From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from eggs.gnu.org ([208.118.235.92]:42510) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1Swnoh-00066M-9i for qemu-devel@nongnu.org; Thu, 02 Aug 2012 01:19:24 -0400 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71) (envelope-from ) id 1Swnog-0001tN-3I for qemu-devel@nongnu.org; Thu, 02 Aug 2012 01:19:23 -0400 Received: from mx3-phx2.redhat.com ([209.132.183.24]:40860) by eggs.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1Swnof-0001tH-RI for qemu-devel@nongnu.org; Thu, 02 Aug 2012 01:19:22 -0400 Date: Thu, 2 Aug 2012 01:19:20 -0400 (EDT) From: Miroslav Rezanina Message-ID: <2067492686.7392228.1343884760982.JavaMail.root@redhat.com> In-Reply-To: <50192D3F.7010303@redhat.com> MIME-Version: 1.0 Content-Type: text/plain; charset=utf-8 Content-Transfer-Encoding: 7bit Subject: Re: [Qemu-devel] [PATCH][RFC] Add compare subcommand for qemu-img List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , To: Eric Blake Cc: Paolo Bonzini , qemu-devel@nongnu.org ----- Original Message ----- > From: "Eric Blake" > To: "Miroslav Rezanina" > Cc: qemu-devel@nongnu.org, "Paolo Bonzini" > Sent: Wednesday, August 1, 2012 3:21:03 PM > Subject: Re: [Qemu-devel] [PATCH][RFC] Add compare subcommand for qemu-img > > On 08/01/2012 04:03 AM, Miroslav Rezanina wrote: > > This patch adds compare subcommand that compares two images. > > Compare has following criteria: > > - only data part is compared > > - unallocated sectors are not read > > - in case of different image size, exceeding part of bigger disk > > has to be zeroed/unallocated to compare rest > > - qemu-img returns: > > - 0 if images are identical > > - 1 if images differ > > - 2 on error > > > > Signed-off-by: Miroslav Rezanina > > > > > +++ b/qemu-img.c > > @@ -96,7 +96,9 @@ static void help(void) > > " '-a' applies a snapshot (revert disk to saved > > state)\n" > > " '-c' creates a snapshot\n" > > " '-d' deletes a snapshot\n" > > - " '-l' lists all snapshots in the given image\n"; > > + " '-l' lists all snapshots in the given image\n" > > + "Parameters to compare subcommand:\n" > > + " '-g' Second image format (in case it differs from > > first image)\n"; > > As written, this sounds like: > > No -f, no -g => probe both > -f, no -g => -f applies to both > no -f, -g => probe first, use -g for second > -f, -g => use given formats for both > > Is that really what you meant, or do we actually get: > Yes, this is what I meant. > -f, no -g => -f applies to first, probe second > > I think both interpretations could make sense, but I'd prefer having > the > omission of -g imply probing the second file type regardless of the > presence or absence of -f, for consistency. > I was evaluating both approach. For me only -f means 'Use this format for input'. In case of compare input is equal to both files so the -f value will be used for both of them. There's only one subcommand allowing to specify multiple input files - convert and it use -f value for all input commands. However, unlike the compare it does not allow to specify different format for different files and has variable number of input files. So I decided to use -f as specification for both files. I do not have problem with using it for first image only but prefer this handling. I can switch handling in next version in case of more votes for it. > -- > Eric Blake eblake@redhat.com +1-919-301-3266 > Libvirt virtualization library http://libvirt.org > > Regards, Miroslav Rezanina