Since we use the Clang from Xcode's toolchain now, the OS version is
not relevant.
In practice this means we will use clang for Xcode 4.2 and up, which
means it's possible to use clang also on Mac OS 10.6 (Snow Leopard),
where Xcode 4.3 is not available.
Change-Id: I9817e237cdd82d10b93aaaa3c90e35767cdca751
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@digia.com>
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
For Mac OS X we currently specify build tools without an absolute path,
which means we end up using the ones in /usr/bin. This is wrong, we
should be using the tools from the toolchain of the chosen SDK.
For iOS we do specify an absolute path, by resolving the toolchain
path in the iOS makespecs.
To solve the situation on Mac OS X, we move the logic of resolving the
toolchain path to sdk.prf, and share it between OSX and iOS.
For configure we need to duplicate some of the logic from sdk.prf, as
configure pulls out QMAKE_CC and QMAKE_CXX for running some initial
tests and building qmake. The new macSDKify function also solves
the issue of missing sysroot and deployment version in the flags.
Change-Id: Ib1d239c9904cf3ccee5214b313cf6205869a1462
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@digia.com>
Is needed now that we write QMAKE_MAC_SDK with a !host_build scope
in qdevice.pri.
Change-Id: I298cc660b496460190337c175aef684a5522d5cb
Reviewed-by: Tor Arne Vestbø <tor.arne.vestbo@digia.com>
qdevice.pri is not target/host specific yet, so we were loading
it and overriding QMAKE_MAC_SDK to iphonesimulator eg., even
for host tools such as moc.
Change-Id: I10277e60e1da84dda239e32a6f19b40dc48f084a
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@digia.com>
Reviewed-by: Morten Johan Sørvig <morten.sorvig@digia.com>
Version 3.1, shipped with Xcode 4.3, has the same issues compiling Qt's
AVX code.
Change-Id: Icb778fbd9d61f01aa84365661af050c9442d4d7f
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Based on the Necessitas project by Bogdan Vatra.
Contributors to the Qt5 project:
BogDan Vatra <bogdan@kde.org>
Eskil Abrahamsen Blomfeldt <eskil.abrahamsen-blomfeldt@digia.com>
hjk <hjk121@nokiamail.com>
Oswald Buddenhagen <oswald.buddenhagen@digia.com>
Paul Olav Tvete <paul.tvete@digia.com>
Robin Burchell <robin+qt@viroteck.net>
Samuel Rødal <samuel.rodal@digia.com>
Yoann Lopes <yoann.lopes@digia.com>
The full history of the Qt5 port can be found in refs/old-heads/android,
SHA-1 249ca9ca2c7d876b91b31df9434dde47f9065d0d
Change-Id: Iff1a7b2dbb707c986f2639e65e39ed8f22430120
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@digia.com>
Reviewed-by: Lars Knoll <lars.knoll@digia.com>
If the command is quoted, it can contain anything but quotes (we do not
support escaped quotes, so single quotes have to be used). If the
command is unquoted we look for the first closing parenthesis. We used
to do this using .*?, but the greedy modifier '?' didn't seem to work,
so we now use an inverse character set.
Change-Id: I40660ce7aef6a6b6d480292d28da1b079bb161da
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@digia.com>
Instead of setting it explicitly to 10.6. Before 736e4258a this was
not a problem, as we were hard-coding the version flags as part of
the C/CXX_FLAGS for each makespec, so getQMakeConf would pick
them up automatically. When 736e4258a introduced sdk.prf as the
place to resolve these flags in a single place, it broke the
qmake build for macx-clang-libc++, as we were then falling
back to the 10.6 target hard-coded in configure.
We fix this by duplicating some of the logic from sdk.prf, by
pulling out the QMAKE_MACOSX_DEPLOYMENT_TARGET variable
and adding to the C/CXX_FLAGS.
Change-Id: I04afc7525031727c2504588c70dd3f7892cc8e42
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@digia.com>
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
This is enabled only for -developer-builds and only for certain
compiler-version combinations that are in a whitelist.
It also requires each library, plugin or tool to declare whether it is
supposedly clean of warnings. When most targets are clean, we can
consider inverting.
Change-Id: I17b5c4e45aee5078f9788e846a45d619c144095a
Reviewed-by: Lars Knoll <lars.knoll@digia.com>
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@digia.com>
Should be applied to Qt modules. Not interesting for third party
users.
Change-Id: I8fce821af397e3ace011a426c762319f6d30004f
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@digia.com>
Reviewed-by: Eskil Abrahamsen Blomfeldt <eskil.abrahamsen-blomfeldt@digia.com>
The dbus_watch_get_fd function was deprecated in D-Bus 1.2 (technically,
in 1.1.1, but that was a development release) because it had a bad name.
Sockets on Windows have file descriptors, but they are not shared from
the same pool as the CRT library's file descriptors.
This commit raises the minimum required version of D-Bus to 1.2. This is
the first requirement raise since this code was introduced in 2006. For
some reason, the D-Bus 1.2.0 release seems to be missing, but 1.2.1 was
released on 04-Apr-2008. That's ancient enough for all distributions
Qt 5 is supposed to run on.
Change-Id: Ia6bbc137fffbb27c77290ed3e32d3380f0ae3c54
Reviewed-by: Lorn Potter <lorn.potter@jollamobile.com>
The LLVM version in Xcode 4.5 and below does not handle the GNU
assembler syntax in the pixman assembly file.
A possible alternative workaround is to include a preprocessed assembly
file using https://github.com/hollylee/gas-preprocessor.
Change-Id: Id95add669c60d3a7da823e5975afdd1f88f71977
Reviewed-by: Morten Johan Sørvig <morten.sorvig@digia.com>
Clarify in the configure --help documentation that the text between
parentheses in the -make option indicates the default parts, not all
the parts that can be chosen from.
Task-number: QTBUG-28826
Change-Id: Iac9cf294b8054823ecfaf262aeafab7779ce4c8b
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@digia.com>
Must use ; as the path delimiter instead of :
Change-Id: I549e1652ef5bbae09c8fddec3e83ac9f52cec3a4
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@digia.com>
this makes it possible to exclude modules from the build without moving
their sources out of the way. substitutes the much-requested -no-webkit.
not adding a symmetrical option, as it is relatively pointless:
to build only specific "leaf" modules, you only need to run
"make module-qt<module> ..." once you configured. and removing
particular "intermediate" modules is achieved with this very option.
Task-number: QTBUG-26697
Change-Id: I25cebdbd029885a2c653c4cde696f9bb78691768
Reviewed-by: Tuukka Turunen <tuukka.turunen@digia.com>
Reviewed-by: Joerg Bornemann <joerg.bornemann@digia.com>
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>
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>
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>
... instead of as a fallback in default_post.
it was this way in qt4, and it requires less code to be written in the
end. we are already doing it for debug/release as well.
Change-Id: I6e02849d61d14a18375cf64a5990768931ebac48
Reviewed-by: Joerg Bornemann <joerg.bornemann@digia.com>
The chosen makespec may spit out errors for various reasons, which will
influence steps such as detecting pkg-config or the architecture test.
It's preferable to catch any issues with the makespec as early as
possible and exit configure before the errors propagate to tests.
Change-Id: Iecbf3217c36dea9f5e0677c58171b72cb6ce1e0b
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@digia.com>
Reviewed-by: Tor Arne Vestbø <tor.arne.vestbo@digia.com>
Passing -sdk to configure should affect the target build, not host tools.
We skip passing MAC_SDK_FLAG for the host arch test as well, since as for
qmake the SDK should only apply to the target.
Change-Id: I3902355715234b9300a65d3095b9c925d9492311
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@digia.com>
Reviewed-by: Tor Arne Vestbø <tor.arne.vestbo@digia.com>
The exact same hunk is present 100 lines below.
Change-Id: Ia3d61037b29186368e30f95f4162282d38bca972
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@digia.com>
This allows for example the iOS mkspecs to use xcode-select to get the
path to Xcode's /Developer directory, which is then in turn used to
set up QMAKE_CXX (which is needed by configure).
Coprocess is GNU awk specific, so we're just using getline.
http://www.gnu.org/software/gawk/manual/html_node/Getline_002fPipe.html#Getline_002fPipe
Change-Id: I7e88112bea68f039361d6e96ee581eedf129ab02
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@digia.com>
Change the default libexec directory to 'lib' on Windows for default
configurations. This makes sure that e.g. qtwebprocess can actually
find & load the right Qt libs.
A separate libexec directory was introduced to avoid conflicts between
a system-wide Qt in PATH and a custom Qt installation. However, since
there are no system wide Qt installations on Windows this isn't an
issue, but getting the executables to find its Qt libraries is.
Alternatively we could have also placed it in the 'bin' directory.
However, putting it in the 'lib' folder carries the notion that,
unlike the binaries in the bin, the libexec executables might have
to be deployed with a target application.
The exception is when a separate archdata prefix is specified, and
$$archdata/lib won't contain Qt libraries. In this case we use libexec
like on the other platforms too.
ChangeLog: [Qt for Windows] Fixed launching of QtWebProcess.exe by
changing default libexec directory to 'lib'.
Task-number: QTBUG-29661
Change-Id: Icbb15fb98474d6fef8ac9310f2e2b482d3282f79
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@digia.com>
Reviewed-by: Jocelyn Turcotte <jocelyn.turcotte@digia.com>
Reviewed-by: Joerg Bornemann <joerg.bornemann@digia.com>
On Windows, it's completely different and we don't currently support that
compiler anyway.
Change-Id: Ie3365ea103c93c63e79ebb1d4908c361173ac449
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@digia.com>
The current code matches *g++*, which matches "clang++" and detects
Clang as GCC. That's highly incorrect.
Change-Id: Ifd85bbd35aa130be3094fc75d471614d06ca23bd
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@digia.com>
Reviewed-by: Morten Johan Sørvig <morten.sorvig@digia.com>
Keeping accessibility and only disabling the bridge will
enable more builds to work.
Warning about disabling accessibility disabled is needed
because in QStyle it is used to discover semantics about widgets
(if a toolbutton is in a toolbar).
Change-Id: Iae4e6ab63479743bdd70cba4b1954ec7cf3f88e9
Reviewed-by: Jan Arve Sæther <jan-arve.saether@digia.com>
eventfd(7) uses less resources than a pipe, as it only needs to store a
single 64-bit integer, as opposed to a full buffer.
It was introduced first on Linux version 2.6.22 and glibc 2.7. However,
both the configure-time test and the runtime usage require the use of
EFD_CLOEXEC for thread-safety, so this code will be enabled only for
Linux 2.6.27 and up as well as glibc 2.9 and up.
Change-Id: Ic7e10b28d7b1d4ca24be614ed84055c4429a68e4
Reviewed-by: Robin Burchell <robin+qt@viroteck.net>
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@digia.com>
On Mac, GCC and Clang print "(framework directory)" at the end of the
directory listing for includes if that's a framework dir. The directory
with "(framework directory)" at the end does not exist, so it doesn't
affect anything by being there.
But we don't have to keep the list longer just for that.
Change-Id: I3d031d3d15c75801ec0d6112b2c913bd63e5def3
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@digia.com>
Make the settings path to be relative to $prefix, instead of absolute.
And update the configure -help output to match the actual code.
Change-Id: I71e4ad6e3db046fec95ef057ae7f7bc566bc5794
Reviewed-by: Sune Vuorela <sune@vuorela.dk>
Reviewed-by: Lars Knoll <lars.knoll@digia.com>
