The application does not update its screens when there are
no windows on which a WM_DPICHANGE could be received. Add a
check for it to the tray window procedure and trigger an
update from there if no top levels are present.
Fixes: QTBUG-79248
Change-Id: I0b1c4db560662ecf2b473304942da373be6fdc73
Reviewed-by: André de la Rocha <andre.rocha@qt.io>
Reviewed-by: Morten Johan Sørvig <morten.sorvig@qt.io>
This is so that we can do simple changes to the docker images
in provisioning, without needing to update the tag here.
If a backwards-incompatible change in the docker images needs to be committed
in provisioning and here, it is possible because the images retain their old
unique SHA1 tag, in addition to being tagged as latest. See comment for more
details.
Requires the change in qt5 repository, with commit sha:
e4f9ac5607a329bae045567a339d36469bc4fff6
Task-number: QTBUG-79867
Change-Id: I1bc72edec62487530575d7e113a25afe16d09129
Reviewed-by: Volker Hilsheimer <volker.hilsheimer@qt.io>
This matches the default collection behavior of NSWindows.
Change-Id: I363ed211daf6c6c2e579eb11c7294ff509d53e91
Fixes: QTBUG-63829
Reviewed-by: Timur Pocheptsov <timur.pocheptsov@qt.io>
We only read this for desktop environments that have traditionally used
these to set settings for other toolkits.
Fixes: QTBUG-80323
Change-Id: Ifa8c2682301e69c2770d3734115080a0e6b4e85c
Reviewed-by: Morten Johan Sørvig <morten.sorvig@qt.io>
Reviewed-by: Gatis Paeglis <gatis.paeglis@qt.io>
The call to data() on a non-const QString led to a detach(), which is
unexpected and unwanted from QString::lastIndexOf() const.
Found by looking at why QFileSystemEntry::fileName() was expensive, in
the hotspot profiler.
The solution is to instanciate QLastIndexOf with QStringView() rather
than QString(). I added a deleted QString overload to make sure nobody
ever instanciates it with a QString argument again.
Change-Id: I06a1b2f937425e83f0779eb215e099aef78c50a7
Reviewed-by: Anton Kudryavtsev <antkudr@mail.ru>
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
When drawing a QMenu which is checkable but does not have an icon
somewhere, the width of the (possible) checkmark was not considered
during drawing and the text was drawn over the checkmark. Also the wrong
state was checked for drawing the checked icon (if one was given).
Fixes: QTBUG-80506
Task-number: QTBUG-78238
Change-Id: Icf8aa37aab424564054d3549defee93eb0d7c1a4
Reviewed-by: Eirik Aavitsland <eirik.aavitsland@qt.io>
QEvent::EnabledChange did not update the palette of the internal
QPlainTextEditControl which lead to a wrong text color when the
QPlainTextEdit was disabled e.g. due to a QGroupBox.
Fix it the same way it is done in QTextEdit - set the new palette also
to the internal control when QEvent::EnabledChange is received.
Fixes: QTBUG-80150
Change-Id: Icbeddf3d6cd4877a3d8d4a06b2da69383dd776d2
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@qt.io>
Reviewed-by: Richard Moe Gustavsen <richard.gustavsen@qt.io>
QTextStream, reading a file with CR+LF newlines from an stdio FILE
handle that was opened without "b", will always return false in atEnd().
Changing the open mode from "r" to "rb" works around the issue.
Task-number: QTBUG-80443
Change-Id: Ib2eafc0c4c6a6d2bcaeea3036474549d2d9e1511
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Reviewed-by: Joerg Bornemann <joerg.bornemann@qt.io>
When an application only has a trayicon and is killed by `taskkill /IM
binary.exe` the trayicon's HWND will receive a WM_CLOSE message. If we
don't handle this, the tray icon will close anyway, but the app still
runs in the task manager.
Fixes: QTBUG-43855
Change-Id: I5f82a068df9c40360bd565a2681e1b37ff114e44
Reviewed-by: Tor Arne Vestbø <tor.arne.vestbo@qt.io>
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@qt.io>
Apparently this is our preferred style for Objective C member references.
Change-Id: I8b2bbaabadbea2cfa74f209372e77cee79e3c895
Reviewed-by: Tor Arne Vestbø <tor.arne.vestbo@qt.io>
Reviewed-by: Qt CI Bot <qt_ci_bot@qt-project.org>
Inline systemTimeZoneId() in it to save the need for init().
We thus save the lookup by name for a time-zone object, when that
object is what we took the name from anyway.
Do some minor tidy-up in the other constructors and add an assert to
systemTimeZoneId() to match the new one in the default constructor.
Change-Id: Ib70acf31bdb4a4fa1306eebd1fd5f00ad6b89bcc
Reviewed-by: Timur Pocheptsov <timur.pocheptsov@qt.io>
The code was assuming that if the parsing of the value worked,
then it must be a list of 4 variants. But in this case it's just
a single length.
This came from <td> using 4 values for border-width
while other elements use a single value. But the storage
is shared. So the fix is to use 4 values everywhere.
When reading 4 and there's only one, it gets duplicated,
so the caller can just use the first one in that case.
Task-number: QTBUG-80496
Change-Id: I682244b6e3781c4d673a62d5e6511dac263c58e8
Reviewed-by: Nils Jeisecke <nils.jeisecke@saltation.com>
Reviewed-by: Shawn Rutledge <shawn.rutledge@qt.io>
The dark edit field is now centered within the frame around it, with a
thin border on all sides, including between input field and button.
Change-Id: I27e853289e9048c21fdc81e45fadacba9665b49e
Fixes: QTBUG-63454
Reviewed-by: Tor Arne Vestbø <tor.arne.vestbo@qt.io>
The JavaScript implementation of that function has
the following code:
if (typeof JSEvents === 'object')
JSEvents.removeAllHandlersOnTarget(GL.contexts[contextHandle].GLctx.canvas); // Release all
// JS event handlers on the DOM element that the GL context is associated with since the context
// is now deleted.
This breaks mouse/keyboard events, etc.
Disable this logic by temporarily setting the JSEvents object
to undefined, for the duration of the emscripten_webgl_destroy_context
call.
Fixes: QTBUG-74850
Change-Id: Ied3177b0ca6e63e8ea07143bf7d6a850b0bce35a
Reviewed-by: Lorn Potter <lorn.potter@gmail.com>
There's no qt documentation module for 'qtrhi' - instead, it should be added
to the qtgui documentation module.
Fixes: QTBUG-80489
Change-Id: Iea61b907811cd2135c2f1258599d9868d2218679
Reviewed-by: Laszlo Agocs <laszlo.agocs@qt.io>
Two little tool functions had come between it and the function it
actually describes.
Change-Id: Ib49d1623833275ea79c7916fece29aed9503aa40
Reviewed-by: Paul Wicking <paul.wicking@qt.io>
These were kept around to keep Qt Quick compiling, but the
migration there has been done a long time ago. Remove these
leftovers now.
Change-Id: Ibd47381b410b11b5475a85c7ed3cb05c22f7adbb
Reviewed-by: Paul Olav Tvete <paul.tvete@qt.io>
AppKit will initialize NSScreens nowadays, so we don't need to manually
trigger it.
Task-number: QTBUG-80193
Change-Id: Ic0251a1b978b9d4ff53f20e67902787cf529fa87
Reviewed-by: Timur Pocheptsov <timur.pocheptsov@qt.io>
Reviewed-by: Volker Hilsheimer <volker.hilsheimer@qt.io>
License file (LICENSE.QT-LICENSE-AGREEMENT-4.0) was recently updated to
version 4.2 and that broke the configure.
Remove qt license file version number from file name to avoid configure
update need when license file is updated
Change-Id: I77b4a4e7c6e590bbbce79d1c86cbcfb965841eae
Reviewed-by: Lars Knoll <lars.knoll@qt.io>
- start from index -1 each time when we iterate. Each time when we add a
FolderIterator to the stack we MUST reset the index (-1) otherwise it will
continue from it's last position. To fix it we are cloning the FolderIterator
to set the index to -1. The index must be -1 in order to set it to 0 when we
first call next() method.
- introduce "fileType" static method for a more reliable also much faster
file type lookup. The old version of checking if a file exists, is a folder or
a file was buggy that's why it skipped some file randomly.
Fixes: QTBUG-80178
Change-Id: I4b28e4616399b1bff35d792b55ded1bf19b62dd9
Reviewed-by: Eskil Abrahamsen Blomfeldt <eskil.abrahamsen-blomfeldt@qt.io>
When doing a multi ABI build in one go, if the TARGET is the same,
we'll endup with only one library which is not okay.
Task-number: QTBUG-80351
Change-Id: I0c5ff24480446d671e59dbd5a045f9889dff39e9
Reviewed-by: Eskil Abrahamsen Blomfeldt <eskil.abrahamsen-blomfeldt@qt.io>
When connecting/disconnecting/reconfiguring screens Qt needs to know
about the changes before they are propagated by the OS in other ways,
so that we can reflect the changes in the list of QScreens as soon as
they happen.
Unfortunately the canonical notifications for this in AppKit,
NSApplicationDidChangeScreenParametersNotification, is delivered
after AppKit itself reacts to the change, which results in receiving
NSWindowDidChangeScreenNotification, NSWindowWillMoveNotification,
and others, while the list of QScreens is stale.
To work around this we adopted the lower layer Quartz Display
Services API in 3976df2805 to notify us when there are changes to
the screen configuration.
Unfortunately the window server on macOS is not consistent in how
it orders events during screen reconfiguration, and we can't rely
on the NSScreen list being up to date when we get our callbacks
from Quartz.
To work around this we still hook into Quartz, so that we get the
callbacks as early as possible, but then track the state of the
AppKit NSScreen list and update our own QScreens as soon as we
see a change.
We now also include sleeping displays in the list of QScreens,
which matches the behavior of NSScreen.screens. Similarly we
exclude displays that are mirroring another display.
Task-number: QTBUG-80193
Change-Id: I6b1958d6ee61373b2861e05a0d971d2300596f3e
Reviewed-by: Timur Pocheptsov <timur.pocheptsov@qt.io>
Reviewed-by: Morten Johan Sørvig <morten.sorvig@qt.io>
QVulkanWindow support has long been removed from the Vulkan backend.
The include is a leftover from those times.
Change-Id: Ie68ac3611b24310f2b6111a72dd0679adafdc74d
Reviewed-by: Johan Helsing <johan.helsing@qt.io>
Otherwise combobox with editable text field won't work.
Change-Id: I135c3a63cf8fba66d724e140a5a63828853e154e
Reviewed-by: Morten Johan Sørvig <morten.sorvig@qt.io>
We need to reset the pressed index when the decoration was pressed on
otherwise if the mouse ends up over an already selected item that was
previously clicked on. This prevents it from thinking that the mouse
has been released on this item right after pressing on it.
Fixes: QTBUG-59067
Change-Id: Iab372ae20db3682ab0812661f86533079ba4083c
Reviewed-by: Christian Ehrlicher <ch.ehrlicher@gmx.de>
Instead of OID as string, OpenSSL returns a human-readable name.
No need to fail then, use such a name in QVERIFY if OID not found.
Fixes: QTBUG-80329
Change-Id: I4392fc9f37367f7c34c05d172db015b44e2a8497
Reviewed-by: Mårten Nordheim <marten.nordheim@qt.io>
The WebGL 2.0 specification explicitly does not support texture swizzles. Therefore,
disabling it when targeting WASM. This fixes "WebGL: INVALID_ENUM: texParameter:
invalid parameter name" when running in Chrome or Firefox.
Change-Id: Ic7e22e0f623095245274924095cb63fd0ff7e8c2
Reference: https://www.khronos.org/registry/webgl/specs/latest/2.0/#5.19
Fixes: QTBUG-80287
Reviewed-by: Laszlo Agocs <laszlo.agocs@qt.io>
Add the equivalent intel warning macro in public header where there
was already the macro for -Wfloat-equal
Change-Id: I8f20400f0b95c8f3857fa7a0a33464c8c34d5c0e
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
The aim is to make it suitable to test for High DPI
bugs, ideally removing the need to provide bug report
examples.
- Add descriptive window titles/output
- Add options to force scaling on/off
- Change the updating of the text to be done in screenChanged()
and log the signal.
- Rearrange the layout and show the descriptions as labels
Task-number: QTBUG-80323
Change-Id: Ia44c184c2b38cb18045c40b440fe785c6c17925f
Reviewed-by: Shawn Rutledge <shawn.rutledge@qt.io>
- Use C++ constructs like range-based for, member initialization
- Fix formatting in a few places
- Silence clang warnings:
- Add override
- Make member variables private
Task-number: QTBUG-80323
Change-Id: I5b0fda06acb6c8054aafa4dd934b763f8493a6b3
Reviewed-by: Shawn Rutledge <shawn.rutledge@qt.io>
Unlike renderbufferStorage, renderbufferStorageMultisample is not
guaranteed to accept the unsized GL_DEPTH_STENCIL internalformat. For the
former, WebGL 2 guarantees it for compatibility for WebGL 1, but the
multisample version does not exist in WebGL 1, so from the specs it is not
given at all that the unsized format would be accepted. So use the ES 3.0
sized format instead, like we would on a "real" ES 3.0 implementation.
Fixes: QTBUG-80296
Change-Id: I822ae382097085c0a3279c16bb69a173dbf15093
Reviewed-by: Andy Nichols <andy.nichols@qt.io>
macOS fails to create a zone for the name its own systemTimeZone
claims to have (see new comment). So make sure we do consistently
recognize the name systemTimeZoneId() returns, using systemTimeZone
from which we got its name.
Add minimal testing of system time-zone.
Fixes: QTBUG-80173
Change-Id: I42f21efbd7c439158fee954d555414bb180e7f8f
Reviewed-by: Tor Arne Vestbø <tor.arne.vestbo@qt.io>
We hard-coded the assumption the import lib naming scheme is always
basename + ".lib" which is wrong for MinGW.
This amends commit a131d610.
Fixes: QTBUG-80366
Change-Id: Ibefb8a54483cc62743b8783530644b03e720262c
Reviewed-by: Ulf Hermann <ulf.hermann@qt.io>
We introduced a work-around for iOS 11 which breaks on more
recent OS versions because we try to request meta-fonts by
name instead of using the special system font descriptors.
This would cause warnings on the console and Times New Roman
when requesting e.g. the system fixed width font.
When testing on iOS 12 without the work-around, we are no
longer able to reproduce the original issue, so the
assumption is that this problem has been resolved. Since
iOS 11 is not a supported target for Qt 5.14 we can remove
the work-around entirely.
[ChangeLog][macOS/iOS] Fixed a bug where QFontDatabase::systemFont()
would return the wrong fonts on macOS 10.15 and iOS 13.
Fixes: QTBUG-79900
Change-Id: Ie375c8c2ab877d6d66e3696662c4939f639a6e9e
Reviewed-by: Tor Arne Vestbø <tor.arne.vestbo@qt.io>
Fix an issue where qmake on macOS will generate a Makefile with a path
to a macOS bundle instead of a bundle-less executable in the Makefile's
check rule if cmdline is specified before testcase in the CONFIG
options.
Fixes: QTBUG-80280
Change-Id: Icc9ee1355b0641981ce79526b36f29957e1afb00
Reviewed-by: Joerg Bornemann <joerg.bornemann@qt.io>
Introduced by fe6e54fb1f. The probability -> 0,
meaning malloc must fail to trigger it, but it is still a leak. We now use
std::unique_ptr which improves the code in general a bit and fixes a leak.
Change-Id: I6c0fa36953196d3235fb60354dc9ad2396d8dfcb
Reviewed-by: Mårten Nordheim <marten.nordheim@qt.io>
When the geometry of a hidden widget was set with setGeometry(),
WA_PendingMoveEvent and WA_PendingResizeEvent were set unconditionally
even if the crect already had the correct value. This lead to
unneeded Move/Resize events within sendPendingMoveAndResizeEvents().
Fixes: QTBUG-75475
Fixes: QTBUG-79906
Change-Id: Ibbe03882f039948b6b7c04887420741ed2e9c0f7
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@qt.io>
Reviewed-by: Shawn Rutledge <shawn.rutledge@qt.io>
Reviewed-by: Richard Moe Gustavsen <richard.gustavsen@qt.io>
It set the time-zone member sensibly to the default zone, but then
called init("UTC"), which over-wrote that default with UTC. This had
no visible effect (as the default-constructed object is only used to
access methods that (though virtual) are effectively static), but was
needlessly complicated.
Tidied up systemTimeZoneId() at the same time.
Change-Id: I897aff16855c28487a1029bef50c75ebc1ff5b55
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
It was setting the system zone ID if it somehow managed to find a
match for an empty zone name; that makes no sense.
A case-insensitive comparison seems reasonable for the "this isn't
just a default zone object, used because the name I asked for isn't
recognized" check.
It set m_id after it had checked everything, where it could just as
well have used m_id as the variable in which to record whether its
check has succeeded already.
It was using the name it was asked for, rather than the one this ended
up being mapped to, which is probably a better name to use for
it. (This should only differ in case.)
Split a long line.
Change-Id: I41a3b01ba99522ee68f3d7941c532019b9ebf946
Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
The ones we reject used a zero offset while the one that does parse
(though it shouldn't - revised comment) has a one hour offset. Made
them all use that offset and added a partner test that has no invalid
characters, so ensure the success of the invalid character tests isn't
due to falsely rejecting the valid date/time text to which the invalid
characters are added.
Task-number: QTBUG-80038
Change-Id: I6e3dd79b981af6803e60877229c56599cfd719cb
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
