From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1756951Ab2FFQSL (ORCPT ); Wed, 6 Jun 2012 12:18:11 -0400 Received: from terminus.zytor.com ([198.137.202.10]:40728 "EHLO terminus.zytor.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1754113Ab2FFQSJ (ORCPT ); Wed, 6 Jun 2012 12:18:09 -0400 Date: Wed, 6 Jun 2012 09:17:56 -0700 From: tip-bot for Andi Kleen Message-ID: Cc: linux-kernel@vger.kernel.org, hpa@zytor.com, mingo@kernel.org, a.p.zijlstra@chello.nl, ak@linux.intel.com, tglx@linutronix.de Reply-To: mingo@kernel.org, hpa@zytor.com, linux-kernel@vger.kernel.org, a.p.zijlstra@chello.nl, ak@linux.intel.com, tglx@linutronix.de In-Reply-To: <1338944211-28275-2-git-send-email-andi@firstfloor.org> References: <1338944211-28275-2-git-send-email-andi@firstfloor.org> To: linux-tip-commits@vger.kernel.org Subject: [tip:perf/core] perf/x86: Don' t assume there can be only 4 PEBS events Git-Commit-ID: 70ab7003dec58afeae7f5d681dfa309b3a259f03 X-Mailer: tip-git-log-daemon Robot-ID: Robot-Unsubscribe: Contact to get blacklisted from these emails MIME-Version: 1.0 Content-Transfer-Encoding: 8bit Content-Type: text/plain; charset=UTF-8 Content-Disposition: inline X-Greylist: Sender IP whitelisted, not delayed by milter-greylist-4.2.6 (terminus.zytor.com [127.0.0.1]); Wed, 06 Jun 2012 09:18:02 -0700 (PDT) Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Commit-ID: 70ab7003dec58afeae7f5d681dfa309b3a259f03 Gitweb: http://git.kernel.org/tip/70ab7003dec58afeae7f5d681dfa309b3a259f03 Author: Andi Kleen AuthorDate: Tue, 5 Jun 2012 17:56:48 -0700 Committer: Ingo Molnar CommitDate: Wed, 6 Jun 2012 17:23:40 +0200 perf/x86: Don't assume there can be only 4 PEBS events On Sandy Bridge in non HT mode there are 8 counters available. Since every counter can write a PEBS record assuming there are 4 max is incorrect. Use the reported counter number -- with an upper limit for a static array -- instead. Also I made the warning messages a bit more informational. Signed-off-by: Andi Kleen Signed-off-by: Peter Zijlstra Link: http://lkml.kernel.org/r/1338944211-28275-2-git-send-email-andi@firstfloor.org Signed-off-by: Ingo Molnar --- arch/x86/kernel/cpu/perf_event.h | 3 ++- arch/x86/kernel/cpu/perf_event_intel.c | 2 ++ arch/x86/kernel/cpu/perf_event_intel_ds.c | 8 ++++---- 3 files changed, 8 insertions(+), 5 deletions(-) diff --git a/arch/x86/kernel/cpu/perf_event.h b/arch/x86/kernel/cpu/perf_event.h index 23b5710..3df3de9 100644 --- a/arch/x86/kernel/cpu/perf_event.h +++ b/arch/x86/kernel/cpu/perf_event.h @@ -69,7 +69,7 @@ struct amd_nb { }; /* The maximal number of PEBS events: */ -#define MAX_PEBS_EVENTS 4 +#define MAX_PEBS_EVENTS 8 /* * A debug store configuration. @@ -378,6 +378,7 @@ struct x86_pmu { void (*drain_pebs)(struct pt_regs *regs); struct event_constraint *pebs_constraints; void (*pebs_aliases)(struct perf_event *event); + int max_pebs_events; /* * Intel LBR diff --git a/arch/x86/kernel/cpu/perf_event_intel.c b/arch/x86/kernel/cpu/perf_event_intel.c index 187c294..e23e71f 100644 --- a/arch/x86/kernel/cpu/perf_event_intel.c +++ b/arch/x86/kernel/cpu/perf_event_intel.c @@ -1800,6 +1800,8 @@ __init int intel_pmu_init(void) x86_pmu.events_maskl = ebx.full; x86_pmu.events_mask_len = eax.split.mask_length; + x86_pmu.max_pebs_events = min_t(unsigned, MAX_PEBS_EVENTS, x86_pmu.num_counters); + /* * Quirk: v2 perfmon does not report fixed-purpose events, so * assume at least 3 events: diff --git a/arch/x86/kernel/cpu/perf_event_intel_ds.c b/arch/x86/kernel/cpu/perf_event_intel_ds.c index 35e2192..026373e 100644 --- a/arch/x86/kernel/cpu/perf_event_intel_ds.c +++ b/arch/x86/kernel/cpu/perf_event_intel_ds.c @@ -620,7 +620,7 @@ static void intel_pmu_drain_pebs_core(struct pt_regs *iregs) * Should not happen, we program the threshold at 1 and do not * set a reset value. */ - WARN_ON_ONCE(n > 1); + WARN_ONCE(n > 1, "bad leftover pebs %d\n", n); at += n - 1; __intel_pmu_pebs_event(event, iregs, at); @@ -651,10 +651,10 @@ static void intel_pmu_drain_pebs_nhm(struct pt_regs *iregs) * Should not happen, we program the threshold at 1 and do not * set a reset value. */ - WARN_ON_ONCE(n > MAX_PEBS_EVENTS); + WARN_ONCE(n > x86_pmu.max_pebs_events, "Unexpected number of pebs records %d\n", n); for ( ; at < top; at++) { - for_each_set_bit(bit, (unsigned long *)&at->status, MAX_PEBS_EVENTS) { + for_each_set_bit(bit, (unsigned long *)&at->status, x86_pmu.max_pebs_events) { event = cpuc->events[bit]; if (!test_bit(bit, cpuc->active_mask)) continue; @@ -670,7 +670,7 @@ static void intel_pmu_drain_pebs_nhm(struct pt_regs *iregs) break; } - if (!event || bit >= MAX_PEBS_EVENTS) + if (!event || bit >= x86_pmu.max_pebs_events) continue; __intel_pmu_pebs_event(event, iregs, at);