From: Thomas Gleixner <tglx@linutronix.de>
To: Torvald Riegel <triegel@redhat.com>
Cc: "Michael Kerrisk (man-pages)" <mtk.manpages@gmail.com>,
"Carlos O'Donell" <carlos@redhat.com>,
Darren Hart <dvhart@linux.intel.com>, Ingo Molnar <mingo@elte.hu>,
Jakub Jelinek <jakub@redhat.com>,
"linux-man@vger.kernel.org" <linux-man@vger.kernel.org>,
lkml <linux-kernel@vger.kernel.org>,
Davidlohr Bueso <davidlohr.bueso@hp.com>,
Arnd Bergmann <arnd@arndb.de>,
Steven Rostedt <rostedt@goodmis.org>,
Peter Zijlstra <peterz@infradead.org>,
Linux API <linux-api@vger.kernel.org>,
Darren Hart <dvhart@infradead.org>,
Anton Blanchard <anton@samba.org>, Petr Baudis <pasky@suse.cz>,
Eric Dumazet <edumazet@google.com>,
bill o gallmeister <bgallmeister@gmail.com>,
Jan Kiszka <jan.kiszka@siemens.com>,
Daniel Wagner <wagi@monom.org>, Rich Felker <dalias@libc.org>
Subject: Re: futex(2) man page update help request
Date: Sat, 24 Jan 2015 12:35:01 +0100 (CET) [thread overview]
Message-ID: <alpine.DEB.2.11.1501241116160.5526@nanos> (raw)
In-Reply-To: <1422037788.29655.0.camel@triegel.csb>
On Fri, 23 Jan 2015, Torvald Riegel wrote:
> Second, the current documentation for EINTR is that it can happen due to
> receiving a signal *or* due to a spurious wake-up. This is difficult to
I don't think so. I went through all callchains again with a fine comb.
futex_wait()
retry:
ret = futex_wait_setup();
if (ret) {
/*
* Possible return codes related to uaddr:
* -EINVAL: Not u32 aligned uaddr
* -EFAULT: No mapping, no RW
* -ENOMEM: Paging ran out of memory
* -EHWPOISON: Memory hardware error
*
* Others:
* -EWOULDBLOCK: value at uaddr has changed
*/
return ret;
}
futex_wait_queue_me();
if (woken by futex_wake/requeue)
return 0;
if (timeout)
return -ETIMEOUT;
/*
* Spurious wakeup, i.e. no signal pending
*/
if (!signal_pending())
goto retry;
/* Handled in the low level syscall exit code */
if (!timed_wait)
return -ERESTARTSYS;
else
return -ERESTARTBLOCK;
Now in the low level syscall exit we try to deliver the signal
if (!signal_delivered())
restart_syscall();
if (sigaction->flags & SA_RESTART)
restart_syscall();
ret_to_userspace -EINTR;
So we should never see -EINTR in the case of a spurious wakeup here.
But, here is the not so good news:
I did some archaeology. The restart handling of futex_wait() got
introduced in kernel 2.6.22, so anything older than that will have
the spurious -EINTR issues.
futex_wait_pi() always had the restart handling and glibc folks back
then (2006) requested that it should never return -EINTR, so it
unconditionally restarts the syscall whether a signal had been
delivered or not.
So kernels >= 2.6.22 should never return -EINTR spuriously. If that
happens it's a bug and needs to be fixed.
> Third, I think it would be useful to -- somewhere -- explain which
> behavior the futex operations would have conceptually when expressed by
> C11 code. We currently say that they wake up, sleep, etc, and which
> values they return. But we never say how to properly synchronize with
> them on the userspace side. The C11 memory model is probably the best
> model to use on the userspace side, so that's why I'm arguing for this.
> Basically, I think we need to (1) tell people that they should use
> memory_order_relaxed accesses to the futex variable (ie, the memory
> location associated with the whole futex construct on the kernel side --
> or do we have another name for this?), and (2) give some conceptual
> guarantees for the kernel-side synchronization so that one use this to
> derive how to use them correctly in userspace.
>
> The man pages might not be the right place for this, and maybe we just
> need a revision of "Futexes are tricky". If you have other suggestions
> for where to document this, or on the content, let me know. (I'm also
> willing to spend time on this :) ).
The current futex code in the kernel has gained documentation about
the required memory ordering recently. That should be a good starting
point.
Thanks,
tglx
next prev parent reply other threads:[~2015-01-24 11:35 UTC|newest]
Thread overview: 80+ messages / expand[flat|nested] mbox.gz Atom feed top
2014-05-14 10:35 futex(2) man page update help request Michael Kerrisk (man-pages)
2014-05-14 16:18 ` Darren Hart
2014-05-14 19:03 ` Michael Kerrisk (man-pages)
2014-05-14 19:59 ` Darren Hart
2014-05-14 20:23 ` Carlos O'Donell
2014-05-14 20:44 ` Andy Lutomirski
2014-05-14 23:34 ` Thomas Gleixner
2014-05-15 3:12 ` Carlos O'Donell
2014-05-15 4:49 ` Michael Kerrisk (man-pages)
2014-05-15 4:53 ` Michael Kerrisk (man-pages)
2014-05-15 14:14 ` Thomas Gleixner
2014-05-15 20:19 ` Michael Kerrisk (man-pages)
2014-08-04 14:46 ` Carlos O'Donell
2014-05-15 20:35 ` Darren Hart
2015-01-15 15:12 ` Michael Kerrisk (man-pages)
2015-01-17 1:33 ` Darren Hart
2015-01-17 9:16 ` Michael Kerrisk (man-pages)
2015-01-17 19:26 ` Darren Hart
2015-01-18 10:18 ` Michael Kerrisk (man-pages)
2015-01-15 15:10 ` Michael Kerrisk (man-pages)
2015-01-15 22:23 ` Thomas Gleixner
2015-01-16 15:17 ` Michael Kerrisk (man-pages)
2015-01-16 15:20 ` Thomas Gleixner
2015-01-16 20:54 ` Michael Kerrisk (man-pages)
2015-01-17 0:46 ` Darren Hart
2015-01-19 10:45 ` Thomas Gleixner
2015-01-19 14:07 ` Michael Kerrisk (man-pages)
2015-01-23 18:19 ` Torvald Riegel
2015-01-24 10:05 ` Thomas Gleixner
2015-01-24 12:58 ` Torvald Riegel
2015-01-24 16:25 ` Thomas Gleixner
2015-01-17 0:56 ` Davidlohr Bueso
2015-01-17 1:11 ` Darren Hart
2015-01-23 18:29 ` Torvald Riegel
2015-01-24 11:35 ` Thomas Gleixner [this message]
2015-01-24 13:12 ` Torvald Riegel
2015-01-27 7:48 ` Michael Kerrisk (man-pages)
2015-02-05 19:57 ` Darren Hart
2014-05-15 8:13 ` Peter Zijlstra
2014-05-15 15:43 ` Darren Hart
2014-05-15 8:14 ` Peter Zijlstra
2014-05-15 13:18 ` Carlos O'Donell
2014-05-15 13:22 ` Peter Zijlstra
2014-05-15 13:49 ` Michael Kerrisk (man-pages)
2014-05-15 13:55 ` Peter Zijlstra
2014-05-15 14:39 ` Carlos O'Donell
2014-05-15 15:11 ` Peter Zijlstra
2014-05-14 20:56 ` Davidlohr Bueso
2014-05-14 21:03 ` Darren Hart
2014-05-14 22:21 ` Paul E. McKenney
2014-05-15 0:28 ` H. Peter Anvin
2014-05-15 0:35 ` Andy Lutomirski
2014-05-15 0:41 ` H. Peter Anvin
2014-05-15 19:10 ` Carlos O'Donell
2014-05-14 21:05 ` Davidlohr Bueso
2014-05-15 15:15 ` Joseph S. Myers
2014-05-15 0:18 ` H. Peter Anvin
2014-05-15 5:21 ` Darren Hart
2014-05-15 8:23 ` Peter Zijlstra
2014-05-15 13:46 ` Michael Kerrisk (man-pages)
2014-05-15 14:59 ` H. Peter Anvin
2014-05-15 15:42 ` chrubis
2014-05-15 15:52 ` H. Peter Anvin
2014-05-15 16:01 ` chrubis
2014-05-15 16:07 ` H. Peter Anvin
2014-05-15 16:17 ` chrubis
2014-05-15 16:56 ` H. Peter Anvin
2014-05-15 17:06 ` chrubis
2014-05-15 15:47 ` Darren Hart
2014-05-15 15:35 ` chrubis
2014-05-15 15:28 ` chrubis
2014-05-15 15:40 ` Steven Rostedt
2014-05-15 16:14 ` Darren Hart
2014-05-15 16:30 ` chrubis
2014-05-15 18:17 ` Darren Hart
2014-05-15 19:05 ` chrubis
2014-05-15 19:38 ` Darren Hart
2014-08-11 10:19 ` chrubis
2014-11-26 13:41 ` Cyril Hrubis
2015-02-16 13:14 ` Cyril Hrubis
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=alpine.DEB.2.11.1501241116160.5526@nanos \
--to=tglx@linutronix.de \
--cc=anton@samba.org \
--cc=arnd@arndb.de \
--cc=bgallmeister@gmail.com \
--cc=carlos@redhat.com \
--cc=dalias@libc.org \
--cc=davidlohr.bueso@hp.com \
--cc=dvhart@infradead.org \
--cc=dvhart@linux.intel.com \
--cc=edumazet@google.com \
--cc=jakub@redhat.com \
--cc=jan.kiszka@siemens.com \
--cc=linux-api@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-man@vger.kernel.org \
--cc=mingo@elte.hu \
--cc=mtk.manpages@gmail.com \
--cc=pasky@suse.cz \
--cc=peterz@infradead.org \
--cc=rostedt@goodmis.org \
--cc=triegel@redhat.com \
--cc=wagi@monom.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).