From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from eggs.gnu.org ([2001:4830:134:3::10]:52914) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1VO3Sq-00019o-OB for qemu-devel@nongnu.org; Mon, 23 Sep 2013 06:34:08 -0400 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71) (envelope-from ) id 1VO3Sj-0004UV-Df for qemu-devel@nongnu.org; Mon, 23 Sep 2013 06:34:00 -0400 Received: from ex-e-2.perimeter.fzi.de ([141.21.8.251]:15819) by eggs.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1VO3Sj-0004U6-6r for qemu-devel@nongnu.org; Mon, 23 Sep 2013 06:33:53 -0400 Message-ID: <5240190E.5050204@fzi.de> Date: Mon, 23 Sep 2013 12:33:50 +0200 From: Sebastian Ottlik MIME-Version: 1.0 References: <1379341429-29141-1-git-send-email-ottlik@fzi.de> <52371BCB.4090004@redhat.com> <52371F58.1030303@fzi.de> <5239DBA8.8070608@weilnetz.de> In-Reply-To: <5239DBA8.8070608@weilnetz.de> Content-Type: text/plain; charset="ISO-8859-1"; format=flowed Content-Transfer-Encoding: 7bit Subject: Re: [Qemu-devel] [PATCH v5 0/5] Do not set SO_REUSEADDR on Windows List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , To: Stefan Weil Cc: Paolo Bonzini , Anthony Liguori , qemu-devel@nongnu.org, Stefan Hajnoczi , Jan Kiszka On 18.09.2013 18:58, Stefan Weil wrote: > Am 16.09.2013 17:10, schrieb Sebastian Ottlik: >> On 16.09.2013 16:55, Paolo Bonzini wrote: >>> Il 16/09/2013 16:23, Sebastian Ottlik ha scritto: >>>> - Added the silent flag to socket_set_fast_reuse controlling error >>>> reporting >>>> One location where SO_REUSEADDR was set would report errors if >>>> setting the >>>> option failed. Keeping the reporting code there would be somewhat >>>> unclean, so >>>> I moved it to socket_set_fast_reuse. A side effect of this was >>>> that the error >>>> reporting was added for all locations that now use >>>> socket_set_fast_reuse. Here >>>> a new flag is added to control error reporting, which means this >>>> patchset >>>> won't change QEMU behaviour (except for not setting SO_REUSEADDR >>>> on Windows). >>> Is there actually a case where setting SO_REUSEADDR could fail? >>> >>> Paolo >> Yes, but its very unlikely. E.g. the first parameter is not a valid >> socket. >> > If failures only happen when something is very wrong (like an invalid > socket id), > an assertion might be better, and we could remove the 'silent' parameter. > > Stefan > IMO for debug builds this is a good idea. However, in production use it is probably preferable to keep QEMU running, as a failure won't be too critical. From a quick grep it looks like NDEBUG is not set so assertions wont be removed for non-debug builds. I don't feel acquainted enough with the source code to decide about this kind of change in functionality, which is why I was waiting so long to reply.