On Fri, May 22, 2020 at 07:07:31PM +0800, Dinghao Liu wrote: > When arizona_request_irq() returns an error code, a > pairing runtime PM usage counter decrement is needed > to keep the counter balanced. For error paths after > this function, things are the same. > > Signed-off-by: Dinghao Liu <dinghao....@zju.edu.cn> > ---
Good spot on the bug thank you. > drivers/extcon/extcon-arizona.c | 1 + > 1 file changed, 1 insertion(+) > > diff --git a/drivers/extcon/extcon-arizona.c b/drivers/extcon/extcon-arizona.c > index 7401733db08b..470bbc8e5089 100644 > --- a/drivers/extcon/extcon-arizona.c > +++ b/drivers/extcon/extcon-arizona.c > @@ -1744,6 +1744,7 @@ static int arizona_extcon_probe(struct platform_device > *pdev) > err_rise: > arizona_free_irq(arizona, jack_irq_rise, info); > err_gpio: > + pm_runtime_put(&pdev->dev); However, I don't think this works as a fix. Firstly, the err_gpio label is used before the call to pm_runtime_get_sync, this might be ok since pm_runtime_enable hasn't been called yet but probably better to add a new label for it. Secondly, following the err_hpdet error path will also result in a double put. In that case I don't think there is any reason why we need to put before calling input_device_register so it might just be simplest to move that put until after registering the input device. Thanks, Charles > gpiod_put(info->micd_pol_gpio); > err_register: > pm_runtime_disable(&pdev->dev); > -- > 2.17.1 >
