[TUHS] fork

[Warner Losh via TUHS]

On Tue, May 12, 2026 at 5:19 AM Peter Jeremy via TUHS <tuhs at tuhs.org> wrote:

> On 2026-May-12 10:19:25 +0200, Folkert van Heusden via TUHS <tuhs at tuhs.org>
> wrote:
> >Now something unexpected is happening: after a request, a "<defunct>"
> proces
> >is left in the process list. Usually that is caused by a fork() not being
> >join()ed afterwards. But in main() of httpdp I invoke signal(SIGCHLD,
> >SIG_IGN); so I thought I took care of that?
>
> For "modern" Unixes (e.g. FreeBSD and Linux), there's a SA_NOCLDWAIT flag
> for sigaction(2) that explicitly says setting SIGCHLD to SIG_IGN means that
> child processes will not become zombies but will instead be implicitly
> reaped immediately.
>

Yes. This was defined in System V (I think r4, but I'm not 100% sure).
4.4BSD
didn't have it. FreeBSD added it in 1997. Linux had it earlier due to its
affinity
to System V, though I've not done a deep dive on this yet.


> This doesn't exist in BSD2.11 and, as far as I can tell from looking
> through the sources, there's no equivalent behaviour: When a process exits,
> the process becomes a zombie and the parent is sent a SIGCHLD (see
> kern_exit.c::exit()).  In psignal(), if a signal is marked SIG_IGN, the
> signal is completely ignored - there's no special handling of SIGCHLD.
> There's no obvious code to clean up zombie processes if the parent doesn't
> wait() for them.
>

Over simplifying a bit, 2.11BSD was 7th Edition, with lots of add-ons from
4.2BSD
and 4.3BSD, plus a number of pdp-11 only things that post-dated the 7th
edition (like
overlay support). Most of the kernel-side components originated from 4.2BSD
with a few updates from 4.3BSD. Since 4BSD at CSRG never had SA_NOCLDWAIT,
2.11 doesn't have it.

The zombies are only cleaned up when the parent exits and they wind up at
init.

Other reapers came much later.

Warner