In some CI configurations this fails from time to time. Have to disable
it unless we can make it more robust somehow.
Change-Id: Iadd8904d7223a6aeff53dafa36b94df3f60e1ad8
Reviewed-by: Robin Burchell <robin.burchell@viroteck.net>
[ChangeLog][Important Behavior Changes] HarfBuzz-NG is now the default
shaper on all platforms. This results in a better shaping results
for various languages, better performance, and lower memory consumption.
Task-number: QTBUG-18980
Change-Id: I4d9454fc37e9050873df3857e52369dfc7f191b2
Reviewed-by: Eskil Abrahamsen Blomfeldt <eskil.abrahamsen-blomfeldt@digia.com>
Takes the total time for this test from ~1.2 seconds to ~0.42 seconds for me.
Change-Id: I426e600a7afe01d7343108b432eda8b83d6f3d85
Reviewed-by: Joerg Bornemann <joerg.bornemann@digia.com>
This takes 300ms off the total time for this test.
Change-Id: I230b8a58315797a81579f6a07da282b0dc40637d
Reviewed-by: Joerg Bornemann <joerg.bornemann@digia.com>
QExposeEvent::region() reports a region in a random coordinate system.
The behavior is undocumented and the platform plugins do different things.
xcb, offscreen and ios are correct. These set the region in local coordinates,
which is the most logical interpretation of the expose region.
windows is almost correct, except for one occurrence.
cocoa and others need changes: passing in geometry() as the exposed region is
always wrong.
The patch documents the expected behavior both for QExposeEvent and
internally in QWindowSystemInterface. The problematic plugins are fixed to
use local coordinates.
Task-number: QTBUG-40470
Change-Id: I6ded3154d14254fa71d4292d8e1b5e6cf696c81a
Reviewed-by: Gunnar Sletta <gunnar.sletta@jollamobile.com>
Instead of abusing inheritance to provide convenience constructors,
use simple inline constructor functions. The name got a lower-case
q to indicate a free function. The usual fromXYZ() static methods
were deemed not fitting in this case in the initial round of review,
since they implied some kind of conversion while these functions are
simply constructors of formsts, which contain no data.
This also solves the problem that some of these ctors could have been
called with just one argument and were therefore candidates for hidden
QPixelFormat temporary injection.
QPixelFormatRgb was renamed to qPixelFormatRgba to explain the third
argument at the call site better.
There seem to be no users of this class in qt5.git at this time.
Change-Id: Ib4fe8ceb2d30744127b116a748724a3406400eb8
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
QOpenGLBuffer::map() and related helpers are becoming useless in
OpenGL ES 3.0 and up: instead of the old GL_OES_map_buffer,
glMapBufferRange, but not glMapBuffer, is now part of the standard.
On desktop GL_ARB_map_buffer_range is present by default in OpenGL 3.0
and newer.
[ChangeLog][QtGui] Added QOpenGLBuffer::mapBufferRange().
Task-number: QTBUG-38168
Change-Id: I4e9bbe8ced9ee4d535ac32849a8c08c26d79cb49
Reviewed-by: Gunnar Sletta <gunnar.sletta@jollamobile.com>
Add defaultFormat() and setDefaultFormat() statics to QSurfaceFormat.
These define the default values for the requestedFormat members of
QOpenGLContext, QWindow and QOpenGLWidget (and implicitly QOpenGLWindow,
QQuickWindow, etc.)
This replaces QQuickWindow::setDefaultFormat() which can now be removed.
The main inspiration here is not the convenience (avoiding setFormat() calls
for all windows/widgets), but robustness: by setting the format once at the start
of the application, all windows and contexts, including the internal share context
used by QOpenGLWidget and QQuickWidget, will use the same format, eliminating
the possibility of failing due to trying to share between incompatible contexts.
Furthermore, since such a functionality is anyway mandatory for QQuickWindow
(due to the possibility of creating windows from QML code), extending it to
QSurfaceFormat and QOpenGLContext/QWindow is the next logical step.
Change-Id: Ie94486adc489d17fecfcebb7050fecedffd2688b
Reviewed-by: Gunnar Sletta <gunnar.sletta@jollamobile.com>
It isn't a thorough check of all the mutable methods for detaching, but a start.
Change-Id: I523fd30d3459186654e12fd25c384ed990ab7a00
Reviewed-by: J-P Nurmi <jpnurmi@digia.com>
As with many other implicitly shared classes, efficient move semantics
requires setting the d-pointer to nullptr, which then needs to be checked
for in the dtor and the copy assignment operator.
Change-Id: I654d181a1dfdd9a16e2f9fb96b57475cdd0b4561
Reviewed-by: J-P Nurmi <jpnurmi@digia.com>
Reviewed-by: Olivier Goffart <ogoffart@woboq.com>
A simple convenience class providing a QWindow that has a
paintEvent and supports opening a painter on itself. It
behaves exactly like QOpenGLWindow in this respect, which is not
surprising since they share the same base class (QPaintDeviceWindow).
QRasterWindow does not however have any OpenGL dependencies and will
be present in -no-opengl builds too.
[ChangeLog] Added QRasterWindow, a thin convenience wrapper for a
QWindow on which a QPainter can be opened.
Done-with: Jorgen Lind <jorgen.lind@digia.com>
Change-Id: I37e82720492945d7b85d5f713eea8d5f7556e511
Reviewed-by: Allan Sandfeld Jensen <allan.jensen@digia.com>
[ChangeLog] Added QOpenGLWindow. This serves as a convenience class for
creating windows showing OpenGL content via an API similar to QGLWidget
and without any widget dependencies.
Done-with: Jorgen Lind <jorgen.lind@digia.com>
Task-number: QTBUG-36899
Change-Id: I52e9bc61acb129dbfd3841b3adeffab2dbcf7f05
Reviewed-by: Gunnar Sletta <gunnar.sletta@jollamobile.com>
QImage::invertPixels may produce invalid data after inversions of
images with premultiplied alpha, because the inverted colors will be
larger than the alpha.
This patch converts any image with a premultiplied alpha channel to
ARGB32 before inverting the pixels, and then back to the original
format after the inversion.
Support is added for correct inversion of RGBA8888 and RGB30 formats.
Task-number: QTBUG-39901
Change-Id: Ief24c55f495e67ef2ad6429b5b418d02963a64dd
Reviewed-by: Gunnar Sletta <gunnar.sletta@jollamobile.com>
Use sizes relative to the test window size; move windows relative
to the top left point of the available screen geometry.
The test now passes on Windows using a 4K monitor.
Task-number: QTBUG-38858
Change-Id: Ia8d992f2a9bfa1cb1deacaf918ed0cfff7616959
Reviewed-by: Oliver Wolff <oliver.wolff@digia.com>
Reviewed-by: Joerg Bornemann <joerg.bornemann@digia.com>
This patch adds support for binding RGB30 images as textures, and as
internal format of framebuffer objects. Together with the
QOpenGLPaintDevice
this provides support for rendering to and from RGB30 in full precision.
[ChangeLog][QtGui][QOpenGLFramebufferObject] Support 10-bit per color
channels formats as the internal framebuffer format, making it possible
to render in that precision.
Change-Id: I06de2d12dfe1c1adc466d574fdffbc77f88f4f16
Reviewed-by: Laszlo Agocs <laszlo.agocs@digia.com>
This function has been inlined in commit 4f50be8d,
but wasn't removed.
Change-Id: I1a4eb7399ab2dd87a9fb15c4c886004d777487a8
Reviewed-by: Jędrzej Nowacki <jedrzej.nowacki@digia.com>
Several tests require test data to be deployed with the
application. The easiest way to achieve this on Android
is to add them to a qrc file and use the QFINDTESTDATA macro
to look up the files. This fixes several test failures
in the gui/text subdirectory for Android.
Change-Id: If944bb1fc93434a1b2d6487da829d21bd6b84e87
Reviewed-by: BogDan Vatra <bogdan@kde.org>
I've also updated the qfont test so it passes on Android now. Note that
there are no suitable cursive/fantasy fonts on Android, so the regular
default "Roboto" (or "Droid Sans" on older devices) will be picked here
instead.
[ChangeLog][Android] Fixed font selection to prefer "Droid Serif" when
Serif style hint is set on QFont.
Change-Id: I294eebcc4d79410e435bdddce552acc6044753b2
Reviewed-by: Christian Stromme <christian.stromme@digia.com>
Remove old message handler hack as we have a new and better api to
ignore warning messages.
Change-Id: Id967b2672fe3e3638db9977500118a19c2afb730
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@digia.com>
When the platform plugin gives us specific fallbacks for a
specific script, we need to respect this, and load the correct
font family regardless of which writing systems it supports.
This is especially important since the common script is adapted
to match surrounding, proper scripts, so characters such as
digits next to e.g. Hebrew text will be marked as Hebrew. On
stock Android, there is a single Hebrew font, and this would
previously be put in all fallback slots for Hebrew regardless of
what fallback fonts were dictated by the platform plugin. Since
this font does not support the digits, they would show up as boxes.
[ChangeLog][Android] Fixed common characters like digits and
punctuation showing as boxes when positioned next to non-latin
scripts.
Task-number: QTBUG-39377
Change-Id: I1555e208a8ddc587c0bbdbfff1600cafdd9442e9
Reviewed-by: Konstantin Ritt <ritt.ks@gmail.com>
At some point since Qt 4.8, the nbsp character has acquired
the whitespace attribute, causing it to be treated exactly
like an ordinary space. To account for this, we add an extra
check in the layout code to avoid breaking on non-breaking
spaces even if they have the whiteSpace flag set.
This is a temporary fix for the regression. The line breaking
algorithm needs to be refactored and support Unicode tr14
properly, which it currently doesn't.
[ChangeLog][Text] Fixed lines breaking on non-breaking spaces.
Task-number: QTBUG-39832
Change-Id: Ibd7e1a11ce4b82c611ecda1542c8638a67bf3cae
Reviewed-by: Konstantin Ritt <ritt.ks@gmail.com>
This is trickier than the GLX and EGL implementations due to the
way pixel formats, windows and contexts work. Apart from some
restrictions, it should be fully functional nonetheless.
Add also some proper documentation.
Change-Id: Ia6e3eb1ab2701e439b8621b9092c2b0934ff2151
Reviewed-by: Jørgen Lind <jorgen.lind@digia.com>
Currently QOpenGLTextureCache::bindTexture always convert any uploaded
image to RGBA8888 before uploading. This is quite inefficient when
OpenGL natively supports uploading formats in the original format.
This patch adds support for uploading a few native QImage formats. This
also get the performance of QOpenGLTextureCache::bindTexture on par with
QGLContext::bindTexture.
The texture brush used by QOpenGLPaintEngine is also converted to QImage,
since bindTexture will convert it to QImage anyway, and going over QPixmap
may cause an unnecessary conversion.
[ChangeLog][QtGui][QOpenGLTextureCache] Support uploading common QImage
formats directly to OpenGL when supported.
Change-Id: I828a763126441a98e4547c32ef52dddf7c129a32
Reviewed-by: Gunnar Sletta <gunnar.sletta@jollamobile.com>
Adds basic support for 10-bit per color channel formats to QImage
and the XCB plugin. This will make it possible to paint to and from
these formats, but only at 8-bit per color channel accuracy.
This also fixes Qt5 applications on X11 with native 30bit depth.
[ChangeLog][QtGui][QImage] Added support for 10-bit per color channel image formats.
Task-number: QTBUG-25998
Change-Id: I93ccd3c74bfbb0bd94b352476e5fe58a94119e1f
Reviewed-by: Gunnar Sletta <gunnar.sletta@jollamobile.com>
Support CF_DIBV5 for these images so that transparency is preserved.
Task-number: QTBUG-11463
Change-Id: I51881ae8bfbd05b92abd309766f4da9a7ea26c2d
Reviewed-by: Joerg Bornemann <joerg.bornemann@digia.com>
There is no guarantee that the OES extension is present and the standard
functions are not required to be dynamically resolvable on ES.
By performing the ARB-style lookup for the suffixless function names we
can also support ES3 compatible contexts on desktop GL.
This also fixes the problem of picking up the APPLE extension functions
instead of ARB when both are available. vaoFuncsType was set to ARB correctly
but the helper has to take the preference of ARB into account too.
Task-number: QTBUG-38168
Change-Id: If7402320e8f96891017674f3c43bc57c4b5d29f3
Reviewed-by: Sean Harmer <sean.harmer@kdab.com>
This allows QQuickSystemPalette to listen to palette changes without
installing an expensive event filter on the application object.
Change-Id: I8b693e047d993c444e393d7a714a5709692c3560
Reviewed-by: Shawn Rutledge <shawn.rutledge@digia.com>
Reviewed-by: Robin Burchell <robin+qt@viroteck.net>
Conflicts:
mkspecs/qnx-x86-qcc/qplatformdefs.h
src/corelib/global/qglobal.h
src/network/socket/qnativesocketengine_winrt.cpp
src/plugins/platforms/android/androidjniaccessibility.cpp
src/plugins/platforms/windows/qwindowswindow.cpp
Manually adjusted:
mkspecs/qnx-armle-v7-qcc/qplatformdefs.h
to include 9ce697f2d5
Thanks goes to Sergio for the qnx mkspecs adjustments.
Change-Id: I53b1fd6bc5bc884e5ee2c2b84975f58171a1cb8e
Use SetWindowPlacement() to set the normal position when applicable
as is done in Qt 4.
Task-number: QTBUG-39544
Change-Id: Ia158b968ea15361d9937619f07b56eb8a0312a13
Reviewed-by: Oliver Wolff <oliver.wolff@digia.com>
This allows QQuickApplication to listen to layout direction changes
without installing an expensive event filter on the application object.
Change-Id: I2d7d8906acecbc092657c4bd918bbdc9aad9744c
Reviewed-by: Shawn Rutledge <shawn.rutledge@digia.com>
The local files in availableSizes() were not resolved.
Introduce member variables for the file names, resolve them
in the constructor and add initTestCase() to verify.
Task-number: QTBUG-39287
Change-Id: If841e904700fe76b6c9265124ccba7764911fdc9
Reviewed-by: aavit <eirik.aavitsland@digia.com>
QImage::fill(uint) was incorrectly performing ARGB->RGBA conversion when
called on RGBA8888 formated images.
This patch moves the color conversion to QImage::fill(QColor) where it
belongs so that fill(uint) can behave consistent with documentation and
how it treats other formats.
The fill(uint) method had no automated tests, and this patch adds one.
[ChangeLog][QtGui][QImage] QImage::fill(uint) now fills the given pixel
value unconverted when used on RGBA8888 image, making it consistent with
the documentation and treatment of all other image formats.
Change-Id: I00a9d810c61d350dbdd7c4b9ad09e5ce11896b6d
Reviewed-by: Gunnar Sletta <gunnar.sletta@jollamobile.com>
This simple patch was an unfortunate victim of the Gitorious to
Gerrit transition. (https://qt.gitorious.org/qt/qt/merge_requests/1179)
As noted in the Gitorious review, a small bug in readHeader is also fixed: ||
instead of &&.
Done-with: Pali Rohár
Task-number: QTBUG-12684
Change-Id: I1fe16359b9b68c10e518904c6a5c58b00fb7379b
Reviewed-by: Yoann Lopes <yoann.lopes@digia.com>
The sanity check added in d16508a285
didn't actually catch the case where the invalid data is large
enough to contain the offset table and table directory. Added sanity
checks to all the code that accesses the font data now, so this
should fix crashes with partial data as well as invalid data.
Task-number: QTBUG-37190
Change-Id: Ie43f10d8cf0b09007783b9b1c4d91bfed8c6b0f0
Reviewed-by: Konstantin Ritt <ritt.ks@gmail.com>
