summaryrefslogtreecommitdiffstats
path: root/src
Commit message (Collapse)AuthorAgeFilesLines
* ucl: fix shared buildBoris Nagaev2015-11-091-0/+52
| | | | | | | | | | | | | | | | | | | | | | | libtool could not find -lwinmm and produced the following message: *** Warning: linker path does not have real file for library -lwinmm. *** I have the capability to make that library automatically link in when *** you link to this library. But I can only do this if you have a *** shared version of the library, which you do not appear to have *** because I did check the linker path looking for a file starting *** with libwinmm and none of the candidates passed a file format test *** using a file magic. Last file checked: /usr/lib/mxe/usr/lib/gcc/i686-w64-mingw32.shared/5.2.0/../../../../i686-w64-mingw32.shared/lib//libwinmm.a *** The inter-library dependencies that have been dropped here will be *** automatically added whenever a program is linked with this library *** or is declared to -dlopen it. *** Since this library must not contain undefined symbols, *** because either the platform does not support them or *** it was explicitly requested with -no-undefined, *** libtool will only create a static version of it. After that, it created a static library instead of shared one. Even -no-undefined did not help.
* add package uclBoris Nagaev2015-11-091-0/+25
|
* gmp: updateMark Brand2015-11-051-5/+4
|
* sqlite: updateMark Brand2015-11-021-2/+2
|
* enbale MXE_GCC_THREADS and allow target spec <triplet>.<link>.<threads>Tony Theodore2015-10-311-1/+1
| | | | | | | * enables optional c++11 <thread>, <mutex>, and <future> in libgcc * avoid double build advice: http://sigrok.org/gitweb/?p=sigrok-util.git;a=blob;f=cross-compile/mingw/README;h=248af546f939a81855dda33142d834eecf2979c2;hb=HEAD#l44 * not heavily tested but should be released for feedback
* gcc pthreads: build pthreads and mingw-w64 headers inline and enable libgompTony Theodore2015-10-3115-94/+36
| | | | | | | | | * fixes "~winpthreads changes ~pthread_signal.h" etc. reported by build-pkg mingw-w64 installs dummy headers if winpthreads isn't built * enables libgomp to avoid double-build (see #331) * no change in openmp-validation (still 20 failures - taken with a grain of salt as the batch file test runner isn't a reliable perl substitute) * pthreads virtual package kept for future testing
* mxe-conf: add CMAKE_PREFIX_PATH to toolchain fileTony Theodore2015-10-271-0/+1
| | | | see #904
* dbus: updateMark Brand2015-10-271-2/+2
|
* postgresql: fix socket error checking for mingwEngine Feeder2015-10-271-0/+152
| | | | | | | | | | | | | | | | | | When the client library of PostgreSQL is compiled with a MinGW-w64 environment the resulting libpq does not work. This has been reported upstream with two bug reports and a patch was submitted by Michael Cronenworth <mike(at)cchtml(dot)com> (1) but it has not been picked up by upstream. This is confirmed to build on Arch Linux x64 with targets: x86_64-w64-mingw32.static and i686-w64-mingw32.static and works under Wine, Windows 7 and Windows 10 (both 32/64bit binaries) (1) http://www.postgresql.org/message-id/520EBC3D.1010103@cchtml.com fixes #690
* aspell: disable curses supportTony Theodore2015-10-271-0/+1
|
* gendef: install to target specific prefixTony Theodore2015-10-251-1/+1
|
* gcc: cleanup lib installationTony Theodore2015-10-251-1/+7
|
* pkgconf yasm: enable native builds and consolidate rulesTony Theodore2015-10-252-22/+19
|
* mxe-conf: consolidate rules and install cmake modules under usrTony Theodore2015-10-251-12/+14
|
* gcc and deps: use native builds and consolidate rulesTony Theodore2015-10-255-11/+31
|
* remove gcc-* virtual packagesTony Theodore2015-10-255-112/+1
|
* box2d: fix case in subdirTony Theodore2015-10-241-1/+1
| | | | closes #954
* qtbase: rebase patchesMark Brand2015-10-231-22/+49
|
* vamp-plugin-sdk: use MAKE_SHARED_FROM_STATIC for shared buildTony Theodore2015-10-221-3/+8
|
* liblaxjson: update and use cmake build/install commandsTony Theodore2015-10-212-12/+40
| | | | still needs some work to provide static/shared as an option
* box2d: move to github and updateTony Theodore2015-10-211-13/+9
| | | | also fixes dll installation
* liblaxjson: fix installation directoriesTony Theodore2015-10-211-1/+8
|
* add CMake module FindFriBiDi.cmakeBoris Nagaev2015-10-191-0/+15
| | | | | | Now the following CMake code finds FriBiDi: find_package(FriBiDi)
* qtbase: fix log corruption with many zero bytesBoris Nagaev2015-10-181-0/+33
| | | | close #938
* openscenegraph-3: Use uintptr_tTimothy Gu2015-10-181-1/+1
| | | Prevents theoretical undefined (or implementation-defined?) behavior on 32-bit machines.
* sqlite: updateMark Brand2015-10-171-2/+2
|
* qt5: remove download related variablesTony Theodore2015-10-171-9/+1
| | | | in rare cases, it would try downloading qtbase again
* Revert "boost pthreads qt sdl: disable cmake tests"Tony Theodore2015-10-174-8/+4
| | | | This reverts commit f1dd70802a4e20c0f8c849aa44f24651895c7cfc.
* qtbase: improve update macroMark Brand2015-10-161-1/+2
|
* harfbuzz: updateMark Brand2015-10-161-2/+2
|
* Merge pull request #940 from LuaAndC/decrease-cmake-versionTony Theodore2015-10-161-4/+5
|\ | | | | Decrease cmake version and check for policies used
| * cmake test: use policy CMP0020 only if availableBoris Nagaev2015-10-161-1/+3
| | | | | | | | | | | | | | | | | | Policy CMP0020 was introduced in cmake 2.8.11. Debian CMake version 2.8.9 is sufficient for building Qt. See https://github.com/LuaAndC/mxe/commit/6133f6ec74aee1b5fa531782e585710 close #909
| * cmake test: decrease min version 2.8.11 -> 2.8.9Boris Nagaev2015-10-161-3/+2
| | | | | | | | | | | | | | Debian Wheezy has CMake version 2.8.9 and it is sufficient for building Qt (version 2.8.11 was chosen as it is recommended by Qt docs). see #909
* | Merge pull request #933 from LuaAndC/fix-cmake-buildTony Theodore2015-10-161-3/+15
|\ \ | |/ |/| cmake wrapper: don't pass -DCMAKE_TOOLCHAIN_FILE to cmake when with --build
| * cmake: no -DCMAKE_TOOLCHAIN_FILE with some optionsBoris Nagaev2015-10-151-3/+15
| | | | | | | | | | | | | | | | | | | | CMake with a toolchain file doesn't work properly in some modes. * --build * --system-information * -E close #932
* | Merge pull request #937 from saiarcot895/upstream-fix-qt-mysql-detectionBoris Nagaev2015-10-161-0/+1
|\ \ | |/ |/| Ensure that our mysql-config is used and not the system mysql-config …
| * Ensure that our mysql-config is used and not the system mysql-config when ↵Saikrishna Arcot2015-10-151-0/+1
| | | | | | | | building Qt 5.
* | Merge pull request #936 from saiarcot895/upstream-fix-osg-patchTimothy Gu2015-10-151-1/+1
|\ \ | | | | | | Fix OpenSceneGraph patch.
| * | Fix OpenSceneGraph patch.Saikrishna Arcot2015-10-151-1/+1
| |/
* | update qt5 modulesMark Brand2015-10-1525-202/+38
|/
* sqlite: updateMark Brand2015-10-142-40/+2
|
* Merge pull request #930 from kolewu/fix-vmimeTimothy Gu2015-10-141-1/+1
|\ | | | | fix project path for vmime extra file
| * fix project path for vmime extra fileUwe Koloska2015-10-141-1/+1
| | | | | | | | | | | | vmime uses a file from the src-directory of mxe but references it using a relative path that is only working if the temporary build directory is inside the mxe project path at the fixed position.
* | use better type for pointer castUwe Koloska2015-10-141-2/+2
|/
* libgsf: portability fix for sedTony Theodore2015-10-141-0/+1
|
* Merge pull request #926 from LuaAndC/fix-blasTony Theodore2015-10-141-4/+3
|\ | | | | update & fix blas
| * update blasBoris Nagaev2015-10-131-4/+3
| | | | | | | | | | see #922 see http://lists.nongnu.org/archive/html/mingw-cross-env-list/2015-10/msg00009.html
* | harfbuzz: updateMark Brand2015-10-131-2/+2
|/
* vmime: fix update macroMark Brand2015-10-121-1/+1
| | | | | The subdir and filename are based on the first 7 characters of the sha1, so it's convenient to treat this as the version.
* qtbase: consolidate patchesMark Brand2015-10-122-96/+89
|
generated by cgit v0.12 at 2025-03-23 21:40:46 (GMT)