On Sat, Apr 10, 2010 at 09:57:34AM +0200, Corentin Chary wrote: > On Sat, Apr 10, 2010 at 9:35 AM, Carlos Corbacho > <carlos@xxxxxxxxxxxxxxxxxxx> wrote: > > On Saturday 10 April 2010 01:16:50 Yong Wang wrote: > >> On Fri, Apr 09, 2010 at 08:11:07PM +0100, Carlos Corbacho wrote: > >> > On Friday 09 April 2010 18:49:02 Dmitry Torokhov wrote: > >> > > On Fri, Apr 09, 2010 at 11:03:20PM +0800, Yong Wang wrote: > >> > > > Currently there is no easy way to associate driver data kind of thing > >> > > > with wmi drivers thus global variables have to be used to pass that > >> > > > information between functions. Add a pair of functions to support > >> > > > that. > >> > > >> > NAK > >> > > >> > > It looks like once you compete the conversion and have platform device > >> > > you can attach to it instead of needing driver data in WMI. > >> > >> The problem is that how can I retrieve the pointer to the platform device > >> itself in eeepc_wmi_exit to platform_device_unregister it? Is there any > >> other way to do it without having to define the platform device as > >> global variable? > > > > I don't think so. But then why is this a problem, since this is what other > > device drivers do? > > > > (The child devices of the platform driver could then be stored in the platform > > devices data, if you wanted). > > Maybe it's not that bad to get a static platform_device > *platform_device; if we don't > have the choice, then put eeepc_wmi in platform_device, and only use > get_drvdata in > platform_probe and eeepc_wmi_exit(). > OK, I will do it this way. Thanks to Carlos and Corentin for the suggestion. -Yong -- To unsubscribe from this list: send the line "unsubscribe platform-driver-x86" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
