According to Oracle's Multithreaded Programming Guide, fork()
should be safe-to-use inside signal handlers. But my process got stuck inside signal handler with to following back trace:
#0 __lll_lock_wait_private () at ../nptl/sysdeps/unix/sysv/linux/x86_64/lowlevellock.S:95
#1 0x00007f86e6a9990d in _L_lock_48 () from /lib/x86_64-linux- gnu/libc.so.6
#2 0x00007f86e6a922ec in ptmalloc_lock_all () at arena.c:242
#3 0x00007f86e6ad5e82 in __libc_fork () at ./nptl/sysdeps/unix/sysv/linux/x86_64/../fork.c:95
#4 0x00007f86e7d9f125 in __fork () at ./nptl/sysdeps/unix/sysv/linux/pt-fork.c:25
....
#7 signal handler called
So as malloc
is not safe to be use in signal handler how fork
can be?
Thanks in advance.
This is now listed as a bug by RedHat:
Bug 1422161 - glibc: fork is not async-signal-safe
...
+++ This bug was initially created as a clone of Bug #1422159 +++
POSIX requires that fork is async-signal-safe. Our current
implementation is not.
fork() will start a new process by copying some of memory of parents, but both are separate processes. So it is safe to use inside signal handler. Below is example of running child....
#include <stdio.h>
#include <unistd.h>
#include <signal.h>
#include <stdlib.h>
int SHOULD_RUN = 1;
void int_sig_handler(int signal){
printf("SIG_INT\n");
if(fork() == 0){
// child code
printf("I am Child");
//signal(SIGINT, int_sig_handler);
SHOULD_RUN = 1;
while(SHOULD_RUN){
printf("I am Running Still...\n");
sleep(1);
}
exit(0);
}else{
// parent code
printf("parent");
}
SHOULD_RUN = 0;
}
int main(int argc, const char *argv[]){
signal(SIGINT, int_sig_handler);
while(SHOULD_RUN){
sleep(1);
}
return 0;
}