From mboxrd@z Thu Jan 1 00:00:00 1970 From: Ian Campbell Subject: Re: [PATCH v8 5/5] tools, docs: add total/local memory bandwith monitoring Date: Wed, 28 Jan 2015 14:12:52 +0000 Message-ID: <1422454372.5187.3.camel@citrix.com> References: <1422432286-7276-1-git-send-email-chao.p.peng@linux.intel.com> <1422432286-7276-6-git-send-email-chao.p.peng@linux.intel.com> <1422453899.14124.68.camel@citrix.com> <21704.60889.831165.117853@mariner.uk.xensource.com> Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit Return-path: In-Reply-To: <21704.60889.831165.117853@mariner.uk.xensource.com> List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Sender: xen-devel-bounces@lists.xen.org Errors-To: xen-devel-bounces@lists.xen.org To: Ian Jackson Cc: wei.liu2@citrix.com, keir@xen.org, stefano.stabellini@eu.citrix.com, andrew.cooper3@citrix.com, xen-devel@lists.xen.org, will.auld@intel.com, JBeulich@suse.com, Chao Peng List-Id: xen-devel@lists.xenproject.org On Wed, 2015-01-28 at 14:10 +0000, Ian Jackson wrote: > Ian Campbell writes ("Re: [PATCH v8 5/5] tools, docs: add total/local memory bandwith monitoring"): > ... > > Another alternative would be to expose the instantaneous values to the > > libxl user and let it sleep as it wishes and calculate the bandwidth > > itself. > > I think that would be better. > > Suppose the libxl caller wants to record the mean bandwidth over a > longer period: they shouldn't have to keep calling this function and > adding up the results (and if they did their algorithm would be racy). True. >>From the code it seems like there is some h/w constraint on roll over of the underlying counter, which might complicate this somewhat. At the very least the libxl interface docs would need to be quite extensive. Ian.