Summary:
While developing on a project it is often convenient to be able to invoke its
build manually, rather than always needing to re-run `getdeps.py`.  This
updates the CMakeBuilder to also emit a script that can be used to manually
run CMake outside of `getdeps.py`.

The CMakeBuilder is the only builder that this really matters for right now,
as pretty much all of the projects where we do first-party development use
CMake for their build system.

Reviewed By: pkaush

Differential Revision: D16477399

fbshipit-source-id: c8a14af158af7b32d6c799ef685b037e68b748ff

Summary:
Move code that computes project hashes to ManifestLoader.  ManifestLoader is
the only class that has all of the information necessary to compute the
project hashes correctly.  The ManifestLoader object can also cache previously
computed hashes, so that we don't have to keep computing hashes for projects
over and over again.  Previously the `BuildOptions.compute_dirs()` function
would end up re-computing hashes for all dependencies each time it was called.

Reviewed By: strager

Differential Revision: D16477401

fbshipit-source-id: ce03642114f91ce4f859f612e6b2e747cf1653be

Summary:
The ManifestLoader contains all of the state needed to create a fetcher
object, so define a helper method on this object to create a fetcher.

Reviewed By: strager

Differential Revision: D16477395

fbshipit-source-id: 6de0942fe6b8de26c18c82bf99343f5467dc006a

Summary:
Add a new ManifestLoader class to handle loading manifests and computing
dependencies.

For now the main thing this class does is maintain the `manifest_by_name`
mapping.  In subsequent diffs we should be able to move some additional logic
into this class, which will help clean up the code and eliminate some redudant
work.  In particular, we can have this class cache project hashes, which will
avoid re-computing hashes over and over again for the same projects as we do
in many cases today.  We should also be able to save and re-use some of the
project dependency ordering computation in some cases as well.

Reviewed By: strager

Differential Revision: D16477400

fbshipit-source-id: f06f62f77d8443fccaa69fe4c1306e39c395b325

Summary:
Update `BuildOpts.compute_dirs()` to use the correct project-specific manifest
context when computing project hashes.  Previously it was incorrectly using
the initial project's context when evaluating all dependencies.  This would
result in some projects potentially seeing the wrong values for variables that
may change from project to project (like `test`).

Reviewed By: pkaush

Differential Revision: D16477398

fbshipit-source-id: 6c23f5e5e19b2402000a138b3920b79044446041

Summary:
Add a ContextGenerator class so that we actually use the correct per-project
context when loading projects and computing dependencies.

Previously commands like `build` and `test` would change the contexts for each
project as they iterated through and performed the build.  However, they did
not do this when first loading the projects.  This could cause them to use
different context values when loading dependencies than when performing the
build.  For instance, this could cause issues if a project depends on
`googletest` only when testing is enabled, as the code previously did not set
the "test" parameter when evaluating dependencies.

Reviewed By: pkaush

Differential Revision: D16477396

fbshipit-source-id: c1e055f07de1cb960861d19594e3bda20a2ccd87

Summary:
Check that all variable names are valid when loading manifest files.
This ensures that getdeps.py will error out if someone makes a typo in a
variable name, rather than treating it as never equal to anything.

Reviewed By: pkaush

Differential Revision: D16477397

fbshipit-source-id: 030e0642ff4a08db8eb74a0a0223e03d53e4880f

Summary:
This allows us to use a custom deleter to track the lifetime of exceptions in
the case of smart pointers which may be persisted outside catch blocks.

Reviewed By: yfeldblum

Differential Revision: D16572937

fbshipit-source-id: ad2f0a981a2b31ef4b17a902719b684f538c29c7

Reviewed By: JunqiWang

Differential Revision: D16497024

fbshipit-source-id: aa0e42b4bfba3c70769514131daa30df84851126

Summary: bring this in line with other facebook opensource projects

Reviewed By: saifhhasan

Differential Revision: D16577367

fbshipit-source-id: d762658505f824cc180c55ea4485cecf525b8fdc

Summary: Add convenience aliases for std::pmr versions of folly::sorted_vector_set and folly::sorted_vector_map.  I also add the traditional one-argument constructor taking the allocator.

Reviewed By: yfeldblum

Differential Revision: D16523576

fbshipit-source-id: 33514db62eb8dc1b603f45204977751091e528d9

Differential Revision:
D16568423

Original commit changeset: 533353ef623a

fbshipit-source-id: fa2a719d34ee422a3e20a6e69f3c5edfb0f00257

Summary: [Folly] Fix `-Wshadow` violations found by gcc.

Reviewed By: meyering

Differential Revision: D16578896

fbshipit-source-id: 3ffe6089ad7b78d9c01d085029b55ec94d6adcff

Reviewed By: yfeldblum, VitalyKalinkin

Differential Revision: D16568423

fbshipit-source-id: 533353ef623a4bd2629a3c1782b6228f598d6c01

Summary:
Simply linking `GMOCK_LIBS` into a binary was not telling CMake that the binary depends on `gmock` being built. So, let's add that dependency explicitly.

This wasn't breaking in production because we typically build with `-j 4`, and `gmock` was getting built before the first dependent binary would attempt to link.

Also, since `rsocket` bundles its own `gmock`, it is just a waste of time to compile a system-level gmock. It's not a real dependency.

NB: The change in `fbcode_builder.py` is needed because now that `rsocket` no longer depends on anything on Github, driver programs that were unconditionally setting `projects_dir` started to fail to build `rsocket`.

Reviewed By: simpkins

Differential Revision: D16461572

fbshipit-source-id: 1e95654e96256e7ed37d42e702b5433bf2fe5328

Summary:
This diff adds unit tests that print out the wall, user, and
system time for two backend implementations of CacheLocality gathering.

Reviewed By: yfeldblum

Differential Revision: D16547404

fbshipit-source-id: 5b5467965eb21bc50ad7248c0badd814edc71d68

Summary: [Folly] Fix non-portable annotation syntax in `F14TestUtil.h`, using portable `[[noreturn]]` v.s. non-portable `__annotation__((__noreturn__))`. The non-portable syntax works with GCC and Clang, but is not recognized by MSVC.

Reviewed By: jkedgar, terrelln, Orvid

Differential Revision: D16553166

fbshipit-source-id: a652d31d31eca773009885f648f44ea3798e1460

Summary: It's desirable to know if a given server socket is accepting or if a regular UDP socket is reading.

Reviewed By: dddmello

Differential Revision: D16529348

fbshipit-source-id: 2e9a2e505f97344d21449aed7c9effc298877e41

Summary:
Cache locality information under /sys is dispersed across a
very large number of files. This is a problem for short-lived processes
due to direct overheads and lock contention in the kernel. This diff
switches to a heuristic strategy that infers the interference pattern from
/proc/cpuinfo instead of computing it exactly. This doesn't necessarily
produce exactly the correct cache hierarchy info, but it yields the
correct topological sort for machines that have only core-local and
socket-local cache locality.

Differential Revision: D16459331

fbshipit-source-id: a322c126d1a4775d015bfb81451dbc6ad6fcc0fd

Summary:
flake8 (in Python 2.7 mode) complains that `typing` is mentioned in type annotations but is not defined:

```
fbcode_builder/getdeps/buildopts.py:251:21: F821 undefined name 'typing'
    subst_mapping,  # type: typing.Mapping[str, str]
                    ^
fbcode_builder/getdeps/buildopts.py:253:5: F821 undefined name 'typing'
    # type: (...) -> typing.Optional[str]
    ^
2     F821 undefined name 'typing'
2
```

Import `typing` explicitly to silence this warning.

Because `typing` may be unavailable, import it conditionally. (Because it's only referenced in comments, failing to import `typing` should have no effect at run time.)

Reviewed By: snarkmaster

Differential Revision: D16435696

fbshipit-source-id: 78a4a7b07acc46aa998f02b54b1a6e52c1daafde

Summary:
We need to wrap coro_.resume() with a RequestContextScopeGuard so that
if the task changes the request context, we restore it after the coroutine
suspends itself.

Reviewed By: lewissbaker

Differential Revision: D16497375

fbshipit-source-id: 439a6f41b2294e56d2ec1b0013a6013524b5ae6a

Summary:
Allow builds for 64-bit ARM (MSVC) to use the same implementation of `__builtin_clzll` as ARM and x86 architectures.
Pull Request resolved: https://github.com/facebook/folly/pull/1191

Test Plan: Imported from GitHub, without a `Test Plan:` line.

Reviewed By: yfeldblum

Differential Revision: D16350573

Pulled By: Orvid

fbshipit-source-id: f25087b3a53de09767f505ee8c7d971a6d957189

Summary:
Note that we only provide the strong exception guarantee for single key insertion and emplacement when `std::is_nothrow_constructible<value_type>::value == true`. We could `static_assert()` that in `sorted_vector_set`, since it seems to hold for all contbuilds. But for `sorted_vector_map` there are quite a few contbuilds broken by that static assert. So I've chosen not to static assert in `sorted_vector_set` for consistency.

`insert()`, `emplace*()` and `erase()` take `const_iterator` in compliance with C++11. `erase()` has an overload that takes an `iterator` in C++17, but since we are backed by a vector I don't think that is necessary.

Add `emplace_hint()` variants.

Reviewed By: yfeldblum

Differential Revision: D16427231

fbshipit-source-id: 4bcf6fd01d9b1320548d12b152d1cef9291c2dd2

Summary:
While wrapping my head around the TDigest code, I made a few small
drive-by clean-ups.

Reviewed By: yfeldblum

Differential Revision: D16478284

fbshipit-source-id: aaf289a57ef633f0f1527263beffb7c0edf596d1

Summary:
The standard requires that operator== for associative containers
check key equality using operator==, which may be a refinement of the
container's key_eq.  (The requirement is expressed via equal_range and
std::is_permutation, but has the same effect.)  The straightforward way to
implement this results in duplicate key comparisons.  This diff adds a
containsEqualValue method that can avoid the second check, and uses
it for map and set operator==.

Reviewed By: yfeldblum

Differential Revision: D16455735

fbshipit-source-id: 8ccd0743f8c11bee5d91c065214cead96dd8b72c

Summary: Eden no longer depends on fboss/common/ 🌈

Reviewed By: simpkins

Differential Revision: D16062879

fbshipit-source-id: e6cf6eed408ca6e688ff2b9b532b15aa96902ac1

Summary: Add a dependency from the eden open source build to the fb303 open source build and switch EdenServiceHandler to BaseService.

Reviewed By: simpkins

Differential Revision: D15528156

fbshipit-source-id: 2ca5c31dd9fcc9bac43fd399b27f33b6f2c5ebfc

Summary: Given a cpp folly::coro::AsyncGenerator, a python async generator/iterator may be built and bridged to it. Users may use `async for ...` syntax to consume data from it.

Reviewed By: yfeldblum

Differential Revision: D16404979

fbshipit-source-id: dd2dfd52d00bd1a3e30e838359a27458744f6461

Summary:
Some people want to use getdeps with Python 2.7. This looks easy to do, so take a step toward Python 2.7 support by fixing getdeps' tests when run with Python 2.7.

For Python 3, this diff should not change behavior.

This diff should address https://github.com/facebook/bistro/issues/35.

Reviewed By: snarkmaster

Differential Revision: D16435667

fbshipit-source-id: f5c262b12995b609263341c4de26dac7f9b12b70

Summary:
This is just https://github.com/facebook/bistro/pull/37/, backported to `fbcode_builder`.

snarkmaster
Pull Request resolved: https://github.com/facebook/bistro/pull/37

Test Plan: Push and watch Travis

Reviewed By: simpkins

Differential Revision: D16453080

Pulled By: snarkmaster

fbshipit-source-id: a15eaead931f046c41e50f8e3b412ef68b172d65

Summary:
`dynamic::hash()` would copy every key-value pair in the object accumulator hash because of two bugs in the code:
1. The lambda took `auto` instead of `auto const&`
2. The hasher was `hash<pair<dynamic, dynamic>>` not `hash<pair<dynamic const, dynamic>>` meaning a conversion was needed.

These bugs together caused 2 copies for each sub-object. Since the copies are recursive, each object got copied 2*depth times.

Reviewed By: yfeldblum

Differential Revision: D16452213

fbshipit-source-id: 64a55e1640abb022c148183646e9f9720fd8482e

Summary:
This is a proposed fix for issue: https://github.com/facebook/folly/issues/1193
Using `folly::resizeWithoutInitialization` with libc++ and the address sanitizer causes false positive 'container overflow' failures.

This PR only addresses the problem for resizing `std::vector`. Presumably the version of `folly::resizeWithoutInitialization` for `std::string` has the same problem, but I have not verified this.

### Potential issues
This PR assumes that the `__sanitizer_annotate_contiguous_container` is available in libc++, and that the `std::vector` implementation has been annotated with it. The latter occurred in ~2014:
http://llvm.org/viewvc/llvm-project/libcxx/trunk/include/vector?r1=191987&r2=208319&pathrev=208319

If there are people using C++14 compilers with versions of libc++ predating the addition of the `std::vector<T>::__annotate_contiguous_container` member function then this fix will fail to compile.

I'm not aware of any sane way to detect the exact version of libc++ being used. Perhaps a suitable mitigation for those (few?) people with a C++14 compiler and pre-2015 libc++ implementation would be to provide a `FOLLY_` macro to disable the annotation.
Pull Request resolved: https://github.com/facebook/folly/pull/1194

Reviewed By: yfeldblum

Differential Revision: D16445890

Pulled By: nbronson

fbshipit-source-id: ef33bbd5cc12c71822985c5174bf8e47e6eb3757

Summary: [Folly] Extract type aliases in `LtHash` members to avoid unnatural line breaks and indentation.

Reviewed By: modocache

Differential Revision: D16453375

fbshipit-source-id: 8d68458ba3ba7bf79fef82986f4f170e3b74ddd6

Summary: [Folly] Apply `clang-format` to `folly/experimental/crypto/`.

Reviewed By: modocache

Differential Revision: D16453122

fbshipit-source-id: 5a24b20dba3ab472cd490afaaa24e7ab1c1295ec

Summary:
getdeps allowed common arguments to be specified either before or after
the subcommand name.  However, the arguments were only actually honored
if they were specified on the command line after the subcommand name.
Specifying any of these arguments before the subcommand would simply be
ignored, and the default value would be used.  This happened since the
subcommand argument's default value overwrote the value specified by the
user from the main command parser.

This fixes the behavior by using `argparse.SUPPRESS` so that argparse
won't set the value at all if it was not specified.  This allows the
value from the main parser to be used correctly if the argument was not
seen by the subcommand parser.

Reviewed By: chadaustin

Differential Revision: D16435358

fbshipit-source-id: f3893ef7a27fb018247f249f313d6438472e0140

Summary: F14NodeSet::insert() leaks memory when construction throws.

Reviewed By: yfeldblum, nbronson

Differential Revision: D16399317

fbshipit-source-id: 4c9385288dd212469dfbd10b3e990b3533dc66d3

Summary: eachToTuple: use to<fbstring>(delim) instead of delim

Reviewed By: yfeldblum

Differential Revision: D16419866

fbshipit-source-id: 7a57f41ea210c569284f6c619784adc95f03b26e

Reviewed By: yfeldblum

Differential Revision: D16426555

fbshipit-source-id: 0211c58b6919dc412b9cd7a029f0543d0060b40b

Differential Revision: D16368120

fbshipit-source-id: 0511e2f84b3bf8d1affbfbeecd05c04cb0269037

Summary: This file is included in `coro/TimedWait.h` and `coro/Collect.h`, so inclusion of those two headers is causing redefinition errors.

Reviewed By: yfeldblum

Differential Revision: D16414841

fbshipit-source-id: 103fa87a7c978d256e58eb67089b4cb984d362e4