This is a live mirror of the Perl 5 development currently hosted at https://github.com/perl/perl5
Fix a side-effect of bug #24674 in the perl debugger.
[perl5.git] / pod / perlsec.pod
CommitLineData
a0d0e21e
LW
1=head1 NAME
2
3perlsec - Perl security
4
5=head1 DESCRIPTION
6
425e5e39
PP
7Perl is designed to make it easy to program securely even when running
8with extra privileges, like setuid or setgid programs. Unlike most
54310121 9command line shells, which are based on multiple substitution passes on
425e5e39
PP
10each line of the script, Perl uses a more conventional evaluation scheme
11with fewer hidden snags. Additionally, because the language has more
54310121 12builtin functionality, it can rely less upon external (and possibly
425e5e39 13untrustworthy) programs to accomplish its purposes.
a0d0e21e 14
425e5e39
PP
15Perl automatically enables a set of special security checks, called I<taint
16mode>, when it detects its program running with differing real and effective
17user or group IDs. The setuid bit in Unix permissions is mode 04000, the
18setgid bit mode 02000; either or both may be set. You can also enable taint
5f05dabc 19mode explicitly by using the B<-T> command line flag. This flag is
425e5e39 20I<strongly> suggested for server programs and any program run on behalf of
fb73857a
PP
21someone else, such as a CGI script. Once taint mode is on, it's on for
22the remainder of your script.
a0d0e21e 23
1e422769
PP
24While in this mode, Perl takes special precautions called I<taint
25checks> to prevent both obvious and subtle traps. Some of these checks
26are reasonably simple, such as verifying that path directories aren't
27writable by others; careful programmers have always used checks like
28these. Other checks, however, are best supported by the language itself,
fb73857a 29and it is these checks especially that contribute to making a set-id Perl
425e5e39
PP
30program more secure than the corresponding C program.
31
fb73857a
PP
32You may not use data derived from outside your program to affect
33something else outside your program--at least, not by accident. All
34command line arguments, environment variables, locale information (see
d929ce6f 35L<perllocale>), results of certain system calls (readdir(),
41d6edb2
JH
36readlink(), the variable of shmread(), the messages returned by
37msgrcv(), the password, gcos and shell fields returned by the
38getpwxxx() calls), and all file input are marked as "tainted".
39Tainted data may not be used directly or indirectly in any command
40that invokes a sub-shell, nor in any command that modifies files,
b7ee89ce
AP
41directories, or processes, B<with the following exceptions>:
42
43=over 4
44
45=item *
46
b7ee89ce
AP
47Arguments to C<print> and C<syswrite> are B<not> checked for taintedness.
48
7f6513c1
JH
49=item *
50
51Symbolic methods
52
53 $obj->$method(@args);
54
55and symbolic sub references
56
57 &{$foo}(@args);
58 $foo->(@args);
59
60are not checked for taintedness. This requires extra carefulness
61unless you want external data to affect your control flow. Unless
62you carefully limit what these symbolic values are, people are able
63to call functions B<outside> your Perl code, such as POSIX::system,
64in which case they are able to run arbitrary external code.
65
b7ee89ce
AP
66=back
67
ee556d55
MG
68The value of an expression containing tainted data will itself be
69tainted, even if it is logically impossible for the tainted data to
70affect the value.
71
d929ce6f
JH
72Because taintedness is associated with each scalar value, some
73elements of an array can be tainted and others not.
a0d0e21e 74
a0d0e21e
LW
75For example:
76
425e5e39
PP
77 $arg = shift; # $arg is tainted
78 $hid = $arg, 'bar'; # $hid is also tainted
79 $line = <>; # Tainted
8ebc5c01
PP
80 $line = <STDIN>; # Also tainted
81 open FOO, "/home/me/bar" or die $!;
82 $line = <FOO>; # Still tainted
a0d0e21e 83 $path = $ENV{'PATH'}; # Tainted, but see below
425e5e39 84 $data = 'abc'; # Not tainted
a0d0e21e 85
425e5e39 86 system "echo $arg"; # Insecure
7de90c4d 87 system "/bin/echo", $arg; # Considered insecure
bbd7eb8a 88 # (Perl doesn't know about /bin/echo)
425e5e39
PP
89 system "echo $hid"; # Insecure
90 system "echo $data"; # Insecure until PATH set
a0d0e21e 91
425e5e39 92 $path = $ENV{'PATH'}; # $path now tainted
a0d0e21e 93
54310121 94 $ENV{'PATH'} = '/bin:/usr/bin';
c90c0ff4 95 delete @ENV{'IFS', 'CDPATH', 'ENV', 'BASH_ENV'};
a0d0e21e 96
425e5e39
PP
97 $path = $ENV{'PATH'}; # $path now NOT tainted
98 system "echo $data"; # Is secure now!
a0d0e21e 99
425e5e39
PP
100 open(FOO, "< $arg"); # OK - read-only file
101 open(FOO, "> $arg"); # Not OK - trying to write
a0d0e21e 102
bbd7eb8a 103 open(FOO,"echo $arg|"); # Not OK
425e5e39 104 open(FOO,"-|")
7de90c4d 105 or exec 'echo', $arg; # Also not OK
a0d0e21e 106
425e5e39 107 $shout = `echo $arg`; # Insecure, $shout now tainted
a0d0e21e 108
425e5e39
PP
109 unlink $data, $arg; # Insecure
110 umask $arg; # Insecure
a0d0e21e 111
bbd7eb8a 112 exec "echo $arg"; # Insecure
7de90c4d
RD
113 exec "echo", $arg; # Insecure
114 exec "sh", '-c', $arg; # Very insecure!
a0d0e21e 115
3a4b19e4
GS
116 @files = <*.c>; # insecure (uses readdir() or similar)
117 @files = glob('*.c'); # insecure (uses readdir() or similar)
7bac28a0 118
3f7d42d8
JH
119 # In Perl releases older than 5.6.0 the <*.c> and glob('*.c') would
120 # have used an external program to do the filename expansion; but in
121 # either case the result is tainted since the list of filenames comes
122 # from outside of the program.
123
ee556d55
MG
124 $bad = ($arg, 23); # $bad will be tainted
125 $arg, `true`; # Insecure (although it isn't really)
126
a0d0e21e 127If you try to do something insecure, you will get a fatal error saying
7de90c4d 128something like "Insecure dependency" or "Insecure $ENV{PATH}".
425e5e39
PP
129
130=head2 Laundering and Detecting Tainted Data
131
3f7d42d8
JH
132To test whether a variable contains tainted data, and whose use would
133thus trigger an "Insecure dependency" message, you can use the
134tainted() function of the Scalar::Util module, available in your
135nearby CPAN mirror, and included in Perl starting from the release 5.8.0.
136Or you may be able to use the following I<is_tainted()> function.
425e5e39
PP
137
138 sub is_tainted {
61890e45 139 return ! eval { eval("#" . substr(join("", @_), 0, 0)); 1 };
425e5e39
PP
140 }
141
142This function makes use of the fact that the presence of tainted data
143anywhere within an expression renders the entire expression tainted. It
144would be inefficient for every operator to test every argument for
145taintedness. Instead, the slightly more efficient and conservative
146approach is used that if any tainted value has been accessed within the
147same expression, the whole expression is considered tainted.
148
5f05dabc 149But testing for taintedness gets you only so far. Sometimes you have just
425e5e39 150to clear your data's taintedness. The only way to bypass the tainting
54310121 151mechanism is by referencing subpatterns from a regular expression match.
425e5e39
PP
152Perl presumes that if you reference a substring using $1, $2, etc., that
153you knew what you were doing when you wrote the pattern. That means using
154a bit of thought--don't just blindly untaint anything, or you defeat the
a034a98d
DD
155entire mechanism. It's better to verify that the variable has only good
156characters (for certain values of "good") rather than checking whether it
157has any bad characters. That's because it's far too easy to miss bad
158characters that you never thought of.
425e5e39
PP
159
160Here's a test to make sure that the data contains nothing but "word"
161characters (alphabetics, numerics, and underscores), a hyphen, an at sign,
162or a dot.
163
54310121 164 if ($data =~ /^([-\@\w.]+)$/) {
425e5e39
PP
165 $data = $1; # $data now untainted
166 } else {
3a2263fe 167 die "Bad data in '$data'"; # log this somewhere
425e5e39
PP
168 }
169
5f05dabc 170This is fairly secure because C</\w+/> doesn't normally match shell
425e5e39
PP
171metacharacters, nor are dot, dash, or at going to mean something special
172to the shell. Use of C</.+/> would have been insecure in theory because
173it lets everything through, but Perl doesn't check for that. The lesson
174is that when untainting, you must be exceedingly careful with your patterns.
19799a22 175Laundering data using regular expression is the I<only> mechanism for
425e5e39
PP
176untainting dirty data, unless you use the strategy detailed below to fork
177a child of lesser privilege.
178
a034a98d
DD
179The example does not untaint $data if C<use locale> is in effect,
180because the characters matched by C<\w> are determined by the locale.
181Perl considers that locale definitions are untrustworthy because they
182contain data from outside the program. If you are writing a
183locale-aware program, and want to launder data with a regular expression
184containing C<\w>, put C<no locale> ahead of the expression in the same
185block. See L<perllocale/SECURITY> for further discussion and examples.
186
3a52c276
CS
187=head2 Switches On the "#!" Line
188
189When you make a script executable, in order to make it usable as a
190command, the system will pass switches to perl from the script's #!
54310121 191line. Perl checks that any command line switches given to a setuid
3a52c276 192(or setgid) script actually match the ones set on the #! line. Some
54310121 193Unix and Unix-like environments impose a one-switch limit on the #!
3a52c276 194line, so you may need to use something like C<-wU> instead of C<-w -U>
54310121
PP
195under such systems. (This issue should arise only in Unix or
196Unix-like environments that support #! and setuid or setgid scripts.)
3a52c276 197
588f7210
SB
198=head2 Taint mode and @INC
199
200When the taint mode (C<-T>) is in effect, the "." directory is removed
201from C<@INC>, and the environment variables C<PERL5LIB> and C<PERLLIB>
202are ignored by Perl. You can still adjust C<@INC> from outside the
203program by using the C<-I> command line option as explained in
204L<perlrun>. The two environment variables are ignored because
205they are obscured, and a user running a program could be unaware that
206they are set, whereas the C<-I> option is clearly visible and
207therefore permitted.
208
209Another way to modify C<@INC> without modifying the program, is to use
210the C<lib> pragma, e.g.:
211
212 perl -Mlib=/foo program
213
214The benefit of using C<-Mlib=/foo> over C<-I/foo>, is that the former
215will automagically remove any duplicated directories, while the later
216will not.
217
425e5e39
PP
218=head2 Cleaning Up Your Path
219
1fef88e7 220For "Insecure C<$ENV{PATH}>" messages, you need to set C<$ENV{'PATH'}> to a
1e422769
PP
221known value, and each directory in the path must be non-writable by others
222than its owner and group. You may be surprised to get this message even
223if the pathname to your executable is fully qualified. This is I<not>
224generated because you didn't supply a full path to the program; instead,
225it's generated because you never set your PATH environment variable, or
226you didn't set it to something that was safe. Because Perl can't
227guarantee that the executable in question isn't itself going to turn
228around and execute some other program that is dependent on your PATH, it
54310121 229makes sure you set the PATH.
a0d0e21e 230
a3cb178b
GS
231The PATH isn't the only environment variable which can cause problems.
232Because some shells may use the variables IFS, CDPATH, ENV, and
233BASH_ENV, Perl checks that those are either empty or untainted when
234starting subprocesses. You may wish to add something like this to your
235setid and taint-checking scripts.
236
237 delete @ENV{qw(IFS CDPATH ENV BASH_ENV)}; # Make %ENV safer
238
a0d0e21e
LW
239It's also possible to get into trouble with other operations that don't
240care whether they use tainted values. Make judicious use of the file
241tests in dealing with any user-supplied filenames. When possible, do
fb73857a
PP
242opens and such B<after> properly dropping any special user (or group!)
243privileges. Perl doesn't prevent you from opening tainted filenames for reading,
a0d0e21e
LW
244so be careful what you print out. The tainting mechanism is intended to
245prevent stupid mistakes, not to remove the need for thought.
246
425e5e39
PP
247Perl does not call the shell to expand wild cards when you pass B<system>
248and B<exec> explicit parameter lists instead of strings with possible shell
249wildcards in them. Unfortunately, the B<open>, B<glob>, and
54310121
PP
250backtick functions provide no such alternate calling convention, so more
251subterfuge will be required.
425e5e39
PP
252
253Perl provides a reasonably safe way to open a file or pipe from a setuid
254or setgid program: just create a child process with reduced privilege who
255does the dirty work for you. First, fork a child using the special
256B<open> syntax that connects the parent and child by a pipe. Now the
257child resets its ID set and any other per-process attributes, like
258environment variables, umasks, current working directories, back to the
259originals or known safe values. Then the child process, which no longer
260has any special permissions, does the B<open> or other system call.
261Finally, the child passes the data it managed to access back to the
5f05dabc 262parent. Because the file or pipe was opened in the child while running
425e5e39
PP
263under less privilege than the parent, it's not apt to be tricked into
264doing something it shouldn't.
265
54310121 266Here's a way to do backticks reasonably safely. Notice how the B<exec> is
425e5e39
PP
267not called with a string that the shell could expand. This is by far the
268best way to call something that might be subjected to shell escapes: just
fb73857a 269never call the shell at all.
cb1a09d0 270
a1ce9542 271 use English '-no_match_vars';
e093bcf0
GW
272 die "Can't fork: $!" unless defined($pid = open(KID, "-|"));
273 if ($pid) { # parent
274 while (<KID>) {
275 # do something
276 }
277 close KID;
278 } else {
279 my @temp = ($EUID, $EGID);
280 my $orig_uid = $UID;
281 my $orig_gid = $GID;
282 $EUID = $UID;
283 $EGID = $GID;
284 # Drop privileges
285 $UID = $orig_uid;
286 $GID = $orig_gid;
287 # Make sure privs are really gone
288 ($EUID, $EGID) = @temp;
289 die "Can't drop privileges"
290 unless $UID == $EUID && $GID eq $EGID;
291 $ENV{PATH} = "/bin:/usr/bin"; # Minimal PATH.
292 # Consider sanitizing the environment even more.
293 exec 'myprog', 'arg1', 'arg2'
294 or die "can't exec myprog: $!";
295 }
425e5e39 296
fb73857a
PP
297A similar strategy would work for wildcard expansion via C<glob>, although
298you can use C<readdir> instead.
425e5e39
PP
299
300Taint checking is most useful when although you trust yourself not to have
301written a program to give away the farm, you don't necessarily trust those
302who end up using it not to try to trick it into doing something bad. This
fb73857a 303is the kind of security checking that's useful for set-id programs and
425e5e39
PP
304programs launched on someone else's behalf, like CGI programs.
305
306This is quite different, however, from not even trusting the writer of the
307code not to try to do something evil. That's the kind of trust needed
308when someone hands you a program you've never seen before and says, "Here,
309run this." For that kind of safety, check out the Safe module,
310included standard in the Perl distribution. This module allows the
311programmer to set up special compartments in which all system operations
312are trapped and namespace access is carefully controlled.
313
314=head2 Security Bugs
315
316Beyond the obvious problems that stem from giving special privileges to
fb73857a 317systems as flexible as scripts, on many versions of Unix, set-id scripts
425e5e39
PP
318are inherently insecure right from the start. The problem is a race
319condition in the kernel. Between the time the kernel opens the file to
fb73857a 320see which interpreter to run and when the (now-set-id) interpreter turns
425e5e39
PP
321around and reopens the file to interpret it, the file in question may have
322changed, especially if you have symbolic links on your system.
323
324Fortunately, sometimes this kernel "feature" can be disabled.
325Unfortunately, there are two ways to disable it. The system can simply
fb73857a
PP
326outlaw scripts with any set-id bit set, which doesn't help much.
327Alternately, it can simply ignore the set-id bits on scripts. If the
425e5e39
PP
328latter is true, Perl can emulate the setuid and setgid mechanism when it
329notices the otherwise useless setuid/gid bits on Perl scripts. It does
330this via a special executable called B<suidperl> that is automatically
54310121 331invoked for you if it's needed.
425e5e39 332
fb73857a
PP
333However, if the kernel set-id script feature isn't disabled, Perl will
334complain loudly that your set-id script is insecure. You'll need to
335either disable the kernel set-id script feature, or put a C wrapper around
425e5e39
PP
336the script. A C wrapper is just a compiled program that does nothing
337except call your Perl program. Compiled programs are not subject to the
fb73857a 338kernel bug that plagues set-id scripts. Here's a simple wrapper, written
425e5e39
PP
339in C:
340
341 #define REAL_PATH "/path/to/script"
54310121 342 main(ac, av)
425e5e39
PP
343 char **av;
344 {
345 execv(REAL_PATH, av);
54310121 346 }
cb1a09d0 347
54310121
PP
348Compile this wrapper into a binary executable and then make I<it> rather
349than your script setuid or setgid.
425e5e39 350
425e5e39
PP
351In recent years, vendors have begun to supply systems free of this
352inherent security bug. On such systems, when the kernel passes the name
fb73857a 353of the set-id script to open to the interpreter, rather than using a
425e5e39
PP
354pathname subject to meddling, it instead passes I</dev/fd/3>. This is a
355special file already opened on the script, so that there can be no race
356condition for evil scripts to exploit. On these systems, Perl should be
357compiled with C<-DSETUID_SCRIPTS_ARE_SECURE_NOW>. The B<Configure>
358program that builds Perl tries to figure this out for itself, so you
359should never have to specify this yourself. Most modern releases of
360SysVr4 and BSD 4.4 use this approach to avoid the kernel race condition.
361
0325b4c4
JH
362Prior to release 5.6.1 of Perl, bugs in the code of B<suidperl> could
363introduce a security hole.
68dc0745
PP
364
365=head2 Protecting Your Programs
366
367There are a number of ways to hide the source to your Perl programs,
368with varying levels of "security".
369
370First of all, however, you I<can't> take away read permission, because
371the source code has to be readable in order to be compiled and
372interpreted. (That doesn't mean that a CGI script's source is
373readable by people on the web, though.) So you have to leave the
5a964f20
TC
374permissions at the socially friendly 0755 level. This lets
375people on your local system only see your source.
68dc0745 376
5a964f20 377Some people mistakenly regard this as a security problem. If your program does
68dc0745
PP
378insecure things, and relies on people not knowing how to exploit those
379insecurities, it is not secure. It is often possible for someone to
380determine the insecure things and exploit them without viewing the
381source. Security through obscurity, the name for hiding your bugs
382instead of fixing them, is little security indeed.
383
83df6a1d
JH
384You can try using encryption via source filters (Filter::* from CPAN,
385or Filter::Util::Call and Filter::Simple since Perl 5.8).
386But crackers might be able to decrypt it. You can try using the byte
387code compiler and interpreter described below, but crackers might be
388able to de-compile it. You can try using the native-code compiler
68dc0745
PP
389described below, but crackers might be able to disassemble it. These
390pose varying degrees of difficulty to people wanting to get at your
391code, but none can definitively conceal it (this is true of every
392language, not just Perl).
393
394If you're concerned about people profiting from your code, then the
395bottom line is that nothing but a restrictive licence will give you
396legal security. License your software and pepper it with threatening
397statements like "This is unpublished proprietary software of XYZ Corp.
398Your access to it does not give you permission to use it blah blah
399blah." You should see a lawyer to be sure your licence's wording will
400stand up in court.
5a964f20 401
0d7c09bb
JH
402=head2 Unicode
403
404Unicode is a new and complex technology and one may easily overlook
405certain security pitfalls. See L<perluniintro> for an overview and
406L<perlunicode> for details, and L<perlunicode/"Security Implications
407of Unicode"> for security implications in particular.
408
504f80c1
JH
409=head2 Algorithmic Complexity Attacks
410
411Certain internal algorithms used in the implementation of Perl can
412be attacked by choosing the input carefully to consume large amounts
413of either time or space or both. This can lead into the so-called
414I<Denial of Service> (DoS) attacks.
415
416=over 4
417
418=item *
419
420Hash Function - the algorithm used to "order" hash elements has been
421changed several times during the development of Perl, mainly to be
422reasonably fast. In Perl 5.8.1 also the security aspect was taken
423into account.
424
425In Perls before 5.8.1 one could rather easily generate data that as
426hash keys would cause Perl to consume large amounts of time because
4546b9e6
JH
427internal structure of hashes would badly degenerate. In Perl 5.8.1
428the hash function is randomly perturbed by a pseudorandom seed which
429makes generating such naughty hash keys harder.
430See L<perlrun/PERL_HASH_SEED> for more information.
431
432The random perturbation is done by default but if one wants for some
433reason emulate the old behaviour one can set the environment variable
434PERL_HASH_SEED to zero (or any other integer). One possible reason
435for wanting to emulate the old behaviour is that in the new behaviour
436consecutive runs of Perl will order hash keys differently, which may
437confuse some applications (like Data::Dumper: the outputs of two
438different runs are no more identical).
504f80c1 439
7b3f7037
JH
440B<Perl has never guaranteed any ordering of the hash keys>, and the
441ordering has already changed several times during the lifetime of
442Perl 5. Also, the ordering of hash keys has always been, and
443continues to be, affected by the insertion order.
444
445Also note that while the order of the hash elements might be
446randomised, this "pseudoordering" should B<not> be used for
447applications like shuffling a list randomly (use List::Util::shuffle()
448for that, see L<List::Util>, a standard core module since Perl 5.8.0;
449or the CPAN module Algorithm::Numerical::Shuffle), or for generating
450permutations (use e.g. the CPAN modules Algorithm::Permute or
451Algorithm::FastPermute), or for any cryptographic applications.
452
504f80c1
JH
453=item *
454
455Regular expressions - Perl's regular expression engine is so called
456NFA (Non-Finite Automaton), which among other things means that it can
457rather easily consume large amounts of both time and space if the
458regular expression may match in several ways. Careful crafting of the
459regular expressions can help but quite often there really isn't much
460one can do (the book "Mastering Regular Expressions" is required
461reading, see L<perlfaq2>). Running out of space manifests itself by
462Perl running out of memory.
463
464=item *
465
466Sorting - the quicksort algorithm used in Perls before 5.8.0 to
467implement the sort() function is very easy to trick into misbehaving
468so that it consumes a lot of time. Nothing more is required than
469resorting a list already sorted. Starting from Perl 5.8.0 a different
470sorting algorithm, mergesort, is used. Mergesort is insensitive to
471its input data, so it cannot be similarly fooled.
472
473=back
474
475See L<http://www.cs.rice.edu/~scrosby/hash/> for more information,
476and any computer science text book on the algorithmic complexity.
477
5a964f20
TC
478=head1 SEE ALSO
479
480L<perlrun> for its description of cleaning up environment variables.