Messages in this thread | | | Date | Fri, 21 Nov 2008 12:05:20 -0500 | From | Dmitry Torokhov <> | Subject | Re: [patch 08/14] input: PCAP2 based touchscreen driver |
| |
On Fri, Nov 21, 2008 at 05:04:11PM +0100, stefan@datenfreihafen.org wrote: > Touchscreen driver based on the PCAP2 multi function device. > > Signed-off-by: Daniel Ribeiro <drwyrm@gmail.com> > > + > +static int __devexit pcap_ts_remove(struct platform_device *pdev) > +{ > + ezx_pcap_unregister_event(PCAP_IRQ_TS); > + > + del_timer_sync(&pcap_ts->timer); > + > + input_unregister_device(pcap_ts->input); > + kfree(pcap_ts); > +
Could pcap_ts->work be still running at this point?
> + return 0; > +} > + > +} > +#else > + > +#define pcap_ts_suspend NULL > +#define pcap_ts_resume NULL > + > +#endif > + > +static struct platform_driver pcap_ts_driver = { > + .probe = pcap_ts_probe, > + .remove = __devexit_p(pcap_ts_remove), > + .suspend = pcap_ts_suspend, > + .resume = pcap_ts_resume,
I think it is preferred that .suspend and .resume assigments are guarded by #ifdef CONFIG_PM, not creating NULL stubs as this will (theoretically) allow removing .suspend and .resume pointers from driver stucture when compiling without PM support.
Also try running through checkpatch.pl - there are a few warnings I'd like to be fixed as well. Otherwise looks pretty good.
Oh, and you need to add your sign-off please. Thanks!
-- Dmitry
| |