Summary: Per Boris's report.

Reviewed By: @yfeldblum

Differential Revision: D2357299

Summary: Originally #247, which was closed before being merged, this doesn't change the behavior of the formatting function now.
Closes #297

Reviewed By: @yfeldblum

Differential Revision: D2340712

Pulled By: @sgolemon

Summary: [Folly] EnvVarSaver.

    TEST(SomeBigClassTest, ChangesEnvVars) {
      folly::EnvVarSaver saver;
      setenv("USER", "root", 1);
      BigClass().doSomethingWithUser();
    }

Reviewed By: @Gownta

Differential Revision: D2354679

Summary: This adds support to Acceptor to be able to switch between
multiple protcols when SSL is being negotiated using MSG_PEEK.

The motivation for this is to be able to try out multiple protocols.

Reviewed By: @djwatson

Differential Revision: D2327946

Summary: Fixes compiling crc32c function on older hardware. (I believe this problem was introduced by 7ec1fc0e).

Also ensures is in final library since the API suggests it should be. This should correct a number of build problems for other projects like wdt and hhvm when they are compiled using a shared folly library rather than borrowing sources at compile time.
Closes #296

Reviewed By: @yfeldblum

Differential Revision: D2340699

Pulled By: @sgolemon

Summary: MSVC spotted this very suspicious line when I was running it with /analyze, and complained about using an unitialized variable. This variable isn't used anywhere in the function, and is initializing itself with itself. I have no idea how this compiled in the first place, but apparently it does.
This removes the line in question.
Closes #295

Reviewed By: @yfeldblum, @paulbiss

Differential Revision: D2340688

Pulled By: @sgolemon

Summary: This uses `<random>` to implement it, because there is no `/dev/urandom` on Windows.
Closes #250

Reviewed By: @yfeldblum

Differential Revision: D2282887

Pulled By: @sgolemon

Summary: [Folly] Extract folly/io/async/test/RequestContextTest.cpp (from Thrift).

Reviewed By: @haijunz

Differential Revision: D2350908

Summary: When we FATAL with a "Double registration of singleton" error, make it abundantly clear why such a thing could have happended so that the developer doesn't feel lost.

Reviewed By: @chipturner

Differential Revision: D2345728

Summary: Today it's hardcoded to use the flag, which could result
in some problem. We should allow callers to choose.

Reviewed By: @djwatson

Differential Revision: D2345036

Summary: Remove recently introduced dependency on boost_thread by breaking EventBaseLocal off to its own rule

Introduced in D2203063

Reviewed By: @​mzlee

Differential Revision: D2301943

Summary: [Folly] Fix AtomicHashArray::defaultConfig SIOF.

May fix https://github.com/facebook/folly/issues/300.

Reviewed By: @paulbiss

Differential Revision: D2343162

Summary: - This mirrors the behavior in AsyncServerSocket

Reviewed By: @yfeldblum

Differential Revision: D2338665

Summary: The fix in D2283221 was specific to a bug in gcc 4.8 but the macro was failing
to filter out clang compilers.

Reviewed By: @yfeldblum

Differential Revision: D2340160

Summary: Because we don't even have `fork` to begin with under MSVC.
Closes #278

Reviewed By: @yfeldblum

Differential Revision: D2306518

Pulled By: @sgolemon

Summary: This adds MSVC support to the detection of `MaxAlign` in `Portability.h`.
Closes #256

Reviewed By: @yfeldblum

Differential Revision: D2283221

Pulled By: @sgolemon

Summary: No idea why these are the only two places it complains about, but it does.
This just explicitly scopes them.
Closes #253

Reviewed By: @yfeldblum

Differential Revision: D2282985

Pulled By: @sgolemon

Summary: [Folly] CodeMod apache::thrift::transport::TSocketAddress to folly::SocketAddress.

    yes | codemod --extensions h,tcc,cpp '^#include [<"]thrift/lib/cpp/transport/TSocketAddress\.h[>"]$' '#include <folly/SocketAddress.h>'
    yes | codemod --extensions h,tcc '\b(((apache::)?thrift::)?transport::)?TSocketAddress\b' 'folly::SocketAddress'
    yes | codemod --extensions cpp '^using apache::thrift::transport::TSocketAddress;$' 'using folly::SocketAddress;'
    yes | codemod --file-list "$(git ls-files "**/*.cpp" | xargs grep -P '^using (namespace folly|folly::SocketAddress);$' | cut -d: -f1 | paste -s -d,)" '\b(((apache::)?thrift::)?transport::)?TSocketAddress\b' 'SocketAddress'
    yes | codemod --extensions cpp '\b(((apache::)?thrift::)?transport::)?TSocketAddress\b' 'folly::SocketAddress'
    yes | codemod --extensions h,tcc,cpp -m '^\s*typedef folly::SocketAddress folly::SocketAddress;\n' ''
    yes | codemod --file-list thrift/lib/cpp/Makefile.am -m '^\s*transport/TSocketAddress\.h \\\n' ''
    rm thrift/lib/cpp/transport/TSocketAddress.h

Reviewed By: @Invalid auth token., @​luk

Differential Revision: D2325523

Summary: MSVC will never be able to load VDSO, and doesn't even have dlopen, so just return `nullptr`, and let the fallback mechanisms handle it from there.
Closes #282

Reviewed By: @yfeldblum

Differential Revision: D2307461

Pulled By: @sgolemon

Summary: Perfect forwarding only works if the function is templated on the same type, not if the type is a class template.

Discovered by @lbrandy.

Reviewed By: @yfeldblum, @lbrandy

Differential Revision: D2333005

Summary: Not aligning `lambdaBuf_` can lead to unaligned reads, which can be a problem on some platforms.

We use `lambdaBuf_` to store a type of up to `lambdaBufSize`, so aligning it to that should always be safe.

Reviewed By: @alexshap

Differential Revision: D2319347

Summary: [Folly] Emplacement in folly::padded::Adaptor, if the adapted class allows.

Reviewed By: @Gownta

Differential Revision: D2157162

Summary: Implemented the given functions by porting similar functionality from the twitter Future utility.

Reviewed By: @​hannesr

Differential Revision: D2303701

Summary: I added this to support logging with varying indentation levels, but
it could also be useful in other situations. Examples are in the
test/documentation.

Reviewed By: @tudor

Differential Revision: D2322206

Summary: [Folly] Avoid incorrect constexpr in folly/AtomicHashArray.h.

It's actually not transitively constexpr, because it uses const values that are not themselves constexpr. Depending on the compiler, it could theoretically fail to build.

Reviewed By: @Gownta

Differential Revision: D2322143

Summary: Mcrouter still enables these by default.  Tracking down a perf regression in another code path that uses fibers.

Reviewed By: @JohnRambo

Differential Revision: D2320772

Summary: Add a golden struct for checking the size of Core.
It makes this test pass on OSX
(now failing because of the different size of std::function).

Reviewed By: @​hannesr

Differential Revision: D2315429

Summary: [Folly] ScopedEventBaseThread should support an EventBaseManager context.

Reviewed By: @Gownta

Differential Revision: D2314049

Summary: [Folly] Use static const keys in AtomicHashArray::Config.

From https://github.com/facebook/folly/pull/264/, MSVC may have a hard time with certain expressions. D2284130 factored them out as `static constexpr` members, but Clang had a hard time with that in HPHP. This adds a test case that triggers the same failure to the Folly build. Not quite sure how this impacts MSVC though.

Reviewed By: @Gownta

Differential Revision: D2304346

Summary: Because `.native()` returns a wide string on MSVC, but a normal string is expected by the rest of the code.
Closes #283

Reviewed By: @yfeldblum

Differential Revision: D2307488

Pulled By: @sgolemon

Summary: MSVC has some odd accessibility rules,
so this makes a couple of things public rather than private.
Closes #284

Reviewed By: @yfeldblum

Differential Revision: D2307578

Pulled By: @sgolemon

Summary: Make RetryingTest more robust by giving it multiple chances to
meeting timing criteria.

Reviewed By: @yfeldblum

Differential Revision: D2306245

Summary: [Folly] Refactor extract FOLLY_SYNCHRONIZED_HAVE_TIMED_MUTEX.

The same complex check is done in two places. Ick. Extract it to a single place.

Reviewed By: @​mzlee

Differential Revision: D2302885

Summary: Fix incorrect using of pthread in folly::ThreadLocal.

The root of the trouble (see man pthread_key_create):
"At thread exit, if a key value has a non-NULL destructor pointer,
and the thread has a non-NULL value associated with that key,
the value of the key is set to NULL, and then the function pointed to is called
with the previously associated value as its sole argument."

At thread exit we need to recover the thread-specific threadEntry
otherwise the subsequent calls of getThreadEntry may recreate it
(what actually DOES  happen in the test ThreadLocalPtr.CreateOnThreadExit)
and the newly created threadEntry will "leak". It will live longer than the corresponding
thread violating the internal invariant of StaticMeta and also it will break
the code of the method onThreadExit. In particular this bug causes the failure
of the test ThreadLocalPtr.CreateOnThreadExit on OSX.

Reviewed By: @lbrandy

Differential Revision: D2304950

Summary: It was originally unconditionally useing `__attribute__` syntax, so this makes it conditionally use the `__declspec` syntax.
Closes #268

Reviewed By: @yfeldblum

Differential Revision: D2283909

Pulled By: @sgolemon

Summary: [Folly] Fix ASAN failure in folly/io/async/test/NotificationQueueTest.cpp.

Reviewed By: @djwatson

Differential Revision: D2299112

Summary: D2284130 broke the HHVM clang build.

Reviewed By: @mxw

Differential Revision: D2303914

Summary: Temp remove EventBaseLocal on mobile.

Reviewed By: @djwatson

Differential Revision: D2302322