From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from eggs.gnu.org ([2001:4830:134:3::10]:43232) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1aJurk-0007WE-AY for qemu-devel@nongnu.org; Thu, 14 Jan 2016 22:15:57 -0500 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71) (envelope-from ) id 1aJurg-0006sq-UX for qemu-devel@nongnu.org; Thu, 14 Jan 2016 22:15:56 -0500 Received: from mx1.redhat.com ([209.132.183.28]:46753) by eggs.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1aJurg-0006sk-Od for qemu-devel@nongnu.org; Thu, 14 Jan 2016 22:15:52 -0500 References: <1452750444-17750-1-git-send-email-jasowang@redhat.com> <20160114112703-mutt-send-email-mst@redhat.com> From: Jason Wang Message-ID: <5698645C.2070504@redhat.com> Date: Fri, 15 Jan 2016 11:15:40 +0800 MIME-Version: 1.0 In-Reply-To: <20160114112703-mutt-send-email-mst@redhat.com> Content-Type: text/plain; charset=windows-1252 Content-Transfer-Encoding: 7bit Subject: Re: [Qemu-devel] [PATCH] intel_iommu: large page support List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , To: "Michael S. Tsirkin" Cc: qemu-devel@nongnu.org, pbonzini@redhat.com, ehabkost@redhat.com, rth@twiddle.net On 01/14/2016 05:28 PM, Michael S. Tsirkin wrote: > On Thu, Jan 14, 2016 at 12:47:24AM -0500, Jason Wang wrote: >> > Current intel_iommu only supports 4K page which may not be sufficient >> > to cover guest working set. This patch tries to enable 2M and 1G mapping >> > for intel_iommu. This is also useful for future device IOTLB >> > implementation to have a better hit rate. >> > >> > Major work is adding a page mask field on IOTLB entry to make it >> > support large page. And also use the slpte level as key to do IOTLB >> > lookup. MAMV was increased to 18 to support direct invalidation for 1G >> > mapping. >> > >> > Cc: Michael S. Tsirkin >> > Cc: Paolo Bonzini >> > Cc: Richard Henderson >> > Cc: Eduardo Habkost >> > Signed-off-by: Jason Wang > Looks good, thanks! > > I was going to comment that changes such as MAMV would > have to be versioned, when I noticed that this device > is unmigrateable ATM. > > So no issue, but we do need to fix migration for it. > Yes, it was in my TODO list. Will first see if there're more caps that needs to be added. (Otherwise, we may have a long property list for iommu).