From mboxrd@z Thu Jan  1 00:00:00 1970
From: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
Subject: Re: Noticeably poor Intel GPU performance on 3.3 and
 3.4 dom0 kernels
Date: Wed, 30 May 2012 17:21:45 -0400
Message-ID: <20120530212145.GA30035@phenom.dumpdata.com>
References: <4FC5F2EF.7000004@invisiblethingslab.com>
	<20120530142830.GE3207@phenom.dumpdata.com>
	<4FC6353C.3090908@invisiblethingslab.com>
Mime-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Content-Transfer-Encoding: 7bit
Return-path: <xen-devel-bounces@lists.xen.org>
Content-Disposition: inline
In-Reply-To: <4FC6353C.3090908@invisiblethingslab.com>
List-Unsubscribe: <http://lists.xen.org/cgi-bin/mailman/options/xen-devel>,
	<mailto:xen-devel-request@lists.xen.org?subject=unsubscribe>
List-Post: <mailto:xen-devel@lists.xen.org>
List-Help: <mailto:xen-devel-request@lists.xen.org?subject=help>
List-Subscribe: <http://lists.xen.org/cgi-bin/mailman/listinfo/xen-devel>,
	<mailto:xen-devel-request@lists.xen.org?subject=subscribe>
Sender: xen-devel-bounces@lists.xen.org
Errors-To: xen-devel-bounces@lists.xen.org
To: Marek Marczykowski <marmarek@invisiblethingslab.com>
Cc: "xen-devel@lists.xensource.com" <xen-devel@lists.xensource.com>, Joanna Rutkowska <joanna@invisiblethingslab.com>
List-Id: xen-devel@lists.xenproject.org

On Wed, May 30, 2012 at 04:57:00PM +0200, Marek Marczykowski wrote:
> On 30.05.2012 16:28, Konrad Rzeszutek Wilk wrote:
> > On Wed, May 30, 2012 at 12:14:07PM +0200, Joanna Rutkowska wrote:
> >> Hello,
> >>
> >> I've recently been some newer Dom0 kernels (3.3.5, 3.4) under Xen 4.1.2,
> >> and have noticed that those kernel noticeably downgrade performance of
> >> (at least) Intel GPUs. This is easily noticeable even on such trivial
> >> tasks as scrolling a webpage in Firefox or Chrome. This slow GPU
> >> performance on 3.3 and 3.4 kernels is in stark contrast with what I see
> >> on 3.2.7 Dom0 kernel, where graphics works just great...
> >>
> >> In order to make sure that this is not caused by power management set
> >> too strictly, I played with xenpm and made sure to set the following:
> > 
> > And are you building with CONFIG_INTEL_IOMMU=y?
> 
> Yes:
> CONFIG_GART_IOMMU=y
> # CONFIG_CALGARY_IOMMU is not set
> CONFIG_SWIOTLB=y
> CONFIG_IOMMU_HELPER=y
> (...)
> CONFIG_IOMMU_API=y
> CONFIG_IOMMU_SUPPORT=y
> CONFIG_AMD_IOMMU=y
> # CONFIG_AMD_IOMMU_STATS is not set
> CONFIG_AMD_IOMMU_V2=m
> CONFIG_DMAR_TABLE=y
> CONFIG_INTEL_IOMMU=y
> CONFIG_INTEL_IOMMU_DEFAULT_ON=y
> CONFIG_INTEL_IOMMU_FLOPPY_WA=y
> CONFIG_IRQ_REMAP=y

In that case I think you might need to do some git bisection to figure
out which patch caused the slow-down. Are there any obvious things in the
kernel logs (you might have to boot it with loglevel=8 debug)?