I've been playing around with code size when linking to libstd recently, and these were some findings I found that really helped code size. I started out by eliminating all I/O implementations from libnative and instead just return an unimplemented error.
In doing so, a `fn main() {}` executable was ~378K before this patch, and about 170K after the patch. These size wins are all pretty minor, but they all seemed pretty reasonable to me. With native I/O not stubbed out, this takes the size of an LTO executable from 675K to 400K.
This function is a tiny wrapper that LLVM doesn't want to inline, and it ends up
causing more bloat than necessary. The bloat is pretty small, but it's a win of
at least 7k for small executables, and I imagine that the number goes up as
there are more calls to fail!().
This removes all usage of Poly in format strings from libstd. This doesn't
prevent more future strings from coming in, but it at least removes the ones for
now.
Most of these are unnecessary because we're only looking at static strings. This
also moves to Vec in a few places instead of ~[T].
This didn't end up getting much of a code size win (update_log_settings is the
third largest function in the executables I'm looking at), but this seems like a
generally nice improvement regardless.
This lowers the #[allow(missing_doc)] directive into some of the lower modules
which are less mature. Most I/O modules now require comprehensive documentation.
This commit alters the diagnostic emission machinery to be focused around a
Writer for emitting errors. This allows it to not hard-code emission of errors
to stderr (useful for other applications).
This lowers the #[allow(missing_doc)] directive into some of the lower modules
which are less mature. Most I/O modules now require comprehensive documentation.
Previously, format!("{a}{b}", a=foo(), b=bar()) has foo() and bar() run in a
nondeterminisc order. This is clearly a non-desirable property, so this commit
uses iteration over a list instead of iteration over a hash map to provide
deterministic code generation of these format arguments.
This new SVH is used to uniquely identify all crates as a snapshot in time of
their ABI/API/publicly reachable state. This current calculation is just a hash
of the entire crate's AST. This is obviously incorrect, but it is currently the
reality for today.
This change threads through the new Svh structure which originates from crate
dependencies. The concept of crate id hash is preserved to provide efficient
matching on filenames for crate loading. The inspected hash once crate metadata
is opened has been changed to use the new Svh.
The goal of this hash is to identify when upstream crates have changed but
downstream crates have not been recompiled. This will prevent the def-id drift
problem where upstream crates were recompiled, thereby changing their metadata,
but downstream crates were not recompiled.
In the future this hash can be expanded to exclude contents of the AST like doc
comments, but limitations in the compiler prevent this change from being made at
this time.
Closes#10207
The previous code passed around a {name,version} pair everywhere, but this is
better expressed as a CrateId. This patch changes these paths to store and pass
around crate ids instead of these pairs of name/version. This also prepares the
code to change the type of hash that is stored in crates.
This is a ubiquitous type in concurrent code, and the assertions are causing
significant code bloat for simple operations such as reading the pointer
(injecting a failure point, etc).
I am testing executable sizes with no I/O implementations (everything stubbed
out to return nothing), and this took the size of a libnative executable from
328K to 207K (37% reduction in size), so I think that this is one assertion
that's well worth configuring off for now.
These are types that are in exported type signatures, but are not
exported themselves, e.g.
struct Foo { ... }
pub fn bar() -> Foo { ... }
will warn about the Foo.
Such types are not listed in documentation, and cannot be named outside
the crate in which they are declared, which is very user-unfriendly.
cc #10573.
There's a lot of these types in the compiler libraries, and a few of the
older or private stdlib ones. Some types are obviously meant to be
public, others not so much.
These are types that are in exported type signatures, but are not
exported themselves, e.g.
struct Foo { ... }
pub fn bar() -> Foo { ... }
will warn about the Foo.
Such types are not listed in documentation, and cannot be named outside
the crate in which they are declared, which is very user-unfriendly.
cc #10573
This PR allows `HashMap`s to work with custom hashers. Also with this patch are:
* a couple generic implementations of `Hash` for a variety of types.
* added `Default`, `Clone` impls to the hashers.
* added a `HashMap::with_hasher()` constructor.
Closes#12546 (Add new target 'make dist-osx' to create a .pkg installer for OS X) r=brson
Closes#12575 (rustc: Move local native libs back in link-args) r=brson
Closes#12587 (Provide a more helpful error for tests that fail due to noexec) r=brson
Closes#12589 (rustc: Remove codemap and reachable from metadata encoder) r=alexcrichton
Closes#12591 (Fix syntax::ext::deriving{,::*} docs formatting.) r=huonw
Closes#12592 (Miscellaneous Vim improvements) r=alexcrichton
Closes#12596 (path: Implement windows::make_non_verbatim()) r=alexcrichton
Closes#12598 (Improve the ctags function regular expression) r=alexcrichton
Closes#12599 (Tutorial improvement (new variant of PR #12472).) r=pnkfelix
Closes#12603 (std: Export the select! macro) r=pcwalton
Closes#12605 (Fix typo in doc of Binary trait in std::fmt) r=alexcrichton
Closes#12613 (Fix bytepos_to_file_charpos) r=brson
This PR includes:
- Create an iterator for ```List<T>``` called ```Items<T>```;
- Move all list operations inside ```List<T>``` impl;
- Removed functions that are already provided by ```Iterator``` trait;
- Refactor on ```len()``` and ```is_empty``` using ```Container``` trait;
- Bunch of minor fixes;
A replacement for using @ is intended, but still in discussion.
Closes#12344.
Make bytepos_to_charpos relative to the start of the filemap rather than its previous behaviour which was to be realtive to the start of the codemap, but ignoring multi-byte chars in earlier filemaps. Rename to bytepos_to_file_charpos. Add tests for multi-byte chars.
Mark it as #[experimental] for now. In theory this attribute will be read in the
future. I believe that the implementation is solid enough for general use,
although I would not be surprised if there were bugs in it still. I think that
it's at the point now where public usage of it will start to uncover hopefully
the last few remaining bugs.
Closes#12044
Refactoring examples on implementation of generics for linked list.
Fixing typo of 'Note's for coherancy.
Adding internal links inside the tutorial example with traits,
generics etc...
Before it would only catch lines starting `fn` or `pub fn`.
Now it can cope with:
- attributes (e.g. `#[test] fn`)
- external functions (e.g. `extern fn`, `extern "C" fn`)
- unsafe functions (e.g. `unsafe fn`)
… and any correct combination of these
(e.g. `#[test] extern "C" unsafe fn`).
Get rid of the unnecessary parenthesies that crept into some macros.
Remove a FIXME that was already fixed.
Fix a comment that wasn't rendering correctly in rustdoc.
(Expressed another way: make `[[` et al. work with the curly brace at
the end of a line as is standard Rust style, not just at the start is it
is by default in Vim, from K&R style.)
This came out of #11492, where a simpler but less effective technique
was initially proposed; some discussion of the techniques, ways and
means can be found there.
There are still a few caveats:
- Operator-pending mode behaves differently to the standard behaviour:
if inside curly braces, it should delete up to and including the
closing of the outermost curly brace (that doesn't seem to me
consistent with documented behaviour, but it's what it does). Actual
behaviour (the more logical and consistent, in my opinion): up to the
start of the next outermost curly brace.
- With folding enabled (`set fdm=syntax`), `[[` and `]]` do not behave
as they should: the default behaviour treats an entire closed fold as
one line for these purposes while this code does not (I explicitly
`set nofoldenable` in the function—the side-effects are worse with
folds enabled), leading to unexpected behaviour, the worst of which is
`[[` and/or `]]` not working in visual mode on a closed fold (visual
mode keeps it at the extreme end of the region line of the folded
region, so it's always going back to the opening line of that fold and
immediately being shoved back to the end by visual mode).
- `[[` and `]]` are operating inside comments, whereas the standard
behaviour skips comments.
- The viewport position is sometimes changed when it should not be
necessary.
Just like the bare keyword `crate` is highlighted as Error (a little
dubious, actually, given macros), `mod` is invalid after `extern`: it's
obsolete syntax.
The most significant fix is for `syntax::ext::deriving::encodable`,
where one of the blocks of code, auspiciously containing `<S>` (recall
that Markdown allows arbitrary HTML to be contained inside it), was not
formatted as a code block, with a fun but messy effect.
The rustdoc tests create and execute a file in a temporary directory. By
default on UNIX-like platforms this is in `/tmp`, which some users mount
with the `noexec` option. In those cases, the tests fail in a mysterious
way. This change adds a note that suggests what the problem might be, if
the error looks like it could have been caused by the `noexec` setup.
Closes#12558
With linkers on unix systems, libraries on the right of the command line are
used to resolve symbols in those on the left of the command line. This means
that arguments must have a right-to-left dependency chain (things on the left
depend on things on the right).
This is currently done by ordering the linker arguments as
1. Local object
2. Local native libraries
3. Upstream rust libraries
4. Upstream native libraries
This commit swaps the order of 2 and 3 so upstream rust libraries have access to
local native libraries. It has been seen that some upstream crates don't specify
the library that they link to because the name varies per platform (e.g.
lua/glfw/etc).
This commit enables building these libraries by allowing the upstream rust crate
to have access to local native libraries. I believe that the failure mode for
this scheme is when an upstream rust crate depends on a symbol in an upstream
library which is then redefined in a local library. This failure mode is
incredibly uncommon, and the failure mode also varies per platform (OSX behaves
differently), so I believe that a change like this is fine to make.
Closes#12446
This pull request partially addresses the 2 issues listed before. As part of the work required for this PR, `NonCopyable` was completely removed.
This PR also replaces the content of `type_is_pod` with `TypeContents::is_pod`, although `type_is_content` is currently not being used anywhere. I kept it for consistency with the other functions that exist in this module.
cc #10834
cc #10577
Proposed static restrictions
=====================
Taken from [this](https://github.com/mozilla/rust/pull/11979#issuecomment-35768249) comment.
I expect some code that, at a high-level, works like this:
- For each *mutable* static item, check that the **type**:
- cannot own any value whose type has a dtor
- cannot own any values whose type is an owned pointer
- For each *immutable* static item, check that the **value**:
- does not contain any ~ or box expressions (including ~[1, 2, 3] sort of things, for now)
- does not contain a struct literal or call to an enum variant / struct constructor where
- the type of the struct/enum is freeze
- the type of the struct/enum has a dtor
The compiler itself doesn't necessarily need any features of green threading
such as spawning tasks and lots of I/O, so libnative is slightly more
appropriate for rustc to use itself.
This should also help the rusti bot which is currently incompatible with libuv.
This commit splits the file implementation into file_unix and file_win32. The
two implementations have diverged to the point that they share almost 0 code at
this point, so it's easier to maintain as separate files.
The other major change accompanied with this commit is that file::open is no
longer based on libc's open function on windows, but rather windows's CreateFile
function. This fixes dealing with binary files on windows (test added in
previous commit).
This also changes the read/write functions to use ReadFile and WriteFile instead
of libc's read/write.
Closes#12406
This weeds out a bunch of warnings building stdtest on windows, and it also adds
a check! macro to the io::fs tests to help diagnose errors that are cropping up
on windows platforms as well.
cc #12516
- For each *mutable* static item, check that the **type**:
- cannot own any value whose type has a dtor
- cannot own any values whose type is an owned pointer
- For each *immutable* static item, check that the **value**:
- does not contain any ~ or box expressions
(including ~[1, 2, 3] sort of things)
- does not contain a struct literal or call to an enum
variant / struct constructor where
- the type of the struct/enum has a dtor
I've added details in the description of each comment as to what it does, which I won't redundantly repeat here in the PR. They all relate to indentation in the emacs rust-mode.
What I will note here is that this closes#8787. It addresses the last remaining case (not in the original issue description but in a comment), of indenting `match` statements. With the changes here, I believe every problem described in the issue description or comments of #8787 is addressed.
The pretty printer was treating block comments with more than two
asterisks after the first slash (e.g. `/***`) as doc comments (which are
attributes), whereas in actual fact they are just regular comments.
The travis builds have been breaking recently because LLVM 3.5 upstream is
changing. This looks like it's likely to continue, so it would be more useful
for us if we could lock ourselves to a system LLVM version that is not changing.
This commit has the support to bring our C++ glue to LLVM back in line with what
was possible back in LLVM 3.{3,4}. I don't think we're going to be able to
reasonably protect against regressions in the future, but this kind of code is a
good sign that we can continue to use the system LLVM for simple-ish things.
Codegen for ARM won't work and it won't have some of the perf improvements we
have, but using the system LLVM should work well enough for development.
This is inspired by the [question](http://www.reddit.com/r/rust/comments/1yy57k/unsolved_question_from_irc/) (re-)posted to /r/rust. The error message in this question correctly states that one should add `'static` to the trait bounds, but does not state which trait bounds. This PR makes that explicit by appending two words.
This also renames `check_durable` to `check_static` and removes the outdated comment as a cleanup.
- "Lending an immutable pointer" might be confusing. It was not discussed why borrowed pointers are immutable in the first place.
- Make it clear that the borrowed pointers are immutable even if the variable was declared with `mut`.
- Make it clear that we cannot even assign anything to the variable while its value is being borrowed.
tutorial: change "--" to an em-dash.
tutorial: change instances of "--" to em-dash.
The printing of the error message on stack overflow had two sometimes false
assumptions previously. The first is that a local task was always available (it
called Local::take) and the second is that it used `println!` instead of
manually writing.
The first assumption isn't necessarily true because while stack overflow will
likely only be detected in situations that a local task is available, it's not
guaranteed to always be in TLS. For example, during a `println!` call a task
may be blocking, causing it to be unavailable. By using Local::try_take(), we
can be resilient against these occurrences.
The second assumption could lead to odd behavior because the stdout logger can
be overwritten to run arbitrary code. Currently this should be possible, but the
utility is much diminished because a stack overflow translates to an abort()
instead of a failure.
Upstream LLVM has changed slightly such that our PassWrapper.cpp no longer
comiles (travis errors). This updates the bundled LLVM to the latest nightly
which will hopefully fix the travis errors we're seeing.
The printing of the error message on stack overflow had two sometimes false
assumptions previously. The first is that a local task was always available (it
called Local::take) and the second is that it used println! instead of
manually writing.
The first assumption isn't necessarily true because while stack overflow will
likely only be detected in situations that a local task is available, it's not
guaranteed to always be in TLS. For example, during a println! call a task
may be blocking, causing it to be unavailable. By using Local::try_take(), we
can be resilient against these occurrences.
The second assumption could lead to odd behavior because the stdout logger can
be overwritten to run arbitrary code. Currently this should be possible, but the
utility is much diminished because a stack overflow translates to an abort()
instead of a failure.
This updates a number of ignore-test tests, and removes a few completely
outdated tests due to the feature being tested no longer being supported.
This brings a number of bench/shootout tests up to date so they're compiling
again. I make no claims to the performance of these benchmarks, it's just nice
to not have bitrotted code.
Closes#2604Closes#9407
Apparently weak linkage and dlopen aren't quite working out for applications
like servo on android. There appears to be a bug or two in how android loads
dynamic libraries and for some reason libservo.so isn't being found.
As a temporary solution, add an extern "C" function to libstd which can be
called if you have a handle to the crate map manually. When crawling the crate
map, we then check this manual symbol before falling back to the old solutions.
cc #11731
Upstream LLVM has changed slightly such that our PassWrapper.cpp no longer
comiles (travis errors). This updates the bundled LLVM to the latest nightly
which will hopefully fix the travis errors we're seeing.
- "Lending an immutable pointer" might be confusing. It was not discussed why borrowed pointers are immutable in the first place.
- Make it clear that the borrowed pointers are immutable even if the variable was declared with `mut`.
- Make it clear that we cannot even assign anything to the variable while its value is being borrowed.
tutorial: change "--" to an em-dash.
tutorial: change instances of "--" to em-dash.
Apparently weak linkage and dlopen aren't quite working out for applications
like servo on android. There appears to be a bug or two in how android loads
dynamic libraries and for some reason libservo.so isn't being found.
As a temporary solution, add an extern "C" function to libstd which can be
called if you have a handle to the crate map manually. When crawling the crate
map, we then check this manual symbol before falling back to the old solutions.
cc #11731
This updates a number of ignore-test tests, and removes a few completely
outdated tests due to the feature being tested no longer being supported.
This brings a number of bench/shootout tests up to date so they're compiling
again. I make no claims to the performance of these benchmarks, it's just nice
to not have bitrotted code.
Closes#2604Closes#9407
