Subj : process vs. thread ids on linux
To   : comp.programming.threads
From : jpknott@gmail.com
Date : Thu Jun 02 2005 10:59 am

I've stumbled across something that violates some of my basic notions
about posix threads and process ids... Hopefully somebody will be able
to set me straight.

This is on a Redhat ES box, by the way, running a 2.4.21 kernel build.

I have been under the impression that when you start a new thread
(using pthread_create), it 'inherits' the pid of the 'parent' process.
  This would seem to be verified by calling getpid()... If I call this
inside the new thread, it returns the same value as the original
process.  Indeed, if I create 10 threads within a process, getpid()
returns the same pid for all of them.

However, as I was reading the man page for the 'ps' command, I
discovered the -m option, which purports to allow you to display 'all
threads'.   But if I do a 'ps -am', what I am apparently seeing is that
each thread has a different, unique pid!

For example, here I have a process called ttest.  The original process
has a pid of 19103.  It spawns 10 threads.    In my debugging logs,
each thread reports a pid of 19103 (as returned from getpid()).
However, a ps -am gives me this:

19103 pts/2    00:00:00 ttest
19104 pts/2    00:00:00 ttest
19105 pts/2    00:00:00 ttest
19106 pts/2    00:00:00 ttest
19107 pts/2    00:00:00 ttest
19108 pts/2    00:00:00 ttest
19109 pts/2    00:00:00 ttest
19110 pts/2    00:00:00 ttest
19111 pts/2    00:00:00 ttest
19112 pts/2    00:00:00 ttest
19113 pts/2    00:00:00 ttest


So, what is going on here?   If the kernel is indeed assigning a new
pid to each thread, how can I get to that id (since getpid() doesn't
work).  Why doesn't getpid() work, anyway?    And can I then use these
pids to send signals directly to a particular thread (which I was under
the impression was not possible)?

.