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 (Safe 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
171 =head2 Handling the SIGHUP Signal in Daemons
173 A process that usually starts when the system boots and shuts down
174 when the system is shut down is called a daemon (Disk And Execution
175 MONitor). If a daemon process has a configuration file which is
176 modified after the process has been started, there should be a way to
177 tell that process to re-read its configuration file, without stopping
178 the process. Many daemons provide this mechanism using the C<SIGHUP>
179 signal handler. When you want to tell the daemon to re-read the file
180 you simply send it the C<SIGHUP> signal.
182 Not all platforms automatically reinstall their (native) signal
183 handlers after a signal delivery. This means that the handler works
184 only the first time the signal is sent. The solution to this problem
185 is to use C<POSIX> signal handlers if available, their behaviour
188 The following example implements a simple daemon, which restarts
189 itself every time the C<SIGHUP> signal is received. The actual code is
190 located in the subroutine C<code()>, which simply prints some debug
191 info to show that it works and should be replaced with the real code.
197 use File::Basename ();
198 use File::Spec::Functions;
202 # make the daemon cross-platform, so exec always calls the script
203 # itself with the right path, no matter how the script was invoked.
204 my $script = File::Basename::basename($0);
205 my $SELF = catfile $FindBin::Bin, $script;
207 # POSIX unmasks the sigprocmask properly
208 my $sigset = POSIX::SigSet->new();
209 my $action = POSIX::SigAction->new('sigHUP_handler',
212 POSIX::sigaction(&POSIX::SIGHUP, $action);
215 print "got SIGHUP\n";
216 exec($SELF, @ARGV) or die "Couldn't restart: $!\n";
223 print "ARGV: @ARGV\n";
235 A named pipe (often referred to as a FIFO) is an old Unix IPC
236 mechanism for processes communicating on the same machine. It works
237 just like a regular, connected anonymous pipes, except that the
238 processes rendezvous using a filename and don't have to be related.
240 To create a named pipe, use the Unix command mknod(1) or on some
241 systems, mkfifo(1). These may not be in your normal path.
243 # system return val is backwards, so && not ||
245 $ENV{PATH} .= ":/etc:/usr/etc";
246 if ( system('mknod', $path, 'p')
247 && system('mkfifo', $path) )
249 die "mk{nod,fifo} $path failed";
253 A fifo is convenient when you want to connect a process to an unrelated
254 one. When you open a fifo, the program will block until there's something
257 For example, let's say you'd like to have your F<.signature> file be a
258 named pipe that has a Perl program on the other end. Now every time any
259 program (like a mailer, news reader, finger program, etc.) tries to read
260 from that file, the reading program will block and your program will
261 supply the new signature. We'll use the pipe-checking file test B<-p>
262 to find out whether anyone (or anything) has accidentally removed our fifo.
265 $FIFO = '.signature';
266 $ENV{PATH} .= ":/etc:/usr/games";
271 system('mknod', $FIFO, 'p')
272 && die "can't mknod $FIFO: $!";
275 # next line blocks until there's a reader
276 open (FIFO, "> $FIFO") || die "can't write $FIFO: $!";
277 print FIFO "John Smith (smith\@host.org)\n", `fortune -s`;
279 sleep 2; # to avoid dup signals
282 =head2 Deferred Signals (Safe signals)
284 In Perls before Perl 5.7.3 by installing Perl code to deal with
285 signals, you were exposing yourself to danger from two things. First,
286 few system library functions are re-entrant. If the signal interrupts
287 while Perl is executing one function (like malloc(3) or printf(3)),
288 and your signal handler then calls the same function again, you could
289 get unpredictable behavior--often, a core dump. Second, Perl isn't
290 itself re-entrant at the lowest levels. If the signal interrupts Perl
291 while Perl is changing its own internal data structures, similarly
292 unpredictable behaviour may result.
294 There were two things you could do, knowing this: be paranoid or be
295 pragmatic. The paranoid approach was to do as little as possible in your
296 signal handler. Set an existing integer variable that already has a
297 value, and return. This doesn't help you if you're in a slow system call,
298 which will just restart. That means you have to C<die> to longjump(3) out
299 of the handler. Even this is a little cavalier for the true paranoiac,
300 who avoids C<die> in a handler because the system I<is> out to get you.
301 The pragmatic approach was to say ``I know the risks, but prefer the
302 convenience'', and to do anything you wanted in your signal handler,
303 and be prepared to clean up core dumps now and again.
305 In Perl 5.7.3 and later to avoid these problems signals are
306 "deferred"-- that is when the signal is delivered to the process by
307 the system (to the C code that implements Perl) a flag is set, and the
308 handler returns immediately. Then at strategic "safe" points in the
309 Perl interpreter (e.g. when it is about to execute a new opcode) the
310 flags are checked and the Perl level handler from %SIG is
311 executed. The "deferred" scheme allows much more flexibility in the
312 coding of signal handler as we know Perl interpreter is in a safe
313 state, and that we are not in a system library function when the
314 handler is called. However the implementation does differ from
315 previous Perls in the following ways:
319 =item Long running opcodes
321 As Perl interpreter only looks at the signal flags when it about to
322 execute a new opcode if a signal arrives during a long running opcode
323 (e.g. a regular expression operation on a very large string) then
324 signal will not be seen until operation completes.
326 =item Interrupting IO
328 When a signal is delivered (e.g. INT control-C) the operating system
329 breaks into IO operations like C<read> (used to implement Perls
330 E<lt>E<gt> operator). On older Perls the handler was called
331 immediately (and as C<read> is not "unsafe" this worked well). With
332 the "deferred" scheme the handler is not called immediately, and if
333 Perl is using system's C<stdio> library that library may re-start the
334 C<read> without returning to Perl and giving it a chance to call the
335 %SIG handler. If this happens on your system the solution is to use
336 C<:perlio> layer to do IO - at least on those handles which you want
337 to be able to break into with signals. (The C<:perlio> layer checks
338 the signal flags and calls %SIG handlers before resuming IO operation.)
340 Note that the default in Perl 5.7.3 and later is to automatically use
341 the C<:perlio> layer.
343 Note that some networking library functions like gethostbyname() are
344 known to have their own implementations of timeouts which may conflict
345 with your timeouts. If you are having problems with such functions,
346 you can try using the POSIX sigaction() function, which bypasses the
347 Perl safe signals (note that this means subjecting yourself to
348 possible memory corruption, as described above). Instead of setting
349 C<$SIG{ALRM}> try something like the following:
352 sigaction SIGALRM, new POSIX::SigAction sub { die "alarm\n" }
353 or die "Error setting SIGALRM handler: $!\n";
355 =item Restartable system calls
357 On systems that supported it, older versions of Perl used the
358 SA_RESTART flag when installing %SIG handlers. This meant that
359 restartable system calls would continue rather than returning when
360 a signal arrived. In order to deliver deferred signals promptly,
361 Perl 5.7.3 and later do I<not> use SA_RESTART. Consequently,
362 restartable system calls can fail (with $! set to C<EINTR>) in places
363 where they previously would have succeeded.
365 Note that the default C<:perlio> layer will retry C<read>, C<write>
366 and C<close> as described above and that interrupted C<wait> and
367 C<waitpid> calls will always be retried.
369 =item Signals as "faults"
371 Certain signals e.g. SEGV, ILL, BUS are generated as a result of
372 virtual memory or other "faults". These are normally fatal and there
373 is little a Perl-level handler can do with them. (In particular the
374 old signal scheme was particularly unsafe in such cases.) However if
375 a %SIG handler is set the new scheme simply sets a flag and returns as
376 described above. This may cause the operating system to try the
377 offending machine instruction again and - as nothing has changed - it
378 will generate the signal again. The result of this is a rather odd
379 "loop". In future Perl's signal mechanism may be changed to avoid this
380 - perhaps by simply disallowing %SIG handlers on signals of that
381 type. Until then the work-round is not to set a %SIG handler on those
382 signals. (Which signals they are is operating system dependant.)
384 =item Signals triggered by operating system state
386 On some operating systems certain signal handlers are supposed to "do
387 something" before returning. One example can be CHLD or CLD which
388 indicates a child process has completed. On some operating systems the
389 signal handler is expected to C<wait> for the completed child
390 process. On such systems the deferred signal scheme will not work for
391 those signals (it does not do the C<wait>). Again the failure will
392 look like a loop as the operating system will re-issue the signal as
393 there are un-waited-for completed child processes.
397 If you want the old signal behaviour back regardless of possible
398 memory corruption, set the environment variable C<PERL_SIGNALS> to
399 C<"unsafe"> (a new feature since Perl 5.8.1).
401 =head1 Using open() for IPC
403 Perl's basic open() statement can also be used for unidirectional
404 interprocess communication by either appending or prepending a pipe
405 symbol to the second argument to open(). Here's how to start
406 something up in a child process you intend to write to:
408 open(SPOOLER, "| cat -v | lpr -h 2>/dev/null")
409 || die "can't fork: $!";
410 local $SIG{PIPE} = sub { die "spooler pipe broke" };
411 print SPOOLER "stuff\n";
412 close SPOOLER || die "bad spool: $! $?";
414 And here's how to start up a child process you intend to read from:
416 open(STATUS, "netstat -an 2>&1 |")
417 || die "can't fork: $!";
419 next if /^(tcp|udp)/;
422 close STATUS || die "bad netstat: $! $?";
424 If one can be sure that a particular program is a Perl script that is
425 expecting filenames in @ARGV, the clever programmer can write something
428 % program f1 "cmd1|" - f2 "cmd2|" f3 < tmpfile
430 and irrespective of which shell it's called from, the Perl program will
431 read from the file F<f1>, the process F<cmd1>, standard input (F<tmpfile>
432 in this case), the F<f2> file, the F<cmd2> command, and finally the F<f3>
433 file. Pretty nifty, eh?
435 You might notice that you could use backticks for much the
436 same effect as opening a pipe for reading:
438 print grep { !/^(tcp|udp)/ } `netstat -an 2>&1`;
439 die "bad netstat" if $?;
441 While this is true on the surface, it's much more efficient to process the
442 file one line or record at a time because then you don't have to read the
443 whole thing into memory at once. It also gives you finer control of the
444 whole process, letting you to kill off the child process early if you'd
447 Be careful to check both the open() and the close() return values. If
448 you're I<writing> to a pipe, you should also trap SIGPIPE. Otherwise,
449 think of what happens when you start up a pipe to a command that doesn't
450 exist: the open() will in all likelihood succeed (it only reflects the
451 fork()'s success), but then your output will fail--spectacularly. Perl
452 can't know whether the command worked because your command is actually
453 running in a separate process whose exec() might have failed. Therefore,
454 while readers of bogus commands return just a quick end of file, writers
455 to bogus command will trigger a signal they'd better be prepared to
458 open(FH, "|bogus") or die "can't fork: $!";
459 print FH "bang\n" or die "can't write: $!";
460 close FH or die "can't close: $!";
462 That won't blow up until the close, and it will blow up with a SIGPIPE.
463 To catch it, you could use this:
465 $SIG{PIPE} = 'IGNORE';
466 open(FH, "|bogus") or die "can't fork: $!";
467 print FH "bang\n" or die "can't write: $!";
468 close FH or die "can't close: status=$?";
472 Both the main process and any child processes it forks share the same
473 STDIN, STDOUT, and STDERR filehandles. If both processes try to access
474 them at once, strange things can happen. You may also want to close
475 or reopen the filehandles for the child. You can get around this by
476 opening your pipe with open(), but on some systems this means that the
477 child process cannot outlive the parent.
479 =head2 Background Processes
481 You can run a command in the background with:
485 The command's STDOUT and STDERR (and possibly STDIN, depending on your
486 shell) will be the same as the parent's. You won't need to catch
487 SIGCHLD because of the double-fork taking place (see below for more
490 =head2 Complete Dissociation of Child from Parent
492 In some cases (starting server processes, for instance) you'll want to
493 completely dissociate the child process from the parent. This is
494 often called daemonization. A well behaved daemon will also chdir()
495 to the root directory (so it doesn't prevent unmounting the filesystem
496 containing the directory from which it was launched) and redirect its
497 standard file descriptors from and to F</dev/null> (so that random
498 output doesn't wind up on the user's terminal).
503 chdir '/' or die "Can't chdir to /: $!";
504 open STDIN, '/dev/null' or die "Can't read /dev/null: $!";
505 open STDOUT, '>/dev/null'
506 or die "Can't write to /dev/null: $!";
507 defined(my $pid = fork) or die "Can't fork: $!";
509 setsid or die "Can't start a new session: $!";
510 open STDERR, '>&STDOUT' or die "Can't dup stdout: $!";
513 The fork() has to come before the setsid() to ensure that you aren't a
514 process group leader (the setsid() will fail if you are). If your
515 system doesn't have the setsid() function, open F</dev/tty> and use the
516 C<TIOCNOTTY> ioctl() on it instead. See L<tty(4)> for details.
518 Non-Unix users should check their Your_OS::Process module for other
521 =head2 Safe Pipe Opens
523 Another interesting approach to IPC is making your single program go
524 multiprocess and communicate between (or even amongst) yourselves. The
525 open() function will accept a file argument of either C<"-|"> or C<"|-">
526 to do a very interesting thing: it forks a child connected to the
527 filehandle you've opened. The child is running the same program as the
528 parent. This is useful for safely opening a file when running under an
529 assumed UID or GID, for example. If you open a pipe I<to> minus, you can
530 write to the filehandle you opened and your kid will find it in his
531 STDIN. If you open a pipe I<from> minus, you can read from the filehandle
532 you opened whatever your kid writes to his STDOUT.
534 use English '-no_match_vars';
538 $pid = open(KID_TO_WRITE, "|-");
539 unless (defined $pid) {
540 warn "cannot fork: $!";
541 die "bailing out" if $sleep_count++ > 6;
544 } until defined $pid;
547 print KID_TO_WRITE @some_data;
548 close(KID_TO_WRITE) || warn "kid exited $?";
550 ($EUID, $EGID) = ($UID, $GID); # suid progs only
551 open (FILE, "> /safe/file")
552 || die "can't open /safe/file: $!";
554 print FILE; # child's STDIN is parent's KID
556 exit; # don't forget this
559 Another common use for this construct is when you need to execute
560 something without the shell's interference. With system(), it's
561 straightforward, but you can't use a pipe open or backticks safely.
562 That's because there's no way to stop the shell from getting its hands on
563 your arguments. Instead, use lower-level control to call exec() directly.
565 Here's a safe backtick or pipe open for read:
567 # add error processing as above
568 $pid = open(KID_TO_READ, "-|");
571 while (<KID_TO_READ>) {
572 # do something interesting
574 close(KID_TO_READ) || warn "kid exited $?";
577 ($EUID, $EGID) = ($UID, $GID); # suid only
578 exec($program, @options, @args)
579 || die "can't exec program: $!";
584 And here's a safe pipe open for writing:
586 # add error processing as above
587 $pid = open(KID_TO_WRITE, "|-");
588 $SIG{PIPE} = sub { die "whoops, $program pipe broke" };
594 close(KID_TO_WRITE) || warn "kid exited $?";
597 ($EUID, $EGID) = ($UID, $GID);
598 exec($program, @options, @args)
599 || die "can't exec program: $!";
603 Since Perl 5.8.0, you can also use the list form of C<open> for pipes :
606 open KID_PS, "-|", "ps", "aux" or die $!;
608 forks the ps(1) command (without spawning a shell, as there are more than
609 three arguments to open()), and reads its standard output via the
610 C<KID_PS> filehandle. The corresponding syntax to read from command
611 pipes (with C<"|-"> in place of C<"-|">) is also implemented.
613 Note that these operations are full Unix forks, which means they may not be
614 correctly implemented on alien systems. Additionally, these are not true
615 multithreading. If you'd like to learn more about threading, see the
616 F<modules> file mentioned below in the SEE ALSO section.
618 =head2 Bidirectional Communication with Another Process
620 While this works reasonably well for unidirectional communication, what
621 about bidirectional communication? The obvious thing you'd like to do
622 doesn't actually work:
624 open(PROG_FOR_READING_AND_WRITING, "| some program |")
626 and if you forget to use the C<use warnings> pragma or the B<-w> flag,
627 then you'll miss out entirely on the diagnostic message:
629 Can't do bidirectional pipe at -e line 1.
631 If you really want to, you can use the standard open2() library function
632 to catch both ends. There's also an open3() for tridirectional I/O so you
633 can also catch your child's STDERR, but doing so would then require an
634 awkward select() loop and wouldn't allow you to use normal Perl input
637 If you look at its source, you'll see that open2() uses low-level
638 primitives like Unix pipe() and exec() calls to create all the connections.
639 While it might have been slightly more efficient by using socketpair(), it
640 would have then been even less portable than it already is. The open2()
641 and open3() functions are unlikely to work anywhere except on a Unix
642 system or some other one purporting to be POSIX compliant.
644 Here's an example of using open2():
648 $pid = open2(*Reader, *Writer, "cat -u -n" );
649 print Writer "stuff\n";
652 The problem with this is that Unix buffering is really going to
653 ruin your day. Even though your C<Writer> filehandle is auto-flushed,
654 and the process on the other end will get your data in a timely manner,
655 you can't usually do anything to force it to give it back to you
656 in a similarly quick fashion. In this case, we could, because we
657 gave I<cat> a B<-u> flag to make it unbuffered. But very few Unix
658 commands are designed to operate over pipes, so this seldom works
659 unless you yourself wrote the program on the other end of the
662 A solution to this is the nonstandard F<Comm.pl> library. It uses
663 pseudo-ttys to make your program behave more reasonably:
666 $ph = open_proc('cat -n');
668 print $ph "a line\n";
669 print "got back ", scalar <$ph>;
672 This way you don't have to have control over the source code of the
673 program you're using. The F<Comm> library also has expect()
674 and interact() functions. Find the library (and we hope its
675 successor F<IPC::Chat>) at your nearest CPAN archive as detailed
676 in the SEE ALSO section below.
678 The newer Expect.pm module from CPAN also addresses this kind of thing.
679 This module requires two other modules from CPAN: IO::Pty and IO::Stty.
680 It sets up a pseudo-terminal to interact with programs that insist on
681 using talking to the terminal device driver. If your system is
682 amongst those supported, this may be your best bet.
684 =head2 Bidirectional Communication with Yourself
686 If you want, you may make low-level pipe() and fork()
687 to stitch this together by hand. This example only
688 talks to itself, but you could reopen the appropriate
689 handles to STDIN and STDOUT and call other processes.
692 # pipe1 - bidirectional communication using two pipe pairs
693 # designed for the socketpair-challenged
694 use IO::Handle; # thousands of lines just for autoflush :-(
695 pipe(PARENT_RDR, CHILD_WTR); # XXX: failure?
696 pipe(CHILD_RDR, PARENT_WTR); # XXX: failure?
697 CHILD_WTR->autoflush(1);
698 PARENT_WTR->autoflush(1);
701 close PARENT_RDR; close PARENT_WTR;
702 print CHILD_WTR "Parent Pid $$ is sending this\n";
703 chomp($line = <CHILD_RDR>);
704 print "Parent Pid $$ just read this: `$line'\n";
705 close CHILD_RDR; close CHILD_WTR;
708 die "cannot fork: $!" unless defined $pid;
709 close CHILD_RDR; close CHILD_WTR;
710 chomp($line = <PARENT_RDR>);
711 print "Child Pid $$ just read this: `$line'\n";
712 print PARENT_WTR "Child Pid $$ is sending this\n";
713 close PARENT_RDR; close PARENT_WTR;
717 But you don't actually have to make two pipe calls. If you
718 have the socketpair() system call, it will do this all for you.
721 # pipe2 - bidirectional communication using socketpair
722 # "the best ones always go both ways"
725 use IO::Handle; # thousands of lines just for autoflush :-(
726 # We say AF_UNIX because although *_LOCAL is the
727 # POSIX 1003.1g form of the constant, many machines
728 # still don't have it.
729 socketpair(CHILD, PARENT, AF_UNIX, SOCK_STREAM, PF_UNSPEC)
730 or die "socketpair: $!";
733 PARENT->autoflush(1);
737 print CHILD "Parent Pid $$ is sending this\n";
738 chomp($line = <CHILD>);
739 print "Parent Pid $$ just read this: `$line'\n";
743 die "cannot fork: $!" unless defined $pid;
745 chomp($line = <PARENT>);
746 print "Child Pid $$ just read this: `$line'\n";
747 print PARENT "Child Pid $$ is sending this\n";
752 =head1 Sockets: Client/Server Communication
754 While not limited to Unix-derived operating systems (e.g., WinSock on PCs
755 provides socket support, as do some VMS libraries), you may not have
756 sockets on your system, in which case this section probably isn't going to do
757 you much good. With sockets, you can do both virtual circuits (i.e., TCP
758 streams) and datagrams (i.e., UDP packets). You may be able to do even more
759 depending on your system.
761 The Perl function calls for dealing with sockets have the same names as
762 the corresponding system calls in C, but their arguments tend to differ
763 for two reasons: first, Perl filehandles work differently than C file
764 descriptors. Second, Perl already knows the length of its strings, so you
765 don't need to pass that information.
767 One of the major problems with old socket code in Perl was that it used
768 hard-coded values for some of the constants, which severely hurt
769 portability. If you ever see code that does anything like explicitly
770 setting C<$AF_INET = 2>, you know you're in for big trouble: An
771 immeasurably superior approach is to use the C<Socket> module, which more
772 reliably grants access to various constants and functions you'll need.
774 If you're not writing a server/client for an existing protocol like
775 NNTP or SMTP, you should give some thought to how your server will
776 know when the client has finished talking, and vice-versa. Most
777 protocols are based on one-line messages and responses (so one party
778 knows the other has finished when a "\n" is received) or multi-line
779 messages and responses that end with a period on an empty line
780 ("\n.\n" terminates a message/response).
782 =head2 Internet Line Terminators
784 The Internet line terminator is "\015\012". Under ASCII variants of
785 Unix, that could usually be written as "\r\n", but under other systems,
786 "\r\n" might at times be "\015\015\012", "\012\012\015", or something
787 completely different. The standards specify writing "\015\012" to be
788 conformant (be strict in what you provide), but they also recommend
789 accepting a lone "\012" on input (but be lenient in what you require).
790 We haven't always been very good about that in the code in this manpage,
791 but unless you're on a Mac, you'll probably be ok.
793 =head2 Internet TCP Clients and Servers
795 Use Internet-domain sockets when you want to do client-server
796 communication that might extend to machines outside of your own system.
798 Here's a sample TCP client using Internet-domain sockets:
803 my ($remote,$port, $iaddr, $paddr, $proto, $line);
805 $remote = shift || 'localhost';
806 $port = shift || 2345; # random port
807 if ($port =~ /\D/) { $port = getservbyname($port, 'tcp') }
808 die "No port" unless $port;
809 $iaddr = inet_aton($remote) || die "no host: $remote";
810 $paddr = sockaddr_in($port, $iaddr);
812 $proto = getprotobyname('tcp');
813 socket(SOCK, PF_INET, SOCK_STREAM, $proto) || die "socket: $!";
814 connect(SOCK, $paddr) || die "connect: $!";
815 while (defined($line = <SOCK>)) {
819 close (SOCK) || die "close: $!";
822 And here's a corresponding server to go along with it. We'll
823 leave the address as INADDR_ANY so that the kernel can choose
824 the appropriate interface on multihomed hosts. If you want sit
825 on a particular interface (like the external side of a gateway
826 or firewall machine), you should fill this in with your real address
831 BEGIN { $ENV{PATH} = '/usr/ucb:/bin' }
834 my $EOL = "\015\012";
836 sub logmsg { print "$0 $$: @_ at ", scalar localtime, "\n" }
838 my $port = shift || 2345;
839 my $proto = getprotobyname('tcp');
841 ($port) = $port =~ /^(\d+)$/ or die "invalid port";
843 socket(Server, PF_INET, SOCK_STREAM, $proto) || die "socket: $!";
844 setsockopt(Server, SOL_SOCKET, SO_REUSEADDR,
845 pack("l", 1)) || die "setsockopt: $!";
846 bind(Server, sockaddr_in($port, INADDR_ANY)) || die "bind: $!";
847 listen(Server,SOMAXCONN) || die "listen: $!";
849 logmsg "server started on port $port";
853 $SIG{CHLD} = \&REAPER;
855 for ( ; $paddr = accept(Client,Server); close Client) {
856 my($port,$iaddr) = sockaddr_in($paddr);
857 my $name = gethostbyaddr($iaddr,AF_INET);
859 logmsg "connection from $name [",
860 inet_ntoa($iaddr), "]
863 print Client "Hello there, $name, it's now ",
864 scalar localtime, $EOL;
867 And here's a multithreaded version. It's multithreaded in that
868 like most typical servers, it spawns (forks) a slave server to
869 handle the client request so that the master server can quickly
870 go back to service a new client.
874 BEGIN { $ENV{PATH} = '/usr/ucb:/bin' }
877 my $EOL = "\015\012";
879 sub spawn; # forward declaration
880 sub logmsg { print "$0 $$: @_ at ", scalar localtime, "\n" }
882 my $port = shift || 2345;
883 my $proto = getprotobyname('tcp');
885 ($port) = $port =~ /^(\d+)$/ or die "invalid port";
887 socket(Server, PF_INET, SOCK_STREAM, $proto) || die "socket: $!";
888 setsockopt(Server, SOL_SOCKET, SO_REUSEADDR,
889 pack("l", 1)) || die "setsockopt: $!";
890 bind(Server, sockaddr_in($port, INADDR_ANY)) || die "bind: $!";
891 listen(Server,SOMAXCONN) || die "listen: $!";
893 logmsg "server started on port $port";
898 use POSIX ":sys_wait_h";
901 while (($waitedpid = waitpid(-1,WNOHANG)) > 0) {
902 logmsg "reaped $waitedpid" . ($? ? " with exit $?" : '');
904 $SIG{CHLD} = \&REAPER; # loathe sysV
907 $SIG{CHLD} = \&REAPER;
909 for ( $waitedpid = 0;
910 ($paddr = accept(Client,Server)) || $waitedpid;
911 $waitedpid = 0, close Client)
913 next if $waitedpid and not $paddr;
914 my($port,$iaddr) = sockaddr_in($paddr);
915 my $name = gethostbyaddr($iaddr,AF_INET);
917 logmsg "connection from $name [",
918 inet_ntoa($iaddr), "]
923 print "Hello there, $name, it's now ", scalar localtime, $EOL;
924 exec '/usr/games/fortune' # XXX: `wrong' line terminators
925 or confess "can't exec fortune: $!";
933 unless (@_ == 0 && $coderef && ref($coderef) eq 'CODE') {
934 confess "usage: spawn CODEREF";
938 if (!defined($pid = fork)) {
939 logmsg "cannot fork: $!";
943 return; # I'm the parent
945 # else I'm the child -- go spawn
947 open(STDIN, "<&Client") || die "can't dup client to stdin";
948 open(STDOUT, ">&Client") || die "can't dup client to stdout";
949 ## open(STDERR, ">&STDOUT") || die "can't dup stdout to stderr";
953 This server takes the trouble to clone off a child version via fork() for
954 each incoming request. That way it can handle many requests at once,
955 which you might not always want. Even if you don't fork(), the listen()
956 will allow that many pending connections. Forking servers have to be
957 particularly careful about cleaning up their dead children (called
958 "zombies" in Unix parlance), because otherwise you'll quickly fill up your
961 We suggest that you use the B<-T> flag to use taint checking (see L<perlsec>)
962 even if we aren't running setuid or setgid. This is always a good idea
963 for servers and other programs run on behalf of someone else (like CGI
964 scripts), because it lessens the chances that people from the outside will
965 be able to compromise your system.
967 Let's look at another TCP client. This one connects to the TCP "time"
968 service on a number of different machines and shows how far their clocks
969 differ from the system on which it's being run:
975 my $SECS_of_70_YEARS = 2208988800;
976 sub ctime { scalar localtime(shift) }
978 my $iaddr = gethostbyname('localhost');
979 my $proto = getprotobyname('tcp');
980 my $port = getservbyname('time', 'tcp');
981 my $paddr = sockaddr_in(0, $iaddr);
985 printf "%-24s %8s %s\n", "localhost", 0, ctime(time());
987 foreach $host (@ARGV) {
988 printf "%-24s ", $host;
989 my $hisiaddr = inet_aton($host) || die "unknown host";
990 my $hispaddr = sockaddr_in($port, $hisiaddr);
991 socket(SOCKET, PF_INET, SOCK_STREAM, $proto) || die "socket: $!";
992 connect(SOCKET, $hispaddr) || die "bind: $!";
994 read(SOCKET, $rtime, 4);
996 my $histime = unpack("N", $rtime) - $SECS_of_70_YEARS ;
997 printf "%8d %s\n", $histime - time, ctime($histime);
1000 =head2 Unix-Domain TCP Clients and Servers
1002 That's fine for Internet-domain clients and servers, but what about local
1003 communications? While you can use the same setup, sometimes you don't
1004 want to. Unix-domain sockets are local to the current host, and are often
1005 used internally to implement pipes. Unlike Internet domain sockets, Unix
1006 domain sockets can show up in the file system with an ls(1) listing.
1009 srw-rw-rw- 1 root 0 Oct 31 07:23 /dev/log
1011 You can test for these with Perl's B<-S> file test:
1013 unless ( -S '/dev/log' ) {
1014 die "something's wicked with the log system";
1017 Here's a sample Unix-domain client:
1022 my ($rendezvous, $line);
1024 $rendezvous = shift || '/tmp/catsock';
1025 socket(SOCK, PF_UNIX, SOCK_STREAM, 0) || die "socket: $!";
1026 connect(SOCK, sockaddr_un($rendezvous)) || die "connect: $!";
1027 while (defined($line = <SOCK>)) {
1032 And here's a corresponding server. You don't have to worry about silly
1033 network terminators here because Unix domain sockets are guaranteed
1034 to be on the localhost, and thus everything works right.
1041 BEGIN { $ENV{PATH} = '/usr/ucb:/bin' }
1042 sub spawn; # forward declaration
1043 sub logmsg { print "$0 $$: @_ at ", scalar localtime, "\n" }
1045 my $NAME = '/tmp/catsock';
1046 my $uaddr = sockaddr_un($NAME);
1047 my $proto = getprotobyname('tcp');
1049 socket(Server,PF_UNIX,SOCK_STREAM,0) || die "socket: $!";
1051 bind (Server, $uaddr) || die "bind: $!";
1052 listen(Server,SOMAXCONN) || die "listen: $!";
1054 logmsg "server started on $NAME";
1058 use POSIX ":sys_wait_h";
1061 while (($waitedpid = waitpid(-1,WNOHANG)) > 0) {
1062 logmsg "reaped $waitedpid" . ($? ? " with exit $?" : '');
1064 $SIG{CHLD} = \&REAPER; # loathe sysV
1067 $SIG{CHLD} = \&REAPER;
1070 for ( $waitedpid = 0;
1071 accept(Client,Server) || $waitedpid;
1072 $waitedpid = 0, close Client)
1075 logmsg "connection on $NAME";
1077 print "Hello there, it's now ", scalar localtime, "\n";
1078 exec '/usr/games/fortune' or die "can't exec fortune: $!";
1083 my $coderef = shift;
1085 unless (@_ == 0 && $coderef && ref($coderef) eq 'CODE') {
1086 confess "usage: spawn CODEREF";
1090 if (!defined($pid = fork)) {
1091 logmsg "cannot fork: $!";
1094 logmsg "begat $pid";
1095 return; # I'm the parent
1097 # else I'm the child -- go spawn
1099 open(STDIN, "<&Client") || die "can't dup client to stdin";
1100 open(STDOUT, ">&Client") || die "can't dup client to stdout";
1101 ## open(STDERR, ">&STDOUT") || die "can't dup stdout to stderr";
1105 As you see, it's remarkably similar to the Internet domain TCP server, so
1106 much so, in fact, that we've omitted several duplicate functions--spawn(),
1107 logmsg(), ctime(), and REAPER()--which are exactly the same as in the
1110 So why would you ever want to use a Unix domain socket instead of a
1111 simpler named pipe? Because a named pipe doesn't give you sessions. You
1112 can't tell one process's data from another's. With socket programming,
1113 you get a separate session for each client: that's why accept() takes two
1116 For example, let's say that you have a long running database server daemon
1117 that you want folks from the World Wide Web to be able to access, but only
1118 if they go through a CGI interface. You'd have a small, simple CGI
1119 program that does whatever checks and logging you feel like, and then acts
1120 as a Unix-domain client and connects to your private server.
1122 =head1 TCP Clients with IO::Socket
1124 For those preferring a higher-level interface to socket programming, the
1125 IO::Socket module provides an object-oriented approach. IO::Socket is
1126 included as part of the standard Perl distribution as of the 5.004
1127 release. If you're running an earlier version of Perl, just fetch
1128 IO::Socket from CPAN, where you'll also find modules providing easy
1129 interfaces to the following systems: DNS, FTP, Ident (RFC 931), NIS and
1130 NISPlus, NNTP, Ping, POP3, SMTP, SNMP, SSLeay, Telnet, and Time--just
1133 =head2 A Simple Client
1135 Here's a client that creates a TCP connection to the "daytime"
1136 service at port 13 of the host name "localhost" and prints out everything
1137 that the server there cares to provide.
1141 $remote = IO::Socket::INET->new(
1143 PeerAddr => "localhost",
1144 PeerPort => "daytime(13)",
1146 or die "cannot connect to daytime port at localhost";
1147 while ( <$remote> ) { print }
1149 When you run this program, you should get something back that
1152 Wed May 14 08:40:46 MDT 1997
1154 Here are what those parameters to the C<new> constructor mean:
1160 This is which protocol to use. In this case, the socket handle returned
1161 will be connected to a TCP socket, because we want a stream-oriented
1162 connection, that is, one that acts pretty much like a plain old file.
1163 Not all sockets are this of this type. For example, the UDP protocol
1164 can be used to make a datagram socket, used for message-passing.
1168 This is the name or Internet address of the remote host the server is
1169 running on. We could have specified a longer name like C<"www.perl.com">,
1170 or an address like C<"204.148.40.9">. For demonstration purposes, we've
1171 used the special hostname C<"localhost">, which should always mean the
1172 current machine you're running on. The corresponding Internet address
1173 for localhost is C<"127.1">, if you'd rather use that.
1177 This is the service name or port number we'd like to connect to.
1178 We could have gotten away with using just C<"daytime"> on systems with a
1179 well-configured system services file,[FOOTNOTE: The system services file
1180 is in I</etc/services> under Unix] but just in case, we've specified the
1181 port number (13) in parentheses. Using just the number would also have
1182 worked, but constant numbers make careful programmers nervous.
1186 Notice how the return value from the C<new> constructor is used as
1187 a filehandle in the C<while> loop? That's what's called an indirect
1188 filehandle, a scalar variable containing a filehandle. You can use
1189 it the same way you would a normal filehandle. For example, you
1190 can read one line from it this way:
1194 all remaining lines from is this way:
1198 and send a line of data to it this way:
1200 print $handle "some data\n";
1202 =head2 A Webget Client
1204 Here's a simple client that takes a remote host to fetch a document
1205 from, and then a list of documents to get from that host. This is a
1206 more interesting client than the previous one because it first sends
1207 something to the server before fetching the server's response.
1211 unless (@ARGV > 1) { die "usage: $0 host document ..." }
1212 $host = shift(@ARGV);
1215 foreach $document ( @ARGV ) {
1216 $remote = IO::Socket::INET->new( Proto => "tcp",
1218 PeerPort => "http(80)",
1220 unless ($remote) { die "cannot connect to http daemon on $host" }
1221 $remote->autoflush(1);
1222 print $remote "GET $document HTTP/1.0" . $BLANK;
1223 while ( <$remote> ) { print }
1227 The web server handing the "http" service, which is assumed to be at
1228 its standard port, number 80. If the web server you're trying to
1229 connect to is at a different port (like 1080 or 8080), you should specify
1230 as the named-parameter pair, C<< PeerPort => 8080 >>. The C<autoflush>
1231 method is used on the socket because otherwise the system would buffer
1232 up the output we sent it. (If you're on a Mac, you'll also need to
1233 change every C<"\n"> in your code that sends data over the network to
1234 be a C<"\015\012"> instead.)
1236 Connecting to the server is only the first part of the process: once you
1237 have the connection, you have to use the server's language. Each server
1238 on the network has its own little command language that it expects as
1239 input. The string that we send to the server starting with "GET" is in
1240 HTTP syntax. In this case, we simply request each specified document.
1241 Yes, we really are making a new connection for each document, even though
1242 it's the same host. That's the way you always used to have to speak HTTP.
1243 Recent versions of web browsers may request that the remote server leave
1244 the connection open a little while, but the server doesn't have to honor
1247 Here's an example of running that program, which we'll call I<webget>:
1249 % webget www.perl.com /guanaco.html
1250 HTTP/1.1 404 File Not Found
1251 Date: Thu, 08 May 1997 18:02:32 GMT
1252 Server: Apache/1.2b6
1254 Content-type: text/html
1256 <HEAD><TITLE>404 File Not Found</TITLE></HEAD>
1257 <BODY><H1>File Not Found</H1>
1258 The requested URL /guanaco.html was not found on this server.<P>
1261 Ok, so that's not very interesting, because it didn't find that
1262 particular document. But a long response wouldn't have fit on this page.
1264 For a more fully-featured version of this program, you should look to
1265 the I<lwp-request> program included with the LWP modules from CPAN.
1267 =head2 Interactive Client with IO::Socket
1269 Well, that's all fine if you want to send one command and get one answer,
1270 but what about setting up something fully interactive, somewhat like
1271 the way I<telnet> works? That way you can type a line, get the answer,
1272 type a line, get the answer, etc.
1274 This client is more complicated than the two we've done so far, but if
1275 you're on a system that supports the powerful C<fork> call, the solution
1276 isn't that rough. Once you've made the connection to whatever service
1277 you'd like to chat with, call C<fork> to clone your process. Each of
1278 these two identical process has a very simple job to do: the parent
1279 copies everything from the socket to standard output, while the child
1280 simultaneously copies everything from standard input to the socket.
1281 To accomplish the same thing using just one process would be I<much>
1282 harder, because it's easier to code two processes to do one thing than it
1283 is to code one process to do two things. (This keep-it-simple principle
1284 a cornerstones of the Unix philosophy, and good software engineering as
1285 well, which is probably why it's spread to other systems.)
1292 my ($host, $port, $kidpid, $handle, $line);
1294 unless (@ARGV == 2) { die "usage: $0 host port" }
1295 ($host, $port) = @ARGV;
1297 # create a tcp connection to the specified host and port
1298 $handle = IO::Socket::INET->new(Proto => "tcp",
1301 or die "can't connect to port $port on $host: $!";
1303 $handle->autoflush(1); # so output gets there right away
1304 print STDERR "[Connected to $host:$port]\n";
1306 # split the program into two processes, identical twins
1307 die "can't fork: $!" unless defined($kidpid = fork());
1309 # the if{} block runs only in the parent process
1311 # copy the socket to standard output
1312 while (defined ($line = <$handle>)) {
1315 kill("TERM", $kidpid); # send SIGTERM to child
1317 # the else{} block runs only in the child process
1319 # copy standard input to the socket
1320 while (defined ($line = <STDIN>)) {
1321 print $handle $line;
1325 The C<kill> function in the parent's C<if> block is there to send a
1326 signal to our child process (current running in the C<else> block)
1327 as soon as the remote server has closed its end of the connection.
1329 If the remote server sends data a byte at time, and you need that
1330 data immediately without waiting for a newline (which might not happen),
1331 you may wish to replace the C<while> loop in the parent with the
1335 while (sysread($handle, $byte, 1) == 1) {
1339 Making a system call for each byte you want to read is not very efficient
1340 (to put it mildly) but is the simplest to explain and works reasonably
1343 =head1 TCP Servers with IO::Socket
1345 As always, setting up a server is little bit more involved than running a client.
1346 The model is that the server creates a special kind of socket that
1347 does nothing but listen on a particular port for incoming connections.
1348 It does this by calling the C<< IO::Socket::INET->new() >> method with
1349 slightly different arguments than the client did.
1355 This is which protocol to use. Like our clients, we'll
1356 still specify C<"tcp"> here.
1361 port in the C<LocalPort> argument, which we didn't do for the client.
1362 This is service name or port number for which you want to be the
1363 server. (Under Unix, ports under 1024 are restricted to the
1364 superuser.) In our sample, we'll use port 9000, but you can use
1365 any port that's not currently in use on your system. If you try
1366 to use one already in used, you'll get an "Address already in use"
1367 message. Under Unix, the C<netstat -a> command will show
1368 which services current have servers.
1372 The C<Listen> parameter is set to the maximum number of
1373 pending connections we can accept until we turn away incoming clients.
1374 Think of it as a call-waiting queue for your telephone.
1375 The low-level Socket module has a special symbol for the system maximum, which
1380 The C<Reuse> parameter is needed so that we restart our server
1381 manually without waiting a few minutes to allow system buffers to
1386 Once the generic server socket has been created using the parameters
1387 listed above, the server then waits for a new client to connect
1388 to it. The server blocks in the C<accept> method, which eventually accepts a
1389 bidirectional connection from the remote client. (Make sure to autoflush
1390 this handle to circumvent buffering.)
1392 To add to user-friendliness, our server prompts the user for commands.
1393 Most servers don't do this. Because of the prompt without a newline,
1394 you'll have to use the C<sysread> variant of the interactive client above.
1396 This server accepts one of five different commands, sending output
1397 back to the client. Note that unlike most network servers, this one
1398 only handles one incoming client at a time. Multithreaded servers are
1399 covered in Chapter 6 of the Camel.
1401 Here's the code. We'll
1405 use Net::hostent; # for OO version of gethostbyaddr
1407 $PORT = 9000; # pick something not in use
1409 $server = IO::Socket::INET->new( Proto => 'tcp',
1411 Listen => SOMAXCONN,
1414 die "can't setup server" unless $server;
1415 print "[Server $0 accepting clients]\n";
1417 while ($client = $server->accept()) {
1418 $client->autoflush(1);
1419 print $client "Welcome to $0; type help for command list.\n";
1420 $hostinfo = gethostbyaddr($client->peeraddr);
1421 printf "[Connect from %s]\n", $hostinfo ? $hostinfo->name : $client->peerhost;
1422 print $client "Command? ";
1423 while ( <$client>) {
1424 next unless /\S/; # blank line
1425 if (/quit|exit/i) { last; }
1426 elsif (/date|time/i) { printf $client "%s\n", scalar localtime; }
1427 elsif (/who/i ) { print $client `who 2>&1`; }
1428 elsif (/cookie/i ) { print $client `/usr/games/fortune 2>&1`; }
1429 elsif (/motd/i ) { print $client `cat /etc/motd 2>&1`; }
1431 print $client "Commands: quit date who cookie motd\n";
1434 print $client "Command? ";
1439 =head1 UDP: Message Passing
1441 Another kind of client-server setup is one that uses not connections, but
1442 messages. UDP communications involve much lower overhead but also provide
1443 less reliability, as there are no promises that messages will arrive at
1444 all, let alone in order and unmangled. Still, UDP offers some advantages
1445 over TCP, including being able to "broadcast" or "multicast" to a whole
1446 bunch of destination hosts at once (usually on your local subnet). If you
1447 find yourself overly concerned about reliability and start building checks
1448 into your message system, then you probably should use just TCP to start
1451 Note that UDP datagrams are I<not> a bytestream and should not be treated
1452 as such. This makes using I/O mechanisms with internal buffering
1453 like stdio (i.e. print() and friends) especially cumbersome. Use syswrite(),
1454 or better send(), like in the example below.
1456 Here's a UDP program similar to the sample Internet TCP client given
1457 earlier. However, instead of checking one host at a time, the UDP version
1458 will check many of them asynchronously by simulating a multicast and then
1459 using select() to do a timed-out wait for I/O. To do something similar
1460 with TCP, you'd have to use a different socket handle for each host.
1467 my ( $count, $hisiaddr, $hispaddr, $histime,
1468 $host, $iaddr, $paddr, $port, $proto,
1469 $rin, $rout, $rtime, $SECS_of_70_YEARS);
1471 $SECS_of_70_YEARS = 2208988800;
1473 $iaddr = gethostbyname(hostname());
1474 $proto = getprotobyname('udp');
1475 $port = getservbyname('time', 'udp');
1476 $paddr = sockaddr_in(0, $iaddr); # 0 means let kernel pick
1478 socket(SOCKET, PF_INET, SOCK_DGRAM, $proto) || die "socket: $!";
1479 bind(SOCKET, $paddr) || die "bind: $!";
1482 printf "%-12s %8s %s\n", "localhost", 0, scalar localtime time;
1486 $hisiaddr = inet_aton($host) || die "unknown host";
1487 $hispaddr = sockaddr_in($port, $hisiaddr);
1488 defined(send(SOCKET, 0, 0, $hispaddr)) || die "send $host: $!";
1492 vec($rin, fileno(SOCKET), 1) = 1;
1494 # timeout after 10.0 seconds
1495 while ($count && select($rout = $rin, undef, undef, 10.0)) {
1497 ($hispaddr = recv(SOCKET, $rtime, 4, 0)) || die "recv: $!";
1498 ($port, $hisiaddr) = sockaddr_in($hispaddr);
1499 $host = gethostbyaddr($hisiaddr, AF_INET);
1500 $histime = unpack("N", $rtime) - $SECS_of_70_YEARS ;
1501 printf "%-12s ", $host;
1502 printf "%8d %s\n", $histime - time, scalar localtime($histime);
1506 Note that this example does not include any retries and may consequently
1507 fail to contact a reachable host. The most prominent reason for this
1508 is congestion of the queues on the sending host if the number of
1509 list of hosts to contact is sufficiently large.
1513 While System V IPC isn't so widely used as sockets, it still has some
1514 interesting uses. You can't, however, effectively use SysV IPC or
1515 Berkeley mmap() to have shared memory so as to share a variable amongst
1516 several processes. That's because Perl would reallocate your string when
1517 you weren't wanting it to.
1519 Here's a small example showing shared memory usage.
1521 use IPC::SysV qw(IPC_PRIVATE IPC_RMID S_IRWXU);
1524 $id = shmget(IPC_PRIVATE, $size, S_IRWXU) || die "$!";
1525 print "shm key $id\n";
1527 $message = "Message #1";
1528 shmwrite($id, $message, 0, 60) || die "$!";
1529 print "wrote: '$message'\n";
1530 shmread($id, $buff, 0, 60) || die "$!";
1531 print "read : '$buff'\n";
1533 # the buffer of shmread is zero-character end-padded.
1534 substr($buff, index($buff, "\0")) = '';
1535 print "un" unless $buff eq $message;
1538 print "deleting shm $id\n";
1539 shmctl($id, IPC_RMID, 0) || die "$!";
1541 Here's an example of a semaphore:
1543 use IPC::SysV qw(IPC_CREAT);
1546 $id = semget($IPC_KEY, 10, 0666 | IPC_CREAT ) || die "$!";
1547 print "shm key $id\n";
1549 Put this code in a separate file to be run in more than one process.
1550 Call the file F<take>:
1552 # create a semaphore
1555 $id = semget($IPC_KEY, 0 , 0 );
1556 die if !defined($id);
1562 # wait for semaphore to be zero
1564 $opstring1 = pack("s!s!s!", $semnum, $semop, $semflag);
1566 # Increment the semaphore count
1568 $opstring2 = pack("s!s!s!", $semnum, $semop, $semflag);
1569 $opstring = $opstring1 . $opstring2;
1571 semop($id,$opstring) || die "$!";
1573 Put this code in a separate file to be run in more than one process.
1574 Call this file F<give>:
1576 # 'give' the semaphore
1577 # run this in the original process and you will see
1578 # that the second process continues
1581 $id = semget($IPC_KEY, 0, 0);
1582 die if !defined($id);
1587 # Decrement the semaphore count
1589 $opstring = pack("s!s!s!", $semnum, $semop, $semflag);
1591 semop($id,$opstring) || die "$!";
1593 The SysV IPC code above was written long ago, and it's definitely
1594 clunky looking. For a more modern look, see the IPC::SysV module
1595 which is included with Perl starting from Perl 5.005.
1597 A small example demonstrating SysV message queues:
1599 use IPC::SysV qw(IPC_PRIVATE IPC_RMID IPC_CREAT S_IRWXU);
1601 my $id = msgget(IPC_PRIVATE, IPC_CREAT | S_IRWXU);
1603 my $sent = "message";
1609 if (msgsnd($id, pack("l! a*", $type_sent, $sent), 0)) {
1610 if (msgrcv($id, $rcvd, 60, 0, 0)) {
1611 ($type_rcvd, $rcvd) = unpack("l! a*", $rcvd);
1612 if ($rcvd eq $sent) {
1618 die "# msgrcv failed\n";
1621 die "# msgsnd failed\n";
1623 msgctl($id, IPC_RMID, 0) || die "# msgctl failed: $!\n";
1625 die "# msgget failed\n";
1630 Most of these routines quietly but politely return C<undef> when they
1631 fail instead of causing your program to die right then and there due to
1632 an uncaught exception. (Actually, some of the new I<Socket> conversion
1633 functions croak() on bad arguments.) It is therefore essential to
1634 check return values from these functions. Always begin your socket
1635 programs this way for optimal success, and don't forget to add B<-T>
1636 taint checking flag to the #! line for servers:
1645 All these routines create system-specific portability problems. As noted
1646 elsewhere, Perl is at the mercy of your C libraries for much of its system
1647 behaviour. It's probably safest to assume broken SysV semantics for
1648 signals and to stick with simple TCP and UDP socket operations; e.g., don't
1649 try to pass open file descriptors over a local UDP datagram socket if you
1650 want your code to stand a chance of being portable.
1652 As mentioned in the signals section, because few vendors provide C
1653 libraries that are safely re-entrant, the prudent programmer will do
1654 little else within a handler beyond setting a numeric variable that
1655 already exists; or, if locked into a slow (restarting) system call,
1656 using die() to raise an exception and longjmp(3) out. In fact, even
1657 these may in some cases cause a core dump. It's probably best to avoid
1658 signals except where they are absolutely inevitable. This
1659 will be addressed in a future release of Perl.
1663 Tom Christiansen, with occasional vestiges of Larry Wall's original
1664 version and suggestions from the Perl Porters.
1668 There's a lot more to networking than this, but this should get you
1671 For intrepid programmers, the indispensable textbook is I<Unix
1672 Network Programming, 2nd Edition, Volume 1> by W. Richard Stevens
1673 (published by Prentice-Hall). Note that most books on networking
1674 address the subject from the perspective of a C programmer; translation
1675 to Perl is left as an exercise for the reader.
1677 The IO::Socket(3) manpage describes the object library, and the Socket(3)
1678 manpage describes the low-level interface to sockets. Besides the obvious
1679 functions in L<perlfunc>, you should also check out the F<modules> file
1680 at your nearest CPAN site. (See L<perlmodlib> or best yet, the F<Perl
1681 FAQ> for a description of what CPAN is and where to get it.)
1683 Section 5 of the F<modules> file is devoted to "Networking, Device Control
1684 (modems), and Interprocess Communication", and contains numerous unbundled
1685 modules numerous networking modules, Chat and Expect operations, CGI
1686 programming, DCE, FTP, IPC, NNTP, Proxy, Ptty, RPC, SNMP, SMTP, Telnet,
1687 Threads, and ToolTalk--just to name a few.