Auto merge of #136507 - matthiaskrgr:rollup-uzwv9mo, r=matthiaskrgr

Rollup of 8 pull requests

Successful merges:

 - #136289 (OnceCell & OnceLock docs: Using (un)initialized consistently)
 - #136299 (Ignore NLL boring locals in polonius diagnostics)
 - #136411 (Omit argument names from function pointers that do not have argument names)
 - #136430 (Use the type-level constant value `ty::Value` where needed)
 - #136476 (Remove generic `//@ ignore-{wasm,wasm32,emscripten}` in tests)
 - #136484 (Notes on types/traits used for in-memory query caching)
 - #136493 (platform-support: document CPU baseline for x86-32 targets)
 - #136498 (Update books)

r? `@ghost`
`@rustbot` modify labels: rollup

1 files changed, 13 insertions, 0 deletions

diff --git a/compiler/rustc_data_structures/src/vec_cache.rs b/compiler/rustc_data_structures/src/vec_cache.rs
index eb251b587c8..2ff60ab7f36 100644
--- a/compiler/rustc_data_structures/src/vec_cache.rs
+++ b/compiler/rustc_data_structures/src/vec_cache.rs
@@ -206,6 +206,19 @@ impl SlotIndex {
     }
 }
 
+/// In-memory cache for queries whose keys are densely-numbered IDs
+/// (e.g `CrateNum`, `LocalDefId`), and can therefore be used as indices
+/// into a dense vector of cached values.
+///
+/// (As of [#124780] the underlying storage is not an actual `Vec`, but rather
+/// a series of increasingly-large buckets, for improved performance when the
+/// parallel frontend is using multiple threads.)
+///
+/// Each entry in the cache stores the query's return value (`V`), and also
+/// an associated index (`I`), which in practice is a `DepNodeIndex` used for
+/// query dependency tracking.
+///
+/// [#124780]: https://github.com/rust-lang/rust/pull/124780
 pub struct VecCache<K: Idx, V, I> {
     // Entries per bucket:
     // Bucket  0:       4096 2^12