summaryrefslogtreecommitdiff
diff options
context:
space:
mode:
Diffstat
-rw-r--r--Documentation/admin-guide/pm/intel_uncore_frequency_scaling.rst60
-rw-r--r--Documentation/admin-guide/pm/working-state.rst1
-rw-r--r--arch/x86/kernel/acpi/sleep.c23
-rw-r--r--drivers/acpi/sleep.c11
-rw-r--r--drivers/base/power/domain.c42
-rw-r--r--drivers/base/power/main.c10
-rw-r--r--drivers/base/power/wakeirq.c2
-rw-r--r--drivers/base/power/wakeup.c4
-rw-r--r--drivers/pci/pci-driver.c14
-rw-r--r--drivers/pnp/driver.c2
-rw-r--r--drivers/usb/core/hcd-pci.c4
-rw-r--r--include/linux/acpi.h2
-rw-r--r--include/linux/pm.h8
-rw-r--r--kernel/power/hibernate.c6
-rw-r--r--kernel/power/suspend_test.c8
-rw-r--r--kernel/power/swap.c8
16 files changed, 146 insertions, 59 deletions
diff --git a/Documentation/admin-guide/pm/intel_uncore_frequency_scaling.rst b/Documentation/admin-guide/pm/intel_uncore_frequency_scaling.rst
new file mode 100644
index 000000000000..09169d935835
--- /dev/null
+++ b/Documentation/admin-guide/pm/intel_uncore_frequency_scaling.rst
@@ -0,0 +1,60 @@
+.. SPDX-License-Identifier: GPL-2.0
+.. include:: <isonum.txt>
+
+==============================
+Intel Uncore Frequency Scaling
+==============================
+
+:Copyright: |copy| 2022 Intel Corporation
+
+:Author: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
+
+Introduction
+------------
+
+The uncore can consume significant amount of power in Intel's Xeon servers based
+on the workload characteristics. To optimize the total power and improve overall
+performance, SoCs have internal algorithms for scaling uncore frequency. These
+algorithms monitor workload usage of uncore and set a desirable frequency.
+
+It is possible that users have different expectations of uncore performance and
+want to have control over it. The objective is similar to allowing users to set
+the scaling min/max frequencies via cpufreq sysfs to improve CPU performance.
+Users may have some latency sensitive workloads where they do not want any
+change to uncore frequency. Also, users may have workloads which require
+different core and uncore performance at distinct phases and they may want to
+use both cpufreq and the uncore scaling interface to distribute power and
+improve overall performance.
+
+Sysfs Interface
+---------------
+
+To control uncore frequency, a sysfs interface is provided in the directory:
+`/sys/devices/system/cpu/intel_uncore_frequency/`.
+
+There is one directory for each package and die combination as the scope of
+uncore scaling control is per die in multiple die/package SoCs or per
+package for single die per package SoCs. The name represents the
+scope of control. For example: 'package_00_die_00' is for package id 0 and
+die 0.
+
+Each package_*_die_* contains the following attributes:
+
+``initial_max_freq_khz``
+ Out of reset, this attribute represent the maximum possible frequency.
+ This is a read-only attribute. If users adjust max_freq_khz,
+ they can always go back to maximum using the value from this attribute.
+
+``initial_min_freq_khz``
+ Out of reset, this attribute represent the minimum possible frequency.
+ This is a read-only attribute. If users adjust min_freq_khz,
+ they can always go back to minimum using the value from this attribute.
+
+``max_freq_khz``
+ This attribute is used to set the maximum uncore frequency.
+
+``min_freq_khz``
+ This attribute is used to set the minimum uncore frequency.
+
+``current_freq_khz``
+ This attribute is used to get the current uncore frequency.
diff --git a/Documentation/admin-guide/pm/working-state.rst b/Documentation/admin-guide/pm/working-state.rst
index 5d2757e2de65..ee45887811ff 100644
--- a/Documentation/admin-guide/pm/working-state.rst
+++ b/Documentation/admin-guide/pm/working-state.rst
@@ -15,3 +15,4 @@ Working-State Power Management
cpufreq_drivers
intel_epb
intel-speed-select
+ intel_uncore_frequency_scaling
diff --git a/arch/x86/kernel/acpi/sleep.c b/arch/x86/kernel/acpi/sleep.c
index 1e97f944b47d..3b7f4cdbf2e0 100644
--- a/arch/x86/kernel/acpi/sleep.c
+++ b/arch/x86/kernel/acpi/sleep.c
@@ -15,6 +15,7 @@
#include <asm/desc.h>
#include <asm/cacheflush.h>
#include <asm/realmode.h>
+#include <asm/hypervisor.h>
#include <linux/ftrace.h>
#include "../../realmode/rm/wakeup.h"
@@ -140,9 +141,9 @@ static int __init acpi_sleep_setup(char *str)
acpi_realmode_flags |= 4;
#ifdef CONFIG_HIBERNATION
if (strncmp(str, "s4_hwsig", 8) == 0)
- acpi_check_s4_hw_signature(1);
+ acpi_check_s4_hw_signature = 1;
if (strncmp(str, "s4_nohwsig", 10) == 0)
- acpi_check_s4_hw_signature(0);
+ acpi_check_s4_hw_signature = 0;
#endif
if (strncmp(str, "nonvs", 5) == 0)
acpi_nvs_nosave();
@@ -160,3 +161,21 @@ static int __init acpi_sleep_setup(char *str)
}
__setup("acpi_sleep=", acpi_sleep_setup);
+
+#if defined(CONFIG_HIBERNATION) && defined(CONFIG_HYPERVISOR_GUEST)
+static int __init init_s4_sigcheck(void)
+{
+ /*
+ * If running on a hypervisor, honour the ACPI specification
+ * by default and trigger a clean reboot when the hardware
+ * signature in FACS is changed after hibernation.
+ */
+ if (acpi_check_s4_hw_signature == -1 &&
+ !hypervisor_is_type(X86_HYPER_NATIVE))
+ acpi_check_s4_hw_signature = 1;
+
+ return 0;
+}
+/* This must happen before acpi_init() which is a subsys initcall */
+arch_initcall(init_s4_sigcheck);
+#endif
diff --git a/drivers/acpi/sleep.c b/drivers/acpi/sleep.c
index d4fbea91ab6b..74e54fd44b8e 100644
--- a/drivers/acpi/sleep.c
+++ b/drivers/acpi/sleep.c
@@ -869,12 +869,7 @@ static inline void acpi_sleep_syscore_init(void) {}
#ifdef CONFIG_HIBERNATION
static unsigned long s4_hardware_signature;
static struct acpi_table_facs *facs;
-static int sigcheck = -1; /* Default behaviour is just to warn */
-
-void __init acpi_check_s4_hw_signature(int check)
-{
- sigcheck = check;
-}
+int acpi_check_s4_hw_signature = -1; /* Default behaviour is just to warn */
static int acpi_hibernation_begin(pm_message_t stage)
{
@@ -999,7 +994,7 @@ static void acpi_sleep_hibernate_setup(void)
hibernation_set_ops(old_suspend_ordering ?
&acpi_hibernation_ops_old : &acpi_hibernation_ops);
sleep_states[ACPI_STATE_S4] = 1;
- if (!sigcheck)
+ if (!acpi_check_s4_hw_signature)
return;
acpi_get_table(ACPI_SIG_FACS, 1, (struct acpi_table_header **)&facs);
@@ -1011,7 +1006,7 @@ static void acpi_sleep_hibernate_setup(void)
*/
s4_hardware_signature = facs->hardware_signature;
- if (sigcheck > 0) {
+ if (acpi_check_s4_hw_signature > 0) {
/*
* If we're actually obeying the ACPI specification
* then the signature is written out as part of the
diff --git a/drivers/base/power/domain.c b/drivers/base/power/domain.c
index 5db704f02e71..1ee878d126fd 100644
--- a/drivers/base/power/domain.c
+++ b/drivers/base/power/domain.c
@@ -636,6 +636,18 @@ static int genpd_power_off(struct generic_pm_domain *genpd, bool one_dev_on,
atomic_read(&genpd->sd_count) > 0)
return -EBUSY;
+ /*
+ * The children must be in their deepest (powered-off) states to allow
+ * the parent to be powered off. Note that, there's no need for
+ * additional locking, as powering on a child, requires the parent's
+ * lock to be acquired first.
+ */
+ list_for_each_entry(link, &genpd->parent_links, parent_node) {
+ struct generic_pm_domain *child = link->child;
+ if (child->state_idx < child->state_count - 1)
+ return -EBUSY;
+ }
+
list_for_each_entry(pdd, &genpd->dev_list, list_node) {
enum pm_qos_flags_status stat;
@@ -1073,6 +1085,13 @@ static void genpd_sync_power_off(struct generic_pm_domain *genpd, bool use_lock,
|| atomic_read(&genpd->sd_count) > 0)
return;
+ /* Check that the children are in their deepest (powered-off) state. */
+ list_for_each_entry(link, &genpd->parent_links, parent_node) {
+ struct generic_pm_domain *child = link->child;
+ if (child->state_idx < child->state_count - 1)
+ return;
+ }
+
/* Choose the deepest state when suspending */
genpd->state_idx = genpd->state_count - 1;
if (_genpd_power_off(genpd, false))
@@ -2058,9 +2077,9 @@ static int genpd_remove(struct generic_pm_domain *genpd)
kfree(link);
}
- genpd_debug_remove(genpd);
list_del(&genpd->gpd_list_node);
genpd_unlock(genpd);
+ genpd_debug_remove(genpd);
cancel_work_sync(&genpd->power_off_work);
if (genpd_is_cpu_domain(genpd))
free_cpumask_var(genpd->cpus);
@@ -2248,12 +2267,8 @@ int of_genpd_add_provider_simple(struct device_node *np,
/* Parse genpd OPP table */
if (genpd->set_performance_state) {
ret = dev_pm_opp_of_add_table(&genpd->dev);
- if (ret) {
- if (ret != -EPROBE_DEFER)
- dev_err(&genpd->dev, "Failed to add OPP table: %d\n",
- ret);
- return ret;
- }
+ if (ret)
+ return dev_err_probe(&genpd->dev, ret, "Failed to add OPP table\n");
/*
* Save table for faster processing while setting performance
@@ -2312,9 +2327,8 @@ int of_genpd_add_provider_onecell(struct device_node *np,
if (genpd->set_performance_state) {
ret = dev_pm_opp_of_add_table_indexed(&genpd->dev, i);
if (ret) {
- if (ret != -EPROBE_DEFER)
- dev_err(&genpd->dev, "Failed to add OPP table for index %d: %d\n",
- i, ret);
+ dev_err_probe(&genpd->dev, ret,
+ "Failed to add OPP table for index %d\n", i);
goto error;
}
@@ -2672,12 +2686,8 @@ static int __genpd_dev_pm_attach(struct device *dev, struct device *base_dev,
ret = genpd_add_device(pd, dev, base_dev);
mutex_unlock(&gpd_list_lock);
- if (ret < 0) {
- if (ret != -EPROBE_DEFER)
- dev_err(dev, "failed to add to PM domain %s: %d",
- pd->name, ret);
- return ret;
- }
+ if (ret < 0)
+ return dev_err_probe(dev, ret, "failed to add to PM domain %s\n", pd->name);
dev->pm_domain->detach = genpd_dev_pm_detach;
dev->pm_domain->sync = genpd_dev_pm_sync;
diff --git a/drivers/base/power/main.c b/drivers/base/power/main.c
index 08c8a69d7b81..c50139207794 100644
--- a/drivers/base/power/main.c
+++ b/drivers/base/power/main.c
@@ -485,7 +485,7 @@ static int dpm_run_callback(pm_callback_t cb, struct device *dev,
trace_device_pm_callback_start(dev, info, state.event);
error = cb(dev);
trace_device_pm_callback_end(dev, error);
- suspend_report_result(cb, error);
+ suspend_report_result(dev, cb, error);
initcall_debug_report(dev, calltime, cb, error);
@@ -1568,7 +1568,7 @@ static int legacy_suspend(struct device *dev, pm_message_t state,
trace_device_pm_callback_start(dev, info, state.event);
error = cb(dev, state);
trace_device_pm_callback_end(dev, error);
- suspend_report_result(cb, error);
+ suspend_report_result(dev, cb, error);
initcall_debug_report(dev, calltime, cb, error);
@@ -1855,7 +1855,7 @@ unlock:
device_unlock(dev);
if (ret < 0) {
- suspend_report_result(callback, ret);
+ suspend_report_result(dev, callback, ret);
pm_runtime_put(dev);
return ret;
}
@@ -1960,10 +1960,10 @@ int dpm_suspend_start(pm_message_t state)
}
EXPORT_SYMBOL_GPL(dpm_suspend_start);
-void __suspend_report_result(const char *function, void *fn, int ret)
+void __suspend_report_result(const char *function, struct device *dev, void *fn, int ret)
{
if (ret)
- pr_err("%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);
diff --git a/drivers/base/power/wakeirq.c b/drivers/base/power/wakeirq.c
index 0004db4a9d3b..d487a6bac630 100644
--- a/drivers/base/power/wakeirq.c
+++ b/drivers/base/power/wakeirq.c
@@ -289,7 +289,7 @@ EXPORT_SYMBOL_GPL(dev_pm_disable_wake_irq);
*
* Enables wakeirq conditionally. We need to enable wake-up interrupt
* lazily on the first rpm_suspend(). This is needed as the consumer device
- * starts in RPM_SUSPENDED state, and the the first pm_runtime_get() would
+ * starts in RPM_SUSPENDED state, and the first pm_runtime_get() would
* otherwise try to disable already disabled wakeirq. The wake-up interrupt
* starts disabled with IRQ_NOAUTOEN set.
*
diff --git a/drivers/base/power/wakeup.c b/drivers/base/power/wakeup.c
index 8666590201c9..a57d469676ca 100644
--- a/drivers/base/power/wakeup.c
+++ b/drivers/base/power/wakeup.c
@@ -587,7 +587,7 @@ static bool wakeup_source_not_registered(struct wakeup_source *ws)
* @ws: Wakeup source to handle.
*
* Update the @ws' statistics and, if @ws has just been activated, notify the PM
- * core of the event by incrementing the counter of of wakeup events being
+ * core of the event by incrementing the counter of the wakeup events being
* processed.
*/
static void wakeup_source_activate(struct wakeup_source *ws)
@@ -733,7 +733,7 @@ static void wakeup_source_deactivate(struct wakeup_source *ws)
/*
* Increment the counter of registered wakeup events and decrement the
- * couter of wakeup events in progress simultaneously.
+ * counter of wakeup events in progress simultaneously.
*/
cec = atomic_add_return(MAX_IN_PROGRESS, &combined_event_count);
trace_wakeup_source_deactivate(ws->name, cec);
diff --git a/drivers/pci/pci-driver.c b/drivers/pci/pci-driver.c
index 588588cfda48..415f7664b010 100644
--- a/drivers/pci/pci-driver.c
+++ b/drivers/pci/pci-driver.c
@@ -596,7 +596,7 @@ static int pci_legacy_suspend(struct device *dev, pm_message_t state)
int error;
error = drv->suspend(pci_dev, state);
- suspend_report_result(drv->suspend, error);
+ suspend_report_result(dev, drv->suspend, error);
if (error)
return error;
@@ -775,7 +775,7 @@ static int pci_pm_suspend(struct device *dev)
int error;
error = pm->suspend(dev);
- suspend_report_result(pm->suspend, error);
+ suspend_report_result(dev, pm->suspend, error);
if (error)
return error;
@@ -821,7 +821,7 @@ static int pci_pm_suspend_noirq(struct device *dev)
int error;
error = pm->suspend_noirq(dev);
- suspend_report_result(pm->suspend_noirq, error);
+ suspend_report_result(dev, pm->suspend_noirq, error);
if (error)
return error;
@@ -1010,7 +1010,7 @@ static int pci_pm_freeze(struct device *dev)
int error;
error = pm->freeze(dev);
- suspend_report_result(pm->freeze, error);
+ suspend_report_result(dev, pm->freeze, error);
if (error)
return error;
}
@@ -1030,7 +1030,7 @@ static int pci_pm_freeze_noirq(struct device *dev)
int error;
error = pm->freeze_noirq(dev);
- suspend_report_result(pm->freeze_noirq, error);
+ suspend_report_result(dev, pm->freeze_noirq, error);
if (error)
return error;
}
@@ -1116,7 +1116,7 @@ static int pci_pm_poweroff(struct device *dev)
int error;
error = pm->poweroff(dev);
- suspend_report_result(pm->poweroff, error);
+ suspend_report_result(dev, pm->poweroff, error);
if (error)
return error;
}
@@ -1154,7 +1154,7 @@ static int pci_pm_poweroff_noirq(struct device *dev)
int error;
error = pm->poweroff_noirq(dev);
- suspend_report_result(pm->poweroff_noirq, error);
+ suspend_report_result(dev, pm->poweroff_noirq, error);
if (error)
return error;
}
diff --git a/drivers/pnp/driver.c b/drivers/pnp/driver.c
index cc6757dfa3f1..c02e7bf643a6 100644
--- a/drivers/pnp/driver.c
+++ b/drivers/pnp/driver.c
@@ -171,7 +171,7 @@ static int __pnp_bus_suspend(struct device *dev, pm_message_t state)
if (pnp_drv->driver.pm && pnp_drv->driver.pm->suspend) {
error = pnp_drv->driver.pm->suspend(dev);
- suspend_report_result(pnp_drv->driver.pm->suspend, error);
+ suspend_report_result(dev, pnp_drv->driver.pm->suspend, error);
if (error)
return error;
}
diff --git a/drivers/usb/core/hcd-pci.c b/drivers/usb/core/hcd-pci.c
index d630cccd2e6e..dd44e37a454a 100644
--- a/drivers/usb/core/hcd-pci.c
+++ b/drivers/usb/core/hcd-pci.c
@@ -446,7 +446,7 @@ static int suspend_common(struct device *dev, bool do_wakeup)
HCD_WAKEUP_PENDING(hcd->shared_hcd))
return -EBUSY;
retval = hcd->driver->pci_suspend(hcd, do_wakeup);
- suspend_report_result(hcd->driver->pci_suspend, retval);
+ suspend_report_result(dev, hcd->driver->pci_suspend, retval);
/* Check again in case wakeup raced with pci_suspend */
if ((retval == 0 && do_wakeup && HCD_WAKEUP_PENDING(hcd)) ||
@@ -556,7 +556,7 @@ static int hcd_pci_suspend_noirq(struct device *dev)
dev_dbg(dev, "--> PCI %s\n",
pci_power_name(pci_dev->current_state));
} else {
- suspend_report_result(pci_prepare_to_sleep, retval);
+ suspend_report_result(dev, pci_prepare_to_sleep, retval);
return retval;
}
diff --git a/include/linux/acpi.h b/include/linux/acpi.h
index 6274758648e3..766dbcb82df1 100644
--- a/include/linux/acpi.h
+++ b/include/linux/acpi.h
@@ -526,7 +526,7 @@ acpi_status acpi_release_memory(acpi_handle handle, struct resource *res,
int acpi_resources_are_enforced(void);
#ifdef CONFIG_HIBERNATION
-void __init acpi_check_s4_hw_signature(int check);
+extern int acpi_check_s4_hw_signature;
#endif
#ifdef CONFIG_PM_SLEEP
diff --git a/include/linux/pm.h b/include/linux/pm.h
index f7d2be686359..e65b3ab28377 100644
--- a/include/linux/pm.h
+++ b/include/linux/pm.h
@@ -770,11 +770,11 @@ extern int dpm_suspend_late(pm_message_t state);
extern int dpm_suspend(pm_message_t state);
extern int dpm_prepare(pm_message_t state);
-extern void __suspend_report_result(const char *function, void *fn, int ret);
+extern void __suspend_report_result(const char *function, struct device *dev, void *fn, int ret);
-#define suspend_report_result(fn, ret) \
+#define suspend_report_result(dev, fn, ret) \
do { \
- __suspend_report_result(__func__, fn, ret); \
+ __suspend_report_result(__func__, dev, fn, ret); \
} while (0)
extern int device_pm_wait_for_dev(struct device *sub, struct device *dev);
@@ -814,7 +814,7 @@ static inline int dpm_suspend_start(pm_message_t state)
return 0;
}
-#define suspend_report_result(fn, ret) do {} while (0)
+#define suspend_report_result(dev, fn, ret) do {} while (0)
static inline int device_pm_wait_for_dev(struct device *a, struct device *b)
{
diff --git a/kernel/power/hibernate.c b/kernel/power/hibernate.c
index e6af502c2fd7..0ac805b753e5 100644
--- a/kernel/power/hibernate.c
+++ b/kernel/power/hibernate.c
@@ -689,8 +689,10 @@ static int load_image_and_restore(void)
lock_device_hotplug();
error = create_basic_memory_bitmaps();
- if (error)
+ if (error) {
+ swsusp_close(FMODE_READ | FMODE_EXCL);
goto Unlock;
+ }
error = swsusp_read(&flags);
swsusp_close(FMODE_READ | FMODE_EXCL);
@@ -1328,7 +1330,7 @@ static int __init resumedelay_setup(char *str)
int rc = kstrtouint(str, 0, &resume_delay);
if (rc)
- return rc;
+ pr_warn("resumedelay: bad option string '%s'\n", str);
return 1;
}
diff --git a/kernel/power/suspend_test.c b/kernel/power/suspend_test.c
index d20526c5be15..b663a97f5867 100644
--- a/kernel/power/suspend_test.c
+++ b/kernel/power/suspend_test.c
@@ -157,22 +157,22 @@ static int __init setup_test_suspend(char *value)
value++;
suspend_type = strsep(&value, ",");
if (!suspend_type)
- return 0;
+ return 1;
repeat = strsep(&value, ",");
if (repeat) {
if (kstrtou32(repeat, 0, &test_repeat_count_max))
- return 0;
+ return 1;
}
for (i = PM_SUSPEND_MIN; i < PM_SUSPEND_MAX; i++)
if (!strcmp(pm_labels[i], suspend_type)) {
test_state_label = pm_labels[i];
- return 0;
+ return 1;
}
printk(warn_bad_state, suspend_type);
- return 0;
+ return 1;
}
__setup("test_suspend", setup_test_suspend);
diff --git a/kernel/power/swap.c b/kernel/power/swap.c
index ad10359030a4..c51f5507b34f 100644
--- a/kernel/power/swap.c
+++ b/kernel/power/swap.c
@@ -89,7 +89,7 @@ struct swap_map_page_list {
struct swap_map_page_list *next;
};
-/**
+/*
* The swap_map_handle structure is used for handling swap in
* a file-alike way
*/
@@ -117,7 +117,7 @@ struct swsusp_header {
static struct swsusp_header *swsusp_header;
-/**
+/*
* The following functions are used for tracing the allocated
* swap pages, so that they can be freed in case of an error.
*/
@@ -171,7 +171,7 @@ static int swsusp_extents_insert(unsigned long swap_offset)
return 0;
}
-/**
+/*
* alloc_swapdev_block - allocate a swap page and register that it has
* been allocated, so that it can be freed in case of an error.
*/
@@ -190,7 +190,7 @@ sector_t alloc_swapdev_block(int swap)
return 0;
}
-/**
+/*
* free_all_swap_pages - free swap pages allocated for saving image data.
* It also frees the extents used to register which swap entries had been
* allocated.
generated by cgit v1.2.3 (git 2.39.1) at 2025-04-26 18:01:15 +0000