From: Darren Hart <dvhart@linux.intel.com>
To: Eric Dumazet <eric.dumazet@gmail.com>
Cc: Peter Zijlstra <peterz@infradead.org>,
David Oliver <david@rgmadvisors.com>,
linux-kernel@vger.kernel.org,
Shawn Bohrer <sbohrer@rgmadvisors.com>,
Zachary Vonler <zvonler@rgmadvisors.com>,
KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com>,
Hugh Dickins <hughd@google.com>,
Thomas Gleixner <tglx@linutronix.de>, Ingo Molnar <mingo@elte.hu>
Subject: Re: Change in functionality of futex() system call.
Date: Fri, 24 Jun 2011 17:00:49 -0700 [thread overview]
Message-ID: <4E052531.3010603@linux.intel.com> (raw)
In-Reply-To: <1307384871.3098.98.camel@edumazet-laptop>
Hi Eric,
I'm finally getting time to review this in depth and try to help Shawn
get his fix upstream. Trying to make sure I have all the facets of this
straight in my head... or on paper at least ;-)
On 06/06/2011 11:27 AM, Eric Dumazet wrote:
> Le lundi 06 juin 2011 à 20:23 +0200, Peter Zijlstra a écrit :
>
>>
>> That's really not the point, what do we do when the COW happens during
>> the FUTEX_WAIT? At that point the process vaddr changes mapping and we
>> cannot continue the wait on the old page, since that would expose
>> invisible information, nor can we switch to the new page since we queued
>> on the old page.
>>
>> Therefore we have to force the COW and queue on the private copy, it
>> really is the only semi sane semantic.
>
> The point is we dont necessarly have to COW the page. If you attempt
> this COW, you shoot on user that did not expect to have a COW.
>
> Take this program : COW is not allowed, still this worked on 2.6.18 (it
> waits until another process change the value in file and call
> futex_wait())
>
> Using PROT_READ | PROT_WRITE instead of PROT_READ was OK too.
>
> (If we use PROT_READ | PROT_WRITE, then after your patch, program doesnt
> work anymore since this process gets a private page after your hidden
> COW : It'll wait forever)
As I understand MMAP(2), this is working due to undefined behavior as
Stephen pointed out earlier:
"It is unspecified whether changes made to the file after the mmap()
call are visible in the mapped region."
I don't think we are under any obligation to keep that working.
--
Darren
>
> #include <errno.h>
> #include <fcntl.h>
> #include <stdint.h>
> typedef uint32_t u32; // for futex.h
> #include <linux/futex.h>
> #include <sys/mman.h>
> #include <sys/syscall.h>
> #include <unistd.h>
>
>
> int main(int argc, char *argv[]) {
> int fd, *futex, rc, val = 42;
>
> fd = open("/tmp/futex_test", O_RDWR|O_CREAT, 0644);
> write(fd, &val, 4);
> futex = (int *)mmap(0, sizeof(int), PROT_READ, MAP_PRIVATE, fd, 0);
> rc = syscall(SYS_futex, futex, FUTEX_WAIT, val, 0, 0, 0);
> printf("rc=%d errno=%d\n", rc, errno);
> }
>
>
>
> --
> To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at http://vger.kernel.org/majordomo-info.html
> Please read the FAQ at http://www.tux.org/lkml/
--
Darren Hart
Intel Open Source Technology Center
Yocto Project - Linux Kernel
next prev parent reply other threads:[~2011-06-25 0:00 UTC|newest]
Thread overview: 99+ messages / expand[flat|nested] mbox.gz Atom feed top
2011-06-06 14:28 Change in functionality of futex() system call David Oliver
2011-06-06 15:23 ` Eric Dumazet
2011-06-06 15:56 ` Shawn Bohrer
2011-06-06 16:11 ` Peter Zijlstra
2011-06-06 16:16 ` Peter Zijlstra
2011-06-06 16:22 ` Eric Dumazet
2011-06-06 16:29 ` Peter Zijlstra
2011-06-06 16:42 ` Eric Dumazet
2011-06-06 17:05 ` Peter Zijlstra
2011-06-06 17:11 ` Eric Dumazet
2011-06-06 17:27 ` Steven Rostedt
2011-06-06 17:56 ` Darren Hart
2011-06-06 18:23 ` Peter Zijlstra
2011-06-06 18:27 ` Eric Dumazet
2011-06-25 0:00 ` Darren Hart [this message]
2011-06-27 16:48 ` Shawn Bohrer
2011-06-06 17:53 ` Darren Hart
2011-06-06 18:11 ` Eric Dumazet
2011-06-07 3:13 ` Darren Hart
2011-06-07 3:49 ` Eric Dumazet
2011-06-07 14:44 ` Andy Lutomirski
2011-06-07 15:56 ` Darren Hart
2011-06-07 15:58 ` Eric Dumazet
2011-06-07 18:43 ` Andrew Lutomirski
2011-06-07 19:01 ` Darren Hart
2011-06-07 19:04 ` Andrew Lutomirski
2011-06-07 19:06 ` Eric Dumazet
2011-06-07 19:10 ` David Oliver
2011-06-07 19:19 ` Andrew Lutomirski
2011-06-07 19:33 ` David Oliver
2011-06-07 19:53 ` Andrew Lutomirski
2011-06-07 20:04 ` David Oliver
2011-06-07 20:12 ` Andrew Lutomirski
2011-06-07 22:26 ` Kyle Moffett
2011-06-08 15:20 ` David Oliver
2011-06-08 15:21 ` Andrew Lutomirski
2011-06-08 16:21 ` Darren Hart
2011-06-09 11:37 ` KOSAKI Motohiro
2011-06-09 12:05 ` Peter Zijlstra
2011-06-09 17:58 ` Peter Zijlstra
2011-06-10 3:30 ` KOSAKI Motohiro
2011-06-10 3:26 ` KOSAKI Motohiro
2011-06-07 18:30 ` Joel Becker
2011-06-09 12:05 ` Peter Zijlstra
2011-06-10 12:10 ` KOSAKI Motohiro
2011-06-10 17:29 ` Darren Hart
2011-06-13 2:11 ` KOSAKI Motohiro
2011-06-13 15:50 ` Darren Hart
2011-06-15 18:50 ` Shawn Bohrer
2011-06-15 18:54 ` Darren Hart
2011-06-17 13:40 ` Shawn Bohrer
2011-06-22 19:19 ` [PATCH RFC] futex: Fix regression with read only mappings Shawn Bohrer
2011-06-22 20:14 ` Darren Hart
2011-06-23 2:51 ` KOSAKI Motohiro
2011-06-23 15:26 ` Darren Hart
2011-06-23 19:49 ` Shawn Bohrer
2011-06-24 15:59 ` [PATCH v2] " Shawn Bohrer
2011-06-25 0:37 ` Darren Hart
2011-06-25 15:10 ` KOSAKI Motohiro
2011-06-27 16:40 ` Shawn Bohrer
2011-06-27 18:15 ` Peter Zijlstra
2011-06-27 20:41 ` Darren Hart
2011-06-27 21:08 ` Shawn Bohrer
2011-06-27 21:39 ` Darren Hart
2011-06-27 22:14 ` Shawn Bohrer
2011-06-27 23:17 ` Darren Hart
2011-06-27 22:22 ` [PATCH v3] " Shawn Bohrer
2011-06-28 10:54 ` Peter Zijlstra
2011-06-28 14:52 ` Darren Hart
2011-06-28 17:38 ` Shawn Bohrer
2011-06-28 20:58 ` Darren Hart
2011-06-28 23:55 ` Darren Hart
2011-06-29 14:56 ` Shawn Bohrer
2011-06-29 15:17 ` [PATCH v4] " Shawn Bohrer
2011-06-29 18:41 ` Darren Hart
2011-06-29 23:38 ` Thomas Gleixner
2011-06-30 4:19 ` Darren Hart
2011-06-30 14:02 ` David C. Oliver
2011-06-30 15:41 ` Darren Hart
2011-06-30 16:21 ` [PATCH v5] " Shawn Bohrer
2011-07-12 15:27 ` Shawn Bohrer
2011-07-25 15:20 ` Shawn Bohrer
2011-07-25 19:28 ` Thomas Gleixner
2011-07-26 19:04 ` [tip:core/urgent] " tip-bot for Shawn Bohrer
2011-06-28 10:50 ` [PATCH v2] " Peter Zijlstra
2011-06-28 14:19 ` Darren Hart
2011-06-28 14:23 ` Peter Zijlstra
2011-06-23 3:58 ` [PATCH RFC] " Shawn Bohrer
2011-06-23 3:23 ` Change in functionality of futex() system call KOSAKI Motohiro
-- strict thread matches above, loose matches on Subject: below --
2011-06-09 0:44 George Spelvin
2011-06-09 3:02 ` Darren Hart
2011-06-09 3:38 ` Andrew Lutomirski
2011-06-09 3:54 ` Eric Dumazet
2011-06-09 4:10 ` Andrew Lutomirski
2011-06-09 5:11 ` Eric Dumazet
2011-06-09 12:12 ` Andrew Lutomirski
2011-06-09 4:43 ` George Spelvin
2011-06-09 5:25 ` Eric Dumazet
2011-06-09 4:44 ` Kyle Moffett
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=4E052531.3010603@linux.intel.com \
--to=dvhart@linux.intel.com \
--cc=david@rgmadvisors.com \
--cc=eric.dumazet@gmail.com \
--cc=hughd@google.com \
--cc=kosaki.motohiro@jp.fujitsu.com \
--cc=linux-kernel@vger.kernel.org \
--cc=mingo@elte.hu \
--cc=peterz@infradead.org \
--cc=sbohrer@rgmadvisors.com \
--cc=tglx@linutronix.de \
--cc=zvonler@rgmadvisors.com \
/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