Commit Graph

57486 Commits

Author SHA1 Message Date
Joe Neeman
902e11dd8f Change error message for intrinsic signature.
Makes it so the signature of the intrinsic in the user's code is
"found," while the signature that rustc knows about is "expected."
2016-09-22 23:58:35 +02:00
Jeffrey Seyfried
173d5b339f Split up DefCollector::root(). 2016-09-22 21:14:05 +00:00
Jeffrey Seyfried
1ee64e44a9 Use Resolver::visit_expansion only with monotonic expansions. 2016-09-22 21:14:03 +00:00
Nicholas Nethercote
cf50f5f965 [breaking-change] Remove TypedArena::with_capacity.
The function is unused by rustc. Also, it doesn't really follow the
usual meaning of a `with_capacity` function because the first chunk
allocation is now delayed until the first `alloc` call.

This change reduces the size of `TypedArena` by one `usize`.
2016-09-23 06:40:22 +10:00
bors
4c04e36f94 Auto merge of #36651 - jonathandturner:rollup, r=jonathandturner
Rollup of 7 pull requests

- Successful merges: #36330, #36423, #36539, #36571, #36589, #36600, #36632
- Failed merges:
2016-09-22 13:11:04 -07:00
Nick Cameron
3863834d9c reviewer comments and rebasing 2016-09-23 07:19:31 +12:00
Jonathan Turner
c1e39380fb Rollup merge of #36632 - CryZe:patch-3, r=sfackler
Fix outdated Doc Comment on BufReader::seek

A long time ago non-panicking `unwrap` methods were renamed to `into_inner` in this Pull Request: https://github.com/rust-lang/rust/pull/19149

Looks like this doc comment was not updated however.
2016-09-22 11:25:02 -07:00
Jonathan Turner
bc7991a54e Rollup merge of #36600 - caipre:sort-by-key-comment-fix, r=apasel422
Minor correction in `sort_by_key` doc comment

The parameter to `sort_by_key` is named `f`, not `key`.
2016-09-22 11:25:02 -07:00
Jonathan Turner
5d4c0edb69 Rollup merge of #36589 - jpadkins:option_docs_safety_wording_fix, r=bluss
fixed the safety header/wording in option.rs

Fixes #36581

screenshot of the rendered documentation: http://imgur.com/14kLVrA

r? @steveklabnik
2016-09-22 11:25:01 -07:00
Jonathan Turner
a86c83f08a Rollup merge of #36571 - kmcallister:gh-29372-rc-docs, r=GuillaumeGomez
Tweak std::rc docs

Fixes #29372.

r? @GuillaumeGomez
2016-09-22 11:25:01 -07:00
Jonathan Turner
3056630d8f Rollup merge of #36539 - mikhail-m1:36530, r=jonathandturner
fix top level attr spans

issue #36530
?r @jonathandturner
2016-09-22 11:25:01 -07:00
Jonathan Turner
b60fc5d16a Rollup merge of #36423 - GuillaumeGomez:eq_impl, r=pnkfelix
Add missing Eq implementations

Part of #36301.
2016-09-22 11:25:01 -07:00
Jonathan Turner
b474c82add Rollup merge of #36330 - aclarry:e0560-update, r=jonathandturner
Updated E0560 to new error format

Added a small bit of code to add a label for E0560.

Also renamed src/test/compile-fail/E560.rs file to E0560.rs (not sure why it was named E560).
Updated all test cases which check this error.

Closes #36199
2016-09-22 11:25:00 -07:00
Corey Farwell
6b697a375e Implement Debug for DirEntry. 2016-09-22 14:08:03 -04:00
Eduard Burtescu
795b6adc9a Don't let a type parameter named "Self" unchanged past HIR lowering. 2016-09-22 20:03:28 +03:00
bors
458f411213 Auto merge of #36261 - parched:soft-float, r=pnkfelix
Fix documentation with 'soft-float' codegen option

This option doesn't cause software FP routines
to be called, it only changes the float ABI.

Additionally, this option is ignored by all targets,
except the ARM eabihf ones.
2016-09-22 09:50:42 -07:00
mcarton
bfa3433283 Fix compile-fail syntax in error file 2016-09-22 18:32:40 +02:00
bors
a09cb57a73 Auto merge of #36618 - jseyfried:crate_root_attr_invoc, r=nrc
macros: allow attribute invocations at the crate root

Fixes #36617.
r? @nrc
2016-09-22 06:29:23 -07:00
Gianni Ciccarelli
a35275c89b #36553 specialisation error 502 is misleading
this commit correct E0502 error text.
See referenced issue for further info
2016-09-22 12:28:46 +02:00
bors
b2627b080a Auto merge of #36592 - nnethercote:TypedArena, r=bluss
Lazily allocate TypedArena's first chunk

Currently `TypedArena` allocates its first chunk, which is usually 4096
bytes, as soon as it is created. If no allocations are ever made from
the arena then this allocation (and the corresponding deallocation) is
wasted effort.

This commit changes `TypedArena` so it doesn't allocate the first chunk
until the first allocation is made.

This change speeds up rustc by a non-trivial amount because rustc uses
`TypedArena` heavily: compilation speed (producing debug builds) on
several of the rustc-benchmarks increases by 1.02--1.06x. The change
should never cause a slow-down because the hot `alloc` function is
unchanged. It does increase the size of `TypedArena` by one `usize`
field, however.

The commit also fixes some out-of-date comments.
2016-09-22 03:13:50 -07:00
Jeffrey Seyfried
272cf4e61d Refactor no_implicit_prelude: Cell<bool> -> no_implicit_prelude: bool. 2016-09-22 07:45:11 +00:00
Jeffrey Seyfried
dc7ed303f7 Refactor out parse_struct_expr. 2016-09-22 07:34:03 +00:00
Jeffrey Seyfried
85bfd82be6 Avoid reconstructing the BuildReducedGraphVisitor. 2016-09-22 07:18:16 +00:00
Jeffrey Seyfried
a344f14caa Refactor out BuildReducedGraphVisitor::visit_trait_item. 2016-09-22 07:17:23 +00:00
Jeffrey Seyfried
9a0e88a339 Refactor away ParentLink. 2016-09-22 07:13:54 +00:00
bors
6ad10844db Auto merge of #36585 - jonathandturner:misc_error_touchups, r=nrc
Add the ability to merge spans to codemap

This PR adds the ability to merge Spans.  To do so, it builds on the Codemap's ability to verify the locations of spans, namely that following can be verified:

* the expn_id of both spans much match
* the lhs span needs to end on the same line the rhs span begins
* the lhs span must start at or before the rhs span

If all of these are met, a new span is returned that is min(lo), max(hi) of the two spans.

This PR also removes an older Span merge, as this new functionality subsumes it.

r? @nrc
2016-09-21 22:40:51 -07:00
bors
1cf592fa40 Auto merge of #36551 - eddyb:meta-games, r=nikomatsakis
Refactor away RBML from rustc_metadata.

RBML and `ty{en,de}code` have had their long-overdue purge. Summary of changes:
* Metadata is now a tree encoded in post-order and with relative backward references pointing to children nodes. With auto-deriving and type safety, this makes maintenance and adding new information to metadata painless and bug-free by default. It's also more compact and cache-friendly (cache misses should be proportional to the depth of the node being accessed, not the number of siblings as in EBML/RBML).
* Metadata sizes have been reduced, for `libcore` it went down 16% (`8.38MB` -> `7.05MB`) and for `libstd` 14% (`3.53MB` -> `3.03MB`), while encoding more or less the same information
* Specialization is used in the bundled `libserialize` (crates.io `rustc_serialize` remains unaffected) to customize the encoding (and more importantly, decoding) of various types, most notably those interned in the `TyCtxt`. Some of this abuses a soundness hole pending a fix (cc @aturon), but when that fix arrives, we'll move to macros 1.1 `#[derive]` and custom `TyCtxt`-aware serialization traits.
* Enumerating children of modules from other crates is now orthogonal to describing those items via `Def` - this is a step towards bridging crate-local HIR and cross-crate metadata
* `CrateNum` has been moved to `rustc` and both it and `NodeId` are now newtypes instead of `u32` aliases, for specializing their decoding. This is `[syntax-breaking]` (cc @Manishearth ).

cc @rust-lang/compiler
2016-09-21 19:17:24 -07:00
bors
4f9812a59a Auto merge of #36496 - pnkfelix:workaround-issue-34427, r=eddyb
Workaround #34427 by using memset of 0 on ARM to set the discriminant.

Workaround #34427 by using memset of 0 on ARM to set the discriminant.
2016-09-21 15:58:39 -07:00
Jeffrey Seyfried
f4fa62f4f2 Add regression test. 2016-09-21 21:39:41 +00:00
Nick Cameron
6a2d2c9495 Adds a ProcMacro form of syntax extension
This commit adds syntax extension forms matching the types for procedural macros 2.0 (RFC #1566), these still require the usual syntax extension boiler plate, but this is a first step towards proper implementation and should be useful for macros 1.1 stuff too.

Supports both attribute-like and function-like macros.
2016-09-22 08:47:57 +12:00
Christopher Serr
e107c8b849 Fix outdated Doc Comment on BufReader::seek
A long time ago non-panicking `unwrap` methods were renamed to `into_inner` in this Pull Request: https://github.com/rust-lang/rust/pull/19149

Looks like this doc comment was not updated however.
2016-09-21 21:11:53 +02:00
Keegan McAllister
c316ae56e6 Tweak std::rc docs
Fixes #29372.
2016-09-21 10:25:01 -07:00
bors
53f9730291 Auto merge of #35074 - ashleygwilliams:assert_ne, r=steveklabnik
add debug_assert_ne + assert_ne

~~ work in progress, please do not merge ~~

fixes #35073
2016-09-21 08:12:37 -07:00
Ashley Williams
3d8d55787b add assert_ne and debug_assert_ne macros 2016-09-21 14:38:26 +02:00
Jeffrey Seyfried
4db157af71 Allow attribute macro invocations at the crate root. 2016-09-21 09:24:06 +00:00
bors
5cc6c6b1b7 Auto merge of #36524 - michaelwoerister:trans-inline-only-on-demand, r=nikomatsakis
trans: Only instantiate #[inline] functions in codegen units referencing them

This PR changes how `#[inline]` functions are translated. Before, there was one "master instance" of the function with `external` linkage and a number of on-demand instances with `available_externally` linkage in each codegen unit that referenced the function. This had two downsides:

* Public functions marked with `#[inline]` would be present in machine code of libraries unnecessarily (see #36280 for an example)
* LLVM would crash on `i686-pc-windows-msvc` due to what I suspect to be a bug in LLVM's Win32 exception handling code, because it doesn't like `available_externally` there (#36309).

This PR changes the behavior, so that there is no master instance and only on-demand instances with `internal` linkage. The downside of this is potential code-bloat if LLVM does not completely inline away the `internal` instances because then there'd be N instances of the function instead of 1. However, this can only become a problem when using more than one codegen unit per crate.

cc @rust-lang/compiler
2016-09-21 01:33:37 -07:00
Jonathan Turner
e4b18422ad Check for overlapping and simplify unit test 2016-09-20 15:39:05 -07:00
aclarry
23efc5453d Update E0560 to include label 2016-09-20 16:59:48 -04:00
Jorge Aparicio
adf85d35c7 libtest: add a --skip flag to the test runner
This flag takes a FILTER argument and instructs the test runner to skip
the tests whose names contain the word FILTER. --skip can be used
several times.
2016-09-20 14:27:48 -05:00
Eduard Burtescu
4ac30013c3 rustc_trans: don't do on-demand drop glue instantiation. 2016-09-20 20:30:55 +03:00
Eduard Burtescu
a2726f4a54 rustc: allow less and handle fn pointers in the type hashing algorithm. 2016-09-20 20:30:55 +03:00
Eduard Burtescu
ade79d7609 rustc_trans: simplify vtable and symbol handling. 2016-09-20 20:30:52 +03:00
Eduard Burtescu
521d3ea193 rustc_resolve: bring back "struct called like a function" cross-crate. 2016-09-20 20:08:07 +03:00
Eduard Burtescu
564f2ee33c rustc_metadata: don't die with --test because CFG_VERSION is missing. 2016-09-20 20:08:07 +03:00
Eduard Burtescu
a23b8cafec rustc: don't hash the --extern crate name, but the original one, in DefPath. 2016-09-20 20:08:06 +03:00
Eduard Burtescu
221d1a97e5 serialize: allow specifying the default behavior for specializations. 2016-09-20 20:08:06 +03:00
Eduard Burtescu
b01d4891e2 rustc: don't recurse through nested items in decoded HIR fragments. 2016-09-20 20:08:06 +03:00
Eduard Burtescu
f2283a7be0 rustc_metadata: fix for the new ? ambiguity around collect. 2016-09-20 20:08:06 +03:00
Eduard Burtescu
dadbaa48ac rustc_metadata: move opt_item_name to TyCtxt::item_name. 2016-09-20 20:08:06 +03:00
Eduard Burtescu
dc26a23301 rustc_metadata: reduce Lazy{,Seq} overhead by using a relative encoding. 2016-09-20 20:08:05 +03:00