Re: [PATCH 0/5] replace signal() with sigaction()

From: Johannes Sixt <j.sixt@viscovery.net>
Date: May 28, 2014, 07:40 UTC
Message-ID: <53859307.8080506@viscovery.net>
In-Reply-To: <1401257655-6043-1-git-send-email-jmmahler@gmail.com>

Am 5/28/2014 8:14, schrieb Jeremiah Mahler:

Show 9 quoted lines

> From signal(2)
> 
>   The behavior of signal() varies across UNIX versions, and has also var‐
>   ied historically across different versions of Linux.   Avoid  its  use:
>   use sigaction(2) instead.  See Portability below.
> 
> This patch set replaces calls to signal() with sigaction() in all files
> except sigchain.c.  sigchain.c is a bit more complicated than the others
> and will be done in a separate patch.

In compat/mingw.c we have:

int sigaction(int sig, struct sigaction *in, struct sigaction *out)
{
	if (sig != SIGALRM)
		return errno = EINVAL,
			error("sigaction only implemented for SIGALRM");
	if (out != NULL)
		return errno = EINVAL,
			error("sigaction: param 3 != NULL not implemented");

	timer_fn = in->sa_handler;
	return 0;
}

Notice "only implemented for SIGALRM". Are adding the missing signals somewhere (here or in a later patch)?

Show 11 quoted lines

> Jeremiah Mahler (5):
>   progress.c: replace signal() with sigaction()
>   daemon.c run_service(): replace signal() with sigaction()
>   daemon.c child_handler(): replace signal() with sigaction()
>   daemon.c service_loop(): replace signal() with sigaction()
>   connect.c: replace signal() with sigaction()
> 
>  connect.c  |  5 ++++-
>  daemon.c   | 15 ++++++++++++---
>  progress.c |  6 +++++-
>  3 files changed, 21 insertions(+), 5 deletions(-)

Isn't this a bit too much of code churn, given that there were no bug reports where signal handling is identified as the culprit despite the warning you cited above?

-- Hannes