From mboxrd@z Thu Jan  1 00:00:00 1970
Received: from eggs.gnu.org ([208.118.235.92]:50082)
	by lists.gnu.org with esmtp (Exim 4.71)
	(envelope-from <kraxel@redhat.com>) id 1SleCb-00054j-HC
	for qemu-devel@nongnu.org; Mon, 02 Jul 2012 06:50:05 -0400
Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71)
	(envelope-from <kraxel@redhat.com>) id 1SleCW-0006fc-Q1
	for qemu-devel@nongnu.org; Mon, 02 Jul 2012 06:49:57 -0400
Received: from mx1.redhat.com ([209.132.183.28]:39703)
	by eggs.gnu.org with esmtp (Exim 4.71)
	(envelope-from <kraxel@redhat.com>) id 1SleCW-0006fH-Hu
	for qemu-devel@nongnu.org; Mon, 02 Jul 2012 06:49:52 -0400
Message-ID: <4FF17CC4.5010608@redhat.com>
Date: Mon, 02 Jul 2012 12:49:40 +0200
From: Gerd Hoffmann <kraxel@redhat.com>
MIME-Version: 1.0
References: <CAEgOgz56OypbOE+ejwftpbBuL0eooZBcytN4D8p+149FS_aadw@mail.gmail.com>
	<4FE8371B.4020503@suse.de> <4FE874EC.3090307@codemonkey.ws>
In-Reply-To: <4FE874EC.3090307@codemonkey.ws>
Content-Type: text/plain; charset=ISO-8859-1
Content-Transfer-Encoding: 7bit
Subject: Re: [Qemu-devel] [RFC] Attaching EHCI to sysbus.
List-Id: <qemu-devel.nongnu.org>
List-Unsubscribe: <https://lists.nongnu.org/mailman/options/qemu-devel>,
	<mailto:qemu-devel-request@nongnu.org?subject=unsubscribe>
List-Archive: <http://lists.nongnu.org/archive/html/qemu-devel>
List-Post: <mailto:qemu-devel@nongnu.org>
List-Help: <mailto:qemu-devel-request@nongnu.org?subject=help>
List-Subscribe: <https://lists.nongnu.org/mailman/listinfo/qemu-devel>,
	<mailto:qemu-devel-request@nongnu.org?subject=subscribe>
To: Anthony Liguori <anthony@codemonkey.ws>
Cc: Anthony Liguori <aliguori@us.ibm.com>, qemu-devel <qemu-devel@nongnu.org>, Peter Crosthwaite <peter.crosthwaite@petalogix.com>, "Edgar E. Iglesias" <edgar.iglesias@gmail.com>, Paolo Bonzini <pbonzini@redhat.com>, =?ISO-8859-1?Q?Andreas_F=E4rber?= <afaerber@suse.de>, John Williams <john.williams@petalogix.com>, David Gibson <david@gibson.dropbear.id.au>

  Hi,

> Does it make sense to have an EHCI bus type that inherits from USBBus?
> 
> That way we could change USBPortOps into methods of the USBBus that the
> subclass overrides.

I don't think this is useful.  USBPortOps should be identical for both
cases.

> That would strongly decouple the EHCI code from the PCI device.  Then
> the ehci-pci device just needs to setup the EHCI bus and forward MMIO
> requests appropriately.

I think what we need to do is:

  (1) Create EHCIPCIState, which holds just PCIDevice and EHCIState.
  (2) Setup dma context in pci init function, then switch over all
      memory access from pci_* to dma_* (simliar to ohci).

Generic ehci code should not have any pci references any more then.  Now:

  (3) create EHCISysbusState, hook up sysbus init function which does
      mmio registration, dma context setup and irq windup the sysbus
      way.

cheers,
  Gerd