Beforehand, it was unclear whether rust was performing the "recommended set" of
optimizations provided by LLVM for code. This commit changes the way we run
passes to closely mirror that of clang, which in theory does it correctly. The
notable changes include:

* Passes are no longer explicitly added one by one. This would be difficult to
  keep up with as LLVM changes and we don't guaranteed always know the best
  order in which to run passes
* Passes are now managed by LLVM's PassManagerBuilder object. This is then used
  to populate the various pass managers run.
* We now run both a FunctionPassManager and a module-wide PassManager. This is
  what clang does, and I presume that we *may* see a speed boost from the
  module-wide passes just having to do less work. I have no measured this.
* The codegen pass manager has been extracted to its own separate pass manager
  to not get mixed up with the other passes
* All pass managers now include passes for target-specific data layout and
  analysis passes

Some new features include:

* You can now print all passes being run with `-Z print-llvm-passes`
* When specifying passes via `--passes`, the passes are now appended to the
  default list of passes instead of overwriting them.
* The output of `--passes list` is now generated by LLVM instead of maintaining
  a list of passes ourselves
* Loop vectorization is turned on by default as an optimization pass and can be
  disabled with `-Z no-vectorize-loops`

...bytes.

(removing previous note about eff-eye-ex'ing #5516 since it actually does not do so, it just gets us half-way.)

Fixes #8536.

Recent improvements to `&mut Trait` have made this work possible, and it solidifies that `ifmt` doesn't always have to return a string, but rather it's based around writers.

The method names in std::rt::io::extensions::WriterByteConversions are
the same as those in std::io::WriterUtils and a resolve error causes
rustc to fail after trying to find an impl of io::Writer instead of
trying to look for rt::io::Writer as well.

These new macros are all based on format! instead of fmt! and purely exist for
bootstrapping purposes. After the next snapshot, all uses of logging will be
migrated to these macros, and then after the next snapshot after that we can
drop the `2` suffix on everything

See #8599

Add OwnedStr::into_bytes

I'm not comfortable turning off rtassert! yet

vec::unshift uses this to add elements, scheduler queues use unshift,
and this was causing a lot of reallocation

These aren't used for anything at the moment and cause some TLS hits
on some perf-critical code paths. Will need to put better thought into
it in the future.

It's not a huge win but it does reduce the amount of time spent
contesting the message queue when the schedulers are under load

This makes the lock much less contended. In the test I'm running the
number of times it's contended goes from ~100000 down to ~1000.

My primary use case here is sending strings across the wire where the
intermediate storage is a byte array. The new method ends up avoiding a
copy.

Force line ending of '.in' files in jemalloc to LF

This fixes issue #8731 . On Windows, if the git option core.autocrlf is true (github recommends users to turn on this flag - see https://help.github.com/articles/dealing-with-line-endings), the headers generated from the .in files will have syntatic errors, causing compilation to fail.

This documents how to use trait bounds in a (hopefully) user-friendly way, in the containers tutorial, and also documents the task watching implementation for runtime developers in kill.rs.

r anybody

to_str, to_pretty_str, to_writer, and to_pretty_writer were at the top
level of extra::json, this moves them into an impl for Json to match
with what's been done for the rest of libextra and libstd.  (or at least for vec and str)

Also meant changing some tests.

Closes #8676.

Also added regression test for the byte vs codepoint issues that the
previous commit fixes.

This makes it relatively easy for us to split testsuite load between machines in buildbot. I've added buildbot-side support for setting up builders with -a.b suffixes (eg. linux-64-opt-vg-0.5, linux-64-opt-vg-1.5, linux-64-opt-vg-2.5, linux-64-opt-vg-3.5, linux-64-opt-vg-4.5 causes the valgrind-supervised testsuite to split 5 ways across hosts).

r?

Naturally, and sadly, turning off sanity checks in the runtime is
a noticable performance win. The particular test I'm running goes from
~1.5 s to ~1.3s.

Sanity checks are turned *on* when not optimizing, or when cfg
includes `rtdebug` or `rtassert`.

Small, but whatever 🎱