summaryrefslogtreecommitdiffstats
path: root/mkspecs
Commit message (Collapse)AuthorAgeFilesLines
* Add Mesa VC4 based RPi3 device specLaszlo Agocs2016-06-142-0/+78
| | | | | | | | The rpi3 spec added previously is only half of the story now that we have the option to use another GL driver. Change-Id: I1b1edde77bcc6d2f382f1021de9c594c27c34d6f Reviewed-by: Andy Nichols <andy.nichols@qt.io>
* BSD mkspec cleanup for FreeBSD, NetBSD and OpenBSD to common namespaceRalf Nolden2016-06-1110-209/+136
| | | | | | | | | | | | | BSD OS mkspecs share mainly the same configurations except very few differences. Merge into a common/bsd/bsd.conf file to be used across all BSD OSes in their respective qmake.conf and add a qplatformdefs.h that contain the common defines to be re-used in the BSD mkspecs. The change includes the usage of <sys/param.h> also on NetBSD through the common qplatformdefs.h, which is intended for using NetBSD's version defines. Change-Id: Ibb0ac9e4c8bb5aff7d0febdcab1a4b9600a61117 Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
* fix ios device&simulator builds againOswald Buddenhagen2016-06-106-8/+16
| | | | | | | | | | | | | the assumption stated in b67a0836d is actually invalid - configure sets build_all without debug_and_release there. debug_and_release does actually imply build_all, though. to make things less confusing, don't let configure inject iphonesimulator_and_iphoneos into all projects, but handle it like debug_and_release instead. Change-Id: Ib7acdc63308a538862fc603428f81aba60bca08e Reviewed-by: Jake Petroules <jake.petroules@qt.io>
* Remove a.out defines in OpenBSD and NetBSD mkspecRalf Nolden2016-06-082-18/+0
| | | | | | | | | | | | | | | | | | | | OpenBSD made the switch from a.out to ELF with release 3.4, published 2003, see http://www.openbsd.org/34.html. In preparation to cleaning up the mkspecs for OpenBSD, remove the a.out defines as only recent versions of OpenBSD (5.8 onward) will be supported. NetBSD switched from a.out to ELF file format with release 1.5, see https://www.netbsd.org/releases/formal-1.5/NetBSD-1.5.html in the year 2000. Remove the defines for older a.out systems now as we are at release 7.0 and only 6.0 onwards releases are currently supported anyway. While cleaning up, remove the old comments for QT_SOCKLEN_T values of outdated OpenBSD and NetBSD releases as well. Change-Id: I8519eab7bcd4af19c0e9f628657878f32b2bd602 Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
* Remove OpenBSD _POSIX_THREAD_SAFE_FUNCTIONS undefineRalf Nolden2016-06-081-15/+0
| | | | | | | | | | | | | | | The qplatformdefs.h for OpenBSD included an undefine of _POSIX_THREAD_SAFE_FUNCTIONS because of (at the time) missing implementations of the required _r() functions. After checking these functions http://www.unix.org/whitepapers/reentrant.html against the OpenBSD man pages http://man.openbsd.org/OpenBSD-4.4/cat3/getpwuid_r.0 they are all complete with OpenBSD 4.4 released in 2008. As OpenBSD only supports the current and the last release before (5.8 and 5.9 now), it is safe to assume that this undefine can go away now for sure. Change-Id: I341bcae77d1bd7249ac3fdeaefce9c5eb595eca7 Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
* Delete unsupported/freebsd-g++-46 mkspecRalf Nolden2016-06-082-78/+0
| | | | | | | | | | | The freebsd-g++46 mkspec was intended to specifically use gcc 4.6 from FreeBSD ports collection. However, the ports collection moves its "standard" gcc version and uses symlinks to the standard commands (e.g. gcc -> gcc48) on installing gcc. The current gcc in ports is 4.8.5, so the mkspec is not useful at all. Change-Id: I041325d05c7dc3f47f4a774d6f46ba24a601bf3e Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
* FreeBSD: remove -D_THREAD_SAFE from pthread usageRalf Nolden2016-06-082-2/+2
| | | | | | | | | | | | On FreeBSD, the gcc manpages until Release 5.1 mentioned using the -D_THREAD_SAFE define when using -pthread. This has been obsoleted and the gcc manpages from Release 5.2 onwards have this removed. Now we finally remove this historic relic here, too. Change-Id: I00a5b688c56f46b938c0806fb44b72d5afe5079e Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@theqtcompany.com> Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
* OpenBSD: simplify mkspec and match portsRalf Nolden2016-06-081-42/+5
| | | | | | | | | | | | | | | | By using commmon/gcc-base-unix.conf and common/g++-unix.conf most of the contents of the qmake.conf for OpenBSD can go. The QMAKE_LFLAGS_UNDEF are reset because the linker can't handle environ in libraries, which is only added to libc in OpenBSD 6 (http://www.openbsd.org/faq/current.html) even when explicitly adding -lc to QMAKE_LIBS. See http://lists.llvm.org/pipermail/llvm-commits/Week-of-Mon-20151102/309783.html for similar questions on using -Wl,-z,defs instead. Change-Id: I6c725fb3bc7ae63270912b07f230a480c9157dfd Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@theqtcompany.com> Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
* NetBSD: simplify mkspec and match supported OS versions (6.x-7.x)Ralf Nolden2016-06-081-46/+9
| | | | | | | | | | | | | | | | | | | | | | By using common/gcc-base-unix.conf and common/g++-unix.conf most of the mkspec details can go, leaving us with a minimized qmake.conf. Change the default X11 location to point to /usr/X11R7 after NetBSD switched to X.org in Release 5.0 by default. Also change the default location for addon software on NetBSD, it uses pkgsrc as its ports system which installs to /usr/pkg by default (the default value for $LOCALBASE of pkgsrc on NetBSD), see https://www.netbsd.org/docs/pkgsrc/using.html chapter 4.1. These default values mirror the current state of NetBSD usage to get the prerequisites to compile Qt with a minimum amount of configure parameters. Tested with NetBSD 7.0 using gcc 4.8.4-nb2 (system compiler) Change-Id: Ic50757af9f070c0383c7356302e7e7686f031740 Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
* Initialize the value before using it.BogDan Vatra2016-05-271-7/+7
| | | | | | | LIBGCC_PATH was used before we initialized it, somehow it magically worked ... Change-Id: I0a9a748ffbfc641f60736b0cf85d23f499b64e66 Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@theqtcompany.com>
* Android: Removed super old NDK checksBogDan Vatra2016-05-271-8/+0
| | | | | | Change-Id: I8c16dca1272f0413ac486fd1d54b6c678a8c9dd8 Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@theqtcompany.com> Reviewed-by: Eskil Abrahamsen Blomfeldt <eskil.abrahamsen-blomfeldt@qt.io>
* Compile fix: add -lexecinfo on NetBSD and OpenBSDRalf Nolden2016-05-232-0/+2
| | | | | | | | | | Add -lexecinfo for backtrace(3) on NetBSD and OpenBSD. Without, qlogging.cpp linking will result in undefined references to backtrace and backtrace_symbols. The behavior required is identical to FreeBSD so no additional fixes are required (see src/corelib/global/global.pri:41) Change-Id: I3cfd1d75f1fb5b8505c08a880f91e7b39a5a650d Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
* Move freebsd-g++ mkspecs back for supported FreeBSD 9.3 and upRalf Nolden2016-05-232-4/+4
| | | | | | | | | | | | | | | | FreeBSD 9.3 is still supported and uses gcc as the default compiler, therefore FreeBSD ports require patching the mkspecs back. To avoid patching, move the mkspecs back to the right place and adapt the path in the qmake.conf/qplatformdefs.h [ChangeLog][FreeBSD] The freebsd-g++ mkspec was moved back and no longer requires the "unsupported/" prefix, matching the FreeBSD ports tree, as FreeBSD 9.3 still defaults to using GCC. Users of GCC that did not previously use the ports patch will need to adapt their build scripts and drop the "unsupported/" prefix. Change-Id: Ideda4a33cccf5381000f6f50b6ae92a5c24ba9d4 Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
* Enable -separate-debug-info for mingw toolchainsUlf Hermann2016-05-231-0/+18
| | | | | | | | | | Having the debug info in separate files is very helpful, especially on memory constrained 32bit systems. When the debug info is compiled into the object files, processes can run out of memory when trying to load them. Change-Id: I1808a32e855d6a62e17e1b734c511b31db5f03f1 Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@theqtcompany.com>
* Remove obsolete freebsd-icc mkspecsRalf Nolden2016-05-202-129/+0
| | | | | | | | | | | The freebsd-icc mkspec was intended to be used with the Intel compiler, which nowadays is still only available in version 8.1.038 (https://www.freshports.org/lang/icc/) and for i386 only while the current version is around 17. On FreeBSD the port was actually never supported by Intel anyway. Change-Id: I03b1939be63eef4363bd6b074a6b739365f495ac Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
* qpa: Use LIBS_PRIVATE, not LIBS when linking in font database bits.Robin Burchell2016-05-192-2/+2
| | | | | | | | | LIBS_PRIVATE is more correct (these shouldn't be part of the public API), and additionally, using LIBS_PRIVATE also ensures these libraries come last on the link command, which fixes builds using -Wl,--as-needed. Change-Id: I9f2e4e78855a30a53ce9013623597b840a2832bd Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@theqtcompany.com>
* Merge remote-tracking branch 'origin/5.6.1' into 5.6Liang Qi2016-05-194-1/+21
|\ | | | | | | | | | | | | Conflicts: src/network/socket/qnativesocketengine_winrt.cpp Change-Id: I8edb72f8ba958d80c3d7993b3feaaae782ca8d9c
| * enable example installs for all modulesOswald Buddenhagen2016-05-181-0/+1
| | | | | | | | | | | | | | | | | | so far, each module had to do it in its .qmake.conf. by now, all modules have been adjusted to the expected structure, so we can enable it centrally. Change-Id: I16b4e7581e51bfc54e71d2f7f852858ae0b47281 Reviewed-by: Joerg Bornemann <joerg.bornemann@qt.io>
| * complain about examples which don't install themselves properlyOswald Buddenhagen2016-05-181-0/+14
| | | | | | | | | | Change-Id: I6555d449430111639b084ddc3e4b4bc14b61bc30 Reviewed-by: Joerg Bornemann <joerg.bornemann@qt.io>
| * Restore the missing -P option for lex handlingAndy Shaw2016-05-101-1/+5
| | | | | | | | | | | | | | | | | | We don't know why it works, but we've been told that it fixes some builds and we have no reason to think it will break others. Task-number: QTBUG-52998 Change-Id: Iacaa1fb3c49e4d7aafd4167e591ff3724489872e Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@theqtcompany.com>
| * make zlib_dependency auto-add QtCore as a private depOswald Buddenhagen2016-05-091-0/+1
| | | | | | | | | | | | | | | | our zlib header includes qglobal.h, so we need the qtcore include dirs, and qtcore is also where the actual code is compiled into. Change-Id: I09f530a1b4e6160438215a6d7223c0771ce94f05 Reviewed-by: Simon Hausmann <simon.hausmann@theqtcompany.com>
* | qmake: Introduce CE_PLATFORMNAME to fix creation of vcxproj filesOliver Wolff2016-05-181-0/+1
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | While previous SDKs used the form "CE_SDK (CE_ARCH)" in their configuration/platform names, this is not true fo the Toradex SDK. Inside Visual Studio the platform is only called "Toradex_CE800" instead of "Toradex_CE800 (ARMV7)". In order not to break other SDKs CE_PLATFORMNAME is introduced and used in the wince80colibri-armv7-msvc2012 mkspec. If the variable is set qmake uses it as the platform name in its vcproj generator. Change-Id: Icb501bf6446a9f617745a0d2c7a80b240680b043 Reviewed-by: Joerg Bornemann <joerg.bornemann@qt.io> Reviewed-by: Andreas Holzammer <andreas.holzammer@kdab.com>
* | Add notes and an example configure line to the imx6 makespecLaszlo Agocs2016-05-171-0/+17
| | | | | | | | | | | | | | | | Other popular specs have this as well. Change-Id: I44245f37857d476b9ee53ecad021261b94214b2f Reviewed-by: Johan Helsing <johan.helsing@qt.io> Reviewed-by: Louai Al-Khanji <louai.al-khanji@qt.io>
* | QT_CONFIG simplification re debug_and_release and build_allOswald Buddenhagen2016-05-175-11/+7
| | | | | | | | | | | | | | | | don't pretend that these two flags can be set separately - the configures set them in tandem. Change-Id: Ib0beae0152de09026d4627fd3ae0feabd9ce1b81 Reviewed-by: Joerg Bornemann <joerg.bornemann@qt.io>
* | Fix running of configure tests when cross compiling on WindowsSimon Hausmann2016-05-171-3/+1
| | | | | | | | | | | | | | | | | | When compiling on Windows to for example Android, we may be using the unix generator, but the make command may be mingw32-make. Task-number: QTBUG-53038 Change-Id: If8d1e1bea733cf1b4ed837c423a34ecfb3cafd1f Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@theqtcompany.com>
* | Add -Wzero-as-null-pointer-constant to headerclean checkMarc Mutz2016-05-131-0/+10
|/ | | | | | | | | Requested by user(s). Change-Id: Id2c7d67a8cd50f4f83f141cff41ba01e21c3a603 Task-number: QTBUG-45291 Reviewed-by: Lars Knoll <lars.knoll@theqtcompany.com> Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
* automate the installation of some more example source filesOswald Buddenhagen2016-05-031-0/+5
| | | | | Change-Id: I61f9f4882c5a7743817059824250958f5ce4a603 Reviewed-by: Joerg Bornemann <joerg.bornemann@theqtcompany.com>
* support directories in COPIESOswald Buddenhagen2016-05-021-4/+20
| | | | | | | | | | this makes it fully consistent with INSTALLS. note that this also removes the file name from the target path when copying files, also for consistency. Change-Id: I69042c9aa1e2cc81f8ff982343ba25688a04abfd Reviewed-by: Joerg Bornemann <joerg.bornemann@theqtcompany.com>
* add some missing filename expandos to extra compilersOswald Buddenhagen2016-05-021-1/+1
| | | | | | | | | | | - QMAKE_FILE_IN_EXT as an alias for QMAKE_FILE_EXT, for consistency with QMAKE_FILE_IN_BASE - QMAKE_FILE_IN_NAME to make pairing _EXT/_BASE to get a full name unnecessary (finally ...), and make use of it - QMAKE_FILE_OUT_PATH, because i'll need it Change-Id: I3d91ddb84f9cce52a665d562da11d165c92550c8 Reviewed-by: Joerg Bornemann <joerg.bornemann@theqtcompany.com>
* remove redundant OTHER_FILES assignmentsOswald Buddenhagen2016-05-021-3/+0
| | | | | | | | newer versions of qt creator understand QMAKE_EXTRA_COMPILERS and INSTALLS, so there is no need to list the files twice. Change-Id: Iccf3cc3248daf3422b8c366c2eb2d2f46c5f08d1 Reviewed-by: Joerg Bornemann <joerg.bornemann@theqtcompany.com>
* don't apply warning options to host buildsOswald Buddenhagen2016-04-291-0/+4
| | | | | | | | | | the host compiler's version is not determined, so we cannot make any assumptions about its capabilities. Task-number: QTBUG-53017 Change-Id: I939fd7402b5daac73f2195b9d8763b9008bc7ea4 Reviewed-by: Thiago Macieira <thiago.macieira@intel.com> Reviewed-by: Rolf Eike Beer <eb@emlix.com>
* make use of COPIESOswald Buddenhagen2016-04-271-17/+2
| | | | | | Change-Id: I34965cada901fce3cf300a99fa65d4d3f9bd4578 Reviewed-by: Liang Qi <liang.qi@qt.io> Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@theqtcompany.com>
* add support for copying arbitrary filesOswald Buddenhagen2016-04-272-1/+42
| | | | | | | | | | | | | | | | | this adds file_copies.prf, which understands the variable COPIES, which works analogously to INSTALLS. i've been holding this off for a long time, as it is not without caveats. however, similar hacks are proliferating all over the code base, so it's time we formalized it. in fact, it's the easiest way to fix some nasty shadow build problems, which is why i'm adding this on the stable branch. Task-number: QTBUG-52256 Change-Id: Icbe3b9fbb79c952546aad2d467a438d3a69d749f Reviewed-by: Liang Qi <liang.qi@qt.io> Reviewed-by: Joerg Bornemann <joerg.bornemann@theqtcompany.com>
* Fix separate_debug_info with no_plugin_name_prefixJake Petroules2016-04-141-1/+1
| | | | | | Task-number: QTBUG-51782 Change-Id: I5499139c732f2ff474095e30bc1f7a22743f4b12 Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@theqtcompany.com>
* qmake: Add support for separate debug info on Apple platforms.Jake Petroules2016-04-1411-16/+229
| | | | | | | | | | This makes the -separate-debug-info configure optional functional, which generates dSYM debug info bundles for Qt libraries on Apple platforms. Task-number: QTBUG-37952 Done-with: Oswald Buddenhagen <oswald.buddenhagen@theqtcompany.com> Change-Id: Ia247674740bf450130a15db926df07fa9007e2ca Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@theqtcompany.com>
* Clang 3.8 is clean of warningsThiago Macieira2016-04-131-2/+2
| | | | | Change-Id: Ibc83b9f7e3bc4962ae35ffff1425c8e686579e15 Reviewed-by: Jake Petroules <jake.petroules@theqtcompany.com>
* remove empty dylib.prfOswald Buddenhagen2016-04-121-1/+0
| | | | | | | it doesn't appear to be referenced in any way, either. Change-Id: Ifd30b435e3e628cd5e48ae24e9aef01c662d6d61 Reviewed-by: Joerg Bornemann <joerg.bornemann@theqtcompany.com>
* Fix bugs in accuracy of target path calculation in resolve_target.Jake Petroules2016-04-111-2/+21
| | | | | | | | | | | | Incidentally, this introduces QMAKE_RESOLVED_BUNDLE, which can be used to determine the path of the bundle wrapper itself as well as the executable target. This is necessary for a subsequent patch adding support for -separate-debug-info on Apple platforms. Change-Id: Ia11430026b8e3f171e5db6677b190b8356832805 Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@theqtcompany.com>
* Introduce RPI3 aarch32 mkspecDonald Carr2016-04-082-0/+75
| | | | | Change-Id: Ia76f7f5b1793a16ac17285d19429f70d722675cf Reviewed-by: Laszlo Agocs <laszlo.agocs@theqtcompany.com>
* Adjust egl config.test to pass sysrooted libsDonald Carr2016-04-081-1/+1
| | | | | | | | | | | We should really start using -L=/foo and -I=/foo inside of sysroots, this test was preventing us from doing so (while arguably buying us nothing). Change-Id: If6e67631c585493871231e5d8a9354fa72e07343 Reviewed-by: Laszlo Agocs <laszlo.agocs@theqtcompany.com> Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@theqtcompany.com>
* remove bogus no_cxx_module conditionalOswald Buddenhagen2016-04-081-1/+1
| | | | | | | | | | this flag specifies that the plugin does not track a qt module. this is entirely unrelated to whether the plugin should be versioned. amends f54a3d783. Change-Id: Ibd3e9bedf488dc58e6354ccf7dd33d974e5f52c2 Reviewed-by: Joerg Bornemann <joerg.bornemann@theqtcompany.com>
* winrt: Add capabilities as specified by modulesMaurice Kalinowski2016-04-074-2/+27
| | | | | | | | | | | | | | | | | | | | | | | So far no capabilities (but internetClient for Windows 10) were added by default, which forced developers to always manually edit the WINRT_MANIFEST.capabilities(_device) property. This allowed to leave out non-required capabilities and keep the created manifest clean, examples being microphone for multimedia. However, this also breaks first user experience as deeper knowledge about this topic is required. Furthermore this is inconsistent with other platforms like Android, where all capabilities are set by default and developers need to edit the manifest manually in any case. With this change, modules can define the capability set to enable all features in the module. If developers want to disable some again, they need to adapt the generated manifest. From our experience this needs to be done in any case, latest at publishing stage when the store manipulates the manifest. Task-number: QTBUG-38802 Change-Id: I6d522268ee0afbfa00a30dbdd5e6ec9f415bebf3 Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@theqtcompany.com>
* winrt: Enable windeployqt by default for Visual Studio projectsMaurice Kalinowski2016-03-231-0/+2
| | | | | | | | | | | | | | | Using Visual Studio a user very seldom wants to disable the automatic invocation of windeployqt. Hence switch from opt-in behavior to opt-out. This also fixes first user experience to invoke qmake –tp vc and then hit run on examples. [ChangeLog][Platform Specific Changes][qmake] qmake-generated Visual Studio projects now automatically invoke windeployqt by default. Task-number: QTBUG-52008 Change-Id: Iee1607269c38c7f6c726f554978ac05477bebe5e Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@theqtcompany.com>
* win32-icc: Make it possible to rename the ANGLE libraries.Friedemann Kleint2016-03-221-0/+4
| | | | | | | | | | | Adapt the mkspec which was forgotten in change b0ec05f27bff6065c97e6e4a0f1606da7472d18c. Task-number: QTBUG-48431 Task-number: QTBUG-51686 Change-Id: Ie95e650de8b7a7027979ec637fb77c7f0357a598 Reviewed-by: Oliver Wolff <oliver.wolff@theqtcompany.com> Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@theqtcompany.com>
* Merge remote-tracking branch 'origin/5.6.0' into 5.6Frederik Gladhorn2016-03-171-1/+1
|\ | | | | | | Change-Id: Iac8ff05cd76cbacf859138a73e8e2ed0a979c75a
| * fix non-git non-prefix shadow buildsOswald Buddenhagen2016-03-021-1/+1
| | | | | | | | | | | | | | | | | | | | | | | | | | turns out we need forwarding .pris in this case: without them, QT_MODULE_INCLUDE_BASE points into the build dir, so we fail to find the pre-generated headers. an alternative would be writing primary module .pris which already take that into account, but that would just add even more arcane code paths. Task-number: QTBUG-51521 Change-Id: I59f2a0d3f2095c9dfa0a8d1cabfc007a30bd2d23 Reviewed-by: Simon Hausmann <simon.hausmann@theqtcompany.com>
* | support relative paths in configure -ROswald Buddenhagen2016-03-163-15/+24
| | | | | | | | | | | | | | | | | | [ChangeLog][configure][Unix] configure -R now supports paths relative to -libdir. Change-Id: Ie56264a6dedcbaf5577c7ef44b056c8a7870ef48 Reviewed-by: Joerg Bornemann <joerg.bornemann@theqtcompany.com> Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@theqtcompany.com>
* | make -D/-I/-L/-l/-R not affect bootstrapped toolsOswald Buddenhagen2016-03-161-0/+4
| | | | | | | | | | | | | | | | | | it's likely that these will be wrong, and the bootstrapped tools usually don't need them anyway. should they turn out necessary after all, we need to add -H* variants of the flags. Change-Id: I15c54c5e25d20ebd474073a530f00254842f515d Reviewed-by: Joerg Bornemann <joerg.bornemann@theqtcompany.com>
* | delay application of configure -D/-I/-L/-l/-R flagsOswald Buddenhagen2016-03-162-0/+30
| | | | | | | | | | | | | | | | | | | | | | | | it is important that the flags coming from the current qt build appear first, as otherwise a pre-existing qt installation may interfere with the build. the windows configure does not have any of this magic to start with. Task-number: QTBUG-6351 Change-Id: Iacc1d9b5aa9eed9a5f0513baef9f6c6ffcef0735 Reviewed-by: Joerg Bornemann <joerg.bornemann@theqtcompany.com>
* | build with explicitlib after allOswald Buddenhagen2016-03-161-0/+4
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | unlike speculated in 2fe363514, this is not a workaround at all: it causes that libraries' public link interfaces (LIBS) are exported in the first place. unlike with staticlib, this does not export LIBS_PRIVATE, so it wouldn't even be a particularly effective workaround for rpath brokenness anyway. the problem was pretty well hidden by the qt module system, which at the level of libraries is pretty redundant with the .prl file handling, which shows just how stupid the whole "design" is. unlike before, we now enable explicitlib for all libraries, not just qt modules - we enable create_prl for all of them as well, after all. an immediate effect of this change is that it fixes linking on RaspPI: the qtcore headers make the user code require linking libatomic, so we must add it to our public link interface. Task-number: QTBUG-51621 Change-Id: I5742c88694db8e8a9b79d17222dc6df2b38e5ab2 Reviewed-by: Joerg Bornemann <joerg.bornemann@theqtcompany.com> Reviewed-by: Allan Sandfeld Jensen <allan.jensen@theqtcompany.com>
generated by cgit v1.2.3 (git 2.25.1) at 2024-05-09 02:33:59 +0000