Commit Graph

368 Commits

Author SHA1 Message Date
Patrick Walton
a8f039a085 librustc: Make monomorphic newtype structs work cross-crate 2013-02-11 20:09:11 -08:00
Jed Davis
3742b62f64 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.
2013-02-09 22:28:43 -08:00
bors
d6442e97a5 auto merge of #4854 : thestinger/rust/oldmap, r=catamorphism 2013-02-09 15:02:51 -08:00
bors
d90abd5879 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
2013-02-09 10:20:55 -08:00
Daniel Micay
a32c5c73ee oldmap: get rid of legacy _ref suffixes 2013-02-08 19:12:51 -05:00
Niko Matsakis
a380df809c 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.
2013-02-08 07:20:39 -08:00
Jed Davis
9318babf6c 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.
2013-02-08 01:52:55 -08:00
Jed Davis
dd7a81d025 Let llsize_of be a ConstantInt 2013-02-08 01:52:47 -08:00
Jed Davis
e89d9853a6 Fix vector indexing in consts so it works with the enum vector const fix. 2013-02-07 18:31:12 -08:00
Jed Davis
de8dc02634 Make tuple-like structs containing enums work as constants. 2013-02-07 18:31:12 -08:00
Jed Davis
30aae3d910 Fix const enum type issues for structs. 2013-02-07 18:31:12 -08:00
Jed Davis
29b99669dc Fix pointer consts to work with enums 2013-02-07 18:28:31 -08:00
Jed Davis
9ad616a102 Make vectors of enums work as constants 2013-02-07 18:28:11 -08:00
Jed Davis
6cdc283415 Also need to pad out "C-like" enum consts (paths as well as calls). 2013-02-07 17:57:02 -08:00
Jed Davis
877fc8d891 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.
2013-02-07 17:57:02 -08:00
Jed Davis
52cf61fd3b Fix const enums better: let them have the same alignment as an "opaque enum". 2013-02-07 17:57:02 -08:00
Patrick Walton
472797b04a librustc: Lots of de-muting. rs=demuting 2013-02-07 16:17:39 -08:00
bors
2bc9655bc1 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...
2013-02-07 15:20:16 -08:00
Niko Matsakis
a32498d846 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.
2013-02-07 05:53:30 -08:00
bors
82d7396333 auto merge of #4823 : pcwalton/rust/enum-variant-discriminants, r=graydon
r? @graydon
2013-02-07 04:02:51 -08:00
bors
f13ea4121e auto merge of #4795 : catamorphism/rust/less-copy, r=catamorphism 2013-02-06 17:06:15 -08:00
Patrick Walton
cf6c3d96fb librustc: Attempt to put out burning tree by fixing translation of unary negation in boolean constants. rs=burningtree 2013-02-06 15:38:23 -08:00
Patrick Walton
6d13c90256 librustc: Stop loading enum variant discriminants from memory 2013-02-06 15:08:33 -08:00
Patrick Walton
b34f871dda librustc: Change i1 to i8 for bools. Attempts to put out burning tree. rs=burningtree 2013-02-06 14:28:02 -08:00
Tim Chevalier
739e5ba369 rustc: Less copy 2013-02-06 12:09:11 -08:00
Patrick Walton
801f3225b2 oldmap: use &K instead of K in find and get
This reverts commit a4250a96fd.

This is not the cause of the nonexhaustive-match failure.
2013-02-05 19:41:45 -08:00
Graydon Hoare
a4250a96fd Revert "oldmap: use &K instead of K in find and get"
This reverts commit 8e643525d4.
2013-02-05 14:30:53 -08:00
Alex Crichton
b368cb2341 Don't warn when imported traits are indeed used 2013-02-05 12:41:19 -05:00
Alex Crichton
6a4483ec7a Fix some bugs with -W unused-imports
1. Don't warn about anything not used in the prelude which is autmoatically
   injected, accomplished with a test that the span is equal to a dummy span.
2. Don't warn about unused imports from the injected intrinsic module,
   accomplished by testing against the name of the imported module
3. If anything is used from a glob import, don't warn about the glob import.
4. If an import imports more than one thing, and none of them are used, only
   issue a warning once

Also updated the unused-imports-warn test to have stricter requirements on
error messages.
2013-02-05 12:40:59 -05:00
bors
52817092a9 auto merge of #4747 : sanxiyn/rust/integral-type, r=graydon
`ty::type_is_integral` returns `true` for `ty_bool`. This causes `-true` to compile, instead of resulting in a type error.
2013-02-04 19:59:06 -08:00
bors
bffe3088c7 auto merge of pull req #4777 from thestinger/rust, r=graydon 2013-02-04 17:24:40 -08:00
bors
3c6b9c2363 automated merge 2013-02-04 15:39:10 -08:00
Seo Sanghyeon
6d5428cb4e Let type_is_integral return false for bool type 2013-02-04 23:01:47 +09:00
ILyoan
6105f09cc1 Use x86-64 CABI for foreign function on arm target 2013-02-04 19:30:17 +09:00
Erick Tryzelaar
9adfa59d8e core: convert ToStr::to_str to take explicit &self 2013-02-03 20:47:26 -08:00
Daniel Micay
8e643525d4 oldmap: use &K instead of K in find and get 2013-02-03 23:30:56 -05:00
Brian Anderson
3b396d17d6 Merge remote-tracking branch 'thestinger/old_map' into incoming
Conflicts:
	src/test/bench/core-map.rs
2013-02-03 17:56:49 -08:00
Patrick Walton
04eb9b4eb0 librustc: Fix bugs regarding to impl privacy. rs=bugfix 2013-02-03 17:50:27 -08:00
Marvin Löbel
eb19462104 Converted libcore/uint-template.rs to the new string functions.
- Moved ToStr implementation of unsigned integers to uint-template.rs.
- Marked the `str()` function as deprecated.
- Forwarded all conversion functions to `core::num::to_str_common()`
  and `core::num::from_str_common()`.
- Fixed most places in the codebase where `to_str()` is being used.
- Added uint-template to_str and from_str overflow tests.
2013-02-03 15:37:24 -08:00
Marvin Löbel
26e72bf92b Converted libcore/int-template.rs to the new string functions.
- Moved ToStr implementation of integers to int-template.rs.
- Marked the `str()` function as deprecated.
- Forwarded all conversion functions to `core::num::to_str_common()`
  and `core::num::from_str_common()`.
- Fixed most places in the codebase where `to_str()` is being used.
- Added int-template to_str and from_str overflow tests.
2013-02-03 15:37:24 -08:00
Daniel Micay
4fd9264875 oldmap: &K instead of K for the remove parameter 2013-02-03 18:20:59 -05:00
Daniel Micay
81b4f36d49 oldmap: remove legacy each method 2013-02-03 15:55:11 -05:00
Daniel Micay
2e496818a5 oldmap: get rid of the legacy each_key method 2013-02-03 15:55:10 -05:00
Daniel Micay
88d9d417a1 oldmap: remove the legacy each_value method 2013-02-03 15:55:10 -05:00
Daniel Micay
1b4eb145f9 oldmap: implement core::container::Container 2013-02-03 15:55:10 -05:00
Daniel Micay
f4a27b2c7d oldmap: get rid of the legacy contains_key method 2013-02-03 15:55:10 -05:00
Daniel Micay
319eeb1c79 rename map -> oldmap and mark it as deprecated
LinearMap is quite a bit faster, and is fully owned/sendable without
requiring copies. The older std::map also doesn't use explicit self and
relies on mutable fields.
2013-02-03 15:55:10 -05:00
Brian Anderson
9dc6938292 Merge remote-tracking branch 'nickdesaulniers/issue4524' into nocomm1 2013-02-01 11:18:58 -08:00
Nick Desaulniers
7868b6bf55 Remove fail keyword from lexer & parser and clean up remaining calls to
fail

Fix merge conflicts - Issue 4524
2013-02-01 00:15:42 -08:00
John Clements
53688addaa test cases, cleanup 2013-01-31 23:05:12 -08:00