| Age | Commit message (Collapse) | Author | Files | Lines |
|---|
|
|
|
Now you don't need the latest Wayland installed to build with
newer protocols supported, as they'll build correctly; even if
your system can't use them, we can make intelligent decisions
at runtime about what's available on the current machine anyhow.
This also simplifies some logic and possible failure cases in
the configure and CMake scripts.
Fixes Bugzilla #4207.
|
|
This work was done by Jacob Lifshay and Mark Callow; I'm just merging it
into revision control.
|
|
libdir
Marcus von Appen
The LT_LDFLAGS in Makefile.in contain the $(DESTDIR) in -rpath, which instructs libtool to take a wrong path into account for linking.
The issue arises, if DESTDIR is passed at build time and installation time.
-rpath only should use $(libdir) for both SDL 1.2 and SDL 2.x.
|
|
Rafal Muzylo
"if we're already using libtool, why aren't we using it ?"; they've been inspired by the fact, that at that mark, neither libSDL2_test.a nor libSDL2main.a were being built correctly (not sure if it's fully broken or just because I've tested the out-of-tree build)
|
|
tschwinger@elitemail.org
Most ironically, although autoconf/automake-based builds install (pretty half-assed) CMake package configuration files, they're missing in installations resulting from CMake-based builds entirely.
A proper configuration file typically also loads target exports (implemented in patch 3572, also fixing this issue - see my comment on that issue for details).
I believe it would be best to let the dinosaurs go extinct and redirect all build efforts to the CMake end for two reasons:
1. It potentially provides the best user experience, but you'd have to give it some love and ship with less quirky buildfiles.
2. It would force distros to build SDL via CMake and thus would ensure target exports are actually available everywhere.
Various CMake patches I submitted today in summary (directly converted from the HG commits and `am`d onto a fork of a git mirror that happened to be on `tip`).
https://github.com/tschw/SDL/commits/patched
Fixing #2576 #3572, #3613, and this fresh ticket, which is almost entirely advertisement ;).
These already do to make SDL much less of a quirky fella to have in your dependency tree...
|
|
|
|
Since we are loading shared objects dynamically, build our own version of the
core protocol symbols, so that we in the future can include protocol
extensions.
|
|
|
|
Leonardo
Despite having a CMakeLists.txt file, SDL2 seems to be mainly built using the autotools system by package-managers. It is nice to have it but it is only useful if you are building SDL2 by yourself.
People that want to use CMake to find their already-installed SDL2 are using a FindSDL2.cmake module based on the old FindSDL.cmake. This is not deprecated but it is discouraged by the CMake devs (see http://www.cmake.org/Bug/view.php?id=14826), as these modules needs maintainers to be included in the official CMake installation.
To fix that, SDL and its accompanying libraries could configure a simple sdl2-config.cmake file, much like it does with sdl2.pc.in. We don't need to configure a full sdl2-config.cmake as in this post, http://forums.libsdl.org/viewtopic.php?t=10068&sid=ccf8abbf0d73fb03ae9cded991e60959 (because it depends on it being built with CMake). Using something as simple as http://www.cmake.org/Wiki/CMake/Tutorials/Packaging#Package_Configuration_Files is deemed enough. Here is another: http://quickgit.kde.org/?p=prison.git&a=blob&f=lib%2Fprison%2FPrisonConfig.cmake.in
|
|
|
|
|
|
Add V=1 to the make command line will show the full commands but by default
we just show the tool-type and the output file. This is generally much easier
on the eye and makes warnings and errors more clearly visible.
|
|
|
|
|
|
Setting the tag type will let libtool work even when it cannot infer
the type of the code being built. One way libtool may fail to infer
the tag type is if one uses a mock compiler (such as for static
analysis).
|
|
|
|
Will remove this again at some point in the future, though.
|
|
Fixes Bugzilla #1945.
|
|
|
|
so they would open properly on all systems.
Updated Makefile.in so "make dist" includes everything necessary for release.
|
|
|
|
|
|
- Fixes bug 1712 by not overwriting SDL_SubsystemRefCount in SDL_Init.
- Removes the SDL_initialized variable, and makes SDL_SubsystemRefCount
the canonical source of truth for whether or not a subsystem has been
initialized.
- Refactors SDL_InitSubSystem and SDL_QuitSubSystem to use helper
functions to manage refcount.
- Adds automated tests for SDL_Init/Quit*.
- Adds SDL_bits.h which contains SDL_MostSignificantBitIndex.
|
|
|
|
of Alfred Reynolds
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
It was very confusing to have configure generate an SDL_config.h and then not have it be used when building on Mac OS X or Windows. I'll just have to remember to use SDL_config_windows.h when building official releases that are supposed to be ABI compatible with Visual Studio.
|
|
Windows so that command line builds and IDE builds have ABI compatibility.
Make sure we don't clobber SDL_revision.h when building from Mercurial
|
|
|
|
|
|
|
|
|
|
|
|
Cleaned up dependencies on generating SDL_revision.h
Fixed 'make install' if you are not building from a Mercurial repository
|
|
from the repo directory and it confuses people, especially if they have no way of regenerating SDL_revision.h :)
|
|
|
|
|
|
|
|
|
|
steps.
The trick is that if you're using configure and you don't want to have SDL_config.h and SDL_revision.h to show up as modified, you need to configure and build from a separate directory.
You also need to include SDL_revision.h directly if you want to use the SDL_REVISION constant, as a side effect of these changes.
|
|
|
|
application.
|
