This simplifies some things, clarifies some things, and also allows
for the possibility of RWops that offer non-blocking i/o (although
none of the current built-in ones do, intentionally, we could add this
later if we choose, or people could provide things like network socket
RWops implementations now, etc.
Fixes#6729.
On window focus, look up the pressed modifier keys via keycodes instead of scancodes to handle the case of remapped keys, as xkb remapping changes the associated keycode, not the scancode.
The annotations have been added to SDL_mutex.h and have been made public so applications can enable this for their own code.
Clang assumes that locking and unlocking can't fail, but SDL has the concept of a NULL mutex, so the mutex functions have been changed not to report errors if a mutex hasn't been initialized. We do have mutexes that might be accessed when they are NULL, notably in the event system, so this is an important change.
This commit cleans up a bunch of rare race conditions in the joystick and game controller code so now everything should be completely protected by the joystick lock.
To test this, change the compiler to "clang -Wthread-safety -Werror=thread-safety -DSDL_THREAD_SAFETY_ANALYSIS"
It was previously thought that these function calls were unnecessary as the initial bug and reproduction case that necessitated their addition seemed to be fixed, but apparently there are still cases where this needs to be set explicitly. Set the xdg surface geometry at creation time and when the window size changes.
Partially reverts #6361. This is not needed in the libdecor path, as libdecor calls this for the content surface internally.
* N3DS: Make Sem waits cooperative friendly.
The 3DS has a cooperative threading model. Sleeping after TryWait and
WaitTimeout avoid starving other threads. It inccurs a runtime penalty,
but it's better than having to hard reset your console to recover from
a deadlock.
Num Lock and Caps Lock always need to be explicitly handled by the modifier handler function, or they won't be correctly set if active at application startup, or if the lock state is changed while the application lacks focus since a key press for these keys will never be received. In these cases, the internal SDL modifier state can end up the inverse of the actual modifier state.
This reverts commit 39eab4bf44.
From @Kontrabant:
Making windows use the window title as the app ID by default doesn't seem like the right direction. The app ID is used by window managers to group windows from the same application together for switching between with alt-tilde and such functionality, and giving each window it's own app ID would break this.
Add the protocol for high-resolution timestamp events and subscribe to them if available.
Event timestamps are now handled natively in nanoseconds as much as possible to avoid error-prone conversions.
Variables have been appended with _ms or _ns where appropriate, to avoid ambiguity.
If you care about timestamps you'll also want to catch all of the sensor events instead of just polling the current state. For example, Nintendo Switch controllers generate 3 sensor events with distinct values for each polling interval.
If a program built against one version of SDL is run in an
environment where there is an earlier version of the SDL .so library
installed, the result varies depending on platform configuration; in
the best case, it won't start at all, at worst it aborts in the
middle of the user doing "something important" (systems implementing
lazy symbol resolution). verdefs on the other hand are always checked
on startup.
The dependency information present in programs and shared libraries
is not only of value to the dynamic linker, but also to a
distribution's package management. If the dynamic linker is able to
tell that a program is not runnable per the above, a package manager
is able to come to the same conclusion — and block the installation
of a nonfunctional program+library ensemble.
Because there are a lot more symbols than there are libraries (I am
going to throw in "10^4 to 1 or worse"), package managers generally
do not evaluate symbols, but only e.g. the SONAME, NEEDED and VERNEED
fields/blocks. Because the SONAME is the same between two SDL
versions like 2.0.24, and 2.0.26, everything rests on having verdefs.
This patch proposes the addition of verdefs.
In SDL3 we plan to make more use of shaders in the 2D render API, and this minimizes the number of platforms we have to consider for new features. OpenGL ES 2.0 or newer is supported on all modern iOS and Android devices.
I updated .clang-format and ran clang-format 14 over the src and test directories to standardize the code base.
In general I let clang-format have it's way, and added markup to prevent formatting of code that would break or be completely unreadable if formatted.
The script I ran for the src directory is added as build-scripts/clang-format-src.sh
This fixes:
#6592#6593#6594
The new protocol adds support for more native communication of
fractional scaling.
Everything in the wayland backend already existed only our fractional
scale was calculated implicitly through a combination of output size
guesswork for fullscreen windows.
This new protocol makes that explicit, providing a more robust solution
and a solution for non-fullscreen surfaces. The fallback code is still
left in place for now whilst compositors gain support.
- do not overwrite error message set by SDL_InitFormat (SDL_AllocFormat)
- set proper error message (Cocoa_Metal_CreateView)
- protect against allocation failure (UIKit_Metal_CreateView)
* Remove depth field from SDL_CreateRGBSurfaceWithFormat and SDL_CreateRGBSurfaceWithFormatFrom
* Removed unused 'flags' parameter from SDL_CreateRGBSurface and SDL_CreateRGBSurfaceWithFormat
* Removed unused 'flags' parameter from SDL_ConvertSurface and SDL_ConvertSurfaceFormat
This behavior more closely matches other platform where the window is borderless but still interacts with the window manager (e.g. task bar shows above it, it can be resized to fit within usable desktop area, etc.)
If you want a borderless window the size of the desktop that looks like a fullscreen window, then you should use the SDL_WINDOW_FULLSCREEN_DESKTOP flag.
Fixes https://github.com/libsdl-org/SDL/issues/6659
`argv[0]`/`getexexname()` are not always absolute paths by default and can be modified to anything the developer wants them to be.
Consider using `readSymLink("/proc/self/path/a.out")` instead and `getexecname()` as the fallback, since the symlink will always be the correct absolute path (unless /proc is ot mounted, but it is by default on Solaris and Illumos platforms).
* Add braces after if conditions
* More add braces after if conditions
* Add braces after while() conditions
* Fix compilation because of macro being modified
* Add braces to for loop
* Add braces after if/goto
* Move comments up
* Remove extra () in the 'return ...;' statements
* More remove extra () in the 'return ...;' statements
* More remove extra () in the 'return ...;' statements after merge
* Fix inconsistent patterns are xxx == NULL vs !xxx
* More "{}" for "if() break;" and "if() continue;"
* More "{}" after if() short statement
* More "{}" after "if () return;" statement
* More fix inconsistent patterns are xxx == NULL vs !xxx
* Revert some modificaion on SDL_RLEaccel.c
* SDL_RLEaccel: no short statement
* Cleanup 'if' where the bracket is in a new line
* Cleanup 'while' where the bracket is in a new line
* Cleanup 'for' where the bracket is in a new line
* Cleanup 'else' where the bracket is in a new line
I ran this script in the include directory:
```sh
sed -i '' -e 's,#include "\(SDL.*\)",#include <SDL3/\1>,' *.h
```
I ran this script in the src directory:
```sh
for i in ../include/SDL3/SDL*.h
do hdr=$(basename $i)
if [ x"$(echo $hdr | egrep 'SDL_main|SDL_name|SDL_test|SDL_syswm|SDL_opengl|SDL_egl|SDL_vulkan')" != x ]; then
find . -type f -exec sed -i '' -e 's,#include "\('$hdr'\)",#include <SDL3/\1>,' {} \;
else
find . -type f -exec sed -i '' -e '/#include "'$hdr'"/d' {} \;
fi
done
```
Fixes https://github.com/libsdl-org/SDL/issues/6575
SDL_HINT_VIDEO_X11_FORCE_EGL was deprecated in favor of the more general SDL_HINT_VIDEO_FORCE_EGL, and Xinerama and Xvidmode support was previously removed from SDL, rendering their associated hints nonfunctional.
Explicitly include the Wayland protocol headers when statically linking against the Wayland libraries or older system headers might be used instead of the local versions.
* The header is no longer dependent on SDL build configuration
* The structures are versioned separately from the rest of SDL
* SDL_GetWindowWMInfo() now returns a standard result code and is passed the version expected by the application
* Updated WhatsNew.txt and docs/README-migration.md with the first API changes in SDL 3.0
We don't want the G29 to show up as a gamepad, Steam will create a virtual Xbox controller for it, which breaks racing games.
(cherry picked from commit dce6ed56d7)
This reverts commit 33a68f575f.
We don't want the G29 to show up as a gamepad, Steam will create a virtual Xbox controller for it, which breaks racing games.
(cherry picked from commit e3d430b83e)
When changing the window state from non-floating to floating (e.g. leaving fullscreen), libdecor can send bogus content sizes that are +/- the height of the window title bar and start 'walking' the window height in one direction or the other with every transition.
The floating window size is known, so use the cached value instead of the size reported by libdecor when restoring the floating state.
Some applications (and embarrassingly, testime is one of them) call
SDL_StartTextInput() or SDL_SetTextInputRect() every frame. On KDE/KWin
with fcitx5, this causes there to be several preedit events every frame
(particularly given some of the workarounds in Wayland_StartTextInput),
which slows testime down to an unusable crawl.
Instead, make SDL_StartTextInput() a no-op if text input is already
enabled, and cache the input rect, only changing it when the new rect is
actually different.
With these changes, we only get preedit events (and hence
SDL_TEXTEDITING events) when the preedit string actually changes. This
matches the behaviour under XWayland, and works very smoothly.
This needs to check what our deployment target is, not what SDK
is available, since this is a linker symbol and not an enum
value or whatever.
Also removed a copy/paste error that mentioned CoreAudio in
the haptic subsystem.
Fixes#6534.
The XKB_KEY_* and XK_* macros resolve to the same constant values, so use the raw values and note what keys they correspond to in the comments, as is done for the other keysym values in this file.
This completely eliminates the need for any X or XKB system headers along with the if/else defines.
Most of these are probably harmless, but the changes to SDL_immdevice.c and SDL_pixels.c appear to have fixed genuine bugs.
SDL_audiocvt.c: By separating the calculation of the divisor, I got rid of the suspicion that dividing a double by an integer led to loss of precision.
SDL_immdevice.c: Added a missing test, one that could have otherwise led to dereferencing a null pointer.
SDL_events.c, SDL_gamecontroller.c, SDL_joystick.c, SDL_malloc.c, SDL_video.c: Made it clear the return values weren't used.
SDL_hidapi_shield.c: The size is zero, so nothing bad would have happened, but the SDL_memset() was still being given an address outside of the array's range.
SDL_dinputjoystick.c: Initialize local data, just in case IDirectInputDevice8_GetProperty() isn't guaranteed to write to it.
SDL_render_sw.c: drawstate.viewport could be null (as seen on line 691).
SDL.c: SDL_MostSignificantBitIndex32() could return -1, though I don't know if you want to cope with that (what I did) or SDL_assert() that it can't happen.
SDL_hints.c: Replaced boolean tests on pointer values with comparisons to NULL.
SDL_pixels.c: Looks like the switch is genuinely missing a break!
SDL_rect_impl.h: The MacOS static checker pointed out issues with the X comparisons that were handled by assertions; I added assertions for the Y comparisons.
SDL_yuv.c, SDL_windowskeyboard.c, SDL_windowswindow.c: Checked error-result returns.
This reverts commit 04b50f6c6b.
It turns out OpenGL vsync has broken again in macOS 12, so
we're reintroducing our CVDisplayLink code to deal with it,
again.
Reference Issue #4918.
SDL is built around the concept of keyboards having a fixed layout with scancodes that correspond to physical keys no matter what linguistic layout is used. Virtual keyboards don't have this concept and can present an arbitrary layout of keys with arbitrary scancodes and names, which don't fit the SDL model. When one of these keyboards is encountered, it requires special handling: use the keysym of the pressed keys to derive their ANSI keyboard scancode equivalents for control keys and ASCII characters. All other characters are passed through as text events only.
Add a helper function to get the keycode for a scancode from the default lookup table. Unlike SDL_GetKeyFromScancode(), this is not affected by the set keymap.
It turns out that we can safely create a Metal view on an existing window, and that avoids issues with the window being recreated with the wrong orientation in iOS 16.
Fixes https://github.com/libsdl-org/SDL/issues/6289
Clamp the wl_seat max version to 5 if being built against a version of libwayland below 1.21.0, or containers that bundle newer versions of SDL with older versions of libwayland can break if the compositor advertises support for a protocol version above 5.
Depending on the underlying EGL library, it may be desirable to
conditionally set some specific EGL attributes depending on available
extensions and other application state.
SDL's EGL usage makes this a little bit complicated because:
- there are multiple functions used to set up a working EGL context
- some of these functions take different types of EGL attributes
(EGLAttrib vs EGLint)
- the EGL extension list before creating an EGLDisplay differs from the
extension list after (i.e. display vs client extensions)
- all of the above happens in a single SDL_CreateWindow call
This leaves no place for the application to discover what EGL extensions
are available and provide custom attribute lists.
Until now, if a developer wants to add a custom EGL attribute for
eglGetPlatformDisplay, eglCreateWindowSurface or eglCreateContext, they
needed to patch SDL itself. This is very undesirable, since such
developers would have to disable the SDL dynapi in order to maintain
compatibility with their needs.
This patch implements some callbacks which developers can use to
dynamically generate custom EGL attributes for SDL to use during
SDL_CreateWindow.
These functions allow applications to call EGL functions against the SDL
EGL context. For example, applications can use an EGL API loader via
SDL_EGL_GetCurrentDisplay and SDL_EGL_GetProcAddress, and can call
functions such as eglQuerySurface against the internal EGLSurface and
EGLDisplay.
The EGL 1.5 specification says that eglGetPlatformDisplay (unlike
eglGetPlatformDisplayExt) takes an EGLAttrib* rather than EGLint* for
the attribute list.
This fallback is undesirable when using ANGLE, because it will end up
using some default configuration (e.g. on Windows it defaults to the
D3D11 backend).
This implements a new SDL_GL_EGL_PLATFORM attribute to set the
"platform" argument for SDL_EGL_LoadLibrary on Windows, macOS, and
Linux. I've limited it to those three operating systems because that's
what I've been able to test against.
This adds support for forcing the use of EGL on Windows and MacOS. The
SDL_HINT_VIDEO_X11_FORCE_EGL hint is retained for backwards
compatibility but is now deprecated.
- use _Interlocked(Compare)ExchangePointer in case of _M_IX86 as well
- improve assertions:
1. add assertions to SDL_AtomicAdd/SDL_AtomicSet and SDL_AtomicCAS
2. use sizeof(a->value) instead of sizeof(int)
Sending rumble to the Amazon Luna controller on macOS gets there, but IOHIDDeviceSetReport() blocks for a long time and eventually fails.
This appears to be a bug in the macOS Bluetooth stack, ref rdar://99265496
- SDL_EventQ.active is a bool variable -> do not use SDL_AtomicGet/Set, it does not help in any way
- protect SDL_EventQ.active with SDL_EventQ.lock
- set SDL_EventQ.active to FALSE by default
This is a USB adapter for controllers shipped with Nintendo's NES-mini and
SNES-mini consoles.
Tested with both NES and SNES controllers, buttons map as expected on both.
Most of this code is disabled out for now.
- For mouse cursors we have a wl_surface for both system and custom
cursors which needs recreating.
- The other patch is about nullification after deletions
This works around udev event nodes arriving before hidraw nodes and the controller being opened twice - once using the Linux driver and once by the HIDAPI driver.
This also fixes a kernel panic on Steam Link hardware due to trying to open the hidraw device node too early.
A delay of 10 ms seems to be a good value, tested on Steam Link hardware.
If relative mouse mode is explicitly enabled, don't modify the capture flag on button events or the window might report having lost mouse focus if a button is pressed while moving the cursor.
This is a functional state for some window managers (tested using stock Ubuntu 22.04.1), and removing that state, e.g. using SDL_RestoreWindow(), results in a window centered and floating, and not visually covering the rest of the desktop.
This uses a newer browser API to get physical scancodes, but still
uses the (deprecated) event field that we were already using for
scancodes, but for keycodes instead now, which appears to be more
accurate.
Since keyboard layout isn't (generally) available to web apps, this
adds an internal interface to send key events with both scancode
and keycode to SDL's internals, instead of sending just scancodes and
expecting SDL to use its own keymap to generate keycodes.
Future work in this area would be to use the keyboard layout APIs
on browsers that support them, which would allow us to use SDL's
usual keymap code and not rely on a deprecated browser API, but
until we get there, this patch gives significantly more correct
results than we would have before.
Fixes#2098.
Exit the fullscreen sequence sooner if it is requested that a popup window be fullscreen.
The surface commit formerly in this path is irrelevant and can be removed as previous changes made it so that SetFullscreen() is no longer called from anywhere except Wayland_SetWindowFullscreen().
The controller can use either hat or buttons for the D-Pad, depending on what Linux driver is in use. The automatic mapping in LINUX_JoystickGetGamepadMapping() will do the right thing based on the exposed capability bits.
I'm sure this is the case for other controllers as well, so we might be removing more mappings over time.
Clear the window to black on the initial window draw, to avoid a really
obnoxious white flash. This doesn't always eliminate it, but it
definitely minimizes it.
This makes the colorspace match across different graphics APIs. By
default, OpenGL was getting a much more saturated colorspace (maybe
Display P3?) and it was looking very different from the rendering done
by Metal or MoltenVK.
- SDL_LoadObject: upon failure, strip the .dll extension and retry,
but only if module name has no path.
- SDL_LoadFunction: upon failure, retry with an underscore prepended,
e.g. for gcc-built dlls.
- strlcpy was passed a wrong buffer length parameter. has worked so
far by luck.
- use memcpy instead of strlcpy for simplicity.
- 'append' has been a typo: should be 'prepend'.
Otherwise the thread might block for a long time (more than 10 seconds!).
It's not clear to me why this happens, or why its safe to do this with a
resource that's still in use, but we have, until recently, always
disposed of the AudioQueue first, so changing back is probably okay.
Also changed the disposal to allow in-flight buffers to reach hardware;
otherwise you lose the last little bit of audio that's already been queued
but not played, which you can hear clearly in the loopwave test program.
Fixes#6377.
