From: Dexuan Cui <decui@microsoft.com>
To: KY Srinivasan <kys@microsoft.com>,
        "gregkh@linuxfoundation.org" <gregkh@linuxfoundation.org>,
        "linux-kernel@vger.kernel.org" <linux-kernel@vger.kernel.org>,
        "devel@linuxdriverproject.org" <devel@linuxdriverproject.org>,
        "olaf@aepfle.de" <olaf@aepfle.de>,
        "apw@canonical.com" <apw@canonical.com>,
        "vkuznets@redhat.com" <vkuznets@redhat.com>
Subject: RE: [PATCH 0/6] Drivers: hv: vmbus 
Thread-Topic: [PATCH 0/6] Drivers: hv: vmbus 
Thread-Index: AQHQSVEy+lX2DyRKVkO+0QvUCJOaj5zylQAAgAArCgCAAAUxAA==
Date: Mon, 16 Feb 2015 09:01:07 +0000
Message-ID: <F792CF86EFE20D4AB8064279AFBA51C61F34F56D@HKNPRD3002MB017.064d.mgd.msft.net>
References: <1424031079-21708-1-git-send-email-kys@microsoft.com>
 <F792CF86EFE20D4AB8064279AFBA51C61F34F0AD@HKNPRD3002MB017.064d.mgd.msft.net>
 <BY2PR0301MB0711FEC2B58456BF4B0DAC80A02E0@BY2PR0301MB0711.namprd03.prod.outlook.com>
In-Reply-To: <BY2PR0301MB0711FEC2B58456BF4B0DAC80A02E0@BY2PR0301MB0711.namprd03.prod.outlook.com>
Accept-Language: en-US
Content-Language: en-US
Content-Type: text/plain; charset="us-ascii"
Content-Transfer-Encoding: 8BIT
MIME-Version: 1.0
X-MS-Exchange-CrossTenant-OriginalArrivalTime: 16 Feb 2015 09:01:10.4902
 (UTC)
X-MS-Exchange-CrossTenant-Id: 72f988bf-86f1-41af-91ab-2d7cd011db47
X-MS-Exchange-CrossTenant-OriginalAttributedTenantConnectingIp: TenantId=72f988bf-86f1-41af-91ab-2d7cd011db47;Ip=[206.191.230.4]
X-MS-Exchange-CrossTenant-FromEntityHeader: HybridOnPrem
X-MS-Exchange-Transport-CrossTenantHeadersStamped: BL2PR03MB388
Sender: linux-kernel-owner@vger.kernel.org
List-ID: <linux-kernel.vger.kernel.org>
X-Mailing-List: linux-kernel@vger.kernel.org
Content-Length: 3748
Lines: 102

> -----Original Message-----
> From: KY Srinivasan
> Sent: Monday, February 16, 2015 13:28 PM
> To: Dexuan Cui; gregkh@linuxfoundation.org; linux-kernel@vger.kernel.org;
> devel@linuxdriverproject.org; olaf@aepfle.de; apw@canonical.com;
> vkuznets@redhat.com
> Subject: RE: [PATCH 0/6] Drivers: hv: vmbus
> > -----Original Message-----
> > From: Dexuan Cui
> > Sent: Sunday, February 15, 2015 7:19 PM
> > To: KY Srinivasan; gregkh@linuxfoundation.org; linux-
> > kernel@vger.kernel.org; devel@linuxdriverproject.org; olaf@aepfle.de;
> > apw@canonical.com; vkuznets@redhat.com
> > Subject: RE: [PATCH 0/6] Drivers: hv: vmbus
> >
> > > -----Original Message-----
> > > From: devel [mailto:driverdev-devel-bounces@linuxdriverproject.org] On
> > > Behalf Of K. Y. Srinivasan
> > > Sent: Monday, February 16, 2015 4:11 AM
> > > To: gregkh@linuxfoundation.org; linux-kernel@vger.kernel.org;
> > > devel@linuxdriverproject.org; olaf@aepfle.de; apw@canonical.com;
> > > vkuznets@redhat.com
> > > Subject: [PATCH 0/6] Drivers: hv: vmbus
> > >
> > > The host can rescind an offer any time after the offer has been made
> > > to the guest. This patch-set cleans up how we handle rescind messages
> > > from the host.
> > >
> > >
> > > K. Y. Srinivasan (6):
> > >   Drivers: hv: vmbus: Properly handle child device remove
> > >   Drivers: hv: vmbus: Introduce a function to remove a rescinded offer
> > >   Drivers: hv: vmbus: Handle both rescind and offer messages in the
> > >     same context
> > >   Drivers: hv: vmbus: Remove the channel from the channel list(s) on
> > >     failure
> > >   Drivers: hv: util: On device remove, close the channel after
> > >     de-initializing the service
> > >   Drivers: hv: vmbus: Get rid of some unnecessary messages
> > >
> > >  drivers/hv/channel.c      |    9 ++++
> > >  drivers/hv/channel_mgmt.c |   95 ++++++++++++++++++++-----------------
> > -------
> > >  drivers/hv/connection.c   |    7 +---
> > >  drivers/hv/hv_util.c      |    2 +-
> > >  drivers/hv/vmbus_drv.c    |   26 +++++++++---
> > >  include/linux/hyperv.h    |    1 +
> > >  6 files changed, 74 insertions(+), 66 deletions(-)
> > >
> > > --
> >
> > The patchset seems good to me.
> > Reviewed-by: Dexuan Cui <decui@microsoft.com>
> 
> Dexuan,
> 
> Thank you for the review.
> >
> > BTW, IMO we need one more patch to remove the queue_work() in
> > free_channel() -- just make it an ordinary synchronous function call:
> >
> > vmbus_process_offer() can invoke free_channel() on error path, and
> > vmbus_process_rescind() can invoke free_channel() too.
> > We should exclude the possible race.
> 
> 
> I don't see the race; free_channel is only called after ensuring the channel
KY, You're correct.
Sorry for my misreading.

> cannot be discovered
> by any other context. Note that we are now executing both rescind and the
> offer message in the
> same work context. With this patch-set, there are only 3 call sites for
> free_channel:
> 1.  hv_process_channel_removal()
> 2. vmbus_free_channels()
> 3. vmbus_process_offer()
> 
> If vmbus_process_offer() calls free_channel, the channel cannot be discovered
> via its ID as we remove
> The chanel from the global as well as the per-cpu lists. In this case, the channel
> is destroyed and nobody can get a reference to it.
Yeah, I got this now.

-- Dexuan

> >
> > And now the controlwq and work fields of struct vmbus_channel are useless
> > now.
> 
> Yes; we can get rid of this now. I will have that in a separate patch.
> 
> Regards,
> 
> K. Y

--
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/