The NEON implementation uses rsqrt and thus can not be taken on 0, so
replace the minimum with something close to zero instead of zero.
Task-number: QTBUG-59961
Change-Id: Ia39e45be675b056c1e22900495ce9ba4e8b70e5f
Reviewed-by: Erik Verbruggen <erik.verbruggen@qt.io>
This was originally added so that you could replace a T with
QAtomicInteger<T> in the same class and still keep ABI. However, for
legacy reasons, on 32-bit x86, types larger than 4 bytes keep an old
1990s alignment of only 4 bytes, but modern std::atomic<T> for those 8-
byte types enforces an alignment of 8 bytes. Therefore, the requirement
to keep alignment is not possible to guarantee.
In other words: you may not replace T with QAtomicInteger<T> or
std::atomic<T> and assume no ABI breakages in all platforms.
This is a requirement to implement atomicity. An 8-byte type aligned to
only a 4-byte boundary could cross a 16-byte boundary or, worse, cross a
cacheline boundary. Crossing the 16-byte boundary could be bad on some
processors, but crossing the cacheline boundary (addresses ending in
0x3C, 0x7C, 0xCC and 0xFC, or 4 out of 64 possible addresses or 6.25%)
is always bad: the CPUs cannot guarantee an atomic load or store
operation.
See also <https://gcc.gnu.org/bugzilla/show_bug.cgi?id=71660>.
Task-number: QTBUG-67858
Change-Id: If90a92b041d3442fa0a4fffd15283e4615474582
Reviewed-by: Frederik Gladhorn <frederik.gladhorn@qt.io>
That check is flaky on Windows. It doesn't seem to be testing Qt functionality.
I also don't see CreateFile2() documentation mentioning any guarantees that
opening the same file twice would give the same HANDLE each time.
Change-Id: Ica2e60571ae9fc39bf822803a2a9dd6add8323d7
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Reviewed-by: Joerg Bornemann <joerg.bornemann@qt.io>
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@qt.io>
When a key press is received which is not spontaneous then it needs to
be manually sent as a shortcut override event to ensure that any
matching shortcut is triggered first. This enables emulation/playback
of recorded events to still have the same effect.
[ChangeLog][QtWidgets] Sending a key press event with sendEvent() now
sends a ShortCutOverride event first to the widget to trigger any
shortcuts set first.
Task-number: QTBUG-48325
Change-Id: Iafcc2cdb1773bffe89edaeb0abc44cd5a51088e6
Reviewed-by: Richard Moe Gustavsen <richard.gustavsen@qt.io>
The drive has to be defined for every Windows configuration (also
including winrt).
Change-Id: I94a3131b8aec20cda97dc78f55b1d87aa10240e4
Reviewed-by: Maurice Kalinowski <maurice.kalinowski@qt.io>
This static method can be called before QGuiApplication is created. At
that point there is yet no primary screen, so the implementation
needs to guard against dereferencing a nullptr.
Task-number: QTBUG-67309
Change-Id: I6b7b9e97b1c3c79bf2f9c6d6247c3b10f39f7a55
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@qt.io>
To read data from a named pipe, QWindowsPipeReader uses the ReadFileEx()
function which runs asynchronously. When reading is completed and the
thread is in an alertable wait state, the notified() callback is called
by the system, reporting a completion status of that operation. Then the
callback queues a readyRead signal and starts a new sequence. The latter
is skipped if the pipe is broken or the read buffer is full.
Thus, if an application does not run the event loop, the next call to
QWindowsPipeReader::waitForReadyRead() should emit the queued signal
and report true to the caller even if no new read operation was started.
Change-Id: I37102dbb1c00191d93365bfc2e94e743d9f3962a
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@qt.io>
1. If a request was redirected or some error was encountered, we
try to reset the uploading byte-device.
2. Disconnecting from the byte-device is not enough, since we have a
queued connection, _q_uploadDataReadyRead() gets called even if
byte-device was deleted and thus sender() can return null -
we have to check this condition.
3. Update auto-test with a case where our server immediately
replies with a redirect status code.
Task-number: QTBUG-67469
Task-number: QTBUG-66913
Change-Id: I9b364cf3dee1717940ddbe50cba37c3398cc9c95
Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
This failed (occasionally) in finding the editor widget, so wait for the window
to be shown properly. Also enter the event loop with QTRY_VERIFY and not for a
fixed time of 1 s.
This however just moved the point of failure. Now the test fails since the
clearing and setting of focus does not dismiss the editor widget sometimes, so
still blacklist it.
Task-number: QTBUG-67282
Task-number: QTBUG-66216
Change-Id: Iec598609fce23a25d7b955082d0973685d612715
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@qt.io>
Reviewed-by: Gatis Paeglis <gatis.paeglis@qt.io>
Use QTRY_COMPARE instead of a wait followed by QCOMPARE. Unless the condition
doesn't change and the wait provides the opportunity for things to go wrong.
Task-number: QTBUG-63992
Change-Id: I13474a45f1f4df9d77a418729a5748235ae0dd1a
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@qt.io>
Reviewed-by: Gatis Paeglis <gatis.paeglis@qt.io>
For the time being, no more stretched QComboBox. They
were already looking bad before and nobody complained,
so it's a non-issue. In the future, we might use square
combo boxes in the same way as we do for push buttons
but for non-editable QComboBox only.
Removes what is now dead code, including some leftovers.
tst_QMacStyle updated to reflect size changes to accom-
modate the focus ring.
Change-Id: I60fac86b9acb52cc96373bca0b3cad598ec4f1e7
Reviewed-by: Morten Johan Sørvig <morten.sorvig@qt.io>
Properly handle single protocol TLS configurations. Previously,
due to the use of generic (non version-specific) client/server method
they worked as ranges of protocols instead. This also fixes a couple
of previously broken tests.
Task-number: QTBUG-67584
Change-Id: Ied23113a4fab6b407a34c953e3bd33eab153bb67
Reviewed-by: Mårten Nordheim <marten.nordheim@qt.io>
Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
This reduces the timeout from 10 seconds to the default 5, but the tested events
(resize events, painting) seem like something that should happen in 5 seconds.
Change-Id: I1d12372f37264bf7be64096e43813cd03f567102
Reviewed-by: Laszlo Agocs <laszlo.agocs@qt.io>
BuiltIn inputs shouldnt be translated into a code input. We comment the
input line for this variables
Change-Id: I3b7d2fd06afc6a122d90a999fc5a5246fc93234e
Reviewed-by: Rafael Roquetto <rafael.roquetto@kdab.com>
The test was failing because the files it expected to create were not actually
visible on the disk.
This was because 189e9c93d7 made QTemporaryFile
use unnamed files if the file name is not actually requested. Fix by forcing the
temporary file to be made with the requested name.
CI has not caught this because the test executable is marked insignificant_test
on Linux.
Change-Id: Ibe0244fdfcc66acf3be4314da51b4c0b3b520b71
Reviewed-by: Simon Hausmann <simon.hausmann@qt.io>
Except RHEL-6.6 and 7.4
It was blacklisted in f3939d943e, along
with a lot of other entries. No specifics are known about why it was
blacklisted originally, but now it only fails on RHEL because they
use OpenSSL 1.0.1.
Change-Id: I6d1d1b7b7bf5386b2115b8780163550cf03bbad7
Reviewed-by: Gatis Paeglis <gatis.paeglis@qt.io>
Reviewed-by: Timur Pocheptsov <timur.pocheptsov@qt.io>
On Wayland, manipulating the clipboard requires an input event serial, which is
not possible to get unless the compositor sends us input events.
Change-Id: If5231a5db85f8d6ad988ea93f240cee0c3466f9f
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@qt.io>
This makes the tests pass on Wayland compositors that don't automatically give
focus to newly created windows. Such as a headless Weston, which is very useful
for testing.
Task-number: QTBUG-66846
Change-Id: I502504b333499c89be193a3ebc19d41264a13580
Reviewed-by: Shawn Rutledge <shawn.rutledge@qt.io>
Reviewed-by: Jan Arve Sæther <jan-arve.saether@qt.io>
Instead of an unconditional wait.
The QSKIP is necessary because the PDF format is non-native and the Windows
implementation of QPrintDialog doesn't work then.
The macOS implementation has the same warning, so add the check for that OS
as well.
The test has previously passed, since it hasn't seen not actually opening the
dialog as an issue.
Change-Id: Ib53e378b232580358f13e0c5206404412f9e6ee2
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@qt.io>
The TZ database has recently revised its ccount of when they skipped a
day to cross the international date line, from skipping Jan 1st 1995
to skipping December 31st 1994. So Move the before-days check to
December 30th; and correct the Feb 2nd that was meant to be Jan 2nd
(and does need to remain so, for compatibility with systems with out
of date data).
Task-number: QTBUG-67497
Change-Id: I5b9483c553205817f995f91793662a5a85e03192
Reviewed-by: Liang Qi <liang.qi@qt.io>
OCIBindByPos2 is only needed when using execBatch(), binding data that
is longer than USHRT_MAX works for exec() so this is left unchanged.
Change-Id: Ifdcf91939d184f225d24c13052ea0b81611ecf91
Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
There is no need to qWait() before a QTRY_VERIFY. qWait() will also
intermittently handle events while waiting, so calling it in a loop isn't
necessary.
Change-Id: Ica7fbf18c03e673213dd9b72f31f71937cdcb145
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@qt.io>
The Q_ASSERT(mimePrivate.fromCache) at qmimedatabase.cpp:218
which I added in commit 7a5644d648, was being triggered when calling comment()
for invalid mimetypes such as db.mimeTypeForName("").
Change-Id: I8037041a4b435d2a5ba24ec94b7858e38b2f0bf2
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
If the db isn't valid, then that's the actual issue, not the fact that we're
getting the same invalid db (with the same driver QSqlNullDriver) in
multiple threads.
Change-Id: I95490818ed78e741c3823e115f139c2cff01b0b1
Reviewed-by: Andy Shaw <andy.shaw@qt.io>
Also use data-driven test to reduce duplication.
Change-Id: I9516e52267cb3c7b239030fd73dbbf23ac8f52f7
Reviewed-by: Sami Nurmenniemi <sami.nurmenniemi@qt.io>
Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
This makes an irreversible global change: tests that do it will mess
with other tests. So make sure they're all last. This required
splitting up one test; and revealed another that secretly depended on
being run with C as default locale.
Task-number: QTBUG-67276
Change-Id: Ic24ef48b2c9bd5c37c1f11260b437628019624ca
Reviewed-by: Kari Oikarinen <kari.oikarinen@qt.io>
Do not let a global qtlogging.ini interfere with an autotest.
This works around an issue on Ubuntu 17.10
Task-number: QTBUG-67385
Change-Id: I0d02835eb7a561b43fe0b98f4383c170c6d51303
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Reviewed-by: Joni Jäntti <joni.jantti@qt.io>
Uses the scripts and tests we already have for lancelot as a painting
benchmark.
Change-Id: Idf8a55e2261162e619f6dbb567dc19f8dc96da4e
Reviewed-by: Eirik Aavitsland <eirik.aavitsland@qt.io>
Instead use QSignalSpy to wait directly for the expected events.
Change-Id: I319302ea7177fe690b5d885347c505454904518e
Reviewed-by: Sami Nurmenniemi <sami.nurmenniemi@qt.io>
Reviewed-by: Erik Verbruggen <erik.verbruggen@qt.io>
It is suspected that the fault actually lies in CI infra.
Amends e3cf2a1ae9.
Task-number: QTBUG-66311
Change-Id: I967da283f0b94be1d0b99481d0cbd15ca7f98d45
Reviewed-by: Sami Nurmenniemi <sami.nurmenniemi@qt.io>
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
These two places were sort of manually implementing QTRY_VERIFY except that they
never time out.
Change-Id: I136e6c7400194327c0475c6acfc019825ccec1b5
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@qt.io>
Reviewed-by: Olivier Goffart (Woboq GmbH) <ogoffart@woboq.com>
Reviewed-by: Sami Nurmenniemi <sami.nurmenniemi@qt.io>
Use QSignalSpy::wait or QTRY_VERIFY instead. This shaved off ~200 ms of the
running time of the test and is more reliable.
Some unconditional qWait()s still remain in this test. They are giving an
opportunity for the wrong thing to happen and thus are not waiting for any
specific condition to be fulfilled.
Task-number: QTBUG-63992
Change-Id: I25a4470fe8d6a5b8b5039b3ed77321d24faa1707
Reviewed-by: Morten Johan Sørvig <morten.sorvig@qt.io>
Reviewed-by: Sami Nurmenniemi <sami.nurmenniemi@qt.io>
Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
Use QTRY_VERIFY instead of hard-coded wait for the tooltip to appear.
Also use QTRY_VERIFY to verify that no top level widgets are left over
to account for tooltips and effect windows.
Change-Id: Ia9835fdc480c6abb034f6fc4ad3d6b32751ee536
Reviewed-by: Sami Nurmenniemi <sami.nurmenniemi@qt.io>
Reviewed-by: Kari Oikarinen <kari.oikarinen@qt.io>
sync-up with qt-creator; no effect on qmake.
comment on cherry-pick: this is actually a lot more than a cherry-pick,
because the dual VFS needs to deal with the file ids which were
concurrently introduced on the qmake side.
Change-Id: I2c1eb16c97526fa275a1c6a2eae9266d385859ac
(cherry picked from qtcreator/424639ecac9d2e404d2bfaff7f46b45ed98664b8)
(cherry picked from qtcreator/a8010b0fff47d903d4a1f80e3adb1a2ef41beb33)
Reviewed-by: Joerg Bornemann <joerg.bornemann@qt.io>
when the QFile object is already constructed, querying whether the file
exists is actually cheap, so do it right away instead of later on
demand. that makes the calling code a bit cleaner.
fwiw, that we need to explicitly query the file's existence at all is a
result of QFile's completely useless error "codes" (which merely say
which function failed, as if the caller would not know).
Change-Id: Ifec39d05b1713d8128046f679287e510f10e45dc
Reviewed-by: Tobias Hunger <tobias.hunger@qt.io>
(cherry picked from qtcreator/5ba32e3484ead2e35cc7732dcd59a97e7459dbfd)
Reviewed-by: Joerg Bornemann <joerg.bornemann@qt.io>
Has been flaky on those platforms in CI.
Task-number: QTBUG-67254
Task-number: QTBUG-66216
Change-Id: Ia1a718a23b1992fcc0e85bf49b714bc43acc4ce2
Reviewed-by: Ville Voutilainen <ville.voutilainen@qt.io>
Otherwise the ::debug() test fails when a build does not print qDebug()
messages.
Change-Id: I3f3c4b3c7d74004abe5ed8d7ac52164d4f88ef1f
Reviewed-by: Joerg Bornemann <joerg.bornemann@qt.io>
Also expands the tst_qsqlquery::batchExec() test to account for this
case and generally test the functionality. In addition it is made to be
more robust to avoid any discrepencies with the testing data. The test
in general is also cleaned up to enable more of it being tested with
the different database drivers where possible.
An expected fail is added for MySQL due to the fact that it has a bug
where null timestamp entries are being converted to the current
datetime when adding it as a bind value.
Change-Id: I0061bd1c69ae35b4858afc49420f13ce59cf48ae
Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
Clipping enabled state would not always be correctly restored for the
raster engine (other engines work fine). The raster engine's QClipData
object is sometimes shared between painter state objects on the
save/restore stack. QClipData has its own enabled flag, and this could
then come out of sync. Fix by making sure we sync the enabled state on
restore.
Task-number: QTBUG-58789
Change-Id: I81e6254ebb93df6e153bbef58e32a885273e3224
Reviewed-by: Allan Sandfeld Jensen <allan.jensen@qt.io>
