf",
.version = XTABLES_VERSION,
.size = XT_ALIGN(sizeof(struct ipt_osf_info)),
.userspacesize = XT_ALIGN(sizeof(struct ipt_osf_info)),
.help = osf_help,
.init = osf_init,
.parse = osf_parse,
.print = osf_print,
.final_check = osf_final_check,
.save = osf_save,
.extra_opts = osf_opts
};Needs .family = NFPROTO_IPV4.
Quoted 23 lines · Evgeniy Polyakov
net/netfilter/xt_osf.c new file +21
-0+struct ipt_osf_finger { +struct rcu_head rcu_head; +struct list_head finger_entry; +struct ipt_osf_user_finger finger; +}; +struct ipt_osf_finger_storage +{ +struct list_head finger_list; +spinlock_t finger_lock; +}; + +/* +* Indexed by dont-fragment bit. +* It is the only constant value in the fingerprint. +*/ +struct ipt_osf_finger_storage ipt_osf_fingers[2 ]; + +static DEFINE_PER_CPU(struct ipt_osf_message, ipt_osf_mbuf); +static void __exit ipt_osf_fini(void) +{ +struct ipt_osf_finger *f;
net/netfilter/xt_osf.c
Quoted 10 lines
net/netfilter/xt_osf.c new file +10
-0+int i; + +cn_del_callback (&cn_osf_id); +xt_unregister_match (&ipt_osf_match); + +rcu_read_lock (); +for (i=0 ; i<ARRAY_SIZE (ipt_osf_fingers); ++i) { +struct ipt_osf_finger_storage *st = &ipt_osf_fingers[i]; + +list_for_each_entry_rcu (f, &st->finger_list, finger_entry) {
spin_lock(&st->finger_lock); //???
net/netfilter/xt_osf.c line 459
Quoted 1 line
net/netfilter/xt_osf.c new file +1
-0+list_del_rcu (&f->finger_entry);
spin_unlock(&st->finger_lock);
net/netfilter/xt_osf.c line 463
Quoted 4 lines
net/netfilter/xt_osf.c new file +4
-0+call_rcu (&f->rcu_head, ipt_osf_finger_free_rcu); + } + } +rcu_read_unlock ();
Should the list_del_rcu() not be protected by a spinlock?
net/netfilter/xt_osf.c line 465
Quoted 1 line
net/netfilter/xt_osf.c new file +1
-0+rcu_barrier ();
In some of my code I call synchronize_net(), is it enough to call rcu_barrier()?
What is the difference between:
synchronize_rcu()
synchronize_net()
rcu_barrier()
Hilsen
Jesper Brouer
--
-------------------------------------------------------------------
MSc. Master of Computer Science
Dept. of Computer Science, University of Copenhagen Author of http://www.adsl-optimizer.dk
-------------------------------------------------------------------
Hi.
Quoted 18 lines · Jesper Dangaard Brouer
net/netfilter/xt_osf.c line 458
Quoted 13 lines
net/netfilter/xt_osf.c new file +13
-0+static void __exit ipt_osf_fini(void) +{ +struct ipt_osf_finger *f; +int i; + +cn_del_callback (&cn_osf_id); +xt_unregister_match (&ipt_osf_match); + +rcu_read_lock (); +for (i=0 ; i<ARRAY_SIZE (ipt_osf_fingers); ++i) { +struct ipt_osf_finger_storage *st = &ipt_osf_fingers[i]; + +list_for_each_entry_rcu (f, &st->finger_list, finger_entry) {
spin_lock(&st->finger_lock); //???net/netfilter/xt_osf.c line 459
Quoted 1 line
net/netfilter/xt_osf.c new file +1
-0+list_del_rcu (&f->finger_entry);
spin_unlock(&st->finger_lock);
Quoted 6 lines
net/netfilter/xt_osf.c line 463
Quoted 4 lines
net/netfilter/xt_osf.c new file +4
-0+call_rcu (&f->rcu_head, ipt_osf_finger_free_rcu); + } + } +rcu_read_unlock ();Should the
list_del_rcu()not be protected by a spinlock?
Not required at this place - all users are already unregistered and no code can access this list except module exit path.
Quoted 4 lines
net/netfilter/xt_osf.c line 465
Quoted 1 line
net/netfilter/xt_osf.c new file +1
-0+rcu_barrier ();In some of my code I call
synchronize_net(), is it enough to callrcu_barrier()?
It is enough here, rcu_barrier() will wait until all scheduled call_rcu() are completed, that's what we need. But in some cases we should only wait for the whole grace period to elapse, then one has to use synchronize_rcu() and friends. rcu_barrier() will wait for the callbacks to be executed, while they are executed after grace period has elapsed, so it implicitly includes synchronize_rcu(), but effectively they are the same: both functions register rcu callback and wait for the completion, rcu_barrier() is a bit more enhanced, since it has several types.
Quoted 4 lines
What is the difference between:
synchronize_rcu()
synchronize_net()
Those are essentially the same - synchronize_net() has additional might_sleep() call. Both will wait until grace period elapced - i.e. all currently RCU protected sections completed.
Quoted 1 line
rcu_barrier()
It will wait until all scheduled rcu callbacks are executed.
So from the description they look different, but implementation suggestes that effectively they are the same, except that there are a bit different invocation types for the barrier.
Signature · 1 line
Evgeniy Polyakov
Quoted 12 lines · Jesper Dangaard Brouer
net/netfilter/xt_osf.c line 463
Quoted 4 lines · Evgeniy Polyakov
net/netfilter/xt_osf.c new file +4
-0+call_rcu (&f->rcu_head, ipt_osf_finger_free_rcu); + } + } +rcu_read_unlock ();Should the
list_del_rcu()not be protected by a spinlock?Quoted 1 line
+
rcu_barrier();
Quoted 7 lines
In some of my code I call
synchronize_net(), is it enough to callrcu_barrier()?What is the difference between:
synchronize_rcu()
synchronize_net()
rcu_barrier()
synchronize_net() is just a call to synchronize_rcu(), so their functionality is equivalent. synchronize_net() is however only supposed to synchronize with RX packet processing, which is usually not enough for netfilter. So I prefer synchronize_rcu() for clarity.
Pablo Neira Ayuso
maintainerQuoted 11 lines · Evgeniy Polyakov
Hi.
Passive OS fingerprinting netfilter module allows to passively detect remote OS and perform various netfilter actions based on that knowledge. This module compares some data (WS, MSS, options and it's order, ttl, df and others) from packets with SYN bit set with dynamically loaded OS fingerprints.
Fingerprint matching rules can be downloaded from OpenBSD source tree and loaded via netlink connector into the kernel via special util found in archive. It will also listen for events about matching packets.
I like this feature. We have nfnetlink so I don't see why we should use the netlink connector instead.
BTW, is there any difference with regards to userspace p0f apart from having this integrated into iptables?
Signature · 1 line
"Los honestos son inadaptados sociales" -- Les Luthiers
Hi Pablo.
Quoted 6 lines · Pablo Neira Ayuso
Quoted 3 lines
Fingerprint matching rules can be downloaded from OpenBSD source tree and loaded via netlink connector into the kernel via special util found in archive. It will also listen for events about matching packets.
I like this feature. We have nfnetlink so I don't see why we should use the netlink connector instead.
OSF exists about 6 years already, netlink configuration was added in 2005, I do not remember if nfnetlink existed those days (IIRC it did not, since I reused ULOG netlink first), right now I just cleanup what was written before.
Quoted 2 lines
BTW, is there any difference with regards to userspace p0f apart from having this integrated into iptables?
There should be no major differences, there are some tweaks for the MTU comparison, maybe something else.
Signature · 1 line
Evgeniy Polyakov
Quoted 8 lines · Evgeniy Polyakov
Hi Pablo.
Quoted 5 lines · Pablo Neira Ayuso
Quoted 3 lines
Fingerprint matching rules can be downloaded from OpenBSD source tree and loaded via netlink connector into the kernel via special util found in archive. It will also listen for events about matching packets.
I like this feature. We have nfnetlink so I don't see why we should use the netlink connector instead.
I fully agree.
Quoted 4 lines
OSF exists about 6 years already, netlink configuration was added in 2005, I do not remember if nfnetlink existed those days (IIRC it did not, since I reused ULOG netlink first), right now I just cleanup what was written before.
We do have nfnetlink today however, so this argument does no longer apply. I don't mind the order in which things are fixed up of course, but before merging, it needs to be converted to nfnetlink.