Unify the behavior for eglfs, linuxfb and kms. The relevant code is
now moved from kms into fbconvenience.
From now on, on all three platforms, terminal keyboard input is turned
off by default. This feature can be disabled by setting
QT_QPA_ENABLE_TERMINAL_KEYBOARD to 1. This is similar to what the
evdev-based keyboard handler did in QWS in Qt4.
[ChangeLog][QtGui] The main Embedded Linux platform plugins (eglfs,
linuxfb, kms) are changed to behave identically with regards to
terminal keyboard input: it is turned off by default on all of these
platforms. If this feature is not desired, it can be disabled by
setting the environment variable QT_QPA_ENABLE_TERMINAL_KEYBOARD.
Task-number: QTBUG-36394
Change-Id: I69e47ed7580464dc5e703e9a0e23891c7c8b7790
Reviewed-by: Andy Nichols <andy.nichols@digia.com>
On boards that provide their own custom ways of querying the screen size
the resolution of widget apps became somewhat incorrect after the recent
eglfs/eglconvenience refactor. This is because the physical size query
helper was not able to access the screenSize() provided by the hooks,
it was instead falling back to the default fb query which in turn returned
the default screen size (e.g. 800x600).
Change-Id: I46d487b61341d69dd9cb76d93198b1f44b64f195
Reviewed-by: Andy Nichols <andy.nichols@digia.com>
Uploading the pixel data for the entire window on every update
is not nice. Not sure why the line in question was added, it is
most likely a leftover from some debugging session.
Change-Id: I161af536c42ac24b299fc674039f2318f0fb8c4f
Reviewed-by: Andy Nichols <andy.nichols@digia.com>
Prevent artifacts on raster windows by properly mapping
the coordinates to [-1,1].
Task-number: QTBUG-36370
Change-Id: I95d0d03952f597ef4ce6a950c6533a3af2df964a
Reviewed-by: Andy Nichols <andy.nichols@digia.com>
This didn't work for a loooong time already;
not it is possible to re-introduce this optimization once again.
Change-Id: I35c40e68933227bda6cbc038c2d8f6709db62451
Reviewed-by: Lars Knoll <lars.knoll@digia.com>
The MSG structure is supposed to contain screen coordinates of the mouse
position. Use GET_X/Y_LPARAM for mouse events and transform for
client coordinates. Use GetCursorPos() for other events.
Task-number: QTBUG-36337
Change-Id: I3ad4de20e1a460ee58f22645a4339a2444d129ed
Reviewed-by: Joerg Bornemann <joerg.bornemann@digia.com>
It was only used to access the QWindow which can be retrieved
using QPlatformBackingStore::window(), potentially causing a crash
when QWindowsBackingStore::resize() was called before the window
is shown.
Change-Id: I7c1bfce0f2c371d5d8847fd15fd1dc760b769a8f
Reviewed-by: Oliver Wolff <oliver.wolff@digia.com>
The correct rounding that respects ForceIntegerMetrics flag
is done just a few lines below.
Change-Id: I8e79ebaded90753092ecd3e99eece8df4c824362
Reviewed-by: Lars Knoll <lars.knoll@digia.com>
This feature (package_manifest) generates a basic application manifest
from a template provided by the mkspec or the developer. It is meant to
deliver an out-of-the-box build experience without attempting to
exhaustively cover all manifest options. It is meant to be a starting
point which allows the developer to customize the manifest further. It
also becomes the default package manifest generator for Windows Phone,
replacing autogen_wmappmanifest.
Common variables, such as the target executable, are populated by qmake
in the newly created manifest. Default icons are also created if needed,
as the build will fail without them. The input manifest can be set by
assigning a file name to WINRT_MANIFEST. Additional options are
documented in the .prf file. If an existing (non-generated) manifest is
already in the directory, it will not be overwritten.
Task-number: QTBUG-35328
Change-Id: I57576a17ff9d2b564c0828f815949cb26d276bfd
Reviewed-by: Oliver Wolff <oliver.wolff@digia.com>
When calling end the refcount should be greater than zero, as calls to
begin/end should match.
Change-Id: I49adbf6f9f1c21cb60985ba4e02574e6c0841b74
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@digia.com>
also add a boolean flag indicating if xsettings is initialized and can
be used
Change-Id: I9dae83b8fef224171744d2b940767cbf54ce4f55
Reviewed-by: Gabriel de Dietrich <gabriel.dedietrich@digia.com>
The fix in b8fe5e1bbc (for bug
QTBUG-27277) created two sets of begin() and end() functions for
QVector. That was required because QTypedArrayData::iterator and
QTypedArrayData::const_iterator have non-implicit copy constructors,
which means the variants returning a strict iterator need to pass the
storage pointer as an implicit first parameter. With the fix, the
compiler would emit functions with two different names for each
variant, which couldn't be merged.
If we remove those copy constructors, the compiler might be able to
generate the same code (no implicit first parameter) for both
functions.
Now, enter MSVC. Due to QPolygon, QVector<QPoint> and QVector<QPointF>
are "extern templates". That is, the compiler is not allowed to inline
anything, it must generate calls into QtCore (which we must fix for Qt
6, if we can). That means QtCore would only have one set of begin() /
end() functions. If an application tried to use the other set by
defining QT_STRICT_ITERATORS, you'd get a linker error for:
?begin@?$QVector@VQPoint@@@@QEAA?AViterator@?$QTypedArrayData@VQPoint@@@@V23@@Z
(class QTypedArrayData<class QPoint>::iterator QGenericArray<class QPoint>::begin(class QTypedArrayData<class QPoint>::iterator))
Change-Id: I5c10a7d0a4855f4ba84056d313c6a800ecdcfe37
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@digia.com>
Reviewed-by: Olivier Goffart <ogoffart@woboq.com>
Add ref qualified versions of QImage::convertToFormat, so that a
temporary QImage can be converted in-place to a format of equal depth.
[ChangeLog][QtGui][QImage]Added rvalue qualified convertToFormat method
for in-place conversion
Change-Id: I2eed5ffd63f5aea4ffa1147bf7607b02a49d9c5d
Reviewed-by: Gunnar Sletta <gunnar.sletta@jollamobile.com>
Adds example paths for the various locations returned on Android.
Change-Id: I55a4e23570c252cbf643596d166c7c43f023ba9c
Reviewed-by: BogDan Vatra <bogdan@kde.org>
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Setting this property makes the editor display a grayed-out
placeholder text as long as the document() is empty.
Change-Id: I997edb867419613ff7cedc760a87c684a2ded711
Reviewed-by: David Faure <david.faure@kdab.com>
According to my profiling of Qt Creator, qHash and the SHA-1 calculation
are the hottest spots remaining in QtCore. The current qHash function is
not really vectorizable. We could come up with a different algorithm
that is more SIMD-friendly, but since we have the CRC32 instruction that
can read 32- and 64-bit entities, we're set.
This commit also updates the benchmark for QHash and benchmarks both
the hashing function itself and the QHash class. The updated
benchmarks for the CRC32 on my machine shows that the hashing function
is *always* improved, but the hashing isn't always. In particular, the
current algorithm is better for the "numbers" case, for which the data
sample differs in very few bits. The new code is 33% slower for that
particular case.
On average, the improvement (including the "numbers" case) is:
compared to qHash only QHash
Qt 5.0 function 2.54x 1.06x
Qt 4.x function 4.34x 1.34x
Java function 2.71x 1.11x
Test machine: Sandybridge Core i7-2620M @ 2.66 GHz with turbo disabled
for the benchmarks
Change-Id: Ia80b98c0e20d785816f7a7f6ddf40b4b302c7297
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@digia.com>
Reviewed-by: Lars Knoll <lars.knoll@digia.com>
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Opacity was not set initially and not committed to libscreen when
changed.
Change-Id: Icf24be833de28bd72275cf93a3ab2fb344726023
Reviewed-by: Kevin Krammer <kevin.krammer@kdab.com>
Reviewed-by: Fabian Bumberger <fbumberger@rim.com>
clang is the default compiler on FreeBSD 10 (if not earlier).
Let's keep it unsupported for now. Can be promoted later.
Change-Id: I909953c986a3da09ce19d8f9f9ee2cc22c417abd
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@digia.com>
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Allows us to sanity check the iOS build in the CI.
Change-Id: I16f9bfafef3988dcab6efd3155503ca0d0b4d1d8
Reviewed-by: Morten Johan Sørvig <morten.sorvig@digia.com>
This tests
* default spacing
* explicit positive spacing
* negative spacings
Negative spacings that causes the layout to shrink down to 0 or
a negative spacing are not supported.
(But make sure they don't crash the whole engine)
This is a partial cherry picked from
qtquickcontrols.git/bc973dcf2163b25f2db74d974b252384bbee8d80)
Only the Q_ASSERT had to be cherry-picked. This is because of that the
qgridlayoutengine* is moved to qtbase.
Therefore, any changes to qgridlayoutengine* won't be merged properly to
dev (they will be lost).
Task-number: QTBUG-35741
Reviewed-by: Jens Bache-Wiig <jens.bache-wiig@digia.com>
Change-Id: I834f1efc11aca59a93c8d870f1a776f2f8810485
Reviewed-by: Caroline Chao <caroline.chao@digia.com>
Introduce logging categories and change most debug statements
to qCDebug().
Retrieve the filter rules from the environment variable
QT_QPA_VERBOSE (should be removed once the logging framework
offers a more convenient way of configuring).
Replace the old per-category variables controlling verboseness
of the output by a single variable which can be set on the command
line.
Change-Id: Iae55ecdb0e150efd165b7d3e90b371f72853f194
Reviewed-by: Joerg Bornemann <joerg.bornemann@digia.com>
The gutter position was hardcoded to 28 which fits when drawing
on the QMenu with a left margin of 3. It became offset when drawing
on a custom widget with no margin. Calculate position relative
to margin for left to right.
[ChangeLog][QtWidgets][QMenu][Windows] Fixed position of menu gutter when using a custom widget action.
Task-number: QTBUG-36218
Change-Id: Ibb34a9fba2bab7257482b80d2e51e76a9755345c
Reviewed-by: Joerg Bornemann <joerg.bornemann@digia.com>
Moc should check full scope of any related objects or
gadget when it constructs extra data.
Change-Id: Ibd1b607a389cd4e788c0916984464cd9103d9c59
Reviewed-by: Olivier Goffart <ogoffart@woboq.com>
Add support for using qCDebug and friends in the 'printf style' way.
This allows an almost mechanical conversion of existing qDebug, qWarning,
qCritical macros, and allows avoiding the size overhead the streaming
style incurs (mostly due to inlined QDebug code).
To handle this gracefully we require variadic macros (part of
C++11/C99). For compilers not supporting variadic macros we fall back
to checking the category in QMessageLogger.
[ChangeLog][QtCore][Logging] Allow qCDebug macros to be used in a printf
style.
Change-Id: I5a8fb135dca504e1d621bb67bf4b2a50c73d41b9
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
[ChangeLog][QtWidgets][QMenu][Windows] Correctly use text color set by style sheet for menu items.
Task-number: QTBUG-36142
Change-Id: I142f2fd82164771b45bf599a6afa1121fcfd2187
Reviewed-by: Jens Bache-Wiig <jens.bache-wiig@digia.com>
Some debug info and a #warning had somehow made it through :)
Change-Id: I420a7637bfd5d1a79a342eacefcc35430994feff
Reviewed-by: Alex Blasche <alexander.blasche@digia.com>
Task-number: QTBUG-35743
[ChangeLog][QtTest][Windows] Use correct UTF-8 encoding for XML test results on platforms with different console encoding.
Change-Id: Ice9d03192098f931e5dac358928e0c4421ab715e
Reviewed-by: Frederik Gladhorn <frederik.gladhorn@digia.com>
Any cipher that is < 128 bits is excluded from the default SSL
configuration. These ciphers are still included in the list
of availableCiphers() and can be used by applications if required.
Calling QSslSocket::setDefaultCiphers(QSslSocket::availableCiphers())
will restore the old behavior.
Note that in doing so I spotted that calling defaultCiphers() before
doing other actions with SSL had an existing bug that I've addressed
as part of the change.
[ChangeLog][Important Behavior Changes] The default set of
ciphers used by QSslSocket has been changed to exclude ciphers that are
using key lengths smaller than 128 bits. These ciphers are still available
and can be enabled by applications if required.
Change-Id: If2241dda67b624e5febf788efa1369f38c6b1dba
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Instead of merely intercepting logging output from stdout/stderr from the
journal side of started processes, this has the advantage of meaning that the
origin process name will be correct.
fprintf won't work, because if a process starts children (like e.g. a
homescreen does), then their stdout/stderr are merged into their parent, and
journal has no way of differentiating the origin.
We are also able to store information about the context of logging, which might
be useful in post-mortem cases.
[ChangeLog][Platform Specific Changes][Linux] Systems with systemd may now pass
-journald to configure to send logging output to journald. Logging will still be
sent to stderr for interactive applications (run from a tty) or with
QT_NO_JOURNALD_LOG set to a non-empty value.
Change-Id: Ib260cec1ea87390bf44f267d217d795583407d00
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
timezone is defined differently in FreeBSD compared to other
(more or less) POSIX-compliant Unices.
Task-number: QTBUG-36080
Change-Id: I4ad1a5ccd0b9ddbadb9fdd90edc26cf0c7252dfb
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
The event must be freed in all cases, also when a motion or configure
notify is skipped.
Change-Id: Ie55e62d0c3ab25bd9b20cc85098ea337d0b1ac5d
Reviewed-by: Uli Schlachter <psychon@znc.in>
Reviewed-by: Shawn Rutledge <shawn.rutledge@digia.com>
For some reason the Raspberry Pi hook includes qeglfscursor.h even
though it is not necessary. Remove this because the file got moved to
eglconvenience.
Change-Id: Ia65f5a8366d750f93eacee49004219e664b52af2
Reviewed-by: Andy Nichols <andy.nichols@digia.com>
Needed to run the XKB extension where we used to expect X11 to be
installed in /usr. In FreeBSD, for instance, X11 is installed in
/usr/local like all the other not out-of-the-box packages.
Same thing goes for the compose platform input context plugin.
Change-Id: Ib7ace3117eaacc6e150394450d1330cba654ba8a
Reviewed-by: Gatis Paeglis <gatis.paeglis@digia.com>
That way we can avoid name conflict with a namespace defined in
a different moc test
Change-Id: Id631d7c5556c9d6940e16dc53eb438dbcd0095eb
Reviewed-by: Olivier Goffart <ogoffart@woboq.com>
The qRemovePostRoutine() function has been added as publicly exported
function in Qt 2.2.0 with behavior equivalent to that of today. It has
never been documented for unknown reasons, possibly simply forgotten.
This function provides needed symmetry for the already documented
qAddPostRoutine() function and is used in some real world applications.
Change-Id: Ied4709505d8335c883e9791ea634df8fb406d995
Reviewed-by: Martin Smith <martin.smith@digia.com>
Reviewed-by: Venugopal Shivashankar <venugopal.shivashankar@digia.com>
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Reviewed-by: Topi Reiniö <topi.reinio@digia.com>
Reviewed-by: Nico Vertriest <nico.vertriest@digia.com>
There is no point in saving \0 twice. Any code that could relay on it
is broken anyway, because moc saves \0\0 for an empty string.
Change-Id: I28fa4f78aae8c883c088df43ec89d608a99b3bdd
Reviewed-by: Olivier Goffart <ogoffart@woboq.com>
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Values from different enums were compared.
Change-Id: I2faf73976e4b24abbdc915e8445256fe7a5be5bd
Reviewed-by: Stephen Kelly <stephen.kelly@kdab.com>
Reviewed-by: Jędrzej Nowacki <jedrzej.nowacki@digia.com>
