In my quest to make ubsan failures a build failure I've added all
current ubsan errors by type to a new suppression file which we will
then also use with CI.
The idea is to only add known ubsan errors to this file, so this is not
a blanket exclusion of thirdparty/ this means that if we get new or
unexpected ubsan in thirdparty/ builds would fail.
These misaligned accesses are shown in all of our CI hooks. It turned
out to not be difficult to fix.
It is likely that this will improve performance for aarch64.
- Change TFM and LangVersion
- Better exception throwing (CA1510, CA1512, CA1513)
- Better exception utility method definition (CA1859)
- Prefer comparing `.Count` over calling `.Any()` (CA1860)
- Prefer `.AsSpan()` over `.Substring()` (CA1846)
- Add a few more `scoped`
- Use `RuntimeHelpers.GetUninitializedObject()` instead of `FormatterServices.GetUninitializedObject()`
- Use delegate instead of delegate pointer in variant generic conversions
- Enable EnforceExtendedAnalyzerRules in source generator projects
- Disable CS8981 on structs named movable in Godot.NativeInterop
Before this commit, ccache where only used on Mac when
`OSXCROSS_ROOT` was defined. Now, it could be used even
when that envirnment variable is not defined.
Include the needed .NET jar in the Godot templates so it's always available, then we don't need to include the jar from a .NET publish which could fail when exporting to multiple architectures because it would attempt to add the same jar for each architecture.
Update doc/classes/TextureButton.xml
Co-authored-by: Micky <66727710+Mickeon@users.noreply.github.com>
Improve explanation of texture fallback behaviors for TextureButton
Remove a bit of incorrect information from TextureButton documentation
Most other state textures are not overlaid (only `texture_focused`)
Apply suggestions from code review
We now cache the Node*<>TreeItem* mapping in the SceneTreeEditor. This
allows us to make targeted updates to the Tree used to display the scene
tree in the editor.
Previously on almost all changes to the scene tree the editor would
rebuild the entire widget, causing a large number of deallocations an
allocations. We now carefully manipulate the Tree widget in-situ saving
a large number of these allocations.
In order to know what Nodes need to be updated we add a
editor_state_changed signal to Node, this is a TOOLS_ENABLED,
editor-only signal fired when changes to Node happen that are relevant
to editor state.
We also now make sure that when nodes are moved/renamed we don't check
expensive properties that cannot contain NodePaths. This saves a lot of
time when SceneTreeDock renames a node in a scene with a lot of
MeshInstances. This makes renaming nodes go from ~27 seconds to ~2
seconds on large scenes.
SceneTreeEditor instances will now also not do all of the potentially
expensive update work if they are invisible. This behavior is turned off
by default so it won't affect existing users. This change allows the
editor to only update SceneTreeEditors that actually in view. In
practice this means that for most changes instead of updating 6
SceneTreeEditors we only update 1 instantly, and the others only when
they become visible.
There is definitely more that could be done, but this is already a
massive improvement. In complex scenes we see an improvement of 10x,
things that used to take ~30 seconds now only take 2.
This fixes#83460
I want to thank KoBeWi, TokisanGames, a-johnston, aniel080400 for
their tireless testing. And AeioMuch for their testing and providing a
fix for the hover issue.
Setting it only for release templates on Windows and macOS was inconsistent,
and Jolt requires it as a minimum.
Drop the `-mxsave` flag from the raycast module, this doesn't seem to be
used explicitly by Embree, and unnecessarily makes our config and baseline
muddy.
Because of ref safety changes in the languages, all methods that return an interop struct have to have all other reference parameters marked as scoped to signal the the method does not capture that reference.
The variant change is necessary, because for some reason a type of the exact shape godot_variant is in, crashes the .NET 7 JIT, but when changing it to be sequential with the same effective layout it works.
Thaddeus Crews
HP van Braam
clayjohn
Lars Pettersson
Hugo Locurcio
Lukas Tenbrink
Rémi Verschelde
Paul Joannon
Pablo Andres Fuente
Pedro J. Estébanez
Florent Guiocheau
Raul Santos
Mikael Hermansson
Malcolm Anderson
smix8
Rémi Verschelde
kobewi
Markus Sauermann
BlueCube3310
RedworkDE
havi05
Matias N. Goldberg
Yufeng Ying
Robert Yevdokimov
David Snopek