]>
Commit | Line | Data |
---|---|---|
28f540f4 RM |
1 | @node Process Startup |
2 | @chapter Process Startup and Termination | |
3 | ||
4 | @cindex process | |
5 | @dfn{Processes} are the primitive units for allocation of system | |
6 | resources. Each process has its own address space and (usually) one | |
7 | thread of control. A process executes a program; you can have multiple | |
8 | processes executing the same program, but each process has its own copy | |
9 | of the program within its own address space and executes it | |
10 | independently of the other copies. | |
11 | ||
12 | This chapter explains what your program should do to handle the startup | |
13 | of a process, to terminate its process, and to receive information | |
14 | (arguments and the environment) from the parent process. | |
15 | ||
16 | @menu | |
17 | * Program Arguments:: Parsing your program's command-line arguments. | |
18 | * Environment Variables:: How to access parameters inherited from | |
19 | a parent process. | |
20 | * Program Termination:: How to cause a process to terminate and | |
21 | return status information to its parent. | |
22 | @end menu | |
23 | ||
24 | @node Program Arguments | |
25 | @section Program Arguments | |
26 | @cindex program arguments | |
27 | @cindex command line arguments | |
28 | @cindex arguments, to program | |
29 | ||
30 | @cindex program startup | |
31 | @cindex startup of program | |
32 | @cindex invocation of program | |
33 | @cindex @code{main} function | |
34 | @findex main | |
35 | The system starts a C program by calling the function @code{main}. It | |
36 | is up to you to write a function named @code{main}---otherwise, you | |
37 | won't even be able to link your program without errors. | |
38 | ||
f65fd747 | 39 | In @w{ISO C} you can define @code{main} either to take no arguments, or to |
28f540f4 RM |
40 | take two arguments that represent the command line arguments to the |
41 | program, like this: | |
42 | ||
43 | @smallexample | |
44 | int main (int @var{argc}, char *@var{argv}[]) | |
45 | @end smallexample | |
46 | ||
47 | @cindex argc (program argument count) | |
48 | @cindex argv (program argument vector) | |
49 | The command line arguments are the whitespace-separated tokens given in | |
50 | the shell command used to invoke the program; thus, in @samp{cat foo | |
51 | bar}, the arguments are @samp{foo} and @samp{bar}. The only way a | |
52 | program can look at its command line arguments is via the arguments of | |
53 | @code{main}. If @code{main} doesn't take arguments, then you cannot get | |
54 | at the command line. | |
55 | ||
56 | The value of the @var{argc} argument is the number of command line | |
57 | arguments. The @var{argv} argument is a vector of C strings; its | |
58 | elements are the individual command line argument strings. The file | |
59 | name of the program being run is also included in the vector as the | |
60 | first element; the value of @var{argc} counts this element. A null | |
61 | pointer always follows the last element: @code{@var{argv}[@var{argc}]} | |
62 | is this null pointer. | |
63 | ||
64 | For the command @samp{cat foo bar}, @var{argc} is 3 and @var{argv} has | |
65 | three elements, @code{"cat"}, @code{"foo"} and @code{"bar"}. | |
66 | ||
28f540f4 RM |
67 | In Unix systems you can define @code{main} a third way, using three arguments: |
68 | ||
69 | @smallexample | |
70 | int main (int @var{argc}, char *@var{argv}[], char *@var{envp}) | |
71 | @end smallexample | |
72 | ||
73 | The first two arguments are just the same. The third argument | |
74 | @var{envp} gives the process's environment; it is the same as the value | |
75 | of @code{environ}. @xref{Environment Variables}. POSIX.1 does not | |
76 | allow this three-argument form, so to be portable it is best to write | |
77 | @code{main} to take two arguments, and use the value of @code{environ}. | |
78 | ||
79 | @menu | |
2064087b | 80 | * Argument Syntax:: By convention, options start with a hyphen. |
b0de3e9e | 81 | * Parsing Program Arguments:: Ways to parse program options and arguments. |
28f540f4 RM |
82 | @end menu |
83 | ||
84 | @node Argument Syntax | |
85 | @subsection Program Argument Syntax Conventions | |
86 | @cindex program argument syntax | |
87 | @cindex syntax, for program arguments | |
88 | @cindex command argument syntax | |
89 | ||
90 | POSIX recommends these conventions for command line arguments. | |
b0de3e9e UD |
91 | @code{getopt} (@pxref{Getopt}) and @code{argp_parse} (@pxref{Argp}) make |
92 | it easy to implement them. | |
28f540f4 RM |
93 | |
94 | @itemize @bullet | |
95 | @item | |
96 | Arguments are options if they begin with a hyphen delimiter (@samp{-}). | |
97 | ||
98 | @item | |
99 | Multiple options may follow a hyphen delimiter in a single token if | |
100 | the options do not take arguments. Thus, @samp{-abc} is equivalent to | |
101 | @samp{-a -b -c}. | |
102 | ||
103 | @item | |
104 | Option names are single alphanumeric characters (as for @code{isalnum}; | |
105 | see @ref{Classification of Characters}). | |
106 | ||
107 | @item | |
108 | Certain options require an argument. For example, the @samp{-o} command | |
109 | of the @code{ld} command requires an argument---an output file name. | |
110 | ||
111 | @item | |
112 | An option and its argument may or may not appear as separate tokens. (In | |
113 | other words, the whitespace separating them is optional.) Thus, | |
114 | @w{@samp{-o foo}} and @samp{-ofoo} are equivalent. | |
115 | ||
116 | @item | |
117 | Options typically precede other non-option arguments. | |
118 | ||
b0de3e9e UD |
119 | The implementations of @code{getopt} and @code{argp_parse} in the GNU C |
120 | library normally make it appear as if all the option arguments were | |
121 | specified before all the non-option arguments for the purposes of | |
122 | parsing, even if the user of your program intermixed option and | |
123 | non-option arguments. They do this by reordering the elements of the | |
124 | @var{argv} array. This behavior is nonstandard; if you want to suppress | |
125 | it, define the @code{_POSIX_OPTION_ORDER} environment variable. | |
126 | @xref{Standard Environment}. | |
28f540f4 RM |
127 | |
128 | @item | |
129 | The argument @samp{--} terminates all options; any following arguments | |
130 | are treated as non-option arguments, even if they begin with a hyphen. | |
131 | ||
132 | @item | |
133 | A token consisting of a single hyphen character is interpreted as an | |
134 | ordinary non-option argument. By convention, it is used to specify | |
135 | input from or output to the standard input and output streams. | |
136 | ||
137 | @item | |
138 | Options may be supplied in any order, or appear multiple times. The | |
139 | interpretation is left up to the particular application program. | |
140 | @end itemize | |
141 | ||
142 | @cindex long-named options | |
143 | GNU adds @dfn{long options} to these conventions. Long options consist | |
144 | of @samp{--} followed by a name made of alphanumeric characters and | |
145 | dashes. Option names are typically one to three words long, with | |
146 | hyphens to separate words. Users can abbreviate the option names as | |
147 | long as the abbreviations are unique. | |
148 | ||
149 | To specify an argument for a long option, write | |
150 | @samp{--@var{name}=@var{value}}. This syntax enables a long option to | |
151 | accept an argument that is itself optional. | |
152 | ||
153 | Eventually, the GNU system will provide completion for long option names | |
154 | in the shell. | |
155 | ||
b0de3e9e UD |
156 | @node Parsing Program Arguments |
157 | @subsection Parsing Program Arguments | |
158 | ||
28f540f4 RM |
159 | @cindex program arguments, parsing |
160 | @cindex command arguments, parsing | |
161 | @cindex parsing program arguments | |
b0de3e9e UD |
162 | If the syntax for the command line arguments to your program is simple |
163 | enough, you can simply pick the arguments off from @var{argv} by hand. | |
164 | But unless your program takes a fixed number of arguments, or all of the | |
165 | arguments are interpreted in the same way (as file names, for example), | |
166 | you are usually better off using @code{getopt} (@pxref{Getopt}) or | |
167 | @code{argp_parse} (@pxref{Argp}) to do the parsing. | |
28f540f4 | 168 | |
b0de3e9e UD |
169 | @code{getopt} is more standard (the short-option only version of it is a |
170 | part of the POSIX standard), but using @code{argp_parse} is often | |
171 | easier, both for very simple and very complex option structures, because | |
172 | it does more of the dirty work for you. | |
28f540f4 | 173 | |
b0de3e9e UD |
174 | @menu |
175 | * Getopt:: Parsing program options using @code{getopt}. | |
176 | * Argp:: Parsing program options using @code{argp_parse}. | |
177 | * Suboptions:: Some programs need more detailed options. | |
178 | * Suboptions Example:: This shows how it could be done for @code{mount}. | |
179 | @end menu | |
28f540f4 | 180 | |
b0de3e9e UD |
181 | @c Getopt and argp start at the @section level so that there's |
182 | @c enough room for their internal hierarchy (mostly a problem with | |
183 | @c argp). -Miles | |
28f540f4 | 184 | |
b0de3e9e UD |
185 | @include getopt.texi |
186 | @include argp.texi | |
28f540f4 | 187 | |
b0de3e9e UD |
188 | @node Suboptions, Suboptions Example, Argp, Parsing Program Arguments |
189 | @c This is a @section so that it's at the same level as getopt and argp | |
190 | @section Parsing of Suboptions | |
2064087b RM |
191 | |
192 | Having a single level of options is sometimes not enough. There might | |
193 | be too many options which have to be available or a set of options is | |
194 | closely related. | |
195 | ||
196 | For this case some programs use suboptions. One of the most prominent | |
197 | programs is certainly @code{mount}(8). The @code{-o} option take one | |
198 | argument which itself is a comma separated list of options. To ease the | |
199 | programming of code like this the function @code{getsubopt} is | |
200 | available. | |
201 | ||
202 | @comment stdlib.h | |
203 | @deftypefun int getsubopt (char **@var{optionp}, const char* const *@var{tokens}, char **@var{valuep}) | |
204 | ||
205 | The @var{optionp} parameter must be a pointer to a variable containing | |
206 | the address of the string to process. When the function returns the | |
207 | reference is updated to point to the next suboption or to the | |
208 | terminating @samp{\0} character if there is no more suboption available. | |
209 | ||
210 | The @var{tokens} parameter references an array of strings containing the | |
211 | known suboptions. All strings must be @samp{\0} terminated and to mark | |
212 | the end a null pointer must be stored. When @code{getsubopt} finds a | |
213 | possible legal suboption it compares it with all strings available in | |
214 | the @var{tokens} array and returns the index in the string as the | |
215 | indicator. | |
216 | ||
217 | In case the suboption has an associated value introduced by a @samp{=} | |
218 | character, a pointer to the value is returned in @var{valuep}. The | |
219 | string is @samp{\0} terminated. If no argument is available | |
220 | @var{valuep} is set to the null pointer. By doing this the caller can | |
221 | check whether a necessary value is given or whether no unexpected value | |
222 | is present. | |
223 | ||
224 | In case the next suboption in the string is not mentioned in the | |
225 | @var{tokens} array the starting address of the suboption including a | |
226 | possible value is returned in @var{valuep} and the return value of the | |
227 | function is @samp{-1}. | |
228 | @end deftypefun | |
229 | ||
b0de3e9e | 230 | @node Suboptions Example, , Suboptions, Parsing Program Arguments |
2064087b RM |
231 | @subsection Parsing of Suboptions Example |
232 | ||
233 | The code which might appear in the @code{mount}(8) program is a perfect | |
234 | example of the use of @code{getsubopt}: | |
235 | ||
236 | @smallexample | |
237 | @include subopt.c.texi | |
238 | @end smallexample | |
239 | ||
240 | ||
28f540f4 RM |
241 | @node Environment Variables |
242 | @section Environment Variables | |
243 | ||
244 | @cindex environment variable | |
245 | When a program is executed, it receives information about the context in | |
246 | which it was invoked in two ways. The first mechanism uses the | |
247 | @var{argv} and @var{argc} arguments to its @code{main} function, and is | |
248 | discussed in @ref{Program Arguments}. The second mechanism uses | |
249 | @dfn{environment variables} and is discussed in this section. | |
250 | ||
251 | The @var{argv} mechanism is typically used to pass command-line | |
252 | arguments specific to the particular program being invoked. The | |
253 | environment, on the other hand, keeps track of information that is | |
254 | shared by many programs, changes infrequently, and that is less | |
255 | frequently used. | |
256 | ||
257 | The environment variables discussed in this section are the same | |
258 | environment variables that you set using assignments and the | |
259 | @code{export} command in the shell. Programs executed from the shell | |
260 | inherit all of the environment variables from the shell. | |
261 | @c !!! xref to right part of bash manual when it exists | |
262 | ||
263 | @cindex environment | |
264 | Standard environment variables are used for information about the user's | |
265 | home directory, terminal type, current locale, and so on; you can define | |
266 | additional variables for other purposes. The set of all environment | |
267 | variables that have values is collectively known as the | |
268 | @dfn{environment}. | |
269 | ||
270 | Names of environment variables are case-sensitive and must not contain | |
271 | the character @samp{=}. System-defined environment variables are | |
272 | invariably uppercase. | |
273 | ||
274 | The values of environment variables can be anything that can be | |
275 | represented as a string. A value must not contain an embedded null | |
276 | character, since this is assumed to terminate the string. | |
277 | ||
278 | ||
279 | @menu | |
2064087b | 280 | * Environment Access:: How to get and set the values of |
40a55d20 | 281 | environment variables. |
2064087b | 282 | * Standard Environment:: These environment variables have |
40a55d20 | 283 | standard interpretations. |
28f540f4 RM |
284 | @end menu |
285 | ||
286 | @node Environment Access | |
287 | @subsection Environment Access | |
288 | @cindex environment access | |
289 | @cindex environment representation | |
290 | ||
291 | The value of an environment variable can be accessed with the | |
292 | @code{getenv} function. This is declared in the header file | |
40a55d20 UD |
293 | @file{stdlib.h}. All of the following functions can be safely used in |
294 | multi-threaded programs. It is made sure that concurrent modifications | |
295 | to the environment do not lead to errors. | |
28f540f4 RM |
296 | @pindex stdlib.h |
297 | ||
298 | @comment stdlib.h | |
f65fd747 | 299 | @comment ISO |
28f540f4 RM |
300 | @deftypefun {char *} getenv (const char *@var{name}) |
301 | This function returns a string that is the value of the environment | |
302 | variable @var{name}. You must not modify this string. In some non-Unix | |
303 | systems not using the GNU library, it might be overwritten by subsequent | |
304 | calls to @code{getenv} (but not by any other library function). If the | |
305 | environment variable @var{name} is not defined, the value is a null | |
306 | pointer. | |
307 | @end deftypefun | |
308 | ||
309 | ||
310 | @comment stdlib.h | |
311 | @comment SVID | |
312 | @deftypefun int putenv (const char *@var{string}) | |
313 | The @code{putenv} function adds or removes definitions from the environment. | |
314 | If the @var{string} is of the form @samp{@var{name}=@var{value}}, the | |
315 | definition is added to the environment. Otherwise, the @var{string} is | |
316 | interpreted as the name of an environment variable, and any definition | |
317 | for this variable in the environment is removed. | |
318 | ||
40a55d20 UD |
319 | This function is part of the extended Unix interface. Since it was also |
320 | available in old SVID libraries you should define either | |
321 | @var{_XOPEN_SOURCE} or @var{_SVID_SOURCE} before including any header. | |
322 | @end deftypefun | |
323 | ||
324 | ||
325 | @comment stdlib.h | |
326 | @comment BSD | |
327 | @deftypefun int setenv (const char *@var{name}, const char *@var{value}, int @var{replace}) | |
328 | The @code{setenv} function can be used to add a new definition to the | |
329 | environment. The entry with the name @var{name} is replaced by the | |
330 | value @samp{@var{name}=@var{value}}. Please note that this is also true | |
331 | if @var{value} is the empty string. A null pointer for the @var{value} | |
332 | parameter is illegal. If the environment already contains an entry with | |
333 | key @var{name} the @var{replace} parameter controls the action. If | |
334 | replace is zero, nothing happens. otherwise the old entry is replaced | |
335 | by the new one. | |
336 | ||
337 | Please note that you cannot remove an entry completely using this function. | |
338 | ||
339 | This function is part of the BSD library. The GNU C Library provides | |
340 | this function for compatibility but it may not be available on other | |
341 | systems. | |
342 | @end deftypefun | |
343 | ||
344 | @comment stdlib.h | |
345 | @comment BSD | |
346 | @deftypefun void unsetenv (const char *@var{name}) | |
347 | Using this function one can remove an entry completely from the | |
348 | environment. If the environment contains an entry with the key | |
349 | @var{name} this whole entry is removed. A call to this function is | |
350 | equivalent to a call to @code{putenv} when the @var{value} part of the | |
351 | string is empty. | |
352 | ||
353 | This function is part of the BSD library. The GNU C Library provides | |
354 | this function for compatibility but it may not be available on other | |
355 | systems. | |
356 | @end deftypefun | |
357 | ||
358 | There is one more function to modify the whole environment. This | |
359 | function is said to be used in the POSIX.9 (POSIX bindings for Fortran | |
360 | 77) and so one should expect it did made it into POSIX.1. But this | |
361 | never happened. But we still provide this function as a GNU extension | |
362 | to enable writing standard compliant Fortran environments. | |
363 | ||
364 | @comment stdlib.h | |
365 | @comment GNU | |
366 | @deftypefun int clearenv (void) | |
367 | The @code{clearenv} function removes all entries from the environment. | |
368 | Using @code{putenv} and @code{setenv} new entries can be added again | |
369 | later. | |
370 | ||
371 | If the function is successful it returns @code{0}. Otherwise the return | |
372 | value is nonzero. | |
28f540f4 RM |
373 | @end deftypefun |
374 | ||
28f540f4 RM |
375 | |
376 | You can deal directly with the underlying representation of environment | |
377 | objects to add more variables to the environment (for example, to | |
378 | communicate with another program you are about to execute; see | |
1b82a4a8 | 379 | @ref{Executing a File}). |
28f540f4 RM |
380 | |
381 | @comment unistd.h | |
382 | @comment POSIX.1 | |
383 | @deftypevar {char **} environ | |
384 | The environment is represented as an array of strings. Each string is | |
385 | of the format @samp{@var{name}=@var{value}}. The order in which | |
386 | strings appear in the environment is not significant, but the same | |
387 | @var{name} must not appear more than once. The last element of the | |
388 | array is a null pointer. | |
389 | ||
390 | This variable is declared in the header file @file{unistd.h}. | |
391 | ||
392 | If you just want to get the value of an environment variable, use | |
393 | @code{getenv}. | |
394 | @end deftypevar | |
395 | ||
396 | Unix systems, and the GNU system, pass the initial value of | |
397 | @code{environ} as the third argument to @code{main}. | |
398 | @xref{Program Arguments}. | |
399 | ||
400 | @node Standard Environment | |
401 | @subsection Standard Environment Variables | |
402 | @cindex standard environment variables | |
403 | ||
404 | These environment variables have standard meanings. This doesn't mean | |
405 | that they are always present in the environment; but if these variables | |
406 | @emph{are} present, they have these meanings. You shouldn't try to use | |
407 | these environment variable names for some other purpose. | |
408 | ||
409 | @comment Extra blank lines make it look better. | |
410 | @table @code | |
411 | @item HOME | |
412 | @cindex HOME environment variable | |
413 | @cindex home directory | |
414 | ||
415 | This is a string representing the user's @dfn{home directory}, or | |
416 | initial default working directory. | |
417 | ||
418 | The user can set @code{HOME} to any value. | |
419 | If you need to make sure to obtain the proper home directory | |
420 | for a particular user, you should not use @code{HOME}; instead, | |
421 | look up the user's name in the user database (@pxref{User Database}). | |
422 | ||
423 | For most purposes, it is better to use @code{HOME}, precisely because | |
424 | this lets the user specify the value. | |
425 | ||
426 | @c !!! also USER | |
427 | @item LOGNAME | |
428 | @cindex LOGNAME environment variable | |
429 | ||
430 | This is the name that the user used to log in. Since the value in the | |
431 | environment can be tweaked arbitrarily, this is not a reliable way to | |
432 | identify the user who is running a process; a function like | |
433 | @code{getlogin} (@pxref{Who Logged In}) is better for that purpose. | |
434 | ||
435 | For most purposes, it is better to use @code{LOGNAME}, precisely because | |
436 | this lets the user specify the value. | |
437 | ||
438 | @item PATH | |
439 | @cindex PATH environment variable | |
440 | ||
441 | A @dfn{path} is a sequence of directory names which is used for | |
442 | searching for a file. The variable @code{PATH} holds a path used | |
443 | for searching for programs to be run. | |
444 | ||
445 | The @code{execlp} and @code{execvp} functions (@pxref{Executing a File}) | |
446 | use this environment variable, as do many shells and other utilities | |
447 | which are implemented in terms of those functions. | |
448 | ||
449 | The syntax of a path is a sequence of directory names separated by | |
1b82a4a8 | 450 | colons. An empty string instead of a directory name stands for the |
28f540f4 RM |
451 | current directory (@pxref{Working Directory}). |
452 | ||
453 | A typical value for this environment variable might be a string like: | |
454 | ||
455 | @smallexample | |
456 | :/bin:/etc:/usr/bin:/usr/new/X11:/usr/new:/usr/local/bin | |
457 | @end smallexample | |
458 | ||
459 | This means that if the user tries to execute a program named @code{foo}, | |
460 | the system will look for files named @file{foo}, @file{/bin/foo}, | |
461 | @file{/etc/foo}, and so on. The first of these files that exists is | |
462 | the one that is executed. | |
463 | ||
464 | @c !!! also TERMCAP | |
465 | @item TERM | |
466 | @cindex TERM environment variable | |
467 | ||
468 | This specifies the kind of terminal that is receiving program output. | |
469 | Some programs can make use of this information to take advantage of | |
470 | special escape sequences or terminal modes supported by particular kinds | |
471 | of terminals. Many programs which use the termcap library | |
472 | (@pxref{Finding a Terminal Description,Find,,termcap,The Termcap Library | |
473 | Manual}) use the @code{TERM} environment variable, for example. | |
474 | ||
475 | @item TZ | |
476 | @cindex TZ environment variable | |
477 | ||
478 | This specifies the time zone. @xref{TZ Variable}, for information about | |
479 | the format of this string and how it is used. | |
480 | ||
481 | @item LANG | |
482 | @cindex LANG environment variable | |
483 | ||
484 | This specifies the default locale to use for attribute categories where | |
485 | neither @code{LC_ALL} nor the specific environment variable for that | |
486 | category is set. @xref{Locales}, for more information about | |
487 | locales. | |
488 | ||
489 | @ignore | |
490 | @c I doubt this really exists | |
491 | @item LC_ALL | |
492 | @cindex LC_ALL environment variable | |
493 | ||
494 | This is similar to the @code{LANG} environment variable. However, its | |
495 | value takes precedence over any values provided for the individual | |
496 | attribute category environment variables, or for the @code{LANG} | |
497 | environment variable. | |
498 | @end ignore | |
499 | ||
40a55d20 UD |
500 | @item LC_ALL |
501 | @cindex LC_ALL environment variable | |
502 | ||
503 | If this environment variable is set it overrides the selection for all | |
504 | the locales done using the other @code{LC_*} environment variables. The | |
505 | value of the other @code{LC_*} environment variables is simply ignored | |
506 | in this case. | |
507 | ||
28f540f4 RM |
508 | @item LC_COLLATE |
509 | @cindex LC_COLLATE environment variable | |
510 | ||
511 | This specifies what locale to use for string sorting. | |
512 | ||
513 | @item LC_CTYPE | |
514 | @cindex LC_CTYPE environment variable | |
515 | ||
516 | This specifies what locale to use for character sets and character | |
517 | classification. | |
518 | ||
40a55d20 UD |
519 | @item LC_MESSAGES |
520 | @cindex LC_MESSAGES environment variable | |
521 | ||
522 | This specifies what locale to use for printing messages and to parse | |
f2ea0f5b | 523 | responses. |
40a55d20 | 524 | |
28f540f4 RM |
525 | @item LC_MONETARY |
526 | @cindex LC_MONETARY environment variable | |
527 | ||
528 | This specifies what locale to use for formatting monetary values. | |
529 | ||
530 | @item LC_NUMERIC | |
531 | @cindex LC_NUMERIC environment variable | |
532 | ||
533 | This specifies what locale to use for formatting numbers. | |
534 | ||
535 | @item LC_TIME | |
536 | @cindex LC_TIME environment variable | |
537 | ||
538 | This specifies what locale to use for formatting date/time values. | |
539 | ||
40a55d20 UD |
540 | @item NLSPATH |
541 | @cindex NLSPATH environment variable | |
542 | ||
543 | This specifies the directories in which the @code{catopen} function | |
544 | looks for message translation catalogs. | |
545 | ||
28f540f4 RM |
546 | @item _POSIX_OPTION_ORDER |
547 | @cindex _POSIX_OPTION_ORDER environment variable. | |
548 | ||
549 | If this environment variable is defined, it suppresses the usual | |
b0de3e9e UD |
550 | reordering of command line arguments by @code{getopt} and |
551 | @code{argp_parse}. @xref{Argument Syntax}. | |
28f540f4 RM |
552 | |
553 | @c !!! GNU also has COREFILE, CORESERVER, EXECSERVERS | |
554 | @end table | |
555 | ||
556 | @node Program Termination | |
557 | @section Program Termination | |
558 | @cindex program termination | |
559 | @cindex process termination | |
560 | ||
561 | @cindex exit status value | |
562 | The usual way for a program to terminate is simply for its @code{main} | |
563 | function to return. The @dfn{exit status value} returned from the | |
564 | @code{main} function is used to report information back to the process's | |
565 | parent process or shell. | |
566 | ||
567 | A program can also terminate normally by calling the @code{exit} | |
568 | function. | |
569 | ||
570 | In addition, programs can be terminated by signals; this is discussed in | |
571 | more detail in @ref{Signal Handling}. The @code{abort} function causes | |
572 | a signal that kills the program. | |
573 | ||
574 | @menu | |
575 | * Normal Termination:: If a program calls @code{exit}, a | |
576 | process terminates normally. | |
1b82a4a8 RM |
577 | * Exit Status:: The @code{exit status} provides information |
578 | about why the process terminated. | |
28f540f4 | 579 | * Cleanups on Exit:: A process can run its own cleanup |
1b82a4a8 | 580 | functions upon normal termination. |
28f540f4 | 581 | * Aborting a Program:: The @code{abort} function causes |
1b82a4a8 | 582 | abnormal program termination. |
28f540f4 RM |
583 | * Termination Internals:: What happens when a process terminates. |
584 | @end menu | |
585 | ||
586 | @node Normal Termination | |
587 | @subsection Normal Termination | |
588 | ||
589 | A process terminates normally when the program calls @code{exit}. | |
590 | Returning from @code{main} is equivalent to calling @code{exit}, and | |
591 | the value that @code{main} returns is used as the argument to @code{exit}. | |
592 | ||
593 | @comment stdlib.h | |
f65fd747 | 594 | @comment ISO |
28f540f4 RM |
595 | @deftypefun void exit (int @var{status}) |
596 | The @code{exit} function terminates the process with status | |
597 | @var{status}. This function does not return. | |
598 | @end deftypefun | |
599 | ||
600 | Normal termination causes the following actions: | |
601 | ||
602 | @enumerate | |
1b82a4a8 | 603 | @item |
28f540f4 RM |
604 | Functions that were registered with the @code{atexit} or @code{on_exit} |
605 | functions are called in the reverse order of their registration. This | |
606 | mechanism allows your application to specify its own ``cleanup'' actions | |
607 | to be performed at program termination. Typically, this is used to do | |
608 | things like saving program state information in a file, or unlocking | |
609 | locks in shared data bases. | |
610 | ||
1b82a4a8 | 611 | @item |
28f540f4 RM |
612 | All open streams are closed, writing out any buffered output data. See |
613 | @ref{Closing Streams}. In addition, temporary files opened | |
614 | with the @code{tmpfile} function are removed; see @ref{Temporary Files}. | |
615 | ||
1b82a4a8 | 616 | @item |
28f540f4 RM |
617 | @code{_exit} is called, terminating the program. @xref{Termination Internals}. |
618 | @end enumerate | |
619 | ||
620 | @node Exit Status | |
621 | @subsection Exit Status | |
622 | @cindex exit status | |
623 | ||
624 | When a program exits, it can return to the parent process a small | |
625 | amount of information about the cause of termination, using the | |
626 | @dfn{exit status}. This is a value between 0 and 255 that the exiting | |
627 | process passes as an argument to @code{exit}. | |
628 | ||
629 | Normally you should use the exit status to report very broad information | |
630 | about success or failure. You can't provide a lot of detail about the | |
631 | reasons for the failure, and most parent processes would not want much | |
632 | detail anyway. | |
633 | ||
634 | There are conventions for what sorts of status values certain programs | |
635 | should return. The most common convention is simply 0 for success and 1 | |
636 | for failure. Programs that perform comparison use a different | |
637 | convention: they use status 1 to indicate a mismatch, and status 2 to | |
638 | indicate an inability to compare. Your program should follow an | |
639 | existing convention if an existing convention makes sense for it. | |
640 | ||
641 | A general convention reserves status values 128 and up for special | |
642 | purposes. In particular, the value 128 is used to indicate failure to | |
643 | execute another program in a subprocess. This convention is not | |
644 | universally obeyed, but it is a good idea to follow it in your programs. | |
645 | ||
646 | @strong{Warning:} Don't try to use the number of errors as the exit | |
647 | status. This is actually not very useful; a parent process would | |
648 | generally not care how many errors occurred. Worse than that, it does | |
649 | not work, because the status value is truncated to eight bits. | |
650 | Thus, if the program tried to report 256 errors, the parent would | |
651 | receive a report of 0 errors---that is, success. | |
652 | ||
653 | For the same reason, it does not work to use the value of @code{errno} | |
654 | as the exit status---these can exceed 255. | |
655 | ||
656 | @strong{Portability note:} Some non-POSIX systems use different | |
657 | conventions for exit status values. For greater portability, you can | |
658 | use the macros @code{EXIT_SUCCESS} and @code{EXIT_FAILURE} for the | |
659 | conventional status value for success and failure, respectively. They | |
660 | are declared in the file @file{stdlib.h}. | |
661 | @pindex stdlib.h | |
662 | ||
663 | @comment stdlib.h | |
f65fd747 | 664 | @comment ISO |
28f540f4 RM |
665 | @deftypevr Macro int EXIT_SUCCESS |
666 | This macro can be used with the @code{exit} function to indicate | |
667 | successful program completion. | |
668 | ||
669 | On POSIX systems, the value of this macro is @code{0}. On other | |
670 | systems, the value might be some other (possibly non-constant) integer | |
671 | expression. | |
672 | @end deftypevr | |
673 | ||
674 | @comment stdlib.h | |
f65fd747 | 675 | @comment ISO |
28f540f4 RM |
676 | @deftypevr Macro int EXIT_FAILURE |
677 | This macro can be used with the @code{exit} function to indicate | |
678 | unsuccessful program completion in a general sense. | |
679 | ||
680 | On POSIX systems, the value of this macro is @code{1}. On other | |
681 | systems, the value might be some other (possibly non-constant) integer | |
f166d865 | 682 | expression. Other nonzero status values also indicate failures. Certain |
28f540f4 RM |
683 | programs use different nonzero status values to indicate particular |
684 | kinds of "non-success". For example, @code{diff} uses status value | |
685 | @code{1} to mean that the files are different, and @code{2} or more to | |
686 | mean that there was difficulty in opening the files. | |
687 | @end deftypevr | |
688 | ||
689 | @node Cleanups on Exit | |
690 | @subsection Cleanups on Exit | |
691 | ||
692 | Your program can arrange to run its own cleanup functions if normal | |
693 | termination happens. If you are writing a library for use in various | |
694 | application programs, then it is unreliable to insist that all | |
695 | applications call the library's cleanup functions explicitly before | |
696 | exiting. It is much more robust to make the cleanup invisible to the | |
697 | application, by setting up a cleanup function in the library itself | |
698 | using @code{atexit} or @code{on_exit}. | |
699 | ||
700 | @comment stdlib.h | |
f65fd747 | 701 | @comment ISO |
28f540f4 RM |
702 | @deftypefun int atexit (void (*@var{function}) (void)) |
703 | The @code{atexit} function registers the function @var{function} to be | |
704 | called at normal program termination. The @var{function} is called with | |
705 | no arguments. | |
706 | ||
707 | The return value from @code{atexit} is zero on success and nonzero if | |
1b82a4a8 | 708 | the function cannot be registered. |
28f540f4 RM |
709 | @end deftypefun |
710 | ||
711 | @comment stdlib.h | |
712 | @comment SunOS | |
713 | @deftypefun int on_exit (void (*@var{function})(int @var{status}, void *@var{arg}), void *@var{arg}) | |
714 | This function is a somewhat more powerful variant of @code{atexit}. It | |
715 | accepts two arguments, a function @var{function} and an arbitrary | |
716 | pointer @var{arg}. At normal program termination, the @var{function} is | |
717 | called with two arguments: the @var{status} value passed to @code{exit}, | |
718 | and the @var{arg}. | |
719 | ||
720 | This function is included in the GNU C library only for compatibility | |
721 | for SunOS, and may not be supported by other implementations. | |
722 | @end deftypefun | |
723 | ||
724 | Here's a trivial program that illustrates the use of @code{exit} and | |
725 | @code{atexit}: | |
726 | ||
727 | @smallexample | |
728 | @include atexit.c.texi | |
729 | @end smallexample | |
730 | ||
731 | @noindent | |
732 | When this program is executed, it just prints the message and exits. | |
733 | ||
734 | @node Aborting a Program | |
735 | @subsection Aborting a Program | |
736 | @cindex aborting a program | |
737 | ||
738 | You can abort your program using the @code{abort} function. The prototype | |
739 | for this function is in @file{stdlib.h}. | |
740 | @pindex stdlib.h | |
741 | ||
742 | @comment stdlib.h | |
f65fd747 | 743 | @comment ISO |
28f540f4 RM |
744 | @deftypefun void abort (void) |
745 | The @code{abort} function causes abnormal program termination. This | |
746 | does not execute cleanup functions registered with @code{atexit} or | |
747 | @code{on_exit}. | |
748 | ||
749 | This function actually terminates the process by raising a | |
750 | @code{SIGABRT} signal, and your program can include a handler to | |
751 | intercept this signal; see @ref{Signal Handling}. | |
752 | @end deftypefun | |
753 | ||
754 | @c Put in by rms. Don't remove. | |
755 | @cartouche | |
756 | @strong{Future Change Warning:} Proposed Federal censorship regulations | |
757 | may prohibit us from giving you information about the possibility of | |
758 | calling this function. We would be required to say that this is not an | |
759 | acceptable way of terminating a program. | |
760 | @end cartouche | |
761 | ||
762 | @node Termination Internals | |
763 | @subsection Termination Internals | |
764 | ||
765 | The @code{_exit} function is the primitive used for process termination | |
766 | by @code{exit}. It is declared in the header file @file{unistd.h}. | |
767 | @pindex unistd.h | |
768 | ||
769 | @comment unistd.h | |
770 | @comment POSIX.1 | |
771 | @deftypefun void _exit (int @var{status}) | |
772 | The @code{_exit} function is the primitive for causing a process to | |
773 | terminate with status @var{status}. Calling this function does not | |
774 | execute cleanup functions registered with @code{atexit} or | |
775 | @code{on_exit}. | |
776 | @end deftypefun | |
777 | ||
778 | When a process terminates for any reason---either by an explicit | |
779 | termination call, or termination as a result of a signal---the | |
780 | following things happen: | |
781 | ||
782 | @itemize @bullet | |
783 | @item | |
784 | All open file descriptors in the process are closed. @xref{Low-Level I/O}. | |
785 | Note that streams are not flushed automatically when the process | |
786 | terminates; @xref{I/O on Streams}. | |
787 | ||
788 | @item | |
789 | The low-order 8 bits of the return status code are saved to be reported | |
790 | back to the parent process via @code{wait} or @code{waitpid}; see | |
791 | @ref{Process Completion}. | |
792 | ||
793 | @item | |
794 | Any child processes of the process being terminated are assigned a new | |
795 | parent process. (On most systems, including GNU, this is the @code{init} | |
796 | process, with process ID 1.) | |
797 | ||
798 | @item | |
799 | A @code{SIGCHLD} signal is sent to the parent process. | |
800 | ||
801 | @item | |
802 | If the process is a session leader that has a controlling terminal, then | |
803 | a @code{SIGHUP} signal is sent to each process in the foreground job, | |
804 | and the controlling terminal is disassociated from that session. | |
805 | @xref{Job Control}. | |
806 | ||
807 | @item | |
808 | If termination of a process causes a process group to become orphaned, | |
809 | and any member of that process group is stopped, then a @code{SIGHUP} | |
810 | signal and a @code{SIGCONT} signal are sent to each process in the | |
811 | group. @xref{Job Control}. | |
812 | @end itemize |