Change childWidget->windowHandle() to childWidget->internalWinId() in
q_createNativeChildrenAndSetParent() to determine whether should we call
childWidget->winId().
This is because in some circumstances Qt will crash due to accessing
deleted QWidgetWindow object if we use windowHandle(). Think about the
following scenario:
1) create a widget A without parent and add two child widgets B and C to A
2) create a native widget D as the child of B, note that when we set
Qt::WA_NativeWindow attribute to it, its QWidgetWindow will be created
which means its windowHandle() is not null.
3) create a top level widget E as the child of C and show it. This will
make Qt call createWinId() to A and then
q_createNativeChildrenAndSetParent() will be called to create A's native
children recursively and finally make D's QWidgetWindow object become a
child of A's QWidgetWindow object. Please note here that B will not become
a native widget just because at that moment windowHandle() of D is not
null and Qt will not call winId() to its parent B
4) Set A's parent to another widget which has been shown, setParent_sys()
will be called to A and then Qt will call destroy() to A. in destroy() Qt
will try to call destroy() to its children recursively with a condition that
the child has Qt::WA_NativeWindow been set. But D's parent B is not a native
widget right now so B and D is not destroyed. Qt will then deleted the
QWidgetWindow object of A, since E's QWidgetWindow object is a child of
A's QWidgetWindow object, it will also be deleted. Now E hold a deleted
pointer of QWidgetWindow object. This is the source of crash later.
Task-number: QTBUG-35600
Change-Id: I97a20a68e626ee62b15bb4eae580e26f8948923b
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@digia.com>
Reviewed-by: Jørgen Lind <jorgen.lind@digia.com>
QtWayland uses QtWaylandScanner to generate c++ files which will be
subclassed. This works ok in QtCompositor as it implements the pimple
pattern. However, QtWayland has to expose a library called
QtWaylandClient as a way to create specialized wayland plugins. These
classes don't promise an ABI, and has limited API guarantees, so they
subclass the classes generated by the QtWaylandScanner directly. For
this library to be possible, we have to install the generated files.
Also, it should be possible to use the private API, and that's not
possible without the generated files.
Change-Id: Id55200fe23652587390dabf84ca846e4f39cf70e
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@digia.com>
Reviewed-by: Laszlo Agocs <laszlo.agocs@digia.com>
Reviewed-by: Andy Nichols <andy.nichols@digia.com>
Hand-roll a string table here, since there's opportunity for
sharing string data between string entries.
Effects on Linux AMD64 GCC 4.9-trunk release stripped:
text: -88B
data: -64B
relocs: -5
Maintainability is hurt somewhat, but it is expected that the
contents of this string table do not change much, so the
overall effect is still very low.
Change-Id: I2a22da4c8548c53ef31c33319b4652f3cb6f62f5
Reviewed-by: Simon Hausmann <simon.hausmann@digia.com>
QAtomic will now depend on QTypeInfo, so shuffle the order to make
sure the includes are right.
The first set do not require anything but the macros from qglobal.h,
then qflags.h requires qtypeinfo.h and qtypetraits.h, qatomic.h
requires qtypeinfo.h and qglobalstatic.h requires qatomic.h.
Of those, only qatomic.h isn't in global/, so we could even consider
using "" includes, but I recall MSVC not honoring them correctly.
Change-Id: I76c150254a99cdea8857445bdebf2e724eaa76bd
Reviewed-by: Olivier Goffart <ogoffart@woboq.com>
Reviewed-by: Marc Mutz <marc.mutz@kdab.com>
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@digia.com>
Reviewed-by: Jędrzej Nowacki <jedrzej.nowacki@digia.com>
Right now, the sublib targets, if any, show up between clean and
distclean targets. That's silly.
I doubt anyone is using sublib targets anyway, but...
Change-Id: I2beffc69f68fa7626ff4aa4a7cc1169b2c6c69a7
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@digia.com>
Then it's possible to embed native cocoa widgets without crashing.
Task-number: QTBUG-35629
Change-Id: Ic212c36178282b4d090c6ce0470012adc8e0c2bb
Reviewed-by: Morten Johan Sørvig <morten.sorvig@digia.com>
Reviewed-by: Gabriel de Dietrich <gabriel.dedietrich@digia.com>
- thread was duplicated
- x86 & ppc are obsolete and don't actually do anything
- incremental was just plain nonsense (it does something entirely
different, and it's better to hide this "feature" from public view)
- resources is basically an implementation detail (it's on by default if
qtcore is used)
Change-Id: I9163af6e8db7988382ccf993d4be280f7faec1f2
Reviewed-by: Leena Miettinen <riitta-leena.miettinen@digia.com>
Reviewed-by: Joerg Bornemann <joerg.bornemann@digia.com>
while the dependencies in the manual projects are crappy anyway, it is
still worth to cover the case of the user changing the install paths.
Change-Id: I0a7ca5c8ba660c689d6d7af6b65d878390d6456f
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Reviewed-by: Joerg Bornemann <joerg.bornemann@digia.com>
Done-with: Leena Miettinen <riitta-leena.miettinen@digia.com>
Change-Id: I1e031402bc3d857cf29782957e5340e3c82f1ed2
Reviewed-by: Leena Miettinen <riitta-leena.miettinen@digia.com>
Some database systems (like PostgreSQL) use alphanumeric error codes.
Introduce a new method nativeErrorCode() which replaces number().
If the error code cannot be converted to int, number() will return 0.
Task-number: QTBUG-142
Change-Id: Ic7fba841737674b75c0c01c2263f51d2041da497
Reviewed-by: Mark Brand <mabrand@mabrand.nl>
Those flags are required to pass the Windows Store App Certification
process. Otherwise apps are not allowed to be published.
The SAFESEH option is only required for x86.
According to documentation APPCONTAINER only talks about the
executable, but when running through the certification, the Qt modules
are reported to be errornous as well.
Change-Id: I5450687dcd5bc537149e331332e253c4617df55d
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@digia.com>
Reviewed-by: Oliver Wolff <oliver.wolff@digia.com>
Reviewed-by: Andrew Knight <andrew.knight@digia.com>
When we get a WM_CHAR message we do not need to add anything to the key
map as these messages already contain the character code in question.
Task-number: QTBUG-35532
Change-Id: If73993cd873b7c616876b9ba52ceba6b27a6a7b1
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@digia.com>
This allows us to add new members without breaking BC.
Change-Id: I59236cd5eeacdcfc490849a18b3bdf70d7a501bf
Reviewed-by: Mark Brand <mabrand@mabrand.nl>
Task-number: QTBUG-34749
Corrected link to drag and drop example.
Corrected link to {mandatory fields}
Created target referring to macros for defining plugins
Change-Id: I387a2d9bab428b2eacd8d371f08c72f42f7e2be2
Reviewed-by: Topi Reiniö <topi.reinio@digia.com>
Querying the "extra" fallback fonts (which are hard-coded names, actually)
for family only makes sense for scripts that cover more than just a single WS;
these currently are Common (<-Inherited) and Han.
Change-Id: I37682f1ae51a476a21b933607de48b01255cf1c3
Reviewed-by: Pierre Rossi <pierre.rossi@gmail.com>
Reviewed-by: Konstantin Ritt <ritt.ks@gmail.com>
In a slot connected to readyRead, if the app detects that the
buffer size is too small and increases it, it expects that
readyRead() will be emitted again.
setReadBufferSize() doesn't re-enable the socket notifier when
calling from within readyRead, and readyRead itself was missing
the code to do it.
Change-Id: Ia00a3066ad3ba09d5cfae0716adc5691ae96c3fa
Done-with: Thiago
Reviewed-by: Peter Hartmann <phartmann@blackberry.com>
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Use of mkpath/mkdir is removed in order to make the proxy less
invasive when there is no compiler service running. Creation of the
directory structure is fully the service's responsibility.
Service availability is now only checked once, at the first invocation of
D3DCompile, as it is expected to be started before the application.
Change-Id: Ib8c4f062c418497c2253daf524654e1db30dae47
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@digia.com>
MSVC warns about multiple symbols when using extern "C", but MinGW
requires it. #ifdef it for MinGW.
Change-Id: I8b16ab4ea129312787090e042115d8f709fdd814
Reviewed-by: Oliver Wolff <oliver.wolff@digia.com>
Do exactly what windows font database does to determine
writingSystems in case of non-truetype fonts.
Task-number: QTBUG-13585
Change-Id: I4a1a48d74c36e403c31f20847cf80295d89a34bc
Reviewed-by: Lars Knoll <lars.knoll@digia.com>
Most devices that use tslib have the device path defined by
the TSLIB_TSDEVICE env. variable.
Change-Id: I89fc17cb2c7fba7741de2fd348aacd608cb21a2a
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@digia.com>
Reviewed-by: Laszlo Agocs <laszlo.agocs@digia.com>
For the linuxfb platform the screen was registered
before its geometry was calculated resulting that
the qpa controller will deal with a screen instance
that has geometry of QRect(0,0,1,1) .
This masks pretty much all the painting and the input.
The fix is to call screen init before registering
the screen instance to the qpa controller.
Change-Id: If84daca6587dc368301b578cdbf7b2a65a56a5b5
Reviewed-by: Laszlo Agocs <laszlo.agocs@digia.com>
QMacPasteboardMime is a public class in Qt 4. Qt 5 has
two copies, one public QtMacExtras and an internal
one in the cocoa plugin. This causes a symbol collision
when building statically.
Rename the internal copy to QMacInternalPasteboardMime.
Task-number: QTBUG-35310
Change-Id: I891787b451a0b342ed85aa7196e606bc11623e21
Reviewed-by: Gabriel de Dietrich <gabriel.dedietrich@digia.com>
The QVarLengthArray inside qlocale_blackberry.cpp was being preallocated with
512 bytes, that means internally it could grow up to 512 bytes using the
stack before switching to the heap, but its actual semantic size was still 0.
After qt_safe_read(... buffer.data() ...) was being called, data was written
to the QVarLengthArray buffer, but its semantic size was still 0, since it was
not resized or anything. This triggered an assertion when buffer[bytes] = '\0'
was assigned, since 'bytes' > buffer.size() ( == 0) despite buffer.capacity()
== 512.
Change-Id: I5503ee9b02413794f67730700fba05a4c194d465
Reviewed-by: Mehdi Fekari <mfekari@blackberry.com>
Reviewed-by: Giuseppe D'Angelo <giuseppe.dangelo@kdab.com>
- Explain that values() is unnecessary and causes an extra copy.
- Replace "are fastest" with "are faster".
- Briefly explain the snippet.
Change-Id: Iff1d3a0b6140fca1d78157aeeca3d538c4d78b84
Reviewed-by: Albert Astals Cid <albert.astals@canonical.com>
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
The "Fixed CE build of sqlite3" patch is preserved in this change.
(ea70ec8711)
Change-Id: I41a268bd077e396810965ca27cd572cef7259d58
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@digia.com>
Reviewed-by: Mark Brand <mabrand@mabrand.nl>
It is annoying that HB maintains binary compatibility but does not
really care about source compatibility.
Change-Id: Ife7b011919f28527b77858ad8398fe723c65b1b3
Reviewed-by: Lars Knoll <lars.knoll@digia.com>
Windows 7 and later have LOCALE_SSHORTTIME, which is what we need.
Task-number: QTBUG-33718
Change-Id: I4c3f113d17102a37fb752de56f06b312f27c7887
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@digia.com>
Using "default" means "reset to default compiler settings", not "revert
to command-line settings". And even if it did, it's possible some other
headers by the user toggled the settings. We shouldn't affect that.
Task-number: QTBUG-35622
Change-Id: If7c4bd0cc8c210f88ee89e3fa2dd3b7194ad77bb
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@digia.com>
When using the visual studio generators, a moc parameters file
for each configuration is generated. The content differs because
the -DQT_NO_DEBUG may be added or not, depending on the configuration.
Change-Id: I52eae536289d451df72df2e4dba709851279694a
Reviewed-by: Stephen Kelly <stephen.kelly@kdab.com>
