]> git.lizzy.rs Git - rust.git/log
rust.git
11 years agort: remove last_os_error from rustrt.def.in
Luqman Aden [Tue, 12 Feb 2013 05:22:58 +0000 (00:22 -0500)]
rt: remove last_os_error from rustrt.def.in

11 years agoLong lines.
Luqman Aden [Tue, 12 Feb 2013 04:49:20 +0000 (23:49 -0500)]
Long lines.

11 years agort: remove last_os_error and adjust tests.
Luqman Aden [Tue, 12 Feb 2013 04:27:12 +0000 (23:27 -0500)]
rt: remove last_os_error and adjust tests.

11 years agocore: Rewrite last_os_error in Rust for windows as well.
Luqman Aden [Tue, 12 Feb 2013 04:07:29 +0000 (23:07 -0500)]
core: Rewrite last_os_error in Rust for windows as well.

11 years agocore: Rewrite last_os_error in Rust for unix and provide access to errno (unix) and...
Luqman Aden [Tue, 12 Feb 2013 01:07:25 +0000 (20:07 -0500)]
core: Rewrite last_os_error in Rust for unix and provide access to errno (unix) and GetLastError (windows).

11 years agolibrustc: Make monomorphic newtype structs work cross-crate
Patrick Walton [Tue, 12 Feb 2013 00:28:39 +0000 (16:28 -0800)]
librustc: Make monomorphic newtype structs work cross-crate

11 years agostd: Fix uv_tcp_t size on i686-apple-darwin
Brian Anderson [Tue, 12 Feb 2013 02:25:11 +0000 (18:25 -0800)]
std: Fix uv_tcp_t size on i686-apple-darwin

11 years agoMerge remote-tracking branch 'thestinger/treemap'
Brian Anderson [Tue, 12 Feb 2013 00:13:50 +0000 (16:13 -0800)]
Merge remote-tracking branch 'thestinger/treemap'

11 years agoAdd Mikko Perttunen to AUTHORS
Brian Anderson [Mon, 11 Feb 2013 23:02:16 +0000 (15:02 -0800)]
Add Mikko Perttunen to AUTHORS

11 years agoFix license block
Brian Anderson [Mon, 11 Feb 2013 21:06:37 +0000 (13:06 -0800)]
Fix license block

11 years agoMerge remote-tracking branch 'cyndis/expansion-span'
Brian Anderson [Mon, 11 Feb 2013 21:02:24 +0000 (13:02 -0800)]
Merge remote-tracking branch 'cyndis/expansion-span'

11 years agoMerge pull request #4873 from jld/rust-mode-wants-cl-when
Brian Anderson [Mon, 11 Feb 2013 21:06:57 +0000 (13:06 -0800)]
Merge pull request #4873 from jld/rust-mode-wants-cl-when

rust-mode.el uses the 'cl macros, so it should actually require them

11 years agoUpdate copyright years
Mikko Perttunen [Mon, 11 Feb 2013 19:02:36 +0000 (21:02 +0200)]
Update copyright years

11 years agoAdd test for using line! in a macro
Mikko Perttunen [Mon, 11 Feb 2013 18:26:40 +0000 (20:26 +0200)]
Add test for using line! in a macro

11 years agoUse topmost span for macro expansion location. Fixes behaviour of file!, line! and...
Mikko Perttunen [Mon, 11 Feb 2013 16:13:18 +0000 (18:13 +0200)]
Use topmost span for macro expansion location. Fixes behaviour of file!, line! and col!

11 years agotreemap: cut down on swap_unwrap in remove
Daniel Micay [Mon, 11 Feb 2013 03:03:26 +0000 (22:03 -0500)]
treemap: cut down on swap_unwrap in remove

Performance before:

    std::treemap::TreeMap
                   sequential_ints 0.083971 s
                       random_ints 0.095861 s
                       delete_ints 0.083931 s
                sequential_strings 0.278272 s
                    random_strings 0.240286 s
                    delete_strings 0.173581 s

Performance after:

    std::treemap::TreeMap
                   sequential_ints 0.083297 s
                       random_ints 0.097644 s
                       delete_ints 0.052602 s
                sequential_strings 0.287326 s
                    random_strings 0.242372 s
                    delete_strings 0.142269 s

11 years agoavoid explicit reborrow in heir_swap
Daniel Micay [Mon, 11 Feb 2013 02:11:33 +0000 (21:11 -0500)]
avoid explicit reborrow in heir_swap

11 years agotreemap: avoid swap_unwrap in insert
Daniel Micay [Mon, 11 Feb 2013 01:44:15 +0000 (20:44 -0500)]
treemap: avoid swap_unwrap in insert

Performance before:

    std::treemap::TreeMap
                   sequential_ints 0.151877 s
                       random_ints 0.160926 s
                       delete_ints 0.08694 s
                sequential_strings 0.316458 s
                    random_strings 0.290778 s
                    delete_strings 0.169892 s

Performance after:

    std::treemap::TreeMap
                   sequential_ints 0.083971 s
                       random_ints 0.095861 s
                       delete_ints 0.083931 s
                sequential_strings 0.278272 s
                    random_strings 0.240286 s
                    delete_strings 0.173581 s

11 years agotreemap: use an &mut parameter for skew and split
Daniel Micay [Mon, 11 Feb 2013 01:37:21 +0000 (20:37 -0500)]
treemap: use an &mut parameter for skew and split

results in a small performance improvement and reduces the compiled
code size

11 years agostd: fix libuv structs on macos
Jeff Olson [Sun, 10 Feb 2013 22:30:07 +0000 (14:30 -0800)]
std: fix libuv structs on macos

11 years agobuild: add link flag for freebsd libuv build
Jeff Olson [Fri, 8 Feb 2013 05:08:26 +0000 (21:08 -0800)]
build: add link flag for freebsd libuv build

11 years agobuild: change libuv to point at joyent's repo and make unpatched build libuv work...
Jeff Olson [Fri, 8 Feb 2013 00:50:43 +0000 (16:50 -0800)]
build: change libuv to point at joyent's repo and make unpatched build libuv work on mingw

11 years agostd: strip sillyness from debug comment in iotask teardown
Jeff Olson [Tue, 5 Feb 2013 02:00:21 +0000 (18:00 -0800)]
std: strip sillyness from debug comment in iotask teardown

11 years agobuild: ifdef for mingw/non-mingw builds
Jeff Olson [Sat, 2 Feb 2013 20:14:04 +0000 (12:14 -0800)]
build: ifdef for mingw/non-mingw builds

11 years agobuild: point libuv at olsonjeffery/libuv, temporarily
Jeff Olson [Mon, 4 Feb 2013 22:37:48 +0000 (14:37 -0800)]
build: point libuv at olsonjeffery/libuv, temporarily

11 years agobuild: tweak rt.mk, as per graydon.. CFLAGS had to stay
Jeff Olson [Sat, 2 Feb 2013 15:40:40 +0000 (07:40 -0800)]
build: tweak rt.mk, as per graydon.. CFLAGS had to stay

11 years agort/std: update of libuv API glue for libuv submodule update
Jeff Olson [Wed, 30 Jan 2013 18:28:27 +0000 (10:28 -0800)]
rt/std: update of libuv API glue for libuv submodule update

11 years agoetc: rework of how libuv is integrated into the build
Jeff Olson [Wed, 30 Jan 2013 17:56:54 +0000 (09:56 -0800)]
etc: rework of how libuv is integrated into the build

- thanks to work in libuv's upstream, we can call libuv's Makefile directly
with parameters, instead of descending in gyp-uv madness and generating
our own.

11 years agorust-mode.el uses the 'cl macros, so it should actually require them
Jed Davis [Sun, 10 Feb 2013 19:16:54 +0000 (11:16 -0800)]
rust-mode.el uses the 'cl macros, so it should actually require them

Without this change, rust-mode doesn't work if 'cl hasn't been required
by something else, apparently.  I'm not entirely sure what changed such
that I started seeing this problem instead of not, but maybe the emacs
world has been making progress towards not loading 'cl at runtime if
it's only needed at compile time.

11 years agoauto merge of #4866 : jld/rust/enum-unitlike, r=graydon
bors [Sun, 10 Feb 2013 18:24:17 +0000 (10:24 -0800)]
auto merge of #4866 : jld/rust/enum-unitlike, r=graydon

If an enum is isomorphic to unit, there's no need to use any bits to
represent it.  The only obvious reason this wasn't the case was because
the enum could be C-like and have a user-specified discriminant -- but
that value is constant, so it doesn't need to be stored.

This change means that all newtype-like enums have the same size (and
layout) as their underlying type, which might be a useful property to
have, at least in terms of making programs' low-level behavior less
surprising.

11 years agoOmit discriminant from nullary univariant enums.
Jed Davis [Mon, 7 Jan 2013 11:42:49 +0000 (03:42 -0800)]
Omit discriminant from nullary univariant enums.

If an enum is isomorphic to unit, there's no need to use any bits to
represent it.  The only obvious reason this wasn't the case was because
the enum could be C-like and have a user-specified discriminant -- but
that value is constant, so it doesn't need to be stored.

This change means that all newtype-like enums have the same size (and
layout) as their underlying type, which might be a useful property to
have, at least in terms of making programs' low-level behavior less
surprising.

11 years agoauto merge of #4861 : sethpink/rust/incoming, r=catamorphism
bors [Sun, 10 Feb 2013 02:37:02 +0000 (18:37 -0800)]
auto merge of #4861 : sethpink/rust/incoming, r=catamorphism

Fix for issue #4830.

11 years agoauto merge of #4827 : mcpherrinm/rust/master, r=catamorphism
bors [Sun, 10 Feb 2013 01:28:04 +0000 (17:28 -0800)]
auto merge of #4827 : mcpherrinm/rust/master, r=catamorphism

It seems to me the library needs more work to be done, but having a non-compilable sample program seems like bad news.

11 years agoauto merge of #4854 : thestinger/rust/oldmap, r=catamorphism
bors [Sat, 9 Feb 2013 23:02:51 +0000 (15:02 -0800)]
auto merge of #4854 : thestinger/rust/oldmap, r=catamorphism

11 years agoauto merge of #4853 : Thiez/rust/incoming, r=catamorphism
bors [Sat, 9 Feb 2013 22:13:59 +0000 (14:13 -0800)]
auto merge of #4853 : Thiez/rust/incoming, r=catamorphism

A simple fix for issue 2174.

11 years agoauto merge of #4852 : Caseus/rust/issue-4836, r=brson
bors [Sat, 9 Feb 2013 19:07:13 +0000 (11:07 -0800)]
auto merge of #4852 : Caseus/rust/issue-4836, r=brson

Simple removal of rust_compare_and_swap_ptr from rt, as per Issue #4836

11 years agoauto merge of #4848 : nikomatsakis/rust/issue-4821-bad-kind-computation, r=catamorphism
bors [Sat, 9 Feb 2013 18:20:55 +0000 (10:20 -0800)]
auto merge of #4848 : nikomatsakis/rust/issue-4821-bad-kind-computation, r=catamorphism

...ear

values to be copied.  Rewrite kind computation so that instead of directly
computing the kind it computes what kinds of values are present in the type,
and then derive kinds based on that. I find this easier to think about.

Fixes #4821.

r? @catamorphism

11 years agoIssue #4830 fix
Seth Pink [Sat, 9 Feb 2013 13:00:55 +0000 (23:00 +1000)]
Issue #4830 fix

11 years agolibrustc/back/rpath.rs: oldmap -> LinearSet
Daniel Micay [Sat, 9 Feb 2013 02:04:28 +0000 (21:04 -0500)]
librustc/back/rpath.rs: oldmap -> LinearSet

11 years agolibsyntax/attr.rs: switch from oldmap to LinearSet
Daniel Micay [Sat, 9 Feb 2013 01:54:30 +0000 (20:54 -0500)]
libsyntax/attr.rs: switch from oldmap to LinearSet

11 years agoRemove trailing whitespace
Matthew McPherrin [Sat, 9 Feb 2013 01:17:50 +0000 (20:17 -0500)]
Remove trailing whitespace

11 years agooldmap: rm unneeded unsafe
Daniel Micay [Sat, 9 Feb 2013 00:28:37 +0000 (19:28 -0500)]
oldmap: rm unneeded unsafe

11 years agooldmap: separate out the methods that need Copy
Daniel Micay [Sat, 9 Feb 2013 00:20:36 +0000 (19:20 -0500)]
oldmap: separate out the methods that need Copy

11 years agort: remove unused 'rust_compare_and_swap_ptr'. Closes #4836
Mark Vian [Sat, 9 Feb 2013 00:12:03 +0000 (18:12 -0600)]
rt: remove unused 'rust_compare_and_swap_ptr'. Closes #4836

11 years agooldmap: get rid of legacy _ref suffixes
Daniel Micay [Fri, 8 Feb 2013 22:08:02 +0000 (17:08 -0500)]
oldmap: get rid of legacy _ref suffixes

11 years agoAdded related FIXME for 3260
Matthijs Hofstra [Sat, 9 Feb 2013 00:01:39 +0000 (01:01 +0100)]
Added related FIXME for 3260

11 years agoFix for issue 2174
Matthijs Hofstra [Fri, 8 Feb 2013 23:45:00 +0000 (00:45 +0100)]
Fix for issue 2174

The function that formats and prints the squigly line that hilights
errors counted tabs as spaces, which resulted in incorrect error
messages when tabs were used for indentation. This change compares
the highlight line with the previous line and inserts a tab instead
of a space whenever such a tab exists on the previous line. Note
that error messages will still highlight incorrectly when the
previous line include characters that require more than one utf8
code point, as mentioned in issue 3260.

11 years agoauto merge of #4847 : nikomatsakis/rust/region-syntax, r=graydon
bors [Fri, 8 Feb 2013 22:25:52 +0000 (14:25 -0800)]
auto merge of #4847 : nikomatsakis/rust/region-syntax, r=graydon

cc #4846

r? @pcwalton

11 years agoMerge branch 'master' of https://github.com/mozilla/rust
Matthew McPherrin [Fri, 8 Feb 2013 20:21:46 +0000 (15:21 -0500)]
Merge branch 'master' of https://github.com/mozilla/rust

11 years agoauto merge of #4844 : jld/rust/const-index-fix, r=graydon
bors [Fri, 8 Feb 2013 18:53:16 +0000 (10:53 -0800)]
auto merge of #4844 : jld/rust/const-index-fix, r=graydon

Fixes #3169 and uses that to fix const array indexing, which I broke (in a way that doesn't break tests but will greatly confuse users) while trying to fix const enum vectors.

11 years agoauto merge of #4845 : osaut/rust/master, r=graydon
bors [Fri, 8 Feb 2013 16:47:45 +0000 (08:47 -0800)]
auto merge of #4845 : osaut/rust/master, r=graydon

There was a simple typo on recursive in doc/tutorial-macros.md: s/recusive/recursive.

11 years agoFix subtle error in caching during kind computation that could cause linear
Niko Matsakis [Fri, 8 Feb 2013 03:33:12 +0000 (19:33 -0800)]
Fix subtle error in caching during kind computation that could cause linear
values to be copied.  Rewrite kind computation so that instead of directly
computing the kind it computes what kinds of values are present in the type,
and then derive kinds based on that. I find this easier to think about.

Fixes #4821.

11 years agoAdd and lex LIFETIME tokens
Niko Matsakis [Fri, 8 Feb 2013 14:02:35 +0000 (06:02 -0800)]
Add and lex LIFETIME tokens

cc #4846

11 years agoTypo s/recusive/recursive/
Olivier Saut [Fri, 8 Feb 2013 10:45:47 +0000 (11:45 +0100)]
Typo s/recusive/recursive/

11 years agoFix const array index limit calculation.
Jed Davis [Fri, 8 Feb 2013 09:52:55 +0000 (01:52 -0800)]
Fix const array index limit calculation.

The number of operands of the LLVM node initializing the array
underlying a const vector isn't always the array length -- if the
array is of a sufficiently primitive type and all the elements' values
are known (or something like that), LLVM uses a specialized Constant
subclass that stores the data packed, and thus has no operands.  Oops.

But, because llsize_of now gives us a ConstantInt, we can just fix
mozilla/rust#3169 and this all goes away.

11 years agoLet llsize_of be a ConstantInt
Jed Davis [Fri, 8 Feb 2013 09:52:47 +0000 (01:52 -0800)]
Let llsize_of be a ConstantInt

11 years agoauto merge of #4843 : ILyoan/rust/i1531, r=catamorphism
bors [Fri, 8 Feb 2013 08:10:49 +0000 (00:10 -0800)]
auto merge of #4843 : ILyoan/rust/i1531, r=catamorphism

Fix for #1531

11 years agoauto merge of #4839 : catamorphism/rust/rm-structural-records, r=catamorphism
bors [Fri, 8 Feb 2013 07:16:42 +0000 (23:16 -0800)]
auto merge of #4839 : catamorphism/rust/rm-structural-records, r=catamorphism

11 years agocore: Remove structural records from the rest of core, except pipes
Tim Chevalier [Fri, 8 Feb 2013 02:23:47 +0000 (18:23 -0800)]
core: Remove structural records from the rest of core, except pipes

That will, sadly, require one more snapshot.

11 years agoauto merge of #4835 : thestinger/rust/iterator, r=graydon
bors [Fri, 8 Feb 2013 05:37:59 +0000 (21:37 -0800)]
auto merge of #4835 : thestinger/rust/iterator, r=graydon

This adds a `BaseIter` impl to `PriorityQueue`, `TreeMap`, `LinearMap` and `SmallIntMap`, and introduces a `ReverseIter` trait + implementations for `TreeMap`, `TreeSet` and `SmallIntMap`.

11 years agoFixed #1531
ILyoan [Fri, 8 Feb 2013 05:10:55 +0000 (14:10 +0900)]
Fixed #1531

11 years agofix map tests
Daniel Micay [Fri, 8 Feb 2013 03:58:16 +0000 (22:58 -0500)]
fix map tests

11 years agoauto merge of #4834 : veddan/rust/zflags, r=graydon
bors [Fri, 8 Feb 2013 03:50:03 +0000 (19:50 -0800)]
auto merge of #4834 : veddan/rust/zflags, r=graydon

Converted --static, --gc, --jit, -g and --xg to -Z flags.

11 years agoimplement BaseIter for LinearMap
Daniel Micay [Fri, 8 Feb 2013 02:03:13 +0000 (21:03 -0500)]
implement BaseIter for LinearMap

11 years agoauto merge of #4745 : jld/rust/constenum-bis, r=graydon
bors [Fri, 8 Feb 2013 03:04:44 +0000 (19:04 -0800)]
auto merge of #4745 : jld/rust/constenum-bis, r=graydon

There were a bunch of problems with consts where an enum was contained within some other type (vector, tuple, struct, etc.); some of these would cause LLVM assertion failures, and some would silently read from the wrong address.  These changes should fix all of that.

It would be good if someone with access to a win32 host could do the equivalent of `make check-stageN-rpass TESTNAME=enum` on that platform before merging this.

11 years agoimplement ReverseIter for SmallIntMap
Daniel Micay [Fri, 8 Feb 2013 01:20:39 +0000 (20:20 -0500)]
implement ReverseIter for SmallIntMap

11 years agoimplement ReverseIter for TreeMap and TreeSet
Daniel Micay [Fri, 8 Feb 2013 00:39:46 +0000 (19:39 -0500)]
implement ReverseIter for TreeMap and TreeSet

11 years agoadd a ReverseIter trait
Daniel Micay [Fri, 8 Feb 2013 00:35:18 +0000 (19:35 -0500)]
add a ReverseIter trait

11 years agoimplement BaseIter for SmallIntMap
Daniel Micay [Fri, 8 Feb 2013 00:13:24 +0000 (19:13 -0500)]
implement BaseIter for SmallIntMap

11 years agomake Option's map and map_default use a lifetime
Daniel Micay [Thu, 7 Feb 2013 23:30:34 +0000 (18:30 -0500)]
make Option's map and map_default use a lifetime

11 years agoimplement BaseIter for TreeMap
Daniel Micay [Thu, 7 Feb 2013 23:05:27 +0000 (18:05 -0500)]
implement BaseIter for TreeMap

11 years agorm each method from the Map trait
Daniel Micay [Thu, 7 Feb 2013 22:49:57 +0000 (17:49 -0500)]
rm each method from the Map trait

the map types should implement BaseIter instead

11 years agoFix vector indexing in consts so it works with the enum vector const fix.
Jed Davis [Thu, 31 Jan 2013 07:59:50 +0000 (23:59 -0800)]
Fix vector indexing in consts so it works with the enum vector const fix.

11 years agoMake tuple-like structs containing enums work as constants.
Jed Davis [Thu, 31 Jan 2013 07:59:06 +0000 (23:59 -0800)]
Make tuple-like structs containing enums work as constants.

11 years agoFix const enum type issues for structs.
Jed Davis [Thu, 31 Jan 2013 01:51:57 +0000 (17:51 -0800)]
Fix const enum type issues for structs.

11 years agoAdd a test for vstore, which already works
Jed Davis [Thu, 31 Jan 2013 00:52:01 +0000 (16:52 -0800)]
Add a test for vstore, which already works

11 years agoFix pointer consts to work with enums
Jed Davis [Thu, 31 Jan 2013 00:42:46 +0000 (16:42 -0800)]
Fix pointer consts to work with enums

11 years agoMake vectors of enums work as constants
Jed Davis [Thu, 31 Jan 2013 00:36:07 +0000 (16:36 -0800)]
Make vectors of enums work as constants

11 years agocore: Remove transitional code
Tim Chevalier [Fri, 8 Feb 2013 02:22:58 +0000 (18:22 -0800)]
core: Remove transitional code

11 years agoAlso need to pad out "C-like" enum consts (paths as well as calls).
Jed Davis [Thu, 31 Jan 2013 00:08:39 +0000 (16:08 -0800)]
Also need to pad out "C-like" enum consts (paths as well as calls).

11 years agoPad out enum consts to the expected size; makes enums in tuples work.
Jed Davis [Wed, 30 Jan 2013 23:34:52 +0000 (15:34 -0800)]
Pad out enum consts to the expected size; makes enums in tuples work.

This is wasted space if the const is just an enum, but optimizing that
case without breaking everything else is an issue that can be addressed
separately.

11 years agoFix const enums better: let them have the same alignment as an "opaque enum".
Jed Davis [Wed, 30 Jan 2013 23:17:23 +0000 (15:17 -0800)]
Fix const enums better: let them have the same alignment as an "opaque enum".

11 years agoadd a BaseIter implementation for PriorityQueue
Daniel Micay [Thu, 7 Feb 2013 22:08:04 +0000 (17:08 -0500)]
add a BaseIter implementation for PriorityQueue

11 years agoauto merge of #4833 : pcwalton/rust/demuting, r=pcwalton
bors [Fri, 8 Feb 2013 00:23:59 +0000 (16:23 -0800)]
auto merge of #4833 : pcwalton/rust/demuting, r=pcwalton

11 years agolibrustc: Lots of de-muting. rs=demuting
Patrick Walton [Mon, 4 Feb 2013 22:02:01 +0000 (14:02 -0800)]
librustc: Lots of de-muting. rs=demuting

11 years agoAdded a few missing 'pub's
Viktor Dahl [Thu, 7 Feb 2013 23:57:31 +0000 (00:57 +0100)]
Added a few missing 'pub's

11 years agoauto merge of #4803 : alexcrichton/rust/fix-unused-imports, r=graydon
bors [Thu, 7 Feb 2013 23:20:16 +0000 (15:20 -0800)]
auto merge of #4803 : alexcrichton/rust/fix-unused-imports, r=graydon

The first commit message has most of the comments, but this pull request basically fixes a lot of issues surrounding the `unused_imports` warning/deny attribute.

Before this patch there were these problems:

1. Unused imports from `prelude.rs` were warned about with dummy spans, leading to a large number of confusing warnings.
2. Unused imports from `intrinsic.rs` were warned about with the file `<intrinsic>` which couldn't be forced to go away
3. Methods used from imported traites (like `io::WriterUtil`) resulted in an unused warning of the import even though it was used.
4. If one `use` statement imported N modules, M of which weren't used, M warning statements were issued.
5. If a glob import statement was used, each public export of the target module which wasn't used had a warning issued.

This patch deals with all these cases by doing:

1. Ignore unused imports from `prelude.rs` (indicated by a dummy span of 0)
2. Ignore unused imports from `intrinsic.rs` (test on the imported module name, is there a better way?)
3. Track when imported modules are used as candidates for methods, and just assume they're used. This may not end up being the actual case, but in theory not warning about an unused thing is worse than warning about a used thing.
4. Only issue one warning statement
5. Only issue one warning statement.

This is the first time I've edited the compiler itself, and I tried to keep up with the style around, but I may have missed something here or there...

11 years agoMoved experimental compiler flags to -Z (#4740)
Viktor Dahl [Thu, 7 Feb 2013 23:14:17 +0000 (00:14 +0100)]
Moved experimental compiler flags to -Z (#4740)

11 years agoauto merge of #4791 : jbclements/rust/demodeing-and-deGCing, r=jbclements,brson
bors [Thu, 7 Feb 2013 21:58:32 +0000 (13:58 -0800)]
auto merge of #4791 : jbclements/rust/demodeing-and-deGCing, r=jbclements,brson

r?

It looks to me like the string_reader and tt_reader structs are
GC pointers only because they predate the modern borrow system.
This commit leaves the type names string_reader and tt_reader alone
(they still refer to GC-ed pointers), but internally the functions
now use borrowed pointers to refer to these structures. My guess
would be that it's possible to move this change outward and not
use the GCed pointers at all, but that change looks like it could be
a larger one. Actually, I'm delighted at how quick this change was.

11 years agoMerge pull request #4619 from brson/exchange
Brian Anderson [Thu, 7 Feb 2013 21:46:10 +0000 (13:46 -0800)]
Merge pull request #4619 from brson/exchange

Some work on freestanding Rust: foreign calls, exchange allocator

11 years agoauto merge of #4832 : sanxiyn/rust/pretty-exact, r=graydon
bors [Thu, 7 Feb 2013 20:17:32 +0000 (12:17 -0800)]
auto merge of #4832 : sanxiyn/rust/pretty-exact, r=graydon

11 years agoauto merge of #4814 : brson/rust/check, r=nikomatsakis
bors [Thu, 7 Feb 2013 18:46:34 +0000 (10:46 -0800)]
auto merge of #4814 : brson/rust/check, r=nikomatsakis

r? @nikomatsakis The third commit here is the one that makes the test suite remember which steps have completed successfully.

11 years agoChange pretty-exact to pp-exact
Seo Sanghyeon [Thu, 7 Feb 2013 17:56:49 +0000 (02:56 +0900)]
Change pretty-exact to pp-exact

11 years agoauto merge of #4831 : bjz/rust/incoming, r=pcwalton
bors [Thu, 7 Feb 2013 16:32:09 +0000 (08:32 -0800)]
auto merge of #4831 : bjz/rust/incoming, r=pcwalton

This is useful for comparing more complex types that include floats.

11 years agoFix broken tests
Brendan Zabarauskas [Thu, 7 Feb 2013 15:41:23 +0000 (02:41 +1100)]
Fix broken tests

11 years agoauto merge of #4810 : nikomatsakis/rust/owned-fn-noncopyable, r=nikomatsakis
bors [Thu, 7 Feb 2013 14:05:10 +0000 (06:05 -0800)]
auto merge of #4810 : nikomatsakis/rust/owned-fn-noncopyable, r=nikomatsakis

Part of #3678.  Fixes #2828, #3904.

r? @brson

11 years agoMake ~fn non-copyable, make &fn copyable, split barefn/closure types,
Niko Matsakis [Fri, 1 Feb 2013 01:12:29 +0000 (17:12 -0800)]
Make ~fn non-copyable, make &fn copyable, split barefn/closure types,
correct handling of moves for struct-record update.

Part of #3678.  Fixes #2828, #3904, #4719.

11 years agoauto merge of #4823 : pcwalton/rust/enum-variant-discriminants, r=graydon
bors [Thu, 7 Feb 2013 12:02:51 +0000 (04:02 -0800)]
auto merge of #4823 : pcwalton/rust/enum-variant-discriminants, r=graydon

r? @graydon

11 years agoMerge branch 'incoming' of https://github.com/mozilla/rust into incoming
Brendan Zabarauskas [Thu, 7 Feb 2013 11:55:23 +0000 (22:55 +1100)]
Merge branch 'incoming' of https://github.com/mozilla/rust into incoming

11 years agoAdd type parameter for epsilon value
Brendan Zabarauskas [Thu, 7 Feb 2013 11:54:52 +0000 (22:54 +1100)]
Add type parameter for epsilon value