Commit | Line | Data |
a0d0e21e |
1 | =head1 NAME |
2 | |
3 | perlsec - Perl security |
4 | |
5 | =head1 DESCRIPTION |
6 | |
425e5e39 |
7 | Perl is designed to make it easy to program securely even when running |
8 | with extra privileges, like setuid or setgid programs. Unlike most |
54310121 |
9 | command line shells, which are based on multiple substitution passes on |
425e5e39 |
10 | each line of the script, Perl uses a more conventional evaluation scheme |
11 | with fewer hidden snags. Additionally, because the language has more |
54310121 |
12 | builtin functionality, it can rely less upon external (and possibly |
425e5e39 |
13 | untrustworthy) programs to accomplish its purposes. |
a0d0e21e |
14 | |
425e5e39 |
15 | Perl automatically enables a set of special security checks, called I<taint |
16 | mode>, when it detects its program running with differing real and effective |
17 | user or group IDs. The setuid bit in Unix permissions is mode 04000, the |
18 | setgid bit mode 02000; either or both may be set. You can also enable taint |
5f05dabc |
19 | mode explicitly by using the B<-T> command line flag. This flag is |
425e5e39 |
20 | I<strongly> suggested for server programs and any program run on behalf of |
fb73857a |
21 | someone else, such as a CGI script. Once taint mode is on, it's on for |
22 | the remainder of your script. |
a0d0e21e |
23 | |
1e422769 |
24 | While in this mode, Perl takes special precautions called I<taint |
25 | checks> to prevent both obvious and subtle traps. Some of these checks |
26 | are reasonably simple, such as verifying that path directories aren't |
27 | writable by others; careful programmers have always used checks like |
28 | these. Other checks, however, are best supported by the language itself, |
fb73857a |
29 | and it is these checks especially that contribute to making a set-id Perl |
425e5e39 |
30 | program more secure than the corresponding C program. |
31 | |
fb73857a |
32 | You may not use data derived from outside your program to affect |
33 | something else outside your program--at least, not by accident. All |
34 | command line arguments, environment variables, locale information (see |
d929ce6f |
35 | L<perllocale>), results of certain system calls (readdir(), |
41d6edb2 |
36 | readlink(), the variable of shmread(), the messages returned by |
37 | msgrcv(), the password, gcos and shell fields returned by the |
38 | getpwxxx() calls), and all file input are marked as "tainted". |
39 | Tainted data may not be used directly or indirectly in any command |
40 | that invokes a sub-shell, nor in any command that modifies files, |
b7ee89ce |
41 | directories, or processes, B<with the following exceptions>: |
42 | |
43 | =over 4 |
44 | |
45 | =item * |
46 | |
ee556d55 |
47 | If you pass more than one argument to either C<system> or C<exec>, |
48 | the arguments are B<not> checked for taintedness. |
b7ee89ce |
49 | |
50 | =item * |
51 | |
52 | Arguments to C<print> and C<syswrite> are B<not> checked for taintedness. |
53 | |
54 | =back |
55 | |
ee556d55 |
56 | The value of an expression containing tainted data will itself be |
57 | tainted, even if it is logically impossible for the tainted data to |
58 | affect the value. |
59 | |
d929ce6f |
60 | Because taintedness is associated with each scalar value, some |
61 | elements of an array can be tainted and others not. |
a0d0e21e |
62 | |
a0d0e21e |
63 | For example: |
64 | |
425e5e39 |
65 | $arg = shift; # $arg is tainted |
66 | $hid = $arg, 'bar'; # $hid is also tainted |
67 | $line = <>; # Tainted |
8ebc5c01 |
68 | $line = <STDIN>; # Also tainted |
69 | open FOO, "/home/me/bar" or die $!; |
70 | $line = <FOO>; # Still tainted |
a0d0e21e |
71 | $path = $ENV{'PATH'}; # Tainted, but see below |
425e5e39 |
72 | $data = 'abc'; # Not tainted |
a0d0e21e |
73 | |
425e5e39 |
74 | system "echo $arg"; # Insecure |
75 | system "/bin/echo", $arg; # Secure (doesn't use sh) |
76 | system "echo $hid"; # Insecure |
77 | system "echo $data"; # Insecure until PATH set |
a0d0e21e |
78 | |
425e5e39 |
79 | $path = $ENV{'PATH'}; # $path now tainted |
a0d0e21e |
80 | |
54310121 |
81 | $ENV{'PATH'} = '/bin:/usr/bin'; |
c90c0ff4 |
82 | delete @ENV{'IFS', 'CDPATH', 'ENV', 'BASH_ENV'}; |
a0d0e21e |
83 | |
425e5e39 |
84 | $path = $ENV{'PATH'}; # $path now NOT tainted |
85 | system "echo $data"; # Is secure now! |
a0d0e21e |
86 | |
425e5e39 |
87 | open(FOO, "< $arg"); # OK - read-only file |
88 | open(FOO, "> $arg"); # Not OK - trying to write |
a0d0e21e |
89 | |
425e5e39 |
90 | open(FOO,"echo $arg|"); # Not OK, but... |
91 | open(FOO,"-|") |
92 | or exec 'echo', $arg; # OK |
a0d0e21e |
93 | |
425e5e39 |
94 | $shout = `echo $arg`; # Insecure, $shout now tainted |
a0d0e21e |
95 | |
425e5e39 |
96 | unlink $data, $arg; # Insecure |
97 | umask $arg; # Insecure |
a0d0e21e |
98 | |
3f7d42d8 |
99 | exec "echo $arg"; # Insecure (uses the shell) |
425e5e39 |
100 | exec "echo", $arg; # Secure (doesn't use the shell) |
101 | exec "sh", '-c', $arg; # Considered secure, alas! |
a0d0e21e |
102 | |
3a4b19e4 |
103 | @files = <*.c>; # insecure (uses readdir() or similar) |
104 | @files = glob('*.c'); # insecure (uses readdir() or similar) |
7bac28a0 |
105 | |
3f7d42d8 |
106 | # In Perl releases older than 5.6.0 the <*.c> and glob('*.c') would |
107 | # have used an external program to do the filename expansion; but in |
108 | # either case the result is tainted since the list of filenames comes |
109 | # from outside of the program. |
110 | |
ee556d55 |
111 | $bad = ($arg, 23); # $bad will be tainted |
112 | $arg, `true`; # Insecure (although it isn't really) |
113 | |
a0d0e21e |
114 | If you try to do something insecure, you will get a fatal error saying |
62f468fc |
115 | something like "Insecure dependency" or "Insecure $ENV{PATH}". Note that you |
425e5e39 |
116 | can still write an insecure B<system> or B<exec>, but only by explicitly |
a3cb178b |
117 | doing something like the "considered secure" example above. |
425e5e39 |
118 | |
119 | =head2 Laundering and Detecting Tainted Data |
120 | |
3f7d42d8 |
121 | To test whether a variable contains tainted data, and whose use would |
122 | thus trigger an "Insecure dependency" message, you can use the |
123 | tainted() function of the Scalar::Util module, available in your |
124 | nearby CPAN mirror, and included in Perl starting from the release 5.8.0. |
125 | Or you may be able to use the following I<is_tainted()> function. |
425e5e39 |
126 | |
127 | sub is_tainted { |
54310121 |
128 | return ! eval { |
129 | join('',@_), kill 0; |
130 | 1; |
425e5e39 |
131 | }; |
132 | } |
133 | |
134 | This function makes use of the fact that the presence of tainted data |
135 | anywhere within an expression renders the entire expression tainted. It |
136 | would be inefficient for every operator to test every argument for |
137 | taintedness. Instead, the slightly more efficient and conservative |
138 | approach is used that if any tainted value has been accessed within the |
139 | same expression, the whole expression is considered tainted. |
140 | |
5f05dabc |
141 | But testing for taintedness gets you only so far. Sometimes you have just |
425e5e39 |
142 | to clear your data's taintedness. The only way to bypass the tainting |
54310121 |
143 | mechanism is by referencing subpatterns from a regular expression match. |
425e5e39 |
144 | Perl presumes that if you reference a substring using $1, $2, etc., that |
145 | you knew what you were doing when you wrote the pattern. That means using |
146 | a bit of thought--don't just blindly untaint anything, or you defeat the |
a034a98d |
147 | entire mechanism. It's better to verify that the variable has only good |
148 | characters (for certain values of "good") rather than checking whether it |
149 | has any bad characters. That's because it's far too easy to miss bad |
150 | characters that you never thought of. |
425e5e39 |
151 | |
152 | Here's a test to make sure that the data contains nothing but "word" |
153 | characters (alphabetics, numerics, and underscores), a hyphen, an at sign, |
154 | or a dot. |
155 | |
54310121 |
156 | if ($data =~ /^([-\@\w.]+)$/) { |
425e5e39 |
157 | $data = $1; # $data now untainted |
158 | } else { |
159 | die "Bad data in $data"; # log this somewhere |
160 | } |
161 | |
5f05dabc |
162 | This is fairly secure because C</\w+/> doesn't normally match shell |
425e5e39 |
163 | metacharacters, nor are dot, dash, or at going to mean something special |
164 | to the shell. Use of C</.+/> would have been insecure in theory because |
165 | it lets everything through, but Perl doesn't check for that. The lesson |
166 | is that when untainting, you must be exceedingly careful with your patterns. |
19799a22 |
167 | Laundering data using regular expression is the I<only> mechanism for |
425e5e39 |
168 | untainting dirty data, unless you use the strategy detailed below to fork |
169 | a child of lesser privilege. |
170 | |
a034a98d |
171 | The example does not untaint $data if C<use locale> is in effect, |
172 | because the characters matched by C<\w> are determined by the locale. |
173 | Perl considers that locale definitions are untrustworthy because they |
174 | contain data from outside the program. If you are writing a |
175 | locale-aware program, and want to launder data with a regular expression |
176 | containing C<\w>, put C<no locale> ahead of the expression in the same |
177 | block. See L<perllocale/SECURITY> for further discussion and examples. |
178 | |
3a52c276 |
179 | =head2 Switches On the "#!" Line |
180 | |
181 | When you make a script executable, in order to make it usable as a |
182 | command, the system will pass switches to perl from the script's #! |
54310121 |
183 | line. Perl checks that any command line switches given to a setuid |
3a52c276 |
184 | (or setgid) script actually match the ones set on the #! line. Some |
54310121 |
185 | Unix and Unix-like environments impose a one-switch limit on the #! |
3a52c276 |
186 | line, so you may need to use something like C<-wU> instead of C<-w -U> |
54310121 |
187 | under such systems. (This issue should arise only in Unix or |
188 | Unix-like environments that support #! and setuid or setgid scripts.) |
3a52c276 |
189 | |
425e5e39 |
190 | =head2 Cleaning Up Your Path |
191 | |
1fef88e7 |
192 | For "Insecure C<$ENV{PATH}>" messages, you need to set C<$ENV{'PATH'}> to a |
1e422769 |
193 | known value, and each directory in the path must be non-writable by others |
194 | than its owner and group. You may be surprised to get this message even |
195 | if the pathname to your executable is fully qualified. This is I<not> |
196 | generated because you didn't supply a full path to the program; instead, |
197 | it's generated because you never set your PATH environment variable, or |
198 | you didn't set it to something that was safe. Because Perl can't |
199 | guarantee that the executable in question isn't itself going to turn |
200 | around and execute some other program that is dependent on your PATH, it |
54310121 |
201 | makes sure you set the PATH. |
a0d0e21e |
202 | |
a3cb178b |
203 | The PATH isn't the only environment variable which can cause problems. |
204 | Because some shells may use the variables IFS, CDPATH, ENV, and |
205 | BASH_ENV, Perl checks that those are either empty or untainted when |
206 | starting subprocesses. You may wish to add something like this to your |
207 | setid and taint-checking scripts. |
208 | |
209 | delete @ENV{qw(IFS CDPATH ENV BASH_ENV)}; # Make %ENV safer |
210 | |
a0d0e21e |
211 | It's also possible to get into trouble with other operations that don't |
212 | care whether they use tainted values. Make judicious use of the file |
213 | tests in dealing with any user-supplied filenames. When possible, do |
fb73857a |
214 | opens and such B<after> properly dropping any special user (or group!) |
215 | privileges. Perl doesn't prevent you from opening tainted filenames for reading, |
a0d0e21e |
216 | so be careful what you print out. The tainting mechanism is intended to |
217 | prevent stupid mistakes, not to remove the need for thought. |
218 | |
425e5e39 |
219 | Perl does not call the shell to expand wild cards when you pass B<system> |
220 | and B<exec> explicit parameter lists instead of strings with possible shell |
221 | wildcards in them. Unfortunately, the B<open>, B<glob>, and |
54310121 |
222 | backtick functions provide no such alternate calling convention, so more |
223 | subterfuge will be required. |
425e5e39 |
224 | |
225 | Perl provides a reasonably safe way to open a file or pipe from a setuid |
226 | or setgid program: just create a child process with reduced privilege who |
227 | does the dirty work for you. First, fork a child using the special |
228 | B<open> syntax that connects the parent and child by a pipe. Now the |
229 | child resets its ID set and any other per-process attributes, like |
230 | environment variables, umasks, current working directories, back to the |
231 | originals or known safe values. Then the child process, which no longer |
232 | has any special permissions, does the B<open> or other system call. |
233 | Finally, the child passes the data it managed to access back to the |
5f05dabc |
234 | parent. Because the file or pipe was opened in the child while running |
425e5e39 |
235 | under less privilege than the parent, it's not apt to be tricked into |
236 | doing something it shouldn't. |
237 | |
54310121 |
238 | Here's a way to do backticks reasonably safely. Notice how the B<exec> is |
425e5e39 |
239 | not called with a string that the shell could expand. This is by far the |
240 | best way to call something that might be subjected to shell escapes: just |
fb73857a |
241 | never call the shell at all. |
cb1a09d0 |
242 | |
e093bcf0 |
243 | use English; |
244 | die "Can't fork: $!" unless defined($pid = open(KID, "-|")); |
245 | if ($pid) { # parent |
246 | while (<KID>) { |
247 | # do something |
248 | } |
249 | close KID; |
250 | } else { |
251 | my @temp = ($EUID, $EGID); |
252 | my $orig_uid = $UID; |
253 | my $orig_gid = $GID; |
254 | $EUID = $UID; |
255 | $EGID = $GID; |
256 | # Drop privileges |
257 | $UID = $orig_uid; |
258 | $GID = $orig_gid; |
259 | # Make sure privs are really gone |
260 | ($EUID, $EGID) = @temp; |
261 | die "Can't drop privileges" |
262 | unless $UID == $EUID && $GID eq $EGID; |
263 | $ENV{PATH} = "/bin:/usr/bin"; # Minimal PATH. |
264 | # Consider sanitizing the environment even more. |
265 | exec 'myprog', 'arg1', 'arg2' |
266 | or die "can't exec myprog: $!"; |
267 | } |
425e5e39 |
268 | |
fb73857a |
269 | A similar strategy would work for wildcard expansion via C<glob>, although |
270 | you can use C<readdir> instead. |
425e5e39 |
271 | |
272 | Taint checking is most useful when although you trust yourself not to have |
273 | written a program to give away the farm, you don't necessarily trust those |
274 | who end up using it not to try to trick it into doing something bad. This |
fb73857a |
275 | is the kind of security checking that's useful for set-id programs and |
425e5e39 |
276 | programs launched on someone else's behalf, like CGI programs. |
277 | |
278 | This is quite different, however, from not even trusting the writer of the |
279 | code not to try to do something evil. That's the kind of trust needed |
280 | when someone hands you a program you've never seen before and says, "Here, |
281 | run this." For that kind of safety, check out the Safe module, |
282 | included standard in the Perl distribution. This module allows the |
283 | programmer to set up special compartments in which all system operations |
284 | are trapped and namespace access is carefully controlled. |
285 | |
286 | =head2 Security Bugs |
287 | |
288 | Beyond the obvious problems that stem from giving special privileges to |
fb73857a |
289 | systems as flexible as scripts, on many versions of Unix, set-id scripts |
425e5e39 |
290 | are inherently insecure right from the start. The problem is a race |
291 | condition in the kernel. Between the time the kernel opens the file to |
fb73857a |
292 | see which interpreter to run and when the (now-set-id) interpreter turns |
425e5e39 |
293 | around and reopens the file to interpret it, the file in question may have |
294 | changed, especially if you have symbolic links on your system. |
295 | |
296 | Fortunately, sometimes this kernel "feature" can be disabled. |
297 | Unfortunately, there are two ways to disable it. The system can simply |
fb73857a |
298 | outlaw scripts with any set-id bit set, which doesn't help much. |
299 | Alternately, it can simply ignore the set-id bits on scripts. If the |
425e5e39 |
300 | latter is true, Perl can emulate the setuid and setgid mechanism when it |
301 | notices the otherwise useless setuid/gid bits on Perl scripts. It does |
302 | this via a special executable called B<suidperl> that is automatically |
54310121 |
303 | invoked for you if it's needed. |
425e5e39 |
304 | |
fb73857a |
305 | However, if the kernel set-id script feature isn't disabled, Perl will |
306 | complain loudly that your set-id script is insecure. You'll need to |
307 | either disable the kernel set-id script feature, or put a C wrapper around |
425e5e39 |
308 | the script. A C wrapper is just a compiled program that does nothing |
309 | except call your Perl program. Compiled programs are not subject to the |
fb73857a |
310 | kernel bug that plagues set-id scripts. Here's a simple wrapper, written |
425e5e39 |
311 | in C: |
312 | |
313 | #define REAL_PATH "/path/to/script" |
54310121 |
314 | main(ac, av) |
425e5e39 |
315 | char **av; |
316 | { |
317 | execv(REAL_PATH, av); |
54310121 |
318 | } |
cb1a09d0 |
319 | |
54310121 |
320 | Compile this wrapper into a binary executable and then make I<it> rather |
321 | than your script setuid or setgid. |
425e5e39 |
322 | |
425e5e39 |
323 | In recent years, vendors have begun to supply systems free of this |
324 | inherent security bug. On such systems, when the kernel passes the name |
fb73857a |
325 | of the set-id script to open to the interpreter, rather than using a |
425e5e39 |
326 | pathname subject to meddling, it instead passes I</dev/fd/3>. This is a |
327 | special file already opened on the script, so that there can be no race |
328 | condition for evil scripts to exploit. On these systems, Perl should be |
329 | compiled with C<-DSETUID_SCRIPTS_ARE_SECURE_NOW>. The B<Configure> |
330 | program that builds Perl tries to figure this out for itself, so you |
331 | should never have to specify this yourself. Most modern releases of |
332 | SysVr4 and BSD 4.4 use this approach to avoid the kernel race condition. |
333 | |
0325b4c4 |
334 | Prior to release 5.6.1 of Perl, bugs in the code of B<suidperl> could |
335 | introduce a security hole. |
68dc0745 |
336 | |
337 | =head2 Protecting Your Programs |
338 | |
339 | There are a number of ways to hide the source to your Perl programs, |
340 | with varying levels of "security". |
341 | |
342 | First of all, however, you I<can't> take away read permission, because |
343 | the source code has to be readable in order to be compiled and |
344 | interpreted. (That doesn't mean that a CGI script's source is |
345 | readable by people on the web, though.) So you have to leave the |
5a964f20 |
346 | permissions at the socially friendly 0755 level. This lets |
347 | people on your local system only see your source. |
68dc0745 |
348 | |
5a964f20 |
349 | Some people mistakenly regard this as a security problem. If your program does |
68dc0745 |
350 | insecure things, and relies on people not knowing how to exploit those |
351 | insecurities, it is not secure. It is often possible for someone to |
352 | determine the insecure things and exploit them without viewing the |
353 | source. Security through obscurity, the name for hiding your bugs |
354 | instead of fixing them, is little security indeed. |
355 | |
83df6a1d |
356 | You can try using encryption via source filters (Filter::* from CPAN, |
357 | or Filter::Util::Call and Filter::Simple since Perl 5.8). |
358 | But crackers might be able to decrypt it. You can try using the byte |
359 | code compiler and interpreter described below, but crackers might be |
360 | able to de-compile it. You can try using the native-code compiler |
68dc0745 |
361 | described below, but crackers might be able to disassemble it. These |
362 | pose varying degrees of difficulty to people wanting to get at your |
363 | code, but none can definitively conceal it (this is true of every |
364 | language, not just Perl). |
365 | |
366 | If you're concerned about people profiting from your code, then the |
367 | bottom line is that nothing but a restrictive licence will give you |
368 | legal security. License your software and pepper it with threatening |
369 | statements like "This is unpublished proprietary software of XYZ Corp. |
370 | Your access to it does not give you permission to use it blah blah |
371 | blah." You should see a lawyer to be sure your licence's wording will |
372 | stand up in court. |
5a964f20 |
373 | |
374 | =head1 SEE ALSO |
375 | |
376 | L<perlrun> for its description of cleaning up environment variables. |