Invoke threadsafe_function during the same tick and avoid marshalling
costs between threads and/or churning event loop if either:
1. There's a queued call already
2. `Push()` is called while the main thread was running
threadsafe_function
PR-URL: https://github.com/nodejs/node/pull/38506
Reviewed-By: Anna Henningsen <anna@addaleax.net>
Reviewed-By: Rich Trott <rtrott@gmail.com>
Reviewed-By: James M Snell <jasnell@gmail.com>
PR-URL: https://github.com/nodejs/node/pull/38220
Reviewed-By: Richard Lau <rlau@redhat.com>
Reviewed-By: Michaël Zasso <targos@protonmail.com>
Reviewed-By: Beth Griggs <bgriggs@redhat.com>
Reviewed-By: James M Snell <jasnell@gmail.com>
Reviewed-By: Colin Ihrig <cjihrig@gmail.com>
Reviewed-By: Luigi Pinca <luigipinca@gmail.com>
Reviewed-By: Darshan Sen <raisinten@gmail.com>
Currently, a reference is being unlinked from the list of references
tracked by the environment when `v8impl::Reference::Delete` is called.
This causes a leak when deletion must be deferred because the finalizer
hasn't yet run, but the finalizer does not run because environment
teardown is in progress, and so no more gc runs will happen, and the
`FinalizeAll` run that happens during environment teardown does not
catch the reference because it's no longer in the list. The test below
will fail when running with ASAN:
```
./node ./test/node-api/test_worker_terminate_finalization/test.js
```
OTOH if, to address the above leak, we make a special case to not
unlink a reference during environment teardown, we run into a
situation where the reference gets deleted by
`v8impl::Reference::Delete` but does not get unlinked because it's
environment teardown time. This leaves a stale pointer in the linked
list which will result in a use-after-free in `FinalizeAll` during
environment teardown. The test below will fail if we make the above
change:
```
./node -e "require('./test/node-api/test_instance_data/build/Release/test_ref_then_set.node');"
```
Thus, we unlink a reference precisely when we destroy it – in its
destructor.
Refs: https://github.com/nodejs/node/issues/34731
Refs: https://github.com/nodejs/node/pull/34839
Refs: https://github.com/nodejs/node/issues/35620
Refs: https://github.com/nodejs/node/issues/35777
Fixes: https://github.com/nodejs/node/issues/35778
Signed-off-by: Gabriel Schulhof <gabriel.schulhof@intel.com>
PR-URL: https://github.com/nodejs/node/pull/35933
Reviewed-By: Rich Trott <rtrott@gmail.com>
Reviewed-By: Michael Dawson <midawson@redhat.com>
Reviewed-By: Zeyu Yang <himself65@outlook.com>
Fixes: https://github.com/nodejs/node/issues/35620
This reverts commit a6b655614f which
changed finalization behavior related to N-API. We will investigate
the original issue with the test separately.
PR-URL: https://github.com/nodejs/node/pull/35777
Reviewed-By: Gireesh Punathil <gpunathi@in.ibm.com>
Reviewed-By: Rich Trott <rtrott@gmail.com>
Reviewed-By: Juan José Arboleda <soyjuanarbol@gmail.com>
Reviewed-By: Chengzhong Wu <legendecas@gmail.com>
Reviewed-By: Gerhard Stöbich <deb2001-github@yahoo.de>
Enforce `//` for multiline comments. Some tests mixed and matched, and
at least one did so in a (to me) surprising way.
PR-URL: https://github.com/nodejs/node/pull/35485
Reviewed-By: Colin Ihrig <cjihrig@gmail.com>
Reviewed-By: James M Snell <jasnell@gmail.com>
* Avoid passing core `void*` and function pointers into add-on.
* Document `napi_async_cleanup_hook_handle` type.
* Render receipt of the handle mandatory from the point where the
hook gets called. Removal of the handle remains mandatory.
Fixes: https://github.com/nodejs/node/issues/34715
Signed-off-by: Gabriel Schulhof <gabriel.schulhof@intel.com>
Co-authored-by: Anna Henningsen <github@addaleax.net>
PR-URL: https://github.com/nodejs/node/pull/34819
Reviewed-By: Michael Dawson <michael_dawson@ca.ibm.com>
Reviewed-By: Zeyu Yang <himself65@outlook.com>
When deleting a weak reference that has no finalizer we must not defer
deletion until the non-existent finalizer gets called.
Fixes: https://github.com/nodejs/node/issues/34731
Signed-off-by: Gabriel Schulhof <gabriel.schulhof@intel.com>
PR-URL: https://github.com/nodejs/node/pull/34839
Reviewed-By: Anna Henningsen <anna@addaleax.net>
Reviewed-By: Gus Caplan <me@gus.host>
Reviewed-By: James M Snell <jasnell@gmail.com>
Reviewed-By: Rich Trott <rtrott@gmail.com>
The test was missing an initialization of the global `ref` variable
because there was also an unused local one, leading to failures
like the one seen in https://github.com/nodejs/node/pull/34625.
PR-URL: https://github.com/nodejs/node/pull/34726
Reviewed-By: Richard Lau <riclau@uk.ibm.com>
Reviewed-By: James M Snell <jasnell@gmail.com>
Reviewed-By: Ricky Zhou <0x19951125@gmail.com>
Sometimes addons need to perform cleanup actions, for example
closing libuv handles or waiting for requests to finish, that
cannot be performed synchronously.
Add C++ API and N-API functions that allow providing such
asynchronous cleanup hooks.
Fixes: https://github.com/nodejs/node/issues/34567
PR-URL: https://github.com/nodejs/node/pull/34572
Reviewed-By: James M Snell <jasnell@gmail.com>
Reviewed-By: Gabriel Schulhof <gabriel.schulhof@intel.com>
* Prefix functions with `static` to make them local
* Remove anonymous namespaces
* `nullptr` -> `NULL`
* .cc -> .c and update binding.gyp
* `static_cast<x>()` -> `(x)()`
* Replace `new`/`delete` with `malloc()`/`free()`
(only in test_callback_scope)
* Move lambda out and convert to local function
(only in test_callback_scope)
* Remove superfluous `#include <vector>`
(only in test_callback_scope_recurse)
Some tests are best left as C++.
```bash
ls -l test/{node-api,js-native-api}/*/*.cc
```
for those remaining as C++ tests.
Signed-off-by: Gabriel Schulhof <gabriel.schulhof@intel.com>
PR-URL: https://github.com/nodejs/node/pull/34615
Reviewed-By: James M Snell <jasnell@gmail.com>
Reviewed-By: Anna Henningsen <anna@addaleax.net>
Reviewed-By: Jiawen Geng <technicalcute@gmail.com>
Reviewed-By: Joyee Cheung <joyeec9h3@gmail.com>
Reviewed-By: Gerhard Stöbich <deb2001-github@yahoo.de>
This feature does not work when a module is imported using ECMAScript
modules specification, therefore it is deprecated.
Fixes: https://github.com/nodejs/modules/issues/469
Backport-PR-URL: https://github.com/nodejs/node/pull/33533
PR-URL: https://github.com/nodejs/node/pull/32217
Reviewed-By: Gus Caplan <me@gus.host>
Reviewed-By: Bradley Farias <bradley.meck@gmail.com>
Reviewed-By: Anna Henningsen <anna@addaleax.net>
Reviewed-By: James M Snell <jasnell@gmail.com>
Reviewed-By: Matteo Collina <matteo.collina@gmail.com>
Reviewed-By: Tobias Nießen <tniessen@tnie.de>
Reviewed-By: Ruben Bridgewater <ruben@bridgewater.de>
This reverts commit d26ca06c16 because
it breaks running the tests in debug mode, as
`v8::Isolate::GetCurrent()` is not allowed if no `Isolate` is active
on the current thread.
Refs: https://github.com/nodejs/node/pull/33276
Refs: https://github.com/nodejs/node/pull/32860
PR-URL: https://github.com/nodejs/node/pull/33453
Reviewed-By: Daniel Bevenius <daniel.bevenius@gmail.com>
Reviewed-By: Colin Ihrig <cjihrig@gmail.com>
Reviewed-By: James M Snell <jasnell@gmail.com>
Reviewed-By: David Carlier <devnexen@gmail.com>
Reviewed-By: Gireesh Punathil <gpunathi@in.ibm.com>
Reviewed-By: Gabriel Schulhof <gabriel.schulhof@intel.com>
Release `Buffer` and `ArrayBuffer` instances that were created through
our addon APIs and have finalizers attached to them only after V8 has
called the deleter callback passed to the `BackingStore`, instead of
relying on our own GC callback(s).
This fixes the following race condition:
1. Addon code allocates pointer P via `malloc`.
2. P is passed into `napi_create_external_buffer` with a finalization
callback which calls `free(P)`. P is inserted into V8’s global array
buffer table for tracking.
3. The finalization callback is executed on GC. P is freed and returned
to the allocator. P is not yet removed from V8’s global array
buffer table. (!)
4. Addon code attempts to allocate memory once again. The allocator
returns P, as it is now available.
5. P is passed into `napi_create_external_buffer`. P still has not been
removed from the v8 global array buffer table.
6. The world ends with `Check failed: result.second`.
Since our API contract is to call the finalizer on the JS thread on
which the `ArrayBuffer` was created, but V8 may call the `BackingStore`
deleter callback on another thread, fixing this requires posting
a task back to the JS thread.
Refs: https://github.com/nodejs/node/issues/32463#issuecomment-625877175
Fixes: https://github.com/nodejs/node/issues/32463
PR-URL: https://github.com/nodejs/node/pull/33321
Reviewed-By: James M Snell <jasnell@gmail.com>
`sizeof(data)` does not return the correct result here, as it measures
the size of the `data` variable, not what it points to.
PR-URL: https://github.com/nodejs/node/pull/33115
Reviewed-By: Colin Ihrig <cjihrig@gmail.com>
Reviewed-By: James M Snell <jasnell@gmail.com>
PR-URL: https://github.com/nodejs/node/pull/33114
Reviewed-By: Chengzhong Wu <legendecas@gmail.com>
Reviewed-By: Colin Ihrig <cjihrig@gmail.com>
Reviewed-By: James M Snell <jasnell@gmail.com>
Reviewed-By: Gerhard Stöbich <deb2001-github@yahoo.de>
napi_async_context should be an opaque type and not be used as same as
node::async_context.
PR-URL: https://github.com/nodejs/node/pull/32928
Reviewed-By: Michael Dawson <michael_dawson@ca.ibm.com>
Reviewed-By: Anna Henningsen <anna@addaleax.net>
We introduce status `napi_would_deadlock` to be used as a return status
by `napi_call_threadsafe_function` if the call is made with
`napi_tsfn_blocking` on the main thread and the queue is full.
Fixes: https://github.com/nodejs/node/issues/32615
Signed-off-by: Gabriel Schulhof <gabriel.schulhof@intel.com>
PR-URL: https://github.com/nodejs/node/pull/32860
Reviewed-By: Ben Noordhuis <info@bnoordhuis.nl>
Reviewed-By: Michael Dawson <michael_dawson@ca.ibm.com>
Reviewed-By: Zeyu Yang <himself65@outlook.com>
We introduce status `napi_would_deadlock` to be used as a return status
by `napi_call_threadsafe_function` if the call is made with
`napi_tsfn_blocking` on the main thread and the queue is full.
PR-URL: https://github.com/nodejs/node/pull/32689
Fixes: https://github.com/nodejs/node/issues/32615
Signed-off-by: Gabriel Schulhof <gabriel.schulhof@intel.com>
Reviewed-By: James M Snell <jasnell@gmail.com>
Reviewed-By: Anna Henningsen <anna@addaleax.net>
Reviewed-By: Colin Ihrig <cjihrig@gmail.com>
Reviewed-By: David Carlier <devnexen@gmail.com>
Reviewed-By: Chengzhong Wu <legendecas@gmail.com>
Reviewed-By: Michael Dawson <michael_dawson@ca.ibm.com>
Mark all N-APIs that have been added since version 5 as stable.
PR-URL: https://github.com/nodejs/node/pull/32058
Fixes: https://github.com/nodejs/abi-stable-node/issues/393
Co-Authored-By: legendecas <legendecas@gmail.com>
Signed-off-by: Gabriel Schulhof <gabriel.schulhof@intel.com>
Reviewed-By: Chengzhong Wu <legendecas@gmail.com>
Reviewed-By: Ben Noordhuis <info@bnoordhuis.nl>
Reviewed-By: Colin Ihrig <cjihrig@gmail.com>
Reviewed-By: Ruben Bridgewater <ruben@bridgewater.de>
Reviewed-By: Michael Dawson <michael_dawson@ca.ibm.com>
Instance data associated with a `napi_env` is no longer stored on the
env itself but is instead rendered as a reference. Since
`v8impl::Reference` is tied to a JS object, this modification factors
out the `v8impl::Reference` refcounting and the deletion process into
a base class for `v8impl::Reference`, called `v8impl::RefBase`. The
instance data is then stored as a `v8impl::RefBase`, along with other
references, preventing a segfault that arises from the fact that, up
until now, upon `napi_env` destruction, the instance data was freed
after all references had already been forcefully freed. If the addon
freed a reference during the `napi_set_instance_data` finalizer
callback, such a reference had already been freed during environment
teardown, causing a double free.
Re: https://github.com/nodejs/node-addon-api/pull/663
PR-URL: https://github.com/nodejs/node/pull/31638
Reviewed-By: Anna Henningsen <anna@addaleax.net>
Reviewed-By: David Carlier <devnexen@gmail.com>
This ports `test/addons/worker-buffer-callback` to N-API,
with the small exception of using external `ArrayBuffer`s rather
than external Node.js `Buffer`s.
PR-URL: https://github.com/nodejs/node/pull/30551
Reviewed-By: Gabriel Schulhof <gabriel.schulhof@intel.com>
Reviewed-By: Colin Ihrig <cjihrig@gmail.com>
Reviewed-By: Denys Otrishko <shishugi@gmail.com>
Reviewed-By: Michael Dawson <michael_dawson@ca.ibm.com>
PR-URL: https://github.com/nodejs/node/pull/29401
Reviewed-By: Ben Noordhuis <info@bnoordhuis.nl>
Reviewed-By: James M Snell <jasnell@gmail.com>
Reviewed-By: Colin Ihrig <cjihrig@gmail.com>
Reviewed-By: Michael Dawson <michael_dawson@ca.ibm.com>
Adds `napi_set_instance_data()` and `napi_get_instance_data()`, which
allow native addons to store their data on and retrieve their data from
`napi_env`. `napi_set_instance_data()` accepts a finalizer which is
called when the `node::Environment()` is destroyed.
This entails rendering the `napi_env` local to each add-on.
Fixes: https://github.com/nodejs/abi-stable-node/issues/378
PR-URL: https://github.com/nodejs/node/pull/28682
Reviewed-By: Ben Noordhuis <info@bnoordhuis.nl>
Reviewed-By: Michael Dawson <michael_dawson@ca.ibm.com>
The `Execute()` callback is not allowed to call into JS, so
we should use `assert()` instead of potentially throwing JS errors.
PR-URL: https://github.com/nodejs/node/pull/28423
Fixes: https://github.com/nodejs/help/issues/1998
Reviewed-By: Gireesh Punathil <gpunathi@in.ibm.com>
Reviewed-By: Yongsheng Zhang <zyszys98@gmail.com>
Reviewed-By: Colin Ihrig <cjihrig@gmail.com>
Reviewed-By: Gabriel Schulhof <gabriel.schulhof@intel.com>
Reviewed-By: Rich Trott <rtrott@gmail.com>
Reviewed-By: Luigi Pinca <luigipinca@gmail.com>
Reviewed-By: Michael Dawson <michael_dawson@ca.ibm.com>
We have a test that verifies that JS execution from the Buffer
finalizer is accepted, and that errors thrown are passed
down synchronously.
However, since the finalizer executes during GC, this is behaviour is
fundamentally invalid and, for good reasons, disallowed by the
JS engine. This leaves us with the options of either finding a way
to allow JS execution from the callback, or explicitly forbidding it on
the N-API side as well.
This commit implements the former option, since it is the more
backwards-compatible one, in the sense that the current situation
sometimes appears to work as well and we should not break that
behaviour if we don’t have to, but rather try to actually make it
work reliably.
Since GC timing is largely unobservable anyway, this commit moves
the callback into a `SetImmediate()`, as we do elsewhere in the code,
and a second pass callback is not an easily implemented option,
as the API is supposed to wrap around Node’s `Buffer` API.
In this case, exceptions are handled like other uncaught exceptions.
Two tests have to be adjusted to account for the timing difference.
This is unfortunate, but unavoidable if we want to conform to the
JS engine API contract and keep all tests.
Fixes: https://github.com/nodejs/node/issues/26754
PR-URL: https://github.com/nodejs/node/pull/28082
Reviewed-By: Michael Dawson <michael_dawson@ca.ibm.com>
Reviewed-By: Gabriel Schulhof <gabriel.schulhof@intel.com>
Reviewed-By: James M Snell <jasnell@gmail.com>
Cover import of a *.node file with a policy manifest. Add invalid
integrity test case.
PR-URL: https://github.com/nodejs/node/pull/27903
Reviewed-By: Anna Henningsen <anna@addaleax.net>
Reviewed-By: Rich Trott <rtrott@gmail.com>
The test_threadsafe_function doesn't seem to be flaky anymore on
Windows. Optimistically removing the flaky designation in the relevant
status file.
Refs: https://github.com/nodejs/node/issues/23621#issuecomment-468938980
PR-URL: https://github.com/nodejs/node/pull/26403
Reviewed-By: Benjamin Gruenbaum <benjamingr@gmail.com>
Reviewed-By: Richard Lau <riclau@uk.ibm.com>
Reviewed-By: Anna Henningsen <anna@addaleax.net>
Reviewed-By: Colin Ihrig <cjihrig@gmail.com>
Reviewed-By: James M Snell <jasnell@gmail.com>
Reviewed-By: Ruben Bridgewater <ruben@bridgewater.de>
Check whether calling into JS is allowed before doing so.
PR-URL: https://github.com/nodejs/node/pull/26127
Reviewed-By: Gus Caplan <me@gus.host>
Reviewed-By: Michael Dawson <michael_dawson@ca.ibm.com>
Partition test/addons-napi into test/js-native-api and test/node-api to
isolate the Node.js-agnostic portion of the N-API tests from the
Node.js-specific portion.
PR-URL: https://github.com/nodejs/node/pull/24557
Reviewed-By: Refael Ackermann <refack@gmail.com>
Reviewed-By: Daniel Bevenius <daniel.bevenius@gmail.com>
