Q_ASSERT will only catch a violation of the precondition of the
checkReverseDirection function in a debug build. This commit replaces
the Q_ASSERT with QVERIFY'ing the precondition (that the widget has a
layout) before calling the function.
Change-Id: Ibaa2b52b7d67fa20784e0df022e8c8169da0e50e
Task-number: QTBUG-17582
Reviewed-by: Rohan McGovern
(cherry picked from commit d1011fc9328fa335bc69a065b29e77ce60608c55)
Sanity-check the test data when using it rather than when creating it.
Change-Id: Ie5d5e1ff3fc439d196096f17cb6f97680927a90c
Task-number: QTBUG-17582
Reviewed-by: Rohan McGovern
(cherry picked from commit 2c64404e2edd3b792b8a1d04c773860d40c04c4c)
Fix regression introduced in 3ca51d87296bcb3098d7de14294fef9294fd5ed4.
The warning is only generated in builds where QT_NO_DEBUG is not
defined *and* the target is not Mac OS X.
Change-Id: I7245f2eab8ea47ab7495fef874b2f8a29186b659
Task-number: QTBUG-17582
Reviewed-by: Rohan McGovern
(cherry picked from commit 60e5ed805af1e11aaec426b823df209b4c895c29)
Make the meaning of the code more obvious and avoid a compiler
warning about the variable "w" being overridden.
Change-Id: Ib76d3aa1cae46e263b2ab61b675d9ef74032aacc
Reviewed-by: Rohan McGovern
(cherry picked from commit 59634e825f1efa6f32befc8767cafedfeb9b7d59)
The function generating data blocks was filling the block to 16 bytes
short of the blockSize, then appending three 8 byte values, causing the
block to grow 8 bytes beyond blockSize and then truncating it back to
blockSize.
This commit makes the code fill the block to 24 bytes short of the
blockSize, so that the block will always end up at the correct size and
truncation is not needed.
Change-Id: I9fe6e6d6cf7bc445513b53e0a910d205c4c8002f
Reviewed-by: Rohan McGovern
(cherry picked from commit b1eb564830ff1b754de14919ce5c1547e9758f7c)
The Q_ASSERT here should really be a QVERIFY. Expanded the trivial
helper function to make it more obvious what the test does.
Change-Id: Ia6017fe820a83b4d6101e5edcfcb68993eef233a
Task-number: QTBUG-17582
Reviewed-by: Rohan McGovern
(cherry picked from commit 6181805f9ea66b37ee164bd67bdac2ac9d53fb65)
The assert and the last four parameters of makeList() are not needed.
The function is always used to make lists of three integers and the data
are all >= 0.
Change-Id: I93ae5a5f541cde2ff61bd9dd21164ed4a9b57403
Task-number: QTBUG-17582
Reviewed-by: Rohan McGovern
(cherry picked from commit e754832eff8e1a35fc2681ae69b6e3b1ad666ddb)
This code would assert in a debug build and potentially crash in a
release build. (The behaviour of QStringList::at() with an out-of-bounds
index is undefined.) This commit makes the program exit with a useful
error message in all builds.
Change-Id: Ia57a2e5693eb25d3eb0b9ba701ed485dfbc1e846
Task-number: QTBUG-17582
Reviewed-by: Rohan McGovern
(cherry picked from commit 5a530f6171a9569d1a35cd8cd1015ec796b048cc)
If a combobox had no completer, this test would assert in a debug build
and crash in a release build. This commit makes the test report a test
failure and avoid terminating in all builds.
Change-Id: Ib2924412e3d55f10875675cb65079259cd4ef552
Task-number: QTBUG-17582
Reviewed-by: Rohan McGovern
(cherry picked from commit a5bd4abed834ce3670b2f28d4f932d612a033b36)
Fix regression introduced in 3ca51d87296bcb3098d7de14294fef9294fd5ed4.
The warning is only generated in builds where QT_NO_DEBUG is not
defined.
Change-Id: I9aa3db369094f4046062b5dc1dc694342208ee45
Task-number: QTBUG-17582
Reviewed-by: Rohan McGovern
(cherry picked from commit 67bc4990f7541cca089882a8c3be1b2b8d98cd4c)
Replace obscure failure on bad test data in debug builds with an
informative warning in all builds.
Change-Id: I9001820f34de2f78bf296a2f0e095ce73d9ac4bd
Task-number: QTBUG-17582
Reviewed-by: Rohan McGovern
(cherry picked from commit 34c077faae538adc12d4aa28113640570dfc8728)
Some code in these tests is not intended to be executed, but rather to
detect compiler warnings. Instead of having an obscure fatal error if
this code is run in a debug build, it now has a clear fatal error if run
in any build.
Change-Id: I1c9d27bb14ebf3313865b68e3e57668ba1d14e25
Task-number: QTBUG-17582
Reviewed-by: Rohan McGovern
(cherry picked from commit 70f42ced8e01ddf0098b143e5804c22d55ebf16f)
Only build the autotest when the feature to be tested is in the Qt
build. This is better than building and running an empty test.
Change-Id: I67721f5f48296afcca64f761d12325f8e040f2d8
Reviewed-by: Rohan McGovern
(cherry picked from commit cf911bc0d297ed30e615fd115b0d3ae574cb2412)
Conflicts:
tests/auto/qaccessibility/tst_qaccessibility.cpp
Instead of crashing when we can't get a QAccessibleInterface for a
widget, output a useful warning and return a value that makes the test
fail gracefully.
Change-Id: I0b292cdd8f5a59e26bdc9b0b67cea2b58591df7d
Task-number: QTBUG-17582
Reviewed-by: Rohan McGovern
(cherry picked from commit d8dc5bfdff25d72e8276c6e80080d72315abe7e0)
Make the test report a useful warning, rather than terminating with an
unintelligible assertion failure, if the GestureState enum is ever
extended.
Change-Id: Ib876a5f2986cbea4e181678a83a72e0d1444f1ee
Task-number: QTBUG-17582
Reviewed-by: Rohan McGovern
(cherry picked from commit ed5dd84582881bc30a5ce85902b37aae9c243978)
Sort test data alphabetically and remove excess whitespace.
Change-Id: I0e244efca9e9adbe747a375a28a63f70992ef4f8
Reviewed-by: Trust Me
(cherry picked from commit 10f3f270c8a319205590ae13bdeab063d5216441)
This test relies on some API specific to MSVC's debug runtime, which is
not always guaranteed to be available.
Change-Id: Ib0ae4694ad51f59198dadfce802f0dfdf0522002
Reviewed-by: Rohan McGovern
(cherry picked from commit 01dfb9ca25bfaec5784f404dcd4bf41bbc2b6fcb)
Re-enabling for object types that don't crash the test.
Change-Id: I8c2f0d02171c973bf1ede227d4139b52cac5939f
Reviewed-by: Rohan McGovern
(cherry picked from commit 3788a90338e9328710b9d3335a63fb85ab7d97fc)
Fixes warning from tst_maketestselftest about test class name not
matching TARGET.
Change-Id: I7a81d87bb64d2902e842d097248e33eacb388746
Reviewed-by: Rohan McGovern
(cherry picked from commit 069e4441f2e14a82b4ae39601734acb71c909bd9)
Re-enable this test, with the two test functions that fail on Linux
disabled until the failures can be diagnosed.
Change-Id: I915e1a0d675cb71a80086e89f9799a4f9f6b600c
Reviewed-by: Rohan McGovern
(cherry picked from commit ee55dec1efe9c67518bf3e27f81b0696075f7153)
With modularized Qt, using QT_CONFIG is dangerous, because the behavior
changes depending on the order in which modules are qmake'd.
For example, an autotest doing:
contains(QT_CONFIG,svg):QT += svg
...will depend on libQtSvg if (and only if) the autotest is qmake'd
_after_ qtsvg is qmake'd.
This makes the tested functionality unpredictable.
Also, if the above example occurs within qtbase, it causes the test to
sometimes have a circular dependency: if qtsvg is qmake'd before the
test is qmake'd, the test in qtbase depends on qtsvg which depends on
qtbase.
Tests must avoid functionality tests via QT_CONFIG except where all the
tested modules are dependencies of the current module.
Usage of QT_CONFIG with qt3support was entirely removed since Qt5
will not retain qt3support.
Reviewed-by: Jason McDonald
Change-Id: I5a5013b3ec7e1f38fb78864763c9e7586c15e70b
Recursive QMAKE_EXTRA_TARGETS were omitting the `-f' option to make.
This would break in the case where the correct makefile was not named
`Makefile'. The included autotest demonstrates the problem.
Note that this was fixed for normal targets back in 2005
by faac7bd178654fd67a6f3f9cf4f6f2605071448d (p4 202370), but was not
fixed for extra targets.
Reviewed-by: ossi
(cherry picked from commit 96a3bf7a8bbc1e5361e16cbeeceb4be674b88c30)
(It did not really fix the issue.)
This reverts commit 33f525e636ef8fa64a15d3e66c56adaea0075bda.
Conflicts:
src/gui/graphicsview/qgraphicslinearlayout.cpp
tests/auto/qgraphicswidget/tst_qgraphicswidget.cpp
(cherry picked from commit fee052e3e37b3335fe563cb8a1881bf59f9e25d0)
* Do not call invalidate from activateRecursive().
This resulted in that a layout was invalidated as many times as there
were items in the layout.
* Several improvements. Do not call resize(size()) too often.
Calling resize() from the widgetEvent() is not very nice though...
* Remove commented out code
* make sure layout is activated even if the widget does not change size
* activate the layout if the resize is same as size()
* In order to not break existing apps, make this an opt-in feature
with QGraphicsLayout::setInstantInvalidatePropagation(true);
Reviewed-by: Frederik Gladhorn
Reviewed-by: John Tapsell
Make sure the resulting image / pixmap is valid if the source was valid.
Task-number: QTBUG-19157
Reviewed-by: Kim
Reviewed-by: Benjamin Poulain
(cherry picked from commit 1c5da7207a21cc44a4a08d291c290ffcd9b958fd)
Older commit history is in KDE svn:
http://websvn.kde.org/trunk/KDE/kdelibs/kdeui/itemviews/kidentityproxymodel.cpp?view=log
Ammended to update the license headers.
Merge-request: 900
Reviewed-by: Gabriel de Dietrich <gabriel.dietrich-de@nokia.com>
Conflicts:
doc/src/frameworks-technologies/model-view-programming.qdoc
tests/auto/headers/tst_headers.cpp
(cherry picked from b00089261eafbdf5f92ed94d7fb20b402bfcaeb2)
Widgets are left in a transitional (and incosistent) state while being
re-parented, which caused QWidget::effectiveWinId() to crash in certain
circumstances. See patch for more details.
Auto test included.
Reviewed-by: ogoffart
(cherry picked from commit 6db0153cd7e35e4a919a76ae2aadbf2d2510bfb7)
select code for open C file/socket descriptors was crashing in FD_SET
if a QSocketNotifier was created with an invalid descriptor.
Added two autotests to QSocketNotifier, one to check notifiers with
bogus socket descriptors don't crash, the other to check that notifiers
with posix socket descriptors do work. (symbian socket engine doesn't
use them so they are not implicitly tested)
Reviewed-by: mread
Task-Number: QTBUG-18138
(cherry picked from commit 8a9a6afcf02f089f932bc81431ab46a60af32134)
Ignore warning when the test intentionally sets an invalid socket descriptor.
Make sure to set content type on all http post tests in tst_qnetworkreply.
Run test with enough capabilities to avoid platsec errors when accessing
certificate store.
Reviewed-By: Markus Goetz
(cherry picked from commit 9632fdefa9012ca11cd1345d66bafd6f417de88e)
* 'master' of git://scm.dev.nokia.troll.no/qt/qtbase-staging:
Fix tst_QVariant::invalidColor
Move QTextCursor::MoveStyle to Qt namespace
Add some QRawFont related low level functions to avoid extra copying
Since a423ff5474b89028eeca95b254f5184311c8223b, the warning message is
no longer produced.
Reviewed-by: Samuel Rødal
(cherry picked from commit 8a5e82732be3aac37d14ef85c6974add46c6b65f)
We cannot use QTextCursor::MoveStyle enums in QTextLine because
QTextCursor is not a QObject, while referring to that enum in
Q_PROPERTY requires it to be. That's why we need to move the
enums in Qt namespace.
Reviewed-by: David Boddie
(cherry picked from commit 5eba82b752e85a5d6cb3a893214ed2646d75f362)
tst_lancelot did QT+=svg, but didn't actually use QtSvg.
qlocalsocket/lackey uses QtScript, which is not available when compiling
qtbase.
Change-Id: Ic3cc5a6f74a58ee8f2fdc48b9c852d9551b85f68
Any test which needs private headers from some Qt module must do:
QT += modulename-private
Reviewed-by: Jason McDonald
Change-Id: I6924a577a960e4990f4379b02bca4822d8248fb4
Jason McDonald
Rohan McGovern
Bjørn Erik Nilsen
Eckhart Koppen
Olivier Goffart
Jan-Arve Sæther
Gabriel de Dietrich
Samuel Rødal
Stephen Kelly
Shane Kearns
Qt Continuous Integration System
Jiang Jiang