xref: /netbsd-src/lib/libc/sys/execve.2 (revision 7330f729ccf0bd976a06f95fad452fe774fc7fd1)
1.\"	$NetBSD: execve.2,v 1.45 2019/09/18 04:57:53 wiz Exp $
2.\"
3.\" Copyright (c) 1980, 1991, 1993
4.\"	The Regents of the University of California.  All rights reserved.
5.\"
6.\" Redistribution and use in source and binary forms, with or without
7.\" modification, are permitted provided that the following conditions
8.\" are met:
9.\" 1. Redistributions of source code must retain the above copyright
10.\"    notice, this list of conditions and the following disclaimer.
11.\" 2. Redistributions in binary form must reproduce the above copyright
12.\"    notice, this list of conditions and the following disclaimer in the
13.\"    documentation and/or other materials provided with the distribution.
14.\" 3. Neither the name of the University nor the names of its contributors
15.\"    may be used to endorse or promote products derived from this software
16.\"    without specific prior written permission.
17.\"
18.\" THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
19.\" ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
20.\" IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
21.\" ARE DISCLAIMED.  IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
22.\" FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
23.\" DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
24.\" OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
25.\" HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
26.\" LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
27.\" OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
28.\" SUCH DAMAGE.
29.\"
30.\"     @(#)execve.2	8.5 (Berkeley) 6/1/94
31.\"
32.Dd September 16, 2019
33.Dt EXECVE 2
34.Os
35.Sh NAME
36.Nm execve ,
37.Nm fexecve
38.Nd execute a file
39.Sh LIBRARY
40.Lb libc
41.Sh SYNOPSIS
42.In unistd.h
43.Ft int
44.Fn execve "const char *path" "char *const argv[]" "char *const envp[]"
45.Ft int
46.Fn fexecve "int fd" "char *const argv[]" "char *const envp[]"
47.Sh DESCRIPTION
48The
49.Fn execve
50system call
51transforms the calling process into a new process.
52The new process is constructed from an ordinary file,
53whose name is pointed to by
54.Fa path ,
55called the
56.Em new process file .
57The
58.Fn fexecve
59system call is equivalent to
60.Fn execve
61except that the file to be executed is determined by the file
62descriptor
63.Fa fd
64instead of a
65.Fa path .
66This file is either an executable object file,
67or a file of data for an interpreter.
68An executable object file consists of an identifying header,
69followed by pages of data representing the initial program (text)
70and initialized data pages.
71Additional pages may be specified
72by the header to be initialized with zero data; see
73.Xr elf 5
74and
75.Xr a.out 5 .
76.Pp
77An interpreter file begins with a line of the form:
78.Pp
79.Bd -ragged -offset indent -compact
80.Sy \&#!
81.Em interpreter
82.Bq Em arg
83.Ed
84.Pp
85When an interpreter file is
86.Sy execve Ar d ,
87the system actually
88.Sy execve Ap s
89the specified
90.Em interpreter .
91If the optional
92.Em arg
93is specified, it becomes the first argument to the
94.Em interpreter ,
95and the name of the originally
96.Sy execve Ap d
97file becomes the second argument;
98otherwise, the name of the originally
99.Sy execve Ap d
100file becomes the first argument.
101The original arguments are shifted over to become the subsequent arguments.
102The zeroth argument, normally the name of the
103.Fn execve Ns d
104file, is left unchanged.
105The interpreter named by
106.Em interpreter
107must not itself be an interpreter file.
108(See
109.Xr script 7
110for a detailed discussion of interpreter file execution.)
111.Pp
112The argument
113.Fa argv
114is a pointer to a null-terminated array of
115character pointers to null-terminated character strings.
116These strings construct the argument list to be made available to the new
117process.
118At least one argument must be present in
119the array; by custom, the first element should be
120the name of the executed program (for example, the last component of
121.Fa path ) .
122.Pp
123The argument
124.Fa envp
125is also a pointer to a null-terminated array of
126character pointers to null-terminated strings.
127A pointer to this array is normally stored in the global variable
128.Va environ .
129These strings pass information to the
130new process that is not directly an argument to the command (see
131.Xr environ 7 ) .
132.Pp
133File descriptors open in the calling process image remain open in
134the new process image, except for those for which the close-on-exec
135flag is set (see
136.Xr close 2
137and
138.Xr fcntl 2 ) .
139Descriptors that remain open are unaffected by
140.Fn execve .
141.Pp
142In the case of a new setuid or setgid executable being executed, if
143file descriptors 0, 1, or 2 (representing stdin, stdout, and stderr)
144are currently unallocated, these descriptors will be opened to point to
145some system file like
146.Pa /dev/null .
147The intent is to ensure these descriptors are not unallocated, since
148many libraries make assumptions about the use of these 3 file descriptors.
149.Pp
150Signals set to be ignored in the calling process are set to be ignored in
151the new process.
152Signals which are set to be caught in the calling process image
153are set to default action in the new process image.
154Blocked signals remain blocked regardless of changes to the signal action.
155The signal stack is reset to be undefined (see
156.Xr sigaction 2
157for more information).
158.Pp
159If the set-user-ID mode bit of the new process image file is set
160(see
161.Xr chmod 2 ) ,
162the effective user ID of the new process image is set to the owner ID
163of the new process image file.
164If the set-group-ID mode bit of the new process image file is set,
165the effective group ID of the new process image is set to the group ID
166of the new process image file.
167(The effective group ID is the first element of the group list.)
168The real user ID, real group ID and
169other group IDs of the new process image remain the same as the calling
170process image.
171After any set-user-ID and set-group-ID processing,
172the effective user ID is recorded as the saved set-user-ID,
173and the effective group ID is recorded as the saved set-group-ID.
174These values may be used in changing the effective IDs later (see
175.Xr setuid 2 ) .
176The set-ID bits are not honored if the respective file system has the
177.Cm nosuid
178option enabled or if the new process file is an interpreter file.
179Syscall
180tracing is disabled if effective IDs are changed.
181.Pp
182The new process also inherits the following attributes from
183the calling process:
184.Pp
185.Bl -column parent_process_ID -offset indent -compact
186.It process ID Ta see Xr getpid 2
187.It parent process ID Ta see Xr getppid 2
188.It process group ID Ta see Xr getpgrp 2
189.It access groups Ta see Xr getgroups 2
190.It working directory Ta see Xr chdir 2
191.It root directory Ta see Xr chroot 2
192.It control terminal Ta see Xr termios 4
193.It resource usages Ta see Xr getrusage 2
194.It interval timers Ta see Xr getitimer 2
195.It resource limits Ta see Xr getrlimit 2
196.It file mode mask Ta see Xr umask 2
197.It signal mask Ta see Xr sigaction 2 ,
198.Xr sigprocmask 2
199.El
200.Pp
201When a program is executed as a result of an
202.Fn execve
203system call, it is entered as follows:
204.Bd -literal -offset indent
205main(argc, argv, envp)
206int argc;
207char **argv, **envp;
208.Ed
209.Pp
210where
211.Fa argc
212is the number of elements in
213.Fa argv
214(the
215.Dq arg count )
216and
217.Fa argv
218points to the array of character pointers
219to the arguments themselves.
220.Pp
221The
222.Fn fexecve
223function ignores the file offset of
224.Fa fd .
225Since execute permission is checked by
226.Fn fexecve ,
227the file descriptor
228.Fa fd
229need not have been opened with the
230.Dv O_EXEC
231flag.
232However, if the file to be executed denies read permission for the process
233preparing to do the exec, the only way to provide the
234.Fa fd
235to
236.Fn fexecve
237is to use the
238.Dv O_EXEC
239flag when opening
240.Fa fd .
241Note that the file to be executed can not be open for writing.
242.Sh RETURN VALUES
243As the
244.Fn execve
245system call overlays the current process image
246with a new process image the successful call
247has no process to return to.
248If
249.Fn execve
250does return to the calling process an error has occurred; the
251return value will be \-1 and the global variable
252.Va errno
253is set to indicate the error.
254.Sh ERRORS
255The
256.Fn execve
257system call
258will fail and return to the calling process if:
259.Bl -tag -width Er
260.It Bq Er E2BIG
261The number of bytes in the new process' argument list
262is larger than the system-imposed limit.
263The default compile time limit is 262144 bytes and is specified
264in the variable
265.Dv NCARGS
266in
267.Aq Pa sys/param.h
268and get be read from the
269.Xr sysctl 3
270MIB variable
271.Dv KERN_ARGMAX .
272.It Bq Er EACCES
273Search permission is denied for a component of the path prefix,
274the new process file is not an ordinary file,
275its file mode denies execute permission, or
276it is on a file system mounted with execution
277disabled
278.Dv ( MNT_NOEXEC
279in
280.Ao Pa sys/mount.h Ac ) .
281.It Bq Er EAGAIN
282A
283.Xr setuid 7
284process has exceeded the current resource limit for the number of
285processes it is allowed to run concurrently.
286.It Bq Er EFAULT
287The new process file is not as long as indicated by
288the size values in its header; or
289the
290.Fa path ,
291.Fa argv ,
292or
293.Fa envp
294arguments point to an illegal address.
295.It Bq Er EIO
296An I/O error occurred while reading from the file system.
297.It Bq Er ELOOP
298Too many symbolic links were encountered in translating the pathname.
299.It Bq Er ENAMETOOLONG
300A component of a pathname exceeded
301.Brq Dv NAME_MAX
302characters, or an entire path name exceeded
303.Brq Dv PATH_MAX
304characters.
305.It Bq Er ENOENT
306The new process file does not exist, or
307the new process file is a script starting with
308.Li #!
309and the script interpreter does not exist.
310.It Bq Er ENOEXEC
311The new process file has the appropriate access
312permission, but has an invalid magic number in its header.
313.It Bq Er ENOMEM
314The new process requires more virtual memory than
315is allowed by the imposed maximum
316.Pq Xr getrlimit 2 .
317.It Bq Er ENOTDIR
318A component of the path prefix is not a directory.
319.It Bq Er ETXTBSY
320The new process file is a pure procedure (shared text)
321file that is currently open for writing or reading by some process.
322.El
323.Pp
324In addition, the
325.Fn fexecve
326will fail and return to the calling process if:
327.Bl -tag -width Er
328.It Bq Er EBADF
329The
330.Fa fd
331argument is not a valid file descriptor open for executing.
332.El
333.Sh SEE ALSO
334.Xr _exit 2 ,
335.Xr fork 2 ,
336.Xr open 2 ,
337.Xr execl 3 ,
338.Xr exit 3 ,
339.Xr sysctl 3 ,
340.Xr a.out 5 ,
341.Xr elf 5 ,
342.\" .Xr fdescfs 5 ,
343.Xr environ 7 ,
344.Xr script 7 ,
345.Xr mount 8
346.Sh STANDARDS
347The
348.Fn execve
349system call conforms to
350.St -p1003.1-2001 .
351with the exception of reopening descriptors 0, 1, and/or 2 in certain
352circumstances.
353A future update of the Standard is expected to require this behavior,
354and it may become the default for non-privileged processes as well.
355.\" NB: update this caveat when TC1 is blessed.
356The support for executing interpreted programs is an extension.
357The
358.Fn fexecve
359system call conforms to The Open Group Extended API Set 2 specification.
360.Sh HISTORY
361The
362.Fn execve
363function call first appeared in
364.At v7 .
365The
366.Fn fexecve
367system call appeared in
368.Nx 10.0 .
369.Sh BUGS
370If a program is
371.Em setuid
372to a non-super-user, but is executed when
373the real
374.Em uid
375is
376.Dq root ,
377then the program has some of the powers of a super-user as well.
378.\" .Pp
379.\" When executing an interpreted program through
380.\" .Fn fexecve ,
381.\" kernel supplies
382.\" .Pa /dev/fd/n
383.\" as a second argument to the interpreter,
384.\" where
385.\" .Ar n
386.\" is the file descriptor passed in the
387.\" .Fa fd
388.\" argument to
389.\" .Fn fexecve .
390.\" For this construction to work correctly, the
391.\" .Xr fdescfs 5
392.\" filesystem shall be mounted on
393.\" .Pa /dev/fd .
394