From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1753422Ab1HHL55 (ORCPT ); Mon, 8 Aug 2011 07:57:57 -0400 Received: from casper.infradead.org ([85.118.1.10]:37940 "EHLO casper.infradead.org" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752870Ab1HHL54 convert rfc822-to-8bit (ORCPT ); Mon, 8 Aug 2011 07:57:56 -0400 Subject: Re: [PATCH 1/4] perf: Add memory load/store events generic code From: Peter Zijlstra To: Anton Blanchard Cc: Paul Mackerras , Lin Ming , Ingo Molnar , Andi Kleen , Stephane Eranian , Arnaldo Carvalho de Melo , linux-kernel Date: Mon, 08 Aug 2011 13:57:30 +0200 In-Reply-To: <1310138308.3282.708.camel@twins> References: <1309766525-14089-1-git-send-email-ming.m.lin@intel.com> <1309766525-14089-2-git-send-email-ming.m.lin@intel.com> <1309768420.3282.8.camel@twins> <1309769066.3282.14.camel@twins> <1309867418.3282.73.camel@twins> <20110705230217.GA9584@bloggs.ozlabs.ibm.com> <1309960718.3282.289.camel@twins> <20110708171836.6465a12e@kryten> <1310138308.3282.708.camel@twins> Content-Type: text/plain; charset="UTF-8" Content-Transfer-Encoding: 8BIT X-Mailer: Evolution 3.0.2- Message-ID: <1312804650.10488.20.camel@twins> Mime-Version: 1.0 Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On Fri, 2011-07-08 at 17:18 +0200, Peter Zijlstra wrote: > On Fri, 2011-07-08 at 17:18 +1000, Anton Blanchard wrote: > > Hi Peter, > > > > > The thing we're talking about is Intel PEBS Load Latency/Precise Store > > > and AMD IBS where together with a mem op retired event (mem loads > > > retired for Load-Latency, mem stores retired for Precise Store) > > > provides an additional field describing where the load/store was > > > sourced from. > > > > > > Such additional data would require the addition of a > > > PERF_SAMPLE_SOURCE field or similar, for some reason or other I was > > > under the impression some of the PPC chips had something similar. But > > > if not, it saves us having to worry about that. > > > > It does sound a lot like our event vector, where we can have up to > > 64 bits of information that goes with a sample. A lot of the fields > > relate to loads and stores, but there are other fields (eg pipeline > > information at the point the sample was taken). > > > > So we could definitely use a field to capture this. > > Happen to have a ref to some docs about that? We'd want to make sure our > definition is wide enough to also work for ppc. Anton, Paul, any word on this? I'd love to see the specs for that power-pmu event vector thing.. It would be a terrible shame if we now include an abstraction that somewhat matches your needs but not quite, requiring us to either introduce another abi component later or.