Commit Graph

8960 Commits

Author SHA1 Message Date
Jonas Ådahl
37f4c644d3 gdk/surface: Reset position after hiding
The x and y position of GdkSurface is only valid when visible, so reset
them to their inital state after hiding.
2020-02-19 09:47:18 +01:00
Jonas Ådahl
46d40cd4bd wayland: Simplify logic deciding whether to use xdg_popup
From a handful of hueristics and guess-work, to a single
`surface->surface_type == GDK_SURFACE_POPUP`.
2020-02-19 09:47:18 +01:00
Jonas Ådahl
11dbc384ab wayland: Fix top-most-popup check
We can map a non-grabbing popup wherever, it's just the grabbing
popup-chain that needs to be ensured not to break any ordering rules.

Fix this by managing two lists; one of open popups, and another for
grabbing ones.
2020-02-19 09:47:18 +01:00
Jonas Ådahl
ebd3ee9880 gdk: Move GdkGeometry declaration to gdktypes.h 2020-02-19 09:47:18 +01:00
Jonas Ådahl
339a04da7c gdk: Move GdkGravity to gdktypes.h 2020-02-19 09:47:18 +01:00
Jonas Ådahl
40693412d3 x11: Make gdk_surface_get_position() return relative position
The returned position should be relative to the parent surface, but
GdkSurface::x,y were only managed properly for O-R windows. This makes
it correct for regular windows too.
2020-02-19 09:47:18 +01:00
Jonas Ådahl
2bb0eddfb5 wayland: Simplify popup parent discovery
Now popups surfaces are always created with the parent set, so we don't
need to implement vorious guess work to try to find what the parent
might be. Remove that code and just use GdkSurface::parent which is
where the parent set during construction ends up at.
2020-02-19 09:47:18 +01:00
Jonas Ådahl
49fcbc27ca wayland: Remove old "gdk-attached-grab-surface" hackery
The corresponding code in gtk/ is long gone, so it's luckly unused these
days.
2020-02-19 09:47:18 +01:00
Jonas Ådahl
7fafa5133b wayland/surface: Add per surface configuration event queues
Add event queues specifically for surface configuration events
(xdg_surface.configure, xdg_toplevel.configure, xdg_popup.configure etc)
so that a configuration can be completed without having side effects on
other surfaces. This will be used to synchronously configure specific
GdkSurfaces, as is needed by the Gtk layout mechanisms.
2020-02-19 09:47:18 +01:00
Jonas Ådahl
0d086872ae gdk/wayland/surface: Use dedicated bool for frame callback freezing
The freezing is conditioned on various state, so lets make the thawing a
bit more robust. Without this there was a risk that we'd thaw too many
times if there was a frame callback requested while the conditions for
the freezing were not met.
2020-02-19 09:47:18 +01:00
Jonas Ådahl
f284e1d483 wayland: Remove position method split
Position can only be done via gdk_surface_queue_relayout(); the legacy
way is no longer possible, so remove the legacy positioning code.
2020-02-19 09:47:18 +01:00
Jonas Ådahl
4e78dd2ae0 gdk/surface: Use enum type for surface type 2020-02-19 09:47:18 +01:00
Benjamin Otte
47c8026e38 Merge branch 'wip/otte/dnd' into 'master'
some DND fixes

See merge request GNOME/gtk!1440
2020-02-15 23:45:25 +00:00
Benjamin Otte
46bbb397e1 wayland: Use a magic mime type for local DND
Otherwise the compositor gets all confused when it's trying to make
drag happen but we know it's not going to happen.

After all, we exchange data behind its back, we just need to keep it
informed.
2020-02-15 22:33:57 +01:00
Benjamin Otte
a824da66d9 drop: Ensure all relevant formats are available
1. GdkDrop does deserialization, so add the deserialize formats

2. If the drop is local, we can copy straight from the drag, so we can
   also copy all its formats. This fixes cases where the backend would
   drop formats it doesn't support.
2020-02-15 20:02:44 +01:00
Matthias Clasen
e5d83eea30 events: Make proximity and scroll events have tools
The Wayland backend tries to set device tools on these
events, and it was just an oversight that they don't
carry them.
2020-02-15 09:44:34 -05:00
Matthias Clasen
8ef7013165 Drop gdk_surface_new_temp
GTK is not using this API anymore, and we want
to consolidate our surface types to just toplevels
and popups.
2020-02-14 21:20:12 +01:00
Matthias Clasen
b6cb1a6eb3 Merge branch 'master-subtract-base-size' into 'master'
gdk: Subtract base size when checking aspect ratio

See merge request GNOME/gtk!1331
2020-02-14 13:19:42 +00:00
David Hogan
c4111dad31 Fixed OpenGL extension detection for extensions promoted to OpenGL core.
For a given OpenGL context, macOS in particular does not support enumeration / detection of OpenGL features that have been promoted to core OpenGL functionality. It is possible other drivers are the same. This change assumes support for GL_ARB_texture_non_power_of_two with OpenGL 2.0+, GL_ARB_texture_rectangle with OpenGL 3.1+ and GL_EXT_framebuffer_blit with OpenGL 3.0+. I failed to find definitive information on whether GL_GREMEDY_frame_terminator has been promoted to OpenGL core, or whether GL_ANGLE_framebuffer_blit or GL_EXT_unpack_subimage have been promoted to core in OpenGL ES. This change results in a significant GtkGLArea performance boost on macOS.

Closes #2428
2020-02-14 05:41:01 +11:00
Emmanuele Bassi
cdd7e4c5ec Add a notify function to GdkContentProvider
The callback-based content providers need a GDestroyNotify function to
free the data passed to them on construction, otherwise they are going
to leak.
2020-02-13 14:53:25 +00:00
Benjamin Otte
052d0f6e60 gdk: Remove gdk_cairo_get_clip_rectangle()
The function is fundamentally broken for unbounded surfaces.
If a surface is unbounded, we cannot represent this as a
cairo_rectangle_int_t, and using the return value doesn't work because
it's already used for something else.

In GTK3, unbounded surfaces aren't a problem, but GTK4 uses recording
surfaces.

So better remove that function before we keep using it and using it
wrong.
2020-02-13 07:36:38 +01:00
Alexander Larsson
6e3dd6dd62 profile: Use separate names for frameclock marks instead of using details
The marks are averaged based on the name, so this makes more sense.

Also rename the map/unmap marks to have the same capitalization as
everything else.
2020-02-12 14:29:31 +01:00
Alexander Larsson
d9c59a7c8c profiler: Add G_GNUC_PRINTF markers to silence warnings
I was getting CI failures like:
 ../gdk/gdkprofiler.c: In function ‘add_markvf’:
../gdk/gdkprofiler.c:111:3: error: function ‘add_markvf’ might be a candidate for ‘gnu_printf’ format attribute [-Werror=suggest-attribute=format]
2020-02-12 13:36:04 +01:00
Alexander Larsson
324202a56c profiler: Clean up profiler marks for frameclock
This drops the marks for before/after-paint as they are internal
things that very rarely use any time, and also flush/resume-events
as any events reported here will get separate marks so will be easy
to see anyway.

Also, we rename the entire frameclock cycle to "frameclock cycle"
rather than "paint_idle" which is rather cryptic.
2020-02-12 13:20:32 +01:00
Alexander Larsson
e3a67385f4 profiler: Add profiler marks for when surfaces are mapped and unmapped
This is done on the X11 side so we can tell when the map/unmap takes
effect and how long it took since the map request.
2020-02-12 12:44:43 +01:00
Alexander Larsson
2890cd849f profiler: Add _end_mark() version of _add_mark()
These don't take a duration, instead they call g_get_monotonic_time() to
and subtract the start time for it.

Almost all our calls are like this, and this makes the callsites clearer
and avoids inlining the clock call into the call site.
2020-02-12 11:25:34 +01:00
Alexander Larsson
01d5ad2056 profiler: Make profiler-is-running a macro
When we use if (GDK_PROFILER_IS_RUNNING) this means we get an
inlined if (FALSE) when the compiler support is not compiled in, which
gets rid of all the related code completely.

We also expand to  G_UNLIKELY(gdk_profiler_is_running ()) in the supported
case which might cause somewhat better code generation.
2020-02-12 11:05:01 +01:00
Alexander Larsson
cc643df88b Convert all profiler times from nsec to usec
usec is the scale of the monotonic timer which is where we get almost
all the times from. The only actual source of nsec is the opengl
GPU time (but who knows what the actual resulution of that is).

Changing this to usec allows us to get rid of " * 1000" in a *lot* of
places all over the codebase, which are ugly and confusing.
2020-02-12 10:44:17 +01:00
Alexander Larsson
685fd3c33d fixup! Add gdk_profiler_add_markf() to do printf formating 2020-02-12 10:11:53 +01:00
Alexander Larsson
8a1c51db1c Use the new gdk_profiler_add_markf() function 2020-02-12 09:59:03 +01:00
Alexander Larsson
6de85c0a68 Add gdk_profiler_add_markf() to do printf formating
This allows us to avoid hand-rolling g_strdup_printf calls,
but also moves the printf into the called function where
it doesn't bloat the code of the calling function if the profiler
is not running.
2020-02-12 09:58:52 +01:00
Alexander Larsson
36808e0c81 profiler: Always mark events with the kind 2020-02-11 19:51:49 +01:00
Emmanuele Bassi
c69595a3a9 vulkan: Handle VK_ERROR_UNKNOWN
A newly added Vulkan 1.2 error.
2020-02-11 14:47:22 +00:00
Emmanuele Bassi
1806af1e83 Declare global counter only in debug builds 2020-02-11 14:47:22 +00:00
Emmanuele Bassi
62352dd1fc Fix compiler warnings in the release build
We have a bunch of debug-only variables that get set and never used
outside of debug code paths, and the compiler is not happy about it.
2020-02-11 13:39:52 +00:00
Emmanuele Bassi
02ac7ff0aa Initialise out variable
Otherwise we might return an undefined value, and the compiler will be
very cross at us.
2020-02-11 13:39:52 +00:00
Alexander Larsson
23211bcae8 hidpi: Rescan monitors when screen scale factor changes
All the code in e.g. init_randr15() divides the physical resolutions with
the screen scale, however if we get the screen scale from xsettings
rather than e.g. GDK_SCALE the initial setup is using the wrong value.

So, whenever the screen scale size is changed we need to trigger
a re-read of the randr data
2020-02-10 15:34:32 +01:00
Matthias Clasen
07988fc17f win32: One more build fix 2020-02-09 23:04:10 -05:00
Matthias Clasen
4f79676b0b win32: Fix the build 2020-02-09 22:58:53 -05:00
Matthias Clasen
900ff743f1 Merge branch 'main-cleanup' into 'master'
Cleanup in gtkmain.h

See merge request GNOME/gtk!1403
2020-02-10 03:55:11 +00:00
Benjamin Otte
fce37b137d gdk: Make gdk_keyval_name() return a const char
The string is const, so return it like that.
2020-02-09 16:50:22 -05:00
Matthias Clasen
ba825dce29 docs: Remove a mention of gtk_main_do_event 2020-02-09 10:37:24 -05:00
Matthias Clasen
be2ba89900 Fix a doc comment
Signals take two colons.
2020-02-08 19:27:32 -05:00
Matthias Clasen
8703c7b7ac Merge branch 'better-test-isolation-2' into 'master'
Better test isolation

See merge request GNOME/gtk!1391
2020-02-06 04:53:07 +00:00
Benjamin Otte
0ad791277a Merge branch 'wip/otte/css' into 'master'
Some CSS cleanup

See merge request GNOME/gtk!1389
2020-02-05 14:04:12 +00:00
Matthias Clasen
9a446f2041 frameclock: cosmetics
Redo the profiler marks for the frame clock signals
to look more uniform.
2020-02-05 06:28:39 -05:00
Matthias Clasen
9e6ae7acd8 gdk: Silence some compiler warnings
These cause unused variable warnings in release builds.
2020-02-05 06:28:39 -05:00
Benjamin Otte
546a748cd8 widget: Replace style-updated signal with css_changed vfunc
1. Rename the thing
2. Turn it from a signal to a vfunc
3. Pass the GtkCssStyleChange to it

We don't export any public API about the GtkCssStyleChange yet, it's
just a boring opaque struct.
2020-02-05 02:46:13 +01:00
Benjamin Otte
7e8d8218b9 Revert "Break out cairo color matrix recoloring to gdk_cairo_image_surface_recolor()"
We're not in the business of adding Cairo APIs. That's Cairo's job.

Also, we don't need this API anywhere like the original commit claimed,
so there's no need to make it available in any way.

This reverts commit afa6cc2369.
2020-02-04 16:43:59 +01:00
Alexander Larsson
85e59220a5 filechooser portal: Free portal proxy when bus connection closes
This was blocking the clean exit from the testdbus shutdown in
the defaultvalues test. The proxy was keeping the connection alive
which blocks g_test_dbus_down().
2020-02-04 14:08:10 +01:00