PM / devfreq: Don't fail devfreq_dev_release if not in list
authorLeonard Crestez <leonard.crestez@nxp.com>
Wed, 13 Nov 2019 23:21:31 +0000 (01:21 +0200)
committerChanwoo Choi <cw00.choi@samsung.com>
Mon, 9 Dec 2019 03:19:16 +0000 (12:19 +0900)
Right now devfreq_dev_release will print a warning and abort the rest of
the cleanup if the devfreq instance is not part of the global
devfreq_list. But this is a valid scenario, for example it can happen if
the governor can't be found or on any other init error that happens
after device_register.

Initialize devfreq->node to an empty list head in devfreq_add_device so
that list_del becomes a safe noop inside devfreq_dev_release and we can
continue the rest of the cleanup.

Signed-off-by: Leonard Crestez <leonard.crestez@nxp.com>
Reviewed-by: Matthias Kaehlcke <mka@chromium.org>
Reviewed-by: Chanwoo Choi <cw00.choi@samsung.com>
Signed-off-by: Chanwoo Choi <cw00.choi@samsung.com>
drivers/devfreq/devfreq.c

index 65a4b6c..59f9fa8 100644 (file)
@@ -637,11 +637,6 @@ static void devfreq_dev_release(struct device *dev)
        struct devfreq *devfreq = to_devfreq(dev);
 
        mutex_lock(&devfreq_list_lock);
-       if (IS_ERR(find_device_devfreq(devfreq->dev.parent))) {
-               mutex_unlock(&devfreq_list_lock);
-               dev_warn(&devfreq->dev, "releasing devfreq which doesn't exist\n");
-               return;
-       }
        list_del(&devfreq->node);
        mutex_unlock(&devfreq_list_lock);
 
@@ -696,6 +691,7 @@ struct devfreq *devfreq_add_device(struct device *dev,
        devfreq->dev.parent = dev;
        devfreq->dev.class = devfreq_class;
        devfreq->dev.release = devfreq_dev_release;
+       INIT_LIST_HEAD(&devfreq->node);
        devfreq->profile = profile;
        strncpy(devfreq->governor_name, governor_name, DEVFREQ_NAME_LEN);
        devfreq->previous_freq = profile->initial_freq;