2 .\" Copyright (c) 1994,1995 Mike Battersby <mib@deakin.edu.au>
3 .\" and Copyright 2004, 2005 Michael Kerrisk <mtk.manpages@gmail.com>
4 .\" based on work by faith@cs.unc.edu
6 .\" %%%LICENSE_START(VERBATIM)
7 .\" Permission is granted to make and distribute verbatim copies of this
8 .\" manual provided the copyright notice and this permission notice are
9 .\" preserved on all copies.
11 .\" Permission is granted to copy and distribute modified versions of this
12 .\" manual under the conditions for verbatim copying, provided that the
13 .\" entire resulting derived work is distributed under the terms of a
14 .\" permission notice identical to this one.
16 .\" Since the Linux kernel and libraries are constantly changing, this
17 .\" manual page may be incorrect or out-of-date. The author(s) assume no
18 .\" responsibility for errors or omissions, or for damages resulting from
19 .\" the use of the information contained herein. The author(s) may not
20 .\" have taken the same level of care in the production of this manual,
21 .\" which is licensed free of charge, as they might when working
24 .\" Formatted or processed versions of this manual, if unaccompanied by
25 .\" the source, must acknowledge the copyright and authors of this work.
28 .\" Modified, aeb, 960424
29 .\" Modified Fri Jan 31 17:31:20 1997 by Eric S. Raymond <esr@thyrsus.com>
30 .\" Modified Thu Nov 26 02:12:45 1998 by aeb - add SIGCHLD stuff.
31 .\" Modified Sat May 8 17:40:19 1999 by Matthew Wilcox
32 .\" add POSIX.1b signals
33 .\" Modified Sat Dec 29 01:44:52 2001 by Evan Jones <ejones@uwaterloo.ca>
35 .\" Modified 2004-11-11 by Michael Kerrisk <mtk.manpages@gmail.com>
36 .\" Added mention of SIGCONT under SA_NOCLDSTOP
37 .\" Added SA_NOCLDWAIT
38 .\" Modified 2004-11-17 by Michael Kerrisk <mtk.manpages@gmail.com>
39 .\" Updated discussion for POSIX.1-2001 and SIGCHLD and sa_flags.
41 .\" 2004-12-09, mtk, added SI_TKILL + other minor changes
42 .\" 2005-09-15, mtk, split sigpending(), sigprocmask(), sigsuspend()
43 .\" out of this page into separate pages.
44 .\" 2010-06-11 Andi Kleen, add hwpoison signal extensions
45 .\" 2010-06-11 mtk, improvements to discussion of various siginfo_t fields.
47 .TH SIGACTION 2 2014-04-27 "Linux" "Linux Programmer's Manual"
49 sigaction \- examine and change a signal action
52 .B #include <signal.h>
54 .BI "int sigaction(int " signum ", const struct sigaction *" act ,
55 .BI " struct sigaction *" oldact );
59 Feature Test Macro Requirements for glibc (see
60 .BR feature_test_macros (7)):
65 _POSIX_C_SOURCE\ >=\ 1 || _XOPEN_SOURCE || _POSIX_SOURCE
68 _POSIX_C_SOURCE >= 199309L
73 system call is used to change the action taken by a process on
74 receipt of a specific signal.
77 for an overview of signals.)
80 specifies the signal and can be any valid signal except
87 is non-NULL, the new action for signal
93 is non-NULL, the previous action is saved in
98 structure is defined as something like:
103 void (*sa_handler)(int);
104 void (*sa_sigaction)(int, siginfo_t *, void *);
107 void (*sa_restorer)(void);
112 On some architectures a union is involved: do not assign to both
119 element is obsolete and should not be used.
120 POSIX does not specify a
125 specifies the action to be associated with
129 for the default action,
131 to ignore this signal, or a pointer to a signal handling function.
132 This function receives the signal number as its only argument.
142 specifies the signal-handling function for
144 This function receives the signal number as its first argument, a
147 as its second argument and a pointer to a
149 (cast to \fIvoid\ *\fP) as its third argument.
150 (Commonly, the handler function doesn't make any use of the third argument.
153 for further information about
157 specifies a mask of signals which should be blocked
158 (i.e., added to the signal mask of the thread in which
159 the signal handler is invoked)
160 during execution of the signal handler.
161 In addition, the signal which triggered the handler
162 will be blocked, unless the
167 specifies a set of flags which modify the behavior of the signal.
168 It is formed by the bitwise OR of zero or more of the following:
176 do not receive notification when child processes stop (i.e., when they
178 .BR SIGSTOP ", " SIGTSTP ", " SIGTTIN ", "
181 or resume (i.e., they receive
185 This flag is meaningful only when establishing a handler for
188 .BR SA_NOCLDWAIT " (since Linux 2.6)"
189 .\" To be precise: Linux 2.5.60 -- MTK
194 do not transform children into zombies when they terminate.
197 This flag is meaningful only when establishing a handler for
199 or when setting that signal's disposition to
204 flag is set when establishing a handler for
206 POSIX.1 leaves it unspecified whether a
208 signal is generated when a child process terminates.
211 signal is generated in this case;
212 on some other implementations, it is not.
215 Do not prevent the signal from being received from within its own signal
217 This flag is meaningful only when establishing a signal handler.
219 is an obsolete, nonstandard synonym for this flag.
222 Call the signal handler on an alternate signal stack provided by
224 If an alternate stack is not available, the default stack will be used.
225 This flag is meaningful only when establishing a signal handler.
228 Restore the signal action to the default upon entry to the signal handler.
229 This flag is meaningful only when establishing a signal handler.
231 is an obsolete, nonstandard synonym for this flag.
234 Provide behavior compatible with BSD signal semantics by making certain
235 system calls restartable across signals.
236 This flag is meaningful only when establishing a signal handler.
239 for a discussion of system call restarting.
241 .BR SA_SIGINFO " (since Linux 2.2)"
242 The signal handler takes three arguments, not one.
245 should be set instead of
247 This flag is meaningful only when establishing a signal handler.
250 .\" field was added in Linux 2.1.86.)
257 is a struct with the following elements:
262 int si_signo; /* Signal number */
263 int si_errno; /* An errno value */
264 int si_code; /* Signal code */
265 int si_trapno; /* Trap number that caused
266 hardware-generated signal
267 (unused on most architectures) */
269 .\" si_trapno seems to be only used on SPARC and Alpha;
270 .\" this page could use a little more detail on its purpose there.
271 pid_t si_pid; /* Sending process ID */
272 uid_t si_uid; /* Real user ID of sending process */
273 int si_status; /* Exit value or signal */
274 clock_t si_utime; /* User time consumed */
275 clock_t si_stime; /* System time consumed */
276 sigval_t si_value; /* Signal value */
277 int si_int; /* POSIX.1b signal */
278 void *si_ptr; /* POSIX.1b signal */
279 int si_overrun; /* Timer overrun count; POSIX.1b timers */
280 int si_timerid; /* Timer ID; POSIX.1b timers */
281 .\" In the kernel: si_tid
282 void *si_addr; /* Memory location which caused fault */
283 long si_band; /* Band event (was \fIint\fP in
284 glibc 2.3.2 and earlier) */
285 int si_fd; /* File descriptor */
286 short si_addr_lsb; /* Least significant bit of address
287 (since Linux 2.6.32) */
292 .IR si_signo ", " si_errno " and " si_code
293 are defined for all signals.
295 is generally unused on Linux.)
296 The rest of the struct may be a union, so that one should
297 read only the fields that are meaningful for the given signal:
304 .IR si_pid " and " si_uid .
305 In addition, signals sent with
308 .IR si_int " and " si_ptr
309 with the values specified by the sender of the signal;
314 Signals sent by POSIX.1b timers (since Linux 2.6) fill in
320 field is an internal ID used by the kernel to identify
321 the timer; it is not the same as the timer ID returned by
322 .BR timer_create (2).
325 field is the timer overrun count;
326 this is the same information as is obtained by a call to
327 .BR timer_getoverrun (2).
328 These fields are nonstandard Linux extensions.
330 Signals sent for message queue notification (see the description of
335 .IR si_int / si_ptr ,
341 with the process ID of the message sender; and
343 with the real user ID of the message sender.
347 .IR si_pid ", " si_uid ", " si_status ", " si_utime ", and " si_stime ,
348 providing information about the child.
351 field is the process ID of the child;
353 is the child's real user ID.
356 field contains the exit status of the child (if
360 or the signal number that caused the process to change state.
365 contain the user and system CPU time used by the child process;
366 these fields do not include the times used by waited-for children (unlike
370 In kernels up to 2.6, and since 2.6.27, these fields report
372 .IR sysconf(_SC_CLK_TCK) .
373 In 2.6 kernels before 2.6.27,
374 a bug meant that these fields reported time in units
375 of the (configurable) system jiffy (see
378 .\" When si_utime and si_stime where originally implemented, the
379 .\" measurement unit was HZ, which was the same as clock ticks
380 .\" (sysconf(_SC_CLK_TCK)). In 2.6, HZ became configurable, and
381 .\" was *still* used as the unit to return the info these fields,
382 .\" with the result that the field values depended on the the
383 .\" configured HZ. Of course, the should have been measured in
384 .\" USER_HZ instead, so that sysconf(_SC_CLK_TCK) could be used to
385 .\" convert to seconds. I have a queued patch to fix this:
386 .\" http://thread.gmane.org/gmane.linux.kernel/698061/ .
387 .\" This patch made it into 2.6.27.
388 .\" But note that these fields still don't return the times of
389 .\" waited-for children (as is done by getrusage() and times()
390 .\" and wait4()). Solaris 8 does include child times.
400 with the address of the fault.
401 .\" FIXME SIGTRAP also sets the following for ptrace_notify() ?
402 .\" info.si_code = exit_code;
403 .\" info.si_pid = task_pid_vnr(current);
404 .\" info.si_uid = current_uid(); /* Real UID */
405 On some architectures,
406 these signals also fill in the
417 This field indicates the least significant bit of the reported address
418 and therefore the extent of the corruption.
419 For example, if a full page was corrupted,
422 .IR log2(sysconf(_SC_PAGESIZE)) .
426 are Linux-specific extensions.
429 (the two names are synonyms on Linux)
431 .IR si_band " and " si_fd .
434 event is a bit mask containing the same values as are filled in the
440 field indicates the file descriptor for which the I/O event occurred.
443 is a value (not a bit mask)
444 indicating why this signal was sent.
445 The following list shows the values which can be placed in
447 for any signal, along with reason that the signal was generated.
463 POSIX message queue state changed (since Linux 2.6.6); see
472 (only in kernels up to Linux 2.2; from Linux 2.4 onward
483 .\" SI_DETHREAD is defined in 2.6.9 sources, but isn't implemented
484 .\" It appears to have been an idea that was tried during 2.5.6
485 .\" through to 2.5.24 and then was backed out.
488 The following values can be placed in
502 illegal addressing mode
520 The following values can be placed in
528 integer divide by zero
534 floating-point divide by zero
537 floating-point overflow
540 floating-point underflow
543 floating-point inexact result
546 floating-point invalid operation
549 subscript out of range
552 The following values can be placed in
560 address not mapped to object
563 invalid permissions for mapped object
566 The following values can be placed in
574 invalid address alignment
577 nonexistent physical address
580 object-specific hardware error
582 .BR BUS_MCEERR_AR " (since Linux 2.6.32)"
583 Hardware memory error consumed on a machine check; action required.
585 .BR BUS_MCEERR_AO " (since Linux 2.6.32)"
586 Hardware memory error detected in process but not consumed; action optional.
589 The following values can be placed in
602 .BR TRAP_BRANCH " (since Linux 2.4)"
603 process taken branch trap
605 .BR TRAP_HWBKPT " (since Linux 2.4)"
606 hardware breakpoint/watchpoint
609 The following values can be placed in
623 child terminated abnormally
626 traced child has trapped
632 stopped child has continued (since Linux 2.6.9)
635 The following values can be placed in
646 output buffers available
649 input message available
655 high priority input available
662 returns 0 on success; on error, \-1 is returned, and
664 is set to indicate the error.
668 .IR act " or " oldact
669 points to memory which is not a valid part of the process address space.
672 An invalid signal was specified.
673 This will also be generated if an attempt
674 is made to change the action for
675 .BR SIGKILL " or " SIGSTOP ", "
676 which cannot be caught or ignored.
679 .\" SVr4 does not document the EINTR condition.
683 inherits a copy of its parent's signal dispositions.
686 the dispositions of handled signals are reset to the default;
687 the dispositions of ignored signals are left unchanged.
689 According to POSIX, the behavior of a process is undefined after it
695 signal that was not generated by
699 Integer division by zero has undefined result.
700 On some architectures it will generate a
703 (Also dividing the most negative integer by \-1 may generate
705 Ignoring this signal might lead to an endless loop.
707 POSIX.1-1990 disallowed setting the action for
711 POSIX.1-2001 allows this possibility, so that ignoring
713 can be used to prevent the creation of zombies (see
715 Nevertheless, the historical BSD and System\ V behaviors for ignoring
717 differ, so that the only completely portable method of ensuring that
718 terminated children do not become zombies is to catch the
724 POSIX.1-1990 specified only
732 Use of these latter values in
734 may be less portable in applications intended for older
735 UNIX implementations.
739 flag is compatible with the SVr4 flag of the same name.
743 flag is compatible with the SVr4 flag of the same name under kernels
745 On older kernels the Linux implementation
746 allowed the receipt of any signal, not just the one we are installing
747 (effectively overriding any
752 can be called with a NULL second argument to query the current signal
754 It can also be used to check whether a given signal is valid for
755 the current machine by calling it with NULL second and third arguments.
757 It is not possible to block
758 .BR SIGKILL " or " SIGSTOP
759 (by specifying them in
761 Attempts to do so are silently ignored.
765 for details on manipulating signal sets.
769 for a list of the async-signal-safe functions that can be
770 safely called inside from inside a signal handler.
772 Before the introduction of
774 it was also possible to get some additional information,
777 with second argument of type
778 .IR "struct sigcontext".
779 See the relevant Linux kernel sources for details.
780 This use is obsolete now.
782 In kernels up to and including 2.6.13, specifying
786 prevents not only the delivered signal from being masked during
787 execution of the handler, but also the signals specified in
789 This bug was fixed in kernel 2.6.14.
798 .BR restart_syscall (2),
807 .BR siginterrupt (3),
814 This page is part of release 3.67 of the Linux
817 A description of the project,
818 information about reporting bugs,
819 and the latest version of this page,
821 \%http://www.kernel.org/doc/man\-pages/.