To handle network events, QEventDispatcherWin32 uses I/O model
based on notifications through the window message queue. Having
successfully posted notification of a particular event to an
application window, no further messages for that network event
will be posted to the application window until the application
makes the function call that implicitly re-enables notification
of that network event. With these semantics, an application need
not read all available data in response to an FD_READ message:
a single recv in response to each FD_READ message is appropriate.
If an application issues multiple recv calls in response to a
single FD_READ, it can receive multiple FD_READ messages
(including spurious).
To solve this issue, this patch always disables the notifier
after getting a notification, and re-enables it only when the
message queue is empty.
Task-number: QTBUG-46552
Change-Id: I05df67032911cd1f5927fa7912f7864bfbf8711e
Reviewed-by: Joerg Bornemann <joerg.bornemann@theqtcompany.com>
There was a small amount of time between the last readDatagram() call
and disabling a read notifier in case the socket had a pending
datagram. If a new datagram arrived in this period, this qualified as
absence of a datagram reader. Do not change the read notifier state
because it is disabled on canReadNotification() entry and always enabled
by the datagram reader.
Thanks to Peter Seiderer, who investigated the same: "Querying
hasPendingDatagrams() for enabling/disabling setReadNotificationEnabled()
is racy (a new datagram could arrive after readDatagam() is called and
before hasPendingDatagrams() is checked). But for unbuffered sockets the
ReadNotification is already disabled before the readReady signal is
emitted and should be re-enabled when calling read() or readDatagram()
from the user."
However, this patch does not completely solve the problem under Windows,
as the socket notifier may emit spurious notifications.
Task-number: QTBUG-46552
Change-Id: If7295d53ae2c788c39e86303502f38135c4d6180
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@theqtcompany.com>
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Disable input and cursor for QGuiApplication instances used in
autotest to initialize it properly.
Change-Id: I78dc9b776269c082c20f244a51f858289129275d
Reviewed-by: Laszlo Agocs <laszlo.agocs@theqtcompany.com>
An invalid mime magic definition could lead to an assert. Replaced with
a qWarning. Move all checking to the QMimeMagicRule constructor, and do
keep invalid rules since they are need to parse child rules.
Unit test added, with QTest::ignoreMessage when using the XML backend
(there's no warning from update-mime-database when using the cache).
Also make it easier to add more shared mime info files for tests.
Task-number: QTBUG-44319
Done-with: Eike Ziller <eike.ziller@theqtcompany.com>
Change-Id: Ie39a160a106b650cdcee88778fa7eff9e932a988
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
At the edge of the view, a line segment could end up as not producing
any pixels even if not clipped by the floating-point clip
routine. Make sure the starting point for the next line is still
updated correctly for any significant segment lengths.
Change-Id: I381a4efb81ce6006f3da4c67abf279aea79e4663
Reviewed-by: Allan Sandfeld Jensen <allan.jensen@theqtcompany.com>
If Qt is not running its own event loop (e.g. if Qt is a plugin running
in a non-Qt host application with its own event loop, a call to
sendPostedEvents() should process all events by default, and not depend
on the flags passed to the last call to processEvents()
We also modify sendPostedEvents() to call its base implementation instead
of directly calling QCoreApplication::sendPostedEvents(). (The behavior of
the base implementation is the same, so no behavior change there).
This also adds a test for QWindow event handling without Qts event loop is
running. This is a black box test, just to ensure that basic functionality
is working. It can be extended later.
Task-number: QTBUG-45956
Change-Id: I7d688c0c6dec5f133fb495f07526debdde5389af
Reviewed-by: Jørgen Lind <jorgen.lind@theqtcompany.com>
QDateTime::addSecs needs to do something similar, but not identical
because it needs the number of days too. And then there are daylight
savings transitions...
Task-number: QTBUG-47717
Change-Id: I7de033f80b0e4431b7f1ffff13f976f4f5e5a059
Reviewed-by: Allan Sandfeld Jensen <allan.jensen@theqtcompany.com>
Playing with the offset argument of pcre_exec is not equivalent to
adjusting the pointer to the subject string. In particular, PCRE
can go behind the offset to check for lookbehinds or "transition"
metacharacters (\b, \B, etc.).
This made the code that deals with QStringRefs not matching in behavior
with the corresponding code dealing with QStrings. For instance,
QString subject("Miss");
QRegularExpression re("(?<=M)iss");
re.match(subject.mid(1)); // doesn't match
re.match(subject.midRef(1)); // matches!!!
Instead, actually adjust the pointer to the subject string so that
the behavior is identical. A broken test that relied on the
equivalence is also removed.
Change-Id: If96333241ef59621d7f5a6a170ebd0a186844874
Reviewed-by: Volker Krause <volker.krause@kdab.com>
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Ensure QAction::setVisible() is
handled and reposition if visibility changes.
Task-number: QTBUG-39660
Change-Id: I14f0659aedc8dc89ddef3159d3a500b40b1563ff
Reviewed-by: David Faure <david.faure@kdab.com>
We don't replay mouse events after closing popups on X11.
This leads to the bug when the menu doesn't show after
clicking three times on the menu bar. It can be fixed by
reverting 78d7192338, but
then we need an alternative fix for QTBUG-32807 on Windows.
So don't replay mouse events for the menu bar on all
platforms.
Change-Id: I3db8e24a6de6f35b0a17dffac6a131f1cad42e6d
Reviewed-by: Shawn Rutledge <shawn.rutledge@theqtcompany.com>
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@theqtcompany.com>
Reviewed-by: Jørgen Lind <jorgen.lind@theqtcompany.com>
Give setStandardOutputFile2 a sensible name, move it to where it
belongs and remove bogus Q_OS_WINCE ifdef.
Change-Id: I5c843e8b6cb626979966f3e61f7a7c720173bb28
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@theqtcompany.com>
Do not call bytesAvailableInChannel if the source pipe end is
invalid. This is the case when redirecting channels on Windows.
The assertions in bytesAvailableInChannel were triggered whenever
an output process or output file was set and waitForBytesWritten
was called.
Task-number: QTBUG-45548
Change-Id: I225dfea2c5e27e122f75008a3a06d425554e00fe
Reviewed-by: Alex Trotsenko <alex1973tr@gmail.com>
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@theqtcompany.com>
The generated xml file is now lowercase.
This was changed in shared-mime-info 3805d0bcf2.
It led to runtime warnings "No file found for ...", which helped notice the bug.
Change-Id: I31f0fc7f0fe8a098c3f79c0bcbeeb1909d2cc05a
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
The buffer may have been left dirty if we were unable to write all the
data to the child process in the previous run. So ensure we clear it
before starting a new one. We already did that for stdout and stderr,
for some reason.
Task-number: QTBUG-44517
Change-Id: I1a800c709d3543699131ffff13c419da3bbffacf
Reviewed-by: Kai Koehne <kai.koehne@theqtcompany.com>
Reviewed-by: Joerg Bornemann <joerg.bornemann@theqtcompany.com>
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@theqtcompany.com>
It failed on Windows due to readonly files copied from the resource,
until adding a setPermission call.
Change-Id: I1d42b53763583aca73d011e0f2bbf061ef6aa891
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@theqtcompany.com>
On Windows and OS X, where QStandardPaths does not use XDG_DATA_DIRS/
_HOME and shared-mime-info is not installed, the tests that require
additional shared mime info xml files were never run.
Mend that by using QStandardPaths' test mode instead of setting
XDG_DATA_HOME.
Change-Id: I53b75c293c41c4dac63986dcb88972c2b54d5428
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@theqtcompany.com>
Reviewed-by: David Faure <david.faure@kdab.com>
The test is very flakey in 5.5 integrations and the OS X CI machines have
notorious problems with networking stability. So the previously listed tests
are not really at fault, it's an infrastructure problem, that we choose to
ignore for the time being.
Change-Id: I7fbfa7b3778daa6b5e60d95b822847c92927122f
Reviewed-by: Richard J. Moore <rich@kde.org>
Reviewed-by: Lars Knoll <lars.knoll@theqtcompany.com>
The former didn't account for ascent and descent.
Change-Id: If741f22f7e79ac3c13e58f2966358010d9f9ec81
Reviewed-by: Frederik Gladhorn <frederik.gladhorn@theqtcompany.com>
[ChangeLog][Important Behavior Changes] QDebug output for QStrings
changed compared to Qt 5.5.0 to more closely match the output of
previous Qt versions. Like Qt 5.5.0, QDebug will escape non-printable
characters, the backslash and quote characters, but will no longer
escape the printable characters.
Task-number: QTBUG-47316
Change-Id: I52dd43c12685407bb9a6ffff13f62ef68cbc80c5
Reviewed-by: Kai Koehne <kai.koehne@theqtcompany.com>
Otherwise, on some systems (Windows), we'll always create the same dirs.
Change-Id: Id3d5c7bf4d4c45069621ffff13f79ba91e0f974b
Reviewed-by: Christopher Adams <chris.adams@jollamobile.com>
[ChangeLog][QtCore][QLoggingCategory] Fixed behavior of default
severity passed to constructor or Q_LOGGING_CATEGORY with regards to
QtInfoMsg, which was previously treated as being more severe than
QtFatalMsg.
This is because the code was using the numeric value of QtMsgType as a
proxy for severity (via the <= operator), but the value of QtInfoMsg is
greater than QtFatalMsg. Instead, the severity ordering must be dealt
with explicitly.
Change-Id: I5f178afc735221b00cb67c2cea4fa964bd9079ce
Reviewed-by: Kai Koehne <kai.koehne@theqtcompany.com>
Build it only in -developer-build mode for tests that might depend
on exact-matching font behavior.
Return earlier to avoid doing any useless job.
Change-Id: I966ee5689f03403e45f4c957b63e3113f0467803
Reviewed-by: Eskil Abrahamsen Blomfeldt <eskil.abrahamsen-blomfeldt@theqtcompany.com>
The QSslCertificate tests only covered certificates with RSA keys, this
extends the test coverage to DSA and EC keys.
Change-Id: Ibee26f449cf6c1d97cbac6b511972eb44d6f0bd2
Reviewed-by: Richard J. Moore <rich@kde.org>
All overloads of QProcess::start will now check whether the program
string is empty and in that case
- set error to FailedToStart,
- set errorString to "No program defined",
- emit error.
Until now only one of the three overloads behaved like this.
As a side effect, start(QString(), QStringList()) will not crash on
Windows anymore.
Task-number: QTBUG-47404
Change-Id: I2f93657204fe3643b1d74a74817843c05fc4a96b
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@theqtcompany.com>
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Reviewed-by: Jake Petroules <jake.petroules@petroules.com>
Reviewed-by: Kai Koehne <kai.koehne@theqtcompany.com>
The comment about non-OpenSSL backends not reproting a specific error
for self-signed certificates contained a typo, this fixes it.
Change-Id: I3010981d5d87d68ebf5e984c003b8bbbfb019b96
Reviewed-by: Timur Pocheptsov <Timur.Pocheptsov@digia.com>
Non-OpenSSL backends are not able to report a specific error code
for self-signed certificates.
Change-Id: I56bf130335b2afa65cf2bd5248a40ac0e32f74c2
Reviewed-by: Timur Pocheptsov <Timur.Pocheptsov@digia.com>
Reviewed-by: Richard J. Moore <rich@kde.org>
Set QFileDevice::WriteUser on all files extracted from resources. These
are read-only, which is preserved by QFile::copy(). This caused the
deletion of the temporary directory to fail on Windows.
Change-Id: Id99de9160471c38bcec68025c89cfabbe209bdbe
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@theqtcompany.com>
Reviewed-by: David Faure <david.faure@kdab.com>
Otherwise, if I open tests/auto/dbus/dbus.pro in Qt Creator, it shows me
"test", "test2", "test3", "test4" and it's very hard to know which test
is which.
Change-Id: Iee8cbc07c4434ce9b560ffff13d0654696c025b7
Reviewed-by: Frederik Gladhorn <frederik.gladhorn@theqtcompany.com>
I'm getting an error with WSARecvFrom in nativeBytesAvailable() and I
don't know why. The number of bytes obtained is correct and the test
works for 2 bytes, so let's use that.
The Windows nativeBytesAvailable() function has a comment saying that
WSAIoctl sometimes indicates 1 byte available when there's a pending
error notification, so that function proceeds to do a WSARecvFrom to
peek the number of bytes. Somehow that function in this test is getting
a SOCKET_ERROR I can't explain.
Change-Id: Ic5b19e556e572a72a9df9a405b1fee3b7efb8b24
Reviewed-by: Richard J. Moore <rich@kde.org>
On Windows, having read-only files in a directory can cause removal
to fail. When file deletion fails, check on the permissions, set
write permissions and retry.
Split apart code paths by OS in tst_QDir::removeRecursivelyFailure();
deletion of the read-only directory on UNIX should still fail.
Change-Id: I36e54be5229a7b552e90fd5f42722b868fa0b6ee
Reviewed-by: David Faure <david.faure@kdab.com>
The test showed failures on OS X:
FAIL! : tst_QListView::batchedMode() Compared values are not the same
Actual (ba.size()): 2
Expected (3) : 3
Loc: [tst_qlistview.cpp(848)]
Use QTRY_COMPARE() to count the number of visible indexes with
a helper function instead of using hard-coded timeouts. Item 3
appears with a little delay on OS X.
Change-Id: I2fb2ff5ebdf9dbe85bdc79401375ad6f47b7b12b
Reviewed-by: Simon Hausmann <simon.hausmann@theqtcompany.com>
The test checks whether a child window receives
mouse events synthesizes from touch. Enlarge the child
window so that it fully covers the parent and move the
touch point a bit inside so that it is not affected
by window manager positioning issues.
Use QTRY_VERIFY.
FAIL! : tst_QApplication::touchEventPropagation() 'widget.seenMouseEvent' returned FALSE. ()
Loc: [/work/build/qt/qtbase/tests/auto/widgets/kernel/qapplication/tst_qapplication.cpp(2107)]
Change-Id: Ic08e68b1e547cc7148cd8994464fdc2a14ac507b
Reviewed-by: Simon Hausmann <simon.hausmann@theqtcompany.com>
The test sends touch events to the root item at the top left
corner which fails if the views starts to scroll.
Make the view sufficiently large to prevent scrolling and align
at top left.
FAIL! : tst_QTouchEvent::touchBeginWithGraphicsWidget() Compared values are not the same
Actual (((root->touchBeginCounter))): 0
Expected (1) : 1
Loc: [/work/build/qt/qtbase/tests/auto/gui/kernel/qtouchevent/tst_qtouchevent.cpp(1471)]
Change-Id: I357322ccc809ddb5cb587febf3c75cbe497e59d8
Reviewed-by: Simon Hausmann <simon.hausmann@theqtcompany.com>
An object that throws in its constructor cannot be reentered. This
violates both C++11 and C++98. It's also a regression from MSVC 2013.
The unit test is renamed to indicate what it really does, as opposed to
a misleading name that was probably a "thinko" on my part.
Task-number: QTBUG-47224
Change-Id: Ib306f8f647014b399b87ffff13f132436d0578ef
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@theqtcompany.com>
QTemporaryFileEngine does not store the pattern, so it needs to get it
again from QTemporaryFilePrivate prior to reopening the file. It's
possible to lose the pattern when remove() is called on the object.
Task-number: QTBUG-46156
Change-Id: I66a35ce5f88941f29aa6ffff13dfc7f83d4fa3a2
Reviewed-by: Olivier Goffart (Woboq GmbH) <ogoffart@woboq.com>
Reviewed-by: David Faure <david.faure@kdab.com>
This is a fix-up for cff39fba10.
That patch lead to some internal state issues that lead to the QTBUG-47048
or to QNetworkReply objects erroring with "Connection Closed" when
the server closed the Keep-Alive connection.
This patch changes the QNAM socket slot connections to be DirectConnection.
We don't close the socket anymore in slots where it is anyway in a closed state
afterwards. This prevents event/stack recursions.
We also flush QSslSocket/QTcpSocket receive buffers when receiving a disconnect
so that the developer always gets the full decrypted data from the buffers.
[ChangeLog][QtNetwork] Fix HTTP issues with "Unknown Error" and "Connection Closed"
[ChangeLog][QtNetwork][Sockets] Read OS/encrypted read buffers when connection
closed by server.
Change-Id: Ib4d6a2d0d988317e3a5356f36e8dbcee4590beed
Task-number: QTBUG-47048
Reviewed-by: Kai Koehne <kai.koehne@theqtcompany.com>
Reviewed-by: Richard J. Moore <rich@kde.org>
On Windows, t->timeout was updated only once, at creation time, so the
timer would always show as "overdue" after the first activation.
The timer is updated to indicate the full remaining time during the slot
activation, which is the behavior of the Unix and Glib dispatchers.
Task-number: QTBUG-46940
Change-Id: I255870833a024a36adf6ffff13ecadb021c4358c
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@theqtcompany.com>
Reviewed-by: Joerg Bornemann <joerg.bornemann@theqtcompany.com>
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@theqtcompany.com>
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
