Callers of this function were passing in -1, -1 for the width and height if
they just wanted the window moving - rather than the size changing. We need to
respect that behaviour and don't try and set the width/height to those
dimensions.
Now pop-up windows (ala combo-boxes) work!
"The child window MUST also have the _NET_SUPPORTING_WM_CHECK property set to
the ID of the child window. […] If the _NET_SUPPORTING_WM_CHECK window on the
client window is missing or not properly set, clients SHOULD assume that no
conforming Window Manager is present."
This commit implements that, which allows us to not have to do a
XGetWindowProperty() every N seconds when running under a compliant WM.
This is also a more correct fix for the bug handled in commit
daf29bffed.
https://bugzilla.gnome.org/show_bug.cgi?id=666921
The part in gdkwindow-x11.c which uses XIEvent and friends is "protected"
by HAVE_XGENERICEVENTS, i.e., XGetEventData() has been found. (Xlib.h)
XIEvent and friends are defined in <X11/extension/XInput2.h> which is
included by gdkdisplay-x11.h if XINPUT_2 is defined.
The patch makes sure XIEvent is only used if XINPUT_2 is defined.
https://bugzilla.gnome.org/show_bug.cgi?id=667534
When we receive a configure event from Wayland telling us to resize our
surface we check against the geometry constraints for the window to ensure we
do not resize below the minimum and maximum limits.
Now that grab and ungrab vfuncs are implemented on GdkDevice then we can use
gdk_device_ungrab to break the implicit grab created by the button press that
triggered the resize and move.
An implicit grab is created inside GTK+ when the button is pressed down on a
window. The semantics of wl_shell_surface_resize means that you don't get a
corresponding release event that would ordinarily break the implicit grab. So
we must do it as part of the resize request.
This change follows on from a change in semantics in Wayland where calling
wl_input_device_attach with nil would make the compositor set the pointer
sprite to it's default cursor sprite.
When the X server does not support the shape extension (as some
vnc implementations seem to), our DND code was always seeing
an empty input shape, so drops always missed their target.
http://bugzilla.gnome.org/show_bug.cgi?id=620240
Apparently the deprecation warning macros are placed into gdkconfig.h
during the configure stage, so put these in the pre-configured
gdkconfig.h.win32 as well, as their definitions are needed for all builds.
Reading a card32 property into a long may lead to undefined high
bits, so mask them off. Also, make the conditions for setting and
unsetting the stick flag opposites, to avoid unintended changes.
Patch by John Lindgren, bug 666842
The current EWMH has added 'source indication' fields to a number
of client messages. Set these to 1 to indicate a regular client.
Also fill the button field of _NET_WM_MOVERESIZE.
==23282== 64 bytes in 2 blocks are definitely lost in loss record 8,069 of 13,389
==23282== at 0x4A074CD: malloc (vg_replace_malloc.c:236)
==23282== by 0x39A1C3E2EA: cairo_region_create (cairo-region.c:196)
==23282== by 0x6D9AF3D: recompute_visible_regions_internal (gdkwindow.c:964)
==23282== by 0x6D9B4B8: recompute_visible_regions (gdkwindow.c:1126)
==23282== by 0x6DA3450: gdk_window_hide (gdkwindow.c:5689)
==23282== by 0x6D9CED9: _gdk_window_destroy_hierarchy (gdkwindow.c:2042)
==23282== by 0x6D9D040: gdk_window_destroy (gdkwindow.c:2109)
==23282== by 0x655B5E4: gtk_entry_unrealize (gtkentry.c:3012)
==23282== by 0x7068BF3: g_cclosure_marshal_VOID__VOID (gmarshal.c:85)
==23282== by 0x706710B: g_type_class_meta_marshal (gclosure.c:885)
==23282== by 0x7066DF9: g_closure_invoke (gclosure.c:774)
==23282== by 0x7080585: signal_emit_unlocked_R (gsignal.c:3340)
==23282== by 0x707F619: g_signal_emit_valist (gsignal.c:3033)
==23282== by 0x707FB71: g_signal_emit (gsignal.c:3090)
==23282== by 0x679E243: gtk_widget_unrealize (gtkwidget.c:4458)
==23282== by 0x64E83C7: gtk_bin_forall (gtkbin.c:172)
==23282== by 0x6548BBD: gtk_container_forall (gtkcontainer.c:2014)
==23282== by 0x67A966D: gtk_widget_real_unrealize (gtkwidget.c:10253)
==23282== by 0x672D002: gtk_tool_item_unrealize (gtktoolitem.c:474)
==23282== by 0x7068BF3: g_cclosure_marshal_VOID__VOID (gmarshal.c:85)
https://bugzilla.gnome.org/show_bug.cgi?id=666552
This is a boolean property that will be set to TRUE if the current
desktop environment is capable of displaying the application menu as
part of the desktop shell.
If it is FALSE then the application will need to display the menu for
itself.
We want to avoid handling focus events for the private focus window,
otherwise the keyboard grab taken by for example buttons will cause a
spurious FOCUS_OUT/FOCUS_IN on the toplevel.
The code that did this seems to have been lost in the XI2 transition for
GTK3.
https://bugzilla.gnome.org/show_bug.cgi?id=657578
The function returns the part of a monitors area that should be
used for positioning popups, menus, etc. The only non-trivial
implementation atm is in the X backend, all the other backends
just return the full monitor area. The X implementation is
currently suboptimal, since it requires roundtrips to collect
the necessary information. It should be changed to monitor
the properties for changes, when XFixes allows to monitor
individual properties.
https://bugzilla.gnome.org/show_bug.cgi?id=641999
Dealing with disabled devices may turn into hierarchy/device
changed events on device IDs with no backing GdkDevice yet,
so protect against that. The device attachment will be handled
correctly when the device is enabled later.
For maximized windows, titlebars cannot be used to reposition or
scale the window, so if an application does not use it to convey
useful information (other than the application name), the screen
space occupied by titlebars could be put to better use.
Add a new window property which requests from the window manager
to hide titlebars when windows are maximized to account for this.
https://bugzilla.gnome.org/show_bug.cgi?id=665616
Setup listener functions for the drag and drop events as well as the selection
events. Then create and save a data structure representing the data offer from
the other client.
This implementation is based on a hash table. The hard coded GtkSelection
atoms are preloaded into the hash table at the correct values. User generated
atoms start after the last preloaded atom.
The data device manager is a global object that provides the support
infrastructure around data devices. These data device objects are the basis
for handling drag and drop as well as selections in Wayland.
As pointed out in bug 665999, these were just not right.
Before this commit, the nicks were 'output' and 'only'.
After this commit, they are 'input-output' and 'input-only'.
This means we don't have to have hardcoded "/usr/share/wayland" to find the
cursors.
This change also fixes up the warning messages for when loading fails.
When an implicit paint is flushed during expose, e.g. because a
non-double buffered widget is painting, make sure to copy the existing
data from the window surface we rendered before flushing back to the
paint surface, instead of using an empty base.
Code was already handling that (and said so in the comment), but only
when no implicit paint was used at all, and not in the case when it's
flushed mid-expose.
Previously all the commands that acted on the shell took the surface that was
to be acted on as parameter. Now we retrieve an object from the shell that
represents its state for the surface. With that wl_shell_surface object we can
then call methods on that.
Signed-off-by: Kristian Høgsberg <krh@bitplanet.net>
gdk_window_get_update_area is supposed to get the area where things
need painting, and remove them from the update areas. However, if
some area is covered by other windows with an alpha background we
can't just expect whatever the app choses to render in the update
area as correct, so we don't actually remove these areas, meaning
they will get correctly rendered when we get to the expose handlers.
gdk_window_move_region doesn't move children, so we can't copy
transparent child window regions with copyarea, so we remove these
from the copy region.
We track the areas that have alpha coverage so that we can
avoid using these as sources when copying window contents.
We also don't remove such areas from the clipping regions so
that they are painted both by parent and child.
This cleans up the expose handling a bit by using the existing
clip regions, and it allows us later to use painters algorithm
to do transparent windows.
This turns GdkWin32DragContext into a proper GdkDragContext subclass.
Because we now correctly initialize GdkWin32DragContext in
gdk_drag_context_new, we no longer crash immediatly when a DnD
operation is initialized (the find_window, drag_status, ... vfuncs
where all pointing to 0x0 instead of their proper win32 implementations).
We now try to consistently refer to GdkDragContext as "context",
GdkWin32DragContext as "win32_context" and the ole2 related
target_drag_context and source_drag_context as "ctx".
Members of GdkWin32DragContext only used by the ole2 DnD codepaths
are now explicitly marked with a ole2_dnd_ prefix.
gdk_cairo_region_create_from_surface doesn't work correctly on PPC.
This is most prominently seen with the GTK window resize grip, the
shape of which is mirrored every eight pixels horizontally.
At the same time, use an A1 surface for the resize grip shape to
eliminates an A8->A1 surface conversion.
Add GDK_MODIFIER_INTENT_SHIFT_GROUP to enum GdkModifierIntent
and handle it in gdk_keymap_get_modifier_mask(). Add an X11
impl of the method and return keymap_x11->group_switch_mask.
Return 0 from the default impl because we don't know.
If the keyboard group shifting modifier is *also* a normal
accelerator modifier, we need to special case it when calling
gdk_keymap_translate_keyboard_state(), so we get the right
key symbol for accelerators (for example we want Option-O,
not Option-Ø displayed in menu items). This patch should only
affect quartz where the Alt key both shifts the group and can
be used as accel modifier, and not X11 or Win32 where AltGr
is not used for accelerators.
- fix quartz' gdk_keymap_translate_keyboard_state() to return
the right consumed_modifiers
- add _gtk_translate_keyboard_accel_state() which does the
special casing
- use it everywhere instead of gdk_keymap_translate_keyboard_state()
g_checksum_get_digest checks to ensure that the passed digest_len is long
enough to hold the digest, before setting it to the actual length of the
digest returned. Digest_len is uninitialized in the code, so if you're
lucky it will be larger than 20 and everything will work fine. If you're
unlucky, g_checksum_get_digest will return either -1 or some number less
than 20, and the g_assert(digest_len==20) will fail.
Wake up the run loop unconditionally (don't check if it is waiting) because
it might go into waiting state right after we checked for it. Fixes GIMP
startup (which has a lot of GIOChannel I/O but zero NSEvents) from several
minutes to a few seconds.
(cherry picked from commit 0729cdc9a1)
We now do proper handling of existing invalid regions, and
we use ScrollDC which allows us to specify the right clip
region rather than just the bounding rect.
Positioning windows at 0,0 post creation failed, because it
was mapped with CW_USEDFAULT, but private->x/y still said 0,
so moving it to 0,0 did nothing. We now always position the
window at the right place, even when not mapped, but we
create it at CW_USEDEFAULT initially and store that position
before moving it to the right place.
This fixes the window sizing test in testgtk and the inital
position for the gimp toolbar.
The synaptics trackpad driver has some weird behaviour on scroll.
It pops up a window over the mouse pointer (looking like a scrollbar).
This has two problems:
* We get extra enter/leave events for the trackpad window
* We get back the trackpad window when we look for the window
under the mouse to deliver the mousewheel message.
So, we add some trackpad specific hacks to avoid this (sigh) based
on the trackpad window window class.
This fixes bug #542777 and was partially based on a patch there
from Peter Clifton.
gdk_flush() should gdk_display_sync() on all open displays.
Both for display_flush and display_sync it seems useful to call
GdiFlush, but we don't have anything extra to do for display_sync,
as there is no inherent roundtrip on win32.
This should close bug #84314
Windows with transients: center on parent
Splash screens: center on monitor
Also properly ignores initial moves of unmapped
windows that are not override redirect or HINT_POS
Fixes bugs #324254 and #612359
There were still cases where we didn't get a WINDOWPOSCHANGED after
a SetWindowPos() call, like e.g. with a larger minimum size than
the set size (bug #574935)
So, we revert the previous fix and now just always manually emit
a configure notify after the move_resize call. Also, we inhibit
the WINDOWPOSCHANGED configure event during the move_resize operation
to avoid multiple Configures.
There are some cases where we don't get a WINDOWPOSCHANGE such that
we generate a configure event, even if we called gdk_window_move_resize()
or similar. For instance:
* The window is fullscreen
* The window is maximized
* The specified pos/size is the same as the current one
However, as per X11 ConfigureNotify semantics we *always* want one, or
we could run into issue like e.g. bug #537296 where we're waiting for
the CONFIGURE to call gdk_window_thaw_toplevel_updates_libgtk_only().
We always get the WM_DESTROY message anyway, and we remove it there.
Bug #336416 even claims this could be a leak if the WM_DESTROY
message was not seen before the DestroyWindow call returned, as
the WM_DESTROY message could not be handled later without the
window in the handle table. I'm not sure this can happen, but we
might as well remove it.
There is no particular reason to special case this, we want to handle all
sort of normal events. The only special thing we keep is that
as an optimization we pump the message loop extra during a WINPOSCHANGED
in a modal operation as that will cause us to repaint faster.
Also, bump the arbitrary number of mainloop iterations for the timer.
I don't see why we need it at all, but at least doing more than one
iteration if needed should be nice.
When you start a window resize or move via the window menu and
don't actually change anything we're not getting an exitsizemove.
In order to work around this we also look for WM_CAPTURECHANGED.
This moves all the code from WM_SIZE, WM_MOVE, and WM_SHOWWINDOW into
one place, cleans up the code and makes sure we only send a single
configure event even if both size and position changes.
We don't pass in raise anymore, but already_mapped.
Also, already_mapped must be used rather than MAPPED, as we already
synthesize the MAPPED in the generic code (and thus we don't have
to synthesize it again).
Calling PeekMessage can cause reentrant calls into the window procedure
for sent (as opposed to posted) messages, so its not safe to call
when we're not expecting reentrancy. Instead we call GetQueueStatus
when we're just looking for availible messages.
Fixes https://bugzilla.gnome.org/show_bug.cgi?id=552041
By reverting gdk_drag_find_window_for_screen logic to what it was
before eb21a7df29.
The old logic knew how to ignore drag_window when searching
for dest_window, but that code was removed (I guess by accident).
https://bugzilla.gnome.org/show_bug.cgi?id=616544
Commit 5ebb32d1ff didn't add the correct
code to find the toplevel window. The WindowFromPoint() function does
not return the toplevel window in the hierarchy, it returns the deepest
non-disabled, non-invisible child. As we don't use invisible or disabled
windows, we don't actually need to use the ChildWindowFromPoint walk for
the non get_toplevel case, so we can remove that code path.
To find a toplevel, we need to start from the desktop and work up, using
ChildWindowFromPointEx (to ignore invisible and disabled windows). If we
don't ignore invisible and disabled windows (as is the case with the
ChildWindowFromPoint call, we are liable to get returns of hidden or
disabled children of the desktop which don't belong to us, but notionally
occupy the same area under the pointer.
An alternative might be to start our walk with one of the children of the
desktop owned by our process and thread - which we can enumerate using,
the EnumThreadWindows call, or (presumably) determine internally. This
would not work when we are inside a GtkSocket though, as the children of
the desktop would belong to the process owning the GtkPlug - we would
have to rely on our own list of windows.
For correctness, this commit adds tests to ensure that we don't try to
return either x or y window coordinates if that corresponding pointer is
NULL.
https://bugzilla.gnome.org/show_bug.cgi?id=658842
The button highlighting in testgtk works again, even with
GDK_NATIVE_WINDOWS. Unfortunately testgtk:menus still does
not work for the forced-native-window-case.
In _gdk_window_move_resize_child it tries to decide whether to pass
SWP_NOSIZE and SWP_NOMOVE based on whether the new size and position
is different from the window's existing position. However it seems
that GDK now ends up updating the window's position before calling
_gdk_window_move_resize_child so this would mean it would think the
window never changes size or position so SWP_NOSIZE|SWP_NOMOVE would
always be set. This causes child windows to never be resized.
This patch changes it so that it never passes either flag to
SetWindowPos. I don't know whether this will cause any side effects
but you'd think it shouldn't do any harm to reassert the current size.
https://bugzilla.gnome.org/show_bug.cgi?id=628049
Signed-off-by: Hans Breuer <hans@breuer.org>
Allows more modern browsers eg. firefox 5+ to use gtk/broadway
Auto-detects protocol version, and can switch between them at
as you connect a different browser.
This works to some extent, but seems to hang sometimes, for
instance the "button box" test in testgtk never shows up.
Don't try to remember the current keyboard modifier and mouse button
states from the last event, because that isn't always right, and don't
set event.state = 0 for generated events. Instead, add private functions
to get the current states, and implement them with API that retrieves
these states independently from an event.
_NET_WM_STATE_FOCUSED is a new _NET_WM_STATE hint which allows us to
implement a meaningful GDK_WINDOW_STATE_FOCUSED under X11. If the window
manager doesn't support this hint we keep GDK_WINDOW_STATE_FOCUSED set since
that is what gtk+ implicitly assumed historically.
https://bugzilla.gnome.org/show_bug.cgi?id=661428
This state means that the toplevel window is presented as focused to the user,
i.e with active decorations under an X11 window manager.
If the GDK backend doesn't implement this flag, it will just remain set after
mapping the window.
https://bugzilla.gnome.org/show_bug.cgi?id=661428
which effectively nails down the MOD1 == ALT assumption that is valid
in all other parts of GTK+. After the modifier abstraction fixes for
OSX, the virtual mapping is now (correctly) used in more places, and
caused problems with the common default PC keyboard layout on X11 that
colocates ALT and META on the same key.
Fixes e.g. crashs when dropping from finder.
Turn the "getting_events" boolean into a counter to handle poll_func()
being called recursively, and track the loop depth correctly by
changing its counter before bailing out in run_loop_observer_callback().
This way we reallocate our autorelease pool at the right time, and
don't kill memory that is still in use by outer run loops.
Also drain, not release the pool, just for some defensive forward
compatibility.
(cherry picked from commit ef9a92d225)
When an NSEvent does not have the window field set, we already assumed
the event was not for us and discarded it. But for NSMouseMoved events
we now make an exception, because such events generated after
using/clicking the main menu bar have the window field set to NULL while
the application window still has focus.
We used to experience a loss of motion events after using the menu bar,
this could be seen in buttons that stopped prelighting and first
clicks often being ignored unless you clicked somewhere else first.
These issues are fixed by this patch.
Since the wmspec_check_window doesn't have a corresponding GdkWindow we can't
rely on the get_event_window() return value to get the XID from. Just use the
XID from the XEvent directly.
https://bugzilla.gnome.org/show_bug.cgi?id=662953
The new file defines GDK_DISABLE_DEPRECATION_WARNINGS so it can happily
use deprecated APIs.
This commit moves those functions there that use deprecated functions
and currently cause warnings.
With this commit, GDK compiles without deprecation warnings.
After consulting with Carlos, we agreed that it should be enough to grab
the core pointer instead of doing a full grab. If it turns out that's
wrong, we need to adapt the internal API for resizes to take the device
doing the resize.
When compiled with older SDKs, the original change for this bug caused a
compiler warning about NSWindow not being able to handle a setStyleMask
message. This tricks the compiler into thinking that it can.
In 2.x, the !HAVE_XCONVERTCASE fallback of keyval_convert_case() was
implicitly used as implementation for all !X11 backends.
In 3.x, when this function was virtualized in GdkDisplayManager,
this fallback was moved to the X11 backend and the other backends
"equipped" with /* FIXME implement */ implementations of
keyval_convert_case() which don't convert anything.
Move the fallback code back to gdk/ as default implementation
of GdkDisplayManager::keyval_convert_case() and remove its
implementations is all backends but X11. Also remove the
implementation in Wayland which was a plain copy of what
is now the default implementation.
(cherry picked from commit f46c1b76d8)
gdk_x11_device_manager_core_list_devices returns a new allocated
list, which has to be freed.
valgrind output:
==18686== 160,176 (80,088 direct, 80,088 indirect) bytes in 3,337 blocks are definitely lost in loss record 25,347 of 25,378
==18686== at 0x4C256DD: malloc (in /usr/lib64/valgrind/vgpreload_memcheck-amd64-linux.so)
==18686== by 0x6CD7752: g_malloc (in /lib64/libglib-2.0.so.0.3000.0)
==18686== by 0x6CEE2B6: g_slice_alloc (in /lib64/libglib-2.0.so.0.3000.0)
==18686== by 0x6CCB37D: g_list_prepend (in /lib64/libglib-2.0.so.0.3000.0)
==18686== by 0x654CADA: gdk_x11_device_manager_core_list_devices (gdkdevicemanager-core-x11.c:836)
==18686== by 0x6531489: gdk_display_pointer_is_grabbed (gdkdisplay.c:1270)
==18686== by 0x5162E1E: filter_func (ui.c:140)
==18686== by 0x6558B50: gdk_event_apply_filters (gdkeventsource.c:83)
==18686== by 0x6558CB3: _gdk_x11_display_queue_events (gdkeventsource.c:197)
==18686== by 0x6530680: gdk_display_get_event (gdkdisplay.c:311)
==18686== by 0x65589F1: gdk_event_source_dispatch (gdkeventsource.c:356)
==18686== by 0x6CD0A0E: g_main_context_dispatch (in /lib64/libglib-2.0.so.0.3000.0)
https://bugzilla.gnome.org/show_bug.cgi?id=660676
Those if() blocks don't have any reason being there, as x and y are not
pointers. If the window is destroyed, just set the out values to zero
and return.
As seen in valgrind:
==3306== Conditional jump or move depends on uninitialised value(s)
==3306== at 0x624C74F: gdk_window_get_root_coords (gdkwindow.c:6933)
==3306== by 0x5E193C3: gtk_tooltip_show_tooltip (gtktooltip.c:1160)
==3306== by 0x5E19C05: tooltip_popup_timeout (gtktooltip.c:1282)
==3306== by 0x623B102: gdk_threads_dispatch (gdk.c:754)
==3306== by 0x8592F3A: g_timeout_dispatch (gmain.c:3907)
==3306== by 0x859174C: g_main_context_dispatch (gmain.c:2441)
==3306== by 0x8591F47: g_main_context_iterate (gmain.c:3089)
==3306== by 0x8592494: g_main_loop_run (gmain.c:3297)
==3306== by 0x5D2E501: gtk_main (gtkmain.c:1362)
==3306== by 0x5C5652F: gtk_application_run_mainloop
(gtkapplication.c:115)
==3306== by 0x7C47C9D: g_application_run (gapplication.c:1323)
==3306== by 0x447B5F: main (nautilus-main.c:102)
==3306== Uninitialised value was created by a stack allocation
==3306== at 0x624D48A: gdk_window_get_device_position
(gdkwindow.c:4952)
gdk_unicode_to_keyval(uc) returning (uc | 0x01000000) is not an
error return value but simply the way to encode 24-bit unicode
characters directly as keyvals.
Add enum GdkModifierIntent which identifies use cases for modifier masks
and GdkKeyMap::get_modifier_mask(). Add a default implementation which returns
what is currently hardcoded all over GTK+, and an implementation in the
quartz backend. Also add gtk_widget_get_modifier_mask() which simplifies
things by doing widget->display->keymap->get_modifier_mask().
Handle dead keys in special_ucs_table and have them converted by
UCKeyTranslate(), so all dead key combinations can be entered.
Later, this should be handled in the input method, just as it's
done for X11/Win32.
Fixes bug #658379 - Disabled devices still added to list on startup,
spotted by Bastien Nocera. Do not create GdkDevices for disabled
devices on device manager construction, leading to a confusing initial
state.
This commit introduces a new setting, gtk-visible-focus, backed
by the Gtk/VisibleFocus X setting. Its three values control how
focus rectangles are displayed.
'always' is equivalent to the traditional GTK+ behaviour of always
rendering focus rectangles.
'never' does what it says, and is intended for keyboardless
situations, e.g. tablets.
'automatic' hides focus rectangles initially, until the user
interacts with the keyboard, at which point focus rectangles
become visible.
https://bugzilla.gnome.org/show_bug.cgi?id=649567
Functions dealing with native Xlib types were (skip)ed because
gobject-introspection did not have correct Xlib types declarations.
They are corrected now, so these GdkX11 functions can be enabled back
again.
https://bugzilla.gnome.org/show_bug.cgi?id=655495
I tried to suppress compiler warnings on pre-10.6 machines this way,
but it defeats its purpose when you compile for pre-10.6 machines on
a 10.6 machine. For now, we have to live with the warnings when
compiling on/for pre-10.6 machines, there does not seem an easy and proper
way to suppress the warnings.
https://bugzilla.gnome.org/show_bug.cgi?id=653947
It could happen that a cookie event has been already allocated/freed
in an event filter, as it can't be allocated a second time, all GDK
can do is skipping the event. Spotted by Guillaume Desmottes.
This function can be used to find the GdkDevice wrapping
an XInput2 device ID. For core devices, the Virtual Core
Pointer/Keyboard IDs (2/3) may be used.
This function can be used to find out the XInput2 device ID
behind a GdkDevice, mostly useful when you need to interact
with say Clutter, or raw libXi calls.
Fixes Bug 645993 - XIM has wierd behaviors. Some XIM modules
filter every key event, possibly replacing it with their own
one. These events usually have serial=0, so make
GdkDeviceManagerXI2 also listen on these.
For client-side windows, we need to queue a repaint when the background
changes. For native windows, the windowing system does take care of it,
but client-side windows are our own, so we gotta do it manually.
https://bugzilla.gnome.org/show_bug.cgi?id=652102
This is already done in gdk_event_source_get_filter_window(), and
could lead to wrong event assignment if an event translator happens
to return a window for an event it doesn't handle.
This method can be implemented by event translators so they
return the right window from XGenericEventCookie events, as
ev->xany.window isn't meaningful for these.
GdkEventSource now also uses this to find out the right window
filters to apply.
XKB and GDK both add "internal" bits to GdkModifierType. In C,
this typically doesn't cause problems as bitfields are just integers,
and there's no validation. However for bindings, it's normal to
convert enumerations to "native" enumeration types, which don't
support unknown bits. See bug 597292.
https://bugzilla.gnome.org/show_bug.cgi?id=634994
