This more than halves the total runtime of this function since the
previous commit, from 8.36% to 4.02%, and is most likely memory
bandwidth limited on this specific board now.
I tried to do a SSE2 version as well, but couldn’t find any equivalent
of the LD4/ST4 ARM instruction.
On x86 on a Kaby Lake CPU, this makes it go from 6.63% of the total
execution time (loading some PNGs using the cairo backend) down to
3.20%.
On ARM on a Cortex-A7, on the same workload, this makes it go from 57%
to 8.36%.
We want our tracking area to be limited to the input region so that we
don't pass along events outside of them for the window. This improves the
chances we click-out of a popover with a large shadow.
This still doesn't let us pass-through clicks for large shadows on top-
level windows though.
We only should be asserting in static functions. Furthermore, this function
did not need to have GDK_BEGIN_MACOS_ALLOC_POOL as nothing is being
allocated there which would cause pooling to get used.
This needs to handle the boundary case where the value is exactly equal
to the edge of a rectangle (which gdk_rectangle_contains_point() does not
consider to be containing). However, if there is a monitor in the list
that is a better match, we still want to prefer it.
When using an external mouse on MacOS, the scrolling behavior is
reversed from the user's scrolling preference. Additionally, it is
noticeably sluggish.
This commit fixes both issues by negating the deltas and multiplying
them by 32 before constructing a new scroll event. 32 seems to be the
"traditional" scaling factor according to [Druid], but I'm not sure
where that value actually comes from. Regardless, scaling the deltas by
this amount makes scrolling feel a lot more responsive in the GTK demos.
Scrolling with a trackpad is not affected by either issue because it
triggers a different code path that uses more precise deltas, and
already negates them.
[Druid]: https://linebender.gitbook.io/linebender-graphics-wiki/mouse-wheel#external-mouse-wheel-vs-trackpad
Some Windows keymaps have bogus mappings for the Ctrl modifier. !4423 attempted
to fix this by ignoring the Ctrl layer, but that was not enough. We also need to
ignore combinations of Ctrl with other modifiers, i.e. Ctrl + Shift. For example,
Ctrl + Shift + 6 is mapped to the character 0x1E on a US keyboard (but it should
be treated as Ctrl + ^). Basically, always ignore Ctrl unless it is used in
conjunction with Alt, i.e. as part of AltGr.
Related issue: #4667
`free` is defined in `stdlib.h`, see for example
<https://pubs.opengroup.org/onlinepubs/009604499/functions/free.html>. Without
this include compilation can fail with the following error:
```
../gdk/loaders/gdkjpeg.c: In function ‘gdk_save_jpeg’:
../gdk/loaders/gdkjpeg.c:264:7: warning: implicit declaration of function ‘free’ [-Wimplicit-function-declaration]
free (data);
^
../gdk/loaders/gdkjpeg.c:264:7: warning: incompatible implicit declaration of built-in function ‘free’
../gdk/loaders/gdkjpeg.c:264:7: note: include ‘<stdlib.h>’ or provide a declaration of ‘free’
../gdk/loaders/gdkjpeg.c:302:67: error: ‘free’ undeclared (first use in this function)
return g_bytes_new_with_free_func (data, size, (GDestroyNotify) free, NULL);
^
../gdk/loaders/gdkjpeg.c:302:67: note: each undeclared identifier is reported only once for each function it appears in
../gdk/loaders/gdkjpeg.c:303:1: warning: control reaches end of non-void function [-Wreturn-type]
}
^
```
We don't want to risk having something really weird come out if we have a
WCG colorspace, so instead only do the performance hack on systems where
the output is likely reasonable.
We will want to eventually just be drawing in the appropriate colorspace,
but that is not available yet.
When using software rendering w/ cairo, assume we're drawing in
the best-monitor's colorspace rather than RGB to avoid colorspace
conversions on every frame.
Instead of relying on cairo_t to perform drawing from our backing
image surface to the Core Graphics context, we can convert the
cairo_image_surface_t into a CGImageRef without having to copy
data if we are certain of the alignment of the image up front.
Without this, there are many situations, based on the size of the
window that could cause cairo to take a slow path and malloc/copy
the data to ensure that alignment.
The previous commit titled "macos: align image surface rowstride to
16-bytes" ensures that this invariant is true so that our drawing
code can assume we can reference the framebuffer from the
cairo_image_surface_t using a CGDataProvider.
Since GdkMacosCairoContext and GdkMacosCairoSubview are coordinating,
we can also setup the transformation/scale early when drawing the
cairo_image_surface_t instead of when copying it to Core Graphics.
Furthermore, the CGImageRef is created with an RGB colorspace so
that we are not performing colorspace conversion to the output
device. We don't get color matching between displays, but we don't
expect that anyway, particularly with the software renderer.
When creating a cairo_image_surface_t we want both the framebuffer pointer
and each row to be aligned to 16-bytes so that Core Graphics will use more
optimal paths.
However, cairo_image_surface_create() will not guarantee that the rowstride
is aligned to 16-bytes so we must do that ourselves.
We need to avoid conflating the managing of frame callbacks from
the freeze/thaw mechanics and ensure we don't perform extra thaw
requests at the wrong time.
Some keymaps on Windows contain bogus mappings for Ctrl+key for certain
keys, e.g. Ctrl+Backspace = Delete, or Ctrl+[ = 0x1B. These are never
used on Windows, so we should ignore them.
Fixes#4667
GTK's old key symbol list is missing a few symbols like the per mille
sign that is included in some keyboard layouts. This commit updates
gdkkeyuni.c to match libxkbcommon's current key symbol list.
This change is done for 2 reasons:
- The logic to request this phase when compressing scroll events is
slightly broken. If there are multiple scroll events that are
coalesced into one, the surface frame clock will not get this request.
The worst case is having >= 2 scroll events on every frame, as the
compressed event will be left in the queue, and be further compressed
on future events.
- Even scroll events aside, this phase is requested in oddly specific
places that are not enough to cover all events, others do rely on
unrelated GdkFrameClock activity that happens to flush the events
as well.
Unify this phase request so it explicitly happens on the arrival of any
event. This ensures that events (compressed or not) will be handled
promptly after arrival.
As per Benjamin's suggestions, cleanup the previous implementation on
initializing the GLES context on Windows, so that we use more items that are
already in GDK proper and integrate two functions into one.
Instead of first trying to explicitly ask for a WGL 4.1 context, ask for
the WGL context version that matches what is reported via
epoxy_gl_version(), so that we get the maximum WGL version that is
supported by the graphics drivers, and make sure any WGL contexts that
are shared with this (initial) WGL context are created likewise.
We can try to do a default-bog-standard 3.2 core WGL context creation
if the need arises, but let's leave that alone for now.
The EGL context that we are actually creating must have matching OpenGL/ES
versions and allowed GL API set with the previously-created EGL context
that will be shared with it so that they can interoperate together, if
applicable.
This will fix the situation by making sure that we request for the
OpenGL/ES version and OpenGL API set that match with what we have in our shared
EGL context. Otherwise, the newly-created EGL context assumed a OpenGL/ES 2.0
context that supported desktop OpenGL, which may not be what we wanted, such as
in the case of libANGLE.
We are now able to create EGL contexts properly on Windows, but not GLES. This
tries to fix things by doing the following:
* Record the GL context type in a more proper fashion, using an Enum. This
makes things a bit cleaner.
* Force GLES-3.0+ contexts, since libANGLE requires this to properly work with
the shaders-its 2.0 contexts don't work well with our shaders.
We only save the size when we transition from floating to fixed, so that
we can restore the size to the one prior to being fixed.
However, we should not restore to this size whenever we see a 0x0 size
from xdg_toplevel, as it can do that any time it doesn't care about the
size, e.g. when the surface is floating and just changing state.
Fix this by only using the saved size when transitioning from fixed to
floating, not when staying floating while previously floating.
Closes: https://gitlab.gnome.org/GNOME/gtk/-/issues/4634
The actual code that does the IM context code handling on Windows now uses the
native Windows APIs to handle keystrokes, so this patch is no longer needed, as
it was found that it instead caused issues.
Pointed out in issue #2865.
This reverts commit fd6ce9975e.
gdk_wayland_toplevel_inhibit_idle() contained a contradictory assert
that always fail. More specifically, in the branch that is supposed to
create the idle inhibitor, there is an assertion that it must already
exist and that the refcount must be greater than zero. This causes a
crash on WMs/DEs that use the ZWP idle inhibit manager protocol such as
KDE Plasma and Sway. Fix this by just asserting that the refcount is
zero instead.
This makes the hotspot of DND surfaces work when using the Vulkan and
OpenGL renderers.
This bumps the CI image used to the newly built image. This is needed to
install a new enough libwayland-client.so needed for wl_surface.offset.
This is done by adding wayland as a meson subproject, building it
on-demand if the version in the system is not new enough. As
libwayland-client.so is pulled in implicitly when linking to gtk4, the
compile step needs LD_LIBRARY_PATH set to make ld find the right library
to link to.
For some users, GetKeyboardLayoutNameA() returns an alias instead of the
fully resolved keyboard layout identifier. In that case, we have to
query the registry to resolve the alias before we can look up the DLL
path.
See comments under https://gitlab.gnome.org/GNOME/gtk/-/issues/4610
Contrary to what you can read on the internet, SGCAPS keys don't work
by having capslock toggle the KBDCTRL bit, they actually have two
consecutive table entries, the first of which is for the normal
version and the second of which is for the capslocked version.
Background: SGCAPS is short for Swiss German caps because Swiss German
was the first layout to use this feature. For keys with the SGCAPS flag,
capslock has a different effect than pressing shift. For example:
Shift + ü = è, CapsLock + ü = Ü, CapsLock + Shift + ü = È
DLL loading failures should not happen under normal circumstances, but
we should at least try not to crash and and print better diagnostic
messages if they do happen.
See https://gitlab.gnome.org/GNOME/gtk/-/issues/4610
Previously, we treated CapsLock and KanaLock as part of the global
keyboard state, much like NumLock and ScrollLock, rather than using
the supplied modifier mask. This was because GDK does not have a
modifier mask for KanaLock, only for CapsLock, so it would not have been
possible to properly support it.
However, this approach ended up causing problems, with certain keyboard
shortcuts not registering when capslock was active. This was first
observed in Inkscape [0] and appears to affect shortcuts consisting of a
single key (like 'a') with no additional modifiers (wheareas shortcuts
like 'ctrl+a' work).
So now we are using the supplied GDK_LOCK_MASK instead, and dropped
support for KanaLock, which we probably don't need anyway (since regular
text input should be handled by the IME input module -- the keymap is
mainly for shortcuts and keybindings, where you don't really want
KanaLock).
[0] https://gitlab.com/inkscape/inkscape/-/issues/3082
The old code used repeated calls to `ToUnicodeEx` to populate
the translation table, which is slow and buggy. The new code
directly loads the layout driver DLLs from Windows.
See https://gitlab.gnome.org/GNOME/gtk/-/merge_requests/4338
libpng wants to receive samples in either RGB or RGBA order, whether
each sample is big-endian or not. This resolves test failures in
testsuite/gdk/memorytexture.c (and a lot of reftests) on s390x, and
probably the PowerPC family too.
Modifying the test to show the color in use and write out the PNG bytes
to a file, and running the memorytexture test on s390x, produces a PNG
that loads with the correct color values in GIMP (on an x86_64 machine),
which seems like evidence that this is the correct change and not just
compensating errors.
Resolves: https://gitlab.gnome.org/GNOME/gtk/-/issues/4616
Signed-off-by: Simon McVittie <smcv@debian.org>
If we ended up on no output at all, keep the HiDPI scale as is, as it
likely means we were on a workspace that was switched away from. By
keeping the same scale, we avoid unnecessary scale changes that would
otherwise take place if the scale when on monitors would end up being
more than 1.
Instead of just passing major/minor, pass them twice, once for GL and
once for GLES. This way, we don't need to check for GL and GLES
separately.
If something is supported unconditionally, passing 0/0 works fine.
That said, I'd like to group the arguments somehow, because otherwise
it's just a confusing list of numbers - but I have no idea how to do
that.
We want critical GL debug messages to be critical, so that the testsuite
sudokus itself when they appear.
This is relevant in particular for GLES warnings in the GLES runner,
because its warnings can cause crashes on GL drivers less forgiving than
Mesa.
Related: #4571
When destroying the EGLSurface or GLXDrawable of a GdkSurface, make sure
the current context is not still bound to it.
If it is, clear the current context.
Fixes#4554
We now have a boolean setting that determines whether the high-contrast
theme should be used. Support it by automatically setting the existing
`gtk-theme-name` and `gtk-icon-theme-name` properties when enabled.
With that, it is no longer necessary to change the regular theme settings
for high-contrast, so toggling between high-contrast and a non-default
theme finally works reliably.
It makes sense to connect the begin/update/end events
for touchpad swipes and pinches in a sequence. This
commit adds the plumbing for it, but not backends
are setting sequences yet.
We finish the write to the output stream long after the stream has been
closed, so we want to keep the event handler around to do just that.
Instead, remove the handler on finalize.
The OutputStream needs to write a 0 byte end of stream Property. We need
to track if that has been written, and we do that with that new
property.
We also use that property to always request flushes when the stream is
being closed, so that we don't wait for another flush() call.
We need to be very careful when writing data, because if we aren't, sync
functions will be called on the output stream and X11 does not like that
at all.
I saw this coming across through a ffi boundary in Sysprof, and we wanted
to keep most things within GDK using native marshalling to improve
profiler results when frame pointers are not used.
.. when creating the surface (with the HWND associated with the
newly-created surface) as well as destroying the surface (with NULL,
since the HWND is going to be destroyed), so that we can tie the EGL
calls to the HWND that we want to do the EGL stuff.
In some circumstances (e.g. activating with a stylus something that
closes a window), we can receive zwp_tablet_tool.proximity_out without
receiving a zwp_tablet_tool.up beforehand.
In those cases, we are not expecting neither .up nor .button, so
reset the stylus device button modifiers on proximity_out.
Fixes: https://gitlab.gnome.org/GNOME/gtk/-/issues/4103
We are looking up the seat logical pointer modifiers (i.e. the wl_pointer),
not the ones for the tablet tool device. This breaks accounting further
along in GTK leaving stuck implicit grabs.
Fixes: https://gitlab.gnome.org/GNOME/gtk/-/issues/4102
Update the functions that were updated in the previous commit to have all
GdkSurface variables named as 'surface' instead of the GTK-3.x-era window, to
make things more consistent across the board. Also fix formatting a bit.
Make the toplevel surface respond to size computations unless it is just being
created, or maximized, made fullscreen or underwent an AeroSnap operation.
This will ensure that the surface size is properly computed in time, so that
surfaces can be resized as needed.
This will fix issues 3728 and 3799.
As far as I can tell, the code here is redundant and probably ended up
this way for historical reasons. A drag surface without
`->is_drag_surface` would be created if `gdk_display_create_surface`
were called with `GDK_SURFACE_TEMP`, but drag surfaces never seem to be
created that way.
In `gtk4-demos`, drag and drop and popovers seem to be working normally
with this.
Ensure that we take the DPI scaling into account so that surfaces will
be placed at their correct positions upon an AeroSnap operation on HiDPI
displays.
Also, use the X coordinate of the surface as-is during snap up so that
we do not inadvertently move the surface to the very left. Also fix the
AeroSnap indicator drawing for snap up so that it is drawn at the
correct places.
Since we are updating these functions, make the old GdkWindow-era
variable names to match better the names we use nowadays.
