[v1] AMD Pstate Driver Core Performance Boost

[PATCH 3/7] cpufreq: amd-pstate: implement cpb_boost sysfs entry for boost control

Posted by Perry Yuan 1 year, 10 months ago

From: Perry Yuan <Perry.Yuan@amd.com>

With this new sysfs entry `cpb_boost`created, user can change CPU boost
state dynamically under `active` and `passive` modes.
And the highest perf and frequency will also be updated as the boost
state changing.

0: check current boost state
cat /sys/devices/system/cpu/amd_pstate/cpb_boost

1: disable CPU boost
sudo bash -c "echo 0 > /sys/devices/system/cpu/amd_pstate/cpb_boost"

2: enable CPU boost
sudo bash -c "echo 1 > /sys/devices/system/cpu/amd_pstate/cpb_boost"

Link: https://bugzilla.kernel.org/show_bug.cgi?id=217931
Link: https://bugzilla.kernel.org/show_bug.cgi?id=217618

Signed-off-by: Perry Yuan <Perry.Yuan@amd.com>
---
 drivers/cpufreq/amd-pstate.c | 97 ++++++++++++++++++++++++++++++++++++
 1 file changed, 97 insertions(+)

diff --git a/drivers/cpufreq/amd-pstate.c b/drivers/cpufreq/amd-pstate.c
index 0dc9124140d4..b37bea7440b9 100644
--- a/drivers/cpufreq/amd-pstate.c
+++ b/drivers/cpufreq/amd-pstate.c
@@ -1036,6 +1036,101 @@ static ssize_t status_store(struct device *a, struct device_attribute *b,
 	return ret < 0 ? ret : count;
 }
 
+static int amd_cpu_boost_update(struct amd_cpudata *cpudata, u32 on)
+{
+	struct cpufreq_policy *policy = cpufreq_cpu_acquire(cpudata->cpu);
+	struct cppc_perf_ctrls perf_ctrls;
+	u32 highest_perf, nominal_perf;
+	int ret;
+
+	if (!policy)
+		return -ENODATA;
+
+	highest_perf = READ_ONCE(cpudata->highest_perf);
+	nominal_perf = READ_ONCE(cpudata->nominal_perf);
+
+	if (boot_cpu_has(X86_FEATURE_CPPC)) {
+		u64 value = READ_ONCE(cpudata->cppc_req_cached);
+
+		value &= ~GENMASK_ULL(7, 0);
+		value |= on ? highest_perf : nominal_perf;
+		WRITE_ONCE(cpudata->cppc_req_cached, value);
+
+		wrmsrl_on_cpu(cpudata->cpu, MSR_AMD_CPPC_REQ, value);
+
+	} else {
+		perf_ctrls.max_perf = on ? highest_perf : nominal_perf;
+		ret = cppc_set_epp_perf(cpudata->cpu, &perf_ctrls, 1);
+		if (ret) {
+			pr_debug("failed to set energy perf value (%d)\n", ret);
+			return ret;
+		}
+	}
+
+	if (on)
+		policy->cpuinfo.max_freq = cpudata->max_freq;
+	else
+		policy->cpuinfo.max_freq = cpudata->nominal_freq;
+
+	policy->max = policy->cpuinfo.max_freq;
+
+	if (cppc_state == AMD_PSTATE_PASSIVE) {
+		ret = freq_qos_update_request(&cpudata->req[1],
+				      policy->cpuinfo.max_freq);
+	}
+
+	cpufreq_cpu_release(policy);
+
+	return ret;
+}
+
+static ssize_t cpb_boost_show(struct device *dev,
+			   struct device_attribute *attr, char *buf)
+{
+	return sysfs_emit(buf, "%u\n", global.cpb_boost);
+}
+
+static ssize_t cpb_boost_store(struct device *dev, struct device_attribute *b,
+			    const char *buf, size_t count)
+{
+	bool new_state;
+	ssize_t ret;
+	int cpu;
+
+	mutex_lock(&amd_pstate_driver_lock);
+	if (!global.cpb_supported) {
+		pr_err("Boost mode is not supported by this processor or SBIOS\n");
+		return -EINVAL;
+	}
+
+	ret = kstrtobool(buf, &new_state);
+	if (ret)
+		return -EINVAL;
+
+	global.cpb_boost = !!new_state;
+
+	for_each_possible_cpu(cpu) {
+
+		struct cpufreq_policy *policy = cpufreq_cpu_get(cpu);
+		struct amd_cpudata *cpudata = policy->driver_data;
+
+		if (!cpudata) {
+			pr_err("cpudata is NULL\n");
+			ret = -ENODATA;
+			cpufreq_cpu_put(policy);
+			goto err_exit;
+		}
+
+		amd_cpu_boost_update(cpudata, global.cpb_boost);
+		refresh_frequency_limits(policy);
+		cpufreq_cpu_put(policy);
+	}
+
+err_exit:
+	mutex_unlock(&amd_pstate_driver_lock);
+	return ret < 0 ? ret : count;
+}
+
 cpufreq_freq_attr_ro(amd_pstate_max_freq);
 cpufreq_freq_attr_ro(amd_pstate_lowest_nonlinear_freq);
 
@@ -1043,6 +1138,7 @@ cpufreq_freq_attr_ro(amd_pstate_highest_perf);
 cpufreq_freq_attr_rw(energy_performance_preference);
 cpufreq_freq_attr_ro(energy_performance_available_preferences);
 static DEVICE_ATTR_RW(status);
+static DEVICE_ATTR_RW(cpb_boost);
 
 static struct freq_attr *amd_pstate_attr[] = {
 	&amd_pstate_max_freq,
@@ -1062,6 +1158,7 @@ static struct freq_attr *amd_pstate_epp_attr[] = {
 
 static struct attribute *pstate_global_attributes[] = {
 	&dev_attr_status.attr,
+	&dev_attr_cpb_boost.attr,
 	NULL
 };
 
-- 
2.34.1

Re: [PATCH 3/7] cpufreq: amd-pstate: implement cpb_boost sysfs entry for boost control

Posted by Mario Limonciello 1 year, 10 months ago

On 1/26/2024 02:08, Perry Yuan wrote:
> From: Perry Yuan <Perry.Yuan@amd.com>
> 
> With this new sysfs entry `cpb_boost`created, user can change CPU boost
> state dynamically under `active` and `passive` modes.

What about guided mode?

> And the highest perf and frequency will also be updated as the boost
> state changing.
> 
> 0: check current boost state
> cat /sys/devices/system/cpu/amd_pstate/cpb_boost
> 
> 1: disable CPU boost
> sudo bash -c "echo 0 > /sys/devices/system/cpu/amd_pstate/cpb_boost"
> 
> 2: enable CPU boost
> sudo bash -c "echo 1 > /sys/devices/system/cpu/amd_pstate/cpb_boost"
> 
> Link: https://bugzilla.kernel.org/show_bug.cgi?id=217931
> Link: https://bugzilla.kernel.org/show_bug.cgi?id=217618
> 
> Signed-off-by: Perry Yuan <Perry.Yuan@amd.com>
> ---
>   drivers/cpufreq/amd-pstate.c | 97 ++++++++++++++++++++++++++++++++++++
>   1 file changed, 97 insertions(+)
> 
> diff --git a/drivers/cpufreq/amd-pstate.c b/drivers/cpufreq/amd-pstate.c
> index 0dc9124140d4..b37bea7440b9 100644
> --- a/drivers/cpufreq/amd-pstate.c
> +++ b/drivers/cpufreq/amd-pstate.c
> @@ -1036,6 +1036,101 @@ static ssize_t status_store(struct device *a, struct device_attribute *b,
>   	return ret < 0 ? ret : count;
>   }
>   
> +static int amd_cpu_boost_update(struct amd_cpudata *cpudata, u32 on)
> +{
> +	struct cpufreq_policy *policy = cpufreq_cpu_acquire(cpudata->cpu);
> +	struct cppc_perf_ctrls perf_ctrls;
> +	u32 highest_perf, nominal_perf;
> +	int ret;
> +
> +	if (!policy)
> +		return -ENODATA;
> +
> +	highest_perf = READ_ONCE(cpudata->highest_perf);
> +	nominal_perf = READ_ONCE(cpudata->nominal_perf);
> +
> +	if (boot_cpu_has(X86_FEATURE_CPPC)) {
> +		u64 value = READ_ONCE(cpudata->cppc_req_cached);
> +
> +		value &= ~GENMASK_ULL(7, 0);

I think it's better to use the macro AMD_CPPC_MAX_PERF in this case.
The less magic masks in the code, the easier it is to follow.

> +		value |= on ? highest_perf : nominal_perf;
> +		WRITE_ONCE(cpudata->cppc_req_cached, value);
> +
> +		wrmsrl_on_cpu(cpudata->cpu, MSR_AMD_CPPC_REQ, value);
> +
> +	} else {
> +		perf_ctrls.max_perf = on ? highest_perf : nominal_perf;
> +		ret = cppc_set_epp_perf(cpudata->cpu, &perf_ctrls, 1);
> +		if (ret) {
> +			pr_debug("failed to set energy perf value (%d)\n", ret);
> +			return ret;
> +		}
> +	}
> +
> +	if (on)
> +		policy->cpuinfo.max_freq = cpudata->max_freq;
> +	else
> +		policy->cpuinfo.max_freq = cpudata->nominal_freq;
> +
> +	policy->max = policy->cpuinfo.max_freq;
> +
> +	if (cppc_state == AMD_PSTATE_PASSIVE) {
> +		ret = freq_qos_update_request(&cpudata->req[1],
> +				      policy->cpuinfo.max_freq);
> +	}
> +
> +	cpufreq_cpu_release(policy);
> +
> +	return ret;
> +}
> +
> +static ssize_t cpb_boost_show(struct device *dev,
> +			   struct device_attribute *attr, char *buf)
> +{
> +	return sysfs_emit(buf, "%u\n", global.cpb_boost);
> +}
> +
> +static ssize_t cpb_boost_store(struct device *dev, struct device_attribute *b,
> +			    const char *buf, size_t count)
> +{
> +	bool new_state;
> +	ssize_t ret;
> +	int cpu;
> +
> +	mutex_lock(&amd_pstate_driver_lock);
> +	if (!global.cpb_supported) {
> +		pr_err("Boost mode is not supported by this processor or SBIOS\n");
> +		return -EINVAL;
> +	}
> +
> +	ret = kstrtobool(buf, &new_state);
> +	if (ret)
> +		return -EINVAL;
> +
> +	global.cpb_boost = !!new_state;
> +
> +	for_each_possible_cpu(cpu) {
> +
> +		struct cpufreq_policy *policy = cpufreq_cpu_get(cpu);
> +		struct amd_cpudata *cpudata = policy->driver_data;
> +
> +		if (!cpudata) {
> +			pr_err("cpudata is NULL\n");
> +			ret = -ENODATA;
> +			cpufreq_cpu_put(policy);
> +			goto err_exit;
> +		}
> +
> +		amd_cpu_boost_update(cpudata, global.cpb_boost);
> +		refresh_frequency_limits(policy);
> +		cpufreq_cpu_put(policy);
> +	}
> +
> +err_exit:
> +	mutex_unlock(&amd_pstate_driver_lock);
> +	return ret < 0 ? ret : count;
> +}
> +
>   cpufreq_freq_attr_ro(amd_pstate_max_freq);
>   cpufreq_freq_attr_ro(amd_pstate_lowest_nonlinear_freq);
>   
> @@ -1043,6 +1138,7 @@ cpufreq_freq_attr_ro(amd_pstate_highest_perf);
>   cpufreq_freq_attr_rw(energy_performance_preference);
>   cpufreq_freq_attr_ro(energy_performance_available_preferences);
>   static DEVICE_ATTR_RW(status);
> +static DEVICE_ATTR_RW(cpb_boost);
>   
>   static struct freq_attr *amd_pstate_attr[] = {
>   	&amd_pstate_max_freq,
> @@ -1062,6 +1158,7 @@ static struct freq_attr *amd_pstate_epp_attr[] = {
>   
>   static struct attribute *pstate_global_attributes[] = {
>   	&dev_attr_status.attr,
> +	&dev_attr_cpb_boost.attr,
>   	NULL
>   };
>

[PATCH 1/7] cpufreq: amd-pstate: remove set_boost callback for passive mode
[PATCH 2/7] cpufreq: amd-pstate: initialize new core precision boost state
[PATCH 3/7] cpufreq: amd-pstate: implement cpb_boost sysfs entry for boost control
[PATCH 4/7] cpufreq: amd-pstate: fix max_perf calculation for amd_get_max_freq()
[PATCH 5/7] cpufreq: amd-pstate: fix the MSR highest perf will be reset issue while cpb boost off
[PATCH 6/7] cpufreq:amd-pstate: add suspend and resume callback for passive mode
[PATCH 7/7] Documentation: cpufreq: amd-pstate: introduce the new cpu boost control method