instead of having each generator do its own magic (little surprisingly,
with different outcomes), add "stuff" to the search path in one place
used by all generators. this has several consequences:
- (unless disabled via CONFIG+=no_include_pwd) $$PWD is now consistently
prepended by all generators. most notably, this was entirely missing
from the MSVC generators (both nmake and VS) - despite them needing it
most. this also affects Xcode projects.
- $$OUT_PWD (if different from $$PWD) is now added right after $$PWD,
not at the end. this precedence clarification only makes sense, given
that qmake tries to make shadow builds as transparent as possible.
- the qmakespec's dir is now consistently appended. the UNIX and PBX
generators prepended it, while the rest already appended. few files
actually include qplatformdefs.h, so having it late in the search path
seems reasonable.
- the effect of CONFIG+=depend_includepath is now fully consistent with
the actual include path.
Change-Id: I5f7570183351ade29342ea74fef706a0738842bf
Reviewed-by: Simon Hausmann <simon.hausmann@digia.com>
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
neither qmake_getpwd()'s return value nor a fileFixify()'d version of it
can be empty.
Change-Id: Ic3b7d20becc57209b9dbe71ad9dc8e7547d435b1
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
The outer QNativeSocketEngine::bind() function will call
fetchConnectionParameters() as soon as nativeBind() returns, so don't
bother copying localAddress.
Change-Id: Ice13e507ccb9c575a7d3bdf0b41394f35230b746
Reviewed-by: Richard J. Moore <rich@kde.org>
Also make the fragment shader a bit simpler
Change-Id: Ie50940da0a4f896504c9f5962cdb6c455983302f
Reviewed-by: Laszlo Agocs <laszlo.agocs@theqtcompany.com>
- After reset a surface we must call makeCurrent before we are usign
swapBuffers.
- No need to set the surface in QPA when surfaceCreated are called in
QtSurface.java, some time the OpenGL surface is not fully initialized at
this stage. Is better to wait for surfaceChanged which is always fired
at least once.
- DO NOT reset m_surfaceId to 1 when there is no surface. The problem
is that if we have one surface and when we distory it we don't (need to)
wait for its surfaceChanged/surfaceDestroyed notifications, and if we
create another one quicly it will have the same id (1).
Task-number: QTBUG-39712
Change-Id: I2aa31e5b59d81ef3b03624d4636a4381eea6d543
Reviewed-by: Eskil Abrahamsen Blomfeldt <eskil.abrahamsen-blomfeldt@theqtcompany.com>
Android does not decorate a floating QDockWidget, leaving the user with
no option to close or move such a widget. This changes activates Qt's
own drawing of window decorations for this case (similar to Win CE).
[ChangeLog][Android][QtWidgets] Enable QDockWidget window decorations.
Change-Id: Id46a994a8be9b8f3e7b530af118315cd36b5fb0e
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@theqtcompany.com>
This shotgun-surgery approach is motivated by trying to get a
clean(er) build for -Wnoexcept on GCC, so it is expected that
for any class touched here, there will be more operations that
can be marked nothrow. But they don't show up in conditional
noexcept clauses, yet, so they are deferred to some later
commit.
Change-Id: I0eb10d75a26c361fb22cf785399e83b434bdf233
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
The way swapping is supposed to work is:
1. Each type supplies a swap() function or function template in its
namespace. Any good STL implementation will find it there through
ADL. As will the primary qSwap() template.
2. Each use of swap() in Qt, in particular in template code, should
use qSwap() instead of std::swap() or the using+swap-trick, because
qSwap() automatically enables ADL. It also has a sophisticated
conditional noexcept specification that can be used in the
custom swap() functions' own noexcept clause.
This change also allows us to convert implicitly-shared classes'
member-swap functions to noexcept one at a time, because the
specialization will no longer be in conflict with the primary
template regarding exception specifications. The primary's
specification could, of course, be reused here, but it's complex
and if the machinery around it is changed later on, it will not
affect Q_DECLARE_SHARED classes.
Change-Id: I3389a655a9fd8de370f363c8fcef60269a9f506c
Reviewed-by: Olivier Goffart <ogoffart@woboq.com>
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Yes, this is necessary. The noexcept operator looks for noexcept tagging,
not at the contents of the function to determine whether to return true.
The more conditionally-noexcept functions are used, the more important it
becomes that low-level functions are correctly marked noexcept. In that, it
is like constexpr.
Change-Id: I4bca178444d1fd7caf3a92f996b1536eebdb5014
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
This is mostly straight-forward, but some things are worth noting:
1. Yes, this is necessary. The noexcept operator looks for noexcept tagging,
not at the contents of the function to determine whether to return true.
The more conditionally-noexcept functions are used, the more important it
becomes that low-level classes are correctly marked noexcept. In that, it
is like constexpr.
2. The functions that actually call into the model (data(), flags(), sibling(),
...) can throw (bad_alloc, if nothing else).
Consequently, they're not marked nothrow. They're the only ones.
Change-Id: Id0413212b0f1c049a339480ee449a53c3ca9fea0
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
An empty ring buffer always has a pre-cached byte array in its
container. In this case size() indicates that no data is
available, but readPointer() tries to resolve this byte array.
To avoid an illegal pointer as a result, return Q_NULLPTR instead.
Change-Id: Icc5f08f071a8f02a14c112b6e1adbe5373bd9466
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@theqtcompany.com>
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
There's no need to allocate memory for the special address.
Change-Id: I5f3760565807731ab595e91fc934c21d10df212a
Reviewed-by: Richard J. Moore <rich@kde.org>
Commit d44781730c disabled use of pkg-
config on Mac, which in turn automatically disables all of the tests
that depend on pkg-config too, like D-Bus. If you then use the -dbus
flag, configure would error out telling you that it couldn't find D-Bus,
despite it being there.
Now we print some helpful information on error and extra information for
finding out why it may have got disabled.
Change-Id: Ic4e41c58d62d80eb1a0f3ca1c93529e049aaf231
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@theqtcompany.com>
This is not used or referenced anywhere
Change-Id: I02e1aa76631627f64e5d1f9b36a13cdb5677e93f
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@theqtcompany.com>
2013 and 2015 compile this fine. I didn't test 2012. I wouldn't have
fixed if the objective weren't to enable QtDBus by default on all
architectures: since it is, we can't have Qt fail to compile from
sources on MSVC 2010.
qdbus_symbols.cpp(92) : fatal error C1001: An internal error has
occurred in the compiler
Change-Id: I42b930bc37c4e478a66725d83c8a73836fbf567c
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@theqtcompany.com>
Reviewed-by: Simon Hausmann <simon.hausmann@digia.com>
we are in an #else of #ifndef QT_BOOTSTRAPPED here already.
Change-Id: I02c4ff2959490110c21ad1016c664b7ddcfea7c0
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
On Windows mouse grabbing doesn't work across processes, which
means we're interacting with other windows when picking colors.
Workaround that by having a transparent 1x1 window below the cursor
at all times so we catch the mouse click. Clicking before the window is below
the cursor won't happen because our timer interval is 30ms, so it's quite fast.
It's hacky but it's what we can do for a feature which was very broken on Windows.
Task-number: QTBUG-43663
Change-Id: I295378e033ddc6d9c2230335f0953a727c21e1dd
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@theqtcompany.com>
It was not possible to determine whether Qt is being built as
a shared or static library in the config tests, as that information
was not added to QMAKE_CONFIG used in the compileTest function.
For linking to ICU on Windows, there is logic which depends on
it in config.tests/unix/icu/icu.pro, but was broken for the static
target, as the expected library name was not found (icuXX -> sicuXX).
Change-Id: I22c2d543ea14feebbad36961c9a29e99f66323ff
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@theqtcompany.com>
The window size was too small on high-DPI screens.
Save the default text color instead of setting it to black.
Change-Id: I78b50624110be0cb1d077d3782d421eb323f4fb0
Reviewed-by: Paul Olav Tvete <paul.tvete@theqtcompany.com>
A mismatch between requiresPretransformedGlyphPositions and
shouldDrawCachedGlyphs in QRasterPaintEngine will cause the text
position to be transformed twice when using drawGlyphRun on a QPainter
with a perspective transform. Since this case falls back to drawing
text as paths there is no reason to require any special treatment.
Change-Id: Ib1c14aee4cc6774dd8feadc5748f0b0ee59633b9
Reviewed-by: Tor Arne Vestbø <tor.arne.vestbo@theqtcompany.com>
Reviewed-by: Gunnar Sletta <gunnar@sletta.org>
On e.g. Android, the tabs will be scaled by DPI. This breaks the
logic in the QTextLayout tests for tabs. It's not possible to
just scale the expected sizes either, since the whole text layout
will be affected by font sizes and scaling, and it's difficult to
predict where characters will land on different platforms and
resolutions.
To avoid breaking this test on other platforms, we just skip them
when we know they will break. Since the code tested is
cross-platform, this will hopefully not have any significant
impact on our coverage.
Change-Id: I65d6c33c9c6724665983a17f99eadcf1baedcc20
Reviewed-by: Tony Sarajärvi <tony.sarajarvi@digia.com>
Reviewed-by: BogDan Vatra <bogdan@kde.org>
Include test data on Android.
Change-Id: Ic0890495c0bf8d2e38595585226a03073b043d57
Reviewed-by: Tony Sarajärvi <tony.sarajarvi@digia.com>
Reviewed-by: BogDan Vatra <bogdan@kde.org>
Call destroyScreen(), like other platform plugins do since
9b4fbe85d2.
Change-Id: Idbddb026d7a964f15c00acfacf70064405dc3f16
Reviewed-by: Giulio Camuffo <giulio.camuffo@jollamobile.com>
MSVC prints warnings if all slots are actually static methods:
moc_xxx.cpp:71: warning: C4189: '_t' : local variable is initialized but not referenced
This can be easily avoided by using Q_UNUSED.
Change-Id: Ice134a8a8d3e6aef1df1e7a89e9a2867281580fd
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@theqtcompany.com>
Reviewed-by: Olivier Goffart <ogoffart@woboq.com>
Move compiler warning 4996 from level 3 to 4. This is needed to avoid
warnings about the use of C++ standard library functions like std::copy,
which is used e.g. in qvector.h (since c6752c5aa1):
'std::copy': Function call with parameters that may be unsafe -
this call relies on the caller to check that the passed values are
correct. To disable this warning, use -D_SCL_SECURE_NO_WARNINGS. See
documentation on how to use Visual C++ 'Checked Iterators'
Because the warning has to be disabled before any standard C++ header
is included one cannot just fix this locally in qvector.h.
Change-Id: I929f1535656bca9f5beb7fd0d557178370c232c6
Reviewed-by: Marc Mutz <marc.mutz@kdab.com>
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@theqtcompany.com>
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@theqtcompany.com>
Just use QT_NO_QDEBUG_MACRO, like we do already for qInfo.
Change-Id: I4b2ef68427fbe6f253fe02a3ab161fa25186e834
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Add an 'info' message type that can be used for messages that are neither
warnings (QtWarningMsg), nor for debugging only (QtDebugMsg). This is
useful mainly for applications that do not have to adhere to the
'do not print anything by default' paradigm that we have for
the Qt libraries itself.
[ChangeLog][QtCore][Logging] QtInfoMsg got added as a new QtMsgType.
Use the new qInfo(), qCInfo() macros to log to it.
Change-Id: I810995d63de46c41a9a99a34d37c0d417fa87a05
Reviewed-by: Jason McDonald <macadder1@gmail.com>
ConfigLocation was erroneously inconsistent, by adding the org name
and app name on Windows (unintentionally) and not on Unix (while having
subdirs in ~/.config is actually common practice for some XDG desktops)
Therefore this adds AppConfigLocation, which always has the org name
and app name (while GenericConfigLocation never does).
[ChangeLog][QtCore][QStandardPaths] Added QStandardPaths::AppConfigLocation,
for application-specific configuration directory. ConfigLocation was inconsistent.
Task-number: QTBUG-38872
Task-number: QTBUG-38845
Change-Id: I80fb98f296436261f3996c9df87571c29b06ab35
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@theqtcompany.com>
Reviewed-by: Joerg Bornemann <joerg.bornemann@theqtcompany.com>
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Using QPointers (or any type that makes a QPointer part of its identity) as a key
in any associative container is wrong. They get externally set to nullptr,
violating the associative container's class invariants, which could lead to
data corruption, even though bucket-based hash implementations are less susceptible
than binary trees.
To fix, write a new class that acts much like the old QPair<QPointer<>,QByteArray>,
but uses the QPointer only as a guard, not as part of its identity. To preseve
identity, also saves the naked pointer originally passed and uses that for op==
and qHash().
Change-Id: I4fa5a6bf86bad8fe7f5abe53d7c7f3ad3754d8d6
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
The old code creates a RestorableId from the passed (QObject*, QByteArray)
and used it for lookup in a hash table (ok) and as a container for the
(QObject*, QByteArray), satisfying later references to those parameters
from the RestorableId instance instead of using the parameters directly.
Now, RestorableId holds the QObject* in a QPointer, so the code might have
wanted to detect the object being destroyed as part of the operation, BUT:
a) the operation is a read-only one, and b) the code didn't check for
nullness before dereferencing the QObject*.
Fix by moving the creation of the RestorableId into the scope it's used
and otherwise using the parameters directly.
Change-Id: Iaf12f722fe6841ee1a55037fe33a6115aa699aca
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
qHashRange() takes an (input iterator) range and hashes each element, combining
the hash values using the hash combiner from Boost/N1837 with the magic number
0x9e3779b9, as described here:
http://stackoverflow.com/questions/4948780/magic-number-in-boosthash-combine
qHashRangeCommutative() does the same but with a cummutative combiner (unsigned
addition) to create hash values that are order-independent, e.g. for hashed
containers. The obvious combiner, XOR, is a bad one because it eliminates
duplicate elements. Signed addition cannot be used, since signed overflow
leads to undefined behavior.
[ChangeLog][QtCore] Added qHashRange() and qHashRangeCommutative() functions to aid
implementing qHash() overloads for custom types.
Change-Id: I3c2bbc9ce4bd0455262a70e0cf248486525e534f
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
This is the modern, correct way of providing a swap
operation. See http://stackoverflow.com/a/8439357/134841 for more
information. By changing this, we also fix Qt building with ICC
and libc++.
This patch also adds a noexcept() rule to match what the C++11 standard
requires.
Change-Id: I18f22fe7c92cf253e94168e1312346b4c183f536
Reviewed-by: Olivier Goffart <ogoffart@woboq.com>
Reviewed-by: Marc Mutz <marc.mutz@kdab.com>
This is primarily to get a cleaner build of src/tools
under -Wnoexcept.
Change-Id: I0dea21e70aad56b25675fc59fac0327b55ee83e3
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Change-Id: I22f1eaa892cba23c498ae210a9a483e468268581
Reviewed-by: Richard Moe Gustavsen <richard.gustavsen@theqtcompany.com>
Reviewed-by: Tor Arne Vestbø <tor.arne.vestbo@theqtcompany.com>
On iOS 6 and above, [UIViewController supportedInterfaceOrientations]
needs to return 0 for [UIApplication setStatusBarOrientation] to work.
This means once you report a content orientation other than the primary
orientation, you'll disable auto-rotation. Reporting the orientation as
Qt::PrimaryOrientation restores the auto-rotation behavior.
Change-Id: I1b8c765c507728fdbc5b828e0b4215324014e221
Reviewed-by: Richard Moe Gustavsen <richard.gustavsen@theqtcompany.com>
Reviewed-by: Tor Arne Vestbø <tor.arne.vestbo@theqtcompany.com>
If reset is disabled then POST and PUT requests can not be authenticated
as the upload device can not be reset. There shouldn't be any reason
that shouldn't be allowed if the QIODevice given supports resetting.
The disableReset feature of QNonContiguousByteDevice is removed as it
is not used anywhere else, and is redundant when reset can indicate
success or failure.
Task-number: QTBUG-43628
Change-Id: If941a98fd3f797872351c10bdca6aa6745dbefea
Reviewed-by: Jocelyn Turcotte <jocelyn.turcotte@digia.com>
