Reduce the number of stat() system calls that require() makes by caching
the results more aggressively.
To avoid unbounded growth without implementing a LRU cache, scope the
cache to the lifetime of the first call to require(). Recursive calls
(i.e. require() calls in the included code) transparently profit from
the cache.
The benchmarked application is the loopback-sample-app[0] and it sees
the number of stat calls at start-up go down by 40%, from 4736 to 2810.
[0] https://github.com/strongloop/loopback-sample-app
PR-URL: https://github.com/nodejs/node/pull/4575
Reviewed-By: Colin Ihrig <cjihrig@gmail.com>
Reviewed-By: James M Snell <jasnell@gmail.com>
It's not guaranteed that the socket data is received in the same chunk
as the upgrade response. Listen for the `data` event to make sure all
the data is received.
PR-URL: https://github.com/nodejs/node/pull/4602
Reviewed-By: Brian White <mscdex@mscdex.net>
Reviewed-By: James M Snell <jasnell@gmail.com>
It's not guaranteed that the socket data is received in the same chunk
as the upgrade response. Listen for the `data` event to make sure all
the data is received.
PR-URL: https://github.com/nodejs/node/pull/4520
Reviewed-By: Rich Trott <rtrott@gmail.com>
Reviewed-By: James M Snell <jasnell@gmail.com>
`fork` is imported twice in a row. Remove duplication.
PR-URL: https://github.com/nodejs/node/pull/4634
Reviewed-By: Johan Bergström <bugs@bergstroem.nu>
Reviewed-By: Colin Ihrig <cjihrig@gmail.com>
Reviewed-By: Brian White <mscdex@mscdex.net>
Reviewed-By: Michaël Zasso <mic.besace@gmail.com>
Add support to fs.createWriteStream and fs.createWriteStream for an autoClose
option that behaves similarly to the autoClose option supported by
fs.createReadStream and fs.ReadStream.
When an instance of fs.createWriteStream created with autoClose === false finishes,
it is not destroyed. Its underlying fd is not closed and it is the
responsibility of the user to close it.
PR-URL: https://github.com/nodejs/node/pull/3679
Reviewed-By: James M Snell <jasnell@gmail.com>
Reviewed-By: Colin Ihrig <cjihrig@gmail.com>
Two tests were requiring the common module twice. This removes the
duplicate require statement in the tests.
PR-URL: https://github.com/nodejs/node/pull/4611
Reviewed-By: Brian White <mscdex@mscdex.net>
Reviewed-By: Colin Ihrig <cjihrig@gmail.com>
Reviewed-By: Johan Bergström <bugs@bergstroem.nu>
Reviewed-By: James M Snell <jasnell@gmail.com>
Previously, Error objects were formatted as the result of a `toString`
call bounded by square brackets. They are now formatted as the stack
trace for the given error object. The intention initially was to emulate
how browsers do `console.error` but since that would also impact
`console.warn`, `console.log`, etc, it was decided to make the change at
`util.inspect` level which is in turn used by the `console` package.
Fixes: nodejs#4452
PR-URL: https://github.com/nodejs/node/pull/4582
Reviewed-By: Ben Noordhuis <info@bnoordhuis.nl>
Reviewed-By: Trevor Norris <trev.norris@gmail.com>
Reviewed-By: Colin Ihrig <cjihrig@gmail.com>
Reviewed-By: James M Snell <jasnell@gmail.com>
Remove unnecessary `setImmediate()` that causes a minor race condition.
Stop the test after 3 occurrences rather than 5 to allow for slower
hosts running the test in parallel with other tests.
Fixes: https://github.com/nodejs/node/issues/4559
PR-URL: https://github.com/nodejs/node/pull/4599
Reviewed-By: Johan Bergström <bugs@bergstroem.nu>
Make default `clientError` behavior (close socket immediately)
overridable. With this APIs it is possible to write a custom error
handler, and to send, for example, a 400 HTTP response.
http.createServer(...).on('clientError', function(err, socket) {
socket.end('HTTP/1.1 400 Bad Request\r\n\r\n');
socket.destroy();
});
Fix: #4543
PR-URL: https://github.com/nodejs/node/pull/4557
Reviewed-By: Brian White <mscdex@mscdex.net>
`clientError` will have `http.Server`-specific behavior, and we don't
want to shadow it in `tls.Server`.
PR-URL: https://github.com/nodejs/node/pull/4557
Reviewed-By: Brian White <mscdex@mscdex.net>
Remove unused vars in tests
PR-URL: https://github.com/nodejs/node/pull/4536
Reviewed-By: Rod Vagg <rod@vagg.org>
Reviewed-By: Michaël Zasso <mic.besace@gmail.com>
Reviewed-By: Colin Ihrig <cjihrig@gmail.com>
1c85849973 "fixed"
test-domain-exit-dispose-again by changing its logic to test that
process.domain was cleared properly in case an error was thrown from a
timer's callback.
However, it became clear when reviewing a recent change that refactors
lib/timers.js that it was not quite the intention of the original test.
Thus, this change adds the original implementation of
test-domain-exit-dispose-again back, with comments that make its
implementation easier to understand.
It also preserve the changes made by
1c85849973, but it moves them to a new
test file named test-timers-reset-process-domain-on-throw.js.
PR: #4256
PR-URL: https://github.com/nodejs/node/pull/4256
Reviewed-By: Jeremiah Senkpiel <fishrock123@rocketmail.com>
Remove all remaining unused variables from tests in test/parallel.
PR-URL: https://github.com/nodejs/node/pull/4511
Reviewed-By: James M Snell<jasnell@gmail.com>
Reviewed-By: Johan Bergström <bugs@bergstroem.nu>
This change adds a new event handler to the `error` event of the socket
after it has been used by the http_client.
The purpose of this change is to catch errors on *keep alived*
connections from idle sockets that otherwise will cause an uncaugh error
event on the application.
Fix: #3595
PR-URL: https://github.com/nodejs/node/pull/4482
Reviewed-By: Fedor Indutny <fedor@indutny.com>
Use common.platformTimeout() to fix flaky
test-stream2-readable-empty-buffer-no-eofi on Raspberry Pis.
Fixes: https://github.com/nodejs/node/issues/4493
PR-URL: https://github.com/nodejs/node/pull/4516
Reviewed-By: Johan Bergström <bugs@bergstroem.nu>
Reviewed-By: Evan Lucas <evanlucas@me.com>
Reviewed-By: James M Snell<jasnell@gmail.com>
Three tests designated as flaky on Linux have not failed on the
continuous integration server in a long time. Removing flaky designation
for these tests.
Fixes: https://github.com/nodejs/node/issues/4446
PR-URL: https://github.com/nodejs/node/pull/4519
Reviewed-By: Johan Bergström <bugs@bergstroem.nu>
Reviewed-By: James M Snell <jasnell@gmail.com>
Reviewed-By: Brian White <mscdex@mscdex.net>
`setTimeout()`, `setInterval()` and `setIntermediate` currently
throw errors when receiving non-function objects as their first
argument, but only do so when trying to execute the callback,
i.e. after the waited time has passed. This may complicate
debugging when a lot of calls to `setTimeout()`/etc. are involved,
so failing as early as possible seems like a good idea.
`setTimeout()` historically ignored an falsy first
argument, while the other functions do not and throw instead.
This patch changes this behaviour to make all three match and
adds remarks in the corresponding documentation.
PR-URL: https://github.com/nodejs/node/pull/4362
Reviewed-By: Jeremiah Senkpiel <fishrock123@rocketmail.com>
Reviewed-By: Colin Ihrig <cjihrig@gmail.com>
Reviewed-By: Brian White <mscdex@mscdex.net>
Reviewed-By: Trevor Norris <trev.norris@gmail.com>
Reviewed-By: Rod Vagg <rod@vagg.org>
test-child-process-fork-net2.js checks that things happen within
certain time constraints, thus doubling as a benchmark test in addition
to a functionality test.
This change removes the time check, as it was causing the test to fail
on SmartOS and Windows (and possibly elsewhere) when the tests were
run in parallel on CI. There is no guarantee that other tests won't
consume enough resources to slow this test down, so don't check the time
constraints (beyond the generous timeout that the test is given by
test.py in the first place, of course).
If we want to do benchmark/performance tests, we should keep them
separate from pure functionality tests. The time check may have been a
remnant of the distant past when Node.js was much slower. It predates
io.js
Ref: https://github.com/nodejs/node/pull/4476
PR-URL: https://github.com/nodejs/node/pull/4494
Reviewed-By: Colin Ihrig <cjihrig@gmail.com>
Reviewed-By: Johan Bergström <bugs@bergstroem.nu>
Refactor test to remove unnecessary booleans and one unnecesary timer.
Instead, throw Error objects where appropriate and rely on
common.mustCall().
The timer seemed to be the source of an issue when parallelizing tests.
Ref: https://github.com/nodejs/node/pull/4476#issuecomment-168080875
PR-URL: https://github.com/nodejs/node/pull/4490
Reviewed-By: Johan Bergström <bugs@bergstroem.nu>
Reviewed-By: Colin Ihrig <cjihrig@gmail.com>
test-fs-realpath.js was writing files to the fixture dir. This changes
it to use the temp directory instead. This also replaces some of the
string concatenation for paths with uses of path.join() and
path.relative().
PR-URL: https://github.com/nodejs/node/pull/4489
Reviewed-By: Johan Bergström <bugs@bergstroem.nu>
Many tests use require() to import modules that subsequently never gets
used. This removes those imports and, in a few cases, removes other
unused variables from tests.
PR-URL: https://github.com/nodejs/node/pull/4475
Reviewed-By: Ben Noordhuis <info@bnoordhuis.nl>
Reviewed-By: Johan Bergström <bugs@bergstroem.nu>
Add a comment to clarify how the tests work and their purpose.
Also removes unnecessary assignment of domain module to a variable.
PR-URL: https://github.com/nodejs/node/pull/4474
Reviewed-By: Colin Ihrig <cjihrig@gmail.com>
Reviewed-By: Julien Gilli <jgilli@fastmail.fm>
Improves the message when an assertion fires in the
test-net-pipe-connect-errors so that it indicates the incorrect value
received rather than merely reporting that the value is incorrect.
PR-URL: https://github.com/nodejs/node/pull/4461
Reviewed-By: Johan Bergström <bugs@bergstroem.nu>
Reviewed-By: Colin Ihrig <cjihrig@gmail.com>
Reviewed-By: Myles Borins <myles.borins@gmail.com>
This fixes CI failures for test-net-pipe-connect-errors on Raspberry Pi
devices.
PR-URL: https://github.com/nodejs/node/pull/4478
Reviewed-By: Johan Bergström <bugs@bergstroem.nu>
Reviewed-By: James M Snell <jasnell@gmail.com>
Many test modules load assert but do not use it. This change removes
those instances.
It also removes a handful of other unused variables when they were
nearby.
PR-URL: https://github.com/nodejs/node/pull/4438
Reviewed-By: Michaël Zasso <mic.besace@gmail.com>
Reviewed-By: Brian White <mscdex@mscdex.net>
Reviewed-By: Colin Ihrig <cjihrig@gmail.com>
Since headers are stored in an empty literal object ({}) instead
of an object created with Object.create(null), care must be taken
with property names inherited from Object. Currently there are
only functions inherited, so we can safely check for existing
strings instead.
Fixes: https://github.com/nodejs/node/issues/4456
PR-URL: https://github.com/nodejs/node/pull/4460
Reviewed-By: Fedor Indutny <fedor.indutny@gmail.com>
Reviewed-By: Rich Trott <rtrott@gmail.com>
Reviewed-By: Colin Ihrig <cjihrig@gmail.com>
Reviewed-By: Myles Borins <myles.borins@gmail.com>
Reviewed-By: Minwoo Jung <jmwsoft@gmail.com>
With the introduction of temporary paths in the test runner
realpath tests would bail in scenarios where the temporary folder
wasn't in the same directory as the source code.
PR-URL: https://github.com/nodejs/node/pull/4477
Reviewed-By: Rich Trott <rtrott@gmail.com>
A few tests assumed that temp dirs always lived in the same
parent folder as fixtures. Make these use `common.tmpDir` instead.
PR-URL: https://github.com/nodejs/node/pull/3325
Reviewed-By: Joao Reis <reis@janeasystems.com>
Remove a handful of variables that are declared but never used in the
tests for the net module.
PR-URL: https://github.com/nodejs/node/pull/4430
Reviewed-By: Michaël Zasso <mic.besace@gmail.com>
Reviewed-By: Colin Ihrig <cjihrig@gmail.com>
In addition to removing unused vars, this also fixes an instance where
booleans were set presumably to check something but then never used.
This now confirms that the events that were setting the booleans are
fired.
PR-URL: https://github.com/nodejs/node/pull/4425
Reviewed-By: Johan Bergström <bugs@bergstroem.nu>
Reviewed-By: Colin Ihrig <cjihrig@gmail.com>
On single core Windows systems, process.send() would cause an
EPIPE because of the ordering of the IPC channel disconnect and
the process.send().
The test was originally only relevant for non-Windows platforms,
so this commit merely skips the test on Windows.
Fixes: https://github.com/nodejs/node/issues/4450
PR-URL: https://github.com/nodejs/node/pull/4457
Reviewed-By: Rich Trott <rtrott@gmail.com>
Before this commit, it was possible on Windows for the server's
'connection' handler to be called *after* the client socket's
'connect' handler. This caused the 'message' event to be missed
and the test would never end (timing out in CI). This problem
was more easily reproducible on a low resource (slow CPU)
Windows (2012r2) installation.
This commit waits until both handlers have been called before
sending the handle to the master process.
Fixes: https://github.com/nodejs/node/issues/3957
PR-URL: https://github.com/nodejs/node/pull/4444
Reviewed-By: Rich Trott <rtrott@gmail.com>
Windows would die with ECONNRESET most times when running
this particular test. This commit makes handling these errors
more tolerable.
PR-URL: https://github.com/nodejs/node/pull/4442
Reviewed-By: Fedor Indutny <fedor.indutny@gmail.com>
Reviewed-By: Rich Trott <rtrott@gmail.com>
Some of the TLS tests have variables that do not get used. This removes
those variables.
PR-URL: https://github.com/nodejs/node/pull/4424
Reviewed-By: Johan Bergström <bugs@bergstroem.nu>
The http tests seem especially prone to including unused variables.
This change removes them.
PR-URL: https://github.com/nodejs/node/pull/4422
Reviewed-By: Johan Bergström <bugs@bergstroem.nu>
It is possible that the internal hnadleMessage() might try to send to
a channel that has been closed. The result can be an AssertionError.
Guard against this.
Fixes: https://github.com/nodejs/node/issues/4205
PR-URL: https://github.com/nodejs/node/pull/4418
Reviewed-By: Brian White <mscdex@mscdex.net>
Reviewed-By: Johan Bergström <bugs@bergstroem.nu>
common.js needs to be loaded in all tests so that there is checking
for variable leaks and possibly other things. However, it does not
need to be assigned to a variable if nothing in common.js is referred
to elsewhere in the test.
PR-URL: https://github.com/nodejs/node/pull/4408
Reviewed-By: James M Snell <jasnell@gmail.com>
`aethrow` is defined as a function returned from makeBlock() but the
function is never used and the call to makeBlock() does not run any
tests.
PR-URL: https://github.com/nodejs/node/pull/4405
Reviewed-By: James M Snell <jasnell@gmail.com>
A number of tests in `test/parallel` were importing the `util` module
via `require()` but not using `util` for anything. This removes those
`require()` statements.
PR-URL: https://github.com/nodejs/node/pull/4397
Reviewed-By: Colin Ihrig <cjihrig@gmail.com>
Reviewed-By: Rod Vagg <rod@vagg.org>
Make sure the server is not closed until both responses have been
received.
PR-URL: https://github.com/nodejs/node/pull/3958
Reviewed-By: Rich Trott <rtrott@gmail.com>
Reviewed-By: Johan Bergström <bugs@bergstroem.nu>
Check the stderr output in the `close` event as it's not guaranteed to
be fully available when the `exit` event is fired.
PR: #4364
PR-URL: https://github.com/nodejs/node/pull/4364
Reviewed-By: Julien Gilli <jgilli@fastmail.fm>
Without these changes, the pi1-raspbian-wheezy CI node was timing
out on these tests.
PR-URL: https://github.com/nodejs/node/pull/4387
Reviewed-By: Ben Noordhuis <info@bnoordhuis.nl>
Reviewed-By: Colin Ihrig <cjihrig@gmail.com>
Rich Trott
Ben Noordhuis
Santiago Gimeno
Saquib
Mudit Ameta
Fedor Indutny
Feross Aboukhadijeh
Julien Gilli
José F. Romaniello
Anna Henningsen
Brian White
Johan Bergström
Max Barinov