It was not possible to extract data from the archive on OSX
which was created on Windows platform because of wrong
separators. Archive was created on Windows via QZipWriter and
opened on OSX with QZipReader. It consisted of a lots directories
and subdirectories with files. The solution is to use '/' separator
for internal representation.
Change-Id: Ic0837ca184bb6188129d53b587a5df2ec61e4e05
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@digia.com>
Activate the window of the widget under mouse pointer before replay mouse
press event.
Change-Id: I9e699374accf108aa49b2a3c73d5e76631100dfd
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@digia.com>
Reviewed-by: Samuel Rødal <samuel.rodal@digia.com>
Reviewed-by: Gunnar Sletta <gunnar.sletta@digia.com>
When QProcess->start() is called, Qt creates a pipe to the process to
get its exit value and output.
It does this with qt_create_pipe, which calls qt_safe_pipe.
qt_safe_pipe, on failure, returns 1. qt_create_pipe then
sets errno and returns void.
The calling function, QProcessPrivate::startProcess, does not check
errno, and thus continues to fork the process, assuming the pipe has been
created successfully.
The child process then has no way to pass its exit value to the calling
process, since the communication pipes it would normally use do not
exist, and thus when it exits it becomes a zombie.
As a bonus, if waitForFinished is called on a broken process, a crash
results because it is trying to wait on a pipe which does not exist.
The fix makes qt_create_pipe return an integer, and QProcess::startProcess
check the return value, set processError and not create the child
process.
Task-Number: QTBUG-18934
Change-Id: I2e1effdd0617be5b8c5492bcbcf5f2b1584b2241
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Building against the local /System can cause build issues when for example
the headers have not been updated to reflect the system version. The system
headers are updated as part of installing the command line tools from within
Xcode, not as part of the system update process, so we might think we are
on 10.8, but the system headers will not reflect that, and we get build
breaks. It's preferable to always build against an SDK, so that we have
a known state for the OS X libraries and headers.
We choose the latests SDK by default, as recommended by Apple.
Change-Id: I79028217ff3a9cbe45aa4cb05ed6dd90388dee50
Reviewed-by: Tor Arne Vestbø <tor.arne.vestbo@digia.com>
And update test to use the clang makespec now that it's the default.
Change-Id: Ifdd34c4220ad76f60b91fd6ef39d189f0f6525f9
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@digia.com>
Add an encrypted signal to QNAM and QNetworkReply to allow applications
to perform additional checks on the certificate chain beyond those done
as part of the standard SSL validation. This allows things like
certificate change notification to be implemented for QNAM as they can
be for QSSLSocket currently.
Change-Id: I693e3e6fec8b7040379b7e7f1f819550e6b2617f
Reviewed-by: Peter Hartmann <phartmann@rim.com>
Add intermediate certificates to our server sockets, and to our client
certs.
Change-Id: Ib5aa575473f9e84f337bebe35099506dd7d7e2ba
Task-Number: QTBUG-19825
Task-Number: QTBUG-13281
Reviewed-by: Peter Hartmann <phartmann@rim.com>
Instead of storing a single QSslCertificate for a the local cert, store
a list of them. This will allow us to handle server sockets that use a
certificate that is not issued directly from the CA root in future.
Change-Id: I9a36b9a99daa9c0bdd17f61b4ce1a7da746f2e96
Reviewed-by: Peter Hartmann <phartmann@rim.com>
QEglFSWindow does not request highestPixelFormat, and can obtain a 16-bit
config when requested. QEGLPlatformContext does request highestPixelFormat,
and can obtain a 32-bit config when a 16-bit config is requested, leading
to this error.
Change-Id: I418c09a35ec19b2e9ca372b32599034e02384e44
Reviewed-by: Samuel Rødal <samuel.rodal@digia.com>
The "mediumfile" is actually 9.6 MB large, and makes the test slow and
even fail on mobile platforms.
Change-Id: I694d9d38e4d09c8e90a03c1141320fe2a757fdd9
Reviewed-by: Richard J. Moore <rich@kde.org>
Replace the old sed / template @FOO@ method with echo.
Enable MSYS bash to build qmake.exe
Use qmake/Makefile.unix for all win32-g++ builds.
Change-Id: I6e27d69b28d27131838bbbb3a4ee5a08b470f31b
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@digia.com>
No users/file permissions. Open function in VxWorks DKM
requires always three parameters.
Change-Id: I93fb075d82aa57e210e224bfe4ede40def82c275
Reviewed-by: Samuel Rødal <samuel.rodal@digia.com>
By default we still use QIconLoaderEngine but now platform theme plugins
have the opportunity to override that. It is in particular planned to be
used in a WIP platform theme plugin for KDE sessions.
Change-Id: I07a82dc91daea44709b3a790f3f6e2a7a090d108
Reviewed-by: Olivier Goffart <ogoffart@woboq.com>
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@digia.com>
Reviewed-by: David Faure (KDE) <faure@kde.org>
Reviewed-by: Samuel Rødal <samuel.rodal@digia.com>
Reviewed-by: Frederik Gladhorn <frederik.gladhorn@digia.com>
If FBIOGET_VSCREENINFO doesn't give us sensible values we need to
default to something instead. Refactor the code that queries the
resolution and depth to behave the same way.
Change-Id: Id2b3fc41349a74610856273b10281f744612890b
Reviewed-by: Gunnar Sletta <gunnar.sletta@digia.com>
Instead of setting -isysroot in both arch.test, compile.test, the various
mkspecs, and sdk.prf, we now propgate the chosen SDK as the qmake
variable QMAKE_MAC_SDK, which is then handled exclusivly in sdk.prf.
The QMAKE_MAC_SDK variable, and -sdk argument to configure, is expected
to be of the short-form name, eg macosx or iphoneos, not a full path, as
that's what Xcode also expects. We take care of translating that into
a full path for -isysroot/-syslibroot in sdk.prf, using xcodebuild as
a helper.
Change-Id: I281655b2fa5180c6e78ffdce36824e4a91447570
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@digia.com>
Exhausting the symbol list while looking for the
final right parenthesis means it is missing.
Task-number: QTBUG-29308
Change-Id: Iccf5897b0f5eb719699fd12d6c8e4a16ff189d9b
Reviewed-by: Simon Hausmann <simon.hausmann@digia.com>
We no longer support OS X < 10.6, so there's no need to check for it.
Change-Id: I2628984846de0c0c19ea86b3ba6d00fc370ddae7
Reviewed-by: Tor Arne Vestbø <tor.arne.vestbo@digia.com>
Instead of after parsing the makespecs, were we assumed QMAKE_PLATFORM
had been set by the makespec and did an explicit update.
Allows loading platform specific features from within a makespec after
updating QMAKE_PLATFORM.
Change-Id: I0eb3b7fb88ce48b00a384850c5d87223c06234d7
Reviewed-by: Tor Arne Vestbø <tor.arne.vestbo@digia.com>
So that mkspecs and features may rely on the host_build test.
Change-Id: I18fee4820d9e2904285afcc7ddb8f1cc3d025fef
Reviewed-by: Richard Moe Gustavsen <richard.gustavsen@digia.com>
Reviewed-by: Tor Arne Vestbø <tor.arne.vestbo@digia.com>
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
The function we use to get the actual vsync on cocoa is documented
to return 0 if the monitor is not a CRT monitor. A refreshrate
of 0 means we have vsync deltas of 1000/0 which cause problems
elsewhere. It is better to use the "default" value in this case
as it will be closer to correct than 0.
Change-Id: Id08007e40a9af5e42f13a07628fcad5fd3a7d0dc
Reviewed-by: Morten Johan Sørvig <morten.sorvig@digia.com>
Reviewed-by: Samuel Rødal <samuel.rodal@digia.com>
Inherits QSurface and allows to use OpenGL from an arbitrary thread.
Platform plugins can implement QPlatformOffscreenSurface, otherwise an
invisible QWindow is used by QOffscreenSurface.
This patch includes an implementation of QOffscreenSurface for XCB
and EglFS platform plugins using pbuffers.
Change-Id: I57b4fc1db417331f34826dcfa754b7698782fde4
Reviewed-by: Lars Knoll <lars.knoll@digia.com>
Reviewed-by: Samuel Rødal <samuel.rodal@digia.com>
This should select the best mode (likely the currently running mode)
for the display instead of the first one found.
The built-in mode was left as a fallback.
Change-Id: I4e1bc798df6f310b001566ab76cb9def3224a7ed
Reviewed-by: Andy Nichols <andy.nichols@digia.com>
GBM provides a way to write directly to a buffer object that gets set
as the drmMode cursor. This eliminates the need to create a GL texture
and opens this class up to platforms that support GBM but not OpenGL.
Change-Id: I7297827387ef9a717a5287b5484f14c9987b4158
Reviewed-by: Andy Nichols <andy.nichols@digia.com>
- Use qtHaveModule to check for OpenGL
- Add __GBM__ to pick up the correct native types in the Mesa EGL headers
Change-Id: Idfc0e81e95672b08ba8f259b9d7edf2b25fd1bad
Reviewed-by: Andy Nichols <andy.nichols@digia.com>
A QGraphicsProxyWidget embeds a focusable widget (e.g., QComboBox). When
deleting QGraphicsProxyWidget, the QWidget will be deleted. The QWidget
clears focus, and QWidget::hasFocus() is nice enough to check if its
embedder QGraphicsProxyWidget has focus - because if it does, it wants
to clear focus from that item too. QGraphicsItem's destructor already
calls clearFocus() however, so this call is unnecessary; we can simply
stop clearing the QWidget's focus in its destructor if the widget is
embedded.
QWidget::hasFocus checks QGraphicsItem::hasFocus (on the proxy widget
that is being deleted), which checks its d_ptr, which is gone. It's
generally unfavorable for an object deleting a child to have the child
go back and poke at the parent object, which is in many ways what's
happening here.
Task-number: QTBUG-29684
Change-Id: I1e52bf28f47b2824752de28dff2d0de13733ee48
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@digia.com>
The crash is deep inside QGraphicsSceneIndex, which calls boundingRect()
on the item that is being destroyed. The vtable is busted, resulting in
a pure virtual function call. There's a more proper fix for this lying
around somewhere but in this particular case we can get the test to
not crash by guarding based on whether the item has a cursor set. This
also happens to speed up QGraphicsItem destruction a bit so I figured
it's a win-win situation to fix it. This case will still crash if the
item actually had a cursor set but that makes the case even more narrow.
Generally speaking, creating objects partially on the stack and
partially on the heap, mixing parent/child relationships and then
deleting one of the heap objects is quite sketchy and I doubt it happens
much outside of this unit test.
Change-Id: I25393d2cafb1256269ab6681519bd554cc380bfd
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@digia.com>
The former was introduced with Mac OS 10.6 and can also be used
for checking iOS versions, so it's preferable. We still include
the old availability header, and use it in various places in Qt,
and so does the Mac OS frameworks, so there's no need to phase
it out, but for new platform checks we want to use the updated
macros of the form:
#if __MAC_OS_X_VERSION_MAX_ALLOWED > __MAC_10_7
Ideally you should not use the named version macro, and use 1070
instead, in case you build against an SDK that does not define
the named version yet, but we take care of defining these
in qsystemdetection.h for convenience.
Change-Id: I9cfa72e37816583f28ff9643793f111e155b7789
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
AvailabilityMacros.h will pick up the MACOSX_DEPLOYMENT_TARGET environment
variable, as well as the -mmacosx-min-version= command line flag, and
set the MAC_OS_X_VERSION_MIN_REQUIRED based on that. By setting the
define before including AvailabilityMacros.h we essentially skipped
that whole logic and always set it to 10.6.
Only in the case where there's no deployment target specified on the
command line do we want to ensure that it's at least 10.6
Change-Id: Ic558ff4deb77937ea805b048d83949815b273bcc
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Qt5 requires Mac OS 10.6, so we can remove checks such as
if MAC_OS_X_VERSION_MAX_ALLOWED >= MAC_OS_X_VERSION_10_6
Change-Id: Iea21727a277291148704ecf9677ed0b68c24920f
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
J-P Nurmi
Gabriel de Dietrich
Pavel Mogylevskiy
jian liang
Oliver Wolff
Friedemann Kleint
Simeon Bird
Tor Arne Vestbø
Tor Arne Vestbø
Richard Moore
Michael Brasser
Peter Hartmann
Erik Verbruggen
Ray Donnelly
Pasi Petäjäjärvi
Kevin Ottens
Samuel Rødal
Takumi ASAKI
Keith Gardner
Frederik Gladhorn
Shawn Rutledge
Gunnar Sletta
Yoann Lopes
anknight
Andreas Aardal Hanssen