From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from eggs.gnu.org ([2001:4830:134:3::10]:51323) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1ZoVOV-0005Ka-N0 for qemu-devel@nongnu.org; Tue, 20 Oct 2015 07:47:56 -0400 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71) (envelope-from ) id 1ZoVOP-0003Kg-PP for qemu-devel@nongnu.org; Tue, 20 Oct 2015 07:47:55 -0400 Received: from mail-vk0-f50.google.com ([209.85.213.50]:34659) by eggs.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1ZoVOP-0003KO-Kw for qemu-devel@nongnu.org; Tue, 20 Oct 2015 07:47:49 -0400 Received: by vkat63 with SMTP id t63so8102269vka.1 for ; Tue, 20 Oct 2015 04:47:49 -0700 (PDT) MIME-Version: 1.0 In-Reply-To: <20151020131952-mutt-send-email-mst@redhat.com> References: <1444403861-3240-1-git-send-email-marcandre.lureau@redhat.com> <20151018111716-mutt-send-email-mst@redhat.com> <20151019183538-mutt-send-email-mst@redhat.com> <20151020001038-mutt-send-email-mst@redhat.com> <20151020131952-mutt-send-email-mst@redhat.com> Date: Tue, 20 Oct 2015 13:47:49 +0200 Message-ID: From: Thibaut Collet Content-Type: text/plain; charset=UTF-8 Subject: Re: [Qemu-devel] [PATCH v8 00/27] vhost-user: add migration support List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , To: "Michael S. Tsirkin" Cc: Linhaifeng , marcandre lureau , Jason Wang , qemu-devel , Paolo Bonzini On Tue, Oct 20, 2015 at 12:21 PM, Michael S. Tsirkin wrote: > On Tue, Oct 20, 2015 at 08:30:49AM +0200, Thibaut Collet wrote: >> On Mon, Oct 19, 2015 at 11:12 PM, Michael S. Tsirkin wrote: >> > On Mon, Oct 19, 2015 at 06:42:43PM +0200, Thibaut Collet wrote: >> >> > >> >> > Can you pls check refs/heads/for_thibaut? >> >> > It should have your patch as the latest commit. >> >> >> >> I do not see yet my patch on the for_thibaut branch >> > >> > Ouch. I meant refs/tags/for_thibaut. >> > Sorry about that. >> > >> > >> >> Sorry for the incorrect wording (I write to quickly my email and use >> the word branch rather than tag). I use the for_thibaut tag for my >> live migration test. The fixup for the double definition of >> vhost_kernel_get_vq_index function id for this tag. >> To do successfully live migration in any conditions I have removed >> this double definition and apply the recent sending patch "vhost: set >> the correct queue index in case of migration with multiqueue" >> >> When you say " It should have your patch as the latest commit." you >> think about which patch ? The >> "0001-FIXUP-vhost-user-add-support-of-live-migration.patch" one or the >> "vhost: set the correct queue index in case of migration with >> multiqueue" one ? >> >> Regards. >> >> Thibaut. > > > This is where for_thibaut points at for me: > > commit bf6830e2416f67571ee2e7196f3625725adec170 > Author: Thibaut Collet > Date: Mon Oct 19 14:59:27 2015 +0200 > > vhost: set the correct queue index in case of migration with multiqueue > > When a live migration is started the log address to mark dirty pages is provided > to the vhost backend through the vhost_dev_set_log function. > This function is called for each queue pairs but the queue index is wrongly set: > always set to the first queue pair. Then vhost backend lost descriptor addresses > of the queue pairs greater than 1 and behaviour of the vhost backend is > unpredictable. > > The queue index is computed by taking account of the vq_index (to retrieve the > queue pair index) and calling the vhost_get_vq_index method of the backend. > > Signed-off-by: Thibaut Collet > > > hash might change if I find any issues. > > If this is not what you see, you need to re-fetch the tag. Ok I got and tested the updated branch. The problem, with my previous attempt, is there are a tag and a branch called for_thibaut. This morning after re-fetch operation I do a "git checkout for_thibaut" that takes the tag and not the branch. I realize it with the commit sha-1 of the "vhost: set the correct queue index in case of migration with multiqueue" patch (sha-1 of commit of for_thibaut tag is e20cff854) > > Please let me know whether this tag works for you. > I have tested this version (commit bf6830e24) with live migration and everything is OK. tested-by: Thibaut Collet > -- > MST