The Linux Programmer’s Guide

Sven Goldt
Sven van der Meer
Scott Burkett
Matt Welsh

Version 0.4
March 1995

0
...Our continuing mission: to seek out knowledge of C, to explore strange unix
commands, and to boldly code where no one has man page 4.
2
Contents

1 The Linux operating system 7

2 The Linux kernel 9

3 The Linux libc package 11

4 System calls 13

5 The “swiss army knife” ioctl 15

6 Linux Interprocess Communications 17

6.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
6.2 Half-duplex UNIX Pipes . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
6.2.1 Basic Concepts . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
6.2.2 Creating Pipes in C . . . . . . . . . . . . . . . . . . . . . . . . . . 19
6.2.3 Pipes the Easy Way! . . . . . . . . . . . . . . . . . . . . . . . . . 23
6.2.4 Atomic Operations with Pipes . . . . . . . . . . . . . . . . . . . . 27
6.2.5 Notes on half-duplex pipes: . . . . . . . . . . . . . . . . . . . . . 27
6.3 Named Pipes (FIFOs - First In First Out) . . . . . . . . . . . . . . . . . . . 27
6.3.1 Basic Concepts . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
6.3.2 Creating a FIFO . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
6.3.3 FIFO Operations . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
6.3.4 Blocking Actions on a FIFO . . . . . . . . . . . . . . . . . . . . . 30
6.3.5 The Infamous SIGPIPE Signal . . . . . . . . . . . . . . . . . . . . 30
6.4 System V IPC . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
6.4.1 Fundamental Concepts . . . . . . . . . . . . . . . . . . . . . . . . 30
6.4.2 Message Queues . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
6.4.3 Semaphores . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46
6.4.4 Shared Memory . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62

7 Sound Programming 69
7.1 Programming the internal speaker . . . . . . . . . . . . . . . . . . . . . . 69
7.2 Programming a sound card . . . . . . . . . . . . . . . . . . . . . . . . . . 69

8 Character Cell Graphics 71

8.1 I/O Function in libc . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72
8.1.1 Formatted Output . . . . . . . . . . . . . . . . . . . . . . . . . . . 72
8.1.2 Formatted Input . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73
8.2 The Termcap Library . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74
8.2.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74
8.2.2 Find a Terminal Description . . . . . . . . . . . . . . . . . . . . . 75
8.2.3 Look at a Terminal Description . . . . . . . . . . . . . . . . . . . 75
8.2.4 Termcap Capabilities . . . . . . . . . . . . . . . . . . . . . . . . . 76

3
4 CONTENTS

8.3 Ncurses - Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80

8.4 Initializing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82
8.5 Windows . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82
8.6 Output . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85
8.6.1 Formatted Output . . . . . . . . . . . . . . . . . . . . . . . . . . . 86
8.6.2 Insert Characters/Lines . . . . . . . . . . . . . . . . . . . . . . . . 86
8.6.3 Delete Characters/Lines . . . . . . . . . . . . . . . . . . . . . . . 86
8.6.4 Boxes and Lines . . . . . . . . . . . . . . . . . . . . . . . . . . . 87
8.6.5 Background Character . . . . . . . . . . . . . . . . . . . . . . . . 88
8.7 Input . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88
8.7.1 Formated Input . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89
8.8 Options . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89
8.8.1 Input Options . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90
8.8.2 Terminal Attributes . . . . . . . . . . . . . . . . . . . . . . . . . . 91
8.8.3 Use Options . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92
8.9 Clear Window and Lines . . . . . . . . . . . . . . . . . . . . . . . . . . . 93
8.10 Updating the Terminal . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94
8.11 Video Attributes and Color . . . . . . . . . . . . . . . . . . . . . . . . . . 95
8.12 Cursor and Window Coordinates . . . . . . . . . . . . . . . . . . . . . . . 98
8.13 Scrolling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99
8.14 Pads . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100
8.15 Soft-labels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101
8.16 Miscellaneous . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101
8.17 Low-level Access . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 102
8.18 Screen Dump . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 102
8.19 Termcap Emulation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 102
8.20 Terminfo Functions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103
8.21 Debug Function . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 104
8.22 Terminfo Capabilities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 104
8.22.1 Boolean Capabilities . . . . . . . . . . . . . . . . . . . . . . . . . 104
8.22.2 Numbers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105
8.22.3 Strings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105
8.23 [N]Curses Function Overview . . . . . . . . . . . . . . . . . . . . . . . . 112

9 Programming I/O ports 115

9.1 Mouse Programming . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 116
9.2 Modem Programming . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 117
9.3 Printer Programming . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 117
9.4 Joystick Programming . . . . . . . . . . . . . . . . . . . . . . . . . . . . 117

10 Porting Applications to Linux 119

10.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 119
10.2 Signal handling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 119
10.2.1 Signals under SVR4, BSD, and POSIX.1 . . . . . . . . . . . . . . 120
10.2.2 Linux signal options . . . . . . . . . . . . . . . . . . . . . . . . . 120
10.2.3 signal under Linux . . . . . . . . . . . . . . . . . . . . . . . . . . 121
10.2.4 Signals supported by Linux . . . . . . . . . . . . . . . . . . . . . 121
10.3 Terminal I/O . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 121
10.4 Process information and control . . . . . . . . . . . . . . . . . . . . . . . 122
10.4.1 kvm routines . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 122
10.4.2 ptrace and the /proc filesystem . . . . . . . . . . . . . . . . . . . . 122
10.4.3 Process control under Linux . . . . . . . . . . . . . . . . . . . . . 122
10.5 Portable conditional compilation . . . . . . . . . . . . . . . . . . . . . . . 123
10.6 Additional Comments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 124
CONTENTS 5

11 Systemcalls in alphabetical order 125

12 Abbreviations 131
• Copyright
The Linux Programmer’s Guide is c 1994, 1995 by Sven Goldt
Sven Goldt, Sachsendamm 47b, 10829 Berlin, Germany
< [email protected] − berlin.de > .
Chapter 8 is
c 1994, 1995 by Sven van der Meer < [email protected] − berlin.de > .
Chapter 6 is
c 1995 Scott Burkett < scottb@IntN et.net > .
Chapter 10 is c 1994, 1995 Matt Welsh < [email protected] > .
Special thanks goes to John D. Harper < [email protected] > for proofreading
this guide.
Permission to reproduce this document in whole or in part is subject to the following
conditions:
1. The copyright notice remains intact and is included.
2. If you make money with it the authors want a share.
3. The authors are not responsible for any harm that might arise by the use of it.

• Preface
This guide is far from being complete.
The first release started at version 0.1 in September 1994. It concentrated on system
calls because of lack of manpower and information. Planned are the description of
library functions and major kernel changes as well as excursions into important areas
like networking, sound, graphics and asynchronous I/O. Maybe some hints about
how to build shared libraries and pointers to useful toolkits will later be included.
This guide will only be a success with generous help in the form of information or
perhaps even submission of whole chapters.
• Introduction
Once upon a time I installed Linux on my PC to learn more about system administra-
tion. I tried to install a slip server but it didn’t work with shadow and mgetty. I had
to patch sliplogin and it worked until the new Linux 1.1 releases. No one could tell
me what had happened. There was no documentation about changes since the 0.99
kernel except the kernel change summaries from Russ Nelson, but they didn’t help
me very much in solving problems.
The Linux Programmer’s Guide is meant to do what the name implies— It is to help
Linux programmers understand the peculiarities of Linux. By its nature, this also
means that it should be useful when porting programs from other operating systems
to Linux. Therefore, this guide must describe the system calls and the major kernel
changes which have effects on older programs like serial I/O and networking.

Sven Goldt The Linux Programmer’s Guide

6
Chapter 1

The Linux operating system

In March 1991 Linus Benedict Torvalds bought the multitasking system Minix for his AT
386. He used it to develop his own multitasking system which he called Linux. In Septem-
ber 1991 he released the first prototype by e-mail to some other Minix users on the internet,
thus beginning the Linux project. Many programmers from that point on have supported
Linux. They have added device drivers, developed applications, and aimed for POSIX
compliance. Today Linux is very powerful, but what is best is that it’s free. Work is beeing
done to port Linux to other platforms.

7
8 CHAPTER 1. THE LINUX OPERATING SYSTEM
Chapter 2

The Linux kernel

The base of Linux is the kernel. You could replace each and every library, but as long
as the Linux kernel remained, it would still be Linux. The kernel contains device drivers,
memory management, process management and communication management. The kernel
hacker gurus follow POSIX guidelines which sometimes makes programming easier and
sometimes harder. If your program behaves differently on a new Linux kernel release,
chances are that a new POSIX guideline has been implemented. For programming infor-
mation about the Linux kernel, read the Linux Kernel Hacker’s Guide.

9
10 CHAPTER 2. THE LINUX KERNEL
Chapter 3

The Linux libc package

libc: ISO 8859.1, < linux/param.h >, YP functions, crypt functions,

some basic shadow routines (by default not included), ...
old routines for compatibility in libcompat (by default not activated),
english, french or german error messages,
bsd 4.4lite compatible screen handling routines in libcurses,
bsd compatible routines in libbsd, screen handling routines in libtermcap,
database management routines in libdbm,
mathematic routines in libm, entry to execute programs in crt0.o ???,
byte sex information in libieee ??? (could someone give some infos
instead of laughing ?), user space profiling in libgmon.
I wish someone of the Linux libc developers would write this chapter.
All i can say now that there is going to be a change from the
a.out executable format to the elf (executable and linkable format)
which also means a change in building shared libraries.
Currently both formats (a.out and elf) are supported.
Most parts of the Linux libc package are under the Library GNU Public License, though
some are under a special exception copyright like crt0.o. For commercial binary distribu-
tions this means a restriction that forbids statically linked executables. Dynamically linked
executables are again a special exception and Richard Stallman of the FSF said:
[...] But it seems to me that we should unambiguously permit distribution of a dynam-
ically linked executable *without* accompanying libraries, provided that the object files
that make up the executable are themselves unrestricted according to section 5 [...] So I’ll
make the decision now to permit this. Actually updating the LGPL will have to wait for
when I have time to make and check a new version.
Sven Goldt The Linux Programmer’s Guide

11
12 CHAPTER 3. THE LINUX LIBC PACKAGE
Chapter 4

System calls

A system call is usually a request to the operating system (kernel) to do a hardware/system-

specific or privileged operation. As of Linux-1.2, 140 system calls have been defined.
System calls like close() are implemented in the Linux libc. This implementation often
involves calling a macro which eventually calls syscall(). Parameters passed to syscall()
are the number of the system call followed by the needed arguments. The actual system
call numbers can be found in < linux/unistd.h > while < sys/syscall.h > gets updated
with a new libc. If new calls appear that don’t have a stub in libc yet, you can use syscall().
As an example, you can close a file using syscall() like this (not advised):

#include <syscall.h>

extern int syscall(int, ...);

int my_close(int filedescriptor)

{
return syscall(SYS_close, filedescriptor);
}

On the i386 architecture, system calls are limited to 5 arguments besides the system
call number because of the number of hardware registers. If you use Linux on another
architecture you can check < asm/unistd.h > for the syscall macros to see how many
arguments your hardware supports or how many the developers chose to support. These
syscall macros can be used instead of syscall(), but this is not recommended since such
a macro expands to a full function which might already exist in a library. Therefore, only
kernel hackers should play with the syscall macros. To demonstrate, here is the close()
example using a syscall macro.

#include <linux/unistd.h>

_syscall1(int, close, int, filedescriptor);

The syscall1 macro expands revealing the close() function. Thus we have close()
twice–once in libc and once in our program. The return value of syscall() or a syscall
macro is -1 if the system call failed and 0 or greater on success. Take a look at the global
variable errno to see what happened if a system call failed.
The following system calls that are available on BSD and SYS V are not available on
Linux:
audit(), auditon(), auditsvc(), fchroot(), getauid(), getdents(), getmsg(), mincore(), poll(),
putmsg(), setaudit(), setauid().
Sven Goldt The Linux Programmer’s Guide

13
14 CHAPTER 4. SYSTEM CALLS
Chapter 5

The “swiss army knife” ioctl

ioctl stands for input/output control and is used to manipulate a character device via a
filedescriptor. The format of ioctl is
ioctl(unsigned int fd, unsigned int request, unsigned long argument).
The return value is -1 if an error occured and a value greater or equal than 0 if the request
succeeded just like other system calls. The kernel distinguishes special and regular files.
Special files are mainly found in /dev and /proc. They differ from regular files in that way
that they hide an interface to a driver and not to a real (regular) file that contains text or
binary data. This is the UNIX philosophy and allows to use normal read/write operations
on every file. But if you need to do more with a special file or a regular file you can do it
with ... yes, ioctl. You more often need ioctl for special files than for regular files, but it’s
possible to use ioctl on regular files as well.

15
16 CHAPTER 5. THE “SWISS ARMY KNIFE” IOCTL
Chapter 6

Linux Interprocess
Communications

B. Scott Burkett, [email protected] v1.0, 29 March 1995

6.1 Introduction
The Linux IPC (Inter-process communication) facilities provide a method for multiple pro-
cesses to communicate with one another. There are several methods of IPC available to
Linux C programmers:

• Half-duplex UNIX pipes

• FIFOs (named pipes)

• SYSV style message queues

• SYSV style semaphore sets

• SYSV style shared memory segments

• Networking sockets (Berkeley style) (not covered in this paper)

• Full-duplex pipes (STREAMS pipes) (not covered in this paper)

These facilities, when used effectively, provide a solid framework for client/server de-
velopment on any UNIX system (including Linux).

6.2 Half-duplex UNIX Pipes

6.2.1 Basic Concepts
Simply put, a pipe is a method of connecting the standard output of one process to the
standard input of another. Pipes are the eldest of the IPC tools, having been around since
the earliest incarnations of the UNIX operating system. They provide a method of one-way
communications (hence the term half-duplex) between processes.
This feature is widely used, even on the UNIX command line (in the shell).

ls | sort | lp

17
18 CHAPTER 6. LINUX INTERPROCESS COMMUNICATIONS

The above sets up a pipeline, taking the output of ls as the input of sort, and the output
of sort as the input of lp. The data is running through a half duplex pipe, traveling (visually)
left to right through the pipeline.
Although most of us use pipes quite religiously in shell script programming, we often
do so without giving a second thought to what transpires at the kernel level.
When a process creates a pipe, the kernel sets up two file descriptors for use by the
pipe. One descriptor is used to allow a path of input into the pipe (write), while the other
is used to obtain data from the pipe (read). At this point, the pipe is of little practical use,
as the creating process can only use the pipe to communicate with itself. Consider this
representation of a process and the kernel after a pipe has been created:

From the above diagram, it is easy to see how the descriptors are connected together. If
the process sends data through the pipe (fd0), it has the ability to obtain (read) that infor-
mation from fd1. However, there is a much larger objective of the simplistic sketch above.
While a pipe initially connects a process to itself, data traveling through the pipe moves
through the kernel. Under Linux, in particular, pipes are actually represented internally
with a valid inode. Of course, this inode resides within the kernel itself, and not within the
bounds of any physical file system. This particular point will open up some pretty handy
I/O doors for us, as we will see a bit later on.
At this point, the pipe is fairly useless. After all, why go to the trouble of creating a
pipe if we are only going to talk to ourself? At this point, the creating process typically
forks a child process. Since a child process will inherit any open file descriptors from the
parent, we now have the basis for multiprocess communication (between parent and child).
Consider this updated version of our simple sketch:

Above, we see that both processes now have access to the file descriptors which consti-
tute the pipeline. It is at this stage, that a critical decision must be made. In which direction
do we desire data to travel? Does the child process send information to the parent, or vice-
versa? The two processes mutually agree on this issue, and proceed to “close” the end
of the pipe that they are not concerned with. For discussion purposes, let’s say the child
performs some processing, and sends information back through the pipe to the parent. Our
newly revised sketch would appear as such:
6.2. HALF-DUPLEX UNIX PIPES 19

Construction of the pipeline is now complete! The only thing left to do is make use of
the pipe. To access a pipe directly, the same system calls that are used for low-level file I/O
can be used (recall that pipes are actually represented internally as a valid inode).
To send data to the pipe, we use the write() system call, and to retrieve data from the
pipe, we use the read() system call. Remember, low-level file I/O system calls work with
file descriptors! However, keep in mind that certain system calls, such as lseek(), do not
work with descriptors to pipes.

6.2.2 Creating Pipes in C

Creating “pipelines” with the C programming language can be a bit more involved than our
simple shell example. To create a simple pipe with C, we make use of the pipe() system
call. It takes a single argument, which is an array of two integers, and if successful, the
array will contain two new file descriptors to be used for the pipeline. After creating a
pipe, the process typically spawns a new process (remember the child inherits open file
descriptors).

SYSTEM CALL: pipe();

PROTOTYPE: int pipe( int fd[2] );

RETURNS: 0 on success
-1 on error: errno = EMFILE (no free descriptors)
EMFILE (system file table is full)
EFAULT (fd array is not valid)

NOTES: fd[0] is set up for reading, fd[1] is set up for writing

The first integer in the array (element 0) is set up and opened for reading, while the
second integer (element 1) is set up and opened for writing. Visually speaking, the output
of fd1 becomes the input for fd0. Once again, all data traveling through the pipe moves
through the kernel.

#include <stdio.h>
#include <unistd.h>
#include <sys/types.h>

main()
{
int fd[2];

pipe(fd);
.
.
}
20 CHAPTER 6. LINUX INTERPROCESS COMMUNICATIONS

Remember that an array name in C decays into a pointer to its first member. Above,
fd is equivalent to &fd[0]. Once we have established the pipeline, we then fork our new
child process:
#include <stdio.h>
#include <unistd.h>
#include <sys/types.h>

main()
{
int fd[2];
pid_t childpid;

pipe(fd);

if((childpid = fork()) == -1)

{
perror("fork");
exit(1);
}
.
.
}

If the parent wants to receive data from the child, it should close fd1, and the child
should close fd0. If the parent wants to send data to the child, it should close fd0, and
the child should close fd1. Since descriptors are shared between the parent and child, we
should always be sure to close the end of pipe we aren’t concerned with. On a technical
note, the EOF will never be returned if the unnecessary ends of the pipe are not explicitly
closed.
#include <stdio.h>
#include <unistd.h>
#include <sys/types.h>

main()
{
int fd[2];
pid_t childpid;

pipe(fd);

if((childpid = fork()) == -1)

{
perror("fork");
exit(1);
}

if(childpid == 0)
{
/* Child process closes up input side of pipe */
close(fd[0]);
}
else
{
6.2. HALF-DUPLEX UNIX PIPES 21

/* Parent process closes up output side of pipe */

close(fd[1]);
}
.
.
}

As mentioned previously, once the pipeline has been established, the file descriptors
may be treated like descriptors to normal files.

/*****************************************************************************
Excerpt from "Linux Programmer’s Guide - Chapter 6"
(C)opyright 1994-1995, Scott Burkett
*****************************************************************************
MODULE: pipe.c
*****************************************************************************/

#include <stdio.h>
#include <unistd.h>
#include <sys/types.h>

int main(void)
{
int fd[2], nbytes;
pid_t childpid;
char string[] = "Hello, world!\n";
char readbuffer[80];

pipe(fd);

if((childpid = fork()) == -1)

{
perror("fork");
exit(1);
}

if(childpid == 0)
{
/* Child process closes up input side of pipe */
close(fd[0]);

/* Send "string" through the output side of pipe */

write(fd[1], string, strlen(string));
exit(0);
}
else
{
/* Parent process closes up output side of pipe */
close(fd[1]);

/* Read in a string from the pipe */

nbytes = read(fd[0], readbuffer, sizeof(readbuffer));
printf("Received string: %s", readbuffer);
22 CHAPTER 6. LINUX INTERPROCESS COMMUNICATIONS

return(0);
}

Often, the descriptors in the child are duplicated onto standard input or output. The
child can then exec() another program, which inherits the standard streams. Let’s look at
the dup() system call:

SYSTEM CALL: dup();

PROTOTYPE: int dup( int oldfd );

RETURNS: new descriptor on success
-1 on error: errno = EBADF (oldfd is not a valid descriptor)
EBADF (newfd is out of range)
EMFILE (too many descriptors for the process)

NOTES: the old descriptor is not closed! Both may be used interchangeably

Although the old descriptor and the newly created descriptor can be used interchange-
ably, we will typically close one of the standard streams first. The dup() system call uses
the lowest-numbered, unused descriptor for the new one.
Consider:

.
.
childpid = fork();

if(childpid == 0)
{
/* Close up standard input of the child */
close(0);

/* Duplicate the input side of pipe to stdin */

dup(fd[0]);
execlp("sort", "sort", NULL);
.
}

Since file descriptor 0 (stdin) was closed, the call to dup() duplicated the input descrip-
tor of the pipe (fd0) onto its standard input. We then make a call to execlp(), to overlay
the child’s text segment (code) with that of the sort program. Since newly exec’d programs
inherit standard streams from their spawners, it actually inherits the input side of the pipe
as its standard input! Now, anything that the original parent process sends to the pipe, goes
into the sort facility.
There is another system call, dup2(), which can be used as well. This particular call
originated with Version 7 of UNIX, and was carried on through the BSD releases and is
now required by the POSIX standard.

SYSTEM CALL: dup2();

PROTOTYPE: int dup2( int oldfd, int newfd );

RETURNS: new descriptor on success
-1 on error: errno = EBADF (oldfd is not a valid descriptor)
6.2. HALF-DUPLEX UNIX PIPES 23

EBADF (newfd is out of range)

EMFILE (too many descriptors for the process)

NOTES: the old descriptor is closed with dup2()!

With this particular call, we have the close operation, and the actual descriptor dupli-
cation, wrapped up in one system call. In addition, it is guaranteed to be atomic, which
essentially means that it will never be interrupted by an arriving signal. The entire oper-
ation will transpire before returning control to the kernel for signal dispatching. With the
original dup() system call, programmers had to perform a close() operation before call-
ing it. That resulted in two system calls, with a small degree of vulnerability in the brief
amount of time which elapsed between them. If a signal arrived during that brief instance,
the descriptor duplication would fail. Of course, dup2() solves this problem for us.
Consider:
.
.
childpid = fork();

if(childpid == 0)
{
/* Close stdin, duplicate the input side of pipe to stdin */
dup2(0, fd[0]);
execlp("sort", "sort", NULL);
.
.
}

6.2.3 Pipes the Easy Way!

If all of the above ramblings seem like a very round-about way of creating and utilizing
pipes, there is an alternative.
LIBRARY FUNCTION: popen();

PROTOTYPE: FILE *popen ( char *command, char *type);

RETURNS: new file stream on success
NULL on unsuccessful fork() or pipe() call

NOTES: creates a pipe, and performs fork/exec operations using "command"

This standard library function creates a half-duplex pipeline by calling pipe() internally.
It then forks a child process, execs the Bourne shell, and executes the ”command” argument
within the shell. Direction of data flow is determined by the second argument, ”type”. It
can be ”r” or ”w”, for ”read” or ”write”. It cannot be both! Under Linux, the pipe will be
opened up in the mode specified by the first character of the ”type” argument. So, if you
try to pass ”rw”, it will only open it up in ”read” mode.
While this library function performs quite a bit of the dirty work for you, there is a
substantial tradeoff. You lose the fine control you once had by using the pipe() system
call, and handling the fork/exec yourself. However, since the Bourne shell is used directly,
shell metacharacter expansion (including wildcards) is permissible within the ”command”
argument.
Pipes which are created with popen() must be closed with pclose(). By now, you have
probably realized that popen/pclose share a striking resemblance to the standard file stream
I/O functions fopen() and fclose().
24 CHAPTER 6. LINUX INTERPROCESS COMMUNICATIONS

LIBRARY FUNCTION: pclose();

PROTOTYPE: int pclose( FILE *stream );

RETURNS: exit status of wait4() call
-1 if "stream" is not valid, or if wait4() fails

NOTES: waits on the pipe process to terminate, then closes the stream.

The pclose() function performs a wait4() on the process forked by popen(). When it
returns, it destroys the pipe and the file stream. Once again, it is synonymous with the
fclose() function for normal stream-based file I/O.
Consider this example, which opens up a pipe to the sort command, and proceeds to
sort an array of strings:

/*****************************************************************************
Excerpt from "Linux Programmer’s Guide - Chapter 6"
(C)opyright 1994-1995, Scott Burkett
*****************************************************************************
MODULE: popen1.c
*****************************************************************************/

#include <stdio.h>

#define MAXSTRS 5

int main(void)
{
int cntr;
FILE *pipe_fp;
char *strings[MAXSTRS] = { "echo", "bravo", "alpha",
"charlie", "delta"};

/* Create one way pipe line with call to popen() */

if (( pipe_fp = popen("sort", "w")) == NULL)
{
perror("popen");
exit(1);
}

/* Processing loop */
for(cntr=0; cntr<MAXSTRS; cntr++) {
fputs(strings[cntr], pipe_fp);
fputc(’\n’, pipe_fp);
}

/* Close the pipe */

pclose(pipe_fp);

return(0);
}

Since popen() uses the shell to do its bidding, all shell expansion characters and
metacharacters are available for use! In addition, more advanced techniques such as redi-
6.2. HALF-DUPLEX UNIX PIPES 25

rection, and even output piping, can be utilized with popen(). Consider the following
sample calls:

popen("ls ˜scottb", "r");

popen("sort > /tmp/foo", "w");
popen("sort | uniq | more", "w");

As another example of popen(), consider this small program, which opens up two pipes
(one to the ls command, the other to sort):

/*****************************************************************************
Excerpt from "Linux Programmer’s Guide - Chapter 6"
(C)opyright 1994-1995, Scott Burkett
*****************************************************************************
MODULE: popen2.c
*****************************************************************************/

#include <stdio.h>

int main(void)
{
FILE *pipein_fp, *pipeout_fp;
char readbuf[80];

/* Create one way pipe line with call to popen() */

if (( pipein_fp = popen("ls", "r")) == NULL)
{
perror("popen");
exit(1);
}

/* Create one way pipe line with call to popen() */

if (( pipeout_fp = popen("sort", "w")) == NULL)
{
perror("popen");
exit(1);
}

/* Processing loop */
while(fgets(readbuf, 80, pipein_fp))
fputs(readbuf, pipeout_fp);

/* Close the pipes */

pclose(pipein_fp);
pclose(pipeout_fp);

return(0);
}

For our final demonstration of popen(), let’s create a generic program that opens up a
pipeline between a passed command and filename:

/*****************************************************************************
26 CHAPTER 6. LINUX INTERPROCESS COMMUNICATIONS

Excerpt from "Linux Programmer’s Guide - Chapter 6"

(C)opyright 1994-1995, Scott Burkett
*****************************************************************************
MODULE: popen3.c
*****************************************************************************/

#include <stdio.h>

int main(int argc, char *argv[])

{
FILE *pipe_fp, *infile;
char readbuf[80];

if( argc != 3) {
fprintf(stderr, "USAGE: popen3 [command] [filename]\n");
exit(1);
}

/* Open up input file */

if (( infile = fopen(argv[2], "rt")) == NULL)
{
perror("fopen");
exit(1);
}

/* Create one way pipe line with call to popen() */

if (( pipe_fp = popen(argv[1], "w")) == NULL)
{
perror("popen");
exit(1);
}

/* Processing loop */
do {
fgets(readbuf, 80, infile);
if(feof(infile)) break;

fputs(readbuf, pipe_fp);
} while(!feof(infile));

fclose(infile);
pclose(pipe_fp);

return(0);
}

Try this program out, with the following invocations:

popen3 sort popen3.c

popen3 cat popen3.c
popen3 more popen3.c
popen3 cat popen3.c | grep main
6.3. NAMED PIPES (FIFOS - FIRST IN FIRST OUT) 27

6.2.4 Atomic Operations with Pipes

In order for an operation to be considered “atomic”, it must not be interrupted for any
reason at all. The entire operation occurs at once. The POSIX standard dictates in
/usr/include/posix1 lim.h that the maximum buffer size for an atomic operation on a pipe
is:

#define _POSIX_PIPE_BUF 512

Up to 512 bytes can be written or retrieved from a pipe atomically. Anything that
crosses this threshold will be split, and not atomic. Under Linux, however, the atomic
operational limit is defined in “linux/limits.h” as:

#define PIPE_BUF 4096

As you can see, Linux accommodates the minimum number of bytes required by
POSIX, quite considerably I might add. The atomicity of a pipe operation becomes im-
portant when more than one process is involved (FIFOS). For example, if the number of
bytes written to a pipe exceeds the atomic limit for a single operation, and multiple pro-
cesses are writing to the pipe, the data will be “interleaved” or “chunked”. In other words,
one process may insert data into the pipeline between the writes of another.

6.2.5 Notes on half-duplex pipes:

• Two way pipes can be created by opening up two pipes, and properly reassigning the
file descriptors in the child process.

• The pipe() call must be made BEFORE a call to fork(), or the descriptors will not be
inherited by the child! (same for popen()).

• With half-duplex pipes, any connected processes must share a related ancestry. Since
the pipe resides within the confines of the kernel, any process that is not in the ances-
try for the creator of the pipe has no way of addressing it. This is not the case with
named pipes (FIFOS).

6.3 Named Pipes (FIFOs - First In First Out)

6.3.1 Basic Concepts
A named pipe works much like a regular pipe, but does have some noticeable differences.

• Named pipes exist as a device special file in the file system.

• Processes of different ancestry can share data through a named pipe.

• When all I/O is done by sharing processes, the named pipe remains in the file system
for later use.

6.3.2 Creating a FIFO

There are several ways of creating a named pipe. The first two can be done directly from
the shell.

mknod MYFIFO p
mkfifo a=rw MYFIFO
28 CHAPTER 6. LINUX INTERPROCESS COMMUNICATIONS

The above two commands perform identical operations, with one exception. The mk-
fifo command provides a hook for altering the permissions on the FIFO file directly after
creation. With mknod, a quick call to the chmod command will be necessary.
FIFO files can be quickly identified in a physical file system by the “p” indicator seen
here in a long directory listing:

$ ls -l MYFIFO
prw-r--r-- 1 root root 0 Dec 14 22:15 MYFIFO|

Also notice the vertical bar (“pipe sign”) located directly after the file name. Another
great reason to run Linux, eh?
To create a FIFO in C, we can make use of the mknod() system call:

LIBRARY FUNCTION: mknod();

PROTOTYPE: int mknod( char *pathname, mode_t mode, dev_t dev);

RETURNS: 0 on success,
-1 on error: errno = EFAULT (pathname invalid)
EACCES (permission denied)
ENAMETOOLONG (pathname too long)
ENOENT (invalid pathname)
ENOTDIR (invalid pathname)
(see man page for mknod for others)

NOTES: Creates a filesystem node (file, device file, or FIFO)

I will leave a more detailed discussion of mknod() to the man page, but let’s consider a
simple example of FIFO creation from C:

mknod("/tmp/MYFIFO", S_IFIFO|0666, 0);

In this case, the file “/tmp/MYFIFO” is created as a FIFO file. The requested permis-
sions are “0666”, although they are affected by the umask setting as follows:

final_umask = requested_permissions & ˜original_umask

A common trick is to use the umask() system call to temporarily zap the umask value:

umask(0);
mknod("/tmp/MYFIFO", S_IFIFO|0666, 0);

In addition, the third argument to mknod() is ignored unless we are creating a device
file. In that instance, it should specify the major and minor numbers of the device file.

6.3.3 FIFO Operations

I/O operations on a FIFO are essentially the same as for normal pipes, with once major
exception. An “open” system call or library function should be used to physically open up
a channel to the pipe. With half-duplex pipes, this is unnecessary, since the pipe resides
in the kernel and not on a physical filesystem. In our examples, we will treat the pipe as a
stream, opening it up with fopen(), and closing it with fclose().
Consider a simple server process:
6.3. NAMED PIPES (FIFOS - FIRST IN FIRST OUT) 29

/*****************************************************************************
Excerpt from "Linux Programmer’s Guide - Chapter 6"
(C)opyright 1994-1995, Scott Burkett
*****************************************************************************
MODULE: fifoserver.c
*****************************************************************************/

#include <stdio.h>
#include <stdlib.h>
#include <sys/stat.h>
#include <unistd.h>

#include <linux/stat.h>

#define FIFO_FILE "MYFIFO"

int main(void)
{
FILE *fp;
char readbuf[80];

/* Create the FIFO if it does not exist */

umask(0);
mknod(FIFO_FILE, S_IFIFO|0666, 0);

while(1)
{
fp = fopen(FIFO_FILE, "r");
fgets(readbuf, 80, fp);
printf("Received string: %s\n", readbuf);
fclose(fp);
}

return(0);
}

Since a FIFO blocks by default, run the server in the background after you compile it:

$ fifoserver&

We will discuss a FIFO’s blocking action in a moment. First, consider the following
simple client frontend to our server:

/*****************************************************************************
Excerpt from "Linux Programmer’s Guide - Chapter 6"
(C)opyright 1994-1995, Scott Burkett
*****************************************************************************
MODULE: fifoclient.c
*****************************************************************************/

#include <stdio.h>
#include <stdlib.h>

#define FIFO_FILE "MYFIFO"

30 CHAPTER 6. LINUX INTERPROCESS COMMUNICATIONS

int main(int argc, char *argv[])

{
FILE *fp;

if ( argc != 2 ) {
printf("USAGE: fifoclient [string]\n");
exit(1);
}

if((fp = fopen(FIFO_FILE, "w")) == NULL) {

perror("fopen");
exit(1);
}

fputs(argv[1], fp);

fclose(fp);
return(0);
}

6.3.4 Blocking Actions on a FIFO

Normally, blocking occurs on a FIFO. In other words, if the FIFO is opened for reading,
the process will ”block” until some other process opens it for writing. This action works
vice-versa as well. If this behavior is undesirable, the O NONBLOCK flag can be used in
an open() call to disable the default blocking action.
In the case with our simple server, we just shoved it into the background, and let it do
its blocking there. The alternative would be to jump to another virtual console and run the
client end, switching back and forth to see the resulting action.

6.3.5 The Infamous SIGPIPE Signal

On a last note, pipes must have a reader and a writer. If a process tries to write to a pipe
that has no reader, it will be sent the SIGPIPE signal from the kernel. This is imperative
when more than two processes are involved in a pipeline.

6.4 System V IPC

6.4.1 Fundamental Concepts
With System V, AT&T introduced three new forms of IPC facilities (message queues,
semaphores, and shared memory). While the POSIX committee has not yet completed
its standardization of these facilities, most implementations do support these. In addition,
Berkeley (BSD) uses sockets as its primary form of IPC, rather than the System V elements.
Linux has the ability to use both forms of IPC (BSD and System V), although we will not
discuss sockets until a later chapter.
The Linux implementation of System V IPC was authored by Krishna Balasubrama-
nian, at [email protected].

IPC Identifiers
Each IPC object has a unique IPC identifier associated with it. When we say “IPC object”,
we are speaking of a single message queue, semaphore set, or shared memory segment.
6.4. SYSTEM V IPC 31

This identifier is used within the kernel to uniquely identify an IPC object. For example, to
access a particular shared memory segment, the only item you need is the unique ID value
which has been assigned to that segment.
The uniqueness of an identifier is relevant to the type of object in question. To illustrate
this, assume a numeric identifier of “12345”. While there can never be two message queues
with this same identifier, there exists the distinct possibility of a message queue and, say, a
shared memory segment, which have the same numeric identifier.

IPC Keys

To obtain a unique ID, a key must be used. The key must be mutually agreed upon by both
client and server processes. This represents the first step in constructing a client/server
framework for an application.
When you use a telephone to call someone, you must know their number. In addition,
the phone company must know how to relay your outgoing call to its final destination. Once
the other party responds by answering the telephone call, the connection is made.
In the case of System V IPC facilities, the “telephone” correllates directly with the type
of object being used. The “phone company”, or routing method, can be directly associated
with an IPC key.
The key can be the same value every time, by hardcoding a key value into an applica-
tion. This has the disadvantage of the key possibly being in use already. Often, the ftok()
function is used to generate key values for both the client and the server.

LIBRARY FUNCTION: ftok();

PROTOTYPE: key_t ftok ( char *pathname, char proj );

RETURNS: new IPC key value if successful
-1 if unsuccessful, errno set to return of stat() call

The returned key value from ftok() is generated by combining the inode number and
minor device number from the file in argument one, with the one character project inden-
tifier in the second argument. This doesn’t guarantee uniqueness, but an application can
check for collisions and retry the key generation.

key_t mykey;
mykey = ftok("/tmp/myapp", ’a’);

In the above snippet, the directory /tmp/myapp is combined with the one letter iden-
tifier of ’a’. Another common example is to use the current directory:

key_t mykey;
mykey = ftok(".", ’a’);

The key generation algorithm used is completely up to the discretion of the application
programmer. As long as measures are in place to prevent race conditions, deadlocks, etc,
any method is viable. For our demonstration purposes, we will use the ftok() approach. If
we assume that each client process will be running from a unique “home” directory, the
keys generated should suffice for our needs.
The key value, however it is obtained, is used in subsequent IPC system calls to create
or gain access to IPC objects.
32 CHAPTER 6. LINUX INTERPROCESS COMMUNICATIONS

The ipcs Command

The ipcs command can be used to obtain the status of all System V IPC objects. The
Linux version of this tool was also authored by Krishna Balasubramanian.

ipcs -q: Show only message queues

ipcs -s: Show only semaphores
ipcs -m: Show only shared memory
ipcs --help: Additional arguments

By default, all three categories of objects are shown. Consider the following sample
output of ipcs:

------ Shared Memory Segments --------

shmid owner perms bytes nattch status

------ Semaphore Arrays --------

semid owner perms nsems status

------ Message Queues --------

msqid owner perms used-bytes messages
0 root 660 5 1

Here we see a single message queue which has an identifier of “0”. It is owned by the
user root, and has octal permissions of 660, or -rw-rw---. There is one message in the
queue, and that message has a total size of 5 bytes.
The ipcs command is a very powerful tool which provides a peek into the kernel’s
storage mechanisms for IPC objects. Learn it, use it, revere it.

The ipcrm Command

The ipcrm command can be used to remove an IPC object from the kernel. While IPC
objects can be removed via system calls in user code (we’ll see how in a moment), the need
often arises, especially under development environments, to remove IPC objects manually.
Its usage is simple:

ipcrm <msg | sem | shm> <IPC ID>

Simply specify whether the object to be deleted is a message queue (msg), a semaphore
set (sem), or a shared memory segment (shm). The IPC ID can be obtained by the ipcs
command. You have to specify the type of object, since identifiers are unique among the
same type (recall our discussion of this earlier).

6.4.2 Message Queues

Basic Concepts
Message queues can be best described as an internal linked list within the kernel’s address-
ing space. Messages can be sent to the queue in order and retrieved from the queue in
several different ways. Each message queue (of course) is uniquely identified by an IPC
identifier.
6.4. SYSTEM V IPC 33

Internal and User Data Structures

The key to fully understanding such complex topics as System V IPC is to become inti-
mately familiar with the various internal data structures that reside within the confines of
the kernel itself. Direct access to some of these structures is necessary for even the most
primitive operations, while others reside at a much lower level.

Message buffer The first structure we’ll visit is the msgbuf structure. This particular
data structure can be thought of as a template for message data. While it is up to the
programmer to define structures of this type, it is imperative that you understand that there
is actually a structure of type msgbuf. It is declared in linux/msg.h as follows:

/* message buffer for msgsnd and msgrcv calls */

struct msgbuf {
long mtype; /* type of message */
char mtext[1]; /* message text */
};

There are two members in the msgbuf structure:

mtype
The message type, represented in a positive number. This must be a positive number!
mtext
The message data itself.

The ability to assign a given message a type, essentially gives you the capability to
multiplex messages on a single queue. For instance, client processes could be assigned a
magic number, which could be used as the message type for messages sent from a server
process. The server itself could use some other number, which clients could use to send
messages to it. In another scenario, an application could mark error messages as having a
message type of 1, request messages could be type 2, etc. The possibilities are endless.
On another note, do not be misled by the almost too-descriptive name assigned to the
message data element (mtext). This field is not restricted to holding only arrays of char-
acters, but any data, in any form. The field itself is actually completely arbitrary, since this
structure gets redefined by the application programmer. Consider this redefinition:

struct my_msgbuf {
long mtype; /* Message type */
long request_id; /* Request identifier */
struct client info; /* Client information structure */
};

Here we see the message type, as before, but the remainder of the structure has been
replaced by two other elements, one of which is another structure! This is the beauty of
message queues. The kernel makes no translations of data whatsoever. Any information
can be sent.
There does exist an internal limit, however, of the maximum size of a given message.
In Linux, this is defined in linux/msg.h as follows:

#define MSGMAX 4056 /* <= 4056 */ /* max size of message (bytes) */

Messages can be no larger than 4,056 bytes in total size, including the mtype member,
which is 4 bytes in length (long).
34 CHAPTER 6. LINUX INTERPROCESS COMMUNICATIONS

Kernel msg structure The kernel stores each message in the queue within the framework
of the msg structure. It is defined for us in linux/msg.h as follows:

/* one msg structure for each message */

struct msg {
struct msg *msg_next; /* next message on queue */
long msg_type;
char *msg_spot; /* message text address */
short msg_ts; /* message text size */
};

msg next
This is a pointer to the next message in the queue. They are stored as a singly linked
list within kernel addressing space.
msg type
This is the message type, as assigned in the user structure msgbuf.
msg spot
A pointer to the beginning of the message body.
msg ts
The length of the message text, or body.

Kernel msqid ds structure Each of the three types of IPC objects has an internal data
structure which is maintained by the kernel. For message queues, this is the msqid ds
structure. The kernel creates, stores, and maintains an instance of this structure for every
message queue created on the system. It is defined in linux/msg.h as follows:

/* one msqid structure for each queue on the system */

struct msqid_ds {
struct ipc_perm msg_perm;
struct msg *msg_first; /* first message on queue */
struct msg *msg_last; /* last message in queue */
time_t msg_stime; /* last msgsnd time */
time_t msg_rtime; /* last msgrcv time */
time_t msg_ctime; /* last change time */
struct wait_queue *wwait;
struct wait_queue *rwait;
ushort msg_cbytes;
ushort msg_qnum;
ushort msg_qbytes; /* max number of bytes on queue */
ushort msg_lspid; /* pid of last msgsnd */
ushort msg_lrpid; /* last receive pid */
};

While you will rarely have to concern yourself with most of the members of this struc-
ture, a brief description of each is in order to complete our tour:

msg perm
An instance of the ipc perm structure, which is defined for us in linux/ipc.h.
This holds the permission information for the message queue, including the access
permissions, and information about the creator of the queue (uid, etc).
6.4. SYSTEM V IPC 35

msg first
Link to the first message in the queue (the head of the list).
msg last
Link to the last message in the queue (the tail of the list).
msg stime
Timestamp (time t) of the last message that was sent to the queue.
msg rtime
Timestamp of the last message retrieved from the queue.
msg ctime
Timestamp of the last “change” made to the queue (more on this later).
wwait
and
rwait
Pointers into the kernel’s wait queue. They are used when an operation on a message
queue deems the process go into a sleep state (i.e. queue is full and the process is
waiting for an opening).
msg cbytes
Total number of bytes residing on the queue (sum of the sizes of all messages).
msg qnum
Number of messages currently in the queue.
msg qbytes
Maximum number of bytes on the queue.
msg lspid
The PID of the process who sent the last message.
msg lrpid
The PID of the process who retrieved the last message.

Kernel ipc perm structure The kernel stores permission information for IPC objects
in a structure of type ipc perm. For example, in the internal structure for a message
queue described above, the msg perm member is of this type. It is declared for us in
linux/ipc.h as follows:

struct ipc_perm
{
key_t key;
ushort uid; /* owner euid and egid */
ushort gid;
ushort cuid; /* creator euid and egid */
ushort cgid;
ushort mode; /* access modes see mode flags below */
ushort seq; /* slot usage sequence number */
};
36 CHAPTER 6. LINUX INTERPROCESS COMMUNICATIONS

All of the above are fairly self-explanatory. Stored along with the IPC key of the object
is information about both the creator and owner of the object (they may be different). The
octal access modes are also stored here, as an unsigned short. Finally, the slot usage
sequence number is stored at the end. Each time an IPC object is closed via a system call
(destroyed), this value gets incremented by the maximum number of IPC objects that can
reside in a system. Will you have to concern yourself with this value? No.
NOTE:There is an excellent discussion on this topic, and the security reasons as to its
existence and behavior, in Richard Stevens’ UNIX Network Programming book, pp. 125.

SYSTEM CALL: msgget()

In order to create a new message queue, or access an existing queue, the msgget() system
call is used.

SYSTEM CALL: msgget();

PROTOTYPE: int msgget ( key_t key, int msgflg );

RETURNS: message queue identifier on success
-1 on error: errno = EACCESS (permission denied)
EEXIST (Queue exists, cannot create)
EIDRM (Queue is marked for deletion)
ENOENT (Queue does not exist)
ENOMEM (Not enough memory to create queue)
ENOSPC (Maximum queue limit exceeded)
NOTES:

The first argument to msgget() is the key value (in our case returned by a call to
ftok()). This key value is then compared to existing key values that exist within the
kernel for other message queues. At that point, the open or access operation is dependent
upon the contents of the msgflg argument.

IPC CREAT
Create the queue if it doesn’t already exist in the kernel.

IPC EXCL
When used with IPC CREAT, fail if queue already exists.

If IPC CREAT is used alone, msgget() either returns the message queue identifier
for a newly created message queue, or returns the identifier for a queue which exists with
the same key value. If IPC EXCL is used along with IPC CREAT, then either a new queue
is created, or if the queue exists, the call fails with -1. IPC EXCL is useless by itself, but
when combined with IPC CREAT, it can be used as a facility to guarantee that no existing
queue is opened for access.
An optional octal mode may be OR’d into the mask, since each IPC object has permis-
sions that are similar in functionality to file permissions on a UNIX file system!
Let’s create a quick wrapper function for opening or creating message queue:

int open_queue( key_t keyval )

{
int qid;

if((qid = msgget( keyval, IPC_CREAT | 0660 )) == -1)

{
6.4. SYSTEM V IPC 37

return(-1);
}

return(qid);
}

Note the use of the explicit permissions of 0660. This small function either returns a
message queue identifier (int), or -1 on error. The key value must be passed to it as its
only argument.

SYSTEM CALL: msgsnd()

Once we have the queue identifier, we can begin performing operations on it. To deliver a
message to a queue, you use the msgsnd system call:

SYSTEM CALL: msgsnd();

PROTOTYPE: int msgsnd ( int msqid, struct msgbuf *msgp, int msgsz, int msgflg )
RETURNS: 0 on success
-1 on error: errno = EAGAIN (queue is full, and IPC_NOWAIT was asser
EACCES (permission denied, no write permission)
EFAULT (msgp address isn’t accessable - invalid
EIDRM (The message queue has been removed)
EINTR (Received a signal while waiting to writ
EINVAL (Invalid message queue identifier, nonpo
message type, or invalid message size)
ENOMEM (Not enough memory to copy message buffe
NOTES:
The first argument to msgsnd is our queue identifier, returned by a previous call to
msgget. The second argument, msgp, is a pointer to our redeclared and loaded message
buffer. The msgsz argument contains the size of the message in bytes, excluding the length
of the message type (4 byte long).
The msgflg argument can be set to 0 (ignored), or:
IPC NOWAIT
If the message queue is full, then the message is not written to the queue, and con-
trol is returned to the calling process. If not specified, then the calling process will
suspend (block) until the message can be written.
Let’s create another wrapper function for sending messages:

int send_message( int qid, struct mymsgbuf *qbuf )

{
int result, length;

/* The length is essentially the size of the structure minus sizeof(mtype

length = sizeof(struct mymsgbuf) - sizeof(long);

if((result = msgsnd( qid, qbuf, length, 0)) == -1)

{
return(-1);
}

return(result);
}
38 CHAPTER 6. LINUX INTERPROCESS COMMUNICATIONS

This small function attempts to send the message residing at the passed address (qbuf)
to the message queue designated by the passed queue identifier (qid). Here is a sample
code snippet utilizing the two wrapper functions we have developed so far:

#include <stdio.h>
#include <stdlib.h>
#include <linux/ipc.h>
#include <linux/msg.h>

main()
{
int qid;
key_t msgkey;
struct mymsgbuf {
long mtype; /* Message type */
int request; /* Work request number */
double salary; /* Employee’s salary */
} msg;

/* Generate our IPC key value */

msgkey = ftok(".", ’m’);

/* Open/create the queue */

if(( qid = open_queue( msgkey)) == -1) {
perror("open_queue");
exit(1);
}

/* Load up the message with arbitrary test data */

msg.mtype = 1; /* Message type must be a positive number! */
msg.request = 1; /* Data element #1 */
msg.salary = 1000.00; /* Data element #2 (my yearly salary!) */

/* Bombs away! */
if((send_message( qid, &msg )) == -1) {
perror("send_message");
exit(1);
}
}

After creating/opening our message queue, we proceed to load up the message buffer
with test data (note the lack of character data to illustrate our point about sending binary
information). A quick call to send message merrily distributes our message out to the
message queue.
Now that we have a message on our queue, try the ipcs command to view the status
of your queue. Now let’s turn the discussion to actually retrieving the message from the
queue. To do this, you use the msgrcv() system call:

SYSTEM CALL: msgrcv();

PROTOTYPE: int msgrcv ( int msqid, struct msgbuf *msgp, int msgsz, long mtype
RETURNS: Number of bytes copied into message buffer
-1 on error: errno = E2BIG (Message length is greater than msgsz,
6.4. SYSTEM V IPC 39

EACCES (No read permission)

EFAULT (Address pointed to by msgp is invalid)
EIDRM (Queue was removed during retrieval)
EINTR (Interrupted by arriving signal)
EINVAL (msgqid invalid, or msgsz less than 0)
ENOMSG (IPC_NOWAIT asserted, and no message exi
in the queue to satisfy the request)
NOTES:

Obviously, the first argument is used to specify the queue to be used during the message
retrieval process (should have been returned by an earlier call to msgget). The second ar-
gument (msgp) represents the address of a message buffer variable to store the retrieved
message at. The third argument (msgsz) represents the size of the message buffer struc-
ture, excluding the length of the mtype member. Once again, this can easily be calculated
as:

msgsz = sizeof(struct mymsgbuf) - sizeof(long);

The fourth argument (mtype) specifies the type of message to retrieve from the queue.
The kernel will search the queue for the oldest message having a matching type, and will
return a copy of it in the address pointed to by the msgp argument. One special case exists.
If the mtype argument is passed with a value of zero, then the oldest message on the queue
is returned, regardless of type.
If IPC NOWAIT is passed as a flag, and no messages are available, the call returns
ENOMSG to the calling process. Otherwise, the calling process blocks until a message
arrives in the queue that satisfies the msgrcv() parameters. If the queue is deleted while
a client is waiting on a message, EIDRM is returned. EINTR is returned if a signal is
caught while the process is in the middle of blocking, and waiting for a message to arrive.
Let’s examine a quick wrapper function for retrieving a message from our queue:

int read_message( int qid, long type, struct mymsgbuf *qbuf )

{
int result, length;

/* The length is essentially the size of the structure minus sizeof(mtype

length = sizeof(struct mymsgbuf) - sizeof(long);

if((result = msgrcv( qid, qbuf, length, type, 0)) == -1)

{
return(-1);
}

return(result);
}

After successfully retrieving a message from the queue, the message entry within the
queue is destroyed.
The MSG NOERROR bit in the msgflg argument provides some additional capabili-
ties. If the size of the physical message data is greater than msgsz, and MSG NOERROR
is asserted, then the message is truncated, and only msgsz bytes are returned. Normally,
the msgrcv() system call returns -1 (E2BIG), and the message will remain on the queue
for later retrieval. This behavior can used to create another wrapper function, which will
allow us to “peek” inside the queue, to see if a message has arrived that satisfies our request:
40 CHAPTER 6. LINUX INTERPROCESS COMMUNICATIONS

int peek_message( int qid, long type )

{
int result, length;

if((result = msgrcv( qid, NULL, 0, type, IPC_NOWAIT)) == -1)

{
if(errno == E2BIG)
return(TRUE);
}

return(FALSE);
}

Above, you will notice the lack of a buffer address and a length. In this particular
case, we want the call to fail. However, we check for the return of E2BIG which indicates
that a message does exist which matches our requested type. The wrapper function returns
TRUE on success, FALSE otherwise. Also note the use of the IPC NOWAIT flag, which
prevents the blocking behavior described earlier.

SYSTEM CALL: msgctl()

Through the development of the wrapper functions presented earlier, you now have a sim-
ple, somewhat elegant approach to creating and utilizing message queues in your appli-
cations. Now, we will turn the discussion to directly manipulating the internal structures
associated with a given message queue.
To perform control operations on a message queue, you use the msgctl() system
call.

SYSTEM CALL: msgctl();

PROTOTYPE: int msgctl ( int msgqid, int cmd, struct msqid_ds *buf );
RETURNS: 0 on success
-1 on error: errno = EACCES (No read permission and cmd is IPC_STA
EFAULT (Address pointed to by buf is invalid
IPC_STAT commands)
EIDRM (Queue was removed during retrieval)
EINVAL (msgqid invalid, or msgsz less than 0)
EPERM (IPC_SET or IPC_RMID command was issue
calling process does not have write (
access to the queue)
NOTES:

Now, common sense dictates that direct manipulation of the internal kernel data struc-
tures could lead to some late night fun. Unfortunately, the resulting duties on the part of
the programmer could only be classified as fun if you like trashing the IPC subsystem. By
using msgctl() with a selective set of commands, you have the ability to manipulate
those items which are less likely to cause grief. Let’s look at these commands:

IPC STAT
Retrieves the msqid ds structure for a queue, and stores it in the address of the buf
argument.

IPC SET
Sets the value of the ipc perm member of the msqid ds structure for a queue. Takes
the values from the buf argument.
6.4. SYSTEM V IPC 41

IPC RMID
Removes the queue from the kernel.

Recall our discussion about the internal data structure for message queues (msqid ds).
The kernel maintains an instance of this structure for each queue which exists in the system.
By using the IPC STAT command, we can retrieve a copy of this structure for examination.
Let’s look at a quick wrapper function that will retrieve the internal structure and copy it
into a passed address:

int get_queue_ds( int qid, struct msgqid_ds *qbuf )

{
if( msgctl( qid, IPC_STAT, qbuf) == -1)
{
return(-1);
}

return(0);
}

If we are unable to copy the internal buffer, -1 is returned to the calling function. If all
went well, a value of 0 (zero) is returned, and the passed buffer should contain a copy of
the internal data structure for the message queue represented by the passed queue identifier
(qid).
Now that we have a copy of the internal data structure for a queue, what attributes
can be manipulated, and how can we alter them? The only modifiable item in the data
structure is the ipc perm member. This contains the permissions for the queue, as well
as information about the owner and creator. However, the only members of the ipc perm
structure that are modifiable are mode, uid, and gid. You can change the owner’s user
id, the owner’s group id, and the access permissions for the queue.
Let’s create a wrapper function designed to change the mode of a queue. The mode
must be passed in as a character array (i.e. “660”).

int change_queue_mode( int qid, char *mode )

{
struct msqid_ds tmpbuf;

/* Retrieve a current copy of the internal data structure */

get_queue_ds( qid, &tmpbuf);

/* Change the permissions using an old trick */

sscanf(mode, "%ho", &tmpbuf.msg_perm.mode);

/* Update the internal data structure */

if( msgctl( qid, IPC_SET, &tmpbuf) == -1)
{
return(-1);
}

return(0);
}

We retrieve a current copy of the internal data structure by a quick call to our
get queue ds wrapper function. We then make a call to sscanf() to alter the mode
member of the associated msg perm structure. No changes take place, however, until
42 CHAPTER 6. LINUX INTERPROCESS COMMUNICATIONS

the new copy is used to update the internal version. This duty is performed by a call to
msgctl() using the IPC SET command.
BE CAREFUL! It is possible to alter the permissions on a queue, and in doing so,
inadvertently lock yourself out! Remember, these IPC objects don’t go away unless they
are properly removed, or the system is rebooted. So, even if you can’t see a queue with
ipcs doesn’t mean that it isn’t there.
To illustrate this point, a somewhat humorous anecdote seems to be in
order. While teaching a class on UNIX internals at the University of South
Florida, I ran into a rather embarrassing stumbling block. I had dialed into
their lab server the night before, in order to compile and test the labwork to
be used in the week-long class. In the process of my testing, I realized that
I had made a typo in the code used to alter the permissions on a message
queue. I created a simple message queue, and tested the sending and receiving
capabilities with no incident. However, when I attempted to change the mode
of the queue from “660” to “600”, the resulting action was that I was locked
out of my own queue! As a result, I could not test the message queue labwork
in the same area of my source directory. Since I used the ftok() function to
create the IPC key, I was trying to access a queue that I did not have proper
permissions for. I ended up contacting the local system administrator on the
morning of the class, only to spend an hour explaining to him what a message
queue was, and why I needed him to run the ipcrm command for me. grrrr.
After successfully retrieving a message from a queue, the message is removed. How-
ever, as mentioned earlier, IPC objects remain in the system unless explicitly removed, or
the system is rebooted. Therefore, our message queue still exists within the kernel, avail-
able for use long after a single message disappears. To complete the life cycle of a message
queue, they should be removed with a call to msgctl(), using the IPC RMID command:
int remove_queue( int qid )
{
if( msgctl( qid, IPC_RMID, 0) == -1)
{
return(-1);
}

return(0);
}
This wrapper function returns 0 if the queue was removed without incident, else a value
of -1. The removal of the queue is atomic in nature, and any subsequent accesses to the
queue for whatever purpose will fail miserably.

msgtool: An interactive message queue manipulator

Few can deny the immediate benefit of having accurate technical information readily avail-
able. Such materials provide a tremendous mechanism for learning and exploring new
areas. On the same note, having real world examples to accompany any technical informa-
tion will speed up and reinforce the learning process.
Until now, the only useful examples which have been presented were the wrapper func-
tions for manipulating message queues. While they are extremely useful, they have not
been presented in a manner which would warrant further study and experimentation. To
remedy this, you will be presented with msgtool, an interactive command line utility for
manipulating IPC message queues. While it certainly functions as an adequate tool for ed-
ucation reinforcement, it can be applied directly into real world assignments, by providing
message queue functionality via standard shell scripts.
6.4. SYSTEM V IPC 43

Background The msgtool program relies on command line arguments to determine

its behavior. This is what makes it especially useful when called from a shell script. All
of the capabilities are provided, from creating, sending, and retrieving, to changing the
permissions and finally removing a queue. Currently, it uses a character array for data,
allowing you to send textual messages. Changing it to facilitate additional data types is left
as an exercise to the reader.

Command Line Syntax

Sending Messages

msgtool s (type) "text"

Retrieving Messages

msgtool r (type)

Changing the Permissions (mode)

msgtool m (mode)

Deleting a Queue

msgtool d

Examples

msgtool s 1 test
msgtool s 5 test
msgtool s 1 "This is a test"
msgtool r 1
msgtool d
msgtool m 660

The Source The following is the source code for the msgtool facility. It should compile
clean on any recent (decent) kernel revision that supports System V IPC. Be sure to enable
System V IPC in your kernel when doing a rebuild!
On a side note, this utility will create a message queue if it does not exist, no matter
what type of action is requested.

NOTE: Since this tool uses the ftok() function to generate IPC key values,
you may encounter directory conflicts. If you change directories at any point
in your script, it probably won’t work. Another solution would be to hardcode
a more complete path into msgtool, such as “/tmp/msgtool”, or possibly even
allow the path to be passed on the command line, along with the operational
arguments.

/*****************************************************************************
Excerpt from "Linux Programmer’s Guide - Chapter 6"
(C)opyright 1994-1995, Scott Burkett
*****************************************************************************
MODULE: msgtool.c
*****************************************************************************
44 CHAPTER 6. LINUX INTERPROCESS COMMUNICATIONS

A command line tool for tinkering with SysV style Message Queues
*****************************************************************************/

#include <stdio.h>
#include <stdlib.h>
#include <ctype.h>
#include <sys/types.h>
#include <sys/ipc.h>
#include <sys/msg.h>

#define MAX_SEND_SIZE 80

struct mymsgbuf {
long mtype;
char mtext[MAX_SEND_SIZE];
};

void send_message(int qid, struct mymsgbuf *qbuf, long type, char *text);
void read_message(int qid, struct mymsgbuf *qbuf, long type);
void remove_queue(int qid);
void change_queue_mode(int qid, char *mode);
void usage(void);

int main(int argc, char *argv[])

{
key_t key;
int msgqueue_id;
struct mymsgbuf qbuf;

if(argc == 1)
usage();

/* Create unique key via call to ftok() */

key = ftok(".", ’m’);

/* Open the queue - create if necessary */

if((msgqueue_id = msgget(key, IPC_CREAT|0660)) == -1) {
perror("msgget");
exit(1);
}

switch(tolower(argv[1][0]))
{
case ’s’: send_message(msgqueue_id, (struct mymsgbuf *)&qbuf,
atol(argv[2]), argv[3]);
break;
case ’r’: read_message(msgqueue_id, &qbuf, atol(argv[2]));
break;
case ’d’: remove_queue(msgqueue_id);
break;
case ’m’: change_queue_mode(msgqueue_id, argv[2]);
break;
6.4. SYSTEM V IPC 45

default: usage();

return(0);
}

void send_message(int qid, struct mymsgbuf *qbuf, long type, char *text)
{
/* Send a message to the queue */
printf("Sending a message ...\n");
qbuf->mtype = type;
strcpy(qbuf->mtext, text);

if((msgsnd(qid, (struct msgbuf *)qbuf,

strlen(qbuf->mtext)+1, 0)) ==-1)
{
perror("msgsnd");
exit(1);
}
}

void read_message(int qid, struct mymsgbuf *qbuf, long type)

{
/* Read a message from the queue */
printf("Reading a message ...\n");
qbuf->mtype = type;
msgrcv(qid, (struct msgbuf *)qbuf, MAX_SEND_SIZE, type, 0);

printf("Type: %ld Text: %s\n", qbuf->mtype, qbuf->mtext);

}

void remove_queue(int qid)

{
/* Remove the queue */
msgctl(qid, IPC_RMID, 0);
}

void change_queue_mode(int qid, char *mode)

{
struct msqid_ds myqueue_ds;

/* Get current info */

msgctl(qid, IPC_STAT, &myqueue_ds);

/* Convert and load the mode */

sscanf(mode, "%ho", &myqueue_ds.msg_perm.mode);

/* Update the mode */

msgctl(qid, IPC_SET, &myqueue_ds);
}

void usage(void)
{
46 CHAPTER 6. LINUX INTERPROCESS COMMUNICATIONS

fprintf(stderr, "msgtool - A utility for tinkering with msg queues\n");

fprintf(stderr, "\nUSAGE: msgtool (s)end <type> <messagetext>\n");
fprintf(stderr, " (r)ecv <type>\n");
fprintf(stderr, " (d)elete\n");
fprintf(stderr, " (m)ode <octal mode>\n");
exit(1);
}

6.4.3 Semaphores
Basic Concepts
Semaphores can best be described as counters used to control access to shared resources by
multiple processes. They are most often used as a locking mechanism to prevent processes
from accessing a particular resource while another process is performing operations on it.
Semaphores are often dubbed the most difficult to grasp of the three types of System V
IPC objects. In order to fully understand semaphores, we’ll discuss them briefly before
engaging any system calls and operational theory.
The name semaphore is actually an old railroad term, referring to the crossroad “arms”
that prevent cars from crossing the tracks at intersections. The same can be said about a
simple semaphore set. If the semaphore is on (the arms are up), then a resource is available
(cars may cross the tracks). However, if the semaphore is off (the arms are down), then
resources are not available (the cars must wait).
While this simple example may stand to introduce the concept, it is important to realize
that semaphores are actually implemented as sets, rather than as single entities. Of course,
a given semaphore set might only have one semaphore, as in our railroad example.
Perhaps another approach to the concept of semaphores, is to think of them as resource
counters. Let’s apply this concept to another real world scenario. Consider a print spooler,
capable of handling multiple printers, with each printer handling multiple print requests.
A hypothetical print spool manager will utilize semaphore sets to monitor access to each
printer.
Assume that in our corporate print room, we have 5 printers online. Our print spool
manager allocates a semaphore set with 5 semaphores in it, one for each printer on the
system. Since each printer is only physically capable of printing one job at a time, each of
our five semaphores in our set will be initialized to a value of 1 (one), meaning that they
are all online, and accepting requests.
John sends a print request to the spooler. The print manager looks at the semaphore set,
and finds the first semaphore which has a value of one. Before sending John’s request to the
physical device, the print manager decrements the semaphore for the corresponding printer
by a value of negative one (-1). Now, that semaphore’s value is zero. In the world of System
V semaphores, a value of zero represents 100% resource utilization on that semaphore. In
our example, no other request can be sent to that printer until it is no longer equal to zero.
When John’s print job has completed, the print manager increments the value of the
semaphore which corresponds to the printer. Its value is now back up to one (1), which
means it is available again. Naturally, if all 5 semaphores had a value of zero, that would
indicate that they are all busy printing requests, and that no printers are available.
Although this was a simple example, please do not be confused by the initial value of
one (1) which was assigned to each semaphore in the set. Semaphores, when thought of as
resource counters, may be initialized to any positive integer value, and are not limited to
either being zero or one. If it were possible for each of our five printers to handle 10 print
jobs at a time, we could initialize each of our semaphores to 10, decrementing by one for
every new job, and incrementing by one whenever a print job was finished. As you will
discover in the next chapter, semaphores have a close working relationship with shared
6.4. SYSTEM V IPC 47

memory segments, acting as a watchdog to prevent multiple writes to the same memory
segment.
Before delving into the associated system calls, lets take a brief tour through the various
internal data structures utilized during semaphore operations.

Internal Data Structures

Let’s briefly look at data structures maintained by the kernel for semaphore sets.

Kernel semid ds structure As with message queues, the kernel maintains a special
internal data structure for each semaphore set which exists within its addressing space.
This structure is of type semid ds, and is defined in linux/sem.h as follows:

/* One semid data structure for each set of semaphores in the system. */
struct semid_ds {
struct ipc_perm sem_perm; /* permissions .. see ipc.h */
time_t sem_otime; /* last semop time */
time_t sem_ctime; /* last change time */
struct sem *sem_base; /* ptr to first semaphore in arra
struct wait_queue *eventn;
struct wait_queue *eventz;
struct sem_undo *undo; /* undo requests on this array */
ushort sem_nsems; /* no. of semaphores in array */
};

As with message queues, operations on this structure are performed by a special system
call, and should not be tinkered with directly. Here are descriptions of the more pertinent
fields:

sem perm
This is an instance of the ipc perm structure, which is defined for us in
linux/ipc.h. This holds the permission information for the semaphore set, in-
cluding the access permissions, and information about the creator of the set (uid,
etc).

sem otime
Time of the last semop() operation (more on this in a moment)

sem ctime
Time of the last change to this structure (mode change, etc)

sem base
Pointer to the first semaphore in the array (see next structure)

sem undo
Number of undo requests in this array (more on this in a moment)

sem nsems
Number of semaphores in the semaphore set (the array)
48 CHAPTER 6. LINUX INTERPROCESS COMMUNICATIONS

Kernel sem structure In the semid ds structure, there exists a pointer to the base of
the semaphore array itself. Each array member is of the sem structure type. It is also
defined in linux/sem.h:

/* One semaphore structure for each semaphore in the system. */

struct sem {
short sempid; /* pid of last operation */
ushort semval; /* current value */
ushort semncnt; /* num procs awaiting increase in semva
ushort semzcnt; /* num procs awaiting semval = 0 */
};

sem pid
The PID (process ID) that performed the last operation

sem semval
The current value of the semaphore

sem semncnt
Number of processes waiting for resources to become available

sem semzcnt
Number of processes waiting for 100% resource utilization

SYSTEM CALL: semget()

In order to create a new semaphore set, or access an existing set, the semget() system
call is used.

SYSTEM CALL: semget();

PROTOTYPE: int semget ( key_t key, int nsems, int semflg );

RETURNS: semaphore set IPC identifier on success
-1 on error: errno = EACCESS (permission denied)
EEXIST (set exists, cannot create (IPC_EXCL))
EIDRM (set is marked for deletion)
ENOENT (set does not exist, no IPC_CREAT was
ENOMEM (Not enough memory to create new set)
ENOSPC (Maximum set limit exceeded)
NOTES:

The first argument to semget() is the key value (in our case returned by a call to
ftok()). This key value is then compared to existing key values that exist within the
kernel for other semaphore sets. At that point, the open or access operation is dependent
upon the contents of the semflg argument.

IPC CREAT
Create the semaphore set if it doesn’t already exist in the kernel.

IPC EXCL
When used with IPC CREAT, fail if semaphore set already exists.
6.4. SYSTEM V IPC 49

If IPC CREAT is used alone, semget() either returns the semaphore set identifier
for a newly created set, or returns the identifier for a set which exists with the same key
value. If IPC EXCL is used along with IPC CREAT, then either a new set is created, or
if the set exists, the call fails with -1. IPC EXCL is useless by itself, but when combined
with IPC CREAT, it can be used as a facility to guarantee that no existing semaphore set
is opened for access.
As with the other forms of System V IPC, an optional octal mode may be OR’d into the
mask to form the permissions on the semaphore set.
The nsems argument specifies the number of semaphores that should be created in a
new set. This represents the number of printers in our fictional print room described earlier.
The maximum number of semaphores in a set is defined in “linux/sem.h” as:

#define SEMMSL 32 /* <=512 max num of semaphores per id */

Note that the nsems argument is ignored if you are explicitly opening an existing set.
Let’s create a wrapper function for opening or creating semaphore sets:

int open_semaphore_set( key_t keyval, int numsems )

{
int sid;

if ( ! numsems )
return(-1);

if((sid = semget( mykey, numsems, IPC_CREAT | 0660 )) == -1)

{
return(-1);
}

return(sid);
}

Note the use of the explicit permissions of 0660. This small function either returns a
semaphore set identifier (int), or -1 on error. The key value must be passed to it, as well
as the number of semaphores to allocate space for if creating a new set. In the example
presented at the end of this section, notice the use of the IPC EXCL flag to determine
whether or not the semaphore set exists or not.

SYSTEM CALL: semop()

SYSTEM CALL: semop();
PROTOTYPE: int semop ( int semid, struct sembuf *sops, unsigned nsops);
RETURNS: 0 on success (all operations performed)
-1 on error: errno = E2BIG (nsops greater than max number of ops all
EACCESS (permission denied)
EAGAIN (IPC_NOWAIT asserted, operation could no
EFAULT (invalid address pointed to by sops argu
EIDRM (semaphore set was removed)
EINTR (Signal received while sleeping)
EINVAL (set doesn’t exist, or semid is invalid)
ENOMEM (SEM_UNDO asserted, not enough memory to
undo structure necessary)
ERANGE (semaphore value out of range)
NOTES:
50 CHAPTER 6. LINUX INTERPROCESS COMMUNICATIONS

The first argument to semget() is the key value (in our case returned by a call to
semget). The second argument (sops) is a pointer to an array of operations to be per-
formed on the semaphore set, while the third argument (nsops) is the number of opera-
tions in that array.
The sops argument points to an array of type sembuf. This structure is declared in
linux/sem.h as follows:

/* semop system call takes an array of these */

struct sembuf {
ushort sem_num; /* semaphore index in array */
short sem_op; /* semaphore operation */
short sem_flg; /* operation flags */
};

sem num
The number of the semaphore you wish to deal with

sem op
The operation to perform (positive, negative, or zero)

sem flg
Operational flags

If sem op is negative, then its value is subtracted from the semaphore. This cor-
relates with obtaining resources that the semaphore controls or monitors access of. If
IPC NOWAIT is not specified, then the calling process sleeps until the requested amount
of resources are available in the semaphore (another process has released some).
If sem op is positive, then it’s value is added to the semaphore. This correlates with
returning resources back to the application’s semaphore set. Resources should always be
returned to a semaphore set when they are no longer needed!
Finally, if sem op is zero (0), then the calling process will sleep() until the semaphore’s
value is 0. This correlates to waiting for a semaphore to reach 100% utilization. A good
example of this would be a daemon running with superuser permissions that could dynam-
ically adjust the size of the semaphore set if it reaches full utilization.
In order to explain the semop call, let’s revisit our print room scenario. Let’s assume
only one printer, capable of only one job at a time. We create a semaphore set with only
one semaphore in it (only one printer), and initialize that one semaphore to a value of one
(only one job at a time).
Each time we desire to send a job to this printer, we need to first make sure that the
resource is available. We do this by attempting to obtain one unit from the semaphore.
Let’s load up a sembuf array to perform the operation:

struct sembuf sem_lock = { 0, -1, IPC_NOWAIT };

Translation of the above initialized structure dictates that a value of “-1” will be added
to semaphore number 0 in the semaphore set. In other words, one unit of resources will be
obtained from the only semaphore in our set (0th member). IPC NOWAIT is specified, so
the call will either go through immediately, or fail if another print job is currently printing.
Here is an example of using this initialized sembuf structure with the semop system call:

if((semop(sid, &sem_lock, 1) == -1)

perror("semop");
6.4. SYSTEM V IPC 51

The third argument (nsops) says that we are only performing one (1) operation (there
is only one sembuf structure in our array of operations). The sid argument is the IPC
identifier for our semaphore set.
When our print job has completed, we must return the resources back to the semaphore
set, so that others may use the printer.

struct sembuf sem_unlock = { 0, 1, IPC_NOWAIT };

Translation of the above initialized structure dictates that a value of “1” will be added
to semaphore number 0 in the semaphore set. In other words, one unit of resources will be
returned to the set.

SYSTEM CALL: semctl()

SYSTEM CALL: semctl();
PROTOTYPE: int semctl ( int semid, int semnum, int cmd, union semun arg );
RETURNS: positive integer on success
-1 on error: errno = EACCESS (permission denied)
EFAULT (invalid address pointed to by arg argum
EIDRM (semaphore set was removed)
EINVAL (set doesn’t exist, or semid is invalid)
EPERM (EUID has no privileges for cmd in arg)
ERANGE (semaphore value out of range)
NOTES: Performs control operations on a semaphore set

The semctl system call is used to perform control operations on a semaphore set. This
call is analogous to the msgctl system call which is used for operations on message queues.
If you compare the argument lists of the two system calls, you will notice that the list for
semctl varies slightly from that of msgctl. Recall that semaphores are actually implemented
as sets, rather than as single entities. With semaphore operations, not only does the IPC
key need to be passed, but the target semaphore within the set as well.
Both system calls utilize a cmd argument, for specification of the command to be per-
formed on the IPC object. The remaining difference lies in the final argument to both
calls. In msgctl, the final argument represents a copy of the internal data structure used
by the kernel. Recall that we used this structure to retrieve internal information about a
message queue, as well as to set or change permissions and ownership of the queue. With
semaphores, additional operational commands are supported, thus requiring a more com-
plex data type as the final argument. The use of a union confuses many neophyte semaphore
programmers to a substantial degree. We will dissect this structure carefully, in an effort to
prevent any confusion.
The first argument to semctl() is the key value (in our case returned by a call to
semget). The second argument (semun) is the semaphore number that an operation is
targeted towards. In essence, this can be thought of as an index into the semaphore set, with
the first semaphore (or only one) in the set being represented by a value of zero (0).
The cmd argument represents the command to be performed against the set. As you
can see, the familiar IPC STAT/IPC SET commands are present, along with a wealth of
additional commands specific to semaphore sets:

IPC STAT
Retrieves the semid ds structure for a set, and stores it in the address of the buf
argument in the semun union.
52 CHAPTER 6. LINUX INTERPROCESS COMMUNICATIONS

IPC SET
Sets the value of the ipc perm member of the semid ds structure for a set. Takes the
values from the buf argument of the semun union.
IPC RMID
Removes the set from the kernel.
GETALL
Used to obtain the values of all semaphores in a set. The integer values are stored in
an array of unsigned short integers pointed to by the array member of the union.
GETNCNT
Returns the number of processes currently waiting for resources.
GETPID
Returns the PID of the process which performed the last semop call.
GETVAL
Returns the value of a single semaphore within the set.
GETZCNT
Returns the number of processes currently waiting for 100% resource utilization.
SETALL
Sets all semaphore values with a set to the matching values contained in the array
member of the union.
SETVAL
Sets the value of an individual semaphore within the set to the val member of the
union.

The arg argument represents an instance of type semun. This particular union is
declared in linux/sem.h as follows:

/* arg for semctl system calls. */

union semun {
int val; /* value for SETVAL */
struct semid_ds *buf; /* buffer for IPC_STAT & IPC_SET */
ushort *array; /* array for GETALL & SETALL */
struct seminfo *__buf; /* buffer for IPC_INFO */
void *__pad;
};

val
Used when the SETVAL command is performed. Specifies the value to set the
semaphore to.
buf
Used in the IPC STAT/IPC SET commands. Represents a copy of the internal
semaphore data structure used in the kernel.
array
A pointer used in the GETALL/SETALL commands. Should point to an array of
integer values to be used in setting or retrieving all semaphore values in a set.
6.4. SYSTEM V IPC 53

The remaining arguments buf and pad are used internally in the semaphore code
within the kernel, and are of little or no use to the application developer. As a matter of
fact, these two arguments are specific to the Linux operating system, and are not found in
other UNIX implementations.
Since this particular system call is arguably the most difficult to grasp of all the System
V IPC calls, we’ll examine multiple examples of it in action.
The following snippet returns the value of the passed semaphore. The final argument
(the union) is ignored when the GETVAL command is used:

int get_sem_val( int sid, int semnum )

{
return( semctl(sid, semnum, GETVAL, 0));
}

To revisit the printer example, let’s say the status of all five printers was required:

#define MAX_PRINTERS 5

printer_usage()
{
int x;

for(x=0; x<MAX_PRINTERS; x++)

printf("Printer %d: %d\n\r", x, get_sem_val( sid, x ));
}

Consider the following function, which could be used to initialize a new semaphore
value:

void init_semaphore( int sid, int semnum, int initval)

{
union semun semopts;

semopts.val = initval;
semctl( sid, semnum, SETVAL, semopts);
}

Note that the final argument of semctl is a copy of the union, rather than a pointer to it.
While we’re on the subject of the union as an argument, allow me to demonstrate a rather
common mistake when using this system call.
Recall from the msgtool project that the IPC STAT and IPC SET commands were used
to alter permissions on the queue. While these commands are supported in the semaphore
implementation, their usage is a bit different, as the internal data structure is retrieved and
copied from a member of the union, rather than as a single entity. Can you locate the bug
in this code?

/* Required permissions should be passed in as text (ex: "660") */

void changemode(int sid, char *mode)

{
int rc;
struct semid_ds mysemds;
54 CHAPTER 6. LINUX INTERPROCESS COMMUNICATIONS

/* Get current values for internal data structure */

if((rc = semctl(sid, 0, IPC_STAT, semopts)) == -1)
{
perror("semctl");
exit(1);
}

printf("Old permissions were %o\n", semopts.buf->sem_perm.mode);

/* Change the permissions on the semaphore */

sscanf(mode, "%o", &semopts.buf->sem_perm.mode);

/* Update the internal data structure */

semctl(sid, 0, IPC_SET, semopts);

printf("Updated...\n");
}

The code is attempting to make a local copy of the internal data structure for the set,
modify the permissions, and IPC SET them back to the kernel. However, the first call to
semctl promptly returns EFAULT, or bad address for the last argument (the union!). In
addition, if we hadn’t checked for errors from that call, we would have gotten a memory
fault. Why?
Recall that the IPC SET/IPC STAT commands use the buf member of the union, which
is a pointer to a type semid ds. Pointers are pointers are pointers are pointers! The buf
member must point to some valid storage location in order for our function to work prop-
erly. Consider this revamped version:

void changemode(int sid, char *mode)

{
int rc;
struct semid_ds mysemds;

/* Get current values for internal data structure */

/* Point to our local copy first! */

semopts.buf = &mysemds;

/* Let’s try this again! */

if((rc = semctl(sid, 0, IPC_STAT, semopts)) == -1)
{
perror("semctl");
exit(1);
}

printf("Old permissions were %o\n", semopts.buf->sem_perm.mode);

/* Change the permissions on the semaphore */

sscanf(mode, "%o", &semopts.buf->sem_perm.mode);

/* Update the internal data structure */

semctl(sid, 0, IPC_SET, semopts);
6.4. SYSTEM V IPC 55

printf("Updated...\n");
}

semtool: An interactive semaphore manipulator

Background The semtool program relies on command line arguments to determine its
behavior. This is what makes it especially useful when called from a shell script. All of the
capabilities are provided, from creating and manipulating, to changing the permissions and
finally removing a semaphore set. It can be used to control shared resources via standard
shell scripts.

Command Line Syntax

Creating a Semaphore Set

semtool c (number of semaphores in set)

Locking a Semaphore

semtool l (semaphore number to lock)

Unlocking a Semaphore

semtool u (semaphore number to unlock)

Changing the Permissions (mode)

semtool m (mode)

Deleting a Semaphore Set

semtool d

Examples

semtool c 5
semtool l
semtool u
semtool m 660
semtool d

The Source

/*****************************************************************************
Excerpt from "Linux Programmer’s Guide - Chapter 6"
(C)opyright 1994-1995, Scott Burkett
*****************************************************************************
MODULE: semtool.c
*****************************************************************************
A command line tool for tinkering with SysV style Semaphore Sets

*****************************************************************************/
56 CHAPTER 6. LINUX INTERPROCESS COMMUNICATIONS

#include <stdio.h>
#include <ctype.h>
#include <stdlib.h>
#include <sys/types.h>
#include <sys/ipc.h>
#include <sys/sem.h>

#define SEM_RESOURCE_MAX 1 /* Initial value of all semaphores */

void opensem(int *sid, key_t key);

void createsem(int *sid, key_t key, int members);
void locksem(int sid, int member);
void unlocksem(int sid, int member);
void removesem(int sid);
unsigned short get_member_count(int sid);
int getval(int sid, int member);
void dispval(int sid, int member);
void changemode(int sid, char *mode);
void usage(void);

int main(int argc, char *argv[])

{
key_t key;
int semset_id;

if(argc == 1)
usage();

/* Create unique key via call to ftok() */

key = ftok(".", ’s’);

switch(tolower(argv[1][0]))
{
case ’c’: if(argc != 3)
usage();
createsem(&semset_id, key, atoi(argv[2]));
break;
case ’l’: if(argc != 3)
usage();
opensem(&semset_id, key);
locksem(semset_id, atoi(argv[2]));
break;
case ’u’: if(argc != 3)
usage();
opensem(&semset_id, key);
unlocksem(semset_id, atoi(argv[2]));
break;
case ’d’: opensem(&semset_id, key);
removesem(semset_id);
break;
case ’m’: opensem(&semset_id, key);
changemode(semset_id, argv[2]);
break;
default: usage();
6.4. SYSTEM V IPC 57

return(0);
}

void opensem(int *sid, key_t key)

{
/* Open the semaphore set - do not create! */

if((*sid = semget(key, 0, 0666)) == -1)

{
printf("Semaphore set does not exist!\n");
exit(1);
}

void createsem(int *sid, key_t key, int members)

{
int cntr;
union semun semopts;

if(members > SEMMSL) {

printf("Sorry, max number of semaphores in a set is %d\n",
SEMMSL);
exit(1);
}

printf("Attempting to create new semaphore set with %d members\n",

members);

if((*sid = semget(key, members, IPC_CREAT|IPC_EXCL|0666))

== -1)
{
fprintf(stderr, "Semaphore set already exists!\n");
exit(1);
}

semopts.val = SEM_RESOURCE_MAX;

/* Initialize all members (could be done with SETALL) */

for(cntr=0; cntr<members; cntr++)
semctl(*sid, cntr, SETVAL, semopts);
}

void locksem(int sid, int member)

{
struct sembuf sem_lock={ 0, -1, IPC_NOWAIT};

if( member<0 || member>(get_member_count(sid)-1))

{
fprintf(stderr, "semaphore member %d out of range\n", member);
return;
58 CHAPTER 6. LINUX INTERPROCESS COMMUNICATIONS

/* Attempt to lock the semaphore set */

if(!getval(sid, member))
{
fprintf(stderr, "Semaphore resources exhausted (no lock)!\n");
exit(1);
}

sem_lock.sem_num = member;

if((semop(sid, &sem_lock, 1)) == -1)

{
fprintf(stderr, "Lock failed\n");
exit(1);
}
else
printf("Semaphore resources decremented by one (locked)\n");

dispval(sid, member);
}

void unlocksem(int sid, int member)

{
struct sembuf sem_unlock={ member, 1, IPC_NOWAIT};
int semval;

if( member<0 || member>(get_member_count(sid)-1))

{
fprintf(stderr, "semaphore member %d out of range\n", member);
return;
}

/* Is the semaphore set locked? */

semval = getval(sid, member);
if(semval == SEM_RESOURCE_MAX) {
fprintf(stderr, "Semaphore not locked!\n");
exit(1);
}

sem_unlock.sem_num = member;

/* Attempt to lock the semaphore set */

if((semop(sid, &sem_unlock, 1)) == -1)
{
fprintf(stderr, "Unlock failed\n");
exit(1);
}
else
printf("Semaphore resources incremented by one (unlocked)\n");

dispval(sid, member);
}
6.4. SYSTEM V IPC 59

void removesem(int sid)

{
semctl(sid, 0, IPC_RMID, 0);
printf("Semaphore removed\n");
}

unsigned short get_member_count(int sid)

{
union semun semopts;
struct semid_ds mysemds;

semopts.buf = &mysemds;

/* Return number of members in the semaphore set */

return(semopts.buf->sem_nsems);
}

int getval(int sid, int member)

{
int semval;

semval = semctl(sid, member, GETVAL, 0);

return(semval);
}

void changemode(int sid, char *mode)

{
int rc;
union semun semopts;
struct semid_ds mysemds;

/* Get current values for internal data structure */

semopts.buf = &mysemds;

rc = semctl(sid, 0, IPC_STAT, semopts);

if (rc == -1) {
perror("semctl");
exit(1);
}

printf("Old permissions were %o\n", semopts.buf->sem_perm.mode);

/* Change the permissions on the semaphore */

sscanf(mode, "%ho", &semopts.buf->sem_perm.mode);

/* Update the internal data structure */

semctl(sid, 0, IPC_SET, semopts);

printf("Updated...\n");

void dispval(int sid, int member)

60 CHAPTER 6. LINUX INTERPROCESS COMMUNICATIONS

{
int semval;

semval = semctl(sid, member, GETVAL, 0);

printf("semval for member %d is %d\n", member, semval);
}

void usage(void)
{
fprintf(stderr, "semtool - A utility for tinkering with semaphores\n");
fprintf(stderr, "\nUSAGE: semtool4 (c)reate <semcount>\n");
fprintf(stderr, " (l)ock <sem #>\n");
fprintf(stderr, " (u)nlock <sem #>\n");
fprintf(stderr, " (d)elete\n");
fprintf(stderr, " (m)ode <mode>\n");
exit(1);
}

semstat: A semtool companion program

As an added bonus, the source code to a companion program called semstat is provided
next. The semstat program displays the values of each of the semaphores in the set
created by semtool.

/*****************************************************************************
Excerpt from "Linux Programmer’s Guide - Chapter 6"
(C)opyright 1994-1995, Scott Burkett
*****************************************************************************
MODULE: semstat.c
*****************************************************************************
A companion command line tool for the semtool package. semstat displays
the current value of all semaphores in the set created by semtool.
*****************************************************************************/

#include <stdio.h>
#include <stdlib.h>
#include <sys/types.h>
#include <sys/ipc.h>
#include <sys/sem.h>

int get_sem_count(int sid);

void show_sem_usage(int sid);
int get_sem_count(int sid);
void dispval(int sid);

int main(int argc, char *argv[])

{
key_t key;
int semset_id;

/* Create unique key via call to ftok() */

key = ftok(".", ’s’);
6.4. SYSTEM V IPC 61

/* Open the semaphore set - do not create! */

if((semset_id = semget(key, 1, 0666)) == -1)
{
printf("Semaphore set does not exist\n");
exit(1);
}

show_sem_usage(semset_id);
return(0);
}

void show_sem_usage(int sid)

{
int cntr=0, maxsems, semval;

maxsems = get_sem_count(sid);

while(cntr < maxsems) {

semval = semctl(sid, cntr, GETVAL, 0);
printf("Semaphore #%d: --> %d\n", cntr, semval);
cntr++;
}
}

int get_sem_count(int sid)

{
int rc;
struct semid_ds mysemds;
union semun semopts;

/* Get current values for internal data structure */

semopts.buf = &mysemds;

if((rc = semctl(sid, 0, IPC_STAT, semopts)) == -1) {

perror("semctl");
exit(1);
}

/* return number of semaphores in set */

return(semopts.buf->sem_nsems);
}

void dispval(int sid)

{
int semval;

semval = semctl(sid, 0, GETVAL, 0);

printf("semval is %d\n", semval);
}
62 CHAPTER 6. LINUX INTERPROCESS COMMUNICATIONS

6.4.4 Shared Memory

Basic Concepts
Shared memory can best be described as the mapping of an area (segment) of memory that
will be mapped and shared by more than one process. This is by far the fastest form of IPC,
because there is no intermediation (i.e. a pipe, a message queue, etc). Instead, information
is mapped directly from a memory segment, and into the addressing space of the calling
process. A segment can be created by one process, and subsequently written to and read
from by any number of processes.

Internal and User Data Structures

Let’s briefly look at data structures maintained by the kernel for shared memory segments.

Kernel shmid ds structure As with message queues and semaphore sets, the kernel
maintains a special internal data structure for each shared memory segment which ex-
ists within its addressing space. This structure is of type shmid ds, and is defined in
linux/shm.h as follows:

/* One shmid data structure for each shared memory segment in the syste
struct shmid_ds {
struct ipc_perm shm_perm; /* operation perms */
int shm_segsz; /* size of segment (bytes) */
time_t shm_atime; /* last attach time */
time_t shm_dtime; /* last detach time */
time_t shm_ctime; /* last change time */
unsigned short shm_cpid; /* pid of creator */
unsigned short shm_lpid; /* pid of last operator */
short shm_nattch; /* no. of current attaches */

/* the following are private *

unsigned short shm_npages; /* size of segment (pages) */

unsigned long *shm_pages; /* array of ptrs to frames ->
struct vm_area_struct *attaches; /* descriptors for attaches */
};

Operations on this structure are performed by a special system call, and should not be
tinkered with directly. Here are descriptions of the more pertinent fields:

shm perm
This is an instance of the ipc perm structure, which is defined for us in
linux/ipc.h. This holds the permission information for the segment, includ-
ing the access permissions, and information about the creator of the segment (uid,
etc).
shm segsz
Size of the segment (measured in bytes).
shm atime
Time the last process attached the segment.
shm dtime
Time the last process detached the segment.
6.4. SYSTEM V IPC 63

shm ctime
Time of the last change to this structure (mode change, etc).

shm cpid
The PID of the creating process.

shm lpid
The PID of the last process to operate on the segment.

shm nattch
Number of processes currently attached to the segment.

SYSTEM CALL: shmget()

In order to create a new message queue, or access an existing queue, the shmget() system
call is used.

SYSTEM CALL: shmget();

PROTOTYPE: int shmget ( key_t key, int size, int shmflg );

RETURNS: shared memory segment identifier on success
-1 on error: errno = EINVAL (Invalid segment size specified)
EEXIST (Segment exists, cannot create)
EIDRM (Segment is marked for deletion, or was r
ENOENT (Segment does not exist)
EACCES (Permission denied)
ENOMEM (Not enough memory to create segment)
NOTES:

This particular call should almost seem like old news at this point. It is strikingly similar
to the corresponding get calls for message queues and semaphore sets.
The first argument to shmget() is the key value (in our case returned by a call to
ftok()). This key value is then compared to existing key values that exist within the
kernel for other shared memory segments. At that point, the open or access operation is
dependent upon the contents of the shmflg argument.

IPC CREAT
Create the segment if it doesn’t already exist in the kernel.

IPC EXCL
When used with IPC CREAT, fail if segment already exists.

If IPC CREAT is used alone, shmget() either returns the segment identifier for a
newly created segment, or returns the identifier for a segment which exists with the same
key value. If IPC EXCL is used along with IPC CREAT, then either a new segment is
created, or if the segment exists, the call fails with -1. IPC EXCL is useless by itself, but
when combined with IPC CREAT, it can be used as a facility to guarantee that no existing
segment is opened for access.
Once again, an optional octal mode may be OR’d into the mask.
Let’s create a wrapper function for locating or creating a shared memory segment :
64 CHAPTER 6. LINUX INTERPROCESS COMMUNICATIONS

int open_segment( key_t keyval, int segsize )

{
int shmid;

if((shmid = shmget( keyval, segsize, IPC_CREAT | 0660 )) == -1)

{
return(-1);
}

return(shmid);
}

Note the use of the explicit permissions of 0660. This small function either returns
a shared memory segment identifier (int), or -1 on error. The key value and requested
segment size (in bytes) are passed as arguments.
Once a process has a valid IPC identifier for a given segment, the next step is for the
process to attach or map the segment into its own addressing space.

SYSTEM CALL: shmat()

SYSTEM CALL: shmat();

PROTOTYPE: int shmat ( int shmid, char *shmaddr, int shmflg);

RETURNS: address at which segment was attached to the process, or
-1 on error: errno = EINVAL (Invalid IPC ID value or attach addres
ENOMEM (Not enough memory to attach segment)
EACCES (Permission denied)
NOTES:

If the addr argument is zero (0), the kernel tries to find an unmapped region. This is the
recommended method. An address can be specified, but is typically only used to facilitate
proprietary hardware or to resolve conflicts with other apps. The SHM RND flag can be
OR’d into the flag argument to force a passed address to be page aligned (rounds down to
the nearest page size).
In addition, if the SHM RDONLY flag is OR’d in with the flag argument, then the
shared memory segment will be mapped in, but marked as readonly.
This call is perhaps the simplest to use. Consider this wrapper function, which is passed
a valid IPC identifier for a segment, and returns the address that the segment was attached
to:

char *attach_segment( int shmid )

{
return(shmat(shmid, 0, 0));
}

Once a segment has been properly attached, and a process has a pointer to the start of
that segment, reading and writing to the segment become as easy as simply referencing or
dereferencing the pointer! Be careful not to lose the value of the original pointer! If this
happens, you will have no way of accessing the base (start) of the segment.

SYSTEM CALL: shmctl()

SYSTEM CALL: shmctl();
PROTOTYPE: int shmctl ( int shmqid, int cmd, struct shmid_ds *buf );
RETURNS: 0 on success
6.4. SYSTEM V IPC 65

-1 on error: errno = EACCES (No read permission and cmd is IPC_STAT)

EFAULT (Address pointed to by buf is invalid wi
IPC_STAT commands)
EIDRM (Segment was removed during retrieval)
EINVAL (shmqid invalid)
EPERM (IPC_SET or IPC_RMID command was issued,
calling process does not have write (al
access to the segment)
NOTES:

This particular call is modeled directly after the msgctl call for message queues. In light
of this fact, it won’t be discussed in too much detail. Valid command values are:

IPC STAT
Retrieves the shmid ds structure for a segment, and stores it in the address of the buf
argument
IPC SET
Sets the value of the ipc perm member of the shmid ds structure for a segment. Takes
the values from the buf argument.
IPC RMID
Marks a segment for removal.

The IPC RMID command doesn’t actually remove a segment from the kernel. Rather,
it marks the segment for removal. The actual removal itself occurs when the last process
currently attached to the segment has properly detached it. Of course, if no processes are
currently attached to the segment, the removal seems immediate.
To properly detach a shared memory segment, a process calls the shmdt system call.

SYSTEM CALL: shmdt()

SYSTEM CALL: shmdt();

PROTOTYPE: int shmdt ( char *shmaddr );

RETURNS: -1 on error: errno = EINVAL (Invalid attach address passed)

After a shared memory segment is no longer needed by a process, it should be detached

by calling this system call. As mentioned earlier, this is not the same as removing the
segment from the kernel! After a detach is successful, the shm nattch member of the as-
sociates shmid ds structure is decremented by one. When this value reaches zero (0), the
kernel will physically remove the segment.

shmtool: An interactive shared memory manipulator

Background Our final example of System V IPC objects will be shmtool, which is a
command line tool for creating, reading, writing, and deleting shared memory segments.
Once again, like the previous examples, the segment is created during any operation, if it
did not previously exist.

Command Line Syntax

Writing strings to the segment

shmtool w "text"
66 CHAPTER 6. LINUX INTERPROCESS COMMUNICATIONS

Retrieving strings from the segment

shmtool r

Changing the Permissions (mode)

shmtool m (mode)

Deleting the segment

shmtool d

Examples
shmtool w test
shmtool w "This is a test"
shmtool r
shmtool d
shmtool m 660

The Source
#include <stdio.h>
#include <sys/types.h>
#include <sys/ipc.h>
#include <sys/shm.h>

#define SEGSIZE 100

main(int argc, char *argv[])

{
key_t key;
int shmid, cntr;
char *segptr;

if(argc == 1)
usage();

/* Create unique key via call to ftok() */

key = ftok(".", ’S’);

/* Open the shared memory segment - create if necessary */

if((shmid = shmget(key, SEGSIZE, IPC_CREAT|IPC_EXCL|0666)) == -1)
{
printf("Shared memory segment exists - opening as client\n");

/* Segment probably already exists - try as a client */

if((shmid = shmget(key, SEGSIZE, 0)) == -1)
{
perror("shmget");
exit(1);
}
}
else
{
6.4. SYSTEM V IPC 67

printf("Creating new shared memory segment\n");

}

/* Attach (map) the shared memory segment into the current process */
if((segptr = shmat(shmid, 0, 0)) == -1)
{
perror("shmat");
exit(1);
}

switch(tolower(argv[1][0]))
{
case ’w’: writeshm(shmid, segptr, argv[2]);
break;
case ’r’: readshm(shmid, segptr);
break;
case ’d’: removeshm(shmid);
break;
case ’m’: changemode(shmid, argv[2]);
break;
default: usage();

}
}

writeshm(int shmid, char *segptr, char *text)

{
strcpy(segptr, text);
printf("Done...\n");
}

readshm(int shmid, char *segptr)

{
printf("segptr: %s\n", segptr);
}

removeshm(int shmid)
{
shmctl(shmid, IPC_RMID, 0);
printf("Shared memory segment marked for deletion\n");
}

changemode(int shmid, char *mode)

{
struct shmid_ds myshmds;

/* Get current values for internal data structure */

shmctl(shmid, IPC_STAT, &myshmds);

/* Display old permissions */

printf("Old permissions were: %o\n", myshmds.shm_perm.mode);

/* Convert and load the mode */

sscanf(mode, "%o", &myshmds.shm_perm.mode);
68 CHAPTER 6. LINUX INTERPROCESS COMMUNICATIONS

/* Update the mode */

shmctl(shmid, IPC_SET, &myshmds);

printf("New permissions are : %o\n", myshmds.shm_perm.mode);

}

usage()
{
fprintf(stderr, "shmtool - A utility for tinkering with shared memory\n
fprintf(stderr, "\nUSAGE: shmtool (w)rite <text>\n");
fprintf(stderr, " (r)ead\n");
fprintf(stderr, " (d)elete\n");
fprintf(stderr, " (m)ode change <octal mode>\n");
exit(1);
}

Sven Goldt The Linux Programmer’s Guide

Chapter 7

Sound Programming

A PC has at least one sound device: the internal speaker. But, you can also buy a sound
card to plug into your PC to provide a more sophisticated sound device. Look at the Linux
Sound User’s Guide or the Sound-HOWTO for supported sound cards.

7.1 Programming the internal speaker

Believe it or not, your PC speaker is part of the Linux console and thus a character device.
Therefore, ioctl() requests exist to manipulate it. For the internal speaker the following 2
requests exist:

1. KDMKTONE
Generates a beep for a specified time using the kernel timer.
Example: ioctl (fd, KDMKTONE,(long) argument).

2. KIOCSOUND
Generates an endless beep or stops a currently sounding beep.
Example: ioctl(fd,KIOCSOUND,(int) tone).

The argument consists of the tone value in the low word and the duration in the
high word. The tone value is not the frequency. The PC mainboard timer 8254 is clocked
at 1.19 MHz and so it’s 1190000/frequency. The duration is measured in timer ticks. Both
ioctl calls return immediately so you can this way produce beeps without blocking the
program.
KDMKTONE should be used for warning signals because you don’t have to worry about
stopping the tone.
KIOCSOUND can be used to play melodies as demonstrated in the example program splay
(please send more .sng files to me). To stop the beep you have to use the tone value 0.

7.2 Programming a sound card

For you as a programmer, it is important to know if the current Linux system has a sound
card plugged in. One way to check is to examine /dev/sndstat. If opening /dev/sndstat fails
and errno=ENODEV then no sound driver is activated which means you will get no help
from the kernel sound driver. The same result might be achieved by trying to open /dev/dsp
as long as it is not a link to the pcsnd driver in which case open() will not fail.
If you want to mess with a sound card at the hardware level you know that some com-
bination of outb() and inb() calls will detect the sound card you are looking for.

69
70 CHAPTER 7. SOUND PROGRAMMING

By using the sound driver for your programs, chances are that they will work on other
i386 systems as well, since some clever people decided to use the same driver for Linux,
isc, FreeBSD and most other i386 based systems. It will aid in porting programs if Linux
on other architectures offers the same sound device interface. A sound card is not part of
the Linux console, but is a special device. A sound card mostly offers three main features:

• Digital sample input/output

• Frequency modulation output

• A midi interface

Each of these features have their own device driver interface. For digital samples it
is /dev/dsp, for the frequency modulation it is /dev/sequencer and for the midi interface it
is /dev/midi. The sound settings (like volume, balance or bass) can be controlled via the
/dev/mixer interface. For compatibility reasons a /dev/audio device exists which can read
SUN µ-law sound data, but it maps to the digital sample device.
You are right if you guessed that you use ioctl() to manipulate these devices. The ioctl()
requests are defined in < linux/soundcard.h > and begin with SNDCTL .

Since I don’t own a soundcard someone else has to continue here

Sven van der Meer v0.3.3, 19 Jan 1995
Chapter 8

Character Cell Graphics

This chapter deals with screen input and output that is not pixel based, but character based.
When we say character, we mean a composition of pixels that can be changed depending
on a charset. Your graphic card already offers one or more charsets and operates by default
in text (charset) mode because text can be processed much faster than pixel graphic. There
is more to do with terminals than to use them as simple (dumb) and boring text displays. I
will describe how to use the special features that your linux terminal, especially the linux
console, offers.

• printf, sprintf, fprintf, scanf, sscanf, fscanf

With these functions from libc you can output formatted strings to stdout (standard
output), stderr (standard error) or other streams defined as FILE *stream (files,
for example). Scanf(...) provides a similar way to read formatted input from stdin.
• termcap
The TERMinal CAPabilitie database is a set of terminal description entries in the
ASCII file /etc/termcap. Here you can find information about how to display spe-
cial characters, how to perform operations (delete, insert characters or lines etc) and
how to initialize a terminal. The database is used, for example, by the editor vi.
There are view library functions to read and use the terminal capabilities (see term-
cap(3x)). With this database, programs can work with a variety of terminals with
the same code. Using the termcap database and library functions provides only low
level access to the terminal. Changing attributes or colors, parameterized output and
optimization must be done by the programmer himself.
• terminfo database
The TERMinal INFOrmation database is based on the termcap database and also
describes terminal capabilities, but on a higher level than termcap. Using terminfo,
the program can easily change screen attributes, use special keys such as function
keys and more. The database can be found in /usr/lib/terminfo/[A-z,0-9]*. Every file
describes one terminal.
• curses
Terminfo is a good base to use for terminal handling in a program. The (BSD-
)CURSES library gives you high level access to the terminal and is based on the
terminfo database. Curses allows you to open and manipulate windows on the screen,
provides a complete set of input and output functions, and can alter video attributes
in a terminal independent manner on more than 150 terminals. The curses library
can be found in /usr/lib/libcurses.a. This is the BSD version of curses.
• ncurses
Ncurses is the next improvement. In version 1.8.6it should be compatible with AT&T

71
72 CHAPTER 8. CHARACTER CELL GRAPHICS

curses as defined in SYSVR4 and has some extensions such as color manipulation,
special optimization for output, terminal specific optimizations, and more. It has
been tested on a lot of systems such as Sun-OS, HP and Linux. I recommend using
ncurses instead of the others. On SYSV Unix systems (such as Sun’s Solaris) there
should exist a curses library with the same functionality as ncurses (actually the
solaris curses has some more functions and mouse support).
In the following sections I will describe how to use the different packages to access a
terminal. With Linux we have the GNU-version of termcap and we can use ncurses instead
of curses.

8.1 I/O Function in libc

8.1.1 Formatted Output
The printf(...) functions in libc provide formatted output and allow transformations of the
arguments.
• int fprintf(FILE *stream, const char *format, ...),
will transform the output (arguments to fill in ...) and write it to stream. The
format defined in format will be written, too. The function will return the number
of written characters or a negative value on error.
format contains two kinds of objects
1. normal characters for the output and
2. information how to transform or format the arguments.
Format information must begin with % followed by values for the format followed
by a character for the translation (to print % by itself use %%). Possible values for
the format are:

– Flags
∗ -
The formatted argument will be printed on the left margin (default is the
right margin in the argument field).
∗ +
Every number will be printed with a sign, e.g. +12 or -2.32.
– Blank
When the first character is not a sign, a blank will be inserted.
– 0
For numeric transformation the field width will be filled up with 0’s on the left
side.
– #
Alternate output depending on the transformation for the argument
∗ For o the first number is a 0.
∗ For x or X 0x or 0X will be printed in front of the argument.
∗ For e, E, f or F the output has a decimal point.
∗ For g or G zeroes on the end of the argument are printed.
– A number for the minimal field width.
The transformed argument is printed in a field which is at least as big as the
argument itself. With a number you can make the field width bigger. If the
formatted argument is smaller, then the field width will be filled with zeroes or
blanks.
8.1. I/O FUNCTION IN LIBC 73

Table 8.1: Libc - printf transformations

Character Formatted to
d,i int signed, decimal
o int unsigned, octal, without leading 0
x,X int unsigned, hexadecimal without leading 0x
u int unsigned, decimal
c int (unsigned) single character
s char * up to \0
f double as [-]mmm.ddd
e,E double as [-]m.dddddde±xx
g,G double using %e or %f as needed
p void *
n int *
% %

– A point to separate the field width and the precision.

– A number for the precision.

Possible values for the transformation are in table 8.1 on page 73.

• int printf(const char *format, ...)

Same as fprintf(stdout, ...).

• int sprintf(char *s, const char *format, ...)

Same as printf(...), except that the output will be written to the the character pointer
s (with a following \0).
(Note: You must allocate enough memory for s.)

• vprintf(const char *format, va list arg)

vfprintf(FILE *stream, const char *format, va list arg)
vsprintf(char *s, const char *format, va list arg)
The same as the functions above, only the argument list is set to arg.

8.1.2 Formatted Input

Just as printf(...) is used for formatted output you can use scanf(...) for formatted input.

• int fscanf(FILE *stream, const char *format, ...)

fscanf(...) reads from stream and will transform the input with the rules defined
in format. The results will be placed in the arguments given by ...(Note: the
arguments must be pointer.). The read ends, when no more transformation rules are
in format. fscanf(...) will return EOF when the first transformation reached the
file end or some error occured. Otherwise it will return the number of transformed
arguments.
format can include rules on how to format the input arguments (see table 8.2 on
page 74). It can also include:

– Spaces or tabs, which are ignored.

– any normal character (except %). The characters must be in the input on the
corresponding position.
74 CHAPTER 8. CHARACTER CELL GRAPHICS

Table 8.2: Libc - scanf transformations

Character Input - Argumenttyp
d decimal integer - int *
i integer - int * (input can be octal or hex)
o octal integer - int * (with or without leading 0)
u decimal, unsigned - unsigned int *
x hex integer - int * (with or without leading 0x)
c one or more characters - char * (without \0)
s characters(without space, tab ...) - char * (with \0)
e,f,gf float - float * (e.g [-]m.dddddde±xx)
p pointer - void *
n number of transformed arguments - int *
[...] nonempty characters in input - char *
[ˆ...] exclude those nonempty characters - char *
% %
h can be before d,i,n,o,u and x when the pointer is short
l can be before d,i,n,o,u and x when the pointer is long
l can be before e,f and g when the pointer is double
L can be before e,f and g when the pointer is long double

– transformation rules, which assembled with a %, the optional character * (this

will permit fscanf(...) to assign to an argument), an optional number, an op-
tional character h, l or L (this is for the length of the target) and the transforma-
tion character.

• int scanf(const char *format, ...)

The same as fscanf(stdin,...).

• int sscanf(char *str, const char *format, ...)

As scanf(...), but the input comes from str.

8.2 The Termcap Library

8.2.1 Introduction
The termcap library is an API to the termcap database which can be found in /etc/termcap/.
The library functions allow the following actions:

• Get a description of the current terminal: tgetent(...).

• Search the description for information: tgetnum(...), tgetflag(...), tgetstr(...).

• Encode numeric parameters in a terminal specific form: tparam(...), tgoto(...).

• Compute and perform padding tputs(...).

Programs using the termcap library must include termcap.h and should be linked with:

gcc [flags] files -ltermcap

Termcap functions are terminal independent routines but only give the programmer low
level access to the terminal. For a higher level package, curses or ncurses should be used.
8.2. THE TERMCAP LIBRARY 75

8.2.2 Find a Terminal Description

• int tgetent(void *buffer, const char *termtype)
On the Linux operating system the current terminal name is contained in the envi-
ronment variable TERM. So, termtype is the result of a call to getenv(3).
For buffer, no memory has to be allocated when using the GNU version of term-
cap. This is what we can assume under Linux! Otherwise, you’ll have to allocate
2048 Bytes. (Formerly, buffer only needed to be 1024 Bytes, but the size has
doubled).
tgetent(...) returns 1 on success and 0 when the database is found but has no entry
for TERM. Other errors will return different values.
The following example should explain how to use tgetent(...):

#define buffer 0
char *termtype=getenv("TERM");
int ok;

ok=tgetent(buffer,termtype);
if(ok==1)
/* all right, we have the entry */
else if(ok==0)
/* uups, something wrong with TERM
* check termtype first, then termcap database
*/
else
/* huuu, fatal error */

By, default termcap uses /etc/termcap/ as the database. If the environment vari-
able TERMCAP is set, with $HOME/mytermcap for instance, all functions will use
mytermcap instead of /etc/termcap. With no leading slash in TERMCAP, the defined
value is used as a name for a terminal.

8.2.3 Look at a Terminal Description

Every piece of information is called a capability, every capability is a two letter code, and
every two letter code is followed by the value for the capability. Possible types are:

• Numeric: For instance co – number of columns

• Boolean or Flag: For instance hc – hardcopy terminal

• String: For instance st – set tab stop

Each capability is associated with a single value type. (co is always numeric, hc is
always a flag and st is always a string). There are three different types of values, so there
are also three functions to interrogate them. char *name is the two letter code for the
capability.

• int tgetnum(char *name)

Get a capability value that is numeric, such as co. tgetnum(...) returns the numeric
value if the capability is available, otherwise 1. (Note: the returned value is not
negative.)

• int tgetflag(char *name)

Get a capability value that is boolean (or flag). Returns 1 if the flag is present, 0
otherwise.
76 CHAPTER 8. CHARACTER CELL GRAPHICS

• char *tgetstr(char *name, char **area)

Get a capability value that is a string. Returns a pointer to the string or NULL if not
present. In the GNU version, if area is NULL, termcap will allocate memory by
itself. Termcap will never refer to this pointer again, so don’t forget to free name
before leaving the program. This method is preferred, because you don’t know how
much space is needed for the pointer, so let termcap do this for you.

char *clstr, *cmstr;

int lines, cols;

void term_caps()
{
char *tmp;

clstr=tgetstr("cl",0); /* clear screen */

cmstr=tgetstr("cm",0); /* move y,x */

lines=tgetnum("li"); /* terminal lines */

cols=tgetnum("co"); /* terminal columns */

tmp=tgetstr("pc",0); /* padding character */

PC=tmp ? *tmp : 0;
BC=tgetstr("le",0); /* cursor left one char */
UP=tgetstr("up",0); /* cursor up one line */
}

8.2.4 Termcap Capabilities

Boolean Capabilities
5i Printer will not echo on screen
am Automatic margins which means automatic line wrap
bs Control-H (8 dec.) performs a backspace
bw Backspace on left margin wraps to previous line and right margin
da Display retained above screen
db Display retained below screen
eo A space erases all characters at cursor position
es Escape sequences and special characters work in status line
gn Generic device
hc This is a hardcopy terminal
HC The cursor is hard to see when not on bottom line
hs Has a status line
hz Hazel tine bug, the terminal can not print tilde characters
in Terminal inserts nulls, not spaces, to fill whitespace
km Terminal has a meta key
mi Cursor movement works in insert mode
ms Cursor movement works in standout/underline mode
NP No pad character
NR ti does not reverse te
nx No padding, must use XON/XOFF
os Terminal can overstrike
ul Terminal underlines although it can not overstrike
xb Beehive glitch, f1 sends ESCAPE, f2 sends ˆC
xn Newline/wraparound glitch
xo Terminal uses xon/xoff protocol
xs Text typed over standout text will be displayed in standout
xt Teleray glitch, destructive tabs and odd standout mode
8.2. THE TERMCAP LIBRARY 77

Numeric Capabilities

co Number of columns lh Height of soft labels

dB Delay in milliseconds for backspace lm Lines of memory
on hardcopy terminals lw Width of soft labels
dC Delay in milliseconds for carriage li Number of lines
return on hardcopy terminals Nl Number of soft labels
dF Delay in milliseconds for form feed pb Lowest baud rate which needs
on hardcopy terminals padding
dN Delay in milliseconds for new line sg Standout glitch
on hardcopy terminals ug Underline glitch
dT Delay in milliseconds for tab stop vt virtual terminal number
on hardcopy terminals ws Width of status line if
dV Delay in milliseconds for vertical different from screen width
tab stop on hardcopy terminals
it Difference between tab positions

String Capabilities

!1 shifted save key &9 shifted begin key

!2 shifted suspend key *0 shifted find key
!3 shifted undo key *1 shifted command key
#1 shifted help key *2 shifted copy key
#2 shifted home key *3 shifted create key
#3 shifted input key *4 shifted delete character
#4 shifted cursor left key *5 shifted delete line
%0 redo key *6 select key
%1 help key *7 shifted end key
%2 mark key *8 shifted clear line key
%3 message key *9 shifted exit key
%4 move key 0 find key
%5 next-object key 1 begin key
%6 open key 2 cancel key
%7 options key 3 close key
%8 previous-object key 4 command key
%9 print key 5 copy key
%a shifted message key 6 create key
%b shifted move key 7 end key
%c shifted next key 8 enter/send key
%d shifted options key 9 exit key
%e shifted previous key al Insert one line
%f shifted print key AL Insert %1 lines
%g shifted redo key ac Pairs of block graphic characters to
%h shifted replace key map alternate character set
%i shifted cursor right key ae End alternative character set
%j shifted resume key as Start alternative character set
&0 shifted cancel key for block graphic characters
&1 reference key bc Backspace, if not ˆH
&2 refresh key bl Audio bell
&3 replace key bt Move to previous tab stop
&4 restart key cb Clear from beginning of line
&5 resume key to cursor
&6 save key cc Dummy command character
&7 suspend key cd Clear to end of screen
&8 undo key ce Clear to end of line
78 CHAPTER 8. CHARACTER CELL GRAPHICS

ch Move cursor horizontally only to K4 bottom left key on keypad

column %1 K5 bottom right key on keypad
cl Clear screen and cursor home k0 Function key 0
cm Cursor move to row %1 and k1 Function key 1
column %2 (on screen) k2 Function key 2
CM Move cursor to row %1 and k3 Function key 3
column %2 (in memory) k4 Function key 4
cr Carriage return k5 Function key 5
cs Scroll region from line %1 to %2 k6 Function key 6
ct Clear tabs k7 Function key 7
cv Move cursor vertically only to k8 Function key 8
line %1 k9 Function key 9
dc Delete one character k; Function key 10
DC Delete %1 characters ka Clear all tabs key
dl Delete one line kA Insert line key
DL Delete %1 lines kb Backspace key
dm Begin delete mode kB Back tab stop
do Cursor down one line kC Clear screen key
DO Cursor down #1 lines kd Cursor down key
ds Disable status line kD Key for delete character
eA Enable alternate character set under cursor
ec Erase %1 characters starting at ke turn keypad off
cursor kE Key for clear to end of line
ed End delete mode kF Key for scrolling forward/down
ei End insert mode kh Cursor home key
ff Form-feed character on hardcopy kH Cursor home down key
terminals kI Insert character/Insert mode key
fs Return character to its position kl Cursor left key
before going to status line kL Key for delete line
F1 The string sent by function key f11 kM Key for exit insert mode
F2 The string sent by function key f12 kN Key for next page
F3 The string sent by function key f13 kP Key for previous page
... ... kr Cursor right key
F9 The string sent by function key f19 kR Key for scrolling backward/up
FA The string sent by function key f20 ks Turn keypad on
FB The string sent by function key f21 kS Clear to end of screen key
... ... kt Clear this tab key
FZ The string sent by function key f45 kT Set tab here key
Fa The string sent by function key f46 ku Cursor up key
Fb The string sent by function key f47 l0 Label of zeroth function key,
... ... if not f0
Fr The string sent by function key f63 l1 Label of first function key,
hd Move cursor a half line down if not f1
ho Cursor home l2 Label of first function key,
hu Move cursor a half line up if not f2
i1 Initialization string 1 at login ... ...
i3 Initialization string 3 at login la Label of tenth function key,
is Initialization string 2 at login if not f10
ic Insert one character le Cursor left one character
IC Insert %1 characters ll Move cursor to lower left corner
if Initialization file LE Cursor left %1 characters
im Begin insert mode LF Turn soft labels off
ip Insert pad time and needed special LO Turn soft labels on
characters after insert mb Start blinking
iP Initialization program MC Clear soft margins
K1 upper left key on keypad md Start bold mode
K2 center key on keypad me End all mode like so, us, mb,
K3 upper right key on keypad md and mr
8.2. THE TERMCAP LIBRARY 79

mh Start half bright mode sa Set %1 %2 %3 %4 %5 %6 %7 %8

mk Dark mode (Characters invisible) %9 attributes
ML Set left soft margin SA enable automatic margins
mm Put terminal in meta mode sc Save cursor position
mo Put terminal out of meta mode se End standout mode
mp Turn on protected attribute sf Normal scroll one line
mr Start reverse mode SF Normal scroll %1 lines
MR Set right soft margin so Start standout mode
nd Cursor right one character sr Reverse scroll
nw Carriage return command SR scroll back %1 lines
pc Padding character st Set tabulator stop in all rows at
pf Turn printer off current column
pk Program key %1 to send string %2 SX Turn on XON/XOFF flow control
as if typed by user ta move to next hardware tab
pl Program key %1 to execute string tc Read in terminal description from
%2 in local mode another entry
pn Program soft label %1 to to show te End program that uses
string %2 cursor motion
po Turn the printer on ti Begin program that uses
pO Turn the printer on for %1 cursor motion
(<256) bytes ts Move cursor to column %1 of
ps Print screen contents on printer status line
px Program key %1 to send string uc Underline character under cursor
%2 to computer and move cursor right
r1 Reset string 1, set sane modes ue End underlining
r2 Reset string 2, set sane modes up Cursor up one line
r3 Reset string 3, set sane modes UP Cursor up %1 lines
RA disable automatic margins us Start underlining
rc Restore saved cursor position vb Visible bell
rf Reset string file name ve Normal cursor visible
RF Request for input from terminal vi Cursor invisible
RI Cursor right %1 characters vs Standout cursor
rp Repeat character %1 for %2 times wi Set window from line %1 to %2
rP Padding after character sent in and column %3 to %4
replace mode XF XOFF character if not ˆS
rs Reset string
RX Turn off XON/XOFF flow control
80 CHAPTER 8. CHARACTER CELL GRAPHICS

8.3 Ncurses - Introduction

The following terminology will be used in this chapter:

• window - is an internal representation containing an image of a part of the screen.

WINDOW is defined in ncurses.h.

• screen - is a window with the size of the entire screen (from the upper left to the
lower right). Stdscr and curscr are screens.

• terminal - is a special screen with information about what the screen currently looks
like.

• variables - the following variables and constants defined in ncurses.h

– WINDOW *curscr - current screen

– WINDOW *stdscr - standard screen
– int LINES - lines on the terminal
– int COLS - columns on the terminal
– bool TRUE - true flag, 1
– bool FALSE - false flag, 0
– int ERR - error flag, -1
– int OK - ok flag, 0

• functions - in the function description the arguments are of the following type:

– win - WINDOW*
– bf - bool
– ch - chtype
– str - char*
– chstr - chtype*
– fmt - char*
– otherwise int

Usually a program using the ncurses library looks like this:

#include <ncurses.h>
...
main()
{
...
initscr();
/* ncurses function calls */
endwin();
...
}

Including ncurses.h will define variables and types for ncurses, such as WINDOW and
function prototypes. It automatically includes stdio.h, stdarg.h, termios.h and unctrl.h.
initscr() is used to initialize the ncurses data structures and to read the proper terminfo
file. Memory for stdscr and curscr will be allocated. If an error occurs, initscr will return
ERR, otherwise a pointer to stdscr will be returned. Additionally, the screen will be erased
and LINES and COLS will be initialized.
8.3. NCURSES - INTRODUCTION 81

endwin() will clean up all allocated resources from ncurses and restore the tty modes
to the status they had before calling initscr(). It must be called before any other function
from the ncurses library and endwin() must be called before your program exits. When you
want to do output to more than one terminal, you can use newterm(...) instead of initscr().
Compile the program with:

gcc [flags] files -lncurses

In flags you can include anything you like (see gcc(1)). Since the path for ncurses.h has
changed you have to include the following line:

-I/usr/include/ncurses

Otherwise, ncurses.h, nterm.h, termcap.h and unctrl.h will not be found. Possible other
flags for Linux are:

-O2 -ansi -Wall -m486

O2 tells gcc to do some optimization, -ansi is for ansi conformant c-code, -Wall will
print out all warnings, -m486 will use optimized code for an Intel 486 (the binary can be
used on an Intel 386, too).
The ncurses library can be found in /usr/lib/. There are three versions of the ncurses
library:
• libncurses.a the normal ncurses library.
• libdcurses.a ncurses for debugging.
• libpcurses.a ncurses for profiling (since 1.8.6libpcurses.a exists no longer ?).
• libcurses.a No fourth version, but the original BSD curses (in my slackware 2.1.0 it
is the bsd package).
The data structures for the screen are called windows as defined in ncurses.h. A win-
dow is something like a character array in memory which the programmer can manipulate
without output to the terminal. The default window is stdscr with the size of the terminal.
You can create other windows with newwin(...).
To update the physical terminal optimally, ncurses has another window declared, curscr.
This is an image of how the terminal actually looks and stdscr is an image of how the
terminal should look. The output will be done when you call refresh(). Ncurses will then
update curscr and the physical terminal with the information in stdscr. The library functions
will use internal optimization for the update process so you can change different windows
and then update the screen at once in the most optimal way.
With the ncurses functions you can manipulate the data structure window. Functions
beginning with w allow you to specify a window, while others will usually affect stdscr.
Functions beginning with mv will move the cursor to the position y,x first.
A character has the type chtype which is long unsigned int to store additional informa-
tion about it (attributes etc.).
Ncurses use the terminfo database. Normally the database is located in usr/lib/terminfo/
and ncurses will look there for local terminal definitions. If you want to test some other def-
initions for a terminal without changing the original terminfo, set the environment variable
TERMINFO. Ncurses will check this variable and use the definitions stored there instead
of /usr/lib/terminfo/.
Current ncurses version is 1.8.6().
At the end of this chapter you can find a table with an overview for the BSD-Curses,
ncurses and the curses from Sun-OS 5.4. Refer to it when you want to look for a specific
function and where it is implemented.
82 CHAPTER 8. CHARACTER CELL GRAPHICS

8.4 Initializing
• WINDOW *initscr()
This is the first function usually called from a program using ncurses. In some
cases it is useful to call slk init(int), filter(), ripoffline(...) or use env(bf) before
initscr(). When using multiple terminals (or perhaps testing capabilities), you can
use newterm(...) instead of initscr().
initscr() will read the proper terminfo file and initialize the ncurses data structures,
allocate memory for curscr and stdscr and set LINES and COLS to the values the
terminal has. It will return a pointer to stdscr or ERR when an error has occured.
You don’t need to initialize the pointer with:
stdscr=initscr();

initscr() will do this for you. If the return value is ERR, your program should exit
because no ncurses function will work:
if(!(initscr())){
fprintf(stderr,"type: initscr() failed\n\n");
exit (1);
}

• SCREEN *newterm(char *type, FILE *outfd, FILE *infd)

For multiple terminal output call newterm(...) for every terminal you would ac-
cess with ncurses instead of initscr(). type is the name of the terminal as con-
tained in $TERM (ansi, xterm, vt100, for example), outfd is the output pointer and
infd is the pointer used for input. Call endwin() for every terminal opened with
newterm(...).

• SCREEN *set term(SCREEN *new)

With set term(SCREEN) you can switch the current terminal. All functions will
affect the current terminal which is set with set term(SCREEN).

• int endwin()
endwin() will do the cleanup, restore the terminal modes in the state they had before
calling initscr() and move the cursor to the lower left corner. Don’t forget to close
all opened windows before you call endwin() to exit your program.
An additional call to refresh() after endwin() will restore the terminal to the status
it had before calling initscr() (visual-mode) otherwise it will be cleared (non-visual-
mode).

• int isendwin()
Returns TRUE if endwin() was called with a following refresh(), otherwise FALSE.

• void delscreen(SCREEN* sp)

After endwin() call delscreen(SCREEN) to free up all occupied resources, when
SCREEN is no longer needed. (Note: not implemented yet.)

8.5 Windows
Windows can be created, deleted, moved, copied, touched, duplicated and more.

• WINDOW *newwin(nlines, ncols, begy, begx)

begy and begx are the window coordinates of the upper left corner. nlines is
an integer with the number of lines and ncols is an integer with the number of
columns.
8.5. WINDOWS 83

Figure 8.1: Ncurses - scheme for newwin

0 begx

0 COLS -

. . . . . . . . ncols
. . . . . . . . . . -
.
begy
.6
.
.
.
.
.
nlines . newwin(nlines, ncols, begy, begx)
.
.
.
.
.
?

LINES
?

WINDOW *mywin;
mywin=newwin(10,60,10,10);

The upper left corner of our window is in line 10 and column 10 and the window has
10 lines and 60 columns. If nlines is zero, the window will have LIN ES − begy
rows. In the same way the, window will have COLS − begx columns when ncols
is zero.
When you call newwin(...) with all argument zero:
WINDOW *mywin;
mywin=newwin(0,0,0,0);

the opened window will have the size of the screen.

With LINES and COLS we can open windows in the middle of the screen, whatever
dimension it has:

#define MYLINE (int) ((LINES-22)/2)

#define MYCOL ((COLS-70)/2)
#define MYLINES 22
#define MYCOLS 70
...
WINDOW *win;
...
if(!(initscr())){
fprintf(stderr,"type: initscr() failed\n\n");
exit(1);
}
...
if ((LINES<22)||(COLS<70)){
84 CHAPTER 8. CHARACTER CELL GRAPHICS

fprintf(stderr,"screen to small\n\n");
endwin(); exit (1);
}
win=newwin(MYLINES,MYCOLS,MYLINE,MYCOL);
...

This will open a window with 22 lines and 70 rows in the middle of the screen.
Check the screen size before opening windows. In the Linux console we have 25 or
more lines and 80 or more columns, but in xterms this may not be the case (they’re
resizable).
Alternatively, use LINES and COLS to adapt two windows to the screen size:

#define MYROWS (int) (LINES/2+LINES/4)

#define MYCOLS (int) (COLS/2+COLS/4)
#define LEFTROW (int) ((LINES-MYROWS)/2)
#define LEFTCOL (int) (((COLS-2)-MYCOLS)/2)
#define RIGHTROW (int) (LEFTROW)
#define RIGHTCOL (int) (LEFTROW+2+MYCOLS)
#define WCOLS (int) (MYCOLS/2)
...
WINDOW *leftwin, *rightwin;
...
leftwin=newwin(MYROWS, WCOLS, LEFTROW, LEFTCOL);
rightwin=newwin(MYROWS, WCOLS, RIGHTROW, RIGHTCOL);
...

See screen.c in the example directory for more explanations.

• int delwin(win)
Delete the window win. When there are subwindows delete them before win. It
will free up all resources occupied by win. Delete all windows you have opened
before calling endwin().
• int mvwin(win, by, bx)
Will move a window to the coordinates by,bx. If this means moving the window
beyond the edges of the screen, nothing is done, and ERR is returned.
• WINDOW *subwin(origwin, nlines, ncols, begy, begx)
Returns a subwindow in the middle of origwin. When you change one of the two
windows (origwin or the new one) this change will be reflected in both windows.
Call touchwin(origwin) before the next refresh().
begx and begy are relative to the screen, not to origwin.
• WINDOW *derwin(origwin, nlines, ncols, begy, begx)
The same as subwin(...) except that begx and begy are relative to the window
origwin than to the screen.
• int mvderwin(win, y, x)
Will move win inside its parent window. (Note: not implemented yet.)
• WINDOW *dupwin(win)
Duplicate the window win.
• int syncok(win, bf)
void wsyncup(win)
void wcursyncup(win)
void wsyncdown(win)
(Note: not implemented yet.)
8.6. OUTPUT 85

• int overlay(win1, win2)

int overwrite(win1, win2)
overlay(...) will copy all text from win1 to win2 without copying blanks. over-
write(...) does the same, but copies blanks, too.

• int copywin(win1, win2, sminrow, smincol, dminrow,

dmincol,
dmaxrow, dmaxcol, overlay)
Similar to overlay(...) and overwrite(...), but provides control over what region of
the window to copy.

8.6 Output
• int addch(ch)
int waddch(win, ch)
int mvaddch(y, x, ch)
int mvwaddch(win, y, x, ch)
These functions are used for character output to a window. They will manipulate
the window and you will have to call refresh() to put it on screen. addch(...) and
waddch(...) put the character ch in the window stdscr or win. mvaddch(...) and
mvwaddch(...) do the same except that they move the cursor to position y,x first.

• int addstr(str)
int addnstr(str, n)
int waddstr(win, str)
int waddnstr(win, str, n)
int mvaddstr(y, x, str)
int mvaddnstr(y, x, str, n)
int mvwaddstr(win, y, x, str)
int mvwaddnstr(win, y, x, str, n)
These functions write a string to a window and are equivalent to series of calls to
addch(...). str is a null terminated string (”blafoo\0”). Functions with w write the
string str to the window win, while other funcions write to stdscr. Functions with
n write n characters of str. If n is -1, the entire string str is written.

• int addchstr(chstr)
int addchnstr(chstr, n)
int waddchstr(win, chstr)
int waddchnstr(win, chstr, n)
int mvaddchstr(y, x, chstr)
int mvaddchnstr(y, x, chstr, n)
int mvwaddchstr(win, y, x, chstr)
int mvwaddchnstr(win, y, x, chstr, n)
These functions copy chstr to the window image (stdscr or win). The starting
position is the current cursor position. Functions with n write n characters of chstr.
If n is -1, the entire string chstr is written. The cursor is not moved and no control
character check is done. These functions are faster than the addstr(...) routines.
chstr is a pointer to an array of chtype.

• int echochar(ch)
int wechochar(win, ch)
The same as call addch(...) (waddch(...) followed by refresh() (wrefresh(win).
86 CHAPTER 8. CHARACTER CELL GRAPHICS

8.6.1 Formatted Output

• int printw(fmt, ...)
int wprintw(win, fmt, ...)
int mvprintw(y, x, fmt, ...)
int mvwprintw(win, y, x, fmt, ...)
int vwprintw(win, fmt, va list)
These functions correspond to printf(...) and its counterparts from libc.
In the libc package printf(...) is used for formatted output. You can define an output
string and include variables of different types in it. See section 8.1.1 on page 72 for
more.
For the use of vwprintw(...) you have to include also varargs.h.

8.6.2 Insert Characters/Lines

• int insch(c)
int winsch(win, c)
int mvinsch(y,x,c)
int mvwinsch(win,y,x,c)
Character ch is inserted to the left of the cursor and all characters are moved one
position to the right. The character on the right end of the line may be lost).
• int insertln()
int winsertln(win)
Insert a blank line above the current one. (The bottom line will be lost).
• int insdelln(n)
int winsdelln(win, n)
For positive n these functions will insert n lines above the cursor in the appropriate
window (so the n bottom lines will be lost). When n is negative, n lines under the
cursor will be deleted and the rest will moved up.
• int insstr(str)
int insnstr(str, n)
int winsstr(win, str)
int winsnstr(win, str, n)
int mvinsstr(y, x, str)
int mvinsnstr(y, x, str, n)
int mvwinsstr(win, y, x, str)
int mvwinsnstr(win, y, x, str, n)
These functions will insert str in the current line left from the cursor (as many
characters as fit to the line). The characters on the right of the cursor are moved
right and will be lost when the end of the line is reached. The cursor position is not
changed.
y and x are the coordinates to which the cursor is moved before str will be inserted,
n is the number of characters to insert (with n=0 the entire string is inserted).

8.6.3 Delete Characters/Lines

• int delch()
int wdelch(win)
int mvdelch(y, x)
int mvwdelch(win, y, x)
Delete the character under the cursor and move the remaining characters to the right
of the cursor one position to the left.
8.6. OUTPUT 87

Table 8.3: Ncurses - border characters

Character Position Default
tl top left ACS ULCORNER
ts top side ACS HLINE
tr top right ACS URCORNER
ls left side ACS VLINE
rs right side ACS VLINE
bl bottom left ACS LLCORNER
bs bottom side ACS HLINE
br bottom right ACS LRCORNER
rt right tee ACS RTEE
lt left tee ACS LTEE
tt top tee ACS TTEE
bt bottom tee ACS BTEE

Figure 8.2: Ncurses - box characters

tl ts tt ts tr

ls rs

lt rt

ls rs

bl bs bt bs br

y and x are the coordinates to which the cursor will be moved to before deleting.
• int deleteln()
int wdeleteln(win)
Delete the line under the cursor and move all other lines below one position up.
Additionally, the bottom line of the window will be erased.

8.6.4 Boxes and Lines

• int border(ls, rs, ts, bs, tl, tr, bl, br)
int wborder(win, ls, rs, ts, bs, tl, tr, bl, br)
int box(win, vert, hor)
Draw a border around the edges of a window (stdscr or win). In the following table
you see the characters and their default values when zero in a call to box(...). In the
picture you can see the position from the characters in a box.
• int vline(ch, n)
int wvline(win, ch, n)
88 CHAPTER 8. CHARACTER CELL GRAPHICS

int hline(ch, n)
int whline(win, ch, n)
These functions draw a vertical or horizontal line starting at the current cursor posi-
tion. ch is the character to use and n is the number of characters to draw. The cursor
position is not advanced.

8.6.5 Background Character

• void bkgdset(ch)
void wbkgdset(win, ch)
Set the background character and attribute for the screen or a window. The attribute
in ch will be ORed with every non blank character in the window. The background
is then part of the window and will not be changed with scrolling and in- or output.

• int bkgd(ch)
int wbkgd(win, ch)
Will change the background character and attribute to ch.

8.7 Input
• int getch()
int wgetch(win)
int mvgetch(y, x)
int mvwgetch(win, y, x)
getch() will read input from the terminal in a manner depending on whether delay
mode is set or not. If delay is on, getch() will wait until a key is pressed, otherwise
it will return the key in the input buffer or ERR if this buffer is empty. mvgetch(...)
and mvwgetch(...) will move the cursor to position y,x first. The w functions read
input from the terminal related to the window win, getch() and mvgetch(...) from
the terminal related to stdscr.
With keypad(...) enabled, getch() will return a code defined in ncurses.h as KEY *
macros when a function key is pressed. When ESCAPE is pressed (which can be the
beginning of a function key) ncurses will start a one second timer. If the remainder
of the keystroke is not finished in this second, the key is returned. Otherwise, the
function key value is returned. (If necessary, use notimeout() to disable the second
timer).

• int ungetch(ch)
Will put the character ch back to the input buffer.

• int getstr(str)
int wgetstr(win, str)
int mvgetstr(y, x, str)
int mvwgetstr(win, y, x, str)
int wgetnstr(win, str, n)
These functions will do a series of calls to getch() until a newline is received. The
characters are placed in str (so don’t forget to allocate memory for your character
pointer before calling getstr(...)). If echo is enabled the string is echoed (use noe-
cho() to disable echo) and the user’s kill and delete characters will be interpreted.

• chtype inch()
chtype winch(win)
chtype mvinch(y, x)
chtype mvwinch(win, y, x)
8.8. OPTIONS 89

These functions return a character from the screen or window. Because the type of
the return value is chtype attribute information is included. This information can
be extracted from the character using the A * constants (see table 8.4 on page 96).

• int instr(str)
int innstr(str, n)
int winstr(win, str)
int winnstr(win, str, n)
int mvinstr(y, x, str)
int mvinnstr(y, x, str, n)
int mvwinstr(win, y, x, str)
int mvwinnstr(win, y, x, str, n)
Return a character string from the screen or a window. (Note: not implemented yet.)

• int inchstr(chstr)
int inchnstr(chstr, n)
int winchstr(win, chstr)
int winchnstr(win, chstr, n)
int mvinchstr(y, x, chstr)
int mvinchnstr(y, x, chstr, n)
int mvwinchstr(win, y, x, chstr)
int mvwinchnstr(win, y, x, chstr, n)
Return a chtype string from the screen or window. In the string, attribute information
is included for every character. (Note: not implemented yet, lib inchstr not included
in the ncurses lib.)

8.7.1 Formated Input

• int scanw(fmt, ...)
int wscanw(win, fmt, ...)
int mvscanw(y, x, fmt, ...)
int mvwscanw(win, y, x, fmt, ...)
int vwscanw(win, fmt, va list)
These are similar to scanf(...) from libc(see section 8.1.2 on page 73). wgetstr(...) is
called and the results will be used as input for the scan.

8.8 Options
Output Options

• int idlok(win, bf)

void idcok(win, bf)
Enable or disable terminal’s insert/delete features for the window (idlok(...) for lines
and idcok(...) for characters). (Note: idcok(...) not implemented yet.)

• void immedok(win, bf)

If set TRUE, every change to the window win will cause a refresh to the physical
screen. This can decrease the performance of a program, so the default value is
FALSE. (Note: not implemented yet.)

• int clearok(win, bf)

If bf is TRUE the next call to wrefresh(win) will clear the screen and redraw it
completely. (as in the editor vi when you press CTRL+L).
90 CHAPTER 8. CHARACTER CELL GRAPHICS

??? KEY KEY NUM / * -

HOME PPAGE

CTRL KEY KEY KEY KEY KEY

+D END NPAGE HOME UP PPAGE +
KEY ??? KEY
LEFT RIGHT

KEY KEY KEY KEY

UP END DOWN NPAGE CTRL

KEY KEY KEY ??? KEY +M

LEFT DOWN RIGHT DC

• int leaveok(win, bf)

The default behavior is for ncurses to leave the physical cursor in the same place
it was on the last refresh of the window. Programs which don’t use the cursor can
set leaveok(...) TRUE and save the time normally required for cursor motion. In
addition, ncurses will try to make the terminal cursor invisible.

• int nl()
int nonl()
Control the translation for newline. Turned on with nl() will translate a newline in
carriage return and line feed on output. nonl() will turn translation off. With disabled
translation ncurses can do faster cursor motion.

8.8.1 Input Options

• int keypad(win, bf)
If TRUE, it enables the keypad on the keyboard of the user’s terminal when waiting
for input. Ncurses will then return a key code defined in ncurses.h as KEY * for
the function and arrow keys on the keypad. This is very useful for a PC keyboard
because you can enable the numerical block and the cursor keys.

• int meta(win, bf)

If TRUE, the key codes returned from getch() are 8-bit-clean (the highest bit will not
be stripped).

• int cbreak()
int nocbreak()
int crmode()
int nocrmode()
cbreak() and nocbreak() will turn the terminal CBREAK mode on or off. When
CBREAK is on, input from a read will be immediately available to the program,
when off the input will be buffered until newline occurs. (Note: crmode() and
nocrmode() are for upward compatibility, don’t use them.)

• int raw()
int noraw()
Turn RAW mode on or off. RAW is the same as CBREAK, except that in RAW
mode no special character processing will be done.

• int echo()
int noecho()
Set echo() to echo input typed by the user and noecho() to be silent about it.
8.8. OPTIONS 91

• int halfdelay(t)
As cbreak() with a delay of t seconds.

• int nodelay(win, bf)

Terminal will be set to no blocking mode. cetch() will return ERR if no input is
ready. If set to FALSE, getch() will wait until a key is pressed.

• int timeout(t)
int wtimeout(win, t)
It is recommended to use these functions instead of halfdelay(t) and node-
lay(win,bf). The result of getch() depends on the value of t. If t is positive, the
read is blocked for t milliseconds, if t is zero, no blocking is done, and when t is
negative the program blocks until input is available.

• int notimeout(win, bf)

If bf is TRUE, getch() will use a special timer (of one second length) to interpret
and input sequence beginning with keys as ESCAPE etc.

• int typeahead(fd)
If fd is -1 no typeahead check will be done, else ncurses will use the file descriptor
fd instead of stdin for these checks.

• int intrflush(win, bf)

When enabled with bf TRUE an interrupt key pressed on the terminal (quit, break
...) will flush all output in the tty driver queue.

• void noqiflush()
void qiflush()
(Note: not implemented yet.)

8.8.2 Terminal Attributes

• int baudrate()
Returns the terminal speed in bps.

• char erasechar()
Returns the current erase character.

• char killchar()
Returns the current kill character.

• int has ic()

int has il()
has ic() returns TRUE if the terminal has insert/delete character capability, has il()
returns TRUE when the terminal has insert/delete line capability. Otherwise the func-
tions return ERR. (Note: not implemented yet.)

• char *longname()
The returned pointer gives access to the description of the current terminal.

• chtype termattrs()
(Note: not implemented yet.)

• char *termname()
Returns the contents of TERM from the users environment. (Note: not implemented
yet.)
92 CHAPTER 8. CHARACTER CELL GRAPHICS

8.8.3 Use Options

Now we have seen the window options and terminal modes it is time to describe their use.
First, on Linux you should enable the keypad. This will allow use of the cursor keys
and the numeric block on the PC keyboard.

keypad(stdscr,TRUE);

Now, there are two main types of input.

1. The program wants the user to enter a key and then will call a function depend on
this key. (For example, something like ”press ’q’ for quit” and wait for q)
2. The program wants a string of characters typed by the user in a mask on the screen.
For example: a directory or an address in a database.

For the first we use the following options and modes and the while loop will work
correctly.

char c;

noecho();
timeout(-1);
nonl();
cbreak();
keypad(stdscr,TRUE);
while(c=getch()){
switch(c){
case ’q’: your_quit_function();
default: break;
}
}

The program will hang until a key is pressed. If the key was q we call our quit function
else we wait for other input.
The switch statement can be expanded until we have an input function that fits our
wishes. Use the KEY * macros to check special keys, for instance

KEY_UP KEY_RIGHT KEY_A1 KEY_B2 KEY_C1

KEY_DOWN KEY_LEFT KEY_A3 KEY_C3

for the cursor keys on the keyboard. For a file viewer the loop can look like this:

int loop=TRUE;
char c;
enum{UP,DOWN,RIGHT,LEFT};

noecho();
timeout(-1);
nonl();
cbreak();
keypad(stdscr,TRUE);
while(loop==TRUE){
c=getch();
switch(c){
case KEY_UP:
case ’u’:
case ’U’: scroll_s(UP);
break;
8.9. CLEAR WINDOW AND LINES 93

case KEY_DOWN:
case ’d’:
case ’D’: scroll_s(DOWN);
break;
case KEY_LEFT:
case ’l’:
case ’L’: scroll_s(LEFT);
break;
case KEY_RIGHT
case ’r’:
case ’R’: scroll_s(RIGHT);
break;
case ’q’:
case ’Q’: loop=FALSE;
default: break;
}
}

For the second, we only need to set echo() and the characters typed by the user will be
printed to the screen. To have the characters printed on the position you want, use move(...)
or wmove(...).
Or, we could open a window with a mask in it (some other colors than those of the
window will do this) and ask the user to input a string:

WINDOW *maskwin;
WINDOW *mainwin;
char *ptr=(char *)malloc(255);
...
mainwin=newwin(3,37,9,21);
maskwin=newwin(1,21,10,35);
...
werase(mainwin);
werase(maskwin);
...
box(mainwin,0,0);
mvwaddstr(mainwin,1,2,"Inputstring: ");
...
wnoutrefresh(mainwin);
wnoutrefresh(maskwin);
doupdate();
...
mvwgetstr(maskwin,0,0,ptr);
...
delwin(maskwin);
delwin(mainwin);
endwin();
free(ptr);

See input.c in the example directory for more explanation.

8.9 Clear Window and Lines

• int erase()
int werase(win)
werase(...) and erase() will copy blanks to every position on the window win or
stdscr. For instance, when you set color attributes to a window and call werase()
the window would be colored. I had some problems with COLOR PAIRS when I
94 CHAPTER 8. CHARACTER CELL GRAPHICS

defined other attributes then black on white so I wrote my own erase function (this is
a low level access to the WINDOW structure):
void NewClear(WINDOW *win)
{
int y,x;

for ( y = 0 ; y <= win -> _maxy ; y++ )

for ( x = 0 ; x <= win -> _maxx ; x++ )
(chtype *) win-> _line[y][x] = ’ ’|win-> _attrs;
win -> _curx = win -> _cury = 0;
touchwin(win);
}

The problem is, that ncurses sometimes makes no use of the window attributes when
blanking the screen. For instance, in lib clrtoeol.c, is BLANK defined as
#define BLANK ’ ’|A_NORMAL

so that the other window attributes get lost while the line is erased.
• int clear()
int wclear(win)
The same as erase(), but will also set clearok() (the screen will be cleared with the
next refresh).
• int clrtobot()
int wclrtobot(win)
Clearing the current cursor line (start is one character right from the cursor) and the
line below the cursor.
• int clrtoeol()
int wclrtoeol(win)
Clear the current line right from the cursor up to its end.

8.10 Updating the Terminal

As written in the overview, ncurses windows are images in memory. This means that any
change to a window is not printed to the physical screen until a refresh is done. This
optimizes the output to the screen because you can do a lot of manipulations and then,
once, call refresh to print it to screen. Otherwise, every change would be printed to the
terminal and decrease the performance of your programs.
• int refresh()
int wrefresh(win)
refresh() copies stdscr to the terminal and wrefresh(win) copies the window image
to stdscr and then makes curscr looks like stdscr.
• int wnoutrefresh(win)
int doupdate()
wnoutrefresh(win) copies the window win to stdscr only. This means that no output
to the terminal is done but the virtual screen stdscr actually looks like the programmer
wanted. doupdate() will do the output to the terminal. A program can change various
windows, call wnoutrefresh(win) for every window and then call doupdate() to
update the physical screen only once.
For instance, we have the following program with two windows. We change both
windows by altering some lines of text. We can write changewin(win) with wre-
fresh(win).
8.11. VIDEO ATTRIBUTES AND COLOR 95

main() changewin(WINDOW *win)

{ {
WINDOW *win1,*win2; ... /* here we change */
... ... /* the lines */
changewin(win1); wrefresh(win);
changewin(win2); return;
... }
}

This will cause ncurses to update the terminal twice and slow down our execution.
With doupdate() we change changewin(win) and our main function and will get
better a performance.

main() changewin(WINDOW *win)

{ {
WINDOW *win1,*win2; ... /* here we change */
... ... /* the lines */
changewin(win1); wnoutrefresh(win);
changewin(win2); return;
doupdate(); }
...
}

• int redrawwin(win)
int wredrawln(win, bline, nlines)
Use these functions when some lines or the entire screen should thrown away before
writing anything new in it (may be when the lines are trashed or so).

• int touchwin(win)
int touchline(win, start, count)
int wtouchln(win, y, n, changed)
int untouchwin(win)
Tells ncurses that the whole window win or the lines from start up to
start+count have been manipulated. For instance, when you have some over-
lapping windows (as in the example type.c) a change to one window will not affect
the image from the other.
wtouchln(...) will touch n lines starting at y. If change is TRUE the lines are
touched, otherwise untouched (changed or unchanged).
untouchwin(win) will mark the window win as unchanged since the last call to
refresh().

• int is linetouched(win, line)

int is wintouched(win)
With these functions you can check if the line line or the window win has been
touched since the last call to refresh().

8.11 Video Attributes and Color

Attributes are special terminal capabilities used when printing characters to the screen.
Characters can be printed bold, underlined, blinking, etc. In ncurses you have the ability
to turn attributes on or off to get better looking output. Possible attributes are listed in the
following table.
Ncurses defines eight colors you can use on a terminal with color support. First, ini-
tialize the color data structures with start color(), then check the terminal capabilities with
96 CHAPTER 8. CHARACTER CELL GRAPHICS

Table 8.4: Ncurses - attributes

Definition Attribute
A ATTRIBUTES mask for attributes (chtype)
A NORMAL normal, reset all other
A STANDOUT best highlighting mode
A UNDERLINE underline
A REVERSE reverse video
A BLINK blinking
A DIM dim or half bright
A BOLD bold or extra bright
A ALTCHARSET use alternate character set
A INVIS invisible
A PROTECT ???
A CHARTEXT mask for actual character (chtype)
A COLOR mask for color
COLOR PAIR(n) set color-pair to that stored in n
PAIR NUMBER(a) get color-pair stored in attribute a

Table 8.5: Ncurses - colors

Definition Color
COLOR BLACK black
COLOR RED red
COLOR GREEN green
COLOR YELLOW yellow
COLOR BLUE blue
COLOR MAGENTA magenta
COLOR CYAN cyan
COLOR WHITE white
8.11. VIDEO ATTRIBUTES AND COLOR 97

has colors(). start color() will initialize COLORS, the maximum colors the terminal sup-
ports, and COLOR PAIR, the maximum number of color pairs you can define.
The attributes can be combined with the OR operator ’|’ so that you can produce bold
blinking output with

A_BOLD|A_BLINK

When you set a window to attribute attr, all characters printed to this window will
get this property until you change the window attribute. It will not get lost when you scroll
or move the window or anything else.
When you write programs for ncurses and BSD curses be careful with colors because
BSD curses has no color support. (Also, old SYS V versions of curses do not have color
support). So you have to use #ifdef operations when you compile for both libraries.

• int attroff(attr)
int wattroff(win, attr)
int attron(attr)
int wattron(win, attr)
Turn on or off the specified attribute attr without reflecting the other attributes in
a window (stdscr or win).
• int attrset(attr)
int wattrset(win, attr)
Set the attribute on stdscr or win to attr.
• int standout()
int standend()
int wstandout(win)
int wstandend(win)
Turn on standout attribute for the window (stdscr or win).
• chtype getattrs(win)
Return the current attributes for window win.
• bool has colors()
Returns TRUE if the terminal has colors. Before you use colors check the terminal
with has colors(), and before this initialize colors with start color()).
• bool can change color()
TRUE if the terminal can redefine colors.
• int start color()
Color initializing. This function has to be called before using colors!
• int init pair(pair, fg, bg)
When you use colors as attributes for windows you have first to define a color pair
with init pair(...). fg is the foreground color and bg the background color for pair.
This is a value from 1 to COLOR P AIRS − 1 (No fault, but 0 is reserved for black
on white). Once defined you can use pair like an attribute. For instance when you
want to have red characters on a blue screen do:

init_pair(1,COLOR_RED,COLOR_BLUE);

Now use wattr(...) to set win to our new color pair:

wattr(win,COLOR_PAIR(1));

Or, combine color pairs with other attributes, such as:

98 CHAPTER 8. CHARACTER CELL GRAPHICS

wattr(win ,A_BOLD|COLOR_PAIR(1));
wattr(win1,A_STANDOUT|COLOR_PAIR(1));

The first will invoke the color pair and set the attribute BOLD and the second will
turn on standout mode, so that you get highlighted red on a blue screen.
• int pair content(pair, f, b)
Will return the foreground and background color from pair.
• int init color(color, r, g, b)
Change the color components r, g and b for color. r, g and b can have values
from 1 to COLORS − 1.
• int color content(color, r, g, b)
Get the color components r, g and b for color.

And how to combine attributes and colors? Some terminals, as the console in Linux,
have colors and some not (xterm, vs100 etc). The following code should solve the problem:

void CheckColor(WINDOW *win1, WINDOW *win2)

{
start_color();
if (has_colors()){
/* fine, we have colors, define color_pairs for foreground
* and background colors
*/
init_pair(1,COLOR_BLUE,COLOR_WHITE);
init_pair(2,COLOR_WHITE,COLOR_RED);
/* now use the defined color_pairs for the windows */
wattrset(win1,COLOR_PAIR(2));
wattrset(win2,COLOR_PAIR(1));
}
else{
/* Ohh, no color (maybe a vt100 or xterm). OK, we’ll
* use black/white attributes instead.
*/
wattrset(win1,A_REVERSE);
wattrset(win2,A_BOLD);
}
return;
}

First, the function CheckColor initializes the colors with start color(), then the function
has colors() will return TRUE if the current terminal has colors. We check this and call
init pair(...) to combine foreground and background colors and wattrset(...) to set these
pairs for the specified window. Alternatively, we can use wattrset(...) alone to set attributes
if we have a black and white terminal.
To get colors in an xterm the best way I found out is to use the ansi xterm with the
patched terminfo entries from the Midnight Commander. Just get the sources of ansi xterm
and Midnight Commander (mc-x.x.tar.gz). Then compile the ansi xterm and use tic with
xterm.ti and vt100.ti from the mc-x.x.tar.gz archive. Execute ansi xterm and test it out.

8.12 Cursor and Window Coordinates

• int move(y, x)
int wmove(win, y, x)
move() moves the cursor from stdscr, wmove(win) the cursor from window win.
8.13. SCROLLING 99

For input/output functions, additional macros are defined which move the cursor
before the specified function is called.
• int curs set(bf)
This will turn the cursor visibility on or off, if the terminal has this capability.
• void getyx(win, y, x)
getyx(...) will return the current cursor position. (Note: this is a macro.)
• void getparyx(win, y, x)
When win is a sub window, getparyx(...) will store the window coordinates relative
to the parent window in y and x. Otherwise y and x are -1. (Note: not implemented
yet.)
• void getbegyx(win, y, x)
void getmaxyx(win, y, x)
int getmaxx(win)
int getmaxy(win)
Store the begin and size coordinates for win in y and x.
• int getsyx(int y, int x)
int setsyx(int y, int x)
Store the virtual screen cursor in y and x or set this cursor. When y and x are -1 and
you call getsyx(...) leaveok will be set.

8.13 Scrolling
• int scrollok(win, bf)
If TRUE, the text in the window win will be scrolled up one line when the cursor is
on the lower right corner and a character is typed (or newline). If FALSE, the cursor
is left in the same position.
When turned on the contents of a window can be scrolled with the following func-
tions. (Note: It would be also scrolled, if you print a new line in the last line of the
window. So, be careful with scrollok(...) or you will get unreasonable results.)
• int scroll(win)
This function will scroll up the window (and the lines in the data structure) one line.
• int scrl(n)
int wscrl(win, n)
These functions will scroll the window stdscr or win up or down depending on
the value of the integer n. If n is positive the window will be scrolled up n lines,
otherwise if n is negative the window will be scrolled down n lines.
• int setscrreg(t, b)
int wsetscrreg(win, t, b)
Set a software scrolling region.

The following code should explain how to get the effect of scrolling a text on the screen.
Look also in type.c in the example directory.
We have a window with 18 lines and 66 columns and want to scroll a text in it. S[] is
a character array with the text. Max s is the number of the last line in s[]. Clear line will
print blank characters from the current cursor position up to the end of the line using the
current attributes from the window (not A NORMAL as clrtoeol does). Beg is the last line
from s[] currently shown on the screen. Scroll is an enumerate to tell the function what to
do, show the NEXT or PREVious line from the text.
100 CHAPTER 8. CHARACTER CELL GRAPHICS

enum{PREV,NEXT)};

void scroll_s(WINDOW *win, int scroll)

{
/* test if we should scroll down and if there is
* anything to scroll down
*/
if((scroll==NEXT)&&(beg<=(max_s-18))){
/* one line down */
beg++;
/* give permissions to scroll */
scrollok(win, TRUE);
/* scroll */
wscrl(win, +1);
/* deny permission to scroll */
scrollok(win, FALSE);
/* set the new string in the last line */
mvwaddnstr(win,17,0,s[beg+17],66);
/* clear the last line from the last character up to end
* of line. Otherwise the attributes will be garbaged.
*/
clear_line(66,win);
}
else if((scroll==PREV)&&(beg>0)){
beg--;
scrollok(win, TRUE);
wscrl(win, -1);
scrollok(win, FALSE);
mvwaddnstr(win,0,0,s[beg],66);
clear_line(66,win);
}
wrefresh(win);
return;
}

8.14 Pads
• WINDOW *newpad(nlines, ncols)

• WINDOW *subpad(orig, nlines, ncols, begy, begx)

• int prefresh(pad, pminrow, pmincol, sminrow, smincol,

smaxrow, smaxcol)

• int pnoutrefresh(pad, pminrow, pmincol, sminrow,

smincol, smaxrow, smaxcol)

• int pechochar(pad, ch)

8.15. SOFT-LABELS 101

8.15 Soft-labels
• int slk init(int fmt)

• int slk set(int labnum, char *label, int fmt)

• int slk refresh()

• int slk noutrefresh()

• char *slk label(int labnum)

• int slk clear()

• int slk restore()

• int slk touch()

• int slk attron(chtype attr)

int slk attrset(chtype attr)
int slk attroff(chtype attr)
These functions correspond to attron(attr), attrset(attr) and attroff(attr). Not im-
plemented yet.

8.16 Miscellaneous
• int beep()

• int flash()

• char *unctrl(chtype c)

• char *keyname(int c)

• int filter()
(Note: not implemented yet.)

• void use env(bf)

• int putwin(WINDOW *win, FILE *filep)

(Note: not implemented yet.)

• WINDOW *getwin(FILE *filep)

(Note: not implemented yet.)
102 CHAPTER 8. CHARACTER CELL GRAPHICS

• int delay output(int ms)

• int flushinp()

8.17 Low-level Access

• int def prog mode()

• int def shell mode()

• int reset prog mode()

• int reset shell mode()

• int resetty()

• int savetty()

• int ripoffline(int line, int (*init)(WINDOW *, int))

• int napms(int ms)

8.18 Screen Dump

• int scr dump(char *filename)
(Note: not implemented yet.)

• int scr restore(char *filename)

(Note: not implemented yet.)

• int scr init(char *filename)

(Note: not implemented yet.)

• int scr set(char *filename)

(Note: not implemented yet.)

8.19 Termcap Emulation

• int tgetent(char *bp, char *name)

• int tgetflag(char id[2])

• int tgetnum(char id[2])

8.20. TERMINFO FUNCTIONS 103

• char *tgetstr(char id[2], char **area)

• char *tgoto(char *cap, int col, int row)

• int tputs(char *str, int affcnt, int (*putc)())

8.20 Terminfo Functions

• int setupterm(char *term, int fildes, int *errret)

• int setterm(char *term)

• int set curterm(TERMINAL *nterm)

• int del curterm(TERMINAL *oterm)

• int restartterm(char *term, int fildes, int *errret)

(Note: not implemented yet.)

• char *tparm(char *str, p1, p2, p3, p4, p5, p6, p7, p8,
p9)
p1 - p9 long int.

• int tputs(char *str, int affcnt, int (*putc)(char))

• int putp(char *str)

• int vidputs(chtype attr, int (*putc)(char))

• int vidattr(chtype attr)

• int mvcur(int oldrow, int oldcol, int newrow, int

newcol)

• int tigetflag(char *capname)

• int tigetnum(char *capname)

• int tigetstr(char *capname)

104 CHAPTER 8. CHARACTER CELL GRAPHICS

8.21 Debug Function

• void init trace()

• void tracef(char *, ...)

• char * traceattr(mode)

• void traceon()

• void traceoff()

8.22 Terminfo Capabilities

8.22.1 Boolean Capabilities
Variable Cap. Int. Description
Name Code
auto left margin bw bw cub1 wraps from column 0 to last column
auto right margin am am Terminal has automatic margins
back color erase bce ut screen erased with background color
can change ccc cc terminal can re-define exiting colors
ceol standout glitch xhp xs Standout not erased by overwriting (hp)
col addr glitch xhpa YA only positive motion for hpa/mhpa caps
cpi changes res cpix YF changing character pitch changes resolution
cr cancels micro mode crxm YB using cr turns off micro mode
eat newline glitch xenl xn newline ignored after 80 cols (Concept)
erase overstrike eo eo Can erase overstrikes with a blank
generic type gn gn Generic line type (e.g.,, dialup, switch).
hard copy hc hc Hardcopy terminal
hard cursor chts HC cursor is hard to see
has meta key km km Has a meta key (shift, sets parity bit)
has print wheel daisy YC printer needs operator to change character set
has status line hs hs Has extra ”status line”
hue lightness saturation hls hl terminal uses only HLS color notation
(Tektronix)
insert null glitch in in Insert mode distinguishes nulls
lpi changes res lpix YG changing line pitch changes resolution
memory above da da Display may be retained above the screen
memory below db db Display may be retained below the screen
move insert mode mir mi Safe to move while in insert mode
move standout mode msgr ms Safe to move in standout modes
needs xon xoff nxon nx padding won’t work, xon/xoff required
no esc ctl c xsb xb Beehive (f1=escape, f2=ctrl C)
non rev rmcup nrrmc NR smcup does not reverse rmcup
no pad char npc NP pad character does not exist
non dest scroll region ndscr ND scrolling region is non-destructive
over strike os os Terminal overstrikes
prtr silent mc5i 5i printer won’t echo on screen
row addr glitch xvpa YD only positive motion for vhp/mvpa caps
semi auto right margin sam YE printing in last column causes cr
status line esc ok eslok es Escape can be used on the status line
dest tabs magic smso xt xt Tabs ruin, magic so char (Teleray 1061)
8.22. TERMINFO CAPABILITIES 105

tilde glitch hz hz Hazel-tine; can not print ˜’s

transparent underline ul ul underline character overstrikes
xon xoff xon xo Terminal uses xon/xoff handshaking

8.22.2 Numbers
Variable Cap. Int. Description
Name Code
bit image entwining bitwin Yo Undocumented in SYSV
buffer capacity bufsz Ya numbers of bytes buffered before printing
columns cols co Number of columns in a line
dot vert spacing spinv Yb spacing of dots horizontally in dots per inch
dot horz spacing spinh Yc spacing of pins vertically in pins per inch
init tabs it it Tabs initially every # spaces
label height lh lh rows in each label
label width lw lw columns in each label
lines lines li Number of lines on screen or page
lines of memory lm lm Lines of memory if ¿ lines. 0 means varies
magic cookie glitch xmc sg Number of blank chars left by smso or rmso
max colors colors Co maximum numbers of colors on screen
max micro address maddr Yd maximum value in micro ... address
max micro jump mjump Ye maximum value in parm ... micro
max pairs pairs pa maximum number of color-pairs on the screen
micro col size mcs Yf Character step size when in micro mode
micro line size mls Yg Line step size when in micro mode
no color video ncv NC video attributes that can’t be used with colors
number of pins npins Yh numbers of pins in print-head
num labels nlab Nl number of labels on screen
output res char orc Yi horizontal resolution in units per line
output res line orl Yj vertical resolution in units per line
output res horz inch orhi Yk horizontal resolution in units per inch
output res vert inch orvi Yl vertical resolution in units per inch
padding baud rate pb pb Lowest baud where cr/nl padding is needed
virtual terminal vt vt Virtual terminal number (UNIX system)
width status line wsl ws No. columns in status line

(The following numeric capabilities are present in the SYSV term structure, but are not yet docu-
mented in the man page. Comments are from the term structure header.)
bit image type bitype Yp Type of bit-image device
buttons btns BT Number of mouse buttons
max attributes ma ma Max combined attributes terminal can handle
maximum windows wnum MW Max number of definable windows
print rate cps Ym Print rate in chars per second
wide char size widcs Yn Char step size in double wide mode

8.22.3 Strings
Variable Cap. Int. Description
Name Code
acs chars acsc ac Graphics charset pairs - def=vt100
alt scancode esc scesa S8 Alternate esc for scancode emulation
(default is vt100)
back tab cbt bt Back tab (P)
bell bel bl Audible signal (bell) (P)
106 CHAPTER 8. CHARACTER CELL GRAPHICS

bit image repeat birep Xy Repeat bit image cell #1 #2 times

(use tparm)
bit image newline binel Zz Move to next row of the bit image
(use tparm)
bit image carriage return bicr Yv Move to beginning of same row
(use tparm)
carriage return cr cr Carriage return (P*)
change char pitch cpi ZA Change # chars per inch
change line pitch lpi ZB Change # lines per inch
change res horz chr ZC Change horizontal resolution
change res vert cvr ZD Change vertical resolution
change scroll region csr cs Change to lines #1 through #2
(vt100) (PG)
char padding rmp rP Like ip but when in insert mode
char set names csnm Zy List of character set names
clear all tabs tbc ct Clear all tab stops (P)
clear margins mgc MC Clear all margins (top, bottom, and sides)
clear screen clear cl Clear screen and home cursor (P*)
clr bol el1 cb Clear to beginning of line
clr eol el ce Clear to end of line (P)
clr eos ed cd Clear to end of display (P*)
code set init csin ci Init sequence for multiple code sets
color names colornm Yw Give name for color #1
column address hpa ch Set cursor column (PG)
command character cmdch CC Term. settable cmd char in prototype
cursor address cup cm Screen rel. cursor motion row #1 col #2
(PG)
cursor down cud1 do Down one line
cursor home home ho Home cursor (if no cup)
cursor invisible civis vi Make cursor invisible
cursor left cub1 le Move cursor left one space
cursor mem address mrcup CM Memory relative cursor addressing
cursor normal cnorm ve Make cursor appear normal (undo vs/vi)
cursor right cuf1 nd Non-destructive space (cursor right)
cursor to ll ll ll Last line, first column (if no cup)
cursor up cuu1 up Upline (cursor up)
cursor visible cvvis vs Make cursor very visible
define bit image region defbi Yx Define rectangular bit image region
(use tparm)
define char defc ZE Define character in a character set
delete character dch1 dc Delete character (P*)
delete line dl1 dl Delete line (P*)
device type devt dv Indicate language/codeset support
dis status line dsl ds Disable status line
display pc char dispc S1 Display PC character
down half line hd hd Half-line down (forward 1/2 linefeed)
ena acs enacs eA enable alternate char set
end bit image region endbi Yy End bit image region (use tparm)
enter alt charset mode smacs as Start alternate character set (P)
enter am mode smam SA turn on automatic margins
enter blink mode blink mb Turn on blinking
enter bold mode bold md Turn on bold (extra bright) mode
enter ca mode smcup ti String to begin programs that use cup
enter delete mode smdc dm Delete mode (enter)
enter dim mode dim mh Turn on half-bright mode
enter doublewide mode swidm ZF Enable double-wide mode
enter draft quality sdrfq ZG Set draft-quality printing
enter insert mode smir im Insert mode (enter);
8.22. TERMINFO CAPABILITIES 107

enter italics mode sitm ZH Enable italics mode

enter leftward mode slm ZI Enable leftward carriage motion
enter micro mode smicm ZJ Enable micro-motion capabilities
enter near letter quality snlq ZK Set NLQ printing
enter normal quality snrmq ZL Set normal quality printing
enter pc charset mode smpch S2 Enter PC character display mode
enter protected mode prot mp Turn on protected mode
enter reverse mode rev mr Turn on reverse video mode
enter scancode mode smsc S4 Enter PC scancode mode
enter secure mode invis mk Turn on blank mode (chars invisible)
enter shadow mode sshm ZM Enable shadow-mode printing
enter standout mode smso so Begin stand out mode
enter subscript mode ssubm ZN Enable subscript printing
enter superscript mode ssupm ZO Enable superscript printing
enter underline mode smul us Start underscore mode
enter upward mode sum ZP Enable upward carriage motion
enter xon mode smxon SX Turn on xon/xoff handshaking
erase chars ech ec Erase #1 characters (PG)
exit alt charset mode rmacs ae End alternate character set (P)
exit am mode rmam RA Turn off automatic margins
exit attribute mode sgr0 me Turn off all attributes
exit ca mode rmcup te String to end programs that use cup
exit delete mode rmdc ed End delete mode
exit doublewide mode rwidm ZQ Disable doublewide printing
exit insert mode rmir ei End insert mode
exit italics mode ritm ZR Disable italic printing
exit leftward mode rlm ZS Enable rightward (normal) carriage
motion
exit micro mode rmicm ZT Disable micro motion capabilities
exit pc charset mode rmpch S3 Disable PC character display
exit scancode mode rmsc S5 Disable PC scancode mode
exit shadow mode rshm ZU Disable shadow printing
exit standout mode rmso se End stand out mode
exit subscript mode rsubm ZV Disable subscript printing
exit superscript mode rsupm ZW Disable superscript printing
exit underline mode rmul ue End underscore mode
exit upward mode rum ZX Enable downward (normal) carriage
motion
exit xon mode rmxon RX turn off xon/xoff handshaking
flash screen flash vb Visible bell (may not move cursor)
form feed ff ff Hardcopy terminal page eject (P*)
from status line fsl fs Return from status line
init 1string is1 i1 Terminal initialization string
init 2string is2 i2 Terminal initialization string
init 3string is3 i3 Terminal initialization string
init file if if Name of file containing is
init prog iprog iP Path name of program for init
initialize color initc Ic Initialize the definition of color
initialize pair initp Ip Initialize color-pair
insert character ich1 ic Insert character (P)
insert line il1 al Add new blank line (P*)
insert padding ip ip Insert pad after character inserted (p*)
key a1 ka1 K1 Upper left of keypad
key a3 ka3 K3 Upper right of keypad
key b2 kb2 K2 Center of keypad
key backspace kbs kb Sent by backspace key
key beg kbeg 1 begin key
key btab kcbt kB back-tab key
108 CHAPTER 8. CHARACTER CELL GRAPHICS

key c1 kc1 K4 Lower left of keypad

key c3 kc3 K5 Lower right of keypad
key cancel kcan 2 cancel key
key catab ktbc ka Sent by clear-all-tabs key
key clear kclr kC Sent by clear screen or erase key
key close kclo 3 close key
key command kcmd 4 command key
key copy kcpy 5 copy key
key create kcrt 6 create key
key ctab kctab kt Sent by clear-tab key
key dc kdch1 kD Sent by delete character key
key dl kdl1 kL Sent by delete line key
key down kcud1 kd Sent by terminal down arrow key
key eic krmir kM Sent by rmir or smir in insert mode
key end kend 7 end key
key enter kent 8 enter/send key
key eol kel kE Sent by clear-to-end-of-line key
key eos ked kS Sent by clear-to-end-of-screen key
key exit kext 9 exit key

key f0 kf0 k0 F00 function key key f32 kf32FM F32 function key
key f1 kf1 k1 F01 function key key f33 kf33FN F33 function key
key f2 kf2 k2 F02 function key key f34 kf34FO F34 function key
key f3 kf3 k3 F03 function key key f35 kf35FP F35 function key
key f4 kf4 k4 F04 function key key f36 kf36FQ F36 function key
key f5 kf5 k5 F05 function key key f37 kf37FR F37 function key
key f6 kf6 k6 F06 function key key f38 kf38FS F38 function key
key f7 kf7 k7 F07 function key key f39 kf39FT F39 function key
key f8 kf8 k8 F08 function key key f40 kf40FU F40 function key
key f9 kf9 k9 F09 function key key f41 kf41FV F41 function key
key f10 kf10k; F10 function key key f42 kf42FW F42 function key
key f11 kf11F1 F11 function key key f43 kf43FX F43 function key
key f12 kf12F2 F12 function key key f44 kf44FY F44 function key
key f13 kf13F3 F13 function key key f45 kf45FZ F45 function key
key f14 kf14F4 F14 function key key f46 kf46Fa F46 function key
key f15 kf15F5 F15 function key key f47 kf47Fb F47 function key
key f16 kf16F6 F16 function key key f48 kf48Fc F48 function key
key f17 kf17F7 F17 function key key f49 kf49Fd F49 function key
key f18 kf18F8 F18 function key key f50 kf50Fe F50 function key
key f19 kf19F9 F19 function key key f51 kf51Ff F51 function key
key f20 kf20FA F20 function key key f52 kf52Fg F52 function key
key f21 kf21FB F21 function key key f53 kf53Fh F53 function key
key f22 kf22FC F22 function key key f54 kf54Fi F54 function key
key f23 kf23FD F23 function key key f55 kf55Fj F55 function key
key f24 kf24FE F24 function key key f56 kf56Fk F56 function key
key f25 kf25FF F25 function key key f57 kf57Fl F57 function key
key f26 kf26FG F26 function key key f58 kf58Fm F58 function key
key f27 kf27FH F27 function key key f59 kf59Fn F59 function key
key f28 kf28FI F28 function key key f60 kf60Fo F60 function key
key f29 kf29FJ F29 function key key f61 kf61Fp F61 function key
key f30 kf30FK F30 function key key f62 kf62Fq F62 function key
key f31 kf31FL F31 function key key f63 kf63Fr F63 function key

key find kfnd 0 find key

key help khlp %1 help key
key home khome kh Sent by home key
key ic kich1 kI Sent by ins char/enter ins mode key
key il kil1 kA Sent by insert line
8.22. TERMINFO CAPABILITIES 109

key left kcub1 kl Sent by terminal left arrow key

key ll kll kH Sent by home-down key
key mark kmrk %2 mark key
key message kmsg %3 message key
key move kmov %4 move key
key next knxt %5 next key
key npage knp kN Sent by next-page key
key open kopn %6 open key
key options kopt %7 options key
key ppage kpp kP Sent by previous-page key
key previous kprv %8 previous key
key print kprt %9 print key
key redo krdo %0 redo key
key reference kref &1 reference key
key refresh krfr &2 refresh key
key replace krpl &3 replace key
key restart krst &4 restart key
key resume kres &5 resume key
key right kcuf1 kr Sent by terminal right arrow key
key save ksav &6 save key
key sbeg kBEG &9 shifted begin key
key scancel kCAN &0 shifted cancel key
key scommand kCMD *1 shifted command key
key scopy kCPY *2 shifted copy key
key screate kCRT *3 shifted create key
key sdc kDC *4 shifted delete char key
key sdl kDL *5 shifted delete line key
key select kslt *6 select key
key send kEND *7 shifted end key
key seol kEOL *8 shifted end of line key
key sexit kEXT *9 shifted exit key
key sf kind kF Sent by scroll-forward/down key
key sfind kFND *0 shifted find key
key shelp kHLP #1 shifted help key
key shome kHOM #2 shifted home key
key sic kIC #3 shifted insert char key
key sleft kLFT #4 shifted left key
key smessage kMSG %a shifted message key
key smove kMOV %b shifted move key
key snext kNXT %c shifted next key
key soptions kOPT %d shifted options key
key sprevious kPRV %e shifted previous key
key sprint kPRT %f shifted print key
key sr kri kR Sent by scroll-backward/up key
key sredo kRDO %g shifted redo key
key sreplace kRPL %h shifted replace key
key sright kRIT %i shifted right key
key srsume kRES %j shifted resume key
key ssave kSAV !1 shifted save key
key ssuspend kSPD !2 shifted suspend key
key stab khts kT Sent by set-tab key
key sundo kUND !3 shifted undo key
key suspend kspd &7 suspend key
key undo kund &8 undo key
key up kcuu1 ku Sent by terminal up arrow key
keypad local rmkx ke Out of ”keypad transmit” mode
keypad xmit smkx ks Put terminal in ”keypad transmit” mode
lab f0 lf0 l0 Labels on function key f0 if not f0
110 CHAPTER 8. CHARACTER CELL GRAPHICS

lab f1 lf1 l1 Labels on function key f1 if not f1

lab f2 lf2 l2 Labels on function key f2 if not f2
lab f3 lf3 l3 Labels on function key f3 if not f3
lab f4 lf4 l4 Labels on function key f4 if not f4
lab f5 lf5 l5 Labels on function key f5 if not f5
lab f6 lf6 l6 Labels on function key f6 if not f6
lab f7 lf7 l7 Labels on function key f7 if not f7
lab f8 lf8 l8 Labels on function key f8 if not f8
lab f9 lf9 l9 Labels on function key f9 if not f9
lab f10 lf10 la Labels on function key f10 if not f10
label on smln LO turn on soft labels
label off rmln LF turn off soft labels
meta off rmm mo Turn off ”meta mode”
meta on smm mm Turn on ”meta mode” (8th bit)
micro column address mhpa ZY Like column address for micro adjustment
micro down mcud1 ZZ Like cursor down for micro adjustment
micro left mcub1 Za Like cursor left for micro adjustment
micro right mcuf1 Zb Like cursor right for micro adjustment
micro row address mvpa Zc Like row address for micro adjustment
micro up mcuu1 Zd Like cursor up for micro adjustment
newline nel nw Newline (behaves like cr followed by lf)
order of pins porder Ze Matches software buts to print-head pins
orig colors oc oc Reset all color pairs
orig pair op op Set default color-pair to original one
pad char pad pc Pad character (rather than null)
parm dch dch DC Delete #1 chars (PG*)
parm delete line dl DL Delete #1 lines (PG*)
parm down cursor cud DO Move cursor down #1 lines (PG*)
parm down micro mcud Zf Like cud for micro adjust
parm ich ich IC Insert #1 blank chars (PG*)
parm index indn SF Scroll forward #1 lines (PG)
parm insert line il AL Add #1 new blank lines (PG*)
parm left cursor cub LE Move cursor left #1 spaces (PG)
parm left micro mcub Zg Like cul for micro adjust
parm right cursor cuf RI Move cursor right #1 spaces (PG*)
parm right micro mcuf Zh Like cuf for micro adjust
parm rindex rin SR Scroll backward #1 lines (PG)
parm up cursor cuu UP Move cursor up #1 lines (PG*)
parm up micro mcuu Zi Like cuu for micro adjust
pkey key pfkey pk Prog funct key #1 to type string #2
pkey local pfloc pl Prog funct key #1 to execute string #2
pkey xmit pfx px Prog funct key #1 to xmit string #2
pkey plab pfxl xl Program key #1 to xmit #2 and show #3
plab norm pln pn program label #1 to show string #2
print screen mc0 ps Print contents of the screen
prtr non mc5p pO Turn on the printer for #1 bytes
prtr off mc4 pf Turn off the printer
prtr on mc5 po Turn on the printer
repeat char rep rp Repeat char #1 #2 times. (PG*)
req for input rfi RF request for input
reset 1string rs1 r1 Reset terminal completely to sane modes.
reset 2string rs2 r2 Reset terminal completely to sane modes.
reset 3string rs3 r3 Reset terminal completely to sane modes.
reset file rf rf Name of file containing reset string
restore cursor rc rc Restore cursor to position of last sc
row address vpa cv Vertical position absolute (set row) (PG)
save cursor sc sc Save cursor position (P)
scancode escape scesc S7 Escape for scancode emulation
8.22. TERMINFO CAPABILITIES 111

scroll forward ind sf Scroll text up (P)

scroll reverse ri sr Scroll text down (P)
select char set scs Zj Select character set
set0 des seq s0ds s0 Shift to codeset 0 (EUC set 0, ASCII)
set1 des seq s1ds s1 Shift to codeset 1
set2 des seq s2ds s2 Shift to codeset 2
set3 des seq s3ds s3 Shift to codeset 3
set a background setab AB Set background color using ANSI escape
set a foreground setaf AF Set foreground color using ANSI escape
set attributes sgr sa Define the video attributes (PG9)
set background setb Sb Set current background color
set bottom margin smgb Zk Set bottom margin at current line
set bottom margin parm smgbp Zl Set bottom line at line #1 or #2
lines from bottom
set color band setcolorYz Change to ribbon color #1
set color pair scp sp Set current color pair
set foreground setf Sf Set current foreground color
set left margin smgl ML Set left margin at current line
set left margin parm smglp Zm Set left (right) margin at #1 (#2)
set lr margin smglr ML Set both left and right margins
set page length slines YZ Set page length to #1 lines (use tparm)
set right margin smgr MR Set right margin at current column
set right margin parm smgrp Zn Set right margin at column #1
set tab hts st Set a tab in all rows, current column
set tb margin smgtb MT Sets both top and bottom margins
set top margin smgt Zo Set top margin at current line
set top margin parm smgtp Zp Set top margin at line #1
set window wind wi Current window is lines #1-#2 cols #3-#4
start bit image sbim Zq Start printing bit image graphics
start char set def scsd Zr Start definition of a character set
stop bit image rbim Zs End printing bit image graphics
stop char set def rcsd Zt End definition of character set
subscript characters subcs Zu List of subscriptable chars
superscript characters supcs Zv List of superscriptable chars
tab ht ta Tab to next 8 space hardware tab stop
these cause cr docr Zw These characters cause a CR
to status line tsl ts Go to status line, column #1
underline char uc uc Underscore one char and move past it
up half line hu hu Half-line up (reverse 1/2 linefeed)
xoff character xoffc XF XON character
xon character xonc XN XOFF character

(The following string capabilities are present in the SYSVr term structure, but are not documented in
the man page. Comments are from the term structure header.)

label format fln Lf ??

set clock sclk SC Set time-of-day clock
display clock dclk DK Display time-of-day clock
remove clock rmclk RC Remove time-of-day clock??
create window cwin CW Define win #1 to go from #2,#3 to #4,#5
goto window wingo WG Goto window #1
hangup hup HU Hang up phone
dial phone dial DI Dial phone number #1
quick dial qdial QD Dial phone number #1, without
progress detection
tone tone TO Select touch tone dialing
pulse pulse PU Select pulse dialing
flash hook hook fh Flash the switch hook
fixed pause pause PA Pause for 2-3 seconds
112 CHAPTER 8. CHARACTER CELL GRAPHICS

wait tone wait WA Wait for dial tone

user0 u0 u0 User string # 0
user1 u1 u1 User string # 1
user2 u2 u2 User string # 2
user3 u3 u3 User string # 3
user4 u4 u4 User string # 4
user5 u5 u5 User string # 5
user6 u6 u6 User string # 6
user7 u7 u7 User string # 7
user8 u8 u8 User string # 8
user9 u9 u9 User string # 9
get mouse getm Gm Curses should get button events
key mouse kmous Km ??
mouse info minfo Mi Mouse status information
pc term options pctrm S6 PC terminal options
req mouse pos reqmp RQ Request mouse position report
zero motion zerom Zx No motion for the subsequent character

8.23 [N]Curses Function Overview

In the following text you will find an overview over the different (n)curses packages. In the
first column is the bsd-curses (as it is in slackware 2.1.0 and in Sun-OS 4.x), in the second
is the sysv-curses (in Sun-OS 5.4 / Solaris 2) and in the third is the ncurses (version 1.8.6).
In the fourth column is a reference to the page in the text where the function is described
(if it is actually described).

x package has this function

n function not yet implemented

Function BSD SYSV Nc. Page def shell mode() x x 102

init trace() x 104 del curterm(...) x x 103
traceattr(mode) x 104 delay output(ms) x x 102
tracef(char *, ...) x 104 delch() x x x 86
addbytes(...) x deleteln() x x x 87
addch(ch) x x x 85 delscreen(...) x x,n 82
addchnstr(...) x x 85 delwin(win) x x x 84
addchstr(chstr) x x 85 derwin(...) x x 84
addnstr(...) x x 85 doupdate() x x 94
addnwstr(...) x drainio(int) x
addstr(str) x x x 85 dupwin(win) x x 84
addwch(...) x echo() x x x 90
addwchnstr(...) x echochar(ch) x x 85
addwchstr(...) x echowchar(ch) x
addwstr(...) x endwin() x x x 82
adjcurspos() x erase() x x x 93
attroff(attr) x x 97 erasechar() x x x 91
attron(attr) x x 97 filter() x x 101
attrset(attr) x x 97 flash() x x 101
baudrate() x x x 91 flushinp() x x 102
beep() x x 101 flushok(...) x
bkgd(ch) x x 88 garbagedlines(...) x
bkgdset(ch) x x 88 garbagedwin(win) x
border(...) x x 87 getattrs(win) x x 97
box(...) x x x 87 getbegyx(...) x x 99
can change color() x x 97 getbkgd(win) x
cbreak() x x x 90 getbmap() x
clear() x x x 94 getcap(str) x
clearok(...) x x x 89 getch() x x x 88
clrtobot() x x x 94 getmaxx(win) x x 99
clrtoeol() x x x 94 getmaxy(win) x x 99
color content(...) x x 98 getmaxyx(...) x x 99
copywin(...) x x 85 getmouse() x
crmode() x x x 90 getnwstr(...) x
curs set(bf) x x 99 getparyx(...) x x 99
curserr() x getstr(str) x x x 88
def prog mode() x x 102 getsyx(...) x x 99
8.23. [N]CURSES FUNCTION OVERVIEW 113

gettmode() x x mvinsstr(...) x x 86
getwch(...) x mvinstr(...) x x,n 89
getwin(...) x mvinswch(...) x
getwin(FILE *) x x,n 101 mvinswstr(...) x
getwstr(...) x mvinwch(...) x
getyx(...) x x x 99 mvinwchnstr(...) x
halfdelay(t) x x 91 mvinwchstr(...) x
has colors() x x 97 mvinwstr(...) x
has ic() x x,n 91 mvprintw(...) x x x 86
has il() x x,n 91 mvscanw(...) x x x 89
hline(...) x x 88 mvvline(...) x
idcok(...) x x,n 89 mvwaddbytes(...) x
idlok(...) x x x 89 mvwaddch(...) x x x 85
immedok(...) x x 89 mvwaddchnstr(...) x x 85
inch() x x x 88 mvwaddchstr(...) x x 85
inchnstr(...) x x,n 89 mvwaddnstr(...) x x 85
inchstr(...) x x,n 89 mvwaddnwstr(...) x
init color(...) x x 98 mvwaddstr(...) x x x 85
init pair(...) x x 97 mvwaddwch(...) x
initscr() x x x 82 mvwaddwchnstr(...) x
innstr(...) x x,n 89 mvwaddwchstr(...) x
innwstr(...) x mvwaddwstr(...) x
insch(c) x x x 86 mvwdelch(...) x x x 86
insdelln(n) x x 86 mvwgetch(...) x x x 88
insertln() x x x 86 mvwgetnwstr(...) x
insnstr(...) x x 86 mvwgetstr(...) x x x 88
insstr(str) x x 86 mvwgetwch(...) x
instr(str) x x,n 89 mvwgetwstr(...) x
inswch(...) x mvwhline(...) x
inswstr(...) x mvwin(...) x x x 84
intrflush(...) x x 91 mvwinch(...) x x x 88
inwch(...) x mvwinchnstr(...) x x,n 89
inwchnstr(...) x mvwinchstr(...) x x,n 89
inwchstr(...) x mvwinnstr(...) x x,n 89
inwchstr(...) x mvwinnwstr(...) x
inwstr(...) x mvwinsch(...) x x x 86
is linetouched(...) x x 95 mvwinsnstr(...) x x 86
is wintouched(win) x x 95 mvwinsstr(...) x x 86
isendwin() x x 82 mvwinstr(...) x x,n 89
keyname(c) x x 101 mvwinswch(...) x
keypad(...) x x 90 mvwinswstr(...) x
killchar() x x x 91 mvwinwch(...) x
leaveok(...) x x x 90 mvwinwchnstr(...) x
longname() x x x 91 mvwinwchstr(...) x
map button(long) x mvwinwstr(...) x
meta(...) x x 90 mvwprintw(...) x x x 86
mouse off(long) x mvwscanw(...) x x x 89
mouse on(long) x mvwvline(...) x
mouse set(long) x napms(ms) x x 102
move(...) x x x 98 newkey(...) x
movenextch() x newpad(...) x x 100
moveprevch() x newscreen(...) x
mvaddbytes(...) x newterm(...) x x 82
mvaddch(...) x x x 85 newwin(...) x x x 82
mvaddchnstr(...) x x 85 nl() x x x 90
mvaddchstr(...) x x 85 nocbreak() x x x 90
mvaddnstr(...) x x 85 nocrmode() x x x 90
mvaddnwstr(...) x nodelay(...) x x 91
mvaddstr(...) x x x 85 noecho() x x x 90
mvaddwch(...) x nonl() x x x 90
mvaddwchnstr(...) x noqiflush() x x,n 91
mvaddwchstr(...) x noraw() x x x 90
mvaddwstr(...) x notimeout(...) x x 91
mvcur(...) x x x 103 overlay(...) x x x 85
mvdelch(...) x x x 86 overwrite(...) x x x 85
mvderwin(...) x x,n 84 pair content(...) x x 98
mvgetch(...) x x x 88 pechochar(...) x x 100
mvgetnwstr(...) x pechowchar(...) x
mvgetstr(...) x x x 88 pnoutrefresh(...) x x 100
mvgetwch(...) x prefresh(...) x x 100
mvgetwstr(...) x printw(...) x x x 86
mvhline(...) x putp(char *) x x 103
mvinch(...) x x x 88 putwin(...) x x,n 101
mvinchnstr(...) x x,n 89 qiflush() x x,n 91
mvinchstr(...) x x,n 89 raw() x x x 90
mvinnstr(...) x x,n 89 redrawwin(win) x x 95
mvinnwstr(...) x refresh() x x x 94
mvinsch(...) x x x 86 request mouse pos() x
mvinsnstr(...) x x 86 reset prog mode() x x 102
mvinsnwstr(...) x reset shell mode() x x 102
114 CHAPTER 8. CHARACTER CELL GRAPHICS

resetty() x x x 102 waddnwstr(...) x

restartterm(...) x x,n 103 waddstr(...) x x x 85
ripoffline(...) x x 102 waddwch(...) x
savetty() x x x 102 waddwchnstr(...) x
scanw(...) x x x 89 waddwchstr(...) x
scr dump(char *) x x,n 102 waddwstr(...) x
scr init(char *) x x,n 102 wadjcurspos(win) x
scr restore(char *) x x,n 102 wattroff(...) x x 97
scr set(char *) x x,n 102 wattron(...) x x 97
scrl(n) x x 99 wattrset(...) x x 97
scroll(win) x x x 99 wbkgd(...) x x 88
scrollok(...) x x x 99 wbkgdset(...) x x 88
set curterm(...) x x 103 wborder(...) x x 87
set term(...) x x 82 wclear(win) x x x 94
setcurscreen(SCREEN *) x wclrtobot(win) x x x 94
setscrreg(...) x x 99 wclrtoeol(win) x x x 94
setsyx(...) x x 99 wcursyncup(win) x x,n 84
setterm(char *) x x x 103 wdelch(win) x x x 86
setupterm(...) x x 103 wdeleteln(win) x x x 87
slk attroff(attr) x x,n 101 wechochar(...) x x 85
slk attron(attr) x x,n 101 wechowchar(...) x
slk attrset(attr) x x,n 101 werase(win) x x x 93
slk clear() x x 101 wgetch(win) x x x 88
slk init(fmt) x x 101 wgetnstr(...) x x 88
slk label(labnum) x x 101 wgetnwstr(...) x
slk noutrefresh() x x 101 wgetstr(...) x x x 88
slk refresh() x x 101 wgetwch(...) x
slk restore() x x 101 wgetwstr(...) x
slk set(...) x x 101 whline() x
slk touch() x x 101 whline(...) x
standend() x x x 97 whline(...) x x 88
standout() x x x 97 winch(win) x x x 88
start color() x x 97 winchnstr(...) x x,n 89
subpad(...) x x 100 winchstr(...) x x,n 89
subwin(...) x x x 84 winnstr(...) x x,n 89
syncok(...) x x,n 84 winnwstr(...) x
termattrs() x x,n 91 winsch(...) x x x 86
termname() x x,n 91 winsdelln(...) x x x 86
tgetent(...) x x 102 winsertln(win) x x 86
tgetflag(char [2]) x x 102 winsnstr(...) x x 86
tgetnum(char [2]) x x 102 winsnwstr(...) x
tgetstr(...) x x 103 winsstr(...) x x 86
tgoto(...) x x 103 winstr(...) x x,n 89
tigetflag(...) x x 103 winswch(...) x
tigetnum(...) x x 103 winswstr(...) x
tigetstr(...) x x 103 winwch(...) x
timeout(t) x x 91 winwchnstr(...) x
touchline(...) x x x 95 winwchstr(...) x
touchwin(win) x x x 95 winwstr(...) x
tparm(...) x x 103 wmouse position(...) x
tputs(...) x 103 wmove(...) x x x 98
traceoff() x x 104 wmovenextch(win) x
traceon() x x 104 wmoveprevch(win) x
typeahead(fd) x x 91 wnoutrefresh(win) x x 94
unctrl(chtype c) x x 101 wprintw(...) x x x 86
ungetch(ch) x x 88 wredrawln(...) x x 95
ungetwch(c) x wrefresh(win) x x x 94
untouchwin(win) x x 95 wscanw(...) x x x 89
use env(bf) x x 101 wscrl(...) x x 99
vidattr(...) x x 103 wsetscrreg(...) x x 99
vidputs(...) x x 103 wstandend(win) x x x 97
vidupdate(...) x wstandout(win) x x x 97
vline(...) x x 87 wsyncdown(win) x x,n 84
vwprintw(...) x x 86 wsyncup(win) x x,n 84
vwscanw(...) x x 89 wtimeout(...) x x 91
waddbytes(...) x wtouchln(...) x x 95
waddch(...) x x x 85 wvline() x
waddchnstr(...) x x 85 wvline(...) x
waddchstr(...) x x 85 wvline(...) x x 87
waddnstr(...) x x 85

To be continued...
Sven Goldt The Linux Programmer’s Guide
Chapter 9

Programming I/O ports

Usually a PC at least has 2 serial and 1 parallel interfaces. These interfaces are special
devices and are mapped as follows:
• /dev/ttyS0 − /dev/ttySn
these are the RS232 serial devices 0-n where n depends on your hardware.
• /dev/cua0 − /dev/cuan
these are the RS232 serial devices 0-n where n depends on your hardware.
• /dev/lp0 − /dev/lpn
these are the parallel devices 0-n where n depends on your hardware.
• /dev/js0 − /dev/jsn
these are the joystick devices 0-n where 0 <= n <= 1.
The difference between the /dev/ttyS∗ and /dev/cua∗ devices is how a call to open()
is handled. The /dev/cua∗ devices are supposed to be used as callout devices and thus
get other default settings by a call to open() than the /dev/ttyS∗ devices which will be
initalized for incoming and outgoing calls. By default the devices are controlling devices
for the process that opened them. Normally ioctl() requests should handle all these special
devices, but POSIX preferred to define new functions to handle asynchronous terminals
heavily depending on the struct termios. Both methods require including < termios.h >.

1. method ioctl:
TCSBRK, TCSBRKP, TCGETA (get attributes), TCSETA (set attributes)
Terminal I/O control (TIOC) requests:
TIOCGSOFTCAR (set soft carrier), TIOCSSOFTCAR (get soft carrier), TIOC-
SCTTY (set controlling tty), TIOCMGET (get modemlines), TIOCMSET (set
modemlines), TIOCGSERIAL, TIOCSSERIAL, TIOCSERCONFIG, TIOCSERG-
WILD, TIOCSERSWILD, TIOCSERGSTRUCT, TIOCMBIS, TIOCMBIC, ...

2. method POSIX:
tcgetattr(), tcsetattr(), tcsendbreak(), tcdrain(), tcflush(), tcflow(), tcgetpgrp(),
tcsetpgrp()
cfsetispeed(), cfgetispeed(), cfsetospeed(), cfgetospeed()

3. other methods:
outb,inb for hardware near programming like using the printer port not for a printer.

115
116 CHAPTER 9. PROGRAMMING I/O PORTS

9.1 Mouse Programming

A mouse is either connected to a serial port or directly to the AT bus and different types
of mouse send distinct kinds of data, which makes mouse programming a bit harder. But,
Andrew Haylett was so kind as to put a generous copyright on his selection program which
means you can use his mouse routines for your own programs. Included in this guide you
can find the pre-release of selection-1.8 with the COPYRIGHT notice. X11 already offers
a comfortable mouse API, so Andrew’s routines should be used for non-X11 applications
only.
You only need the modules mouse.c and mouse.h from the selection package. To get
mouse events you basically have to call ms init() and get ms event(). ms init needs the
following 10 arguments:

1. int acceleration
is the acceleration factor. If you move the mouse more than delta pixels, motion
becomes faster depending on this value.

2. int baud
is the bps rate your mouse uses (normally 1200).

3. int delta
this is the number of pixels that you have to move the mouse before the acceleration
starts.

4. char *device
is the name of your mouse device (e.g. /dev/mouse).

5. int toggle
toggle the DTR, RTS or both DTR and RTS mouse modem lines on initialization
(normally 0).

6. int sample
the resolution (dpi) of your mouse (normally 100).

7. mouse type mouse

the identifier of the connected mouse like P MSC (Mouse Systems Corp.) for my
mouse ;).

8. int slack
amount of slack for wraparound which means if slack is -1 a try to move the mouse
over the screen border will leave the mouse at the border. Values >= 0 mean that the
mouse cursor will wrap to the other end after moving the mouse slack pixels against
the border.

9. int maxx
the resolution of your current terminal in x direction. With the default font, a char is
10 pixels wide, and therefore the overall x screen resolution is 10*80-1.

10. int maxy

the resolution of your current terminal in y direction. With the default font, a char is
12 pixels high and therefore the overall y screen resolution is 12*25-1.

get ms event() just needs a pointer to a struct ms event. If get ms event() returns -1, an
error occured. On success, it returns 0, and the struct ms event will contain the actual
mouse state.
9.2. MODEM PROGRAMMING 117

9.2 Modem Programming

See example miniterm.c
Use termios to control rs232 port.
Use Hayes Commands to control modem.

9.3 Printer Programming

See example checklp.c
Don’t use termios to control printer port. Use ioctl and inb/outb if necessary.
Use Epson, Postscript, PCL, etc. commands to control printer.
< linux/lp.h >
ioctl calls: LPCHAR, LPTIME, LPABORT, LPSETIRQ, LPGETIRQ, LPWAIT
inb/outb for status and control port.

9.4 Joystick Programming

See example js.c in the joystick loadable kernel module package.
< linux/joystick.h >
ioctl calls: JS SET CAL, JS GET CAL, JS SET TIMEOUT, JS GET TIMEOUT,
JS SET TIMELIMIT, JS GET TIMELIMIT, JS GET ALL, JS SET ALL. A read oper-
ation on /dev/jsn will return the struct JS DATA TYPE.
118 CHAPTER 9. PROGRAMMING I/O PORTS
Chapter 10

Porting Applications to Linux

Matt Welsh
[email protected] 26 January 1995

10.1 Introduction
Porting UNIX applications to the Linux operating system is remarkably easy. Linux, and
the GNU C library used by it, have been designed with applications portability in mind,
meaning that many applications will compile simply by issuing make. Those which don’t
generally use some obscure feature of a particular implementation, or rely strongly on
undocumented or undefined behavior of, say, a particular system call.
Linux is mostly compliant with IEEE Std 1003.1-1988 (POSIX.1), but has not actually
been certified as such. Similarly, Linux also implements many features found in the SVID
and BSD strains of UNIX, but again does not necessarily adhere to them in all cases. In
general, Linux has been designed to be compatible with other UNIX implementations,
to make applications porting easier, and in a number of instances has improved upon or
corrected behavior found in those implementations.
As an example, the timeout argument passed to the select system call is actually decre-
mented during the poll operation by Linux. Other implementations don’t modify this value
at all, and applications which aren’t expecting this could break when compiled under Linux.
The BSD and SunOS man pages for select warn that in a “future implementation”, the sys-
tem call may modify the timeout pointer. Unfortunately, many applications still assume
that the value will be untouched.
The goal of this paper is to provide an overview of the major issues associated with port-
ing applications to Linux, highlighting the differences between Linux, POSIX.1, SVID, and
BSD in the following areas: signal handling, terminal I/O, process control and information
gathering, and portable conditional compilation.

10.2 Signal handling

Over the years, the definition and semantics of signals have been modified in various ways
by different implementations of UNIX. Today, there are two major classes of symbols: un-
reliable and reliable. Unreliable signals are those for which the signal handler does not
remain installed once called. These “one-shot” signals must re-install the signal handler
within the signal handler itself, if the program wishes the signal to remain installed. Be-
cause of this, there is a race condition in which the signal can arrive again before the handler
is re-installed, which can cause the signal to either be lost or for the original behavior of the
signal to be triggered (such as killing the process). Therefore, these signals are “unreliable”
because the signal catching and handler re-installation operations are nonatomic.

119
120 CHAPTER 10. PORTING APPLICATIONS TO LINUX

Under unreliable signal semantics, system calls are not restarted automatically when
interrupted by a signal. Therefore, in order for a program to account for all cases, the
program would need to check the value of errno after every system call, and reissue the
system call if its value is EINTR.
Along similar lines, unreliable signal semantics don’t provide an easy way to get an
atomic pause operation (put the process to sleep until a signal arrives). Because of the
unreliable nature of reinstalling signal handlers, there are cases in which a signal can arrive
without the program realizing this.
Under reliable signal semantics, on the other hand, the signal handler remains installed
when called, and the race condition for reinstallation is avoided. Also, certain system calls
can be restarted, and an atomic pause operation is available via the POSIX sigsuspend
function.

10.2.1 Signals under SVR4, BSD, and POSIX.1

The SVR4 implementation of signals incorporates the functions signal, sigset, sighold,
sigrelse, sigignore, and sigpause. The signal function under SVR4 is identical to the clas-
sic UNIX V7 signals, providing only unreliable signals. The other functions do provide
signals with automatic reinstallation of the signal handler, but no system call restarting is
supported.
Under BSD, the functions signal, sigvec, sigblock, sigsetmask, and sigpause are sup-
ported. All of the functions provide reliable signals with system call restarting by default,
but that behavior can be disabled if the programmer wishes.
Under POSIX.1, sigaction, sigprocmask, sigpending, and sigsuspend are provided.
Note that there is no signal function, and according to POSIX.1 it is depreciated. These
functions provide reliable signals, but system call restart behavior is not defined by POSIX.
If sigaction is used under SVR4 and BSD, system call restarting is disabled by default, but
it can be turned on if the signal flag SA RESTART is specified.
Therefore, the “best” way to use signals in a program is to use sigaction, which allows
you to explicitly specify the behavior of signal handlers. However, signal is still used
in many applications, and as we can see above signal provides different semantics under
SVR4 and BSD.

10.2.2 Linux signal options

The following values for the sa flags member of the sigaction structure are defined
for Linux.
• SA NOCLDSTOP: Don’t send SIGCHLD when a child process is stopped.
• SA RESTART: Force restart of certain system calls when interrupted by a signal
handler.
• SA NOMASK: Disable signal mask (which blocks signals during execution of a signal
handler).
• SA ONESHOT: Clear signal handler after execution. Note that SVR4 uses
SA RESETHAND to mean the same thing.
• SA INTERRUPT: Defined under Linux, but unused. Under SunOS, system calls
were automatically restarted, and this flag disabled that behavior.
• SA STACK: Currently a no-op, to be used for signal stacks.
Note that POSIX.1 defines only SA NOCLDSTOP, and there are several other options
defined by SVR4 not available under Linux. When porting applications which use sigac-
tion, you may have to modify the values of sa flags to get the appropriate behavior.
10.3. TERMINAL I/O 121

10.2.3 signal under Linux

Under Linux, the signal function is equivalent to using sigaction with the SA ONESHOT
and SA NOMASK options; that is, it corresponds to the classic, unreliable signal semantics
as used under SVR4.
If you wish signal to use BSD semantics, most Linux systems provide a BSD compati-
bility library which can be linked with. To use this library, you could add the options

-I/usr/include/bsd -lbsd

to the compilation command line. When porting applications using signal, pay close atten-
tion to what assumptions the program makes about use of signal handlers, and modify the
code (or compile with the appropriate definitions) to get the right behavior.

10.2.4 Signals supported by Linux

Linux supports nearly every signal provided by SVR4, BSD, and POSIX, with few excep-
tions:

• SIGEMT is not supported; it corresponds to a hardware fault under SVR4 and BSD.

• SIGINFO is not supported; it is used for keyboard information requests under SVR4.

• SIGSYS is not supported; it refers to an invalid system call in SVR4 and BSD. If
you link with libbsd, this signal is redefined to SIGUNUSED.

• SIGABRT and SIGIOT are identical.

• SIGIO, SIGPOLL, and SIGURG are identical.

• SIGBUS is defined as SIGUNUSED. Technically there is no “bus error” in the Linux

environment.

10.3 Terminal I/O

As with signals, terminal I/O control has three different implementations under SVR4,
BSD, and POSIX.1.
SVR4 uses the termio structure, and various ioctl calls (such as TCSETA, TCGETA,
and so forth) on a terminal device to obtain and set parameters with the termio structure.
This structure looks like:

struct termio {
unsigned short c_iflag; /* Input modes */
unsigned short c_oflag; /* Output modes */
unsigned short c_cflag; /* Control modes */
unsigned short c_lflag; /* Line discipline modes */
char c_line; /* Line discipline */
unsigned char c_cc[NCC]; /* Control characters */
};

Under BSD, the sgtty structure is used with various ioctl calls, such as TIOCGETP,
TIOCSETP, and so forth.
Under POSIX, the termios struct is used, along with various functions defined by
POSIX.1, such as tcsetattr and tcgetattr. The termios structure is identical to
struct termio used by SVR4, but the types are renamed (such as tcflag t instead
of unsigned short), and NCCS is used for the size of the c cc array.
122 CHAPTER 10. PORTING APPLICATIONS TO LINUX

Under Linux, both POSIX.1 termios and SVR4 termio are supported directly by the
kernel. This means that if your program uses either of these methods for accessing terminal
I/O, it should compile directly under Linux. If you’re ever in doubt, it’s easy to modify code
using termio to use termios, using a bit of knowledge of both methods. Hopefully, this
shouldn’t ever be necessary. But, do pay attention if a program attempts to use the c line
field in the termio structure. For nearly all applications, this should be N TTY, and if the
program assumes that some other line discipline is available you might have trouble.
If your program uses the BSD sgtty implementation, you can link against libbsd.a as
described above. This will provide a replacement for ioctl which will resubmit the terminal
I/O requests in terms of the POSIX termios calls used by the kernel. When compiling
such a program, if symbols such as TIOCGETP are undefined, you will need to link against
libbsd.

10.4 Process information and control

Programs such as ps, top, and free must have some way to obtain information from the
kernel about a processes and system resources. Similarly, debuggers and other like tools
need the ability to control and inspect a running process. These features have been provided
by a number of interfaces by different versions of UNIX, and nearly all of them are either
machine-specific or tied to a particular kernel design. So far, there has been no universally-
accepted interface for this kind of process-kernel interaction.

10.4.1 kvm routines

Many systems use routines such as kvm open, kvm nlist, and kvm read to access kernel
data structures directly via the /dev/kmem device. In general, these programs will open
/dev/kmem, read the kernel’s symbol table, locate data in the running kernel with this table,
and read the appropriate addresses in the kernel address space with these routines. Be-
cause this requires the user program and the kernel to agree upon the size and format of
data structures read in this fashion, such programs often have to be rebuilt for each kernel
revision, CPU type, and so forth.

10.4.2 ptrace and the /proc filesystem

The ptrace system call is used in 4.3BSD and SVID to control a process and read informa-
tion from it. It is classically used by debuggers to, say, trap execution of a running process
or examine its state. Under SVR4, ptrace is superseded by the /proc filesystem, which
appears as a directory containing a single file entry for each running process, named by
process ID. The user program can open the file corresponding to the process of interest and
issue various ioctl calls on it to control its execution or obtain information from the ker-
nel on the process. Similarly, the program can read or write data directly in the process’s
address space through the file descriptor into the /proc filesystem.

10.4.3 Process control under Linux

Under Linux, the ptrace system call is supported for process control, and it works as in
4.3BSD. To obtain process and system information, Linux also provides a /proc filesystem,
but with very different semantics. Under Linux, /proc consists of a number of files pro-
viding general system information, such as memory usage, load average, loaded module
statistics, and network statistics. These files are generally accessed using read and write
and their contents can be parsed using scanf. The /proc filesystem under Linux also pro-
vides a directory entry for each running process, named by process ID, which contains file
entries for information such as the command line, links to the current working directory
10.5. PORTABLE CONDITIONAL COMPILATION 123

and executable file, open file descriptors, and so forth. The kernel provides all of this in-
formation on the fly in response to read requests. This implementation is not unlike the
/proc filesystem found in Plan 9, but it does have its drawbacks—for example, for a tool
such as ps to list a table of information on all running processes, many directories must be
traversed and many files opened and read. By comparison, the kvm routines used on other
UNIX systems read kernel data structures directly with only a few system calls.
Obviously, each implementation is so vastly different that porting applications which
use them can prove to be a real task. It should be pointed out that the SVR4 /proc filesystem
is a very different beast than that found in Linux, and they may not be used in the same
context. Arguably, any program which uses the kvm routines or SVR4 /proc filesystem
is not really portable, and those sections of code should be rewritten for each operating
system.
The Linux ptrace call is nearly identical to that found in BSD, but there are a few
differences:
• The requests PTRACE PEEKUSER and PTRACE POKEUSER under BSD are named
PTRACE PEEKUSR and PTRACE POKEUSR, respectively, under Linux.
• Process registers can be set using the PTRACE POKEUSR request with offsets found
in /usr/include/linux/ptrace.h.
• The SunOS requests PTRACE {READ,WRITE}{TEXT,DATA} are not supported,
nor are PTRACE SETACBKPT, PTRACE SETWRBKPT, PTRACE CLRBKPT, or
PTRACE DUMPCORE. These missing requests should only affect a small number
of existing programs.
Linux does not provide the kvm routines for reading the kernel address space from a
user program, but some programs (most notably kmem ps) implement their own versions
of these routines. In general, these are not portable, and any code which uses the kvm
routines is probably depending upon the availability of certain symbols or data structures
in the kernel—not a safe assumption to make. Use of kvm routines should be considered
architecture-specific.

10.5 Portable conditional compilation

If you need to make modifications to existing code in order to port it to Linux, you may
need to use ifdef. . . endif pairs to surround parts of Linux-specific code—or, for that
matter, code corresponding to other implementations. No real standard for selecting por-
tions of code to be compiled based on the operating system exists, but many programs use
a convention such as defining SVR4 for System V code, BSD for BSD code, and linux
for Linux-specific code.
The GNU C library used by Linux allows you to turn on various features of the library
by defining various macros at compile time. These are:
• STRICT ANSI : For ANSI C features only
• POSIX SOURCE: For POSIX.1 features
• POSIX C SOURCE: If defined as 1, POSIX.1 features; if defined as 2, POSIX.2
features.
• BSD SOURCE: ANSI, POSIX, and BSD features.
• SVID SOURCE: ANSI, POSIX, and System V features.
• GNU SOURCE: ANSI, POSIX, BSD, SVID, and GNU extensions. This is the default
if none of the above are defined.
124 CHAPTER 10. PORTING APPLICATIONS TO LINUX

If you define BSD SOURCE yourself, the additional definition FAVOR BSD will be
defined for the library. This will cause BSD behavior for certain things to be selected over
POSIX or SVR4. For example, if FAVOR BSD is defined, setjmp and longjmp will save
and restore the signal mask, and getpgrp will accept a PID argument. Note that you must
still link against libbsd to get BSD-like behavior for the features mentioned earlier in
this paper.
Under Linux, gcc defines a number of macros automatically which you can use in your
program. These are:
• GNUC (major GNU C version, e.g., 2)

• GNUC MINOR (minor GNU C version, e.g., 5)

• unix

• i386

• linux

• unix

• i386

• linux

• unix

• i386

• linux

Many programs use

#ifdef linux

to surround Linux-specific code. Using these compile-time macros you can easily adapt
existing code to include or exclude changes necessary to port the program to Linux, Note
that because Linux supports more System V-like features in general, the best code base to
start from with a program written for both System V and BSD is probably the System V
version. Alternately, you can start from the BSD base and link against libbsd.

10.6 Additional Comments

1
This chapter covers most of the porting issues except the missing system calls that are
named in the system calls chapter and the yet missing streams (rumors say a loadable
stream module should exist at ftp.uni-stuttgart.de in pub/systems/linux/isdn).

1 Added by Sven Goldt

Chapter 11

Systemcalls in alphabetical order

Sven Goldt The Linux Programmer’s Guide

exit - like exit but with fewer actions (m+c)

accept - accept a connection on a socket (m+c!)
access - check user’s permissions for a file (m+c)
acct - not yet implemented (mc)
adjtimex - set/get kernel time variables (-c)
afs syscall - reserved andrew filesystem call (-)
alarm - send SIGALRM at a specified time (m+c)
bdflush - flush dirty buffers to disk (-c)
bind - name a socket for interprocess communication (m!c)
break - not yet implemented (-)
brk - change data segment size (mc)
chdir - change working directory (m+c)
chmod - change file attributes (m+c)
chown - change ownership of a file (m+c)
chroot - set a new root directory (mc)
clone - see fork (m-)
close - close a file by reference (m+c)
connect - link 2 sockets (m!c)
creat - create a file (m+c)
create module - allocate space for a loadable kernel module (-)
delete module - unload a kernel module (-)
dup - create a file descriptor duplicate (m+c)
dup2 - duplicate a file descriptor (m+c)
execl, execlp, execle, ... - see execve (m+!c)
execve - execute a file (m+c)
exit - terminate a program (m+c)
fchdir - change working directory by reference ()
fchmod - see chmod (mc)
fchown - change ownership of a file (mc)
fclose - close a file by reference (m+!c)
fcntl - file/filedescriptor control (m+c)
flock - change file locking (m!c)
fork - create a child process (m+c)
fpathconf - get info about a file by reference (m+!c)
fread - read array of binary data from stream (m+!c)
fstat - get file status (m+c)
fstatfs - get filesystem status by reference (mc)

125
126 CHAPTER 11. SYSTEMCALLS IN ALPHABETICAL ORDER

fsync - write file cache to disk (mc)

ftime - get timezone+seconds since 1.1.1970 (m!c)
ftruncate - change file size (mc)
fwrite - write array of binary datas to stream (m+!c)
get kernel syms - get kernel symbol table or its size (-)
getdomainname - get system’s domainname (m!c)
getdtablesize - get filedescriptor table size (m!c)
getegid - get effective group id (m+c)
geteuid - get effective user id (m+c)
getgid - get real group id (m+c)
getgroups - get supplemental groups (m+c)
gethostid - get unique host identifier (m!c)
gethostname - get system’s hostname (m!c)
getitimer - get value of interval timer (mc)
getpagesize - get size of a system page (m-!c)
getpeername - get address of a connected peer socket (m!c)
getpgid - get parent group id of a process (+c)
getpgrp - get parent group id of current process (m+c)
getpid - get process id of current process (m+c)
getppid - get process id of the parent process (m+c)
getpriority - get a process/group/user priority (mc)
getrlimit - get resource limits (mc)
getrusage - get usage of resources (m)
getsockname - get the adress of a socket (m!c)
getsockopt - get option settings of a socket (m!c)
gettimeofday - get timezone+seconds since 1.1.1970 (mc)
getuid - get real uid (m+c)
gtty - not yet implemented ()
idle - make a process a candidate for swap (mc)
init module - insert a loadable kernel module (-)
ioctl - manipulate a character device (mc)
ioperm - set some i/o port’s permissions (m-c)
iopl - set all i/o port’s permissions (m-c)
ipc - interprocess communication (-c)
kill - send a signal to a process (m+c)
killpg - send a signal to a process group (mc!)
klog - see syslog (-!)
link - create a hardlink for an existing file (m+c)
listen - listen for socket connections (m!c)
llseek - lseek for large files (-)
lock - not implemented yet ()
lseek - change the position ptr of a file descriptor (m+c)
lstat - get file status (mc)
mkdir - create a directory (m+c)
mknod - create a device (mc)
mmap - map a file into memory (mc)
modify ldt - read or write local descriptor table (-)
mount - mount a filesystem (mc)
mprotect - read, write or execute protect memory (-)
mpx - not implemented yet ()
msgctl - ipc message control (m!c)
msgget - get an ipc message queue id (m!c)
msgrcv - receive an ipc message (m!c)
msgsnd - send an ipc message (m!c)
 127

munmap - unmap a file from memory (mc)

nice - change process priority (mc)
oldfstat - no longer existing
oldlstat - no longer existing
oldolduname - no longer existing
oldstat - no longer existing
olduname - no longer existing
open - open a file (m+c)
pathconf - get information about a file (m+!c)
pause - sleep until signal (m+c)
personality - change current execution domain for ibcs (-)
phys - not implemented yet (m)
pipe - create a pipe (m+c)
prof - not yet implemented ()
profil - execution time profile (m!c)
ptrace - trace a child process (mc)
quotactl - not implemented yet ()
read - read data from a file (m+c)
readv - read datablocks from a file (m!c)
readdir - read a directory (m+c)
readlink - get content of a symbolic link (mc)
reboot - reboot or toggle vulcan death grip (-mc)
recv - receive a message from a connected socket (m!c)
recvfrom - receive a message from a socket (m!c)
rename - move/rename a file (m+c)
rmdir - delete an empty directory (m+c)
sbrk - see brk (mc!)
select - sleep until action on a filedescriptor (mc)
semctl - ipc semaphore control (m!c)
semget - ipc get a semaphore set identifier (m!c)
semop - ipc operation on semaphore set members (m!c)
send - send a message to a connected socket (m!c)
sendto - send a message to a socket (m!c)
setdomainname - set system’s domainname (mc)
setfsgid - set filesystem group id ()
setfsuid - set filesystem user id ()
setgid - set real group id (m+c)
setgroups - set supplemental groups (mc)
sethostid - set unique host identifier (mc)
sethostname - set the system’s hostname (mc)
setitimer - set interval timer (mc)
setpgid - set process group id (m+c)
setpgrp - has no effect (mc!)
setpriority - set a process/group/user priority (mc)
setregid - set real and effective group id (mc)
setreuid - set real and effective user id (mc)
setrlimit - set resource limit (mc)
setsid - create a session (+c)
setsockopt - change options of a socket (mc)
settimeofday - set timezone+seconds since 1.1.1970 (mc)
setuid - set real user id (m+c)
setup - initialize devices and mount root (-)
sgetmask - see siggetmask (m)
shmat - attach shared memory to data segment (m!c)
128 CHAPTER 11. SYSTEMCALLS IN ALPHABETICAL ORDER

shmctl - ipc manipulate shared memory (m!c)

shmdt - detach shared memory from data segment (m!c)
shmget - get/create shared memory segment (m!c)
shutdown - shutdown a socket (m!c)
sigaction - set/get signal handler (m+c)
sigblock - block signals (m!c)
siggetmask - get signal blocking of current process (!c)
signal - setup a signal handler (mc)
sigpause - use a new signal mask until a signal (mc)
sigpending - get pending, but blocked signals (m+c)
sigprocmask - set/get signal blocking of current process (+c)
sigreturn - not yet used ()
sigsetmask - set signal blocking of current process (c!)
sigsuspend - replacement for sigpause (m+c)
sigvec - see sigaction (m!)
socket - create a socket communication endpoint (m!c)
socketcall - socket call multiplexer (-)
socketpair - create 2 connected sockets (m!c)
ssetmask - see sigsetmask (m)
stat - get file status (m+c)
statfs - get filesystem status (mc)
stime - set seconds since 1.1.1970 (mc)
stty - not yet implemented ()
swapoff - stop swapping to a file/device (m-c)
swapon - start swapping to a file/device (m-c)
symlink - create a symbolic link to a file (m+c)
sync - sync memory and disk buffers (mc)
syscall - execute a systemcall by number (-!c)
sysconf - get value of a system variable (m+!c)
sysfs - get infos about configured filesystems ()
sysinfo - get Linux system infos (m-)
syslog - manipulate system logging (m-c)
system - execute a shell command (m!c)
time - get seconds since 1.1.1970 (m+c)
times - get process times (m+c)
truncate - change file size (mc)
ulimit - get/set file limits (c!)
umask - set file creation mask (m+c)
umount - unmount a filesystem (mc)
uname - get system information (m+c)
unlink - remove a file when not busy (m+c)
uselib - use a shared library (m-c)
ustat - not yet implemented (c)
utime - modify inode time entries (m+c)
utimes - see utime (m!c)
vfork - see fork (m!c)
vhangup - virtually hang up current tty (m-c)
vm86 - enter virtual 8086 mode (m-c)
wait - wait for process termination (m+!c)
wait3 - bsd wait for a specified process (m!c)
wait4 - bsd wait for a specified process (mc)
waitpid - wait for a specified process (m+c)
write - write data to a file (m+c)
writev - write datablocks to a file (m!c)
 129

(m) manual page exists.

(+) POSIX compliant.
(-) Linux specific.
(c) in libc.
(!) not a sole system call.uses a different system call.

Sven Goldt The Linux Programmer’s Guide

130 CHAPTER 11. SYSTEMCALLS IN ALPHABETICAL ORDER
Chapter 12

Abbreviations

ANSI American National Standard for Information Systems

API Application Programming Interface
ASCII American Standard Code for Information Interchange
AT 386 Advanced Technology Intel 80386 based PC
FIPS Federal Information Processing Standard
FSF Free Software Foundation
IEEE Institute of Electrical and Electronics Engineers, Inc.
IPC Inter Process Communication
ISO International Organization for Standards
POSIX Portable Operating System Interface for uniX
POSIX.1 IEEE Std. 1003.1-1990 Standard for Information Technology -
Portable Operating System Interface (POSIX) - Part 1:
System Application Programming Interface (API)

131