This PR changes the `emit_opaque` and `read_opaque` methods in the RBML library to use a space-efficient binary encoder that does not emit any tags and uses the LEB128 variable-length integer format for all numbers it emits.
The space savings are nice, albeit a bit underwhelming, especially for dynamic libraries where metadata is already compressed.
| RLIBs | NEW | OLD |
|--------------|--------|-----------|
|libstd | 8.8 MB | 10.5 MB |
|libcore |15.6 MB | 19.7 MB |
|libcollections| 3.7 MB | 4.8 MB |
|librustc |34.0 MB | 37.8 MB |
|libsyntax |28.3 MB | 32.1 MB |
| SOs | NEW | OLD |
|---------------|-----------|--------|
| libstd | 4.8 MB | 5.1 MB |
| librustc | 8.6 MB | 9.2 MB |
| libsyntax | 7.8 MB | 8.4 MB |
At least this should make up for the size increase caused recently by also storing MIR in crate metadata.
Can this be a breaking change for anyone?
cc @rust-lang/compiler
This PR is a rebase of the original PR by @eddyb https://github.com/rust-lang/rust/pull/21836 with some unrebasable parts manually reapplied, feature gate added + type equality restriction added as described below.
This implementation is partial because the type equality restriction is applied to all type ascription expressions and not only those in lvalue contexts. Thus, all difficulties with detection of these contexts and translation of coercions having effect in runtime are avoided.
So, you can't write things with coercions like `let slice = &[1, 2, 3]: &[u8];`. It obviously makes type ascription less useful than it should be, but it's still much more useful than not having type ascription at all.
In particular, things like `let v = something.iter().collect(): Vec<_>;` and `let u = t.into(): U;` work as expected and I'm pretty happy with these improvements alone.
Part of https://github.com/rust-lang/rust/issues/23416
Make RFC 1214 warnings into errors, and rip out the "warn or err"
associated machinery. Future such attempts should go through lints
anyhow.
There is a fair amount of fallout in the compile-fail tests, as WF
checking now occurs earlier in the process.
r? @arielb1
associated machinery. Future such attempts should go through lints
anyhow.
There is a fair amount of fallout in the compile-fail tests, as WF
checking now occurs earlier in the process.
previously the error was erased and a `non-const path` error was emitted at the location of the field access instead of at the overflow location (as can be seen in the playground: http://is.gd/EuAF5F )
previously the error was erased and a `non-const path` error was emitted at the location of the field access instead of at the overflow location (as can be seen in the playground: http://is.gd/EuAF5F )
This fixes a bug in which unused imports can get wrongly marked as used when checking for unused qualifications in `resolve_path` (issue #30078), and it removes unused imports that were previously undetected because of the bug.
Ensure borrows of fn/closure params do not outlive invocations.
Does this by adding a new CallSiteScope to the region (or rather code extent) hierarchy, which outlives even the ParameterScope (which in turn outlives the DestructionScope of a fn/closure's body).
Fix#29793
r? @nikomatsakis
Currently, a coherence error based on overlapping impls simply mentions
the trait, and points to the two conflicting impls:
```
error: conflicting implementations for trait `Foo`
```
With this commit, the error will include all input types to the
trait (including the `Self` type) after unification between the
overlapping impls. In other words, the error message will provide
feedback with full type details, like:
```
error: conflicting implementations of trait `Foo<u32>` for type `u8`:
```
When the `Self` type for the two impls unify to an inference variable,
it is elided in the output, since "for type `_`" is just noise in that
case.
Closes#23980
r? @nikomatsakis
Currently, a coherence error based on overlapping impls simply mentions
the trait, and points to the two conflicting impls:
```
error: conflicting implementations for trait `Foo`
```
With this commit, the error will include all input types to the
trait (including the `Self` type) after unification between the
overlapping impls. In other words, the error message will provide
feedback with full type details, like:
```
error: conflicting implementations of trait `Foo<u32>` for type `u8`:
```
When the `Self` type for the two impls unify to an inference variable,
it is elided in the output, since "for type `_`" is just noise in that
case.
Closes#23980
Turns out that calling `resolve_type_variables_if_possible` in a O(n^2)
loop is a bad idea. Now we just resolve each copy of the region variable
to its lowest name each time (we resolve the region variable to its lowest
name, rather than to its unify-table name to avoid the risk of
the unify-table name changing infinitely many times. That may be
not a problem in practice, but I am not sure of it).
We can now handle name resolution errors and get past type checking (if we're a bit lucky). This is the first step towards doing code completion for partial programs (we need error recovery in the parser and early access to save-analysis).
Closes https://github.com/rust-lang/rust/issues/29935
The attributes `deprecated` and `rustc_deprecated` are completely independent in this implementation and it leads to some noticeable code duplication. Representing `deprecated` as
```
Stability {
level: Stable { since: "" },
feature: "",
depr: Some(Deprecation),
}
```
or, contrariwise, splitting rustc_deprecation from stability makes most of the duplication go away.
I can do this refactoring, but before doing it I must be sure, that further divergence of `deprecated` and `rustc_deprecated` is certainly not a goal.
cc @llogiq
resolve_lifetime.rs: Switch from BlockScope to FnScope in ScopeChain
construction. Lifetimes introduced by a fn signature are scoped to the
call-site for that fn. (Note `add_scope_and_walk_fn` must only add
FnScope for the walk of body, *not* of the fn signature.)
region.rs: Introduce new CodeExtentData::CallSiteScope variant. Use
CodeExtentData as the cx.parent, rather than just a NodeId. Change
DestructionScopeData to CallSiteScopeData.
regionck.rs: Thread call_site_scope via Rcx; constrain fn return
values.
(update; incorporated review feedback from niko.)
This PR reverts #29543 and instead implements proper support for "=*m" and "+*m" indirect output operands. This provides a framework on top of which support for plain memory operands ("m", "=m" and "+m") can be implemented.
This also fixes the liveness analysis pass not handling read/write operands correctly.
Turns out that calling `resolve_type_variables_if_possible` in a O(n^2)
loop is a bad idea. Now we just resolve each copy of the region variable
to its lowest name each time (we resolve the region variable to its lowest
name, rather than to its unify-table name to avoid the risk of
the unify-table name changing infinitely many times. That may be
not a problem in practice, but I am not sure of it).
Fixes#29844
I would prefer to
(a) make some performance measurements
(b) use the unification table in a few more places
before committing further, but this is probably good enough for beta.
r? @nikomatsakis
With this commit, metadata encoding and decoding can make use of thread-local encoding and decoding contexts. These allow implementers of `serialize::Encodable` and `Decodable` to access information and
datastructures that would otherwise not be available to them. For example, we can automatically translate def-id and span information during decoding because the decoding context knows which crate the data is decoded from. Or it allows to make `ty::Ty` decodable because the context has access to the `ty::ctxt` that is needed for creating `ty::Ty` instances.
Some notes:
- `tls::with_encoding_context()` and `tls::with_decoding_context()` (as opposed to their unsafe versions) try to prevent the TLS data getting out-of-sync by making sure that the encoder/decoder passed in is actually the same as the one stored in the context. This should prevent accidentally reading from the wrong decoder.
- There are no real tests in this PR. I had a unit tests for some of the core aspects of the TLS implementation but it was kind of brittle, a lot of code for mocking `ty::ctxt`, `crate_metadata`, etc and did actually test not so much. The code will soon be tested by the first incremental compilation auto-tests that rely on MIR being properly serialized. However, if people think that some tests should be added before this can land, I'll try to provide some that make sense.
r? @nikomatsakis
With this commit, metadata encoding and decoding can make use of
thread-local encoding and decoding contexts. These allow implementers
of serialize::Encodable and Decodable to access information and
datastructures that would otherwise not be available to them. For
example, we can automatically translate def-id and span information
during decoding because the decoding context knows which crate the
data is decoded from. Or it allows to make ty::Ty decodable because
the context has access to the ty::ctxt that is needed for creating
ty::Ty instances.
Instead of `ast::Ident`, bindings, paths and labels in HIR now keep a new structure called `hir::Ident` containing mtwt-renamed `name` and the original not-renamed `unhygienic_name`. `name` is supposed to be used by default, `unhygienic_name` is rarely used.
This is not ideal, but better than the status quo for two reasons:
- MTWT tables can be cleared immediately after lowering to HIR
- This is less bug-prone, because it is impossible now to forget applying `mtwt::resolve` to a name. It is still possible to use `name` instead of `unhygienic_name` by mistake, but `unhygienic_name`s are used only in few very special circumstances, so it shouldn't be a problem.
Besides name resolution `unhygienic_name` is used in some lints and debuginfo. `unhygienic_name` can be very well approximated by "reverse renaming" `token::intern(name.as_str())` or even plain string `name.as_str()`, except that it would break gensyms like `iter` in desugared `for` loops. This approximation is likely good enough for lints and debuginfo, but not for name resolution, unfortunately (see https://github.com/rust-lang/rust/issues/27639), so `unhygienic_name` has to be kept.
cc https://github.com/rust-lang/rust/issues/29782
r? @nrc
I've measured the time/memory consumption before and after - the difference is lost in statistical noise, so it's mostly a code simplification.
Sizes of `enum`s are not affected.
r? @nrc
I wonder if AST/HIR visitors could run faster if `P`s are systematically removed (except for cases where they control `enum` sizes). Theoretically they should.
Remaining unnecessary `P`s can't be easily removed because many folders accept `P<X>`s as arguments, but these folders can be converted to accept `X`s instead without loss of efficiency.
When I have a mood for some mindless refactoring again, I'll probably try to convert the folders, remove remaining `P`s and measure again.
This commit is the standard API stabilization commit for the 1.6 release cycle.
The list of issues and APIs below have all been through their cycle-long FCP and
the libs team decisions are listed below
Stabilized APIs
* `Read::read_exact`
* `ErrorKind::UnexpectedEof` (renamed from `UnexpectedEOF`)
* libcore -- this was a bit of a nuanced stabilization, the crate itself is now
marked as `#[stable]` and the methods appearing via traits for primitives like
`char` and `str` are now also marked as stable. Note that the extension traits
themeselves are marked as unstable as they're imported via the prelude. The
`try!` macro was also moved from the standard library into libcore to have the
same interface. Otherwise the functions all have copied stability from the
standard library now.
* `fs::DirBuilder`
* `fs::DirBuilder::new`
* `fs::DirBuilder::recursive`
* `fs::DirBuilder::create`
* `os::unix::fs::DirBuilderExt`
* `os::unix::fs::DirBuilderExt::mode`
* `vec::Drain`
* `vec::Vec::drain`
* `string::Drain`
* `string::String::drain`
* `vec_deque::Drain`
* `vec_deque::VecDeque::drain`
* `collections::hash_map::Drain`
* `collections::hash_map::HashMap::drain`
* `collections::hash_set::Drain`
* `collections::hash_set::HashSet::drain`
* `collections::binary_heap::Drain`
* `collections::binary_heap::BinaryHeap::drain`
* `Vec::extend_from_slice` (renamed from `push_all`)
* `Mutex::get_mut`
* `Mutex::into_inner`
* `RwLock::get_mut`
* `RwLock::into_inner`
* `Iterator::min_by_key` (renamed from `min_by`)
* `Iterator::max_by_key` (renamed from `max_by`)
Deprecated APIs
* `ErrorKind::UnexpectedEOF` (renamed to `UnexpectedEof`)
* `OsString::from_bytes`
* `OsStr::to_cstring`
* `OsStr::to_bytes`
* `fs::walk_dir` and `fs::WalkDir`
* `path::Components::peek`
* `slice::bytes::MutableByteVector`
* `slice::bytes::copy_memory`
* `Vec::push_all` (renamed to `extend_from_slice`)
* `Duration::span`
* `IpAddr`
* `SocketAddr::ip`
* `Read::tee`
* `io::Tee`
* `Write::broadcast`
* `io::Broadcast`
* `Iterator::min_by` (renamed to `min_by_key`)
* `Iterator::max_by` (renamed to `max_by_key`)
* `net::lookup_addr`
New APIs (still unstable)
* `<[T]>::sort_by_key` (added to mirror `min_by_key`)
Closes#27585Closes#27704Closes#27707Closes#27710Closes#27711Closes#27727Closes#27740Closes#27744Closes#27799Closes#27801
cc #27801 (doesn't close as `Chars` is still unstable)
Closes#28968
This commit is the standard API stabilization commit for the 1.6 release cycle.
The list of issues and APIs below have all been through their cycle-long FCP and
the libs team decisions are listed below
Stabilized APIs
* `Read::read_exact`
* `ErrorKind::UnexpectedEof` (renamed from `UnexpectedEOF`)
* libcore -- this was a bit of a nuanced stabilization, the crate itself is now
marked as `#[stable]` and the methods appearing via traits for primitives like
`char` and `str` are now also marked as stable. Note that the extension traits
themeselves are marked as unstable as they're imported via the prelude. The
`try!` macro was also moved from the standard library into libcore to have the
same interface. Otherwise the functions all have copied stability from the
standard library now.
* The `#![no_std]` attribute
* `fs::DirBuilder`
* `fs::DirBuilder::new`
* `fs::DirBuilder::recursive`
* `fs::DirBuilder::create`
* `os::unix::fs::DirBuilderExt`
* `os::unix::fs::DirBuilderExt::mode`
* `vec::Drain`
* `vec::Vec::drain`
* `string::Drain`
* `string::String::drain`
* `vec_deque::Drain`
* `vec_deque::VecDeque::drain`
* `collections::hash_map::Drain`
* `collections::hash_map::HashMap::drain`
* `collections::hash_set::Drain`
* `collections::hash_set::HashSet::drain`
* `collections::binary_heap::Drain`
* `collections::binary_heap::BinaryHeap::drain`
* `Vec::extend_from_slice` (renamed from `push_all`)
* `Mutex::get_mut`
* `Mutex::into_inner`
* `RwLock::get_mut`
* `RwLock::into_inner`
* `Iterator::min_by_key` (renamed from `min_by`)
* `Iterator::max_by_key` (renamed from `max_by`)
Deprecated APIs
* `ErrorKind::UnexpectedEOF` (renamed to `UnexpectedEof`)
* `OsString::from_bytes`
* `OsStr::to_cstring`
* `OsStr::to_bytes`
* `fs::walk_dir` and `fs::WalkDir`
* `path::Components::peek`
* `slice::bytes::MutableByteVector`
* `slice::bytes::copy_memory`
* `Vec::push_all` (renamed to `extend_from_slice`)
* `Duration::span`
* `IpAddr`
* `SocketAddr::ip`
* `Read::tee`
* `io::Tee`
* `Write::broadcast`
* `io::Broadcast`
* `Iterator::min_by` (renamed to `min_by_key`)
* `Iterator::max_by` (renamed to `max_by_key`)
* `net::lookup_addr`
New APIs (still unstable)
* `<[T]>::sort_by_key` (added to mirror `min_by_key`)
Closes#27585Closes#27704Closes#27707Closes#27710Closes#27711Closes#27727Closes#27740Closes#27744Closes#27799Closes#27801
cc #27801 (doesn't close as `Chars` is still unstable)
Closes#28968
See https://github.com/rust-lang/rfcs/pull/16 and https://github.com/rust-lang/rust/issues/15701
- Added syntax support for attributes on expressions and all syntax nodes in statement position.
- Extended `#[cfg]` folder to allow removal of statements, and
of expressions in optional positions like expression lists and trailing
block expressions.
- Extended lint checker to recognize lint levels on expressions and
locals.
- As per RFC, attributes are not yet accepted on `if` expressions.
Examples:
```rust
let x = y;
{
...
}
assert_eq!((1, #[cfg(unset)] 2, 3), (1, 3));
let FOO = 0;
```
Implementation wise, there are a few rough corners and open questions:
- The parser work ended up a bit ugly.
- The pretty printer change was based mostly on guessing.
- Similar to the `if` case, there are some places in the grammar where a new `Expr` node starts,
but where it seemed weird to accept attributes and hence the parser doesn't. This includes:
- const expressions in patterns
- in the middle of an postfix operator chain (that is, after `.`, before indexing, before calls)
- on range expressions, since `#[attr] x .. y` parses as `(#[attr] x) .. y`, which is inconsistent with
`#[attr] .. y` which would parse as `#[attr] (.. y)`
- Attributes are added as additional `Option<Box<Vec<Attribute>>>` fields in expressions and locals.
- Memory impact has not been measured yet.
- A cfg-away trailing expression in a block does not currently promote the previous `StmtExpr` in a block to a new trailing expr. That is to say, this won't work:
```rust
let x = {
#[cfg(foo)]
Foo { data: x }
#[cfg(not(foo))]
Foo { data: y }
};
```
- One-element tuples can have their inner expression removed to become Unit, but just Parenthesis can't. Eg, `(#[cfg(unset)] x,) == ()` but `(#[cfg(unset)] x) == error`. This seemed reasonable to me since tuples and unit are type constructors, but could probably be argued either way.
- Attributes on macro nodes are currently unconditionally dropped during macro expansion, which seemed fine since macro disappear at that point?
- Attributes on `ast::ExprParens` will be prepend-ed to the inner expression in the hir folder.
- The work on pretty printer tests for this did trigger, but not fix errors regarding macros:
- expression `foo![]` prints as `foo!()`
- expression `foo!{}` prints as `foo!()`
- statement `foo![];` prints as `foo!();`
- statement `foo!{};` prints as `foo!();`
- statement `foo!{}` triggers a `None` unwrap ICE.
The local item-path includes the local crates path to the extern crate
declaration which breaks cross-crate rustdoc links if the extern crate
is not linked into the crate root or renamed via `extern foo as bar`.
This PR allows the constant evaluation of index operations on constant arrays and repeat expressions. This allows index expressions to appear in the expression path of the length expression of a repeat expression or an array type.
An example is
```rust
const ARR: [usize; 5] = [1, 2, 3, 4, 5];
const ARR2: [usize; ARR[1]] = [42, 99];
```
In most other locations llvm's const evaluator figures it out already. This is not specific to index expressions and could be remedied in the future.
Fixes https://github.com/rust-lang/rust/issues/28692
Fixes https://github.com/rust-lang/rust/issues/28992
Fixes some other similar issues (see the tests)
[breaking-change], needs crater run (cc @brson or @alexcrichton )
The pattern with parens `UnitVariant(..)` for unit variants seems to be popular in rustc (see the second commit), but mostly used by one person (@nikomatsakis), according to git blame. If it causes breakage on crates.io I'll add an exceptional case for it.
nodes in statement position.
Extended #[cfg] folder to allow removal of statements, and
of expressions in optional positions like expression lists and trailing
block expressions.
Extended lint checker to recognize lint levels on expressions and
locals.
Trait references are always invariant, so all uses of subtyping between
them are equivalent to using equality.
Moreover, the overlap check was previously performed twice per impl
pair, once in each direction. It is now performed only once, and
internally uses the equality check.
On glium, a crate that spends some time in coherence, this change sped
up coherence checking by a few percent (not very significant).
r? @nikomatsakis
This patch implements the plan described in https://internals.rust-lang.org/t/privacy-and-its-interaction-with-docs-lints-and-stability/2880 with one deviation.
It turns out, that rustdoc needs the "directly public" set for its docs inlining logic, so the privacy pass have to produce three sets and not two. Three is arguably too many, so I merged them in one map:
`public_items/exported_items/reachable_items: NodeSet => access_levels: NodeMap<AccessLevel>`
r? @alexcrichton
Trait references are always invariant, so all uses of subtyping between
them are equivalent to using equality.
Moreover, the overlap check was previously performed twice per impl
pair, once in each direction. It is now performed only once, and
internally uses the equality check.
On glium, a crate that spends some time in coherence, this change sped
up coherence checking by a few percent (not very significant).
