Created attachment 248875 [details] [review]
backend-x11: Remove bad branch prediction

This will only get once, at in Clutter initialization time.

Created attachment 248876 [details] [review]
x11: Remove support for XInput 1

Now we either use core X11 or XInput 2.

Created attachment 248877 [details] [review]
device-manager: Don't pass the event mask around

There's no point in doing this, as we always use a constant event mask.
Simply do what everything else does.

Created attachment 248878 [details] [review]
device-manager: Select for events on XIAllMasterDevices

This removes a bit of work that we have to do for every device, and makes it
easy for mutter to patch out parts of the event mask it doesn't want.

Review of attachment 248875 [details] [review]:

okay

Review of attachment 248876 [details] [review]:

::: clutter/x11/clutter-input-device-core-x11.c
@@ +36,1 @@
 #define MAX_DEVICE_CLASSES      13

this can also go away, since it's only used inside the HAVE_XINPUT code paths.

@@ +45,1 @@
   gint *axis_data;

axis_data can go away as well.

@@ +62,1 @@
   g_free (device_x11->axis_data);

I think this should have also been included inside the HAVE_XINPUT check, since it's only going to be set to be !NULL there. the whole dispose() implementation should go away.

Review of attachment 248877 [details] [review]:

okay.

Review of attachment 248878 [details] [review]:

looks good.

Attachment 248875 [details] pushed as 0b32f99 - backend-x11: Remove bad branch prediction
Attachment 248876 [details] pushed as e38ea7a - x11: Remove support for XInput 1
Attachment 248877 [details] pushed as 0326888 - device-manager: Don't pass the event mask around
Attachment 248878 [details] pushed as e62cf47 - device-manager: Select for events on XIAllMasterDevices