This commit brings the local_data api up to modern rust standards with a few key
improvements:
* All functionality is now exposed as a method on the keys themselves. Instead
of importing std::local_data, you now use "key.set()" and "key.get()".
* All closures have been removed in favor of RAII functionality. This means that
get() and get_mut() no long require closures, but rather return
Option<SmartPointer> where the smart pointer takes care of relinquishing the
borrow and also implements the necessary Deref traits
* The modify() function was removed to cut the local_data interface down to its
bare essentials (similarly to how RefCell removed set/get).
[breaking-change]
This commit brings the local_data api up to modern rust standards with a few key
improvements:
* The `pop` and `set` methods have been combined into one method, `replace`
* The `get_mut` method has been removed. All interior mutability should be done
through `RefCell`.
* All functionality is now exposed as a method on the keys themselves. Instead
of importing std::local_data, you now use "key.replace()" and "key.get()".
* All closures have been removed in favor of RAII functionality. This means that
get() and get_mut() no long require closures, but rather return
Option<SmartPointer> where the smart pointer takes care of relinquishing the
borrow and also implements the necessary Deref traits
* The modify() function was removed to cut the local_data interface down to its
bare essentials (similarly to how RefCell removed set/get).
[breaking-change]
Previously, the parser would not allow you to simultaneously implement a
function with a different abi as well as being unsafe at the same time. This
extends the parser to allow functions of the form:
unsafe extern fn foo() {
// ...
}
The closure type grammar was also changed to reflect this reversal, types
previously written as "extern unsafe fn()" must now be written as
"unsafe extern fn()". The parser currently has a hack which allows the old
style, but this will go away once a snapshot has landed.
Closes#10025
[breaking-change]
for `~str`/`~[]`.
Note that `~self` still remains, since I forgot to add support for
`Box<self>` before the snapshot.
How to update your code:
* Instead of `~EXPR`, you should write `box EXPR`.
* Instead of `~TYPE`, you should write `Box<Type>`.
* Instead of `~PATTERN`, you should write `box PATTERN`.
[breaking-change]
Previously, the parser would not allow you to simultaneously implement a
function with a different abi as well as being unsafe at the same time. This
extends the parser to allow functions of the form:
unsafe extern fn foo() {
// ...
}
The closure type grammar was also changed to reflect this reversal, types
previously written as "extern unsafe fn()" must now be written as
"unsafe extern fn()". The parser currently has a hack which allows the old
style, but this will go away once a snapshot has landed.
Closes#10025
[breaking-change]
Currently, rustc requires that a linkage be a product of 100% rlibs or 100%
dylibs. This is to satisfy the requirement that each object appear at most once
in the final output products. This is a bit limiting, and the upcoming libcore
library cannot exist as a dylib, so these rules must change.
The goal of this commit is to enable *some* use cases for mixing rlibs and
dylibs, primarily libcore's use case. It is not targeted at allowing an
exhaustive number of linkage flavors.
There is a new dependency_format module in rustc which calculates what format
each upstream library should be linked as in each output type of the current
unit of compilation. The module itself contains many gory details about what's
going on here.
cc #10729
The code in resolve erroneously assumed that private enums weren't visited, so
the logic was adjusted to check to see if the enum definition itself was public.
Closes#11680
By carefully distinguishing falling back to the default arm from moving
on to the next pattern, this patch adjusts the codegen logic for range
and guarded arms of pattern matching expression. It is a more
appropriate way of fixing #12582 and #13027 without causing regressions
such as #13867.
Closes#13867
This patch fixes issue #13186.
When generating constant expression for enum, it is possible that
alignment of expression may be not equal to alignment of type. In that
case space after last struct field must be padded to match size of value
and size of struct. This commit adds that padding.
See detailed explanation in src/test/run-pass/trans-tag-static-padding.rs
This was removed because these could alias with `&const T` or `@mut T`
and those are now gone from the language. There are still aliasing
issues within local scopes, but this is correct for function parameters.
This also removes the no-op `noalias` marker on proc (not a pointer) and
leaves out the mention of #6750 because real type-based alias analysis
is not within the scope of best effort usage of the `noalias` attribute.
Test case:
pub fn foo(x: &mut &mut u32) {
**x = 5;
**x = 5;
}
Before:
define void @_ZN3foo20h0ce94c9671b0150bdaa4v0.0E(i32** nocapture readonly) unnamed_addr #0 {
entry-block:
%1 = load i32** %0, align 8
store i32 5, i32* %1, align 4
%2 = load i32** %0, align 8
store i32 5, i32* %2, align 4
ret void
}
After:
define void @_ZN3foo20h0ce94c9671b0150bdaa4v0.0E(i32** noalias nocapture readonly) unnamed_addr #0 {
entry-block:
%1 = load i32** %0, align 8
store i32 5, i32* %1, align 4
ret void
}
Closes#12436
By carefully distinguishing falling back to the default arm from moving
on to the next pattern, this patch adjusts the codegen logic for range
and guarded arms of pattern matching expression. It is a more
appropriate way of fixing #12582 and #13027 without causing regressions
such as #13867.
Closes#13867
Some cases were not correctly handled by this lint, for instance `let a = 42u8; a < 0` and `let a = 42u8; a > 255`.
It led to the discovery of two useless comparisons, which I removed.
This has long since not been too relevant since the introduction of many crate
type outputs. This commit removes the flag entirely, adjusting all logic to do
the most reasonable thing when building both a library and an executable.
Closes#13337
This was removed because these could alias with `&const T` or `@mut T`
and those are now gone from the language. There are still aliasing
issues within local scopes, but this is correct for function parameters.
This also removes the no-op `noalias` marker on proc (not a pointer) and
leaves out the mention of #6750 because real type-based alias analysis
is not within the scope of best effort usage of the `noalias` attribute.
Test case:
pub fn foo(x: &mut &mut u32) {
**x = 5;
**x = 5;
}
Before:
define void @_ZN3foo20h0ce94c9671b0150bdaa4v0.0E(i32** nocapture readonly) unnamed_addr #0 {
entry-block:
%1 = load i32** %0, align 8
store i32 5, i32* %1, align 4
%2 = load i32** %0, align 8
store i32 5, i32* %2, align 4
ret void
}
After:
define void @_ZN3foo20h0ce94c9671b0150bdaa4v0.0E(i32** noalias nocapture readonly) unnamed_addr #0 {
entry-block:
%1 = load i32** %0, align 8
store i32 5, i32* %1, align 4
ret void
}
Closes#12436
This ensures that private functions exported through static initializers will
actually end up being public in the object file (so other objects can continue
to reference the function).
Closes#13620
Closes#7575.
I don't think the change from a contains lookup to an iteration of the HashSet in the resolver should be much of a burden as the set of methods with the same name should be relatively small.
This is a first patch towards an opt-in built-in trait world. This patch removes the restriction on built-in traits and allows such traits to be derived.
[RFC#3]
cc #13231
@nikomatsakis r?
This ensures that private functions exported through static initializers will
actually end up being public in the object file (so other objects can continue
to reference the function).
Closes#13620
This has long since not been too relevant since the introduction of many crate
type outputs. This commit removes the flag entirely, adjusting all logic to do
the most reasonable thing when building both a library and an executable.
Closes#13337
Currently, rustc requires that a linkage be a product of 100% rlibs or 100%
dylibs. This is to satisfy the requirement that each object appear at most once
in the final output products. This is a bit limiting, and the upcoming libcore
library cannot exist as a dylib, so these rules must change.
The goal of this commit is to enable *some* use cases for mixing rlibs and
dylibs, primarily libcore's use case. It is not targeted at allowing an
exhaustive number of linkage flavors.
There is a new dependency_format module in rustc which calculates what format
each upstream library should be linked as in each output type of the current
unit of compilation. The module itself contains many gory details about what's
going on here.
cc #10729
Pre-step towards issue #12624 and others: Introduce ExprUseVisitor, remove the
moves computation. ExprUseVisitor is a visitor that walks the AST for a
function and calls a delegate to inform it where borrows, copies, and moves
occur.
In this patch, I rewrite the gather_loans visitor to use ExprUseVisitor, but in
future patches, I think we could rewrite regionck, check_loans, and possibly
other passes to use it as well. This would refactor the repeated code between
those places that tries to determine where copies/moves/etc occur.
r? @alexcrichton
In the process, `Splits` got changed to be more like `CharSplits` in `str` to present the DEI interface.
Note that `treemap` still has a `rev_iter` function because it seems like it would be a significant interface change to expose a DEI - the iterator would have to gain an extra pointer, the completion checks would be more complicated, and it isn't easy to check that such an implementation is correct due to the use of unsafety to subvert the aliasing properties of `&mut`.
This fixes#9391.
This commit deprecates rev_iter, mut_rev_iter, move_rev_iter everywhere (except treemap) and also
deprecates related functions like rsplit, rev_components, and rev_str_components. In every case,
these functions can be replaced with the non-reversed form followed by a call to .rev(). To make this
more concrete, a translation table for all functional changes necessary follows:
* container.rev_iter() -> container.iter().rev()
* container.mut_rev_iter() -> container.mut_iter().rev()
* container.move_rev_iter() -> container.move_iter().rev()
* sliceorstr.rsplit(sep) -> sliceorstr.split(sep).rev()
* path.rev_components() -> path.components().rev()
* path.rev_str_components() -> path.str_components().rev()
In terms of the type system, this change also deprecates any specialized reversed iterator types (except
in treemap), opting instead to use Rev directly if any type annotations are needed. However, since
methods directly returning reversed iterators are now discouraged, the need for such annotations should
be small. However, in those cases, the general pattern for conversion is to take whatever follows Rev in
the original reversed name and surround it with Rev<>:
* RevComponents<'a> -> Rev<Components<'a>>
* RevStrComponents<'a> -> Rev<StrComponents<'a>>
* RevItems<'a, T> -> Rev<Items<'a, T>>
* etc.
The reasoning behind this change is that it makes the standard API much simpler without reducing readability,
performance, or power. The presence of functions such as rev_iter adds more boilerplate code to libraries
(all of which simply call .iter().rev()), clutters up the documentation, and only helps code by saving two
characters. Additionally, the numerous type synonyms that were used to make the type signatures look nice
like RevItems add even more boilerplate and clutter up the docs even more. With this change, all that cruft
goes away.
[breaking-change]
Similar to my recent changes to ~[T]/&[T], these changes remove the vstore abstraction and represent str types as ~(str) and &(str). The Option<uint> in ty_str is the length of the string, None if the string is dynamically sized.
It reflected the obsolete syntax `use a, b, c;` and did not make past the parser (though it was a non-fatal error so we can continue). This legacy affected many portions of rustc and rustdoc as well, so this PR cleans them up altogether.
As a side effect of cleanup, we now have `SCHEMA_VERSION` in `rustdoc::clean` (instead of the crate root), so it has a better chance to be updated when `rustdoc::clean` gets updated.
it reflected the obsolete syntax `use a, b, c;` and did not make
past the parser (though it was a non-fatal error so we can continue).
this legacy affected many portions of rustc and rustdoc as well,
so this commit cleans them up altogether.
This addresses the ICE from #13763, but it does not allow the test to compile,
due to #13768. An alternate test was checked in in the meantime.
Closes#13763
moves computation. ExprUseVisitor is a visitor that walks the AST for a
function and calls a delegate to inform it where borrows, copies, and moves
occur.
In this patch, I rewrite the gather_loans visitor to use ExprUseVisitor, but in
future patches, I think we could rewrite regionck, check_loans, and possibly
other passes to use it as well. This would refactor the repeated code between
those places that tries to determine where copies/moves/etc occur.
Specifically, the method parameter cardinality mismatch or missing
method error message span now gets method itself exactly. It was the
whole expression.
Closes#9390Closes#13684Closes#13709
This patch removes the special auto-rooting for `@` from the borrow checker. With `@` moving into a library, it doesn't make sense to keep this code around anymore. It also simplifies `trans` by removing root checking from there
@nikomatsakis
Closes: #11586
Specifically, the method parameter cardinality mismatch or missing
method error message span now gets method itself exactly. It was the
whole expression.
Closes#9390Closes#13684Closes#13709
This alters the borrow checker's requirements on invoking closures from
requiring an immutable borrow to requiring a unique immutable borrow. This means
that it is illegal to invoke a closure through a `&` pointer because there is no
guarantee that is not aliased. This does not mean that a closure is required to
be in a mutable location, but rather a location which can be proven to be
unique (often through a mutable pointer).
For example, the following code is unsound and is no longer allowed:
type Fn<'a> = ||:'a;
fn call(f: |Fn|) {
f(|| {
f(|| {})
});
}
fn main() {
call(|a| {
a();
});
}
There is no replacement for this pattern. For all closures which are stored in
structures, it was previously allowed to invoke the closure through `&self` but
it now requires invocation through `&mut self`.
The standard library has a good number of violations of this new rule, but the
fixes will be separated into multiple breaking change commits.
Closes#12224
This alters the borrow checker's requirements on invoking closures from
requiring an immutable borrow to requiring a unique immutable borrow. This means
that it is illegal to invoke a closure through a `&` pointer because there is no
guarantee that is not aliased. This does not mean that a closure is required to
be in a mutable location, but rather a location which can be proven to be
unique (often through a mutable pointer).
For example, the following code is unsound and is no longer allowed:
type Fn<'a> = ||:'a;
fn call(f: |Fn|) {
f(|| {
f(|| {})
});
}
fn main() {
call(|a| {
a();
});
}
There is no replacement for this pattern. For all closures which are stored in
structures, it was previously allowed to invoke the closure through `&self` but
it now requires invocation through `&mut self`.
The standard library has a good number of violations of this new rule, but the
fixes will be separated into multiple breaking change commits.
Closes#12224
[breaking-change]
`@` pointers used to have special rooting and regions management. With `@`
moving to standalone library, we don't need to keep that special
treatment around. This patch modifies the way `@` pointers are treated by
treating them as if they were `~` pointers
Region checker and borrow checker were modified in this patch.
Closes#11586
[breaking-change]
This exposes volatile versions of the memset/memmove/memcpy intrinsics.
The volatile parameter must be constant, so this can't simply be a
parameter to our intrinsics.
Now with proper checking of enums and allows unsized fields as the last field in a struct or variant. This PR only checks passing of unsized types and distinguishing them from sized ones. To be safe we also need to control storage.
Closes issues #12969 and #13121, supersedes #13375 (all the discussion there is valid here too).
This currently requires linking against a library like libquadmath (or
libgcc), because compiler-rt barely has any support for this and most
hardware does not yet have 128-bit precision floating point. For this
reason, it's currently hidden behind a feature gate.
When compiler-rt is updated to trunk, some tests can be added for
constant evaluation since there will be support for the comparison
operators.
Closes#13381
This exposes volatile versions of the memset/memmove/memcpy intrinsics.
The volatile parameter must be constant, so this can't simply be a
parameter to our intrinsics.
This patch adds a special rule for `Unsafe<T>` and makes it `Share`
regardless of whether T is `Share`.
[breaking-change]
Closes#13125
cc @nikomatsakis
When reporting "consider removing this semicolon" hint message, the
offending semicolon may come from macro call site instead of macro
itself. Using the more appropriate span makes the hint more helpful.
Closes#13428.
This gives a better NOTE error message when a privacy error is encountered with
a static method. Previously no note was emitted (due to lack of support), but
now a note is emitted indicating that the struct/enum itself is private.
Closes#13641
Refactors all uses of ty_vec and associated things to remove the vstore abstraction (still used for strings, for now). Pointers to vectors are stored as ty_rptr or ty_uniq wrapped around a ty_vec. There are no user-facing changes. Existing behaviour is preserved by special-casing many instances of pointers containing vectors. Hopefully with DST most of these hacks will go away. For now it is useful to leave them hanging around rather than abstracting them into a method or something.
Closes#13554.
This gives a better NOTE error message when a privacy error is encountered with
a static method. Previously no note was emitted (due to lack of support), but
now a note is emitted indicating that the struct/enum itself is private.
Closes#13641
This commit removes the compiler support for floating point modulus operations,
as well as from the language. An implementation for this operator is now
required to be provided by libraries.
Floating point modulus is rarely used, doesn't exist in C, and is always lowered
to an fmod library call by LLVM, and LLVM is considering removing support
entirely.
Closes#12278
Refactores all uses of ty_vec and associated things to remove the vstore abstraction (still used for strings, for now). Pointers to vectors are stored as ty_rptr or ty_uniq wrapped around a ty_vec. There are no user-facing changes. Existing behaviour is preserved by special-casing many instances of pointers containing vectors. Hopefully with DST most of these hacks will go away. For now it is useful to leave them hanging around rather than abstracting them into a method or something.
Closes#13554.
In upgrading LLVM, only rust functions had the "split-stack" attribute added.
This commit changes the addition of LLVM's "split-stack" attribute to *always*
occur and then we remove it sometimes if the "no_split_stack" rust attribute is
present.
Closes#13625
This is a bit of an interesting upgrade to LLVM. Upstream LLVM has started using C++11 features, so they require a C++11 compiler to build. I've updated all the bots to have a C++11 compiler, and they appear to be building LLVM successfully:
* Linux bots - I added gcc/g++ 4.7 (good enough)
* Android bots - same as the linux ones
* Mac bots - I installed the most recent command line tools for Lion which gives us clang 3.2, but LLVM wouldn't build unless it was explicitly asked to link to `libc++` instead of `libstdc++`. This involved tweaking `mklldeps.py` and the `configure` script to get things to work out
* Windows bots - mingw-w64 has gcc 4.8.1 which is sufficient for building LLVM (hurray!)
* BSD bots - I updated FreeBSD to 10.0 which brought with it a relevant version of clang.
The largest fallout I've seen so far is that the test suite doesn't work at all on FreeBSD 10. We've already stopped gating on FreeBSD due to #13427 (we used to be on freebsd 9), so I don't think this puts us in too bad of a situation. I will continue to attempt to fix FreeBSD and the breakage on there.
The LLVM update brings with it all of the recently upstreamed LLVM patches. We only have one local patch now which is just an optimization, and isn't required to use upstream LLVM. I want to maintain compatibility with LLVM 3.3 and 3.4 while we can, and this upgrade is keeping us up to date with the 3.5 release. Once 3.5 is release we will in theory no longer require a bundled LLVM.
When reporting "consider removing this semicolon" hint message, the
offending semicolon may come from macro call site instead of macro
itself. Using the more appropriate span makes the hint more helpful.
Closes#13428.
This version of `is_to_be_inferred` double-checks the result from
`inferred_map` by querying the `named_region_map` and `ast_map` and
then asserts that the `inferred_map` state is consistent with its own
findings. (See issue 13261 for further discussion of the approaches).
This comes with a number of fixes to be compatible with upstream LLVM:
* Previously all monomorphizations of "mem::size_of()" would receive the same
symbol. In the past LLVM would silently rename duplicated symbols, but it
appears to now be dropping the duplicate symbols and functions now. The symbol
names of monomorphized functions are now no longer solely based on the type of
the function, but rather the type and the unique hash for the
monomorphization.
* Split stacks are no longer a global feature controlled by a flag in LLVM.
Instead, they are opt-in on a per-function basis through a function attribute.
The rust #[no_split_stack] attribute will disable this, otherwise all
functions have #[split_stack] attached to them.
* The compare and swap instruction now takes two atomic orderings, one for the
successful case and one for the failure case. LLVM internally has an
implementation of calculating the appropriate failure ordering given a
particular success ordering (previously only a success ordering was
specified), and I copied that into the intrinsic translation so the failure
ordering isn't supplied on a source level for now.
* Minor tweaks to LLVM's API in terms of debuginfo, naming, c++11 conventions,
etc.
Part of this required added an override of `fold_type_method` in the
Folder for Ctx impl; it follows the same pattern as `fold_method`.
Also, as a drive-by fix, I moved all of the calls to `folder.new_id`
in syntax::fold's no-op default traversal to really be the first
statement in each function.
* This is to uphold the invariant that `folder.new_id` is always
called first (an unfortunate requirement of the current `ast_map`
code), an invariant that we seemingly were breaking in e.g. the
previous `noop_fold_block`.
* Now it should be easier to see when adding new code that this
invariant must be upheld.
* (note that the breakage in `noop_fold_block` may not have mattered
so much previously, since the only thing that blocks can bind are
lifetimes, which I am only adding support for now.)
Before adding a variance constrant for a given early-bound param,
check if it was meant to be inferred.
To support the above, added `fn is_to_be_inferred` to
`variance::ConstraintContext`.
When instantiating trait default methods for certain implementation,
`typeck` correctly combined type parameters from trait bound with those
from method bound, but didn't do so for lifetime parameters. Applies
the same logic to lifetime parameters.
Closes#13204
This commit changes the way move errors are reported when some value is
captured by a PatIdent. First, we collect all of the "cannot move out
of" errors before reporting them, and those errors with the same "move
source" are reported together. If the move is caused by a PatIdent (that
binds by value), we add a note indicating where it is and suggest the
user to put `ref` if they don't want the value to move. This makes the
"cannot move out of" error in match expression nicer (though the extra
note may not feel that helpful in other places :P). For example, with
the following code snippet,
```rust
enum Foo {
Foo1(~u32, ~u32),
Foo2(~u32),
Foo3,
}
fn main() {
let f = &Foo1(~1u32, ~2u32);
match *f {
Foo1(num1, num2) => (),
Foo2(num) => (),
Foo3 => ()
}
}
```
Errors before the change:
```rust
test.rs:10:9: 10:25 error: cannot move out of dereference of `&`-pointer
test.rs:10 Foo1(num1, num2) => (),
^~~~~~~~~~~~~~~~
test.rs:10:9: 10:25 error: cannot move out of dereference of `&`-pointer
test.rs:10 Foo1(num1, num2) => (),
^~~~~~~~~~~~~~~~
test.rs:11:9: 11:18 error: cannot move out of dereference of `&`-pointer
test.rs:11 Foo2(num) => (),
^~~~~~~~~
```
After:
```rust
test.rs:9:11: 9:13 error: cannot move out of dereference of `&`-pointer
test.rs:9 match *f {
^~
test.rs:10:14: 10:18 note: attempting to move value to here (to prevent the move, use `ref num1` or `ref mut num1` to capture value by reference)
test.rs:10 Foo1(num1, num2) => (),
^~~~
test.rs:10:20: 10:24 note: and here (use `ref num2` or `ref mut num2`)
test.rs:10 Foo1(num1, num2) => (),
^~~~
test.rs:11:14: 11:17 note: and here (use `ref num` or `ref mut num`)
test.rs:11 Foo2(num) => (),
^~~
```
Close#8064
When instantiating trait default methods for certain implementation,
`typeck` correctly combined type parameters from trait bound with those
from method bound, but didn't do so for lifetime parameters. Applies
the same logic to lifetime parameters.
Closes#13204
This removes the `priv` keyword from the language and removes private enum
variants as a result. The remaining use cases of private enum variants were all
updated to be a struct with one private field that is a private enum.
RFC: 0006-remove-priv
Closes#13535
