From mboxrd@z Thu Jan  1 00:00:00 1970
From: Ian Campbell <Ian.Campbell@citrix.com>
Subject: Re: domU guest for xcp 0.1.1
Date: Thu, 18 Mar 2010 14:32:18 +0000
Message-ID: <1268922738.10129.39708.camel@zakaz.uk.xensource.com>
References: <29b32d341003141329v3c6a73c0y8030cf3c6736634c@mail.gmail.com>
	<29b32d341003150646o4d215dfel1c2a2bc6d287edc@mail.gmail.com>
	<29b32d341003151904m2a992be5s74697d6047b864c6@mail.gmail.com>
	<1268725069.8652.4179.camel@localhost.localdomain>
	<20100316152259.GO1878@reaktio.net>
	<29b32d341003161625w2969f2daib3408561920793d3@mail.gmail.com>
	<29b32d341003171144k628a7741x17b6658731f96e4b@mail.gmail.com>
	<1268903824.10129.38718.camel@zakaz.uk.xensource.com>
	<29b32d341003180643o326f7f43h97c25bd104f3b658@mail.gmail.com>
	<1268921503.10129.39640.camel@zakaz.uk.xensource.com>
	<29b32d341003180723l652adb4bvb3b3aba3d8f3a529@mail.gmail.com>
	<1268922424.10129.39687.camel@zakaz.uk.xensource.com>
Mime-Version: 1.0
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: 7bit
Return-path: <xen-devel-bounces@lists.xensource.com>
In-Reply-To: <1268922424.10129.39687.camel@zakaz.uk.xensource.com>
List-Unsubscribe: <http://lists.xensource.com/mailman/listinfo/xen-devel>,
	<mailto:xen-devel-request@lists.xensource.com?subject=unsubscribe>
List-Post: <mailto:xen-devel@lists.xensource.com>
List-Help: <mailto:xen-devel-request@lists.xensource.com?subject=help>
List-Subscribe: <http://lists.xensource.com/mailman/listinfo/xen-devel>,
	<mailto:xen-devel-request@lists.xensource.com?subject=subscribe>
Sender: xen-devel-bounces@lists.xensource.com
Errors-To: xen-devel-bounces@lists.xensource.com
To: Ritu kaur <ritu.kaur.us@gmail.com>
Cc: "xen-devel@lists.xensource.com" <xen-devel@lists.xensource.com>
List-Id: xen-devel@lists.xenproject.org

On Thu, 2010-03-18 at 14:27 +0000, Ian Campbell wrote:
> On Thu, 2010-03-18 at 14:23 +0000, Ritu kaur wrote:
> > Thanks Ian, could you please clarify me how are interrupts handled
> for
> > pci passthrough device. I thought it is delivered via evtchn from
> > pciback to pcifront i.e
> > InterruptDescTable->Hypervisor->pciback->pcifront->actual_device is
> > this correct?
> 
> Yes, but not on the event channel you found previously, there is a
> separate one for each device IRQ. 

Actually, not quite. pciback sets up an evtchn which pcifront binds to
but pciback then arranges for the hypervisor to directly inject the
interrupt on that event channel rather than bouncing through pciback for
every interrupt so the actual path is more like:

InterruptDescTable->Hypervisor->pcifront->actual_device

And really once the device is setup pcifront isn't really involved
either, the interrupt goes straight from the event channel to the device
driver's interrupt handlerm pciback doesn't do anything for this part.

Ian.