Jump to content United States-English
HP.com Home Products and Services Support and Drivers Solutions How to Buy
» Contact HP
More options
HP.com home
HP-UX Reference > E

exec(2)

HP-UX 11i Version 3: February 2007
» 

Technical documentation

» Feedback
Content starts here

 » Table of Contents

 » Index

NAME

exec: execl(), execle(), execlp(), execv(), execve(), execvp() — execute a file

SYNOPSIS

#include <unistd.h> extern char **environ; int execl(const char *path, const char *arg0, ... /* * [const char *arg1, ..., const char *argn,] */ (char *)0); int execle(const char *path, const char *arg0, ... /* * [const char *arg1, ..., const char *argn,] */ (char *)0, char * const envp[]); int execlp(const char *file, const char *arg0, ... /* * [const char *arg1, ..., const char *argn,] */ (char *)0); int execv(const char *path, char * const argv[]); int execve(const char *path, char * const argv[], char * const envp[]); int execvp(const char *file, char * const argv[]);

Remarks

The ANSI C ", ... " construct denotes a variable length argument list whose optional and required members are given in the associated comment (/* */).

DESCRIPTION

The exec*() system calls, in all their forms, load a program from an ordinary executable file into the current process, replacing the current program. The path or file argument refers to either an executable object file or a file of data for an interpreter. In the latter case, the file of data is also called a script file.

If the calling process is multi-threaded, a call to any of the exec functions will cause all threads and light weight processes in the calling process to be terminated and the new executable image to be loaded and executed. No thread specific data destructor functions are called. If the exec() function fails and returns to the caller, threads and light weight processes (LWPs) in the calling process will not be terminated.

An executable object file consists of a header (see a.out(4)), text segment, and data segment. The data segment contains an initialized portion and an uninitialized portion (bss). For execlp() and execvp() the POSIX shell (see sh-posix(1)) can be loaded to interpret a script instead. A successful call to exec*() does not return because the new program overwrites the calling program.

When a C program is executed, it is called as follows:

main (int argc, char **argv, char **envp)

where argc is the argument count and argv is the address of an array of character pointers to the arguments themselves. As indicated, argc usually has a value of at least one, and the first member of the array points to a string containing the name of the file. Exit conditions from main are discussed in exit(2).

path points to a path name that identifies the executable file containing the new program.

file (in execlp() or execvp()) points to a file name identifying the executable file containing the new program. The path prefix for this file is obtained by searching the directories passed in the environment variable PATH (see environ(5)). The environment is supplied by the shell (see sh(1)). If file does not have an executable magic number (see magic(4)), it is passed to the POSIX shell as a shell script.

arg0, ..., argn are one or more pointers to null-terminated character strings. These strings constitute the argument list available to the new program. By convention, at least arg0 must be present and point to a string identical to path or to path's last component.

argv is an array of character pointers to null-terminated strings. These strings constitute the argument list available to the new program. By convention, argv must have at least one member, and must point to a string that is identical to path or path's last component. argv is terminated by a null pointer.

envp is an array of character pointers to null-terminated strings. These strings constitute the environment in which the new program runs. envp is terminated by a null pointer. For execle() and execve(), the C run-time start-off routine places a pointer to the environment of the calling program in the global cell:

extern char **environ;

and it is used to pass the environment of the calling program to the new program.

Multi-threaded applications should not use the environ variable to access or modify any environment variable while another thread is concurrently modifying any environment variable. Calling any function which is dependent upon any environment variable is considered a use of the environ variable to access that environment variable.

Open file descriptors remain open, except for those whose close-on-exec flag is set (see fcntl(2)). The file offset, access mode, and status flags of open file descriptors are unchanged.

Note that normal executable files are open only briefly when they start execution. Other executable file types can be kept open for a long time, or even indefinitely under some circumstances.

The processing of signals by the process is unchanged by exec*(), except that signals caught by the process are set to their default values (see signal(2)).

If the set-user-ID mode bit of the executable file pointed to by path or file is set (see chmod(2)), exec*() sets the effective user ID of the new process to the user ID of the executable file. Similarly, if the set-group-ID mode bit of the executable file is set, the effective group ID of the process is set to the group ID of the executable file. The real user ID and real group ID of the process are unchanged. Normally, the set-user-ID and set-group-ID functions do not apply to scripts; thus, if execlp() or execvp() executes a script, the set-user-ID and set-group-ID bits are ignored, even if they are set. This behavior can be overridden by setting the kernel tunable secure_sid_scripts to zero.

The saved user ID and saved group ID of the process are always set to the effective user ID and effective group ID, respectively, of the process at the end of the exec*(), whether or not set-user-ID or set-group-ID is in effect.

The shared memory segments attached to the calling program are not attached to the new program (see shmop(2)).

Text and data segment memory locks are not passed on to the new program (see plock(2)).

Profiling is disabled for the new process (see profil(2)).

The process also retains the following attributes:

  • current working directory

  • file creation mode mask (see umask(2))

  • file locks (see fcntl(2)), except for files closed-on-execution

  • file size limit (see ulimit(2))

  • interval timers (see getitimer(2))

  • nice value (see nice(2))

  • nice value (see parent process ID)

  • pending signals

  • process ID

  • process group ID

  • real user ID

  • real group ID

  • process start time

  • real-time priority (see rtprio(2))

  • root directory (see chroot(2))

  • semadj values (see semop(2))

  • session membership

  • signal mask (see sigprocmask(2))

  • supplementary group IDs

  • time left until an alarm clock signal (see alarm(2))

  • trace flag (see ttrace(2) request TT_PROC_SETTRC)

  • tms_utime, tms_stime, tms_cutime, and tms_cstime (see times(2))

For a script file, the initial line of a script file must begin with #! as the first two bytes, followed by zero or more spaces, followed by interpreter or interpreter argument, as in:

#! interpreter [argument]

One or more spaces or tabs must separate interpreter and argument. The first line should end with either a newline or a null character.

When the script file is executed, the system executes the specified interpreter as an executable object file. Even in the case of execlp() or execvp(), no path searching is done of the interpreter name.

The argument is anything that follows interpreter and tabs or spaces. If an argument is given, it is passed to the interpreter as argv[1], and the name of the script file is passed as argv[2]. Otherwise, the name of the script file is passed as argv[1]. argv[0] is passed as specified in the exec*() call. All other arguments specified in the exec*() call are passed following the name of the script file (that is, beginning at argv[3] if there is an argument; otherwise, at argv[2]).

Some interpreters process the interpreter and the argument internally, and do not provide the interpreter and the argument to the users script.

If the initial line of the script file exceeds a system-defined maximum number of characters, exec*() fails. The minimum value for this limit is 32.

The set-user-ID and set-group-ID bits are honored for the script but not for the interpreter.

For an executable object file, the arguments are passed as argv[1], ..., argv[n]. argv[0] is passed as specified in the exec*() call, unless either argv or argv[0] is null as specified, in which case a pointer to a null string is passed as argv[0].

Security Restrictions

Some or all of the actions associated with this system call may require privileges. See privileges(5) for more information.

RETURN VALUE

If exec*() returns to the calling program, an error has occurred; the return value is -1 and errno is set to indicate the error.

ERRORS

If exec*() fails and returns to the calling program, errno is set to one of the following values:

E2BIG

The number of bytes in the new program's argument list plus environment is greater than the system-imposed limit. This limit is at least 5120 bytes on HP-UX systems.

EACCES

Read permission is denied for the executable file or interpreter, and the trace flag (see ttrace(2) request TT_PROC_SETTRC) of the process is set.

EACCES

Search permission is denied for a directory listed in the executable file's or the interpreter's path prefix.

EACCES

The executable file or the interpreter is not an ordinary file.

EACCES

The file described by path or file is not executable. A user with the DACREAD privilege cannot execute a file unless at least one access permission bit or entry in its access control list has an execute bit set.

EFAULT

path, argv, or envp point to an illegal address. The reliable detection of this error is implementation dependent.

EINTR

A signal was caught during the exec*() system call.

EINVAL

The executable file is incompatible with the architecture on which the exec*() has been performed, and is presumed to be for a different architecture. It is not guaranteed that every architecture's executable files will be recognized.

EINVAL

argv points to NULL and null pointer dereferencing is allowed. In this case, NULL is a valid address, but is considered an invalid argument.

ELOOP

Too many symbolic links were encountered in translating the path name.

ENAMETOOLONG

The executable file's path name or the interpreter's path name exceeds PATH_MAX bytes, or the length of a component of the path name exceeds NAME_MAX bytes while _POSIX_NO_TRUNC is in effect.

ENOENT

path points to an empty string.

ENOENT

One or more components of the executable file's path name or the interpreter's path name do not exist.

ENOEXEC

The executable file is shorter than indicated by the size values in its header, or is otherwise inconsistent. The reliable detection of this error is implementation dependent.

ENOEXEC

The function call is not execlp() or execvp(), and the executable file has the appropriate access permission, but there is neither a valid magic number nor the characters #! as the first two bytes of the file's initial line.

ENOEXEC

The number of bytes in the initial line of a script file exceeds the system's maximum.

ENOMEM

The new process requires more memory than is available or allowed by the system-imposed maximum.

ENOTDIR

A component of the executable file's path prefix or the interpreter's path prefix is not a directory.

EPERM

The process does not have sufficient privileges.

ETXTBSY

The executable file is currently open for writing.

WARNINGS

Unsharable executable files are not supported. These are files whose EXEC_MAGIC magic number was produced with the -N option of ld (see ld(1)).

It is recommended to use the execve() call in multi-threaded applications to avoid possible deadlocks.

DEPENDENCIES

HP Process Resource Manager

If the optional HP Process Resource Manager (PRM) software is installed and configured, the process's process resource group ID is not changed by exec*(). See prmconfig(1) for a description of how to configure HP PRM, and prmconf(4) for the definition of process resource group.

SEE ALSO

sh(1), sh-posix(1), kctune(1M), alarm(2), exit(2), fork(2), nice(2), ttrace(2), semop(2), signal(2), times(2), ulimit(2), umask(2), a.out(4), acl(5), environ(5), privileges(5), signal(5), thread_safety(5).

HP Process Resource Manager: prmconfig(1), prmconf(4) in HP Process Resource Manager User's Guide.

STANDARDS CONFORMANCE

environ: AES, SVID2, SVID3, XPG2, XPG3, XPG4, FIPS 151-2, POSIX.1

execl(): AES, SVID2, SVID3, XPG2, XPG3, XPG4, FIPS 151-2, POSIX.1

execle(): AES, SVID2, SVID3, XPG2, XPG3, XPG4, FIPS 151-2, POSIX.1

execlp(): AES, SVID2, SVID3, XPG2, XPG3, XPG4, FIPS 151-2, POSIX.1

execv(): AES, SVID2, SVID3, XPG2, XPG3, XPG4, FIPS 151-2, POSIX.1

execve(): AES, SVID2, SVID3, XPG2, XPG3, XPG4, FIPS 151-2, POSIX.1

execvp(): AES, SVID2, SVID3, XPG2, XPG3, XPG4, FIPS 151-2, POSIX.1

Printable version
Privacy statement Using this site means you accept its terms Feedback to webmaster
© 1983-2007 Hewlett-Packard Development Company, L.P.