From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1756273Ab3A1QQS (ORCPT ); Mon, 28 Jan 2013 11:16:18 -0500 Received: from one.firstfloor.org ([213.235.205.2]:54523 "EHLO one.firstfloor.org" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751882Ab3A1QQR (ORCPT ); Mon, 28 Jan 2013 11:16:17 -0500 Date: Mon, 28 Jan 2013 17:16:16 +0100 From: Andi Kleen To: Stephane Eranian Cc: Andi Kleen , Ingo Molnar , LKML , Peter Zijlstra , Andrew Morton , Arnaldo Carvalho de Melo , Jiri Olsa , Namhyung Kim , Andi Kleen Subject: Re: [PATCH 02/12] perf, x86: Basic Haswell PMU support v2 Message-ID: <20130128161616.GS30577@one.firstfloor.org> References: <1359151245-13179-1-git-send-email-andi@firstfloor.org> <1359151245-13179-3-git-send-email-andi@firstfloor.org> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: User-Agent: Mutt/1.4.2.2i Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org > > +static struct event_constraint intel_hsw_event_constraints[] = > > +{ > > + FIXED_EVENT_CONSTRAINT(0x00c0, 0), /* INST_RETIRED.ANY */ > > + FIXED_EVENT_CONSTRAINT(0x003c, 1), /* CPU_CLK_UNHALTED.CORE */ > > + FIXED_EVENT_CONSTRAINT(0x0300, 2), /* CPU_CLK_UNHALTED.REF */ > > + INTEL_EVENT_CONSTRAINT(0x48, 0x4), /* L1D_PEND_MISS.PENDING */ > > You're covering the entire event here, so comment should be: L1D_PEND_MISS.* Ok. > > > + INTEL_UEVENT_CONSTRAINT(0x01c0, 0x2), /* INST_RETIRED.PREC_DIST */ > > + INTEL_EVENT_CONSTRAINT(0xcd, 0x8), /* MEM_TRANS_RETIRED.LOAD_LATENCY */ > > + EVENT_CONSTRAINT_END > > You are missing constraint on the following public events: > - CYCLE_ACTIVITY (0xa3) AFAIK that's not a PEBS event. > > > +}; > > + > > static u64 intel_pmu_event_map(int hw_event) > > { > > return intel_perfmon_event_map[hw_event]; > > @@ -2107,6 +2118,24 @@ __init int intel_pmu_init(void) > > break; > > > > > > + case 60: /* Haswell Client */ > > + case 70: > > + case 71: > > + memcpy(hw_cache_event_ids, snb_hw_cache_event_ids, > > + sizeof(hw_cache_event_ids)); > > + > > + intel_pmu_lbr_init_nhm(); > > + > I suspect this one should be intel_pmu_lbr_init_snb(), otherwise > you inherit the broken filter workarounds. Good point. -Andi -- ak@linux.intel.com -- Speaking for myself only.