From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from eggs.gnu.org ([2001:4830:134:3::10]:38486) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1WsTzB-0006SQ-4N for qemu-devel@nongnu.org; Thu, 05 Jun 2014 05:29:34 -0400 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71) (envelope-from ) id 1WsTz2-0002V4-1q for qemu-devel@nongnu.org; Thu, 05 Jun 2014 05:29:25 -0400 Received: from mx-v6.kamp.de ([2a02:248:0:51::16]:52791 helo=mx01.kamp.de) by eggs.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1WsTz1-0002Sq-FV for qemu-devel@nongnu.org; Thu, 05 Jun 2014 05:29:15 -0400 Message-ID: <53903868.6070500@kamp.de> Date: Thu, 05 Jun 2014 11:29:12 +0200 From: Peter Lieven MIME-Version: 1.0 References: <1401384704-14103-1-git-send-email-pl@kamp.de> <538772C9.8080500@redhat.com> <53877F41.7090805@kamp.de> <538787DC.10306@redhat.com> <5387979F.1070103@kamp.de> <53881C32.2040703@redhat.com> <5388498C.9040007@kamp.de> <538849D8.10806@redhat.com> In-Reply-To: <538849D8.10806@redhat.com> Content-Type: text/plain; charset=ISO-8859-15; format=flowed Content-Transfer-Encoding: 7bit Subject: Re: [Qemu-devel] [PATCH] block/iscsi: handle BUSY condition List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , To: Paolo Bonzini , qemu-devel@nongnu.org Cc: kwolf@redhat.com, famz@redhat.com, stefanha@redhat.com, ronniesahlberg@gmail.com On 30.05.2014 11:05, Paolo Bonzini wrote: > Il 30/05/2014 11:04, Peter Lieven ha scritto: >> Am 30.05.2014 07:50, schrieb Paolo Bonzini: >>> Il 29/05/2014 22:25, Peter Lieven ha scritto: >>>> Am 29.05.2014 21:17, schrieb Paolo Bonzini: >>>>> Il 29/05/2014 20:41, Peter Lieven ha scritto: >>>>>> static inline unsigned geo_rand_range(double m, double n) { >>>>>> return exp((log(m) + (double)rand() / >>>>>> (RAND_MAX / (log(n) - log(m) + log(1)) + log(1)))); >>>>>> } >>>>> >>>>> Yes, exactly (except that log(1) is zero, and was only needed in your case because you were using integers). >>>> >>>> Regarding aio_timer_new vs. aio_timer_init. Shall we really have the complete QEMU_TIMER struct in each >>>> iTask? We need this retry_timer only in the really rare case of a BUSY. >>> >>> Ok, makes sense. Though the overall cost of IscsiTasks should be less than 4k per block device with a separate QemuTimer, and 8k if you embed it. >>> >>> Otherwise the patch looks good, thanks! >> >> Okay, when are you preparing the next pull request? > > After the dataplane patches go in. But you can send the patch now, I'll rebase. It might be that we need a LIBS+= -lm to the Makefile in the root directory. I had strange problems when linking qemu-img/qemu-io/qemu-nbd. Peter