A deadlock can occur if the user does
QLocalSocket *ls = new QLocalSocket;
ls->moveToThread(t);
...
delete ls;
Then QLocalSocket calls QWindowsPipeReader::stop() in a different thread
than the I/O operation is running in. The waitForNotified(-1) call would
then wait indefinitely until the I/O thread is in alertable wait state
again. Especially on application shut down this might never be the case,
and the application would deadlock.
Solve this by detaching the Overlapped object from the
QWindowsPipe{Reader|Writer} in stop() and delete it in the callback.
Task-number: QTBUG-61643
Change-Id: Ie262d75c5fd92ac7cf7dfcdbf1519050be9fd3c4
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@qt.io>
We forgot to update the warnings when removing qt_cancelIo.
Also, use %p instead of %x, because HANDLE is void*.
This amends commit fade2958.
Change-Id: Ia11d7d094aa6beb939e0be4bbe4ab3654eaa1c02
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@qt.io>
The classes themselves and their equality operators are used in
constexpr functions/ctors (in QKeyValueIterator) so Visual Studio 2017 expects
them to be marked constexpr as well. Currently this causes a compilation error
when instantiating a QKeyValueIterator using either of these iterators.
Change-Id: I2e3eeaf3b3f11f381a63875e6575dfd82fe56fcb
Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
The references to the this pointer look somewhat alien in the
documentation, because it isn't part of the signature. Rather make
the relationship explicit.
Change-Id: I6de516e165ea6e9c4ee2898836e9490fbaf4545c
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
The documentation claims that QGuiApplication forwards the top-level
windows. However this appears to be a search & replace mistake from the
QApplication <> QGuiApplication separation times. Only QApplication
forwards the event to top-level widgets and changeEvent() is a virtual
method of QWidget. Nothing is implemented for plain QGuiApplication and
QWindow.
Change-Id: I71b05ecebc90f7c28e150590764438ebaa90e88f
Reviewed-by: Michael Brasser <michael.brasser@live.com>
QVariant claims to be able to QVariantHash and QVariantMap, but the
actual conversion implementation is missing.
Task-number: QTBUG-61471
Change-Id: I0cba74642aa77dc423effed289bc7619922a89eb
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Older versions of CMake expanded `@var@` in CMake code, so this could be
expanded at this location rather than the `string(CONFIGURE)` call if
`module` were set inadvertently. Instead, hide the literal `@` symbol
from CMake, but not from the string.
This avoids a CMP0053 warning for projects using Qt5 with a minimum
version set lower than 3.1 and silent bugs with projects explicitly
setting CMP0053 to OLD.
Change-Id: I0e4a86469fdf69b8706387799ab9b17498b8d1ca
Reviewed-by: Stephen Kelly <steveire@gmail.com>
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@qt.io>
Bail out of QStorageInfoPrivate::doStats() should an error occur
and set the ready/valid flags accordingly.
Task-number: QTBUG-6039
Change-Id: Id5354b31329d951599ae991aa7edde0515c90514
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
We can't open a file for reading if the file is open by another process
(or by ourselves) without sharing permitted. So ask for no access just
so we can get a handle to it.
Change-Id: I998653739e1cec2a58a07a6593b6ff87c1d59dd1
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
QFileSystemEngine::id() will stat(2)/CreateFile in order to get the ID
of the file anyway, so we don't need to use QFile::exists() to check if
the destination exists. Instead, rely on id() returning a null value to
indicate error. On Windows, it's possible that the calls to either
GetFileInformationByHandle or GetFileInformationByHandleEx might fail,
but we ignore those.
Change-Id: I1eba2b016de74620bfc8fffd14ccaebcbed64419
Reviewed-by: Jędrzej Nowacki <jedrzej.nowacki@qt.io>
Doesn't affect our current builds since it's just a marker for the
linker on what sections should be merged. Unless you're mixing
namespaced and non-namespaced static builds into one executable.
Change-Id: Ia53158e207a94bf49489fffd14c7bc294fccf8f9
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@qt.io>
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Code like:
QElapsedTimer timer;
timer.start();
QTest::qWait(30);
QVERIFY(timer.elapsed() >= 30);
is failing, because qWait sleeps in increments of 10 ms and the last
chunk may be off by less than one millisecond, so we end up sleeping too
little and thus returning before 30 ms have elapsed.
This matches the QElapsedTimer::elapsed() code that rounds down:
return nsecsElapsed() / Q_INT64_C(1000000);
Task-number: QTBUG-61741
Change-Id: Ic3a088f9f08a4fd7ae91fffd14cea4a91d3f51a8
Reviewed-by: Allan Sandfeld Jensen <allan.jensen@qt.io>
[ChangeLog][QtCore][QLocale] Fixed the conversion of QTime to string
form and parsing from string form to always treat the value as the
decimal fraction of the seconds component. That is, the string format
".z" produces/parses ".2" for 200 milliseconds and ".002" for 2
milliseconds. Use of "z" or "zzz" is discouraged outside decimal
fractions to avoid surprises.
Task-number: QTBUG-53565
Change-Id: Ia19de85ad35e4eb7bb95fffd14792caf9b4a5156
Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
QWindowsFileSystemWatcherEngine uses one change notification
per directory to watch directories or files within that
directory. Adding files and their directories in a sequence
caused the value in QWindowsFileSystemWatcherEngineThread::HandleForDirHash
to be overwritten.
Relax the check for the flags (watcher attributes) to use >= and recreate
the change notification of a directory should its flags be insufficient.
This triggers when a file is added after its directory since files
require more attributes.
Task-number: QTBUG-61792
Change-Id: I371a72f1934fa82c53aaf84beb907825031f1c81
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
This protects against the file having been renamed or deleted. We'll
still operate on the open file, regardless the name it may have on the
filesystem.
Change-Id: I1eba2b016de74620bfc8fffd14cca85cfd672e6d
Reviewed-by: Lars Knoll <lars.knoll@qt.io>
It's not dangling only because of QFileSystemMetaData's construction:
the nativeFilePath() function returns a member variable. Since
QByteArray COWs, the pointer that we stored would not be freed.
But this was dangerous, since any change to the "entry" variable could
cause it to invalidate the member variable and the pointer to become
dangling. This line is only as old as this entire file is.
Change-Id: I8d96dea9955d4c749b99fffd14cda4d8e2cc5e5b
Reviewed-by: Lars Knoll <lars.knoll@qt.io>
The deserialized user type is now shown to the user to figure which
QMetaType registration is missing.
Change-Id: I4b7624827e479b1bea67065ce3542183b7355165
Reviewed-by: Jędrzej Nowacki <jedrzej.nowacki@qt.io>
First, it can never return, so we can mark it Q_NORETURN and add an
std::termianate at the end. Though if it did, we'd end up in a null-
pointer dereference crash in the caller.
Second, add Q_DECL_NOTHROW to it. It can't throw, but it terminates
execution. This also prevents both puts and fprintf from escaping via
pthread asynchronous cancellation on Linux/glibc.
Third, don't use QMessageLogger, since that allocates memory and
actually uses QString. If we really are in an OOM situation, then
QString's failed allocation would recurse back into qt_check_pointer.
We'd compound the OOM situation with a stack overflow...
Change-Id: Ia53158e207a94bf49489fffd14c81c47971f4e82
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@qt.io>
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
qt_assert and qt_check_pointer get the function name and line number
from the caller (the functions are called from the Q_ASSERT and
Q_CHECK_PTR macros, respectively), so we don't need to capture the
context from those two functions. Instead, pass the context to
QMessageLogger for proper logging.
I've left the file name and line number in the assertions, for users who
did not add them to their message log pattern, but I've removed from the
almost never used qt_check_pointer function.
Note: how useful is it that we allocate memory in response to failing to
allocate memory?
Change-Id: Ia53158e207a94bf49489fffd14c81b359c5b6537
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@qt.io>
Reviewed-by: Kai Koehne <kai.koehne@qt.io>
This compiler seems to require explicit initialization of all member
variables in a constexpr constructor, even if they have an implicit
default constructor of their own. We probably fixed the rest of Qt a
couple of years ago, but not these two places because they were arrays
and those require the C++11 syntax for uniform initialization.
All compilers that support constexpr do support uniform initialization.
MSVC 2015 fixed our issues with it on the same update.
Change-Id: Ibc1eb23e3ae093f5c6928ded3a041be35eb9baae
Reviewed-by: Marc Mutz <marc.mutz@kdab.com>
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Since the result is an actual zero, this section of code looking for
underflows kicks in. But we forgot to take the capital letter into
account when parsing the number.
Task-number: QTBUG-61350
Change-Id: Ia53158e207a94bf49489fffd14c6abbd21f0bac0
Reviewed-by: Olivier Goffart (Woboq GmbH) <ogoffart@woboq.com>
The code was trying to open a local file with the same path
as the remote URL, which is unnecessary and wrong in the unlikely
case where such a file would exist.
Spotted by Christoph Feck when reading the code.
Change-Id: I1d77e5781cf606b025d2877f48a9914dd1e36b1d
Reviewed-by: Giuseppe D'Angelo <giuseppe.dangelo@kdab.com>
This provides a helper function which does the check on the string.
QFileInfo::isRoot() in addition checks for the existence of
the directory, which can cause hangs with network drives.
Use the new function in appropriate places in QtWidgets.
Task-number: QTBUG-6039
Change-Id: I54d0d860713e82b28fa4069a5345b042337f9c52
Reviewed-by: Joerg Bornemann <joerg.bornemann@qt.io>
A deadline timer is more correct for timeouts.
Also starts the timer before trying to acquire the mutex for more
precise timeouts.
Task-number: QTBUG-58745
Change-Id: I230266a3a5d7b7af33981efed4e6882e5727a41c
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Add a note saying that invalid input to QByteArray::fromPercentEncoding
and QUrl::fromPercentEncoding will produce invalid output, and provide
an example.
Change-Id: Icc68f59c23cf199640b646cd4a6ca8e4808a3f71
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
In theory, there's nothing wrong with having it in the child process. In
practice, we've found that strerror/malloc can hang: if an application-
wide lock was held by another thread before fork(), the child process
could wait forever for an unlocking that will not happen (no threads
running). See https://sourceware.org/bugzilla/show_bug.cgi?id=19431
As an added bonus, we now use qt_error_string(), which may produce
slightly different text from strerror.
[ChangeLog][QtCore][QProcess] Added a workaround for a rare race-
condition bug in some C libraries that caused the child process started
by QProcess to hang after trying to launch a non-existent executable or
change to a non-existent directory.
Task-number: QTBUG-61634
Change-Id: I1eba2b016de74620bfc8fffd14cbce4b9f9af69b
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@qt.io>
According to qcompilerdetection.h, the Dinkumware C++ library does not
have std::initializer_list, even though the compiler supports it.
Add the missing Q_COMPILER_ guards.
Change-Id: I84a7d5054c00dba38bcde15e277ceb0ee05e6cd7
Reviewed-by: Jake Petroules <jake.petroules@qt.io>
Reviewed-by: James McDonnell <jmcdonnell@blackberry.com>
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
The values do not need to be 2 x 32-bits. Eight bits suffice for a long time to come.
Should save 189 bytes in text size on Android builds.
Change-Id: I78e31e7caa7a698f41c66d7bbac58a766c6e8834
Reviewed-by: Jake Petroules <jake.petroules@qt.io>
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
The copy constructor is not disabled by any user-defined move or copy
special member function, and thus does not need to be = default'ed.
Change-Id: I90586d25756885ac77f0946c147079efb5d1b1e0
Reviewed-by: Jake Petroules <jake.petroules@qt.io>
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Fix developer build with GCC 7, which complains:
codecs\qgb18030codec.cpp:387:37: error: 'buf[1]' may be used uninitialized in this function [-Werror=maybe-uninitialized]
codecs\qgb18030codec.cpp:8908:17: error: 'buf[0]' may be used uninitialized in this function [-Werror=maybe-uninitialized]
codecs\qgb18030codec.cpp:536:37: error: 'buf[1]' may be used uninitialized in this function [-Werror=maybe-uninitialized]
codecs\qgb18030codec.cpp:535:37: error: 'buf[0]' may be used uninitialized in this function [-Werror=maybe-uninitialized]
Change-Id: I7e98355c625276c74792707f22d4318c0cc9be6a
Reviewed-by: Marc Mutz <marc.mutz@kdab.com>
The type-cast to int for qle_signedbitfield was wrong for all cases
where width + pos != 32. The class is currently only used two places
though, both where that happen to apply.
Change-Id: I108c565b75c9f29dd49b5e2e39f84910d17ead85
Reviewed-by: Lars Knoll <lars.knoll@qt.io>
it contains no code.
Change-Id: Ie8a43abb2db3d040f7046206adf2bf555960dd9c
Reviewed-by: Lars Knoll <lars.knoll@qt.io>
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
The motivation for this change is to make it simple to pass a
correctly sorted environment block to Win32 CreateProcess(). It is
also nice in other contexts that the environment variables are
sorted. The change is made for all platforms. This keeps it simple and
the only ill effect is slightly slower lookups.
Concerning the environment block passed to Win32 CreateProcess:
The environment block that is passed to CreateProcess() must be sorted
case-insensitively and without regard to locale. See
https://msdn.microsoft.com/en-us/library/windows/desktop/ms682009(v=vs.85).aspx
The need for sorting the environment block is also mentioned in the
CreateProcess() documentation, but with less details:
https://msdn.microsoft.com/en-us/library/windows/desktop/ms682425(v=vs.85).aspx
Task-number: QTBUG-61315
Change-Id: Ie1edd443301de79cf5f699d45beab01b7c0f9de3
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Also use auto for iterators to vars. This is a small refactoring in
preparation for changing type of vars to QMap.
Task-number: QTBUG-61315
Change-Id: I5731d7916b6f54a0da5be2da378c09a7688bd870
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
When punctuation is ignored then the kUCCollatePunctionSignificantMask
should not be set. This was originally thought to not be working due to
a bug on the Apple platforms, but this is not the case.
[ChangeLog][Platform Specific Changes][macOS][iOS] QCollator now
respects the ignorePunctuation property on Apple based platforms
correctly.
Task-number: QTBUG-41978
Change-Id: I62044076387d6e4479f4aaef3c2f48f49dbd160e
Reviewed-by: Lars Knoll <lars.knoll@qt.io>
Commit 418184c2a0 set some extra defines
that Clang and GCC do set so that MSVC and ICC builds would properly get
the features detected. But that meant we set them with Clang and GCC
(technically, set them again, but to the same value so no warning was
printed).
Don't do that. This commit allows me to use "-march=native -mno-rdrnd"
to disable the unconditional use of RDRAND instruction. That's required
to valgrind any applications, as the current version (3.12) does not
have support for that instruction.
vex amd64->IR: unhandled instruction bytes: 0x48 0xF 0xC7 0xF0 0x48 0x8B 0x55 0xE8 0x48 0x89
vex amd64->IR: REX=1 REX.W=1 REX.R=0 REX.X=0 REX.B=0
vex amd64->IR: VEX=0 VEX.L=0 VEX.nVVVV=0x0 ESC=0F
vex amd64->IR: PFX.66=0 PFX.F2=0 PFX.F3=0
==78321== valgrind: Unrecognised instruction at address 0x4ef159c.
==78321== at 0x4EF159C: _rdrand64_step (immintrin.h:208)
==78321== by 0x4EF159C: qt_random_cpu(void*, long long) (qrandom.cpp:95)
Change-Id: Ia3e896da908f42939148fffd14c6884501de4fa4
Reviewed-by: Allan Sandfeld Jensen <allan.jensen@qt.io>
QTimeZone("UTC") should be valid, as "UTC" appears in the list of
availableTimeZoneIds(), and tst_QTimeZone::dataStreamTest() constructs
timezones like this, which are considered valid.
The internal representation of a QTimeZone("UTC") as created by
QTimeZone::QTimeZone(const QByteArray &ianaId) is a QUtcTimeZonePrivate
which isValid(), so the containing QTimeZone isValid() too.
When QTimeZone is serialized into a QDataStream, it calls
tz.d->serialize(ds) which is QUtcTimeZonePrivate::serialize. This
writes QStringLiteral("OffsetFromUtc") followed by the IANA ID and
the offset (etc.) to the datastream.
When QTimeZone is deserialized it looks for this marker string, and if
present, it passed all of the parameters to the QTimeZone constructor
(not just the name). However, that constructor does not support standard
IANA timezones (only custom ones), and when it detects that the supplied
IANA ID is actually listed in availableTimeZoneIds(), it leaves the
pointer to the QTimeZonePrivate uninitialized (NULL), which leaves
the QTimeZone invalid (isValid() returns false).
Thus, a valid timezone which was serialized and then deserialized has
become invalid. This also affects serialization of QDateTimes with
timezones.
Fixed by calling the name-only constructor first, which works (only) for
IANA standard timezones and leaves the QTimeZone invalid (isValid()
returns false) otherwise. In which case, we can call the many-argument
contructor to create a custom timezone with the same offset as the one
which was originally serialized.
[ChangeLog][QtCore][QTimeZone] Fixed sending IANA standard UTC-offset
QTimeZones through QDataStream, which previously came out invalid after
deserialization.
Task-number: QTBUG-60595
Change-Id: Id9c47e8bda701faae4d800e012afb6db545b2fe9
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@qt.io>
Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
Change-Id: Id6533c8a444854f6215f6e47000875ef9751905b
Reviewed-by: Gabriel de Dietrich <gabriel.dedietrich@qt.io>
Reviewed-by: Lars Knoll <lars.knoll@qt.io>
Commit d56c6cf7a4 was incorrect. It was a
nice try, but on a 64-bit Mac machine (x86_64 CPU), it returned
hw.cputype = 7, which is CPU_TYPE_X86. CPU_TYPE_X86_64 is only used in
Mach-O slices for fat binaries and does not reflect hw.cputype.
Task-number: QTBUG-61205
Change-Id: Ia3e896da908f42939148fffd14c54b3050b8e64b
Reviewed-by: Jake Petroules <jake.petroules@qt.io>
The hang happend when using QtAndroidPrivate::requestPermissions before
the QApplication::exec. Android UI calls "sendRequestPermissionsResult"
which was blocking until the event is delivered, but the qt main loop is
blocked and waits for the main surface to be created by the Android UI
thread which is already blocked.
With this patch sendRequestPermissionsResult won't block for the result
to be delivered.
Change-Id: I48ada65fe9ea63471ab46d8a9d839ba1b91d17b3
Reviewed-by: Christian Stromme <christian.stromme@qt.io>
Clang's definition of the __EXCEPTIONS macro is inconsistent across
platforms. When compiling for Darwin, Clang 3.6 and newer will set the
token when exceptions are enabled in either C++ or ObjC. This change
adds the reliable check described in the Clang 3.6 release notes to
ensure that QT_NO_EXCEPTIONS is defined when required.
The check requires the use of the Clang-specific __has_feature()
syntax for which a new proxy macro QT_HAS_FEATURE(x) is added in
qcompilerdetection.h
Task-number: QTBUG-61034
Change-Id: Ie7b482dfa1a4a5b700a6b97562c26b626be1fc04
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Analysis proves this is a false positive:
qarraydataops.h:69:17: error: ‘void* memset(void*, int, size_t)’: specified size between 18446744056529682436 and
18446744065119617024 exceeds maximum object size 9223372036854775807 [-Werror=stringop-overflow=]
Change-Id: I7814054a102a407d876ffffd14b6ab0be9e222fc
Reviewed-by: Ville Voutilainen <ville.voutilainen@qt.io>
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Various editorial fixes. Also, in 5.9 QStringLiteral does
not fall back to fromUtf8 any longer, but guarantees
a compile-time construction.
Change-Id: Ida4698cf8e32a6e3de97b2c16b997fc9630c9db9
Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
The documentation says that it's equivalent to
qgetenv(varName).toInt()
But the implementation wasn't. QByteArray::toInt() verifies that the
entire string was consumed, so QByteArray("1a").toInt() == 0, but
qstrtoll alone doesn't. That is, qstrtoll("1a", ...) == 1.
The implementation also detected the base, a behavior I kept. Instead, I
updated the documentation.
Change-Id: I0031aa609e714ae983c3fffd14676ea6061a9268
Reviewed-by: Marc Mutz <marc.mutz@kdab.com>
This method now returns -1 by default, due to commit 6255cb893d
which mistakenly replaced -1 with Qt::IgnoreAction (0x0).
As a result, dropping is forbidden in a number of applications
(I detected this in zanshin).
Change-Id: I4922451216e08d5d3fe36f8ba87364a361b691bf
Reviewed-by: Giuseppe D'Angelo <giuseppe.dangelo@kdab.com>
This forces the API not to wait for synchronous I/O from the filesystems
and get the information, but instead just use the cached information
directly. It's a good idea if we have an unresponsive FS, like NFS with
an unreachable server.
Task-number: QTBUG-61096
Change-Id: Iddeeffb6f4ad4a2894a2fffd14c32f6e90664a63
Reviewed-by: Tobias C. Berner
Reviewed-by: Shawn Rutledge <shawn.rutledge@qt.io>
Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
Fixup of commit b4995eb749.
We can't call QWidgetPrivate::get(w) on a deleted QWidget, because of
the call to the member function QWidget::d_func. We can however call
QObjectPrivate::get since we still are in the QObject destructor.
tst_qstackedlayout now pass without ubsan Warnings.
Change-Id: I4e839a97ddbd1cf21435a8fca76523b98a1f7d9b
Reviewed-by: Marc Mutz <marc.mutz@kdab.com>
QList::erase(it) is a linear operation. Calling it in a loop
constitutes quadratic behavior.
Fix by using the erase-remove idiom.
Write a generic function object QtFunctionObjects::IsNotValid, instead
of a lambda, because I am building up a collection of such function
objects in my tree, to be submitted any time now™, so this will reduce
the churn once that gets merged. Wrap it in the unnamed namespace to
avoid ODR violations in case this pattern repeats.
Replace an existing erase-remove idiom's isNotValid lambda function
with the new function object.
Change-Id: I4f0af04e7b201eb168e79beeda332eee3c6f33c3
Reviewed-by: Anton Kudryavtsev <antkudr@mail.ru>
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@qt.io>
This reverts commit ac0184d608. That
commit added the use of qHash in qt_safe_ftok, which made ftok even more
unsafe than previously. Since the algorithm in qHash can change across
Qt versions, we need a stable algorithm instead.
[ChangeLog][Important Behavior Changes] Reverted a change that caused Qt
5.6 through 5.8 not to connect to QSystemSemaphore and QSharedMemory
created by running applications using Qt earlier than
5.6. Unfortunately, this means that Qt 5.9 will not connect to 5.6-5.8.
Task-number: QTBUG-60771
Change-Id: Ibc3472e1c11d46358357fffd14bf51aeb48ef2c8
Reviewed-by: Lars Knoll <lars.knoll@qt.io>
This is one of the main use cases for QDirIterator, but it wasn't
obvious enough that it's possible.
Change-Id: Idae11cfe75dd0e16f1a960bba2470b1695d11241
Reviewed-by: Robin Burchell <robin.burchell@crimson.no>
This commit fixes two bugs:
1) Two ranges should not be merged if they are of different columns.
The old code would have merged (0,0) with (1, 1). Tranforming a selection
of just two indexes in a rectangle of four indexes.
2) The QItemSelectionRange appended had wrong column and worked only for
indexes of the first column. For example if 'tl' was (0, 1) than br was (0, 1)
so the QItemSelectionRange would have be ((0,1), (0, 1-1)) so ((0,1), (0,0)).
This QItemSelectionRange is invalid because topLeft columns is greater than
bottomRight column. The fix take in consideration the bottomRight column.
Task-number: QTBUG-58871
Change-Id: I591ef0bcc63926f24a7b1ced002af9b7737a4b6e
Reviewed-by: Olivier Goffart (Woboq GmbH) <ogoffart@woboq.com>
Reviewed-by: Qt CI Bot <qt_ci_bot@qt-project.org>
... and update the unittest accordingly.
Compared to the 1.8 release there is one change in freedesktop.org.xml,
the magic for application/x-java-keystore was changed from host32 to
big32, as done upstream, see https://bugs.freedesktop.org/show_bug.cgi?id=99328
Task-number: QTBUG-60608
Change-Id: I47de71c9396cfc3eabc884d5679c73a3e4850a17
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
The operator double() and operator long double() members of qfloat16
are causing cast ambiguities. This removes them, leaving only
operator float() which seems to be adequate.
Also, additional arithmetic operator tests were added which without
this removal fail to compile.
Change-Id: Id52a101b318fd754969b3de13c1e528d0aac2387
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
They're deprecated since C++11 and removed in C++17.
Change-Id: Ia2acd9312707bfee96838743645a04ae1780e5dd
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
QLocaleData::unsLongLongToString uses qulltoa, which will allocate a
zero-length QArrayData. Then with padding a single 0 was put in a
QString, which gets prepended to the result. By taking care of this
special case, we can now also fast-path the common case where base=10
and no flags nor precision was provided.
Change-Id: Ia893b0ea4c77634c24e7cef5aafb06d0ef44c507
Reviewed-by: Lars Knoll <lars.knoll@qt.io>
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
The existing QHash::operator== does not work when the same
keys appear in different order between the two hashes being compared.
However, relying on iteration order on a QHash is (as usual) a bad
idea and one should never do it.
Task-number: QTBUG-60395
Change-Id: Ifb39a6779230e26bbd6fdba82ccc0247b9cdc6ed
Reviewed-by: Lars Knoll <lars.knoll@qt.io>
Compilation and link times in CPU seconds with GCC 7, using precompiled
headers (not including moc, rcc, uic, etc. steps or headersclean):
Before After
Debug -O0 198,1 180,3
Debug -Og 240,7 229,2
Release -O3 267,1 249,2
Release LTO 239,4 229,8
QtCore required a little manual adjusting because some files are
bootstrapped into moc itself and into qmake.
Change-Id: I84e363d735b443cb9beefffd14b8b57c10e7da36
Reviewed-by: Lars Knoll <lars.knoll@qt.io>
Reviewed-by: Marc Mutz <marc.mutz@kdab.com>
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@qt.io>
There's no need to derive. This fixes the build with MSVC 2017 under
/permissive-. I don't know what was wrong (ICC, Clang and GCC don't
complain), but it must be related to "Lookup members in dependent base"
in [1].
[1] https://docs.microsoft.com/en-us/cpp/build/reference/permissive-standards-conformance
Change-Id: I9ad33fff8b634979bdbafffd14bb8016f5dc98b3
Reviewed-by: Marc Mutz <marc.mutz@kdab.com>
Reviewed-by: Giuseppe D'Angelo <giuseppe.dangelo@kdab.com>
Change-Id: I26e8c5caca31e842adc7a09151b6de2cc17698ed
Reviewed-by: David Faure <david.faure@kdab.com>
Reviewed-by: Leena Miettinen <riitta-leena.miettinen@qt.io>
We can't depend on QT_HAS_INCLUDE for such an important functionality in
QtQml, so detect at configure time.
alloca() is not a POSIX function (it apparently first appeared in
Version 32V AT&T UNIX), so the actual header that defines it varies from
system to system. Clearly, if alloca.h exists, that's the one, so we try
it first. On most other systems that don't define it, it's in stdlib.h.
The only exception is Windows, where it's actually defined in malloc.h.
Task-number: QTBUG-59700
Started-by: Thiago Macieira <thiago.macieira@intel.com>
Change-Id: Icd0e0d4b27cb4e5eb892fffd14b4b2b389a4684e
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Reviewed-by: Lars Knoll <lars.knoll@qt.io>
Even if a callback type is not automatically re-enabled, callbacks are
implicitly enabled when the source has been added to the run loop.
In this case, calling CFSocketEnableCallBacks() could produce an extra
notification if there is a pending event in the queue.
The bug is quite unstable and completely depends on the internal OS
delays. So, it can't be tested inside Qt.
Task-number: QTBUG-59930
Change-Id: I751b8b8cf99cb86b80055f2214a42a638f01abe4
Reviewed-by: Timur Pocheptsov <timur.pocheptsov@qt.io>
It's an iterator, not a const_iterator. Let QDoc figure out the correct one.
Change-Id: I7ddd1568adbf811b801c170794465ba14ceed05e
Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
There were still two cases where spurious notifications would be
possible:
- user calls hasPendingDatagrams()/pendingDatagramSize() on UDP
socket somewhere outside the slot connected to readyRead()
signal (::WSARecvFrom posts FD_READ notification, even if
a notification for incoming datagram already exists in the
message queue);
- a socket was registered to receive several types of event and
WM_QT_ACTIVATENOTIFIERS message is located between the
different events for this socket in the queue.
Provided patch ensures that the message queue is synchronized with
the Qt event processing mechanism and adds a way to detect spurious
notifications inside the window procedure.
Task-number: QTBUG-58214
Change-Id: I49609dace601f300de09875ff1653617efabd72f
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@qt.io>
Reviewed-by: Lars Knoll <lars.knoll@qt.io>
Reviewed-by: Oliver Wolff <oliver.wolff@qt.io>
Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
Reviewed-by: Peter Seiderer <ps.report@gmx.net>
So that any objects autoreleased during application initialization are
released. Otherwise they will end up in the root level pool and only
be released when the application exits and the application goes out
of scope.
Change-Id: If02d24fd70098f9b4b1b0ea3218e0a15e438b9db
Reviewed-by: Morten Johan Sørvig <morten.sorvig@qt.io>
If the user calls QLocalSocket::setReadBufferSize() with a value less
than the current size of the pipe buffer, startAsyncRead() would call
ReadFileEx() with invalid parameters:
ReadFileEx(handle, nullptr, some_big_value, ...);
Change-Id: I3d153e3ec34f8038dc001c1c896aeceb666a8979
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
... as this would require us to remove usage of QAbstractEventDispatcher::flush
from Qt source code as well. We can not do this already in Qt 5.9 as there are
few event dispatchers in Qt that actually do override the ::flush (with non empty
bodies). Removing this code could result in behavior changes (not very likely) for
some user code. This method will be removed in Qt6.
Instead mark it with a well known "### Qt{Version}" comment.
This patch amends 41eefd7. The warning was:
qcoreapplication.cpp: In static member function ‘static void QCoreApplication::flush()’:
qcoreapplication.cpp:733:48: warning: ‘virtual void QAbstractEventDispatcher::flush()’ is deprecated [-Wdeprecated-declarations]
self->d_func()->eventDispatcher->flush();
Change-Id: I48a1c68b84ff93268956205e1205e6d4b5d48664
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
These are in public classes, so we can't remove them until we get to
Qt 6.
qtimer.h:231:20: warning: private field 'del' is not used [-Wunused-private-field]
Change-Id: I84e363d735b443cb9beefffd14b8b5fb720c9e41
Reviewed-by: Olivier Goffart (Woboq GmbH) <ogoffart@woboq.com>
ICC 17 does not yet implement C++17 [[fallthrough]] nor does it support
GCC's __attribute__((fallthrough))
Change-Id: I84e363d735b443cb9beefffd14b8aba60a7e3f81
Reviewed-by: Giuseppe D'Angelo <giuseppe.dangelo@kdab.com>
We guarded against the Unicode form being invalid and did not produce an
encoded form. But we did not guard against proper Punycode sequences
that decode to forms that had not passed the proper Nameprep stage. So
check for that and, if it fails, just keep the label in the form we
found it in (it's valid STD3 anyway).
[ChangeLog][QtCore][QUrl] Fixed a bug that caused certain domain names
that look like Internationalized Domain Names to become corrupt in
decoded forms of QUrl, notably toString() and toDisplayString().
Task-number: QTBUG-60364
Change-Id: Iadfecb6f28984634979dfffd14b833142cca8d0d
Reviewed-by: Lars Knoll <lars.knoll@qt.io>
The whitelist is kept in ACE form, so if the TLD came in Unicode, we
need to run ToASCII before we can check the whitelist. This is slightly
inefficient because we'll run the same operation later in this domain.
Change-Id: Iadfecb6f28984634979dfffd14b831f37b0f4818
Reviewed-by: Lars Knoll <lars.knoll@qt.io>
Reviewed-by: Florian Bruhin <qt-project.org@the-compiler.org>
... as it has outlived its original purpose:
Qt3 implementation on X11:
void QApplication::flush() { flushX(); }
void QApplication::flushX() { if (appDpy) XFlush( appDpy ); }
Qt4 implementation on X11:
Did nothing when QApplication::flush() was called (the flush()
overrides in {unix,glib} event dispatchers with empty bodies).
In Qt5 this function somehow has been repurposed (inconsistently)
to do what QCoreApplication::sendPostedEvents already does:
QAbstractEventDispatcher::flush() = 0;
=> QCocoaEventDispatcher::flush() {}
=> QEventDispatcherCoreFoundation::flush() {}
=> QIOSEventDispatcher (does not override ::flush())
=> QEventDispatcherGlib::flush() {}
=> QPAEventDispatcherGlib (does not override ::flush())
=> QEventDispatcherUNIX::flush() {}
=> QUnixEventDispatcherQPA (when QT_NO_GLIB=true)
::flush() { if (qApp) qApp->sendPostedEvents(); })
==> QAndroidEventDispatcher (does not override ::flush())
=> QEventDispatcherWin32::flush() {}
=> QOffscreenEventDispatcher::flush() {
if (qApp) qApp->sendPostedEvents();
QEventDispatcherWin32::flush();
}
=> QWindowsGuiEventDispatcher (does not override ::flush())
=> QWindowsDirect2DEventDispatcher (does not override ::flush())
=> QEventDispatcherWinRT::flush() {}
=> QOffscreenEventDispatcher::flush() {
if (qApp) qApp->sendPostedEvents();
QEventDispatcherWinRT::flush();
}
=> QWinRTEventDispatcher (qminimaleglintegration.cpp) (does not override ::flush())
=> QWinRTEventDispatcher (qwinrteventdispatcher.h) (does not override ::flush())
Whatever this function was doing on macOS in Qt3 and Qt4 also has been
dropped in Qt5. It appears that the other event dispatchers in Qt5 that
have overrides for flush() have simply copy-pasted this logic.
Clearly the documentation of QCoreApplication::flush() is outdated and
has nothing to do with the actual implementation in Qt5.
This function is rarely used in Qt5 sources. It should be safe to remove
the calls to QCoreApplication::flush() from Qt source code, as this
function has been doing nothing on most platforms anyways. Repurposing
it even broke handling of posted events (see QTBUG-48717).
[ChangeLog][QtCore][Event loop] QCoreApplication::flush() is now
deprecated. Use QCoreApplication::processEvents() and
QCoreApplication::sendPostedEvents() instead.
Task-number: QTBUG-33489
Task-number: QTBUG-48717
Change-Id: Icc7347ff203024b7153ea74be6bf527dd07ce821
Reviewed-by: Tor Arne Vestbø <tor.arne.vestbo@qt.io>
Reviewed-by: David Faure <david.faure@kdab.com>
Reviewed-by: Gunnar Sletta <gunnar@crimson.no>
This was simply not working for two reasons:
- The index passed to QMetaObject::metacall was not right (there was an offset
because of the return type)
- If the registration succeeded, the arguments were not even initialized.
The tests in tst_moc always called QMetaMethod::parameterType before calling invoke,
which was properly registering the type. So this was not seen in the tests before.
[ChangeLog][QtCore][QMetaMethod] Fixed crash in invoke() with QueuedConnection and
types whose metatype gets automatically registered.
Task-number: QTBUG-60185
Change-Id: I4247628484214fba0a8acc1813ed8f112f59c888
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
qstringbuilder.cpp:75: warning: Command '\li' outside of '\list' and '\table'
Change-Id: I2353462cfd14a4f7cf60d5064ecb069155d1cd34
Reviewed-by: Martin Smith <martin.smith@qt.io>
It only contained a concatenation of the individual rule sets,
probably to fix their order in a central place, as well as
simplifying iteration in defaultCategoryFilter().
Fix these two issues differently, but introducing a RuleSet
enum that lists rule sets in the order in which they should
be applied by defaultCategoryFilter(), and turn individual
rule sets vectors into a C array of vectors.
This enables two nested loops in defaultCategoryFilter to
replace the one loop over 'rules'. Apart from building up
'rules' in updateRules(), this was the only access to that
member. That leaves updateRules() with just the task of
running defaultCategoryFilter() on the new rule sets.
Consequently, a call to updateRules() can now replace the
identical loop in installFilter().
Performance should not suffer. Iterating over a fixed-size
array of vectors is hardly any slower than iterating over
a single vector, and while the construction of 'rules'
was probably a one-off task in most programs, this way
of keeping the rules also saves memory because rules are
not kept in two different vectors.
It is also more maintainable, of course.
Change-Id: Ibc132d096c8137dd02b034752646212e51208637
Reviewed-by: Kai Koehne <kai.koehne@qt.io>
Reviewed-by: Olivier Goffart (Woboq GmbH) <ogoffart@woboq.com>
Since commit bf2160e72c, we can rely on
charNN_t support in all compilers except MSVC 2013, and since that
commit, we use (in 5.10, not 5.9, yet)
!defined(Q_OS_WIN) || defined(Q_COMPILER_UNICODE_STRINGS)
when we only need charNN_t, the type, as opposed to its library
support (u16string, char_traits<char16_t>, ...).
This patch splits the Q_C_UNICODE_STRINGS macro into two, adding
Q_STDLIB_UNICODE_STRINGS for when we need std::uNNstring, leaving
Q_C_UNICODE_STRINGS for when we need just charNN_t support.
In QDebug, when constructing a QChar out of a char16_t, cast to ushort
first, since QChar(char16_t) was only officially introduced in Qt 5.10.
[ChangeLog][Potentially Source-Incompatible Changes] The internal
Q_COMPILER_UNICODE_STRINGS macro is now defined if the compiler
supports charNN_t, even if the standard library does not. To check for
availability of std::uNNstring, use the new Q_STDLIB_UNICODE_STRINGS
macro.
Change-Id: I8f210fd7f1799fe21faf54506475a759b1f76a59
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
That's before the return type or static, inline, constexpr or such
keywords (if any).
Perl Script:
s/^(\s+)(.*) Q_REQUIRED_RESULT(;)?(\s*\/\/.*)?$/\1Q_REQUIRED_RESULT \2\3\4/
Change-Id: I7814054a102a407d876ffffd14b6a16182f159e2
Reviewed-by: Olivier Goffart (Woboq GmbH) <ogoffart@woboq.com>
They're not affected by the GCC bug noted in the comment.
Change-Id: I7814054a102a407d876ffffd14b69e8a8e2527f1
Reviewed-by: Olivier Goffart (Woboq GmbH) <ogoffart@woboq.com>
This reverts commit 47cc9e23a3.
We use QCoreApplication::applicationDirPath in the logging initialization to find
a possible qtlogging.ini file. Because QCoreApplication::applicationDirPath requires
a QCoreApplication instance this leads to a qWarning, which in turn leads to a
recursive call to the logging initialization, and in turn to a recursive mutex deadlock.
Task-number: QTCREATORBUG-18031
Change-Id: Ic75e1e8c062eb647991725378489bf87c9648cca
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@qt.io>
Allows categorized logging before QCoreApplication has been created,
which otherwise would silently fail to output anything because the
category would never be enabled, despite QT_LOGGING_RULES being set.
Change-Id: Ia733105c5b6f28e22af511ced5271e45782da12b
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Initialize a deleter for a new object, created by
QSharedPointer::create(), only after the object is actually
constructed.
[ChangeLog][QtCore][QSharedPointer] Fixed undefined behavior when
creating an object with QSharedPointer::create() and its conscructor
throws an exception.
Task-number: QTBUG-49824
Change-Id: I07f77a78ff468d9b45b8ef133278e8cdd96a0647
Reviewed-by: Olivier Goffart (Woboq GmbH) <ogoffart@woboq.com>
Clang 3.8 has support for __attribute__((target(xxx))) and its SIMD
headers can be included unconditionally.
Change-Id: Ic15b7ff417c8412893e5fffd14b5b42b950b48d7
Reviewed-by: Allan Sandfeld Jensen <allan.jensen@qt.io>
_mm_cvtps_ph is usually defined as a macro:
qfloat16.h:122:37: error: use of old-style cast [-Werror=old-style-cast]
Change-Id: Icd0e0d4b27cb4e5eb892fffd14b516ec47826c0c
Reviewed-by: Ville Voutilainen <ville.voutilainen@qt.io>
Reviewed-by: Marc Mutz <marc.mutz@kdab.com>
Seems like an obvious omission.
[ChangeLog][QtCore][QVariant] Can now convert QUuid to and from
QByteArray, not just QString.
Change-Id: Ib56ae86ca0c27adaf1e095b6b85e64fe64ea8d18
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
- Mention you can build QByteArrays, too
- Nicer list of types that can be used, separate for QByteArray and
QString
Change-Id: Ia91445f0cb4872bab12a55f4812c283e9c38dba4
Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
Reviewed-by: Olivier Goffart (Woboq GmbH) <ogoffart@woboq.com>
QSysInfo::productType() returned "osx" for all versions of macOS, even
10.12. Change 3e2bde3578 was incorrect.
[ChangeLog][Important Behavior Changes] QSysInfo::productType() and
QFileSelector behavior on macOS was restored to match what Qt used to
return in version 5.7.0 and earlier. The behavior found in Qt 5.6.2,
5.7.1 and 5.8.0 is removed.
[ChangeLog][Future Compatibility Notice] The identifiers that
QSysInfo::productType() and QFileSelector will use to identify macOS
systems will change in Qt 6.0 to match the Apple naming guidelines which
will be current then.
Task-number: QTBUG-59849
Change-Id: Ib0e40a7a3ebc44329f23fffd14b2b39392210c4f
Reviewed-by: Jake Petroules <jake.petroules@qt.io>
Turns out that different versions of ICC use different warning numbers.
The Linux and Windows compilers emit 1786, but the macOS one emits 1478.
Don't ask me why.
Change-Id: I523b0abacd5148b2bf08fffd14b475a4c4d89ba1
Reviewed-by: Jake Petroules <jake.petroules@qt.io>
This fixes a regression introduced in ff19ebcc
Task-number: QTBUG-60046
Change-Id: I47c357433b25f07011a7a3a64d3150591785b206
Reviewed-by: Ulf Hermann <ulf.hermann@qt.io>
POSIX documents that localtime() ensures that tzset() has been called,
but the wording could be understood to mean that it only needs to do so
the first time. Anyway, we're sure that the MS runtime only gets the
timezone information from the Control Panel once. That means Qt-based
applications will not react to a change in the timezone.
Attempt to do that by moving tzset() out of the #if, to apply to all
operating systems.
Task-number: QTBUG-60043
Change-Id: I6ab535fb61094af19fc1fffd14b413541fe5a64c
Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
The developer should see what's wrong even on release builds.
That code hides any mistakes we do in JNI which are pretty critical
for developers because they can't see what's wrong with their code. e.g.
QtAndroid::activity().callMethod<void>("wrongMethodName")
*silently* fails, which is so wrong!
Change-Id: I8b6a24946dfef716fcd86ab9bba82666974e3991
Reviewed-by: Eskil Abrahamsen Blomfeldt <eskil.abrahamsen-blomfeldt@qt.io>
Factor the line parsing into a separate function, parseNextLine(),
taking a QStringRef.
In setContent(QTextStream&), use the new readLineInto() function to
re-use the capacity of a single QString for all lines.
In setContent(QString), use splitRef() to split the lines.
In either function, pass each line to parseNextLine().
In order to port all the parsing to QStringRef, I needed to make some
semantic changes: the old code removed all whitespace right at the
beginning. This is not possible with QStringRef. It also didn't feel
right, since a line like
[ r u l e s ]
would successfully parse as the section named "rules".
I added trimmed() calls at the beginning, and around the valueStr and
pattern extraction, which should be good enough.
Also, when a section is found, don't store it anymore. Instead, only
store whether it was the [rules] section, because that's all we'll
test for. That way, we don't have to convert QStringRefs to QString
just to store them across parseNextLine() calls.
Replace the setSection() function with setImplicitRulesSection(),
because "rules" is all that was ever passed.
This is private API, we can bring back some of the dropped flexibility
later, as needed.
[ChangeLog][Important Behavior Changes] Logging rules can no longer
contain arbitrary whitespace such as within a category identifier.
Change-Id: Ic26cd23c71f5c810b37ef4b972354ac31d3408fe
Reviewed-by: Kai Koehne <kai.koehne@qt.io>
- toUuid(): QUuid is a built-in type, so use type(), not userType()
- canConvert()/toUuid(): QUuid converts to and from QString
Change-Id: I5262ff7ab093040cb943b6ab9cfffe95491d2b9b
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
The end() pointer, like in all other containers, is a sentinel value
that must never be dereferenced. But unlike array-based containers,
end() in QMap is not "last element plus one", but points to a base class
of Node, not a full Node.
Therefore, the casting from QMapNodeBase to QMapNode must not be a
static_cast, reinterpret_cast is required.
libstdc++-v3's red-black tree had the exact same problem:
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=60734
Change-Id: I43f05fedf0b44314a2dafffd14b33697861ae589
Reviewed-by: Marc Mutz <marc.mutz@kdab.com>
It is incorrect to collapse a "symlink/.." segment because the parent
directory of the symlink's target may not be the directory where the
symlink itself is located.
[ChangeLog][QtCore][QDir] Fixed a bug that caused QDir::mkpath() to
create the wrong directory if the requested path contained a symbolic
link and "../".
Change-Id: Iaddbecfbba5441c8b2e4fffd14a3e367730a1e24
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@qt.io>
Reviewed-by: David Faure <david.faure@kdab.com>
std::nullptr_t is nullary: it accepts only one value, nullptr. So we
don't need to read or write anything. This commit simply adds the two
operators that allow generic code to operate on std::nullptr_t if
required.
This commit also adds the actual use to QMetaType::load/save, even
though there's no change in behavior.
[ChangeLog][QtCore][QDataStream] Added operator<< and operator>>
overloads that take std::nullptr_t, to facilitate generic code.
Change-Id: Iae839f6a131a4f0784bffffd14aa37e7f62d2740
Reviewed-by: Marc Mutz <marc.mutz@kdab.com>
Reviewed-by: Lars Knoll <lars.knoll@qt.io>
Add a small table to illustrate the results exactMatch() and split
out the part on partial matching to a separate section since it
is less common.
Change-Id: Ifbd5c3cbd1d8c0ee9e8b2d58ed13f40776b03762
Reviewed-by: Giuseppe D'Angelo <giuseppe.dangelo@kdab.com>
Reviewed-by: Leena Miettinen <riitta-leena.miettinen@qt.io>
Neither the Intel compiler nor Visual C++ have a dedicated switch to
enable F16C support, like GCC and Clang do. So we used the AVX switch
for that in commit 8241d51f70, as it was
the closest, lowest denominator. That was incorrect and insufficient.
The Intel compiler silently miscompiles the intrinsics with -xAVX,
making calls to out-of-line functions like _mm_cvtps_ph, which don't
exist. So we actually have to use AVX2 support to generate correct code.
That might be a problem later, since Ivy Bridge supports F16C but not
AVX2.
Visual C++ is able to generate F16C code with just -arch:AVX.
Either way, since there's no dedicated command-line switch, there's also
no dedicated preprocessor macro. We're using __AVX2__ for both
compilers, as that's a sufficient condition to indicate a processor that
supports F16C.
Change-Id: I27b55fdf514247549455fffd14b205b8d8b86da7
Reviewed-by: Allan Sandfeld Jensen <allan.jensen@qt.io>
MSVC, Apple's Clang and Clang prior to 3.9 do not recognize _cvtss_sh
and _cvtsh_ss. So expand the operation to use directly the packed
intrinsics.
Change-Id: I27b55fdf514247549455fffd14b2046fd638593d
Reviewed-by: Allan Sandfeld Jensen <allan.jensen@qt.io>
qsysinfo.h(235): error #1786: enum "QSysInfo::MacVersion" (declared at line 156) was declared deprecated ("Use QOperatingSystemVersion")
Take this opportunity to merge the two groups.
Change-Id: I27b55fdf514247549455fffd14b1c2a1d8eab869
Reviewed-by: Jake Petroules <jake.petroules@qt.io>
Aligns with EPERM (Operation not permitted) and ENOENT (No such file
or directory), and is what errno is set to on macOS when opening the
/dev/tty device when running inside e.g. Xcode, where isatty() will
return true.
Change-Id: I09b88eaa3ff611d95ab37f0ff4df9aaaca52747d
Reviewed-by: Simon Hausmann <simon.hausmann@qt.io>
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
More fallthrough-are-errors fixed.
Change-Id: I9a6cb6efe988400ed3f9cb95d1e426dac317e6c4
Reviewed-by: Allan Sandfeld Jensen <allan.jensen@qt.io>
Reviewed-by: Marc Mutz <marc.mutz@kdab.com>
When we call realloc, the alignment of the new block may be different
from the old one. When that happens, we need to memmove the data to the
new position, before we start overwriting things.
Task-number: QTBUG-59804
Change-Id: I27b55fdf514247549455fffd14b07ea78918a3d0
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@qt.io>
GCC 7 defines __has_cpp_attribute even when invoked as "gcc" (possibly,
Clang does the same, according to a comment in the code, did not test
myself).
Hence, define the fallthrough declaration (as C++11 attributes)
only when compiling as C++, otherwise we pick them up even in C mode,
and they cause build failures.
Change-Id: I3f13205e014bb1dea59ee3664b29111521a7eae3
Reviewed-by: Allan Sandfeld Jensen <allan.jensen@qt.io>
Clang implements the _cvtss_sh intrinsic by way of a macro, which
uses a C99 extension and that's not allowed in C++ mode:
float16.h:119:11: error: compound literals are a C99-specific feature [-Werror,-Wc99-extensions]
/usr/bin/../lib64/clang/3.9.1/include/f16cintrin.h:76:55: note: expanded from macro '_cvtss_sh'
Reported at https://bugs.llvm.org/show_bug.cgi?id=32491.
Change-Id: I27b55fdf514247549455fffd14b170df75dd4e1f
Reviewed-by: Marc Mutz <marc.mutz@kdab.com>
Some compilers are known to complain about this with a warning. GCC
complains about const on return values on -Wignored-qualifiers (enabled
at -Wextra), so it's not too much of a jump to assume that others do
too. Besides, this is not Qt Library API policy. As maintainer for
QtCore, I'm exercising my prerrogative in specifying certain unspecified
parts of the coding style, like I've done for constructor initializer
lists.
Since all the classes involved are exported (including QVector, through
derived classes), we can't remove the qualifier until Qt 6, since there
are compilers known to encode the qualifier in the mangled name
(suncc). I'm not introducing #ifdef to silence unknown compilers unless
we get an actual complaint.
Change-Id: I33850dcdb2ce4a47878efffd14a876edef843c46
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
The SHA3 family is a modified version of Keccak. We were
incorrectly calculating Keccak (and even *testing* Keccak!),
but claiming it was SHA3.
To actually calculate SHA3, we need invoke Keccak on the original
message followed by the two bits sequence 0b01, cf. §6.1 [1].
[1] http://dx.doi.org/10.6028/NIST.FIPS.202
[ChangeLog][QtCore][QCryptographicHash] QCryptographicHash now
properly calculates SHA3 message digests. Before, when asked
to calculate a SHA3 digest, it calculated a Keccak digest instead.
Task-number: QTBUG-59770
Change-Id: Iae694d1a1668aa676922e3e00a292cddc30d3e0d
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
