Merge branches 'pm-cpuidle', 'pm-sleep' and 'pm-powercap'

Merge cpuidle updates, changes related to system sleep and power capping
updates for 6.10:

 - Fix kerneldoc description of ladder_do_selection() (Jeff Johnson).

 - Convert the cpuidle kirkwood driver to platform remove callback
   returning void (Yangtao Li).

 - Replace deprecated strncpy() with strscpy() in the hibernation core
   code (Justin Stitt).

 - Use %ps to simplify debug output in the core system-wide suspend and
   resume code (Len Brown).

 - Remove unnecessary else from device_init_wakeup() and make
   device_wakeup_disable() return void (Dhruva Gole).

 - Enable PMU support in the Intel TPMI RAPL driver (Zhang Rui).

 - Add support for ArrowLake-H platform to the Intel RAPL driver (Zhang
   Rui).

 - Avoid explicit cpumask allocation on stack in DTPM (Dawei Li).

* pm-cpuidle:
  cpuidle: ladder: fix ladder_do_selection() kernel-doc
  cpuidle: kirkwood: Convert to platform remove callback returning void

* pm-sleep:
  PM: hibernate: replace deprecated strncpy() with strscpy()
  PM: sleep: Take advantage of %ps to simplify debug output
  PM: wakeup: Remove unnecessary else from device_init_wakeup()
  PM: wakeup: make device_wakeup_disable() return void

* pm-powercap:
  powercap: intel_rapl_tpmi: Enable PMU support
  powercap: intel_rapl: Introduce APIs for PMU support
  powercap: intel_rapl: Sort header files
  powercap: intel_rapl: Add support for ArrowLake-H platform
  powercap: DTPM: Avoid explicit cpumask allocation on stack

drivers/base/power/main.c

@@ -208,7 +208,7 @@ static ktime_t initcall_debug_start(struct device *dev, void *cb)
 	if (!pm_print_times_enabled)
 		return 0;
 
-	dev_info(dev, "calling %pS @ %i, parent: %s\n", cb,
+	dev_info(dev, "calling %ps @ %i, parent: %s\n", cb,
 		 task_pid_nr(current),
 		 dev->parent ? dev_name(dev->parent) : "none");
 	return ktime_get();
@@ -223,7 +223,7 @@ static void initcall_debug_report(struct device *dev, ktime_t calltime,
 		return;
 
 	rettime = ktime_get();
-	dev_info(dev, "%pS returned %d after %Ld usecs\n", cb, error,
+	dev_info(dev, "%ps returned %d after %Ld usecs\n", cb, error,
 		 (unsigned long long)ktime_us_delta(rettime, calltime));
 }
 
@@ -1927,7 +1927,7 @@ EXPORT_SYMBOL_GPL(dpm_suspend_start);
 void __suspend_report_result(const char *function, struct device *dev, void *fn, int ret)
 {
 	if (ret)
-		dev_err(dev, "%s(): %pS returns %d\n", function, fn, ret);
+		dev_err(dev, "%s(): %ps returns %d\n", function, fn, ret);
 }
 EXPORT_SYMBOL_GPL(__suspend_report_result);
 

drivers/base/power/wakeup.c

@@ -451,16 +451,15 @@ static struct wakeup_source *device_wakeup_detach(struct device *dev)
  * Detach the @dev's wakeup source object from it, unregister this wakeup source
  * object and destroy it.
  */
-int device_wakeup_disable(struct device *dev)
+void device_wakeup_disable(struct device *dev)
 {
 	struct wakeup_source *ws;
 
 	if (!dev || !dev->power.can_wakeup)
-		return -EINVAL;
+		return;
 
 	ws = device_wakeup_detach(dev);
 	wakeup_source_unregister(ws);
-	return 0;
 }
 EXPORT_SYMBOL_GPL(device_wakeup_disable);
 
@@ -502,7 +501,11 @@ EXPORT_SYMBOL_GPL(device_set_wakeup_capable);
  */
 int device_set_wakeup_enable(struct device *dev, bool enable)
 {
-	return enable ? device_wakeup_enable(dev) : device_wakeup_disable(dev);
+	if (enable)
+		return device_wakeup_enable(dev);
+
+	device_wakeup_disable(dev);
+	return 0;
 }
 EXPORT_SYMBOL_GPL(device_set_wakeup_enable);
 

drivers/cpuidle/cpuidle-kirkwood.c

@@ -59,15 +59,14 @@ static int kirkwood_cpuidle_probe(struct platform_device *pdev)
 	return cpuidle_register(&kirkwood_idle_driver, NULL);
 }
 
-static int kirkwood_cpuidle_remove(struct platform_device *pdev)
+static void kirkwood_cpuidle_remove(struct platform_device *pdev)
 {
 	cpuidle_unregister(&kirkwood_idle_driver);
-	return 0;
 }
 
 static struct platform_driver kirkwood_cpuidle_driver = {
 	.probe = kirkwood_cpuidle_probe,
-	.remove = kirkwood_cpuidle_remove,
+	.remove_new = kirkwood_cpuidle_remove,
 	.driver = {
 		   .name = "kirkwood_cpuidle",
 		   },

drivers/cpuidle/governors/ladder.c

@@ -44,6 +44,7 @@ static DEFINE_PER_CPU(struct ladder_device, ladder_devices);
 
 /**
  * ladder_do_selection - prepares private data for a state change
+ * @dev: the CPU
  * @ldev: the ladder device
  * @old_idx: the current state index
  * @new_idx: the new target state index

drivers/mmc/host/sdhci-pci-core.c

@@ -63,7 +63,7 @@ static int sdhci_pci_init_wakeup(struct sdhci_pci_chip *chip)
 	if ((pm_flags & MMC_PM_KEEP_POWER) && (pm_flags & MMC_PM_WAKE_SDIO_IRQ))
 		return device_wakeup_enable(&chip->pdev->dev);
 	else if (!cap_cd_wake)
-		return device_wakeup_disable(&chip->pdev->dev);
+		device_wakeup_disable(&chip->pdev->dev);
 
 	return 0;
 }

drivers/powercap/dtpm_cpu.c

@@ -43,13 +43,11 @@ static u64 set_pd_power_limit(struct dtpm *dtpm, u64 power_limit)
 	struct dtpm_cpu *dtpm_cpu = to_dtpm_cpu(dtpm);
 	struct em_perf_domain *pd = em_cpu_get(dtpm_cpu->cpu);
 	struct em_perf_state *table;
-	struct cpumask cpus;
 	unsigned long freq;
 	u64 power;
 	int i, nr_cpus;
 
-	cpumask_and(&cpus, cpu_online_mask, to_cpumask(pd->cpus));
-	nr_cpus = cpumask_weight(&cpus);
+	nr_cpus = cpumask_weight_and(cpu_online_mask, to_cpumask(pd->cpus));
 
 	rcu_read_lock();
 	table = em_perf_state_from_pd(pd);
@@ -123,11 +121,9 @@ static int update_pd_power_uw(struct dtpm *dtpm)
 	struct dtpm_cpu *dtpm_cpu = to_dtpm_cpu(dtpm);
 	struct em_perf_domain *em = em_cpu_get(dtpm_cpu->cpu);
 	struct em_perf_state *table;
-	struct cpumask cpus;
 	int nr_cpus;
 
-	cpumask_and(&cpus, cpu_online_mask, to_cpumask(em->cpus));
-	nr_cpus = cpumask_weight(&cpus);
+	nr_cpus = cpumask_weight_and(cpu_online_mask, to_cpumask(em->cpus));
 
 	rcu_read_lock();
 	table = em_perf_state_from_pd(em);

drivers/powercap/intel_rapl_common.c

@@ -5,27 +5,29 @@
  */
 #define pr_fmt(fmt) KBUILD_MODNAME ": " fmt
 
+#include <linux/bitmap.h>
 #include <linux/cleanup.h>
+#include <linux/cpu.h>
+#include <linux/delay.h>
+#include <linux/device.h>
+#include <linux/intel_rapl.h>
 #include <linux/kernel.h>
-#include <linux/module.h>
 #include <linux/list.h>
-#include <linux/types.h>
-#include <linux/device.h>
-#include <linux/slab.h>
 #include <linux/log2.h>
-#include <linux/bitmap.h>
-#include <linux/delay.h>
-#include <linux/sysfs.h>
-#include <linux/cpu.h>
+#include <linux/module.h>
+#include <linux/nospec.h>
+#include <linux/perf_event.h>
+#include <linux/platform_device.h>
 #include <linux/powercap.h>
-#include <linux/suspend.h>
-#include <linux/intel_rapl.h>
 #include <linux/processor.h>
-#include <linux/platform_device.h>
+#include <linux/slab.h>
+#include <linux/suspend.h>
+#include <linux/sysfs.h>
+#include <linux/types.h>
 
-#include <asm/iosf_mbi.h>
 #include <asm/cpu_device_id.h>
 #include <asm/intel-family.h>
+#include <asm/iosf_mbi.h>
 
 /* bitmasks for RAPL MSRs, used by primitive access functions */
 #define ENERGY_STATUS_MASK      0xffffffff
@@ -1263,6 +1265,7 @@ static const struct x86_cpu_id rapl_ids[] __initconst = {
 	X86_MATCH_INTEL_FAM6_MODEL(SAPPHIRERAPIDS_X,	&rapl_defaults_spr_server),
 	X86_MATCH_INTEL_FAM6_MODEL(EMERALDRAPIDS_X,	&rapl_defaults_spr_server),
 	X86_MATCH_INTEL_FAM6_MODEL(LUNARLAKE_M,		&rapl_defaults_core),
+	X86_MATCH_INTEL_FAM6_MODEL(ARROWLAKE_H,		&rapl_defaults_core),
 	X86_MATCH_INTEL_FAM6_MODEL(ARROWLAKE,		&rapl_defaults_core),
 	X86_MATCH_INTEL_FAM6_MODEL(LAKEFIELD,		&rapl_defaults_core),
 
@@ -1506,6 +1509,586 @@ static int rapl_detect_domains(struct rapl_package *rp)
 	return 0;
 }
 
+#ifdef CONFIG_PERF_EVENTS
+
+/*
+ * Support for RAPL PMU
+ *
+ * Register a PMU if any of the registered RAPL Packages have the requirement
+ * of exposing its energy counters via Perf PMU.
+ *
+ * PMU Name:
+ *	power
+ *
+ * Events:
+ *	Name		Event id	RAPL Domain
+ *	energy_cores	0x01		RAPL_DOMAIN_PP0
+ *	energy_pkg	0x02		RAPL_DOMAIN_PACKAGE
+ *	energy_ram	0x03		RAPL_DOMAIN_DRAM
+ *	energy_gpu	0x04		RAPL_DOMAIN_PP1
+ *	energy_psys	0x05		RAPL_DOMAIN_PLATFORM
+ *
+ * Unit:
+ *	Joules
+ *
+ * Scale:
+ *	2.3283064365386962890625e-10
+ *	The same RAPL domain in different RAPL Packages may have different
+ *	energy units. Use 2.3283064365386962890625e-10 (2^-32) Joules as
+ *	the fixed unit for all energy counters, and covert each hardware
+ *	counter increase to N times of PMU event counter increases.
+ *
+ * This is fully compatible with the current MSR RAPL PMU. This means that
+ * userspace programs like turbostat can use the same code to handle RAPL Perf
+ * PMU, no matter what RAPL Interface driver (MSR/TPMI, etc) is running
+ * underlying on the platform.
+ *
+ * Note that RAPL Packages can be probed/removed dynamically, and the events
+ * supported by each TPMI RAPL device can be different. Thus the RAPL PMU
+ * support is done on demand, which means
+ * 1. PMU is registered only if it is needed by a RAPL Package. PMU events for
+ *    unsupported counters are not exposed.
+ * 2. PMU is unregistered and registered when a new RAPL Package is probed and
+ *    supports new counters that are not supported by current PMU.
+ * 3. PMU is unregistered when all registered RAPL Packages don't need PMU.
+ */
+
+struct rapl_pmu {
+	struct pmu pmu;			/* Perf PMU structure */
+	u64 timer_ms;			/* Maximum expiration time to avoid counter overflow */
+	unsigned long domain_map;	/* Events supported by current registered PMU */
+	bool registered;		/* Whether the PMU has been registered or not */
+};
+
+static struct rapl_pmu rapl_pmu;
+
+/* PMU helpers */
+
+static int get_pmu_cpu(struct rapl_package *rp)
+{
+	int cpu;
+
+	if (!rp->has_pmu)
+		return nr_cpu_ids;
+
+	/* Only TPMI RAPL is supported for now */
+	if (rp->priv->type != RAPL_IF_TPMI)
+		return nr_cpu_ids;
+
+	/* TPMI RAPL uses any CPU in the package for PMU */
+	for_each_online_cpu(cpu)
+		if (topology_physical_package_id(cpu) == rp->id)
+			return cpu;
+
+	return nr_cpu_ids;
+}
+
+static bool is_rp_pmu_cpu(struct rapl_package *rp, int cpu)
+{
+	if (!rp->has_pmu)
+		return false;
+
+	/* Only TPMI RAPL is supported for now */
+	if (rp->priv->type != RAPL_IF_TPMI)
+		return false;
+
+	/* TPMI RAPL uses any CPU in the package for PMU */
+	return topology_physical_package_id(cpu) == rp->id;
+}
+
+static struct rapl_package_pmu_data *event_to_pmu_data(struct perf_event *event)
+{
+	struct rapl_package *rp = event->pmu_private;
+
+	return &rp->pmu_data;
+}
+
+/* PMU event callbacks */
+
+static u64 event_read_counter(struct perf_event *event)
+{
+	struct rapl_package *rp = event->pmu_private;
+	u64 val;
+	int ret;
+
+	/* Return 0 for unsupported events */
+	if (event->hw.idx < 0)
+		return 0;
+
+	ret = rapl_read_data_raw(&rp->domains[event->hw.idx], ENERGY_COUNTER, false, &val);
+
+	/* Return 0 for failed read */
+	if (ret)
+		return 0;
+
+	return val;
+}
+
+static void __rapl_pmu_event_start(struct perf_event *event)
+{
+	struct rapl_package_pmu_data *data = event_to_pmu_data(event);
+
+	if (WARN_ON_ONCE(!(event->hw.state & PERF_HES_STOPPED)))
+		return;
+
+	event->hw.state = 0;
+
+	list_add_tail(&event->active_entry, &data->active_list);
+
+	local64_set(&event->hw.prev_count, event_read_counter(event));
+	if (++data->n_active == 1)
+		hrtimer_start(&data->hrtimer, data->timer_interval,
+			      HRTIMER_MODE_REL_PINNED);
+}
+
+static void rapl_pmu_event_start(struct perf_event *event, int mode)
+{
+	struct rapl_package_pmu_data *data = event_to_pmu_data(event);
+	unsigned long flags;
+
+	raw_spin_lock_irqsave(&data->lock, flags);
+	__rapl_pmu_event_start(event);
+	raw_spin_unlock_irqrestore(&data->lock, flags);
+}
+
+static u64 rapl_event_update(struct perf_event *event)
+{
+	struct hw_perf_event *hwc = &event->hw;
+	struct rapl_package_pmu_data *data = event_to_pmu_data(event);
+	u64 prev_raw_count, new_raw_count;
+	s64 delta, sdelta;
+
+	/*
+	 * Follow the generic code to drain hwc->prev_count.
+	 * The loop is not expected to run for multiple times.
+	 */
+	prev_raw_count = local64_read(&hwc->prev_count);
+	do {
+		new_raw_count = event_read_counter(event);
+	} while (!local64_try_cmpxchg(&hwc->prev_count,
+		&prev_raw_count, new_raw_count));
+
+
+	/*
+	 * Now we have the new raw value and have updated the prev
+	 * timestamp already. We can now calculate the elapsed delta
+	 * (event-)time and add that to the generic event.
+	 */
+	delta = new_raw_count - prev_raw_count;
+
+	/*
+	 * Scale delta to smallest unit (2^-32)
+	 * users must then scale back: count * 1/(1e9*2^32) to get Joules
+	 * or use ldexp(count, -32).
+	 * Watts = Joules/Time delta
+	 */
+	sdelta = delta * data->scale[event->hw.flags];
+
+	local64_add(sdelta, &event->count);
+
+	return new_raw_count;
+}
+
+static void rapl_pmu_event_stop(struct perf_event *event, int mode)
+{
+	struct rapl_package_pmu_data *data = event_to_pmu_data(event);
+	struct hw_perf_event *hwc = &event->hw;
+	unsigned long flags;
+
+	raw_spin_lock_irqsave(&data->lock, flags);
+
+	/* Mark event as deactivated and stopped */
+	if (!(hwc->state & PERF_HES_STOPPED)) {
+		WARN_ON_ONCE(data->n_active <= 0);
+		if (--data->n_active == 0)
+			hrtimer_cancel(&data->hrtimer);
+
+		list_del(&event->active_entry);
+
+		WARN_ON_ONCE(hwc->state & PERF_HES_STOPPED);
+		hwc->state |= PERF_HES_STOPPED;
+	}
+
+	/* Check if update of sw counter is necessary */
+	if ((mode & PERF_EF_UPDATE) && !(hwc->state & PERF_HES_UPTODATE)) {
+		/*
+		 * Drain the remaining delta count out of a event
+		 * that we are disabling:
+		 */
+		rapl_event_update(event);
+		hwc->state |= PERF_HES_UPTODATE;
+	}
+
+	raw_spin_unlock_irqrestore(&data->lock, flags);
+}
+
+static int rapl_pmu_event_add(struct perf_event *event, int mode)
+{
+	struct rapl_package_pmu_data *data = event_to_pmu_data(event);
+	struct hw_perf_event *hwc = &event->hw;
+	unsigned long flags;
+
+	raw_spin_lock_irqsave(&data->lock, flags);
+
+	hwc->state = PERF_HES_UPTODATE | PERF_HES_STOPPED;
+
+	if (mode & PERF_EF_START)
+		__rapl_pmu_event_start(event);
+
+	raw_spin_unlock_irqrestore(&data->lock, flags);
+
+	return 0;
+}
+
+static void rapl_pmu_event_del(struct perf_event *event, int flags)
+{
+	rapl_pmu_event_stop(event, PERF_EF_UPDATE);
+}
+
+/* RAPL PMU event ids, same as shown in sysfs */
+enum perf_rapl_events {
+	PERF_RAPL_PP0 = 1,	/* all cores */
+	PERF_RAPL_PKG,		/* entire package */
+	PERF_RAPL_RAM,		/* DRAM */
+	PERF_RAPL_PP1,		/* gpu */
+	PERF_RAPL_PSYS,		/* psys */
+	PERF_RAPL_MAX
+};
+#define RAPL_EVENT_MASK GENMASK(7, 0)
+
+static const int event_to_domain[PERF_RAPL_MAX] = {
+	[PERF_RAPL_PP0]		= RAPL_DOMAIN_PP0,
+	[PERF_RAPL_PKG]		= RAPL_DOMAIN_PACKAGE,
+	[PERF_RAPL_RAM]		= RAPL_DOMAIN_DRAM,
+	[PERF_RAPL_PP1]		= RAPL_DOMAIN_PP1,
+	[PERF_RAPL_PSYS]	= RAPL_DOMAIN_PLATFORM,
+};
+
+static int rapl_pmu_event_init(struct perf_event *event)
+{
+	struct rapl_package *pos, *rp = NULL;
+	u64 cfg = event->attr.config & RAPL_EVENT_MASK;
+	int domain, idx;
+
+	/* Only look at RAPL events */
+	if (event->attr.type != event->pmu->type)
+		return -ENOENT;
+
+	/* Check for supported events only */
+	if (!cfg || cfg >= PERF_RAPL_MAX)
+		return -EINVAL;
+
+	if (event->cpu < 0)
+		return -EINVAL;
+
+	/* Find out which Package the event belongs to */
+	list_for_each_entry(pos, &rapl_packages, plist) {
+		if (is_rp_pmu_cpu(pos, event->cpu)) {
+			rp = pos;
+			break;
+		}
+	}
+	if (!rp)
+		return -ENODEV;
+
+	/* Find out which RAPL Domain the event belongs to */
+	domain = event_to_domain[cfg];
+
+	event->event_caps |= PERF_EV_CAP_READ_ACTIVE_PKG;
+	event->pmu_private = rp;	/* Which package */
+	event->hw.flags = domain;	/* Which domain */
+
+	event->hw.idx = -1;
+	/* Find out the index in rp->domains[] to get domain pointer */
+	for (idx = 0; idx < rp->nr_domains; idx++) {
+		if (rp->domains[idx].id == domain) {
+			event->hw.idx = idx;
+			break;
+		}
+	}
+
+	return 0;
+}
+
+static void rapl_pmu_event_read(struct perf_event *event)
+{
+	rapl_event_update(event);
+}
+
+static enum hrtimer_restart rapl_hrtimer_handle(struct hrtimer *hrtimer)
+{
+	struct rapl_package_pmu_data *data =
+		container_of(hrtimer, struct rapl_package_pmu_data, hrtimer);
+	struct perf_event *event;
+	unsigned long flags;
+
+	if (!data->n_active)
+		return HRTIMER_NORESTART;
+
+	raw_spin_lock_irqsave(&data->lock, flags);
+
+	list_for_each_entry(event, &data->active_list, active_entry)
+		rapl_event_update(event);
+
+	raw_spin_unlock_irqrestore(&data->lock, flags);
+
+	hrtimer_forward_now(hrtimer, data->timer_interval);
+
+	return HRTIMER_RESTART;
+}
+
+/* PMU sysfs attributes */
+
+/*
+ * There are no default events, but we need to create "events" group (with
+ * empty attrs) before updating it with detected events.
+ */
+static struct attribute *attrs_empty[] = {
+	NULL,
+};
+
+static struct attribute_group pmu_events_group = {
+	.name = "events",
+	.attrs = attrs_empty,
+};
+
+static ssize_t cpumask_show(struct device *dev,
+			    struct device_attribute *attr, char *buf)
+{
+	struct rapl_package *rp;
+	cpumask_var_t cpu_mask;
+	int cpu;
+	int ret;
+
+	if (!alloc_cpumask_var(&cpu_mask, GFP_KERNEL))
+		return -ENOMEM;
+
+	cpus_read_lock();
+
+	cpumask_clear(cpu_mask);
+
+	/* Choose a cpu for each RAPL Package */
+	list_for_each_entry(rp, &rapl_packages, plist) {
+		cpu = get_pmu_cpu(rp);
+		if (cpu < nr_cpu_ids)
+			cpumask_set_cpu(cpu, cpu_mask);
+	}
+	cpus_read_unlock();
+
+	ret = cpumap_print_to_pagebuf(true, buf, cpu_mask);
+
+	free_cpumask_var(cpu_mask);
+
+	return ret;
+}
+
+static DEVICE_ATTR_RO(cpumask);
+
+static struct attribute *pmu_cpumask_attrs[] = {
+	&dev_attr_cpumask.attr,
+	NULL
+};
+
+static struct attribute_group pmu_cpumask_group = {
+	.attrs = pmu_cpumask_attrs,
+};
+
+PMU_FORMAT_ATTR(event, "config:0-7");
+static struct attribute *pmu_format_attr[] = {
+	&format_attr_event.attr,
+	NULL
+};
+
+static struct attribute_group pmu_format_group = {
+	.name = "format",
+	.attrs = pmu_format_attr,
+};
+
+static const struct attribute_group *pmu_attr_groups[] = {
+	&pmu_events_group,
+	&pmu_cpumask_group,
+	&pmu_format_group,
+	NULL
+};
+
+#define RAPL_EVENT_ATTR_STR(_name, v, str)					\
+static struct perf_pmu_events_attr event_attr_##v = {				\
+	.attr		= __ATTR(_name, 0444, perf_event_sysfs_show, NULL),	\
+	.event_str	= str,							\
+}
+
+RAPL_EVENT_ATTR_STR(energy-cores,	rapl_cores,	"event=0x01");
+RAPL_EVENT_ATTR_STR(energy-pkg,		rapl_pkg,	"event=0x02");
+RAPL_EVENT_ATTR_STR(energy-ram,		rapl_ram,	"event=0x03");
+RAPL_EVENT_ATTR_STR(energy-gpu,		rapl_gpu,	"event=0x04");
+RAPL_EVENT_ATTR_STR(energy-psys,	rapl_psys,	"event=0x05");
+
+RAPL_EVENT_ATTR_STR(energy-cores.unit,	rapl_unit_cores,	"Joules");
+RAPL_EVENT_ATTR_STR(energy-pkg.unit,	rapl_unit_pkg,		"Joules");
+RAPL_EVENT_ATTR_STR(energy-ram.unit,	rapl_unit_ram,		"Joules");
+RAPL_EVENT_ATTR_STR(energy-gpu.unit,	rapl_unit_gpu,		"Joules");
+RAPL_EVENT_ATTR_STR(energy-psys.unit,	rapl_unit_psys,		"Joules");
+
+RAPL_EVENT_ATTR_STR(energy-cores.scale,	rapl_scale_cores,	"2.3283064365386962890625e-10");
+RAPL_EVENT_ATTR_STR(energy-pkg.scale,	rapl_scale_pkg,		"2.3283064365386962890625e-10");
+RAPL_EVENT_ATTR_STR(energy-ram.scale,	rapl_scale_ram,		"2.3283064365386962890625e-10");
+RAPL_EVENT_ATTR_STR(energy-gpu.scale,	rapl_scale_gpu,		"2.3283064365386962890625e-10");
+RAPL_EVENT_ATTR_STR(energy-psys.scale,	rapl_scale_psys,	"2.3283064365386962890625e-10");
+
+#define RAPL_EVENT_GROUP(_name, domain)			\
+static struct attribute *pmu_attr_##_name[] = {		\
+	&event_attr_rapl_##_name.attr.attr,		\
+	&event_attr_rapl_unit_##_name.attr.attr,	\
+	&event_attr_rapl_scale_##_name.attr.attr,	\
+	NULL						\
+};							\
+static umode_t is_visible_##_name(struct kobject *kobj, struct attribute *attr, int event)	\
+{											\
+	return rapl_pmu.domain_map & BIT(domain) ? attr->mode : 0;	\
+}							\
+static struct attribute_group pmu_group_##_name = {	\
+	.name  = "events",				\
+	.attrs = pmu_attr_##_name,			\
+	.is_visible = is_visible_##_name,		\
+}
+
+RAPL_EVENT_GROUP(cores,	RAPL_DOMAIN_PP0);
+RAPL_EVENT_GROUP(pkg,	RAPL_DOMAIN_PACKAGE);
+RAPL_EVENT_GROUP(ram,	RAPL_DOMAIN_DRAM);
+RAPL_EVENT_GROUP(gpu,	RAPL_DOMAIN_PP1);
+RAPL_EVENT_GROUP(psys,	RAPL_DOMAIN_PLATFORM);
+
+static const struct attribute_group *pmu_attr_update[] = {
+	&pmu_group_cores,
+	&pmu_group_pkg,
+	&pmu_group_ram,
+	&pmu_group_gpu,
+	&pmu_group_psys,
+	NULL
+};
+
+static int rapl_pmu_update(struct rapl_package *rp)
+{
+	int ret = 0;
+
+	/* Return if PMU already covers all events supported by current RAPL Package */
+	if (rapl_pmu.registered && !(rp->domain_map & (~rapl_pmu.domain_map)))
+		goto end;
+
+	/* Unregister previous registered PMU */
+	if (rapl_pmu.registered)
+		perf_pmu_unregister(&rapl_pmu.pmu);
+
+	rapl_pmu.registered = false;
+	rapl_pmu.domain_map |= rp->domain_map;
+
+	memset(&rapl_pmu.pmu, 0, sizeof(struct pmu));
+	rapl_pmu.pmu.attr_groups = pmu_attr_groups;
+	rapl_pmu.pmu.attr_update = pmu_attr_update;
+	rapl_pmu.pmu.task_ctx_nr = perf_invalid_context;
+	rapl_pmu.pmu.event_init = rapl_pmu_event_init;
+	rapl_pmu.pmu.add = rapl_pmu_event_add;
+	rapl_pmu.pmu.del = rapl_pmu_event_del;
+	rapl_pmu.pmu.start = rapl_pmu_event_start;
+	rapl_pmu.pmu.stop = rapl_pmu_event_stop;
+	rapl_pmu.pmu.read = rapl_pmu_event_read;
+	rapl_pmu.pmu.module = THIS_MODULE;
+	rapl_pmu.pmu.capabilities = PERF_PMU_CAP_NO_EXCLUDE | PERF_PMU_CAP_NO_INTERRUPT;
+	ret = perf_pmu_register(&rapl_pmu.pmu, "power", -1);
+	if (ret) {
+		pr_info("Failed to register PMU\n");
+		return ret;
+	}
+
+	rapl_pmu.registered = true;
+end:
+	rp->has_pmu = true;
+	return ret;
+}
+
+int rapl_package_add_pmu(struct rapl_package *rp)
+{
+	struct rapl_package_pmu_data *data = &rp->pmu_data;
+	int idx;
+
+	if (rp->has_pmu)
+		return -EEXIST;
+
+	guard(cpus_read_lock)();
+
+	for (idx = 0; idx < rp->nr_domains; idx++) {
+		struct rapl_domain *rd = &rp->domains[idx];
+		int domain = rd->id;
+		u64 val;
+
+		if (!test_bit(domain, &rp->domain_map))
+			continue;
+
+		/*
+		 * The RAPL PMU granularity is 2^-32 Joules
+		 * data->scale[]: times of 2^-32 Joules for each ENERGY COUNTER increase
+		 */
+		val = rd->energy_unit * (1ULL << 32);
+		do_div(val, ENERGY_UNIT_SCALE * 1000000);
+		data->scale[domain] = val;
+
+		if (!rapl_pmu.timer_ms) {
+			struct rapl_primitive_info *rpi = get_rpi(rp, ENERGY_COUNTER);
+
+			/*
+			 * Calculate the timer rate:
+			 * Use reference of 200W for scaling the timeout to avoid counter
+			 * overflows.
+			 *
+			 * max_count = rpi->mask >> rpi->shift + 1
+			 * max_energy_pj = max_count * rd->energy_unit
+			 * max_time_sec = (max_energy_pj / 1000000000) / 200w
+			 *
+			 * rapl_pmu.timer_ms = max_time_sec * 1000 / 2
+			 */
+			val = (rpi->mask >> rpi->shift) + 1;
+			val *= rd->energy_unit;
+			do_div(val, 1000000 * 200 * 2);
+			rapl_pmu.timer_ms = val;
+
+			pr_debug("%llu ms overflow timer\n", rapl_pmu.timer_ms);
+		}
+
+		pr_debug("Domain %s: hw unit %lld * 2^-32 Joules\n", rd->name, data->scale[domain]);
+	}
+
+	/* Initialize per package PMU data */
+	raw_spin_lock_init(&data->lock);
+	INIT_LIST_HEAD(&data->active_list);
+	data->timer_interval = ms_to_ktime(rapl_pmu.timer_ms);
+	hrtimer_init(&data->hrtimer, CLOCK_MONOTONIC, HRTIMER_MODE_REL);
+	data->hrtimer.function = rapl_hrtimer_handle;
+
+	return rapl_pmu_update(rp);
+}
+EXPORT_SYMBOL_GPL(rapl_package_add_pmu);
+
+void rapl_package_remove_pmu(struct rapl_package *rp)
+{
+	struct rapl_package *pos;
+
+	if (!rp->has_pmu)
+		return;
+
+	guard(cpus_read_lock)();
+
+	list_for_each_entry(pos, &rapl_packages, plist) {
+		/* PMU is still needed */
+		if (pos->has_pmu && pos != rp)
+			return;
+	}
+
+	perf_pmu_unregister(&rapl_pmu.pmu);
+	memset(&rapl_pmu, 0, sizeof(struct rapl_pmu));
+}
+EXPORT_SYMBOL_GPL(rapl_package_remove_pmu);
+#endif
+
 /* called from CPU hotplug notifier, hotplug lock held */
 void rapl_remove_package_cpuslocked(struct rapl_package *rp)
 {

drivers/powercap/intel_rapl_tpmi.c

@@ -302,6 +302,8 @@ static int intel_rapl_tpmi_probe(struct auxiliary_device *auxdev,
 		goto err;
 	}
 
+	rapl_package_add_pmu(trp->rp);
+
 	auxiliary_set_drvdata(auxdev, trp);
 
 	return 0;
@@ -314,6 +316,7 @@ static void intel_rapl_tpmi_remove(struct auxiliary_device *auxdev)
 {
 	struct tpmi_rapl_package *trp = auxiliary_get_drvdata(auxdev);
 
+	rapl_package_remove_pmu(trp->rp);
 	rapl_remove_package(trp->rp);
 	trp_release(trp);
 }

include/linux/intel_rapl.h

@@ -158,6 +158,26 @@ struct rapl_if_priv {
 	void *rpi;
 };
 
+#ifdef CONFIG_PERF_EVENTS
+/**
+ * struct rapl_package_pmu_data: Per package data for PMU support
+ * @scale:		Scale of 2^-32 Joules for each energy counter increase.
+ * @lock:		Lock to protect n_active and active_list.
+ * @n_active:		Number of active events.
+ * @active_list:	List of active events.
+ * @timer_interval:	Maximum timer expiration time before counter overflow.
+ * @hrtimer:		Periodically update the counter to prevent overflow.
+ */
+struct rapl_package_pmu_data {
+	u64 scale[RAPL_DOMAIN_MAX];
+	raw_spinlock_t lock;
+	int n_active;
+	struct list_head active_list;
+	ktime_t timer_interval;
+	struct hrtimer hrtimer;
+};
+#endif
+
 /* maximum rapl package domain name: package-%d-die-%d */
 #define PACKAGE_DOMAIN_NAME_LENGTH 30
 
@@ -176,6 +196,10 @@ struct rapl_package {
 	struct cpumask cpumask;
 	char name[PACKAGE_DOMAIN_NAME_LENGTH];
 	struct rapl_if_priv *priv;
+#ifdef CONFIG_PERF_EVENTS
+	bool has_pmu;
+	struct rapl_package_pmu_data pmu_data;
+#endif
 };
 
 struct rapl_package *rapl_find_package_domain_cpuslocked(int id, struct rapl_if_priv *priv,
@@ -188,4 +212,12 @@ struct rapl_package *rapl_find_package_domain(int id, struct rapl_if_priv *priv,
 struct rapl_package *rapl_add_package(int id, struct rapl_if_priv *priv, bool id_is_cpu);
 void rapl_remove_package(struct rapl_package *rp);
 
+#ifdef CONFIG_PERF_EVENTS
+int rapl_package_add_pmu(struct rapl_package *rp);
+void rapl_package_remove_pmu(struct rapl_package *rp);
+#else
+static inline int rapl_package_add_pmu(struct rapl_package *rp) { return 0; }
+static inline void rapl_package_remove_pmu(struct rapl_package *rp) { }
+#endif
+
 #endif /* __INTEL_RAPL_H__ */

include/linux/pm_wakeup.h

@@ -107,7 +107,7 @@ extern void wakeup_sources_read_unlock(int idx);
 extern struct wakeup_source *wakeup_sources_walk_start(void);
 extern struct wakeup_source *wakeup_sources_walk_next(struct wakeup_source *ws);
 extern int device_wakeup_enable(struct device *dev);
-extern int device_wakeup_disable(struct device *dev);
+extern void device_wakeup_disable(struct device *dev);
 extern void device_set_wakeup_capable(struct device *dev, bool capable);
 extern int device_set_wakeup_enable(struct device *dev, bool enable);
 extern void __pm_stay_awake(struct wakeup_source *ws);
@@ -154,10 +154,9 @@ static inline int device_wakeup_enable(struct device *dev)
 	return 0;
 }
 
-static inline int device_wakeup_disable(struct device *dev)
+static inline void device_wakeup_disable(struct device *dev)
 {
 	dev->power.should_wakeup = false;
-	return 0;
 }
 
 static inline int device_set_wakeup_enable(struct device *dev, bool enable)
@@ -235,11 +234,10 @@ static inline int device_init_wakeup(struct device *dev, bool enable)
 	if (enable) {
 		device_set_wakeup_capable(dev, true);
 		return device_wakeup_enable(dev);
-	} else {
-		device_wakeup_disable(dev);
-		device_set_wakeup_capable(dev, false);
-		return 0;
 	}
+	device_wakeup_disable(dev);
+	device_set_wakeup_capable(dev, false);
+	return 0;
 }
 
 #endif /* _LINUX_PM_WAKEUP_H */

kernel/power/hibernate.c

@@ -1361,7 +1361,7 @@ static int __init resume_setup(char *str)
 	if (noresume)
 		return 1;
 
-	strncpy(resume_file, str, 255);
+	strscpy(resume_file, str);
 	return 1;
 }