Commit Graph

35367 Commits

Author SHA1 Message Date
Alex Crichton
d08600b189 std: Move the panic flag to its own thread local
This flag is somewhat tied to the `unwind` module rather than the `thread_info`
module, so this commit moves it into that module as well as allowing the same OS
thread to call `unwind::try` multiple times. Previously once a thread panicked
its panic flag was never reset, even after exiting the panic handler.
2014-12-18 23:35:52 -08:00
Alex Crichton
5759cff48e std: Lower abstractions for thread_local/at_exit
The current implementations use `std::sync` primitives, but these primitives
currently end up relying on `thread_info` and a local `Thread` being available
(mainly for checking the panicking flag).

To get around this, this commit lowers the abstractions used by the windows
thread_local implementation as well as the at_exit_imp module. Both of these
modules now use a `sys::Mutex` and a `static mut` and manage the
allocation/locking manually.
2014-12-18 23:35:52 -08:00
Aaron Turon
a27fbac868 Revise std::thread API to join by default
This commit is part of a series that introduces a `std::thread` API to
replace `std::task`.

In the new API, `spawn` returns a `JoinGuard`, which by default will
join the spawned thread when dropped. It can also be used to join
explicitly at any time, returning the thread's result. Alternatively,
the spawned thread can be explicitly detached (so no join takes place).

As part of this change, Rust processes now terminate when the main
thread exits, even if other detached threads are still running, moving
Rust closer to standard threading models. This new behavior may break code
that was relying on the previously implicit join-all.

In addition to the above, the new thread API also offers some built-in
support for building blocking abstractions in user space; see the module
doc for details.

Closes #18000

[breaking-change]
2014-12-18 23:31:52 -08:00
Aaron Turon
13f302d0c5 Update doc comment for std::rt 2014-12-18 23:31:52 -08:00
Alex Crichton
4ffd9f49c3 Avoid .take().unwrap() with FnOnce closures 2014-12-18 23:31:52 -08:00
Alex Crichton
a7061d02e1 Tweak the startup routine to pass on linux
We need to be sure to init thread_info before we init args for example because
args is grabbing locks which may entail looking at the local thread eventually.
2014-12-18 23:31:52 -08:00
Alex Crichton
7a6c54c46e Fix compilation on linux 2014-12-18 23:31:52 -08:00
Alex Crichton
9644d60cc4 Fix the capture_stderr test
There's always a fun time having two sets of standard libraries when testing!
2014-12-18 23:31:52 -08:00
Aaron Turon
ced2239852 Disable capture_stderr test for now 2014-12-18 23:31:51 -08:00
Aaron Turon
43ae4b3301 Fallout from new thread API 2014-12-18 23:31:51 -08:00
Aaron Turon
14c1a103bc Revise rt::unwind 2014-12-18 23:31:51 -08:00
Aaron Turon
d8e4780b0b Remove rt::{mutex, exclusive} 2014-12-18 23:31:51 -08:00
Aaron Turon
7fd7ce682d Add blocking support module for channels 2014-12-18 23:31:51 -08:00
Aaron Turon
84cb6cd938 Remove rt::{local, local_data, thread_local_storage} 2014-12-18 23:31:51 -08:00
Aaron Turon
cac133c9a8 Introduce std::thread
Also removes:

* `std::task`
* `std::rt::task`
* `std::rt::thread`

Notes for the new API are in a follow-up commit.

Closes #18000
2014-12-18 23:31:35 -08:00
Aaron Turon
9b03b72d7f Remove rt::bookkeeping
This commit removes the runtime bookkeeping previously used to ensure
that all Rust tasks were joined before the runtime was shut down.

This functionality will be replaced by an RAII style `Thread` API, that
will also offer a detached mode.

Since this changes the semantics of shutdown, it is a:

[breaking-change]
2014-12-18 23:31:35 -08:00
Aaron Turon
c009bfdf94 Make at_exit initialize lazily 2014-12-18 23:31:34 -08:00
Aaron Turon
b66681cd31 Allow args to work without rt initialization 2014-12-18 23:31:34 -08:00
Aaron Turon
74d0769993 Refactor std::os to use sys::os 2014-12-18 23:31:34 -08:00
Aaron Turon
2b3477d373 libs: merge librustrt into libstd
This commit merges the `rustrt` crate into `std`, undoing part of the
facade. This merger continues the paring down of the runtime system.

Code relying on the public API of `rustrt` will break; some of this API
is now available through `std::rt`, but is likely to change and/or be
removed very soon.

[breaking-change]
2014-12-18 23:31:34 -08:00
bors
840de07208 auto merge of #19955 : Gankro/rust/kill-all-code, r=aturon
EnumSet lives on in libcollections so that librustc can still use it. This adds a direct dependency on libcollections to librustc and libserialize.

Should not be merged until https://github.com/rust-lang/rfcs/pull/509 is accepted.

All of these collections have already been moved to collect-rs where they will ideally be maintained and experimented with, or will be replaced by something better: https://github.com/Gankro/collect-rs/

[breaking-change]

r? @aturon @alexcrichton
2014-12-19 02:52:01 +00:00
Alexis Beingessner
67d3823fc3 enumset fallout 2014-12-18 16:20:32 -05:00
Alexis Beingessner
0bd4dc68e6 s/Tree/BTree 2014-12-18 16:20:32 -05:00
Alexis Beingessner
6c00f9c5ff remove TreeMap, TreeSet, TrieMap, TrieSet, LruCache. deprecate EnumSet's std re-export 2014-12-18 16:20:31 -05:00
bors
99d6956c3b auto merge of #19896 : ktossell/rust/allow-nodoc-install, r=alexcrichton
If you configure with `--disable-docs`, the `doc` directory does not get generated, so the
`cp -r doc dist/` step fails when you `make dist{,-tar-bins,-doc}` or `make install`.
2014-12-18 20:32:07 +00:00
bors
f9a48492a8 auto merge of #19984 : japaric/rust/macro-expressions, r=alexcrichton
followed by a semicolon.

This allows code like `vec![1i, 2, 3].len();` to work.

This breaks code that uses macros as statements without putting
semicolons after them, such as:

    fn main() {
        ...
        assert!(a == b)
        assert!(c == d)
        println(...);
    }

It also breaks code that uses macros as items without semicolons:

    local_data_key!(foo)

    fn main() {
        println("hello world")
    }

Add semicolons to fix this code. Those two examples can be fixed as
follows:

    fn main() {
        ...
        assert!(a == b);
        assert!(c == d);
        println(...);
    }

    local_data_key!(foo);

    fn main() {
        println("hello world")
    }

RFC #378.

Closes #18635.

[breaking-change]

---

Rebased version of #18958
r? @alexcrichton 
cc @pcwalton
2014-12-18 17:32:07 +00:00
Patrick Walton
ddb2466f6a librustc: Always parse macro!()/macro![] as expressions if not
followed by a semicolon.

This allows code like `vec![1i, 2, 3].len();` to work.

This breaks code that uses macros as statements without putting
semicolons after them, such as:

    fn main() {
        ...
        assert!(a == b)
        assert!(c == d)
        println(...);
    }

It also breaks code that uses macros as items without semicolons:

    local_data_key!(foo)

    fn main() {
        println("hello world")
    }

Add semicolons to fix this code. Those two examples can be fixed as
follows:

    fn main() {
        ...
        assert!(a == b);
        assert!(c == d);
        println(...);
    }

    local_data_key!(foo);

    fn main() {
        println("hello world")
    }

RFC #378.

Closes #18635.

[breaking-change]
2014-12-18 12:09:07 -05:00
bors
c4d58ce15b auto merge of #19819 : vadimcn/rust/fix-demangle, r=alexcrichton
Windows dbghelp strips leading underscores from symbols, and I could not find a way to turn this off.  So let's accept  "ZN...E" form too.
Also, print PC displacement from symbols.  This is helpful in gauging whether the PC was indeed within the function displayed in the backtrace, or whether it just happened to be the closest public symbol in the module.
2014-12-18 14:42:06 +00:00
bors
2a231594c4 auto merge of #19780 : jakub-/rust/e-needstest, r=alexcrichton
Closes #5988.
Closes #10176.
Closes #10456.
Closes #12744.
Closes #13264.
Closes #13324.
Closes #14182.
Closes #15381.
Closes #15444.
Closes #15480.
Closes #15756.
Closes #16822.
Closes #16966.
Closes #17351.
Closes #17503.
Closes #17545.
Closes #17771.
Closes #17816.
Closes #17897.
Closes #17905.
Closes #18188.
Closes #18232.
Closes #18345.
Closes #18389.
Closes #18400.
Closes #18502.
Closes #18611.
Closes #18783.
Closes #19009.
Closes #19081.
Closes #19098.
Closes #19127.
Closes #19135.
2014-12-18 12:21:57 +00:00
bors
c0b2885ee1 auto merge of #19769 : nick29581/rust/coerce-if, r=nikomatsakis
r? @nikomatsakis 

We discussed coercions for `if` and `match` expressions. `if` seems to work already, was there some specific behaviour which wasn't working?
2014-12-18 06:41:59 +00:00
Nick Cameron
46eb72453f Propagate coercions through match expressions. 2014-12-18 13:48:05 +13:00
Ken Tossell
cbf80f3a68 Only try to install the doc directory if it exists.
If you configure with `--disable-docs`, the `doc` directory does not get generated, so
`cp -r doc dist/` fails when you `make dist{,-tar-bins,-doc}` or `make install`
2014-12-17 19:24:40 -05:00
Jakub Bukaj
62d80df0c9 Add tests for E-needstest issues 2014-12-17 23:00:32 +01:00
bors
22a9f250b5 auto merge of #19958 : alexcrichton/rust/rollup, r=alexcrichton 2014-12-17 21:33:15 +00:00
Alex Crichton
5c98952409 Test fixes and rebase conflicts 2014-12-17 11:50:32 -08:00
Alex Crichton
76fbb5d527 rollup merge of #19947: csouth3/arc-borrowfrom
Closes #19937.
2014-12-17 11:50:31 -08:00
Alex Crichton
7ffbd8bad5 rollup merge of #19943: steveklabnik/doc_std_vec
Random improvements to the `std::vec` docs.
2014-12-17 11:50:31 -08:00
Alex Crichton
cbc3cf7a9e rollup merge of #19942: steveklabnik/doc_std_mem 2014-12-17 11:50:31 -08:00
Alex Crichton
1b9b647b07 rollup merge of #19940: phi-gamma/canonical-master
Signed-off-by: Philipp Gesang <phg@phi-gamma.net>
2014-12-17 11:50:31 -08:00
Alex Crichton
a7bb4795d3 rollup merge of #19935: cgaebel/hashmap-tuple-indexing
r? @Gankro @pczarn
2014-12-17 11:50:31 -08:00
Alex Crichton
42dbee098b rollup merge of #19930: cllns/lowercase-if
On the [guide site](http://doc.rust-lang.org/guide.html#if) I was confused when I got to "5 If". It looked like "5 LF" in lowercase.

Changing the if to lowercase solves this problem. I know titles are all capitalized, but I think it makes sense in this case to keep it lowercase, since `if` is a reserved word. I'd also be open to making it ```    `if`   ``` but I'm not sure how that would look on the site.

Before:
![screen shot 2014-12-16 at 12 58 01](https://cloud.githubusercontent.com/assets/632942/5458866/cb34c006-8523-11e4-89ef-3a3964bcedfc.png)
After:
![screen shot 2014-12-16 at 12 58 14](https://cloud.githubusercontent.com/assets/632942/5458865/cb33c444-8523-11e4-8d95-d377ed583ed6.png)
2014-12-17 11:50:30 -08:00
Alex Crichton
b56d4bf916 rollup merge of #19923: pnkfelix/fix-make-tags.emacs
Fix `make TAGS.emacs`.

@nikomatsakis has been complaining to me about this.  (I had not noticed since I drive `ctags` with a separate script.)

(Suitable for a rollup build.)
2014-12-17 11:50:30 -08:00
Alex Crichton
a02885e167 rollup merge of #19918: pnkfelix/ast-refactor-make-place-in-exprbox-an-option
This is to allow us to migrate away from UnUniq in a followup commit,
and thus unify the code paths related to all forms of `box`.
2014-12-17 11:50:30 -08:00
Alex Crichton
b496adaefb rollup merge of #19912: P1start/fn-formatting
This is to encourage the use of the sugary syntax instead of the `<>` syntax, which will not be usable post-1.0. Rustdoc [still uses the `<>` syntax](https://github.com/rust-lang/rust/issues/19909), so if a rustdoc wizard is looking for something to do, it would be nice to use the parenthetical syntax there as well. (I tried to patch rustdoc as well, but failed…)
2014-12-17 11:50:30 -08:00
Alex Crichton
1b929090ac rollup merge of #19911: mdinger/tuple_tests2
Fixes some tuple indexing deprecation warnings. Didn't test. Don't see how it could fail unless I need to modify a makefile somewhere...

r? @alexcrichton
2014-12-17 11:50:30 -08:00
Alex Crichton
cce3fc043e rollup merge of #19905: AaronFriel/patch-1
Was testing rustup on a very minimal Debian installation and got errors during the install process (error occurred in `install.sh` of the Rust nightly.)

Noticed that Rustup was downloading the i686 nightly instead of x86-64. Installing `file` fixed the problem, and this patch adds the probe to ensure file is installed before attempting to use it.

There may still be an issue with the i686 installation, I did not investigate further.
2014-12-17 11:50:30 -08:00
Alex Crichton
3369b33a20 rollup merge of #19902: alexcrichton/second-pass-mem
This commit stabilizes the `mem` and `default` modules of std.
2014-12-17 11:50:29 -08:00
Alex Crichton
cada943f14 rollup merge of #19895: jbranchaud/add-string-add-doctest 2014-12-17 11:50:29 -08:00
Alex Crichton
1f2a18c7a6 rollup merge of #19893: JacobEdelman/patch-1
FIxed the spelling of the word "specific".
2014-12-17 11:50:29 -08:00
Alex Crichton
4a7757038a rollup merge of #19892: pnkfelix/region-graphviz
Added -Z print-region-graph debugging option; produces graphviz visualization of region inference constraint graph.

Optionally uses environment variables `RUST_REGION_GRAPH=<path_template>` and `RUST_REGION_GRAPH_NODE=<node-id>` to select which file to output to and which AST node to print.
2014-12-17 11:50:29 -08:00