From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from eggs.gnu.org ([2001:4830:134:3::10]:54138) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1agBDU-0000Ce-RV for qemu-devel@nongnu.org; Wed, 16 Mar 2016 09:10:28 -0400 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71) (envelope-from ) id 1agBDQ-0003WS-UC for qemu-devel@nongnu.org; Wed, 16 Mar 2016 09:10:24 -0400 Received: from mail-db5eur01on0104.outbound.protection.outlook.com ([104.47.2.104]:23648 helo=EUR01-DB5-obe.outbound.protection.outlook.com) by eggs.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1agBDQ-0003Q2-A6 for qemu-devel@nongnu.org; Wed, 16 Mar 2016 09:10:20 -0400 References: <1457954501-26528-1-git-send-email-den@openvz.org> <1457954501-26528-7-git-send-email-den@openvz.org> <56E6CBD1.2060509@redhat.com> <20160314143857.GE21198@redhat.com> <56E6D323.5080803@openvz.org> <20160314152604.GG21198@redhat.com> <20160314161153.GH21198@redhat.com> <56E6E3DB.2060109@openvz.org> <56E6E984.1030000@redhat.com> From: "Denis V. Lunev" Message-ID: <56E95B22.9070209@openvz.org> Date: Wed, 16 Mar 2016 16:09:54 +0300 MIME-Version: 1.0 In-Reply-To: <56E6E984.1030000@redhat.com> Content-Type: text/plain; charset="utf-8"; format=flowed Content-Transfer-Encoding: 7bit Subject: Re: [Qemu-devel] [PATCH 6/9] log: log QMP commands and replies List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , To: Paolo Bonzini , "Daniel P. Berrange" Cc: Markus Armbruster , Luiz Capitulino , qemu-devel@nongnu.org, Stefan Hajnoczi , Pavel Butsykin On 03/14/2016 07:40 PM, Paolo Bonzini wrote: > > On 14/03/2016 17:16, Denis V. Lunev wrote: >>> BTW, I'm not against adding this to QEMU in general - just pointing out >>> that in the context of libvirt usage reference in the commit message, >>> it is not really needed. I'm totally ambivalent wrt adding this to QEMU >>> for benefit of non-libvirt usage. >> I see. This just requires very clear commit message from my side. >> Thank you for a discussion. > I think this is not a compelling usage of the logging subsystem; QMP is > always used through a character device, and character devices already > have a logging mechanism. > > Adding a mechanism that looks for '\n' in the data written to a > character device and emits a timestamp on the next character would be a > useful addition. I'd be entirely happy with adding it now, it should be > safe enough that we can do it during soft freeze. > > Paolo really interesting approach with a good place to get in. But we will need to modify this a infrastructure a bit. Af far as I can see only data from VM to outside world is logged. Data being 'read' should be logged too as far as I could understand. Den