[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Message-ID: <alpine.DEB.1.10.0812131026120.28976@alien.or.mcafeemobile.com>
Date: Sat, 13 Dec 2008 10:39:13 -0800 (PST)
From: Davide Libenzi <davidel@...ilserver.org>
To: Scott James Remnant <scott@...split.com>
cc: Casey Dahlin <cdahlin@...hat.com>,
Linux Kernel <linux-kernel@...r.kernel.org>
Subject: Re: [RFC PATCH] waitfd: file descriptor to wait on child processes
On Sat, 13 Dec 2008, Scott James Remnant wrote:
> On Fri, 2008-12-12 at 20:29 -0800, Davide Libenzi wrote:
>
> > And how about this?
> >
> > sfd = signalfd(SIGCHLD);
> >
> > for (;;) {
> > poll(sfd, POLLIN);
> > while ((pid = waitpid(0, &status, WNOHANG)) != -1)
> > process_child_death(pid);
> > }
> >
> At this point, why have signalfd()'s read() return siginfo_t at all?
>
> You have to discard the entire information since it's only ever the
> first signal that matched, all subsequent ones are thrown away.
Signalfd is a wakeup gate available to the poll/select/epoll subsystems.
Signals that are queued inside the kernel, get de-queueing behavior, while
signals that are overlapping, get overlapping behavior.
The code above enable you to wait for signals inside a poll/select/epoll
dispatch loop, and get all the info waitpid() can return to you - w/out
missing anything.
In a real app case, you're probably waiting for more than simple SIGCHLD,
so you likely will have one (or more) read(2) of the signalfd to find out
which signal you got, and do the waitpid(WNOHANG) loop in case of SIGCHLD.
- Davide
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
Powered by blists - more mailing lists