diff --git a/Documentation/kernel-parameters.txt b/Documentation/kernel-parameters.txt
index eacb2e0397ae3f4e80428ef239d5b28faaf622b3..10b8cc1bda8df74b3ad523c56b39eb9e4d7a3cec 100644
--- a/Documentation/kernel-parameters.txt
+++ b/Documentation/kernel-parameters.txt
@@ -1446,6 +1446,15 @@ bytes respectively. Such letter suffixes can also be entirely omitted.
disable
Do not enable intel_pstate as the default
scaling driver for the supported processors
+ force
+ Enable intel_pstate on systems that prohibit it by default
+ in favor of acpi-cpufreq. Forcing the intel_pstate driver
+ instead of acpi-cpufreq may disable platform features, such
+ as thermal controls and power capping, that rely on ACPI
+ P-States information being indicated to OSPM and therefore
+ should be used with caution. This option does not work with
+ processors that aren't supported by the intel_pstate driver
+ or on platforms that use pcc-cpufreq instead of acpi-cpufreq.
no_hwp
Do not enable hardware P state control (HWP)
if available.
diff --git a/MAINTAINERS b/MAINTAINERS
index c8927bc7748e54842826a0181506c6bae3c2c85e..1c63cfecc6bfa5ab66e03c410ab5a24a30ab62b6 100644
--- a/MAINTAINERS
+++ b/MAINTAINERS
@@ -4869,6 +4869,12 @@ T: git git://git.kernel.org/pub/scm/linux/kernel/git/lenb/linux.git
S: Supported
F: drivers/idle/intel_idle.c
+INTEL PSTATE DRIVER
+M: Kristen Carlson Accardi <kristen@linux.intel.com>
+L: linux-pm@vger.kernel.org
+S: Supported
+F: drivers/cpufreq/intel_pstate.c
+
INTEL FRAMEBUFFER DRIVER (excluding 810 and 815)
M: Maik Broemme <mbroemme@plusserver.de>
L: linux-fbdev@vger.kernel.org
diff --git a/drivers/base/power/opp.c b/drivers/base/power/opp.c
index 1bbef8e838e75d6d2db27e1299ec76d2a8373038..d24dd614a0bd3208c0d88fd79c35397db5be8a71 100644
--- a/drivers/base/power/opp.c
+++ b/drivers/base/power/opp.c
@@ -84,7 +84,11 @@ struct dev_pm_opp {
*
* This is an internal data structure maintaining the link to opps attached to
* a device. This structure is not meant to be shared to users as it is
- * meant for book keeping and private to OPP library
+ * meant for book keeping and private to OPP library.
+ *
+ * Because the opp structures can be used from both rcu and srcu readers, we
+ * need to wait for the grace period of both of them before freeing any
+ * resources. And so we have used kfree_rcu() from within call_srcu() handlers.
*/
struct device_opp {
struct list_head node;
@@ -382,12 +386,34 @@ struct dev_pm_opp *dev_pm_opp_find_freq_floor(struct device *dev,
}
EXPORT_SYMBOL_GPL(dev_pm_opp_find_freq_floor);
+static struct device_opp *add_device_opp(struct device *dev)
+{
+ struct device_opp *dev_opp;
+
+ /*
+ * Allocate a new device OPP table. In the infrequent case where a new
+ * device is needed to be added, we pay this penalty.
+ */
+ dev_opp = kzalloc(sizeof(*dev_opp), GFP_KERNEL);
+ if (!dev_opp)
+ return NULL;
+
+ dev_opp->dev = dev;
+ srcu_init_notifier_head(&dev_opp->srcu_head);
+ INIT_LIST_HEAD(&dev_opp->opp_list);
+
+ /* Secure the device list modification */
+ list_add_rcu(&dev_opp->node, &dev_opp_list);
+ return dev_opp;
+}
+
static int dev_pm_opp_add_dynamic(struct device *dev, unsigned long freq,
unsigned long u_volt, bool dynamic)
{
struct device_opp *dev_opp = NULL;
struct dev_pm_opp *opp, *new_opp;
struct list_head *head;
+ int ret;
/* allocate new OPP node */
new_opp = kzalloc(sizeof(*new_opp), GFP_KERNEL);
@@ -408,27 +434,12 @@ static int dev_pm_opp_add_dynamic(struct device *dev, unsigned long freq,
/* Check for existing list for 'dev' */
dev_opp = find_device_opp(dev);
if (IS_ERR(dev_opp)) {
- /*
- * Allocate a new device OPP table. In the infrequent case
- * where a new device is needed to be added, we pay this
- * penalty.
- */
- dev_opp = kzalloc(sizeof(struct device_opp), GFP_KERNEL);
+ dev_opp = add_device_opp(dev);
if (!dev_opp) {
- mutex_unlock(&dev_opp_list_lock);
- kfree(new_opp);
- dev_warn(dev,
- "%s: Unable to create device OPP structure\n",
- __func__);
- return -ENOMEM;
+ ret = -ENOMEM;
+ goto free_opp;
}
- dev_opp->dev = dev;
- srcu_init_notifier_head(&dev_opp->srcu_head);
- INIT_LIST_HEAD(&dev_opp->opp_list);
-
- /* Secure the device list modification */
- list_add_rcu(&dev_opp->node, &dev_opp_list);
head = &dev_opp->opp_list;
goto list_add;
}
@@ -447,15 +458,13 @@ static int dev_pm_opp_add_dynamic(struct device *dev, unsigned long freq,
/* Duplicate OPPs ? */
if (new_opp->rate == opp->rate) {
- int ret = opp->available && new_opp->u_volt == opp->u_volt ?
+ ret = opp->available && new_opp->u_volt == opp->u_volt ?
0 : -EEXIST;
dev_warn(dev, "%s: duplicate OPPs detected. Existing: freq: %lu, volt: %lu, enabled: %d. New: freq: %lu, volt: %lu, enabled: %d\n",
__func__, opp->rate, opp->u_volt, opp->available,
new_opp->rate, new_opp->u_volt, new_opp->available);
- mutex_unlock(&dev_opp_list_lock);
- kfree(new_opp);
- return ret;
+ goto free_opp;
}
list_add:
@@ -469,6 +478,11 @@ list_add:
*/
srcu_notifier_call_chain(&dev_opp->srcu_head, OPP_EVENT_ADD, new_opp);
return 0;
+
+free_opp:
+ mutex_unlock(&dev_opp_list_lock);
+ kfree(new_opp);
+ return ret;
}
/**
@@ -511,10 +525,11 @@ static void kfree_device_rcu(struct rcu_head *head)
{
struct device_opp *device_opp = container_of(head, struct device_opp, rcu_head);
- kfree(device_opp);
+ kfree_rcu(device_opp, rcu_head);
}
-void __dev_pm_opp_remove(struct device_opp *dev_opp, struct dev_pm_opp *opp)
+static void __dev_pm_opp_remove(struct device_opp *dev_opp,
+ struct dev_pm_opp *opp)
{
/*
* Notify the changes in the availability of the operable
@@ -592,7 +607,7 @@ EXPORT_SYMBOL_GPL(dev_pm_opp_remove);
static int opp_set_availability(struct device *dev, unsigned long freq,
bool availability_req)
{
- struct device_opp *tmp_dev_opp, *dev_opp = ERR_PTR(-ENODEV);
+ struct device_opp *dev_opp;
struct dev_pm_opp *new_opp, *tmp_opp, *opp = ERR_PTR(-ENODEV);
int r = 0;
@@ -606,12 +621,7 @@ static int opp_set_availability(struct device *dev, unsigned long freq,
mutex_lock(&dev_opp_list_lock);
/* Find the device_opp */
- list_for_each_entry(tmp_dev_opp, &dev_opp_list, node) {
- if (dev == tmp_dev_opp->dev) {
- dev_opp = tmp_dev_opp;
- break;
- }
- }
+ dev_opp = find_device_opp(dev);
if (IS_ERR(dev_opp)) {
r = PTR_ERR(dev_opp);
dev_warn(dev, "%s: Device OPP not found (%d)\n", __func__, r);
diff --git a/drivers/cpufreq/intel_pstate.c b/drivers/cpufreq/intel_pstate.c
index 1405b393c93d90f68ac77f3ba9e23e223bcb80a3..742eefba12c2101d8bc8e035958d7bb59596ef93 100644
--- a/drivers/cpufreq/intel_pstate.c
+++ b/drivers/cpufreq/intel_pstate.c
@@ -199,7 +199,14 @@ static signed int pid_calc(struct _pid *pid, int32_t busy)
pid->integral += fp_error;
- /* limit the integral term */
+ /*
+ * We limit the integral here so that it will never
+ * get higher than 30. This prevents it from becoming
+ * too large an input over long periods of time and allows
+ * it to get factored out sooner.
+ *
+ * The value of 30 was chosen through experimentation.
+ */
integral_limit = int_tofp(30);
if (pid->integral > integral_limit)
pid->integral = integral_limit;
@@ -616,6 +623,11 @@ static void intel_pstate_get_min_max(struct cpudata *cpu, int *min, int *max)
if (limits.no_turbo || limits.turbo_disabled)
max_perf = cpu->pstate.max_pstate;
+ /*
+ * performance can be limited by user through sysfs, by cpufreq
+ * policy, or by cpu specific default values determined through
+ * experimentation.
+ */
max_perf_adj = fp_toint(mul_fp(int_tofp(max_perf), limits.max_perf));
*max = clamp_t(int, max_perf_adj,
cpu->pstate.min_pstate, cpu->pstate.turbo_pstate);
@@ -717,11 +729,29 @@ static inline int32_t intel_pstate_get_scaled_busy(struct cpudata *cpu)
u32 duration_us;
u32 sample_time;
+ /*
+ * core_busy is the ratio of actual performance to max
+ * max_pstate is the max non turbo pstate available
+ * current_pstate was the pstate that was requested during
+ * the last sample period.
+ *
+ * We normalize core_busy, which was our actual percent
+ * performance to what we requested during the last sample
+ * period. The result will be a percentage of busy at a
+ * specified pstate.
+ */
core_busy = cpu->sample.core_pct_busy;
max_pstate = int_tofp(cpu->pstate.max_pstate);
current_pstate = int_tofp(cpu->pstate.current_pstate);
core_busy = mul_fp(core_busy, div_fp(max_pstate, current_pstate));
+ /*
+ * Since we have a deferred timer, it will not fire unless
+ * we are in C0. So, determine if the actual elapsed time
+ * is significantly greater (3x) than our sample interval. If it
+ * is, then we were idle for a long enough period of time
+ * to adjust our busyness.
+ */
sample_time = pid_params.sample_rate_ms * USEC_PER_MSEC;
duration_us = (u32) ktime_us_delta(cpu->sample.time,
cpu->last_sample_time);
@@ -948,6 +978,7 @@ static struct cpufreq_driver intel_pstate_driver = {
static int __initdata no_load;
static int __initdata no_hwp;
+static unsigned int force_load;
static int intel_pstate_msrs_not_valid(void)
{
@@ -1094,7 +1125,8 @@ static bool intel_pstate_platform_pwr_mgmt_exists(void)
case PSS:
return intel_pstate_no_acpi_pss();
case PPC:
- return intel_pstate_has_acpi_ppc();
+ return intel_pstate_has_acpi_ppc() &&
+ (!force_load);
}
}
@@ -1175,6 +1207,8 @@ static int __init intel_pstate_setup(char *str)
no_load = 1;
if (!strcmp(str, "no_hwp"))
no_hwp = 1;
+ if (!strcmp(str, "force"))
+ force_load = 1;
return 0;
}
early_param("intel_pstate", intel_pstate_setup);
diff --git a/tools/power/cpupower/utils/cpuidle-info.c b/tools/power/cpupower/utils/cpuidle-info.c
index 458d69b444ad7d7a075de31a0b1fa8cb40ba8c77..75e66de7e7a7fc9934ec005667c6857cff42ff04 100644
--- a/tools/power/cpupower/utils/cpuidle-info.c
+++ b/tools/power/cpupower/utils/cpuidle-info.c
@@ -22,13 +22,13 @@
static void cpuidle_cpu_output(unsigned int cpu, int verbose)
{
- int idlestates, idlestate;
+ unsigned int idlestates, idlestate;
char *tmp;
printf(_ ("Analyzing CPU %d:\n"), cpu);
idlestates = sysfs_get_idlestate_count(cpu);
- if (idlestates < 1) {
+ if (idlestates == 0) {
printf(_("CPU %u: No idle states\n"), cpu);
return;
}
@@ -100,10 +100,10 @@ static void cpuidle_general_output(void)
static void proc_cpuidle_cpu_output(unsigned int cpu)
{
long max_allowed_cstate = 2000000000;
- int cstate, cstates;
+ unsigned int cstate, cstates;
cstates = sysfs_get_idlestate_count(cpu);
- if (cstates < 1) {
+ if (cstates == 0) {
printf(_("CPU %u: No C-states info\n"), cpu);
return;
}