From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from eggs.gnu.org ([208.118.235.92]:43961) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1TiViN-0007t2-2A for qemu-devel@nongnu.org; Tue, 11 Dec 2012 14:42:04 -0500 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71) (envelope-from ) id 1TiViI-000272-R9 for qemu-devel@nongnu.org; Tue, 11 Dec 2012 14:42:02 -0500 Received: from mx1.redhat.com ([209.132.183.28]:61272) by eggs.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1TiViI-00026q-Jh for qemu-devel@nongnu.org; Tue, 11 Dec 2012 14:41:58 -0500 Date: Tue, 11 Dec 2012 17:41:56 -0200 From: Luiz Capitulino Message-ID: <20121211174156.6e3e4d78@doriath.home> In-Reply-To: <24E144B8C0207547AD09C467A8259F755788045F@lisa.maurer-it.com> References: <1355168173-14571-1-git-send-email-lcapitulino@redhat.com> <24E144B8C0207547AD09C467A8259F7557870AEE@lisa.maurer-it.com> <20121211094513.29fa32bd@doriath.home> <24E144B8C0207547AD09C467A8259F755787AC78@lisa.maurer-it.com> <20121211105947.1b92f411@doriath.home> <24E144B8C0207547AD09C467A8259F755787E011@lisa.maurer-it.com> <20121211173808.GH30686@vm> <20121211162840.4b01701e@doriath.home> <24E144B8C0207547AD09C467A8259F755788040D@lisa.maurer-it.com> <20121211172355.3ee96cd4@doriath.home> <24E144B8C0207547AD09C467A8259F755788045F@lisa.maurer-it.com> Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Subject: Re: [Qemu-devel] [PATCH 0/3] re-enable balloon stats List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , To: Dietmar Maurer Cc: "qemu-devel@nongnu.org" , "aliguori@us.ibm.com" , mdroth , "agl@us.ibm.com" On Tue, 11 Dec 2012 19:39:40 +0000 Dietmar Maurer wrote: > > > > but it would silently fail if the feature is never negotiated (as > > > > you said above). > > > > > > Why don't we raise the error when we query the values? > > > > Hmm, that's a good idea. > > > > Only small nit is that, today old stats will remain available to be queried even > > if you disable the timer. If we do what you suggest, old stats won't be > > available if the module is unloaded (well, I *guess* the guest will unset the > > feature bit on module removal). > > What is the problem with that? It would be no problem to simulate the > old behavior, but why do you want that? I actually just thought that it's a nice to have.