Move lib/internal/assert.js to lib/internal/assert/assertion_error.js.
This is in preparation for making lib/internal/assert.js a tiny module
for use in Node.js built-ins so that we can use `assert()` without
having to load the entire ~1200 line `assert` module.
PR-URL: https://github.com/nodejs/node/pull/25956
Reviewed-By: Ruben Bridgewater <ruben@bridgewater.de>
Reviewed-By: Minwoo Jung <minwoo@nodesource.com>
`Watchdog::Async` features only once for the
async callback, so make it a lambda.
PR-URL: https://github.com/nodejs/node/pull/25945
Reviewed-By: James M Snell <jasnell@gmail.com>
Reviewed-By: Anna Henningsen <anna@addaleax.net>
Reviewed-By: Colin Ihrig <cjihrig@gmail.com>
Execute JS code in worker through same vm context
while exiting from the main thread at arbitrary
execution points, and make sure that the workers
quiesce without crashing.
`worker_threads` are not necessarily the subject of
testing, those are used for easy simulation of
multi-thread scenarios.
Refs: https://github.com/nodejs/node/issues/25007
PR-URL: https://github.com/nodejs/node/pull/25085
Reviewed-By: Anna Henningsen <anna@addaleax.net>
Instead of using
`internalBinding('config').shouldAbortOnUncaughtException`.
Also removes that property from the binding.
PR-URL: https://github.com/nodejs/node/pull/25862
Reviewed-By: Richard Lau <riclau@uk.ibm.com>
Reviewed-By: Luigi Pinca <luigipinca@gmail.com>
Reviewed-By: Gus Caplan <me@gus.host>
Reviewed-By: Anna Henningsen <anna@addaleax.net>
Reviewed-By: Colin Ihrig <cjihrig@gmail.com>
Since this is only necessary when user code execution is expected.
PR-URL: https://github.com/nodejs/node/pull/25825
Reviewed-By: Anna Henningsen <anna@addaleax.net>
Reviewed-By: Daniel Bevenius <daniel.bevenius@gmail.com>
Reviewed-By: James M Snell <jasnell@gmail.com>
Moves the `process.on()` and `promise.emit()` calls happened during
bootstrap for signal events into `bootstrap/node.js` so it's easier
to tell the side effects.
Drive-by changes:
- Moves the signal event re-arming to a point later during
the bootstrap - as early as it were it's unlikely that there
could be any existing signal events to re-arm for node-report.
- Use a Map instead of an Object for signal wraps since it is used
as a deletable dictionary anyway.
PR-URL: https://github.com/nodejs/node/pull/25859
Reviewed-By: Anna Henningsen <anna@addaleax.net>
Signed-off-by: gengjiawen <technicalcute@gmail.com>
PR-URL: https://github.com/nodejs/node/pull/25923
Reviewed-By: Anna Henningsen <anna@addaleax.net>
Reviewed-By: James M Snell <jasnell@gmail.com>
Reviewed-By: Richard Lau <riclau@uk.ibm.com>
There is no need to establish a TCP connection. It is sufficient to
test that the listener that forwards the `'timeout'` event from the
socket to the `ClientRequest` instance is added to the socket.
PR-URL: https://github.com/nodejs/node/pull/25886
Reviewed-By: Gireesh Punathil <gpunathi@in.ibm.com>
Reviewed-By: Anna Henningsen <anna@addaleax.net>
Signed-off-by: gengjiawen <technicalcute@gmail.com>
PR-URL: https://github.com/nodejs/node/pull/25914
Reviewed-By: Anna Henningsen <anna@addaleax.net>
Reviewed-By: James M Snell <jasnell@gmail.com>
All benchmark code uses trailing commas on multi-line arrays and arrow
functions for anonymous callbacks. Apply lint rules to that effect.
PR-URL: https://github.com/nodejs/node/pull/25944
Reviewed-By: Anto Aravinth <anto.aravinth.cse@gmail.com>
Reviewed-By: James M Snell <jasnell@gmail.com>
Code in benchmark directory sometimes uses `function () {}` for
anonymous callbacks and sometimes uses `() => {}`. Multi-line arrays
sometimes have a trailing comma and sometimes do not. Update to always
use arrow functions for anonymous callbacks and trailing commas for
multiline arrays.
PR-URL: https://github.com/nodejs/node/pull/25944
Reviewed-By: Anto Aravinth <anto.aravinth.cse@gmail.com>
Reviewed-By: James M Snell <jasnell@gmail.com>
Currently the following compiler warnings are generated on Linux:
../src/node_buffer.cc:
In function 'void node::Buffer::{anonymous}::StringSlice(
const v8::FunctionCallbackInfo<v8::Value>&)
[with node::encoding encoding = (node::encoding)1]':
../src/node_buffer.cc:54:20: warning:
'start' may be used uninitialized in this function
[-Wmaybe-uninitialized]
if (end < start) end = start;
^~~
../src/node_buffer.cc:50:10: note: 'start' was declared here
size_t start;
^~~~~
This commit initializes start and end to zero to avoid these warnings.
PR-URL: https://github.com/nodejs/node/pull/25665
Reviewed-By: Jeremiah Senkpiel <fishrock123@rocketmail.com>
Reviewed-By: Richard Lau <riclau@uk.ibm.com>
Reviewed-By: Michael Dawson <michael_dawson@ca.ibm.com>
Reviewed-By: James M Snell <jasnell@gmail.com>
When adding to the REPL API recently, I was unsure what to put as a
version number for the new method I added. This change adds a reference
to `releases.md` where the use of `REPLACEME` is discussed.
PR-URL: https://github.com/nodejs/node/pull/25961
Reviewed-By: Anna Henningsen <anna@addaleax.net>
Reviewed-By: Ruben Bridgewater <ruben@bridgewater.de>
Reviewed-By: James M Snell <jasnell@gmail.com>
Reviewed-By: Rich Trott <rtrott@gmail.com>
Reviewed-By: Richard Lau <riclau@uk.ibm.com>
Revise the "Reverting commits" section of the Collaborator Guide. Reduce
passive voice. Increase clarity.
PR-URL: https://github.com/nodejs/node/pull/25942
Reviewed-By: James M Snell <jasnell@gmail.com>
Reviewed-By: Luigi Pinca <luigipinca@gmail.com>
Reviewed-By: Vse Mozhet Byt <vsemozhetbyt@gmail.com>
Fix perplexing comment. It's not that TLS "clients" don't support
'secureConnect', it's that client sockets created with `new TLSSocket`
(as opposed to `tls.connect()`) don't support that event.
PR-URL: https://github.com/nodejs/node/pull/25508
Reviewed-By: Anna Henningsen <anna@addaleax.net>
Reviewed-By: James M Snell <jasnell@gmail.com>
Instead of pushing state into global arrays and checking the results
before exit, use common.mustCall() and make the checks immediately.
PR-URL: https://github.com/nodejs/node/pull/25508
Reviewed-By: Anna Henningsen <anna@addaleax.net>
Reviewed-By: James M Snell <jasnell@gmail.com>
Connection is known to be completely setup only after data has
exchanged, so wait unil data echo before sending a bad record.
Otherwise, the bad record could interrupt completion of the server's
handshake, and whether the error is emitted on the connection or server
is a matter of timing.
Also, assert that server errors do not occur. 'error' would crash node
with and unhandled event, but 'tlsClientError' is ignored by default.
PR-URL: https://github.com/nodejs/node/pull/25508
Reviewed-By: Anna Henningsen <anna@addaleax.net>
Reviewed-By: James M Snell <jasnell@gmail.com>
The timing of destroy between client/server is undefined, but end is
guaranteed to occur.
PR-URL: https://github.com/nodejs/node/pull/25508
Reviewed-By: Anna Henningsen <anna@addaleax.net>
Reviewed-By: James M Snell <jasnell@gmail.com>
This test has a dependency on the order in which the TCP connection is
made, and TLS server handshake completes. It assumes those server side
events occur before the client side write callback, which is not
guaranteed by the TLS API. It usually passes with TLS1.3, but TLS1.3
didn't exist at the time the bug existed.
Pin the test to TLS1.2, since the test shouldn't be changed in a way that
doesn't trigger a segfault in 7.7.3:
- https://github.com/nodejs/node/issues/13184#issuecomment-303700377
PR-URL: https://github.com/nodejs/node/pull/25508
Reviewed-By: Anna Henningsen <anna@addaleax.net>
Reviewed-By: James M Snell <jasnell@gmail.com>
Existing code assumed that the server completed the handshake before the
client rejected the certificate, and destroyed the socket. This
assumption is fragile, remove it, and instead check explicitly that data
can or cannot be exchanged via TLS, whichever is expected.
PR-URL: https://github.com/nodejs/node/pull/25508
Reviewed-By: Anna Henningsen <anna@addaleax.net>
Reviewed-By: James M Snell <jasnell@gmail.com>
Do not assume that server handshake event happens before client, it is
not guaranteed.
PR-URL: https://github.com/nodejs/node/pull/25508
Reviewed-By: Anna Henningsen <anna@addaleax.net>
Reviewed-By: James M Snell <jasnell@gmail.com>
Test assumed that server got the the connection before the client
destroys it, but that is not guaranteed. Also, the test was closing the
TCP connection 3 times, effectively:
1. on the server side, right after TLS connection occurs (if it does)
2. on the client side, internal to tls, when the cert is rejected
3. again on the client side, in the error event which is emitted by
the internal tls destroy from 2
This is too often, and the dependency on 1 occurring is fragile.
PR-URL: https://github.com/nodejs/node/pull/25508
Reviewed-By: Anna Henningsen <anna@addaleax.net>
Reviewed-By: James M Snell <jasnell@gmail.com>
Test assumed server gets a handshake before the client destroyed it, and
didn't assert that dns.lookup() callback occurred.
PR-URL: https://github.com/nodejs/node/pull/25508
Reviewed-By: Anna Henningsen <anna@addaleax.net>
Reviewed-By: James M Snell <jasnell@gmail.com>
Test for deep strict equality when prototype and toStringTag
have been modified in surprising ways.
PR-URL: https://github.com/nodejs/node/pull/25932
Reviewed-By: Ruben Bridgewater <ruben@bridgewater.de>
Reviewed-By: Anto Aravinth <anto.aravinth.cse@gmail.com>
Reviewed-By: Colin Ihrig <cjihrig@gmail.com>
Reviewed-By: James M Snell <jasnell@gmail.com>
Add some BigInt tests for test-util-isDeepStrictEqual to get 100%
coverage for lib/internal/util/comparisons.js.
PR-URL: https://github.com/nodejs/node/pull/25932
Reviewed-By: Ruben Bridgewater <ruben@bridgewater.de>
Reviewed-By: Anto Aravinth <anto.aravinth.cse@gmail.com>
Reviewed-By: Colin Ihrig <cjihrig@gmail.com>
Reviewed-By: James M Snell <jasnell@gmail.com>
Previously a GC prologue callback and a GC epilogue callback
are always unconditionally enabled during bootstrap when
the `performance` binding is loaded, even when the user does
not use the performance timeline API to enable GC tracking.
This patch makes the callback addition conditional and only
enables them when the user explicitly requests
`observer.observe(['gc'])` to avoid the overhead.
PR-URL: https://github.com/nodejs/node/pull/25853
Reviewed-By: Anna Henningsen <anna@addaleax.net>
Reviewed-By: Gus Caplan <me@gus.host>
Reviewed-By: James M Snell <jasnell@gmail.com>
Allow undefined as a callback, but do not allow null.
PR-URL: https://github.com/nodejs/node/pull/25929
Reviewed-By: Colin Ihrig <cjihrig@gmail.com>
Reviewed-By: Anatoli Papirovski <apapirovski@mac.com>
Reviewed-By: Ruben Bridgewater <ruben@bridgewater.de>
Reviewed-By: Luigi Pinca <luigipinca@gmail.com>
Reviewed-By: Minwoo Jung <minwoo@nodesource.com>
Reviewed-By: James M Snell <jasnell@gmail.com>
Remove setting of a 'close' event handler on MessagePort through the use
of an `.onclose` property. We don't use this convention anywhere else in
our codebase for 'close' events, this feature is undocumented, and we
don't test it.
PR-URL: https://github.com/nodejs/node/pull/25904
Reviewed-By: Benjamin Gruenbaum <benjamingr@gmail.com>
Reviewed-By: Yuta Hiroto <hello@hiroppy.me>
Reviewed-By: Ujjwal Sharma <usharma1998@gmail.com>
* Confusing (but correct) regex using `A-z` character range by
accident
* Add the status argument to afterShutdown
PR-URL: https://github.com/nodejs/node/pull/25873
Reviewed-By: Refael Ackermann <refack@gmail.com>
Reviewed-By: Anna Henningsen <anna@addaleax.net>
Reviewed-By: Sam Roberts <vieuxtech@gmail.com>
PR-URL: https://github.com/nodejs/node/pull/25906
Reviewed-By: Colin Ihrig <cjihrig@gmail.com>
Reviewed-By: Richard Lau <riclau@uk.ibm.com>
Reviewed-By: James M Snell <jasnell@gmail.com>
when t0 and t1 are spawned with t0's outputstream [1, 2] is piped into
t1's input, a new pipe is created which uses a copy of the t0's fd.
This leaves the original copy in Node parent, unattended. Net result is
that when t0 produces data, it gets bifurcated into both the copies
Detect the passed handle to be of 'wrap' type and close after the
native spawn invocation by which time piping would have been over.
Fixes: https://github.com/nodejs/node/issues/9413
Fixes: https://github.com/nodejs/node/issues/18016
PR-URL: https://github.com/nodejs/node/pull/21209
Reviewed-By: Matteo Collina <matteo.collina@gmail.com>
Reviewed-By: James M Snell <jasnell@gmail.com>
Reviewed-By: Anna Henningsen <anna@addaleax.net>
The removed line does not add anything of value to the test. It was
removed to simplify the test.
PR-URL: https://github.com/nodejs/node/pull/25731
Reviewed-By: Michaël Zasso <targos@protonmail.com>
Reviewed-By: Ben Noordhuis <info@bnoordhuis.nl>
It is checked if an command buffer exists or not. This code branch
can only be reached if none exist, so there's no need to clear that
buffer again.
PR-URL: https://github.com/nodejs/node/pull/25731
Reviewed-By: Michaël Zasso <targos@protonmail.com>
Reviewed-By: Ben Noordhuis <info@bnoordhuis.nl>
In case no error has occurred during the evaluation of some code,
`undefined` has been returned in some cases as error argument instead
of `null`. This is fixed by this patch.
PR-URL: https://github.com/nodejs/node/pull/25731
Reviewed-By: Michaël Zasso <targos@protonmail.com>
Reviewed-By: Ben Noordhuis <info@bnoordhuis.nl>
The completion lists used a hand crafted list of global entries that
was redundant due to also using the actual global properties for tab
completion. Those entries ended up in an separated completion group
which did not seem useful.
PR-URL: https://github.com/nodejs/node/pull/25731
Reviewed-By: Michaël Zasso <targos@protonmail.com>
Reviewed-By: Ben Noordhuis <info@bnoordhuis.nl>
Signed-off-by: gengjiawen <technicalcute@gmail.com>
PR-URL: https://github.com/nodejs/node/pull/25849
Reviewed-By: Anna Henningsen <anna@addaleax.net>
Reviewed-By: Colin Ihrig <cjihrig@gmail.com>
Reviewed-By: James M Snell <jasnell@gmail.com>
Signed-off-by: gengjiawen <technicalcute@gmail.com>
PR-URL: https://github.com/nodejs/node/pull/25850
Reviewed-By: Luigi Pinca <luigipinca@gmail.com>
Reviewed-By: James M Snell <jasnell@gmail.com>
Because the part that is shared by `process.reallyExit` and the
Node.js teardown is `WaitForInspectorDisconnect()`, move that
into node_internals.h instead, and move the C++ binding code
into `node_process_methods.cc` since that's the only place
it's needed.
PR-URL: https://github.com/nodejs/node/pull/25860
Reviewed-By: Anna Henningsen <anna@addaleax.net>
Reviewed-By: Minwoo Jung <minwoo@nodesource.com>
Reviewed-By: Jeremiah Senkpiel <fishrock123@rocketmail.com>
Signed-off-by: gengjiawen <technicalcute@gmail.com>
PR-URL: https://github.com/nodejs/node/pull/25889
Reviewed-By: Anna Henningsen <anna@addaleax.net>
Reviewed-By: Daniel Bevenius <daniel.bevenius@gmail.com>
Reviewed-By: Colin Ihrig <cjihrig@gmail.com>
Reviewed-By: James M Snell <jasnell@gmail.com>
Some process methods are not supported in workers. This commit
adds stubs that throw more informative errors.
PR-URL: https://github.com/nodejs/node/pull/25587
Fixes: https://github.com/nodejs/node/issues/25448
Reviewed-By: Anna Henningsen <anna@addaleax.net>
Reviewed-By: Refael Ackermann <refack@gmail.com>
Reviewed-By: Franziska Hinkelmann <franziska.hinkelmann@gmail.com>
Reviewed-By: James M Snell <jasnell@gmail.com>
