From patchwork Tue Sep 1 17:29:55 2020 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: "Rafael J. Wysocki" X-Patchwork-Id: 256876 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org X-Spam-Level: X-Spam-Status: No, score=-10.0 required=3.0 tests=BAYES_00, HEADER_FROM_DIFFERENT_DOMAINS, INCLUDES_PATCH, MAILING_LIST_MULTI, SIGNED_OFF_BY, SPF_HELO_NONE, SPF_PASS, URIBL_BLOCKED autolearn=unavailable autolearn_force=no version=3.4.0 Received: from mail.kernel.org (mail.kernel.org [198.145.29.99]) by smtp.lore.kernel.org (Postfix) with ESMTP id 052EAC433E6 for ; Tue, 1 Sep 2020 17:40:19 +0000 (UTC) Received: from vger.kernel.org (vger.kernel.org [23.128.96.18]) by mail.kernel.org (Postfix) with ESMTP id E37FB207D3 for ; Tue, 1 Sep 2020 17:40:18 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1728900AbgIARjz (ORCPT ); Tue, 1 Sep 2020 13:39:55 -0400 Received: from cloudserver094114.home.pl ([79.96.170.134]:59386 "EHLO cloudserver094114.home.pl" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1728192AbgIARjy (ORCPT ); Tue, 1 Sep 2020 13:39:54 -0400 Received: from 89-64-88-247.dynamic.chello.pl (89.64.88.247) (HELO kreacher.localnet) by serwer1319399.home.pl (79.96.170.134) with SMTP (IdeaSmtpServer 0.83.459) id 4763ca42f0cd1aa2; Tue, 1 Sep 2020 19:39:52 +0200 From: "Rafael J. Wysocki" To: Linux PM Cc: Srinivas Pandruvada , LKML , Doug Smythies , Artem Bityutskiy Subject: [PATCH v4 3/5] cpufreq: intel_pstate: Tweak the EPP sysfs interface Date: Tue, 01 Sep 2020 19:29:55 +0200 Message-ID: <5280698.oOqPgdve7T@kreacher> In-Reply-To: <3748218.V0HrpZKF9g@kreacher> References: <3748218.V0HrpZKF9g@kreacher> MIME-Version: 1.0 Sender: linux-pm-owner@vger.kernel.org Precedence: bulk List-ID: X-Mailing-List: linux-pm@vger.kernel.org From: "Rafael J. Wysocki" Modify the EPP sysfs interface to reject attempts to change the EPP to values different from 0 ("performance") in the active mode with the "performance" policy (ie. scaling_governor set to "performance"), to avoid situations in which the kernel appears to discard data passed to it via the EPP sysfs attribute. Signed-off-by: Rafael J. Wysocki Reviewed-by: Artem Bityutskiy --- v2 -> v3: New patch v3 -> v4: Add the R-by from Artem --- Documentation/admin-guide/pm/intel_pstate.rst | 4 +++- drivers/cpufreq/intel_pstate.c | 8 ++++++++ 2 files changed, 11 insertions(+), 1 deletion(-) diff --git a/Documentation/admin-guide/pm/intel_pstate.rst b/Documentation/admin-guide/pm/intel_pstate.rst index cdd1a9a7f9a2..5072e7064d13 100644 --- a/Documentation/admin-guide/pm/intel_pstate.rst +++ b/Documentation/admin-guide/pm/intel_pstate.rst @@ -123,7 +123,9 @@ Energy-Performance Bias (EPB) knob (otherwise), which means that the processor's internal P-state selection logic is expected to focus entirely on performance. This will override the EPP/EPB setting coming from the ``sysfs`` interface -(see `Energy vs Performance Hints`_ below). +(see `Energy vs Performance Hints`_ below). Moreover, any attempts to change +the EPP/EPB to a value different from 0 ("performance") via ``sysfs`` in this +configuration will be rejected. Also, in this configuration the range of P-states available to the processor's internal P-state selection logic is always restricted to the upper boundary diff --git a/drivers/cpufreq/intel_pstate.c b/drivers/cpufreq/intel_pstate.c index e540448e0bd0..b308c39b6204 100644 --- a/drivers/cpufreq/intel_pstate.c +++ b/drivers/cpufreq/intel_pstate.c @@ -684,6 +684,14 @@ static int intel_pstate_set_energy_pref_index(struct cpudata *cpu_data, else if (epp == -EINVAL) epp = epp_values[pref_index - 1]; + /* + * To avoid confusion, refuse to set EPP to any values different + * from 0 (performance) if the current policy is "performance", + * because those values would be overridden. + */ + if (epp > 0 && cpu_data->policy == CPUFREQ_POLICY_PERFORMANCE) + return -EBUSY; + ret = intel_pstate_set_epp(cpu_data, epp); } else { if (epp == -EINVAL) From patchwork Tue Sep 1 17:39:32 2020 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: "Rafael J. Wysocki" X-Patchwork-Id: 256875 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org X-Spam-Level: X-Spam-Status: No, score=-10.0 required=3.0 tests=BAYES_00, HEADER_FROM_DIFFERENT_DOMAINS, INCLUDES_PATCH, MAILING_LIST_MULTI, SIGNED_OFF_BY, SPF_HELO_NONE, SPF_PASS, URIBL_BLOCKED autolearn=ham autolearn_force=no version=3.4.0 Received: from mail.kernel.org (mail.kernel.org [198.145.29.99]) by smtp.lore.kernel.org (Postfix) with ESMTP id 609D0C43461 for ; Tue, 1 Sep 2020 17:40:20 +0000 (UTC) Received: from vger.kernel.org (vger.kernel.org [23.128.96.18]) by mail.kernel.org (Postfix) with ESMTP id 4ECD8207D3 for ; Tue, 1 Sep 2020 17:40:20 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1728896AbgIARkS (ORCPT ); Tue, 1 Sep 2020 13:40:18 -0400 Received: from cloudserver094114.home.pl ([79.96.170.134]:60874 "EHLO cloudserver094114.home.pl" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1726116AbgIARjz (ORCPT ); Tue, 1 Sep 2020 13:39:55 -0400 Received: from 89-64-88-247.dynamic.chello.pl (89.64.88.247) (HELO kreacher.localnet) by serwer1319399.home.pl (79.96.170.134) with SMTP (IdeaSmtpServer 0.83.459) id 0a7eba54eafc963d; Tue, 1 Sep 2020 19:39:49 +0200 From: "Rafael J. Wysocki" To: Linux PM Cc: Srinivas Pandruvada , LKML , Doug Smythies , Artem Bityutskiy Subject: [PATCH v4 5/5] cpufreq: intel_pstate: Free memory only when turning off Date: Tue, 01 Sep 2020 19:39:32 +0200 Message-ID: <28055847.QciAqr1Adi@kreacher> In-Reply-To: <3748218.V0HrpZKF9g@kreacher> References: <3748218.V0HrpZKF9g@kreacher> MIME-Version: 1.0 Sender: linux-pm-owner@vger.kernel.org Precedence: bulk List-ID: X-Mailing-List: linux-pm@vger.kernel.org From: "Rafael J. Wysocki" When intel_pstate switches the operation mode from "active" to "passive" or the other way around, freeing its data structures representing CPUs and allocating them again from scratch is not necessary and wasteful. Moreover, if these data structures are preserved, the cached HWP Request MSR value from there may be written to the MSR to start with to reinitialize it and help to restore the EPP value set previously (it is set to 0xFF when CPUs go offline to allow their SMT siblings to use the full range of EPP values and that also happens when the driver gets unregistered). Accordingly, modify the driver to only do a full cleanup on driver object registration errors and when its status is changed to "off" via sysfs and to write the cached HWP Request MSR value back to the MSR on CPU init if the data structure representing the given CPU is still there. Signed-off-by: Rafael J. Wysocki --- -> v2: Rearrange intel_pstate_init_cpu() to restore some of the previous behavior of it to retain the current active-mode EPP management. v2 -> v3: * Rebase (it was [4/5] previously). v3 -> v4: * Re-enable HWP in "init" even if the data structures are in there. --- drivers/cpufreq/intel_pstate.c | 57 ++++++++++++++-------------------- 1 file changed, 24 insertions(+), 33 deletions(-) diff --git a/drivers/cpufreq/intel_pstate.c b/drivers/cpufreq/intel_pstate.c index 8181a1f1dc79..c92c085fc495 100644 --- a/drivers/cpufreq/intel_pstate.c +++ b/drivers/cpufreq/intel_pstate.c @@ -2116,25 +2116,31 @@ static int intel_pstate_init_cpu(unsigned int cpunum) all_cpu_data[cpunum] = cpu; - cpu->epp_default = -EINVAL; - cpu->epp_powersave = -EINVAL; - } + cpu->cpu = cpunum; - cpu = all_cpu_data[cpunum]; + cpu->epp_default = -EINVAL; - cpu->cpu = cpunum; + if (hwp_active) { + const struct x86_cpu_id *id; - if (hwp_active) { - const struct x86_cpu_id *id; + intel_pstate_hwp_enable(cpu); - intel_pstate_hwp_enable(cpu); - cpu->epp_policy = 0; - - id = x86_match_cpu(intel_pstate_hwp_boost_ids); - if (id && intel_pstate_acpi_pm_profile_server()) - hwp_boost = true; + id = x86_match_cpu(intel_pstate_hwp_boost_ids); + if (id && intel_pstate_acpi_pm_profile_server()) + hwp_boost = true; + } + } else if (hwp_active) { + /* + * Re-enable HWP in case this happens after a resume from ACPI + * S3 if the CPU was offline during the whole system/resume + * cycle. + */ + intel_pstate_hwp_reenable(cpu); } + cpu->epp_powersave = -EINVAL; + cpu->epp_policy = 0; + intel_pstate_get_cpu_pstates(cpu); pr_debug("controlling: cpu %d\n", cpunum); @@ -2730,9 +2736,6 @@ static void intel_pstate_driver_cleanup(void) } put_online_cpus(); - if (intel_pstate_driver == &intel_pstate) - intel_pstate_sysfs_hide_hwp_dynamic_boost(); - intel_pstate_driver = NULL; } @@ -2758,14 +2761,6 @@ static int intel_pstate_register_driver(struct cpufreq_driver *driver) return 0; } -static int intel_pstate_unregister_driver(void) -{ - cpufreq_unregister_driver(intel_pstate_driver); - intel_pstate_driver_cleanup(); - - return 0; -} - static ssize_t intel_pstate_show_status(char *buf) { if (!intel_pstate_driver) @@ -2777,8 +2772,6 @@ static ssize_t intel_pstate_show_status(char *buf) static int intel_pstate_update_status(const char *buf, size_t size) { - int ret; - if (size == 3 && !strncmp(buf, "off", size)) { if (!intel_pstate_driver) return -EINVAL; @@ -2786,7 +2779,8 @@ static int intel_pstate_update_status(const char *buf, size_t size) if (hwp_active) return -EBUSY; - return intel_pstate_unregister_driver(); + cpufreq_unregister_driver(intel_pstate_driver); + intel_pstate_driver_cleanup(); } if (size == 6 && !strncmp(buf, "active", size)) { @@ -2794,9 +2788,7 @@ static int intel_pstate_update_status(const char *buf, size_t size) if (intel_pstate_driver == &intel_pstate) return 0; - ret = intel_pstate_unregister_driver(); - if (ret) - return ret; + cpufreq_unregister_driver(intel_pstate_driver); } return intel_pstate_register_driver(&intel_pstate); @@ -2807,9 +2799,8 @@ static int intel_pstate_update_status(const char *buf, size_t size) if (intel_pstate_driver == &intel_cpufreq) return 0; - ret = intel_pstate_unregister_driver(); - if (ret) - return ret; + cpufreq_unregister_driver(intel_pstate_driver); + intel_pstate_sysfs_hide_hwp_dynamic_boost(); } return intel_pstate_register_driver(&intel_cpufreq);