bors [Fri, 29 May 2015 19:24:40 +0000 (19:24 +0000)]
Auto merge of #25494 - alexcrichton:stdio-from-raw, r=aturon
This commit implements a number of standard traits for the standard library's
process I/O handles. The `FromRaw{Fd,Handle}` traits are now implemented for the
`Stdio` type and the `AsRaw{Fd,Handle}` traits are now implemented for the
`Child{Stdout,Stdin,Stderr}` types.
The stability markers for these implementations mention that they are stable for
1.1 as I will nominate this commit for cherry-picking to beta.
bors [Fri, 29 May 2015 17:38:40 +0000 (17:38 +0000)]
Auto merge of #25880 - nikomatsakis:const-fn-feature-gate-calls, r=alexcrichton
The previous feature gate assumed we would not define any (stable) const fns. But then @eddyb went and cleaned up the code. So this now extends the feature-gate to prohibit calls; but calls inside of macros are considered ok.
bors [Fri, 29 May 2015 07:07:07 +0000 (07:07 +0000)]
Auto merge of #25747 - SimonSapin:map_ref, r=alexcrichton
For slightly complex data structures like `rustc_serialize::json::Json`, it is often convenient to have helper methods like `Json::as_string(&self) -> Option<&str>` that return a borrow of some component of `&self`.
However, when `RefCell`s are involved, keeping a `Ref` around is required to hold a borrow to the insides of a `RefCell`. But `Ref` so far only references the entirety of the contents of a `RefCell`, not a component. But there is no reason it couldn’t: `Ref` internally contains just a data reference and a borrow count reference. The two can be dissociated.
This adds a `map_ref` function that creates a new `Ref` for some other data, but borrowing the same `RefCell` as an existing `Ref`.
bors [Thu, 28 May 2015 21:48:40 +0000 (21:48 +0000)]
Auto merge of #25744 - SimonSapin:cell-eq, r=alexcrichton
`core::cell::Cell<T>` and `core::cell::RefCell<T>` currently implement `PartialEq` when `T` does, and just defer to comparing `T` values. There is no reason the same shouldn’t apply to `Eq`.
This enables `#[derive(Eq, PartialEq)]` on e.g. structs that have a `RefCell` field.
r? @alexcrichton
I’m unsure what to do with `#[stable]` attributes on `impl`s. `impl`s generated by `#[derive]` don’t have them.
bors [Thu, 28 May 2015 20:16:08 +0000 (20:16 +0000)]
Auto merge of #25856 - bluss:binary-heap-hole, r=Gankro
collections: Make BinaryHeap panic safe in sift_up / sift_down
Use a struct called Hole that keeps track of an invalid location
in the vector and fills the hole on drop.
I include a run-pass test that the current BinaryHeap fails, and the new
one passes.
NOTE: The BinaryHeap will still be inconsistent after a comparison fails. It will
not have the heap property. What we fix is just that elements will be valid
values.
This is actually a performance win -- the new code does not bother to write in `zeroed()`
values in the holes, it just leaves them as they were.
Net result is something like a 5% decrease in runtime for `BinaryHeap::from_vec`. This
can be further improved by using unchecked indexing (I confirmed it makes a difference,
not a surprise with the non-sequential access going on), but let's leave that for another PR.
Safety first :wink:
bors [Thu, 28 May 2015 02:17:48 +0000 (02:17 +0000)]
Auto merge of #25824 - alexcrichton:fix-deadlocking-test-on-windows, r=nikomatsakis
Windows tests can often deadlock if a child thread continues after the main
thread and then panics, and a `println!` executed in a child thread after the
main thread has exited is at risk of panicking.
Rollup merge of #25792 - koobs:patch-1, r=alexcrichton
Currently, there are two conditional blocks that exist to check for "clang or gcc"
On line 866:
```
if [ -z "$CFG_ENABLE_CLANG" -a -z "$CFG_GCC" ]
then
err "either clang or gcc is required"
fi
```
and on line 1019:
```
if [ -z "$CC" -a -z "$CFG_ENABLE_CLANG" -a -z "$CFG_GCC" ]
then
err "either clang or gcc is required"
fi
```
Given the order of the clauses, this results in the "either clang or gcc is required" error from the earlier block, (even) when CC is set.
The expected behaviour is to honour user-flags, in this case CC.
Aside from removing all hand-holdy compiler checks in favour of actual compiler *feature* checks, this change removes the redundant former block in favour of the latter block, which appears designed to allow the expected behaviour.
bors [Wed, 27 May 2015 17:25:00 +0000 (17:25 +0000)]
Auto merge of #25799 - alexcrichton:fix-link-in-mk, r=luqmana
The changes scaled back in 4cc025d8 were a little too aggressive and broke a
bunch of cross compilations by not defining the `LINK_$(1)` variable for all
targets. This commit ensures that the variable is defined for all targets by
defaulting it to the normal compiler if it's not already defined (it's only
defined specially for MSVC).
Alex Crichton [Wed, 27 May 2015 16:15:19 +0000 (09:15 -0700)]
test: Join child threads on windows in tests
Windows tests can often deadlock if a child thread continues after the main
thread and then panics, and a `println!` executed in a child thread after the
main thread has exited is at risk of panicking.
bors [Wed, 27 May 2015 10:21:11 +0000 (10:21 +0000)]
Auto merge of #25762 - dotdash:codegen_test, r=alexcrichton
The current codegen tests only compare IR line counts between similar
rust and C programs, the latter getting compiled with clang. That looked
like a good idea back then, but actually things like lifetime intrinsics
mean that less IR isn't always better, so the metric isn't really
helpful.
Instead, we can start doing tests that check specific aspects of the
generated IR, like attributes or metadata. To do that, we can use LLVM's
FileCheck tool which has a number of useful features for such tests.
To start off, I created some tests for a few things that were recently
added and/or broken.
Björn Steinbrink [Sun, 24 May 2015 16:07:52 +0000 (18:07 +0200)]
Revamp codegen tests to check IR quality instead of quantity
The current codegen tests only compare IR line counts between similar
rust and C programs, the latter getting compiled with clang. That looked
like a good idea back then, but actually things like lifetime intrinsics
mean that less IR isn't always better, so the metric isn't really
helpful.
Instead, we can start doing tests that check specific aspects of the
generated IR, like attributes or metadata. To do that, we can use LLVM's
FileCheck tool which has a number of useful features for such tests.
To start off, I created some tests for a few things that were recently
added and/or broken.
bors [Wed, 27 May 2015 07:14:45 +0000 (07:14 +0000)]
Auto merge of #25791 - barosl:use-paths-as-items, r=alexcrichton
Currently, for `use` declarations with multiple paths, only the `use` item itself is saved in the AST map, not the individual path nodes. This can lead to a problem when a span of a specific path node is needed.
For example, #24818 caused an ICE because of this, in `ImportResolver::check_for_conflicting_import()`.
bors [Wed, 27 May 2015 00:42:55 +0000 (00:42 +0000)]
Auto merge of #25713 - Stebalien:pattern, r=alexcrichton
Needed to support:
```rust
match X {
pattern if Y ...
}
for pattern in Y {}
```
IMO, this shouldn't require an RFC because it can't interfere with any future language changes (because `pattern if` and `pattern in` are already legal in rust) and can't cause any ambiguity.
bors [Tue, 26 May 2015 23:10:06 +0000 (23:10 +0000)]
Auto merge of #25688 - nham:E0055_E0192, r=alexcrichton
- Adds explanations for E0055, E0089, E0192, E0261-E0263, E0318
- Improves explanations for E0250, E0368, E0372.
- Converts 15 diagnostics to have error codes (E0380-E0394). Adds an explanation for E0380.
- The E0087-E0090 messages currently look like "expected {} parameter(s) found {} parameter(s)". This changes them to either use "parameter" or "parameters", based on the number.
bors [Tue, 26 May 2015 18:30:42 +0000 (18:30 +0000)]
Auto merge of #25593 - nham:E0210_E0326, r=alexcrichton
Changes:
- adds explanations for E0185, E0186, E0202, E0326
- fixes the explanation for E0053. The previous description was too narrow; there are other error cases.
- changes the error message for E0202 to be specific for associated types, since it seems inherent associated constants are implemented.
Alex Crichton [Tue, 26 May 2015 16:48:24 +0000 (09:48 -0700)]
mk: Ensure LINK_$(1) is defined for all targets
The changes scaled back in 4cc025d8 were a little too aggressive and broke a
bunch of cross compilations by not defining the `LINK_$(1)` variable for all
targets. This commit ensures that the variable is defined for all targets by
defaulting it to the normal compiler if it's not already defined (it's only
defined specially for MSVC).
bors [Tue, 26 May 2015 16:58:07 +0000 (16:58 +0000)]
Auto merge of #25091 - quantheory:trait_associated_const_fixes, r=nikomatsakis
Closes #25046 (by rejecting the code that causes the ICE) and #24946. I haven't been able to deal with the array size or recursion issues yet for associated consts, though my hope was that the change I made for range match patterns might help with array sizes, too.
bors [Tue, 26 May 2015 14:08:24 +0000 (14:08 +0000)]
Auto merge of #25411 - alexcrichton:issue-18913, r=brson
This flag indicates that when files are being replaced or added to archives (the
`r` flag) that the new file should not be inserted if it is not newer than the
file that already exists in the archive. The compiler never actually has a use
case of *not* wanting to insert a file because it already exists, and this
causes rlibs to not be updated in some cases when the compiler was re-run too
quickly.
Alex Crichton [Thu, 14 May 2015 18:53:37 +0000 (11:53 -0700)]
rustc_back: Don't pass 'u' to ar invocations
This flag indicates that when files are being replaced or added to archives (the
`r` flag) that the new file should not be inserted if it is not newer than the
file that already exists in the archive. The compiler never actually has a use
case of *not* wanting to insert a file because it already exists, and this
causes rlibs to not be updated in some cases when the compiler was re-run too
quickly.
bors [Tue, 26 May 2015 12:15:30 +0000 (12:15 +0000)]
Auto merge of #25785 - arielb1:cleanup-201505, r=eddyb
The caching essentially eliminates "stability checking" time (my attempt to clean-up junk got tangled up with stability, so I added the caching while I was at it).
bors [Tue, 26 May 2015 09:49:36 +0000 (09:49 +0000)]
Auto merge of #25778 - econoplas:master, r=pnkfelix
A regression was introduced by commit https://github.com/rust-lang/rust/commit/7b1916d25347913fce3e336517ef22025ccd875f #25612. Negative signed integer literals less than -9223372036854775808i64 were no longer properly reported as #[warn(overflowing_literals)].
Also adding missing test cases to test/compile-fail/lint-type-overflow.rs which could have detected the regression.
Further explanation:
The expression `(negative && v > max as u64 + 1)` relies on the fact that algebraically speaking `-min == max + 1` to avoid negation and removing the need for `min` completely.
If i128 or i256 are ever added, it should also work for these types without requiring a change to `min != i64::MIN &&` also simplifying maintenance.
Barosl Lee [Tue, 26 May 2015 02:27:15 +0000 (11:27 +0900)]
Associate each path in a `use` declaration with the item in the AST map
Currently, for `use` declarations with multiple paths, only the `use`
item itself is saved in the AST map, not the individual path nodes. This
can lead to a problem when a span of a specific path node is needed.
For example, #24818 caused an ICE because of this, in
`ImportResolver::check_for_conflicting_import()`.