From mboxrd@z Thu Jan 1 00:00:00 1970 From: Avi Kivity Subject: Re: [PATCH v2] Driver for Inter-VM shared memory device for KVM supporting interrupts. Date: Wed, 20 May 2009 12:11:57 +0300 Message-ID: <4A13C95D.1020402@redhat.com> References: <1241713567-17256-1-git-send-email-cam@cs.ualberta.ca> <200905200933.01736.borntraeger@de.ibm.com> <4A13C33E.3090705@redhat.com> <200905201107.00930.borntraeger@de.ibm.com> Mime-Version: 1.0 Content-Type: text/plain; charset=UTF-8; format=flowed Content-Transfer-Encoding: QUOTED-PRINTABLE Cc: Rusty Russell , Cam Macdonell , kvm@vger.kernel.org, Christian Ehrhardt , Anthony Liguori To: =?UTF-8?B?Q2hyaXN0aWFuIEJvcm50csOkZ2Vy?= Return-path: Received: from mx2.redhat.com ([66.187.237.31]:54389 "EHLO mx2.redhat.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1754058AbZETJMF (ORCPT ); Wed, 20 May 2009 05:12:05 -0400 In-Reply-To: <200905201107.00930.borntraeger@de.ibm.com> Sender: kvm-owner@vger.kernel.org List-ID: Christian Borntr=C3=A4ger wrote: > Am Mittwoch 20 Mai 2009 10:45:50 schrieb Avi Kivity: > =20 >> Christian Borntr=C3=A4ger wrote: >> =20 >>> o shared guest kernels: The CMS operating system is build as a boot= able >>> DCSS (called named-saved-segments NSS). All guests have the same ho= st >>> pages for the read-only parts of the CMS kernel. The local data is = stored >>> in exclusive-write parts of the same NSS. Linux on System z is also >>> capable of using this feature (CONFIG_SHARED_KERNEL). The kernel li= nkage >>> is changed in a way to separate the read-only text segment from the= other >>> parts with segment size alignment >>> =20 >> How does patching (smp, kprobes/jprobes, markers/ftrace) work with t= his? >> =20 > It does not. :-)=20 > Because of that and since most distro kernels are fully modular and k= ernel=20 > updates are another problem this feature is not used very often for L= inux. It=20 > is used heavily in CMS, though. > Actually, we could do COW in the host but then it is really not worth= the=20 > effort. > =20 ksm on low throttle would solve all of those problems. > Yes, KSM is easier and it even finds duplicate data pages. > On the other hand it does only provide memory saving. It does not spe= edup=20 > application startup like execute-in-place (major page faults become m= inor page=20 > faults for text pages if the page is already backed by the host) > I am not claiming that KSM is useless. Depending on the scenario you = might=20 > want the one or the other or even both. For typical desktop use, KSM = is very=20 > likely the better approach If ksm shares pagecache, then doesn't it become effectively XIP? We could also hook virtio dma to preemptively share pages somehow. --=20 error compiling committee.c: too many arguments to function