Hi,
On 10/23/2012 11:20 AM, Rajanikanth HV wrote:
>>> diff --git a/drivers/power/ab8500_fg.c b/drivers/power/ab8500_fg.c
>>> index bf02225..ff64dd4 100644
>>> --- a/drivers/power/ab8500_fg.c
>>> +++ b/drivers/power/ab8500_fg.c
>>> @@ -22,15 +22,16 @@
> [...]
>>>
>>> #define MILLI_TO_MICRO 1000
>>> #define FG_LSB_IN_MA 1627
>>> @@ -212,7 +213,6 @@ struct ab8500_fg {
>>> struct ab8500_fg_avg_cap avg_cap;
>>> struct ab8500 *parent;
>>> struct ab8500_gpadc *gpadc;
>>> - struct abx500_fg_platform_data *pdata;
>>> struct abx500_bm_data *bat;
>>> struct power_supply fg_psy;
>>> struct workqueue_struct *fg_wq;
>>> @@ -2416,6 +2416,8 @@ static int __devexit ab8500_fg_remove(struct
>>> platform_device *pdev)
>>> int ret = 0;
>>> struct ab8500_fg *di = platform_get_drvdata(pdev);
>>>
>>> + of_node_put(pdev->dev.of_node);
>>
>> This is wrong, the probe function doesn't increment the refcount of this
>> node, so you don't have to decrement it here.
> you hinted in one of your earlier comments as:
> "
> Also, if I'm not mistaken we have a leak here, because the refcount of
> these nodes is never decremented, not even in the driver remove routine.
> " while referring bmdevs_of_probe(..), nodes being referred,
> np and np_bat_supply
Whenever you call of_find_node_*(), or of_parse_phandle(), the returned
node has its refcount incremented, so of_node_put() should be called
after you are done with the node. In your previous patch, you called
of_find_node_by_phandle() multiple times, without calling of_node_put()
on the corresponding nodes.
In your last patch, you call of_parse_phandle() once, and then call
of_node_put() on the returned node: that's all you have to do. You don't
have to call of_node_put() on pdev->dev.of_node, because you never
incremented the refcount of that node.
--
Francesco
_______________________________________________
linaro-dev mailing list
linaro-dev@lists.linaro.org
http://lists.linaro.org/mailman/listinfo/linaro-dev
