On Mon, Oct 04, 2010 at 06:54:00PM +0530, Rahul Ruikar wrote:
quoted text
> Dan,
> 
> Following things I tried to do with this change.
> there can be 3 cases where one need to handle device_register()
> 1) reaching error path but device_register() is never called..( ie,
> error occurred before calling device_register() function call, in this
> case "dev->reg_status" will have value "0" and in error handler
> device_unregister() or put_device() will not be called.
> 

Good point.  This is a bug in the original code.

quoted text
> 2) error occurred at device_register() ( ie. it fails and calls error
> handler, this way "dev->reg_status" will have value "2" and in error
> handler, based on this value "put_device() will be called.
> 

Yeah.  But device_register() sucks.  It shouldn't require such byzantine
error handling.  Someone is working on this.

regards,
dan carpenter

quoted text
> 3) error occured after device_register() success. ( ie, error occured
> due to some other function, ) in this case "dev->reg_status" will have
> value "1" and in error handler, based on this value
> "device_unregister" will be called.
> 
> but anyways someone has now proposed to change it to new usb
> interface, so this patch is of no use now.
> 
> Thanks
> Rahul Ruikar
> 

--
unsubscribe notice
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/