Permissions

The Unix operating system (and likewise, Linux) differs from other computing environments in that it is not only a multitasking system but it is also a multi-user system as well.

What exactly does this mean? It means that more than one user can be operating the computer at the same time. While your computer will only have one keyboard and monitor, it can still be used by more than one user. For example, if your computer is attached to a network, or the Internet, remote users can log in via telnet or ssh (secure shell) and operate the computer. In fact, remote users can execute X applications and have the graphical output displayed on a remote computer. The X Windows system supports this.

The multi-user capability of Unix is not a recent "innovation," but rather a feature that is deeply ingrained into the design of the operating system. If you remember the environment in which Unix was created, this makes perfect sense. Years ago before computers were "personal," they were large, expensive, and centralized. A typical university computer system consisted of a large mainframe computer located in some building on campus and terminals were located throughout the campus, each connected to the large central computer. The computer would support many users at the same time.

In order to make this practical, a method had to be devised to protect the users from each other. After all, you could not allow the actions of one user to crash the computer, nor could you allow one user to interfere with the files belonging to another user.

This lesson will cover the following commands:

• chmod - modify file access rights
• su - temporarily become the superuser
• chown - change file ownership
• chgrp - change a file's group ownership

File permissions

Linux uses the same permissions scheme as Unix. Each file and directory on your system is assigned access rights for the owner of the file, the members of a group of related users, and everybody else. Rights can be assigned to read a file, to write a file, and to execute a file (i.e., run the file as a program).

To see the permission settings for a file, we can use the ls command as follows:

-rw-rw-r-- 1 me   me   1097374 Sep 26 18:48 some_file

We can determine a lot from examining the results of this command:

• The file "some_file" is owned by user "me"
• User "me" has the right to read and write this file
• The file is owned by the group "me"
• Members of the group "me" can also read and write this file
• Everybody else can read this file

Let's try another example. We will look at the bash program which is located in the /bin directory:

-rwxr-xr-x 1 root root  316848 Feb 27  2000 /bin/bash

Here we can see:

• The file "/bin/bash" is owned by user "root"
• The superuser has the right to read, write, and execute this file
• The file is owned by the group "root"
• Members of the group "root" can also read and execute this file
• Everybody else can read and execute this file

In the diagram below, we see how the first portion of the listing is interpreted. It consists of a character indicating the file type, followed by three sets of three characters that convey the reading, writing and execution permission for the owner, group, and everybody else.

chmod

The chmod command is used to change the permissions of a file or directory. To use it, you specify the desired permission settings and the file or files that you wish to modify. There are two ways to specify the permissions, but I am only going to teach one way.

It is easy to think of the permission settings as a series of bits (which is how the computer thinks about them). Here's how it works:

rwx rwx rwx = 111 111 111
rw- rw- rw- = 110 110 110
rwx --- --- = 111 000 000

and so on...

rwx = 111 in binary = 7
rw- = 110 in binary = 6
r-x = 101 in binary = 5
r-- = 100 in binary = 4

Now, if you represent each of the three sets of permissions (owner, group, and other) as a single digit, you have a pretty convenient way of expressing the possible permissions settings. For example, if we wanted to set some_file to have read and write permission for the owner, but wanted to keep the file private from others, we would:

Here is a table of numbers that covers all the common settings. The ones beginning with "7" are used with programs (since they enable execution) and the rest are for other kinds of files.

Directory permissions

The chmod command can also be used to control the access permissions for directories. In most ways, the permissions scheme for directories works the same way as they do with files. However, the execution permission is used in a different way. It provides control for access to file listing and other things. Here are some useful settings for directories:

Becoming the superuser for a short while

It is often useful to become the superuser to perform important system administration tasks, but as you have been warned (and not just by me!), you should not stay logged on as the superuser. In most distributions, there is a program that can give you temporary access to the superuser's privileges. This program is called su (short for substitute user) and can be used in those cases when you need to be the superuser for a small number of tasks. To become the superuser, simply type the su command. You will be prompted for the superuser's password:

After executing the su command, you have a new shell session as the superuser. To exit the superuser session, type exit and you will return to your previous session.

In some distributions, most notably Ubuntu, an alternate method is used. Rather than using su, these systems employ the sudo command instead. With sudo, one or more users are granted superuser privileges on an as needed basis. To execute a command as the superuser, the desired command is simply preceeded with the sudo command. After the command is entered, the user is prompted for the user's password rather than the superuser's:

Changing file ownership

You can change the owner of a file by using the chown command. Here's an example: Suppose I wanted to change the owner of some_file from "me" to "you". I could:

Notice that in order to change the owner of a file, you must be the superuser. To do this, our example employed the su command, then we executed chown, and finally we typed exit to return to our previous session.

chown works the same way on directories as it does on files.

Changing group ownership

The group ownership of a file or directory may be changed with chgrp. This command is used like this:

In the example above, we changed the group ownership of some_file from its previous group to "new_group". You must be the owner of the file or directory to perform a chgrp.

Job Control

In the previous lesson, we looked at some of the implications of Linux being a multi-user operating system. In this lesson, we will examine the multitasking nature of Linux, and how this is manipulated with the command line interface.

As with any multitasking operating system, Linux executes multiple, simultaneous processes. Well, they appear simultaneous, anyway. Actually, a single processor computer can only execute one process at time but the Linux kernel manages to give each process its turn at the processor and each appears to be running at the same time.

There are several commands that can be used to control processes. They are:

• ps - list the processes running on the system
• kill - send a signal to one or more processes (usually to "kill" a process)
• jobs - an alternate way of listing your own processes
• bg - put a process in the background
• fg - put a process in the forground

A practical example

While it may seem that this subject is rather obscure, it can be very practical for the average user who mostly works with the graphical user interface. You might not know this, but most (if not all) of the graphical programs can be launched from the command line. Here's an example: there is a small program supplied with the X Windows system called xload which displays a graph representing system load. You can excute this program by typing the following:

Notice that the small xload window appears and begins to display the system load graph. Notice also that your prompt did not reappear after the program launched. The shell is waiting for the program to finish before control returns to you. If you close the xload window, the xload program terminates and the prompt returns.

Putting a program in the background

Now, in order to make life a little easier, we are going to launch the xload program again, but this time we will put it in the background so that the prompt will return. To do this, we execute xload like this:

In this case, the prompt returned because the process was put in the background.

Now imagine that you forgot to use the "&" symbol to put the program into the background. There is still hope. You can type control-z and the process will be suspended. The process still exists, but is idle. To resume the process in the background, type the bg command (short for background). Here is an example:

Listing your processes

Now that we have a process in the background, it would be helpful to display a list of the processes we have launched. To do this, we can use either the jobs command or the more powerful ps command.

Killing a process

Suppose that you have a program that becomes unresponsive (hmmm...Netscape comes to mind ; how do you get rid of it? You use the kill command, of course. Let's try this out on xload. First, you need to identify the process you want to kill. You can use either jobs or ps, to do this. If you use jobs you will get back a job number. With ps, you are given a process id (PID). We will do it both ways:

A little more about kill

While the kill command is used to "kill" processes, its real purpose is to send signals to processes. Most of the time the signal is intended to tell the process to go away, but there is more to it than that. Programs (if they are properly written) listen for signals from the operating system and respond to them, most often to allow some graceful method of terminating. For example, a text editor might listen for any signal that indicates that the user is logging off, or that the computer is shutting down. When it receives this signal, it saves the work in progress before it exits. The kill command can send a variety of signals to processes. Typing:

kill -l

will give you a list of the signals it supports. Most are rather obscure, but several are useful to know:

Now let's suppose that you have a program that is hopelessly hung (Netscape, maybe) and you want to get rid of it. Here's what you do:

1. Use the ps command to get the process id (PID) of the process you want to terminate.
2. Issue a kill command for that PID.
3. If the process refuses to terminate (i.e., it is ignoring the signal), send increasingly harsh signals until it does terminate.

In the example above I used the kill command in the formal way. In actual practice, it is more common to do it in the following way since the default signal sent by kill is SIGTERM and kill can also use the signal number instead of the signal name:

Then, if the process does not terminate, force it with the SIGKILL signal:

That's it!

This concludes the "Learning the shell" series of lessons. In the next series, "Writing shell scripts," we will look at how to automate tasks with the shell.