summaryrefslogtreecommitdiffstats
Commit message (Collapse)AuthorAgeFilesLines
* Remove deprecated method QOpenGLFunctions::initializeGLFunctions()Marcel Krems2020-09-082-11/+0
| | | | | Change-Id: I0c02b22630f8c9d463f48f6402a1930b60d3f818 Reviewed-by: Laszlo Agocs <laszlo.agocs@qt.io>
* Remove deprecated API from QSurfaceFormatMarcel Krems2020-09-082-43/+0
| | | | | Change-Id: I8dacdaa18cea967a85e8835c2440ba53ee5df2e1 Reviewed-by: Tor Arne Vestbø <tor.arne.vestbo@qt.io>
* Remove deprecated method QTextList::isEmpty()Marcel Krems2020-09-082-20/+0
| | | | | Change-Id: I7acc0bb110c2bed55374576f59c24451afa7742a Reviewed-by: Volker Hilsheimer <volker.hilsheimer@qt.io>
* Interbase: Correctly read/write arrays to the databaseAndy Shaw2020-09-082-17/+32
| | | | | | | | | | | The fix ensures that it can find the column for the array correctly when reading/writing and also handles the integer typed arrays correctly too. Pick-to: 5.15 Fixes: QTBUG-83409 Change-Id: I92d982bdf0927e6ebc6dce84fec9ad6c44c26c25 Reviewed-by: Christian Ehrlicher <ch.ehrlicher@gmx.de>
* macOS: Don't assume we will get didBecomeKey for non-Qt NSWindowsTor Arne Vestbø2020-09-081-9/+12
| | | | | | | | | | | | The optimization resulted in losing out on window focus changes when for example a native file dialog was shown, resulting in the cursor blinking both in the parent window, and in the native file dialog. Pick-to: 5.15 Pick-to: 5.12 Change-Id: I9c1f9df20fbc5c4b80f906ded70d9a2658b70438 Reviewed-by: Volker Hilsheimer <volker.hilsheimer@qt.io> Reviewed-by: Timur Pocheptsov <timur.pocheptsov@qt.io>
* PostgreSQL: Attempt to subscribe even if it is already addedAndy Shaw2020-09-071-9/+8
| | | | | | | | | | | | As the connection could be lost and then reconnected for the same driver instance then it should just do the LISTEN query as it will not do anything if it is already subscribed. Fixes: QTBUG-84356 Change-Id: I6179bca3991c3828ccee066fd96a6e5003c522be Pick-to: 5.15 Reviewed-by: Lisandro Damián Nicanor Pérez Meyer <perezmeyer@gmail.com> Reviewed-by: Volker Hilsheimer <volker.hilsheimer@qt.io>
* Deliver Quit event when calling QCoreApplication::quit()Tor Arne Vestbø2020-09-070-0/+0
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Instead of QCoreApplication::quit() directly calling exit(0), which would leave QGuiApplication and client code out of the loop, we now send the Quit event, and let it pass through event delivery, before finally ending up in QCoreApplication::event(), where we call exit(0). This has the advantage that QGuiApplication can ensure all windows are closed before quitting, and if any of those windows ignore the close event the quit will be aborted. This aligns the behavior of synthetic quits via QCoreApplication::quit() with spontaneous quits from the platform via QGuiApplicationPrivate::processApplicationTermination. Clients who wish to exit the application without any event delivery or potential user interaction can call the lower level exit() function directly. [ChangeLog][QtGui] Application termination via qApp->quit() will now deliver Quit events to the application, which in turn will result in application windows being closed as part of the application quit, with an option to cancel the application quit by ignoring the close event. Clients who explicitly want to exit the application without any user interaction should call QCoreApplication::exit() explicitly. Task-number: QTBUG-45262 Task-number: QTBUG-33235 Task-number: QTBUG-72013 Task-number: QTBUG-59782 Change-Id: Id4b3907e329b9ecfd936fe9a5f8a70cb66b76bb7 Reviewed-by: Volker Hilsheimer <volker.hilsheimer@qt.io> Reviewed-by: Lars Knoll <lars.knoll@qt.io>
* tst_qshortcut (x2): use op | instead of op + to combine keysGiuseppe D'Angelo2020-09-070-0/+0
| | | | | | | | | | | | | | Following the introduction of QKeyCombination, reduce the number of warnings raised by the test. Drive-by, remove some pointless math like Qt::SHIFT+0, which does not make any sense and would actually fail to compile (shortly). Refactoring the test to fully use QKeyCombination (instead of ints) is left as a future exercise; some QKeyCombination->int warnings are still around. Change-Id: If825bc4c369986623447927bb11493c4f58b544f Reviewed-by: Volker Hilsheimer <volker.hilsheimer@qt.io>
* tst_qmenubar: stop doing math on an enumGiuseppe D'Angelo2020-09-070-0/+0
| | | | | | | Cast to the right datatypes. Change-Id: Ia1ba9be1b0530263e810b50408f432dca485744c Reviewed-by: Volker Hilsheimer <volker.hilsheimer@qt.io>
* Huffman compression: handle QByteArray's elements as unsigned charTimur Pocheptsov2020-09-070-0/+0
| | | | | | | Fixes: QTBUG-86418 Pick-to: 5.15 Change-Id: Ie57ecad2b1fda7c2fc0de60cbdb4304c00e645d0 Reviewed-by: Mårten Nordheim <marten.nordheim@qt.io>
* Update QList's documentation bitsAndrei Golubev2020-09-070-0/+0
| | | | | | | | Fixed some QList documentation that described old API/behavior Change-Id: I9101ebb7bed9bcac328509765f8e9b85d63d305b Reviewed-by: Lars Knoll <lars.knoll@qt.io> Reviewed-by: Mårten Nordheim <marten.nordheim@qt.io>
* QCommonArrayOps: append when inserting into empty containerAndrei Golubev2020-09-070-0/+0
| | | | | | | | | | | Ensured append is chosen instead of prepend for corner cases when inserting into an empty container Mirrored the logic in shouldGrowBeforeInsert function Task-number: QTBUG-84320 Change-Id: I1c963a2588c331029e450fe55001bbf324f65fb4 Reviewed-by: Lars Knoll <lars.knoll@qt.io>
* QList/QString/QByteArray: no prepend optimized allocation when emptyAndrei Golubev2020-09-070-0/+0
| | | | | | | | | | | | | | | | | | Scoped prepend optimized allocation to only work when prepending into a non-empty container. Otherwise, even appends would be considered prepends since d.size == 0 when container is empty This is, of course, not good for prepend cases but we prefer appends over prepends. My proposal is to figure out what's the best strategy based on use cases and performance measurements. For now, let's just make sure appends are not additionally pessimized Anyhow, this is an implementation detail and should not be considered behavior change (at least not the one that is user noticeable) Task-number: QTBUG-84320 Change-Id: Ibed616a2afa9bc24f78252f15a617bf92e2c6ea3 Reviewed-by: Lars Knoll <lars.knoll@qt.io>
* QCommonArrayOps: append when inserting into empty containerAndrei Golubev2020-09-071-3/+3
| | | | | | | | | | | Ensured append is chosen instead of prepend for corner cases when inserting into an empty container Mirrored the logic in shouldGrowBeforeInsert function Task-number: QTBUG-84320 Change-Id: I1c963a2588c331029e450fe55001bbf324f65fb4 Reviewed-by: Lars Knoll <lars.knoll@qt.io>
* QList/QString/QByteArray: no prepend optimized allocation when emptyAndrei Golubev2020-09-073-7/+6
| | | | | | | | | | | | | | | | | | Scoped prepend optimized allocation to only work when prepending into a non-empty container. Otherwise, even appends would be considered prepends since d.size == 0 when container is empty This is, of course, not good for prepend cases but we prefer appends over prepends. My proposal is to figure out what's the best strategy based on use cases and performance measurements. For now, let's just make sure appends are not additionally pessimized Anyhow, this is an implementation detail and should not be considered behavior change (at least not the one that is user noticeable) Task-number: QTBUG-84320 Change-Id: Ibed616a2afa9bc24f78252f15a617bf92e2c6ea3 Reviewed-by: Lars Knoll <lars.knoll@qt.io>
* Update QList's documentation bitsAndrei Golubev2020-09-071-44/+43
| | | | | | | | Fixed some QList documentation that described old API/behavior Change-Id: I9101ebb7bed9bcac328509765f8e9b85d63d305b Reviewed-by: Lars Knoll <lars.knoll@qt.io> Reviewed-by: Mårten Nordheim <marten.nordheim@qt.io>
* tst_qmenubar: stop doing math on an enumGiuseppe D'Angelo2020-09-071-1/+1
| | | | | | | Cast to the right datatypes. Change-Id: Ia1ba9be1b0530263e810b50408f432dca485744c Reviewed-by: Volker Hilsheimer <volker.hilsheimer@qt.io>
* Huffman compression: handle QByteArray's elements as unsigned charTimur Pocheptsov2020-09-071-2/+4
| | | | | | | Fixes: QTBUG-86418 Pick-to: 5.15 Change-Id: Ie57ecad2b1fda7c2fc0de60cbdb4304c00e645d0 Reviewed-by: Mårten Nordheim <marten.nordheim@qt.io>
* tst_qshortcut (x2): use op | instead of op + to combine keysGiuseppe D'Angelo2020-09-072-54/+54
| | | | | | | | | | | | | | Following the introduction of QKeyCombination, reduce the number of warnings raised by the test. Drive-by, remove some pointless math like Qt::SHIFT+0, which does not make any sense and would actually fail to compile (shortly). Refactoring the test to fully use QKeyCombination (instead of ints) is left as a future exercise; some QKeyCombination->int warnings are still around. Change-Id: If825bc4c369986623447927bb11493c4f58b544f Reviewed-by: Volker Hilsheimer <volker.hilsheimer@qt.io>
* Deliver Quit event when calling QCoreApplication::quit()Tor Arne Vestbø2020-09-0711-25/+103
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Instead of QCoreApplication::quit() directly calling exit(0), which would leave QGuiApplication and client code out of the loop, we now send the Quit event, and let it pass through event delivery, before finally ending up in QCoreApplication::event(), where we call exit(0). This has the advantage that QGuiApplication can ensure all windows are closed before quitting, and if any of those windows ignore the close event the quit will be aborted. This aligns the behavior of synthetic quits via QCoreApplication::quit() with spontaneous quits from the platform via QGuiApplicationPrivate::processApplicationTermination. Clients who wish to exit the application without any event delivery or potential user interaction can call the lower level exit() function directly. [ChangeLog][QtGui] Application termination via qApp->quit() will now deliver Quit events to the application, which in turn will result in application windows being closed as part of the application quit, with an option to cancel the application quit by ignoring the close event. Clients who explicitly want to exit the application without any user interaction should call QCoreApplication::exit() explicitly. Task-number: QTBUG-45262 Task-number: QTBUG-33235 Task-number: QTBUG-72013 Task-number: QTBUG-59782 Change-Id: Id4b3907e329b9ecfd936fe9a5f8a70cb66b76bb7 Reviewed-by: Volker Hilsheimer <volker.hilsheimer@qt.io> Reviewed-by: Lars Knoll <lars.knoll@qt.io>
* Clean up QTextBoundaryFinder and qunicodetoolsLars Knoll2020-09-076-200/+187
| | | | | | | | | | | Make QTBF ready for Qt6 by using qsizetype in the API and use QStringView where it makes sense. Change the exported API of qunicodetools to use QStringView as well and use char16_t internally. Change-Id: I853537bcabf40546a8e60fdf2ee7d751bc371761 Reviewed-by: Mårten Nordheim <marten.nordheim@qt.io>
* Make QStringMatcher ready for Qt 6Lars Knoll2020-09-072-48/+31
| | | | | | | | | | Use qsizetype for string indices everywhere. Clean up the data structures and remove some Qt 3 or Qt 4 left-overs. This reduces the size of the QStringMatcher from 1056 to 288 bytes. Change-Id: Icc351da8e3aad10a6c940196f52c39f8d2f5bf80 Reviewed-by: Fabian Kosmale <fabian.kosmale@qt.io>
* Remove deprecated method QQuaternion::conjugateMarcel Krems2020-09-073-24/+4
| | | | | Change-Id: If539e47618357ef329925033842f266447632e6f Reviewed-by: Volker Hilsheimer <volker.hilsheimer@qt.io>
* QStyleOptionGraphicsItem: remove deprecated member matrix/levelOfDetailChristian Ehrlicher2020-09-066-64/+23
| | | | | | | | QStyleOptionGraphicsItem::matrix and levelOfDetail are deprecated and not used anywhere anymore. Change-Id: Id4b0fa8ac3ac53d771e858ecda5524a5e690342d Reviewed-by: Volker Hilsheimer <volker.hilsheimer@qt.io>
* Use C++17 [[maybe_unused]]Allan Sandfeld Jensen2020-09-0611-187/+192
| | | | | | | In some places needs to be ordered before const/constexpr though. Change-Id: I57a521ac0ad22b5a018761c4d52befbef69d64c0 Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
* Round float->qfloat16 to evenAllan Sandfeld Jensen2020-09-064-14/+629
| | | | | | | | Calibrated to match F16C and ARM-FP16 hardware conversions. Change-Id: I3bdd4d3db3046fee4aeb24e4ce8b9bc9a06e0397 Reviewed-by: Edward Welbourne <edward.welbourne@qt.io> Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
* Inline QTest::qSleep()Lars Knoll2020-09-064-45/+8
| | | | | | | Simply use QThread::msleep() to implement it. Change-Id: I37c255fc70951715edc489d9f67669b01af380b1 Reviewed-by: Tor Arne Vestbø <tor.arne.vestbo@qt.io>
* Get rid of hasPendingEvents() and flush()Lars Knoll2020-09-0625-152/+6
| | | | | | | They are unused. Change-Id: I77383f2be45551401ed9c2f88285511134cc8b0d Reviewed-by: Tor Arne Vestbø <tor.arne.vestbo@qt.io>
* Remove two obsolete #definesLars Knoll2020-09-061-5/+0
| | | | | Change-Id: Icc13408cfdb8ce0db6f274904c3e44f8376cd1e5 Reviewed-by: Tor Arne Vestbø <tor.arne.vestbo@qt.io>
* Cleanup QSemaphore and make it always 64bit largeLars Knoll2020-09-052-104/+60
| | | | | | | | This simplifies the implementation so that we don't need separate code paths for 32 and 64 bit anymore. Change-Id: I823612865e7d648fb0bd1632385ce67b5a452b8a Reviewed-by: Volker Hilsheimer <volker.hilsheimer@qt.io>
* Cleanup a ### Qt 6Lars Knoll2020-09-052-6/+5
| | | | | | Change-Id: If9dd04e25ddc5164cc23ac7d5684254e38b7eda6 Reviewed-by: Volker Hilsheimer <volker.hilsheimer@qt.io> Reviewed-by: Mårten Nordheim <marten.nordheim@qt.io>
* Remove /INCREMENTAL:YES from linker flagsMarcel Krems2020-09-051-2/+0
| | | | | | | | | This flag doesn't exist in MSVC 2015+: LINK : warning LNK4224: /INCREMENTAL:YES is no longer supported; ignored https://docs.microsoft.com/en-us/cpp/build/reference/incremental-link-incrementally?view=vs-2015 Change-Id: I3b2aace619351e2dc9cc86e1df28facc21b7d9fe Reviewed-by: Volker Hilsheimer <volker.hilsheimer@qt.io>
* tst_qdbuscpp2xml: stop doing math on enumerationsGiuseppe D'Angelo2020-09-051-1/+1
| | | | | | | Cast it to int to do math. Change-Id: Ia553cccb7e160df50e77fba63d66d0ed45c246c2 Reviewed-by: Volker Hilsheimer <volker.hilsheimer@qt.io>
* QCompleterItemDelegate: replace QItemDelegate with QStyledItemDelegateWang Peng2020-09-051-5/+4
| | | | | | | Resolve remaining Qt6 TODOs Change-Id: Ice8dae2f1cbd3ed452ab78d5d7735127d7662900 Reviewed-by: Volker Hilsheimer <volker.hilsheimer@qt.io>
* QCalendarDelegate: replace QItemDelegate with QStyledItemDelegateWang Peng2020-09-051-6/+5
| | | | | | | | Resolve remaining Qt6 TODOs Change-Id: I018ef5e3761816dce052a78be3dcd07ede866135 Reviewed-by: JiDe Zhang <zccrs@live.com> Reviewed-by: Volker Hilsheimer <volker.hilsheimer@qt.io>
* QMetaType: optimize comparisonFabian Kosmale2020-09-051-1/+1
| | | | | | | We can skip the id calls if we know that the pointers are equal. Change-Id: I62f9cac557d7b82b640a143965f9056a8cd46028 Reviewed-by: Ulf Hermann <ulf.hermann@qt.io>
* Use QList instead of QVector in qmakeJarek Kobus2020-09-0511-22/+22
| | | | | | | Task-number: QTBUG-84469 Change-Id: I4a3da94702f1dad1ee10b1ba3c6712b6f40338c9 Reviewed-by: Lars Knoll <lars.knoll@qt.io> Reviewed-by: Friedemann Kleint <Friedemann.Kleint@qt.io>
* When finding the engine for a fallback font, then clear the familiesAndy Shaw2020-09-051-0/+1
| | | | | | | | | | | If it is a fallback font, then it should just have family set to that font so it does not get interference from what the families were originally set to. Fixes: QTBUG-85560 Change-Id: I6232f3d2ae12052fa3b0b3bc0e7f106e239a585d Pick-to: 5.15 Reviewed-by: Eskil Abrahamsen Blomfeldt <eskil.abrahamsen-blomfeldt@qt.io>
* Get rid of the MSVC specific fileLars Knoll2020-09-053-61/+0
| | | | | | | | | It used to contain the MSVC specific export hack for QVector, but that one is not required in Qt 6 anymore and the file was not doing anything anymore. Change-Id: Ic8b4aa355a8934beb6abcf10235d218344a294cc Reviewed-by: Mårten Nordheim <marten.nordheim@qt.io>
* QStyle: remove deprecated SH_SpellCheckUnderlineStyleChristian Ehrlicher2020-09-054-10/+0
| | | | | | | | It was deprecated with 88e6f8cff2974c46b1262f3a1a61e1440c664e0c and is not used anywhere anymore so it can be safely removed. Change-Id: If4050ac8bf116fb31491b3b08096554c5ea3e4d5 Reviewed-by: Volker Hilsheimer <volker.hilsheimer@qt.io>
* Make test pass on machines with many coresLars Knoll2020-09-051-1/+1
| | | | | | | | | | Just starting 20 threads to test this won't cut it if the machine you're testing on has an ideal thread count of 16 or larger. Change-Id: Icba8f00aa836fec6da41c71b318e9e17bdd47c0e Reviewed-by: Volker Hilsheimer <volker.hilsheimer@qt.io> Reviewed-by: Mårten Nordheim <marten.nordheim@qt.io>
* Remove some comments that are no longer relevantEdward Welbourne2020-09-054-7/+2
| | | | | | | Task-number: QTBUG-85700 Change-Id: I5ce368e8edca2b9483a0f0ef34bc9eb6b4e44574 Reviewed-by: Mitch Curtis <mitch.curtis@qt.io> Reviewed-by: Volker Hilsheimer <volker.hilsheimer@qt.io>
* Remove some obsolete codeLars Knoll2020-09-053-95/+0
| | | | | | | Those checks aren't required anymore for Qt 6. Change-Id: Ida9f93544f42cb03723659a81e0094aa0cf14799 Reviewed-by: Volker Hilsheimer <volker.hilsheimer@qt.io>
* QWindowsKeyMapper: fix mixed math between int and enum/flagsGiuseppe D'Angelo2020-09-051-3/+3
| | | | | | | | | Either do it between ints, or cast to the right types before doing it. This should enable removing operator+ and - between int and QFlags. Change-Id: I30ecc03566cff8ce880b048ba3a9d7d50f3c8509 Reviewed-by: Friedemann Kleint <Friedemann.Kleint@qt.io>
* Change QWidget::enterEvent signature to take a QEnterEventVolker Hilsheimer2020-09-0514-22/+21
| | | | | | | | | | | | | | | | | | | This is a source incompatible change for widget implementors. Leaving the old enterEvent as a virtual overload is problematic due to shadowing. Best to make a clean cut, widget reimplementors will get a compile time warning if they mark their override as such, or if they try to call the parent class implementation. Addresses ### Qt 6 comment. [ChangeLog][QtWidgets][QWidget] The virtual enterEvent handler now receives a QEnterEvent, which contains information about mouse position and button states, rather than a plain QEvent. Change-Id: I233f594fd79c0c090983b3db8532913d00132fde Reviewed-by: Shawn Rutledge <shawn.rutledge@qt.io> Reviewed-by: Friedemann Kleint <Friedemann.Kleint@qt.io> Reviewed-by: Lars Knoll <lars.knoll@qt.io>
* Change translation context of strings to new classVolker Hilsheimer2020-09-051-13/+6
| | | | | | | | | | | The strings were in QFileDialog context even though they were in QFileIconProvider, probably for historical reasons. Now that the strings have been moved into QtGui, using a QtWidgets class as the context makes no sense. Change-Id: Ia4f3bd18abaab2a5fbbb94e945782f4d2d94e7d1 Reviewed-by: Lars Knoll <lars.knoll@qt.io>
* Fix compiler warnings from using deprecated qChecksum overloadVolker Hilsheimer2020-09-052-4/+4
| | | | | Change-Id: Ib7a6d0e752dc662a42abe38f2010da3f2b4626f1 Reviewed-by: Mårten Nordheim <marten.nordheim@qt.io>
* Fix compiler warnings after QKeyCombination introductionVolker Hilsheimer2020-09-052-11/+10
| | | | | Change-Id: If5d7c5e037b99c14c51d83adf8b1e20d6b924bc5 Reviewed-by: Tor Arne Vestbø <tor.arne.vestbo@qt.io>
* Fix compiler warning from unused private field in QSequentialIterableVolker Hilsheimer2020-09-051-0/+1
| | | | | Change-Id: I9c330b4f99f77404eacda2506d264f917f9e1161 Reviewed-by: Ulf Hermann <ulf.hermann@qt.io>
* xcb: drop Xinerama supportLiang Qi2020-09-0411-74/+5
| | | | | | | | | | [ChangeLog][Important Behavior Changes][X11] Xinerama is no longer supported. Fixes: QTBUG-86082 Change-Id: Ieb57d9035e1659fc22bf8333247fc3573fb62992 Reviewed-by: Tor Arne Vestbø <tor.arne.vestbo@qt.io> Reviewed-by: Gatis Paeglis <gatis.paeglis@qt.io>
generated by cgit v1.2.3 (git 2.25.1) at 2024-06-02 03:52:36 +0000