diff options
| -rw-r--r-- | Documentation/power/index.rst | 1 | ||||
| -rw-r--r-- | Documentation/power/powercap/dtpm.rst | 212 | ||||
| -rw-r--r-- | drivers/powercap/Kconfig | 13 | ||||
| -rw-r--r-- | drivers/powercap/Makefile | 2 | ||||
| -rw-r--r-- | drivers/powercap/dtpm.c | 480 | ||||
| -rw-r--r-- | drivers/powercap/dtpm_cpu.c | 257 | ||||
| -rw-r--r-- | drivers/powercap/intel_rapl_common.c | 9 | ||||
| -rw-r--r-- | include/asm-generic/vmlinux.lds.h | 11 | ||||
| -rw-r--r-- | include/linux/cpuhotplug.h | 1 | ||||
| -rw-r--r-- | include/linux/dtpm.h | 77 | ||||
| -rw-r--r-- | include/linux/units.h | 4 | ||||
| -rw-r--r-- | kernel/power/Kconfig | 12 |
12 files changed, 1067 insertions, 12 deletions
diff --git a/Documentation/power/index.rst b/Documentation/power/index.rst index ced8a8007434..a0f5244fb427 100644 --- a/Documentation/power/index.rst +++ b/Documentation/power/index.rst @@ -30,6 +30,7 @@ Power Management userland-swsusp powercap/powercap + powercap/dtpm regulator/consumer regulator/design diff --git a/Documentation/power/powercap/dtpm.rst b/Documentation/power/powercap/dtpm.rst new file mode 100644 index 000000000000..a38dee3d815b --- /dev/null +++ b/Documentation/power/powercap/dtpm.rst @@ -0,0 +1,212 @@ +.. SPDX-License-Identifier: GPL-2.0 + +========================================== +Dynamic Thermal Power Management framework +========================================== + +On the embedded world, the complexity of the SoC leads to an +increasing number of hotspots which need to be monitored and mitigated +as a whole in order to prevent the temperature to go above the +normative and legally stated 'skin temperature'. + +Another aspect is to sustain the performance for a given power budget, +for example virtual reality where the user can feel dizziness if the +performance is capped while a big CPU is processing something else. Or +reduce the battery charging because the dissipated power is too high +compared with the power consumed by other devices. + +The user space is the most adequate place to dynamically act on the +different devices by limiting their power given an application +profile: it has the knowledge of the platform. + +The Dynamic Thermal Power Management (DTPM) is a technique acting on +the device power by limiting and/or balancing a power budget among +different devices. + +The DTPM framework provides an unified interface to act on the +device power. + +Overview +======== + +The DTPM framework relies on the powercap framework to create the +powercap entries in the sysfs directory and implement the backend +driver to do the connection with the power manageable device. + +The DTPM is a tree representation describing the power constraints +shared between devices, not their physical positions. + +The nodes of the tree are a virtual description aggregating the power +characteristics of the children nodes and their power limitations. + +The leaves of the tree are the real power manageable devices. + +For instance:: + + SoC + | + `-- pkg + | + |-- pd0 (cpu0-3) + | + `-- pd1 (cpu4-5) + +The pkg power will be the sum of pd0 and pd1 power numbers:: + + SoC (400mW - 3100mW) + | + `-- pkg (400mW - 3100mW) + | + |-- pd0 (100mW - 700mW) + | + `-- pd1 (300mW - 2400mW) + +When the nodes are inserted in the tree, their power characteristics are propagated to the parents:: + + SoC (600mW - 5900mW) + | + |-- pkg (400mW - 3100mW) + | | + | |-- pd0 (100mW - 700mW) + | | + | `-- pd1 (300mW - 2400mW) + | + `-- pd2 (200mW - 2800mW) + +Each node have a weight on a 2^10 basis reflecting the percentage of power consumption along the siblings:: + + SoC (w=1024) + | + |-- pkg (w=538) + | | + | |-- pd0 (w=231) + | | + | `-- pd1 (w=794) + | + `-- pd2 (w=486) + + Note the sum of weights at the same level are equal to 1024. + +When a power limitation is applied to a node, then it is distributed along the children given their weights. For example, if we set a power limitation of 3200mW at the 'SoC' root node, the resulting tree will be:: + + SoC (w=1024) <--- power_limit = 3200mW + | + |-- pkg (w=538) --> power_limit = 1681mW + | | + | |-- pd0 (w=231) --> power_limit = 378mW + | | + | `-- pd1 (w=794) --> power_limit = 1303mW + | + `-- pd2 (w=486) --> power_limit = 1519mW + + +Flat description +---------------- + +A root node is created and it is the parent of all the nodes. This +description is the simplest one and it is supposed to give to user +space a flat representation of all the devices supporting the power +limitation without any power limitation distribution. + +Hierarchical description +------------------------ + +The different devices supporting the power limitation are represented +hierarchically. There is one root node, all intermediate nodes are +grouping the child nodes which can be intermediate nodes also or real +devices. + +The intermediate nodes aggregate the power information and allows to +set the power limit given the weight of the nodes. + +User space API +============== + +As stated in the overview, the DTPM framework is built on top of the +powercap framework. Thus the sysfs interface is the same, please refer +to the powercap documentation for further details. + + * power_uw: Instantaneous power consumption. If the node is an + intermediate node, then the power consumption will be the sum of all + children power consumption. + + * max_power_range_uw: The power range resulting of the maximum power + minus the minimum power. + + * name: The name of the node. This is implementation dependent. Even + if it is not recommended for the user space, several nodes can have + the same name. + + * constraint_X_name: The name of the constraint. + + * constraint_X_max_power_uw: The maximum power limit to be applicable + to the node. + + * constraint_X_power_limit_uw: The power limit to be applied to the + node. If the value contained in constraint_X_max_power_uw is set, + the constraint will be removed. + + * constraint_X_time_window_us: The meaning of this file will depend + on the constraint number. + +Constraints +----------- + + * Constraint 0: The power limitation is immediately applied, without + limitation in time. + +Kernel API +========== + +Overview +-------- + +The DTPM framework has no power limiting backend support. It is +generic and provides a set of API to let the different drivers to +implement the backend part for the power limitation and create the +power constraints tree. + +It is up to the platform to provide the initialization function to +allocate and link the different nodes of the tree. + +A special macro has the role of declaring a node and the corresponding +initialization function via a description structure. This one contains +an optional parent field allowing to hook different devices to an +already existing tree at boot time. + +For instance:: + + struct dtpm_descr my_descr = { + .name = "my_name", + .init = my_init_func, + }; + + DTPM_DECLARE(my_descr); + +The nodes of the DTPM tree are described with dtpm structure. The +steps to add a new power limitable device is done in three steps: + + * Allocate the dtpm node + * Set the power number of the dtpm node + * Register the dtpm node + +The registration of the dtpm node is done with the powercap +ops. Basically, it must implements the callbacks to get and set the +power and the limit. + +Alternatively, if the node to be inserted is an intermediate one, then +a simple function to insert it as a future parent is available. + +If a device has its power characteristics changing, then the tree must +be updated with the new power numbers and weights. + +Nomenclature +------------ + + * dtpm_alloc() : Allocate and initialize a dtpm structure + + * dtpm_register() : Add the dtpm node to the tree + + * dtpm_unregister() : Remove the dtpm node from the tree + + * dtpm_update_power() : Update the power characteristics of the dtpm node diff --git a/drivers/powercap/Kconfig b/drivers/powercap/Kconfig index bc228725346b..20b4325c6161 100644 --- a/drivers/powercap/Kconfig +++ b/drivers/powercap/Kconfig @@ -43,4 +43,17 @@ config IDLE_INJECT CPUs for power capping. Idle period can be injected synchronously on a set of specified CPUs or alternatively on a per CPU basis. + +config DTPM + bool "Power capping for Dynamic Thermal Power Management" + help + This enables support for the power capping for the dynamic + thermal power management userspace engine. + +config DTPM_CPU + bool "Add CPU power capping based on the energy model" + depends on DTPM && ENERGY_MODEL + help + This enables support for CPU power limitation based on + energy model. endif diff --git a/drivers/powercap/Makefile b/drivers/powercap/Makefile index 7255c94ec61c..fabcf388a8d3 100644 --- a/drivers/powercap/Makefile +++ b/drivers/powercap/Makefile @@ -1,4 +1,6 @@ # SPDX-License-Identifier: GPL-2.0-only +obj-$(CONFIG_DTPM) += dtpm.o +obj-$(CONFIG_DTPM_CPU) += dtpm_cpu.o obj-$(CONFIG_POWERCAP) += powercap_sys.o obj-$(CONFIG_INTEL_RAPL_CORE) += intel_rapl_common.o obj-$(CONFIG_INTEL_RAPL) += intel_rapl_msr.o diff --git a/drivers/powercap/dtpm.c b/drivers/powercap/dtpm.c new file mode 100644 index 000000000000..5a51cd34a7e8 --- /dev/null +++ b/drivers/powercap/dtpm.c @@ -0,0 +1,480 @@ +// SPDX-License-Identifier: GPL-2.0-only +/* + * Copyright 2020 Linaro Limited + * + * Author: Daniel Lezcano <daniel.lezcano@linaro.org> + * + * The powercap based Dynamic Thermal Power Management framework + * provides to the userspace a consistent API to set the power limit + * on some devices. + * + * DTPM defines the functions to create a tree of constraints. Each + * parent node is a virtual description of the aggregation of the + * children. It propagates the constraints set at its level to its + * children and collect the children power information. The leaves of + * the tree are the real devices which have the ability to get their + * current power consumption and set their power limit. + */ +#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt + +#include <linux/dtpm.h> +#include <linux/init.h> +#include <linux/kernel.h> +#include <linux/powercap.h> +#include <linux/slab.h> +#include <linux/mutex.h> + +#define DTPM_POWER_LIMIT_FLAG 0 + +static const char *constraint_name[] = { + "Instantaneous", +}; + +static DEFINE_MUTEX(dtpm_lock); +static struct powercap_control_type *pct; +static struct dtpm *root; + +static int get_time_window_us(struct powercap_zone *pcz, int cid, u64 *window) +{ + return -ENOSYS; +} + +static int set_time_window_us(struct powercap_zone *pcz, int cid, u64 window) +{ + return -ENOSYS; +} + +static int get_max_power_range_uw(struct powercap_zone *pcz, u64 *max_power_uw) +{ + struct dtpm *dtpm = to_dtpm(pcz); + + mutex_lock(&dtpm_lock); + *max_power_uw = dtpm->power_max - dtpm->power_min; + mutex_unlock(&dtpm_lock); + + return 0; +} + +static int __get_power_uw(struct dtpm *dtpm, u64 *power_uw) +{ + struct dtpm *child; + u64 power; + int ret = 0; + + if (dtpm->ops) { + *power_uw = dtpm->ops->get_power_uw(dtpm); + return 0; + } + + *power_uw = 0; + + list_for_each_entry(child, &dtpm->children, sibling) { + ret = __get_power_uw(child, &power); + if (ret) + break; + *power_uw += power; + } + + return ret; +} + +static int get_power_uw(struct powercap_zone *pcz, u64 *power_uw) +{ + struct dtpm *dtpm = to_dtpm(pcz); + int ret; + + mutex_lock(&dtpm_lock); + ret = __get_power_uw(dtpm, power_uw); + mutex_unlock(&dtpm_lock); + + return ret; +} + +static void __dtpm_rebalance_weight(struct dtpm *dtpm) +{ + struct dtpm *child; + + list_for_each_entry(child, &dtpm->children, sibling) { + + pr_debug("Setting weight '%d' for '%s'\n", + child->weight, child->zone.name); + + child->weight = DIV64_U64_ROUND_CLOSEST( + child->power_max * 1024, dtpm->power_max); + + __dtpm_rebalance_weight(child); + } +} + +static void __dtpm_sub_power(struct dtpm *dtpm) +{ + struct dtpm *parent = dtpm->parent; + + while (parent) { + parent->power_min -= dtpm->power_min; + parent->power_max -= dtpm->power_max; + parent->power_limit -= dtpm->power_limit; + parent = parent->parent; + } + + __dtpm_rebalance_weight(root); +} + +static void __dtpm_add_power(struct dtpm *dtpm) +{ + struct dtpm *parent = dtpm->parent; + + while (parent) { + parent->power_min += dtpm->power_min; + parent->power_max += dtpm->power_max; + parent->power_limit += dtpm->power_limit; + parent = parent->parent; + } + + __dtpm_rebalance_weight(root); +} + +/** + * dtpm_update_power - Update the power on the dtpm + * @dtpm: a pointer to a dtpm structure to update + * @power_min: a u64 representing the new power_min value + * @power_max: a u64 representing the new power_max value + * + * Function to update the power values of the dtpm node specified in + * parameter. These new values will be propagated to the tree. + * + * Return: zero on success, -EINVAL if the values are inconsistent + */ +int dtpm_update_power(struct dtpm *dtpm, u64 power_min, u64 power_max) +{ + int ret = 0; + + mutex_lock(&dtpm_lock); + + if (power_min == dtpm->power_min && power_max == dtpm->power_max) + goto unlock; + + if (power_max < power_min) { + ret = -EINVAL; + goto unlock; + } + + __dtpm_sub_power(dtpm); + + dtpm->power_min = power_min; + dtpm->power_max = power_max; + if (!test_bit(DTPM_POWER_LIMIT_FLAG, &dtpm->flags)) + dtpm->power_limit = power_max; + + __dtpm_add_power(dtpm); + +unlock: + mutex_unlock(&dtpm_lock); + + return ret; +} + +/** + * dtpm_release_zone - Cleanup when the node is released + * @pcz: a pointer to a powercap_zone structure + * + * Do some housecleaning and update the weight on the tree. The + * release will be denied if the node has children. This function must + * be called by the specific release callback of the different + * backends. + * + * Return: 0 on success, -EBUSY if there are children + */ +int dtpm_release_zone(struct powercap_zone *pcz) +{ + struct dtpm *dtpm = to_dtpm(pcz); + struct dtpm *parent = dtpm->parent; + + mutex_lock(&dtpm_lock); + + if (!list_empty(&dtpm->children)) { + mutex_unlock(&dtpm_lock); + return -EBUSY; + } + + if (parent) + list_del(&dtpm->sibling); + + __dtpm_sub_power(dtpm); + + mutex_unlock(&dtpm_lock); + + if (dtpm->ops) + dtpm->ops->release(dtpm); + + kfree(dtpm); + + return 0; +} + +static int __get_power_limit_uw(struct dtpm *dtpm, int cid, u64 *power_limit) +{ + *power_limit = dtpm->power_limit; + return 0; +} + +static int get_power_limit_uw(struct powercap_zone *pcz, + int cid, u64 *power_limit) +{ + struct dtpm *dtpm = to_dtpm(pcz); + int ret; + + mutex_lock(&dtpm_lock); + ret = __get_power_limit_uw(dtpm, cid, power_limit); + mutex_unlock(&dtpm_lock); + + return ret; +} + +/* + * Set the power limit on the nodes, the power limit is distributed + * given the weight of the children. + * + * The dtpm node lock must be held when calling this function. + */ +static int __set_power_limit_uw(struct dtpm *dtpm, int cid, u64 power_limit) +{ + struct dtpm *child; + int ret = 0; + u64 power; + + /* + * A max power limitation means we remove the power limit, + * otherwise we set a constraint and flag the dtpm node. + */ + if (power_limit == dtpm->power_max) { + clear_bit(DTPM_POWER_LIMIT_FLAG, &dtpm->flags); + } else { + set_bit(DTPM_POWER_LIMIT_FLAG, &dtpm->flags); + } + + pr_debug("Setting power limit for '%s': %llu uW\n", + dtpm->zone.name, power_limit); + + /* + * Only leaves of the dtpm tree has ops to get/set the power + */ + if (dtpm->ops) { + dtpm->power_limit = dtpm->ops->set_power_uw(dtpm, power_limit); + } else { + dtpm->power_limit = 0; + + list_for_each_entry(child, &dtpm->children, sibling) { + + /* + * Integer division rounding will inevitably + * lead to a different min or max value when + * set several times. In order to restore the + * initial value, we force the child's min or + * max power every time if the constraint is + * at the boundaries. + */ + if (power_limit == dtpm->power_max) { + power = child->power_max; + } else if (power_limit == dtpm->power_min) { + power = child->power_min; + } else { + power = DIV_ROUND_CLOSEST_ULL( + power_limit * child->weight, 1024); + } + + pr_debug("Setting power limit for '%s': %llu uW\n", + child->zone.name, power); + + ret = __set_power_limit_uw(child, cid, power); + if (!ret) + ret = __get_power_limit_uw(child, cid, &power); + + if (ret) + break; + + dtpm->power_limit += power; + } + } + + return ret; +} + +static int set_power_limit_uw(struct powercap_zone *pcz, + int cid, u64 power_limit) +{ + struct dtpm *dtpm = to_dtpm(pcz); + int ret; + + mutex_lock(&dtpm_lock); + + /* + * Don't allow values outside of the power range previously + * set when initializing the power numbers. + */ + power_limit = clamp_val(power_limit, dtpm->power_min, dtpm->power_max); + + ret = __set_power_limit_uw(dtpm, cid, power_limit); + + pr_debug("%s: power limit: %llu uW, power max: %llu uW\n", + dtpm->zone.name, dtpm->power_limit, dtpm->power_max); + + mutex_unlock(&dtpm_lock); + + return ret; +} + +static const char *get_constraint_name(struct powercap_zone *pcz, int cid) +{ + return constraint_name[cid]; +} + +static int get_max_power_uw(struct powercap_zone *pcz, int id, u64 *max_power) +{ + struct dtpm *dtpm = to_dtpm(pcz); + + mutex_lock(&dtpm_lock); + *max_power = dtpm->power_max; + mutex_unlock(&dtpm_lock); + + return 0; +} + +static struct powercap_zone_constraint_ops constraint_ops = { + .set_power_limit_uw = set_power_limit_uw, + .get_power_limit_uw = get_power_limit_uw, + .set_time_window_us = set_time_window_us, + .get_time_window_us = get_time_window_us, + .get_max_power_uw = get_max_power_uw, + .get_name = get_constraint_name, +}; + +static struct powercap_zone_ops zone_ops = { + .get_max_power_range_uw = get_max_power_range_uw, + .get_power_uw = get_power_uw, + .release = dtpm_release_zone, +}; + +/** + * dtpm_alloc - Allocate and initialize a dtpm struct + * @name: a string specifying the name of the node + * + * Return: a struct dtpm pointer, NULL in case of error + */ +struct dtpm *dtpm_alloc(struct dtpm_ops *ops) +{ + struct dtpm *dtpm; + + dtpm = kzalloc(sizeof(*dtpm), GFP_KERNEL); + if (dtpm) { + INIT_LIST_HEAD(&dtpm->children); + INIT_LIST_HEAD(&dtpm->sibling); + dtpm->weight = 1024; + dtpm->ops = ops; + } + + return dtpm; +} + +/** + * dtpm_unregister - Unregister a dtpm node from the hierarchy tree + * @dtpm: a pointer to a dtpm structure corresponding to the node to be removed + * + * Call the underlying powercap unregister function. That will call + * the release callback of the powercap zone. + */ +void dtpm_unregister(struct dtpm *dtpm) +{ + powercap_unregister_zone(pct, &dtpm->zone); + + pr_info("Unregistered dtpm node '%s'\n", dtpm->zone.name); +} + +/** + * dtpm_register - Register a dtpm node in the hierarchy tree + * @name: a string specifying the name of the node + * @dtpm: a pointer to a dtpm structure corresponding to the new node + * @parent: a pointer to a dtpm structure corresponding to the parent node + * + * Create a dtpm node in the tree. If no parent is specified, the node + * is the root node of the hierarchy. If the root node already exists, + * then the registration will fail. The powercap controller must be + * initialized before calling this function. + * + * The dtpm structure must be initialized with the power numbers + * before calling this function. + * + * Return: zero on success, a negative value in case of error: + * -EAGAIN: the function is called before the framework is initialized. + * -EBUSY: the root node is already inserted + * -EINVAL: * there is no root node yet and @parent is specified + * * no all ops are defined + * * parent have ops which are reserved for leaves + * Other negative values are reported back from the powercap framework + */ +int dtpm_register(const char *name, struct dtpm *dtpm, struct dtpm *parent) +{ + struct powercap_zone *pcz; + + if (!pct) + return -EAGAIN; + + if (root && !parent) + return -EBUSY; + + if (!root && parent) + return -EINVAL; + + if (parent && parent->ops) + return -EINVAL; + + if (!dtpm) + return -EINVAL; + + if (dtpm->ops && !(dtpm->ops->set_power_uw && + dtpm->ops->get_power_uw && + dtpm->ops->release)) + return -EINVAL; + + pcz = powercap_register_zone(&dtpm->zone, pct, name, + parent ? &parent->zone : NULL, + &zone_ops, MAX_DTPM_CONSTRAINTS, + &constraint_ops); + if (IS_ERR(pcz)) + return PTR_ERR(pcz); + + mutex_lock(&dtpm_lock); + + if (parent) { + list_add_tail(&dtpm->sibling, &parent->children); + dtpm->parent = parent; + } else { + root = dtpm; + } + + __dtpm_add_power(dtpm); + + pr_info("Registered dtpm node '%s' / %llu-%llu uW, \n", + dtpm->zone.name, dtpm->power_min, dtpm->power_max); + + mutex_unlock(&dtpm_lock); + + return 0; +} + +static int __init dtpm_init(void) +{ + struct dtpm_descr **dtpm_descr; + + pct = powercap_register_control_type(NULL, "dtpm", NULL); + if (IS_ERR(pct)) { + pr_err("Failed to register control type\n"); + return PTR_ERR(pct); + } + + for_each_dtpm_table(dtpm_descr) + (*dtpm_descr)->init(*dtpm_descr); + + return 0; +} +late_initcall(dtpm_init); diff --git a/drivers/powercap/dtpm_cpu.c b/drivers/powercap/dtpm_cpu.c new file mode 100644 index 000000000000..51c366938acd --- /dev/null +++ b/drivers/powercap/dtpm_cpu.c @@ -0,0 +1,257 @@ +// SPDX-License-Identifier: GPL-2.0-only +/* + * Copyright 2020 Linaro Limited + * + * Author: Daniel Lezcano <daniel.lezcano@linaro.org> + * + * The DTPM CPU is based on the energy model. It hooks the CPU in the + * DTPM tree which in turns update the power number by propagating the + * power number from the CPU energy model information to the parents. + * + * The association between the power and the performance state, allows + * to set the power of the CPU at the OPP granularity. + * + * The CPU hotplug is supported and the power numbers will be updated + * if a CPU is hot plugged / unplugged. + */ +#include <linux/cpumask.h> +#include <linux/cpufreq.h> +#include <linux/cpuhotplug.h> +#include <linux/dtpm.h> +#include <linux/energy_model.h> +#include <linux/pm_qos.h> +#include <linux/slab.h> +#include <linux/units.h> + +static struct dtpm *__parent; + +static DEFINE_PER_CPU(struct dtpm *, dtpm_per_cpu); + +struct dtpm_cpu { + struct freq_qos_request qos_req; + int cpu; +}; + +/* + * When a new CPU is inserted at hotplug or boot time, add the power + * contribution and update the dtpm tree. + */ +static int power_add(struct dtpm *dtpm, struct em_perf_domain *em) +{ + u64 power_min, power_max; + + power_min = em->table[0].power; + power_min *= MICROWATT_PER_MILLIWATT; + power_min += dtpm->power_min; + + power_max = em->table[em->nr_perf_states - 1].power; + power_max *= MICROWATT_PER_MILLIWATT; + power_max += dtpm->power_max; + + return dtpm_update_power(dtpm, power_min, power_max); +} + +/* + * When a CPU is unplugged, remove its power contribution from the + * dtpm tree. + */ +static int power_sub(struct dtpm *dtpm, struct em_perf_domain *em) +{ + u64 power_min, power_max; + + power_min = em->table[0].power; + power_min *= MICROWATT_PER_MILLIWATT; + power_min = dtpm->power_min - power_min; + + power_max = em->table[em->nr_perf_states - 1].power; + power_max *= MICROWATT_PER_MILLIWATT; + power_max = dtpm->power_max - power_max; + + return dtpm_update_power(dtpm, power_min, power_max); +} + +static u64 set_pd_power_limit(struct dtpm *dtpm, u64 power_limit) +{ + struct dtpm_cpu *dtpm_cpu = dtpm->private; + struct em_perf_domain *pd; + struct cpumask cpus; + unsigned long freq; + u64 power; + int i, nr_cpus; + + pd = em_cpu_get(dtpm_cpu->cpu); + + cpumask_and(&cpus, cpu_online_mask, to_cpumask(pd->cpus)); + + nr_cpus = cpumask_weight(&cpus); + + for (i = 0; i < pd->nr_perf_states; i++) { + + power = pd->table[i].power * MICROWATT_PER_MILLIWATT * nr_cpus; + + if (power > power_limit) + break; + } + + freq = pd->table[i - 1].frequency; + + freq_qos_update_request(&dtpm_cpu->qos_req, freq); + + power_limit = pd->table[i - 1].power * + MICROWATT_PER_MILLIWATT * nr_cpus; + + return power_limit; +} + +static u64 get_pd_power_uw(struct dtpm *dtpm) +{ + struct dtpm_cpu *dtpm_cpu = dtpm->private; + struct em_perf_domain *pd; + struct cpumask cpus; + unsigned long freq; + int i, nr_cpus; + + pd = em_cpu_get(dtpm_cpu->cpu); + freq = cpufreq_quick_get(dtpm_cpu->cpu); + cpumask_and(&cpus, cpu_online_mask, to_cpumask(pd->cpus)); + nr_cpus = cpumask_weight(&cpus); + + for (i = 0; i < pd->nr_perf_states; i++) { + + if (pd->table[i].frequency < freq) + continue; + + return pd->table[i].power * + MICROWATT_PER_MILLIWATT * nr_cpus; + } + + return 0; +} + +static void pd_release(struct dtpm *dtpm) +{ + struct dtpm_cpu *dtpm_cpu = dtpm->private; + + if (freq_qos_request_active(&dtpm_cpu->qos_req)) + freq_qos_remove_request(&dtpm_cpu->qos_req); + + kfree(dtpm_cpu); +} + +static struct dtpm_ops dtpm_ops = { + .set_power_uw = set_pd_power_limit, + .get_power_uw = get_pd_power_uw, + .release = pd_release, +}; + +static int cpuhp_dtpm_cpu_offline(unsigned int cpu) +{ + struct cpufreq_policy *policy; + struct em_perf_domain *pd; + struct dtpm *dtpm; + + policy = cpufreq_cpu_get(cpu); + + if (!policy) + return 0; + + pd = em_cpu_get(cpu); + if (!pd) + return -EINVAL; + + dtpm = per_cpu(dtpm_per_cpu, cpu); + + power_sub(dtpm, pd); + + if (cpumask_weight(policy->cpus) != 1) + return 0; + + for_each_cpu(cpu, policy->related_cpus) + per_cpu(dtpm_per_cpu, cpu) = NULL; + + dtpm_unregister(dtpm); + + return 0; +} + +static int cpuhp_dtpm_cpu_online(unsigned int cpu) +{ + struct dtpm *dtpm; + struct dtpm_cpu *dtpm_cpu; + struct cpufreq_policy *policy; + struct em_perf_domain *pd; + char name[CPUFREQ_NAME_LEN]; + int ret = -ENOMEM; + + policy = cpufreq_cpu_get(cpu); + + if (!policy) + return 0; + + pd = em_cpu_get(cpu); + if (!pd) + return -EINVAL; + + dtpm = per_cpu(dtpm_per_cpu, cpu); + if (dtpm) + return power_add(dtpm, pd); + + dtpm = dtpm_alloc(&dtpm_ops); + if (!dtpm) + return -EINVAL; + + dtpm_cpu = kzalloc(sizeof(*dtpm_cpu), GFP_KERNEL); + if (!dtpm_cpu) + goto out_kfree_dtpm; + + dtpm->private = dtpm_cpu; + dtpm_cpu->cpu = cpu; + + for_each_cpu(cpu, policy->related_cpus) + per_cpu(dtpm_per_cpu, cpu) = dtpm; + + sprintf(name, "cpu%d", dtpm_cpu->cpu); + + ret = dtpm_register(name, dtpm, __parent); + if (ret) + goto out_kfree_dtpm_cpu; + + ret = power_add(dtpm, pd); + if (ret) + goto out_dtpm_unregister; + + ret = freq_qos_add_request(&policy->constraints, + &dtpm_cpu->qos_req, FREQ_QOS_MAX, + pd->table[pd->nr_perf_states - 1].frequency); + if (ret) + goto out_power_sub; + + return 0; + +out_power_sub: + power_sub(dtpm, pd); + +out_dtpm_unregister: + dtpm_unregister(dtpm); + dtpm_cpu = NULL; + dtpm = NULL; + +out_kfree_dtpm_cpu: + for_each_cpu(cpu, policy->related_cpus) + per_cpu(dtpm_per_cpu, cpu) = NULL; + kfree(dtpm_cpu); + +out_kfree_dtpm: + kfree(dtpm); + return ret; +} + +int dtpm_register_cpu(struct dtpm *parent) +{ + __parent = parent; + + return cpuhp_setup_state(CPUHP_AP_DTPM_CPU_ONLINE, + "dtpm_cpu:online", + cpuhp_dtpm_cpu_online, + cpuhp_dtpm_cpu_offline); +} diff --git a/drivers/powercap/intel_rapl_common.c b/drivers/powercap/intel_rapl_common.c index c9e57237d778..fdda2a737186 100644 --- a/drivers/powercap/intel_rapl_common.c +++ b/drivers/powercap/intel_rapl_common.c @@ -547,7 +547,7 @@ static void rapl_init_domains(struct rapl_package *rp) if (i == RAPL_DOMAIN_PLATFORM && rp->id > 0) { snprintf(rd->name, RAPL_DOMAIN_NAME_LENGTH, "psys-%d", - cpu_data(rp->lead_cpu).phys_proc_id); + topology_physical_package_id(rp->lead_cpu)); } else snprintf(rd->name, RAPL_DOMAIN_NAME_LENGTH, "%s", rapl_domain_names[i]); @@ -1049,6 +1049,7 @@ static const struct x86_cpu_id rapl_ids[] __initconst = { X86_MATCH_INTEL_FAM6_MODEL(TIGERLAKE, &rapl_defaults_core), X86_MATCH_INTEL_FAM6_MODEL(ROCKETLAKE, &rapl_defaults_core), X86_MATCH_INTEL_FAM6_MODEL(ALDERLAKE, &rapl_defaults_core), + X86_MATCH_INTEL_FAM6_MODEL(ALDERLAKE_L, &rapl_defaults_core), X86_MATCH_INTEL_FAM6_MODEL(SAPPHIRERAPIDS_X, &rapl_defaults_spr_server), X86_MATCH_INTEL_FAM6_MODEL(LAKEFIELD, &rapl_defaults_core), @@ -1309,7 +1310,6 @@ struct rapl_package *rapl_add_package(int cpu, struct rapl_if_priv *priv) { int id = topology_logical_die_id(cpu); struct rapl_package *rp; - struct cpuinfo_x86 *c = &cpu_data(cpu); int ret; if (!rapl_defaults) @@ -1326,10 +1326,11 @@ struct rapl_package *rapl_add_package(int cpu, struct rapl_if_priv *priv) if (topology_max_die_per_package() > 1) snprintf(rp->name, PACKAGE_DOMAIN_NAME_LENGTH, - "package-%d-die-%d", c->phys_proc_id, c->cpu_die_id); + "package-%d-die-%d", + topology_physical_package_id(cpu), topology_die_id(cpu)); else snprintf(rp->name, PACKAGE_DOMAIN_NAME_LENGTH, "package-%d", - c->phys_proc_id); + topology_physical_package_id(cpu)); /* check if the package contains valid domains */ if (rapl_detect_domains(rp, cpu) || rapl_defaults->check_unit(rp, cpu)) { diff --git a/include/asm-generic/vmlinux.lds.h b/include/asm-generic/vmlinux.lds.h index b2b3d81b1535..b3e4e0740089 100644 --- a/include/asm-generic/vmlinux.lds.h +++ b/include/asm-generic/vmlinux.lds.h @@ -316,6 +316,16 @@ #define THERMAL_TABLE(name) #endif +#ifdef CONFIG_DTPM +#define DTPM_TABLE() \ + . = ALIGN(8); \ + __dtpm_table = .; \ + KEEP(*(__dtpm_table)) \ + __dtpm_table_end = .; +#else +#define DTPM_TABLE() +#endif + #define KERNEL_DTB() \ STRUCT_ALIGN(); \ __dtb_start = .; \ @@ -733,6 +743,7 @@ ACPI_PROBE_TABLE(irqchip) \ ACPI_PROBE_TABLE(timer) \ THERMAL_TABLE(governor) \ + DTPM_TABLE() \ EARLYCON_TABLE() \ LSM_TABLE() \ EARLY_LSM_TABLE() \ diff --git a/include/linux/cpuhotplug.h b/include/linux/cpuhotplug.h index 0042ef362511..ee09a39627d6 100644 --- a/include/linux/cpuhotplug.h +++ b/include/linux/cpuhotplug.h @@ -193,6 +193,7 @@ enum cpuhp_state { CPUHP_AP_ONLINE_DYN_END = CPUHP_AP_ONLINE_DYN + 30, CPUHP_AP_X86_HPET_ONLINE, CPUHP_AP_X86_KVM_CLK_ONLINE, + CPUHP_AP_DTPM_CPU_ONLINE, CPUHP_AP_ACTIVE, CPUHP_ONLINE, }; diff --git a/include/linux/dtpm.h b/include/linux/dtpm.h new file mode 100644 index 000000000000..e80a332e3d8a --- /dev/null +++ b/include/linux/dtpm.h @@ -0,0 +1,77 @@ +/* SPDX-License-Identifier: GPL-2.0-only */ +/* + * Copyright (C) 2020 Linaro Ltd + * + * Author: Daniel Lezcano <daniel.lezcano@linaro.org> + */ +#ifndef ___DTPM_H__ +#define ___DTPM_H__ + +#include <linux/powercap.h> + +#define MAX_DTPM_DESCR 8 +#define MAX_DTPM_CONSTRAINTS 1 + +struct dtpm { + struct powercap_zone zone; + struct dtpm *parent; + struct list_head sibling; + struct list_head children; + struct dtpm_ops *ops; + unsigned long flags; + u64 power_limit; + u64 power_max; + u64 power_min; + int weight; + void *private; +}; + +struct dtpm_ops { + u64 (*set_power_uw)(struct dtpm *, u64); + u64 (*get_power_uw)(struct dtpm *); + void (*release)(struct dtpm *); +}; + +struct dtpm_descr; + +typedef int (*dtpm_init_t)(struct dtpm_descr *); + +struct dtpm_descr { + struct dtpm *parent; + const char *name; + dtpm_init_t init; +}; + +/* Init section thermal table */ +extern struct dtpm_descr *__dtpm_table[]; +extern struct dtpm_descr *__dtpm_table_end[]; + +#define DTPM_TABLE_ENTRY(name) \ + static typeof(name) *__dtpm_table_entry_##name \ + __used __section("__dtpm_table") = &name + +#define DTPM_DECLARE(name) DTPM_TABLE_ENTRY(name) + +#define for_each_dtpm_table(__dtpm) \ + for (__dtpm = __dtpm_table; \ + __dtpm < __dtpm_table_end; \ + __dtpm++) + +static inline struct dtpm *to_dtpm(struct powercap_zone *zone) +{ + return container_of(zone, struct dtpm, zone); +} + +int dtpm_update_power(struct dtpm *dtpm, u64 power_min, u64 power_max); + +int dtpm_release_zone(struct powercap_zone *pcz); + +struct dtpm *dtpm_alloc(struct dtpm_ops *ops); + +void dtpm_unregister(struct dtpm *dtpm); + +int dtpm_register(const char *name, struct dtpm *dtpm, struct dtpm *parent); + +int dtpm_register_cpu(struct dtpm *parent); + +#endif diff --git a/include/linux/units.h b/include/linux/units.h index 5c115c809507..dcc30a53fa93 100644 --- a/include/linux/units.h +++ b/include/linux/units.h @@ -4,6 +4,10 @@ #include <linux/math.h> +#define MILLIWATT_PER_WATT 1000L +#define MICROWATT_PER_MILLIWATT 1000L +#define MICROWATT_PER_WATT 1000000L + #define ABSOLUTE_ZERO_MILLICELSIUS -273150 static inline long milli_kelvin_to_millicelsius(long t) diff --git a/kernel/power/Kconfig b/kernel/power/Kconfig index a7320f07689d..6bfe3ead10ad 100644 --- a/kernel/power/Kconfig +++ b/kernel/power/Kconfig @@ -139,7 +139,6 @@ config PM_SLEEP_SMP_NONZERO_CPU config PM_AUTOSLEEP bool "Opportunistic sleep" depends on PM_SLEEP - default n help Allow the kernel to trigger a system transition into a global sleep state automatically whenever there are no active wakeup sources. @@ -147,7 +146,6 @@ config PM_AUTOSLEEP config PM_WAKELOCKS bool "User space wakeup sources interface" depends on PM_SLEEP - default n help Allow user space to create, activate and deactivate wakeup source objects with the help of a sysfs-based interface. @@ -293,7 +291,6 @@ config PM_GENERIC_DOMAINS config WQ_POWER_EFFICIENT_DEFAULT bool "Enable workqueue power-efficient mode by default" depends on PM - default n help Per-cpu workqueues are generally preferred because they show better performance thanks to cache locality; unfortunately, @@ -322,15 +319,14 @@ config CPU_PM bool config ENERGY_MODEL - bool "Energy Model for CPUs" + bool "Energy Model for devices with DVFS (CPUs, GPUs, etc)" depends on SMP depends on CPU_FREQ - default n help Several subsystems (thermal and/or the task scheduler for example) - can leverage information about the energy consumed by CPUs to make - smarter decisions. This config option enables the framework from - which subsystems can access the energy models. + can leverage information about the energy consumed by devices to + make smarter decisions. This config option enables the framework + from which subsystems can access the energy models. The exact usage of the energy model is subsystem-dependent. |