Summary:
Clang 3.5 performs more rigorous static analysis than compilers we've previously used, and accordingly comes up with some additional warnings. This diff fixes some of the issues clang-3.5 warns about.

Facebook:

Test Plan:
# after pulling in changes that allow clang-3.5 to be used
fbconfig --clang --with-project-version clang:3.5 -r common/memory && fbmake runtests_opt
fbconfig --clang --with-project-version clang:3.5 -r common/strings && fbmake runtests_opt
fbconfig --clang --with-project-version clang:3.5 folly/test && fbmake runtests_opt

Tests for common/strings and folly/test pass. common/memory has one test failure, which also occurs with clang-3.4, while eliminating a couple of test failures that clang-3.4 had.

Reviewed By: meyering@fb.com

Subscribers: fbcode-common-diffs@, sdwilsh, mathieubaudet, folly-diffs@

FB internal diff: D1761101

Signature: t1:1761101:1419976252:ff60166dc8986e1239f88012b5e552a7d204aede

Summary:
emutls as far as StaticMeta needs to use it is broken on
Android as well due to unspecified pthread_key cleanup order between the
emulated __thread and our manual uses of it. Use the alternative that we
use on __APPLE__ there as well.

Test Plan: compiled and ran consuming code on Android

Reviewed By: dancol@fb.com

Subscribers: fma, shikong, kmdent, benyluo, ranjeeth, subodh, folly-diffs@

FB internal diff: D1760569

Tasks: 5265754, 5907613

Signature: t1:1760569:1419955865:afe8f35dadda85393492ac9331e9f62a74f4fdad

Summary: same as title

Test Plan: - fbconfig -r folly && fbmake runtests

Reviewed By: andrei.alexandrescu@fb.com

Subscribers: folly-diffs@

FB internal diff: D1754577

Signature: t1:1754577:1419618119:4c1a59cd19b23de9a9f6574341944e5e80530d62

Summary: Add a test to check that the accept IOPool can be the same as the IO worker thread pool

Test Plan: unittest

Reviewed By: jsedgwick@fb.com

Subscribers: doug, fugalh, njormrod, folly-diffs@

FB internal diff: D1710628

Signature: t1:1710628:1417469919:6d81a9426b61e9f1b804114f895ed541e547110f

Summary: kma asked

Test Plan: tests added

Reviewed By: andrei.bajenov@fb.com

Subscribers: alandau, bmatheny, mshneer, folly-diffs@

FB internal diff: D1756731

Signature: t1:1756731:1419442987:afdccce166ef1f1e609d8894ea587915f6fea8e7

Summary:
(almost) all the necessary code has been moved to folly.  A couple last functions in SSLUtils, moved them to SSLContext

I think this is enough so that open source mcrouter doesn't need an fbthrift dep anymore - although stuff in mcrouter/facebook still uses thrift.

Test Plan:
fbconfig -r mcrouter; fbmake runtests_dev

contbuild will probably show some other projects to fix up for SSLUtils

Reviewed By: alikhtarov@fb.com

Subscribers: doug, ps, alandau, bmatheny, alikhtarov, mshneer, folly-diffs@

FB internal diff: D1747133

Signature: t1:1747133:1418946064:1c30a60e43f017213e5514d462f267f91761abbe

Test Plan: jenkins

Reviewed By: ajitb@fb.com

Subscribers: exa, folly-diffs@

FB internal diff: D1758358

Tasks: 5880251

Signature: t1:1758358:1419640865:c9e9ea5922a602c44042c358777067e025800e6e

Summary: Add some forgotten header and cpp files to `Makefile.am` for OSS build of folly.

Test Plan: jenkins

Reviewed By: ajitb@fb.com

Subscribers: exa, folly-diffs@

FB internal diff: D1758318

Tasks: 5880251

Signature: t1:1758318:1419639812:47f527fdb2b3d2c49e7557b4176c6ab2478091ae

Summary:
Add basic timeout functionality. This adds `futures::sleep` which makes an async Future that finishes after the given duration, and `Future::get` which blocks on the result and takes an optional timeout.

Introducing the folly::wangle::futures namespace (soon to be just folly::futures) which will hold our wangle utility functions, the things that live in the Future object in Twitter's scala code. We'll probably move when* and wait-ish methods in here too, and perhaps alias makeFuture-ish methods too, though James has me mostly convinced not to deprecate them at the folly::wangle level (because they're basically Future constructors and Future lives at folly::wangle)

`Future::delayed` after Twitter's helper of the same name

Test Plan: new and old unit tests

Reviewed By: davejwatson@fb.com

Subscribers: trunkagent, fugalh, exa, folly-diffs@

FB internal diff: D1748894

Tasks: 4548494

Signature: t1:1748894:1419363496:1f4a62ec8455989c1fcce845695ace1d01c101c8

Summary: Provide a way to massacre outstanding timers.

Test Plan: new test, and using in wangle (different diff)

Reviewed By: davejwatson@fb.com

Subscribers: jsedgwick, exa, folly-diffs@

FB internal diff: D1753866

Tasks: 4548494

Signature: t1:1753866:1419361846:58ff6ca4d01b0d546495b032b454c8bf0fdb0277

Summary:
futexWaitUntil could generate an invalid timespec when presented
with a time_point from before the epoch, which resulted in an EINVAL from
the futex syscall.  Debug builds crashed on this unexpected return value,
while release builds soldiered on.  This path happened to be exercised
by the unit test.  This diff fixes the unintentional deadline overflow
in the test, adds explicit testing of overflow behavior, and makes
futexWaitUntil handle invalid time_points in a sensible manner.

Test Plan:
1. new unit tests
2. fbmake runtests_dbg

Reviewed By: mssarang@fb.com

Subscribers: strager, njormrod, folly-diffs@

FB internal diff: D1747972

Tasks: 5853949

Signature: t1:1747972:1419382193:862c193a13428d96acb33c85f962f59203661f40

Summary: This reverts commit fc83e983e2376bccc80dc4b21890e3a41dbc211f.

Test Plan: none

Reviewed By: lucian@fb.com, robot9@fb.com

Subscribers: ps, bmatheny, folly-diffs@

FB internal diff: D1756753

Tasks: 5883894, 5886688

Signature: t1:1756753:1419374764:977160379cfa49530babb38022b9d4cc80573a7d

Summary:
Sometimes when the socket is destroyed from a destructor, we wouldn't
want further callbacks on shutdown. We can unregister the readCallback_
by calling setReadCB(nullptr), but if the state is CONNECTING, we can
still get connectErr() callback. I found an ASAN trace (https://phabricator.fb.com/P18837265)
that turned out to be because of this inability to cancel this callback.
This provides a way to unregister the connect callback as well.

Test Plan: fbconfig -r folly && fbmake runtests

Reviewed By: afrind@fb.com

Subscribers: folly-diffs@

FB internal diff: D1751778

Tasks: 5852935

Signature: t1:1751778:1419363638:26967c2d4fc5819e4d65ae706d217a954dfd784f

Summary:
The 'test' command uses one equal sign for string equality, not
two.  bash apparently is okay with two, but zsh isn't, causing this test
to pass if your shell was bash and fail if it was zsh.  The test now
passes with bash, bash in sh mode, zsh, and even dash.

Test Plan:
runtests, also, these pass (zsh failed before):

SHELL=/bin/bash _build/dbg/folly/test/subprocess_test
SHELL=/bin/sh _build/dbg/folly/test/subprocess_test
SHELL=/bin/zsh _build/dbg/folly/test/subprocess_test
SHELL=/bin/dash _build/dbg/folly/test/subprocess_test

Reviewed By: njormrod@fb.com

Subscribers: lins, anca, folly-diffs@

FB internal diff: D1756090

Signature: t1:1756090:1419360674:8576e61a6e0ee102612c5eae0e1fbd79cc397bfa

Summary:
Simplify the logic slightly to better handle joining paths.
This makes DwarfTest.cpp pass.

Test Plan: runtests

Reviewed By: njormrod@fb.com

Subscribers: lins, anca, folly-diffs@

FB internal diff: D1756036

Tasks: 5871014

Signature: t1:1756036:1419359254:123cc508b8836ea0b1485abe361b1c158538aa08

Summary: advancePastEmpty() was called for all created iterators. It only makes sense for begin(). For find() it's harmful: find() shouldn't return an iterator to the next element if the key was removed. I suspect that the same race condition was possible for insert(), but I haven't tried to reproduce it.

Test Plan: Added a test for it. It fails without these changes.

Reviewed By: delong.j@fb.com

Subscribers: folly-diffs@, lovro

FB internal diff: D1751280

Tasks: 5841499

Signature: t1:1751280:1419107193:71311ff68d92d0a4dcf1941dacdfdc23c25255cc

Summary: Can't connect from ipv6-only cluster to ipv4/ipv6 service which uses different ports.

Test Plan:
```
./fastcopy_server --dir . &

# netstat -atnlp | grep LISTEN | grep fastc
tcp        0      0 0.0.0.0:65478               0.0.0.0:*                   LISTEN      9348/./fastcopy_ser
tcp        0      0 :::52793                    :::*                        LISTEN      9348/./fastcopy_ser
```

Reviewed By: philipp@fb.com, sdoroshenko@fb.com

Subscribers: ps, bmatheny, folly-diffs@

FB internal diff: D1752846

Tasks: 5868818

Signature: t1:1752846:1419043494:7cc0646882249f17258ade5ce7ae5619b13148a0

Summary: Some overloads, particularly with rvalue references, were missing. Also this fixes https://our.intern.facebook.com/intern/tasks/?t=5849579

Test Plan: unittests

Reviewed By: simpkins@fb.com

Subscribers: trunkagent, las, net-systems@, njormrod, folly-diffs@

FB internal diff: D1746319

Tasks: 5849579

Signature: t1:1746319:1418875228:febb965cf52710a5e76b7c1cce5aec601086ad90

Summary: Allow conversion from folly::StringPiece to folly::StringPiece

Test Plan: - fbconfig -r folly && fbmake runtests

Reviewed By: mpawlowski@fb.com

Subscribers: folly-diffs@

FB internal diff: D1752609

Signature: t1:1752609:1419037172:04846f658901306b53fb60c9eafbdf0f4d4d7005

Summary:
For servers using GLOG, all thread ID are get from gettid system call. So it
is very useful if we print this ID so during debugging I can find the log
of crashing thread.

I'm not sure who should review this diff. Just found you guys from previous
diffs touched this file :)

Test Plan:
See the crash log.
*** Aborted at 1419029404 (Unix time, try 'date -d @1419029404') ***
*** Signal 11 (SIGSEGV) (0x18) received by PID 18852 (pthread TID 0x7f52a1fff700) (linux TID 27908), stack trace: ***

Reviewed By: tudorb@fb.com

Subscribers: seanc, folly-diffs@

FB internal diff: D1752301

Signature: t1:1752301:1419032887:e43c2ace6914e54812149a8b8e84eaffaf42a09c

Summary:
- Allow both `baseDir` and `subDir` to be empty.  This can happen
if DW_AT_comp_dir is `.` and gets simplified to the empty string.
(The DWARF standard doesn't appear to disallow relative dirs here,
but I couldn't find anthing definitive).
- Fix `./` prefix  stripping to avoid making paths like `.///hello` absolute.
- Fix `/` suffix stripping to avoid making the root dir reference (`/`) empty.
- Do `baseDir` and `subDir` swapping after simplification.

Test Plan: Added unittests.

Reviewed By: njormrod@fb.com

Subscribers: trunkagent, sdwilsh, folly-diffs@

FB internal diff: D1747978

Signature: t1:1747978:1419014942:87f14cb31b8c19b524d7a95b14d63cf5661a8634

Test Plan: no, it's a comment

Reviewed By: zanfur@fb.com

Subscribers: folly-diffs@

FB internal diff: D1752850

Tasks: 5823969

Signature: t1:1752850:1419043661:673c68c0daad28b3fde9709e767b528b0dec8ad1

Summary: A potential "real life" scenario that maybe can hit this bug is if we erase almost all elements and then iterate over the whole map.

Test Plan: Added a test for it.

Reviewed By: mwilliams@fb.com

Subscribers: folly-diffs@, lovro

FB internal diff: D1751455

Tasks: 5866368

Signature: t1:1751455:1419016611:b44c41d348f54397844460cb38002ad0d9704972

Summary:
D1746899 enforced the constraint (previously in a comment) on the
constructor `Range(const Range&, size_t, size_t)` that `Iter` is a
`const char*`. There is however no reason for this constraint.

This patch generalizes and simplifies the constructor, and since it
has the same semantics as `subpiece`, the latter is implemented in
terms of the constructor.

Test Plan:
fbconfig -r folly && fbmake runtests_opt

Reviewed By: soren@fb.com

Subscribers: trunkagent, folly-diffs@

FB internal diff: D1747958

Signature: t1:1747958:1418930360:fcd6beeda34e64ec8a34b9491a57674ae2265596

Summary:
This temporary class will be useful to cut dependencies on thrift
while (roughly) maintaining the old interface.

Fixup various test classes that depend on the interface.  Next diff will be to cut fbthrift dep in proxygen/mcrouter

Test Plan: unit tests, jenkins

Reviewed By: hans@fb.com

Subscribers: fbcode-common-diffs@, andrewcox, alandau, bmatheny, alikhtarov, mshneer, trunkagent, doug, njormrod, folly-diffs@, davejwatson

FB internal diff: D1696094

Signature: t1:1696094:1418762312:6028815cce2845ac5f8e754132a038fabf5fd957

Summary:
We've discussed these a bunch and here they are. I stole a bunch of Hannes' magic from https://www.facebook.com/groups/715931878455430/permalink/772854686096482/ to make this easier

Test Plan: added lots of unit tests

Reviewed By: hans@fb.com

Subscribers: fugalh, folly-diffs@, hannesr

FB internal diff: D1748418

Signature: t1:1748418:1418945606:e14c7d6a31245e222bc6a0d259d0e2b9ddd5a830

Summary:
Various TARGETS could definitely use some rearranging but I'd rather wait until we're cut over to the new repo

Test Plan: wait for contbuild

Reviewed By: davejwatson@fb.com

Subscribers: ptarjan, joelm, trunkagent, hphp-diffs@, ps, fbcode-common-diffs@, fugalh, alandau, bmatheny, everstore-dev@, mwa, jgehring, fuegen, mshneer, folly-diffs@, hannesr

FB internal diff: D1740858

Tasks: 5802833

Signature: t1:1740858:1418752569:4d7d9c5b955e4d9fab4b322cf08a3d285e3db7ce

Summary: autotools is fun.  rule name has to match

Test Plan: No contbuild for this, only a jenkins hourly build: https://ci-builds.fb.com/job/folly/

Reviewed By: noamler@fb.com

Subscribers: doug, net-systems@, folly-diffs@

FB internal diff: D1747221

Signature: t1:1747221:1418929497:a29320b5e21a657fe7e0221d2bea8b332626627b

Summary:
In preparation for a global IOThreadPoolExecutor - we can't override the workerFactory if we are using a global one.

Instead, add a new getEventBases() to IOTHreadPoolExecutro and just run code on each eventBase.  For shutdown, explicitly add a stop(), instead of doing the work after loop() as exited

Cpp2Worker required minor changes to be able to use EventBaseManager's eventbase vs. their own

Test Plan: fbconfig -r thrift/lib/cpp2; fbmake runtests

Reviewed By: jsedgwick@fb.com

Subscribers: trunkagent, andrewcox, doug, fugalh, alandau, bmatheny, mshneer, folly-diffs@

FB internal diff: D1734795

Signature: t1:1734795:1418759367:b45f419e5cbc482a64a5873d9e00e3bcb071db5a

Summary:
`Range<Iter>` has an implicit constructors from strings for any
`Iter`, however such constructors are invalid (compilation fails)
if `Iter` is not `[const] char *`.

This can be an issue for overload resolution: for example

struct IsString {
bool operator()(folly::Range<int*>) { return false; }
bool operator()(folly::StringPiece) { return true; }
};

IsString()(std::string());

fails to compile because the overload is ambiguous, even if the
conversion to `ByteRange` is invalid.

This patch disables all the invalid constructors from
`[const] char*`, `std::string`, and `fbstring`.

Test Plan:
fbconfig -r folly && fbmake runtests_opt

Reviewed By: philipp@fb.com

Subscribers: folly-diffs@

FB internal diff: D1746899

Signature: t1:1746899:1418868361:50784c4993df0bd96eeb62c09c659d5e53964d9b

Summary: see attached task. concurrency is hard.

Test Plan: stress ran the hell out of it, induced failures before, can't after

Reviewed By: hans@fb.com

Subscribers: fugalh, folly-diffs@

FB internal diff: D1746879

Tasks: 5589311

Signature: t1:1746879:1418860467:0ec3f51a398eee8a836532770682a4c58758ce93

Summary: Fix the path to detail/SpinLockImpl.h in Makefile.am

Test Plan: Will confirm the Jenkins builds pass for fbthrift and proxygen.

Reviewed By: davejwatson@fb.com

Subscribers: doug, net-systems@, exa, folly-diffs@

FB internal diff: D1745795

Signature: t1:1745795:1418841636:8a861275a62b018a8181541c7d70ba0890d7129f

fix bug with elf object left with file open upon error causing Entered fatal signal handler recursively later

Summary: fix bug with elf object left with file open upon error causing Entered fatal signal handler recursively later

Test Plan: run publisher at the limit

Reviewed By: tudorb@fb.com

Subscribers: trunkagent, chip, folly-diffs@

FB internal diff: D1743052

Tasks: 5752190

Signature: t1:1743052:1418780889:ad2ee0e6999cf417c2b9fd1ef6bc82d2086774a1

Summary:
This makes several improvements to the SpinLock code:

- Add a SpinLock implementation using pthread_spinlock_t.  On non-x86_64
platforms this is preferred over the pthread_mutex_t implementation
when available.

- For the pthread implementations, throw std::system_error on error,
rather than completely aborting the program using glog's CHECK()
macros.

- Update the pthread_mutex_t implementation to call
pthread_mutex_destroy() on destruction.

- Always unit test all implementations that can be compiled on the
current build platform, even though only a single implementation will
be selected as folly::SpinLock.  This way x86_64 builds will still
unit test the pthread-based implementations.

Test Plan: Ran the unit tests.

Reviewed By: seanc@fb.com

Subscribers: trunkagent, doug, net-systems@, exa, folly-diffs@

FB internal diff: D1735770

Signature: t1:1735770:1418445953:b238aa8fb835a8d55e6e98e20c4615ae1938b98f

Summary:
Remove the PortableSpinLock header file now that all call sites have
been migrated to the new SpinLock name.

Test Plan: Built and ran unit tests for folly, thrift, and proxygen.

Reviewed By: seanc@fb.com

Subscribers: trunkagent, doug, net-systems@, exa, folly-diffs@

FB internal diff: D1734685

Signature: t1:1734685:1418335405:e834f64d7282393e942a1de3638f4cf67410eaab

Summary:
Update call sites to use the new folly::SpinLock name, instead of
folly::io::PortableSpinLock.

Test Plan: Built and ran the folly unit tests.

Reviewed By: davejwatson@fb.com

Subscribers: doug, net-systems@, exa, folly-diffs@

FB internal diff: D1734647

Signature: t1:1734647:1418334925:1bbcffccd06907224de2a102f8d4bfbe1bd62fd1

Summary:
folly::io::PortableSpinLock seems generally useful outside of the io
code.  This moves it into the base folly namespace, and renames it from
PortableSpinLock to just SpinLock.

For most users, the main difference between MicroSpinLock and SpinLock
is that SpinLock provides a constructor that does the right thing, while
MicroSpinLock has to be explicitly initialized.

Test Plan:
Added some new unit tests, and tested both the MicroSpinLock and
pthread_spinlock_t implementations.  I didn't test the Mac OS version,
although that code remains unchanged.

Reviewed By: seanc@fb.com

Subscribers: trunkagent, doug, net-systems@, exa, folly-diffs@

FB internal diff: D1734632

Signature: t1:1734632:1418394033:99f6fbe73b485a1d631a2ef7d1b39ea6f49ecb0b

Summary: This adds API which makes folly::Singleton as performant as Meyers/static-object singletons.

Test Plan:
unit test + benchmark

============================================================================
folly/experimental/test/SingletonTest.cpp       relative  time/iter  iters/s
============================================================================
NormalSingleton                                            333.32ps    3.00G
MeyersSingleton                                  100.00%   333.33ps    3.00G
FollySingletonSlow                                 0.35%    94.36ns   10.60M
FollySingletonFast                                99.43%   335.24ps    2.98G
FollySingletonFastWeak                             0.62%    53.74ns   18.61M
============================================================================

Reviewed By: alikhtarov@fb.com

Subscribers: trunkagent, folly-diffs@

FB internal diff: D1741961

Signature: t1:1741961:1418765462:d9806f1bf5275bfbe2c4c53a41b735bda93753fe