Commit Graph

80673 Commits

Author SHA1 Message Date
ljedrz
296e72f11c Deny bare trait objects in in src/liballoc 2018-07-10 20:45:16 +02:00
Anirudh Balaji
4b51484fed
Change wording for {copy, clone}_from_slice 2018-07-10 10:56:58 -07:00
bors
90bd83c9fc Auto merge of #52196 - ollie27:rustdoc_ctor_imports, r=QuietMisdreavus
rustdoc: Hide struct and enum variant constructor imports

This is fallout from #51425. The duplicate variant imports can be seen [here](https://doc.rust-lang.org/nightly/std/prelude/v1/index.html) for example.

This is fixing a regression so could be backported to beta.

r? @QuietMisdreavus
2018-07-10 17:29:30 +00:00
ljedrz
d1e3567250 Deny bare trait objects in src/bootstrap. 2018-07-10 18:10:05 +02:00
bors
ce45cbb053 Auto merge of #52191 - SimonSapin:alloc_error_handler, r=alexcrichton
Implement #[alloc_error_handler]

This to-be-stable attribute is equivalent to `#[lang = "oom"]`. It is required when using the `alloc` crate without the `std` crate. It is called by `handle_alloc_error`, which is in turned called by "infallible" allocations APIs such as `Vec::push`.
2018-07-10 15:20:17 +00:00
NODA, Kai
97d0bc3f04
bootstrap: our best to achieve atomic rename on Win32
This is a tricky operation to implement on Win32; see
https://ci.appveyor.com/project/nodakai/python-win-behavior

Signed-off-by: NODA, Kai <nodakai@gmail.com>
2018-07-10 22:50:20 +08:00
NODA, Kai
bcb8a06ef7 bootstrap: write texts to a .tmp file first for atomicity
If you are using a hard-linked file as your config.toml, this change will affect the way other instances of the file is modified.
The original version would modify all other instances whereas the new version will leave others unchanged, reducing the ref count by one.

Signed-off-by: NODA, Kai <nodakai@gmail.com>
2018-07-10 22:50:00 +08:00
steveklabnik
0df68d850c Remove rustdoc plugins
See CVE-2018-1000622.
2018-07-10 10:41:39 -04:00
bors
fc491526dd Auto merge of #52168 - nikomatsakis:nll-region-name, r=estebank
find and highlight the `&` or `'_` in `region_name`

Before:

```
   --> $DIR/dyn-trait-underscore.rs:18:5
    |
 LL | fn a<T>(items: &[T]) -> Box<dyn Iterator<Item=&T>> {
-   |         ----- lifetime `'1` appears in this argument
 LL |     Box::new(items.iter()) //~ ERROR cannot infer an appropriate lifetime
    |     ^^^^^^^^^^^^^^^^^^^^^^ cast requires that `'1` must outlive `'static`
```

After:

```
   --> $DIR/dyn-trait-underscore.rs:18:5
    |
 LL | fn a<T>(items: &[T]) -> Box<dyn Iterator<Item=&T>> {
+   |                - let's call the lifetime of this reference `'1`
 LL |     Box::new(items.iter()) //~ ERROR cannot infer an appropriate lifetime
    |     ^^^^^^^^^^^^^^^^^^^^^^ cast requires that `'1` must outlive `'static`
```

Not intended as the final end point necessarily in any sense. I intentionally left some to-do points to fill in later:

- Does not apply to upvars in closures yet (should be relatively easy)
- Does not handle the case where we can't find a precise match very well
- And of course we can still tweak wording

but shows the basic idea of how to make the `Ty` and `hir::Ty` to find a good spot to highlight.

r? @estebank
cc @davidtwco
2018-07-10 11:19:31 +00:00
bors
e46bfa2879 Auto merge of #52204 - zackmdavis:and_the_crate_of_the_missing_module, r=oli-obk
correct import suggestions for edition 2018

For #52202.
2018-07-10 09:05:46 +00:00
Ralf Jung
f511c5ea4a improve error message shown for unsafe operations: explain why undefined behavior could arise
Inspired by @gnzlbg at https://github.com/rust-lang/rust/issues/46043#issuecomment-381544673
2018-07-10 11:01:44 +02:00
bors
77117e3836 Auto merge of #52177 - ljedrz:bare_gcc_warning, r=alexcrichton
Warn windows-gnu users that the bundled gcc can't compile

Add a `DO NOT USE THIS gcc.exe FOR COMPILATION.txt` file to `lib\rustlib\*-pc-windows-gnu\bin` folders in `windows-gnu` installations in order to warn against attempting to use the bundled `gcc.exe` as a C compiler. I'm pretty sure that location is usually found manually, so this should be easily noticeable.

This mistake has been made plenty of times and has lead to misunderstandings:
Rust: [Bundled gcc (windows x64) is unable to build any c file](https://github.com/rust-lang/rust/issues/24418)
gtk-rs:    [Compiling on windows](https://github.com/gtk-rs/gtk/issues/625)
bzip2-rs: [Build failure at gcc level: blocksort.c not found](https://github.com/alexcrichton/bzip2-rs/issues/30)

Alternatives: rename the bundled `gcc.exe` to e.g. `rustc-gcc.exe` or `gcc-linker.exe`. This might require a more comprehensive change or break crates already using it as a linker.

r? @alexcrichton
2018-07-10 06:52:20 +00:00
Zack M. Davis
96b151bd9c in which use suggestions meet edition 2018
The intent here is to resolve #52202.
2018-07-09 23:21:29 -07:00
bors
b3e7d70ce7 Auto merge of #51583 - cuviper:packed_pair-bool, r=Mark-Simulacrum
Store scalar pair bools as i8 in memory

We represent `bool` as `i1` in a `ScalarPair`, unlike other aggregates,
to optimize IR for checked operators and the like.  With this patch, we
still do so when the pair is an immediate value, but we use the `i8`
memory type when the value is loaded or stored as an LLVM aggregate.

So `(bool, bool)` looks like an `{ i1, i1 }` immediate, but `{ i8, i8 }`
in memory.  When a pair is a direct function argument, `PassMode::Pair`,
it is still passed using the immediate `i1` type, but as a return value
it will use the `i8` memory type.  Also, `bool`-like` enum tags will now
use scalar pairs when possible, where they were previously excluded due
to optimization issues.

Fixes #51516.
Closes #51566.

r? @eddyb
cc @nox
2018-07-10 03:08:47 +00:00
bors
295858eba7 Auto merge of #52100 - nielx:fix/rust_driver-stacklimit, r=cramertj
Haiku: work around the lack of setrlimit

The default Unix codepath fails, because Haiku does not implement
setrlimit for stack size. Thus we create an additional path.

By default, Haiku has the desired 16 MB stack, therefore in general
we do not have to spawn a new thread. The code has been written in
such a way that any changes in Haiku or in Rust will be adapted to.
2018-07-10 01:07:23 +00:00
Emerentius
39fcfa8ccb step_by: leave time of item skip unspecified
this gives us some leeway when optimizing
2018-07-10 01:44:27 +02:00
Oliver Middleton
6b1d584ed8 rustdoc: Hide struct and enum variant constructor imports 2018-07-10 00:16:18 +01:00
bors
5b525916ee Auto merge of #51899 - gnzlbg:llvm501, r=alexcrichton
bump minimum LLVM version to 5.0

Closes #51878 .

r? @alexcrichton

--

cc @cuviper @infinity0
2018-07-09 23:08:51 +00:00
Alex Crichton
bc09fcb990 rustc: Avoid /tmp/ in graphviz writing
This issue was reported to security@rust-lang.org by Sebastien Marie following
our recent [security advisory][1]. Because `/tmp` is typically globally writable
it's possible for one user to place symlinks in `/tmp` pointing to files in
another user's directories, causing `rustc` to overwrite the contents of
innocent files by accident.

This patch instead defaults the output path here to the cwd which should avoid
this issue.

[1]: https://blog.rust-lang.org/2018/07/06/security-advisory-for-rustdoc.html
2018-07-09 15:26:57 -07:00
Simon Sapin
239ec7d2dc Implement #[alloc_error_handler]
This to-be-stable attribute is equivalent to `#[lang = "oom"]`.
It is required when using the alloc crate without the std crate.
It is called by `handle_alloc_error`, which is in turned called
by "infallible" allocations APIs such as `Vec::push`.
2018-07-09 23:13:24 +02:00
Simon Sapin
872effa118 Add a test for the default allocation error hook 2018-07-09 23:13:24 +02:00
Simon Sapin
620599e886 Remove extern on the pub fn rust_oom lang item in libstd, to match ABI of the declaration in liballoc
This turned out to be important on Windows.
Calling `handle_alloc_error(Layout:🆕:<[u8; 42]>())` caused:

```
Exception thrown at 0x00007FF7C70DC399 in a.exe: 0xC0000005:
Access violation reading location 0x000000000000002A.
```

0x2A equals 42, so it looks like the `Layout::size` field of type `usize`
was interpreted as a pointer to read from.
2018-07-09 23:13:24 +02:00
bors
9bd8458c92 Auto merge of #50250 - csmoe:wf_traitref, r=scalexm
Chalk lowering rule: WellFormed-TraitRef

Address chalk lowering "Implemented-From-Env" as part of #49177.
r? @nikomatsakis
2018-07-09 21:01:58 +00:00
Anirudh Balaji
c90a821d43
Add "or destination" to {copy, clone}_from_slice example 2018-07-09 13:41:46 -07:00
ashtneoi
dc8ae26c1e Fix issue #51515 and update test 2018-07-09 13:33:57 -07:00
ashtneoi
a49b75d2f3 Add test case from issue #51515 2018-07-09 13:16:02 -07:00
ljedrz
232e77e007 Change gcc warning file name, remove unnecessary reference 2018-07-09 21:25:16 +02:00
scalexm
37c5c0bf9c Change wording 2018-07-09 21:20:26 +02:00
bors
bdd185c897 Auto merge of #51956 - GuillaumeGomez:shutdown-doc-lints, r=oli-obk
Fix rustdoc run failures by shutting down definitely some lints

Fixes #51661.

cc @oli-obk @arielb1 @eddyb
2018-07-09 18:50:37 +00:00
Guillaume Gomez
66beb4e5b4 add comment about lints whitelisting 2018-07-09 18:10:08 +02:00
gnzlbg
3b36ce64a5 revert travis-ci changes 2018-07-09 17:25:00 +02:00
bors
c6807bb1b2 Auto merge of #52159 - SimonSapin:alloc-prelude, r=alexcrichton
Add the `alloc::prelude` module

It contains the re-exports that are in `std::prelude::v1` but not in `core::prelude::v1`.

Calling it prelude is somewhat of a misnomer since (unlike those modules in `std` or `core`) its contents are never implicitly imported in modules. Rather it is intended to be used with an explicit glob import like `use alloc::prelude::*;`. However there is precedent for the same misnomer with `std::io::prelude`, for example.

This new module is unstable with the same feature name as the `alloc` care. They are proposed for stabilization together in RFC https://github.com/rust-lang/rfcs/pull/2480.
2018-07-09 14:29:59 +00:00
ljedrz
6b6a7a4691 Warn windows-gnu users that the bundled gcc can't compile 2018-07-09 15:28:12 +02:00
Clément RENAULT
c8f0e6f210
Fix the documentation of Option::replace 2018-07-09 14:52:32 +02:00
Clément RENAULT
af87a3594a
Add a basic test to Option::replace 2018-07-09 14:50:54 +02:00
gnzlbg
fe75f617c3 run asmjs and emscripten builds on CI (temporary) 2018-07-09 11:35:53 +02:00
gnzlbg
23dfb42ab3 still support LLVM4 for emscripten 2018-07-09 11:35:53 +02:00
gnzlbg
52d7740277 bump llvm version of failing codegen test 2018-07-09 11:35:53 +02:00
gnzlbg
4ff90c7e0a bump minimum LLVM version to 5.0 2018-07-09 11:35:52 +02:00
bors
c30acc7187 Auto merge of #52160 - euclio:include-macros, r=oli-obk
add regression test for #48835

Fixes #48835.

The underlying issue was fixed in #51978.
2018-07-09 08:41:43 +00:00
Oliver Schneider
7a8e085835 Also distribute cargo clippy 2018-07-09 10:15:30 +02:00
Oliver Schneider
68620b14a2 Inject clippy into the rls again
Also makes sure we actually point to the local rls
2018-07-09 10:10:42 +02:00
Ben Harris
aac0d914b2 Correct some codegen stats counters 2018-07-09 09:40:16 +02:00
bors
ec039c7cb1 Auto merge of #52066 - benjaminp:obligation-select, r=Mark-Simulacrum
Remove obsolete documentation from FufillmentContext::select comment.

The `only_new_obligations` parameter has not existed since 43756934d2.
2018-07-09 06:38:47 +00:00
bors
a80a610a4c Auto merge of #52166 - orlp:master, r=joshtriplett
Performance improvement of Vec's swap_remove.

The old implementation *literally* swapped and then removed, which resulted in unnecessary move instructions. The new implementation does use unsafe code, but is easy to see that it is correct.

Fixes https://github.com/rust-lang/rust/issues/52150.
2018-07-09 04:42:27 +00:00
Orson Peters
e529dfd590 Removed a single trailing space. Oops. 2018-07-09 06:31:24 +02:00
Niko Matsakis
a6adb1ebff find and highlight the & or '_ in region_name 2018-07-09 00:20:36 -04:00
Orson Peters
6faa295cec Reimplemented Vec's swap_remove to not rely on pop. 2018-07-09 06:13:58 +02:00
Orson Peters
295768ae8f Performance improvement of Vec's swap_remove. 2018-07-09 05:01:39 +02:00
Andy Russell
1be1d90662
add regression test for #48835
Fixes #48835.

The underlying issue was fixed in #51978.
2018-07-08 16:45:01 -04:00