Re: [PATCH] proposed PnP layer changes/fixes and cleanups

[Greg KH]

On Sun, Dec 29, 2002 at 12:20:15PM +0100, Jaroslav Kysela wrote:
> Hi,
> 
> 	I've revised the PnP code and found numerous of simple bugs. Also,
> my opinion is that the resctriction of resources to only ones hardcoded by
> hardware vendors is not a good idea. Hardware usually supports more
> configurations (especially ISA PnP devices).
> 	I've tried to implement a configuration template, so driver can
> pass it's own resource map in the probe phase. Also, I've added a new
> member to the driver structures - flags. The flag
> PNP_DRIVER_DO_NOT_ACTIVATE instructs the pnp code that the device
> shouldn't be activated before the probe() callback is entered. It's
> necessary to implement this behaviour to allow passing of a new
> configuration template (thus calling pnp_activate_dev() from the probe()
> callback).
> 	Also, the resources can be locked now (I've changed the ro flag to 
> lock_resources).

These changes look good, I'll send them on to Linus.

thanks,

greg k-h