1system(3) Library Functions Manual system(3)
2
3
4
6 system - execute a shell command
7
9 Standard C library (libc, -lc)
10
12 #include <stdlib.h>
13
14 int system(const char *command);
15
17 The system() library function behaves as if it used fork(2) to create a
18 child process that executed the shell command specified in command us‐
19 ing execl(3) as follows:
20
21 execl("/bin/sh", "sh", "-c", command, (char *) NULL);
22
23 system() returns after the command has been completed.
24
25 During execution of the command, SIGCHLD will be blocked, and SIGINT
26 and SIGQUIT will be ignored, in the process that calls system().
27 (These signals will be handled according to their defaults inside the
28 child process that executes command.)
29
30 If command is NULL, then system() returns a status indicating whether a
31 shell is available on the system.
32
34 The return value of system() is one of the following:
35
36 • If command is NULL, then a nonzero value if a shell is available, or
37 0 if no shell is available.
38
39 • If a child process could not be created, or its status could not be
40 retrieved, the return value is -1 and errno is set to indicate the
41 error.
42
43 • If a shell could not be executed in the child process, then the re‐
44 turn value is as though the child shell terminated by calling
45 _exit(2) with the status 127.
46
47 • If all system calls succeed, then the return value is the termina‐
48 tion status of the child shell used to execute command. (The termi‐
49 nation status of a shell is the termination status of the last com‐
50 mand it executes.)
51
52 In the last two cases, the return value is a "wait status" that can be
53 examined using the macros described in waitpid(2). (i.e., WIFEXITED(),
54 WEXITSTATUS(), and so on).
55
56 system() does not affect the wait status of any other children.
57
59 system() can fail with any of the same errors as fork(2).
60
62 For an explanation of the terms used in this section, see at‐
63 tributes(7).
64
65 ┌────────────────────────────────────────────┬───────────────┬─────────┐
66 │Interface │ Attribute │ Value │
67 ├────────────────────────────────────────────┼───────────────┼─────────┤
68 │system() │ Thread safety │ MT-Safe │
69 └────────────────────────────────────────────┴───────────────┴─────────┘
70
72 C11, POSIX.1-2008.
73
75 POSIX.1-2001, C89.
76
78 system() provides simplicity and convenience: it handles all of the
79 details of calling fork(2), execl(3), and waitpid(2), as well as the
80 necessary manipulations of signals; in addition, the shell performs the
81 usual substitutions and I/O redirections for command. The main cost of
82 system() is inefficiency: additional system calls are required to
83 create the process that runs the shell and to execute the shell.
84
85 If the _XOPEN_SOURCE feature test macro is defined (before including
86 any header files), then the macros described in waitpid(2)
87 (WEXITSTATUS(), etc.) are made available when including <stdlib.h>.
88
89 As mentioned, system() ignores SIGINT and SIGQUIT. This may make
90 programs that call it from a loop uninterruptible, unless they take
91 care themselves to check the exit status of the child. For example:
92
93 while (something) {
94 int ret = system("foo");
95
96 if (WIFSIGNALED(ret) &&
97 (WTERMSIG(ret) == SIGINT || WTERMSIG(ret) == SIGQUIT))
98 break;
99 }
100
101 According to POSIX.1, it is unspecified whether handlers registered us‐
102 ing pthread_atfork(3) are called during the execution of system(). In
103 the glibc implementation, such handlers are not called.
104
105 Before glibc 2.1.3, the check for the availability of /bin/sh was not
106 actually performed if command was NULL; instead it was always assumed
107 to be available, and system() always returned 1 in this case. Since
108 glibc 2.1.3, this check is performed because, even though POSIX.1-2001
109 requires a conforming implementation to provide a shell, that shell may
110 not be available or executable if the calling program has previously
111 called chroot(2) (which is not specified by POSIX.1-2001).
112
113 It is possible for the shell command to terminate with a status of 127,
114 which yields a system() return value that is indistinguishable from the
115 case where a shell could not be executed in the child process.
116
117 Caveats
118 Do not use system() from a privileged program (a set-user-ID or set-
119 group-ID program, or a program with capabilities) because strange val‐
120 ues for some environment variables might be used to subvert system in‐
121 tegrity. For example, PATH could be manipulated so that an arbitrary
122 program is executed with privilege. Use the exec(3) family of func‐
123 tions instead, but not execlp(3) or execvp(3) (which also use the PATH
124 environment variable to search for an executable).
125
126 system() will not, in fact, work properly from programs with set-user-
127 ID or set-group-ID privileges on systems on which /bin/sh is bash ver‐
128 sion 2: as a security measure, bash 2 drops privileges on startup.
129 (Debian uses a different shell, dash(1), which does not do this when
130 invoked as sh.)
131
132 Any user input that is employed as part of command should be carefully
133 sanitized, to ensure that unexpected shell commands or command options
134 are not executed. Such risks are especially grave when using system()
135 from a privileged program.
136
138 If the command name starts with a hyphen, sh(1) interprets the command
139 name as an option, and the behavior is undefined. (See the -c option
140 to sh(1).) To work around this problem, prepend the command with a
141 space as in the following call:
142
143 system(" -unfortunate-command-name");
144
146 sh(1), execve(2), fork(2), sigaction(2), sigprocmask(2), wait(2),
147 exec(3), signal(7)
148
149
150
151Linux man-pages 6.05 2023-07-20 system(3)