3 perlipc - Perl interprocess communication (signals, fifos, pipes, safe subprocesses, sockets, and semaphores)
7 The basic IPC facilities of Perl are built out of the good old Unix
8 signals, named pipes, pipe opens, the Berkeley socket routines, and SysV
9 IPC calls. Each is used in slightly different situations.
13 Perl uses a simple signal handling model: the %SIG hash contains names or
14 references of user-installed signal handlers. These handlers will be called
15 with an argument which is the name of the signal that triggered it. A
16 signal may be generated intentionally from a particular keyboard sequence like
17 control-C or control-Z, sent to you from another process, or
18 triggered automatically by the kernel when special events transpire, like
19 a child process exiting, your process running out of stack space, or
20 hitting file size limit.
22 For example, to trap an interrupt signal, set up a handler like this.
23 Do as little as you possibly can in your handler; notice how all we do is
24 set a global variable and then raise an exception. That's because on most
25 systems, libraries are not re-entrant; particularly, memory allocation and
26 I/O routines are not. That means that doing nearly I<anything> in your
27 handler could in theory trigger a memory fault and subsequent core dump.
32 die "Somebody sent me a SIG$signame";
34 $SIG{INT} = 'catch_zap'; # could fail in modules
35 $SIG{INT} = \&catch_zap; # best strategy
37 The names of the signals are the ones listed out by C<kill -l> on your
38 system, or you can retrieve them from the Config module. Set up an
39 @signame list indexed by number to get the name and a %signo table
40 indexed by name to get the number:
43 defined $Config{sig_name} || die "No sigs?";
44 foreach $name (split(' ', $Config{sig_name})) {
50 So to check whether signal 17 and SIGALRM were the same, do just this:
52 print "signal #17 = $signame[17]\n";
54 print "SIGALRM is $signo{ALRM}\n";
57 You may also choose to assign the strings C<'IGNORE'> or C<'DEFAULT'> as
58 the handler, in which case Perl will try to discard the signal or do the
61 On most Unix platforms, the C<CHLD> (sometimes also known as C<CLD>) signal
62 has special behavior with respect to a value of C<'IGNORE'>.
63 Setting C<$SIG{CHLD}> to C<'IGNORE'> on such a platform has the effect of
64 not creating zombie processes when the parent process fails to C<wait()>
65 on its child processes (i.e. child processes are automatically reaped).
66 Calling C<wait()> with C<$SIG{CHLD}> set to C<'IGNORE'> usually returns
67 C<-1> on such platforms.
69 Some signals can be neither trapped nor ignored, such as
70 the KILL and STOP (but not the TSTP) signals. One strategy for
71 temporarily ignoring signals is to use a local() statement, which will be
72 automatically restored once your block is exited. (Remember that local()
73 values are "inherited" by functions called from within that block.)
76 local $SIG{INT} = 'IGNORE';
80 # interrupts still ignored, for now...
83 Sending a signal to a negative process ID means that you send the signal
84 to the entire Unix process-group. This code sends a hang-up signal to all
85 processes in the current process group (and sets $SIG{HUP} to IGNORE so
86 it doesn't kill itself):
89 local $SIG{HUP} = 'IGNORE';
91 # snazzy writing of: kill('HUP', -$$)
94 Another interesting signal to send is signal number zero. This doesn't
95 actually affect another process, but instead checks whether it's alive
96 or has changed its UID.
98 unless (kill 0 => $kid_pid) {
99 warn "something wicked happened to $kid_pid";
102 You might also want to employ anonymous functions for simple signal
105 $SIG{INT} = sub { die "\nOutta here!\n" };
107 But that will be problematic for the more complicated handlers that need
108 to reinstall themselves. Because Perl's signal mechanism is currently
109 based on the signal(3) function from the C library, you may sometimes be so
110 misfortunate as to run on systems where that function is "broken", that
111 is, it behaves in the old unreliable SysV way rather than the newer, more
112 reasonable BSD and POSIX fashion. So you'll see defensive people writing
113 signal handlers like this:
117 # loathe sysV: it makes us not only reinstate
118 # the handler, but place it after the wait
119 $SIG{CHLD} = \&REAPER;
121 $SIG{CHLD} = \&REAPER;
122 # now do something that forks...
124 or even the more elaborate:
126 use POSIX ":sys_wait_h";
129 while ($child = waitpid(-1,WNOHANG)) {
130 $Kid_Status{$child} = $?;
132 $SIG{CHLD} = \&REAPER; # still loathe sysV
134 $SIG{CHLD} = \&REAPER;
135 # do something that forks...
137 Signal handling is also used for timeouts in Unix, While safely
138 protected within an C<eval{}> block, you set a signal handler to trap
139 alarm signals and then schedule to have one delivered to you in some
140 number of seconds. Then try your blocking operation, clearing the alarm
141 when it's done but not before you've exited your C<eval{}> block. If it
142 goes off, you'll use die() to jump out of the block, much as you might
143 using longjmp() or throw() in other languages.
148 local $SIG{ALRM} = sub { die "alarm clock restart" };
150 flock(FH, 2); # blocking write lock
153 if ($@ and $@ !~ /alarm clock restart/) { die }
155 For more complex signal handling, you might see the standard POSIX
156 module. Lamentably, this is almost entirely undocumented, but
157 the F<t/lib/posix.t> file from the Perl source distribution has some
162 A named pipe (often referred to as a FIFO) is an old Unix IPC
163 mechanism for processes communicating on the same machine. It works
164 just like a regular, connected anonymous pipes, except that the
165 processes rendezvous using a filename and don't have to be related.
167 To create a named pipe, use the Unix command mknod(1) or on some
168 systems, mkfifo(1). These may not be in your normal path.
170 # system return val is backwards, so && not ||
172 $ENV{PATH} .= ":/etc:/usr/etc";
173 if ( system('mknod', $path, 'p')
174 && system('mkfifo', $path) )
176 die "mk{nod,fifo} $path failed";
180 A fifo is convenient when you want to connect a process to an unrelated
181 one. When you open a fifo, the program will block until there's something
184 For example, let's say you'd like to have your F<.signature> file be a
185 named pipe that has a Perl program on the other end. Now every time any
186 program (like a mailer, news reader, finger program, etc.) tries to read
187 from that file, the reading program will block and your program will
188 supply the new signature. We'll use the pipe-checking file test B<-p>
189 to find out whether anyone (or anything) has accidentally removed our fifo.
192 $FIFO = '.signature';
193 $ENV{PATH} .= ":/etc:/usr/games";
198 system('mknod', $FIFO, 'p')
199 && die "can't mknod $FIFO: $!";
202 # next line blocks until there's a reader
203 open (FIFO, "> $FIFO") || die "can't write $FIFO: $!";
204 print FIFO "John Smith (smith\@host.org)\n", `fortune -s`;
206 sleep 2; # to avoid dup signals
211 By installing Perl code to deal with signals, you're exposing yourself
212 to danger from two things. First, few system library functions are
213 re-entrant. If the signal interrupts while Perl is executing one function
214 (like malloc(3) or printf(3)), and your signal handler then calls the
215 same function again, you could get unpredictable behavior--often, a
216 core dump. Second, Perl isn't itself re-entrant at the lowest levels.
217 If the signal interrupts Perl while Perl is changing its own internal
218 data structures, similarly unpredictable behaviour may result.
220 There are two things you can do, knowing this: be paranoid or be
221 pragmatic. The paranoid approach is to do as little as possible in your
222 signal handler. Set an existing integer variable that already has a
223 value, and return. This doesn't help you if you're in a slow system call,
224 which will just restart. That means you have to C<die> to longjump(3) out
225 of the handler. Even this is a little cavalier for the true paranoiac,
226 who avoids C<die> in a handler because the system I<is> out to get you.
227 The pragmatic approach is to say ``I know the risks, but prefer the
228 convenience'', and to do anything you want in your signal handler,
229 prepared to clean up core dumps now and again.
231 To forbid signal handlers altogether would bars you from
232 many interesting programs, including virtually everything in this manpage,
233 since you could no longer even write SIGCHLD handlers. Their dodginess
234 is expected to be addresses in the 5.005 release.
237 =head1 Using open() for IPC
239 Perl's basic open() statement can also be used for unidirectional interprocess
240 communication by either appending or prepending a pipe symbol to the second
241 argument to open(). Here's how to start something up in a child process you
244 open(SPOOLER, "| cat -v | lpr -h 2>/dev/null")
245 || die "can't fork: $!";
246 local $SIG{PIPE} = sub { die "spooler pipe broke" };
247 print SPOOLER "stuff\n";
248 close SPOOLER || die "bad spool: $! $?";
250 And here's how to start up a child process you intend to read from:
252 open(STATUS, "netstat -an 2>&1 |")
253 || die "can't fork: $!";
255 next if /^(tcp|udp)/;
258 close STATUS || die "bad netstat: $! $?";
260 If one can be sure that a particular program is a Perl script that is
261 expecting filenames in @ARGV, the clever programmer can write something
264 % program f1 "cmd1|" - f2 "cmd2|" f3 < tmpfile
266 and irrespective of which shell it's called from, the Perl program will
267 read from the file F<f1>, the process F<cmd1>, standard input (F<tmpfile>
268 in this case), the F<f2> file, the F<cmd2> command, and finally the F<f3>
269 file. Pretty nifty, eh?
271 You might notice that you could use backticks for much the
272 same effect as opening a pipe for reading:
274 print grep { !/^(tcp|udp)/ } `netstat -an 2>&1`;
275 die "bad netstat" if $?;
277 While this is true on the surface, it's much more efficient to process the
278 file one line or record at a time because then you don't have to read the
279 whole thing into memory at once. It also gives you finer control of the
280 whole process, letting you to kill off the child process early if you'd
283 Be careful to check both the open() and the close() return values. If
284 you're I<writing> to a pipe, you should also trap SIGPIPE. Otherwise,
285 think of what happens when you start up a pipe to a command that doesn't
286 exist: the open() will in all likelihood succeed (it only reflects the
287 fork()'s success), but then your output will fail--spectacularly. Perl
288 can't know whether the command worked because your command is actually
289 running in a separate process whose exec() might have failed. Therefore,
290 while readers of bogus commands return just a quick end of file, writers
291 to bogus command will trigger a signal they'd better be prepared to
294 open(FH, "|bogus") or die "can't fork: $!";
295 print FH "bang\n" or die "can't write: $!";
296 close FH or die "can't close: $!";
298 That won't blow up until the close, and it will blow up with a SIGPIPE.
299 To catch it, you could use this:
301 $SIG{PIPE} = 'IGNORE';
302 open(FH, "|bogus") or die "can't fork: $!";
303 print FH "bang\n" or die "can't write: $!";
304 close FH or die "can't close: status=$?";
308 Both the main process and any child processes it forks share the same
309 STDIN, STDOUT, and STDERR filehandles. If both processes try to access
310 them at once, strange things can happen. You may also want to close
311 or reopen the filehandles for the child. You can get around this by
312 opening your pipe with open(), but on some systems this means that the
313 child process cannot outlive the parent.
315 =head2 Background Processes
317 You can run a command in the background with:
321 The command's STDOUT and STDERR (and possibly STDIN, depending on your
322 shell) will be the same as the parent's. You won't need to catch
323 SIGCHLD because of the double-fork taking place (see below for more
326 =head2 Complete Dissociation of Child from Parent
328 In some cases (starting server processes, for instance) you'll want to
329 completely dissociate the child process from the parent. This is
330 often called daemonization. A well behaved daemon will also chdir()
331 to the root directory (so it doesn't prevent unmounting the filesystem
332 containing the directory from which it was launched) and redirect its
333 standard file descriptors from and to F</dev/null> (so that random
334 output doesn't wind up on the user's terminal).
339 chdir '/' or die "Can't chdir to /: $!";
340 open STDIN, '/dev/null' or die "Can't read /dev/null: $!";
341 open STDOUT, '>/dev/null'
342 or die "Can't write to /dev/null: $!";
343 defined(my $pid = fork) or die "Can't fork: $!";
345 setsid or die "Can't start a new session: $!";
346 open STDERR, '>&STDOUT' or die "Can't dup stdout: $!";
349 The fork() has to come before the setsid() to ensure that you aren't a
350 process group leader (the setsid() will fail if you are). If your
351 system doesn't have the setsid() function, open F</dev/tty> and use the
352 C<TIOCNOTTY> ioctl() on it instead. See L<tty(4)> for details.
354 Non-Unix users should check their Your_OS::Process module for other
357 =head2 Safe Pipe Opens
359 Another interesting approach to IPC is making your single program go
360 multiprocess and communicate between (or even amongst) yourselves. The
361 open() function will accept a file argument of either C<"-|"> or C<"|-">
362 to do a very interesting thing: it forks a child connected to the
363 filehandle you've opened. The child is running the same program as the
364 parent. This is useful for safely opening a file when running under an
365 assumed UID or GID, for example. If you open a pipe I<to> minus, you can
366 write to the filehandle you opened and your kid will find it in his
367 STDIN. If you open a pipe I<from> minus, you can read from the filehandle
368 you opened whatever your kid writes to his STDOUT.
374 $pid = open(KID_TO_WRITE, "|-");
375 unless (defined $pid) {
376 warn "cannot fork: $!";
377 die "bailing out" if $sleep_count++ > 6;
380 } until defined $pid;
383 print KID_TO_WRITE @some_data;
384 close(KID_TO_WRITE) || warn "kid exited $?";
386 ($EUID, $EGID) = ($UID, $GID); # suid progs only
387 open (FILE, "> /safe/file")
388 || die "can't open /safe/file: $!";
390 print FILE; # child's STDIN is parent's KID
392 exit; # don't forget this
395 Another common use for this construct is when you need to execute
396 something without the shell's interference. With system(), it's
397 straightforward, but you can't use a pipe open or backticks safely.
398 That's because there's no way to stop the shell from getting its hands on
399 your arguments. Instead, use lower-level control to call exec() directly.
401 Here's a safe backtick or pipe open for read:
403 # add error processing as above
404 $pid = open(KID_TO_READ, "-|");
407 while (<KID_TO_READ>) {
408 # do something interesting
410 close(KID_TO_READ) || warn "kid exited $?";
413 ($EUID, $EGID) = ($UID, $GID); # suid only
414 exec($program, @options, @args)
415 || die "can't exec program: $!";
420 And here's a safe pipe open for writing:
422 # add error processing as above
423 $pid = open(KID_TO_WRITE, "|-");
424 $SIG{ALRM} = sub { die "whoops, $program pipe broke" };
430 close(KID_TO_WRITE) || warn "kid exited $?";
433 ($EUID, $EGID) = ($UID, $GID);
434 exec($program, @options, @args)
435 || die "can't exec program: $!";
439 Note that these operations are full Unix forks, which means they may not be
440 correctly implemented on alien systems. Additionally, these are not true
441 multithreading. If you'd like to learn more about threading, see the
442 F<modules> file mentioned below in the SEE ALSO section.
444 =head2 Bidirectional Communication with Another Process
446 While this works reasonably well for unidirectional communication, what
447 about bidirectional communication? The obvious thing you'd like to do
448 doesn't actually work:
450 open(PROG_FOR_READING_AND_WRITING, "| some program |")
452 and if you forget to use the B<-w> flag, then you'll miss out
453 entirely on the diagnostic message:
455 Can't do bidirectional pipe at -e line 1.
457 If you really want to, you can use the standard open2() library function
458 to catch both ends. There's also an open3() for tridirectional I/O so you
459 can also catch your child's STDERR, but doing so would then require an
460 awkward select() loop and wouldn't allow you to use normal Perl input
463 If you look at its source, you'll see that open2() uses low-level
464 primitives like Unix pipe() and exec() calls to create all the connections.
465 While it might have been slightly more efficient by using socketpair(), it
466 would have then been even less portable than it already is. The open2()
467 and open3() functions are unlikely to work anywhere except on a Unix
468 system or some other one purporting to be POSIX compliant.
470 Here's an example of using open2():
474 $pid = open2(*Reader, *Writer, "cat -u -n" );
475 print Writer "stuff\n";
478 The problem with this is that Unix buffering is really going to
479 ruin your day. Even though your C<Writer> filehandle is auto-flushed,
480 and the process on the other end will get your data in a timely manner,
481 you can't usually do anything to force it to give it back to you
482 in a similarly quick fashion. In this case, we could, because we
483 gave I<cat> a B<-u> flag to make it unbuffered. But very few Unix
484 commands are designed to operate over pipes, so this seldom works
485 unless you yourself wrote the program on the other end of the
488 A solution to this is the nonstandard F<Comm.pl> library. It uses
489 pseudo-ttys to make your program behave more reasonably:
492 $ph = open_proc('cat -n');
494 print $ph "a line\n";
495 print "got back ", scalar <$ph>;
498 This way you don't have to have control over the source code of the
499 program you're using. The F<Comm> library also has expect()
500 and interact() functions. Find the library (and we hope its
501 successor F<IPC::Chat>) at your nearest CPAN archive as detailed
502 in the SEE ALSO section below.
504 The newer Expect.pm module from CPAN also addresses this kind of thing.
505 This module requires two other modules from CPAN: IO::Pty and IO::Stty.
506 It sets up a pseudo-terminal to interact with programs that insist on
507 using talking to the terminal device driver. If your system is
508 amongst those supported, this may be your best bet.
510 =head2 Bidirectional Communication with Yourself
512 If you want, you may make low-level pipe() and fork()
513 to stitch this together by hand. This example only
514 talks to itself, but you could reopen the appropriate
515 handles to STDIN and STDOUT and call other processes.
518 # pipe1 - bidirectional communication using two pipe pairs
519 # designed for the socketpair-challenged
520 use IO::Handle; # thousands of lines just for autoflush :-(
521 pipe(PARENT_RDR, CHILD_WTR); # XXX: failure?
522 pipe(CHILD_RDR, PARENT_WTR); # XXX: failure?
523 CHILD_WTR->autoflush(1);
524 PARENT_WTR->autoflush(1);
527 close PARENT_RDR; close PARENT_WTR;
528 print CHILD_WTR "Parent Pid $$ is sending this\n";
529 chomp($line = <CHILD_RDR>);
530 print "Parent Pid $$ just read this: `$line'\n";
531 close CHILD_RDR; close CHILD_WTR;
534 die "cannot fork: $!" unless defined $pid;
535 close CHILD_RDR; close CHILD_WTR;
536 chomp($line = <PARENT_RDR>);
537 print "Child Pid $$ just read this: `$line'\n";
538 print PARENT_WTR "Child Pid $$ is sending this\n";
539 close PARENT_RDR; close PARENT_WTR;
543 But you don't actually have to make two pipe calls. If you
544 have the socketpair() system call, it will do this all for you.
547 # pipe2 - bidirectional communication using socketpair
548 # "the best ones always go both ways"
551 use IO::Handle; # thousands of lines just for autoflush :-(
552 # We say AF_UNIX because although *_LOCAL is the
553 # POSIX 1003.1g form of the constant, many machines
554 # still don't have it.
555 socketpair(CHILD, PARENT, AF_UNIX, SOCK_STREAM, PF_UNSPEC)
556 or die "socketpair: $!";
559 PARENT->autoflush(1);
563 print CHILD "Parent Pid $$ is sending this\n";
564 chomp($line = <CHILD>);
565 print "Parent Pid $$ just read this: `$line'\n";
569 die "cannot fork: $!" unless defined $pid;
571 chomp($line = <PARENT>);
572 print "Child Pid $$ just read this: `$line'\n";
573 print PARENT "Child Pid $$ is sending this\n";
578 =head1 Sockets: Client/Server Communication
580 While not limited to Unix-derived operating systems (e.g., WinSock on PCs
581 provides socket support, as do some VMS libraries), you may not have
582 sockets on your system, in which case this section probably isn't going to do
583 you much good. With sockets, you can do both virtual circuits (i.e., TCP
584 streams) and datagrams (i.e., UDP packets). You may be able to do even more
585 depending on your system.
587 The Perl function calls for dealing with sockets have the same names as
588 the corresponding system calls in C, but their arguments tend to differ
589 for two reasons: first, Perl filehandles work differently than C file
590 descriptors. Second, Perl already knows the length of its strings, so you
591 don't need to pass that information.
593 One of the major problems with old socket code in Perl was that it used
594 hard-coded values for some of the constants, which severely hurt
595 portability. If you ever see code that does anything like explicitly
596 setting C<$AF_INET = 2>, you know you're in for big trouble: An
597 immeasurably superior approach is to use the C<Socket> module, which more
598 reliably grants access to various constants and functions you'll need.
600 If you're not writing a server/client for an existing protocol like
601 NNTP or SMTP, you should give some thought to how your server will
602 know when the client has finished talking, and vice-versa. Most
603 protocols are based on one-line messages and responses (so one party
604 knows the other has finished when a "\n" is received) or multi-line
605 messages and responses that end with a period on an empty line
606 ("\n.\n" terminates a message/response).
608 =head2 Internet Line Terminators
610 The Internet line terminator is "\015\012". Under ASCII variants of
611 Unix, that could usually be written as "\r\n", but under other systems,
612 "\r\n" might at times be "\015\015\012", "\012\012\015", or something
613 completely different. The standards specify writing "\015\012" to be
614 conformant (be strict in what you provide), but they also recommend
615 accepting a lone "\012" on input (but be lenient in what you require).
616 We haven't always been very good about that in the code in this manpage,
617 but unless you're on a Mac, you'll probably be ok.
619 =head2 Internet TCP Clients and Servers
621 Use Internet-domain sockets when you want to do client-server
622 communication that might extend to machines outside of your own system.
624 Here's a sample TCP client using Internet-domain sockets:
629 my ($remote,$port, $iaddr, $paddr, $proto, $line);
631 $remote = shift || 'localhost';
632 $port = shift || 2345; # random port
633 if ($port =~ /\D/) { $port = getservbyname($port, 'tcp') }
634 die "No port" unless $port;
635 $iaddr = inet_aton($remote) || die "no host: $remote";
636 $paddr = sockaddr_in($port, $iaddr);
638 $proto = getprotobyname('tcp');
639 socket(SOCK, PF_INET, SOCK_STREAM, $proto) || die "socket: $!";
640 connect(SOCK, $paddr) || die "connect: $!";
641 while (defined($line = <SOCK>)) {
645 close (SOCK) || die "close: $!";
648 And here's a corresponding server to go along with it. We'll
649 leave the address as INADDR_ANY so that the kernel can choose
650 the appropriate interface on multihomed hosts. If you want sit
651 on a particular interface (like the external side of a gateway
652 or firewall machine), you should fill this in with your real address
657 BEGIN { $ENV{PATH} = '/usr/ucb:/bin' }
662 sub logmsg { print "$0 $$: @_ at ", scalar localtime, "\n" }
664 my $port = shift || 2345;
665 my $proto = getprotobyname('tcp');
666 $port = $1 if $port =~ /(\d+)/; # untaint port number
668 socket(Server, PF_INET, SOCK_STREAM, $proto) || die "socket: $!";
669 setsockopt(Server, SOL_SOCKET, SO_REUSEADDR,
670 pack("l", 1)) || die "setsockopt: $!";
671 bind(Server, sockaddr_in($port, INADDR_ANY)) || die "bind: $!";
672 listen(Server,SOMAXCONN) || die "listen: $!";
674 logmsg "server started on port $port";
678 $SIG{CHLD} = \&REAPER;
680 for ( ; $paddr = accept(Client,Server); close Client) {
681 my($port,$iaddr) = sockaddr_in($paddr);
682 my $name = gethostbyaddr($iaddr,AF_INET);
684 logmsg "connection from $name [",
685 inet_ntoa($iaddr), "]
688 print Client "Hello there, $name, it's now ",
689 scalar localtime, $EOL;
692 And here's a multithreaded version. It's multithreaded in that
693 like most typical servers, it spawns (forks) a slave server to
694 handle the client request so that the master server can quickly
695 go back to service a new client.
699 BEGIN { $ENV{PATH} = '/usr/ucb:/bin' }
704 sub spawn; # forward declaration
705 sub logmsg { print "$0 $$: @_ at ", scalar localtime, "\n" }
707 my $port = shift || 2345;
708 my $proto = getprotobyname('tcp');
709 $port = $1 if $port =~ /(\d+)/; # untaint port number
711 socket(Server, PF_INET, SOCK_STREAM, $proto) || die "socket: $!";
712 setsockopt(Server, SOL_SOCKET, SO_REUSEADDR,
713 pack("l", 1)) || die "setsockopt: $!";
714 bind(Server, sockaddr_in($port, INADDR_ANY)) || die "bind: $!";
715 listen(Server,SOMAXCONN) || die "listen: $!";
717 logmsg "server started on port $port";
724 $SIG{CHLD} = \&REAPER; # loathe sysV
725 logmsg "reaped $waitedpid" . ($? ? " with exit $?" : '');
728 $SIG{CHLD} = \&REAPER;
730 for ( $waitedpid = 0;
731 ($paddr = accept(Client,Server)) || $waitedpid;
732 $waitedpid = 0, close Client)
734 next if $waitedpid and not $paddr;
735 my($port,$iaddr) = sockaddr_in($paddr);
736 my $name = gethostbyaddr($iaddr,AF_INET);
738 logmsg "connection from $name [",
739 inet_ntoa($iaddr), "]
743 print "Hello there, $name, it's now ", scalar localtime, $EOL;
744 exec '/usr/games/fortune' # XXX: `wrong' line terminators
745 or confess "can't exec fortune: $!";
753 unless (@_ == 0 && $coderef && ref($coderef) eq 'CODE') {
754 confess "usage: spawn CODEREF";
758 if (!defined($pid = fork)) {
759 logmsg "cannot fork: $!";
763 return; # I'm the parent
765 # else I'm the child -- go spawn
767 open(STDIN, "<&Client") || die "can't dup client to stdin";
768 open(STDOUT, ">&Client") || die "can't dup client to stdout";
769 ## open(STDERR, ">&STDOUT") || die "can't dup stdout to stderr";
773 This server takes the trouble to clone off a child version via fork() for
774 each incoming request. That way it can handle many requests at once,
775 which you might not always want. Even if you don't fork(), the listen()
776 will allow that many pending connections. Forking servers have to be
777 particularly careful about cleaning up their dead children (called
778 "zombies" in Unix parlance), because otherwise you'll quickly fill up your
781 We suggest that you use the B<-T> flag to use taint checking (see L<perlsec>)
782 even if we aren't running setuid or setgid. This is always a good idea
783 for servers and other programs run on behalf of someone else (like CGI
784 scripts), because it lessens the chances that people from the outside will
785 be able to compromise your system.
787 Let's look at another TCP client. This one connects to the TCP "time"
788 service on a number of different machines and shows how far their clocks
789 differ from the system on which it's being run:
795 my $SECS_of_70_YEARS = 2208988800;
796 sub ctime { scalar localtime(shift) }
798 my $iaddr = gethostbyname('localhost');
799 my $proto = getprotobyname('tcp');
800 my $port = getservbyname('time', 'tcp');
801 my $paddr = sockaddr_in(0, $iaddr);
805 printf "%-24s %8s %s\n", "localhost", 0, ctime(time());
807 foreach $host (@ARGV) {
808 printf "%-24s ", $host;
809 my $hisiaddr = inet_aton($host) || die "unknown host";
810 my $hispaddr = sockaddr_in($port, $hisiaddr);
811 socket(SOCKET, PF_INET, SOCK_STREAM, $proto) || die "socket: $!";
812 connect(SOCKET, $hispaddr) || die "bind: $!";
814 read(SOCKET, $rtime, 4);
816 my $histime = unpack("N", $rtime) - $SECS_of_70_YEARS ;
817 printf "%8d %s\n", $histime - time, ctime($histime);
820 =head2 Unix-Domain TCP Clients and Servers
822 That's fine for Internet-domain clients and servers, but what about local
823 communications? While you can use the same setup, sometimes you don't
824 want to. Unix-domain sockets are local to the current host, and are often
825 used internally to implement pipes. Unlike Internet domain sockets, Unix
826 domain sockets can show up in the file system with an ls(1) listing.
829 srw-rw-rw- 1 root 0 Oct 31 07:23 /dev/log
831 You can test for these with Perl's B<-S> file test:
833 unless ( -S '/dev/log' ) {
834 die "something's wicked with the print system";
837 Here's a sample Unix-domain client:
842 my ($rendezvous, $line);
844 $rendezvous = shift || '/tmp/catsock';
845 socket(SOCK, PF_UNIX, SOCK_STREAM, 0) || die "socket: $!";
846 connect(SOCK, sockaddr_un($rendezvous)) || die "connect: $!";
847 while (defined($line = <SOCK>)) {
852 And here's a corresponding server. You don't have to worry about silly
853 network terminators here because Unix domain sockets are guaranteed
854 to be on the localhost, and thus everything works right.
861 BEGIN { $ENV{PATH} = '/usr/ucb:/bin' }
862 sub logmsg { print "$0 $$: @_ at ", scalar localtime, "\n" }
864 my $NAME = '/tmp/catsock';
865 my $uaddr = sockaddr_un($NAME);
866 my $proto = getprotobyname('tcp');
868 socket(Server,PF_UNIX,SOCK_STREAM,0) || die "socket: $!";
870 bind (Server, $uaddr) || die "bind: $!";
871 listen(Server,SOMAXCONN) || die "listen: $!";
873 logmsg "server started on $NAME";
879 $SIG{CHLD} = \&REAPER; # loathe sysV
880 logmsg "reaped $waitedpid" . ($? ? " with exit $?" : '');
883 $SIG{CHLD} = \&REAPER;
886 for ( $waitedpid = 0;
887 accept(Client,Server) || $waitedpid;
888 $waitedpid = 0, close Client)
891 logmsg "connection on $NAME";
893 print "Hello there, it's now ", scalar localtime, "\n";
894 exec '/usr/games/fortune' or die "can't exec fortune: $!";
898 As you see, it's remarkably similar to the Internet domain TCP server, so
899 much so, in fact, that we've omitted several duplicate functions--spawn(),
900 logmsg(), ctime(), and REAPER()--which are exactly the same as in the
903 So why would you ever want to use a Unix domain socket instead of a
904 simpler named pipe? Because a named pipe doesn't give you sessions. You
905 can't tell one process's data from another's. With socket programming,
906 you get a separate session for each client: that's why accept() takes two
909 For example, let's say that you have a long running database server daemon
910 that you want folks from the World Wide Web to be able to access, but only
911 if they go through a CGI interface. You'd have a small, simple CGI
912 program that does whatever checks and logging you feel like, and then acts
913 as a Unix-domain client and connects to your private server.
915 =head1 TCP Clients with IO::Socket
917 For those preferring a higher-level interface to socket programming, the
918 IO::Socket module provides an object-oriented approach. IO::Socket is
919 included as part of the standard Perl distribution as of the 5.004
920 release. If you're running an earlier version of Perl, just fetch
921 IO::Socket from CPAN, where you'll also find find modules providing easy
922 interfaces to the following systems: DNS, FTP, Ident (RFC 931), NIS and
923 NISPlus, NNTP, Ping, POP3, SMTP, SNMP, SSLeay, Telnet, and Time--just
926 =head2 A Simple Client
928 Here's a client that creates a TCP connection to the "daytime"
929 service at port 13 of the host name "localhost" and prints out everything
930 that the server there cares to provide.
934 $remote = IO::Socket::INET->new(
936 PeerAddr => "localhost",
937 PeerPort => "daytime(13)",
939 or die "cannot connect to daytime port at localhost";
940 while ( <$remote> ) { print }
942 When you run this program, you should get something back that
945 Wed May 14 08:40:46 MDT 1997
947 Here are what those parameters to the C<new> constructor mean:
953 This is which protocol to use. In this case, the socket handle returned
954 will be connected to a TCP socket, because we want a stream-oriented
955 connection, that is, one that acts pretty much like a plain old file.
956 Not all sockets are this of this type. For example, the UDP protocol
957 can be used to make a datagram socket, used for message-passing.
961 This is the name or Internet address of the remote host the server is
962 running on. We could have specified a longer name like C<"www.perl.com">,
963 or an address like C<"204.148.40.9">. For demonstration purposes, we've
964 used the special hostname C<"localhost">, which should always mean the
965 current machine you're running on. The corresponding Internet address
966 for localhost is C<"127.1">, if you'd rather use that.
970 This is the service name or port number we'd like to connect to.
971 We could have gotten away with using just C<"daytime"> on systems with a
972 well-configured system services file,[FOOTNOTE: The system services file
973 is in I</etc/services> under Unix] but just in case, we've specified the
974 port number (13) in parentheses. Using just the number would also have
975 worked, but constant numbers make careful programmers nervous.
979 Notice how the return value from the C<new> constructor is used as
980 a filehandle in the C<while> loop? That's what's called an indirect
981 filehandle, a scalar variable containing a filehandle. You can use
982 it the same way you would a normal filehandle. For example, you
983 can read one line from it this way:
987 all remaining lines from is this way:
991 and send a line of data to it this way:
993 print $handle "some data\n";
995 =head2 A Webget Client
997 Here's a simple client that takes a remote host to fetch a document
998 from, and then a list of documents to get from that host. This is a
999 more interesting client than the previous one because it first sends
1000 something to the server before fetching the server's response.
1004 unless (@ARGV > 1) { die "usage: $0 host document ..." }
1005 $host = shift(@ARGV);
1008 foreach $document ( @ARGV ) {
1009 $remote = IO::Socket::INET->new( Proto => "tcp",
1011 PeerPort => "http(80)",
1013 unless ($remote) { die "cannot connect to http daemon on $host" }
1014 $remote->autoflush(1);
1015 print $remote "GET $document HTTP/1.0" . $BLANK;
1016 while ( <$remote> ) { print }
1020 The web server handing the "http" service, which is assumed to be at
1021 its standard port, number 80. If your the web server you're trying to
1022 connect to is at a different port (like 1080 or 8080), you should specify
1023 as the named-parameter pair, C<PeerPort =E<gt> 8080>. The C<autoflush>
1024 method is used on the socket because otherwise the system would buffer
1025 up the output we sent it. (If you're on a Mac, you'll also need to
1026 change every C<"\n"> in your code that sends data over the network to
1027 be a C<"\015\012"> instead.)
1029 Connecting to the server is only the first part of the process: once you
1030 have the connection, you have to use the server's language. Each server
1031 on the network has its own little command language that it expects as
1032 input. The string that we send to the server starting with "GET" is in
1033 HTTP syntax. In this case, we simply request each specified document.
1034 Yes, we really are making a new connection for each document, even though
1035 it's the same host. That's the way you always used to have to speak HTTP.
1036 Recent versions of web browsers may request that the remote server leave
1037 the connection open a little while, but the server doesn't have to honor
1040 Here's an example of running that program, which we'll call I<webget>:
1042 % webget www.perl.com /guanaco.html
1043 HTTP/1.1 404 File Not Found
1044 Date: Thu, 08 May 1997 18:02:32 GMT
1045 Server: Apache/1.2b6
1047 Content-type: text/html
1049 <HEAD><TITLE>404 File Not Found</TITLE></HEAD>
1050 <BODY><H1>File Not Found</H1>
1051 The requested URL /guanaco.html was not found on this server.<P>
1054 Ok, so that's not very interesting, because it didn't find that
1055 particular document. But a long response wouldn't have fit on this page.
1057 For a more fully-featured version of this program, you should look to
1058 the I<lwp-request> program included with the LWP modules from CPAN.
1060 =head2 Interactive Client with IO::Socket
1062 Well, that's all fine if you want to send one command and get one answer,
1063 but what about setting up something fully interactive, somewhat like
1064 the way I<telnet> works? That way you can type a line, get the answer,
1065 type a line, get the answer, etc.
1067 This client is more complicated than the two we've done so far, but if
1068 you're on a system that supports the powerful C<fork> call, the solution
1069 isn't that rough. Once you've made the connection to whatever service
1070 you'd like to chat with, call C<fork> to clone your process. Each of
1071 these two identical process has a very simple job to do: the parent
1072 copies everything from the socket to standard output, while the child
1073 simultaneously copies everything from standard input to the socket.
1074 To accomplish the same thing using just one process would be I<much>
1075 harder, because it's easier to code two processes to do one thing than it
1076 is to code one process to do two things. (This keep-it-simple principle
1077 a cornerstones of the Unix philosophy, and good software engineering as
1078 well, which is probably why it's spread to other systems.)
1085 my ($host, $port, $kidpid, $handle, $line);
1087 unless (@ARGV == 2) { die "usage: $0 host port" }
1088 ($host, $port) = @ARGV;
1090 # create a tcp connection to the specified host and port
1091 $handle = IO::Socket::INET->new(Proto => "tcp",
1094 or die "can't connect to port $port on $host: $!";
1096 $handle->autoflush(1); # so output gets there right away
1097 print STDERR "[Connected to $host:$port]\n";
1099 # split the program into two processes, identical twins
1100 die "can't fork: $!" unless defined($kidpid = fork());
1102 # the if{} block runs only in the parent process
1104 # copy the socket to standard output
1105 while (defined ($line = <$handle>)) {
1108 kill("TERM", $kidpid); # send SIGTERM to child
1110 # the else{} block runs only in the child process
1112 # copy standard input to the socket
1113 while (defined ($line = <STDIN>)) {
1114 print $handle $line;
1118 The C<kill> function in the parent's C<if> block is there to send a
1119 signal to our child process (current running in the C<else> block)
1120 as soon as the remote server has closed its end of the connection.
1122 If the remote server sends data a byte at time, and you need that
1123 data immediately without waiting for a newline (which might not happen),
1124 you may wish to replace the C<while> loop in the parent with the
1128 while (sysread($handle, $byte, 1) == 1) {
1132 Making a system call for each byte you want to read is not very efficient
1133 (to put it mildly) but is the simplest to explain and works reasonably
1136 =head1 TCP Servers with IO::Socket
1138 As always, setting up a server is little bit more involved than running a client.
1139 The model is that the server creates a special kind of socket that
1140 does nothing but listen on a particular port for incoming connections.
1141 It does this by calling the C<IO::Socket::INET-E<gt>new()> method with
1142 slightly different arguments than the client did.
1148 This is which protocol to use. Like our clients, we'll
1149 still specify C<"tcp"> here.
1154 port in the C<LocalPort> argument, which we didn't do for the client.
1155 This is service name or port number for which you want to be the
1156 server. (Under Unix, ports under 1024 are restricted to the
1157 superuser.) In our sample, we'll use port 9000, but you can use
1158 any port that's not currently in use on your system. If you try
1159 to use one already in used, you'll get an "Address already in use"
1160 message. Under Unix, the C<netstat -a> command will show
1161 which services current have servers.
1165 The C<Listen> parameter is set to the maximum number of
1166 pending connections we can accept until we turn away incoming clients.
1167 Think of it as a call-waiting queue for your telephone.
1168 The low-level Socket module has a special symbol for the system maximum, which
1173 The C<Reuse> parameter is needed so that we restart our server
1174 manually without waiting a few minutes to allow system buffers to
1179 Once the generic server socket has been created using the parameters
1180 listed above, the server then waits for a new client to connect
1181 to it. The server blocks in the C<accept> method, which eventually an
1182 bidirectional connection to the remote client. (Make sure to autoflush
1183 this handle to circumvent buffering.)
1185 To add to user-friendliness, our server prompts the user for commands.
1186 Most servers don't do this. Because of the prompt without a newline,
1187 you'll have to use the C<sysread> variant of the interactive client above.
1189 This server accepts one of five different commands, sending output
1190 back to the client. Note that unlike most network servers, this one
1191 only handles one incoming client at a time. Multithreaded servers are
1192 covered in Chapter 6 of the Camel.
1194 Here's the code. We'll
1198 use Net::hostent; # for OO version of gethostbyaddr
1200 $PORT = 9000; # pick something not in use
1202 $server = IO::Socket::INET->new( Proto => 'tcp',
1204 Listen => SOMAXCONN,
1207 die "can't setup server" unless $server;
1208 print "[Server $0 accepting clients]\n";
1210 while ($client = $server->accept()) {
1211 $client->autoflush(1);
1212 print $client "Welcome to $0; type help for command list.\n";
1213 $hostinfo = gethostbyaddr($client->peeraddr);
1214 printf "[Connect from %s]\n", $hostinfo->name || $client->peerhost;
1215 print $client "Command? ";
1216 while ( <$client>) {
1217 next unless /\S/; # blank line
1218 if (/quit|exit/i) { last; }
1219 elsif (/date|time/i) { printf $client "%s\n", scalar localtime; }
1220 elsif (/who/i ) { print $client `who 2>&1`; }
1221 elsif (/cookie/i ) { print $client `/usr/games/fortune 2>&1`; }
1222 elsif (/motd/i ) { print $client `cat /etc/motd 2>&1`; }
1224 print $client "Commands: quit date who cookie motd\n";
1227 print $client "Command? ";
1232 =head1 UDP: Message Passing
1234 Another kind of client-server setup is one that uses not connections, but
1235 messages. UDP communications involve much lower overhead but also provide
1236 less reliability, as there are no promises that messages will arrive at
1237 all, let alone in order and unmangled. Still, UDP offers some advantages
1238 over TCP, including being able to "broadcast" or "multicast" to a whole
1239 bunch of destination hosts at once (usually on your local subnet). If you
1240 find yourself overly concerned about reliability and start building checks
1241 into your message system, then you probably should use just TCP to start
1244 Here's a UDP program similar to the sample Internet TCP client given
1245 earlier. However, instead of checking one host at a time, the UDP version
1246 will check many of them asynchronously by simulating a multicast and then
1247 using select() to do a timed-out wait for I/O. To do something similar
1248 with TCP, you'd have to use a different socket handle for each host.
1255 my ( $count, $hisiaddr, $hispaddr, $histime,
1256 $host, $iaddr, $paddr, $port, $proto,
1257 $rin, $rout, $rtime, $SECS_of_70_YEARS);
1259 $SECS_of_70_YEARS = 2208988800;
1261 $iaddr = gethostbyname(hostname());
1262 $proto = getprotobyname('udp');
1263 $port = getservbyname('time', 'udp');
1264 $paddr = sockaddr_in(0, $iaddr); # 0 means let kernel pick
1266 socket(SOCKET, PF_INET, SOCK_DGRAM, $proto) || die "socket: $!";
1267 bind(SOCKET, $paddr) || die "bind: $!";
1270 printf "%-12s %8s %s\n", "localhost", 0, scalar localtime time;
1274 $hisiaddr = inet_aton($host) || die "unknown host";
1275 $hispaddr = sockaddr_in($port, $hisiaddr);
1276 defined(send(SOCKET, 0, 0, $hispaddr)) || die "send $host: $!";
1280 vec($rin, fileno(SOCKET), 1) = 1;
1282 # timeout after 10.0 seconds
1283 while ($count && select($rout = $rin, undef, undef, 10.0)) {
1285 ($hispaddr = recv(SOCKET, $rtime, 4, 0)) || die "recv: $!";
1286 ($port, $hisiaddr) = sockaddr_in($hispaddr);
1287 $host = gethostbyaddr($hisiaddr, AF_INET);
1288 $histime = unpack("N", $rtime) - $SECS_of_70_YEARS ;
1289 printf "%-12s ", $host;
1290 printf "%8d %s\n", $histime - time, scalar localtime($histime);
1296 While System V IPC isn't so widely used as sockets, it still has some
1297 interesting uses. You can't, however, effectively use SysV IPC or
1298 Berkeley mmap() to have shared memory so as to share a variable amongst
1299 several processes. That's because Perl would reallocate your string when
1300 you weren't wanting it to.
1302 Here's a small example showing shared memory usage.
1304 use IPC::SysV qw(IPC_PRIVATE IPC_RMID S_IRWXU S_IRWXG S_IRWXO);
1307 $key = shmget(IPC_PRIVATE, $size, S_IRWXU|S_IRWXG|S_IRWXO) || die "$!";
1308 print "shm key $key\n";
1310 $message = "Message #1";
1311 shmwrite($key, $message, 0, 60) || die "$!";
1312 print "wrote: '$message'\n";
1313 shmread($key, $buff, 0, 60) || die "$!";
1314 print "read : '$buff'\n";
1316 # the buffer of shmread is zero-character end-padded.
1317 substr($buff, index($buff, "\0")) = '';
1318 print "un" unless $buff eq $message;
1321 print "deleting shm $key\n";
1322 shmctl($key, IPC_RMID, 0) || die "$!";
1324 Here's an example of a semaphore:
1326 use IPC::SysV qw(IPC_CREAT);
1329 $key = semget($IPC_KEY, 10, 0666 | IPC_CREAT ) || die "$!";
1330 print "shm key $key\n";
1332 Put this code in a separate file to be run in more than one process.
1333 Call the file F<take>:
1335 # create a semaphore
1338 $key = semget($IPC_KEY, 0 , 0 );
1339 die if !defined($key);
1345 # wait for semaphore to be zero
1347 $opstring1 = pack("sss", $semnum, $semop, $semflag);
1349 # Increment the semaphore count
1351 $opstring2 = pack("sss", $semnum, $semop, $semflag);
1352 $opstring = $opstring1 . $opstring2;
1354 semop($key,$opstring) || die "$!";
1356 Put this code in a separate file to be run in more than one process.
1357 Call this file F<give>:
1359 # 'give' the semaphore
1360 # run this in the original process and you will see
1361 # that the second process continues
1364 $key = semget($IPC_KEY, 0, 0);
1365 die if !defined($key);
1370 # Decrement the semaphore count
1372 $opstring = pack("sss", $semnum, $semop, $semflag);
1374 semop($key,$opstring) || die "$!";
1376 The SysV IPC code above was written long ago, and it's definitely
1377 clunky looking. For a more modern look, see the IPC::SysV module
1378 which is included with Perl starting from Perl 5.005.
1382 Most of these routines quietly but politely return C<undef> when they
1383 fail instead of causing your program to die right then and there due to
1384 an uncaught exception. (Actually, some of the new I<Socket> conversion
1385 functions croak() on bad arguments.) It is therefore essential to
1386 check return values from these functions. Always begin your socket
1387 programs this way for optimal success, and don't forget to add B<-T>
1388 taint checking flag to the #! line for servers:
1397 All these routines create system-specific portability problems. As noted
1398 elsewhere, Perl is at the mercy of your C libraries for much of its system
1399 behaviour. It's probably safest to assume broken SysV semantics for
1400 signals and to stick with simple TCP and UDP socket operations; e.g., don't
1401 try to pass open file descriptors over a local UDP datagram socket if you
1402 want your code to stand a chance of being portable.
1404 As mentioned in the signals section, because few vendors provide C
1405 libraries that are safely re-entrant, the prudent programmer will do
1406 little else within a handler beyond setting a numeric variable that
1407 already exists; or, if locked into a slow (restarting) system call,
1408 using die() to raise an exception and longjmp(3) out. In fact, even
1409 these may in some cases cause a core dump. It's probably best to avoid
1410 signals except where they are absolutely inevitable. This
1411 will be addressed in a future release of Perl.
1415 Tom Christiansen, with occasional vestiges of Larry Wall's original
1416 version and suggestions from the Perl Porters.
1420 There's a lot more to networking than this, but this should get you
1423 For intrepid programmers, the indispensable textbook is I<Unix Network
1424 Programming> by W. Richard Stevens (published by Addison-Wesley). Note
1425 that most books on networking address networking from the perspective of
1426 a C programmer; translation to Perl is left as an exercise for the reader.
1428 The IO::Socket(3) manpage describes the object library, and the Socket(3)
1429 manpage describes the low-level interface to sockets. Besides the obvious
1430 functions in L<perlfunc>, you should also check out the F<modules> file
1431 at your nearest CPAN site. (See L<perlmodlib> or best yet, the F<Perl
1432 FAQ> for a description of what CPAN is and where to get it.)
1434 Section 5 of the F<modules> file is devoted to "Networking, Device Control
1435 (modems), and Interprocess Communication", and contains numerous unbundled
1436 modules numerous networking modules, Chat and Expect operations, CGI
1437 programming, DCE, FTP, IPC, NNTP, Proxy, Ptty, RPC, SNMP, SMTP, Telnet,
1438 Threads, and ToolTalk--just to name a few.