From mboxrd@z Thu Jan 1 00:00:00 1970 From: =?windows-1252?Q?Roger_Pau_Monn=E9?= Subject: Re: [PATCH v8 00/15] Migration v2 (libxc) Date: Wed, 8 Apr 2015 17:44:28 +0200 Message-ID: <55254CDC.9070504@citrix.com> References: <1427999563-22132-1-git-send-email-andrew.cooper3@citrix.com> Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit Return-path: In-Reply-To: <1427999563-22132-1-git-send-email-andrew.cooper3@citrix.com> List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Sender: xen-devel-bounces@lists.xen.org Errors-To: xen-devel-bounces@lists.xen.org To: Andrew Cooper , Xen-devel Cc: Wei Liu , Ian Campbell , Wen Congyang , Tim Deegan , Ian Jackson , Ross Lagerwall , David Vrabel , Shriram Rajagopalan , Hongyang Yang , Vijay Kilari List-Id: xen-devel@lists.xenproject.org El 02/04/15 a les 20.32, Andrew Cooper ha escrit: > Presented here is v8 of the Migration v2 series (libxc subset), which is able > to function when transparently inserted under an unmodified xl/libxl. > > There are numerous tweaks in v8 (fix build with MiniOS, tweak some code to > appease Coverity, style fixes, clarification of some error messages) but no > major changes from the previous version. It is my belief that the libxc > changes are functionally complete. > > This series can be found on a 'saverestore2-v8' at > http://xenbits.xen.org/git-http/people/andrewcoop/xen.git > > To experiment, simply set XG_MIGRATION_V2 in xl's environment. For migration, > the easiest way is to tweak libxl-save-helper to be a shell script > > root@vitruvias:/home# cat /usr/lib/xen/bin/libxl-save-helper > #!/bin/bash > export XG_MIGRATION_V2=x > exec /usr/lib/xen/bin/libxl-save-helper.bin "$@" > > which will ensure that XG_MIGRATION_V2 gets set in the environment for both > the source and destination of migration. > > Patches 1 and 2 are prerequisites > Patches 3 to 13 are the meat of the series > Patch 14 is the design document > Patch 15 is the compat layer for unmodified libxl > > Please experiment! As requested by Andrew I've tested them on a PVH Dom0 and it removes the need to enable the MMU_MACHPHYS_UPDATE hypercall on a PVH Dom0 in order to migrate PV guests. Roger.