adding shared install paths to QMAKE_{INCDIR,LIBDIR} in the spec has the
tiny side effect that they are searched _first_, which is generally a
really bad idea - they should be _last_.
for that purpose, make QMAKE_{INCDIR,LIBDIR}_POST live up to their names
(i.e., search them actually last) and migrate all affected specs to use
them.
Task-number: QTBUG-40825
Change-Id: Ie0de81c3cc49e193186d2fedd7d6c77590c8ef79
Reviewed-by: Joerg Bornemann <joerg.bornemann@qt.io>
Because we obviously don't support Microsoft's extensions in C++. This
is required because some MS headers have code that isn't proper C++,
like iso646.h:
#if !defined(__cplusplus) || defined(_MSC_EXTENSIONS)
#define and &&
Do not pass /Za to MSVC to generate moc_predefs.h, because this option
is incompatible with compiler options like /fp:fast that may be
user-specified.
This reverts commit e1a70ce4 and re-fixes the issue similarly to commit
d72ac3f3.
Task-number: QTBUG-58391
Change-Id: I5c0143283afed09f98200806c87259c039c00ae1
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@qt.io>
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@qt.io>
Reviewed-by: Olivier Goffart (Woboq GmbH) <ogoffart@woboq.com>
As Qt applications using OpenGL are linked against these libs, merging
them into QtANGLE by default (780105f906)
was a binary incompatible change. This change restores the default
behavior to the one before given change.
If the user wants the libraries to be merged, he can pass
combined-angle-lib to configure.
Task-number: QTBUG-60373
Change-Id: Iedbd3f2ce9284fdde924cfae8d915d6d5fef00db
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@qt.io>
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@qt.io>
Reviewed-by: Jan Arve Sæther <jan-arve.saether@qt.io>
Reviewed-by: Joerg Bornemann <joerg.bornemann@qt.io>
prefix it to the purpose field. this makes the presentation more
consistent with the graphical tool, and avoids the need to be redundant
in the description.
Change-Id: Iea58885637e0518aa1ec2f69ff05090b7e1e77a9
Reviewed-by: Stephan Binner <stephan.binner@basyskom.com>
Reviewed-by: Joerg Bornemann <joerg.bornemann@qt.io>
As the directory installation command also works with files as a source
we can unify the external commands, resulting in simpler command lines.
Change-Id: I65013626eedbdb3ce1c77ed230d46edd1603b986
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@qt.io>
VS 2017 still uses vclibs version 14/140 and it also strictly requires
the vclibs' publisher to be set in the manifest. As we only support VS
2015 and 2017 the condition for adding additional vclib dependency
information can be dropped.
Change-Id: I813a9ad976339e347fd8d9283d92282e86b2791b
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@qt.io>
Reviewed-by: Maurice Kalinowski <maurice.kalinowski@qt.io>
Use the new qmake install command to copy files around, in order to
avoid the '+' trap of the Windows copy command.
Task-number: QTBUG-60214
Change-Id: I7f588dbbfcdd89b7e98dbef7757944ca856815aa
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@qt.io>
Adds default off configure flag to use compiler optimizations
for size instead of the default speed/size trade-off.
Change-Id: I36702064ef2cc743d2d03a386adf5cefd5371b6e
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@qt.io>
the options are mutually exclusive, with the last one winning.
Task-number: QTBUG-60382
Change-Id: Ie6d888ac8be6b22b204fcad93c184477baa38965
Reviewed-by: Joerg Bornemann <joerg.bornemann@qt.io>
delaying the logging setup until the cache use is determined (which
depends on xspec determination) causes too much trouble.
as already explained in 7ac15ab0f, there is a bit of a gray area when
exactly the log should be cleared anyway.
a more complete solution would cache the tests' output along with the
results and re-log it (with an appropriate marker) upon re-use.
Task-number: QTBUG-59565
Change-Id: I17d457598d885bceafd6505cad5ff074c4ace502
Reviewed-by: Joerg Bornemann <joerg.bornemann@qt.io>
They have not been tested with Qt 5, not even once to make sure
that the mkspec passes the sanity check.
This removes for OSes:
HP-UX (running on PA-RISC -- Itanium still supported)
SGI IRIX
SCO Unix
Tru64 Unix
Unixware
And compilers:
IBM xlC (Visual Age C++)
HP aCC compiler (PA-RISC and Itanium)
PGCC, cxx and kcc on Linux
There were a couple more OSes detected in the configure script that were
lacking even the mkspec. Those have also been removed.
Of those, only hpuxi-acc and aix-xlc have been tested in the last 9
years, though only with Qt 4 and never tried with C++11 support. IRIX
was last tested over 10 years ago and PA-RISC is definitely not
supported due to its lack of atomic operations.
Support for HP-UXi and AIX is now only possible with GCC (assuming GCC
supports those). Support files for Oracle's Sun Studio compiler are left
behind, but its state is unknown.
Change-Id: I7814054a102a407d876ffffd14b69c796b97c972
Reviewed-by: Ville Voutilainen <ville.voutilainen@qt.io>
Reviewed-by: Lars Knoll <lars.knoll@qt.io>
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@qt.io>
The module might add QT_RESTRICTED_CAST_FROM_ASCII to DEFINES to
be able to use string literals under the assumption they are UTF8
without warnings from QT_ASCII_CAST_WARN, but this conflicts with
QT_NO_CAST_FROM_ASCII which is added for the headersclean build
step.
Change-Id: Ic1d7b5415350477d751a6c15219d4f8feb816a3c
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
This change restores usability of Qt with older versions of CMake which
do not provide a language feature used in the implementation of the new
Qt5::<ModuleName>Private targets.
Task-number: QTBUG-60229
Change-Id: I6a6a1b09be3edca200692250bb8f0d7f6a1b8f7a
Reviewed-by: Stephen Kelly <steveire@gmail.com>
While a previous commit added the logic for handling arm from a qmake
perspective, the responsible mkspec was forgotten.
Change-Id: I5f36d411f2d2c22d8b048e0931da50d911e7e1b4
Reviewed-by: Jake Petroules <jake.petroules@qt.io>
Creating a precompiled header exclusively for a one- or two-file module
is wasteful. The time that it takes to build the precompiled source is
on the same order as a regular compilation, so enabling precompiled
headers for those modules just makes the build slower.
Also make it possible to override the precompiled header by just setting
PRECOMPILED_HEADER to empty.
Change-Id: I0e1a09998253489388abfffd14b5f221288c4826
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@qt.io>
Let's not allow any new code that uses non-conforming syntaxes. With
GCC and like, we already use -std=c++11 instead of -std=gnu++11 for that
very reason.
Change-Id: I4a7dc1fe14154695b968fffd14aba9f8cea69c47
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@qt.io>
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@qt.io>
Reviewed-by: Oliver Wolff <oliver.wolff@qt.io>
The original commit only added support for GCC and Clang, but not ICC.
Amends 73331eeb
Change-Id: Id7638cf1b538edb1008fb3aa10754c1f517a994f
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@qt.io>
Otherwise, in separate_debug_info.prf, we try to process the original
target, which is not what we actually build.
Task-number: QTBUG-59779
Change-Id: I2b2d2e7b5f87041cc51075da7c3a5b690f94f1f3
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@qt.io>
Neither the Intel compiler nor Visual C++ have a dedicated switch to
enable F16C support, like GCC and Clang do. So we used the AVX switch
for that in commit 8241d51f70, as it was
the closest, lowest denominator. That was incorrect and insufficient.
The Intel compiler silently miscompiles the intrinsics with -xAVX,
making calls to out-of-line functions like _mm_cvtps_ph, which don't
exist. So we actually have to use AVX2 support to generate correct code.
That might be a problem later, since Ivy Bridge supports F16C but not
AVX2.
Visual C++ is able to generate F16C code with just -arch:AVX.
Either way, since there's no dedicated command-line switch, there's also
no dedicated preprocessor macro. We're using __AVX2__ for both
compilers, as that's a sufficient condition to indicate a processor that
supports F16C.
Change-Id: I27b55fdf514247549455fffd14b205b8d8b86da7
Reviewed-by: Allan Sandfeld Jensen <allan.jensen@qt.io>
The .pro file requires the QMAKE_CFLAGS_F16C to be set to something. So
set it to AVX, as the instructions require the VEX prefix anyway (ICC
has no dedicated option for just F16C).
Change-Id: I27b55fdf514247549455fffd14b171940afd35a2
Reviewed-by: Allan Sandfeld Jensen <allan.jensen@qt.io>
The configure-time detection (cxx11default) isn't enough if the compiler
can be changed. This is especially necessary if Qt is compiled with a
compiler that defaults to >= C++11 (e.g., GCC 6) and then the user
selects a compiler another compiler (e.g., Clang) via -spec option. In
that case, we'd miss adding the -std=c++11 or -std=gnu++11 option to the
command-line, causing the compilation to fail.
As a nice side-effect, even moc without moc_predefs.h will now get the
__cplusplus setting.
Task-number: QTBUG-58321
Change-Id: I74966ed02f674a7295f8fffd14a8be35da9640e1
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@qt.io>
Reviewed-by: Marc Mutz <marc.mutz@kdab.com>
The compiler is mostly GCC in disguise, but the libraries are not. Since
the toolchain is not open, it's difficult to fix issues in it.
Task-number: QTBUG-59671
Task-number: QTBUG-59672
Change-Id: Id92f4a61915b49ddaee6fffd14aea2639153f073
Reviewed-by: Jake Petroules <jake.petroules@qt.io>
the borrowing of headers always happens from "proper" modules which are
actually built as frameworks if so requested. that means that even
though the borrowing module itself never is a framework, it needs a
framework path and include paths that point into frameworks.
amends 20c7ab44.
Change-Id: Ic582060dd179cc592e9be7792ff02cebdfabd772
Reviewed-by: Ulf Hermann <ulf.hermann@qt.io>
Reviewed-by: Joerg Bornemann <joerg.bornemann@qt.io>
Add the option to omit code generated by wayland-scanner, to
avoid the case where the Qt versions replace driver-specific
functionality.
Task-number: QTBUG-58299
Change-Id: I508b18b9392dbd9e2b8233399301c06410f9f5ba
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@qt.io>
Reviewed-by: Johan Helsing <johan.helsing@qt.io>
It is ignored (and is unnecessary to begin with) in that case,
and emits an annoying warning which this patch silences.
Change-Id: I6059969724b203d6e0e2eea81ad3e3e8f8d536d6
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@qt.io>
This fixes an issue which caused certain iOS projects to fail to link
when building with Xcode, and also fixes an issue where projects were
only ever built with -fembed-bitcode instead of -fembed-bitcode-marker
for debug builds with the makefile generator.
Task-number: QTBUG-58754
Change-Id: Icf0c9f0d64dbc0b38d6c48bf635c5383a78bd6d5
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@qt.io>
Reviewed-by: Gabriel de Dietrich <gabriel.dedietrich@qt.io>
otherwise, the invalid "--qpa xcb" would complain about "xcb" being
unknown, rather than "--qpa" missing an argument.
actual booleans are handled by the type-specific callback, just like
-foo would be.
Task-number: QTBUG-59062
Change-Id: I96632dacfb721cfcbf223b76f6c5c38c810e8d0e
Reviewed-by: Joerg Bornemann <joerg.bornemann@qt.io>
Sometimes, users need to add compiler flags to QMAKE_CXXFLAGS that
aren't supposed to be passed to the predefs dump. That's especially true
for -include options, as that would change completely what's
defined. Not to mention that -include is a preprocessor option and
shouldn't be in CXXFLAGS in the first place (Automake has CPPFLAGS, but
qmake only has INCLUDEPATH and DEFINES).
[ChangeLog][qmake] Added the ability to suppress the collection of the
compiler predefined macros for moc's use. To disable the collection, use
CONFIG += no_moc_predefs.
Task-number: QTBUG-58857
Change-Id: I4139d5f93dcb4b429ae9fffd14a34d49825d9b85
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@qt.io>
these statements are assumed to provide dependencies for the currently
detected library. this implies that their resolved content must be
passed to the linker after the to be detected static library.
Change-Id: Ifaaee2ac71bf176e8a0033765fb979fe119deaba
Reviewed-by: Lars Knoll <lars.knoll@qt.io>
this went unnoticed, because the only cross-module 'use' so far is that
of egl, for which the code path provided for modular builds happens to
(mostly) work due to the specs already providing the library definition.
amends cc842ca4.
Change-Id: I58c638d896eabd26f27d5cd90e3a7f8eeece9bc0
Reviewed-by: Lars Knoll <lars.knoll@qt.io>
we derive the keys from the last fragment of the parent directory, which
is potentially not globally unique.
Change-Id: I57cf13394984e6e3d902c0f1bb495bd3920bfc75
Reviewed-by: Lars Knoll <lars.knoll@qt.io>
if a library uses another library which happens to be absent, then the
former must also fail.
amends cc842ca4.
Change-Id: I91f157a6d1ed40b66e196340a282ebe493fcf40e
Reviewed-by: Lars Knoll <lars.knoll@qt.io>
the outdated ones remain for backwards compatibility; some remain
unchanged.
Task-number: QTBUG-30083
Change-Id: Ia596b854d26b00fcb4f48df0da7ad893650ac1c8
Reviewed-by: Lars Knoll <lars.knoll@qt.io>
currently mostly for debugging purposes (especially with -verbose).
Change-Id: I8af32c61df0b19861aa79bc4bbdd3f6095dbe9b7
Reviewed-by: Lars Knoll <lars.knoll@qt.io>
also actually deletes qfeatures.txt, which was already claimed by
a668c6a6, but not actually done.
Task-number: QTBUG-58411
Change-Id: I686760632fee7c10b01bd2e83f2481b01bc2b774
Reviewed-by: Lars Knoll <lars.knoll@qt.io>
a header-only library in a default location would produce no variables
at all, making it appear undefined. fix this by forcing the writeout of
the QMAKE_LIBS_* variable, and use its definedness (rather than
non-emptiness) as a signifier.
this works for both QMAKE_USE and configure tests'/libraries' 'use'
entries.
Change-Id: Id7a1e23725caba1a91ea4db448b4aeb7fe632393
Reviewed-by: Lars Knoll <lars.knoll@qt.io>
if multiple library entries provide the same export (as openssl is going
to), it makes sense to make them recognize the same input variables, as
it would be rather counterproductive to require different configure
arguments for each.
Change-Id: Ia32842e95294296d50220297f85689bc92de2d05
Reviewed-by: Lars Knoll <lars.knoll@qt.io>
pass the library name instead of a library source object.
improves overall legibility at a marginal runtime cost.
Change-Id: I248ee9622af2b2c37daa2dbc0cc0bca5701d7925
Reviewed-by: Lars Knoll <lars.knoll@qt.io>
icc defaults to -O2, so this was causing debug code to be built like
that, making debugging very hard. This change also hardcodes -O2 for
release builds, just in case.
Change-Id: Ibc5c715fda334a75bd2efffd14a478ce539a3a3f
Reviewed-by: Robert Pancoast <RobertPancoast77@gmail.com>
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@qt.io>
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
They are the default with icc.
Change-Id: Ibc5c715fda334a75bd2efffd14a478c20b527d7c
Reviewed-by: Robert Pancoast <RobertPancoast77@gmail.com>
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@qt.io>
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Use F16C or ARM FP16 if available at compile time.
Configure check added because older clang compilers have F16C defines
and flags but not all the intrinsics.
Change-Id: I71f358b8fd003e70ab8fcf35097414591e485112
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
This is an old error of the gstrip binutils. The bug has been corrected
and re-introduced.
The command *elfdump -d xxx* on the ELF does bring lines like those :
.SUNW_syminfo: invalid sh_info: 0
Task-number: QTBUG-58814
Change-Id: I330c4031dcf4ba64297df4b333b41cf0a003914f
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@qt.io>
The value _XOPEN_SOURCE=600 should be used for C99 as we compile at
least with C++11. By doing so the compilation reaches another error in
a third library. Simply removing the option makes the compilation
working normally.
Task-number: QTBUG-56293
Change-Id: Ie040325936591958d05cc0a2d43643fa5d0c43b5
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@qt.io>
The path match OpenIndiana distribution based on Illumos.
Task-number: QTBUG-56293
Change-Id: I44e7defa63809dc4f413b46329481b53e5e74c30
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@qt.io>
Use of gcc-base.conf, g++-base.conf files and creation of solaris.conf.
The content of solaris.conf should follow the content of linux.conf.
Task-number: QTBUG-56293
Change-Id: I59cf9efa82ab0a2b22ea1a58f6339280460e5f92
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@qt.io>
there should be only one, at the level of the isolating .qmake.conf.
Change-Id: I25f05864d6f5c1bb5caf2fb4138adb4bb9cc2f22
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@qt.io>
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@qt.io>
don't fail to set up cross_compile early enough. otherwise, we'd
populate the cache for target builds with data for the host.
amends 6b8666c7 and 5060740f.
conversely, pass on extra flags to configure tests when not cross
building.
amends d8be8110 (and 2c5eb3e6).
Task-number: QTBUG-58556
Change-Id: I531d71e06204a0b17ae6dabf017a52e0f2efd9a7
Reviewed-by: Joerg Bornemann <joerg.bornemann@qt.io>
these are not referenced anywhere, and neither should they, as they are
not relevant to _using_ these modules in any way.
notably, QT.*.bins remains exported - this is because dlls reside there,
so it is necessary for setting up a launch environment.
Change-Id: I7a33c72be6e4789ea29a2fbbcac9588213900b6e
Reviewed-by: Joerg Bornemann <joerg.bornemann@qt.io>
the only users of module versions in the first place are found within
qt's own prfs; even qbs' qt module importer ignores them. but arguably,
the information makes sense.
however, exporting the same barely useful information redundantly is
plain over the top, so remove the pre-split representation.
Change-Id: Iaee69c86d8b7c8b8ef4f3580b8da333aeb8ade2c
Reviewed-by: Joerg Bornemann <joerg.bornemann@qt.io>
in a framework build, the headers are inside the "library", so it's
obviously not very wise to suppress its installation on the basis of
it not being there.
Task-number: QTBUG-57656
Change-Id: I026a3e486a2aad6ee0b8e0d264af4385af945e42
Reviewed-by: Jake Petroules <jake.petroules@qt.io>
Reviewed-by: Harald Fernengel <harryf@gmx.com>
Building qt with msvc would fail with the -developer-build configure
option turned on under simplified chinese locale.
This is because msvc will emit warnings for source files with utf-8
characters which are not representable in CP936, and -developer-build
implies treating warnings as errors.
This patch turn on -utf-8 compiler option for msvc2015 update 2
and up only for building qt.
Task-number: QTBUG-58161
Change-Id: If38ea11eb1f39f8e08efa1cccb92e0eea50daf92
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@qt.io>
Match the logic that is used for building.
Change-Id: I9ce6db2357b9b3e19a1e932fae57d376aae406f7
Reviewed-by: Eskil Abrahamsen Blomfeldt <eskil.abrahamsen-blomfeldt@qt.io>
Nothing uses this and there's no incremental_off.prf since the MSVC6 purge
in 99f390f9 (dec 2009).
Change-Id: I2bc52f3c7a574209b213fffd149b4bf378805969
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@qt.io>
There's no reason to have that. Let the compiler decide whether it wants
to align or not. Current versions seem not to want to align beyond
multiple of 2.
Change-Id: Iab7c358838e1487387a2fffd149d73e0a218d3c2
Reviewed-by: Dominik Haumann <dhaumann@kde.org>
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@qt.io>
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
The uvfd flag is implicitly defined when -Olink is used.
This causes the compiler to generate a warning for every
file being compiled in release mode.
Change-Id: I75759151864da7cf2f6d9c812e466a52c1208444
Reviewed-by: Nikola Velinov <nvelinov@ghs.com>
Reviewed-by: Lars Knoll <lars.knoll@qt.io>
This fixes a regression introduced in the merge 318b5856.
Due to the removal of actual simulator_and_device in 5.8 (397f345a6),
conditions using it have become meaningless.
Task-number: QTBUG-58440
Change-Id: I9f874f9f85efa590c40602dbcd07793ff17d35f5
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@qt.io>
Reviewed-by: Liang Qi <liang.qi@qt.io>
This ensures at compile-time that Qt libraries do not use any APIs that
are not safe for use in application extensions, and fixes warning
messages that appear when linking to Qt libraries that are not built
with this flag, when used in an application extension.
This is especially important on watchOS where *all* "applications" are
actually application extensions, and on other Apple platforms if
application extensions are developed using Qt.
Task-number: QTBUG-40101
Change-Id: I022046f2584e0222253d33052b0abc221d7c93d6
Reviewed-by: Tor Arne Vestbø <tor.arne.vestbo@qt.io>
The patch fixes a number of bugs in code, and removes dead logic
clarifying that MIPS DSP, like ARM NEON, has no runtime detecton.
Change-Id: If2f4eea68da5b2eaa80b8e9c8258206d8c1b7173
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@qt.io>
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
The compiler command line used to populate QMAKE_DEFAULT_INCDIRS must
include the sysroot in order to generate the correct paths list.
This fixes a regression introduced in afd8263 which in turn attempted to
fix an earlier regression making it impossible to override the
deployment target in user project files.
Task-number: QTBUG-58325
Change-Id: I93e6b7ef90b2744dd2f03c77da31c692cb194976
Reviewed-by: Gabriel de Dietrich <gabriel.dedietrich@qt.io>
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@qt.io>
Enables optimizing with -Og if GCC has the option available, this
should produce faster debug binaries without compromising debugability.
Is a privateConfig to limit it to the default Qt build.
Includes two fixes for false positives of maybe_uninitialized triggered
by -Og on gcc 4.9.
Change-Id: I466d7a4070295714189024369312e6cbd36cfacf
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@qt.io>
It only applied to Windows (not MSVC, like the help said) and the build
was broken with this option. So remove it, as we clearly never test
this.
[ChangeLog][Windows] The -no-rtti configure option was removed, as Qt
5.8 fails to build under that condition. To disable RTTI on user code,
add to your .pro file: CONFIG += rtti_off.
Change-Id: I2bc52f3c7a574209b213fffd149aae1b8d0cf9df
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@qt.io>
The problem isn't the compiler, but the linker, so we ought to turn this
flag on for all ELF-based systems where we compile Qt with -Bsymbolic,
but so far only the Intel compiler has been affected. We can turn it on
as needed for other systems.
The cause of the problem is the way that ICC materializes the pointer-
to-member-function: it stores the full 2*sizeof(void*) data in an
anonymous variable, so that it can load it in one go, such as one 16-
byte SSE aligned load on 64-bit systems. That relocation in a data
variable gets turned into a fixed-position copy relocation by the
linker, which breaks the signal-identification mechanism.
GCC and Clang are likely to be affected if anyone did:
static const auto destroyed = &QObject::destroyed;
QObject::connect(obj, destroyed, [](){});
Task-number: QTBUG-52439
Change-Id: I0d69eaf61af149db9574fffd149d0cfb08459c33
Reviewed-by: Olivier Goffart (Woboq GmbH) <ogoffart@woboq.com>
This was used solely for building freetype. It was made obsolete as of
2eaf0cf8fd which upgraded freetype to a
newer version which does not rely on that define, first included in 5.5.
Change-Id: Iaaea8d6783032d784f0a370f8404972967fa7a06
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@qt.io>
This implies that using tools like windeployqt and macdeployqt will not
generate a working standalone application.
Change-Id: I002cf6e527e479ccbee2f18df8766648196d6232
Reviewed-by: Andy Shaw <andy.shaw@qt.io>
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@qt.io>
Without this, building a project with qmake -Wall will
always produce the following warning:
mkspecs/features/resources.prf:22: Unescaped backslashes are deprecated
Change-Id: I0aeedbf470958ab458651a263e3f804ea2d1a0f0
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@qt.io>
don't ignore detected features for host tools when we're not actually
cross-building.
Change-Id: Id62a3c1c6b7ae422b14efb4fbea0892b05a047cc
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Reviewed-by: René J.V. Bertin <rjvbertin@gmail.com>
This was missed in dd9ec1564, leading to errors for example when building
with separate_debug_info.
Change-Id: Ibeb8020abe32690bcc691c1ca139508775c91db2
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@qt.io>
Renaming these specs should cause less trouble than usual since
most people are likely using the rasp-pi2 spec even for RPi3.
[ChangeLog][Platform Specific Changes][Linux] The Raspberry Pi 3
device specs have been renamed from linux-rpi3-* to linux-rasp-pi3-*
in order to follow the pattern of the RPi 1 and 2 specs.
Change-Id: I7d1ed6a4665078af9a4cbbe4c38b5fac760d03f9
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@qt.io>
