Age | Commit message (Collapse) | Author | |
---|---|---|---|
2016-12-21 | [cmake] Move to root folder: Move files | h.udo | |
2016-12-20 | Merge pull request #11224 from cjmayo/gtk-update-icon-cache | Wolfgang Schupp | |
[cmake] linux: don't run gtk-update-icon-cache | |||
2016-12-20 | Merge pull request #11160 from Rechi/cmakeTestHeaders | Wolfgang Schupp | |
[cmake] core_add_test_library also add headers to IDE | |||
2016-12-19 | [cmake] linux: don't run gtk-update-icon-cache | Chris Mayo | |
2016-12-18 | [win] define __SSE__ and __SSE2__ for the whole platform | Rechi | |
2016-12-16 | Merge pull request #11144 from garbear/fix-texturepacker-install | jenkins4kodi | |
2016-12-14 | Merge pull request #11172 from AlwinEsch/remove-shared-addon-lib-games | Alwin Esch | |
[retroplayer] rework to new code style without shared lib | |||
2016-12-14 | Merge pull request #11157 from wsnipex/cmake_mmal | Wolfgang Schupp | |
Cmake mmal | |||
2016-12-14 | [retroplayer] rework to new code style without shared lib | Alwin Esch | |
2016-12-13 | Merge pull request #11154 from AlwinEsch/remove-shared-addon-lib-peripheral | Alwin Esch | |
[peripheral] bye, bye libKODI_peripheral.cpp | |||
2016-12-13 | Merge pull request #11151 from cjmayo/docdir | jenkins4kodi | |
2016-12-13 | [cmake] core_add_test_library also add headers to IDE (same as in ↵ | Rechi | |
core_add_library) | |||
2016-12-13 | [cmake] quote vars in string functions | wsnipex | |
fixes errors when the vars are empty | |||
2016-12-13 | Merge pull request #11088 from notspiff/fix_toolchain | Wolfgang Schupp | |
fixed: the assumption of toolchain = crosscompile cause issues | |||
2016-12-13 | [cmake] linux: report docdir and install docs into docdir | Chris Mayo | |
2016-12-13 | [peripheral] change libKODI_peripheral.h to style of the others | Alwin Esch | |
2016-12-12 | [depends] remove rtmp as it's included through ffmpeg or as binary addon | Martijn Kaijser | |
2016-12-12 | fixed: the assumption of toolchain = crosscompile cause issues | notspiff | |
ref https://github.com/gentoo/gentoo/pull/3027 | |||
2016-12-11 | CMake: Fix TexturePacker installation directory | Garrett Brown | |
2016-12-10 | Merge pull request #10509 from fetzerch/cmake_xcode_ios | Christian Fetzer | |
[cmake/ios] Enable Xcode generator for IOS | |||
2016-12-09 | Merge pull request #10946 from Kwiboo/cmake-win32 | Philipp Temminghoff | |
[cmake/win32] fixes for building with abnormal configuration | |||
2016-12-09 | [cmake/ios] Enable Xcode generator for IOS | Christian Fetzer | |
- Adapt Toolchain file to build a bundles with xcode. This change has to be done in the Toolchain file because CMake otherwise fails to run some compile checks. - Unify path where bundle is built for Makefiles and Xcode to build/$(CONFIGURATION)-$(EFFECTIVE_PLATFORM_NAME) - Set only one architecture to prevent building multi-arch bundles. - Simplify Install.cmake. Existing scrips are called now as post build steps so that the bundle can be executed with Xcode - Apply xcconfig settings. Note: Two adaptions to the existing "autotools/Xcode" scripts were needed. - CMake cannot have the bundle name be different to the binary and we cannot call the binary "Kodi.bin". The script has been adapted to adapt the library execution path directly in place. - CMake calls mkdeb.sh with "Debug-iphoneos", hence the script has been adapted to match strings starting with "debug" or "release". Reference: - https://cmake.org/Bug/view.php?id=15329 - http://stackoverflow.com/questions/33660608/how-can-i-disable-xcode-bitcode-within-a-cmake-project | |||
2016-12-09 | [cmake] Set correct output directory for core_add_shared_library | Christian Fetzer | |
Set the correct output-directory for shared libraries on all platforms (module vs. shared library with different extensions on Darwin) and for all configurations (Debug, Release, RelWithDebInfo, MinSizeRelease). | |||
2016-12-08 | [cmake] fix: command not found "GIT_SHALLOW 1" | wsnipex | |
2016-12-02 | Enable VAAPI support for Mir using DRM | Brandon Schaefer | |
2016-12-02 | [cmake] linux: install game headers | wsnipex | |
2016-12-02 | Merge pull request #10898 from BrandonSchaefer/master | Rainer Hochecker | |
Mir windowing system for Kodi | |||
2016-12-01 | [retroplayer] Game add-ons | Garrett Brown | |
Thanks to Themaister for rewind functionality, fetzerch for mouse support, file length check and cmake modifications, topfs2 for fixing a crash when loading game clients, eibma for fixing linux compilation errors, a1rwulf for catching a missing callback symbol and fixing some rebase errors, and to notspiff for helping with the rebrand and cmake. | |||
2016-12-01 | Mir windowing system | Brandon Schaefer | |
2016-11-30 | [cmake] Support @MINGW_TOOLCHAIN_FILE@ in dependency's flags.txt | montellese | |
2016-11-30 | [binary addons] Support platform specific deps.txt | montellese | |
2016-11-30 | [cmake] Move downloading of patch.exe tool into generic tool script | montellese | |
2016-11-30 | [addon/depends] Clone only the last commit of dependencies | Christian Fetzer | |
This has been introduced with CMake 3.6 but is ignored on older versions. | |||
2016-11-30 | [addon/depends] Handle autocrlf for depends | Christian Fetzer | |
On Windows, if we have addon dependencies, we need to forcefully disable autocrlf because otherwise the patches will not apply. Patches are typically in LF format (and the build system automatically adds --binary if necessary). But the patches will also not apply if the source files are in CLRF which can happen if the user has core.autocrlf=true in the global git config. A better way of fixing would be to pass the config already to clone: https://gitlab.kitware.com/cmake/cmake/issues/15799 | |||
2016-11-30 | [cmake] fix build error when version tag is empty | Jonas Karlman | |
2016-11-27 | Merge pull request #10988 from wsnipex/texturepacker-ext | Wolfgang Schupp | |
[cmake] allow using an existing Texturepacker binary | |||
2016-11-25 | [cmake] allow using an existing Texturepacker binary | wsnipex | |
2016-11-25 | Merge pull request #10967 from Memphiz/fix_addon_platforms_txt | Cristiano A. Silva | |
Fixed missing binary addons | |||
2016-11-23 | fixed: probe for inotify in build system | Arne Morten Kvarving | |
2016-11-23 | [cmake] fix linking wrapped libs with gold linker in debug builds | wsnipex | |
make sure our custom command uses the correct linker flags | |||
2016-11-21 | [cmake] - fixe wrong parsing of platforms.txt (platforms might be on ↵ | Memphiz | |
seperate lines...) after 897094feeef8357eaf7ce2a4328a062b1af9d151 | |||
2016-11-18 | Merge pull request #10865 from fetzerch/cmake_addon_depends | Christian Fetzer | |
[binary-addons] Rename DEPENDS_PATH to ADDON_DEPENDS_PATH | |||
2016-11-16 | Merge pull request #10863 from fetzerch/cmake_crosscompile | Christian Fetzer | |
[cmake] Unify usage of CMAKE_CROSSCOMPILE throughout the project | |||
2016-11-09 | [cmake] separate_arguments() is stupid | h.udo | |
2016-11-07 | Merge pull request #10785 from ronie/privacypolicy | Martijn Kaijser | |
add privacy policy | |||
2016-11-07 | update build files | ronie | |
2016-11-06 | [cmake/tests] Introduce variable to determine if we can execute build artefacts | Christian Fetzer | |
CMAKE_CROSSCOMPILING is set on all platforms if a Toolchain file is used. Even if we use a toolchain file, we still want to be able to execute unit tests on the build host for Linux and OSX. Introduce a new variable for this. | |||
2016-11-05 | [binary-addons] Rename DEPENDS_PATH into ADDON_DEPENDS_PATH | Christian Fetzer | |
DEPENDS_PATH is used twice in the code build system it's set to the system depends path (via toolchain file). In the addon system it's used to point to the place where addon dependencies should be built. This change renames the variable for binary-addons to ADDON_DEPENDS_PATH. The toolchain file can then be used for local addon builds. | |||
2016-11-03 | Merge pull request #10837 from fetzerch/cmake_cleanup | Christian Fetzer | |
[cmake/cleanup] file(REMOVE/REMOVE_RECURSE) ignores non existing files | |||
2016-11-01 | [cmake/cleanup] file(REMOVE/REMOVE_RECURSE) ignores non existing files | Christian Fetzer | |
There's no need to protect `file(REMOVE/REMOVE_RECURSE)` calls by `if(EXISTS)`. |