From 045149e6a22119e5bf0d16a0b24a4173a2abb71d Mon Sep 17 00:00:00 2001 From: "Rafael J. Wysocki" Date: Thu, 23 Nov 2017 01:23:16 +0100 Subject: cpufreq: Clean up cpufreq_parse_governor() Drop an unnecessary local variable from cpufreq_parse_governor() and rearrange the code in there to make it easier to follow. Signed-off-by: Rafael J. Wysocki Acked-by: Viresh Kumar Signed-off-by: Rafael J. Wysocki --- drivers/cpufreq/cpufreq.c | 32 +++++++++++++++++--------------- 1 file changed, 17 insertions(+), 15 deletions(-) diff --git a/drivers/cpufreq/cpufreq.c b/drivers/cpufreq/cpufreq.c index 41d148af7748..4d76b7c57b7a 100644 --- a/drivers/cpufreq/cpufreq.c +++ b/drivers/cpufreq/cpufreq.c @@ -604,16 +604,15 @@ static struct cpufreq_governor *find_governor(const char *str_governor) static int cpufreq_parse_governor(char *str_governor, unsigned int *policy, struct cpufreq_governor **governor) { - int err = -EINVAL; - if (cpufreq_driver->setpolicy) { if (!strncasecmp(str_governor, "performance", CPUFREQ_NAME_LEN)) { *policy = CPUFREQ_POLICY_PERFORMANCE; - err = 0; - } else if (!strncasecmp(str_governor, "powersave", - CPUFREQ_NAME_LEN)) { + return 0; + } + + if (!strncasecmp(str_governor, "powersave", CPUFREQ_NAME_LEN)) { *policy = CPUFREQ_POLICY_POWERSAVE; - err = 0; + return 0; } } else { struct cpufreq_governor *t; @@ -621,26 +620,29 @@ static int cpufreq_parse_governor(char *str_governor, unsigned int *policy, mutex_lock(&cpufreq_governor_mutex); t = find_governor(str_governor); - - if (t == NULL) { + if (!t) { int ret; mutex_unlock(&cpufreq_governor_mutex); + ret = request_module("cpufreq_%s", str_governor); + if (ret) + return -EINVAL; + mutex_lock(&cpufreq_governor_mutex); - if (ret == 0) - t = find_governor(str_governor); + t = find_governor(str_governor); } - if (t != NULL) { + mutex_unlock(&cpufreq_governor_mutex); + + if (t) { *governor = t; - err = 0; + return 0; } - - mutex_unlock(&cpufreq_governor_mutex); } - return err; + + return -EINVAL; } /** -- cgit 1.4.1 From ae0ff89f36b282ef32ff0f73e847352ea625464c Mon Sep 17 00:00:00 2001 From: "Rafael J. Wysocki" Date: Thu, 23 Nov 2017 01:24:05 +0100 Subject: cpufreq: Pass policy pointer to cpufreq_parse_governor() Pass policy pointer to cpufreq_parse_governor() instead of passing pointers to two members of it so as to make the code slightly more straightforward. Signed-off-by: Rafael J. Wysocki Acked-by: Viresh Kumar Signed-off-by: Rafael J. Wysocki --- drivers/cpufreq/cpufreq.c | 16 +++++++--------- 1 file changed, 7 insertions(+), 9 deletions(-) diff --git a/drivers/cpufreq/cpufreq.c b/drivers/cpufreq/cpufreq.c index 4d76b7c57b7a..8f356c4befda 100644 --- a/drivers/cpufreq/cpufreq.c +++ b/drivers/cpufreq/cpufreq.c @@ -601,17 +601,17 @@ static struct cpufreq_governor *find_governor(const char *str_governor) /** * cpufreq_parse_governor - parse a governor string */ -static int cpufreq_parse_governor(char *str_governor, unsigned int *policy, - struct cpufreq_governor **governor) +static int cpufreq_parse_governor(char *str_governor, + struct cpufreq_policy *policy) { if (cpufreq_driver->setpolicy) { if (!strncasecmp(str_governor, "performance", CPUFREQ_NAME_LEN)) { - *policy = CPUFREQ_POLICY_PERFORMANCE; + policy->policy = CPUFREQ_POLICY_PERFORMANCE; return 0; } if (!strncasecmp(str_governor, "powersave", CPUFREQ_NAME_LEN)) { - *policy = CPUFREQ_POLICY_POWERSAVE; + policy->policy = CPUFREQ_POLICY_POWERSAVE; return 0; } } else { @@ -637,7 +637,7 @@ static int cpufreq_parse_governor(char *str_governor, unsigned int *policy, mutex_unlock(&cpufreq_governor_mutex); if (t) { - *governor = t; + policy->governor = t; return 0; } } @@ -762,8 +762,7 @@ static ssize_t store_scaling_governor(struct cpufreq_policy *policy, if (ret != 1) return -EINVAL; - if (cpufreq_parse_governor(str_governor, &new_policy.policy, - &new_policy.governor)) + if (cpufreq_parse_governor(str_governor, &new_policy)) return -EINVAL; ret = cpufreq_set_policy(policy, &new_policy); @@ -1046,8 +1045,7 @@ static int cpufreq_init_policy(struct cpufreq_policy *policy) if (policy->last_policy) new_policy.policy = policy->last_policy; else - cpufreq_parse_governor(gov->name, &new_policy.policy, - NULL); + cpufreq_parse_governor(gov->name, &new_policy); } /* set default policy */ return cpufreq_set_policy(policy, &new_policy); -- cgit 1.4.1 From 70d1ff71161b1c56c6d025e6a957bc878dfd940b Mon Sep 17 00:00:00 2001 From: "Rafael J. Wysocki" Date: Thu, 23 Nov 2017 01:30:16 +0100 Subject: cpufreq: Drop pointless return statement Drop a pointless return statement from cpufreq_unregister_governor(). Signed-off-by: Rafael J. Wysocki Acked-by: Viresh Kumar Signed-off-by: Rafael J. Wysocki --- drivers/cpufreq/cpufreq.c | 1 - 1 file changed, 1 deletion(-) diff --git a/drivers/cpufreq/cpufreq.c b/drivers/cpufreq/cpufreq.c index 8f356c4befda..d2a22de4e4d2 100644 --- a/drivers/cpufreq/cpufreq.c +++ b/drivers/cpufreq/cpufreq.c @@ -2160,7 +2160,6 @@ void cpufreq_unregister_governor(struct cpufreq_governor *governor) mutex_lock(&cpufreq_governor_mutex); list_del(&governor->governor_list); mutex_unlock(&cpufreq_governor_mutex); - return; } EXPORT_SYMBOL_GPL(cpufreq_unregister_governor); -- cgit 1.4.1 From a8b149d32b663c1a4105273295184b78f53d33cf Mon Sep 17 00:00:00 2001 From: "Rafael J. Wysocki" Date: Thu, 23 Nov 2017 14:27:07 +0100 Subject: cpufreq: Fix governor module removal race It is possible to remove a cpufreq governor module after cpufreq_parse_governor() has returned success in store_scaling_governor() and before cpufreq_set_policy() acquires a reference to it, because the governor list is not protected during that period and nothing prevents the governor from being unregistered then. Prevent that from happening by acquiring an extra reference to the governor module temporarily in cpufreq_parse_governor(), under cpufreq_governor_mutex, and dropping it in store_scaling_governor(), when cpufreq_set_policy() returns. Note that the second cpufreq_parse_governor() call site is fine, because it only cares about the policy member of new_policy. Signed-off-by: Rafael J. Wysocki Acked-by: Viresh Kumar Signed-off-by: Rafael J. Wysocki --- drivers/cpufreq/cpufreq.c | 6 ++++++ 1 file changed, 6 insertions(+) diff --git a/drivers/cpufreq/cpufreq.c b/drivers/cpufreq/cpufreq.c index d2a22de4e4d2..421f318c0e66 100644 --- a/drivers/cpufreq/cpufreq.c +++ b/drivers/cpufreq/cpufreq.c @@ -633,6 +633,8 @@ static int cpufreq_parse_governor(char *str_governor, t = find_governor(str_governor); } + if (t && !try_module_get(t->owner)) + t = NULL; mutex_unlock(&cpufreq_governor_mutex); @@ -766,6 +768,10 @@ static ssize_t store_scaling_governor(struct cpufreq_policy *policy, return -EINVAL; ret = cpufreq_set_policy(policy, &new_policy); + + if (new_policy.governor) + module_put(new_policy.governor->owner); + return ret ? ret : count; } -- cgit 1.4.1 From f5f263fed66f75a4482d7ad49392b4283a05885a Mon Sep 17 00:00:00 2001 From: Viresh Kumar Date: Tue, 5 Dec 2017 11:02:43 +0530 Subject: cpu_cooling: Make of_cpufreq_power_cooling_register() parse DT All the callers of of_cpufreq_power_cooling_register() have almost identical code and it makes more sense to move that code into the helper as its all about reading DT properties. This got rid of lot of redundant code. Acked-by: Eduardo Valentin Signed-off-by: Viresh Kumar Signed-off-by: Rafael J. Wysocki --- Documentation/thermal/cpu-cooling-api.txt | 7 ++--- drivers/cpufreq/arm_big_little.c | 23 +-------------- drivers/cpufreq/cpufreq-dt.c | 27 +---------------- drivers/cpufreq/mediatek-cpufreq.c | 22 +------------- drivers/cpufreq/qoriq-cpufreq.c | 14 +-------- drivers/thermal/cpu_cooling.c | 49 +++++++++++++++++++------------ include/linux/cpu_cooling.h | 15 ++-------- 7 files changed, 41 insertions(+), 116 deletions(-) diff --git a/Documentation/thermal/cpu-cooling-api.txt b/Documentation/thermal/cpu-cooling-api.txt index 71653584cd03..4f6f5e9bb4d6 100644 --- a/Documentation/thermal/cpu-cooling-api.txt +++ b/Documentation/thermal/cpu-cooling-api.txt @@ -51,8 +51,7 @@ Dynamic power). "plat_static_func" is a function to calculate the static power consumed by these cpus (See 2.2 Static power). 1.1.4 struct thermal_cooling_device *of_cpufreq_power_cooling_register( - struct device_node *np, const struct cpumask *clip_cpus, u32 capacitance, - get_static_t plat_static_func) + struct cpufreq_policy *policy) Similar to cpufreq_power_cooling_register, this function register a cpufreq cooling device with power extensions using the device tree @@ -76,8 +75,8 @@ cpu. If you are using CONFIG_CPUFREQ_DT then the device. The `plat_static_func` parameter of `cpufreq_power_cooling_register()` -and `of_cpufreq_power_cooling_register()` is optional. If you don't -provide it, only dynamic power will be considered. +is optional. If you don't provide it, only dynamic power will be +considered. 2.1 Dynamic power diff --git a/drivers/cpufreq/arm_big_little.c b/drivers/cpufreq/arm_big_little.c index 65ec5f01aa8d..3d5ed4ef3927 100644 --- a/drivers/cpufreq/arm_big_little.c +++ b/drivers/cpufreq/arm_big_little.c @@ -526,34 +526,13 @@ static int bL_cpufreq_exit(struct cpufreq_policy *policy) static void bL_cpufreq_ready(struct cpufreq_policy *policy) { - struct device *cpu_dev = get_cpu_device(policy->cpu); int cur_cluster = cpu_to_cluster(policy->cpu); - struct device_node *np; /* Do not register a cpu_cooling device if we are in IKS mode */ if (cur_cluster >= MAX_CLUSTERS) return; - np = of_node_get(cpu_dev->of_node); - if (WARN_ON(!np)) - return; - - if (of_find_property(np, "#cooling-cells", NULL)) { - u32 power_coefficient = 0; - - of_property_read_u32(np, "dynamic-power-coefficient", - &power_coefficient); - - cdev[cur_cluster] = of_cpufreq_power_cooling_register(np, - policy, power_coefficient, NULL); - if (IS_ERR(cdev[cur_cluster])) { - dev_err(cpu_dev, - "running cpufreq without cooling device: %ld\n", - PTR_ERR(cdev[cur_cluster])); - cdev[cur_cluster] = NULL; - } - } - of_node_put(np); + cdev[cur_cluster] = of_cpufreq_power_cooling_register(policy); } static struct cpufreq_driver bL_cpufreq_driver = { diff --git a/drivers/cpufreq/cpufreq-dt.c b/drivers/cpufreq/cpufreq-dt.c index 545946ad0752..1e7bec7694ab 100644 --- a/drivers/cpufreq/cpufreq-dt.c +++ b/drivers/cpufreq/cpufreq-dt.c @@ -319,33 +319,8 @@ static int cpufreq_exit(struct cpufreq_policy *policy) static void cpufreq_ready(struct cpufreq_policy *policy) { struct private_data *priv = policy->driver_data; - struct device_node *np = of_node_get(priv->cpu_dev->of_node); - if (WARN_ON(!np)) - return; - - /* - * For now, just loading the cooling device; - * thermal DT code takes care of matching them. - */ - if (of_find_property(np, "#cooling-cells", NULL)) { - u32 power_coefficient = 0; - - of_property_read_u32(np, "dynamic-power-coefficient", - &power_coefficient); - - priv->cdev = of_cpufreq_power_cooling_register(np, - policy, power_coefficient, NULL); - if (IS_ERR(priv->cdev)) { - dev_err(priv->cpu_dev, - "running cpufreq without cooling device: %ld\n", - PTR_ERR(priv->cdev)); - - priv->cdev = NULL; - } - } - - of_node_put(np); + priv->cdev = of_cpufreq_power_cooling_register(policy); } static struct cpufreq_driver dt_cpufreq_driver = { diff --git a/drivers/cpufreq/mediatek-cpufreq.c b/drivers/cpufreq/mediatek-cpufreq.c index e0d5090b303d..6ff783e1b18a 100644 --- a/drivers/cpufreq/mediatek-cpufreq.c +++ b/drivers/cpufreq/mediatek-cpufreq.c @@ -310,28 +310,8 @@ static int mtk_cpufreq_set_target(struct cpufreq_policy *policy, static void mtk_cpufreq_ready(struct cpufreq_policy *policy) { struct mtk_cpu_dvfs_info *info = policy->driver_data; - struct device_node *np = of_node_get(info->cpu_dev->of_node); - u32 capacitance = 0; - if (WARN_ON(!np)) - return; - - if (of_find_property(np, "#cooling-cells", NULL)) { - of_property_read_u32(np, DYNAMIC_POWER, &capacitance); - - info->cdev = of_cpufreq_power_cooling_register(np, - policy, capacitance, NULL); - - if (IS_ERR(info->cdev)) { - dev_err(info->cpu_dev, - "running cpufreq without cooling device: %ld\n", - PTR_ERR(info->cdev)); - - info->cdev = NULL; - } - } - - of_node_put(np); + info->cdev = of_cpufreq_power_cooling_register(policy); } static int mtk_cpu_dvfs_info_init(struct mtk_cpu_dvfs_info *info, int cpu) diff --git a/drivers/cpufreq/qoriq-cpufreq.c b/drivers/cpufreq/qoriq-cpufreq.c index 4ada55b8856e..3a665c18e14e 100644 --- a/drivers/cpufreq/qoriq-cpufreq.c +++ b/drivers/cpufreq/qoriq-cpufreq.c @@ -275,20 +275,8 @@ static int qoriq_cpufreq_target(struct cpufreq_policy *policy, static void qoriq_cpufreq_ready(struct cpufreq_policy *policy) { struct cpu_data *cpud = policy->driver_data; - struct device_node *np = of_get_cpu_node(policy->cpu, NULL); - if (of_find_property(np, "#cooling-cells", NULL)) { - cpud->cdev = of_cpufreq_cooling_register(np, policy); - - if (IS_ERR(cpud->cdev) && PTR_ERR(cpud->cdev) != -ENOSYS) { - pr_err("cpu%d is not running as cooling device: %ld\n", - policy->cpu, PTR_ERR(cpud->cdev)); - - cpud->cdev = NULL; - } - } - - of_node_put(np); + cpud->cdev = of_cpufreq_power_cooling_register(policy); } static struct cpufreq_driver qoriq_cpufreq_driver = { diff --git a/drivers/thermal/cpu_cooling.c b/drivers/thermal/cpu_cooling.c index dc63aba092e4..a31eb03c788e 100644 --- a/drivers/thermal/cpu_cooling.c +++ b/drivers/thermal/cpu_cooling.c @@ -873,38 +873,51 @@ EXPORT_SYMBOL(cpufreq_power_cooling_register); /** * of_cpufreq_power_cooling_register() - create cpufreq cooling device with power extensions - * @np: a valid struct device_node to the cooling device device tree node - * @policy: cpufreq policy - * @capacitance: dynamic power coefficient for these cpus - * @plat_static_func: function to calculate the static power consumed by these - * cpus (optional) + * @policy: CPUFreq policy. * * This interface function registers the cpufreq cooling device with * the name "thermal-cpufreq-%x". This api can support multiple * instances of cpufreq cooling devices. Using this API, the cpufreq - * cooling device will be linked to the device tree node provided. + * cooling device will be linked to the device tree node of the provided + * policy's CPU. * Using this function, the cooling device will implement the power * extensions by using a simple cpu power model. The cpus must have * registered their OPPs using the OPP library. * - * An optional @plat_static_func may be provided to calculate the - * static power consumed by these cpus. If the platform's static - * power consumption is unknown or negligible, make it NULL. + * It also takes into account, if property present in policy CPU node, the + * static power consumed by the cpu. * * Return: a valid struct thermal_cooling_device pointer on success, - * on failure, it returns a corresponding ERR_PTR(). + * and NULL on failure. */ struct thermal_cooling_device * -of_cpufreq_power_cooling_register(struct device_node *np, - struct cpufreq_policy *policy, - u32 capacitance, - get_static_t plat_static_func) +of_cpufreq_power_cooling_register(struct cpufreq_policy *policy) { - if (!np) - return ERR_PTR(-EINVAL); + struct device_node *np = of_get_cpu_node(policy->cpu, NULL); + struct thermal_cooling_device *cdev = NULL; + u32 capacitance = 0; + + if (!np) { + pr_err("cpu_cooling: OF node not available for cpu%d\n", + policy->cpu); + return NULL; + } - return __cpufreq_cooling_register(np, policy, capacitance, - plat_static_func); + if (of_find_property(np, "#cooling-cells", NULL)) { + of_property_read_u32(np, "dynamic-power-coefficient", + &capacitance); + + cdev = __cpufreq_cooling_register(np, policy, capacitance, + NULL); + if (IS_ERR(cdev)) { + pr_err("cpu_cooling: cpu%d is not running as cooling device: %ld\n", + policy->cpu, PTR_ERR(cdev)); + cdev = NULL; + } + } + + of_node_put(np); + return cdev; } EXPORT_SYMBOL(of_cpufreq_power_cooling_register); diff --git a/include/linux/cpu_cooling.h b/include/linux/cpu_cooling.h index d4292ebc5c8b..f09d4feb34f4 100644 --- a/include/linux/cpu_cooling.h +++ b/include/linux/cpu_cooling.h @@ -56,10 +56,7 @@ of_cpufreq_cooling_register(struct device_node *np, struct cpufreq_policy *policy); struct thermal_cooling_device * -of_cpufreq_power_cooling_register(struct device_node *np, - struct cpufreq_policy *policy, - u32 capacitance, - get_static_t plat_static_func); +of_cpufreq_power_cooling_register(struct cpufreq_policy *policy); #else static inline struct thermal_cooling_device * of_cpufreq_cooling_register(struct device_node *np, @@ -69,10 +66,7 @@ of_cpufreq_cooling_register(struct device_node *np, } static inline struct thermal_cooling_device * -of_cpufreq_power_cooling_register(struct device_node *np, - struct cpufreq_policy *policy, - u32 capacitance, - get_static_t plat_static_func) +of_cpufreq_power_cooling_register(struct cpufreq_policy *policy) { return NULL; } @@ -105,10 +99,7 @@ of_cpufreq_cooling_register(struct device_node *np, } static inline struct thermal_cooling_device * -of_cpufreq_power_cooling_register(struct device_node *np, - struct cpufreq_policy *policy, - u32 capacitance, - get_static_t plat_static_func) +of_cpufreq_power_cooling_register(struct cpufreq_policy *policy) { return NULL; } -- cgit 1.4.1 From ba0966da208ef0793486502a0e6b929fbd6d4223 Mon Sep 17 00:00:00 2001 From: Viresh Kumar Date: Tue, 5 Dec 2017 11:02:44 +0530 Subject: cpu_cooling: Remove unused cpufreq_power_cooling_register() It isn't used by anyone, drop it. Acked-by: Eduardo Valentin Signed-off-by: Viresh Kumar Signed-off-by: Rafael J. Wysocki --- Documentation/thermal/cpu-cooling-api.txt | 24 +++--------------------- drivers/thermal/cpu_cooling.c | 30 ------------------------------ include/linux/cpu_cooling.h | 10 ---------- 3 files changed, 3 insertions(+), 61 deletions(-) diff --git a/Documentation/thermal/cpu-cooling-api.txt b/Documentation/thermal/cpu-cooling-api.txt index 4f6f5e9bb4d6..ea61e8bf7e2b 100644 --- a/Documentation/thermal/cpu-cooling-api.txt +++ b/Documentation/thermal/cpu-cooling-api.txt @@ -36,28 +36,14 @@ the user. The registration APIs returns the cooling device pointer. np: pointer to the cooling device device tree node clip_cpus: cpumask of cpus where the frequency constraints will happen. -1.1.3 struct thermal_cooling_device *cpufreq_power_cooling_register( - const struct cpumask *clip_cpus, u32 capacitance, - get_static_t plat_static_func) - -Similar to cpufreq_cooling_register, this function registers a cpufreq -cooling device. Using this function, the cooling device will -implement the power extensions by using a simple cpu power model. The -cpus must have registered their OPPs using the OPP library. - -The additional parameters are needed for the power model (See 2. Power -models). "capacitance" is the dynamic power coefficient (See 2.1 -Dynamic power). "plat_static_func" is a function to calculate the -static power consumed by these cpus (See 2.2 Static power). - -1.1.4 struct thermal_cooling_device *of_cpufreq_power_cooling_register( +1.1.3 struct thermal_cooling_device *of_cpufreq_power_cooling_register( struct cpufreq_policy *policy) -Similar to cpufreq_power_cooling_register, this function register a +Similar to cpufreq_cooling_register, this function register a cpufreq cooling device with power extensions using the device tree information supplied by the np parameter. -1.1.5 void cpufreq_cooling_unregister(struct thermal_cooling_device *cdev) +1.1.4 void cpufreq_cooling_unregister(struct thermal_cooling_device *cdev) This interface function unregisters the "thermal-cpufreq-%x" cooling device. @@ -74,10 +60,6 @@ cpu. If you are using CONFIG_CPUFREQ_DT then the `cpufreq_frequency_table` should already be assigned to the cpu device. -The `plat_static_func` parameter of `cpufreq_power_cooling_register()` -is optional. If you don't provide it, only dynamic power will be -considered. - 2.1 Dynamic power The dynamic power consumption of a processor depends on many factors. diff --git a/drivers/thermal/cpu_cooling.c b/drivers/thermal/cpu_cooling.c index a31eb03c788e..10199f7e1196 100644 --- a/drivers/thermal/cpu_cooling.c +++ b/drivers/thermal/cpu_cooling.c @@ -841,36 +841,6 @@ of_cpufreq_cooling_register(struct device_node *np, } EXPORT_SYMBOL_GPL(of_cpufreq_cooling_register); -/** - * cpufreq_power_cooling_register() - create cpufreq cooling device with power extensions - * @policy: cpufreq policy - * @capacitance: dynamic power coefficient for these cpus - * @plat_static_func: function to calculate the static power consumed by these - * cpus (optional) - * - * This interface function registers the cpufreq cooling device with - * the name "thermal-cpufreq-%x". This api can support multiple - * instances of cpufreq cooling devices. Using this function, the - * cooling device will implement the power extensions by using a - * simple cpu power model. The cpus must have registered their OPPs - * using the OPP library. - * - * An optional @plat_static_func may be provided to calculate the - * static power consumed by these cpus. If the platform's static - * power consumption is unknown or negligible, make it NULL. - * - * Return: a valid struct thermal_cooling_device pointer on success, - * on failure, it returns a corresponding ERR_PTR(). - */ -struct thermal_cooling_device * -cpufreq_power_cooling_register(struct cpufreq_policy *policy, u32 capacitance, - get_static_t plat_static_func) -{ - return __cpufreq_cooling_register(NULL, policy, capacitance, - plat_static_func); -} -EXPORT_SYMBOL(cpufreq_power_cooling_register); - /** * of_cpufreq_power_cooling_register() - create cpufreq cooling device with power extensions * @policy: CPUFreq policy. diff --git a/include/linux/cpu_cooling.h b/include/linux/cpu_cooling.h index f09d4feb34f4..c35778960a9c 100644 --- a/include/linux/cpu_cooling.h +++ b/include/linux/cpu_cooling.h @@ -41,10 +41,6 @@ typedef int (*get_static_t)(cpumask_t *cpumask, int interval, struct thermal_cooling_device * cpufreq_cooling_register(struct cpufreq_policy *policy); -struct thermal_cooling_device * -cpufreq_power_cooling_register(struct cpufreq_policy *policy, - u32 capacitance, get_static_t plat_static_func); - /** * of_cpufreq_cooling_register - create cpufreq cooling device based on DT. * @np: a valid struct device_node to the cooling device device tree node. @@ -84,12 +80,6 @@ cpufreq_cooling_register(struct cpufreq_policy *policy) { return ERR_PTR(-ENOSYS); } -static inline struct thermal_cooling_device * -cpufreq_power_cooling_register(struct cpufreq_policy *policy, - u32 capacitance, get_static_t plat_static_func) -{ - return NULL; -} static inline struct thermal_cooling_device * of_cpufreq_cooling_register(struct device_node *np, -- cgit 1.4.1 From 3ebb62ffc4e4817c3288cdf7ed67ccaf453385e3 Mon Sep 17 00:00:00 2001 From: Viresh Kumar Date: Tue, 5 Dec 2017 11:02:45 +0530 Subject: cpu_cooling: Keep only one of_cpufreq*cooling_register() helper of_cpufreq_cooling_register() isn't used by anyone and so can be removed, but then we would be left with two routines: cpufreq_cooling_register() and of_cpufreq_power_cooling_register() that would look odd. Remove current implementation of of_cpufreq_cooling_register() and rename of_cpufreq_power_cooling_register() as of_cpufreq_cooling_register(). This simplifies lots of stuff. Acked-by: Eduardo Valentin Signed-off-by: Viresh Kumar Signed-off-by: Rafael J. Wysocki --- Documentation/thermal/cpu-cooling-api.txt | 14 ++------ drivers/cpufreq/arm_big_little.c | 2 +- drivers/cpufreq/cpufreq-dt.c | 2 +- drivers/cpufreq/mediatek-cpufreq.c | 2 +- drivers/cpufreq/qoriq-cpufreq.c | 2 +- drivers/thermal/cpu_cooling.c | 28 ++-------------- include/linux/cpu_cooling.h | 53 ++++++++----------------------- 7 files changed, 23 insertions(+), 80 deletions(-) diff --git a/Documentation/thermal/cpu-cooling-api.txt b/Documentation/thermal/cpu-cooling-api.txt index ea61e8bf7e2b..7a1c89db0419 100644 --- a/Documentation/thermal/cpu-cooling-api.txt +++ b/Documentation/thermal/cpu-cooling-api.txt @@ -26,24 +26,16 @@ the user. The registration APIs returns the cooling device pointer. clip_cpus: cpumask of cpus where the frequency constraints will happen. 1.1.2 struct thermal_cooling_device *of_cpufreq_cooling_register( - struct device_node *np, const struct cpumask *clip_cpus) + struct cpufreq_policy *policy) This interface function registers the cpufreq cooling device with the name "thermal-cpufreq-%x" linking it with a device tree node, in order to bind it via the thermal DT code. This api can support multiple instances of cpufreq cooling devices. - np: pointer to the cooling device device tree node - clip_cpus: cpumask of cpus where the frequency constraints will happen. - -1.1.3 struct thermal_cooling_device *of_cpufreq_power_cooling_register( - struct cpufreq_policy *policy) - -Similar to cpufreq_cooling_register, this function register a -cpufreq cooling device with power extensions using the device tree -information supplied by the np parameter. + policy: CPUFreq policy. -1.1.4 void cpufreq_cooling_unregister(struct thermal_cooling_device *cdev) +1.1.3 void cpufreq_cooling_unregister(struct thermal_cooling_device *cdev) This interface function unregisters the "thermal-cpufreq-%x" cooling device. diff --git a/drivers/cpufreq/arm_big_little.c b/drivers/cpufreq/arm_big_little.c index 3d5ed4ef3927..c56b57dcfda5 100644 --- a/drivers/cpufreq/arm_big_little.c +++ b/drivers/cpufreq/arm_big_little.c @@ -532,7 +532,7 @@ static void bL_cpufreq_ready(struct cpufreq_policy *policy) if (cur_cluster >= MAX_CLUSTERS) return; - cdev[cur_cluster] = of_cpufreq_power_cooling_register(policy); + cdev[cur_cluster] = of_cpufreq_cooling_register(policy); } static struct cpufreq_driver bL_cpufreq_driver = { diff --git a/drivers/cpufreq/cpufreq-dt.c b/drivers/cpufreq/cpufreq-dt.c index 1e7bec7694ab..de3d104c25d7 100644 --- a/drivers/cpufreq/cpufreq-dt.c +++ b/drivers/cpufreq/cpufreq-dt.c @@ -320,7 +320,7 @@ static void cpufreq_ready(struct cpufreq_policy *policy) { struct private_data *priv = policy->driver_data; - priv->cdev = of_cpufreq_power_cooling_register(policy); + priv->cdev = of_cpufreq_cooling_register(policy); } static struct cpufreq_driver dt_cpufreq_driver = { diff --git a/drivers/cpufreq/mediatek-cpufreq.c b/drivers/cpufreq/mediatek-cpufreq.c index 6ff783e1b18a..f95975b76d98 100644 --- a/drivers/cpufreq/mediatek-cpufreq.c +++ b/drivers/cpufreq/mediatek-cpufreq.c @@ -311,7 +311,7 @@ static void mtk_cpufreq_ready(struct cpufreq_policy *policy) { struct mtk_cpu_dvfs_info *info = policy->driver_data; - info->cdev = of_cpufreq_power_cooling_register(policy); + info->cdev = of_cpufreq_cooling_register(policy); } static int mtk_cpu_dvfs_info_init(struct mtk_cpu_dvfs_info *info, int cpu) diff --git a/drivers/cpufreq/qoriq-cpufreq.c b/drivers/cpufreq/qoriq-cpufreq.c index 3a665c18e14e..0562761a3dec 100644 --- a/drivers/cpufreq/qoriq-cpufreq.c +++ b/drivers/cpufreq/qoriq-cpufreq.c @@ -276,7 +276,7 @@ static void qoriq_cpufreq_ready(struct cpufreq_policy *policy) { struct cpu_data *cpud = policy->driver_data; - cpud->cdev = of_cpufreq_power_cooling_register(policy); + cpud->cdev = of_cpufreq_cooling_register(policy); } static struct cpufreq_driver qoriq_cpufreq_driver = { diff --git a/drivers/thermal/cpu_cooling.c b/drivers/thermal/cpu_cooling.c index 10199f7e1196..3371caf3095c 100644 --- a/drivers/thermal/cpu_cooling.c +++ b/drivers/thermal/cpu_cooling.c @@ -819,7 +819,6 @@ EXPORT_SYMBOL_GPL(cpufreq_cooling_register); /** * of_cpufreq_cooling_register - function to create cpufreq cooling device. - * @np: a valid struct device_node to the cooling device device tree node * @policy: cpufreq policy * * This interface function registers the cpufreq cooling device with the name @@ -827,29 +826,6 @@ EXPORT_SYMBOL_GPL(cpufreq_cooling_register); * cooling devices. Using this API, the cpufreq cooling device will be * linked to the device tree node provided. * - * Return: a valid struct thermal_cooling_device pointer on success, - * on failure, it returns a corresponding ERR_PTR(). - */ -struct thermal_cooling_device * -of_cpufreq_cooling_register(struct device_node *np, - struct cpufreq_policy *policy) -{ - if (!np) - return ERR_PTR(-EINVAL); - - return __cpufreq_cooling_register(np, policy, 0, NULL); -} -EXPORT_SYMBOL_GPL(of_cpufreq_cooling_register); - -/** - * of_cpufreq_power_cooling_register() - create cpufreq cooling device with power extensions - * @policy: CPUFreq policy. - * - * This interface function registers the cpufreq cooling device with - * the name "thermal-cpufreq-%x". This api can support multiple - * instances of cpufreq cooling devices. Using this API, the cpufreq - * cooling device will be linked to the device tree node of the provided - * policy's CPU. * Using this function, the cooling device will implement the power * extensions by using a simple cpu power model. The cpus must have * registered their OPPs using the OPP library. @@ -861,7 +837,7 @@ EXPORT_SYMBOL_GPL(of_cpufreq_cooling_register); * and NULL on failure. */ struct thermal_cooling_device * -of_cpufreq_power_cooling_register(struct cpufreq_policy *policy) +of_cpufreq_cooling_register(struct cpufreq_policy *policy) { struct device_node *np = of_get_cpu_node(policy->cpu, NULL); struct thermal_cooling_device *cdev = NULL; @@ -889,7 +865,7 @@ of_cpufreq_power_cooling_register(struct cpufreq_policy *policy) of_node_put(np); return cdev; } -EXPORT_SYMBOL(of_cpufreq_power_cooling_register); +EXPORT_SYMBOL_GPL(of_cpufreq_cooling_register); /** * cpufreq_cooling_unregister - function to remove cpufreq cooling device. diff --git a/include/linux/cpu_cooling.h b/include/linux/cpu_cooling.h index c35778960a9c..fd0ea8ddca93 100644 --- a/include/linux/cpu_cooling.h +++ b/include/linux/cpu_cooling.h @@ -41,33 +41,6 @@ typedef int (*get_static_t)(cpumask_t *cpumask, int interval, struct thermal_cooling_device * cpufreq_cooling_register(struct cpufreq_policy *policy); -/** - * of_cpufreq_cooling_register - create cpufreq cooling device based on DT. - * @np: a valid struct device_node to the cooling device device tree node. - * @policy: cpufreq policy. - */ -#ifdef CONFIG_THERMAL_OF -struct thermal_cooling_device * -of_cpufreq_cooling_register(struct device_node *np, - struct cpufreq_policy *policy); - -struct thermal_cooling_device * -of_cpufreq_power_cooling_register(struct cpufreq_policy *policy); -#else -static inline struct thermal_cooling_device * -of_cpufreq_cooling_register(struct device_node *np, - struct cpufreq_policy *policy) -{ - return ERR_PTR(-ENOSYS); -} - -static inline struct thermal_cooling_device * -of_cpufreq_power_cooling_register(struct cpufreq_policy *policy) -{ - return NULL; -} -#endif - /** * cpufreq_cooling_unregister - function to remove cpufreq cooling device. * @cdev: thermal cooling device pointer. @@ -81,24 +54,26 @@ cpufreq_cooling_register(struct cpufreq_policy *policy) return ERR_PTR(-ENOSYS); } -static inline struct thermal_cooling_device * -of_cpufreq_cooling_register(struct device_node *np, - struct cpufreq_policy *policy) +static inline +void cpufreq_cooling_unregister(struct thermal_cooling_device *cdev) { - return ERR_PTR(-ENOSYS); + return; } +#endif /* CONFIG_CPU_THERMAL */ +#if defined(CONFIG_THERMAL_OF) && defined(CONFIG_CPU_THERMAL) +/** + * of_cpufreq_cooling_register - create cpufreq cooling device based on DT. + * @policy: cpufreq policy. + */ +struct thermal_cooling_device * +of_cpufreq_cooling_register(struct cpufreq_policy *policy); +#else static inline struct thermal_cooling_device * -of_cpufreq_power_cooling_register(struct cpufreq_policy *policy) +of_cpufreq_cooling_register(struct cpufreq_policy *policy) { return NULL; } - -static inline -void cpufreq_cooling_unregister(struct thermal_cooling_device *cdev) -{ - return; -} -#endif /* CONFIG_CPU_THERMAL */ +#endif /* defined(CONFIG_THERMAL_OF) && defined(CONFIG_CPU_THERMAL) */ #endif /* __CPU_COOLING_H__ */ -- cgit 1.4.1 From 84fe2cab48590e4373978e4ef2031c977de98995 Mon Sep 17 00:00:00 2001 From: Viresh Kumar Date: Tue, 5 Dec 2017 11:02:46 +0530 Subject: cpu_cooling: Drop static-power related stuff No one has used it for the last two and half years (since it was introduced by commit c36cf0717631 (thermal: cpu_cooling: implement the power cooling device API), get rid of it. Acked-by: Eduardo Valentin Signed-off-by: Viresh Kumar Signed-off-by: Rafael J. Wysocki --- drivers/thermal/cpu_cooling.c | 106 +++++------------------------------------ include/linux/cpu_cooling.h | 3 -- include/trace/events/thermal.h | 10 ++-- 3 files changed, 16 insertions(+), 103 deletions(-) diff --git a/drivers/thermal/cpu_cooling.c b/drivers/thermal/cpu_cooling.c index 3371caf3095c..dfd23245f778 100644 --- a/drivers/thermal/cpu_cooling.c +++ b/drivers/thermal/cpu_cooling.c @@ -88,7 +88,6 @@ struct time_in_idle { * @policy: cpufreq policy. * @node: list_head to link all cpufreq_cooling_device together. * @idle_time: idle time stats - * @plat_get_static_power: callback to calculate the static power * * This structure is required for keeping information of each registered * cpufreq_cooling_device. @@ -104,7 +103,6 @@ struct cpufreq_cooling_device { struct cpufreq_policy *policy; struct list_head node; struct time_in_idle *idle_time; - get_static_t plat_get_static_power; }; static DEFINE_IDA(cpufreq_ida); @@ -318,60 +316,6 @@ static u32 get_load(struct cpufreq_cooling_device *cpufreq_cdev, int cpu, return load; } -/** - * get_static_power() - calculate the static power consumed by the cpus - * @cpufreq_cdev: struct &cpufreq_cooling_device for this cpu cdev - * @tz: thermal zone device in which we're operating - * @freq: frequency in KHz - * @power: pointer in which to store the calculated static power - * - * Calculate the static power consumed by the cpus described by - * @cpu_actor running at frequency @freq. This function relies on a - * platform specific function that should have been provided when the - * actor was registered. If it wasn't, the static power is assumed to - * be negligible. The calculated static power is stored in @power. - * - * Return: 0 on success, -E* on failure. - */ -static int get_static_power(struct cpufreq_cooling_device *cpufreq_cdev, - struct thermal_zone_device *tz, unsigned long freq, - u32 *power) -{ - struct dev_pm_opp *opp; - unsigned long voltage; - struct cpufreq_policy *policy = cpufreq_cdev->policy; - struct cpumask *cpumask = policy->related_cpus; - unsigned long freq_hz = freq * 1000; - struct device *dev; - - if (!cpufreq_cdev->plat_get_static_power) { - *power = 0; - return 0; - } - - dev = get_cpu_device(policy->cpu); - WARN_ON(!dev); - - opp = dev_pm_opp_find_freq_exact(dev, freq_hz, true); - if (IS_ERR(opp)) { - dev_warn_ratelimited(dev, "Failed to find OPP for frequency %lu: %ld\n", - freq_hz, PTR_ERR(opp)); - return -EINVAL; - } - - voltage = dev_pm_opp_get_voltage(opp); - dev_pm_opp_put(opp); - - if (voltage == 0) { - dev_err_ratelimited(dev, "Failed to get voltage for frequency %lu\n", - freq_hz); - return -EINVAL; - } - - return cpufreq_cdev->plat_get_static_power(cpumask, tz->passive_delay, - voltage, power); -} - /** * get_dynamic_power() - calculate the dynamic power * @cpufreq_cdev: &cpufreq_cooling_device for this cdev @@ -491,8 +435,8 @@ static int cpufreq_get_requested_power(struct thermal_cooling_device *cdev, u32 *power) { unsigned long freq; - int i = 0, cpu, ret; - u32 static_power, dynamic_power, total_load = 0; + int i = 0, cpu; + u32 total_load = 0; struct cpufreq_cooling_device *cpufreq_cdev = cdev->devdata; struct cpufreq_policy *policy = cpufreq_cdev->policy; u32 *load_cpu = NULL; @@ -522,22 +466,15 @@ static int cpufreq_get_requested_power(struct thermal_cooling_device *cdev, cpufreq_cdev->last_load = total_load; - dynamic_power = get_dynamic_power(cpufreq_cdev, freq); - ret = get_static_power(cpufreq_cdev, tz, freq, &static_power); - if (ret) { - kfree(load_cpu); - return ret; - } + *power = get_dynamic_power(cpufreq_cdev, freq); if (load_cpu) { trace_thermal_power_cpu_get_power(policy->related_cpus, freq, - load_cpu, i, dynamic_power, - static_power); + load_cpu, i, *power); kfree(load_cpu); } - *power = static_power + dynamic_power; return 0; } @@ -561,8 +498,6 @@ static int cpufreq_state2power(struct thermal_cooling_device *cdev, unsigned long state, u32 *power) { unsigned int freq, num_cpus; - u32 static_power, dynamic_power; - int ret; struct cpufreq_cooling_device *cpufreq_cdev = cdev->devdata; /* Request state should be less than max_level */ @@ -572,13 +507,9 @@ static int cpufreq_state2power(struct thermal_cooling_device *cdev, num_cpus = cpumask_weight(cpufreq_cdev->policy->cpus); freq = cpufreq_cdev->freq_table[state].frequency; - dynamic_power = cpu_freq_to_power(cpufreq_cdev, freq) * num_cpus; - ret = get_static_power(cpufreq_cdev, tz, freq, &static_power); - if (ret) - return ret; + *power = cpu_freq_to_power(cpufreq_cdev, freq) * num_cpus; - *power = static_power + dynamic_power; - return ret; + return 0; } /** @@ -606,21 +537,14 @@ static int cpufreq_power2state(struct thermal_cooling_device *cdev, unsigned long *state) { unsigned int cur_freq, target_freq; - int ret; - s32 dyn_power; - u32 last_load, normalised_power, static_power; + u32 last_load, normalised_power; struct cpufreq_cooling_device *cpufreq_cdev = cdev->devdata; struct cpufreq_policy *policy = cpufreq_cdev->policy; cur_freq = cpufreq_quick_get(policy->cpu); - ret = get_static_power(cpufreq_cdev, tz, cur_freq, &static_power); - if (ret) - return ret; - - dyn_power = power - static_power; - dyn_power = dyn_power > 0 ? dyn_power : 0; + power = power > 0 ? power : 0; last_load = cpufreq_cdev->last_load ?: 1; - normalised_power = (dyn_power * 100) / last_load; + normalised_power = (power * 100) / last_load; target_freq = cpu_power_to_freq(cpufreq_cdev, normalised_power); *state = get_level(cpufreq_cdev, target_freq); @@ -671,8 +595,6 @@ static unsigned int find_next_max(struct cpufreq_frequency_table *table, * @policy: cpufreq policy * Normally this should be same as cpufreq policy->related_cpus. * @capacitance: dynamic power coefficient for these cpus - * @plat_static_func: function to calculate the static power consumed by these - * cpus (optional) * * This interface function registers the cpufreq cooling device with the name * "thermal-cpufreq-%x". This api can support multiple instances of cpufreq @@ -684,8 +606,7 @@ static unsigned int find_next_max(struct cpufreq_frequency_table *table, */ static struct thermal_cooling_device * __cpufreq_cooling_register(struct device_node *np, - struct cpufreq_policy *policy, u32 capacitance, - get_static_t plat_static_func) + struct cpufreq_policy *policy, u32 capacitance) { struct thermal_cooling_device *cdev; struct cpufreq_cooling_device *cpufreq_cdev; @@ -755,8 +676,6 @@ __cpufreq_cooling_register(struct device_node *np, } if (capacitance) { - cpufreq_cdev->plat_get_static_power = plat_static_func; - ret = update_freq_table(cpufreq_cdev, capacitance); if (ret) { cdev = ERR_PTR(ret); @@ -813,7 +732,7 @@ free_cdev: struct thermal_cooling_device * cpufreq_cooling_register(struct cpufreq_policy *policy) { - return __cpufreq_cooling_register(NULL, policy, 0, NULL); + return __cpufreq_cooling_register(NULL, policy, 0); } EXPORT_SYMBOL_GPL(cpufreq_cooling_register); @@ -853,8 +772,7 @@ of_cpufreq_cooling_register(struct cpufreq_policy *policy) of_property_read_u32(np, "dynamic-power-coefficient", &capacitance); - cdev = __cpufreq_cooling_register(np, policy, capacitance, - NULL); + cdev = __cpufreq_cooling_register(np, policy, capacitance); if (IS_ERR(cdev)) { pr_err("cpu_cooling: cpu%d is not running as cooling device: %ld\n", policy->cpu, PTR_ERR(cdev)); diff --git a/include/linux/cpu_cooling.h b/include/linux/cpu_cooling.h index fd0ea8ddca93..de0dafb9399d 100644 --- a/include/linux/cpu_cooling.h +++ b/include/linux/cpu_cooling.h @@ -30,9 +30,6 @@ struct cpufreq_policy; -typedef int (*get_static_t)(cpumask_t *cpumask, int interval, - unsigned long voltage, u32 *power); - #ifdef CONFIG_CPU_THERMAL /** * cpufreq_cooling_register - function to create cpufreq cooling device. diff --git a/include/trace/events/thermal.h b/include/trace/events/thermal.h index 78946640fe03..135e5421f003 100644 --- a/include/trace/events/thermal.h +++ b/include/trace/events/thermal.h @@ -94,9 +94,9 @@ TRACE_EVENT(thermal_zone_trip, #ifdef CONFIG_CPU_THERMAL TRACE_EVENT(thermal_power_cpu_get_power, TP_PROTO(const struct cpumask *cpus, unsigned long freq, u32 *load, - size_t load_len, u32 dynamic_power, u32 static_power), + size_t load_len, u32 dynamic_power), - TP_ARGS(cpus, freq, load, load_len, dynamic_power, static_power), + TP_ARGS(cpus, freq, load, load_len, dynamic_power), TP_STRUCT__entry( __bitmask(cpumask, num_possible_cpus()) @@ -104,7 +104,6 @@ TRACE_EVENT(thermal_power_cpu_get_power, __dynamic_array(u32, load, load_len) __field(size_t, load_len ) __field(u32, dynamic_power ) - __field(u32, static_power ) ), TP_fast_assign( @@ -115,13 +114,12 @@ TRACE_EVENT(thermal_power_cpu_get_power, load_len * sizeof(*load)); __entry->load_len = load_len; __entry->dynamic_power = dynamic_power; - __entry->static_power = static_power; ), - TP_printk("cpus=%s freq=%lu load={%s} dynamic_power=%d static_power=%d", + TP_printk("cpus=%s freq=%lu load={%s} dynamic_power=%d", __get_bitmask(cpumask), __entry->freq, __print_array(__get_dynamic_array(load), __entry->load_len, 4), - __entry->dynamic_power, __entry->static_power) + __entry->dynamic_power) ); TRACE_EVENT(thermal_power_cpu_limit, -- cgit 1.4.1 From 1d0d064307cbfd8546841f6e9d94d02c55e45e1e Mon Sep 17 00:00:00 2001 From: Viresh Kumar Date: Thu, 7 Dec 2017 15:15:19 +0530 Subject: cpufreq: longhaul: Revert transition_delay_us to 200 ms The commit e948bc8fbee0 ("cpufreq: Cap the default transition delay value to 10 ms") caused a regression on EPIA-M min-ITX computer where shutdown or reboot hangs occasionally with a print message like: longhaul: Warning: Timeout while waiting for idle PCI bus cpufreq: __target_index: Failed to change cpu frequency: -16 This probably happens because the cpufreq governor tries to change the frequency of the CPU faster than allowed by the hardware. Before the above commit, the default transition delay was set to 200 ms for a transition_latency of 200000 ns. Lets revert back to that transition delay value to fix it. Note that several other transition delay values were tested like 20 ms and 30 ms and none of them have resolved system hang issue completely. Fixes: e948bc8fbee0 (cpufreq: Cap the default transition delay value to 10 ms) Reported-by: Meelis Roos Suggested-by: Rafael J. Wysocki Signed-off-by: Viresh Kumar Signed-off-by: Rafael J. Wysocki --- drivers/cpufreq/longhaul.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/cpufreq/longhaul.c b/drivers/cpufreq/longhaul.c index c46a12df40dd..5faa37c5b091 100644 --- a/drivers/cpufreq/longhaul.c +++ b/drivers/cpufreq/longhaul.c @@ -894,7 +894,7 @@ static int longhaul_cpu_init(struct cpufreq_policy *policy) if ((longhaul_version != TYPE_LONGHAUL_V1) && (scale_voltage != 0)) longhaul_setup_voltagescaling(); - policy->cpuinfo.transition_latency = 200000; /* nsec */ + policy->transition_delay_us = 200000; /* usec */ return cpufreq_table_validate_and_show(policy, longhaul_table); } -- cgit 1.4.1 From a9596dbc3582c19da0958df25b3852696a0f5ae5 Mon Sep 17 00:00:00 2001 From: Andrew-sh Cheng Date: Fri, 8 Dec 2017 14:07:55 +0800 Subject: cpufreq: mediatek: add mt2712 into compatible list Support mt2712 in mediatek-cpufreq.c Signed-off-by: Andrew-sh Cheng Acked-by: Viresh Kumar Signed-off-by: Rafael J. Wysocki --- drivers/cpufreq/mediatek-cpufreq.c | 1 + 1 file changed, 1 insertion(+) diff --git a/drivers/cpufreq/mediatek-cpufreq.c b/drivers/cpufreq/mediatek-cpufreq.c index e0d5090b303d..b783919f063d 100644 --- a/drivers/cpufreq/mediatek-cpufreq.c +++ b/drivers/cpufreq/mediatek-cpufreq.c @@ -574,6 +574,7 @@ static struct platform_driver mtk_cpufreq_platdrv = { /* List of machines supported by this driver */ static const struct of_device_id mtk_cpufreq_machines[] __initconst = { { .compatible = "mediatek,mt2701", }, + { .compatible = "mediatek,mt2712", }, { .compatible = "mediatek,mt7622", }, { .compatible = "mediatek,mt7623", }, { .compatible = "mediatek,mt817x", }, -- cgit 1.4.1 From 6066998cbd2b1012a8d5bc9a2957cfd0ad53150e Mon Sep 17 00:00:00 2001 From: Andrew-sh Cheng Date: Fri, 8 Dec 2017 14:07:56 +0800 Subject: cpufreq: mediatek: add mediatek related projects into blacklist mediatek projects will use mediate-cpufreq.c as cpufreq driver, instead of using cpufreq_dt.c Add mediatek related projects into cpufreq-dt blacklist Signed-off-by: Andrew-sh Cheng Acked-by: Viresh Kumar Signed-off-by: Rafael J. Wysocki --- drivers/cpufreq/cpufreq-dt-platdev.c | 8 ++++++++ 1 file changed, 8 insertions(+) diff --git a/drivers/cpufreq/cpufreq-dt-platdev.c b/drivers/cpufreq/cpufreq-dt-platdev.c index ecc56e26f8f6..3b585e4bfac5 100644 --- a/drivers/cpufreq/cpufreq-dt-platdev.c +++ b/drivers/cpufreq/cpufreq-dt-platdev.c @@ -108,6 +108,14 @@ static const struct of_device_id blacklist[] __initconst = { { .compatible = "marvell,armadaxp", }, + { .compatible = "mediatek,mt2701", }, + { .compatible = "mediatek,mt2712", }, + { .compatible = "mediatek,mt7622", }, + { .compatible = "mediatek,mt7623", }, + { .compatible = "mediatek,mt817x", }, + { .compatible = "mediatek,mt8173", }, + { .compatible = "mediatek,mt8176", }, + { .compatible = "nvidia,tegra124", }, { .compatible = "st,stih407", }, -- cgit 1.4.1 From b17d2f8d373ef648a05ad7b894722f154d6660f4 Mon Sep 17 00:00:00 2001 From: Gregory CLEMENT Date: Wed, 13 Dec 2017 18:05:35 +0100 Subject: cpufreq: ARM: sort the Kconfig menu Group all the related big LITTLE configuration together and sort the other entries in alphabetic order. Also fixing tab vs space issue while mofifying these entries. Acked-by: Viresh Kumar Signed-off-by: Gregory CLEMENT Signed-off-by: Rafael J. Wysocki --- drivers/cpufreq/Kconfig.arm | 81 ++++++++++++++++++++++----------------------- 1 file changed, 40 insertions(+), 41 deletions(-) diff --git a/drivers/cpufreq/Kconfig.arm b/drivers/cpufreq/Kconfig.arm index bdce4488ded1..beb8826afbb1 100644 --- a/drivers/cpufreq/Kconfig.arm +++ b/drivers/cpufreq/Kconfig.arm @@ -2,6 +2,22 @@ # ARM CPU Frequency scaling drivers # +config ACPI_CPPC_CPUFREQ + tristate "CPUFreq driver based on the ACPI CPPC spec" + depends on ACPI_PROCESSOR + select ACPI_CPPC_LIB + help + This adds a CPUFreq driver which uses CPPC methods + as described in the ACPIv5.1 spec. CPPC stands for + Collaborative Processor Performance Controls. It + is based on an abstract continuous scale of CPU + performance values which allows the remote power + processor to flexibly optimize for power and + performance. CPPC relies on power management firmware + support for its operation. + + If in doubt, say N. + # big LITTLE core layer and glue drivers config ARM_BIG_LITTLE_CPUFREQ tristate "Generic ARM big LITTLE CPUfreq driver" @@ -12,6 +28,30 @@ config ARM_BIG_LITTLE_CPUFREQ help This enables the Generic CPUfreq driver for ARM big.LITTLE platforms. +config ARM_DT_BL_CPUFREQ + tristate "Generic probing via DT for ARM big LITTLE CPUfreq driver" + depends on ARM_BIG_LITTLE_CPUFREQ && OF + help + This enables probing via DT for Generic CPUfreq driver for ARM + big.LITTLE platform. This gets frequency tables from DT. + +config ARM_SCPI_CPUFREQ + tristate "SCPI based CPUfreq driver" + depends on ARM_BIG_LITTLE_CPUFREQ && ARM_SCPI_PROTOCOL && COMMON_CLK_SCPI + help + This adds the CPUfreq driver support for ARM big.LITTLE platforms + using SCPI protocol for CPU power management. + + This driver uses SCPI Message Protocol driver to interact with the + firmware providing the CPU DVFS functionality. + +config ARM_VEXPRESS_SPC_CPUFREQ + tristate "Versatile Express SPC based CPUfreq driver" + depends on ARM_BIG_LITTLE_CPUFREQ && ARCH_VEXPRESS_SPC + help + This add the CPUfreq driver support for Versatile Express + big.LITTLE platforms using SPC for power management. + config ARM_BRCMSTB_AVS_CPUFREQ tristate "Broadcom STB AVS CPUfreq driver" depends on ARCH_BRCMSTB || COMPILE_TEST @@ -33,20 +73,6 @@ config ARM_BRCMSTB_AVS_CPUFREQ_DEBUG If in doubt, say N. -config ARM_DT_BL_CPUFREQ - tristate "Generic probing via DT for ARM big LITTLE CPUfreq driver" - depends on ARM_BIG_LITTLE_CPUFREQ && OF - help - This enables probing via DT for Generic CPUfreq driver for ARM - big.LITTLE platform. This gets frequency tables from DT. - -config ARM_VEXPRESS_SPC_CPUFREQ - tristate "Versatile Express SPC based CPUfreq driver" - depends on ARM_BIG_LITTLE_CPUFREQ && ARCH_VEXPRESS_SPC - help - This add the CPUfreq driver support for Versatile Express - big.LITTLE platforms using SPC for power management. - config ARM_EXYNOS5440_CPUFREQ tristate "SAMSUNG EXYNOS5440" depends on SOC_EXYNOS5440 @@ -205,16 +231,6 @@ config ARM_SA1100_CPUFREQ config ARM_SA1110_CPUFREQ bool -config ARM_SCPI_CPUFREQ - tristate "SCPI based CPUfreq driver" - depends on ARM_BIG_LITTLE_CPUFREQ && ARM_SCPI_PROTOCOL && COMMON_CLK_SCPI - help - This adds the CPUfreq driver support for ARM big.LITTLE platforms - using SCPI protocol for CPU power management. - - This driver uses SCPI Message Protocol driver to interact with the - firmware providing the CPU DVFS functionality. - config ARM_SPEAR_CPUFREQ bool "SPEAr CPUFreq support" depends on PLAT_SPEAR @@ -275,20 +291,3 @@ config ARM_PXA2xx_CPUFREQ This add the CPUFreq driver support for Intel PXA2xx SOCs. If in doubt, say N. - -config ACPI_CPPC_CPUFREQ - tristate "CPUFreq driver based on the ACPI CPPC spec" - depends on ACPI_PROCESSOR - select ACPI_CPPC_LIB - default n - help - This adds a CPUFreq driver which uses CPPC methods - as described in the ACPIv5.1 spec. CPPC stands for - Collaborative Processor Performance Controls. It - is based on an abstract continuous scale of CPU - performance values which allows the remote power - processor to flexibly optimize for power and - performance. CPPC relies on power management firmware - support for its operation. - - If in doubt, say N. -- cgit 1.4.1 From 16630642f174cfc9b169c39cc74ec8e1360a9c55 Mon Sep 17 00:00:00 2001 From: Gregory CLEMENT Date: Wed, 13 Dec 2017 18:05:36 +0100 Subject: cpufreq: sort the drivers in ARM part Keep the driver files alphabetically sorted. Acked-by: Viresh Kumar Signed-off-by: Gregory CLEMENT Signed-off-by: Rafael J. Wysocki --- drivers/cpufreq/Makefile | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) diff --git a/drivers/cpufreq/Makefile b/drivers/cpufreq/Makefile index 812f9e0d01a3..d762e76887e7 100644 --- a/drivers/cpufreq/Makefile +++ b/drivers/cpufreq/Makefile @@ -53,22 +53,24 @@ obj-$(CONFIG_ARM_BIG_LITTLE_CPUFREQ) += arm_big_little.o obj-$(CONFIG_ARM_DT_BL_CPUFREQ) += arm_big_little_dt.o obj-$(CONFIG_ARM_BRCMSTB_AVS_CPUFREQ) += brcmstb-avs-cpufreq.o +obj-$(CONFIG_ACPI_CPPC_CPUFREQ) += cppc_cpufreq.o obj-$(CONFIG_ARCH_DAVINCI) += davinci-cpufreq.o obj-$(CONFIG_ARM_EXYNOS5440_CPUFREQ) += exynos5440-cpufreq.o obj-$(CONFIG_ARM_HIGHBANK_CPUFREQ) += highbank-cpufreq.o obj-$(CONFIG_ARM_IMX6Q_CPUFREQ) += imx6q-cpufreq.o obj-$(CONFIG_ARM_KIRKWOOD_CPUFREQ) += kirkwood-cpufreq.o obj-$(CONFIG_ARM_MEDIATEK_CPUFREQ) += mediatek-cpufreq.o +obj-$(CONFIG_MACH_MVEBU_V7) += mvebu-cpufreq.o obj-$(CONFIG_ARM_OMAP2PLUS_CPUFREQ) += omap-cpufreq.o obj-$(CONFIG_ARM_PXA2xx_CPUFREQ) += pxa2xx-cpufreq.o obj-$(CONFIG_PXA3xx) += pxa3xx-cpufreq.o -obj-$(CONFIG_ARM_S3C24XX_CPUFREQ) += s3c24xx-cpufreq.o -obj-$(CONFIG_ARM_S3C24XX_CPUFREQ_DEBUGFS) += s3c24xx-cpufreq-debugfs.o obj-$(CONFIG_ARM_S3C2410_CPUFREQ) += s3c2410-cpufreq.o obj-$(CONFIG_ARM_S3C2412_CPUFREQ) += s3c2412-cpufreq.o obj-$(CONFIG_ARM_S3C2416_CPUFREQ) += s3c2416-cpufreq.o obj-$(CONFIG_ARM_S3C2440_CPUFREQ) += s3c2440-cpufreq.o obj-$(CONFIG_ARM_S3C64XX_CPUFREQ) += s3c64xx-cpufreq.o +obj-$(CONFIG_ARM_S3C24XX_CPUFREQ) += s3c24xx-cpufreq.o +obj-$(CONFIG_ARM_S3C24XX_CPUFREQ_DEBUGFS) += s3c24xx-cpufreq-debugfs.o obj-$(CONFIG_ARM_S5PV210_CPUFREQ) += s5pv210-cpufreq.o obj-$(CONFIG_ARM_SA1100_CPUFREQ) += sa1100-cpufreq.o obj-$(CONFIG_ARM_SA1110_CPUFREQ) += sa1110-cpufreq.o @@ -81,8 +83,6 @@ obj-$(CONFIG_ARM_TEGRA124_CPUFREQ) += tegra124-cpufreq.o obj-$(CONFIG_ARM_TEGRA186_CPUFREQ) += tegra186-cpufreq.o obj-$(CONFIG_ARM_TI_CPUFREQ) += ti-cpufreq.o obj-$(CONFIG_ARM_VEXPRESS_SPC_CPUFREQ) += vexpress-spc-cpufreq.o -obj-$(CONFIG_ACPI_CPPC_CPUFREQ) += cppc_cpufreq.o -obj-$(CONFIG_MACH_MVEBU_V7) += mvebu-cpufreq.o ################################################################################## -- cgit 1.4.1 From b3371600926638842c460d652599c4dddef72da6 Mon Sep 17 00:00:00 2001 From: Gregory CLEMENT Date: Wed, 13 Dec 2017 18:29:13 +0100 Subject: cpufreq: mvebu: Free the clock reference in the normal path In case of error the clock reference was freed but not in normal path once it was nor more used. This patch fixes it. Signed-off-by: Gregory CLEMENT Acked-by: Viresh Kumar Signed-off-by: Rafael J. Wysocki --- drivers/cpufreq/mvebu-cpufreq.c | 1 + 1 file changed, 1 insertion(+) diff --git a/drivers/cpufreq/mvebu-cpufreq.c b/drivers/cpufreq/mvebu-cpufreq.c index ed915ee85dd9..c043aad8e3a0 100644 --- a/drivers/cpufreq/mvebu-cpufreq.c +++ b/drivers/cpufreq/mvebu-cpufreq.c @@ -99,6 +99,7 @@ static int __init armada_xp_pmsu_cpufreq_init(void) if (ret) dev_err(cpu_dev, "%s: failed to mark OPPs as shared: %d\n", __func__, ret); + clk_put(clk); } platform_device_register_simple("cpufreq-dt", -1, NULL, 0); -- cgit 1.4.1 From 3f4590a4a371006f66385eb5b739d5d01249ee3f Mon Sep 17 00:00:00 2001 From: Gregory CLEMENT Date: Wed, 13 Dec 2017 18:29:14 +0100 Subject: cpufreq: mvebu: Free opp if registering failed Since the introduction of this driver, the functions to remove the opp were added. So stop claiming we can't remove opp and use one of them in case of failure. Signed-off-by: Gregory CLEMENT Acked-by: Viresh Kumar Signed-off-by: Rafael J. Wysocki --- drivers/cpufreq/mvebu-cpufreq.c | 15 ++++++++------- 1 file changed, 8 insertions(+), 7 deletions(-) diff --git a/drivers/cpufreq/mvebu-cpufreq.c b/drivers/cpufreq/mvebu-cpufreq.c index c043aad8e3a0..31513bd42705 100644 --- a/drivers/cpufreq/mvebu-cpufreq.c +++ b/drivers/cpufreq/mvebu-cpufreq.c @@ -76,12 +76,6 @@ static int __init armada_xp_pmsu_cpufreq_init(void) return PTR_ERR(clk); } - /* - * In case of a failure of dev_pm_opp_add(), we don't - * bother with cleaning up the registered OPP (there's - * no function to do so), and simply cancel the - * registration of the cpufreq device. - */ ret = dev_pm_opp_add(cpu_dev, clk_get_rate(clk), 0); if (ret) { clk_put(clk); @@ -91,7 +85,8 @@ static int __init armada_xp_pmsu_cpufreq_init(void) ret = dev_pm_opp_add(cpu_dev, clk_get_rate(clk) / 2, 0); if (ret) { clk_put(clk); - return ret; + dev_err(cpu_dev, "Failed to register OPPs\n"); + goto opp_register_failed; } ret = dev_pm_opp_set_sharing_cpus(cpu_dev, @@ -104,5 +99,11 @@ static int __init armada_xp_pmsu_cpufreq_init(void) platform_device_register_simple("cpufreq-dt", -1, NULL, 0); return 0; + +opp_register_failed: + /* As registering has failed remove all the opp for all cpus */ + dev_pm_opp_cpumask_remove_table(cpu_possible_mask); + + return ret; } device_initcall(armada_xp_pmsu_cpufreq_init); -- cgit 1.4.1 From 7a62f48b271919b78f23c216ede6ac5d2b8b0368 Mon Sep 17 00:00:00 2001 From: Gregory CLEMENT Date: Thu, 14 Dec 2017 16:00:03 +0100 Subject: dt-bindings: marvell: Add documentation for the North Bridge PM on Armada 37xx Extend the documentation of the Armada 37xx SoC with the the North Bridge Power Management component. Reviewed-by: Rob Herring Signed-off-by: Gregory CLEMENT Signed-off-by: Rafael J. Wysocki --- .../devicetree/bindings/arm/marvell/armada-37xx.txt | 19 +++++++++++++++++++ 1 file changed, 19 insertions(+) diff --git a/Documentation/devicetree/bindings/arm/marvell/armada-37xx.txt b/Documentation/devicetree/bindings/arm/marvell/armada-37xx.txt index 51336e5fc761..35c3c3460d17 100644 --- a/Documentation/devicetree/bindings/arm/marvell/armada-37xx.txt +++ b/Documentation/devicetree/bindings/arm/marvell/armada-37xx.txt @@ -14,3 +14,22 @@ following property before the previous one: Example: compatible = "marvell,armada-3720-db", "marvell,armada3720", "marvell,armada3710"; + + +Power management +---------------- + +For power management (particularly DVFS and AVS), the North Bridge +Power Management component is needed: + +Required properties: +- compatible : should contain "marvell,armada-3700-nb-pm", "syscon"; +- reg : the register start and length for the North Bridge + Power Management + +Example: + +nb_pm: syscon@14000 { + compatible = "marvell,armada-3700-nb-pm", "syscon"; + reg = <0x14000 0x60>; +} -- cgit 1.4.1 From d71f617af7849e65884e2a5bd590807ca6b905b8 Mon Sep 17 00:00:00 2001 From: Gregory CLEMENT Date: Thu, 14 Dec 2017 16:00:04 +0100 Subject: MAINTAINERS: add new entries for Armada 37xx cpufreq driver This new driver belongs to the mvebu family, update the MAINTAINER file to document it. Acked-by: Viresh Kumar Signed-off-by: Gregory CLEMENT Signed-off-by: Rafael J. Wysocki --- MAINTAINERS | 1 + 1 file changed, 1 insertion(+) diff --git a/MAINTAINERS b/MAINTAINERS index d4fdcb12616c..8b275dd1105a 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -1582,6 +1582,7 @@ F: arch/arm/boot/dts/kirkwood* F: arch/arm/configs/mvebu_*_defconfig F: arch/arm/mach-mvebu/ F: arch/arm64/boot/dts/marvell/armada* +F: drivers/cpufreq/armada-37xx-cpufreq.c F: drivers/cpufreq/mvebu-cpufreq.c F: drivers/irqchip/irq-armada-370-xp.c F: drivers/irqchip/irq-mvebu-* -- cgit 1.4.1 From 92ce45fb875d7c3e021cc454482fe0687ff54f29 Mon Sep 17 00:00:00 2001 From: Gregory CLEMENT Date: Thu, 14 Dec 2017 16:00:05 +0100 Subject: cpufreq: Add DVFS support for Armada 37xx This patch adds DVFS support for the Armada 37xx SoCs There are up to four CPU frequency loads for Armada 37xx controlled by the hardware. This driver associates the CPU load level to a frequency, then the hardware will switch while selecting a load level. The hardware also can associate a voltage for each level (AVS support) but it is not yet supported Tested-by: Andre Heider Acked-by: Viresh Kumar Signed-off-by: Gregory CLEMENT Signed-off-by: Rafael J. Wysocki --- drivers/cpufreq/Kconfig.arm | 7 + drivers/cpufreq/Makefile | 1 + drivers/cpufreq/armada-37xx-cpufreq.c | 241 ++++++++++++++++++++++++++++++++++ 3 files changed, 249 insertions(+) create mode 100644 drivers/cpufreq/armada-37xx-cpufreq.c diff --git a/drivers/cpufreq/Kconfig.arm b/drivers/cpufreq/Kconfig.arm index beb8826afbb1..3a88e33b0cfe 100644 --- a/drivers/cpufreq/Kconfig.arm +++ b/drivers/cpufreq/Kconfig.arm @@ -18,6 +18,13 @@ config ACPI_CPPC_CPUFREQ If in doubt, say N. +config ARM_ARMADA_37XX_CPUFREQ + tristate "Armada 37xx CPUFreq support" + depends on ARCH_MVEBU + help + This adds the CPUFreq driver support for Marvell Armada 37xx SoCs. + The Armada 37xx PMU supports 4 frequency and VDD levels. + # big LITTLE core layer and glue drivers config ARM_BIG_LITTLE_CPUFREQ tristate "Generic ARM big LITTLE CPUfreq driver" diff --git a/drivers/cpufreq/Makefile b/drivers/cpufreq/Makefile index d762e76887e7..e07715ce8844 100644 --- a/drivers/cpufreq/Makefile +++ b/drivers/cpufreq/Makefile @@ -52,6 +52,7 @@ obj-$(CONFIG_ARM_BIG_LITTLE_CPUFREQ) += arm_big_little.o # LITTLE drivers, so that it is probed last. obj-$(CONFIG_ARM_DT_BL_CPUFREQ) += arm_big_little_dt.o +obj-$(CONFIG_ARM_ARMADA_37XX_CPUFREQ) += armada-37xx-cpufreq.o obj-$(CONFIG_ARM_BRCMSTB_AVS_CPUFREQ) += brcmstb-avs-cpufreq.o obj-$(CONFIG_ACPI_CPPC_CPUFREQ) += cppc_cpufreq.o obj-$(CONFIG_ARCH_DAVINCI) += davinci-cpufreq.o diff --git a/drivers/cpufreq/armada-37xx-cpufreq.c b/drivers/cpufreq/armada-37xx-cpufreq.c new file mode 100644 index 000000000000..c6ebc88a7d8d --- /dev/null +++ b/drivers/cpufreq/armada-37xx-cpufreq.c @@ -0,0 +1,241 @@ +// SPDX-License-Identifier: GPL-2.0+ +/* + * CPU frequency scaling support for Armada 37xx platform. + * + * Copyright (C) 2017 Marvell + * + * Gregory CLEMENT + */ + +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include + +/* Power management in North Bridge register set */ +#define ARMADA_37XX_NB_L0L1 0x18 +#define ARMADA_37XX_NB_L2L3 0x1C +#define ARMADA_37XX_NB_TBG_DIV_OFF 13 +#define ARMADA_37XX_NB_TBG_DIV_MASK 0x7 +#define ARMADA_37XX_NB_CLK_SEL_OFF 11 +#define ARMADA_37XX_NB_CLK_SEL_MASK 0x1 +#define ARMADA_37XX_NB_CLK_SEL_TBG 0x1 +#define ARMADA_37XX_NB_TBG_SEL_OFF 9 +#define ARMADA_37XX_NB_TBG_SEL_MASK 0x3 +#define ARMADA_37XX_NB_VDD_SEL_OFF 6 +#define ARMADA_37XX_NB_VDD_SEL_MASK 0x3 +#define ARMADA_37XX_NB_CONFIG_SHIFT 16 +#define ARMADA_37XX_NB_DYN_MOD 0x24 +#define ARMADA_37XX_NB_CLK_SEL_EN BIT(26) +#define ARMADA_37XX_NB_TBG_EN BIT(28) +#define ARMADA_37XX_NB_DIV_EN BIT(29) +#define ARMADA_37XX_NB_VDD_EN BIT(30) +#define ARMADA_37XX_NB_DFS_EN BIT(31) +#define ARMADA_37XX_NB_CPU_LOAD 0x30 +#define ARMADA_37XX_NB_CPU_LOAD_MASK 0x3 +#define ARMADA_37XX_DVFS_LOAD_0 0 +#define ARMADA_37XX_DVFS_LOAD_1 1 +#define ARMADA_37XX_DVFS_LOAD_2 2 +#define ARMADA_37XX_DVFS_LOAD_3 3 + +/* + * On Armada 37xx the Power management manages 4 level of CPU load, + * each level can be associated with a CPU clock source, a CPU + * divider, a VDD level, etc... + */ +#define LOAD_LEVEL_NR 4 + +struct armada_37xx_dvfs { + u32 cpu_freq_max; + u8 divider[LOAD_LEVEL_NR]; +}; + +static struct armada_37xx_dvfs armada_37xx_dvfs[] = { + {.cpu_freq_max = 1200*1000*1000, .divider = {1, 2, 4, 6} }, + {.cpu_freq_max = 1000*1000*1000, .divider = {1, 2, 4, 5} }, + {.cpu_freq_max = 800*1000*1000, .divider = {1, 2, 3, 4} }, + {.cpu_freq_max = 600*1000*1000, .divider = {2, 4, 5, 6} }, +}; + +static struct armada_37xx_dvfs *armada_37xx_cpu_freq_info_get(u32 freq) +{ + int i; + + for (i = 0; i < ARRAY_SIZE(armada_37xx_dvfs); i++) { + if (freq == armada_37xx_dvfs[i].cpu_freq_max) + return &armada_37xx_dvfs[i]; + } + + pr_err("Unsupported CPU frequency %d MHz\n", freq/1000000); + return NULL; +} + +/* + * Setup the four level managed by the hardware. Once the four level + * will be configured then the DVFS will be enabled. + */ +static void __init armada37xx_cpufreq_dvfs_setup(struct regmap *base, + struct clk *clk, u8 *divider) +{ + int load_lvl; + struct clk *parent; + + for (load_lvl = 0; load_lvl < LOAD_LEVEL_NR; load_lvl++) { + unsigned int reg, mask, val, offset = 0; + + if (load_lvl <= ARMADA_37XX_DVFS_LOAD_1) + reg = ARMADA_37XX_NB_L0L1; + else + reg = ARMADA_37XX_NB_L2L3; + + if (load_lvl == ARMADA_37XX_DVFS_LOAD_0 || + load_lvl == ARMADA_37XX_DVFS_LOAD_2) + offset += ARMADA_37XX_NB_CONFIG_SHIFT; + + /* Set cpu clock source, for all the level we use TBG */ + val = ARMADA_37XX_NB_CLK_SEL_TBG << ARMADA_37XX_NB_CLK_SEL_OFF; + mask = (ARMADA_37XX_NB_CLK_SEL_MASK + << ARMADA_37XX_NB_CLK_SEL_OFF); + + /* + * Set cpu divider based on the pre-computed array in + * order to have balanced step. + */ + val |= divider[load_lvl] << ARMADA_37XX_NB_TBG_DIV_OFF; + mask |= (ARMADA_37XX_NB_TBG_DIV_MASK + << ARMADA_37XX_NB_TBG_DIV_OFF); + + /* Set VDD divider which is actually the load level. */ + val |= load_lvl << ARMADA_37XX_NB_VDD_SEL_OFF; + mask |= (ARMADA_37XX_NB_VDD_SEL_MASK + << ARMADA_37XX_NB_VDD_SEL_OFF); + + val <<= offset; + mask <<= offset; + + regmap_update_bits(base, reg, mask, val); + } + + /* + * Set cpu clock source, for all the level we keep the same + * clock source that the one already configured. For this one + * we need to use the clock framework + */ + parent = clk_get_parent(clk); + clk_set_parent(clk, parent); +} + +static void __init armada37xx_cpufreq_disable_dvfs(struct regmap *base) +{ + unsigned int reg = ARMADA_37XX_NB_DYN_MOD, + mask = ARMADA_37XX_NB_DFS_EN; + + regmap_update_bits(base, reg, mask, 0); +} + +static void __init armada37xx_cpufreq_enable_dvfs(struct regmap *base) +{ + unsigned int val, reg = ARMADA_37XX_NB_CPU_LOAD, + mask = ARMADA_37XX_NB_CPU_LOAD_MASK; + + /* Start with the highest load (0) */ + val = ARMADA_37XX_DVFS_LOAD_0; + regmap_update_bits(base, reg, mask, val); + + /* Now enable DVFS for the CPUs */ + reg = ARMADA_37XX_NB_DYN_MOD; + mask = ARMADA_37XX_NB_CLK_SEL_EN | ARMADA_37XX_NB_TBG_EN | + ARMADA_37XX_NB_DIV_EN | ARMADA_37XX_NB_VDD_EN | + ARMADA_37XX_NB_DFS_EN; + + regmap_update_bits(base, reg, mask, mask); +} + +static int __init armada37xx_cpufreq_driver_init(void) +{ + struct armada_37xx_dvfs *dvfs; + struct platform_device *pdev; + unsigned int cur_frequency; + struct regmap *nb_pm_base; + struct device *cpu_dev; + int load_lvl, ret; + struct clk *clk; + + nb_pm_base = + syscon_regmap_lookup_by_compatible("marvell,armada-3700-nb-pm"); + + if (IS_ERR(nb_pm_base)) + return -ENODEV; + + /* Before doing any configuration on the DVFS first, disable it */ + armada37xx_cpufreq_disable_dvfs(nb_pm_base); + + /* + * On CPU 0 register the operating points supported (which are + * the nominal CPU frequency and full integer divisions of + * it). + */ + cpu_dev = get_cpu_device(0); + if (!cpu_dev) { + dev_err(cpu_dev, "Cannot get CPU\n"); + return -ENODEV; + } + + clk = clk_get(cpu_dev, 0); + if (IS_ERR(clk)) { + dev_err(cpu_dev, "Cannot get clock for CPU0\n"); + return PTR_ERR(clk); + } + + /* Get nominal (current) CPU frequency */ + cur_frequency = clk_get_rate(clk); + if (!cur_frequency) { + dev_err(cpu_dev, "Failed to get clock rate for CPU\n"); + return -EINVAL; + } + + dvfs = armada_37xx_cpu_freq_info_get(cur_frequency); + if (!dvfs) + return -EINVAL; + + armada37xx_cpufreq_dvfs_setup(nb_pm_base, clk, dvfs->divider); + + for (load_lvl = ARMADA_37XX_DVFS_LOAD_0; load_lvl < LOAD_LEVEL_NR; + load_lvl++) { + unsigned long freq = cur_frequency / dvfs->divider[load_lvl]; + + ret = dev_pm_opp_add(cpu_dev, freq, 0); + if (ret) { + /* clean-up the already added opp before leaving */ + while (load_lvl-- > ARMADA_37XX_DVFS_LOAD_0) { + freq = cur_frequency / dvfs->divider[load_lvl]; + dev_pm_opp_remove(cpu_dev, freq); + } + return ret; + } + } + + /* Now that everything is setup, enable the DVFS at hardware level */ + armada37xx_cpufreq_enable_dvfs(nb_pm_base); + + pdev = platform_device_register_simple("cpufreq-dt", -1, NULL, 0); + + return PTR_ERR_OR_ZERO(pdev); +} +/* late_initcall, to guarantee the driver is loaded after A37xx clock driver */ +late_initcall(armada37xx_cpufreq_driver_init); + +MODULE_AUTHOR("Gregory CLEMENT "); +MODULE_DESCRIPTION("Armada 37xx cpufreq driver"); +MODULE_LICENSE("GPL"); -- cgit 1.4.1 From db410b2b3839e962f9df4bc87b4fea9a2996047c Mon Sep 17 00:00:00 2001 From: Dave Gerlach Date: Thu, 14 Dec 2017 22:25:25 -0600 Subject: cpufreq: ti-cpufreq: Convert to module_platform_driver ti-cpufreq will be responsible for calling dev_pm_opp_set_regulators on platforms that require AVS and ABB regulator support so we must be able to defer probe if regulators are not yet available, so change ti-cpufreq to be a module_platform_driver to allow for probe defer. Acked-by: Viresh Kumar Signed-off-by: Dave Gerlach Signed-off-by: Rafael J. Wysocki --- drivers/cpufreq/ti-cpufreq.c | 23 +++++++++++++++++++++-- 1 file changed, 21 insertions(+), 2 deletions(-) diff --git a/drivers/cpufreq/ti-cpufreq.c b/drivers/cpufreq/ti-cpufreq.c index 923317f03b4b..b1c230a1e2aa 100644 --- a/drivers/cpufreq/ti-cpufreq.c +++ b/drivers/cpufreq/ti-cpufreq.c @@ -17,6 +17,7 @@ #include #include #include +#include #include #include #include @@ -195,7 +196,7 @@ static const struct of_device_id ti_cpufreq_of_match[] = { {}, }; -static int ti_cpufreq_init(void) +static int ti_cpufreq_probe(struct platform_device *pdev) { u32 version[VERSION_COUNT]; struct device_node *np; @@ -269,4 +270,22 @@ free_opp_data: return ret; } -device_initcall(ti_cpufreq_init); + +static int ti_cpufreq_init(void) +{ + platform_device_register_simple("ti-cpufreq", -1, NULL, 0); + return 0; +} +module_init(ti_cpufreq_init); + +static struct platform_driver ti_cpufreq_driver = { + .probe = ti_cpufreq_probe, + .driver = { + .name = "ti-cpufreq", + }, +}; +module_platform_driver(ti_cpufreq_driver); + +MODULE_DESCRIPTION("TI CPUFreq/OPP hw-supported driver"); +MODULE_AUTHOR("Dave Gerlach "); +MODULE_LICENSE("GPL v2"); -- cgit 1.4.1 From c8343e83d45fb074da5dfc304e101995d19f05b2 Mon Sep 17 00:00:00 2001 From: Dave Gerlach Date: Thu, 14 Dec 2017 22:25:26 -0600 Subject: cpufreq: ti-cpufreq: Add support for multiple regulators Some platforms, like those in the DRA7 and AM57 families, require the scaling of multiple regulators in order to properly support higher OPPs. Let the ti-cpufreq driver determine when this is required and pass the appropriate regulator names to the OPP core so that they can be properly managed. Acked-by: Viresh Kumar Signed-off-by: Dave Gerlach Signed-off-by: Rafael J. Wysocki --- drivers/cpufreq/ti-cpufreq.c | 28 ++++++++++++++++++++++++---- 1 file changed, 24 insertions(+), 4 deletions(-) diff --git a/drivers/cpufreq/ti-cpufreq.c b/drivers/cpufreq/ti-cpufreq.c index b1c230a1e2aa..a099b7bf74cd 100644 --- a/drivers/cpufreq/ti-cpufreq.c +++ b/drivers/cpufreq/ti-cpufreq.c @@ -51,6 +51,7 @@ struct ti_cpufreq_soc_data { unsigned long efuse_mask; unsigned long efuse_shift; unsigned long rev_offset; + bool multi_regulator; }; struct ti_cpufreq_data { @@ -58,6 +59,7 @@ struct ti_cpufreq_data { struct device_node *opp_node; struct regmap *syscon; const struct ti_cpufreq_soc_data *soc_data; + struct opp_table *opp_table; }; static unsigned long amx3_efuse_xlate(struct ti_cpufreq_data *opp_data, @@ -96,6 +98,7 @@ static struct ti_cpufreq_soc_data am3x_soc_data = { .efuse_offset = 0x07fc, .efuse_mask = 0x1fff, .rev_offset = 0x600, + .multi_regulator = false, }; static struct ti_cpufreq_soc_data am4x_soc_data = { @@ -104,6 +107,7 @@ static struct ti_cpufreq_soc_data am4x_soc_data = { .efuse_offset = 0x0610, .efuse_mask = 0x3f, .rev_offset = 0x600, + .multi_regulator = false, }; static struct ti_cpufreq_soc_data dra7_soc_data = { @@ -112,6 +116,7 @@ static struct ti_cpufreq_soc_data dra7_soc_data = { .efuse_mask = 0xf80000, .efuse_shift = 19, .rev_offset = 0x204, + .multi_regulator = true, }; /** @@ -201,7 +206,9 @@ static int ti_cpufreq_probe(struct platform_device *pdev) u32 version[VERSION_COUNT]; struct device_node *np; const struct of_device_id *match; + struct opp_table *ti_opp_table; struct ti_cpufreq_data *opp_data; + const char * const reg_names[] = {"vdd", "vbb"}; int ret; np = of_find_node_by_path("/"); @@ -248,16 +255,29 @@ static int ti_cpufreq_probe(struct platform_device *pdev) if (ret) goto fail_put_node; - ret = PTR_ERR_OR_ZERO(dev_pm_opp_set_supported_hw(opp_data->cpu_dev, - version, VERSION_COUNT)); - if (ret) { + ti_opp_table = dev_pm_opp_set_supported_hw(opp_data->cpu_dev, + version, VERSION_COUNT); + if (IS_ERR(ti_opp_table)) { dev_err(opp_data->cpu_dev, "Failed to set supported hardware\n"); + ret = PTR_ERR(ti_opp_table); goto fail_put_node; } - of_node_put(opp_data->opp_node); + opp_data->opp_table = ti_opp_table; + + if (opp_data->soc_data->multi_regulator) { + ti_opp_table = dev_pm_opp_set_regulators(opp_data->cpu_dev, + reg_names, + ARRAY_SIZE(reg_names)); + if (IS_ERR(ti_opp_table)) { + dev_pm_opp_put_supported_hw(opp_data->opp_table); + ret = PTR_ERR(ti_opp_table); + goto fail_put_node; + } + } + of_node_put(opp_data->opp_node); register_cpufreq_dt: platform_device_register_simple("cpufreq-dt", -1, NULL, 0); -- cgit 1.4.1 From 212b7287ae6098337fffa9c0cd7e139dceb98125 Mon Sep 17 00:00:00 2001 From: Dave Gerlach Date: Thu, 14 Dec 2017 22:25:27 -0600 Subject: dt-bindings: opp: Introduce ti-opp-supply bindings Document the devicetree bindings that describe Texas Instruments opp-supply which allow a platform to describe multiple regulators and additional information, such as registers containing data needed to program aforementioned regulators. Signed-off-by: Dave Gerlach Acked-by: Viresh Kumar Acked-by: Rob Herring Signed-off-by: Rafael J. Wysocki --- .../bindings/opp/ti-omap5-opp-supply.txt | 63 ++++++++++++++++++++++ 1 file changed, 63 insertions(+) create mode 100644 Documentation/devicetree/bindings/opp/ti-omap5-opp-supply.txt diff --git a/Documentation/devicetree/bindings/opp/ti-omap5-opp-supply.txt b/Documentation/devicetree/bindings/opp/ti-omap5-opp-supply.txt new file mode 100644 index 000000000000..832346e489a3 --- /dev/null +++ b/Documentation/devicetree/bindings/opp/ti-omap5-opp-supply.txt @@ -0,0 +1,63 @@ +Texas Instruments OMAP compatible OPP supply description + +OMAP5, DRA7, and AM57 family of SoCs have Class0 AVS eFuse registers which +contain data that can be used to adjust voltages programmed for some of their +supplies for more efficient operation. This binding provides the information +needed to read these values and use them to program the main regulator during +an OPP transitions. + +Also, some supplies may have an associated vbb-supply which is an Adaptive Body +Bias regulator which much be transitioned in a specific sequence with regards +to the vdd-supply and clk when making an OPP transition. By supplying two +regulators to the device that will undergo OPP transitions we can make use +of the multi regulator binding that is part of the OPP core described here [1] +to describe both regulators needed by the platform. + +[1] Documentation/devicetree/bindings/opp/opp.txt + +Required Properties for Device Node: +- vdd-supply: phandle to regulator controlling VDD supply +- vbb-supply: phandle to regulator controlling Body Bias supply + (Usually Adaptive Body Bias regulator) + +Required Properties for opp-supply node: +- compatible: Should be one of: + "ti,omap-opp-supply" - basic OPP supply controlling VDD and VBB + "ti,omap5-opp-supply" - OMAP5+ optimized voltages in efuse(class0)VDD + along with VBB + "ti,omap5-core-opp-supply" - OMAP5+ optimized voltages in efuse(class0) VDD + but no VBB. +- reg: Address and length of the efuse register set for the device (mandatory + only for "ti,omap5-opp-supply") +- ti,efuse-settings: An array of u32 tuple items providing information about + optimized efuse configuration. Each item consists of the following: + volt: voltage in uV - reference voltage (OPP voltage) + efuse_offseet: efuse offset from reg where the optimized voltage is stored. +- ti,absolute-max-voltage-uv: absolute maximum voltage for the OPP supply. + +Example: + +/* Device Node (CPU) */ +cpus { + cpu0: cpu@0 { + device_type = "cpu"; + + ... + + vdd-supply = <&vcc>; + vbb-supply = <&abb_mpu>; + }; +}; + +/* OMAP OPP Supply with Class0 registers */ +opp_supply_mpu: opp_supply@4a003b20 { + compatible = "ti,omap5-opp-supply"; + reg = <0x4a003b20 0x8>; + ti,efuse-settings = < + /* uV offset */ + 1060000 0x0 + 1160000 0x4 + 1210000 0x8 + >; + ti,absolute-max-voltage-uv = <1500000>; +}; -- cgit 1.4.1 From 9a835fa6e47f27b1ae71390b6f12efce7335aaac Mon Sep 17 00:00:00 2001 From: Dave Gerlach Date: Thu, 14 Dec 2017 22:25:28 -0600 Subject: PM / OPP: Add ti-opp-supply driver Introduce a ti-opp-supply driver that will use new multiple regulator support that is part of the OPP core This is needed on TI platforms like DRA7/AM57 in order to control both CPU regulator and Adaptive Body Bias (ABB) regulator. These regulators must be scaled in sequence during an OPP transition depending on whether or not the frequency is being scaled up or down. This driver also implements AVS Class0 for these parts by looking up the required values from registers in the SoC and programming adjusted optimal voltage values for each OPP. Signed-off-by: Dave Gerlach Acked-by: Viresh Kumar Signed-off-by: Rafael J. Wysocki --- drivers/opp/Makefile | 1 + drivers/opp/ti-opp-supply.c | 425 ++++++++++++++++++++++++++++++++++++++++++++ 2 files changed, 426 insertions(+) create mode 100644 drivers/opp/ti-opp-supply.c diff --git a/drivers/opp/Makefile b/drivers/opp/Makefile index e70ceb406fe9..6ce6aefacc81 100644 --- a/drivers/opp/Makefile +++ b/drivers/opp/Makefile @@ -2,3 +2,4 @@ ccflags-$(CONFIG_DEBUG_DRIVER) := -DDEBUG obj-y += core.o cpu.o obj-$(CONFIG_OF) += of.o obj-$(CONFIG_DEBUG_FS) += debugfs.o +obj-$(CONFIG_ARM_TI_CPUFREQ) += ti-opp-supply.o diff --git a/drivers/opp/ti-opp-supply.c b/drivers/opp/ti-opp-supply.c new file mode 100644 index 000000000000..44dae3e51aac --- /dev/null +++ b/drivers/opp/ti-opp-supply.c @@ -0,0 +1,425 @@ +// SPDX-License-Identifier: GPL-2.0 +/* + * Copyright (C) 2016-2017 Texas Instruments Incorporated - http://www.ti.com/ + * Nishanth Menon + * Dave Gerlach + * + * TI OPP supply driver that provides override into the regulator control + * for generic opp core to handle devices with ABB regulator and/or + * SmartReflex Class0. + */ +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include + +/** + * struct ti_opp_supply_optimum_voltage_table - optimized voltage table + * @reference_uv: reference voltage (usually Nominal voltage) + * @optimized_uv: Optimized voltage from efuse + */ +struct ti_opp_supply_optimum_voltage_table { + unsigned int reference_uv; + unsigned int optimized_uv; +}; + +/** + * struct ti_opp_supply_data - OMAP specific opp supply data + * @vdd_table: Optimized voltage mapping table + * @num_vdd_table: number of entries in vdd_table + * @vdd_absolute_max_voltage_uv: absolute maximum voltage in UV for the supply + */ +struct ti_opp_supply_data { + struct ti_opp_supply_optimum_voltage_table *vdd_table; + u32 num_vdd_table; + u32 vdd_absolute_max_voltage_uv; +}; + +static struct ti_opp_supply_data opp_data; + +/** + * struct ti_opp_supply_of_data - device tree match data + * @flags: specific type of opp supply + * @efuse_voltage_mask: mask required for efuse register representing voltage + * @efuse_voltage_uv: Are the efuse entries in micro-volts? if not, assume + * milli-volts. + */ +struct ti_opp_supply_of_data { +#define OPPDM_EFUSE_CLASS0_OPTIMIZED_VOLTAGE BIT(1) +#define OPPDM_HAS_NO_ABB BIT(2) + const u8 flags; + const u32 efuse_voltage_mask; + const bool efuse_voltage_uv; +}; + +/** + * _store_optimized_voltages() - store optimized voltages + * @dev: ti opp supply device for which we need to store info + * @data: data specific to the device + * + * Picks up efuse based optimized voltages for VDD unique per device and + * stores it in internal data structure for use during transition requests. + * + * Return: If successful, 0, else appropriate error value. + */ +static int _store_optimized_voltages(struct device *dev, + struct ti_opp_supply_data *data) +{ + void __iomem *base; + struct property *prop; + struct resource *res; + const __be32 *val; + int proplen, i; + int ret = 0; + struct ti_opp_supply_optimum_voltage_table *table; + const struct ti_opp_supply_of_data *of_data = dev_get_drvdata(dev); + + /* pick up Efuse based voltages */ + res = platform_get_resource(to_platform_device(dev), IORESOURCE_MEM, 0); + if (!res) { + dev_err(dev, "Unable to get IO resource\n"); + ret = -ENODEV; + goto out_map; + } + + base = ioremap_nocache(res->start, resource_size(res)); + if (!base) { + dev_err(dev, "Unable to map Efuse registers\n"); + ret = -ENOMEM; + goto out_map; + } + + /* Fetch efuse-settings. */ + prop = of_find_property(dev->of_node, "ti,efuse-settings", NULL); + if (!prop) { + dev_err(dev, "No 'ti,efuse-settings' property found\n"); + ret = -EINVAL; + goto out; + } + + proplen = prop->length / sizeof(int); + data->num_vdd_table = proplen / 2; + /* Verify for corrupted OPP entries in dt */ + if (data->num_vdd_table * 2 * sizeof(int) != prop->length) { + dev_err(dev, "Invalid 'ti,efuse-settings'\n"); + ret = -EINVAL; + goto out; + } + + ret = of_property_read_u32(dev->of_node, "ti,absolute-max-voltage-uv", + &data->vdd_absolute_max_voltage_uv); + if (ret) { + dev_err(dev, "ti,absolute-max-voltage-uv is missing\n"); + ret = -EINVAL; + goto out; + } + + table = kzalloc(sizeof(*data->vdd_table) * + data->num_vdd_table, GFP_KERNEL); + if (!table) { + ret = -ENOMEM; + goto out; + } + data->vdd_table = table; + + val = prop->value; + for (i = 0; i < data->num_vdd_table; i++, table++) { + u32 efuse_offset; + u32 tmp; + + table->reference_uv = be32_to_cpup(val++); + efuse_offset = be32_to_cpup(val++); + + tmp = readl(base + efuse_offset); + tmp &= of_data->efuse_voltage_mask; + tmp >>= __ffs(of_data->efuse_voltage_mask); + + table->optimized_uv = of_data->efuse_voltage_uv ? tmp : + tmp * 1000; + + dev_dbg(dev, "[%d] efuse=0x%08x volt_table=%d vset=%d\n", + i, efuse_offset, table->reference_uv, + table->optimized_uv); + + /* + * Some older samples might not have optimized efuse + * Use reference voltage for those - just add debug message + * for them. + */ + if (!table->optimized_uv) { + dev_dbg(dev, "[%d] efuse=0x%08x volt_table=%d:vset0\n", + i, efuse_offset, table->reference_uv); + table->optimized_uv = table->reference_uv; + } + } +out: + iounmap(base); +out_map: + return ret; +} + +/** + * _free_optimized_voltages() - free resources for optvoltages + * @dev: device for which we need to free info + * @data: data specific to the device + */ +static void _free_optimized_voltages(struct device *dev, + struct ti_opp_supply_data *data) +{ + kfree(data->vdd_table); + data->vdd_table = NULL; + data->num_vdd_table = 0; +} + +/** + * _get_optimal_vdd_voltage() - Finds optimal voltage for the supply + * @dev: device for which we need to find info + * @data: data specific to the device + * @reference_uv: reference voltage (OPP voltage) for which we need value + * + * Return: if a match is found, return optimized voltage, else return + * reference_uv, also return reference_uv if no optimization is needed. + */ +static int _get_optimal_vdd_voltage(struct device *dev, + struct ti_opp_supply_data *data, + int reference_uv) +{ + int i; + struct ti_opp_supply_optimum_voltage_table *table; + + if (!data->num_vdd_table) + return reference_uv; + + table = data->vdd_table; + if (!table) + return -EINVAL; + + /* Find a exact match - this list is usually very small */ + for (i = 0; i < data->num_vdd_table; i++, table++) + if (table->reference_uv == reference_uv) + return table->optimized_uv; + + /* IF things are screwed up, we'd make a mess on console.. ratelimit */ + dev_err_ratelimited(dev, "%s: Failed optimized voltage match for %d\n", + __func__, reference_uv); + return reference_uv; +} + +static int _opp_set_voltage(struct device *dev, + struct dev_pm_opp_supply *supply, + int new_target_uv, struct regulator *reg, + char *reg_name) +{ + int ret; + unsigned long vdd_uv, uv_max; + + if (new_target_uv) + vdd_uv = new_target_uv; + else + vdd_uv = supply->u_volt; + + /* + * If we do have an absolute max voltage specified, then we should + * use that voltage instead to allow for cases where the voltage rails + * are ganged (example if we set the max for an opp as 1.12v, and + * the absolute max is 1.5v, for another rail to get 1.25v, it cannot + * be achieved if the regulator is constrainted to max of 1.12v, even + * if it can function at 1.25v + */ + if (opp_data.vdd_absolute_max_voltage_uv) + uv_max = opp_data.vdd_absolute_max_voltage_uv; + else + uv_max = supply->u_volt_max; + + if (vdd_uv > uv_max || + vdd_uv < supply->u_volt_min || + supply->u_volt_min > uv_max) { + dev_warn(dev, + "Invalid range voltages [Min:%lu target:%lu Max:%lu]\n", + supply->u_volt_min, vdd_uv, uv_max); + return -EINVAL; + } + + dev_dbg(dev, "%s scaling to %luuV[min %luuV max %luuV]\n", reg_name, + vdd_uv, supply->u_volt_min, + uv_max); + + ret = regulator_set_voltage_triplet(reg, + supply->u_volt_min, + vdd_uv, + uv_max); + if (ret) { + dev_err(dev, "%s failed for %luuV[min %luuV max %luuV]\n", + reg_name, vdd_uv, supply->u_volt_min, + uv_max); + return ret; + } + + return 0; +} + +/** + * ti_opp_supply_set_opp() - do the opp supply transition + * @data: information on regulators and new and old opps provided by + * opp core to use in transition + * + * Return: If successful, 0, else appropriate error value. + */ +int ti_opp_supply_set_opp(struct dev_pm_set_opp_data *data) +{ + struct dev_pm_opp_supply *old_supply_vdd = &data->old_opp.supplies[0]; + struct dev_pm_opp_supply *old_supply_vbb = &data->old_opp.supplies[1]; + struct dev_pm_opp_supply *new_supply_vdd = &data->new_opp.supplies[0]; + struct dev_pm_opp_supply *new_supply_vbb = &data->new_opp.supplies[1]; + struct device *dev = data->dev; + unsigned long old_freq = data->old_opp.rate, freq = data->new_opp.rate; + struct clk *clk = data->clk; + struct regulator *vdd_reg = data->regulators[0]; + struct regulator *vbb_reg = data->regulators[1]; + int vdd_uv; + int ret; + + vdd_uv = _get_optimal_vdd_voltage(dev, &opp_data, + new_supply_vbb->u_volt); + + /* Scaling up? Scale voltage before frequency */ + if (freq > old_freq) { + ret = _opp_set_voltage(dev, new_supply_vdd, vdd_uv, vdd_reg, + "vdd"); + if (ret) + goto restore_voltage; + + ret = _opp_set_voltage(dev, new_supply_vbb, 0, vbb_reg, "vbb"); + if (ret) + goto restore_voltage; + } + + /* Change frequency */ + dev_dbg(dev, "%s: switching OPP: %lu Hz --> %lu Hz\n", + __func__, old_freq, freq); + + ret = clk_set_rate(clk, freq); + if (ret) { + dev_err(dev, "%s: failed to set clock rate: %d\n", __func__, + ret); + goto restore_voltage; + } + + /* Scaling down? Scale voltage after frequency */ + if (freq < old_freq) { + ret = _opp_set_voltage(dev, new_supply_vbb, 0, vbb_reg, "vbb"); + if (ret) + goto restore_freq; + + ret = _opp_set_voltage(dev, new_supply_vdd, vdd_uv, vdd_reg, + "vdd"); + if (ret) + goto restore_freq; + } + + return 0; + +restore_freq: + ret = clk_set_rate(clk, old_freq); + if (ret) + dev_err(dev, "%s: failed to restore old-freq (%lu Hz)\n", + __func__, old_freq); +restore_voltage: + /* This shouldn't harm even if the voltages weren't updated earlier */ + if (old_supply_vdd->u_volt) { + ret = _opp_set_voltage(dev, old_supply_vbb, 0, vbb_reg, "vbb"); + if (ret) + return ret; + + ret = _opp_set_voltage(dev, old_supply_vdd, 0, vdd_reg, + "vdd"); + if (ret) + return ret; + } + + return ret; +} + +static const struct ti_opp_supply_of_data omap_generic_of_data = { +}; + +static const struct ti_opp_supply_of_data omap_omap5_of_data = { + .flags = OPPDM_EFUSE_CLASS0_OPTIMIZED_VOLTAGE, + .efuse_voltage_mask = 0xFFF, + .efuse_voltage_uv = false, +}; + +static const struct ti_opp_supply_of_data omap_omap5core_of_data = { + .flags = OPPDM_EFUSE_CLASS0_OPTIMIZED_VOLTAGE | OPPDM_HAS_NO_ABB, + .efuse_voltage_mask = 0xFFF, + .efuse_voltage_uv = false, +}; + +static const struct of_device_id ti_opp_supply_of_match[] = { + {.compatible = "ti,omap-opp-supply", .data = &omap_generic_of_data}, + {.compatible = "ti,omap5-opp-supply", .data = &omap_omap5_of_data}, + {.compatible = "ti,omap5-core-opp-supply", + .data = &omap_omap5core_of_data}, + {}, +}; +MODULE_DEVICE_TABLE(of, ti_opp_supply_of_match); + +static int ti_opp_supply_probe(struct platform_device *pdev) +{ + struct device *dev = &pdev->dev; + struct device *cpu_dev = get_cpu_device(0); + const struct of_device_id *match; + const struct ti_opp_supply_of_data *of_data; + int ret = 0; + + match = of_match_device(ti_opp_supply_of_match, dev); + if (!match) { + /* We do not expect this to happen */ + dev_err(dev, "%s: Unable to match device\n", __func__); + return -ENODEV; + } + if (!match->data) { + /* Again, unlikely.. but mistakes do happen */ + dev_err(dev, "%s: Bad data in match\n", __func__); + return -EINVAL; + } + of_data = match->data; + + dev_set_drvdata(dev, (void *)of_data); + + /* If we need optimized voltage */ + if (of_data->flags & OPPDM_EFUSE_CLASS0_OPTIMIZED_VOLTAGE) { + ret = _store_optimized_voltages(dev, &opp_data); + if (ret) + return ret; + } + + ret = PTR_ERR_OR_ZERO(dev_pm_opp_register_set_opp_helper(cpu_dev, + ti_opp_supply_set_opp)); + if (ret) + _free_optimized_voltages(dev, &opp_data); + + return ret; +} + +static struct platform_driver ti_opp_supply_driver = { + .probe = ti_opp_supply_probe, + .driver = { + .name = "ti_opp_supply", + .owner = THIS_MODULE, + .of_match_table = of_match_ptr(ti_opp_supply_of_match), + }, +}; +module_platform_driver(ti_opp_supply_driver); + +MODULE_DESCRIPTION("Texas Instruments OMAP OPP Supply driver"); +MODULE_AUTHOR("Texas Instruments Inc."); +MODULE_LICENSE("GPL v2"); -- cgit 1.4.1 From 3eff5f67a21997b4a86e0e5062fc72c2347e25bf Mon Sep 17 00:00:00 2001 From: Wei Yongjun Date: Wed, 20 Dec 2017 05:44:40 +0000 Subject: PM / OPP: Make local function ti_opp_supply_set_opp() static Fixes the following sparse warning: drivers/opp/ti-opp-supply.c:276:5: warning: symbol 'ti_opp_supply_set_opp' was not declared. Should it be static? Signed-off-by: Wei Yongjun Signed-off-by: Rafael J. Wysocki --- drivers/opp/ti-opp-supply.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/opp/ti-opp-supply.c b/drivers/opp/ti-opp-supply.c index 44dae3e51aac..370eff3acd8a 100644 --- a/drivers/opp/ti-opp-supply.c +++ b/drivers/opp/ti-opp-supply.c @@ -273,7 +273,7 @@ static int _opp_set_voltage(struct device *dev, * * Return: If successful, 0, else appropriate error value. */ -int ti_opp_supply_set_opp(struct dev_pm_set_opp_data *data) +static int ti_opp_supply_set_opp(struct dev_pm_set_opp_data *data) { struct dev_pm_opp_supply *old_supply_vdd = &data->old_opp.supplies[0]; struct dev_pm_opp_supply *old_supply_vbb = &data->old_opp.supplies[1]; -- cgit 1.4.1 From 2332bd04199353b06bf35f14f972d518907f08e0 Mon Sep 17 00:00:00 2001 From: Dong Aisheng Date: Sat, 23 Dec 2017 12:53:52 +0800 Subject: cpufreq: imx6q: switch to Use clk_bulk_get() to refine clk operations Use clk_bulk_get() to simplify the driver's clocks handling. Acked-by: Viresh Kumar Signed-off-by: Dong Aisheng Signed-off-by: Rafael J. Wysocki --- drivers/cpufreq/imx6q-cpufreq.c | 125 ++++++++++++++++++---------------------- 1 file changed, 56 insertions(+), 69 deletions(-) diff --git a/drivers/cpufreq/imx6q-cpufreq.c b/drivers/cpufreq/imx6q-cpufreq.c index d9b2c2de49c4..8bfb0775662b 100644 --- a/drivers/cpufreq/imx6q-cpufreq.c +++ b/drivers/cpufreq/imx6q-cpufreq.c @@ -25,15 +25,29 @@ static struct regulator *arm_reg; static struct regulator *pu_reg; static struct regulator *soc_reg; -static struct clk *arm_clk; -static struct clk *pll1_sys_clk; -static struct clk *pll1_sw_clk; -static struct clk *step_clk; -static struct clk *pll2_pfd2_396m_clk; - -/* clk used by i.MX6UL */ -static struct clk *pll2_bus_clk; -static struct clk *secondary_sel_clk; +enum IMX6_CPUFREQ_CLKS { + ARM, + PLL1_SYS, + STEP, + PLL1_SW, + PLL2_PFD2_396M, + /* MX6UL requires two more clks */ + PLL2_BUS, + SECONDARY_SEL, +}; +#define IMX6Q_CPUFREQ_CLK_NUM 5 +#define IMX6UL_CPUFREQ_CLK_NUM 7 + +static int num_clks; +static struct clk_bulk_data clks[] = { + { .id = "arm" }, + { .id = "pll1_sys" }, + { .id = "step" }, + { .id = "pll1_sw" }, + { .id = "pll2_pfd2_396m" }, + { .id = "pll2_bus" }, + { .id = "secondary_sel" }, +}; static struct device *cpu_dev; static bool free_opp; @@ -53,7 +67,7 @@ static int imx6q_set_target(struct cpufreq_policy *policy, unsigned int index) new_freq = freq_table[index].frequency; freq_hz = new_freq * 1000; - old_freq = clk_get_rate(arm_clk) / 1000; + old_freq = clk_get_rate(clks[ARM].clk) / 1000; opp = dev_pm_opp_find_freq_ceil(cpu_dev, &freq_hz); if (IS_ERR(opp)) { @@ -112,29 +126,31 @@ static int imx6q_set_target(struct cpufreq_policy *policy, unsigned int index) * voltage of 528MHz, so lower the CPU frequency to one * half before changing CPU frequency. */ - clk_set_rate(arm_clk, (old_freq >> 1) * 1000); - clk_set_parent(pll1_sw_clk, pll1_sys_clk); - if (freq_hz > clk_get_rate(pll2_pfd2_396m_clk)) - clk_set_parent(secondary_sel_clk, pll2_bus_clk); + clk_set_rate(clks[ARM].clk, (old_freq >> 1) * 1000); + clk_set_parent(clks[PLL1_SW].clk, clks[PLL1_SYS].clk); + if (freq_hz > clk_get_rate(clks[PLL2_PFD2_396M].clk)) + clk_set_parent(clks[SECONDARY_SEL].clk, + clks[PLL2_BUS].clk); else - clk_set_parent(secondary_sel_clk, pll2_pfd2_396m_clk); - clk_set_parent(step_clk, secondary_sel_clk); - clk_set_parent(pll1_sw_clk, step_clk); + clk_set_parent(clks[SECONDARY_SEL].clk, + clks[PLL2_PFD2_396M].clk); + clk_set_parent(clks[STEP].clk, clks[SECONDARY_SEL].clk); + clk_set_parent(clks[PLL1_SW].clk, clks[STEP].clk); } else { - clk_set_parent(step_clk, pll2_pfd2_396m_clk); - clk_set_parent(pll1_sw_clk, step_clk); - if (freq_hz > clk_get_rate(pll2_pfd2_396m_clk)) { - clk_set_rate(pll1_sys_clk, new_freq * 1000); - clk_set_parent(pll1_sw_clk, pll1_sys_clk); + clk_set_parent(clks[STEP].clk, clks[PLL2_PFD2_396M].clk); + clk_set_parent(clks[PLL1_SW].clk, clks[STEP].clk); + if (freq_hz > clk_get_rate(clks[PLL2_PFD2_396M].clk)) { + clk_set_rate(clks[PLL1_SYS].clk, new_freq * 1000); + clk_set_parent(clks[PLL1_SW].clk, clks[PLL1_SYS].clk); } else { /* pll1_sys needs to be enabled for divider rate change to work. */ pll1_sys_temp_enabled = true; - clk_prepare_enable(pll1_sys_clk); + clk_prepare_enable(clks[PLL1_SYS].clk); } } /* Ensure the arm clock divider is what we expect */ - ret = clk_set_rate(arm_clk, new_freq * 1000); + ret = clk_set_rate(clks[ARM].clk, new_freq * 1000); if (ret) { dev_err(cpu_dev, "failed to set clock rate: %d\n", ret); regulator_set_voltage_tol(arm_reg, volt_old, 0); @@ -143,7 +159,7 @@ static int imx6q_set_target(struct cpufreq_policy *policy, unsigned int index) /* PLL1 is only needed until after ARM-PODF is set. */ if (pll1_sys_temp_enabled) - clk_disable_unprepare(pll1_sys_clk); + clk_disable_unprepare(clks[PLL1_SYS].clk); /* scaling down? scale voltage after frequency */ if (new_freq < old_freq) { @@ -174,7 +190,7 @@ static int imx6q_cpufreq_init(struct cpufreq_policy *policy) { int ret; - policy->clk = arm_clk; + policy->clk = clks[ARM].clk; ret = cpufreq_generic_init(policy, freq_table, transition_latency); policy->suspend_freq = policy->max; @@ -266,28 +282,15 @@ static int imx6q_cpufreq_probe(struct platform_device *pdev) return -ENOENT; } - arm_clk = clk_get(cpu_dev, "arm"); - pll1_sys_clk = clk_get(cpu_dev, "pll1_sys"); - pll1_sw_clk = clk_get(cpu_dev, "pll1_sw"); - step_clk = clk_get(cpu_dev, "step"); - pll2_pfd2_396m_clk = clk_get(cpu_dev, "pll2_pfd2_396m"); - if (IS_ERR(arm_clk) || IS_ERR(pll1_sys_clk) || IS_ERR(pll1_sw_clk) || - IS_ERR(step_clk) || IS_ERR(pll2_pfd2_396m_clk)) { - dev_err(cpu_dev, "failed to get clocks\n"); - ret = -ENOENT; - goto put_clk; - } - if (of_machine_is_compatible("fsl,imx6ul") || - of_machine_is_compatible("fsl,imx6ull")) { - pll2_bus_clk = clk_get(cpu_dev, "pll2_bus"); - secondary_sel_clk = clk_get(cpu_dev, "secondary_sel"); - if (IS_ERR(pll2_bus_clk) || IS_ERR(secondary_sel_clk)) { - dev_err(cpu_dev, "failed to get clocks specific to imx6ul\n"); - ret = -ENOENT; - goto put_clk; - } - } + of_machine_is_compatible("fsl,imx6ull")) + num_clks = IMX6UL_CPUFREQ_CLK_NUM; + else + num_clks = IMX6Q_CPUFREQ_CLK_NUM; + + ret = clk_bulk_get(cpu_dev, num_clks, clks); + if (ret) + goto put_node; arm_reg = regulator_get(cpu_dev, "arm"); pu_reg = regulator_get_optional(cpu_dev, "pu"); @@ -424,22 +427,11 @@ put_reg: regulator_put(pu_reg); if (!IS_ERR(soc_reg)) regulator_put(soc_reg); -put_clk: - if (!IS_ERR(arm_clk)) - clk_put(arm_clk); - if (!IS_ERR(pll1_sys_clk)) - clk_put(pll1_sys_clk); - if (!IS_ERR(pll1_sw_clk)) - clk_put(pll1_sw_clk); - if (!IS_ERR(step_clk)) - clk_put(step_clk); - if (!IS_ERR(pll2_pfd2_396m_clk)) - clk_put(pll2_pfd2_396m_clk); - if (!IS_ERR(pll2_bus_clk)) - clk_put(pll2_bus_clk); - if (!IS_ERR(secondary_sel_clk)) - clk_put(secondary_sel_clk); + + clk_bulk_put(num_clks, clks); +put_node: of_node_put(np); + return ret; } @@ -453,13 +445,8 @@ static int imx6q_cpufreq_remove(struct platform_device *pdev) if (!IS_ERR(pu_reg)) regulator_put(pu_reg); regulator_put(soc_reg); - clk_put(arm_clk); - clk_put(pll1_sys_clk); - clk_put(pll1_sw_clk); - clk_put(step_clk); - clk_put(pll2_pfd2_396m_clk); - clk_put(pll2_bus_clk); - clk_put(secondary_sel_clk); + + clk_bulk_put(num_clks, clks); return 0; } -- cgit 1.4.1 From ac89c400ebb146604e718b3fa168c15592e73a8c Mon Sep 17 00:00:00 2001 From: Viresh Kumar Date: Tue, 2 Jan 2018 10:51:34 +0530 Subject: cpu_cooling: Remove static-power related documentation commit 84fe2cab4859 ("cpu_cooling: Drop static-power related stuff") removed support for static-power in kernel, but it missed reflecting the same in documentation. Remove the static power related documentation bits as well. Reported-by: Javi Merino Signed-off-by: Viresh Kumar Signed-off-by: Rafael J. Wysocki --- Documentation/thermal/cpu-cooling-api.txt | 82 +------------------------------ 1 file changed, 2 insertions(+), 80 deletions(-) diff --git a/Documentation/thermal/cpu-cooling-api.txt b/Documentation/thermal/cpu-cooling-api.txt index 7a1c89db0419..7df567eaea1a 100644 --- a/Documentation/thermal/cpu-cooling-api.txt +++ b/Documentation/thermal/cpu-cooling-api.txt @@ -44,16 +44,14 @@ the user. The registration APIs returns the cooling device pointer. 2. Power models The power API registration functions provide a simple power model for -CPUs. The current power is calculated as dynamic + (optionally) -static power. This power model requires that the operating-points of +CPUs. The current power is calculated as dynamic power (static power isn't +supported currently). This power model requires that the operating-points of the CPUs are registered using the kernel's opp library and the `cpufreq_frequency_table` is assigned to the `struct device` of the cpu. If you are using CONFIG_CPUFREQ_DT then the `cpufreq_frequency_table` should already be assigned to the cpu device. -2.1 Dynamic power - The dynamic power consumption of a processor depends on many factors. For a given processor implementation the primary factors are: @@ -92,79 +90,3 @@ mW/MHz/uVolt^2. Typical values for mobile CPUs might lie in range from 100 to 500. For reference, the approximate values for the SoC in ARM's Juno Development Platform are 530 for the Cortex-A57 cluster and 140 for the Cortex-A53 cluster. - - -2.2 Static power - -Static leakage power consumption depends on a number of factors. For a -given circuit implementation the primary factors are: - -- Time the circuit spends in each 'power state' -- Temperature -- Operating voltage -- Process grade - -The time the circuit spends in each 'power state' for a given -evaluation period at first order means OFF or ON. However, -'retention' states can also be supported that reduce power during -inactive periods without loss of context. - -Note: The visibility of state entries to the OS can vary, according to -platform specifics, and this can then impact the accuracy of a model -based on OS state information alone. It might be possible in some -cases to extract more accurate information from system resources. - -The temperature, operating voltage and process 'grade' (slow to fast) -of the circuit are all significant factors in static leakage power -consumption. All of these have complex relationships to static power. - -Circuit implementation specific factors include the chosen silicon -process as well as the type, number and size of transistors in both -the logic gates and any RAM elements included. - -The static power consumption modelling must take into account the -power managed regions that are implemented. Taking the example of an -ARM processor cluster, the modelling would take into account whether -each CPU can be powered OFF separately or if only a single power -region is implemented for the complete cluster. - -In one view, there are others, a static power consumption model can -then start from a set of reference values for each power managed -region (e.g. CPU, Cluster/L2) in each state (e.g. ON, OFF) at an -arbitrary process grade, voltage and temperature point. These values -are then scaled for all of the following: the time in each state, the -process grade, the current temperature and the operating voltage. -However, since both implementation specific and complex relationships -dominate the estimate, the appropriate interface to the model from the -cpu cooling device is to provide a function callback that calculates -the static power in this platform. When registering the cpu cooling -device pass a function pointer that follows the `get_static_t` -prototype: - - int plat_get_static(cpumask_t *cpumask, int interval, - unsigned long voltage, u32 &power); - -`cpumask` is the cpumask of the cpus involved in the calculation. -`voltage` is the voltage at which they are operating. The function -should calculate the average static power for the last `interval` -milliseconds. It returns 0 on success, -E* on error. If it -succeeds, it should store the static power in `power`. Reading the -temperature of the cpus described by `cpumask` is left for -plat_get_static() to do as the platform knows best which thermal -sensor is closest to the cpu. - -If `plat_static_func` is NULL, static power is considered to be -negligible for this platform and only dynamic power is considered. - -The platform specific callback can then use any combination of tables -and/or equations to permute the estimated value. Process grade -information is not passed to the model since access to such data, from -on-chip measurement capability or manufacture time data, is platform -specific. - -Note: the significance of static power for CPUs in comparison to -dynamic power is highly dependent on implementation. Given the -potential complexity in implementation, the importance and accuracy of -its inclusion when using cpu cooling devices should be assessed on a -case by case basis. - -- cgit 1.4.1 From ee1f4a7dafa997816ff3de96155c6f3edc21c1e6 Mon Sep 17 00:00:00 2001 From: "Gautham R. Shenoy" Date: Wed, 13 Dec 2017 12:27:39 +0530 Subject: powernv-cpufreq: Add helper to extract pstate from PMSR On POWERNV platform, the fields for pstates in the Power Management Status Register (PMSR) and the Power Management Control Register (PMCR) are 8-bits wide. On POWER8 the pstates are negatively numbered while on POWER9 they are positively numbered. The device-tree exports pstates as 32-bit entries. The device-tree implementation sign-extends the 8-bit pstate values to obtain the corresponding 32-bit entry. Eg: On POWER8, a pstate value 0x82 [-126] is represented in the device-tree as 0xfffffff82 while on POWER9, the same value 0x82 [130] is represented in the device-tree as 0x00000082. The powernv-cpufreq driver implementation represents pstates using the integer type. In multiple places in the driver, the code interprets the pstates extracted from the PMSR as a signed byte and assigns it to a integer variable to get the sign-extention. On POWER9 platforms which have greater than 128 pstates, this results in the driver performing incorrect sign-extention, and thereby treating a legitimate pstate (say 130) as an invalid pstates (since it is interpreted as -126). This patch fixes the issue by implementing a helper function to extract Pstates from PMSR register, and correctly sign-extend it to be consistent with the values provided by the device-tree. Signed-off-by: Gautham R. Shenoy Acked-by: Balbir Singh Signed-off-by: Rafael J. Wysocki --- drivers/cpufreq/powernv-cpufreq.c | 37 +++++++++++++++++++++++-------------- 1 file changed, 23 insertions(+), 14 deletions(-) diff --git a/drivers/cpufreq/powernv-cpufreq.c b/drivers/cpufreq/powernv-cpufreq.c index b6d7c4c98d0a..f46b60fb3084 100644 --- a/drivers/cpufreq/powernv-cpufreq.c +++ b/drivers/cpufreq/powernv-cpufreq.c @@ -41,11 +41,9 @@ #define POWERNV_MAX_PSTATES 256 #define PMSR_PSAFE_ENABLE (1UL << 30) #define PMSR_SPR_EM_DISABLE (1UL << 31) -#define PMSR_MAX(x) ((x >> 32) & 0xFF) +#define MAX_PSTATE_SHIFT 32 #define LPSTATE_SHIFT 48 #define GPSTATE_SHIFT 56 -#define GET_LPSTATE(x) (((x) >> LPSTATE_SHIFT) & 0xFF) -#define GET_GPSTATE(x) (((x) >> GPSTATE_SHIFT) & 0xFF) #define MAX_RAMP_DOWN_TIME 5120 /* @@ -94,6 +92,7 @@ struct global_pstate_info { }; static struct cpufreq_frequency_table powernv_freqs[POWERNV_MAX_PSTATES+1]; +u32 pstate_sign_prefix; static bool rebooting, throttled, occ_reset; static const char * const throttle_reason[] = { @@ -148,6 +147,20 @@ static struct powernv_pstate_info { bool wof_enabled; } powernv_pstate_info; +static inline int extract_pstate(u64 pmsr_val, unsigned int shift) +{ + int ret = ((pmsr_val >> shift) & 0xFF); + + if (!ret) + return ret; + + return (pstate_sign_prefix | ret); +} + +#define extract_local_pstate(x) extract_pstate(x, LPSTATE_SHIFT) +#define extract_global_pstate(x) extract_pstate(x, GPSTATE_SHIFT) +#define extract_max_pstate(x) extract_pstate(x, MAX_PSTATE_SHIFT) + /* Use following macros for conversions between pstate_id and index */ static inline int idx_to_pstate(unsigned int i) { @@ -278,6 +291,9 @@ next: powernv_pstate_info.nr_pstates = nr_pstates; pr_debug("NR PStates %d\n", nr_pstates); + + pstate_sign_prefix = pstate_min & ~0xFF; + for (i = 0; i < nr_pstates; i++) { u32 id = be32_to_cpu(pstate_ids[i]); u32 freq = be32_to_cpu(pstate_freqs[i]); @@ -438,17 +454,10 @@ struct powernv_smp_call_data { static void powernv_read_cpu_freq(void *arg) { unsigned long pmspr_val; - s8 local_pstate_id; struct powernv_smp_call_data *freq_data = arg; pmspr_val = get_pmspr(SPRN_PMSR); - - /* - * The local pstate id corresponds bits 48..55 in the PMSR. - * Note: Watch out for the sign! - */ - local_pstate_id = (pmspr_val >> 48) & 0xFF; - freq_data->pstate_id = local_pstate_id; + freq_data->pstate_id = extract_local_pstate(pmspr_val); freq_data->freq = pstate_id_to_freq(freq_data->pstate_id); pr_debug("cpu %d pmsr %016lX pstate_id %d frequency %d kHz\n", @@ -522,7 +531,7 @@ static void powernv_cpufreq_throttle_check(void *data) chip = this_cpu_read(chip_info); /* Check for Pmax Capping */ - pmsr_pmax = (s8)PMSR_MAX(pmsr); + pmsr_pmax = extract_max_pstate(pmsr); pmsr_pmax_idx = pstate_to_idx(pmsr_pmax); if (pmsr_pmax_idx != powernv_pstate_info.max) { if (chip->throttled) @@ -645,8 +654,8 @@ void gpstate_timer_handler(struct timer_list *t) * value. Hence, read from PMCR to get correct data. */ val = get_pmspr(SPRN_PMCR); - freq_data.gpstate_id = (s8)GET_GPSTATE(val); - freq_data.pstate_id = (s8)GET_LPSTATE(val); + freq_data.gpstate_id = extract_global_pstate(val); + freq_data.pstate_id = extract_local_pstate(val); if (freq_data.gpstate_id == freq_data.pstate_id) { reset_gpstates(policy); spin_unlock(&gpstates->gpstate_lock); -- cgit 1.4.1 From 332f0a01f0dd669dcd208e4e9666d80dffd62e7b Mon Sep 17 00:00:00 2001 From: "Gautham R. Shenoy" Date: Wed, 13 Dec 2017 12:27:40 +0530 Subject: powernv-cpufreq: Fix pstate_to_idx() to handle non-continguous pstates The code in powernv-cpufreq, makes the following two assumptions which are not guaranteed by the device-tree bindings: 1) Pstate ids are continguous: This is used in pstate_to_idx() to obtain the reverse map from a pstate to it's corresponding entry into the cpufreq frequency table. 2) Every Pstate should always lie between the max and the min pstates that are explicitly reported in the device tree: This is used to determine whether a pstate reported by the PMSR is out of bounds. Both these assumptions are unwarranted and can change on future platforms. In this patch, we maintain the reverse map from a pstate to it's index in the cpufreq frequency table and use this in pstate_to_idx(). This does away with the assumptions (1) mentioned above, and will work with non continguous pstate ids. If no entry exists for a particular pstate, then such a pstate is treated as being out of bounds. This gets rid of assumption (2). On all the existing platforms, where the pstates are 8-bit long values, the new implementation of pstate_to_idx() takes constant time. Signed-off-by: Gautham R. Shenoy Signed-off-by: Rafael J. Wysocki --- drivers/cpufreq/powernv-cpufreq.c | 85 +++++++++++++++++++++++++++++---------- 1 file changed, 63 insertions(+), 22 deletions(-) diff --git a/drivers/cpufreq/powernv-cpufreq.c b/drivers/cpufreq/powernv-cpufreq.c index f46b60fb3084..8e3dbcaee286 100644 --- a/drivers/cpufreq/powernv-cpufreq.c +++ b/drivers/cpufreq/powernv-cpufreq.c @@ -29,6 +29,7 @@ #include #include #include +#include #include #include @@ -38,7 +39,8 @@ #include #include -#define POWERNV_MAX_PSTATES 256 +#define POWERNV_MAX_PSTATES_ORDER 8 +#define POWERNV_MAX_PSTATES (1UL << (POWERNV_MAX_PSTATES_ORDER)) #define PMSR_PSAFE_ENABLE (1UL << 30) #define PMSR_SPR_EM_DISABLE (1UL << 31) #define MAX_PSTATE_SHIFT 32 @@ -92,6 +94,27 @@ struct global_pstate_info { }; static struct cpufreq_frequency_table powernv_freqs[POWERNV_MAX_PSTATES+1]; + +DEFINE_HASHTABLE(pstate_revmap, POWERNV_MAX_PSTATES_ORDER); +/** + * struct pstate_idx_revmap_data: Entry in the hashmap pstate_revmap + * indexed by a function of pstate id. + * + * @pstate_id: pstate id for this entry. + * + * @cpufreq_table_idx: Index into the powernv_freqs + * cpufreq_frequency_table for frequency + * corresponding to pstate_id. + * + * @hentry: hlist_node that hooks this entry into the pstate_revmap + * hashtable + */ +struct pstate_idx_revmap_data { + int pstate_id; + unsigned int cpufreq_table_idx; + struct hlist_node hentry; +}; + u32 pstate_sign_prefix; static bool rebooting, throttled, occ_reset; @@ -161,39 +184,47 @@ static inline int extract_pstate(u64 pmsr_val, unsigned int shift) #define extract_global_pstate(x) extract_pstate(x, GPSTATE_SHIFT) #define extract_max_pstate(x) extract_pstate(x, MAX_PSTATE_SHIFT) -/* Use following macros for conversions between pstate_id and index */ +/* Use following functions for conversions between pstate_id and index */ + +/** + * idx_to_pstate : Returns the pstate id corresponding to the + * frequency in the cpufreq frequency table + * powernv_freqs indexed by @i. + * + * If @i is out of bound, this will return the pstate + * corresponding to the nominal frequency. + */ static inline int idx_to_pstate(unsigned int i) { if (unlikely(i >= powernv_pstate_info.nr_pstates)) { - pr_warn_once("index %u is out of bound\n", i); + pr_warn_once("idx_to_pstate: index %u is out of bound\n", i); return powernv_freqs[powernv_pstate_info.nominal].driver_data; } return powernv_freqs[i].driver_data; } -static inline unsigned int pstate_to_idx(int pstate) +/** + * pstate_to_idx : Returns the index in the cpufreq frequencytable + * powernv_freqs for the frequency whose corresponding + * pstate id is @pstate. + * + * If no frequency corresponding to @pstate is found, + * this will return the index of the nominal + * frequency. + */ +static unsigned int pstate_to_idx(int pstate) { - int min = powernv_freqs[powernv_pstate_info.min].driver_data; - int max = powernv_freqs[powernv_pstate_info.max].driver_data; + unsigned int key = pstate % POWERNV_MAX_PSTATES; + struct pstate_idx_revmap_data *revmap_data; - if (min > 0) { - if (unlikely((pstate < max) || (pstate > min))) { - pr_warn_once("pstate %d is out of bound\n", pstate); - return powernv_pstate_info.nominal; - } - } else { - if (unlikely((pstate > max) || (pstate < min))) { - pr_warn_once("pstate %d is out of bound\n", pstate); - return powernv_pstate_info.nominal; - } + hash_for_each_possible(pstate_revmap, revmap_data, hentry, key) { + if (revmap_data->pstate_id == pstate) + return revmap_data->cpufreq_table_idx; } - /* - * abs() is deliberately used so that is works with - * both monotonically increasing and decreasing - * pstate values - */ - return abs(pstate - idx_to_pstate(powernv_pstate_info.max)); + + pr_warn_once("pstate_to_idx: pstate %d not found\n", pstate); + return powernv_pstate_info.nominal; } static inline void reset_gpstates(struct cpufreq_policy *policy) @@ -297,11 +328,21 @@ next: for (i = 0; i < nr_pstates; i++) { u32 id = be32_to_cpu(pstate_ids[i]); u32 freq = be32_to_cpu(pstate_freqs[i]); + struct pstate_idx_revmap_data *revmap_data; + unsigned int key; pr_debug("PState id %d freq %d MHz\n", id, freq); powernv_freqs[i].frequency = freq * 1000; /* kHz */ powernv_freqs[i].driver_data = id; + revmap_data = (struct pstate_idx_revmap_data *) + kmalloc(sizeof(*revmap_data), GFP_KERNEL); + + revmap_data->pstate_id = id; + revmap_data->cpufreq_table_idx = i; + key = id % POWERNV_MAX_PSTATES; + hash_add(pstate_revmap, &revmap_data->hentry, key); + if (id == pstate_max) powernv_pstate_info.max = i; else if (id == pstate_nominal) -- cgit 1.4.1 From 967b87fd81d513de7aac247320f02b1645f9ca64 Mon Sep 17 00:00:00 2001 From: "Gautham R. Shenoy" Date: Wed, 13 Dec 2017 12:27:41 +0530 Subject: powernv-cpufreq: Treat pstates as opaque 8-bit values On POWER8 and POWER9, the PMSR and the PMCR registers define pstates to be 8-bit wide values. The device-tree exports pstates as 32-bit wide values of which the lower byte is the actual pstate. The current implementation in the kernel treats pstates as integer type, since it used to use the sign of the pstate for performing some boundary-checks. This is no longer required after the patch "powernv-cpufreq: Fix pstate_to_idx() to handle non-continguous pstates". So, in this patch, we modify the powernv-cpufreq driver to uniformly treat pstates as opaque 8-bit values obtained from the device-tree or the PMCR. This simplifies the extract_pstate() helper function since we no longer no longer require to worry about the sign-extentions. Signed-off-by: Gautham R. Shenoy Signed-off-by: Rafael J. Wysocki --- drivers/cpufreq/powernv-cpufreq.c | 47 ++++++++++++++++----------------------- 1 file changed, 19 insertions(+), 28 deletions(-) diff --git a/drivers/cpufreq/powernv-cpufreq.c b/drivers/cpufreq/powernv-cpufreq.c index 8e3dbcaee286..8a4e2ce0804c 100644 --- a/drivers/cpufreq/powernv-cpufreq.c +++ b/drivers/cpufreq/powernv-cpufreq.c @@ -110,12 +110,11 @@ DEFINE_HASHTABLE(pstate_revmap, POWERNV_MAX_PSTATES_ORDER); * hashtable */ struct pstate_idx_revmap_data { - int pstate_id; + u8 pstate_id; unsigned int cpufreq_table_idx; struct hlist_node hentry; }; -u32 pstate_sign_prefix; static bool rebooting, throttled, occ_reset; static const char * const throttle_reason[] = { @@ -170,14 +169,9 @@ static struct powernv_pstate_info { bool wof_enabled; } powernv_pstate_info; -static inline int extract_pstate(u64 pmsr_val, unsigned int shift) +static inline u8 extract_pstate(u64 pmsr_val, unsigned int shift) { - int ret = ((pmsr_val >> shift) & 0xFF); - - if (!ret) - return ret; - - return (pstate_sign_prefix | ret); + return ((pmsr_val >> shift) & 0xFF); } #define extract_local_pstate(x) extract_pstate(x, LPSTATE_SHIFT) @@ -194,7 +188,7 @@ static inline int extract_pstate(u64 pmsr_val, unsigned int shift) * If @i is out of bound, this will return the pstate * corresponding to the nominal frequency. */ -static inline int idx_to_pstate(unsigned int i) +static inline u8 idx_to_pstate(unsigned int i) { if (unlikely(i >= powernv_pstate_info.nr_pstates)) { pr_warn_once("idx_to_pstate: index %u is out of bound\n", i); @@ -213,7 +207,7 @@ static inline int idx_to_pstate(unsigned int i) * this will return the index of the nominal * frequency. */ -static unsigned int pstate_to_idx(int pstate) +static unsigned int pstate_to_idx(u8 pstate) { unsigned int key = pstate % POWERNV_MAX_PSTATES; struct pstate_idx_revmap_data *revmap_data; @@ -223,7 +217,7 @@ static unsigned int pstate_to_idx(int pstate) return revmap_data->cpufreq_table_idx; } - pr_warn_once("pstate_to_idx: pstate %d not found\n", pstate); + pr_warn_once("pstate_to_idx: pstate 0x%x not found\n", pstate); return powernv_pstate_info.nominal; } @@ -291,7 +285,7 @@ static int init_powernv_pstates(void) powernv_pstate_info.wof_enabled = true; next: - pr_info("cpufreq pstate min %d nominal %d max %d\n", pstate_min, + pr_info("cpufreq pstate min 0x%x nominal 0x%x max 0x%x\n", pstate_min, pstate_nominal, pstate_max); pr_info("Workload Optimized Frequency is %s in the platform\n", (powernv_pstate_info.wof_enabled) ? "enabled" : "disabled"); @@ -323,8 +317,6 @@ next: powernv_pstate_info.nr_pstates = nr_pstates; pr_debug("NR PStates %d\n", nr_pstates); - pstate_sign_prefix = pstate_min & ~0xFF; - for (i = 0; i < nr_pstates; i++) { u32 id = be32_to_cpu(pstate_ids[i]); u32 freq = be32_to_cpu(pstate_freqs[i]); @@ -333,14 +325,14 @@ next: pr_debug("PState id %d freq %d MHz\n", id, freq); powernv_freqs[i].frequency = freq * 1000; /* kHz */ - powernv_freqs[i].driver_data = id; + powernv_freqs[i].driver_data = id & 0xFF; revmap_data = (struct pstate_idx_revmap_data *) kmalloc(sizeof(*revmap_data), GFP_KERNEL); - revmap_data->pstate_id = id; + revmap_data->pstate_id = id & 0xFF; revmap_data->cpufreq_table_idx = i; - key = id % POWERNV_MAX_PSTATES; + key = (revmap_data->pstate_id) % POWERNV_MAX_PSTATES; hash_add(pstate_revmap, &revmap_data->hentry, key); if (id == pstate_max) @@ -364,14 +356,13 @@ next: } /* Returns the CPU frequency corresponding to the pstate_id. */ -static unsigned int pstate_id_to_freq(int pstate_id) +static unsigned int pstate_id_to_freq(u8 pstate_id) { int i; i = pstate_to_idx(pstate_id); if (i >= powernv_pstate_info.nr_pstates || i < 0) { - pr_warn("PState id %d outside of PState table, " - "reporting nominal id %d instead\n", + pr_warn("PState id 0x%x outside of PState table, reporting nominal id 0x%x instead\n", pstate_id, idx_to_pstate(powernv_pstate_info.nominal)); i = powernv_pstate_info.nominal; } @@ -477,8 +468,8 @@ static inline void set_pmspr(unsigned long sprn, unsigned long val) */ struct powernv_smp_call_data { unsigned int freq; - int pstate_id; - int gpstate_id; + u8 pstate_id; + u8 gpstate_id; }; /* @@ -501,9 +492,9 @@ static void powernv_read_cpu_freq(void *arg) freq_data->pstate_id = extract_local_pstate(pmspr_val); freq_data->freq = pstate_id_to_freq(freq_data->pstate_id); - pr_debug("cpu %d pmsr %016lX pstate_id %d frequency %d kHz\n", - raw_smp_processor_id(), pmspr_val, freq_data->pstate_id, - freq_data->freq); + pr_debug("cpu %d pmsr %016lX pstate_id 0x%x frequency %d kHz\n", + raw_smp_processor_id(), pmspr_val, freq_data->pstate_id, + freq_data->freq); } /* @@ -565,7 +556,7 @@ static void powernv_cpufreq_throttle_check(void *data) struct chip *chip; unsigned int cpu = smp_processor_id(); unsigned long pmsr; - int pmsr_pmax; + u8 pmsr_pmax; unsigned int pmsr_pmax_idx; pmsr = get_pmspr(SPRN_PMSR); @@ -579,7 +570,7 @@ static void powernv_cpufreq_throttle_check(void *data) goto next; chip->throttled = true; if (pmsr_pmax_idx > powernv_pstate_info.nominal) { - pr_warn_once("CPU %d on Chip %u has Pmax(%d) reduced below nominal frequency(%d)\n", + pr_warn_once("CPU %d on Chip %u has Pmax(0x%x) reduced below that of nominal frequency(0x%x)\n", cpu, chip->id, pmsr_pmax, idx_to_pstate(powernv_pstate_info.nominal)); chip->throttle_sub_turbo++; -- cgit 1.4.1 From d476ec4f7f5aba47b0a570cbf659d7330d7e71cf Mon Sep 17 00:00:00 2001 From: Viresh Kumar Date: Thu, 4 Jan 2018 08:53:54 +0530 Subject: cpufreq: stats: Change return type of cpufreq_stats_update() as void It always returns 0 and none of its callers check its return value. Make it return void. Signed-off-by: Viresh Kumar Signed-off-by: Rafael J. Wysocki --- drivers/cpufreq/cpufreq_stats.c | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-) diff --git a/drivers/cpufreq/cpufreq_stats.c b/drivers/cpufreq/cpufreq_stats.c index 1e55b5790853..1572129844a5 100644 --- a/drivers/cpufreq/cpufreq_stats.c +++ b/drivers/cpufreq/cpufreq_stats.c @@ -27,7 +27,7 @@ struct cpufreq_stats { unsigned int *trans_table; }; -static int cpufreq_stats_update(struct cpufreq_stats *stats) +static void cpufreq_stats_update(struct cpufreq_stats *stats) { unsigned long long cur_time = get_jiffies_64(); @@ -35,7 +35,6 @@ static int cpufreq_stats_update(struct cpufreq_stats *stats) stats->time_in_state[stats->last_index] += cur_time - stats->last_time; stats->last_time = cur_time; spin_unlock(&cpufreq_stats_lock); - return 0; } static void cpufreq_stats_clear_table(struct cpufreq_stats *stats) -- cgit 1.4.1 From c9619bb293c9ab758ba298f039cf4b820dd8436f Mon Sep 17 00:00:00 2001 From: Anson Huang Date: Mon, 8 Jan 2018 10:04:50 +0800 Subject: ARM: dts: imx6ul: add 696MHz operating point Add 696MHz operating point according to datasheet (Rev. 0, 12/2015). Signed-off-by: Anson Huang Reviewed-by: Fabio Estevam Acked-by: Dong Aisheng Signed-off-by: Rafael J. Wysocki --- arch/arm/boot/dts/imx6ul.dtsi | 2 ++ 1 file changed, 2 insertions(+) diff --git a/arch/arm/boot/dts/imx6ul.dtsi b/arch/arm/boot/dts/imx6ul.dtsi index d5181f85ca9c..963e1698fe1d 100644 --- a/arch/arm/boot/dts/imx6ul.dtsi +++ b/arch/arm/boot/dts/imx6ul.dtsi @@ -68,12 +68,14 @@ clock-latency = <61036>; /* two CLK32 periods */ operating-points = < /* kHz uV */ + 696000 1275000 528000 1175000 396000 1025000 198000 950000 >; fsl,soc-operating-points = < /* KHz uV */ + 696000 1275000 528000 1175000 396000 1175000 198000 1175000 -- cgit 1.4.1 From 5028f5d2b38ea68531d6b265b64e1741a141a828 Mon Sep 17 00:00:00 2001 From: Anson Huang Date: Mon, 8 Jan 2018 10:04:51 +0800 Subject: cpufreq: imx6q: add 696MHz operating point for i.mx6ul Add 696MHz operating point for i.MX6UL, only for those parts with speed grading fuse set to 2b'10 supports 696MHz operating point, so, speed grading check is also added for i.MX6UL in this patch, the clock tree for each operating point are as below: 696MHz: pll1 696000000 pll1_bypass 696000000 pll1_sys 696000000 pll1_sw 696000000 arm 696000000 528MHz: pll2 528000000 pll2_bypass 528000000 pll2_bus 528000000 ca7_secondary_sel 528000000 step 528000000 pll1_sw 528000000 arm 528000000 396MHz: pll2_pfd2_396m 396000000 ca7_secondary_sel 396000000 step 396000000 pll1_sw 396000000 arm 396000000 198MHz: pll2_pfd2_396m 396000000 ca7_secondary_sel 396000000 step 396000000 pll1_sw 396000000 arm 198000000 Signed-off-by: Anson Huang Reviewed-by: Fabio Estevam Acked-by: Viresh Kumar Acked-by: Dong Aisheng Signed-off-by: Rafael J. Wysocki --- drivers/cpufreq/imx6q-cpufreq.c | 46 ++++++++++++++++++++++++++++++++++++++++- 1 file changed, 45 insertions(+), 1 deletion(-) diff --git a/drivers/cpufreq/imx6q-cpufreq.c b/drivers/cpufreq/imx6q-cpufreq.c index 8bfb0775662b..741f22e5cee3 100644 --- a/drivers/cpufreq/imx6q-cpufreq.c +++ b/drivers/cpufreq/imx6q-cpufreq.c @@ -136,6 +136,10 @@ static int imx6q_set_target(struct cpufreq_policy *policy, unsigned int index) clks[PLL2_PFD2_396M].clk); clk_set_parent(clks[STEP].clk, clks[SECONDARY_SEL].clk); clk_set_parent(clks[PLL1_SW].clk, clks[STEP].clk); + if (freq_hz > clk_get_rate(clks[PLL2_BUS].clk)) { + clk_set_rate(clks[PLL1_SYS].clk, new_freq * 1000); + clk_set_parent(clks[PLL1_SW].clk, clks[PLL1_SYS].clk); + } } else { clk_set_parent(clks[STEP].clk, clks[PLL2_PFD2_396M].clk); clk_set_parent(clks[PLL1_SW].clk, clks[STEP].clk); @@ -260,6 +264,43 @@ put_node: of_node_put(np); } +#define OCOTP_CFG3_6UL_SPEED_696MHZ 0x2 + +static void imx6ul_opp_check_speed_grading(struct device *dev) +{ + struct device_node *np; + void __iomem *base; + u32 val; + + np = of_find_compatible_node(NULL, NULL, "fsl,imx6ul-ocotp"); + if (!np) + return; + + base = of_iomap(np, 0); + if (!base) { + dev_err(dev, "failed to map ocotp\n"); + goto put_node; + } + + /* + * Speed GRADING[1:0] defines the max speed of ARM: + * 2b'00: Reserved; + * 2b'01: 528000000Hz; + * 2b'10: 696000000Hz; + * 2b'11: Reserved; + * We need to set the max speed of ARM according to fuse map. + */ + val = readl_relaxed(base + OCOTP_CFG3); + val >>= OCOTP_CFG3_SPEED_SHIFT; + val &= 0x3; + if (val != OCOTP_CFG3_6UL_SPEED_696MHZ) + if (dev_pm_opp_disable(dev, 696000000)) + dev_warn(dev, "failed to disable 696MHz OPP\n"); + iounmap(base); +put_node: + of_node_put(np); +} + static int imx6q_cpufreq_probe(struct platform_device *pdev) { struct device_node *np; @@ -314,7 +355,10 @@ static int imx6q_cpufreq_probe(struct platform_device *pdev) goto put_reg; } - imx6q_opp_check_speed_grading(cpu_dev); + if (of_machine_is_compatible("fsl,imx6ul")) + imx6ul_opp_check_speed_grading(cpu_dev); + else + imx6q_opp_check_speed_grading(cpu_dev); /* Because we have added the OPPs here, we must free them */ free_opp = true; -- cgit 1.4.1 From dbd49b85eec7eb6d7ae61bad8306d5cdd85c142d Mon Sep 17 00:00:00 2001 From: Srinivas Pandruvada Date: Wed, 10 Jan 2018 11:38:51 -0800 Subject: cpufreq: intel_pstate: Replace bxt_funcs with core_funcs Since core_funcs and bxt_funcs have same set of callbacks, replace bxt_funcs with core_funcs. Signed-off-by: Srinivas Pandruvada Signed-off-by: Rafael J. Wysocki --- drivers/cpufreq/intel_pstate.c | 13 ++----------- 1 file changed, 2 insertions(+), 11 deletions(-) diff --git a/drivers/cpufreq/intel_pstate.c b/drivers/cpufreq/intel_pstate.c index 93a0e88bef76..3b6616b15c59 100644 --- a/drivers/cpufreq/intel_pstate.c +++ b/drivers/cpufreq/intel_pstate.c @@ -1595,15 +1595,6 @@ static const struct pstate_funcs knl_funcs = { .get_val = core_get_val, }; -static const struct pstate_funcs bxt_funcs = { - .get_max = core_get_max_pstate, - .get_max_physical = core_get_max_pstate_physical, - .get_min = core_get_min_pstate, - .get_turbo = core_get_turbo_pstate, - .get_scaling = core_get_scaling, - .get_val = core_get_val, -}; - #define ICPU(model, policy) \ { X86_VENDOR_INTEL, 6, model, X86_FEATURE_APERFMPERF,\ (unsigned long)&policy } @@ -1627,8 +1618,8 @@ static const struct x86_cpu_id intel_pstate_cpu_ids[] = { ICPU(INTEL_FAM6_BROADWELL_XEON_D, core_funcs), ICPU(INTEL_FAM6_XEON_PHI_KNL, knl_funcs), ICPU(INTEL_FAM6_XEON_PHI_KNM, knl_funcs), - ICPU(INTEL_FAM6_ATOM_GOLDMONT, bxt_funcs), - ICPU(INTEL_FAM6_ATOM_GEMINI_LAKE, bxt_funcs), + ICPU(INTEL_FAM6_ATOM_GOLDMONT, core_funcs), + ICPU(INTEL_FAM6_ATOM_GEMINI_LAKE, core_funcs), {} }; MODULE_DEVICE_TABLE(x86cpu, intel_pstate_cpu_ids); -- cgit 1.4.1 From d8de7a44e11f98f2c2a4c2e12e79ba9ffb839306 Mon Sep 17 00:00:00 2001 From: Srinivas Pandruvada Date: Wed, 10 Jan 2018 11:38:52 -0800 Subject: cpufreq: intel_pstate: Add Skylake servers support Currently intel_pstate can function only in HWP mode on Skylake servers. When HWP feature is not enabled on the processor then acpi-cpufreq is driver is used. Based on the power and performance tests using intel_pstate scaling algorithm the results are comparable. But intel_pstate brings in additional features: - Display of turbo frequency range, which many users like to see - Place limits in the turbo frequency range when platform allows Since these tests are done only using non PID algorithm introduced in kernel version 4.14, this patch is not a backport candidate. So each user has to carefully weigh the benefits before he backports. Signed-off-by: Srinivas Pandruvada Signed-off-by: Rafael J. Wysocki --- drivers/cpufreq/intel_pstate.c | 1 + 1 file changed, 1 insertion(+) diff --git a/drivers/cpufreq/intel_pstate.c b/drivers/cpufreq/intel_pstate.c index 3b6616b15c59..7edf7a0e5a96 100644 --- a/drivers/cpufreq/intel_pstate.c +++ b/drivers/cpufreq/intel_pstate.c @@ -1620,6 +1620,7 @@ static const struct x86_cpu_id intel_pstate_cpu_ids[] = { ICPU(INTEL_FAM6_XEON_PHI_KNM, knl_funcs), ICPU(INTEL_FAM6_ATOM_GOLDMONT, core_funcs), ICPU(INTEL_FAM6_ATOM_GEMINI_LAKE, core_funcs), + ICPU(INTEL_FAM6_SKYLAKE_X, core_funcs), {} }; MODULE_DEVICE_TABLE(x86cpu, intel_pstate_cpu_ids); -- cgit 1.4.1 From 3fa4680b860bf48b437d6a2c039789c4abe202ae Mon Sep 17 00:00:00 2001 From: Shilpasri G Bhat Date: Fri, 12 Jan 2018 12:43:53 +0530 Subject: cpufreq: powernv: Dont assume distinct pstate values for nominal and pmin Some OpenPOWER boxes can have same pstate values for nominal and pmin pstates. In these boxes the current code will not initialize 'powernv_pstate_info.min' variable and result in erroneous CPU frequency reporting. This patch fixes this problem. Fixes: 09ca4c9b5958 (cpufreq: powernv: Replacing pstate_id with frequency table index) Reported-by: Alvin Wang Signed-off-by: Shilpasri G Bhat Acked-by: Viresh Kumar Cc: 4.8+ # 4.8+ Signed-off-by: Rafael J. Wysocki --- drivers/cpufreq/powernv-cpufreq.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/drivers/cpufreq/powernv-cpufreq.c b/drivers/cpufreq/powernv-cpufreq.c index 8a4e2ce0804c..29cdec198657 100644 --- a/drivers/cpufreq/powernv-cpufreq.c +++ b/drivers/cpufreq/powernv-cpufreq.c @@ -337,9 +337,9 @@ next: if (id == pstate_max) powernv_pstate_info.max = i; - else if (id == pstate_nominal) + if (id == pstate_nominal) powernv_pstate_info.nominal = i; - else if (id == pstate_min) + if (id == pstate_min) powernv_pstate_info.min = i; if (powernv_pstate_info.wof_enabled && id == pstate_turbo) { -- cgit 1.4.1 From dff4113d5e3753c23e8b5bb6818f5829ccd0d06a Mon Sep 17 00:00:00 2001 From: Sudeep Holla Date: Wed, 10 Jan 2018 16:44:14 +0000 Subject: drivers: psci: remove cluster terminology and dependency on physical_package_id Since the definition of the term "cluster" is not well defined in the architecture, we should avoid using it. Also the physical package id is currently mapped to so called "clusters" in ARM/ARM64 platforms which is already argumentative. Currently PSCI checker uses the physical package id assuming that CPU power domains map to "clusters" and the physical package id in the code as it stands also maps to cluster boundaries. It does that trying to test "cluster" idle states to its best. However the CPU power domain often but not always maps directly to the processor topology. This patch removes the dependency on physical_package_id from the topology in this PSCI checker. Also it replaces all the occurences of clusters to cpu_groups which is derived from core_sibling_mask and may not directly map to physical "cluster". Acked-by: Lorenzo Pieralisi Signed-off-by: Sudeep Holla Signed-off-by: Rafael J. Wysocki --- drivers/firmware/psci_checker.c | 46 ++++++++++++++++++++--------------------- 1 file changed, 22 insertions(+), 24 deletions(-) diff --git a/drivers/firmware/psci_checker.c b/drivers/firmware/psci_checker.c index f3f4f810e5df..bb1c068bff19 100644 --- a/drivers/firmware/psci_checker.c +++ b/drivers/firmware/psci_checker.c @@ -77,8 +77,8 @@ static int psci_ops_check(void) return 0; } -static int find_clusters(const struct cpumask *cpus, - const struct cpumask **clusters) +static int find_cpu_groups(const struct cpumask *cpus, + const struct cpumask **cpu_groups) { unsigned int nb = 0; cpumask_var_t tmp; @@ -88,11 +88,11 @@ static int find_clusters(const struct cpumask *cpus, cpumask_copy(tmp, cpus); while (!cpumask_empty(tmp)) { - const struct cpumask *cluster = + const struct cpumask *cpu_group = topology_core_cpumask(cpumask_any(tmp)); - clusters[nb++] = cluster; - cpumask_andnot(tmp, tmp, cluster); + cpu_groups[nb++] = cpu_group; + cpumask_andnot(tmp, tmp, cpu_group); } free_cpumask_var(tmp); @@ -170,24 +170,24 @@ static int hotplug_tests(void) { int err; cpumask_var_t offlined_cpus; - int i, nb_cluster; - const struct cpumask **clusters; + int i, nb_cpu_group; + const struct cpumask **cpu_groups; char *page_buf; err = -ENOMEM; if (!alloc_cpumask_var(&offlined_cpus, GFP_KERNEL)) return err; - /* We may have up to nb_available_cpus clusters. */ - clusters = kmalloc_array(nb_available_cpus, sizeof(*clusters), - GFP_KERNEL); - if (!clusters) + /* We may have up to nb_available_cpus cpu_groups. */ + cpu_groups = kmalloc_array(nb_available_cpus, sizeof(*cpu_groups), + GFP_KERNEL); + if (!cpu_groups) goto out_free_cpus; page_buf = (char *)__get_free_page(GFP_KERNEL); if (!page_buf) - goto out_free_clusters; + goto out_free_cpu_groups; err = 0; - nb_cluster = find_clusters(cpu_online_mask, clusters); + nb_cpu_group = find_cpu_groups(cpu_online_mask, cpu_groups); /* * Of course the last CPU cannot be powered down and cpu_down() should @@ -197,24 +197,22 @@ static int hotplug_tests(void) err += down_and_up_cpus(cpu_online_mask, offlined_cpus); /* - * Take down CPUs by cluster this time. When the last CPU is turned - * off, the cluster itself should shut down. + * Take down CPUs by cpu group this time. When the last CPU is turned + * off, the cpu group itself should shut down. */ - for (i = 0; i < nb_cluster; ++i) { - int cluster_id = - topology_physical_package_id(cpumask_any(clusters[i])); + for (i = 0; i < nb_cpu_group; ++i) { ssize_t len = cpumap_print_to_pagebuf(true, page_buf, - clusters[i]); + cpu_groups[i]); /* Remove trailing newline. */ page_buf[len - 1] = '\0'; - pr_info("Trying to turn off and on again cluster %d " - "(CPUs %s)\n", cluster_id, page_buf); - err += down_and_up_cpus(clusters[i], offlined_cpus); + pr_info("Trying to turn off and on again group %d (CPUs %s)\n", + i, page_buf); + err += down_and_up_cpus(cpu_groups[i], offlined_cpus); } free_page((unsigned long)page_buf); -out_free_clusters: - kfree(clusters); +out_free_cpu_groups: + kfree(cpu_groups); out_free_cpus: free_cpumask_var(offlined_cpus); return err; -- cgit 1.4.1 From 343a8d17fa8d6dd97f408e8fedbcef12073f3774 Mon Sep 17 00:00:00 2001 From: Sudeep Holla Date: Wed, 10 Jan 2018 16:44:15 +0000 Subject: cpufreq: scpi: remove arm_big_little dependency The dependency on physical_package_id from the topology to get the cluster identifier is wrong. The concept of cluster used in ARM topology is unfortunately not well defined in the architecture, we should avoid using it. Further the frequency domain need not be mapped to so called "clusters" one to one. SCPI already provides means to obtain the frequency domain id from the device tree. In order to support some new topologies(e.g. DSU which contains 2 frequency domains within the physical cluster), pseudo clusters are created to make this driver work which is wrong again. In order to solve those issues and also remove dependency of topological physical id for frequency domain, this patch removes the arm_big_little dependency from scpi driver. Acked-by: Viresh Kumar Signed-off-by: Sudeep Holla Signed-off-by: Rafael J. Wysocki --- drivers/cpufreq/scpi-cpufreq.c | 193 +++++++++++++++++++++++++++++++++++++---- 1 file changed, 178 insertions(+), 15 deletions(-) diff --git a/drivers/cpufreq/scpi-cpufreq.c b/drivers/cpufreq/scpi-cpufreq.c index 05d299052c5c..247fcbfa4cb5 100644 --- a/drivers/cpufreq/scpi-cpufreq.c +++ b/drivers/cpufreq/scpi-cpufreq.c @@ -18,27 +18,89 @@ #define pr_fmt(fmt) KBUILD_MODNAME ": " fmt +#include #include #include +#include +#include +#include #include -#include +#include #include #include +#include #include -#include "arm_big_little.h" +struct scpi_data { + struct clk *clk; + struct device *cpu_dev; + struct thermal_cooling_device *cdev; +}; static struct scpi_ops *scpi_ops; -static int scpi_get_transition_latency(struct device *cpu_dev) +static unsigned int scpi_cpufreq_get_rate(unsigned int cpu) +{ + struct cpufreq_policy *policy = cpufreq_cpu_get_raw(cpu); + struct scpi_data *priv = policy->driver_data; + unsigned long rate = clk_get_rate(priv->clk); + + return rate / 1000; +} + +static int +scpi_cpufreq_set_target(struct cpufreq_policy *policy, unsigned int index) +{ + struct scpi_data *priv = policy->driver_data; + u64 rate = policy->freq_table[index].frequency * 1000; + int ret; + + ret = clk_set_rate(priv->clk, rate); + if (!ret && (clk_get_rate(priv->clk) != rate)) + ret = -EIO; + + return ret; +} + +static int +scpi_get_sharing_cpus(struct device *cpu_dev, struct cpumask *cpumask) { - return scpi_ops->get_transition_latency(cpu_dev); + int cpu, domain, tdomain; + struct device *tcpu_dev; + + domain = scpi_ops->device_domain_id(cpu_dev); + if (domain < 0) + return domain; + + for_each_possible_cpu(cpu) { + if (cpu == cpu_dev->id) + continue; + + tcpu_dev = get_cpu_device(cpu); + if (!tcpu_dev) + continue; + + tdomain = scpi_ops->device_domain_id(tcpu_dev); + if (tdomain == domain) + cpumask_set_cpu(cpu, cpumask); + } + + return 0; } -static int scpi_init_opp_table(const struct cpumask *cpumask) +static int scpi_cpufreq_init(struct cpufreq_policy *policy) { int ret; - struct device *cpu_dev = get_cpu_device(cpumask_first(cpumask)); + unsigned int latency; + struct device *cpu_dev; + struct scpi_data *priv; + struct cpufreq_frequency_table *freq_table; + + cpu_dev = get_cpu_device(policy->cpu); + if (!cpu_dev) { + pr_err("failed to get cpu%d device\n", policy->cpu); + return -ENODEV; + } ret = scpi_ops->add_opps_to_device(cpu_dev); if (ret) { @@ -46,32 +108,133 @@ static int scpi_init_opp_table(const struct cpumask *cpumask) return ret; } - ret = dev_pm_opp_set_sharing_cpus(cpu_dev, cpumask); - if (ret) + ret = scpi_get_sharing_cpus(cpu_dev, policy->cpus); + if (ret) { + dev_warn(cpu_dev, "failed to get sharing cpumask\n"); + return ret; + } + + ret = dev_pm_opp_set_sharing_cpus(cpu_dev, policy->cpus); + if (ret) { dev_err(cpu_dev, "%s: failed to mark OPPs as shared: %d\n", __func__, ret); + return ret; + } + + ret = dev_pm_opp_get_opp_count(cpu_dev); + if (ret <= 0) { + dev_dbg(cpu_dev, "OPP table is not ready, deferring probe\n"); + ret = -EPROBE_DEFER; + goto out_free_opp; + } + + priv = kzalloc(sizeof(*priv), GFP_KERNEL); + if (!priv) { + ret = -ENOMEM; + goto out_free_opp; + } + + ret = dev_pm_opp_init_cpufreq_table(cpu_dev, &freq_table); + if (ret) { + dev_err(cpu_dev, "failed to init cpufreq table: %d\n", ret); + goto out_free_priv; + } + + priv->cpu_dev = cpu_dev; + priv->clk = clk_get(cpu_dev, NULL); + if (IS_ERR(priv->clk)) { + dev_err(cpu_dev, "%s: Failed to get clk for cpu: %d\n", + __func__, cpu_dev->id); + goto out_free_cpufreq_table; + } + + policy->driver_data = priv; + + ret = cpufreq_table_validate_and_show(policy, freq_table); + if (ret) { + dev_err(cpu_dev, "%s: invalid frequency table: %d\n", __func__, + ret); + goto out_put_clk; + } + + /* scpi allows DVFS request for any domain from any CPU */ + policy->dvfs_possible_from_any_cpu = true; + + latency = scpi_ops->get_transition_latency(cpu_dev); + if (!latency) + latency = CPUFREQ_ETERNAL; + + policy->cpuinfo.transition_latency = latency; + + policy->fast_switch_possible = false; + return 0; + +out_put_clk: + clk_put(priv->clk); +out_free_cpufreq_table: + dev_pm_opp_free_cpufreq_table(cpu_dev, &freq_table); +out_free_priv: + kfree(priv); +out_free_opp: + dev_pm_opp_cpumask_remove_table(policy->cpus); + return ret; } -static const struct cpufreq_arm_bL_ops scpi_cpufreq_ops = { - .name = "scpi", - .get_transition_latency = scpi_get_transition_latency, - .init_opp_table = scpi_init_opp_table, - .free_opp_table = dev_pm_opp_cpumask_remove_table, +static int scpi_cpufreq_exit(struct cpufreq_policy *policy) +{ + struct scpi_data *priv = policy->driver_data; + + cpufreq_cooling_unregister(priv->cdev); + clk_put(priv->clk); + dev_pm_opp_free_cpufreq_table(priv->cpu_dev, &policy->freq_table); + kfree(priv); + dev_pm_opp_cpumask_remove_table(policy->related_cpus); + + return 0; +} + +static void scpi_cpufreq_ready(struct cpufreq_policy *policy) +{ + struct scpi_data *priv = policy->driver_data; + struct thermal_cooling_device *cdev; + + cdev = of_cpufreq_cooling_register(policy); + if (!IS_ERR(cdev)) + priv->cdev = cdev; +} + +static struct cpufreq_driver scpi_cpufreq_driver = { + .name = "scpi-cpufreq", + .flags = CPUFREQ_STICKY | CPUFREQ_HAVE_GOVERNOR_PER_POLICY | + CPUFREQ_NEED_INITIAL_FREQ_CHECK, + .verify = cpufreq_generic_frequency_table_verify, + .attr = cpufreq_generic_attr, + .get = scpi_cpufreq_get_rate, + .init = scpi_cpufreq_init, + .exit = scpi_cpufreq_exit, + .ready = scpi_cpufreq_ready, + .target_index = scpi_cpufreq_set_target, }; static int scpi_cpufreq_probe(struct platform_device *pdev) { + int ret; + scpi_ops = get_scpi_ops(); if (!scpi_ops) return -EIO; - return bL_cpufreq_register(&scpi_cpufreq_ops); + ret = cpufreq_register_driver(&scpi_cpufreq_driver); + if (ret) + dev_err(&pdev->dev, "%s: registering cpufreq failed, err: %d\n", + __func__, ret); + return ret; } static int scpi_cpufreq_remove(struct platform_device *pdev) { - bL_cpufreq_unregister(&scpi_cpufreq_ops); + cpufreq_unregister_driver(&scpi_cpufreq_driver); scpi_ops = NULL; return 0; } -- cgit 1.4.1