| Commit message (Collapse) | Author | Age | Files | Lines |
|
|
|
|
|
|
|
|
|
| |
Removes qmake as a build system for building Qt Creator itself.
Keep them for some tests that are not completely moved to CMake yet.
Change-Id: I846c6ef65626b6dfae6375fdc85d00677aa8c2fb
Reviewed-by: hjk <hjk@qt.io>
Reviewed-by: Christian Kandeler <christian.kandeler@qt.io>
Reviewed-by: Christian Stenger <christian.stenger@qt.io>
|
|
|
|
|
|
|
|
|
|
|
| |
Don't install qtcreator.sh, it cannot be used.
Don't install application into bin/ subfolder. App-bundles do not follow
the UNIX bin/, lib/, share/ convention, so don't even try. (This makes
bundling it into a 7z consistent with what we do on the other
platforms.)
Change-Id: Ib3faa2909a9b30a11193586f279528f1b8bd96fa
Reviewed-by: Cristian Adam <cristian.adam@qt.io>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Based on Tobias Hunger's work from a few months ago.
The CMake configuration needs libclang and Qt paths specified as
CMAKE_PREFIX_PATH.
Auto tests are run with "ctest". At the moment the pass rate is 87%.
Change-Id: Iba98e39bf22077d52706dce6c85986be67a6eab0
Reviewed-by: Alessandro Portale <alessandro.portale@qt.io>
Reviewed-by: Tobias Hunger <tobias.hunger@qt.io>
Reviewed-by: Eike Ziller <eike.ziller@qt.io>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
We used to set QT_OPENGL, currently set LD_LIBRARY_PATH when qtcreator.sh
is used, and might set other variables in the future,
but these environment modifications should not be passed on
to user applications or when we run tools that are not shipped with
Qt Creator.
For LD_LIBRARY_PATH there already was a hack for
Environment::systemEnvironment.
For environment variables that we might set in main()
in the future, this patch caches the system environment for
Environment::systemEnvironment early before any modifications are made.
The previous hack for LD_LIBRARY_PATH no longer works, since it
used QCoreApplication::applicationDirPath() which cannot be used at that
point in time (before the QApplication has been created). Instead pass
the correct user LD_LIBRARY_PATH directly from qtcreator.sh to
Qt Creator through a command line option, which is cleaner anyhow.
Task-number: QTCREATORBUG-20808
Change-Id: I6674a5e0537e1b37fd7dcbff371b542fa24bce69
Reviewed-by: Ulf Hermann <ulf.hermann@qt.io>
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@qt.io>
Reviewed-by: Tobias Hunger <tobias.hunger@qt.io>
|
|
|
|
|
|
|
|
|
|
|
| |
it's pointless and causes an unnecessary delay.
what's worse, we wouldn't actually know whether we found a compatible
(gnu) readlink, so the fallback would never get activated. this would be
the case on macos (the problem wasn't noticed so far, because on macos
we build a bundle which doesn't need the wrapper in the first place).
Change-Id: Ibf4c370d5d9e25064e80faeaf12eb5fa1bc68c20
Reviewed-by: Eike Ziller <eike.ziller@qt.io>
|
|
|
|
|
|
|
|
| |
the base directory for making the symlink target absolute is of course
the dirname of the original file, not the target itself.
Change-Id: If5171ead8da0aa2afcaa8ccd6365771441440146
Reviewed-by: Eike Ziller <eike.ziller@qt.io>
|
|
|
|
|
|
|
|
|
|
|
| |
We need to deploy Qt 5.6 into its own directory (lib/Qt/) in the
Qt Creator packages, so the LD_LIBRARY_PATH that is set in
qtcreator.sh needs that too.
Task-number: QTCREATORBUG-15748
Change-Id: I637322dfe5eb669b6447aa2f2b52e3ba2fe2979f
Reviewed-by: Edward Welbourne <edward.welbourne@theqtcompany.com>
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@theqtcompany.com>
|
|
|
|
|
|
|
|
|
|
| |
Define variables in qtcreator.pri for the various
install locations, instead of hardcoding the paths
everywhere where they are needed.
Change-Id: Ia34de711a7f2be317272ede023e2a8b9cdcc30fe
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@theqtcompany.com>
Reviewed-by: Christian Kandeler <christian.kandeler@theqtcompany.com>
|
|
|
|
|
|
|
| |
it's the correct variable to use.
Change-Id: I9ec7795099097715a64969516cd3fae9e213a0f4
Reviewed-by: Daniel Teske <daniel.teske@digia.com>
|
|
|
|
|
|
|
| |
same trick as in static.pro
Change-Id: I1d3c30f8d03179fcff1fa463bf7f44af0290f4f0
Reviewed-by: Eike Ziller <eike.ziller@digia.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
'make install' now installs to $(INSTALL_ROOT)$$QTC_PREFIX/...
This is used for giving the contents of the 7zips an additional prefix.
(previously done by doing an additional copying step in bindistHelper).
QTC_PREFIX can also be used to give Qt Creator a different install path
at qmake time, and defaults to /usr/local on Linux.
On Windows and Mac there is no default for QTC_PREFIX.
Usage: qmake -r QTC_PREFIX=/qtcreator-2.6.0 && make &&
INSTALL_ROOT=/tmp/creator-dist make install
Change-Id: Id30781e14bfdde52531800f22b22e39f0459e806
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@digia.com>
|
|
|
|
|
|
|
|
|
|
| |
It depends on and mixes in changes made in another change on gerrit
that is not merged, so it breaks the package build.
This reverts commit c29bf6f6525d435d69936576761b4a29102cd056.
Change-Id: Ibb251150909271f3e119f05a1691832aae8ac633
Reviewed-by: Eike Ziller <eike.ziller@nokia.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Reason: $INSTALL_ROOT is only meant to be used by packagers
to temporarily put the contents into a different location,
which is needed for fakeroot packaging.
QTC_PREFIX is not a qmake variable, and defaults to
/usr/local. On Windows the default prefix is "QtCreator",
since "make install" is expected to be used in
a packaging context only where either INSTALL_ROOT
should be used or QTC_PREFIX should be set to the
absolute destination path where e.g. an installer
generator will pick it up.
Change-Id: Ifa4950340e58e34726c53f5417adcc7b50828ce1
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@nokia.com>
|
|
|
|
|
|
|
|
|
| |
yet another hack
Change-Id: Ic041e8bdfb7331d6934620c8fbb91435f3495ba7
Reviewed-on: http://codereview.qt-project.org/4790
Reviewed-by: Qt Sanity Bot <qt_sanity_bot@ovi.com>
Reviewed-by: Daniel Molkentin <daniel.molkentin@nokia.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
we need to override RUNPATHs from possible plugins.
due to d7d23226, this should not re-introduce QTCREATORBUG-1646.
unlike in the previous incarnation, we don't replace the binary with the
script, but give the script an extension. this is nicer for packagers
and less confusing in creator's own run configurations. the .desktop
files created by our installers need to be adjusted for that.
Task-number: QTCREATORBUG-5565
Change-Id: Icd2fa55456754a05257376b8288e8bdf423c62db
Reviewed-on: http://codereview.qt.nokia.com/3180
Reviewed-by: Qt Sanity Bot <qt_sanity_bot@ovi.com>
Reviewed-by: Daniel Molkentin <daniel.molkentin@nokia.com>
Reviewed-by: Christian Kamm <christian.d.kamm@nokia.com>
|
|
|
|
|
|
|
|
| |
Christian Kamm convinced me that it should not have any effect.
Reviewed-By: ckamm
Task-Nr: QTCREATORBUG-1646
|
|
|
|
| |
Task-Nr: QTCREATOR-116
|
|
|
|
| |
./bin/qtcreator: 3: Syntax error: "(" unexpected
|
|
|
|
|
|
|
|
|
| |
First, there's no $( ) for subshells. You have to use backticks.
Second problem, "VAR=value exec command" doesn't work. Split into
variable setting, exporting and then exec'ing the program.
Third, Solaris doesn't come with readlink, so parse the output of ls -l.
|
| |
|
| |
|
|
|
|
|
|
|
|
|
| |
now bin and share are not pegged to linking the executable any more,
but are proper subdirs targets of their own.
unforturnately, qmake is a bit tenacious, so we hack around a bit. :)
more or less in passing by, i removed some superfluous trailing slashes
and added path normalization as well.
|
|
|
|
|
|
| |
Bug reported and fix suggested by catanzag on Qt Labs blog. Thanks!
Reviewed By: Thorbjorn Lindeijer <thorbjorn.lindeijer@nokia.com>
|
| |
|
| |
|
| |
|
|
|
|
| |
RevBy: hjk
|
| |
|
|
|
|
| |
Details: LD_LIBRARY_PATH setting still missing
|
|
|
|
|
|
|
| |
Details: Moving files so that the local layout corresponds to the layout pushed by make install.
Adjust Make target, modify the resource path of the loader to find the resources in the
right place. All pathes are still looked up relative to binaryPath(), libs are found via
relative rpath.
|
| |
|
| |
|
| |
|
| |
|
| |
|
| |
|
|
|
|
| |
to the "user space" (gdbmacros.cpp)
|
| |
|
|
|
|
| |
qualification like 'const char *'
|
| |
|
| |
|
| |
|
| |
|
| |
|
| |
|
| |
|
| |
|
| |
|
| |
|
| |
|