From: Paolo Bonzini <pbonzini@redhat.com>
To: Stefan Hajnoczi <stefanha@gmail.com>
Cc: Kevin Wolf <kwolf@redhat.com>,
Supriya Kannery <supriyak@linux.vnet.ibm.com>,
Anthony Liguori <aliguori@us.ibm.com>,
qemu-devel <qemu-devel@nongnu.org>
Subject: Re: [Qemu-devel] Safely reopening image files by stashing fds
Date: Fri, 05 Aug 2011 11:55:24 +0200 [thread overview]
Message-ID: <4E3BBE0C.4030109@redhat.com> (raw)
In-Reply-To: <CAJSP0QXJpgtE+JjfcQoSkCtTndT9Uma9ZFte6T-mwvCM8ojQRw@mail.gmail.com>
On 08/05/2011 11:27 AM, Stefan Hajnoczi wrote:
>>> First call bdrv_extract_fds() to stash the file descriptors, then close
>>> the block device. Try opening the new image but if that fails, reopen using
>>> the stashed file descriptors.
>>
>> Why not do the latter unconditionally?
>
> Because you cannot change O_DIRECT on an open fd:(. This is why
> we're going through this pain.
>
> The only method I've found that works is to open("/proc/self/fd/X",
> new_flags) but that's non-portable.
Maybe I'm missing something obvious, but so is O_DIRECT, no? :)
So for Linux you can dup the stashed file descriptors using
/proc/self/fd and change flags directly, and for other OSes you can dup
them using dup2 and change flags with F_SETFL. In any case, reopening
can always be done using the stashed descriptors (or BlockDriverStates).
Paolo
next prev parent reply other threads:[~2011-08-05 9:55 UTC|newest]
Thread overview: 39+ messages / expand[flat|nested] mbox.gz Atom feed top
2011-08-05 8:40 [Qemu-devel] Safely reopening image files by stashing fds Stefan Hajnoczi
2011-08-05 9:04 ` Paolo Bonzini
2011-08-05 9:27 ` Stefan Hajnoczi
2011-08-05 9:55 ` Paolo Bonzini [this message]
2011-08-05 13:03 ` Stefan Hajnoczi
2011-08-05 13:12 ` Daniel P. Berrange
2011-08-05 14:28 ` Christoph Hellwig
2011-08-05 15:24 ` Stefan Hajnoczi
2011-08-05 15:43 ` Kevin Wolf
2011-08-05 15:49 ` Anthony Liguori
2011-08-08 7:02 ` Supriya Kannery
2011-08-08 8:12 ` Kevin Wolf
2011-08-09 9:22 ` supriya kannery
2011-08-09 9:51 ` Kevin Wolf
2011-08-09 9:32 ` supriya kannery
2011-08-16 19:18 ` [Qemu-devel] [RFC] " Supriya Kannery
2011-08-16 19:18 ` Supriya Kannery
2011-08-17 14:35 ` Kevin Wolf
2011-10-10 18:28 ` [Qemu-devel] " Kevin Wolf
2011-10-11 5:21 ` Supriya Kannery
2011-08-05 14:27 ` Christoph Hellwig
2011-08-05 9:07 ` Kevin Wolf
2011-08-05 9:29 ` Stefan Hajnoczi
2011-08-05 9:48 ` Kevin Wolf
2011-08-08 14:49 ` Stefan Hajnoczi
2011-08-08 15:16 ` Kevin Wolf
2011-08-09 10:25 ` Stefan Hajnoczi
2011-08-09 10:35 ` Kevin Wolf
2011-08-09 10:50 ` Stefan Hajnoczi
2011-08-09 10:56 ` Stefan Hajnoczi
2011-08-09 11:39 ` Kevin Wolf
2011-08-09 12:00 ` Stefan Hajnoczi
2011-08-09 12:24 ` Kevin Wolf
2011-08-09 19:39 ` Blue Swirl
2011-08-10 7:58 ` Kevin Wolf
2011-08-10 17:20 ` Blue Swirl
2011-08-11 7:37 ` Kevin Wolf
2011-08-11 16:21 ` Blue Swirl
2011-08-05 20:16 ` Blue Swirl
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=4E3BBE0C.4030109@redhat.com \
--to=pbonzini@redhat.com \
--cc=aliguori@us.ibm.com \
--cc=kwolf@redhat.com \
--cc=qemu-devel@nongnu.org \
--cc=stefanha@gmail.com \
--cc=supriyak@linux.vnet.ibm.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).