2005-04-29 17:23:29 +02:00
|
|
|
/* audit.c -- Auditing support
|
2005-04-17 00:20:36 +02:00
|
|
|
* Gateway between the kernel (e.g., selinux) and the user-space audit daemon.
|
|
|
|
* System-call specific features have moved to auditsc.c
|
|
|
|
*
|
2007-01-19 20:39:55 +01:00
|
|
|
* Copyright 2003-2007 Red Hat Inc., Durham, North Carolina.
|
2005-04-17 00:20:36 +02:00
|
|
|
* All Rights Reserved.
|
|
|
|
*
|
|
|
|
* 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
|
|
|
|
*
|
|
|
|
* Written by Rickard E. (Rik) Faith <faith@redhat.com>
|
|
|
|
*
|
2008-03-01 21:01:11 +01:00
|
|
|
* Goals: 1) Integrate fully with Security Modules.
|
2005-04-17 00:20:36 +02:00
|
|
|
* 2) Minimal run-time overhead:
|
|
|
|
* a) Minimal when syscall auditing is disabled (audit_enable=0).
|
|
|
|
* b) Small when syscall auditing is enabled and no audit record
|
|
|
|
* is generated (defer as much work as possible to record
|
|
|
|
* generation time):
|
|
|
|
* i) context is allocated,
|
|
|
|
* ii) names from getname are stored without a copy, and
|
|
|
|
* iii) inode information stored from path_lookup.
|
|
|
|
* 3) Ability to disable syscall auditing at boot time (audit=0).
|
|
|
|
* 4) Usable by other parts of the kernel (if audit_log* is called,
|
|
|
|
* then a syscall record will be generated automatically for the
|
|
|
|
* current syscall).
|
|
|
|
* 5) Netlink interface to user-space.
|
|
|
|
* 6) Support low-overhead kernel-based filtering to minimize the
|
|
|
|
* information that must be passed to user-space.
|
|
|
|
*
|
2005-04-29 17:23:29 +02:00
|
|
|
* Example user-space utilities: http://people.redhat.com/sgrubb/audit/
|
2005-04-17 00:20:36 +02:00
|
|
|
*/
|
|
|
|
|
|
|
|
#include <linux/init.h>
|
|
|
|
#include <asm/types.h>
|
2006-01-19 02:44:07 +01:00
|
|
|
#include <asm/atomic.h>
|
2005-04-17 00:20:36 +02:00
|
|
|
#include <linux/mm.h>
|
|
|
|
#include <linux/module.h>
|
include cleanup: Update gfp.h and slab.h includes to prepare for breaking implicit slab.h inclusion from percpu.h
percpu.h is included by sched.h and module.h and thus ends up being
included when building most .c files. percpu.h includes slab.h which
in turn includes gfp.h making everything defined by the two files
universally available and complicating inclusion dependencies.
percpu.h -> slab.h dependency is about to be removed. Prepare for
this change by updating users of gfp and slab facilities include those
headers directly instead of assuming availability. As this conversion
needs to touch large number of source files, the following script is
used as the basis of conversion.
http://userweb.kernel.org/~tj/misc/slabh-sweep.py
The script does the followings.
* Scan files for gfp and slab usages and update includes such that
only the necessary includes are there. ie. if only gfp is used,
gfp.h, if slab is used, slab.h.
* When the script inserts a new include, it looks at the include
blocks and try to put the new include such that its order conforms
to its surrounding. It's put in the include block which contains
core kernel includes, in the same order that the rest are ordered -
alphabetical, Christmas tree, rev-Xmas-tree or at the end if there
doesn't seem to be any matching order.
* If the script can't find a place to put a new include (mostly
because the file doesn't have fitting include block), it prints out
an error message indicating which .h file needs to be added to the
file.
The conversion was done in the following steps.
1. The initial automatic conversion of all .c files updated slightly
over 4000 files, deleting around 700 includes and adding ~480 gfp.h
and ~3000 slab.h inclusions. The script emitted errors for ~400
files.
2. Each error was manually checked. Some didn't need the inclusion,
some needed manual addition while adding it to implementation .h or
embedding .c file was more appropriate for others. This step added
inclusions to around 150 files.
3. The script was run again and the output was compared to the edits
from #2 to make sure no file was left behind.
4. Several build tests were done and a couple of problems were fixed.
e.g. lib/decompress_*.c used malloc/free() wrappers around slab
APIs requiring slab.h to be added manually.
5. The script was run on all .h files but without automatically
editing them as sprinkling gfp.h and slab.h inclusions around .h
files could easily lead to inclusion dependency hell. Most gfp.h
inclusion directives were ignored as stuff from gfp.h was usually
wildly available and often used in preprocessor macros. Each
slab.h inclusion directive was examined and added manually as
necessary.
6. percpu.h was updated not to include slab.h.
7. Build test were done on the following configurations and failures
were fixed. CONFIG_GCOV_KERNEL was turned off for all tests (as my
distributed build env didn't work with gcov compiles) and a few
more options had to be turned off depending on archs to make things
build (like ipr on powerpc/64 which failed due to missing writeq).
* x86 and x86_64 UP and SMP allmodconfig and a custom test config.
* powerpc and powerpc64 SMP allmodconfig
* sparc and sparc64 SMP allmodconfig
* ia64 SMP allmodconfig
* s390 SMP allmodconfig
* alpha SMP allmodconfig
* um on x86_64 SMP allmodconfig
8. percpu.h modifications were reverted so that it could be applied as
a separate patch and serve as bisection point.
Given the fact that I had only a couple of failures from tests on step
6, I'm fairly confident about the coverage of this conversion patch.
If there is a breakage, it's likely to be something in one of the arch
headers which should be easily discoverable easily on most builds of
the specific arch.
Signed-off-by: Tejun Heo <tj@kernel.org>
Guess-its-ok-by: Christoph Lameter <cl@linux-foundation.org>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: Lee Schermerhorn <Lee.Schermerhorn@hp.com>
2010-03-24 09:04:11 +01:00
|
|
|
#include <linux/slab.h>
|
2005-05-19 11:56:58 +02:00
|
|
|
#include <linux/err.h>
|
|
|
|
#include <linux/kthread.h>
|
2005-04-17 00:20:36 +02:00
|
|
|
|
|
|
|
#include <linux/audit.h>
|
|
|
|
|
|
|
|
#include <net/sock.h>
|
2006-02-07 18:05:27 +01:00
|
|
|
#include <net/netlink.h>
|
2005-04-17 00:20:36 +02:00
|
|
|
#include <linux/skbuff.h>
|
2011-06-30 13:31:57 +02:00
|
|
|
#ifdef CONFIG_SECURITY
|
|
|
|
#include <linux/security.h>
|
|
|
|
#endif
|
2005-04-17 00:20:36 +02:00
|
|
|
#include <linux/netlink.h>
|
2006-12-07 05:34:23 +01:00
|
|
|
#include <linux/freezer.h>
|
Audit: add TTY input auditing
Add TTY input auditing, used to audit system administrator's actions. This is
required by various security standards such as DCID 6/3 and PCI to provide
non-repudiation of administrator's actions and to allow a review of past
actions if the administrator seems to overstep their duties or if the system
becomes misconfigured for unknown reasons. These requirements do not make it
necessary to audit TTY output as well.
Compared to an user-space keylogger, this approach records TTY input using the
audit subsystem, correlated with other audit events, and it is completely
transparent to the user-space application (e.g. the console ioctls still
work).
TTY input auditing works on a higher level than auditing all system calls
within the session, which would produce an overwhelming amount of mostly
useless audit events.
Add an "audit_tty" attribute, inherited across fork (). Data read from TTYs
by process with the attribute is sent to the audit subsystem by the kernel.
The audit netlink interface is extended to allow modifying the audit_tty
attribute, and to allow sending explanatory audit events from user-space (for
example, a shell might send an event containing the final command, after the
interactive command-line editing and history expansion is performed, which
might be difficult to decipher from the TTY input alone).
Because the "audit_tty" attribute is inherited across fork (), it would be set
e.g. for sshd restarted within an audited session. To prevent this, the
audit_tty attribute is cleared when a process with no open TTY file
descriptors (e.g. after daemon startup) opens a TTY.
See https://www.redhat.com/archives/linux-audit/2007-June/msg00000.html for a
more detailed rationale document for an older version of this patch.
[akpm@linux-foundation.org: build fix]
Signed-off-by: Miloslav Trmac <mitr@redhat.com>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Cc: Alan Cox <alan@lxorguk.ukuu.org.uk>
Cc: Paul Fulghum <paulkf@microgate.com>
Cc: Casey Schaufler <casey@schaufler-ca.com>
Cc: Steve Grubb <sgrubb@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-07-16 08:40:56 +02:00
|
|
|
#include <linux/tty.h>
|
2006-03-11 01:14:06 +01:00
|
|
|
|
|
|
|
#include "audit.h"
|
2005-04-17 00:20:36 +02:00
|
|
|
|
2008-11-05 18:47:09 +01:00
|
|
|
/* No auditing will take place until audit_initialized == AUDIT_INITIALIZED.
|
2005-04-17 00:20:36 +02:00
|
|
|
* (Initialization happens after skb_init is called.) */
|
2008-11-05 18:47:09 +01:00
|
|
|
#define AUDIT_DISABLED -1
|
|
|
|
#define AUDIT_UNINITIALIZED 0
|
|
|
|
#define AUDIT_INITIALIZED 1
|
2005-04-17 00:20:36 +02:00
|
|
|
static int audit_initialized;
|
|
|
|
|
2008-01-07 23:09:31 +01:00
|
|
|
#define AUDIT_OFF 0
|
|
|
|
#define AUDIT_ON 1
|
|
|
|
#define AUDIT_LOCKED 2
|
2005-04-17 00:20:36 +02:00
|
|
|
int audit_enabled;
|
2008-01-08 23:38:31 +01:00
|
|
|
int audit_ever_enabled;
|
2005-04-17 00:20:36 +02:00
|
|
|
|
2011-01-18 06:48:12 +01:00
|
|
|
EXPORT_SYMBOL_GPL(audit_enabled);
|
|
|
|
|
2005-04-17 00:20:36 +02:00
|
|
|
/* Default state when kernel boots without any parameters. */
|
|
|
|
static int audit_default;
|
|
|
|
|
|
|
|
/* If auditing cannot proceed, audit_failure selects what happens. */
|
|
|
|
static int audit_failure = AUDIT_FAIL_PRINTK;
|
|
|
|
|
audit: netlink socket can be auto-bound to pid other than current->pid (v2)
From: Pavel Emelyanov <xemul@openvz.org>
This patch is based on the one from Thomas.
The kauditd_thread() calls the netlink_unicast() and passes
the audit_pid to it. The audit_pid, in turn, is received from
the user space and the tool (I've checked the audit v1.6.9)
uses getpid() to pass one in the kernel. Besides, this tool
doesn't bind the netlink socket to this id, but simply creates
it allowing the kernel to auto-bind one.
That's the preamble.
The problem is that netlink_autobind() _does_not_ guarantees
that the socket will be auto-bound to the current pid. Instead
it uses the current pid as a hint to start looking for a free
id. So, in case of conflict, the audit messages can be sent
to a wrong socket. This can happen (it's unlikely, but can be)
in case some task opens more than one netlink sockets and then
the audit one starts - in this case the audit's pid can be busy
and its socket will be bound to another id.
The proposal is to introduce an audit_nlk_pid in audit subsys,
that will point to the netlink socket to send packets to. It
will most often be equal to audit_pid. The socket id can be
got from the skb's netlink CB right in the audit_receive_msg.
The audit_nlk_pid reset to 0 is not required, since all the
decisions are taken based on audit_pid value only.
Later, if the audit tools will bind the socket themselves, the
kernel will have to provide a way to setup the audit_nlk_pid
as well.
A good side effect of this patch is that audit_pid can later
be converted to struct pid, as it is not longer safe to use
pid_t-s in the presence of pid namespaces. But audit code still
uses the tgid from task_struct in the audit_signal_info and in
the audit_filter_syscall.
Signed-off-by: Thomas Graf <tgraf@suug.ch>
Signed-off-by: Pavel Emelyanov <xemul@openvz.org>
Acked-by: Eric Paris <eparis@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2008-03-20 23:39:41 +01:00
|
|
|
/*
|
|
|
|
* If audit records are to be written to the netlink socket, audit_pid
|
|
|
|
* contains the pid of the auditd process and audit_nlk_pid contains
|
|
|
|
* the pid to use to send netlink messages to that process.
|
|
|
|
*/
|
2005-05-06 13:38:39 +02:00
|
|
|
int audit_pid;
|
audit: netlink socket can be auto-bound to pid other than current->pid (v2)
From: Pavel Emelyanov <xemul@openvz.org>
This patch is based on the one from Thomas.
The kauditd_thread() calls the netlink_unicast() and passes
the audit_pid to it. The audit_pid, in turn, is received from
the user space and the tool (I've checked the audit v1.6.9)
uses getpid() to pass one in the kernel. Besides, this tool
doesn't bind the netlink socket to this id, but simply creates
it allowing the kernel to auto-bind one.
That's the preamble.
The problem is that netlink_autobind() _does_not_ guarantees
that the socket will be auto-bound to the current pid. Instead
it uses the current pid as a hint to start looking for a free
id. So, in case of conflict, the audit messages can be sent
to a wrong socket. This can happen (it's unlikely, but can be)
in case some task opens more than one netlink sockets and then
the audit one starts - in this case the audit's pid can be busy
and its socket will be bound to another id.
The proposal is to introduce an audit_nlk_pid in audit subsys,
that will point to the netlink socket to send packets to. It
will most often be equal to audit_pid. The socket id can be
got from the skb's netlink CB right in the audit_receive_msg.
The audit_nlk_pid reset to 0 is not required, since all the
decisions are taken based on audit_pid value only.
Later, if the audit tools will bind the socket themselves, the
kernel will have to provide a way to setup the audit_nlk_pid
as well.
A good side effect of this patch is that audit_pid can later
be converted to struct pid, as it is not longer safe to use
pid_t-s in the presence of pid namespaces. But audit code still
uses the tgid from task_struct in the audit_signal_info and in
the audit_filter_syscall.
Signed-off-by: Thomas Graf <tgraf@suug.ch>
Signed-off-by: Pavel Emelyanov <xemul@openvz.org>
Acked-by: Eric Paris <eparis@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2008-03-20 23:39:41 +01:00
|
|
|
static int audit_nlk_pid;
|
2005-04-17 00:20:36 +02:00
|
|
|
|
2005-09-13 21:47:11 +02:00
|
|
|
/* If audit_rate_limit is non-zero, limit the rate of sending audit records
|
2005-04-17 00:20:36 +02:00
|
|
|
* to that number per second. This prevents DoS attacks, but results in
|
|
|
|
* audit records being dropped. */
|
|
|
|
static int audit_rate_limit;
|
|
|
|
|
|
|
|
/* Number of outstanding audit_buffers allowed. */
|
|
|
|
static int audit_backlog_limit = 64;
|
2005-07-02 15:08:48 +02:00
|
|
|
static int audit_backlog_wait_time = 60 * HZ;
|
|
|
|
static int audit_backlog_wait_overflow = 0;
|
2005-04-17 00:20:36 +02:00
|
|
|
|
2005-05-06 13:38:39 +02:00
|
|
|
/* The identity of the user shutting down the audit system. */
|
|
|
|
uid_t audit_sig_uid = -1;
|
|
|
|
pid_t audit_sig_pid = -1;
|
2006-05-25 16:19:47 +02:00
|
|
|
u32 audit_sig_sid = 0;
|
2005-05-06 13:38:39 +02:00
|
|
|
|
2005-04-17 00:20:36 +02:00
|
|
|
/* Records can be lost in several ways:
|
|
|
|
0) [suppressed in audit_alloc]
|
|
|
|
1) out of memory in audit_log_start [kmalloc of struct audit_buffer]
|
|
|
|
2) out of memory in audit_log_move [alloc_skb]
|
|
|
|
3) suppressed due to audit_rate_limit
|
|
|
|
4) suppressed due to audit_backlog_limit
|
|
|
|
*/
|
|
|
|
static atomic_t audit_lost = ATOMIC_INIT(0);
|
|
|
|
|
|
|
|
/* The netlink socket. */
|
|
|
|
static struct sock *audit_sock;
|
|
|
|
|
[PATCH] audit: path-based rules
In this implementation, audit registers inotify watches on the parent
directories of paths specified in audit rules. When audit's inotify
event handler is called, it updates any affected rules based on the
filesystem event. If the parent directory is renamed, removed, or its
filesystem is unmounted, audit removes all rules referencing that
inotify watch.
To keep things simple, this implementation limits location-based
auditing to the directory entries in an existing directory. Given
a path-based rule for /foo/bar/passwd, the following table applies:
passwd modified -- audit event logged
passwd replaced -- audit event logged, rules list updated
bar renamed -- rule removed
foo renamed -- untracked, meaning that the rule now applies to
the new location
Audit users typically want to have many rules referencing filesystem
objects, which can significantly impact filtering performance. This
patch also adds an inode-number-based rule hash to mitigate this
situation.
The patch is relative to the audit git tree:
http://kernel.org/git/?p=linux/kernel/git/viro/audit-current.git;a=summary
and uses the inotify kernel API:
http://lkml.org/lkml/2006/6/1/145
Signed-off-by: Amy Griffis <amy.griffis@hp.com>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
2006-04-07 22:55:56 +02:00
|
|
|
/* Hash for inode-based rules */
|
|
|
|
struct list_head audit_inode_hash[AUDIT_INODE_BUCKETS];
|
|
|
|
|
2005-05-19 11:56:58 +02:00
|
|
|
/* The audit_freelist is a list of pre-allocated audit buffers (if more
|
2005-04-17 00:20:36 +02:00
|
|
|
* than AUDIT_MAXFREE are in use, the audit buffer is freed instead of
|
|
|
|
* being placed on the freelist). */
|
|
|
|
static DEFINE_SPINLOCK(audit_freelist_lock);
|
2005-09-13 21:47:11 +02:00
|
|
|
static int audit_freelist_count;
|
2005-04-17 00:20:36 +02:00
|
|
|
static LIST_HEAD(audit_freelist);
|
|
|
|
|
2005-05-19 11:56:58 +02:00
|
|
|
static struct sk_buff_head audit_skb_queue;
|
2008-04-18 16:02:28 +02:00
|
|
|
/* queue of skbs to send to auditd when/if it comes back */
|
|
|
|
static struct sk_buff_head audit_skb_hold_queue;
|
2005-05-19 11:56:58 +02:00
|
|
|
static struct task_struct *kauditd_task;
|
|
|
|
static DECLARE_WAIT_QUEUE_HEAD(kauditd_wait);
|
2005-06-22 16:04:33 +02:00
|
|
|
static DECLARE_WAIT_QUEUE_HEAD(audit_backlog_wait);
|
2005-04-17 00:20:36 +02:00
|
|
|
|
[PATCH] audit: path-based rules
In this implementation, audit registers inotify watches on the parent
directories of paths specified in audit rules. When audit's inotify
event handler is called, it updates any affected rules based on the
filesystem event. If the parent directory is renamed, removed, or its
filesystem is unmounted, audit removes all rules referencing that
inotify watch.
To keep things simple, this implementation limits location-based
auditing to the directory entries in an existing directory. Given
a path-based rule for /foo/bar/passwd, the following table applies:
passwd modified -- audit event logged
passwd replaced -- audit event logged, rules list updated
bar renamed -- rule removed
foo renamed -- untracked, meaning that the rule now applies to
the new location
Audit users typically want to have many rules referencing filesystem
objects, which can significantly impact filtering performance. This
patch also adds an inode-number-based rule hash to mitigate this
situation.
The patch is relative to the audit git tree:
http://kernel.org/git/?p=linux/kernel/git/viro/audit-current.git;a=summary
and uses the inotify kernel API:
http://lkml.org/lkml/2006/6/1/145
Signed-off-by: Amy Griffis <amy.griffis@hp.com>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
2006-04-07 22:55:56 +02:00
|
|
|
/* Serialize requests from userspace. */
|
2009-06-24 06:02:38 +02:00
|
|
|
DEFINE_MUTEX(audit_cmd_mutex);
|
2005-04-17 00:20:36 +02:00
|
|
|
|
|
|
|
/* AUDIT_BUFSIZ is the size of the temporary buffer used for formatting
|
|
|
|
* audit records. Since printk uses a 1024 byte buffer, this buffer
|
|
|
|
* should be at least that large. */
|
|
|
|
#define AUDIT_BUFSIZ 1024
|
|
|
|
|
|
|
|
/* AUDIT_MAXFREE is the number of empty audit_buffers we keep on the
|
|
|
|
* audit_freelist. Doing so eliminates many kmalloc/kfree calls. */
|
|
|
|
#define AUDIT_MAXFREE (2*NR_CPUS)
|
|
|
|
|
|
|
|
/* The audit_buffer is used when formatting an audit record. The caller
|
|
|
|
* locks briefly to get the record off the freelist or to allocate the
|
|
|
|
* buffer, and locks briefly to send the buffer to the netlink layer or
|
|
|
|
* to place it on a transmit queue. Multiple audit_buffers can be in
|
|
|
|
* use simultaneously. */
|
|
|
|
struct audit_buffer {
|
|
|
|
struct list_head list;
|
2005-05-06 16:54:17 +02:00
|
|
|
struct sk_buff *skb; /* formatted skb ready to send */
|
2005-04-17 00:20:36 +02:00
|
|
|
struct audit_context *ctx; /* NULL or associated context */
|
2005-10-21 09:22:03 +02:00
|
|
|
gfp_t gfp_mask;
|
2005-04-17 00:20:36 +02:00
|
|
|
};
|
|
|
|
|
2008-04-18 16:11:04 +02:00
|
|
|
struct audit_reply {
|
|
|
|
int pid;
|
|
|
|
struct sk_buff *skb;
|
|
|
|
};
|
|
|
|
|
2005-05-13 19:17:42 +02:00
|
|
|
static void audit_set_pid(struct audit_buffer *ab, pid_t pid)
|
|
|
|
{
|
2008-01-08 00:14:19 +01:00
|
|
|
if (ab) {
|
|
|
|
struct nlmsghdr *nlh = nlmsg_hdr(ab->skb);
|
|
|
|
nlh->nlmsg_pid = pid;
|
|
|
|
}
|
2005-05-13 19:17:42 +02:00
|
|
|
}
|
|
|
|
|
2005-11-03 18:15:16 +01:00
|
|
|
void audit_panic(const char *message)
|
2005-04-17 00:20:36 +02:00
|
|
|
{
|
|
|
|
switch (audit_failure)
|
|
|
|
{
|
|
|
|
case AUDIT_FAIL_SILENT:
|
|
|
|
break;
|
|
|
|
case AUDIT_FAIL_PRINTK:
|
2008-01-24 04:55:05 +01:00
|
|
|
if (printk_ratelimit())
|
|
|
|
printk(KERN_ERR "audit: %s\n", message);
|
2005-04-17 00:20:36 +02:00
|
|
|
break;
|
|
|
|
case AUDIT_FAIL_PANIC:
|
2008-02-21 21:53:05 +01:00
|
|
|
/* test audit_pid since printk is always losey, why bother? */
|
|
|
|
if (audit_pid)
|
|
|
|
panic("audit: %s\n", message);
|
2005-04-17 00:20:36 +02:00
|
|
|
break;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
static inline int audit_rate_check(void)
|
|
|
|
{
|
|
|
|
static unsigned long last_check = 0;
|
|
|
|
static int messages = 0;
|
|
|
|
static DEFINE_SPINLOCK(lock);
|
|
|
|
unsigned long flags;
|
|
|
|
unsigned long now;
|
|
|
|
unsigned long elapsed;
|
|
|
|
int retval = 0;
|
|
|
|
|
|
|
|
if (!audit_rate_limit) return 1;
|
|
|
|
|
|
|
|
spin_lock_irqsave(&lock, flags);
|
|
|
|
if (++messages < audit_rate_limit) {
|
|
|
|
retval = 1;
|
|
|
|
} else {
|
|
|
|
now = jiffies;
|
|
|
|
elapsed = now - last_check;
|
|
|
|
if (elapsed > HZ) {
|
|
|
|
last_check = now;
|
|
|
|
messages = 0;
|
|
|
|
retval = 1;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
spin_unlock_irqrestore(&lock, flags);
|
|
|
|
|
|
|
|
return retval;
|
|
|
|
}
|
|
|
|
|
2005-09-13 21:47:11 +02:00
|
|
|
/**
|
|
|
|
* audit_log_lost - conditionally log lost audit message event
|
|
|
|
* @message: the message stating reason for lost audit message
|
|
|
|
*
|
|
|
|
* Emit at least 1 message per second, even if audit_rate_check is
|
|
|
|
* throttling.
|
|
|
|
* Always increment the lost messages counter.
|
|
|
|
*/
|
2005-04-17 00:20:36 +02:00
|
|
|
void audit_log_lost(const char *message)
|
|
|
|
{
|
|
|
|
static unsigned long last_msg = 0;
|
|
|
|
static DEFINE_SPINLOCK(lock);
|
|
|
|
unsigned long flags;
|
|
|
|
unsigned long now;
|
|
|
|
int print;
|
|
|
|
|
|
|
|
atomic_inc(&audit_lost);
|
|
|
|
|
|
|
|
print = (audit_failure == AUDIT_FAIL_PANIC || !audit_rate_limit);
|
|
|
|
|
|
|
|
if (!print) {
|
|
|
|
spin_lock_irqsave(&lock, flags);
|
|
|
|
now = jiffies;
|
|
|
|
if (now - last_msg > HZ) {
|
|
|
|
print = 1;
|
|
|
|
last_msg = now;
|
|
|
|
}
|
|
|
|
spin_unlock_irqrestore(&lock, flags);
|
|
|
|
}
|
|
|
|
|
|
|
|
if (print) {
|
2008-01-24 04:55:05 +01:00
|
|
|
if (printk_ratelimit())
|
|
|
|
printk(KERN_WARNING
|
|
|
|
"audit: audit_lost=%d audit_rate_limit=%d "
|
|
|
|
"audit_backlog_limit=%d\n",
|
|
|
|
atomic_read(&audit_lost),
|
|
|
|
audit_rate_limit,
|
|
|
|
audit_backlog_limit);
|
2005-04-17 00:20:36 +02:00
|
|
|
audit_panic(message);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2008-01-07 23:09:31 +01:00
|
|
|
static int audit_log_config_change(char *function_name, int new, int old,
|
2008-04-18 16:09:25 +02:00
|
|
|
uid_t loginuid, u32 sessionid, u32 sid,
|
|
|
|
int allow_changes)
|
2005-04-17 00:20:36 +02:00
|
|
|
{
|
2008-01-07 23:09:31 +01:00
|
|
|
struct audit_buffer *ab;
|
|
|
|
int rc = 0;
|
2006-04-02 01:29:34 +02:00
|
|
|
|
2008-01-07 23:09:31 +01:00
|
|
|
ab = audit_log_start(NULL, GFP_KERNEL, AUDIT_CONFIG_CHANGE);
|
2008-04-18 16:09:25 +02:00
|
|
|
audit_log_format(ab, "%s=%d old=%d auid=%u ses=%u", function_name, new,
|
|
|
|
old, loginuid, sessionid);
|
2006-04-02 01:29:34 +02:00
|
|
|
if (sid) {
|
|
|
|
char *ctx = NULL;
|
|
|
|
u32 len;
|
2008-01-07 23:09:31 +01:00
|
|
|
|
Audit: use new LSM hooks instead of SELinux exports
Stop using the following exported SELinux interfaces:
selinux_get_inode_sid(inode, sid)
selinux_get_ipc_sid(ipcp, sid)
selinux_get_task_sid(tsk, sid)
selinux_sid_to_string(sid, ctx, len)
kfree(ctx)
and use following generic LSM equivalents respectively:
security_inode_getsecid(inode, secid)
security_ipc_getsecid*(ipcp, secid)
security_task_getsecid(tsk, secid)
security_sid_to_secctx(sid, ctx, len)
security_release_secctx(ctx, len)
Call security_release_secctx only if security_secid_to_secctx
succeeded.
Signed-off-by: Casey Schaufler <casey@schaufler-ca.com>
Signed-off-by: Ahmed S. Darwish <darwish.07@gmail.com>
Acked-by: James Morris <jmorris@namei.org>
Reviewed-by: Paul Moore <paul.moore@hp.com>
2008-03-01 20:54:38 +01:00
|
|
|
rc = security_secid_to_secctx(sid, &ctx, &len);
|
2008-01-07 23:09:31 +01:00
|
|
|
if (rc) {
|
|
|
|
audit_log_format(ab, " sid=%u", sid);
|
|
|
|
allow_changes = 0; /* Something weird, deny request */
|
|
|
|
} else {
|
|
|
|
audit_log_format(ab, " subj=%s", ctx);
|
Audit: use new LSM hooks instead of SELinux exports
Stop using the following exported SELinux interfaces:
selinux_get_inode_sid(inode, sid)
selinux_get_ipc_sid(ipcp, sid)
selinux_get_task_sid(tsk, sid)
selinux_sid_to_string(sid, ctx, len)
kfree(ctx)
and use following generic LSM equivalents respectively:
security_inode_getsecid(inode, secid)
security_ipc_getsecid*(ipcp, secid)
security_task_getsecid(tsk, secid)
security_sid_to_secctx(sid, ctx, len)
security_release_secctx(ctx, len)
Call security_release_secctx only if security_secid_to_secctx
succeeded.
Signed-off-by: Casey Schaufler <casey@schaufler-ca.com>
Signed-off-by: Ahmed S. Darwish <darwish.07@gmail.com>
Acked-by: James Morris <jmorris@namei.org>
Reviewed-by: Paul Moore <paul.moore@hp.com>
2008-03-01 20:54:38 +01:00
|
|
|
security_release_secctx(ctx, len);
|
2008-01-07 23:09:31 +01:00
|
|
|
}
|
2007-01-19 20:39:55 +01:00
|
|
|
}
|
2008-01-07 23:09:31 +01:00
|
|
|
audit_log_format(ab, " res=%d", allow_changes);
|
|
|
|
audit_log_end(ab);
|
2007-01-19 20:39:55 +01:00
|
|
|
return rc;
|
2005-04-17 00:20:36 +02:00
|
|
|
}
|
|
|
|
|
2008-01-07 23:09:31 +01:00
|
|
|
static int audit_do_config_change(char *function_name, int *to_change,
|
2008-04-18 16:09:25 +02:00
|
|
|
int new, uid_t loginuid, u32 sessionid,
|
|
|
|
u32 sid)
|
2005-04-17 00:20:36 +02:00
|
|
|
{
|
2008-01-07 23:09:31 +01:00
|
|
|
int allow_changes, rc = 0, old = *to_change;
|
2007-01-19 20:39:55 +01:00
|
|
|
|
|
|
|
/* check if we are locked */
|
2008-01-07 23:09:31 +01:00
|
|
|
if (audit_enabled == AUDIT_LOCKED)
|
|
|
|
allow_changes = 0;
|
2007-01-19 20:39:55 +01:00
|
|
|
else
|
2008-01-07 23:09:31 +01:00
|
|
|
allow_changes = 1;
|
2006-04-02 01:29:34 +02:00
|
|
|
|
2008-01-07 23:09:31 +01:00
|
|
|
if (audit_enabled != AUDIT_OFF) {
|
2008-04-18 16:09:25 +02:00
|
|
|
rc = audit_log_config_change(function_name, new, old, loginuid,
|
|
|
|
sessionid, sid, allow_changes);
|
2008-01-07 23:09:31 +01:00
|
|
|
if (rc)
|
|
|
|
allow_changes = 0;
|
2007-01-19 20:39:55 +01:00
|
|
|
}
|
|
|
|
|
|
|
|
/* If we are allowed, make the change */
|
2008-01-07 23:09:31 +01:00
|
|
|
if (allow_changes == 1)
|
|
|
|
*to_change = new;
|
2007-01-19 20:39:55 +01:00
|
|
|
/* Not allowed, update reason */
|
|
|
|
else if (rc == 0)
|
|
|
|
rc = -EPERM;
|
|
|
|
return rc;
|
2005-04-17 00:20:36 +02:00
|
|
|
}
|
|
|
|
|
2008-04-18 16:09:25 +02:00
|
|
|
static int audit_set_rate_limit(int limit, uid_t loginuid, u32 sessionid,
|
|
|
|
u32 sid)
|
2005-04-17 00:20:36 +02:00
|
|
|
{
|
2008-01-07 23:09:31 +01:00
|
|
|
return audit_do_config_change("audit_rate_limit", &audit_rate_limit,
|
2008-04-18 16:09:25 +02:00
|
|
|
limit, loginuid, sessionid, sid);
|
2008-01-07 23:09:31 +01:00
|
|
|
}
|
2006-04-02 01:29:34 +02:00
|
|
|
|
2008-04-18 16:09:25 +02:00
|
|
|
static int audit_set_backlog_limit(int limit, uid_t loginuid, u32 sessionid,
|
|
|
|
u32 sid)
|
2008-01-07 23:09:31 +01:00
|
|
|
{
|
|
|
|
return audit_do_config_change("audit_backlog_limit", &audit_backlog_limit,
|
2008-04-18 16:09:25 +02:00
|
|
|
limit, loginuid, sessionid, sid);
|
2008-01-07 23:09:31 +01:00
|
|
|
}
|
2007-01-19 20:39:55 +01:00
|
|
|
|
2008-04-18 16:09:25 +02:00
|
|
|
static int audit_set_enabled(int state, uid_t loginuid, u32 sessionid, u32 sid)
|
2008-01-07 23:09:31 +01:00
|
|
|
{
|
2008-01-08 23:38:31 +01:00
|
|
|
int rc;
|
2008-01-07 23:09:31 +01:00
|
|
|
if (state < AUDIT_OFF || state > AUDIT_LOCKED)
|
|
|
|
return -EINVAL;
|
2007-01-19 20:39:55 +01:00
|
|
|
|
2008-01-08 23:38:31 +01:00
|
|
|
rc = audit_do_config_change("audit_enabled", &audit_enabled, state,
|
2008-04-18 16:09:25 +02:00
|
|
|
loginuid, sessionid, sid);
|
2008-01-08 23:38:31 +01:00
|
|
|
|
|
|
|
if (!rc)
|
|
|
|
audit_ever_enabled |= !!state;
|
|
|
|
|
|
|
|
return rc;
|
2005-04-17 00:20:36 +02:00
|
|
|
}
|
|
|
|
|
2008-04-18 16:09:25 +02:00
|
|
|
static int audit_set_failure(int state, uid_t loginuid, u32 sessionid, u32 sid)
|
2005-04-17 00:20:36 +02:00
|
|
|
{
|
|
|
|
if (state != AUDIT_FAIL_SILENT
|
|
|
|
&& state != AUDIT_FAIL_PRINTK
|
|
|
|
&& state != AUDIT_FAIL_PANIC)
|
|
|
|
return -EINVAL;
|
2006-04-02 01:29:34 +02:00
|
|
|
|
2008-01-07 23:09:31 +01:00
|
|
|
return audit_do_config_change("audit_failure", &audit_failure, state,
|
2008-04-18 16:09:25 +02:00
|
|
|
loginuid, sessionid, sid);
|
2005-04-17 00:20:36 +02:00
|
|
|
}
|
|
|
|
|
2008-04-18 16:02:28 +02:00
|
|
|
/*
|
|
|
|
* Queue skbs to be sent to auditd when/if it comes back. These skbs should
|
|
|
|
* already have been sent via prink/syslog and so if these messages are dropped
|
|
|
|
* it is not a huge concern since we already passed the audit_log_lost()
|
|
|
|
* notification and stuff. This is just nice to get audit messages during
|
|
|
|
* boot before auditd is running or messages generated while auditd is stopped.
|
|
|
|
* This only holds messages is audit_default is set, aka booting with audit=1
|
|
|
|
* or building your kernel that way.
|
|
|
|
*/
|
|
|
|
static void audit_hold_skb(struct sk_buff *skb)
|
|
|
|
{
|
|
|
|
if (audit_default &&
|
|
|
|
skb_queue_len(&audit_skb_hold_queue) < audit_backlog_limit)
|
|
|
|
skb_queue_tail(&audit_skb_hold_queue, skb);
|
|
|
|
else
|
|
|
|
kfree_skb(skb);
|
|
|
|
}
|
|
|
|
|
2009-06-11 20:31:35 +02:00
|
|
|
/*
|
|
|
|
* For one reason or another this nlh isn't getting delivered to the userspace
|
|
|
|
* audit daemon, just send it to printk.
|
|
|
|
*/
|
|
|
|
static void audit_printk_skb(struct sk_buff *skb)
|
|
|
|
{
|
|
|
|
struct nlmsghdr *nlh = nlmsg_hdr(skb);
|
|
|
|
char *data = NLMSG_DATA(nlh);
|
|
|
|
|
|
|
|
if (nlh->nlmsg_type != AUDIT_EOE) {
|
|
|
|
if (printk_ratelimit())
|
|
|
|
printk(KERN_NOTICE "type=%d %s\n", nlh->nlmsg_type, data);
|
|
|
|
else
|
|
|
|
audit_log_lost("printk limit exceeded\n");
|
|
|
|
}
|
|
|
|
|
|
|
|
audit_hold_skb(skb);
|
|
|
|
}
|
|
|
|
|
2008-04-18 16:02:28 +02:00
|
|
|
static void kauditd_send_skb(struct sk_buff *skb)
|
|
|
|
{
|
|
|
|
int err;
|
|
|
|
/* take a reference in case we can't send it and we want to hold it */
|
|
|
|
skb_get(skb);
|
|
|
|
err = netlink_unicast(audit_sock, skb, audit_nlk_pid, 0);
|
|
|
|
if (err < 0) {
|
2009-12-18 21:40:42 +01:00
|
|
|
BUG_ON(err != -ECONNREFUSED); /* Shouldn't happen */
|
2008-04-18 16:02:28 +02:00
|
|
|
printk(KERN_ERR "audit: *NO* daemon at audit_pid=%d\n", audit_pid);
|
2010-10-22 17:43:17 +02:00
|
|
|
audit_log_lost("auditd disappeared\n");
|
2008-04-18 16:02:28 +02:00
|
|
|
audit_pid = 0;
|
|
|
|
/* we might get lucky and get this in the next auditd */
|
|
|
|
audit_hold_skb(skb);
|
|
|
|
} else
|
|
|
|
/* drop the extra reference if sent ok */
|
2010-07-20 08:45:56 +02:00
|
|
|
consume_skb(skb);
|
2008-04-18 16:02:28 +02:00
|
|
|
}
|
|
|
|
|
2006-01-08 10:02:17 +01:00
|
|
|
static int kauditd_thread(void *dummy)
|
2005-05-19 11:56:58 +02:00
|
|
|
{
|
|
|
|
struct sk_buff *skb;
|
|
|
|
|
2007-07-17 13:03:35 +02:00
|
|
|
set_freezable();
|
2006-10-06 09:43:48 +02:00
|
|
|
while (!kthread_should_stop()) {
|
2008-04-18 16:02:28 +02:00
|
|
|
/*
|
|
|
|
* if auditd just started drain the queue of messages already
|
|
|
|
* sent to syslog/printk. remember loss here is ok. we already
|
|
|
|
* called audit_log_lost() if it didn't go out normally. so the
|
|
|
|
* race between the skb_dequeue and the next check for audit_pid
|
|
|
|
* doesn't matter.
|
|
|
|
*
|
|
|
|
* if you ever find kauditd to be too slow we can get a perf win
|
|
|
|
* by doing our own locking and keeping better track if there
|
|
|
|
* are messages in this queue. I don't see the need now, but
|
|
|
|
* in 5 years when I want to play with this again I'll see this
|
|
|
|
* note and still have no friggin idea what i'm thinking today.
|
|
|
|
*/
|
|
|
|
if (audit_default && audit_pid) {
|
|
|
|
skb = skb_dequeue(&audit_skb_hold_queue);
|
|
|
|
if (unlikely(skb)) {
|
|
|
|
while (skb && audit_pid) {
|
|
|
|
kauditd_send_skb(skb);
|
|
|
|
skb = skb_dequeue(&audit_skb_hold_queue);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2005-05-19 11:56:58 +02:00
|
|
|
skb = skb_dequeue(&audit_skb_queue);
|
2005-06-22 16:04:33 +02:00
|
|
|
wake_up(&audit_backlog_wait);
|
2005-05-19 11:56:58 +02:00
|
|
|
if (skb) {
|
2008-04-18 16:02:28 +02:00
|
|
|
if (audit_pid)
|
|
|
|
kauditd_send_skb(skb);
|
2009-06-11 20:31:35 +02:00
|
|
|
else
|
|
|
|
audit_printk_skb(skb);
|
2005-05-19 11:56:58 +02:00
|
|
|
} else {
|
|
|
|
DECLARE_WAITQUEUE(wait, current);
|
|
|
|
set_current_state(TASK_INTERRUPTIBLE);
|
|
|
|
add_wait_queue(&kauditd_wait, &wait);
|
|
|
|
|
2005-12-12 09:37:22 +01:00
|
|
|
if (!skb_queue_len(&audit_skb_queue)) {
|
|
|
|
try_to_freeze();
|
2005-05-19 11:56:58 +02:00
|
|
|
schedule();
|
2005-12-12 09:37:22 +01:00
|
|
|
}
|
2005-05-19 11:56:58 +02:00
|
|
|
|
|
|
|
__set_current_state(TASK_RUNNING);
|
|
|
|
remove_wait_queue(&kauditd_wait, &wait);
|
|
|
|
}
|
|
|
|
}
|
2006-10-06 09:43:48 +02:00
|
|
|
return 0;
|
2005-05-19 11:56:58 +02:00
|
|
|
}
|
|
|
|
|
2008-04-18 16:09:25 +02:00
|
|
|
static int audit_prepare_user_tty(pid_t pid, uid_t loginuid, u32 sessionid)
|
Audit: add TTY input auditing
Add TTY input auditing, used to audit system administrator's actions. This is
required by various security standards such as DCID 6/3 and PCI to provide
non-repudiation of administrator's actions and to allow a review of past
actions if the administrator seems to overstep their duties or if the system
becomes misconfigured for unknown reasons. These requirements do not make it
necessary to audit TTY output as well.
Compared to an user-space keylogger, this approach records TTY input using the
audit subsystem, correlated with other audit events, and it is completely
transparent to the user-space application (e.g. the console ioctls still
work).
TTY input auditing works on a higher level than auditing all system calls
within the session, which would produce an overwhelming amount of mostly
useless audit events.
Add an "audit_tty" attribute, inherited across fork (). Data read from TTYs
by process with the attribute is sent to the audit subsystem by the kernel.
The audit netlink interface is extended to allow modifying the audit_tty
attribute, and to allow sending explanatory audit events from user-space (for
example, a shell might send an event containing the final command, after the
interactive command-line editing and history expansion is performed, which
might be difficult to decipher from the TTY input alone).
Because the "audit_tty" attribute is inherited across fork (), it would be set
e.g. for sshd restarted within an audited session. To prevent this, the
audit_tty attribute is cleared when a process with no open TTY file
descriptors (e.g. after daemon startup) opens a TTY.
See https://www.redhat.com/archives/linux-audit/2007-June/msg00000.html for a
more detailed rationale document for an older version of this patch.
[akpm@linux-foundation.org: build fix]
Signed-off-by: Miloslav Trmac <mitr@redhat.com>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Cc: Alan Cox <alan@lxorguk.ukuu.org.uk>
Cc: Paul Fulghum <paulkf@microgate.com>
Cc: Casey Schaufler <casey@schaufler-ca.com>
Cc: Steve Grubb <sgrubb@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-07-16 08:40:56 +02:00
|
|
|
{
|
|
|
|
struct task_struct *tsk;
|
|
|
|
int err;
|
|
|
|
|
2009-12-09 15:19:31 +01:00
|
|
|
rcu_read_lock();
|
2008-04-18 22:30:15 +02:00
|
|
|
tsk = find_task_by_vpid(pid);
|
2009-12-09 15:19:31 +01:00
|
|
|
if (!tsk) {
|
|
|
|
rcu_read_unlock();
|
|
|
|
return -ESRCH;
|
|
|
|
}
|
|
|
|
get_task_struct(tsk);
|
|
|
|
rcu_read_unlock();
|
|
|
|
err = tty_audit_push_task(tsk, loginuid, sessionid);
|
|
|
|
put_task_struct(tsk);
|
Audit: add TTY input auditing
Add TTY input auditing, used to audit system administrator's actions. This is
required by various security standards such as DCID 6/3 and PCI to provide
non-repudiation of administrator's actions and to allow a review of past
actions if the administrator seems to overstep their duties or if the system
becomes misconfigured for unknown reasons. These requirements do not make it
necessary to audit TTY output as well.
Compared to an user-space keylogger, this approach records TTY input using the
audit subsystem, correlated with other audit events, and it is completely
transparent to the user-space application (e.g. the console ioctls still
work).
TTY input auditing works on a higher level than auditing all system calls
within the session, which would produce an overwhelming amount of mostly
useless audit events.
Add an "audit_tty" attribute, inherited across fork (). Data read from TTYs
by process with the attribute is sent to the audit subsystem by the kernel.
The audit netlink interface is extended to allow modifying the audit_tty
attribute, and to allow sending explanatory audit events from user-space (for
example, a shell might send an event containing the final command, after the
interactive command-line editing and history expansion is performed, which
might be difficult to decipher from the TTY input alone).
Because the "audit_tty" attribute is inherited across fork (), it would be set
e.g. for sshd restarted within an audited session. To prevent this, the
audit_tty attribute is cleared when a process with no open TTY file
descriptors (e.g. after daemon startup) opens a TTY.
See https://www.redhat.com/archives/linux-audit/2007-June/msg00000.html for a
more detailed rationale document for an older version of this patch.
[akpm@linux-foundation.org: build fix]
Signed-off-by: Miloslav Trmac <mitr@redhat.com>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Cc: Alan Cox <alan@lxorguk.ukuu.org.uk>
Cc: Paul Fulghum <paulkf@microgate.com>
Cc: Casey Schaufler <casey@schaufler-ca.com>
Cc: Steve Grubb <sgrubb@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-07-16 08:40:56 +02:00
|
|
|
return err;
|
|
|
|
}
|
|
|
|
|
2006-05-22 07:09:24 +02:00
|
|
|
int audit_send_list(void *_dest)
|
|
|
|
{
|
|
|
|
struct audit_netlink_list *dest = _dest;
|
|
|
|
int pid = dest->pid;
|
|
|
|
struct sk_buff *skb;
|
|
|
|
|
|
|
|
/* wait for parent to finish and send an ACK */
|
[PATCH] audit: path-based rules
In this implementation, audit registers inotify watches on the parent
directories of paths specified in audit rules. When audit's inotify
event handler is called, it updates any affected rules based on the
filesystem event. If the parent directory is renamed, removed, or its
filesystem is unmounted, audit removes all rules referencing that
inotify watch.
To keep things simple, this implementation limits location-based
auditing to the directory entries in an existing directory. Given
a path-based rule for /foo/bar/passwd, the following table applies:
passwd modified -- audit event logged
passwd replaced -- audit event logged, rules list updated
bar renamed -- rule removed
foo renamed -- untracked, meaning that the rule now applies to
the new location
Audit users typically want to have many rules referencing filesystem
objects, which can significantly impact filtering performance. This
patch also adds an inode-number-based rule hash to mitigate this
situation.
The patch is relative to the audit git tree:
http://kernel.org/git/?p=linux/kernel/git/viro/audit-current.git;a=summary
and uses the inotify kernel API:
http://lkml.org/lkml/2006/6/1/145
Signed-off-by: Amy Griffis <amy.griffis@hp.com>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
2006-04-07 22:55:56 +02:00
|
|
|
mutex_lock(&audit_cmd_mutex);
|
|
|
|
mutex_unlock(&audit_cmd_mutex);
|
2006-05-22 07:09:24 +02:00
|
|
|
|
|
|
|
while ((skb = __skb_dequeue(&dest->q)) != NULL)
|
|
|
|
netlink_unicast(audit_sock, skb, pid, 0);
|
|
|
|
|
|
|
|
kfree(dest);
|
|
|
|
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
struct sk_buff *audit_make_reply(int pid, int seq, int type, int done,
|
2010-10-21 02:23:50 +02:00
|
|
|
int multi, const void *payload, int size)
|
2006-05-22 07:09:24 +02:00
|
|
|
{
|
|
|
|
struct sk_buff *skb;
|
|
|
|
struct nlmsghdr *nlh;
|
|
|
|
void *data;
|
|
|
|
int flags = multi ? NLM_F_MULTI : 0;
|
|
|
|
int t = done ? NLMSG_DONE : type;
|
|
|
|
|
2009-06-11 20:31:35 +02:00
|
|
|
skb = nlmsg_new(size, GFP_KERNEL);
|
2006-05-22 07:09:24 +02:00
|
|
|
if (!skb)
|
|
|
|
return NULL;
|
|
|
|
|
2009-06-11 20:31:35 +02:00
|
|
|
nlh = NLMSG_NEW(skb, pid, seq, t, size, flags);
|
|
|
|
data = NLMSG_DATA(nlh);
|
2006-05-22 07:09:24 +02:00
|
|
|
memcpy(data, payload, size);
|
|
|
|
return skb;
|
|
|
|
|
2009-06-11 20:31:35 +02:00
|
|
|
nlmsg_failure: /* Used by NLMSG_NEW */
|
2006-05-22 07:09:24 +02:00
|
|
|
if (skb)
|
|
|
|
kfree_skb(skb);
|
|
|
|
return NULL;
|
|
|
|
}
|
|
|
|
|
2008-04-18 16:11:04 +02:00
|
|
|
static int audit_send_reply_thread(void *arg)
|
|
|
|
{
|
|
|
|
struct audit_reply *reply = (struct audit_reply *)arg;
|
|
|
|
|
|
|
|
mutex_lock(&audit_cmd_mutex);
|
|
|
|
mutex_unlock(&audit_cmd_mutex);
|
|
|
|
|
|
|
|
/* Ignore failure. It'll only happen if the sender goes away,
|
|
|
|
because our timeout is set to infinite. */
|
|
|
|
netlink_unicast(audit_sock, reply->skb, reply->pid, 0);
|
|
|
|
kfree(reply);
|
|
|
|
return 0;
|
|
|
|
}
|
2005-09-13 21:47:11 +02:00
|
|
|
/**
|
|
|
|
* audit_send_reply - send an audit reply message via netlink
|
|
|
|
* @pid: process id to send reply to
|
|
|
|
* @seq: sequence number
|
|
|
|
* @type: audit message type
|
|
|
|
* @done: done (last) flag
|
|
|
|
* @multi: multi-part message flag
|
|
|
|
* @payload: payload data
|
|
|
|
* @size: payload size
|
|
|
|
*
|
|
|
|
* Allocates an skb, builds the netlink message, and sends it to the pid.
|
|
|
|
* No failure notifications.
|
|
|
|
*/
|
2010-10-21 02:23:50 +02:00
|
|
|
static void audit_send_reply(int pid, int seq, int type, int done, int multi,
|
|
|
|
const void *payload, int size)
|
2005-04-17 00:20:36 +02:00
|
|
|
{
|
2008-04-18 16:11:04 +02:00
|
|
|
struct sk_buff *skb;
|
|
|
|
struct task_struct *tsk;
|
|
|
|
struct audit_reply *reply = kmalloc(sizeof(struct audit_reply),
|
|
|
|
GFP_KERNEL);
|
|
|
|
|
|
|
|
if (!reply)
|
|
|
|
return;
|
|
|
|
|
2006-05-22 07:09:24 +02:00
|
|
|
skb = audit_make_reply(pid, seq, type, done, multi, payload, size);
|
2005-04-17 00:20:36 +02:00
|
|
|
if (!skb)
|
2008-05-15 01:11:48 +02:00
|
|
|
goto out;
|
2008-04-18 16:11:04 +02:00
|
|
|
|
|
|
|
reply->pid = pid;
|
|
|
|
reply->skb = skb;
|
|
|
|
|
|
|
|
tsk = kthread_run(audit_send_reply_thread, reply, "audit_send_reply");
|
2008-05-15 01:11:48 +02:00
|
|
|
if (!IS_ERR(tsk))
|
|
|
|
return;
|
|
|
|
kfree_skb(skb);
|
|
|
|
out:
|
|
|
|
kfree(reply);
|
2005-04-17 00:20:36 +02:00
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Check for appropriate CAP_AUDIT_ capabilities on incoming audit
|
|
|
|
* control messages.
|
|
|
|
*/
|
2006-06-27 22:26:11 +02:00
|
|
|
static int audit_netlink_ok(struct sk_buff *skb, u16 msg_type)
|
2005-04-17 00:20:36 +02:00
|
|
|
{
|
|
|
|
int err = 0;
|
|
|
|
|
|
|
|
switch (msg_type) {
|
|
|
|
case AUDIT_GET:
|
|
|
|
case AUDIT_LIST:
|
2006-02-07 18:05:27 +01:00
|
|
|
case AUDIT_LIST_RULES:
|
2005-04-17 00:20:36 +02:00
|
|
|
case AUDIT_SET:
|
|
|
|
case AUDIT_ADD:
|
2006-02-07 18:05:27 +01:00
|
|
|
case AUDIT_ADD_RULE:
|
2005-04-17 00:20:36 +02:00
|
|
|
case AUDIT_DEL:
|
2006-02-07 18:05:27 +01:00
|
|
|
case AUDIT_DEL_RULE:
|
2005-05-06 13:38:39 +02:00
|
|
|
case AUDIT_SIGNAL_INFO:
|
Audit: add TTY input auditing
Add TTY input auditing, used to audit system administrator's actions. This is
required by various security standards such as DCID 6/3 and PCI to provide
non-repudiation of administrator's actions and to allow a review of past
actions if the administrator seems to overstep their duties or if the system
becomes misconfigured for unknown reasons. These requirements do not make it
necessary to audit TTY output as well.
Compared to an user-space keylogger, this approach records TTY input using the
audit subsystem, correlated with other audit events, and it is completely
transparent to the user-space application (e.g. the console ioctls still
work).
TTY input auditing works on a higher level than auditing all system calls
within the session, which would produce an overwhelming amount of mostly
useless audit events.
Add an "audit_tty" attribute, inherited across fork (). Data read from TTYs
by process with the attribute is sent to the audit subsystem by the kernel.
The audit netlink interface is extended to allow modifying the audit_tty
attribute, and to allow sending explanatory audit events from user-space (for
example, a shell might send an event containing the final command, after the
interactive command-line editing and history expansion is performed, which
might be difficult to decipher from the TTY input alone).
Because the "audit_tty" attribute is inherited across fork (), it would be set
e.g. for sshd restarted within an audited session. To prevent this, the
audit_tty attribute is cleared when a process with no open TTY file
descriptors (e.g. after daemon startup) opens a TTY.
See https://www.redhat.com/archives/linux-audit/2007-June/msg00000.html for a
more detailed rationale document for an older version of this patch.
[akpm@linux-foundation.org: build fix]
Signed-off-by: Miloslav Trmac <mitr@redhat.com>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Cc: Alan Cox <alan@lxorguk.ukuu.org.uk>
Cc: Paul Fulghum <paulkf@microgate.com>
Cc: Casey Schaufler <casey@schaufler-ca.com>
Cc: Steve Grubb <sgrubb@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-07-16 08:40:56 +02:00
|
|
|
case AUDIT_TTY_GET:
|
|
|
|
case AUDIT_TTY_SET:
|
[PATCH] audit: watching subtrees
New kind of audit rule predicates: "object is visible in given subtree".
The part that can be sanely implemented, that is. Limitations:
* if you have hardlink from outside of tree, you'd better watch
it too (or just watch the object itself, obviously)
* if you mount something under a watched tree, tell audit
that new chunk should be added to watched subtrees
* if you umount something in a watched tree and it's still mounted
elsewhere, you will get matches on events happening there. New command
tells audit to recalculate the trees, trimming such sources of false
positives.
Note that it's _not_ about path - if something mounted in several places
(multiple mount, bindings, different namespaces, etc.), the match does
_not_ depend on which one we are using for access.
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
2007-07-22 14:04:18 +02:00
|
|
|
case AUDIT_TRIM:
|
|
|
|
case AUDIT_MAKE_EQUIV:
|
2006-06-27 22:26:11 +02:00
|
|
|
if (security_netlink_recv(skb, CAP_AUDIT_CONTROL))
|
2005-04-17 00:20:36 +02:00
|
|
|
err = -EPERM;
|
|
|
|
break;
|
2005-05-21 01:18:37 +02:00
|
|
|
case AUDIT_USER:
|
2007-05-08 09:29:20 +02:00
|
|
|
case AUDIT_FIRST_USER_MSG ... AUDIT_LAST_USER_MSG:
|
|
|
|
case AUDIT_FIRST_USER_MSG2 ... AUDIT_LAST_USER_MSG2:
|
2006-06-27 22:26:11 +02:00
|
|
|
if (security_netlink_recv(skb, CAP_AUDIT_WRITE))
|
2005-04-17 00:20:36 +02:00
|
|
|
err = -EPERM;
|
|
|
|
break;
|
|
|
|
default: /* bad msg */
|
|
|
|
err = -EINVAL;
|
|
|
|
}
|
|
|
|
|
|
|
|
return err;
|
|
|
|
}
|
|
|
|
|
2008-01-08 00:14:19 +01:00
|
|
|
static int audit_log_common_recv_msg(struct audit_buffer **ab, u16 msg_type,
|
2008-04-18 16:09:25 +02:00
|
|
|
u32 pid, u32 uid, uid_t auid, u32 ses,
|
|
|
|
u32 sid)
|
2008-01-08 00:14:19 +01:00
|
|
|
{
|
|
|
|
int rc = 0;
|
|
|
|
char *ctx = NULL;
|
|
|
|
u32 len;
|
|
|
|
|
|
|
|
if (!audit_enabled) {
|
|
|
|
*ab = NULL;
|
|
|
|
return rc;
|
|
|
|
}
|
|
|
|
|
|
|
|
*ab = audit_log_start(NULL, GFP_KERNEL, msg_type);
|
2008-04-18 16:09:25 +02:00
|
|
|
audit_log_format(*ab, "user pid=%d uid=%u auid=%u ses=%u",
|
|
|
|
pid, uid, auid, ses);
|
2008-01-08 00:14:19 +01:00
|
|
|
if (sid) {
|
Audit: use new LSM hooks instead of SELinux exports
Stop using the following exported SELinux interfaces:
selinux_get_inode_sid(inode, sid)
selinux_get_ipc_sid(ipcp, sid)
selinux_get_task_sid(tsk, sid)
selinux_sid_to_string(sid, ctx, len)
kfree(ctx)
and use following generic LSM equivalents respectively:
security_inode_getsecid(inode, secid)
security_ipc_getsecid*(ipcp, secid)
security_task_getsecid(tsk, secid)
security_sid_to_secctx(sid, ctx, len)
security_release_secctx(ctx, len)
Call security_release_secctx only if security_secid_to_secctx
succeeded.
Signed-off-by: Casey Schaufler <casey@schaufler-ca.com>
Signed-off-by: Ahmed S. Darwish <darwish.07@gmail.com>
Acked-by: James Morris <jmorris@namei.org>
Reviewed-by: Paul Moore <paul.moore@hp.com>
2008-03-01 20:54:38 +01:00
|
|
|
rc = security_secid_to_secctx(sid, &ctx, &len);
|
2008-01-08 00:14:19 +01:00
|
|
|
if (rc)
|
|
|
|
audit_log_format(*ab, " ssid=%u", sid);
|
Audit: use new LSM hooks instead of SELinux exports
Stop using the following exported SELinux interfaces:
selinux_get_inode_sid(inode, sid)
selinux_get_ipc_sid(ipcp, sid)
selinux_get_task_sid(tsk, sid)
selinux_sid_to_string(sid, ctx, len)
kfree(ctx)
and use following generic LSM equivalents respectively:
security_inode_getsecid(inode, secid)
security_ipc_getsecid*(ipcp, secid)
security_task_getsecid(tsk, secid)
security_sid_to_secctx(sid, ctx, len)
security_release_secctx(ctx, len)
Call security_release_secctx only if security_secid_to_secctx
succeeded.
Signed-off-by: Casey Schaufler <casey@schaufler-ca.com>
Signed-off-by: Ahmed S. Darwish <darwish.07@gmail.com>
Acked-by: James Morris <jmorris@namei.org>
Reviewed-by: Paul Moore <paul.moore@hp.com>
2008-03-01 20:54:38 +01:00
|
|
|
else {
|
2008-01-08 00:14:19 +01:00
|
|
|
audit_log_format(*ab, " subj=%s", ctx);
|
Audit: use new LSM hooks instead of SELinux exports
Stop using the following exported SELinux interfaces:
selinux_get_inode_sid(inode, sid)
selinux_get_ipc_sid(ipcp, sid)
selinux_get_task_sid(tsk, sid)
selinux_sid_to_string(sid, ctx, len)
kfree(ctx)
and use following generic LSM equivalents respectively:
security_inode_getsecid(inode, secid)
security_ipc_getsecid*(ipcp, secid)
security_task_getsecid(tsk, secid)
security_sid_to_secctx(sid, ctx, len)
security_release_secctx(ctx, len)
Call security_release_secctx only if security_secid_to_secctx
succeeded.
Signed-off-by: Casey Schaufler <casey@schaufler-ca.com>
Signed-off-by: Ahmed S. Darwish <darwish.07@gmail.com>
Acked-by: James Morris <jmorris@namei.org>
Reviewed-by: Paul Moore <paul.moore@hp.com>
2008-03-01 20:54:38 +01:00
|
|
|
security_release_secctx(ctx, len);
|
|
|
|
}
|
2008-01-08 00:14:19 +01:00
|
|
|
}
|
|
|
|
|
|
|
|
return rc;
|
|
|
|
}
|
|
|
|
|
2005-04-17 00:20:36 +02:00
|
|
|
static int audit_receive_msg(struct sk_buff *skb, struct nlmsghdr *nlh)
|
|
|
|
{
|
2006-04-03 15:08:13 +02:00
|
|
|
u32 uid, pid, seq, sid;
|
2005-04-17 00:20:36 +02:00
|
|
|
void *data;
|
|
|
|
struct audit_status *status_get, status_set;
|
|
|
|
int err;
|
2005-05-13 19:17:42 +02:00
|
|
|
struct audit_buffer *ab;
|
2005-04-17 00:20:36 +02:00
|
|
|
u16 msg_type = nlh->nlmsg_type;
|
2005-04-29 17:27:17 +02:00
|
|
|
uid_t loginuid; /* loginuid of sender */
|
2008-04-18 16:09:25 +02:00
|
|
|
u32 sessionid;
|
2006-05-25 16:19:47 +02:00
|
|
|
struct audit_sig_info *sig_data;
|
2008-01-08 00:14:19 +01:00
|
|
|
char *ctx = NULL;
|
2006-05-25 16:19:47 +02:00
|
|
|
u32 len;
|
2005-04-17 00:20:36 +02:00
|
|
|
|
2006-06-27 22:26:11 +02:00
|
|
|
err = audit_netlink_ok(skb, msg_type);
|
2005-04-17 00:20:36 +02:00
|
|
|
if (err)
|
|
|
|
return err;
|
|
|
|
|
2005-09-13 21:47:11 +02:00
|
|
|
/* As soon as there's any sign of userspace auditd,
|
|
|
|
* start kauditd to talk to it */
|
2005-05-19 11:56:58 +02:00
|
|
|
if (!kauditd_task)
|
|
|
|
kauditd_task = kthread_run(kauditd_thread, NULL, "kauditd");
|
|
|
|
if (IS_ERR(kauditd_task)) {
|
|
|
|
err = PTR_ERR(kauditd_task);
|
|
|
|
kauditd_task = NULL;
|
|
|
|
return err;
|
|
|
|
}
|
|
|
|
|
2005-04-17 00:20:36 +02:00
|
|
|
pid = NETLINK_CREDS(skb)->pid;
|
|
|
|
uid = NETLINK_CREDS(skb)->uid;
|
2011-03-03 19:55:40 +01:00
|
|
|
loginuid = audit_get_loginuid(current);
|
|
|
|
sessionid = audit_get_sessionid(current);
|
|
|
|
security_task_getsecid(current, &sid);
|
2005-04-17 00:20:36 +02:00
|
|
|
seq = nlh->nlmsg_seq;
|
|
|
|
data = NLMSG_DATA(nlh);
|
|
|
|
|
|
|
|
switch (msg_type) {
|
|
|
|
case AUDIT_GET:
|
|
|
|
status_set.enabled = audit_enabled;
|
|
|
|
status_set.failure = audit_failure;
|
|
|
|
status_set.pid = audit_pid;
|
|
|
|
status_set.rate_limit = audit_rate_limit;
|
|
|
|
status_set.backlog_limit = audit_backlog_limit;
|
|
|
|
status_set.lost = atomic_read(&audit_lost);
|
2005-05-19 11:56:58 +02:00
|
|
|
status_set.backlog = skb_queue_len(&audit_skb_queue);
|
2005-04-17 00:20:36 +02:00
|
|
|
audit_send_reply(NETLINK_CB(skb).pid, seq, AUDIT_GET, 0, 0,
|
|
|
|
&status_set, sizeof(status_set));
|
|
|
|
break;
|
|
|
|
case AUDIT_SET:
|
|
|
|
if (nlh->nlmsg_len < sizeof(struct audit_status))
|
|
|
|
return -EINVAL;
|
|
|
|
status_get = (struct audit_status *)data;
|
|
|
|
if (status_get->mask & AUDIT_STATUS_ENABLED) {
|
2006-04-02 01:29:34 +02:00
|
|
|
err = audit_set_enabled(status_get->enabled,
|
2008-04-18 16:09:25 +02:00
|
|
|
loginuid, sessionid, sid);
|
2008-07-31 04:11:19 +02:00
|
|
|
if (err < 0)
|
|
|
|
return err;
|
2005-04-17 00:20:36 +02:00
|
|
|
}
|
|
|
|
if (status_get->mask & AUDIT_STATUS_FAILURE) {
|
2006-04-02 01:29:34 +02:00
|
|
|
err = audit_set_failure(status_get->failure,
|
2008-04-18 16:09:25 +02:00
|
|
|
loginuid, sessionid, sid);
|
2008-07-31 04:11:19 +02:00
|
|
|
if (err < 0)
|
|
|
|
return err;
|
2005-04-17 00:20:36 +02:00
|
|
|
}
|
|
|
|
if (status_get->mask & AUDIT_STATUS_PID) {
|
2008-01-07 23:09:31 +01:00
|
|
|
int new_pid = status_get->pid;
|
|
|
|
|
|
|
|
if (audit_enabled != AUDIT_OFF)
|
|
|
|
audit_log_config_change("audit_pid", new_pid,
|
|
|
|
audit_pid, loginuid,
|
2008-04-18 16:09:25 +02:00
|
|
|
sessionid, sid, 1);
|
2008-01-07 23:09:31 +01:00
|
|
|
|
|
|
|
audit_pid = new_pid;
|
audit: netlink socket can be auto-bound to pid other than current->pid (v2)
From: Pavel Emelyanov <xemul@openvz.org>
This patch is based on the one from Thomas.
The kauditd_thread() calls the netlink_unicast() and passes
the audit_pid to it. The audit_pid, in turn, is received from
the user space and the tool (I've checked the audit v1.6.9)
uses getpid() to pass one in the kernel. Besides, this tool
doesn't bind the netlink socket to this id, but simply creates
it allowing the kernel to auto-bind one.
That's the preamble.
The problem is that netlink_autobind() _does_not_ guarantees
that the socket will be auto-bound to the current pid. Instead
it uses the current pid as a hint to start looking for a free
id. So, in case of conflict, the audit messages can be sent
to a wrong socket. This can happen (it's unlikely, but can be)
in case some task opens more than one netlink sockets and then
the audit one starts - in this case the audit's pid can be busy
and its socket will be bound to another id.
The proposal is to introduce an audit_nlk_pid in audit subsys,
that will point to the netlink socket to send packets to. It
will most often be equal to audit_pid. The socket id can be
got from the skb's netlink CB right in the audit_receive_msg.
The audit_nlk_pid reset to 0 is not required, since all the
decisions are taken based on audit_pid value only.
Later, if the audit tools will bind the socket themselves, the
kernel will have to provide a way to setup the audit_nlk_pid
as well.
A good side effect of this patch is that audit_pid can later
be converted to struct pid, as it is not longer safe to use
pid_t-s in the presence of pid namespaces. But audit code still
uses the tgid from task_struct in the audit_signal_info and in
the audit_filter_syscall.
Signed-off-by: Thomas Graf <tgraf@suug.ch>
Signed-off-by: Pavel Emelyanov <xemul@openvz.org>
Acked-by: Eric Paris <eparis@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2008-03-20 23:39:41 +01:00
|
|
|
audit_nlk_pid = NETLINK_CB(skb).pid;
|
2005-04-17 00:20:36 +02:00
|
|
|
}
|
2008-07-31 04:11:19 +02:00
|
|
|
if (status_get->mask & AUDIT_STATUS_RATE_LIMIT) {
|
2006-04-27 23:45:14 +02:00
|
|
|
err = audit_set_rate_limit(status_get->rate_limit,
|
2008-04-18 16:09:25 +02:00
|
|
|
loginuid, sessionid, sid);
|
2008-07-31 04:11:19 +02:00
|
|
|
if (err < 0)
|
|
|
|
return err;
|
|
|
|
}
|
2005-04-17 00:20:36 +02:00
|
|
|
if (status_get->mask & AUDIT_STATUS_BACKLOG_LIMIT)
|
2006-04-27 23:45:14 +02:00
|
|
|
err = audit_set_backlog_limit(status_get->backlog_limit,
|
2008-04-18 16:09:25 +02:00
|
|
|
loginuid, sessionid, sid);
|
2005-04-17 00:20:36 +02:00
|
|
|
break;
|
2005-05-21 01:18:37 +02:00
|
|
|
case AUDIT_USER:
|
2007-05-08 09:29:20 +02:00
|
|
|
case AUDIT_FIRST_USER_MSG ... AUDIT_LAST_USER_MSG:
|
|
|
|
case AUDIT_FIRST_USER_MSG2 ... AUDIT_LAST_USER_MSG2:
|
2005-06-22 15:56:47 +02:00
|
|
|
if (!audit_enabled && msg_type != AUDIT_USER_AVC)
|
|
|
|
return 0;
|
|
|
|
|
2008-05-20 03:13:02 +02:00
|
|
|
err = audit_filter_user(&NETLINK_CB(skb));
|
2005-06-22 15:56:47 +02:00
|
|
|
if (err == 1) {
|
|
|
|
err = 0;
|
Audit: add TTY input auditing
Add TTY input auditing, used to audit system administrator's actions. This is
required by various security standards such as DCID 6/3 and PCI to provide
non-repudiation of administrator's actions and to allow a review of past
actions if the administrator seems to overstep their duties or if the system
becomes misconfigured for unknown reasons. These requirements do not make it
necessary to audit TTY output as well.
Compared to an user-space keylogger, this approach records TTY input using the
audit subsystem, correlated with other audit events, and it is completely
transparent to the user-space application (e.g. the console ioctls still
work).
TTY input auditing works on a higher level than auditing all system calls
within the session, which would produce an overwhelming amount of mostly
useless audit events.
Add an "audit_tty" attribute, inherited across fork (). Data read from TTYs
by process with the attribute is sent to the audit subsystem by the kernel.
The audit netlink interface is extended to allow modifying the audit_tty
attribute, and to allow sending explanatory audit events from user-space (for
example, a shell might send an event containing the final command, after the
interactive command-line editing and history expansion is performed, which
might be difficult to decipher from the TTY input alone).
Because the "audit_tty" attribute is inherited across fork (), it would be set
e.g. for sshd restarted within an audited session. To prevent this, the
audit_tty attribute is cleared when a process with no open TTY file
descriptors (e.g. after daemon startup) opens a TTY.
See https://www.redhat.com/archives/linux-audit/2007-June/msg00000.html for a
more detailed rationale document for an older version of this patch.
[akpm@linux-foundation.org: build fix]
Signed-off-by: Miloslav Trmac <mitr@redhat.com>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Cc: Alan Cox <alan@lxorguk.ukuu.org.uk>
Cc: Paul Fulghum <paulkf@microgate.com>
Cc: Casey Schaufler <casey@schaufler-ca.com>
Cc: Steve Grubb <sgrubb@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-07-16 08:40:56 +02:00
|
|
|
if (msg_type == AUDIT_USER_TTY) {
|
2008-04-18 16:09:25 +02:00
|
|
|
err = audit_prepare_user_tty(pid, loginuid,
|
|
|
|
sessionid);
|
Audit: add TTY input auditing
Add TTY input auditing, used to audit system administrator's actions. This is
required by various security standards such as DCID 6/3 and PCI to provide
non-repudiation of administrator's actions and to allow a review of past
actions if the administrator seems to overstep their duties or if the system
becomes misconfigured for unknown reasons. These requirements do not make it
necessary to audit TTY output as well.
Compared to an user-space keylogger, this approach records TTY input using the
audit subsystem, correlated with other audit events, and it is completely
transparent to the user-space application (e.g. the console ioctls still
work).
TTY input auditing works on a higher level than auditing all system calls
within the session, which would produce an overwhelming amount of mostly
useless audit events.
Add an "audit_tty" attribute, inherited across fork (). Data read from TTYs
by process with the attribute is sent to the audit subsystem by the kernel.
The audit netlink interface is extended to allow modifying the audit_tty
attribute, and to allow sending explanatory audit events from user-space (for
example, a shell might send an event containing the final command, after the
interactive command-line editing and history expansion is performed, which
might be difficult to decipher from the TTY input alone).
Because the "audit_tty" attribute is inherited across fork (), it would be set
e.g. for sshd restarted within an audited session. To prevent this, the
audit_tty attribute is cleared when a process with no open TTY file
descriptors (e.g. after daemon startup) opens a TTY.
See https://www.redhat.com/archives/linux-audit/2007-June/msg00000.html for a
more detailed rationale document for an older version of this patch.
[akpm@linux-foundation.org: build fix]
Signed-off-by: Miloslav Trmac <mitr@redhat.com>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Cc: Alan Cox <alan@lxorguk.ukuu.org.uk>
Cc: Paul Fulghum <paulkf@microgate.com>
Cc: Casey Schaufler <casey@schaufler-ca.com>
Cc: Steve Grubb <sgrubb@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-07-16 08:40:56 +02:00
|
|
|
if (err)
|
|
|
|
break;
|
|
|
|
}
|
2008-01-08 00:14:19 +01:00
|
|
|
audit_log_common_recv_msg(&ab, msg_type, pid, uid,
|
2008-04-18 16:09:25 +02:00
|
|
|
loginuid, sessionid, sid);
|
2008-01-08 00:14:19 +01:00
|
|
|
|
|
|
|
if (msg_type != AUDIT_USER_TTY)
|
|
|
|
audit_log_format(ab, " msg='%.1024s'",
|
|
|
|
(char *)data);
|
|
|
|
else {
|
|
|
|
int size;
|
|
|
|
|
|
|
|
audit_log_format(ab, " msg=");
|
|
|
|
size = nlmsg_len(nlh);
|
2009-03-19 14:52:47 +01:00
|
|
|
if (size > 0 &&
|
|
|
|
((unsigned char *)data)[size - 1] == '\0')
|
|
|
|
size--;
|
2008-04-18 16:12:59 +02:00
|
|
|
audit_log_n_untrustedstring(ab, data, size);
|
2005-06-22 15:56:47 +02:00
|
|
|
}
|
2008-01-08 00:14:19 +01:00
|
|
|
audit_set_pid(ab, pid);
|
|
|
|
audit_log_end(ab);
|
2005-06-19 20:35:50 +02:00
|
|
|
}
|
2005-04-17 00:20:36 +02:00
|
|
|
break;
|
|
|
|
case AUDIT_ADD:
|
|
|
|
case AUDIT_DEL:
|
2006-02-07 18:05:27 +01:00
|
|
|
if (nlmsg_len(nlh) < sizeof(struct audit_rule))
|
2005-04-17 00:20:36 +02:00
|
|
|
return -EINVAL;
|
2008-01-07 23:09:31 +01:00
|
|
|
if (audit_enabled == AUDIT_LOCKED) {
|
2008-01-08 00:14:19 +01:00
|
|
|
audit_log_common_recv_msg(&ab, AUDIT_CONFIG_CHANGE, pid,
|
2008-04-18 16:09:25 +02:00
|
|
|
uid, loginuid, sessionid, sid);
|
2008-01-08 00:14:19 +01:00
|
|
|
|
|
|
|
audit_log_format(ab, " audit_enabled=%d res=0",
|
|
|
|
audit_enabled);
|
|
|
|
audit_log_end(ab);
|
2007-01-19 20:39:55 +01:00
|
|
|
return -EPERM;
|
|
|
|
}
|
2005-04-17 00:20:36 +02:00
|
|
|
/* fallthrough */
|
|
|
|
case AUDIT_LIST:
|
2008-05-16 04:15:04 +02:00
|
|
|
err = audit_receive_filter(msg_type, NETLINK_CB(skb).pid,
|
2006-02-07 18:05:27 +01:00
|
|
|
uid, seq, data, nlmsg_len(nlh),
|
2008-04-18 16:09:25 +02:00
|
|
|
loginuid, sessionid, sid);
|
2006-02-07 18:05:27 +01:00
|
|
|
break;
|
|
|
|
case AUDIT_ADD_RULE:
|
|
|
|
case AUDIT_DEL_RULE:
|
|
|
|
if (nlmsg_len(nlh) < sizeof(struct audit_rule_data))
|
|
|
|
return -EINVAL;
|
2008-01-07 23:09:31 +01:00
|
|
|
if (audit_enabled == AUDIT_LOCKED) {
|
2008-01-08 00:14:19 +01:00
|
|
|
audit_log_common_recv_msg(&ab, AUDIT_CONFIG_CHANGE, pid,
|
2008-04-18 16:09:25 +02:00
|
|
|
uid, loginuid, sessionid, sid);
|
2008-01-08 00:14:19 +01:00
|
|
|
|
|
|
|
audit_log_format(ab, " audit_enabled=%d res=0",
|
|
|
|
audit_enabled);
|
|
|
|
audit_log_end(ab);
|
2007-01-19 20:39:55 +01:00
|
|
|
return -EPERM;
|
|
|
|
}
|
2006-02-07 18:05:27 +01:00
|
|
|
/* fallthrough */
|
|
|
|
case AUDIT_LIST_RULES:
|
2008-05-16 04:15:04 +02:00
|
|
|
err = audit_receive_filter(msg_type, NETLINK_CB(skb).pid,
|
2006-02-07 18:05:27 +01:00
|
|
|
uid, seq, data, nlmsg_len(nlh),
|
2008-04-18 16:09:25 +02:00
|
|
|
loginuid, sessionid, sid);
|
2005-04-17 00:20:36 +02:00
|
|
|
break;
|
[PATCH] audit: watching subtrees
New kind of audit rule predicates: "object is visible in given subtree".
The part that can be sanely implemented, that is. Limitations:
* if you have hardlink from outside of tree, you'd better watch
it too (or just watch the object itself, obviously)
* if you mount something under a watched tree, tell audit
that new chunk should be added to watched subtrees
* if you umount something in a watched tree and it's still mounted
elsewhere, you will get matches on events happening there. New command
tells audit to recalculate the trees, trimming such sources of false
positives.
Note that it's _not_ about path - if something mounted in several places
(multiple mount, bindings, different namespaces, etc.), the match does
_not_ depend on which one we are using for access.
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
2007-07-22 14:04:18 +02:00
|
|
|
case AUDIT_TRIM:
|
|
|
|
audit_trim_trees();
|
2008-01-08 00:14:19 +01:00
|
|
|
|
|
|
|
audit_log_common_recv_msg(&ab, AUDIT_CONFIG_CHANGE, pid,
|
2008-04-18 16:09:25 +02:00
|
|
|
uid, loginuid, sessionid, sid);
|
2008-01-08 00:14:19 +01:00
|
|
|
|
[PATCH] audit: watching subtrees
New kind of audit rule predicates: "object is visible in given subtree".
The part that can be sanely implemented, that is. Limitations:
* if you have hardlink from outside of tree, you'd better watch
it too (or just watch the object itself, obviously)
* if you mount something under a watched tree, tell audit
that new chunk should be added to watched subtrees
* if you umount something in a watched tree and it's still mounted
elsewhere, you will get matches on events happening there. New command
tells audit to recalculate the trees, trimming such sources of false
positives.
Note that it's _not_ about path - if something mounted in several places
(multiple mount, bindings, different namespaces, etc.), the match does
_not_ depend on which one we are using for access.
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
2007-07-22 14:04:18 +02:00
|
|
|
audit_log_format(ab, " op=trim res=1");
|
|
|
|
audit_log_end(ab);
|
|
|
|
break;
|
|
|
|
case AUDIT_MAKE_EQUIV: {
|
|
|
|
void *bufp = data;
|
|
|
|
u32 sizes[2];
|
2008-04-27 11:39:56 +02:00
|
|
|
size_t msglen = nlmsg_len(nlh);
|
[PATCH] audit: watching subtrees
New kind of audit rule predicates: "object is visible in given subtree".
The part that can be sanely implemented, that is. Limitations:
* if you have hardlink from outside of tree, you'd better watch
it too (or just watch the object itself, obviously)
* if you mount something under a watched tree, tell audit
that new chunk should be added to watched subtrees
* if you umount something in a watched tree and it's still mounted
elsewhere, you will get matches on events happening there. New command
tells audit to recalculate the trees, trimming such sources of false
positives.
Note that it's _not_ about path - if something mounted in several places
(multiple mount, bindings, different namespaces, etc.), the match does
_not_ depend on which one we are using for access.
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
2007-07-22 14:04:18 +02:00
|
|
|
char *old, *new;
|
|
|
|
|
|
|
|
err = -EINVAL;
|
2008-04-27 11:39:56 +02:00
|
|
|
if (msglen < 2 * sizeof(u32))
|
[PATCH] audit: watching subtrees
New kind of audit rule predicates: "object is visible in given subtree".
The part that can be sanely implemented, that is. Limitations:
* if you have hardlink from outside of tree, you'd better watch
it too (or just watch the object itself, obviously)
* if you mount something under a watched tree, tell audit
that new chunk should be added to watched subtrees
* if you umount something in a watched tree and it's still mounted
elsewhere, you will get matches on events happening there. New command
tells audit to recalculate the trees, trimming such sources of false
positives.
Note that it's _not_ about path - if something mounted in several places
(multiple mount, bindings, different namespaces, etc.), the match does
_not_ depend on which one we are using for access.
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
2007-07-22 14:04:18 +02:00
|
|
|
break;
|
|
|
|
memcpy(sizes, bufp, 2 * sizeof(u32));
|
|
|
|
bufp += 2 * sizeof(u32);
|
2008-04-27 11:39:56 +02:00
|
|
|
msglen -= 2 * sizeof(u32);
|
|
|
|
old = audit_unpack_string(&bufp, &msglen, sizes[0]);
|
[PATCH] audit: watching subtrees
New kind of audit rule predicates: "object is visible in given subtree".
The part that can be sanely implemented, that is. Limitations:
* if you have hardlink from outside of tree, you'd better watch
it too (or just watch the object itself, obviously)
* if you mount something under a watched tree, tell audit
that new chunk should be added to watched subtrees
* if you umount something in a watched tree and it's still mounted
elsewhere, you will get matches on events happening there. New command
tells audit to recalculate the trees, trimming such sources of false
positives.
Note that it's _not_ about path - if something mounted in several places
(multiple mount, bindings, different namespaces, etc.), the match does
_not_ depend on which one we are using for access.
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
2007-07-22 14:04:18 +02:00
|
|
|
if (IS_ERR(old)) {
|
|
|
|
err = PTR_ERR(old);
|
|
|
|
break;
|
|
|
|
}
|
2008-04-27 11:39:56 +02:00
|
|
|
new = audit_unpack_string(&bufp, &msglen, sizes[1]);
|
[PATCH] audit: watching subtrees
New kind of audit rule predicates: "object is visible in given subtree".
The part that can be sanely implemented, that is. Limitations:
* if you have hardlink from outside of tree, you'd better watch
it too (or just watch the object itself, obviously)
* if you mount something under a watched tree, tell audit
that new chunk should be added to watched subtrees
* if you umount something in a watched tree and it's still mounted
elsewhere, you will get matches on events happening there. New command
tells audit to recalculate the trees, trimming such sources of false
positives.
Note that it's _not_ about path - if something mounted in several places
(multiple mount, bindings, different namespaces, etc.), the match does
_not_ depend on which one we are using for access.
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
2007-07-22 14:04:18 +02:00
|
|
|
if (IS_ERR(new)) {
|
|
|
|
err = PTR_ERR(new);
|
|
|
|
kfree(old);
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
/* OK, here comes... */
|
|
|
|
err = audit_tag_tree(old, new);
|
|
|
|
|
2008-01-08 00:14:19 +01:00
|
|
|
audit_log_common_recv_msg(&ab, AUDIT_CONFIG_CHANGE, pid,
|
2008-04-18 16:09:25 +02:00
|
|
|
uid, loginuid, sessionid, sid);
|
2008-01-08 00:14:19 +01:00
|
|
|
|
[PATCH] audit: watching subtrees
New kind of audit rule predicates: "object is visible in given subtree".
The part that can be sanely implemented, that is. Limitations:
* if you have hardlink from outside of tree, you'd better watch
it too (or just watch the object itself, obviously)
* if you mount something under a watched tree, tell audit
that new chunk should be added to watched subtrees
* if you umount something in a watched tree and it's still mounted
elsewhere, you will get matches on events happening there. New command
tells audit to recalculate the trees, trimming such sources of false
positives.
Note that it's _not_ about path - if something mounted in several places
(multiple mount, bindings, different namespaces, etc.), the match does
_not_ depend on which one we are using for access.
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
2007-07-22 14:04:18 +02:00
|
|
|
audit_log_format(ab, " op=make_equiv old=");
|
|
|
|
audit_log_untrustedstring(ab, old);
|
|
|
|
audit_log_format(ab, " new=");
|
|
|
|
audit_log_untrustedstring(ab, new);
|
|
|
|
audit_log_format(ab, " res=%d", !err);
|
|
|
|
audit_log_end(ab);
|
|
|
|
kfree(old);
|
|
|
|
kfree(new);
|
|
|
|
break;
|
|
|
|
}
|
2005-05-06 13:38:39 +02:00
|
|
|
case AUDIT_SIGNAL_INFO:
|
2009-09-23 19:46:00 +02:00
|
|
|
len = 0;
|
|
|
|
if (audit_sig_sid) {
|
|
|
|
err = security_secid_to_secctx(audit_sig_sid, &ctx, &len);
|
|
|
|
if (err)
|
|
|
|
return err;
|
|
|
|
}
|
2006-05-25 16:19:47 +02:00
|
|
|
sig_data = kmalloc(sizeof(*sig_data) + len, GFP_KERNEL);
|
|
|
|
if (!sig_data) {
|
2009-09-23 19:46:00 +02:00
|
|
|
if (audit_sig_sid)
|
|
|
|
security_release_secctx(ctx, len);
|
2006-05-25 16:19:47 +02:00
|
|
|
return -ENOMEM;
|
|
|
|
}
|
|
|
|
sig_data->uid = audit_sig_uid;
|
|
|
|
sig_data->pid = audit_sig_pid;
|
2009-09-23 19:46:00 +02:00
|
|
|
if (audit_sig_sid) {
|
|
|
|
memcpy(sig_data->ctx, ctx, len);
|
|
|
|
security_release_secctx(ctx, len);
|
|
|
|
}
|
2007-10-18 12:06:10 +02:00
|
|
|
audit_send_reply(NETLINK_CB(skb).pid, seq, AUDIT_SIGNAL_INFO,
|
2006-05-25 16:19:47 +02:00
|
|
|
0, 0, sig_data, sizeof(*sig_data) + len);
|
|
|
|
kfree(sig_data);
|
2005-05-06 13:38:39 +02:00
|
|
|
break;
|
Audit: add TTY input auditing
Add TTY input auditing, used to audit system administrator's actions. This is
required by various security standards such as DCID 6/3 and PCI to provide
non-repudiation of administrator's actions and to allow a review of past
actions if the administrator seems to overstep their duties or if the system
becomes misconfigured for unknown reasons. These requirements do not make it
necessary to audit TTY output as well.
Compared to an user-space keylogger, this approach records TTY input using the
audit subsystem, correlated with other audit events, and it is completely
transparent to the user-space application (e.g. the console ioctls still
work).
TTY input auditing works on a higher level than auditing all system calls
within the session, which would produce an overwhelming amount of mostly
useless audit events.
Add an "audit_tty" attribute, inherited across fork (). Data read from TTYs
by process with the attribute is sent to the audit subsystem by the kernel.
The audit netlink interface is extended to allow modifying the audit_tty
attribute, and to allow sending explanatory audit events from user-space (for
example, a shell might send an event containing the final command, after the
interactive command-line editing and history expansion is performed, which
might be difficult to decipher from the TTY input alone).
Because the "audit_tty" attribute is inherited across fork (), it would be set
e.g. for sshd restarted within an audited session. To prevent this, the
audit_tty attribute is cleared when a process with no open TTY file
descriptors (e.g. after daemon startup) opens a TTY.
See https://www.redhat.com/archives/linux-audit/2007-June/msg00000.html for a
more detailed rationale document for an older version of this patch.
[akpm@linux-foundation.org: build fix]
Signed-off-by: Miloslav Trmac <mitr@redhat.com>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Cc: Alan Cox <alan@lxorguk.ukuu.org.uk>
Cc: Paul Fulghum <paulkf@microgate.com>
Cc: Casey Schaufler <casey@schaufler-ca.com>
Cc: Steve Grubb <sgrubb@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-07-16 08:40:56 +02:00
|
|
|
case AUDIT_TTY_GET: {
|
|
|
|
struct audit_tty_status s;
|
|
|
|
struct task_struct *tsk;
|
2009-12-09 15:19:41 +01:00
|
|
|
unsigned long flags;
|
Audit: add TTY input auditing
Add TTY input auditing, used to audit system administrator's actions. This is
required by various security standards such as DCID 6/3 and PCI to provide
non-repudiation of administrator's actions and to allow a review of past
actions if the administrator seems to overstep their duties or if the system
becomes misconfigured for unknown reasons. These requirements do not make it
necessary to audit TTY output as well.
Compared to an user-space keylogger, this approach records TTY input using the
audit subsystem, correlated with other audit events, and it is completely
transparent to the user-space application (e.g. the console ioctls still
work).
TTY input auditing works on a higher level than auditing all system calls
within the session, which would produce an overwhelming amount of mostly
useless audit events.
Add an "audit_tty" attribute, inherited across fork (). Data read from TTYs
by process with the attribute is sent to the audit subsystem by the kernel.
The audit netlink interface is extended to allow modifying the audit_tty
attribute, and to allow sending explanatory audit events from user-space (for
example, a shell might send an event containing the final command, after the
interactive command-line editing and history expansion is performed, which
might be difficult to decipher from the TTY input alone).
Because the "audit_tty" attribute is inherited across fork (), it would be set
e.g. for sshd restarted within an audited session. To prevent this, the
audit_tty attribute is cleared when a process with no open TTY file
descriptors (e.g. after daemon startup) opens a TTY.
See https://www.redhat.com/archives/linux-audit/2007-June/msg00000.html for a
more detailed rationale document for an older version of this patch.
[akpm@linux-foundation.org: build fix]
Signed-off-by: Miloslav Trmac <mitr@redhat.com>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Cc: Alan Cox <alan@lxorguk.ukuu.org.uk>
Cc: Paul Fulghum <paulkf@microgate.com>
Cc: Casey Schaufler <casey@schaufler-ca.com>
Cc: Steve Grubb <sgrubb@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-07-16 08:40:56 +02:00
|
|
|
|
2009-12-09 15:19:41 +01:00
|
|
|
rcu_read_lock();
|
2008-04-18 22:30:15 +02:00
|
|
|
tsk = find_task_by_vpid(pid);
|
2009-12-09 15:19:41 +01:00
|
|
|
if (tsk && lock_task_sighand(tsk, &flags)) {
|
Audit: add TTY input auditing
Add TTY input auditing, used to audit system administrator's actions. This is
required by various security standards such as DCID 6/3 and PCI to provide
non-repudiation of administrator's actions and to allow a review of past
actions if the administrator seems to overstep their duties or if the system
becomes misconfigured for unknown reasons. These requirements do not make it
necessary to audit TTY output as well.
Compared to an user-space keylogger, this approach records TTY input using the
audit subsystem, correlated with other audit events, and it is completely
transparent to the user-space application (e.g. the console ioctls still
work).
TTY input auditing works on a higher level than auditing all system calls
within the session, which would produce an overwhelming amount of mostly
useless audit events.
Add an "audit_tty" attribute, inherited across fork (). Data read from TTYs
by process with the attribute is sent to the audit subsystem by the kernel.
The audit netlink interface is extended to allow modifying the audit_tty
attribute, and to allow sending explanatory audit events from user-space (for
example, a shell might send an event containing the final command, after the
interactive command-line editing and history expansion is performed, which
might be difficult to decipher from the TTY input alone).
Because the "audit_tty" attribute is inherited across fork (), it would be set
e.g. for sshd restarted within an audited session. To prevent this, the
audit_tty attribute is cleared when a process with no open TTY file
descriptors (e.g. after daemon startup) opens a TTY.
See https://www.redhat.com/archives/linux-audit/2007-June/msg00000.html for a
more detailed rationale document for an older version of this patch.
[akpm@linux-foundation.org: build fix]
Signed-off-by: Miloslav Trmac <mitr@redhat.com>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Cc: Alan Cox <alan@lxorguk.ukuu.org.uk>
Cc: Paul Fulghum <paulkf@microgate.com>
Cc: Casey Schaufler <casey@schaufler-ca.com>
Cc: Steve Grubb <sgrubb@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-07-16 08:40:56 +02:00
|
|
|
s.enabled = tsk->signal->audit_tty != 0;
|
2009-12-09 15:19:41 +01:00
|
|
|
unlock_task_sighand(tsk, &flags);
|
|
|
|
} else
|
|
|
|
err = -ESRCH;
|
|
|
|
rcu_read_unlock();
|
2009-12-09 15:19:35 +01:00
|
|
|
|
|
|
|
if (!err)
|
|
|
|
audit_send_reply(NETLINK_CB(skb).pid, seq,
|
|
|
|
AUDIT_TTY_GET, 0, 0, &s, sizeof(s));
|
Audit: add TTY input auditing
Add TTY input auditing, used to audit system administrator's actions. This is
required by various security standards such as DCID 6/3 and PCI to provide
non-repudiation of administrator's actions and to allow a review of past
actions if the administrator seems to overstep their duties or if the system
becomes misconfigured for unknown reasons. These requirements do not make it
necessary to audit TTY output as well.
Compared to an user-space keylogger, this approach records TTY input using the
audit subsystem, correlated with other audit events, and it is completely
transparent to the user-space application (e.g. the console ioctls still
work).
TTY input auditing works on a higher level than auditing all system calls
within the session, which would produce an overwhelming amount of mostly
useless audit events.
Add an "audit_tty" attribute, inherited across fork (). Data read from TTYs
by process with the attribute is sent to the audit subsystem by the kernel.
The audit netlink interface is extended to allow modifying the audit_tty
attribute, and to allow sending explanatory audit events from user-space (for
example, a shell might send an event containing the final command, after the
interactive command-line editing and history expansion is performed, which
might be difficult to decipher from the TTY input alone).
Because the "audit_tty" attribute is inherited across fork (), it would be set
e.g. for sshd restarted within an audited session. To prevent this, the
audit_tty attribute is cleared when a process with no open TTY file
descriptors (e.g. after daemon startup) opens a TTY.
See https://www.redhat.com/archives/linux-audit/2007-June/msg00000.html for a
more detailed rationale document for an older version of this patch.
[akpm@linux-foundation.org: build fix]
Signed-off-by: Miloslav Trmac <mitr@redhat.com>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Cc: Alan Cox <alan@lxorguk.ukuu.org.uk>
Cc: Paul Fulghum <paulkf@microgate.com>
Cc: Casey Schaufler <casey@schaufler-ca.com>
Cc: Steve Grubb <sgrubb@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-07-16 08:40:56 +02:00
|
|
|
break;
|
|
|
|
}
|
|
|
|
case AUDIT_TTY_SET: {
|
|
|
|
struct audit_tty_status *s;
|
|
|
|
struct task_struct *tsk;
|
2009-12-09 15:19:41 +01:00
|
|
|
unsigned long flags;
|
Audit: add TTY input auditing
Add TTY input auditing, used to audit system administrator's actions. This is
required by various security standards such as DCID 6/3 and PCI to provide
non-repudiation of administrator's actions and to allow a review of past
actions if the administrator seems to overstep their duties or if the system
becomes misconfigured for unknown reasons. These requirements do not make it
necessary to audit TTY output as well.
Compared to an user-space keylogger, this approach records TTY input using the
audit subsystem, correlated with other audit events, and it is completely
transparent to the user-space application (e.g. the console ioctls still
work).
TTY input auditing works on a higher level than auditing all system calls
within the session, which would produce an overwhelming amount of mostly
useless audit events.
Add an "audit_tty" attribute, inherited across fork (). Data read from TTYs
by process with the attribute is sent to the audit subsystem by the kernel.
The audit netlink interface is extended to allow modifying the audit_tty
attribute, and to allow sending explanatory audit events from user-space (for
example, a shell might send an event containing the final command, after the
interactive command-line editing and history expansion is performed, which
might be difficult to decipher from the TTY input alone).
Because the "audit_tty" attribute is inherited across fork (), it would be set
e.g. for sshd restarted within an audited session. To prevent this, the
audit_tty attribute is cleared when a process with no open TTY file
descriptors (e.g. after daemon startup) opens a TTY.
See https://www.redhat.com/archives/linux-audit/2007-June/msg00000.html for a
more detailed rationale document for an older version of this patch.
[akpm@linux-foundation.org: build fix]
Signed-off-by: Miloslav Trmac <mitr@redhat.com>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Cc: Alan Cox <alan@lxorguk.ukuu.org.uk>
Cc: Paul Fulghum <paulkf@microgate.com>
Cc: Casey Schaufler <casey@schaufler-ca.com>
Cc: Steve Grubb <sgrubb@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-07-16 08:40:56 +02:00
|
|
|
|
|
|
|
if (nlh->nlmsg_len < sizeof(struct audit_tty_status))
|
|
|
|
return -EINVAL;
|
|
|
|
s = data;
|
|
|
|
if (s->enabled != 0 && s->enabled != 1)
|
|
|
|
return -EINVAL;
|
2009-12-09 15:19:41 +01:00
|
|
|
rcu_read_lock();
|
2008-04-18 22:30:15 +02:00
|
|
|
tsk = find_task_by_vpid(pid);
|
2009-12-09 15:19:41 +01:00
|
|
|
if (tsk && lock_task_sighand(tsk, &flags)) {
|
Audit: add TTY input auditing
Add TTY input auditing, used to audit system administrator's actions. This is
required by various security standards such as DCID 6/3 and PCI to provide
non-repudiation of administrator's actions and to allow a review of past
actions if the administrator seems to overstep their duties or if the system
becomes misconfigured for unknown reasons. These requirements do not make it
necessary to audit TTY output as well.
Compared to an user-space keylogger, this approach records TTY input using the
audit subsystem, correlated with other audit events, and it is completely
transparent to the user-space application (e.g. the console ioctls still
work).
TTY input auditing works on a higher level than auditing all system calls
within the session, which would produce an overwhelming amount of mostly
useless audit events.
Add an "audit_tty" attribute, inherited across fork (). Data read from TTYs
by process with the attribute is sent to the audit subsystem by the kernel.
The audit netlink interface is extended to allow modifying the audit_tty
attribute, and to allow sending explanatory audit events from user-space (for
example, a shell might send an event containing the final command, after the
interactive command-line editing and history expansion is performed, which
might be difficult to decipher from the TTY input alone).
Because the "audit_tty" attribute is inherited across fork (), it would be set
e.g. for sshd restarted within an audited session. To prevent this, the
audit_tty attribute is cleared when a process with no open TTY file
descriptors (e.g. after daemon startup) opens a TTY.
See https://www.redhat.com/archives/linux-audit/2007-June/msg00000.html for a
more detailed rationale document for an older version of this patch.
[akpm@linux-foundation.org: build fix]
Signed-off-by: Miloslav Trmac <mitr@redhat.com>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Cc: Alan Cox <alan@lxorguk.ukuu.org.uk>
Cc: Paul Fulghum <paulkf@microgate.com>
Cc: Casey Schaufler <casey@schaufler-ca.com>
Cc: Steve Grubb <sgrubb@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-07-16 08:40:56 +02:00
|
|
|
tsk->signal->audit_tty = s->enabled != 0;
|
2009-12-09 15:19:41 +01:00
|
|
|
unlock_task_sighand(tsk, &flags);
|
|
|
|
} else
|
|
|
|
err = -ESRCH;
|
|
|
|
rcu_read_unlock();
|
Audit: add TTY input auditing
Add TTY input auditing, used to audit system administrator's actions. This is
required by various security standards such as DCID 6/3 and PCI to provide
non-repudiation of administrator's actions and to allow a review of past
actions if the administrator seems to overstep their duties or if the system
becomes misconfigured for unknown reasons. These requirements do not make it
necessary to audit TTY output as well.
Compared to an user-space keylogger, this approach records TTY input using the
audit subsystem, correlated with other audit events, and it is completely
transparent to the user-space application (e.g. the console ioctls still
work).
TTY input auditing works on a higher level than auditing all system calls
within the session, which would produce an overwhelming amount of mostly
useless audit events.
Add an "audit_tty" attribute, inherited across fork (). Data read from TTYs
by process with the attribute is sent to the audit subsystem by the kernel.
The audit netlink interface is extended to allow modifying the audit_tty
attribute, and to allow sending explanatory audit events from user-space (for
example, a shell might send an event containing the final command, after the
interactive command-line editing and history expansion is performed, which
might be difficult to decipher from the TTY input alone).
Because the "audit_tty" attribute is inherited across fork (), it would be set
e.g. for sshd restarted within an audited session. To prevent this, the
audit_tty attribute is cleared when a process with no open TTY file
descriptors (e.g. after daemon startup) opens a TTY.
See https://www.redhat.com/archives/linux-audit/2007-June/msg00000.html for a
more detailed rationale document for an older version of this patch.
[akpm@linux-foundation.org: build fix]
Signed-off-by: Miloslav Trmac <mitr@redhat.com>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Cc: Alan Cox <alan@lxorguk.ukuu.org.uk>
Cc: Paul Fulghum <paulkf@microgate.com>
Cc: Casey Schaufler <casey@schaufler-ca.com>
Cc: Steve Grubb <sgrubb@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-07-16 08:40:56 +02:00
|
|
|
break;
|
|
|
|
}
|
2005-04-17 00:20:36 +02:00
|
|
|
default:
|
|
|
|
err = -EINVAL;
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
|
|
|
|
return err < 0 ? err : 0;
|
|
|
|
}
|
|
|
|
|
2005-09-13 21:47:11 +02:00
|
|
|
/*
|
2009-06-11 20:31:35 +02:00
|
|
|
* Get message from skb. Each message is processed by audit_receive_msg.
|
|
|
|
* Malformed skbs with wrong length are discarded silently.
|
2005-09-13 21:47:11 +02:00
|
|
|
*/
|
2005-05-03 23:55:09 +02:00
|
|
|
static void audit_receive_skb(struct sk_buff *skb)
|
2005-04-17 00:20:36 +02:00
|
|
|
{
|
2009-06-11 20:31:35 +02:00
|
|
|
struct nlmsghdr *nlh;
|
|
|
|
/*
|
|
|
|
* len MUST be signed for NLMSG_NEXT to be able to dec it below 0
|
|
|
|
* if the nlmsg_len was not aligned
|
|
|
|
*/
|
|
|
|
int len;
|
|
|
|
int err;
|
|
|
|
|
|
|
|
nlh = nlmsg_hdr(skb);
|
|
|
|
len = skb->len;
|
|
|
|
|
|
|
|
while (NLMSG_OK(nlh, len)) {
|
|
|
|
err = audit_receive_msg(skb, nlh);
|
|
|
|
/* if err or if this message says it wants a response */
|
|
|
|
if (err || (nlh->nlmsg_flags & NLM_F_ACK))
|
2005-04-17 00:20:36 +02:00
|
|
|
netlink_ack(skb, nlh, err);
|
2009-06-11 20:31:35 +02:00
|
|
|
|
|
|
|
nlh = NLMSG_NEXT(nlh, len);
|
2005-04-17 00:20:36 +02:00
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
/* Receive messages from netlink socket. */
|
2007-10-11 06:15:29 +02:00
|
|
|
static void audit_receive(struct sk_buff *skb)
|
2005-04-17 00:20:36 +02:00
|
|
|
{
|
[PATCH] audit: path-based rules
In this implementation, audit registers inotify watches on the parent
directories of paths specified in audit rules. When audit's inotify
event handler is called, it updates any affected rules based on the
filesystem event. If the parent directory is renamed, removed, or its
filesystem is unmounted, audit removes all rules referencing that
inotify watch.
To keep things simple, this implementation limits location-based
auditing to the directory entries in an existing directory. Given
a path-based rule for /foo/bar/passwd, the following table applies:
passwd modified -- audit event logged
passwd replaced -- audit event logged, rules list updated
bar renamed -- rule removed
foo renamed -- untracked, meaning that the rule now applies to
the new location
Audit users typically want to have many rules referencing filesystem
objects, which can significantly impact filtering performance. This
patch also adds an inode-number-based rule hash to mitigate this
situation.
The patch is relative to the audit git tree:
http://kernel.org/git/?p=linux/kernel/git/viro/audit-current.git;a=summary
and uses the inotify kernel API:
http://lkml.org/lkml/2006/6/1/145
Signed-off-by: Amy Griffis <amy.griffis@hp.com>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
2006-04-07 22:55:56 +02:00
|
|
|
mutex_lock(&audit_cmd_mutex);
|
2007-10-11 06:15:29 +02:00
|
|
|
audit_receive_skb(skb);
|
[PATCH] audit: path-based rules
In this implementation, audit registers inotify watches on the parent
directories of paths specified in audit rules. When audit's inotify
event handler is called, it updates any affected rules based on the
filesystem event. If the parent directory is renamed, removed, or its
filesystem is unmounted, audit removes all rules referencing that
inotify watch.
To keep things simple, this implementation limits location-based
auditing to the directory entries in an existing directory. Given
a path-based rule for /foo/bar/passwd, the following table applies:
passwd modified -- audit event logged
passwd replaced -- audit event logged, rules list updated
bar renamed -- rule removed
foo renamed -- untracked, meaning that the rule now applies to
the new location
Audit users typically want to have many rules referencing filesystem
objects, which can significantly impact filtering performance. This
patch also adds an inode-number-based rule hash to mitigate this
situation.
The patch is relative to the audit git tree:
http://kernel.org/git/?p=linux/kernel/git/viro/audit-current.git;a=summary
and uses the inotify kernel API:
http://lkml.org/lkml/2006/6/1/145
Signed-off-by: Amy Griffis <amy.griffis@hp.com>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
2006-04-07 22:55:56 +02:00
|
|
|
mutex_unlock(&audit_cmd_mutex);
|
2005-04-17 00:20:36 +02:00
|
|
|
}
|
|
|
|
|
|
|
|
/* Initialize audit support at boot time. */
|
|
|
|
static int __init audit_init(void)
|
|
|
|
{
|
[PATCH] audit: path-based rules
In this implementation, audit registers inotify watches on the parent
directories of paths specified in audit rules. When audit's inotify
event handler is called, it updates any affected rules based on the
filesystem event. If the parent directory is renamed, removed, or its
filesystem is unmounted, audit removes all rules referencing that
inotify watch.
To keep things simple, this implementation limits location-based
auditing to the directory entries in an existing directory. Given
a path-based rule for /foo/bar/passwd, the following table applies:
passwd modified -- audit event logged
passwd replaced -- audit event logged, rules list updated
bar renamed -- rule removed
foo renamed -- untracked, meaning that the rule now applies to
the new location
Audit users typically want to have many rules referencing filesystem
objects, which can significantly impact filtering performance. This
patch also adds an inode-number-based rule hash to mitigate this
situation.
The patch is relative to the audit git tree:
http://kernel.org/git/?p=linux/kernel/git/viro/audit-current.git;a=summary
and uses the inotify kernel API:
http://lkml.org/lkml/2006/6/1/145
Signed-off-by: Amy Griffis <amy.griffis@hp.com>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
2006-04-07 22:55:56 +02:00
|
|
|
int i;
|
|
|
|
|
2008-11-05 18:47:09 +01:00
|
|
|
if (audit_initialized == AUDIT_DISABLED)
|
|
|
|
return 0;
|
|
|
|
|
2005-04-17 00:20:36 +02:00
|
|
|
printk(KERN_INFO "audit: initializing netlink socket (%s)\n",
|
|
|
|
audit_default ? "enabled" : "disabled");
|
2007-09-12 13:05:38 +02:00
|
|
|
audit_sock = netlink_kernel_create(&init_net, NETLINK_AUDIT, 0,
|
|
|
|
audit_receive, NULL, THIS_MODULE);
|
2005-04-17 00:20:36 +02:00
|
|
|
if (!audit_sock)
|
|
|
|
audit_panic("cannot initialize netlink socket");
|
2006-03-07 04:40:05 +01:00
|
|
|
else
|
|
|
|
audit_sock->sk_sndtimeo = MAX_SCHEDULE_TIMEOUT;
|
2005-04-17 00:20:36 +02:00
|
|
|
|
2005-05-19 11:56:58 +02:00
|
|
|
skb_queue_head_init(&audit_skb_queue);
|
2008-04-18 16:02:28 +02:00
|
|
|
skb_queue_head_init(&audit_skb_hold_queue);
|
2008-11-05 18:47:09 +01:00
|
|
|
audit_initialized = AUDIT_INITIALIZED;
|
2005-04-17 00:20:36 +02:00
|
|
|
audit_enabled = audit_default;
|
2008-01-08 23:38:31 +01:00
|
|
|
audit_ever_enabled |= !!audit_default;
|
2006-03-11 01:14:06 +01:00
|
|
|
|
2005-06-22 16:04:33 +02:00
|
|
|
audit_log(NULL, GFP_KERNEL, AUDIT_KERNEL, "initialized");
|
[PATCH] audit: path-based rules
In this implementation, audit registers inotify watches on the parent
directories of paths specified in audit rules. When audit's inotify
event handler is called, it updates any affected rules based on the
filesystem event. If the parent directory is renamed, removed, or its
filesystem is unmounted, audit removes all rules referencing that
inotify watch.
To keep things simple, this implementation limits location-based
auditing to the directory entries in an existing directory. Given
a path-based rule for /foo/bar/passwd, the following table applies:
passwd modified -- audit event logged
passwd replaced -- audit event logged, rules list updated
bar renamed -- rule removed
foo renamed -- untracked, meaning that the rule now applies to
the new location
Audit users typically want to have many rules referencing filesystem
objects, which can significantly impact filtering performance. This
patch also adds an inode-number-based rule hash to mitigate this
situation.
The patch is relative to the audit git tree:
http://kernel.org/git/?p=linux/kernel/git/viro/audit-current.git;a=summary
and uses the inotify kernel API:
http://lkml.org/lkml/2006/6/1/145
Signed-off-by: Amy Griffis <amy.griffis@hp.com>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
2006-04-07 22:55:56 +02:00
|
|
|
|
|
|
|
for (i = 0; i < AUDIT_INODE_BUCKETS; i++)
|
|
|
|
INIT_LIST_HEAD(&audit_inode_hash[i]);
|
|
|
|
|
2005-04-17 00:20:36 +02:00
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
__initcall(audit_init);
|
|
|
|
|
|
|
|
/* Process kernel command-line parameter at boot time. audit=0 or audit=1. */
|
|
|
|
static int __init audit_enable(char *str)
|
|
|
|
{
|
|
|
|
audit_default = !!simple_strtol(str, NULL, 0);
|
2008-11-05 18:47:09 +01:00
|
|
|
if (!audit_default)
|
|
|
|
audit_initialized = AUDIT_DISABLED;
|
|
|
|
|
|
|
|
printk(KERN_INFO "audit: %s", audit_default ? "enabled" : "disabled");
|
|
|
|
|
|
|
|
if (audit_initialized == AUDIT_INITIALIZED) {
|
2005-04-17 00:20:36 +02:00
|
|
|
audit_enabled = audit_default;
|
2008-01-08 23:38:31 +01:00
|
|
|
audit_ever_enabled |= !!audit_default;
|
2008-11-05 18:47:09 +01:00
|
|
|
} else if (audit_initialized == AUDIT_UNINITIALIZED) {
|
|
|
|
printk(" (after initialization)");
|
|
|
|
} else {
|
|
|
|
printk(" (until reboot)");
|
2008-01-08 23:38:31 +01:00
|
|
|
}
|
2008-11-05 18:47:09 +01:00
|
|
|
printk("\n");
|
|
|
|
|
2006-03-31 12:30:33 +02:00
|
|
|
return 1;
|
2005-04-17 00:20:36 +02:00
|
|
|
}
|
|
|
|
|
|
|
|
__setup("audit=", audit_enable);
|
|
|
|
|
2005-05-06 16:53:34 +02:00
|
|
|
static void audit_buffer_free(struct audit_buffer *ab)
|
|
|
|
{
|
|
|
|
unsigned long flags;
|
|
|
|
|
2005-05-06 16:54:17 +02:00
|
|
|
if (!ab)
|
|
|
|
return;
|
|
|
|
|
2005-05-06 16:54:53 +02:00
|
|
|
if (ab->skb)
|
|
|
|
kfree_skb(ab->skb);
|
2005-05-19 11:56:58 +02:00
|
|
|
|
2005-05-06 16:53:34 +02:00
|
|
|
spin_lock_irqsave(&audit_freelist_lock, flags);
|
2006-04-27 23:45:14 +02:00
|
|
|
if (audit_freelist_count > AUDIT_MAXFREE)
|
2005-05-06 16:53:34 +02:00
|
|
|
kfree(ab);
|
2006-04-27 23:45:14 +02:00
|
|
|
else {
|
|
|
|
audit_freelist_count++;
|
2005-05-06 16:53:34 +02:00
|
|
|
list_add(&ab->list, &audit_freelist);
|
2006-04-27 23:45:14 +02:00
|
|
|
}
|
2005-05-06 16:53:34 +02:00
|
|
|
spin_unlock_irqrestore(&audit_freelist_lock, flags);
|
|
|
|
}
|
|
|
|
|
2005-05-13 19:17:42 +02:00
|
|
|
static struct audit_buffer * audit_buffer_alloc(struct audit_context *ctx,
|
2005-10-07 08:46:04 +02:00
|
|
|
gfp_t gfp_mask, int type)
|
2005-05-06 16:53:34 +02:00
|
|
|
{
|
|
|
|
unsigned long flags;
|
|
|
|
struct audit_buffer *ab = NULL;
|
2005-05-13 19:17:42 +02:00
|
|
|
struct nlmsghdr *nlh;
|
2005-05-06 16:53:34 +02:00
|
|
|
|
|
|
|
spin_lock_irqsave(&audit_freelist_lock, flags);
|
|
|
|
if (!list_empty(&audit_freelist)) {
|
|
|
|
ab = list_entry(audit_freelist.next,
|
|
|
|
struct audit_buffer, list);
|
|
|
|
list_del(&ab->list);
|
|
|
|
--audit_freelist_count;
|
|
|
|
}
|
|
|
|
spin_unlock_irqrestore(&audit_freelist_lock, flags);
|
|
|
|
|
|
|
|
if (!ab) {
|
2005-05-06 16:59:57 +02:00
|
|
|
ab = kmalloc(sizeof(*ab), gfp_mask);
|
2005-05-06 16:53:34 +02:00
|
|
|
if (!ab)
|
2005-05-06 16:54:17 +02:00
|
|
|
goto err;
|
2005-05-06 16:53:34 +02:00
|
|
|
}
|
2005-05-06 16:54:17 +02:00
|
|
|
|
2005-05-19 11:56:58 +02:00
|
|
|
ab->ctx = ctx;
|
2005-06-22 16:04:33 +02:00
|
|
|
ab->gfp_mask = gfp_mask;
|
2009-06-11 20:31:35 +02:00
|
|
|
|
|
|
|
ab->skb = nlmsg_new(AUDIT_BUFSIZ, gfp_mask);
|
|
|
|
if (!ab->skb)
|
|
|
|
goto nlmsg_failure;
|
|
|
|
|
|
|
|
nlh = NLMSG_NEW(ab->skb, 0, 0, type, 0, 0);
|
|
|
|
|
2005-05-06 16:53:34 +02:00
|
|
|
return ab;
|
2009-06-11 20:31:35 +02:00
|
|
|
|
|
|
|
nlmsg_failure: /* Used by NLMSG_NEW */
|
|
|
|
kfree_skb(ab->skb);
|
|
|
|
ab->skb = NULL;
|
2005-05-06 16:54:17 +02:00
|
|
|
err:
|
|
|
|
audit_buffer_free(ab);
|
|
|
|
return NULL;
|
2005-05-06 16:53:34 +02:00
|
|
|
}
|
2005-04-17 00:20:36 +02:00
|
|
|
|
2005-09-13 21:47:11 +02:00
|
|
|
/**
|
|
|
|
* audit_serial - compute a serial number for the audit record
|
|
|
|
*
|
|
|
|
* Compute a serial number for the audit record. Audit records are
|
2005-05-21 22:08:09 +02:00
|
|
|
* written to user-space as soon as they are generated, so a complete
|
|
|
|
* audit record may be written in several pieces. The timestamp of the
|
|
|
|
* record and this serial number are used by the user-space tools to
|
|
|
|
* determine which pieces belong to the same audit record. The
|
|
|
|
* (timestamp,serial) tuple is unique for each syscall and is live from
|
|
|
|
* syscall entry to syscall exit.
|
|
|
|
*
|
|
|
|
* NOTE: Another possibility is to store the formatted records off the
|
|
|
|
* audit context (for those records that have a context), and emit them
|
|
|
|
* all at syscall exit. However, this could delay the reporting of
|
|
|
|
* significant errors until syscall exit (or never, if the system
|
2005-09-13 21:47:11 +02:00
|
|
|
* halts).
|
|
|
|
*/
|
2005-05-21 22:08:09 +02:00
|
|
|
unsigned int audit_serial(void)
|
|
|
|
{
|
2006-06-27 11:53:55 +02:00
|
|
|
static DEFINE_SPINLOCK(serial_lock);
|
2005-07-15 13:56:03 +02:00
|
|
|
static unsigned int serial = 0;
|
|
|
|
|
|
|
|
unsigned long flags;
|
|
|
|
unsigned int ret;
|
2005-05-21 22:08:09 +02:00
|
|
|
|
2005-07-15 13:56:03 +02:00
|
|
|
spin_lock_irqsave(&serial_lock, flags);
|
2005-05-21 22:08:09 +02:00
|
|
|
do {
|
2005-07-18 20:24:46 +02:00
|
|
|
ret = ++serial;
|
|
|
|
} while (unlikely(!ret));
|
2005-07-15 13:56:03 +02:00
|
|
|
spin_unlock_irqrestore(&serial_lock, flags);
|
2005-05-21 22:08:09 +02:00
|
|
|
|
2005-07-15 13:56:03 +02:00
|
|
|
return ret;
|
2005-05-21 22:08:09 +02:00
|
|
|
}
|
|
|
|
|
2007-10-18 12:06:10 +02:00
|
|
|
static inline void audit_get_stamp(struct audit_context *ctx,
|
2005-05-21 22:08:09 +02:00
|
|
|
struct timespec *t, unsigned int *serial)
|
|
|
|
{
|
2008-12-06 07:05:50 +01:00
|
|
|
if (!ctx || !auditsc_get_stamp(ctx, t, serial)) {
|
2005-05-21 22:08:09 +02:00
|
|
|
*t = CURRENT_TIME;
|
|
|
|
*serial = audit_serial();
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2005-04-17 00:20:36 +02:00
|
|
|
/* Obtain an audit buffer. This routine does locking to obtain the
|
|
|
|
* audit buffer, but then no locking is required for calls to
|
|
|
|
* audit_log_*format. If the tsk is a task that is currently in a
|
|
|
|
* syscall, then the syscall is marked as auditable and an audit record
|
|
|
|
* will be written at syscall exit. If there is no associated task, tsk
|
|
|
|
* should be NULL. */
|
2005-06-22 16:04:33 +02:00
|
|
|
|
2005-09-13 21:47:11 +02:00
|
|
|
/**
|
|
|
|
* audit_log_start - obtain an audit buffer
|
|
|
|
* @ctx: audit_context (may be NULL)
|
|
|
|
* @gfp_mask: type of allocation
|
|
|
|
* @type: audit message type
|
|
|
|
*
|
|
|
|
* Returns audit_buffer pointer on success or NULL on error.
|
|
|
|
*
|
|
|
|
* Obtain an audit buffer. This routine does locking to obtain the
|
|
|
|
* audit buffer, but then no locking is required for calls to
|
|
|
|
* audit_log_*format. If the task (ctx) is a task that is currently in a
|
|
|
|
* syscall, then the syscall is marked as auditable and an audit record
|
|
|
|
* will be written at syscall exit. If there is no associated task, then
|
|
|
|
* task context (ctx) should be NULL.
|
|
|
|
*/
|
2005-10-21 09:22:03 +02:00
|
|
|
struct audit_buffer *audit_log_start(struct audit_context *ctx, gfp_t gfp_mask,
|
2005-06-22 16:04:33 +02:00
|
|
|
int type)
|
2005-04-17 00:20:36 +02:00
|
|
|
{
|
|
|
|
struct audit_buffer *ab = NULL;
|
|
|
|
struct timespec t;
|
2008-01-10 20:02:39 +01:00
|
|
|
unsigned int uninitialized_var(serial);
|
2005-06-22 16:04:33 +02:00
|
|
|
int reserve;
|
2005-07-02 15:08:48 +02:00
|
|
|
unsigned long timeout_start = jiffies;
|
2005-04-17 00:20:36 +02:00
|
|
|
|
2008-11-05 18:47:09 +01:00
|
|
|
if (audit_initialized != AUDIT_INITIALIZED)
|
2005-04-17 00:20:36 +02:00
|
|
|
return NULL;
|
|
|
|
|
2005-11-03 17:12:36 +01:00
|
|
|
if (unlikely(audit_filter_type(type)))
|
|
|
|
return NULL;
|
|
|
|
|
2005-06-22 16:04:33 +02:00
|
|
|
if (gfp_mask & __GFP_WAIT)
|
|
|
|
reserve = 0;
|
|
|
|
else
|
2007-10-18 12:06:10 +02:00
|
|
|
reserve = 5; /* Allow atomic callers to go up to five
|
2005-06-22 16:04:33 +02:00
|
|
|
entries over the normal backlog limit */
|
|
|
|
|
|
|
|
while (audit_backlog_limit
|
|
|
|
&& skb_queue_len(&audit_skb_queue) > audit_backlog_limit + reserve) {
|
2005-07-02 15:08:48 +02:00
|
|
|
if (gfp_mask & __GFP_WAIT && audit_backlog_wait_time
|
|
|
|
&& time_before(jiffies, timeout_start + audit_backlog_wait_time)) {
|
|
|
|
|
2005-06-22 16:04:33 +02:00
|
|
|
/* Wait for auditd to drain the queue a little */
|
|
|
|
DECLARE_WAITQUEUE(wait, current);
|
|
|
|
set_current_state(TASK_INTERRUPTIBLE);
|
|
|
|
add_wait_queue(&audit_backlog_wait, &wait);
|
|
|
|
|
|
|
|
if (audit_backlog_limit &&
|
|
|
|
skb_queue_len(&audit_skb_queue) > audit_backlog_limit)
|
2005-07-02 15:08:48 +02:00
|
|
|
schedule_timeout(timeout_start + audit_backlog_wait_time - jiffies);
|
2005-06-22 16:04:33 +02:00
|
|
|
|
|
|
|
__set_current_state(TASK_RUNNING);
|
|
|
|
remove_wait_queue(&audit_backlog_wait, &wait);
|
2005-07-02 15:08:48 +02:00
|
|
|
continue;
|
2005-06-22 16:04:33 +02:00
|
|
|
}
|
2008-01-24 04:55:05 +01:00
|
|
|
if (audit_rate_check() && printk_ratelimit())
|
2005-05-19 15:55:56 +02:00
|
|
|
printk(KERN_WARNING
|
|
|
|
"audit: audit_backlog=%d > "
|
|
|
|
"audit_backlog_limit=%d\n",
|
|
|
|
skb_queue_len(&audit_skb_queue),
|
|
|
|
audit_backlog_limit);
|
|
|
|
audit_log_lost("backlog limit exceeded");
|
2005-07-02 15:08:48 +02:00
|
|
|
audit_backlog_wait_time = audit_backlog_wait_overflow;
|
|
|
|
wake_up(&audit_backlog_wait);
|
2005-05-19 15:55:56 +02:00
|
|
|
return NULL;
|
|
|
|
}
|
|
|
|
|
2005-06-22 16:04:33 +02:00
|
|
|
ab = audit_buffer_alloc(ctx, gfp_mask, type);
|
2005-04-17 00:20:36 +02:00
|
|
|
if (!ab) {
|
|
|
|
audit_log_lost("out of memory in audit_log_start");
|
|
|
|
return NULL;
|
|
|
|
}
|
|
|
|
|
2005-05-21 22:08:09 +02:00
|
|
|
audit_get_stamp(ab->ctx, &t, &serial);
|
2005-05-11 11:54:05 +02:00
|
|
|
|
2005-04-17 00:20:36 +02:00
|
|
|
audit_log_format(ab, "audit(%lu.%03lu:%u): ",
|
|
|
|
t.tv_sec, t.tv_nsec/1000000, serial);
|
|
|
|
return ab;
|
|
|
|
}
|
|
|
|
|
2005-05-06 16:54:17 +02:00
|
|
|
/**
|
2005-05-06 16:54:53 +02:00
|
|
|
* audit_expand - expand skb in the audit buffer
|
2005-05-06 16:54:17 +02:00
|
|
|
* @ab: audit_buffer
|
2005-09-13 21:47:11 +02:00
|
|
|
* @extra: space to add at tail of the skb
|
2005-05-06 16:54:17 +02:00
|
|
|
*
|
|
|
|
* Returns 0 (no space) on failed expansion, or available space if
|
|
|
|
* successful.
|
|
|
|
*/
|
2005-05-10 19:56:08 +02:00
|
|
|
static inline int audit_expand(struct audit_buffer *ab, int extra)
|
2005-05-06 16:54:17 +02:00
|
|
|
{
|
2005-05-06 16:54:53 +02:00
|
|
|
struct sk_buff *skb = ab->skb;
|
2008-01-29 05:47:09 +01:00
|
|
|
int oldtail = skb_tailroom(skb);
|
|
|
|
int ret = pskb_expand_head(skb, 0, extra, ab->gfp_mask);
|
|
|
|
int newtail = skb_tailroom(skb);
|
|
|
|
|
2005-05-06 16:54:53 +02:00
|
|
|
if (ret < 0) {
|
|
|
|
audit_log_lost("out of memory in audit_expand");
|
2005-05-06 16:54:17 +02:00
|
|
|
return 0;
|
2005-05-06 16:54:53 +02:00
|
|
|
}
|
2008-01-29 05:47:09 +01:00
|
|
|
|
|
|
|
skb->truesize += newtail - oldtail;
|
|
|
|
return newtail;
|
2005-05-06 16:54:17 +02:00
|
|
|
}
|
2005-04-17 00:20:36 +02:00
|
|
|
|
2005-09-13 21:47:11 +02:00
|
|
|
/*
|
|
|
|
* Format an audit message into the audit buffer. If there isn't enough
|
2005-04-17 00:20:36 +02:00
|
|
|
* room in the audit buffer, more room will be allocated and vsnprint
|
|
|
|
* will be called a second time. Currently, we assume that a printk
|
2005-09-13 21:47:11 +02:00
|
|
|
* can't format message larger than 1024 bytes, so we don't either.
|
|
|
|
*/
|
2005-04-17 00:20:36 +02:00
|
|
|
static void audit_log_vformat(struct audit_buffer *ab, const char *fmt,
|
|
|
|
va_list args)
|
|
|
|
{
|
|
|
|
int len, avail;
|
2005-05-06 16:54:53 +02:00
|
|
|
struct sk_buff *skb;
|
2005-05-10 19:58:51 +02:00
|
|
|
va_list args2;
|
2005-04-17 00:20:36 +02:00
|
|
|
|
|
|
|
if (!ab)
|
|
|
|
return;
|
|
|
|
|
2005-05-06 16:54:53 +02:00
|
|
|
BUG_ON(!ab->skb);
|
|
|
|
skb = ab->skb;
|
|
|
|
avail = skb_tailroom(skb);
|
|
|
|
if (avail == 0) {
|
2005-05-10 19:56:08 +02:00
|
|
|
avail = audit_expand(ab, AUDIT_BUFSIZ);
|
2005-05-06 16:54:17 +02:00
|
|
|
if (!avail)
|
|
|
|
goto out;
|
2005-04-17 00:20:36 +02:00
|
|
|
}
|
2005-05-10 19:58:51 +02:00
|
|
|
va_copy(args2, args);
|
2007-04-20 05:29:13 +02:00
|
|
|
len = vsnprintf(skb_tail_pointer(skb), avail, fmt, args);
|
2005-04-17 00:20:36 +02:00
|
|
|
if (len >= avail) {
|
|
|
|
/* The printk buffer is 1024 bytes long, so if we get
|
|
|
|
* here and AUDIT_BUFSIZ is at least 1024, then we can
|
|
|
|
* log everything that printk could have logged. */
|
2005-09-13 21:47:11 +02:00
|
|
|
avail = audit_expand(ab,
|
|
|
|
max_t(unsigned, AUDIT_BUFSIZ, 1+len-avail));
|
2005-05-06 16:54:17 +02:00
|
|
|
if (!avail)
|
|
|
|
goto out;
|
2007-04-20 05:29:13 +02:00
|
|
|
len = vsnprintf(skb_tail_pointer(skb), avail, fmt, args2);
|
2005-04-17 00:20:36 +02:00
|
|
|
}
|
2008-01-10 20:02:40 +01:00
|
|
|
va_end(args2);
|
2005-05-19 11:24:22 +02:00
|
|
|
if (len > 0)
|
|
|
|
skb_put(skb, len);
|
2005-05-06 16:54:17 +02:00
|
|
|
out:
|
|
|
|
return;
|
2005-04-17 00:20:36 +02:00
|
|
|
}
|
|
|
|
|
2005-09-13 21:47:11 +02:00
|
|
|
/**
|
|
|
|
* audit_log_format - format a message into the audit buffer.
|
|
|
|
* @ab: audit_buffer
|
|
|
|
* @fmt: format string
|
|
|
|
* @...: optional parameters matching @fmt string
|
|
|
|
*
|
|
|
|
* All the work is done in audit_log_vformat.
|
|
|
|
*/
|
2005-04-17 00:20:36 +02:00
|
|
|
void audit_log_format(struct audit_buffer *ab, const char *fmt, ...)
|
|
|
|
{
|
|
|
|
va_list args;
|
|
|
|
|
|
|
|
if (!ab)
|
|
|
|
return;
|
|
|
|
va_start(args, fmt);
|
|
|
|
audit_log_vformat(ab, fmt, args);
|
|
|
|
va_end(args);
|
|
|
|
}
|
|
|
|
|
2005-09-13 21:47:11 +02:00
|
|
|
/**
|
|
|
|
* audit_log_hex - convert a buffer to hex and append it to the audit skb
|
|
|
|
* @ab: the audit_buffer
|
|
|
|
* @buf: buffer to convert to hex
|
|
|
|
* @len: length of @buf to be converted
|
|
|
|
*
|
|
|
|
* No return value; failure to expand is silently ignored.
|
|
|
|
*
|
|
|
|
* This function will take the passed buf and convert it into a string of
|
|
|
|
* ascii hex digits. The new string is placed onto the skb.
|
|
|
|
*/
|
2008-04-18 16:12:59 +02:00
|
|
|
void audit_log_n_hex(struct audit_buffer *ab, const unsigned char *buf,
|
2005-05-19 11:24:22 +02:00
|
|
|
size_t len)
|
2005-04-29 16:54:44 +02:00
|
|
|
{
|
2005-05-19 11:24:22 +02:00
|
|
|
int i, avail, new_len;
|
|
|
|
unsigned char *ptr;
|
|
|
|
struct sk_buff *skb;
|
|
|
|
static const unsigned char *hex = "0123456789ABCDEF";
|
|
|
|
|
2006-09-07 23:03:02 +02:00
|
|
|
if (!ab)
|
|
|
|
return;
|
|
|
|
|
2005-05-19 11:24:22 +02:00
|
|
|
BUG_ON(!ab->skb);
|
|
|
|
skb = ab->skb;
|
|
|
|
avail = skb_tailroom(skb);
|
|
|
|
new_len = len<<1;
|
|
|
|
if (new_len >= avail) {
|
|
|
|
/* Round the buffer request up to the next multiple */
|
|
|
|
new_len = AUDIT_BUFSIZ*(((new_len-avail)/AUDIT_BUFSIZ) + 1);
|
|
|
|
avail = audit_expand(ab, new_len);
|
|
|
|
if (!avail)
|
|
|
|
return;
|
|
|
|
}
|
2005-04-29 16:54:44 +02:00
|
|
|
|
2007-04-20 05:29:13 +02:00
|
|
|
ptr = skb_tail_pointer(skb);
|
2005-05-19 11:24:22 +02:00
|
|
|
for (i=0; i<len; i++) {
|
|
|
|
*ptr++ = hex[(buf[i] & 0xF0)>>4]; /* Upper nibble */
|
|
|
|
*ptr++ = hex[buf[i] & 0x0F]; /* Lower nibble */
|
|
|
|
}
|
|
|
|
*ptr = 0;
|
|
|
|
skb_put(skb, len << 1); /* new string is twice the old string */
|
2005-04-29 16:54:44 +02:00
|
|
|
}
|
|
|
|
|
2006-06-09 05:19:31 +02:00
|
|
|
/*
|
|
|
|
* Format a string of no more than slen characters into the audit buffer,
|
|
|
|
* enclosed in quote marks.
|
|
|
|
*/
|
2008-04-18 16:12:59 +02:00
|
|
|
void audit_log_n_string(struct audit_buffer *ab, const char *string,
|
|
|
|
size_t slen)
|
2006-06-09 05:19:31 +02:00
|
|
|
{
|
|
|
|
int avail, new_len;
|
|
|
|
unsigned char *ptr;
|
|
|
|
struct sk_buff *skb;
|
|
|
|
|
2006-09-07 23:03:02 +02:00
|
|
|
if (!ab)
|
|
|
|
return;
|
|
|
|
|
2006-06-09 05:19:31 +02:00
|
|
|
BUG_ON(!ab->skb);
|
|
|
|
skb = ab->skb;
|
|
|
|
avail = skb_tailroom(skb);
|
|
|
|
new_len = slen + 3; /* enclosing quotes + null terminator */
|
|
|
|
if (new_len > avail) {
|
|
|
|
avail = audit_expand(ab, new_len);
|
|
|
|
if (!avail)
|
|
|
|
return;
|
|
|
|
}
|
2007-04-20 05:29:13 +02:00
|
|
|
ptr = skb_tail_pointer(skb);
|
2006-06-09 05:19:31 +02:00
|
|
|
*ptr++ = '"';
|
|
|
|
memcpy(ptr, string, slen);
|
|
|
|
ptr += slen;
|
|
|
|
*ptr++ = '"';
|
|
|
|
*ptr = 0;
|
|
|
|
skb_put(skb, slen + 2); /* don't include null terminator */
|
|
|
|
}
|
|
|
|
|
2008-01-07 20:31:58 +01:00
|
|
|
/**
|
|
|
|
* audit_string_contains_control - does a string need to be logged in hex
|
2008-03-28 22:15:56 +01:00
|
|
|
* @string: string to be checked
|
|
|
|
* @len: max length of the string to check
|
2008-01-07 20:31:58 +01:00
|
|
|
*/
|
|
|
|
int audit_string_contains_control(const char *string, size_t len)
|
|
|
|
{
|
|
|
|
const unsigned char *p;
|
2009-03-19 14:48:27 +01:00
|
|
|
for (p = string; p < (const unsigned char *)string + len; p++) {
|
2008-07-22 23:06:13 +02:00
|
|
|
if (*p == '"' || *p < 0x21 || *p > 0x7e)
|
2008-01-07 20:31:58 +01:00
|
|
|
return 1;
|
|
|
|
}
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
2005-09-13 21:47:11 +02:00
|
|
|
/**
|
Audit: add TTY input auditing
Add TTY input auditing, used to audit system administrator's actions. This is
required by various security standards such as DCID 6/3 and PCI to provide
non-repudiation of administrator's actions and to allow a review of past
actions if the administrator seems to overstep their duties or if the system
becomes misconfigured for unknown reasons. These requirements do not make it
necessary to audit TTY output as well.
Compared to an user-space keylogger, this approach records TTY input using the
audit subsystem, correlated with other audit events, and it is completely
transparent to the user-space application (e.g. the console ioctls still
work).
TTY input auditing works on a higher level than auditing all system calls
within the session, which would produce an overwhelming amount of mostly
useless audit events.
Add an "audit_tty" attribute, inherited across fork (). Data read from TTYs
by process with the attribute is sent to the audit subsystem by the kernel.
The audit netlink interface is extended to allow modifying the audit_tty
attribute, and to allow sending explanatory audit events from user-space (for
example, a shell might send an event containing the final command, after the
interactive command-line editing and history expansion is performed, which
might be difficult to decipher from the TTY input alone).
Because the "audit_tty" attribute is inherited across fork (), it would be set
e.g. for sshd restarted within an audited session. To prevent this, the
audit_tty attribute is cleared when a process with no open TTY file
descriptors (e.g. after daemon startup) opens a TTY.
See https://www.redhat.com/archives/linux-audit/2007-June/msg00000.html for a
more detailed rationale document for an older version of this patch.
[akpm@linux-foundation.org: build fix]
Signed-off-by: Miloslav Trmac <mitr@redhat.com>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Cc: Alan Cox <alan@lxorguk.ukuu.org.uk>
Cc: Paul Fulghum <paulkf@microgate.com>
Cc: Casey Schaufler <casey@schaufler-ca.com>
Cc: Steve Grubb <sgrubb@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-07-16 08:40:56 +02:00
|
|
|
* audit_log_n_untrustedstring - log a string that may contain random characters
|
2005-09-13 21:47:11 +02:00
|
|
|
* @ab: audit_buffer
|
2008-03-28 22:15:56 +01:00
|
|
|
* @len: length of string (not including trailing null)
|
2005-09-13 21:47:11 +02:00
|
|
|
* @string: string to be logged
|
|
|
|
*
|
|
|
|
* This code will escape a string that is passed to it if the string
|
|
|
|
* contains a control character, unprintable character, double quote mark,
|
2005-05-19 11:24:22 +02:00
|
|
|
* or a space. Unescaped strings will start and end with a double quote mark.
|
2005-09-13 21:47:11 +02:00
|
|
|
* Strings that are escaped are printed in hex (2 digits per char).
|
2006-06-09 05:19:31 +02:00
|
|
|
*
|
|
|
|
* The caller specifies the number of characters in the string to log, which may
|
|
|
|
* or may not be the entire string.
|
2005-09-13 21:47:11 +02:00
|
|
|
*/
|
2008-04-18 16:12:59 +02:00
|
|
|
void audit_log_n_untrustedstring(struct audit_buffer *ab, const char *string,
|
|
|
|
size_t len)
|
2005-04-29 16:54:44 +02:00
|
|
|
{
|
2008-01-07 20:31:58 +01:00
|
|
|
if (audit_string_contains_control(string, len))
|
2008-04-18 16:12:59 +02:00
|
|
|
audit_log_n_hex(ab, string, len);
|
2008-01-07 20:31:58 +01:00
|
|
|
else
|
2008-04-18 16:12:59 +02:00
|
|
|
audit_log_n_string(ab, string, len);
|
2005-04-29 16:54:44 +02:00
|
|
|
}
|
|
|
|
|
2006-06-09 05:19:31 +02:00
|
|
|
/**
|
Audit: add TTY input auditing
Add TTY input auditing, used to audit system administrator's actions. This is
required by various security standards such as DCID 6/3 and PCI to provide
non-repudiation of administrator's actions and to allow a review of past
actions if the administrator seems to overstep their duties or if the system
becomes misconfigured for unknown reasons. These requirements do not make it
necessary to audit TTY output as well.
Compared to an user-space keylogger, this approach records TTY input using the
audit subsystem, correlated with other audit events, and it is completely
transparent to the user-space application (e.g. the console ioctls still
work).
TTY input auditing works on a higher level than auditing all system calls
within the session, which would produce an overwhelming amount of mostly
useless audit events.
Add an "audit_tty" attribute, inherited across fork (). Data read from TTYs
by process with the attribute is sent to the audit subsystem by the kernel.
The audit netlink interface is extended to allow modifying the audit_tty
attribute, and to allow sending explanatory audit events from user-space (for
example, a shell might send an event containing the final command, after the
interactive command-line editing and history expansion is performed, which
might be difficult to decipher from the TTY input alone).
Because the "audit_tty" attribute is inherited across fork (), it would be set
e.g. for sshd restarted within an audited session. To prevent this, the
audit_tty attribute is cleared when a process with no open TTY file
descriptors (e.g. after daemon startup) opens a TTY.
See https://www.redhat.com/archives/linux-audit/2007-June/msg00000.html for a
more detailed rationale document for an older version of this patch.
[akpm@linux-foundation.org: build fix]
Signed-off-by: Miloslav Trmac <mitr@redhat.com>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Cc: Alan Cox <alan@lxorguk.ukuu.org.uk>
Cc: Paul Fulghum <paulkf@microgate.com>
Cc: Casey Schaufler <casey@schaufler-ca.com>
Cc: Steve Grubb <sgrubb@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-07-16 08:40:56 +02:00
|
|
|
* audit_log_untrustedstring - log a string that may contain random characters
|
2006-06-09 05:19:31 +02:00
|
|
|
* @ab: audit_buffer
|
|
|
|
* @string: string to be logged
|
|
|
|
*
|
Audit: add TTY input auditing
Add TTY input auditing, used to audit system administrator's actions. This is
required by various security standards such as DCID 6/3 and PCI to provide
non-repudiation of administrator's actions and to allow a review of past
actions if the administrator seems to overstep their duties or if the system
becomes misconfigured for unknown reasons. These requirements do not make it
necessary to audit TTY output as well.
Compared to an user-space keylogger, this approach records TTY input using the
audit subsystem, correlated with other audit events, and it is completely
transparent to the user-space application (e.g. the console ioctls still
work).
TTY input auditing works on a higher level than auditing all system calls
within the session, which would produce an overwhelming amount of mostly
useless audit events.
Add an "audit_tty" attribute, inherited across fork (). Data read from TTYs
by process with the attribute is sent to the audit subsystem by the kernel.
The audit netlink interface is extended to allow modifying the audit_tty
attribute, and to allow sending explanatory audit events from user-space (for
example, a shell might send an event containing the final command, after the
interactive command-line editing and history expansion is performed, which
might be difficult to decipher from the TTY input alone).
Because the "audit_tty" attribute is inherited across fork (), it would be set
e.g. for sshd restarted within an audited session. To prevent this, the
audit_tty attribute is cleared when a process with no open TTY file
descriptors (e.g. after daemon startup) opens a TTY.
See https://www.redhat.com/archives/linux-audit/2007-June/msg00000.html for a
more detailed rationale document for an older version of this patch.
[akpm@linux-foundation.org: build fix]
Signed-off-by: Miloslav Trmac <mitr@redhat.com>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Cc: Alan Cox <alan@lxorguk.ukuu.org.uk>
Cc: Paul Fulghum <paulkf@microgate.com>
Cc: Casey Schaufler <casey@schaufler-ca.com>
Cc: Steve Grubb <sgrubb@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-07-16 08:40:56 +02:00
|
|
|
* Same as audit_log_n_untrustedstring(), except that strlen is used to
|
2006-06-09 05:19:31 +02:00
|
|
|
* determine string length.
|
|
|
|
*/
|
2008-01-07 20:31:58 +01:00
|
|
|
void audit_log_untrustedstring(struct audit_buffer *ab, const char *string)
|
2006-06-09 05:19:31 +02:00
|
|
|
{
|
2008-04-18 16:12:59 +02:00
|
|
|
audit_log_n_untrustedstring(ab, string, strlen(string));
|
2006-06-09 05:19:31 +02:00
|
|
|
}
|
|
|
|
|
2005-05-19 11:24:22 +02:00
|
|
|
/* This is a helper-function to print the escaped d_path */
|
2005-04-17 00:20:36 +02:00
|
|
|
void audit_log_d_path(struct audit_buffer *ab, const char *prefix,
|
2008-02-15 04:38:33 +01:00
|
|
|
struct path *path)
|
2005-04-17 00:20:36 +02:00
|
|
|
{
|
2008-02-15 04:38:33 +01:00
|
|
|
char *p, *pathname;
|
2005-04-17 00:20:36 +02:00
|
|
|
|
2005-05-06 16:54:17 +02:00
|
|
|
if (prefix)
|
|
|
|
audit_log_format(ab, " %s", prefix);
|
2005-04-17 00:20:36 +02:00
|
|
|
|
2005-05-19 11:24:22 +02:00
|
|
|
/* We will allow 11 spaces for ' (deleted)' to be appended */
|
2008-02-15 04:38:33 +01:00
|
|
|
pathname = kmalloc(PATH_MAX+11, ab->gfp_mask);
|
|
|
|
if (!pathname) {
|
2009-03-10 23:00:14 +01:00
|
|
|
audit_log_string(ab, "<no_memory>");
|
2005-05-19 11:24:22 +02:00
|
|
|
return;
|
2005-04-17 00:20:36 +02:00
|
|
|
}
|
2008-02-15 04:38:44 +01:00
|
|
|
p = d_path(path, pathname, PATH_MAX+11);
|
2005-05-19 11:24:22 +02:00
|
|
|
if (IS_ERR(p)) { /* Should never happen since we send PATH_MAX */
|
|
|
|
/* FIXME: can we save some information here? */
|
2009-03-10 23:00:14 +01:00
|
|
|
audit_log_string(ab, "<too_long>");
|
2007-10-18 12:06:10 +02:00
|
|
|
} else
|
2005-05-19 11:24:22 +02:00
|
|
|
audit_log_untrustedstring(ab, p);
|
2008-02-15 04:38:33 +01:00
|
|
|
kfree(pathname);
|
2005-04-17 00:20:36 +02:00
|
|
|
}
|
|
|
|
|
2009-06-11 20:31:37 +02:00
|
|
|
void audit_log_key(struct audit_buffer *ab, char *key)
|
|
|
|
{
|
|
|
|
audit_log_format(ab, " key=");
|
|
|
|
if (key)
|
|
|
|
audit_log_untrustedstring(ab, key);
|
|
|
|
else
|
|
|
|
audit_log_format(ab, "(null)");
|
|
|
|
}
|
|
|
|
|
2005-09-13 21:47:11 +02:00
|
|
|
/**
|
|
|
|
* audit_log_end - end one audit record
|
|
|
|
* @ab: the audit_buffer
|
|
|
|
*
|
|
|
|
* The netlink_* functions cannot be called inside an irq context, so
|
|
|
|
* the audit buffer is placed on a queue and a tasklet is scheduled to
|
2005-04-17 00:20:36 +02:00
|
|
|
* remove them from the queue outside the irq context. May be called in
|
2005-09-13 21:47:11 +02:00
|
|
|
* any context.
|
|
|
|
*/
|
2005-05-19 11:56:58 +02:00
|
|
|
void audit_log_end(struct audit_buffer *ab)
|
2005-04-17 00:20:36 +02:00
|
|
|
{
|
|
|
|
if (!ab)
|
|
|
|
return;
|
|
|
|
if (!audit_rate_check()) {
|
|
|
|
audit_log_lost("rate limit exceeded");
|
|
|
|
} else {
|
2008-02-21 22:59:22 +01:00
|
|
|
struct nlmsghdr *nlh = nlmsg_hdr(ab->skb);
|
2008-04-18 16:02:28 +02:00
|
|
|
nlh->nlmsg_len = ab->skb->len - NLMSG_SPACE(0);
|
|
|
|
|
2005-05-19 11:56:58 +02:00
|
|
|
if (audit_pid) {
|
|
|
|
skb_queue_tail(&audit_skb_queue, ab->skb);
|
|
|
|
wake_up_interruptible(&kauditd_wait);
|
2008-04-18 16:02:28 +02:00
|
|
|
} else {
|
2009-06-11 20:31:35 +02:00
|
|
|
audit_printk_skb(ab->skb);
|
2005-05-19 11:56:58 +02:00
|
|
|
}
|
2008-04-18 16:02:28 +02:00
|
|
|
ab->skb = NULL;
|
2005-04-17 00:20:36 +02:00
|
|
|
}
|
2005-05-06 16:53:34 +02:00
|
|
|
audit_buffer_free(ab);
|
2005-04-17 00:20:36 +02:00
|
|
|
}
|
|
|
|
|
2005-09-13 21:47:11 +02:00
|
|
|
/**
|
|
|
|
* audit_log - Log an audit record
|
|
|
|
* @ctx: audit context
|
|
|
|
* @gfp_mask: type of allocation
|
|
|
|
* @type: audit message type
|
|
|
|
* @fmt: format string to use
|
|
|
|
* @...: variable parameters matching the format string
|
|
|
|
*
|
|
|
|
* This is a convenience function that calls audit_log_start,
|
|
|
|
* audit_log_vformat, and audit_log_end. It may be called
|
|
|
|
* in any context.
|
|
|
|
*/
|
2007-10-18 12:06:10 +02:00
|
|
|
void audit_log(struct audit_context *ctx, gfp_t gfp_mask, int type,
|
2005-06-22 16:04:33 +02:00
|
|
|
const char *fmt, ...)
|
2005-04-17 00:20:36 +02:00
|
|
|
{
|
|
|
|
struct audit_buffer *ab;
|
|
|
|
va_list args;
|
|
|
|
|
2005-06-22 16:04:33 +02:00
|
|
|
ab = audit_log_start(ctx, gfp_mask, type);
|
2005-04-17 00:20:36 +02:00
|
|
|
if (ab) {
|
|
|
|
va_start(args, fmt);
|
|
|
|
audit_log_vformat(ab, fmt, args);
|
|
|
|
va_end(args);
|
|
|
|
audit_log_end(ab);
|
|
|
|
}
|
|
|
|
}
|
2006-03-09 00:33:47 +01:00
|
|
|
|
2011-06-30 13:31:57 +02:00
|
|
|
#ifdef CONFIG_SECURITY
|
|
|
|
/**
|
|
|
|
* audit_log_secctx - Converts and logs SELinux context
|
|
|
|
* @ab: audit_buffer
|
|
|
|
* @secid: security number
|
|
|
|
*
|
|
|
|
* This is a helper function that calls security_secid_to_secctx to convert
|
|
|
|
* secid to secctx and then adds the (converted) SELinux context to the audit
|
|
|
|
* log by calling audit_log_format, thus also preventing leak of internal secid
|
|
|
|
* to userspace. If secid cannot be converted audit_panic is called.
|
|
|
|
*/
|
|
|
|
void audit_log_secctx(struct audit_buffer *ab, u32 secid)
|
|
|
|
{
|
|
|
|
u32 len;
|
|
|
|
char *secctx;
|
|
|
|
|
|
|
|
if (security_secid_to_secctx(secid, &secctx, &len)) {
|
|
|
|
audit_panic("Cannot convert secid to context");
|
|
|
|
} else {
|
|
|
|
audit_log_format(ab, " obj=%s", secctx);
|
|
|
|
security_release_secctx(secctx, len);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
EXPORT_SYMBOL(audit_log_secctx);
|
|
|
|
#endif
|
|
|
|
|
2006-03-09 00:33:47 +01:00
|
|
|
EXPORT_SYMBOL(audit_log_start);
|
|
|
|
EXPORT_SYMBOL(audit_log_end);
|
|
|
|
EXPORT_SYMBOL(audit_log_format);
|
|
|
|
EXPORT_SYMBOL(audit_log);
|