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17 .\" manual page may be incorrect or out-of-date.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.
46 .\" 2015-01-17, Kees Cook <keescook@chromium.org>
47 .\" Added notes on ptrace SIGTRAP and SYS_SECCOMP.
49 .TH SIGACTION 2 2014-12-31 "Linux" "Linux Programmer's Manual"
51 sigaction \- examine and change a signal action
54 .B #include <signal.h>
56 .BI "int sigaction(int " signum ", const struct sigaction *" act ,
57 .BI " struct sigaction *" oldact );
61 Feature Test Macro Requirements for glibc (see
62 .BR feature_test_macros (7)):
67 _POSIX_C_SOURCE\ >=\ 1 || _XOPEN_SOURCE || _POSIX_SOURCE
70 _POSIX_C_SOURCE >= 199309L
75 system call is used to change the action taken by a process on
76 receipt of a specific signal.
79 for an overview of signals.)
82 specifies the signal and can be any valid signal except
89 is non-NULL, the new action for signal
95 is non-NULL, the previous action is saved in
100 structure is defined as something like:
105 void (*sa_handler)(int);
106 void (*sa_sigaction)(int, siginfo_t *, void *);
109 void (*sa_restorer)(void);
114 On some architectures a union is involved: do not assign to both
121 field is not intended for application use.
122 (POSIX does not specify a
125 Some further details of purpose of this field can be found in
129 specifies the action to be associated with
133 for the default action,
135 to ignore this signal, or a pointer to a signal handling function.
136 This function receives the signal number as its only argument.
146 specifies the signal-handling function for
148 This function receives the signal number as its first argument, a
151 as its second argument and a pointer to a
153 (cast to \fIvoid\ *\fP) as its third argument.
154 (Commonly, the handler function doesn't make any use of the third argument.
157 for further information about
161 specifies a mask of signals which should be blocked
162 (i.e., added to the signal mask of the thread in which
163 the signal handler is invoked)
164 during execution of the signal handler.
165 In addition, the signal which triggered the handler
166 will be blocked, unless the
171 specifies a set of flags which modify the behavior of the signal.
172 It is formed by the bitwise OR of zero or more of the following:
180 do not receive notification when child processes stop (i.e., when they
182 .BR SIGSTOP ", " SIGTSTP ", " SIGTTIN ", "
185 or resume (i.e., they receive
189 This flag is meaningful only when establishing a handler for
192 .BR SA_NOCLDWAIT " (since Linux 2.6)"
193 .\" To be precise: Linux 2.5.60 -- MTK
198 do not transform children into zombies when they terminate.
201 This flag is meaningful only when establishing a handler for
203 or when setting that signal's disposition to
208 flag is set when establishing a handler for
210 POSIX.1 leaves it unspecified whether a
212 signal is generated when a child process terminates.
215 signal is generated in this case;
216 on some other implementations, it is not.
219 Do not prevent the signal from being received from within its own signal
221 This flag is meaningful only when establishing a signal handler.
223 is an obsolete, nonstandard synonym for this flag.
226 Call the signal handler on an alternate signal stack provided by
228 If an alternate stack is not available, the default stack will be used.
229 This flag is meaningful only when establishing a signal handler.
232 Restore the signal action to the default upon entry to the signal handler.
233 This flag is meaningful only when establishing a signal handler.
235 is an obsolete, nonstandard synonym for this flag.
238 Provide behavior compatible with BSD signal semantics by making certain
239 system calls restartable across signals.
240 This flag is meaningful only when establishing a signal handler.
243 for a discussion of system call restarting.
246 .IR "Not intended for application use" .
247 This flag is used by C libraries to indicate that the
249 field contains the address of a "signal trampoline".
254 .BR SA_SIGINFO " (since Linux 2.2)"
255 The signal handler takes three arguments, not one.
258 should be set instead of
260 This flag is meaningful only when establishing a signal handler.
263 .\" field was added in Linux 2.1.86.)
270 is a struct with the following fields:
275 int si_signo; /* Signal number */
276 int si_errno; /* An errno value */
277 int si_code; /* Signal code */
278 int si_trapno; /* Trap number that caused
279 hardware-generated signal
280 (unused on most architectures) */
282 .\" The siginfo_t 'si_trapno' field seems to be used only on SPARC and Alpha;
283 .\" this page could use a little more detail on its purpose there.
284 pid_t si_pid; /* Sending process ID */
285 uid_t si_uid; /* Real user ID of sending process */
286 int si_status; /* Exit value or signal */
287 clock_t si_utime; /* User time consumed */
288 clock_t si_stime; /* System time consumed */
289 sigval_t si_value; /* Signal value */
290 int si_int; /* POSIX.1b signal */
291 void *si_ptr; /* POSIX.1b signal */
292 int si_overrun; /* Timer overrun count;
294 int si_timerid; /* Timer ID; POSIX.1b timers */
295 .\" In the kernel: si_tid
296 void *si_addr; /* Memory location which caused fault */
297 long si_band; /* Band event (was \fIint\fP in
298 glibc 2.3.2 and earlier) */
299 int si_fd; /* File descriptor */
300 short si_addr_lsb; /* Least significant bit of address
301 (since Linux 2.6.32) */
302 void *si_call_addr; /* Address of system call instruction
304 int si_syscall; /* Number of attempted system call
306 unsigned int si_arch; /* Architecture of attempted system call
312 .IR si_signo ", " si_errno " and " si_code
313 are defined for all signals.
315 is generally unused on Linux.)
316 The rest of the struct may be a union, so that one should
317 read only the fields that are meaningful for the given signal:
324 .IR si_pid " and " si_uid .
325 In addition, signals sent with
328 .IR si_int " and " si_ptr
329 with the values specified by the sender of the signal;
334 Signals sent by POSIX.1b timers (since Linux 2.6) fill in
340 field is an internal ID used by the kernel to identify
341 the timer; it is not the same as the timer ID returned by
342 .BR timer_create (2).
345 field is the timer overrun count;
346 this is the same information as is obtained by a call to
347 .BR timer_getoverrun (2).
348 These fields are nonstandard Linux extensions.
350 Signals sent for message queue notification (see the description of
355 .IR si_int / si_ptr ,
361 with the process ID of the message sender; and
363 with the real user ID of the message sender.
367 .IR si_pid ", " si_uid ", " si_status ", " si_utime ", and " si_stime ,
368 providing information about the child.
371 field is the process ID of the child;
373 is the child's real user ID.
376 field contains the exit status of the child (if
380 or the signal number that caused the process to change state.
385 contain the user and system CPU time used by the child process;
386 these fields do not include the times used by waited-for children (unlike
390 In kernels up to 2.6, and since 2.6.27, these fields report
392 .IR sysconf(_SC_CLK_TCK) .
393 In 2.6 kernels before 2.6.27,
394 a bug meant that these fields reported time in units
395 of the (configurable) system jiffy (see
398 .\" When si_utime and si_stime where originally implemented, the
399 .\" measurement unit was HZ, which was the same as clock ticks
400 .\" (sysconf(_SC_CLK_TCK)). In 2.6, HZ became configurable, and
401 .\" was *still* used as the unit to return the info these fields,
402 .\" with the result that the field values depended on the the
403 .\" configured HZ. Of course, the should have been measured in
404 .\" USER_HZ instead, so that sysconf(_SC_CLK_TCK) could be used to
405 .\" convert to seconds. I have a queued patch to fix this:
406 .\" http://thread.gmane.org/gmane.linux.kernel/698061/ .
407 .\" This patch made it into 2.6.27.
408 .\" But note that these fields still don't return the times of
409 .\" waited-for children (as is done by getrusage() and times()
410 .\" and wait4()). Solaris 8 does include child times.
420 with the address of the fault.
421 On some architectures,
422 these signals also fill in the
433 This field indicates the least significant bit of the reported address
434 and therefore the extent of the corruption.
435 For example, if a full page was corrupted,
438 .IR log2(sysconf(_SC_PAGESIZE)) .
441 is delivered in response to a
443 event (PTRACE_EVENT_foo),
445 is not populated, but
449 are populated with the respective process ID and user ID responsible for
453 the tracee will be shown as delivering the event.
457 are Linux-specific extensions.
460 (the two names are synonyms on Linux)
462 .IR si_band " and " si_fd .
465 event is a bit mask containing the same values as are filled in the
471 field indicates the file descriptor for which the I/O event occurred.
474 generated (since Linux 3.5)
475 .\" commit a0727e8ce513fe6890416da960181ceb10fbfae6
476 when a seccomp filter returns
477 .BR SECCOMP_RET_TRAP ,
483 and other fields as described in
487 is a value (not a bit mask) indicating why this signal was sent.
494 and have the ptrace event in the high byte:
497 (SIGTRAP | PTRACE_EVENT_foo << 8).
500 For a regular signal, the following list shows the values which can be
503 for any signal, along with reason that the signal was generated.
518 .BR SI_MESGQ " (since Linux 2.6.6)"
519 POSIX message queue state changed; see
528 (only in kernels up to Linux 2.2; from Linux 2.4 onward
534 .BR SI_TKILL " (since Linux 2.4.19)"
538 .\" SI_DETHREAD is defined in 2.6.9 sources, but isn't implemented
539 .\" It appears to have been an idea that was tried during 2.5.6
540 .\" through to 2.5.24 and then was backed out.
543 The following values can be placed in
557 Illegal addressing mode.
572 Internal stack error.
575 The following values can be placed in
583 Integer divide by zero.
589 Floating-point divide by zero.
592 Floating-point overflow.
595 Floating-point underflow.
598 Floating-point inexact result.
601 Floating-point invalid operation.
604 Subscript out of range.
607 The following values can be placed in
615 Address not mapped to object.
618 Invalid permissions for mapped object.
621 The following values can be placed in
629 Invalid address alignment.
632 Nonexistent physical address.
635 Object-specific hardware error.
637 .BR BUS_MCEERR_AR " (since Linux 2.6.32)"
638 Hardware memory error consumed on a machine check; action required.
640 .BR BUS_MCEERR_AO " (since Linux 2.6.32)"
641 Hardware memory error detected in process but not consumed; action optional.
644 The following values can be placed in
657 .BR TRAP_BRANCH " (since Linux 2.4)"
658 Process taken branch trap.
660 .BR TRAP_HWBKPT " (since Linux 2.4)"
661 Hardware breakpoint/watchpoint.
664 The following values can be placed in
678 Child terminated abnormally.
681 Traced child has trapped.
686 .BR CLD_CONTINUED " (since Linux 2.6.9)"
687 Stopped child has continued.
690 The following values can be placed in
698 Data input available.
701 Output buffers available.
704 Input message available.
710 High priority input available.
716 The following value can be placed in
723 .BR SYS_SECCOMP " (since Linux 3.5)"
730 returns 0 on success; on error, \-1 is returned, and
732 is set to indicate the error.
736 .IR act " or " oldact
737 points to memory which is not a valid part of the process address space.
740 An invalid signal was specified.
741 This will also be generated if an attempt
742 is made to change the action for
743 .BR SIGKILL " or " SIGSTOP ", "
744 which cannot be caught or ignored.
747 .\" SVr4 does not document the EINTR condition.
751 inherits a copy of its parent's signal dispositions.
754 the dispositions of handled signals are reset to the default;
755 the dispositions of ignored signals are left unchanged.
757 According to POSIX, the behavior of a process is undefined after it
763 signal that was not generated by
767 Integer division by zero has undefined result.
768 On some architectures it will generate a
771 (Also dividing the most negative integer by \-1 may generate
773 Ignoring this signal might lead to an endless loop.
775 POSIX.1-1990 disallowed setting the action for
779 POSIX.1-2001 allows this possibility, so that ignoring
781 can be used to prevent the creation of zombies (see
783 Nevertheless, the historical BSD and System\ V behaviors for ignoring
785 differ, so that the only completely portable method of ensuring that
786 terminated children do not become zombies is to catch the
792 POSIX.1-1990 specified only
800 Use of these latter values in
802 may be less portable in applications intended for older
803 UNIX implementations.
807 flag is compatible with the SVr4 flag of the same name.
811 flag is compatible with the SVr4 flag of the same name under kernels
813 On older kernels the Linux implementation
814 allowed the receipt of any signal, not just the one we are installing
815 (effectively overriding any
820 can be called with a NULL second argument to query the current signal
822 It can also be used to check whether a given signal is valid for
823 the current machine by calling it with NULL second and third arguments.
825 It is not possible to block
826 .BR SIGKILL " or " SIGSTOP
827 (by specifying them in
829 Attempts to do so are silently ignored.
833 for details on manipulating signal sets.
837 for a list of the async-signal-safe functions that can be
838 safely called inside from inside a signal handler.
840 Before the introduction of
842 it was also possible to get some additional information,
845 with second argument of type
846 .IR "struct sigcontext".
847 See the relevant Linux kernel sources for details.
848 This use is obsolete now.
850 In kernels up to and including 2.6.13, specifying
854 prevents not only the delivered signal from being masked during
855 execution of the handler, but also the signals specified in
857 This bug was fixed in kernel 2.6.14.
866 .BR restart_syscall (2),
877 .BR siginterrupt (3),