8

In Linux, when a child process terminates and it's parent has not yet waited on it, it becomes a zombie process. The child's exit code is stored in the pid descriptor.

If a SIGKILL is sent to the child, there is not supposed to be any effect.

Does this mean that the exit code will not be modified by the SIGKILL or will the exit code be modified to indicate that the child exited because it received a SIGKILL?

11

To answer that question, you have to understand how signals are sent to a process and how a process exists in the kernel.

Each process is represented as a task_struct inside the kernel (the definition is in the sched.h header file and beginns here). That struct holds information about the process; for instance the pid. The important information is in line 1566 where the associated signal is stored. This is set only if a signal is sent to the process.

A dead process or a zombie process still has a task_struct. The struct remains, until the parent process (natural or by adoption) has called wait() after recieving SIGCHLD to reap its child process. When a signal is sent, the signal_struct is set. It doesn't matter if the signal is a catchable one or not, in this case.

Signals are evaluated every time when the process runs. Or to be exact, before the process would run. The process is then in the TASK_RUNNING state. The kernel runs the schedule() routine which determines the next running process according to its scheduling algorithm. Assuming this process is the next running process. Then the value of the signal_struct is evaluated, whether there is a waiting signal to be handled or not. If a signal handler is manually defined (via signal() or sigaction()), the registered function is exeuted, if not the signal's default action is executed. The default action depends on the signal being sent.

For instance, the SIGSTOP signal's default handler will change the current process's state to TASK_STOPPED and then run schedule() to select a new process to run. Notice, SIGSTOP is not catchable (like SIGKILL), therefore there is no possibility to register a manual signal handler. In case of an uncatchable signal, the default action will always be executed.


To your question:

A defunct or dead process will never be determined by the scheduler to be in the TASK_RUNNING state again. Thus the kernel will never run the signal handler (default or defined) for the corresponding signal, whichever signal is was. Therefore the exit_signal will never be set again. The signal is "delievered" to the process by setting the signal_struct in task_struct of the process, but nothing else will happen, because the process will never run again. There is no code to run, all that remains of the process is that process struct.

If then the parent process however reaps its children by wait(), the exit code it recieves, is the one when the process "initially" died. It doesn't matter if there is a signal waiting to be handled.

8

A zombie process is basically already dead. The only thing is that nobody has acknowledged its death yet so it continues occupying an entry in the process table as well as a control block (the structure the Linux kernel maintains for every thread in activity). Other resources like mandatory locks on files, shared memory segments, semaphores, etc. are reclaimed.

You cannot signal them because nobody can act upon this signal. Even fatal signals like KILL are useless since the process has already terminated its execution. You can try yourself:

#include <stdio.h>
#include <stdlib.h>
#include <sys/types.h>
#include <unistd.h>
#include <sys/wait.h>
int main(void)
{
    pid_t pid = fork();

    if (pid == -1)
        exit(-1);

    if (pid > 0) {
        //parent
        printf("[parent]: I'm the parent, the pid of my child is %i\n"
            "I'll start waiting for it in 10 seconds.\n", pid);
        sleep(10);
        int status;
        wait(&status);

        if (WIFSIGNALED(status)) {
            printf("[parent]: My child has died from a signal: %i\n", WTERMSIG(status));
        } else if (WIFEXITED(status)) {
            printf("[parent]: My child has died from natural death\n");
        } else {
            printf("[parent]: I don't know what happened to my child\n");
        }
    } else {
        //child
        printf("[child]: I'm dying soon, try to kill me.\n");
        sleep(5);
        printf("[child]: Dying now!\n");
    }

    return 0;
}

Here, I start a process that forks and sleeps before waiting for its child. The child does nothing but sleep a little. You can kill the child when it's sleeping or just after it exits to see the difference:

$ make zombie 
cc     zombie.c   -o zombie

$ ./zombie    
[parent]: I'm the parent, the pid of my child is 16693
I'll start waiting for it in 10 seconds.
[child]: I'm dying soon, try to kill me.
# Here, I did "kill -15 16693" in another console
[parent]: My child has died from a signal: 15

$ ./zombie
[parent]: I'm the parent, the pid of my child is 16717
I'll start waiting for it in 10 seconds.
[child]: I'm dying soon, try to kill me.
[child]: Dying now!
# Here, I did "kill -15 16717" in another console
[parent]: My child has died from natural death
  • I'd like to find the relevant passage in the kernel source code for you but I'm struggling to find it... – lgeorget Jan 22 '16 at 8:25
  • @Igeorget Thanks but that's ok, I don't need to see the kernel code. – user137481 Jan 22 '16 at 16:26

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.