Changes since version 0.11.15:

* fsevents: remove kFSEventStreamCreateFlagNoDefer polyfill (ci-innoq)

* libuv: add more getaddrinfo errors (Steven Kabbes)

* unix: fix accept() EMFILE error handling (Ben Noordhuis)

* linux: fix up SO_REUSEPORT back-port (Ben Noordhuis)

* fsevents: fix subfolder check (Fedor Indutny)

* fsevents: fix invalid memory access (huxingyi)

* windows/timer: fix uv_hrtime discontinuity (Bert Belder)

* unix: fix various memory leaks and undef behavior (Fedor Indutny)

* unix, windows: always update loop time (Saúl Ibarra Corretgé)

* windows: translate system errors in uv_spawn (Alexis Campailla)

* windows: uv_spawn code refactor (Alexis Campailla)

* unix, windows: detect errors in uv_ip4/6_addr (Yorkie)

* stream: introduce uv_try_write(...) (Fedor Indutny)

`uv_try_write(stream, buf, size)` acts like `uv_write()`,
but without queueing actual write until UV_POLLOUT (or IOCP completion).
This is useful for doing writes using on-stack `uv_write_t` requests.

fix #1025

Conflicts:
	.mailmap
	ChangeLog
	src/version.c

We weren't always translating system errors for in uv_spawn.

This is half a fix for node's unit test
 test-child-process-cwd.js on win.
 The rest will go into node lib.
 It also fixes test-child-process-exec-error

Changes since version 0.10.19:

* linux: fix up SO_REUSEPORT back-port (Ben Noordhuis)

* fs-event: fix invalid memory access (huxingyi)

Fixes #846

Also, don't check for a specific pkg-config version, since libuv doesn't
use pkg-config itself

When using configure, there are situations where libuv will attempt
to build uv-dtrace.h, even if it is configured with --disable-dtrace.
For example, if libuv is first configured with dtrace enabled, then
built, the .deps files will contain references to include/uv-dtrace.h.
After a make clean and configure --disable-dtrace, the build will still
attempt to create include/uv-dtrace.h and fail. make will see the
dependency reference (which survives the make clean), use the rule
(which is always added to the Makefile), and fail since DTRACE is not
defined.

This commit protects the rules to make uv-dtrace.h with the proper
conditionals to ensure the rules are not written if --disable-dtrace
is chosen.

Fix #963.

Kindly suggested by `cppcheck`.

The link to nodejs.org in README.md was broken.

Large performance counter frequency values would cause overflows, even
when 64-bit integers were used to do the multiplication with NANOSEC.
Fix this by using floating point math instead.

Fixes #850

file_info->FileName is not null terminated.

Conflicts:
	src/unix/udp.c

The eclipse backend (like the ninja backend) does not support the
--generator_output switch.

First of all, a bit of explanation of what happens there:

1. FSEvents emits absolute paths to changed files or directories
2. We cut off the first part of such paths, which is equal to handle's
   real path ('/dir/subdir/subsubdir`, without trailing slash)
3. Then, if we are running in non-recursive mode, we discard paths that
   have slashes ('/') as a non-first character in them

Commit 3d2c820a back-ports a patch from the master branch that disables
the use of SO_REUSEPORT on Linux for reasons mentioned in the commit
log.

Unfortunately, the back-port was incomplete; another setsockopt() call
site in src/unix/udp.c was overlooked.  This commit rectifies that.

Hat tip to Luca Bruno for helping troubleshoot the issue.

The gyp build on Windows produces a *.pyc file as of commit 991409e4.

Conflicts:
	build.mk
	test/test-emfile.c
	test/test-list.h

Back-port the test from commit 27795cfc from the master branch ("unix:
fix accept() EMFILE error handling").

Fix a bug that was introduced in commit 3ee4d3f1 ("unix, windows:
return error codes directly") and add a regression test for good
measure.

Hat tip to Fedor for pointing out the issue.

Fixes #1007.

The kFSEventStreamCreateFlagNoDefer flag is already
defined in CarbonCore/FSEvents.h since OS X 10.5.

Fixes #1000.

Changes since version 0.11.14:

* fsevents: report errors to user (Fedor Indutny)

* include: UV_FS_EVENT_RECURSIVE is a flag (Fedor Indutny)

* linux: use CLOCK_MONOTONIC_COARSE if available (Ben Noordhuis)

* build: make systemtap probes work with gyp build (Ben Noordhuis)

* unix: update events from pevents between polls (Fedor Indutny)

* fsevents: support japaneese characters in path (Chris Bank)

* linux: don't turn on SO_REUSEPORT socket option (Ben Noordhuis)

* queue: strengthen type checks (Ben Noordhuis)

* include: remove uv_strlcat() and uv_strlcpy() (Ben Noordhuis)

* build: fix windows smp build with gyp (Geert Jansen)

* unix: return exec errors from uv_spawn, not async (Alex Crichton)

* fsevents: use native character encoding file paths (Ben Noordhuis)

* linux: handle EPOLLHUP without EPOLLIN/EPOLLOUT (Ben Noordhuis)

* windows: use _snwprintf(), not swprintf() (Ben Noordhuis)

* fsevents: use FlagNoDefer for FSEventStreamCreate (Fedor Indutny)

* unix: fix reopened fd bug (Fedor Indutny)

* core: fix fake watcher list and count preservation (Fedor Indutny)

* unix: set close-on-exec flag on received fds (Ben Noordhuis)

* netbsd, openbsd: enable futimes() wrapper (Ben Noordhuis)

* unix: nicer error message when kqueue() fails (Ben Noordhuis)

* samples: add socks5 proxy sample application (Ben Noordhuis)

Print the error message rather than just the errno.  The fact that the
errno is 24 is only informative to people that have their operating
system's error codes memorized.

OpenBSD and NetBSD have working futimes() implementations.  This commit
makes uv_fs_futime() work on those platforms.

OpenBSD won't let you do that, the connect() system call fails with
an EINVAL error.  Connect to 127.0.0.1 instead.

Set the close-on-exec flag on file descriptors that we've received with
recvmsg() so we don't leak them when calling fork() afterwards.

On Linux, we use the MSG_CMSG_CLOEXEC flag when supported (2.6.23 and
up.)

On older Linux versions and other platforms, we walk the received file
descriptors and set the close-on-exec flag for each fd manually.  That
won't entirely avoid race conditions when other threads call fork() or
clone() but at least we're less likely to leak file descriptors now.

Create a lot of `FSEventStream`s to make `FSEventStreamStart()` fail.
Fetch the error code and verify that it is correct.

Add test/test-tcp-close-accept.c to the dependency list of the
`make check` target of the autotools build.