This removes the `attr` matcher and adds a `meta` matcher. The previous `attr`
matcher is now ambiguous because it doesn't disambiguate whether it means inner
attribute or outer attribute.

The new behavior can still be achieved by taking an argument of the form
`#[$foo:meta]` (the brackets are part of the macro pattern).

Closes #13067

Some unix platforms will send a SIGPIPE signal instead of returning EPIPE from a
syscall by default. The native runtime doesn't install a SIGPIPE handler,
causing the program to die immediately in this case. This brings the behavior in
line with libgreen by ignoring SIGPIPE and propagating EPIPE upwards to the
application in the form of an IoError.

Closes #13123

Some unix platforms will send a SIGPIPE signal instead of returning EPIPE from a
syscall by default. The native runtime doesn't install a SIGPIPE handler,
causing the program to die immediately in this case. This brings the behavior in
line with libgreen by ignoring SIGPIPE and propagating EPIPE upwards to the
application in the form of an IoError.

Closes #13123

It turns out that on linux, and possibly other platforms, child processes will
continue to accept signals until they have been *reaped*. This means that once
the child has exited, it will succeed to receive signals until waitpid() has
been invoked on it.

This is unfortunate behavior, and differs from what is seen on OSX and windows.
This commit changes the behavior of Process::signal() to be the same across
platforms, and updates the documentation of Process::kill() to note that when
signaling a foreign process it may accept signals until reaped.

Implementation-wise, this invokes waitpid() with WNOHANG before each signal to
the child to ensure that if the child has exited that we will reap it. Other
possibilities include installing a SIGCHLD signal handler, but at this time I
believe that that's too complicated.

Closes #13124

It turns out that on linux, and possibly other platforms, child processes will
continue to accept signals until they have been *reaped*. This means that once
the child has exited, it will succeed to receive signals until waitpid() has
been invoked on it.

This is unfortunate behavior, and differs from what is seen on OSX and windows.
This commit changes the behavior of Process::signal() to be the same across
platforms, and updates the documentation of Process::kill() to note that when
signaling a foreign process it may accept signals until reaped.

Implementation-wise, this invokes waitpid() with WNOHANG before each signal to
the child to ensure that if the child has exited that we will reap it. Other
possibilities include installing a SIGCHLD signal handler, but at this time I
believe that that's too complicated.

Closes #13124

So far, we've used the term POD "Plain Old Data" to refer to types that
can be safely copied. However, this term is not consistent with the
other built-in bounds that use verbs instead. This patch renames the `Pod`
kind into `Copy`.

RFC: 0003-opt-in-builtin-traits

r? @nikomatsakis 

HashMap and HashSet require keys to implement TotalEq. This makes it possible to use TypeId as a HashMap key again.

Question for reviewers: assuming we want to support `HashMap<TypeId, whatever>`, would it make sense to add a relevant test? If so, should it go to libcollections or libstd?

Summary:
So far, we've used the term POD "Plain Old Data" to refer to types that
can be safely copied. However, this term is not consistent with the
other built-in bounds that use verbs instead. This patch renames the Pod
kind into Copy.

RFC: 0003-opt-in-builtin-traits

Test Plan: make check

Reviewers: cmr

Differential Revision: http://phabricator.octayn.net/D3

All of Decoder and Encoder's methods now return a Result.

Encodable.encode() and Decodable.decode() return a Result as well.

fixes #12292

Fixes #13032 

... be stripped out"

This reverts commit 7180b5de.

We don't particularly need this, I've never seen it clutter up the docs, it just seemed nice at the time. Sadly it caused a regression for reexported methods.

Closes #13091

All of Decoder and Encoder's methods now return a Result.

Encodable.encode() and Decodable.decode() return a Result as well.

fixes #12292

The previous dependency calculation was based on an arbitrary set of asterisks
at an arbitrary depth, but using the recursive version should be much more
robust in figuring out what's dependent.

This reverts commit 7180b5de.

See #10296 for the rationale, and commits for the implementation.

libuv.a always looks out of date to the makefile, causing make to
always descend into the libuv makefile, even when there's nothing
to build.

A variety of stuff here, mostly aimed at making `make install` work correctly with `--libdir` and `--mandir`. `make install` again goes through `install.sh`.

This is all purely fallout of getting the previous commit to compile.

This commit removes implicitly adding the Send bound to ~Trait objects and
procedure types. It will now be manually required to specify that a procedure
or trait must be send-able.

Closes #10296

This bench is meant to exercise libgreen, not libnative. It recently caused the
auto-linux-32-nopt-t bot to fail as no output was produced for an hour.

Closes #6751

We really do *not* want TCO to kick in. If it does, we'll never blow the
stack, and never trigger the condition the test is checking for. To that end,
do a meaningless alloc that serves only to get a destructor to run. The
addition of nocapture/noalias seems to have let LLVM do more TCO, which
hurt this testcase.

The `_match.rs` takes advantage of passes prior to `trans` and
aggressively prunes the sub-match tree based on exact equality. When it
comes to literal or range, the strategy may lead to wrong result if
there's guard function or multiple patterns inside tuple.

Closes #12582.
Closes #13027.

Closes #6751

The `_match.rs` takes advantage of passes prior to `trans` and
aggressively prunes the sub-match tree based on exact equality. When it
comes to literal or range, the strategy may lead to wrong result if
there's guard function or multiple patterns inside tuple.

Closes #12582.
Closes #13027.

The previous syntax was `Foo:Bound<trait-parameters>`, but this is a little
ambiguous because it was being parsed as `Foo: (Bound<trait-parameters)` rather
than `Foo: (Bound) <trait-parameters>`

This commit changes the syntax to `Foo<trait-parameters>: Bound` in order to be
clear where the trait parameters are going.

Closes #9265

Instead of just looking for its presence we need to see if it actually
contains rust stuffs.

This change prepares `rustc` to accept private fields by default. These changes will have to go through a snapshot before the rest of the changes can happen.

The installed manifest is a different file, so they should have
different names. This should prevent various wierd conflicts in the future.