tree_arena: Use hashbrown::HashMap as drop-in replacement to optimi…

…ze tree-access (#774)

As noted in
#772 (review),
this seems to be a really cheap but quite effective optimization, as we
have `hashbrown` already in our dependency tree.
I've noticed speedups of the rewrite passes of up to >100%.
On average about 20-50% and is also definitely noticeable (smoother and
less latency).

Another quite interesting observation with my (admittedly not really
reproducible) performance tests,
is that the safe tree arena is actually faster than the unsafe version
(roughly 10-30%), and tells yet another time that benches are really
important while optimizing...

Rough overview of my performance tests/benches:
I've mostly hovered over 10000 buttons in the mason example while
testing, but every second or so a new button is spawned and also this is
quite a bit faster.
I did a quick'n dirty `Instant::elapsed` in the event loop (so rendering
is not included, but I think there's also a speedup as we're iterating
the tree).

Cargo.toml

@@ -117,6 +117,7 @@ peniko = "0.2.0"
 winit = "0.30.4"
 tracing = { version = "0.1.40", default-features = false }
 smallvec = "1.13.2"
+hashbrown = "0.15.2"
 dpi = "0.1.1"
 image = { version = "0.25.2", default-features = false }
 web-time = "1.1.0"

tree_arena/Cargo.toml

@@ -21,6 +21,7 @@ targets = []
 
 
 [dependencies]
+hashbrown.workspace = true
 
 [features]
 # This crate contains two implementations of a tree for use in masonry, one safe and the other unsafe.

tree_arena/src/tree_arena_safe.rs

@@ -13,7 +13,7 @@
 
 use super::NodeId;
 
-use std::collections::HashMap;
+use hashbrown::HashMap;
 
 #[derive(Debug)]
 struct TreeNode<T> {

tree_arena/src/tree_arena_unsafe.rs

@@ -5,7 +5,8 @@
 use super::NodeId;
 
 use std::cell::UnsafeCell;
-use std::collections::HashMap;
+
+use hashbrown::HashMap;
 
 #[derive(Debug)]
 struct TreeNode<T> {