isaacs [Wed, 18 Apr 2012 01:16:52 +0000 (18:16 -0700)]
Reapply patches floating on V8
isaacs [Wed, 18 Apr 2012 01:16:27 +0000 (18:16 -0700)]
Upgrade V8 to 3.9.24.9
isaacs [Wed, 18 Apr 2012 00:14:25 +0000 (17:14 -0700)]
Upgrade npm to 1.1.17
isaacs [Tue, 17 Apr 2012 22:52:57 +0000 (15:52 -0700)]
lint
isaacs [Tue, 17 Apr 2012 22:45:48 +0000 (15:45 -0700)]
uv: Cherry-pick joyent/libuv@3c41597
isaacs [Tue, 17 Apr 2012 22:30:22 +0000 (15:30 -0700)]
test: Minor change to message/undefined_reference_in_new_context
Dave Pacheco [Tue, 17 Apr 2012 20:50:09 +0000 (13:50 -0700)]
Fix 64-bit SmartOS build
isaacs [Tue, 17 Apr 2012 20:36:29 +0000 (13:36 -0700)]
Merge branch 'domains3'
isaacs [Thu, 12 Apr 2012 15:16:30 +0000 (08:16 -0700)]
Add Todo comments about domain-ifying crypto
isaacs [Tue, 10 Apr 2012 21:59:21 +0000 (14:59 -0700)]
Add switches to http_simple bench to use domains
isaacs [Fri, 13 Apr 2012 23:27:23 +0000 (16:27 -0700)]
Domain hooks in ReqWrap<T> and MakeCallback
isaacs [Fri, 6 Apr 2012 23:26:18 +0000 (16:26 -0700)]
Domain feature
This is a squashed commit of the main work done on the domains-wip branch.
The original commit messages are preserved for posterity:
* Implicitly add EventEmitters to active domain
* Implicitly add timers to active domain
* domain: add members, remove ctor cb
* Don't hijack bound callbacks for Domain error events
* Add dispose method
* Add domain.remove(ee) method
* A test of multiple domains in process at once
* Put the active domain on the process object
* Only intercept error arg if explicitly requested
* Typo
* Don't auto-add new domains to the current domain
While an automatic parent/child relationship is sort of neat,
and leads to some nice error-bubbling characteristics, it also
results in keeping a reference to every EE and timer created,
unless domains are explicitly disposed of.
* Explicitly adding one domain to another is still fine, of course.
* Don't allow circular domain->domain memberships
* Disposing of a domain removes it from its parent
* Domain disposal turns functions into no-ops
* More documentation of domains
* More thorough dispose() semantics
* An example using domains in an HTTP server
* Don't handle errors on a disposed domain
* Need to push, even if the same domain is entered multiple times
* Array.push is too slow for the EE Ctor
* lint domain
* domain: docs
* Also call abort and destroySoon to clean up event emitters
* domain: Wrap destroy methods in a try/catch
* Attach tick callbacks to active domain
* domain: Only implicitly bind timers, not explicitly
* domain: Don't fire timers when disposed.
* domain: Simplify naming so that MakeCallback works on Timers
* Add setInterval and nextTick to domain test
* domain: Make stack private
isaacs [Thu, 12 Apr 2012 23:03:47 +0000 (16:03 -0700)]
MakeCallback: Consistent symbol usage
isaacs [Thu, 12 Apr 2012 05:29:36 +0000 (22:29 -0700)]
MakeCallback: Use in node_signal_watcher
isaacs [Thu, 12 Apr 2012 05:29:18 +0000 (22:29 -0700)]
MakeCallback: Use in node_io_watcher
isaacs [Thu, 12 Apr 2012 05:28:44 +0000 (22:28 -0700)]
MakeCallback: Use in node_crypto
isaacs [Sun, 8 Apr 2012 07:48:10 +0000 (00:48 -0700)]
MakeCallback: Use in node_file.cc
isaacs [Tue, 17 Apr 2012 04:27:12 +0000 (21:27 -0700)]
MakeCallback: abort() if not a function
isaacs [Fri, 13 Apr 2012 23:34:48 +0000 (16:34 -0700)]
MakeCallback: Return the callback return value
isaacs [Fri, 13 Apr 2012 23:33:09 +0000 (16:33 -0700)]
MakeCallback: Accept Function or Symbol argument
isaacs [Sat, 7 Apr 2012 02:23:16 +0000 (19:23 -0700)]
Report errors thrown from uncaughtException handers
Nathan Rajlich [Tue, 17 Apr 2012 18:49:31 +0000 (11:49 -0700)]
test: lowercase setRawMode() test case filename
Nathan Rajlich [Tue, 17 Apr 2012 18:44:54 +0000 (11:44 -0700)]
doc: update the `readline` docs for the "close" event behavior
Nathan Rajlich [Tue, 17 Apr 2012 18:35:11 +0000 (11:35 -0700)]
repl: update the `repl` for the new `readline` behavior
This fixes the failing REPL tests.
Nathan Rajlich [Tue, 17 Apr 2012 18:34:37 +0000 (11:34 -0700)]
test: check for multiple "emit" calls in repl-end-emits-exit.js
Nathan Rajlich [Tue, 17 Apr 2012 18:31:07 +0000 (11:31 -0700)]
readline: re-add the Interface#close() method; rename "end" to "close"
The idea here is to reduce the number of times that `setRawMode()` is called
on the `input` stream, since it is expensive, and simply pause()/resume()
should not call it.
So now `setRawMode()` only gets called at the beginning of the Interface
instance, and then when `Interface#close()` is called.
Test case included.
Nathan Rajlich [Tue, 17 Apr 2012 18:17:54 +0000 (11:17 -0700)]
doc: change `in` to `input` in the readline docs
Also compacting some long lines.
Kyle Robinson Young [Tue, 17 Apr 2012 05:15:51 +0000 (22:15 -0700)]
doc: add require.extensions to globals
Closes #3028
Kyle Robinson Young [Mon, 16 Apr 2012 21:52:44 +0000 (14:52 -0700)]
doc: add cache argument to fs.realpath()
domenic [Mon, 9 Apr 2012 05:46:55 +0000 (01:46 -0400)]
doc: note that `stream.pause` is advisory
ssuda [Sun, 8 Apr 2012 15:24:53 +0000 (20:54 +0530)]
Fixing ClientRequest setTimeout EventEmitter Leak
This will fix #3068
Zachary Scott [Mon, 16 Apr 2012 20:43:02 +0000 (16:43 -0400)]
docs: fix broken links in zlib docs
Zachary Scott [Sun, 8 Apr 2012 16:07:28 +0000 (12:07 -0400)]
docs: url.format uses host for hostname and port, not auth
Fixes #3062.
Zachary Scott [Sat, 7 Apr 2012 15:10:22 +0000 (11:10 -0400)]
docs: url.host doesn't include auth
Fixes #3062.
fukayatsu [Mon, 16 Apr 2012 18:37:32 +0000 (03:37 +0900)]
tls: remove duplicate line
Shigeki Ohtsu [Fri, 13 Apr 2012 07:45:38 +0000 (16:45 +0900)]
tcp: make getsockname() return address family as string
Ben Noordhuis [Mon, 16 Apr 2012 15:57:41 +0000 (17:57 +0200)]
test: don't make request until server is listening
Preemptively fixes simple/test-https-timeout on platforms where binding to an
interface is not an instantaneous action.
ssuda [Fri, 13 Apr 2012 23:35:28 +0000 (05:05 +0530)]
http, https: fix .setTimeout()
Fixes #3107.
Kyle Robinson Young [Mon, 16 Apr 2012 05:24:25 +0000 (22:24 -0700)]
docs: more explicit argument names in fs
Maciej Małecki [Sun, 15 Apr 2012 09:09:37 +0000 (11:09 +0200)]
events: remove misleading comment
Ben Noordhuis [Sat, 14 Apr 2012 21:51:03 +0000 (21:51 +0000)]
deps: fix -DOPENSSL_NO_SOCK on sunos
The OPENSSL_NO_SOCK macro in OpenSSL missed a couple of networking functions
that called other functions that OPENSSL_NO_SOCK *had* filtered out. None of
the functions (filtered or not) were actually used but it was enough to trip
up the Solaris linker.
Mustansir Golawala [Thu, 12 Apr 2012 08:29:15 +0000 (01:29 -0700)]
os: add cross platform EOL character
isaacs [Fri, 13 Apr 2012 23:53:51 +0000 (16:53 -0700)]
Set errno in fake-failing child-process kill test
Ben Noordhuis [Fri, 13 Apr 2012 12:46:35 +0000 (14:46 +0200)]
dns: remove unused functions
Zachary Scott [Fri, 13 Apr 2012 00:24:35 +0000 (20:24 -0400)]
docs: rewrite "addons" docs to use node-gyp
Closes #3100.
Closes #3101.
Aaron Jacobs [Thu, 15 Mar 2012 02:26:35 +0000 (13:26 +1100)]
typed arrays: unexport SizeOfArrayElementForType()
It isn't used anywhere else, so made it an implementation detail in
v8_typed_array.cc.
Ben Noordhuis [Thu, 12 Apr 2012 17:13:04 +0000 (19:13 +0200)]
net: honor 'enable' flag in .setNoDelay()
Fixes #3096.
Andreas Madsen [Tue, 13 Mar 2012 16:22:47 +0000 (17:22 +0100)]
child_process: emit error when .kill fails
isaacs [Thu, 12 Apr 2012 01:26:52 +0000 (18:26 -0700)]
gyp: Apply 'argument too long' fix in another place
For some reason,
aa5961a445acbd2b533ef870eb19733be7b7ede5 caused
'make test' to rebuild the entire project every time. Applying
the fix to the other place where gyp chops up the argument list
makes it behave properly.
Bert Belder [Wed, 11 Apr 2012 16:20:54 +0000 (18:20 +0200)]
Disable OpenSSL UI
Ben Noordhuis [Tue, 10 Apr 2012 23:05:10 +0000 (01:05 +0200)]
test: merge tls-ext-key-usage into tls-securepair-client
Ben Noordhuis [Tue, 10 Apr 2012 22:47:52 +0000 (00:47 +0200)]
test: fix openssl tests
Don't assume that the libcrypto and libssl that we're linked against is the same
version as the openssl command line tool. This is important because the tool has
a bug in all pre-1.0.0 versions that makes it unusable for these tests.
Ben Noordhuis [Tue, 10 Apr 2012 21:41:34 +0000 (21:41 +0000)]
build: configure openssl
* compile with -DOPENSSL_NO_SOCK and -DOPENSSL_NO_DGRAM, we don't need it
* compile with -DOPENSSL_NO_GOST and -DOPENSSL_NO_HW_PADLOCK, works around the
brain dead linker on solaris and maybe others
* compile with -DTERMIOS, OS X doesn't have <termio.h>
* compile with -D__EXTENSIONS__ on solaris, makes siginfo_t available
* compile without -ansi on linux, it hides a number of POSIX declarations
(sigaction, NI_MAXHOST, etc.)
Ben Noordhuis [Tue, 10 Apr 2012 14:04:49 +0000 (16:04 +0200)]
deps: reapply 0a34755 to bundled openssl
Ben Noordhuis [Tue, 10 Apr 2012 14:08:50 +0000 (16:08 +0200)]
deps: reapply 0110c90 to bundled openssl
Ben Noordhuis [Tue, 10 Apr 2012 13:57:21 +0000 (15:57 +0200)]
deps: upgrade openssl to 1.0.0f
Ben Noordhuis [Tue, 10 Apr 2012 18:01:34 +0000 (20:01 +0200)]
gyp: fix 'argument list too long' build error
Ben Noordhuis [Wed, 11 Apr 2012 22:59:38 +0000 (00:59 +0200)]
Revert "deps: upgrade libuv to 3c41597"
This reverts commit
0db4dc0024eaa538bf4913d6bf256a18126de2ba.
This commit makes a lot of tests fail due to reference counting errors. It's
not worth it to debug because the reference counting scheme is due to change
soon anyway.
Bert Belder [Wed, 11 Apr 2012 20:02:12 +0000 (22:02 +0200)]
Slab allocator: don't attempt to shrink a non-buffer
Nathan Rajlich [Wed, 11 Apr 2012 18:16:47 +0000 (11:16 -0700)]
configure: output a newline at the end of config.gypi
Nathan Rajlich [Wed, 11 Apr 2012 18:16:11 +0000 (11:16 -0700)]
configure: don't use "with" for Python 2.5 and older
isaacs [Wed, 11 Apr 2012 01:29:25 +0000 (18:29 -0700)]
Fix #3089 Build changelog.html for website
isaacs [Wed, 11 Apr 2012 01:22:42 +0000 (18:22 -0700)]
Makefile: minor nit
Nathan Rajlich [Tue, 10 Apr 2012 21:39:29 +0000 (14:39 -0700)]
build: add comment explaining MACOSX_DEPLOYMENT_TARGET.
Nathan Rajlich [Tue, 10 Apr 2012 21:36:17 +0000 (14:36 -0700)]
Re-apply "build: target OSX 10.5 when building on darwin"
This reverts commit
93eca95aece9804d22b012cc89487f4b0064a86d.
Fixes #3072 (once again).
Ben Noordhuis [Tue, 10 Apr 2012 21:26:28 +0000 (23:26 +0200)]
deps: upgrade libuv to 3c41597
isaacs [Mon, 9 Apr 2012 15:39:13 +0000 (08:39 -0700)]
fs.WriteStream: Handle modifications to fs.open
If the fs.open method is modified via AOP-style extension, in between
the creation of an fs.WriteStream and the processing of its action
queue, then the test of whether or not the method === fs.open will fail,
because fs.open has been replaced.
The solution is to save a reference to fs.open on the stream itself when
the action is placed in the queue.
This fixes isaacs/node-graceful-fs#6.
isaacs [Mon, 9 Apr 2012 05:31:28 +0000 (22:31 -0700)]
Revert "build: target OSX 10.5 when building on darwin"
This reverts commit
b6d6a54f8057d0adad13d4bcc11eb3f443079ddc,
which fixed #3072, so we'll have to figure out another way
to make that work.
Nathan Rajlich [Fri, 6 Apr 2012 21:33:58 +0000 (14:33 -0700)]
readline: use StringDecoder for decoding "normal" data
The fix from #3059 was not handling multi-byte utf8 data properly.
Nathan Rajlich [Fri, 6 Apr 2012 20:50:34 +0000 (13:50 -0700)]
test: make the ArrayStream in repl tests write a '\n'
This is more correct.
Fixes them from failing with the updated readline behavior.
Nathan Rajlich [Fri, 6 Apr 2012 20:49:27 +0000 (13:49 -0700)]
test: make repl-autolibs check that the callback was invoked
Nathan Rajlich [Fri, 6 Apr 2012 19:54:48 +0000 (12:54 -0700)]
repl: make the completer use newlines
Fixes the repl.complete() function when terminal is false, since it
is now explicitly looking for a '\n' char.
Nathan Rajlich [Fri, 6 Apr 2012 19:20:01 +0000 (12:20 -0700)]
test: fix failing test-repl.js
Nathan Friedly [Fri, 6 Apr 2012 18:41:59 +0000 (11:41 -0700)]
readline: buffer data to only emit 'line' on '\n'
In "terminal: false" mode.
(And fire it multiple times if multiple lines arrive at once.)
This is necessary because the Windows telnet client sends every single
keystroke as it's typed.
See: http://stackoverflow.com/questions/9962197/node-js-readline-not-waiting-for-a-full-line-on-socket-connections
Closes #3059.
isaacs [Fri, 6 Apr 2012 21:16:09 +0000 (14:16 -0700)]
child_process: Emit err rather than throw for IPC write failure
Zachary Scott [Mon, 2 Apr 2012 02:55:09 +0000 (22:55 -0400)]
docs: grammar and spelling on lib/cluster.js
Nathan Rajlich [Thu, 5 Apr 2012 23:30:33 +0000 (16:30 -0700)]
build: target OSX 10.5 when building on darwin
Bert Belder [Wed, 4 Apr 2012 16:06:33 +0000 (18:06 +0200)]
Windows: installer shows license agreement dialog
Bert Belder [Wed, 4 Apr 2012 16:06:00 +0000 (18:06 +0200)]
Windows: add build step that generates license.rtf from LICENSE
Bert Belder [Tue, 3 Apr 2012 01:07:31 +0000 (03:07 +0200)]
Disable V8 postmortem debugging on Windows
It is not supported by V8.
Ryan Dahl [Mon, 2 Apr 2012 22:14:52 +0000 (15:14 -0700)]
Mac installer shows license
ref #3056
Ryan Dahl [Mon, 2 Apr 2012 22:12:23 +0000 (15:12 -0700)]
Include text of licenses in LICENSE file
For easy inclusion in binary distributions. ref #3056
Ben Noordhuis [Mon, 2 Apr 2012 21:53:05 +0000 (23:53 +0200)]
build: define _DARWIN_USE_64_BIT_INODE=1 on OS X
Fixes a segmentation fault on some OS X systems due to sizeof(struct stat)
mismatches.
Fixes #2061.
Ryan Dahl [Mon, 2 Apr 2012 21:36:23 +0000 (14:36 -0700)]
Fix links to libev and libeio licenses
isaacs [Sun, 1 Apr 2012 04:01:55 +0000 (21:01 -0700)]
Fix #3052 Handle errors properly in zlib
Dave Pacheco [Wed, 28 Mar 2012 17:26:10 +0000 (10:26 -0700)]
build: add support for DTrace and postmortem
* fixes #2110
* includes V8 postmortem metadata in Solaris builds
* adds GYP support for DTrace probes and ustack helper
* ustack helper derives constants dynamically from libv8_base.a
* build with DTrace support by default on SunOS
Ben Noordhuis [Sat, 31 Mar 2012 23:50:02 +0000 (23:50 +0000)]
zlib: fix uninitialized variable compiler warning
Ben Noordhuis [Sat, 31 Mar 2012 23:15:17 +0000 (01:15 +0200)]
node: provide snprintf implementation on windows
_snprintf() doesn't zero-terminate the buffer on overflow.
Ben Noordhuis [Sat, 31 Mar 2012 21:23:46 +0000 (23:23 +0200)]
node: don't check return value of unsetenv()
It returns void on some platforms, notably FreeBSD.
Ben Noordhuis [Sat, 31 Mar 2012 20:27:42 +0000 (22:27 +0200)]
Alias _snprintf to snprintf, fix Windows build.
isaacs [Sat, 31 Mar 2012 00:36:58 +0000 (17:36 -0700)]
Revert "tty: add keypress event for backwards-compat"
In this case, backwards compatibility is not worth the API
inconsistency. We can just document the change.
This reverts commit
b521ff3b4f8e79c5e28a1f4915d4203d8bcb0ab7.
Ben Noordhuis [Fri, 30 Mar 2012 20:54:13 +0000 (22:54 +0200)]
doc: fix grammar error in cluster docs
isaacs [Fri, 30 Mar 2012 19:24:46 +0000 (12:24 -0700)]
cluster: Rename 'death' back to 'exit'
Bert Belder [Fri, 30 Mar 2012 19:50:15 +0000 (21:50 +0200)]
Revert "Re-add top-level v8::Locker"
The locker makes node crash in debug mode sometimes.
For example, test/simple/test-repl.js triggers it.
This reverts commit
9a6012edd9330296b7476bc6b7fbda2cd5c8165d.
Conflicts:
src/node.cc
Ben Noordhuis [Fri, 30 Mar 2012 14:00:39 +0000 (16:00 +0200)]
udp_wrap: simplify AddressToJS() function
Ben Noordhuis [Fri, 30 Mar 2012 13:54:06 +0000 (15:54 +0200)]
udp_wrap: use new slab allocator
Ben Noordhuis [Thu, 29 Mar 2012 01:04:35 +0000 (03:04 +0200)]
stream_wrap: use new slab allocator
Ben Noordhuis [Wed, 28 Mar 2012 00:21:56 +0000 (02:21 +0200)]
core: add reusable slab allocator
Ben Noordhuis [Tue, 27 Mar 2012 13:42:51 +0000 (15:42 +0200)]
core: add ROUND_UP() macro
Ben Noordhuis [Mon, 26 Mar 2012 14:05:13 +0000 (07:05 -0700)]
typed arrays: root JS objects in HandleScope with Local<>