From 21469df4676e89cb9fd4be489d9a2cc4af09db8c Mon Sep 17 00:00:00 2001 From: Viresh Kumar Date: Mon, 7 Jan 2019 12:28:54 +0530 Subject: [PATCH 01/46] cpufreq: Don't update new_policy on failures The local variable "new_policy" hasn't been used in the error path of cpufreq_online() since commit f9f41e3ef99a (cpufreq: Remove policy create/remove notifiers). Don't update it in that error path. Signed-off-by: Viresh Kumar [ rjw: Changelog ] Signed-off-by: Rafael J. Wysocki --- drivers/cpufreq/cpufreq.c | 2 -- 1 file changed, 2 deletions(-) diff --git a/drivers/cpufreq/cpufreq.c b/drivers/cpufreq/cpufreq.c index e35a886e00bc..a8fa684f5f90 100644 --- a/drivers/cpufreq/cpufreq.c +++ b/drivers/cpufreq/cpufreq.c @@ -1305,8 +1305,6 @@ static int cpufreq_online(unsigned int cpu) if (ret) { pr_err("%s: Failed to initialize policy for cpu: %d (%d)\n", __func__, cpu, ret); - /* cpufreq_policy_free() will notify based on this */ - new_policy = false; goto out_destroy_policy; } From 4944514e6c7e613b578aaaafb2a2dd1a54bcf538 Mon Sep 17 00:00:00 2001 From: "Gustavo A. R. Silva" Date: Mon, 7 Jan 2019 11:33:43 -0600 Subject: [PATCH 02/46] cpufreq: e_powersaver: Use struct_size() in kzalloc() One of the more common cases of allocation size calculations is finding the size of a structure that has a zero-sized array at the end, along with memory for some number of elements for that array. For example: struct foo { int stuff; void *entry[]; }; instance = kzalloc(sizeof(struct foo) + sizeof(void *) * count, GFP_KERNEL); Instead of leaving these open-coded and prone to type mistakes, we can now use the new struct_size() helper: instance = kzalloc(struct_size(instance, entry, count), GFP_KERNEL); This code was detected with the help of Coccinelle. Signed-off-by: Gustavo A. R. Silva Acked-by: Viresh Kumar Signed-off-by: Rafael J. Wysocki --- drivers/cpufreq/e_powersaver.c | 5 ++--- 1 file changed, 2 insertions(+), 3 deletions(-) diff --git a/drivers/cpufreq/e_powersaver.c b/drivers/cpufreq/e_powersaver.c index 60bea302abbe..2d3ef208dd70 100644 --- a/drivers/cpufreq/e_powersaver.c +++ b/drivers/cpufreq/e_powersaver.c @@ -323,9 +323,8 @@ static int eps_cpu_init(struct cpufreq_policy *policy) states = 2; /* Allocate private data and frequency table for current cpu */ - centaur = kzalloc(sizeof(*centaur) - + (states + 1) * sizeof(struct cpufreq_frequency_table), - GFP_KERNEL); + centaur = kzalloc(struct_size(centaur, freq_table, states + 1), + GFP_KERNEL); if (!centaur) return -ENOMEM; eps_cpu[0] = centaur; From 8321be6a9df5c5cfbf3fb5f716caf8698a5a7016 Mon Sep 17 00:00:00 2001 From: Amit Kucheria Date: Mon, 21 Jan 2019 14:17:37 +0530 Subject: [PATCH 03/46] cpufreq: Replace open-coded << with BIT() Minor clean-up to use BIT() and keep checkpatch happy. Clean up the comment formatting while we're at it to make it easier to read. Signed-off-by: Amit Kucheria Reviewed-by: Stephen Boyd Signed-off-by: Rafael J. Wysocki --- include/linux/cpufreq.h | 25 +++++++++++++------------ 1 file changed, 13 insertions(+), 12 deletions(-) diff --git a/include/linux/cpufreq.h b/include/linux/cpufreq.h index c86d6d8bdfed..bd7fbd6a4478 100644 --- a/include/linux/cpufreq.h +++ b/include/linux/cpufreq.h @@ -346,14 +346,15 @@ struct cpufreq_driver { }; /* flags */ -#define CPUFREQ_STICKY (1 << 0) /* driver isn't removed even if - all ->init() calls failed */ -#define CPUFREQ_CONST_LOOPS (1 << 1) /* loops_per_jiffy or other - kernel "constants" aren't - affected by frequency - transitions */ -#define CPUFREQ_PM_NO_WARN (1 << 2) /* don't warn on suspend/resume - speed mismatches */ + +/* driver isn't removed even if all ->init() calls failed */ +#define CPUFREQ_STICKY BIT(0) + +/* loops_per_jiffy or other kernel "constants" aren't affected by frequency transitions */ +#define CPUFREQ_CONST_LOOPS BIT(1) + +/* don't warn on suspend/resume speed mismatches */ +#define CPUFREQ_PM_NO_WARN BIT(2) /* * This should be set by platforms having multiple clock-domains, i.e. @@ -361,14 +362,14 @@ struct cpufreq_driver { * be created in cpu/cpu/cpufreq/ directory and so they can use the same * governor with different tunables for different clusters. */ -#define CPUFREQ_HAVE_GOVERNOR_PER_POLICY (1 << 3) +#define CPUFREQ_HAVE_GOVERNOR_PER_POLICY BIT(3) /* * Driver will do POSTCHANGE notifications from outside of their ->target() * routine and so must set cpufreq_driver->flags with this flag, so that core * can handle them specially. */ -#define CPUFREQ_ASYNC_NOTIFICATION (1 << 4) +#define CPUFREQ_ASYNC_NOTIFICATION BIT(4) /* * Set by drivers which want cpufreq core to check if CPU is running at a @@ -377,13 +378,13 @@ struct cpufreq_driver { * from the table. And if that fails, we will stop further boot process by * issuing a BUG_ON(). */ -#define CPUFREQ_NEED_INITIAL_FREQ_CHECK (1 << 5) +#define CPUFREQ_NEED_INITIAL_FREQ_CHECK BIT(5) /* * Set by drivers to disallow use of governors with "dynamic_switching" flag * set. */ -#define CPUFREQ_NO_AUTO_DYNAMIC_SWITCHING (1 << 6) +#define CPUFREQ_NO_AUTO_DYNAMIC_SWITCHING BIT(6) int cpufreq_register_driver(struct cpufreq_driver *driver_data); int cpufreq_unregister_driver(struct cpufreq_driver *driver_data); From 625c85a62cb7d3c79f6e16de3cfa972033658250 Mon Sep 17 00:00:00 2001 From: Viresh Kumar Date: Fri, 25 Jan 2019 12:53:07 +0530 Subject: [PATCH 04/46] cpufreq: Use struct kobj_attribute instead of struct global_attr The cpufreq_global_kobject is created using kobject_create_and_add() helper, which assigns the kobj_type as dynamic_kobj_ktype and show/store routines are set to kobj_attr_show() and kobj_attr_store(). These routines pass struct kobj_attribute as an argument to the show/store callbacks. But all the cpufreq files created using the cpufreq_global_kobject expect the argument to be of type struct attribute. Things work fine currently as no one accesses the "attr" argument. We may not see issues even if the argument is used, as struct kobj_attribute has struct attribute as its first element and so they will both get same address. But this is logically incorrect and we should rather use struct kobj_attribute instead of struct global_attr in the cpufreq core and drivers and the show/store callbacks should take struct kobj_attribute as argument instead. This bug is caught using CFI CLANG builds in android kernel which catches mismatch in function prototypes for such callbacks. Reported-by: Donghee Han Reported-by: Sangkyu Kim Signed-off-by: Viresh Kumar Signed-off-by: Rafael J. Wysocki --- drivers/cpufreq/cpufreq.c | 6 +++--- drivers/cpufreq/intel_pstate.c | 23 ++++++++++++----------- include/linux/cpufreq.h | 12 ++---------- 3 files changed, 17 insertions(+), 24 deletions(-) diff --git a/drivers/cpufreq/cpufreq.c b/drivers/cpufreq/cpufreq.c index a8fa684f5f90..3eff158d9750 100644 --- a/drivers/cpufreq/cpufreq.c +++ b/drivers/cpufreq/cpufreq.c @@ -545,13 +545,13 @@ EXPORT_SYMBOL_GPL(cpufreq_policy_transition_delay_us); * SYSFS INTERFACE * *********************************************************************/ static ssize_t show_boost(struct kobject *kobj, - struct attribute *attr, char *buf) + struct kobj_attribute *attr, char *buf) { return sprintf(buf, "%d\n", cpufreq_driver->boost_enabled); } -static ssize_t store_boost(struct kobject *kobj, struct attribute *attr, - const char *buf, size_t count) +static ssize_t store_boost(struct kobject *kobj, struct kobj_attribute *attr, + const char *buf, size_t count) { int ret, enable; diff --git a/drivers/cpufreq/intel_pstate.c b/drivers/cpufreq/intel_pstate.c index dd66decf2087..5ab6a4fe93aa 100644 --- a/drivers/cpufreq/intel_pstate.c +++ b/drivers/cpufreq/intel_pstate.c @@ -895,7 +895,7 @@ static void intel_pstate_update_policies(void) /************************** sysfs begin ************************/ #define show_one(file_name, object) \ static ssize_t show_##file_name \ - (struct kobject *kobj, struct attribute *attr, char *buf) \ + (struct kobject *kobj, struct kobj_attribute *attr, char *buf) \ { \ return sprintf(buf, "%u\n", global.object); \ } @@ -904,7 +904,7 @@ static ssize_t intel_pstate_show_status(char *buf); static int intel_pstate_update_status(const char *buf, size_t size); static ssize_t show_status(struct kobject *kobj, - struct attribute *attr, char *buf) + struct kobj_attribute *attr, char *buf) { ssize_t ret; @@ -915,7 +915,7 @@ static ssize_t show_status(struct kobject *kobj, return ret; } -static ssize_t store_status(struct kobject *a, struct attribute *b, +static ssize_t store_status(struct kobject *a, struct kobj_attribute *b, const char *buf, size_t count) { char *p = memchr(buf, '\n', count); @@ -929,7 +929,7 @@ static ssize_t store_status(struct kobject *a, struct attribute *b, } static ssize_t show_turbo_pct(struct kobject *kobj, - struct attribute *attr, char *buf) + struct kobj_attribute *attr, char *buf) { struct cpudata *cpu; int total, no_turbo, turbo_pct; @@ -955,7 +955,7 @@ static ssize_t show_turbo_pct(struct kobject *kobj, } static ssize_t show_num_pstates(struct kobject *kobj, - struct attribute *attr, char *buf) + struct kobj_attribute *attr, char *buf) { struct cpudata *cpu; int total; @@ -976,7 +976,7 @@ static ssize_t show_num_pstates(struct kobject *kobj, } static ssize_t show_no_turbo(struct kobject *kobj, - struct attribute *attr, char *buf) + struct kobj_attribute *attr, char *buf) { ssize_t ret; @@ -998,7 +998,7 @@ static ssize_t show_no_turbo(struct kobject *kobj, return ret; } -static ssize_t store_no_turbo(struct kobject *a, struct attribute *b, +static ssize_t store_no_turbo(struct kobject *a, struct kobj_attribute *b, const char *buf, size_t count) { unsigned int input; @@ -1045,7 +1045,7 @@ static ssize_t store_no_turbo(struct kobject *a, struct attribute *b, return count; } -static ssize_t store_max_perf_pct(struct kobject *a, struct attribute *b, +static ssize_t store_max_perf_pct(struct kobject *a, struct kobj_attribute *b, const char *buf, size_t count) { unsigned int input; @@ -1075,7 +1075,7 @@ static ssize_t store_max_perf_pct(struct kobject *a, struct attribute *b, return count; } -static ssize_t store_min_perf_pct(struct kobject *a, struct attribute *b, +static ssize_t store_min_perf_pct(struct kobject *a, struct kobj_attribute *b, const char *buf, size_t count) { unsigned int input; @@ -1107,12 +1107,13 @@ static ssize_t store_min_perf_pct(struct kobject *a, struct attribute *b, } static ssize_t show_hwp_dynamic_boost(struct kobject *kobj, - struct attribute *attr, char *buf) + struct kobj_attribute *attr, char *buf) { return sprintf(buf, "%u\n", hwp_boost); } -static ssize_t store_hwp_dynamic_boost(struct kobject *a, struct attribute *b, +static ssize_t store_hwp_dynamic_boost(struct kobject *a, + struct kobj_attribute *b, const char *buf, size_t count) { unsigned int input; diff --git a/include/linux/cpufreq.h b/include/linux/cpufreq.h index bd7fbd6a4478..c19142911554 100644 --- a/include/linux/cpufreq.h +++ b/include/linux/cpufreq.h @@ -254,20 +254,12 @@ __ATTR(_name, 0644, show_##_name, store_##_name) static struct freq_attr _name = \ __ATTR(_name, 0200, NULL, store_##_name) -struct global_attr { - struct attribute attr; - ssize_t (*show)(struct kobject *kobj, - struct attribute *attr, char *buf); - ssize_t (*store)(struct kobject *a, struct attribute *b, - const char *c, size_t count); -}; - #define define_one_global_ro(_name) \ -static struct global_attr _name = \ +static struct kobj_attribute _name = \ __ATTR(_name, 0444, show_##_name, NULL) #define define_one_global_rw(_name) \ -static struct global_attr _name = \ +static struct kobj_attribute _name = \ __ATTR(_name, 0644, show_##_name, store_##_name) From afa1f2ab43d48d0e1fa1bda524a0cf53e4cd6c87 Mon Sep 17 00:00:00 2001 From: Amit Kucheria Date: Tue, 29 Jan 2019 10:25:07 +0530 Subject: [PATCH 05/46] thermal: cpu_cooling: Require thermal core to be compiled in The CPU cooling driver (cpu_cooling.c) allows the platform's cpufreq driver to register as a cooling device and cool down the platform by throttling the CPU frequency. In order to be able to auto-register a cpufreq driver as a cooling device from the cpufreq core, we need access to code inside cpu_cooling.c which, in turn, accesses code inside thermal core. CPU_FREQ is a bool while THERMAL is tristate. In some configurations (e.g. allmodconfig), CONFIG_THERMAL ends up as a module while CONFIG_CPU_FREQ is compiled in. This leads to following error: drivers/cpufreq/cpufreq.o: In function `cpufreq_offline': cpufreq.c:(.text+0x407c): undefined reference to `cpufreq_cooling_unregister' drivers/cpufreq/cpufreq.o: In function `cpufreq_online': cpufreq.c:(.text+0x70c0): undefined reference to `of_cpufreq_cooling_register' Given that platforms using CPU_THERMAL usually want it compiled-in so it is available early in boot, make CPU_THERMAL depend on THERMAL being compiled-in instead of allowing it to be a module. As a result of this change, get rid of the ugly (!CPU_THERMAL || THERMAL) dependency in all cpufreq drivers using CPU_THERMAL. Suggested-by: Rafael J. Wysocki Signed-off-by: Amit Kucheria Acked-by: Viresh Kumar Signed-off-by: Rafael J. Wysocki --- drivers/cpufreq/Kconfig | 3 --- drivers/cpufreq/Kconfig.arm | 5 ----- drivers/thermal/Kconfig | 1 + 3 files changed, 1 insertion(+), 8 deletions(-) diff --git a/drivers/cpufreq/Kconfig b/drivers/cpufreq/Kconfig index 608af20a3494..b22e6bba71f1 100644 --- a/drivers/cpufreq/Kconfig +++ b/drivers/cpufreq/Kconfig @@ -207,8 +207,6 @@ comment "CPU frequency scaling drivers" config CPUFREQ_DT tristate "Generic DT based cpufreq driver" depends on HAVE_CLK && OF - # if CPU_THERMAL is on and THERMAL=m, CPUFREQ_DT cannot be =y: - depends on !CPU_THERMAL || THERMAL select CPUFREQ_DT_PLATDEV select PM_OPP help @@ -327,7 +325,6 @@ endif config QORIQ_CPUFREQ tristate "CPU frequency scaling driver for Freescale QorIQ SoCs" depends on OF && COMMON_CLK && (PPC_E500MC || ARM || ARM64) - depends on !CPU_THERMAL || THERMAL select CLK_QORIQ help This adds the CPUFreq driver support for Freescale QorIQ SoCs diff --git a/drivers/cpufreq/Kconfig.arm b/drivers/cpufreq/Kconfig.arm index 688f10227793..ca8567c3152c 100644 --- a/drivers/cpufreq/Kconfig.arm +++ b/drivers/cpufreq/Kconfig.arm @@ -29,8 +29,6 @@ config ARM_ARMADA_37XX_CPUFREQ config ARM_BIG_LITTLE_CPUFREQ tristate "Generic ARM big LITTLE CPUfreq driver" depends on ARM_CPU_TOPOLOGY && HAVE_CLK - # if CPU_THERMAL is on and THERMAL=m, ARM_BIT_LITTLE_CPUFREQ cannot be =y - depends on !CPU_THERMAL || THERMAL select PM_OPP help This enables the Generic CPUfreq driver for ARM big.LITTLE platforms. @@ -38,7 +36,6 @@ config ARM_BIG_LITTLE_CPUFREQ config ARM_SCPI_CPUFREQ tristate "SCPI based CPUfreq driver" depends on ARM_SCPI_PROTOCOL && COMMON_CLK_SCPI - depends on !CPU_THERMAL || THERMAL help This adds the CPUfreq driver support for ARM platforms using SCPI protocol for CPU power management. @@ -93,7 +90,6 @@ config ARM_KIRKWOOD_CPUFREQ config ARM_MEDIATEK_CPUFREQ tristate "CPU Frequency scaling support for MediaTek SoCs" depends on ARCH_MEDIATEK && REGULATOR - depends on !CPU_THERMAL || THERMAL select PM_OPP help This adds the CPUFreq driver support for MediaTek SoCs. @@ -233,7 +229,6 @@ config ARM_SA1110_CPUFREQ config ARM_SCMI_CPUFREQ tristate "SCMI based CPUfreq driver" depends on ARM_SCMI_PROTOCOL || COMPILE_TEST - depends on !CPU_THERMAL || THERMAL select PM_OPP help This adds the CPUfreq driver support for ARM platforms using SCMI diff --git a/drivers/thermal/Kconfig b/drivers/thermal/Kconfig index 30323426902e..58bb7d72dc2b 100644 --- a/drivers/thermal/Kconfig +++ b/drivers/thermal/Kconfig @@ -152,6 +152,7 @@ config CPU_THERMAL bool "generic cpu cooling support" depends on CPU_FREQ depends on THERMAL_OF + depends on THERMAL=y help This implements the generic cpu cooling mechanism through frequency reduction. An ACPI version of this already exists From 5c238a8b599f1ae25eaeb08ad0e9e13e2b9eb023 Mon Sep 17 00:00:00 2001 From: Amit Kucheria Date: Wed, 30 Jan 2019 10:52:01 +0530 Subject: [PATCH 06/46] cpufreq: Auto-register the driver as a thermal cooling device if asked All cpufreq drivers do similar things to register as a cooling device. Provide a cpufreq driver flag so drivers can just ask the cpufreq core to register the cooling device on their behalf. This allows us to get rid of duplicated code in the drivers. In order to allow this, we add a struct thermal_cooling_device pointer to struct cpufreq_policy so that drivers don't need to store it in a private data structure. Suggested-by: Stephen Boyd Suggested-by: Viresh Kumar Signed-off-by: Amit Kucheria Reviewed-by: Matthias Kaehlcke Tested-by: Matthias Kaehlcke Acked-by: Viresh Kumar Reviewed-by: Daniel Lezcano Signed-off-by: Rafael J. Wysocki --- drivers/cpufreq/cpufreq.c | 11 +++++++++++ include/linux/cpufreq.h | 9 +++++++++ 2 files changed, 20 insertions(+) diff --git a/drivers/cpufreq/cpufreq.c b/drivers/cpufreq/cpufreq.c index 3eff158d9750..96a69c67a545 100644 --- a/drivers/cpufreq/cpufreq.c +++ b/drivers/cpufreq/cpufreq.c @@ -19,6 +19,7 @@ #include #include +#include #include #include #include @@ -1316,6 +1317,10 @@ static int cpufreq_online(unsigned int cpu) if (cpufreq_driver->ready) cpufreq_driver->ready(policy); + if (IS_ENABLED(CONFIG_CPU_THERMAL) && + cpufreq_driver->flags & CPUFREQ_IS_COOLING_DEV) + policy->cdev = of_cpufreq_cooling_register(policy); + pr_debug("initialization complete\n"); return 0; @@ -1403,6 +1408,12 @@ static int cpufreq_offline(unsigned int cpu) goto unlock; } + if (IS_ENABLED(CONFIG_CPU_THERMAL) && + cpufreq_driver->flags & CPUFREQ_IS_COOLING_DEV) { + cpufreq_cooling_unregister(policy->cdev); + policy->cdev = NULL; + } + if (cpufreq_driver->stop_cpu) cpufreq_driver->stop_cpu(policy); diff --git a/include/linux/cpufreq.h b/include/linux/cpufreq.h index c19142911554..9db074ecbbd7 100644 --- a/include/linux/cpufreq.h +++ b/include/linux/cpufreq.h @@ -151,6 +151,9 @@ struct cpufreq_policy { /* For cpufreq driver's internal use */ void *driver_data; + + /* Pointer to the cooling device if used for thermal mitigation */ + struct thermal_cooling_device *cdev; }; /* Only for ACPI */ @@ -378,6 +381,12 @@ struct cpufreq_driver { */ #define CPUFREQ_NO_AUTO_DYNAMIC_SWITCHING BIT(6) +/* + * Set by drivers that want the core to automatically register the cpufreq + * driver as a thermal cooling device. + */ +#define CPUFREQ_IS_COOLING_DEV BIT(7) + int cpufreq_register_driver(struct cpufreq_driver *driver_data); int cpufreq_unregister_driver(struct cpufreq_driver *driver_data); From 4c5ff1c8320d209c038f3698d8fa91fe946a3818 Mon Sep 17 00:00:00 2001 From: Amit Kucheria Date: Tue, 29 Jan 2019 10:25:09 +0530 Subject: [PATCH 07/46] cpufreq: qcom-hw: Register as a cpufreq cooling device Add the CPUFREQ_IS_COOLING_DEV flag to allow the cpufreq core to auto-register the driver as a cooling device. Signed-off-by: Amit Kucheria Reviewed-by: Matthias Kaehlcke Tested-by: Matthias Kaehlcke Reviewed-by: Stephen Boyd Acked-by: Viresh Kumar Reviewed-by: Daniel Lezcano Signed-off-by: Rafael J. Wysocki --- drivers/cpufreq/qcom-cpufreq-hw.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/drivers/cpufreq/qcom-cpufreq-hw.c b/drivers/cpufreq/qcom-cpufreq-hw.c index d83939a1b3d4..c88b51304d89 100644 --- a/drivers/cpufreq/qcom-cpufreq-hw.c +++ b/drivers/cpufreq/qcom-cpufreq-hw.c @@ -231,7 +231,8 @@ static struct freq_attr *qcom_cpufreq_hw_attr[] = { static struct cpufreq_driver cpufreq_qcom_hw_driver = { .flags = CPUFREQ_STICKY | CPUFREQ_NEED_INITIAL_FREQ_CHECK | - CPUFREQ_HAVE_GOVERNOR_PER_POLICY, + CPUFREQ_HAVE_GOVERNOR_PER_POLICY | + CPUFREQ_IS_COOLING_DEV, .verify = cpufreq_generic_frequency_table_verify, .target_index = qcom_cpufreq_hw_target_index, .get = qcom_cpufreq_hw_get, From 4b498869268ed631c7a383c50a6a6048b280dcd7 Mon Sep 17 00:00:00 2001 From: Amit Kucheria Date: Tue, 29 Jan 2019 10:25:10 +0530 Subject: [PATCH 08/46] cpufreq: imx6q: Use auto-registration of thermal cooling device Use the CPUFREQ_IS_COOLING_DEV flag to allow cpufreq core to automatically register as a thermal cooling device. This allows removal of boiler plate code from the driver. Signed-off-by: Amit Kucheria Acked-by: Viresh Kumar Reviewed-by: Daniel Lezcano Signed-off-by: Rafael J. Wysocki --- drivers/cpufreq/imx6q-cpufreq.c | 24 ++---------------------- 1 file changed, 2 insertions(+), 22 deletions(-) diff --git a/drivers/cpufreq/imx6q-cpufreq.c b/drivers/cpufreq/imx6q-cpufreq.c index 9fedf627e000..ca955713e070 100644 --- a/drivers/cpufreq/imx6q-cpufreq.c +++ b/drivers/cpufreq/imx6q-cpufreq.c @@ -9,7 +9,6 @@ #include #include #include -#include #include #include #include @@ -52,7 +51,6 @@ static struct clk_bulk_data clks[] = { }; static struct device *cpu_dev; -static struct thermal_cooling_device *cdev; static bool free_opp; static struct cpufreq_frequency_table *freq_table; static unsigned int max_freq; @@ -193,16 +191,6 @@ static int imx6q_set_target(struct cpufreq_policy *policy, unsigned int index) return 0; } -static void imx6q_cpufreq_ready(struct cpufreq_policy *policy) -{ - cdev = of_cpufreq_cooling_register(policy); - - if (!cdev) - dev_err(cpu_dev, - "running cpufreq without cooling device: %ld\n", - PTR_ERR(cdev)); -} - static int imx6q_cpufreq_init(struct cpufreq_policy *policy) { int ret; @@ -214,22 +202,14 @@ static int imx6q_cpufreq_init(struct cpufreq_policy *policy) return ret; } -static int imx6q_cpufreq_exit(struct cpufreq_policy *policy) -{ - cpufreq_cooling_unregister(cdev); - - return 0; -} - static struct cpufreq_driver imx6q_cpufreq_driver = { - .flags = CPUFREQ_NEED_INITIAL_FREQ_CHECK, + .flags = CPUFREQ_NEED_INITIAL_FREQ_CHECK | + CPUFREQ_IS_COOLING_DEV, .verify = cpufreq_generic_frequency_table_verify, .target_index = imx6q_set_target, .get = cpufreq_generic_get, .init = imx6q_cpufreq_init, - .exit = imx6q_cpufreq_exit, .name = "imx6q-cpufreq", - .ready = imx6q_cpufreq_ready, .attr = cpufreq_generic_attr, .suspend = cpufreq_generic_suspend, }; From e248d8d35cff6a8351a10544588deeca345bbbc6 Mon Sep 17 00:00:00 2001 From: Amit Kucheria Date: Tue, 29 Jan 2019 10:25:11 +0530 Subject: [PATCH 09/46] cpufreq: cpufreq-dt: Use auto-registration of thermal cooling device Use the CPUFREQ_IS_COOLING_DEV flag to allow cpufreq core to automatically register as a thermal cooling device. This allows removal of boiler plate code from the driver. Signed-off-by: Amit Kucheria Acked-by: Viresh Kumar Reviewed-by: Daniel Lezcano Signed-off-by: Rafael J. Wysocki --- drivers/cpufreq/cpufreq-dt.c | 14 ++------------ 1 file changed, 2 insertions(+), 12 deletions(-) diff --git a/drivers/cpufreq/cpufreq-dt.c b/drivers/cpufreq/cpufreq-dt.c index e58bfcb1169e..7ba392911cd0 100644 --- a/drivers/cpufreq/cpufreq-dt.c +++ b/drivers/cpufreq/cpufreq-dt.c @@ -13,7 +13,6 @@ #include #include -#include #include #include #include @@ -30,7 +29,6 @@ struct private_data { struct opp_table *opp_table; struct device *cpu_dev; - struct thermal_cooling_device *cdev; const char *reg_name; bool have_static_opps; }; @@ -301,7 +299,6 @@ static int cpufreq_exit(struct cpufreq_policy *policy) { struct private_data *priv = policy->driver_data; - cpufreq_cooling_unregister(priv->cdev); dev_pm_opp_free_cpufreq_table(priv->cpu_dev, &policy->freq_table); if (priv->have_static_opps) dev_pm_opp_of_cpumask_remove_table(policy->related_cpus); @@ -314,21 +311,14 @@ static int cpufreq_exit(struct cpufreq_policy *policy) return 0; } -static void cpufreq_ready(struct cpufreq_policy *policy) -{ - struct private_data *priv = policy->driver_data; - - priv->cdev = of_cpufreq_cooling_register(policy); -} - static struct cpufreq_driver dt_cpufreq_driver = { - .flags = CPUFREQ_STICKY | CPUFREQ_NEED_INITIAL_FREQ_CHECK, + .flags = CPUFREQ_STICKY | CPUFREQ_NEED_INITIAL_FREQ_CHECK | + CPUFREQ_IS_COOLING_DEV, .verify = cpufreq_generic_frequency_table_verify, .target_index = set_target, .get = cpufreq_generic_get, .init = cpufreq_init, .exit = cpufreq_exit, - .ready = cpufreq_ready, .name = "cpufreq-dt", .attr = cpufreq_dt_attr, .suspend = cpufreq_generic_suspend, From 0db60d6b89b921c26d6dac4ec7b35e0102d8f9f8 Mon Sep 17 00:00:00 2001 From: Amit Kucheria Date: Tue, 29 Jan 2019 10:25:12 +0530 Subject: [PATCH 10/46] cpufreq: mediatek: Use auto-registration of thermal cooling device Use the CPUFREQ_IS_COOLING_DEV flag to allow cpufreq core to automatically register as a thermal cooling device. This allows removal of boiler plate code from the driver. Signed-off-by: Amit Kucheria Acked-by: Viresh Kumar Reviewed-by: Daniel Lezcano Signed-off-by: Rafael J. Wysocki --- drivers/cpufreq/mediatek-cpufreq.c | 14 ++------------ 1 file changed, 2 insertions(+), 12 deletions(-) diff --git a/drivers/cpufreq/mediatek-cpufreq.c b/drivers/cpufreq/mediatek-cpufreq.c index eb8920d39818..4229fcc31310 100644 --- a/drivers/cpufreq/mediatek-cpufreq.c +++ b/drivers/cpufreq/mediatek-cpufreq.c @@ -14,7 +14,6 @@ #include #include -#include #include #include #include @@ -48,7 +47,6 @@ struct mtk_cpu_dvfs_info { struct regulator *sram_reg; struct clk *cpu_clk; struct clk *inter_clk; - struct thermal_cooling_device *cdev; struct list_head list_head; int intermediate_voltage; bool need_voltage_tracking; @@ -307,13 +305,6 @@ static int mtk_cpufreq_set_target(struct cpufreq_policy *policy, #define DYNAMIC_POWER "dynamic-power-coefficient" -static void mtk_cpufreq_ready(struct cpufreq_policy *policy) -{ - struct mtk_cpu_dvfs_info *info = policy->driver_data; - - info->cdev = of_cpufreq_cooling_register(policy); -} - static int mtk_cpu_dvfs_info_init(struct mtk_cpu_dvfs_info *info, int cpu) { struct device *cpu_dev; @@ -472,7 +463,6 @@ static int mtk_cpufreq_exit(struct cpufreq_policy *policy) { struct mtk_cpu_dvfs_info *info = policy->driver_data; - cpufreq_cooling_unregister(info->cdev); dev_pm_opp_free_cpufreq_table(info->cpu_dev, &policy->freq_table); return 0; @@ -480,13 +470,13 @@ static int mtk_cpufreq_exit(struct cpufreq_policy *policy) static struct cpufreq_driver mtk_cpufreq_driver = { .flags = CPUFREQ_STICKY | CPUFREQ_NEED_INITIAL_FREQ_CHECK | - CPUFREQ_HAVE_GOVERNOR_PER_POLICY, + CPUFREQ_HAVE_GOVERNOR_PER_POLICY | + CPUFREQ_IS_COOLING_DEV, .verify = cpufreq_generic_frequency_table_verify, .target_index = mtk_cpufreq_set_target, .get = cpufreq_generic_get, .init = mtk_cpufreq_init, .exit = mtk_cpufreq_exit, - .ready = mtk_cpufreq_ready, .name = "mtk-cpufreq", .attr = cpufreq_generic_attr, }; From 17170ec17109bdcd26bf542392a5bafcb6a5732f Mon Sep 17 00:00:00 2001 From: Amit Kucheria Date: Tue, 29 Jan 2019 10:25:13 +0530 Subject: [PATCH 11/46] cpufreq: qoriq: Use auto-registration of thermal cooling device Use the CPUFREQ_IS_COOLING_DEV flag to allow cpufreq core to automatically register as a thermal cooling device. This allows removal of boiler plate code from the driver. Signed-off-by: Amit Kucheria Acked-by: Viresh Kumar Reviewed-by: Daniel Lezcano Signed-off-by: Rafael J. Wysocki --- drivers/cpufreq/qoriq-cpufreq.c | 15 ++------------- 1 file changed, 2 insertions(+), 13 deletions(-) diff --git a/drivers/cpufreq/qoriq-cpufreq.c b/drivers/cpufreq/qoriq-cpufreq.c index 3d773f64b4df..4295e5476264 100644 --- a/drivers/cpufreq/qoriq-cpufreq.c +++ b/drivers/cpufreq/qoriq-cpufreq.c @@ -13,7 +13,6 @@ #include #include #include -#include #include #include #include @@ -31,7 +30,6 @@ struct cpu_data { struct clk **pclk; struct cpufreq_frequency_table *table; - struct thermal_cooling_device *cdev; }; /* @@ -239,7 +237,6 @@ static int qoriq_cpufreq_cpu_exit(struct cpufreq_policy *policy) { struct cpu_data *data = policy->driver_data; - cpufreq_cooling_unregister(data->cdev); kfree(data->pclk); kfree(data->table); kfree(data); @@ -258,23 +255,15 @@ static int qoriq_cpufreq_target(struct cpufreq_policy *policy, return clk_set_parent(policy->clk, parent); } - -static void qoriq_cpufreq_ready(struct cpufreq_policy *policy) -{ - struct cpu_data *cpud = policy->driver_data; - - cpud->cdev = of_cpufreq_cooling_register(policy); -} - static struct cpufreq_driver qoriq_cpufreq_driver = { .name = "qoriq_cpufreq", - .flags = CPUFREQ_CONST_LOOPS, + .flags = CPUFREQ_CONST_LOOPS | + CPUFREQ_IS_COOLING_DEV, .init = qoriq_cpufreq_cpu_init, .exit = qoriq_cpufreq_cpu_exit, .verify = cpufreq_generic_frequency_table_verify, .target_index = qoriq_cpufreq_target, .get = cpufreq_generic_get, - .ready = qoriq_cpufreq_ready, .attr = cpufreq_generic_attr, }; From 5da7af9a94a735c683de6271f5c1704e8e756695 Mon Sep 17 00:00:00 2001 From: Amit Kucheria Date: Tue, 29 Jan 2019 10:25:14 +0530 Subject: [PATCH 12/46] cpufreq: scmi: Use auto-registration of thermal cooling device Use the CPUFREQ_IS_COOLING_DEV flag to allow cpufreq core to automatically register as a thermal cooling device. This allows removal of boiler plate code from the driver. Signed-off-by: Amit Kucheria Acked-by: Sudeep Holla Acked-by: Viresh Kumar Reviewed-by: Daniel Lezcano Signed-off-by: Rafael J. Wysocki --- drivers/cpufreq/scmi-cpufreq.c | 14 ++------------ 1 file changed, 2 insertions(+), 12 deletions(-) diff --git a/drivers/cpufreq/scmi-cpufreq.c b/drivers/cpufreq/scmi-cpufreq.c index 242c3370544e..66b633b48eb1 100644 --- a/drivers/cpufreq/scmi-cpufreq.c +++ b/drivers/cpufreq/scmi-cpufreq.c @@ -11,7 +11,6 @@ #include #include #include -#include #include #include #include @@ -22,7 +21,6 @@ struct scmi_data { int domain_id; struct device *cpu_dev; - struct thermal_cooling_device *cdev; }; static const struct scmi_handle *handle; @@ -185,7 +183,6 @@ static int scmi_cpufreq_exit(struct cpufreq_policy *policy) { struct scmi_data *priv = policy->driver_data; - cpufreq_cooling_unregister(priv->cdev); dev_pm_opp_free_cpufreq_table(priv->cpu_dev, &policy->freq_table); kfree(priv); dev_pm_opp_remove_all_dynamic(priv->cpu_dev); @@ -193,17 +190,11 @@ static int scmi_cpufreq_exit(struct cpufreq_policy *policy) return 0; } -static void scmi_cpufreq_ready(struct cpufreq_policy *policy) -{ - struct scmi_data *priv = policy->driver_data; - - priv->cdev = of_cpufreq_cooling_register(policy); -} - static struct cpufreq_driver scmi_cpufreq_driver = { .name = "scmi", .flags = CPUFREQ_STICKY | CPUFREQ_HAVE_GOVERNOR_PER_POLICY | - CPUFREQ_NEED_INITIAL_FREQ_CHECK, + CPUFREQ_NEED_INITIAL_FREQ_CHECK | + CPUFREQ_IS_COOLING_DEV, .verify = cpufreq_generic_frequency_table_verify, .attr = cpufreq_generic_attr, .target_index = scmi_cpufreq_set_target, @@ -211,7 +202,6 @@ static struct cpufreq_driver scmi_cpufreq_driver = { .get = scmi_cpufreq_get_rate, .init = scmi_cpufreq_init, .exit = scmi_cpufreq_exit, - .ready = scmi_cpufreq_ready, }; static int scmi_cpufreq_probe(struct scmi_device *sdev) From cb772b8ce4b9f5353c34f4d38ab0c697c0071625 Mon Sep 17 00:00:00 2001 From: Amit Kucheria Date: Tue, 29 Jan 2019 10:25:15 +0530 Subject: [PATCH 13/46] cpufreq: scpi: Use auto-registration of thermal cooling device Use the CPUFREQ_IS_COOLING_DEV flag to allow cpufreq core to automatically register as a thermal cooling device. This allows removal of boiler plate code from the driver. Signed-off-by: Amit Kucheria Acked-by: Sudeep Holla Acked-by: Viresh Kumar Reviewed-by: Daniel Lezcano Signed-off-by: Rafael J. Wysocki --- drivers/cpufreq/scpi-cpufreq.c | 14 ++------------ 1 file changed, 2 insertions(+), 12 deletions(-) diff --git a/drivers/cpufreq/scpi-cpufreq.c b/drivers/cpufreq/scpi-cpufreq.c index 99449738faa4..1db2f6927e13 100644 --- a/drivers/cpufreq/scpi-cpufreq.c +++ b/drivers/cpufreq/scpi-cpufreq.c @@ -22,7 +22,6 @@ #include #include #include -#include #include #include #include @@ -34,7 +33,6 @@ struct scpi_data { struct clk *clk; struct device *cpu_dev; - struct thermal_cooling_device *cdev; }; static struct scpi_ops *scpi_ops; @@ -186,7 +184,6 @@ 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); @@ -195,23 +192,16 @@ static int scpi_cpufreq_exit(struct cpufreq_policy *policy) return 0; } -static void scpi_cpufreq_ready(struct cpufreq_policy *policy) -{ - struct scpi_data *priv = policy->driver_data; - - priv->cdev = of_cpufreq_cooling_register(policy); -} - static struct cpufreq_driver scpi_cpufreq_driver = { .name = "scpi-cpufreq", .flags = CPUFREQ_STICKY | CPUFREQ_HAVE_GOVERNOR_PER_POLICY | - CPUFREQ_NEED_INITIAL_FREQ_CHECK, + CPUFREQ_NEED_INITIAL_FREQ_CHECK | + CPUFREQ_IS_COOLING_DEV, .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, }; From 10b818211d04805da731087e596fc7c6ce199dea Mon Sep 17 00:00:00 2001 From: Viresh Kumar Date: Fri, 1 Feb 2019 11:45:44 +0530 Subject: [PATCH 14/46] cpufreq: stats: Declare freq-attr right after their callbacks Freq attribute for "trans_table" is defined right after its callback (without any blank line between them), but the others are defined separately later on. Keep this consistent and define all attributes right after their callbacks. Signed-off-by: Viresh Kumar Signed-off-by: Rafael J. Wysocki --- drivers/cpufreq/cpufreq_stats.c | 7 +++---- 1 file changed, 3 insertions(+), 4 deletions(-) diff --git a/drivers/cpufreq/cpufreq_stats.c b/drivers/cpufreq/cpufreq_stats.c index 1572129844a5..941e63e3e652 100644 --- a/drivers/cpufreq/cpufreq_stats.c +++ b/drivers/cpufreq/cpufreq_stats.c @@ -51,6 +51,7 @@ static ssize_t show_total_trans(struct cpufreq_policy *policy, char *buf) { return sprintf(buf, "%d\n", policy->stats->total_trans); } +cpufreq_freq_attr_ro(total_trans); static ssize_t show_time_in_state(struct cpufreq_policy *policy, char *buf) { @@ -69,6 +70,7 @@ static ssize_t show_time_in_state(struct cpufreq_policy *policy, char *buf) } return len; } +cpufreq_freq_attr_ro(time_in_state); static ssize_t store_reset(struct cpufreq_policy *policy, const char *buf, size_t count) @@ -77,6 +79,7 @@ static ssize_t store_reset(struct cpufreq_policy *policy, const char *buf, cpufreq_stats_clear_table(policy->stats); return count; } +cpufreq_freq_attr_wo(reset); static ssize_t show_trans_table(struct cpufreq_policy *policy, char *buf) { @@ -126,10 +129,6 @@ static ssize_t show_trans_table(struct cpufreq_policy *policy, char *buf) } cpufreq_freq_attr_ro(trans_table); -cpufreq_freq_attr_ro(total_trans); -cpufreq_freq_attr_ro(time_in_state); -cpufreq_freq_attr_wo(reset); - static struct attribute *default_attrs[] = { &total_trans.attr, &time_in_state.attr, From 9795607dc41e7606e90e48bffe6927bee32cd336 Mon Sep 17 00:00:00 2001 From: Viresh Kumar Date: Fri, 1 Feb 2019 11:45:45 +0530 Subject: [PATCH 15/46] cpufreq: stats: Fix concurrency issues while resetting stats It is possible for cpufreq_stats_clear_table() and cpufreq_stats_record_transition() to get called concurrently and they will try to update same variables simultaneously and may lead to corruption of data. Prevent that with the help of existing spinlock. Signed-off-by: Viresh Kumar Signed-off-by: Rafael J. Wysocki --- drivers/cpufreq/cpufreq_stats.c | 9 +++++++-- 1 file changed, 7 insertions(+), 2 deletions(-) diff --git a/drivers/cpufreq/cpufreq_stats.c b/drivers/cpufreq/cpufreq_stats.c index 941e63e3e652..e2db5581489a 100644 --- a/drivers/cpufreq/cpufreq_stats.c +++ b/drivers/cpufreq/cpufreq_stats.c @@ -31,20 +31,20 @@ static void cpufreq_stats_update(struct cpufreq_stats *stats) { unsigned long long cur_time = get_jiffies_64(); - spin_lock(&cpufreq_stats_lock); stats->time_in_state[stats->last_index] += cur_time - stats->last_time; stats->last_time = cur_time; - spin_unlock(&cpufreq_stats_lock); } static void cpufreq_stats_clear_table(struct cpufreq_stats *stats) { unsigned int count = stats->max_state; + spin_lock(&cpufreq_stats_lock); memset(stats->time_in_state, 0, count * sizeof(u64)); memset(stats->trans_table, 0, count * count * sizeof(int)); stats->last_time = get_jiffies_64(); stats->total_trans = 0; + spin_unlock(&cpufreq_stats_lock); } static ssize_t show_total_trans(struct cpufreq_policy *policy, char *buf) @@ -62,7 +62,10 @@ static ssize_t show_time_in_state(struct cpufreq_policy *policy, char *buf) if (policy->fast_switch_enabled) return 0; + spin_lock(&cpufreq_stats_lock); cpufreq_stats_update(stats); + spin_unlock(&cpufreq_stats_lock); + for (i = 0; i < stats->state_num; i++) { len += sprintf(buf + len, "%u %llu\n", stats->freq_table[i], (unsigned long long) @@ -239,9 +242,11 @@ void cpufreq_stats_record_transition(struct cpufreq_policy *policy, if (old_index == -1 || new_index == -1 || old_index == new_index) return; + spin_lock(&cpufreq_stats_lock); cpufreq_stats_update(stats); stats->last_index = new_index; stats->trans_table[old_index * stats->max_state + new_index]++; stats->total_trans++; + spin_unlock(&cpufreq_stats_lock); } From 285881b51eb58dacae78073763e782e56e2fb253 Mon Sep 17 00:00:00 2001 From: Viresh Kumar Date: Thu, 31 Jan 2019 13:53:25 +0530 Subject: [PATCH 16/46] PM / OPP: Remove unused parameter of _generic_set_opp_clk_only() The previous frequency value isn't getting used in the routine _generic_set_opp_clk_only(), drop it. Reviewed-by: Stephen Boyd Signed-off-by: Viresh Kumar --- drivers/opp/core.c | 11 +++++------ 1 file changed, 5 insertions(+), 6 deletions(-) diff --git a/drivers/opp/core.c b/drivers/opp/core.c index e5507add8f04..57d88a275de0 100644 --- a/drivers/opp/core.c +++ b/drivers/opp/core.c @@ -533,9 +533,8 @@ static int _set_opp_voltage(struct device *dev, struct regulator *reg, return ret; } -static inline int -_generic_set_opp_clk_only(struct device *dev, struct clk *clk, - unsigned long old_freq, unsigned long freq) +static inline int _generic_set_opp_clk_only(struct device *dev, struct clk *clk, + unsigned long freq) { int ret; @@ -572,7 +571,7 @@ static int _generic_set_opp_regulator(const struct opp_table *opp_table, } /* Change frequency */ - ret = _generic_set_opp_clk_only(dev, opp_table->clk, old_freq, freq); + ret = _generic_set_opp_clk_only(dev, opp_table->clk, freq); if (ret) goto restore_voltage; @@ -586,7 +585,7 @@ static int _generic_set_opp_regulator(const struct opp_table *opp_table, return 0; restore_freq: - if (_generic_set_opp_clk_only(dev, opp_table->clk, freq, old_freq)) + if (_generic_set_opp_clk_only(dev, opp_table->clk, old_freq)) dev_err(dev, "%s: failed to restore old-freq (%lu Hz)\n", __func__, old_freq); restore_voltage: @@ -759,7 +758,7 @@ int dev_pm_opp_set_rate(struct device *dev, unsigned long target_freq) opp->supplies); } else { /* Only frequency scaling */ - ret = _generic_set_opp_clk_only(dev, clk, old_freq, freq); + ret = _generic_set_opp_clk_only(dev, clk, freq); } /* Scaling down? Configure required OPPs after frequency */ From a4f342b9607d8c2034d3135cbbb11b4028be3678 Mon Sep 17 00:00:00 2001 From: Quentin Perret Date: Mon, 4 Feb 2019 11:09:48 +0000 Subject: [PATCH 17/46] PM / OPP: Introduce a power estimation helper The Energy Model (EM) framework provides an API to let drivers register the active power of CPUs. The drivers are expected to provide a callback method which estimates the power consumed by a CPU at each available performance levels. How exactly this should be implemented, however, depends on the platform. On some systems, PM_OPP knows the voltage and frequency at which CPUs can run. When coupled with the CPU 'capacitance' (as provided by the 'dynamic-power-coefficient' devicetree binding), it is possible to estimate the dynamic power consumption of a CPU as P = C * V^2 * f, with C its capacitance and V and f respectively the voltage and frequency of the OPP. The Intelligent Power Allocator (IPA) thermal governor already implements that estimation method, in the thermal framework. However, this power estimation method can be applied to any platform where all the parameters are known (C, V and f), and not only those suffering thermal issues. As such, the code implementing this feature can be re-used to also populate the EM framework now used by EAS. As a first step, introduce in PM_OPP a helper function which CPUFreq drivers can use to register into the EM framework. This duplicates the power estimation done in IPA until it can be migrated to using the EM framework. This will be done later, once the EM framework has support for at least all platforms currently supported by IPA. Signed-off-by: Quentin Perret Tested-by: Matthias Kaehlcke Reviewed-by: Matthias Kaehlcke Signed-off-by: Viresh Kumar --- drivers/opp/of.c | 99 ++++++++++++++++++++++++++++++++++++++++++ include/linux/pm_opp.h | 6 +++ 2 files changed, 105 insertions(+) diff --git a/drivers/opp/of.c b/drivers/opp/of.c index 06f0f632ec47..cd58959e5158 100644 --- a/drivers/opp/of.c +++ b/drivers/opp/of.c @@ -20,6 +20,7 @@ #include #include #include +#include #include "opp.h" @@ -1047,3 +1048,101 @@ struct device_node *dev_pm_opp_get_of_node(struct dev_pm_opp *opp) return of_node_get(opp->np); } EXPORT_SYMBOL_GPL(dev_pm_opp_get_of_node); + +/* + * Callback function provided to the Energy Model framework upon registration. + * This computes the power estimated by @CPU at @kHz if it is the frequency + * of an existing OPP, or at the frequency of the first OPP above @kHz otherwise + * (see dev_pm_opp_find_freq_ceil()). This function updates @kHz to the ceiled + * frequency and @mW to the associated power. The power is estimated as + * P = C * V^2 * f with C being the CPU's capacitance and V and f respectively + * the voltage and frequency of the OPP. + * + * Returns -ENODEV if the CPU device cannot be found, -EINVAL if the power + * calculation failed because of missing parameters, 0 otherwise. + */ +static int __maybe_unused _get_cpu_power(unsigned long *mW, unsigned long *kHz, + int cpu) +{ + struct device *cpu_dev; + struct dev_pm_opp *opp; + struct device_node *np; + unsigned long mV, Hz; + u32 cap; + u64 tmp; + int ret; + + cpu_dev = get_cpu_device(cpu); + if (!cpu_dev) + return -ENODEV; + + np = of_node_get(cpu_dev->of_node); + if (!np) + return -EINVAL; + + ret = of_property_read_u32(np, "dynamic-power-coefficient", &cap); + of_node_put(np); + if (ret) + return -EINVAL; + + Hz = *kHz * 1000; + opp = dev_pm_opp_find_freq_ceil(cpu_dev, &Hz); + if (IS_ERR(opp)) + return -EINVAL; + + mV = dev_pm_opp_get_voltage(opp) / 1000; + dev_pm_opp_put(opp); + if (!mV) + return -EINVAL; + + tmp = (u64)cap * mV * mV * (Hz / 1000000); + do_div(tmp, 1000000000); + + *mW = (unsigned long)tmp; + *kHz = Hz / 1000; + + return 0; +} + +/** + * dev_pm_opp_of_register_em() - Attempt to register an Energy Model + * @cpus : CPUs for which an Energy Model has to be registered + * + * This checks whether the "dynamic-power-coefficient" devicetree property has + * been specified, and tries to register an Energy Model with it if it has. + */ +void dev_pm_opp_of_register_em(struct cpumask *cpus) +{ + struct em_data_callback em_cb = EM_DATA_CB(_get_cpu_power); + int ret, nr_opp, cpu = cpumask_first(cpus); + struct device *cpu_dev; + struct device_node *np; + u32 cap; + + cpu_dev = get_cpu_device(cpu); + if (!cpu_dev) + return; + + nr_opp = dev_pm_opp_get_opp_count(cpu_dev); + if (nr_opp <= 0) + return; + + np = of_node_get(cpu_dev->of_node); + if (!np) + return; + + /* + * Register an EM only if the 'dynamic-power-coefficient' property is + * set in devicetree. It is assumed the voltage values are known if that + * property is set since it is useless otherwise. If voltages are not + * known, just let the EM registration fail with an error to alert the + * user about the inconsistent configuration. + */ + ret = of_property_read_u32(np, "dynamic-power-coefficient", &cap); + of_node_put(np); + if (ret || !cap) + return; + + em_register_perf_domain(cpus, nr_opp, &em_cb); +} +EXPORT_SYMBOL_GPL(dev_pm_opp_of_register_em); diff --git a/include/linux/pm_opp.h b/include/linux/pm_opp.h index 0a2a88e5a383..1470c57933cf 100644 --- a/include/linux/pm_opp.h +++ b/include/linux/pm_opp.h @@ -322,6 +322,7 @@ int dev_pm_opp_of_get_sharing_cpus(struct device *cpu_dev, struct cpumask *cpuma struct device_node *dev_pm_opp_of_get_opp_desc_node(struct device *dev); struct device_node *dev_pm_opp_get_of_node(struct dev_pm_opp *opp); int of_get_required_opp_performance_state(struct device_node *np, int index); +void dev_pm_opp_of_register_em(struct cpumask *cpus); #else static inline int dev_pm_opp_of_add_table(struct device *dev) { @@ -360,6 +361,11 @@ static inline struct device_node *dev_pm_opp_get_of_node(struct dev_pm_opp *opp) { return NULL; } + +static inline void dev_pm_opp_of_register_em(struct cpumask *cpus) +{ +} + static inline int of_get_required_opp_performance_state(struct device_node *np, int index) { return -ENOTSUPP; From 70e6e7d92bca041b66dac4512821a3aa72dda0cd Mon Sep 17 00:00:00 2001 From: Lukas Bulwahn Date: Wed, 16 Jan 2019 07:37:02 +0100 Subject: [PATCH 18/46] MAINTAINERS: use common indentation Commit 46e2856b8e18 ("cpufreq: Add Kryo CPU scaling driver") slips in some formatting with spaces instead of tabs, which are used in the common format for the MAINTAINERS file. Also update to Ilia's new email address, as Ilia requested. Fixes: 46e2856b8e18 ("cpufreq: Add Kryo CPU scaling driver") Signed-off-by: Lukas Bulwahn Signed-off-by: Viresh Kumar --- MAINTAINERS | 10 +++++----- 1 file changed, 5 insertions(+), 5 deletions(-) diff --git a/MAINTAINERS b/MAINTAINERS index 32d444476a90..6957548f73a4 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -12593,11 +12593,11 @@ F: Documentation/media/v4l-drivers/qcom_camss.rst F: drivers/media/platform/qcom/camss/ QUALCOMM CPUFREQ DRIVER MSM8996/APQ8096 -M: Ilia Lin -L: linux-pm@vger.kernel.org -S: Maintained -F: Documentation/devicetree/bindings/opp/kryo-cpufreq.txt -F: drivers/cpufreq/qcom-cpufreq-kryo.c +M: Ilia Lin +L: linux-pm@vger.kernel.org +S: Maintained +F: Documentation/devicetree/bindings/opp/kryo-cpufreq.txt +F: drivers/cpufreq/qcom-cpufreq-kryo.c QUALCOMM EMAC GIGABIT ETHERNET DRIVER M: Timur Tabi From e0e5b2b4f427cf6f403c6edfb52ce2b0e3acae23 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Pawe=C5=82=20Chmiel?= Date: Sun, 13 Jan 2019 20:57:54 +0100 Subject: [PATCH 19/46] cpufreq: s5pv210: Defer probe if getting regulators fail MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit There is possibility, that when probing driver, regulators are not yet initialized. In this case we should return EPROBE_DEFER and wait till they're initialized, since they're required currently for cpufreq driver to work. Also move regulator initialization code at beginning of probe, so we can defer as fast as posibble. Signed-off-by: Paweł Chmiel Reviewed-by: Krzysztof Kozlowski Signed-off-by: Viresh Kumar --- drivers/cpufreq/s5pv210-cpufreq.c | 67 ++++++++++++++++++++++--------- 1 file changed, 48 insertions(+), 19 deletions(-) diff --git a/drivers/cpufreq/s5pv210-cpufreq.c b/drivers/cpufreq/s5pv210-cpufreq.c index dbecd7667db2..5b4289460bc9 100644 --- a/drivers/cpufreq/s5pv210-cpufreq.c +++ b/drivers/cpufreq/s5pv210-cpufreq.c @@ -584,7 +584,7 @@ static struct notifier_block s5pv210_cpufreq_reboot_notifier = { static int s5pv210_cpufreq_probe(struct platform_device *pdev) { struct device_node *np; - int id; + int id, result = 0; /* * HACK: This is a temporary workaround to get access to clock @@ -594,18 +594,39 @@ static int s5pv210_cpufreq_probe(struct platform_device *pdev) * this whole driver as soon as S5PV210 gets migrated to use * cpufreq-dt driver. */ + arm_regulator = regulator_get(NULL, "vddarm"); + if (IS_ERR(arm_regulator)) { + if (PTR_ERR(arm_regulator) == -EPROBE_DEFER) + pr_debug("vddarm regulator not ready, defer\n"); + else + pr_err("failed to get regulator vddarm\n"); + return PTR_ERR(arm_regulator); + } + + int_regulator = regulator_get(NULL, "vddint"); + if (IS_ERR(int_regulator)) { + if (PTR_ERR(int_regulator) == -EPROBE_DEFER) + pr_debug("vddint regulator not ready, defer\n"); + else + pr_err("failed to get regulator vddint\n"); + result = PTR_ERR(int_regulator); + goto err_int_regulator; + } + np = of_find_compatible_node(NULL, NULL, "samsung,s5pv210-clock"); if (!np) { pr_err("%s: failed to find clock controller DT node\n", __func__); - return -ENODEV; + result = -ENODEV; + goto err_clock; } clk_base = of_iomap(np, 0); of_node_put(np); if (!clk_base) { pr_err("%s: failed to map clock registers\n", __func__); - return -EFAULT; + result = -EFAULT; + goto err_clock; } for_each_compatible_node(np, NULL, "samsung,s5pv210-dmc") { @@ -614,7 +635,8 @@ static int s5pv210_cpufreq_probe(struct platform_device *pdev) pr_err("%s: failed to get alias of dmc node '%pOFn'\n", __func__, np); of_node_put(np); - return id; + result = id; + goto err_clk_base; } dmc_base[id] = of_iomap(np, 0); @@ -622,33 +644,40 @@ static int s5pv210_cpufreq_probe(struct platform_device *pdev) pr_err("%s: failed to map dmc%d registers\n", __func__, id); of_node_put(np); - return -EFAULT; + result = -EFAULT; + goto err_dmc; } } for (id = 0; id < ARRAY_SIZE(dmc_base); ++id) { if (!dmc_base[id]) { pr_err("%s: failed to find dmc%d node\n", __func__, id); - return -ENODEV; + result = -ENODEV; + goto err_dmc; } } - arm_regulator = regulator_get(NULL, "vddarm"); - if (IS_ERR(arm_regulator)) { - pr_err("failed to get regulator vddarm\n"); - return PTR_ERR(arm_regulator); - } - - int_regulator = regulator_get(NULL, "vddint"); - if (IS_ERR(int_regulator)) { - pr_err("failed to get regulator vddint\n"); - regulator_put(arm_regulator); - return PTR_ERR(int_regulator); - } - register_reboot_notifier(&s5pv210_cpufreq_reboot_notifier); return cpufreq_register_driver(&s5pv210_driver); + +err_dmc: + for (id = 0; id < ARRAY_SIZE(dmc_base); ++id) + if (dmc_base[id]) { + iounmap(dmc_base[id]); + dmc_base[id] = NULL; + } + +err_clk_base: + iounmap(clk_base); + +err_clock: + regulator_put(int_regulator); + +err_int_regulator: + regulator_put(arm_regulator); + + return result; } static struct platform_driver s5pv210_cpufreq_platdrv = { From 8e3151d16c7d8d3576fee23a6c898dd54ead4e0a Mon Sep 17 00:00:00 2001 From: Gregory CLEMENT Date: Fri, 18 Jan 2019 15:11:39 +0100 Subject: [PATCH 20/46] MAINTAINERS: add new entries for Armada 8K cpufreq driver This new driver belongs to the mvebu family, update the MAINTAINER file to document it. Signed-off-by: Gregory CLEMENT Signed-off-by: Viresh Kumar --- MAINTAINERS | 1 + 1 file changed, 1 insertion(+) diff --git a/MAINTAINERS b/MAINTAINERS index 6957548f73a4..fabbd9a59ab5 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -1737,6 +1737,7 @@ 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/armada-8k-cpufreq.c F: drivers/cpufreq/mvebu-cpufreq.c F: drivers/irqchip/irq-armada-370-xp.c F: drivers/irqchip/irq-mvebu-* From f525a670533d961fd72ab748e3aac002d7b3d1b9 Mon Sep 17 00:00:00 2001 From: Gregory CLEMENT Date: Fri, 18 Jan 2019 15:51:01 +0100 Subject: [PATCH 21/46] cpufreq: ap806: add cpufreq driver for Armada 8K Add cpufreq driver for Marvell AP-806 found on Aramda 8K. The AP-806 has DFS (Dynamic Frequency Scaling) with coupled clock domain for two clusters, so this driver will directly use generic cpufreq-dt driver as backend. Based on the work of Omri Itach . Signed-off-by: Gregory CLEMENT Signed-off-by: Viresh Kumar --- drivers/cpufreq/Kconfig.arm | 11 ++ drivers/cpufreq/Makefile | 1 + drivers/cpufreq/armada-8k-cpufreq.c | 204 ++++++++++++++++++++++++++++ 3 files changed, 216 insertions(+) create mode 100644 drivers/cpufreq/armada-8k-cpufreq.c diff --git a/drivers/cpufreq/Kconfig.arm b/drivers/cpufreq/Kconfig.arm index 688f10227793..10bc5c798d17 100644 --- a/drivers/cpufreq/Kconfig.arm +++ b/drivers/cpufreq/Kconfig.arm @@ -25,6 +25,17 @@ config ARM_ARMADA_37XX_CPUFREQ This adds the CPUFreq driver support for Marvell Armada 37xx SoCs. The Armada 37xx PMU supports 4 frequency and VDD levels. +config ARM_ARMADA_8K_CPUFREQ + tristate "Armada 8K CPUFreq driver" + depends on ARCH_MVEBU && CPUFREQ_DT + help + This enables the CPUFreq driver support for Marvell + Armada8k SOCs. + Armada8K device has the AP806 which supports scaling + to any full integer divider. + + If in doubt, say N. + # 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 08c071be2491..689b26c6f949 100644 --- a/drivers/cpufreq/Makefile +++ b/drivers/cpufreq/Makefile @@ -50,6 +50,7 @@ obj-$(CONFIG_X86_SFI_CPUFREQ) += sfi-cpufreq.o obj-$(CONFIG_ARM_BIG_LITTLE_CPUFREQ) += arm_big_little.o obj-$(CONFIG_ARM_ARMADA_37XX_CPUFREQ) += armada-37xx-cpufreq.o +obj-$(CONFIG_ARM_ARMADA_8K_CPUFREQ) += armada-8k-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-8k-cpufreq.c b/drivers/cpufreq/armada-8k-cpufreq.c new file mode 100644 index 000000000000..8a5ddb93fc58 --- /dev/null +++ b/drivers/cpufreq/armada-8k-cpufreq.c @@ -0,0 +1,204 @@ +// SPDX-License-Identifier: GPL-2.0+ +/* + * CPUFreq support for Armada 8K + * + * Copyright (C) 2018 Marvell + * + * Omri Itach + * Gregory Clement + */ + +#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt + +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include + +/* + * Setup the opps list with the divider for the max frequency, that + * will be filled at runtime. + */ +static const int opps_div[] __initconst = {1, 2, 3, 4}; + +static struct platform_device *armada_8k_pdev; + +struct freq_table { + struct device *cpu_dev; + unsigned int freq[ARRAY_SIZE(opps_div)]; +}; + +/* If the CPUs share the same clock, then they are in the same cluster. */ +static void __init armada_8k_get_sharing_cpus(struct clk *cur_clk, + struct cpumask *cpumask) +{ + int cpu; + + for_each_possible_cpu(cpu) { + struct device *cpu_dev; + struct clk *clk; + + cpu_dev = get_cpu_device(cpu); + if (!cpu_dev) { + pr_warn("Failed to get cpu%d device\n", cpu); + continue; + } + + clk = clk_get(cpu_dev, 0); + if (IS_ERR(clk)) { + pr_warn("Cannot get clock for CPU %d\n", cpu); + } else { + if (clk_is_match(clk, cur_clk)) + cpumask_set_cpu(cpu, cpumask); + + clk_put(clk); + } + } +} + +static int __init armada_8k_add_opp(struct clk *clk, struct device *cpu_dev, + struct freq_table *freq_tables, + int opps_index) +{ + unsigned int cur_frequency; + unsigned int freq; + int i, ret; + + /* Get nominal (current) CPU frequency. */ + cur_frequency = clk_get_rate(clk); + if (!cur_frequency) { + dev_err(cpu_dev, "Failed to get clock rate for this CPU\n"); + return -EINVAL; + } + + freq_tables[opps_index].cpu_dev = cpu_dev; + + for (i = 0; i < ARRAY_SIZE(opps_div); i++) { + freq = cur_frequency / opps_div[i]; + + ret = dev_pm_opp_add(cpu_dev, freq, 0); + if (ret) + return ret; + + freq_tables[opps_index].freq[i] = freq; + } + + return 0; +} + +static void armada_8k_cpufreq_free_table(struct freq_table *freq_tables) +{ + int opps_index, nb_cpus = num_possible_cpus(); + + for (opps_index = 0 ; opps_index <= nb_cpus; opps_index++) { + int i; + + /* If cpu_dev is NULL then we reached the end of the array */ + if (!freq_tables[opps_index].cpu_dev) + break; + + for (i = 0; i < ARRAY_SIZE(opps_div); i++) { + /* + * A 0Hz frequency is not valid, this meant + * that it was not yet initialized so there is + * no more opp to free + */ + if (freq_tables[opps_index].freq[i] == 0) + break; + + dev_pm_opp_remove(freq_tables[opps_index].cpu_dev, + freq_tables[opps_index].freq[i]); + } + } + + kfree(freq_tables); +} + +static int __init armada_8k_cpufreq_init(void) +{ + int ret = 0, opps_index = 0, cpu, nb_cpus; + struct freq_table *freq_tables; + struct device_node *node; + struct cpumask cpus; + + node = of_find_compatible_node(NULL, NULL, "marvell,ap806-cpu-clock"); + if (!node || !of_device_is_available(node)) + return -ENODEV; + + nb_cpus = num_possible_cpus(); + freq_tables = kcalloc(nb_cpus, sizeof(*freq_tables), GFP_KERNEL); + cpumask_copy(&cpus, cpu_possible_mask); + + /* + * For each CPU, this loop registers the operating points + * supported (which are the nominal CPU frequency and full integer + * divisions of it). + */ + for_each_cpu(cpu, &cpus) { + struct cpumask shared_cpus; + struct device *cpu_dev; + struct clk *clk; + + cpu_dev = get_cpu_device(cpu); + + if (!cpu_dev) { + pr_err("Cannot get CPU %d\n", cpu); + continue; + } + + clk = clk_get(cpu_dev, 0); + + if (IS_ERR(clk)) { + pr_err("Cannot get clock for CPU %d\n", cpu); + ret = PTR_ERR(clk); + goto remove_opp; + } + + ret = armada_8k_add_opp(clk, cpu_dev, freq_tables, opps_index); + if (ret) { + clk_put(clk); + goto remove_opp; + } + + opps_index++; + cpumask_clear(&shared_cpus); + armada_8k_get_sharing_cpus(clk, &shared_cpus); + dev_pm_opp_set_sharing_cpus(cpu_dev, &shared_cpus); + cpumask_andnot(&cpus, &cpus, &shared_cpus); + clk_put(clk); + } + + armada_8k_pdev = platform_device_register_simple("cpufreq-dt", -1, + NULL, 0); + ret = PTR_ERR_OR_ZERO(armada_8k_pdev); + if (ret) + goto remove_opp; + + platform_set_drvdata(armada_8k_pdev, freq_tables); + + return 0; + +remove_opp: + armada_8k_cpufreq_free_table(freq_tables); + return ret; +} +module_init(armada_8k_cpufreq_init); + +static void __exit armada_8k_cpufreq_exit(void) +{ + struct freq_table *freq_tables = platform_get_drvdata(armada_8k_pdev); + + platform_device_unregister(armada_8k_pdev); + armada_8k_cpufreq_free_table(freq_tables); +} +module_exit(armada_8k_cpufreq_exit); + +MODULE_AUTHOR("Gregory Clement "); +MODULE_DESCRIPTION("Armada 8K cpufreq driver"); +MODULE_LICENSE("GPL"); From 0dc10eac6583602335c275c7dc5975946da30901 Mon Sep 17 00:00:00 2001 From: Baruch Siach Date: Thu, 24 Jan 2019 11:53:11 +0200 Subject: [PATCH 22/46] MAINTAINERS: Update the active pm tree for ARM The Linaro hosted git tree is no longer active. Update the cpufreq entry. Signed-off-by: Baruch Siach Signed-off-by: Viresh Kumar --- MAINTAINERS | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/MAINTAINERS b/MAINTAINERS index fabbd9a59ab5..f0400a466870 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -3958,7 +3958,7 @@ M: Viresh Kumar L: linux-pm@vger.kernel.org S: Maintained T: git git://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pm.git -T: git git://git.linaro.org/people/vireshk/linux.git (For ARM Updates) +T: git git://git.kernel.org/pub/scm/linux/kernel/git/vireshk/pm.git (For ARM Updates) B: https://bugzilla.kernel.org F: Documentation/cpu-freq/ F: Documentation/devicetree/bindings/cpufreq/ From 50c0b12f098fb3cfac7abcc0f2b5409f6bac5fa2 Mon Sep 17 00:00:00 2001 From: Yangtao Li Date: Mon, 4 Feb 2019 01:13:10 -0500 Subject: [PATCH 23/46] cpufreq: qcom-kryo: make some variables static The variables are local to the source and do not need to be in global scope, so make them static. Signed-off-by: Yangtao Li Signed-off-by: Viresh Kumar --- drivers/cpufreq/qcom-cpufreq-kryo.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/cpufreq/qcom-cpufreq-kryo.c b/drivers/cpufreq/qcom-cpufreq-kryo.c index 2a3675c24032..1c8583cc06a2 100644 --- a/drivers/cpufreq/qcom-cpufreq-kryo.c +++ b/drivers/cpufreq/qcom-cpufreq-kryo.c @@ -42,7 +42,7 @@ enum _msm8996_version { NUM_OF_MSM8996_VERSIONS, }; -struct platform_device *cpufreq_dt_pdev, *kryo_cpufreq_pdev; +static struct platform_device *cpufreq_dt_pdev, *kryo_cpufreq_pdev; static enum _msm8996_version qcom_cpufreq_kryo_get_msm_id(void) { From 446fae2bb5395f3028d8e3aae1508737e5a72ea1 Mon Sep 17 00:00:00 2001 From: Yangtao Li Date: Mon, 4 Feb 2019 02:48:54 -0500 Subject: [PATCH 24/46] cpufreq: tegra124: add missing of_node_put() of_cpu_device_node_get() will increase the refcount of device_node, it is necessary to call of_node_put() at the end to release the refcount. Fixes: 9eb15dbbfa1a2 ("cpufreq: Add cpufreq driver for Tegra124") Cc: # 4.4+ Signed-off-by: Yangtao Li Acked-by: Thierry Reding Signed-off-by: Viresh Kumar --- drivers/cpufreq/tegra124-cpufreq.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/drivers/cpufreq/tegra124-cpufreq.c b/drivers/cpufreq/tegra124-cpufreq.c index 43530254201a..4bb154f6c54c 100644 --- a/drivers/cpufreq/tegra124-cpufreq.c +++ b/drivers/cpufreq/tegra124-cpufreq.c @@ -134,6 +134,8 @@ static int tegra124_cpufreq_probe(struct platform_device *pdev) platform_set_drvdata(pdev, priv); + of_node_put(np); + return 0; out_switch_to_pllx: From f896d06665ec8ae09f6412a47b97863d9ef8ff64 Mon Sep 17 00:00:00 2001 From: Amit Kucheria Date: Thu, 10 Jan 2019 05:30:53 +0530 Subject: [PATCH 25/46] cpufreq: qcom-hw: Move to device_initcall subsys_initcall causes problems registering the driver as a thermal cooling device. If "faster boot" is the main reason for doing subsys_initcall, this should be handled in the bootloader or another boot constraint framework. Signed-off-by: Amit Kucheria Signed-off-by: Viresh Kumar --- drivers/cpufreq/qcom-cpufreq-hw.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/cpufreq/qcom-cpufreq-hw.c b/drivers/cpufreq/qcom-cpufreq-hw.c index d83939a1b3d4..649dddd72749 100644 --- a/drivers/cpufreq/qcom-cpufreq-hw.c +++ b/drivers/cpufreq/qcom-cpufreq-hw.c @@ -296,7 +296,7 @@ static int __init qcom_cpufreq_hw_init(void) { return platform_driver_register(&qcom_cpufreq_hw_driver); } -subsys_initcall(qcom_cpufreq_hw_init); +device_initcall(qcom_cpufreq_hw_init); static void __exit qcom_cpufreq_hw_exit(void) { From 55538fbc79e926f159a5abcc571ba1c02529a3d1 Mon Sep 17 00:00:00 2001 From: Taniya Das Date: Thu, 31 Jan 2019 23:02:50 +0530 Subject: [PATCH 26/46] cpufreq: qcom: Read voltage LUT and populate OPP Add support to read the voltage look up table and populate OPP for all corresponding CPUS for consumers like the energy model could use the frequency and voltage from the OPP tables. Also update the logic to not add duplicate OPPs. Tested-by: Matthias Kaehlcke Reviewed-by: Matthias Kaehlcke Signed-off-by: Matthias Kaehlcke Signed-off-by: Taniya Das Signed-off-by: Viresh Kumar --- drivers/cpufreq/qcom-cpufreq-hw.c | 46 +++++++++++++++++++++++++------ 1 file changed, 37 insertions(+), 9 deletions(-) diff --git a/drivers/cpufreq/qcom-cpufreq-hw.c b/drivers/cpufreq/qcom-cpufreq-hw.c index 649dddd72749..72223536ca4e 100644 --- a/drivers/cpufreq/qcom-cpufreq-hw.c +++ b/drivers/cpufreq/qcom-cpufreq-hw.c @@ -10,18 +10,21 @@ #include #include #include +#include #include #define LUT_MAX_ENTRIES 40U #define LUT_SRC GENMASK(31, 30) #define LUT_L_VAL GENMASK(7, 0) #define LUT_CORE_COUNT GENMASK(18, 16) +#define LUT_VOLT GENMASK(11, 0) #define LUT_ROW_SIZE 32 #define CLK_HW_DIV 2 /* Register offsets */ #define REG_ENABLE 0x0 -#define REG_LUT_TABLE 0x110 +#define REG_FREQ_LUT 0x110 +#define REG_VOLT_LUT 0x114 #define REG_PERF_STATE 0x920 static unsigned long cpu_hw_rate, xo_rate; @@ -70,11 +73,12 @@ static unsigned int qcom_cpufreq_hw_fast_switch(struct cpufreq_policy *policy, return policy->freq_table[index].frequency; } -static int qcom_cpufreq_hw_read_lut(struct device *dev, +static int qcom_cpufreq_hw_read_lut(struct device *cpu_dev, struct cpufreq_policy *policy, void __iomem *base) { u32 data, src, lval, i, core_count, prev_cc = 0, prev_freq = 0, freq; + u32 volt; unsigned int max_cores = cpumask_weight(policy->cpus); struct cpufreq_frequency_table *table; @@ -83,23 +87,28 @@ static int qcom_cpufreq_hw_read_lut(struct device *dev, return -ENOMEM; for (i = 0; i < LUT_MAX_ENTRIES; i++) { - data = readl_relaxed(base + REG_LUT_TABLE + i * LUT_ROW_SIZE); + data = readl_relaxed(base + REG_FREQ_LUT + + i * LUT_ROW_SIZE); src = FIELD_GET(LUT_SRC, data); lval = FIELD_GET(LUT_L_VAL, data); core_count = FIELD_GET(LUT_CORE_COUNT, data); + data = readl_relaxed(base + REG_VOLT_LUT + + i * LUT_ROW_SIZE); + volt = FIELD_GET(LUT_VOLT, data) * 1000; + if (src) freq = xo_rate * lval / 1000; else freq = cpu_hw_rate / 1000; - /* Ignore boosts in the middle of the table */ - if (core_count != max_cores) { - table[i].frequency = CPUFREQ_ENTRY_INVALID; - } else { + if (freq != prev_freq && core_count == max_cores) { table[i].frequency = freq; - dev_dbg(dev, "index=%d freq=%d, core_count %d\n", i, + dev_pm_opp_add(cpu_dev, freq * 1000, volt); + dev_dbg(cpu_dev, "index=%d freq=%d, core_count %d\n", i, freq, core_count); + } else { + table[i].frequency = CPUFREQ_ENTRY_INVALID; } /* @@ -116,6 +125,7 @@ static int qcom_cpufreq_hw_read_lut(struct device *dev, if (prev_cc != max_cores) { prev->frequency = prev_freq; prev->flags = CPUFREQ_BOOST_FREQ; + dev_pm_opp_add(cpu_dev, prev_freq * 1000, volt); } break; @@ -127,6 +137,7 @@ static int qcom_cpufreq_hw_read_lut(struct device *dev, table[i].frequency = CPUFREQ_TABLE_END; policy->freq_table = table; + dev_pm_opp_set_sharing_cpus(cpu_dev, policy->cpus); return 0; } @@ -159,10 +170,18 @@ static int qcom_cpufreq_hw_cpu_init(struct cpufreq_policy *policy) struct device *dev = &global_pdev->dev; struct of_phandle_args args; struct device_node *cpu_np; + struct device *cpu_dev; struct resource *res; void __iomem *base; int ret, index; + cpu_dev = get_cpu_device(policy->cpu); + if (!cpu_dev) { + pr_err("%s: failed to get cpu%d device\n", __func__, + policy->cpu); + return -ENODEV; + } + cpu_np = of_cpu_device_node_get(policy->cpu); if (!cpu_np) return -EINVAL; @@ -199,12 +218,19 @@ static int qcom_cpufreq_hw_cpu_init(struct cpufreq_policy *policy) policy->driver_data = base + REG_PERF_STATE; - ret = qcom_cpufreq_hw_read_lut(dev, policy, base); + ret = qcom_cpufreq_hw_read_lut(cpu_dev, policy, base); if (ret) { dev_err(dev, "Domain-%d failed to read LUT\n", index); goto error; } + ret = dev_pm_opp_get_opp_count(cpu_dev); + if (ret <= 0) { + dev_err(cpu_dev, "Failed to add OPPs\n"); + ret = -ENODEV; + goto error; + } + policy->fast_switch_possible = true; return 0; @@ -215,8 +241,10 @@ error: static int qcom_cpufreq_hw_cpu_exit(struct cpufreq_policy *policy) { + struct device *cpu_dev = get_cpu_device(policy->cpu); void __iomem *base = policy->driver_data - REG_PERF_STATE; + dev_pm_opp_remove_all_dynamic(cpu_dev); kfree(policy->freq_table); devm_iounmap(&global_pdev->dev, base); From dab535052f67db0ff48b1b23e714b58650d1a787 Mon Sep 17 00:00:00 2001 From: Matthias Kaehlcke Date: Tue, 5 Feb 2019 09:52:24 -0800 Subject: [PATCH 27/46] cpufreq: qcom-hw: Register an Energy Model Try and register an Energy Model from qcom-cpufreq-hw to allow interested sub-systems like the task scheduler to use the provided information. Signed-off-by: Matthias Kaehlcke [ Viresh: Rebased over cpufreq related changes ] Signed-off-by: Viresh Kumar --- drivers/cpufreq/qcom-cpufreq-hw.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/drivers/cpufreq/qcom-cpufreq-hw.c b/drivers/cpufreq/qcom-cpufreq-hw.c index 72223536ca4e..c647cfb1f5e6 100644 --- a/drivers/cpufreq/qcom-cpufreq-hw.c +++ b/drivers/cpufreq/qcom-cpufreq-hw.c @@ -231,6 +231,8 @@ static int qcom_cpufreq_hw_cpu_init(struct cpufreq_policy *policy) goto error; } + dev_pm_opp_of_register_em(policy->cpus); + policy->fast_switch_possible = true; return 0; From 91a12e91dc39137906d929a4ff6f9c32c59697fa Mon Sep 17 00:00:00 2001 From: Viresh Kumar Date: Tue, 12 Feb 2019 16:36:04 +0530 Subject: [PATCH 28/46] cpufreq: Allow light-weight tear down and bring up of CPUs The cpufreq core doesn't remove the cpufreq policy anymore on CPU offline operation, rather that happens when the CPU device gets unregistered from the kernel. This allows faster recovery when the CPU comes back online. This is also very useful during system wide suspend/resume where we offline all non-boot CPUs during suspend and then bring them back on resume. This commit takes the same idea a step ahead to allow drivers to do light weight tear-down and bring-up during CPU offline and online operations. A new set of callbacks is introduced, online/offline(). online() gets called when the first CPU of an inactive policy is brought up and offline() gets called when all the CPUs of a policy are offlined. The existing init/exit() callback get called on policy creation/destruction. They also get called instead of online/offline() callbacks if the online/offline() callbacks aren't provided. This also moves around some code to get executed only for the new-policy case going forward. Signed-off-by: Viresh Kumar Signed-off-by: Rafael J. Wysocki --- drivers/cpufreq/cpufreq.c | 58 +++++++++++++++++++++++++-------------- include/linux/cpufreq.h | 2 ++ 2 files changed, 40 insertions(+), 20 deletions(-) diff --git a/drivers/cpufreq/cpufreq.c b/drivers/cpufreq/cpufreq.c index 96a69c67a545..55e9795801a4 100644 --- a/drivers/cpufreq/cpufreq.c +++ b/drivers/cpufreq/cpufreq.c @@ -1201,28 +1201,39 @@ static int cpufreq_online(unsigned int cpu) return -ENOMEM; } - cpumask_copy(policy->cpus, cpumask_of(cpu)); + if (!new_policy && cpufreq_driver->online) { + ret = cpufreq_driver->online(policy); + if (ret) { + pr_debug("%s: %d: initialization failed\n", __func__, + __LINE__); + goto out_exit_policy; + } - /* call driver. From then on the cpufreq must be able - * to accept all calls to ->verify and ->setpolicy for this CPU - */ - ret = cpufreq_driver->init(policy); - if (ret) { - pr_debug("initialization failed\n"); - goto out_free_policy; - } + /* Recover policy->cpus using related_cpus */ + cpumask_copy(policy->cpus, policy->related_cpus); + } else { + cpumask_copy(policy->cpus, cpumask_of(cpu)); - ret = cpufreq_table_validate_and_sort(policy); - if (ret) - goto out_exit_policy; + /* + * Call driver. From then on the cpufreq must be able + * to accept all calls to ->verify and ->setpolicy for this CPU. + */ + ret = cpufreq_driver->init(policy); + if (ret) { + pr_debug("%s: %d: initialization failed\n", __func__, + __LINE__); + goto out_free_policy; + } - down_write(&policy->rwsem); + ret = cpufreq_table_validate_and_sort(policy); + if (ret) + goto out_exit_policy; - if (new_policy) { /* related_cpus should at least include policy->cpus. */ cpumask_copy(policy->related_cpus, policy->cpus); } + down_write(&policy->rwsem); /* * affected cpus must always be the one, which are online. We aren't * managing offline cpus here. @@ -1421,11 +1432,12 @@ static int cpufreq_offline(unsigned int cpu) cpufreq_exit_governor(policy); /* - * Perform the ->exit() even during light-weight tear-down, - * since this is a core component, and is essential for the - * subsequent light-weight ->init() to succeed. + * Perform the ->offline() during light-weight tear-down, as + * that allows fast recovery when the CPU comes back. */ - if (cpufreq_driver->exit) { + if (cpufreq_driver->offline) { + cpufreq_driver->offline(policy); + } else if (cpufreq_driver->exit) { cpufreq_driver->exit(policy); policy->freq_table = NULL; } @@ -1454,8 +1466,13 @@ static void cpufreq_remove_dev(struct device *dev, struct subsys_interface *sif) cpumask_clear_cpu(cpu, policy->real_cpus); remove_cpu_dev_symlink(policy, dev); - if (cpumask_empty(policy->real_cpus)) + if (cpumask_empty(policy->real_cpus)) { + /* We did light-weight exit earlier, do full tear down now */ + if (cpufreq_driver->offline) + cpufreq_driver->exit(policy); + cpufreq_policy_free(policy); + } } /** @@ -2488,7 +2505,8 @@ int cpufreq_register_driver(struct cpufreq_driver *driver_data) driver_data->target) || (driver_data->setpolicy && (driver_data->target_index || driver_data->target)) || - (!!driver_data->get_intermediate != !!driver_data->target_intermediate)) + (!!driver_data->get_intermediate != !!driver_data->target_intermediate) || + (!driver_data->online != !driver_data->offline)) return -EINVAL; pr_debug("trying to register driver %s\n", driver_data->name); diff --git a/include/linux/cpufreq.h b/include/linux/cpufreq.h index 9db074ecbbd7..b160e98076e3 100644 --- a/include/linux/cpufreq.h +++ b/include/linux/cpufreq.h @@ -325,6 +325,8 @@ struct cpufreq_driver { /* optional */ int (*bios_limit)(int cpu, unsigned int *limit); + int (*online)(struct cpufreq_policy *policy); + int (*offline)(struct cpufreq_policy *policy); int (*exit)(struct cpufreq_policy *policy); void (*stop_cpu)(struct cpufreq_policy *policy); int (*suspend)(struct cpufreq_policy *policy); From 263abfe74b5f1058887b14e7668b0d737e040e42 Mon Sep 17 00:00:00 2001 From: Viresh Kumar Date: Tue, 12 Feb 2019 12:36:47 +0530 Subject: [PATCH 29/46] cpufreq: dt: Implement online/offline() callbacks Implement the light-weight tear down and bring up helpers to reduce the amount of work to do on CPU offline/online operation. Signed-off-by: Viresh Kumar Signed-off-by: Rafael J. Wysocki --- drivers/cpufreq/cpufreq-dt.c | 17 +++++++++++++++++ 1 file changed, 17 insertions(+) diff --git a/drivers/cpufreq/cpufreq-dt.c b/drivers/cpufreq/cpufreq-dt.c index 7ba392911cd0..1aefaa1b0ca2 100644 --- a/drivers/cpufreq/cpufreq-dt.c +++ b/drivers/cpufreq/cpufreq-dt.c @@ -295,6 +295,21 @@ out_put_clk: return ret; } +static int cpufreq_online(struct cpufreq_policy *policy) +{ + /* We did light-weight tear down earlier, nothing to do here */ + return 0; +} + +static int cpufreq_offline(struct cpufreq_policy *policy) +{ + /* + * Preserve policy->driver_data and don't free resources on light-weight + * tear down. + */ + return 0; +} + static int cpufreq_exit(struct cpufreq_policy *policy) { struct private_data *priv = policy->driver_data; @@ -319,6 +334,8 @@ static struct cpufreq_driver dt_cpufreq_driver = { .get = cpufreq_generic_get, .init = cpufreq_init, .exit = cpufreq_exit, + .online = cpufreq_online, + .offline = cpufreq_offline, .name = "cpufreq-dt", .attr = cpufreq_dt_attr, .suspend = cpufreq_generic_suspend, From 076b862c7e4409d2dcacfda19f7eaf8d07ab9200 Mon Sep 17 00:00:00 2001 From: Erwan Velu Date: Wed, 13 Feb 2019 09:58:35 +0100 Subject: [PATCH 30/46] cpufreq: intel_pstate: Add reasons for failure and debug messages The init code path has several exceptions where the driver can decide not to load. As CONFIG_X86_INTEL_PSTATE is generally set to Y, the return code is not reachable. The initialization code is neither verbose of the reason why it did choose to prematurely exit, so it is difficult for a user to determine, on a given platform, why the driver didn't load properly. This patch is about reporting to the user the reason/context of why the driver failed to load. That is a precious hint when debugging a platform. Signed-off-by: Erwan Velu [ rjw: Subject & changelog, minor fixups ] Signed-off-by: Rafael J. Wysocki --- drivers/cpufreq/intel_pstate.c | 27 +++++++++++++++++++++------ 1 file changed, 21 insertions(+), 6 deletions(-) diff --git a/drivers/cpufreq/intel_pstate.c b/drivers/cpufreq/intel_pstate.c index 5ab6a4fe93aa..75a5408e6981 100644 --- a/drivers/cpufreq/intel_pstate.c +++ b/drivers/cpufreq/intel_pstate.c @@ -2476,6 +2476,7 @@ static bool __init intel_pstate_no_acpi_pss(void) kfree(pss); } + pr_debug("ACPI _PSS not found\n"); return true; } @@ -2486,9 +2487,14 @@ static bool __init intel_pstate_no_acpi_pcch(void) status = acpi_get_handle(NULL, "\\_SB", &handle); if (ACPI_FAILURE(status)) - return true; + goto not_found; - return !acpi_has_method(handle, "PCCH"); + if (acpi_has_method(handle, "PCCH")) + return false; + +not_found: + pr_debug("ACPI PCCH not found\n"); + return true; } static bool __init intel_pstate_has_acpi_ppc(void) @@ -2503,6 +2509,7 @@ static bool __init intel_pstate_has_acpi_ppc(void) if (acpi_has_method(pr->handle, "_PPC")) return true; } + pr_debug("ACPI _PPC not found\n"); return false; } @@ -2540,8 +2547,10 @@ static bool __init intel_pstate_platform_pwr_mgmt_exists(void) id = x86_match_cpu(intel_pstate_cpu_oob_ids); if (id) { rdmsrl(MSR_MISC_PWR_MGMT, misc_pwr); - if ( misc_pwr & (1 << 8)) + if (misc_pwr & (1 << 8)) { + pr_debug("Bit 8 in the MISC_PWR_MGMT MSR set\n"); return true; + } } idx = acpi_match_platform_list(plat_info); @@ -2607,22 +2616,28 @@ static int __init intel_pstate_init(void) } } else { id = x86_match_cpu(intel_pstate_cpu_ids); - if (!id) + if (!id) { + pr_info("CPU ID not supported\n"); return -ENODEV; + } copy_cpu_funcs((struct pstate_funcs *)id->driver_data); } - if (intel_pstate_msrs_not_valid()) + if (intel_pstate_msrs_not_valid()) { + pr_info("Invalid MSRs\n"); return -ENODEV; + } hwp_cpu_matched: /* * The Intel pstate driver will be ignored if the platform * firmware has its own power management modes. */ - if (intel_pstate_platform_pwr_mgmt_exists()) + if (intel_pstate_platform_pwr_mgmt_exists()) { + pr_info("P-states controlled by the platform\n"); return -ENODEV; + } if (!hwp_active && hwp_only) return -ENOTSUPP; From a9a22b570bd512eb1ed946042bb88f11ecccca0c Mon Sep 17 00:00:00 2001 From: Viresh Kumar Date: Thu, 14 Feb 2019 16:16:21 +0530 Subject: [PATCH 31/46] cpufreq: Replace double NOT (!!) with single NOT (!) Double NOT (!!) operation is normally done to convert a non-zero value to 1 and keep zero as is, but that isn't the requirement in this case. All we wanted was to make sure that only one of the two routines isn't set, i.e. either both function pointers are set or both are unset. This can be done with a single NOT (!) operation as well. Signed-off-by: Viresh Kumar Signed-off-by: Rafael J. Wysocki --- drivers/cpufreq/cpufreq.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/cpufreq/cpufreq.c b/drivers/cpufreq/cpufreq.c index 55e9795801a4..ad4e9991c3cc 100644 --- a/drivers/cpufreq/cpufreq.c +++ b/drivers/cpufreq/cpufreq.c @@ -2505,7 +2505,7 @@ int cpufreq_register_driver(struct cpufreq_driver *driver_data) driver_data->target) || (driver_data->setpolicy && (driver_data->target_index || driver_data->target)) || - (!!driver_data->get_intermediate != !!driver_data->target_intermediate) || + (!driver_data->get_intermediate != !driver_data->target_intermediate) || (!driver_data->online != !driver_data->offline)) return -EINVAL; From cd284ae36b6a43d478261bcb7c001db07e3a0ed5 Mon Sep 17 00:00:00 2001 From: Yangtao Li Date: Sat, 16 Feb 2019 10:55:26 -0500 Subject: [PATCH 32/46] cpufreq: pcc-cpufreq: remove unneeded semicolon The semicolon is unneeded, so remove it. Signed-off-by: Yangtao Li Signed-off-by: Viresh Kumar --- drivers/cpufreq/pcc-cpufreq.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/cpufreq/pcc-cpufreq.c b/drivers/cpufreq/pcc-cpufreq.c index 099a849396f6..1e5e64643c3a 100644 --- a/drivers/cpufreq/pcc-cpufreq.c +++ b/drivers/cpufreq/pcc-cpufreq.c @@ -268,7 +268,7 @@ static int pcc_get_offset(int cpu) if (!pccp || pccp->type != ACPI_TYPE_PACKAGE) { ret = -ENODEV; goto out_free; - }; + } offset = &(pccp->package.elements[0]); if (!offset || offset->type != ACPI_TYPE_INTEGER) { From 2814335cb3c8b364dcc740b6317bd3297e81995c Mon Sep 17 00:00:00 2001 From: Yangtao Li Date: Sat, 16 Feb 2019 10:55:27 -0500 Subject: [PATCH 33/46] cpufreq: longhaul: remove unneeded semicolon The semicolon is unneeded, so remove it. Signed-off-by: Yangtao Li Signed-off-by: Viresh Kumar --- 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 279bd9e9fa95..fb546e0d0356 100644 --- a/drivers/cpufreq/longhaul.c +++ b/drivers/cpufreq/longhaul.c @@ -851,7 +851,7 @@ static int longhaul_cpu_init(struct cpufreq_policy *policy) case TYPE_POWERSAVER: pr_cont("Powersaver supported\n"); break; - }; + } /* Doesn't hurt */ longhaul_setup_southbridge(); From 5ae06c237fd05a1e9acf6f35891cfd85de792521 Mon Sep 17 00:00:00 2001 From: Yangtao Li Date: Sat, 16 Feb 2019 12:06:23 -0500 Subject: [PATCH 34/46] cpufreq: powernv: fix missing check of return value in init_powernv_pstates() kmalloc() could fail, so insert a check of its return value. And if it fails, returns -ENOMEM. And remove (struct pstate_idx_revmap_data *) to fix coccinelle WARNING by the way. WARNING: casting value returned by memory allocation function to (struct pstate_idx_revmap_data *) is useless. Signed-off-by: Yangtao Li Signed-off-by: Viresh Kumar --- drivers/cpufreq/powernv-cpufreq.c | 10 +++++++--- 1 file changed, 7 insertions(+), 3 deletions(-) diff --git a/drivers/cpufreq/powernv-cpufreq.c b/drivers/cpufreq/powernv-cpufreq.c index 7e7ad3879c4e..d2230812fa4b 100644 --- a/drivers/cpufreq/powernv-cpufreq.c +++ b/drivers/cpufreq/powernv-cpufreq.c @@ -244,6 +244,7 @@ static int init_powernv_pstates(void) u32 len_ids, len_freqs; u32 pstate_min, pstate_max, pstate_nominal; u32 pstate_turbo, pstate_ultra_turbo; + int rc = -ENODEV; power_mgt = of_find_node_by_path("/ibm,opal/power-mgt"); if (!power_mgt) { @@ -327,8 +328,11 @@ next: powernv_freqs[i].frequency = freq * 1000; /* kHz */ powernv_freqs[i].driver_data = id & 0xFF; - revmap_data = (struct pstate_idx_revmap_data *) - kmalloc(sizeof(*revmap_data), GFP_KERNEL); + revmap_data = kmalloc(sizeof(*revmap_data), GFP_KERNEL); + if (!revmap_data) { + rc = -ENOMEM; + goto out; + } revmap_data->pstate_id = id & 0xFF; revmap_data->cpufreq_table_idx = i; @@ -357,7 +361,7 @@ next: return 0; out: of_node_put(power_mgt); - return -ENODEV; + return rc; } /* Returns the CPU frequency corresponding to the pstate_id. */ From d6c8e086e9d98a591a6b515078cb0e05fb538b5c Mon Sep 17 00:00:00 2001 From: Yangtao Li Date: Sat, 16 Feb 2019 11:15:01 -0500 Subject: [PATCH 35/46] cpufreq: speedstep: convert BUG() to BUG_ON() To fix coccinelle WARNING. WARNING: Use BUG_ON instead of if condition followed by BUG. Signed-off-by: Yangtao Li Signed-off-by: Viresh Kumar --- drivers/cpufreq/speedstep-ich.c | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-) diff --git a/drivers/cpufreq/speedstep-ich.c b/drivers/cpufreq/speedstep-ich.c index fbbcb88db061..5d8a09b82efb 100644 --- a/drivers/cpufreq/speedstep-ich.c +++ b/drivers/cpufreq/speedstep-ich.c @@ -243,8 +243,7 @@ static unsigned int speedstep_get(unsigned int cpu) unsigned int speed; /* You're supposed to ensure CPU is online. */ - if (smp_call_function_single(cpu, get_freq_data, &speed, 1) != 0) - BUG(); + BUG_ON(smp_call_function_single(cpu, get_freq_data, &speed, 1)); pr_debug("detected %u kHz as current frequency\n", speed); return speed; From 40b46b3b2f098e3740f65024099a7c55ff4b9866 Mon Sep 17 00:00:00 2001 From: Bartosz Golaszewski Date: Thu, 14 Feb 2019 18:05:05 +0100 Subject: [PATCH 36/46] cpufreq: davinci: move configuration to include/linux/platform_data The header containing the configuration structure for davinci cpufreq driver lives in mach-davinci/include/mach/. This is fine for now but if we want to make davinci part of the multi_v5 build, no code external to mach-davinci should include machine-specific headers. Move the configuration structure to include/linux/platform_data. While we're at it: convert the GPL-2.0 boilerplate to a proper SPDX license identifier. Signed-off-by: Bartosz Golaszewski Acked-by: Sekhar Nori Signed-off-by: Viresh Kumar --- arch/arm/mach-davinci/da850.c | 2 +- arch/arm/mach-davinci/include/mach/cpufreq.h | 26 ------------------- drivers/cpufreq/davinci-cpufreq.c | 5 +--- include/linux/platform_data/davinci-cpufreq.h | 19 ++++++++++++++ 4 files changed, 21 insertions(+), 31 deletions(-) delete mode 100644 arch/arm/mach-davinci/include/mach/cpufreq.h create mode 100644 include/linux/platform_data/davinci-cpufreq.h diff --git a/arch/arm/mach-davinci/da850.c b/arch/arm/mach-davinci/da850.c index e7b78df2bfef..a02ff431ba47 100644 --- a/arch/arm/mach-davinci/da850.c +++ b/arch/arm/mach-davinci/da850.c @@ -21,6 +21,7 @@ #include #include #include +#include #include #include #include @@ -29,7 +30,6 @@ #include #include -#include #include #include #include diff --git a/arch/arm/mach-davinci/include/mach/cpufreq.h b/arch/arm/mach-davinci/include/mach/cpufreq.h deleted file mode 100644 index 3c089cfb6cd6..000000000000 --- a/arch/arm/mach-davinci/include/mach/cpufreq.h +++ /dev/null @@ -1,26 +0,0 @@ -/* - * TI DaVinci CPUFreq platform support. - * - * Copyright (C) 2009 Texas Instruments, Inc. http://www.ti.com/ - * - * This program is free software; you can redistribute it and/or - * modify it under the terms of the GNU General Public License as - * published by the Free Software Foundation version 2. - * - * This program is distributed "as is" WITHOUT ANY WARRANTY of any - * kind, whether express or implied; without even the implied warranty - * of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the - * GNU General Public License for more details. - */ -#ifndef _MACH_DAVINCI_CPUFREQ_H -#define _MACH_DAVINCI_CPUFREQ_H - -#include - -struct davinci_cpufreq_config { - struct cpufreq_frequency_table *freq_table; - int (*set_voltage) (unsigned int index); - int (*init) (void); -}; - -#endif diff --git a/drivers/cpufreq/davinci-cpufreq.c b/drivers/cpufreq/davinci-cpufreq.c index d54a27c99121..940fe85db97a 100644 --- a/drivers/cpufreq/davinci-cpufreq.c +++ b/drivers/cpufreq/davinci-cpufreq.c @@ -23,13 +23,10 @@ #include #include #include +#include #include #include -#include -#include -#include - struct davinci_cpufreq { struct device *dev; struct clk *armclk; diff --git a/include/linux/platform_data/davinci-cpufreq.h b/include/linux/platform_data/davinci-cpufreq.h new file mode 100644 index 000000000000..3fbf9f2793b5 --- /dev/null +++ b/include/linux/platform_data/davinci-cpufreq.h @@ -0,0 +1,19 @@ +/* SPDX-License-Identifier: GPL-2.0 */ +/* + * TI DaVinci CPUFreq platform support. + * + * Copyright (C) 2009 Texas Instruments, Inc. http://www.ti.com/ + */ + +#ifndef _MACH_DAVINCI_CPUFREQ_H +#define _MACH_DAVINCI_CPUFREQ_H + +#include + +struct davinci_cpufreq_config { + struct cpufreq_frequency_table *freq_table; + int (*set_voltage)(unsigned int index); + int (*init)(void); +}; + +#endif /* _MACH_DAVINCI_CPUFREQ_H */ From 1757d05f3112acc5c0cdbcccad3afdee99655bf9 Mon Sep 17 00:00:00 2001 From: Xiongfeng Wang Date: Sun, 17 Feb 2019 11:54:14 +0800 Subject: [PATCH 37/46] ACPI / CPPC: Add a helper to get desired performance This patch add a helper to get the value of desired performance register. Signed-off-by: Xiongfeng Wang [ rjw: More white space ] Signed-off-by: Rafael J. Wysocki --- drivers/acpi/cppc_acpi.c | 42 ++++++++++++++++++++++++++++++++++++++++ include/acpi/cppc_acpi.h | 1 + 2 files changed, 43 insertions(+) diff --git a/drivers/acpi/cppc_acpi.c b/drivers/acpi/cppc_acpi.c index 217a782c3e55..1b207fca1420 100644 --- a/drivers/acpi/cppc_acpi.c +++ b/drivers/acpi/cppc_acpi.c @@ -1050,6 +1050,48 @@ static int cpc_write(int cpu, struct cpc_register_resource *reg_res, u64 val) return ret_val; } +/** + * cppc_get_desired_perf - Get the value of desired performance register. + * @cpunum: CPU from which to get desired performance. + * @desired_perf: address of a variable to store the returned desired performance + * + * Return: 0 for success, -EIO otherwise. + */ +int cppc_get_desired_perf(int cpunum, u64 *desired_perf) +{ + struct cpc_desc *cpc_desc = per_cpu(cpc_desc_ptr, cpunum); + int pcc_ss_id = per_cpu(cpu_pcc_subspace_idx, cpunum); + struct cpc_register_resource *desired_reg; + struct cppc_pcc_data *pcc_ss_data = NULL; + + desired_reg = &cpc_desc->cpc_regs[DESIRED_PERF]; + + if (CPC_IN_PCC(desired_reg)) { + int ret = 0; + + if (pcc_ss_id < 0) + return -EIO; + + pcc_ss_data = pcc_data[pcc_ss_id]; + + down_write(&pcc_ss_data->pcc_lock); + + if (send_pcc_cmd(pcc_ss_id, CMD_READ) >= 0) + cpc_read(cpunum, desired_reg, desired_perf); + else + ret = -EIO; + + up_write(&pcc_ss_data->pcc_lock); + + return ret; + } + + cpc_read(cpunum, desired_reg, desired_perf); + + return 0; +} +EXPORT_SYMBOL_GPL(cppc_get_desired_perf); + /** * cppc_get_perf_caps - Get a CPUs performance capabilities. * @cpunum: CPU from which to get capabilities info. diff --git a/include/acpi/cppc_acpi.h b/include/acpi/cppc_acpi.h index 4f34734e7f36..ba6fd7202775 100644 --- a/include/acpi/cppc_acpi.h +++ b/include/acpi/cppc_acpi.h @@ -137,6 +137,7 @@ struct cppc_cpudata { cpumask_var_t shared_cpu_map; }; +extern int cppc_get_desired_perf(int cpunum, u64 *desired_perf); extern int cppc_get_perf_ctrs(int cpu, struct cppc_perf_fb_ctrs *perf_fb_ctrs); extern int cppc_set_perf(int cpu, struct cppc_perf_ctrls *perf_ctrls); extern int cppc_get_perf_caps(int cpu, struct cppc_perf_caps *caps); From 6c8d750f9784cef32a8cffdad74c8a351b4ca3a6 Mon Sep 17 00:00:00 2001 From: Xiongfeng Wang Date: Sun, 17 Feb 2019 11:54:15 +0800 Subject: [PATCH 38/46] cpufreq / cppc: Work around for Hisilicon CPPC cpufreq Hisilicon chips do not support delivered performance counter register and reference performance counter register. But the platform can calculate the real performance using its own method. We reuse the desired performance register to store the real performance calculated by the platform. After the platform finished the frequency adjust, it gets the real performance and writes it into desired performance register. Os can use it to calculate the real frequency. Signed-off-by: Xiongfeng Wang [ rjw: Drop unnecessary braces ] Signed-off-by: Rafael J. Wysocki --- drivers/cpufreq/cppc_cpufreq.c | 65 ++++++++++++++++++++++++++++++++++ 1 file changed, 65 insertions(+) diff --git a/drivers/cpufreq/cppc_cpufreq.c b/drivers/cpufreq/cppc_cpufreq.c index fd25c21cee72..2ae978d27e61 100644 --- a/drivers/cpufreq/cppc_cpufreq.c +++ b/drivers/cpufreq/cppc_cpufreq.c @@ -42,6 +42,66 @@ */ static struct cppc_cpudata **all_cpu_data; +struct cppc_workaround_oem_info { + char oem_id[ACPI_OEM_ID_SIZE +1]; + char oem_table_id[ACPI_OEM_TABLE_ID_SIZE + 1]; + u32 oem_revision; +}; + +static bool apply_hisi_workaround; + +static struct cppc_workaround_oem_info wa_info[] = { + { + .oem_id = "HISI ", + .oem_table_id = "HIP07 ", + .oem_revision = 0, + }, { + .oem_id = "HISI ", + .oem_table_id = "HIP08 ", + .oem_revision = 0, + } +}; + +static unsigned int cppc_cpufreq_perf_to_khz(struct cppc_cpudata *cpu, + unsigned int perf); + +/* + * HISI platform does not support delivered performance counter and + * reference performance counter. It can calculate the performance using the + * platform specific mechanism. We reuse the desired performance register to + * store the real performance calculated by the platform. + */ +static unsigned int hisi_cppc_cpufreq_get_rate(unsigned int cpunum) +{ + struct cppc_cpudata *cpudata = all_cpu_data[cpunum]; + u64 desired_perf; + int ret; + + ret = cppc_get_desired_perf(cpunum, &desired_perf); + if (ret < 0) + return -EIO; + + return cppc_cpufreq_perf_to_khz(cpudata, desired_perf); +} + +static void cppc_check_hisi_workaround(void) +{ + struct acpi_table_header *tbl; + acpi_status status = AE_OK; + int i; + + status = acpi_get_table(ACPI_SIG_PCCT, 0, &tbl); + if (ACPI_FAILURE(status) || !tbl) + return; + + for (i = 0; i < ARRAY_SIZE(wa_info); i++) { + if (!memcmp(wa_info[i].oem_id, tbl->oem_id, ACPI_OEM_ID_SIZE) && + !memcmp(wa_info[i].oem_table_id, tbl->oem_table_id, ACPI_OEM_TABLE_ID_SIZE) && + wa_info[i].oem_revision == tbl->oem_revision) + apply_hisi_workaround = true; + } +} + /* Callback function used to retrieve the max frequency from DMI */ static void cppc_find_dmi_mhz(const struct dmi_header *dm, void *private) { @@ -334,6 +394,9 @@ static unsigned int cppc_cpufreq_get_rate(unsigned int cpunum) struct cppc_cpudata *cpu = all_cpu_data[cpunum]; int ret; + if (apply_hisi_workaround) + return hisi_cppc_cpufreq_get_rate(cpunum); + ret = cppc_get_perf_ctrs(cpunum, &fb_ctrs_t0); if (ret) return ret; @@ -386,6 +449,8 @@ static int __init cppc_cpufreq_init(void) goto out; } + cppc_check_hisi_workaround(); + ret = cpufreq_register_driver(&cppc_cpufreq_driver); if (ret) goto out; From fa93b51c55099660c2583ec918ef27df18fc511a Mon Sep 17 00:00:00 2001 From: "Rafael J. Wysocki" Date: Fri, 15 Feb 2019 13:15:32 +0100 Subject: [PATCH 39/46] cpufreq: intel_pstate: Avoid redundant initialization of local vars After commit 1a4fe38add8b ("cpufreq: intel_pstate: Remove max/min fractions to limit performance") the initial value of the pstate local variable in intel_pstate_max_within_limits() and the initial value of the max_pstate local variable in intel_pstate_prepare_request() are both immediately discarded, so initialize both these variables to their target values upfront. No intentional changes of behavior. Signed-off-by: Rafael J. Wysocki --- drivers/cpufreq/intel_pstate.c | 10 +++------- 1 file changed, 3 insertions(+), 7 deletions(-) diff --git a/drivers/cpufreq/intel_pstate.c b/drivers/cpufreq/intel_pstate.c index 75a5408e6981..3a4d15a48c1f 100644 --- a/drivers/cpufreq/intel_pstate.c +++ b/drivers/cpufreq/intel_pstate.c @@ -1471,11 +1471,9 @@ static void intel_pstate_set_min_pstate(struct cpudata *cpu) static void intel_pstate_max_within_limits(struct cpudata *cpu) { - int pstate; + int pstate = max(cpu->pstate.min_pstate, cpu->max_perf_ratio); update_turbo_state(); - pstate = intel_pstate_get_base_pstate(cpu); - pstate = max(cpu->pstate.min_pstate, cpu->max_perf_ratio); intel_pstate_set_pstate(cpu, pstate); } @@ -1716,11 +1714,9 @@ static inline int32_t get_target_pstate(struct cpudata *cpu) static int intel_pstate_prepare_request(struct cpudata *cpu, int pstate) { - int max_pstate = intel_pstate_get_base_pstate(cpu); - int min_pstate; + int min_pstate = max(cpu->pstate.min_pstate, cpu->min_perf_ratio); + int max_pstate = max(min_pstate, cpu->max_perf_ratio); - min_pstate = max(cpu->pstate.min_pstate, cpu->min_perf_ratio); - max_pstate = max(min_pstate, cpu->max_perf_ratio); return clamp_t(int, pstate, min_pstate, max_pstate); } From a8e1942d97dcc44d1425807c71a4252f9e3b53b6 Mon Sep 17 00:00:00 2001 From: "Rafael J. Wysocki" Date: Fri, 15 Feb 2019 13:16:40 +0100 Subject: [PATCH 40/46] cpufreq: intel_pstate: Eliminate intel_pstate_get_base_pstate() There is only one caller of intel_pstate_get_base_pstate() and it is more straightforward to carry out the computation directly in the caller, so do that and drop intel_pstate_get_base_pstate(). No intentional changes of behavior. Signed-off-by: Rafael J. Wysocki --- drivers/cpufreq/intel_pstate.c | 9 ++------- 1 file changed, 2 insertions(+), 7 deletions(-) diff --git a/drivers/cpufreq/intel_pstate.c b/drivers/cpufreq/intel_pstate.c index 3a4d15a48c1f..9a6cb3a1be50 100644 --- a/drivers/cpufreq/intel_pstate.c +++ b/drivers/cpufreq/intel_pstate.c @@ -1445,12 +1445,6 @@ static int knl_get_turbo_pstate(void) return ret; } -static int intel_pstate_get_base_pstate(struct cpudata *cpu) -{ - return global.no_turbo || global.turbo_disabled ? - cpu->pstate.max_pstate : cpu->pstate.turbo_pstate; -} - static void intel_pstate_set_pstate(struct cpudata *cpu, int pstate) { trace_cpu_frequency(pstate * cpu->pstate.scaling, cpu->cpu); @@ -1973,7 +1967,8 @@ static void intel_pstate_update_perf_limits(struct cpufreq_policy *policy, if (hwp_active) { intel_pstate_get_hwp_max(cpu->cpu, &turbo_max, &max_state); } else { - max_state = intel_pstate_get_base_pstate(cpu); + max_state = global.no_turbo || global.turbo_disabled ? + cpu->pstate.max_pstate : cpu->pstate.turbo_pstate; turbo_max = cpu->pstate.turbo_pstate; } From b8bd1581aa6110eb234c0d424eccd3f32d7317e6 Mon Sep 17 00:00:00 2001 From: "Rafael J. Wysocki" Date: Thu, 7 Feb 2019 12:51:04 +0100 Subject: [PATCH 41/46] cpufreq: intel_pstate: Rework iowait boosting to be less aggressive The current iowait boosting mechanism in intel_pstate_update_util() is quite aggressive, as it goes to the maximum P-state right away, and may cause excessive amounts of energy to be used, which is not desirable and arguably isn't necessary too. Follow commit a5a0809bc58e ("cpufreq: schedutil: Make iowait boost more energy efficient") that reworked the analogous iowait boost mechanism in the schedutil governor and make the iowait boosting in intel_pstate_update_util() work along the same lines. Signed-off-by: Rafael J. Wysocki --- drivers/cpufreq/intel_pstate.c | 36 +++++++++++++++++----------------- 1 file changed, 18 insertions(+), 18 deletions(-) diff --git a/drivers/cpufreq/intel_pstate.c b/drivers/cpufreq/intel_pstate.c index 9a6cb3a1be50..002f5169d4eb 100644 --- a/drivers/cpufreq/intel_pstate.c +++ b/drivers/cpufreq/intel_pstate.c @@ -50,6 +50,8 @@ #define int_tofp(X) ((int64_t)(X) << FRAC_BITS) #define fp_toint(X) ((X) >> FRAC_BITS) +#define ONE_EIGHTH_FP ((int64_t)1 << (FRAC_BITS - 3)) + #define EXT_BITS 6 #define EXT_FRAC_BITS (EXT_BITS + FRAC_BITS) #define fp_ext_toint(X) ((X) >> EXT_FRAC_BITS) @@ -1671,17 +1673,14 @@ static inline int32_t get_avg_pstate(struct cpudata *cpu) static inline int32_t get_target_pstate(struct cpudata *cpu) { struct sample *sample = &cpu->sample; - int32_t busy_frac, boost; + int32_t busy_frac; int target, avg_pstate; busy_frac = div_fp(sample->mperf << cpu->aperf_mperf_shift, sample->tsc); - boost = cpu->iowait_boost; - cpu->iowait_boost >>= 1; - - if (busy_frac < boost) - busy_frac = boost; + if (busy_frac < cpu->iowait_boost) + busy_frac = cpu->iowait_boost; sample->busy_scaled = busy_frac * 100; @@ -1758,29 +1757,30 @@ static void intel_pstate_update_util(struct update_util_data *data, u64 time, if (smp_processor_id() != cpu->cpu) return; + delta_ns = time - cpu->last_update; if (flags & SCHED_CPUFREQ_IOWAIT) { - cpu->iowait_boost = int_tofp(1); - cpu->last_update = time; - /* - * The last time the busy was 100% so P-state was max anyway - * so avoid overhead of computation. - */ - if (fp_toint(cpu->sample.busy_scaled) == 100) - return; - - goto set_pstate; + /* Start over if the CPU may have been idle. */ + if (delta_ns > TICK_NSEC) { + cpu->iowait_boost = ONE_EIGHTH_FP; + } else if (cpu->iowait_boost) { + cpu->iowait_boost <<= 1; + if (cpu->iowait_boost > int_tofp(1)) + cpu->iowait_boost = int_tofp(1); + } else { + cpu->iowait_boost = ONE_EIGHTH_FP; + } } else if (cpu->iowait_boost) { /* Clear iowait_boost if the CPU may have been idle. */ - delta_ns = time - cpu->last_update; if (delta_ns > TICK_NSEC) cpu->iowait_boost = 0; + else + cpu->iowait_boost >>= 1; } cpu->last_update = time; delta_ns = time - cpu->sample.time; if ((s64)delta_ns < INTEL_PSTATE_SAMPLING_INTERVAL) return; -set_pstate: if (intel_pstate_sample(cpu, time)) intel_pstate_adjust_pstate(cpu); } From a0dbb819b84f87b3cb35083e264e9e584630e1f7 Mon Sep 17 00:00:00 2001 From: "Rafael J. Wysocki" Date: Wed, 20 Feb 2019 00:22:51 +0100 Subject: [PATCH 42/46] cpufreq: Add kerneldoc comments for two core functions Add kerneldoc comments describing cpufreq_set_policy() and cpufreq_update_policy() as they have not been properly documented so far and they really need to be documented. While at it, fix white space around the cpufreq_set_policy() header. Signed-off-by: Rafael J. Wysocki Acked-by: Viresh Kumar --- drivers/cpufreq/cpufreq.c | 32 ++++++++++++++++++++++++-------- 1 file changed, 24 insertions(+), 8 deletions(-) diff --git a/drivers/cpufreq/cpufreq.c b/drivers/cpufreq/cpufreq.c index ad4e9991c3cc..e4a7dbee84fc 100644 --- a/drivers/cpufreq/cpufreq.c +++ b/drivers/cpufreq/cpufreq.c @@ -2218,12 +2218,25 @@ int cpufreq_get_policy(struct cpufreq_policy *policy, unsigned int cpu) } EXPORT_SYMBOL(cpufreq_get_policy); -/* - * policy : current policy. - * new_policy: policy to be set. +/** + * cpufreq_set_policy - Modify cpufreq policy parameters. + * @policy: Policy object to modify. + * @new_policy: New policy data. + * + * Pass @new_policy to the cpufreq driver's ->verify() callback, run the + * installed policy notifiers for it with the CPUFREQ_ADJUST value, pass it to + * the driver's ->verify() callback again and run the notifiers for it again + * with the CPUFREQ_NOTIFY value. Next, copy the min and max parameters + * of @new_policy to @policy and either invoke the driver's ->setpolicy() + * callback (if present) or carry out a governor update for @policy. That is, + * run the current governor's ->limits() callback (if the governor field in + * @new_policy points to the same object as the one in @policy) or replace the + * governor for @policy with the new one stored in @new_policy. + * + * The cpuinfo part of @policy is not updated by this function. */ static int cpufreq_set_policy(struct cpufreq_policy *policy, - struct cpufreq_policy *new_policy) + struct cpufreq_policy *new_policy) { struct cpufreq_governor *old_gov; int ret; @@ -2319,11 +2332,14 @@ static int cpufreq_set_policy(struct cpufreq_policy *policy, } /** - * cpufreq_update_policy - re-evaluate an existing cpufreq policy - * @cpu: CPU which shall be re-evaluated + * cpufreq_update_policy - Re-evaluate an existing cpufreq policy. + * @cpu: CPU to re-evaluate the policy for. * - * Useful for policy notifiers which have different necessities - * at different times. + * Update the current frequency for the cpufreq policy of @cpu and use + * cpufreq_set_policy() to re-apply the min and max limits saved in the + * user_policy sub-structure of that policy, which triggers the evaluation + * of policy notifiers and the cpufreq driver's ->verify() callback for the + * policy in question, among other things. */ void cpufreq_update_policy(unsigned int cpu) { From 348a2ec5f5a5af15509252dfc6a30fa190ac0203 Mon Sep 17 00:00:00 2001 From: "Rafael J. Wysocki" Date: Wed, 20 Feb 2019 00:24:25 +0100 Subject: [PATCH 43/46] cpufreq: Reorder and simplify cpufreq_update_policy() In cpufreq_update_policy(), instead of updating new_policy.cur separately, which is kind of confusing, because cpufreq_set_policy() doesn't take that value into account directly anyway, make the copy of the existing policy after calling cpufreq_update_current_freq(). No intentional changes of behavior. Signed-off-by: Rafael J. Wysocki Acked-by: Viresh Kumar --- drivers/cpufreq/cpufreq.c | 19 +++++++------------ 1 file changed, 7 insertions(+), 12 deletions(-) diff --git a/drivers/cpufreq/cpufreq.c b/drivers/cpufreq/cpufreq.c index e4a7dbee84fc..fd5221007ade 100644 --- a/drivers/cpufreq/cpufreq.c +++ b/drivers/cpufreq/cpufreq.c @@ -2354,23 +2354,18 @@ void cpufreq_update_policy(unsigned int cpu) if (policy_is_inactive(policy)) goto unlock; - pr_debug("updating policy for CPU %u\n", cpu); - memcpy(&new_policy, policy, sizeof(*policy)); - new_policy.min = policy->user_policy.min; - new_policy.max = policy->user_policy.max; - /* * BIOS might change freq behind our back * -> ask driver for current freq and notify governors about a change */ - if (cpufreq_driver->get && !cpufreq_driver->setpolicy) { - if (cpufreq_suspended) - goto unlock; + if (cpufreq_driver->get && !cpufreq_driver->setpolicy && + (cpufreq_suspended || WARN_ON(!cpufreq_update_current_freq(policy)))) + goto unlock; - new_policy.cur = cpufreq_update_current_freq(policy); - if (WARN_ON(!new_policy.cur)) - goto unlock; - } + pr_debug("updating policy for CPU %u\n", cpu); + memcpy(&new_policy, policy, sizeof(*policy)); + new_policy.min = policy->user_policy.min; + new_policy.max = policy->user_policy.max; cpufreq_set_policy(policy, &new_policy); From 2bb4059e075dcb8d5a2f8689bb661aa76c487ab0 Mon Sep 17 00:00:00 2001 From: "Rafael J. Wysocki" Date: Wed, 20 Feb 2019 00:25:18 +0100 Subject: [PATCH 44/46] cpufreq: Fix two debug messages in cpufreq_set_policy() Remove the redundant "cpufreq:" prefix from two debug messages in cpufreq_set_policy(). Signed-off-by: Rafael J. Wysocki Acked-by: Viresh Kumar --- drivers/cpufreq/cpufreq.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/drivers/cpufreq/cpufreq.c b/drivers/cpufreq/cpufreq.c index fd5221007ade..06b1a954d6e4 100644 --- a/drivers/cpufreq/cpufreq.c +++ b/drivers/cpufreq/cpufreq.c @@ -2290,7 +2290,7 @@ static int cpufreq_set_policy(struct cpufreq_policy *policy, } if (new_policy->governor == policy->governor) { - pr_debug("cpufreq: governor limits update\n"); + pr_debug("governor limits update\n"); cpufreq_governor_limits(policy); return 0; } @@ -2311,7 +2311,7 @@ static int cpufreq_set_policy(struct cpufreq_policy *policy, if (!ret) { ret = cpufreq_start_governor(policy); if (!ret) { - pr_debug("cpufreq: governor change\n"); + pr_debug("governor change\n"); sched_cpufreq_governor_change(policy, old_gov); return 0; } From 167a38dcd5caaa85994ac8b7d2d1c20a71cd947b Mon Sep 17 00:00:00 2001 From: "Rafael J. Wysocki" Date: Wed, 20 Feb 2019 00:26:30 +0100 Subject: [PATCH 45/46] cpufreq: Pass updated policy to driver ->setpolicy() callback The invocation of the ->setpolicy() cpufreq driver callback should be equivalent to calling cpufreq_governor_limits(policy) for drivers with internal governors, but in fact it isn't so, because the temporary new_policy object is passed to it instead of the updated policy. That is a bit confusing, so make cpufreq_set_policy() pass the updated policy to the driver ->setpolicy() callback. No intentional changes of behavior. Signed-off-by: Rafael J. Wysocki Acked-by: Viresh Kumar --- drivers/cpufreq/cpufreq.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/cpufreq/cpufreq.c b/drivers/cpufreq/cpufreq.c index 06b1a954d6e4..0e626b00053b 100644 --- a/drivers/cpufreq/cpufreq.c +++ b/drivers/cpufreq/cpufreq.c @@ -2286,7 +2286,7 @@ static int cpufreq_set_policy(struct cpufreq_policy *policy, if (cpufreq_driver->setpolicy) { policy->policy = new_policy->policy; pr_debug("setting range\n"); - return cpufreq_driver->setpolicy(new_policy); + return cpufreq_driver->setpolicy(policy); } if (new_policy->governor == policy->governor) { From 1222d527f314c86a3b59a522115d62facc5a7965 Mon Sep 17 00:00:00 2001 From: Erwan Velu Date: Wed, 20 Feb 2019 11:10:17 +0100 Subject: [PATCH 46/46] cpufreq: acpi-cpufreq: Report if CPU doesn't support boost technologies There is some rare cases where CPB (and possibly IDA) are missing on processors. This is the case fixed by commit f7f3dc00f612 ("x86/cpu/AMD: Fix erratum 1076 (CPB bit)") and following. In such context, the boost status isn't reported by /sys/devices/system/cpu/cpufreq/boost. This commit is about printing a message to report that the CPU doesn't expose the boost capabilities. This message could help debugging platforms hit by this phenomena. Signed-off-by: Erwan Velu [ rjw: Change the message text somewhat ] Signed-off-by: Rafael J. Wysocki --- drivers/cpufreq/acpi-cpufreq.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) diff --git a/drivers/cpufreq/acpi-cpufreq.c b/drivers/cpufreq/acpi-cpufreq.c index d62fd374d5c7..c72258a44ba4 100644 --- a/drivers/cpufreq/acpi-cpufreq.c +++ b/drivers/cpufreq/acpi-cpufreq.c @@ -916,8 +916,10 @@ static void __init acpi_cpufreq_boost_init(void) { int ret; - if (!(boot_cpu_has(X86_FEATURE_CPB) || boot_cpu_has(X86_FEATURE_IDA))) + if (!(boot_cpu_has(X86_FEATURE_CPB) || boot_cpu_has(X86_FEATURE_IDA))) { + pr_debug("Boost capabilities not present in the processor\n"); return; + } acpi_cpufreq_driver.set_boost = set_boost; acpi_cpufreq_driver.boost_enabled = boost_state(0);