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
14 or references of user-installed signal handlers. These handlers will
15 be called with an argument which is the name of the signal that
16 triggered it. A signal may be generated intentionally from a
17 particular keyboard sequence like control-C or control-Z, sent to you
18 from another process, or triggered automatically by the kernel when
19 special events transpire, like a child process exiting, your process
20 running out of stack space, or hitting file size limit.
22 For example, to trap an interrupt signal, set up a handler like this:
27 die "Somebody sent me a SIG$signame";
29 $SIG{INT} = 'catch_zap'; # could fail in modules
30 $SIG{INT} = \&catch_zap; # best strategy
32 Prior to Perl 5.7.3 it was necessary to do as little as you possibly
33 could in your handler; notice how all we do is set a global variable
34 and then raise an exception. That's because on most systems,
35 libraries are not re-entrant; particularly, memory allocation and I/O
36 routines are not. That meant that doing nearly I<anything> in your
37 handler could in theory trigger a memory fault and subsequent core
38 dump - see L<Deferred Signals> below.
40 The names of the signals are the ones listed out by C<kill -l> on your
41 system, or you can retrieve them from the Config module. Set up an
42 @signame list indexed by number to get the name and a %signo table
43 indexed by name to get the number:
46 defined $Config{sig_name} || die "No sigs?";
47 foreach $name (split(' ', $Config{sig_name})) {
53 So to check whether signal 17 and SIGALRM were the same, do just this:
55 print "signal #17 = $signame[17]\n";
57 print "SIGALRM is $signo{ALRM}\n";
60 You may also choose to assign the strings C<'IGNORE'> or C<'DEFAULT'> as
61 the handler, in which case Perl will try to discard the signal or do the
64 On most Unix platforms, the C<CHLD> (sometimes also known as C<CLD>) signal
65 has special behavior with respect to a value of C<'IGNORE'>.
66 Setting C<$SIG{CHLD}> to C<'IGNORE'> on such a platform has the effect of
67 not creating zombie processes when the parent process fails to C<wait()>
68 on its child processes (i.e. child processes are automatically reaped).
69 Calling C<wait()> with C<$SIG{CHLD}> set to C<'IGNORE'> usually returns
70 C<-1> on such platforms.
72 Some signals can be neither trapped nor ignored, such as
73 the KILL and STOP (but not the TSTP) signals. One strategy for
74 temporarily ignoring signals is to use a local() statement, which will be
75 automatically restored once your block is exited. (Remember that local()
76 values are "inherited" by functions called from within that block.)
79 local $SIG{INT} = 'IGNORE';
83 # interrupts still ignored, for now...
86 Sending a signal to a negative process ID means that you send the signal
87 to the entire Unix process-group. This code sends a hang-up signal to all
88 processes in the current process group (and sets $SIG{HUP} to IGNORE so
89 it doesn't kill itself):
92 local $SIG{HUP} = 'IGNORE';
94 # snazzy writing of: kill('HUP', -$$)
97 Another interesting signal to send is signal number zero. This doesn't
98 actually affect another process, but instead checks whether it's alive
99 or has changed its UID.
101 unless (kill 0 => $kid_pid) {
102 warn "something wicked happened to $kid_pid";
105 You might also want to employ anonymous functions for simple signal
108 $SIG{INT} = sub { die "\nOutta here!\n" };
110 But that will be problematic for the more complicated handlers that need
111 to reinstall themselves. Because Perl's signal mechanism is currently
112 based on the signal(3) function from the C library, you may sometimes be so
113 misfortunate as to run on systems where that function is "broken", that
114 is, it behaves in the old unreliable SysV way rather than the newer, more
115 reasonable BSD and POSIX fashion. So you'll see defensive people writing
116 signal handlers like this:
120 # loathe sysV: it makes us not only reinstate
121 # the handler, but place it after the wait
122 $SIG{CHLD} = \&REAPER;
124 $SIG{CHLD} = \&REAPER;
125 # now do something that forks...
129 use POSIX ":sys_wait_h";
132 # If a second child dies while in the signal handler caused by the
133 # first death, we won't get another signal. So must loop here else
134 # we will leave the unreaped child as a zombie. And the next time
135 # two children die we get another zombie. And so on.
136 while (($child = waitpid(-1,WNOHANG)) > 0) {
137 $Kid_Status{$child} = $?;
139 $SIG{CHLD} = \&REAPER; # still loathe sysV
141 $SIG{CHLD} = \&REAPER;
142 # do something that forks...
144 Signal handling is also used for timeouts in Unix, While safely
145 protected within an C<eval{}> block, you set a signal handler to trap
146 alarm signals and then schedule to have one delivered to you in some
147 number of seconds. Then try your blocking operation, clearing the alarm
148 when it's done but not before you've exited your C<eval{}> block. If it
149 goes off, you'll use die() to jump out of the block, much as you might
150 using longjmp() or throw() in other languages.
155 local $SIG{ALRM} = sub { die "alarm clock restart" };
157 flock(FH, 2); # blocking write lock
160 if ($@ and $@ !~ /alarm clock restart/) { die }
162 If the operation being timed out is system() or qx(), this technique
163 is liable to generate zombies. If this matters to you, you'll
164 need to do your own fork() and exec(), and kill the errant child process.
166 For more complex signal handling, you might see the standard POSIX
167 module. Lamentably, this is almost entirely undocumented, but
168 the F<t/lib/posix.t> file from the Perl source distribution has some
173 A named pipe (often referred to as a FIFO) is an old Unix IPC
174 mechanism for processes communicating on the same machine. It works
175 just like a regular, connected anonymous pipes, except that the
176 processes rendezvous using a filename and don't have to be related.
178 To create a named pipe, use the Unix command mknod(1) or on some
179 systems, mkfifo(1). These may not be in your normal path.
181 # system return val is backwards, so && not ||
183 $ENV{PATH} .= ":/etc:/usr/etc";
184 if ( system('mknod', $path, 'p')
185 && system('mkfifo', $path) )
187 die "mk{nod,fifo} $path failed";
191 A fifo is convenient when you want to connect a process to an unrelated
192 one. When you open a fifo, the program will block until there's something
195 For example, let's say you'd like to have your F<.signature> file be a
196 named pipe that has a Perl program on the other end. Now every time any
197 program (like a mailer, news reader, finger program, etc.) tries to read
198 from that file, the reading program will block and your program will
199 supply the new signature. We'll use the pipe-checking file test B<-p>
200 to find out whether anyone (or anything) has accidentally removed our fifo.
203 $FIFO = '.signature';
204 $ENV{PATH} .= ":/etc:/usr/games";
209 system('mknod', $FIFO, 'p')
210 && die "can't mknod $FIFO: $!";
213 # next line blocks until there's a reader
214 open (FIFO, "> $FIFO") || die "can't write $FIFO: $!";
215 print FIFO "John Smith (smith\@host.org)\n", `fortune -s`;
217 sleep 2; # to avoid dup signals
220 =head2 Deferred Signals
222 In Perls before Perl 5.7.3 by installing Perl code to deal with
223 signals, you were exposing yourself to danger from two things. First,
224 few system library functions are re-entrant. If the signal interrupts
225 while Perl is executing one function (like malloc(3) or printf(3)),
226 and your signal handler then calls the same function again, you could
227 get unpredictable behavior--often, a core dump. Second, Perl isn't
228 itself re-entrant at the lowest levels. If the signal interrupts Perl
229 while Perl is changing its own internal data structures, similarly
230 unpredictable behaviour may result.
232 There were two things you could do, knowing this: be paranoid or be
233 pragmatic. The paranoid approach was to do as little as possible in your
234 signal handler. Set an existing integer variable that already has a
235 value, and return. This doesn't help you if you're in a slow system call,
236 which will just restart. That means you have to C<die> to longjump(3) out
237 of the handler. Even this is a little cavalier for the true paranoiac,
238 who avoids C<die> in a handler because the system I<is> out to get you.
239 The pragmatic approach was to say ``I know the risks, but prefer the
240 convenience'', and to do anything you wanted in your signal handler,
241 and be prepared to clean up core dumps now and again.
243 In Perl 5.7.3 and later to avoid these problems signals are
244 "deferred"-- that is when the signal is delivered to the process by
245 the system (to the C code that implements Perl) a flag is set, and the
246 handler returns immediately. Then at strategic "safe" points in the
247 Perl interpreter (e.g. when it is about to execute a new opcode) the
248 flags are checked and the Perl level handler from %SIG is
249 executed. The "deferred" scheme allows much more flexibility in the
250 coding of signal handler as we know Perl interpreter is in a safe
251 state, and that we are not in a system library function when the
252 handler is called. However the implementation does differ from
253 previous Perls in the following ways:
257 =item Long running opcodes
259 As Perl interpreter only looks at the signal flags when it about to
260 execute a new opcode if a signal arrives during a long running opcode
261 (e.g. a regular expression operation on a very large string) then
262 signal will not be seen until operation completes.
264 =item Interrupting IO
266 When a signal is delivered (e.g. INT control-C) the operating system
267 breaks into IO operations like C<read> (used to implement Perls
268 E<lt>E<gt> operator). On older Perls the handler was called
269 immediately (and as C<read> is not "unsafe" this worked well). With
270 the "deferred" scheme the handler is not called immediately, and if
271 Perl is using system's C<stdio> library that library may re-start the
272 C<read> without returning to Perl and giving it a chance to call the
273 %SIG handler. If this happens on your system the solution is to use
274 C<:perlio> layer to do IO - at least on those handles which you want
275 to be able to break into with signals. (The C<:perlio> layer checks
276 the signal flags and calls %SIG handlers before resuming IO operation.)
278 Note that the default in Perl 5.7.3 and later is to automatically use
279 the C<:perlio> layer.
281 =item Signals as "faults"
283 Certain signals e.g. SEGV, ILL, BUS are generated as a result of
284 virtual memory or other "faults". These are normally fatal and there
285 is little a Perl-level handler can do with them. (In particular the
286 old signal scheme was particularly unsafe in such cases.) However if
287 a %SIG handler is set the new scheme simply sets a flag and returns as
288 described above. This may cause the operating system to try the
289 offending machine instruction again and - as nothing has changed - it
290 will generate the signal again. The result of this is a rather odd
291 "loop". In future Perl's signal mechanism may be changed to avoid this
292 - perhaps by simply disallowing %SIG handlers on signals of that
293 type. Until then the work-round is not to set a %SIG handler on those
294 signals. (Which signals they are is operating system dependant.)
296 =item Signals triggered by operating system state
298 On some operating systems certain signal handlers are supposed to "do
299 something" before returning. One example can be CHLD or CLD which
300 indicates a child process has completed. On some operating systems the
301 signal handler is expected to C<wait> for the completed child
302 process. On such systems the deferred signal scheme will not work for
303 those signals (it does not do the C<wait>). Again the failure will
304 look like a loop as the operating system will re-issue the signal as
305 there are un-waited-for completed child processes.
309 =head1 Using open() for IPC
311 Perl's basic open() statement can also be used for unidirectional
312 interprocess communication by either appending or prepending a pipe
313 symbol to the second argument to open(). Here's how to start
314 something up in a child process you intend to write to:
316 open(SPOOLER, "| cat -v | lpr -h 2>/dev/null")
317 || die "can't fork: $!";
318 local $SIG{PIPE} = sub { die "spooler pipe broke" };
319 print SPOOLER "stuff\n";
320 close SPOOLER || die "bad spool: $! $?";
322 And here's how to start up a child process you intend to read from:
324 open(STATUS, "netstat -an 2>&1 |")
325 || die "can't fork: $!";
327 next if /^(tcp|udp)/;
330 close STATUS || die "bad netstat: $! $?";
332 If one can be sure that a particular program is a Perl script that is
333 expecting filenames in @ARGV, the clever programmer can write something
336 % program f1 "cmd1|" - f2 "cmd2|" f3 < tmpfile
338 and irrespective of which shell it's called from, the Perl program will
339 read from the file F<f1>, the process F<cmd1>, standard input (F<tmpfile>
340 in this case), the F<f2> file, the F<cmd2> command, and finally the F<f3>
341 file. Pretty nifty, eh?
343 You might notice that you could use backticks for much the
344 same effect as opening a pipe for reading:
346 print grep { !/^(tcp|udp)/ } `netstat -an 2>&1`;
347 die "bad netstat" if $?;
349 While this is true on the surface, it's much more efficient to process the
350 file one line or record at a time because then you don't have to read the
351 whole thing into memory at once. It also gives you finer control of the
352 whole process, letting you to kill off the child process early if you'd
355 Be careful to check both the open() and the close() return values. If
356 you're I<writing> to a pipe, you should also trap SIGPIPE. Otherwise,
357 think of what happens when you start up a pipe to a command that doesn't
358 exist: the open() will in all likelihood succeed (it only reflects the
359 fork()'s success), but then your output will fail--spectacularly. Perl
360 can't know whether the command worked because your command is actually
361 running in a separate process whose exec() might have failed. Therefore,
362 while readers of bogus commands return just a quick end of file, writers
363 to bogus command will trigger a signal they'd better be prepared to
366 open(FH, "|bogus") or die "can't fork: $!";
367 print FH "bang\n" or die "can't write: $!";
368 close FH or die "can't close: $!";
370 That won't blow up until the close, and it will blow up with a SIGPIPE.
371 To catch it, you could use this:
373 $SIG{PIPE} = 'IGNORE';
374 open(FH, "|bogus") or die "can't fork: $!";
375 print FH "bang\n" or die "can't write: $!";
376 close FH or die "can't close: status=$?";
380 Both the main process and any child processes it forks share the same
381 STDIN, STDOUT, and STDERR filehandles. If both processes try to access
382 them at once, strange things can happen. You may also want to close
383 or reopen the filehandles for the child. You can get around this by
384 opening your pipe with open(), but on some systems this means that the
385 child process cannot outlive the parent.
387 =head2 Background Processes
389 You can run a command in the background with:
393 The command's STDOUT and STDERR (and possibly STDIN, depending on your
394 shell) will be the same as the parent's. You won't need to catch
395 SIGCHLD because of the double-fork taking place (see below for more
398 =head2 Complete Dissociation of Child from Parent
400 In some cases (starting server processes, for instance) you'll want to
401 completely dissociate the child process from the parent. This is
402 often called daemonization. A well behaved daemon will also chdir()
403 to the root directory (so it doesn't prevent unmounting the filesystem
404 containing the directory from which it was launched) and redirect its
405 standard file descriptors from and to F</dev/null> (so that random
406 output doesn't wind up on the user's terminal).
411 chdir '/' or die "Can't chdir to /: $!";
412 open STDIN, '/dev/null' or die "Can't read /dev/null: $!";
413 open STDOUT, '>/dev/null'
414 or die "Can't write to /dev/null: $!";
415 defined(my $pid = fork) or die "Can't fork: $!";
417 setsid or die "Can't start a new session: $!";
418 open STDERR, '>&STDOUT' or die "Can't dup stdout: $!";
421 The fork() has to come before the setsid() to ensure that you aren't a
422 process group leader (the setsid() will fail if you are). If your
423 system doesn't have the setsid() function, open F</dev/tty> and use the
424 C<TIOCNOTTY> ioctl() on it instead. See L<tty(4)> for details.
426 Non-Unix users should check their Your_OS::Process module for other
429 =head2 Safe Pipe Opens
431 Another interesting approach to IPC is making your single program go
432 multiprocess and communicate between (or even amongst) yourselves. The
433 open() function will accept a file argument of either C<"-|"> or C<"|-">
434 to do a very interesting thing: it forks a child connected to the
435 filehandle you've opened. The child is running the same program as the
436 parent. This is useful for safely opening a file when running under an
437 assumed UID or GID, for example. If you open a pipe I<to> minus, you can
438 write to the filehandle you opened and your kid will find it in his
439 STDIN. If you open a pipe I<from> minus, you can read from the filehandle
440 you opened whatever your kid writes to his STDOUT.
442 use English '-no_match_vars';
446 $pid = open(KID_TO_WRITE, "|-");
447 unless (defined $pid) {
448 warn "cannot fork: $!";
449 die "bailing out" if $sleep_count++ > 6;
452 } until defined $pid;
455 print KID_TO_WRITE @some_data;
456 close(KID_TO_WRITE) || warn "kid exited $?";
458 ($EUID, $EGID) = ($UID, $GID); # suid progs only
459 open (FILE, "> /safe/file")
460 || die "can't open /safe/file: $!";
462 print FILE; # child's STDIN is parent's KID
464 exit; # don't forget this
467 Another common use for this construct is when you need to execute
468 something without the shell's interference. With system(), it's
469 straightforward, but you can't use a pipe open or backticks safely.
470 That's because there's no way to stop the shell from getting its hands on
471 your arguments. Instead, use lower-level control to call exec() directly.
473 Here's a safe backtick or pipe open for read:
475 # add error processing as above
476 $pid = open(KID_TO_READ, "-|");
479 while (<KID_TO_READ>) {
480 # do something interesting
482 close(KID_TO_READ) || warn "kid exited $?";
485 ($EUID, $EGID) = ($UID, $GID); # suid only
486 exec($program, @options, @args)
487 || die "can't exec program: $!";
492 And here's a safe pipe open for writing:
494 # add error processing as above
495 $pid = open(KID_TO_WRITE, "|-");
496 $SIG{ALRM} = sub { die "whoops, $program pipe broke" };
502 close(KID_TO_WRITE) || warn "kid exited $?";
505 ($EUID, $EGID) = ($UID, $GID);
506 exec($program, @options, @args)
507 || die "can't exec program: $!";
511 Note that these operations are full Unix forks, which means they may not be
512 correctly implemented on alien systems. Additionally, these are not true
513 multithreading. If you'd like to learn more about threading, see the
514 F<modules> file mentioned below in the SEE ALSO section.
516 =head2 Bidirectional Communication with Another Process
518 While this works reasonably well for unidirectional communication, what
519 about bidirectional communication? The obvious thing you'd like to do
520 doesn't actually work:
522 open(PROG_FOR_READING_AND_WRITING, "| some program |")
524 and if you forget to use the C<use warnings> pragma or the B<-w> flag,
525 then you'll miss out entirely on the diagnostic message:
527 Can't do bidirectional pipe at -e line 1.
529 If you really want to, you can use the standard open2() library function
530 to catch both ends. There's also an open3() for tridirectional I/O so you
531 can also catch your child's STDERR, but doing so would then require an
532 awkward select() loop and wouldn't allow you to use normal Perl input
535 If you look at its source, you'll see that open2() uses low-level
536 primitives like Unix pipe() and exec() calls to create all the connections.
537 While it might have been slightly more efficient by using socketpair(), it
538 would have then been even less portable than it already is. The open2()
539 and open3() functions are unlikely to work anywhere except on a Unix
540 system or some other one purporting to be POSIX compliant.
542 Here's an example of using open2():
546 $pid = open2(*Reader, *Writer, "cat -u -n" );
547 print Writer "stuff\n";
550 The problem with this is that Unix buffering is really going to
551 ruin your day. Even though your C<Writer> filehandle is auto-flushed,
552 and the process on the other end will get your data in a timely manner,
553 you can't usually do anything to force it to give it back to you
554 in a similarly quick fashion. In this case, we could, because we
555 gave I<cat> a B<-u> flag to make it unbuffered. But very few Unix
556 commands are designed to operate over pipes, so this seldom works
557 unless you yourself wrote the program on the other end of the
560 A solution to this is the nonstandard F<Comm.pl> library. It uses
561 pseudo-ttys to make your program behave more reasonably:
564 $ph = open_proc('cat -n');
566 print $ph "a line\n";
567 print "got back ", scalar <$ph>;
570 This way you don't have to have control over the source code of the
571 program you're using. The F<Comm> library also has expect()
572 and interact() functions. Find the library (and we hope its
573 successor F<IPC::Chat>) at your nearest CPAN archive as detailed
574 in the SEE ALSO section below.
576 The newer Expect.pm module from CPAN also addresses this kind of thing.
577 This module requires two other modules from CPAN: IO::Pty and IO::Stty.
578 It sets up a pseudo-terminal to interact with programs that insist on
579 using talking to the terminal device driver. If your system is
580 amongst those supported, this may be your best bet.
582 =head2 Bidirectional Communication with Yourself
584 If you want, you may make low-level pipe() and fork()
585 to stitch this together by hand. This example only
586 talks to itself, but you could reopen the appropriate
587 handles to STDIN and STDOUT and call other processes.
590 # pipe1 - bidirectional communication using two pipe pairs
591 # designed for the socketpair-challenged
592 use IO::Handle; # thousands of lines just for autoflush :-(
593 pipe(PARENT_RDR, CHILD_WTR); # XXX: failure?
594 pipe(CHILD_RDR, PARENT_WTR); # XXX: failure?
595 CHILD_WTR->autoflush(1);
596 PARENT_WTR->autoflush(1);
599 close PARENT_RDR; close PARENT_WTR;
600 print CHILD_WTR "Parent Pid $$ is sending this\n";
601 chomp($line = <CHILD_RDR>);
602 print "Parent Pid $$ just read this: `$line'\n";
603 close CHILD_RDR; close CHILD_WTR;
606 die "cannot fork: $!" unless defined $pid;
607 close CHILD_RDR; close CHILD_WTR;
608 chomp($line = <PARENT_RDR>);
609 print "Child Pid $$ just read this: `$line'\n";
610 print PARENT_WTR "Child Pid $$ is sending this\n";
611 close PARENT_RDR; close PARENT_WTR;
615 But you don't actually have to make two pipe calls. If you
616 have the socketpair() system call, it will do this all for you.
619 # pipe2 - bidirectional communication using socketpair
620 # "the best ones always go both ways"
623 use IO::Handle; # thousands of lines just for autoflush :-(
624 # We say AF_UNIX because although *_LOCAL is the
625 # POSIX 1003.1g form of the constant, many machines
626 # still don't have it.
627 socketpair(CHILD, PARENT, AF_UNIX, SOCK_STREAM, PF_UNSPEC)
628 or die "socketpair: $!";
631 PARENT->autoflush(1);
635 print CHILD "Parent Pid $$ is sending this\n";
636 chomp($line = <CHILD>);
637 print "Parent Pid $$ just read this: `$line'\n";
641 die "cannot fork: $!" unless defined $pid;
643 chomp($line = <PARENT>);
644 print "Child Pid $$ just read this: `$line'\n";
645 print PARENT "Child Pid $$ is sending this\n";
650 =head1 Sockets: Client/Server Communication
652 While not limited to Unix-derived operating systems (e.g., WinSock on PCs
653 provides socket support, as do some VMS libraries), you may not have
654 sockets on your system, in which case this section probably isn't going to do
655 you much good. With sockets, you can do both virtual circuits (i.e., TCP
656 streams) and datagrams (i.e., UDP packets). You may be able to do even more
657 depending on your system.
659 The Perl function calls for dealing with sockets have the same names as
660 the corresponding system calls in C, but their arguments tend to differ
661 for two reasons: first, Perl filehandles work differently than C file
662 descriptors. Second, Perl already knows the length of its strings, so you
663 don't need to pass that information.
665 One of the major problems with old socket code in Perl was that it used
666 hard-coded values for some of the constants, which severely hurt
667 portability. If you ever see code that does anything like explicitly
668 setting C<$AF_INET = 2>, you know you're in for big trouble: An
669 immeasurably superior approach is to use the C<Socket> module, which more
670 reliably grants access to various constants and functions you'll need.
672 If you're not writing a server/client for an existing protocol like
673 NNTP or SMTP, you should give some thought to how your server will
674 know when the client has finished talking, and vice-versa. Most
675 protocols are based on one-line messages and responses (so one party
676 knows the other has finished when a "\n" is received) or multi-line
677 messages and responses that end with a period on an empty line
678 ("\n.\n" terminates a message/response).
680 =head2 Internet Line Terminators
682 The Internet line terminator is "\015\012". Under ASCII variants of
683 Unix, that could usually be written as "\r\n", but under other systems,
684 "\r\n" might at times be "\015\015\012", "\012\012\015", or something
685 completely different. The standards specify writing "\015\012" to be
686 conformant (be strict in what you provide), but they also recommend
687 accepting a lone "\012" on input (but be lenient in what you require).
688 We haven't always been very good about that in the code in this manpage,
689 but unless you're on a Mac, you'll probably be ok.
691 =head2 Internet TCP Clients and Servers
693 Use Internet-domain sockets when you want to do client-server
694 communication that might extend to machines outside of your own system.
696 Here's a sample TCP client using Internet-domain sockets:
701 my ($remote,$port, $iaddr, $paddr, $proto, $line);
703 $remote = shift || 'localhost';
704 $port = shift || 2345; # random port
705 if ($port =~ /\D/) { $port = getservbyname($port, 'tcp') }
706 die "No port" unless $port;
707 $iaddr = inet_aton($remote) || die "no host: $remote";
708 $paddr = sockaddr_in($port, $iaddr);
710 $proto = getprotobyname('tcp');
711 socket(SOCK, PF_INET, SOCK_STREAM, $proto) || die "socket: $!";
712 connect(SOCK, $paddr) || die "connect: $!";
713 while (defined($line = <SOCK>)) {
717 close (SOCK) || die "close: $!";
720 And here's a corresponding server to go along with it. We'll
721 leave the address as INADDR_ANY so that the kernel can choose
722 the appropriate interface on multihomed hosts. If you want sit
723 on a particular interface (like the external side of a gateway
724 or firewall machine), you should fill this in with your real address
729 BEGIN { $ENV{PATH} = '/usr/ucb:/bin' }
732 my $EOL = "\015\012";
734 sub logmsg { print "$0 $$: @_ at ", scalar localtime, "\n" }
736 my $port = shift || 2345;
737 my $proto = getprotobyname('tcp');
739 ($port) = $port =~ /^(\d+)$/ or die "invalid port";
741 socket(Server, PF_INET, SOCK_STREAM, $proto) || die "socket: $!";
742 setsockopt(Server, SOL_SOCKET, SO_REUSEADDR,
743 pack("l", 1)) || die "setsockopt: $!";
744 bind(Server, sockaddr_in($port, INADDR_ANY)) || die "bind: $!";
745 listen(Server,SOMAXCONN) || die "listen: $!";
747 logmsg "server started on port $port";
751 $SIG{CHLD} = \&REAPER;
753 for ( ; $paddr = accept(Client,Server); close Client) {
754 my($port,$iaddr) = sockaddr_in($paddr);
755 my $name = gethostbyaddr($iaddr,AF_INET);
757 logmsg "connection from $name [",
758 inet_ntoa($iaddr), "]
761 print Client "Hello there, $name, it's now ",
762 scalar localtime, $EOL;
765 And here's a multithreaded version. It's multithreaded in that
766 like most typical servers, it spawns (forks) a slave server to
767 handle the client request so that the master server can quickly
768 go back to service a new client.
772 BEGIN { $ENV{PATH} = '/usr/ucb:/bin' }
775 my $EOL = "\015\012";
777 sub spawn; # forward declaration
778 sub logmsg { print "$0 $$: @_ at ", scalar localtime, "\n" }
780 my $port = shift || 2345;
781 my $proto = getprotobyname('tcp');
783 ($port) = $port =~ /^(\d+)$/ or die "invalid port";
785 socket(Server, PF_INET, SOCK_STREAM, $proto) || die "socket: $!";
786 setsockopt(Server, SOL_SOCKET, SO_REUSEADDR,
787 pack("l", 1)) || die "setsockopt: $!";
788 bind(Server, sockaddr_in($port, INADDR_ANY)) || die "bind: $!";
789 listen(Server,SOMAXCONN) || die "listen: $!";
791 logmsg "server started on port $port";
796 use POSIX ":sys_wait_h";
799 while (($waitedpid = waitpid(-1,WNOHANG)) > 0) {
800 logmsg "reaped $waitedpid" . ($? ? " with exit $?" : '');
802 $SIG{CHLD} = \&REAPER; # loathe sysV
805 $SIG{CHLD} = \&REAPER;
807 for ( $waitedpid = 0;
808 ($paddr = accept(Client,Server)) || $waitedpid;
809 $waitedpid = 0, close Client)
811 next if $waitedpid and not $paddr;
812 my($port,$iaddr) = sockaddr_in($paddr);
813 my $name = gethostbyaddr($iaddr,AF_INET);
815 logmsg "connection from $name [",
816 inet_ntoa($iaddr), "]
821 print "Hello there, $name, it's now ", scalar localtime, $EOL;
822 exec '/usr/games/fortune' # XXX: `wrong' line terminators
823 or confess "can't exec fortune: $!";
831 unless (@_ == 0 && $coderef && ref($coderef) eq 'CODE') {
832 confess "usage: spawn CODEREF";
836 if (!defined($pid = fork)) {
837 logmsg "cannot fork: $!";
841 return; # I'm the parent
843 # else I'm the child -- go spawn
845 open(STDIN, "<&Client") || die "can't dup client to stdin";
846 open(STDOUT, ">&Client") || die "can't dup client to stdout";
847 ## open(STDERR, ">&STDOUT") || die "can't dup stdout to stderr";
851 This server takes the trouble to clone off a child version via fork() for
852 each incoming request. That way it can handle many requests at once,
853 which you might not always want. Even if you don't fork(), the listen()
854 will allow that many pending connections. Forking servers have to be
855 particularly careful about cleaning up their dead children (called
856 "zombies" in Unix parlance), because otherwise you'll quickly fill up your
859 We suggest that you use the B<-T> flag to use taint checking (see L<perlsec>)
860 even if we aren't running setuid or setgid. This is always a good idea
861 for servers and other programs run on behalf of someone else (like CGI
862 scripts), because it lessens the chances that people from the outside will
863 be able to compromise your system.
865 Let's look at another TCP client. This one connects to the TCP "time"
866 service on a number of different machines and shows how far their clocks
867 differ from the system on which it's being run:
873 my $SECS_of_70_YEARS = 2208988800;
874 sub ctime { scalar localtime(shift) }
876 my $iaddr = gethostbyname('localhost');
877 my $proto = getprotobyname('tcp');
878 my $port = getservbyname('time', 'tcp');
879 my $paddr = sockaddr_in(0, $iaddr);
883 printf "%-24s %8s %s\n", "localhost", 0, ctime(time());
885 foreach $host (@ARGV) {
886 printf "%-24s ", $host;
887 my $hisiaddr = inet_aton($host) || die "unknown host";
888 my $hispaddr = sockaddr_in($port, $hisiaddr);
889 socket(SOCKET, PF_INET, SOCK_STREAM, $proto) || die "socket: $!";
890 connect(SOCKET, $hispaddr) || die "bind: $!";
892 read(SOCKET, $rtime, 4);
894 my $histime = unpack("N", $rtime) - $SECS_of_70_YEARS ;
895 printf "%8d %s\n", $histime - time, ctime($histime);
898 =head2 Unix-Domain TCP Clients and Servers
900 That's fine for Internet-domain clients and servers, but what about local
901 communications? While you can use the same setup, sometimes you don't
902 want to. Unix-domain sockets are local to the current host, and are often
903 used internally to implement pipes. Unlike Internet domain sockets, Unix
904 domain sockets can show up in the file system with an ls(1) listing.
907 srw-rw-rw- 1 root 0 Oct 31 07:23 /dev/log
909 You can test for these with Perl's B<-S> file test:
911 unless ( -S '/dev/log' ) {
912 die "something's wicked with the log system";
915 Here's a sample Unix-domain client:
920 my ($rendezvous, $line);
922 $rendezvous = shift || '/tmp/catsock';
923 socket(SOCK, PF_UNIX, SOCK_STREAM, 0) || die "socket: $!";
924 connect(SOCK, sockaddr_un($rendezvous)) || die "connect: $!";
925 while (defined($line = <SOCK>)) {
930 And here's a corresponding server. You don't have to worry about silly
931 network terminators here because Unix domain sockets are guaranteed
932 to be on the localhost, and thus everything works right.
939 BEGIN { $ENV{PATH} = '/usr/ucb:/bin' }
940 sub spawn; # forward declaration
941 sub logmsg { print "$0 $$: @_ at ", scalar localtime, "\n" }
943 my $NAME = '/tmp/catsock';
944 my $uaddr = sockaddr_un($NAME);
945 my $proto = getprotobyname('tcp');
947 socket(Server,PF_UNIX,SOCK_STREAM,0) || die "socket: $!";
949 bind (Server, $uaddr) || die "bind: $!";
950 listen(Server,SOMAXCONN) || die "listen: $!";
952 logmsg "server started on $NAME";
956 use POSIX ":sys_wait_h";
959 while (($waitedpid = waitpid(-1,WNOHANG)) > 0) {
960 logmsg "reaped $waitedpid" . ($? ? " with exit $?" : '');
962 $SIG{CHLD} = \&REAPER; # loathe sysV
965 $SIG{CHLD} = \&REAPER;
968 for ( $waitedpid = 0;
969 accept(Client,Server) || $waitedpid;
970 $waitedpid = 0, close Client)
973 logmsg "connection on $NAME";
975 print "Hello there, it's now ", scalar localtime, "\n";
976 exec '/usr/games/fortune' or die "can't exec fortune: $!";
983 unless (@_ == 0 && $coderef && ref($coderef) eq 'CODE') {
984 confess "usage: spawn CODEREF";
988 if (!defined($pid = fork)) {
989 logmsg "cannot fork: $!";
993 return; # I'm the parent
995 # else I'm the child -- go spawn
997 open(STDIN, "<&Client") || die "can't dup client to stdin";
998 open(STDOUT, ">&Client") || die "can't dup client to stdout";
999 ## open(STDERR, ">&STDOUT") || die "can't dup stdout to stderr";
1003 As you see, it's remarkably similar to the Internet domain TCP server, so
1004 much so, in fact, that we've omitted several duplicate functions--spawn(),
1005 logmsg(), ctime(), and REAPER()--which are exactly the same as in the
1008 So why would you ever want to use a Unix domain socket instead of a
1009 simpler named pipe? Because a named pipe doesn't give you sessions. You
1010 can't tell one process's data from another's. With socket programming,
1011 you get a separate session for each client: that's why accept() takes two
1014 For example, let's say that you have a long running database server daemon
1015 that you want folks from the World Wide Web to be able to access, but only
1016 if they go through a CGI interface. You'd have a small, simple CGI
1017 program that does whatever checks and logging you feel like, and then acts
1018 as a Unix-domain client and connects to your private server.
1020 =head1 TCP Clients with IO::Socket
1022 For those preferring a higher-level interface to socket programming, the
1023 IO::Socket module provides an object-oriented approach. IO::Socket is
1024 included as part of the standard Perl distribution as of the 5.004
1025 release. If you're running an earlier version of Perl, just fetch
1026 IO::Socket from CPAN, where you'll also find modules providing easy
1027 interfaces to the following systems: DNS, FTP, Ident (RFC 931), NIS and
1028 NISPlus, NNTP, Ping, POP3, SMTP, SNMP, SSLeay, Telnet, and Time--just
1031 =head2 A Simple Client
1033 Here's a client that creates a TCP connection to the "daytime"
1034 service at port 13 of the host name "localhost" and prints out everything
1035 that the server there cares to provide.
1039 $remote = IO::Socket::INET->new(
1041 PeerAddr => "localhost",
1042 PeerPort => "daytime(13)",
1044 or die "cannot connect to daytime port at localhost";
1045 while ( <$remote> ) { print }
1047 When you run this program, you should get something back that
1050 Wed May 14 08:40:46 MDT 1997
1052 Here are what those parameters to the C<new> constructor mean:
1058 This is which protocol to use. In this case, the socket handle returned
1059 will be connected to a TCP socket, because we want a stream-oriented
1060 connection, that is, one that acts pretty much like a plain old file.
1061 Not all sockets are this of this type. For example, the UDP protocol
1062 can be used to make a datagram socket, used for message-passing.
1066 This is the name or Internet address of the remote host the server is
1067 running on. We could have specified a longer name like C<"www.perl.com">,
1068 or an address like C<"204.148.40.9">. For demonstration purposes, we've
1069 used the special hostname C<"localhost">, which should always mean the
1070 current machine you're running on. The corresponding Internet address
1071 for localhost is C<"127.1">, if you'd rather use that.
1075 This is the service name or port number we'd like to connect to.
1076 We could have gotten away with using just C<"daytime"> on systems with a
1077 well-configured system services file,[FOOTNOTE: The system services file
1078 is in I</etc/services> under Unix] but just in case, we've specified the
1079 port number (13) in parentheses. Using just the number would also have
1080 worked, but constant numbers make careful programmers nervous.
1084 Notice how the return value from the C<new> constructor is used as
1085 a filehandle in the C<while> loop? That's what's called an indirect
1086 filehandle, a scalar variable containing a filehandle. You can use
1087 it the same way you would a normal filehandle. For example, you
1088 can read one line from it this way:
1092 all remaining lines from is this way:
1096 and send a line of data to it this way:
1098 print $handle "some data\n";
1100 =head2 A Webget Client
1102 Here's a simple client that takes a remote host to fetch a document
1103 from, and then a list of documents to get from that host. This is a
1104 more interesting client than the previous one because it first sends
1105 something to the server before fetching the server's response.
1109 unless (@ARGV > 1) { die "usage: $0 host document ..." }
1110 $host = shift(@ARGV);
1113 foreach $document ( @ARGV ) {
1114 $remote = IO::Socket::INET->new( Proto => "tcp",
1116 PeerPort => "http(80)",
1118 unless ($remote) { die "cannot connect to http daemon on $host" }
1119 $remote->autoflush(1);
1120 print $remote "GET $document HTTP/1.0" . $BLANK;
1121 while ( <$remote> ) { print }
1125 The web server handing the "http" service, which is assumed to be at
1126 its standard port, number 80. If the web server you're trying to
1127 connect to is at a different port (like 1080 or 8080), you should specify
1128 as the named-parameter pair, C<< PeerPort => 8080 >>. The C<autoflush>
1129 method is used on the socket because otherwise the system would buffer
1130 up the output we sent it. (If you're on a Mac, you'll also need to
1131 change every C<"\n"> in your code that sends data over the network to
1132 be a C<"\015\012"> instead.)
1134 Connecting to the server is only the first part of the process: once you
1135 have the connection, you have to use the server's language. Each server
1136 on the network has its own little command language that it expects as
1137 input. The string that we send to the server starting with "GET" is in
1138 HTTP syntax. In this case, we simply request each specified document.
1139 Yes, we really are making a new connection for each document, even though
1140 it's the same host. That's the way you always used to have to speak HTTP.
1141 Recent versions of web browsers may request that the remote server leave
1142 the connection open a little while, but the server doesn't have to honor
1145 Here's an example of running that program, which we'll call I<webget>:
1147 % webget www.perl.com /guanaco.html
1148 HTTP/1.1 404 File Not Found
1149 Date: Thu, 08 May 1997 18:02:32 GMT
1150 Server: Apache/1.2b6
1152 Content-type: text/html
1154 <HEAD><TITLE>404 File Not Found</TITLE></HEAD>
1155 <BODY><H1>File Not Found</H1>
1156 The requested URL /guanaco.html was not found on this server.<P>
1159 Ok, so that's not very interesting, because it didn't find that
1160 particular document. But a long response wouldn't have fit on this page.
1162 For a more fully-featured version of this program, you should look to
1163 the I<lwp-request> program included with the LWP modules from CPAN.
1165 =head2 Interactive Client with IO::Socket
1167 Well, that's all fine if you want to send one command and get one answer,
1168 but what about setting up something fully interactive, somewhat like
1169 the way I<telnet> works? That way you can type a line, get the answer,
1170 type a line, get the answer, etc.
1172 This client is more complicated than the two we've done so far, but if
1173 you're on a system that supports the powerful C<fork> call, the solution
1174 isn't that rough. Once you've made the connection to whatever service
1175 you'd like to chat with, call C<fork> to clone your process. Each of
1176 these two identical process has a very simple job to do: the parent
1177 copies everything from the socket to standard output, while the child
1178 simultaneously copies everything from standard input to the socket.
1179 To accomplish the same thing using just one process would be I<much>
1180 harder, because it's easier to code two processes to do one thing than it
1181 is to code one process to do two things. (This keep-it-simple principle
1182 a cornerstones of the Unix philosophy, and good software engineering as
1183 well, which is probably why it's spread to other systems.)
1190 my ($host, $port, $kidpid, $handle, $line);
1192 unless (@ARGV == 2) { die "usage: $0 host port" }
1193 ($host, $port) = @ARGV;
1195 # create a tcp connection to the specified host and port
1196 $handle = IO::Socket::INET->new(Proto => "tcp",
1199 or die "can't connect to port $port on $host: $!";
1201 $handle->autoflush(1); # so output gets there right away
1202 print STDERR "[Connected to $host:$port]\n";
1204 # split the program into two processes, identical twins
1205 die "can't fork: $!" unless defined($kidpid = fork());
1207 # the if{} block runs only in the parent process
1209 # copy the socket to standard output
1210 while (defined ($line = <$handle>)) {
1213 kill("TERM", $kidpid); # send SIGTERM to child
1215 # the else{} block runs only in the child process
1217 # copy standard input to the socket
1218 while (defined ($line = <STDIN>)) {
1219 print $handle $line;
1223 The C<kill> function in the parent's C<if> block is there to send a
1224 signal to our child process (current running in the C<else> block)
1225 as soon as the remote server has closed its end of the connection.
1227 If the remote server sends data a byte at time, and you need that
1228 data immediately without waiting for a newline (which might not happen),
1229 you may wish to replace the C<while> loop in the parent with the
1233 while (sysread($handle, $byte, 1) == 1) {
1237 Making a system call for each byte you want to read is not very efficient
1238 (to put it mildly) but is the simplest to explain and works reasonably
1241 =head1 TCP Servers with IO::Socket
1243 As always, setting up a server is little bit more involved than running a client.
1244 The model is that the server creates a special kind of socket that
1245 does nothing but listen on a particular port for incoming connections.
1246 It does this by calling the C<< IO::Socket::INET->new() >> method with
1247 slightly different arguments than the client did.
1253 This is which protocol to use. Like our clients, we'll
1254 still specify C<"tcp"> here.
1259 port in the C<LocalPort> argument, which we didn't do for the client.
1260 This is service name or port number for which you want to be the
1261 server. (Under Unix, ports under 1024 are restricted to the
1262 superuser.) In our sample, we'll use port 9000, but you can use
1263 any port that's not currently in use on your system. If you try
1264 to use one already in used, you'll get an "Address already in use"
1265 message. Under Unix, the C<netstat -a> command will show
1266 which services current have servers.
1270 The C<Listen> parameter is set to the maximum number of
1271 pending connections we can accept until we turn away incoming clients.
1272 Think of it as a call-waiting queue for your telephone.
1273 The low-level Socket module has a special symbol for the system maximum, which
1278 The C<Reuse> parameter is needed so that we restart our server
1279 manually without waiting a few minutes to allow system buffers to
1284 Once the generic server socket has been created using the parameters
1285 listed above, the server then waits for a new client to connect
1286 to it. The server blocks in the C<accept> method, which eventually accepts a
1287 bidirectional connection from the remote client. (Make sure to autoflush
1288 this handle to circumvent buffering.)
1290 To add to user-friendliness, our server prompts the user for commands.
1291 Most servers don't do this. Because of the prompt without a newline,
1292 you'll have to use the C<sysread> variant of the interactive client above.
1294 This server accepts one of five different commands, sending output
1295 back to the client. Note that unlike most network servers, this one
1296 only handles one incoming client at a time. Multithreaded servers are
1297 covered in Chapter 6 of the Camel.
1299 Here's the code. We'll
1303 use Net::hostent; # for OO version of gethostbyaddr
1305 $PORT = 9000; # pick something not in use
1307 $server = IO::Socket::INET->new( Proto => 'tcp',
1309 Listen => SOMAXCONN,
1312 die "can't setup server" unless $server;
1313 print "[Server $0 accepting clients]\n";
1315 while ($client = $server->accept()) {
1316 $client->autoflush(1);
1317 print $client "Welcome to $0; type help for command list.\n";
1318 $hostinfo = gethostbyaddr($client->peeraddr);
1319 printf "[Connect from %s]\n", $hostinfo->name || $client->peerhost;
1320 print $client "Command? ";
1321 while ( <$client>) {
1322 next unless /\S/; # blank line
1323 if (/quit|exit/i) { last; }
1324 elsif (/date|time/i) { printf $client "%s\n", scalar localtime; }
1325 elsif (/who/i ) { print $client `who 2>&1`; }
1326 elsif (/cookie/i ) { print $client `/usr/games/fortune 2>&1`; }
1327 elsif (/motd/i ) { print $client `cat /etc/motd 2>&1`; }
1329 print $client "Commands: quit date who cookie motd\n";
1332 print $client "Command? ";
1337 =head1 UDP: Message Passing
1339 Another kind of client-server setup is one that uses not connections, but
1340 messages. UDP communications involve much lower overhead but also provide
1341 less reliability, as there are no promises that messages will arrive at
1342 all, let alone in order and unmangled. Still, UDP offers some advantages
1343 over TCP, including being able to "broadcast" or "multicast" to a whole
1344 bunch of destination hosts at once (usually on your local subnet). If you
1345 find yourself overly concerned about reliability and start building checks
1346 into your message system, then you probably should use just TCP to start
1349 Note that UDP datagrams are I<not> a bytestream and should not be treated
1350 as such. This makes using I/O mechanisms with internal buffering
1351 like stdio (i.e. print() and friends) especially cumbersome. Use syswrite(),
1352 or better send(), like in the example below.
1354 Here's a UDP program similar to the sample Internet TCP client given
1355 earlier. However, instead of checking one host at a time, the UDP version
1356 will check many of them asynchronously by simulating a multicast and then
1357 using select() to do a timed-out wait for I/O. To do something similar
1358 with TCP, you'd have to use a different socket handle for each host.
1365 my ( $count, $hisiaddr, $hispaddr, $histime,
1366 $host, $iaddr, $paddr, $port, $proto,
1367 $rin, $rout, $rtime, $SECS_of_70_YEARS);
1369 $SECS_of_70_YEARS = 2208988800;
1371 $iaddr = gethostbyname(hostname());
1372 $proto = getprotobyname('udp');
1373 $port = getservbyname('time', 'udp');
1374 $paddr = sockaddr_in(0, $iaddr); # 0 means let kernel pick
1376 socket(SOCKET, PF_INET, SOCK_DGRAM, $proto) || die "socket: $!";
1377 bind(SOCKET, $paddr) || die "bind: $!";
1380 printf "%-12s %8s %s\n", "localhost", 0, scalar localtime time;
1384 $hisiaddr = inet_aton($host) || die "unknown host";
1385 $hispaddr = sockaddr_in($port, $hisiaddr);
1386 defined(send(SOCKET, 0, 0, $hispaddr)) || die "send $host: $!";
1390 vec($rin, fileno(SOCKET), 1) = 1;
1392 # timeout after 10.0 seconds
1393 while ($count && select($rout = $rin, undef, undef, 10.0)) {
1395 ($hispaddr = recv(SOCKET, $rtime, 4, 0)) || die "recv: $!";
1396 ($port, $hisiaddr) = sockaddr_in($hispaddr);
1397 $host = gethostbyaddr($hisiaddr, AF_INET);
1398 $histime = unpack("N", $rtime) - $SECS_of_70_YEARS ;
1399 printf "%-12s ", $host;
1400 printf "%8d %s\n", $histime - time, scalar localtime($histime);
1404 Note that this example does not include any retries and may consequently
1405 fail to contact a reachable host. The most prominent reason for this
1406 is congestion of the queues on the sending host if the number of
1407 list of hosts to contact is sufficiently large.
1411 While System V IPC isn't so widely used as sockets, it still has some
1412 interesting uses. You can't, however, effectively use SysV IPC or
1413 Berkeley mmap() to have shared memory so as to share a variable amongst
1414 several processes. That's because Perl would reallocate your string when
1415 you weren't wanting it to.
1417 Here's a small example showing shared memory usage.
1419 use IPC::SysV qw(IPC_PRIVATE IPC_RMID S_IRWXU);
1422 $id = shmget(IPC_PRIVATE, $size, S_IRWXU) || die "$!";
1423 print "shm key $id\n";
1425 $message = "Message #1";
1426 shmwrite($id, $message, 0, 60) || die "$!";
1427 print "wrote: '$message'\n";
1428 shmread($id, $buff, 0, 60) || die "$!";
1429 print "read : '$buff'\n";
1431 # the buffer of shmread is zero-character end-padded.
1432 substr($buff, index($buff, "\0")) = '';
1433 print "un" unless $buff eq $message;
1436 print "deleting shm $id\n";
1437 shmctl($id, IPC_RMID, 0) || die "$!";
1439 Here's an example of a semaphore:
1441 use IPC::SysV qw(IPC_CREAT);
1444 $id = semget($IPC_KEY, 10, 0666 | IPC_CREAT ) || die "$!";
1445 print "shm key $id\n";
1447 Put this code in a separate file to be run in more than one process.
1448 Call the file F<take>:
1450 # create a semaphore
1453 $id = semget($IPC_KEY, 0 , 0 );
1454 die if !defined($id);
1460 # wait for semaphore to be zero
1462 $opstring1 = pack("s!s!s!", $semnum, $semop, $semflag);
1464 # Increment the semaphore count
1466 $opstring2 = pack("s!s!s!", $semnum, $semop, $semflag);
1467 $opstring = $opstring1 . $opstring2;
1469 semop($id,$opstring) || die "$!";
1471 Put this code in a separate file to be run in more than one process.
1472 Call this file F<give>:
1474 # 'give' the semaphore
1475 # run this in the original process and you will see
1476 # that the second process continues
1479 $id = semget($IPC_KEY, 0, 0);
1480 die if !defined($id);
1485 # Decrement the semaphore count
1487 $opstring = pack("s!s!s!", $semnum, $semop, $semflag);
1489 semop($id,$opstring) || die "$!";
1491 The SysV IPC code above was written long ago, and it's definitely
1492 clunky looking. For a more modern look, see the IPC::SysV module
1493 which is included with Perl starting from Perl 5.005.
1495 A small example demonstrating SysV message queues:
1497 use IPC::SysV qw(IPC_PRIVATE IPC_RMID IPC_CREAT S_IRWXU);
1499 my $id = msgget(IPC_PRIVATE, IPC_CREAT | S_IRWXU);
1501 my $sent = "message";
1507 if (msgsnd($id, pack("l! a*", $type_sent, $sent), 0)) {
1508 if (msgrcv($id, $rcvd, 60, 0, 0)) {
1509 ($type_rcvd, $rcvd) = unpack("l! a*", $rcvd);
1510 if ($rcvd eq $sent) {
1516 die "# msgrcv failed\n";
1519 die "# msgsnd failed\n";
1521 msgctl($id, IPC_RMID, 0) || die "# msgctl failed: $!\n";
1523 die "# msgget failed\n";
1528 Most of these routines quietly but politely return C<undef> when they
1529 fail instead of causing your program to die right then and there due to
1530 an uncaught exception. (Actually, some of the new I<Socket> conversion
1531 functions croak() on bad arguments.) It is therefore essential to
1532 check return values from these functions. Always begin your socket
1533 programs this way for optimal success, and don't forget to add B<-T>
1534 taint checking flag to the #! line for servers:
1543 All these routines create system-specific portability problems. As noted
1544 elsewhere, Perl is at the mercy of your C libraries for much of its system
1545 behaviour. It's probably safest to assume broken SysV semantics for
1546 signals and to stick with simple TCP and UDP socket operations; e.g., don't
1547 try to pass open file descriptors over a local UDP datagram socket if you
1548 want your code to stand a chance of being portable.
1550 As mentioned in the signals section, because few vendors provide C
1551 libraries that are safely re-entrant, the prudent programmer will do
1552 little else within a handler beyond setting a numeric variable that
1553 already exists; or, if locked into a slow (restarting) system call,
1554 using die() to raise an exception and longjmp(3) out. In fact, even
1555 these may in some cases cause a core dump. It's probably best to avoid
1556 signals except where they are absolutely inevitable. This
1557 will be addressed in a future release of Perl.
1561 Tom Christiansen, with occasional vestiges of Larry Wall's original
1562 version and suggestions from the Perl Porters.
1566 There's a lot more to networking than this, but this should get you
1569 For intrepid programmers, the indispensable textbook is I<Unix Network
1570 Programming> by W. Richard Stevens (published by Addison-Wesley). Note
1571 that most books on networking address networking from the perspective of
1572 a C programmer; translation to Perl is left as an exercise for the reader.
1574 The IO::Socket(3) manpage describes the object library, and the Socket(3)
1575 manpage describes the low-level interface to sockets. Besides the obvious
1576 functions in L<perlfunc>, you should also check out the F<modules> file
1577 at your nearest CPAN site. (See L<perlmodlib> or best yet, the F<Perl
1578 FAQ> for a description of what CPAN is and where to get it.)
1580 Section 5 of the F<modules> file is devoted to "Networking, Device Control
1581 (modems), and Interprocess Communication", and contains numerous unbundled
1582 modules numerous networking modules, Chat and Expect operations, CGI
1583 programming, DCE, FTP, IPC, NNTP, Proxy, Ptty, RPC, SNMP, SMTP, Telnet,
1584 Threads, and ToolTalk--just to name a few.