From: Jeff Moyer <jmoyer@redhat.com>
To: Jan Kara <jack@suse.cz>
Cc: Michael Kerrisk <mtk.manpages@gmail.com>,
LKML <linux-kernel@vger.kernel.org>,
linux-man@vger.kernel.org, linux-mm@kvack.org, mgorman@suse.de
Subject: Re: [PATCH] Describe race of direct read and fork for unaligned buffers
Date: Mon, 30 Apr 2012 09:41:50 -0400 [thread overview]
Message-ID: <x49haw1jq01.fsf@segfault.boston.devel.redhat.com> (raw)
In-Reply-To: <1335778207-6511-1-git-send-email-jack@suse.cz> (Jan Kara's message of "Mon, 30 Apr 2012 11:30:07 +0200")
Jan Kara <jack@suse.cz> writes:
> This is a long standing problem (or a surprising feature) in our implementation
> of get_user_pages() (used by direct IO). Since several attempts to fix it
> failed (e.g.
> http://linux.derkeiler.com/Mailing-Lists/Kernel/2009-04/msg06542.html, or
> http://lkml.indiana.edu/hypermail/linux/kernel/0903.1/01498.html refused in
> http://comments.gmane.org/gmane.linux.kernel.mm/31569) and it's not completely
> clear whether we really want to fix it given the costs, let's at least document
> it.
>
> CC: mgorman@suse.de
> CC: Jeff Moyer <jmoyer@redhat.com>
> Signed-off-by: Jan Kara <jack@suse.cz>
> ---
>
> --- a/man2/open.2 2012-04-27 00:07:51.736883092 +0200
> +++ b/man2/open.2 2012-04-27 00:29:59.489892980 +0200
> @@ -769,7 +769,12 @@
> and the file offset must all be multiples of the logical block size
> of the file system.
> Under Linux 2.6, alignment to 512-byte boundaries
> -suffices.
> +suffices. However, if the user buffer is not page aligned and direct read
> +runs in parallel with a
> +.BR fork (2)
> +of the reader process, it may happen that the read data is split between
> +pages owned by the original process and its child. Thus effectively read
> +data is corrupted.
> .LP
> The
> .B O_DIRECT
I think this sufficiently distills the problem. Thanks, Jan.
Acked-by: Jeff Moyer <jmoyer@redhat.com>
next prev parent reply other threads:[~2012-04-30 13:42 UTC|newest]
Thread overview: 31+ messages / expand[flat|nested] mbox.gz Atom feed top
2012-04-30 9:30 [PATCH] Describe race of direct read and fork for unaligned buffers Jan Kara
2012-04-30 13:41 ` Jeff Moyer [this message]
2012-04-30 14:30 ` Mel Gorman
2012-05-01 5:50 ` Michael Kerrisk (man-pages)
2012-05-01 6:49 ` Nick Piggin
2012-05-01 14:31 ` KOSAKI Motohiro
2012-05-01 14:37 ` KOSAKI Motohiro
2012-05-01 15:11 ` Jeff Moyer
2012-05-01 15:34 ` KOSAKI Motohiro
2012-05-01 15:38 ` Jeff Moyer
2012-05-01 15:50 ` Nick Piggin
2012-05-01 23:51 ` Andrea Arcangeli
2012-05-02 8:17 ` Jan Kara
2012-05-02 9:09 ` Nick Piggin
2012-05-02 9:18 ` Jan Kara
2012-05-02 19:14 ` KOSAKI Motohiro
2012-05-02 19:23 ` Jan Kara
2012-05-02 19:25 ` KOSAKI Motohiro
2012-05-05 11:28 ` Michael Kerrisk (man-pages)
2012-05-05 15:29 ` KOSAKI Motohiro
2012-05-08 23:10 ` Nick Piggin
2012-05-09 5:35 ` Michael Kerrisk (man-pages)
2012-05-09 7:01 ` Nick Piggin
2012-05-09 7:18 ` Michael Kerrisk (man-pages)
2012-05-10 15:00 ` Jan Kara
2012-05-01 16:15 ` KOSAKI Motohiro
2012-05-01 17:56 ` Michael Kerrisk (man-pages)
2012-05-02 0:34 ` Nick Piggin
2012-05-02 3:04 ` Hugh Dickins
2012-05-02 3:10 ` Nick Piggin
2012-05-02 9:20 ` Jan Kara
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=x49haw1jq01.fsf@segfault.boston.devel.redhat.com \
--to=jmoyer@redhat.com \
--cc=jack@suse.cz \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-man@vger.kernel.org \
--cc=linux-mm@kvack.org \
--cc=mgorman@suse.de \
--cc=mtk.manpages@gmail.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;
as well as URLs for NNTP newsgroup(s).