-
Notifications
You must be signed in to change notification settings - Fork 66
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Global Task Pools #54
Conversation
# Objective Fixes #3183. Requiring a `&TaskPool` parameter is sort of meaningless if the only correct one is to use the one provided by `Res<ComputeTaskPool>` all the time. ## Solution Have `QueryState` save a clone of the `ComputeTaskPool` which is used for all `par_for_each` functions. ~~Adds a small overhead of the internal `Arc` clone as a part of the startup, but the ergonomics win should be well worth this hardly-noticable overhead.~~ Updated the docs to note that it will panic the task pool is not present as a resource. # Future Work If bevyengine/rfcs#54 is approved, we can replace these resource lookups with a static function call instead to get the `ComputeTaskPool`. --- ## Changelog Removed: The `task_pool` parameter of `Query(State)::par_for_each(_mut)`. These calls will use the `World`'s `ComputeTaskPool` resource instead. ## Migration Guide The `task_pool` parameter for `Query(State)::par_for_each(_mut)` has been removed. Remove these parameters from all calls to these functions. Before: ```rust fn parallel_system( task_pool: Res<ComputeTaskPool>, query: Query<&MyComponent>, ) { query.par_for_each(&task_pool, 32, |comp| { ... }); } ``` After: ```rust fn parallel_system(query: Query<&MyComponent>) { query.par_for_each(32, |comp| { ... }); } ``` If using `Query(State)` outside of a system run by the scheduler, you may need to manually configure and initialize a `ComputeTaskPool` as a resource in the `World`.
# Objective Fixes #4791. `ParallelExecutor` inserts a default `CompteTaskPool` if there isn't one stored as a resource, including when it runs on a different world. When spawning the render sub-app, the main world's `ComputeTaskPool` is not cloned and inserted into the render app's, which causes a second `ComputeTaskPool` with the default configuration to be spawned. This results in an excess number of threads being spawned. ## Solution Copy the task pools from the main world to the subapps upon creating them. ## Alternative An alternative to this would be to make the task pools global, as seen in #2250 or bevyengine/rfcs#54.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
I'm convinced! The ergonomics wins are hard to ignore and given that this is pretty much the pattern adopted everywhere, it seems like a safe bet. Globals do create "conflict" problems and feel more than a little bit dirty. I generally don't want to embrace this pattern in Bevy, but task pools do seem like an exceptional case. As we've recently discussed, I do think that the unified task pool in bevyengine/bevy#4740 is the right general direction, but I'm comfortable merging this in the current state as it feels orthogonal from a design standpoint.
# Objective Fixes bevyengine#3183. Requiring a `&TaskPool` parameter is sort of meaningless if the only correct one is to use the one provided by `Res<ComputeTaskPool>` all the time. ## Solution Have `QueryState` save a clone of the `ComputeTaskPool` which is used for all `par_for_each` functions. ~~Adds a small overhead of the internal `Arc` clone as a part of the startup, but the ergonomics win should be well worth this hardly-noticable overhead.~~ Updated the docs to note that it will panic the task pool is not present as a resource. # Future Work If bevyengine/rfcs#54 is approved, we can replace these resource lookups with a static function call instead to get the `ComputeTaskPool`. --- ## Changelog Removed: The `task_pool` parameter of `Query(State)::par_for_each(_mut)`. These calls will use the `World`'s `ComputeTaskPool` resource instead. ## Migration Guide The `task_pool` parameter for `Query(State)::par_for_each(_mut)` has been removed. Remove these parameters from all calls to these functions. Before: ```rust fn parallel_system( task_pool: Res<ComputeTaskPool>, query: Query<&MyComponent>, ) { query.par_for_each(&task_pool, 32, |comp| { ... }); } ``` After: ```rust fn parallel_system(query: Query<&MyComponent>) { query.par_for_each(32, |comp| { ... }); } ``` If using `Query(State)` outside of a system run by the scheduler, you may need to manually configure and initialize a `ComputeTaskPool` as a resource in the `World`.
# Objective Fixes bevyengine#4791. `ParallelExecutor` inserts a default `CompteTaskPool` if there isn't one stored as a resource, including when it runs on a different world. When spawning the render sub-app, the main world's `ComputeTaskPool` is not cloned and inserted into the render app's, which causes a second `ComputeTaskPool` with the default configuration to be spawned. This results in an excess number of threads being spawned. ## Solution Copy the task pools from the main world to the subapps upon creating them. ## Alternative An alternative to this would be to make the task pools global, as seen in bevyengine#2250 or bevyengine/rfcs#54.
# Objective Fixes bevyengine#3183. Requiring a `&TaskPool` parameter is sort of meaningless if the only correct one is to use the one provided by `Res<ComputeTaskPool>` all the time. ## Solution Have `QueryState` save a clone of the `ComputeTaskPool` which is used for all `par_for_each` functions. ~~Adds a small overhead of the internal `Arc` clone as a part of the startup, but the ergonomics win should be well worth this hardly-noticable overhead.~~ Updated the docs to note that it will panic the task pool is not present as a resource. # Future Work If bevyengine/rfcs#54 is approved, we can replace these resource lookups with a static function call instead to get the `ComputeTaskPool`. --- ## Changelog Removed: The `task_pool` parameter of `Query(State)::par_for_each(_mut)`. These calls will use the `World`'s `ComputeTaskPool` resource instead. ## Migration Guide The `task_pool` parameter for `Query(State)::par_for_each(_mut)` has been removed. Remove these parameters from all calls to these functions. Before: ```rust fn parallel_system( task_pool: Res<ComputeTaskPool>, query: Query<&MyComponent>, ) { query.par_for_each(&task_pool, 32, |comp| { ... }); } ``` After: ```rust fn parallel_system(query: Query<&MyComponent>) { query.par_for_each(32, |comp| { ... }); } ``` If using `Query(State)` outside of a system run by the scheduler, you may need to manually configure and initialize a `ComputeTaskPool` as a resource in the `World`.
# Objective Fixes bevyengine#4791. `ParallelExecutor` inserts a default `CompteTaskPool` if there isn't one stored as a resource, including when it runs on a different world. When spawning the render sub-app, the main world's `ComputeTaskPool` is not cloned and inserted into the render app's, which causes a second `ComputeTaskPool` with the default configuration to be spawned. This results in an excess number of threads being spawned. ## Solution Copy the task pools from the main world to the subapps upon creating them. ## Alternative An alternative to this would be to make the task pools global, as seen in bevyengine#2250 or bevyengine/rfcs#54.
RENDERED
This proposes changes to the way the three main TaskPools are initialized and used in Bevy.