platform/upstream/nodejs.git
9 years agodeps: upgrade v8 to 4.1.0.27
Ben Noordhuis [Fri, 27 Mar 2015 22:06:05 +0000 (23:06 +0100)]
deps: upgrade v8 to 4.1.0.27

PR-URL: https://github.com/iojs/io.js/pull/1289
Reviewed-By: Johan Bergström <bugs@bergstroem.nu>
9 years agostreams: use strict on _stream_wrap
Brendan Ashworth [Thu, 26 Mar 2015 22:20:27 +0000 (15:20 -0700)]
streams: use strict on _stream_wrap

A Mostly Harmless™ change to enable 'use strict' mode in _stream_wrap, bringing it in line with /all/ the other modules.

PR-URL: https://github.com/iojs/io.js/pull/1279
Reviewed-By: Brian White (@mscdex) <mscdex@mscdex.net>
Reviewed-By: Roman Reiss (@silverwind) <me@silverwind.io>
Reviewed-By: Yosuke Furukawa (@yosuke-furukawa)
<yosuke.furukawa@gmail.com>

9 years agotimers: make Timer.close idempotent
Petka Antonov [Fri, 27 Mar 2015 18:43:25 +0000 (20:43 +0200)]
timers: make Timer.close idempotent

fixes #1287

PR-URL: https://github.com/iojs/io.js/pull/1288
Reviewed-By: Ben Noordhuis <info@bnoordhuis.nl>
9 years agodoc: fix spelling error in feature flags
Phillip Lamplugh [Fri, 27 Mar 2015 17:16:32 +0000 (10:16 -0700)]
doc: fix spelling error in feature flags

PR-URL: https://github.com/iojs/io.js/pull/1286
Reviewed-By: Ben Noordhuis <info@bnoordhuis.nl>
9 years agodebugger: don't spawn child process in remote mode
Jackson Tian [Fri, 27 Mar 2015 02:28:24 +0000 (10:28 +0800)]
debugger: don't spawn child process in remote mode

When debug in remote mode with host:port or pid, the interface
spawn child process also. If the debugger agent is running, will
get following output:

```
< Error: listen EADDRINUSE :::5858
<     at Object.exports._errnoException (util.js:734:11)
<     at exports._exceptionWithHostPort (util.js:757:20)
<     at Agent.Server._listen2 (net.js:1155:14)
<     at listen (net.js:1181:10)
<     at Agent.Server.listen (net.js:1268:5)
<     at Object.start (_debug_agent.js:21:9)
<     at startup (node.js:68:9)
<     at node.js:799:3
```

This fix won't spawn child process and no more error message was
shown.

When use `iojs debug`, the tip information just like this:

```
Usage: iojs debug script.js
```

This fix will display the advance usage also:

```
Usage: iojs debug script.js
       iojs debug <host>:<port>
       iojs debug -p <pid>
```

Fixes: https://github.com/iojs/io.js/issues/889
PR-URL: https://github.com/iojs/io.js/pull/1282
Reviewed-By: Ben Noordhuis <info@bnoordhuis.nl>
9 years agotest: reduce sequential/test-fs-watch flakiness
Roman Reiss [Thu, 26 Mar 2015 19:01:17 +0000 (20:01 +0100)]
test: reduce sequential/test-fs-watch flakiness

The fs.watch test's write events sometimes aren't produced on OS X,
possibly because of a fsevents race condition. This patch gives delays
the writing a total of 20ms, which makes the test pass consistently.

PR-URL: https://github.com/iojs/io.js/pull/1275
Reviewed-By: Ben Noordhuis <info@bnoordhuis.nl>
9 years agojs2c: fix module id generation on windows
Ben Noordhuis [Fri, 27 Mar 2015 00:07:07 +0000 (01:07 +0100)]
js2c: fix module id generation on windows

Fix a regression that was introduced in commit 2db758c ("iojs: introduce
internal modules") where the computed id for "config.gypi" on Windows
was not "config" but an empty string.

With an empty string, the build succeeds but the binary is unusable:
startup.processConfig() in src/node.js chokes on the missing .config
property.

PR-URL: https://github.com/iojs/io.js/pull/1281
Reviewed-By: Vladimir Kurchatkin <vladimir.kurchatkin@gmail.com>
9 years agosrc: don't lazy-load timer globals
Ben Noordhuis [Thu, 26 Mar 2015 22:56:26 +0000 (23:56 +0100)]
src: don't lazy-load timer globals

Don't lazy-load setInterval(), setTimeout(), etc.  Most applications are
going to need them and routing every call through NativeModule.require()
and Function#apply() is not exactly efficient.

PR-URL: https://github.com/iojs/io.js/pull/1280
Reviewed-By: Jeremiah Senkpiel <fishrock123@rocketmail.com>
9 years agoRevert "doc: clarify real name requirement"
Jeremiah Senkpiel [Thu, 26 Mar 2015 19:03:11 +0000 (15:03 -0400)]
Revert "doc: clarify real name requirement"

This reverts commit 4e9bf93e9c645f33446f7d47005868ffcb3d281e.

PR-URL: https://github.com/iojs/io.js/pull/1276
Reviewed-By: Roman Reiss <me@silverwind.io>
Reviewed-By: Rod Vagg <rod@vagg.org>
9 years agowin,node-gyp: make delay-load hook optional
Bert Belder [Thu, 26 Mar 2015 03:53:35 +0000 (20:53 -0700)]
win,node-gyp: make delay-load hook optional

The delay-load hook that was landed in 3d46fef to make compiled addons
work on Windows regardless of the iojs.exe/node.exe filename causes
issues with a small amount of compiled addons.

Therefore this patch makes it an opt-in feature. An addon may set the
'win_delay_load_hook' option to 'true' in its binding.gyp to enable this
feature.

Example:

```
{
  'targets': [
    {
      'target_name': 'ernie',
      'win_delay_load_hook': 'true',
      ...
```

Refs: https://github.com/iojs/io.js/pull/1251
PR-URL: https://github.com/iojs/io.js/pull/1266
Reviewed-By: Ben Noordhuis <info@bnoordhuis.nl>
9 years agotimers: cleanup interval handling
Jeremiah Senkpiel [Thu, 26 Mar 2015 15:52:36 +0000 (11:52 -0400)]
timers: cleanup interval handling

Uses `null` as the false-y value for `_repeat` as like other properties.
Removes un-reachable statement in setInterval’s `wrapper()`.

PR-URL: https://github.com/iojs/io.js/pull/1272
Reviewed-by: Trevor Norris <trev.norris@gmail.com>
Reviewed-By: Ben Noordhuis <info@bnoordhuis.nl>
9 years agotimers: assure setTimeout callback only runs once
Roman Reiss [Sat, 21 Mar 2015 16:39:35 +0000 (17:39 +0100)]
timers: assure setTimeout callback only runs once

Calling this.unref() during the callback of SetTimeout caused the
callback to get executed twice because unref() didn't expect to be
called during that time and did not stop the ref()ed Timeout but
did start a new timer. This commit prevents the new timer creation
when the callback was already called.

Fixes: https://github.com/iojs/io.js/issues/1191
Reviewed-by: Trevor Norris <trev.norris@gmail.com>
Reviewed-By: Jeremiah Senkpiel <fishrock123@rocketmail.com>
PR-URL: https://github.com/iojs/io.js/pull/1231

9 years agotls_wrap: fix this incredibly stupid leak
Fedor Indutny [Wed, 25 Mar 2015 03:21:32 +0000 (20:21 -0700)]
tls_wrap: fix this incredibly stupid leak

Always call `Done` on the WriteWrap, and ensure that `EncOut` will
consume all data in clear_in_ and invoke queued callbacks.

Fix: https://github.com/iojs/io.js/issues/1075
PR-URL: https://github.com/iojs/io.js/pull/1244
Reviewed-By: Ben Noordhuis <info@bnoordhuis.nl>
9 years agotls_wrap: fix BIO leak on SSL error
Fedor Indutny [Tue, 24 Mar 2015 03:07:03 +0000 (20:07 -0700)]
tls_wrap: fix BIO leak on SSL error

Fix: https://github.com/iojs/io.js/issues/1075
PR-URL: https://github.com/iojs/io.js/pull/1244
Reviewed-By: Ben Noordhuis <info@bnoordhuis.nl>
9 years agoiojs: introduce internal modules
Vladimir Kurchatkin [Sat, 14 Feb 2015 19:53:34 +0000 (22:53 +0300)]
iojs: introduce internal modules

Internal modules can be used to share private code between
public modules without risk to expose private APIs to the
user.

PR-URL: https://github.com/iojs/io.js/pull/848
Reviewed-By: Trevor Norris <trev.norris@gmail.com>
Reviewed-By: Ben Noordhuis <info@bnoordhuis.nl>
9 years agodoc: fix format docs discrepancy
Brendan Ashworth [Wed, 25 Mar 2015 05:44:08 +0000 (22:44 -0700)]
doc: fix format docs discrepancy

Only objects and symbols use `util.inspect`, others are simply
concatenated.

Fixes: https://github.com/iojs/io.js/issues/935
PR-URL: https://github.com/iojs/io.js/pull/1255
Reviewed-By: Colin Ihrig <cjihrig@gmail.com>
9 years agostring_decoder: optimize write()
Brian White [Thu, 19 Mar 2015 21:31:34 +0000 (17:31 -0400)]
string_decoder: optimize write()

By limiting property getting/setting to only where they are
absolutely necessary, we can achieve greater performance
especially with small utf8 inputs and any size base64 inputs.

PR-URL: https://github.com/iojs/io.js/pull/1209
Reviewed-By: Rod Vagg <rod@vagg.org>
Reviewed-By: Nicu Micleușanu <micnic90@gmail.com>
Reviewed-By: Chris Dickinson <christopher.s.dickinson@gmail.com>
9 years agowin,node-gyp: allow node.exe/iojs.exe to be renamed
Bert Belder [Wed, 25 Mar 2015 04:00:12 +0000 (21:00 -0700)]
win,node-gyp: allow node.exe/iojs.exe to be renamed

On Windows, when node or io.js attempts to dynamically load a compiled
addon, the compiled addon tries to load node.exe or iojs.exe again -
depending on which import library the module used when it was linked.
This makes it impossible to rename node.exe or iojs.exe, because when
that happens the module can't find its dependencies.

With this patch, a delay-load hook is added to all modules that are
compiled with node-gyp. The delay-load hook ensures that whenever a
module tries to load imports from node.exe/iojs.exe, it'll just refer
back to the process image, thus making it possible to rename the
iojs/node binary.

Bug: https://github.com/iojs/io.js/issues/751
Bug: https://github.com/iojs/io.js/issues/965
Upstream PR: https://github.com/TooTallNate/node-gyp/pull/599

PR-URL: https://github.com/iojs/io.js/pull/1251
Reviewed-By: Rod Vagg <rod@vagg.org>
9 years agofs: fix corruption in writeFile and writeFileSync
Olov Lassus [Thu, 5 Mar 2015 00:03:34 +0000 (01:03 +0100)]
fs: fix corruption in writeFile and writeFileSync

1. writeFileSync bumps position incorrectly, causing it to drift in
iteration three and onwards.

2. Append mode files will get corrupted in the middle if writeFile or
writeFileSync iterates multiple times, unless running on Linux. position
starts out as null so first write is OK, but then position will refer to
a location inside an existing file, corrupting that data. Linux ignores
position for append mode files so it doesn't happen there.

This commit fixes these two related issues by bumping position correctly
and by always using null as the position argument to write/writeSync for
append mode files.

PR-URL: https://github.com/iojs/io.js/pull/1063
Reviewed-By: Bert Belder <bertbelder@gmail.com>
9 years agodoc: clarify real name requirement
Roman Reiss [Tue, 24 Mar 2015 21:53:31 +0000 (22:53 +0100)]
doc: clarify real name requirement

CONTRIBUTING.md didn't make our rule for requiring real names from
contributors clear enough. This commit shall clarify that part.

Reviewed-By: Jeremiah Senkpiel <fishrock123@rocketmail.com>
PR-URL: https://github.com/iojs/io.js/pull/1250

9 years agodoc: document repl on-demand module loading
Roman Reiss [Tue, 24 Mar 2015 21:06:27 +0000 (22:06 +0100)]
doc: document repl on-demand module loading

Fixes: https://github.com/iojs/io.js/issues/992
Reviewed-By: Jeremiah Senkpiel <fishrock123@rocketmail.com>
Reviewed-By: Brendan Ashworth <brendan.ashworth@me.com>
PR-URL: https://github.com/iojs/io.js/pull/1249

9 years agolib: add missing `new` for errors lib/*.js
Mayhem [Tue, 24 Mar 2015 14:15:57 +0000 (15:15 +0100)]
lib: add missing `new` for errors lib/*.js

Not including `new` adds a useless frame and removes a potentially
useful frame.

PR-URL: https://github.com/iojs/io.js/pull/1246
Reviewed-By: Petka Antonov <petka_antonov@hotmail.com>
Reviewed-By: Roman Reiss <me@silverwind.io>
Reviewed-By: Brendan Ashworth <brendan.ashworth@me.com>
9 years agoassert: simplify logic of testing buffer equality
Alex Yursha [Tue, 17 Mar 2015 14:12:51 +0000 (17:12 +0300)]
assert: simplify logic of testing buffer equality

Delegate buffer equality check to `buffer.equals()`

PR-URL: https://github.com/iojs/io.js/pull/1171
Reviewed-By: Brendan Ashworth <brendan.ashworth@me.com>
Reviewed-By: Christian Vaagland Tellnes <christian@tellnes.com>
9 years agodeps: make node-gyp work with io.js
cjihrig [Fri, 27 Feb 2015 18:40:58 +0000 (13:40 -0500)]
deps: make node-gyp work with io.js

Every npm version bump requires a few patches to be floated on
node-gyp for io.js compatibility. These patches are found in
03d199276e21c1fa08d8df14eeb654c90cc5aa20,
5de334c23096492014a097ff487f07ad8eaee6d2, and
da730c76e98fb9fd18dac445dafbbec74d79f802. This commit squashes
them into a single commit.

PR-URL: https://github.com/iojs/io.js/pull/990
Reviewed-By: Ben Noordhuis <info@bnoordhuis.nl>
9 years agodeps: upgrade npm to 2.7.3
Forrest L Norvell [Fri, 20 Mar 2015 08:40:50 +0000 (01:40 -0700)]
deps: upgrade npm to 2.7.3

PR-URL: https://github.com/iojs/io.js/pull/1219
Reviewed-By: Jeremiah Senkpiel <fishrock123@rocketmail.com>
Reviewed-By: Rod Vagg <rod@vagg.org>
9 years agosrc: remove unnecessary environment lookups
Ben Noordhuis [Sun, 22 Mar 2015 23:38:42 +0000 (00:38 +0100)]
src: remove unnecessary environment lookups

Remove some unnecessary environment lookups and delete a few superfluous
HandleScope variables.

PR-URL: https://github.com/iojs/io.js/pull/1238
Reviewed-By: Fedor Indutny <fedor@indutny.com>
9 years agosrc: make accessors immune to context confusion
Ben Noordhuis [Sun, 22 Mar 2015 23:26:59 +0000 (00:26 +0100)]
src: make accessors immune to context confusion

It's possible for an accessor or named interceptor to get called with
a different execution context than the one it lives in, see the test
case for an example using the debug API.

This commit fortifies against that by passing the environment as a
data property instead of looking it up through the current context.

Fixes: https://github.com/iojs/io.js/issues/1190 (again)
PR-URL: https://github.com/iojs/io.js/pull/1238
Reviewed-By: Fedor Indutny <fedor@indutny.com>
9 years agoWorking on v1.6.3
Rod Vagg [Mon, 23 Mar 2015 05:28:45 +0000 (16:28 +1100)]
Working on v1.6.3

9 years ago2015-03-23 io.js v1.6.2 Release v1.6.2
Rod Vagg [Mon, 23 Mar 2015 05:26:28 +0000 (16:26 +1100)]
2015-03-23 io.js v1.6.2 Release

Notable Changes:

* Windows: The ongoing work in improving the state of Windows support
  has resulted in full test suite passes once again. As noted in the
  release notes for v1.4.2, CI system and configuration problems
  prevented it from properly reporting problems with the Windows
  tests, the problems with the CI and the codebase appear to have been
  fully resolved.
* FreeBSD: A kernel bug impacting io.js/Node.js was discovered and a
  patch has been introduced to prevent it causing problems for io.js
  (Fedor Indutny) #1218.
* module: you can now require('.') instead of having to require('./'),
  this is considered a bugfix (Michaël Zasso) #1185.
* v8: updated to 4.1.0.25 including patches for --max_old_space_size
  values above 4096 and Solaris support, both of which are already
  included in io.js.

9 years agoutil: Check input to util.inherits
Connor Peet [Mon, 23 Mar 2015 00:45:16 +0000 (20:45 -0400)]
util: Check input to util.inherits

PR-URL: https://github.com/iojs/io.js/pull/1240
Reviewed-By: Brendan Ashworth <brendan.ashworth@me.com>
Reviewed-By: Petka Antonov <petka_antonov@hotmail.com>
9 years agodeps: upgrade v8 to 4.1.0.25
Johan Bergström [Sat, 21 Mar 2015 01:31:48 +0000 (12:31 +1100)]
deps: upgrade v8 to 4.1.0.25

PR-URL: https://github.com/iojs/io.js/pull/1224
Reviewed-By: Fedor Indutny <fedor@indutny.com>
Reviewed-By: Rod Vagg <rod@vagg.org>
Reviewed-By: Ben Noordhuis <info@bnoordhuis.nl>
9 years agotest: move sequential/test-signal-unregister
Ben Noordhuis [Fri, 20 Mar 2015 22:57:11 +0000 (23:57 +0100)]
test: move sequential/test-signal-unregister

Move sequential/test-signal-unregister to test/parallel, it doesn't
need to run in sequential mode.

PR-URL: https://github.com/iojs/io.js/pull/1227
Reviewed-By: Johan Bergström <bugs@bergstroem.nu>
9 years agotest: fix timing issue in signal test
Ben Noordhuis [Fri, 20 Mar 2015 22:55:37 +0000 (23:55 +0100)]
test: fix timing issue in signal test

Change sequential/test-signal-unregister so it doesn't use fixed
timeouts for sending the signal and expecting the child to quit.

Fixes: https://github.com/iojs/io.js/issues/1223
PR-URL: https://github.com/iojs/io.js/pull/1227
Reviewed-By: Johan Bergström <bugs@bergstroem.nu>
9 years agovm: fix crash on fatal error in debug context
Ben Noordhuis [Sat, 21 Mar 2015 12:41:16 +0000 (13:41 +0100)]
vm: fix crash on fatal error in debug context

Ensure that the debug context has an Environment assigned in case
a fatal error is raised.

The fatal exception handler in node.cc is not equipped to deal with
contexts that don't have one and can't easily be taught that due to
a deficiency in the V8 API: there is no way for the embedder to tell
if the data index is in use.

Fixes: https://github.com/iojs/io.js/issues/1190
PR-URL: https://github.com/iojs/io.js/pull/1229
Reviewed-By: Fedor Indutny <fedor@indutny.com>
9 years agodoc: update AUTHORS list
Rod Vagg [Sun, 22 Mar 2015 06:34:07 +0000 (17:34 +1100)]
doc: update AUTHORS list

Update AUTHORS file using tools/update-authors.sh

PR-URL: https://github.com/iojs/io.js/pull/1234
Reviewed-By: Roman Reiss <me@silverwind.io>
Reviewed-By: Ben Noordhuis <info@bnoordhuis.nl>
9 years agodoc: fix typo in CHANGELOG
Mathieu Darse [Sat, 21 Mar 2015 13:37:47 +0000 (14:37 +0100)]
doc: fix typo in CHANGELOG

PR-URL: https://github.com/iojs/io.js/pull/1230
Reviewed-By: Rod Vagg <rod@vagg.org>
9 years agonode: ensure that streams2 won't `.end()` stdin
Fedor Indutny [Sat, 21 Mar 2015 23:19:04 +0000 (16:19 -0700)]
node: ensure that streams2 won't `.end()` stdin

Stdin is purely read-only stream. Although, `net.Socket` might be used
to create it if stdin is in fact a Pipe or TCP socket, the
`stream.Duplex` should not try to call `.end()` on it.

Fix: https://github.com/iojs/io.js/issues/1068
PR-URL: https://github.com/iojs/io.js/pull/1233
Reviewed-By: Chris Dickinson <christopher.s.dickinson@gmail.com>
9 years agotest: fix crypto-binary-default bad crypto check
Brendan Ashworth [Fri, 13 Mar 2015 08:42:21 +0000 (01:42 -0700)]
test: fix crypto-binary-default bad crypto check

This commit fixes a small bug introduced in 671fbd5
that caused the test to not be run. crypto was properly
checked, but since tls was not imported, a TypeError
would be thrown in the `try {} catch {}` block and
falsely reported as no crypto.

This is now fixed.

PR-URL: https://github.com/iojs/io.js/pull/1141
Reviewed-By: Johan Bergström <bugs@bergstroem.nu>
9 years agodoc: call js function in null context
Ben Noordhuis [Wed, 11 Mar 2015 13:19:21 +0000 (14:19 +0100)]
doc: call js function in null context

It's good practice now to call JS functions that don't execute in a
specific scope with v8::Null() as the receiver.  Update the addons
documentation.

PR-URL: https://github.com/iojs/io.js/pull/1125
Reviewed-By: Rod Vagg <rod@vagg.org>
Reviewed-By: Trevor Norris <trev.norris@gmail.com>
9 years agodoc: don't use `using namespace v8`
Ben Noordhuis [Wed, 11 Mar 2015 13:14:18 +0000 (14:14 +0100)]
doc: don't use `using namespace v8`

Wholesale importing an entire namespace with `using namespace` is a bad
practice.  Remove it from the addons documentation and replace it with
proper `using` directives.  Wrap code in a namespace while we are here.

PR-URL: https://github.com/iojs/io.js/pull/1125
Reviewed-By: Rod Vagg <rod@vagg.org>
Reviewed-By: Trevor Norris <trev.norris@gmail.com>
9 years agodoc: replace v8::Handle<T> with v8::Local<T>
Ben Noordhuis [Wed, 11 Mar 2015 13:03:02 +0000 (14:03 +0100)]
doc: replace v8::Handle<T> with v8::Local<T>

v8::Handle is on its way out, to be replaced with v8::Local.  Update the
addons documentation accordingly.

PR-URL: https://github.com/iojs/io.js/pull/1125
Reviewed-By: Rod Vagg <rod@vagg.org>
Reviewed-By: Trevor Norris <trev.norris@gmail.com>
9 years agodoc: remove unnecessary v8::HandleScopes
Ben Noordhuis [Wed, 11 Mar 2015 12:56:37 +0000 (13:56 +0100)]
doc: remove unnecessary v8::HandleScopes

Remove unnecessary v8::HandleScope uses from the addons documentation.

C++ API callbacks run in an implicit v8::HandleScope, there is no need
to declare one in the callback function.

PR-URL: https://github.com/iojs/io.js/pull/1125
Reviewed-By: Rod Vagg <rod@vagg.org>
Reviewed-By: Trevor Norris <trev.norris@gmail.com>
9 years agodoc: remove uses of v8::Isolate::GetCurrent()
Ben Noordhuis [Wed, 11 Mar 2015 12:41:32 +0000 (13:41 +0100)]
doc: remove uses of v8::Isolate::GetCurrent()

v8::Isolate::GetCurrent() is slated for deprecation.  Replace its uses
in the addons documentation with v8::Object::GetIsolate(), etc.

PR-URL: https://github.com/iojs/io.js/pull/1125
Reviewed-By: Rod Vagg <rod@vagg.org>
Reviewed-By: Trevor Norris <trev.norris@gmail.com>
9 years agotest: add setTimeout/setInterval multi-arg tests
Ben Noordhuis [Fri, 20 Mar 2015 16:35:52 +0000 (17:35 +0100)]
test: add setTimeout/setInterval multi-arg tests

It turns out we have little to no test coverage for setTimeout() and
setInterval() calls with optional arguments.  Now we do.

PR-URL: https://github.com/iojs/io.js/pull/1221
Reviewed-By: Trevor Norris <trev.norris@gmail.com>
9 years agolib: don't penalize setInterval() common case
Ben Noordhuis [Fri, 20 Mar 2015 16:20:43 +0000 (17:20 +0100)]
lib: don't penalize setInterval() common case

The common case is where setInterval() is called with two arguments,
the callback and the timeout.  Specifying optional arguments in
the parameter list forces common case calls to go through an arguments
adaptor stack frame.

PR-URL: https://github.com/iojs/io.js/pull/1221
Reviewed-By: Trevor Norris <trev.norris@gmail.com>
9 years agolib: don't penalize setTimeout() common case
Ben Noordhuis [Fri, 20 Mar 2015 15:55:19 +0000 (16:55 +0100)]
lib: don't penalize setTimeout() common case

The common case is where setTimeout() is called with two arguments,
the callback and the timeout.  Specifying optional arguments in the
parameter list forces common case calls to go through an arguments
adaptor stack frame.

PR-URL: https://github.com/iojs/io.js/pull/1221
Reviewed-By: Trevor Norris <trev.norris@gmail.com>
9 years agosrc: reset signal handler to SIG_DFL on FreeBSD
Fedor Indutny [Fri, 20 Mar 2015 05:03:34 +0000 (22:03 -0700)]
src: reset signal handler to SIG_DFL on FreeBSD

FreeBSD has a nasty bug with SA_RESETHAND reseting the SA_SIGINFO,
that is in turn set for a libthr wrapper. This leads to a crash.
Work around the issue by manually setting SIG_DFL in the signal
handler.

Fix: https://github.com/joyent/node/issues/9326
PR-URL: https://github.com/iojs/io.js/pull/1218
Reviewed-By: Johan Bergström <bugs@bergstroem.nu>
Reviewed-By: Ben Noordhuis <info@bnoordhuis.nl>
9 years agomodule: allow require('.')
Michaël Zasso [Wed, 18 Mar 2015 07:38:00 +0000 (08:38 +0100)]
module: allow require('.')

Previously, the minimal argument to require the current directory was
require('./'). This commits allows to skip the trailing slash.

Fixes: https://github.com/iojs/io.js/issues/1178
PR-URL: https://github.com/iojs/io.js/pull/1185
Reviewed-By: Chris Dickinson <christopher.s.dickinson@gmail.com>
Reviewed-By: Christian Tellnes <christian@tellnes.no>
Reviewed-By: Roman Reiss <me@silverwind.io>
9 years agoWorking on v1.6.2
Rod Vagg [Fri, 20 Mar 2015 02:24:09 +0000 (13:24 +1100)]
Working on v1.6.2

9 years ago2015-03-20 io.js v1.6.1 Release v1.6.1
Rod Vagg [Fri, 20 Mar 2015 02:19:18 +0000 (13:19 +1100)]
2015-03-20 io.js v1.6.1 Release

Notable Changes:

* path: New type-checking on path.resolve()
  <https://github.com/iojs/io.js/pull/1153> uncovered some edge-cases
  being relied upon in the wild, most notably path.dirname(undefined).
  Type-checking has been loosened for path.dirname(), path.basename(),
  and path.extname(), (Colin Ihrig)
  <https://github.com/iojs/io.js/pull/1216>.
* querystring: Internal optimizations in querystring.parse() and
  querystring.stringify() <https://github.com/iojs/io.js/pull/847>
  prevented Number literals from being properly converted via
  querystring.escape() <https://github.com/iojs/io.js/pull/1208>,
  exposing a blind-spot in the test suite. The bug and the tests have
  now been fixed (Jeremiah Senkpiel)
  <https://github.com/iojs/io.js/pull/1213>.

9 years agoquerystring: fix broken stringifyPrimitive
Jeremiah Senkpiel [Fri, 20 Mar 2015 01:33:04 +0000 (21:33 -0400)]
querystring: fix broken stringifyPrimitive

stringifyPrimitive has always failed to stringify numbers since its
introduction in 422d3c9. This went uncaught due to encodeURIComponent's
string coercion.

Fixes: https://github.com/iojs/io.js/issues/1208
PR-URL: https://github.com/iojs/io.js/pull/1213
Reviewed-By: Rod Vagg <rod@vagg.org>
Reviewed-By: Brian White <mscdex@mscdex.net>
9 years agoquerystring: parse numbers correctly
Jeremiah Senkpiel [Thu, 19 Mar 2015 22:51:55 +0000 (18:51 -0400)]
querystring: parse numbers correctly

Fixes a number parsing regression introduced in 85a92a3

Fixes: https://github.com/iojs/io.js/issues/1208
PR-URL: https://github.com/iojs/io.js/pull/1213
Reviewed-By: Rod Vagg <rod@vagg.org>
Reviewed-By: Brian White <mscdex@mscdex.net>
9 years agopath: reduce type checking on some methods
cjihrig [Fri, 20 Mar 2015 00:55:31 +0000 (20:55 -0400)]
path: reduce type checking on some methods

a465840313f548b913eb2bd8ea3d26c2ab5dcebb added strict type
checking for the methods in the path module. However, dirname(),
basename(), and extname() actually had some undocumented uses
in the wild. This commit loosens the type checking on those
methods.

Fixes: https://github.com/iojs/io.js/issues/1215
PR-URL: https://github.com/iojs/io.js/pull/1216
Reviewed-By: Rod Vagg <rod@vagg.org>
9 years agobuild: make check aliases test
Johan Bergström [Thu, 19 Mar 2015 22:15:54 +0000 (09:15 +1100)]
build: make check aliases test

A common convention in auto* is to call make check rather than
make test. Let one alias the other

PR-URL: https://github.com/iojs/io.js/pull/1211
Reviewed-By: Ben Noordhuis <info@bnoordhuis.nl>
9 years agoconfigure: use cc and c++ as defaults on os x
Ben Noordhuis [Thu, 19 Mar 2015 21:57:11 +0000 (22:57 +0100)]
configure: use cc and c++ as defaults on os x

Commit 8b2363d ("configure: use gcc and g++ as CC and CXX defaults")
switches the CC and CXX defaults but it turns out that GYP uses cc
and c++ as defaults on OS X.

It also made the configure script complain about old compilers because
the xcode gcc identifies as v4.2.1, whereas cc is less ambiguous about
it being a clang hybrid.

PR-URL: https://github.com/iojs/io.js/pull/1210
Reviewed-By: Jeremiah Senkpiel <fishrock123@rocketmail.com>
Reviewed-By: Johan Bergström <bugs@bergstroem.nu>
9 years agosmalloc: don't mix malloc() and new char[]
Ben Noordhuis [Thu, 19 Mar 2015 12:35:47 +0000 (13:35 +0100)]
smalloc: don't mix malloc() and new char[]

It's technically undefined behavior to mix malloc with delete[] and
new char[] with free().  smalloc was using new char[] in one place and
malloc() in another but in both cases the memory was freed with free().

PR-URL: https://github.com/iojs/io.js/pull/1205
Reviewed-By: Trevor Norris <trev.norris@gmail.com>
9 years agoWorking on v1.6.1
Chris Dickinson [Thu, 19 Mar 2015 18:28:03 +0000 (11:28 -0700)]
Working on v1.6.1

9 years ago2015-03-19 io.js v1.6.0 Release v1.6.0
Chris Dickinson [Thu, 19 Mar 2015 18:12:05 +0000 (11:12 -0700)]
2015-03-19 io.js v1.6.0 Release

Notable Changes:

* node: a new -r or --require command-line option can be used to
  pre-load modules at start-up (Ali Ijaz Sheikh)
* querystring: parse() and stringify() are now faster (Brian White)
* http: the http.ClientRequest#flush() method has been deprecated and
  replaced with http.ClientRequest#flushHeaders() to match the same
  change now in Node.js v0.12 as per
  https://github.com/joyent/node/pull/9048 (Yosuke Furukawa)
* net: allow server.listen() to accept a String option for port, e.g.
  { port: "1234" }, to match the same option being accepted in
  net.connect() as of https://github.com/joyent/node/pull/9268 (Ben
  Noordhuis)
* tls: further work on the reported memory leak although there appears
  to be a minor leak remaining for the use-case in question, track
  progress at https://github.com/iojs/io.js/issues/1075.
* v8: backport a fix for an integer overflow when --max_old_space_size
  values above 4096 are used (Ben Noordhuis)
* platforms: the io.js CI system now reports passes on FreeBSD and
  SmartOS (_Solaris_).
* npm: upgrade npm to 2.7.1. See the npm CHANGELOG.md for details.
  https://github.com/npm/npm/blob/master/CHANGELOG.md#v271-2015-03-05

9 years agodeps: upgrade to openssl-1.0.1m
Shigeki Ohtsu [Thu, 19 Mar 2015 14:04:41 +0000 (23:04 +0900)]
deps: upgrade to openssl-1.0.1m

All sources are just extracted from tarball into deps/openssl/openssl.

change all openssl/include/openssl/*.h to include resolved symbolic
links and openssl/crypto/opensslconf.h to refer config/opensslconf.h

sha256-x86_64.pl does not exist in the origin openssl distribution. It
was copied from sha512-x86_64.pl and both sha256/sha512 scripts were
modified so as to generates only one asm file specified as its key
hash length.

`x86masm.pl` was mistakenly using .486 instruction set, why `cpuid`
(and perhaps others) are requiring .686.

removed vpaesni-x86_64.asm in x64-win32-masm - it is no longer used.

Fixes: https://github.com/iojs/io.js/issues/1186
PR-URL: https://github.com/iojs/io.js/pull/1206
Reviewed-By: Fedor Indutny <fedor@indutny.com>
Reviewed-By: Ben Noordhuis <info@bnoordhuis.nl>
9 years agodoc: fix a broken collaborator github link
Aleksanteri Negru-Vode [Thu, 19 Mar 2015 13:21:33 +0000 (15:21 +0200)]
doc: fix a broken collaborator github link

PR-URL: https://github.com/iojs/io.js/pull/1204
Reviewed-By: Jeremiah Senkpiel <fishrock123@rocketmail.com>
9 years agobuffer: removing duplicate code
Thorsten Lorenz [Fri, 13 Mar 2015 19:24:01 +0000 (15:24 -0400)]
buffer: removing duplicate code

- using an overload of Alloc that does the same that was being done
  inside `Buffer::New`

The overload we now call inside `smalloc.cc` takes care of the same as
the code that was removed:

    if (length == 0)
      return Alloc(env, obj, nullptr, length, type);

    char* data = static_cast<char*>(malloc(length));
    if (data == nullptr) {
      FatalError("node::smalloc::Alloc(v8::Handle<v8::Object>, size_t,"
                  " v8::ExternalArrayType)", "Out Of Memory");
    }

    Alloc(env, obj, data, length, type);

PR-URL: https://github.com/iojs/io.js/pull/1144
Reviewed-By: Ben Noordhuis <info@bnoordhuis.nl>
Reviewed-By: Trevor Norris <trev.norris@gmail.com>
9 years agosrc: add -r/--require flags for preloading modules
Ali Ijaz Sheikh [Tue, 17 Feb 2015 22:37:37 +0000 (14:37 -0800)]
src: add -r/--require flags for preloading modules

This patch adds a command line option (-r/--require) that allows one
to provide modules on the command line that will be 'required' during
node startup. This can be useful for debugging, tracing, memory leak
analysis etc. to be preloaded without explicit changes to the user
script. The option can be repeated to preload multiple modules.

PR-URL: https://github.com/iojs/io.js/pull/881
Reviewed-By: Trevor Norris <trev.norris@gmail.com>
Reviewed-By: Chris Dickinson <christopher.s.dickinson@gmail.com>
Reviewed-By: Rod Vagg <rod@vagg.org>
9 years agotest: cache lazy properties, fix style nits
Rod Vagg [Wed, 18 Mar 2015 22:27:26 +0000 (09:27 +1100)]
test: cache lazy properties, fix style nits

inFreeBSDJail involves an execSync() and is used by localhost_ipv4 so
will be unnecessarily expensive, so cache both values and reuse
rather than re-evaluate each time.

Renamed localhost_ipv4 to localhostIPv4 for style consistency.

PR-URL: https://github.com/iojs/io.js/pull/1196
Reviewed-By: Johan Bergström <bugs@bergstroem.nu>
9 years agotest: double timeout in tls-wrap-timeout.js
Fedor Indutny [Thu, 19 Mar 2015 03:15:06 +0000 (20:15 -0700)]
test: double timeout in tls-wrap-timeout.js

The test is timing dependent, ensure that it won't fail on the busy CI
boxes.

Fix: https://github.com/iojs/io.js/issues/1200
PR-URL: https://github.com/iojs/io.js/pull/1201
Reviewed-By: Johan Bergström <bugs@bergstroem.nu>
9 years agobuild: remove incorrect argument in vcbuild.bat
Jeremiah Senkpiel [Thu, 19 Mar 2015 01:03:14 +0000 (21:03 -0400)]
build: remove incorrect argument in vcbuild.bat

This change had been incorrectly committed in f19e9b6

PR-URL: https://github.com/iojs/io.js/pull/1198
Reviewed-By: Rod Vagg <rod@vagg.org>
9 years agolib: don't error in repl when cwd doesn't exist
Ben Noordhuis [Wed, 18 Mar 2015 21:31:16 +0000 (22:31 +0100)]
lib: don't error in repl when cwd doesn't exist

The current working directory may not exist when the REPL starts up.
Don't treat that as an error because it's still possible to do many
useful things.  This is like the previous commit but for the REPL.

Fixes: https://github.com/iojs/io.js/issues/1184
PR-URL: https://github.com/iojs/io.js/pull/1194
Reviewed-By: Johan Bergström <bugs@bergstroem.nu>
Reviewed-By: Rod Vagg <rod@vagg.org>
9 years agosrc: don't error at startup when cwd doesn't exist
Ben Noordhuis [Wed, 18 Mar 2015 21:11:14 +0000 (22:11 +0100)]
src: don't error at startup when cwd doesn't exist

The current working directory may not exist when iojs starts up.  Don't
treat that as an error because it's still possible to do many useful
things, like evaluating a command line script or starting a REPL.

This commit also fixes an age-old Windows bug where process.argv[0] was
not properly expanded, that's why the parallel/test-process-argv-0 test
gets an update as well.

Fixes: https://github.com/iojs/io.js/issues/1184
PR-URL: https://github.com/iojs/io.js/pull/1194
Reviewed-By: Johan Bergström <bugs@bergstroem.nu>
Reviewed-By: Rod Vagg <rod@vagg.org>
9 years agotest: Introduce knowledge of FreeBSD jails
Johan Bergström [Tue, 17 Mar 2015 01:06:48 +0000 (12:06 +1100)]
test: Introduce knowledge of FreeBSD jails

FreeBSD jails act differently than your average vm or similar
application container. All routing passes through one ip address,
which makes things like localhost or 0.0.0.0 resolve differently.

Introduce a helper that allows us to verify if we're in a jail
and another one for returning an ip address for localhost.

Also, skip one test instead of trading additional complexity
in common.js for one specific user scenario.

PR-URL: https://github.com/iojs/io.js/pull/1167
Reviewed-By: Ben Noordhuis <info@bnoordhuis.nl>
9 years agosrc: fix crypto bio integer wraparound on 32 bits
Ben Noordhuis [Wed, 18 Mar 2015 18:20:23 +0000 (19:20 +0100)]
src: fix crypto bio integer wraparound on 32 bits

Fix a bug where a size_t was negated and passed to a function that takes
an int64_t.  It works by accident when sizeof(size_t) == sizeof(int64_t)
but it causes the value to underflow when size_t is a 32 bits type.

v8::Isolate::AdjustAmountOfExternalAllocatedMemory() is the function I'm
talking about.  The goal of that call is to tell V8 that some memory has
been freed but due to that underflow, we were actually reporting that we
had just allocated gigabytes of memory.  It set off a garbage collector
frenzy and essentially brought the VM to a standstill.

Fixes: https://github.com/iojs/io.js/issues/1188
PR-URL: https://github.com/iojs/io.js/pull/1192
Reviewed-By: Fedor Indutny <fedor@indutny.com>
9 years agodoc: add yosuke-furukawa as collaborator
Yosuke Furukawa [Wed, 18 Mar 2015 00:33:39 +0000 (09:33 +0900)]
doc: add yosuke-furukawa as collaborator

PR-URL: https://github.com/iojs/io.js/pull/1183
Reviewed-By: Ben Noordhuis <info@bnoordhuis.nl>
Reviewed-By: Rod Vagg <rod@vagg.org>
Reviewed-By: Shigeki Ohtsu <ohtsu@iij.ad.jp>
9 years agodoc: update test section in CONTRIBUTING.md
Ben Noordhuis [Tue, 17 Mar 2015 19:46:32 +0000 (20:46 +0100)]
doc: update test section in CONTRIBUTING.md

test/simple no longer exists, tell contributors to add their tests to
test/parallel.

PR-URL: https://github.com/iojs/io.js/pull/1181
Reviewed-By: Jeremiah Senkpiel <fishrock123@rocketmail.com>
9 years agodoc: add petkaantonov as collaborator
Petka Antonov [Tue, 17 Mar 2015 19:53:57 +0000 (21:53 +0200)]
doc: add petkaantonov as collaborator

PR-URL: https://github.com/iojs/io.js/pull/1179
Reviewed-By: Ben Noordhuis <info@bnoordhuis.nl>
Reviewed-By: cjihrig <cjihrig@gmail.com>
9 years agodoc: add silverwind as collaborator
Roman Reiss [Tue, 17 Mar 2015 18:24:31 +0000 (19:24 +0100)]
doc: add silverwind as collaborator

PR-URL: https://github.com/iojs/io.js/pull/1176
Reviewed-By: Jeremiah Senkpiel <fishrock123@rocketmail.com>
9 years agodoc: add jbergstroem as collaborator
Johan Bergström [Tue, 17 Mar 2015 17:52:48 +0000 (04:52 +1100)]
doc: add jbergstroem as collaborator

PR-URL: https://github.com/iojs/io.js/pull/1175
Reviewed-By: Chris Dickinson <christopher.s.dickinson@gmail.com>
9 years agoconfigure: use gcc and g++ as CC and CXX defaults
Ben Noordhuis [Tue, 17 Mar 2015 15:36:36 +0000 (16:36 +0100)]
configure: use gcc and g++ as CC and CXX defaults

It matches what GYP's Makefile generator does and it should improve
compiler detection because cc and c++ are not always gcc and g++.

Fixes: https://github.com/iojs/io.js/issues/1173
PR-URL: https://github.com/iojs/io.js/pull/1174
Reviewed-By: Johan Bergström <bugs@bergstroem.nu>
9 years agodoc: fix typo in buffer module documentation
Alex Yursha [Tue, 17 Mar 2015 10:28:36 +0000 (13:28 +0300)]
doc: fix typo in buffer module documentation

PR-URL: https://github.com/iojs/io.js/pull/1169
Reviewed-By: Ben Noordhuis <info@bnoordhuis.nl>
9 years agobenchmark: add output format option [csv]
Brendan Ashworth [Tue, 10 Feb 2015 03:36:52 +0000 (19:36 -0800)]
benchmark: add output format option [csv]

This commit adds an `OUTPUT_FORMAT` environment variable option for
all benchmark tests that allow either 'csv' or 'default' output. Default
output has been left unchanged, and csv output prints out the csv
headers along with the csv formatted per-test output, each test also
seperated by a newline.

It can be used like the following:
$ OUTPUT_FORMAT=csv iojs benchmark/common.js http

Not specifying the OUTPUT_FORMAT env var will default to 'default'.
Specifying a bad value will throw an error.

PR-URL: https://github.com/iojs/io.js/pull/777
Reviewed-By: Ben Noordhuis <info@bnoordhuis.nl>
9 years agobenchmark: add plot_csv R graphing script
Brendan Ashworth [Tue, 10 Feb 2015 04:25:47 +0000 (20:25 -0800)]
benchmark: add plot_csv R graphing script

This commit adds a graphing script (in R) for graphing the CSV output
of a benchmark. It can be run like this:

```
$ OUTPUT_FORMAT=csv iojs benchmark/http/client-request-body.js >
data.csv
$ ./benchmark/plot_csv.R data.csv graph.png bytes type
```

This will graph the output to `graph.png`, using the output's `bytes`
value as X and the result value for each as Y. Output will be grouped
by `type`.

Running as the example yields a beautiful graph like this:
http://pbrd.co/1vBhUfy.

PR-URL: https://github.com/iojs/io.js/pull/777
Reviewed-By: Ben Noordhuis <info@bnoordhuis.nl>
9 years agov8: fix --max_old_space_size=4096 integer overflow
Ben Noordhuis [Mon, 9 Feb 2015 07:00:32 +0000 (23:00 -0800)]
v8: fix --max_old_space_size=4096 integer overflow

See https://code.google.com/p/v8/issues/detail?id=3857 for the bug
report and https://codereview.chromium.org/897543002 for the CL.

PR-URL: https://github.com/iojs/io.js/pull/1166
Reviewed-By: Fedor Indutny <fedor@indutny.com>
9 years agohttp: add flushHeaders and deprecate flush
Yosuke Furukawa [Sat, 14 Mar 2015 07:24:07 +0000 (16:24 +0900)]
http: add flushHeaders and deprecate flush

PR-URL: https://github.com/iojs/io.js/pull/1156
Reviewed-By: Ben Noordhuis <info@bnoordhuis.nl>
Reviewed-By: Christian Tellnes <christian@tellnes.no>
Reviewed-By: Colin Ihrig <cjihrig@gmail.com>
Reviewed-By: Jeremiah Senkpiel <fishrock123@rocketmail.com>
9 years agomake: remove node_dtrace from cpplint excludes
Julien Gilli [Mon, 17 Nov 2014 20:00:11 +0000 (12:00 -0800)]
make: remove node_dtrace from cpplint excludes

PR-URL: https://github.com/joyent/node/pull/8741
Reviewed-By: Fedor Indutny <fedor@indutny.com>
9 years agonet: use cached peername to resolve remote fields
James Hartig [Tue, 10 Mar 2015 20:48:19 +0000 (16:48 -0400)]
net: use cached peername to resolve remote fields

Allows socket.remote* properties to still be accessed even after the
socket is closed.

Fixes: https://github.com/joyent/node/issues/9287
PR-URL: https://github.com/joyent/node/pull/9366
Reviewed-By: Colin Ihrig <cjihrig@gmail.com>
9 years agodoc: fix '\\' typos on Windows
Steven Vercruysse [Sun, 15 Mar 2015 11:29:18 +0000 (12:29 +0100)]
doc: fix '\\' typos on Windows

This commit changes the Windows examples in path.markdown to
correctly display '\\'.

PR-URL: https://github.com/joyent/node/pull/9412
Reviewed-By: Colin Ihrig <cjihrig@gmail.com>
9 years agobuild: allow custom PackageMaker path
Julien Gilli [Wed, 11 Mar 2015 18:45:57 +0000 (11:45 -0700)]
build: allow custom PackageMaker path

Make PACKAGEMAKER customizable because PackageMaker is not necessarily
installed in /Developer on OSX anymore.

PR-URL: https://github.com/joyent/node/pull/9377
Reviewed-By: Timothy J Fontaine <tjfontaine@gmail.com>
9 years agolib: remove broken NODE_MODULE_CONTEXTS feature
Ben Noordhuis [Mon, 16 Mar 2015 11:34:39 +0000 (12:34 +0100)]
lib: remove broken NODE_MODULE_CONTEXTS feature

This feature has no tests and has been broken for ages, see for example
https://github.com/iojs/io.js/pull/1160.  Don't bother fixing it, it's
pretty much broken by design and there can't be too many users because
it's almost undocumented.  A quick Google search suggests that it causes
more grief than joy to the few that do use it.  Remove it.

PR-URL: https://github.com/iojs/io.js/pull/1162
Reviewed-By: Colin Ihrig <cjihrig@gmail.com>
9 years agosrc: use Number::New() for heapTotal/heapUsed
Ben Noordhuis [Fri, 13 Mar 2015 20:54:40 +0000 (21:54 +0100)]
src: use Number::New() for heapTotal/heapUsed

With --max_old_space_size=12345 it's possible to create a JS heap that
is larger than what fits in an unsigned int so use Number::New() rather
than Integer::NewFromUnsigned().

Performance-wise, it doesn't matter much.  If V8 can fit the double in
a SMI, it will.

PR-URL: https://github.com/iojs/io.js/pull/1148
Reviewed-By: Trevor Norris <trev.norris@gmail.com>
9 years agosrc: don't create js string twice on error
Ben Noordhuis [Fri, 13 Mar 2015 18:15:18 +0000 (19:15 +0100)]
src: don't create js string twice on error

Rewrite ErrnoException() so that it doesn't turn the file path into a
string twice.

PR-URL: https://github.com/iojs/io.js/pull/1148
Reviewed-By: Trevor Norris <trev.norris@gmail.com>
9 years agopath: add type checking for path inputs
cjihrig [Sat, 14 Mar 2015 02:21:47 +0000 (22:21 -0400)]
path: add type checking for path inputs

This commit adds type checking of path inputs to exported methods
in the path module. The exception is _makeLong(), which seems to
explicitly support any data type.

Fixes: https://github.com/iojs/io.js/issues/1139
PR-URL: https://github.com/iojs/io.js/pull/1153
Reviewed-By: Ben Noordhuis <info@bnoordhuis.nl>
Reviewed-By: Jeremiah Senkpiel <fishrock123@rocketmail.com>
Reviewed-By: Brendan Ashworth <brendan.ashworth@me.com>
9 years agodoc: reflect new require('events') behaviour
Alex Yursha [Thu, 26 Feb 2015 21:47:35 +0000 (00:47 +0300)]
doc: reflect new require('events') behaviour

We don't need to do `require('events').EventEmitter` any longer.

PR-URL: https://github.com/iojs/io.js/pull/975
Reviewed-By: Ben Noordhuis <info@bnoordhuis.nl>
Reviewed-By: Brendan Ashworth <brendan.ashworth@me.com>
9 years agoquerystring: optimize parse and stringify
Brian White [Thu, 19 Feb 2015 14:53:04 +0000 (09:53 -0500)]
querystring: optimize parse and stringify

parse optimizations:

* Move try-catch to separate function to keep entire function from
being deoptimized.
* Use key array lookup instead of using hasOwnProperty.
* Avoid decoding known empty strings.
* Avoid possibly unnecessary switch to slower decoder for values if
key decoding throws.

stringify optimizations:

* Use manual loop for default encoder instead of encodeURIComponent.
* Use string concatenation instead of joining an array of strings.
* Avoid caching result of typeof.

PR-URL: https://github.com/iojs/io.js/pull/847
Reviewed-By: Trevor Norris <trev.norris@gmail.com>
9 years agodeps: make node-gyp work with io.js
cjihrig [Fri, 27 Feb 2015 18:40:58 +0000 (13:40 -0500)]
deps: make node-gyp work with io.js

Every npm version bump requires a few patches to be floated on
node-gyp for io.js compatibility. These patches are found in
03d199276e21c1fa08d8df14eeb654c90cc5aa20,
5de334c23096492014a097ff487f07ad8eaee6d2, and
da730c76e98fb9fd18dac445dafbbec74d79f802. This commit squashes
them into a single commit.

PR-URL: https://github.com/iojs/io.js/pull/990
Reviewed-By: Ben Noordhuis <info@bnoordhuis.nl>
9 years agodeps: upgrade npm to 2.7.1
Forrest L Norvell [Fri, 13 Mar 2015 09:07:27 +0000 (02:07 -0700)]
deps: upgrade npm to 2.7.1

PR-URL: https://github.com/iojs/io.js/pull/1142
Reviewed-By: Jeremiah Senkpiel <fishrock123@rocketmail.com>
Reviewed-By: Rod Vagg <rod@vagg.org>
9 years agotls: re-enable `.writev()` on TLSWrap
Fedor Indutny [Sat, 14 Mar 2015 05:41:26 +0000 (22:41 -0700)]
tls: re-enable `.writev()` on TLSWrap

Fix the `parallel/test-tls-over-http-tunnel.js` on Windows by
re-enabling the accidentally disabled `.writev()` method on TLSWrap.

It appears that there is some subtle issue with shutdown timing and it
manifests itself when the chunks are written in separate packets. This
leads to concurrent `shutdown`/`destroy`, which breaks the test.

PR-URL: https://github.com/iojs/io.js/pull/1155
Reviewed-By: Bert Belder <bertbelder@gmail.com>
9 years agotls: fix leak on `DoWrite()` errors
Fedor Indutny [Sat, 14 Mar 2015 04:40:48 +0000 (21:40 -0700)]
tls: fix leak on `DoWrite()` errors

It is very unlikely to happen, but still the write request should be
disposed in case of immediate failure.

PR-URL: https://github.com/iojs/io.js/pull/1154
Reviewed-By: Ben Noordhuis <info@bnoordhuis.nl>
9 years agosrc: revert -r/--require flags
Chris Dickinson [Fri, 13 Mar 2015 21:57:11 +0000 (14:57 -0700)]
src: revert -r/--require flags

This reverts commit 7bde3f1a8f53d82992a8fed73e5f93159bb400b3.

The added test (test/parallel/test-preload.js) fails on Windows.

PR-URL: https://github.com/iojs/io.js/pull/1150
Reviewed-By: Jeremiah Senkpiel <fishrock123@rocketmail.com>
9 years agodoc: fix vm module examples
FangDun Cai [Fri, 13 Mar 2015 21:07:31 +0000 (05:07 +0800)]
doc: fix vm module examples

PR-URL: https://github.com/iojs/io.js/pull/1147
Reviewed-By: Ben Noordhuis <info@bnoordhuis.nl>
9 years agosrc: add -r/--require flags for preloading modules
Ali Ijaz Sheikh [Tue, 17 Feb 2015 22:37:37 +0000 (14:37 -0800)]
src: add -r/--require flags for preloading modules

-r/--require can be used to preload modules on node startup. The option
takes a single module name. The option can be repeated as necessary to
preload multiple modules.

This patch allows 'vendors' (such a cloud host) to inject functionality
that gets executed at application startup without requiring an explicit
require from the user's application. This can be useful to load vendor
specific application monitoring APIs transparently.

PR-URL: https://github.com/iojs/io.js/pull/881
Reviewed-By: Trevor Norris <trev.norris@gmail.com>
Reviewed-By: Chris Dickinson <christopher.s.dickinson@gmail.com>
9 years agotest: fix test-http-content-length
Jeremiah Senkpiel [Fri, 13 Mar 2015 19:26:27 +0000 (15:26 -0400)]
test: fix test-http-content-length

Previously the test did not allow the last request to complete.

Fixes: https://github.com/iojs/io.js/pull/1137
PR-URL: https://github.com/iojs/io.js/pull/1145
Reviewed-By: Brian White <mscdex@mscdex.net>
Reviewed-By: Bert Belder <bertbelder@gmail.com>
9 years agocrypto: add deprecated ValiCert CA for cross cert
Shigeki Ohtsu [Fri, 13 Mar 2015 00:55:29 +0000 (09:55 +0900)]
crypto: add deprecated ValiCert CA for cross cert

The host of melissadata.net has a cross root certification between
Starfield Class 2 and ValiCert Class 2. OpenSSL-1.0.1 only looks up
a cert chain to the deprecated ValiCert Class 2 CA and causes
untrusted error. We add it for a short-term remedy and it is to be
removed after upgrading OpenSSSL-1.0.2 and applying private patches
to support alternative cert chains.
See #402 and #589.

Fixes: https://github.com/iojs/io.js/issues/923
PR-URL: https://github.com/iojs/io.js/pull/1135
Reviewed-By: Ben Noordhuis <info@bnoordhuis.nl>
9 years agotest: fix ext commands to be double quoted
Shigeki Ohtsu [Thu, 12 Mar 2015 01:16:26 +0000 (10:16 +0900)]
test: fix ext commands to be double quoted

Paths used on the Windows command line need to be enclosed in double
quotes, or they'll be parsed incorrectly when there are spaces in the
path.

PR-URL: https://github.com/iojs/io.js/pull/1122
Reviewed-by: Bert Belder <bertbelder@gmail.com>