- Implements new `KeyValuePairs` and `KeyValuePairAt` internal calls
to get the `key` and the `value` in one call.
- Caches the `DictionaryEntry` to reuse properties without repeating
internal calls.
(cherry picked from commit 2deefd938f)
For the time being we don't support writing a description for those, preferring
having all details in the method's description.
Using self-closing tags saves half the lines, and prevents contributors from
thinking that they should write the argument or return documentation there.
(cherry picked from commit 7adf4cc9b5)
Implemented some basic caching to avoid unnecessary AOT compilation
of unchanged assemblies that were already compiled previously.
This reduces iOS export times considerably for subsequent builds
since many dependencies never change, such as framework assemblies
and the Godot bindings.
The AOT compiler asm output and object files are now placed in
`res://.mono/temp/obj/<CONFIG>/godot-aot-cache/` instead of a
temporary directory.
Implement Octahedral Compression for normal/tangent vectors
*Oct32 for uncompressed vectors
*Oct16 for compressed vectors
Reduces vertex size for each attribute by
*Uncompressed: 12 bytes, vec4<float32> -> vec2<unorm16>
*Compressed: 2 bytes, vec4<unorm8> -> vec2<unorm8>
Binormal sign is encoded in the y coordinate of the encoded tangent
Added conversion functions to go from octahedral mapping to cartesian
for normal and tangent vectors
sprite_3d and soft_body meshes write to their vertex buffer memory
directly and need to convert their normals and tangents to the new oct
format before writing
Created a new mesh flag to specify whether a mesh is using octahedral
compression or not
Updated documentation to discuss new flag/defaults
Created shader flags to specify whether octahedral or cartesian vectors
are being used
Updated importers to use octahedral representation as the default format
for importing meshes
Updated ShaderGLES2 to support 64 bit version codes as we hit the limit
of the 32-bit integer that was previously used as a bitset to store
enabled/disabled flags
Due to a quirk in CSG Shapes, updating is usually deferred to the next frame. This is problematic as we need to read back the geometry on the first frame when converting levels.
This PR adds a function to CSGShape to force immediate updating (if dirty), and calls it during room conversion.
Use `System.Array.Empty<T>` to get an empty array instead of allocating
a new one every time. Since arrays are immutable there is no need to
allocate them every time.
(cherry picked from commit accd05f4ad)
Higher values will make indirect lighting brighter.
A value of 1.0 represents physically accurate behavior, but higher values
can be used to make indirect lighting propagate more visibly when using
a low number of bounces.
This can be used to speed up bake times by lowering the number of bounces
then increasing the bounce indirect energy. Unlike BakedLightmapData's
energy property, this property does not affect direct lighting
emitted by light nodes or emissive materials.
This PR improves and streamlines the workflow for VisualScriptFunctionNodes Call Set Get
Uniform design.
Drag in set-get from tree is now working.
port 'pass' not backported to 3.x to keep script backwards compatibility
While there are still various bugs to solve and features to implement, the C#
support as of Godot 3.4 is fairly mature and already used by a number of users
in production. Now that we default to dotnet CLI as build tool, it also seems
to be more reliable than MSBuild.
The documentation can (and does for the most part) point out some caveats that
users should be aware of, but this info dialog has outlived its intended
purpose.
(cherry picked from commit 671467b888)
We already had `MODULE_*_ENABLED` defines but only in the modules
environment, and a few custom `*_ENABLED` defines in the main env
when we needed the information in core.
Now this is defined in a single header which can be included in the
files that need this information.
(cherry picked from commit b7297fb39c)
Clean: remove duplicate and interior vertices (uses Bullet algorithm)
Simplify: modify the geometry for further simplification (uses VHACD
algorithm)
In the editor, single convex hull now uses the clean option.
Added a new editor entry to create a simplified convex hull, can be
useful for creating convex hull from highly tessellated triangle meshes.
Specific change for 3.x:
Add support for Vector<Vector3> and PoolVector<Vector3> in the convex hull generator.
Allows customization of the maximum time a client is allowed to stay in
the the "pending" state (i.e. awaiting HTTP handshake).
This used to be 1 second by before, the new default is 3 seconds.
(cherry picked from commit 458437edef)
Setting `server_relay = false` prevents the server from letting clients
communicate with each other, but without this fix, the server would also
ignore broadcast packets.
With this change, the server still does not relay messages to other
clients, but will correctly process broadcast messages (and "exclusive"
messages) as if they were directed to just the server.
(cherry picked from commit fc255bde29)
Added `SystemConfiguration.framework` to the Xcode project to fix
undefined symbols errors building without the interpreter, like:
`_SCNetworkReachabilityScheduleWithRunLoop`.
Added explicit static constructors to the generated `NativeCalls`
class to avoid a `TypeInitializationException` at startup when
Godot attempts to read the static fields (like `godot_api_hash`)
from this class.
This seems to be an issue with Mono's AOT compiler and classes
with the `beforefieldinit` attribute. Not sure if it only happens
when the fields are only accessed via reflection as was our case.
Explicitly declaring the static constructor makes the C# compiler
not add the `beforefieldinit` attribute to the class.
We use `Mono.Cecil` to search for P/Invoke methods in assemblies in
order to collect symbols that we must prevent from being stripped.
We could pass the symbols as `-u` linker arguments (`-Wl,-u,symbol`)
for the native target (not for the project), but it was simpler to
generate referencing code and avoid changes to Godot's iOS exporter.
Replaced obsolete preprocessor check for simulator/device in C code.
Architecture can no longer be used to determine this with Apple Silicon.
The new code uses `TARGET_OS_SIMULATOR` from `TargetConditionals.h`.
We have some mono libs which can only be used in devide builds.
We were adding them as static libs. Previously it was only causing
warnings because missing arch for the simulator, but now this
is treated as an error.
To fix this we turn them into xcframeworks with dummy static libs
for the simulator and the actual ones for devices.
Fix issue when two skeletons end up directly parented.
Prevent animating TRS for skinned Mesh node.
Fix animating weights on meshes with targets but no weights.
Since Embree v3.13.0 supports AARCH64, switch back to the
official repo instead of using Embree-aarch64.
`thirdparty/embree/patches/godot-changes.patch` should now contain
an accurate diff of the changes done to the library.
(cherry picked from commit 767e374dce)
The code is based on the current version of thirdparty/vhacd and modified to use Godot's types and code style.
Additional changes:
- backported and extended PagedAllocator to allow leaked objects
- applied patch from https://github.com/bulletphysics/bullet3/pull/3037
This changes the types of a big number of variables.
General rules:
- Using `uint64_t` in general. We also considered `int64_t` but eventually
settled on keeping it unsigned, which is also closer to what one would expect
with `size_t`/`off_t`.
- We only keep `int64_t` for `seek_end` (takes a negative offset from the end)
and for the `Variant` bindings, since `Variant::INT` is `int64_t`. This means
we only need to guard against passing negative values in `core_bind.cpp`.
- Using `uint32_t` integers for concepts not needing such a huge range, like
pages, blocks, etc.
In addition:
- Improve usage of integer types in some related places; namely, `DirAccess`,
core binds.
Note:
- On Windows, `_ftelli64` reports invalid values when using 32-bit MinGW with
version < 8.0. This was an upstream bug fixed in 8.0. It breaks support for
big files on 32-bit Windows builds made with that toolchain. We might add a
workaround.
Fixes#44363.
Fixesgodotengine/godot-proposals#400.
Co-authored-by: Rémi Verschelde <rverschelde@gmail.com>
PR #45096 put the backported code in the wrong case, it should be for `type`
and not for `hint`.
Also synced `hint` enum values with `Object.PropertyHint`.
Fixes#48698.
Splits the URL into (scheme, host, port, path).
Supports both literal IPv4 and IPv6.
Strip credentials when present (e.g. http://user:pass@example.com/).
Use that function in both HTTPRequest and WebSocketClient.
(cherry picked from commit 3bb40669d5)
Since we clone the environments to build thirdparty code, we don't get an
explicit dependency on the build objects produced by that environment.
So when we update thirdparty code, Godot code using it is not necessarily
rebuilt (I think it is for changed headers, but not for changed .c/.cpp files),
which can lead to an invalid compilation output (linking old Godot .o files
with a newer, potentially ABI breaking version of thirdparty code).
This was only seen as really problematic with bullet updates (leading to
crashes when rebuilding Godot after a bullet update without cleaning .o files),
but it's safer to fix it everywhere, even if it's a LOT of hacky boilerplate.
(cherry picked from commit c7b53c03ae)
We've been using standard C library functions `memcpy`/`memset` for these since
2016 with 67f65f6639.
There was still the possibility for third-party platform ports to override the
definitions with a custom header, but this doesn't seem useful anymore.
Backport of #48239.
We need to propagate the hacky checks from the raycast config to the
lightmapper config, as the failure of a `can_build()` check is not notified to
other modules (which might even be checked further depending on the processing
order in SConstruct).
A more thorough fix would be to change SConstruct to do two loops on modules:
one to check `can_build()` and disable modules which can't build, then another
one to rechecked `can_build()` with the new lineup and do further config.
But there would be more risk for regressions than with this ad hoc hack.
Similar story for the `platform/x11/detect.py` change... oh my eyes :(
* Better handling of the scene's environment energy in the lightmapper
bakes.
* Fixed a bug where ProceduralSky::get_panorama() returned a reference
instead of a copy.
* Removed includes to Embree's internal header files.
Seeking isn't implemented in built-in video formats and can only
be supported in GDNative-provided video formats.
(cherry picked from commit ea46639e22)
- `Texture::~Texture` expects `props` to be dynamically allocated.
- `GetPropertyTable` returned a pointer to an existing `PropertyTable`
but is expected to return a newly, dynamically allocated one.
- `PropertyTable::PropertyTable()` suggests that an empty `element`
property is valid.
fix#46876fix#45573
(cherry picked from commit 09bda3f140)
Fixes#45997.
Setting a GDNativeLibrary config file as null or any other object but a ConfigFile will now cause an error.
(cherry picked from commit 618dd892f5)
We decided to rename the upcoming 3.2.4 release to 3.3 to better reflect that
it is a significant feature release, and not a maintenance update.
The `3.2` branch was also renamed to `3.x` and will now be the development
branch for future 3.x releases (3.3, 3.4, etc.).
This is a tricky one, it used to work, but it was wrong, because in such
a scenario instead of passing NULL as required by the API, it would pass
a buffer containing the `\0` terminator.
This stopped working on a specific miniupnpc version, when they fixed
some network endianess issue on Windows, to which we made a workaround,
which in turn would probably result in failures when the interface is
specified.
This commit address the issue properly, by checking the specified
interface string size, and correctly passing NULL instead of the empty
string when necessary.
Also reverts the commit that introduced the bogus workaround:
388adac947
One of those PR when the explanation is much longer then code changes
:).
`Main::cleanup()` prints warnings if it finds `StringName`s still alive.
We need the `BindingsGenerator` to be destructed before calling cleanup.
(cherry picked from commit d9603b2d73)
- Fix objects with no material being considered as fully transparent by the lightmapper.
- Added "environment_min_light" property: gives artistic control over the shadow color.
- Fixed "Custom Color" environment mode, it was ignored before.
- Added "interior" property to BakedLightmapData: controls whether dynamic capture objects receive environment light or not.
- Automatically update dynamic capture objects when the capture data changes (also works for "energy" which used to require object movement to trigger the update).
- Added "use_in_baked_light" property to GridMap: controls whether the GridMap will be included in BakedLightmap bakes.
- Set "flush zero" and "denormal zero" mode for SSE2 instructions in the Embree raycaster. According to Embree docs it should give a performance improvement.
Previously godot would try to access
`CollisionObjectBullet::bt_collision_object` even if it was null.
Fixes#46651
(cherry picked from commit c47070e165)
codeLensProvider was false, it may not be a boolean like some of the other providers can, only optionally present.
(cherry picked from commit 91ce8d5fff)
The `dev=yes` and `production=yes` options work as aliases to set a number of
options, while still aiming to allow overriding specific options if the user
wishes so. (E.g. `production=yes use_lto=no` should work to enable production
defaults *but* disable LTO.)
That wasn't working as `ARGUMENTS.get()` returns a string and not a boolean as
expected by `BoolVariable`, and this wasn't flagged as a bug... So added a
helper method using SCons' `BoolVariable._text2bool` to do the conversion
manually.
Fixes#45687.
This is really just a band-aid, our current buildsystem doesn't work well for
cross-compilation and needs a thorough refactoring to do so.
- Based on C++11's `atomic`
- Reworked `SafeRefCount` (based on the rewrite by @hpvb)
- Replaced free atomic functions by the new `SafeNumeric<T>`
- Replaced wrong cases of `volatile` by the new `SafeFlag`
- Platform-specific implementations no longer needed
Co-authored-by: Hein-Pieter van Braam-Stewart <hp@tmm.cx>
- Based on C++11's `thread` and `thread_local`
- No more need to allocate-deallocate or check for null
- No pointer anymore, just a member variable
- Platform-specific implementations no longer needed (except for the few cases of non-portable functions)
- Simpler for `NO_THREADS`
- Thread ids are now the same across platforms (main is 1; others follow)
- Based on C++11's `mutex` and `condition_variable`
- No more need to allocate-deallocate or check for null
- No pointer anymore, just a member variable
- Platform-specific implementations no longer needed
- Simpler for `NO_THREADS`
- Based on C++11's `mutex`
- No more need to allocate-deallocate or check for null
- No pointer anymore, just a member variable
- Platform-specific implementations no longer needed
- Simpler for `NO_THREADS`
- `BinaryMutex` added for special cases as the non-recursive version
- `MutexLock` now takes a reference. At this point the cases of null `Mutex`es are rare. If you ever need that, just don't use `MutexLock`.
- `ScopedMutexLock` is dropped and replaced by `MutexLock`, because they were pretty much the same.
Added BakedLightmap.use_hdr and BakedLightmap.use_color properties
that can reduce the flie size of lightmap texture at the expense of quality.
Changed the denoiser to work in a single buffer, reducing RAM
usage. Also added the `-mstackrealign` flag in the denoiser compilation
for MinGW builds. This flag helped fix a bug in Embree, so I want to see
if it will help fix GH #45296.
The input to smoothstep is not actually a weight, and the decscription
of smoothstep was pretty hard to understand and easy to misinterpret.
Clarified what it means to be approximately equal.
nearest_po2 does not do what the descriptions says it does. For one,
it returns the same power if the input is a power of 2. Second, it
returns 0 if the input is negative or 0, while the smallest possible
integral power of 2 actually is 1 (2^0 = 1). Due to the implementation
and how it is used in a lot of places, it does not seem wise to change
such a core function however, and I decided it is better to alter the
description of the built-in.
Added a few examples/clarifications/edge-cases.
(cherry picked from commit 7f9bfee0ac)
- Fix crash when a ray hits a texel with a UV2 coordinate exactly
equal to 1.0.
- Take BakedLightmap extents into account.
- Clear capture data between bakes.
- Fix minor issues with seam correction.
