| Commit message (Collapse) | Author | Age | Files | Lines |
|\
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
into master-integration
* 'staging-master' of scm.dev.nokia.troll.no:qt/qt-lighthouse: (101 commits)
Fixes detected by the static-check
Fixup license headers in Lighthouse plugins
Mac: add autorelease pools to create_sys
Don't skip remaining events on unhandled events in xcb backend.
Lighthouse: Xcb, compile fix for the glx path
Lighthouse: Make sure we don't use xkbcommon
Lighthouse: Add pkg-config check for opengl es2
Rename qnaclunimplemented -> qfunctions_nacl
Rename qconfig-minimal-system-depenencies.h
Add QT_NO_FILESYSTEMITERATOR.
Add QT_NO_DYNAMIC_LIBRARY.
Compile.
NaCl: Dont force release builds.
NaCl: disable libtiff.
posix platformdefs: Add QT_NO_SOCKET_H.
qpnghandler: Compile with QT_NO_IMAGE_TEXT.
Lighthouse: Support Q_NO_CLIPBOARD.
Compile.
Fix CoreServices framework addition for qpa.
Update NaCl mkspecs.
...
|
| |
| |
| |
| |
| |
| | |
This included added a license header on a moc file. Its nasty, but has
to stay like that until we have a better sollution for generating moc
files on uikit
|
| | |
|
| |
| |
| |
| | |
Fixes a couple of on-startup leaks.
|
| | |
|
| |\
| | |
| | |
| | |
| | | |
Conflicts:
mkspecs/qws/macx-nacl-g++/qplatformdefs.h
|
| | | |
|
| | |
| | |
| | |
| | | |
when we don't have it
|
| | | |
|
| | |
| | |
| | | |
(Match the existing files for wince and vxworks.)
|
| | |
| | |
| | |
| | |
| | | |
Rename to qconfig.nacl.h. This reflects that the
config now enables all dependencies NaCl can
support, not necessarily the minimal set.
|
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
This makes it possible to use the file-related classes
on platforms that does not have a native QFileSystemIterator
implementation.
One use case is that we want to use the Qt
resource system even if the platform file system
support is limited.
Reviewed-by: joao
|
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
Set it for nacl, replace the Q_OS defines vxworks
already has in place.
Motivation: Support static plugins on platforms
that does not support dynamic plugins.
Static plugin support is implemented in QPluginLoader,
which has a QLibraryPrivate d pointer. The easiest
way to untangle this seems to be to compile in
QLibrary and then disable the ports of it that uses
dlopen.
Reviewed-By: Harald Fernengel
|
| | |
| | |
| | |
| | |
| | | |
Respect the system-zlib QT_CONFIG setting.
Reviewed-by: Jørgen Lind
|
| | | |
|
| | |
| | |
| | | |
(libtiff doesn't compile)
|
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
Prevents including <sys/socket.h>
Reviewed-by: Jørgen Lind
|
| | | |
|
| | |
| | |
| | | |
Reviewed-by: Jørgen Lind
|
| | | |
|
| | |
| | |
| | |
| | |
| | |
| | | |
Test against the coreservices QT_CONFIG on qpa,
unconditionally add CoreFoundation for mac and
darwin
|
| | |
| | |
| | |
| | |
| | | |
Less duplication, use the common unix and gcc
conf files.
|
| | | |
|
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
Make sure to also flush output when waiting for an ack from the
compositor, in case we haven't sent the frame request yet.
Reviewed-by: Jørgen
|
| | | |
|
| | | |
|
| | | |
|
| | |
| | |
| | |
| | | |
See the README in the uikit platform plugin for details.
|
| | |
| | |
| | |
| | | |
Reviewed-by: Jørgen Lind
|
| | |
| | |
| | |
| | |
| | | |
Reviewed-by: Jedrzej Nowacki
Reviewed-by: Simon Hausmann
|
| | |
| | |
| | |
| | | |
Reviewed-by: Morten Johan Sørvig
|
| | |
| | |
| | |
| | |
| | | |
in xcbplugin. I also reoredered how eglInitialize was executed,
(its not so lazy anymore)
|
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
Cocoa.h can't be included after qscroller_p.h, but we need to have
Q_WS_MAC defined before the ifdef.
Reviewed-by: Morten Johan Sørvig
|
| | |
| | |
| | |
| | |
| | |
| | | |
QPA sets CaseSensitive unconditionally, so if OS_MAC tries to use
CaseInsensitive QSettings will assert. The patch leaves the non-QPA
case unchanged.
|
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
Processing posted Qt events before the event loop integration is started
may lead to the process being killed for taking too long time before the
native event loop is running.
Since event loop integrations are required to install a timer that does
processEvents anyhow, the Qt events will be processed even without doing
it explicitly in advance.
|
| | | |
|
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
The initialial readable iteration, needs to be done before
initializing egl, or it will be done inside egl.
The readable iteration after initialzing egl would block
forever otherwise.
|
| | | |
|
| | | |
|
| |\ \ |
|
| | | |
| | | |
| | | |
| | | | |
Reviewed-by: paul
|
| | | |
| | | |
| | | |
| | | | |
Reviewed-by: Jørgen
|
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
Add qmake.conf variables, and use the generic font plugin.
Reviewed-by: Jørgen
|
| | | | |
|
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
We realised that we couldn't keep on adding virtual functions for all
specific needs so we made one virtual function which takes a string
parameter which enables the implementation to return anything.
It seems a bit odd to have a virtual function as an accessor for a class
with only 1 virtual function. But the idea is that we don't want a
void * in the core lighthouse api. Also, the implementations should
decleare more functions to do the heavy lifting of getting the handles
Reviewed-by: paul
|
| | | | |
|
| | | |
| | | |
| | | |
| | | | |
Reviewed-by: Jørgen Lind
|
| | | | |
|
| |\ \ \ |
|
| | | | | |
|