Summary:
Disable TSAN instrumentation across StaticMetaBase::onThreadExit() to avoid a shortcoming in TSAN on Linux that produces sprurious failures.

This was recently worked around by ignoring TSAN failures in ld-linux's _dl_deallocate_tsd() (D20597534), but that turns out to be insufficient, because other TSAN failures can occur on the TLS area after the dynamic linker has called free() on it.  This diff backs out the change from D20597534, and works around the problem in a more direct & complete way.

Reviewed By: yfeldblum

Differential Revision: D20671620

fbshipit-source-id: 05a74e709b620391f18467c7a5e4d14dcf9f4a1e

Summary: Fixes a copy-paste mistake from the `collectAnySemiFuture` deprecation message.

Reviewed By: Orvid

Differential Revision: D21199692

fbshipit-source-id: e61b740ea238dcbd6ca8481e1ba578a5073fe7bf

Summary: [Folly] Remove unused virtual inheritance in `AsyncSocket`.

Differential Revision: D20966977

fbshipit-source-id: 502ab00438eebed0d8ebcf32ef602af50b05df47

Summary:
Resolves these warnings:
```
folly\Singleton.cpp(102,1): warning: function declared 'noreturn' should not return [-Winvalid-noreturn]
}
^
folly\Singleton.cpp(108,1): warning: function declared 'noreturn' should not return [-Winvalid-noreturn]
}
^
folly\Singleton.cpp(126,1): warning: function declared 'noreturn' should not return [-Winvalid-noreturn]
}
^
folly\Singleton.cpp(134,1): warning: function declared 'noreturn' should not return [-Winvalid-noreturn]
}
^
folly\Singleton.cpp(145,1): warning: function declared 'noreturn' should not return [-Winvalid-noreturn]
}
```

Reviewed By: mzlee, yfeldblum

Differential Revision: D21130719

fbshipit-source-id: a0f198d69c4fde44f9a5250b6915e36b6ee68642

Summary: Android doesn't expose `malloc_usable_size` below API-level 17. Correctly reflect this in folly-config.h, and then ensure that all code paths that use `malloc_usable_size` are unreachable when compiling without it.

Reviewed By: yfeldblum

Differential Revision: D20721326

fbshipit-source-id: d60b3caa1910dc2bf8ed0844177f5cb93a9fd067

Summary: Relying on `erase()`, which must be able to move elements when erasing in front of other valid elements, causes compilation failures when trying to erase non-assignable elements at the end of the vector even though no elements actually need to be moved (and thus assigned) in that case. Here we help the compiler out and extract the destruction/resizing logic away from the moving logic.

Reviewed By: yfeldblum, ot, mkatsevVR

Differential Revision: D20718712

fbshipit-source-id: 51c4e6636bc33577494c0c10e95da5f92b58d4b6

Summary:
I think swap() is nothrow as long as value_type is_nothrow_move_constructible.  We already do this on the move c'tor.

(Note: this ignores all push blocking failures!)

Reviewed By: yfeldblum

Differential Revision: D21136202

fbshipit-source-id: a5df772c5cafa592fce95bc5ce4149dc738a984c

Summary:
`sizeof(symbolizer::FrameArray<kMaxStackTraceDepth>) = 17608`, it is quite big to put on stack. When using folly::fiber we usually only have a few KB of stack space, and will overflow when calling this function.

By putting FrameArray on heap we can avoid this issue.

Reviewed By: yfeldblum

Differential Revision: D21150720

fbshipit-source-id: 7ae1d60d47b1aa3551bf96044aadad5cc21bf914

Summary: Resolve issue here: https://fb.workplace.com/groups/560979627394613/permalink/1763417530484144/

Differential Revision: D21147181

fbshipit-source-id: 82acb7eb910deff28a722026d1f43bae855f2942

Differential Revision: D21059558

fbshipit-source-id: 588e6ec2a6fdaf70ea5c03810358c84dedf81911

Summary:
This one DCHECK is the only thing keeping Range from
being free of external deps. Lets make it not be a thing, so
people are more inclined to pull in folly because Range is an
easy gateway class to the rest of folly.

Reviewed By: yfeldblum

Differential Revision: D21121743

fbshipit-source-id: 2a54e94074135c951526c50d4bb63ebf2076a72f

Summary:
The README stated "folly requires gcc 5.1+ and a version of boost..." which
some people might thing means that folly only supports gcc.

Edit the README with more details about supported compilers and platforms.
I tried to strike a balance between informative (FB uses folly on many
platforms), conservative (the CMake build might not support all of them),
and not mentioning specific versions (likely to get outdated.)

Reviewed By: yfeldblum

Differential Revision: D21093741

fbshipit-source-id: 3a541f22b30736201361d2a10d861d61bd278e47

Summary: Unbreak -fno-exceptions support for this file.

Reviewed By: yfeldblum

Differential Revision: D21043400

fbshipit-source-id: aceea45c32d143c723ad6bae68a7403f14d7e0f2

Summary:
There are some duplicated unit test in TokenBucketTest.cpp. So this pr is intent to delete these unused unit test.
Pull Request resolved: https://github.com/facebook/folly/pull/1354

Reviewed By: yfeldblum

Differential Revision: D21061401

Pulled By: Orvid

fbshipit-source-id: ea486fc244e2be68196ef015e020b8664bb3670a

Summary: Repurpose the existing `folly::ssl::SSLSession` to be an abstraction for SSL sessions, with the internal implementation hidden. The intention is to soon move to using these abstracted sessions instead of directly using OpenSSL sessions.

Reviewed By: yfeldblum, mingtaoy

Differential Revision: D20143848

fbshipit-source-id: 01e175668c0afcf8b91d19a289368496717d3150

Summary:
Fix a bug in overwrite of null data.

Null old data should be erased so that the insert of the new data can succeed.

Reviewed By: davidtgoldblatt

Differential Revision: D20908574

fbshipit-source-id: 6873d9c76daf1289f1d5b97b001c21473c41ea3f

Summary:
Fix typo

(Note: this ignores all push blocking failures!)

Reviewed By: mokomull

Differential Revision: D21056053

fbshipit-source-id: 3bf4295f9c6be1c50cdc64f6bbdf1b34cce6bece

Summary:
- Indicate that the EDEN_VERSION_OVERRIDE environment variable affects the
  build.
- Exclude eden/config.py from the shipit  path map, as this file is
  auto-generated in open source builds.
- Include pexpect as a dependency, as this is needed for some of the
  integration tests.

Reviewed By: genevievehelsel

Differential Revision: D21000163

fbshipit-source-id: 8eec378f66487229c995f637c4787eae530c7845

Summary: We want to use observer API to correlate tasks across same Thrift request. `RequestContext::getRootId` is the thing for it.

Reviewed By: yfeldblum

Differential Revision: D20600485

fbshipit-source-id: 1a79bde0f6c5b3692b4fb3805c98f377b1b317c9

Summary: Use pre-existing cached local contexts, rather than creating our own.

Reviewed By: bimbashrestha

Differential Revision: D18890550

fbshipit-source-id: d53d0bdf35ebfcfa5dfb45da6b99a886022d5f8b

Summary:
Fix PollIoBackend compile error

(Note: this ignores all push blocking failures!)

Reviewed By: yfeldblum, kevin-vigor

Differential Revision: D20955224

fbshipit-source-id: a2c7a34d9129cb282d3652607e79dcee1ce1cc78

Summary:
This diff extracts the fbsource commit hash and the date of that
commit and maintains that in place of just the commit hash that
we were previously extracting.

This data is exported into the environment that we pass on to
builders so that it is available if they choose to use it.

In a follow on diff I'll use this to default a version number
in the watchman project.

Reviewed By: fanzeyi

Differential Revision: D20949666

fbshipit-source-id: dc12bffe5f0efc4297b15ba0140c4c67a23ab0fd

Summary:
Allow a T that derives from EnableMasterFromThis<T> to use masterLockFromThis() to get a non-owning shared_ptr to this and to use masterRefFromThis() to get a MasterPtrRef<> from this.

Adds MasterPtr::cleanup() that returns SemiFuture<Unit>. join() just does a blocking wait on the SemiFuture returned from cleanup().

Allows a T to provide a T::cleanup() method that will be composed into the MasterPtr::cleanup() work.

MasterPtr now uses SemiFuture<Unit> instead of Baton. This allows users of MasterPtr::cleanup() to compose cleanup work with other tasks.

Andrii suggested that the cleanup feature be extracted out of MasterPtr

Adds cleanup traits (that MasterPtr satisfies) and a Cleanup type that satisfies the cleanup traits and allows objects that are not heap-allocated to participate in structured concurrency by deriving from Cleanup.

Reviewed By: andriigrynenko

Differential Revision: D19584561

fbshipit-source-id: aa2d608effe613ec84b08f902a1c61561f3458bb

Summary: [Folly] Simplify a few cases of writing a loop to test each element of the loop in small-locks and spin-locks unit-tests.

Reviewed By: markisaa

Differential Revision: D20936029

fbshipit-source-id: fd8a922d8e7b0a00ccd524c9905c641fd97a32f9

Add coroutine-frame memory allocation hooks to enable identifying async coroutine allocations in traces

Summary:
Customises the coroutine-frame allocation for all async coroutine types to call through
the new `folly_coro_async_malloc` and `folly_coro_async_free` functions when heap-allocating
coroutine frames.

This should allow the identification of samples/traces that are calling into memory allocation
for allocating coroutine frames and should enable better quantification of the CPU cycles
and memory usage attributable to coroutine-frame allocations by filtering to samples that
contain these functions in their stack-traces.

Reviewed By: davidtgoldblatt

Differential Revision: D20929042

fbshipit-source-id: a94691377b92fab42736942ec8c7316c82a4205d

Summary: Folly OSS does not list range-v3 as a dependency. As a result, test/RangeTest.cpp fails to compile because it tries to include a range-v3 header. This change gates the inclusion of the header with `__has_include`, making the dependency optional.

Reviewed By: yfeldblum

Differential Revision: D20839733

fbshipit-source-id: 2d5e9c17427a0913f40f1ee80e868e27080a68f0

Summary: Improve IOBuf::clone and IOBuf::cloneOne() perf

Reviewed By: yfeldblum

Differential Revision: D20882203

fbshipit-source-id: ce37c2bf3c0fbdbe3c316856dc431a0108e4b9bf

Summary:
Fix folly::AsyncUDPSocket::gro_.value() check - GRO is enabled if the value is > 0

(Note: this ignores all push blocking failures!)

Reviewed By: kevin-vigor

Differential Revision: D20930544

fbshipit-source-id: 6672b65cafe3f9859f86ad34a6c3a2d65c2dc355

Summary: This prevents one acquire/release.

Reviewed By: lewissbaker

Differential Revision: D20924353

fbshipit-source-id: 8c5135bd2ebc03b736b1ea3dbb385e01218cffbc

Summary:
This is a follow up from D16427231; after testing the code in a similar situation on a mac (with `clang` and `libc++`) I realized there were still standards compliance issues.

In that diff, the guarantee we provide is:

```
 *   - insert() single key variants, emplace(), and emplace_hint() only provide
 *     the strong exception guarantee (unchanged when exception is thrown) when
 *     std::is_nothrow_move_constructible<value_type>::value is true.
```

The implementation is eventually backed by `std::vector`.

The guarantees provided by `std::vector` in the standard are slightly weaker (https://en.cppreference.com/w/cpp/container/vector/insert):

```
If an exception is thrown when inserting a single element at the end, and T is CopyInsertable or std::is_nothrow_move_constructible<T>::value is true, there are no effects (strong exception guarantee).
```

i.e. it only provides this safety guarantee if the element is inserted at the end, and that's exactly what the new test shows.

After discussion with nbronson we discovered that `libstdc++` and `libc++` are using different techniques to handle the corner case of `v.insert(0, v[1])` for `insert(iter, value_type const&)`. `libstdc++` makes the copy into a temporary, and then moves it in, so it does in fact provide a stronger guarantee than necessary. `libc++` checks the address of the reference it is given, and if it is in the moved range it adjusts it (!) and then uses the copy assignment operator. This means on copy assignment exception there will be a moved-from element at the desired index.

Section 26.2.1 para 11 says that `insert` or `emplace` will have no effects if an exception is thrown, unless otherwise specified. 26.3.11.5 is the `vector` modifiers, but that only mentions `insert`. That means we can fix this by merely switch our uses of `std::vector::insert` to `std::vector::emplace` inside `sorted_vector_map/set`. On `libstdc++` this should't have any effect, since both insert and emplace make a temporary copy for the non-end insertion case (in different but similar code paths). On `libc++` this will avoid the address-adjusting trick and use the temporary copy technique like in `libstdc++`.

Reviewed By: yfeldblum

Differential Revision: D20785383

fbshipit-source-id: 290c3c7c061dedf1660da9534f4b9cc338da6224

Summary: This should reduce refcounting

Reviewed By: lewissbaker

Differential Revision: D20911114

fbshipit-source-id: be941a5e0a095652c916e4704b0dd29874e0979d

Summary:
Migration from Future-returning executor-erasing collectX forms to
SemiFuture-returning forms, that are less risky in particular with coroutines.

Earlier changes added collectXSemiFuture and collectXUnsafe as a migration
path. We then migrated collectX callsites to collectXSemiFuture or
collectXUnsafe and switched the implementation of collectX to the SemiFuture
form.

This diff adds deprecation flag to collectXSemiFuture during the migration to
collectX and and removes from folly tests that are not specific to testing those
operations.

Reviewed By: yfeldblum

Differential Revision: D20841855

fbshipit-source-id: 30818711362c88c3296cfc976cd75fbd94e1c15c

Summary:
VS2019 changes the internal implementation of basic_string and vector.
This diff adds support for the new internals, and fixes a linkage
problem on VS2017.  It also adds the UninitializedMemoryHacksTest to
the getdeps-built tests.

For basic_string a single version handles both old and new.
basic_string::_Eos changed from public to private, so code that uses
the template specialization hack to invoke it when it is private will
work for the older public method as well.

vector does not have any suitable internal methods, so we need to adapt
to the new internal structure.

This diff fixes the issues addressed by
https://github.com/facebook/folly/pull/1345 , but uses an alternate
strategy to avoid reinterpret_cast.

Reviewed By: yfeldblum

Differential Revision: D20838799

fbshipit-source-id: eba7db2bd6feade1349d51be224c481a9156732b

Reviewed By: simpkins

Differential Revision: D20885314

fbshipit-source-id: 8c3a5ccbfd6630107b421b0d6953f17a93da2412

Summary:
Reduce the cost of copying by keeping a vector of cleared RequestData references instead of keeping a hash map of the state of all RequestData references.

Also, added FOLLY_ALWAYS_INLINE to functions called in copying of request contexts.

Reviewed By: yfeldblum

Differential Revision: D20863318

fbshipit-source-id: 29cc38ac6e8abde3acaffcf8fa4d1e7ccbd45e87

Summary: Use memcpy for copying maps with the same capacity instead of rehashing.

Reviewed By: yfeldblum

Differential Revision: D20863257

fbshipit-source-id: 11e52f9f2246440fb724a94c2611e44e253e462d

Summary:
When incrementally adding in MasterPtr to an existing code-base there are some shared_ptr<> (eg. returned from factories) that need to be stored in MasterPtr.

since unique_ptr converts to shared_ptr, just take and store a shared_ptr

Yes, this removes the guarantee that ~T() will run in join(). I am open to suggestions

Reviewed By: andriigrynenko

Differential Revision: D19583122

fbshipit-source-id: 3daf0436d56b9c396b5e82d397ebaf93ed1a2a2d

Summary:
when T derives from EnableSharedFromThis<T> and T is placed in a MasterPtr<T> then users of T have access to MasterPtr<T> functionality.

`EnableSharedFromThis<T>::masterLockFromThis()`
`EnableSharedFromThis<T>::masterRefFromThis()`

Reviewed By: andriigrynenko

Differential Revision: D19583025

fbshipit-source-id: ee12e9de30fd844b5be36c39d7e1ade830e27bbb

Summary:
The read of globalConfigData outside of sm.lock() races with any other writer that is modifying
globalConfigData inside of sm.lock(). This is a documentation only change.
Pull Request resolved: https://github.com/facebook/folly/pull/1348

Reviewed By: paulmckrcu

Differential Revision: D20864073

Pulled By: yfeldblum

fbshipit-source-id: 31b7ed469716acb35913df9a71585c2a59e50bea

Summary:
Add microbenchmark for ShallowCopyRequestContextScopeGuard that involves copying RequestContext-s that hold references to multiple RequestData objects.

Also, removed microbenchmark results for less common patterns, and updated results after switching to the hazard pointer-based implementation (D19145252).

Reviewed By: A5he

Differential Revision: D20862753

fbshipit-source-id: b4fc36a548761d4281c7c55ec97f731c7b4bc98d