Mirrors/bevy
2
0
Fork 0
mirror of https://github.com/bevyengine/bevy synced 2024-11-22 12:43:34 +00:00
Code Issues Projects Releases Packages Wiki Activity

Update bevy_window::PresentMode to mirror wgpu::PresentMode (#9230)

Browse source 
# Objective

- Update `bevy_window::PresentMode` to mirror `wgpu::PresentMode`, Fixes
#9151.

## Solution

Add `bevy_window::PresentMode::FifoRelaxed` to
`bevy_window::PresentMode`, add documents.

---

## Changelog

### Added
- Add `bevy_window::PresentMode::FifoRelaxed` to
`bevy_window::PresentMode`.


## Migration Guide

- Handle `bevy_window::PresentMode::FifoRelaxed` when tweaking window
present mode manually.
This commit is contained in:
VitalyR 2023-07-22 02:40:08 +08:00 committed by GitHub
parent ad011d0455
commit 6093385b31
No known key found for this signature in database
GPG key ID: 4AEE18F83AFDEB23
2 changed files with 56 additions and 17 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

1
crates/bevy_render/src/view/window.rs
View file

@ -266,6 +266,7 @@ pub fn prepare_windows(
usage: wgpu::TextureUsages::RENDER_ATTACHMENT,
present_mode: match window.present_mode {
PresentMode::Fifo => wgpu::PresentMode::Fifo,
PresentMode::FifoRelaxed => wgpu::PresentMode::FifoRelaxed,
PresentMode::Mailbox => wgpu::PresentMode::Mailbox,
PresentMode::Immediate => wgpu::PresentMode::Immediate,
PresentMode::AutoVsync => wgpu::PresentMode::AutoVsync,

72
crates/bevy_window/src/window.rs
View file

@ -801,6 +801,7 @@ pub enum MonitorSelection {
/// [`Immediate`] or [`Mailbox`] will panic if not supported by the platform.
///
/// [`Fifo`]: PresentMode::Fifo
/// [`FifoRelaxed`]: PresentMode::FifoRelaxed
/// [`Immediate`]: PresentMode::Immediate
/// [`Mailbox`]: PresentMode::Mailbox
/// [`AutoVsync`]: PresentMode::AutoVsync
@ -819,30 +820,67 @@ pub enum PresentMode {
/// Chooses FifoRelaxed -> Fifo based on availability.
///
/// Because of the fallback behavior, it is supported everywhere.
AutoVsync = 0,
AutoVsync = 0, // NOTE: The explicit ordinal values mirror wgpu.
/// Chooses Immediate -> Mailbox -> Fifo (on web) based on availability.
///
/// Because of the fallback behavior, it is supported everywhere.
AutoNoVsync = 1,
/// The presentation engine does **not** wait for a vertical blanking period and
/// the request is presented immediately. This is a low-latency presentation mode,
/// but visible tearing may be observed. Not optimal for mobile.
/// Presentation frames are kept in a First-In-First-Out queue approximately 3 frames
/// long. Every vertical blanking period, the presentation engine will pop a frame
/// off the queue to display. If there is no frame to display, it will present the same
/// frame again until the next vblank.
///
/// Selecting this variant will panic if not supported, it is preferred to use
/// [`PresentMode::AutoNoVsync`].
Immediate = 2,
/// The presentation engine waits for the next vertical blanking period to update
/// the current image, but frames may be submitted without delay. This is a low-latency
/// presentation mode and visible tearing will **not** be observed. Not optimal for mobile.
/// When a present command is executed on the gpu, the presented image is added on the queue.
///
/// Selecting this variant will panic if not supported, it is preferred to use
/// [`PresentMode::AutoNoVsync`].
Mailbox = 3,
/// The presentation engine waits for the next vertical blanking period to update
/// the current image. The framerate will be capped at the display refresh rate,
/// corresponding to the `VSync`. Tearing cannot be observed. Optimal for mobile.
/// No tearing will be observed.
///
/// Calls to get_current_texture will block until there is a spot in the queue.
///
/// Supported on all platforms.
///
/// If you don't know what mode to choose, choose this mode. This is traditionally called "Vsync On".
#[default]
Fifo = 4, // NOTE: The explicit ordinal values mirror wgpu.
Fifo = 2,
/// Presentation frames are kept in a First-In-First-Out queue approximately 3 frames
/// long. Every vertical blanking period, the presentation engine will pop a frame
/// off the queue to display. If there is no frame to display, it will present the
/// same frame until there is a frame in the queue. The moment there is a frame in the
/// queue, it will immediately pop the frame off the queue.
///
/// When a present command is executed on the gpu, the presented image is added on the queue.
///
/// Tearing will be observed if frames last more than one vblank as the front buffer.
///
/// Calls to get_current_texture will block until there is a spot in the queue.
///
/// Supported on AMD on Vulkan.
///
/// This is traditionally called "Adaptive Vsync"
FifoRelaxed = 3,
/// Presentation frames are not queued at all. The moment a present command
/// is executed on the GPU, the presented image is swapped onto the front buffer
/// immediately.
///
/// Tearing can be observed.
///
/// Supported on most platforms except older DX12 and Wayland.
///
/// This is traditionally called "Vsync Off".
Immediate = 4,
/// Presentation frames are kept in a single-frame queue. Every vertical blanking period,
/// the presentation engine will pop a frame from the queue. If there is no frame to display,
/// it will present the same frame again until the next vblank.
///
/// When a present command is executed on the gpu, the frame will be put into the queue.
/// If there was already a frame in the queue, the new frame will _replace_ the old frame
/// on the queue.
///
/// No tearing will be observed.
///
/// Supported on DX11/12 on Windows 10, NVidia on Vulkan and Wayland on Vulkan.
///
/// This is traditionally called "Fast Vsync"
Mailbox = 5,
}
/// Specifies how the alpha channel of the textures should be handled during compositing, for a [`Window`].