signal(7): should it mention that SIGCHLD is also sent when child is continued?

signal(7): should it mention that SIGCHLD is also sent when child is continued?

[Arkadiusz Drabczyk]

Currently it says:

> SIGCHLD      P1990      Ign     Child stopped or terminated

It's the definition from POSIX 1990 which is referred here but the
modern POSIX says that it's also sent when child continues
https://pubs.opengroup.org/onlinepubs/9699919799/basedefs/signal.h.html:

> Child process terminated, stopped, or continued.

It's supported on Linux - in sigaction(2) a flag is described that
controls this and it correctly says that it's sent on stop and resume:

> SA_NOCLDSTOP
>
> If signum is SIGCHLD, do not receive notification when child processes
> stop (i.e., when they receive one of SIGSTOP, SIGTSTP, SIGTTIN, or
> SIGTTOU) or resume (i.e., they receive SIGCONT) (see wait(2)).  This
> flag is meaningful only when establishing a handler for SIGCHLD.

The question about wording in signal(7) has been originally asked by a
confused user here
https://unix.stackexchange.com/q/790116/72304. Should it be changed?

-- 
Arkadiusz Drabczyk <arkadiusz@drabczyk.org>

Re: signal(7): should it mention that SIGCHLD is also sent when child is continued?

[Alejandro Colomar]

[-- Attachment #1: Type: text/plain, Size: 1374 bytes --]

Hi Arkadiusz,

On Sat, Jan 25, 2025 at 07:58:34PM +0100, Arkadiusz Drabczyk wrote:
> Currently it says:
> 
> > SIGCHLD      P1990      Ign     Child stopped or terminated
> 
> It's the definition from POSIX 1990 which is referred here but the
> modern POSIX says that it's also sent when child continues
> https://pubs.opengroup.org/onlinepubs/9699919799/basedefs/signal.h.html:
> 
> > Child process terminated, stopped, or continued.
> 
> It's supported on Linux - in sigaction(2) a flag is described that
> controls this and it correctly says that it's sent on stop and resume:
> 
> > SA_NOCLDSTOP
> >
> > If signum is SIGCHLD, do not receive notification when child processes
> > stop (i.e., when they receive one of SIGSTOP, SIGTSTP, SIGTTIN, or
> > SIGTTOU) or resume (i.e., they receive SIGCONT) (see wait(2)).  This
> > flag is meaningful only when establishing a handler for SIGCHLD.
> 
> The question about wording in signal(7) has been originally asked by a
> confused user here
> https://unix.stackexchange.com/q/790116/72304. Should it be changed?

Yes, we should change the comment in signal(7) to say that a continued
child also generates that signal, and the comment should also mention
the POSIX version where this was introduced.  Would you mind sending a
patch?


Thanks!
Alex

-- 
<https://www.alejandro-colomar.es/>

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 833 bytes --]

Re: signal(7): should it mention that SIGCHLD is also sent when child is continued?

[Arkadiusz Drabczyk]

On Sun, Feb 02, 2025 at 12:06:09PM +0100, Alejandro Colomar wrote:
> Yes, we should change the comment in signal(7) to say that a continued
> child also generates that signal, and the comment should also mention
> the POSIX version where this was introduced.  Would you mind sending a
> patch?

No, I've just sent a patch.

-- 
Arkadiusz Drabczyk <arkadiusz@drabczyk.org>