Sponsored Content
Full Discussion: Nanosleep in signal call
Top Forums Programming Nanosleep in signal call Post 302963454 by jim mcnamara on Wednesday 30th of December 2015 06:33:56 PM
Old 12-30-2015
I believe rand() is not async safe as well - you cannot have interruptible calls in your signal handler. Well make that should not have. It uses global variables which is another problem. Plus you are doing a lot in your signal handler, so also consider blocking SIGUSR1 as the first step in your handler. Multiple signals can mess up your code otherwise - especially the rand() call. New applications should use the POSIX signal API (sigaction(2), sigprocmask(2), etc.) which is what you did and that is good.

Next:
Code:
signal_name = "SIGUSR1";

could SEGFAULT, but in any case it is undefined behavior: const *signal_name;

Last edited by jim mcnamara; 12-30-2015 at 07:39 PM..
 

9 More Discussions You Might Find Interesting

1. Programming

sigwait system call in UNIX signal

Hi Everybody, I have gone through man of sigwait and new to UNIX signals. Could anyone explain me about the following lines mentioned in sigwait man help ? "The selection of a signal in set is independent of the signal mask of the calling thread or LWP. This means a thread or LWP can ... (1 Reply)
Discussion started by: md7ahuja
1 Replies

2. Programming

nanosleep

Hello, Can anyone, please, guide me on the use of nanosleep. I'm learning threads. I want to introduce a delay (not nested for loops, something more customizable). Nanosleep looked useful (or any other form of customizable and easy-to-use delay). Sleep is too long. :) (2 Replies)
Discussion started by: ameya
2 Replies

3. UNIX for Advanced & Expert Users

how to differentiate system call from library call

Hi, Ho do I differentiate system call from library call? for example if I am using chmod , how do I find out if it is a system call or library call? Thanks Muru (2 Replies)
Discussion started by: muru
2 Replies

4. Programming

nanosleep returns prematurely, with return value 0

Hi, I have encountered the following problem on Solaris 10: I have a thread that is asleep on nanosleep (set to 24 hours). Something that happens on another thread, causes the nanosleep to exit, even though the time has not elapsed. The returned value is 0 (so it doesn't look like it... (1 Reply)
Discussion started by: MeMyself
1 Replies

5. Programming

Problem with signal handler and interrupted system call

Hi, I have a daq program that runs in an infinite loop until it receives SIGINT. A handler catches the signal and sets a flag to stop the while loop. After the loop some things have to be cleaned up. The problem is that I want my main while loop to wait until the next full second begins, to... (2 Replies)
Discussion started by: soeckel
2 Replies

6. Infrastructure Monitoring

diffrence between method call and function call in perl

Hello, I have a problem with package and name space. require "/Mehran/DSGateEngineLib/general.pl"; use strict; sub System_Status_Main_Service_Status_Intrusion_Prevention { my %idpstatus; my @result; &General_ReadHash("/var/dsg/idp/settings",\%idpstatus); #print... (4 Replies)
Discussion started by: Zaxon
4 Replies

7. UNIX for Dummies Questions & Answers

Why do I need to call make if I call gcc ?

Why do I need to call make if I call gcc ? I thought gcc already compiles the sources. thanks (1 Reply)
Discussion started by: aneuryzma
1 Replies

8. Programming

Where is nanosleep?

I use nanosleep under solaris10,like follows: #include <pthread.h> #include <time.h> #include <sys/time.h> struct to_info{ void (*to_fn)(void *); void *to_arg; struct timespec to_wait; }; void *timeout_helper(void *arg){ struct to_info *tip; nanosleep(&tip->to_wait,NULL);... (1 Reply)
Discussion started by: konvalo
1 Replies

9. Programming

nanosleep

Application runs on both solaris 6 and 10. solaris 6 having only posix4.so library, solaris 10 having libposix4.so and librt.so Can we link application to lposix4 instead of lrt for nanosleep, sothat application will run in both machines? (2 Replies)
Discussion started by: satish@123
2 Replies
BSD_SIGNAL(3)						     Linux Programmer's Manual						     BSD_SIGNAL(3)

NAME
bsd_signal - signal handling with BSD semantics SYNOPSIS
#define _XOPEN_SOURCE /* See feature_test_macros(7) */ #include <signal.h> typedef void (*sighandler_t)(int); sighandler_t bsd_signal(int signum, sighandler_t handler); DESCRIPTION
The bsd_signal() function takes the same arguments, and performs the same task, as signal(2). The difference between the two is that bsd_signal() is guaranteed to provide reliable signal semantics, that is: a) the disposition of the signal is not reset to the default when the handler is invoked; b) delivery of further instances of the signal is blocked while the signal handler is executing; and c) if the handler interrupts a blocking system call, then the system call is automatically restarted. A portable application cannot rely on signal(2) to provide these guarantees. RETURN VALUE
The bsd_signal() function returns the previous value of the signal handler, or SIG_ERR on error. ERRORS
As for signal(2). CONFORMING TO
4.2BSD, POSIX.1-2001. POSIX.1-2008 removes the specification of bsd_signal(), recommending the use of sigaction(2) instead. NOTES
Use of bsd_signal() should be avoided; use sigaction(2) instead. On modern Linux systems, bsd_signal() and signal(2) are equivalent. But on older systems, signal(2) provided unreliable signal semantics; see signal(2) for details. The use of sighandler_t is a GNU extension; this type is defined only if the _GNU_SOURCE feature test macro is defined. SEE ALSO
sigaction(2), signal(2), sysv_signal(3), signal(7) COLOPHON
This page is part of release 3.53 of the Linux man-pages project. A description of the project, and information about reporting bugs, can be found at http://www.kernel.org/doc/man-pages/. 2009-03-15 BSD_SIGNAL(3)
All times are GMT -4. The time now is 07:26 AM.
Unix & Linux Forums Content Copyright 1993-2022. All Rights Reserved.
Privacy Policy