The duo Q_OBJECT_BINDABLE_PROPERTY and QObjectBindableProperty
can only be documented together. The documentation is now with
QObjectBindableProperty. This patch adds a documentation entry
for Q_OBJECT_BINDABLE_PROPERTY which links the user to
QObjectBindableProperty.
Task-number: QTBUG-90511
Pick-to: 6.0 6.1
Change-Id: I9af4a99d49f4b02ee9645a2cc9a9a024a6a1a552
Reviewed-by: Fabian Kosmale <fabian.kosmale@qt.io>
Reviewed-by: Paul Wicking <paul.wicking@qt.io>
Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
And refactor the TMP to use std::conjunction and use variable template
specialization instead of template class specialization for the base
cases.
Change-Id: Iea6a03f13ea3443a0fa7365af21c496670c1e07f
Reviewed-by: Qt CI Bot <qt_ci_bot@qt-project.org>
Reviewed-by: Lars Knoll <lars.knoll@qt.io>
These types can be copied via memcpy, so use the correct tag.
Change-Id: Icba37d42dd8658225ce4889284661d108122d489
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Reviewed-by: Lars Knoll <lars.knoll@qt.io>
Reviewed-by: Andrei Golubev <andrei.golubev@qt.io>
This harmonizes the execution between start() and startDetached(). Both
did QStandardDirs::findExecutable(), but duplicated code. However, only
start() supported launching .app bundles on Mac.
[ChangeLog][QtCore][QProcess] startDetached() now supports launching
.app executable bundles on macOS / iOS systems, like start() already
did.
Change-Id: Ic90d8429a0eb4837971dfffd1664f776b2c0edfd
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@gmx.de>
qregularexpression.h requires FEATURE_regularexpression to be enabled.
Add a condition to the 'qt_pch.h' header to fix the build when PCH
are enabled and FEATURE_regularexpression is disabled.
It seems that the filter implementation of QSortFilterProxyModel
has QRegularExpression as its base. It's necessary to make
sortfilterproxymodel dependent on the regularexpression feature.
Fix the precompiler condition for the extractSections function in
the QString implementation. Use the same precompiler condition for
the QString::section and extractSections functions, since
QString::section depends on extractSections.
Change-Id: I5b775e0842a0aa1a8d47f8dded376bdfcf63b5bf
Reviewed-by: Joerg Bornemann <joerg.bornemann@qt.io>
Like for LibraryExecutables, we must default to "bin" on Windows for
HostLibraryExecutables in our generated qconfig.cpp.
Pick-to: 6.1
Fixes: QTBUG-91496
Change-Id: Ib5a4b3b3fd6192bd953e615058b482e67ad19462
Reviewed-by: Cristian Adam <cristian.adam@qt.io>
This feature only existed to trigger the evaluation of the libs gdi32,
kernel32 and friends. Those libs only were relevant for the qmake build
of Qt and can be removed now.
Task-number: QTBUG-83932
Change-Id: Idfd4d8f70b6dbd1067412fecc1115b504b1ad347
Reviewed-by: Alexey Edelev <alexey.edelev@qt.io>
Reviewed-by: Qt CI Bot <qt_ci_bot@qt-project.org>
Reviewed-by: Kai Koehne <kai.koehne@qt.io>
One of the changes done in 902505a058
results in a compilation error: somehow an expression
"int * enum value (with underlying type qint64)" has result type
"long int" and thus the compiler cannot find matching add_overflow
Return the qint64 cast back to overcome this
Compiler: gcc 7.5.0-3ubuntu1~18.04
Change-Id: Iaca882762e812bef69ec325df5f59e02082a0130
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Follow up of 612a01be65
(deprecating QSP::take()): for the same reasons, swap()
functions do not belong to QScopedPointer, or they would
allow the pointer to escape:
QScopedPointer a;
{
QScopedPointer b = ~~~;
qSwap(a, b);
}
// b's pointer escaped its scope
Deprecate them as well.
[ChangeLog][QtCore][QScopedPointer] QScopedPointer swapping
functions have been deprecated, as they would allow the
managed pointer to escape the scope. If you need those semantics,
use std::unique_ptr instead.
Change-Id: I2b0938b62f2ef5a3561f61f595a3fb4c505a8f08
Reviewed-by: Allan Sandfeld Jensen <allan.jensen@qt.io>
Reviewed-by: Volker Hilsheimer <volker.hilsheimer@qt.io>
QMetaType::construct will return nullptr if a copy is requested but
if there is no copy constructor. In that case, the memory allocated in
QMetaType::create will be leaked.
This case is covered in the unit tests in
template<>
void testCreateHelper<QMetaType::Void>()
so we can't assert in construct that a copy constructor is present if
a copy is requested.
Fixes static analyzer warning 618b1e6877f737df57c39fd35ad91cd4.
Pick-to: 6.1
Change-Id: Ic439b36b1f120b9d0f888efaeab2801d8acc13de
Reviewed-by: Fabian Kosmale <fabian.kosmale@qt.io>
QMetaType::convert does have code paths where nothing is written to the
'to' out-variable. Those code paths shouldn't run in this particular
case, but it's good practice to initialize out-variables anyway.
Fixes static analyzer warning 1453920f5abe1e30471eca6aa01d22d1.
Change-Id: I6f6d1044759866c405e61cb51147ca0490c4f3a4
Reviewed-by: Fabian Kosmale <fabian.kosmale@qt.io>
When a foreign event loop that does not enter an alertable wait state
is running (which is also the case when a native dialog window is
modal), pipe handlers would freeze temporarily due to their APC
callbacks not being invoked.
We address this problem by moving the I/O callbacks to the Windows
thread pool, and only posting completion events to the main loop
from there. That makes the actual I/O completely independent from
any main loop, while the signal delivery works also with foreign
loops (because Qt event delivery uses Windows messages, which foreign
loops typically handle correctly).
As a nice side effect, performance (and in particular scalability)
is improved.
Several other approaches have been tried:
1) Using QWinEventNotifier was about a quarter slower and scaled much
worse. Additionally, it also required a rather egregious hack to
handle the (pathological) case of a single thread talking to both
ends of a QLocalSocket synchronously.
2) Queuing APCs from the thread pool to the main thread and also
posting wake-up events to its event loop, and handling I/O on the
main thread; this performed roughly like this solution, but scaled
half as well, and the separate wake-up path was still deemed hacky.
3) Only posting wake-up events to the main thread from the thread pool,
and still handling I/O on the main thread; this still performed
comparably to 2), and the pathological case was not handled at all.
4) Using this approach for reads and that of 3) for writes was slightly
faster with big amounts of data, but scaled slightly worse, and the
diverging implementations were deemed not desirable.
Fixes: QTBUG-64443
Change-Id: I66443c3021d6ba98639a214c3e768be97d2cf14b
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@gmx.de>
... which makes me wonder, why isn't this stuff bumped automatically
when a minor version branch is created?
Change-Id: Ia43f898163a4baa0896a09bd13d65cf534fe1df5
Reviewed-by: Volker Hilsheimer <volker.hilsheimer@qt.io>
Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
Since the QT_CONFIGURE_HOSTBINDIR_TO_HOSTPREFIX_PATH and
QT_CONFIGURE_HOSTBINDIR_TO_EXTPREFIX_PATH definitions keep the same
value in modern CMake build, no need to have special handling in cases
where these values are used in qmake. Also it will be useful to
specify the relative path to the prefix directory from the directories
different of 'bin' when use QMakeLibraryInfo.
Task-number: QTBUG-75870
Change-Id: I5a777001eb334dcf05e22853a514d4257352d59b
Reviewed-by: Joerg Bornemann <joerg.bornemann@qt.io>
QT_CONFIGURE_CROSSBUILD is always '0', so it's not necessary to keep
conditional compilation in QMakeLibraryInfo.
Also the QT_CONFIGURE_SYSROOTIFY_PREFIX definition is never used
in the project.
Task-number: QTBUG-75870
Change-Id: I1fe42dce40fddc6a72254736883f93aa4727f6b3
Reviewed-by: Joerg Bornemann <joerg.bornemann@qt.io>
Previously, *datetime was only written to if the parse was a success.
When parsing a date-time that's invalid by virtue of falling in a
spring-forward gap, the parser returns a date-time that is invalid but
has a toMSecsSinceEpoch() suitable for use in creating a sensible
interpretation of the parsed string (in offset by the width of the gap
from the specified position in the gap). It is more useful to return
this value than a default-created QDateTime.
Change-Id: I89f39e729b1f9fede1532d8b82f6f676477ddadb
Reviewed-by: Mårten Nordheim <marten.nordheim@qt.io>
Reviewed-by: Andrei Golubev <andrei.golubev@qt.io>
When refreshing a QDateTime(,, Qt::LocalTime) we call mktime on data
obtained from it, passing in the DST status (when known; this keeps
two otherwise identical times in a fall-back distinct). One of the
tests relies on changing zone under the feet of such a date-time,
created in Hawaiian standard time; it serializes it, the reads it back
in Western Australian Daylight-saving time and expects the results to
be equal. However, the two differ in DST-ness, which leads to mktime()
failing for the Hawaiian original, with unwelcome results.
Notice this case, failure with DST-ness claimed known, and retry
without the claim, so as to correct the DST-ness.
Change-Id: Id0278df53130f76fc587769efe946ca9af1adc26
Reviewed-by: Mårten Nordheim <marten.nordheim@qt.io>
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
When computing a recent and imminent time, to bracket the time for
which we want data, take care not to cycle round to the other end of
the range of representable times.
Rephrased comments on this function, in the process, to more
accurately reflect what we're doing.
Change-Id: Iacd36186abc6b19d0ca03981aec80b2c5af077b3
Reviewed-by: Mårten Nordheim <marten.nordheim@qt.io>
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
The addition of a sanity-assertion revealed that, for an invalid
time-zone, refreshZonedDateTime() left epochMSecs unset but computed
offsetFromUtc from it none the less. Leave it as zero in that case, or
any other where the conversion to UTC didn't give valid date and time.
Change-Id: I0ebd955798532e91e7e211bf065667e313ee5c2d
Reviewed-by: Mårten Nordheim <marten.nordheim@qt.io>
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
This take more lines but makes the condition clearer and the lines
shorter, even after converting to use the names for constants in the
condition.
Change-Id: I9e5b7b79ff62095ed11b8723be238444fd32d9c1
Reviewed-by: Mårten Nordheim <marten.nordheim@qt.io>
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Various constants used in qdatetime.cpp were cast to qint64() where
used. There ware also some Q_INT64_C()s, two of which should have used
members of this enum; the third suggested a new addition to the enum.
Adding that and basing the enum on qint64 eliminates the need for
casting, although one asprintf() does now require a cast back to int.
There were also some redundant casts to qint64(), so I removed those.
Change-Id: Ia51ad8020f037badb1506ca379da19098a8655f8
Reviewed-by: Mårten Nordheim <marten.nordheim@qt.io>
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
QPointF operator== is fuzzy, hence it can't be hashed efficiently.
Prevent the erroneous addition of the overload by client code
by providing it as deleted.
Change-Id: I2cfaaf5c2c8896ec9a7929f7c7bf52a912d0450f
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Before this change, the QMetaType for T const* where T is derived from
QObject would not store the static QMetaObject of T. This commit changes
this. As a consequence, the metatype system can now convert between
const and non-const pointers to QObject. Note that this allows casting
const away; but so does C++ with const_cast.
In addition, a new flag, QMetaType::IsImmutable is introduced, and used
to tag the metatypes of pointer to const types. This allows code to
discern between pointers to mutable and const QObjects, which is
relevant for the QML engine.
Task-number: QTBUG-82354
Change-Id: I3e4e4f39f565bd99a65e161528ce5304df73d6d6
Reviewed-by: Shawn Rutledge <shawn.rutledge@qt.io>
Reviewed-by: Ulf Hermann <ulf.hermann@qt.io>
Getters and setters of bindable properties have to be carefully
written to avoid several problems. This patch adds documentation
for this.
Task-number: QTBUG-89505
Task-number: QTBUG-90511
Change-Id: Ib25590b3d8d95c490d9555c0f258f48cb6cfe4a9
Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
Reviewed-by: Karsten Heimrich <karsten.heimrich@qt.io>
Reviewed-by: Fabian Kosmale <fabian.kosmale@qt.io>
Instead of intercepting the WM_QT_SENDPOSTEDEVENTS message in the hook
procedure, we can handle it at receive points, making
qt_GetMessageHook() unnecessary.
Including general performance improvements, this patch fixes the issue
where some applications (e.g. Ableton Live) do not call a chain of
nested hooks for plugins.
Pick-to: 6.1 5.15
Fixes: QTBUG-90949
Change-Id: If8e96848392c6f10d45af2aac0567707d16af673
Reviewed-by: Oliver Wolff <oliver.wolff@qt.io>
Reviewed-by: Mårten Nordheim <marten.nordheim@qt.io>
Reviewed-by: Volker Hilsheimer <volker.hilsheimer@qt.io>
Reviewed-by: Nuno Santos <nunosantos@imaginando.net>
After QPoint(F), it's now the time of QSize(F) and QVectorND,
which can be unambiguously decomposed.
[ChangeLog][QtCore][QSize] QSize is now usable in a structured
binding declaration.
[ChangeLog][QtCore][QSizeF] QSizeF is now usable in a structured
binding declaration.
[ChangeLog][QtGui][QVector2D] QVector2D is now usable in a
structured binding declaration.
[ChangeLog][QtGui][QVector3D] QVector3D is now usable in a
structured binding declaration.
[ChangeLog][QtGui][QVector4D] QVector4D is now usable in a
structured binding declaration.
Change-Id: I67bb152f4210f2be27607179cd2ec522174cc483
Reviewed-by: Volker Hilsheimer <volker.hilsheimer@qt.io>
Move the qmake-specific logic of the QLibraryInfo class to
qmake internals. 'qconfig.cpp.in' now stores information about
the library info entries to keep them consistent between qmake
and the Core library. qmake requires specific features enabled
in the Core library, so building qmake will be skipped if the
features are not enabled.
All flags directly related to the qmake have been removed from
Core lib.
Remove all bootstrap related sections from qmake CMakeLists.txt
Task-number: QTBUG-89369
Change-Id: I26de157d3bfd4a5526699296e9d46e1c180b89ae
Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
Reviewed-by: Kai Koehne <kai.koehne@qt.io>
Be more precise about attaching a continuation with the default
(QtFuture::Launch::Sync) launch policy after a continuation with
context.
Pick-to: 6.1
Change-Id: I5b80063df2443e5742033864ba012bf34ed4cdf7
Reviewed-by: Ivan Solovev <ivan.solovev@qt.io>
Reviewed-by: Jarek Kobus <jaroslaw.kobus@qt.io>
Reviewed-by: Leena Miettinen <riitta-leena.miettinen@qt.io>
Reviewed-by: Mårten Nordheim <marten.nordheim@qt.io>
setCurrentAnimation(0) is expected to set the currentAnimation pointer -
which it does, as long as there are any animations.
Fixes static analyzer warning 106daf11ddfe9bb4fa7cf98c9097962c
Pick-to: 6.1
Change-Id: I3739bcfda103db391d21f135485f0b700c48ead1
Reviewed-by: Ulf Hermann <ulf.hermann@qt.io>
In QTimeZonePrivate::dataForLocalTime(), mistrust the Android
backend's hasDaylightTime(), as it has a comment saying it only knows
about future transitions, not past. This caller of it really needs to
query "has ever had a transition", which this doesn't answer. Many
zones that have no plans for future transitions have had transitions
in the past; these were failing the transitionEachZone() test.
In the process, refine the test itself, making sure we catch some
quirk cases that shouldn't arise and making the debug message on
failure more informative (while eliding the zone name, as this is part
of the test name anyway, so added to the output by qDebug() itself).
Fixes: QTBUG-69131
Change-Id: I88a0528182c247acb8b6327b40516178e455bcc0
Reviewed-by: Andrei Golubev <andrei.golubev@qt.io>
Reviewed-by: Mårten Nordheim <marten.nordheim@qt.io>
If called by QMetaType::canConvert with two nullptr values,
the QMETATYPE_CONVERTER_ASSIGN macro will expand to code
dereferencing both 'to' and 'from' pointers.
Assert that others callers provide two valid pointers.
Fixes static analyzer warning 02dc34cc2ad1d4c3c6e55b44e08983f2
Pick-to: 6.1
Change-Id: I24de914faa25dc7cb1da5eae09a125506caac389
Reviewed-by: Fabian Kosmale <fabian.kosmale@qt.io>
Some plugins need to append additional processing to targets that use
them. This is most commonly needed for static builds to identify
additional plugins that need to be linked, but which cannot be
determined automatically by CMake as part of its transitive
dependency handling. Check for finalizers on directly linked targets
as a way to help automate common scenarios.
Because we need to lookup dependency targets and these may have
visibility limited to the depender target's scope, we now formally
document that finalizers should be called from the same scope as
the one in which the target was created. This was loosely expected
before, but is now made explicit.
Task-number: QTBUG-90819
Task-number: QTBUG-86669
Pick-to: 6.1
Change-Id: I065a3f3a4ab6eeff6a98a655835427949e2fe0f1
Reviewed-by: Joerg Bornemann <joerg.bornemann@qt.io>
[ChangeLog][Build System] Tools that are called by the build system and
are unlikely to be called by the user are now installed to the libexec
directory.
This is a step towards easier co-installability of different Qt
versions.
Pick-to: 6.1
Task-number: QTBUG-88791
Change-Id: Id19575b5ba27795f7715e4ea6a09391b26dd4942
Reviewed-by: Kai Koehne <kai.koehne@qt.io>
We haven't used the spawn functionality on QNX since Qt 5.7 (commit
005a8bfbf0) because that's when we dropped
support for QNX 6.5.0.
Change-Id: Ic90d8429a0eb4837971dfffd1664f9712bdce2d8
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@gmx.de>
Use a structure that will automatically close them for us. This doesn't
apply to startProcess() because the pipes there are long-lived (though
each of them in QProcessPrivate could be an AutoPipe...).
The destructor only runs in the parent process, so the child processes
don't need to worry about setting file descriptors to -1.
Pick-to: 6.1
Change-Id: Ic90d8429a0eb4837971dfffd1664ed98f3d74d1c
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@gmx.de>
This is unnecessary because we can only get SIGPIPE if the reading end
of the pipe is closed. And that can only happen if the parent process
has exited, meaning there's no one to read our message anyway.
Change-Id: Ic90d8429a0eb4837971dfffd1664ec6821993ada
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@gmx.de>
This should have been SIG_DFL, as we're about to execute a child
process. It's the child's responsibility to ignore SIGPIPE if it wants
to, or get killed by it when it writes to an pipe with no readers.
Qt itself does this for its own purposes (see qt_ignore_sigpipe() [until
I can get some time to teach Linux about O_NOSIGPIPE]). Therefore, we
ought to reset what we've done.
Change-Id: Ic90d8429a0eb4837971dfffd166585a686790dde
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@gmx.de>
And set *pid to -1.
[ChangeLog][QtCore][QProcess] If a startDetached() fails to start the
target application, the QProcess object should now have a proper error
string in errorString().
Pick-to: 6.1
Change-Id: Ic90d8429a0eb4837971dfffd1664e825ffcb923e
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@gmx.de>
QAndroidApplication provides the Android specific app context() and
isActivityContext() to determine whether the context is an Activity or
otherwise a Service.
Task-number: QTBUG-90499
Change-Id: Iae2eef7ec44859a89825b09f52f09506b20b5420
Reviewed-by: Tor Arne Vestbø <tor.arne.vestbo@qt.io>
That qWarning cannot be in the child process (we don't know if a user
logger is fork-no-exec-safe) and the failure to chdir() should be
reported as a failure in QProcess::startDetached() instead.
Change-Id: Ic90d8429a0eb4837971dfffd1664e7577c81610b
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@gmx.de>
Reviewed-by: David Llewellyn-Jones <david.llewellyn-jones@jolla.com>
That only created an opportunity for qWarning(), which should never be
in the child process in the first place.
Change-Id: Ic90d8429a0eb4837971dfffd1664e57a2291ea78
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@gmx.de>
Reviewed-by: David Llewellyn-Jones <david.llewellyn-jones@jolla.com>
This will be used to access host tools that are installed in
${prefix}/libexec instead of ${prefix}/bin.
Pick-to: 6.1
Change-Id: I36c4b5736330f8229d267a117c65d55cd5e12758
Reviewed-by: Kai Koehne <kai.koehne@qt.io>
It was comparing time->second() to MSECS_PER_DAY - 1, but
time->second() is the second within its minute, so is at most 59.
It should be comparing seconds into the day to SECS_PER_DAY - 1.
Prompted by a PVS-studio article.
Pick-to: 6.1 6.0 5.15
Change-Id: I1802c49fa18818f4b4fe74f187da5f799df7d1de
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Registration by ID allows for detection of duplicate instantiation of
built-in back-ends, which can be detected and flagged by setting the
ID to ~size_t(0) instead of the enum value for which it sought to be
registered. A new method, calendarId(), is provided to access this;
while the old calendarSystem() becomes non-virtual, as it can be
inferred (when registration was successful) from the ID.
Make registration by name or alias conditional on successful
registration by ID. Previously, failed registration by name precluded
registration by ID, which now becomes the authoritative registration.
This incidentally makes it possible to add a QCalendar constructor
taking the unique ID of a backend, for use in conjunction with custom
calendar implementations.
Change-Id: Ib22925a8ac3ef9439a09ec3855f6231cf9b91c21
Reviewed-by: Mårten Nordheim <marten.nordheim@qt.io>
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Trivial details picked up during the course of investigating a
time-zone issue.
Change-Id: I4d6e7ab1787a2500bd950e7f12ed8618a31f1f8e
Reviewed-by: Mårten Nordheim <marten.nordheim@qt.io>
Reviewed-by: Andrei Golubev <andrei.golubev@qt.io>
QAndroidTimeZonePrivate overrode the transition-related methods of its
base, but there was no point in doing so, since our Android backend
has no access to transition data, just the same as the base
implementation.
Change-Id: Ie4ff375381b463078b412f50e8ddc925ab1587a3
Reviewed-by: Andrei Golubev <andrei.golubev@qt.io>
Reviewed-by: Mårten Nordheim <marten.nordheim@qt.io>
Fix:
qtbase/src/testlib/qtestelementattribute.cpp:89: (qdoc) warning: Undocumented enum item 'LET_Message' in QTest::LogElementType
qtbase/src/testlib/qtestelementattribute.cpp:89: (qdoc) warning: No such enum item 'LET_Error' in QTest::LogElementType
qtbase/src/testlib/qtestelementattribute.cpp:89: (qdoc) warning: Undocumented enum item 'LET_SystemOutput' in QTest::LogElementType
qtbase/src/network/ssl/qsslsocket.cpp:1666: (qdoc) warning: Unknown command '\cl'
qtbase/src/corelib/kernel/qproperty.cpp:883: (qdoc) warning: Unknown command '\T'
qtbase/src/corelib/kernel/qproperty.cpp:799: (qdoc) warning: Undocumented return value (hint: use 'return' or 'returns' in the text
qtbase/src/corelib/kernel/qjnienvironment.cpp:250: (qdoc) warning: Undocumented return value (hint: use 'return' or 'returns' in the text
Pick-to: 6.1
Change-Id: I116f5d8ace2c29ba7b6b93256d5761591e01296a
Reviewed-by: Assam Boudjelthia <assam.boudjelthia@qt.io>
Reviewed-by: Paul Wicking <paul.wicking@qt.io>
Reviewed-by: Nico Vertriest <nico.vertriest@qt.io>
This inlines the fastpath of QMetaType::id and splits the slowpath into
its own function. With that change, we can also use id in operator==,
simplifying the code there.
Change-Id: I286fe173b43a495dbda8faa151a93895b4fd22e4
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Fixes error: explicit specialization of class "QTypeInfo<QRingChunk>"
must precede its first use
Change-Id: Ib60f8fc1feb97953f5b0eeed63734e81f067d9b3
Reviewed-by: Janne Koskinen <janne.p.koskinen@qt.io>
Reviewed-by: Ville Voutilainen <ville.voutilainen@qt.io>
That class was only really meant for the QML compiler, and it is dubious
if event that one needs it. The current implementation is also broken.
Pick-to: 6.1 6.0
Change-Id: Ie40d282707f3fabc8079bee9e98f082aeb9d30b3
Reviewed-by: Ulf Hermann <ulf.hermann@qt.io>
QQmlProperyBinding needs the ability to suspend binding evaluation, and
needs access to the propertyDataPtr.
Change-Id: If82079ffdf28fb277c6e5083714c28478f6e1729
Reviewed-by: Ulf Hermann <ulf.hermann@qt.io>
Reviewed-by: Andrei Golubev <andrei.golubev@qt.io>
This adds a public function to check whether the QPropertyBindingPrivate
is a normal binding, or a QQmlPropertyBinding. In addition, this check
is used so that the source location function doesn't return garbage, but
instead indicates that the binding was set up from QML.
A function to retrieve the binding location from C++ can be added to
declarative at a later point.
Change-Id: Ica0f70780735fe9c60d01c2b21057d59714079e0
Reviewed-by: Ulf Hermann <ulf.hermann@qt.io>
Reviewed-by: Andrei Golubev <andrei.golubev@qt.io>
Looks like the corelib_add_mimetypes_resources call is redundant and
deprecated. It duplicates the mime-type database in the Core library.
This also fixes the static build of the qmimedatabase tests, since
avoids propagation of the resource symbols by the Core library.
Fixes: QTBUG-89952
Pick-to: 6.0
Pick-to: 6.1
Change-Id: I5c0dbd761b7726589fdf6970cd546af89d2ff837
Reviewed-by: Joerg Bornemann <joerg.bornemann@qt.io>
A recent change ( a7ca8b1a28 )
led to failure of binding removal in setInterval().
This was fixed by introducing setterScope.
This patch add unit tests for this regression.
Change-Id: Ic8da1f2d82ad6c8ccd81c9b1eff72d42cf75f28a
Reviewed-by: Fabian Kosmale <fabian.kosmale@qt.io>
The generated code contained many repeated blank lines.
Most main()s were preceded by one, but not all.
Only include blank lines for actual empty entries in lists, where
configure.json specified them, plus one before main if anything else
precedes it.
Change-Id: I6e6c34940e08712a1aa848a3c9ad1b0fb5806d77
Reviewed-by: Joerg Bornemann <joerg.bornemann@qt.io>
Apparently we used to have (back in 2007, only on Windows,
incompatibly with what we were then using on Unix) a TextDate format
(only for QDateTime, QDate used what it still uses) that put the
day-of-month number, with a dot after it, before the month's short
name. We have retained parsing of this format, on all platforms, ever
since.
It no longer matches the format we now use (since 5.2, in 2013, commit
61b56a89a1, which harmonised the format
with Unix and QDate); now seems like a good time to stop complicating
our parser for its sake.
[ChangeLog][QtCore][QDateTime] The parsing of Qt::TextDate in
QDateTime::fromString() no longer supports the old TextDate format
used (only) on Windows by Qt < 5.2 ("ddd d. MMM yyyy" with an
"HH:mm:ss" time either appended or inserted before "yyyy").
Change-Id: I73a798ab78f187543e415119cc4a11f1cfd73820
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Since fromString() can't parse the (ambiguous at the best of times;
also backend-dependent and thus potentially system-locale-dependent)
abbreviations currently produced (since 5.9) and can parse UTC-based
offsets, the OffsetName of the zone is a more robust format for the
zone-suffix. This also makes it possible to consistently use the C
locale, compatibly with everything else about post-6.0 date-time
serialization.
[ChangeLog][QtCore][QDateTime] When spec is Qt::TimeZone, the
offset-suffix now used for the toString(Qt::TextDate) format is now a
UTC-based offset string, compatible with the parsing (now) supported
by fromString(). The zone-abbreviation suffix in use since 5.9 was not
parseable.
Change-Id: I4024ae87980c6d3590c68a67b8d1c8f433e36855
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
There are GMT-offset zones whose convention for the sign of the offset
is the reverse of what we are (still) using, which is the usual
convention for UTC-offset zone: for example, the Olson Database's
Etc/GMT+3 has offset -3 hours in the UTC-based system we use, so we
give it suffix GMT-0300. The UTC-based suffix is also what we use as
the abbreviation for OffsetFromUTC() in toString().
For now this only adds support for parsing a planned future form: the
old form using GMT is retained, to give client code some chance to
prepare for a backwards-compatible transition. Although the GMT prefix
is matched case-insensitively, only match UTC if fully upper-case;
there is no meaningful precedent for case-insensitive usage here.
[ChangeLog][QtCore][QDateTime] The Qt::TextDate format now recognizes
UTC-based offset suffixes in addition to suffixes based on the
deprecated alias GMT. This prepares for toString() to use such
UTC-based suffixes for time-zones (fromString() cannot parse the
present abbreviation suffix). A future release of Qt shall use
UTC-based suffixes in place of the present GMT-based suffixes (which
conflict with GMT-based IANA zone names) for Qt::LocalTime and
Qt::OffsetFromUTC time-specs. Client code is encouraged to use and
recognize UTC-based zone suffixes in preparation for that transition,
unless compatibility with versions before 6.2 is required.
Change-Id: I5a42a488f1232a30f4b427b7954759283423b9b3
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
The main(void) signature obviates the need for unused parameters and
has existed since (at least) C89; so use that instead of the
old-fashioned argc/argv arguments we don't use in any of these tests.
Change-Id: Ibfe850a1fce378673c9781011475ea623fd75ad4
Reviewed-by: Joerg Bornemann <joerg.bornemann@qt.io>
Running the script updated them, without updating the configure.cmake
files they relate to.
Change-Id: I050d2131fa575811ae9ff4b51ef786a073c554c9
Reviewed-by: Joerg Bornemann <joerg.bornemann@qt.io>
Windows 8 added this pair of functions that can be used to implement the
same functionality as we have on Linux. For ease of understanding, I'm
calling them "futex" on Windows too.
From Qt 6 our minimum platform is Windows 10 so we can use this
unconditionally.
Change-Id: Ifea6e497f11a461db432ffff1448c6806ecfc36c
Reviewed-by: Qt CI Bot <qt_ci_bot@qt-project.org>
Reviewed-by: Fabian Kosmale <fabian.kosmale@qt.io>
Include Qt::TimeZone as a possible spec, use a \list for the spec
values. Incidentally use the public QTimeZone::abbreviation() instead
of digging around in its privates to achieve the same effect.
Change-Id: Ibabbeac9b085b4d09de46bda911356c20faadae8
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
The test verified that a LocalTime's time since Epoch changes when the
system time-zone changes. This works when the QDateTime object is in
short form and recomputes its offset from UTC every time it is needed,
but fails with a pimpled QDateTime, as this caches its offset from UTC
when it is created, saving the recomputation which - in the far more
usual case where the system time-zone does not change in the lifetime
of a QDateTime object - would normally produce the same result.
Changed the test to use a newly-created QDateTime constructed with the
same parameters, which doesn't have the cached out-of-date knowledge
of its zone offset. Removed the XFAIL. Made the test data-driven and
added test-cases: one so close to the Epoch that it should be short
even on 32-bit systems, one so far that it's pimpled even on 64-bit
systems (used in reproducing the issue in order to debug it).
This then revealed that Android 5 doesn't seem to support the POSIX
zone IDs used by this test, so it now verifies that LocalTime has the
expected offset from UTC after zone changes, QSKIP()ping if not.
Documented that the behavior of LocalTime is undefined after a change
to the system time-zone. Cleaned up the existing doc of Qt::TimeSpec
in the process.
Fixes: QTBUG-89889
Pick-to: 6.1 6.0 5.15
Change-Id: I1058f47a1ff3ee1c326f3579ac80bd8bab242e28
Reviewed-by: Mårten Nordheim <marten.nordheim@qt.io>
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
This reverts a limited part of commit ad2da2d27a
that deleted .pro files for snippets compilation.
Some .pro files which contain snippets used in the documentation
itself should be restored.
Task-number: QTBUG-90483
Change-Id: I1b03833c8ff17b5fca43a5b6c5417e8545b1711b
Reviewed-by: Topi Reiniö <topi.reinio@qt.io>
A recent patch made it so we don't need to keep patching the includes
of the sources. However it changed how we use double-conversion ourselves
which meant that system installs would no longer work. Keeping an extra
directory and adding another INCLUDE_DIRECTORY fixes both of our problems.
Amends dca74b45a3
Fixes: QTBUG-90961
Change-Id: If17aa1670535867343374eec846055441592f36b
Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
Reviewed-by: Joerg Bornemann <joerg.bornemann@qt.io>
Use {} instead of \, this fix a badly generated block of text.
Task-number: QTBUG-89632
Pick-to: 6.1
Change-Id: Ia4aad4c06285eb016a092f4340669fcbef1a6780
Reviewed-by: Paul Wicking <paul.wicking@qt.io>
In a small example program using HTTPS (on Windows with Schannel) >40%
of the time was spent initializing the backend, attempting to find the
time zone used in various certificates.
By adding an early check to see if the requested time zone is UTC (or an
alias ('Z')) we can skip most of the lookup that was required.
In the example program this cut away ~200ms of a total of ~550ms.
Change-Id: I4d29568653e02b64feebbf329469899eb7be1494
Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
Cuts off 1 second from the timezone test locally
Change-Id: I184728e97bcd65ca0362df4c26a3407576e12dfb
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
Add a private function to handle checking/clearing and deleting the
local reference to jobject before returning a QJniObject.
Task-number: QTBUG-89633
Pick-to: 6.1
Change-Id: I0ea28c8ba4da0bfc1e341c6b4c1f61fecfec87a6
Reviewed-by: Ville Voutilainen <ville.voutilainen@qt.io>
The initial implementation and the commit
c00ab6f8ea was wrong:
* env->findClass() in fact returns a global reference, and in any
case we shouldn't be calling that, instead QJniObject would be
enough.
* The size param provided to env->RegisterNatives was wrong.
* A test for registerNativeMethods() is added to ensure such break
is not repeated again.
Task-number: QTBUG-89633
Pick-to: 6.1
Change-Id: I4d3a6a9270755f465c40add25521fb750dd4de0a
Reviewed-by: Mårten Nordheim <marten.nordheim@qt.io>
Reviewed-by: Ville Voutilainen <ville.voutilainen@qt.io>
The system locale of a macOS application is not affected by environment
variables like LANG. Yet, we were reporting a name determined from
environment variables as the fallbackUiLocale(), rather than one based
on the language and country of the actual system locale.
This lead, via the usual CLDR likely-subtag fallback, to claiming the
system locale's name, language, script and country were those obtained
from these environment variables, even when they were at odds with the
actual locale being used by the system, which was being used for some
queries.
Worse yet, any data not supplied by these queries was being obtained
from the same CLDR locale as the name, making for an inconsistent mix
of locale data.
While we cannot avoid the likely-subtag fallback step for fallback
data, it is more consistent to use the actual system locale's name
as start-point for that fallback.
At the same time, add support for the language, script and country
queries, so that the QLocale::system() describes itself faithfully,
instead of claiming to be the locale that results from that fallback.
If we want to support LANG or other environment variable overrides,
they should be handled by the layer above the system locale, by
changing the default locale of the Qt application, as if the user
had called QLocale::setDefault().
[ChangeLog][QtCore][QLocale] QLocale::system() on macOS no longer
pretends to support LANG or other environment variables as overrides,
as this is not a feature that the system locale on macOS supports.
To override the locale of an application, use QLocale::setDefault(),
or pass -AppleLocale en_US.
Fixes: QTBUG-90971
Change-Id: Ibdaf5ff9a2050f61233a88eabf3c29094f7757f1
Reviewed-by: Volker Hilsheimer <volker.hilsheimer@qt.io>
Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
The standard user defaults are initialized from an ordered list of domains,
as documented by NSUserDefaults.standardUserDefaults. This includes e.g.
parsing command line arguments, such as -AppleLocale "fr_FR", as well as
global defaults. AppKit does this during [NSUserDefaults init], which in
turn initializes the locale returned by CFLocaleCopyCurrent() and
NSLocale.currentLocale.
If those functions are called before NSUserDefaults does its thing the
locale will already have been created, and the logic in NSUserDefaults
won't have any effect -- nor is there any way for us to set/override
the locale after this.
To ensure that the -AppleLocale command line override is available through
the lower level Core Foundation preferences APIs, we need to initialize
the user defaults as early as possible via the Foundation-API.
Change-Id: I906a5a8b05a7216e60020ec45f8da725b801d2c5
Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
Reviewed-by: Volker Hilsheimer <volker.hilsheimer@qt.io>
Qt CMake uses a fake_prefix as install dir when running tests instead of
the main qt install path, this will throw androiddeployqt off since it
expects the real qt install path which has gradle and java sources.
Fixes: QTBUG-88579
Pick-to: 6.0 6.1
Change-Id: I6580470840ae14d4a4a68a95f217b30408d7ab44
Reviewed-by: Craig Scott <craig.scott@qt.io>
Reviewed-by: Joerg Bornemann <joerg.bornemann@qt.io>
Somehow QList::fill(t, newSize) introduced a regression in Qt6:
when newSize < QList::size() we should resize to the newSize.
This is aligned with QVector::fill() in 5.15 and std::vector::assign()
While 6.0 is already out, picking it to 6.0.x could save someone who
haven't migrated yet as well as fix some accidental bugs in Qt's code
[ChangeLog][QtCore][QList] Fixed QList::fill() regression introduced in
6.0: calling fill() with size < current list size wouldn't truncate the
list
Fixes: QTBUG-91042
Pick-to: 6.0 6.1
Change-Id: Ic166e2c5e42390b61df1030f7c705e344433f7f2
Reviewed-by: Lars Knoll <lars.knoll@qt.io>
Other affected rows have also been fixed．
Change-Id: Ie0a32f724bd2e40e7bfacfaa43a78190b58e4a21
Reviewed-by: Volker Hilsheimer <volker.hilsheimer@qt.io>
Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
The "2" is meaningless. We retain the old name as alias for source
compatibility, but we don't document it anymore.
Task-number: QTBUG-85064
Change-Id: Id10d58f6d39a0faaec042304b374ae6b23469362
Reviewed-by: Joerg Bornemann <joerg.bornemann@qt.io>
Reviewed-by: Mitch Curtis <mitch.curtis@qt.io>
This reverts commit 74f549b402.
Reason for revert: This breaks transitive dependencies.
The object file we're adding with target_sources will not be
propagated transitively.
For example, qmltestrunner depends on QuickTest, and
QuickTest depends on Quick.
The 'qrc_scenegraph.cpp.o' object file will not be propagated to
qmltestrunner if added with target_source(Quick INTERFACE ...).
Fixes: QTBUG-90933
Pick-to: 6.0
Change-Id: I11269537ae75de14189f8b5f59094b467e2eee24
Reviewed-by: Jani Heikkinen <jani.heikkinen@qt.io>
After a "QtCore: WARNING:" intro (separated to avoid being mistaken
for a footer) syncqt.pl said:
.../qtbase/src/corelib/kernel/qjnienvironment.h includes QScopedPointer when it should include QtCore/QScopedPointer
Change-Id: Iaa1f025367b7321af9c5de27196ebf271f9179df
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
We can avoid the overhead of the out-of-line function calls if the id is
already set.
As a drive-by, use relaxed loads in QMetaType::id (was: acquire).
Change-Id: I1d53dbf7e58eae80776ff36334ebf0642ba7842f
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
The various spatial-vector, line, point, region and margin types have
all long taken for granted that their co-ordinates are finite and, in
particular, not NaN. Make this expectation explicit in the
documentation. Added assertions where (chiefly) noexcept and (where
the assertion would be a simple qIsFinite() call) constexpr didn't
preclude doing so. Also assert against zero divisors that would lead
to non-finite results.
Make minor clean-ups to docs in the process. QMarginsF had several
methods whose declaration, definition and docs weren't consistent in
their parameter names.
Task-number: QTBUG-89010
Change-Id: I601a830959d13a73dcb17ce31a1202a1486c8f7f
Reviewed-by: Laszlo Agocs <laszlo.agocs@qt.io>
Reviewed-by: Qt CI Bot <qt_ci_bot@qt-project.org>
Being able to iterate h = h.add(next) over a sequence of numbers makes
it possible to implement arbitrarily-many parameter functions that
need hypot, e.g. the implementation of V4's Math.hypot().
This amends commit b08368d99f.
Change-Id: I76bb65bd5f318e915d42fa4c935129850e258132
Reviewed-by: Andrei Golubev <andrei.golubev@qt.io>
Reviewed-by: Ulf Hermann <ulf.hermann@qt.io>
This simplifies the include path addition we need when compiling and
eliminates the need to rewrite all the #include lines on update.
Change-Id: I14e301d277f1eb49286721b012559f0ec84ea520
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Reviewed-by: Joerg Bornemann <joerg.bornemann@qt.io>
Setters of QObjectCompatProperty need to remove the binding, but
only if it is not in the wrapper. This patch adds this method to
the public interface.
Change-Id: Ibc96418fe15b0aeeb2059411ea910ba1f733c255
Reviewed-by: Fabian Kosmale <fabian.kosmale@qt.io>
A static build of Qt with QT_BUILD_TESTS=ON would fail with errors like
this:
FAILED: tests/auto/corelib/io/qprocess/testProcessEcho2/testProcessEcho2
undefined reference to `qRegisterResourceData(...)'
We create object libraries for resources. In this case, for the
mimetypes of QtCore.
For static builds we added the object files to INTERFACE_LINK_LIBRARIES
of the Qt module the resource belongs to:
target_link_libraries(Core INTERFACE
"$<TARGET_OBJECTS:$<TARGET_NAME:${resource_target}>>")
That is, however, a problem for projects that link only against QtCore,
like testProcessEcho2 above. On the linker command line Qt6Core.a is
before qrc_mimetypes.cpp.o and that doesn't work, because the latter
needs symbols of the former.
To have the resource object files before any other libraries on the
linker's command line, we now treat them as source files.
Pick-to: 6.0
Fixes: QTBUG-90582
Change-Id: Ia2786beb5f7c2eba194e8f00ae8785572df1cf6e
Reviewed-by: Tor Arne Vestbø <tor.arne.vestbo@qt.io>
registerNativeMethods() should be using the local jclass from
findClass() instead of the global reference.
Task-number: QTBUG-89633
Change-Id: I469a9a1ecff95ab9948421baa5c88735ba7ad776
Reviewed-by: Alex Blasche <alexander.blasche@qt.io>
Dating from the origins of our support for the zoneinfo file format,
the mapping of POSIX's day-numbering (0 = Sunday through 6 = Saturday,
see [*]) to Qt's (1 = Monday through 7 = Sunday) was done by mapping 0
to 1, when it should have been 7.
[*] http://www.gnu.org/software/libc/manual/html_node/TZ-Variable.html
Corrected a QTimeZone test that trusted the results it got without
checking which day of the week those were: they were all Mondays.
Verified that the corrected dates are in fact Sundays.
Checked the zone abbreviations, too.
Fixes: QTBUG-90553
Pick-to: 6.0 5.15
Change-Id: I84b4b14f9892ff687918cd3c42c7c9807e45313c
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Reviewed-by: Qt CI Bot <qt_ci_bot@qt-project.org>
Reviewed-by: Andrei Golubev <andrei.golubev@qt.io>
As we call operator new on create(), we also need to delete on
destroy(). Otherwise we leak memory.
Change-Id: Ib80fe96c4173cba6fa474d3c81d88fe603d1ded2
Reviewed-by: Fabian Kosmale <fabian.kosmale@qt.io>
Reviewed-by: Mårten Nordheim <marten.nordheim@qt.io>
In order for other Qt modules to be able to reliably use this property,
it needs to be defined under all supported scenarios. Previously, it
would only have been defined for a per-repo build and not for a top
level superbuild. The only place it was being set was in the
Qt6CoreConfigExtras.cmake file and only when QT_NO_CREATE_TARGETS was
false (so it wouldn't be defined for examples even in a per-repo build).
It was also noticed that INTERFACE_QT_MAJOR_VERSION was also being
handled in Qt6CoreConfigExtras.cmake, but this was not needed. This was
already being done as part of a call to qt_autogen_tools_initial_setup()
which occurs inside qt_internal_add_module().
Fixes: QTBUG-90617
Change-Id: I589316ffb8138bd2df8a1d49fdd6335d2e2836f1
Reviewed-by: Joerg Bornemann <joerg.bornemann@qt.io>
The QT_BUILD_QMAKE_BOOTSTRAP and QT_BUILD_QMAKE_LIBRARY definitions
are never defined in the project since migrating to CMake build.
Change-Id: I7108c92387005a2fde6ebdb2d74843e3efc6413e
Reviewed-by: Joerg Bornemann <joerg.bornemann@qt.io>
Turned our local diffs into a side-branch in the upstream repo, got
material changes upstreamed (leaving only our #include changes),
rebased it to master, copied back to our version. Upstream has made no
new release since 3.1.5 in 2019 May, but there is substantial change
since then, including acceptance of our upstreaming. Our patches have
been refined in the process of upstreaming.
The list of source files is slightly changed. The double-conversion.*
files are split in two, string-to-double.* and double-to-string.*, but
the old double-conversion.h header is retained, simply pulling in its
two halves. These thus need to be present in the include directory.
Changed the patches/ directory to be in git format-patch form, for
sending or applying to the upstream repo, before copying files into
the qtbase source tree. This should make future upstreaming easier.
Task-number: QTBUG-90214
Change-Id: Id94198f58ef7bdf02af117c35cb9678b5c34ac0e
Reviewed-by: Lars Knoll <lars.knoll@qt.io>
Introduction of QObjectCompatProperty requires every write to
the property to be examined whether it is OK or should be replaced
by a setValueBypassingBindings/markDirty combination. The existence
of operator= make this difficult as it is easy to miss places where
it is written. By not having operator=, we can help developers
make sure they had a conscious decision about each write to the
property.
Change-Id: Ia61ea4722eb0bab26ce7684b85dd03d710cd1751
Reviewed-by: Fabian Kosmale <fabian.kosmale@qt.io>
It is quite common for a socket notifier to be disabled in a slot
connected to its activated() signal. But the event dispatcher did not
update the position in the list of registered notifiers, which caused
the next notifier to be omitted. Of course, on the next iteration of
the event loop, the omitted notifier would fire again, but this reduced
the scalability of applications that use a large number of sockets.
To solve the problem, we update the current position in the list when
a notifier becomes unregistered.
Change-Id: I6565bf23500d9e38ce84b34784d89d227fa166e1
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@gmx.de>
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
This reverts commit 8f8405e046.
Reason for revert: Appears not entirely thread-safe and caused QTBUG-90705
Change-Id: I390c0b1a555a18e6a095b52010371d017071e26b
Fixes: QTBUG-90705
Reviewed-by: Sona Kurazyan <sona.kurazyan@qt.io>
There is no reason for QVariant to go through QMetaType when it can use
the QMetaTypeInterface directly. Without LTO, the QMetaType method calls
are opaque, and we therefore risk to lose optimizations.
Additionally, avoid constructing a QMetaType from a type id if we
already have the QMetaType.
Fixes: QTBUG-90673
Change-Id: I7069ff6aff70d5baecdf5cf5760014c3dda81784
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
This is in line with QML where
import QtQuick 2.15
Rectangle {
width: 100
height: 100
color: "red"
Rectangle {
id: inner
x: 10
y: x
width: 50
height: 50
onYChanged: { console.log("hey"); inner.x = 10}
TapHandler {
onTapped: inner.x = 20
}
}
}
results in a binding loop warning when the tap handler triggers. While
the change handler would only run once, we cannot statically determine
if we need to loop once, twice, or if there actually is a diverging
loop. Thus we unconditionally warn about the binding loop and stop
executing the binding.
As a drive-by, verify in the related test that a change handler which
overwrites its properties binding itself removes the binding.
Change-Id: I5372019c2389ab724c49cd7489ecbd3ebced1c69
Reviewed-by: Ulf Hermann <ulf.hermann@qt.io>
So far, all documentation about bindable properties was in the
corresponding classes QProperty and QObjectBindableProperty.
This patch creates one documentation page for a general
introduction to bindable properties and information
concerning all of those classes.
Change-Id: Ib718dbeb385c3899fb34cc2ce03bec7a8d43a034
Reviewed-by: Fabian Kosmale <fabian.kosmale@qt.io>
There are two methods which set the interval and do things
afterwards. Setting the interval might trigger other code. This
patch uses setValueBypassingBindings and marks the interval dirty
later, just to be safe.
Change-Id: I2cae3e4a9f040007dfd246c424250034e8df10a0
Reviewed-by: Fabian Kosmale <fabian.kosmale@qt.io>
Reviewed-by: Ivan Solovev <ivan.solovev@qt.io>
Same reason as for the version without args. Amends
1461df2ce4.
Change-Id: Ib9f0d8f27d858ff35134001c3cd7945b17668ae4
Reviewed-by: Sona Kurazyan <sona.kurazyan@qt.io>
To minimize code duplication, move the socket notifier deletion to the
closeChannel() function, where the pipe descriptor will be closed.
Change-Id: If75ba1c955c706ae6e2b3d9f53f7a25e4aa32fa7
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@gmx.de>
As part of Qt 6 restructring for the extras modules, this change exposes
the Jni APIs which are very important for Android platform. This patch
adds the APIs QJniObject, QJniEnvironment, QJniExceptionCleaner based
from private QtCore and QtAndroidExtras.
The Jni interface is cross-platform which justifies the name, but
currently, this API is used mainly for Android, and the naming comes
generic without Android keyword to avoid any future limitation on
supporting other platforms.
[ChangeLog][QtCore] Add new QJniObject, QJniEnvironment and
QJniExceptionCleaner APIs.
Task-number: QTBUG-89482
Fixes: QTBUG-89633
Change-Id: I4382dd53a225375759b9d042f6035a4a9810572b
Reviewed-by: Ville Voutilainen <ville.voutilainen@qt.io>
We missed takeBinding as a supported operation on Q(Untyped)Bindable.
To avoid adding version checks to code dealing with QBindableInterface,
we simply synthesize takeBinding as a combination of binding to retrieve
the binding and setBinding with a default-constructed
QUntypedPropertyBinding.
Change-Id: I43803a0dfe210353d0235f0373d2257f75ffe534
Reviewed-by: Lars Knoll <lars.knoll@qt.io>
We were using the first abbreviation in the list, where the current
time's one is probably more apt.
We look up displayName() using ICU, when in use, but that abbreviatio
may be unknown to it. So ensure, when using abbreviation in place of
empty id, that we get the system zone for ICU, for use if we're asked
for the display name.
This is a follow up to b12d6c6a8a.
Pick-to: 6.0 5.15
Change-Id: I177db55de1ffbc763def8a0423642f2b3da74fa6
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
We have plenty of places where we add some squares and take a square
root; this may be done more accurately and faster by hypot().
Introduce QHypotHelper to handle hypot with more than 3 parameters,
and with 3 when the C++17 version is missing (which it never should
be). Include an overload taking arbitrarily many valus and ensure that
we can use qHypot() with qfloat16. Illustrate with some example uses,
add some tests.
[ChangeLog][QtCore][QMath] Header <QMath> now provides qHypot(), an
implementation of std::hypot() taking arbitrarily many numeric values,
including support for qfloat16, while avoiding the overflow and
underflow problems that arise when naively taking the square root of a
sum of squares.
Change-Id: Ia4e3913fe83fc27d17d8e7f1a52f03ad445c1fed
Reviewed-by: Andrei Golubev <andrei.golubev@qt.io>
Reviewed-by: Mårten Nordheim <marten.nordheim@qt.io>
No change to QLocale's data, one addition to the Windows time-zone
data. What was formerly "Us Mountain Standard time / Canada" is now
Yukon Standard Time.
Fixes: QTBUG-89784
Pick-to: 6.0 5.15
Change-Id: I4c9a23620e74ea379be8a4c5ba0896d35fe9b594
Reviewed-by: Mårten Nordheim <marten.nordheim@qt.io>
Ported all properties, except activeThreadCount. Marking it dirty may
cause a re-evaluation of properties depending on it, which may reault in
a deadlock in case of trying to read activeThreadCount property which is
being marked as dirty.
Task-number: QTBUG-85520
Change-Id: Id073b0895c89a9e6b05b57ad520db994e550a1c9
Reviewed-by: Fabian Kosmale <fabian.kosmale@qt.io>
At least some of them don't like the fact we have classes with
non-trivial destructors but trivial move/copy constructors -
disable them explicitly, hopefully those analysers are smart
enough to stop complaining.
Change-Id: I285711a2d21bff89661af928ba326ecd5b69823f
Reviewed-by: Volker Hilsheimer <volker.hilsheimer@qt.io>
Reviewed-by: Mårten Nordheim <marten.nordheim@qt.io>
Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
perf indicates that we spend quite some time in ~QVariant in the
referenced task's benchmark. With this patch, we significantly reduce
the amount of work we do in the destructor for trivial types: Instead of
calling an out-of-line function, which does a few checks and then does
an indirect call through a function pointer to a do-nothing dtor before
freeing the memory, we now simply free the memory.
We achieve this by changing QMetaTypeInterface to leave the dtor
function pointer null if the type is trivially destructible. Then, in
QVariant we use the QMetaTypeInterface directly instead of going through
QMetaType.
Task-number: QTBUG-90560
Change-Id: Iaf87036710e1680323842e1ba703a5d3d0e5027a
Reviewed-by: Lars Knoll <lars.knoll@qt.io>
- Add move constructors to QRegularExpression, QRegularExpressionMatch
and QRegularExpressionMatchIterator.
- Update the documentation to explicitly state that only destructor
and assignment operators can be called for a moved-from object
Task-number: QTBUG-86634
Change-Id: I06b4f54e300541033a9a18339c97338717a06da0
Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
Reviewed-by: Alex Blasche <alexander.blasche@qt.io>
Constructing a QMetaType from an id requires a lookup which isn't free,
even for builtin types. When constructing a QVariant from an integer,
this leads to a noticeable performance loss.
In the mirco-benchmark of the linked bug, we improve from
100000000 965
to
100000000 789
on Linux, Intel i5-4690, clang 11.01 with a CMake release build.
This however still does not recover the performance we had in Qt 5.
Task-number: QTBUG-90560
Change-Id: I26440bdc04ef131ce91d3698957f0df43a4c4017
Reviewed-by: Lars Knoll <lars.knoll@qt.io>
Reviewed-by: Andrei Golubev <andrei.golubev@qt.io>
Otherwise ASAN complains about mismatched new/delete.
Change-Id: I76ddcc388309e20055c93c68b9b6fa8a11b884e1
Reviewed-by: Andrei Golubev <andrei.golubev@qt.io>
Reviewed-by: Ville Voutilainen <ville.voutilainen@qt.io>
And add an entry about the API changes to qt6-changed.qdoc.
Pick-to: 6.0
Task-number: QTBUG-89429
Change-Id: I146574acd3e4aa5ed6f1629fc68888f009472f19
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Make stricter alignment requirements for the allocated header
This strict alignment allows reallocateUnaligned() to property account
for the padding occurring in cases when
alignof(QArrayData) < alignof(T) <= alignof(std::max_align_t), which
happens to be the case on e.g. 32-bit platforms with specific alignment
requirements.
This adds 4 bytes (the difference between alignof(std::max_align_t) and
sizeof(QArrayData)) of overhead for QString, QByteArray and certain QLists
on 32-bit systems.
Task-number: QTBUG-90359
Pick-to: 6.0
Change-Id: I8176a4cc79f100ee772b09425e88fe8ff3ae226a
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Reviewed-by: Qt CI Bot <qt_ci_bot@qt-project.org>
Commit 370324f6e2 changed the call to 'xml' to 'xmlstarlet' because
some Linux distributions do not provide a binary called xml. Anyhow,
the official Windows packages of xmlstarlet only provide 'xml.exe'.
Therefore just check for both.
Pick-to: 5.15 6.0
Change-Id: I4381b256850e4101b760df21f2b0baffb4414620
Reviewed-by: Eric Lemanissier <eric.lemanissier@gmail.com>
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
This complements patch a148c7b5d71d244, where languageToCode(),
countryToCode() scriptToCode() methods were introduced, with matching
codeToLanguage(), codeToCountry(), and codeToScript() methods.
This allows us to remove the use of private Qt Core API in Qt Linguist.
[ChangeLog][QtCore][QLocale] Added static codeToLanguage(),
codeToCountry(), codeToScript() methods that convert ISO code strings
to the respective enum values.
Change-Id: If5c0843a718c006ade086a6f74ceb86ac6e0fce4
Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
Previously, different threads instantiating the same back-end could
collide in the register, with various unwelcome results.
Give Registry an atomic status flag to track whether it has been
populated or is being destroyed; and protect it with a mutex to ensure
distinct threads do not collide during registration or attempt to
register while the registry is being destroyed.
Document the correct way to instantiate custom backends, and that no
code other than the QCalendar implementation should instantiate the
built-in ones. Instantiators that follow these rules should be safe
from failed registrations, provided they don't pick a name that
conflicts with some other backend. They can also use the recent change
to semantics of registerAlias() to verify that registration *has*
succeeded.
Done-with: Giuseppe D'Angelo <giuseppe.dangelo@kdab.com>
Pick-to: 6.0 5.15
Task-number: QTBUG-88815
Task-number: QTBUG-85692
Fixes: QTBUG-84575
Change-Id: Ie78e700e71d610454152c05cafb38f6f713649ad
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
This fixes INTEGRITY build error:
cannot access position 53 in array of 49 elements
constexpr const char *begin = func + prefix;
and many "expression must have a constant value" errors.
Change-Id: I0e043028fbcb479c6657b59ea144b4536cc65651
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
... where it belongs.
To avoid the loop, introduce the drainAndStop() function, which allows
QWindowsPipeReader to flush the pipe itself. It determines the number of
bytes pending and blocks until the remainder of the process output is
received.
Note that the loop in drainOutputPipes() didn't actually have to
interleave the two pipes (because we're presuming that the operations
will finish instantly), so we don't do it now. Also, the code violated
the API contract: 'true' was returned when the 'wrong' channel received
data; this is now fixed as a side effect.
Change-Id: I38ed4861a238e39e793c3716e856e5bfdeed3d74
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@gmx.de>
Restore holding of a getID() call's return in a QJNIObjectPrivate
eliminated in commit 78cde1bfd94521bbe4972f31a79c959d0990ea77; this
keeps the Java internal object alive for its lifetime, where it might
otherwise be garbage-collected before we're done with it.
Fixes: QTBUG-88610
Pick-to: 5.15 6.0
Change-Id: Id65b059012f7bd3377a6562c1b647feb75a13679
Reviewed-by: BogDan Vatra <bogdan@kdab.com>
To optimize certain operations, it can be useful to know whether we are
currently evaluating a binding. For instance, we have properties whose
storage is only alloctaed on-demand when they are set. However, we would
also allocate them if they are used in a binding context, as we would
otherwise not properly track the dependency. Using
isAnyBindingEvaluating in the getter, we can detect this
situation, and avoid the allocation if it returns false.
This API is private for now, as it exposes some internals of the
property system and should be used with care. As it needs to access the
TLS variable, it also has a non-negligible cost.
Change-Id: I373aabee644fe7020b2ffba7d6a0ad9a1e1b4ec0
Reviewed-by: Ulf Hermann <ulf.hermann@qt.io>
It previously fell back on abbreviation, a very poor choice.
The base class does better than that, so use its implementation instead.
Pick-to: 6.0
Change-Id: I47cbfce815cd8b2b533d9c6aeebcf398e5852d02
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Avoid complaints from static analyzers that the pointer returned
by malloc might be null.
Change-Id: I3ec3ba03d0b5283dd569200a3040a5fe5990f763
Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
So that static analyzers don't get confused by its semantics.
In builds with exceptions disabled, it's not actually called
by client code (e.g. Q_CHECK_PTR will just terminate in that case),
but we still need to make it not return -- add another path that
callss std::terminate(), otherwise we'd have a noreturn function
returning.
Change-Id: Ia8c4ce3e9d971f1757e9c273051cb3dedf23c61f
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Add INTERFACE_COMPILE_DEFINITIONS of Qt6::Core to generated rcc object
libraries. This propagates QT_NAMESPACE definition to the object
library.
Fixes: QTBUG-85620
Change-Id: I252d1aaee7b19a49bc321fdd271a5d85a34bf67f
Reviewed-by: Tor Arne Vestbø <tor.arne.vestbo@qt.io>
Reviewed-by: Joerg Bornemann <joerg.bornemann@qt.io>
Reviewed-by: Qt CI Bot <qt_ci_bot@qt-project.org>
It used QString.compare() and assumed it was returning a bool true on
equality, when it actually returns an int that compares to 0 as the
given strings compare. So it should use compare() == 0.
This fixes several of QTimeZone's blacklisted tests on Android and a
crasher, which we dodged with a QSKIP. Added an id-comparison to a
test. Gave two local variables more informative names, made an early
return into a QSKIP so it explains itself.
Fixes: QTBUG-89905
Fixes: QTBUG-69122
Fixes: QTBUG-69132
Fixes: QTBUG-87435
Pick-to: 6.0 5.15
Change-Id: Icf18ed5a810143d6e65d36e34a70e82faac10b8e
Reviewed-by: Qt CI Bot <qt_ci_bot@qt-project.org>
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Reviewed-by: Andrei Golubev <andrei.golubev@qt.io>
When creating a time-zone from a UTC+offset name that isn't known to
the system, QTimeZone (since the fix to QTBUG-77738 in 5.15.0) falls
back to constructing a suitable UTC-offset backend; however, the id of
this is not guaranteed to match the id passed in to the constructor.
In all other cases, the id of a QTimeZone does match the id passed to
its constructor.
Some utcOffsetId testcases had different id() than the id passed to
the constructor, due to mismatches where a zone was constructed using
the fall-back but the generated id included its minutes (as :00) or
omitted its seconds. The omission of seconds is clearly a bug, but we
also don't want to include :00 for seconds when it's not needed. So
change QTimeZonePrivate::isoOffsetFormat() to accept a
QTimeZone::NameType to configure how much we include in an id. Its
callers other than the relevant constructor (from offset) still get
minutes, even when :00, but will also get seconds added if that isn't
zero; and the constructor from offset now gets the short form obtained
by omitting all trailing zeros.
Since all valid whole-hour offset names that do include :00 for the
minutes field are in fact known standard offset names, the elision of
minutes will only affect zones created by ID in the case of a
whole-hour offset given without :00 minutes specifier, so these shall
necessarily in fact get the ID passed to the constructor. Creating by
UTC-offset with a name that specifies zero seconds will result in a
QTimeZone instance whose id() differs from what was passed to its
constructor (eliding the :00 seconds and potentially also minutes, if
also zero) but this should be the only case where a QTimeZone's id
doesn't match the one passed to the constructor, when constructed by
id.
Fixed inconsistency between the offset-constructor's declaration
(taking offset as int) and definition (taking qint32) in the process.
Added an id check to the utcOffsetId() testcase. Amended two tests of
offset-derived time-zones' IDs, added comments to make clear how one
of those differs from a matching standard name test and converted two
uses of QCOMPARE(, true) to QVERIFY().
[ChangeLog][QtCore][QTimeZone] QTimeZone instances created by offset
from UTC (in seconds) shall now only include minutes in their ID when
the offset is not a whole number of hours. They shall also include the
seconds in their ID when the offset is not a whole number of minutes.
Pick-to: 6.0 5.15
Task-number: QTBUG-87435
Change-Id: I610e0a78e2aca51e12bfe003497434a998e93dc7
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Original QML-specific connection mechanism ignores the receiver argument
and uses sender as receiver. This causes uncontrollable memory growth
in certain cases as connections on receiver persist even after receiver
is destroyed
New connect() with receiver parameter uses underlying API correctly,
disconnect is provided for the symmetry (not sure it's really needed)
Task-number: QTBUG-86368
Pick-to: 5.15 6.0
Change-Id: I4580d75b617cb2c4dfb971a4dfb8e943e325572b
Reviewed-by: Fabian Kosmale <fabian.kosmale@qt.io>
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Currently the codes are only exposed in aggregated form, i.e. through
name(), bcp47Name(). There are use cases though where you are only
interested in either language, country, or script codes. One example
is in Qt Linguist.
This patch therefore exposes the static languageToCode(),
countryToCode(), scriptToCode() methods that were so far only available
in the private API also in the public API.
[ChangeLog][QtCore][QLocale] Added static languageToCode(),
countryToCode() scriptToCode() methods that convert enum values
to the respective ISO code strings.
Fixes: QTBUG-39542
Fixes: QTBUG-64942
Change-Id: Ib1d5c3293e2f53245ba4c1fc8159275bcb290080
Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
In Qt 5, QVariant::fromValue<T> would not compile unless
Q_DECLARE_METATYPE(T) was used, and Q_DECLARE_METATYPE(T) would lead to
a compile error if T were not copy constructible.
In Qt 6, we do not require Q_DECLARE_METATYPE before using fromValue,
and QMetaType itself works with non-copy constructible types just fine.
However, QVariant still requires it, thus we need to now enforce this in
fromValue itself.
Change-Id: Ib6964a438d8c46033dd3a037b9d871de2b42e175
Reviewed-by: Lars Knoll <lars.knoll@qt.io>
They have been reported wrong since Qt 4, and not fixed for behavior
compatibility, fixing it for Qt 6.
[ChangeLog][X11] XF86LaunchXX keys have been remapped, so the Qt names
and X11 names match, instead of being 2 off.
Pick-to: 6.0
Fixes: QTBUG-25261
Change-Id: Ie3a8676439ae3e93a78218c9e7f4443565e84356
Reviewed-by: Gatis Paeglis <gatis.paeglis@qt.io>
When writing out a float value, the output string is encoded as QVariant
for no reason. Looks like an oversight when QMetaType::Float was added a
long time ago.
Fixes: QTBUG-21156
Change-Id: I7f5d31e15892d700c1b1e5e731b7733ce3a15730
Reviewed-by: Lars Knoll <lars.knoll@qt.io>
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
[ChangeLog][QtCore] qIsInf(), qIsNaN() and qIsFinite() now, like
std::isinf() and friends, accept integral types (returning false,
false and true, respectively) as well as floating-point ones.
Change-Id: I1e2b7f033f0e8166c0b21e31a62b3e6d37b9344a
Reviewed-by: Volker Hilsheimer <volker.hilsheimer@qt.io>
Functions in QtFlagHandlingHelpers.cmake try to update the CMake cache
variables for compiler and linker flags. These were using the current
value of those variables and writing the modified ones back to the
cache every time CMake ran. If a toolchain file sets or modifies any of
these variables, that updated value is used and written back into the
cache instead of the original cache variable's value. The next time
CMake executes, the toolchain file re-applies the same change and the
variable grows longer each time with flags repeated. With Ninja, this
causes a complete rebuild every time CMake is re-run. The Android NDK
toolchain file is one example where this behavior is triggered (the
fault is shared, one could argue that the NDK should only be setting
..._INIT variables, but that's out of our control).
Another related bug in the previous implementation was that the flags
used to build after the first CMake execution could be different to
those used for all builds after the second and later CMake runs. This
is because the CMake cache was being updated, but not always the
calling scope of the functions that modified them. If a toolchain file
set any of the compiler or linker flag variables as non-cache
variables, then updating the cache variable would have no effect on
the calling scope. The non-cache variable would continue to take
precedence for that scope for that run. The next time CMake executes
though, the updated cache variable would now have been used by the
toolchain file and the change *will* be part of the non-cache
variable's value.
The above are examples of why you should try to avoid updating these
cache variables from project code. We could leave the cache alone and
always update only non-cache variables, but then a developer looking
at the cache may wonder why the values they see there don't match the
values being used in builds. Or worse, they think the cache values
are being used and don't realize the builds are using something
different. Ultimately, we have to choose which downside we are happy
to live with. The changes here preserve the previous intent of
updating the cache, but it's still a bit fragile.
Fixes: QTBUG-89821
Task-number: QTBUG-85992
Task-number: QTBUG-86866
Change-Id: I8a16753e159bde338e6d2e5dd2ce91fc8ac5c39d
Reviewed-by: Joerg Bornemann <joerg.bornemann@qt.io>
To enable the current Qt version to use a qt.conf that is independent
from the previous version, a file named 'qt<major-version>.conf' is
searched for first. If this file does not exist, a file with the
name 'qt.conf' is searched for.
Task-number: QTBUG-88635
Change-Id: If75b63f72a7bdbdf1de7729ea4ecb00810b58d12
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@gmx.de>
The delta was clearly intended to be used on the total (and still is)
but it also wound up getting stored in the cache, which wouldn't be a
big problem unless the object was removed, in which case we could
incidentally 'free up more space' than intended.
Pick-to: 6.0
Change-Id: Ib2b0f072d30da6d16a93dce60e4c5f6080c109fc
Reviewed-by: Lars Knoll <lars.knoll@qt.io>
In the readFromInotify() function, add a check whether
the ioctl function is successful or not. In the case of
failure, continuing to execute the following code would cause
errors in the read function because there is no data in the buffer.
Change-Id: Id53037e9e48c97c9eb75835048143875275b6370
Reviewed-by: Lars Knoll <lars.knoll@qt.io>
The porting script could not handle this special case.
To achieve the same thing, we'll use an intermediate 'OBJECT' library
which will get the '-fno-lto' build flag if the conditions are met.
Fixes: QTBUG-89426
Pick-to: 6.0
Change-Id: Ied62502ce9c7f7fd6b89fab166f7bccfcd6e7433
Reviewed-by: Joerg Bornemann <joerg.bornemann@qt.io>
Those serve no purpose anymore, now that the .pro files are gone.
Task-number: QTBUG-88742
Change-Id: I39943327b8c9871785b58e9973e4e7602371793e
Reviewed-by: Cristian Adam <cristian.adam@qt.io>
Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
Reviewed-by: Kai Koehne <kai.koehne@qt.io>
In QPropertyObserverPointer::notify and its calling sites,
variable names "alreadyKnownToHaveChanged",
"knownIfPropertyChanged", "propertyChanged", and at its calling site
"knownIfChanged" are used. This is confusing. This patch
changes those four to "knownToHaveChanged". For the logic
implemented it is not necessary to track whether we
have knowledge about having changed and whether it has
actually changed (if we have knowledge) separately.
Change-Id: I90b86b276ab67b2ed70dba4e456cd90220588870
Reviewed-by: Fabian Kosmale <fabian.kosmale@qt.io>
If we create a QBindable from a const property, we should obtain a
read-only interface. Besides implementing this feature, this patch adds
a isReadOnly method to Q(Untyped)Bindable which can be used to check
whether one can modify the property via the bindable interface.
Task-number: QTBUG-89505
Task-number: QTBUG-89469
Change-Id: Ic36949a5b84c5119e0060ed0a1cf4ac94a66f341
Reviewed-by: Ulf Hermann <ulf.hermann@qt.io>
The snippets do not talk about Qt Widgets specifically.
Pick-to: 6.0 5.15
Change-Id: Icc4f243c93189f08124074e4918d535aa37e6560
Reviewed-by: Leena Miettinen <riitta-leena.miettinen@qt.io>
There are still other parts of the CMake API that are not yet
documented. This change only addresses qt_add_executable() and the
Android-related commands it uses.
Fixes: QTBUG-88839
Task-number: QTBUG-84482
Pick-to: 6.0
Change-Id: I761b5ce908d1f62284baabe2d414cd37a0efe83d
Reviewed-by: Joerg Bornemann <joerg.bornemann@qt.io>
Reviewed-by: Leena Miettinen <riitta-leena.miettinen@qt.io>
Reviewed-by: Assam Boudjelthia <assam.boudjelthia@qt.io>
To avoid the mostly hypothetical possibility of failure, delete the
processFinishedNotifier before closing the handle on which it operates.
Previously, because of this, we explicitly disabled the notifier in the
processFinished() function, which made the code unclear. Now, we can
remove that safely, because cleanup() works correctly, and doing it
before calling findExitCode() was not necessary to start with.
Change-Id: Ia7095ded2c7eba8f4d738c6b87c7be41aa3cbbc8
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@gmx.de>
if pid == nullptr, the child process has already exited and the pipe
readers have been stopped. So the call to drainOutputPipes() is
unnecessary here.
Change-Id: I0bed90d08ac879bb0ae178a1cdc37afb9d825314
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@gmx.de>
Otherwise, the user may receive the readyRead() signal just before
started().
Change-Id: I8d6fd18fdfcef0580a3e609100198b03b18b1175
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@gmx.de>
The completion of the child process can take place asynchronously or in
one of the waitFor...() functions. In both cases, we used the same
handler (_q_processDied()), which caused several problems:
a. technically, waitForReadyRead() should have taken into account the
result of the calls to _q_canRead...() slots inside the
_q_processDied() function:
- the user calls waitForReadyRead();
- forkfd descriptor becomes signaled, while a grandchild
process is still alive;
- as readyRead() signal has not been emitted, _q_processDied()
is called;
- the grandchild process writes to stdout pipe;
- now data arrives, and _q_processDied() will collect it, but
won't report it.
b. we had a bug with recursions on Unix:
- death notification comes asynchronously;
- waitForDeadChild() closes forkfd;
- _q_canRead...() emits readyRead();
- a slot connected to readyRead() calls waitForFinished();
- waitForFinished() hangs (forkfd == -1).
c. for blocking functions, drainOutputPipes() was called twice on
Windows.
By introducing a new processFinished() function, we leave the read
operations in the _q_processDied() slot, while the process completion
code is guaranteed to run only once.
Change-Id: I5f9d09bc68a058169de4d9e490b48fc0b35e94cd
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@gmx.de>
In construction of QPropertyBindingPrivate was an initialization
of inlineDependencyObservers with a comment that this is necessary
because of a union. But inlineDependencyObservers is not inside a
union. The comment is outdated and the explicit initialization
superfluous. This patch removes it.
Change-Id: I06544a816533c41af8fc5da3ef44ddd7c18ca86e
Reviewed-by: Lars Knoll <lars.knoll@qt.io>
This is an addendum to commit 0da5726a43b; making the note about paths
starting with ':' being treated as absolute paths by the underlying
QFileSystemEngine, more visible, so that users of the class don't hit
that pitfall.
Add similar docs to QFileInfo.
Pick-to: 5.15 6.0
Change-Id: Ib34d066d31bb673d340ec41422ed4daf911765b8
Reviewed-by: Paul Wicking <paul.wicking@qt.io>
There is no reason to have the startup notifier and the death notifier
be active at the same time, as the former will detect death as well.
Previously, these notifiers were racing, but _q_processDied() ordered
signals by calling _q_startupNotification() manually. Thus, the
started()/finished() sequence was always emitted if the child process
was killed anywhere. Now this ordering is simply not necessary anymore.
This makes it possible to reuse the startup notifier for death
notification.
Change-Id: I5ebed9b5f28b19fe56c80498977a3b21be9288fd
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@gmx.de>
This is assumed in various places, so best checked explicitly.
Change-Id: I475dcf5cbb60c0272dec560acd2893cff73872ce
Reviewed-by: Lars Knoll <lars.knoll@qt.io>
Andreas Buhr
Wang Yicun
Fabian Kosmale
Paul Wicking
Giuseppe D'Angelo
Wang Yu
Thiago Macieira
Alexey Edelev
Joerg Bornemann
Andrei Golubev
Volker Hilsheimer
Alex Trotsenko
Hou Lei
David Faure
Friedemann Kleint
Edward Welbourne
Sona Kurazyan
Craig Scott
Assam Boudjelthia
Nico Vertriest
Ulf Hermann
Mårten Nordheim
Kimmo Ollila
Juha Vuolle
JiDe Zhang
Tor Arne Vestbø
Alex Blasche
Mitch Curtis
Robert Loehning
Allan Sandfeld Jensen
Timur Pocheptsov
Ivan Solovev
Lars Knoll
Kai Köhne
Ville Voutilainen
Nicolas Fella
Christian Ehrlicher
Ralf Habacker
Lu Zhen
Christophe Giboudeaux
Ahmad Samir