# Objective
- Implement the foundations of automatic batching/instancing of draw
commands as the next step from #89
- NOTE: More performance improvements will come when more data is
managed and bound in ways that do not require rebinding such as mesh,
material, and texture data.
## Solution
- The core idea for batching of draw commands is to check whether any of
the information that has to be passed when encoding a draw command
changes between two things that are being drawn according to the sorted
render phase order. These should be things like the pipeline, bind
groups and their dynamic offsets, index/vertex buffers, and so on.
- The following assumptions have been made:
- Only entities with prepared assets (pipelines, materials, meshes) are
queued to phases
- View bindings are constant across a phase for a given draw function as
phases are per-view
- `batch_and_prepare_render_phase` is the only system that performs this
batching and has sole responsibility for preparing the per-object data.
As such the mesh binding and dynamic offsets are assumed to only vary as
a result of the `batch_and_prepare_render_phase` system, e.g. due to
having to split data across separate uniform bindings within the same
buffer due to the maximum uniform buffer binding size.
- Implement `GpuArrayBuffer` for `Mesh2dUniform` to store Mesh2dUniform
in arrays in GPU buffers rather than each one being at a dynamic offset
in a uniform buffer. This is the same optimisation that was made for 3D
not long ago.
- Change batch size for a range in `PhaseItem`, adding API for getting
or mutating the range. This is more flexible than a size as the length
of the range can be used in place of the size, but the start and end can
be otherwise whatever is needed.
- Add an optional mesh bind group dynamic offset to `PhaseItem`. This
avoids having to do a massive table move just to insert
`GpuArrayBufferIndex` components.
## Benchmarks
All tests have been run on an M1 Max on AC power. `bevymark` and
`many_cubes` were modified to use 1920x1080 with a scale factor of 1. I
run a script that runs a separate Tracy capture process, and then runs
the bevy example with `--features bevy_ci_testing,trace_tracy` and
`CI_TESTING_CONFIG=../benchmark.ron` with the contents of
`../benchmark.ron`:
```rust
(
exit_after: Some(1500)
)
```
...in order to run each test for 1500 frames.
The recent changes to `many_cubes` and `bevymark` added reproducible
random number generation so that with the same settings, the same rng
will occur. They also added benchmark modes that use a fixed delta time
for animations. Combined this means that the same frames should be
rendered both on main and on the branch.
The graphs compare main (yellow) to this PR (red).
### 3D Mesh `many_cubes --benchmark`
<img width="1411" alt="Screenshot 2023-09-03 at 23 42 10"
src="https://github.com/bevyengine/bevy/assets/302146/2088716a-c918-486c-8129-090b26fd2bc4">
The mesh and material are the same for all instances. This is basically
the best case for the initial batching implementation as it results in 1
draw for the ~11.7k visible meshes. It gives a ~30% reduction in median
frame time.
The 1000th frame is identical using the flip tool:
```
Mean: 0.000000
Weighted median: 0.000000
1st weighted quartile: 0.000000
3rd weighted quartile: 0.000000
Min: 0.000000
Max: 0.000000
Evaluation time: 0.4615 seconds
```
### 3D Mesh `many_cubes --benchmark --material-texture-count 10`
<img width="1404" alt="Screenshot 2023-09-03 at 23 45 18"
src="https://github.com/bevyengine/bevy/assets/302146/5ee9c447-5bd2-45c6-9706-ac5ff8916daf">
This run uses 10 different materials by varying their textures. The
materials are randomly selected, and there is no sorting by material
bind group for opaque 3D so any batching is 'random'. The PR produces a
~5% reduction in median frame time. If we were to sort the opaque phase
by the material bind group, then this should be a lot faster. This
produces about 10.5k draws for the 11.7k visible entities. This makes
sense as randomly selecting from 10 materials gives a chance that two
adjacent entities randomly select the same material and can be batched.
The 1000th frame is identical in flip:
```
Mean: 0.000000
Weighted median: 0.000000
1st weighted quartile: 0.000000
3rd weighted quartile: 0.000000
Min: 0.000000
Max: 0.000000
Evaluation time: 0.4537 seconds
```
### 3D Mesh `many_cubes --benchmark --vary-per-instance`
<img width="1394" alt="Screenshot 2023-09-03 at 23 48 44"
src="https://github.com/bevyengine/bevy/assets/302146/f02a816b-a444-4c18-a96a-63b5436f3b7f">
This run varies the material data per instance by randomly-generating
its colour. This is the worst case for batching and that it performs
about the same as `main` is a good thing as it demonstrates that the
batching has minimal overhead when dealing with ~11k visible mesh
entities.
The 1000th frame is identical according to flip:
```
Mean: 0.000000
Weighted median: 0.000000
1st weighted quartile: 0.000000
3rd weighted quartile: 0.000000
Min: 0.000000
Max: 0.000000
Evaluation time: 0.4568 seconds
```
### 2D Mesh `bevymark --benchmark --waves 160 --per-wave 1000 --mode
mesh2d`
<img width="1412" alt="Screenshot 2023-09-03 at 23 59 56"
src="https://github.com/bevyengine/bevy/assets/302146/cb02ae07-237b-4646-ae9f-fda4dafcbad4">
This spawns 160 waves of 1000 quad meshes that are shaded with
ColorMaterial. Each wave has a different material so 160 waves currently
should result in 160 batches. This results in a 50% reduction in median
frame time.
Capturing a screenshot of the 1000th frame main vs PR gives:
```
Mean: 0.001222
Weighted median: 0.750432
1st weighted quartile: 0.453494
3rd weighted quartile: 0.969758
Min: 0.000000
Max: 0.990296
Evaluation time: 0.4255 seconds
```
So they seem to produce the same results. I also double-checked the
number of draws. `main` does 160000 draws, and the PR does 160, as
expected.
### 2D Mesh `bevymark --benchmark --waves 160 --per-wave 1000 --mode
mesh2d --material-texture-count 10`
<img width="1392" alt="Screenshot 2023-09-04 at 00 09 22"
src="https://github.com/bevyengine/bevy/assets/302146/4358da2e-ce32-4134-82df-3ab74c40849c">
This generates 10 textures and generates materials for each of those and
then selects one material per wave. The median frame time is reduced by
50%. Similar to the plain run above, this produces 160 draws on the PR
and 160000 on `main` and the 1000th frame is identical (ignoring the fps
counter text overlay).
```
Mean: 0.002877
Weighted median: 0.964980
1st weighted quartile: 0.668871
3rd weighted quartile: 0.982749
Min: 0.000000
Max: 0.992377
Evaluation time: 0.4301 seconds
```
### 2D Mesh `bevymark --benchmark --waves 160 --per-wave 1000 --mode
mesh2d --vary-per-instance`
<img width="1396" alt="Screenshot 2023-09-04 at 00 13 53"
src="https://github.com/bevyengine/bevy/assets/302146/b2198b18-3439-47ad-919a-cdabe190facb">
This creates unique materials per instance by randomly-generating the
material's colour. This is the worst case for 2D batching. Somehow, this
PR manages a 7% reduction in median frame time. Both main and this PR
issue 160000 draws.
The 1000th frame is the same:
```
Mean: 0.001214
Weighted median: 0.937499
1st weighted quartile: 0.635467
3rd weighted quartile: 0.979085
Min: 0.000000
Max: 0.988971
Evaluation time: 0.4462 seconds
```
### 2D Sprite `bevymark --benchmark --waves 160 --per-wave 1000 --mode
sprite`
<img width="1396" alt="Screenshot 2023-09-04 at 12 21 12"
src="https://github.com/bevyengine/bevy/assets/302146/8b31e915-d6be-4cac-abf5-c6a4da9c3d43">
This just spawns 160 waves of 1000 sprites. There should be and is no
notable difference between main and the PR.
### 2D Sprite `bevymark --benchmark --waves 160 --per-wave 1000 --mode
sprite --material-texture-count 10`
<img width="1389" alt="Screenshot 2023-09-04 at 12 36 08"
src="https://github.com/bevyengine/bevy/assets/302146/45fe8d6d-c901-4062-a349-3693dd044413">
This spawns the sprites selecting a texture at random per instance from
the 10 generated textures. This has no significant change vs main and
shouldn't.
### 2D Sprite `bevymark --benchmark --waves 160 --per-wave 1000 --mode
sprite --vary-per-instance`
<img width="1401" alt="Screenshot 2023-09-04 at 12 29 52"
src="https://github.com/bevyengine/bevy/assets/302146/762c5c60-352e-471f-8dbe-bbf10e24ebd6">
This sets the sprite colour as being unique per instance. This can still
all be drawn using one batch. There should be no difference but the PR
produces median frame times that are 4% higher. Investigation showed no
clear sources of cost, rather a mix of give and take that should not
happen. It seems like noise in the results.
### Summary
| Benchmark | % change in median frame time |
| ------------- | ------------- |
| many_cubes | 🟩 -30% |
| many_cubes 10 materials | 🟩 -5% |
| many_cubes unique materials | 🟩 ~0% |
| bevymark mesh2d | 🟩 -50% |
| bevymark mesh2d 10 materials | 🟩 -50% |
| bevymark mesh2d unique materials | 🟩 -7% |
| bevymark sprite | 🟥 2% |
| bevymark sprite 10 materials | 🟥 0.6% |
| bevymark sprite unique materials | 🟥 4.1% |
---
## Changelog
- Added: 2D and 3D mesh entities that share the same mesh and material
(same textures, same data) are now batched into the same draw command
for better performance.
---------
Co-authored-by: robtfm <50659922+robtfm@users.noreply.github.com>
Co-authored-by: Nicola Papale <nico@nicopap.ch>
I'm adopting this ~~child~~ PR.
# Objective
- Working with exclusive world access is not always easy: in many cases,
a standard system or three is more ergonomic to write, and more
modularly maintainable.
- For small, one-off tasks (commonly handled with scripting), running an
event-reader system incurs a small but flat overhead cost and muddies
the schedule.
- Certain forms of logic (e.g. turn-based games) want very fine-grained
linear and/or branching control over logic.
- SystemState is not automatically cached, and so performance can suffer
and change detection breaks.
- Fixes https://github.com/bevyengine/bevy/issues/2192.
- Partial workaround for https://github.com/bevyengine/bevy/issues/279.
## Solution
- Adds a SystemRegistry resource to the World, which stores initialized
systems keyed by their SystemSet.
- Allows users to call world.run_system(my_system) and
commands.run_system(my_system), without re-initializing or losing state
(essential for change detection).
- Add a Callback type to enable convenient use of dynamic one shot
systems and reduce the mental overhead of working with Box<dyn
SystemSet>.
- Allow users to run systems based on their SystemSet, enabling more
complex user-made abstractions.
## Future work
- Parameterized one-shot systems would improve reusability and bring
them closer to events and commands. The API could be something like
run_system_with_input(my_system, my_input) and use the In SystemParam.
- We should evaluate the unification of commands and one-shot systems
since they are two different ways to run logic on demand over a World.
### Prior attempts
- https://github.com/bevyengine/bevy/pull/2234
- https://github.com/bevyengine/bevy/pull/2417
- https://github.com/bevyengine/bevy/pull/4090
- https://github.com/bevyengine/bevy/pull/7999
This PR continues the work done in
https://github.com/bevyengine/bevy/pull/7999.
---------
Co-authored-by: Alice Cecile <alice.i.cecile@gmail.com>
Co-authored-by: Federico Rinaldi <gisquerin@gmail.com>
Co-authored-by: MinerSebas <66798382+MinerSebas@users.noreply.github.com>
Co-authored-by: Aevyrie <aevyrie@gmail.com>
Co-authored-by: Alejandro Pascual Pozo <alejandro.pascual.pozo@gmail.com>
Co-authored-by: Rob Parrett <robparrett@gmail.com>
Co-authored-by: François <mockersf@gmail.com>
Co-authored-by: Dmytro Banin <banind@cs.washington.edu>
Co-authored-by: James Liu <contact@jamessliu.com>
# Objective
Fix#9747
## Solution
Linkers don't like what we're doing with CowArc (I'm guessing it has
something to do with `?Sized`). Weirdly the `Reflect` derive on
`AssetPath` doesn't fail, despite `CowArc` not implementing `Reflect`.
To resolve this, we manually implement "reflect value" for
`AssetPath<'static>`. It sadly cannot use `impl_reflect_value` because
that macro doesn't support static lifetimes.
---------
Co-authored-by: Martin Dickopp <martin@zero-based.org>
# Objective
The `AssetServer` and `AssetProcessor` do a lot of `AssetPath` cloning
(across many threads). To store the path on the handle, to store paths
in dependency lists, to pass an owned path to the offloaded thread, to
pass a path to the LoadContext, etc , etc. Cloning multiple string
allocations multiple times like this will add up. It is worth optimizing
this.
Referenced in #9714
## Solution
Added a new `CowArc<T>` type to `bevy_util`, which behaves a lot like
`Cow<T>`, but the Owned variant is an `Arc<T>`. Use this in place of
`Cow<str>` and `Cow<Path>` on `AssetPath`.
---
## Changelog
- `AssetPath` now internally uses `CowArc`, making clone operations much
cheaper
- `AssetPath` now serializes as `AssetPath("some_path.extension#Label")`
instead of as `AssetPath { path: "some_path.extension", label:
Some("Label) }`
## Migration Guide
```rust
// Old
AssetPath::new("logo.png", None);
// New
AssetPath::new("logo.png");
// Old
AssetPath::new("scene.gltf", Some("Mesh0");
// New
AssetPath::new("scene.gltf").with_label("Mesh0");
```
`AssetPath` now serializes as `AssetPath("some_path.extension#Label")`
instead of as `AssetPath { path: "some_path.extension", label:
Some("Label) }`
---------
Co-authored-by: Pascal Hertleif <killercup@gmail.com>
# Objective
Any time we wish to transform the output of a system, we currently use
system piping to do so:
```rust
my_system.pipe(|In(x)| do_something(x))
```
Unfortunately, system piping is not a zero cost abstraction. Each call
to `.pipe` requires allocating two extra access sets: one for the second
system and one for the combined accesses of both systems. This also adds
extra work to each call to `update_archetype_component_access`, which
stacks as one adds multiple layers of system piping.
## Solution
Add the `AdapterSystem` abstraction: similar to `CombinatorSystem`, this
allows you to implement a trait to generically control how a system is
run and how its inputs and outputs are processed. Unlike
`CombinatorSystem`, this does not have any overhead when computing world
accesses which makes it ideal for simple operations such as inverting or
ignoring the output of a system.
Add the extension method `.map(...)`: this is similar to `.pipe(...)`,
only it accepts a closure as an argument instead of an `In<T>` system.
```rust
my_system.map(do_something)
```
This has the added benefit of making system names less messy: a system
that ignores its output will just be called `my_system`, instead of
`Pipe(my_system, ignore)`
---
## Changelog
TODO
## Migration Guide
The `system_adapter` functions have been deprecated: use `.map` instead,
which is a lightweight alternative to `.pipe`.
```rust
// Before:
my_system.pipe(system_adapter::ignore)
my_system.pipe(system_adapter::unwrap)
my_system.pipe(system_adapter::new(T::from))
// After:
my_system.map(std::mem::drop)
my_system.map(Result::unwrap)
my_system.map(T::from)
// Before:
my_system.pipe(system_adapter::info)
my_system.pipe(system_adapter::dbg)
my_system.pipe(system_adapter::warn)
my_system.pipe(system_adapter::error)
// After:
my_system.map(bevy_utils::info)
my_system.map(bevy_utils::dbg)
my_system.map(bevy_utils::warn)
my_system.map(bevy_utils::error)
```
---------
Co-authored-by: Alice Cecile <alice.i.cecile@gmail.com>
# Objective
Fixes#9509
## Solution
We use the assumption, that enum types are uppercase in contrast to
module names.
[`collapse_type_name`](crates/bevy_util/src/short_names) is now
retaining the second last segment, if it starts with a uppercase
character.
---------
Co-authored-by: Emi <emanuel.boehm@gmail.com>
Co-authored-by: Nicola Papale <nicopap@users.noreply.github.com>
CI-capable version of #9086
---------
Co-authored-by: Bevy Auto Releaser <41898282+github-actions[bot]@users.noreply.github.com>
Co-authored-by: François <mockersf@gmail.com>
# Objective
Fix typos throughout the project.
## Solution
[`typos`](https://github.com/crate-ci/typos) project was used for
scanning, but no automatic corrections were applied. I checked
everything by hand before fixing.
Most of the changes are documentation/comments corrections. Also, there
are few trivial changes to code (variable name, pub(crate) function name
and a few error/panic messages).
## Unsolved
`bevy_reflect_derive` has
[typo](1b51053f19/crates/bevy_reflect/bevy_reflect_derive/src/type_path.rs (L76))
in enum variant name that I didn't fix. Enum is `pub(crate)`, so there
shouldn't be any trouble if fixed. However, code is tightly coupled with
macro usage, so I decided to leave it for more experienced contributor
just in case.
I created this manually as Github didn't want to run CI for the
workflow-generated PR. I'm guessing we didn't hit this in previous
releases because we used bors.
Co-authored-by: Bevy Auto Releaser <41898282+github-actions[bot]@users.noreply.github.com>
# Objective
Reduce missing docs warning noise when building examples for wasm
## Solution
Added "#[allow(missing_docs)]" on the wasm specific version of
BoxedFuture
# Objective
Methods for interacting with world schedules currently have two
variants: one that takes `impl ScheduleLabel` and one that takes `&dyn
ScheduleLabel`. Operations such as `run_schedule` or `schedule_scope`
only use the label by reference, so there is little reason to have an
owned variant of these functions.
## Solution
Decrease maintenance burden by merging the `ref` variants of these
functions with the owned variants.
---
## Changelog
- Deprecated `World::run_schedule_ref`. It is now redundant, since
`World::run_schedule` can take values by reference.
## Migration Guide
The method `World::run_schedule_ref` has been deprecated, and will be
removed in the next version of Bevy. Use `run_schedule` instead.
# Objective
Label traits such as `ScheduleLabel` currently have a major footgun: the
trait is implemented for `Box<dyn ScheduleLabel>`, but the
implementation does not function as one would expect since `Box<T>` is
considered to be a distinct type from `T`. This is because the behavior
of the `ScheduleLabel` trait is specified mainly through blanket
implementations, which prevents `Box<dyn ScheduleLabel>` from being
properly special-cased.
## Solution
Replace the blanket-implemented behavior with a series of methods
defined on `ScheduleLabel`. This allows us to fully special-case
`Box<dyn ScheduleLabel>` .
---
## Changelog
Fixed a bug where boxed label types (such as `Box<dyn ScheduleLabel>`)
behaved incorrectly when compared with concretely-typed labels.
## Migration Guide
The `ScheduleLabel` trait has been refactored to no longer depend on the
traits `std::any::Any`, `bevy_utils::DynEq`, and `bevy_utils::DynHash`.
Any manual implementations will need to implement new trait methods in
their stead.
```rust
impl ScheduleLabel for MyType {
// Before:
fn dyn_clone(&self) -> Box<dyn ScheduleLabel> { ... }
// After:
fn dyn_clone(&self) -> Box<dyn ScheduleLabel> { ... }
fn as_dyn_eq(&self) -> &dyn DynEq {
self
}
// No, `mut state: &mut` is not a typo.
fn dyn_hash(&self, mut state: &mut dyn Hasher) {
self.hash(&mut state);
// Hashing the TypeId isn't strictly necessary, but it prevents collisions.
TypeId::of::<Self>().hash(&mut state);
}
}
```
# Objective
The clippy lint `type_complexity` is known not to play well with bevy.
It frequently triggers when writing complex queries, and taking the
lint's advice of using a type alias almost always just obfuscates the
code with no benefit. Because of this, this lint is currently ignored in
CI, but unfortunately it still shows up when viewing bevy code in an
IDE.
As someone who's made a fair amount of pull requests to this repo, I
will say that this issue has been a consistent thorn in my side. Since
bevy code is filled with spurious, ignorable warnings, it can be very
difficult to spot the *real* warnings that must be fixed -- most of the
time I just ignore all warnings, only to later find out that one of them
was real after I'm done when CI runs.
## Solution
Suppress this lint in all bevy crates. This was previously attempted in
#7050, but the review process ended up making it more complicated than
it needs to be and landed on a subpar solution.
The discussion in https://github.com/rust-lang/rust-clippy/pull/10571
explores some better long-term solutions to this problem. Since there is
no timeline on when these solutions may land, we should resolve this
issue in the meantime by locally suppressing these lints.
### Unresolved issues
Currently, these lints are not suppressed in our examples, since that
would require suppressing the lint in every single source file. They are
still ignored in CI.
# Objective
The function `SyncUnsafeCell::from_mut` returns `&SyncUnsafeCell<T>`,
even though it could return `&mut SyncUnsafeCell<T>`. This means it is
not possible to call `get_mut` on the returned value, so you need to use
unsafe code to get exclusive access back.
## Solution
Return `&mut Self` instead of `&Self` in `SyncUnsafeCell::from_mut`.
This is consistent with my proposal for `UnsafeCell::from_mut`:
https://github.com/rust-lang/libs-team/issues/198.
Replace an unsafe pointer dereference with a safe call to `get_mut`.
---
## Changelog
+ The function `bevy_utils::SyncUnsafeCell::get_mut` now returns a value
of type `&mut SyncUnsafeCell<T>`. Previously, this returned an immutable
reference.
## Migration Guide
The function `bevy_utils::SyncUnsafeCell::get_mut` now returns a value
of type `&mut SyncUnsafeCell<T>`. Previously, this returned an immutable
reference.
# Objective
While working on #7442 i discovered that `get_short_name` does not work well with sub paths after closing brackets. It currently turns `bevy_asset::assets::Assets<bevy_scene::dynamic_scene::DynamicScene>::asset_event_system` into `Assets<DynamicScene>asset_event_system`. This PR fixes that.
## Solution
- Retain `::` after a closing bracket like `>`, `)` or `]`.
- Add a test for all sub path after closing bracket cases.
# Objective
The type `SyncCell<T>` (added in #5483) is used to force any wrapped type to be `Sync`, by only allowing exclusive access to the wrapped value. This restriction is unnecessary for types which are already `Sync`.
---
## Changelog
+ Added the method `read` to `SyncCell`, which allows shared access to values that already implement the `Sync` trait.
# Objective
- Fixes#5432
- Fixes#6680
## Solution
- move code responsible for generating the `impl TypeUuid` from `type_uuid_derive` into a new function, `gen_impl_type_uuid`.
- this allows the new proc macro, `impl_type_uuid`, to call the code for generation.
- added struct `TypeUuidDef` and implemented `syn::Parse` to allow parsing of the input for the new macro.
- finally, used the new macro `impl_type_uuid` to implement `TypeUuid` for the standard library (in `crates/bevy_reflect/src/type_uuid_impl.rs`).
- fixes#6680 by doing a wrapping add of the param's index to its `TYPE_UUID`
Co-authored-by: dis-da-moe <84386186+dis-da-moe@users.noreply.github.com>
Profiles show that in extremely hot loops, like the draw loops in the renderer, invoking the trace! macro has noticeable overhead, even if the trace log level is not enabled.
Solve this by introduce a 'wrapper' detailed_trace macro around trace, that wraps the trace! log statement in a trivially false if statement unless a cargo feature is enabled
# Objective
- Eliminate significant overhead observed with trace-level logging in render hot loops, even when trace log level is not enabled.
- This is an alternative solution to the one proposed in #7223
## Solution
- Introduce a wrapper around the `trace!` macro called `detailed_trace!`. This macro wraps the `trace!` macro with an if statement that is conditional on a new cargo feature, `detailed_trace`. When the feature is not enabled (the default), then the if statement is trivially false and should be optimized away at compile time.
- Convert the observed hot occurrences of trace logging in `TrackedRenderPass` with this new macro.
Testing the results of
```
cargo run --profile stress-test --features bevy/trace_tracy --example many_cubes -- spheres
```
shows significant improvement of the `main_opaque_pass_3d` of the renderer, a median time decrease from 6.0ms to 3.5ms.
---
## Changelog
- For performance reasons, some detailed renderer trace logs now require the use of cargo feature `detailed_trace` in addition to setting the log level to `TRACE` in order to be shown.
## Migration Guide
- Some detailed bevy trace events now require the use of the cargo feature `detailed_trace` in addition to enabling `TRACE` level logging to view. Should you wish to see these logs, please compile your code with the bevy feature `detailed_trace`. Currently, the only logs that are affected are the renderer logs pertaining to `TrackedRenderPass` functions
Huge thanks to @maniwani, @devil-ira, @hymm, @cart, @superdump and @jakobhellermann for the help with this PR.
# Objective
- Followup #6587.
- Minimal integration for the Stageless Scheduling RFC: https://github.com/bevyengine/rfcs/pull/45
## Solution
- [x] Remove old scheduling module
- [x] Migrate new methods to no longer use extension methods
- [x] Fix compiler errors
- [x] Fix benchmarks
- [x] Fix examples
- [x] Fix docs
- [x] Fix tests
## Changelog
### Added
- a large number of methods on `App` to work with schedules ergonomically
- the `CoreSchedule` enum
- `App::add_extract_system` via the `RenderingAppExtension` trait extension method
- the private `prepare_view_uniforms` system now has a public system set for scheduling purposes, called `ViewSet::PrepareUniforms`
### Removed
- stages, and all code that mentions stages
- states have been dramatically simplified, and no longer use a stack
- `RunCriteriaLabel`
- `AsSystemLabel` trait
- `on_hierarchy_reports_enabled` run criteria (now just uses an ad hoc resource checking run condition)
- systems in `RenderSet/Stage::Extract` no longer warn when they do not read data from the main world
- `RunCriteriaLabel`
- `transform_propagate_system_set`: this was a nonstandard pattern that didn't actually provide enough control. The systems are already `pub`: the docs have been updated to ensure that the third-party usage is clear.
### Changed
- `System::default_labels` is now `System::default_system_sets`.
- `App::add_default_labels` is now `App::add_default_sets`
- `CoreStage` and `StartupStage` enums are now `CoreSet` and `StartupSet`
- `App::add_system_set` was renamed to `App::add_systems`
- The `StartupSchedule` label is now defined as part of the `CoreSchedules` enum
- `.label(SystemLabel)` is now referred to as `.in_set(SystemSet)`
- `SystemLabel` trait was replaced by `SystemSet`
- `SystemTypeIdLabel<T>` was replaced by `SystemSetType<T>`
- The `ReportHierarchyIssue` resource now has a public constructor (`new`), and implements `PartialEq`
- Fixed time steps now use a schedule (`CoreSchedule::FixedTimeStep`) rather than a run criteria.
- Adding rendering extraction systems now panics rather than silently failing if no subapp with the `RenderApp` label is found.
- the `calculate_bounds` system, with the `CalculateBounds` label, is now in `CoreSet::Update`, rather than in `CoreSet::PostUpdate` before commands are applied.
- `SceneSpawnerSystem` now runs under `CoreSet::Update`, rather than `CoreStage::PreUpdate.at_end()`.
- `bevy_pbr::add_clusters` is no longer an exclusive system
- the top level `bevy_ecs::schedule` module was replaced with `bevy_ecs::scheduling`
- `tick_global_task_pools_on_main_thread` is no longer run as an exclusive system. Instead, it has been replaced by `tick_global_task_pools`, which uses a `NonSend` resource to force running on the main thread.
## Migration Guide
- Calls to `.label(MyLabel)` should be replaced with `.in_set(MySet)`
- Stages have been removed. Replace these with system sets, and then add command flushes using the `apply_system_buffers` exclusive system where needed.
- The `CoreStage`, `StartupStage, `RenderStage` and `AssetStage` enums have been replaced with `CoreSet`, `StartupSet, `RenderSet` and `AssetSet`. The same scheduling guarantees have been preserved.
- Systems are no longer added to `CoreSet::Update` by default. Add systems manually if this behavior is needed, although you should consider adding your game logic systems to `CoreSchedule::FixedTimestep` instead for more reliable framerate-independent behavior.
- Similarly, startup systems are no longer part of `StartupSet::Startup` by default. In most cases, this won't matter to you.
- For example, `add_system_to_stage(CoreStage::PostUpdate, my_system)` should be replaced with
- `add_system(my_system.in_set(CoreSet::PostUpdate)`
- When testing systems or otherwise running them in a headless fashion, simply construct and run a schedule using `Schedule::new()` and `World::run_schedule` rather than constructing stages
- Run criteria have been renamed to run conditions. These can now be combined with each other and with states.
- Looping run criteria and state stacks have been removed. Use an exclusive system that runs a schedule if you need this level of control over system control flow.
- For app-level control flow over which schedules get run when (such as for rollback networking), create your own schedule and insert it under the `CoreSchedule::Outer` label.
- Fixed timesteps are now evaluated in a schedule, rather than controlled via run criteria. The `run_fixed_timestep` system runs this schedule between `CoreSet::First` and `CoreSet::PreUpdate` by default.
- Command flush points introduced by `AssetStage` have been removed. If you were relying on these, add them back manually.
- Adding extract systems is now typically done directly on the main app. Make sure the `RenderingAppExtension` trait is in scope, then call `app.add_extract_system(my_system)`.
- the `calculate_bounds` system, with the `CalculateBounds` label, is now in `CoreSet::Update`, rather than in `CoreSet::PostUpdate` before commands are applied. You may need to order your movement systems to occur before this system in order to avoid system order ambiguities in culling behavior.
- the `RenderLabel` `AppLabel` was renamed to `RenderApp` for clarity
- `App::add_state` now takes 0 arguments: the starting state is set based on the `Default` impl.
- Instead of creating `SystemSet` containers for systems that run in stages, simply use `.on_enter::<State::Variant>()` or its `on_exit` or `on_update` siblings.
- `SystemLabel` derives should be replaced with `SystemSet`. You will also need to add the `Debug`, `PartialEq`, `Eq`, and `Hash` traits to satisfy the new trait bounds.
- `with_run_criteria` has been renamed to `run_if`. Run criteria have been renamed to run conditions for clarity, and should now simply return a bool.
- States have been dramatically simplified: there is no longer a "state stack". To queue a transition to the next state, call `NextState::set`
## TODO
- [x] remove dead methods on App and World
- [x] add `App::add_system_to_schedule` and `App::add_systems_to_schedule`
- [x] avoid adding the default system set at inappropriate times
- [x] remove any accidental cycles in the default plugins schedule
- [x] migrate benchmarks
- [x] expose explicit labels for the built-in command flush points
- [x] migrate engine code
- [x] remove all mentions of stages from the docs
- [x] verify docs for States
- [x] fix uses of exclusive systems that use .end / .at_start / .before_commands
- [x] migrate RenderStage and AssetStage
- [x] migrate examples
- [x] ensure that transform propagation is exported in a sufficiently public way (the systems are already pub)
- [x] ensure that on_enter schedules are run at least once before the main app
- [x] re-enable opt-in to execution order ambiguities
- [x] revert change to `update_bounds` to ensure it runs in `PostUpdate`
- [x] test all examples
- [x] unbreak directional lights
- [x] unbreak shadows (see 3d_scene, 3d_shape, lighting, transparaency_3d examples)
- [x] game menu example shows loading screen and menu simultaneously
- [x] display settings menu is a blank screen
- [x] `without_winit` example panics
- [x] ensure all tests pass
- [x] SubApp doc test fails
- [x] runs_spawn_local tasks fails
- [x] [Fix panic_when_hierachy_cycle test hanging](https://github.com/alice-i-cecile/bevy/pull/120)
## Points of Difficulty and Controversy
**Reviewers, please give feedback on these and look closely**
1. Default sets, from the RFC, have been removed. These added a tremendous amount of implicit complexity and result in hard to debug scheduling errors. They're going to be tackled in the form of "base sets" by @cart in a followup.
2. The outer schedule controls which schedule is run when `App::update` is called.
3. I implemented `Label for `Box<dyn Label>` for our label types. This enables us to store schedule labels in concrete form, and then later run them. I ran into the same set of problems when working with one-shot systems. We've previously investigated this pattern in depth, and it does not appear to lead to extra indirection with nested boxes.
4. `SubApp::update` simply runs the default schedule once. This sucks, but this whole API is incomplete and this was the minimal changeset.
5. `time_system` and `tick_global_task_pools_on_main_thread` no longer use exclusive systems to attempt to force scheduling order
6. Implemetnation strategy for fixed timesteps
7. `AssetStage` was migrated to `AssetSet` without reintroducing command flush points. These did not appear to be used, and it's nice to remove these bottlenecks.
8. Migration of `bevy_render/lib.rs` and pipelined rendering. The logic here is unusually tricky, as we have complex scheduling requirements.
## Future Work (ideally before 0.10)
- Rename schedule_v3 module to schedule or scheduling
- Add a derive macro to states, and likely a `EnumIter` trait of some form
- Figure out what exactly to do with the "systems added should basically work by default" problem
- Improve ergonomics for working with fixed timesteps and states
- Polish FixedTime API to match Time
- Rebase and merge #7415
- Resolve all internal ambiguities (blocked on better tools, especially #7442)
- Add "base sets" to replace the removed default sets.
# Objective
I found several words in code and docs are incorrect. This should be fixed.
## Solution
- Fix several minor typos
Co-authored-by: Chris Ohk <utilforever@gmail.com>
# Objective
Complete the first part of the migration detailed in bevyengine/rfcs#45.
## Solution
Add all the new stuff.
### TODO
- [x] Impl tuple methods.
- [x] Impl chaining.
- [x] Port ambiguity detection.
- [x] Write docs.
- [x] ~~Write more tests.~~(will do later)
- [ ] Write changelog and examples here?
- [x] ~~Replace `petgraph`.~~ (will do later)
Co-authored-by: james7132 <contact@jamessliu.com>
Co-authored-by: Michael Hsu <mike.hsu@gmail.com>
Co-authored-by: Mike Hsu <mike.hsu@gmail.com>
# Objective
- The function `BlobVec::replace_unchecked` has informal use of safety comments.
- This function does strange things with `OwningPtr` in order to get around the borrow checker.
## Solution
- Put safety comments in front of each unsafe operation. Describe the specific invariants of each operation and how they apply here.
- Added a guard type `OnDrop`, which is used to simplify ownership transfer in case of a panic.
---
## Changelog
+ Added the guard type `bevy_utils::OnDrop`.
+ Added conversions from `Ptr`, `PtrMut`, and `OwningPtr` to `NonNull<u8>`.
# Objective
Partially address #3492.
## Solution
Document the remaining undocumented members of `bevy_utils` and set `warn(missing_docs)` on the crate level. Also enabled `clippy::undocumented_unsafe_blocks` as a warning on the crate to keep it in sync with `bevy_ecs`'s warnings.
# Objective
Currently, `Local` has a `Sync` bound. Theoretically this is unnecessary as a local can only ever be accessed from its own system, ensuring exclusive access on one thread. This PR removes this restriction.
## Solution
- By removing the `Resource` bound from `Local` and adding the new `SyncCell` threading primative, `Local` can have the `Sync` bound removed.
## Changelog
### Added
- Added `SyncCell` to `bevy_utils`
### Changed
- Removed `Resource` bound from `Local`
- `Local` is now wrapped in a `SyncCell`
## Migration Guide
- Any code relying on `Local<T>` having `T: Resource` may have to be changed, but this is unlikely.
Co-authored-by: PROMETHIA-27 <42193387+PROMETHIA-27@users.noreply.github.com>
# Objective
I noticed while working on #5366 that the documentation for label types wasn't working correctly. Having experimented with this for a few weeks, I believe that generating docs in macros is more effort than it's worth.
## Solution
Add more boilerplate, copy-paste and edit the docs across types. This also lets us add custom doctests for specific types. Also, we don't need `concat_idents` as a dependency anymore.
# Objective
- Closes#4954
- Reduce the complexity of the `{System, App, *}Label` APIs.
## Solution
For the sake of brevity I will only refer to `SystemLabel`, but everything applies to all of the other label types as well.
- Add `SystemLabelId`, a lightweight, `copy` struct.
- Convert custom types into `SystemLabelId` using the trait `SystemLabel`.
## Changelog
- String literals implement `SystemLabel` for now, but this should be changed with #4409 .
## Migration Guide
- Any previous use of `Box<dyn SystemLabel>` should be replaced with `SystemLabelId`.
- `AsSystemLabel` trait has been modified.
- No more output generics.
- Method `as_system_label` now returns `SystemLabelId`, removing an unnecessary level of indirection.
- If you *need* a label that is determined at runtime, you can use `Box::leak`. Not recommended.
## Questions for later
* Should we generate a `Debug` impl along with `#[derive(*Label)]`?
* Should we rename `as_str()`?
* Should we remove the extra derives (such as `Hash`) from builtin `*Label` types?
* Should we automatically derive types like `Clone, Copy, PartialEq, Eq`?
* More-ergonomic comparisons between `Label` and `LabelId`.
* Move `Dyn{Eq, Hash,Clone}` somewhere else.
* Some API to make interning dynamic labels easier.
* Optimize string representation
* Empty string for unit structs -- no debug info but faster comparisons
* Don't show enum types -- same tradeoffs as asbove.
# Objective
- Slight documentation tweak to make it more clear that `FloatOrd` also implements `Hash` and `Eq`, not just `Ord`.
- I know that it does show that Hash is implemented in the docs, but I had missed it after reading the description and assuming it didn't do it, so hopefully this helps other people who might miss it like I did. :)
## Solution
- Just mention in the Hash and Eq implementation in the docstring.
# Objective
`SAFETY` comments are meant to be placed before `unsafe` blocks and should contain the reasoning of why in this case the usage of unsafe is okay. This is useful when reading the code because it makes it clear which assumptions are required for safety, and makes it easier to spot possible unsoundness holes. It also forces the code writer to think of something to write and maybe look at the safety contracts of any called unsafe methods again to double-check their correct usage.
There's a clippy lint called `undocumented_unsafe_blocks` which warns when using a block without such a comment.
## Solution
- since clippy expects `SAFETY` instead of `SAFE`, rename those
- add `SAFETY` comments in more places
- for the last remaining 3 places, add an `#[allow()]` and `// TODO` since I wasn't comfortable enough with the code to justify their safety
- add ` #![warn(clippy::undocumented_unsafe_blocks)]` to `bevy_ecs`
### Note for reviewers
The first commit only renames `SAFETY` to `SAFE` so it doesn't need a thorough review.
cb042a416e..55cef2d6fa is the diff for all other changes.
### Safety comments where I'm not too familiar with the code
774012ece5/crates/bevy_ecs/src/entity/mod.rs (L540-L546)774012ece5/crates/bevy_ecs/src/world/entity_ref.rs (L249-L252)
### Locations left undocumented with a `TODO` comment
5dde944a30/crates/bevy_ecs/src/schedule/executor_parallel.rs (L196-L199)5dde944a30/crates/bevy_ecs/src/world/entity_ref.rs (L287-L289)5dde944a30/crates/bevy_ecs/src/world/entity_ref.rs (L413-L415)
Co-authored-by: Jakob Hellermann <hellermann@sipgate.de>
# Summary
This method strips a long type name like `bevy::render:📷:PerspectiveCameraBundle` down into the bare type name (`PerspectiveCameraBundle`). This is generally useful utility method, needed by #4299 and #5121.
As a result:
- This method was moved to `bevy_utils` for easier reuse.
- The legibility and robustness of this method has been significantly improved.
- Harder test cases have been added.
This change was split out of #4299 to unblock it and make merging / reviewing the rest of those changes easier.
## Changelog
- added `bevy_utils::get_short_name`, which strips the path from a type name for convenient display.
- removed the `TypeRegistry::get_short_name` method. Use the function in `bevy_utils` instead.
# Objective
- Update hashbrown to 0.12
## Solution
- Replace #4004
- As the 0.12 is already in Bevy dependency tree, it shouldn't be an issue to update
- The exception for the 0.11 should be removed once https://github.com/zakarumych/gpu-descriptor/pull/21 is merged and released
- Also removed a few exceptions that weren't needed anymore
# Objective
Fixes#4353. Fixes#4431. Picks up fixes for a panic for `gilrs` when `getGamepads()` is not available.
## Solution
Update the `gilrs` to `v0.9.0`. Changelog can be seen here: dba36f9186
EDIT: Updated `uuid` to 1.1 to avoid duplicate dependencies. Added `nix`'s two dependencies as exceptions until `rodio` updates their deps.
# Objective
We have some macros that are public but only used internally for now. They fail on user's code due to the use of crate names like `bevy_utils`, while the user only has `bevy::utils`. There are two affected macros.
- `bevy_utils::define_label`: it may be useful in user's code for defining custom kinds of label traits (this is why I made this PR).
- `bevy_asset::load_internal_asset`: not useful currently due to limitations of the debug asset server, but this may change in the future.
## Solution
We can make them work by using `$crate` instead of names of their own crates, which can refer to the macro's defining crate regardless of the user's setup. Even though our objective is rather low-priority here, the solution adds no maintenance cost so it is still worthwhile.
# Objective
Reduce from scratch build time.
## Solution
Reduce the size of the critical path by removing dependencies between crates where not necessary. For `cargo check --no-default-features` this reduced build time from ~51s to ~45s. For some commits I am not completely sure if the tradeoff between build time reduction and convenience caused by the commit is acceptable. If not, I can drop them.
# Objective
Reduce the catch-all grab-bag of functionality in bevy_core by moving FloatOrd to bevy_utils.
A step in addressing #2931 and splitting bevy_core into more specific locations.
## Solution
Move FloatOrd into bevy_utils. Fix the compile errors.
As a result, bevy_core_pipeline, bevy_pbr, bevy_sprite, bevy_text, and bevy_ui no longer depend on bevy_core (they were only using it for `FloatOrd` previously).
# Objective
- Debug logs are useful in release builds, but `tracing` logs are hard-capped (`release_max_level_info`) at the `info` level by `bevy_utils`.
## Solution
- This PR simply removes the limit in `bevy_utils` with no further actions.
- If any out-of-the box performance regressions arise, the steps to enable this `tracing` feature should be documented in a user guide in the future.
This PR closes#4069 and closes#1206.
## Alternatives considered
- Instruct the user to build with `debug-assertions` enabled: this is just a workaround, as it obviously enables all `debug-assertions` that affect more than logging itself.
- Re-exporting the feature from `tracing` and enabling it by default: I believe it just adds complexity and confusion, the `tracing` feature can also be re-enabled with one line in userland.
---
## Changelog
### Fixed
- Log level is not hard capped at `info` for release builds anymore.
## Migration Guide
- Maximum log levels for release builds is not enforced by Bevy anymore, to omit "debug" and "trace" level logs entirely from release builds, `tracing` must be added as a dependency with its `release_max_level_info` feature enabled in `Cargo.toml`. (`tracing = { version = "0.1", features = ["release_max_level_info"] }`)
