The new takeRow() functions return a pair of pointers to
QLayoutItems and, as the name particle 'take' suggests,
releases ownership of these layout items. Which in turn
means that the caller of the function is supposed to deal
with them.
This was not done here.
To fix, write a RAII class that takes ownership of the
returned layout items, deleting them when it goes out of
scope or gets a new value assigned (only move special
member functions are implemented, making the class move
-only).
Deleting the QLayoutItems is not so easy, though:
QFormLayout has a special function for clearing the
QLayoutItems out, so it appears that just calling their
destructors is not going to fly (though I don't know off
the top of the head why that should be a problem).
Solve this, for now, by adding the layout items back into
a temporary QFormLayout for destruction.
Change-Id: If862989207b20f1e3f757c19ec9d498c4491184f
Reviewed-by: Samuel Gaist <samuel.gaist@edeltech.ch>
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
HTTP/2 allows a server to pre-emptively send (or "push") responses (along
with corresponding "promised" requests) to a client in association with a
previous client-initiated request. This can be useful when the server
knows the client will need to have those responses available in order
to fully process the response to the original request.
Server push is semantically equivalent to a server responding to a request;
however, in this case, that request is also sent by the server, as a
PUSH_PROMISE frame.
The PUSH_PROMISE frame includes a header block that contains a complete set
of request header fields that the server attributes to the request.
After sending the PUSH_PROMISE frame, the server can begin delivering the
pushed response as a response on a server-initiated stream that uses the
promised stream identifier.
This patch:
- fixes the HPACK decompression of PUSH_PROMISE frames;
- allows a user to enable PUSH_PROMISE;
- processes and caches pushed data for promised streams;
- updates auto-test - emulates a simple PUSH_PROMISE
scenario.
Change-Id: Ic4850863a5e3895320baac3871a723fc091b4aca
Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
The Qt::ISODate format strips milliseconds, so a new format is introduced
that keeps the milliseconds. A new format was chosen over fixing the
existing format due to the behavioral change of suddenly having ms
as part of Qt::ISODate.
Change-Id: If8b852daed068cce8eee9b61a7cd4576bc763443
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
By templating on the <chrono> types and unconditionally using
duration_cast to coerce the duration into a milliseconds, we
violate a principal design rule of <chrono>, namely that non-
narrowing conversions are implicit, but narrowing conversions
need duration_cast. By accepting any duration, we allow non-
sensical code such as
QTimer::singleShot(10us, ...)
to compile, which is misleading, since it's actually a zero-
timeout timer.
Overloading a non-template with a template also has adverse
effects: it breaks qOverload().
Fix by replacing the function templates with functions that
just take std::chrono::milliseconds. This way, benign code
such as
QTimer::singleShot(10s, ...)
QTimer::singleShot(10min, ...)
QTimer::singleShot(1h, ...)
work as expected, but attempts to use sub-millisecond
resolution fails to compile / needs an explicit user-
provided duration_cast.
To allow future extension to more precise timers, forcibly
inline the functions, so they don't partake in the ABI of the
class and we can later support sub-millisecond resolution by
simply taking micro- or nano- instead of milliseconds.
Change-Id: I12c9a98bdabefcd8ec18a9eb09f87ad908d889de
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
There's a conflict between QGtk3Menu and QDbusPlatformMenuBar. The
problem is that on Unity the type of the platform menu instance must
be different depending on whether the menu is in the global menubar or
a standalone context menu.
Since QMenu creates a platform menu instance at construction time, it
does not yet know whether it will be added into a menubar. QMenuBar
checks that the QMenu already has a platform menu instance, and passes
it to the platform menubar. As a result, a QGtk3Menu instance is passed
to QDbusPlatformMenuBar.
Currently, a standalone QMenu does not use the native platform menu
instance. Only menus that are added to a QMenuBar do. Therefore we
don't need to create the platform instance when QMenu is constructed,
but only after it is added to QMenuBar. The platform menu instance
creation is implemented in QMenuBarPrivate::getPlatformMenu(), and
QMenu::setPlatformMenu() calls syncPlatformMenu() to take care of
syncing the QMenu properties and actions to the new platform menu
instance.
The macOS-specific methods QMenu::toNSMenu() and QMenu::setAsDockMenu()
rely on the platform menu instance, and must therefore create it on
demand.
This is a hot fix for the release blocker, not a long term solution.
In the future, if standalone QMenus are made to use native platform
menu instances, the instance must be created lazily when the menu is
about to be made visible.
Task-number: QTBUG-56526
Change-Id: I044933cabb1639406fe47908dfc4b1903af214d1
Reviewed-by: Dmitry Shachnev <mitya57@gmail.com>
Reviewed-by: Lars Knoll <lars.knoll@qt.io>
Fixes the test for width of condensed fonts so it doesn't depend
on the presence of the Liberation font on the system, and adds
another test that condensed sub-families can be matched
consistently. The latter will however not work on Windows until
QTBUG-53458 is solved.
Task-number: QTBUG-51335.
Change-Id: Id6d046274fa21b2dce0ad6b32dce7f1c8a92a4f4
Reviewed-by: Eskil Abrahamsen Blomfeldt <eskil.abrahamsen-blomfeldt@qt.io>
lumping together all kinds of unrelated stuff has caused problems with
spurious dependencies from the beginning. as the modularization infra is
now in a state which supports many small private libraries just fine,
take advantage of it.
Change-Id: Ic40f47ce76a308bbfd32deae281f6f064fe1ef4c
Reviewed-by: Jake Petroules <jake.petroules@qt.io>
Commit c5db8fc74 changed all instances of Q_WS_FOO to have the prefix
Q_DEAD_CODE_FROM_QT4 instead, to make it clearer when reading the code
that the code in question was a left-over from Qt4, when we used
Q_WS_ defines instead of Q_OS_ defines.
This worked well for cases of #ifdef Q_DEAD_CODE_FROM_QT4, but less so
for cases of #ifndef Q_DEAD_CODE_FROM_QT4, where the code was actually
unconditionally included.
To make this even clearer, the defines have been replaced by checks for
1 or 0, with a comment describing how the code used to look in Qt4. The
use of constants in the check also makes it easier for editors to parse
the condition and show visually that the code is defined out.
Change-Id: I152070d87334df7259b417cd5e17d7b7950379b7
Reviewed-by: Lars Knoll <lars.knoll@qt.io>
Adds missing image-formats so it doesn't segfault.
Also changes the exclusion of rare formats to lists of included ones
Change-Id: I1d00562cf8e96baa03121a0b996764224911e06a
Reviewed-by: Gunnar Sletta <gunnar@sletta.org>
The previous sync() of the specific scope and domain may have failed due
to not having the necessary permissions, but the saved value may still be
cached, so we need to check both.
This was observed on macOS Sierra, where a failed sync() will result in
marking the CFPrefsPlistSource as read-only, eg:
2016-10-04 13:14:11.713271 tst_qsettings[88537:767733] [User Defaults] attempt to set
<private> for key in <private> in read-only (due to a previous failed write) preferences
domain CFPrefsPlistSource<0x6180000e1780> (Domain: org.software.KillerAPP,
User: kCFPreferencesAnyUser, ByHost: Yes, Container: (null))
Change-Id: I8976c1c4acfe2cb0d5510298d5c585faca9607f6
Reviewed-by: Erik Verbruggen <erik.verbruggen@qt.io>
WinRT is not the only sandboxed platform. Since it doesn't hurt to keep
the test data in a well known location, we enable the code for all
platforms. We also make sure to mkpath the location, since writableLocation
doesn't guarantee that the location exists.
Change-Id: Ie8d90c5fbdf3b7fbf85ba6be25372b0ef7c4da55
Reviewed-by: Maurice Kalinowski <maurice.kalinowski@qt.io>
Reviewed-by: Richard Moe Gustavsen <richard.gustavsen@qt.io>
Before commit f1e9076809, mousePressEvent
called the virtual method setSelection(const QRect&, SelectionFlags)
with the 1x1 rectangle which contains only the clicked QPoint, unless
the SelectionFlag "Current" was set because Shift was pressed during
the mouse press.
Since that commit, the behavior has been changed such that the
rectangle is the one that is spanned by the center of the clicked item
and the clicked pixel. In theory, the result should be the same (i.e.,
only the clicked item should be selected), but
* the code path in QListView::setSelection for 1x1 QRects is more
efficient, and
* using a larger QRect can cause problems with custom views, see the
comments in QTBUG-18009
This commit ensures that the 1x1 QRect is used again, unless the
SelectionFlag "Current" is used.
Change-Id: I70dd70c083c20a3af6cd6095aa89a489756b505f
Task-number: QTBUG-18009
Reviewed-by: Olivier Goffart (Woboq GmbH) <ogoffart@woboq.com>
Reviewed-by: Thorbjørn Lund Martsum <tmartsum@gmail.com>
... and avoid detach()ing potentially large data for just
preserving the QPlatformPixmap::pixelType().
A QBitmap differs from a QPixmap (its base class, urgh)
by always having a data != nullptr and a Bitmap pixel
type, yet load() was unconditionally setting 'data' to
nullptr on failure, turning a QBitmap into a non-QBitmap.
Fix by move-assigning a null QBitmap instead of resetting
'data'.
Add some tests.
Change-Id: Ida58b3b24d96472a5f9d0f18f81cc763edcf3c16
Reviewed-by: Anton Kudryavtsev <a.kudryavtsev@netris.ru>
Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
Reviewed-by: Ulf Hermann <ulf.hermann@qt.io>
By using _MIN for signed values, and _MAX for unsigned values, we may
detect conversion issues when serializing QVariants using QSettings.
Change-Id: I3ce58ba4b93f791f75c7ae44d1fd5030f07b2f25
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
This leaves a clearer separation between the foreign-window and non-foreign
window use-cases, where a single QCococaWindow can only be in one mode,
which is determined in the constructor and doesn't change after that.
There are no source or binary compatibility guarantees for the QPA classes,
meaning the helper function in QPlatformNativeInterface can be removed.
Change-Id: I3232aedca1d98c49a8f54e16750832187f9dc69a
Reviewed-by: Morten Johan Sørvig <morten.sorvig@qt.io>
The fix is trivial, but the patch adds a new QSettings tests that iterates most
of the QMetaTypes and verifies that storing and retrieving them again gives the
same value. This is a more complete test than the testVariantTypes tests, which
is limited to a subset of the QVariant types. The new tests borrows logic from
the QMetaType test machinery.
QSettings has been Q_ENUM'ified in the process, for improved debug output.
Note that on backends such as the INI backend, the metatype of the QVariant read
from the settings will be a string, so it won't match the input QVariant type,
but the result of converting that to the original value type should still work.
Task-number: QTBUG-56124
Change-Id: Ib03a26abf77c9fb449b94160d28bc4baeb095f25
Reviewed-by: Jake Petroules <jake.petroules@qt.io>
The slot is invoked from QObject::destroyed(), which is emitted
from ~QObject. By that time the object is no longer a QShortcut,
so the static_cast it invalid.
Found by UBSan:
tst_qshortcut.cpp:1210:53: runtime error: downcast of address 0x6020000289d0 which does not point to an object of type 'QShortcut'
0x6020000289d0: note: object is of type 'QObject'
10 00 80 17 c0 ce 63 df 93 2b 00 00 b0 02 00 00 d0 60 00 00 02 00 00 00 ff ff ff 04 04 00 00 00
^~~~~~~~~~~~~~~~~~~~~~~
vptr for 'QObject'
#0 0x42b3bb in tst_QShortcut::shortcutDestroyed(QObject*) tst_qshortcut.cpp:1210
#1 0x446cc9 in tst_QShortcut::qt_static_metacall(QObject*, QMetaObject::Call, int, void**) .moc/tst_qshortcut.moc:186
#2 0x2b93dba52c86 in QMetaObject::activate(QObject*, int, int, void**) qobject.cpp:3787
#3 0x2b93dba55400 in QObject::destroyed(QObject*) .moc/moc_qobject.cpp:213
#4 0x2b93dba8d80d in QObject::~QObject() qobject.cpp:967
#5 0x2b93c6b6e032 in QShortcut::~QShortcut() qshortcut.cpp:476
#6 0x2b93c6b6e370 in QShortcut::~QShortcut() qshortcut.cpp:481
#7 0x42a5de in void qDeleteAll<QList<QShortcut*>::const_iterator>(QList<QShortcut*>::const_iterator, QList<QShortcut*>::const_iterator) qalgorithms.h:317
#8 0x42a5de in void qDeleteAll<QList<QShortcut*> >(QList<QShortcut*> const&) qalgorithms.h:325
#9 0x42a5de in tst_QShortcut::clearAllShortcuts() tst_qshortcut.cpp:1136
Fix by replacing QVector::replaceAll() with the erase-remove idiom,
which does not require the cast, because it can perform mixed-type
lookups.
Change-Id: I4251c1895fa4398023f489dbfd7108d90c1a6c94
Reviewed-by: Giuseppe D'Angelo <giuseppe.dangelo@kdab.com>
To keep the change minimal, keep 'sw' as a pointer variable,
but back it by a stack-allocated QStackedWidget instead of a
heap-allocated one that's never deleted.
Change-Id: I9e2a8c07979b861eb7e7040c144d8e75c90d0bc9
Reviewed-by: Giuseppe D'Angelo <giuseppe.dangelo@kdab.com>
Saving strings with embedded zero-bytes (\0) as CFStrings would
sometimes fail, and only write the part of the string leading
up to the first zero-byte, instead of all the way to the final
zero-terminator. This bug was revealed by the code-path that
falls back to storing e.g. QTime as strings, via the helper
method QSettingsPrivate::variantToString().
We now use the same approach as on platforms such as Windows
and WinRT, where the string produced by variantToString() is
checked for null-bytes, and if so, stored using a binary
representation instead of as a string. For our case that
means we fall back to CFData when detecting the null-byte.
To separate strings from regular byte arrays, new logic has
been added to variantToString() that wraps the null-byte
strings in @String(). That way we can implement a fast-path
when converting back from CFData, that doesn't go via the
slow and lossy conversion via UTF8, and the resulting QVariant
will be of type QVariant::ByteArray. The reason for using
UTF-8 as the binary representation of the string is that
in the case of storing a QByteArray("@foo") we need to
still be able to convert it back to the same byte array,
which doesn't work if the on-disk format is UTF-16.
Task-number: QTBUG-56124
Change-Id: Iab2f71cf96cf3225de48dc5e71870d74b6dde1e8
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
A fair amount of tests are skipped if we can't write to the system scope,
eg on iOS. Without this detection they will fail.
Change-Id: I8257f1f24e69dae88925c20d2bff851e81701405
Reviewed-by: Jake Petroules <jake.petroules@qt.io>
it does not appear that the tests actually use these flags in any way;
they don't include any (actual) d-bus headers and have no ifdefs.
and the qdbus module already pulls in the flags via QMAKE_USE (in the
case where they are defined at all, i.e., dbus-linked).
Change-Id: Ie6bc6da7d1dd96da7b73f2d0fe45576936715874
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
The layout isn't actually created until endLayout() or setLineWidth() is
called. So in the case where this was not done, the height of the line
would be 0, thus multiple lines would be placed on top of each other, at
y == 0.
[ChangeLog][QtGui][Text] Fixed QStaticText when manually breaking lines
and no text width was set.
Task-number: QTBUG-56346
Change-Id: I7f6ed6260545882f05fe39b21134315eca7401b9
Reviewed-by: Konstantin Ritt <ritt.ks@gmail.com>
this actually fixes the build when ssl is not enabled, as the openssl
features are in the not included network-private module.
Change-Id: Ibafae9867af493da184a45cf3981628d475d37a6
Reviewed-by: Jake Petroules <jake.petroules@qt.io>
It fails on iOS as well, and likely also tvOS and watchOS.
Change-Id: Idfce98a5aeccb5680f6b4c6e66b526dd7922156d
Reviewed-by: Jake Petroules <jake.petroules@qt.io>
Turn the list of newRow() calls into a for loop over a constexpr
data structure.
Fixes the GCC note:
tst_qgraphicsview_2.cpp:47:13: note: variable tracking size limit exceeded with -fvar-tracking-assignments, retrying without
and speeds up compilation of the file from 13s to 2.5s on my
machine.
Task-number: QTBUG-38890
Change-Id: I4f0b3565c7df64b286d1d32eb3f3d6bf4df92609
Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
Added the binary compatibility test data files for Qt 5.7 for QtBase
Change-Id: I5b19571f5e266c52622027d820062afa5fd4fbf3
Reviewed-by: Sergio Ahumada <sahumada@texla.cl>
CFAbsoluteTime is measured in seconds, represented by a double,
so when converting milliseconds to CFAbsoluteTime we may get a
slight error due to missing precision in double to represent
the milliseconds exactly. By rounding to the closest millisecond
when converting back, we avoid truncating and being one ms off.
Change-Id: If1e99f97b000fb8cb893ddfc5d7ba81096c0ea88
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Reviewed-by: Jake Petroules <jake.petroules@qt.io>
In tst_QCompleter, two completers were leaked because they
had no parent and setCompleter() calls don't reparent.
Fixed by giving them parents.
In tst_QUndo*, fix lots of leaked QActions by storing them
in a QScopedPointer. There were some half-hearted attempts
to clean them up with manual deletes, but I ported these to
scoped pointers, too, to make the code more robust in the
face of failures.
This fixes the remaining errors in GCC 6.1 Linux ASan runs of
tests/auto/widgets/util.
Change-Id: Icc5248cc9cf4514540915924df1c4d9e09c071fa
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Reviewed-by: Olivier Goffart (Woboq GmbH) <ogoffart@woboq.com>
The generated data files for the binary compatibility test
updated for QtBase.
Change-Id: Idae703c83f55ff17cada4419db742ea12b22bf86
Reviewed-by: Sergio Ahumada <sahumada@texla.cl>
Reviewed-by: Tony Sarajärvi <tony.sarajarvi@qt.io>
The usual:
- delete styles
Either by using QScopedPointer.
This fixes the remaining errors in GCC 6.1 Linux ASan runs of
tests/auto/widgets/styles.
Change-Id: Ifba59085c057d474bf964cbb93010c408d773a61
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Reviewed-by: Olivier Goffart (Woboq GmbH) <ogoffart@woboq.com>
GCC warned:
tst_qabstractslider.cpp:858:89: warning: dereferencing type-punned pointer will break strict-aliasing rules [-Wstrict-aliasing]
Qt::Orientation orientation = *reinterpret_cast<Qt::Orientation*>(&sliderOrientation);
^
tst_qabstractslider.cpp:867:72: warning: dereferencing type-punned pointer will break strict-aliasing rules [-Wstrict-aliasing]
orientation = *reinterpret_cast<Qt::Orientation*>(&wheelOrientation);
^
The solution, of course, would be to use a static_cast here, but
why go via int in the first place? Qt::Orientation can perfectly
well be used in QFETCH, as proven by tst_qmainwindow, among other
things.
Change-Id: I97916a50405e16d114837bc52580ce6666d74b17
Reviewed-by: Olivier Goffart (Woboq GmbH) <ogoffart@woboq.com>
Restore Qt::AA_DontUseNativeDialogs that is disabled in the beginning
of the test function.
Change-Id: I4ff8eab4ecc458c478337824e66b5a59fbdd7c65
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@qt.io>
... and move toString() implementations there which were repeated in
network tests, or #ifdef'ed in qtest.h.
Since the functions moved from network tests are now in a public header,
had to massage them a bit to pass headersclean:
- replace Q_FOREACH with C++11 range-for
- avoid implicit conversion from QByteArray -> const char* (done by
re-using toString(QByteArray) instead of calling strdup() manually)
Also made the functions overloads instead of specializations. This
allows to pass the enum by value instead of by const-&.
Like the existing QHostAddress, the newly-added toString() overloads are
marked as \internal. We can decide later whether to turn them into
public API.
Change-Id: I8c23db7a0a6575273567017d42d7b2a957acece8
Reviewed-by: Jędrzej Nowacki <jedrzej.nowacki@theqtcompany.com>
Reviewed-by: Timur Pocheptsov <timur.pocheptsov@theqtcompany.com>
After e109b8a0f3, it is possible
that the cache will be flushed as a result of inserting a new
font rather than just when the timer event triggers. When doing
superscript and subscript text layouts, we would first get
a regular font engine, then a scaled one, and then reference
the regular font engine *after* getting the scaled one. If the
regular font engine was deleted as a result of inserting the scaled
one, we would get a dangling pointer and crash.
The situation was improved by 49926bb9ef.
You would now to switch between 256 different fonts in the layout
in order to trigger it. The test in the commit will trigger the
crash even with this change.
[ChangeLog][Qt Gui][Text] Fixed a crash that could happen if you
were doing many different text layouts with different fonts
and superscript or subscript alignment.
Task-number: QTBUG-53911
Change-Id: Ia33108252e030eff25924ef1b7c10b9d59b5bc8c
Reviewed-by: Lars Knoll <lars.knoll@qt.io>
Nested namespaces are quite common, therefore moc should support them.
Task-number: QTBUG-55415
Change-Id: I756cab36d498eb4342b402d255836d5d30f07b30
Reviewed-by: Olivier Goffart (Woboq GmbH) <ogoffart@woboq.com>
We need to delete the style returned from QStyleFactory::create()
ourselves, so put them into a QScopedPointer.
The alternative would have been to create this once, as a member
of tst_QWidget, but this is the minimal approach that ensures
behavior just as the old code, but without the leak.
Change-Id: I527f1031c57be6f05942f4acc057e7dae1af2571
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Found by UBSan:
tst_qapplication.cpp:1754:48: runtime error: member access within address 0x7ffda11f2220 which does not point to an object of type 'SpontaneousEvent'
0x7ffda11f2220: note: object is of type 'QMouseEvent'
The code attempted to model the layout of a QEvent with another class
that allows public access to the memory location that (hopefully)
corresponds to QEvent::spont, gaining access by casting a QEvent
object to that specifically-crafted class.
Fix by the using the existing QSpontaneKeyEvent::setSpontaneous()
call, which, despite its name, works for all QEvent subclasses, and
which has already been fixed to not invoke UB (in bc087db).
Change-Id: I7db8b8a8a823f7d61ab17375142d19dc3874fea5
Reviewed-by: Olivier Goffart (Woboq GmbH) <ogoffart@woboq.com>
Adds two formats that does not have optimized code-paths in qdrawhelper
to ensure the generic path has coverage.
This has already uncovered one bug fixed before this patch could go in.
Change-Id: I0e0a1a873555b27f6438f69a76982b8e06263dcf
Reviewed-by: Marc Mutz <marc.mutz@kdab.com>
QDateTimeEdit ignores the time-spec of its date-time value, using its
own time-spec instead; mostly, this works because it first conforms
the value to its own time-spec. However, during construction, before
doing this, it set up its display data, which could leave it with a
different time (rather than a different representation of the given
time) than it was asked to use.
Moved the updateTimeSpec() calls to immediately after setting value in
QDateTimeEditPrivate::init() to ensure correct handling. Added test.
Task-number: QTBUG-54781
Change-Id: I3b07c10997abb858fc0b40558bff96e3fdabbd83
Reviewed-by: Jesus Fernandez <jesus.fernandez@qt.io>
Reviewed-by: Marc Mutz <marc.mutz@kdab.com>
1. QSslSocketBackendPrivate::transmit was ignoring 'readBufferMaxSize';
as a result, we can have a user trying to set read buffer's size
to a small value (and more important - reading slowly in a small
chunks from this socket), but SSL itself socket reading 'too fast',
potentially growing its internal buffer to a huge size. This also
results in auto-tests failing - whenever we're trying to limit read
rate in some test.
2. Update qsslsocket auto-test.
Task-number: QTBUG-43388
Task-number: QTBUG-55170
Change-Id: Iedece26df0ac5b3b7cad62cc8c98aedc28e7ca5b
Reviewed-by: Richard J. Moore <rich@kde.org>
