On Tue, Oct 05, 2010 at 02:27:27PM -0400, Nathaniel McCallum wrote: > On 10/05/2010 02:25 PM, Matthew Garrett wrote: > > The range of ways that lid switches can be broken is large. One machine > > I've seen tries to read from a GPIO that's off by 16, because Intel's > > GPIO/GPE numbering is complicated. Another generates an event on lid > > close and generates a different event on lid open, but there's no > > handler for the open event and so the state variable never gets updated. > > The world is full of bad hardware and the kernel can't always be there > > to shield userspace from reality. > > Of course, but where possible it should sanitize. Userspace is like a > teenager, you expose them to just enough reality to know what's out > there, but in general they have no idea just how scary it really is. :) And it's not possible to sanitise in 100% of cases here, so userspace has to cope with the breakage. And if userspace has to cope, there's no point in sanitising any of it. -- Matthew Garrett | mjg59@xxxxxxxxxxxxx -- devel mailing list devel@xxxxxxxxxxxxxxxxxxxxxxx https://admin.fedoraproject.org/mailman/listinfo/devel
