kmx git

Commit	Date	Message
0ad4b0b6	2018-05-21T12:00:21	thread: fixed compiler warnings on non-Linux systems that use pthread. (static function rtkit_setpriority was unused, moved it in with rest of __LINUX__ section.)
49881861	2018-05-21T11:54:09	audio: Patched to compile on Visual Studio. (It gets upset at the -2147483648, thinking this should be an unsigned value because 2147483648 is too large for an int32, so the negative sign upsets the compiler.)
4df859c5	2018-05-21T11:35:42	cpuinfo: Added SDL_HasAVX512F(). This checks for the "foundation" AVX-512 instructions (that all AVX-512 compatible CPUs support).
8543ad7d	2018-05-21T11:34:57	cpuinfo: Added some internal SIMD-aligned allocation functions. Fixes Bugzilla #4150 (sort of).
999af809	2018-05-18T13:09:30	Merged latest changes from Steam Link app
e2619f1d	2018-05-17T12:50:46	dynapi: don't let system loader resolve the initializer to the wrong version. Fixes problems launching Firewatch on Linux (which statically links SDL but also dynamically loads a system-wide copy from a plugin shared library) with a newer SDL build.
b7e88aaa	2018-05-16T02:03:06	audio: Added ARM NEON versions of audio converters. These are _much_ faster than the scalar equivalents on the Raspberry Pi that I tested on. Often 3x to 4x as fast!
cb0e614f	2018-05-15T02:29:35	audio: SSE2 float-to-int converters should clamp input. The scalar versions already do this.
a07e5815	2018-05-15T01:40:05	audio: Fix range on float-to-int data clamping. I can't tell if there was a good reason for this or it was just me getting numbers wrong due to exhaustion.
7832cb65	2018-05-15T01:35:53	audio: float to int converters should clamp inclusively. If we have to test if a sample is > 1.0f anyhow, we might as well use this to avoid the unnecessary multiplication when it's == 1.0f, too. (etc).
e2ec1eb1	2018-05-15T01:04:11	audio: converting int32 to/from float shouldn't use doubles. The concern is that a massive int sample, like 0x7FFFFFFF, won't fit in a float32, which doesn't have enough bits to hold a whole number this large, just to divide it to get a value between 0 and 1. Previously we would convert to double, to get more bits, do the division, and cast back to a float, but this is expensive. Casting to double is more accurate, but it's 2x to 3x slower. Shifting out the least significant byte of an int32, so it'll definitely fit in a float, and dividing by 0x7FFFFF is still accurate to about 5 decimal places, and the difference doesn't appear to be perceptable.
ed4fe4c9	2018-05-15T00:04:02	testresample: correctly output .wav files that have floating point audio.
4718791f	2018-05-14T00:03:39	cpuinfo: Make CPU flags easier to read and add to.
425149f4	2018-05-11T09:37:00	remove testvulkan.vcproj (was a VS2008 left-over.)
652d59fb	2018-05-10T09:02:39	make sure SDL_vsnprintf() nul terminates if it is using _vsnprintf The change makes sure that SDL_vsnprintf() nul terminates if it is using _vsnprintf() for the job. I made this patch for Watcom, whose _vsnprintf() doesn't guarantee nul termination. The preprocessor check can be extended to windows in general too, if required. Closes bug #3769.
c11ae93a	2018-05-10T08:28:00	SDL_stdinc.h: move the alloca() includes before begin_code.h
fe032ff4	2018-05-10T08:25:23	do the direct3d tap dance for overscan hint only if SDL_VIDEO_RENDER_D3D == 1
a4d0571e	2018-05-07T20:10:12	Reverted change for bug 4152 - restrict the win10 mouse bug workaround to win10 v1709 only Daniel Gibson Sorry, but it seems like Microsoft didn't fix the issue properly. I just updated my Win10 machine, it now is Version 1803, Build 17134.1 I tested with SDL2 2.0.7 (my workaround was released with 2.0.8) and still got lots of events that directly undid the prior "real" events - just like before. (See simple testcase in attachement) By default it sets SDL_HINT_MOUSE_RELATIVE_MODE_WARP - which triggered (and on my machine still triggers) the buggy behavior. You can start it with -raw, then it'll not set that hint and the events will be as expected. The easiest way to see the difference is looking at the window title, which shows accumulated X and Y values: If you just move your mouse to the right, in -raw mode the number just increases. In non-raw mode (using mouse warping) it stays around 0. I also had a WinAPI-only testcase: https://gist.github.com/DanielGibson/b5b033c67b9137f0280af9fc53352c68 It just calls SetCursorPos(320,240); on each WM_MOUSEMOVE event, and it also logs all those events to a mouseevents.log textfile. This log indeed looks a bit different since the latest Win10 update: It seems like all those events with x=320 y=240 do arrive - but only after I stopped moving the mouse - even though the cursor seems to be moved back every frame (or so). So moving the mouse to the right gives X coordinates like 330, 325, 333, 340, 330, ... and then when stopping movement I get lots of events with X coordinate 320
eb14b635	2018-05-07T19:52:25	Fixed bug 4134 - Render targets lose scale quality after minimizing a fullscreen window Olli-Samuli Lehmus If one creates a window with the SDL_WINDOW_FULLSCREEN_DESKTOP flag, and creates a render target with SDL_SetHint(SDL_HINT_RENDER_SCALE_QUALITY, "linear"), and afterwards sets SDL_SetHint(SDL_HINT_RENDER_SCALE_QUALITY, "nearest"), after minimizing the window, the scale quality hint is lost on the render target. Textures however do keep their interpolation modes.
c04dca0d	2018-05-07T19:26:02	Fixed bug 4159 - Windows headers are included after packing alignment change lectem The SDL_syswm.h header includes the windows.h header after including begin_code.h which changes the structure packing alignment. It seems this is not safe as suggested by the following warning : warning C4121: 'JOBOBJECT_IO_RATE_CONTROL_INFORMATION_NATIVE_V2': alignment of a member was sensitive to packing
0bade9b4	2018-05-05T10:31:03	Fixed bug 4154 - Fix three empty variable tests in configure Zack Middleton Running top-level SDL configure on macOS 10.11 resulted in the errors below because automake removed the brackets about the tests. ./configure: line 15756: : command not found ./configure: line 15759: -Iinclude -I/Users/zack/SDL/include -idirafter /Users/zack/SDL/src/video/khronos : No such file or directory ./configure: line 15763: : command not found
606c5a58	2018-05-05T10:27:53	Fixed bug 4152 - Windows 10 v1803 update seems to have fixed the jumping mouse bug (see bug #3931.)
386790ef	2018-04-23T22:29:14	Improved error messages when Vulkan isn't configured (thanks Daniel Gibson!)
8f780e76	2018-04-23T22:17:56	Fixed build
f521b22e	2018-04-23T22:07:56	Added SDL_THREAD_PRIORITY_TIME_CRITICAL
db94dfb1	2018-04-23T21:55:59	Fixed bug 4144 - CMake complains about trailing spaces in sdl2.pc Azamat H. Hackimov When you try use SDL2 2.0.8 in CMake project in Linux, it complains about trailing spaces in sdl2.pc: CMake Error at CMakeLists.txt:147 (add_executable): Target "TestSimpleMain" links to item "-L/usr/lib64 -lSDL2 " which has leading or trailing whitespace. This is now an error according to policy CMP0004.
a1b8fa60	2018-04-23T21:50:03	TryLockMutex: Fix error handling for TryLockMutex Christian Herzig pthread_mutex_trylock() and by the way, pthread_mutex_lock() do not set errno. Pthread-methods directly return error code as int. See related man-pages for details.
a9ae1b50	2018-04-23T20:24:12	Handle NULL return from SDL_DBus_GetContext()
b5d231ee	2018-04-23T19:20:12	Handle NULL return from SDL_DBus_GetContext()
43231256	2018-04-23T19:18:52	Added SDL_LinuxSetThreadPriority() to directly set the priority of a Linux thread (tid) This function tries using RealtimeKit connecting over DBUS as needed.
816a6e68	2018-04-23T17:10:36	Added support for adjusting thread priorities using Linux RealtimeKit Michael Sartain This is a quick pass at adding Linux RealtimeKit thread priority support to SDL. It allows me to bump the thread priority to high without root privileges or setting any caps, etc. rtkit readme here: http://git.0pointer.net/rtkit.git/tree/README
dc8b55e5	2018-04-16T02:11:09	coreaudio: Use the standard SDL audio thread instead of spinning a new one. Fixes corner cases, like the audio callback not firing if the device is disconnected, etc.
2df59062	2018-04-15T17:42:09	wayland: zxdg_shell_v6 needs a configure event before using a surface at all. Fixes Bugzilla #4109. Fixes Bugzilla #4119.
5e8c8167	2018-04-15T09:37:51	Fixed bug 4135 - Broken symlink libSDL2.so since rev11940 Tiago O. Symlink points to the wrong folder, and target will always have debug postfix, so it'll be broken for other build types.
b7228bc5	2018-04-11T18:28:03	Added support for the GameSir G3w
9856d967	2018-04-11T06:16:23	Fix the include path in the installed CMake target import file Previously the include path was {INSTALL_PREFIX}/include, it is now {INSTALL_PREFIX}/include/SDL2 to be consistent with the other build and package configuration systems. Fixes #4128.
6b5ed0fd	2018-04-10T08:03:54	Added debug postfix to install command and fixed library path
6a0ef0cd	2018-04-09T10:37:31	SDL: On Windows, have SDL_ShowWindow() not activate the window if the window has the WS_EX_NOACTIVATE window flag.
4d78a995	2018-03-26T12:38:29	Fixed bug where an SDL window that was activated while hidden could never be shown. Test code: { SDL_Window *win = SDL_CreateWindow( "Dummy", SDL_WINDOWPOS_CENTERED, SDL_WINDOWPOS_CENTERED, 128, 128, SDL_WINDOW_HIDDEN ); SDL_SysWMinfo info; SDL_VERSION( &info.version ); SDL_GetWindowWMInfo( win, &info ); SetActiveWindow( info.info.win.window ); { DWORD then = SDL_GetTicks(); while ( ( SDL_GetTicks() - then ) < 3000 ) { SDL_Event evt; SDL_PollEvent( &evt ); } SDL_ShowWindow( win ); then = SDL_GetTicks(); while ( ( SDL_GetTicks() - then ) < 3000 ) { SDL_Event evt; SDL_PollEvent( &evt ); } } SDL_DestroyWindow( win ); }
e14278ef	2018-03-24T10:26:40	Fixed bug 3804 - Message box on Windows truncates button ID Simon Hug I just wanted to fix a simple compiler warning in SDL_ShowMessageBox on Windows (which Sam fixed recently) and ended up finding some issues. Attached patch fixes these issues: - Because Windows only reports the lower 16 bits of the control identifier that was pushed, the button IDs used by SDL (C type int, most likely 32 bits) can get cut off. - The documentation states (somewhat ambiguously) that the button ID will be -1 if the dialog was closed, but the current code sets 0. For SDL 2.1, I think this should be a return code of SDL_ShowMessageBox itself. That will free up the button ID and it seems a more appropriate place for signaling this event. - Ampersands in controls will create mnemonics on Windows (underlined letters that, if combined with the Alt key, will push the button). I was thinking of adding a hint or flag to let the users enable it, but that might have unexpected results. - When the size of the text gets calculated, it doesn't use the same parameters as the static control. This can cut off text or wrap it weirdly. - On Windows, the Tab key is used to switch between control groups and sometimes between buttons in dialogs. This didn't seem to work correctly. Attached patch also adds: - Icons. Just the system ones that can be loaded with the ordinals IDI_ERROR, IDI_WARNING and IDI_INFORMATION. - A button limit of 2^16 - 101. - Some more specific error messages, but they never reach the user because how SDL_ShowMessageBox handles them if an implementation returns with an error.
b41b9d34	2018-03-19T14:52:53	Fixed return value
8e062f69	2018-03-19T14:42:51	Generalized the handling of instantaneous guide button presses so there's a minimum of 100 ms between guide button press and release. This happens with at least the following controllers: All Apple MFI controllers, ASUS Gamepad, XiaoMi Bluetooth Controller
d529b001	2018-03-19T13:16:11	Added mapping for the ASUS Gamepad removing the guide button, which doesn't generate events even though it's reported in the Android APIs.
f536fbea	2018-03-16T11:08:53	Reimplemented Android cursor API support using reflection so it builds with older SDKs
e20d4173	2018-03-15T18:22:48	Added Android custom cursor implementation This is commented out in SDLActivity.java, with the note #CURSORIMPLEENTATION because it requires API 24, which is higher than the minimum required SDK
1cfbe664	2018-03-12T18:41:06	Added Mac OpenGL ES configure support
d401a77e	2018-03-10T21:22:42	Fixed bug 4102 - define _WIN32_WINNT_WIN7 if not already defined Ozkan Sezer The following patch defines _WIN32_WINNT_WIN7 if it is not already defined in core/windows/SDL_windows.c, similar to what is already there for _WIN32_WINNT_VISTA.
6d39e67b	2018-03-10T21:20:25	Fixed bug 4103 - SDL_cpuinfo.h provokes warnings with -Wundef on non-x86 hosts Felix Geyer Forwarding from https://bugs.debian.org/892087 quoting verbatim: The SDL2 header SDL_cpuinfo.h generates gcc warnings if the program using it compiles with the -Wundef warning. (In particular, this means that QEMU builds using it fail on at least sparc hosts, since QEMU dev builds use both -Wundef and -Werror.). /usr/include/SDL2/SDL_cpuinfo.h:63:5: warning: "HAVE_IMMINTRIN_H" is not defined, evaluates to 0 [-Wundef] #if HAVE_IMMINTRIN_H && !defined(SDL_DISABLE_IMMINTRIN_H)
0a5c1065	2018-03-10T21:16:14	Fix ARM builds with MSVC
cc7b2fc5	2018-03-10T21:13:50	Temporary fix for bug 3432 - macOS 10.12: small scrolls (1 wheel notch) don't generate events Eric Wasylishen This bug was reintroduced by https://hg.libsdl.org/SDL/rev/fcf24b38a28a The steps to reproduce are the same: run the "testrelative" SDL demo with "--info all", connect a USB mouse with a scroll wheel, and roll the scroll wheel one "notch". You'll get log output like: testdraw2[1644:67222] INFO: SDL EVENT: Mouse: wheel scrolled 0 in x and 0 in y (reversed: 1) in window 1 As far as I can tell macOS doesn't have an API for getting the number of "wheel notches"; I get a deltaY of 0.100006 for one "notch", and it's heavily accelerated (if you roll the wheel quickly you'll get large deltas). So NSEvent's deltaY is only meant to be used for scrolling a scroll view, with the given distance in points, not something like selecting an item in a game. Here's a temporary patch that at restores the foor/ceil in Cocoa_HandleMouseWheel. Not ideal, but at least it restores the ability to scroll one notch of a mousewheel.
129431b4	2018-03-08T16:32:22	Delay delivery of the pause button release on MFI controllers so it doesn't happen in the same frame as the button press
92847022	2018-03-07T18:10:01	Added a mapping for the latest firmware for the Xbox One S controller on Android
a2c1d83c	2018-03-07T18:09:58	Include a USB VID/PID for Apple MFI controllers This is just placeholder VID/PID, but allows code that works with VID/PID to identify the MFI controllers easily.
a8ac5885	2018-03-07T13:30:40	Added SDL_GameControllerMappingForDeviceIndex() to get the mapping for a controller before it's opened
9e651b69	2018-03-06T14:51:50	Try to dynamically create a default Android game controller mapping based on the buttons and axes on the controller. Include the controller USB VID/PID in the GUID where possible, as we do on other platforms.
2419d267	2018-03-02T22:53:25	Progress fixing bug 4100 - errors and warnings after changeset 11917 Ozkan Sezer 2018-03-02 20:02:37 UTC http://hg.libsdl.org/SDL/rev/d702b0c54e52 resulted in an error and two warnings when compiled with mingw. 1. Error from SDL_windowstaskdialog.h: In file included from src/video/windows/SDL_windowsmessagebox.c:29:0: src/video/windows/SDL_windowstaskdialog.h:23:54: error: expected ')' before 'HWND' This is fixed by removing unnecessary annotations: 2. Warning from SDL_assert.c: src/SDL_assert.c: In function 'SDL_ExitProcess': src/SDL_assert.c:138:1: warning: 'noreturn' function does return Indeed ExitProcess() is prototyped with DECLSPEC_NORETURN, but TerminateProcess() is not. This can be rectified by adding an exit() call in there. Do NOTE, however, that requires building with a libc: 3. Warning from SDL_windowsmessagebox.c: src/video/windows/SDL_windowsmessagebox.c: In function 'WIN_ShowMessageBox': src/video/windows/SDL_windowsmessagebox.c:513:9: warning: 'nCancelButton' may be used uninitialized in this function My lazy solution was manually initializing nCancelButton to 0.
e9b29ed0	2018-03-02T22:48:15	Fixed bug 4101 - configure needs regenerating after changeset 11894
ac2d1f67	2018-03-02T12:08:18	Fixed setting the layer drawable size Without this change the drawable had a size of 0 and the metal renderer asserted because the projection matrix wasn't set.
cef1c1c2	2018-03-02T14:10:25	windows: Restore patches for Task Dialogs and TerminateProcess(). 2.0.8 has shipped, these can live in revision control now!
003c0dce	2018-03-02T10:56:21	Use the real controller name for game controllers on iOS and Apple TV
4f695bc5	2018-03-01T09:32:53	Added tag release-2.0.8 for changeset dfbae17e87ac
e9fc66a0	2018-03-01T08:26:10	Added patch note for Android Studio support
0df7fe84	2018-03-01T08:22:56	Temporarily disabled fullscreen switching code on Android, until we can resolve bug 4096 - Enabling fullscreen on Android causes the app to toggle fullscreen mode continuously in a loop
eac37666	2018-03-01T08:08:20	Added a patch note for OpenGL ES support on Mac
3537c3e7	2018-02-28T10:39:41	Back out Task Dialog and TerminateProcess patches for 2.0.8. These can return to revision control once we ship.
6a1cfcce	2018-02-28T02:14:15	windows: dos2unix'd messagebox code, and (hopefully) fixed on MingW.
a749035f	2018-02-28T01:54:22	windows: Message boxes use Task Dialogs if possible (thanks, Jack!). This lets the message box have an icon. Unless the app has opted-in to using the v6 common controls, though, this will fall back to the usual SDL message boxes.
8536130a	2018-02-28T01:23:49	assert: Use TerminateProcess() on Windows, vs ExitProcess (thanks, Jack!). "What I have done is use TerminateProcess rather than ExitProcess. ExitProcess will cause Microsoft's leak detection to continue, TerminateProcess won't. It is also technically wrong to use ExitProcess in the case of aborting the application. Jack Powell Twitter @jack9267"
99683f7c	2018-02-26T08:39:25	Added patch note about the Vulkan Mac SDK
be6ca785	2018-02-25T23:02:09	Support official Vulkan SDK for macOS. This tries to load vulkan.framework or libvulkan.1.dylib before MoltenVK.framework or libMoltenVK.dylib. In the previous version, layers would not work for applications run-time loading the default library.
f9f45d0b	2018-02-25T19:51:34	Fixed bug 4097 - Segmentation fault by SDL_CreateThreadWithStackSize Dongsun Kim Normal case 1. [thread 1] SDL_CreateThreadWithStackSize calls SDL_SYS_CreateThread. 2. [thread 1] If successful, it calls SDL_SemWait. 3. [thread 2] SDL_RunThread calls SDL_SYS_SetupThread, SDL_ThreadID, SDL_SemPost. 4. [thread 1] SDL_CreateThreadWithStackSize calls SDL_DestroySemaphore, SDL_free. Crash case (Segmentation fault) 1. [thread 1] SDL_CreateThreadWithStackSize calls SDL_SYS_CreateThread. 2. [thread 1] If successful, it calls SDL_SemWait. --> Error return due to SIGNAL(SYSTEM or Real Time) at sem_wait(pthread). 3. [thread 1] SDL_CreateThreadWithStackSize calls SDL_DestroySemaphore, SDL_free. 4. [thread 2] SDL_RunThread calls SDL_SYS_SetupThread, SDL_ThreadID, SDL_SemPost. --> Segmentation fault at strlen or sem_post.
5cbb90da	2018-02-25T19:22:47	cmake: Set debug library name suffix per target instead of setting it globally. Don't add a suffix on android
3ea093cc	2018-02-25T10:15:00	SDL_test_fuzzer.c: fix strict aliasing warnings by using a union.
cbfee82d	2018-02-24T09:09:20	Moved Android patch note into the Android section
6cf4d0e4	2018-02-24T08:59:58	Fixed bug 4092 - CMake support for building everything in the "test" directory Eric Wasylishen Patch to support building the tests with cmake. Disabled by default, use: "cmake .. -DSDL_TEST=YES" to enable the tests. Tested on macOS 10.13 with the ninja, makefile, and Xcode generators, and Windows 10 with the Visual Studio 2017 generator.
849d042f	2018-02-24T08:58:22	Fixed bug 4091 - Undefined references to Android audio functions when SDL_AUDIO_DISABLED is on Manuel Sabogal If SDL is compiled with the Audio subsystem disabled there are some undefined references to the functions ANDROIDAUDIO_ResumeDevices and ANDROIDAUDIO_PauseDevices in the file src/video/android/SDL_androidevents.c.
8e4afb55	2018-02-24T08:42:04	Added patch note for Ryan's work
b33dbd27	2018-02-24T08:40:30	Added missing file to tvOS build
99a0c0f0	2018-02-24T08:23:44	Fixed MinGW-w64 build
5adc1b7d	2018-02-24T08:18:09	Updated documentation with API changes in SDL 2.0.8
0626486e	2018-02-23T19:12:04	Backout the vulkan change in d449dea10fc8 breaks Mir builds
e17c3219	2018-02-23T11:24:26	mir: Disable Mir by default as Mir supports Wayland clients Also remove enabling VK support for Mir
8891f591	2018-02-21T22:53:52	Backed out changeset 6c8521d53507 Apparently this still triggers a compiler warning, have to dig further.
fd8f12d2	2018-02-21T22:35:17	android: apparently they fixed this header at some point.
ac309df7	2018-02-21T22:27:09	vulkan: Possibly fix a compiler warning (-Wstrict-prototypes).
ed64d54d	2018-02-21T21:36:10	windows: added WIN_IsWindows7OrGreater().
7e1fa0ce	2018-02-21T21:34:35	wasapi: fixed typo in an assert message.
c7e43665	2018-02-21T21:34:06	wasapi: let Windows do the resampling for us if possible.
69958441	2018-02-21T09:58:21	Fix high-dpi support on macOS and simplify it and iOS variant. The detault drawableSize for a CAMetalLayer is its bounds x its scale. So it is sufficient to set the layer's scale to the desired value.
6e3d0a13	2018-02-21T12:42:30	Use new XInput mapping for Win10+ (Bugzilla #3960)
a0687a9c	2018-02-21T09:40:47	Fixed bug 4034 - Don't include _DllMainCRTStartup() if SDL_STATIC_LIB is defined.
c00858bf	2018-02-20T09:04:31	Fixed bug 4088 - Fix Metal link errors with test programs in SDLTest.xcodeproj Eric Wasylishen The following patch adds Metal.framework to the "link binary with libraries" section of each test program, with "status" set to "optional", which fixes link errors on all of the test programs. I'm not sure if this is a correct fix - the fact that this was necessary might indicate the static SDL2.a library has a hard dependency on Metal.framework (?) - but it gets the test programs working in Xcode again. It also adds testyuv_cvt.c to the testoverlay2 target, fixing a link error.
58f9be12	2018-02-18T09:09:56	Actually, this is needed for building with Visual Studio with both /MT and /MD. With the previous change, I get: 1> Creating library C:\projects\SDL\VisualC\Win32\Debug\SDL2.lib and object C:\projects\SDL\VisualC\Win32\Debug\SDL2.exp 1>LINK : error LNK2001: unresolved external symbol __DllMainCRTStartup@12
3c9d3336	2018-02-18T08:57:01	Fixed bug 4034 - Do we really need _DllMainCRTStartup() in every Windows build? Andreas Falkenhahn In src/SDL.c there is this code: _DllMainCRTStartup(HANDLE hModule, ... The comment says that this is needed on Watcom C for some reason but why is it included then when building with Visual C as well? Shouldn't it be only included when compiling on Watcom C then? I'm asking because this code caused me a lot of headaches because I'm building a DLL that contains SDL and I link using /MT and the _DllMainCRTStartup() symbol obviously led to lots of trouble but it wasn't clear to me where the problem was because all I got from the linker was: LNK2019: unresolved external symbol _main referenced in function ___tmainCRTStartup So I had to got through each and every object to see what the culprit was. See here for the full story: https://stackoverflow.com/questions/25067151/lnk2019-unresolved-external-symbol-main-referenced-in-function-tmaincrtstar/48177067#48177067 So if it isn't necessary on Visual C, please just leave that symbol out on Visual C so that it no longer leads to any trouble. Thanks.
75a58303	2018-02-17T23:57:57	pthread: fix error code checks (thanks, Andreas!). Most pthread functions return 0 on success and non-zero on error, but those errors might be positive or negative, so checking for return values in the Unix style, where errors are less than zero, is a bug. Fixes Bugzilla #4039.
2ea4419a	2018-02-17T20:18:48	yuv: patched to compile.
7c0c2c22	2018-02-17T20:10:13	yuv: fixed variable declaration shadowing warnings. Fixes Bugzilla #4062.
97494f53	2018-02-17T18:30:21	pulseaudio: Just read/dump captured data in FlushCapture. Apparently pa_stream_flush() doesn't work as expected: https://lists.freedesktop.org/archives/pulseaudio-discuss/2012-April/013328.html Fixes Bugzilla #4087.
6867f618	2018-02-16T14:56:28	video: put a spinlock around a global linked list. This should only contend if you're allocating or freeing surfaces from multiple threads at once, and then just for a short time. Fixes Bugzilla #4084.
8ddebfa0	2018-02-16T10:23:10	Fixed bug 4085 - X11: Allow configuring _NET_WM_BYPASS_COMPOSITOR through SDL hints Callum McGing This patch allows the user to disable the behaviour that blocks the compositor through a new hint: SDL_VIDEO_X11_NET_WM_BYPASS_COMPOSITOR. This allows tools or other windowed applications to behave properly under KWin.

0ad4b0b6

2018-05-21T12:00:21

thread: fixed compiler warnings on non-Linux systems that use pthread. (static function rtkit_setpriority was unused, moved it in with rest of __LINUX__ section.)

49881861

2018-05-21T11:54:09

audio: Patched to compile on Visual Studio. (It gets upset at the -2147483648, thinking this should be an unsigned value because 2147483648 is too large for an int32, so the negative sign upsets the compiler.)

4df859c5

2018-05-21T11:35:42

cpuinfo: Added SDL_HasAVX512F(). This checks for the "foundation" AVX-512 instructions (that all AVX-512 compatible CPUs support).

8543ad7d

2018-05-21T11:34:57

cpuinfo: Added some internal SIMD-aligned allocation functions. Fixes Bugzilla #4150 (sort of).

999af809

2018-05-18T13:09:30

Merged latest changes from Steam Link app

e2619f1d

2018-05-17T12:50:46

dynapi: don't let system loader resolve the initializer to the wrong version. Fixes problems launching Firewatch on Linux (which statically links SDL but also dynamically loads a system-wide copy from a plugin shared library) with a newer SDL build.

b7e88aaa

2018-05-16T02:03:06

audio: Added ARM NEON versions of audio converters. These are _much_ faster than the scalar equivalents on the Raspberry Pi that I tested on. Often 3x to 4x as fast!

cb0e614f

2018-05-15T02:29:35

audio: SSE2 float-to-int converters should clamp input. The scalar versions already do this.

a07e5815

2018-05-15T01:40:05

audio: Fix range on float-to-int data clamping. I can't tell if there was a good reason for this or it was just me getting numbers wrong due to exhaustion.

7832cb65

2018-05-15T01:35:53

audio: float to int converters should clamp inclusively. If we have to test if a sample is > 1.0f anyhow, we might as well use this to avoid the unnecessary multiplication when it's == 1.0f, too. (etc).

e2ec1eb1

2018-05-15T01:04:11

audio: converting int32 to/from float shouldn't use doubles. The concern is that a massive int sample, like 0x7FFFFFFF, won't fit in a float32, which doesn't have enough bits to hold a whole number this large, just to divide it to get a value between 0 and 1. Previously we would convert to double, to get more bits, do the division, and cast back to a float, but this is expensive. Casting to double is more accurate, but it's 2x to 3x slower. Shifting out the least significant byte of an int32, so it'll definitely fit in a float, and dividing by 0x7FFFFF is still accurate to about 5 decimal places, and the difference doesn't appear to be perceptable.

ed4fe4c9

2018-05-15T00:04:02

testresample: correctly output .wav files that have floating point audio.

4718791f

2018-05-14T00:03:39

cpuinfo: Make CPU flags easier to read and add to.

425149f4

2018-05-11T09:37:00

remove testvulkan.vcproj (was a VS2008 left-over.)

652d59fb

2018-05-10T09:02:39

make sure SDL_vsnprintf() nul terminates if it is using _vsnprintf The change makes sure that SDL_vsnprintf() nul terminates if it is using _vsnprintf() for the job. I made this patch for Watcom, whose _vsnprintf() doesn't guarantee nul termination. The preprocessor check can be extended to windows in general too, if required. Closes bug #3769.

c11ae93a

2018-05-10T08:28:00

SDL_stdinc.h: move the alloca() includes before begin_code.h

fe032ff4

2018-05-10T08:25:23

do the direct3d tap dance for overscan hint only if SDL_VIDEO_RENDER_D3D == 1

a4d0571e

2018-05-07T20:10:12

Reverted change for bug 4152 - restrict the win10 mouse bug workaround to win10 v1709 only Daniel Gibson Sorry, but it seems like Microsoft didn't fix the issue properly. I just updated my Win10 machine, it now is Version 1803, Build 17134.1 I tested with SDL2 2.0.7 (my workaround was released with 2.0.8) and still got lots of events that directly undid the prior "real" events - just like before. (See simple testcase in attachement) By default it sets SDL_HINT_MOUSE_RELATIVE_MODE_WARP - which triggered (and on my machine still triggers) the buggy behavior. You can start it with -raw, then it'll not set that hint and the events will be as expected. The easiest way to see the difference is looking at the window title, which shows accumulated X and Y values: If you just move your mouse to the right, in -raw mode the number just increases. In non-raw mode (using mouse warping) it stays around 0. I also had a WinAPI-only testcase: https://gist.github.com/DanielGibson/b5b033c67b9137f0280af9fc53352c68 It just calls SetCursorPos(320,240); on each WM_MOUSEMOVE event, and it also logs all those events to a mouseevents.log textfile. This log indeed looks a bit different since the latest Win10 update: It seems like all those events with x=320 y=240 do arrive - but only after I stopped moving the mouse - even though the cursor seems to be moved back every frame (or so). So moving the mouse to the right gives X coordinates like 330, 325, 333, 340, 330, ... and then when stopping movement I get lots of events with X coordinate 320

eb14b635

2018-05-07T19:52:25

Fixed bug 4134 - Render targets lose scale quality after minimizing a fullscreen window Olli-Samuli Lehmus If one creates a window with the SDL_WINDOW_FULLSCREEN_DESKTOP flag, and creates a render target with SDL_SetHint(SDL_HINT_RENDER_SCALE_QUALITY, "linear"), and afterwards sets SDL_SetHint(SDL_HINT_RENDER_SCALE_QUALITY, "nearest"), after minimizing the window, the scale quality hint is lost on the render target. Textures however do keep their interpolation modes.

c04dca0d

2018-05-07T19:26:02

Fixed bug 4159 - Windows headers are included after packing alignment change lectem The SDL_syswm.h header includes the windows.h header after including begin_code.h which changes the structure packing alignment. It seems this is not safe as suggested by the following warning : warning C4121: 'JOBOBJECT_IO_RATE_CONTROL_INFORMATION_NATIVE_V2': alignment of a member was sensitive to packing

0bade9b4

2018-05-05T10:31:03

Fixed bug 4154 - Fix three empty variable tests in configure Zack Middleton Running top-level SDL configure on macOS 10.11 resulted in the errors below because automake removed the brackets about the tests. ./configure: line 15756: : command not found ./configure: line 15759: -Iinclude -I/Users/zack/SDL/include -idirafter /Users/zack/SDL/src/video/khronos : No such file or directory ./configure: line 15763: : command not found

606c5a58

2018-05-05T10:27:53

Fixed bug 4152 - Windows 10 v1803 update seems to have fixed the jumping mouse bug (see bug #3931.)

386790ef

2018-04-23T22:29:14

Improved error messages when Vulkan isn't configured (thanks Daniel Gibson!)

8f780e76

2018-04-23T22:17:56

Fixed build

f521b22e

2018-04-23T22:07:56

Added SDL_THREAD_PRIORITY_TIME_CRITICAL

db94dfb1

2018-04-23T21:55:59

Fixed bug 4144 - CMake complains about trailing spaces in sdl2.pc Azamat H. Hackimov When you try use SDL2 2.0.8 in CMake project in Linux, it complains about trailing spaces in sdl2.pc: CMake Error at CMakeLists.txt:147 (add_executable): Target "TestSimpleMain" links to item "-L/usr/lib64 -lSDL2 " which has leading or trailing whitespace. This is now an error according to policy CMP0004.

a1b8fa60

2018-04-23T21:50:03

TryLockMutex: Fix error handling for TryLockMutex Christian Herzig pthread_mutex_trylock() and by the way, pthread_mutex_lock() do not set errno. Pthread-methods directly return error code as int. See related man-pages for details.

a9ae1b50

2018-04-23T20:24:12

Handle NULL return from SDL_DBus_GetContext()

b5d231ee

2018-04-23T19:20:12

Handle NULL return from SDL_DBus_GetContext()

43231256

2018-04-23T19:18:52

Added SDL_LinuxSetThreadPriority() to directly set the priority of a Linux thread (tid) This function tries using RealtimeKit connecting over DBUS as needed.

816a6e68

2018-04-23T17:10:36

Added support for adjusting thread priorities using Linux RealtimeKit Michael Sartain This is a quick pass at adding Linux RealtimeKit thread priority support to SDL. It allows me to bump the thread priority to high without root privileges or setting any caps, etc. rtkit readme here: http://git.0pointer.net/rtkit.git/tree/README

dc8b55e5

2018-04-16T02:11:09

coreaudio: Use the standard SDL audio thread instead of spinning a new one. Fixes corner cases, like the audio callback not firing if the device is disconnected, etc.

2df59062

2018-04-15T17:42:09

wayland: zxdg_shell_v6 needs a configure event before using a surface at all. Fixes Bugzilla #4109. Fixes Bugzilla #4119.

5e8c8167

2018-04-15T09:37:51

Fixed bug 4135 - Broken symlink libSDL2.so since rev11940 Tiago O. Symlink points to the wrong folder, and target will always have debug postfix, so it'll be broken for other build types.

b7228bc5

2018-04-11T18:28:03

Added support for the GameSir G3w

9856d967

2018-04-11T06:16:23

Fix the include path in the installed CMake target import file Previously the include path was {INSTALL_PREFIX}/include, it is now {INSTALL_PREFIX}/include/SDL2 to be consistent with the other build and package configuration systems. Fixes #4128.

6b5ed0fd

2018-04-10T08:03:54

Added debug postfix to install command and fixed library path

6a0ef0cd

2018-04-09T10:37:31

SDL: On Windows, have SDL_ShowWindow() not activate the window if the window has the WS_EX_NOACTIVATE window flag.

4d78a995

2018-03-26T12:38:29

Fixed bug where an SDL window that was activated while hidden could never be shown. Test code: { SDL_Window *win = SDL_CreateWindow( "Dummy", SDL_WINDOWPOS_CENTERED, SDL_WINDOWPOS_CENTERED, 128, 128, SDL_WINDOW_HIDDEN ); SDL_SysWMinfo info; SDL_VERSION( &info.version ); SDL_GetWindowWMInfo( win, &info ); SetActiveWindow( info.info.win.window ); { DWORD then = SDL_GetTicks(); while ( ( SDL_GetTicks() - then ) < 3000 ) { SDL_Event evt; SDL_PollEvent( &evt ); } SDL_ShowWindow( win ); then = SDL_GetTicks(); while ( ( SDL_GetTicks() - then ) < 3000 ) { SDL_Event evt; SDL_PollEvent( &evt ); } } SDL_DestroyWindow( win ); }

e14278ef

2018-03-24T10:26:40

Fixed bug 3804 - Message box on Windows truncates button ID Simon Hug I just wanted to fix a simple compiler warning in SDL_ShowMessageBox on Windows (which Sam fixed recently) and ended up finding some issues. Attached patch fixes these issues: - Because Windows only reports the lower 16 bits of the control identifier that was pushed, the button IDs used by SDL (C type int, most likely 32 bits) can get cut off. - The documentation states (somewhat ambiguously) that the button ID will be -1 if the dialog was closed, but the current code sets 0. For SDL 2.1, I think this should be a return code of SDL_ShowMessageBox itself. That will free up the button ID and it seems a more appropriate place for signaling this event. - Ampersands in controls will create mnemonics on Windows (underlined letters that, if combined with the Alt key, will push the button). I was thinking of adding a hint or flag to let the users enable it, but that might have unexpected results. - When the size of the text gets calculated, it doesn't use the same parameters as the static control. This can cut off text or wrap it weirdly. - On Windows, the Tab key is used to switch between control groups and sometimes between buttons in dialogs. This didn't seem to work correctly. Attached patch also adds: - Icons. Just the system ones that can be loaded with the ordinals IDI_ERROR, IDI_WARNING and IDI_INFORMATION. - A button limit of 2^16 - 101. - Some more specific error messages, but they never reach the user because how SDL_ShowMessageBox handles them if an implementation returns with an error.

b41b9d34

2018-03-19T14:52:53

Fixed return value

8e062f69

2018-03-19T14:42:51

Generalized the handling of instantaneous guide button presses so there's a minimum of 100 ms between guide button press and release. This happens with at least the following controllers: All Apple MFI controllers, ASUS Gamepad, XiaoMi Bluetooth Controller

d529b001

2018-03-19T13:16:11

Added mapping for the ASUS Gamepad removing the guide button, which doesn't generate events even though it's reported in the Android APIs.

f536fbea

2018-03-16T11:08:53

Reimplemented Android cursor API support using reflection so it builds with older SDKs

e20d4173

2018-03-15T18:22:48

Added Android custom cursor implementation This is commented out in SDLActivity.java, with the note #CURSORIMPLEENTATION because it requires API 24, which is higher than the minimum required SDK

1cfbe664

2018-03-12T18:41:06

Added Mac OpenGL ES configure support

d401a77e

2018-03-10T21:22:42

Fixed bug 4102 - define _WIN32_WINNT_WIN7 if not already defined Ozkan Sezer The following patch defines _WIN32_WINNT_WIN7 if it is not already defined in core/windows/SDL_windows.c, similar to what is already there for _WIN32_WINNT_VISTA.

6d39e67b

2018-03-10T21:20:25

Fixed bug 4103 - SDL_cpuinfo.h provokes warnings with -Wundef on non-x86 hosts Felix Geyer Forwarding from https://bugs.debian.org/892087 quoting verbatim: The SDL2 header SDL_cpuinfo.h generates gcc warnings if the program using it compiles with the -Wundef warning. (In particular, this means that QEMU builds using it fail on at least sparc hosts, since QEMU dev builds use both -Wundef and -Werror.). /usr/include/SDL2/SDL_cpuinfo.h:63:5: warning: "HAVE_IMMINTRIN_H" is not defined, evaluates to 0 [-Wundef] #if HAVE_IMMINTRIN_H && !defined(SDL_DISABLE_IMMINTRIN_H)

0a5c1065

2018-03-10T21:16:14

Fix ARM builds with MSVC

cc7b2fc5

2018-03-10T21:13:50

Temporary fix for bug 3432 - macOS 10.12: small scrolls (1 wheel notch) don't generate events Eric Wasylishen This bug was reintroduced by https://hg.libsdl.org/SDL/rev/fcf24b38a28a The steps to reproduce are the same: run the "testrelative" SDL demo with "--info all", connect a USB mouse with a scroll wheel, and roll the scroll wheel one "notch". You'll get log output like: testdraw2[1644:67222] INFO: SDL EVENT: Mouse: wheel scrolled 0 in x and 0 in y (reversed: 1) in window 1 As far as I can tell macOS doesn't have an API for getting the number of "wheel notches"; I get a deltaY of 0.100006 for one "notch", and it's heavily accelerated (if you roll the wheel quickly you'll get large deltas). So NSEvent's deltaY is only meant to be used for scrolling a scroll view, with the given distance in points, not something like selecting an item in a game. Here's a temporary patch that at restores the foor/ceil in Cocoa_HandleMouseWheel. Not ideal, but at least it restores the ability to scroll one notch of a mousewheel.

129431b4

2018-03-08T16:32:22

Delay delivery of the pause button release on MFI controllers so it doesn't happen in the same frame as the button press

92847022

2018-03-07T18:10:01

Added a mapping for the latest firmware for the Xbox One S controller on Android

a2c1d83c

2018-03-07T18:09:58

Include a USB VID/PID for Apple MFI controllers This is just placeholder VID/PID, but allows code that works with VID/PID to identify the MFI controllers easily.

a8ac5885

2018-03-07T13:30:40

Added SDL_GameControllerMappingForDeviceIndex() to get the mapping for a controller before it's opened

9e651b69

2018-03-06T14:51:50

Try to dynamically create a default Android game controller mapping based on the buttons and axes on the controller. Include the controller USB VID/PID in the GUID where possible, as we do on other platforms.

2419d267

2018-03-02T22:53:25

Progress fixing bug 4100 - errors and warnings after changeset 11917 Ozkan Sezer 2018-03-02 20:02:37 UTC http://hg.libsdl.org/SDL/rev/d702b0c54e52 resulted in an error and two warnings when compiled with mingw. 1. Error from SDL_windowstaskdialog.h: In file included from src/video/windows/SDL_windowsmessagebox.c:29:0: src/video/windows/SDL_windowstaskdialog.h:23:54: error: expected ')' before 'HWND' This is fixed by removing unnecessary annotations: 2. Warning from SDL_assert.c: src/SDL_assert.c: In function 'SDL_ExitProcess': src/SDL_assert.c:138:1: warning: 'noreturn' function does return Indeed ExitProcess() is prototyped with DECLSPEC_NORETURN, but TerminateProcess() is not. This can be rectified by adding an exit() call in there. Do NOTE, however, that requires building with a libc: 3. Warning from SDL_windowsmessagebox.c: src/video/windows/SDL_windowsmessagebox.c: In function 'WIN_ShowMessageBox': src/video/windows/SDL_windowsmessagebox.c:513:9: warning: 'nCancelButton' may be used uninitialized in this function My lazy solution was manually initializing nCancelButton to 0.

e9b29ed0

2018-03-02T22:48:15

Fixed bug 4101 - configure needs regenerating after changeset 11894

ac2d1f67

2018-03-02T12:08:18

Fixed setting the layer drawable size Without this change the drawable had a size of 0 and the metal renderer asserted because the projection matrix wasn't set.

cef1c1c2

2018-03-02T14:10:25

windows: Restore patches for Task Dialogs and TerminateProcess(). 2.0.8 has shipped, these can live in revision control now!

003c0dce

2018-03-02T10:56:21

Use the real controller name for game controllers on iOS and Apple TV

4f695bc5

2018-03-01T09:32:53

Added tag release-2.0.8 for changeset dfbae17e87ac

e9fc66a0

2018-03-01T08:26:10

Added patch note for Android Studio support

0df7fe84

2018-03-01T08:22:56

Temporarily disabled fullscreen switching code on Android, until we can resolve bug 4096 - Enabling fullscreen on Android causes the app to toggle fullscreen mode continuously in a loop

eac37666

2018-03-01T08:08:20

Added a patch note for OpenGL ES support on Mac

3537c3e7

2018-02-28T10:39:41

Back out Task Dialog and TerminateProcess patches for 2.0.8. These can return to revision control once we ship.

6a1cfcce

2018-02-28T02:14:15

windows: dos2unix'd messagebox code, and (hopefully) fixed on MingW.

a749035f

2018-02-28T01:54:22

windows: Message boxes use Task Dialogs if possible (thanks, Jack!). This lets the message box have an icon. Unless the app has opted-in to using the v6 common controls, though, this will fall back to the usual SDL message boxes.

8536130a

2018-02-28T01:23:49

assert: Use TerminateProcess() on Windows, vs ExitProcess (thanks, Jack!). "What I have done is use TerminateProcess rather than ExitProcess. ExitProcess will cause Microsoft's leak detection to continue, TerminateProcess won't. It is also technically wrong to use ExitProcess in the case of aborting the application. Jack Powell Twitter @jack9267"

99683f7c

2018-02-26T08:39:25

Added patch note about the Vulkan Mac SDK

be6ca785

2018-02-25T23:02:09

Support official Vulkan SDK for macOS. This tries to load vulkan.framework or libvulkan.1.dylib before MoltenVK.framework or libMoltenVK.dylib. In the previous version, layers would not work for applications run-time loading the default library.

f9f45d0b

2018-02-25T19:51:34

Fixed bug 4097 - Segmentation fault by SDL_CreateThreadWithStackSize Dongsun Kim Normal case 1. [thread 1] SDL_CreateThreadWithStackSize calls SDL_SYS_CreateThread. 2. [thread 1] If successful, it calls SDL_SemWait. 3. [thread 2] SDL_RunThread calls SDL_SYS_SetupThread, SDL_ThreadID, SDL_SemPost. 4. [thread 1] SDL_CreateThreadWithStackSize calls SDL_DestroySemaphore, SDL_free. Crash case (Segmentation fault) 1. [thread 1] SDL_CreateThreadWithStackSize calls SDL_SYS_CreateThread. 2. [thread 1] If successful, it calls SDL_SemWait. --> Error return due to SIGNAL(SYSTEM or Real Time) at sem_wait(pthread). 3. [thread 1] SDL_CreateThreadWithStackSize calls SDL_DestroySemaphore, SDL_free. 4. [thread 2] SDL_RunThread calls SDL_SYS_SetupThread, SDL_ThreadID, SDL_SemPost. --> Segmentation fault at strlen or sem_post.

5cbb90da

2018-02-25T19:22:47

cmake: Set debug library name suffix per target instead of setting it globally. Don't add a suffix on android

3ea093cc

2018-02-25T10:15:00

SDL_test_fuzzer.c: fix strict aliasing warnings by using a union.

cbfee82d

2018-02-24T09:09:20

Moved Android patch note into the Android section

6cf4d0e4

2018-02-24T08:59:58

Fixed bug 4092 - CMake support for building everything in the "test" directory Eric Wasylishen Patch to support building the tests with cmake. Disabled by default, use: "cmake .. -DSDL_TEST=YES" to enable the tests. Tested on macOS 10.13 with the ninja, makefile, and Xcode generators, and Windows 10 with the Visual Studio 2017 generator.

849d042f

2018-02-24T08:58:22

Fixed bug 4091 - Undefined references to Android audio functions when SDL_AUDIO_DISABLED is on Manuel Sabogal If SDL is compiled with the Audio subsystem disabled there are some undefined references to the functions ANDROIDAUDIO_ResumeDevices and ANDROIDAUDIO_PauseDevices in the file src/video/android/SDL_androidevents.c.

8e4afb55

2018-02-24T08:42:04

Added patch note for Ryan's work

b33dbd27

2018-02-24T08:40:30

Added missing file to tvOS build

99a0c0f0

2018-02-24T08:23:44

Fixed MinGW-w64 build

5adc1b7d

2018-02-24T08:18:09

Updated documentation with API changes in SDL 2.0.8

0626486e

2018-02-23T19:12:04

Backout the vulkan change in d449dea10fc8 breaks Mir builds

e17c3219

2018-02-23T11:24:26

mir: Disable Mir by default as Mir supports Wayland clients Also remove enabling VK support for Mir

8891f591

2018-02-21T22:53:52

Backed out changeset 6c8521d53507 Apparently this still triggers a compiler warning, have to dig further.

fd8f12d2

2018-02-21T22:35:17

android: apparently they fixed this header at some point.

ac309df7

2018-02-21T22:27:09

vulkan: Possibly fix a compiler warning (-Wstrict-prototypes).

ed64d54d

2018-02-21T21:36:10

windows: added WIN_IsWindows7OrGreater().

7e1fa0ce

2018-02-21T21:34:35

wasapi: fixed typo in an assert message.

c7e43665

2018-02-21T21:34:06

wasapi: let Windows do the resampling for us if possible.

69958441

2018-02-21T09:58:21

Fix high-dpi support on macOS and simplify it and iOS variant. The detault drawableSize for a CAMetalLayer is its bounds x its scale. So it is sufficient to set the *layer's* scale to the desired value.

6e3d0a13

2018-02-21T12:42:30

Use new XInput mapping for Win10+ (Bugzilla #3960)

a0687a9c

2018-02-21T09:40:47

Fixed bug 4034 - Don't include _DllMainCRTStartup() if SDL_STATIC_LIB is defined.

c00858bf

2018-02-20T09:04:31

Fixed bug 4088 - Fix Metal link errors with test programs in SDLTest.xcodeproj Eric Wasylishen The following patch adds Metal.framework to the "link binary with libraries" section of each test program, with "status" set to "optional", which fixes link errors on all of the test programs. I'm not sure if this is a correct fix - the fact that this was necessary might indicate the static SDL2.a library has a hard dependency on Metal.framework (?) - but it gets the test programs working in Xcode again. It also adds testyuv_cvt.c to the testoverlay2 target, fixing a link error.

58f9be12

2018-02-18T09:09:56

Actually, this is needed for building with Visual Studio with both /MT and /MD. With the previous change, I get: 1> Creating library C:\projects\SDL\VisualC\Win32\Debug\SDL2.lib and object C:\projects\SDL\VisualC\Win32\Debug\SDL2.exp 1>LINK : error LNK2001: unresolved external symbol __DllMainCRTStartup@12

3c9d3336

2018-02-18T08:57:01

Fixed bug 4034 - Do we really need _DllMainCRTStartup() in every Windows build? Andreas Falkenhahn In src/SDL.c there is this code: _DllMainCRTStartup(HANDLE hModule, ... The comment says that this is needed on Watcom C for some reason but why is it included then when building with Visual C as well? Shouldn't it be only included when compiling on Watcom C then? I'm asking because this code caused me a lot of headaches because I'm building a DLL that contains SDL and I link using /MT and the _DllMainCRTStartup() symbol obviously led to lots of trouble but it wasn't clear to me where the problem was because all I got from the linker was: LNK2019: unresolved external symbol _main referenced in function ___tmainCRTStartup So I had to got through each and every object to see what the culprit was. See here for the full story: https://stackoverflow.com/questions/25067151/lnk2019-unresolved-external-symbol-main-referenced-in-function-tmaincrtstar/48177067#48177067 So if it isn't necessary on Visual C, please just leave that symbol out on Visual C so that it no longer leads to any trouble. Thanks.

75a58303

2018-02-17T23:57:57

pthread: fix error code checks (thanks, Andreas!). Most pthread functions return 0 on success and non-zero on error, but those errors might be positive or negative, so checking for return values in the Unix style, where errors are less than zero, is a bug. Fixes Bugzilla #4039.

2ea4419a

2018-02-17T20:18:48

yuv: patched to compile.

7c0c2c22

2018-02-17T20:10:13

yuv: fixed variable declaration shadowing warnings. Fixes Bugzilla #4062.

97494f53

2018-02-17T18:30:21

pulseaudio: Just read/dump captured data in FlushCapture. Apparently pa_stream_flush() doesn't work as expected: https://lists.freedesktop.org/archives/pulseaudio-discuss/2012-April/013328.html Fixes Bugzilla #4087.

6867f618

2018-02-16T14:56:28

video: put a spinlock around a global linked list. This should only contend if you're allocating or freeing surfaces from multiple threads at once, and then just for a short time. Fixes Bugzilla #4084.

8ddebfa0

2018-02-16T10:23:10

Fixed bug 4085 - X11: Allow configuring _NET_WM_BYPASS_COMPOSITOR through SDL hints Callum McGing This patch allows the user to disable the behaviour that blocks the compositor through a new hint: SDL_VIDEO_X11_NET_WM_BYPASS_COMPOSITOR. This allows tools or other windowed applications to behave properly under KWin.

kc3-lang/SDL

Log