Weekly edition	Kernel	Security	Distributions	Contact Us	Search
Archives	Calendar	Subscribe	Write for LWN	LWN.net FAQ	Sponsors

Why not do both?

Posted Aug 7, 2008 2:40 UTC (Thu) by smitty_one_each (subscriber, #28989)
Parent article: Can user-space bugs be kernel regressions?

If the driver has a bug, fix it.
Then patch the kernel.
A hybrid approach, after the Paris Hilton model.

(Log in to post comments)

Why not do both?

Posted Aug 7, 2008 9:35 UTC (Thu) by k3ninho (subscriber, #50375) [Link]

There's a third step: mark the kernel patch as Deprecated and schedule its eventual removal
for a point when the driver should be reasonably well-available in the wild.

K3n.

Why not do both?

Posted Aug 7, 2008 13:00 UTC (Thu) by bangert (subscriber, #28342) [Link]

exactly. was the driver fixed?

Why not do both?

Posted Aug 7, 2008 13:54 UTC (Thu) by jcristau (subscriber, #41237) [Link]

yes.
http://cgit.freedesktop.org/xorg/driver/xf86-input-synapt...