Build on wgpu 30

Two majors, and the renderer is under everything else that is left to
extract, so it goes before the slices that would otherwise be written twice.
`image` 0.25.6 -> 0.25.10 rides along; `winit` stays on 0.30.12, since 0.31
is only a prerelease.

What the API asked for:

- An instance now takes the display it will present on, and GLES on Wayland
  needs it -- so the window the surface is made from is handed over with it.
- `get_current_texture` returns a `CurrentSurfaceTexture` rather than a
  `Result`, which replaces an `unwrap` that would have panicked on a resize
  or an occluded window: reconfigure when the surface is outdated, lost or
  suboptimal, and skip the frame when there is nothing to draw into.
- Presenting moved to the queue, after `pre_present_notify` as before.
- Bind group layouts and vertex buffer layouts are sparse now, so a pipeline
  states `Some(layout)` per slot.

Verified with the three render checks and `tests/draw_cost.rs`: identical
pictures, and 33.6/167/587/2855 us per frame at 8/64/256/1024 layers against
33.3/161/588/2903 on wgpu 28.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>

src/default/mod.rs

@@ -1,6 +1,10 @@
 use crate::prelude::*;
 use arboard::Clipboard;
-use std::{marker::PhantomData, sync::Arc, time::Instant};
+use std::{
+    marker::{PhantomData, Sized},
+    sync::Arc,
+    time::Instant,
+};
 use winit::{
     event::{Ime, WindowEvent},
     event_loop::{ActiveEventLoop, EventLoopProxy},

src/default/render.rs

@@ -24,6 +24,8 @@ impl UiRenderer {
     pub fn draw(&mut self) {
         let output = match self.surface.get_current_texture() {
             CurrentSurfaceTexture::Success(texture) => texture,
+            // Still drawable; the surface has just changed under us, and
+            // reconfiguring is what the next frame wants rather than this one.
             CurrentSurfaceTexture::Suboptimal(texture) => {
                 self.surface.configure(&self.device, &self.config);
                 texture
@@ -32,6 +34,7 @@ impl UiRenderer {
                 self.surface.configure(&self.device, &self.config);
                 return;
             }
+            // Nothing to draw into this frame.
             CurrentSurfaceTexture::Timeout
             | CurrentSurfaceTexture::Occluded
             | CurrentSurfaceTexture::Validation => return,
@@ -78,6 +81,8 @@ impl UiRenderer {
     pub fn new(window: Arc<Window>) -> Self {
         let size = window.inner_size();
 
+        // The display handle is what GLES needs to present on Wayland, and it
+        // has to be the one the surface is made from.
         let instance = Instance::new(InstanceDescriptor {
             backends: Backends::PRIMARY,
             display: Some(Box::new(window.clone())),

src/widget/position/mod.rs

@@ -4,7 +4,7 @@ mod max_size;
 mod offset;
 mod pad;
 mod scroll;
-mod set_size;
+mod sized;
 mod span;
 mod stack;
 
@@ -14,6 +14,6 @@ pub use max_size::*;
 pub use offset::*;
 pub use pad::*;
 pub use scroll::*;
-pub use set_size::*;
+pub use sized::*;
 pub use span::*;
 pub use stack::*;

src/widget/position/sized.rs

@@ -1,12 +1,12 @@
 use crate::prelude::*;
 
-pub struct SetSize {
+pub struct Sized {
     pub inner: StrongWidget,
     pub x: Option<Len>,
     pub y: Option<Len>,
 }
 
-impl SetSize {
+impl Sized {
     fn apply_to_outer(&self, ctx: &mut SizeCtx) {
         if let Some(x) = self.x {
             ctx.outer.x.select_len(x.apply_rest());
@@ -17,7 +17,7 @@ impl SetSize {
     }
 }
 
-impl Widget for SetSize {
+impl Widget for Sized {
     fn draw(&mut self, painter: &mut Painter) {
         painter.widget(&self.inner);
     }

src/widget/ptr.rs

@@ -1,5 +1,5 @@
 use crate::prelude::*;
-use std::marker::Unsize;
+use std::marker::{Sized, Unsize};
 
 pub struct WidgetPtr {
     pub inner: Option<StrongWidget>,

src/widget/text/build.rs

@@ -1,5 +1,5 @@
 use crate::prelude::*;
-use std::marker::PhantomData;
+use std::marker::{PhantomData, Sized};
 
 pub struct TextBuilder<State, O = TextOutput, H: WidgetOption<State> = ()> {
     pub content: String,

src/widget/trait_fns.rs

@@ -31,9 +31,9 @@ widget_trait! {
         }
     }
 
-    fn sized(self, size: impl Into<Size>) -> impl WidgetFn<Rsc, SetSize> {
+    fn sized(self, size: impl Into<Size>) -> impl WidgetFn<Rsc, Sized> {
         let size = size.into();
-        move |state| SetSize {
+        move |state| Sized {
             inner: self.add_strong(state),
             x: Some(size.x),
             y: Some(size.y),
@@ -58,18 +58,18 @@ widget_trait! {
         }
     }
 
-    fn width(self, len: impl Into<Len>) -> impl WidgetFn<Rsc, SetSize> {
+    fn width(self, len: impl Into<Len>) -> impl WidgetFn<Rsc, Sized> {
         let len = len.into();
-        move |state| SetSize {
+        move |state| Sized {
             inner: self.add_strong(state),
             x: Some(len),
             y: None,
         }
     }
 
-    fn height(self, len: impl Into<Len>) -> impl WidgetFn<Rsc, SetSize> {
+    fn height(self, len: impl Into<Len>) -> impl WidgetFn<Rsc, Sized> {
         let len = len.into();
-        move |state| SetSize {
+        move |state| Sized {
             inner: self.add_strong(state),
             x: None,
             y: Some(len),

tests/draw_cost.rs

@@ -4,19 +4,20 @@
 //!
 //!     cargo test --release --test draw_cost -- --ignored --nocapture
 //!
-//! Wall time is the wrong number to read for anything under a few percent --
+//! **Read the instruction count, not the clock.** Wall time here swings by 2x
-//! it varied by 2x between runs of one unchanged binary where instructions
+//! between runs of one binary on this machine -- more under `cargo test` than
-//! retired varied by 0.1%. Count those instead:
+//! run directly -- while instructions retired are stable to 0.1%:
 //!
 //!     perf stat -e instructions:u target/release/.../draw_cost-* --ignored
 //!
-//! That is how `PrimitiveRender` was measured against a match in the renderer:
+//! Measured that way on 2026-09-13, drawing each primitive through its own
-//! 6 instructions per list drawn, against the ~5,400 wgpu spends recording
+//! `PrimitiveRender` rather than a match in the renderer costs **6
-//! one.
+//! instructions per list drawn**, which is 0.1% of a frame at both 256 and
+//! 1024 layers. Recording one list into the pass costs wgpu ~5,400.
 //!
-//! The instance is leaked deliberately. A Vulkan loader may unload the driver
+//! The instance is leaked on purpose. Dropping the last one makes the Vulkan
-//! when the last one drops, which can fault as a thread that used it exits --
+//! loader unload Mesa's ICD, which faults when a thread that touched Vulkan
-//! and every test runs on a spawned thread.
+//! exits -- and libtest runs every test on a spawned thread.
 
 use std::time::Instant;
 
@@ -29,13 +30,13 @@ use wgpu::{Color as GpuColor, *};
 
 const SIZE: u32 = 1024;
 const FRAMES: u32 = 200;
-/// Reported as the best of this many batches, since the mean moves by more
+/// Reported as the best of this many batches. The mean moves by 15% between
-/// than the thing being measured.
+/// runs on this machine, which is more than the thing being measured.
 const BATCHES: u32 = 8;
 
 fn gpu() -> Option<(Device, Queue)> {
-    // Probed rather than assumed: there may be no Vulkan adapter, and GL is
+    // Probed rather than assumed: this machine's Vulkan device comes and goes,
-    // what is left when there is not.
+    // and GL is what is left when it is gone.
     let all = Instance::new(InstanceDescriptor::new_without_display_handle());
     let instance = match pollster::block_on(all.request_adapter(&RequestAdapterOptions::default()))
     {