Fedora Core 5 SELinux FAQ

SELinux (Security-Enhanced Linux) in Fedora Core is an implementation of mandatory access control in the Linux kernel using the Linux Security Modules (LSM) framework. Standard Linux security is a discretionary access control model.

In a DAC model, file and resource decisions are based solely on user identity and ownership of the objects. Each user and program run by that user has complete discretion over the user's objects. Malicious or flawed software can do anything with the files and resources it controls through the user that started the process. If the user is the super-user or the application is setuid or setgid to root, the process can have root level control over the entire file system.

A MAC system does not suffer from these problems. First, you can administratively define a security policy over all processes and objects. Second, you control all processes and objects, in the case of SELinux through the kernel. Third, decisions are based on all the security relevant information available, and not just authenticated user identity.

MAC under SELinux allows you to provide granular permissions for all subjects (users, programs, processes) and objects (files, devices). In practice, think of subjects as processes, and objects as the target of a process operation. You can safely grant a process only the permissions it needs to perform its function, and no more.

The SELinux implementation uses role-based access control (RBAC), which provides abstracted user-level control based on roles, and Type Enforcement® (TE). TE uses a table, or matrix to handle access controls, enforcing policy rules based on the types of processes and objects. Process types are called domains, and a cross-reference on the matrix of the process's domain and the object's type defines their interaction. This system provides extremely granular control for actors in a Linux system.

The SELinux policy describes the access permissions for all subjects and objects, that is, the entire system of users, programs, and processes and the files and devices they act upon. Fedora Core policy is delivered in a package, with an associated source package. Current shipping policy packages are:

More information on the different policies available in SELinux can be found at http://fedoraproject.org/wiki/SELinux/Policies.

When SELinux was initially introduced in Fedora Core, it enforced the NSA strict policy. For testing purposes, this effectively exposed hundreds of problems in the strict policy. In addition, it demonstrated that applying a single strict policy to the many environments of Fedora users was not feasible. To manage a single strict policy for anything other than default installation would require local expertise.

At this point, the SELinux developers reviewed their choices, and decided to try a different strategy. They decided to create a targeted policy that locks down specific daemons, especially those vulnerable to attack or which could devastate a system if broken or compromised. The rest of the system runs exactly as it would under standard Linux DAC security.

Under the targeted policy, most processes run in the unconfined_t domain. As the name implies, these processes are mostly unconfined by the SELinux policy. They are still governed by standard Linux DAC security, however.

Those network daemons which are addressed in the targeted policy make a transition to the targeted policy when the application starts. For example, at system boot, init runs under the unconfined_t policy. When named starts, it makes a transition to the named_t domain and is locked down by the appropriate policy.

For more information on enabling or disabling targeted policy on each of the specific daemons, refer to How to use system-config-securitylevel.

More information on the different policies available in SELinux can be found at http://fedoraproject.org/wiki/SELinux/Policies.

Currently, the list of programs is approximately:

accton, amanda, httpd (apache), arpwatch, pam, automount, avahi, named, bluez, lilo, grub, canna, comsat, cpucontrol, cpuspeed, cups, cvs, cyrus, dbskkd, dbus, dhcpd, dictd, dmidecode, dovecot, fetchmail, fingerd, ftpd (vsftpd, proftpd, and muddleftpd), gpm, hald, hotplug, howl, innd, kerberos, ktalkd, openldap, auditd, syslog, logwatch, lpd, lvm, mailman, module-init-tools, mount, mysql, NetworkManager, NIS, nscd, ntp, pegasus, portmap, postfix, postgresql, pppd, pptp, privoxy, procmail, radiusd, radvd, rlogin, nfs, rsync, samba, saslauthd, snmpd, spamd, squid, stunnel, dhcpc, ifconfig, sysstat, tcp wrappers, telnetd, tftpd, updfstab, user management (passwd, useradd, etc.), crack, uucpd, vpnc, webalizer, xend, xfs, zebra

The strict policy does work on Fedora Core. It is challenged by the unique environments of different users. To use the strict policy in your environment, you may need to fine-tune both the policy and your systems.

To make the strict policy easier to use, SELinux developers have tried to make the change from one policy to the other easier. For example, system-config-securitylevel builds a relabel into the startup scripts.

More information on the different policies available in SELinux can be found at http://fedoraproject.org/wiki/SELinux/Policies.

The mls policy is similar to the strict policy, but adds an additional field to security contexts for separating levels. SELinux can use these levels to separate data in an environment that calls for strict hierarchical separation. A typical example is a military setting, where data is classified at a certain level. This policy is geared toward this sort of environment, and is probably not useful to you unless you fall into this category.

More information on the different policies available in SELinux can be found at http://fedoraproject.org/wiki/SELinux/Policies.

The Reference Policy is a new project maintained by Tresys Technology (http://www.tresys.com/) designed to rewrite the entire SELinux policy in a way that is easier to use and understand. To do this, it uses the concepts of modularity, abstraction, and well-defined interfaces. Refer to http://serefpolicy.sourceforge.net/ for more information on the Reference Policy.

Note that Reference Policy is not a new type of policy, like targeted or strict. Rather, it is a new base that policies can be built from. Targeted, strict, and mls policies can all be built from Reference Policy. In fact, one of the design goals of Reference Policy is to have a single unified source tree for the different policy variants.

Fedora policies at version 1.x are based on the traditional example policy. Version 2.x policies (as used in Fedora Core 5) are based on the Reference Policy.

File contexts are used by the setfiles command to generate persistent labels which describe the security context for a file or directory.

Fedora Core ships with the fixfiles script, which supports three options: check, restore, and relabel. This script allows users to relabel the file system without having the selinux-policy-targeted-sources package installed. The command line usage is more friendly than the standard setfiles command.

The new option -Z is the short method for displaying the context of a subject or object:

ls -alZ file.foo 
id -Z 
ps -eZ

There is no difference between a domain and a type, although domain is sometimes used to refer to the type of a process. The use of domain in this way stems from Domain and Type Enforcement (DTE) models, where domains and types are separate.

Prior to Fedora Core 5, SELinux policies were monolithic, meaning making a change required getting the entire policy source, modifying it, compiling it, and replacing the current policy with it. With Fedora Core 5, the policy is now modular. This means that third party developers can ship policy modules with their applications, and then they can be added to the policy without having to switch out the entire policy. The new module is then added to the module store, which results in a new policy binary that is a combination of the previous policy and the new module.

This actually works by separating out compile and link steps in the policy build procedure. Policy modules are compiled from source, and linked when installed into the module store (see Managed Policy). This linked policy is then loaded into the kernel for enforcement.

The primary command for dealing with modules is semodule, which lets you perform basic functions such as installing, upgrading, or removing modules. Other useful commands include checkmodule, which is the module compiler and is installed with the checkpolicy rpm, as well as semodule_package, which creates a policy package file (.pp) from a compiled policy module.

Modules are usually stored as policy package file (.pp extension) in /usr/share/selinux/policyname/. There you should at least find the base.pp, which is the base module.

To see how to write a simple policy module, check out Local Policy Customizations.

Prior to Fedora Core 5, SELinux policies were handled as user-editable config files in etc. Unfortunately, this made it difficult to address many of the usability issues arising with SELinux. So, a new library, libsemanage, was added to provide userspace tools an interface to making policy management easier. All policy management should use this library to access the policy store. The policy store holds all the policy information, and is found at /etc/selinux/policyname/modules/.

You should never have to edit the store directly. Instead, you should use tools that link against libsemanage. One example tool is semanage, which is a command line tool for managing much of the policy such as SELinux user mappings, SELinux port mappings, and file contexts entries. Other examples of tools that use libsemanage include semodule which uses it to manage the SELinux policy modules installed to the policy store and setsebool which uses it manage SELinux policy booleans. Additionally, graphical tools are currently being developed to utilize the functionality provided by libsemanage.

The installer follows the choice you make in the Firewall Configuration screen. The default running policy is the targeted policy, and it is on by default.

The answer might be nothing. There are many Fedora users that don't even realize that they are using SELinux. SELinux provides protection for their systems with an out-of-the-box configuration. That said, there are a couple of things an administrator might want to do to configure their system. These include:

Use system-config-securitylevel, also known as the Security Level Configuration graphical tool, to control the Boolean values of specific daemons. For example, if you need to disable SELinux for Apache to run correctly in your environment, you can disable the value in system-config-securitylevel. This change disables the transition to the policy defined in apache.te, allowing httpd to remain under regular Linux DAC security.

Since Fedora Core 5 uses a modular policy, you don't have to have the complete policy source any more. Now, you can just create a local policy module for your local policy customizations. To do this, follow these steps.

	Module are uniquely identified by name
	This means that if you later insert another local.pp, it will replace the one you just loaded. So, you should keep this local.te around, and just add to it if you need to make later policy customizations. If you lose it, but want to keep your previous policy around, just call the new local policy module something else (say local2.te).

If you have specific AVC messages you can use audit2allow to generate a Type Enforcement file that is ready to load as a policy module.

audit2allow -M local < /tmp/avcs

This creates a local.pp which you can then load into the kernel using semodule -i local.pp. You can also edit the local.te to make additional customizations. To create a module allowing all the denials since the last reboot that you can then customize, execute the following:

audit2allow -m local -l -i /var/log/messages > local.te

Note that the above assumes you are not using the audit daemon. If you were using the audit daemon, then you should use /var/log/audit/audit.log instead of /var/log/messages as your log file. This generates a local.te file, that looks similar to the following:

module local 1.0;

require {
        class file { append execute execute_no_trans getattr ioctl read write };
        type httpd_t;
        type httpd_w3c_script_exec_t;
 };


allow httpd_t httpd_w3c_script_exec_t:file { execute execute_no_trans getattr ioctl read };

You can hand edit this file, removing allow statements that you don't want to allow, and then recompile and reload it using

	Important
	In order to load this newly created policy package into the kernel, you are required to execute semodule -i local.pp Note that if you later install another module called local, it will replace this module. If you want to keep these rules around, then you either need to append future customizations to this local.te, or give future customizations a different name.

Your help is definitely appreciated.

Also, since the Fedora Core 5 policy is based on the Reference Policy, you should look at the documentation on its project page. Another excellent source of information is the example policy files in /usr/share/doc/selinux-policy->version< and /usr/share/selinux/devel.

If you want to create a new policy domain, you can look at the interface files in the /usr/share/selinux/devel sub-directories. There is also a tool there to help you get started. The following procedure is an example:

Now you can collect avc messages. You can use audit2allow to translate the avc messages to allow rules and begin updating your mydaemon.te file. You should search for interface macros in the /usr/share/selinux/devel/include directory and use these instead of using the allow rules directly, whenever possible. audit2allow -R will attempt to find interfaces that match the allow rule. If you want more examples of policy, you could always install the selinux-policy src rpm, which contains all of the policy te files for the reference policy.

	Use caution when switching policy
	Other than trying out a new policy on a test machine for research purposes, you should seriously consider your situation before switching to a different policy on a production system. The act of switching is straightforward. This method is fairly safe, but you should try it first on a test system.

To use the automated method, run the Security Level Configuration tool. From the GUI Main Menu, select Desktop → System Settings → Security level, or from a terminal, run system-config-securitylevel. Change the policy as desired and ensure that the Relabel on next reboot option is enabled.

You can also perform these steps manually with the following procedure:

Use the star utility, which supports the extended attributes that store the security context labels. Specify the -xattr and -H=exustar options when creating archives.

ls -Z /var/log/maillog
-rw-------  root   root    system_u:object_r:var_log_t   /var/log/maillog
cd /var/log
star -xattr -H=exustar -c -f maillog.star ./maillog*

Absolute paths can overwrite existing data

If you use an absolute path, such as /var/log/maillog, when you unpack the archive with star -c -f, the files are restored on the same path they were archived with. The maillog file attempts to write to /var/log/maillog. You should received a warning from star if the files about to be overwritten have a later date, but you cannot rely on this behavior. Consider carefully how you construct your archiving argument.

This process presumes that you have enabled user public HTML directories in your Apache configuration file, /etc/httpd/conf/httpd.conf. This process only covers serving static Web content. For more information about Apache HTTP and SELinux, refer to http://fedora.redhat.com/docs/selinux-apache-fc3/.

Set SELINUX=disabled in /etc/selinux/config.

Alternatively, you can add selinux=0 to your kernel boot parameters. However, this option is not recommended.

	Be careful when disabling SELinux
	If you boot with selinux=0, any files you create while SELinux is disabled do not have SELinux context information. The file system is marked for relabeling at the next boot. If an unforeseen problem prevents you from rebooting normally, you may need to boot in single-user mode for recovery. Add the option emergency to your kernel boot parameters.

You can specify the SELinux mode using the configuration file /etc/sysconfig/selinux.

# This file controls the state of SELinux on the system.
# SELINUX= can take one of these three values:
#       enforcing - SELinux security policy is enforced.
#       permissive - SELinux prints warnings instead of enforcing.
#       disabled - No SELinux policy is loaded.
SELINUX=enforcing
# SELINUXTYPE= type of policy in use. Possible values are:
#       targeted - Only targeted network daemons are protected.
#       strict - Full SELinux protection.
SELINUXTYPE=targeted

Setting the value to enforcing is the same as adding enforcing=1 to the kernel boot parameters. Setting the value to permissive is the same as adding enforcing=0 to the kernel boot parameters.

However, setting the value to disabled is not the same as the selinux=0 kernel boot parameter. Rather than fully disabling SELinux in the kernel, the disabled setting instead turns enforcing off and skips loading a policy.

	SELinux Configuration Precedence
	The command line kernel parameter overrides the configuration file.

Occasionally you may need to perform an action that is normally prevented by policy. Run the command setenforce 0 to turn off enforcing mode in real time. When you are finished, run setenforce 1 to turn enforcing back on.

	sysadm_r Role Required for strict policy
	You must issue the setenforce command with the sysadm_r role if you are using strict policy. If you are using the standard targeted policy, then this is not necessary. Use the newrole command to assume this role.

Add audit=1 to your kernel command line to turn system call auditing on. Add audit=0 to your kernel command line to turn system call auditing off.

System-call auditing is on by default. When on, it provides information about the system call that was executing when SELinux generated a denied message. The error message is helpful when debugging policy.

Run auditctl -e 0. Note that this command does not affect auditing of SELinux AVC denials.

As root, execute the command /usr/sbin/sestatus -v. For more information, refer to the sestatus(8) manual page.

Very few domains in the SELinux world are allowed to read the /etc/shadow file. There are constraint rules that prevent policy writers from writing code like

allow mydomain_t shadow_t:file read;

In RHEL4 you can setup your domain to use the unix_chkpwd command. The easiest way is to use the unix_chkpwd attribute. So if you were writing policy for an ftpd daemon you would write something like

daemon_domain(vsftpd, `auth_chkpwd')

This would create a context where vsftpd_t -> chkpwd_exec_t -> system_chkpwd_t which can read /etc/shadow, while vsftpd_t is not able to read it.

In Fedora Core 5/RHEL5, add the rule

auth_domtrans_chk_passwd(vsftpd_t)

You need to execute the command semodule -i myapp.pp. This modifies the policy that is stored on the machine. Your policy module now is loaded with the rest of the policy. You can even remove the pp file from the system.

semodule -l lists the currently loaded modules.

#semodule -i 
myapp   1.2.1

If you later would like to remove the policy package, you can execute semodule -r myapp.

In Fedora Core 2 and 3, SELinux AVC messages could be found in /var/log/messages. In Fedora Core 4, the audit daemon was added, and these messages moved to /var/log/audit/audit.log. In Fedora Core 5, the audit daemon is not installed by default, and consequently these messages can be found in /var/log/messages unless you choose to install and enable the audit daemon, in which case AVC messages will be in /var/log/audit/audit.log.

This message means that the current SELinux policy is not allowing the application to do something. There are a number of reasons this could happen.

First, one of the files the application is trying to access could be mislabeled. If the AVC message refers to a specific file, inspect its current label with ls -alZ /path/to/file. If it seems wrong, use the command restorecon -v /path/to/file to restore the file's default context. If you have a large number of denials related to files, you may want to use fixfiles relabel, or run restorecon -R /path to recursively relabel a directory path.

Denials are sometimes due to a configuration change in the program that triggered the denial message. For example, if you change Apache to also listen on port 8800, you must also change the security policy, apache.te. Refer to External Link List for more information about writing policy.

If you are having trouble getting a specific application like Apache to work, refer to How to use system-config-securitylevel for information on disabling enforcement just for that application.

Your /home partition is not labeled correctly. You can easily fix this two different ways.

If you just want to relabel /home recursively:

/sbin/restorecon -v -R /home

If you want to be sure there are no other files incorrectly labeled, you can relabel the entire file system:

/sbin/fixfiles relabel

You must have the policycoreutils package installed to use fixfiles.

You can read the files from a non-SELinux distribution, or one with SELinux disabled. However, files created by a system not using SELinux systems do not have a security context, nor do any files you remove and recreate. This could be a challenge with files such as ~/.bashrc. You may have to relabel /home when you reboot the SELinux enabled Fedora Core system.

Just as NFS transparently supports many file system types, it can be used to share directories between SELinux and non-SELinux systems.

When you mount a non-SELinux file system via NFS, by default SELinux treats all the files in the share as having a context of nfs_t. You can override the default context by setting it manually, using the context= option. The following command makes the files in the NFS mounted directory appear to have a context of system_u:object_r:tmp_t to SELinux:

mount -t nfs -o context=system_u:object_r:tmp_t server:/shared/foo /mnt/foo

When SELinux exports a file system via NFS, newly created files have the context of the directory they were created in. In other words, the presence of SELinux on the remote mounting system has no effect on the local security contexts.

You can create your new user with the standard useradd command. First you must become root. Under the strict policy you need to change role to sysadm_r with the following command:

newrole -r sysadm_r

For the targeted policy you do not need to switch roles, staying in unconfined_t:

su - root
id -Z
root:system_r:unconfined_t
useradd auser
ls -Z /home
drwx------  auser   auser   root:object_r:user_home_dir_t /home/auser

The initial context for a new user directory has an identity of root. Subsequent relabeling of the file system changes the identity to system_u. These are functionally the same since the role and type are identical (object_r:user_home_dir_t.)

In previous versions of Fedora Core, security context transitions were integrated into the su via pam_selinux. This turned out to be more trouble than it was worth, and is quite unnecessary on a system running targeted policy. So, this is no longer the case. Now, su/sudo only change the Linux identy. You will need to use newrole to change the SELinux identity, role, or level.

Other forms of Linux/UNIX® identity change, for example setuid(2), also do not cause an SELinux identity change.

If you wanted to not audit dmesg, for example, you would put this in your dmesg.te file:

dontaudit dmesg_t userdomain:fd { use };

This eliminates the error output to the terminal for all user domains, including user, staff and sysadm.

In a non-enforcing mode, you should actually receive more messages than in enforcing mode. The kernel logs each access denial as if you were in an enforcing mode. Since you are not restricted by policy enforcement, you can perform more actions, which results in more denials being logged.

If an application running under an enforcing mode is denied access to read a number of files in a directory, it is stopped once at the beginning of the action. In a non-enforcing mode, the application is not stopped from traversing the directory tree, and generates a denial message for each file read in the directory.

The most common reason for a silent denial is when the policy contains an explicit dontaudit rule to suppress audit messages. The dontaudit rule is often used this way when a benign denial is filling the audit logs.

To look for your particular denial, enable auditing of all dontaudit rules:

semodule -b /usr/share/selinux/targeted/enableaudit.pp

	Enabled dontaudit output is verbose
	Enabling auditing of all dontaudit rules likely produce a large amount of audit information, most of which is irrelevant to your denial. Use this technique only if you are specifically looking for an audit message for a denial that seems to occur silently. You want to re-enable dontaudit rules as soon as possible.

Once you have found your problem you can reset to the default mode by executing

semodule -b /usr/share/selinux/targeted/base.pp

SELinux intentionally disables access to the tty devices to stop daemons from communicating back with the controlling terminal. This communication is a potential security hole because such daemons could insert commands into the controlling terminal. A broken or compromised program could use this hole to cause serious problems.

There are a few ways you can capture standard output from daemons. One method is to pipe the output to the cat command.

snmpd -v | cat

When debugging a daemon, you may want to turn off the transition of the daemon to its specific domain. You can do this using system-config-securitylevel or setsebool on the command line.

A final option is to turn off enforcing mode while debugging. Issue the command setenforce 0 to turn off enforcing mode, and use the command setenforce 1 to re-enable SELinux when you are finished debugging.

Policy reloads itself when the package is updated. This behavior replaces the manual make load.

In certain situations, you may need to relabel the file system. This might occur as part of an SELinux bug fix where file contexts become invalid, or when the policy update makes changes to the file /etc/selinux/targeted/contexts/files/file_contexts.

After the file system is relabeled, a reboot is not required, but is useful in ensuring every process and program is running in the proper domain. This is highly dependent on the changes in the updated policy.

To relabel, you have several options. You may use the fixfiles command:

fixfiles relabel
reboot

Alternately, use the /.autorelabel mechanism:

touch /.autorelabel
reboot

Yes. The security contexts for the files owned by the package are stored in the header data for the package. The file contexts are set directly after the cpio copy, as the package files are being put on the disk.

When you install a policy package, pre-compiled binary policy files are put directly into /etc/selinux. The different build environments will make target files that have different sizes and MD5 checksums.

There is a possibility that changes in the policy package or in the policy shipping with an application package can cause errors, more denials, or other unknown behaviors. You can discover which package caused the breakage by reverting policy and application packages one at a time. If you don't want to return to the previous package, the older version of the configuration files will be saved with the extension .rpmsave. Use the mailing lists, bugzilla, and IRC to help you work through your problem. If you are able, write or fix policy to resolve your problem.

To regain useful control, turn off kernel messages to the console with this command:

dmesg -n 1

You can test SELinux default policy by installing just the selinux-policy-policyname and policycoreutils packages. Without the policy source installed, the fixfiles command automates the file system relabeling.

The command fixfiles relabel is the equivalent of make relabel. During the relabeling, it will delete all of the files in /tmp, cleaning up files which may have old file context labels.

Other commands are fixfiles check, which checks for mislabeled files, and fixfiles restore, which fixes the mislabeled files but does not delete the files in /tmp. The fixfiles command does not take a list of directories as an argument, because it relabels the entire file system. If you need to relabel a specific directory path, use restorecon.

KDE executables always appear as kdeinit, which limits what can be done with SELinux policy. This is because every KDE application runs in the domain for kdeinit.

Problems often arise when installing SELinux because it is not possible to relabel /tmp and /var/tmp. There is no good method of determining which file should have which context.

The solution is to fully log out of KDE and remove all KDE temporary files:

rm -rf	/var/tmp/kdecache-<username>
rm -rf /var/tmp/<other_kde_files>

At your next login, your problem should be fixed.

Be careful of white space in the file /etc/sysconfig/selinux. The code is very sensitive to white space, even trailing space.

We have begun to confine the unconfined_t domain somewhat. SELinux restricts certain memory protection operation. Following is a list of those denials, as well as possible reasons and solutions for those denials. For more information on these restrictions, see http://people.redhat.com/drepper/selinux-mem.html.

These show up in /var/log/messages (or /var/log/audit/audit.log if using the audit daemon) as avc denials. These can also show up when running programs with errors like

error while loading shared libraries: /usr/lib/libavutil.so.49:
cannot restore segment prot after reloc: Permission denied

which indicates that the library is trying to perform a text relocation and failing. Text relocations are bad, but can be allowed via the first hint below. Below are the SELinux memory permissions that are denied, as well as hints at how to address these denials.

restorecon reset /etc/modprobe.conf context system_u:object_r:etc_runtime_t->system_u:object_r:modules_conf_t
restorecon reset /etc/cups/ppd/homehp.ppd context user_u:object_r:cupsd_etc_t->system_u:object_r:cupsd_rw_etc_t

During the update process, the selinux package runs restorecon on the difference between the previously install policy file_context and the newly install policy context. This maintains the correct file context on disk.

libsepol.sepol_genbools_array: boolean hidd_disable_trans no longer in policy

This indicates that the updated policy has removed the boolean from policy.

You can use the semanage command to define additional ports. So say you want httpd to be able to listen on port 8082. You could enter the command.

semanage port -a -p tcp -t http_port_t 8082

Translations are handled through libsemanage. Use semanage translation -l to list all current translations.

# semanage translation -l
Level                     Translation

s0
s0-s0:c0.c255             SystemLow-SystemHigh
s0:c0.c255                SystemHigh

Now pick an unused category. Say you wanted to add Payroll as a translation, and s0:c6 is unused.

# semanage translation -a -T Payroll s0:c6
# semanage translation -l
Level                     Translation

s0
s0-s0:c0.c255             SystemLow-SystemHigh
s0:c0.c255                SystemHigh
s0:c6                     Payroll

You can modify the range of categories a user can login with by using semanage, as seen in this example.

# semanage login -a -r s0-Payroll csellers
# semanage login -l

Login Name                SELinux User              MLS/MCS Range            

__default__               user_u                    s0                       
csellers                  user_u                    s0-Payroll               
root                      root                      SystemLow-SystemHigh

In the above example, the user csellers was given access to the Payroll category with the first command, as indicated in the listing output from the second command.

First, you should never allow a system service to execute anything it can write. This gives an attacker the ability to upload malicious code to the server and then execute it, which is something we want to prevent.

If you merely need to allow your script to create (non-executable) files, this is possible. That said, you should avoid having system applications writing to the /tmp directory, since users tend to use the /tmp directory also. It would be better to create a directory elsewhere which could be owned by the apache process and allow your script to write to it. You should label the directory httpd_sys_script_rw_t, which will allow apache to read and write files to that directory. This directory could be located anywhere that apache can get to (even $HOME/public_html/).

You need to identify the swapfile to SELinux by setting its file context to swapfile_t.

chcon -t swapfile_t SWAPFILE

For files, relabelfrom means "Can domain D relabel a file from (i.e. currently in) type T1?" and relabelto means "Can domain D relabel a file to type T2?", so both checks are applied upon a file relabeling, where T1 is the original type of the type and T2 is the new type specified by the program.

Useful documents to look at:

The file system must support xattr labels in the right security.* namespace. In addition to ext2/ext3, XFS has recently added support for the necessary labels.

Note that XFS SELinux support is broken in upstream kernel 2.6.14 and 2.6.15, but fixed (worked around) in 2.6.16. Your kernel must include this fix if you choose to use XFS with SELinux.