| Commit message (Collapse) | Author | Age | Files | Lines |
|
|
|
|
|
|
|
|
| |
that server was found out not to support HTTP pipelining.
tested manually; for more information see the task.
Reviewed-by: Markus Goetz
Task-number: QTBUG-21369
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
If a HTTP server responds with gzip-encoded empty content without defining
Content-Length in the response header QHttpNetworkReplyPrivate::gunzipBodyPartiallyEnd
will crash because it calls zlib inflateEnd for an uninitialized stream.
- Fixed the crash by adding a check if the stream is initialized to gunzipBodyPartiallyEnd.
- Added a regression test tst_QNetworkReply::nb279420gzipNoContentLengthEmptyContentDisconnect
PMO 279420
Task-number: QTBUG-22660
Signed-off-by: Sami Rosendahl <sami.rosendahl@reaktor.fi>
Merge-request: 1465
Reviewed-by: Peter Hartmann <peter.hartmann@nokia.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Calling setTextModeEnabled on a QIODevice that is not open generates
a warning as a result of QTBUG-20905. Previously, it caused incorrect
behaviour when called with true.
Now only call it if the QIODevice passed in is already opened.
If not open, then it is opened by the xml reader without text mode,
so there is no difference in behaviour.
xml autotests still pass, although this removes several QWARN messages
from the qdom autotest.
Task-number: QTBUG-22659
Change-Id: Ie3137e44fe776bff6371ebd008428110168717ec
Reviewed-by: João Abecasis <joao.abecasis@nokia.com>
(cherry picked from commit d457f148cf1efbea3db5fdc87c418cc9647e2baf)
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
The Proxy-Connection header is a non standard header, but is widely
used so forming a de-facto standard.
Some proxies use the official Connection header, so we should check
for that in responses. Otherwise https connections over http proxy
fail in case the proxy sends "Connection: close" with the 407 reply.
Task-number: QTBUG-22177
Change-Id: If6cfa4ebb7ac9d97d65b6ddcc8257aee20ac0448
Reviewed-by: Peter Hartmann <peter.hartmann@nokia.com>
(cherry picked from commit 0ad18e18d1223b173d4a0d374b70ec08c3b22b11)
|
|
|
|
| |
Reviewed-by: Olivier Goffart <ogoffart@woboq.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Two problems:
- The signal cacheCredidentials was not connected in the synchronous
case while it must be connected. (Regression when the threaded http
was merged)
- We cannot cache the credidentials when we proceed the url because
at that point, we do not know the realm (this basically reverts
9bc5a32b875b812c3a706034c8c27614f86bd138)
Merge-request: 1459
Reviewed-by: Peter Hartmann <peter.hartmann@nokia.com>
Task-number: QTBUG-18411
|
|
|
|
|
|
|
| |
... as did browser vendors.
Tested manually with affected CA certificates.
Reviewed-by: Richard J. Moore <rich@kde.org>
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
| |
Calling setTextMode() before open() would make the device appear to be
already open and cause later errors.
Added a qWarning and documentation update to prevent this API misuse
Task-number: QTBUG-20905
Change-Id: I2e06cd8e79f4afcf27417ac0eae6ebef980a17aa
Reviewed-by: Thiago Macieira (Intel) <thiago.macieira@intel.com>
Reviewed-by: João Abecasis <joao.abecasis@nokia.com>
(cherry picked from commit 29c30a20bab4c4ea892b95c08c71bb5f136bb82c)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
There are lots of buggy SSL servers around and to connect to them you
need to disable various features. This commit adds the ability to
disable the SSL ticket extension, the ability to disable the insertion
of empty fragments, and the ability to disable compression.
Task-number: QTBUG-21906
Change-Id: I3e1d0347a46e9030b889bbf15b2aad19b8513b73
Merge-request: 68
Reviewed-by: Peter Hartmann <peter.hartmann@nokia.com>
(cherry picked from commit 78d02e93aca5325fc5be9bfd275862795207abaa)
(commit was cherry-picked from Qt 5 to 4.8 after agreeing with the
author because the merge request was filed against Qt 5.)
|
|
|
|
|
|
|
|
|
|
|
|
| |
This reverts commit cf17b743d2fe84ab259b7232ab07b58a1872e18e, which
changed QMutex to use a Mach semaphore_t on Mac OS X. We now use
pthread_mutex_t on Mac OS X as well. Contention performance is mostly
unchanged, but the new constructionQMutex() benchmark added in this
commit shows that creating/destroying a semaphore_t is about 20 times
slower than pthread_mutex_t.
Reviewed-by: Olivier Goffart
Reviewed-by: João Abecasis
|
|
|
|
|
|
|
|
| |
Working directory (as opposed to applicationDirPath) was
on the Z drive before, change it to be the same.
Task-Number: QTBUG-22024
Reviewed-By: mread
|
|
|
|
|
|
|
|
| |
For the addition of QApplicationPrivate::symbianHandleLiteModeStartup.
Two other unfrozen functions also added.
Task-number: QT-4933
Reviewed-by: TrustMe
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
QtAppBooster can partially start apps and hide them in the background
before they are requested by the user. This gives an apparently faster
start for these apps.
Qt is now detecting these app starts and bringing the construction of
the status pane and softkeys forward so that they happen in the
background. This cuts their construction time off the time between the
user clicking on the app and it being ready to use. On a Nokia E7, this
gain was measured at an average of 127ms.
Task-number: QT-4933
Reviewed-by: Sami Merila
|
|
|
|
|
|
|
|
|
| |
The native side seems to return StaCon pane height as zero in Belle.
If no StaCon pane height is available, try to fetch Status Pane height
and use that one.
Task-number: QTBUG-22022
Reviewed-by: Miikka Heikkinen
|
|\ |
|
| |\
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
master-integration
* 'master' of scm.dev.nokia.troll.no:qt/qt-fire-staging:
Fixes: libpng symbols exported from QtGui.dll on Windows
omit unassigned (and too recent codepoints) from the text Normalization process
Fixed broken window surface flush when depth is 24 and bpp is not 32.
|
| | |\ |
|
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
> handle any code points that were not defined in the earlier version
> as if they were unassigned: such code points will not decompose or compose,
> and their Canonical_Combining_Class value will be zero.
see http://www.unicode.org/reports/tr15/#Guaranteeing_Process_Stability for details
we should respect that the QChar::Unicode_Unassigned == 0
and thus less than any other QChar::UnicodeVersion value
Merge-request: 952
Reviewed-by: Jiang Jiang <jiang.jiang@nokia.com>
|
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
Some X servers use a compact representation of 24 depth visuals. In that
case we can't use the shared memory or XPutImage paths, as Qt's RGB32
does not match the internal memory layout.
Also fixed QPixmap::fromImage() to work in this case to prevent the red
and blue channels from being swapped.
Task-number: QTBUG-21754
Reviewed-by: Alberto Mardegan
|
| | |/
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
The upgrade of the bundled libpng to version 1.5 had the side effect
of causing the png functions to be exported from the QtGUi DLL
when built with the MS Visual C++ compiler.
Task-number: QTBUG-21455
Reviewed-by: Kim
|
| |\ \
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
Conflicts:
doc/src/getting-started/installation.qdoc
doc/src/platforms/platform-notes.qdoc
src/corelib/tools/qlocale_symbian.cpp
src/gui/kernel/qwidget_p.h
src/network/access/qnetworkaccesshttpbackend.cpp
src/opengl/qgl.cpp
src/plugins/bearer/symbian/qnetworksession_impl.cpp
|
| | |\ \
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
4.7-integration
* '4.7' of scm.dev.nokia.troll.no:qt/qt-s60-public:
Do not let QGLContext get out of sync when doing mixed VG-GL rendering
|
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
The default graphics system is openvg, this means that apps using
QGLWidget (despite the official recommendation), without explicitly
setting opengl as the graphics system will use both the openvg and
opengl modules to perform rendering. This presents some problems
because QtOpenGL's comes with its own wrapper over QEglContext,
and making a foreign QEglContext (e.g. from VG) current will cause
its state to get out of sync, resulting in various drawing issues
due to skipped makeCurrent and similar calls.
The patch works this around by querying the current context from
EGL in QGLContext::currentContext() and returning null if the
EGLContext does not match. This somewhat ugly workaround ensures
that at least QGLContext::currentContext() will not lie: It will
not return a valid pointer when the underlying QEglContext's
underlying EGLContext is not actually the current context. This
will allow e.g. QGLPaintDevice::beginPaint to correctly recognize
the need for making the GL widget's context current and thus
avoid various rendering issues in certain games.
Task-number: QTTH-1553
Reviewed-by: Jani Hautakangas
|
| | |\ \ \
| | | |/ /
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
4.7-integration
* '4.7' of scm.dev.nokia.troll.no:qt/qt-s60-public:
Create auto test for http HEAD request
QNAM - fix poor performance of HEAD request with authentication
|
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
QHttpNetworkReply was waiting for a body to be sent for 401 and 407
responses, whereas with a HTTP HEAD request, there will be no body.
This delayed the authentication signal until after the http channel
is closed by the server after a timeout. For example with the server
used for autotesting, the authentication signal is delayed 15 seconds.
When the server has a very long timeout, the authentication signal may
not be emitted at all.
Task-Number: QT-5304
Reviewed-By: Martin Petersson
|
| | |\ \ \
| | | |/ /
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
4.7-integration
* '4.7' of scm.dev.nokia.troll.no:qt/qt-s60-public:
symbian - add support for ELangKazakh and ELangEnglish_India
|
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
I have not cherry picked this from 4.8 because it's tied up with a
feature implementation that refactored the code somewhat.
Task-number: QT-5298
Reviewed-by: mread
|
| | |\ \ \
| | | |/ /
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
4.7-integration
* '4.7' of scm.dev.nokia.troll.no:qt/qt-s60-public:
QAuthenticator::setUser() parse user name in form user@domain
|
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
Task-number: QTBUG-19894
Change-Id: I063dbc66e5f47a83cc1c0aee8913062b4b5e42bc
(cherry picked from commit d3313bc2f095dc15718b2a1ed613489ab1ae0927)
|
| | |\ \ \
| | | |/ /
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
4.7-integration
* '4.7' of scm.dev.nokia.troll.no:qt/qt-s60-public:
A patch for 'Fix to QtOpenGL crash'
Always recreate backing store when TLW transparency changes
Crash in QDeclarativeCompiler::indexOfProperty
Fix to QtOpenGL crash
|
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
EGL API was used also on non-EGL Symbian platforms
which generated compile errors when compiling Qt
on those platforms. This patch #ifdefs EGL usage
correctly for EGL enabled platforms only.
Reviewed-by: TRUSTME
|
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
When using either the opengl or openvg graphics system on Symbian,
if a TLW is made transparent, the widget switches to using raster
rendering, because EGL surface transparency is currently not
supported by the platform.
This patch enables the reverse to occur: when the widget is
subsequently made opaque, rendering switches back to using GL/VG.
Task-number: QTBUG-21211
Reviewed-by: Jani Hautakangas
Reviewed-by: Laszlo Agocs
Reviewed-by: Sami Merila
|
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
In QDeclarativePropertyCache, plain integer bitfield overrideIndex is
initialized with -1 in class constructor. Unfortunately, ARM compiler
treats bitfields as unsigned, unless explicitly defined as signed [1].
Therefore, overrideIndex actually gets initial value of 2147483647,
which causes array operations done with the index to fail.
As a fix, define overrideIndex as signed int bitfield.
[1] http://www.keil.com/support/man/docs/armccref/armccref_Babjddhe.htm
Under bitfields/Note: "A plain bitfield, declared without either
signed or unsigned qualifiers, is treated as unsigned"
Task-number: QT-5285
Reviewed-by: Aaron Kennedy
|
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
Exiting the native video recorder on Symbian and going
back to Qt app which runs on opengl graphics system
crashed on some devices because there was not enough GPU
memory and also because low GPU mem device environment
detection was flawed. This patch fixes 32MB GPU memory
detection and adds wait/retrial to EGL surface creation
if it fails.
Task-number: QTBUG-21499
Reviewed-by: Laszlo Agocs
|
| | |\ \ \
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | | |
4.7-integration
* '4.7' of scm.dev.nokia.troll.no:qt/oslo-staging-1:
Incorrect 'break' in switch-case was replaced by 'return'
|
| | | | | | |
|
| | |\ \ \ \
| | | | |/ /
| | | |/| |
| | | | | |
| | | | | |
| | | | | |
| | | | | | |
4.7-integration
* '4.7' of scm.dev.nokia.troll.no:qt/qt-s60-public:
QLineControl: selectedTextChanged() signal is omitted in some cases
|
| | | |\ \ \ |
|
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | | |
QLineControl does not emit selectionChanged() signal when
selection is changed due to handling of input method query.
As a fix, the signal is emitted when internal values for selection
end and beginning are changed in processInputMethodEvent().
Additionally, added deselect() test case for both QDeclarativeTextEdit
and QDeclarativeTextInput to help out autotesting of text selections.
Note that there are some differences in the QLineControl and QTextControl
implementations. This causes the different signal counts even if the
test steps are similar.
Task-number: QTBUG-21429
Reviewed-by: Miikka Heikkinen
|
| | |\ \ \ \ \
| | | | |/ / /
| | | |/| | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | | |
4.7-integration
* '4.7' of scm.dev.nokia.troll.no:qt/qt-s60-public:
QML import path puts Qt dirs in correct Symbian order
|
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | | |
The QML import path contains an entry for the standard
import location on each drive on Symbian. These were
being ordered so that Z:, the ROM drive, was listed
first. In fact the drive order was the reverse of the
normal Symbian drive search order. This had the effect
that upgraded versions of QML imports installed on
the device were generally ignored for older ROM based
versions.
The import path has now been changed to match the normal
Symbian drive search order. This will make it more likely
for upgraded QML imports to be picked up and used.
Task-number: QTBUG-21409
Reviewed-by: Shane Kearns
Reviewed-by: Martin Jones
|
| | |\ \ \ \ \
| | | | | | | |
| | | | | | | |
| | | | | | | |
| | | | | | | |
| | | | | | | |
| | | | | | | | |
4.7-integration
* '4.7' of scm.dev.nokia.troll.no:qt/oslo-staging-2:
Avoid runtime warning about interlace handling from libpng 1.5
|
| | | | | | | |
| | | | | | | |
| | | | | | | |
| | | | | | | |
| | | | | | | | |
Task-number: QTBUG-13298, QTBUG-21408
(cherry picked from commit c8851ce8e109fb6b403ae80036d9f319f1b69d7a)
|
| | |\ \ \ \ \ \
| | | | |/ / / /
| | | |/| | | |
| | | | | | | |
| | | | | | | |
| | | | | | | |
| | | | | | | | |
4.7-integration
* '4.7' of scm.dev.nokia.troll.no:qt/qt-s60-public:
symbian bearer: fix tst_qnetworksession test failures
|
| | | | |/ / /
| | | |/| | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | | |
Symbian is not reporting the expected KLinkLayerClosed event on the
last RConnection handle. Instead, it reports the event
KConfigDaemonFinishedDeregistrationStop, which indicates the nif
plugin is being destroyed (this is followed by unload events)
To fix, this event is treated the same as a link down event, and
results in the state change to disconnected.
Either there has been a behaviour change in the platform, or the
test was previously run with another user of the API running in the
background (e.g. WLAN wizard, connection monitor) which prevented
the unload.
Task-Number: QTBUG-21382
Reviewed-By: mread
(cherry picked from commit a0ad54acd33873316908080f432768fdf3e3bfc2)
|
| | |\ \ \ \ \
| | | | |_|/ /
| | | |/| | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | | |
4.7-integration
* '4.7' of scm.dev.nokia.troll.no:qt/oslo-staging-1:
Add missing checks for httpReply in QNetworkHttpBackend.
|
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | | |
httpReply is not guaranteed to be non-null. Adding missing checks for methods
that can be called from outside the class.
Reviewed-By: Peter Hartmann
|
| | |\ \ \ \ \
| | | | |/ / /
| | | |/| | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | | |
4.7-integration
* '4.7' of scm.dev.nokia.troll.no:qt/qt-s60-public:
Refactor dangerous multiple inheritance
Fix RConnection handle leak in symbian bearer plugin
Prevent crash when cache is changed on the fly
|
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | |
| | | | | | | |
QObject and CBase both expect to be the root class of the object hierarchy
so it can cause problems if they are used in multiple inheritance.
Refactored the CActive used for starting RConnection into a helper class.
Reviewed-by: Markus Goetz
(cherry picked from commit 5a6365f14006ab50854e41a5927645c7e9966756)
|