Re: [PATCH 04/14] cpufreq/amd-pstate: Overhaul locking
From: Dhananjay Ugwekar
Date: Wed Feb 12 2025 - 00:15:39 EST
On 2/12/2025 3:24 AM, Mario Limonciello wrote:
> On 2/10/2025 23:02, Dhananjay Ugwekar wrote:
>> On 2/7/2025 3:26 AM, Mario Limonciello wrote:
>>> From: Mario Limonciello <mario.limonciello@xxxxxxx>
>>>
>>> amd_pstate_cpu_boost_update() and refresh_frequency_limits() both
>>> update the policy state and have nothing to do with the amd-pstate
>>> driver itself.
>>>
>>> A global "limits" lock doesn't make sense because each CPU can have
>>> policies changed independently. Instead introduce locks into to the
>>> cpudata structure and lock each CPU independently.
>>>
>>> The remaining "global" driver lock is used to ensure that only one
>>> entity can change driver modes at a given time.
>>>
>>> Signed-off-by: Mario Limonciello <mario.limonciello@xxxxxxx>
>>> ---
>>> drivers/cpufreq/amd-pstate.c | 27 +++++++++++++++++----------
>>> drivers/cpufreq/amd-pstate.h | 2 ++
>>> 2 files changed, 19 insertions(+), 10 deletions(-)
>>>
>>> diff --git a/drivers/cpufreq/amd-pstate.c b/drivers/cpufreq/amd-pstate.c
>>> index 77bc6418731ee..dd230ed3b9579 100644
>>> --- a/drivers/cpufreq/amd-pstate.c
>>> +++ b/drivers/cpufreq/amd-pstate.c
>>> @@ -196,7 +196,6 @@ static inline int get_mode_idx_from_str(const char *str, size_t size)
>>> return -EINVAL;
>>> }
>>> -static DEFINE_MUTEX(amd_pstate_limits_lock);
>>> static DEFINE_MUTEX(amd_pstate_driver_lock);
>>> static u8 msr_get_epp(struct amd_cpudata *cpudata)
>>> @@ -283,6 +282,8 @@ static int msr_set_epp(struct amd_cpudata *cpudata, u8 epp)
>>> u64 value, prev;
>>> int ret;
>>> + lockdep_assert_held(&cpudata->lock);
>>
>> After making the perf_cached variable writes atomic, do we still need a cpudata->lock ?
>
> My concern was specifically that userspace could interact with multiple sysfs files that influence the atomic perf variable (and the HW) at the same time. So you would not have a deterministic behavior if they raced. But if you take the mutex on all the paths that this could happen it will be a FIFO.
I guess, the lock still wont guarantee the ordering right? It will just ensure that one thread executes
that code path for a specific CPU at a time. And do we even care about the ordering ? I'm having a hard
time thinking of a scenario where we'll need the lock. Can you or Gautham think of any such scenario?
>
>>
>> Regards,
>> Dhananjay
>>
>>> +
>>> value = prev = READ_ONCE(cpudata->cppc_req_cached);
>>> value &= ~AMD_CPPC_EPP_PERF_MASK;
>>> value |= FIELD_PREP(AMD_CPPC_EPP_PERF_MASK, epp);
>>> @@ -315,6 +316,8 @@ static int shmem_set_epp(struct amd_cpudata *cpudata, u8 epp)
>>> int ret;
>>> struct cppc_perf_ctrls perf_ctrls;
>>> + lockdep_assert_held(&cpudata->lock);
>>> +
>>> if (epp == cpudata->epp_cached)
>>> return 0;
>>> @@ -335,6 +338,8 @@ static int amd_pstate_set_energy_pref_index(struct cpufreq_policy *policy,
>>> struct amd_cpudata *cpudata = policy->driver_data;
>>> u8 epp;
>>> + guard(mutex)(&cpudata->lock);
>>> +
>>> if (!pref_index)
>>> epp = cpudata->epp_default;
>>> else
>>> @@ -750,7 +755,6 @@ static int amd_pstate_set_boost(struct cpufreq_policy *policy, int state)
>>> pr_err("Boost mode is not supported by this processor or SBIOS\n");
>>> return -EOPNOTSUPP;
>>> }
>>> - guard(mutex)(&amd_pstate_driver_lock);
>>> ret = amd_pstate_cpu_boost_update(policy, state);
>>> refresh_frequency_limits(policy);
>>> @@ -973,6 +977,9 @@ static int amd_pstate_cpu_init(struct cpufreq_policy *policy)
>>> cpudata->cpu = policy->cpu;
>>> + mutex_init(&cpudata->lock);
>>> + guard(mutex)(&cpudata->lock);
>>> +
>>> ret = amd_pstate_init_perf(cpudata);
>>> if (ret)
>>> goto free_cpudata1;
>>> @@ -1179,8 +1186,6 @@ static ssize_t store_energy_performance_preference(
>>> if (ret < 0)
>>> return -EINVAL;
>>> - guard(mutex)(&amd_pstate_limits_lock);
>>> -
>>> ret = amd_pstate_set_energy_pref_index(policy, ret);
>>> return ret ? ret : count;
>>> @@ -1353,8 +1358,10 @@ int amd_pstate_update_status(const char *buf, size_t size)
>>> if (mode_idx < 0 || mode_idx >= AMD_PSTATE_MAX)
>>> return -EINVAL;
>>> - if (mode_state_machine[cppc_state][mode_idx])
>>> + if (mode_state_machine[cppc_state][mode_idx]) {
>>> + guard(mutex)(&amd_pstate_driver_lock);
>>> return mode_state_machine[cppc_state][mode_idx](mode_idx);
>>> + }
>>> return 0;
>>> }
>>> @@ -1375,7 +1382,6 @@ static ssize_t status_store(struct device *a, struct device_attribute *b,
>>> char *p = memchr(buf, '\n', count);
>>> int ret;
>>> - guard(mutex)(&amd_pstate_driver_lock);
>>> ret = amd_pstate_update_status(buf, p ? p - buf : count);
>>> return ret < 0 ? ret : count;
>>> @@ -1472,6 +1478,9 @@ static int amd_pstate_epp_cpu_init(struct cpufreq_policy *policy)
>>> cpudata->cpu = policy->cpu;
>>> + mutex_init(&cpudata->lock);
>>> + guard(mutex)(&cpudata->lock);
>>> +
>>> ret = amd_pstate_init_perf(cpudata);
>>> if (ret)
>>> goto free_cpudata1;
>>> @@ -1558,6 +1567,8 @@ static int amd_pstate_epp_update_limit(struct cpufreq_policy *policy)
>>> union perf_cached perf;
>>> u8 epp;
>>> + guard(mutex)(&cpudata->lock);
>>> +
>>> amd_pstate_update_min_max_limit(policy);
>>> if (cpudata->policy == CPUFREQ_POLICY_PERFORMANCE)
>>> @@ -1646,8 +1657,6 @@ static int amd_pstate_epp_cpu_offline(struct cpufreq_policy *policy)
>>> if (cpudata->suspended)
>>> return 0;
>>> - guard(mutex)(&amd_pstate_limits_lock);
>>> -
>>> if (trace_amd_pstate_epp_perf_enabled()) {
>>> trace_amd_pstate_epp_perf(cpudata->cpu, perf.highest_perf,
>>> AMD_CPPC_EPP_BALANCE_POWERSAVE,
>>> @@ -1684,8 +1693,6 @@ static int amd_pstate_epp_resume(struct cpufreq_policy *policy)
>>> struct amd_cpudata *cpudata = policy->driver_data;
>>> if (cpudata->suspended) {
>>> - guard(mutex)(&amd_pstate_limits_lock);
>>> -
>>> /* enable amd pstate from suspend state*/
>>> amd_pstate_epp_reenable(policy);
>>> diff --git a/drivers/cpufreq/amd-pstate.h b/drivers/cpufreq/amd-pstate.h
>>> index a140704b97430..6d776c3e5712a 100644
>>> --- a/drivers/cpufreq/amd-pstate.h
>>> +++ b/drivers/cpufreq/amd-pstate.h
>>> @@ -96,6 +96,8 @@ struct amd_cpudata {
>>> bool boost_supported;
>>> bool hw_prefcore;
>>> + struct mutex lock;
>>> +
>>> /* EPP feature related attributes*/
>>> u8 epp_cached;
>>> u32 policy;
>>
>