GNOME Bugzilla – Bug 762820
Remove "core_pointer" from the GdkDisplay struct
Last modified: 2016-03-10 06:42:44 UTC
This patch sets tries to cleanup all backends to not need display->core_pointer anymore. The same core device can be obtained from the device manager and the seat, so I think we should try to clean things up to make all this less confusing. Note: I can only test the x11 and win32 backends
Created attachment 322580 [details] [review] gdkwindow: get the pointer device from the default seat
Created attachment 322581 [details] [review] gdkdisplay: get the pointer device from the default seat
Created attachment 322582 [details] [review] broadway: get the core pointer from the device manager
Created attachment 322583 [details] [review] quartz: get the core pointer from the device manager
Created attachment 322584 [details] [review] gdk: remove the core_pointer field from GdkDisplay It is not used anymore, so no need to set it
Review of attachment 322580 [details] [review]: Looks fine
Review of attachment 322581 [details] [review]: sure
Review of attachment 322582 [details] [review]: ok
Review of attachment 322583 [details] [review]: Looks ok
Review of attachment 322584 [details] [review]: ok
thanks for the review. I pushed everything, but as mentioned I have no idea how to test some of the backends. If there are problems ping me.
Paolo, ping - this broke compilation of the Quartz backend. GdkDisplay->core_pointer is still used in gdkevents-quartz.c. Here's a patch that just copies what you did in gdkdisplay-quartz.c, it seems to work but I'm not too familiar with this part of the code.
Created attachment 323589 [details] [review] quartz: Get core pointer from device manager Some spots missed in the earlier refactor removing core_pointer from the GdkDisplay struct.
Thanks for fixing this up. Please push
Attachment 323589 [details] pushed as b3860e4 - quartz: Get core pointer from device manager