This commit guards all calls to `DepGraphThreadData::enqueue` with a
check to make sure it is enabled. This requires distinguishing between a
"fully enabled" and an "enqueue-enabled" graph.
This change avoids some useless allocation and vector manipulations when
the graph is disabled (i.e. when incremental compilation is off) which
improves speed by ~1% on some of the rustc-benchmarks.
macros: clean up scopes of expanded `#[macro_use]` imports
This PR changes the scope of macro-expanded `#[macro_use]` imports to match that of unexpanded `#[macro_use]` imports. For example, this would be allowed:
```rust
example!();
macro_rules! m { () => { #[macro_use(example)] extern crate example_crate; } }
m!();
```
This PR also enforces the full shadowing restrictions from RFC 1560 on `#[macro_use]` imports (currently, we only enforce the weakened restrictions from #36767).
This is a [breaking-change], but I believe it is highly unlikely to cause breakage in practice.
r? @nrc
Error monitor should emit error to stderr instead of stdout
We are pretty consistent about emitting to stderr, except for when there is actually an error, in which case we emit to stdout. This seems a bit backwards. This PR just changes that exception to emit to stderr. This is useful for the RLS since the LS protocol uses stdout (grrr).
r? @alexcrichton
Avoid allocations in `Decoder::read_str`.
`opaque::Decoder::read_str` is very hot within `rustc` due to its use in
the reading of crate metadata, and it currently returns a `String`. This
commit changes it to instead return a `Cow<str>`, which avoids a heap
allocation.
This change reduces the number of calls to `malloc` by almost 10% in
some benchmarks.
This is a [breaking-change] to libserialize.
Add comparison operators to boolean const eval.
I think it might be worth adding tests here, but since I don't know how or where to do that, I have not done so yet. Willing to do so if asked and given an explanation as to how.
Fixes#37047.
Changed 0 into '0'
Right now `0` is an undefined production rule.
[Documentation following the grammar specification](https://doc.rust-lang.org/nightly/std/fmt/#sign0) strongly suggests `'0'` is meant as it is used as a character literal.
r? @steveklabnik
ICH: Enable some cases in trait definition hashing.
Enable some test cases originally written by @eulerdisk. The tests can be enabled now because @MathieuBordere has fixed the underlying problem in #36974.
r? @nikomatsakis
Merge `Printer::token` and `Printer::size`.
Logically, it's a vector of pairs, so might as well represent it that
way.
The commit also changes `scan_stack` so that it is initialized with the
default size, instead of the excessive `55 * linewidth` size, which it
usually doesn't get even close to reaching.
Book: Be very explicit of lifetimes being descriptive
... not prescriptive. Pointed out in https://users.rust-lang.org/t/what-if-i-get-lifetimes-wrong/7535/4, which was a revelation to me and made me think this should be more clear in the book. I'm not sure if I got this entirely right or if the wording is good, but I figured a PR is more helpful than a simple issue.
r? @steveklabnik
Small Note: There's also https://github.com/rust-lang/book, should I have sent the PR there? It doesn't coincide with the online book though, so I figured it's better of here.
Add method str::repeat(self, usize) -> String
It is relatively simple to repeat a string n times:
`(0..n).map(|_| s).collect::<String>()`. It becomes slightly more
complicated to do it “right” (sizing the allocation up front), which
warrants a method that does it for us.
This method is useful in writing testcases, or when generating text.
`format!()` can be used to repeat single characters, but not repeating
strings like this.
rustdoc: print non-self arguments of bare functions and struct methods on their own line
This change alters the formatting rustdoc uses when it creates function and struct method documentation. For bare functions, each argument is printed on its own line. For struct methods, non-self arguments are printed on their own line. In both cases, no line breaks are introduced if there are no arguments, and for struct methods, no line breaks are introduced if there is only a single self argument. This should aid readability of long function signatures and allow for greater comprehension of these functions.
I've run rustdoc with these changes on my crate egg-mode and its set of dependencies and put the result [on my server](https://shiva.icesoldier.me/doc-custom/egg_mode/). Of note, here are a few shortcut links that highlight the changes:
* [Bare function with a long signature](https://shiva.icesoldier.me/doc-custom/egg_mode/place/fn.reverse_geocode.html)
* [Struct methods, with single self argument and with self and non-self arguments](https://shiva.icesoldier.me/doc-custom/egg_mode/tweet/struct.Timeline.html#method.reset)
* [Bare functions with no arguments](https://shiva.icesoldier.me/doc-custom/rand/fn.thread_rng.html) and [struct methods with no arguments](https://shiva.icesoldier.me/doc-custom/hyper/client/struct.Client.html#method.new) are left unchanged.
This PR consists of two commits: one for bare functions and one for struct methods.
Turn compatibility lint `match_of_unit_variant_via_paren_dotdot` into a hard error
The lint was introduced 10 months ago and made deny-by-default 7 months ago.
In case someone is still using it, https://github.com/rust-lang/rust/pull/36868 contains a stable replacement.
r? @nikomatsakis
configure: Add options for separate musl roots
This allows using the `./configure` script to enable rustbuild to compile
multiple musl targets at once. We'll hopefully use this soon on our bots to
produce a bunch of targets.
