On Sun, 2011-01-09 at 20:17 +0300, Sergei Shtylyov wrote: > > Which means that gadget drivers testing for valid struct usb_request * > > pointers against NULL will be fooled thinking it got a valid struct > > usb_request *, which is a mistake. > > How, if the 'request' field is the first field in the 'struct musb_request'? someone could add a field before it. > > So the potential for kernel oops is valid, my commit log might not have > > been the brightest though :-p > > Only a potential -- if we actually move the 'request' field. .. exactly. The code should work even if request isn't the first field. Ideally it would always be. It would actually make container_of() be optimized to a simple type cast. -- balbi -- To unsubscribe from this list: send the line "unsubscribe linux-usb" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
