{ "content": [ { "type": "text", "text": "# auditctl (man)\n\n## NAME\n\nauditctl - a utility to assist controlling the kernel's audit system\n\n## SYNOPSIS\n\nauditctl [options]\n\n## DESCRIPTION\n\nThe auditctl program is used to configure kernel options related to auditing, to see status\nof the configuration, and to load discretionary audit rules.\n\n## TLDR\n\n> Utility to control the behavior, get status and manage rules of the Linux Auditing System.\n\n- Display the [s]tatus of the audit system:\n `sudo auditctl -s`\n- [l]ist all currently loaded audit rules:\n `sudo auditctl -l`\n- [D]elete all audit rules:\n `sudo auditctl -D`\n- [e]nable/disable the audit system:\n `sudo auditctl -e {{1|0}}`\n- Watch a file for changes:\n `sudo auditctl -a always,exit -F arch=b64 -F path=/{{path/to/file}} -F perm=wa`\n- Recursively watch a directory for changes:\n `sudo auditctl -a always,exit -F arch=b64 -F dir=/{{path/to/directory}}/ -F perm=wa`\n- Display [h]elp:\n `auditctl -h`\n\n*Source: tldr-pages*\n\n## Sections\n\n- **NAME**\n- **SYNOPSIS**\n- **DESCRIPTION**\n- **CONFIGURATION OPTIONS** (14 subsections)\n- **STATUS OPTIONS** (4 subsections)\n- **RULE OPTIONS** (10 subsections)\n- **PERFORMANCE TIPS** (4 subsections)\n- **EXAMPLES** (10 subsections)\n- **DISABLED BY DEFAULT**\n- **FILES**\n- **SEE ALSO**\n- **AUTHOR**\n\nUse structuredContent.sections for detailed options, examples, and full documentation.\n" } ], "structuredContent": { "command": "auditctl", "section": "", "mode": "man", "summary": "auditctl - a utility to assist controlling the kernel's audit system", "synopsis": "auditctl [options]", "tldr_summary": "Utility to control the behavior, get status and manage rules of the Linux Auditing System.", "tldr_examples": [ { "description": "Display the [s]tatus of the audit system", "command": "sudo auditctl -s" }, { "description": "[l]ist all currently loaded audit rules", "command": "sudo auditctl -l" }, { "description": "[D]elete all audit rules", "command": "sudo auditctl -D" }, { "description": "[e]nable/disable the audit system", "command": "sudo auditctl -e {{1|0}}" }, { "description": "Watch a file for changes", "command": "sudo auditctl -a always,exit -F arch=b64 -F path=/{{path/to/file}} -F perm=wa" }, { "description": "Recursively watch a directory for changes", "command": "sudo auditctl -a always,exit -F arch=b64 -F dir=/{{path/to/directory}}/ -F perm=wa" }, { "description": "Display [h]elp", "command": "auditctl -h" } ], "tldr_source": "official", "flags": [ { "flag": "-b", "long": null, "arg": null, "description": "Set max number (limit) of outstanding audit buffers allowed (Kernel Default=64) If all buffers are full, the failure flag is consulted by the kernel for action. --backlogwaittime waittime Set the time for the kernel to wait (Kernel Default 60*HZ) when the backlog limit is reached before queuing more audit events to be transferred to auditd. The number must be greater than or equal to zero and less than 10 times the default value." }, { "flag": "", "long": "--reset_backlog_wait_time_actual", "arg": null, "description": "Reset the actual backlog wait time counter shown by the status command." }, { "flag": "-c", "long": null, "arg": null, "description": "the rules. The exit code will not be success if any rule fails to load." }, { "flag": "-D", "long": null, "arg": null, "description": "" }, { "flag": "-e", "long": null, "arg": null, "description": "Set enabled flag. When 0 is passed, this can be used to temporarily disable auditing. When 1 is passed as an argument, it will enable auditing. To lock the audit configura‐ tion so that it can't be changed, pass a 2 as the argument. Locking the configuration is intended to be the last command in audit.rules for anyone wishing this feature to be active. Any attempt to change the configuration in this mode will be audited and denied. The configuration can only be changed by rebooting the machine." }, { "flag": "-f", "long": null, "arg": null, "description": "Set failure mode 0=silent 1=printk 2=panic. This option lets you determine how you want the kernel to handle critical errors. Example conditions where this mode may have an effect includes: transmission errors to userspace audit daemon, backlog limit ex‐ ceeded, out of kernel memory, and rate limit exceeded. The default value is 1. Secure environments will probably want to set this to 2." }, { "flag": "-h", "long": null, "arg": null, "description": "" }, { "flag": "-i", "long": null, "arg": null, "description": "ditctl to always return a success exit code. If passed as an argument to -s then it gives an interpretation of the numbers to human readable words if possible." }, { "flag": "", "long": "--loginuid-immutable", "arg": null, "description": "This option tells the kernel to make loginuids unchangeable once they are set. Chang‐ ing loginuids requires CAPAUDITCONTROL. So, its not something that can be done by unprivileged users. Setting this makes loginuid tamper-proof, but can cause some prob‐ lems in certain kinds of containers." }, { "flag": "-q", "long": null, "arg": null, "description": "If you have an existing directory watch and bind or move mount another subtree in the watched subtree, you need to tell the kernel to make the subtree being mounted equiva‐ lent to the directory being watched. If the subtree is already mounted at the time the directory watch is issued, the subtree is automatically tagged for watching. Please note the comma separating the two values. Omitting it will cause errors." }, { "flag": "-r", "long": null, "arg": null, "description": "Set limit in messages/sec (0=none). If this rate is non-zero and is exceeded, the failure flag is consulted by the kernel for action. The default value is 0." }, { "flag": "", "long": "--reset-lost", "arg": null, "description": "Reset the lost record counter shown by the status command." }, { "flag": "-R", "long": null, "arg": null, "description": "Read and execute auditctl commands from a file. The commands are executed line-by- line, in the order that they appear in the file. The file must be owned by root and not readable by other users, or else it will be rejected. Empty lines are skipped. Lines starting with the '#' character are treated as comment lines. Each line is exe‐ cuted as if it was provided to auditctl as command line arguments. Since auditctl is the one reading the file and not a shell such as bash, do not escape special shell characters. See the EXAMPLES section for an example. --signal signal Send a signal to the audit daemon. You must have privileges to do this. Supported sig‐ nals are TERM, HUP, USR1, USR2, CONT and user friendly versions stop, reload, rotate, resume, state." }, { "flag": "-t", "long": null, "arg": null, "description": "" }, { "flag": "-l", "long": null, "arg": null, "description": "either a key option (-k) to list rules that match a key or a (-i) to have a0 through a3 interpreted to help determine the syscall argument values are correct ." }, { "flag": "-m", "long": null, "arg": null, "description": "Send a user space message into the audit system. This can only be done if you have CAPAUDITWRITE capability (normally the root user has this). The resulting event will be the USER type." }, { "flag": "-s", "long": null, "arg": null, "description": "can be set by -e, -f, -r, and -b options. The pid value is the process number of the audit daemon. Note that a pid of 0 indicates that the audit daemon is not running. The lost entry will tell you how many event records that have been discarded due to the kernel audit queue overflowing. The backlog field tells how many event records are currently queued waiting for auditd to read them. This option can be followed by the -i to get a couple fields interpreted." }, { "flag": "-v", "long": null, "arg": null, "description": "" }, { "flag": "-a", "long": null, "arg": null, "description": "Append rule to the end of list with action. Please note the comma separating the two values. Omitting it will cause errors. The fields may be in either order. It could be list,action or action,list. The following describes the valid list names: task Add a rule to the per task list. This rule list is used only at the time a task is created -- when fork() or clone() are called by the parent task. When using this list, you should only use fields that are known at task creation time, such as the uid, gid, etc. exit Add a rule to the syscall exit list. This list is used upon exit from a system call to determine if an audit event should be created. user Add a rule to the user message filter list. This list is used by the ker‐ nel to filter events originating in user space before relaying them to the audit daemon. It should be noted that the only fields that are valid are: uid, auid, gid, pid, subjuser, subjrole, subjtype, subjsen, subjclr, msgtype, and executable name. All other fields will be treated as non- matching. It should be understood that any event originating from user space from a process that has CAPAUDITWRITE will be recorded into the audit trail. This means that the most likely use for this filter is with rules that have an action of never since nothing has to be done to allow events to be recorded. exclude Add a rule to the event type exclusion filter list. This list is used to filter events that you do not want to see. For example, if you do not want to see any avc messages, you would using this list to record that. Events can be excluded by process ID, user ID, group ID, login user ID, message type, subject context, or executable name. The action is ignored and uses its default of \"never\". filesystem Add a rule that will be applied to a whole filesystem. The filesystem must be identified with a fstype field. Normally this filter is used to exclude any events for a whole filesystem such as tracefs or debugfs. iouring Add a rule to the iouring syscall filter. Rules against this filter spec‐ ify the syscall operation using the -S syscall notion explained below. You can add a key field to the rule so that it may be grouped with other rules watching the same underlying syscall. The following describes the valid actions for the rule: never No audit records will be generated. This can be used to suppress event generation. In general, you want suppressions at the top of the list in‐ stead of the bottom. This is because the event triggers on the first matching rule. always Allocate an audit context, always fill it in at syscall entry time, and always write out a record at syscall exit time." }, { "flag": "-A", "long": null, "arg": null, "description": "Add rule to the beginning list with action." }, { "flag": "-C", "long": null, "arg": null, "description": "Build an inter-field comparison rule: field, operation, field. You may pass multiple comparisons on a single command line. Each one must start with -C. Each inter-field equation is anded with each other as well as equations starting with -F to trigger an audit record. There are 2 operators supported - equal, and not equal. Valid fields are: auid, uid, euid, suid, fsuid, objuid; and gid, egid, sgid, fsgid, objgid The two groups of uid and gid cannot be mixed. But any comparison within the group can be made. The objuid/gid fields are collected from the object of the event such as a file or directory." }, { "flag": "-d", "long": null, "arg": null, "description": "Delete rule from list with action. The rule is deleted only if it exactly matches syscall name(s) and every field name and value." }, { "flag": "-F", "long": null, "arg": null, "description": "Build a rule field: name, operation, value. You may have up to 64 fields passed on a single command line. Each one must start with -F. Each field equation is anded with each other (as well as equations starting with -C) to trigger an audit record. There are 8 operators supported - equal, not equal, less than, greater than, less than or equal, and greater than or equal, bit mask, and bit test respectively. Bit test will \"and\" the values and check that they are equal, bit mask just \"ands\" the values. Fields that take a user ID may instead have the user's name; the program will convert the name to user ID. The same is true of group names. Valid fields are: a0, a1, a2, a3 Respectively, the first 4 arguments to a syscall. Note that string argu‐ ments are not supported. This is because the kernel is passed a pointer to the string. Triggering on a pointer address value is not likely to work. So, when using this, you should only use on numeric values. This is most likely to be used on platforms that multiplex socket or IPC operations. arch The CPU architecture of the syscall. The arch can be found doing 'uname -m'. If you do not know the arch of your machine but you want to use the 32 bit syscall table and your machine supports 32 bit, you can also use b32 for the arch. The same applies to the 64 bit syscall table, you can use b64. In this way, you can write rules that are somewhat arch indepen‐ dent because the family type will be auto detected. However, syscalls can be arch specific and what is available on x8664, may not be available on ppc. The arch directive should precede the -S option so that auditctl knows which internal table to use to look up the syscall numbers. auid The original ID the user logged in with. Its an abbreviation of audit uid. Sometimes its referred to as loginuid. Either the user account text or number may be used. devmajor Device Major Number devminor Device Minor Number dir Full Path of Directory to watch. This will place a recursive watch on the directory and its whole subtree. It can only be used on exit list. See \"-w\". egid Effective Group ID. May be numeric or the groups name. euid Effective User ID. May be numeric or the user account name. exe Absolute path to application that while executing this rule will apply to. It supports = and != operators. Note that you can only use this once for each rule. exit Exit value from a syscall. If the exit code is an errno, you may use the text representation, too. fsgid Filesystem Group ID. May be numeric or the groups name. fstype File system type. This is used with the filesystem rule list. The only values supported are debugfs and tracefs. fsuid Filesystem User ID. May be numeric or the user account name. filetype The target file's type. Can be either file, dir, socket, link, character, block, or fifo. gid Group ID. May be numeric or the groups name. inode Inode Number key This is another way of setting a filter key. See discussion above for -k option. msgtype This is used to match the event's record type. It should only be used on the exclude or user filter lists. objuid Object's UID objgid Object's GID objuser Resource's SE Linux User objrole Resource's SE Linux Role objtype Resource's SE Linux Type objlevlow Resource's SE Linux Low Level objlevhigh Resource's SE Linux High Level path Full Path of File to watch. It can only be used on exit list. perm Permission filter for file operations. See \"-p\". It can only be used on exit list. You can use this without specifying a syscall and the kernel will select the syscalls that satisfy the permissions being requested. pers OS Personality Number pid Process ID ppid Parent's Process ID saddrfam Address family number as found in /usr/include/bits/socket.h. For example, IPv4 would be 2 and IPv6 would be 10. sessionid User's login session ID subjuser Program's SE Linux User subjrole Program's SE Linux Role subjtype Program's SE Linux Type subjsen Program's SE Linux Sensitivity subjclr Program's SE Linux Clearance sgid Saved Group ID. See getresgid(2) man page. success If the exit value is >= 0 this is true/yes otherwise its false/no. When writing a rule, use a 1 for true/yes and a 0 for false/no suid Saved User ID. See getresuid(2) man page. uid User ID. May be numeric or the user account name." }, { "flag": "-k", "long": null, "arg": null, "description": "can be up to 31 bytes long. It can uniquely identify the audit records produced by a rule. Typical use is for when you have several rules that together satisfy a security requirement. The key value can be searched on with ausearch so that no matter which rule triggered the event, you can find its results. The key can also be used on delete all (-D) and list rules (-l) to select rules with a specific key. You may have more than one key on a rule if you want to be able to search logged events in multiple ways or if you have an auditd plugin that uses a key to aid its analysis." }, { "flag": "-p", "long": null, "arg": null, "description": "Describe the permission access type that a file system watch will trigger on. r=read, w=write, x=execute, a=attribute change. These permissions are not the standard file permissions, but rather the kind of syscall that would do this kind of thing. The read & write syscalls are omitted from this set since they would overwhelm the logs. But rather for reads or writes, the open flags are looked at to see what permission was requested." }, { "flag": "-S", "long": null, "arg": null, "description": "Any syscall name or number may be used. The word 'all' may also be used. If the given syscall is made by a program, then start an audit record. If a field rule is given and no syscall is specified, it will default to all syscalls. You may also specify multi‐ ple syscalls in the same rule by using multiple -S options in the same rule. Doing so improves performance since fewer rules need to be evaluated. Alternatively, you may pass a comma separated list of syscall names. If you are on a bi-arch system, like x8664, you should be aware that auditctl simply takes the text, looks it up for the native arch (in this case b64) and sends that rule to the kernel. If there are no ad‐ ditional arch directives, IT WILL APPLY TO BOTH 32 & 64 BIT SYSCALLS. This can have undesirable effects since there is no guarantee that any syscall has the same number on both 32 and 64 bit interfaces. You will likely want to control this and write 2 rules, one with arch equal to b32 and one with b64 to make sure the kernel finds the events that you intend. See the arch field discussion for more info." }, { "flag": "-w", "long": null, "arg": null, "description": "Insert a watch for the file system object at path. You cannot insert a watch to the top level directory. This is prohibited by the kernel. Wildcards are not supported ei‐ ther and will generate a warning. The way that watches work is by tracking the inode internally. If you place a watch on a file, its the same as using the -F path option on a syscall rule. If you place a watch on a directory, its the same as using the -F dir option on a syscall rule. The -w form of writing watches is for backwards compati‐ bility and the syscall based form is more expressive. Unlike most syscall auditing rules, watches do not impact performance based on the number of rules sent to the ker‐ nel. The only valid options when using a watch are the -p and -k. If you need to do anything fancy like audit a specific user accessing a file, then use the syscall au‐ diting form with the path or dir fields. See the EXAMPLES section for an example of converting one form to another." }, { "flag": "-W", "long": null, "arg": null, "description": "Remove a watch for the file system object at path. The rule must match exactly. See -d discussion for more info." }, { "flag": "-D", "long": null, "arg": null, "description": "# Never record sudo invocations -A exclude,always -F exe=/usr/bin/sudo" } ], "examples": [ "To see all syscalls made by a specific program:", "# By pid:", "# By executable path", "To see files opened by a specific user:", "To see unsuccessful openat calls:", "To watch a file for changes (2 ways to express):", "To recursively watch a directory for changes (2 ways to express):", "To see if an admin is accessing other user's files:", "This is an example rules file:", "# Remove all existing rules", "# Never record sudo invocations", "-A exclude,always -F exe=/usr/bin/sudo" ], "see_also": [ { "name": "audit.rules", "section": "7", "url": "https://www.chedong.com/phpMan.php/man/audit.rules/7/json" }, { "name": "ausearch", "section": "8", "url": "https://www.chedong.com/phpMan.php/man/ausearch/8/json" }, { "name": "aureport", "section": "8", "url": "https://www.chedong.com/phpMan.php/man/aureport/8/json" }, { "name": "auditd", "section": "8", "url": "https://www.chedong.com/phpMan.php/man/auditd/8/json" } ], "section_outline": [ { "name": "NAME", "lines": 2, "subsections": [] }, { "name": "SYNOPSIS", "lines": 2, "subsections": [] }, { "name": "DESCRIPTION", "lines": 3, "subsections": [] }, { "name": "CONFIGURATION OPTIONS", "lines": 1, "subsections": [ { "name": "-b _", "lines": 8, "flag": "-b" }, { "name": "--reset_backlog_wait_time_actual", "lines": 2, "long": "--reset_backlog_wait_time_actual" }, { "name": "-c", "lines": 2, "flag": "-c" }, { "name": "-D", "lines": 1, "flag": "-D" }, { "name": "-e", "lines": 7, "flag": "-e" }, { "name": "-f", "lines": 6, "flag": "-f" }, { "name": "-h", "lines": 1, "flag": "-h" }, { "name": "-i", "lines": 3, "flag": "-i" }, { "name": "--loginuid-immutable", "lines": 5, "long": "--loginuid-immutable" }, { "name": "-q _", "lines": 6, "flag": "-q" }, { "name": "-r _", "lines": 3, "flag": "-r" }, { "name": "--reset-lost", "lines": 2, "long": "--reset-lost" }, { "name": "-R _", "lines": 13, "flag": "-R" }, { "name": "-t", "lines": 1, "flag": "-t" } ] }, { "name": "STATUS OPTIONS", "lines": 1, "subsections": [ { "name": "-l", "lines": 3, "flag": "-l" }, { "name": "-m _", "lines": 4, "flag": "-m" }, { "name": "-s", "lines": 7, "flag": "-s" }, { "name": "-v", "lines": 1, "flag": "-v" } ] }, { "name": "RULE OPTIONS", "lines": 1, "subsections": [ { "name": "-a _", "lines": 49, "flag": "-a" }, { "name": "-A _", "lines": 2, "flag": "-A" }, { "name": "-C", "lines": 13, "flag": "-C" }, { "name": "-d _", "lines": 3, "flag": "-d" }, { "name": "-F", "lines": 120, "flag": "-F" }, { "name": "-k _", "lines": 8, "flag": "-k" }, { "name": "-p", "lines": 7, "flag": "-p" }, { "name": "-S", "lines": 14, "flag": "-S" }, { "name": "-w _", "lines": 13, "flag": "-w" }, { "name": "-W _", "lines": 3, "flag": "-W" } ] }, { "name": "PERFORMANCE TIPS", "lines": 5, "subsections": [ { "name": "auditctl -a always,exit -F arch=b64 -S openat -F success=0", "lines": 1 }, { "name": "auditctl -a always,exit -F arch=b64 -S truncate -F success=0", "lines": 2 }, { "name": "auditctl -a always,exit -F arch=b64 -S openat -S truncate -F success=0", "lines": 5 }, { "name": "auditctl -a always,exit -S openat -S truncate -F dir=/etc -F success=0", "lines": 4 } ] }, { "name": "EXAMPLES", "lines": 3, "subsections": [ { "name": "auditctl -a always,exit -S all -F pid=1005", "lines": 1 }, { "name": "auditctl -a always,exit -S all -F exe=/usr/bin/ls", "lines": 2 }, { "name": "auditctl -a always,exit -S openat -F auid=510", "lines": 2 }, { "name": "auditctl -a always,exit -S openat -F success=0", "lines": 2 }, { "name": "auditctl -w /etc/shadow -p wa", "lines": 1 }, { "name": "auditctl -a always,exit -F path=/etc/shadow -F perm=wa", "lines": 2 }, { "name": "auditctl -w /etc/ -p wa", "lines": 1 }, { "name": "auditctl -a always,exit -F dir=/etc/ -F perm=wa", "lines": 2 }, { "name": "auditctl -a always,exit -F dir=/home/ -F uid=0 -C auid!=obj_uid", "lines": 3 }, { "name": "-D", "lines": 4, "flag": "-D" } ] }, { "name": "DISABLED BY DEFAULT", "lines": 10, "subsections": [] }, { "name": "FILES", "lines": 3, "subsections": [] }, { "name": "SEE ALSO", "lines": 3, "subsections": [] }, { "name": "AUTHOR", "lines": 3, "subsections": [] } ], "sections": { "NAME": { "content": "auditctl - a utility to assist controlling the kernel's audit system\n", "subsections": [] }, "SYNOPSIS": { "content": "auditctl [options]\n", "subsections": [] }, "DESCRIPTION": { "content": "The auditctl program is used to configure kernel options related to auditing, to see status\nof the configuration, and to load discretionary audit rules.\n", "subsections": [] }, "CONFIGURATION OPTIONS": { "content": "", "subsections": [ { "name": "-b _", "content": "Set max number (limit) of outstanding audit buffers allowed (Kernel Default=64) If all\nbuffers are full, the failure flag is consulted by the kernel for action.\n\n--backlogwaittime waittime\nSet the time for the kernel to wait (Kernel Default 60*HZ) when the backlog limit is\nreached before queuing more audit events to be transferred to auditd. The number must\nbe greater than or equal to zero and less than 10 times the default value.\n", "flag": "-b" }, { "name": "--reset_backlog_wait_time_actual", "content": "Reset the actual backlog wait time counter shown by the status command.\n", "long": "--reset_backlog_wait_time_actual" }, { "name": "-c", "content": "the rules. The exit code will not be success if any rule fails to load.\n", "flag": "-c" }, { "name": "-D", "content": "", "flag": "-D" }, { "name": "-e", "content": "Set enabled flag. When 0 is passed, this can be used to temporarily disable auditing.\nWhen 1 is passed as an argument, it will enable auditing. To lock the audit configura‐\ntion so that it can't be changed, pass a 2 as the argument. Locking the configuration\nis intended to be the last command in audit.rules for anyone wishing this feature to\nbe active. Any attempt to change the configuration in this mode will be audited and\ndenied. The configuration can only be changed by rebooting the machine.\n", "flag": "-e" }, { "name": "-f", "content": "Set failure mode 0=silent 1=printk 2=panic. This option lets you determine how you\nwant the kernel to handle critical errors. Example conditions where this mode may have\nan effect includes: transmission errors to userspace audit daemon, backlog limit ex‐\nceeded, out of kernel memory, and rate limit exceeded. The default value is 1. Secure\nenvironments will probably want to set this to 2.\n", "flag": "-f" }, { "name": "-h", "content": "", "flag": "-h" }, { "name": "-i", "content": "ditctl to always return a success exit code. If passed as an argument to -s then it\ngives an interpretation of the numbers to human readable words if possible.\n", "flag": "-i" }, { "name": "--loginuid-immutable", "content": "This option tells the kernel to make loginuids unchangeable once they are set. Chang‐\ning loginuids requires CAPAUDITCONTROL. So, its not something that can be done by\nunprivileged users. Setting this makes loginuid tamper-proof, but can cause some prob‐\nlems in certain kinds of containers.\n", "long": "--loginuid-immutable" }, { "name": "-q _", "content": "If you have an existing directory watch and bind or move mount another subtree in the\nwatched subtree, you need to tell the kernel to make the subtree being mounted equiva‐\nlent to the directory being watched. If the subtree is already mounted at the time the\ndirectory watch is issued, the subtree is automatically tagged for watching. Please\nnote the comma separating the two values. Omitting it will cause errors.\n", "flag": "-q" }, { "name": "-r _", "content": "Set limit in messages/sec (0=none). If this rate is non-zero and is exceeded, the\nfailure flag is consulted by the kernel for action. The default value is 0.\n", "flag": "-r" }, { "name": "--reset-lost", "content": "Reset the lost record counter shown by the status command.\n", "long": "--reset-lost" }, { "name": "-R _", "content": "Read and execute auditctl commands from a file. The commands are executed line-by-\nline, in the order that they appear in the file. The file must be owned by root and\nnot readable by other users, or else it will be rejected. Empty lines are skipped.\nLines starting with the '#' character are treated as comment lines. Each line is exe‐\ncuted as if it was provided to auditctl as command line arguments. Since auditctl is\nthe one reading the file and not a shell such as bash, do not escape special shell\ncharacters. See the EXAMPLES section for an example.\n\n--signal signal\nSend a signal to the audit daemon. You must have privileges to do this. Supported sig‐\nnals are TERM, HUP, USR1, USR2, CONT\nand user friendly versions stop, reload, rotate, resume, state.\n", "flag": "-R" }, { "name": "-t", "content": "", "flag": "-t" } ] }, "STATUS OPTIONS": { "content": "", "subsections": [ { "name": "-l", "content": "either a key option (-k) to list rules that match a key or a (-i) to have a0 through\na3 interpreted to help determine the syscall argument values are correct .\n", "flag": "-l" }, { "name": "-m _", "content": "Send a user space message into the audit system. This can only be done if you have\nCAPAUDITWRITE capability (normally the root user has this). The resulting event will\nbe the USER type.\n", "flag": "-m" }, { "name": "-s", "content": "can be set by -e, -f, -r, and -b options. The pid value is the process number of the\naudit daemon. Note that a pid of 0 indicates that the audit daemon is not running. The\nlost entry will tell you how many event records that have been discarded due to the\nkernel audit queue overflowing. The backlog field tells how many event records are\ncurrently queued waiting for auditd to read them. This option can be followed by the\n-i to get a couple fields interpreted.\n", "flag": "-s" }, { "name": "-v", "content": "", "flag": "-v" } ] }, "RULE OPTIONS": { "content": "", "subsections": [ { "name": "-a _", "content": "Append rule to the end of list with action. Please note the comma separating the two\nvalues. Omitting it will cause errors. The fields may be in either order. It could be\nlist,action or action,list. The following describes the valid list names:\n\ntask Add a rule to the per task list. This rule list is used only at the time a\ntask is created -- when fork() or clone() are called by the parent task.\nWhen using this list, you should only use fields that are known at task\ncreation time, such as the uid, gid, etc.\n\nexit Add a rule to the syscall exit list. This list is used upon exit from a\nsystem call to determine if an audit event should be created.\n\nuser Add a rule to the user message filter list. This list is used by the ker‐\nnel to filter events originating in user space before relaying them to the\naudit daemon. It should be noted that the only fields that are valid are:\nuid, auid, gid, pid, subjuser, subjrole, subjtype, subjsen, subjclr,\nmsgtype, and executable name. All other fields will be treated as non-\nmatching. It should be understood that any event originating from user\nspace from a process that has CAPAUDITWRITE will be recorded into the\naudit trail. This means that the most likely use for this filter is with\nrules that have an action of never since nothing has to be done to allow\nevents to be recorded.\n\nexclude Add a rule to the event type exclusion filter list. This list is used to\nfilter events that you do not want to see. For example, if you do not want\nto see any avc messages, you would using this list to record that. Events\ncan be excluded by process ID, user ID, group ID, login user ID, message\ntype, subject context, or executable name. The action is ignored and uses\nits default of \"never\".\n\nfilesystem Add a rule that will be applied to a whole filesystem. The filesystem must\nbe identified with a fstype field. Normally this filter is used to exclude\nany events for a whole filesystem such as tracefs or debugfs.\n\niouring Add a rule to the iouring syscall filter. Rules against this filter spec‐\nify the syscall operation using the -S syscall notion explained below. You\ncan add a key field to the rule so that it may be grouped with other rules\nwatching the same underlying syscall.\n\nThe following describes the valid actions for the rule:\n\nnever No audit records will be generated. This can be used to suppress event\ngeneration. In general, you want suppressions at the top of the list in‐\nstead of the bottom. This is because the event triggers on the first\nmatching rule.\n\nalways Allocate an audit context, always fill it in at syscall entry time, and\nalways write out a record at syscall exit time.\n", "flag": "-a" }, { "name": "-A _", "content": "Add rule to the beginning list with action.\n", "flag": "-A" }, { "name": "-C", "content": "Build an inter-field comparison rule: field, operation, field. You may pass multiple\ncomparisons on a single command line. Each one must start with -C. Each inter-field\nequation is anded with each other as well as equations starting with -F to trigger an\naudit record. There are 2 operators supported - equal, and not equal. Valid fields\nare:\n\nauid, uid, euid, suid, fsuid, objuid; and gid, egid, sgid, fsgid, objgid\n\nThe two groups of uid and gid cannot be mixed. But any comparison within the group can\nbe made. The objuid/gid fields are collected from the object of the event such as a\nfile or directory.\n\n", "flag": "-C" }, { "name": "-d _", "content": "Delete rule from list with action. The rule is deleted only if it exactly matches\nsyscall name(s) and every field name and value.\n", "flag": "-d" }, { "name": "-F", "content": "Build a rule field: name, operation, value. You may have up to 64 fields passed on a\nsingle command line. Each one must start with -F. Each field equation is anded with\neach other (as well as equations starting with -C) to trigger an audit record. There\nare 8 operators supported - equal, not equal, less than, greater than, less than or\nequal, and greater than or equal, bit mask, and bit test respectively. Bit test will\n\"and\" the values and check that they are equal, bit mask just \"ands\" the values.\nFields that take a user ID may instead have the user's name; the program will convert\nthe name to user ID. The same is true of group names. Valid fields are:\n\na0, a1, a2, a3\nRespectively, the first 4 arguments to a syscall. Note that string argu‐\nments are not supported. This is because the kernel is passed a pointer to\nthe string. Triggering on a pointer address value is not likely to work.\nSo, when using this, you should only use on numeric values. This is most\nlikely to be used on platforms that multiplex socket or IPC operations.\n\narch The CPU architecture of the syscall. The arch can be found doing 'uname\n-m'. If you do not know the arch of your machine but you want to use the\n32 bit syscall table and your machine supports 32 bit, you can also use\nb32 for the arch. The same applies to the 64 bit syscall table, you can\nuse b64. In this way, you can write rules that are somewhat arch indepen‐\ndent because the family type will be auto detected. However, syscalls can\nbe arch specific and what is available on x8664, may not be available on\nppc. The arch directive should precede the -S option so that auditctl\nknows which internal table to use to look up the syscall numbers.\n\nauid The original ID the user logged in with. Its an abbreviation of audit uid.\nSometimes its referred to as loginuid. Either the user account text or\nnumber may be used.\n\ndevmajor Device Major Number\n\ndevminor Device Minor Number\n\ndir Full Path of Directory to watch. This will place a recursive watch on the\ndirectory and its whole subtree. It can only be used on exit list. See\n\"-w\".\n\negid Effective Group ID. May be numeric or the groups name.\n\neuid Effective User ID. May be numeric or the user account name.\n\nexe Absolute path to application that while executing this rule will apply to.\nIt supports = and != operators. Note that you can only use this once for\neach rule.\n\nexit Exit value from a syscall. If the exit code is an errno, you may use the\ntext representation, too.\n\nfsgid Filesystem Group ID. May be numeric or the groups name.\n\nfstype File system type. This is used with the filesystem rule list. The only\nvalues supported are debugfs and tracefs.\n\nfsuid Filesystem User ID. May be numeric or the user account name.\n\nfiletype The target file's type. Can be either file, dir, socket, link, character,\nblock, or fifo.\n\ngid Group ID. May be numeric or the groups name.\n\ninode Inode Number\n\nkey This is another way of setting a filter key. See discussion above for -k\noption.\n\nmsgtype This is used to match the event's record type. It should only be used on\nthe exclude or user filter lists.\n\nobjuid Object's UID\n\nobjgid Object's GID\n\nobjuser Resource's SE Linux User\n\nobjrole Resource's SE Linux Role\n\nobjtype Resource's SE Linux Type\n\nobjlevlow Resource's SE Linux Low Level\n\nobjlevhigh\nResource's SE Linux High Level\n\npath Full Path of File to watch. It can only be used on exit list.\n\nperm Permission filter for file operations. See \"-p\". It can only be used on\nexit list. You can use this without specifying a syscall and the kernel\nwill select the syscalls that satisfy the permissions being requested.\n\npers OS Personality Number\n\npid Process ID\n\nppid Parent's Process ID\n\nsaddrfam Address family number as found in /usr/include/bits/socket.h. For example,\nIPv4 would be 2 and IPv6 would be 10.\n\nsessionid User's login session ID\n\nsubjuser Program's SE Linux User\n\nsubjrole Program's SE Linux Role\n\nsubjtype Program's SE Linux Type\n\nsubjsen Program's SE Linux Sensitivity\n\nsubjclr Program's SE Linux Clearance\n\nsgid Saved Group ID. See getresgid(2) man page.\n\nsuccess If the exit value is >= 0 this is true/yes otherwise its false/no. When\nwriting a rule, use a 1 for true/yes and a 0 for false/no\n\nsuid Saved User ID. See getresuid(2) man page.\n\nuid User ID. May be numeric or the user account name.\n", "flag": "-F" }, { "name": "-k _", "content": "can be up to 31 bytes long. It can uniquely identify the audit records produced by a\nrule. Typical use is for when you have several rules that together satisfy a security\nrequirement. The key value can be searched on with ausearch so that no matter which\nrule triggered the event, you can find its results. The key can also be used on delete\nall (-D) and list rules (-l) to select rules with a specific key. You may have more\nthan one key on a rule if you want to be able to search logged events in multiple ways\nor if you have an auditd plugin that uses a key to aid its analysis.\n", "flag": "-k" }, { "name": "-p", "content": "Describe the permission access type that a file system watch will trigger on. r=read,\nw=write, x=execute, a=attribute change. These permissions are not the standard file\npermissions, but rather the kind of syscall that would do this kind of thing. The read\n& write syscalls are omitted from this set since they would overwhelm the logs. But\nrather for reads or writes, the open flags are looked at to see what permission was\nrequested.\n", "flag": "-p" }, { "name": "-S", "content": "Any syscall name or number may be used. The word 'all' may also be used. If the given\nsyscall is made by a program, then start an audit record. If a field rule is given and\nno syscall is specified, it will default to all syscalls. You may also specify multi‐\nple syscalls in the same rule by using multiple -S options in the same rule. Doing so\nimproves performance since fewer rules need to be evaluated. Alternatively, you may\npass a comma separated list of syscall names. If you are on a bi-arch system, like\nx8664, you should be aware that auditctl simply takes the text, looks it up for the\nnative arch (in this case b64) and sends that rule to the kernel. If there are no ad‐\nditional arch directives, IT WILL APPLY TO BOTH 32 & 64 BIT SYSCALLS. This can have\nundesirable effects since there is no guarantee that any syscall has the same number\non both 32 and 64 bit interfaces. You will likely want to control this and write 2\nrules, one with arch equal to b32 and one with b64 to make sure the kernel finds the\nevents that you intend. See the arch field discussion for more info.\n", "flag": "-S" }, { "name": "-w _", "content": "Insert a watch for the file system object at path. You cannot insert a watch to the\ntop level directory. This is prohibited by the kernel. Wildcards are not supported ei‐\nther and will generate a warning. The way that watches work is by tracking the inode\ninternally. If you place a watch on a file, its the same as using the -F path option\non a syscall rule. If you place a watch on a directory, its the same as using the -F\ndir option on a syscall rule. The -w form of writing watches is for backwards compati‐\nbility and the syscall based form is more expressive. Unlike most syscall auditing\nrules, watches do not impact performance based on the number of rules sent to the ker‐\nnel. The only valid options when using a watch are the -p and -k. If you need to do\nanything fancy like audit a specific user accessing a file, then use the syscall au‐\nditing form with the path or dir fields. See the EXAMPLES section for an example of\nconverting one form to another.\n", "flag": "-w" }, { "name": "-W _", "content": "Remove a watch for the file system object at path. The rule must match exactly. See -d\ndiscussion for more info.\n", "flag": "-W" } ] }, "PERFORMANCE TIPS": { "content": "Syscall rules get evaluated for each syscall for every program. If you have 10 syscall rules,\nevery program on your system will delay during a syscall while the audit system evaluates\neach rule. Too many syscall rules will hurt performance. Try to combine as many as you can\nwhenever the filter, action, key, and fields are identical. For example:\n", "subsections": [ { "name": "auditctl -a always,exit -F arch=b64 -S openat -F success=0", "content": "" }, { "name": "auditctl -a always,exit -F arch=b64 -S truncate -F success=0", "content": "could be re-written as one rule:\n" }, { "name": "auditctl -a always,exit -F arch=b64 -S openat -S truncate -F success=0", "content": "Also, try to use file system auditing wherever practical. This improves performance. For ex‐\nample, if you were wanting to capture all failed opens & truncates like above, but were only\nconcerned about files in /etc and didn't care about /usr or /sbin, its possible to use this\nrule:\n" }, { "name": "auditctl -a always,exit -S openat -S truncate -F dir=/etc -F success=0", "content": "This will be higher performance since the kernel will not evaluate it each and every syscall.\nIt will be handled by the filesystem auditing code and only checked on filesystem related\nsyscalls.\n" } ] }, "EXAMPLES": { "content": "To see all syscalls made by a specific program:\n\n# By pid:", "subsections": [ { "name": "auditctl -a always,exit -S all -F pid=1005", "content": "# By executable path" }, { "name": "auditctl -a always,exit -S all -F exe=/usr/bin/ls", "content": "To see files opened by a specific user:\n" }, { "name": "auditctl -a always,exit -S openat -F auid=510", "content": "To see unsuccessful openat calls:\n" }, { "name": "auditctl -a always,exit -S openat -F success=0", "content": "To watch a file for changes (2 ways to express):\n" }, { "name": "auditctl -w /etc/shadow -p wa", "content": "" }, { "name": "auditctl -a always,exit -F path=/etc/shadow -F perm=wa", "content": "To recursively watch a directory for changes (2 ways to express):\n" }, { "name": "auditctl -w /etc/ -p wa", "content": "" }, { "name": "auditctl -a always,exit -F dir=/etc/ -F perm=wa", "content": "To see if an admin is accessing other user's files:\n" }, { "name": "auditctl -a always,exit -F dir=/home/ -F uid=0 -C auid!=obj_uid", "content": "This is an example rules file:\n\n# Remove all existing rules" }, { "name": "-D", "content": "# Never record sudo invocations\n-A exclude,always -F exe=/usr/bin/sudo\n\n", "flag": "-D" } ] }, "DISABLED BY DEFAULT": { "content": "On many systems auditd is configured to install an -a never,task rule by default. This rule\ncauses every new process to skip all audit rule processing. This is usually done to avoid a\nsmall performance overhead imposed by syscall auditing. If you want to use auditd, you need\nto remove that rule by deleting 10-no-audit.rules and adding 10-base-config.rules to the au‐\ndit rules directory.\n\nIf you have defined audit rules that are not matching when they should, check auditctl -l to\nmake sure there is no never,task rule there.\n\n", "subsections": [] }, "FILES": { "content": "/etc/audit/audit.rules /etc/audit/audit-stop.rules\n\n", "subsections": [] }, "SEE ALSO": { "content": "audit.rules(7), ausearch(8), aureport(8), auditd(8).\n\n", "subsections": [] }, "AUTHOR": { "content": "Steve Grubb\n\nRed Hat Feb 2023 AUDITCTL(8)", "subsections": [] } } } }