From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from eggs.gnu.org ([2001:4830:134:3::10]:56488) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1VOSy1-0000wT-Dc for qemu-devel@nongnu.org; Tue, 24 Sep 2013 09:48:01 -0400 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71) (envelope-from ) id 1VOSxs-0000HB-F6 for qemu-devel@nongnu.org; Tue, 24 Sep 2013 09:47:53 -0400 Received: from mail-ea0-x22d.google.com ([2a00:1450:4013:c01::22d]:41347) by eggs.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1VOSxs-0000Ga-7W for qemu-devel@nongnu.org; Tue, 24 Sep 2013 09:47:44 -0400 Received: by mail-ea0-f173.google.com with SMTP id g10so2483765eak.32 for ; Tue, 24 Sep 2013 06:47:43 -0700 (PDT) Date: Tue, 24 Sep 2013 15:47:40 +0200 From: Stefan Hajnoczi Message-ID: <20130924134740.GA13194@stefanha-thinkpad.redhat.com> References: <52387EC2.6000502@siemens.com> <5238810E.1050304@redhat.com> <5238883A.90607@siemens.com> <52388BAA.1060207@redhat.com> <6BF1E4A0-946C-402A-9225-0B1DE04DBC54@alex.org.uk> <52395CD5.7030203@redhat.com> <6DF6BCC2-14BC-4B10-B22E-887D2AFF2012@alex.org.uk> <52397185.9010405@redhat.com> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <52397185.9010405@redhat.com> Subject: Re: [Qemu-devel] [RFC] [PATCHv10 00/31] aio / timers: Add AioContext timers and use ppoll List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , To: Paolo Bonzini Cc: Kevin Wolf , Anthony Liguori , Stefan Hajnoczi , Jan Kiszka , qemu-devel@nongnu.org, liu ping fan , Alex Bligh , MORITA Kazutaka , rth@twiddle.net On Wed, Sep 18, 2013 at 11:25:25AM +0200, Paolo Bonzini wrote: > Il 18/09/2013 11:02, Alex Bligh ha scritto: > > Paolo, > > > > On 18 Sep 2013, at 09:23, Alex Bligh wrote: > > > >>> Yes, that was my understanding too. Can we do it for 1.7? > > > > Whilst we are changing the calling semantics, do you think > > qemu_coroutine_yield() should also run the timers for the > > aio_context? IE should timers always be deferred to the > > next qemu_aio_wait() / aio_poll(), or should we make it > > so they can run whenever there is a 'schedule' between > > coroutines? > > No, they should be left for the next aio_poll(). I agree. Besides the side-effects, I want to avoid coupling coroutines too deeply with event loop entities like timers. They should not know about each other. Stefan