Summary: It's possible for user code to be run in the ObserverManager thread outside of the refresh call (e.g. from Core destructor), so we can't rely on inManagerThread check to guarantee that Dependency recorder is active.

Reviewed By: yfeldblum

Differential Revision: D17805252

fbshipit-source-id: 5a858be0cf204da1342dc182e25797fd130c0458

Summary: See https://fb.workplace.com/groups/osssupport/permalink/2846876118694318/

Reviewed By: wez

Differential Revision: D17750243

fbshipit-source-id: 9e149df1f8e09203820f50d0fcac4a5cecf52e33

Summary: This is the first step towards removing `watchman/thirdparty/tap.{cpp,h}`

Reviewed By: chadaustin

Differential Revision: D17775680

fbshipit-source-id: d6ac32c3b2489e1713fb132b0bb46d848c56811f

Summary: This helps to squash out some flakiness

Reviewed By: pkaush

Differential Revision: D17804696

fbshipit-source-id: decd8e5dd37d802c62cae1168c4f4d72c0fc5c83

Summary: Same as regular groupBy

Reviewed By: ot

Differential Revision: D17843280

fbshipit-source-id: 33d74d7ed84b8c46509bf186bfa52e31d6284015

Summary: As it turns out, several of the `fizz` dependencies require it to have been built with tests enabled, so it's just easier to build them always, IIRC they only waste 1-2 minutes of time.

Reviewed By: lnicco

Differential Revision: D17837758

fbshipit-source-id: dd0c73b3aaf72831ce702dbcecd4e3ff627a4901

Reviewed By: wez

Differential Revision: D17494253

fbshipit-source-id: 45bab057e5cefd8f2808be96ba31814c5196051b

Summary:
Proxygen no longer uses `fbcode_builder` to run its tests, so whatever the purpose of D17158685, these `fbcode_builder` configs no longer affect Proxygen, and can be reverted to their original state.

Since the general design pattern for `fbcode_builder` has been to link everything as `.so`s, let's return to this (which helps fix Bistro's build).

Also, let's not waste time building & linking tests for libraries that are not the library under test. That is:
 - Before: The Bistro build also builds tests for wangle, proxygen, etc. This is a result of some accidental changes in D17158685.
 - After: We explicitly don't build test for any of the 4 dependencies here. This is OK because each project also has its own `fbcode_builder_config.py`, which **does** build tests.

This latter part should result in a build-time reduction.

Reviewed By: lnicco

Differential Revision: D17819858

fbshipit-source-id: 7cad1bed86b2f0c3934b0fc5d6fb33e6a2ee2695

Summary:
This diff introduces a data structure that keeps a set of, well any object,
but it's intended to cache compression contexts, e.g., `ZSTD_CCtx`es. These
objects can be expensive to allocate but cheap to reuse. E.g., HHVM was
spending 70% of its Zstd CPU allocating, memsetting, and freeing contexts
before it started using a structure like this.

Reviewed By: terrelln

Differential Revision: D15787721

fbshipit-source-id: bda33e9028c73c14435e18ce07cfbc8989263d98

Summary: [Folly] Namespace preprocessor names in `folly/CpuId.h`.

Reviewed By: LeeHowes

Differential Revision: D17780995

fbshipit-source-id: 29580e780a4f8639c4247b9af38f47182fc000bb

Summary:
[Folly] Fix `GroupVarint.h` includes and preprocessor namespacing.

* Always include everything rather than conditionally including things.
* Namespace as `FOLLY_HAVE_GROUP_VARINT`.

Reviewed By: LeeHowes

Differential Revision: D17780999

fbshipit-source-id: 109fa38794e312e0c5eba0f5bfab56853df59c92

Summary: s/#if _WIN32/#ifdef _WIN32/

Reviewed By: Orvid

Differential Revision: D17738798

fbshipit-source-id: cbfcf67c55b56a5e5453430d127acc3a7116814c

Summary:
[ Extend D17527441 changes from F15MapTest into F14SetTest ]
The current version of libc++ deallocates the existing capacity when calling operator=(initializer_list), even if the map is empty. This diff relaxes the expectations of F14's allocated memory tests (which fall back to the underlying STL std::unordered_map on some platforms) to accept this implementation.

Reviewed By: yfeldblum

Differential Revision: D17726849

fbshipit-source-id: 4757792a5aefe6d085b5d164ffe7b663f6078d9a

Summary:
We are seeing random segment fault originating from OpenSSL on macOS when
Mononoke fetching is enabled.

The cause is that on macOS we are actually linking against libcurl shipped with
the system instead of ours. That copy of libcurl is linked with macOS's
libcrypto instead of the one we compiles during Eden's build, and it seems that
version of libcrypto does not provide concurrency safety.

The solution is to build curl on macOS and make sure it is linked to our
OpenSSL that has the concurrency callbacks registered.

Reviewed By: wez

Differential Revision: D17657822

fbshipit-source-id: 85abdf3be10b3903a5efc6b3a91624c7258de790

Summary:
This diff teaches the generated Github Actions configuration to copy the
executable outputs from the leaf project into an artifacts dir, and then capture
the artifacts in the build.

This page has details on the artifacts mechanism used by Github Actions:
https://github.com/actions/upload-artifact

This page has the "Artifacts" link on the RHS where an example of the artifacts
can be downloaded:
https://github.com/facebook/watchman/pull/750/checks?check_run_id=244088199

Here's the actual download:
https://github.com/facebook/watchman/suites/247296465/artifacts/74686

And here's what the inside of that zip file contains for the watchman build:

```
$ find . -ls
49685444        0 drwxr-xr-x    6 wez              mygroup      192 Oct  1 18:50 .
49685487       16 -rw-r--r--    1 wez              mygroup     6148 Oct  1 18:51 ./.DS_Store
49685446        0 drwxr-xr-x    5 wez              mygroup      160 Oct  1 18:51 ./mac
49685501       16 -rw-r--r--    1 wez              mygroup     6148 Oct  1 18:51 ./mac/.DS_Store
49685453        0 drwxr-xr-x    3 wez              mygroup       96 Oct  1 18:50 ./mac/bin
49685455     5432 -rwxr-xr-x    1 wez              mygroup  2778500 Oct  2 01:02 ./mac/bin/watchman
49685454        0 drwxr-xr-x    7 wez              mygroup      224 Oct  1 18:50 ./mac/lib
49685457      296 -rwxr-xr-x    1 wez              mygroup   149096 Oct  2 01:02 ./mac/lib/libgflags.2.2.dylib
49685460     1072 -rwxr-xr-x    1 wez              mygroup   546348 Oct  2 01:02 ./mac/lib/libssl.1.1.dylib
49685459      272 -rwxr-xr-x    1 wez              mygroup   135524 Oct  2 01:02 ./mac/lib/liblz4.1.dylib
49685458      352 -rwxr-xr-x    1 wez              mygroup   176900 Oct  2 01:02 ./mac/lib/libglog.0.dylib
49685456     4888 -rwxr-xr-x    1 wez              mygroup  2501012 Oct  2 01:02 ./mac/lib/libcrypto.1.1.dylib
49685445        0 drwxr-xr-x    5 wez              mygroup      160 Oct  1 18:50 ./linux
49685486       16 -rw-r--r--    1 wez              mygroup     6148 Oct  1 18:50 ./linux/.DS_Store
49685448        0 drwxr-xr-x    3 wez              mygroup       96 Oct  1 18:50 ./linux/bin
49685450   196264 -rwxr-xr-x    1 wez              mygroup 100486648 Oct  2 01:16 ./linux/bin/watchman
49685449        0 drwxr-xr-x    4 wez              mygroup      128 Oct  1 18:50 ./linux/lib
49685451     3864 -rwxr-xr-x    1 wez              mygroup  1976888 Oct  2 01:16 ./linux/lib/libgflags.so.2.2
49685452     2208 -rwxr-xr-x    1 wez              mygroup  1127336 Oct  2 01:16 ./linux/lib/libglog.so.0
49685447        0 drwxr-xr-x    4 wez              mygroup      128 Oct  1 18:51 ./windows
49685513       16 -rw-r--r--    1 wez              mygroup     6148 Oct  1 18:51 ./windows/.DS_Store
49685461        0 drwxr-xr-x    5 wez              mygroup      160 Oct  2 01:44 ./windows/bin
49685464     7880 -rwxr-xr-x    1 wez              mygroup  4030976 Oct  2 01:44 ./windows/bin/watchman.exe
49685463      488 -rwxr-xr-x    1 wez              mygroup   249856 Oct  2 01:44 ./windows/bin/glog.dll
49685462      544 -rwxr-xr-x    1 wez              mygroup   278016 Oct  2 01:44 ./windows/bin/gflags.dll
```

The `main.yml` file was updated by running:

```
$ opensource/fbcode_builder/getdeps.py generate-github-actions --output-file watchman/.github/workflows/main.yml watchman
```

Pull Request resolved: https://github.com/facebook/watchman/pull/750

Test Plan: See description above

Reviewed By: pkaush

Differential Revision: D17705148

Pulled By: wez

fbshipit-source-id: 4f3c466180f5a1cb9fa31a9b3b3f117876f8b89d

Summary:
Production-enabled assertions have shown value in detecting
a variety of issues that should be impossible but are observed to
occur. This diff adds some checks to F14 that may fail under some types
of memory corruption, tag missearch, or hash miscomputation.

Reviewed By: yfeldblum

Differential Revision: D17739592

fbshipit-source-id: 92c9839ed5dab030350408d704a01c97573e3123

Summary: In Cython, use `from folly.cast cimport down_cast` to import the function.

Reviewed By: yfeldblum

Differential Revision: D17652305

fbshipit-source-id: 6b038c5ac6c2a8ecd8e95f33ae3a37647a2bfdff

Summary: [Folly] Fix missing qualification in `FOLLY_SAFE_DCHECK` macro.

Reviewed By: nbronson

Differential Revision: D17747688

fbshipit-source-id: 1bab5770891c900252f767252b5d7769e209bcb1

Summary:
[Folly] Export `folly::json::parse_error` v.s. having it be a hidden type.

* Move class body to `.h`.
* Leave helper function, previously the constructor, in anonymous namespace in `.cpp`.

Reviewed By: vitaut

Differential Revision: D17578773

fbshipit-source-id: 0d50fdf662bbf67e844ff0ec9ef24975cea92f62

Summary:
We were troubleshooting an issue with the eden tests on windows
where the boost dlls where not being found during gtest discovery.

When we compute the environment, we were only including INST/bin in the
PATH on windows.  On Windows, the dlls are searched for in the PATH, and
since boost installs those into its `lib` dir we were missing those.

This diff causes `lib` dirs to get added to PATH on windows in the same
manner that we would add them to `LD_LIBRARY_PATH` on linux.

Reviewed By: pkaush

Differential Revision: D17694542

fbshipit-source-id: 143a907e6d30d8c12360caa43c8d9c26ff8c88c6

Summary:
On linux we use `patchelf` to manipulate dynamic deps but it
isn't guaranteed to be installed everywhere.  We have a manifest file
that describes how to build it, but so far nothing has told getdeps
that it should build it.

This diff updates the ELF dep munging code to literally run
`getdeps.py build patchelf` and then use that patchelf binary to
manipulate the object files.

Refs: https://github.com/facebook/watchman/pull/750

Reviewed By: pkaush

Differential Revision: D17705351

Pulled By: wez

fbshipit-source-id: 358ef239edb389fbd51fa023ff553963aa80b6c7

Summary: Mobile phones have more like 16 CPUs and less like 128 CPUs.

Reviewed By: yfeldblum

Differential Revision: D16151292

fbshipit-source-id: 87caed64d9a874f71e85ee2842194405e43568a3

Summary: [Folly] Dead Code: `PriorityMPMCQueue`.

Reviewed By: LeeHowes

Differential Revision: D17718372

fbshipit-source-id: c79c376abadf7149b94d343cd6705f030fd71666

Summary: [Folly] `down_cast` for polymorphic checked-in-debug down-casts.

Reviewed By: nbronson

Differential Revision: D17650399

fbshipit-source-id: 9cdd224b3f97ba40f88013877f968ed311a3ab3c

Summary:
SSL_CTX_get_cipher_list() has different semantics when running under
OpenSSL 1.1.1.

This adjusts the tests that explicitly call SSLContext::setCiphers() to only
check expectations against < TLS 1.3 cipher names.

Reviewed By: knekritz

Differential Revision: D17539611

fbshipit-source-id: c293bd1e6b886dbc575a0677a98776d256b23fb5

Summary:
X509_get0_notBefore returns and takes const parameters.
Signed-off-by: Rosen Penev <rosenp@gmail.com>

ping reanimus
Pull Request resolved: https://github.com/facebook/folly/pull/1235

Test Plan: Build

Reviewed By: knekritz

Differential Revision: D17695324

Pulled By: reanimus

fbshipit-source-id: bebdb11fc9dd0b49354e7efcd1adc4f87a6675bc

Summary:
Fix missing null check.
It is possible for a thread to successfully CAS count from >= threshold to 0 and also pop an empty list.
Example with 2 threads:
- Batch contains Threshold - 1 objects
- Thread 1 pushes an object, finds count == Threshold, CAS-es count to 0 successfully.
- Thread 2 pushes Threshold more objects, finds count == Threshold, CAS-es count to 0 successfully.
- Thread 1 pops all 2xThreshold objects (that's ok)
- Thread 2 pops 0 objects (that's ok too but we need to do null check before dereferencing the head of the list).

Reviewed By: davidtgoldblatt

Differential Revision: D17680391

fbshipit-source-id: a1d68997668077cc0b34cc66405628ed2f95702b

Summary: According to https://github.com/jemalloc/jemalloc/issues/237 and https://fburl.com/aeia78qk, `__attribute__((nothrow))` is added to jemalloc API functions to be in compatible with `stdlib.h`.

Reviewed By: yfeldblum

Differential Revision: D17679036

fbshipit-source-id: 525a076f19719d2406c0dce92ca814f5f5011497

Summary:
Add a manifest to download pexpect-4.7.0 from PyPI, as well as its ptyprocess
dependency.

Reviewed By: wez

Differential Revision: D17669618

fbshipit-source-id: 13395ec07f503f39adb3dc5aa8d0c2d8d0f1d927

Summary:
Correctly emit dependency information when one Python package depends on
another.

Reviewed By: wez

Differential Revision: D17669620

fbshipit-source-id: f51c7851470fe50dc0c17263c94c4d858d6e0921

Summary: [Folly] Assorted nits.

Reviewed By: chadaustin

Differential Revision: D17677721

fbshipit-source-id: a5903a61059cd60e093423d087b9e1d7e9e5d443

Summary: [Folly] `aligned` and `cacheline_aligned`, aligned and implicitly padded value wrapper types.

Reviewed By: nbronson

Differential Revision: D17386115

fbshipit-source-id: 6da33a4d97f40ada59682e02af9d56863ccaff3d

Summary:
```
s/safer to safer to log/safer to log/
```

Reviewed By: yfeldblum

Differential Revision: D17629820

fbshipit-source-id: 8f9c3e412dfe52664ac1ae45e521f7df8815cef1

Summary: Allow propagation of completing executor KA in order to permit inlining

Reviewed By: LeeHowes

Differential Revision: D17123836

fbshipit-source-id: 7524868284f9482f56e5b8cd8b9e33bc4212de74

Reviewed By: dddmello

Differential Revision: D17642516

fbshipit-source-id: 5e9820df831fd907c97fab1238b38712ce2f6e64

Summary: Add a `--mode=tests` behavior for invoking `ctest`

Reviewed By: simpkins

Differential Revision: D17610033

fbshipit-source-id: 6284fe1dbfb7adf38a67820c842a5531eefd72fa

Summary:
Add a function for defining Python unit tests.  This creates the test
executable, and also emits logic to perform test discovery for ctest.

Reviewed By: simpkins

Differential Revision: D17610034

fbshipit-source-id: cdf15b0b04acc1d3e906a1e2a95eb327951176ba

Summary:
Export a property that indicates the path to the test executable.  This is
useful for callers that want to install the binary or run it from other CMake
rules.

Reviewed By: simpkins

Differential Revision: D17647146

fbshipit-source-id: b32e2694e44a07d7c234e53a7a5c8443cb144487

Summary:
This is a little strange: we already build with the system GCC on Debian 9 (Stretch),
which is GCC 6.3 - but these changes are needed on other distributions when building GCC 6.3
verbatim from upstream.

fixes facebook/folly#823

Reviewed By: jjergus

Differential Revision: D17533036

fbshipit-source-id: 32f7d1dc1fb9af3a89fe628185c0b239203e86d5

Add terminateOrKill function to folly::Subprocess to use a combination of SIGTERM/SIGKILL to terminate a process.

Summary:
Add terminateOrKill function to folly::Subprocess to use a
combination of SIGTERM/SIGKILL to terminate a process. The document
has clearly indicated this class is thread-unsafe, and it's strongly
suggested to call signal and wait in the same thread to avoid
unintended consequence, yet there is code that tried to do that in
mutliple threads and ends with potential kill(-1).

By providing a function that basically does `terminate();
waitpid_with_timeout(); kill(); waitpid()`, it should help eliminate
usage errors mentioned above.

Reviewed By: yfeldblum

Differential Revision: D17662349

fbshipit-source-id: 3a68ede5ca09270f804c88839a798ae31b615ae3