From mboxrd@z Thu Jan  1 00:00:00 1970
Received: from eggs.gnu.org ([2001:4830:134:3::10]:49309)
	by lists.gnu.org with esmtp (Exim 4.71)
	(envelope-from <peterx@redhat.com>) id 1auWrY-0007PU-Cj
	for qemu-devel@nongnu.org; Sun, 24 Apr 2016 23:07:05 -0400
Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71)
	(envelope-from <peterx@redhat.com>) id 1auWrV-0007V5-6Z
	for qemu-devel@nongnu.org; Sun, 24 Apr 2016 23:07:04 -0400
Received: from mx1.redhat.com ([209.132.183.28]:51813)
	by eggs.gnu.org with esmtp (Exim 4.71)
	(envelope-from <peterx@redhat.com>) id 1auWrV-0007UZ-0W
	for qemu-devel@nongnu.org; Sun, 24 Apr 2016 23:07:01 -0400
Date: Mon, 25 Apr 2016 11:06:55 +0800
From: Peter Xu <peterx@redhat.com>
Message-ID: <20160425030655.GE3261@pxdev.xzpeter.org>
References: <1461535977-331-1-git-send-email-davidkiarie4@gmail.com>
	<1461535977-331-5-git-send-email-davidkiarie4@gmail.com>
MIME-Version: 1.0
Content-Type: text/plain; charset=utf-8
Content-Disposition: inline
In-Reply-To: <1461535977-331-5-git-send-email-davidkiarie4@gmail.com>
Subject: Re: [Qemu-devel] [V9 4/4] hw/pci-host: Emulate AMD IOMMU
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: David Kiarie <davidkiarie4@gmail.com>
Cc: qemu-devel@nongnu.org, jan.kizska@web.de, mst@redhat.com, valentine.sinitsyn@gmail.com, marcel@redhat.com, imammedo@redhat.com

On Mon, Apr 25, 2016 at 01:12:57AM +0300, David Kiarie wrote:
> Add AMD IOMMU emulation support to q35 chipset
> 
> Signed-off-by: David Kiarie <davidkiarie4@gmail.com>
> ---
>  hw/pci-host/q35.c             | 21 +++++++++++++++++++--
>  include/hw/i386/intel_iommu.h |  2 +-
>  2 files changed, 20 insertions(+), 3 deletions(-)
> 
> diff --git a/hw/pci-host/q35.c b/hw/pci-host/q35.c
> index 70f897e..37f8a84 100644
> --- a/hw/pci-host/q35.c
> +++ b/hw/pci-host/q35.c
> @@ -32,6 +32,7 @@
>  #include "hw/pci-host/q35.h"
>  #include "qapi/error.h"
>  #include "qapi/visitor.h"
> +#include "hw/i386/amd_iommu.h"
>  
>  /****************************************************************************
>   * Q35 host
> @@ -448,6 +449,19 @@ static void mch_init_dmar(MCHPCIState *mch)
>      pci_setup_iommu(pci_bus, q35_host_dma_iommu, mch->iommu);
>  }
>  
> +static void mch_init_amdvi(MCHPCIState *mch)
> +{
> +    AMDIOMMUState *iommu_state;
> +    PCIBus *bus = PCI_BUS(qdev_get_parent_bus(DEVICE(mch)));
> +    PCIDevice *iommu;
> +
> +    iommu = pci_create_simple(bus, 0x20, TYPE_AMD_IOMMU_DEVICE);
> +
> +    iommu_state = AMD_IOMMU_DEVICE(iommu);
> +
> +    pci_setup_iommu(bus, bridge_host_amd_iommu, iommu_state);
> +}
> +
>  static void mch_realize(PCIDevice *d, Error **errp)
>  {
>      int i;
> @@ -506,8 +520,11 @@ static void mch_realize(PCIDevice *d, Error **errp)
>                   mch->pci_address_space, &mch->pam_regions[i+1],
>                   PAM_EXPAN_BASE + i * PAM_EXPAN_SIZE, PAM_EXPAN_SIZE);
>      }
> -    /* Intel IOMMU (VT-d) */
> -    if (object_property_get_bool(qdev_get_machine(), "iommu", NULL)) {
> +
> +    if (object_property_get_bool(qdev_get_machine(), "iommu", NULL) &&
> +        MACHINE(qdev_get_machine())->x_iommu_type) {
> +        mch_init_amdvi(mch);
> +    } else if (object_property_get_bool(qdev_get_machine(), "iommu", NULL)) {
>          mch_init_dmar(mch);
>      }

Like patch 3's comment, I would prefer this:

    if (object_property_get_bool(qdev_get_machine(), "iommu", NULL)) {
        if (MACHINE(qdev_get_machine())->x_iommu_type) {
            mch_init_amdvi(mch);
        } else {
            mch_init_dmar(mch);
        }
    }

Thanks.

-- peterx