Keep it simple, stupid!

Let's read code of two simple programs. First one comes with GNU coreutils in many Linux distributions. It is called "true" and usually installs as /bin/true :

/* Exit with a status code indicating success.
   Copyright (C) 1999-2012 Free Software Foundation, Inc.

   This program is free software: you can redistribute it and/or modify
   it under the terms of the GNU General Public License as published by
   the Free Software Foundation, either version 3 of the License, or
   (at your option) any later version.

   This program is distributed in the hope that it will be useful,
   but WITHOUT ANY WARRANTY; without even the implied warranty of
   MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
   GNU General Public License for more details.

   You should have received a copy of the GNU General Public License
   along with this program.  If not, see <http://www.gnu.org/licenses/>.  */

#include <config.h>
#include <stdio.h>
#include <sys/types.h>
#include <system.h>

/* Act like "true" by default; false.c overrides this.  */
#ifndef EXIT_STATUS
# define EXIT_STATUS EXIT_SUCCESS
#endif

#if EXIT_STATUS == EXIT_SUCCESS
# define PROGRAM_NAME "true"
#else
# define PROGRAM_NAME "false"
#endif

#define AUTHORS proper_name ("Jim Meyering")

void
usage (int status)
{
  printf (_("\
Usage: %s [ignored command line arguments]\n\
  or:  %s OPTION\n\
"),
          program_name, program_name);
  printf ("%s\n\n",
          _(EXIT_STATUS == EXIT_SUCCESS
            ? N_("Exit with a status code indicating success.")
            : N_("Exit with a status code indicating failure.")));
  fputs (HELP_OPTION_DESCRIPTION, stdout);
  fputs (VERSION_OPTION_DESCRIPTION, stdout);
  printf (USAGE_BUILTIN_WARNING, PROGRAM_NAME);
  emit_ancillary_info ();
  exit (status);
}

int
main (int argc, char **argv)
{
  /* Recognize --help or --version only if it's the only command-line
     argument.  */
  if (argc == 2)
    {
      initialize_main (&argc, &argv);
      set_program_name (argv[0]);
      setlocale (LC_ALL, "");
      bindtextdomain (PACKAGE, LOCALEDIR);
      textdomain (PACKAGE);

      atexit (close_stdout);

      if (STREQ (argv[1], "--help"))
        usage (EXIT_STATUS);

      if (STREQ (argv[1], "--version"))
        version_etc (stdout, PROGRAM_NAME, PACKAGE_NAME, Version, AUTHORS,
                     (char *) NULL);
    }

  exit (EXIT_STATUS);
}

Second one belongs to so-called "sbase" by suckless.org.

/* See LICENSE file for copyright and license details. */
#include <stdlib.h>

int
main(void)
{
 return EXIT_SUCCESS;
}

This programs does one thing. Nothing. Just exits, returning successful value. Who follows UNIX ideas closely? Who follows common sense? Answer this questions to yourself.

Links:

• KISS
• GNU Coreutils true
• suckless true

Linux Thread ID

There is some useful and interesting book about Linux programming, freely available in network: Advanced Linux Programming. One of outdated topics of this book is 4.5 GNU/Linux Thread Implementation. Book says that on GNU/Linux threads are implemented as processes. Next code should confirm that, but it is not do it for now:

#include <pthread.h>
#include <stdio.h>
#include <unistd.h>
void *thread_fun(void *arg)
{
    fprintf(stderr, "child thread pid is %d\n", (int)getpid());
    while(1);
    return NULL;
}

int main(void)
{
    pthread_t thread;
    fprintf(stderr, "main thread pid is %d\n", (int)getpid());
    pthread_create(&thread, NULL, &thread_fun, NULL);
    while(1);
    return 0;
}

So, instead of different process ID's we have one:

$ ./alp-thread-pid
main thread pid is 17280
child thread pid is 17280

How can we get thread ID? We can find simple function pthread_t pthread_self(void); in <pthread.h>. We must use this function, if we write portable program. This thread IDs are only guaranteed to be unique within a process, but may be reused after a terminated thread has been joined, or a detached thread has terminated. Value, returned by pthread_self, has opaque type - POSIX.1 allows it to be arithmetic type or a structure. So, we cannot print it if we don't know how implementation represents it.
However, Linux kernel uses its own thread identification. We can get kernel thread ID by call pid_t gettid(void); in <sys/types.h>; But glibc does not provide a wrapper for this system call! So, let's write it ourself and print it:

#include <pthread.h>
#include <stdio.h>
#include <unistd.h>
#include <sys/syscall.h>

pid_t gettid(void)
{
    return syscall(SYS_gettid);
}

void *thread_fun(void *arg)
{
    fprintf(stderr, "child thread pid is %d\n", (int)getpid());
    fprintf(stderr, "child kernel thread tid is %d\n", (int)gettid());
    while(1);
    return NULL;
}

int main(void)
{
    pthread_t thread;
    fprintf(stderr, "main thread pid is %d\n", (int)getpid());
    fprintf(stderr, "main kernel thread tid is %d\n", (int)gettid());
    pthread_create(&thread, NULL, &thread_fun, NULL);
    while(1);
    return 0;
}

Run it:

$ ./thread-pid 
main thread pid is 18866
main kernel thread tid is 18866
child thread pid is 18866
child kernel thread tid is 18867

We found our TID!