From mboxrd@z Thu Jan 1 00:00:00 1970 From: Andi Kleen Subject: Re: Understanding perf mem -t load results Date: Tue, 07 Jan 2014 13:27:01 -0800 Message-ID: <87wqib4h5m.fsf@tassilo.jf.intel.com> References: <1776760155.10395562.1386857164818.JavaMail.root@insa-lyon.fr> <877gb5vmmn.fsf@tassilo.jf.intel.com> <52AE2721.8020501@insa-lyon.fr> <20131215234548.GP21717@two.firstfloor.org> <20131224021822.GH20765@two.firstfloor.org> <52B93352.2030909@gmail.com> <20131224072813.GI20765@two.firstfloor.org> <52B93B01.4020305@gmail.com> <20131224212728.GK20765@two.firstfloor.org> <52BAB2A8.1040709@gmail.com> Mime-Version: 1.0 Content-Type: text/plain Return-path: Received: from mga09.intel.com ([134.134.136.24]:59775 "EHLO mga09.intel.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1753783AbaAGV1C (ORCPT ); Tue, 7 Jan 2014 16:27:02 -0500 In-Reply-To: (Manuel Selva's message of "Tue, 7 Jan 2014 16:06:18 +0100") Sender: linux-perf-users-owner@vger.kernel.org List-ID: To: Manuel Selva Cc: linux-perf-users@vger.kernel.org Manuel Selva writes: > The perf mem record tool use the > MEM_INST_RETIRED.LATENCY_ABOVE_THRESHOLD which I guess belongs to the > memory PEBS events you mentionned (isn't it ?) is reporting Yes. > information about the sources of the events. it seems that for this > purpose the IP information is used, does it mean that what perf mem > reports can be wrong because of the random IP selection mechanism you > mentionned ? The CPU can theoretically execute billions of loads/stores every second. There is no way any reporting mechanism can keep up with that. So the only thing you can do is to sample (only collect every N operations), with a fairly large period. -Andi -- ak@linux.intel.com -- Speaking for myself only