On Sun, 2010-12-12 at 23:56 +0100, Marco d'Itri wrote: > On Dec 12, Ben Hutchings <ben@decadent.org.uk> wrote: > > > > The name is correct, the kernel name change should be backported. > > Please explain (1) why udev knows better than the kernel what the > > canonical name is and > The name has always been input/uinput, but in the last year udev > deprecated changing the device names and instead started to rely on the > kernel providing the correct ones. This driver was not fixed yet at the > time 2.6.32 was released. > > > (2) if it does, why it shouldn't change it quietly. > It does, this is just a warning. Is this summary correct: 1. The kernel called this device 'uinput' for a long time. 2. There is a longstanding udev rule that renames it to 'input/uinput'. 3. udev has deprecated such renaming rules. 4. New kernel versions call it 'input/uinput', so this rule is no longer triggered. I can't see the upstream change (4) so please point it out. Ben. -- Ben Hutchings Once a job is fouled up, anything done to improve it makes it worse.
Attachment:
signature.asc
Description: This is a digitally signed message part