In Prepare mode, "buf" is a null pointer. We never dereference it, but we
still compute an offset from it to obtain a pointer to a (then unused)
QMetaObjectPrivater.
clang's UBSan complains about this, so initialize the pointer to nullptr
instead when in Prepare mode.
Pick-to: 6.7 6.6 6.5
Change-Id: Id9d78058f72bb1b44440d07f565374f3eb3c20fd
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
It's a shared list that, in theory, can be written to by multiple
threads. Protect it with a mutex.
To make the change in a clean way, move all the .contains() checks
into a separate function so we can simply hold the lock in there.
Pick-to: 6.7 6.6 6.5
Change-Id: I99ff185346d52e43a3f59f2910a7b2fa6031e3e4
Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
Add the USE_XCB_<component>_STATIC switches that force using the
static versions of xcb components if necessary.
The xcb components needs to be built with -fPIC option, to make the
static linking possible.
When Qt is built staticaly USE_XCB_<component>_STATIC needs to be
passed to the application command line, since in static case xcb
libraries are resolved and linked when configuring application, but
not when building Qt.
Pick-to: 6.7 6.6 6.5
Task-number: QTBUG-120769
Change-Id: I274e9cf4a6a111dc87ad84bb7f49cad76a3d364a
Reviewed-by: Liang Qi <liang.qi@qt.io>
Silences clazy-copyable-polymorphic.
The Q_DISABLE_COPY_MOVE() disables the implicit default constructor,
so bring it back with =default.
Amends aa8393c94f.
Pick-to: 6.7 6.6 6.5
Change-Id: I44f0b19208d3865fb84e277bcaf65d7f4541d368
Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
Reviewed-by: Ivan Solovev <ivan.solovev@qt.io>
[ChangeLog][QtWidgets][Deprecation Notices][QCheckBox]
stateChanged(int) has been deprecated in favor of
checkStateChanged(Qt:CheckState).
Found in API-review. Amends 5a96d13bb5.
Pick-to: 6.7
Change-Id: I6ff4aa38c2f43622ba4b127420aff83790785455
Reviewed-by: Fabian Kosmale <fabian.kosmale@qt.io>
Reviewed-by: Ivan Solovev <ivan.solovev@qt.io>
Add Q_PROPERTY to follow the current Qt style and simplify the
documentation.
Task-number: QTBUG-120566
Change-Id: Ibbafde35bb5600453ed3a3de36b6bf2b5c560e5f
Reviewed-by: Kai Köhne <kai.koehne@qt.io>
Add Q_PROPERTY to follow the current Qt style and simplify the
documentation.
Task-number: QTBUG-120566
Change-Id: I593ba9f322273991661c1d14fb51dcb23678ea0f
Reviewed-by: Kai Köhne <kai.koehne@qt.io>
Fix the coding style to match the current Qt style.
Pick-to: 6.7
Fixes: QTBUG-68661
Fixes: QTBUG-120909
Change-Id: I314ca9da8a03727e3e0336a23fce1ce9d065d3a4
Reviewed-by: Qt CI Bot <qt_ci_bot@qt-project.org>
Reviewed-by: Kai Köhne <kai.koehne@qt.io>
When a `.moc` file is included in a source file and that source passed
to `qt_wrap_cpp`, Users should add the generated `.moc`s path to the
target's include path. Since we don't share anything about the output
path of generated files by `qt_wrap_cpp`, it makes sense to add in
inside `qt_wrap_cpp`. And also, the generated `.moc` file is added
to target's source to complete the dependency graph. Otherwise, Users
need to get output variable and pass it to target's sources.
* Update docs
* Add test
[ChangeLog][Build System] qt_wrap_cpp will accept .cpp files from now
on. When .cpp a file is passed to qt_wrap_cpp, TARGET parameter
becomes required. Generated .moc files are added to target's sources
inside qt_wrap_cpp. That's why the output parameter will not contain
generated .moc files.
Fixes: QTBUG-113402
Change-Id: I54dd2b1ff8e5c9ba457b1eb4f73b0a8190d9c659
Reviewed-by: Alexandru Croitor <alexandru.croitor@qt.io>
Reviewed-by: Joerg Bornemann <joerg.bornemann@qt.io>
We've previously used the zip-file form, but that's not been published
for CLDR v44.1 - the advice on the list was to use github
instead. That, however, has ↑↑↑ as a special value for fields, meaning
to inherit from a prent locale. So special-case that value. I have
verified that v44 from the zip file produces identical results to v44
from github, with this minor fix. As it happens v44.1 also produces
identical results.
Pick-to: 6.7 6.5
Change-Id: I6eb0aedda7556753cdc83bb9d76652fbb68dc669
Reviewed-by: Ievgenii Meshcheriakov <ievgenii.meshcheriakov@qt.io>
The constructor and update() method for the system locale were using
GetUserDefaultLCID(), where query() and the fallback locale first
checked the LANG environment variable, leading to inconsistent results
if the user set the environment variable to something different from
the system's configured locale.
Break out the logic of parsing %LANG% into a static helper, replace
the existing parsing with a call to this and add a helper to get the
right ID to use, possibly via it, using GetUserDefaultLCID() as
fall-back. Drive-by: initialize substititionType in its declaration.
Also look up %LANG% each time we want it; it's not that expensive,
given how rarely this code is called, and client code could change its
value at runtime. Partially inspired by a patch from Wladimir
Leuschner <wladimir.leuschner@qt.io>
Pick-to: 6.5 6.6 6.7
Fixes: QTBUG-120961
Change-Id: Ie706c7089bd2b3757a3eab627723ec34a5e2b07f
Reviewed-by: Ivan Solovev <ivan.solovev@qt.io>
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
When qfloat16 uses float as an underlying type, the operators cannot be
constexpr, because operator float() is not constexpr.
However, operator NativeType() is, so we can make the relational
operators constexpr when we are using a native 16-bit float as an
underlying type.
To avoid code duplication, introduce new temporary macros for
constexpr and Q_DECLARE_PARTIALLY_ORDERED_LITERAL_TYPE.
Extend the tests to verify that the operators are constexpr when
native float16 type is used.
Task-number: QTBUG-119433
Pick-to: 6.7
Change-Id: I001b087d78c398c71b71a504b65c316199dd4792
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Reviewed-by: Marc Mutz <marc.mutz@qt.io>
The problem with the QTestPrivate::testAllComparisonOperators() and
QTestPrivate::testEqualityOperators() functions is that if they fail,
they point into the helper function, but not into the actual test that
called the helper function. This is specially annoying when some test
calls the helper function multiple times.
This patch introduces the helper macros QT_TEST_ALL_COMPARISON_OPS and
QT_TEST_EQUALITY_OPS that wrap the respective function calls together
with the QTest::currentTestFailed() check. If the test has failed,
the macro generates a meaningful debug message with the original file
name and line number.
This patch also applies the new macros to qtbase.
Task-number: QTBUG-119433
Pick-to: 6.7
Change-Id: Iad709de45e5bf53c82e7afa8e9f51e9275c1e619
Reviewed-by: Marc Mutz <marc.mutz@qt.io>
The compiler knows how large the array is, so pass the size and drop
the terminating null.
Pick-to: 6.7
Change-Id: I5795c89834b8395f04d1777d737d080f19be2a21
Reviewed-by: Eskil Abrahamsen Blomfeldt <eskil.abrahamsen-blomfeldt@qt.io>
Reviewed-by: Konstantin Ritt <ritt.ks@gmail.com>
Static and thread-local objects should be constexpr or constinit if possible.
Task-number: QTBUG-100485
Pick-to: 6.7 6.6 6.5
Change-Id: I29088798a50d6278252c9088e7c191c4214b2e5b
Reviewed-by: Qt CI Bot <qt_ci_bot@qt-project.org>
Reviewed-by: Fabian Kosmale <fabian.kosmale@qt.io>
4c60a11d8f had to be reverted, because it
broke restoring the state of hidden, floating dock widgets.
Extend tst_QDockWdget::hideAndShow() to test this functionality.
Blacklist test function on Apple, because Apple machines on CI show
dock widgets with the main window.
Task-number: QTBUG-120191
Pick-to: 6.7
Change-Id: Ibc6121a99f15f33d62ca0a9376dbe9fafbe20290
Reviewed-by: Volker Hilsheimer <volker.hilsheimer@qt.io>
Reviewed-by: Tor Arne Vestbø <tor.arne.vestbo@qt.io>
This partially reverts eb0abd9789 so we
depend on __SIZEOF_INT128__ in this structure, not on QT_SUPPORTS_INT128
so the Id128Bytes union has a quad-word integer as a member at all
times, even in QtBluetooth's removed_api.cpp.
That's required because on some ABIs, the presence of this member causes
the type to be passed differently, usually in registers (we fixed the
stack problem with the alignas(16) in that commit).
The "Itanium Software Conventions and Runtime Architecture Guide" did
the same back in 2001; section 8.5.1 says integers from 65 to 128 bits
are passed "Next Even" and so are aggregates with alignment of 16 bytes.
This rule seems to also exist in the AArch64's AAPCS64[1], which says,
in rule C.10, that the type should be passed in even registers:
C.10 If the argument has an alignment of 16 then the NGRN is rounded up
to the next even number.
(NGRN = Next General Register Number)
It's unclear whether this applies to the union type as a whole or to its
individual members. If the latter, then Clang may be buggy.
[ChangeLog][Important Binary-Incompatibity Fixes] The QUuid::Id128Bytes
type had a loose definition that could cause it be passed incompatibly
between functions, in some architectures, depending on whether GNU
extensions were allowed. This is now fixed, but may cause code compiled
with Qt 6.6.0 and 6.6.1 to fail when recompiled with 6.6.2 or later.
[1] https://github.com/ARM-software/abi-aa/blob/main/aapcs64/aapcs64.rst#682parameter-passing-rules
Fixes: QTBUG-119248
Pick-to: 6.6 6.7
Change-Id: I6e2677aad2ab45759db2fffd17a86a208f788cb7
Reviewed-by: Qt CI Bot <qt_ci_bot@qt-project.org>
Reviewed-by: Marc Mutz <marc.mutz@qt.io>
When QDeferredDeleteEvent was declared in qcoreevent.h, it
had Q_CORE_EXPORT. After moving to qcoreevent_p.h, it changed
to Q_AUTOTEST_EXPORT. The correct macro should be preserved.
Noticed that this change was needed to make GammaRay work
with Qt 6.7
Pick-to: 6.7
Change-Id: I6fab494d9f60b1261edce6f326a3b96467ff7da1
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
This is a function template defined inside a namespace, `static` makes
no difference here.
Change-Id: I2eb52b4fb4432c269b2449ae37550bd0be0dbbce
Pick-to: 6.7
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
The static function appeared in two places, and in a unity-build
this fails quite visibly.
Pick-to: 6.7
Change-Id: I60000d01194a2c79ca9c101f2a6d3f77f469f1a7
Reviewed-by: Alexey Edelev <alexey.edelev@qt.io>
Reviewed-by: Nodir Temirkhodjaev <nodir.temir@gmail.com>
A QMultiMap is used, while only qset.h is included.
FTBFS w/o precompiled headers.
Include qmap.h as well.
Pick-to: 6.7 6.6 6.5
Change-Id: I5f6e1d204bfa94cc6a81f6883c7a7b0ee1e9963c
Reviewed-by: Ivan Solovev <ivan.solovev@qt.io>
This reverts commit ca2f46c04c.
Reason for revert: <BIC in 6.6 found by Marc Mutz>
Pick-to: 6.7 6.6
Change-Id: Ia5b8849e55ebccb514bc753ce1d31855d91e2406
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@qt.io>
The CHECK_SINGLE_OPERATOR macro is used inside FOR_EACH_CREF macro,
which, like all other FOR_EACH_* macros in Qt (see qdebug.h or
qmetatype.h), expects its parameter to have a scope or a semicolon.
All other CHECK_* macros are *not* used inside FOR_EACH_CREF, so
they can follow a more traditional approach with the
`do { } while (false)` scope without a trailing semicolon.
This allows to use them in the code like:
if (condition)
MACRO(args);
else
OTHER_MACRO(args);
In the example above, if MACRO ended with the semicolon, adding
one more semicolon would have discarded the else part of the
conditon, leading to compilation error.
Amends bfb237d19a.
Task-number: QTBUG-119433
Pick-to: 6.7
Change-Id: I9a7f17416ba7c37a50f022f685b54e2643e4a9e2
Reviewed-by: Marc Mutz <marc.mutz@qt.io>
Apply some documentation improvements that were suggested before FF.
Task-number: QTBUG-119433
Pick-to: 6.7
Change-Id: I9b1d83c69821e25ae4cd8db0cbf3fa6d6330a6dc
Reviewed-by: Dennis Oberst <dennis.oberst@qt.io>
Reviewed-by: Marc Mutz <marc.mutz@qt.io>
If the size of the documents default font is specified in pixels all
text ranges would get wrapped in a span specifying the font size again.
Change-Id: Iae2ea34d79bb6253badd8a0575489d3fb6409eba
Reviewed-by: Eskil Abrahamsen Blomfeldt <eskil.abrahamsen-blomfeldt@qt.io>
CMake has a built in file extension list for "Source Files", that does
not include qml files. We can't extend that file list with a regex,
so we need to add qml files individually as a finalization action.
We also stuff generated files (both auto-gen as well as other generated
files) into a separate source folder, to focus the project on the user's
files.
Pick-to: 6.7
Change-Id: Ic9a4d9daad088b1874bcf81e75c776537f3ff531
Reviewed-by: Alexandru Croitor <alexandru.croitor@qt.io>
It doesn't always pay to be explicit about SMFs. In the present case,
clang-tidy complains that the move ctors of the class aren't
noexcept. It's a false positive, but just removing all the =default'ed
SMFs makes it go away, too.
Amends 0dba3f6b71.
Pick-to: 6.7
Change-Id: Ib641d4eed9e214b41c143b60494e82790da6fda8
Reviewed-by: Volker Hilsheimer <volker.hilsheimer@qt.io>
Reviewed-by: Mårten Nordheim <marten.nordheim@qt.io>
We do have special macros for this case, so use one of them!
Amends c2956f8f76
Found in 6.7 API review.
Pick-to: 6.7
Change-Id: I205966d6f7363abf1a26f59b04555b1a48a4e696
Reviewed-by: Marc Mutz <marc.mutz@qt.io>
And make sure it cannot happen again by adding the
QT_NO_CAST_FROM_ASCII define to the Network module.
Amends 5346404da6
Change-Id: Ib10d7822c514caaf40b6b36ee71947c034daf338
Reviewed-by: Ivan Solovev <ivan.solovev@qt.io>
Previously we would more or less wait for the first connection to finish before
initiating more connections. This was just happening by virtue of
calling startNextRequest when a request was finished.
Since we have already determined the preferred network layer, let's just
try to start more connections. In case we are using https this doesn't
make a difference because we will try http2 upgrade first before allowing
multiple connections. Unless, of course, http2 has been disabled.
Fixes: QTBUG-120619
Pick-to: 6.7 6.6 6.5
Change-Id: Id3c3be59fd07120bff574d63481119e5ed9eb88a
Reviewed-by: Qt CI Bot <qt_ci_bot@qt-project.org>
Reviewed-by: Timur Pocheptsov <timur.pocheptsov@qt.io>
Due to a recent change in QDoc, the outputdir path has changed.
Task-number: QTBUG-119500
Change-Id: I35bc9f79eb0cd82bbd42b39fef37204189c4a1de
Reviewed-by: Topi Reiniö <topi.reinio@qt.io>
Says clang-tidy:
destructor of '(unnamed struct at qmetatype.cpp:966:14)' is public and non-virtual
in file:src/corelib/kernel/qmetatype.cpp line:966 col:14
static const struct : QMetaTypeModuleHelper
Yes, these classes are polymorphic (because the base class is). Yes,
the destructor is non-virtual (because the base class' one isn't, but
it's also protected, so fine).
But these classes are not used as base classes, so suppress the
warning.
Pick-to: 6.7 6.6 6.5 6.2
Change-Id: I75be86bca36a4a0e93d72acb1a0d2fe0dca1c505
Reviewed-by: Fabian Kosmale <fabian.kosmale@qt.io>
Spell out some API names in enum.
Some functions are now protected.
Remove property for autoRenderTarget.
textureFormat -> colorBufferFormat.
Used "fixed" instead of "explicit" and follow
the above naming, so that explicitSize becomes
fixedColorBufferSize.
Pick-to: 6.7
Change-Id: I2fd6ad46033313a3febbb8846146021d5dd11010
Reviewed-by: Volker Hilsheimer <volker.hilsheimer@qt.io>
We can not guarantee 128bit alignment on 32bit windows
Pick-to: 6.7
Change-Id: I16bd3bded16bd29f4cb84b0cdf2671e4d9a19601
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
QWindowsFontDatabaseBase::invalidate() was defined inside an
the declaration was not.
Since the font file loader is only protected by
QT_CONFIG(directwrite), we move it out from the previous block
and add an #if for the contents instead.
Pick-to: 6.7
Change-Id: Iedc02cceb15fe3d25d21b49af1486659a4d6373b
Reviewed-by: Volker Hilsheimer <volker.hilsheimer@qt.io>
All that is required is that we write the declaration as if they weren't
inline, like we used to for C++98 non-constexpr static members. Amends
090991123d.
This is required with MinGW and happens with both GCC and Clang: when a
static constexpr variable found in an exported class is used in a
context that requires getting its address, the compiler emit a DLL-
importing statement. For example:
void f(const std::chrono::milliseconds &ms);
void f()
{
f(QNetworkRequest::DefaultTransferTimeout);
}
Emits with GCC:
_Z1fv:
movq __imp__ZN15QNetworkRequest22DefaultTransferTimeoutE(%rip), %rcx
jmp _Z1fRKNSt6chrono8durationIxSt5ratioILx1ELx1000EEEE
And Clang:
_Z1fv: # @_Z1fv
# %bb.0:
movq __imp__ZN15QNetworkRequest22DefaultTransferTimeoutE(%rip), %rcx
jmp _Z1fRKNSt6chrono8durationIxSt5ratioILx1ELx1000EEEE # TAILCALL
Pick-to: 6.7
Fixes: QTBUG-121135
Change-Id: I76ffba14ece04f24b43efffd17aae1a98071c570
Reviewed-by: Juha Vuolle <juha.vuolle@qt.io>
Reviewed-by: André Klitzing <aklitzing@gmail.com>
Reviewed-by: Mårten Nordheim <marten.nordheim@qt.io>
If a QFont is moved to a different thread (B) than where it belongs (A),
and the font cache is then purged on thread A, the last remaining
reference to the engine will be on thread B. When this QFontEngine is
later replaced with one created on B, we end up deleting A's QFontEngine
on thread B.
This caused QFreetypeFace::release() to be called on the wrong thread and
the face would never be removed from the thread-local Freetype cache in
A. Hence that cache would contain a dangling pointer to the freetype face
which we would later end up fetching.
To avoid this, we make sure the thread-local cache itself increases
the ref count of the face. That way, the only time it will be deleted
on a different thread is when the cache has been destroyed because the
thread has shut down.
If the last reference (except the cache reference) to a face is cleared
on a different thread, we keep it in the cache until later. It will
then be in a "deferred delete" mode and will be deleted as soon as
possible. This is done either when the thread shuts down, when a lookup
causes the "deferred delete" face to be returned, or when release()
is called on any font on the thread (at which point we will purge all
faces that only have the single cache reference.)
Pick-to: 6.7
Fixes: QTBUG-118867
Change-Id: Ifa07a9cb6f4cd3e783e12a73d8b283e70d6fb474
Reviewed-by: Mårten Nordheim <marten.nordheim@qt.io>
Since https://gitlab.kitware.com/cmake/cmake/-/merge_requests/8919 is merged, `qtbase` should be updated accordingly. If `CMake` is equal or greater than `3.29`. `timestamp_$<CONFIG>` is used instead of `timestamp`.
This commit reimplements 8041bfba47 which is reverted with 4b1d96ea1b
Change-Id: I335e3855c1a11c404202ae7b74f0c356b8fe53bd
Reviewed-by: Alexandru Croitor <alexandru.croitor@qt.io>
Reviewed-by: Leena Miettinen <riitta-leena.miettinen@qt.io>
QSystemLocale is setting the global window.onLanguageChanged
property which may cause conflicts in cases where there
are more than one Qt app instance on the same web page.
In addition the QSystemLocale destructor never runs,
which breaks also for repeated QLocale instantiations
on the same page (for example for the QLocale auto-
test)
Pick-to: 6.7
Change-Id: I85af0587527b2922804fc866c83864f677700e6d
Reviewed-by: Tor Arne Vestbø <tor.arne.vestbo@qt.io>
Use the QT_SUPERBUILD procedure when configuring the qtbase
documentation-only build. This only makes sense for qtbase since we
disallow building it using existing Qt installation. Other repositories
should be configured using qt-configure-module script from the existing
Qt installation and simply run 'ninja docs'.
Fixes: QTBUG-120485
Pick-to: 6.7 6.6 6.5
Change-Id: Iafed5c17bea5c61edc239f08045922497215fb73
Reviewed-by: Joerg Bornemann <joerg.bornemann@qt.io>
Mårten Nordheim
Fabian Kosmale
Alexey Edelev
Marc Mutz
Christian Ehrlicher
Orkun Tokdemir
Edward Welbourne
Ivan Solovev
Volker Hilsheimer
Axel Spoerl
Thiago Macieira
Matt Aber
Aleix Pol Gonzalez
Ahmad Samir
Viktor Arvidsson
Tor Arne Vestbø
Paul Wicking
Laszlo Agocs
Allan Sandfeld Jensen
Eskil Abrahamsen Blomfeldt
Morten Sørvig
Liang Qi