thermal: core: Fix race between zone registration and system suspend

[ Upstream commit cdf771ab476bd9acb0948f3088a277d5c3cacc6b ]

If the registration of a thermal zone takes place at the time when
system suspend is started, thermal_pm_notify() can run before the new
thermal zone is added to thermal_tz_list and its "suspended" flag will
not be set.  Consequently, if __thermal_zone_device_update() is called
for that thermal zone, it will not return early as expected which may
cause some destructive interference with the system suspend or resume
flow to occur.

To avoid that, make thermal_zone_init_complete() introduced previously
set the "suspended" flag for new thermal zones if it runs during system
suspend or resume.

Fixes: 4e814173a8c4 ("thermal: core: Fix thermal zone suspend-resume synchronization")
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Link: https://patch.msgid.link/8490245.NyiUUSuA9g@rjwysocki.net
Reviewed-by: Lukasz Luba <lukasz.luba@arm.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>

diff --git a/drivers/thermal/thermal_core.c b/drivers/thermal/thermal_core.c
index 91512a8cb49d958cb0b1ae5a71f71560269daff7..a674469adad192d1421e2ab58b0a6be8f6af7b56 100644 (file)
--- a/drivers/thermal/thermal_core.c
+++ b/drivers/thermal/thermal_core.c
@@ -40,6 +40,8 @@ static DEFINE_MUTEX(thermal_governor_lock);
 
 static struct thermal_governor *def_governor;
 
+static bool thermal_pm_suspended;
+
 /*
  * Governor section: set of functions to handle thermal governors
  *
@@ -1356,6 +1358,14 @@ static void thermal_zone_init_complete(struct thermal_zone_device *tz)
        mutex_lock(&tz->lock);
 
        tz->state &= ~TZ_STATE_FLAG_INIT;
+       /*
+        * If system suspend or resume is in progress at this point, the
+        * new thermal zone needs to be marked as suspended because
+        * thermal_pm_notify() has run already.
+        */
+       if (thermal_pm_suspended)
+               tz->state |= TZ_STATE_FLAG_SUSPENDED;
+
        __thermal_zone_device_update(tz, THERMAL_EVENT_UNSPECIFIED);
 
        mutex_unlock(&tz->lock);
@@ -1544,10 +1554,10 @@ thermal_zone_device_register_with_trips(const char *type,
                }
        }
 
-       mutex_unlock(&thermal_list_lock);
-
        thermal_zone_init_complete(tz);
 
+       mutex_unlock(&thermal_list_lock);
+
        thermal_notify_tz_create(tz);
 
        thermal_debug_tz_add(tz);
@@ -1768,6 +1778,8 @@ static int thermal_pm_notify(struct notifier_block *nb,
        case PM_SUSPEND_PREPARE:
                mutex_lock(&thermal_list_lock);
 
+               thermal_pm_suspended = true;
+
                list_for_each_entry(tz, &thermal_tz_list, node)
                        thermal_zone_pm_prepare(tz);
 
@@ -1778,6 +1790,8 @@ static int thermal_pm_notify(struct notifier_block *nb,
        case PM_POST_SUSPEND:
                mutex_lock(&thermal_list_lock);
 
+               thermal_pm_suspended = false;
+
                list_for_each_entry(tz, &thermal_tz_list, node)
                        thermal_zone_pm_complete(tz);