Convert the function to be data-driven, then add cases for ranges with
limits INT_MIN, INT_MAX and zero to check for overflows in the
painting code. Speed up the function by removing the per-iteration
qWait() call and rely solely on the existing QTRY_VERIFY() to do the
right thing.
Change-Id: I6d2a2c2b7637fca7ddb2a9adc5f6550f8255da14
Reviewed-by: David Faure <david.faure@kdab.com>
Systematic testing in tst_QStringApiSymmetry revealed a bug in
QStringRef::toLatin1(): a null input did not result in a null output,
but an empty one. This is fixed, for consistency with
QString::toLatin1(), and QString(Ref)::toUtf8(), which behaved
correctly already.
The same bug was found in QString(Ref)::toLocal8Bit(), which is
particularly hideous, as it's documented to fall back to toLatin1(),
which preserves null inputs. Fixed, too.
[ChangeLog][QtCore][QString] toLocal8Bit() now preserves nullness of
the input QString (outputs null QByteArray).
[ChangeLog][QtCore][QStringRef] toLocal8Bit() and toLatin1() now
preserve nullness of the input QStringRef (output null QByteArrays).
Change-Id: I7026211922c287e03d07e89edbad2987aa646e51
Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
After it started to fail (somehow it's only OpenSUSE 42.1) again and again
and after a quick re-evaluation it appears the logic testing SETTINGS|ACK
is incorrect. We (client side) start by sending the preface and then
continue to send our request(s). The other side (server) starts from sending its
SETTINGS frame. These settings must be ACKed, but apparently it can happen,
that server receives a requests and sends a reply before it receives SETTINGS|ACK,
resulting in replyFinished (replyFinishedWithError) signal and event loop stopping.
As a result - QVERIFY(serverGotSettingsACK) fails.
Task-number: QTBUG-58758
Change-Id: I8184cf459b2b88f70c646171e0115c184237fad1
Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
tst_QDateTime::operator_insert_extract() was setting the time-zone and
taking care to restore it at the end of the test; however, if the test
were to fail, the restore would be skipped. Package the zone-setting
and restore in a class instance, so that premature return can't bypass
the restore.
Change-Id: I3df63260da17e481ef4d0d107d9f0fdea3e147e7
Reviewed-by: Marc Mutz <marc.mutz@kdab.com>
We keep the runnables from finishing by having them block
on a QSemaphore::acquire() call inside run().
If we fail a test that precedes the call to sem.release()
further into the test, the early return will cause the
thread pool to be destroyed, which will then attempt to
wait for the runnables to finished, which, in turn wait
for the semaphore to be released.
-> dead lock
Fix by introducing a RAII object to release the semaphore
with a sufficiently large number to unblock all runnables.
That number will in some situations be too large, but that
does not matter.
Change-Id: I1ec7e29b37bc36309e93e6e30708cc7db3c9579c
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
In order to get reproducible runs of the test, we need to
wait in the main thread until all runnables have started
executing. Otherwise, what the cancel() loop below actually
does will vary from run to run.
Change-Id: Ib912b0943e7bbd55c9480ae6fd4011ba20ac457e
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Instead of allocating a statically-sized array on the heap,
use an automatic C array instead.
Replace some magic numbers with named constants.
Change-Id: I17d29a76a67c4a413453ac26a5dee8cd54a8a37d
Reviewed-by: David Faure <david.faure@kdab.com>
Manipulating a simple int from multiple threads is a data race,
thus undefined behavior.
Fix by using QAtomicInt and atomic operations instead.
Change-Id: I5418bc260da57fe353a71b8e5c7c1c97adbe7597
Reviewed-by: David Faure <david.faure@kdab.com>
This should reduce the failure rate. We're still doing qSleep of the
same amount of time, but we now only fail if the slip is over 100 ms.
Task-number: QTBUG-58713
Change-Id: I536c32a88bff44dab37afffd14a1afdf0b2e522a
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@qt.io>
Reviewed-by: Marc Mutz <marc.mutz@kdab.com>
currencySymbol == "$" does not mean negative currency values will be formatted
as "($value)". With all locales I have on my mac machines (10.11/10.12) the
result is different from what this test expects. Also, the results are very
different for different locales. Apparently, we never saw this problem before
since in our CI "macs" we never have US Dollar/en_US selected in System Preferences.
Task-number: QTBUG-58784
Change-Id: Ic2c3a3172bf1e715e99092ddee8f461b216d995a
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
The runUpdateMimeDatabase() call was timing out on the CI due to
running out of the default 30s timeout for QProcess::waitForFinished()
(on my machine, that call takes less than 0.5s, reliably...).
Increase to two minutes.
Change-Id: I61b2e060ea9c2508b853847ba7040ad499e0084c
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@qt.io>
Reviewed-by: David Faure <david.faure@kdab.com>
When one of the QCOMPAREs in Consumer::run() fails, the consumer
returns early, leaving the producer deadlocked in a QSemaphore's
acquire() call. Change these to tryAcquire() with a large timeout,
so the producer, too, eventually leaves run().
Change-Id: I7421d43305decd4754e09c8e092363594d1be06b
Reviewed-by: David Faure <david.faure@kdab.com>
This test was determined to be flaky on the CI.
Task-number: QTBUG-58741
Change-Id: I43196d3a27f726fb96b427f5071e726b571a0404
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@qt.io>
Reviewed-by: David Faure <david.faure@kdab.com>
This is a simple RAII class that makes semaphore releasing
reliable in the face of exceptions and early returns.
This code originates from KDTools' KDSemaphoreReleaser[1], but
has been extensively reworked to support C++11 move semantics.
[1] https://docs.kdab.com/kdtools/2.3.0/class_k_d_semaphore_releaser.html
[ChangeLog][QtCore][QSemaphore] Added a new RAII class, QSemaphoreReleaser,
to reliably perform release() calls.
Change-Id: I6aff64d37cc0882b17c4419817bde60b542f34d9
Reviewed-by: Mitch Curtis <mitch.curtis@qt.io>
Reviewed-by: David Faure <david.faure@kdab.com>
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
The verifySessionProtocol() method in the SecureTransport backend did not
properly handle TlsV1_0OrLater, TlsV1_1OrLater and TlsV1_2OrLater.
This commit teaches verifySessionProtocol() about them.
It also adds TlsV1_0OrLater, TlsV1_1OrLater and TlsV1_2OrLater to the
protocolServerSide() test in tst_qsslsocket.
Change-Id: I394766c75dbea4d89bf3d6b9c3f3c89530ac0749
Reviewed-by: Timur Pocheptsov <timur.pocheptsov@qt.io>
We can defer the creation until the window is shown.
Change-Id: I3d5b45ae59ee0925996cf12cd46dd574c8c6ef95
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@qt.io>
Add proper support for 'so' and 'bundle' suffixes. Qt wrongly assumes
.so libraries are not versioned on Apple platforms, which is
wrong. Also, the shared library .bundle which is what Apple recommends
instead of .so, are also versioned (not to be confound with the
different Core Foundation bundles, which are directory hierarchy).
For more info, see http://docstore.mik.ua/orelly/unix3/mac/ch05_03.htm.
Especially the part that reads:
"Loadable modules, called bundles in Mac OS X, have the file type
MH_BUNDLE. Most Unix-based software ports usually produce bundles with a
.so extension, for the sake of consistency across platforms. Although
Apple recommends giving bundles a .bundle extension, it isn't
mandatory."
Task-number: QTBUG-50446
Change-Id: Iacd5136397a12d65d83821434f332eb602550b4b
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Vectorized versions of ucstrncmp work on larger chunks of text
(typically 8 characters and an optional 4 more in some cases), so there
are now 4 extra sets of tests:
- strings of 1-65 characters, all different
- strings of 1-65 characters, all the same
- strings of 1-65 characters, all the same *except* the last character
- strings of 16 characters long, all the same except one, and that one
is different for every string (i.e. first string differs in first
char, second in second char, etc)
This should excercise both 1 or more iterations of the vectorized loop,
the detection logic inside the loop, and off-by-one cases.
The input is all ascii, so the ::compare() test will run them for both
QChar-QChar comparisson and for QChar-latin1 comparisson.
Change-Id: Ifaa7e019c63b581d4af5aef6dcfb3e7456c7d360
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
This reverts commit b0c1e07d64.
The unit tests it introduced trigger errors with the macOS and
Win32 implementations, which means the there's something seriously
wrong with either the API, our implementation or the tests.
Revert for now until this gets fixed. Blacklisting was also the
wrong tool to use. The tests should have used QEXPECT_FAIL.
Change-Id: Ida20c6bbe0c019835a22464535029585e8e1e367
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
There weren't any, at all.
Testing on the CI showed that the implementation is broken on
macOS, and, to a lesser extent, on Windows, so blacklist the
failing tests until the implementation can be fixed. No need
to hold back testing the other implementations.
Task-number: QTBUG-58737
Change-Id: I9ae16ab778dbe2e95a6ca5e0bae00df4bad65cb2
Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
This test was determined to be flaky on the CI, depite attempts
to stabilize it (b750a3786f).
Task-number: QTBUG-58745
Change-Id: I933199cd537002699906147d172bb797f1dc90c1
Reviewed-by: Allan Sandfeld Jensen <allan.jensen@qt.io>
Since 5.7, QIODevice::peek() implementation is based on transaction
mechanism. While technically it's correct, seeking backward on a
buffered random-access device clears the internal buffer that affects
the performance of reading.
To solve the problem, this patch implements peek mode directly inside
the reading procedure.
Task-number: QTBUG-56032
Change-Id: Ic5269f76e44c491a0309e13aba87fa7cf7b9259f
Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
The controls need to reflect the facts that e.g. maximized and fullscreen
can both be set at the same time, the same way a window can be minimized
and fullscreen.
Change-Id: I7f3e354a5efaefb9f51e6b1f24fa35980fe35899
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@qt.io>
Makes operations like minimizing a window happen "behind the scenes"
because the window is hidden during the state change.
Change-Id: I01a00661e57f2dcfa6aef78ee0cfa36fbed1bb03
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@qt.io>
This test was determined to be flaky on the CI.
Task-number: QTBUG-58713
Change-Id: Ie6e6a69b8ea625e3a3102c88d52f1f0fbec242aa
Reviewed-by: Lars Knoll <lars.knoll@qt.io>
I didn't even try to understand what the old code was trying to do;
once you're told by a user that the code is wrong, you see that it is.
Fixed by just using the row as passed to takeRow() instead of trying
to do some storage-index calculations. The m_matrix indexing operator
does it all for us.
Added a test that checks that the expected field widget gets
returned. Fixed expected test data that was wrong, and just checking
that the implementation behaves as implemented, instead of as
documented.
Amends change 8fbae648db.
[ChangeLog][QtWidget][QFormLayout] The functions takeRow() and
removeRow(), new in 5.8.0, now take and remove the correct row.
Task-number: QTBUG-58693
Task-number: QTBUG-15990
Change-Id: I7185ccbc6c03e2579741cad5c0c821d3ed165474
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@qt.io>
When a non-QString-backed mode (via the (QChar*, int) ctor) was added
for Qt 4.5, the author forgot to adjust the setCaseSensitivity()
function. It still uses q_pattern instead of (p.uc, p.len) as the
pattern for which to create the skip-table. Since there is no
setPattern() overload for this mode, the correctness of the matcher is
not harmed by this, but its performance degrades to that of a linear
scan: the skip-table, being filled from an empty pattern, will be
all-zeros, sending bm_find() into the 'possible match' case at every
character.
Since matching is still correct, but slow, it's not possible to write
a test for this. I did, however, leave my attempts in the auto-test,
for when we add QStringView overloads of setPattern() which will then
be able to expose the bug.
Change-Id: I7b803e8624b0352a0a974900affbbfc0c260d93b
Reviewed-by: Giuseppe D'Angelo <giuseppe.dangelo@kdab.com>
Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
The setData method of an item view would get an incorrect value of a
QDialog's result. This patch changes the order of functions called to
fix that.
[ChangeLog][QtWidgets][QDialog] Fixed a bug where accessing the result
of QDialog's result could yield an incorrect value in some situation
like using it as a delegate for item views.
Task-number: QTBUG-6018
Task-number: QTBUG-12156
Task-number: QTBUG-14430
Change-Id: I6ee4b6e8cacf6a806631c05c6c5dbcff925df65e
Reviewed-by: Shawn Rutledge <shawn.rutledge@qt.io>
Reviewed-by: Jesus Fernandez <Jesus.Fernandez@qt.io>
Reviewed-by: Timur Pocheptsov <timur.pocheptsov@qt.io>
Remove the line setting the DPR from the source; the image
is moved.
Task-number: QTBUG-58653
Task-number: QTBUG-58645
Change-Id: I2de94681459dba1d69dee06da44617fb9fa35bcc
Reviewed-by: Allan Sandfeld Jensen <allan.jensen@qt.io>
As explained in
https://blogs.kde.org/2009/03/26/how-crash-almost-every-qtkde-application-and-how-fix-it-0
creating dialogs on the stack is a bad idea if the
application or the dialog's parent window can be closed
by means other than user interaction (such as a timer or
an IPC call). Since we cannot know whether Qt is used to
build such an application, we must assume it is, create
the dialog on the heap, and monitor its lifetime with a
QPointer.
Instead of using manual resource management, add a
minimal implementation of QAutoPointer, and use that in
all static get*() functions.
Task-number: QTBUG-54693
Change-Id: I6157dca18608e02be1ea2c2defbc31641defc9d1
Reviewed-by: Giuseppe D'Angelo <giuseppe.dangelo@kdab.com>
Reviewed-by: David Faure <david.faure@kdab.com>
When a QFile object is reused, the atEnd() method may return incorrect
values. The reason for this is that QFileDevicePrivate::cachedSize is
not cleared. Setting cachedSize = 0 in the close() method fixes this issue.
Task-number: QTBUG-57698
Change-Id: I828a2cf844e98d581098f2c781fa47d2cd3275ce
Reviewed-by: Alex Trotsenko <alex1973tr@gmail.com>
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@qt.io>
Reviewed-by: Lars Knoll <lars.knoll@qt.io>
At 200 ms, the error on first firing could be 10 ms.
Task-number: QTBUG-58519
Change-Id: Ifaee7464122d402991b6fffd14a02a4ce782f11f
Reviewed-by: Allan Sandfeld Jensen <allan.jensen@qt.io>
tst_qsortfilterproxymodel.cpp:3997:14: warning: 'data' overrides a member function but is not marked 'override' [-Winconsistent-missing-override]
tst_qsortfilterproxymodel.cpp:4004:9: warning: 'rowCount' overrides a member function but is not marked 'override' [-Winconsistent-missing-override]
accessiblewidgets.h:101:11: warning: 'interface_cast' overrides a member function but is not marked 'override' [-Winconsistent-missing-override]
accessiblewidgets.h:115:13: warning: 'textBeforeOffset' overrides a member function but is not marked 'override' [-Winconsistent-missing-override]
accessiblewidgets.h:121:13: warning: 'textAtOffset' overrides a member function but is not marked 'override' [-Winconsistent-missing-override]
Change-Id: I75ba00a0109ff25a2a22554980b8e541e661f806
Reviewed-by: Frederik Gladhorn <frederik.gladhorn@qt.io>
On macOS, the code that read the plist is using
QByteArray::fromRawCFData. When we return the data directly
we need to detach the QByteArray so that it does not point
CFData's data that will get deallocated just after the call.
Task-number: QTBUG-58531
Change-Id: If829a304b986c99c8fc2aeeb992f2d539a4eef3a
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Reviewed-by: Tor Arne Vestbø <tor.arne.vestbo@qt.io>
The pattern did not match double-digit version numbers like 5.10.0.
Adapt it to better match the version number and use QRegularExpression.
Change-Id: I2e25b247213cee6853bfeb27871c839135e3970f
Reviewed-by: Jarek Kobus <jaroslaw.kobus@qt.io>
When calculating the width of a text for drawing decorations on top,
we use the effective advance of the whole text after it has been
through the shaper.
However, in the case of QStaticText and QGlyphRun, there is shortcut:
Since we only have the glyph indexes and position of each glyph,
we use the position + advance of the right-most glyph to find the
right-most edge of the decoration. For this, however, we use the
advance of the glyph *out of context* of the rest of the string,
because the whole idea is to avoid doing the shaping of the string
with every draw call. In some rare cases, the advance of the
right-most character, in the context of the string, is different
from the advance of the standalone glyph.
Now, one way of fixing this would be to store the width of the
text in QStaticText and QGlyphRun, but since it is a very rare
artifact which is barely visible, I have opted to just work around
it in the test instead, the workaround being to force integer
metrics so that we don't get the small 0.2 pixel error.
Task-number: QTBUG-55217
Change-Id: I8d16d52f2ef27275cabb7d3865aeeaa31617ba3d
Reviewed-by: Lars Knoll <lars.knoll@qt.io>
Recent HB changed the way of handling ZWJ/ZWNJ to be more in par
with other engines.
Change-Id: I8abacd195e4b247c8fa6d91ef1086e74da0a1efb
Reviewed-by: Eskil Abrahamsen Blomfeldt <eskil.abrahamsen-blomfeldt@qt.io>
The QFontMetricsF version of the test should not truncate the
returned values, as the results may then be wrong.
Change-Id: I17f97f846bb723709e695e8866e437d6888d275b
Reviewed-by: Simon Hausmann <simon.hausmann@qt.io>
