qemu-devel.nongnu.org archive mirror
 help / color / mirror / Atom feed
From: Anthony Liguori <anthony@codemonkey.ws>
To: "Michael S. Tsirkin" <mst@redhat.com>
Cc: qemu-devel@nongnu.org
Subject: [Qemu-devel] Re: [PATCH] net: delay peer host device delete
Date: Mon, 20 Sep 2010 15:20:59 -0500	[thread overview]
Message-ID: <4C97C22B.3010502@codemonkey.ws> (raw)
In-Reply-To: <20100920194415.GK30611@redhat.com>

On 09/20/2010 02:44 PM, Michael S. Tsirkin wrote:
>
>> I think the only workable approach that doesn't involve new commands
>> is to change the semantics of the existing ones.
>>
>> Make netdev_del work regardless of whether the device is still present.
>>
>> You would need to reference count the actual netdev structure and
>> have each device using it unref on delete.  You make netdev_del mark
>> the device as deleted and when a device is deleted, any calls into
>> the device effectively become nops.
>>
>> You have to go through most of the cleanup process to ensure that
>> tap device gets closed even before your reference count goes to
>> zero.
>>      
> I think you mean 'does not get closed': we need the fd to get the flags etc.
>    

No, I actually meant does get closed.

When you do netdev_del, it should result in the fd getting closed.

The actual netdev structure then becomes a zombie that's completely 
useless until the device goes away.

> Note that it will mostly work unless when it'll crash.
> Issue is we don't have any documentation so
> people get the command set by trial and error.
>
> So how can we prove it's a user bug and not qemu bug?
> I guess we should blame ourselves until proven innocent.
>    

Here's what I'm now suggesting:

device_del -> may or may not unplug a device from a guest when it 
returns.  To figure out if it does, you have to run info qdm.

netdev_del -> always destroys a netdev device when it returns.  May be 
called at any point in time.  If you destroy a netdev while the device 
is still using it, all packets go into the bit bucket and the link 
status is modified to be unplugged.

You're suggesting:

netdev_del -> may or may not destroy a netdev depending on when the 
device delete completes.  Eventually, when there's a reset, we will kill 
the device.  Even though the netdev is still active, we'll hide it from 
the management tools.

I think the suggested semantics are totally unusable.  If we can make 
something deterministic for a management tool, that should be the path 
we take.

Regards,

Anthony Liguori

  reply	other threads:[~2010-09-20 20:23 UTC|newest]

Thread overview: 27+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2010-09-20 16:30 [Qemu-devel] [PATCH] net: delay peer host device delete Michael S. Tsirkin
2010-09-20 16:41 ` Anthony Liguori
2010-09-20 16:47   ` Michael S. Tsirkin
2010-09-20 16:56     ` Anthony Liguori
2010-09-20 17:14       ` Michael S. Tsirkin
2010-09-20 18:14         ` Anthony Liguori
2010-09-20 18:19           ` Anthony Liguori
2010-09-20 18:59             ` [Qemu-devel] " Michael S. Tsirkin
2010-09-20 19:22               ` Anthony Liguori
2010-09-20 19:37                 ` Michael S. Tsirkin
2010-09-20 20:15                   ` Anthony Liguori
2010-09-20 20:15                     ` Michael S. Tsirkin
2010-09-21  8:58                   ` Daniel P. Berrange
2010-09-21  9:20                     ` Michael S. Tsirkin
2010-09-21 12:47                       ` Anthony Liguori
2010-09-20 18:24           ` Michael S. Tsirkin
2010-09-20 18:39             ` Anthony Liguori
2010-09-20 19:15               ` Michael S. Tsirkin
2010-09-20 19:28                 ` Anthony Liguori
2010-09-20 19:44                   ` Michael S. Tsirkin
2010-09-20 20:20                     ` Anthony Liguori [this message]
2010-09-20 20:27                       ` Michael S. Tsirkin
2010-09-20 20:38                         ` Anthony Liguori
2010-09-20 20:37                           ` Michael S. Tsirkin
2010-09-20 20:50                             ` Anthony Liguori
2010-09-21  9:18                               ` Michael S. Tsirkin
2010-09-21 12:42                                 ` Anthony Liguori

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=4C97C22B.3010502@codemonkey.ws \
    --to=anthony@codemonkey.ws \
    --cc=mst@redhat.com \
    --cc=qemu-devel@nongnu.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).