Weak-GPU hardware (Raspberry Pi 4 / v3dv, Haswell / hasvk) exposes a real Vulkan adapter, so the old binary wgpu-or-software detection handed it the slowest possible path: femtovg-wgpu over an immature Vulkan driver. Mesa's GL path is the fast one on that class of hardware. - Three-tier chain: wgpu -> femtovg GL -> software. Auto-detect classifies adapters (weak-driver markers incl. hasvk, integrated-on-arm, tiny texture limits); adapters only reachable through wgpu's GL backend never select the wgpu tier (the renderer masks Backends::GL out, so detection must too). renderer-femtovg (GL) feature now compiled on Linux. - wgpu tier requests PowerPreference::LowPower (WGPU_POWER_PREF still wins). - Per-window swapchain alpha in the vendored femtovg renderer: only the miniplayer gets a transparent (blended) swapchain; the main window keeps an Opaque one, sparing a full-window compositor blend per frame. - Settings > Appearance gains a Renderer row (Linux only): Auto (default) / GPU / GPU compatibility / Software, persisted in ui_prefs.json, applied on restart. QBZ_RENDERER env still overrides (now also accepts gl|gles|femtovg). - Auto-revert sentinel: a non-auto choice is armed before backend init and disarmed after the window survives 5s; if the previous start died before that, the setting reverts to Auto with an explanatory toast, so a bad renderer pick can never lock the user out of Settings. - Developer tools > Diagnostics now reports the active renderer, why it was chosen, and the enumerated GPU adapters (panel + markdown report). cc
493 lines
18 KiB
Rust
493 lines
18 KiB
Rust
// Copyright © SixtyFPS GmbH <info@slint.dev>
|
|
// SPDX-License-Identifier: GPL-3.0-only OR LicenseRef-Slint-Royalty-free-2.0 OR LicenseRef-Slint-Software-3.0
|
|
|
|
use std::{cell::RefCell, pin::Pin, rc::Rc};
|
|
|
|
use i_slint_core::platform::PlatformError;
|
|
use i_slint_core::renderer::RendererSealed;
|
|
use i_slint_core::{api::PhysicalSize as PhysicalWindowSize, graphics::RequestedGraphicsAPI};
|
|
|
|
use crate::{FemtoVGRenderer, GraphicsBackend, WindowSurface};
|
|
|
|
use wgpu_28 as wgpu;
|
|
|
|
/// QBZ vendor patch: per-window swapchain alpha opt-in. When true, a
|
|
/// `WGPUBackend` constructed via `new_suspended` captures a preference for a
|
|
/// non-opaque composite alpha mode (needed for the borderless transparent
|
|
/// miniplayer); when false (default) its surfaces stay Opaque/Auto so the
|
|
/// compositor does not alpha-blend the whole main window every frame. The app
|
|
/// sets this from its winit window-attributes hook, which runs on the event
|
|
/// loop thread right before the window ADAPTER — and therefore this backend —
|
|
/// is created. The value is captured PER BACKEND at construction rather than
|
|
/// read in `set_surface`: `set_surface` re-runs at every winit re-realization
|
|
/// (Wayland destroys the window on hide, and even the first realization is
|
|
/// deferred to a later event-loop iteration), long after the hook fired, so
|
|
/// reading the global there would leak the last-created window's preference
|
|
/// into whichever window is (re)realized next.
|
|
static SURFACE_PREFERS_TRANSPARENT: core::sync::atomic::AtomicBool =
|
|
core::sync::atomic::AtomicBool::new(false);
|
|
|
|
/// QBZ vendor patch: mark whether `WGPUBackend`s constructed from now on
|
|
/// prefer a transparent (non-opaque) composite alpha mode for their surfaces.
|
|
pub fn set_surface_prefers_transparent(transparent: bool) {
|
|
SURFACE_PREFERS_TRANSPARENT.store(transparent, core::sync::atomic::Ordering::Relaxed);
|
|
}
|
|
|
|
fn preferred_alpha_mode(
|
|
modes: &[wgpu::CompositeAlphaMode],
|
|
transparent: bool,
|
|
) -> wgpu::CompositeAlphaMode {
|
|
if transparent {
|
|
[
|
|
wgpu::CompositeAlphaMode::PreMultiplied,
|
|
wgpu::CompositeAlphaMode::PostMultiplied,
|
|
wgpu::CompositeAlphaMode::Inherit,
|
|
]
|
|
.into_iter()
|
|
.find(|mode| modes.contains(mode))
|
|
.unwrap_or(wgpu::CompositeAlphaMode::Auto)
|
|
} else if modes.contains(&wgpu::CompositeAlphaMode::Opaque) {
|
|
wgpu::CompositeAlphaMode::Opaque
|
|
} else {
|
|
// Auto is always valid (wgpu resolves it to Opaque or Inherit) and is
|
|
// upstream Slint's behavior, so this is the graceful fallback.
|
|
wgpu::CompositeAlphaMode::Auto
|
|
}
|
|
}
|
|
|
|
pub struct WGPUBackend {
|
|
instance: RefCell<Option<wgpu::Instance>>,
|
|
device: RefCell<Option<wgpu::Device>>,
|
|
queue: RefCell<Option<wgpu::Queue>>,
|
|
surface_config: RefCell<Option<wgpu::SurfaceConfiguration>>,
|
|
surface: RefCell<Option<wgpu::Surface<'static>>>,
|
|
/// QBZ vendor patch: captured from `SURFACE_PREFERS_TRANSPARENT` at
|
|
/// construction — this backend's window keeps the same alpha preference
|
|
/// across every surface re-creation.
|
|
prefers_transparent: bool,
|
|
}
|
|
|
|
pub struct WGPUWindowSurface {
|
|
surface_texture: wgpu::SurfaceTexture,
|
|
}
|
|
|
|
impl WindowSurface<femtovg::renderer::WGPURenderer> for WGPUWindowSurface {
|
|
fn render_output(&self) -> impl Into<femtovg::renderer::WGPURenderOutput> {
|
|
&self.surface_texture.texture
|
|
}
|
|
}
|
|
|
|
impl GraphicsBackend for WGPUBackend {
|
|
type Renderer = femtovg::renderer::WGPURenderer;
|
|
type WindowSurface = WGPUWindowSurface;
|
|
const NAME: &'static str = "WGPU";
|
|
|
|
fn new_suspended() -> Self {
|
|
Self {
|
|
instance: Default::default(),
|
|
device: Default::default(),
|
|
queue: Default::default(),
|
|
surface_config: Default::default(),
|
|
surface: Default::default(),
|
|
prefers_transparent: SURFACE_PREFERS_TRANSPARENT
|
|
.load(core::sync::atomic::Ordering::Relaxed),
|
|
}
|
|
}
|
|
|
|
fn clear_graphics_context(&self) {
|
|
self.surface_config.borrow_mut().take();
|
|
self.surface.borrow_mut().take();
|
|
self.queue.borrow_mut().take();
|
|
self.device.borrow_mut().take();
|
|
}
|
|
|
|
fn begin_surface_rendering(
|
|
&self,
|
|
) -> Result<Self::WindowSurface, Box<dyn std::error::Error + Send + Sync>> {
|
|
let surface = self.surface.borrow();
|
|
let surface = surface.as_ref().unwrap();
|
|
let frame = match surface.get_current_texture() {
|
|
Ok(texture) => texture,
|
|
Err(wgpu::SurfaceError::Timeout) => surface.get_current_texture()?,
|
|
// Outdated or lost: re-configure and try again
|
|
Err(_) => {
|
|
let mut device = self.device.borrow_mut();
|
|
let device = device.as_mut().unwrap();
|
|
|
|
surface.configure(device, self.surface_config.borrow().as_ref().unwrap());
|
|
surface.get_current_texture()?
|
|
}
|
|
};
|
|
Ok(WGPUWindowSurface { surface_texture: frame })
|
|
}
|
|
|
|
fn submit_commands(&self, commands: <Self::Renderer as femtovg::Renderer>::CommandBuffer) {
|
|
self.queue.borrow().as_ref().unwrap().submit(commands);
|
|
}
|
|
|
|
fn present_surface(
|
|
&self,
|
|
surface: Self::WindowSurface,
|
|
) -> Result<(), Box<dyn std::error::Error + Send + Sync>> {
|
|
surface.surface_texture.present();
|
|
Ok(())
|
|
}
|
|
|
|
#[cfg(feature = "unstable-wgpu-28")]
|
|
fn with_graphics_api<R>(
|
|
&self,
|
|
callback: impl FnOnce(Option<i_slint_core::api::GraphicsAPI<'_>>) -> R,
|
|
) -> Result<R, i_slint_core::platform::PlatformError> {
|
|
let instance = self.instance.borrow().clone();
|
|
let device = self.device.borrow().clone();
|
|
let queue = self.queue.borrow().clone();
|
|
if let (Some(instance), Some(device), Some(queue)) = (instance, device, queue) {
|
|
Ok(callback(Some(i_slint_core::graphics::create_graphics_api_wgpu_28(
|
|
instance, device, queue,
|
|
))))
|
|
} else {
|
|
Ok(callback(None))
|
|
}
|
|
}
|
|
|
|
#[cfg(not(feature = "unstable-wgpu-28"))]
|
|
fn with_graphics_api<R>(
|
|
&self,
|
|
callback: impl FnOnce(Option<i_slint_core::api::GraphicsAPI<'_>>) -> R,
|
|
) -> Result<R, i_slint_core::platform::PlatformError> {
|
|
Ok(callback(None))
|
|
}
|
|
|
|
fn resize(
|
|
&self,
|
|
width: std::num::NonZeroU32,
|
|
height: std::num::NonZeroU32,
|
|
) -> Result<(), Box<dyn std::error::Error + Send + Sync>> {
|
|
// Try to get hold of the wgpu types, but if we receive the resize event while suspended, ignore it.
|
|
let mut surface_config = self.surface_config.borrow_mut();
|
|
let Some(surface_config) = surface_config.as_mut() else { return Ok(()) };
|
|
let mut device = self.device.borrow_mut();
|
|
let Some(device) = device.as_mut() else { return Ok(()) };
|
|
let mut surface = self.surface.borrow_mut();
|
|
let Some(surface) = surface.as_mut() else { return Ok(()) };
|
|
|
|
// Prefer FIFO modes over possible Mailbox setting for frame pacing and better energy efficiency.
|
|
surface_config.present_mode = wgpu::PresentMode::AutoVsync;
|
|
surface_config.width = width.get();
|
|
surface_config.height = height.get();
|
|
|
|
surface.configure(device, surface_config);
|
|
Ok(())
|
|
}
|
|
}
|
|
|
|
impl FemtoVGRenderer<WGPUBackend> {
|
|
pub fn set_surface(
|
|
&self,
|
|
surface_target: impl Into<i_slint_core::graphics::wgpu_28::SurfaceTarget>,
|
|
size: PhysicalWindowSize,
|
|
requested_graphics_api: Option<RequestedGraphicsAPI>,
|
|
) -> Result<(), Box<dyn std::error::Error + Send + Sync>> {
|
|
let (instance, adapter, device, queue, surface) =
|
|
i_slint_core::graphics::wgpu_28::init_instance_adapter_device_queue_surface(
|
|
surface_target,
|
|
requested_graphics_api,
|
|
/* rendering artifacts :( */
|
|
wgpu::Backends::GL,
|
|
)?;
|
|
|
|
let mut surface_config =
|
|
surface.get_default_config(&adapter, size.width, size.height).unwrap();
|
|
|
|
let swapchain_capabilities = surface.get_capabilities(&adapter);
|
|
let swapchain_format = swapchain_capabilities
|
|
.formats
|
|
.iter()
|
|
.find(|f| {
|
|
matches!(f, wgpu::TextureFormat::Rgba8Unorm | wgpu::TextureFormat::Bgra8Unorm)
|
|
})
|
|
.copied()
|
|
.unwrap_or_else(|| swapchain_capabilities.formats[0]);
|
|
surface_config.format = swapchain_format;
|
|
let wants_transparent = self.graphics_backend.prefers_transparent;
|
|
surface_config.alpha_mode =
|
|
preferred_alpha_mode(&swapchain_capabilities.alpha_modes, wants_transparent);
|
|
i_slint_core::debug_log!(
|
|
"[qbz-slint] femtovg-wgpu surface alpha modes: {:?}; transparent={}; selected: {:?}",
|
|
swapchain_capabilities.alpha_modes,
|
|
wants_transparent,
|
|
surface_config.alpha_mode
|
|
);
|
|
surface.configure(&device, &surface_config);
|
|
|
|
*self.graphics_backend.instance.borrow_mut() = Some(instance.clone());
|
|
*self.graphics_backend.device.borrow_mut() = Some(device.clone());
|
|
*self.graphics_backend.queue.borrow_mut() = Some(queue.clone());
|
|
*self.graphics_backend.surface_config.borrow_mut() = Some(surface_config);
|
|
*self.graphics_backend.surface.borrow_mut() = Some(surface);
|
|
|
|
let wgpu_renderer = femtovg::renderer::WGPURenderer::new(device, queue);
|
|
let femtovg_canvas = femtovg::Canvas::new_with_text_context(
|
|
wgpu_renderer,
|
|
crate::font_cache::FONT_CACHE.with(|cache| cache.borrow().text_context.clone()),
|
|
)
|
|
.unwrap();
|
|
|
|
let canvas = Rc::new(RefCell::new(femtovg_canvas));
|
|
self.reset_canvas(canvas);
|
|
Ok(())
|
|
}
|
|
}
|
|
|
|
struct TextureWindowSurface {
|
|
render_output: femtovg::renderer::WGPURenderOutput,
|
|
}
|
|
|
|
impl WindowSurface<femtovg::renderer::WGPURenderer> for TextureWindowSurface {
|
|
fn render_output(&self) -> impl Into<femtovg::renderer::WGPURenderOutput> {
|
|
self.render_output.clone()
|
|
}
|
|
}
|
|
|
|
struct WgpuTextureBackend {
|
|
instance: wgpu::Instance,
|
|
device: wgpu::Device,
|
|
queue: wgpu::Queue,
|
|
render_output: RefCell<Option<femtovg::renderer::WGPURenderOutput>>,
|
|
}
|
|
|
|
impl GraphicsBackend for WgpuTextureBackend {
|
|
type Renderer = femtovg::renderer::WGPURenderer;
|
|
type WindowSurface = TextureWindowSurface;
|
|
const NAME: &'static str = "WGPU Texture";
|
|
|
|
fn new_suspended() -> Self {
|
|
panic!("Suspended backend not supported for WgpuTextureBackend (requires device/queue)");
|
|
}
|
|
|
|
fn clear_graphics_context(&self) {
|
|
// Nothing to clear here, we don't own the device/queue/texture
|
|
}
|
|
|
|
fn begin_surface_rendering(
|
|
&self,
|
|
) -> Result<Self::WindowSurface, Box<dyn std::error::Error + Send + Sync>> {
|
|
let render_output =
|
|
self.render_output.borrow().clone().ok_or("No texture set for rendering")?;
|
|
Ok(TextureWindowSurface { render_output })
|
|
}
|
|
|
|
fn submit_commands(&self, commands: <Self::Renderer as femtovg::Renderer>::CommandBuffer) {
|
|
self.queue.submit(commands);
|
|
}
|
|
|
|
fn present_surface(
|
|
&self,
|
|
_surface: Self::WindowSurface,
|
|
) -> Result<(), Box<dyn std::error::Error + Send + Sync>> {
|
|
// No presentation needed - the caller owns the texture and handles presenting it
|
|
Ok(())
|
|
}
|
|
|
|
#[cfg(feature = "unstable-wgpu-28")]
|
|
fn with_graphics_api<R>(
|
|
&self,
|
|
callback: impl FnOnce(Option<i_slint_core::api::GraphicsAPI<'_>>) -> R,
|
|
) -> Result<R, i_slint_core::platform::PlatformError> {
|
|
Ok(callback(Some(i_slint_core::graphics::create_graphics_api_wgpu_28(
|
|
self.instance.clone(),
|
|
self.device.clone(),
|
|
self.queue.clone(),
|
|
))))
|
|
}
|
|
|
|
#[cfg(not(feature = "unstable-wgpu-28"))]
|
|
fn with_graphics_api<R>(
|
|
&self,
|
|
callback: impl FnOnce(Option<i_slint_core::api::GraphicsAPI<'_>>) -> R,
|
|
) -> Result<R, i_slint_core::platform::PlatformError> {
|
|
Ok(callback(None))
|
|
}
|
|
|
|
fn resize(
|
|
&self,
|
|
_width: std::num::NonZeroU32,
|
|
_height: std::num::NonZeroU32,
|
|
) -> Result<(), Box<dyn std::error::Error + Send + Sync>> {
|
|
// No resize needed - texture size is determined by the texture passed to render_to_texture
|
|
Ok(())
|
|
}
|
|
}
|
|
|
|
/// Use the FemtoVG renderer with WGPU when implementing a custom Slint platform where you want the scene to be rendered
|
|
/// into a WGPU texture. The rendering is done using the [FemtoVG](https://github.com/femtovg/femtovg) library.
|
|
pub struct FemtoVGWGPURenderer(FemtoVGRenderer<WgpuTextureBackend>);
|
|
|
|
impl FemtoVGWGPURenderer {
|
|
/// Creates a new FemtoVGWGPURenderer.
|
|
///
|
|
/// The `instance`, `device` and `queue` are the WGPU resources used for rendering.
|
|
/// These are also provided to [`Window::set_rendering_notifier()`](i_slint_core::api::Window::set_rendering_notifier) callbacks via [`GraphicsAPI::WGPU28`](i_slint_core::api::GraphicsAPI::WGPU28).
|
|
pub fn new(
|
|
instance: wgpu::Instance,
|
|
device: wgpu::Device,
|
|
queue: wgpu::Queue,
|
|
) -> Result<Self, PlatformError> {
|
|
let backend = WgpuTextureBackend {
|
|
instance,
|
|
device: device.clone(),
|
|
queue: queue.clone(),
|
|
render_output: RefCell::new(None),
|
|
};
|
|
let renderer = FemtoVGRenderer::new_internal(backend);
|
|
|
|
let wgpu_renderer = femtovg::renderer::WGPURenderer::new(device, queue);
|
|
let femtovg_canvas = femtovg::Canvas::new_with_text_context(
|
|
wgpu_renderer,
|
|
crate::font_cache::FONT_CACHE.with(|cache| cache.borrow().text_context.clone()),
|
|
)
|
|
.map_err(|e| format!("Failed to create femtovg canvas: {:?}", e))?;
|
|
|
|
let canvas = Rc::new(RefCell::new(femtovg_canvas));
|
|
renderer.reset_canvas(canvas);
|
|
Ok(Self(renderer))
|
|
}
|
|
|
|
/// Render the scene to the given texture view, with the specified size and format.
|
|
pub fn render_to_texture_view(
|
|
&self,
|
|
texture_view: &wgpu::TextureView,
|
|
width: u32,
|
|
height: u32,
|
|
format: wgpu::TextureFormat,
|
|
) -> Result<(), PlatformError> {
|
|
*self.0.graphics_backend.render_output.borrow_mut() =
|
|
Some(femtovg::renderer::WGPURenderOutput {
|
|
view: texture_view.clone(),
|
|
width,
|
|
height,
|
|
format,
|
|
});
|
|
let result = self.0.render();
|
|
*self.0.graphics_backend.render_output.borrow_mut() = None;
|
|
result
|
|
}
|
|
|
|
/// Render the scene to the given texture.
|
|
/// This is a convenience method that creates a texture view for the entire texture and calls `render_to_texture_view`.
|
|
pub fn render_to_texture(&self, texture: &wgpu::Texture) -> Result<(), PlatformError> {
|
|
let size = texture.size();
|
|
self.render_to_texture_view(
|
|
&texture.create_view(&wgpu::TextureViewDescriptor::default()),
|
|
size.width,
|
|
size.height,
|
|
texture.format(),
|
|
)
|
|
}
|
|
}
|
|
|
|
#[doc(hidden)]
|
|
impl RendererSealed for FemtoVGWGPURenderer {
|
|
fn text_size(
|
|
&self,
|
|
text_item: Pin<&dyn i_slint_core::item_rendering::RenderString>,
|
|
item_rc: &i_slint_core::items::ItemRc,
|
|
max_width: Option<i_slint_core::lengths::LogicalLength>,
|
|
text_wrap: i_slint_core::items::TextWrap,
|
|
) -> i_slint_core::lengths::LogicalSize {
|
|
self.0.text_size(text_item, item_rc, max_width, text_wrap)
|
|
}
|
|
|
|
fn char_size(
|
|
&self,
|
|
text_item: Pin<&dyn i_slint_core::item_rendering::HasFont>,
|
|
item_rc: &i_slint_core::items::ItemRc,
|
|
ch: char,
|
|
) -> i_slint_core::lengths::LogicalSize {
|
|
self.0.char_size(text_item, item_rc, ch)
|
|
}
|
|
|
|
fn font_metrics(
|
|
&self,
|
|
font_request: i_slint_core::graphics::FontRequest,
|
|
) -> i_slint_core::items::FontMetrics {
|
|
self.0.font_metrics(font_request)
|
|
}
|
|
|
|
fn text_input_byte_offset_for_position(
|
|
&self,
|
|
text_input: Pin<&i_slint_core::items::TextInput>,
|
|
item_rc: &i_slint_core::items::ItemRc,
|
|
pos: i_slint_core::lengths::LogicalPoint,
|
|
) -> usize {
|
|
self.0.text_input_byte_offset_for_position(text_input, item_rc, pos)
|
|
}
|
|
|
|
fn text_input_cursor_rect_for_byte_offset(
|
|
&self,
|
|
text_input: Pin<&i_slint_core::items::TextInput>,
|
|
item_rc: &i_slint_core::items::ItemRc,
|
|
byte_offset: usize,
|
|
) -> i_slint_core::lengths::LogicalRect {
|
|
self.0.text_input_cursor_rect_for_byte_offset(text_input, item_rc, byte_offset)
|
|
}
|
|
|
|
fn register_font_from_memory(
|
|
&self,
|
|
data: &'static [u8],
|
|
) -> Result<(), Box<dyn std::error::Error>> {
|
|
self.0.register_font_from_memory(data)
|
|
}
|
|
|
|
fn register_font_from_path(
|
|
&self,
|
|
path: &std::path::Path,
|
|
) -> Result<(), Box<dyn std::error::Error>> {
|
|
self.0.register_font_from_path(path)
|
|
}
|
|
|
|
fn default_font_size(&self) -> i_slint_core::lengths::LogicalLength {
|
|
self.0.default_font_size()
|
|
}
|
|
|
|
fn set_rendering_notifier(
|
|
&self,
|
|
callback: Box<dyn i_slint_core::api::RenderingNotifier>,
|
|
) -> Result<(), i_slint_core::api::SetRenderingNotifierError> {
|
|
self.0.set_rendering_notifier(callback)
|
|
}
|
|
|
|
fn free_graphics_resources(
|
|
&self,
|
|
component: i_slint_core::item_tree::ItemTreeRef,
|
|
items: &mut dyn Iterator<Item = Pin<i_slint_core::items::ItemRef<'_>>>,
|
|
) -> Result<(), PlatformError> {
|
|
self.0.free_graphics_resources(component, items)
|
|
}
|
|
|
|
fn set_window_adapter(&self, window_adapter: &Rc<dyn i_slint_core::window::WindowAdapter>) {
|
|
self.0.set_window_adapter(window_adapter)
|
|
}
|
|
|
|
fn window_adapter(&self) -> Option<Rc<dyn i_slint_core::window::WindowAdapter>> {
|
|
RendererSealed::window_adapter(&self.0)
|
|
}
|
|
|
|
fn resize(&self, size: i_slint_core::api::PhysicalSize) -> Result<(), PlatformError> {
|
|
self.0.resize(size)
|
|
}
|
|
|
|
fn take_snapshot(
|
|
&self,
|
|
) -> Result<
|
|
i_slint_core::graphics::SharedPixelBuffer<i_slint_core::graphics::Rgba8Pixel>,
|
|
PlatformError,
|
|
> {
|
|
self.0.take_snapshot()
|
|
}
|
|
|
|
fn supports_transformations(&self) -> bool {
|
|
self.0.supports_transformations()
|
|
}
|
|
}
|