From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from eggs.gnu.org ([2001:4830:134:3::10]:37836) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1UupeN-0001mh-DW for qemu-devel@nongnu.org; Thu, 04 Jul 2013 15:57:09 -0400 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71) (envelope-from ) id 1UupXV-0008M1-Tr for qemu-devel@nongnu.org; Thu, 04 Jul 2013 15:50:03 -0400 Received: from mail-ee0-x22a.google.com ([2a00:1450:4013:c00::22a]:53233) by eggs.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1UupXV-0008Ie-Ii for qemu-devel@nongnu.org; Thu, 04 Jul 2013 15:50:01 -0400 Received: by mail-ee0-f42.google.com with SMTP id c4so974195eek.1 for ; Thu, 04 Jul 2013 12:50:00 -0700 (PDT) Sender: Paolo Bonzini Message-ID: <51D5D1DF.4060003@redhat.com> Date: Thu, 04 Jul 2013 21:49:51 +0200 From: Paolo Bonzini MIME-Version: 1.0 References: <1372931597-28115-1-git-send-email-gaowanlong@cn.fujitsu.com> In-Reply-To: <1372931597-28115-1-git-send-email-gaowanlong@cn.fujitsu.com> Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit Subject: Re: [Qemu-devel] [PATCH V4 00/10] Add support for binding guest numa nodes to host numa nodes List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , To: Wanlong Gao Cc: aliguori@us.ibm.com, ehabkost@redhat.com, qemu-devel@nongnu.org, lcapitulino@redhat.com, bsd@redhat.com, y-goto@jp.fujitsu.com, Laszlo Ersek , afaerber@suse.de Il 04/07/2013 11:53, Wanlong Gao ha scritto: > As you know, QEMU can't direct it's memory allocation now, this may cause > guest cross node access performance regression. > And, the worse thing is that if PCI-passthrough is used, > direct-attached-device uses DMA transfer between device and qemu process. > All pages of the guest will be pinned by get_user_pages(). > > KVM_ASSIGN_PCI_DEVICE ioctl > kvm_vm_ioctl_assign_device() > =>kvm_assign_device() > => kvm_iommu_map_memslots() > => kvm_iommu_map_pages() > => kvm_pin_pages() > > So, with direct-attached-device, all guest page's page count will be +1 and > any page migration will not work. AutoNUMA won't too. > > So, we should set the guest nodes memory allocation policy before > the pages are really mapped. > > According to this patch set, we are able to set guest nodes memory policy > like following: > > -numa node,nodeid=0,mem=1024,cpus=0,mem-policy=membind,mem-hostnode=0-1 > -numa node,nodeid=1,mem=1024,cpus=1,mem-policy=interleave,mem-hostnode=1 Did you see my suggestion to use instead something like this: -numa node,nodeid=0,cpus=0 -numa node,nodeid=1,cpus=1 \ -numa mem,nodeid=0,size=1G,policy=membind,hostnode=0-1 -numa mem,nodeid=1,size=2G,policy=interleave,hostnode=1 With an eye to when we'll support memory hotplug, I think it is better. It is not hard to implement it using the OptsVisitor; see 14aa0c2de045a6c2fcfadf38c04434fd15909455 for an example of a complex schema described with OptsVistor. Paolo