qemu-devel.nongnu.org archive mirror
 help / color / mirror / Atom feed
From: Corey Bryant <coreyb@linux.vnet.ibm.com>
To: Eric Blake <eblake@redhat.com>
Cc: kwolf@redhat.com, aliguori@us.ibm.com,
	stefanha@linux.vnet.ibm.com, libvir-list@redhat.com,
	qemu-devel@nongnu.org, lcapitulino@redhat.com
Subject: Re: [Qemu-devel] [PATCH v5 6/6] block: Enable qemu_open/close to work with fd sets
Date: Thu, 02 Aug 2012 11:08:24 -0400	[thread overview]
Message-ID: <501A97E8.9000705@linux.vnet.ibm.com> (raw)
In-Reply-To: <5010C031.9040602@linux.vnet.ibm.com>



On 07/25/2012 11:57 PM, Corey Bryant wrote:
>
>
> On 07/25/2012 03:43 PM, Eric Blake wrote:
>> On 07/23/2012 07:08 AM, Corey Bryant wrote:
>>> When qemu_open is passed a filename of the "/dev/fdset/nnn"
>>> format (where nnn is the fdset ID), an fd with matching access
>>> mode flags will be searched for within the specified monitor
>>> fd set.  If the fd is found, a dup of the fd will be returned
>>> from qemu_open.
>>>
>>> Each fd set has a reference count.  The purpose of the reference
>>> count is to determine if an fd set contains file descriptors that
>>> have open dup() references that have not yet been closed.  It is
>>> incremented on qemu_open and decremented on qemu_close.  It is
>>> not until the refcount is zero that file desriptors in an fd set
>>> can be closed.  If an fd set has dup() references open, then we
>>> must keep the other fds in the fd set open in case a reopen
>>> of the file occurs that requires an fd with a different access
>>> mode.
>>>
>>
>>> +++ b/monitor.c
>>> @@ -2551,6 +2551,91 @@ static void monitor_fdsets_set_in_use(Monitor
>>> *mon, bool in_use)
>>>       }
>>>   }
>>>
>>> +void monitor_fdset_increment_refcount(Monitor *mon, int64_t fdset_id)
>>> +{
>>> +    mon_fdset_t *mon_fdset;
>>> +
>>> +    if (!mon) {
>>> +        return;
>>> +    }
>>
>> Am I reading this code right by stating that 'if there is no monitor, we
>> don't increment the refcount'?  How does a monitor reattach affect
>> things?  Or am I missing something fundamental about the cases when
>> 'mon==NULL' will exist?
>>
>
> Yes you're reading this correctly.
>
> I'm pretty sure that mon will only be NULL if QEMU is started without a
> monitor.
>
> If QEMU has a monitor, and libvirt disconnects it's connection to the
> qemu monitor, then I believe mon will remain non-NULL.

I've verified this to be true and everything is working as expected.

If libvirt's connection to the monitor fd is closed, mon will remain 
non-NULL and the refcount will still be incremented/decremented on 
qemu_open()/qemu_close().  When libvirt reconnects, any fdsets that 
haven't been cleaned up will still be available.  query-fdsets can be 
used to determine what's available.

-- 
Regards,
Corey

  parent reply	other threads:[~2012-08-02 16:01 UTC|newest]

Thread overview: 36+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2012-07-23 13:07 [Qemu-devel] [PATCH v5 0/6] file descriptor passing using fd sets Corey Bryant
2012-07-23 13:08 ` [Qemu-devel] [PATCH v5 1/6] qemu-char: Add MSG_CMSG_CLOEXEC flag to recvmsg Corey Bryant
2012-07-23 22:50   ` Eric Blake
2012-07-24  2:19     ` Corey Bryant
2012-07-23 13:08 ` [Qemu-devel] [PATCH v5 2/6] qapi: Introduce add-fd, remove-fd, query-fdsets Corey Bryant
2012-07-25 18:16   ` Eric Blake
2012-07-26  2:55     ` Corey Bryant
2012-07-23 13:08 ` [Qemu-devel] [PATCH v5 3/6] monitor: Clean up fd sets on monitor disconnect Corey Bryant
2012-07-23 13:08 ` [Qemu-devel] [PATCH v5 4/6] block: Convert open calls to qemu_open Corey Bryant
2012-07-25 19:22   ` Eric Blake
2012-07-26  3:11     ` Corey Bryant
2012-07-23 13:08 ` [Qemu-devel] [PATCH v5 5/6] block: Convert close calls to qemu_close Corey Bryant
2012-07-23 13:08 ` [Qemu-devel] [PATCH v5 6/6] block: Enable qemu_open/close to work with fd sets Corey Bryant
2012-07-23 13:14   ` Corey Bryant
2012-08-02 22:21     ` Corey Bryant
2012-08-06  9:15       ` Kevin Wolf
2012-08-06 13:32         ` Corey Bryant
2012-08-06 13:51           ` Kevin Wolf
2012-08-06 14:15             ` Corey Bryant
2012-08-07 16:43               ` Corey Bryant
2012-07-24 12:07   ` Kevin Wolf
2012-07-25  3:41     ` Corey Bryant
2012-07-25  8:22       ` Kevin Wolf
2012-07-25 19:25         ` Eric Blake
2012-07-26  3:21           ` Corey Bryant
2012-07-26 13:13             ` Eric Blake
2012-07-26 13:16               ` Kevin Wolf
2012-07-27  4:07                 ` Corey Bryant
2012-07-25 19:43   ` Eric Blake
2012-07-26  3:57     ` Corey Bryant
2012-07-26  9:07       ` Kevin Wolf
2012-07-27  3:59         ` Corey Bryant
2012-07-27  4:03         ` Corey Bryant
2012-08-02 15:08       ` Corey Bryant [this message]
2012-07-24 12:09 ` [Qemu-devel] [PATCH v5 0/6] file descriptor passing using " Kevin Wolf
2012-07-25  3:42   ` Corey Bryant

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=501A97E8.9000705@linux.vnet.ibm.com \
    --to=coreyb@linux.vnet.ibm.com \
    --cc=aliguori@us.ibm.com \
    --cc=eblake@redhat.com \
    --cc=kwolf@redhat.com \
    --cc=lcapitulino@redhat.com \
    --cc=libvir-list@redhat.com \
    --cc=qemu-devel@nongnu.org \
    --cc=stefanha@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).