The advantage of making these docs is mostly in pointing out that these
functions all make new allocations and copy/clone/move the source into them.
These docs are on the function, and not the `impl` block, to avoid showing
the "[+] show undocumented items" button.
CC #51430
BTreeMap: disentangle Drop implementation from IntoIter
No longer require every `BTreeMap` to dig up its last leaf edge before dying. This speeds up the `clone_` benchmarks by 25% for normal keys and values (far less for huge values).
r? `@Mark-Simulacrum`
Optimize Vec::retain
Use `copy_non_overlapping` instead of `swap` to reduce memory writes, like what we've done in #44355 and `String::retain`.
#48065 already tried to do this optimization but it is reverted in #67300 due to bad codegen of `DrainFilter::drop`.
This PR re-implement the drop-then-move approach. I did a [benchmark](https://gist.github.com/oxalica/3360eec9376f22533fcecff02798b698) on small-no-drop, small-need-drop, large-no-drop elements with different predicate functions. It turns out that the new implementation is >20% faster in average for almost all cases. Only 2/24 cases are slower by 3% and 5%. See the link above for more detail.
I think regression in may-panic cases is due to drop-guard preventing some optimization. If it's permitted to leak elements when predicate function of element's `drop` panic, the new implementation should be almost always faster than current one.
I'm not sure if we should leak on panic, since there is indeed an issue (#52267) complains about it before.
Make Vec::split_at_spare_mut public
This PR introduces a new method to the public API, under
`vec_split_at_spare` feature gate:
```rust
impl<T, A: Allocator> impl Vec<T, A> {
pub fn split_at_spare_mut(&mut self) -> (&mut [T], &mut [MaybeUninit<T>]);
}
```
The method returns 2 slices, one slice references the content of the vector,
and the other references the remaining spare capacity.
The method was previously implemented while adding `Vec::extend_from_within` in #79015,
and used to implement `Vec::spare_capacity_mut` (as the later is just a
subset of former one).
See also previous [discussion in `Vec::spare_capacity_mut` tracking issue](https://github.com/rust-lang/rust/issues/75017#issuecomment-770381335).
## Unresolved questions
- [ ] Should we consider changing the name? `split_at_spare_mut` doesn't seem like an intuitive name
- [ ] Should we deprecate `Vec::spare_capacity_mut`? Any usecase of `Vec::spare_capacity_mut` can be replaced with `Vec::split_at_spare_mut` (but not vise-versa)
r? `@KodrAus`
Add `Box::into_inner`.
This adds a `Box::into_inner` method to the `Box` type. <del>I actually suggest deprecating the compiler magic of `*b` if this gets stablized in the future.</del>
r? `@m-ou-se`
Rollup of 11 pull requests
Successful merges:
- #72209 (Add checking for no_mangle to unsafe_code lint)
- #80732 (Allow Trait inheritance with cycles on associated types take 2)
- #81697 (Add "every" as a doc alias for "all".)
- #81826 (Prefer match over combinators to make some Box methods inlineable)
- #81834 (Resolve typedef in HashMap lldb pretty-printer only if possible)
- #81841 ([rustbuild] Output rustdoc-json-types docs )
- #81849 (Expand the docs for ops::ControlFlow a bit)
- #81876 (parser: Fix panic in 'const impl' recovery)
- #81882 (⬆️ rust-analyzer)
- #81888 (Fix pretty printer macro_rules with semicolon.)
- #81896 (Remove outdated comment in windows' mutex.rs)
Failed merges:
r? `@ghost`
`@rustbot` modify labels: rollup
Prefer match over combinators to make some Box methods inlineable
Hopefully this patch would make two snippets generated identical code: <https://rust.godbolt.org/z/fjrj4E>.
BTree: remove Ord bound where it is absent elsewhere
Some btree methods don't really need an Ord bound and don't have one, while some methods that more obviously don't need it, do have one.
An example of the former is `iter`, even though it explicitly exposes the work of the Ord implementation (["sorted by key"](https://doc.rust-lang.org/std/collections/struct.BTreeMap.html#method.iter) - but I'm not suggesting it should have the Ord bound). An example of the latter is `new`, which doesn't involve any keys whatsoever.
BTreeMap: make Ord bound explicit, compile-test its absence
Most `BTreeMap` and `BTreeSet` members are subject to an `Ord` bound but a fair number of methods are not. To better convey and perhaps later tune the `Ord` bound, make it stand out in individual `where` clauses, instead of once far away at the beginning of an `impl` block. This PR does not introduce or remove any bounds.
Also adds compilation test cases checking that the bound doesn't creep in unintended on the historically unbounded methods.
This commit introduces a new method to the public API, under
`vec_split_at_spare` feature gate:
```rust
impl<T, A: Allocator> impl Vec<T, A> {
pub fn split_at_spare_mut(&mut self) -> (&mut [T], &mut [MaybeUninit<T>]);
}
```
The method returns 2 slices, one slice references the content of the vector,
and the other references the remaining spare capacity.
The method was previously implemented while adding `Vec::extend_from_within`,
and used to implement `Vec::spare_capacity_mut` (as the later is just a
subset of former one).
Add doc aliases for "delete"
This patch adds doc aliases for "delete". The added aliases are supposed to reference usages `delete` in other programming languages.
- `HashMap::remove`, `BTreeMap::remove` -> `Map#delete` and `delete` keyword in JavaScript.
- `HashSet::remove`, `BTreeSet::remove` -> `Set#delete` in JavaScript.
- `mem::drop` -> `delete` keyword in C++.
- `fs::remove_file`, `fs::remove_dir`, `fs::remove_dir_all`-> `File#delete` in Java, `File#delete` and `Dir#delete` in Ruby.
Before this change, searching for "delete" in documentation returned no results.
add `Vec::extend_from_within` method under `vec_extend_from_within` feature gate
Implement <https://github.com/rust-lang/rfcs/pull/2714>
### tl;dr
This PR adds a `extend_from_within` method to `Vec` which allows copying elements from a range to the end:
```rust
#![feature(vec_extend_from_within)]
let mut vec = vec![0, 1, 2, 3, 4];
vec.extend_from_within(2..);
assert_eq!(vec, [0, 1, 2, 3, 4, 2, 3, 4]);
vec.extend_from_within(..2);
assert_eq!(vec, [0, 1, 2, 3, 4, 2, 3, 4, 0, 1]);
vec.extend_from_within(4..8);
assert_eq!(vec, [0, 1, 2, 3, 4, 2, 3, 4, 0, 1, 4, 2, 3, 4]);
```
### Implementation notes
Originally I've copied `@Shnatsel's` [implementation](690742a0de/src/lib.rs (L74)) with some minor changes to support other ranges:
```rust
pub fn append_from_within<R>(&mut self, src: R)
where
T: Copy,
R: RangeBounds<usize>,
{
let len = self.len();
let Range { start, end } = src.assert_len(len);;
let count = end - start;
self.reserve(count);
unsafe {
// This is safe because `reserve()` above succeeded,
// so `self.len() + count` did not overflow usize
ptr::copy_nonoverlapping(
self.get_unchecked(src.start),
self.as_mut_ptr().add(len),
count,
);
self.set_len(len + count);
}
}
```
But then I've realized that this duplicates most of the code from (private) `Vec::append_elements`, so I've used it instead.
Then I've applied `@KodrAus` suggestions from https://github.com/rust-lang/rust/pull/79015#issuecomment-727200852.
Implement <https://github.com/rust-lang/rfcs/pull/2714>, changes from the RFC:
- Rename the method `append_from_within` => `extend_from_within`
- Loose :Copy bound => :Clone
- Specialize in case of :Copy
This commit also adds `Vec::split_at_spare` private method and use it to implement
`Vec::spare_capacity_mut` and `Vec::extend_from_within`. This method returns 2
slices - initialized elements (same as `&mut vec[..]`) and uninitialized but
allocated space (same as `vec.spare_capacity_mut()`).
Remove const_in_array_repeat
Fixes#80371. Fixes#81315. Fixes#80767. Fixes#75682.
I thought there might be some issue with `Repeats(_, 0)`, but if you increase the items in the array it still ICEs. I'm not sure if this is the best fix but it does fix the given issue.
This patch adds doc aliases for "delete". The added aliases are
supposed to reference usages `delete` in other programming
languages.
- `HashMap::remove`, `BTreeMap::remove` -> `Map#delete` and `delete`
keyword in JavaScript.
- `HashSet::remove`, `BTreeSet::remove` -> `Set#delete` in JavaScript.
- `mem::drop` -> `delete` keyword in C++.
- `fs::remove_file`, `fs::remove_dir`, `fs::remove_dir_all`
-> `File#delete` in Java, `File#delete` and `Dir#delete` in Ruby.
Before this change, searching for "delete" in documentation
returned no results.
Stabilize by-value `[T; N]` iterator `core::array::IntoIter`
Tracking issue: https://github.com/rust-lang/rust/issues/65798
This is unblocked now that `min_const_generics` has been stabilized in https://github.com/rust-lang/rust/pull/79135.
This PR does *not* include the corresponding `IntoIterator` impl, which is https://github.com/rust-lang/rust/pull/65819. Instead, an iterator can be constructed through the `new` method.
`new` would become unnecessary when `IntoIterator` is implemented and might be deprecated then, although it will stay stable.
Stabilize raw ref macros
This stabilizes `raw_ref_macros` (https://github.com/rust-lang/rust/issues/73394), which is possible now that https://github.com/rust-lang/rust/issues/74355 is fixed.
However, as I already said in https://github.com/rust-lang/rust/issues/73394#issuecomment-751342185, I am not particularly happy with the current names of the macros. So I propose we also change them, which means I am proposing to stabilize the following in `core::ptr`:
```rust
pub macro const_addr_of($e:expr) {
&raw const $e
}
pub macro mut_addr_of($e:expr) {
&raw mut $e
}
```
The macro name change means we need another round of FCP. Cc `````@rust-lang/libs`````
Fixes#73394
Add `core::stream::Stream`
[[Tracking issue: #79024](https://github.com/rust-lang/rust/issues/79024)]
This patch adds the `core::stream` submodule and implements `core::stream::Stream` in accordance with [RFC2996](https://github.com/rust-lang/rfcs/pull/2996). The RFC hasn't been merged yet, but as requested by the libs team in https://github.com/rust-lang/rfcs/pull/2996#issuecomment-725696389 I'm filing this PR to get the ball rolling.
## Documentatation
The docs in this PR have been adapted from [`std::iter`](https://doc.rust-lang.org/std/iter/index.html), [`async_std::stream`](https://docs.rs/async-std/1.7.0/async_std/stream/index.html), and [`futures::stream::Stream`](https://docs.rs/futures/0.3.8/futures/stream/trait.Stream.html). Once this PR lands my plan is to follow this up with PRs to add helper methods such as `stream::repeat` which can be used to document more of the concepts that are currently missing. That will allow us to cover concepts such as "infinite streams" and "laziness" in more depth.
## Feature gate
The feature gate for `Stream` is `stream_trait`. This matches the `#[lang = "future_trait"]` attribute name. The intention is that only the APIs defined in RFC2996 will use this feature gate, with future additions such as `stream::repeat` using their own feature gates. This is so we can ensure a smooth path towards stabilizing the `Stream` trait without needing to stabilize all the APIs in `core::stream` at once. But also don't start expanding the API until _after_ stabilization, as was the case with `std::future`.
__edit:__ the feature gate has been changed to `async_stream` to match the feature gate proposed in the RFC.
## Conclusion
This PR introduces `core::stream::{Stream, Next}` and re-exports it from `std` as `std::stream::{Stream, Next}`. Landing `Stream` in the stdlib has been a mult-year process; and it's incredibly exciting for this to finally happen!
---
r? `````@KodrAus`````
cc/ `````@rust-lang/wg-async-foundations````` `````@rust-lang/libs`````
BTreeMap: prevent tree from ever being owned by non-root node
This introduces a new marker type, `Dying`, which is used to note trees which are in the process of deallocation. On such trees, some fields may be in an inconsistent state as we are deallocating the tree. Unfortunately, there's not a great way to express conditional unsafety, so the methods for traversal can cause UB if not invoked correctly, but not marked as such. This is not a regression from the previous state, but rather isolates the destructive methods to solely being called on the dying state.
Trying to shrink_to greater than capacity should be no-op
Per the discussion in https://github.com/rust-lang/rust/issues/56431, `shrink_to` shouldn't panic if you try to make a vector shrink to a capacity greater than its current capacity.
BTreeMap: test all borrowing interfaces and test more chaotic order behavior
Inspired by #81169, test what happens if you mess up order of the type with which you search (as opposed to the key type).
r? `@Mark-Simulacrum`
BTreeMap: bring back the key slice for immutable lookup
Pave the way for binary search, by reverting a bit of #73971, which banned `keys` for misbehaving while it was defined for every `BorrowType`. Adding some `debug_assert`s along the way.
r? `@Mark-Simulacrum`
mark raw_vec::ptr with inline
when a lot of vectors is used in a enum as in the example in #66617 if this function is not inlined and multiple cgus is used this results in huge compile times. with this fix the compile time is 6s from minutes for the example in #66617. I did not have the patience to wait for it to compile for more then 3 min.
Add doc aliases for memory allocations
This patch adds doc aliases for various C allocation functions, making it possible to search for the C-equivalent of a function and finding the (safe) Rust counterpart:
- `Vec::with_capacity` / `Box::new` / `vec!` -> alloc + malloc, allocates memory
- `Box::new_zeroed` -> calloc, allocates zeroed-out memory
- `Vec::{reserve,reserve_exact,try_reserve_exact,shrink_to_fit,shrink_to}` -> realloc, reallocates a previously allocated slice of memory
It's worth noting that `Vec::new` does not allocate, so we don't link to it. Instead people are probably looking for `Vec::with_capacity` or `vec!`. I hope this will allow people comfortable with the system allocation APIs to make it easier to find what they may be looking for.
Thanks!
Enforce statically that `MIN_NON_ZERO_CAP` is calculated at compile time
Previously, it would usually get computed by LLVM, but this enforces it. This removes the need for the comment saying "LLVM is smart enough".
I don't expect this to make a performance difference, but I do think it makes the performance properties easier to reason about.
Fix broken links with `--document-private-items` in the standard library
As it was suggested in #81037 `SpecFromIter` is not
in the scope and therefore we get a warning when we try to
do document private intems in `rust/library/alloc/`.
This addresses #81037 by adding the trait in the scope as ```@jyn514```
suggested and also adding an `allow(unused_imports)` flag so that
the compiler does not complain, Since the trait is not used
per se in the code, it's just needed to have properly documented
docs.
Improve grammar in documentation of format strings
The docs previously were
* using some weird `<` and `>` around some nonterminals
* _correct me if these **did** have any meaning_
* using of a (not explicitly defined) `text` nonterminal that didn’t explicitly disallow productions containing `'{'` or `'}'`
* incorrect in not allowing for `x?` and `X?` productions of `type`
* unnecessarily ambiguous, both
* allowing `type` to be `''`, and
* using an optional `[type]`
* using inconsistent underscore/hyphenation style between `format_string` and `format_spec` vs `maybe-format`
_Rendered:_
_(current docs: https://doc.rust-lang.org/nightly/std/fmt/#syntax)_
```@rustbot``` modify labels: T-doc
Visualize vector while differentiating between stack and heap.
Inspired by cheats.rs, as this is probably the first place beginner go,
they could understand stack and heap, length and capacity with this. Not
sure if adding this means we should add to other places too.
Superseeds #76066
BTreeMap: prefer bulk_steal functions over specialized ones
The `steal_` functions (apart from their return value) are basically specializations of the more general `bulk_steal_` functions. This PR removes the specializations. The library/alloc benchmarks say this is never slower and up to 6% faster.
r? ``@Mark-Simulacrum``
As it was suggested in #81037 `SpecFromIter` is not
in the scope and therefore (even it should fail),
we get a warning when we try do document private
intems in `rust/library/alloc/`.
This fixes#81037 by adding the trait in the scope
and also adding an `allow(unused_imports)` flag so that
the compiler does not complain, Since the trait is not used
per se in the code, it's just needed to have properly documented
docs.
Don't make tools responsible for checking unknown and renamed lints
Previously, clippy (and any other tool emitting lints) had to have their
own separate UNKNOWN_LINTS pass, because the compiler assumed any tool
lint could be valid. Now, as long as any lint starting with the tool
prefix exists, the compiler will warn when an unknown lint is present.
This may interact with the unstable `tool_lint` feature, which I don't entirely understand, but it will take the burden off those external tools to add their own lint pass, which seems like a step in the right direction to me.
- Don't mark `ineffective_unstable_trait_impl` as an internal lint
- Use clippy's more advanced lint suggestions
- Deprecate the `UNKNOWN_CLIPPY_LINTS` pass (and make it a no-op)
- Say 'unknown lint `clippy::x`' instead of 'unknown lint x'
This is tested by existing clippy tests. When https://github.com/rust-lang/rust/pull/80527 merges, it will also be tested in rustdoc tests. AFAIK there is no way to test this with rustc directly.
Force vec![] to expression position only
r? `@oli-obk`
I went with the lazy way of only changing what broke. I moved the test to ui/macros because the diagnostics no longer give suggestions.
Closes#61933
BTreeMap: expose new_internal function and sanitize from_new_internal
`new_internal` is the functional core of the imperative `push_internal_level`, and `from_new_internal` can easily do a proper job instead of returning a half-baked node.
r? `@Mark-Simulacrum`
Re-stabilize Weak::as_ptr and friends for unsized T
As per [T-lang consensus](https://hackmd.io/7r3_is6uTz-163fsOV8Vfg), this uses a branch to handle the dangling case. The discussed optimization of only doing the branch in the T: ?Sized case is left for a followup patch, as doing so is not trivial (as it requires specialization) and not _obviously_ better (as it requires using `wrapping_offset` rather than `offset` more).
<details><summary>Basically said optimization</summary>
Specialize on `T: Sized`:
```rust
fn as_ptr(&self) -> *const T {
if [ T is Sized ] || !is_dangling(ptr) {
(ptr as *mut T).set_ptr_value( (ptr as *mut u8).wrapping_offset(data_offset) )
} else {
ptr::null()
}
}
fn from_raw(*const T) -> Self {
if [ T is Sized ] || !ptr.is_null() {
let ptr = (ptr as *mut RcBox).set_ptr_value( (ptr as *mut u8).wrapping_offset(-data_offset) );
Weak { ptr }
} else {
Weak::new()
}
}
```
(but with more `set_ptr_value` to avoid `Sized` restrictions and maintain metadata.)
Written in this fashion, this is not a correctness-critical specialization (i.e. so long as `[ T is Sized ]` is false for unsized `T`, it can be `rand()` for sized `T` without breaking correctness), but it's still touchy, so I'd rather do it in another PR with separate review.
---
</details>
This effectively reverts #80422 and re-establishes #74160. T-libs [previously signed off](https://github.com/rust-lang/rust/pull/74160#issuecomment-660539373) on this stable API change in #74160.
Clarify what the effects of a 'logic error' are
This clarifies what a 'logic error' is (which is a term used to describe what happens if you put things in a hash table or btree and then use something like a refcell to break the internal ordering). This tries to be as vague as possible, as we don't really want to promise what happens, except "bad things, but not UB". This was discussed in #80657
It's not an internal lint:
- It's not in the rustc::internal lint group
- It's on unconditionally, because it actually lints `staged_api`, not
the compiler
This fixes a bug where `#[deny(rustc::internal)]` would warn that
`rustc::internal` was an unknown lint.
Remove unreachable panics from VecDeque::{front/back}[_mut]
`VecDeque`'s `front`, `front_mut`, `back` and `back_mut` methods are implemented in terms of the index operator, which causes these functions to contain [unreachable panic calls](https://rust.godbolt.org/z/MTnq1o).
This PR reimplements these methods in terms of `get[_mut]` instead.
Remove unstable deprecated Vec::remove_item
Closes#40062
The `Vec::remove_item` method was deprecated in `1.46.0` (in August of 2020). This PR now removes that unstable method entirely.
Deprecate atomic::spin_loop_hint in favour of hint::spin_loop
For https://github.com/rust-lang/rust/issues/55002
We wanted to leave `atomic::spin_loop_hint` alone when stabilizing `hint::spin_loop` so folks had some time to migrate. This now deprecates `atomic_spin_loop_hint`.
Try to avoid locals when cloning into Box/Rc/Arc
For generic `T: Clone`, we can allocate an uninitialized box beforehand,
which gives the optimizer a chance to create the clone directly in the
heap. For `T: Copy`, we can go further and do a simple memory copy,
regardless of optimization level.
The same applies to `Rc`/`Arc::make_mut` when they must clone the data.
Stabilize split_inclusive
### Contents of this MR
This stabilises:
* `slice::split_inclusive`
* `slice::split_inclusive_mut`
* `str::split_inclusive`
Closes#72360.
### A possible concern
The proliferation of `split_*` methods is not particularly pretty. The existence of `split_inclusive` seems to invite the addition of `rsplit_inclusive`, `splitn_inclusive`, etc. We could instead have a more general API, along these kinds of lines maybe:
```
pub fn split_generic('a,P,H>(&'a self, pat: P, how: H) -> ...
where P: Pattern
where H: SplitHow;
pub fn split_generic_mut('a,P,H>(&'a mut self, pat: P, how: H) -> ...
where P: Pattern
where H: SplitHow;
trait SplitHow {
fn reverse(&self) -> bool;
fn inclusive -> bool;
fn limit(&self) -> Option<usize>;
}
pub struct SplitFwd;
...
pub struct SplitRevInclN(pub usize);
```
But maybe that is worse.
### Let us defer that? ###
This seems like a can of worms. I think we can defer opening it now; if and when we have something more general, these two methods can become convenience aliases. But I thought I would mention it so the lang API team can consider it and have an opinion.
As we did with `Box`, we can allocate an uninitialized `Rc` or `Arc`
beforehand, giving the optimizer a chance to skip the local value for
regular clones, or avoid any local altogether for `T: Copy`.
For generic `T: Clone`, we can allocate an uninitialized box beforehand,
which gives the optimizer a chance to create the clone directly in the
heap. For `T: Copy`, we can go further and do a simple memory copy,
regardless of optimization level.
BTreeMap: tougher checking on most uses of copy_nonoverlapping
Miri checks pointer provenance and destination, but we can check it in debug builds already.
Also, we can let Miri confirm we don't mistake imprints of moved keys and values as genuine.
r? `@Mark-Simulacrum`
As per T-lang consensus, this uses a branch to handle the dangling case.
The discussed optimization of only doing the branch in the T: ?Sized
case is left for a followup patch, as doing so is not trivial
(as it requires specialization for correctness, not just optimization).
remove allow(incomplete_features) from std
cc https://github.com/rust-lang/rust/pull/80349#issuecomment-753357123
> Now I am somewhat concerned that the standard library uses some of these features...
I think it is theoretically ok to use incomplete features in the standard library or the compiler if we know that there is an already working subset and we explicitly document what we have to be careful about. Though at that point it is probably better to try and split the incomplete feature into two separate ones, similar to `min_specialization`.
Will be interesting once `feature(const_evaluatable_checked)` works well enough to imo be used in the compiler but not yet well enough to be removed from `INCOMPLETE_FEATURES`.
r? `@RalfJung`
Remove many unnecessary manual link resolves from library
Now that #76934 has merged, we can remove a lot of these! E.g, this is
no longer necessary:
[`Vec<T>`]: Vec
cc `@jyn514`
The return of the GroupBy and GroupByMut iterators on slice
According to https://github.com/rust-lang/rfcs/pull/2477#issuecomment-742034372, I am opening this PR again, this time I implemented it in safe Rust only, it is therefore much easier to read and is completely safe.
This PR proposes to add two new methods to the slice, the `group_by` and `group_by_mut`. These two methods provide a way to iterate over non-overlapping sub-slices of a base slice that are separated by the predicate given by the user (e.g. `Partial::eq`, `|a, b| a.abs() < b.abs()`).
```rust
let slice = &[1, 1, 1, 3, 3, 2, 2, 2];
let mut iter = slice.group_by(|a, b| a == b);
assert_eq!(iter.next(), Some(&[1, 1, 1][..]));
assert_eq!(iter.next(), Some(&[3, 3][..]));
assert_eq!(iter.next(), Some(&[2, 2, 2][..]));
assert_eq!(iter.next(), None);
```
[An RFC](https://github.com/rust-lang/rfcs/pull/2477) was open 2 years ago but wasn't necessary.
Rollup of 9 pull requests
Successful merges:
- #78934 (refactor: removing library/alloc/src/vec/mod.rs ignore-tidy-filelength)
- #79479 (Add `Iterator::intersperse`)
- #80128 (Edit rustc_ast::ast::FieldPat docs)
- #80424 (Don't give an error when creating a file for the first time)
- #80458 (Some Promotion Refactoring)
- #80488 (Do not create dangling &T in Weak<T>::drop)
- #80491 (Miri: make size/align_of_val work for dangling raw ptrs)
- #80495 (Rename kw::Invalid -> kw::Empty)
- #80513 (Add regression test for #80062)
Failed merges:
r? `@ghost`
`@rustbot` modify labels: rollup
Do not create dangling &T in Weak<T>::drop
Since at this point all strong pointers have been dropped, the wrapped `T` has also been dropped. As such, creating a `&T` to the dropped place is negligent at best (language UB at worst). Since we have `Layout::for_value_raw` now, use that instead of `Layout::for_value` to avoid creating the `&T`.
This does have implications for custom (potentially thin) DSTs, though much less severe than those discussed in #80407. Specifically, one of two things has to be true:
- It has to be possible to use a `*const T` to a dropped (potentially custom, potentially thin) unsized tailed object to determine the layout (size/align) of the object. This is what is currently implemented (though with `&T` instead of `&T`). The validity of reading some location after it has been dropped is an open question IIUC (https://github.com/rust-lang/unsafe-code-guidelines/issues/188) (except when the whole type is `Copy`, per `drop_in_place`'s docs).
In this design, custom DSTs would get a `*mut T` and use that to return layout, and must be able to do so while in the "zombie" (post-drop, pre-free) state.
- `RcBox`/`ArcInner` compute and store layout eagerly, so that they don't have to ask the type for its layout after dropping it.
Importantly, this is already true today, as you can construct `Rc<DST>`, create a `Weak<DST>`, and drop the `Rc` before the `Weak`. This PR is a strict improvement over the status quo, and the above question about potentially thin DSTs will need to be resolved by any custom DST proposal.
Add "length" as doc alias to len methods
Currently when searching for `length` there are no results: https://doc.rust-lang.org/std/?search=length. This makes `len` methods appear when searching for `length`.
BTreeMap: clean up access to MaybeUninit arrays
Stop exposing and using immutable access to `MaybeUninit` slices when we need and have exclusive access to the tree.
r? `@Mark-Simulacrum`
BTreeMap: relax the explicit borrow rule to make code shorter and safer
Expressions like `.reborrow_mut().into_len_mut()` are annoyingly long, and kind of dangerous for the reason `reborrow_mut()` is unsafe. By relaxing the single rule, we no longer have to make an exception for functions with a `borrow` name and functions like `as_leaf_mut`. This is largely restoring the declaration style of the btree::node API about a year ago, but with more explanation and consistency.
r? `@Mark-Simulacrum`
Stabilize or_insert_with_key
Stabilizes the `or_insert_with_key` feature from https://github.com/rust-lang/rust/issues/71024. This allows inserting key-derived values when a `HashMap`/`BTreeMap` entry is vacant.
The difference between this and `.or_insert_with(|| ... )` is that this provides a reference to the key to the closure after it is moved with `.entry(key_being_moved)`, avoiding the need to copy or clone the key.
Fix overflow when converting ZST Vec to VecDeque
```rust
let v = vec![(); 100];
let queue = VecDeque::from(v);
println!("{:?}", queue);
```
This code will currently panic with a capacity overflow.
This PR resolves this issue and makes the code run fine.
Resolves#78532
Do not inline finish_grow
Fixes#78471.
Looking at libgkrust.a in Firefox, the sizes for the `gkrust.*.o` file is:
- 18584816 (text) 582418 (data) with unmodified master
- 17937659 (text) 582554 (data) with #72227 reverted
- 17968228 (text) 582858 (data) with `#[inline(never)]` on `grow_amortized` and `grow_exact`, but that has some performance consequences
- 17927760 (text) 582322 (data) with this change
So in terms of size, at least in the case of Firefox, this patch more than undoes the regression. I don't think it should affect performance, but we'll see.
doc(array,vec): add notes about side effects when empty-initializing
Copying some context from a conversation in the Rust discord:
* Both `vec![T; 0]` and `[T; 0]` are syntactically valid, and produce empty containers of their respective types
* Both *also* have side effects:
```rust
fn side_effect() -> String {
println!("side effect!");
"foo".into()
}
fn main() {
println!("before!");
let x = vec![side_effect(); 0];
let y = [side_effect(); 0];
println!("{:?}, {:?}", x, y);
}
```
produces:
```
before!
side effect!
side effect!
[], []
```
This PR just adds two small notes to each's documentation, warning users that side effects can occur.
I've also submitted a clippy proposal: https://github.com/rust-lang/rust-clippy/issues/6439
BTreeMap: clarify comments and panics around choose_parent_kv
Fixes a lie in recent code: `unreachable!("empty non-root node")` should shout "empty internal node", but it might as well be good and keep quiet
r? `@Mark-Simulacrum`
Clarify that String::split_at takes a byte index.
To someone skimming through the `String` docs and only reads the first line, the person could interpret "index" to be "char index". Later on in the docs it clarifies, but by adding "byte" it removes that ambiguity.
Privatize some of libcore unicode_internals
My understanding is that these API are perma unstable, so it doesn't
make sense to pollute docs & IDE completion[1] with them.
[1]: https://github.com/rust-analyzer/rust-analyzer/issues/6738
We also change the specialization of `SpecFromIterNested::from_iter` for
`TrustedLen` to use `Vec::with_capacity` when the iterator has a proper size
hint, instead of `Vec::new`, avoiding calls to `grow_*` and thus
`finish_grow` in some fully inlinable cases, which would regress with
this change.
Fixes#78471.
