Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1754912Ab1BPNtW (ORCPT ); Wed, 16 Feb 2011 08:49:22 -0500 Received: from hera.kernel.org ([140.211.167.34]:40037 "EHLO hera.kernel.org" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1754555Ab1BPNtT (ORCPT ); Wed, 16 Feb 2011 08:49:19 -0500 Date: Wed, 16 Feb 2011 13:48:50 GMT From: tip-bot for Robert Richter Cc: linux-kernel@vger.kernel.org, hpa@zytor.com, mingo@redhat.com, robert.richter@amd.com, a.p.zijlstra@chello.nl, tglx@linutronix.de, mingo@elte.hu Reply-To: mingo@redhat.com, hpa@zytor.com, linux-kernel@vger.kernel.org, robert.richter@amd.com, a.p.zijlstra@chello.nl, tglx@linutronix.de, mingo@elte.hu In-Reply-To: <1296664860-10886-5-git-send-email-robert.richter@amd.com> References: <1296664860-10886-5-git-send-email-robert.richter@amd.com> To: linux-tip-commits@vger.kernel.org Subject: [tip:perf/core] perf, x86: Store perfctr msr addresses in config_base/event_base Message-ID: Git-Commit-ID: 73d6e52206a20354738418625cedc244cbfd5023 X-Mailer: tip-git-log-daemon Robot-ID: Robot-Unsubscribe: Contact to get blacklisted from these emails MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Disposition: inline X-Greylist: Sender IP whitelisted, not delayed by milter-greylist-4.2.3 (hera.kernel.org [127.0.0.1]); Wed, 16 Feb 2011 13:48:51 +0000 (UTC) Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 5807 Lines: 154 Commit-ID: 73d6e52206a20354738418625cedc244cbfd5023 Gitweb: http://git.kernel.org/tip/73d6e52206a20354738418625cedc244cbfd5023 Author: Robert Richter AuthorDate: Wed, 2 Feb 2011 17:40:59 +0100 Committer: Ingo Molnar CommitDate: Wed, 16 Feb 2011 13:30:52 +0100 perf, x86: Store perfctr msr addresses in config_base/event_base Instead of storing the base addresses we can store the counter's msr addresses directly in config_base/event_base of struct hw_perf_event. This avoids recalculating the address with each msr access. The addresses are configured one time. We also need this change to later modify the address calculation. Signed-off-by: Robert Richter Signed-off-by: Peter Zijlstra LKML-Reference: <1296664860-10886-5-git-send-email-robert.richter@amd.com> Signed-off-by: Ingo Molnar --- arch/x86/kernel/cpu/perf_event.c | 21 ++++++++------------- arch/x86/kernel/cpu/perf_event_p4.c | 8 ++++---- arch/x86/kernel/cpu/perf_event_p6.c | 4 ++-- 3 files changed, 14 insertions(+), 19 deletions(-) diff --git a/arch/x86/kernel/cpu/perf_event.c b/arch/x86/kernel/cpu/perf_event.c index ee40c1ad..3161943 100644 --- a/arch/x86/kernel/cpu/perf_event.c +++ b/arch/x86/kernel/cpu/perf_event.c @@ -298,7 +298,7 @@ x86_perf_event_update(struct perf_event *event) */ again: prev_raw_count = local64_read(&hwc->prev_count); - rdmsrl(hwc->event_base + idx, new_raw_count); + rdmsrl(hwc->event_base, new_raw_count); if (local64_cmpxchg(&hwc->prev_count, prev_raw_count, new_raw_count) != prev_raw_count) @@ -655,7 +655,7 @@ static void x86_pmu_disable(struct pmu *pmu) static inline void __x86_pmu_enable_event(struct hw_perf_event *hwc, u64 enable_mask) { - wrmsrl(hwc->config_base + hwc->idx, hwc->config | enable_mask); + wrmsrl(hwc->config_base, hwc->config | enable_mask); } static void x86_pmu_enable_all(int added) @@ -834,15 +834,10 @@ static inline void x86_assign_hw_event(struct perf_event *event, hwc->event_base = 0; } else if (hwc->idx >= X86_PMC_IDX_FIXED) { hwc->config_base = MSR_ARCH_PERFMON_FIXED_CTR_CTRL; - /* - * We set it so that event_base + idx in wrmsr/rdmsr maps to - * MSR_ARCH_PERFMON_FIXED_CTR0 ... CTR2: - */ - hwc->event_base = - MSR_ARCH_PERFMON_FIXED_CTR0 - X86_PMC_IDX_FIXED; + hwc->event_base = MSR_ARCH_PERFMON_FIXED_CTR0; } else { - hwc->config_base = x86_pmu.eventsel; - hwc->event_base = x86_pmu.perfctr; + hwc->config_base = x86_pmu_config_addr(hwc->idx); + hwc->event_base = x86_pmu_event_addr(hwc->idx); } } @@ -932,7 +927,7 @@ static inline void x86_pmu_disable_event(struct perf_event *event) { struct hw_perf_event *hwc = &event->hw; - wrmsrl(hwc->config_base + hwc->idx, hwc->config); + wrmsrl(hwc->config_base, hwc->config); } static DEFINE_PER_CPU(u64 [X86_PMC_IDX_MAX], pmc_prev_left); @@ -985,7 +980,7 @@ x86_perf_event_set_period(struct perf_event *event) */ local64_set(&hwc->prev_count, (u64)-left); - wrmsrl(hwc->event_base + idx, (u64)(-left) & x86_pmu.cntval_mask); + wrmsrl(hwc->event_base, (u64)(-left) & x86_pmu.cntval_mask); /* * Due to erratum on certan cpu we need @@ -993,7 +988,7 @@ x86_perf_event_set_period(struct perf_event *event) * is updated properly */ if (x86_pmu.perfctr_second_write) { - wrmsrl(hwc->event_base + idx, + wrmsrl(hwc->event_base, (u64)(-left) & x86_pmu.cntval_mask); } diff --git a/arch/x86/kernel/cpu/perf_event_p4.c b/arch/x86/kernel/cpu/perf_event_p4.c index ff751a9..3769ac8 100644 --- a/arch/x86/kernel/cpu/perf_event_p4.c +++ b/arch/x86/kernel/cpu/perf_event_p4.c @@ -764,9 +764,9 @@ static inline int p4_pmu_clear_cccr_ovf(struct hw_perf_event *hwc) u64 v; /* an official way for overflow indication */ - rdmsrl(hwc->config_base + hwc->idx, v); + rdmsrl(hwc->config_base, v); if (v & P4_CCCR_OVF) { - wrmsrl(hwc->config_base + hwc->idx, v & ~P4_CCCR_OVF); + wrmsrl(hwc->config_base, v & ~P4_CCCR_OVF); return 1; } @@ -815,7 +815,7 @@ static inline void p4_pmu_disable_event(struct perf_event *event) * state we need to clear P4_CCCR_OVF, otherwise interrupt get * asserted again and again */ - (void)checking_wrmsrl(hwc->config_base + hwc->idx, + (void)checking_wrmsrl(hwc->config_base, (u64)(p4_config_unpack_cccr(hwc->config)) & ~P4_CCCR_ENABLE & ~P4_CCCR_OVF & ~P4_CCCR_RESERVED); } @@ -885,7 +885,7 @@ static void p4_pmu_enable_event(struct perf_event *event) p4_pmu_enable_pebs(hwc->config); (void)checking_wrmsrl(escr_addr, escr_conf); - (void)checking_wrmsrl(hwc->config_base + hwc->idx, + (void)checking_wrmsrl(hwc->config_base, (cccr & ~P4_CCCR_RESERVED) | P4_CCCR_ENABLE); } diff --git a/arch/x86/kernel/cpu/perf_event_p6.c b/arch/x86/kernel/cpu/perf_event_p6.c index 34ba07b..20c097e 100644 --- a/arch/x86/kernel/cpu/perf_event_p6.c +++ b/arch/x86/kernel/cpu/perf_event_p6.c @@ -68,7 +68,7 @@ p6_pmu_disable_event(struct perf_event *event) if (cpuc->enabled) val |= ARCH_PERFMON_EVENTSEL_ENABLE; - (void)checking_wrmsrl(hwc->config_base + hwc->idx, val); + (void)checking_wrmsrl(hwc->config_base, val); } static void p6_pmu_enable_event(struct perf_event *event) @@ -81,7 +81,7 @@ static void p6_pmu_enable_event(struct perf_event *event) if (cpuc->enabled) val |= ARCH_PERFMON_EVENTSEL_ENABLE; - (void)checking_wrmsrl(hwc->config_base + hwc->idx, val); + (void)checking_wrmsrl(hwc->config_base, val); } static __initconst const struct x86_pmu p6_pmu = { -- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/