- Add support for dispatching input on the render thread (UI thread is the current default) when `input_buffering` and `accumulated_input` are disabled. At the expense of latency, this helps prevent 'heavy' applications / games from blocking the UI thread (the default behavior) which may cause the application to ANR.
- Remove GLSurfaceView logic causing the UI thread to wait on the GL thread during lifecycle events. The removed logic would cause the UI thread to ANR when the GL thread is blocked.
As discussed with upstream, the C/C++ standard library is always fully
included when building with MAIN_MODULE=1, so using EMCC_FORCE_STDLIBS
is not necessary in our case.
G-SYNC (NVIDIA's VRR) is known to be buggy on windowed mode in Windows.
While the driver only enables G-SYNC for full screen mode by default,
users can toggle it on for windowed mode too, resulting in unstable
refresh rates during Editor usage.
This patch extends Godot's NVIDIA profile to force the default full
screen mode only G-SYNC with Godot.
If window was maximized before minimizing, restoring the window
afterwards using SW_RESTORE seems to cause the window to forget its
original size and position before being maximized, causing the window to
fill the screen instead.
Before of this patch, as explained in the usual
commented-wall-of-text-longer-than-the-actual-patch-itself™, due to the
multithreaded nature of the Wayland thread, it was possible to commit a
surface while the renderer was doing stuff, which was _very_ wrong.
Initially the consequences of such a sin weren't obvious but, now that
explicit synchronization is becoming more and more common, we can't
commit a buffer randomly without basically guaranteeing a nasty, nasty
crash (and we should have avoided commits altogether in the first place
to ensure atomic surface updates).
We now only trigger a commit _in the main thread_ when low processor usage
mode is on _and_ if we know that we won't be rendering anything as, due to
its intermittent nature, it makes "legacy" (pre xdg_wm_base v6) frame
callback based suspension quite annoying.
- Make warnings print only once per session.
- Tweak the message to be less confusing, and mention that the issue
most likely stems from a graphics driver limitation.
Instead of calling loadDynamicLibraries ourselves, we add the
GDExtension libraries to preload to the "dynamicLibraries" module config
property.
This seems to fix some threading issue with some browsers during the
init phase.
SUPPORT_LONGJMP have changed since emscripten 3.1.32 to default to
"wasm" mode when exceptions are enabled, and "emscripten" mode when
disabled.
While we generally doesn't use exception in core, linked libraries may
need them, and emscripten don't plan to support WASM EH + Emscripten
SjLj in the long term.
