Commit Graph

187 Commits

Author SHA1 Message Date
Jonas Ådahl
98ee85ce82 Let the MetaWindow calculate its main logical monitor itself
There is no point in going via a MetaScreen API to calculate the "main
monitor" of a window, so let the window do that calculation itself.

https://bugzilla.gnome.org/show_bug.cgi?id=777732
2017-01-25 16:28:52 +08:00
Jonas Ådahl
5a8509f895 Move current logical monitor retrieval to the backends
The method used for getting the current logical monitor (the monitor
where the pointer cursor is currently at) depends on the backend type,
so move that logic to the corresponding backends.

https://bugzilla.gnome.org/show_bug.cgi?id=777732
2017-01-25 16:28:52 +08:00
Jonas Ådahl
ecf796f82b window: Keep track of fullscreen monitors directly
Instead of keeping around array indexes, keep track of them by storing
a pointer instead. This also changes from using an array (imitating the
X11 behaviour) to more explicit storing.

https://bugzilla.gnome.org/show_bug.cgi?id=777732
2017-01-25 16:28:51 +08:00
Jonas Ådahl
842ff5d6aa Rename MetaMonitorInfo MetaLogicalMonitor
In preparation for further refactorizations, rename the MetaMonitorInfo
struct to MetaLogicalMonitor. Eventually, part of MetaLogicalMonitor
will be split into a MetaMonitor type.

https://bugzilla.gnome.org/show_bug.cgi?id=777732
2017-01-25 16:28:51 +08:00
Carlos Garnacho
be5c2ebcb5 x11: Unminimize windows with initial IconicState if first known on MapRequest
In that case, the MetaWindow is created, but it should also be unminimized
to satisfy the MapRequest triggered by the client, otherwise these would
stay minimized until they're shown explicitly by the user.

https://bugzilla.gnome.org/show_bug.cgi?id=774333
2016-11-15 18:39:42 +01:00
Carlos Garnacho
ae73e9d84d x11: Double check whether the user time window was previously used
Some applications like Wine may choose to juggle the same user time
window across different toplevels, in that case we receive warnings
when trying to register the window a second time, leading to wrong
accounting.

If the window was already used as the user time window for another
toplevel, unset it from the previous MetaWindow owner, and unregister
so the registration with the new MetaWindow is successful.

https://bugzilla.gnome.org/show_bug.cgi?id=774330
2016-11-15 17:33:46 +01:00
Florian Müllner
e9944350d3 delete: Share some code between X11 and wayland
In order to kill a window, on both X11 and wayland we first try to
kill(3) the corresponding process, so we can add the newly added
get_client_pid() method to share that code.

https://bugzilla.gnome.org/show_bug.cgi?id=772613
2016-10-10 23:35:02 +02:00
Florian Müllner
1fab6e69b7 window: Add get_client_pid() method
It is often useful to identify the client process that created
a particular window, however the existing meta_window_get_pid()
method relies on _NET_WM_PID, which is only available on X11 and
depends on applications to set it correctly (which may not even
be possible when the app runs in its own PID namespace as Flatpak
apps do). So add a get_client_pid() method that uses windowing
system facilities to resolve the PID associated with a particular
window.

https://bugzilla.gnome.org/show_bug.cgi?id=772613
2016-10-10 23:35:01 +02:00
Jonas Ådahl
8942e98e1d Revert "window: Make meta_window_has_pointer() per protocol implemented"
Windows from Xwayland still needs to use the Wayland path, but is
represented an MetaWindowX11, thus the abstraction introduced in
"window: Make meta_window_has_pointer() per protocol implemented"
is wrong. Lets turn back time, and reconsider how this can be
abstracted more correctly in the future.

This reverts commit 9fb891d216.
2016-08-26 06:51:19 +08:00
Jonas Ådahl
9fb891d216 window: Make meta_window_has_pointer() per protocol implemented
https://bugzilla.gnome.org/show_bug.cgi?id=769936
2016-08-25 12:19:40 +08:00
Jonas Ådahl
cd225c4e19 Always use the default screen
GDK doesn't support multiple screens, so effectively we don't either.
Lets stop pretending we do.

This fixes a few -Wdeprecated warnings.

https://bugzilla.gnome.org/show_bug.cgi?id=769070
2016-07-23 10:55:02 +08:00
Jonas Ådahl
95a68854ad configure.ac: Add AC_GNU_SOURCE
We define it here and there already, and if we are to include private
cogl files, we need it in more places. Lets just add it everywhere.

https://bugzilla.gnome.org/show_bug.cgi?id=768976
2016-07-20 14:23:48 +08:00
Rui Matos
d62491f46e x11/window-props: Convert WM_NAME and WM_CLASS to UTF-8
gjs throws exceptions on non UTF-8 strings which, in some cases, crash
gnome-shell. ICCCM string properties are defined to be Latin-1 encoded
so we can try to convert them to avoid it.

Note that _NET_WM_NAME is defined to be UTF-8 and we already validate
it in utf8_string_from_results() .

https://bugzilla.gnome.org/show_bug.cgi?id=752788
2016-06-07 20:17:38 +02:00
Jonas Ådahl
417cb2b213 bell: Let the X11 caller deal with the X11 fallback
To support invoking the system bell on Wayland we shouldn't have paths
that fallback to X11. Let the X11 caller deal with the absence of
libcanberra, and change API to not take any X events.

https://bugzilla.gnome.org/show_bug.cgi?id=763284
2016-03-10 12:34:56 +08:00
Carlos Garnacho
56beedf9f2 core: Refactor startup notification into a separate object
This is kind of in a middle ground at the moment. Even though it
handles sequences not coming from libsn, they're added nowhere at
the moment, we'll rely on the app launch context being in the x11
side at the moment.

Also, even though we do create internal sequence objects, we keep
exposing SnStartupSequences to make gnome-shell happy, we could
consider making this object "public" (and the sequence objects with
it), things stay private at the moment.

https://bugzilla.gnome.org/show_bug.cgi?id=762268
2016-02-19 17:41:03 +01:00
Jasper St. Pierre
6fc51e3723 window-x11: Fix checks for a bounding region input region
When cleaning up the logic in commit c408cf7, I forgot to properly apply
de Morgan's laws to an inverse.

Reported by ricotz on IRC.
2016-01-22 18:28:47 -08:00
Sebastian Keller
a7a376ae1f xprops: Null-terminate property reply values
Some of the mutter code using these properties expects them to be
null-terminated whereas xcb does not use null-terminated strings:

http://xcb.freedesktop.org/XcbRationale/

This was in some cases resulting in the WM_CLASS property containing
garbage data which broke application matching, caused the hot-corner and
window-switcher to stop working, or was exposed as text in the UI.

https://bugzilla.gnome.org/show_bug.cgi?id=759658
2016-01-12 16:30:30 +01:00
Florian Müllner
3aea8d8ce6 Revert "Add "size states" which save window size information"
This reverts commit 2c7ef2269f.
2016-01-10 15:16:09 +01:00
Jasper St. Pierre
2c7ef2269f Add "size states" which save window size information
https://bugzilla.gnome.org/show_bug.cgi?id=751857
2016-01-09 18:08:35 -08:00
Jasper St. Pierre
053f5088df window-x11: Fix unredirection for windows that explicitly set bounding
Some applications, like Chromium, explicitly set their bounding region
to the client area when full-screen. Detect this case, and allow us to
fullscreen when this happens.
2016-01-04 08:32:09 -08:00
Rui Matos
e7390cff83 x11/window: Ensure we send a ConfigureNotify to just mapped windows
When managing a non-OR window we're required by the ICCCM to behave as
if we received a ConfigureRequest which means that we must generate a
synthetic ConfigureNotify even if the window isn't moved or resized
from its current (initial) geometry.

During MetaWindow's x11/wayland split a slight behavior change for x11
windows crept in. Before the code split, MetaWindow->rect was
initialized with the X window's geometry, but now we're not
initializing MetaWindowX11Private->client_rect which causes the checks
for whether it's necessary to move/resize the window in
meta_window_x11_move_resize_internal() to tell us that we do need to
move/resize which means we do an XConfigureWindow() call and don't
send the sythetic ConfigureNotify. But since the X window isn't really
moving, the XConfigureWindow() call doesn't cause the X server to
generate a ConfigureNotify which breaks some clients such as Java's
AWT.

We can fix this by setting MetaWindowX11Privatew->client_rect for both
OR and non-OR windows. We can set buffer_rect for non-OR windows as
well to simplify the code since it will be assigned the correct value
in meta_window_x11_move_resize_internal() .

https://bugzilla.gnome.org/show_bug.cgi?id=759492
2015-12-16 19:46:41 +01:00
Jasper St. Pierre
8cc345fcf5 window: Allow minimizing windows which don't advertise support for it
Wine removes the minimize func from its Motif hints on full-screen
windows, because, as the Win32 API literally says, the minimize button
is indeed not visible on full-screen windows.

Given that this code was added to prevent minimizing a panel by
accident, I don't necessarily think that it's relevant anymore.

https://bugzilla.gnome.org/show_bug.cgi?id=758186
2015-12-05 10:46:21 -08:00
Rui Matos
7606f79a1e x11/window-props: Initialize bypass compositor hint
If a client only ever sets the hint on window creation we'd never pick
the value. Also, include override redirect windows since the hint is
relevant to them too.

https://bugzilla.gnome.org/show_bug.cgi?id=758544
2015-11-23 19:54:48 +01:00
Giovanni Campagna
f2afa7aa6c mutter: don't show the resize popup for 2 x 2 size increments
In a HiDPI environment, all gtk+ apps will report a 2 x 2 size
increment to avoid odd size. But that does not mean they are
resizing in cells like terminals, so they resize popup should
not be shown.

Ideally, we should ignore <= scale x scale increments, but in
practice scale is 1 or 2, and even in a lo-dpi setting a 2 x 2
increment makes little sense so let's keep the patch simple.

https://bugzilla.gnome.org/show_bug.cgi?id=746420
2015-10-19 17:21:59 -07:00
Florian Müllner
2feeb57dee iconcache: Mark surfaces as dirty after changing data
This is required to tell cairo to update its cached areas.
2015-10-16 04:04:31 +02:00
Florian Müllner
3a63d58d9e events: Don't use XIEvent serial numbers
XInput2 uses the raw sequence number for XIEvent serials[0], which only
matches the serial number in XEvents up to 16 bits[1]. So in order to
be able to make reliable comparisons with serials from other events or
calls to XNextRequest(), always use the field from the original XEvent
rather than the XIEvent serial (at least until we can get libXi fixed).

This (partially) reverts commit 35dd1e644d.

[0] http://cgit.freedesktop.org/xorg/lib/libXi/commit?id=5d43d4914dcabb6d
[1] http://cgit.freedesktop.org/xorg/lib/libX11/tree/src/XlibInt.c#n265

https://bugzilla.gnome.org/show_bug.cgi?id=756649
2015-10-15 18:50:51 +02:00
Florian Müllner
a95ae4d178 session: Fix crash when saving sticky windows
Since commit 527c53a2a0, window->workspace is set to %NULL when
the window is sticky (see comment[0]), so don't try to save the
workspace index in that case.

[0] https://git.gnome.org/browse/mutter/tree/src/core/window.c#n4307

https://bugzilla.gnome.org/show_bug.cgi?id=756642
2015-10-15 16:30:43 +02:00
Rui Matos
54557f062e xprops: Fix reading Window and XSyncCounter properties
Both Window and XSyncCounter are XIDs which on 64 bit X clients are 8
bytes wide. But the values on the wire are 32 bit so, for these types,
we always copy 4 bytes into results->prop. As such copying them out
with a cast such as *(Window *) means that we are actually reading 8
bytes which depending on whether the higher addressed 4 bytes are zero
means that sometimes this works while others it gives us a bogus
value.

https://bugzilla.gnome.org/show_bug.cgi?id=756074
2015-10-06 22:33:42 +02:00
Florian Müllner
25a796afc6 Fix constness warnings 2015-09-24 03:01:51 +02:00
Jasper St. Pierre
352cac3850 group: Convert code to XCB to prevent errors
Windows can disappear at any time because X11 is really cool and good,
so just use XCB so we won't crash if the window disappears.
2015-08-05 13:39:23 -07:00
Rui Matos
07f533f617 window: Remove fullscreen_after_placement special case
This was introduced in commit c6793d477a
to prevent window self-maximisation. It turns out that that bug seems
to have been fixed meanwhile in a different way since the reproducer
in https://bugzilla.gnome.org/show_bug.cgi?id=461927#c37 now works
fine with this special handling removed.

In fact, failing to set window->fullscreen immediately when loading
the initial set of X properties causes us to create a UI frame for a
window that sets _NET_WM_STATE_FULLSCREEN.

This, in turn, might cause the fullscreen constrain code to fail if
the window also sets min_width/min_height size hints to be the monitor
size since the UI frame size added to those makes the rectangle too
big to fit the monitor. If the window doesn't set these hints, we
fullscreen it but the window will get sized such that the UI frame is
taken into account while it really shouldn't (see the reproducer
above).

https://bugzilla.gnome.org/show_bug.cgi?id=753020
2015-07-30 15:27:01 +02:00
Florian Müllner
b41873dced window-x11: Remove duplicated code
meta_window_change_workspace_without_transients() already does
exactly that.
2015-07-23 12:13:45 +02:00
Jonas Ådahl
f6c9261bf6 wayland: Scale window geometry rects given the main output
Since we scale surface actors given what main output their toplevel
window is on, also scale the window geometry coordinates and sizes
(window->rect size and window->custom_frame_extents.top/left) in order
to make the window geometry represent what is being rendered on the
stage.

https://bugzilla.gnome.org/show_bug.cgi?id=744934
2015-07-16 11:54:14 +08:00
Jasper St. Pierre
aea71fbd01 group: Select for property notifies on group leaders 2015-06-27 00:53:22 -07:00
Jasper St. Pierre
f041b35b9b xprops: Fix STRING types as well 2015-06-26 13:51:45 -07:00
Jasper St. Pierre
6dbec6f81b xprops: More Xlib / long cleanliness with Motif WM hints
Fixes some CSD windows sometimes getting frames when they totes
shouldn't.
2015-06-24 14:34:48 -07:00
Jasper St. Pierre
3962f1d982 xprops: Apply the same Xlib / 64-bit hygiene to atom lists 2015-06-23 15:37:58 -07:00
Jasper St. Pierre
414be07a69 xprops: Remove unused synchronous property getters
No reason to keep these around..
2015-06-23 15:37:23 -07:00
Jasper St. Pierre
7f3ada7831 xprops: Some more cardinal-related hygiene 2015-06-23 15:32:07 -07:00
Jasper St. Pierre
92d6a69153 xprops: Fix more gulong / 64-bit fallout from the xcb port 2015-06-23 15:09:32 -07:00
Jasper St. Pierre
734402e14d xprops: Cut off UTF-8 strings at n_items characters
There can be junk after here on some X servers.
2015-06-23 11:53:30 -07:00
Jasper St. Pierre
5d360a9bce xprops: Make sure that properties that failed to fetch are ignored 2015-06-23 11:52:49 -07:00
Jasper St. Pierre
f8cf5e373c mutter-Xatomtype: Fix the sizes of these structs
Since we're using xcb now, not Xlib, it doesn't have the stupid silly
thing where it copies all 32-bit data to 64-bit data to match the
long-y-ness of it.
2015-06-23 11:46:29 -07:00
Jonas Ådahl
bbf2b4e60e xprops: Don't pass ULONG_MAX as a uint32_t parameter
It seems the largest possible value is to be passed, so actually pass
that instead. Note that even though the name of the xcb_get_property
parameter is called 'long_length' its actually a uint32_t.

https://bugzilla.gnome.org/show_bug.cgi?id=751136
2015-06-18 11:38:00 +08:00
Jasper St. Pierre
2345b9c6ad xprops: Fix syntax error 2015-06-17 18:19:53 -07:00
Jasper St. Pierre
af7cc87bfa xprops: More XFree => g_free compatibility 2015-06-17 18:18:43 -07:00
Jasper St. Pierre
af3b599cbb xprops: Remove cvtINT32toInt
These are awful compatibility hacks for systems without native 64-bit
support. We can trash these now.
2015-06-17 18:18:03 -07:00
Jasper St. Pierre
0e73ceb4bd x11: Remove async-getprop
xcb can do this for us now.
2015-06-17 18:08:11 -07:00
Carlos Garnacho
4fc1811c15 wayland: Add X11/wayland selection interoperation
This piece of code hooks in both wl_data_device and the relevant X
selection events, an X11 Window is set up so it can act as the clipboard
owner when any wayland client owns the selection, reacting to
SelectionRequest events, and returning the data from the wayland client
FD to any X11 requestor through X properties.

In the opposite direction, SelectionNotify messages are received,
which results in the property contents being converted then written
into the wayland requestor's FD.

This code also takes care of the handling incremental transfers through
the INCR property type, reading/writing data chunk by chunk.

https://bugzilla.gnome.org/show_bug.cgi?id=738312
2015-05-15 17:43:53 +02:00
Jasper St. Pierre
ca6e799b97 Move atomnames.h to x11/
This should *not* be part of the public API.
2015-04-27 16:36:56 -07:00