macros 1.1: Allow proc_macro functions to declare attributes to be mark as used
This PR allows proc macro functions to declare attribute names that should be marked as used when attached to the deriving item. There are a few questions for this PR.
- Currently this uses a separate attribute named `#[proc_macro_attributes(..)]`, is this the best choice?
- In order to make this work, the `check_attribute` function had to be modified to not error on attributes marked as used. This is a pretty large change in semantics, is there a better way to do this?
- I've got a few clones where I don't know if I need them (like turning `item` into a `TokenStream`), can these be avoided?
- Is switching to `MultiItemDecorator` the right thing here?
Also fixes https://github.com/rust-lang/rust/issues/37563.
Vendor all rustbuild dependencies in this repo
This commit vendors all crates.io dependencies into the rust-lang/rust repository using the `cargo-vendor` tool. This is done in an effort to make rustbuild distro-ready by ensuring that our source tarballs are self-contained units which don't need extraneous network downloads.
A new `src/vendor` directory is created with all vendored crates, and Cargo, when using rustbuild, is configured to use this directory. Over time we can deduplicate this directory with the actual src tree (e.g. src/librustc_serialize, src/liblibc, src/libgetopts, ...). For now though that's left to a separate commit.
Document convention for using both fmt::Write and io::Write
Using a trait's methods (like `Write::write_fmt` as used in `writeln!` and other macros) requires importing that trait directly (not just the module containing it). Both `fmt::Write` and `io::Write` provide compatible `Write::write_fmt` methods, and code can use `writeln!` and other macros on both an object implementing `fmt::Write` (such as a `String`) and an object implementing `io::Write` (such as `Stderr`). However, importing both `Write` traits produces an error due to the name conflict.
The convention I've seen renames both of them on import, to `FmtWrite` and `IoWrite` respectively. Document that convention in the Rust documentation for `write!` and `writeln!`, with examples.
Point to type argument span when used as trait
Given the following code:
``` rust
struct Foo<T: Clone>(T);
use std::ops::Add;
impl<T: Clone, Add> Add for Foo<T> {
type Output = usize;
fn add(self, rhs: Self) -> Self::Output {
unimplemented!();
}
}
```
present the following output:
``` nocode
error[E0404]: `Add` is not a trait
--> file3.rs:5:21
|
5 | impl<T: Clone, Add> Add for Okok<T> {
| --- ^^^ expected trait, found type parameter
| |
| type parameter defined here
```
Fixes#35987.
Include type of missing trait methods in error
Provide either a span pointing to the original definition of missing
trait items, or a message with the inferred definitions.
Fixes#24626. Follow up to PR #36371.
If PR #37369 lands, missing trait items that present a multiline span will be able to show the entirety of the item definition on the error itself, instead of just the first line.
Replace FNV with a faster hash function.
Hash table lookups are very hot in rustc profiles and the time taken within `FnvHash` itself is a big part of that. Although FNV is a simple hash, it processes its input one byte at a time. In contrast, Firefox has a homespun hash function that is also simple but works on multiple bytes at a time. So I tried it out and the results are compelling:
```
futures-rs-test 4.326s vs 4.212s --> 1.027x faster (variance: 1.001x, 1.007x)
helloworld 0.233s vs 0.232s --> 1.004x faster (variance: 1.037x, 1.016x)
html5ever-2016- 5.397s vs 5.210s --> 1.036x faster (variance: 1.009x, 1.006x)
hyper.0.5.0 5.018s vs 4.905s --> 1.023x faster (variance: 1.007x, 1.006x)
inflate-0.1.0 4.889s vs 4.872s --> 1.004x faster (variance: 1.012x, 1.007x)
issue-32062-equ 0.347s vs 0.335s --> 1.035x faster (variance: 1.033x, 1.019x)
issue-32278-big 1.717s vs 1.622s --> 1.059x faster (variance: 1.027x, 1.028x)
jld-day15-parse 1.537s vs 1.459s --> 1.054x faster (variance: 1.005x, 1.003x)
piston-image-0. 11.863s vs 11.482s --> 1.033x faster (variance: 1.060x, 1.002x)
regex.0.1.30 2.517s vs 2.453s --> 1.026x faster (variance: 1.011x, 1.013x)
rust-encoding-0 2.080s vs 2.047s --> 1.016x faster (variance: 1.005x, 1.005x)
syntex-0.42.2 32.268s vs 31.275s --> 1.032x faster (variance: 1.014x, 1.022x)
syntex-0.42.2-i 17.629s vs 16.559s --> 1.065x faster (variance: 1.013x, 1.021x)
```
(That's a stage1 compiler doing debug builds. Results for a stage2 compiler are similar.)
The attached commit is not in a state suitable for landing because I changed the implementation of FnvHasher without changing its name (because that would have required touching many lines in the compiler). Nonetheless, it is a good place to start discussions.
Profiles show very clearly that this new hash function is a lot faster to compute than FNV. The quality of the new hash function is less clear -- it seems to do better in some cases and worse in others (judging by the number of instructions executed in `Hash{Map,Set}::get`).
CC @brson, @arthurprs
Partially stabilize RFC 1506 "Clarify relationships between ADTs"
Lifted restrictions on tuple structs/variants are stabilized, i.e. `S{..}` can be used with any structs and empty tuple structs are permitted without feature gate.
Numeric fields in struct expressions/patterns `S { 0: a, 1: b }` are **NOT** stabilized.
This was implemented 1.5 months ago in Rust 1.12, but this is a tiny technical change that could probably go even without RFC/stabilization period.
cc https://github.com/rust-lang/rust/issues/35626https://github.com/rust-lang/rust/pull/36871
r? @nikomatsakis
rustbuild: Fix dependencies of check-error-index
This depends on the error index actually existing rather than just the tool to
generate the error index.
Given the following code:
```rust
struct Foo<T: Clone>(T);
use std::ops::Add;
impl<T: Clone, Add> Add for Foo<T> {
type Output = usize;
fn add(self, rhs: Self) -> Self::Output {
unimplemented!();
}
}
```
present the following output:
```nocode
error[E0404]: `Add` is not a trait
--> file3.rs:5:21
|
5 | impl<T: Clone, Add> Add for Okok<T> {
| --- ^^^ expected trait, found type parameter
| |
| type parameter defined here
```
Add unary and binary tests for incr-comp
This is my draft of tests for unary and binary expressions as desired by #37520 for use in the test suite for hashes in incremental compilation. Feedback would be wonderful, if there's any changes I need to make I would appreciate the code review.
?r @michaelwoerister
book: Removed platform compatibility table, link to the forge
The content is duplicated, and it doesn't need to be in this location.
It's mostly trivia that doesn't apply to most of the audience.
The forge is up to date.
r? @steveklabnik cc @alexcrichton
Add error note to illegal code snippet
Mark intentionally invalid code snippet in documentation as such with a comment. Similar comments used elsewhere in this file.
r? @steveklabnik
Make it clear that the reference isn't normative
Any time someone edits the reference, it has to be taken very seriously,
since it's the closest thing we have to a specification. This commit
adds language which indicates that this is not a normative document,
which makes it easier to make tweaks without worrying about forever
harming the future of Rust by painting ourselves in a corner.
r? @aturon
Add `{into,from}_raw` to Rc and Arc
These methods convert to and from a `*const T` for `Rc` and `Arc` similar to the way they work on `Box`. The only slight complication is that `from_raw` needs to offset the pointer back to find the beginning of the `RcBox`/`ArcInner`.
I felt this is a fairly small addition, filling in a gap (when compared to `Box`) so it wouldn't need an RFC. The motivation is primarily for FFI.
(I'll create an issue and update a PR with the issue number if reviewers agree with the change in principle **Edit: done #37197**)
~~Edit: This was initially `{into,from}_raw` but concerns were raised about the possible footgun if mixed with the methods of the same name of `Box`.~~
Edit: This was went from `{into,from}_raw` to `{into,from}_inner_raw` then back to `{into,from}_raw` during review.
A few changes are included here:
* The `winapi` and `url` dependencies were dropped. The source code for these
projects is pretty weighty, and we're about to vendor them, so let's not
commit to that intake just yet. If necessary we can vendor them later but for
now it shouldn't be necessary.
* The `--frozen` flag is now always passed to Cargo, obviating the need for
tidy's `cargo_lock` check.
* Tidy was updated to not check the vendor directory
Closes#34687
This commit vendors all dependencies when using rustbuild to ensure that we
don't hit the network during a build and can build as a self-contained unit.
By using a second attribute `attributes(Bar)` on
proc_macro_derive, whitelist any attributes with
the name `Bar` in the deriving item. This allows
a proc_macro function to use custom attribtues
without a custom attribute error or unused attribute
lint.
Stabilize `..` in tuple (struct) patterns
I'd like to nominate `..` in tuple and tuple struct patterns for stabilization.
This feature is a relatively small extension to existing stable functionality and doesn't have known blockers.
The feature first appeared in Rust 1.10 6 months ago.
An example of use: https://github.com/rust-lang/rust/pull/36203
Closes https://github.com/rust-lang/rust/issues/33627
r? @nikomatsakis
KNOWN_ATTRIBUTES should really be named BUILT_ATTRIBUTES,
while KNOWN_ATTRIBUTES should be used to mark attributes
as known, similar to USED_ATTRIBUTES.
rustbuild: support MIPS host builds
There is a *little* code duplication, but primarily for sake of "match exhaustiveness". Let's blame Linux/MIPS for not exposing endianness explicitly in `uname -m` (that's user-space interface and as such is frozen).
Currently the build won't work as we have to wait for a new stage0 for the MIPS host compilers, but this paves the way to self-hosted Rust on MIPS. The cross-compiled MIPS binaries are confirmed to work on the Loongson 3A2000 (MIPS64r2-compatible) so we have plenty of confidence that they'll work on other MIPS platforms too, as Linux/MIPS user-space ABI is consistent across machines of the same bitness.
r? @alexcrichton
Remove mention of mipsel target_arch
This is the only remaining instance in rustc. All others were in the libc repo, removed with rust-lang/libc#445.
Actually there's more to clean in `libsyntax/abi.rs`, but let's save that for another commit...
Fix Arc::clone()'s MAX_REFCOUNT check (off-by-one)
Before, the strong count of an `Arc` could be set to
`MAX_REFCOUNT + 1`, because when this happened, `old_size` would
be exactly `MAX_REFCOUNT`. `Arc::clone()` would not abort.
This commit changes the check in `Arc::clone()` to also abort if
the old value is equal to `MAX_REFCOUNT`, because then the new
value will be equal to `MAX_REFCOUNT + 1`.
A test would require allocating memory for `isize::MAX` pointers.
This would probably crash any machine, so no test is submitted
with this commit.
There is a *little* code duplication, but primarily for sake of "match
exhaustiveness". Let's blame Linux/MIPS for not exposing endianness
explicitly in `uname -m` (that's user-space interface and as such is
frozen).
Currently the build won't work as we have to wait for a new stage0 for
the MIPS host compilers, but this paves the way to self-hosted Rust on
MIPS. The cross-compiled MIPS binaries are confirmed to work on the
Loongson 3A2000 (MIPS64r2-compatible) so we have plenty of confidence
that they'll work on other MIPS platforms too, as Linux/MIPS user-space
ABI is consistent across machines of the same bitness.
The `mipsel` `target_arch` was introduced with the initial MIPSel
support (rust-lang/rust@82ec1aef29),
but was subsequently removed with implementation of the Flexible Target
Specification (Rust RFC 0131,
rust-lang/rust@3a8f4ec32a).
This is the only remaining instance in rustc. All others are in the libc
repo, and are fixed in rust-lang/libc@b3676593f6.
macros: improve shadowing checks
This PR improves macro-expanded shadowing checks to work with out-of-(pre)order expansion.
Out-of-order expansion became possible in #37084, so this technically a [breaking-change] for nightly.
The regression test from this PR is an example of code that would break.
r? @nrc
Force static linking of LLVM
Run `llvm-config` with `--link-static` if available, to force static linking of LLVM.
This option was added in LLVM 3.8.
This is my first pull request, any feedback is welcome!
Fixes#36854
See also: #36996
The constant name `MAX_REFCOUNT` suggests that the value is a
_hard_ limit on the amount of references to an `Arc`. This is
a more soft limit however. This commit adds a comment to the
constant to annotate this.
See also: PR #37605
