| Commit message (Collapse) | Author | Age | Files | Lines |
... | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
As a side effect, QString::simplified() will always return a detached
copy, even if it's the same contents.
QStringRef::trimmed() can use the same calculation algorithm but can't
use the trimmed_helper() template function due to its lack of a
constructor taking begin pointer and size (std::string_view could do
it). That constructor can't be added because QStringRef always refers to
an existing QString, not to data in memory.
Change-Id: Ib966c1741819c68c6bac5fcbb00f8ac818b3ccab
Reviewed-by: Olivier Goffart <ogoffart@woboq.com>
Reviewed-by: Jędrzej Nowacki <jedrzej.nowacki@digia.com>
Reviewed-by: Marc Mutz <marc.mutz@kdab.com>
|
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
This is even more common than the QByteArray equivalents.
Qt Qt Creator
const & && const & &&
toLower 71 50 45 26
toUpper 35 8 46 35
Change-Id: I8b797d2321b22ce414c23656c5f1709ac649c423
Reviewed-by: Marc Mutz <marc.mutz@kdab.com>
|
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
This unifies the code for those three functions in one refactored
template function, using QStringIterator. I don't think there's any loss
of performance by doing that refactoring -- this is based on my reading
of the disassembly, without running any benchmarks.
Change-Id: I5893c6ed47462c473886c722a21577b1e8a23841
Reviewed-by: Marc Mutz <marc.mutz@kdab.com>
|
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
Those operations aren't very common with QByteArray but this is easy to
optimize.
Qt Qt Creator
const & && const & &&
toLower 34 10 0 1
toUpper 3 1 0 0
Change-Id: I2097955f4c889ea5a21903c35ddbc0ff27bf62c5
Reviewed-by: Marc Mutz <marc.mutz@kdab.com>
Reviewed-by: Olivier Goffart <ogoffart@woboq.com>
|
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
The latin1 conversion tables that were recently introduced to qbytearray.cpp
to speed up QByteArray::to{Upper,Lower}() can also be used in qstr(n)icmp.
This results in speedups between ~2 for strings with differences in the
first char, to almost 10x for strings with differences only after 8k of
data.
Change-Id: I878ddb4c01c798069d439a9d33e24351fe1039d3
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Reviewed-by: Olivier Goffart <ogoffart@woboq.com>
|
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
Do a check first if we need to transform before doing the transform.
This means we won't detach when transforming data that is already
correct.
And instead of using QChar, use our own hand-rolled table. In a proper
LTO build, the QChar calls would be resolved to a lookup of the Unicode
data, but not many people do LTO builds, Therefore, this means a great
speed-up is achieved by simply avoiding the function call. The extra
gain in performance comes from the simpler translation table instead of
the more complex full-Unicode data.
Also as a consequence, this changes the handling of two characters in
Latin 1: 'ß' should be uppercased to "SS" but we won't do it, and 'ÿ'
can't be uppercased in Latin 1 ('Ÿ' is outside the range).
Benchmarking is included. Comparing the Qt 5.4 algorithm to the new code
is almost 20x faster. Other alternatives are included in the benchmark
and are all faster than the current code, though slower than the new
one. While all of them could compress the tables to be smaller or shared
between uppercasing and lowercasing, they would also expand to more code
(though probably less than the extra bytes required in the full
translation table). In the trade-off, I decided to go with simplicity
and most efficient code.
Change-Id: I002d98318d236de0d27ffbea39d662cbed359985
Reviewed-by: Marc Mutz <marc.mutz@kdab.com>
|
|\| | |
| | | |
| | | |
| | | | |
Change-Id: Ia03a010d1fbfc89c04144c1c106336f0e1c45ec1
|
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
Not all Android versions appear to have f_flags in statfs, but we
can check for the presence using a feature define.
Change-Id: Ib2acf6063d6211b871f462ee491459ac8675aa37
Reviewed-by: Ivan Komissarov <ABBAPOH@gmail.com>
Reviewed-by: Eskil Abrahamsen Blomfeldt <eskil.abrahamsen-blomfeldt@digia.com>
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
|
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
QMimeDataStruct is larger than a pointer, so holding it in a
QList is horribly inefficient.
Fix by marking as movable and holding in a QVector instead.
Change-Id: If285dd31546066db0b240ea0d4d30668f50b5f2c
Reviewed-by: Jędrzej Nowacki <jedrzej.nowacki@digia.com>
Reviewed-by: Olivier Goffart <ogoffart@woboq.com>
|
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
The code contained old Qt4 logic. There is no point in checking values
of QVariant as they are in sync with QMetaType.
Change-Id: I58ace52f69939488e6ea7e7bab98df38419420a8
Reviewed-by: Olivier Goffart <ogoffart@woboq.com>
|
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
An event dispatcher fetching the core dispatcher can deadlock when
looking up the window. Simplify this by only looking up the view, and
treating the event dispatcher as one which will never obtain the core
dispatcher. This fixes a deadlock which occurs when starting network
delegates.
Change-Id: Ic27f4ab6f168eb95f3de70d78d64f035c331a7f5
Reviewed-by: Oliver Wolff <oliver.wolff@digia.com>
|
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
Allows to retrieve information about mounted volumes such as label,
total/available size, filesystem type and so on.
Possible use cases are:
- allows to do checks about filesystem before performing actual
operation (such as available/maximum volume size)
- allows to retrive information about volume that can be shown in file
dialogs
- allows to retrieve volume for specific path and check if two or more
paths belong to the same volume or not
[ChangeLog][QtCore] Added QStorageInfo class to retrive information
about mounted volumes and drives
Change-Id: Ibf9c2e6b53ef39c5605894a4422acdbbca4030c4
Reviewed-by: Shawn Rutledge <shawn.rutledge@digia.com>
|
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
WinRT stores settings inside the app bundle, not in the registry.
Some tests are not fully functional due to errors in the file
implementation (See QTBUG-40588).
QSettings::SystemScope translates to a roaming container on WinRT,
meaning that settings stored inside there should be uploaded and
shared among devices. However, this is untested so far.
The tests have been updated for those platforms which do not store
the order of keys. This has been done on some locations only so
far, but needed to be done on more places for WinRT.
Task-number: QTBUG-33498
Change-Id: Ifd0194387b09c220d31812b4b6fd0ce9a7d84d24
Reviewed-by: Andrew Knight <andrew.knight@digia.com>
|
| |/ /
| | |
| | |
| | |
| | |
| | |
| | | |
One can't dereference the null pointer value. Even to access a static member.
Change-Id: I5edc92d7886ccc7029d1f3c00a12de78b439f080
Reviewed-by: Jędrzej Nowacki <jedrzej.nowacki@digia.com>
|
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
The core dispatcher lookup should be performed from the constructor
in all cases but the first (WinMain) thread, and only rechecked if the
thread changes.
Change-Id: I05f0c15b3e199994aa5d740b2092b42fb8d2f596
Reviewed-by: Oliver Wolff <oliver.wolff@digia.com>
|
| | |
| | |
| | |
| | |
| | |
| | | |
Change-Id: I8fd9ec2ea1138a94478b73a284ea1b550b62bd5a
Reviewed-by: Sergio Ahumada <sahumada@blackberry.com>
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
|
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
Replacing usage QString::split by QString::splitRef saves a few
allocations.
Change-Id: I1cadca296279248b75af6f9f8394c54f13c37c55
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@digia.com>
|
|/ /
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
The automatic registration code that would be added will break
with a reference type.
QObject's template code calls QMetaTypeId2 on the arguments of the
signal. Including on references.
Reference type cannot be register as metatype. They are not supported
in QVariant, or in QueuedConnection.
Change-Id: I83f2e50f9f886909dc24f3809c990e5763012cc2
Reviewed-by: Jędrzej Nowacki <jedrzej.nowacki@digia.com>
|
| |
| |
| |
| |
| |
| |
| |
| | |
Using QStringRef saves few instructions
Change-Id: I0d67a1f6a1c1385c88a8b21cb98b9c7dbdfb66e5
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@digia.com>
Reviewed-by: Jędrzej Nowacki <jedrzej.nowacki@digia.com>
|
| |
| |
| |
| |
| |
| |
| |
| | |
Use QString::splitRef on platforms with case sensitive paths, to
avoid unnecessary allocations.
Change-Id: Iff331a5eb67c4cbaf1333e663e77290261c5b013
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@digia.com>
|
| |
| |
| |
| |
| |
| |
| | |
And correct the comment by listing more types that are available
Change-Id: I9cb4b664f97300357a55d81bc99dd542a29e933b
Reviewed-by: Marc Mutz <marc.mutz@kdab.com>
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
On templates, adding constexpr makes a given instantiation constexpr
if it can be.
This turns qMakePair(0,0), say, into a compile-time constant.
The effects on existing code are small, but exist:
$ size lib/*{-baseline,-paircexp} | sort -nr
6516727 211192 2608 6730527 66b31f lib/libQt5Widgets.so.5.4.0-baseline
6516711 211192 2608 6730511 66b30f lib/libQt5Widgets.so.5.4.0-paircexp
5373720 44492 15976 5434188 52eb4c lib/libQt5Core.so.5.4.0-baseline
5373504 44492 15976 5433972 52ea74 lib/libQt5Core.so.5.4.0-paircexp
5107206 125072 6080 5238358 4fee56 lib/libQt5Gui.so.5.4.0-baseline
5107030 125072 6080 5238182 4feda6 lib/libQt5Gui.so.5.4.0-paircexp
1341290 30180 2600 1374070 14f776 lib/libQt5Network.so.5.4.0-baseline
1341210 30180 2600 1373990 14f726 lib/libQt5Network.so.5.4.0-paircexp
# no other libraries benefit
[ChangeLog][QtCore][QPair] Can now be used in C++11 constexpr contexts.
Change-Id: I3872e6aa33a7d02a168516f4dfa7119efcac8c40
Reviewed-by: Olivier Goffart <ogoffart@woboq.com>
Reviewed-by: Jędrzej Nowacki <jedrzej.nowacki@digia.com>
|
| |
| |
| |
| |
| |
| |
| |
| |
| | |
{,c}{begin,end}() were added post-5.3, in 1a6f490b, but weren't marked with \since 5.4.
Add it.
Change-Id: Ide743833144f784c7d09b125e7a22f9b184ed823
Reviewed-by: Jędrzej Nowacki <jedrzej.nowacki@digia.com>
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
and make QByteArrayList be a simple typedef.
As a side-effect, the constructor taking a QByteArray is no longer
available since I couldn't find a way to add it to QList<T> when T is
QByteArray. My template-foo failed me. I tried:
- QEnableIf<is_same<T, QByteArray>::value, QByteArray>::type
=> makes QList fail to compile for any T that isn't QByteArray
- make the constructor a template member
=> it compiles if the parameter is a QByteArray, but not a const
char[4] like the test was
- inheriting constructors
=> runs into ICC and Clang bugs that I could not work around
Besides, the constructor with std::initializer_list is a superior
solution anyway.
Change-Id: Ic86fbadc1104142bfd907a5c4147199bf839fb89
Reviewed-by: Marc Mutz <marc.mutz@kdab.com>
Reviewed-by: Jędrzej Nowacki <jedrzej.nowacki@digia.com>
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
Initial submission of a new class QByteArrayList with the
purpose of aggregating and then joining QByteArray instances.
[ChangeLog][QtCore] Added new QByteArrayList class.
Done-with: Marc Mutz <marc.mutz@kdab.com>
Change-Id: I503af58f125d7f44fef10360177490c933e5840f
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
The class provides compare operators, stream operators, and hashing
functions. This class aims to be compatible with (but not restricted to)
the Semantic Versioning 2.0 standard (semver.org).
[ChangeLog][QtCore] Added QVersionNumber class
Done-with: Marc Mutz <marc.mutz@kdab.com>
Change-Id: I244c8ccc002909af03987a2df052734d1a8621a9
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@digia.com>
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
QObjectPrivate::Connection::method should never be called when
the Connection was made with the function pointer syntax
This caused valgrind warning about using uninitialized value on such code:
QObject::connect(&o, &Object::aSignal, &o, &Object::aSlot);
o.disconnect(&o, SLOT(aSlot()));
Change-Id: Iaff9ecd3ddfe665db92726b420021493453c4cea
Reviewed-by: Giuseppe D'Angelo <giuseppe.dangelo@kdab.com>
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
|
| |
| |
| |
| |
| |
| |
| |
| |
| | |
These will be needed in some template code that is to come.
Change-Id: I5b93f4320313f7b15a6404de2c98f85485735fda
Reviewed-by: Olivier Goffart <ogoffart@woboq.com>
Reviewed-by: Marc Mutz <marc.mutz@kdab.com>
Reviewed-by: Jędrzej Nowacki <jedrzej.nowacki@digia.com>
|
| |
| |
| |
| |
| |
| |
| | |
Lots of code depended on an indirect includes from qstringlist.h.
Change-Id: I33d0dce33d64302d6c0e49180cc1249b90ab27c5
Reviewed-by: Marc Mutz <marc.mutz@kdab.com>
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
If the calling thread changes when processing events, the dispatcher
will no longer have thread access and event processing will fail. This
can e.g. prevent new threads from being created. To remedy this, the
dispatcher object is re-fetched if the thread is changed.
Change-Id: I519cff521f9b84211db3f28a7a28b532de44a6a4
Reviewed-by: Oliver Wolff <oliver.wolff@digia.com>
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
The goal of this commit to make the code in the test work:
QCommandLineParser parser;
parser.addOptions({
{ "a", "The A option." },
{ { "v", "verbose" }, "The verbose option." },
{ { "i", "infile" }, "The input file.", "value" },
});
For this, QCommandLineParser needs a version of addOption that can
take a list of options. That's what addOptions() is for.
More importantly, the QCommandLineOption ctors mustn't be explicit.
OTOH, any implicit conversion from QString or QStringList to
QCommandLineOption is also undesirable.
To solve this dilemma, add new QCommandLineOption ctors that just
take one argument and are explicit, and make the existing ctors
implicit. In order to avoid ambiguities, remove the default values
of their resp. 2nd arguments. The new ctors are by intention not
\since 5.4, as they are completely transparent to the user.
Et voila, even better than getopt_long(3).
[ChangeLog][QtCore][QCommandLineParser] Added addOptions() method.
Change-Id: I5e779f3406cd0f6c8ec6ecbf6c8074af226de300
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
|
| |
| |
| |
| |
| |
| |
| | |
[ChangeLog][QtCore] Added the QStringRef::split() function
Change-Id: I28709c9761785dea7be4e7d621ecf4e1ae007a72
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
|
| |
| |
| |
| |
| |
| |
| |
| | |
This code hasn't been tested for at least 4 years. It's not maintained
and probably doesn't work.
Change-Id: I4b9a5179e34111b400914f91caa6b741b69771bb
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@digia.com>
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
Properly capitalize the names UTF-16 and UCS-4 and make sure we talk
about UTF-16 and not UCS-2. UCS-2 is not the same and does not support
surrogate pairs.
Task-number: QTBUG-35287
Change-Id: If33270996bacc9ae5d04c87423fa1ee9ddaff230
Reviewed-by: Martin Smith <martin.smith@digia.com>
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
The #undef in qcompilerdetection.h was missing.
And apparently we can detect Neon since Windows Mobile 6 too.
Change-Id: I38a5f71b2704a29a706183e39f43db3a78a729db
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@digia.com>
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@digia.com>
Reviewed-by: Lars Knoll <lars.knoll@digia.com>
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
Background:
It is often necessary/advisable to schedule tasks on thread pools !=
globalInstance(). As Herb Sutter writes in
http://www.drdobbs.com/parallel/use-thread-pools-correctly-keep-tasks-sh/216500409
and the Qt Training Material stresses, tasks you schedule on a (global)
thread pool should be non-blocking, which currently rules out using any of
the QtConcurrent functions for, say, file I/O.
Nonetheless it's often convenient to have thread pools also for file I/O, as
the thumbnail viewer exercise in the Qt Training Material shows. In this
case, you'd use a dedicated thead pool, leaving the global thread pool for
CPU-bound tasks.
Yet, none of the QtConcurrent functions allow to pick the QThreadPool
instance on which to schedule the work created with them.
This patch prepares for them to do so.
This is the first part of the forward-port of
https://qt.gitorious.org/qt/qt/merge_requests/1281.
Implement by using a new QThreadPool* member that defaults to nullptr,
and adding setThreadPool to set this member, then using it in lieu of
QThreadPool::globalInstance() everywhere.
I chose to leave m_pool == nullptr to mean globalInstance() to avoid
creating the global instance whenever a QFuture is created, even if the
future represents the result of a calculation not run on the global thread
pool.
[ChangeLog][QtCore][QFuture] Can now be used with any QThreadPool, not
just globalInstance().
Task-number: QTBUG-17220
Change-Id: I4e1dc18d55cf60141b2fa3d14e2d44a3e9e74858
Reviewed-by: Olivier Goffart <ogoffart@woboq.com>
|
| |
| |
| |
| |
| |
| | |
Change-Id: Ia591805c82e7bf77dc7af04f1054c7daca58eb6a
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Reviewed-by: Olivier Goffart <ogoffart@woboq.com>
|
|\ \ |
|
| |\|
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
Conflicts:
config.tests/arch/arch.cpp
src/opengl/qgl_qpa.cpp
src/widgets/kernel/qapplication.cpp
Change-Id: I80b442a4c2c9632743a5b5c7319ff201ec5bc4fd
|
| | |\
| | | |
| | | |
| | | | |
Change-Id: I2a044d44ca991ba20ddd710053b85afb51e362d3
|
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
The atomic functions on MIPS are based on the sync opcode with an
immediate argument, which is something introduced in the MIPS32
instruction set. This prevent to use Qt on pre-MIPS32 CPU, like the
Loongson 2 CPU.
However some of the pre-MIPS32 CPUs interprets the sync opcode with and
immediate argument as a sync opcode without argument (which is a stronger
ordering than with the argument), and for the others the kernel emulates
it.
It is therefore fine to use the current MIPS atomic functions on
pre-MIPS32 CPU. This patch allows that by temporarily changing the
instruction set to MIPS32 around the sync instruction, so that binutils
doesn't choke on it.
Change-Id: I9cc984bd55b5f172736ce9e638a6f4e271b79fe7
Reviewed-by: Lisandro Damián Nicanor Pérez Meyer <perezmeyer@gmail.com>
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
|
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
It has been working in Debian for some time.
It also adds detection for 64bits Sparc.
Change-Id: Ie4fc0f58b37672b79191ebe51de0caf2eaf8a1d9
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
|
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
When compiling the innerFunction() of the Q_GLOBAL_STATIC expansion, ICC
emits global symbols for the internal "holder" local static variable and
its guard. If there are two global statics of the same name in two
different .cpp files, the linker will incorrectly merge the two "holder"
variables.
This was noted between the "customTypes" global statics of qmetatype.cpp
and qdbusmetatype.cpp in a static build.
The C++ standard requires that local static variables declared in inline
functions must be the same, regardless of whether the body of the
function got inlined or not. The IA-64 C++ ABI does that by requiring
local static symbols for inline functions to be global and mergeable
("link once"). However, two functions in anonymous namespaces in
different files are not considered to be the same function, so their
local statics should not be merged. This is where ICC failed: the local
statics are global and mergeable, even though the function is in an
anonymous namespace. ICC correctly emits the function itself as a local
symbol.
Alternative solutions were:
1) add "static", but you can't use a static symbol in a template
parameter in C++98 mode
2) remove the "inline" keyword, but then GCC 4.8 will not inline
Intel issue ID: 6000058488
Task-number: QTBUG-40053
Change-Id: I307622222499682dde711b2771c8cf7557400799
Reviewed-by: Jędrzej Nowacki <jedrzej.nowacki@digia.com>
|
| | | |
| | | |
| | | |
| | | |
| | | | |
Change-Id: Ic91b9a74d0c28ec225d2708a13aafddd1e2c32af
Reviewed-by: Laszlo Agocs <laszlo.agocs@digia.com>
|
|/ / /
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
qlist.h(133) : warning C4267: 'argument' : conversion from 'size_t' to 'int', possible loss of data
qlist.h(131) : while compiling class template member function 'QList<QString>::QList(std::initializer_list<T>)'
Change-Id: I8a11e298cd10da199490fbd8b269405a9e1cf5f3
Reviewed-by: Olivier Goffart <ogoffart@woboq.com>
|
| | |
| | |
| | |
| | |
| | | |
Change-Id: Idc46a347009556d06db4bc1f7ce3e2613fb1a405
Reviewed-by: Olivier Goffart <ogoffart@woboq.com>
|
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
We currently do not receive events from most multimedia keys on Windows
because they are sent using appcommand events instead of the normal key
events.
This patchs adds support for parsing appcommand events.
Change-Id: I19f568b6fa0d3d136e3a22318df1499a1fab6edb
Reviewed-by: Oliver Wolff <oliver.wolff@digia.com>
|
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
QOpenGLWidget is now public.
In addition Qt::WA_AlwaysStackOnTop is introduced to support the
special case of semi-transparent QOpenGLWidget or QQuickWidget on
top of regular widgets.
hellogl_es2 becomes the qopenglwidget example. This example performs
painting both via QPainter and native GL commands and has the OpenGL
widget combined with other, normal widgets.
The widget stack receives some changes when it comes to renderToTexture
widgets like QQuickWidget and QOpenGLWidget. Calling update() will now
result in a paint event, which is essential for QOpenGLWidget since we
want it to behave like a regular widget. The dirty region handling is
extended specially for such widgets due to performance reasons.
(an OpenGL content update must not result in any backingstore painting,
and is thus handled as a different kind of dirtiness)
[ChangeLog] Added QOpenGLWidget. This widget serves as a replacement for QGLWidget.
Task-number: QTBUG-36899
Task-number: QTBUG-40086
Change-Id: Ibf7f82fea99b39edfffd2fc088e7e0eadbca25cf
Reviewed-by: Paul Olav Tvete <paul.tvete@digia.com>
|
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
This function is used in the named timezone ctor and was using QByteArray::split(),
followed by size checks and a linear scan for invalid chars per section. The use of
split() resulted in a lot of memory allocations and, unsurprisingly, bad performance.
The new code just performs one linear scan through the byte array, calculating
section sizes on the fly.
Benchmark results (with the test data in tst_QTimeZone::isValidId_data()) show
typical speedups of ~10x for valid IDs:
RESULT : tst_QTimeZone::isValidId_bench():"minimal middle":
- 0.00036 msecs per iteration (total: 95, iterations: 262144)
+ 0.000035 msecs per iteration (total: 74, iterations: 2097152)
Even in the sweet-spot case of the old code---a space character anywhere in the
string, checked for before the split---the new code is anywhere between slightly
faster and not much slower:
RESULT : tst_QTimeZone::isValidId_bench():"invalid char ' ' front":
- 0.000011 msecs per iteration (total: 94, iterations: 8388608)
+ 0.000010 msecs per iteration (total: 86, iterations: 8388608)
RESULT : tst_QTimeZone::isValidId_bench():"invalid char ' ' middle":
- 0.000014 msecs per iteration (total: 62, iterations: 4194304)
+ 0.000016 msecs per iteration (total: 69, iterations: 4194304)
RESULT : tst_QTimeZone::isValidId_bench():"invalid char ' ' back":
- 0.000018 msecs per iteration (total: 79, iterations: 4194304)
+ 0.000023 msecs per iteration (total: 98, iterations: 4194304)
This is not surprising, as the space character was singled out for a fast-exit
check before. For any other invalid character, the new version is anywhere from
15x to 35x faster:
RESULT : tst_QTimeZone::isValidId_bench():"invalid char ? front":
- 0.00034 msecs per iteration (total: 91, iterations: 262144)
+ 0.000010 msecs per iteration (total: 87, iterations: 8388608)
RESULT : tst_QTimeZone::isValidId_bench():"invalid char ? middle":
- 0.00036 msecs per iteration (total: 96, iterations: 262144)
+ 0.000016 msecs per iteration (total: 68, iterations: 4194304)
RESULT : tst_QTimeZone::isValidId_bench():"invalid char ? back":
- 0.00035 msecs per iteration (total: 94, iterations: 262144)
+ 0.000021 msecs per iteration (total: 92, iterations: 4194304)
If there was a deeper reason to single out the space character, that fast-exit
path can easily be restored.
This function is often used in conjunction with availableTimeZoneIds(), which
currently vastly dominates the runtime of the function calling both, but I'll
add another optimization for the common use-case of just checking for a time-zone's
existence in a subsequent commit.
Change-Id: Ife1d096fcd39464083ea464c23e49ad98fabf345
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
|
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
The main event loop will be paused when an application is suspended,
this is also the normal behavior of any Android application. When an
application is suspended on Android all its Gl surfaces are destroyed
and can't be used to render anymore. So, we need to pause the main
event loop in order to pause all the timers which might trigger
drawings. The event loop is resumed immediately after the application
is foreground. AndroidManifest.xml contains more info about how to
disable this behavior and what might happen if you do it.
[ChangeLog][Android][Important Behavior Changes] The main event loop is
now stopped when the app is suspended
Task-number: QTBUG-36274
Change-Id: I4c0ba5df9d95f348bca67ea5c76865d6d20775e4
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Reviewed-by: Paul Olav Tvete <paul.tvete@digia.com>
|