From patchwork Thu Jul 7 07:15:52 2022 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Lukasz Luba X-Patchwork-Id: 12909130 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org Received: from bombadil.infradead.org (bombadil.infradead.org [198.137.202.133]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by smtp.lore.kernel.org (Postfix) with ESMTPS id 67A59C433EF for ; Thu, 7 Jul 2022 07:17:46 +0000 (UTC) DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=lists.infradead.org; s=bombadil.20210309; h=Sender: Content-Transfer-Encoding:Content-Type:MIME-Version:List-Subscribe:List-Help: List-Post:List-Archive:List-Unsubscribe:List-Id:References:In-Reply-To: Message-Id:Date:Subject:Cc:To:From:Reply-To:Content-ID:Content-Description: Resent-Date:Resent-From:Resent-Sender:Resent-To:Resent-Cc:Resent-Message-ID: List-Owner; bh=mVw7W92PkQFa8HwpIbJYW+DJZew7x5zDEscWeuVIkZM=; b=1v7z1DrIonsjMR j8g4pqdatnojdjwbCEqcqquPhuzeUgF/merjFOC7Iuq/6HIugVntfApRvGoxZzHrKQ/nT6pp3Z2Ds I8pF6/K4P1KsSHY9CyUcSXaM2RCSpEu4eRZoSUC0Hp1Nr2P3+5R1vbczH35k1UKCPiI06WrtpRcFq NuqkuUZX5LxXozGgxDyLnVyshP6O7YeY0PXiw3NfR4qHb6LZ/0Z9bfseUTCdS/tdUN8E39mwkc7FR iA546u5JXMWPrX4lvacIPLBbZU9Raq8u2QqIFcIWOsBBcvihb2xAa+I5a9BN0JTbyfAFqALfRekft i31+pQOxxw4edk4nLmGw==; Received: from localhost ([::1] helo=bombadil.infradead.org) by bombadil.infradead.org with esmtp (Exim 4.94.2 #2 (Red Hat Linux)) id 1o9Lkc-00ETmL-84; Thu, 07 Jul 2022 07:16:38 +0000 Received: from foss.arm.com ([217.140.110.172]) by bombadil.infradead.org with esmtp (Exim 4.94.2 #2 (Red Hat Linux)) id 1o9LkB-00ETcY-Nr; Thu, 07 Jul 2022 07:16:14 +0000 Received: from usa-sjc-imap-foss1.foss.arm.com (unknown [10.121.207.14]) by usa-sjc-mx-foss1.foss.arm.com (Postfix) with ESMTP id E695F106F; Thu, 7 Jul 2022 00:16:09 -0700 (PDT) Received: from e123648.arm.com (unknown [10.57.11.194]) by usa-sjc-imap-foss1.foss.arm.com (Postfix) with ESMTPA id 858803F70D; Thu, 7 Jul 2022 00:16:06 -0700 (PDT) From: Lukasz Luba To: linux-kernel@vger.kernel.org, linux-pm@vger.kernel.org Cc: daniel.lezcano@linaro.org, amitk@kernel.org, rui.zhang@intel.com, lukasz.luba@arm.com, viresh.kumar@linaro.org, rafael@kernel.org, dietmar.eggemann@arm.com, nm@ti.com, sboyd@kernel.org, sudeep.holla@arm.com, cristian.marussi@arm.com, matthias.bgg@gmail.com, linux-mediatek@lists.infradead.org, linux-arm-kernel@lists.infradead.org Subject: [PATCH v2 1/4] PM: EM: convert power field to micro-Watts precision and align drivers Date: Thu, 7 Jul 2022 08:15:52 +0100 Message-Id: <20220707071555.10085-2-lukasz.luba@arm.com> X-Mailer: git-send-email 2.17.1 In-Reply-To: <20220707071555.10085-1-lukasz.luba@arm.com> References: <20220707071555.10085-1-lukasz.luba@arm.com> X-CRM114-Version: 20100106-BlameMichelson ( TRE 0.8.0 (BSD) ) MR-646709E3 X-CRM114-CacheID: sfid-20220707_001611_941696_28117159 X-CRM114-Status: GOOD ( 33.20 ) X-BeenThere: linux-arm-kernel@lists.infradead.org X-Mailman-Version: 2.1.34 Precedence: list List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , MIME-Version: 1.0 Sender: "linux-arm-kernel" Errors-To: linux-arm-kernel-bounces+linux-arm-kernel=archiver.kernel.org@lists.infradead.org The milli-Watts precision causes rounding errors while calculating efficiency cost for each OPP. This is especially visible in the 'simple' Energy Model (EM), where the power for each OPP is provided from OPP framework. This can cause some OPPs to be marked inefficient, while using micro-Watts precision that might not happen. Update all EM users which access 'power' field and assume the value is in milli-Watts. Solve also an issue with potential overflow in calculation of energy estimation on 32bit machine. It's needed now since the power value (thus the 'cost' as well) are higher. Example calculation which shows the rounding error and impact: power = 'dyn-power-coeff' * volt_mV * volt_mV * freq_MHz power_a_uW = (100 * 600mW * 600mW * 500MHz) / 10^6 = 18000 power_a_mW = (100 * 600mW * 600mW * 500MHz) / 10^9 = 18 power_b_uW = (100 * 605mW * 605mW * 600MHz) / 10^6 = 21961 power_b_mW = (100 * 605mW * 605mW * 600MHz) / 10^9 = 21 max_freq = 2000MHz cost_a_mW = 18 * 2000MHz/500MHz = 72 cost_a_uW = 18000 * 2000MHz/500MHz = 72000 cost_b_mW = 21 * 2000MHz/600MHz = 70 // <- artificially better cost_b_uW = 21961 * 2000MHz/600MHz = 73203 The 'cost_b_mW' (which is based on old milli-Watts) is misleadingly better that the 'cost_b_uW' (this patch uses micro-Watts) and such would have impact on the 'inefficient OPPs' information in the Cpufreq framework. This patch set removes the rounding issue. Signed-off-by: Lukasz Luba Acked-by: Daniel Lezcano Acked-by: Viresh Kumar --- drivers/cpufreq/mediatek-cpufreq-hw.c | 7 ++-- drivers/cpufreq/scmi-cpufreq.c | 6 +++ drivers/opp/of.c | 15 ++++---- drivers/powercap/dtpm_cpu.c | 5 +-- drivers/thermal/cpufreq_cooling.c | 13 ++++++- drivers/thermal/devfreq_cooling.c | 19 ++++++++-- include/linux/energy_model.h | 54 +++++++++++++++++++-------- kernel/power/energy_model.c | 24 ++++++++---- 8 files changed, 100 insertions(+), 43 deletions(-) diff --git a/drivers/cpufreq/mediatek-cpufreq-hw.c b/drivers/cpufreq/mediatek-cpufreq-hw.c index 813cccbfe934..f0e0a35c7f21 100644 --- a/drivers/cpufreq/mediatek-cpufreq-hw.c +++ b/drivers/cpufreq/mediatek-cpufreq-hw.c @@ -51,7 +51,7 @@ static const u16 cpufreq_mtk_offsets[REG_ARRAY_SIZE] = { }; static int __maybe_unused -mtk_cpufreq_get_cpu_power(struct device *cpu_dev, unsigned long *mW, +mtk_cpufreq_get_cpu_power(struct device *cpu_dev, unsigned long *uW, unsigned long *KHz) { struct mtk_cpufreq_data *data; @@ -71,8 +71,9 @@ mtk_cpufreq_get_cpu_power(struct device *cpu_dev, unsigned long *mW, i--; *KHz = data->table[i].frequency; - *mW = readl_relaxed(data->reg_bases[REG_EM_POWER_TBL] + - i * LUT_ROW_SIZE) / 1000; + /* Provide micro-Watts value to the Energy Model */ + *uW = readl_relaxed(data->reg_bases[REG_EM_POWER_TBL] + + i * LUT_ROW_SIZE); return 0; } diff --git a/drivers/cpufreq/scmi-cpufreq.c b/drivers/cpufreq/scmi-cpufreq.c index 6d2a4cf46db7..bfd35583d653 100644 --- a/drivers/cpufreq/scmi-cpufreq.c +++ b/drivers/cpufreq/scmi-cpufreq.c @@ -19,6 +19,7 @@ #include #include #include +#include struct scmi_data { int domain_id; @@ -99,6 +100,7 @@ static int __maybe_unused scmi_get_cpu_power(struct device *cpu_dev, unsigned long *power, unsigned long *KHz) { + bool power_scale_mw = perf_ops->power_scale_mw_get(ph); unsigned long Hz; int ret, domain; @@ -112,6 +114,10 @@ scmi_get_cpu_power(struct device *cpu_dev, unsigned long *power, if (ret) return ret; + /* Provide bigger resolution power to the Energy Model */ + if (power_scale_mw) + *power *= MICROWATT_PER_MILLIWATT; + /* The EM framework specifies the frequency in KHz. */ *KHz = Hz / 1000; diff --git a/drivers/opp/of.c b/drivers/opp/of.c index ea8fc9e1f7e3..74c33cebeb29 100644 --- a/drivers/opp/of.c +++ b/drivers/opp/of.c @@ -1482,12 +1482,12 @@ EXPORT_SYMBOL_GPL(dev_pm_opp_get_of_node); * It provides the power used by @dev at @kHz if it is the frequency of an * existing OPP, or at the frequency of the first OPP above @kHz otherwise * (see dev_pm_opp_find_freq_ceil()). This function updates @kHz to the ceiled - * frequency and @mW to the associated power. + * frequency and @uW to the associated power. * * Returns 0 on success or a proper -EINVAL value in case of error. */ static int __maybe_unused -_get_dt_power(struct device *dev, unsigned long *mW, unsigned long *kHz) +_get_dt_power(struct device *dev, unsigned long *uW, unsigned long *kHz) { struct dev_pm_opp *opp; unsigned long opp_freq, opp_power; @@ -1504,7 +1504,7 @@ _get_dt_power(struct device *dev, unsigned long *mW, unsigned long *kHz) return -EINVAL; *kHz = opp_freq / 1000; - *mW = opp_power / 1000; + *uW = opp_power; return 0; } @@ -1514,14 +1514,14 @@ _get_dt_power(struct device *dev, unsigned long *mW, unsigned long *kHz) * This computes the power estimated by @dev at @kHz if it is the frequency * of an existing OPP, or at the frequency of the first OPP above @kHz otherwise * (see dev_pm_opp_find_freq_ceil()). This function updates @kHz to the ceiled - * frequency and @mW to the associated power. The power is estimated as + * frequency and @uW to the associated power. The power is estimated as * P = C * V^2 * f with C being the device's capacitance and V and f * respectively the voltage and frequency of the OPP. * * Returns -EINVAL if the power calculation failed because of missing * parameters, 0 otherwise. */ -static int __maybe_unused _get_power(struct device *dev, unsigned long *mW, +static int __maybe_unused _get_power(struct device *dev, unsigned long *uW, unsigned long *kHz) { struct dev_pm_opp *opp; @@ -1551,9 +1551,10 @@ static int __maybe_unused _get_power(struct device *dev, unsigned long *mW, return -EINVAL; tmp = (u64)cap * mV * mV * (Hz / 1000000); - do_div(tmp, 1000000000); + /* Provide power in micro-Watts */ + do_div(tmp, 1000000); - *mW = (unsigned long)tmp; + *uW = (unsigned long)tmp; *kHz = Hz / 1000; return 0; diff --git a/drivers/powercap/dtpm_cpu.c b/drivers/powercap/dtpm_cpu.c index f5eced0842b3..61c5ff80bd30 100644 --- a/drivers/powercap/dtpm_cpu.c +++ b/drivers/powercap/dtpm_cpu.c @@ -53,7 +53,7 @@ static u64 set_pd_power_limit(struct dtpm *dtpm, u64 power_limit) for (i = 0; i < pd->nr_perf_states; i++) { - power = pd->table[i].power * MICROWATT_PER_MILLIWATT * nr_cpus; + power = pd->table[i].power * nr_cpus; if (power > power_limit) break; @@ -63,8 +63,7 @@ static u64 set_pd_power_limit(struct dtpm *dtpm, u64 power_limit) freq_qos_update_request(&dtpm_cpu->qos_req, freq); - power_limit = pd->table[i - 1].power * - MICROWATT_PER_MILLIWATT * nr_cpus; + power_limit = pd->table[i - 1].power * nr_cpus; return power_limit; } diff --git a/drivers/thermal/cpufreq_cooling.c b/drivers/thermal/cpufreq_cooling.c index b8151d95a806..dc19e7c80751 100644 --- a/drivers/thermal/cpufreq_cooling.c +++ b/drivers/thermal/cpufreq_cooling.c @@ -21,6 +21,7 @@ #include #include #include +#include #include @@ -101,6 +102,7 @@ static unsigned long get_level(struct cpufreq_cooling_device *cpufreq_cdev, static u32 cpu_freq_to_power(struct cpufreq_cooling_device *cpufreq_cdev, u32 freq) { + unsigned long power_mw; int i; for (i = cpufreq_cdev->max_level - 1; i >= 0; i--) { @@ -108,16 +110,23 @@ static u32 cpu_freq_to_power(struct cpufreq_cooling_device *cpufreq_cdev, break; } - return cpufreq_cdev->em->table[i + 1].power; + power_mw = cpufreq_cdev->em->table[i + 1].power; + power_mw /= MICROWATT_PER_MILLIWATT; + + return power_mw; } static u32 cpu_power_to_freq(struct cpufreq_cooling_device *cpufreq_cdev, u32 power) { + unsigned long em_power_mw; int i; for (i = cpufreq_cdev->max_level; i > 0; i--) { - if (power >= cpufreq_cdev->em->table[i].power) + /* Convert EM power to milli-Watts to make safe comparison */ + em_power_mw = cpufreq_cdev->em->table[i].power; + em_power_mw /= MICROWATT_PER_MILLIWATT; + if (power >= em_power_mw) break; } diff --git a/drivers/thermal/devfreq_cooling.c b/drivers/thermal/devfreq_cooling.c index 8c76f9655e57..8d1260f65061 100644 --- a/drivers/thermal/devfreq_cooling.c +++ b/drivers/thermal/devfreq_cooling.c @@ -200,7 +200,11 @@ static int devfreq_cooling_get_requested_power(struct thermal_cooling_device *cd res = dfc->power_ops->get_real_power(df, power, freq, voltage); if (!res) { state = dfc->capped_state; + + /* Convert EM power into milli-Watts first */ dfc->res_util = dfc->em_pd->table[state].power; + dfc->res_util /= MICROWATT_PER_MILLIWATT; + dfc->res_util *= SCALE_ERROR_MITIGATION; if (*power > 1) @@ -218,8 +222,10 @@ static int devfreq_cooling_get_requested_power(struct thermal_cooling_device *cd _normalize_load(&status); - /* Scale power for utilization */ + /* Convert EM power into milli-Watts first */ *power = dfc->em_pd->table[perf_idx].power; + *power /= MICROWATT_PER_MILLIWATT; + /* Scale power for utilization */ *power *= status.busy_time; *power >>= 10; } @@ -244,6 +250,7 @@ static int devfreq_cooling_state2power(struct thermal_cooling_device *cdev, perf_idx = dfc->max_state - state; *power = dfc->em_pd->table[perf_idx].power; + *power /= MICROWATT_PER_MILLIWATT; return 0; } @@ -254,7 +261,7 @@ static int devfreq_cooling_power2state(struct thermal_cooling_device *cdev, struct devfreq_cooling_device *dfc = cdev->devdata; struct devfreq *df = dfc->devfreq; struct devfreq_dev_status status; - unsigned long freq; + unsigned long freq, em_power_mw; s32 est_power; int i; @@ -279,9 +286,13 @@ static int devfreq_cooling_power2state(struct thermal_cooling_device *cdev, * Find the first cooling state that is within the power * budget. The EM power table is sorted ascending. */ - for (i = dfc->max_state; i > 0; i--) - if (est_power >= dfc->em_pd->table[i].power) + for (i = dfc->max_state; i > 0; i--) { + /* Convert EM power to milli-Watts to make safe comparison */ + em_power_mw = dfc->em_pd->table[i].power; + em_power_mw /= MICROWATT_PER_MILLIWATT; + if (est_power >= em_power_mw) break; + } *state = dfc->max_state - i; dfc->capped_state = *state; diff --git a/include/linux/energy_model.h b/include/linux/energy_model.h index 8419bffb4398..b9caa01dfac4 100644 --- a/include/linux/energy_model.h +++ b/include/linux/energy_model.h @@ -62,7 +62,7 @@ struct em_perf_domain { /* * em_perf_domain flags: * - * EM_PERF_DOMAIN_MILLIWATTS: The power values are in milli-Watts or some + * EM_PERF_DOMAIN_MICROWATTS: The power values are in micro-Watts or some * other scale. * * EM_PERF_DOMAIN_SKIP_INEFFICIENCIES: Skip inefficient states when estimating @@ -71,7 +71,7 @@ struct em_perf_domain { * EM_PERF_DOMAIN_ARTIFICIAL: The power values are artificial and might be * created by platform missing real power information */ -#define EM_PERF_DOMAIN_MILLIWATTS BIT(0) +#define EM_PERF_DOMAIN_MICROWATTS BIT(0) #define EM_PERF_DOMAIN_SKIP_INEFFICIENCIES BIT(1) #define EM_PERF_DOMAIN_ARTIFICIAL BIT(2) @@ -79,22 +79,44 @@ struct em_perf_domain { #define em_is_artificial(em) ((em)->flags & EM_PERF_DOMAIN_ARTIFICIAL) #ifdef CONFIG_ENERGY_MODEL -#define EM_MAX_POWER 0xFFFF +/* + * The max power value in micro-Watts. The limit of 64 Watts is set as + * a safety net to not overflow multiplications on 32bit platforms. The + * 32bit value limit for total Perf Domain power implies a limit of + * maximum CPUs in such domain to 64. + */ +#define EM_MAX_POWER (64000000) /* 64 Watts */ + +/* + * To avoid possible energy estimation overflow on 32bit machines add + * limits to number of CPUs in the Perf. Domain. + * We are safe on 64bit machine, thus some big number. + */ +#ifdef CONFIG_64BIT +#define EM_MAX_NUM_CPUS 4096 +#else +#define EM_MAX_NUM_CPUS 16 +#endif /* - * Increase resolution of energy estimation calculations for 64-bit - * architectures. The extra resolution improves decision made by EAS for the - * task placement when two Performance Domains might provide similar energy - * estimation values (w/o better resolution the values could be equal). + * To avoid an overflow on 32bit machines while calculating the energy + * use a different order in the operation. First divide by the 'cpu_scale' + * which would reduce big value stored in the 'cost' field, then multiply by + * the 'sum_util'. This would allow to handle existing platforms, which have + * e.g. power ~1.3 Watt at max freq, so the 'cost' value > 1mln micro-Watts. + * In such scenario, where there are 4 CPUs in the Perf. Domain the 'sum_util' + * could be 4096, then multiplication: 'cost' * 'sum_util' would overflow. + * This reordering of operations has some limitations, we lose small + * precision in the estimation (comparing to 64bit platform w/o reordering). * - * We increase resolution only if we have enough bits to allow this increased - * resolution (i.e. 64-bit). The costs for increasing resolution when 32-bit - * are pretty high and the returns do not justify the increased costs. + * We are safe on 64bit machine. */ #ifdef CONFIG_64BIT -#define em_scale_power(p) ((p) * 1000) +#define em_estimate_energy(cost, sum_util, scale_cpu) \ + (((cost) * (sum_util)) / (scale_cpu)) #else -#define em_scale_power(p) (p) +#define em_estimate_energy(cost, sum_util, scale_cpu) \ + (((cost) / (scale_cpu)) * (sum_util)) #endif struct em_data_callback { @@ -112,7 +134,7 @@ struct em_data_callback { * and frequency. * * In case of CPUs, the power is the one of a single CPU in the domain, - * expressed in milli-Watts or an abstract scale. It is expected to + * expressed in micro-Watts or an abstract scale. It is expected to * fit in the [0, EM_MAX_POWER] range. * * Return 0 on success. @@ -148,7 +170,7 @@ struct em_perf_domain *em_cpu_get(int cpu); struct em_perf_domain *em_pd_get(struct device *dev); int em_dev_register_perf_domain(struct device *dev, unsigned int nr_states, struct em_data_callback *cb, cpumask_t *span, - bool milliwatts); + bool microwatts); void em_dev_unregister_perf_domain(struct device *dev); /** @@ -273,7 +295,7 @@ static inline unsigned long em_cpu_energy(struct em_perf_domain *pd, * pd_nrg = ------------------------ (4) * scale_cpu */ - return ps->cost * sum_util / scale_cpu; + return em_estimate_energy(ps->cost, sum_util, scale_cpu); } /** @@ -297,7 +319,7 @@ struct em_data_callback {}; static inline int em_dev_register_perf_domain(struct device *dev, unsigned int nr_states, struct em_data_callback *cb, cpumask_t *span, - bool milliwatts) + bool microwatts) { return -EINVAL; } diff --git a/kernel/power/energy_model.c b/kernel/power/energy_model.c index 6c373f2960e7..f82111837b8d 100644 --- a/kernel/power/energy_model.c +++ b/kernel/power/energy_model.c @@ -145,7 +145,7 @@ static int em_create_perf_table(struct device *dev, struct em_perf_domain *pd, /* * The power returned by active_state() is expected to be - * positive and to fit into 16 bits. + * positive and be in range. */ if (!power || power > EM_MAX_POWER) { dev_err(dev, "EM: invalid power: %lu\n", @@ -170,7 +170,7 @@ static int em_create_perf_table(struct device *dev, struct em_perf_domain *pd, goto free_ps_table; } } else { - power_res = em_scale_power(table[i].power); + power_res = table[i].power; cost = div64_u64(fmax * power_res, table[i].frequency); } @@ -201,9 +201,17 @@ static int em_create_pd(struct device *dev, int nr_states, { struct em_perf_domain *pd; struct device *cpu_dev; - int cpu, ret; + int cpu, ret, num_cpus; if (_is_cpu_device(dev)) { + num_cpus = cpumask_weight(cpus); + + /* Prevent max possible energy calculation to not overflow */ + if (num_cpus > EM_MAX_NUM_CPUS) { + dev_err(dev, "EM: too many CPUs, overflow possible\n"); + return -EINVAL; + } + pd = kzalloc(sizeof(*pd) + cpumask_size(), GFP_KERNEL); if (!pd) return -ENOMEM; @@ -314,13 +322,13 @@ EXPORT_SYMBOL_GPL(em_cpu_get); * @cpus : Pointer to cpumask_t, which in case of a CPU device is * obligatory. It can be taken from i.e. 'policy->cpus'. For other * type of devices this should be set to NULL. - * @milliwatts : Flag indicating that the power values are in milliWatts or + * @microwatts : Flag indicating that the power values are in micro-Watts or * in some other scale. It must be set properly. * * Create Energy Model tables for a performance domain using the callbacks * defined in cb. * - * The @milliwatts is important to set with correct value. Some kernel + * The @microwatts is important to set with correct value. Some kernel * sub-systems might rely on this flag and check if all devices in the EM are * using the same scale. * @@ -331,7 +339,7 @@ EXPORT_SYMBOL_GPL(em_cpu_get); */ int em_dev_register_perf_domain(struct device *dev, unsigned int nr_states, struct em_data_callback *cb, cpumask_t *cpus, - bool milliwatts) + bool microwatts) { unsigned long cap, prev_cap = 0; unsigned long flags = 0; @@ -381,8 +389,8 @@ int em_dev_register_perf_domain(struct device *dev, unsigned int nr_states, } } - if (milliwatts) - flags |= EM_PERF_DOMAIN_MILLIWATTS; + if (microwatts) + flags |= EM_PERF_DOMAIN_MICROWATTS; else if (cb->get_cost) flags |= EM_PERF_DOMAIN_ARTIFICIAL; From patchwork Thu Jul 7 07:15:53 2022 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Lukasz Luba X-Patchwork-Id: 12909131 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org Received: from bombadil.infradead.org (bombadil.infradead.org [198.137.202.133]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by smtp.lore.kernel.org (Postfix) with ESMTPS id 1E124C433EF for ; Thu, 7 Jul 2022 07:18:03 +0000 (UTC) DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=lists.infradead.org; s=bombadil.20210309; h=Sender: Content-Transfer-Encoding:Content-Type:MIME-Version:List-Subscribe:List-Help: List-Post:List-Archive:List-Unsubscribe:List-Id:References:In-Reply-To: Message-Id:Date:Subject:Cc:To:From:Reply-To:Content-ID:Content-Description: Resent-Date:Resent-From:Resent-Sender:Resent-To:Resent-Cc:Resent-Message-ID: List-Owner; bh=Bv1YM7Of8YFzKJbeTz4dEOQ7Nj/5+efwyVaWGw2E4Yc=; b=0qzliIu39GtpC8 cR52uRkQuDp7u2s1zOaUG/vJJuc1Zrdo1imbrPKSns9LV9GBCxkw5Di+6SxE/WE4vZx56l5ZyT/uD GZs6h/ysvDRk/0TS3Q9WbE0EkNu9XWJ8mXX24Ff1fsh3DU6NSZOxTRXalRtWTzlPwDu2ZymazVz/1 bCQGs2nK0O9PJSh2abG/h0pzo4rHAtzudXfPkY4GXl0y4XFb9KUCb3Z7vfWhXI1xkN654U8JPrbZM U7J1+8ePh6dOpGKrDXOqTJj2dfmGM0ESptv0fFbOYgWxq8gQRnNlbkbnLtuS4WY9yhdAtvgu8EVXj 76UAnMJg7ZD7NwJvTYTA==; Received: from localhost ([::1] helo=bombadil.infradead.org) by bombadil.infradead.org with esmtp (Exim 4.94.2 #2 (Red Hat Linux)) id 1o9Lkx-00ETu2-57; Thu, 07 Jul 2022 07:16:59 +0000 Received: from foss.arm.com ([217.140.110.172]) by bombadil.infradead.org with esmtp (Exim 4.94.2 #2 (Red Hat Linux)) id 1o9LkF-00ETdx-5m; Thu, 07 Jul 2022 07:16:17 +0000 Received: from usa-sjc-imap-foss1.foss.arm.com (unknown [10.121.207.14]) by usa-sjc-mx-foss1.foss.arm.com (Postfix) with ESMTP id 21360113E; Thu, 7 Jul 2022 00:16:13 -0700 (PDT) Received: from e123648.arm.com (unknown [10.57.11.194]) by usa-sjc-imap-foss1.foss.arm.com (Postfix) with ESMTPA id DFAC73F70D; Thu, 7 Jul 2022 00:16:09 -0700 (PDT) From: Lukasz Luba To: linux-kernel@vger.kernel.org, linux-pm@vger.kernel.org Cc: daniel.lezcano@linaro.org, amitk@kernel.org, rui.zhang@intel.com, lukasz.luba@arm.com, viresh.kumar@linaro.org, rafael@kernel.org, dietmar.eggemann@arm.com, nm@ti.com, sboyd@kernel.org, sudeep.holla@arm.com, cristian.marussi@arm.com, matthias.bgg@gmail.com, linux-mediatek@lists.infradead.org, linux-arm-kernel@lists.infradead.org Subject: [PATCH v2 2/4] Documentation: EM: Switch to micro-Watts scale Date: Thu, 7 Jul 2022 08:15:53 +0100 Message-Id: <20220707071555.10085-3-lukasz.luba@arm.com> X-Mailer: git-send-email 2.17.1 In-Reply-To: <20220707071555.10085-1-lukasz.luba@arm.com> References: <20220707071555.10085-1-lukasz.luba@arm.com> X-CRM114-Version: 20100106-BlameMichelson ( TRE 0.8.0 (BSD) ) MR-646709E3 X-CRM114-CacheID: sfid-20220707_001615_371388_DC6FF347 X-CRM114-Status: GOOD ( 15.58 ) X-BeenThere: linux-arm-kernel@lists.infradead.org X-Mailman-Version: 2.1.34 Precedence: list List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , MIME-Version: 1.0 Sender: "linux-arm-kernel" Errors-To: linux-arm-kernel-bounces+linux-arm-kernel=archiver.kernel.org@lists.infradead.org The EM now uses the micro-Watts scale for the power values. Update related documentation to reflect that fact. Fix also a problematic sentence in the doc "to:" which triggers test scripts complaining about wrong email address. Reviewed-by: Daniel Lezcano Signed-off-by: Lukasz Luba --- Documentation/power/energy-model.rst | 14 +++++++------- 1 file changed, 7 insertions(+), 7 deletions(-) diff --git a/Documentation/power/energy-model.rst b/Documentation/power/energy-model.rst index feb257b7f350..ef341be2882b 100644 --- a/Documentation/power/energy-model.rst +++ b/Documentation/power/energy-model.rst @@ -20,20 +20,20 @@ possible source of information on its own, the EM framework intervenes as an abstraction layer which standardizes the format of power cost tables in the kernel, hence enabling to avoid redundant work. -The power values might be expressed in milli-Watts or in an 'abstract scale'. +The power values might be expressed in micro-Watts or in an 'abstract scale'. Multiple subsystems might use the EM and it is up to the system integrator to check that the requirements for the power value scale types are met. An example can be found in the Energy-Aware Scheduler documentation Documentation/scheduler/sched-energy.rst. For some subsystems like thermal or powercap power values expressed in an 'abstract scale' might cause issues. These subsystems are more interested in estimation of power used in the past, -thus the real milli-Watts might be needed. An example of these requirements can +thus the real micro-Watts might be needed. An example of these requirements can be found in the Intelligent Power Allocation in Documentation/driver-api/thermal/power_allocator.rst. Kernel subsystems might implement automatic detection to check whether EM registered devices have inconsistent scale (based on EM internal flag). Important thing to keep in mind is that when the power values are expressed in -an 'abstract scale' deriving real energy in milli-Joules would not be possible. +an 'abstract scale' deriving real energy in micro-Joules would not be possible. The figure below depicts an example of drivers (Arm-specific here, but the approach is applicable to any architecture) providing power costs to the EM @@ -98,7 +98,7 @@ Drivers are expected to register performance domains into the EM framework by calling the following API:: int em_dev_register_perf_domain(struct device *dev, unsigned int nr_states, - struct em_data_callback *cb, cpumask_t *cpus, bool milliwatts); + struct em_data_callback *cb, cpumask_t *cpus, bool microwatts); Drivers must provide a callback function returning tuples for each performance state. The callback function provided by the driver is free @@ -106,10 +106,10 @@ to fetch data from any relevant location (DT, firmware, ...), and by any mean deemed necessary. Only for CPU devices, drivers must specify the CPUs of the performance domains using cpumask. For other devices than CPUs the last argument must be set to NULL. -The last argument 'milliwatts' is important to set with correct value. Kernel +The last argument 'microwatts' is important to set with correct value. Kernel subsystems which use EM might rely on this flag to check if all EM devices use the same scale. If there are different scales, these subsystems might decide -to: return warning/error, stop working or panic. +to return warning/error, stop working or panic. See Section 3. for an example of driver implementing this callback, or Section 2.4 for further documentation on this API @@ -137,7 +137,7 @@ The .get_cost() allows to provide the 'cost' values which reflect the efficiency of the CPUs. This would allow to provide EAS information which has different relation than what would be forced by the EM internal formulas calculating 'cost' values. To register an EM for such platform, the -driver must set the flag 'milliwatts' to 0, provide .get_power() callback +driver must set the flag 'microwatts' to 0, provide .get_power() callback and provide .get_cost() callback. The EM framework would handle such platform properly during registration. A flag EM_PERF_DOMAIN_ARTIFICIAL is set for such platform. Special care should be taken by other frameworks which are using EM From patchwork Thu Jul 7 07:15:54 2022 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Lukasz Luba X-Patchwork-Id: 12909145 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org Received: from bombadil.infradead.org (bombadil.infradead.org [198.137.202.133]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by smtp.lore.kernel.org (Postfix) with ESMTPS id 084ACC43334 for ; Thu, 7 Jul 2022 07:18:48 +0000 (UTC) DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=lists.infradead.org; s=bombadil.20210309; h=Sender: Content-Transfer-Encoding:Content-Type:MIME-Version:List-Subscribe:List-Help: List-Post:List-Archive:List-Unsubscribe:List-Id:References:In-Reply-To: Message-Id:Date:Subject:Cc:To:From:Reply-To:Content-ID:Content-Description: Resent-Date:Resent-From:Resent-Sender:Resent-To:Resent-Cc:Resent-Message-ID: List-Owner; bh=Jii/jZgAGorJSeZB7GTKvsPrlyIRhtxFcYVWiG3M0VI=; b=DfORRp7pByNjH2 t1oDMQdSU4ULkMPcgNysKI2saNY/OmGHEoVVFZsEsXS8kQZ/HYs+sELSuHSJ3PFlT/dKCEZYLfASF NsbuJnZIld6yrgDnm8bL101+TvAT7TzxwKEykn5YsnSVWr5rZKucoQfhkN2QPIfGImeJft5wxss5W 21dijZ+ku5WtGRN8Llr3Nhgx6D3frVzqGhEf7tkCDaRF6/XZO5XDkiAAFYcfftUo1wvEpjEGTTlYZ gMFWUKdMcXqwbKFBYvE3Zc3AWx7+hDnc2vDutE6avUYdFUFB47ts0AWpyhNkkwtcPP/44aW2GtGSZ Ijqp2YamWviQW+/umCTw==; Received: from localhost ([::1] helo=bombadil.infradead.org) by bombadil.infradead.org with esmtp (Exim 4.94.2 #2 (Red Hat Linux)) id 1o9Llg-00EUDm-Ik; Thu, 07 Jul 2022 07:17:44 +0000 Received: from foss.arm.com ([217.140.110.172]) by bombadil.infradead.org with esmtp (Exim 4.94.2 #2 (Red Hat Linux)) id 1o9LkH-00ETfJ-FP; Thu, 07 Jul 2022 07:16:20 +0000 Received: from usa-sjc-imap-foss1.foss.arm.com (unknown [10.121.207.14]) by usa-sjc-mx-foss1.foss.arm.com (Postfix) with ESMTP id 546F6D6E; Thu, 7 Jul 2022 00:16:16 -0700 (PDT) Received: from e123648.arm.com (unknown [10.57.11.194]) by usa-sjc-imap-foss1.foss.arm.com (Postfix) with ESMTPA id 1AF3A3F70D; Thu, 7 Jul 2022 00:16:12 -0700 (PDT) From: Lukasz Luba To: linux-kernel@vger.kernel.org, linux-pm@vger.kernel.org Cc: daniel.lezcano@linaro.org, amitk@kernel.org, rui.zhang@intel.com, lukasz.luba@arm.com, viresh.kumar@linaro.org, rafael@kernel.org, dietmar.eggemann@arm.com, nm@ti.com, sboyd@kernel.org, sudeep.holla@arm.com, cristian.marussi@arm.com, matthias.bgg@gmail.com, linux-mediatek@lists.infradead.org, linux-arm-kernel@lists.infradead.org Subject: [PATCH v2 3/4] firmware: arm_scmi: Get detailed power scale from perf Date: Thu, 7 Jul 2022 08:15:54 +0100 Message-Id: <20220707071555.10085-4-lukasz.luba@arm.com> X-Mailer: git-send-email 2.17.1 In-Reply-To: <20220707071555.10085-1-lukasz.luba@arm.com> References: <20220707071555.10085-1-lukasz.luba@arm.com> X-CRM114-Version: 20100106-BlameMichelson ( TRE 0.8.0 (BSD) ) MR-646709E3 X-CRM114-CacheID: sfid-20220707_001617_686084_3ECFA980 X-CRM114-Status: GOOD ( 12.68 ) X-BeenThere: linux-arm-kernel@lists.infradead.org X-Mailman-Version: 2.1.34 Precedence: list List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , MIME-Version: 1.0 Sender: "linux-arm-kernel" Errors-To: linux-arm-kernel-bounces+linux-arm-kernel=archiver.kernel.org@lists.infradead.org In SCMI v3.1 the power scale can be in micro-Watts. The upper layers, e.g. cpufreq and EM should handle received power values properly (upscale when needed). Thus, provide an interface which allows to check what is the scale for power values. The old interface allowed to distinguish between bogo-Watts and milli-Watts only (which was good for older SCMI spec). Acked-by: Sudeep Holla Signed-off-by: Lukasz Luba --- drivers/firmware/arm_scmi/perf.c | 18 +++++++++++------- include/linux/scmi_protocol.h | 8 +++++++- 2 files changed, 18 insertions(+), 8 deletions(-) diff --git a/drivers/firmware/arm_scmi/perf.c b/drivers/firmware/arm_scmi/perf.c index bbb0331801ff..92414e53f908 100644 --- a/drivers/firmware/arm_scmi/perf.c +++ b/drivers/firmware/arm_scmi/perf.c @@ -170,8 +170,7 @@ struct perf_dom_info { struct scmi_perf_info { u32 version; int num_domains; - bool power_scale_mw; - bool power_scale_uw; + enum scmi_power_scale power_scale; u64 stats_addr; u32 stats_size; struct perf_dom_info *dom_info; @@ -201,9 +200,13 @@ static int scmi_perf_attributes_get(const struct scmi_protocol_handle *ph, u16 flags = le16_to_cpu(attr->flags); pi->num_domains = le16_to_cpu(attr->num_domains); - pi->power_scale_mw = POWER_SCALE_IN_MILLIWATT(flags); + + if (POWER_SCALE_IN_MILLIWATT(flags)) + pi->power_scale = SCMI_POWER_MILLIWATTS; if (PROTOCOL_REV_MAJOR(pi->version) >= 0x3) - pi->power_scale_uw = POWER_SCALE_IN_MICROWATT(flags); + if (POWER_SCALE_IN_MICROWATT(flags)) + pi->power_scale = SCMI_POWER_MICROWATTS; + pi->stats_addr = le32_to_cpu(attr->stats_addr_low) | (u64)le32_to_cpu(attr->stats_addr_high) << 32; pi->stats_size = le32_to_cpu(attr->stats_size); @@ -792,11 +795,12 @@ static bool scmi_fast_switch_possible(const struct scmi_protocol_handle *ph, return dom->fc_info && dom->fc_info->level_set_addr; } -static bool scmi_power_scale_mw_get(const struct scmi_protocol_handle *ph) +static enum scmi_power_scale +scmi_power_scale_get(const struct scmi_protocol_handle *ph) { struct scmi_perf_info *pi = ph->get_priv(ph); - return pi->power_scale_mw; + return pi->power_scale; } static const struct scmi_perf_proto_ops perf_proto_ops = { @@ -811,7 +815,7 @@ static const struct scmi_perf_proto_ops perf_proto_ops = { .freq_get = scmi_dvfs_freq_get, .est_power_get = scmi_dvfs_est_power_get, .fast_switch_possible = scmi_fast_switch_possible, - .power_scale_mw_get = scmi_power_scale_mw_get, + .power_scale_get = scmi_power_scale_get, }; static int scmi_perf_set_notify_enabled(const struct scmi_protocol_handle *ph, diff --git a/include/linux/scmi_protocol.h b/include/linux/scmi_protocol.h index 704111f63993..a0a246310ba1 100644 --- a/include/linux/scmi_protocol.h +++ b/include/linux/scmi_protocol.h @@ -60,6 +60,12 @@ struct scmi_clock_info { }; }; +enum scmi_power_scale { + SCMI_POWER_BOGOWATTS, + SCMI_POWER_MILLIWATTS, + SCMI_POWER_MICROWATTS +}; + struct scmi_handle; struct scmi_device; struct scmi_protocol_handle; @@ -135,7 +141,7 @@ struct scmi_perf_proto_ops { unsigned long *rate, unsigned long *power); bool (*fast_switch_possible)(const struct scmi_protocol_handle *ph, struct device *dev); - bool (*power_scale_mw_get)(const struct scmi_protocol_handle *ph); + enum scmi_power_scale (*power_scale_get)(const struct scmi_protocol_handle *ph); }; /** From patchwork Thu Jul 7 07:15:55 2022 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Lukasz Luba X-Patchwork-Id: 12909146 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org Received: from bombadil.infradead.org (bombadil.infradead.org [198.137.202.133]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by smtp.lore.kernel.org (Postfix) with ESMTPS id 10D74C43334 for ; Thu, 7 Jul 2022 07:19:10 +0000 (UTC) DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=lists.infradead.org; s=bombadil.20210309; h=Sender: Content-Transfer-Encoding:Content-Type:MIME-Version:List-Subscribe:List-Help: List-Post:List-Archive:List-Unsubscribe:List-Id:References:In-Reply-To: Message-Id:Date:Subject:Cc:To:From:Reply-To:Content-ID:Content-Description: Resent-Date:Resent-From:Resent-Sender:Resent-To:Resent-Cc:Resent-Message-ID: List-Owner; bh=BJTGA3fj4CkY9yAcOx8U3SExPBEhLUthUQPmYl8eMZ4=; b=LUCA9uVMPRD/b1 F9uWgIgUAedUaFSc/MlAlVHHi6k/KAgq9cNOal4VgewVMJ6gb2FsbAiBYpcQj+IaZ0ifWtgmlve/N ffo21+CVSJAzYQosthGMCHkArS+6tDYFfCy6aKh2NHp11EFN6lGd0sLsl38HrHj5omxMFvylMvm0N RpDVmq9APe78FeWYVDw/Btz/krt+WC740ngZJvVsrmiTIU3gh8mEV3UqdnHi+eON83fxn8zgmT1Sv oUbZ1UZKBvkF8Ic4gvFWbF5Hux94oiK5AuAZsXHnZHVUyxDNzJIlm80Pkt6n/BCb++rBPNn8UJDq5 gVNthEwJum2vxAHk3WzQ==; Received: from localhost ([::1] helo=bombadil.infradead.org) by bombadil.infradead.org with esmtp (Exim 4.94.2 #2 (Red Hat Linux)) id 1o9Lm4-00EUS1-2Q; Thu, 07 Jul 2022 07:18:08 +0000 Received: from foss.arm.com ([217.140.110.172]) by bombadil.infradead.org with esmtp (Exim 4.94.2 #2 (Red Hat Linux)) id 1o9LkK-00ETgM-Dq; Thu, 07 Jul 2022 07:16:22 +0000 Received: from usa-sjc-imap-foss1.foss.arm.com (unknown [10.121.207.14]) by usa-sjc-mx-foss1.foss.arm.com (Postfix) with ESMTP id 85EB2106F; Thu, 7 Jul 2022 00:16:19 -0700 (PDT) Received: from e123648.arm.com (unknown [10.57.11.194]) by usa-sjc-imap-foss1.foss.arm.com (Postfix) with ESMTPA id 503463F70D; Thu, 7 Jul 2022 00:16:16 -0700 (PDT) From: Lukasz Luba To: linux-kernel@vger.kernel.org, linux-pm@vger.kernel.org Cc: daniel.lezcano@linaro.org, amitk@kernel.org, rui.zhang@intel.com, lukasz.luba@arm.com, viresh.kumar@linaro.org, rafael@kernel.org, dietmar.eggemann@arm.com, nm@ti.com, sboyd@kernel.org, sudeep.holla@arm.com, cristian.marussi@arm.com, matthias.bgg@gmail.com, linux-mediatek@lists.infradead.org, linux-arm-kernel@lists.infradead.org Subject: [PATCH v2 4/4] cpufreq: scmi: Support the power scale in micro-Watts in SCMI v3.1 Date: Thu, 7 Jul 2022 08:15:55 +0100 Message-Id: <20220707071555.10085-5-lukasz.luba@arm.com> X-Mailer: git-send-email 2.17.1 In-Reply-To: <20220707071555.10085-1-lukasz.luba@arm.com> References: <20220707071555.10085-1-lukasz.luba@arm.com> X-CRM114-Version: 20100106-BlameMichelson ( TRE 0.8.0 (BSD) ) MR-646709E3 X-CRM114-CacheID: sfid-20220707_001620_577131_A029EB68 X-CRM114-Status: GOOD ( 12.36 ) X-BeenThere: linux-arm-kernel@lists.infradead.org X-Mailman-Version: 2.1.34 Precedence: list List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , MIME-Version: 1.0 Sender: "linux-arm-kernel" Errors-To: linux-arm-kernel-bounces+linux-arm-kernel=archiver.kernel.org@lists.infradead.org The SCMI v3.1 adds support for power values in micro-Watts. They are not always in milli-Watts anymore (ignoring the bogo-Watts). Thus, the power must be converted conditionally before sending to Energy Model. Add the logic which handles the needed checks and conversions. Acked-by: Sudeep Holla Signed-off-by: Lukasz Luba Acked-by: Viresh Kumar --- drivers/cpufreq/scmi-cpufreq.c | 15 ++++++++++----- 1 file changed, 10 insertions(+), 5 deletions(-) diff --git a/drivers/cpufreq/scmi-cpufreq.c b/drivers/cpufreq/scmi-cpufreq.c index bfd35583d653..513a071845c2 100644 --- a/drivers/cpufreq/scmi-cpufreq.c +++ b/drivers/cpufreq/scmi-cpufreq.c @@ -100,7 +100,7 @@ static int __maybe_unused scmi_get_cpu_power(struct device *cpu_dev, unsigned long *power, unsigned long *KHz) { - bool power_scale_mw = perf_ops->power_scale_mw_get(ph); + enum scmi_power_scale power_scale = perf_ops->power_scale_get(ph); unsigned long Hz; int ret, domain; @@ -114,8 +114,8 @@ scmi_get_cpu_power(struct device *cpu_dev, unsigned long *power, if (ret) return ret; - /* Provide bigger resolution power to the Energy Model */ - if (power_scale_mw) + /* Convert the power to uW if it is mW (ignore bogoW) */ + if (power_scale == SCMI_POWER_MILLIWATTS) *power *= MICROWATT_PER_MILLIWATT; /* The EM framework specifies the frequency in KHz. */ @@ -255,8 +255,9 @@ static int scmi_cpufreq_exit(struct cpufreq_policy *policy) static void scmi_cpufreq_register_em(struct cpufreq_policy *policy) { struct em_data_callback em_cb = EM_DATA_CB(scmi_get_cpu_power); - bool power_scale_mw = perf_ops->power_scale_mw_get(ph); + enum scmi_power_scale power_scale = perf_ops->power_scale_get(ph); struct scmi_data *priv = policy->driver_data; + bool em_power_scale = false; /* * This callback will be called for each policy, but we don't need to @@ -268,9 +269,13 @@ static void scmi_cpufreq_register_em(struct cpufreq_policy *policy) if (!priv->nr_opp) return; + if (power_scale == SCMI_POWER_MILLIWATTS + || power_scale == SCMI_POWER_MICROWATTS) + em_power_scale = true; + em_dev_register_perf_domain(get_cpu_device(policy->cpu), priv->nr_opp, &em_cb, priv->opp_shared_cpus, - power_scale_mw); + em_power_scale); } static struct cpufreq_driver scmi_cpufreq_driver = {