2005-04-29 23:23:29 +08:00
|
|
|
/* audit.h -- Auditing support
|
2005-04-17 06:20:36 +08:00
|
|
|
*
|
|
|
|
* Copyright 2003-2004 Red Hat Inc., Durham, North Carolina.
|
|
|
|
* All Rights Reserved.
|
|
|
|
*
|
|
|
|
* This program is free software; you can redistribute it and/or modify
|
|
|
|
* it under the terms of the GNU General Public License as published by
|
|
|
|
* the Free Software Foundation; either version 2 of the License, or
|
|
|
|
* (at your option) any later version.
|
|
|
|
*
|
|
|
|
* This program is distributed in the hope that it will be useful,
|
|
|
|
* but WITHOUT ANY WARRANTY; without even the implied warranty of
|
|
|
|
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
|
|
|
|
* GNU General Public License for more details.
|
|
|
|
*
|
|
|
|
* You should have received a copy of the GNU General Public License
|
|
|
|
* along with this program; if not, write to the Free Software
|
|
|
|
* Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
|
|
|
|
*
|
|
|
|
* Written by Rickard E. (Rik) Faith <faith@redhat.com>
|
|
|
|
*
|
|
|
|
*/
|
|
|
|
|
|
|
|
#ifndef _LINUX_AUDIT_H_
|
|
|
|
#define _LINUX_AUDIT_H_
|
|
|
|
|
2007-03-06 17:42:17 +08:00
|
|
|
#include <linux/types.h>
|
2006-04-27 07:11:01 +08:00
|
|
|
#include <linux/elf-em.h>
|
2005-04-29 23:08:28 +08:00
|
|
|
|
2005-05-14 01:17:42 +08:00
|
|
|
/* The netlink messages for the audit system is divided into blocks:
|
|
|
|
* 1000 - 1099 are for commanding the audit system
|
|
|
|
* 1100 - 1199 user space trusted application messages
|
|
|
|
* 1200 - 1299 messages internal to the audit daemon
|
|
|
|
* 1300 - 1399 audit event messages
|
|
|
|
* 1400 - 1499 SE Linux use
|
2005-11-03 23:48:08 +08:00
|
|
|
* 1500 - 1599 kernel LSPP events
|
|
|
|
* 1600 - 1699 kernel crypto events
|
2005-12-03 21:39:35 +08:00
|
|
|
* 1700 - 1799 kernel anomaly records
|
|
|
|
* 1800 - 1999 future kernel use (maybe integrity labels and related events)
|
2005-11-03 23:48:08 +08:00
|
|
|
* 2000 is for otherwise unclassified kernel audit messages (legacy)
|
|
|
|
* 2001 - 2099 unused (kernel)
|
|
|
|
* 2100 - 2199 user space anomaly records
|
|
|
|
* 2200 - 2299 user space actions taken in response to anomalies
|
|
|
|
* 2300 - 2399 user space generated LSPP events
|
|
|
|
* 2400 - 2499 user space crypto events
|
|
|
|
* 2500 - 2999 future user space (maybe integrity labels and related events)
|
2005-05-14 01:17:42 +08:00
|
|
|
*
|
2005-11-03 23:48:08 +08:00
|
|
|
* Messages from 1000-1199 are bi-directional. 1200-1299 & 2100 - 2999 are
|
|
|
|
* exclusively user space. 1300-2099 is kernel --> user space
|
|
|
|
* communication.
|
2005-05-14 01:17:42 +08:00
|
|
|
*/
|
2005-05-06 19:38:39 +08:00
|
|
|
#define AUDIT_GET 1000 /* Get status */
|
|
|
|
#define AUDIT_SET 1001 /* Set status (enable/disable/auditd) */
|
2006-02-08 01:05:27 +08:00
|
|
|
#define AUDIT_LIST 1002 /* List syscall rules -- deprecated */
|
|
|
|
#define AUDIT_ADD 1003 /* Add syscall rule -- deprecated */
|
|
|
|
#define AUDIT_DEL 1004 /* Delete syscall rule -- deprecated */
|
2005-05-14 01:17:42 +08:00
|
|
|
#define AUDIT_USER 1005 /* Message from userspace -- deprecated */
|
2005-05-06 19:38:39 +08:00
|
|
|
#define AUDIT_LOGIN 1006 /* Define the login id and information */
|
2005-05-14 01:17:42 +08:00
|
|
|
#define AUDIT_WATCH_INS 1007 /* Insert file/dir watch entry */
|
|
|
|
#define AUDIT_WATCH_REM 1008 /* Remove file/dir watch entry */
|
|
|
|
#define AUDIT_WATCH_LIST 1009 /* List all file/dir watches */
|
|
|
|
#define AUDIT_SIGNAL_INFO 1010 /* Get info about sender of signal to auditd */
|
2006-02-08 01:05:27 +08:00
|
|
|
#define AUDIT_ADD_RULE 1011 /* Add syscall filtering rule */
|
|
|
|
#define AUDIT_DEL_RULE 1012 /* Delete syscall filtering rule */
|
|
|
|
#define AUDIT_LIST_RULES 1013 /* List syscall filtering rules */
|
[PATCH] audit: watching subtrees
New kind of audit rule predicates: "object is visible in given subtree".
The part that can be sanely implemented, that is. Limitations:
* if you have hardlink from outside of tree, you'd better watch
it too (or just watch the object itself, obviously)
* if you mount something under a watched tree, tell audit
that new chunk should be added to watched subtrees
* if you umount something in a watched tree and it's still mounted
elsewhere, you will get matches on events happening there. New command
tells audit to recalculate the trees, trimming such sources of false
positives.
Note that it's _not_ about path - if something mounted in several places
(multiple mount, bindings, different namespaces, etc.), the match does
_not_ depend on which one we are using for access.
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
2007-07-22 20:04:18 +08:00
|
|
|
#define AUDIT_TRIM 1014 /* Trim junk from watched tree */
|
|
|
|
#define AUDIT_MAKE_EQUIV 1015 /* Append to watched tree */
|
2007-08-23 17:19:53 +08:00
|
|
|
#define AUDIT_TTY_GET 1016 /* Get TTY auditing status */
|
|
|
|
#define AUDIT_TTY_SET 1017 /* Set TTY auditing status */
|
2005-05-14 01:17:42 +08:00
|
|
|
|
2005-06-22 21:56:47 +08:00
|
|
|
#define AUDIT_FIRST_USER_MSG 1100 /* Userspace messages mostly uninteresting to kernel */
|
|
|
|
#define AUDIT_USER_AVC 1107 /* We filter this differently */
|
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
|
|
|
#define AUDIT_USER_TTY 1124 /* Non-ICANON TTY input meaning */
|
2005-05-18 17:21:07 +08:00
|
|
|
#define AUDIT_LAST_USER_MSG 1199
|
2005-11-03 23:48:08 +08:00
|
|
|
#define AUDIT_FIRST_USER_MSG2 2100 /* More user space messages */
|
|
|
|
#define AUDIT_LAST_USER_MSG2 2999
|
2005-05-14 01:17:42 +08:00
|
|
|
|
|
|
|
#define AUDIT_DAEMON_START 1200 /* Daemon startup record */
|
|
|
|
#define AUDIT_DAEMON_END 1201 /* Daemon normal stop record */
|
|
|
|
#define AUDIT_DAEMON_ABORT 1202 /* Daemon error stop record */
|
|
|
|
#define AUDIT_DAEMON_CONFIG 1203 /* Daemon config change */
|
|
|
|
|
|
|
|
#define AUDIT_SYSCALL 1300 /* Syscall event */
|
2006-09-29 23:56:49 +08:00
|
|
|
/* #define AUDIT_FS_WATCH 1301 * Deprecated */
|
2005-05-27 19:17:28 +08:00
|
|
|
#define AUDIT_PATH 1302 /* Filename path information */
|
2005-05-14 01:17:42 +08:00
|
|
|
#define AUDIT_IPC 1303 /* IPC record */
|
2005-05-17 19:08:48 +08:00
|
|
|
#define AUDIT_SOCKETCALL 1304 /* sys_socketcall arguments */
|
2005-05-14 01:17:42 +08:00
|
|
|
#define AUDIT_CONFIG_CHANGE 1305 /* Audit system configuration change */
|
2005-05-17 19:08:48 +08:00
|
|
|
#define AUDIT_SOCKADDR 1306 /* sockaddr copied as syscall arg */
|
2005-05-27 19:17:28 +08:00
|
|
|
#define AUDIT_CWD 1307 /* Current working directory */
|
2006-04-27 02:04:08 +08:00
|
|
|
#define AUDIT_EXECVE 1309 /* execve arguments */
|
2006-04-03 05:07:33 +08:00
|
|
|
#define AUDIT_IPC_SET_PERM 1311 /* IPC new permissions record type */
|
2006-05-25 05:09:55 +08:00
|
|
|
#define AUDIT_MQ_OPEN 1312 /* POSIX MQ open record type */
|
|
|
|
#define AUDIT_MQ_SENDRECV 1313 /* POSIX MQ send/receive record type */
|
|
|
|
#define AUDIT_MQ_NOTIFY 1314 /* POSIX MQ notify record type */
|
|
|
|
#define AUDIT_MQ_GETSETATTR 1315 /* POSIX MQ get/set attribute record type */
|
2006-09-29 23:56:49 +08:00
|
|
|
#define AUDIT_KERNEL_OTHER 1316 /* For use by 3rd party modules */
|
2007-02-07 14:48:00 +08:00
|
|
|
#define AUDIT_FD_PAIR 1317 /* audit record for pipe/socketpair */
|
2007-03-21 01:58:35 +08:00
|
|
|
#define AUDIT_OBJ_PID 1318 /* ptrace target */
|
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
|
|
|
#define AUDIT_TTY 1319 /* Input on an administrative TTY */
|
2008-01-08 02:49:15 +08:00
|
|
|
#define AUDIT_EOE 1320 /* End of multi-record event */
|
2008-11-11 18:48:18 +08:00
|
|
|
#define AUDIT_BPRM_FCAPS 1321 /* Information about fcaps increasing perms */
|
2008-11-11 18:48:22 +08:00
|
|
|
#define AUDIT_CAPSET 1322 /* Record showing argument to sys_capset */
|
2005-05-14 01:17:42 +08:00
|
|
|
|
|
|
|
#define AUDIT_AVC 1400 /* SE Linux avc denial or grant */
|
|
|
|
#define AUDIT_SELINUX_ERR 1401 /* Internal SE Linux Errors */
|
2005-05-21 07:15:52 +08:00
|
|
|
#define AUDIT_AVC_PATH 1402 /* dentry, vfsmount pair from avc */
|
2006-01-04 22:08:39 +08:00
|
|
|
#define AUDIT_MAC_POLICY_LOAD 1403 /* Policy file load */
|
|
|
|
#define AUDIT_MAC_STATUS 1404 /* Changed enforcing,permissive,off */
|
|
|
|
#define AUDIT_MAC_CONFIG_CHANGE 1405 /* Changes to booleans */
|
2006-09-30 08:05:05 +08:00
|
|
|
#define AUDIT_MAC_UNLBL_ALLOW 1406 /* NetLabel: allow unlabeled traffic */
|
|
|
|
#define AUDIT_MAC_CIPSOV4_ADD 1407 /* NetLabel: add CIPSOv4 DOI entry */
|
|
|
|
#define AUDIT_MAC_CIPSOV4_DEL 1408 /* NetLabel: del CIPSOv4 DOI entry */
|
|
|
|
#define AUDIT_MAC_MAP_ADD 1409 /* NetLabel: add LSM domain mapping */
|
|
|
|
#define AUDIT_MAC_MAP_DEL 1410 /* NetLabel: del LSM domain mapping */
|
2007-09-18 02:51:22 +08:00
|
|
|
#define AUDIT_MAC_IPSEC_ADDSA 1411 /* Not used */
|
|
|
|
#define AUDIT_MAC_IPSEC_DELSA 1412 /* Not used */
|
|
|
|
#define AUDIT_MAC_IPSEC_ADDSPD 1413 /* Not used */
|
|
|
|
#define AUDIT_MAC_IPSEC_DELSPD 1414 /* Not used */
|
|
|
|
#define AUDIT_MAC_IPSEC_EVENT 1415 /* Audit an IPSec event */
|
2008-01-29 21:44:23 +08:00
|
|
|
#define AUDIT_MAC_UNLBL_STCADD 1416 /* NetLabel: add a static label */
|
|
|
|
#define AUDIT_MAC_UNLBL_STCDEL 1417 /* NetLabel: del a static label */
|
2005-05-06 19:38:39 +08:00
|
|
|
|
2005-12-03 21:39:35 +08:00
|
|
|
#define AUDIT_FIRST_KERN_ANOM_MSG 1700
|
|
|
|
#define AUDIT_LAST_KERN_ANOM_MSG 1799
|
|
|
|
#define AUDIT_ANOM_PROMISCUOUS 1700 /* Device changed promiscuous mode */
|
2007-04-19 22:28:21 +08:00
|
|
|
#define AUDIT_ANOM_ABEND 1701 /* Process ended abnormally */
|
2005-12-03 21:39:35 +08:00
|
|
|
|
2005-05-06 19:38:39 +08:00
|
|
|
#define AUDIT_KERNEL 2000 /* Asynchronous audit record. NOT A REQUEST. */
|
2005-04-17 06:20:36 +08:00
|
|
|
|
|
|
|
/* Rule flags */
|
2005-06-20 02:35:50 +08:00
|
|
|
#define AUDIT_FILTER_USER 0x00 /* Apply rule to user-generated messages */
|
|
|
|
#define AUDIT_FILTER_TASK 0x01 /* Apply rule at task creation (not syscall) */
|
|
|
|
#define AUDIT_FILTER_ENTRY 0x02 /* Apply rule at syscall entry */
|
|
|
|
#define AUDIT_FILTER_WATCH 0x03 /* Apply rule to file system watches */
|
|
|
|
#define AUDIT_FILTER_EXIT 0x04 /* Apply rule at syscall exit */
|
2005-11-04 00:12:36 +08:00
|
|
|
#define AUDIT_FILTER_TYPE 0x05 /* Apply rule at audit_log_start */
|
2005-06-20 02:35:50 +08:00
|
|
|
|
2005-11-04 00:12:36 +08:00
|
|
|
#define AUDIT_NR_FILTERS 6
|
2005-06-20 02:35:50 +08:00
|
|
|
|
|
|
|
#define AUDIT_FILTER_PREPEND 0x10 /* Prepend to front of list */
|
2005-04-17 06:20:36 +08:00
|
|
|
|
|
|
|
/* Rule actions */
|
|
|
|
#define AUDIT_NEVER 0 /* Do not build context if rule matches */
|
|
|
|
#define AUDIT_POSSIBLE 1 /* Build context if rule matches */
|
|
|
|
#define AUDIT_ALWAYS 2 /* Generate audit record if rule matches */
|
|
|
|
|
|
|
|
/* Rule structure sizes -- if these change, different AUDIT_ADD and
|
|
|
|
* AUDIT_LIST commands must be implemented. */
|
|
|
|
#define AUDIT_MAX_FIELDS 64
|
2008-04-18 22:36:22 +08:00
|
|
|
#define AUDIT_MAX_KEY_LEN 256
|
2005-04-17 06:20:36 +08:00
|
|
|
#define AUDIT_BITMASK_SIZE 64
|
|
|
|
#define AUDIT_WORD(nr) ((__u32)((nr)/32))
|
|
|
|
#define AUDIT_BIT(nr) (1 << ((nr) - AUDIT_WORD(nr)*32))
|
|
|
|
|
2006-07-01 15:56:16 +08:00
|
|
|
#define AUDIT_SYSCALL_CLASSES 16
|
|
|
|
#define AUDIT_CLASS_DIR_WRITE 0
|
|
|
|
#define AUDIT_CLASS_DIR_WRITE_32 1
|
|
|
|
#define AUDIT_CLASS_CHATTR 2
|
|
|
|
#define AUDIT_CLASS_CHATTR_32 3
|
2006-09-01 07:05:56 +08:00
|
|
|
#define AUDIT_CLASS_READ 4
|
|
|
|
#define AUDIT_CLASS_READ_32 5
|
|
|
|
#define AUDIT_CLASS_WRITE 6
|
|
|
|
#define AUDIT_CLASS_WRITE_32 7
|
2007-03-30 06:00:37 +08:00
|
|
|
#define AUDIT_CLASS_SIGNAL 8
|
|
|
|
#define AUDIT_CLASS_SIGNAL_32 9
|
2006-07-01 15:56:16 +08:00
|
|
|
|
[PATCH] Filter rule comparators
Currently, audit only supports the "=" and "!=" operators in the -F
filter rules.
This patch reworks the support for "=" and "!=", and adds support
for ">", ">=", "<", and "<=".
This turned out to be a pretty clean, and simply process. I ended up
using the high order bits of the "field", as suggested by Steve and Amy.
This allowed for no changes whatsoever to the netlink communications.
See the documentation within the patch in the include/linux/audit.h
area, where there is a table that explains the reasoning of the bitmask
assignments clearly.
The patch adds a new function, audit_comparator(left, op, right).
This function will perform the specified comparison (op, which defaults
to "==" for backward compatibility) between two values (left and right).
If the negate bit is on, it will negate whatever that result was. This
value is returned.
Signed-off-by: Dustin Kirkland <dustin.kirkland@us.ibm.com>
Signed-off-by: David Woodhouse <dwmw2@infradead.org>
2005-11-03 23:41:46 +08:00
|
|
|
/* This bitmask is used to validate user input. It represents all bits that
|
|
|
|
* are currently used in an audit field constant understood by the kernel.
|
|
|
|
* If you are adding a new #define AUDIT_<whatever>, please ensure that
|
|
|
|
* AUDIT_UNUSED_BITS is updated if need be. */
|
2007-06-05 05:00:14 +08:00
|
|
|
#define AUDIT_UNUSED_BITS 0x07FFFC00
|
[PATCH] Filter rule comparators
Currently, audit only supports the "=" and "!=" operators in the -F
filter rules.
This patch reworks the support for "=" and "!=", and adds support
for ">", ">=", "<", and "<=".
This turned out to be a pretty clean, and simply process. I ended up
using the high order bits of the "field", as suggested by Steve and Amy.
This allowed for no changes whatsoever to the netlink communications.
See the documentation within the patch in the include/linux/audit.h
area, where there is a table that explains the reasoning of the bitmask
assignments clearly.
The patch adds a new function, audit_comparator(left, op, right).
This function will perform the specified comparison (op, which defaults
to "==" for backward compatibility) between two values (left and right).
If the negate bit is on, it will negate whatever that result was. This
value is returned.
Signed-off-by: Dustin Kirkland <dustin.kirkland@us.ibm.com>
Signed-off-by: David Woodhouse <dwmw2@infradead.org>
2005-11-03 23:41:46 +08:00
|
|
|
|
|
|
|
|
2005-04-17 06:20:36 +08:00
|
|
|
/* Rule fields */
|
|
|
|
/* These are useful when checking the
|
|
|
|
* task structure at task creation time
|
|
|
|
* (AUDIT_PER_TASK). */
|
|
|
|
#define AUDIT_PID 0
|
|
|
|
#define AUDIT_UID 1
|
|
|
|
#define AUDIT_EUID 2
|
|
|
|
#define AUDIT_SUID 3
|
|
|
|
#define AUDIT_FSUID 4
|
|
|
|
#define AUDIT_GID 5
|
|
|
|
#define AUDIT_EGID 6
|
|
|
|
#define AUDIT_SGID 7
|
|
|
|
#define AUDIT_FSGID 8
|
|
|
|
#define AUDIT_LOGINUID 9
|
|
|
|
#define AUDIT_PERS 10
|
2005-04-29 23:08:28 +08:00
|
|
|
#define AUDIT_ARCH 11
|
2005-11-04 00:12:36 +08:00
|
|
|
#define AUDIT_MSGTYPE 12
|
2006-06-30 05:56:39 +08:00
|
|
|
#define AUDIT_SUBJ_USER 13 /* security label user */
|
|
|
|
#define AUDIT_SUBJ_ROLE 14 /* security label role */
|
|
|
|
#define AUDIT_SUBJ_TYPE 15 /* security label type */
|
|
|
|
#define AUDIT_SUBJ_SEN 16 /* security label sensitivity label */
|
|
|
|
#define AUDIT_SUBJ_CLR 17 /* security label clearance label */
|
2006-05-06 20:26:27 +08:00
|
|
|
#define AUDIT_PPID 18
|
2006-06-30 05:56:39 +08:00
|
|
|
#define AUDIT_OBJ_USER 19
|
|
|
|
#define AUDIT_OBJ_ROLE 20
|
|
|
|
#define AUDIT_OBJ_TYPE 21
|
|
|
|
#define AUDIT_OBJ_LEV_LOW 22
|
|
|
|
#define AUDIT_OBJ_LEV_HIGH 23
|
2005-04-17 06:20:36 +08:00
|
|
|
|
|
|
|
/* These are ONLY useful when checking
|
|
|
|
* at syscall exit time (AUDIT_AT_EXIT). */
|
|
|
|
#define AUDIT_DEVMAJOR 100
|
|
|
|
#define AUDIT_DEVMINOR 101
|
|
|
|
#define AUDIT_INODE 102
|
|
|
|
#define AUDIT_EXIT 103
|
|
|
|
#define AUDIT_SUCCESS 104 /* exit >= 0; value ignored */
|
[PATCH] audit: path-based rules
In this implementation, audit registers inotify watches on the parent
directories of paths specified in audit rules. When audit's inotify
event handler is called, it updates any affected rules based on the
filesystem event. If the parent directory is renamed, removed, or its
filesystem is unmounted, audit removes all rules referencing that
inotify watch.
To keep things simple, this implementation limits location-based
auditing to the directory entries in an existing directory. Given
a path-based rule for /foo/bar/passwd, the following table applies:
passwd modified -- audit event logged
passwd replaced -- audit event logged, rules list updated
bar renamed -- rule removed
foo renamed -- untracked, meaning that the rule now applies to
the new location
Audit users typically want to have many rules referencing filesystem
objects, which can significantly impact filtering performance. This
patch also adds an inode-number-based rule hash to mitigate this
situation.
The patch is relative to the audit git tree:
http://kernel.org/git/?p=linux/kernel/git/viro/audit-current.git;a=summary
and uses the inotify kernel API:
http://lkml.org/lkml/2006/6/1/145
Signed-off-by: Amy Griffis <amy.griffis@hp.com>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
2006-04-08 04:55:56 +08:00
|
|
|
#define AUDIT_WATCH 105
|
2006-09-01 07:26:40 +08:00
|
|
|
#define AUDIT_PERM 106
|
[PATCH] audit: watching subtrees
New kind of audit rule predicates: "object is visible in given subtree".
The part that can be sanely implemented, that is. Limitations:
* if you have hardlink from outside of tree, you'd better watch
it too (or just watch the object itself, obviously)
* if you mount something under a watched tree, tell audit
that new chunk should be added to watched subtrees
* if you umount something in a watched tree and it's still mounted
elsewhere, you will get matches on events happening there. New command
tells audit to recalculate the trees, trimming such sources of false
positives.
Note that it's _not_ about path - if something mounted in several places
(multiple mount, bindings, different namespaces, etc.), the match does
_not_ depend on which one we are using for access.
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
2007-07-22 20:04:18 +08:00
|
|
|
#define AUDIT_DIR 107
|
2008-04-28 16:15:49 +08:00
|
|
|
#define AUDIT_FILETYPE 108
|
2005-04-17 06:20:36 +08:00
|
|
|
|
|
|
|
#define AUDIT_ARG0 200
|
|
|
|
#define AUDIT_ARG1 (AUDIT_ARG0+1)
|
|
|
|
#define AUDIT_ARG2 (AUDIT_ARG0+2)
|
|
|
|
#define AUDIT_ARG3 (AUDIT_ARG0+3)
|
|
|
|
|
2006-06-15 06:45:21 +08:00
|
|
|
#define AUDIT_FILTERKEY 210
|
|
|
|
|
[PATCH] Filter rule comparators
Currently, audit only supports the "=" and "!=" operators in the -F
filter rules.
This patch reworks the support for "=" and "!=", and adds support
for ">", ">=", "<", and "<=".
This turned out to be a pretty clean, and simply process. I ended up
using the high order bits of the "field", as suggested by Steve and Amy.
This allowed for no changes whatsoever to the netlink communications.
See the documentation within the patch in the include/linux/audit.h
area, where there is a table that explains the reasoning of the bitmask
assignments clearly.
The patch adds a new function, audit_comparator(left, op, right).
This function will perform the specified comparison (op, which defaults
to "==" for backward compatibility) between two values (left and right).
If the negate bit is on, it will negate whatever that result was. This
value is returned.
Signed-off-by: Dustin Kirkland <dustin.kirkland@us.ibm.com>
Signed-off-by: David Woodhouse <dwmw2@infradead.org>
2005-11-03 23:41:46 +08:00
|
|
|
#define AUDIT_NEGATE 0x80000000
|
|
|
|
|
|
|
|
/* These are the supported operators.
|
2007-06-05 05:00:14 +08:00
|
|
|
* 4 2 1 8
|
|
|
|
* = > < ?
|
|
|
|
* ----------
|
|
|
|
* 0 0 0 0 00 nonsense
|
|
|
|
* 0 0 0 1 08 & bit mask
|
|
|
|
* 0 0 1 0 10 <
|
|
|
|
* 0 1 0 0 20 >
|
|
|
|
* 0 1 1 0 30 !=
|
|
|
|
* 1 0 0 0 40 =
|
|
|
|
* 1 0 0 1 48 &= bit test
|
|
|
|
* 1 0 1 0 50 <=
|
|
|
|
* 1 1 0 0 60 >=
|
|
|
|
* 1 1 1 1 78 all operators
|
[PATCH] Filter rule comparators
Currently, audit only supports the "=" and "!=" operators in the -F
filter rules.
This patch reworks the support for "=" and "!=", and adds support
for ">", ">=", "<", and "<=".
This turned out to be a pretty clean, and simply process. I ended up
using the high order bits of the "field", as suggested by Steve and Amy.
This allowed for no changes whatsoever to the netlink communications.
See the documentation within the patch in the include/linux/audit.h
area, where there is a table that explains the reasoning of the bitmask
assignments clearly.
The patch adds a new function, audit_comparator(left, op, right).
This function will perform the specified comparison (op, which defaults
to "==" for backward compatibility) between two values (left and right).
If the negate bit is on, it will negate whatever that result was. This
value is returned.
Signed-off-by: Dustin Kirkland <dustin.kirkland@us.ibm.com>
Signed-off-by: David Woodhouse <dwmw2@infradead.org>
2005-11-03 23:41:46 +08:00
|
|
|
*/
|
2007-06-05 05:00:14 +08:00
|
|
|
#define AUDIT_BIT_MASK 0x08000000
|
[PATCH] Filter rule comparators
Currently, audit only supports the "=" and "!=" operators in the -F
filter rules.
This patch reworks the support for "=" and "!=", and adds support
for ">", ">=", "<", and "<=".
This turned out to be a pretty clean, and simply process. I ended up
using the high order bits of the "field", as suggested by Steve and Amy.
This allowed for no changes whatsoever to the netlink communications.
See the documentation within the patch in the include/linux/audit.h
area, where there is a table that explains the reasoning of the bitmask
assignments clearly.
The patch adds a new function, audit_comparator(left, op, right).
This function will perform the specified comparison (op, which defaults
to "==" for backward compatibility) between two values (left and right).
If the negate bit is on, it will negate whatever that result was. This
value is returned.
Signed-off-by: Dustin Kirkland <dustin.kirkland@us.ibm.com>
Signed-off-by: David Woodhouse <dwmw2@infradead.org>
2005-11-03 23:41:46 +08:00
|
|
|
#define AUDIT_LESS_THAN 0x10000000
|
|
|
|
#define AUDIT_GREATER_THAN 0x20000000
|
|
|
|
#define AUDIT_NOT_EQUAL 0x30000000
|
|
|
|
#define AUDIT_EQUAL 0x40000000
|
2007-06-05 05:00:14 +08:00
|
|
|
#define AUDIT_BIT_TEST (AUDIT_BIT_MASK|AUDIT_EQUAL)
|
[PATCH] Filter rule comparators
Currently, audit only supports the "=" and "!=" operators in the -F
filter rules.
This patch reworks the support for "=" and "!=", and adds support
for ">", ">=", "<", and "<=".
This turned out to be a pretty clean, and simply process. I ended up
using the high order bits of the "field", as suggested by Steve and Amy.
This allowed for no changes whatsoever to the netlink communications.
See the documentation within the patch in the include/linux/audit.h
area, where there is a table that explains the reasoning of the bitmask
assignments clearly.
The patch adds a new function, audit_comparator(left, op, right).
This function will perform the specified comparison (op, which defaults
to "==" for backward compatibility) between two values (left and right).
If the negate bit is on, it will negate whatever that result was. This
value is returned.
Signed-off-by: Dustin Kirkland <dustin.kirkland@us.ibm.com>
Signed-off-by: David Woodhouse <dwmw2@infradead.org>
2005-11-03 23:41:46 +08:00
|
|
|
#define AUDIT_LESS_THAN_OR_EQUAL (AUDIT_LESS_THAN|AUDIT_EQUAL)
|
|
|
|
#define AUDIT_GREATER_THAN_OR_EQUAL (AUDIT_GREATER_THAN|AUDIT_EQUAL)
|
2007-06-05 05:00:14 +08:00
|
|
|
#define AUDIT_OPERATORS (AUDIT_EQUAL|AUDIT_NOT_EQUAL|AUDIT_BIT_MASK)
|
2005-04-17 06:20:36 +08:00
|
|
|
|
|
|
|
/* Status symbols */
|
|
|
|
/* Mask values */
|
|
|
|
#define AUDIT_STATUS_ENABLED 0x0001
|
|
|
|
#define AUDIT_STATUS_FAILURE 0x0002
|
|
|
|
#define AUDIT_STATUS_PID 0x0004
|
|
|
|
#define AUDIT_STATUS_RATE_LIMIT 0x0008
|
|
|
|
#define AUDIT_STATUS_BACKLOG_LIMIT 0x0010
|
|
|
|
/* Failure-to-log actions */
|
|
|
|
#define AUDIT_FAIL_SILENT 0
|
|
|
|
#define AUDIT_FAIL_PRINTK 1
|
|
|
|
#define AUDIT_FAIL_PANIC 2
|
|
|
|
|
2005-04-29 23:08:28 +08:00
|
|
|
/* distinguish syscall tables */
|
|
|
|
#define __AUDIT_ARCH_64BIT 0x80000000
|
|
|
|
#define __AUDIT_ARCH_LE 0x40000000
|
|
|
|
#define AUDIT_ARCH_ALPHA (EM_ALPHA|__AUDIT_ARCH_64BIT|__AUDIT_ARCH_LE)
|
|
|
|
#define AUDIT_ARCH_ARM (EM_ARM|__AUDIT_ARCH_LE)
|
|
|
|
#define AUDIT_ARCH_ARMEB (EM_ARM)
|
|
|
|
#define AUDIT_ARCH_CRIS (EM_CRIS|__AUDIT_ARCH_LE)
|
|
|
|
#define AUDIT_ARCH_FRV (EM_FRV)
|
|
|
|
#define AUDIT_ARCH_H8300 (EM_H8_300)
|
|
|
|
#define AUDIT_ARCH_I386 (EM_386|__AUDIT_ARCH_LE)
|
|
|
|
#define AUDIT_ARCH_IA64 (EM_IA_64|__AUDIT_ARCH_64BIT|__AUDIT_ARCH_LE)
|
|
|
|
#define AUDIT_ARCH_M32R (EM_M32R)
|
|
|
|
#define AUDIT_ARCH_M68K (EM_68K)
|
|
|
|
#define AUDIT_ARCH_MIPS (EM_MIPS)
|
|
|
|
#define AUDIT_ARCH_MIPSEL (EM_MIPS|__AUDIT_ARCH_LE)
|
|
|
|
#define AUDIT_ARCH_MIPS64 (EM_MIPS|__AUDIT_ARCH_64BIT)
|
|
|
|
#define AUDIT_ARCH_MIPSEL64 (EM_MIPS|__AUDIT_ARCH_64BIT|__AUDIT_ARCH_LE)
|
|
|
|
#define AUDIT_ARCH_PARISC (EM_PARISC)
|
|
|
|
#define AUDIT_ARCH_PARISC64 (EM_PARISC|__AUDIT_ARCH_64BIT)
|
|
|
|
#define AUDIT_ARCH_PPC (EM_PPC)
|
|
|
|
#define AUDIT_ARCH_PPC64 (EM_PPC64|__AUDIT_ARCH_64BIT)
|
|
|
|
#define AUDIT_ARCH_S390 (EM_S390)
|
|
|
|
#define AUDIT_ARCH_S390X (EM_S390|__AUDIT_ARCH_64BIT)
|
|
|
|
#define AUDIT_ARCH_SH (EM_SH)
|
|
|
|
#define AUDIT_ARCH_SHEL (EM_SH|__AUDIT_ARCH_LE)
|
|
|
|
#define AUDIT_ARCH_SH64 (EM_SH|__AUDIT_ARCH_64BIT)
|
|
|
|
#define AUDIT_ARCH_SHEL64 (EM_SH|__AUDIT_ARCH_64BIT|__AUDIT_ARCH_LE)
|
|
|
|
#define AUDIT_ARCH_SPARC (EM_SPARC)
|
2005-07-11 10:29:45 +08:00
|
|
|
#define AUDIT_ARCH_SPARC64 (EM_SPARCV9|__AUDIT_ARCH_64BIT)
|
2005-04-29 23:08:28 +08:00
|
|
|
#define AUDIT_ARCH_X86_64 (EM_X86_64|__AUDIT_ARCH_64BIT|__AUDIT_ARCH_LE)
|
|
|
|
|
2006-09-01 07:26:40 +08:00
|
|
|
#define AUDIT_PERM_EXEC 1
|
|
|
|
#define AUDIT_PERM_WRITE 2
|
|
|
|
#define AUDIT_PERM_READ 4
|
|
|
|
#define AUDIT_PERM_ATTR 8
|
|
|
|
|
2005-04-17 06:20:36 +08:00
|
|
|
struct audit_status {
|
|
|
|
__u32 mask; /* Bit mask for valid entries */
|
2005-05-14 01:35:15 +08:00
|
|
|
__u32 enabled; /* 1 = enabled, 0 = disabled */
|
2005-04-17 06:20:36 +08:00
|
|
|
__u32 failure; /* Failure-to-log action */
|
|
|
|
__u32 pid; /* pid of auditd process */
|
|
|
|
__u32 rate_limit; /* messages rate limit (per second) */
|
|
|
|
__u32 backlog_limit; /* waiting messages limit */
|
|
|
|
__u32 lost; /* messages lost */
|
|
|
|
__u32 backlog; /* messages waiting in queue */
|
|
|
|
};
|
|
|
|
|
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
|
|
|
struct audit_tty_status {
|
|
|
|
__u32 enabled; /* 1 = enabled, 0 = disabled */
|
|
|
|
};
|
|
|
|
|
2006-02-08 01:05:27 +08:00
|
|
|
/* audit_rule_data supports filter rules with both integer and string
|
|
|
|
* fields. It corresponds with AUDIT_ADD_RULE, AUDIT_DEL_RULE and
|
|
|
|
* AUDIT_LIST_RULES requests.
|
|
|
|
*/
|
|
|
|
struct audit_rule_data {
|
|
|
|
__u32 flags; /* AUDIT_PER_{TASK,CALL}, AUDIT_PREPEND */
|
|
|
|
__u32 action; /* AUDIT_NEVER, AUDIT_POSSIBLE, AUDIT_ALWAYS */
|
|
|
|
__u32 field_count;
|
2006-01-09 22:48:17 +08:00
|
|
|
__u32 mask[AUDIT_BITMASK_SIZE]; /* syscall(s) affected */
|
2006-02-08 01:05:27 +08:00
|
|
|
__u32 fields[AUDIT_MAX_FIELDS];
|
|
|
|
__u32 values[AUDIT_MAX_FIELDS];
|
|
|
|
__u32 fieldflags[AUDIT_MAX_FIELDS];
|
|
|
|
__u32 buflen; /* total length of string fields */
|
|
|
|
char buf[0]; /* string fields buffer */
|
|
|
|
};
|
|
|
|
|
|
|
|
/* audit_rule is supported to maintain backward compatibility with
|
|
|
|
* userspace. It supports integer fields only and corresponds to
|
|
|
|
* AUDIT_ADD, AUDIT_DEL and AUDIT_LIST requests.
|
|
|
|
*/
|
2005-04-17 06:20:36 +08:00
|
|
|
struct audit_rule { /* for AUDIT_LIST, AUDIT_ADD, and AUDIT_DEL */
|
|
|
|
__u32 flags; /* AUDIT_PER_{TASK,CALL}, AUDIT_PREPEND */
|
|
|
|
__u32 action; /* AUDIT_NEVER, AUDIT_POSSIBLE, AUDIT_ALWAYS */
|
|
|
|
__u32 field_count;
|
|
|
|
__u32 mask[AUDIT_BITMASK_SIZE];
|
|
|
|
__u32 fields[AUDIT_MAX_FIELDS];
|
|
|
|
__u32 values[AUDIT_MAX_FIELDS];
|
|
|
|
};
|
|
|
|
|
|
|
|
#ifdef __KERNEL__
|
2006-04-27 07:11:01 +08:00
|
|
|
#include <linux/sched.h>
|
2005-04-17 06:20:36 +08:00
|
|
|
|
2005-05-06 19:38:39 +08:00
|
|
|
struct audit_sig_info {
|
|
|
|
uid_t uid;
|
|
|
|
pid_t pid;
|
2006-05-25 22:19:47 +08:00
|
|
|
char ctx[0];
|
2005-05-06 19:38:39 +08:00
|
|
|
};
|
|
|
|
|
2005-04-17 06:20:36 +08:00
|
|
|
struct audit_buffer;
|
|
|
|
struct audit_context;
|
|
|
|
struct inode;
|
2005-06-24 21:14:05 +08:00
|
|
|
struct netlink_skb_parms;
|
2006-04-27 02:04:08 +08:00
|
|
|
struct linux_binprm;
|
2006-05-25 05:09:55 +08:00
|
|
|
struct mq_attr;
|
|
|
|
struct mqstat;
|
2008-03-02 04:03:14 +08:00
|
|
|
struct audit_watch;
|
|
|
|
struct audit_tree;
|
|
|
|
|
|
|
|
struct audit_krule {
|
|
|
|
int vers_ops;
|
|
|
|
u32 flags;
|
|
|
|
u32 listnr;
|
|
|
|
u32 action;
|
|
|
|
u32 mask[AUDIT_BITMASK_SIZE];
|
|
|
|
u32 buflen; /* for data alloc on list rules */
|
|
|
|
u32 field_count;
|
|
|
|
char *filterkey; /* ties events to rules */
|
|
|
|
struct audit_field *fields;
|
|
|
|
struct audit_field *arch_f; /* quick access to arch field */
|
|
|
|
struct audit_field *inode_f; /* quick access to an inode field */
|
|
|
|
struct audit_watch *watch; /* associated watch */
|
|
|
|
struct audit_tree *tree; /* associated watched tree */
|
|
|
|
struct list_head rlist; /* entry in audit_{watch,tree}.rules list */
|
|
|
|
};
|
|
|
|
|
|
|
|
struct audit_field {
|
|
|
|
u32 type;
|
|
|
|
u32 val;
|
|
|
|
u32 op;
|
2008-04-19 07:59:43 +08:00
|
|
|
char *lsm_str;
|
|
|
|
void *lsm_rule;
|
2008-03-02 04:03:14 +08:00
|
|
|
};
|
2005-04-17 06:20:36 +08:00
|
|
|
|
2005-04-29 23:08:28 +08:00
|
|
|
#define AUDITSC_INVALID 0
|
|
|
|
#define AUDITSC_SUCCESS 1
|
|
|
|
#define AUDITSC_FAILURE 2
|
|
|
|
#define AUDITSC_RESULT(x) ( ((long)(x))<0?AUDITSC_FAILURE:AUDITSC_SUCCESS )
|
2006-07-01 15:56:16 +08:00
|
|
|
extern int __init audit_register_class(int class, unsigned *list);
|
2006-09-01 07:26:40 +08:00
|
|
|
extern int audit_classify_syscall(int abi, unsigned syscall);
|
2007-03-30 06:01:04 +08:00
|
|
|
extern int audit_classify_arch(int arch);
|
2005-04-17 06:20:36 +08:00
|
|
|
#ifdef CONFIG_AUDITSYSCALL
|
|
|
|
/* These are defined in auditsc.c */
|
|
|
|
/* Public API */
|
2008-11-13 07:37:41 +08:00
|
|
|
extern void audit_finish_fork(struct task_struct *child);
|
2005-04-17 06:20:36 +08:00
|
|
|
extern int audit_alloc(struct task_struct *task);
|
|
|
|
extern void audit_free(struct task_struct *task);
|
2006-03-30 09:23:36 +08:00
|
|
|
extern void audit_syscall_entry(int arch,
|
2005-04-17 06:20:36 +08:00
|
|
|
int major, unsigned long a0, unsigned long a1,
|
|
|
|
unsigned long a2, unsigned long a3);
|
2006-03-30 09:23:36 +08:00
|
|
|
extern void audit_syscall_exit(int failed, long return_code);
|
2006-05-19 04:01:30 +08:00
|
|
|
extern void __audit_getname(const char *name);
|
2005-04-17 06:20:36 +08:00
|
|
|
extern void audit_putname(const char *name);
|
2007-06-08 00:19:32 +08:00
|
|
|
extern void __audit_inode(const char *name, const struct dentry *dentry);
|
|
|
|
extern void __audit_inode_child(const char *dname, const struct dentry *dentry,
|
2006-07-14 01:16:39 +08:00
|
|
|
const struct inode *parent);
|
2007-03-21 01:58:35 +08:00
|
|
|
extern void __audit_ptrace(struct task_struct *t);
|
|
|
|
|
2006-08-03 22:59:26 +08:00
|
|
|
static inline int audit_dummy_context(void)
|
|
|
|
{
|
|
|
|
void *p = current->audit_context;
|
|
|
|
return !p || *(int *)p;
|
|
|
|
}
|
2006-05-19 04:01:30 +08:00
|
|
|
static inline void audit_getname(const char *name)
|
|
|
|
{
|
2006-07-16 18:38:45 +08:00
|
|
|
if (unlikely(!audit_dummy_context()))
|
2006-05-19 04:01:30 +08:00
|
|
|
__audit_getname(name);
|
|
|
|
}
|
2007-06-08 00:19:32 +08:00
|
|
|
static inline void audit_inode(const char *name, const struct dentry *dentry) {
|
2006-07-16 18:38:45 +08:00
|
|
|
if (unlikely(!audit_dummy_context()))
|
2007-06-08 00:19:32 +08:00
|
|
|
__audit_inode(name, dentry);
|
2005-11-04 00:00:25 +08:00
|
|
|
}
|
|
|
|
static inline void audit_inode_child(const char *dname,
|
2007-06-08 00:19:32 +08:00
|
|
|
const struct dentry *dentry,
|
2006-07-14 01:16:39 +08:00
|
|
|
const struct inode *parent) {
|
2006-07-16 18:38:45 +08:00
|
|
|
if (unlikely(!audit_dummy_context()))
|
2007-06-08 00:19:32 +08:00
|
|
|
__audit_inode_child(dname, dentry, parent);
|
2005-11-04 00:00:25 +08:00
|
|
|
}
|
2007-04-19 22:28:21 +08:00
|
|
|
void audit_core_dumps(long signr);
|
2005-04-17 06:20:36 +08:00
|
|
|
|
2007-03-21 01:58:35 +08:00
|
|
|
static inline void audit_ptrace(struct task_struct *t)
|
|
|
|
{
|
|
|
|
if (unlikely(!audit_dummy_context()))
|
|
|
|
__audit_ptrace(t);
|
|
|
|
}
|
|
|
|
|
2005-04-17 06:20:36 +08:00
|
|
|
/* Private API (for audit.c only) */
|
2005-05-22 04:08:09 +08:00
|
|
|
extern unsigned int audit_serial(void);
|
2008-12-06 14:05:50 +08:00
|
|
|
extern int auditsc_get_stamp(struct audit_context *ctx,
|
2005-05-22 04:08:09 +08:00
|
|
|
struct timespec *t, unsigned int *serial);
|
2005-04-30 00:30:07 +08:00
|
|
|
extern int audit_set_loginuid(struct task_struct *task, uid_t loginuid);
|
2008-01-10 17:53:18 +08:00
|
|
|
#define audit_get_loginuid(t) ((t)->loginuid)
|
2008-01-08 23:06:53 +08:00
|
|
|
#define audit_get_sessionid(t) ((t)->sessionid)
|
2006-11-28 03:11:54 +08:00
|
|
|
extern void audit_log_task_context(struct audit_buffer *ab);
|
2008-12-10 16:40:06 +08:00
|
|
|
extern void __audit_ipc_obj(struct kern_ipc_perm *ipcp);
|
2008-12-10 16:47:15 +08:00
|
|
|
extern void __audit_ipc_set_perm(unsigned long qbytes, uid_t uid, gid_t gid, mode_t mode);
|
2006-04-27 02:04:08 +08:00
|
|
|
extern int audit_bprm(struct linux_binprm *bprm);
|
2008-12-10 16:16:51 +08:00
|
|
|
extern void audit_socketcall(int nargs, unsigned long *args);
|
2005-05-17 19:08:48 +08:00
|
|
|
extern int audit_sockaddr(int len, void *addr);
|
2007-02-07 14:48:00 +08:00
|
|
|
extern int __audit_fd_pair(int fd1, int fd2);
|
2005-11-04 01:15:16 +08:00
|
|
|
extern int audit_set_macxattr(const char *name);
|
2006-05-25 05:09:55 +08:00
|
|
|
extern int __audit_mq_open(int oflag, mode_t mode, struct mq_attr __user *u_attr);
|
2008-12-14 16:46:48 +08:00
|
|
|
extern void __audit_mq_sendrecv(mqd_t mqdes, size_t msg_len, unsigned int msg_prio, const struct timespec *abs_timeout);
|
2008-12-10 20:16:12 +08:00
|
|
|
extern void __audit_mq_notify(mqd_t mqdes, const struct sigevent *notification);
|
2008-12-10 19:58:59 +08:00
|
|
|
extern void __audit_mq_getsetattr(mqd_t mqdes, struct mq_attr *mqstat);
|
CRED: Inaugurate COW credentials
Inaugurate copy-on-write credentials management. This uses RCU to manage the
credentials pointer in the task_struct with respect to accesses by other tasks.
A process may only modify its own credentials, and so does not need locking to
access or modify its own credentials.
A mutex (cred_replace_mutex) is added to the task_struct to control the effect
of PTRACE_ATTACHED on credential calculations, particularly with respect to
execve().
With this patch, the contents of an active credentials struct may not be
changed directly; rather a new set of credentials must be prepared, modified
and committed using something like the following sequence of events:
struct cred *new = prepare_creds();
int ret = blah(new);
if (ret < 0) {
abort_creds(new);
return ret;
}
return commit_creds(new);
There are some exceptions to this rule: the keyrings pointed to by the active
credentials may be instantiated - keyrings violate the COW rule as managing
COW keyrings is tricky, given that it is possible for a task to directly alter
the keys in a keyring in use by another task.
To help enforce this, various pointers to sets of credentials, such as those in
the task_struct, are declared const. The purpose of this is compile-time
discouragement of altering credentials through those pointers. Once a set of
credentials has been made public through one of these pointers, it may not be
modified, except under special circumstances:
(1) Its reference count may incremented and decremented.
(2) The keyrings to which it points may be modified, but not replaced.
The only safe way to modify anything else is to create a replacement and commit
using the functions described in Documentation/credentials.txt (which will be
added by a later patch).
This patch and the preceding patches have been tested with the LTP SELinux
testsuite.
This patch makes several logical sets of alteration:
(1) execve().
This now prepares and commits credentials in various places in the
security code rather than altering the current creds directly.
(2) Temporary credential overrides.
do_coredump() and sys_faccessat() now prepare their own credentials and
temporarily override the ones currently on the acting thread, whilst
preventing interference from other threads by holding cred_replace_mutex
on the thread being dumped.
This will be replaced in a future patch by something that hands down the
credentials directly to the functions being called, rather than altering
the task's objective credentials.
(3) LSM interface.
A number of functions have been changed, added or removed:
(*) security_capset_check(), ->capset_check()
(*) security_capset_set(), ->capset_set()
Removed in favour of security_capset().
(*) security_capset(), ->capset()
New. This is passed a pointer to the new creds, a pointer to the old
creds and the proposed capability sets. It should fill in the new
creds or return an error. All pointers, barring the pointer to the
new creds, are now const.
(*) security_bprm_apply_creds(), ->bprm_apply_creds()
Changed; now returns a value, which will cause the process to be
killed if it's an error.
(*) security_task_alloc(), ->task_alloc_security()
Removed in favour of security_prepare_creds().
(*) security_cred_free(), ->cred_free()
New. Free security data attached to cred->security.
(*) security_prepare_creds(), ->cred_prepare()
New. Duplicate any security data attached to cred->security.
(*) security_commit_creds(), ->cred_commit()
New. Apply any security effects for the upcoming installation of new
security by commit_creds().
(*) security_task_post_setuid(), ->task_post_setuid()
Removed in favour of security_task_fix_setuid().
(*) security_task_fix_setuid(), ->task_fix_setuid()
Fix up the proposed new credentials for setuid(). This is used by
cap_set_fix_setuid() to implicitly adjust capabilities in line with
setuid() changes. Changes are made to the new credentials, rather
than the task itself as in security_task_post_setuid().
(*) security_task_reparent_to_init(), ->task_reparent_to_init()
Removed. Instead the task being reparented to init is referred
directly to init's credentials.
NOTE! This results in the loss of some state: SELinux's osid no
longer records the sid of the thread that forked it.
(*) security_key_alloc(), ->key_alloc()
(*) security_key_permission(), ->key_permission()
Changed. These now take cred pointers rather than task pointers to
refer to the security context.
(4) sys_capset().
This has been simplified and uses less locking. The LSM functions it
calls have been merged.
(5) reparent_to_kthreadd().
This gives the current thread the same credentials as init by simply using
commit_thread() to point that way.
(6) __sigqueue_alloc() and switch_uid()
__sigqueue_alloc() can't stop the target task from changing its creds
beneath it, so this function gets a reference to the currently applicable
user_struct which it then passes into the sigqueue struct it returns if
successful.
switch_uid() is now called from commit_creds(), and possibly should be
folded into that. commit_creds() should take care of protecting
__sigqueue_alloc().
(7) [sg]et[ug]id() and co and [sg]et_current_groups.
The set functions now all use prepare_creds(), commit_creds() and
abort_creds() to build and check a new set of credentials before applying
it.
security_task_set[ug]id() is called inside the prepared section. This
guarantees that nothing else will affect the creds until we've finished.
The calling of set_dumpable() has been moved into commit_creds().
Much of the functionality of set_user() has been moved into
commit_creds().
The get functions all simply access the data directly.
(8) security_task_prctl() and cap_task_prctl().
security_task_prctl() has been modified to return -ENOSYS if it doesn't
want to handle a function, or otherwise return the return value directly
rather than through an argument.
Additionally, cap_task_prctl() now prepares a new set of credentials, even
if it doesn't end up using it.
(9) Keyrings.
A number of changes have been made to the keyrings code:
(a) switch_uid_keyring(), copy_keys(), exit_keys() and suid_keys() have
all been dropped and built in to the credentials functions directly.
They may want separating out again later.
(b) key_alloc() and search_process_keyrings() now take a cred pointer
rather than a task pointer to specify the security context.
(c) copy_creds() gives a new thread within the same thread group a new
thread keyring if its parent had one, otherwise it discards the thread
keyring.
(d) The authorisation key now points directly to the credentials to extend
the search into rather pointing to the task that carries them.
(e) Installing thread, process or session keyrings causes a new set of
credentials to be created, even though it's not strictly necessary for
process or session keyrings (they're shared).
(10) Usermode helper.
The usermode helper code now carries a cred struct pointer in its
subprocess_info struct instead of a new session keyring pointer. This set
of credentials is derived from init_cred and installed on the new process
after it has been cloned.
call_usermodehelper_setup() allocates the new credentials and
call_usermodehelper_freeinfo() discards them if they haven't been used. A
special cred function (prepare_usermodeinfo_creds()) is provided
specifically for call_usermodehelper_setup() to call.
call_usermodehelper_setkeys() adjusts the credentials to sport the
supplied keyring as the new session keyring.
(11) SELinux.
SELinux has a number of changes, in addition to those to support the LSM
interface changes mentioned above:
(a) selinux_setprocattr() no longer does its check for whether the
current ptracer can access processes with the new SID inside the lock
that covers getting the ptracer's SID. Whilst this lock ensures that
the check is done with the ptracer pinned, the result is only valid
until the lock is released, so there's no point doing it inside the
lock.
(12) is_single_threaded().
This function has been extracted from selinux_setprocattr() and put into
a file of its own in the lib/ directory as join_session_keyring() now
wants to use it too.
The code in SELinux just checked to see whether a task shared mm_structs
with other tasks (CLONE_VM), but that isn't good enough. We really want
to know if they're part of the same thread group (CLONE_THREAD).
(13) nfsd.
The NFS server daemon now has to use the COW credentials to set the
credentials it is going to use. It really needs to pass the credentials
down to the functions it calls, but it can't do that until other patches
in this series have been applied.
Signed-off-by: David Howells <dhowells@redhat.com>
Acked-by: James Morris <jmorris@namei.org>
Signed-off-by: James Morris <jmorris@namei.org>
2008-11-14 07:39:23 +08:00
|
|
|
extern int __audit_log_bprm_fcaps(struct linux_binprm *bprm,
|
|
|
|
const struct cred *new,
|
|
|
|
const struct cred *old);
|
|
|
|
extern int __audit_log_capset(pid_t pid, const struct cred *new, const struct cred *old);
|
2006-05-19 04:01:30 +08:00
|
|
|
|
2008-12-10 16:40:06 +08:00
|
|
|
static inline void audit_ipc_obj(struct kern_ipc_perm *ipcp)
|
2006-05-19 04:01:30 +08:00
|
|
|
{
|
2006-07-16 18:38:45 +08:00
|
|
|
if (unlikely(!audit_dummy_context()))
|
2008-12-10 16:40:06 +08:00
|
|
|
__audit_ipc_obj(ipcp);
|
2006-05-19 04:01:30 +08:00
|
|
|
}
|
2007-02-07 14:48:00 +08:00
|
|
|
static inline int audit_fd_pair(int fd1, int fd2)
|
|
|
|
{
|
|
|
|
if (unlikely(!audit_dummy_context()))
|
|
|
|
return __audit_fd_pair(fd1, fd2);
|
|
|
|
return 0;
|
|
|
|
}
|
2008-12-10 16:47:15 +08:00
|
|
|
static inline void audit_ipc_set_perm(unsigned long qbytes, uid_t uid, gid_t gid, mode_t mode)
|
2006-05-19 04:01:30 +08:00
|
|
|
{
|
2006-07-16 18:38:45 +08:00
|
|
|
if (unlikely(!audit_dummy_context()))
|
2008-12-10 16:47:15 +08:00
|
|
|
__audit_ipc_set_perm(qbytes, uid, gid, mode);
|
2006-05-19 04:01:30 +08:00
|
|
|
}
|
2006-05-25 05:09:55 +08:00
|
|
|
static inline int audit_mq_open(int oflag, mode_t mode, struct mq_attr __user *u_attr)
|
|
|
|
{
|
2006-07-16 18:38:45 +08:00
|
|
|
if (unlikely(!audit_dummy_context()))
|
2006-05-25 05:09:55 +08:00
|
|
|
return __audit_mq_open(oflag, mode, u_attr);
|
|
|
|
return 0;
|
|
|
|
}
|
2008-12-14 16:46:48 +08:00
|
|
|
static inline void audit_mq_sendrecv(mqd_t mqdes, size_t msg_len, unsigned int msg_prio, const struct timespec *abs_timeout)
|
2006-05-25 05:09:55 +08:00
|
|
|
{
|
2006-07-16 18:38:45 +08:00
|
|
|
if (unlikely(!audit_dummy_context()))
|
2008-12-14 16:46:48 +08:00
|
|
|
__audit_mq_sendrecv(mqdes, msg_len, msg_prio, abs_timeout);
|
2006-05-25 05:09:55 +08:00
|
|
|
}
|
2008-12-10 20:16:12 +08:00
|
|
|
static inline void audit_mq_notify(mqd_t mqdes, const struct sigevent *notification)
|
2006-05-25 05:09:55 +08:00
|
|
|
{
|
2006-07-16 18:38:45 +08:00
|
|
|
if (unlikely(!audit_dummy_context()))
|
2008-12-10 20:16:12 +08:00
|
|
|
__audit_mq_notify(mqdes, notification);
|
2006-05-25 05:09:55 +08:00
|
|
|
}
|
2008-12-10 19:58:59 +08:00
|
|
|
static inline void audit_mq_getsetattr(mqd_t mqdes, struct mq_attr *mqstat)
|
2006-05-25 05:09:55 +08:00
|
|
|
{
|
2006-07-16 18:38:45 +08:00
|
|
|
if (unlikely(!audit_dummy_context()))
|
2008-12-10 19:58:59 +08:00
|
|
|
__audit_mq_getsetattr(mqdes, mqstat);
|
2006-05-25 05:09:55 +08:00
|
|
|
}
|
2008-11-11 18:48:18 +08:00
|
|
|
|
CRED: Inaugurate COW credentials
Inaugurate copy-on-write credentials management. This uses RCU to manage the
credentials pointer in the task_struct with respect to accesses by other tasks.
A process may only modify its own credentials, and so does not need locking to
access or modify its own credentials.
A mutex (cred_replace_mutex) is added to the task_struct to control the effect
of PTRACE_ATTACHED on credential calculations, particularly with respect to
execve().
With this patch, the contents of an active credentials struct may not be
changed directly; rather a new set of credentials must be prepared, modified
and committed using something like the following sequence of events:
struct cred *new = prepare_creds();
int ret = blah(new);
if (ret < 0) {
abort_creds(new);
return ret;
}
return commit_creds(new);
There are some exceptions to this rule: the keyrings pointed to by the active
credentials may be instantiated - keyrings violate the COW rule as managing
COW keyrings is tricky, given that it is possible for a task to directly alter
the keys in a keyring in use by another task.
To help enforce this, various pointers to sets of credentials, such as those in
the task_struct, are declared const. The purpose of this is compile-time
discouragement of altering credentials through those pointers. Once a set of
credentials has been made public through one of these pointers, it may not be
modified, except under special circumstances:
(1) Its reference count may incremented and decremented.
(2) The keyrings to which it points may be modified, but not replaced.
The only safe way to modify anything else is to create a replacement and commit
using the functions described in Documentation/credentials.txt (which will be
added by a later patch).
This patch and the preceding patches have been tested with the LTP SELinux
testsuite.
This patch makes several logical sets of alteration:
(1) execve().
This now prepares and commits credentials in various places in the
security code rather than altering the current creds directly.
(2) Temporary credential overrides.
do_coredump() and sys_faccessat() now prepare their own credentials and
temporarily override the ones currently on the acting thread, whilst
preventing interference from other threads by holding cred_replace_mutex
on the thread being dumped.
This will be replaced in a future patch by something that hands down the
credentials directly to the functions being called, rather than altering
the task's objective credentials.
(3) LSM interface.
A number of functions have been changed, added or removed:
(*) security_capset_check(), ->capset_check()
(*) security_capset_set(), ->capset_set()
Removed in favour of security_capset().
(*) security_capset(), ->capset()
New. This is passed a pointer to the new creds, a pointer to the old
creds and the proposed capability sets. It should fill in the new
creds or return an error. All pointers, barring the pointer to the
new creds, are now const.
(*) security_bprm_apply_creds(), ->bprm_apply_creds()
Changed; now returns a value, which will cause the process to be
killed if it's an error.
(*) security_task_alloc(), ->task_alloc_security()
Removed in favour of security_prepare_creds().
(*) security_cred_free(), ->cred_free()
New. Free security data attached to cred->security.
(*) security_prepare_creds(), ->cred_prepare()
New. Duplicate any security data attached to cred->security.
(*) security_commit_creds(), ->cred_commit()
New. Apply any security effects for the upcoming installation of new
security by commit_creds().
(*) security_task_post_setuid(), ->task_post_setuid()
Removed in favour of security_task_fix_setuid().
(*) security_task_fix_setuid(), ->task_fix_setuid()
Fix up the proposed new credentials for setuid(). This is used by
cap_set_fix_setuid() to implicitly adjust capabilities in line with
setuid() changes. Changes are made to the new credentials, rather
than the task itself as in security_task_post_setuid().
(*) security_task_reparent_to_init(), ->task_reparent_to_init()
Removed. Instead the task being reparented to init is referred
directly to init's credentials.
NOTE! This results in the loss of some state: SELinux's osid no
longer records the sid of the thread that forked it.
(*) security_key_alloc(), ->key_alloc()
(*) security_key_permission(), ->key_permission()
Changed. These now take cred pointers rather than task pointers to
refer to the security context.
(4) sys_capset().
This has been simplified and uses less locking. The LSM functions it
calls have been merged.
(5) reparent_to_kthreadd().
This gives the current thread the same credentials as init by simply using
commit_thread() to point that way.
(6) __sigqueue_alloc() and switch_uid()
__sigqueue_alloc() can't stop the target task from changing its creds
beneath it, so this function gets a reference to the currently applicable
user_struct which it then passes into the sigqueue struct it returns if
successful.
switch_uid() is now called from commit_creds(), and possibly should be
folded into that. commit_creds() should take care of protecting
__sigqueue_alloc().
(7) [sg]et[ug]id() and co and [sg]et_current_groups.
The set functions now all use prepare_creds(), commit_creds() and
abort_creds() to build and check a new set of credentials before applying
it.
security_task_set[ug]id() is called inside the prepared section. This
guarantees that nothing else will affect the creds until we've finished.
The calling of set_dumpable() has been moved into commit_creds().
Much of the functionality of set_user() has been moved into
commit_creds().
The get functions all simply access the data directly.
(8) security_task_prctl() and cap_task_prctl().
security_task_prctl() has been modified to return -ENOSYS if it doesn't
want to handle a function, or otherwise return the return value directly
rather than through an argument.
Additionally, cap_task_prctl() now prepares a new set of credentials, even
if it doesn't end up using it.
(9) Keyrings.
A number of changes have been made to the keyrings code:
(a) switch_uid_keyring(), copy_keys(), exit_keys() and suid_keys() have
all been dropped and built in to the credentials functions directly.
They may want separating out again later.
(b) key_alloc() and search_process_keyrings() now take a cred pointer
rather than a task pointer to specify the security context.
(c) copy_creds() gives a new thread within the same thread group a new
thread keyring if its parent had one, otherwise it discards the thread
keyring.
(d) The authorisation key now points directly to the credentials to extend
the search into rather pointing to the task that carries them.
(e) Installing thread, process or session keyrings causes a new set of
credentials to be created, even though it's not strictly necessary for
process or session keyrings (they're shared).
(10) Usermode helper.
The usermode helper code now carries a cred struct pointer in its
subprocess_info struct instead of a new session keyring pointer. This set
of credentials is derived from init_cred and installed on the new process
after it has been cloned.
call_usermodehelper_setup() allocates the new credentials and
call_usermodehelper_freeinfo() discards them if they haven't been used. A
special cred function (prepare_usermodeinfo_creds()) is provided
specifically for call_usermodehelper_setup() to call.
call_usermodehelper_setkeys() adjusts the credentials to sport the
supplied keyring as the new session keyring.
(11) SELinux.
SELinux has a number of changes, in addition to those to support the LSM
interface changes mentioned above:
(a) selinux_setprocattr() no longer does its check for whether the
current ptracer can access processes with the new SID inside the lock
that covers getting the ptracer's SID. Whilst this lock ensures that
the check is done with the ptracer pinned, the result is only valid
until the lock is released, so there's no point doing it inside the
lock.
(12) is_single_threaded().
This function has been extracted from selinux_setprocattr() and put into
a file of its own in the lib/ directory as join_session_keyring() now
wants to use it too.
The code in SELinux just checked to see whether a task shared mm_structs
with other tasks (CLONE_VM), but that isn't good enough. We really want
to know if they're part of the same thread group (CLONE_THREAD).
(13) nfsd.
The NFS server daemon now has to use the COW credentials to set the
credentials it is going to use. It really needs to pass the credentials
down to the functions it calls, but it can't do that until other patches
in this series have been applied.
Signed-off-by: David Howells <dhowells@redhat.com>
Acked-by: James Morris <jmorris@namei.org>
Signed-off-by: James Morris <jmorris@namei.org>
2008-11-14 07:39:23 +08:00
|
|
|
static inline int audit_log_bprm_fcaps(struct linux_binprm *bprm,
|
|
|
|
const struct cred *new,
|
|
|
|
const struct cred *old)
|
2008-11-11 18:48:18 +08:00
|
|
|
{
|
|
|
|
if (unlikely(!audit_dummy_context()))
|
CRED: Inaugurate COW credentials
Inaugurate copy-on-write credentials management. This uses RCU to manage the
credentials pointer in the task_struct with respect to accesses by other tasks.
A process may only modify its own credentials, and so does not need locking to
access or modify its own credentials.
A mutex (cred_replace_mutex) is added to the task_struct to control the effect
of PTRACE_ATTACHED on credential calculations, particularly with respect to
execve().
With this patch, the contents of an active credentials struct may not be
changed directly; rather a new set of credentials must be prepared, modified
and committed using something like the following sequence of events:
struct cred *new = prepare_creds();
int ret = blah(new);
if (ret < 0) {
abort_creds(new);
return ret;
}
return commit_creds(new);
There are some exceptions to this rule: the keyrings pointed to by the active
credentials may be instantiated - keyrings violate the COW rule as managing
COW keyrings is tricky, given that it is possible for a task to directly alter
the keys in a keyring in use by another task.
To help enforce this, various pointers to sets of credentials, such as those in
the task_struct, are declared const. The purpose of this is compile-time
discouragement of altering credentials through those pointers. Once a set of
credentials has been made public through one of these pointers, it may not be
modified, except under special circumstances:
(1) Its reference count may incremented and decremented.
(2) The keyrings to which it points may be modified, but not replaced.
The only safe way to modify anything else is to create a replacement and commit
using the functions described in Documentation/credentials.txt (which will be
added by a later patch).
This patch and the preceding patches have been tested with the LTP SELinux
testsuite.
This patch makes several logical sets of alteration:
(1) execve().
This now prepares and commits credentials in various places in the
security code rather than altering the current creds directly.
(2) Temporary credential overrides.
do_coredump() and sys_faccessat() now prepare their own credentials and
temporarily override the ones currently on the acting thread, whilst
preventing interference from other threads by holding cred_replace_mutex
on the thread being dumped.
This will be replaced in a future patch by something that hands down the
credentials directly to the functions being called, rather than altering
the task's objective credentials.
(3) LSM interface.
A number of functions have been changed, added or removed:
(*) security_capset_check(), ->capset_check()
(*) security_capset_set(), ->capset_set()
Removed in favour of security_capset().
(*) security_capset(), ->capset()
New. This is passed a pointer to the new creds, a pointer to the old
creds and the proposed capability sets. It should fill in the new
creds or return an error. All pointers, barring the pointer to the
new creds, are now const.
(*) security_bprm_apply_creds(), ->bprm_apply_creds()
Changed; now returns a value, which will cause the process to be
killed if it's an error.
(*) security_task_alloc(), ->task_alloc_security()
Removed in favour of security_prepare_creds().
(*) security_cred_free(), ->cred_free()
New. Free security data attached to cred->security.
(*) security_prepare_creds(), ->cred_prepare()
New. Duplicate any security data attached to cred->security.
(*) security_commit_creds(), ->cred_commit()
New. Apply any security effects for the upcoming installation of new
security by commit_creds().
(*) security_task_post_setuid(), ->task_post_setuid()
Removed in favour of security_task_fix_setuid().
(*) security_task_fix_setuid(), ->task_fix_setuid()
Fix up the proposed new credentials for setuid(). This is used by
cap_set_fix_setuid() to implicitly adjust capabilities in line with
setuid() changes. Changes are made to the new credentials, rather
than the task itself as in security_task_post_setuid().
(*) security_task_reparent_to_init(), ->task_reparent_to_init()
Removed. Instead the task being reparented to init is referred
directly to init's credentials.
NOTE! This results in the loss of some state: SELinux's osid no
longer records the sid of the thread that forked it.
(*) security_key_alloc(), ->key_alloc()
(*) security_key_permission(), ->key_permission()
Changed. These now take cred pointers rather than task pointers to
refer to the security context.
(4) sys_capset().
This has been simplified and uses less locking. The LSM functions it
calls have been merged.
(5) reparent_to_kthreadd().
This gives the current thread the same credentials as init by simply using
commit_thread() to point that way.
(6) __sigqueue_alloc() and switch_uid()
__sigqueue_alloc() can't stop the target task from changing its creds
beneath it, so this function gets a reference to the currently applicable
user_struct which it then passes into the sigqueue struct it returns if
successful.
switch_uid() is now called from commit_creds(), and possibly should be
folded into that. commit_creds() should take care of protecting
__sigqueue_alloc().
(7) [sg]et[ug]id() and co and [sg]et_current_groups.
The set functions now all use prepare_creds(), commit_creds() and
abort_creds() to build and check a new set of credentials before applying
it.
security_task_set[ug]id() is called inside the prepared section. This
guarantees that nothing else will affect the creds until we've finished.
The calling of set_dumpable() has been moved into commit_creds().
Much of the functionality of set_user() has been moved into
commit_creds().
The get functions all simply access the data directly.
(8) security_task_prctl() and cap_task_prctl().
security_task_prctl() has been modified to return -ENOSYS if it doesn't
want to handle a function, or otherwise return the return value directly
rather than through an argument.
Additionally, cap_task_prctl() now prepares a new set of credentials, even
if it doesn't end up using it.
(9) Keyrings.
A number of changes have been made to the keyrings code:
(a) switch_uid_keyring(), copy_keys(), exit_keys() and suid_keys() have
all been dropped and built in to the credentials functions directly.
They may want separating out again later.
(b) key_alloc() and search_process_keyrings() now take a cred pointer
rather than a task pointer to specify the security context.
(c) copy_creds() gives a new thread within the same thread group a new
thread keyring if its parent had one, otherwise it discards the thread
keyring.
(d) The authorisation key now points directly to the credentials to extend
the search into rather pointing to the task that carries them.
(e) Installing thread, process or session keyrings causes a new set of
credentials to be created, even though it's not strictly necessary for
process or session keyrings (they're shared).
(10) Usermode helper.
The usermode helper code now carries a cred struct pointer in its
subprocess_info struct instead of a new session keyring pointer. This set
of credentials is derived from init_cred and installed on the new process
after it has been cloned.
call_usermodehelper_setup() allocates the new credentials and
call_usermodehelper_freeinfo() discards them if they haven't been used. A
special cred function (prepare_usermodeinfo_creds()) is provided
specifically for call_usermodehelper_setup() to call.
call_usermodehelper_setkeys() adjusts the credentials to sport the
supplied keyring as the new session keyring.
(11) SELinux.
SELinux has a number of changes, in addition to those to support the LSM
interface changes mentioned above:
(a) selinux_setprocattr() no longer does its check for whether the
current ptracer can access processes with the new SID inside the lock
that covers getting the ptracer's SID. Whilst this lock ensures that
the check is done with the ptracer pinned, the result is only valid
until the lock is released, so there's no point doing it inside the
lock.
(12) is_single_threaded().
This function has been extracted from selinux_setprocattr() and put into
a file of its own in the lib/ directory as join_session_keyring() now
wants to use it too.
The code in SELinux just checked to see whether a task shared mm_structs
with other tasks (CLONE_VM), but that isn't good enough. We really want
to know if they're part of the same thread group (CLONE_THREAD).
(13) nfsd.
The NFS server daemon now has to use the COW credentials to set the
credentials it is going to use. It really needs to pass the credentials
down to the functions it calls, but it can't do that until other patches
in this series have been applied.
Signed-off-by: David Howells <dhowells@redhat.com>
Acked-by: James Morris <jmorris@namei.org>
Signed-off-by: James Morris <jmorris@namei.org>
2008-11-14 07:39:23 +08:00
|
|
|
return __audit_log_bprm_fcaps(bprm, new, old);
|
|
|
|
return 0;
|
2008-11-11 18:48:18 +08:00
|
|
|
}
|
|
|
|
|
CRED: Inaugurate COW credentials
Inaugurate copy-on-write credentials management. This uses RCU to manage the
credentials pointer in the task_struct with respect to accesses by other tasks.
A process may only modify its own credentials, and so does not need locking to
access or modify its own credentials.
A mutex (cred_replace_mutex) is added to the task_struct to control the effect
of PTRACE_ATTACHED on credential calculations, particularly with respect to
execve().
With this patch, the contents of an active credentials struct may not be
changed directly; rather a new set of credentials must be prepared, modified
and committed using something like the following sequence of events:
struct cred *new = prepare_creds();
int ret = blah(new);
if (ret < 0) {
abort_creds(new);
return ret;
}
return commit_creds(new);
There are some exceptions to this rule: the keyrings pointed to by the active
credentials may be instantiated - keyrings violate the COW rule as managing
COW keyrings is tricky, given that it is possible for a task to directly alter
the keys in a keyring in use by another task.
To help enforce this, various pointers to sets of credentials, such as those in
the task_struct, are declared const. The purpose of this is compile-time
discouragement of altering credentials through those pointers. Once a set of
credentials has been made public through one of these pointers, it may not be
modified, except under special circumstances:
(1) Its reference count may incremented and decremented.
(2) The keyrings to which it points may be modified, but not replaced.
The only safe way to modify anything else is to create a replacement and commit
using the functions described in Documentation/credentials.txt (which will be
added by a later patch).
This patch and the preceding patches have been tested with the LTP SELinux
testsuite.
This patch makes several logical sets of alteration:
(1) execve().
This now prepares and commits credentials in various places in the
security code rather than altering the current creds directly.
(2) Temporary credential overrides.
do_coredump() and sys_faccessat() now prepare their own credentials and
temporarily override the ones currently on the acting thread, whilst
preventing interference from other threads by holding cred_replace_mutex
on the thread being dumped.
This will be replaced in a future patch by something that hands down the
credentials directly to the functions being called, rather than altering
the task's objective credentials.
(3) LSM interface.
A number of functions have been changed, added or removed:
(*) security_capset_check(), ->capset_check()
(*) security_capset_set(), ->capset_set()
Removed in favour of security_capset().
(*) security_capset(), ->capset()
New. This is passed a pointer to the new creds, a pointer to the old
creds and the proposed capability sets. It should fill in the new
creds or return an error. All pointers, barring the pointer to the
new creds, are now const.
(*) security_bprm_apply_creds(), ->bprm_apply_creds()
Changed; now returns a value, which will cause the process to be
killed if it's an error.
(*) security_task_alloc(), ->task_alloc_security()
Removed in favour of security_prepare_creds().
(*) security_cred_free(), ->cred_free()
New. Free security data attached to cred->security.
(*) security_prepare_creds(), ->cred_prepare()
New. Duplicate any security data attached to cred->security.
(*) security_commit_creds(), ->cred_commit()
New. Apply any security effects for the upcoming installation of new
security by commit_creds().
(*) security_task_post_setuid(), ->task_post_setuid()
Removed in favour of security_task_fix_setuid().
(*) security_task_fix_setuid(), ->task_fix_setuid()
Fix up the proposed new credentials for setuid(). This is used by
cap_set_fix_setuid() to implicitly adjust capabilities in line with
setuid() changes. Changes are made to the new credentials, rather
than the task itself as in security_task_post_setuid().
(*) security_task_reparent_to_init(), ->task_reparent_to_init()
Removed. Instead the task being reparented to init is referred
directly to init's credentials.
NOTE! This results in the loss of some state: SELinux's osid no
longer records the sid of the thread that forked it.
(*) security_key_alloc(), ->key_alloc()
(*) security_key_permission(), ->key_permission()
Changed. These now take cred pointers rather than task pointers to
refer to the security context.
(4) sys_capset().
This has been simplified and uses less locking. The LSM functions it
calls have been merged.
(5) reparent_to_kthreadd().
This gives the current thread the same credentials as init by simply using
commit_thread() to point that way.
(6) __sigqueue_alloc() and switch_uid()
__sigqueue_alloc() can't stop the target task from changing its creds
beneath it, so this function gets a reference to the currently applicable
user_struct which it then passes into the sigqueue struct it returns if
successful.
switch_uid() is now called from commit_creds(), and possibly should be
folded into that. commit_creds() should take care of protecting
__sigqueue_alloc().
(7) [sg]et[ug]id() and co and [sg]et_current_groups.
The set functions now all use prepare_creds(), commit_creds() and
abort_creds() to build and check a new set of credentials before applying
it.
security_task_set[ug]id() is called inside the prepared section. This
guarantees that nothing else will affect the creds until we've finished.
The calling of set_dumpable() has been moved into commit_creds().
Much of the functionality of set_user() has been moved into
commit_creds().
The get functions all simply access the data directly.
(8) security_task_prctl() and cap_task_prctl().
security_task_prctl() has been modified to return -ENOSYS if it doesn't
want to handle a function, or otherwise return the return value directly
rather than through an argument.
Additionally, cap_task_prctl() now prepares a new set of credentials, even
if it doesn't end up using it.
(9) Keyrings.
A number of changes have been made to the keyrings code:
(a) switch_uid_keyring(), copy_keys(), exit_keys() and suid_keys() have
all been dropped and built in to the credentials functions directly.
They may want separating out again later.
(b) key_alloc() and search_process_keyrings() now take a cred pointer
rather than a task pointer to specify the security context.
(c) copy_creds() gives a new thread within the same thread group a new
thread keyring if its parent had one, otherwise it discards the thread
keyring.
(d) The authorisation key now points directly to the credentials to extend
the search into rather pointing to the task that carries them.
(e) Installing thread, process or session keyrings causes a new set of
credentials to be created, even though it's not strictly necessary for
process or session keyrings (they're shared).
(10) Usermode helper.
The usermode helper code now carries a cred struct pointer in its
subprocess_info struct instead of a new session keyring pointer. This set
of credentials is derived from init_cred and installed on the new process
after it has been cloned.
call_usermodehelper_setup() allocates the new credentials and
call_usermodehelper_freeinfo() discards them if they haven't been used. A
special cred function (prepare_usermodeinfo_creds()) is provided
specifically for call_usermodehelper_setup() to call.
call_usermodehelper_setkeys() adjusts the credentials to sport the
supplied keyring as the new session keyring.
(11) SELinux.
SELinux has a number of changes, in addition to those to support the LSM
interface changes mentioned above:
(a) selinux_setprocattr() no longer does its check for whether the
current ptracer can access processes with the new SID inside the lock
that covers getting the ptracer's SID. Whilst this lock ensures that
the check is done with the ptracer pinned, the result is only valid
until the lock is released, so there's no point doing it inside the
lock.
(12) is_single_threaded().
This function has been extracted from selinux_setprocattr() and put into
a file of its own in the lib/ directory as join_session_keyring() now
wants to use it too.
The code in SELinux just checked to see whether a task shared mm_structs
with other tasks (CLONE_VM), but that isn't good enough. We really want
to know if they're part of the same thread group (CLONE_THREAD).
(13) nfsd.
The NFS server daemon now has to use the COW credentials to set the
credentials it is going to use. It really needs to pass the credentials
down to the functions it calls, but it can't do that until other patches
in this series have been applied.
Signed-off-by: David Howells <dhowells@redhat.com>
Acked-by: James Morris <jmorris@namei.org>
Signed-off-by: James Morris <jmorris@namei.org>
2008-11-14 07:39:23 +08:00
|
|
|
static inline int audit_log_capset(pid_t pid, const struct cred *new,
|
|
|
|
const struct cred *old)
|
2008-11-11 18:48:22 +08:00
|
|
|
{
|
|
|
|
if (unlikely(!audit_dummy_context()))
|
CRED: Inaugurate COW credentials
Inaugurate copy-on-write credentials management. This uses RCU to manage the
credentials pointer in the task_struct with respect to accesses by other tasks.
A process may only modify its own credentials, and so does not need locking to
access or modify its own credentials.
A mutex (cred_replace_mutex) is added to the task_struct to control the effect
of PTRACE_ATTACHED on credential calculations, particularly with respect to
execve().
With this patch, the contents of an active credentials struct may not be
changed directly; rather a new set of credentials must be prepared, modified
and committed using something like the following sequence of events:
struct cred *new = prepare_creds();
int ret = blah(new);
if (ret < 0) {
abort_creds(new);
return ret;
}
return commit_creds(new);
There are some exceptions to this rule: the keyrings pointed to by the active
credentials may be instantiated - keyrings violate the COW rule as managing
COW keyrings is tricky, given that it is possible for a task to directly alter
the keys in a keyring in use by another task.
To help enforce this, various pointers to sets of credentials, such as those in
the task_struct, are declared const. The purpose of this is compile-time
discouragement of altering credentials through those pointers. Once a set of
credentials has been made public through one of these pointers, it may not be
modified, except under special circumstances:
(1) Its reference count may incremented and decremented.
(2) The keyrings to which it points may be modified, but not replaced.
The only safe way to modify anything else is to create a replacement and commit
using the functions described in Documentation/credentials.txt (which will be
added by a later patch).
This patch and the preceding patches have been tested with the LTP SELinux
testsuite.
This patch makes several logical sets of alteration:
(1) execve().
This now prepares and commits credentials in various places in the
security code rather than altering the current creds directly.
(2) Temporary credential overrides.
do_coredump() and sys_faccessat() now prepare their own credentials and
temporarily override the ones currently on the acting thread, whilst
preventing interference from other threads by holding cred_replace_mutex
on the thread being dumped.
This will be replaced in a future patch by something that hands down the
credentials directly to the functions being called, rather than altering
the task's objective credentials.
(3) LSM interface.
A number of functions have been changed, added or removed:
(*) security_capset_check(), ->capset_check()
(*) security_capset_set(), ->capset_set()
Removed in favour of security_capset().
(*) security_capset(), ->capset()
New. This is passed a pointer to the new creds, a pointer to the old
creds and the proposed capability sets. It should fill in the new
creds or return an error. All pointers, barring the pointer to the
new creds, are now const.
(*) security_bprm_apply_creds(), ->bprm_apply_creds()
Changed; now returns a value, which will cause the process to be
killed if it's an error.
(*) security_task_alloc(), ->task_alloc_security()
Removed in favour of security_prepare_creds().
(*) security_cred_free(), ->cred_free()
New. Free security data attached to cred->security.
(*) security_prepare_creds(), ->cred_prepare()
New. Duplicate any security data attached to cred->security.
(*) security_commit_creds(), ->cred_commit()
New. Apply any security effects for the upcoming installation of new
security by commit_creds().
(*) security_task_post_setuid(), ->task_post_setuid()
Removed in favour of security_task_fix_setuid().
(*) security_task_fix_setuid(), ->task_fix_setuid()
Fix up the proposed new credentials for setuid(). This is used by
cap_set_fix_setuid() to implicitly adjust capabilities in line with
setuid() changes. Changes are made to the new credentials, rather
than the task itself as in security_task_post_setuid().
(*) security_task_reparent_to_init(), ->task_reparent_to_init()
Removed. Instead the task being reparented to init is referred
directly to init's credentials.
NOTE! This results in the loss of some state: SELinux's osid no
longer records the sid of the thread that forked it.
(*) security_key_alloc(), ->key_alloc()
(*) security_key_permission(), ->key_permission()
Changed. These now take cred pointers rather than task pointers to
refer to the security context.
(4) sys_capset().
This has been simplified and uses less locking. The LSM functions it
calls have been merged.
(5) reparent_to_kthreadd().
This gives the current thread the same credentials as init by simply using
commit_thread() to point that way.
(6) __sigqueue_alloc() and switch_uid()
__sigqueue_alloc() can't stop the target task from changing its creds
beneath it, so this function gets a reference to the currently applicable
user_struct which it then passes into the sigqueue struct it returns if
successful.
switch_uid() is now called from commit_creds(), and possibly should be
folded into that. commit_creds() should take care of protecting
__sigqueue_alloc().
(7) [sg]et[ug]id() and co and [sg]et_current_groups.
The set functions now all use prepare_creds(), commit_creds() and
abort_creds() to build and check a new set of credentials before applying
it.
security_task_set[ug]id() is called inside the prepared section. This
guarantees that nothing else will affect the creds until we've finished.
The calling of set_dumpable() has been moved into commit_creds().
Much of the functionality of set_user() has been moved into
commit_creds().
The get functions all simply access the data directly.
(8) security_task_prctl() and cap_task_prctl().
security_task_prctl() has been modified to return -ENOSYS if it doesn't
want to handle a function, or otherwise return the return value directly
rather than through an argument.
Additionally, cap_task_prctl() now prepares a new set of credentials, even
if it doesn't end up using it.
(9) Keyrings.
A number of changes have been made to the keyrings code:
(a) switch_uid_keyring(), copy_keys(), exit_keys() and suid_keys() have
all been dropped and built in to the credentials functions directly.
They may want separating out again later.
(b) key_alloc() and search_process_keyrings() now take a cred pointer
rather than a task pointer to specify the security context.
(c) copy_creds() gives a new thread within the same thread group a new
thread keyring if its parent had one, otherwise it discards the thread
keyring.
(d) The authorisation key now points directly to the credentials to extend
the search into rather pointing to the task that carries them.
(e) Installing thread, process or session keyrings causes a new set of
credentials to be created, even though it's not strictly necessary for
process or session keyrings (they're shared).
(10) Usermode helper.
The usermode helper code now carries a cred struct pointer in its
subprocess_info struct instead of a new session keyring pointer. This set
of credentials is derived from init_cred and installed on the new process
after it has been cloned.
call_usermodehelper_setup() allocates the new credentials and
call_usermodehelper_freeinfo() discards them if they haven't been used. A
special cred function (prepare_usermodeinfo_creds()) is provided
specifically for call_usermodehelper_setup() to call.
call_usermodehelper_setkeys() adjusts the credentials to sport the
supplied keyring as the new session keyring.
(11) SELinux.
SELinux has a number of changes, in addition to those to support the LSM
interface changes mentioned above:
(a) selinux_setprocattr() no longer does its check for whether the
current ptracer can access processes with the new SID inside the lock
that covers getting the ptracer's SID. Whilst this lock ensures that
the check is done with the ptracer pinned, the result is only valid
until the lock is released, so there's no point doing it inside the
lock.
(12) is_single_threaded().
This function has been extracted from selinux_setprocattr() and put into
a file of its own in the lib/ directory as join_session_keyring() now
wants to use it too.
The code in SELinux just checked to see whether a task shared mm_structs
with other tasks (CLONE_VM), but that isn't good enough. We really want
to know if they're part of the same thread group (CLONE_THREAD).
(13) nfsd.
The NFS server daemon now has to use the COW credentials to set the
credentials it is going to use. It really needs to pass the credentials
down to the functions it calls, but it can't do that until other patches
in this series have been applied.
Signed-off-by: David Howells <dhowells@redhat.com>
Acked-by: James Morris <jmorris@namei.org>
Signed-off-by: James Morris <jmorris@namei.org>
2008-11-14 07:39:23 +08:00
|
|
|
return __audit_log_capset(pid, new, old);
|
2008-11-11 18:48:22 +08:00
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
2006-07-10 20:29:24 +08:00
|
|
|
extern int audit_n_rules;
|
2007-03-30 06:01:04 +08:00
|
|
|
extern int audit_signals;
|
2005-04-17 06:20:36 +08:00
|
|
|
#else
|
2008-11-13 07:37:41 +08:00
|
|
|
#define audit_finish_fork(t)
|
2005-04-17 06:20:36 +08:00
|
|
|
#define audit_alloc(t) ({ 0; })
|
|
|
|
#define audit_free(t) do { ; } while (0)
|
2006-03-30 09:23:36 +08:00
|
|
|
#define audit_syscall_entry(ta,a,b,c,d,e) do { ; } while (0)
|
|
|
|
#define audit_syscall_exit(f,r) do { ; } while (0)
|
2006-08-03 22:59:26 +08:00
|
|
|
#define audit_dummy_context() 1
|
2005-04-17 06:20:36 +08:00
|
|
|
#define audit_getname(n) do { ; } while (0)
|
|
|
|
#define audit_putname(n) do { ; } while (0)
|
2007-06-08 00:19:32 +08:00
|
|
|
#define __audit_inode(n,d) do { ; } while (0)
|
2005-11-04 00:00:25 +08:00
|
|
|
#define __audit_inode_child(d,i,p) do { ; } while (0)
|
2007-06-08 00:19:32 +08:00
|
|
|
#define audit_inode(n,d) do { ; } while (0)
|
2005-11-04 00:00:25 +08:00
|
|
|
#define audit_inode_child(d,i,p) do { ; } while (0)
|
2007-04-19 22:28:21 +08:00
|
|
|
#define audit_core_dumps(i) do { ; } while (0)
|
2008-12-06 14:05:50 +08:00
|
|
|
#define auditsc_get_stamp(c,t,s) (0)
|
2008-01-10 17:20:52 +08:00
|
|
|
#define audit_get_loginuid(t) (-1)
|
2008-01-08 23:06:53 +08:00
|
|
|
#define audit_get_sessionid(t) (-1)
|
2006-11-28 03:11:54 +08:00
|
|
|
#define audit_log_task_context(b) do { ; } while (0)
|
2008-12-10 16:40:06 +08:00
|
|
|
#define audit_ipc_obj(i) ((void)0)
|
2008-12-10 16:47:15 +08:00
|
|
|
#define audit_ipc_set_perm(q,u,g,m) ((void)0)
|
2006-04-27 02:04:08 +08:00
|
|
|
#define audit_bprm(p) ({ 0; })
|
2008-12-10 16:16:51 +08:00
|
|
|
#define audit_socketcall(n,a) ((void)0)
|
2007-02-07 14:48:00 +08:00
|
|
|
#define audit_fd_pair(n,a) ({ 0; })
|
2005-05-17 19:08:48 +08:00
|
|
|
#define audit_sockaddr(len, addr) ({ 0; })
|
2005-11-04 01:15:16 +08:00
|
|
|
#define audit_set_macxattr(n) do { ; } while (0)
|
2006-05-25 05:09:55 +08:00
|
|
|
#define audit_mq_open(o,m,a) ({ 0; })
|
2008-12-14 16:46:48 +08:00
|
|
|
#define audit_mq_sendrecv(d,l,p,t) ((void)0)
|
2008-12-10 20:16:12 +08:00
|
|
|
#define audit_mq_notify(d,n) ((void)0)
|
2008-12-10 19:58:59 +08:00
|
|
|
#define audit_mq_getsetattr(d,s) ((void)0)
|
CRED: Inaugurate COW credentials
Inaugurate copy-on-write credentials management. This uses RCU to manage the
credentials pointer in the task_struct with respect to accesses by other tasks.
A process may only modify its own credentials, and so does not need locking to
access or modify its own credentials.
A mutex (cred_replace_mutex) is added to the task_struct to control the effect
of PTRACE_ATTACHED on credential calculations, particularly with respect to
execve().
With this patch, the contents of an active credentials struct may not be
changed directly; rather a new set of credentials must be prepared, modified
and committed using something like the following sequence of events:
struct cred *new = prepare_creds();
int ret = blah(new);
if (ret < 0) {
abort_creds(new);
return ret;
}
return commit_creds(new);
There are some exceptions to this rule: the keyrings pointed to by the active
credentials may be instantiated - keyrings violate the COW rule as managing
COW keyrings is tricky, given that it is possible for a task to directly alter
the keys in a keyring in use by another task.
To help enforce this, various pointers to sets of credentials, such as those in
the task_struct, are declared const. The purpose of this is compile-time
discouragement of altering credentials through those pointers. Once a set of
credentials has been made public through one of these pointers, it may not be
modified, except under special circumstances:
(1) Its reference count may incremented and decremented.
(2) The keyrings to which it points may be modified, but not replaced.
The only safe way to modify anything else is to create a replacement and commit
using the functions described in Documentation/credentials.txt (which will be
added by a later patch).
This patch and the preceding patches have been tested with the LTP SELinux
testsuite.
This patch makes several logical sets of alteration:
(1) execve().
This now prepares and commits credentials in various places in the
security code rather than altering the current creds directly.
(2) Temporary credential overrides.
do_coredump() and sys_faccessat() now prepare their own credentials and
temporarily override the ones currently on the acting thread, whilst
preventing interference from other threads by holding cred_replace_mutex
on the thread being dumped.
This will be replaced in a future patch by something that hands down the
credentials directly to the functions being called, rather than altering
the task's objective credentials.
(3) LSM interface.
A number of functions have been changed, added or removed:
(*) security_capset_check(), ->capset_check()
(*) security_capset_set(), ->capset_set()
Removed in favour of security_capset().
(*) security_capset(), ->capset()
New. This is passed a pointer to the new creds, a pointer to the old
creds and the proposed capability sets. It should fill in the new
creds or return an error. All pointers, barring the pointer to the
new creds, are now const.
(*) security_bprm_apply_creds(), ->bprm_apply_creds()
Changed; now returns a value, which will cause the process to be
killed if it's an error.
(*) security_task_alloc(), ->task_alloc_security()
Removed in favour of security_prepare_creds().
(*) security_cred_free(), ->cred_free()
New. Free security data attached to cred->security.
(*) security_prepare_creds(), ->cred_prepare()
New. Duplicate any security data attached to cred->security.
(*) security_commit_creds(), ->cred_commit()
New. Apply any security effects for the upcoming installation of new
security by commit_creds().
(*) security_task_post_setuid(), ->task_post_setuid()
Removed in favour of security_task_fix_setuid().
(*) security_task_fix_setuid(), ->task_fix_setuid()
Fix up the proposed new credentials for setuid(). This is used by
cap_set_fix_setuid() to implicitly adjust capabilities in line with
setuid() changes. Changes are made to the new credentials, rather
than the task itself as in security_task_post_setuid().
(*) security_task_reparent_to_init(), ->task_reparent_to_init()
Removed. Instead the task being reparented to init is referred
directly to init's credentials.
NOTE! This results in the loss of some state: SELinux's osid no
longer records the sid of the thread that forked it.
(*) security_key_alloc(), ->key_alloc()
(*) security_key_permission(), ->key_permission()
Changed. These now take cred pointers rather than task pointers to
refer to the security context.
(4) sys_capset().
This has been simplified and uses less locking. The LSM functions it
calls have been merged.
(5) reparent_to_kthreadd().
This gives the current thread the same credentials as init by simply using
commit_thread() to point that way.
(6) __sigqueue_alloc() and switch_uid()
__sigqueue_alloc() can't stop the target task from changing its creds
beneath it, so this function gets a reference to the currently applicable
user_struct which it then passes into the sigqueue struct it returns if
successful.
switch_uid() is now called from commit_creds(), and possibly should be
folded into that. commit_creds() should take care of protecting
__sigqueue_alloc().
(7) [sg]et[ug]id() and co and [sg]et_current_groups.
The set functions now all use prepare_creds(), commit_creds() and
abort_creds() to build and check a new set of credentials before applying
it.
security_task_set[ug]id() is called inside the prepared section. This
guarantees that nothing else will affect the creds until we've finished.
The calling of set_dumpable() has been moved into commit_creds().
Much of the functionality of set_user() has been moved into
commit_creds().
The get functions all simply access the data directly.
(8) security_task_prctl() and cap_task_prctl().
security_task_prctl() has been modified to return -ENOSYS if it doesn't
want to handle a function, or otherwise return the return value directly
rather than through an argument.
Additionally, cap_task_prctl() now prepares a new set of credentials, even
if it doesn't end up using it.
(9) Keyrings.
A number of changes have been made to the keyrings code:
(a) switch_uid_keyring(), copy_keys(), exit_keys() and suid_keys() have
all been dropped and built in to the credentials functions directly.
They may want separating out again later.
(b) key_alloc() and search_process_keyrings() now take a cred pointer
rather than a task pointer to specify the security context.
(c) copy_creds() gives a new thread within the same thread group a new
thread keyring if its parent had one, otherwise it discards the thread
keyring.
(d) The authorisation key now points directly to the credentials to extend
the search into rather pointing to the task that carries them.
(e) Installing thread, process or session keyrings causes a new set of
credentials to be created, even though it's not strictly necessary for
process or session keyrings (they're shared).
(10) Usermode helper.
The usermode helper code now carries a cred struct pointer in its
subprocess_info struct instead of a new session keyring pointer. This set
of credentials is derived from init_cred and installed on the new process
after it has been cloned.
call_usermodehelper_setup() allocates the new credentials and
call_usermodehelper_freeinfo() discards them if they haven't been used. A
special cred function (prepare_usermodeinfo_creds()) is provided
specifically for call_usermodehelper_setup() to call.
call_usermodehelper_setkeys() adjusts the credentials to sport the
supplied keyring as the new session keyring.
(11) SELinux.
SELinux has a number of changes, in addition to those to support the LSM
interface changes mentioned above:
(a) selinux_setprocattr() no longer does its check for whether the
current ptracer can access processes with the new SID inside the lock
that covers getting the ptracer's SID. Whilst this lock ensures that
the check is done with the ptracer pinned, the result is only valid
until the lock is released, so there's no point doing it inside the
lock.
(12) is_single_threaded().
This function has been extracted from selinux_setprocattr() and put into
a file of its own in the lib/ directory as join_session_keyring() now
wants to use it too.
The code in SELinux just checked to see whether a task shared mm_structs
with other tasks (CLONE_VM), but that isn't good enough. We really want
to know if they're part of the same thread group (CLONE_THREAD).
(13) nfsd.
The NFS server daemon now has to use the COW credentials to set the
credentials it is going to use. It really needs to pass the credentials
down to the functions it calls, but it can't do that until other patches
in this series have been applied.
Signed-off-by: David Howells <dhowells@redhat.com>
Acked-by: James Morris <jmorris@namei.org>
Signed-off-by: James Morris <jmorris@namei.org>
2008-11-14 07:39:23 +08:00
|
|
|
#define audit_log_bprm_fcaps(b, ncr, ocr) ({ 0; })
|
|
|
|
#define audit_log_capset(pid, ncr, ocr) ({ 0; })
|
2007-03-21 01:58:35 +08:00
|
|
|
#define audit_ptrace(t) ((void)0)
|
2006-07-10 20:29:24 +08:00
|
|
|
#define audit_n_rules 0
|
2007-03-30 06:01:04 +08:00
|
|
|
#define audit_signals 0
|
2005-04-17 06:20:36 +08:00
|
|
|
#endif
|
|
|
|
|
|
|
|
#ifdef CONFIG_AUDIT
|
|
|
|
/* These are defined in audit.c */
|
|
|
|
/* Public API */
|
2005-10-21 15:22:03 +08:00
|
|
|
extern void audit_log(struct audit_context *ctx, gfp_t gfp_mask,
|
2005-06-22 22:04:33 +08:00
|
|
|
int type, const char *fmt, ...)
|
|
|
|
__attribute__((format(printf,4,5)));
|
2005-04-17 06:20:36 +08:00
|
|
|
|
2005-10-21 15:22:03 +08:00
|
|
|
extern struct audit_buffer *audit_log_start(struct audit_context *ctx, gfp_t gfp_mask, int type);
|
2005-04-17 06:20:36 +08:00
|
|
|
extern void audit_log_format(struct audit_buffer *ab,
|
|
|
|
const char *fmt, ...)
|
|
|
|
__attribute__((format(printf,2,3)));
|
|
|
|
extern void audit_log_end(struct audit_buffer *ab);
|
2008-01-08 03:31:58 +08:00
|
|
|
extern int audit_string_contains_control(const char *string,
|
|
|
|
size_t len);
|
2008-04-18 22:12:59 +08:00
|
|
|
extern void audit_log_n_hex(struct audit_buffer *ab,
|
|
|
|
const unsigned char *buf,
|
|
|
|
size_t len);
|
|
|
|
extern void audit_log_n_string(struct audit_buffer *ab,
|
|
|
|
const char *buf,
|
|
|
|
size_t n);
|
|
|
|
#define audit_log_string(a,b) audit_log_n_string(a, b, strlen(b));
|
|
|
|
extern void audit_log_n_untrustedstring(struct audit_buffer *ab,
|
|
|
|
const char *string,
|
|
|
|
size_t n);
|
2008-01-08 03:31:58 +08:00
|
|
|
extern void audit_log_untrustedstring(struct audit_buffer *ab,
|
2005-04-29 22:54:44 +08:00
|
|
|
const char *string);
|
2005-04-17 06:20:36 +08:00
|
|
|
extern void audit_log_d_path(struct audit_buffer *ab,
|
|
|
|
const char *prefix,
|
2008-02-15 11:38:33 +08:00
|
|
|
struct path *path);
|
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
|
|
|
extern void audit_log_lost(const char *message);
|
2008-03-02 04:03:14 +08:00
|
|
|
extern int audit_update_lsm_rules(void);
|
|
|
|
|
2005-12-16 02:33:52 +08:00
|
|
|
/* Private API (for audit.c only) */
|
2008-05-20 09:13:02 +08:00
|
|
|
extern int audit_filter_user(struct netlink_skb_parms *cb);
|
2005-12-16 02:33:52 +08:00
|
|
|
extern int audit_filter_type(int type);
|
|
|
|
extern int audit_receive_filter(int type, int pid, int uid, int seq,
|
2008-04-18 22:09:25 +08:00
|
|
|
void *data, size_t datasz, uid_t loginuid,
|
|
|
|
u32 sessionid, u32 sid);
|
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
|
|
|
extern int audit_enabled;
|
2005-04-17 06:20:36 +08:00
|
|
|
#else
|
2005-07-02 20:50:40 +08:00
|
|
|
#define audit_log(c,g,t,f,...) do { ; } while (0)
|
2005-07-02 20:49:07 +08:00
|
|
|
#define audit_log_start(c,g,t) ({ NULL; })
|
2005-04-17 06:20:36 +08:00
|
|
|
#define audit_log_vformat(b,f,a) do { ; } while (0)
|
|
|
|
#define audit_log_format(b,f,...) do { ; } while (0)
|
|
|
|
#define audit_log_end(b) do { ; } while (0)
|
2008-04-18 22:12:59 +08:00
|
|
|
#define audit_log_n_hex(a,b,l) do { ; } while (0)
|
|
|
|
#define audit_log_n_string(a,c,l) do { ; } while (0)
|
|
|
|
#define audit_log_string(a,c) do { ; } while (0)
|
2006-06-09 11:19:31 +08:00
|
|
|
#define audit_log_n_untrustedstring(a,n,s) do { ; } while (0)
|
2008-04-18 22:12:59 +08:00
|
|
|
#define audit_log_untrustedstring(a,s) do { ; } while (0)
|
2008-02-15 11:38:33 +08:00
|
|
|
#define audit_log_d_path(b, p, d) do { ; } while (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
|
|
|
#define audit_enabled 0
|
2005-04-17 06:20:36 +08:00
|
|
|
#endif
|
|
|
|
#endif
|
|
|
|
#endif
|