OpenCloudOS-Kernel/kernel/audit.c

2467 lines
65 KiB
C
Raw Normal View History

// SPDX-License-Identifier: GPL-2.0-or-later
/* audit.c -- Auditing support
* Gateway between the kernel (e.g., selinux) and the user-space audit daemon.
* System-call specific features have moved to auditsc.c
*
* Copyright 2003-2007 Red Hat Inc., Durham, North Carolina.
* All Rights Reserved.
*
* Written by Rickard E. (Rik) Faith <faith@redhat.com>
*
* Goals: 1) Integrate fully with Security Modules.
* 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.
*
* Audit userspace, documentation, tests, and bug/issue trackers:
* https://github.com/linux-audit
*/
#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt
#include <linux/file.h>
#include <linux/init.h>
#include <linux/types.h>
#include <linux/atomic.h>
#include <linux/mm.h>
#include <linux/export.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 16:04:11 +08:00
#include <linux/slab.h>
#include <linux/err.h>
#include <linux/kthread.h>
#include <linux/kernel.h>
#include <linux/syscalls.h>
audit: fix auditd/kernel connection state tracking What started as a rather straightforward race condition reported by Dmitry using the syzkaller fuzzer ended up revealing some major problems with how the audit subsystem managed its netlink sockets and its connection with the userspace audit daemon. Fixing this properly had quite the cascading effect and what we are left with is this rather large and complicated patch. My initial goal was to try and decompose this patch into multiple smaller patches, but the way these changes are intertwined makes it difficult to split these changes into meaningful pieces that don't break or somehow make things worse for the intermediate states. The patch makes a number of changes, but the most significant are highlighted below: * The auditd tracking variables, e.g. audit_sock, are now gone and replaced by a RCU/spin_lock protected variable auditd_conn which is a structure containing all of the auditd tracking information. * We no longer track the auditd sock directly, instead we track it via the network namespace in which it resides and we use the audit socket associated with that namespace. In spirit, this is what the code was trying to do prior to this patch (at least I think that is what the original authors intended), but it was done rather poorly and added a layer of obfuscation that only masked the underlying problems. * Big backlog queue cleanup, again. In v4.10 we made some pretty big changes to how the audit backlog queues work, here we haven't changed the queue design so much as cleaned up the implementation. Brought about by the locking changes, we've simplified kauditd_thread() quite a bit by consolidating the queue handling into a new helper function, kauditd_send_queue(), which allows us to eliminate a lot of very similar code and makes the looping logic in kauditd_thread() clearer. * All netlink messages sent to auditd are now sent via auditd_send_unicast_skb(). Other than just making sense, this makes the lock handling easier. * Change the audit_log_start() sleep behavior so that we never sleep on auditd events (unchanged) or if the caller is holding the audit_cmd_mutex (changed). Previously we didn't sleep if the caller was auditd or if the message type fell between a certain range; the type check was a poor effort of doing what the cmd_mutex check now does. Richard Guy Briggs originally proposed not sleeping the cmd_mutex owner several years ago but his patch wasn't acceptable at the time. At least the idea lives on here. * A problem with the lost record counter has been resolved. Steve Grubb and I both happened to notice this problem and according to some quick testing by Steve, this problem goes back quite some time. It's largely a harmless problem, although it may have left some careful sysadmins quite puzzled. Cc: <stable@vger.kernel.org> # 4.10.x- Reported-by: Dmitry Vyukov <dvyukov@google.com> Signed-off-by: Paul Moore <paul@paul-moore.com>
2017-03-21 23:26:35 +08:00
#include <linux/spinlock.h>
#include <linux/rcupdate.h>
#include <linux/mutex.h>
#include <linux/gfp.h>
#include <linux/pid.h>
#include <linux/audit.h>
#include <net/sock.h>
#include <net/netlink.h>
#include <linux/skbuff.h>
#ifdef CONFIG_SECURITY
#include <linux/security.h>
#endif
#include <linux/freezer.h>
#include <linux/pid_namespace.h>
#include <net/netns/generic.h>
#include "audit.h"
/* No auditing will take place until audit_initialized == AUDIT_INITIALIZED.
* (Initialization happens after skb_init is called.) */
#define AUDIT_DISABLED -1
#define AUDIT_UNINITIALIZED 0
#define AUDIT_INITIALIZED 1
static int audit_initialized = AUDIT_UNINITIALIZED;
u32 audit_enabled = AUDIT_OFF;
bool audit_ever_enabled = !!AUDIT_OFF;
EXPORT_SYMBOL_GPL(audit_enabled);
/* Default state when kernel boots without any parameters. */
static u32 audit_default = AUDIT_OFF;
/* If auditing cannot proceed, audit_failure selects what happens. */
static u32 audit_failure = AUDIT_FAIL_PRINTK;
audit: fix auditd/kernel connection state tracking What started as a rather straightforward race condition reported by Dmitry using the syzkaller fuzzer ended up revealing some major problems with how the audit subsystem managed its netlink sockets and its connection with the userspace audit daemon. Fixing this properly had quite the cascading effect and what we are left with is this rather large and complicated patch. My initial goal was to try and decompose this patch into multiple smaller patches, but the way these changes are intertwined makes it difficult to split these changes into meaningful pieces that don't break or somehow make things worse for the intermediate states. The patch makes a number of changes, but the most significant are highlighted below: * The auditd tracking variables, e.g. audit_sock, are now gone and replaced by a RCU/spin_lock protected variable auditd_conn which is a structure containing all of the auditd tracking information. * We no longer track the auditd sock directly, instead we track it via the network namespace in which it resides and we use the audit socket associated with that namespace. In spirit, this is what the code was trying to do prior to this patch (at least I think that is what the original authors intended), but it was done rather poorly and added a layer of obfuscation that only masked the underlying problems. * Big backlog queue cleanup, again. In v4.10 we made some pretty big changes to how the audit backlog queues work, here we haven't changed the queue design so much as cleaned up the implementation. Brought about by the locking changes, we've simplified kauditd_thread() quite a bit by consolidating the queue handling into a new helper function, kauditd_send_queue(), which allows us to eliminate a lot of very similar code and makes the looping logic in kauditd_thread() clearer. * All netlink messages sent to auditd are now sent via auditd_send_unicast_skb(). Other than just making sense, this makes the lock handling easier. * Change the audit_log_start() sleep behavior so that we never sleep on auditd events (unchanged) or if the caller is holding the audit_cmd_mutex (changed). Previously we didn't sleep if the caller was auditd or if the message type fell between a certain range; the type check was a poor effort of doing what the cmd_mutex check now does. Richard Guy Briggs originally proposed not sleeping the cmd_mutex owner several years ago but his patch wasn't acceptable at the time. At least the idea lives on here. * A problem with the lost record counter has been resolved. Steve Grubb and I both happened to notice this problem and according to some quick testing by Steve, this problem goes back quite some time. It's largely a harmless problem, although it may have left some careful sysadmins quite puzzled. Cc: <stable@vger.kernel.org> # 4.10.x- Reported-by: Dmitry Vyukov <dvyukov@google.com> Signed-off-by: Paul Moore <paul@paul-moore.com>
2017-03-21 23:26:35 +08:00
/* private audit network namespace index */
static unsigned int audit_net_id;
/**
* struct audit_net - audit private network namespace data
* @sk: communication socket
*/
struct audit_net {
struct sock *sk;
};
/**
* struct auditd_connection - kernel/auditd connection state
* @pid: auditd PID
* @portid: netlink portid
* @net: the associated network namespace
* @rcu: RCU head
audit: fix auditd/kernel connection state tracking What started as a rather straightforward race condition reported by Dmitry using the syzkaller fuzzer ended up revealing some major problems with how the audit subsystem managed its netlink sockets and its connection with the userspace audit daemon. Fixing this properly had quite the cascading effect and what we are left with is this rather large and complicated patch. My initial goal was to try and decompose this patch into multiple smaller patches, but the way these changes are intertwined makes it difficult to split these changes into meaningful pieces that don't break or somehow make things worse for the intermediate states. The patch makes a number of changes, but the most significant are highlighted below: * The auditd tracking variables, e.g. audit_sock, are now gone and replaced by a RCU/spin_lock protected variable auditd_conn which is a structure containing all of the auditd tracking information. * We no longer track the auditd sock directly, instead we track it via the network namespace in which it resides and we use the audit socket associated with that namespace. In spirit, this is what the code was trying to do prior to this patch (at least I think that is what the original authors intended), but it was done rather poorly and added a layer of obfuscation that only masked the underlying problems. * Big backlog queue cleanup, again. In v4.10 we made some pretty big changes to how the audit backlog queues work, here we haven't changed the queue design so much as cleaned up the implementation. Brought about by the locking changes, we've simplified kauditd_thread() quite a bit by consolidating the queue handling into a new helper function, kauditd_send_queue(), which allows us to eliminate a lot of very similar code and makes the looping logic in kauditd_thread() clearer. * All netlink messages sent to auditd are now sent via auditd_send_unicast_skb(). Other than just making sense, this makes the lock handling easier. * Change the audit_log_start() sleep behavior so that we never sleep on auditd events (unchanged) or if the caller is holding the audit_cmd_mutex (changed). Previously we didn't sleep if the caller was auditd or if the message type fell between a certain range; the type check was a poor effort of doing what the cmd_mutex check now does. Richard Guy Briggs originally proposed not sleeping the cmd_mutex owner several years ago but his patch wasn't acceptable at the time. At least the idea lives on here. * A problem with the lost record counter has been resolved. Steve Grubb and I both happened to notice this problem and according to some quick testing by Steve, this problem goes back quite some time. It's largely a harmless problem, although it may have left some careful sysadmins quite puzzled. Cc: <stable@vger.kernel.org> # 4.10.x- Reported-by: Dmitry Vyukov <dvyukov@google.com> Signed-off-by: Paul Moore <paul@paul-moore.com>
2017-03-21 23:26:35 +08:00
*
* Description:
* This struct is RCU protected; you must either hold the RCU lock for reading
* or the associated spinlock for writing.
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>
2008-03-21 06:39:41 +08:00
*/
struct auditd_connection {
struct pid *pid;
audit: fix auditd/kernel connection state tracking What started as a rather straightforward race condition reported by Dmitry using the syzkaller fuzzer ended up revealing some major problems with how the audit subsystem managed its netlink sockets and its connection with the userspace audit daemon. Fixing this properly had quite the cascading effect and what we are left with is this rather large and complicated patch. My initial goal was to try and decompose this patch into multiple smaller patches, but the way these changes are intertwined makes it difficult to split these changes into meaningful pieces that don't break or somehow make things worse for the intermediate states. The patch makes a number of changes, but the most significant are highlighted below: * The auditd tracking variables, e.g. audit_sock, are now gone and replaced by a RCU/spin_lock protected variable auditd_conn which is a structure containing all of the auditd tracking information. * We no longer track the auditd sock directly, instead we track it via the network namespace in which it resides and we use the audit socket associated with that namespace. In spirit, this is what the code was trying to do prior to this patch (at least I think that is what the original authors intended), but it was done rather poorly and added a layer of obfuscation that only masked the underlying problems. * Big backlog queue cleanup, again. In v4.10 we made some pretty big changes to how the audit backlog queues work, here we haven't changed the queue design so much as cleaned up the implementation. Brought about by the locking changes, we've simplified kauditd_thread() quite a bit by consolidating the queue handling into a new helper function, kauditd_send_queue(), which allows us to eliminate a lot of very similar code and makes the looping logic in kauditd_thread() clearer. * All netlink messages sent to auditd are now sent via auditd_send_unicast_skb(). Other than just making sense, this makes the lock handling easier. * Change the audit_log_start() sleep behavior so that we never sleep on auditd events (unchanged) or if the caller is holding the audit_cmd_mutex (changed). Previously we didn't sleep if the caller was auditd or if the message type fell between a certain range; the type check was a poor effort of doing what the cmd_mutex check now does. Richard Guy Briggs originally proposed not sleeping the cmd_mutex owner several years ago but his patch wasn't acceptable at the time. At least the idea lives on here. * A problem with the lost record counter has been resolved. Steve Grubb and I both happened to notice this problem and according to some quick testing by Steve, this problem goes back quite some time. It's largely a harmless problem, although it may have left some careful sysadmins quite puzzled. Cc: <stable@vger.kernel.org> # 4.10.x- Reported-by: Dmitry Vyukov <dvyukov@google.com> Signed-off-by: Paul Moore <paul@paul-moore.com>
2017-03-21 23:26:35 +08:00
u32 portid;
struct net *net;
struct rcu_head rcu;
};
static struct auditd_connection __rcu *auditd_conn;
static DEFINE_SPINLOCK(auditd_conn_lock);
/* If audit_rate_limit is non-zero, limit the rate of sending audit records
* to that number per second. This prevents DoS attacks, but results in
* audit records being dropped. */
static u32 audit_rate_limit;
/* Number of outstanding audit_buffers allowed.
* When set to zero, this means unlimited. */
static u32 audit_backlog_limit = 64;
#define AUDIT_BACKLOG_WAIT_TIME (60 * HZ)
static u32 audit_backlog_wait_time = AUDIT_BACKLOG_WAIT_TIME;
/* The identity of the user shutting down the audit system. */
static kuid_t audit_sig_uid = INVALID_UID;
static pid_t audit_sig_pid = -1;
static u32 audit_sig_sid;
/* 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);
/* Monotonically increasing sum of time the kernel has spent
* waiting while the backlog limit is exceeded.
*/
static atomic_t audit_backlog_wait_time_actual = ATOMIC_INIT(0);
/* Hash for inode-based rules */
struct list_head audit_inode_hash[AUDIT_INODE_BUCKETS];
static struct kmem_cache *audit_buffer_cache;
/* queue msgs to send via kauditd_task */
static struct sk_buff_head audit_queue;
/* queue msgs due to temporary unicast send problems */
static struct sk_buff_head audit_retry_queue;
/* queue msgs waiting for new auditd connection */
static struct sk_buff_head audit_hold_queue;
/* queue servicing thread */
static struct task_struct *kauditd_task;
static DECLARE_WAIT_QUEUE_HEAD(kauditd_wait);
/* waitqueue for callers who are blocked on the audit backlog */
static DECLARE_WAIT_QUEUE_HEAD(audit_backlog_wait);
static struct audit_features af = {.vers = AUDIT_FEATURE_VERSION,
.mask = -1,
.features = 0,
.lock = 0,};
static char *audit_feature_names[2] = {
"only_unset_loginuid",
"loginuid_immutable",
};
/**
* struct audit_ctl_mutex - serialize requests from userspace
* @lock: the mutex used for locking
* @owner: the task which owns the lock
*
* Description:
* This is the lock struct used to ensure we only process userspace requests
* in an orderly fashion. We can't simply use a mutex/lock here because we
* need to track lock ownership so we don't end up blocking the lock owner in
* audit_log_start() or similar.
*/
static struct audit_ctl_mutex {
struct mutex lock;
void *owner;
} audit_cmd_mutex;
/* 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
/* 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 sk_buff *skb; /* formatted skb ready to send */
struct audit_context *ctx; /* NULL or associated context */
gfp_t gfp_mask;
};
struct audit_reply {
__u32 portid;
struct net *net;
struct sk_buff *skb;
};
audit: fix auditd/kernel connection state tracking What started as a rather straightforward race condition reported by Dmitry using the syzkaller fuzzer ended up revealing some major problems with how the audit subsystem managed its netlink sockets and its connection with the userspace audit daemon. Fixing this properly had quite the cascading effect and what we are left with is this rather large and complicated patch. My initial goal was to try and decompose this patch into multiple smaller patches, but the way these changes are intertwined makes it difficult to split these changes into meaningful pieces that don't break or somehow make things worse for the intermediate states. The patch makes a number of changes, but the most significant are highlighted below: * The auditd tracking variables, e.g. audit_sock, are now gone and replaced by a RCU/spin_lock protected variable auditd_conn which is a structure containing all of the auditd tracking information. * We no longer track the auditd sock directly, instead we track it via the network namespace in which it resides and we use the audit socket associated with that namespace. In spirit, this is what the code was trying to do prior to this patch (at least I think that is what the original authors intended), but it was done rather poorly and added a layer of obfuscation that only masked the underlying problems. * Big backlog queue cleanup, again. In v4.10 we made some pretty big changes to how the audit backlog queues work, here we haven't changed the queue design so much as cleaned up the implementation. Brought about by the locking changes, we've simplified kauditd_thread() quite a bit by consolidating the queue handling into a new helper function, kauditd_send_queue(), which allows us to eliminate a lot of very similar code and makes the looping logic in kauditd_thread() clearer. * All netlink messages sent to auditd are now sent via auditd_send_unicast_skb(). Other than just making sense, this makes the lock handling easier. * Change the audit_log_start() sleep behavior so that we never sleep on auditd events (unchanged) or if the caller is holding the audit_cmd_mutex (changed). Previously we didn't sleep if the caller was auditd or if the message type fell between a certain range; the type check was a poor effort of doing what the cmd_mutex check now does. Richard Guy Briggs originally proposed not sleeping the cmd_mutex owner several years ago but his patch wasn't acceptable at the time. At least the idea lives on here. * A problem with the lost record counter has been resolved. Steve Grubb and I both happened to notice this problem and according to some quick testing by Steve, this problem goes back quite some time. It's largely a harmless problem, although it may have left some careful sysadmins quite puzzled. Cc: <stable@vger.kernel.org> # 4.10.x- Reported-by: Dmitry Vyukov <dvyukov@google.com> Signed-off-by: Paul Moore <paul@paul-moore.com>
2017-03-21 23:26:35 +08:00
/**
* auditd_test_task - Check to see if a given task is an audit daemon
* @task: the task to check
*
* Description:
* Return 1 if the task is a registered audit daemon, 0 otherwise.
*/
int auditd_test_task(struct task_struct *task)
audit: fix auditd/kernel connection state tracking What started as a rather straightforward race condition reported by Dmitry using the syzkaller fuzzer ended up revealing some major problems with how the audit subsystem managed its netlink sockets and its connection with the userspace audit daemon. Fixing this properly had quite the cascading effect and what we are left with is this rather large and complicated patch. My initial goal was to try and decompose this patch into multiple smaller patches, but the way these changes are intertwined makes it difficult to split these changes into meaningful pieces that don't break or somehow make things worse for the intermediate states. The patch makes a number of changes, but the most significant are highlighted below: * The auditd tracking variables, e.g. audit_sock, are now gone and replaced by a RCU/spin_lock protected variable auditd_conn which is a structure containing all of the auditd tracking information. * We no longer track the auditd sock directly, instead we track it via the network namespace in which it resides and we use the audit socket associated with that namespace. In spirit, this is what the code was trying to do prior to this patch (at least I think that is what the original authors intended), but it was done rather poorly and added a layer of obfuscation that only masked the underlying problems. * Big backlog queue cleanup, again. In v4.10 we made some pretty big changes to how the audit backlog queues work, here we haven't changed the queue design so much as cleaned up the implementation. Brought about by the locking changes, we've simplified kauditd_thread() quite a bit by consolidating the queue handling into a new helper function, kauditd_send_queue(), which allows us to eliminate a lot of very similar code and makes the looping logic in kauditd_thread() clearer. * All netlink messages sent to auditd are now sent via auditd_send_unicast_skb(). Other than just making sense, this makes the lock handling easier. * Change the audit_log_start() sleep behavior so that we never sleep on auditd events (unchanged) or if the caller is holding the audit_cmd_mutex (changed). Previously we didn't sleep if the caller was auditd or if the message type fell between a certain range; the type check was a poor effort of doing what the cmd_mutex check now does. Richard Guy Briggs originally proposed not sleeping the cmd_mutex owner several years ago but his patch wasn't acceptable at the time. At least the idea lives on here. * A problem with the lost record counter has been resolved. Steve Grubb and I both happened to notice this problem and according to some quick testing by Steve, this problem goes back quite some time. It's largely a harmless problem, although it may have left some careful sysadmins quite puzzled. Cc: <stable@vger.kernel.org> # 4.10.x- Reported-by: Dmitry Vyukov <dvyukov@google.com> Signed-off-by: Paul Moore <paul@paul-moore.com>
2017-03-21 23:26:35 +08:00
{
int rc;
struct auditd_connection *ac;
audit: fix auditd/kernel connection state tracking What started as a rather straightforward race condition reported by Dmitry using the syzkaller fuzzer ended up revealing some major problems with how the audit subsystem managed its netlink sockets and its connection with the userspace audit daemon. Fixing this properly had quite the cascading effect and what we are left with is this rather large and complicated patch. My initial goal was to try and decompose this patch into multiple smaller patches, but the way these changes are intertwined makes it difficult to split these changes into meaningful pieces that don't break or somehow make things worse for the intermediate states. The patch makes a number of changes, but the most significant are highlighted below: * The auditd tracking variables, e.g. audit_sock, are now gone and replaced by a RCU/spin_lock protected variable auditd_conn which is a structure containing all of the auditd tracking information. * We no longer track the auditd sock directly, instead we track it via the network namespace in which it resides and we use the audit socket associated with that namespace. In spirit, this is what the code was trying to do prior to this patch (at least I think that is what the original authors intended), but it was done rather poorly and added a layer of obfuscation that only masked the underlying problems. * Big backlog queue cleanup, again. In v4.10 we made some pretty big changes to how the audit backlog queues work, here we haven't changed the queue design so much as cleaned up the implementation. Brought about by the locking changes, we've simplified kauditd_thread() quite a bit by consolidating the queue handling into a new helper function, kauditd_send_queue(), which allows us to eliminate a lot of very similar code and makes the looping logic in kauditd_thread() clearer. * All netlink messages sent to auditd are now sent via auditd_send_unicast_skb(). Other than just making sense, this makes the lock handling easier. * Change the audit_log_start() sleep behavior so that we never sleep on auditd events (unchanged) or if the caller is holding the audit_cmd_mutex (changed). Previously we didn't sleep if the caller was auditd or if the message type fell between a certain range; the type check was a poor effort of doing what the cmd_mutex check now does. Richard Guy Briggs originally proposed not sleeping the cmd_mutex owner several years ago but his patch wasn't acceptable at the time. At least the idea lives on here. * A problem with the lost record counter has been resolved. Steve Grubb and I both happened to notice this problem and according to some quick testing by Steve, this problem goes back quite some time. It's largely a harmless problem, although it may have left some careful sysadmins quite puzzled. Cc: <stable@vger.kernel.org> # 4.10.x- Reported-by: Dmitry Vyukov <dvyukov@google.com> Signed-off-by: Paul Moore <paul@paul-moore.com>
2017-03-21 23:26:35 +08:00
rcu_read_lock();
ac = rcu_dereference(auditd_conn);
rc = (ac && ac->pid == task_tgid(task) ? 1 : 0);
audit: fix auditd/kernel connection state tracking What started as a rather straightforward race condition reported by Dmitry using the syzkaller fuzzer ended up revealing some major problems with how the audit subsystem managed its netlink sockets and its connection with the userspace audit daemon. Fixing this properly had quite the cascading effect and what we are left with is this rather large and complicated patch. My initial goal was to try and decompose this patch into multiple smaller patches, but the way these changes are intertwined makes it difficult to split these changes into meaningful pieces that don't break or somehow make things worse for the intermediate states. The patch makes a number of changes, but the most significant are highlighted below: * The auditd tracking variables, e.g. audit_sock, are now gone and replaced by a RCU/spin_lock protected variable auditd_conn which is a structure containing all of the auditd tracking information. * We no longer track the auditd sock directly, instead we track it via the network namespace in which it resides and we use the audit socket associated with that namespace. In spirit, this is what the code was trying to do prior to this patch (at least I think that is what the original authors intended), but it was done rather poorly and added a layer of obfuscation that only masked the underlying problems. * Big backlog queue cleanup, again. In v4.10 we made some pretty big changes to how the audit backlog queues work, here we haven't changed the queue design so much as cleaned up the implementation. Brought about by the locking changes, we've simplified kauditd_thread() quite a bit by consolidating the queue handling into a new helper function, kauditd_send_queue(), which allows us to eliminate a lot of very similar code and makes the looping logic in kauditd_thread() clearer. * All netlink messages sent to auditd are now sent via auditd_send_unicast_skb(). Other than just making sense, this makes the lock handling easier. * Change the audit_log_start() sleep behavior so that we never sleep on auditd events (unchanged) or if the caller is holding the audit_cmd_mutex (changed). Previously we didn't sleep if the caller was auditd or if the message type fell between a certain range; the type check was a poor effort of doing what the cmd_mutex check now does. Richard Guy Briggs originally proposed not sleeping the cmd_mutex owner several years ago but his patch wasn't acceptable at the time. At least the idea lives on here. * A problem with the lost record counter has been resolved. Steve Grubb and I both happened to notice this problem and according to some quick testing by Steve, this problem goes back quite some time. It's largely a harmless problem, although it may have left some careful sysadmins quite puzzled. Cc: <stable@vger.kernel.org> # 4.10.x- Reported-by: Dmitry Vyukov <dvyukov@google.com> Signed-off-by: Paul Moore <paul@paul-moore.com>
2017-03-21 23:26:35 +08:00
rcu_read_unlock();
return rc;
}
/**
* audit_ctl_lock - Take the audit control lock
*/
void audit_ctl_lock(void)
{
mutex_lock(&audit_cmd_mutex.lock);
audit_cmd_mutex.owner = current;
}
/**
* audit_ctl_unlock - Drop the audit control lock
*/
void audit_ctl_unlock(void)
{
audit_cmd_mutex.owner = NULL;
mutex_unlock(&audit_cmd_mutex.lock);
}
/**
* audit_ctl_owner_current - Test to see if the current task owns the lock
*
* Description:
* Return true if the current task owns the audit control lock, false if it
* doesn't own the lock.
*/
static bool audit_ctl_owner_current(void)
{
return (current == audit_cmd_mutex.owner);
}
/**
* auditd_pid_vnr - Return the auditd PID relative to the namespace
*
* Description:
* Returns the PID in relation to the namespace, 0 on failure.
*/
static pid_t auditd_pid_vnr(void)
{
pid_t pid;
const struct auditd_connection *ac;
rcu_read_lock();
ac = rcu_dereference(auditd_conn);
if (!ac || !ac->pid)
pid = 0;
else
pid = pid_vnr(ac->pid);
rcu_read_unlock();
return pid;
}
audit: fix auditd/kernel connection state tracking What started as a rather straightforward race condition reported by Dmitry using the syzkaller fuzzer ended up revealing some major problems with how the audit subsystem managed its netlink sockets and its connection with the userspace audit daemon. Fixing this properly had quite the cascading effect and what we are left with is this rather large and complicated patch. My initial goal was to try and decompose this patch into multiple smaller patches, but the way these changes are intertwined makes it difficult to split these changes into meaningful pieces that don't break or somehow make things worse for the intermediate states. The patch makes a number of changes, but the most significant are highlighted below: * The auditd tracking variables, e.g. audit_sock, are now gone and replaced by a RCU/spin_lock protected variable auditd_conn which is a structure containing all of the auditd tracking information. * We no longer track the auditd sock directly, instead we track it via the network namespace in which it resides and we use the audit socket associated with that namespace. In spirit, this is what the code was trying to do prior to this patch (at least I think that is what the original authors intended), but it was done rather poorly and added a layer of obfuscation that only masked the underlying problems. * Big backlog queue cleanup, again. In v4.10 we made some pretty big changes to how the audit backlog queues work, here we haven't changed the queue design so much as cleaned up the implementation. Brought about by the locking changes, we've simplified kauditd_thread() quite a bit by consolidating the queue handling into a new helper function, kauditd_send_queue(), which allows us to eliminate a lot of very similar code and makes the looping logic in kauditd_thread() clearer. * All netlink messages sent to auditd are now sent via auditd_send_unicast_skb(). Other than just making sense, this makes the lock handling easier. * Change the audit_log_start() sleep behavior so that we never sleep on auditd events (unchanged) or if the caller is holding the audit_cmd_mutex (changed). Previously we didn't sleep if the caller was auditd or if the message type fell between a certain range; the type check was a poor effort of doing what the cmd_mutex check now does. Richard Guy Briggs originally proposed not sleeping the cmd_mutex owner several years ago but his patch wasn't acceptable at the time. At least the idea lives on here. * A problem with the lost record counter has been resolved. Steve Grubb and I both happened to notice this problem and according to some quick testing by Steve, this problem goes back quite some time. It's largely a harmless problem, although it may have left some careful sysadmins quite puzzled. Cc: <stable@vger.kernel.org> # 4.10.x- Reported-by: Dmitry Vyukov <dvyukov@google.com> Signed-off-by: Paul Moore <paul@paul-moore.com>
2017-03-21 23:26:35 +08:00
/**
* audit_get_sk - Return the audit socket for the given network namespace
* @net: the destination network namespace
*
* Description:
* Returns the sock pointer if valid, NULL otherwise. The caller must ensure
* that a reference is held for the network namespace while the sock is in use.
*/
static struct sock *audit_get_sk(const struct net *net)
{
struct audit_net *aunet;
if (!net)
return NULL;
aunet = net_generic(net, audit_net_id);
return aunet->sk;
}
void audit_panic(const char *message)
{
switch (audit_failure) {
case AUDIT_FAIL_SILENT:
break;
case AUDIT_FAIL_PRINTK:
if (printk_ratelimit())
pr_err("%s\n", message);
break;
case AUDIT_FAIL_PANIC:
audit: fix auditd/kernel connection state tracking What started as a rather straightforward race condition reported by Dmitry using the syzkaller fuzzer ended up revealing some major problems with how the audit subsystem managed its netlink sockets and its connection with the userspace audit daemon. Fixing this properly had quite the cascading effect and what we are left with is this rather large and complicated patch. My initial goal was to try and decompose this patch into multiple smaller patches, but the way these changes are intertwined makes it difficult to split these changes into meaningful pieces that don't break or somehow make things worse for the intermediate states. The patch makes a number of changes, but the most significant are highlighted below: * The auditd tracking variables, e.g. audit_sock, are now gone and replaced by a RCU/spin_lock protected variable auditd_conn which is a structure containing all of the auditd tracking information. * We no longer track the auditd sock directly, instead we track it via the network namespace in which it resides and we use the audit socket associated with that namespace. In spirit, this is what the code was trying to do prior to this patch (at least I think that is what the original authors intended), but it was done rather poorly and added a layer of obfuscation that only masked the underlying problems. * Big backlog queue cleanup, again. In v4.10 we made some pretty big changes to how the audit backlog queues work, here we haven't changed the queue design so much as cleaned up the implementation. Brought about by the locking changes, we've simplified kauditd_thread() quite a bit by consolidating the queue handling into a new helper function, kauditd_send_queue(), which allows us to eliminate a lot of very similar code and makes the looping logic in kauditd_thread() clearer. * All netlink messages sent to auditd are now sent via auditd_send_unicast_skb(). Other than just making sense, this makes the lock handling easier. * Change the audit_log_start() sleep behavior so that we never sleep on auditd events (unchanged) or if the caller is holding the audit_cmd_mutex (changed). Previously we didn't sleep if the caller was auditd or if the message type fell between a certain range; the type check was a poor effort of doing what the cmd_mutex check now does. Richard Guy Briggs originally proposed not sleeping the cmd_mutex owner several years ago but his patch wasn't acceptable at the time. At least the idea lives on here. * A problem with the lost record counter has been resolved. Steve Grubb and I both happened to notice this problem and according to some quick testing by Steve, this problem goes back quite some time. It's largely a harmless problem, although it may have left some careful sysadmins quite puzzled. Cc: <stable@vger.kernel.org> # 4.10.x- Reported-by: Dmitry Vyukov <dvyukov@google.com> Signed-off-by: Paul Moore <paul@paul-moore.com>
2017-03-21 23:26:35 +08:00
panic("audit: %s\n", message);
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;
}
/**
* 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.
*/
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) {
if (printk_ratelimit())
pr_warn("audit_lost=%u audit_rate_limit=%u audit_backlog_limit=%u\n",
atomic_read(&audit_lost),
audit_rate_limit,
audit_backlog_limit);
audit_panic(message);
}
}
static int audit_log_config_change(char *function_name, u32 new, u32 old,
int allow_changes)
{
struct audit_buffer *ab;
int rc = 0;
ab = audit_log_start(audit_context(), GFP_KERNEL, AUDIT_CONFIG_CHANGE);
if (unlikely(!ab))
return rc;
audit_log_format(ab, "op=set %s=%u old=%u ", function_name, new, old);
2013-04-30 21:53:34 +08:00
audit_log_session_info(ab);
rc = audit_log_task_context(ab);
if (rc)
allow_changes = 0; /* Something weird, deny request */
audit_log_format(ab, " res=%d", allow_changes);
audit_log_end(ab);
return rc;
}
static int audit_do_config_change(char *function_name, u32 *to_change, u32 new)
{
int allow_changes, rc = 0;
u32 old = *to_change;
/* check if we are locked */
if (audit_enabled == AUDIT_LOCKED)
allow_changes = 0;
else
allow_changes = 1;
if (audit_enabled != AUDIT_OFF) {
rc = audit_log_config_change(function_name, new, old, allow_changes);
if (rc)
allow_changes = 0;
}
/* If we are allowed, make the change */
if (allow_changes == 1)
*to_change = new;
/* Not allowed, update reason */
else if (rc == 0)
rc = -EPERM;
return rc;
}
static int audit_set_rate_limit(u32 limit)
{
return audit_do_config_change("audit_rate_limit", &audit_rate_limit, limit);
}
static int audit_set_backlog_limit(u32 limit)
{
return audit_do_config_change("audit_backlog_limit", &audit_backlog_limit, limit);
}
static int audit_set_backlog_wait_time(u32 timeout)
{
return audit_do_config_change("audit_backlog_wait_time",
&audit_backlog_wait_time, timeout);
}
static int audit_set_enabled(u32 state)
{
int rc;
if (state > AUDIT_LOCKED)
return -EINVAL;
rc = audit_do_config_change("audit_enabled", &audit_enabled, state);
if (!rc)
audit_ever_enabled |= !!state;
return rc;
}
static int audit_set_failure(u32 state)
{
if (state != AUDIT_FAIL_SILENT
&& state != AUDIT_FAIL_PRINTK
&& state != AUDIT_FAIL_PANIC)
return -EINVAL;
return audit_do_config_change("audit_failure", &audit_failure, state);
}
/**
* auditd_conn_free - RCU helper to release an auditd connection struct
* @rcu: RCU head
*
* Description:
* Drop any references inside the auditd connection tracking struct and free
* the memory.
*/
static void auditd_conn_free(struct rcu_head *rcu)
{
struct auditd_connection *ac;
ac = container_of(rcu, struct auditd_connection, rcu);
put_pid(ac->pid);
put_net(ac->net);
kfree(ac);
}
audit: fix auditd/kernel connection state tracking What started as a rather straightforward race condition reported by Dmitry using the syzkaller fuzzer ended up revealing some major problems with how the audit subsystem managed its netlink sockets and its connection with the userspace audit daemon. Fixing this properly had quite the cascading effect and what we are left with is this rather large and complicated patch. My initial goal was to try and decompose this patch into multiple smaller patches, but the way these changes are intertwined makes it difficult to split these changes into meaningful pieces that don't break or somehow make things worse for the intermediate states. The patch makes a number of changes, but the most significant are highlighted below: * The auditd tracking variables, e.g. audit_sock, are now gone and replaced by a RCU/spin_lock protected variable auditd_conn which is a structure containing all of the auditd tracking information. * We no longer track the auditd sock directly, instead we track it via the network namespace in which it resides and we use the audit socket associated with that namespace. In spirit, this is what the code was trying to do prior to this patch (at least I think that is what the original authors intended), but it was done rather poorly and added a layer of obfuscation that only masked the underlying problems. * Big backlog queue cleanup, again. In v4.10 we made some pretty big changes to how the audit backlog queues work, here we haven't changed the queue design so much as cleaned up the implementation. Brought about by the locking changes, we've simplified kauditd_thread() quite a bit by consolidating the queue handling into a new helper function, kauditd_send_queue(), which allows us to eliminate a lot of very similar code and makes the looping logic in kauditd_thread() clearer. * All netlink messages sent to auditd are now sent via auditd_send_unicast_skb(). Other than just making sense, this makes the lock handling easier. * Change the audit_log_start() sleep behavior so that we never sleep on auditd events (unchanged) or if the caller is holding the audit_cmd_mutex (changed). Previously we didn't sleep if the caller was auditd or if the message type fell between a certain range; the type check was a poor effort of doing what the cmd_mutex check now does. Richard Guy Briggs originally proposed not sleeping the cmd_mutex owner several years ago but his patch wasn't acceptable at the time. At least the idea lives on here. * A problem with the lost record counter has been resolved. Steve Grubb and I both happened to notice this problem and according to some quick testing by Steve, this problem goes back quite some time. It's largely a harmless problem, although it may have left some careful sysadmins quite puzzled. Cc: <stable@vger.kernel.org> # 4.10.x- Reported-by: Dmitry Vyukov <dvyukov@google.com> Signed-off-by: Paul Moore <paul@paul-moore.com>
2017-03-21 23:26:35 +08:00
/**
* auditd_set - Set/Reset the auditd connection state
* @pid: auditd PID
* @portid: auditd netlink portid
* @net: auditd network namespace pointer
*
* Description:
* This function will obtain and drop network namespace references as
* necessary. Returns zero on success, negative values on failure.
audit: fix auditd/kernel connection state tracking What started as a rather straightforward race condition reported by Dmitry using the syzkaller fuzzer ended up revealing some major problems with how the audit subsystem managed its netlink sockets and its connection with the userspace audit daemon. Fixing this properly had quite the cascading effect and what we are left with is this rather large and complicated patch. My initial goal was to try and decompose this patch into multiple smaller patches, but the way these changes are intertwined makes it difficult to split these changes into meaningful pieces that don't break or somehow make things worse for the intermediate states. The patch makes a number of changes, but the most significant are highlighted below: * The auditd tracking variables, e.g. audit_sock, are now gone and replaced by a RCU/spin_lock protected variable auditd_conn which is a structure containing all of the auditd tracking information. * We no longer track the auditd sock directly, instead we track it via the network namespace in which it resides and we use the audit socket associated with that namespace. In spirit, this is what the code was trying to do prior to this patch (at least I think that is what the original authors intended), but it was done rather poorly and added a layer of obfuscation that only masked the underlying problems. * Big backlog queue cleanup, again. In v4.10 we made some pretty big changes to how the audit backlog queues work, here we haven't changed the queue design so much as cleaned up the implementation. Brought about by the locking changes, we've simplified kauditd_thread() quite a bit by consolidating the queue handling into a new helper function, kauditd_send_queue(), which allows us to eliminate a lot of very similar code and makes the looping logic in kauditd_thread() clearer. * All netlink messages sent to auditd are now sent via auditd_send_unicast_skb(). Other than just making sense, this makes the lock handling easier. * Change the audit_log_start() sleep behavior so that we never sleep on auditd events (unchanged) or if the caller is holding the audit_cmd_mutex (changed). Previously we didn't sleep if the caller was auditd or if the message type fell between a certain range; the type check was a poor effort of doing what the cmd_mutex check now does. Richard Guy Briggs originally proposed not sleeping the cmd_mutex owner several years ago but his patch wasn't acceptable at the time. At least the idea lives on here. * A problem with the lost record counter has been resolved. Steve Grubb and I both happened to notice this problem and according to some quick testing by Steve, this problem goes back quite some time. It's largely a harmless problem, although it may have left some careful sysadmins quite puzzled. Cc: <stable@vger.kernel.org> # 4.10.x- Reported-by: Dmitry Vyukov <dvyukov@google.com> Signed-off-by: Paul Moore <paul@paul-moore.com>
2017-03-21 23:26:35 +08:00
*/
static int auditd_set(struct pid *pid, u32 portid, struct net *net)
audit: fix auditd/kernel connection state tracking What started as a rather straightforward race condition reported by Dmitry using the syzkaller fuzzer ended up revealing some major problems with how the audit subsystem managed its netlink sockets and its connection with the userspace audit daemon. Fixing this properly had quite the cascading effect and what we are left with is this rather large and complicated patch. My initial goal was to try and decompose this patch into multiple smaller patches, but the way these changes are intertwined makes it difficult to split these changes into meaningful pieces that don't break or somehow make things worse for the intermediate states. The patch makes a number of changes, but the most significant are highlighted below: * The auditd tracking variables, e.g. audit_sock, are now gone and replaced by a RCU/spin_lock protected variable auditd_conn which is a structure containing all of the auditd tracking information. * We no longer track the auditd sock directly, instead we track it via the network namespace in which it resides and we use the audit socket associated with that namespace. In spirit, this is what the code was trying to do prior to this patch (at least I think that is what the original authors intended), but it was done rather poorly and added a layer of obfuscation that only masked the underlying problems. * Big backlog queue cleanup, again. In v4.10 we made some pretty big changes to how the audit backlog queues work, here we haven't changed the queue design so much as cleaned up the implementation. Brought about by the locking changes, we've simplified kauditd_thread() quite a bit by consolidating the queue handling into a new helper function, kauditd_send_queue(), which allows us to eliminate a lot of very similar code and makes the looping logic in kauditd_thread() clearer. * All netlink messages sent to auditd are now sent via auditd_send_unicast_skb(). Other than just making sense, this makes the lock handling easier. * Change the audit_log_start() sleep behavior so that we never sleep on auditd events (unchanged) or if the caller is holding the audit_cmd_mutex (changed). Previously we didn't sleep if the caller was auditd or if the message type fell between a certain range; the type check was a poor effort of doing what the cmd_mutex check now does. Richard Guy Briggs originally proposed not sleeping the cmd_mutex owner several years ago but his patch wasn't acceptable at the time. At least the idea lives on here. * A problem with the lost record counter has been resolved. Steve Grubb and I both happened to notice this problem and according to some quick testing by Steve, this problem goes back quite some time. It's largely a harmless problem, although it may have left some careful sysadmins quite puzzled. Cc: <stable@vger.kernel.org> # 4.10.x- Reported-by: Dmitry Vyukov <dvyukov@google.com> Signed-off-by: Paul Moore <paul@paul-moore.com>
2017-03-21 23:26:35 +08:00
{
unsigned long flags;
struct auditd_connection *ac_old, *ac_new;
audit: fix auditd/kernel connection state tracking What started as a rather straightforward race condition reported by Dmitry using the syzkaller fuzzer ended up revealing some major problems with how the audit subsystem managed its netlink sockets and its connection with the userspace audit daemon. Fixing this properly had quite the cascading effect and what we are left with is this rather large and complicated patch. My initial goal was to try and decompose this patch into multiple smaller patches, but the way these changes are intertwined makes it difficult to split these changes into meaningful pieces that don't break or somehow make things worse for the intermediate states. The patch makes a number of changes, but the most significant are highlighted below: * The auditd tracking variables, e.g. audit_sock, are now gone and replaced by a RCU/spin_lock protected variable auditd_conn which is a structure containing all of the auditd tracking information. * We no longer track the auditd sock directly, instead we track it via the network namespace in which it resides and we use the audit socket associated with that namespace. In spirit, this is what the code was trying to do prior to this patch (at least I think that is what the original authors intended), but it was done rather poorly and added a layer of obfuscation that only masked the underlying problems. * Big backlog queue cleanup, again. In v4.10 we made some pretty big changes to how the audit backlog queues work, here we haven't changed the queue design so much as cleaned up the implementation. Brought about by the locking changes, we've simplified kauditd_thread() quite a bit by consolidating the queue handling into a new helper function, kauditd_send_queue(), which allows us to eliminate a lot of very similar code and makes the looping logic in kauditd_thread() clearer. * All netlink messages sent to auditd are now sent via auditd_send_unicast_skb(). Other than just making sense, this makes the lock handling easier. * Change the audit_log_start() sleep behavior so that we never sleep on auditd events (unchanged) or if the caller is holding the audit_cmd_mutex (changed). Previously we didn't sleep if the caller was auditd or if the message type fell between a certain range; the type check was a poor effort of doing what the cmd_mutex check now does. Richard Guy Briggs originally proposed not sleeping the cmd_mutex owner several years ago but his patch wasn't acceptable at the time. At least the idea lives on here. * A problem with the lost record counter has been resolved. Steve Grubb and I both happened to notice this problem and according to some quick testing by Steve, this problem goes back quite some time. It's largely a harmless problem, although it may have left some careful sysadmins quite puzzled. Cc: <stable@vger.kernel.org> # 4.10.x- Reported-by: Dmitry Vyukov <dvyukov@google.com> Signed-off-by: Paul Moore <paul@paul-moore.com>
2017-03-21 23:26:35 +08:00
if (!pid || !net)
return -EINVAL;
ac_new = kzalloc(sizeof(*ac_new), GFP_KERNEL);
if (!ac_new)
return -ENOMEM;
ac_new->pid = get_pid(pid);
ac_new->portid = portid;
ac_new->net = get_net(net);
spin_lock_irqsave(&auditd_conn_lock, flags);
ac_old = rcu_dereference_protected(auditd_conn,
lockdep_is_held(&auditd_conn_lock));
rcu_assign_pointer(auditd_conn, ac_new);
spin_unlock_irqrestore(&auditd_conn_lock, flags);
if (ac_old)
call_rcu(&ac_old->rcu, auditd_conn_free);
return 0;
audit: fix auditd/kernel connection state tracking What started as a rather straightforward race condition reported by Dmitry using the syzkaller fuzzer ended up revealing some major problems with how the audit subsystem managed its netlink sockets and its connection with the userspace audit daemon. Fixing this properly had quite the cascading effect and what we are left with is this rather large and complicated patch. My initial goal was to try and decompose this patch into multiple smaller patches, but the way these changes are intertwined makes it difficult to split these changes into meaningful pieces that don't break or somehow make things worse for the intermediate states. The patch makes a number of changes, but the most significant are highlighted below: * The auditd tracking variables, e.g. audit_sock, are now gone and replaced by a RCU/spin_lock protected variable auditd_conn which is a structure containing all of the auditd tracking information. * We no longer track the auditd sock directly, instead we track it via the network namespace in which it resides and we use the audit socket associated with that namespace. In spirit, this is what the code was trying to do prior to this patch (at least I think that is what the original authors intended), but it was done rather poorly and added a layer of obfuscation that only masked the underlying problems. * Big backlog queue cleanup, again. In v4.10 we made some pretty big changes to how the audit backlog queues work, here we haven't changed the queue design so much as cleaned up the implementation. Brought about by the locking changes, we've simplified kauditd_thread() quite a bit by consolidating the queue handling into a new helper function, kauditd_send_queue(), which allows us to eliminate a lot of very similar code and makes the looping logic in kauditd_thread() clearer. * All netlink messages sent to auditd are now sent via auditd_send_unicast_skb(). Other than just making sense, this makes the lock handling easier. * Change the audit_log_start() sleep behavior so that we never sleep on auditd events (unchanged) or if the caller is holding the audit_cmd_mutex (changed). Previously we didn't sleep if the caller was auditd or if the message type fell between a certain range; the type check was a poor effort of doing what the cmd_mutex check now does. Richard Guy Briggs originally proposed not sleeping the cmd_mutex owner several years ago but his patch wasn't acceptable at the time. At least the idea lives on here. * A problem with the lost record counter has been resolved. Steve Grubb and I both happened to notice this problem and according to some quick testing by Steve, this problem goes back quite some time. It's largely a harmless problem, although it may have left some careful sysadmins quite puzzled. Cc: <stable@vger.kernel.org> # 4.10.x- Reported-by: Dmitry Vyukov <dvyukov@google.com> Signed-off-by: Paul Moore <paul@paul-moore.com>
2017-03-21 23:26:35 +08:00
}
/**
* kauditd_printk_skb - Print the audit record to the ring buffer
audit: fix auditd/kernel connection state tracking What started as a rather straightforward race condition reported by Dmitry using the syzkaller fuzzer ended up revealing some major problems with how the audit subsystem managed its netlink sockets and its connection with the userspace audit daemon. Fixing this properly had quite the cascading effect and what we are left with is this rather large and complicated patch. My initial goal was to try and decompose this patch into multiple smaller patches, but the way these changes are intertwined makes it difficult to split these changes into meaningful pieces that don't break or somehow make things worse for the intermediate states. The patch makes a number of changes, but the most significant are highlighted below: * The auditd tracking variables, e.g. audit_sock, are now gone and replaced by a RCU/spin_lock protected variable auditd_conn which is a structure containing all of the auditd tracking information. * We no longer track the auditd sock directly, instead we track it via the network namespace in which it resides and we use the audit socket associated with that namespace. In spirit, this is what the code was trying to do prior to this patch (at least I think that is what the original authors intended), but it was done rather poorly and added a layer of obfuscation that only masked the underlying problems. * Big backlog queue cleanup, again. In v4.10 we made some pretty big changes to how the audit backlog queues work, here we haven't changed the queue design so much as cleaned up the implementation. Brought about by the locking changes, we've simplified kauditd_thread() quite a bit by consolidating the queue handling into a new helper function, kauditd_send_queue(), which allows us to eliminate a lot of very similar code and makes the looping logic in kauditd_thread() clearer. * All netlink messages sent to auditd are now sent via auditd_send_unicast_skb(). Other than just making sense, this makes the lock handling easier. * Change the audit_log_start() sleep behavior so that we never sleep on auditd events (unchanged) or if the caller is holding the audit_cmd_mutex (changed). Previously we didn't sleep if the caller was auditd or if the message type fell between a certain range; the type check was a poor effort of doing what the cmd_mutex check now does. Richard Guy Briggs originally proposed not sleeping the cmd_mutex owner several years ago but his patch wasn't acceptable at the time. At least the idea lives on here. * A problem with the lost record counter has been resolved. Steve Grubb and I both happened to notice this problem and according to some quick testing by Steve, this problem goes back quite some time. It's largely a harmless problem, although it may have left some careful sysadmins quite puzzled. Cc: <stable@vger.kernel.org> # 4.10.x- Reported-by: Dmitry Vyukov <dvyukov@google.com> Signed-off-by: Paul Moore <paul@paul-moore.com>
2017-03-21 23:26:35 +08:00
* @skb: audit record
*
* Whatever the reason, this packet may not make it to the auditd connection
* so write it via printk so the information isn't completely lost.
*/
static void kauditd_printk_skb(struct sk_buff *skb)
{
struct nlmsghdr *nlh = nlmsg_hdr(skb);
char *data = nlmsg_data(nlh);
audit: fix auditd/kernel connection state tracking What started as a rather straightforward race condition reported by Dmitry using the syzkaller fuzzer ended up revealing some major problems with how the audit subsystem managed its netlink sockets and its connection with the userspace audit daemon. Fixing this properly had quite the cascading effect and what we are left with is this rather large and complicated patch. My initial goal was to try and decompose this patch into multiple smaller patches, but the way these changes are intertwined makes it difficult to split these changes into meaningful pieces that don't break or somehow make things worse for the intermediate states. The patch makes a number of changes, but the most significant are highlighted below: * The auditd tracking variables, e.g. audit_sock, are now gone and replaced by a RCU/spin_lock protected variable auditd_conn which is a structure containing all of the auditd tracking information. * We no longer track the auditd sock directly, instead we track it via the network namespace in which it resides and we use the audit socket associated with that namespace. In spirit, this is what the code was trying to do prior to this patch (at least I think that is what the original authors intended), but it was done rather poorly and added a layer of obfuscation that only masked the underlying problems. * Big backlog queue cleanup, again. In v4.10 we made some pretty big changes to how the audit backlog queues work, here we haven't changed the queue design so much as cleaned up the implementation. Brought about by the locking changes, we've simplified kauditd_thread() quite a bit by consolidating the queue handling into a new helper function, kauditd_send_queue(), which allows us to eliminate a lot of very similar code and makes the looping logic in kauditd_thread() clearer. * All netlink messages sent to auditd are now sent via auditd_send_unicast_skb(). Other than just making sense, this makes the lock handling easier. * Change the audit_log_start() sleep behavior so that we never sleep on auditd events (unchanged) or if the caller is holding the audit_cmd_mutex (changed). Previously we didn't sleep if the caller was auditd or if the message type fell between a certain range; the type check was a poor effort of doing what the cmd_mutex check now does. Richard Guy Briggs originally proposed not sleeping the cmd_mutex owner several years ago but his patch wasn't acceptable at the time. At least the idea lives on here. * A problem with the lost record counter has been resolved. Steve Grubb and I both happened to notice this problem and according to some quick testing by Steve, this problem goes back quite some time. It's largely a harmless problem, although it may have left some careful sysadmins quite puzzled. Cc: <stable@vger.kernel.org> # 4.10.x- Reported-by: Dmitry Vyukov <dvyukov@google.com> Signed-off-by: Paul Moore <paul@paul-moore.com>
2017-03-21 23:26:35 +08:00
if (nlh->nlmsg_type != AUDIT_EOE && printk_ratelimit())
pr_notice("type=%d %s\n", nlh->nlmsg_type, data);
}
/**
* kauditd_rehold_skb - Handle a audit record send failure in the hold queue
* @skb: audit record
audit: improve audit queue handling when "audit=1" on cmdline When an admin enables audit at early boot via the "audit=1" kernel command line the audit queue behavior is slightly different; the audit subsystem goes to greater lengths to avoid dropping records, which unfortunately can result in problems when the audit daemon is forcibly stopped for an extended period of time. This patch makes a number of changes designed to improve the audit queuing behavior so that leaving the audit daemon in a stopped state for an extended period does not cause a significant impact to the system. - kauditd_send_queue() is now limited to looping through the passed queue only once per call. This not only prevents the function from looping indefinitely when records are returned to the current queue, it also allows any recovery handling in kauditd_thread() to take place when kauditd_send_queue() returns. - Transient netlink send errors seen as -EAGAIN now cause the record to be returned to the retry queue instead of going to the hold queue. The intention of the hold queue is to store, perhaps for an extended period of time, the events which led up to the audit daemon going offline. The retry queue remains a temporary queue intended to protect against transient issues between the kernel and the audit daemon. - The retry queue is now limited by the audit_backlog_limit setting, the same as the other queues. This allows admins to bound the size of all of the audit queues on the system. - kauditd_rehold_skb() now returns records to the end of the hold queue to ensure ordering is preserved in the face of recent changes to kauditd_send_queue(). Cc: stable@vger.kernel.org Fixes: 5b52330bbfe63 ("audit: fix auditd/kernel connection state tracking") Fixes: f4b3ee3c85551 ("audit: improve robustness of the audit queue handling") Reported-by: Gaosheng Cui <cuigaosheng1@huawei.com> Tested-by: Gaosheng Cui <cuigaosheng1@huawei.com> Reviewed-by: Richard Guy Briggs <rgb@redhat.com> Signed-off-by: Paul Moore <paul@paul-moore.com>
2022-01-14 07:54:38 +08:00
* @error: error code (unused)
audit: fix auditd/kernel connection state tracking What started as a rather straightforward race condition reported by Dmitry using the syzkaller fuzzer ended up revealing some major problems with how the audit subsystem managed its netlink sockets and its connection with the userspace audit daemon. Fixing this properly had quite the cascading effect and what we are left with is this rather large and complicated patch. My initial goal was to try and decompose this patch into multiple smaller patches, but the way these changes are intertwined makes it difficult to split these changes into meaningful pieces that don't break or somehow make things worse for the intermediate states. The patch makes a number of changes, but the most significant are highlighted below: * The auditd tracking variables, e.g. audit_sock, are now gone and replaced by a RCU/spin_lock protected variable auditd_conn which is a structure containing all of the auditd tracking information. * We no longer track the auditd sock directly, instead we track it via the network namespace in which it resides and we use the audit socket associated with that namespace. In spirit, this is what the code was trying to do prior to this patch (at least I think that is what the original authors intended), but it was done rather poorly and added a layer of obfuscation that only masked the underlying problems. * Big backlog queue cleanup, again. In v4.10 we made some pretty big changes to how the audit backlog queues work, here we haven't changed the queue design so much as cleaned up the implementation. Brought about by the locking changes, we've simplified kauditd_thread() quite a bit by consolidating the queue handling into a new helper function, kauditd_send_queue(), which allows us to eliminate a lot of very similar code and makes the looping logic in kauditd_thread() clearer. * All netlink messages sent to auditd are now sent via auditd_send_unicast_skb(). Other than just making sense, this makes the lock handling easier. * Change the audit_log_start() sleep behavior so that we never sleep on auditd events (unchanged) or if the caller is holding the audit_cmd_mutex (changed). Previously we didn't sleep if the caller was auditd or if the message type fell between a certain range; the type check was a poor effort of doing what the cmd_mutex check now does. Richard Guy Briggs originally proposed not sleeping the cmd_mutex owner several years ago but his patch wasn't acceptable at the time. At least the idea lives on here. * A problem with the lost record counter has been resolved. Steve Grubb and I both happened to notice this problem and according to some quick testing by Steve, this problem goes back quite some time. It's largely a harmless problem, although it may have left some careful sysadmins quite puzzled. Cc: <stable@vger.kernel.org> # 4.10.x- Reported-by: Dmitry Vyukov <dvyukov@google.com> Signed-off-by: Paul Moore <paul@paul-moore.com>
2017-03-21 23:26:35 +08:00
*
* Description:
* This should only be used by the kauditd_thread when it fails to flush the
* hold queue.
*/
audit: improve audit queue handling when "audit=1" on cmdline When an admin enables audit at early boot via the "audit=1" kernel command line the audit queue behavior is slightly different; the audit subsystem goes to greater lengths to avoid dropping records, which unfortunately can result in problems when the audit daemon is forcibly stopped for an extended period of time. This patch makes a number of changes designed to improve the audit queuing behavior so that leaving the audit daemon in a stopped state for an extended period does not cause a significant impact to the system. - kauditd_send_queue() is now limited to looping through the passed queue only once per call. This not only prevents the function from looping indefinitely when records are returned to the current queue, it also allows any recovery handling in kauditd_thread() to take place when kauditd_send_queue() returns. - Transient netlink send errors seen as -EAGAIN now cause the record to be returned to the retry queue instead of going to the hold queue. The intention of the hold queue is to store, perhaps for an extended period of time, the events which led up to the audit daemon going offline. The retry queue remains a temporary queue intended to protect against transient issues between the kernel and the audit daemon. - The retry queue is now limited by the audit_backlog_limit setting, the same as the other queues. This allows admins to bound the size of all of the audit queues on the system. - kauditd_rehold_skb() now returns records to the end of the hold queue to ensure ordering is preserved in the face of recent changes to kauditd_send_queue(). Cc: stable@vger.kernel.org Fixes: 5b52330bbfe63 ("audit: fix auditd/kernel connection state tracking") Fixes: f4b3ee3c85551 ("audit: improve robustness of the audit queue handling") Reported-by: Gaosheng Cui <cuigaosheng1@huawei.com> Tested-by: Gaosheng Cui <cuigaosheng1@huawei.com> Reviewed-by: Richard Guy Briggs <rgb@redhat.com> Signed-off-by: Paul Moore <paul@paul-moore.com>
2022-01-14 07:54:38 +08:00
static void kauditd_rehold_skb(struct sk_buff *skb, __always_unused int error)
audit: fix auditd/kernel connection state tracking What started as a rather straightforward race condition reported by Dmitry using the syzkaller fuzzer ended up revealing some major problems with how the audit subsystem managed its netlink sockets and its connection with the userspace audit daemon. Fixing this properly had quite the cascading effect and what we are left with is this rather large and complicated patch. My initial goal was to try and decompose this patch into multiple smaller patches, but the way these changes are intertwined makes it difficult to split these changes into meaningful pieces that don't break or somehow make things worse for the intermediate states. The patch makes a number of changes, but the most significant are highlighted below: * The auditd tracking variables, e.g. audit_sock, are now gone and replaced by a RCU/spin_lock protected variable auditd_conn which is a structure containing all of the auditd tracking information. * We no longer track the auditd sock directly, instead we track it via the network namespace in which it resides and we use the audit socket associated with that namespace. In spirit, this is what the code was trying to do prior to this patch (at least I think that is what the original authors intended), but it was done rather poorly and added a layer of obfuscation that only masked the underlying problems. * Big backlog queue cleanup, again. In v4.10 we made some pretty big changes to how the audit backlog queues work, here we haven't changed the queue design so much as cleaned up the implementation. Brought about by the locking changes, we've simplified kauditd_thread() quite a bit by consolidating the queue handling into a new helper function, kauditd_send_queue(), which allows us to eliminate a lot of very similar code and makes the looping logic in kauditd_thread() clearer. * All netlink messages sent to auditd are now sent via auditd_send_unicast_skb(). Other than just making sense, this makes the lock handling easier. * Change the audit_log_start() sleep behavior so that we never sleep on auditd events (unchanged) or if the caller is holding the audit_cmd_mutex (changed). Previously we didn't sleep if the caller was auditd or if the message type fell between a certain range; the type check was a poor effort of doing what the cmd_mutex check now does. Richard Guy Briggs originally proposed not sleeping the cmd_mutex owner several years ago but his patch wasn't acceptable at the time. At least the idea lives on here. * A problem with the lost record counter has been resolved. Steve Grubb and I both happened to notice this problem and according to some quick testing by Steve, this problem goes back quite some time. It's largely a harmless problem, although it may have left some careful sysadmins quite puzzled. Cc: <stable@vger.kernel.org> # 4.10.x- Reported-by: Dmitry Vyukov <dvyukov@google.com> Signed-off-by: Paul Moore <paul@paul-moore.com>
2017-03-21 23:26:35 +08:00
{
audit: improve audit queue handling when "audit=1" on cmdline When an admin enables audit at early boot via the "audit=1" kernel command line the audit queue behavior is slightly different; the audit subsystem goes to greater lengths to avoid dropping records, which unfortunately can result in problems when the audit daemon is forcibly stopped for an extended period of time. This patch makes a number of changes designed to improve the audit queuing behavior so that leaving the audit daemon in a stopped state for an extended period does not cause a significant impact to the system. - kauditd_send_queue() is now limited to looping through the passed queue only once per call. This not only prevents the function from looping indefinitely when records are returned to the current queue, it also allows any recovery handling in kauditd_thread() to take place when kauditd_send_queue() returns. - Transient netlink send errors seen as -EAGAIN now cause the record to be returned to the retry queue instead of going to the hold queue. The intention of the hold queue is to store, perhaps for an extended period of time, the events which led up to the audit daemon going offline. The retry queue remains a temporary queue intended to protect against transient issues between the kernel and the audit daemon. - The retry queue is now limited by the audit_backlog_limit setting, the same as the other queues. This allows admins to bound the size of all of the audit queues on the system. - kauditd_rehold_skb() now returns records to the end of the hold queue to ensure ordering is preserved in the face of recent changes to kauditd_send_queue(). Cc: stable@vger.kernel.org Fixes: 5b52330bbfe63 ("audit: fix auditd/kernel connection state tracking") Fixes: f4b3ee3c85551 ("audit: improve robustness of the audit queue handling") Reported-by: Gaosheng Cui <cuigaosheng1@huawei.com> Tested-by: Gaosheng Cui <cuigaosheng1@huawei.com> Reviewed-by: Richard Guy Briggs <rgb@redhat.com> Signed-off-by: Paul Moore <paul@paul-moore.com>
2022-01-14 07:54:38 +08:00
/* put the record back in the queue */
skb_queue_tail(&audit_hold_queue, skb);
}
/**
* kauditd_hold_skb - Queue an audit record, waiting for auditd
* @skb: audit record
audit: improve audit queue handling when "audit=1" on cmdline When an admin enables audit at early boot via the "audit=1" kernel command line the audit queue behavior is slightly different; the audit subsystem goes to greater lengths to avoid dropping records, which unfortunately can result in problems when the audit daemon is forcibly stopped for an extended period of time. This patch makes a number of changes designed to improve the audit queuing behavior so that leaving the audit daemon in a stopped state for an extended period does not cause a significant impact to the system. - kauditd_send_queue() is now limited to looping through the passed queue only once per call. This not only prevents the function from looping indefinitely when records are returned to the current queue, it also allows any recovery handling in kauditd_thread() to take place when kauditd_send_queue() returns. - Transient netlink send errors seen as -EAGAIN now cause the record to be returned to the retry queue instead of going to the hold queue. The intention of the hold queue is to store, perhaps for an extended period of time, the events which led up to the audit daemon going offline. The retry queue remains a temporary queue intended to protect against transient issues between the kernel and the audit daemon. - The retry queue is now limited by the audit_backlog_limit setting, the same as the other queues. This allows admins to bound the size of all of the audit queues on the system. - kauditd_rehold_skb() now returns records to the end of the hold queue to ensure ordering is preserved in the face of recent changes to kauditd_send_queue(). Cc: stable@vger.kernel.org Fixes: 5b52330bbfe63 ("audit: fix auditd/kernel connection state tracking") Fixes: f4b3ee3c85551 ("audit: improve robustness of the audit queue handling") Reported-by: Gaosheng Cui <cuigaosheng1@huawei.com> Tested-by: Gaosheng Cui <cuigaosheng1@huawei.com> Reviewed-by: Richard Guy Briggs <rgb@redhat.com> Signed-off-by: Paul Moore <paul@paul-moore.com>
2022-01-14 07:54:38 +08:00
* @error: error code
*
* Description:
* Queue the audit record, waiting for an instance of auditd. When this
* function is called we haven't given up yet on sending the record, but things
* are not looking good. The first thing we want to do is try to write the
* record via printk and then see if we want to try and hold on to the record
* and queue it, if we have room. If we want to hold on to the record, but we
* don't have room, record a record lost message.
*/
audit: improve audit queue handling when "audit=1" on cmdline When an admin enables audit at early boot via the "audit=1" kernel command line the audit queue behavior is slightly different; the audit subsystem goes to greater lengths to avoid dropping records, which unfortunately can result in problems when the audit daemon is forcibly stopped for an extended period of time. This patch makes a number of changes designed to improve the audit queuing behavior so that leaving the audit daemon in a stopped state for an extended period does not cause a significant impact to the system. - kauditd_send_queue() is now limited to looping through the passed queue only once per call. This not only prevents the function from looping indefinitely when records are returned to the current queue, it also allows any recovery handling in kauditd_thread() to take place when kauditd_send_queue() returns. - Transient netlink send errors seen as -EAGAIN now cause the record to be returned to the retry queue instead of going to the hold queue. The intention of the hold queue is to store, perhaps for an extended period of time, the events which led up to the audit daemon going offline. The retry queue remains a temporary queue intended to protect against transient issues between the kernel and the audit daemon. - The retry queue is now limited by the audit_backlog_limit setting, the same as the other queues. This allows admins to bound the size of all of the audit queues on the system. - kauditd_rehold_skb() now returns records to the end of the hold queue to ensure ordering is preserved in the face of recent changes to kauditd_send_queue(). Cc: stable@vger.kernel.org Fixes: 5b52330bbfe63 ("audit: fix auditd/kernel connection state tracking") Fixes: f4b3ee3c85551 ("audit: improve robustness of the audit queue handling") Reported-by: Gaosheng Cui <cuigaosheng1@huawei.com> Tested-by: Gaosheng Cui <cuigaosheng1@huawei.com> Reviewed-by: Richard Guy Briggs <rgb@redhat.com> Signed-off-by: Paul Moore <paul@paul-moore.com>
2022-01-14 07:54:38 +08:00
static void kauditd_hold_skb(struct sk_buff *skb, int error)
{
/* at this point it is uncertain if we will ever send this to auditd so
* try to send the message via printk before we go any further */
kauditd_printk_skb(skb);
/* can we just silently drop the message? */
audit: improve audit queue handling when "audit=1" on cmdline When an admin enables audit at early boot via the "audit=1" kernel command line the audit queue behavior is slightly different; the audit subsystem goes to greater lengths to avoid dropping records, which unfortunately can result in problems when the audit daemon is forcibly stopped for an extended period of time. This patch makes a number of changes designed to improve the audit queuing behavior so that leaving the audit daemon in a stopped state for an extended period does not cause a significant impact to the system. - kauditd_send_queue() is now limited to looping through the passed queue only once per call. This not only prevents the function from looping indefinitely when records are returned to the current queue, it also allows any recovery handling in kauditd_thread() to take place when kauditd_send_queue() returns. - Transient netlink send errors seen as -EAGAIN now cause the record to be returned to the retry queue instead of going to the hold queue. The intention of the hold queue is to store, perhaps for an extended period of time, the events which led up to the audit daemon going offline. The retry queue remains a temporary queue intended to protect against transient issues between the kernel and the audit daemon. - The retry queue is now limited by the audit_backlog_limit setting, the same as the other queues. This allows admins to bound the size of all of the audit queues on the system. - kauditd_rehold_skb() now returns records to the end of the hold queue to ensure ordering is preserved in the face of recent changes to kauditd_send_queue(). Cc: stable@vger.kernel.org Fixes: 5b52330bbfe63 ("audit: fix auditd/kernel connection state tracking") Fixes: f4b3ee3c85551 ("audit: improve robustness of the audit queue handling") Reported-by: Gaosheng Cui <cuigaosheng1@huawei.com> Tested-by: Gaosheng Cui <cuigaosheng1@huawei.com> Reviewed-by: Richard Guy Briggs <rgb@redhat.com> Signed-off-by: Paul Moore <paul@paul-moore.com>
2022-01-14 07:54:38 +08:00
if (!audit_default)
goto drop;
/* the hold queue is only for when the daemon goes away completely,
* not -EAGAIN failures; if we are in a -EAGAIN state requeue the
* record on the retry queue unless it's full, in which case drop it
*/
if (error == -EAGAIN) {
if (!audit_backlog_limit ||
skb_queue_len(&audit_retry_queue) < audit_backlog_limit) {
skb_queue_tail(&audit_retry_queue, skb);
return;
}
audit_log_lost("kauditd retry queue overflow");
goto drop;
}
audit: improve audit queue handling when "audit=1" on cmdline When an admin enables audit at early boot via the "audit=1" kernel command line the audit queue behavior is slightly different; the audit subsystem goes to greater lengths to avoid dropping records, which unfortunately can result in problems when the audit daemon is forcibly stopped for an extended period of time. This patch makes a number of changes designed to improve the audit queuing behavior so that leaving the audit daemon in a stopped state for an extended period does not cause a significant impact to the system. - kauditd_send_queue() is now limited to looping through the passed queue only once per call. This not only prevents the function from looping indefinitely when records are returned to the current queue, it also allows any recovery handling in kauditd_thread() to take place when kauditd_send_queue() returns. - Transient netlink send errors seen as -EAGAIN now cause the record to be returned to the retry queue instead of going to the hold queue. The intention of the hold queue is to store, perhaps for an extended period of time, the events which led up to the audit daemon going offline. The retry queue remains a temporary queue intended to protect against transient issues between the kernel and the audit daemon. - The retry queue is now limited by the audit_backlog_limit setting, the same as the other queues. This allows admins to bound the size of all of the audit queues on the system. - kauditd_rehold_skb() now returns records to the end of the hold queue to ensure ordering is preserved in the face of recent changes to kauditd_send_queue(). Cc: stable@vger.kernel.org Fixes: 5b52330bbfe63 ("audit: fix auditd/kernel connection state tracking") Fixes: f4b3ee3c85551 ("audit: improve robustness of the audit queue handling") Reported-by: Gaosheng Cui <cuigaosheng1@huawei.com> Tested-by: Gaosheng Cui <cuigaosheng1@huawei.com> Reviewed-by: Richard Guy Briggs <rgb@redhat.com> Signed-off-by: Paul Moore <paul@paul-moore.com>
2022-01-14 07:54:38 +08:00
/* if we have room in the hold queue, queue the message */
if (!audit_backlog_limit ||
skb_queue_len(&audit_hold_queue) < audit_backlog_limit) {
skb_queue_tail(&audit_hold_queue, skb);
return;
}
/* we have no other options - drop the message */
audit_log_lost("kauditd hold queue overflow");
audit: improve audit queue handling when "audit=1" on cmdline When an admin enables audit at early boot via the "audit=1" kernel command line the audit queue behavior is slightly different; the audit subsystem goes to greater lengths to avoid dropping records, which unfortunately can result in problems when the audit daemon is forcibly stopped for an extended period of time. This patch makes a number of changes designed to improve the audit queuing behavior so that leaving the audit daemon in a stopped state for an extended period does not cause a significant impact to the system. - kauditd_send_queue() is now limited to looping through the passed queue only once per call. This not only prevents the function from looping indefinitely when records are returned to the current queue, it also allows any recovery handling in kauditd_thread() to take place when kauditd_send_queue() returns. - Transient netlink send errors seen as -EAGAIN now cause the record to be returned to the retry queue instead of going to the hold queue. The intention of the hold queue is to store, perhaps for an extended period of time, the events which led up to the audit daemon going offline. The retry queue remains a temporary queue intended to protect against transient issues between the kernel and the audit daemon. - The retry queue is now limited by the audit_backlog_limit setting, the same as the other queues. This allows admins to bound the size of all of the audit queues on the system. - kauditd_rehold_skb() now returns records to the end of the hold queue to ensure ordering is preserved in the face of recent changes to kauditd_send_queue(). Cc: stable@vger.kernel.org Fixes: 5b52330bbfe63 ("audit: fix auditd/kernel connection state tracking") Fixes: f4b3ee3c85551 ("audit: improve robustness of the audit queue handling") Reported-by: Gaosheng Cui <cuigaosheng1@huawei.com> Tested-by: Gaosheng Cui <cuigaosheng1@huawei.com> Reviewed-by: Richard Guy Briggs <rgb@redhat.com> Signed-off-by: Paul Moore <paul@paul-moore.com>
2022-01-14 07:54:38 +08:00
drop:
kfree_skb(skb);
}
/**
* kauditd_retry_skb - Queue an audit record, attempt to send again to auditd
* @skb: audit record
audit: improve audit queue handling when "audit=1" on cmdline When an admin enables audit at early boot via the "audit=1" kernel command line the audit queue behavior is slightly different; the audit subsystem goes to greater lengths to avoid dropping records, which unfortunately can result in problems when the audit daemon is forcibly stopped for an extended period of time. This patch makes a number of changes designed to improve the audit queuing behavior so that leaving the audit daemon in a stopped state for an extended period does not cause a significant impact to the system. - kauditd_send_queue() is now limited to looping through the passed queue only once per call. This not only prevents the function from looping indefinitely when records are returned to the current queue, it also allows any recovery handling in kauditd_thread() to take place when kauditd_send_queue() returns. - Transient netlink send errors seen as -EAGAIN now cause the record to be returned to the retry queue instead of going to the hold queue. The intention of the hold queue is to store, perhaps for an extended period of time, the events which led up to the audit daemon going offline. The retry queue remains a temporary queue intended to protect against transient issues between the kernel and the audit daemon. - The retry queue is now limited by the audit_backlog_limit setting, the same as the other queues. This allows admins to bound the size of all of the audit queues on the system. - kauditd_rehold_skb() now returns records to the end of the hold queue to ensure ordering is preserved in the face of recent changes to kauditd_send_queue(). Cc: stable@vger.kernel.org Fixes: 5b52330bbfe63 ("audit: fix auditd/kernel connection state tracking") Fixes: f4b3ee3c85551 ("audit: improve robustness of the audit queue handling") Reported-by: Gaosheng Cui <cuigaosheng1@huawei.com> Tested-by: Gaosheng Cui <cuigaosheng1@huawei.com> Reviewed-by: Richard Guy Briggs <rgb@redhat.com> Signed-off-by: Paul Moore <paul@paul-moore.com>
2022-01-14 07:54:38 +08:00
* @error: error code (unused)
*
* Description:
* Not as serious as kauditd_hold_skb() as we still have a connected auditd,
* but for some reason we are having problems sending it audit records so
* queue the given record and attempt to resend.
*/
audit: improve audit queue handling when "audit=1" on cmdline When an admin enables audit at early boot via the "audit=1" kernel command line the audit queue behavior is slightly different; the audit subsystem goes to greater lengths to avoid dropping records, which unfortunately can result in problems when the audit daemon is forcibly stopped for an extended period of time. This patch makes a number of changes designed to improve the audit queuing behavior so that leaving the audit daemon in a stopped state for an extended period does not cause a significant impact to the system. - kauditd_send_queue() is now limited to looping through the passed queue only once per call. This not only prevents the function from looping indefinitely when records are returned to the current queue, it also allows any recovery handling in kauditd_thread() to take place when kauditd_send_queue() returns. - Transient netlink send errors seen as -EAGAIN now cause the record to be returned to the retry queue instead of going to the hold queue. The intention of the hold queue is to store, perhaps for an extended period of time, the events which led up to the audit daemon going offline. The retry queue remains a temporary queue intended to protect against transient issues between the kernel and the audit daemon. - The retry queue is now limited by the audit_backlog_limit setting, the same as the other queues. This allows admins to bound the size of all of the audit queues on the system. - kauditd_rehold_skb() now returns records to the end of the hold queue to ensure ordering is preserved in the face of recent changes to kauditd_send_queue(). Cc: stable@vger.kernel.org Fixes: 5b52330bbfe63 ("audit: fix auditd/kernel connection state tracking") Fixes: f4b3ee3c85551 ("audit: improve robustness of the audit queue handling") Reported-by: Gaosheng Cui <cuigaosheng1@huawei.com> Tested-by: Gaosheng Cui <cuigaosheng1@huawei.com> Reviewed-by: Richard Guy Briggs <rgb@redhat.com> Signed-off-by: Paul Moore <paul@paul-moore.com>
2022-01-14 07:54:38 +08:00
static void kauditd_retry_skb(struct sk_buff *skb, __always_unused int error)
{
audit: improve audit queue handling when "audit=1" on cmdline When an admin enables audit at early boot via the "audit=1" kernel command line the audit queue behavior is slightly different; the audit subsystem goes to greater lengths to avoid dropping records, which unfortunately can result in problems when the audit daemon is forcibly stopped for an extended period of time. This patch makes a number of changes designed to improve the audit queuing behavior so that leaving the audit daemon in a stopped state for an extended period does not cause a significant impact to the system. - kauditd_send_queue() is now limited to looping through the passed queue only once per call. This not only prevents the function from looping indefinitely when records are returned to the current queue, it also allows any recovery handling in kauditd_thread() to take place when kauditd_send_queue() returns. - Transient netlink send errors seen as -EAGAIN now cause the record to be returned to the retry queue instead of going to the hold queue. The intention of the hold queue is to store, perhaps for an extended period of time, the events which led up to the audit daemon going offline. The retry queue remains a temporary queue intended to protect against transient issues between the kernel and the audit daemon. - The retry queue is now limited by the audit_backlog_limit setting, the same as the other queues. This allows admins to bound the size of all of the audit queues on the system. - kauditd_rehold_skb() now returns records to the end of the hold queue to ensure ordering is preserved in the face of recent changes to kauditd_send_queue(). Cc: stable@vger.kernel.org Fixes: 5b52330bbfe63 ("audit: fix auditd/kernel connection state tracking") Fixes: f4b3ee3c85551 ("audit: improve robustness of the audit queue handling") Reported-by: Gaosheng Cui <cuigaosheng1@huawei.com> Tested-by: Gaosheng Cui <cuigaosheng1@huawei.com> Reviewed-by: Richard Guy Briggs <rgb@redhat.com> Signed-off-by: Paul Moore <paul@paul-moore.com>
2022-01-14 07:54:38 +08:00
if (!audit_backlog_limit ||
skb_queue_len(&audit_retry_queue) < audit_backlog_limit) {
skb_queue_tail(&audit_retry_queue, skb);
return;
}
/* we have to drop the record, send it via printk as a last effort */
kauditd_printk_skb(skb);
audit_log_lost("kauditd retry queue overflow");
kfree_skb(skb);
}
audit: try harder to send to auditd upon netlink failure There are several reports of the kernel losing contact with auditd when it is, in fact, still running. When this happens, kernel syslogs show: "audit: *NO* daemon at audit_pid=<pid>" although auditd is still running, and is apparently happy, listening on the netlink socket. The pid in the "*NO* daemon" message matches the pid of the running auditd process. Restarting auditd solves this. The problem appears to happen randomly, and doesn't seem to be strongly correlated to the rate of audit events being logged. The problem happens fairly regularly (every few days), but not yet reproduced to order. On production kernels, BUG_ON() is a no-op, so any error will trigger this. Commit 34eab0a7cd45 ("audit: prevent an older auditd shutdown from orphaning a newer auditd startup") eliminates one possible cause. This isn't the case here, since the PID in the error message and the PID of the running auditd match. The primary expected cause of error here is -ECONNREFUSED when the audit daemon goes away, when netlink_getsockbyportid() can't find the auditd portid entry in the netlink audit table (or there is no receive function). If -EPERM is returned, that situation isn't likely to be resolved in a timely fashion without administrator intervention. In both cases, reset the audit_pid. This does not rule out a race condition. SELinux is expected to return zero since this isn't an INET or INET6 socket. Other LSMs may have other return codes. Log the error code for better diagnosis in the future. In the case of -ENOMEM, the situation could be temporary, based on local or general availability of buffers. -EAGAIN should never happen since the netlink audit (kernel) socket is set to MAX_SCHEDULE_TIMEOUT. -ERESTARTSYS and -EINTR are not expected since this kernel thread is not expected to receive signals. In these cases (or any other unexpected ones for now), report the error and re-schedule the thread, retrying up to 5 times. v2: Removed BUG_ON(). Moved comma in pr_*() statements. Removed audit_strerror() text. Reported-by: Vipin Rathor <v.rathor@gmail.com> Reported-by: <ctcard@hotmail.com> Signed-off-by: Richard Guy Briggs <rgb@redhat.com> [PM: applied rgb's fixup patch to correct audit_log_lost() format issues] Signed-off-by: Paul Moore <pmoore@redhat.com>
2015-11-04 21:23:50 +08:00
/**
* auditd_reset - Disconnect the auditd connection
audit: fix a race condition with the auditd tracking code Originally reported by Adam and Dusty, it appears we have a small race window in kauditd_thread(), as documented in the Fedora BZ: * https://bugzilla.redhat.com/show_bug.cgi?id=1459326#c35 "This issue is partly due to the read-copy nature of RCU, and partly due to how we sync the auditd_connection state across kauditd_thread and the audit control channel. The kauditd_thread thread is always running so it can service the record queues and emit the multicast messages, if it happens to be just past the "main_queue" label, but before the "if (sk == NULL || ...)" if-statement which calls auditd_reset() when the new auditd connection is registered it could end up resetting the auditd connection, regardless of if it is valid or not. This is a rather small window and the variable nature of multi-core scheduling explains why this is proving rather difficult to reproduce." The fix is to have functions only call auditd_reset() when they believe that the kernel/auditd connection is still valid, e.g. non-NULL, and to have these callers pass their local copy of the auditd_connection pointer to auditd_reset() where it can be compared with the current connection state before resetting. If the caller has a stale state tracking pointer then the reset is ignored. We also make a small change to kauditd_thread() so that if the kernel/auditd connection is dead we skip the retry queue and send the records straight to the hold queue. This is necessary as we used to rely on auditd_reset() to occasionally purge the retry queue but we are going to be calling the reset function much less now and we want to make sure the retry queue doesn't grow unbounded. Reported-by: Adam Williamson <awilliam@redhat.com> Reported-by: Dusty Mabe <dustymabe@redhat.com> Reviewed-by: Richard Guy Briggs <rgb@redhat.com> Signed-off-by: Paul Moore <paul@paul-moore.com>
2017-06-12 21:35:24 +08:00
* @ac: auditd connection state
*
* Description:
* Break the auditd/kauditd connection and move all the queued records into the
audit: fix a race condition with the auditd tracking code Originally reported by Adam and Dusty, it appears we have a small race window in kauditd_thread(), as documented in the Fedora BZ: * https://bugzilla.redhat.com/show_bug.cgi?id=1459326#c35 "This issue is partly due to the read-copy nature of RCU, and partly due to how we sync the auditd_connection state across kauditd_thread and the audit control channel. The kauditd_thread thread is always running so it can service the record queues and emit the multicast messages, if it happens to be just past the "main_queue" label, but before the "if (sk == NULL || ...)" if-statement which calls auditd_reset() when the new auditd connection is registered it could end up resetting the auditd connection, regardless of if it is valid or not. This is a rather small window and the variable nature of multi-core scheduling explains why this is proving rather difficult to reproduce." The fix is to have functions only call auditd_reset() when they believe that the kernel/auditd connection is still valid, e.g. non-NULL, and to have these callers pass their local copy of the auditd_connection pointer to auditd_reset() where it can be compared with the current connection state before resetting. If the caller has a stale state tracking pointer then the reset is ignored. We also make a small change to kauditd_thread() so that if the kernel/auditd connection is dead we skip the retry queue and send the records straight to the hold queue. This is necessary as we used to rely on auditd_reset() to occasionally purge the retry queue but we are going to be calling the reset function much less now and we want to make sure the retry queue doesn't grow unbounded. Reported-by: Adam Williamson <awilliam@redhat.com> Reported-by: Dusty Mabe <dustymabe@redhat.com> Reviewed-by: Richard Guy Briggs <rgb@redhat.com> Signed-off-by: Paul Moore <paul@paul-moore.com>
2017-06-12 21:35:24 +08:00
* hold queue in case auditd reconnects. It is important to note that the @ac
* pointer should never be dereferenced inside this function as it may be NULL
* or invalid, you can only compare the memory address! If @ac is NULL then
* the connection will always be reset.
*/
audit: fix a race condition with the auditd tracking code Originally reported by Adam and Dusty, it appears we have a small race window in kauditd_thread(), as documented in the Fedora BZ: * https://bugzilla.redhat.com/show_bug.cgi?id=1459326#c35 "This issue is partly due to the read-copy nature of RCU, and partly due to how we sync the auditd_connection state across kauditd_thread and the audit control channel. The kauditd_thread thread is always running so it can service the record queues and emit the multicast messages, if it happens to be just past the "main_queue" label, but before the "if (sk == NULL || ...)" if-statement which calls auditd_reset() when the new auditd connection is registered it could end up resetting the auditd connection, regardless of if it is valid or not. This is a rather small window and the variable nature of multi-core scheduling explains why this is proving rather difficult to reproduce." The fix is to have functions only call auditd_reset() when they believe that the kernel/auditd connection is still valid, e.g. non-NULL, and to have these callers pass their local copy of the auditd_connection pointer to auditd_reset() where it can be compared with the current connection state before resetting. If the caller has a stale state tracking pointer then the reset is ignored. We also make a small change to kauditd_thread() so that if the kernel/auditd connection is dead we skip the retry queue and send the records straight to the hold queue. This is necessary as we used to rely on auditd_reset() to occasionally purge the retry queue but we are going to be calling the reset function much less now and we want to make sure the retry queue doesn't grow unbounded. Reported-by: Adam Williamson <awilliam@redhat.com> Reported-by: Dusty Mabe <dustymabe@redhat.com> Reviewed-by: Richard Guy Briggs <rgb@redhat.com> Signed-off-by: Paul Moore <paul@paul-moore.com>
2017-06-12 21:35:24 +08:00
static void auditd_reset(const struct auditd_connection *ac)
{
unsigned long flags;
struct sk_buff *skb;
struct auditd_connection *ac_old;
/* if it isn't already broken, break the connection */
spin_lock_irqsave(&auditd_conn_lock, flags);
ac_old = rcu_dereference_protected(auditd_conn,
lockdep_is_held(&auditd_conn_lock));
audit: fix a race condition with the auditd tracking code Originally reported by Adam and Dusty, it appears we have a small race window in kauditd_thread(), as documented in the Fedora BZ: * https://bugzilla.redhat.com/show_bug.cgi?id=1459326#c35 "This issue is partly due to the read-copy nature of RCU, and partly due to how we sync the auditd_connection state across kauditd_thread and the audit control channel. The kauditd_thread thread is always running so it can service the record queues and emit the multicast messages, if it happens to be just past the "main_queue" label, but before the "if (sk == NULL || ...)" if-statement which calls auditd_reset() when the new auditd connection is registered it could end up resetting the auditd connection, regardless of if it is valid or not. This is a rather small window and the variable nature of multi-core scheduling explains why this is proving rather difficult to reproduce." The fix is to have functions only call auditd_reset() when they believe that the kernel/auditd connection is still valid, e.g. non-NULL, and to have these callers pass their local copy of the auditd_connection pointer to auditd_reset() where it can be compared with the current connection state before resetting. If the caller has a stale state tracking pointer then the reset is ignored. We also make a small change to kauditd_thread() so that if the kernel/auditd connection is dead we skip the retry queue and send the records straight to the hold queue. This is necessary as we used to rely on auditd_reset() to occasionally purge the retry queue but we are going to be calling the reset function much less now and we want to make sure the retry queue doesn't grow unbounded. Reported-by: Adam Williamson <awilliam@redhat.com> Reported-by: Dusty Mabe <dustymabe@redhat.com> Reviewed-by: Richard Guy Briggs <rgb@redhat.com> Signed-off-by: Paul Moore <paul@paul-moore.com>
2017-06-12 21:35:24 +08:00
if (ac && ac != ac_old) {
/* someone already registered a new auditd connection */
spin_unlock_irqrestore(&auditd_conn_lock, flags);
return;
}
rcu_assign_pointer(auditd_conn, NULL);
spin_unlock_irqrestore(&auditd_conn_lock, flags);
if (ac_old)
call_rcu(&ac_old->rcu, auditd_conn_free);
/* flush the retry queue to the hold queue, but don't touch the main
* queue since we need to process that normally for multicast */
while ((skb = skb_dequeue(&audit_retry_queue)))
audit: improve audit queue handling when "audit=1" on cmdline When an admin enables audit at early boot via the "audit=1" kernel command line the audit queue behavior is slightly different; the audit subsystem goes to greater lengths to avoid dropping records, which unfortunately can result in problems when the audit daemon is forcibly stopped for an extended period of time. This patch makes a number of changes designed to improve the audit queuing behavior so that leaving the audit daemon in a stopped state for an extended period does not cause a significant impact to the system. - kauditd_send_queue() is now limited to looping through the passed queue only once per call. This not only prevents the function from looping indefinitely when records are returned to the current queue, it also allows any recovery handling in kauditd_thread() to take place when kauditd_send_queue() returns. - Transient netlink send errors seen as -EAGAIN now cause the record to be returned to the retry queue instead of going to the hold queue. The intention of the hold queue is to store, perhaps for an extended period of time, the events which led up to the audit daemon going offline. The retry queue remains a temporary queue intended to protect against transient issues between the kernel and the audit daemon. - The retry queue is now limited by the audit_backlog_limit setting, the same as the other queues. This allows admins to bound the size of all of the audit queues on the system. - kauditd_rehold_skb() now returns records to the end of the hold queue to ensure ordering is preserved in the face of recent changes to kauditd_send_queue(). Cc: stable@vger.kernel.org Fixes: 5b52330bbfe63 ("audit: fix auditd/kernel connection state tracking") Fixes: f4b3ee3c85551 ("audit: improve robustness of the audit queue handling") Reported-by: Gaosheng Cui <cuigaosheng1@huawei.com> Tested-by: Gaosheng Cui <cuigaosheng1@huawei.com> Reviewed-by: Richard Guy Briggs <rgb@redhat.com> Signed-off-by: Paul Moore <paul@paul-moore.com>
2022-01-14 07:54:38 +08:00
kauditd_hold_skb(skb, -ECONNREFUSED);
}
/**
audit: fix auditd/kernel connection state tracking What started as a rather straightforward race condition reported by Dmitry using the syzkaller fuzzer ended up revealing some major problems with how the audit subsystem managed its netlink sockets and its connection with the userspace audit daemon. Fixing this properly had quite the cascading effect and what we are left with is this rather large and complicated patch. My initial goal was to try and decompose this patch into multiple smaller patches, but the way these changes are intertwined makes it difficult to split these changes into meaningful pieces that don't break or somehow make things worse for the intermediate states. The patch makes a number of changes, but the most significant are highlighted below: * The auditd tracking variables, e.g. audit_sock, are now gone and replaced by a RCU/spin_lock protected variable auditd_conn which is a structure containing all of the auditd tracking information. * We no longer track the auditd sock directly, instead we track it via the network namespace in which it resides and we use the audit socket associated with that namespace. In spirit, this is what the code was trying to do prior to this patch (at least I think that is what the original authors intended), but it was done rather poorly and added a layer of obfuscation that only masked the underlying problems. * Big backlog queue cleanup, again. In v4.10 we made some pretty big changes to how the audit backlog queues work, here we haven't changed the queue design so much as cleaned up the implementation. Brought about by the locking changes, we've simplified kauditd_thread() quite a bit by consolidating the queue handling into a new helper function, kauditd_send_queue(), which allows us to eliminate a lot of very similar code and makes the looping logic in kauditd_thread() clearer. * All netlink messages sent to auditd are now sent via auditd_send_unicast_skb(). Other than just making sense, this makes the lock handling easier. * Change the audit_log_start() sleep behavior so that we never sleep on auditd events (unchanged) or if the caller is holding the audit_cmd_mutex (changed). Previously we didn't sleep if the caller was auditd or if the message type fell between a certain range; the type check was a poor effort of doing what the cmd_mutex check now does. Richard Guy Briggs originally proposed not sleeping the cmd_mutex owner several years ago but his patch wasn't acceptable at the time. At least the idea lives on here. * A problem with the lost record counter has been resolved. Steve Grubb and I both happened to notice this problem and according to some quick testing by Steve, this problem goes back quite some time. It's largely a harmless problem, although it may have left some careful sysadmins quite puzzled. Cc: <stable@vger.kernel.org> # 4.10.x- Reported-by: Dmitry Vyukov <dvyukov@google.com> Signed-off-by: Paul Moore <paul@paul-moore.com>
2017-03-21 23:26:35 +08:00
* auditd_send_unicast_skb - Send a record via unicast to auditd
* @skb: audit record
*
* Description:
audit: fix auditd/kernel connection state tracking What started as a rather straightforward race condition reported by Dmitry using the syzkaller fuzzer ended up revealing some major problems with how the audit subsystem managed its netlink sockets and its connection with the userspace audit daemon. Fixing this properly had quite the cascading effect and what we are left with is this rather large and complicated patch. My initial goal was to try and decompose this patch into multiple smaller patches, but the way these changes are intertwined makes it difficult to split these changes into meaningful pieces that don't break or somehow make things worse for the intermediate states. The patch makes a number of changes, but the most significant are highlighted below: * The auditd tracking variables, e.g. audit_sock, are now gone and replaced by a RCU/spin_lock protected variable auditd_conn which is a structure containing all of the auditd tracking information. * We no longer track the auditd sock directly, instead we track it via the network namespace in which it resides and we use the audit socket associated with that namespace. In spirit, this is what the code was trying to do prior to this patch (at least I think that is what the original authors intended), but it was done rather poorly and added a layer of obfuscation that only masked the underlying problems. * Big backlog queue cleanup, again. In v4.10 we made some pretty big changes to how the audit backlog queues work, here we haven't changed the queue design so much as cleaned up the implementation. Brought about by the locking changes, we've simplified kauditd_thread() quite a bit by consolidating the queue handling into a new helper function, kauditd_send_queue(), which allows us to eliminate a lot of very similar code and makes the looping logic in kauditd_thread() clearer. * All netlink messages sent to auditd are now sent via auditd_send_unicast_skb(). Other than just making sense, this makes the lock handling easier. * Change the audit_log_start() sleep behavior so that we never sleep on auditd events (unchanged) or if the caller is holding the audit_cmd_mutex (changed). Previously we didn't sleep if the caller was auditd or if the message type fell between a certain range; the type check was a poor effort of doing what the cmd_mutex check now does. Richard Guy Briggs originally proposed not sleeping the cmd_mutex owner several years ago but his patch wasn't acceptable at the time. At least the idea lives on here. * A problem with the lost record counter has been resolved. Steve Grubb and I both happened to notice this problem and according to some quick testing by Steve, this problem goes back quite some time. It's largely a harmless problem, although it may have left some careful sysadmins quite puzzled. Cc: <stable@vger.kernel.org> # 4.10.x- Reported-by: Dmitry Vyukov <dvyukov@google.com> Signed-off-by: Paul Moore <paul@paul-moore.com>
2017-03-21 23:26:35 +08:00
* Send a skb to the audit daemon, returns positive/zero values on success and
* negative values on failure; in all cases the skb will be consumed by this
* function. If the send results in -ECONNREFUSED the connection with auditd
* will be reset. This function may sleep so callers should not hold any locks
* where this would cause a problem.
*/
audit: fix auditd/kernel connection state tracking What started as a rather straightforward race condition reported by Dmitry using the syzkaller fuzzer ended up revealing some major problems with how the audit subsystem managed its netlink sockets and its connection with the userspace audit daemon. Fixing this properly had quite the cascading effect and what we are left with is this rather large and complicated patch. My initial goal was to try and decompose this patch into multiple smaller patches, but the way these changes are intertwined makes it difficult to split these changes into meaningful pieces that don't break or somehow make things worse for the intermediate states. The patch makes a number of changes, but the most significant are highlighted below: * The auditd tracking variables, e.g. audit_sock, are now gone and replaced by a RCU/spin_lock protected variable auditd_conn which is a structure containing all of the auditd tracking information. * We no longer track the auditd sock directly, instead we track it via the network namespace in which it resides and we use the audit socket associated with that namespace. In spirit, this is what the code was trying to do prior to this patch (at least I think that is what the original authors intended), but it was done rather poorly and added a layer of obfuscation that only masked the underlying problems. * Big backlog queue cleanup, again. In v4.10 we made some pretty big changes to how the audit backlog queues work, here we haven't changed the queue design so much as cleaned up the implementation. Brought about by the locking changes, we've simplified kauditd_thread() quite a bit by consolidating the queue handling into a new helper function, kauditd_send_queue(), which allows us to eliminate a lot of very similar code and makes the looping logic in kauditd_thread() clearer. * All netlink messages sent to auditd are now sent via auditd_send_unicast_skb(). Other than just making sense, this makes the lock handling easier. * Change the audit_log_start() sleep behavior so that we never sleep on auditd events (unchanged) or if the caller is holding the audit_cmd_mutex (changed). Previously we didn't sleep if the caller was auditd or if the message type fell between a certain range; the type check was a poor effort of doing what the cmd_mutex check now does. Richard Guy Briggs originally proposed not sleeping the cmd_mutex owner several years ago but his patch wasn't acceptable at the time. At least the idea lives on here. * A problem with the lost record counter has been resolved. Steve Grubb and I both happened to notice this problem and according to some quick testing by Steve, this problem goes back quite some time. It's largely a harmless problem, although it may have left some careful sysadmins quite puzzled. Cc: <stable@vger.kernel.org> # 4.10.x- Reported-by: Dmitry Vyukov <dvyukov@google.com> Signed-off-by: Paul Moore <paul@paul-moore.com>
2017-03-21 23:26:35 +08:00
static int auditd_send_unicast_skb(struct sk_buff *skb)
{
audit: fix auditd/kernel connection state tracking What started as a rather straightforward race condition reported by Dmitry using the syzkaller fuzzer ended up revealing some major problems with how the audit subsystem managed its netlink sockets and its connection with the userspace audit daemon. Fixing this properly had quite the cascading effect and what we are left with is this rather large and complicated patch. My initial goal was to try and decompose this patch into multiple smaller patches, but the way these changes are intertwined makes it difficult to split these changes into meaningful pieces that don't break or somehow make things worse for the intermediate states. The patch makes a number of changes, but the most significant are highlighted below: * The auditd tracking variables, e.g. audit_sock, are now gone and replaced by a RCU/spin_lock protected variable auditd_conn which is a structure containing all of the auditd tracking information. * We no longer track the auditd sock directly, instead we track it via the network namespace in which it resides and we use the audit socket associated with that namespace. In spirit, this is what the code was trying to do prior to this patch (at least I think that is what the original authors intended), but it was done rather poorly and added a layer of obfuscation that only masked the underlying problems. * Big backlog queue cleanup, again. In v4.10 we made some pretty big changes to how the audit backlog queues work, here we haven't changed the queue design so much as cleaned up the implementation. Brought about by the locking changes, we've simplified kauditd_thread() quite a bit by consolidating the queue handling into a new helper function, kauditd_send_queue(), which allows us to eliminate a lot of very similar code and makes the looping logic in kauditd_thread() clearer. * All netlink messages sent to auditd are now sent via auditd_send_unicast_skb(). Other than just making sense, this makes the lock handling easier. * Change the audit_log_start() sleep behavior so that we never sleep on auditd events (unchanged) or if the caller is holding the audit_cmd_mutex (changed). Previously we didn't sleep if the caller was auditd or if the message type fell between a certain range; the type check was a poor effort of doing what the cmd_mutex check now does. Richard Guy Briggs originally proposed not sleeping the cmd_mutex owner several years ago but his patch wasn't acceptable at the time. At least the idea lives on here. * A problem with the lost record counter has been resolved. Steve Grubb and I both happened to notice this problem and according to some quick testing by Steve, this problem goes back quite some time. It's largely a harmless problem, although it may have left some careful sysadmins quite puzzled. Cc: <stable@vger.kernel.org> # 4.10.x- Reported-by: Dmitry Vyukov <dvyukov@google.com> Signed-off-by: Paul Moore <paul@paul-moore.com>
2017-03-21 23:26:35 +08:00
int rc;
u32 portid;
struct net *net;
struct sock *sk;
struct auditd_connection *ac;
audit: fix auditd/kernel connection state tracking What started as a rather straightforward race condition reported by Dmitry using the syzkaller fuzzer ended up revealing some major problems with how the audit subsystem managed its netlink sockets and its connection with the userspace audit daemon. Fixing this properly had quite the cascading effect and what we are left with is this rather large and complicated patch. My initial goal was to try and decompose this patch into multiple smaller patches, but the way these changes are intertwined makes it difficult to split these changes into meaningful pieces that don't break or somehow make things worse for the intermediate states. The patch makes a number of changes, but the most significant are highlighted below: * The auditd tracking variables, e.g. audit_sock, are now gone and replaced by a RCU/spin_lock protected variable auditd_conn which is a structure containing all of the auditd tracking information. * We no longer track the auditd sock directly, instead we track it via the network namespace in which it resides and we use the audit socket associated with that namespace. In spirit, this is what the code was trying to do prior to this patch (at least I think that is what the original authors intended), but it was done rather poorly and added a layer of obfuscation that only masked the underlying problems. * Big backlog queue cleanup, again. In v4.10 we made some pretty big changes to how the audit backlog queues work, here we haven't changed the queue design so much as cleaned up the implementation. Brought about by the locking changes, we've simplified kauditd_thread() quite a bit by consolidating the queue handling into a new helper function, kauditd_send_queue(), which allows us to eliminate a lot of very similar code and makes the looping logic in kauditd_thread() clearer. * All netlink messages sent to auditd are now sent via auditd_send_unicast_skb(). Other than just making sense, this makes the lock handling easier. * Change the audit_log_start() sleep behavior so that we never sleep on auditd events (unchanged) or if the caller is holding the audit_cmd_mutex (changed). Previously we didn't sleep if the caller was auditd or if the message type fell between a certain range; the type check was a poor effort of doing what the cmd_mutex check now does. Richard Guy Briggs originally proposed not sleeping the cmd_mutex owner several years ago but his patch wasn't acceptable at the time. At least the idea lives on here. * A problem with the lost record counter has been resolved. Steve Grubb and I both happened to notice this problem and according to some quick testing by Steve, this problem goes back quite some time. It's largely a harmless problem, although it may have left some careful sysadmins quite puzzled. Cc: <stable@vger.kernel.org> # 4.10.x- Reported-by: Dmitry Vyukov <dvyukov@google.com> Signed-off-by: Paul Moore <paul@paul-moore.com>
2017-03-21 23:26:35 +08:00
/* NOTE: we can't call netlink_unicast while in the RCU section so
* take a reference to the network namespace and grab local
* copies of the namespace, the sock, and the portid; the
* namespace and sock aren't going to go away while we hold a
* reference and if the portid does become invalid after the RCU
* section netlink_unicast() should safely return an error */
rcu_read_lock();
ac = rcu_dereference(auditd_conn);
if (!ac) {
audit: fix auditd/kernel connection state tracking What started as a rather straightforward race condition reported by Dmitry using the syzkaller fuzzer ended up revealing some major problems with how the audit subsystem managed its netlink sockets and its connection with the userspace audit daemon. Fixing this properly had quite the cascading effect and what we are left with is this rather large and complicated patch. My initial goal was to try and decompose this patch into multiple smaller patches, but the way these changes are intertwined makes it difficult to split these changes into meaningful pieces that don't break or somehow make things worse for the intermediate states. The patch makes a number of changes, but the most significant are highlighted below: * The auditd tracking variables, e.g. audit_sock, are now gone and replaced by a RCU/spin_lock protected variable auditd_conn which is a structure containing all of the auditd tracking information. * We no longer track the auditd sock directly, instead we track it via the network namespace in which it resides and we use the audit socket associated with that namespace. In spirit, this is what the code was trying to do prior to this patch (at least I think that is what the original authors intended), but it was done rather poorly and added a layer of obfuscation that only masked the underlying problems. * Big backlog queue cleanup, again. In v4.10 we made some pretty big changes to how the audit backlog queues work, here we haven't changed the queue design so much as cleaned up the implementation. Brought about by the locking changes, we've simplified kauditd_thread() quite a bit by consolidating the queue handling into a new helper function, kauditd_send_queue(), which allows us to eliminate a lot of very similar code and makes the looping logic in kauditd_thread() clearer. * All netlink messages sent to auditd are now sent via auditd_send_unicast_skb(). Other than just making sense, this makes the lock handling easier. * Change the audit_log_start() sleep behavior so that we never sleep on auditd events (unchanged) or if the caller is holding the audit_cmd_mutex (changed). Previously we didn't sleep if the caller was auditd or if the message type fell between a certain range; the type check was a poor effort of doing what the cmd_mutex check now does. Richard Guy Briggs originally proposed not sleeping the cmd_mutex owner several years ago but his patch wasn't acceptable at the time. At least the idea lives on here. * A problem with the lost record counter has been resolved. Steve Grubb and I both happened to notice this problem and according to some quick testing by Steve, this problem goes back quite some time. It's largely a harmless problem, although it may have left some careful sysadmins quite puzzled. Cc: <stable@vger.kernel.org> # 4.10.x- Reported-by: Dmitry Vyukov <dvyukov@google.com> Signed-off-by: Paul Moore <paul@paul-moore.com>
2017-03-21 23:26:35 +08:00
rcu_read_unlock();
kfree_skb(skb);
audit: fix auditd/kernel connection state tracking What started as a rather straightforward race condition reported by Dmitry using the syzkaller fuzzer ended up revealing some major problems with how the audit subsystem managed its netlink sockets and its connection with the userspace audit daemon. Fixing this properly had quite the cascading effect and what we are left with is this rather large and complicated patch. My initial goal was to try and decompose this patch into multiple smaller patches, but the way these changes are intertwined makes it difficult to split these changes into meaningful pieces that don't break or somehow make things worse for the intermediate states. The patch makes a number of changes, but the most significant are highlighted below: * The auditd tracking variables, e.g. audit_sock, are now gone and replaced by a RCU/spin_lock protected variable auditd_conn which is a structure containing all of the auditd tracking information. * We no longer track the auditd sock directly, instead we track it via the network namespace in which it resides and we use the audit socket associated with that namespace. In spirit, this is what the code was trying to do prior to this patch (at least I think that is what the original authors intended), but it was done rather poorly and added a layer of obfuscation that only masked the underlying problems. * Big backlog queue cleanup, again. In v4.10 we made some pretty big changes to how the audit backlog queues work, here we haven't changed the queue design so much as cleaned up the implementation. Brought about by the locking changes, we've simplified kauditd_thread() quite a bit by consolidating the queue handling into a new helper function, kauditd_send_queue(), which allows us to eliminate a lot of very similar code and makes the looping logic in kauditd_thread() clearer. * All netlink messages sent to auditd are now sent via auditd_send_unicast_skb(). Other than just making sense, this makes the lock handling easier. * Change the audit_log_start() sleep behavior so that we never sleep on auditd events (unchanged) or if the caller is holding the audit_cmd_mutex (changed). Previously we didn't sleep if the caller was auditd or if the message type fell between a certain range; the type check was a poor effort of doing what the cmd_mutex check now does. Richard Guy Briggs originally proposed not sleeping the cmd_mutex owner several years ago but his patch wasn't acceptable at the time. At least the idea lives on here. * A problem with the lost record counter has been resolved. Steve Grubb and I both happened to notice this problem and according to some quick testing by Steve, this problem goes back quite some time. It's largely a harmless problem, although it may have left some careful sysadmins quite puzzled. Cc: <stable@vger.kernel.org> # 4.10.x- Reported-by: Dmitry Vyukov <dvyukov@google.com> Signed-off-by: Paul Moore <paul@paul-moore.com>
2017-03-21 23:26:35 +08:00
rc = -ECONNREFUSED;
goto err;
}
net = get_net(ac->net);
audit: fix auditd/kernel connection state tracking What started as a rather straightforward race condition reported by Dmitry using the syzkaller fuzzer ended up revealing some major problems with how the audit subsystem managed its netlink sockets and its connection with the userspace audit daemon. Fixing this properly had quite the cascading effect and what we are left with is this rather large and complicated patch. My initial goal was to try and decompose this patch into multiple smaller patches, but the way these changes are intertwined makes it difficult to split these changes into meaningful pieces that don't break or somehow make things worse for the intermediate states. The patch makes a number of changes, but the most significant are highlighted below: * The auditd tracking variables, e.g. audit_sock, are now gone and replaced by a RCU/spin_lock protected variable auditd_conn which is a structure containing all of the auditd tracking information. * We no longer track the auditd sock directly, instead we track it via the network namespace in which it resides and we use the audit socket associated with that namespace. In spirit, this is what the code was trying to do prior to this patch (at least I think that is what the original authors intended), but it was done rather poorly and added a layer of obfuscation that only masked the underlying problems. * Big backlog queue cleanup, again. In v4.10 we made some pretty big changes to how the audit backlog queues work, here we haven't changed the queue design so much as cleaned up the implementation. Brought about by the locking changes, we've simplified kauditd_thread() quite a bit by consolidating the queue handling into a new helper function, kauditd_send_queue(), which allows us to eliminate a lot of very similar code and makes the looping logic in kauditd_thread() clearer. * All netlink messages sent to auditd are now sent via auditd_send_unicast_skb(). Other than just making sense, this makes the lock handling easier. * Change the audit_log_start() sleep behavior so that we never sleep on auditd events (unchanged) or if the caller is holding the audit_cmd_mutex (changed). Previously we didn't sleep if the caller was auditd or if the message type fell between a certain range; the type check was a poor effort of doing what the cmd_mutex check now does. Richard Guy Briggs originally proposed not sleeping the cmd_mutex owner several years ago but his patch wasn't acceptable at the time. At least the idea lives on here. * A problem with the lost record counter has been resolved. Steve Grubb and I both happened to notice this problem and according to some quick testing by Steve, this problem goes back quite some time. It's largely a harmless problem, although it may have left some careful sysadmins quite puzzled. Cc: <stable@vger.kernel.org> # 4.10.x- Reported-by: Dmitry Vyukov <dvyukov@google.com> Signed-off-by: Paul Moore <paul@paul-moore.com>
2017-03-21 23:26:35 +08:00
sk = audit_get_sk(net);
portid = ac->portid;
audit: fix auditd/kernel connection state tracking What started as a rather straightforward race condition reported by Dmitry using the syzkaller fuzzer ended up revealing some major problems with how the audit subsystem managed its netlink sockets and its connection with the userspace audit daemon. Fixing this properly had quite the cascading effect and what we are left with is this rather large and complicated patch. My initial goal was to try and decompose this patch into multiple smaller patches, but the way these changes are intertwined makes it difficult to split these changes into meaningful pieces that don't break or somehow make things worse for the intermediate states. The patch makes a number of changes, but the most significant are highlighted below: * The auditd tracking variables, e.g. audit_sock, are now gone and replaced by a RCU/spin_lock protected variable auditd_conn which is a structure containing all of the auditd tracking information. * We no longer track the auditd sock directly, instead we track it via the network namespace in which it resides and we use the audit socket associated with that namespace. In spirit, this is what the code was trying to do prior to this patch (at least I think that is what the original authors intended), but it was done rather poorly and added a layer of obfuscation that only masked the underlying problems. * Big backlog queue cleanup, again. In v4.10 we made some pretty big changes to how the audit backlog queues work, here we haven't changed the queue design so much as cleaned up the implementation. Brought about by the locking changes, we've simplified kauditd_thread() quite a bit by consolidating the queue handling into a new helper function, kauditd_send_queue(), which allows us to eliminate a lot of very similar code and makes the looping logic in kauditd_thread() clearer. * All netlink messages sent to auditd are now sent via auditd_send_unicast_skb(). Other than just making sense, this makes the lock handling easier. * Change the audit_log_start() sleep behavior so that we never sleep on auditd events (unchanged) or if the caller is holding the audit_cmd_mutex (changed). Previously we didn't sleep if the caller was auditd or if the message type fell between a certain range; the type check was a poor effort of doing what the cmd_mutex check now does. Richard Guy Briggs originally proposed not sleeping the cmd_mutex owner several years ago but his patch wasn't acceptable at the time. At least the idea lives on here. * A problem with the lost record counter has been resolved. Steve Grubb and I both happened to notice this problem and according to some quick testing by Steve, this problem goes back quite some time. It's largely a harmless problem, although it may have left some careful sysadmins quite puzzled. Cc: <stable@vger.kernel.org> # 4.10.x- Reported-by: Dmitry Vyukov <dvyukov@google.com> Signed-off-by: Paul Moore <paul@paul-moore.com>
2017-03-21 23:26:35 +08:00
rcu_read_unlock();
audit: fix auditd/kernel connection state tracking What started as a rather straightforward race condition reported by Dmitry using the syzkaller fuzzer ended up revealing some major problems with how the audit subsystem managed its netlink sockets and its connection with the userspace audit daemon. Fixing this properly had quite the cascading effect and what we are left with is this rather large and complicated patch. My initial goal was to try and decompose this patch into multiple smaller patches, but the way these changes are intertwined makes it difficult to split these changes into meaningful pieces that don't break or somehow make things worse for the intermediate states. The patch makes a number of changes, but the most significant are highlighted below: * The auditd tracking variables, e.g. audit_sock, are now gone and replaced by a RCU/spin_lock protected variable auditd_conn which is a structure containing all of the auditd tracking information. * We no longer track the auditd sock directly, instead we track it via the network namespace in which it resides and we use the audit socket associated with that namespace. In spirit, this is what the code was trying to do prior to this patch (at least I think that is what the original authors intended), but it was done rather poorly and added a layer of obfuscation that only masked the underlying problems. * Big backlog queue cleanup, again. In v4.10 we made some pretty big changes to how the audit backlog queues work, here we haven't changed the queue design so much as cleaned up the implementation. Brought about by the locking changes, we've simplified kauditd_thread() quite a bit by consolidating the queue handling into a new helper function, kauditd_send_queue(), which allows us to eliminate a lot of very similar code and makes the looping logic in kauditd_thread() clearer. * All netlink messages sent to auditd are now sent via auditd_send_unicast_skb(). Other than just making sense, this makes the lock handling easier. * Change the audit_log_start() sleep behavior so that we never sleep on auditd events (unchanged) or if the caller is holding the audit_cmd_mutex (changed). Previously we didn't sleep if the caller was auditd or if the message type fell between a certain range; the type check was a poor effort of doing what the cmd_mutex check now does. Richard Guy Briggs originally proposed not sleeping the cmd_mutex owner several years ago but his patch wasn't acceptable at the time. At least the idea lives on here. * A problem with the lost record counter has been resolved. Steve Grubb and I both happened to notice this problem and according to some quick testing by Steve, this problem goes back quite some time. It's largely a harmless problem, although it may have left some careful sysadmins quite puzzled. Cc: <stable@vger.kernel.org> # 4.10.x- Reported-by: Dmitry Vyukov <dvyukov@google.com> Signed-off-by: Paul Moore <paul@paul-moore.com>
2017-03-21 23:26:35 +08:00
rc = netlink_unicast(sk, skb, portid, 0);
put_net(net);
if (rc < 0)
goto err;
return rc;
err:
audit: fix a race condition with the auditd tracking code Originally reported by Adam and Dusty, it appears we have a small race window in kauditd_thread(), as documented in the Fedora BZ: * https://bugzilla.redhat.com/show_bug.cgi?id=1459326#c35 "This issue is partly due to the read-copy nature of RCU, and partly due to how we sync the auditd_connection state across kauditd_thread and the audit control channel. The kauditd_thread thread is always running so it can service the record queues and emit the multicast messages, if it happens to be just past the "main_queue" label, but before the "if (sk == NULL || ...)" if-statement which calls auditd_reset() when the new auditd connection is registered it could end up resetting the auditd connection, regardless of if it is valid or not. This is a rather small window and the variable nature of multi-core scheduling explains why this is proving rather difficult to reproduce." The fix is to have functions only call auditd_reset() when they believe that the kernel/auditd connection is still valid, e.g. non-NULL, and to have these callers pass their local copy of the auditd_connection pointer to auditd_reset() where it can be compared with the current connection state before resetting. If the caller has a stale state tracking pointer then the reset is ignored. We also make a small change to kauditd_thread() so that if the kernel/auditd connection is dead we skip the retry queue and send the records straight to the hold queue. This is necessary as we used to rely on auditd_reset() to occasionally purge the retry queue but we are going to be calling the reset function much less now and we want to make sure the retry queue doesn't grow unbounded. Reported-by: Adam Williamson <awilliam@redhat.com> Reported-by: Dusty Mabe <dustymabe@redhat.com> Reviewed-by: Richard Guy Briggs <rgb@redhat.com> Signed-off-by: Paul Moore <paul@paul-moore.com>
2017-06-12 21:35:24 +08:00
if (ac && rc == -ECONNREFUSED)
auditd_reset(ac);
audit: fix auditd/kernel connection state tracking What started as a rather straightforward race condition reported by Dmitry using the syzkaller fuzzer ended up revealing some major problems with how the audit subsystem managed its netlink sockets and its connection with the userspace audit daemon. Fixing this properly had quite the cascading effect and what we are left with is this rather large and complicated patch. My initial goal was to try and decompose this patch into multiple smaller patches, but the way these changes are intertwined makes it difficult to split these changes into meaningful pieces that don't break or somehow make things worse for the intermediate states. The patch makes a number of changes, but the most significant are highlighted below: * The auditd tracking variables, e.g. audit_sock, are now gone and replaced by a RCU/spin_lock protected variable auditd_conn which is a structure containing all of the auditd tracking information. * We no longer track the auditd sock directly, instead we track it via the network namespace in which it resides and we use the audit socket associated with that namespace. In spirit, this is what the code was trying to do prior to this patch (at least I think that is what the original authors intended), but it was done rather poorly and added a layer of obfuscation that only masked the underlying problems. * Big backlog queue cleanup, again. In v4.10 we made some pretty big changes to how the audit backlog queues work, here we haven't changed the queue design so much as cleaned up the implementation. Brought about by the locking changes, we've simplified kauditd_thread() quite a bit by consolidating the queue handling into a new helper function, kauditd_send_queue(), which allows us to eliminate a lot of very similar code and makes the looping logic in kauditd_thread() clearer. * All netlink messages sent to auditd are now sent via auditd_send_unicast_skb(). Other than just making sense, this makes the lock handling easier. * Change the audit_log_start() sleep behavior so that we never sleep on auditd events (unchanged) or if the caller is holding the audit_cmd_mutex (changed). Previously we didn't sleep if the caller was auditd or if the message type fell between a certain range; the type check was a poor effort of doing what the cmd_mutex check now does. Richard Guy Briggs originally proposed not sleeping the cmd_mutex owner several years ago but his patch wasn't acceptable at the time. At least the idea lives on here. * A problem with the lost record counter has been resolved. Steve Grubb and I both happened to notice this problem and according to some quick testing by Steve, this problem goes back quite some time. It's largely a harmless problem, although it may have left some careful sysadmins quite puzzled. Cc: <stable@vger.kernel.org> # 4.10.x- Reported-by: Dmitry Vyukov <dvyukov@google.com> Signed-off-by: Paul Moore <paul@paul-moore.com>
2017-03-21 23:26:35 +08:00
return rc;
}
/**
audit: fix auditd/kernel connection state tracking What started as a rather straightforward race condition reported by Dmitry using the syzkaller fuzzer ended up revealing some major problems with how the audit subsystem managed its netlink sockets and its connection with the userspace audit daemon. Fixing this properly had quite the cascading effect and what we are left with is this rather large and complicated patch. My initial goal was to try and decompose this patch into multiple smaller patches, but the way these changes are intertwined makes it difficult to split these changes into meaningful pieces that don't break or somehow make things worse for the intermediate states. The patch makes a number of changes, but the most significant are highlighted below: * The auditd tracking variables, e.g. audit_sock, are now gone and replaced by a RCU/spin_lock protected variable auditd_conn which is a structure containing all of the auditd tracking information. * We no longer track the auditd sock directly, instead we track it via the network namespace in which it resides and we use the audit socket associated with that namespace. In spirit, this is what the code was trying to do prior to this patch (at least I think that is what the original authors intended), but it was done rather poorly and added a layer of obfuscation that only masked the underlying problems. * Big backlog queue cleanup, again. In v4.10 we made some pretty big changes to how the audit backlog queues work, here we haven't changed the queue design so much as cleaned up the implementation. Brought about by the locking changes, we've simplified kauditd_thread() quite a bit by consolidating the queue handling into a new helper function, kauditd_send_queue(), which allows us to eliminate a lot of very similar code and makes the looping logic in kauditd_thread() clearer. * All netlink messages sent to auditd are now sent via auditd_send_unicast_skb(). Other than just making sense, this makes the lock handling easier. * Change the audit_log_start() sleep behavior so that we never sleep on auditd events (unchanged) or if the caller is holding the audit_cmd_mutex (changed). Previously we didn't sleep if the caller was auditd or if the message type fell between a certain range; the type check was a poor effort of doing what the cmd_mutex check now does. Richard Guy Briggs originally proposed not sleeping the cmd_mutex owner several years ago but his patch wasn't acceptable at the time. At least the idea lives on here. * A problem with the lost record counter has been resolved. Steve Grubb and I both happened to notice this problem and according to some quick testing by Steve, this problem goes back quite some time. It's largely a harmless problem, although it may have left some careful sysadmins quite puzzled. Cc: <stable@vger.kernel.org> # 4.10.x- Reported-by: Dmitry Vyukov <dvyukov@google.com> Signed-off-by: Paul Moore <paul@paul-moore.com>
2017-03-21 23:26:35 +08:00
* kauditd_send_queue - Helper for kauditd_thread to flush skb queues
* @sk: the sending sock
* @portid: the netlink destination
* @queue: the skb queue to process
* @retry_limit: limit on number of netlink unicast failures
* @skb_hook: per-skb hook for additional processing
* @err_hook: hook called if the skb fails the netlink unicast send
*
* Description:
* Run through the given queue and attempt to send the audit records to auditd,
* returns zero on success, negative values on failure. It is up to the caller
* to ensure that the @sk is valid for the duration of this function.
*
*/
audit: fix auditd/kernel connection state tracking What started as a rather straightforward race condition reported by Dmitry using the syzkaller fuzzer ended up revealing some major problems with how the audit subsystem managed its netlink sockets and its connection with the userspace audit daemon. Fixing this properly had quite the cascading effect and what we are left with is this rather large and complicated patch. My initial goal was to try and decompose this patch into multiple smaller patches, but the way these changes are intertwined makes it difficult to split these changes into meaningful pieces that don't break or somehow make things worse for the intermediate states. The patch makes a number of changes, but the most significant are highlighted below: * The auditd tracking variables, e.g. audit_sock, are now gone and replaced by a RCU/spin_lock protected variable auditd_conn which is a structure containing all of the auditd tracking information. * We no longer track the auditd sock directly, instead we track it via the network namespace in which it resides and we use the audit socket associated with that namespace. In spirit, this is what the code was trying to do prior to this patch (at least I think that is what the original authors intended), but it was done rather poorly and added a layer of obfuscation that only masked the underlying problems. * Big backlog queue cleanup, again. In v4.10 we made some pretty big changes to how the audit backlog queues work, here we haven't changed the queue design so much as cleaned up the implementation. Brought about by the locking changes, we've simplified kauditd_thread() quite a bit by consolidating the queue handling into a new helper function, kauditd_send_queue(), which allows us to eliminate a lot of very similar code and makes the looping logic in kauditd_thread() clearer. * All netlink messages sent to auditd are now sent via auditd_send_unicast_skb(). Other than just making sense, this makes the lock handling easier. * Change the audit_log_start() sleep behavior so that we never sleep on auditd events (unchanged) or if the caller is holding the audit_cmd_mutex (changed). Previously we didn't sleep if the caller was auditd or if the message type fell between a certain range; the type check was a poor effort of doing what the cmd_mutex check now does. Richard Guy Briggs originally proposed not sleeping the cmd_mutex owner several years ago but his patch wasn't acceptable at the time. At least the idea lives on here. * A problem with the lost record counter has been resolved. Steve Grubb and I both happened to notice this problem and according to some quick testing by Steve, this problem goes back quite some time. It's largely a harmless problem, although it may have left some careful sysadmins quite puzzled. Cc: <stable@vger.kernel.org> # 4.10.x- Reported-by: Dmitry Vyukov <dvyukov@google.com> Signed-off-by: Paul Moore <paul@paul-moore.com>
2017-03-21 23:26:35 +08:00
static int kauditd_send_queue(struct sock *sk, u32 portid,
struct sk_buff_head *queue,
unsigned int retry_limit,
void (*skb_hook)(struct sk_buff *skb),
audit: improve audit queue handling when "audit=1" on cmdline When an admin enables audit at early boot via the "audit=1" kernel command line the audit queue behavior is slightly different; the audit subsystem goes to greater lengths to avoid dropping records, which unfortunately can result in problems when the audit daemon is forcibly stopped for an extended period of time. This patch makes a number of changes designed to improve the audit queuing behavior so that leaving the audit daemon in a stopped state for an extended period does not cause a significant impact to the system. - kauditd_send_queue() is now limited to looping through the passed queue only once per call. This not only prevents the function from looping indefinitely when records are returned to the current queue, it also allows any recovery handling in kauditd_thread() to take place when kauditd_send_queue() returns. - Transient netlink send errors seen as -EAGAIN now cause the record to be returned to the retry queue instead of going to the hold queue. The intention of the hold queue is to store, perhaps for an extended period of time, the events which led up to the audit daemon going offline. The retry queue remains a temporary queue intended to protect against transient issues between the kernel and the audit daemon. - The retry queue is now limited by the audit_backlog_limit setting, the same as the other queues. This allows admins to bound the size of all of the audit queues on the system. - kauditd_rehold_skb() now returns records to the end of the hold queue to ensure ordering is preserved in the face of recent changes to kauditd_send_queue(). Cc: stable@vger.kernel.org Fixes: 5b52330bbfe63 ("audit: fix auditd/kernel connection state tracking") Fixes: f4b3ee3c85551 ("audit: improve robustness of the audit queue handling") Reported-by: Gaosheng Cui <cuigaosheng1@huawei.com> Tested-by: Gaosheng Cui <cuigaosheng1@huawei.com> Reviewed-by: Richard Guy Briggs <rgb@redhat.com> Signed-off-by: Paul Moore <paul@paul-moore.com>
2022-01-14 07:54:38 +08:00
void (*err_hook)(struct sk_buff *skb, int error))
{
audit: fix auditd/kernel connection state tracking What started as a rather straightforward race condition reported by Dmitry using the syzkaller fuzzer ended up revealing some major problems with how the audit subsystem managed its netlink sockets and its connection with the userspace audit daemon. Fixing this properly had quite the cascading effect and what we are left with is this rather large and complicated patch. My initial goal was to try and decompose this patch into multiple smaller patches, but the way these changes are intertwined makes it difficult to split these changes into meaningful pieces that don't break or somehow make things worse for the intermediate states. The patch makes a number of changes, but the most significant are highlighted below: * The auditd tracking variables, e.g. audit_sock, are now gone and replaced by a RCU/spin_lock protected variable auditd_conn which is a structure containing all of the auditd tracking information. * We no longer track the auditd sock directly, instead we track it via the network namespace in which it resides and we use the audit socket associated with that namespace. In spirit, this is what the code was trying to do prior to this patch (at least I think that is what the original authors intended), but it was done rather poorly and added a layer of obfuscation that only masked the underlying problems. * Big backlog queue cleanup, again. In v4.10 we made some pretty big changes to how the audit backlog queues work, here we haven't changed the queue design so much as cleaned up the implementation. Brought about by the locking changes, we've simplified kauditd_thread() quite a bit by consolidating the queue handling into a new helper function, kauditd_send_queue(), which allows us to eliminate a lot of very similar code and makes the looping logic in kauditd_thread() clearer. * All netlink messages sent to auditd are now sent via auditd_send_unicast_skb(). Other than just making sense, this makes the lock handling easier. * Change the audit_log_start() sleep behavior so that we never sleep on auditd events (unchanged) or if the caller is holding the audit_cmd_mutex (changed). Previously we didn't sleep if the caller was auditd or if the message type fell between a certain range; the type check was a poor effort of doing what the cmd_mutex check now does. Richard Guy Briggs originally proposed not sleeping the cmd_mutex owner several years ago but his patch wasn't acceptable at the time. At least the idea lives on here. * A problem with the lost record counter has been resolved. Steve Grubb and I both happened to notice this problem and according to some quick testing by Steve, this problem goes back quite some time. It's largely a harmless problem, although it may have left some careful sysadmins quite puzzled. Cc: <stable@vger.kernel.org> # 4.10.x- Reported-by: Dmitry Vyukov <dvyukov@google.com> Signed-off-by: Paul Moore <paul@paul-moore.com>
2017-03-21 23:26:35 +08:00
int rc = 0;
audit: improve audit queue handling when "audit=1" on cmdline When an admin enables audit at early boot via the "audit=1" kernel command line the audit queue behavior is slightly different; the audit subsystem goes to greater lengths to avoid dropping records, which unfortunately can result in problems when the audit daemon is forcibly stopped for an extended period of time. This patch makes a number of changes designed to improve the audit queuing behavior so that leaving the audit daemon in a stopped state for an extended period does not cause a significant impact to the system. - kauditd_send_queue() is now limited to looping through the passed queue only once per call. This not only prevents the function from looping indefinitely when records are returned to the current queue, it also allows any recovery handling in kauditd_thread() to take place when kauditd_send_queue() returns. - Transient netlink send errors seen as -EAGAIN now cause the record to be returned to the retry queue instead of going to the hold queue. The intention of the hold queue is to store, perhaps for an extended period of time, the events which led up to the audit daemon going offline. The retry queue remains a temporary queue intended to protect against transient issues between the kernel and the audit daemon. - The retry queue is now limited by the audit_backlog_limit setting, the same as the other queues. This allows admins to bound the size of all of the audit queues on the system. - kauditd_rehold_skb() now returns records to the end of the hold queue to ensure ordering is preserved in the face of recent changes to kauditd_send_queue(). Cc: stable@vger.kernel.org Fixes: 5b52330bbfe63 ("audit: fix auditd/kernel connection state tracking") Fixes: f4b3ee3c85551 ("audit: improve robustness of the audit queue handling") Reported-by: Gaosheng Cui <cuigaosheng1@huawei.com> Tested-by: Gaosheng Cui <cuigaosheng1@huawei.com> Reviewed-by: Richard Guy Briggs <rgb@redhat.com> Signed-off-by: Paul Moore <paul@paul-moore.com>
2022-01-14 07:54:38 +08:00
struct sk_buff *skb = NULL;
struct sk_buff *skb_tail;
audit: improve robustness of the audit queue handling If the audit daemon were ever to get stuck in a stopped state the kernel's kauditd_thread() could get blocked attempting to send audit records to the userspace audit daemon. With the kernel thread blocked it is possible that the audit queue could grow unbounded as certain audit record generating events must be exempt from the queue limits else the system enter a deadlock state. This patch resolves this problem by lowering the kernel thread's socket sending timeout from MAX_SCHEDULE_TIMEOUT to HZ/10 and tweaks the kauditd_send_queue() function to better manage the various audit queues when connection problems occur between the kernel and the audit daemon. With this patch, the backlog may temporarily grow beyond the defined limits when the audit daemon is stopped and the system is under heavy audit pressure, but kauditd_thread() will continue to make progress and drain the queues as it would for other connection problems. For example, with the audit daemon put into a stopped state and the system configured to audit every syscall it was still possible to shutdown the system without a kernel panic, deadlock, etc.; granted, the system was slow to shutdown but that is to be expected given the extreme pressure of recording every syscall. The timeout value of HZ/10 was chosen primarily through experimentation and this developer's "gut feeling". There is likely no one perfect value, but as this scenario is limited in scope (root privileges would be needed to send SIGSTOP to the audit daemon), it is likely not worth exposing this as a tunable at present. This can always be done at a later date if it proves necessary. Cc: stable@vger.kernel.org Fixes: 5b52330bbfe63 ("audit: fix auditd/kernel connection state tracking") Reported-by: Gaosheng Cui <cuigaosheng1@huawei.com> Tested-by: Gaosheng Cui <cuigaosheng1@huawei.com> Reviewed-by: Richard Guy Briggs <rgb@redhat.com> Signed-off-by: Paul Moore <paul@paul-moore.com>
2021-12-10 00:46:07 +08:00
unsigned int failed = 0;
audit: try harder to send to auditd upon netlink failure There are several reports of the kernel losing contact with auditd when it is, in fact, still running. When this happens, kernel syslogs show: "audit: *NO* daemon at audit_pid=<pid>" although auditd is still running, and is apparently happy, listening on the netlink socket. The pid in the "*NO* daemon" message matches the pid of the running auditd process. Restarting auditd solves this. The problem appears to happen randomly, and doesn't seem to be strongly correlated to the rate of audit events being logged. The problem happens fairly regularly (every few days), but not yet reproduced to order. On production kernels, BUG_ON() is a no-op, so any error will trigger this. Commit 34eab0a7cd45 ("audit: prevent an older auditd shutdown from orphaning a newer auditd startup") eliminates one possible cause. This isn't the case here, since the PID in the error message and the PID of the running auditd match. The primary expected cause of error here is -ECONNREFUSED when the audit daemon goes away, when netlink_getsockbyportid() can't find the auditd portid entry in the netlink audit table (or there is no receive function). If -EPERM is returned, that situation isn't likely to be resolved in a timely fashion without administrator intervention. In both cases, reset the audit_pid. This does not rule out a race condition. SELinux is expected to return zero since this isn't an INET or INET6 socket. Other LSMs may have other return codes. Log the error code for better diagnosis in the future. In the case of -ENOMEM, the situation could be temporary, based on local or general availability of buffers. -EAGAIN should never happen since the netlink audit (kernel) socket is set to MAX_SCHEDULE_TIMEOUT. -ERESTARTSYS and -EINTR are not expected since this kernel thread is not expected to receive signals. In these cases (or any other unexpected ones for now), report the error and re-schedule the thread, retrying up to 5 times. v2: Removed BUG_ON(). Moved comma in pr_*() statements. Removed audit_strerror() text. Reported-by: Vipin Rathor <v.rathor@gmail.com> Reported-by: <ctcard@hotmail.com> Signed-off-by: Richard Guy Briggs <rgb@redhat.com> [PM: applied rgb's fixup patch to correct audit_log_lost() format issues] Signed-off-by: Paul Moore <pmoore@redhat.com>
2015-11-04 21:23:50 +08:00
audit: fix auditd/kernel connection state tracking What started as a rather straightforward race condition reported by Dmitry using the syzkaller fuzzer ended up revealing some major problems with how the audit subsystem managed its netlink sockets and its connection with the userspace audit daemon. Fixing this properly had quite the cascading effect and what we are left with is this rather large and complicated patch. My initial goal was to try and decompose this patch into multiple smaller patches, but the way these changes are intertwined makes it difficult to split these changes into meaningful pieces that don't break or somehow make things worse for the intermediate states. The patch makes a number of changes, but the most significant are highlighted below: * The auditd tracking variables, e.g. audit_sock, are now gone and replaced by a RCU/spin_lock protected variable auditd_conn which is a structure containing all of the auditd tracking information. * We no longer track the auditd sock directly, instead we track it via the network namespace in which it resides and we use the audit socket associated with that namespace. In spirit, this is what the code was trying to do prior to this patch (at least I think that is what the original authors intended), but it was done rather poorly and added a layer of obfuscation that only masked the underlying problems. * Big backlog queue cleanup, again. In v4.10 we made some pretty big changes to how the audit backlog queues work, here we haven't changed the queue design so much as cleaned up the implementation. Brought about by the locking changes, we've simplified kauditd_thread() quite a bit by consolidating the queue handling into a new helper function, kauditd_send_queue(), which allows us to eliminate a lot of very similar code and makes the looping logic in kauditd_thread() clearer. * All netlink messages sent to auditd are now sent via auditd_send_unicast_skb(). Other than just making sense, this makes the lock handling easier. * Change the audit_log_start() sleep behavior so that we never sleep on auditd events (unchanged) or if the caller is holding the audit_cmd_mutex (changed). Previously we didn't sleep if the caller was auditd or if the message type fell between a certain range; the type check was a poor effort of doing what the cmd_mutex check now does. Richard Guy Briggs originally proposed not sleeping the cmd_mutex owner several years ago but his patch wasn't acceptable at the time. At least the idea lives on here. * A problem with the lost record counter has been resolved. Steve Grubb and I both happened to notice this problem and according to some quick testing by Steve, this problem goes back quite some time. It's largely a harmless problem, although it may have left some careful sysadmins quite puzzled. Cc: <stable@vger.kernel.org> # 4.10.x- Reported-by: Dmitry Vyukov <dvyukov@google.com> Signed-off-by: Paul Moore <paul@paul-moore.com>
2017-03-21 23:26:35 +08:00
/* NOTE: kauditd_thread takes care of all our locking, we just use
* the netlink info passed to us (e.g. sk and portid) */
audit: improve audit queue handling when "audit=1" on cmdline When an admin enables audit at early boot via the "audit=1" kernel command line the audit queue behavior is slightly different; the audit subsystem goes to greater lengths to avoid dropping records, which unfortunately can result in problems when the audit daemon is forcibly stopped for an extended period of time. This patch makes a number of changes designed to improve the audit queuing behavior so that leaving the audit daemon in a stopped state for an extended period does not cause a significant impact to the system. - kauditd_send_queue() is now limited to looping through the passed queue only once per call. This not only prevents the function from looping indefinitely when records are returned to the current queue, it also allows any recovery handling in kauditd_thread() to take place when kauditd_send_queue() returns. - Transient netlink send errors seen as -EAGAIN now cause the record to be returned to the retry queue instead of going to the hold queue. The intention of the hold queue is to store, perhaps for an extended period of time, the events which led up to the audit daemon going offline. The retry queue remains a temporary queue intended to protect against transient issues between the kernel and the audit daemon. - The retry queue is now limited by the audit_backlog_limit setting, the same as the other queues. This allows admins to bound the size of all of the audit queues on the system. - kauditd_rehold_skb() now returns records to the end of the hold queue to ensure ordering is preserved in the face of recent changes to kauditd_send_queue(). Cc: stable@vger.kernel.org Fixes: 5b52330bbfe63 ("audit: fix auditd/kernel connection state tracking") Fixes: f4b3ee3c85551 ("audit: improve robustness of the audit queue handling") Reported-by: Gaosheng Cui <cuigaosheng1@huawei.com> Tested-by: Gaosheng Cui <cuigaosheng1@huawei.com> Reviewed-by: Richard Guy Briggs <rgb@redhat.com> Signed-off-by: Paul Moore <paul@paul-moore.com>
2022-01-14 07:54:38 +08:00
skb_tail = skb_peek_tail(queue);
while ((skb != skb_tail) && (skb = skb_dequeue(queue))) {
audit: fix auditd/kernel connection state tracking What started as a rather straightforward race condition reported by Dmitry using the syzkaller fuzzer ended up revealing some major problems with how the audit subsystem managed its netlink sockets and its connection with the userspace audit daemon. Fixing this properly had quite the cascading effect and what we are left with is this rather large and complicated patch. My initial goal was to try and decompose this patch into multiple smaller patches, but the way these changes are intertwined makes it difficult to split these changes into meaningful pieces that don't break or somehow make things worse for the intermediate states. The patch makes a number of changes, but the most significant are highlighted below: * The auditd tracking variables, e.g. audit_sock, are now gone and replaced by a RCU/spin_lock protected variable auditd_conn which is a structure containing all of the auditd tracking information. * We no longer track the auditd sock directly, instead we track it via the network namespace in which it resides and we use the audit socket associated with that namespace. In spirit, this is what the code was trying to do prior to this patch (at least I think that is what the original authors intended), but it was done rather poorly and added a layer of obfuscation that only masked the underlying problems. * Big backlog queue cleanup, again. In v4.10 we made some pretty big changes to how the audit backlog queues work, here we haven't changed the queue design so much as cleaned up the implementation. Brought about by the locking changes, we've simplified kauditd_thread() quite a bit by consolidating the queue handling into a new helper function, kauditd_send_queue(), which allows us to eliminate a lot of very similar code and makes the looping logic in kauditd_thread() clearer. * All netlink messages sent to auditd are now sent via auditd_send_unicast_skb(). Other than just making sense, this makes the lock handling easier. * Change the audit_log_start() sleep behavior so that we never sleep on auditd events (unchanged) or if the caller is holding the audit_cmd_mutex (changed). Previously we didn't sleep if the caller was auditd or if the message type fell between a certain range; the type check was a poor effort of doing what the cmd_mutex check now does. Richard Guy Briggs originally proposed not sleeping the cmd_mutex owner several years ago but his patch wasn't acceptable at the time. At least the idea lives on here. * A problem with the lost record counter has been resolved. Steve Grubb and I both happened to notice this problem and according to some quick testing by Steve, this problem goes back quite some time. It's largely a harmless problem, although it may have left some careful sysadmins quite puzzled. Cc: <stable@vger.kernel.org> # 4.10.x- Reported-by: Dmitry Vyukov <dvyukov@google.com> Signed-off-by: Paul Moore <paul@paul-moore.com>
2017-03-21 23:26:35 +08:00
/* call the skb_hook for each skb we touch */
if (skb_hook)
(*skb_hook)(skb);
/* can we send to anyone via unicast? */
if (!sk) {
if (err_hook)
audit: improve audit queue handling when "audit=1" on cmdline When an admin enables audit at early boot via the "audit=1" kernel command line the audit queue behavior is slightly different; the audit subsystem goes to greater lengths to avoid dropping records, which unfortunately can result in problems when the audit daemon is forcibly stopped for an extended period of time. This patch makes a number of changes designed to improve the audit queuing behavior so that leaving the audit daemon in a stopped state for an extended period does not cause a significant impact to the system. - kauditd_send_queue() is now limited to looping through the passed queue only once per call. This not only prevents the function from looping indefinitely when records are returned to the current queue, it also allows any recovery handling in kauditd_thread() to take place when kauditd_send_queue() returns. - Transient netlink send errors seen as -EAGAIN now cause the record to be returned to the retry queue instead of going to the hold queue. The intention of the hold queue is to store, perhaps for an extended period of time, the events which led up to the audit daemon going offline. The retry queue remains a temporary queue intended to protect against transient issues between the kernel and the audit daemon. - The retry queue is now limited by the audit_backlog_limit setting, the same as the other queues. This allows admins to bound the size of all of the audit queues on the system. - kauditd_rehold_skb() now returns records to the end of the hold queue to ensure ordering is preserved in the face of recent changes to kauditd_send_queue(). Cc: stable@vger.kernel.org Fixes: 5b52330bbfe63 ("audit: fix auditd/kernel connection state tracking") Fixes: f4b3ee3c85551 ("audit: improve robustness of the audit queue handling") Reported-by: Gaosheng Cui <cuigaosheng1@huawei.com> Tested-by: Gaosheng Cui <cuigaosheng1@huawei.com> Reviewed-by: Richard Guy Briggs <rgb@redhat.com> Signed-off-by: Paul Moore <paul@paul-moore.com>
2022-01-14 07:54:38 +08:00
(*err_hook)(skb, -ECONNREFUSED);
audit: fix auditd/kernel connection state tracking What started as a rather straightforward race condition reported by Dmitry using the syzkaller fuzzer ended up revealing some major problems with how the audit subsystem managed its netlink sockets and its connection with the userspace audit daemon. Fixing this properly had quite the cascading effect and what we are left with is this rather large and complicated patch. My initial goal was to try and decompose this patch into multiple smaller patches, but the way these changes are intertwined makes it difficult to split these changes into meaningful pieces that don't break or somehow make things worse for the intermediate states. The patch makes a number of changes, but the most significant are highlighted below: * The auditd tracking variables, e.g. audit_sock, are now gone and replaced by a RCU/spin_lock protected variable auditd_conn which is a structure containing all of the auditd tracking information. * We no longer track the auditd sock directly, instead we track it via the network namespace in which it resides and we use the audit socket associated with that namespace. In spirit, this is what the code was trying to do prior to this patch (at least I think that is what the original authors intended), but it was done rather poorly and added a layer of obfuscation that only masked the underlying problems. * Big backlog queue cleanup, again. In v4.10 we made some pretty big changes to how the audit backlog queues work, here we haven't changed the queue design so much as cleaned up the implementation. Brought about by the locking changes, we've simplified kauditd_thread() quite a bit by consolidating the queue handling into a new helper function, kauditd_send_queue(), which allows us to eliminate a lot of very similar code and makes the looping logic in kauditd_thread() clearer. * All netlink messages sent to auditd are now sent via auditd_send_unicast_skb(). Other than just making sense, this makes the lock handling easier. * Change the audit_log_start() sleep behavior so that we never sleep on auditd events (unchanged) or if the caller is holding the audit_cmd_mutex (changed). Previously we didn't sleep if the caller was auditd or if the message type fell between a certain range; the type check was a poor effort of doing what the cmd_mutex check now does. Richard Guy Briggs originally proposed not sleeping the cmd_mutex owner several years ago but his patch wasn't acceptable at the time. At least the idea lives on here. * A problem with the lost record counter has been resolved. Steve Grubb and I both happened to notice this problem and according to some quick testing by Steve, this problem goes back quite some time. It's largely a harmless problem, although it may have left some careful sysadmins quite puzzled. Cc: <stable@vger.kernel.org> # 4.10.x- Reported-by: Dmitry Vyukov <dvyukov@google.com> Signed-off-by: Paul Moore <paul@paul-moore.com>
2017-03-21 23:26:35 +08:00
continue;
}
audit: improve robustness of the audit queue handling If the audit daemon were ever to get stuck in a stopped state the kernel's kauditd_thread() could get blocked attempting to send audit records to the userspace audit daemon. With the kernel thread blocked it is possible that the audit queue could grow unbounded as certain audit record generating events must be exempt from the queue limits else the system enter a deadlock state. This patch resolves this problem by lowering the kernel thread's socket sending timeout from MAX_SCHEDULE_TIMEOUT to HZ/10 and tweaks the kauditd_send_queue() function to better manage the various audit queues when connection problems occur between the kernel and the audit daemon. With this patch, the backlog may temporarily grow beyond the defined limits when the audit daemon is stopped and the system is under heavy audit pressure, but kauditd_thread() will continue to make progress and drain the queues as it would for other connection problems. For example, with the audit daemon put into a stopped state and the system configured to audit every syscall it was still possible to shutdown the system without a kernel panic, deadlock, etc.; granted, the system was slow to shutdown but that is to be expected given the extreme pressure of recording every syscall. The timeout value of HZ/10 was chosen primarily through experimentation and this developer's "gut feeling". There is likely no one perfect value, but as this scenario is limited in scope (root privileges would be needed to send SIGSTOP to the audit daemon), it is likely not worth exposing this as a tunable at present. This can always be done at a later date if it proves necessary. Cc: stable@vger.kernel.org Fixes: 5b52330bbfe63 ("audit: fix auditd/kernel connection state tracking") Reported-by: Gaosheng Cui <cuigaosheng1@huawei.com> Tested-by: Gaosheng Cui <cuigaosheng1@huawei.com> Reviewed-by: Richard Guy Briggs <rgb@redhat.com> Signed-off-by: Paul Moore <paul@paul-moore.com>
2021-12-10 00:46:07 +08:00
retry:
audit: fix auditd/kernel connection state tracking What started as a rather straightforward race condition reported by Dmitry using the syzkaller fuzzer ended up revealing some major problems with how the audit subsystem managed its netlink sockets and its connection with the userspace audit daemon. Fixing this properly had quite the cascading effect and what we are left with is this rather large and complicated patch. My initial goal was to try and decompose this patch into multiple smaller patches, but the way these changes are intertwined makes it difficult to split these changes into meaningful pieces that don't break or somehow make things worse for the intermediate states. The patch makes a number of changes, but the most significant are highlighted below: * The auditd tracking variables, e.g. audit_sock, are now gone and replaced by a RCU/spin_lock protected variable auditd_conn which is a structure containing all of the auditd tracking information. * We no longer track the auditd sock directly, instead we track it via the network namespace in which it resides and we use the audit socket associated with that namespace. In spirit, this is what the code was trying to do prior to this patch (at least I think that is what the original authors intended), but it was done rather poorly and added a layer of obfuscation that only masked the underlying problems. * Big backlog queue cleanup, again. In v4.10 we made some pretty big changes to how the audit backlog queues work, here we haven't changed the queue design so much as cleaned up the implementation. Brought about by the locking changes, we've simplified kauditd_thread() quite a bit by consolidating the queue handling into a new helper function, kauditd_send_queue(), which allows us to eliminate a lot of very similar code and makes the looping logic in kauditd_thread() clearer. * All netlink messages sent to auditd are now sent via auditd_send_unicast_skb(). Other than just making sense, this makes the lock handling easier. * Change the audit_log_start() sleep behavior so that we never sleep on auditd events (unchanged) or if the caller is holding the audit_cmd_mutex (changed). Previously we didn't sleep if the caller was auditd or if the message type fell between a certain range; the type check was a poor effort of doing what the cmd_mutex check now does. Richard Guy Briggs originally proposed not sleeping the cmd_mutex owner several years ago but his patch wasn't acceptable at the time. At least the idea lives on here. * A problem with the lost record counter has been resolved. Steve Grubb and I both happened to notice this problem and according to some quick testing by Steve, this problem goes back quite some time. It's largely a harmless problem, although it may have left some careful sysadmins quite puzzled. Cc: <stable@vger.kernel.org> # 4.10.x- Reported-by: Dmitry Vyukov <dvyukov@google.com> Signed-off-by: Paul Moore <paul@paul-moore.com>
2017-03-21 23:26:35 +08:00
/* grab an extra skb reference in case of error */
skb_get(skb);
rc = netlink_unicast(sk, skb, portid, 0);
if (rc < 0) {
audit: improve robustness of the audit queue handling If the audit daemon were ever to get stuck in a stopped state the kernel's kauditd_thread() could get blocked attempting to send audit records to the userspace audit daemon. With the kernel thread blocked it is possible that the audit queue could grow unbounded as certain audit record generating events must be exempt from the queue limits else the system enter a deadlock state. This patch resolves this problem by lowering the kernel thread's socket sending timeout from MAX_SCHEDULE_TIMEOUT to HZ/10 and tweaks the kauditd_send_queue() function to better manage the various audit queues when connection problems occur between the kernel and the audit daemon. With this patch, the backlog may temporarily grow beyond the defined limits when the audit daemon is stopped and the system is under heavy audit pressure, but kauditd_thread() will continue to make progress and drain the queues as it would for other connection problems. For example, with the audit daemon put into a stopped state and the system configured to audit every syscall it was still possible to shutdown the system without a kernel panic, deadlock, etc.; granted, the system was slow to shutdown but that is to be expected given the extreme pressure of recording every syscall. The timeout value of HZ/10 was chosen primarily through experimentation and this developer's "gut feeling". There is likely no one perfect value, but as this scenario is limited in scope (root privileges would be needed to send SIGSTOP to the audit daemon), it is likely not worth exposing this as a tunable at present. This can always be done at a later date if it proves necessary. Cc: stable@vger.kernel.org Fixes: 5b52330bbfe63 ("audit: fix auditd/kernel connection state tracking") Reported-by: Gaosheng Cui <cuigaosheng1@huawei.com> Tested-by: Gaosheng Cui <cuigaosheng1@huawei.com> Reviewed-by: Richard Guy Briggs <rgb@redhat.com> Signed-off-by: Paul Moore <paul@paul-moore.com>
2021-12-10 00:46:07 +08:00
/* send failed - try a few times unless fatal error */
audit: fix auditd/kernel connection state tracking What started as a rather straightforward race condition reported by Dmitry using the syzkaller fuzzer ended up revealing some major problems with how the audit subsystem managed its netlink sockets and its connection with the userspace audit daemon. Fixing this properly had quite the cascading effect and what we are left with is this rather large and complicated patch. My initial goal was to try and decompose this patch into multiple smaller patches, but the way these changes are intertwined makes it difficult to split these changes into meaningful pieces that don't break or somehow make things worse for the intermediate states. The patch makes a number of changes, but the most significant are highlighted below: * The auditd tracking variables, e.g. audit_sock, are now gone and replaced by a RCU/spin_lock protected variable auditd_conn which is a structure containing all of the auditd tracking information. * We no longer track the auditd sock directly, instead we track it via the network namespace in which it resides and we use the audit socket associated with that namespace. In spirit, this is what the code was trying to do prior to this patch (at least I think that is what the original authors intended), but it was done rather poorly and added a layer of obfuscation that only masked the underlying problems. * Big backlog queue cleanup, again. In v4.10 we made some pretty big changes to how the audit backlog queues work, here we haven't changed the queue design so much as cleaned up the implementation. Brought about by the locking changes, we've simplified kauditd_thread() quite a bit by consolidating the queue handling into a new helper function, kauditd_send_queue(), which allows us to eliminate a lot of very similar code and makes the looping logic in kauditd_thread() clearer. * All netlink messages sent to auditd are now sent via auditd_send_unicast_skb(). Other than just making sense, this makes the lock handling easier. * Change the audit_log_start() sleep behavior so that we never sleep on auditd events (unchanged) or if the caller is holding the audit_cmd_mutex (changed). Previously we didn't sleep if the caller was auditd or if the message type fell between a certain range; the type check was a poor effort of doing what the cmd_mutex check now does. Richard Guy Briggs originally proposed not sleeping the cmd_mutex owner several years ago but his patch wasn't acceptable at the time. At least the idea lives on here. * A problem with the lost record counter has been resolved. Steve Grubb and I both happened to notice this problem and according to some quick testing by Steve, this problem goes back quite some time. It's largely a harmless problem, although it may have left some careful sysadmins quite puzzled. Cc: <stable@vger.kernel.org> # 4.10.x- Reported-by: Dmitry Vyukov <dvyukov@google.com> Signed-off-by: Paul Moore <paul@paul-moore.com>
2017-03-21 23:26:35 +08:00
if (++failed >= retry_limit ||
rc == -ECONNREFUSED || rc == -EPERM) {
sk = NULL;
if (err_hook)
audit: improve audit queue handling when "audit=1" on cmdline When an admin enables audit at early boot via the "audit=1" kernel command line the audit queue behavior is slightly different; the audit subsystem goes to greater lengths to avoid dropping records, which unfortunately can result in problems when the audit daemon is forcibly stopped for an extended period of time. This patch makes a number of changes designed to improve the audit queuing behavior so that leaving the audit daemon in a stopped state for an extended period does not cause a significant impact to the system. - kauditd_send_queue() is now limited to looping through the passed queue only once per call. This not only prevents the function from looping indefinitely when records are returned to the current queue, it also allows any recovery handling in kauditd_thread() to take place when kauditd_send_queue() returns. - Transient netlink send errors seen as -EAGAIN now cause the record to be returned to the retry queue instead of going to the hold queue. The intention of the hold queue is to store, perhaps for an extended period of time, the events which led up to the audit daemon going offline. The retry queue remains a temporary queue intended to protect against transient issues between the kernel and the audit daemon. - The retry queue is now limited by the audit_backlog_limit setting, the same as the other queues. This allows admins to bound the size of all of the audit queues on the system. - kauditd_rehold_skb() now returns records to the end of the hold queue to ensure ordering is preserved in the face of recent changes to kauditd_send_queue(). Cc: stable@vger.kernel.org Fixes: 5b52330bbfe63 ("audit: fix auditd/kernel connection state tracking") Fixes: f4b3ee3c85551 ("audit: improve robustness of the audit queue handling") Reported-by: Gaosheng Cui <cuigaosheng1@huawei.com> Tested-by: Gaosheng Cui <cuigaosheng1@huawei.com> Reviewed-by: Richard Guy Briggs <rgb@redhat.com> Signed-off-by: Paul Moore <paul@paul-moore.com>
2022-01-14 07:54:38 +08:00
(*err_hook)(skb, rc);
audit: improve robustness of the audit queue handling If the audit daemon were ever to get stuck in a stopped state the kernel's kauditd_thread() could get blocked attempting to send audit records to the userspace audit daemon. With the kernel thread blocked it is possible that the audit queue could grow unbounded as certain audit record generating events must be exempt from the queue limits else the system enter a deadlock state. This patch resolves this problem by lowering the kernel thread's socket sending timeout from MAX_SCHEDULE_TIMEOUT to HZ/10 and tweaks the kauditd_send_queue() function to better manage the various audit queues when connection problems occur between the kernel and the audit daemon. With this patch, the backlog may temporarily grow beyond the defined limits when the audit daemon is stopped and the system is under heavy audit pressure, but kauditd_thread() will continue to make progress and drain the queues as it would for other connection problems. For example, with the audit daemon put into a stopped state and the system configured to audit every syscall it was still possible to shutdown the system without a kernel panic, deadlock, etc.; granted, the system was slow to shutdown but that is to be expected given the extreme pressure of recording every syscall. The timeout value of HZ/10 was chosen primarily through experimentation and this developer's "gut feeling". There is likely no one perfect value, but as this scenario is limited in scope (root privileges would be needed to send SIGSTOP to the audit daemon), it is likely not worth exposing this as a tunable at present. This can always be done at a later date if it proves necessary. Cc: stable@vger.kernel.org Fixes: 5b52330bbfe63 ("audit: fix auditd/kernel connection state tracking") Reported-by: Gaosheng Cui <cuigaosheng1@huawei.com> Tested-by: Gaosheng Cui <cuigaosheng1@huawei.com> Reviewed-by: Richard Guy Briggs <rgb@redhat.com> Signed-off-by: Paul Moore <paul@paul-moore.com>
2021-12-10 00:46:07 +08:00
if (rc == -EAGAIN)
rc = 0;
/* continue to drain the queue */
audit: fix auditd/kernel connection state tracking What started as a rather straightforward race condition reported by Dmitry using the syzkaller fuzzer ended up revealing some major problems with how the audit subsystem managed its netlink sockets and its connection with the userspace audit daemon. Fixing this properly had quite the cascading effect and what we are left with is this rather large and complicated patch. My initial goal was to try and decompose this patch into multiple smaller patches, but the way these changes are intertwined makes it difficult to split these changes into meaningful pieces that don't break or somehow make things worse for the intermediate states. The patch makes a number of changes, but the most significant are highlighted below: * The auditd tracking variables, e.g. audit_sock, are now gone and replaced by a RCU/spin_lock protected variable auditd_conn which is a structure containing all of the auditd tracking information. * We no longer track the auditd sock directly, instead we track it via the network namespace in which it resides and we use the audit socket associated with that namespace. In spirit, this is what the code was trying to do prior to this patch (at least I think that is what the original authors intended), but it was done rather poorly and added a layer of obfuscation that only masked the underlying problems. * Big backlog queue cleanup, again. In v4.10 we made some pretty big changes to how the audit backlog queues work, here we haven't changed the queue design so much as cleaned up the implementation. Brought about by the locking changes, we've simplified kauditd_thread() quite a bit by consolidating the queue handling into a new helper function, kauditd_send_queue(), which allows us to eliminate a lot of very similar code and makes the looping logic in kauditd_thread() clearer. * All netlink messages sent to auditd are now sent via auditd_send_unicast_skb(). Other than just making sense, this makes the lock handling easier. * Change the audit_log_start() sleep behavior so that we never sleep on auditd events (unchanged) or if the caller is holding the audit_cmd_mutex (changed). Previously we didn't sleep if the caller was auditd or if the message type fell between a certain range; the type check was a poor effort of doing what the cmd_mutex check now does. Richard Guy Briggs originally proposed not sleeping the cmd_mutex owner several years ago but his patch wasn't acceptable at the time. At least the idea lives on here. * A problem with the lost record counter has been resolved. Steve Grubb and I both happened to notice this problem and according to some quick testing by Steve, this problem goes back quite some time. It's largely a harmless problem, although it may have left some careful sysadmins quite puzzled. Cc: <stable@vger.kernel.org> # 4.10.x- Reported-by: Dmitry Vyukov <dvyukov@google.com> Signed-off-by: Paul Moore <paul@paul-moore.com>
2017-03-21 23:26:35 +08:00
continue;
} else
audit: improve robustness of the audit queue handling If the audit daemon were ever to get stuck in a stopped state the kernel's kauditd_thread() could get blocked attempting to send audit records to the userspace audit daemon. With the kernel thread blocked it is possible that the audit queue could grow unbounded as certain audit record generating events must be exempt from the queue limits else the system enter a deadlock state. This patch resolves this problem by lowering the kernel thread's socket sending timeout from MAX_SCHEDULE_TIMEOUT to HZ/10 and tweaks the kauditd_send_queue() function to better manage the various audit queues when connection problems occur between the kernel and the audit daemon. With this patch, the backlog may temporarily grow beyond the defined limits when the audit daemon is stopped and the system is under heavy audit pressure, but kauditd_thread() will continue to make progress and drain the queues as it would for other connection problems. For example, with the audit daemon put into a stopped state and the system configured to audit every syscall it was still possible to shutdown the system without a kernel panic, deadlock, etc.; granted, the system was slow to shutdown but that is to be expected given the extreme pressure of recording every syscall. The timeout value of HZ/10 was chosen primarily through experimentation and this developer's "gut feeling". There is likely no one perfect value, but as this scenario is limited in scope (root privileges would be needed to send SIGSTOP to the audit daemon), it is likely not worth exposing this as a tunable at present. This can always be done at a later date if it proves necessary. Cc: stable@vger.kernel.org Fixes: 5b52330bbfe63 ("audit: fix auditd/kernel connection state tracking") Reported-by: Gaosheng Cui <cuigaosheng1@huawei.com> Tested-by: Gaosheng Cui <cuigaosheng1@huawei.com> Reviewed-by: Richard Guy Briggs <rgb@redhat.com> Signed-off-by: Paul Moore <paul@paul-moore.com>
2021-12-10 00:46:07 +08:00
goto retry;
audit: fix auditd/kernel connection state tracking What started as a rather straightforward race condition reported by Dmitry using the syzkaller fuzzer ended up revealing some major problems with how the audit subsystem managed its netlink sockets and its connection with the userspace audit daemon. Fixing this properly had quite the cascading effect and what we are left with is this rather large and complicated patch. My initial goal was to try and decompose this patch into multiple smaller patches, but the way these changes are intertwined makes it difficult to split these changes into meaningful pieces that don't break or somehow make things worse for the intermediate states. The patch makes a number of changes, but the most significant are highlighted below: * The auditd tracking variables, e.g. audit_sock, are now gone and replaced by a RCU/spin_lock protected variable auditd_conn which is a structure containing all of the auditd tracking information. * We no longer track the auditd sock directly, instead we track it via the network namespace in which it resides and we use the audit socket associated with that namespace. In spirit, this is what the code was trying to do prior to this patch (at least I think that is what the original authors intended), but it was done rather poorly and added a layer of obfuscation that only masked the underlying problems. * Big backlog queue cleanup, again. In v4.10 we made some pretty big changes to how the audit backlog queues work, here we haven't changed the queue design so much as cleaned up the implementation. Brought about by the locking changes, we've simplified kauditd_thread() quite a bit by consolidating the queue handling into a new helper function, kauditd_send_queue(), which allows us to eliminate a lot of very similar code and makes the looping logic in kauditd_thread() clearer. * All netlink messages sent to auditd are now sent via auditd_send_unicast_skb(). Other than just making sense, this makes the lock handling easier. * Change the audit_log_start() sleep behavior so that we never sleep on auditd events (unchanged) or if the caller is holding the audit_cmd_mutex (changed). Previously we didn't sleep if the caller was auditd or if the message type fell between a certain range; the type check was a poor effort of doing what the cmd_mutex check now does. Richard Guy Briggs originally proposed not sleeping the cmd_mutex owner several years ago but his patch wasn't acceptable at the time. At least the idea lives on here. * A problem with the lost record counter has been resolved. Steve Grubb and I both happened to notice this problem and according to some quick testing by Steve, this problem goes back quite some time. It's largely a harmless problem, although it may have left some careful sysadmins quite puzzled. Cc: <stable@vger.kernel.org> # 4.10.x- Reported-by: Dmitry Vyukov <dvyukov@google.com> Signed-off-by: Paul Moore <paul@paul-moore.com>
2017-03-21 23:26:35 +08:00
} else {
audit: improve robustness of the audit queue handling If the audit daemon were ever to get stuck in a stopped state the kernel's kauditd_thread() could get blocked attempting to send audit records to the userspace audit daemon. With the kernel thread blocked it is possible that the audit queue could grow unbounded as certain audit record generating events must be exempt from the queue limits else the system enter a deadlock state. This patch resolves this problem by lowering the kernel thread's socket sending timeout from MAX_SCHEDULE_TIMEOUT to HZ/10 and tweaks the kauditd_send_queue() function to better manage the various audit queues when connection problems occur between the kernel and the audit daemon. With this patch, the backlog may temporarily grow beyond the defined limits when the audit daemon is stopped and the system is under heavy audit pressure, but kauditd_thread() will continue to make progress and drain the queues as it would for other connection problems. For example, with the audit daemon put into a stopped state and the system configured to audit every syscall it was still possible to shutdown the system without a kernel panic, deadlock, etc.; granted, the system was slow to shutdown but that is to be expected given the extreme pressure of recording every syscall. The timeout value of HZ/10 was chosen primarily through experimentation and this developer's "gut feeling". There is likely no one perfect value, but as this scenario is limited in scope (root privileges would be needed to send SIGSTOP to the audit daemon), it is likely not worth exposing this as a tunable at present. This can always be done at a later date if it proves necessary. Cc: stable@vger.kernel.org Fixes: 5b52330bbfe63 ("audit: fix auditd/kernel connection state tracking") Reported-by: Gaosheng Cui <cuigaosheng1@huawei.com> Tested-by: Gaosheng Cui <cuigaosheng1@huawei.com> Reviewed-by: Richard Guy Briggs <rgb@redhat.com> Signed-off-by: Paul Moore <paul@paul-moore.com>
2021-12-10 00:46:07 +08:00
/* skb sent - drop the extra reference and continue */
audit: fix auditd/kernel connection state tracking What started as a rather straightforward race condition reported by Dmitry using the syzkaller fuzzer ended up revealing some major problems with how the audit subsystem managed its netlink sockets and its connection with the userspace audit daemon. Fixing this properly had quite the cascading effect and what we are left with is this rather large and complicated patch. My initial goal was to try and decompose this patch into multiple smaller patches, but the way these changes are intertwined makes it difficult to split these changes into meaningful pieces that don't break or somehow make things worse for the intermediate states. The patch makes a number of changes, but the most significant are highlighted below: * The auditd tracking variables, e.g. audit_sock, are now gone and replaced by a RCU/spin_lock protected variable auditd_conn which is a structure containing all of the auditd tracking information. * We no longer track the auditd sock directly, instead we track it via the network namespace in which it resides and we use the audit socket associated with that namespace. In spirit, this is what the code was trying to do prior to this patch (at least I think that is what the original authors intended), but it was done rather poorly and added a layer of obfuscation that only masked the underlying problems. * Big backlog queue cleanup, again. In v4.10 we made some pretty big changes to how the audit backlog queues work, here we haven't changed the queue design so much as cleaned up the implementation. Brought about by the locking changes, we've simplified kauditd_thread() quite a bit by consolidating the queue handling into a new helper function, kauditd_send_queue(), which allows us to eliminate a lot of very similar code and makes the looping logic in kauditd_thread() clearer. * All netlink messages sent to auditd are now sent via auditd_send_unicast_skb(). Other than just making sense, this makes the lock handling easier. * Change the audit_log_start() sleep behavior so that we never sleep on auditd events (unchanged) or if the caller is holding the audit_cmd_mutex (changed). Previously we didn't sleep if the caller was auditd or if the message type fell between a certain range; the type check was a poor effort of doing what the cmd_mutex check now does. Richard Guy Briggs originally proposed not sleeping the cmd_mutex owner several years ago but his patch wasn't acceptable at the time. At least the idea lives on here. * A problem with the lost record counter has been resolved. Steve Grubb and I both happened to notice this problem and according to some quick testing by Steve, this problem goes back quite some time. It's largely a harmless problem, although it may have left some careful sysadmins quite puzzled. Cc: <stable@vger.kernel.org> # 4.10.x- Reported-by: Dmitry Vyukov <dvyukov@google.com> Signed-off-by: Paul Moore <paul@paul-moore.com>
2017-03-21 23:26:35 +08:00
consume_skb(skb);
failed = 0;
}
}
audit: fix auditd/kernel connection state tracking What started as a rather straightforward race condition reported by Dmitry using the syzkaller fuzzer ended up revealing some major problems with how the audit subsystem managed its netlink sockets and its connection with the userspace audit daemon. Fixing this properly had quite the cascading effect and what we are left with is this rather large and complicated patch. My initial goal was to try and decompose this patch into multiple smaller patches, but the way these changes are intertwined makes it difficult to split these changes into meaningful pieces that don't break or somehow make things worse for the intermediate states. The patch makes a number of changes, but the most significant are highlighted below: * The auditd tracking variables, e.g. audit_sock, are now gone and replaced by a RCU/spin_lock protected variable auditd_conn which is a structure containing all of the auditd tracking information. * We no longer track the auditd sock directly, instead we track it via the network namespace in which it resides and we use the audit socket associated with that namespace. In spirit, this is what the code was trying to do prior to this patch (at least I think that is what the original authors intended), but it was done rather poorly and added a layer of obfuscation that only masked the underlying problems. * Big backlog queue cleanup, again. In v4.10 we made some pretty big changes to how the audit backlog queues work, here we haven't changed the queue design so much as cleaned up the implementation. Brought about by the locking changes, we've simplified kauditd_thread() quite a bit by consolidating the queue handling into a new helper function, kauditd_send_queue(), which allows us to eliminate a lot of very similar code and makes the looping logic in kauditd_thread() clearer. * All netlink messages sent to auditd are now sent via auditd_send_unicast_skb(). Other than just making sense, this makes the lock handling easier. * Change the audit_log_start() sleep behavior so that we never sleep on auditd events (unchanged) or if the caller is holding the audit_cmd_mutex (changed). Previously we didn't sleep if the caller was auditd or if the message type fell between a certain range; the type check was a poor effort of doing what the cmd_mutex check now does. Richard Guy Briggs originally proposed not sleeping the cmd_mutex owner several years ago but his patch wasn't acceptable at the time. At least the idea lives on here. * A problem with the lost record counter has been resolved. Steve Grubb and I both happened to notice this problem and according to some quick testing by Steve, this problem goes back quite some time. It's largely a harmless problem, although it may have left some careful sysadmins quite puzzled. Cc: <stable@vger.kernel.org> # 4.10.x- Reported-by: Dmitry Vyukov <dvyukov@google.com> Signed-off-by: Paul Moore <paul@paul-moore.com>
2017-03-21 23:26:35 +08:00
return (rc >= 0 ? 0 : rc);
}
/*
* kauditd_send_multicast_skb - Send a record to any multicast listeners
* @skb: audit record
*
* Description:
audit: fix auditd/kernel connection state tracking What started as a rather straightforward race condition reported by Dmitry using the syzkaller fuzzer ended up revealing some major problems with how the audit subsystem managed its netlink sockets and its connection with the userspace audit daemon. Fixing this properly had quite the cascading effect and what we are left with is this rather large and complicated patch. My initial goal was to try and decompose this patch into multiple smaller patches, but the way these changes are intertwined makes it difficult to split these changes into meaningful pieces that don't break or somehow make things worse for the intermediate states. The patch makes a number of changes, but the most significant are highlighted below: * The auditd tracking variables, e.g. audit_sock, are now gone and replaced by a RCU/spin_lock protected variable auditd_conn which is a structure containing all of the auditd tracking information. * We no longer track the auditd sock directly, instead we track it via the network namespace in which it resides and we use the audit socket associated with that namespace. In spirit, this is what the code was trying to do prior to this patch (at least I think that is what the original authors intended), but it was done rather poorly and added a layer of obfuscation that only masked the underlying problems. * Big backlog queue cleanup, again. In v4.10 we made some pretty big changes to how the audit backlog queues work, here we haven't changed the queue design so much as cleaned up the implementation. Brought about by the locking changes, we've simplified kauditd_thread() quite a bit by consolidating the queue handling into a new helper function, kauditd_send_queue(), which allows us to eliminate a lot of very similar code and makes the looping logic in kauditd_thread() clearer. * All netlink messages sent to auditd are now sent via auditd_send_unicast_skb(). Other than just making sense, this makes the lock handling easier. * Change the audit_log_start() sleep behavior so that we never sleep on auditd events (unchanged) or if the caller is holding the audit_cmd_mutex (changed). Previously we didn't sleep if the caller was auditd or if the message type fell between a certain range; the type check was a poor effort of doing what the cmd_mutex check now does. Richard Guy Briggs originally proposed not sleeping the cmd_mutex owner several years ago but his patch wasn't acceptable at the time. At least the idea lives on here. * A problem with the lost record counter has been resolved. Steve Grubb and I both happened to notice this problem and according to some quick testing by Steve, this problem goes back quite some time. It's largely a harmless problem, although it may have left some careful sysadmins quite puzzled. Cc: <stable@vger.kernel.org> # 4.10.x- Reported-by: Dmitry Vyukov <dvyukov@google.com> Signed-off-by: Paul Moore <paul@paul-moore.com>
2017-03-21 23:26:35 +08:00
* Write a multicast message to anyone listening in the initial network
* namespace. This function doesn't consume an skb as might be expected since
* it has to copy it anyways.
*/
static void kauditd_send_multicast_skb(struct sk_buff *skb)
{
struct sk_buff *copy;
audit: fix auditd/kernel connection state tracking What started as a rather straightforward race condition reported by Dmitry using the syzkaller fuzzer ended up revealing some major problems with how the audit subsystem managed its netlink sockets and its connection with the userspace audit daemon. Fixing this properly had quite the cascading effect and what we are left with is this rather large and complicated patch. My initial goal was to try and decompose this patch into multiple smaller patches, but the way these changes are intertwined makes it difficult to split these changes into meaningful pieces that don't break or somehow make things worse for the intermediate states. The patch makes a number of changes, but the most significant are highlighted below: * The auditd tracking variables, e.g. audit_sock, are now gone and replaced by a RCU/spin_lock protected variable auditd_conn which is a structure containing all of the auditd tracking information. * We no longer track the auditd sock directly, instead we track it via the network namespace in which it resides and we use the audit socket associated with that namespace. In spirit, this is what the code was trying to do prior to this patch (at least I think that is what the original authors intended), but it was done rather poorly and added a layer of obfuscation that only masked the underlying problems. * Big backlog queue cleanup, again. In v4.10 we made some pretty big changes to how the audit backlog queues work, here we haven't changed the queue design so much as cleaned up the implementation. Brought about by the locking changes, we've simplified kauditd_thread() quite a bit by consolidating the queue handling into a new helper function, kauditd_send_queue(), which allows us to eliminate a lot of very similar code and makes the looping logic in kauditd_thread() clearer. * All netlink messages sent to auditd are now sent via auditd_send_unicast_skb(). Other than just making sense, this makes the lock handling easier. * Change the audit_log_start() sleep behavior so that we never sleep on auditd events (unchanged) or if the caller is holding the audit_cmd_mutex (changed). Previously we didn't sleep if the caller was auditd or if the message type fell between a certain range; the type check was a poor effort of doing what the cmd_mutex check now does. Richard Guy Briggs originally proposed not sleeping the cmd_mutex owner several years ago but his patch wasn't acceptable at the time. At least the idea lives on here. * A problem with the lost record counter has been resolved. Steve Grubb and I both happened to notice this problem and according to some quick testing by Steve, this problem goes back quite some time. It's largely a harmless problem, although it may have left some careful sysadmins quite puzzled. Cc: <stable@vger.kernel.org> # 4.10.x- Reported-by: Dmitry Vyukov <dvyukov@google.com> Signed-off-by: Paul Moore <paul@paul-moore.com>
2017-03-21 23:26:35 +08:00
struct sock *sock = audit_get_sk(&init_net);
struct nlmsghdr *nlh;
audit: fix auditd/kernel connection state tracking What started as a rather straightforward race condition reported by Dmitry using the syzkaller fuzzer ended up revealing some major problems with how the audit subsystem managed its netlink sockets and its connection with the userspace audit daemon. Fixing this properly had quite the cascading effect and what we are left with is this rather large and complicated patch. My initial goal was to try and decompose this patch into multiple smaller patches, but the way these changes are intertwined makes it difficult to split these changes into meaningful pieces that don't break or somehow make things worse for the intermediate states. The patch makes a number of changes, but the most significant are highlighted below: * The auditd tracking variables, e.g. audit_sock, are now gone and replaced by a RCU/spin_lock protected variable auditd_conn which is a structure containing all of the auditd tracking information. * We no longer track the auditd sock directly, instead we track it via the network namespace in which it resides and we use the audit socket associated with that namespace. In spirit, this is what the code was trying to do prior to this patch (at least I think that is what the original authors intended), but it was done rather poorly and added a layer of obfuscation that only masked the underlying problems. * Big backlog queue cleanup, again. In v4.10 we made some pretty big changes to how the audit backlog queues work, here we haven't changed the queue design so much as cleaned up the implementation. Brought about by the locking changes, we've simplified kauditd_thread() quite a bit by consolidating the queue handling into a new helper function, kauditd_send_queue(), which allows us to eliminate a lot of very similar code and makes the looping logic in kauditd_thread() clearer. * All netlink messages sent to auditd are now sent via auditd_send_unicast_skb(). Other than just making sense, this makes the lock handling easier. * Change the audit_log_start() sleep behavior so that we never sleep on auditd events (unchanged) or if the caller is holding the audit_cmd_mutex (changed). Previously we didn't sleep if the caller was auditd or if the message type fell between a certain range; the type check was a poor effort of doing what the cmd_mutex check now does. Richard Guy Briggs originally proposed not sleeping the cmd_mutex owner several years ago but his patch wasn't acceptable at the time. At least the idea lives on here. * A problem with the lost record counter has been resolved. Steve Grubb and I both happened to notice this problem and according to some quick testing by Steve, this problem goes back quite some time. It's largely a harmless problem, although it may have left some careful sysadmins quite puzzled. Cc: <stable@vger.kernel.org> # 4.10.x- Reported-by: Dmitry Vyukov <dvyukov@google.com> Signed-off-by: Paul Moore <paul@paul-moore.com>
2017-03-21 23:26:35 +08:00
/* NOTE: we are not taking an additional reference for init_net since
* we don't have to worry about it going away */
if (!netlink_has_listeners(sock, AUDIT_NLGRP_READLOG))
return;
/*
* The seemingly wasteful skb_copy() rather than bumping the refcount
* using skb_get() is necessary because non-standard mods are made to
* the skb by the original kaudit unicast socket send routine. The
* existing auditd daemon assumes this breakage. Fixing this would
* require co-ordinating a change in the established protocol between
* the kaudit kernel subsystem and the auditd userspace code. There is
* no reason for new multicast clients to continue with this
* non-compliance.
*/
copy = skb_copy(skb, GFP_KERNEL);
if (!copy)
return;
nlh = nlmsg_hdr(copy);
nlh->nlmsg_len = skb->len;
nlmsg_multicast(sock, copy, 0, AUDIT_NLGRP_READLOG, GFP_KERNEL);
}
/**
audit: fix auditd/kernel connection state tracking What started as a rather straightforward race condition reported by Dmitry using the syzkaller fuzzer ended up revealing some major problems with how the audit subsystem managed its netlink sockets and its connection with the userspace audit daemon. Fixing this properly had quite the cascading effect and what we are left with is this rather large and complicated patch. My initial goal was to try and decompose this patch into multiple smaller patches, but the way these changes are intertwined makes it difficult to split these changes into meaningful pieces that don't break or somehow make things worse for the intermediate states. The patch makes a number of changes, but the most significant are highlighted below: * The auditd tracking variables, e.g. audit_sock, are now gone and replaced by a RCU/spin_lock protected variable auditd_conn which is a structure containing all of the auditd tracking information. * We no longer track the auditd sock directly, instead we track it via the network namespace in which it resides and we use the audit socket associated with that namespace. In spirit, this is what the code was trying to do prior to this patch (at least I think that is what the original authors intended), but it was done rather poorly and added a layer of obfuscation that only masked the underlying problems. * Big backlog queue cleanup, again. In v4.10 we made some pretty big changes to how the audit backlog queues work, here we haven't changed the queue design so much as cleaned up the implementation. Brought about by the locking changes, we've simplified kauditd_thread() quite a bit by consolidating the queue handling into a new helper function, kauditd_send_queue(), which allows us to eliminate a lot of very similar code and makes the looping logic in kauditd_thread() clearer. * All netlink messages sent to auditd are now sent via auditd_send_unicast_skb(). Other than just making sense, this makes the lock handling easier. * Change the audit_log_start() sleep behavior so that we never sleep on auditd events (unchanged) or if the caller is holding the audit_cmd_mutex (changed). Previously we didn't sleep if the caller was auditd or if the message type fell between a certain range; the type check was a poor effort of doing what the cmd_mutex check now does. Richard Guy Briggs originally proposed not sleeping the cmd_mutex owner several years ago but his patch wasn't acceptable at the time. At least the idea lives on here. * A problem with the lost record counter has been resolved. Steve Grubb and I both happened to notice this problem and according to some quick testing by Steve, this problem goes back quite some time. It's largely a harmless problem, although it may have left some careful sysadmins quite puzzled. Cc: <stable@vger.kernel.org> # 4.10.x- Reported-by: Dmitry Vyukov <dvyukov@google.com> Signed-off-by: Paul Moore <paul@paul-moore.com>
2017-03-21 23:26:35 +08:00
* kauditd_thread - Worker thread to send audit records to userspace
* @dummy: unused
*/
static int kauditd_thread(void *dummy)
{
int rc;
audit: fix auditd/kernel connection state tracking What started as a rather straightforward race condition reported by Dmitry using the syzkaller fuzzer ended up revealing some major problems with how the audit subsystem managed its netlink sockets and its connection with the userspace audit daemon. Fixing this properly had quite the cascading effect and what we are left with is this rather large and complicated patch. My initial goal was to try and decompose this patch into multiple smaller patches, but the way these changes are intertwined makes it difficult to split these changes into meaningful pieces that don't break or somehow make things worse for the intermediate states. The patch makes a number of changes, but the most significant are highlighted below: * The auditd tracking variables, e.g. audit_sock, are now gone and replaced by a RCU/spin_lock protected variable auditd_conn which is a structure containing all of the auditd tracking information. * We no longer track the auditd sock directly, instead we track it via the network namespace in which it resides and we use the audit socket associated with that namespace. In spirit, this is what the code was trying to do prior to this patch (at least I think that is what the original authors intended), but it was done rather poorly and added a layer of obfuscation that only masked the underlying problems. * Big backlog queue cleanup, again. In v4.10 we made some pretty big changes to how the audit backlog queues work, here we haven't changed the queue design so much as cleaned up the implementation. Brought about by the locking changes, we've simplified kauditd_thread() quite a bit by consolidating the queue handling into a new helper function, kauditd_send_queue(), which allows us to eliminate a lot of very similar code and makes the looping logic in kauditd_thread() clearer. * All netlink messages sent to auditd are now sent via auditd_send_unicast_skb(). Other than just making sense, this makes the lock handling easier. * Change the audit_log_start() sleep behavior so that we never sleep on auditd events (unchanged) or if the caller is holding the audit_cmd_mutex (changed). Previously we didn't sleep if the caller was auditd or if the message type fell between a certain range; the type check was a poor effort of doing what the cmd_mutex check now does. Richard Guy Briggs originally proposed not sleeping the cmd_mutex owner several years ago but his patch wasn't acceptable at the time. At least the idea lives on here. * A problem with the lost record counter has been resolved. Steve Grubb and I both happened to notice this problem and according to some quick testing by Steve, this problem goes back quite some time. It's largely a harmless problem, although it may have left some careful sysadmins quite puzzled. Cc: <stable@vger.kernel.org> # 4.10.x- Reported-by: Dmitry Vyukov <dvyukov@google.com> Signed-off-by: Paul Moore <paul@paul-moore.com>
2017-03-21 23:26:35 +08:00
u32 portid = 0;
struct net *net = NULL;
struct sock *sk = NULL;
struct auditd_connection *ac;
#define UNICAST_RETRIES 5
set_freezable();
while (!kthread_should_stop()) {
audit: fix auditd/kernel connection state tracking What started as a rather straightforward race condition reported by Dmitry using the syzkaller fuzzer ended up revealing some major problems with how the audit subsystem managed its netlink sockets and its connection with the userspace audit daemon. Fixing this properly had quite the cascading effect and what we are left with is this rather large and complicated patch. My initial goal was to try and decompose this patch into multiple smaller patches, but the way these changes are intertwined makes it difficult to split these changes into meaningful pieces that don't break or somehow make things worse for the intermediate states. The patch makes a number of changes, but the most significant are highlighted below: * The auditd tracking variables, e.g. audit_sock, are now gone and replaced by a RCU/spin_lock protected variable auditd_conn which is a structure containing all of the auditd tracking information. * We no longer track the auditd sock directly, instead we track it via the network namespace in which it resides and we use the audit socket associated with that namespace. In spirit, this is what the code was trying to do prior to this patch (at least I think that is what the original authors intended), but it was done rather poorly and added a layer of obfuscation that only masked the underlying problems. * Big backlog queue cleanup, again. In v4.10 we made some pretty big changes to how the audit backlog queues work, here we haven't changed the queue design so much as cleaned up the implementation. Brought about by the locking changes, we've simplified kauditd_thread() quite a bit by consolidating the queue handling into a new helper function, kauditd_send_queue(), which allows us to eliminate a lot of very similar code and makes the looping logic in kauditd_thread() clearer. * All netlink messages sent to auditd are now sent via auditd_send_unicast_skb(). Other than just making sense, this makes the lock handling easier. * Change the audit_log_start() sleep behavior so that we never sleep on auditd events (unchanged) or if the caller is holding the audit_cmd_mutex (changed). Previously we didn't sleep if the caller was auditd or if the message type fell between a certain range; the type check was a poor effort of doing what the cmd_mutex check now does. Richard Guy Briggs originally proposed not sleeping the cmd_mutex owner several years ago but his patch wasn't acceptable at the time. At least the idea lives on here. * A problem with the lost record counter has been resolved. Steve Grubb and I both happened to notice this problem and according to some quick testing by Steve, this problem goes back quite some time. It's largely a harmless problem, although it may have left some careful sysadmins quite puzzled. Cc: <stable@vger.kernel.org> # 4.10.x- Reported-by: Dmitry Vyukov <dvyukov@google.com> Signed-off-by: Paul Moore <paul@paul-moore.com>
2017-03-21 23:26:35 +08:00
/* NOTE: see the lock comments in auditd_send_unicast_skb() */
rcu_read_lock();
ac = rcu_dereference(auditd_conn);
if (!ac) {
audit: fix auditd/kernel connection state tracking What started as a rather straightforward race condition reported by Dmitry using the syzkaller fuzzer ended up revealing some major problems with how the audit subsystem managed its netlink sockets and its connection with the userspace audit daemon. Fixing this properly had quite the cascading effect and what we are left with is this rather large and complicated patch. My initial goal was to try and decompose this patch into multiple smaller patches, but the way these changes are intertwined makes it difficult to split these changes into meaningful pieces that don't break or somehow make things worse for the intermediate states. The patch makes a number of changes, but the most significant are highlighted below: * The auditd tracking variables, e.g. audit_sock, are now gone and replaced by a RCU/spin_lock protected variable auditd_conn which is a structure containing all of the auditd tracking information. * We no longer track the auditd sock directly, instead we track it via the network namespace in which it resides and we use the audit socket associated with that namespace. In spirit, this is what the code was trying to do prior to this patch (at least I think that is what the original authors intended), but it was done rather poorly and added a layer of obfuscation that only masked the underlying problems. * Big backlog queue cleanup, again. In v4.10 we made some pretty big changes to how the audit backlog queues work, here we haven't changed the queue design so much as cleaned up the implementation. Brought about by the locking changes, we've simplified kauditd_thread() quite a bit by consolidating the queue handling into a new helper function, kauditd_send_queue(), which allows us to eliminate a lot of very similar code and makes the looping logic in kauditd_thread() clearer. * All netlink messages sent to auditd are now sent via auditd_send_unicast_skb(). Other than just making sense, this makes the lock handling easier. * Change the audit_log_start() sleep behavior so that we never sleep on auditd events (unchanged) or if the caller is holding the audit_cmd_mutex (changed). Previously we didn't sleep if the caller was auditd or if the message type fell between a certain range; the type check was a poor effort of doing what the cmd_mutex check now does. Richard Guy Briggs originally proposed not sleeping the cmd_mutex owner several years ago but his patch wasn't acceptable at the time. At least the idea lives on here. * A problem with the lost record counter has been resolved. Steve Grubb and I both happened to notice this problem and according to some quick testing by Steve, this problem goes back quite some time. It's largely a harmless problem, although it may have left some careful sysadmins quite puzzled. Cc: <stable@vger.kernel.org> # 4.10.x- Reported-by: Dmitry Vyukov <dvyukov@google.com> Signed-off-by: Paul Moore <paul@paul-moore.com>
2017-03-21 23:26:35 +08:00
rcu_read_unlock();
goto main_queue;
}
net = get_net(ac->net);
audit: fix auditd/kernel connection state tracking What started as a rather straightforward race condition reported by Dmitry using the syzkaller fuzzer ended up revealing some major problems with how the audit subsystem managed its netlink sockets and its connection with the userspace audit daemon. Fixing this properly had quite the cascading effect and what we are left with is this rather large and complicated patch. My initial goal was to try and decompose this patch into multiple smaller patches, but the way these changes are intertwined makes it difficult to split these changes into meaningful pieces that don't break or somehow make things worse for the intermediate states. The patch makes a number of changes, but the most significant are highlighted below: * The auditd tracking variables, e.g. audit_sock, are now gone and replaced by a RCU/spin_lock protected variable auditd_conn which is a structure containing all of the auditd tracking information. * We no longer track the auditd sock directly, instead we track it via the network namespace in which it resides and we use the audit socket associated with that namespace. In spirit, this is what the code was trying to do prior to this patch (at least I think that is what the original authors intended), but it was done rather poorly and added a layer of obfuscation that only masked the underlying problems. * Big backlog queue cleanup, again. In v4.10 we made some pretty big changes to how the audit backlog queues work, here we haven't changed the queue design so much as cleaned up the implementation. Brought about by the locking changes, we've simplified kauditd_thread() quite a bit by consolidating the queue handling into a new helper function, kauditd_send_queue(), which allows us to eliminate a lot of very similar code and makes the looping logic in kauditd_thread() clearer. * All netlink messages sent to auditd are now sent via auditd_send_unicast_skb(). Other than just making sense, this makes the lock handling easier. * Change the audit_log_start() sleep behavior so that we never sleep on auditd events (unchanged) or if the caller is holding the audit_cmd_mutex (changed). Previously we didn't sleep if the caller was auditd or if the message type fell between a certain range; the type check was a poor effort of doing what the cmd_mutex check now does. Richard Guy Briggs originally proposed not sleeping the cmd_mutex owner several years ago but his patch wasn't acceptable at the time. At least the idea lives on here. * A problem with the lost record counter has been resolved. Steve Grubb and I both happened to notice this problem and according to some quick testing by Steve, this problem goes back quite some time. It's largely a harmless problem, although it may have left some careful sysadmins quite puzzled. Cc: <stable@vger.kernel.org> # 4.10.x- Reported-by: Dmitry Vyukov <dvyukov@google.com> Signed-off-by: Paul Moore <paul@paul-moore.com>
2017-03-21 23:26:35 +08:00
sk = audit_get_sk(net);
portid = ac->portid;
audit: fix auditd/kernel connection state tracking What started as a rather straightforward race condition reported by Dmitry using the syzkaller fuzzer ended up revealing some major problems with how the audit subsystem managed its netlink sockets and its connection with the userspace audit daemon. Fixing this properly had quite the cascading effect and what we are left with is this rather large and complicated patch. My initial goal was to try and decompose this patch into multiple smaller patches, but the way these changes are intertwined makes it difficult to split these changes into meaningful pieces that don't break or somehow make things worse for the intermediate states. The patch makes a number of changes, but the most significant are highlighted below: * The auditd tracking variables, e.g. audit_sock, are now gone and replaced by a RCU/spin_lock protected variable auditd_conn which is a structure containing all of the auditd tracking information. * We no longer track the auditd sock directly, instead we track it via the network namespace in which it resides and we use the audit socket associated with that namespace. In spirit, this is what the code was trying to do prior to this patch (at least I think that is what the original authors intended), but it was done rather poorly and added a layer of obfuscation that only masked the underlying problems. * Big backlog queue cleanup, again. In v4.10 we made some pretty big changes to how the audit backlog queues work, here we haven't changed the queue design so much as cleaned up the implementation. Brought about by the locking changes, we've simplified kauditd_thread() quite a bit by consolidating the queue handling into a new helper function, kauditd_send_queue(), which allows us to eliminate a lot of very similar code and makes the looping logic in kauditd_thread() clearer. * All netlink messages sent to auditd are now sent via auditd_send_unicast_skb(). Other than just making sense, this makes the lock handling easier. * Change the audit_log_start() sleep behavior so that we never sleep on auditd events (unchanged) or if the caller is holding the audit_cmd_mutex (changed). Previously we didn't sleep if the caller was auditd or if the message type fell between a certain range; the type check was a poor effort of doing what the cmd_mutex check now does. Richard Guy Briggs originally proposed not sleeping the cmd_mutex owner several years ago but his patch wasn't acceptable at the time. At least the idea lives on here. * A problem with the lost record counter has been resolved. Steve Grubb and I both happened to notice this problem and according to some quick testing by Steve, this problem goes back quite some time. It's largely a harmless problem, although it may have left some careful sysadmins quite puzzled. Cc: <stable@vger.kernel.org> # 4.10.x- Reported-by: Dmitry Vyukov <dvyukov@google.com> Signed-off-by: Paul Moore <paul@paul-moore.com>
2017-03-21 23:26:35 +08:00
rcu_read_unlock();
/* attempt to flush the hold queue */
audit: fix auditd/kernel connection state tracking What started as a rather straightforward race condition reported by Dmitry using the syzkaller fuzzer ended up revealing some major problems with how the audit subsystem managed its netlink sockets and its connection with the userspace audit daemon. Fixing this properly had quite the cascading effect and what we are left with is this rather large and complicated patch. My initial goal was to try and decompose this patch into multiple smaller patches, but the way these changes are intertwined makes it difficult to split these changes into meaningful pieces that don't break or somehow make things worse for the intermediate states. The patch makes a number of changes, but the most significant are highlighted below: * The auditd tracking variables, e.g. audit_sock, are now gone and replaced by a RCU/spin_lock protected variable auditd_conn which is a structure containing all of the auditd tracking information. * We no longer track the auditd sock directly, instead we track it via the network namespace in which it resides and we use the audit socket associated with that namespace. In spirit, this is what the code was trying to do prior to this patch (at least I think that is what the original authors intended), but it was done rather poorly and added a layer of obfuscation that only masked the underlying problems. * Big backlog queue cleanup, again. In v4.10 we made some pretty big changes to how the audit backlog queues work, here we haven't changed the queue design so much as cleaned up the implementation. Brought about by the locking changes, we've simplified kauditd_thread() quite a bit by consolidating the queue handling into a new helper function, kauditd_send_queue(), which allows us to eliminate a lot of very similar code and makes the looping logic in kauditd_thread() clearer. * All netlink messages sent to auditd are now sent via auditd_send_unicast_skb(). Other than just making sense, this makes the lock handling easier. * Change the audit_log_start() sleep behavior so that we never sleep on auditd events (unchanged) or if the caller is holding the audit_cmd_mutex (changed). Previously we didn't sleep if the caller was auditd or if the message type fell between a certain range; the type check was a poor effort of doing what the cmd_mutex check now does. Richard Guy Briggs originally proposed not sleeping the cmd_mutex owner several years ago but his patch wasn't acceptable at the time. At least the idea lives on here. * A problem with the lost record counter has been resolved. Steve Grubb and I both happened to notice this problem and according to some quick testing by Steve, this problem goes back quite some time. It's largely a harmless problem, although it may have left some careful sysadmins quite puzzled. Cc: <stable@vger.kernel.org> # 4.10.x- Reported-by: Dmitry Vyukov <dvyukov@google.com> Signed-off-by: Paul Moore <paul@paul-moore.com>
2017-03-21 23:26:35 +08:00
rc = kauditd_send_queue(sk, portid,
&audit_hold_queue, UNICAST_RETRIES,
NULL, kauditd_rehold_skb);
if (rc < 0) {
audit: fix auditd/kernel connection state tracking What started as a rather straightforward race condition reported by Dmitry using the syzkaller fuzzer ended up revealing some major problems with how the audit subsystem managed its netlink sockets and its connection with the userspace audit daemon. Fixing this properly had quite the cascading effect and what we are left with is this rather large and complicated patch. My initial goal was to try and decompose this patch into multiple smaller patches, but the way these changes are intertwined makes it difficult to split these changes into meaningful pieces that don't break or somehow make things worse for the intermediate states. The patch makes a number of changes, but the most significant are highlighted below: * The auditd tracking variables, e.g. audit_sock, are now gone and replaced by a RCU/spin_lock protected variable auditd_conn which is a structure containing all of the auditd tracking information. * We no longer track the auditd sock directly, instead we track it via the network namespace in which it resides and we use the audit socket associated with that namespace. In spirit, this is what the code was trying to do prior to this patch (at least I think that is what the original authors intended), but it was done rather poorly and added a layer of obfuscation that only masked the underlying problems. * Big backlog queue cleanup, again. In v4.10 we made some pretty big changes to how the audit backlog queues work, here we haven't changed the queue design so much as cleaned up the implementation. Brought about by the locking changes, we've simplified kauditd_thread() quite a bit by consolidating the queue handling into a new helper function, kauditd_send_queue(), which allows us to eliminate a lot of very similar code and makes the looping logic in kauditd_thread() clearer. * All netlink messages sent to auditd are now sent via auditd_send_unicast_skb(). Other than just making sense, this makes the lock handling easier. * Change the audit_log_start() sleep behavior so that we never sleep on auditd events (unchanged) or if the caller is holding the audit_cmd_mutex (changed). Previously we didn't sleep if the caller was auditd or if the message type fell between a certain range; the type check was a poor effort of doing what the cmd_mutex check now does. Richard Guy Briggs originally proposed not sleeping the cmd_mutex owner several years ago but his patch wasn't acceptable at the time. At least the idea lives on here. * A problem with the lost record counter has been resolved. Steve Grubb and I both happened to notice this problem and according to some quick testing by Steve, this problem goes back quite some time. It's largely a harmless problem, although it may have left some careful sysadmins quite puzzled. Cc: <stable@vger.kernel.org> # 4.10.x- Reported-by: Dmitry Vyukov <dvyukov@google.com> Signed-off-by: Paul Moore <paul@paul-moore.com>
2017-03-21 23:26:35 +08:00
sk = NULL;
audit: fix a race condition with the auditd tracking code Originally reported by Adam and Dusty, it appears we have a small race window in kauditd_thread(), as documented in the Fedora BZ: * https://bugzilla.redhat.com/show_bug.cgi?id=1459326#c35 "This issue is partly due to the read-copy nature of RCU, and partly due to how we sync the auditd_connection state across kauditd_thread and the audit control channel. The kauditd_thread thread is always running so it can service the record queues and emit the multicast messages, if it happens to be just past the "main_queue" label, but before the "if (sk == NULL || ...)" if-statement which calls auditd_reset() when the new auditd connection is registered it could end up resetting the auditd connection, regardless of if it is valid or not. This is a rather small window and the variable nature of multi-core scheduling explains why this is proving rather difficult to reproduce." The fix is to have functions only call auditd_reset() when they believe that the kernel/auditd connection is still valid, e.g. non-NULL, and to have these callers pass their local copy of the auditd_connection pointer to auditd_reset() where it can be compared with the current connection state before resetting. If the caller has a stale state tracking pointer then the reset is ignored. We also make a small change to kauditd_thread() so that if the kernel/auditd connection is dead we skip the retry queue and send the records straight to the hold queue. This is necessary as we used to rely on auditd_reset() to occasionally purge the retry queue but we are going to be calling the reset function much less now and we want to make sure the retry queue doesn't grow unbounded. Reported-by: Adam Williamson <awilliam@redhat.com> Reported-by: Dusty Mabe <dustymabe@redhat.com> Reviewed-by: Richard Guy Briggs <rgb@redhat.com> Signed-off-by: Paul Moore <paul@paul-moore.com>
2017-06-12 21:35:24 +08:00
auditd_reset(ac);
audit: fix auditd/kernel connection state tracking What started as a rather straightforward race condition reported by Dmitry using the syzkaller fuzzer ended up revealing some major problems with how the audit subsystem managed its netlink sockets and its connection with the userspace audit daemon. Fixing this properly had quite the cascading effect and what we are left with is this rather large and complicated patch. My initial goal was to try and decompose this patch into multiple smaller patches, but the way these changes are intertwined makes it difficult to split these changes into meaningful pieces that don't break or somehow make things worse for the intermediate states. The patch makes a number of changes, but the most significant are highlighted below: * The auditd tracking variables, e.g. audit_sock, are now gone and replaced by a RCU/spin_lock protected variable auditd_conn which is a structure containing all of the auditd tracking information. * We no longer track the auditd sock directly, instead we track it via the network namespace in which it resides and we use the audit socket associated with that namespace. In spirit, this is what the code was trying to do prior to this patch (at least I think that is what the original authors intended), but it was done rather poorly and added a layer of obfuscation that only masked the underlying problems. * Big backlog queue cleanup, again. In v4.10 we made some pretty big changes to how the audit backlog queues work, here we haven't changed the queue design so much as cleaned up the implementation. Brought about by the locking changes, we've simplified kauditd_thread() quite a bit by consolidating the queue handling into a new helper function, kauditd_send_queue(), which allows us to eliminate a lot of very similar code and makes the looping logic in kauditd_thread() clearer. * All netlink messages sent to auditd are now sent via auditd_send_unicast_skb(). Other than just making sense, this makes the lock handling easier. * Change the audit_log_start() sleep behavior so that we never sleep on auditd events (unchanged) or if the caller is holding the audit_cmd_mutex (changed). Previously we didn't sleep if the caller was auditd or if the message type fell between a certain range; the type check was a poor effort of doing what the cmd_mutex check now does. Richard Guy Briggs originally proposed not sleeping the cmd_mutex owner several years ago but his patch wasn't acceptable at the time. At least the idea lives on here. * A problem with the lost record counter has been resolved. Steve Grubb and I both happened to notice this problem and according to some quick testing by Steve, this problem goes back quite some time. It's largely a harmless problem, although it may have left some careful sysadmins quite puzzled. Cc: <stable@vger.kernel.org> # 4.10.x- Reported-by: Dmitry Vyukov <dvyukov@google.com> Signed-off-by: Paul Moore <paul@paul-moore.com>
2017-03-21 23:26:35 +08:00
goto main_queue;
}
/* attempt to flush the retry queue */
audit: fix auditd/kernel connection state tracking What started as a rather straightforward race condition reported by Dmitry using the syzkaller fuzzer ended up revealing some major problems with how the audit subsystem managed its netlink sockets and its connection with the userspace audit daemon. Fixing this properly had quite the cascading effect and what we are left with is this rather large and complicated patch. My initial goal was to try and decompose this patch into multiple smaller patches, but the way these changes are intertwined makes it difficult to split these changes into meaningful pieces that don't break or somehow make things worse for the intermediate states. The patch makes a number of changes, but the most significant are highlighted below: * The auditd tracking variables, e.g. audit_sock, are now gone and replaced by a RCU/spin_lock protected variable auditd_conn which is a structure containing all of the auditd tracking information. * We no longer track the auditd sock directly, instead we track it via the network namespace in which it resides and we use the audit socket associated with that namespace. In spirit, this is what the code was trying to do prior to this patch (at least I think that is what the original authors intended), but it was done rather poorly and added a layer of obfuscation that only masked the underlying problems. * Big backlog queue cleanup, again. In v4.10 we made some pretty big changes to how the audit backlog queues work, here we haven't changed the queue design so much as cleaned up the implementation. Brought about by the locking changes, we've simplified kauditd_thread() quite a bit by consolidating the queue handling into a new helper function, kauditd_send_queue(), which allows us to eliminate a lot of very similar code and makes the looping logic in kauditd_thread() clearer. * All netlink messages sent to auditd are now sent via auditd_send_unicast_skb(). Other than just making sense, this makes the lock handling easier. * Change the audit_log_start() sleep behavior so that we never sleep on auditd events (unchanged) or if the caller is holding the audit_cmd_mutex (changed). Previously we didn't sleep if the caller was auditd or if the message type fell between a certain range; the type check was a poor effort of doing what the cmd_mutex check now does. Richard Guy Briggs originally proposed not sleeping the cmd_mutex owner several years ago but his patch wasn't acceptable at the time. At least the idea lives on here. * A problem with the lost record counter has been resolved. Steve Grubb and I both happened to notice this problem and according to some quick testing by Steve, this problem goes back quite some time. It's largely a harmless problem, although it may have left some careful sysadmins quite puzzled. Cc: <stable@vger.kernel.org> # 4.10.x- Reported-by: Dmitry Vyukov <dvyukov@google.com> Signed-off-by: Paul Moore <paul@paul-moore.com>
2017-03-21 23:26:35 +08:00
rc = kauditd_send_queue(sk, portid,
&audit_retry_queue, UNICAST_RETRIES,
NULL, kauditd_hold_skb);
if (rc < 0) {
audit: fix auditd/kernel connection state tracking What started as a rather straightforward race condition reported by Dmitry using the syzkaller fuzzer ended up revealing some major problems with how the audit subsystem managed its netlink sockets and its connection with the userspace audit daemon. Fixing this properly had quite the cascading effect and what we are left with is this rather large and complicated patch. My initial goal was to try and decompose this patch into multiple smaller patches, but the way these changes are intertwined makes it difficult to split these changes into meaningful pieces that don't break or somehow make things worse for the intermediate states. The patch makes a number of changes, but the most significant are highlighted below: * The auditd tracking variables, e.g. audit_sock, are now gone and replaced by a RCU/spin_lock protected variable auditd_conn which is a structure containing all of the auditd tracking information. * We no longer track the auditd sock directly, instead we track it via the network namespace in which it resides and we use the audit socket associated with that namespace. In spirit, this is what the code was trying to do prior to this patch (at least I think that is what the original authors intended), but it was done rather poorly and added a layer of obfuscation that only masked the underlying problems. * Big backlog queue cleanup, again. In v4.10 we made some pretty big changes to how the audit backlog queues work, here we haven't changed the queue design so much as cleaned up the implementation. Brought about by the locking changes, we've simplified kauditd_thread() quite a bit by consolidating the queue handling into a new helper function, kauditd_send_queue(), which allows us to eliminate a lot of very similar code and makes the looping logic in kauditd_thread() clearer. * All netlink messages sent to auditd are now sent via auditd_send_unicast_skb(). Other than just making sense, this makes the lock handling easier. * Change the audit_log_start() sleep behavior so that we never sleep on auditd events (unchanged) or if the caller is holding the audit_cmd_mutex (changed). Previously we didn't sleep if the caller was auditd or if the message type fell between a certain range; the type check was a poor effort of doing what the cmd_mutex check now does. Richard Guy Briggs originally proposed not sleeping the cmd_mutex owner several years ago but his patch wasn't acceptable at the time. At least the idea lives on here. * A problem with the lost record counter has been resolved. Steve Grubb and I both happened to notice this problem and according to some quick testing by Steve, this problem goes back quite some time. It's largely a harmless problem, although it may have left some careful sysadmins quite puzzled. Cc: <stable@vger.kernel.org> # 4.10.x- Reported-by: Dmitry Vyukov <dvyukov@google.com> Signed-off-by: Paul Moore <paul@paul-moore.com>
2017-03-21 23:26:35 +08:00
sk = NULL;
audit: fix a race condition with the auditd tracking code Originally reported by Adam and Dusty, it appears we have a small race window in kauditd_thread(), as documented in the Fedora BZ: * https://bugzilla.redhat.com/show_bug.cgi?id=1459326#c35 "This issue is partly due to the read-copy nature of RCU, and partly due to how we sync the auditd_connection state across kauditd_thread and the audit control channel. The kauditd_thread thread is always running so it can service the record queues and emit the multicast messages, if it happens to be just past the "main_queue" label, but before the "if (sk == NULL || ...)" if-statement which calls auditd_reset() when the new auditd connection is registered it could end up resetting the auditd connection, regardless of if it is valid or not. This is a rather small window and the variable nature of multi-core scheduling explains why this is proving rather difficult to reproduce." The fix is to have functions only call auditd_reset() when they believe that the kernel/auditd connection is still valid, e.g. non-NULL, and to have these callers pass their local copy of the auditd_connection pointer to auditd_reset() where it can be compared with the current connection state before resetting. If the caller has a stale state tracking pointer then the reset is ignored. We also make a small change to kauditd_thread() so that if the kernel/auditd connection is dead we skip the retry queue and send the records straight to the hold queue. This is necessary as we used to rely on auditd_reset() to occasionally purge the retry queue but we are going to be calling the reset function much less now and we want to make sure the retry queue doesn't grow unbounded. Reported-by: Adam Williamson <awilliam@redhat.com> Reported-by: Dusty Mabe <dustymabe@redhat.com> Reviewed-by: Richard Guy Briggs <rgb@redhat.com> Signed-off-by: Paul Moore <paul@paul-moore.com>
2017-06-12 21:35:24 +08:00
auditd_reset(ac);
audit: fix auditd/kernel connection state tracking What started as a rather straightforward race condition reported by Dmitry using the syzkaller fuzzer ended up revealing some major problems with how the audit subsystem managed its netlink sockets and its connection with the userspace audit daemon. Fixing this properly had quite the cascading effect and what we are left with is this rather large and complicated patch. My initial goal was to try and decompose this patch into multiple smaller patches, but the way these changes are intertwined makes it difficult to split these changes into meaningful pieces that don't break or somehow make things worse for the intermediate states. The patch makes a number of changes, but the most significant are highlighted below: * The auditd tracking variables, e.g. audit_sock, are now gone and replaced by a RCU/spin_lock protected variable auditd_conn which is a structure containing all of the auditd tracking information. * We no longer track the auditd sock directly, instead we track it via the network namespace in which it resides and we use the audit socket associated with that namespace. In spirit, this is what the code was trying to do prior to this patch (at least I think that is what the original authors intended), but it was done rather poorly and added a layer of obfuscation that only masked the underlying problems. * Big backlog queue cleanup, again. In v4.10 we made some pretty big changes to how the audit backlog queues work, here we haven't changed the queue design so much as cleaned up the implementation. Brought about by the locking changes, we've simplified kauditd_thread() quite a bit by consolidating the queue handling into a new helper function, kauditd_send_queue(), which allows us to eliminate a lot of very similar code and makes the looping logic in kauditd_thread() clearer. * All netlink messages sent to auditd are now sent via auditd_send_unicast_skb(). Other than just making sense, this makes the lock handling easier. * Change the audit_log_start() sleep behavior so that we never sleep on auditd events (unchanged) or if the caller is holding the audit_cmd_mutex (changed). Previously we didn't sleep if the caller was auditd or if the message type fell between a certain range; the type check was a poor effort of doing what the cmd_mutex check now does. Richard Guy Briggs originally proposed not sleeping the cmd_mutex owner several years ago but his patch wasn't acceptable at the time. At least the idea lives on here. * A problem with the lost record counter has been resolved. Steve Grubb and I both happened to notice this problem and according to some quick testing by Steve, this problem goes back quite some time. It's largely a harmless problem, although it may have left some careful sysadmins quite puzzled. Cc: <stable@vger.kernel.org> # 4.10.x- Reported-by: Dmitry Vyukov <dvyukov@google.com> Signed-off-by: Paul Moore <paul@paul-moore.com>
2017-03-21 23:26:35 +08:00
goto main_queue;
}
audit: efficiency fix 1: only wake up if queue shorter than backlog limit These and similar errors were seen on a patched 3.8 kernel when the audit subsystem was overrun during boot: udevd[876]: worker [887] unexpectedly returned with status 0x0100 udevd[876]: worker [887] failed while handling '/devices/pci0000:00/0000:00:03.0/0000:40:00.0' udevd[876]: worker [880] unexpectedly returned with status 0x0100 udevd[876]: worker [880] failed while handling '/devices/LNXSYSTM:00/LNXPWRBN:00/input/input1/event1' udevadm settle - timeout of 180 seconds reached, the event queue contains: /sys/devices/LNXSYSTM:00/LNXPWRBN:00/input/input1/event1 (3995) /sys/devices/LNXSYSTM:00/LNXSYBUS:00/PNP0A08:00/INT3F0D:00 (4034) audit: audit_backlog=258 > audit_backlog_limit=256 audit: audit_lost=1 audit_rate_limit=0 audit_backlog_limit=256 The change below increases the efficiency of the audit code and prevents it from being overrun: Only issue a wake_up in kauditd if the length of the skb queue is less than the backlog limit. Otherwise, threads waiting in wait_for_auditd() will simply wake up, discover that the queue is still too long for them to proceed, and go back to sleep. This results in wasted context switches and machine cycles. kauditd_thread() is the only function that removes buffers from audit_skb_queue so we can't race. If we did, the timeout in wait_for_auditd() would expire and the waiting thread would continue. See: https://lkml.org/lkml/2013/9/2/479 Signed-off-by: Dan Duval <dan.duval@oracle.com> Signed-off-by: Chuck Anderson <chuck.anderson@oracle.com> Signed-off-by: Richard Guy Briggs <rgb@redhat.com> Signed-off-by: Eric Paris <eparis@redhat.com>
2013-09-16 23:11:12 +08:00
audit: fix auditd/kernel connection state tracking What started as a rather straightforward race condition reported by Dmitry using the syzkaller fuzzer ended up revealing some major problems with how the audit subsystem managed its netlink sockets and its connection with the userspace audit daemon. Fixing this properly had quite the cascading effect and what we are left with is this rather large and complicated patch. My initial goal was to try and decompose this patch into multiple smaller patches, but the way these changes are intertwined makes it difficult to split these changes into meaningful pieces that don't break or somehow make things worse for the intermediate states. The patch makes a number of changes, but the most significant are highlighted below: * The auditd tracking variables, e.g. audit_sock, are now gone and replaced by a RCU/spin_lock protected variable auditd_conn which is a structure containing all of the auditd tracking information. * We no longer track the auditd sock directly, instead we track it via the network namespace in which it resides and we use the audit socket associated with that namespace. In spirit, this is what the code was trying to do prior to this patch (at least I think that is what the original authors intended), but it was done rather poorly and added a layer of obfuscation that only masked the underlying problems. * Big backlog queue cleanup, again. In v4.10 we made some pretty big changes to how the audit backlog queues work, here we haven't changed the queue design so much as cleaned up the implementation. Brought about by the locking changes, we've simplified kauditd_thread() quite a bit by consolidating the queue handling into a new helper function, kauditd_send_queue(), which allows us to eliminate a lot of very similar code and makes the looping logic in kauditd_thread() clearer. * All netlink messages sent to auditd are now sent via auditd_send_unicast_skb(). Other than just making sense, this makes the lock handling easier. * Change the audit_log_start() sleep behavior so that we never sleep on auditd events (unchanged) or if the caller is holding the audit_cmd_mutex (changed). Previously we didn't sleep if the caller was auditd or if the message type fell between a certain range; the type check was a poor effort of doing what the cmd_mutex check now does. Richard Guy Briggs originally proposed not sleeping the cmd_mutex owner several years ago but his patch wasn't acceptable at the time. At least the idea lives on here. * A problem with the lost record counter has been resolved. Steve Grubb and I both happened to notice this problem and according to some quick testing by Steve, this problem goes back quite some time. It's largely a harmless problem, although it may have left some careful sysadmins quite puzzled. Cc: <stable@vger.kernel.org> # 4.10.x- Reported-by: Dmitry Vyukov <dvyukov@google.com> Signed-off-by: Paul Moore <paul@paul-moore.com>
2017-03-21 23:26:35 +08:00
main_queue:
/* process the main queue - do the multicast send and attempt
* unicast, dump failed record sends to the retry queue; if
* sk == NULL due to previous failures we will just do the
audit: fix a race condition with the auditd tracking code Originally reported by Adam and Dusty, it appears we have a small race window in kauditd_thread(), as documented in the Fedora BZ: * https://bugzilla.redhat.com/show_bug.cgi?id=1459326#c35 "This issue is partly due to the read-copy nature of RCU, and partly due to how we sync the auditd_connection state across kauditd_thread and the audit control channel. The kauditd_thread thread is always running so it can service the record queues and emit the multicast messages, if it happens to be just past the "main_queue" label, but before the "if (sk == NULL || ...)" if-statement which calls auditd_reset() when the new auditd connection is registered it could end up resetting the auditd connection, regardless of if it is valid or not. This is a rather small window and the variable nature of multi-core scheduling explains why this is proving rather difficult to reproduce." The fix is to have functions only call auditd_reset() when they believe that the kernel/auditd connection is still valid, e.g. non-NULL, and to have these callers pass their local copy of the auditd_connection pointer to auditd_reset() where it can be compared with the current connection state before resetting. If the caller has a stale state tracking pointer then the reset is ignored. We also make a small change to kauditd_thread() so that if the kernel/auditd connection is dead we skip the retry queue and send the records straight to the hold queue. This is necessary as we used to rely on auditd_reset() to occasionally purge the retry queue but we are going to be calling the reset function much less now and we want to make sure the retry queue doesn't grow unbounded. Reported-by: Adam Williamson <awilliam@redhat.com> Reported-by: Dusty Mabe <dustymabe@redhat.com> Reviewed-by: Richard Guy Briggs <rgb@redhat.com> Signed-off-by: Paul Moore <paul@paul-moore.com>
2017-06-12 21:35:24 +08:00
* multicast send and move the record to the hold queue */
rc = kauditd_send_queue(sk, portid, &audit_queue, 1,
kauditd_send_multicast_skb,
audit: fix a race condition with the auditd tracking code Originally reported by Adam and Dusty, it appears we have a small race window in kauditd_thread(), as documented in the Fedora BZ: * https://bugzilla.redhat.com/show_bug.cgi?id=1459326#c35 "This issue is partly due to the read-copy nature of RCU, and partly due to how we sync the auditd_connection state across kauditd_thread and the audit control channel. The kauditd_thread thread is always running so it can service the record queues and emit the multicast messages, if it happens to be just past the "main_queue" label, but before the "if (sk == NULL || ...)" if-statement which calls auditd_reset() when the new auditd connection is registered it could end up resetting the auditd connection, regardless of if it is valid or not. This is a rather small window and the variable nature of multi-core scheduling explains why this is proving rather difficult to reproduce." The fix is to have functions only call auditd_reset() when they believe that the kernel/auditd connection is still valid, e.g. non-NULL, and to have these callers pass their local copy of the auditd_connection pointer to auditd_reset() where it can be compared with the current connection state before resetting. If the caller has a stale state tracking pointer then the reset is ignored. We also make a small change to kauditd_thread() so that if the kernel/auditd connection is dead we skip the retry queue and send the records straight to the hold queue. This is necessary as we used to rely on auditd_reset() to occasionally purge the retry queue but we are going to be calling the reset function much less now and we want to make sure the retry queue doesn't grow unbounded. Reported-by: Adam Williamson <awilliam@redhat.com> Reported-by: Dusty Mabe <dustymabe@redhat.com> Reviewed-by: Richard Guy Briggs <rgb@redhat.com> Signed-off-by: Paul Moore <paul@paul-moore.com>
2017-06-12 21:35:24 +08:00
(sk ?
kauditd_retry_skb : kauditd_hold_skb));
if (ac && rc < 0)
auditd_reset(ac);
sk = NULL;
audit: fix auditd/kernel connection state tracking What started as a rather straightforward race condition reported by Dmitry using the syzkaller fuzzer ended up revealing some major problems with how the audit subsystem managed its netlink sockets and its connection with the userspace audit daemon. Fixing this properly had quite the cascading effect and what we are left with is this rather large and complicated patch. My initial goal was to try and decompose this patch into multiple smaller patches, but the way these changes are intertwined makes it difficult to split these changes into meaningful pieces that don't break or somehow make things worse for the intermediate states. The patch makes a number of changes, but the most significant are highlighted below: * The auditd tracking variables, e.g. audit_sock, are now gone and replaced by a RCU/spin_lock protected variable auditd_conn which is a structure containing all of the auditd tracking information. * We no longer track the auditd sock directly, instead we track it via the network namespace in which it resides and we use the audit socket associated with that namespace. In spirit, this is what the code was trying to do prior to this patch (at least I think that is what the original authors intended), but it was done rather poorly and added a layer of obfuscation that only masked the underlying problems. * Big backlog queue cleanup, again. In v4.10 we made some pretty big changes to how the audit backlog queues work, here we haven't changed the queue design so much as cleaned up the implementation. Brought about by the locking changes, we've simplified kauditd_thread() quite a bit by consolidating the queue handling into a new helper function, kauditd_send_queue(), which allows us to eliminate a lot of very similar code and makes the looping logic in kauditd_thread() clearer. * All netlink messages sent to auditd are now sent via auditd_send_unicast_skb(). Other than just making sense, this makes the lock handling easier. * Change the audit_log_start() sleep behavior so that we never sleep on auditd events (unchanged) or if the caller is holding the audit_cmd_mutex (changed). Previously we didn't sleep if the caller was auditd or if the message type fell between a certain range; the type check was a poor effort of doing what the cmd_mutex check now does. Richard Guy Briggs originally proposed not sleeping the cmd_mutex owner several years ago but his patch wasn't acceptable at the time. At least the idea lives on here. * A problem with the lost record counter has been resolved. Steve Grubb and I both happened to notice this problem and according to some quick testing by Steve, this problem goes back quite some time. It's largely a harmless problem, although it may have left some careful sysadmins quite puzzled. Cc: <stable@vger.kernel.org> # 4.10.x- Reported-by: Dmitry Vyukov <dvyukov@google.com> Signed-off-by: Paul Moore <paul@paul-moore.com>
2017-03-21 23:26:35 +08:00
/* drop our netns reference, no auditd sends past this line */
if (net) {
put_net(net);
net = NULL;
}
audit: fix auditd/kernel connection state tracking What started as a rather straightforward race condition reported by Dmitry using the syzkaller fuzzer ended up revealing some major problems with how the audit subsystem managed its netlink sockets and its connection with the userspace audit daemon. Fixing this properly had quite the cascading effect and what we are left with is this rather large and complicated patch. My initial goal was to try and decompose this patch into multiple smaller patches, but the way these changes are intertwined makes it difficult to split these changes into meaningful pieces that don't break or somehow make things worse for the intermediate states. The patch makes a number of changes, but the most significant are highlighted below: * The auditd tracking variables, e.g. audit_sock, are now gone and replaced by a RCU/spin_lock protected variable auditd_conn which is a structure containing all of the auditd tracking information. * We no longer track the auditd sock directly, instead we track it via the network namespace in which it resides and we use the audit socket associated with that namespace. In spirit, this is what the code was trying to do prior to this patch (at least I think that is what the original authors intended), but it was done rather poorly and added a layer of obfuscation that only masked the underlying problems. * Big backlog queue cleanup, again. In v4.10 we made some pretty big changes to how the audit backlog queues work, here we haven't changed the queue design so much as cleaned up the implementation. Brought about by the locking changes, we've simplified kauditd_thread() quite a bit by consolidating the queue handling into a new helper function, kauditd_send_queue(), which allows us to eliminate a lot of very similar code and makes the looping logic in kauditd_thread() clearer. * All netlink messages sent to auditd are now sent via auditd_send_unicast_skb(). Other than just making sense, this makes the lock handling easier. * Change the audit_log_start() sleep behavior so that we never sleep on auditd events (unchanged) or if the caller is holding the audit_cmd_mutex (changed). Previously we didn't sleep if the caller was auditd or if the message type fell between a certain range; the type check was a poor effort of doing what the cmd_mutex check now does. Richard Guy Briggs originally proposed not sleeping the cmd_mutex owner several years ago but his patch wasn't acceptable at the time. At least the idea lives on here. * A problem with the lost record counter has been resolved. Steve Grubb and I both happened to notice this problem and according to some quick testing by Steve, this problem goes back quite some time. It's largely a harmless problem, although it may have left some careful sysadmins quite puzzled. Cc: <stable@vger.kernel.org> # 4.10.x- Reported-by: Dmitry Vyukov <dvyukov@google.com> Signed-off-by: Paul Moore <paul@paul-moore.com>
2017-03-21 23:26:35 +08:00
/* we have processed all the queues so wake everyone */
wake_up(&audit_backlog_wait);
/* NOTE: we want to wake up if there is anything on the queue,
* regardless of if an auditd is connected, as we need to
* do the multicast send and rotate records from the
* main queue to the retry/hold queues */
wait_event_freezable(kauditd_wait,
(skb_queue_len(&audit_queue) ? 1 : 0));
}
return 0;
}
int audit_send_list_thread(void *_dest)
{
struct audit_netlink_list *dest = _dest;
struct sk_buff *skb;
audit: fix auditd/kernel connection state tracking What started as a rather straightforward race condition reported by Dmitry using the syzkaller fuzzer ended up revealing some major problems with how the audit subsystem managed its netlink sockets and its connection with the userspace audit daemon. Fixing this properly had quite the cascading effect and what we are left with is this rather large and complicated patch. My initial goal was to try and decompose this patch into multiple smaller patches, but the way these changes are intertwined makes it difficult to split these changes into meaningful pieces that don't break or somehow make things worse for the intermediate states. The patch makes a number of changes, but the most significant are highlighted below: * The auditd tracking variables, e.g. audit_sock, are now gone and replaced by a RCU/spin_lock protected variable auditd_conn which is a structure containing all of the auditd tracking information. * We no longer track the auditd sock directly, instead we track it via the network namespace in which it resides and we use the audit socket associated with that namespace. In spirit, this is what the code was trying to do prior to this patch (at least I think that is what the original authors intended), but it was done rather poorly and added a layer of obfuscation that only masked the underlying problems. * Big backlog queue cleanup, again. In v4.10 we made some pretty big changes to how the audit backlog queues work, here we haven't changed the queue design so much as cleaned up the implementation. Brought about by the locking changes, we've simplified kauditd_thread() quite a bit by consolidating the queue handling into a new helper function, kauditd_send_queue(), which allows us to eliminate a lot of very similar code and makes the looping logic in kauditd_thread() clearer. * All netlink messages sent to auditd are now sent via auditd_send_unicast_skb(). Other than just making sense, this makes the lock handling easier. * Change the audit_log_start() sleep behavior so that we never sleep on auditd events (unchanged) or if the caller is holding the audit_cmd_mutex (changed). Previously we didn't sleep if the caller was auditd or if the message type fell between a certain range; the type check was a poor effort of doing what the cmd_mutex check now does. Richard Guy Briggs originally proposed not sleeping the cmd_mutex owner several years ago but his patch wasn't acceptable at the time. At least the idea lives on here. * A problem with the lost record counter has been resolved. Steve Grubb and I both happened to notice this problem and according to some quick testing by Steve, this problem goes back quite some time. It's largely a harmless problem, although it may have left some careful sysadmins quite puzzled. Cc: <stable@vger.kernel.org> # 4.10.x- Reported-by: Dmitry Vyukov <dvyukov@google.com> Signed-off-by: Paul Moore <paul@paul-moore.com>
2017-03-21 23:26:35 +08:00
struct sock *sk = audit_get_sk(dest->net);
/* wait for parent to finish and send an ACK */
audit_ctl_lock();
audit_ctl_unlock();
while ((skb = __skb_dequeue(&dest->q)) != NULL)
audit: fix auditd/kernel connection state tracking What started as a rather straightforward race condition reported by Dmitry using the syzkaller fuzzer ended up revealing some major problems with how the audit subsystem managed its netlink sockets and its connection with the userspace audit daemon. Fixing this properly had quite the cascading effect and what we are left with is this rather large and complicated patch. My initial goal was to try and decompose this patch into multiple smaller patches, but the way these changes are intertwined makes it difficult to split these changes into meaningful pieces that don't break or somehow make things worse for the intermediate states. The patch makes a number of changes, but the most significant are highlighted below: * The auditd tracking variables, e.g. audit_sock, are now gone and replaced by a RCU/spin_lock protected variable auditd_conn which is a structure containing all of the auditd tracking information. * We no longer track the auditd sock directly, instead we track it via the network namespace in which it resides and we use the audit socket associated with that namespace. In spirit, this is what the code was trying to do prior to this patch (at least I think that is what the original authors intended), but it was done rather poorly and added a layer of obfuscation that only masked the underlying problems. * Big backlog queue cleanup, again. In v4.10 we made some pretty big changes to how the audit backlog queues work, here we haven't changed the queue design so much as cleaned up the implementation. Brought about by the locking changes, we've simplified kauditd_thread() quite a bit by consolidating the queue handling into a new helper function, kauditd_send_queue(), which allows us to eliminate a lot of very similar code and makes the looping logic in kauditd_thread() clearer. * All netlink messages sent to auditd are now sent via auditd_send_unicast_skb(). Other than just making sense, this makes the lock handling easier. * Change the audit_log_start() sleep behavior so that we never sleep on auditd events (unchanged) or if the caller is holding the audit_cmd_mutex (changed). Previously we didn't sleep if the caller was auditd or if the message type fell between a certain range; the type check was a poor effort of doing what the cmd_mutex check now does. Richard Guy Briggs originally proposed not sleeping the cmd_mutex owner several years ago but his patch wasn't acceptable at the time. At least the idea lives on here. * A problem with the lost record counter has been resolved. Steve Grubb and I both happened to notice this problem and according to some quick testing by Steve, this problem goes back quite some time. It's largely a harmless problem, although it may have left some careful sysadmins quite puzzled. Cc: <stable@vger.kernel.org> # 4.10.x- Reported-by: Dmitry Vyukov <dvyukov@google.com> Signed-off-by: Paul Moore <paul@paul-moore.com>
2017-03-21 23:26:35 +08:00
netlink_unicast(sk, skb, dest->portid, 0);
audit: fix auditd/kernel connection state tracking What started as a rather straightforward race condition reported by Dmitry using the syzkaller fuzzer ended up revealing some major problems with how the audit subsystem managed its netlink sockets and its connection with the userspace audit daemon. Fixing this properly had quite the cascading effect and what we are left with is this rather large and complicated patch. My initial goal was to try and decompose this patch into multiple smaller patches, but the way these changes are intertwined makes it difficult to split these changes into meaningful pieces that don't break or somehow make things worse for the intermediate states. The patch makes a number of changes, but the most significant are highlighted below: * The auditd tracking variables, e.g. audit_sock, are now gone and replaced by a RCU/spin_lock protected variable auditd_conn which is a structure containing all of the auditd tracking information. * We no longer track the auditd sock directly, instead we track it via the network namespace in which it resides and we use the audit socket associated with that namespace. In spirit, this is what the code was trying to do prior to this patch (at least I think that is what the original authors intended), but it was done rather poorly and added a layer of obfuscation that only masked the underlying problems. * Big backlog queue cleanup, again. In v4.10 we made some pretty big changes to how the audit backlog queues work, here we haven't changed the queue design so much as cleaned up the implementation. Brought about by the locking changes, we've simplified kauditd_thread() quite a bit by consolidating the queue handling into a new helper function, kauditd_send_queue(), which allows us to eliminate a lot of very similar code and makes the looping logic in kauditd_thread() clearer. * All netlink messages sent to auditd are now sent via auditd_send_unicast_skb(). Other than just making sense, this makes the lock handling easier. * Change the audit_log_start() sleep behavior so that we never sleep on auditd events (unchanged) or if the caller is holding the audit_cmd_mutex (changed). Previously we didn't sleep if the caller was auditd or if the message type fell between a certain range; the type check was a poor effort of doing what the cmd_mutex check now does. Richard Guy Briggs originally proposed not sleeping the cmd_mutex owner several years ago but his patch wasn't acceptable at the time. At least the idea lives on here. * A problem with the lost record counter has been resolved. Steve Grubb and I both happened to notice this problem and according to some quick testing by Steve, this problem goes back quite some time. It's largely a harmless problem, although it may have left some careful sysadmins quite puzzled. Cc: <stable@vger.kernel.org> # 4.10.x- Reported-by: Dmitry Vyukov <dvyukov@google.com> Signed-off-by: Paul Moore <paul@paul-moore.com>
2017-03-21 23:26:35 +08:00
put_net(dest->net);
kfree(dest);
return 0;
}
struct sk_buff *audit_make_reply(int seq, int type, int done,
int multi, const void *payload, int size)
{
struct sk_buff *skb;
struct nlmsghdr *nlh;
void *data;
int flags = multi ? NLM_F_MULTI : 0;
int t = done ? NLMSG_DONE : type;
skb = nlmsg_new(size, GFP_KERNEL);
if (!skb)
return NULL;
nlh = nlmsg_put(skb, 0, seq, t, size, flags);
if (!nlh)
goto out_kfree_skb;
data = nlmsg_data(nlh);
memcpy(data, payload, size);
return skb;
out_kfree_skb:
kfree_skb(skb);
return NULL;
}
static void audit_free_reply(struct audit_reply *reply)
{
if (!reply)
return;
kfree_skb(reply->skb);
if (reply->net)
put_net(reply->net);
kfree(reply);
}
static int audit_send_reply_thread(void *arg)
{
struct audit_reply *reply = (struct audit_reply *)arg;
audit_ctl_lock();
audit_ctl_unlock();
/* Ignore failure. It'll only happen if the sender goes away,
because our timeout is set to infinite. */
netlink_unicast(audit_get_sk(reply->net), reply->skb, reply->portid, 0);
reply->skb = NULL;
audit_free_reply(reply);
return 0;
}
/**
* audit_send_reply - send an audit reply message via netlink
* @request_skb: skb of request we are replying to (used to target the reply)
* @seq: sequence number
* @type: audit message type
* @done: done (last) flag
* @multi: multi-part message flag
* @payload: payload data
* @size: payload size
*
* Allocates a skb, builds the netlink message, and sends it to the port id.
*/
static void audit_send_reply(struct sk_buff *request_skb, int seq, int type, int done,
int multi, const void *payload, int size)
{
struct task_struct *tsk;
struct audit_reply *reply;
reply = kzalloc(sizeof(*reply), GFP_KERNEL);
if (!reply)
return;
reply->skb = audit_make_reply(seq, type, done, multi, payload, size);
if (!reply->skb)
goto err;
reply->net = get_net(sock_net(NETLINK_CB(request_skb).sk));
reply->portid = NETLINK_CB(request_skb).portid;
tsk = kthread_run(audit_send_reply_thread, reply, "audit_send_reply");
if (IS_ERR(tsk))
goto err;
return;
err:
audit_free_reply(reply);
}
/*
* Check for appropriate CAP_AUDIT_ capabilities on incoming audit
* control messages.
*/
static int audit_netlink_ok(struct sk_buff *skb, u16 msg_type)
{
int err = 0;
/* Only support initial user namespace for now. */
AUDIT: Allow login in non-init namespaces It its possible to configure your PAM stack to refuse login if audit messages (about the login) were unable to be sent. This is common in many distros and thus normal configuration of many containers. The PAM modules determine if audit is enabled/disabled in the kernel based on the return value from sending an audit message on the netlink socket. If userspace gets back ECONNREFUSED it believes audit is disabled in the kernel. If it gets any other error else it refuses to let the login proceed. Just about ever since the introduction of namespaces the kernel audit subsystem has returned EPERM if the task sending a message was not in the init user or pid namespace. So many forms of containers have never worked if audit was enabled in the kernel. BUT if the container was not in net_init then the kernel network code would send ECONNREFUSED (instead of the audit code sending EPERM). Thus by pure accident/dumb luck/bug if an admin configured the PAM stack to reject all logins that didn't talk to audit, but then ran the login untility in the non-init_net namespace, it would work!! Clearly this was a bug, but it is a bug some people expected. With the introduction of network namespace support in 3.14-rc1 the two bugs stopped cancelling each other out. Now, containers in the non-init_net namespace refused to let users log in (just like PAM was configfured!) Obviously some people were not happy that what used to let users log in, now didn't! This fix is kinda hacky. We return ECONNREFUSED for all non-init relevant namespaces. That means that not only will the old broken non-init_net setups continue to work, now the broken non-init_pid or non-init_user setups will 'work'. They don't really work, since audit isn't logging things. But it's what most users want. In 3.15 we should have patches to support not only the non-init_net (3.14) namespace but also the non-init_pid and non-init_user namespace. So all will be right in the world. This just opens the doors wide open on 3.14 and hopefully makes users happy, if not the audit system... Reported-by: Andre Tomt <andre@tomt.net> Reported-by: Adam Richter <adam_richter2004@yahoo.com> Signed-off-by: Eric Paris <eparis@redhat.com> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2014-03-31 07:07:54 +08:00
/*
* We return ECONNREFUSED because it tricks userspace into thinking
* that audit was not configured into the kernel. Lots of users
* configure their PAM stack (because that's what the distro does)
* to reject login if unable to send messages to audit. If we return
* ECONNREFUSED the PAM stack thinks the kernel does not have audit
* configured in and will let login proceed. If we return EPERM
* userspace will reject all logins. This should be removed when we
* support non init namespaces!!
*/
if (current_user_ns() != &init_user_ns)
AUDIT: Allow login in non-init namespaces It its possible to configure your PAM stack to refuse login if audit messages (about the login) were unable to be sent. This is common in many distros and thus normal configuration of many containers. The PAM modules determine if audit is enabled/disabled in the kernel based on the return value from sending an audit message on the netlink socket. If userspace gets back ECONNREFUSED it believes audit is disabled in the kernel. If it gets any other error else it refuses to let the login proceed. Just about ever since the introduction of namespaces the kernel audit subsystem has returned EPERM if the task sending a message was not in the init user or pid namespace. So many forms of containers have never worked if audit was enabled in the kernel. BUT if the container was not in net_init then the kernel network code would send ECONNREFUSED (instead of the audit code sending EPERM). Thus by pure accident/dumb luck/bug if an admin configured the PAM stack to reject all logins that didn't talk to audit, but then ran the login untility in the non-init_net namespace, it would work!! Clearly this was a bug, but it is a bug some people expected. With the introduction of network namespace support in 3.14-rc1 the two bugs stopped cancelling each other out. Now, containers in the non-init_net namespace refused to let users log in (just like PAM was configfured!) Obviously some people were not happy that what used to let users log in, now didn't! This fix is kinda hacky. We return ECONNREFUSED for all non-init relevant namespaces. That means that not only will the old broken non-init_net setups continue to work, now the broken non-init_pid or non-init_user setups will 'work'. They don't really work, since audit isn't logging things. But it's what most users want. In 3.15 we should have patches to support not only the non-init_net (3.14) namespace but also the non-init_pid and non-init_user namespace. So all will be right in the world. This just opens the doors wide open on 3.14 and hopefully makes users happy, if not the audit system... Reported-by: Andre Tomt <andre@tomt.net> Reported-by: Adam Richter <adam_richter2004@yahoo.com> Signed-off-by: Eric Paris <eparis@redhat.com> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2014-03-31 07:07:54 +08:00
return -ECONNREFUSED;
switch (msg_type) {
case AUDIT_LIST:
case AUDIT_ADD:
case AUDIT_DEL:
return -EOPNOTSUPP;
case AUDIT_GET:
case AUDIT_SET:
case AUDIT_GET_FEATURE:
case AUDIT_SET_FEATURE:
case AUDIT_LIST_RULES:
case AUDIT_ADD_RULE:
case AUDIT_DEL_RULE:
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 14:40:56 +08:00
case AUDIT_TTY_GET:
case AUDIT_TTY_SET:
case AUDIT_TRIM:
case AUDIT_MAKE_EQUIV:
/* Only support auditd and auditctl in initial pid namespace
* for now. */
if (task_active_pid_ns(current) != &init_pid_ns)
return -EPERM;
if (!netlink_capable(skb, CAP_AUDIT_CONTROL))
err = -EPERM;
break;
case AUDIT_USER:
case AUDIT_FIRST_USER_MSG ... AUDIT_LAST_USER_MSG:
case AUDIT_FIRST_USER_MSG2 ... AUDIT_LAST_USER_MSG2:
if (!netlink_capable(skb, CAP_AUDIT_WRITE))
err = -EPERM;
break;
default: /* bad msg */
err = -EINVAL;
}
return err;
}
static void audit_log_common_recv_msg(struct audit_context *context,
struct audit_buffer **ab, u16 msg_type)
{
uid_t uid = from_kuid(&init_user_ns, current_uid());
pid_t pid = task_tgid_nr(current);
if (!audit_enabled && msg_type != AUDIT_USER_AVC) {
*ab = NULL;
return;
}
*ab = audit_log_start(context, GFP_KERNEL, msg_type);
if (unlikely(!*ab))
return;
audit_log_format(*ab, "pid=%d uid=%u ", pid, uid);
2013-04-30 21:53:34 +08:00
audit_log_session_info(*ab);
audit_log_task_context(*ab);
}
static inline void audit_log_user_recv_msg(struct audit_buffer **ab,
u16 msg_type)
{
audit_log_common_recv_msg(NULL, ab, msg_type);
}
static int is_audit_feature_set(int i)
{
return af.features & AUDIT_FEATURE_TO_MASK(i);
}
static int audit_get_feature(struct sk_buff *skb)
{
u32 seq;
seq = nlmsg_hdr(skb)->nlmsg_seq;
audit_send_reply(skb, seq, AUDIT_GET_FEATURE, 0, 0, &af, sizeof(af));
return 0;
}
static void audit_log_feature_change(int which, u32 old_feature, u32 new_feature,
u32 old_lock, u32 new_lock, int res)
{
struct audit_buffer *ab;
if (audit_enabled == AUDIT_OFF)
return;
ab = audit_log_start(audit_context(), GFP_KERNEL, AUDIT_FEATURE_CHANGE);
if (!ab)
return;
audit_log_task_info(ab);
audit_log_format(ab, " feature=%s old=%u new=%u old_lock=%u new_lock=%u res=%d",
audit_feature_names[which], !!old_feature, !!new_feature,
!!old_lock, !!new_lock, res);
audit_log_end(ab);
}
static int audit_set_feature(struct audit_features *uaf)
{
int i;
BUILD_BUG_ON(AUDIT_LAST_FEATURE + 1 > ARRAY_SIZE(audit_feature_names));
/* if there is ever a version 2 we should handle that here */
for (i = 0; i <= AUDIT_LAST_FEATURE; i++) {
u32 feature = AUDIT_FEATURE_TO_MASK(i);
u32 old_feature, new_feature, old_lock, new_lock;
/* if we are not changing this feature, move along */
if (!(feature & uaf->mask))
continue;
old_feature = af.features & feature;
new_feature = uaf->features & feature;
new_lock = (uaf->lock | af.lock) & feature;
old_lock = af.lock & feature;
/* are we changing a locked feature? */
if (old_lock && (new_feature != old_feature)) {
audit_log_feature_change(i, old_feature, new_feature,
old_lock, new_lock, 0);
return -EPERM;
}
}
/* nothing invalid, do the changes */
for (i = 0; i <= AUDIT_LAST_FEATURE; i++) {
u32 feature = AUDIT_FEATURE_TO_MASK(i);
u32 old_feature, new_feature, old_lock, new_lock;
/* if we are not changing this feature, move along */
if (!(feature & uaf->mask))
continue;
old_feature = af.features & feature;
new_feature = uaf->features & feature;
old_lock = af.lock & feature;
new_lock = (uaf->lock | af.lock) & feature;
if (new_feature != old_feature)
audit_log_feature_change(i, old_feature, new_feature,
old_lock, new_lock, 1);
if (new_feature)
af.features |= feature;
else
af.features &= ~feature;
af.lock |= new_lock;
}
return 0;
}
static int audit_replace(struct pid *pid)
{
pid_t pvnr;
audit: fix auditd/kernel connection state tracking What started as a rather straightforward race condition reported by Dmitry using the syzkaller fuzzer ended up revealing some major problems with how the audit subsystem managed its netlink sockets and its connection with the userspace audit daemon. Fixing this properly had quite the cascading effect and what we are left with is this rather large and complicated patch. My initial goal was to try and decompose this patch into multiple smaller patches, but the way these changes are intertwined makes it difficult to split these changes into meaningful pieces that don't break or somehow make things worse for the intermediate states. The patch makes a number of changes, but the most significant are highlighted below: * The auditd tracking variables, e.g. audit_sock, are now gone and replaced by a RCU/spin_lock protected variable auditd_conn which is a structure containing all of the auditd tracking information. * We no longer track the auditd sock directly, instead we track it via the network namespace in which it resides and we use the audit socket associated with that namespace. In spirit, this is what the code was trying to do prior to this patch (at least I think that is what the original authors intended), but it was done rather poorly and added a layer of obfuscation that only masked the underlying problems. * Big backlog queue cleanup, again. In v4.10 we made some pretty big changes to how the audit backlog queues work, here we haven't changed the queue design so much as cleaned up the implementation. Brought about by the locking changes, we've simplified kauditd_thread() quite a bit by consolidating the queue handling into a new helper function, kauditd_send_queue(), which allows us to eliminate a lot of very similar code and makes the looping logic in kauditd_thread() clearer. * All netlink messages sent to auditd are now sent via auditd_send_unicast_skb(). Other than just making sense, this makes the lock handling easier. * Change the audit_log_start() sleep behavior so that we never sleep on auditd events (unchanged) or if the caller is holding the audit_cmd_mutex (changed). Previously we didn't sleep if the caller was auditd or if the message type fell between a certain range; the type check was a poor effort of doing what the cmd_mutex check now does. Richard Guy Briggs originally proposed not sleeping the cmd_mutex owner several years ago but his patch wasn't acceptable at the time. At least the idea lives on here. * A problem with the lost record counter has been resolved. Steve Grubb and I both happened to notice this problem and according to some quick testing by Steve, this problem goes back quite some time. It's largely a harmless problem, although it may have left some careful sysadmins quite puzzled. Cc: <stable@vger.kernel.org> # 4.10.x- Reported-by: Dmitry Vyukov <dvyukov@google.com> Signed-off-by: Paul Moore <paul@paul-moore.com>
2017-03-21 23:26:35 +08:00
struct sk_buff *skb;
pvnr = pid_vnr(pid);
skb = audit_make_reply(0, AUDIT_REPLACE, 0, 0, &pvnr, sizeof(pvnr));
if (!skb)
return -ENOMEM;
audit: fix auditd/kernel connection state tracking What started as a rather straightforward race condition reported by Dmitry using the syzkaller fuzzer ended up revealing some major problems with how the audit subsystem managed its netlink sockets and its connection with the userspace audit daemon. Fixing this properly had quite the cascading effect and what we are left with is this rather large and complicated patch. My initial goal was to try and decompose this patch into multiple smaller patches, but the way these changes are intertwined makes it difficult to split these changes into meaningful pieces that don't break or somehow make things worse for the intermediate states. The patch makes a number of changes, but the most significant are highlighted below: * The auditd tracking variables, e.g. audit_sock, are now gone and replaced by a RCU/spin_lock protected variable auditd_conn which is a structure containing all of the auditd tracking information. * We no longer track the auditd sock directly, instead we track it via the network namespace in which it resides and we use the audit socket associated with that namespace. In spirit, this is what the code was trying to do prior to this patch (at least I think that is what the original authors intended), but it was done rather poorly and added a layer of obfuscation that only masked the underlying problems. * Big backlog queue cleanup, again. In v4.10 we made some pretty big changes to how the audit backlog queues work, here we haven't changed the queue design so much as cleaned up the implementation. Brought about by the locking changes, we've simplified kauditd_thread() quite a bit by consolidating the queue handling into a new helper function, kauditd_send_queue(), which allows us to eliminate a lot of very similar code and makes the looping logic in kauditd_thread() clearer. * All netlink messages sent to auditd are now sent via auditd_send_unicast_skb(). Other than just making sense, this makes the lock handling easier. * Change the audit_log_start() sleep behavior so that we never sleep on auditd events (unchanged) or if the caller is holding the audit_cmd_mutex (changed). Previously we didn't sleep if the caller was auditd or if the message type fell between a certain range; the type check was a poor effort of doing what the cmd_mutex check now does. Richard Guy Briggs originally proposed not sleeping the cmd_mutex owner several years ago but his patch wasn't acceptable at the time. At least the idea lives on here. * A problem with the lost record counter has been resolved. Steve Grubb and I both happened to notice this problem and according to some quick testing by Steve, this problem goes back quite some time. It's largely a harmless problem, although it may have left some careful sysadmins quite puzzled. Cc: <stable@vger.kernel.org> # 4.10.x- Reported-by: Dmitry Vyukov <dvyukov@google.com> Signed-off-by: Paul Moore <paul@paul-moore.com>
2017-03-21 23:26:35 +08:00
return auditd_send_unicast_skb(skb);
}
static int audit_receive_msg(struct sk_buff *skb, struct nlmsghdr *nlh)
{
u32 seq;
void *data;
int data_len;
int err;
struct audit_buffer *ab;
u16 msg_type = nlh->nlmsg_type;
struct audit_sig_info *sig_data;
char *ctx = NULL;
u32 len;
err = audit_netlink_ok(skb, msg_type);
if (err)
return err;
seq = nlh->nlmsg_seq;
data = nlmsg_data(nlh);
data_len = nlmsg_len(nlh);
switch (msg_type) {
case AUDIT_GET: {
struct audit_status s;
memset(&s, 0, sizeof(s));
s.enabled = audit_enabled;
s.failure = audit_failure;
/* NOTE: use pid_vnr() so the PID is relative to the current
* namespace */
s.pid = auditd_pid_vnr();
s.rate_limit = audit_rate_limit;
s.backlog_limit = audit_backlog_limit;
s.lost = atomic_read(&audit_lost);
s.backlog = skb_queue_len(&audit_queue);
s.feature_bitmap = AUDIT_FEATURE_BITMAP_ALL;
s.backlog_wait_time = audit_backlog_wait_time;
s.backlog_wait_time_actual = atomic_read(&audit_backlog_wait_time_actual);
audit_send_reply(skb, seq, AUDIT_GET, 0, 0, &s, sizeof(s));
break;
}
case AUDIT_SET: {
struct audit_status s;
memset(&s, 0, sizeof(s));
/* guard against past and future API changes */
memcpy(&s, data, min_t(size_t, sizeof(s), data_len));
if (s.mask & AUDIT_STATUS_ENABLED) {
err = audit_set_enabled(s.enabled);
if (err < 0)
return err;
}
if (s.mask & AUDIT_STATUS_FAILURE) {
err = audit_set_failure(s.failure);
if (err < 0)
return err;
}
if (s.mask & AUDIT_STATUS_PID) {
/* NOTE: we are using the vnr PID functions below
* because the s.pid value is relative to the
* namespace of the caller; at present this
* doesn't matter much since you can really only
* run auditd from the initial pid namespace, but
* something to keep in mind if this changes */
pid_t new_pid = s.pid;
audit: fix auditd/kernel connection state tracking What started as a rather straightforward race condition reported by Dmitry using the syzkaller fuzzer ended up revealing some major problems with how the audit subsystem managed its netlink sockets and its connection with the userspace audit daemon. Fixing this properly had quite the cascading effect and what we are left with is this rather large and complicated patch. My initial goal was to try and decompose this patch into multiple smaller patches, but the way these changes are intertwined makes it difficult to split these changes into meaningful pieces that don't break or somehow make things worse for the intermediate states. The patch makes a number of changes, but the most significant are highlighted below: * The auditd tracking variables, e.g. audit_sock, are now gone and replaced by a RCU/spin_lock protected variable auditd_conn which is a structure containing all of the auditd tracking information. * We no longer track the auditd sock directly, instead we track it via the network namespace in which it resides and we use the audit socket associated with that namespace. In spirit, this is what the code was trying to do prior to this patch (at least I think that is what the original authors intended), but it was done rather poorly and added a layer of obfuscation that only masked the underlying problems. * Big backlog queue cleanup, again. In v4.10 we made some pretty big changes to how the audit backlog queues work, here we haven't changed the queue design so much as cleaned up the implementation. Brought about by the locking changes, we've simplified kauditd_thread() quite a bit by consolidating the queue handling into a new helper function, kauditd_send_queue(), which allows us to eliminate a lot of very similar code and makes the looping logic in kauditd_thread() clearer. * All netlink messages sent to auditd are now sent via auditd_send_unicast_skb(). Other than just making sense, this makes the lock handling easier. * Change the audit_log_start() sleep behavior so that we never sleep on auditd events (unchanged) or if the caller is holding the audit_cmd_mutex (changed). Previously we didn't sleep if the caller was auditd or if the message type fell between a certain range; the type check was a poor effort of doing what the cmd_mutex check now does. Richard Guy Briggs originally proposed not sleeping the cmd_mutex owner several years ago but his patch wasn't acceptable at the time. At least the idea lives on here. * A problem with the lost record counter has been resolved. Steve Grubb and I both happened to notice this problem and according to some quick testing by Steve, this problem goes back quite some time. It's largely a harmless problem, although it may have left some careful sysadmins quite puzzled. Cc: <stable@vger.kernel.org> # 4.10.x- Reported-by: Dmitry Vyukov <dvyukov@google.com> Signed-off-by: Paul Moore <paul@paul-moore.com>
2017-03-21 23:26:35 +08:00
pid_t auditd_pid;
struct pid *req_pid = task_tgid(current);
/* Sanity check - PID values must match. Setting
* pid to 0 is how auditd ends auditing. */
if (new_pid && (new_pid != pid_vnr(req_pid)))
return -EINVAL;
audit: fix auditd/kernel connection state tracking What started as a rather straightforward race condition reported by Dmitry using the syzkaller fuzzer ended up revealing some major problems with how the audit subsystem managed its netlink sockets and its connection with the userspace audit daemon. Fixing this properly had quite the cascading effect and what we are left with is this rather large and complicated patch. My initial goal was to try and decompose this patch into multiple smaller patches, but the way these changes are intertwined makes it difficult to split these changes into meaningful pieces that don't break or somehow make things worse for the intermediate states. The patch makes a number of changes, but the most significant are highlighted below: * The auditd tracking variables, e.g. audit_sock, are now gone and replaced by a RCU/spin_lock protected variable auditd_conn which is a structure containing all of the auditd tracking information. * We no longer track the auditd sock directly, instead we track it via the network namespace in which it resides and we use the audit socket associated with that namespace. In spirit, this is what the code was trying to do prior to this patch (at least I think that is what the original authors intended), but it was done rather poorly and added a layer of obfuscation that only masked the underlying problems. * Big backlog queue cleanup, again. In v4.10 we made some pretty big changes to how the audit backlog queues work, here we haven't changed the queue design so much as cleaned up the implementation. Brought about by the locking changes, we've simplified kauditd_thread() quite a bit by consolidating the queue handling into a new helper function, kauditd_send_queue(), which allows us to eliminate a lot of very similar code and makes the looping logic in kauditd_thread() clearer. * All netlink messages sent to auditd are now sent via auditd_send_unicast_skb(). Other than just making sense, this makes the lock handling easier. * Change the audit_log_start() sleep behavior so that we never sleep on auditd events (unchanged) or if the caller is holding the audit_cmd_mutex (changed). Previously we didn't sleep if the caller was auditd or if the message type fell between a certain range; the type check was a poor effort of doing what the cmd_mutex check now does. Richard Guy Briggs originally proposed not sleeping the cmd_mutex owner several years ago but his patch wasn't acceptable at the time. At least the idea lives on here. * A problem with the lost record counter has been resolved. Steve Grubb and I both happened to notice this problem and according to some quick testing by Steve, this problem goes back quite some time. It's largely a harmless problem, although it may have left some careful sysadmins quite puzzled. Cc: <stable@vger.kernel.org> # 4.10.x- Reported-by: Dmitry Vyukov <dvyukov@google.com> Signed-off-by: Paul Moore <paul@paul-moore.com>
2017-03-21 23:26:35 +08:00
/* test the auditd connection */
audit_replace(req_pid);
audit: fix auditd/kernel connection state tracking What started as a rather straightforward race condition reported by Dmitry using the syzkaller fuzzer ended up revealing some major problems with how the audit subsystem managed its netlink sockets and its connection with the userspace audit daemon. Fixing this properly had quite the cascading effect and what we are left with is this rather large and complicated patch. My initial goal was to try and decompose this patch into multiple smaller patches, but the way these changes are intertwined makes it difficult to split these changes into meaningful pieces that don't break or somehow make things worse for the intermediate states. The patch makes a number of changes, but the most significant are highlighted below: * The auditd tracking variables, e.g. audit_sock, are now gone and replaced by a RCU/spin_lock protected variable auditd_conn which is a structure containing all of the auditd tracking information. * We no longer track the auditd sock directly, instead we track it via the network namespace in which it resides and we use the audit socket associated with that namespace. In spirit, this is what the code was trying to do prior to this patch (at least I think that is what the original authors intended), but it was done rather poorly and added a layer of obfuscation that only masked the underlying problems. * Big backlog queue cleanup, again. In v4.10 we made some pretty big changes to how the audit backlog queues work, here we haven't changed the queue design so much as cleaned up the implementation. Brought about by the locking changes, we've simplified kauditd_thread() quite a bit by consolidating the queue handling into a new helper function, kauditd_send_queue(), which allows us to eliminate a lot of very similar code and makes the looping logic in kauditd_thread() clearer. * All netlink messages sent to auditd are now sent via auditd_send_unicast_skb(). Other than just making sense, this makes the lock handling easier. * Change the audit_log_start() sleep behavior so that we never sleep on auditd events (unchanged) or if the caller is holding the audit_cmd_mutex (changed). Previously we didn't sleep if the caller was auditd or if the message type fell between a certain range; the type check was a poor effort of doing what the cmd_mutex check now does. Richard Guy Briggs originally proposed not sleeping the cmd_mutex owner several years ago but his patch wasn't acceptable at the time. At least the idea lives on here. * A problem with the lost record counter has been resolved. Steve Grubb and I both happened to notice this problem and according to some quick testing by Steve, this problem goes back quite some time. It's largely a harmless problem, although it may have left some careful sysadmins quite puzzled. Cc: <stable@vger.kernel.org> # 4.10.x- Reported-by: Dmitry Vyukov <dvyukov@google.com> Signed-off-by: Paul Moore <paul@paul-moore.com>
2017-03-21 23:26:35 +08:00
auditd_pid = auditd_pid_vnr();
if (auditd_pid) {
/* replacing a healthy auditd is not allowed */
if (new_pid) {
audit_log_config_change("audit_pid",
new_pid, auditd_pid, 0);
return -EEXIST;
}
/* only current auditd can unregister itself */
if (pid_vnr(req_pid) != auditd_pid) {
audit_log_config_change("audit_pid",
new_pid, auditd_pid, 0);
return -EACCES;
}
}
audit: fix auditd/kernel connection state tracking What started as a rather straightforward race condition reported by Dmitry using the syzkaller fuzzer ended up revealing some major problems with how the audit subsystem managed its netlink sockets and its connection with the userspace audit daemon. Fixing this properly had quite the cascading effect and what we are left with is this rather large and complicated patch. My initial goal was to try and decompose this patch into multiple smaller patches, but the way these changes are intertwined makes it difficult to split these changes into meaningful pieces that don't break or somehow make things worse for the intermediate states. The patch makes a number of changes, but the most significant are highlighted below: * The auditd tracking variables, e.g. audit_sock, are now gone and replaced by a RCU/spin_lock protected variable auditd_conn which is a structure containing all of the auditd tracking information. * We no longer track the auditd sock directly, instead we track it via the network namespace in which it resides and we use the audit socket associated with that namespace. In spirit, this is what the code was trying to do prior to this patch (at least I think that is what the original authors intended), but it was done rather poorly and added a layer of obfuscation that only masked the underlying problems. * Big backlog queue cleanup, again. In v4.10 we made some pretty big changes to how the audit backlog queues work, here we haven't changed the queue design so much as cleaned up the implementation. Brought about by the locking changes, we've simplified kauditd_thread() quite a bit by consolidating the queue handling into a new helper function, kauditd_send_queue(), which allows us to eliminate a lot of very similar code and makes the looping logic in kauditd_thread() clearer. * All netlink messages sent to auditd are now sent via auditd_send_unicast_skb(). Other than just making sense, this makes the lock handling easier. * Change the audit_log_start() sleep behavior so that we never sleep on auditd events (unchanged) or if the caller is holding the audit_cmd_mutex (changed). Previously we didn't sleep if the caller was auditd or if the message type fell between a certain range; the type check was a poor effort of doing what the cmd_mutex check now does. Richard Guy Briggs originally proposed not sleeping the cmd_mutex owner several years ago but his patch wasn't acceptable at the time. At least the idea lives on here. * A problem with the lost record counter has been resolved. Steve Grubb and I both happened to notice this problem and according to some quick testing by Steve, this problem goes back quite some time. It's largely a harmless problem, although it may have left some careful sysadmins quite puzzled. Cc: <stable@vger.kernel.org> # 4.10.x- Reported-by: Dmitry Vyukov <dvyukov@google.com> Signed-off-by: Paul Moore <paul@paul-moore.com>
2017-03-21 23:26:35 +08:00
if (new_pid) {
audit: fix auditd/kernel connection state tracking What started as a rather straightforward race condition reported by Dmitry using the syzkaller fuzzer ended up revealing some major problems with how the audit subsystem managed its netlink sockets and its connection with the userspace audit daemon. Fixing this properly had quite the cascading effect and what we are left with is this rather large and complicated patch. My initial goal was to try and decompose this patch into multiple smaller patches, but the way these changes are intertwined makes it difficult to split these changes into meaningful pieces that don't break or somehow make things worse for the intermediate states. The patch makes a number of changes, but the most significant are highlighted below: * The auditd tracking variables, e.g. audit_sock, are now gone and replaced by a RCU/spin_lock protected variable auditd_conn which is a structure containing all of the auditd tracking information. * We no longer track the auditd sock directly, instead we track it via the network namespace in which it resides and we use the audit socket associated with that namespace. In spirit, this is what the code was trying to do prior to this patch (at least I think that is what the original authors intended), but it was done rather poorly and added a layer of obfuscation that only masked the underlying problems. * Big backlog queue cleanup, again. In v4.10 we made some pretty big changes to how the audit backlog queues work, here we haven't changed the queue design so much as cleaned up the implementation. Brought about by the locking changes, we've simplified kauditd_thread() quite a bit by consolidating the queue handling into a new helper function, kauditd_send_queue(), which allows us to eliminate a lot of very similar code and makes the looping logic in kauditd_thread() clearer. * All netlink messages sent to auditd are now sent via auditd_send_unicast_skb(). Other than just making sense, this makes the lock handling easier. * Change the audit_log_start() sleep behavior so that we never sleep on auditd events (unchanged) or if the caller is holding the audit_cmd_mutex (changed). Previously we didn't sleep if the caller was auditd or if the message type fell between a certain range; the type check was a poor effort of doing what the cmd_mutex check now does. Richard Guy Briggs originally proposed not sleeping the cmd_mutex owner several years ago but his patch wasn't acceptable at the time. At least the idea lives on here. * A problem with the lost record counter has been resolved. Steve Grubb and I both happened to notice this problem and according to some quick testing by Steve, this problem goes back quite some time. It's largely a harmless problem, although it may have left some careful sysadmins quite puzzled. Cc: <stable@vger.kernel.org> # 4.10.x- Reported-by: Dmitry Vyukov <dvyukov@google.com> Signed-off-by: Paul Moore <paul@paul-moore.com>
2017-03-21 23:26:35 +08:00
/* register a new auditd connection */
err = auditd_set(req_pid,
NETLINK_CB(skb).portid,
sock_net(NETLINK_CB(skb).sk));
if (audit_enabled != AUDIT_OFF)
audit_log_config_change("audit_pid",
new_pid,
auditd_pid,
err ? 0 : 1);
if (err)
return err;
audit: fix auditd/kernel connection state tracking What started as a rather straightforward race condition reported by Dmitry using the syzkaller fuzzer ended up revealing some major problems with how the audit subsystem managed its netlink sockets and its connection with the userspace audit daemon. Fixing this properly had quite the cascading effect and what we are left with is this rather large and complicated patch. My initial goal was to try and decompose this patch into multiple smaller patches, but the way these changes are intertwined makes it difficult to split these changes into meaningful pieces that don't break or somehow make things worse for the intermediate states. The patch makes a number of changes, but the most significant are highlighted below: * The auditd tracking variables, e.g. audit_sock, are now gone and replaced by a RCU/spin_lock protected variable auditd_conn which is a structure containing all of the auditd tracking information. * We no longer track the auditd sock directly, instead we track it via the network namespace in which it resides and we use the audit socket associated with that namespace. In spirit, this is what the code was trying to do prior to this patch (at least I think that is what the original authors intended), but it was done rather poorly and added a layer of obfuscation that only masked the underlying problems. * Big backlog queue cleanup, again. In v4.10 we made some pretty big changes to how the audit backlog queues work, here we haven't changed the queue design so much as cleaned up the implementation. Brought about by the locking changes, we've simplified kauditd_thread() quite a bit by consolidating the queue handling into a new helper function, kauditd_send_queue(), which allows us to eliminate a lot of very similar code and makes the looping logic in kauditd_thread() clearer. * All netlink messages sent to auditd are now sent via auditd_send_unicast_skb(). Other than just making sense, this makes the lock handling easier. * Change the audit_log_start() sleep behavior so that we never sleep on auditd events (unchanged) or if the caller is holding the audit_cmd_mutex (changed). Previously we didn't sleep if the caller was auditd or if the message type fell between a certain range; the type check was a poor effort of doing what the cmd_mutex check now does. Richard Guy Briggs originally proposed not sleeping the cmd_mutex owner several years ago but his patch wasn't acceptable at the time. At least the idea lives on here. * A problem with the lost record counter has been resolved. Steve Grubb and I both happened to notice this problem and according to some quick testing by Steve, this problem goes back quite some time. It's largely a harmless problem, although it may have left some careful sysadmins quite puzzled. Cc: <stable@vger.kernel.org> # 4.10.x- Reported-by: Dmitry Vyukov <dvyukov@google.com> Signed-off-by: Paul Moore <paul@paul-moore.com>
2017-03-21 23:26:35 +08:00
/* try to process any backlog */
wake_up_interruptible(&kauditd_wait);
} else {
if (audit_enabled != AUDIT_OFF)
audit_log_config_change("audit_pid",
new_pid,
auditd_pid, 1);
audit: fix auditd/kernel connection state tracking What started as a rather straightforward race condition reported by Dmitry using the syzkaller fuzzer ended up revealing some major problems with how the audit subsystem managed its netlink sockets and its connection with the userspace audit daemon. Fixing this properly had quite the cascading effect and what we are left with is this rather large and complicated patch. My initial goal was to try and decompose this patch into multiple smaller patches, but the way these changes are intertwined makes it difficult to split these changes into meaningful pieces that don't break or somehow make things worse for the intermediate states. The patch makes a number of changes, but the most significant are highlighted below: * The auditd tracking variables, e.g. audit_sock, are now gone and replaced by a RCU/spin_lock protected variable auditd_conn which is a structure containing all of the auditd tracking information. * We no longer track the auditd sock directly, instead we track it via the network namespace in which it resides and we use the audit socket associated with that namespace. In spirit, this is what the code was trying to do prior to this patch (at least I think that is what the original authors intended), but it was done rather poorly and added a layer of obfuscation that only masked the underlying problems. * Big backlog queue cleanup, again. In v4.10 we made some pretty big changes to how the audit backlog queues work, here we haven't changed the queue design so much as cleaned up the implementation. Brought about by the locking changes, we've simplified kauditd_thread() quite a bit by consolidating the queue handling into a new helper function, kauditd_send_queue(), which allows us to eliminate a lot of very similar code and makes the looping logic in kauditd_thread() clearer. * All netlink messages sent to auditd are now sent via auditd_send_unicast_skb(). Other than just making sense, this makes the lock handling easier. * Change the audit_log_start() sleep behavior so that we never sleep on auditd events (unchanged) or if the caller is holding the audit_cmd_mutex (changed). Previously we didn't sleep if the caller was auditd or if the message type fell between a certain range; the type check was a poor effort of doing what the cmd_mutex check now does. Richard Guy Briggs originally proposed not sleeping the cmd_mutex owner several years ago but his patch wasn't acceptable at the time. At least the idea lives on here. * A problem with the lost record counter has been resolved. Steve Grubb and I both happened to notice this problem and according to some quick testing by Steve, this problem goes back quite some time. It's largely a harmless problem, although it may have left some careful sysadmins quite puzzled. Cc: <stable@vger.kernel.org> # 4.10.x- Reported-by: Dmitry Vyukov <dvyukov@google.com> Signed-off-by: Paul Moore <paul@paul-moore.com>
2017-03-21 23:26:35 +08:00
/* unregister the auditd connection */
audit: fix a race condition with the auditd tracking code Originally reported by Adam and Dusty, it appears we have a small race window in kauditd_thread(), as documented in the Fedora BZ: * https://bugzilla.redhat.com/show_bug.cgi?id=1459326#c35 "This issue is partly due to the read-copy nature of RCU, and partly due to how we sync the auditd_connection state across kauditd_thread and the audit control channel. The kauditd_thread thread is always running so it can service the record queues and emit the multicast messages, if it happens to be just past the "main_queue" label, but before the "if (sk == NULL || ...)" if-statement which calls auditd_reset() when the new auditd connection is registered it could end up resetting the auditd connection, regardless of if it is valid or not. This is a rather small window and the variable nature of multi-core scheduling explains why this is proving rather difficult to reproduce." The fix is to have functions only call auditd_reset() when they believe that the kernel/auditd connection is still valid, e.g. non-NULL, and to have these callers pass their local copy of the auditd_connection pointer to auditd_reset() where it can be compared with the current connection state before resetting. If the caller has a stale state tracking pointer then the reset is ignored. We also make a small change to kauditd_thread() so that if the kernel/auditd connection is dead we skip the retry queue and send the records straight to the hold queue. This is necessary as we used to rely on auditd_reset() to occasionally purge the retry queue but we are going to be calling the reset function much less now and we want to make sure the retry queue doesn't grow unbounded. Reported-by: Adam Williamson <awilliam@redhat.com> Reported-by: Dusty Mabe <dustymabe@redhat.com> Reviewed-by: Richard Guy Briggs <rgb@redhat.com> Signed-off-by: Paul Moore <paul@paul-moore.com>
2017-06-12 21:35:24 +08:00
auditd_reset(NULL);
}
}
if (s.mask & AUDIT_STATUS_RATE_LIMIT) {
err = audit_set_rate_limit(s.rate_limit);
if (err < 0)
return err;
}
if (s.mask & AUDIT_STATUS_BACKLOG_LIMIT) {
err = audit_set_backlog_limit(s.backlog_limit);
if (err < 0)
return err;
}
if (s.mask & AUDIT_STATUS_BACKLOG_WAIT_TIME) {
if (sizeof(s) > (size_t)nlh->nlmsg_len)
return -EINVAL;
if (s.backlog_wait_time > 10*AUDIT_BACKLOG_WAIT_TIME)
return -EINVAL;
err = audit_set_backlog_wait_time(s.backlog_wait_time);
if (err < 0)
return err;
}
if (s.mask == AUDIT_STATUS_LOST) {
u32 lost = atomic_xchg(&audit_lost, 0);
audit_log_config_change("lost", 0, lost, 1);
return lost;
}
if (s.mask == AUDIT_STATUS_BACKLOG_WAIT_TIME_ACTUAL) {
u32 actual = atomic_xchg(&audit_backlog_wait_time_actual, 0);
audit_log_config_change("backlog_wait_time_actual", 0, actual, 1);
return actual;
}
break;
}
case AUDIT_GET_FEATURE:
err = audit_get_feature(skb);
if (err)
return err;
break;
case AUDIT_SET_FEATURE:
if (data_len < sizeof(struct audit_features))
return -EINVAL;
err = audit_set_feature(data);
if (err)
return err;
break;
case AUDIT_USER:
case AUDIT_FIRST_USER_MSG ... AUDIT_LAST_USER_MSG:
case AUDIT_FIRST_USER_MSG2 ... AUDIT_LAST_USER_MSG2:
if (!audit_enabled && msg_type != AUDIT_USER_AVC)
return 0;
/* exit early if there isn't at least one character to print */
if (data_len < 2)
return -EINVAL;
err = audit_filter(msg_type, AUDIT_FILTER_USER);
if (err == 1) { /* match or error */
char *str = data;
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 14:40:56 +08:00
if (msg_type == AUDIT_USER_TTY) {
err = tty_audit_push();
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 14:40:56 +08:00
if (err)
break;
}
audit_log_user_recv_msg(&ab, msg_type);
if (msg_type != AUDIT_USER_TTY) {
/* ensure NULL termination */
str[data_len - 1] = '\0';
audit_log_format(ab, " msg='%.*s'",
AUDIT_MESSAGE_TEXT_MAX,
str);
} else {
audit_log_format(ab, " data=");
if (str[data_len - 1] == '\0')
data_len--;
audit_log_n_untrustedstring(ab, str, data_len);
}
audit_log_end(ab);
}
break;
case AUDIT_ADD_RULE:
case AUDIT_DEL_RULE:
if (data_len < sizeof(struct audit_rule_data))
return -EINVAL;
if (audit_enabled == AUDIT_LOCKED) {
audit_log_common_recv_msg(audit_context(), &ab,
AUDIT_CONFIG_CHANGE);
audit_log_format(ab, " op=%s audit_enabled=%d res=0",
msg_type == AUDIT_ADD_RULE ?
"add_rule" : "remove_rule",
audit_enabled);
audit_log_end(ab);
return -EPERM;
}
err = audit_rule_change(msg_type, seq, data, data_len);
break;
case AUDIT_LIST_RULES:
err = audit_list_rules_send(skb, seq);
break;
case AUDIT_TRIM:
audit_trim_trees();
audit_log_common_recv_msg(audit_context(), &ab,
AUDIT_CONFIG_CHANGE);
audit_log_format(ab, " op=trim res=1");
audit_log_end(ab);
break;
case AUDIT_MAKE_EQUIV: {
void *bufp = data;
u32 sizes[2];
size_t msglen = data_len;
char *old, *new;
err = -EINVAL;
if (msglen < 2 * sizeof(u32))
break;
memcpy(sizes, bufp, 2 * sizeof(u32));
bufp += 2 * sizeof(u32);
msglen -= 2 * sizeof(u32);
old = audit_unpack_string(&bufp, &msglen, sizes[0]);
if (IS_ERR(old)) {
err = PTR_ERR(old);
break;
}
new = audit_unpack_string(&bufp, &msglen, sizes[1]);
if (IS_ERR(new)) {
err = PTR_ERR(new);
kfree(old);
break;
}
/* OK, here comes... */
err = audit_tag_tree(old, new);
audit_log_common_recv_msg(audit_context(), &ab,
AUDIT_CONFIG_CHANGE);
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;
}
case AUDIT_SIGNAL_INFO:
len = 0;
if (audit_sig_sid) {
err = security_secid_to_secctx(audit_sig_sid, &ctx, &len);
if (err)
return err;
}
sig_data = kmalloc(struct_size(sig_data, ctx, len), GFP_KERNEL);
if (!sig_data) {
if (audit_sig_sid)
security_release_secctx(ctx, len);
return -ENOMEM;
}
sig_data->uid = from_kuid(&init_user_ns, audit_sig_uid);
sig_data->pid = audit_sig_pid;
if (audit_sig_sid) {
memcpy(sig_data->ctx, ctx, len);
security_release_secctx(ctx, len);
}
audit_send_reply(skb, seq, AUDIT_SIGNAL_INFO, 0, 0,
sig_data, struct_size(sig_data, ctx, len));
kfree(sig_data);
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 14:40:56 +08:00
case AUDIT_TTY_GET: {
struct audit_tty_status s;
unsigned int t;
t = READ_ONCE(current->signal->audit_tty);
s.enabled = t & AUDIT_TTY_ENABLE;
s.log_passwd = !!(t & AUDIT_TTY_LOG_PASSWD);
audit_send_reply(skb, 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 14:40:56 +08:00
break;
}
case AUDIT_TTY_SET: {
struct audit_tty_status s, old;
struct audit_buffer *ab;
unsigned int t;
memset(&s, 0, sizeof(s));
/* guard against past and future API changes */
memcpy(&s, data, min_t(size_t, sizeof(s), data_len));
/* check if new data is valid */
if ((s.enabled != 0 && s.enabled != 1) ||
(s.log_passwd != 0 && s.log_passwd != 1))
err = -EINVAL;
if (err)
t = READ_ONCE(current->signal->audit_tty);
else {
t = s.enabled | (-s.log_passwd & AUDIT_TTY_LOG_PASSWD);
t = xchg(&current->signal->audit_tty, t);
}
old.enabled = t & AUDIT_TTY_ENABLE;
old.log_passwd = !!(t & AUDIT_TTY_LOG_PASSWD);
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 14:40:56 +08:00
audit_log_common_recv_msg(audit_context(), &ab,
AUDIT_CONFIG_CHANGE);
audit_log_format(ab, " op=tty_set old-enabled=%d new-enabled=%d"
" old-log_passwd=%d new-log_passwd=%d res=%d",
old.enabled, s.enabled, old.log_passwd,
s.log_passwd, !err);
audit_log_end(ab);
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 14:40:56 +08:00
break;
}
default:
err = -EINVAL;
break;
}
return err < 0 ? err : 0;
}
/**
* audit_receive - receive messages from a netlink control socket
* @skb: the message buffer
*
* Parse the provided skb and deal with any messages that may be present,
* malformed skbs are discarded.
*/
static void audit_receive(struct sk_buff *skb)
{
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;
audit_ctl_lock();
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))
netlink_ack(skb, nlh, err, NULL);
nlh = nlmsg_next(nlh, &len);
}
audit_ctl_unlock();
/* can't block with the ctrl lock, so penalize the sender now */
if (audit_backlog_limit &&
(skb_queue_len(&audit_queue) > audit_backlog_limit)) {
DECLARE_WAITQUEUE(wait, current);
/* wake kauditd to try and flush the queue */
wake_up_interruptible(&kauditd_wait);
add_wait_queue_exclusive(&audit_backlog_wait, &wait);
set_current_state(TASK_UNINTERRUPTIBLE);
schedule_timeout(audit_backlog_wait_time);
remove_wait_queue(&audit_backlog_wait, &wait);
}
}
audit: log audit netlink multicast bind and unbind Log information about programs connecting to and disconnecting from the audit netlink multicast socket. This is needed so that during investigations a security officer can tell who or what had access to the audit trail. This helps to meet the FAU_SAR.2 requirement for Common Criteria. Here is the systemd startup event: type=PROCTITLE msg=audit(2020-04-22 10:10:21.787:10) : proctitle=/init type=SYSCALL msg=audit(2020-04-22 10:10:21.787:10) : arch=x86_64 syscall=bind success=yes exit=0 a0=0x19 a1=0x555f4aac7e90 a2=0xc a3=0x7ffcb792ff44 items=0 ppid=0 pid=1 auid=unset uid=root gid=root euid=root suid=root fsuid=root egid=root sgid=root fsgid=root tty=(none) ses=unset comm=systemd exe=/usr/lib/systemd/systemd subj=kernel key=(null) type=UNKNOWN[1335] msg=audit(2020-04-22 10:10:21.787:10) : pid=1 uid=root auid=unset tty=(none) ses=unset subj=kernel comm=systemd exe=/usr/lib/systemd/systemd nl-mcgrp=1 op=connect res=yes And events from the test suite that just uses close(): type=PROCTITLE msg=audit(2020-04-22 11:47:08.501:442) : proctitle=/usr/bin/perl -w amcast_joinpart/test type=SYSCALL msg=audit(2020-04-22 11:47:08.501:442) : arch=x86_64 syscall=bind success=yes exit=0 a0=0x7 a1=0x563004378760 a2=0xc a3=0x0 items=0 ppid=815 pid=818 auid=root uid=root gid=root euid=root suid=root fsuid=root egid=root sgid=root fsgid=root tty=ttyS0 ses=1 comm=perl exe=/usr/bin/perl subj=unconfined_u:unconfined_r:unconfined_t:s0-s0:c0.c1023 key=(null) type=UNKNOWN[1335] msg=audit(2020-04-22 11:47:08.501:442) : pid=818 uid=root auid=root tty=ttyS0 ses=1 subj=unconfined_u:unconfined_r:unconfined_t:s0-s0:c0.c1023 comm=perl exe=/usr/bin/perl nl-mcgrp=1 op=connect res=yes type=UNKNOWN[1335] msg=audit(2020-04-22 11:47:08.501:443) : pid=818 uid=root auid=root tty=ttyS0 ses=1 subj=unconfined_u:unconfined_r:unconfined_t:s0-s0:c0.c1023 comm=perl exe=/usr/bin/perl nl-mcgrp=1 op=disconnect res=yes And the events from the test suite using setsockopt with NETLINK_DROP_MEMBERSHIP: type=PROCTITLE msg=audit(2020-04-22 11:39:53.291:439) : proctitle=/usr/bin/perl -w amcast_joinpart/test type=SYSCALL msg=audit(2020-04-22 11:39:53.291:439) : arch=x86_64 syscall=bind success=yes exit=0 a0=0x7 a1=0x5560877c2d20 a2=0xc a3=0x0 items=0 ppid=772 pid=775 auid=root uid=root gid=root euid=root suid=root fsuid=root egid=root sgid=root fsgid=root tty=ttyS0 ses=1 comm=perl exe=/usr/bin/perl subj=unconfined_u:unconfined_r:unconfined_t:s0-s0:c0.c1023 key=(null) type=UNKNOWN[1335] msg=audit(2020-04-22 11:39:53.291:439) : pid=775 uid=root auid=root tty=ttyS0 ses=1 subj=unconfined_u:unconfined_r:unconfined_t:s0-s0:c0.c1023 comm=perl exe=/usr/bin/perl nl-mcgrp=1 op=connect res=yes type=PROCTITLE msg=audit(2020-04-22 11:39:53.292:440) : proctitle=/usr/bin/perl -w amcast_joinpart/test type=SYSCALL msg=audit(2020-04-22 11:39:53.292:440) : arch=x86_64 syscall=setsockopt success=yes exit=0 a0=0x7 a1=SOL_NETLINK a2=0x2 a3=0x7ffc8366f000 items=0 ppid=772 pid=775 auid=root uid=root gid=root euid=root suid=root fsuid=root egid=root sgid=root fsgid=root tty=ttyS0 ses=1 comm=perl exe=/usr/bin/perl subj=unconfined_u:unconfined_r:unconfined_t:s0-s0:c0.c1023 key=(null) type=UNKNOWN[1335] msg=audit(2020-04-22 11:39:53.292:440) : pid=775 uid=root auid=root tty=ttyS0 ses=1 subj=unconfined_u:unconfined_r:unconfined_t:s0-s0:c0.c1023 comm=perl exe=/usr/bin/perl nl-mcgrp=1 op=disconnect res=yes Please see the upstream issue tracker at https://github.com/linux-audit/audit-kernel/issues/28 With the feature description at https://github.com/linux-audit/audit-kernel/wiki/RFE-Audit-Multicast-Socket-Join-Part The testsuite support is at https://github.com/rgbriggs/audit-testsuite/compare/ghak28-mcast-part-join https://github.com/linux-audit/audit-testsuite/pull/93 And the userspace support patch is at https://github.com/linux-audit/audit-userspace/pull/114 Signed-off-by: Richard Guy Briggs <rgb@redhat.com> Signed-off-by: Paul Moore <paul@paul-moore.com>
2020-04-23 05:37:04 +08:00
/* Log information about who is connecting to the audit multicast socket */
static void audit_log_multicast(int group, const char *op, int err)
{
const struct cred *cred;
struct tty_struct *tty;
char comm[sizeof(current->comm)];
struct audit_buffer *ab;
if (!audit_enabled)
return;
ab = audit_log_start(audit_context(), GFP_KERNEL, AUDIT_EVENT_LISTENER);
if (!ab)
return;
cred = current_cred();
tty = audit_get_tty();
audit_log_format(ab, "pid=%u uid=%u auid=%u tty=%s ses=%u",
task_pid_nr(current),
from_kuid(&init_user_ns, cred->uid),
from_kuid(&init_user_ns, audit_get_loginuid(current)),
tty ? tty_name(tty) : "(none)",
audit_get_sessionid(current));
audit_put_tty(tty);
audit_log_task_context(ab); /* subj= */
audit_log_format(ab, " comm=");
audit_log_untrustedstring(ab, get_task_comm(comm, current));
audit_log_d_path_exe(ab, current->mm); /* exe= */
audit_log_format(ab, " nl-mcgrp=%d op=%s res=%d", group, op, !err);
audit_log_end(ab);
}
/* Run custom bind function on netlink socket group connect or bind requests. */
audit: log audit netlink multicast bind and unbind Log information about programs connecting to and disconnecting from the audit netlink multicast socket. This is needed so that during investigations a security officer can tell who or what had access to the audit trail. This helps to meet the FAU_SAR.2 requirement for Common Criteria. Here is the systemd startup event: type=PROCTITLE msg=audit(2020-04-22 10:10:21.787:10) : proctitle=/init type=SYSCALL msg=audit(2020-04-22 10:10:21.787:10) : arch=x86_64 syscall=bind success=yes exit=0 a0=0x19 a1=0x555f4aac7e90 a2=0xc a3=0x7ffcb792ff44 items=0 ppid=0 pid=1 auid=unset uid=root gid=root euid=root suid=root fsuid=root egid=root sgid=root fsgid=root tty=(none) ses=unset comm=systemd exe=/usr/lib/systemd/systemd subj=kernel key=(null) type=UNKNOWN[1335] msg=audit(2020-04-22 10:10:21.787:10) : pid=1 uid=root auid=unset tty=(none) ses=unset subj=kernel comm=systemd exe=/usr/lib/systemd/systemd nl-mcgrp=1 op=connect res=yes And events from the test suite that just uses close(): type=PROCTITLE msg=audit(2020-04-22 11:47:08.501:442) : proctitle=/usr/bin/perl -w amcast_joinpart/test type=SYSCALL msg=audit(2020-04-22 11:47:08.501:442) : arch=x86_64 syscall=bind success=yes exit=0 a0=0x7 a1=0x563004378760 a2=0xc a3=0x0 items=0 ppid=815 pid=818 auid=root uid=root gid=root euid=root suid=root fsuid=root egid=root sgid=root fsgid=root tty=ttyS0 ses=1 comm=perl exe=/usr/bin/perl subj=unconfined_u:unconfined_r:unconfined_t:s0-s0:c0.c1023 key=(null) type=UNKNOWN[1335] msg=audit(2020-04-22 11:47:08.501:442) : pid=818 uid=root auid=root tty=ttyS0 ses=1 subj=unconfined_u:unconfined_r:unconfined_t:s0-s0:c0.c1023 comm=perl exe=/usr/bin/perl nl-mcgrp=1 op=connect res=yes type=UNKNOWN[1335] msg=audit(2020-04-22 11:47:08.501:443) : pid=818 uid=root auid=root tty=ttyS0 ses=1 subj=unconfined_u:unconfined_r:unconfined_t:s0-s0:c0.c1023 comm=perl exe=/usr/bin/perl nl-mcgrp=1 op=disconnect res=yes And the events from the test suite using setsockopt with NETLINK_DROP_MEMBERSHIP: type=PROCTITLE msg=audit(2020-04-22 11:39:53.291:439) : proctitle=/usr/bin/perl -w amcast_joinpart/test type=SYSCALL msg=audit(2020-04-22 11:39:53.291:439) : arch=x86_64 syscall=bind success=yes exit=0 a0=0x7 a1=0x5560877c2d20 a2=0xc a3=0x0 items=0 ppid=772 pid=775 auid=root uid=root gid=root euid=root suid=root fsuid=root egid=root sgid=root fsgid=root tty=ttyS0 ses=1 comm=perl exe=/usr/bin/perl subj=unconfined_u:unconfined_r:unconfined_t:s0-s0:c0.c1023 key=(null) type=UNKNOWN[1335] msg=audit(2020-04-22 11:39:53.291:439) : pid=775 uid=root auid=root tty=ttyS0 ses=1 subj=unconfined_u:unconfined_r:unconfined_t:s0-s0:c0.c1023 comm=perl exe=/usr/bin/perl nl-mcgrp=1 op=connect res=yes type=PROCTITLE msg=audit(2020-04-22 11:39:53.292:440) : proctitle=/usr/bin/perl -w amcast_joinpart/test type=SYSCALL msg=audit(2020-04-22 11:39:53.292:440) : arch=x86_64 syscall=setsockopt success=yes exit=0 a0=0x7 a1=SOL_NETLINK a2=0x2 a3=0x7ffc8366f000 items=0 ppid=772 pid=775 auid=root uid=root gid=root euid=root suid=root fsuid=root egid=root sgid=root fsgid=root tty=ttyS0 ses=1 comm=perl exe=/usr/bin/perl subj=unconfined_u:unconfined_r:unconfined_t:s0-s0:c0.c1023 key=(null) type=UNKNOWN[1335] msg=audit(2020-04-22 11:39:53.292:440) : pid=775 uid=root auid=root tty=ttyS0 ses=1 subj=unconfined_u:unconfined_r:unconfined_t:s0-s0:c0.c1023 comm=perl exe=/usr/bin/perl nl-mcgrp=1 op=disconnect res=yes Please see the upstream issue tracker at https://github.com/linux-audit/audit-kernel/issues/28 With the feature description at https://github.com/linux-audit/audit-kernel/wiki/RFE-Audit-Multicast-Socket-Join-Part The testsuite support is at https://github.com/rgbriggs/audit-testsuite/compare/ghak28-mcast-part-join https://github.com/linux-audit/audit-testsuite/pull/93 And the userspace support patch is at https://github.com/linux-audit/audit-userspace/pull/114 Signed-off-by: Richard Guy Briggs <rgb@redhat.com> Signed-off-by: Paul Moore <paul@paul-moore.com>
2020-04-23 05:37:04 +08:00
static int audit_multicast_bind(struct net *net, int group)
{
audit: log audit netlink multicast bind and unbind Log information about programs connecting to and disconnecting from the audit netlink multicast socket. This is needed so that during investigations a security officer can tell who or what had access to the audit trail. This helps to meet the FAU_SAR.2 requirement for Common Criteria. Here is the systemd startup event: type=PROCTITLE msg=audit(2020-04-22 10:10:21.787:10) : proctitle=/init type=SYSCALL msg=audit(2020-04-22 10:10:21.787:10) : arch=x86_64 syscall=bind success=yes exit=0 a0=0x19 a1=0x555f4aac7e90 a2=0xc a3=0x7ffcb792ff44 items=0 ppid=0 pid=1 auid=unset uid=root gid=root euid=root suid=root fsuid=root egid=root sgid=root fsgid=root tty=(none) ses=unset comm=systemd exe=/usr/lib/systemd/systemd subj=kernel key=(null) type=UNKNOWN[1335] msg=audit(2020-04-22 10:10:21.787:10) : pid=1 uid=root auid=unset tty=(none) ses=unset subj=kernel comm=systemd exe=/usr/lib/systemd/systemd nl-mcgrp=1 op=connect res=yes And events from the test suite that just uses close(): type=PROCTITLE msg=audit(2020-04-22 11:47:08.501:442) : proctitle=/usr/bin/perl -w amcast_joinpart/test type=SYSCALL msg=audit(2020-04-22 11:47:08.501:442) : arch=x86_64 syscall=bind success=yes exit=0 a0=0x7 a1=0x563004378760 a2=0xc a3=0x0 items=0 ppid=815 pid=818 auid=root uid=root gid=root euid=root suid=root fsuid=root egid=root sgid=root fsgid=root tty=ttyS0 ses=1 comm=perl exe=/usr/bin/perl subj=unconfined_u:unconfined_r:unconfined_t:s0-s0:c0.c1023 key=(null) type=UNKNOWN[1335] msg=audit(2020-04-22 11:47:08.501:442) : pid=818 uid=root auid=root tty=ttyS0 ses=1 subj=unconfined_u:unconfined_r:unconfined_t:s0-s0:c0.c1023 comm=perl exe=/usr/bin/perl nl-mcgrp=1 op=connect res=yes type=UNKNOWN[1335] msg=audit(2020-04-22 11:47:08.501:443) : pid=818 uid=root auid=root tty=ttyS0 ses=1 subj=unconfined_u:unconfined_r:unconfined_t:s0-s0:c0.c1023 comm=perl exe=/usr/bin/perl nl-mcgrp=1 op=disconnect res=yes And the events from the test suite using setsockopt with NETLINK_DROP_MEMBERSHIP: type=PROCTITLE msg=audit(2020-04-22 11:39:53.291:439) : proctitle=/usr/bin/perl -w amcast_joinpart/test type=SYSCALL msg=audit(2020-04-22 11:39:53.291:439) : arch=x86_64 syscall=bind success=yes exit=0 a0=0x7 a1=0x5560877c2d20 a2=0xc a3=0x0 items=0 ppid=772 pid=775 auid=root uid=root gid=root euid=root suid=root fsuid=root egid=root sgid=root fsgid=root tty=ttyS0 ses=1 comm=perl exe=/usr/bin/perl subj=unconfined_u:unconfined_r:unconfined_t:s0-s0:c0.c1023 key=(null) type=UNKNOWN[1335] msg=audit(2020-04-22 11:39:53.291:439) : pid=775 uid=root auid=root tty=ttyS0 ses=1 subj=unconfined_u:unconfined_r:unconfined_t:s0-s0:c0.c1023 comm=perl exe=/usr/bin/perl nl-mcgrp=1 op=connect res=yes type=PROCTITLE msg=audit(2020-04-22 11:39:53.292:440) : proctitle=/usr/bin/perl -w amcast_joinpart/test type=SYSCALL msg=audit(2020-04-22 11:39:53.292:440) : arch=x86_64 syscall=setsockopt success=yes exit=0 a0=0x7 a1=SOL_NETLINK a2=0x2 a3=0x7ffc8366f000 items=0 ppid=772 pid=775 auid=root uid=root gid=root euid=root suid=root fsuid=root egid=root sgid=root fsgid=root tty=ttyS0 ses=1 comm=perl exe=/usr/bin/perl subj=unconfined_u:unconfined_r:unconfined_t:s0-s0:c0.c1023 key=(null) type=UNKNOWN[1335] msg=audit(2020-04-22 11:39:53.292:440) : pid=775 uid=root auid=root tty=ttyS0 ses=1 subj=unconfined_u:unconfined_r:unconfined_t:s0-s0:c0.c1023 comm=perl exe=/usr/bin/perl nl-mcgrp=1 op=disconnect res=yes Please see the upstream issue tracker at https://github.com/linux-audit/audit-kernel/issues/28 With the feature description at https://github.com/linux-audit/audit-kernel/wiki/RFE-Audit-Multicast-Socket-Join-Part The testsuite support is at https://github.com/rgbriggs/audit-testsuite/compare/ghak28-mcast-part-join https://github.com/linux-audit/audit-testsuite/pull/93 And the userspace support patch is at https://github.com/linux-audit/audit-userspace/pull/114 Signed-off-by: Richard Guy Briggs <rgb@redhat.com> Signed-off-by: Paul Moore <paul@paul-moore.com>
2020-04-23 05:37:04 +08:00
int err = 0;
if (!capable(CAP_AUDIT_READ))
audit: log audit netlink multicast bind and unbind Log information about programs connecting to and disconnecting from the audit netlink multicast socket. This is needed so that during investigations a security officer can tell who or what had access to the audit trail. This helps to meet the FAU_SAR.2 requirement for Common Criteria. Here is the systemd startup event: type=PROCTITLE msg=audit(2020-04-22 10:10:21.787:10) : proctitle=/init type=SYSCALL msg=audit(2020-04-22 10:10:21.787:10) : arch=x86_64 syscall=bind success=yes exit=0 a0=0x19 a1=0x555f4aac7e90 a2=0xc a3=0x7ffcb792ff44 items=0 ppid=0 pid=1 auid=unset uid=root gid=root euid=root suid=root fsuid=root egid=root sgid=root fsgid=root tty=(none) ses=unset comm=systemd exe=/usr/lib/systemd/systemd subj=kernel key=(null) type=UNKNOWN[1335] msg=audit(2020-04-22 10:10:21.787:10) : pid=1 uid=root auid=unset tty=(none) ses=unset subj=kernel comm=systemd exe=/usr/lib/systemd/systemd nl-mcgrp=1 op=connect res=yes And events from the test suite that just uses close(): type=PROCTITLE msg=audit(2020-04-22 11:47:08.501:442) : proctitle=/usr/bin/perl -w amcast_joinpart/test type=SYSCALL msg=audit(2020-04-22 11:47:08.501:442) : arch=x86_64 syscall=bind success=yes exit=0 a0=0x7 a1=0x563004378760 a2=0xc a3=0x0 items=0 ppid=815 pid=818 auid=root uid=root gid=root euid=root suid=root fsuid=root egid=root sgid=root fsgid=root tty=ttyS0 ses=1 comm=perl exe=/usr/bin/perl subj=unconfined_u:unconfined_r:unconfined_t:s0-s0:c0.c1023 key=(null) type=UNKNOWN[1335] msg=audit(2020-04-22 11:47:08.501:442) : pid=818 uid=root auid=root tty=ttyS0 ses=1 subj=unconfined_u:unconfined_r:unconfined_t:s0-s0:c0.c1023 comm=perl exe=/usr/bin/perl nl-mcgrp=1 op=connect res=yes type=UNKNOWN[1335] msg=audit(2020-04-22 11:47:08.501:443) : pid=818 uid=root auid=root tty=ttyS0 ses=1 subj=unconfined_u:unconfined_r:unconfined_t:s0-s0:c0.c1023 comm=perl exe=/usr/bin/perl nl-mcgrp=1 op=disconnect res=yes And the events from the test suite using setsockopt with NETLINK_DROP_MEMBERSHIP: type=PROCTITLE msg=audit(2020-04-22 11:39:53.291:439) : proctitle=/usr/bin/perl -w amcast_joinpart/test type=SYSCALL msg=audit(2020-04-22 11:39:53.291:439) : arch=x86_64 syscall=bind success=yes exit=0 a0=0x7 a1=0x5560877c2d20 a2=0xc a3=0x0 items=0 ppid=772 pid=775 auid=root uid=root gid=root euid=root suid=root fsuid=root egid=root sgid=root fsgid=root tty=ttyS0 ses=1 comm=perl exe=/usr/bin/perl subj=unconfined_u:unconfined_r:unconfined_t:s0-s0:c0.c1023 key=(null) type=UNKNOWN[1335] msg=audit(2020-04-22 11:39:53.291:439) : pid=775 uid=root auid=root tty=ttyS0 ses=1 subj=unconfined_u:unconfined_r:unconfined_t:s0-s0:c0.c1023 comm=perl exe=/usr/bin/perl nl-mcgrp=1 op=connect res=yes type=PROCTITLE msg=audit(2020-04-22 11:39:53.292:440) : proctitle=/usr/bin/perl -w amcast_joinpart/test type=SYSCALL msg=audit(2020-04-22 11:39:53.292:440) : arch=x86_64 syscall=setsockopt success=yes exit=0 a0=0x7 a1=SOL_NETLINK a2=0x2 a3=0x7ffc8366f000 items=0 ppid=772 pid=775 auid=root uid=root gid=root euid=root suid=root fsuid=root egid=root sgid=root fsgid=root tty=ttyS0 ses=1 comm=perl exe=/usr/bin/perl subj=unconfined_u:unconfined_r:unconfined_t:s0-s0:c0.c1023 key=(null) type=UNKNOWN[1335] msg=audit(2020-04-22 11:39:53.292:440) : pid=775 uid=root auid=root tty=ttyS0 ses=1 subj=unconfined_u:unconfined_r:unconfined_t:s0-s0:c0.c1023 comm=perl exe=/usr/bin/perl nl-mcgrp=1 op=disconnect res=yes Please see the upstream issue tracker at https://github.com/linux-audit/audit-kernel/issues/28 With the feature description at https://github.com/linux-audit/audit-kernel/wiki/RFE-Audit-Multicast-Socket-Join-Part The testsuite support is at https://github.com/rgbriggs/audit-testsuite/compare/ghak28-mcast-part-join https://github.com/linux-audit/audit-testsuite/pull/93 And the userspace support patch is at https://github.com/linux-audit/audit-userspace/pull/114 Signed-off-by: Richard Guy Briggs <rgb@redhat.com> Signed-off-by: Paul Moore <paul@paul-moore.com>
2020-04-23 05:37:04 +08:00
err = -EPERM;
audit_log_multicast(group, "connect", err);
return err;
}
audit: log audit netlink multicast bind and unbind Log information about programs connecting to and disconnecting from the audit netlink multicast socket. This is needed so that during investigations a security officer can tell who or what had access to the audit trail. This helps to meet the FAU_SAR.2 requirement for Common Criteria. Here is the systemd startup event: type=PROCTITLE msg=audit(2020-04-22 10:10:21.787:10) : proctitle=/init type=SYSCALL msg=audit(2020-04-22 10:10:21.787:10) : arch=x86_64 syscall=bind success=yes exit=0 a0=0x19 a1=0x555f4aac7e90 a2=0xc a3=0x7ffcb792ff44 items=0 ppid=0 pid=1 auid=unset uid=root gid=root euid=root suid=root fsuid=root egid=root sgid=root fsgid=root tty=(none) ses=unset comm=systemd exe=/usr/lib/systemd/systemd subj=kernel key=(null) type=UNKNOWN[1335] msg=audit(2020-04-22 10:10:21.787:10) : pid=1 uid=root auid=unset tty=(none) ses=unset subj=kernel comm=systemd exe=/usr/lib/systemd/systemd nl-mcgrp=1 op=connect res=yes And events from the test suite that just uses close(): type=PROCTITLE msg=audit(2020-04-22 11:47:08.501:442) : proctitle=/usr/bin/perl -w amcast_joinpart/test type=SYSCALL msg=audit(2020-04-22 11:47:08.501:442) : arch=x86_64 syscall=bind success=yes exit=0 a0=0x7 a1=0x563004378760 a2=0xc a3=0x0 items=0 ppid=815 pid=818 auid=root uid=root gid=root euid=root suid=root fsuid=root egid=root sgid=root fsgid=root tty=ttyS0 ses=1 comm=perl exe=/usr/bin/perl subj=unconfined_u:unconfined_r:unconfined_t:s0-s0:c0.c1023 key=(null) type=UNKNOWN[1335] msg=audit(2020-04-22 11:47:08.501:442) : pid=818 uid=root auid=root tty=ttyS0 ses=1 subj=unconfined_u:unconfined_r:unconfined_t:s0-s0:c0.c1023 comm=perl exe=/usr/bin/perl nl-mcgrp=1 op=connect res=yes type=UNKNOWN[1335] msg=audit(2020-04-22 11:47:08.501:443) : pid=818 uid=root auid=root tty=ttyS0 ses=1 subj=unconfined_u:unconfined_r:unconfined_t:s0-s0:c0.c1023 comm=perl exe=/usr/bin/perl nl-mcgrp=1 op=disconnect res=yes And the events from the test suite using setsockopt with NETLINK_DROP_MEMBERSHIP: type=PROCTITLE msg=audit(2020-04-22 11:39:53.291:439) : proctitle=/usr/bin/perl -w amcast_joinpart/test type=SYSCALL msg=audit(2020-04-22 11:39:53.291:439) : arch=x86_64 syscall=bind success=yes exit=0 a0=0x7 a1=0x5560877c2d20 a2=0xc a3=0x0 items=0 ppid=772 pid=775 auid=root uid=root gid=root euid=root suid=root fsuid=root egid=root sgid=root fsgid=root tty=ttyS0 ses=1 comm=perl exe=/usr/bin/perl subj=unconfined_u:unconfined_r:unconfined_t:s0-s0:c0.c1023 key=(null) type=UNKNOWN[1335] msg=audit(2020-04-22 11:39:53.291:439) : pid=775 uid=root auid=root tty=ttyS0 ses=1 subj=unconfined_u:unconfined_r:unconfined_t:s0-s0:c0.c1023 comm=perl exe=/usr/bin/perl nl-mcgrp=1 op=connect res=yes type=PROCTITLE msg=audit(2020-04-22 11:39:53.292:440) : proctitle=/usr/bin/perl -w amcast_joinpart/test type=SYSCALL msg=audit(2020-04-22 11:39:53.292:440) : arch=x86_64 syscall=setsockopt success=yes exit=0 a0=0x7 a1=SOL_NETLINK a2=0x2 a3=0x7ffc8366f000 items=0 ppid=772 pid=775 auid=root uid=root gid=root euid=root suid=root fsuid=root egid=root sgid=root fsgid=root tty=ttyS0 ses=1 comm=perl exe=/usr/bin/perl subj=unconfined_u:unconfined_r:unconfined_t:s0-s0:c0.c1023 key=(null) type=UNKNOWN[1335] msg=audit(2020-04-22 11:39:53.292:440) : pid=775 uid=root auid=root tty=ttyS0 ses=1 subj=unconfined_u:unconfined_r:unconfined_t:s0-s0:c0.c1023 comm=perl exe=/usr/bin/perl nl-mcgrp=1 op=disconnect res=yes Please see the upstream issue tracker at https://github.com/linux-audit/audit-kernel/issues/28 With the feature description at https://github.com/linux-audit/audit-kernel/wiki/RFE-Audit-Multicast-Socket-Join-Part The testsuite support is at https://github.com/rgbriggs/audit-testsuite/compare/ghak28-mcast-part-join https://github.com/linux-audit/audit-testsuite/pull/93 And the userspace support patch is at https://github.com/linux-audit/audit-userspace/pull/114 Signed-off-by: Richard Guy Briggs <rgb@redhat.com> Signed-off-by: Paul Moore <paul@paul-moore.com>
2020-04-23 05:37:04 +08:00
static void audit_multicast_unbind(struct net *net, int group)
{
audit_log_multicast(group, "disconnect", 0);
}
static int __net_init audit_net_init(struct net *net)
{
struct netlink_kernel_cfg cfg = {
.input = audit_receive,
audit: log audit netlink multicast bind and unbind Log information about programs connecting to and disconnecting from the audit netlink multicast socket. This is needed so that during investigations a security officer can tell who or what had access to the audit trail. This helps to meet the FAU_SAR.2 requirement for Common Criteria. Here is the systemd startup event: type=PROCTITLE msg=audit(2020-04-22 10:10:21.787:10) : proctitle=/init type=SYSCALL msg=audit(2020-04-22 10:10:21.787:10) : arch=x86_64 syscall=bind success=yes exit=0 a0=0x19 a1=0x555f4aac7e90 a2=0xc a3=0x7ffcb792ff44 items=0 ppid=0 pid=1 auid=unset uid=root gid=root euid=root suid=root fsuid=root egid=root sgid=root fsgid=root tty=(none) ses=unset comm=systemd exe=/usr/lib/systemd/systemd subj=kernel key=(null) type=UNKNOWN[1335] msg=audit(2020-04-22 10:10:21.787:10) : pid=1 uid=root auid=unset tty=(none) ses=unset subj=kernel comm=systemd exe=/usr/lib/systemd/systemd nl-mcgrp=1 op=connect res=yes And events from the test suite that just uses close(): type=PROCTITLE msg=audit(2020-04-22 11:47:08.501:442) : proctitle=/usr/bin/perl -w amcast_joinpart/test type=SYSCALL msg=audit(2020-04-22 11:47:08.501:442) : arch=x86_64 syscall=bind success=yes exit=0 a0=0x7 a1=0x563004378760 a2=0xc a3=0x0 items=0 ppid=815 pid=818 auid=root uid=root gid=root euid=root suid=root fsuid=root egid=root sgid=root fsgid=root tty=ttyS0 ses=1 comm=perl exe=/usr/bin/perl subj=unconfined_u:unconfined_r:unconfined_t:s0-s0:c0.c1023 key=(null) type=UNKNOWN[1335] msg=audit(2020-04-22 11:47:08.501:442) : pid=818 uid=root auid=root tty=ttyS0 ses=1 subj=unconfined_u:unconfined_r:unconfined_t:s0-s0:c0.c1023 comm=perl exe=/usr/bin/perl nl-mcgrp=1 op=connect res=yes type=UNKNOWN[1335] msg=audit(2020-04-22 11:47:08.501:443) : pid=818 uid=root auid=root tty=ttyS0 ses=1 subj=unconfined_u:unconfined_r:unconfined_t:s0-s0:c0.c1023 comm=perl exe=/usr/bin/perl nl-mcgrp=1 op=disconnect res=yes And the events from the test suite using setsockopt with NETLINK_DROP_MEMBERSHIP: type=PROCTITLE msg=audit(2020-04-22 11:39:53.291:439) : proctitle=/usr/bin/perl -w amcast_joinpart/test type=SYSCALL msg=audit(2020-04-22 11:39:53.291:439) : arch=x86_64 syscall=bind success=yes exit=0 a0=0x7 a1=0x5560877c2d20 a2=0xc a3=0x0 items=0 ppid=772 pid=775 auid=root uid=root gid=root euid=root suid=root fsuid=root egid=root sgid=root fsgid=root tty=ttyS0 ses=1 comm=perl exe=/usr/bin/perl subj=unconfined_u:unconfined_r:unconfined_t:s0-s0:c0.c1023 key=(null) type=UNKNOWN[1335] msg=audit(2020-04-22 11:39:53.291:439) : pid=775 uid=root auid=root tty=ttyS0 ses=1 subj=unconfined_u:unconfined_r:unconfined_t:s0-s0:c0.c1023 comm=perl exe=/usr/bin/perl nl-mcgrp=1 op=connect res=yes type=PROCTITLE msg=audit(2020-04-22 11:39:53.292:440) : proctitle=/usr/bin/perl -w amcast_joinpart/test type=SYSCALL msg=audit(2020-04-22 11:39:53.292:440) : arch=x86_64 syscall=setsockopt success=yes exit=0 a0=0x7 a1=SOL_NETLINK a2=0x2 a3=0x7ffc8366f000 items=0 ppid=772 pid=775 auid=root uid=root gid=root euid=root suid=root fsuid=root egid=root sgid=root fsgid=root tty=ttyS0 ses=1 comm=perl exe=/usr/bin/perl subj=unconfined_u:unconfined_r:unconfined_t:s0-s0:c0.c1023 key=(null) type=UNKNOWN[1335] msg=audit(2020-04-22 11:39:53.292:440) : pid=775 uid=root auid=root tty=ttyS0 ses=1 subj=unconfined_u:unconfined_r:unconfined_t:s0-s0:c0.c1023 comm=perl exe=/usr/bin/perl nl-mcgrp=1 op=disconnect res=yes Please see the upstream issue tracker at https://github.com/linux-audit/audit-kernel/issues/28 With the feature description at https://github.com/linux-audit/audit-kernel/wiki/RFE-Audit-Multicast-Socket-Join-Part The testsuite support is at https://github.com/rgbriggs/audit-testsuite/compare/ghak28-mcast-part-join https://github.com/linux-audit/audit-testsuite/pull/93 And the userspace support patch is at https://github.com/linux-audit/audit-userspace/pull/114 Signed-off-by: Richard Guy Briggs <rgb@redhat.com> Signed-off-by: Paul Moore <paul@paul-moore.com>
2020-04-23 05:37:04 +08:00
.bind = audit_multicast_bind,
.unbind = audit_multicast_unbind,
.flags = NL_CFG_F_NONROOT_RECV,
.groups = AUDIT_NLGRP_MAX,
};
struct audit_net *aunet = net_generic(net, audit_net_id);
audit: fix auditd/kernel connection state tracking What started as a rather straightforward race condition reported by Dmitry using the syzkaller fuzzer ended up revealing some major problems with how the audit subsystem managed its netlink sockets and its connection with the userspace audit daemon. Fixing this properly had quite the cascading effect and what we are left with is this rather large and complicated patch. My initial goal was to try and decompose this patch into multiple smaller patches, but the way these changes are intertwined makes it difficult to split these changes into meaningful pieces that don't break or somehow make things worse for the intermediate states. The patch makes a number of changes, but the most significant are highlighted below: * The auditd tracking variables, e.g. audit_sock, are now gone and replaced by a RCU/spin_lock protected variable auditd_conn which is a structure containing all of the auditd tracking information. * We no longer track the auditd sock directly, instead we track it via the network namespace in which it resides and we use the audit socket associated with that namespace. In spirit, this is what the code was trying to do prior to this patch (at least I think that is what the original authors intended), but it was done rather poorly and added a layer of obfuscation that only masked the underlying problems. * Big backlog queue cleanup, again. In v4.10 we made some pretty big changes to how the audit backlog queues work, here we haven't changed the queue design so much as cleaned up the implementation. Brought about by the locking changes, we've simplified kauditd_thread() quite a bit by consolidating the queue handling into a new helper function, kauditd_send_queue(), which allows us to eliminate a lot of very similar code and makes the looping logic in kauditd_thread() clearer. * All netlink messages sent to auditd are now sent via auditd_send_unicast_skb(). Other than just making sense, this makes the lock handling easier. * Change the audit_log_start() sleep behavior so that we never sleep on auditd events (unchanged) or if the caller is holding the audit_cmd_mutex (changed). Previously we didn't sleep if the caller was auditd or if the message type fell between a certain range; the type check was a poor effort of doing what the cmd_mutex check now does. Richard Guy Briggs originally proposed not sleeping the cmd_mutex owner several years ago but his patch wasn't acceptable at the time. At least the idea lives on here. * A problem with the lost record counter has been resolved. Steve Grubb and I both happened to notice this problem and according to some quick testing by Steve, this problem goes back quite some time. It's largely a harmless problem, although it may have left some careful sysadmins quite puzzled. Cc: <stable@vger.kernel.org> # 4.10.x- Reported-by: Dmitry Vyukov <dvyukov@google.com> Signed-off-by: Paul Moore <paul@paul-moore.com>
2017-03-21 23:26:35 +08:00
aunet->sk = netlink_kernel_create(net, NETLINK_AUDIT, &cfg);
if (aunet->sk == NULL) {
audit_panic("cannot initialize netlink socket in namespace");
return -ENOMEM;
}
audit: improve robustness of the audit queue handling If the audit daemon were ever to get stuck in a stopped state the kernel's kauditd_thread() could get blocked attempting to send audit records to the userspace audit daemon. With the kernel thread blocked it is possible that the audit queue could grow unbounded as certain audit record generating events must be exempt from the queue limits else the system enter a deadlock state. This patch resolves this problem by lowering the kernel thread's socket sending timeout from MAX_SCHEDULE_TIMEOUT to HZ/10 and tweaks the kauditd_send_queue() function to better manage the various audit queues when connection problems occur between the kernel and the audit daemon. With this patch, the backlog may temporarily grow beyond the defined limits when the audit daemon is stopped and the system is under heavy audit pressure, but kauditd_thread() will continue to make progress and drain the queues as it would for other connection problems. For example, with the audit daemon put into a stopped state and the system configured to audit every syscall it was still possible to shutdown the system without a kernel panic, deadlock, etc.; granted, the system was slow to shutdown but that is to be expected given the extreme pressure of recording every syscall. The timeout value of HZ/10 was chosen primarily through experimentation and this developer's "gut feeling". There is likely no one perfect value, but as this scenario is limited in scope (root privileges would be needed to send SIGSTOP to the audit daemon), it is likely not worth exposing this as a tunable at present. This can always be done at a later date if it proves necessary. Cc: stable@vger.kernel.org Fixes: 5b52330bbfe63 ("audit: fix auditd/kernel connection state tracking") Reported-by: Gaosheng Cui <cuigaosheng1@huawei.com> Tested-by: Gaosheng Cui <cuigaosheng1@huawei.com> Reviewed-by: Richard Guy Briggs <rgb@redhat.com> Signed-off-by: Paul Moore <paul@paul-moore.com>
2021-12-10 00:46:07 +08:00
/* limit the timeout in case auditd is blocked/stopped */
aunet->sk->sk_sndtimeo = HZ / 10;
audit: fix auditd/kernel connection state tracking What started as a rather straightforward race condition reported by Dmitry using the syzkaller fuzzer ended up revealing some major problems with how the audit subsystem managed its netlink sockets and its connection with the userspace audit daemon. Fixing this properly had quite the cascading effect and what we are left with is this rather large and complicated patch. My initial goal was to try and decompose this patch into multiple smaller patches, but the way these changes are intertwined makes it difficult to split these changes into meaningful pieces that don't break or somehow make things worse for the intermediate states. The patch makes a number of changes, but the most significant are highlighted below: * The auditd tracking variables, e.g. audit_sock, are now gone and replaced by a RCU/spin_lock protected variable auditd_conn which is a structure containing all of the auditd tracking information. * We no longer track the auditd sock directly, instead we track it via the network namespace in which it resides and we use the audit socket associated with that namespace. In spirit, this is what the code was trying to do prior to this patch (at least I think that is what the original authors intended), but it was done rather poorly and added a layer of obfuscation that only masked the underlying problems. * Big backlog queue cleanup, again. In v4.10 we made some pretty big changes to how the audit backlog queues work, here we haven't changed the queue design so much as cleaned up the implementation. Brought about by the locking changes, we've simplified kauditd_thread() quite a bit by consolidating the queue handling into a new helper function, kauditd_send_queue(), which allows us to eliminate a lot of very similar code and makes the looping logic in kauditd_thread() clearer. * All netlink messages sent to auditd are now sent via auditd_send_unicast_skb(). Other than just making sense, this makes the lock handling easier. * Change the audit_log_start() sleep behavior so that we never sleep on auditd events (unchanged) or if the caller is holding the audit_cmd_mutex (changed). Previously we didn't sleep if the caller was auditd or if the message type fell between a certain range; the type check was a poor effort of doing what the cmd_mutex check now does. Richard Guy Briggs originally proposed not sleeping the cmd_mutex owner several years ago but his patch wasn't acceptable at the time. At least the idea lives on here. * A problem with the lost record counter has been resolved. Steve Grubb and I both happened to notice this problem and according to some quick testing by Steve, this problem goes back quite some time. It's largely a harmless problem, although it may have left some careful sysadmins quite puzzled. Cc: <stable@vger.kernel.org> # 4.10.x- Reported-by: Dmitry Vyukov <dvyukov@google.com> Signed-off-by: Paul Moore <paul@paul-moore.com>
2017-03-21 23:26:35 +08:00
return 0;
}
static void __net_exit audit_net_exit(struct net *net)
{
struct audit_net *aunet = net_generic(net, audit_net_id);
audit: fix auditd/kernel connection state tracking What started as a rather straightforward race condition reported by Dmitry using the syzkaller fuzzer ended up revealing some major problems with how the audit subsystem managed its netlink sockets and its connection with the userspace audit daemon. Fixing this properly had quite the cascading effect and what we are left with is this rather large and complicated patch. My initial goal was to try and decompose this patch into multiple smaller patches, but the way these changes are intertwined makes it difficult to split these changes into meaningful pieces that don't break or somehow make things worse for the intermediate states. The patch makes a number of changes, but the most significant are highlighted below: * The auditd tracking variables, e.g. audit_sock, are now gone and replaced by a RCU/spin_lock protected variable auditd_conn which is a structure containing all of the auditd tracking information. * We no longer track the auditd sock directly, instead we track it via the network namespace in which it resides and we use the audit socket associated with that namespace. In spirit, this is what the code was trying to do prior to this patch (at least I think that is what the original authors intended), but it was done rather poorly and added a layer of obfuscation that only masked the underlying problems. * Big backlog queue cleanup, again. In v4.10 we made some pretty big changes to how the audit backlog queues work, here we haven't changed the queue design so much as cleaned up the implementation. Brought about by the locking changes, we've simplified kauditd_thread() quite a bit by consolidating the queue handling into a new helper function, kauditd_send_queue(), which allows us to eliminate a lot of very similar code and makes the looping logic in kauditd_thread() clearer. * All netlink messages sent to auditd are now sent via auditd_send_unicast_skb(). Other than just making sense, this makes the lock handling easier. * Change the audit_log_start() sleep behavior so that we never sleep on auditd events (unchanged) or if the caller is holding the audit_cmd_mutex (changed). Previously we didn't sleep if the caller was auditd or if the message type fell between a certain range; the type check was a poor effort of doing what the cmd_mutex check now does. Richard Guy Briggs originally proposed not sleeping the cmd_mutex owner several years ago but his patch wasn't acceptable at the time. At least the idea lives on here. * A problem with the lost record counter has been resolved. Steve Grubb and I both happened to notice this problem and according to some quick testing by Steve, this problem goes back quite some time. It's largely a harmless problem, although it may have left some careful sysadmins quite puzzled. Cc: <stable@vger.kernel.org> # 4.10.x- Reported-by: Dmitry Vyukov <dvyukov@google.com> Signed-off-by: Paul Moore <paul@paul-moore.com>
2017-03-21 23:26:35 +08:00
/* NOTE: you would think that we would want to check the auditd
* connection and potentially reset it here if it lives in this
* namespace, but since the auditd connection tracking struct holds a
* reference to this namespace (see auditd_set()) we are only ever
* going to get here after that connection has been released */
audit: fix auditd/kernel connection state tracking What started as a rather straightforward race condition reported by Dmitry using the syzkaller fuzzer ended up revealing some major problems with how the audit subsystem managed its netlink sockets and its connection with the userspace audit daemon. Fixing this properly had quite the cascading effect and what we are left with is this rather large and complicated patch. My initial goal was to try and decompose this patch into multiple smaller patches, but the way these changes are intertwined makes it difficult to split these changes into meaningful pieces that don't break or somehow make things worse for the intermediate states. The patch makes a number of changes, but the most significant are highlighted below: * The auditd tracking variables, e.g. audit_sock, are now gone and replaced by a RCU/spin_lock protected variable auditd_conn which is a structure containing all of the auditd tracking information. * We no longer track the auditd sock directly, instead we track it via the network namespace in which it resides and we use the audit socket associated with that namespace. In spirit, this is what the code was trying to do prior to this patch (at least I think that is what the original authors intended), but it was done rather poorly and added a layer of obfuscation that only masked the underlying problems. * Big backlog queue cleanup, again. In v4.10 we made some pretty big changes to how the audit backlog queues work, here we haven't changed the queue design so much as cleaned up the implementation. Brought about by the locking changes, we've simplified kauditd_thread() quite a bit by consolidating the queue handling into a new helper function, kauditd_send_queue(), which allows us to eliminate a lot of very similar code and makes the looping logic in kauditd_thread() clearer. * All netlink messages sent to auditd are now sent via auditd_send_unicast_skb(). Other than just making sense, this makes the lock handling easier. * Change the audit_log_start() sleep behavior so that we never sleep on auditd events (unchanged) or if the caller is holding the audit_cmd_mutex (changed). Previously we didn't sleep if the caller was auditd or if the message type fell between a certain range; the type check was a poor effort of doing what the cmd_mutex check now does. Richard Guy Briggs originally proposed not sleeping the cmd_mutex owner several years ago but his patch wasn't acceptable at the time. At least the idea lives on here. * A problem with the lost record counter has been resolved. Steve Grubb and I both happened to notice this problem and according to some quick testing by Steve, this problem goes back quite some time. It's largely a harmless problem, although it may have left some careful sysadmins quite puzzled. Cc: <stable@vger.kernel.org> # 4.10.x- Reported-by: Dmitry Vyukov <dvyukov@google.com> Signed-off-by: Paul Moore <paul@paul-moore.com>
2017-03-21 23:26:35 +08:00
netlink_kernel_release(aunet->sk);
}
static struct pernet_operations audit_net_ops __net_initdata = {
.init = audit_net_init,
.exit = audit_net_exit,
.id = &audit_net_id,
.size = sizeof(struct audit_net),
};
/* Initialize audit support at boot time. */
static int __init audit_init(void)
{
int i;
if (audit_initialized == AUDIT_DISABLED)
return 0;
audit_buffer_cache = kmem_cache_create("audit_buffer",
sizeof(struct audit_buffer),
0, SLAB_PANIC, NULL);
skb_queue_head_init(&audit_queue);
skb_queue_head_init(&audit_retry_queue);
skb_queue_head_init(&audit_hold_queue);
for (i = 0; i < AUDIT_INODE_BUCKETS; i++)
INIT_LIST_HEAD(&audit_inode_hash[i]);
mutex_init(&audit_cmd_mutex.lock);
audit_cmd_mutex.owner = NULL;
audit: fix auditd/kernel connection state tracking What started as a rather straightforward race condition reported by Dmitry using the syzkaller fuzzer ended up revealing some major problems with how the audit subsystem managed its netlink sockets and its connection with the userspace audit daemon. Fixing this properly had quite the cascading effect and what we are left with is this rather large and complicated patch. My initial goal was to try and decompose this patch into multiple smaller patches, but the way these changes are intertwined makes it difficult to split these changes into meaningful pieces that don't break or somehow make things worse for the intermediate states. The patch makes a number of changes, but the most significant are highlighted below: * The auditd tracking variables, e.g. audit_sock, are now gone and replaced by a RCU/spin_lock protected variable auditd_conn which is a structure containing all of the auditd tracking information. * We no longer track the auditd sock directly, instead we track it via the network namespace in which it resides and we use the audit socket associated with that namespace. In spirit, this is what the code was trying to do prior to this patch (at least I think that is what the original authors intended), but it was done rather poorly and added a layer of obfuscation that only masked the underlying problems. * Big backlog queue cleanup, again. In v4.10 we made some pretty big changes to how the audit backlog queues work, here we haven't changed the queue design so much as cleaned up the implementation. Brought about by the locking changes, we've simplified kauditd_thread() quite a bit by consolidating the queue handling into a new helper function, kauditd_send_queue(), which allows us to eliminate a lot of very similar code and makes the looping logic in kauditd_thread() clearer. * All netlink messages sent to auditd are now sent via auditd_send_unicast_skb(). Other than just making sense, this makes the lock handling easier. * Change the audit_log_start() sleep behavior so that we never sleep on auditd events (unchanged) or if the caller is holding the audit_cmd_mutex (changed). Previously we didn't sleep if the caller was auditd or if the message type fell between a certain range; the type check was a poor effort of doing what the cmd_mutex check now does. Richard Guy Briggs originally proposed not sleeping the cmd_mutex owner several years ago but his patch wasn't acceptable at the time. At least the idea lives on here. * A problem with the lost record counter has been resolved. Steve Grubb and I both happened to notice this problem and according to some quick testing by Steve, this problem goes back quite some time. It's largely a harmless problem, although it may have left some careful sysadmins quite puzzled. Cc: <stable@vger.kernel.org> # 4.10.x- Reported-by: Dmitry Vyukov <dvyukov@google.com> Signed-off-by: Paul Moore <paul@paul-moore.com>
2017-03-21 23:26:35 +08:00
pr_info("initializing netlink subsys (%s)\n",
audit_default ? "enabled" : "disabled");
register_pernet_subsys(&audit_net_ops);
audit_initialized = AUDIT_INITIALIZED;
kauditd_task = kthread_run(kauditd_thread, NULL, "kauditd");
if (IS_ERR(kauditd_task)) {
int err = PTR_ERR(kauditd_task);
panic("audit: failed to start the kauditd thread (%d)\n", err);
}
audit_log(NULL, GFP_KERNEL, AUDIT_KERNEL,
"state=initialized audit_enabled=%u res=1",
audit_enabled);
return 0;
}
postcore_initcall(audit_init);
/*
* Process kernel command-line parameter at boot time.
* audit={0|off} or audit={1|on}.
*/
static int __init audit_enable(char *str)
{
if (!strcasecmp(str, "off") || !strcmp(str, "0"))
audit_default = AUDIT_OFF;
else if (!strcasecmp(str, "on") || !strcmp(str, "1"))
audit_default = AUDIT_ON;
else {
pr_err("audit: invalid 'audit' parameter value (%s)\n", str);
audit_default = AUDIT_ON;
}
if (audit_default == AUDIT_OFF)
audit_initialized = AUDIT_DISABLED;
if (audit_set_enabled(audit_default))
pr_err("audit: error setting audit state (%d)\n",
audit_default);
pr_info("%s\n", audit_default ?
"enabled (after initialization)" : "disabled (until reboot)");
return 1;
}
__setup("audit=", audit_enable);
audit: add kernel set-up parameter to override default backlog limit The default audit_backlog_limit is 64. This was a reasonable limit at one time. systemd causes so much audit queue activity on startup that auditd doesn't start before the backlog queue has already overflowed by more than a factor of 2. On a system with audit= not set on the kernel command line, this isn't an issue since that history isn't kept for auditd when it is available. On a system with audit=1 set on the kernel command line, kaudit tries to keep that history until auditd is able to drain the queue. This default can be changed by the "-b" option in audit.rules once the system has booted, but won't help with lost messages on boot. One way to solve this would be to increase the default backlog queue size to avoid losing any messages before auditd is able to consume them. This would be overkill to the embedded community and insufficient for some servers. Another way to solve it might be to add a kconfig option to set the default based on the system type. An embedded system would get the current (or smaller) default, while Workstations might get more than now and servers might get more. None of these solutions helps if a system's compiled default is too small to see the lost messages without compiling a new kernel. This patch adds a kernel set-up parameter (audit already has one to enable/disable it) "audit_backlog_limit=<n>" that overrides the default to allow the system administrator to set the backlog limit. Signed-off-by: Richard Guy Briggs <rgb@redhat.com> Signed-off-by: Eric Paris <eparis@redhat.com>
2013-09-18 00:34:52 +08:00
/* Process kernel command-line parameter at boot time.
* audit_backlog_limit=<n> */
static int __init audit_backlog_limit_set(char *str)
{
u32 audit_backlog_limit_arg;
audit: add kernel set-up parameter to override default backlog limit The default audit_backlog_limit is 64. This was a reasonable limit at one time. systemd causes so much audit queue activity on startup that auditd doesn't start before the backlog queue has already overflowed by more than a factor of 2. On a system with audit= not set on the kernel command line, this isn't an issue since that history isn't kept for auditd when it is available. On a system with audit=1 set on the kernel command line, kaudit tries to keep that history until auditd is able to drain the queue. This default can be changed by the "-b" option in audit.rules once the system has booted, but won't help with lost messages on boot. One way to solve this would be to increase the default backlog queue size to avoid losing any messages before auditd is able to consume them. This would be overkill to the embedded community and insufficient for some servers. Another way to solve it might be to add a kconfig option to set the default based on the system type. An embedded system would get the current (or smaller) default, while Workstations might get more than now and servers might get more. None of these solutions helps if a system's compiled default is too small to see the lost messages without compiling a new kernel. This patch adds a kernel set-up parameter (audit already has one to enable/disable it) "audit_backlog_limit=<n>" that overrides the default to allow the system administrator to set the backlog limit. Signed-off-by: Richard Guy Briggs <rgb@redhat.com> Signed-off-by: Eric Paris <eparis@redhat.com>
2013-09-18 00:34:52 +08:00
pr_info("audit_backlog_limit: ");
if (kstrtouint(str, 0, &audit_backlog_limit_arg)) {
pr_cont("using default of %u, unable to parse %s\n",
audit_backlog_limit, str);
audit: add kernel set-up parameter to override default backlog limit The default audit_backlog_limit is 64. This was a reasonable limit at one time. systemd causes so much audit queue activity on startup that auditd doesn't start before the backlog queue has already overflowed by more than a factor of 2. On a system with audit= not set on the kernel command line, this isn't an issue since that history isn't kept for auditd when it is available. On a system with audit=1 set on the kernel command line, kaudit tries to keep that history until auditd is able to drain the queue. This default can be changed by the "-b" option in audit.rules once the system has booted, but won't help with lost messages on boot. One way to solve this would be to increase the default backlog queue size to avoid losing any messages before auditd is able to consume them. This would be overkill to the embedded community and insufficient for some servers. Another way to solve it might be to add a kconfig option to set the default based on the system type. An embedded system would get the current (or smaller) default, while Workstations might get more than now and servers might get more. None of these solutions helps if a system's compiled default is too small to see the lost messages without compiling a new kernel. This patch adds a kernel set-up parameter (audit already has one to enable/disable it) "audit_backlog_limit=<n>" that overrides the default to allow the system administrator to set the backlog limit. Signed-off-by: Richard Guy Briggs <rgb@redhat.com> Signed-off-by: Eric Paris <eparis@redhat.com>
2013-09-18 00:34:52 +08:00
return 1;
}
audit_backlog_limit = audit_backlog_limit_arg;
pr_cont("%d\n", audit_backlog_limit);
audit: add kernel set-up parameter to override default backlog limit The default audit_backlog_limit is 64. This was a reasonable limit at one time. systemd causes so much audit queue activity on startup that auditd doesn't start before the backlog queue has already overflowed by more than a factor of 2. On a system with audit= not set on the kernel command line, this isn't an issue since that history isn't kept for auditd when it is available. On a system with audit=1 set on the kernel command line, kaudit tries to keep that history until auditd is able to drain the queue. This default can be changed by the "-b" option in audit.rules once the system has booted, but won't help with lost messages on boot. One way to solve this would be to increase the default backlog queue size to avoid losing any messages before auditd is able to consume them. This would be overkill to the embedded community and insufficient for some servers. Another way to solve it might be to add a kconfig option to set the default based on the system type. An embedded system would get the current (or smaller) default, while Workstations might get more than now and servers might get more. None of these solutions helps if a system's compiled default is too small to see the lost messages without compiling a new kernel. This patch adds a kernel set-up parameter (audit already has one to enable/disable it) "audit_backlog_limit=<n>" that overrides the default to allow the system administrator to set the backlog limit. Signed-off-by: Richard Guy Briggs <rgb@redhat.com> Signed-off-by: Eric Paris <eparis@redhat.com>
2013-09-18 00:34:52 +08:00
return 1;
}
__setup("audit_backlog_limit=", audit_backlog_limit_set);
static void audit_buffer_free(struct audit_buffer *ab)
{
if (!ab)
return;
kfree_skb(ab->skb);
kmem_cache_free(audit_buffer_cache, ab);
}
static struct audit_buffer *audit_buffer_alloc(struct audit_context *ctx,
gfp_t gfp_mask, int type)
{
struct audit_buffer *ab;
ab = kmem_cache_alloc(audit_buffer_cache, gfp_mask);
if (!ab)
return NULL;
ab->skb = nlmsg_new(AUDIT_BUFSIZ, gfp_mask);
if (!ab->skb)
goto err;
if (!nlmsg_put(ab->skb, 0, 0, type, 0, 0))
goto err;
ab->ctx = ctx;
ab->gfp_mask = gfp_mask;
return ab;
err:
audit_buffer_free(ab);
return NULL;
}
/**
* audit_serial - compute a serial number for the audit record
*
* Compute a serial number for the audit record. Audit records are
* 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
* halts).
*/
unsigned int audit_serial(void)
{
static atomic_t serial = ATOMIC_INIT(0);
return atomic_inc_return(&serial);
}
static inline void audit_get_stamp(struct audit_context *ctx,
struct timespec64 *t, unsigned int *serial)
{
if (!ctx || !auditsc_get_stamp(ctx, t, serial)) {
ktime_get_coarse_real_ts64(t);
*serial = audit_serial();
}
}
/**
* 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.
*/
struct audit_buffer *audit_log_start(struct audit_context *ctx, gfp_t gfp_mask,
int type)
{
struct audit_buffer *ab;
struct timespec64 t;
treewide: Remove uninitialized_var() usage Using uninitialized_var() is dangerous as it papers over real bugs[1] (or can in the future), and suppresses unrelated compiler warnings (e.g. "unused variable"). If the compiler thinks it is uninitialized, either simply initialize the variable or make compiler changes. In preparation for removing[2] the[3] macro[4], remove all remaining needless uses with the following script: git grep '\buninitialized_var\b' | cut -d: -f1 | sort -u | \ xargs perl -pi -e \ 's/\buninitialized_var\(([^\)]+)\)/\1/g; s:\s*/\* (GCC be quiet|to make compiler happy) \*/$::g;' drivers/video/fbdev/riva/riva_hw.c was manually tweaked to avoid pathological white-space. No outstanding warnings were found building allmodconfig with GCC 9.3.0 for x86_64, i386, arm64, arm, powerpc, powerpc64le, s390x, mips, sparc64, alpha, and m68k. [1] https://lore.kernel.org/lkml/20200603174714.192027-1-glider@google.com/ [2] https://lore.kernel.org/lkml/CA+55aFw+Vbj0i=1TGqCR5vQkCzWJ0QxK6CernOU6eedsudAixw@mail.gmail.com/ [3] https://lore.kernel.org/lkml/CA+55aFwgbgqhbp1fkxvRKEpzyR5J8n1vKT1VZdz9knmPuXhOeg@mail.gmail.com/ [4] https://lore.kernel.org/lkml/CA+55aFz2500WfbKXAx8s67wrm9=yVJu65TpLgN_ybYNv0VEOKA@mail.gmail.com/ Reviewed-by: Leon Romanovsky <leonro@mellanox.com> # drivers/infiniband and mlx4/mlx5 Acked-by: Jason Gunthorpe <jgg@mellanox.com> # IB Acked-by: Kalle Valo <kvalo@codeaurora.org> # wireless drivers Reviewed-by: Chao Yu <yuchao0@huawei.com> # erofs Signed-off-by: Kees Cook <keescook@chromium.org>
2020-06-04 04:09:38 +08:00
unsigned int serial;
if (audit_initialized != AUDIT_INITIALIZED)
return NULL;
if (unlikely(!audit_filter(type, AUDIT_FILTER_EXCLUDE)))
return NULL;
audit: fix auditd/kernel connection state tracking What started as a rather straightforward race condition reported by Dmitry using the syzkaller fuzzer ended up revealing some major problems with how the audit subsystem managed its netlink sockets and its connection with the userspace audit daemon. Fixing this properly had quite the cascading effect and what we are left with is this rather large and complicated patch. My initial goal was to try and decompose this patch into multiple smaller patches, but the way these changes are intertwined makes it difficult to split these changes into meaningful pieces that don't break or somehow make things worse for the intermediate states. The patch makes a number of changes, but the most significant are highlighted below: * The auditd tracking variables, e.g. audit_sock, are now gone and replaced by a RCU/spin_lock protected variable auditd_conn which is a structure containing all of the auditd tracking information. * We no longer track the auditd sock directly, instead we track it via the network namespace in which it resides and we use the audit socket associated with that namespace. In spirit, this is what the code was trying to do prior to this patch (at least I think that is what the original authors intended), but it was done rather poorly and added a layer of obfuscation that only masked the underlying problems. * Big backlog queue cleanup, again. In v4.10 we made some pretty big changes to how the audit backlog queues work, here we haven't changed the queue design so much as cleaned up the implementation. Brought about by the locking changes, we've simplified kauditd_thread() quite a bit by consolidating the queue handling into a new helper function, kauditd_send_queue(), which allows us to eliminate a lot of very similar code and makes the looping logic in kauditd_thread() clearer. * All netlink messages sent to auditd are now sent via auditd_send_unicast_skb(). Other than just making sense, this makes the lock handling easier. * Change the audit_log_start() sleep behavior so that we never sleep on auditd events (unchanged) or if the caller is holding the audit_cmd_mutex (changed). Previously we didn't sleep if the caller was auditd or if the message type fell between a certain range; the type check was a poor effort of doing what the cmd_mutex check now does. Richard Guy Briggs originally proposed not sleeping the cmd_mutex owner several years ago but his patch wasn't acceptable at the time. At least the idea lives on here. * A problem with the lost record counter has been resolved. Steve Grubb and I both happened to notice this problem and according to some quick testing by Steve, this problem goes back quite some time. It's largely a harmless problem, although it may have left some careful sysadmins quite puzzled. Cc: <stable@vger.kernel.org> # 4.10.x- Reported-by: Dmitry Vyukov <dvyukov@google.com> Signed-off-by: Paul Moore <paul@paul-moore.com>
2017-03-21 23:26:35 +08:00
/* NOTE: don't ever fail/sleep on these two conditions:
* 1. auditd generated record - since we need auditd to drain the
* queue; also, when we are checking for auditd, compare PIDs using
* task_tgid_vnr() since auditd_pid is set in audit_receive_msg()
* using a PID anchored in the caller's namespace
audit: fix auditd/kernel connection state tracking What started as a rather straightforward race condition reported by Dmitry using the syzkaller fuzzer ended up revealing some major problems with how the audit subsystem managed its netlink sockets and its connection with the userspace audit daemon. Fixing this properly had quite the cascading effect and what we are left with is this rather large and complicated patch. My initial goal was to try and decompose this patch into multiple smaller patches, but the way these changes are intertwined makes it difficult to split these changes into meaningful pieces that don't break or somehow make things worse for the intermediate states. The patch makes a number of changes, but the most significant are highlighted below: * The auditd tracking variables, e.g. audit_sock, are now gone and replaced by a RCU/spin_lock protected variable auditd_conn which is a structure containing all of the auditd tracking information. * We no longer track the auditd sock directly, instead we track it via the network namespace in which it resides and we use the audit socket associated with that namespace. In spirit, this is what the code was trying to do prior to this patch (at least I think that is what the original authors intended), but it was done rather poorly and added a layer of obfuscation that only masked the underlying problems. * Big backlog queue cleanup, again. In v4.10 we made some pretty big changes to how the audit backlog queues work, here we haven't changed the queue design so much as cleaned up the implementation. Brought about by the locking changes, we've simplified kauditd_thread() quite a bit by consolidating the queue handling into a new helper function, kauditd_send_queue(), which allows us to eliminate a lot of very similar code and makes the looping logic in kauditd_thread() clearer. * All netlink messages sent to auditd are now sent via auditd_send_unicast_skb(). Other than just making sense, this makes the lock handling easier. * Change the audit_log_start() sleep behavior so that we never sleep on auditd events (unchanged) or if the caller is holding the audit_cmd_mutex (changed). Previously we didn't sleep if the caller was auditd or if the message type fell between a certain range; the type check was a poor effort of doing what the cmd_mutex check now does. Richard Guy Briggs originally proposed not sleeping the cmd_mutex owner several years ago but his patch wasn't acceptable at the time. At least the idea lives on here. * A problem with the lost record counter has been resolved. Steve Grubb and I both happened to notice this problem and according to some quick testing by Steve, this problem goes back quite some time. It's largely a harmless problem, although it may have left some careful sysadmins quite puzzled. Cc: <stable@vger.kernel.org> # 4.10.x- Reported-by: Dmitry Vyukov <dvyukov@google.com> Signed-off-by: Paul Moore <paul@paul-moore.com>
2017-03-21 23:26:35 +08:00
* 2. generator holding the audit_cmd_mutex - we don't want to block
* while holding the mutex, although we do penalize the sender
* later in audit_receive() when it is safe to block
*/
if (!(auditd_test_task(current) || audit_ctl_owner_current())) {
audit: fix auditd/kernel connection state tracking What started as a rather straightforward race condition reported by Dmitry using the syzkaller fuzzer ended up revealing some major problems with how the audit subsystem managed its netlink sockets and its connection with the userspace audit daemon. Fixing this properly had quite the cascading effect and what we are left with is this rather large and complicated patch. My initial goal was to try and decompose this patch into multiple smaller patches, but the way these changes are intertwined makes it difficult to split these changes into meaningful pieces that don't break or somehow make things worse for the intermediate states. The patch makes a number of changes, but the most significant are highlighted below: * The auditd tracking variables, e.g. audit_sock, are now gone and replaced by a RCU/spin_lock protected variable auditd_conn which is a structure containing all of the auditd tracking information. * We no longer track the auditd sock directly, instead we track it via the network namespace in which it resides and we use the audit socket associated with that namespace. In spirit, this is what the code was trying to do prior to this patch (at least I think that is what the original authors intended), but it was done rather poorly and added a layer of obfuscation that only masked the underlying problems. * Big backlog queue cleanup, again. In v4.10 we made some pretty big changes to how the audit backlog queues work, here we haven't changed the queue design so much as cleaned up the implementation. Brought about by the locking changes, we've simplified kauditd_thread() quite a bit by consolidating the queue handling into a new helper function, kauditd_send_queue(), which allows us to eliminate a lot of very similar code and makes the looping logic in kauditd_thread() clearer. * All netlink messages sent to auditd are now sent via auditd_send_unicast_skb(). Other than just making sense, this makes the lock handling easier. * Change the audit_log_start() sleep behavior so that we never sleep on auditd events (unchanged) or if the caller is holding the audit_cmd_mutex (changed). Previously we didn't sleep if the caller was auditd or if the message type fell between a certain range; the type check was a poor effort of doing what the cmd_mutex check now does. Richard Guy Briggs originally proposed not sleeping the cmd_mutex owner several years ago but his patch wasn't acceptable at the time. At least the idea lives on here. * A problem with the lost record counter has been resolved. Steve Grubb and I both happened to notice this problem and according to some quick testing by Steve, this problem goes back quite some time. It's largely a harmless problem, although it may have left some careful sysadmins quite puzzled. Cc: <stable@vger.kernel.org> # 4.10.x- Reported-by: Dmitry Vyukov <dvyukov@google.com> Signed-off-by: Paul Moore <paul@paul-moore.com>
2017-03-21 23:26:35 +08:00
long stime = audit_backlog_wait_time;
while (audit_backlog_limit &&
(skb_queue_len(&audit_queue) > audit_backlog_limit)) {
/* wake kauditd to try and flush the queue */
wake_up_interruptible(&kauditd_wait);
/* sleep if we are allowed and we haven't exhausted our
* backlog wait limit */
audit: fix auditd/kernel connection state tracking What started as a rather straightforward race condition reported by Dmitry using the syzkaller fuzzer ended up revealing some major problems with how the audit subsystem managed its netlink sockets and its connection with the userspace audit daemon. Fixing this properly had quite the cascading effect and what we are left with is this rather large and complicated patch. My initial goal was to try and decompose this patch into multiple smaller patches, but the way these changes are intertwined makes it difficult to split these changes into meaningful pieces that don't break or somehow make things worse for the intermediate states. The patch makes a number of changes, but the most significant are highlighted below: * The auditd tracking variables, e.g. audit_sock, are now gone and replaced by a RCU/spin_lock protected variable auditd_conn which is a structure containing all of the auditd tracking information. * We no longer track the auditd sock directly, instead we track it via the network namespace in which it resides and we use the audit socket associated with that namespace. In spirit, this is what the code was trying to do prior to this patch (at least I think that is what the original authors intended), but it was done rather poorly and added a layer of obfuscation that only masked the underlying problems. * Big backlog queue cleanup, again. In v4.10 we made some pretty big changes to how the audit backlog queues work, here we haven't changed the queue design so much as cleaned up the implementation. Brought about by the locking changes, we've simplified kauditd_thread() quite a bit by consolidating the queue handling into a new helper function, kauditd_send_queue(), which allows us to eliminate a lot of very similar code and makes the looping logic in kauditd_thread() clearer. * All netlink messages sent to auditd are now sent via auditd_send_unicast_skb(). Other than just making sense, this makes the lock handling easier. * Change the audit_log_start() sleep behavior so that we never sleep on auditd events (unchanged) or if the caller is holding the audit_cmd_mutex (changed). Previously we didn't sleep if the caller was auditd or if the message type fell between a certain range; the type check was a poor effort of doing what the cmd_mutex check now does. Richard Guy Briggs originally proposed not sleeping the cmd_mutex owner several years ago but his patch wasn't acceptable at the time. At least the idea lives on here. * A problem with the lost record counter has been resolved. Steve Grubb and I both happened to notice this problem and according to some quick testing by Steve, this problem goes back quite some time. It's largely a harmless problem, although it may have left some careful sysadmins quite puzzled. Cc: <stable@vger.kernel.org> # 4.10.x- Reported-by: Dmitry Vyukov <dvyukov@google.com> Signed-off-by: Paul Moore <paul@paul-moore.com>
2017-03-21 23:26:35 +08:00
if (gfpflags_allow_blocking(gfp_mask) && (stime > 0)) {
long rtime = stime;
DECLARE_WAITQUEUE(wait, current);
add_wait_queue_exclusive(&audit_backlog_wait,
&wait);
set_current_state(TASK_UNINTERRUPTIBLE);
stime = schedule_timeout(rtime);
atomic_add(rtime - stime, &audit_backlog_wait_time_actual);
remove_wait_queue(&audit_backlog_wait, &wait);
} else {
if (audit_rate_check() && printk_ratelimit())
pr_warn("audit_backlog=%d > audit_backlog_limit=%d\n",
skb_queue_len(&audit_queue),
audit_backlog_limit);
audit_log_lost("backlog limit exceeded");
return NULL;
}
}
}
ab = audit_buffer_alloc(ctx, gfp_mask, type);
if (!ab) {
audit_log_lost("out of memory in audit_log_start");
return NULL;
}
audit_get_stamp(ab->ctx, &t, &serial);
audit: trigger accompanying records when no rules present When there are no audit rules registered, mandatory records (config, etc.) are missing their accompanying records (syscall, proctitle, etc.). This is due to audit context dummy set on syscall entry based on absence of rules that signals that no other records are to be printed. Clear the dummy bit if any record is generated, open coding this in audit_log_start(). The proctitle context and dummy checks are pointless since the proctitle record will not be printed if no syscall records are printed. The fds array is reset to -1 after the first syscall to indicate it isn't valid any more, but was never set to -1 when the context was allocated to indicate it wasn't yet valid. Check ctx->pwd in audit_log_name(). The audit_inode* functions can be called without going through getname_flags() or getname_kernel() that sets audit_names and cwd, so set the cwd in audit_alloc_name() if it has not already been done so due to audit_names being valid and purge all other audit_getcwd() calls. Revert the LSM dump_common_audit_data() LSM_AUDIT_DATA_* cases from the ghak96 patch since they are no longer necessary due to cwd coverage in audit_alloc_name(). Thanks to bauen1 <j2468h@googlemail.com> for reporting LSM situations in which context->cwd is not valid, inadvertantly fixed by the ghak96 patch. Please see upstream github issue https://github.com/linux-audit/audit-kernel/issues/120 This is also related to upstream github issue https://github.com/linux-audit/audit-kernel/issues/96 Signed-off-by: Richard Guy Briggs <rgb@redhat.com> Signed-off-by: Paul Moore <paul@paul-moore.com>
2020-09-22 20:44:50 +08:00
/* cancel dummy context to enable supporting records */
if (ctx)
ctx->dummy = 0;
audit_log_format(ab, "audit(%llu.%03lu:%u): ",
(unsigned long long)t.tv_sec, t.tv_nsec/1000000, serial);
return ab;
}
/**
* audit_expand - expand skb in the audit buffer
* @ab: audit_buffer
* @extra: space to add at tail of the skb
*
* Returns 0 (no space) on failed expansion, or available space if
* successful.
*/
static inline int audit_expand(struct audit_buffer *ab, int extra)
{
struct sk_buff *skb = ab->skb;
int oldtail = skb_tailroom(skb);
int ret = pskb_expand_head(skb, 0, extra, ab->gfp_mask);
int newtail = skb_tailroom(skb);
if (ret < 0) {
audit_log_lost("out of memory in audit_expand");
return 0;
}
skb->truesize += newtail - oldtail;
return newtail;
}
/*
* Format an audit message into the audit buffer. If there isn't enough
* room in the audit buffer, more room will be allocated and vsnprint
* will be called a second time. Currently, we assume that a printk
* can't format message larger than 1024 bytes, so we don't either.
*/
static void audit_log_vformat(struct audit_buffer *ab, const char *fmt,
va_list args)
{
int len, avail;
struct sk_buff *skb;
va_list args2;
if (!ab)
return;
BUG_ON(!ab->skb);
skb = ab->skb;
avail = skb_tailroom(skb);
if (avail == 0) {
avail = audit_expand(ab, AUDIT_BUFSIZ);
if (!avail)
goto out;
}
va_copy(args2, args);
len = vsnprintf(skb_tail_pointer(skb), avail, fmt, args);
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. */
avail = audit_expand(ab,
max_t(unsigned, AUDIT_BUFSIZ, 1+len-avail));
if (!avail)
goto out_va_end;
len = vsnprintf(skb_tail_pointer(skb), avail, fmt, args2);
}
if (len > 0)
skb_put(skb, len);
out_va_end:
va_end(args2);
out:
return;
}
/**
* 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.
*/
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);
}
/**
* audit_log_n_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.
*/
void audit_log_n_hex(struct audit_buffer *ab, const unsigned char *buf,
size_t len)
{
int i, avail, new_len;
unsigned char *ptr;
struct sk_buff *skb;
if (!ab)
return;
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;
}
ptr = skb_tail_pointer(skb);
for (i = 0; i < len; i++)
ptr = hex_byte_pack_upper(ptr, buf[i]);
*ptr = 0;
skb_put(skb, len << 1); /* new string is twice the old string */
}
[PATCH] log more info for directory entry change events When an audit event involves changes to a directory entry, include a PATH record for the directory itself. A few other notable changes: - fixed audit_inode_child() hooks in fsnotify_move() - removed unused flags arg from audit_inode() - added audit log routines for logging a portion of a string Here's some sample output. before patch: type=SYSCALL msg=audit(1149821605.320:26): arch=40000003 syscall=39 success=yes exit=0 a0=bf8d3c7c a1=1ff a2=804e1b8 a3=bf8d3c7c items=1 ppid=739 pid=800 auid=0 uid=0 gid=0 euid=0 suid=0 fsuid=0 egid=0 sgid=0 fsgid=0 tty=ttyS0 comm="mkdir" exe="/bin/mkdir" subj=root:system_r:unconfined_t:s0-s0:c0.c255 type=CWD msg=audit(1149821605.320:26): cwd="/root" type=PATH msg=audit(1149821605.320:26): item=0 name="foo" parent=164068 inode=164010 dev=03:00 mode=040755 ouid=0 ogid=0 rdev=00:00 obj=root:object_r:user_home_t:s0 after patch: type=SYSCALL msg=audit(1149822032.332:24): arch=40000003 syscall=39 success=yes exit=0 a0=bfdd9c7c a1=1ff a2=804e1b8 a3=bfdd9c7c items=2 ppid=714 pid=777 auid=0 uid=0 gid=0 euid=0 suid=0 fsuid=0 egid=0 sgid=0 fsgid=0 tty=ttyS0 comm="mkdir" exe="/bin/mkdir" subj=root:system_r:unconfined_t:s0-s0:c0.c255 type=CWD msg=audit(1149822032.332:24): cwd="/root" type=PATH msg=audit(1149822032.332:24): item=0 name="/root" inode=164068 dev=03:00 mode=040750 ouid=0 ogid=0 rdev=00:00 obj=root:object_r:user_home_dir_t:s0 type=PATH msg=audit(1149822032.332:24): item=1 name="foo" inode=164010 dev=03:00 mode=040755 ouid=0 ogid=0 rdev=00:00 obj=root:object_r:user_home_t:s0 Signed-off-by: Amy Griffis <amy.griffis@hp.com> Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
2006-06-09 11:19:31 +08:00
/*
* Format a string of no more than slen characters into the audit buffer,
* enclosed in quote marks.
*/
void audit_log_n_string(struct audit_buffer *ab, const char *string,
size_t slen)
[PATCH] log more info for directory entry change events When an audit event involves changes to a directory entry, include a PATH record for the directory itself. A few other notable changes: - fixed audit_inode_child() hooks in fsnotify_move() - removed unused flags arg from audit_inode() - added audit log routines for logging a portion of a string Here's some sample output. before patch: type=SYSCALL msg=audit(1149821605.320:26): arch=40000003 syscall=39 success=yes exit=0 a0=bf8d3c7c a1=1ff a2=804e1b8 a3=bf8d3c7c items=1 ppid=739 pid=800 auid=0 uid=0 gid=0 euid=0 suid=0 fsuid=0 egid=0 sgid=0 fsgid=0 tty=ttyS0 comm="mkdir" exe="/bin/mkdir" subj=root:system_r:unconfined_t:s0-s0:c0.c255 type=CWD msg=audit(1149821605.320:26): cwd="/root" type=PATH msg=audit(1149821605.320:26): item=0 name="foo" parent=164068 inode=164010 dev=03:00 mode=040755 ouid=0 ogid=0 rdev=00:00 obj=root:object_r:user_home_t:s0 after patch: type=SYSCALL msg=audit(1149822032.332:24): arch=40000003 syscall=39 success=yes exit=0 a0=bfdd9c7c a1=1ff a2=804e1b8 a3=bfdd9c7c items=2 ppid=714 pid=777 auid=0 uid=0 gid=0 euid=0 suid=0 fsuid=0 egid=0 sgid=0 fsgid=0 tty=ttyS0 comm="mkdir" exe="/bin/mkdir" subj=root:system_r:unconfined_t:s0-s0:c0.c255 type=CWD msg=audit(1149822032.332:24): cwd="/root" type=PATH msg=audit(1149822032.332:24): item=0 name="/root" inode=164068 dev=03:00 mode=040750 ouid=0 ogid=0 rdev=00:00 obj=root:object_r:user_home_dir_t:s0 type=PATH msg=audit(1149822032.332:24): item=1 name="foo" inode=164010 dev=03:00 mode=040755 ouid=0 ogid=0 rdev=00:00 obj=root:object_r:user_home_t:s0 Signed-off-by: Amy Griffis <amy.griffis@hp.com> Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
2006-06-09 11:19:31 +08:00
{
int avail, new_len;
unsigned char *ptr;
struct sk_buff *skb;
if (!ab)
return;
[PATCH] log more info for directory entry change events When an audit event involves changes to a directory entry, include a PATH record for the directory itself. A few other notable changes: - fixed audit_inode_child() hooks in fsnotify_move() - removed unused flags arg from audit_inode() - added audit log routines for logging a portion of a string Here's some sample output. before patch: type=SYSCALL msg=audit(1149821605.320:26): arch=40000003 syscall=39 success=yes exit=0 a0=bf8d3c7c a1=1ff a2=804e1b8 a3=bf8d3c7c items=1 ppid=739 pid=800 auid=0 uid=0 gid=0 euid=0 suid=0 fsuid=0 egid=0 sgid=0 fsgid=0 tty=ttyS0 comm="mkdir" exe="/bin/mkdir" subj=root:system_r:unconfined_t:s0-s0:c0.c255 type=CWD msg=audit(1149821605.320:26): cwd="/root" type=PATH msg=audit(1149821605.320:26): item=0 name="foo" parent=164068 inode=164010 dev=03:00 mode=040755 ouid=0 ogid=0 rdev=00:00 obj=root:object_r:user_home_t:s0 after patch: type=SYSCALL msg=audit(1149822032.332:24): arch=40000003 syscall=39 success=yes exit=0 a0=bfdd9c7c a1=1ff a2=804e1b8 a3=bfdd9c7c items=2 ppid=714 pid=777 auid=0 uid=0 gid=0 euid=0 suid=0 fsuid=0 egid=0 sgid=0 fsgid=0 tty=ttyS0 comm="mkdir" exe="/bin/mkdir" subj=root:system_r:unconfined_t:s0-s0:c0.c255 type=CWD msg=audit(1149822032.332:24): cwd="/root" type=PATH msg=audit(1149822032.332:24): item=0 name="/root" inode=164068 dev=03:00 mode=040750 ouid=0 ogid=0 rdev=00:00 obj=root:object_r:user_home_dir_t:s0 type=PATH msg=audit(1149822032.332:24): item=1 name="foo" inode=164010 dev=03:00 mode=040755 ouid=0 ogid=0 rdev=00:00 obj=root:object_r:user_home_t:s0 Signed-off-by: Amy Griffis <amy.griffis@hp.com> Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
2006-06-09 11:19:31 +08:00
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;
}
ptr = skb_tail_pointer(skb);
[PATCH] log more info for directory entry change events When an audit event involves changes to a directory entry, include a PATH record for the directory itself. A few other notable changes: - fixed audit_inode_child() hooks in fsnotify_move() - removed unused flags arg from audit_inode() - added audit log routines for logging a portion of a string Here's some sample output. before patch: type=SYSCALL msg=audit(1149821605.320:26): arch=40000003 syscall=39 success=yes exit=0 a0=bf8d3c7c a1=1ff a2=804e1b8 a3=bf8d3c7c items=1 ppid=739 pid=800 auid=0 uid=0 gid=0 euid=0 suid=0 fsuid=0 egid=0 sgid=0 fsgid=0 tty=ttyS0 comm="mkdir" exe="/bin/mkdir" subj=root:system_r:unconfined_t:s0-s0:c0.c255 type=CWD msg=audit(1149821605.320:26): cwd="/root" type=PATH msg=audit(1149821605.320:26): item=0 name="foo" parent=164068 inode=164010 dev=03:00 mode=040755 ouid=0 ogid=0 rdev=00:00 obj=root:object_r:user_home_t:s0 after patch: type=SYSCALL msg=audit(1149822032.332:24): arch=40000003 syscall=39 success=yes exit=0 a0=bfdd9c7c a1=1ff a2=804e1b8 a3=bfdd9c7c items=2 ppid=714 pid=777 auid=0 uid=0 gid=0 euid=0 suid=0 fsuid=0 egid=0 sgid=0 fsgid=0 tty=ttyS0 comm="mkdir" exe="/bin/mkdir" subj=root:system_r:unconfined_t:s0-s0:c0.c255 type=CWD msg=audit(1149822032.332:24): cwd="/root" type=PATH msg=audit(1149822032.332:24): item=0 name="/root" inode=164068 dev=03:00 mode=040750 ouid=0 ogid=0 rdev=00:00 obj=root:object_r:user_home_dir_t:s0 type=PATH msg=audit(1149822032.332:24): item=1 name="foo" inode=164010 dev=03:00 mode=040755 ouid=0 ogid=0 rdev=00:00 obj=root:object_r:user_home_t:s0 Signed-off-by: Amy Griffis <amy.griffis@hp.com> Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
2006-06-09 11:19:31 +08:00
*ptr++ = '"';
memcpy(ptr, string, slen);
ptr += slen;
*ptr++ = '"';
*ptr = 0;
skb_put(skb, slen + 2); /* don't include null terminator */
}
/**
* audit_string_contains_control - does a string need to be logged in hex
* @string: string to be checked
* @len: max length of the string to check
*/
bool audit_string_contains_control(const char *string, size_t len)
{
const unsigned char *p;
for (p = string; p < (const unsigned char *)string + len; p++) {
if (*p == '"' || *p < 0x21 || *p > 0x7e)
return true;
}
return false;
}
/**
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 14:40:56 +08:00
* audit_log_n_untrustedstring - log a string that may contain random characters
* @ab: audit_buffer
* @len: length of string (not including trailing null)
* @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,
* or a space. Unescaped strings will start and end with a double quote mark.
* Strings that are escaped are printed in hex (2 digits per char).
[PATCH] log more info for directory entry change events When an audit event involves changes to a directory entry, include a PATH record for the directory itself. A few other notable changes: - fixed audit_inode_child() hooks in fsnotify_move() - removed unused flags arg from audit_inode() - added audit log routines for logging a portion of a string Here's some sample output. before patch: type=SYSCALL msg=audit(1149821605.320:26): arch=40000003 syscall=39 success=yes exit=0 a0=bf8d3c7c a1=1ff a2=804e1b8 a3=bf8d3c7c items=1 ppid=739 pid=800 auid=0 uid=0 gid=0 euid=0 suid=0 fsuid=0 egid=0 sgid=0 fsgid=0 tty=ttyS0 comm="mkdir" exe="/bin/mkdir" subj=root:system_r:unconfined_t:s0-s0:c0.c255 type=CWD msg=audit(1149821605.320:26): cwd="/root" type=PATH msg=audit(1149821605.320:26): item=0 name="foo" parent=164068 inode=164010 dev=03:00 mode=040755 ouid=0 ogid=0 rdev=00:00 obj=root:object_r:user_home_t:s0 after patch: type=SYSCALL msg=audit(1149822032.332:24): arch=40000003 syscall=39 success=yes exit=0 a0=bfdd9c7c a1=1ff a2=804e1b8 a3=bfdd9c7c items=2 ppid=714 pid=777 auid=0 uid=0 gid=0 euid=0 suid=0 fsuid=0 egid=0 sgid=0 fsgid=0 tty=ttyS0 comm="mkdir" exe="/bin/mkdir" subj=root:system_r:unconfined_t:s0-s0:c0.c255 type=CWD msg=audit(1149822032.332:24): cwd="/root" type=PATH msg=audit(1149822032.332:24): item=0 name="/root" inode=164068 dev=03:00 mode=040750 ouid=0 ogid=0 rdev=00:00 obj=root:object_r:user_home_dir_t:s0 type=PATH msg=audit(1149822032.332:24): item=1 name="foo" inode=164010 dev=03:00 mode=040755 ouid=0 ogid=0 rdev=00:00 obj=root:object_r:user_home_t:s0 Signed-off-by: Amy Griffis <amy.griffis@hp.com> Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
2006-06-09 11:19:31 +08:00
*
* The caller specifies the number of characters in the string to log, which may
* or may not be the entire string.
*/
void audit_log_n_untrustedstring(struct audit_buffer *ab, const char *string,
size_t len)
{
if (audit_string_contains_control(string, len))
audit_log_n_hex(ab, string, len);
else
audit_log_n_string(ab, string, len);
}
[PATCH] log more info for directory entry change events When an audit event involves changes to a directory entry, include a PATH record for the directory itself. A few other notable changes: - fixed audit_inode_child() hooks in fsnotify_move() - removed unused flags arg from audit_inode() - added audit log routines for logging a portion of a string Here's some sample output. before patch: type=SYSCALL msg=audit(1149821605.320:26): arch=40000003 syscall=39 success=yes exit=0 a0=bf8d3c7c a1=1ff a2=804e1b8 a3=bf8d3c7c items=1 ppid=739 pid=800 auid=0 uid=0 gid=0 euid=0 suid=0 fsuid=0 egid=0 sgid=0 fsgid=0 tty=ttyS0 comm="mkdir" exe="/bin/mkdir" subj=root:system_r:unconfined_t:s0-s0:c0.c255 type=CWD msg=audit(1149821605.320:26): cwd="/root" type=PATH msg=audit(1149821605.320:26): item=0 name="foo" parent=164068 inode=164010 dev=03:00 mode=040755 ouid=0 ogid=0 rdev=00:00 obj=root:object_r:user_home_t:s0 after patch: type=SYSCALL msg=audit(1149822032.332:24): arch=40000003 syscall=39 success=yes exit=0 a0=bfdd9c7c a1=1ff a2=804e1b8 a3=bfdd9c7c items=2 ppid=714 pid=777 auid=0 uid=0 gid=0 euid=0 suid=0 fsuid=0 egid=0 sgid=0 fsgid=0 tty=ttyS0 comm="mkdir" exe="/bin/mkdir" subj=root:system_r:unconfined_t:s0-s0:c0.c255 type=CWD msg=audit(1149822032.332:24): cwd="/root" type=PATH msg=audit(1149822032.332:24): item=0 name="/root" inode=164068 dev=03:00 mode=040750 ouid=0 ogid=0 rdev=00:00 obj=root:object_r:user_home_dir_t:s0 type=PATH msg=audit(1149822032.332:24): item=1 name="foo" inode=164010 dev=03:00 mode=040755 ouid=0 ogid=0 rdev=00:00 obj=root:object_r:user_home_t:s0 Signed-off-by: Amy Griffis <amy.griffis@hp.com> Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
2006-06-09 11:19:31 +08:00
/**
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 14:40:56 +08:00
* audit_log_untrustedstring - log a string that may contain random characters
[PATCH] log more info for directory entry change events When an audit event involves changes to a directory entry, include a PATH record for the directory itself. A few other notable changes: - fixed audit_inode_child() hooks in fsnotify_move() - removed unused flags arg from audit_inode() - added audit log routines for logging a portion of a string Here's some sample output. before patch: type=SYSCALL msg=audit(1149821605.320:26): arch=40000003 syscall=39 success=yes exit=0 a0=bf8d3c7c a1=1ff a2=804e1b8 a3=bf8d3c7c items=1 ppid=739 pid=800 auid=0 uid=0 gid=0 euid=0 suid=0 fsuid=0 egid=0 sgid=0 fsgid=0 tty=ttyS0 comm="mkdir" exe="/bin/mkdir" subj=root:system_r:unconfined_t:s0-s0:c0.c255 type=CWD msg=audit(1149821605.320:26): cwd="/root" type=PATH msg=audit(1149821605.320:26): item=0 name="foo" parent=164068 inode=164010 dev=03:00 mode=040755 ouid=0 ogid=0 rdev=00:00 obj=root:object_r:user_home_t:s0 after patch: type=SYSCALL msg=audit(1149822032.332:24): arch=40000003 syscall=39 success=yes exit=0 a0=bfdd9c7c a1=1ff a2=804e1b8 a3=bfdd9c7c items=2 ppid=714 pid=777 auid=0 uid=0 gid=0 euid=0 suid=0 fsuid=0 egid=0 sgid=0 fsgid=0 tty=ttyS0 comm="mkdir" exe="/bin/mkdir" subj=root:system_r:unconfined_t:s0-s0:c0.c255 type=CWD msg=audit(1149822032.332:24): cwd="/root" type=PATH msg=audit(1149822032.332:24): item=0 name="/root" inode=164068 dev=03:00 mode=040750 ouid=0 ogid=0 rdev=00:00 obj=root:object_r:user_home_dir_t:s0 type=PATH msg=audit(1149822032.332:24): item=1 name="foo" inode=164010 dev=03:00 mode=040755 ouid=0 ogid=0 rdev=00:00 obj=root:object_r:user_home_t:s0 Signed-off-by: Amy Griffis <amy.griffis@hp.com> Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
2006-06-09 11:19:31 +08:00
* @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 14:40:56 +08:00
* Same as audit_log_n_untrustedstring(), except that strlen is used to
[PATCH] log more info for directory entry change events When an audit event involves changes to a directory entry, include a PATH record for the directory itself. A few other notable changes: - fixed audit_inode_child() hooks in fsnotify_move() - removed unused flags arg from audit_inode() - added audit log routines for logging a portion of a string Here's some sample output. before patch: type=SYSCALL msg=audit(1149821605.320:26): arch=40000003 syscall=39 success=yes exit=0 a0=bf8d3c7c a1=1ff a2=804e1b8 a3=bf8d3c7c items=1 ppid=739 pid=800 auid=0 uid=0 gid=0 euid=0 suid=0 fsuid=0 egid=0 sgid=0 fsgid=0 tty=ttyS0 comm="mkdir" exe="/bin/mkdir" subj=root:system_r:unconfined_t:s0-s0:c0.c255 type=CWD msg=audit(1149821605.320:26): cwd="/root" type=PATH msg=audit(1149821605.320:26): item=0 name="foo" parent=164068 inode=164010 dev=03:00 mode=040755 ouid=0 ogid=0 rdev=00:00 obj=root:object_r:user_home_t:s0 after patch: type=SYSCALL msg=audit(1149822032.332:24): arch=40000003 syscall=39 success=yes exit=0 a0=bfdd9c7c a1=1ff a2=804e1b8 a3=bfdd9c7c items=2 ppid=714 pid=777 auid=0 uid=0 gid=0 euid=0 suid=0 fsuid=0 egid=0 sgid=0 fsgid=0 tty=ttyS0 comm="mkdir" exe="/bin/mkdir" subj=root:system_r:unconfined_t:s0-s0:c0.c255 type=CWD msg=audit(1149822032.332:24): cwd="/root" type=PATH msg=audit(1149822032.332:24): item=0 name="/root" inode=164068 dev=03:00 mode=040750 ouid=0 ogid=0 rdev=00:00 obj=root:object_r:user_home_dir_t:s0 type=PATH msg=audit(1149822032.332:24): item=1 name="foo" inode=164010 dev=03:00 mode=040755 ouid=0 ogid=0 rdev=00:00 obj=root:object_r:user_home_t:s0 Signed-off-by: Amy Griffis <amy.griffis@hp.com> Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
2006-06-09 11:19:31 +08:00
* determine string length.
*/
void audit_log_untrustedstring(struct audit_buffer *ab, const char *string)
[PATCH] log more info for directory entry change events When an audit event involves changes to a directory entry, include a PATH record for the directory itself. A few other notable changes: - fixed audit_inode_child() hooks in fsnotify_move() - removed unused flags arg from audit_inode() - added audit log routines for logging a portion of a string Here's some sample output. before patch: type=SYSCALL msg=audit(1149821605.320:26): arch=40000003 syscall=39 success=yes exit=0 a0=bf8d3c7c a1=1ff a2=804e1b8 a3=bf8d3c7c items=1 ppid=739 pid=800 auid=0 uid=0 gid=0 euid=0 suid=0 fsuid=0 egid=0 sgid=0 fsgid=0 tty=ttyS0 comm="mkdir" exe="/bin/mkdir" subj=root:system_r:unconfined_t:s0-s0:c0.c255 type=CWD msg=audit(1149821605.320:26): cwd="/root" type=PATH msg=audit(1149821605.320:26): item=0 name="foo" parent=164068 inode=164010 dev=03:00 mode=040755 ouid=0 ogid=0 rdev=00:00 obj=root:object_r:user_home_t:s0 after patch: type=SYSCALL msg=audit(1149822032.332:24): arch=40000003 syscall=39 success=yes exit=0 a0=bfdd9c7c a1=1ff a2=804e1b8 a3=bfdd9c7c items=2 ppid=714 pid=777 auid=0 uid=0 gid=0 euid=0 suid=0 fsuid=0 egid=0 sgid=0 fsgid=0 tty=ttyS0 comm="mkdir" exe="/bin/mkdir" subj=root:system_r:unconfined_t:s0-s0:c0.c255 type=CWD msg=audit(1149822032.332:24): cwd="/root" type=PATH msg=audit(1149822032.332:24): item=0 name="/root" inode=164068 dev=03:00 mode=040750 ouid=0 ogid=0 rdev=00:00 obj=root:object_r:user_home_dir_t:s0 type=PATH msg=audit(1149822032.332:24): item=1 name="foo" inode=164010 dev=03:00 mode=040755 ouid=0 ogid=0 rdev=00:00 obj=root:object_r:user_home_t:s0 Signed-off-by: Amy Griffis <amy.griffis@hp.com> Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
2006-06-09 11:19:31 +08:00
{
audit_log_n_untrustedstring(ab, string, strlen(string));
[PATCH] log more info for directory entry change events When an audit event involves changes to a directory entry, include a PATH record for the directory itself. A few other notable changes: - fixed audit_inode_child() hooks in fsnotify_move() - removed unused flags arg from audit_inode() - added audit log routines for logging a portion of a string Here's some sample output. before patch: type=SYSCALL msg=audit(1149821605.320:26): arch=40000003 syscall=39 success=yes exit=0 a0=bf8d3c7c a1=1ff a2=804e1b8 a3=bf8d3c7c items=1 ppid=739 pid=800 auid=0 uid=0 gid=0 euid=0 suid=0 fsuid=0 egid=0 sgid=0 fsgid=0 tty=ttyS0 comm="mkdir" exe="/bin/mkdir" subj=root:system_r:unconfined_t:s0-s0:c0.c255 type=CWD msg=audit(1149821605.320:26): cwd="/root" type=PATH msg=audit(1149821605.320:26): item=0 name="foo" parent=164068 inode=164010 dev=03:00 mode=040755 ouid=0 ogid=0 rdev=00:00 obj=root:object_r:user_home_t:s0 after patch: type=SYSCALL msg=audit(1149822032.332:24): arch=40000003 syscall=39 success=yes exit=0 a0=bfdd9c7c a1=1ff a2=804e1b8 a3=bfdd9c7c items=2 ppid=714 pid=777 auid=0 uid=0 gid=0 euid=0 suid=0 fsuid=0 egid=0 sgid=0 fsgid=0 tty=ttyS0 comm="mkdir" exe="/bin/mkdir" subj=root:system_r:unconfined_t:s0-s0:c0.c255 type=CWD msg=audit(1149822032.332:24): cwd="/root" type=PATH msg=audit(1149822032.332:24): item=0 name="/root" inode=164068 dev=03:00 mode=040750 ouid=0 ogid=0 rdev=00:00 obj=root:object_r:user_home_dir_t:s0 type=PATH msg=audit(1149822032.332:24): item=1 name="foo" inode=164010 dev=03:00 mode=040755 ouid=0 ogid=0 rdev=00:00 obj=root:object_r:user_home_t:s0 Signed-off-by: Amy Griffis <amy.griffis@hp.com> Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
2006-06-09 11:19:31 +08:00
}
/* This is a helper-function to print the escaped d_path */
void audit_log_d_path(struct audit_buffer *ab, const char *prefix,
const struct path *path)
{
char *p, *pathname;
if (prefix)
audit_log_format(ab, "%s", prefix);
/* We will allow 11 spaces for ' (deleted)' to be appended */
pathname = kmalloc(PATH_MAX+11, ab->gfp_mask);
if (!pathname) {
audit_log_format(ab, "\"<no_memory>\"");
return;
}
p = d_path(path, pathname, PATH_MAX+11);
if (IS_ERR(p)) { /* Should never happen since we send PATH_MAX */
/* FIXME: can we save some information here? */
audit_log_format(ab, "\"<too_long>\"");
} else
audit_log_untrustedstring(ab, p);
kfree(pathname);
}
2013-04-30 21:53:34 +08:00
void audit_log_session_info(struct audit_buffer *ab)
{
unsigned int sessionid = audit_get_sessionid(current);
2013-04-30 21:53:34 +08:00
uid_t auid = from_kuid(&init_user_ns, audit_get_loginuid(current));
audit_log_format(ab, "auid=%u ses=%u", auid, sessionid);
2013-04-30 21:53:34 +08:00
}
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)");
}
int audit_log_task_context(struct audit_buffer *ab)
{
char *ctx = NULL;
unsigned len;
int error;
u32 sid;
security_current_getsecid_subj(&sid);
if (!sid)
return 0;
error = security_secid_to_secctx(sid, &ctx, &len);
if (error) {
if (error != -EINVAL)
goto error_path;
return 0;
}
audit_log_format(ab, " subj=%s", ctx);
security_release_secctx(ctx, len);
return 0;
error_path:
audit_panic("error in audit_log_task_context");
return error;
}
EXPORT_SYMBOL(audit_log_task_context);
void audit_log_d_path_exe(struct audit_buffer *ab,
struct mm_struct *mm)
{
struct file *exe_file;
if (!mm)
goto out_null;
exe_file = get_mm_exe_file(mm);
if (!exe_file)
goto out_null;
audit_log_d_path(ab, " exe=", &exe_file->f_path);
fput(exe_file);
return;
out_null:
audit_log_format(ab, " exe=(null)");
}
struct tty_struct *audit_get_tty(void)
{
struct tty_struct *tty = NULL;
unsigned long flags;
spin_lock_irqsave(&current->sighand->siglock, flags);
if (current->signal)
tty = tty_kref_get(current->signal->tty);
spin_unlock_irqrestore(&current->sighand->siglock, flags);
return tty;
}
void audit_put_tty(struct tty_struct *tty)
{
tty_kref_put(tty);
}
void audit_log_task_info(struct audit_buffer *ab)
{
const struct cred *cred;
char comm[sizeof(current->comm)];
struct tty_struct *tty;
if (!ab)
return;
cred = current_cred();
tty = audit_get_tty();
audit_log_format(ab,
" ppid=%d pid=%d auid=%u uid=%u gid=%u"
" euid=%u suid=%u fsuid=%u"
" egid=%u sgid=%u fsgid=%u tty=%s ses=%u",
task_ppid_nr(current),
task_tgid_nr(current),
from_kuid(&init_user_ns, audit_get_loginuid(current)),
from_kuid(&init_user_ns, cred->uid),
from_kgid(&init_user_ns, cred->gid),
from_kuid(&init_user_ns, cred->euid),
from_kuid(&init_user_ns, cred->suid),
from_kuid(&init_user_ns, cred->fsuid),
from_kgid(&init_user_ns, cred->egid),
from_kgid(&init_user_ns, cred->sgid),
from_kgid(&init_user_ns, cred->fsgid),
tty ? tty_name(tty) : "(none)",
audit_get_sessionid(current));
audit_put_tty(tty);
audit_log_format(ab, " comm=");
audit_log_untrustedstring(ab, get_task_comm(comm, current));
audit_log_d_path_exe(ab, current->mm);
audit_log_task_context(ab);
}
EXPORT_SYMBOL(audit_log_task_info);
/**
* audit_log_path_denied - report a path restriction denial
* @type: audit message type (AUDIT_ANOM_LINK, AUDIT_ANOM_CREAT, etc)
* @operation: specific operation name
*/
void audit_log_path_denied(int type, const char *operation)
{
struct audit_buffer *ab;
if (!audit_enabled || audit_dummy_context())
return;
/* Generate log with subject, operation, outcome. */
ab = audit_log_start(audit_context(), GFP_KERNEL, type);
if (!ab)
return;
audit_log_format(ab, "op=%s", operation);
audit_log_task_info(ab);
audit_log_format(ab, " res=0");
audit_log_end(ab);
}
/* global counter which is incremented every time something logs in */
static atomic_t session_id = ATOMIC_INIT(0);
static int audit_set_loginuid_perm(kuid_t loginuid)
{
/* if we are unset, we don't need privs */
if (!audit_loginuid_set(current))
return 0;
/* if AUDIT_FEATURE_LOGINUID_IMMUTABLE means never ever allow a change*/
if (is_audit_feature_set(AUDIT_FEATURE_LOGINUID_IMMUTABLE))
return -EPERM;
/* it is set, you need permission */
if (!capable(CAP_AUDIT_CONTROL))
return -EPERM;
/* reject if this is not an unset and we don't allow that */
if (is_audit_feature_set(AUDIT_FEATURE_ONLY_UNSET_LOGINUID)
&& uid_valid(loginuid))
return -EPERM;
return 0;
}
static void audit_log_set_loginuid(kuid_t koldloginuid, kuid_t kloginuid,
unsigned int oldsessionid,
unsigned int sessionid, int rc)
{
struct audit_buffer *ab;
uid_t uid, oldloginuid, loginuid;
struct tty_struct *tty;
if (!audit_enabled)
return;
ab = audit_log_start(audit_context(), GFP_KERNEL, AUDIT_LOGIN);
if (!ab)
return;
uid = from_kuid(&init_user_ns, task_uid(current));
oldloginuid = from_kuid(&init_user_ns, koldloginuid);
loginuid = from_kuid(&init_user_ns, kloginuid);
tty = audit_get_tty();
audit_log_format(ab, "pid=%d uid=%u", task_tgid_nr(current), uid);
audit_log_task_context(ab);
audit_log_format(ab, " old-auid=%u auid=%u tty=%s old-ses=%u ses=%u res=%d",
oldloginuid, loginuid, tty ? tty_name(tty) : "(none)",
oldsessionid, sessionid, !rc);
audit_put_tty(tty);
audit_log_end(ab);
}
/**
* audit_set_loginuid - set current task's loginuid
* @loginuid: loginuid value
*
* Returns 0.
*
* Called (set) from fs/proc/base.c::proc_loginuid_write().
*/
int audit_set_loginuid(kuid_t loginuid)
{
unsigned int oldsessionid, sessionid = AUDIT_SID_UNSET;
kuid_t oldloginuid;
int rc;
oldloginuid = audit_get_loginuid(current);
oldsessionid = audit_get_sessionid(current);
rc = audit_set_loginuid_perm(loginuid);
if (rc)
goto out;
/* are we setting or clearing? */
if (uid_valid(loginuid)) {
sessionid = (unsigned int)atomic_inc_return(&session_id);
if (unlikely(sessionid == AUDIT_SID_UNSET))
sessionid = (unsigned int)atomic_inc_return(&session_id);
}
current->sessionid = sessionid;
current->loginuid = loginuid;
out:
audit_log_set_loginuid(oldloginuid, loginuid, oldsessionid, sessionid, rc);
return rc;
}
/**
* audit_signal_info - record signal info for shutting down audit subsystem
* @sig: signal value
* @t: task being signaled
*
* If the audit subsystem is being terminated, record the task (pid)
* and uid that is doing that.
*/
int audit_signal_info(int sig, struct task_struct *t)
{
kuid_t uid = current_uid(), auid;
if (auditd_test_task(t) &&
(sig == SIGTERM || sig == SIGHUP ||
sig == SIGUSR1 || sig == SIGUSR2)) {
audit_sig_pid = task_tgid_nr(current);
auid = audit_get_loginuid(current);
if (uid_valid(auid))
audit_sig_uid = auid;
else
audit_sig_uid = uid;
security_current_getsecid_subj(&audit_sig_sid);
}
return audit_signal_info_syscall(t);
}
/**
* audit_log_end - end one audit record
* @ab: the audit_buffer
*
* We can not do a netlink send inside an irq context because it blocks (last
* arg, flags, is not set to MSG_DONTWAIT), so the audit buffer is placed on a
* queue and a kthread is scheduled to remove them from the queue outside the
* irq context. May be called in any context.
*/
void audit_log_end(struct audit_buffer *ab)
{
audit: fix auditd/kernel connection state tracking What started as a rather straightforward race condition reported by Dmitry using the syzkaller fuzzer ended up revealing some major problems with how the audit subsystem managed its netlink sockets and its connection with the userspace audit daemon. Fixing this properly had quite the cascading effect and what we are left with is this rather large and complicated patch. My initial goal was to try and decompose this patch into multiple smaller patches, but the way these changes are intertwined makes it difficult to split these changes into meaningful pieces that don't break or somehow make things worse for the intermediate states. The patch makes a number of changes, but the most significant are highlighted below: * The auditd tracking variables, e.g. audit_sock, are now gone and replaced by a RCU/spin_lock protected variable auditd_conn which is a structure containing all of the auditd tracking information. * We no longer track the auditd sock directly, instead we track it via the network namespace in which it resides and we use the audit socket associated with that namespace. In spirit, this is what the code was trying to do prior to this patch (at least I think that is what the original authors intended), but it was done rather poorly and added a layer of obfuscation that only masked the underlying problems. * Big backlog queue cleanup, again. In v4.10 we made some pretty big changes to how the audit backlog queues work, here we haven't changed the queue design so much as cleaned up the implementation. Brought about by the locking changes, we've simplified kauditd_thread() quite a bit by consolidating the queue handling into a new helper function, kauditd_send_queue(), which allows us to eliminate a lot of very similar code and makes the looping logic in kauditd_thread() clearer. * All netlink messages sent to auditd are now sent via auditd_send_unicast_skb(). Other than just making sense, this makes the lock handling easier. * Change the audit_log_start() sleep behavior so that we never sleep on auditd events (unchanged) or if the caller is holding the audit_cmd_mutex (changed). Previously we didn't sleep if the caller was auditd or if the message type fell between a certain range; the type check was a poor effort of doing what the cmd_mutex check now does. Richard Guy Briggs originally proposed not sleeping the cmd_mutex owner several years ago but his patch wasn't acceptable at the time. At least the idea lives on here. * A problem with the lost record counter has been resolved. Steve Grubb and I both happened to notice this problem and according to some quick testing by Steve, this problem goes back quite some time. It's largely a harmless problem, although it may have left some careful sysadmins quite puzzled. Cc: <stable@vger.kernel.org> # 4.10.x- Reported-by: Dmitry Vyukov <dvyukov@google.com> Signed-off-by: Paul Moore <paul@paul-moore.com>
2017-03-21 23:26:35 +08:00
struct sk_buff *skb;
struct nlmsghdr *nlh;
if (!ab)
return;
audit: fix auditd/kernel connection state tracking What started as a rather straightforward race condition reported by Dmitry using the syzkaller fuzzer ended up revealing some major problems with how the audit subsystem managed its netlink sockets and its connection with the userspace audit daemon. Fixing this properly had quite the cascading effect and what we are left with is this rather large and complicated patch. My initial goal was to try and decompose this patch into multiple smaller patches, but the way these changes are intertwined makes it difficult to split these changes into meaningful pieces that don't break or somehow make things worse for the intermediate states. The patch makes a number of changes, but the most significant are highlighted below: * The auditd tracking variables, e.g. audit_sock, are now gone and replaced by a RCU/spin_lock protected variable auditd_conn which is a structure containing all of the auditd tracking information. * We no longer track the auditd sock directly, instead we track it via the network namespace in which it resides and we use the audit socket associated with that namespace. In spirit, this is what the code was trying to do prior to this patch (at least I think that is what the original authors intended), but it was done rather poorly and added a layer of obfuscation that only masked the underlying problems. * Big backlog queue cleanup, again. In v4.10 we made some pretty big changes to how the audit backlog queues work, here we haven't changed the queue design so much as cleaned up the implementation. Brought about by the locking changes, we've simplified kauditd_thread() quite a bit by consolidating the queue handling into a new helper function, kauditd_send_queue(), which allows us to eliminate a lot of very similar code and makes the looping logic in kauditd_thread() clearer. * All netlink messages sent to auditd are now sent via auditd_send_unicast_skb(). Other than just making sense, this makes the lock handling easier. * Change the audit_log_start() sleep behavior so that we never sleep on auditd events (unchanged) or if the caller is holding the audit_cmd_mutex (changed). Previously we didn't sleep if the caller was auditd or if the message type fell between a certain range; the type check was a poor effort of doing what the cmd_mutex check now does. Richard Guy Briggs originally proposed not sleeping the cmd_mutex owner several years ago but his patch wasn't acceptable at the time. At least the idea lives on here. * A problem with the lost record counter has been resolved. Steve Grubb and I both happened to notice this problem and according to some quick testing by Steve, this problem goes back quite some time. It's largely a harmless problem, although it may have left some careful sysadmins quite puzzled. Cc: <stable@vger.kernel.org> # 4.10.x- Reported-by: Dmitry Vyukov <dvyukov@google.com> Signed-off-by: Paul Moore <paul@paul-moore.com>
2017-03-21 23:26:35 +08:00
if (audit_rate_check()) {
skb = ab->skb;
ab->skb = NULL;
audit: fix auditd/kernel connection state tracking What started as a rather straightforward race condition reported by Dmitry using the syzkaller fuzzer ended up revealing some major problems with how the audit subsystem managed its netlink sockets and its connection with the userspace audit daemon. Fixing this properly had quite the cascading effect and what we are left with is this rather large and complicated patch. My initial goal was to try and decompose this patch into multiple smaller patches, but the way these changes are intertwined makes it difficult to split these changes into meaningful pieces that don't break or somehow make things worse for the intermediate states. The patch makes a number of changes, but the most significant are highlighted below: * The auditd tracking variables, e.g. audit_sock, are now gone and replaced by a RCU/spin_lock protected variable auditd_conn which is a structure containing all of the auditd tracking information. * We no longer track the auditd sock directly, instead we track it via the network namespace in which it resides and we use the audit socket associated with that namespace. In spirit, this is what the code was trying to do prior to this patch (at least I think that is what the original authors intended), but it was done rather poorly and added a layer of obfuscation that only masked the underlying problems. * Big backlog queue cleanup, again. In v4.10 we made some pretty big changes to how the audit backlog queues work, here we haven't changed the queue design so much as cleaned up the implementation. Brought about by the locking changes, we've simplified kauditd_thread() quite a bit by consolidating the queue handling into a new helper function, kauditd_send_queue(), which allows us to eliminate a lot of very similar code and makes the looping logic in kauditd_thread() clearer. * All netlink messages sent to auditd are now sent via auditd_send_unicast_skb(). Other than just making sense, this makes the lock handling easier. * Change the audit_log_start() sleep behavior so that we never sleep on auditd events (unchanged) or if the caller is holding the audit_cmd_mutex (changed). Previously we didn't sleep if the caller was auditd or if the message type fell between a certain range; the type check was a poor effort of doing what the cmd_mutex check now does. Richard Guy Briggs originally proposed not sleeping the cmd_mutex owner several years ago but his patch wasn't acceptable at the time. At least the idea lives on here. * A problem with the lost record counter has been resolved. Steve Grubb and I both happened to notice this problem and according to some quick testing by Steve, this problem goes back quite some time. It's largely a harmless problem, although it may have left some careful sysadmins quite puzzled. Cc: <stable@vger.kernel.org> # 4.10.x- Reported-by: Dmitry Vyukov <dvyukov@google.com> Signed-off-by: Paul Moore <paul@paul-moore.com>
2017-03-21 23:26:35 +08:00
/* setup the netlink header, see the comments in
* kauditd_send_multicast_skb() for length quirks */
nlh = nlmsg_hdr(skb);
nlh->nlmsg_len = skb->len - NLMSG_HDRLEN;
/* queue the netlink packet and poke the kauditd thread */
skb_queue_tail(&audit_queue, skb);
wake_up_interruptible(&kauditd_wait);
} else
audit_log_lost("rate limit exceeded");
audit_buffer_free(ab);
}
/**
* 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.
*/
void audit_log(struct audit_context *ctx, gfp_t gfp_mask, int type,
const char *fmt, ...)
{
struct audit_buffer *ab;
va_list args;
ab = audit_log_start(ctx, gfp_mask, type);
if (ab) {
va_start(args, fmt);
audit_log_vformat(ab, fmt, args);
va_end(args);
audit_log_end(ab);
}
}
EXPORT_SYMBOL(audit_log_start);
EXPORT_SYMBOL(audit_log_end);
EXPORT_SYMBOL(audit_log_format);
EXPORT_SYMBOL(audit_log);