From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from eggs.gnu.org ([2001:4830:134:3::10]:50477) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1fNDjw-0007kp-De for qemu-devel@nongnu.org; Mon, 28 May 2018 04:42:53 -0400 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71) (envelope-from ) id 1fNDjv-0002my-NA for qemu-devel@nongnu.org; Mon, 28 May 2018 04:42:52 -0400 Date: Mon, 28 May 2018 10:42:40 +0200 From: Kevin Wolf Message-ID: <20180528084240.GC4580@localhost.localdomain> References: <20180525163327.23097-1-kwolf@redhat.com> <737e2294-6d40-04a0-3bb8-88ed85d6e224@redhat.com> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <737e2294-6d40-04a0-3bb8-88ed85d6e224@redhat.com> Subject: Re: [Qemu-devel] [PATCH 00/14] block: Make blockdev-create a job and stable API List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , To: Eric Blake Cc: qemu-block@nongnu.org, mreitz@redhat.com, jsnow@redhat.com, pkrempa@redhat.com, qemu-devel@nongnu.org Am 25.05.2018 um 20:13 hat Eric Blake geschrieben: > On 05/25/2018 11:33 AM, Kevin Wolf wrote: > > This changes the x-blockdev-create QMP command so that it doesn't block > > the monitor and the main loop any more, but starts a background job that > > performs the image creation. > > > > The basic job as implemented here is all that is necessary to make image > > creation asynchronous and to provide a QMP interface that can be marked > > stable, but it still lacks a few features that jobs usually provide: The > > job will ignore pause commands and it doesn't publish progress yet (so > > both current-progress and total-progress stay at 0). These features can > > be added later without breaking compatibility. > > Can we at least have total-progress start at 1, and current-progress move > from 0 to 1 at completion? Seeing a 0/1 => 1/1 transition is better than a > divide-by-zero 0/0 ratio throughout the entire job; and libvirt doesn't want > to add any more special-casing of 0/0 than it already has (where it wants to > treat that as "job not yet started" rather than the more usual sense that if > total==current the job is hopefully complete). Sure, I can do that. Kevin