Re: system(3) semantics when SIGCHLD is SIG_IGN'd

[Taylor R Campbell <campbell+netbsd-tech-userlevel%mumble.net@localhost>]

> Date: Sun, 13 Aug 2023 13:03:09 +0200
> From: Rhialto <rhialto%falu.nl@localhost>
> 
> On Sat 12 Aug 2023 at 11:58:36 +0000, Taylor R Campbell wrote:
> > Cons:
> > - POSIX doesn't ask system(3) to work when SIGCHLD is set to SIG_IGN
> >   or when it has SA_NOCLDWAIT set, so this code is nonportable anyway;
> 
> I read through
> https://pubs.opengroup.org/onlinepubs/007908799/xsh/system.html and it
> seems entirely silent on the matter. How does this mean that system(3)
> is allowed to fail in this situation? Since the state is not mentioned
> as a special case, I expect that it is supposed to work.

Newer versions of POSIX give a sample implementation of system(3)
which fails the same way under these circumstances:

https://pubs.opengroup.org/onlinepubs/9699919799/functions/system.html

The only thing it does with SIGCHLD is block it temporarily.