This commit addresses most of the review comments in
https://github.com/nodejs/node/pull/2540, which are kept in this
separate commit so as to better preserve the prior two patches as they
landed in 0.12.
This commit:
- Fixes a bug with unrefActive timers and disposed domains.
- Fixes a bug with unrolling an unrefActive timer from another.
- Adds a test for both above bugs.
- Improves check logic, making it stricter, simpler, or both.
- Optimizes nicer with a smaller, separate function for the try/catch.
Fixes: https://github.com/nodejs/node-convergence-archive/issues/23
Ref: https://github.com/nodejs/node/issues/268
PR-URL: https://github.com/nodejs/node/pull/2540
Reviewed-By: bnoordhuis - Ben Noordhuis <info@bnoordhuis.nl>
Before this change, _unrefActive would keep the unrefList sorted when
adding a new timer.
Because _unrefActive is called extremely frequently, this linear scan
(O(n) at worse) would make _unrefActive show high in the list of
contributors when profiling CPU usage.
This commit changes _unrefActive so that it doesn't try to keep the
unrefList sorted. The insertion thus happens in constant time.
However, when a timer expires, unrefTimeout has to go through the whole
unrefList because it's not ordered anymore.
It is usually not large enough to have a significant impact on
performance because:
- Most of the time, the timers will be removed before unrefTimeout is
called because their users (sockets mainly) cancel them when an I/O
operation takes place.
- If they're not, it means that some I/O took a long time to happen, and
the initiator of subsequents I/O operations that would add more timers
has to wait for them to complete.
With this change, _unrefActive does not show as a significant
contributor in CPU profiling reports anymore.
Fixes: https://github.com/joyent/node/issues/8160
Signed-off-by: Timothy J Fontaine <tjfontaine@gmail.com>
Conflicts:
lib/timers.js
Fixes: https://github.com/nodejs/node-convergence-archive/issues/23
Ref: https://github.com/nodejs/node/issues/268
PR-URL: https://github.com/nodejs/node/pull/2540
Reviewed-By: bnoordhuis - Ben Noordhuis <info@bnoordhuis.nl>
ParseArgs sets the first element of v8_args to argv[0], so v8_argc will
always be at least 1. This change only calls
V8::SetFlagsFromCommandLine if v8_argc > 1, leading to an additional
startup improvement of ~5%.
PR-URL: https://github.com/nodejs/node/pull/2646
Reviewed-By: trevnorris - Trevor Norris <trev.norris@gmail.com>
Both pointer arguments to memcmp are defined as non-null
and compiler optimizes upon that.
PR-URL: https://github.com/nodejs/node/pull/2544
Reviewed-By: trevnorris - Trevor Norris <trev.norris@gmail.com>
Reviewed-By: thefourtheye - Sakthipriyan Vairamani <thechargingvolcano@gmail.com>
String concatenation in the assert messages has drastic impact on test
runtime. Removal of these messages is unlikely to affect debugging if
any breaking changes are made.
Previous time to run:
$ time ./iojs test/parallel/test-stringbytes-external.js
real 0m2.321s
user 0m2.256s
sys 0m0.092s
With fix:
$ time ./iojs test/parallel/test-stringbytes-external.js
real 0m0.518s
user 0m0.508s
sys 0m0.008s
PR-URL: https://github.com/nodejs/node/pull/2544
Reviewed-By: trevnorris - Trevor Norris <trev.norris@gmail.com>
Reviewed-By: thefourtheye - Sakthipriyan Vairamani <thechargingvolcano@gmail.com>
new server has "nodejs" and "iojs" directories, upload to the
appropriate one
PR-URL: https://github.com/nodejs/node/pull/2623
Reviewed-By: jbergstroem - Johan Bergström <bugs@bergstroem.nu>
now need to specify "nodejs" or "iojs", also remove .gpg file
PR-URL: https://github.com/nodejs/node/pull/2623
Reviewed-By: jbergstroem - Johan Bergström <bugs@bergstroem.nu>
The test was failing after adding 'use strict' because the windows CI
uses the autocrlf option of git which converts \r into \r\n on checkout.
Refactored the test to not read itself anymore and create a temp file on
the fly instead to avoid this line-ending issue.
PR-URL: https://github.com/nodejs/node/pull/2494
Reviewed-By: Joao Reis <reis@janeasystems.com>
This is a port of 14db629497.
Original commit message:
Since install is per machine only, installation path should be stored
in local machine instead of current user. The registry stores HKLM in
different places for 32 and 64 bit applications, so the installer
will not suggest the old path when upgrading from 32 to 64 bit
version.
Fixesjoyent/node#5592Fixesjoyent/node#25087
PR-URL: https://github.com/joyent/node/pull/25640
Reviewed-By: Alexis Campailla <alexis@janeasystems.com>
Reviewed-By: Bert Belder <bertbelder@gmail.com>
PR-URL: https://github.com/nodejs/node/pull/2565
Reviewed-By: Alexis Campailla <alexis@janeasystems.com>
This is an adaptation of 8e80528453.
Original commit message:
The MSI install scope was set to the WiX default, which is per-user.
However, with UAC, it could not be installed by a standard user
because InstallPrivileges is elevated by default, hence the install
scope should be set to per-machine. Furthermore, the default install
path is a per-machine location and setting the system path requires
administrator privileges.
By changing the InstallScope to perMachine, Start Menu shortcuts are
placed in ProgramData and not the installing user's AppData folder,
making the shortcuts available to other users. This also fixes the
installation when AppData is a network folder.
The custom action is necessary to allow upgrades. Since a per-machine
MSI cannot upgrade an application installed per-user, the custom
action checks if there is going to be an upgrade to a previous
version installed per-user and sets the installation as per-user to
allow upgrading. Hence, the advantages of installing per-machine will
only apply in fresh installations.
Fixesjoyent/node#5849Fixesjoyent/node#7629
PR-URL: https://github.com/joyent/node/pull/25640
Reviewed-By: Alexis Campailla <alexis@janeasystems.com>
Reviewed-By: Bert Belder <bertbelder@gmail.com>
The original commit was adapted to search all upgrade codes listed in
the upgrade table, as the current installer tries to upgrade from two
different upgrade codes.
PR-URL: https://github.com/nodejs/node/pull/2565
Reviewed-By: Alexis Campailla <alexis@janeasystems.com>
This test was using fixturesDir to create temp files to test. This
patch replaces that with tmpDir and uses `assert` module to test.
Also, this test has been moved to `parallel`, from `sequential` mode.
PR-URL: https://github.com/nodejs/node/pull/2583
Reviewed-By: Ben Noordhuis <info@bnoordhuis.nl>
Reviewed-By: Rich Trott <rtrott@gmail.com>
This reverts commit 6cd0e2664b.
This reverts commit 7a999a1376.
This reverts commit f337595441.
It turns out that on Windows, uv_pipe_getsockname() is a no-op for
client sockets. It slipped through testing because of a CI snafu.
PR-URL: https://github.com/nodejs/node/pull/2584
Reviewed-By: Sakthipriyan Vairamani <thechargingvolcano@gmail.com>
Commit f337595 ("lib,src: add unix socket getsockname/getpeername")
accidentally left in a reference to io.js v1.5.0 in the documentation.
Update it to say "Node.js v4.0.0", the upcoming release.
PR-URL: https://github.com/nodejs/node/pull/2580
Reviewed-By: Jeremiah Senkpiel <fishrock123@rocketmail.com>
Complements the existing net.Socket#remoteFamily property.
PR-URL: https://github.com/nodejs/node/pull/956
Reviewed-By: Sam Roberts <vieuxtech@gmail.com>
Reviewed-By: Trevor Norris <trev.norris@gmail.com>
The implementation is a minor API change in that socket.address() now
returns a `{ address: '/path/to/socket' }` object, like it does for TCP
and UDP sockets. Before this commit, it returned `socket._pipeName`,
which is a string when present.
Change common.PIPE on Windows from '\\\\.\\pipe\\libuv-test' to
'\\\\?\\pipe\\libuv-test'. Windows converts the '.' to a '?' when
creating a named pipe, meaning that common.PIPE didn't match the
result from NtQueryInformationFile().
Fixes: https://github.com/nodejs/node/issues/954
PR-URL: https://github.com/nodejs/node/pull/956
Reviewed-By: Sam Roberts <vieuxtech@gmail.com>
Reviewed-By: Trevor Norris <trev.norris@gmail.com>
Extract the common logic into a template function. No functional
changes, just code cleanup.
PR-URL: https://github.com/nodejs/node/pull/956
Reviewed-By: Sam Roberts <vieuxtech@gmail.com>
Reviewed-By: Trevor Norris <trev.norris@gmail.com>
Refer: https://github.com/nodejs/io.js/issues/1543
When this test fails, it leaves dead processes in the system. This
patch makes sure that the child processes exit first, in case of
errors.
PR-URL: https://github.com/nodejs/node/pull/2206
Reviewed-By: Johan Bergström <bugs@bergstroem.nu>
Reviewed-By: Ben Noordhuis <info@bnoordhuis.nl>
This poplulates the lists of flaky tests with tests that failed recently
in Jenkins.
PR-URL: https://github.com/nodejs/node/pull/2424
Reviewed-By: Ben Noordhuis <info@bnoordhuis.nl>
Reviewed-By: João Reis <reis@janeasystems.com>
Reviewed-By: Sakthipriyan Vairamani <thechargingvolcano@gmail.com>
PR-URL: https://github.com/nodejs/node/pull/2424
Reviewed-By: Ben Noordhuis <info@bnoordhuis.nl>
Reviewed-By: João Reis <reis@janeasystems.com>
Reviewed-By: Sakthipriyan Vairamani <thechargingvolcano@gmail.com>
Adding support for specifying flaky test mode to
the test runner:
- via an environment variable FLAKY_TESTS for Makefile
- via an argument ignore-flaky for vcbuild.bat
Ported from 2d2494cf14
PR-URL: https://github.com/nodejs/node/pull/2424
Reviewed-By: Ben Noordhuis <info@bnoordhuis.nl>
Reviewed-By: João Reis <reis@janeasystems.com>
Reviewed-By: Sakthipriyan Vairamani <thechargingvolcano@gmail.com>
Template configuration files for marking tests as flaky.
PR-URL: https://github.com/nodejs/node/pull/2424
Reviewed-By: Ben Noordhuis <info@bnoordhuis.nl>
Reviewed-By: João Reis <reis@janeasystems.com>
Reviewed-By: Sakthipriyan Vairamani <thechargingvolcano@gmail.com>
Make the test runner return a 0 exit code when only
flaky tests fail and --flaky-tests=dontcare is specified.
Ported from a9b642cf5b
PR-URL: https://github.com/nodejs/node/pull/2424
Reviewed-By: Ben Noordhuis <info@bnoordhuis.nl>
Reviewed-By: João Reis <reis@janeasystems.com>
Reviewed-By: Sakthipriyan Vairamani <thechargingvolcano@gmail.com>
Adding --flaky-tests option, to allow regarding flaky tests failures
as non-fatal.
Currently only observed by the TapProgressIndicator, which will
add a # TODO directive to tests classified as flaky. According to the
TAP specification, the test harness is supposed to treat failures
that have a # TODO directive as non-fatal.
Ported from df3a2b2cf2
PR-URL: https://github.com/nodejs/node/pull/2424
Reviewed-By: Ben Noordhuis <info@bnoordhuis.nl>
Reviewed-By: João Reis <reis@janeasystems.com>
Reviewed-By: Sakthipriyan Vairamani <thechargingvolcano@gmail.com>
There is no type-conversion to be done. Therefore, use the === operator.
PR-URL: https://github.com/nodejs/node/pull/2558
Reviewed-By: Rich Trott <rtrott@gmail.com>
There might be a need to "kick off" the next tick queue and execute
events on it. Normally it is done through the `MakeCallback` interface,
but in case when it is not - we need a way to "kick them off" manually.
PR-URL: https://github.com/nodejs/node/pull/2355
Reviewed-By: Trevor Norris <trev.norris@gmail.com>
included:
* commands to run to import all active keys
* list of previous Node.js releasers key details
PR-URL: https://github.com/nodejs/node/pull/2455
Reviewed-By: Colin Ihrig <cjihrig@gmail.com>
Per feedback from @thefourtheye: (a) improve word on the
description of the `close` event in stream.markdown and
(b) remove `[module section]: modules.html` from
globals.markdown
Reviewed By: Sakthipriyan Vairamani <thechargingvolcano@gmail.com>
PR-URL: https://github.com/nodejs/node/pull/2378
Jeremiah Senkpiel
Julien Gilli
Evan Lucas
Rod Vagg
Saúl Ibarra Corretgé
Karl Skomski
Trevor Norris
Rich Trott
P.S.V.R
Roman Reiss
Bradley Meck
João Reis
Michaël Zasso
Sakthipriyan Vairamani
Ben Noordhuis
Сковорода Никита Андреевич
Alexis Campailla
Minwoo Jung
Fedor Indutny
James M Snell