source: branches/3.2/mindi-busybox/docs/ctty.htm @ 3186

Last change on this file since 3186 was 2725, checked in by Bruno Cornec, 9 years ago
  • Update mindi-busybox to 1.18.3 to avoid problems with the tar command which is now failing on recent versions with busybox 1.7.3
File size: 17.1 KB
Line 
1<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 3.2 Final//EN">
2<html><head>
3 <!-- saved from http://www.win.tue.nl/~aeb/linux/lk/lk-10.html -->
4 <meta name="GENERATOR" content="SGML-Tools 1.0.9"><title>The Linux kernel: Processes</title>
5</head>
6<body>
7<hr>
8<h2><a name="s10">10. Processes</a></h2>
9
10<p>Before looking at the Linux implementation, first a general Unix
11description of threads, processes, process groups and sessions.
12</p><p>A session contains a number of process groups, and a process group
13contains a number of processes, and a process contains a number
14of threads.
15</p><p>A session can have a controlling tty.
16At most one process group in a session can be a foreground process group.
17An interrupt character typed on a tty ("Teletype", i.e., terminal)
18causes a signal to be sent to all members of the foreground process group
19in the session (if any) that has that tty as controlling tty.
20</p><p>All these objects have numbers, and we have thread IDs, process IDs,
21process group IDs and session IDs.
22</p><p>
23</p><h2><a name="ss10.1">10.1 Processes</a>
24</h2>
25
26<p>
27</p><h3>Creation</h3>
28
29<p>A new process is traditionally started using the <code>fork()</code>
30system call:
31</p><blockquote>
32<pre>pid_t p;
33
34p = fork();
35if (p == (pid_t) -1)
36        /* ERROR */
37else if (p == 0)
38        /* CHILD */
39else
40        /* PARENT */
41</pre>
42</blockquote>
43<p>This creates a child as a duplicate of its parent.
44Parent and child are identical in almost all respects.
45In the code they are distinguished by the fact that the parent
46learns the process ID of its child, while <code>fork()</code>
47returns 0 in the child. (It can find the process ID of its
48parent using the <code>getppid()</code> system call.)
49</p><p>
50</p><h3>Termination</h3>
51
52<p>Normal termination is when the process does
53</p><blockquote>
54<pre>exit(n);
55</pre>
56</blockquote>
57
58or
59<blockquote>
60<pre>return n;
61</pre>
62</blockquote>
63
64from its <code>main()</code> procedure. It returns the single byte <code>n</code>
65to its parent.
66<p>Abnormal termination is usually caused by a signal.
67</p><p>
68</p><h3>Collecting the exit code. Zombies</h3>
69
70<p>The parent does
71</p><blockquote>
72<pre>pid_t p;
73int status;
74
75p = wait(&amp;status);
76</pre>
77</blockquote>
78
79and collects two bytes:
80<p>
81<figure>
82<eps file="absent">
83<img src="ctty_files/exit_status.png">
84</eps>
85</figure></p><p>A process that has terminated but has not yet been waited for
86is a <i>zombie</i>. It need only store these two bytes:
87exit code and reason for termination.
88</p><p>On the other hand, if the parent dies first, <code>init</code> (process 1)
89inherits the child and becomes its parent.
90</p><p>
91</p><h3>Signals</h3>
92
93<p>
94</p><h3>Stopping</h3>
95
96<p>Some signals cause a process to stop:
97<code>SIGSTOP</code> (stop!),
98<code>SIGTSTP</code> (stop from tty: probably ^Z was typed),
99<code>SIGTTIN</code> (tty input asked by background process),
100<code>SIGTTOU</code> (tty output sent by background process, and this was
101disallowed by <code>stty tostop</code>).
102</p><p>Apart from ^Z there also is ^Y. The former stops the process
103when it is typed, the latter stops it when it is read.
104</p><p>Signals generated by typing the corresponding character on some tty
105are sent to all processes that are in the foreground process group
106of the session that has that tty as controlling tty. (Details below.)
107</p><p>If a process is being traced, every signal will stop it.
108</p><p>
109</p><h3>Continuing</h3>
110
111<p><code>SIGCONT</code>: continue a stopped process.
112</p><p>
113</p><h3>Terminating</h3>
114
115<p><code>SIGKILL</code> (die! now!),
116<code>SIGTERM</code> (please, go away),
117<code>SIGHUP</code> (modem hangup),
118<code>SIGINT</code> (^C),
119<code>SIGQUIT</code> (^\), etc.
120Many signals have as default action to kill the target.
121(Sometimes with an additional core dump, when such is
122allowed by rlimit.)
123The signals <code>SIGCHLD</code> and <code>SIGWINCH</code>
124are ignored by default.
125All except <code>SIGKILL</code> and <code>SIGSTOP</code> can be
126caught or ignored or blocked.
127For details, see <code>signal(7)</code>.
128</p><p>
129</p><h2><a name="ss10.2">10.2 Process groups</a>
130</h2>
131
132<p>Every process is member of a unique <i>process group</i>,
133identified by its <i>process group ID</i>.
134(When the process is created, it becomes a member of the process group
135of its parent.)
136By convention, the process group ID of a process group
137equals the process ID of the first member of the process group,
138called the <i>process group leader</i>.
139A process finds the ID of its process group using the system call
140<code>getpgrp()</code>, or, equivalently, <code>getpgid(0)</code>.
141One finds the process group ID of process <code>p</code> using
142<code>getpgid(p)</code>.
143</p><p>One may use the command <code>ps j</code> to see PPID (parent process ID),
144PID (process ID), PGID (process group ID) and SID (session ID)
145of processes. With a shell that does not know about job control,
146like <code>ash</code>, each of its children will be in the same session
147and have the same process group as the shell. With a shell that knows
148about job control, like <code>bash</code>, the processes of one pipeline, like
149</p><blockquote>
150<pre>% cat paper | ideal | pic | tbl | eqn | ditroff &gt; out
151</pre>
152</blockquote>
153
154form a single process group.
155<p>
156</p><h3>Creation</h3>
157
158<p>A process <code>pid</code> is put into the process group <code>pgid</code> by
159</p><blockquote>
160<pre>setpgid(pid, pgid);
161</pre>
162</blockquote>
163
164If <code>pgid == pid</code> or <code>pgid == 0</code> then this creates
165a new process group with process group leader <code>pid</code>.
166Otherwise, this puts <code>pid</code> into the already existing
167process group <code>pgid</code>.
168A zero <code>pid</code> refers to the current process.
169The call <code>setpgrp()</code> is equivalent to <code>setpgid(0,0)</code>.
170<p>
171</p><h3>Restrictions on setpgid()</h3>
172
173<p>The calling process must be <code>pid</code> itself, or its parent,
174and the parent can only do this before <code>pid</code> has done
175<code>exec()</code>, and only when both belong to the same session.
176It is an error if process <code>pid</code> is a session leader
177(and this call would change its <code>pgid</code>).
178</p><p>
179</p><h3>Typical sequence</h3>
180
181<p>
182</p><blockquote>
183<pre>p = fork();
184if (p == (pid_t) -1) {
185        /* ERROR */
186} else if (p == 0) {    /* CHILD */
187        setpgid(0, pgid);
188        ...
189} else {                /* PARENT */
190        setpgid(p, pgid);
191        ...
192}
193</pre>
194</blockquote>
195
196This ensures that regardless of whether parent or child is scheduled
197first, the process group setting is as expected by both.
198<p>
199</p><h3>Signalling and waiting</h3>
200
201<p>One can signal all members of a process group:
202</p><blockquote>
203<pre>killpg(pgrp, sig);
204</pre>
205</blockquote>
206<p>One can wait for children in ones own process group:
207</p><blockquote>
208<pre>waitpid(0, &amp;status, ...);
209</pre>
210</blockquote>
211
212or in a specified process group:
213<blockquote>
214<pre>waitpid(-pgrp, &amp;status, ...);
215</pre>
216</blockquote>
217<p>
218</p><h3>Foreground process group</h3>
219
220<p>Among the process groups in a session at most one can be
221the <i>foreground process group</i> of that session.
222The tty input and tty signals (signals generated by ^C, ^Z, etc.)
223go to processes in this foreground process group.
224</p><p>A process can determine the foreground process group in its session
225using <code>tcgetpgrp(fd)</code>, where <code>fd</code> refers to its
226controlling tty. If there is none, this returns a random value
227larger than 1 that is not a process group ID.
228</p><p>A process can set the foreground process group in its session
229using <code>tcsetpgrp(fd,pgrp)</code>, where <code>fd</code> refers to its
230controlling tty, and <code>pgrp</code> is a process group in
231its session, and this session still is associated to the controlling
232tty of the calling process.
233</p><p>How does one get <code>fd</code>? By definition, <code>/dev/tty</code>
234refers to the controlling tty, entirely independent of redirects
235of standard input and output. (There is also the function
236<code>ctermid()</code> to get the name of the controlling terminal.
237On a POSIX standard system it will return <code>/dev/tty</code>.)
238Opening the name of the
239controlling tty gives a file descriptor <code>fd</code>.
240</p><p>
241</p><h3>Background process groups</h3>
242
243<p>All process groups in a session that are not foreground
244process group are <i>background process groups</i>.
245Since the user at the keyboard is interacting with foreground
246processes, background processes should stay away from it.
247When a background process reads from the terminal it gets
248a SIGTTIN signal. Normally, that will stop it, the job control shell
249notices and tells the user, who can say <code>fg</code> to continue
250this background process as a foreground process, and then this
251process can read from the terminal. But if the background process
252ignores or blocks the SIGTTIN signal, or if its process group
253is orphaned (see below), then the read() returns an EIO error,
254and no signal is sent. (Indeed, the idea is to tell the process
255that reading from the terminal is not allowed right now.
256If it wouldn't see the signal, then it will see the error return.)
257</p><p>When a background process writes to the terminal, it may get
258a SIGTTOU signal. May: namely, when the flag that this must happen
259is set (it is off by default). One can set the flag by
260</p><blockquote>
261<pre>% stty tostop
262</pre>
263</blockquote>
264
265and clear it again by
266<blockquote>
267<pre>% stty -tostop
268</pre>
269</blockquote>
270
271and inspect it by
272<blockquote>
273<pre>% stty -a
274</pre>
275</blockquote>
276
277Again, if TOSTOP is set but the background process ignores or blocks
278the SIGTTOU signal, or if its process group is orphaned (see below),
279then the write() returns an EIO error, and no signal is sent.
280<p>
281</p><h3>Orphaned process groups</h3>
282
283<p>The process group leader is the first member of the process group.
284It may terminate before the others, and then the process group is
285without leader.
286</p><p>A process group is called <i>orphaned</i> when <i>the
287parent of every member is either in the process group
288or outside the session</i>.
289In particular, the process group of the session leader
290is always orphaned.
291</p><p>If termination of a process causes a process group to become
292orphaned, and some member is stopped, then all are sent first SIGHUP
293and then SIGCONT.
294</p><p>The idea is that perhaps the parent of the process group leader
295is a job control shell. (In the same session but a different
296process group.) As long as this parent is alive, it can
297handle the stopping and starting of members in the process group.
298When it dies, there may be nobody to continue stopped processes.
299Therefore, these stopped processes are sent SIGHUP, so that they
300die unless they catch or ignore it, and then SIGCONT to continue them.
301</p><p>Note that the process group of the session leader is already
302orphaned, so no signals are sent when the session leader dies.
303</p><p>Note also that a process group can become orphaned in two ways
304by termination of a process: either it was a parent and not itself
305in the process group, or it was the last element of the process group
306with a parent outside but in the same session.
307Furthermore, that a process group can become orphaned
308other than by termination of a process, namely when some
309member is moved to a different process group.
310</p><p>
311</p><h2><a name="ss10.3">10.3 Sessions</a>
312</h2>
313
314<p>Every process group is in a unique <i>session</i>.
315(When the process is created, it becomes a member of the session
316of its parent.)
317By convention, the session ID of a session
318equals the process ID of the first member of the session,
319called the <i>session leader</i>.
320A process finds the ID of its session using the system call
321<code>getsid()</code>.
322</p><p>Every session may have a <i>controlling tty</i>,
323that then also is called the controlling tty of each of
324its member processes.
325A file descriptor for the controlling tty is obtained by
326opening <code>/dev/tty</code>. (And when that fails, there was no
327controlling tty.) Given a file descriptor for the controlling tty,
328one may obtain the SID using <code>tcgetsid(fd)</code>.
329</p><p>A session is often set up by a login process. The terminal
330on which one is logged in then becomes the controlling tty
331of the session. All processes that are descendants of the
332login process will in general be members of the session.
333</p><p>
334</p><h3>Creation</h3>
335
336<p>A new session is created by
337</p><blockquote>
338<pre>pid = setsid();
339</pre>
340</blockquote>
341
342This is allowed only when the current process is not a process group leader.
343In order to be sure of that we fork first:
344<blockquote>
345<pre>p = fork();
346if (p) exit(0);
347pid = setsid();
348</pre>
349</blockquote>
350
351The result is that the current process (with process ID <code>pid</code>)
352becomes session leader of a new session with session ID <code>pid</code>.
353Moreover, it becomes process group leader of a new process group.
354Both session and process group contain only the single process <code>pid</code>.
355Furthermore, this process has no controlling tty.
356<p>The restriction that the current process must not be a process group leader
357is needed: otherwise its PID serves as PGID of some existing process group
358and cannot be used as the PGID of a new process group.
359</p><p>
360</p><h3>Getting a controlling tty</h3>
361
362<p>How does one get a controlling terminal? Nobody knows,
363this is a great mystery.
364</p><p>The System V approach is that the first tty opened by the process
365becomes its controlling tty.
366</p><p>The BSD approach is that one has to explicitly call
367</p><blockquote>
368<pre>ioctl(fd, TIOCSCTTY, 0/1);
369</pre>
370</blockquote>
371
372to get a controlling tty.
373<p>Linux tries to be compatible with both, as always, and this
374results in a very obscure complex of conditions. Roughly:
375</p><p>The <code>TIOCSCTTY</code> ioctl will give us a controlling tty,
376provided that (i) the current process is a session leader,
377and (ii) it does not yet have a controlling tty, and
378(iii) maybe the tty should not already control some other session;
379if it does it is an error if we aren't root, or we steal the tty
380if we are all-powerful.
381[vda: correction: third parameter controls this: if 1, we steal tty from
382any such session, if 0, we don't steal]
383</p><p>Opening some terminal will give us a controlling tty,
384provided that (i) the current process is a session leader, and
385(ii) it does not yet have a controlling tty, and
386(iii) the tty does not already control some other session, and
387(iv) the open did not have the <code>O_NOCTTY</code> flag, and
388(v) the tty is not the foreground VT, and
389(vi) the tty is not the console, and
390(vii) maybe the tty should not be master or slave pty.
391</p><p>
392</p><h3>Getting rid of a controlling tty</h3>
393
394<p>If a process wants to continue as a daemon, it must detach itself
395from its controlling tty. Above we saw that <code>setsid()</code>
396will remove the controlling tty. Also the ioctl TIOCNOTTY does this.
397Moreover, in order not to get a controlling tty again as soon as it
398opens a tty, the process has to fork once more, to assure that it
399is not a session leader. Typical code fragment:
400</p><p>
401</p><pre>        if ((fork()) != 0)
402                exit(0);
403        setsid();
404        if ((fork()) != 0)
405                exit(0);
406</pre>
407<p>See also <code>daemon(3)</code>.
408</p><p>
409</p><h3>Disconnect</h3>
410
411<p>If the terminal goes away by modem hangup, and the line was not local,
412then a SIGHUP is sent to the session leader.
413Any further reads from the gone terminal return EOF.
414(Or possibly -1 with <code>errno</code> set to EIO.)
415</p><p>If the terminal is the slave side of a pseudotty, and the master side
416is closed (for the last time), then a SIGHUP is sent to the foreground
417process group of the slave side.
418</p><p>When the session leader dies, a SIGHUP is sent to all processes
419in the foreground process group. Moreover, the terminal stops being
420the controlling terminal of this session (so that it can become
421the controlling terminal of another session).
422</p><p>Thus, if the terminal goes away and the session leader is
423a job control shell, then it can handle things for its descendants,
424e.g. by sending them again a SIGHUP.
425If on the other hand the session leader is an innocent process
426that does not catch SIGHUP, it will die, and all foreground processes
427get a SIGHUP.
428</p><p>
429</p><h2><a name="ss10.4">10.4 Threads</a>
430</h2>
431
432<p>A process can have several threads. New threads (with the same PID
433as the parent thread) are started using the <code>clone</code> system
434call using the <code>CLONE_THREAD</code> flag. Threads are distinguished
435by a <i>thread ID</i> (TID). An ordinary process has a single thread
436with TID equal to PID. The system call <code>gettid()</code> returns the
437TID. The system call <code>tkill()</code> sends a signal to a single thread.
438</p><p>Example: a process with two threads. Both only print PID and TID and exit.
439(Linux 2.4.19 or later.)
440</p><pre>% cat &lt;&lt; EOF &gt; gettid-demo.c
441#include &lt;unistd.h&gt;
442#include &lt;sys/types.h&gt;
443#define CLONE_SIGHAND   0x00000800
444#define CLONE_THREAD    0x00010000
445#include &lt;linux/unistd.h&gt;
446#include &lt;errno.h&gt;
447_syscall0(pid_t,gettid)
448
449int thread(void *p) {
450        printf("thread: %d %d\n", gettid(), getpid());
451}
452
453main() {
454        unsigned char stack[4096];
455        int i;
456
457        i = clone(thread, stack+2048, CLONE_THREAD | CLONE_SIGHAND, NULL);
458        if (i == -1)
459                perror("clone");
460        else
461                printf("clone returns %d\n", i);
462        printf("parent: %d %d\n", gettid(), getpid());
463}
464EOF
465% cc -o gettid-demo gettid-demo.c
466% ./gettid-demo
467clone returns 21826
468parent: 21825 21825
469thread: 21826 21825
470%
471</pre>
472<p>
473</p><p>
474</p><hr>
475
476</body></html>
Note: See TracBrowser for help on using the repository browser.