On Wed, 2015-07-08 at 12:14 +0100, Mark Brown wrote: > On Wed, Jul 08, 2015 at 04:25:50PM +0800, Koro Chen wrote: > > > The driver does not free irq if snd_soc_register_codec fails. > > It does not return error when request irq failed, either. > > Fix this by using devm_request_threaded_irq(), and returns when error. > > Unfortunately this isn't safe... > > > - if (i2c->irq) > > - free_irq(i2c->irq, rt5645); > > - > > cancel_delayed_work_sync(&rt5645->jack_detect_work); > > This work item is queued up by the interrupt handler so we need to > unregister the interrupt before we cancel any pending work otherwise > it's possible that the interrupt may fire after we cancelled the work. Thank you for seeing this, I didn't notice the delayed work. Do you think I should use devm_request_threaded_irq(), and change free_irq to devm_free_irq in remove? Or I should keep the original request_thread_irq(), and just add a free_irq() during probe failed?
-- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to [email protected] More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/
