Commit Graph

8498 Commits

Author SHA1 Message Date
Sam Lantinga 88e9f7765c Fixed relative mouse mode using warping after 82793ac279 2021-10-14 18:37:27 -07:00
Sam Lantinga a1fabca162 Removed mouse warping for local mice and improved warp handling for mouse over RDP 2021-10-14 16:52:21 -07:00
Sam Lantinga 82793ac279 Fixed mouse warping while in relative mode
We should get a mouse event with an absolute position and no relative motion and shouldn't change the OS cursor position at all
2021-10-14 14:26:21 -07:00
Sylvain 072e3fdfc4
Fixed bug #4534: NEON implementation of Convert51ToStereo (Thanks Ryan!) 2021-10-14 23:17:08 +02:00
Sam Lantinga 5e89b3c89e Don't need to use raw input to track the mouse during mouse capture (thanks Brick!) 2021-10-14 11:46:07 -07:00
Brick 0b6a821188 Messages posted on the same tick are not new 2021-10-14 09:05:38 -07:00
Rémy Tassoux 0789610cfb Add SDL_HINT_WINDOW_NO_ACTIVATION_WHEN_SHOWN 2021-10-14 08:39:48 -07:00
Sam Lantinga 227c33b8ec Add note about SDL_TICKS_PASSED hitting int overflow problems with large timeouts 2021-10-13 09:33:57 -07:00
Sam Lantinga 1fa154fda3 Fix weak enforcement of timeouts in SDL_WaitEventTimeout_Device. This will loop pumping events and waiting for a system event to come in. However not all system events will turn into an SDL event. It's not unusual for a Windows message to be some internal thing that SDL doesn't convert into a message. In that case the loop will simple circle but not exit. As long as such messages are coming in the loop will continue to run regardless of the timeout. When messages finally stop it'll still wait for the full timeout so you can have arbitrarily long delays.
Instead do an absolute elapsed time check since the start of the wait.  If that is exceeded during any iteration the routine exits as the timeout has elapsed.
2021-10-13 09:33:54 -07:00
Sam Lantinga 2423c51471 Work around hang in AAudioStream_write() during extended shared object loading while running in a debugger. Observed on a OnePlus 8T (KB2005) running Oxygen OS 11.0.10.10.KB05AA.
The observed behavior is that any nonzero timeout value would hang until the device was paused and resumed. And a zero timeout value would always return 0 frames written even when audio fragments could be heard. Making a manual timeout system unworkable.
None of the straightforward systems imply that there's a detectable problem before the call to AAudioStream_write(). And the callback set within AAudioStreamBuilder_setErrorCallback() does not get called as we enter the hang state.
I've found that AAudioStream_getTimestamp() will report an error state from another thread. So this change codifies that behavior a bit until a better fix or more root cause can be found.
2021-10-13 09:33:51 -07:00
Sylvain 325ae5c35d
Android: getCurrentOrientation uses getContext() to retrieve the activity (see bug #4825) 2021-10-13 08:35:56 +02:00
Erik Soma 0e294e90ae
Ensure that SDL_InitSubSystem quits subsystems after an error. (#4834)
* Ensure that SDL_InitSubSystem quits subsystems after an error.

* Fix unnecessary change.
2021-10-12 15:55:31 -07:00
Ozkan Sezer 6149e60136 remove IsWin10FCUorNewer() -- not used since commit 40ed9f75c9 2021-10-13 00:03:56 +03:00
Ozkan Sezer 311671a029 fixed build after commit 6e356e20ad 2021-10-12 23:50:02 +03:00
Rémy Tassoux 6e356e20ad Fix mouse focus being set to null when a captured mouse cursor leaves the window. 2021-10-12 13:30:46 -07:00
Cameron Cawley 9ee6942e79 Improve RISC OS implementations of SDL_GetBasePath and SDL_GetPrefPath 2021-10-12 13:07:52 -07:00
Cameron Cawley 3db898c5b6 riscos: Implement SDL_GetPrefPath() 2021-10-12 13:07:52 -07:00
Sylvain 723f5cea43
Fixed bug #4825 - NullPointerException on SDLActivity.getCurrentOrientation() (Thanks Benau!) 2021-10-12 20:58:58 +02:00
Ryan C. Gordon 1b49f09243
include: manually ran wikiheaders.pl and cleaned up the obvious issues. 2021-10-08 20:50:30 -04:00
Ryan C. Gordon 55f60847cb
wikiheaders.pl: Try to handle functions with multiple declarations better. 2021-10-08 20:50:30 -04:00
Ryan C. Gordon 3970a885a5
wikiheaders.pl: Migrate in docs that exist only in the wiki.
This is going to expose some docs that needs cleaning up once
the bridge imports those docs, though.
2021-10-08 15:58:02 -04:00
Ryan C. Gordon 194657a2c6
include: Fix indentation on SDL_SoftStretch doxygen comment.
Otherwise wikiheaders.pl doesn't think it's valid.
2021-10-08 15:58:01 -04:00
Sam Lantinga 40ed9f75c9 Workaround for Windows occasionally ignoring SetCursorPos() calls
Also, since we're flushing mouse motion before and including the warp, we don't need the isWin10FCUorNewer hack to simulate mouse warp motion.

Fixes https://github.com/libsdl-org/SDL/issues/4339 and https://github.com/libsdl-org/SDL/issues/4165
2021-10-08 10:05:27 -07:00
Sam Lantinga 373216ae5b Added support for touchpads in the Linux evdev code (thanks Francisco!) 2021-10-07 18:14:16 -07:00
Sam Lantinga 16aeb8d0f5 Guarantee that we don't dispatch any mouse motion from before or including the last mouse warp 2021-10-07 15:04:06 -07:00
Sam Lantinga 649466f491 Flush any pending mouse motion when warping the mouse cursor
Fixes https://github.com/libsdl-org/SDL/issues/4165
2021-10-07 13:28:44 -07:00
Väinö Mäkelä ad5205739e wayland: Only dispatch frame events in Wayland_GLES_SwapWindow
Dispatching all events in Wayland_GLES_SwapWindow leads to resizes being
acked before the program has a chance to handle the resize. This change
reduces jumping on fullscreen transition with apps that call
SDL_PollEvent before issuing any render calls.
2021-10-07 05:12:00 -04:00
Brick f30e2e438d Fix compiler warnings in testvulkan.c
MSVC gives the following warnings when compiling testvulkan.c:
```
testvulkan.c(656,1): warning C4018: '<': signed/unsigned mismatch
testvulkan.c(656,1): warning C4018: '>': signed/unsigned mismatch
testvulkan.c(660,1): warning C4018: '<': signed/unsigned mismatch
testvulkan.c(660,1): warning C4018: '>': signed/unsigned mismatch
```
2021-10-06 15:24:41 -07:00
Sam Lantinga dd95c9c8a2 Moved focus click check into WIN_UpdateFocus() so we have the correct state when setting keyboard focus
Fixes https://github.com/libsdl-org/SDL/issues/4817
2021-10-06 09:09:39 -07:00
Sam Lantinga 1ec409c214 Don't warp the mouse within a window while it's minimized 2021-10-06 09:09:09 -07:00
Daniel Gibson 87ac373493 Fix Debug-only builds with CMake
"get_filename_component(SDL2_LIBDIR ${sdl2implib} PATH)" failed if
sdl2implib was set from sdl2implibdbg instead of ${sdl2implibdbg}

Thanks to seyuup from the SDL Discourse for pointing this out.
2021-10-05 18:45:53 -07:00
DomGries 0d541e5a88 Revert "Fixed relative mode mouse events stopping if you click on the title bar"
This has been better fixed by b28ed02 or another related relative mouse mode change of @slouken in SDL 2.0.17 and as such can be reverted to reduce unneeded processing in WM_MOUSEMOVE
2021-10-05 15:46:12 -07:00
Sam Lantinga c542de9213 React to WM_NCACTIVATE instead of WM_SETFOCUS or WM_ACTIVATE for focus changes
See https://github.com/libsdl-org/SDL/pull/4293 and https://github.com/libsdl-org/SDL/issues/4450 for details
2021-10-05 14:09:52 -07:00
Ryan C. Gordon ab7910facf
cmake: Move the old, global add_definitions (etc) to an interface library.
Fixes #4150.
2021-10-05 12:34:29 -04:00
Ryan C. Gordon f1793af404
cmake: removed an add_definitions call. 2021-10-05 11:19:30 -04:00
Ryan C. Gordon 25fdba39bb
cmake: fix indentation. 2021-10-05 11:19:14 -04:00
Ryan C. Gordon f306662544
cmake: Correctly set IOS variable.
Fixes #4635.
2021-10-05 11:13:41 -04:00
Ryan C. Gordon 60d59aeb61
cmake: Calculate padding for dislaying options automatically. 2021-10-05 10:54:06 -04:00
Ryan C. Gordon 7850d0cf6f
cmake: Prefix all options with "SDL_"
This makes it clear which options came from SDL's CMake project when
building SDL as a subdirectory of a parent CMake project.

Fixes #4139.
2021-10-05 10:45:26 -04:00
Sylvain 172957387e
PSP: links for setting up, compiling and running on PSP 2021-10-04 22:12:08 +02:00
Sylvain f080273acd
PSP: add implementation for RenderGeometry 2021-10-04 22:01:11 +02:00
Ozkan Sezer e92a639b23 replaced many uses of libc calls with SDL_ counterparts in os2 sources.
TODO: core/os2 and geniconv sources _must_ be updated, as well..
2021-10-04 23:00:28 +03:00
Sylvain dd9b2daf8d
PSP: fixed build 2021-10-04 21:56:42 +02:00
Ozkan Sezer 3ea35fe5bb fix SDL_atoi()
fixes https://github.com/libsdl-org/SDL/issues/4811
2021-10-04 21:32:00 +03:00
Ivan Epifanov 407867cf72 Alias SDL2::SDL2 to SDL2::SDL2-static on static-only builds 2021-10-04 11:17:30 -07:00
Ozkan Sezer cd7ef24cba minor update to os2 makefile. 2021-10-04 00:32:56 +03:00
Sam Lantinga c501d159a9 Fixed range for Steam Controller sensor values 2021-10-03 14:10:21 -07:00
Sam Lantinga ea1a2b94f2 Use the correct update rate for Steam Controller sensors 2021-10-02 09:13:29 -07:00
David Gow eadc8f9355 wayland: Cleanup some SDL_TryLockMutex() calls.
Check the result of these against 0 explicitly, so that it's obvious
we're bailing out on failure, not success.
2021-10-02 11:14:52 -04:00
David Gow 25f9e32b0e wayland: Don't let multiple threads dispatch wayland events at once
wl_display_dispatch() will block if there are no events available, and
while we try to avoid this by using SDL_IOReady() to verify there are
events before calling it, there is a race condition between
SDL_IOReady() and wl_display_dispatch() if multiple threads are
involved.

This is made more likely by the fact that SDL_GL_SwapWindow() calls
wl_display_dispatch() if vsync is enabled, in order to wait for frame
events. Therefore any program which pumps events on a different thread
from SDL_GL_SwapWindow() could end up blocking in one or other of them
until another event arrives.

This change fixes this by wrapping wl_display_dispatch() in a new mutex,
which ensures only one thread can compete for wayland events at a time,
and hence the SDL_IOReady() check should successfully prevent either
from blocking.
2021-10-02 11:03:20 -04:00