[CPUFREQ] Fix ondemand vs suspend deadlock

Rootcaused the bug to a deadlock in cpufreq and ondemand. Due to non-existent
ordering between cpu_hotplug lock and dbs_mutex. Basically a race condition
between cpu_down() and do_dbs_timer().

cpu_down() flow:
* cpu_down() call for CPU 1
* Takes hot plug lock
* Calls pre down notifier
*     cpufreq notifier handler calls cpufreq_driver_target() which takes
      cpu_hotplug lock again. OK as cpu_hotplug lock is recursive in same
      process context
* CPU 1 goes down
* Calls post down notifier
*     cpufreq notifier handler calls ondemand event stop which takes dbs_mutex

So, cpu_hotplug lock is taken before dbs_mutex in this flow.

do_dbs_timer is triggerred by a periodic timer event.
It first takes dbs_mutex and then takes cpu_hotplug lock in
cpufreq_driver_target().
Note the reverse order here compared to above. So, if this timer event happens
at right moment during cpu_down, system will deadlok.

Attached patch fixes the issue for both ondemand and conservative.

Signed-off-by: Venkatesh Pallipadi <venkatesh.pallipadi@intel.com>
Signed-off-by: Dave Jones <davej@redhat.com>

2 files changed, 24 insertions, 0 deletions

diff --git a/drivers/cpufreq/cpufreq_conservative.c b/drivers/cpufreq/cpufreq_conservative.c
index e07a35487bde..8878a154ed43 100644
--- a/drivers/cpufreq/cpufreq_conservative.c
+++ b/drivers/cpufreq/cpufreq_conservative.c
@@ -72,6 +72,14 @@ static DEFINE_PER_CPU(struct cpu_dbs_info_s, cpu_dbs_info);
 
 static unsigned int dbs_enable; /* number of CPUs using this policy */
 
+/*
+ * DEADLOCK ALERT! There is a ordering requirement between cpu_hotplug
+ * lock and dbs_mutex. cpu_hotplug lock should always be held before
+ * dbs_mutex. If any function that can potentially take cpu_hotplug lock
+ * (like __cpufreq_driver_target()) is being called with dbs_mutex taken, then
+ * cpu_hotplug lock should be taken before that. Note that cpu_hotplug lock
+ * is recursive for the same process. -Venki
+ */
 static DEFINE_MUTEX     (dbs_mutex);
 static DECLARE_WORK     (dbs_work, do_dbs_timer, NULL);
 
@@ -414,12 +422,14 @@ static void dbs_check_cpu(int cpu)
 static void do_dbs_timer(void *data)
 { 
         int i;
+        lock_cpu_hotplug();
         mutex_lock(&dbs_mutex);
         for_each_online_cpu(i)
                 dbs_check_cpu(i);
         schedule_delayed_work(&dbs_work, 
                         usecs_to_jiffies(dbs_tuners_ins.sampling_rate));
         mutex_unlock(&dbs_mutex);
+        unlock_cpu_hotplug();
 } 
 
 static inline void dbs_timer_init(void)
@@ -514,6 +524,7 @@ static int cpufreq_governor_dbs(struct cpufreq_policy *policy,
                 break;
 
         case CPUFREQ_GOV_LIMITS:
+                lock_cpu_hotplug();
                 mutex_lock(&dbs_mutex);
                 if (policy->max < this_dbs_info->cur_policy->cur)
                         __cpufreq_driver_target(
@@ -524,6 +535,7 @@ static int cpufreq_governor_dbs(struct cpufreq_policy *policy,
                                         this_dbs_info->cur_policy,
                                         policy->min, CPUFREQ_RELATION_L);
                 mutex_unlock(&dbs_mutex);
+                unlock_cpu_hotplug();
                 break;
         }
         return 0;
diff --git a/drivers/cpufreq/cpufreq_ondemand.c b/drivers/cpufreq/cpufreq_ondemand.c
index 3e6ffcaa5af4..4d308410b60e 100644
--- a/drivers/cpufreq/cpufreq_ondemand.c
+++ b/drivers/cpufreq/cpufreq_ondemand.c
@@ -71,6 +71,14 @@ static DEFINE_PER_CPU(struct cpu_dbs_info_s, cpu_dbs_info);
 
 static unsigned int dbs_enable; /* number of CPUs using this policy */
 
+/*
+ * DEADLOCK ALERT! There is a ordering requirement between cpu_hotplug
+ * lock and dbs_mutex. cpu_hotplug lock should always be held before
+ * dbs_mutex. If any function that can potentially take cpu_hotplug lock
+ * (like __cpufreq_driver_target()) is being called with dbs_mutex taken, then
+ * cpu_hotplug lock should be taken before that. Note that cpu_hotplug lock
+ * is recursive for the same process. -Venki
+ */
 static DEFINE_MUTEX (dbs_mutex);
 static DECLARE_WORK     (dbs_work, do_dbs_timer, NULL);
 
@@ -363,12 +371,14 @@ static void dbs_check_cpu(int cpu)
 static void do_dbs_timer(void *data)
 {
         int i;
+        lock_cpu_hotplug();
         mutex_lock(&dbs_mutex);
         for_each_online_cpu(i)
                 dbs_check_cpu(i);
         queue_delayed_work(dbs_workq, &dbs_work,
                            usecs_to_jiffies(dbs_tuners_ins.sampling_rate));
         mutex_unlock(&dbs_mutex);
+        unlock_cpu_hotplug();
 }
 
 static inline void dbs_timer_init(void)
@@ -469,6 +479,7 @@ static int cpufreq_governor_dbs(struct cpufreq_policy *policy,
                 break;
 
         case CPUFREQ_GOV_LIMITS:
+                lock_cpu_hotplug();
                 mutex_lock(&dbs_mutex);
                 if (policy->max < this_dbs_info->cur_policy->cur)
                         __cpufreq_driver_target(
@@ -479,6 +490,7 @@ static int cpufreq_governor_dbs(struct cpufreq_policy *policy,
                                         this_dbs_info->cur_policy,
                                         policy->min, CPUFREQ_RELATION_L);
                 mutex_unlock(&dbs_mutex);
+                unlock_cpu_hotplug();
                 break;
         }
         return 0;