src: use effective cppgc wrapper id to deduce non-cppgc id by joyeecheung · Pull Request #48660 · nodejs/node

joyeecheung · 2023-07-05T15:48:01Z

deps: V8: cherry-pick 93275031284c

Original commit message:

[cppgc] expose wrapper descriptor on CppHeap
This makes it possible for embedders to:
1. Avoid creating wrapper objects that happen to have a layout that
leads V8 to consider the object cppgc-managed while it's not.
Refs: https://github.com/nodejs/node/pull/43521
2. Create cppgc-managed wrapper objects when they do not own the
CppHeap. Refs: https://github.com/nodejs/node/pull/45704
Bug: v8:13960
Change-Id: If31f4d56c5ead59dc0d56f937494d23d631f7438
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/4598833
Reviewed-by: Michael Lippautz <mlippautz@chromium.org>
Commit-Queue: Michael Lippautz <mlippautz@chromium.org>
Cr-Commit-Position: refs/heads/main@{#88490}

Refs: v8/v8@9327503

src: use effective cppgc wrapper id to deduce non-cppgc id

Previously we hard-code a wrapper id to be used in BaseObjects
to avoid accidentally triggering cppgc on these non-cppgc-managed
objects, but hard-coding can be be hacky and result in mismatch
when we start to create CppHeap ourselves. This patch makes it
more robust by deducing non-cppgc id from the effective cppgc id,
if there is one.

Refs: #48660

nodejs-github-bot · 2023-07-05T15:48:07Z

Review requested:

@nodejs/gyp
@nodejs/startup
@nodejs/v8-update

richardlau · 2023-07-05T15:56:28Z

The V8 cherry-pick should have also bumped v8_embedder_string in common.gypi.

joyeecheung · 2023-07-05T16:14:22Z

@richardlau Thanks for catching it, it probably happened because I created the backport commit before ce4102e landed. I've amended the first commit.

nodejs-github-bot · 2023-07-05T19:04:48Z

CI: https://ci.nodejs.org/job/node-test-pull-request/52623/

nodejs-github-bot · 2023-07-05T21:14:34Z

CI: https://ci.nodejs.org/job/node-test-pull-request/52627/

richardlau · 2023-07-07T12:00:21Z

Did this get a V8 CI run?

joyeecheung · 2023-07-07T14:54:27Z

V8 CI: https://ci.nodejs.org/job/node-test-commit-v8-linux/5433/

nodejs-github-bot · 2023-07-07T14:54:36Z

V8 CI: https://ci.nodejs.org/job/node-test-commit-v8-linux/nodes=rhel8-s390x,v8test=v8test/5433/

nodejs-github-bot · 2023-07-07T14:54:41Z

V8 CI: https://ci.nodejs.org/job/node-test-commit-v8-linux/nodes=benchmark-ubuntu2204-intel-64,v8test=v8test/5433/

nodejs-github-bot · 2023-07-07T14:54:47Z

V8 CI: https://ci.nodejs.org/job/node-test-commit-v8-linux/nodes=rhel8-ppc64le,v8test=v8test/5433/

joyeecheung · 2023-07-08T08:16:15Z

Looks like there is a build error in our build machine, perhaps depot_tools is out of date ?

16:54:59 Cloning into '/home/iojs/build/workspace/node-test-commit-v8-linux/deps/v8/_depot_tools'...
16:55:05 File "/home/iojs/build/workspace/node-test-commit-v8-linux/deps/v8/_depot_tools/gclient.py", line 3461
16:55:05 f'160000,{value},{name}'
16:55:05 ^
16:55:05 SyntaxError: invalid syntax

joyeecheung · 2023-07-08T13:15:17Z

Rebased to resolve merge conflict

nodejs-github-bot · 2023-07-08T13:15:44Z

CI: https://ci.nodejs.org/job/node-test-pull-request/52651/

nodejs-github-bot · 2023-07-12T14:42:53Z

CI: https://ci.nodejs.org/job/node-test-pull-request/52734/

Original commit message: [cppgc] expose wrapper descriptor on CppHeap This makes it possible for embedders to: 1. Avoid creating wrapper objects that happen to have a layout that leads V8 to consider the object cppgc-managed while it's not. Refs: nodejs#43521 2. Create cppgc-managed wrapper objects when they do not own the CppHeap. Refs: nodejs#45704 Bug: v8:13960 Change-Id: If31f4d56c5ead59dc0d56f937494d23d631f7438 Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/4598833 Reviewed-by: Michael Lippautz <mlippautz@chromium.org> Commit-Queue: Michael Lippautz <mlippautz@chromium.org> Cr-Commit-Position: refs/heads/main@{#88490} Refs: v8/v8@9327503 PR-URL: nodejs#48660 Reviewed-By: Chengzhong Wu <legendecas@gmail.com> Reviewed-By: Jiawen Geng <technicalcute@gmail.com>

Previously we hard-code a wrapper id to be used in BaseObjects to avoid accidentally triggering cppgc on these non-cppgc-managed objects, but hard-coding can be be hacky and result in mismatch when we start to create CppHeap ourselves. This patch makes it more robust by deducing non-cppgc id from the effective cppgc id, if there is one. PR-URL: nodejs#48660 Refs: v8/v8@9327503 Reviewed-By: Chengzhong Wu <legendecas@gmail.com> Reviewed-By: Jiawen Geng <technicalcute@gmail.com>

Original commit message: [cppgc] expose wrapper descriptor on CppHeap This makes it possible for embedders to: 1. Avoid creating wrapper objects that happen to have a layout that leads V8 to consider the object cppgc-managed while it's not. Refs: nodejs#43521 2. Create cppgc-managed wrapper objects when they do not own the CppHeap. Refs: nodejs#45704 Bug: v8:13960 Change-Id: If31f4d56c5ead59dc0d56f937494d23d631f7438 Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/4598833 Reviewed-by: Michael Lippautz <mlippautz@chromium.org> Commit-Queue: Michael Lippautz <mlippautz@chromium.org> Cr-Commit-Position: refs/heads/main@{#88490} Refs: v8/v8@9327503 PR-URL: nodejs#48660 Reviewed-By: Chengzhong Wu <legendecas@gmail.com> Reviewed-By: Jiawen Geng <technicalcute@gmail.com>

Previously we hard-code a wrapper id to be used in BaseObjects to avoid accidentally triggering cppgc on these non-cppgc-managed objects, but hard-coding can be be hacky and result in mismatch when we start to create CppHeap ourselves. This patch makes it more robust by deducing non-cppgc id from the effective cppgc id, if there is one. PR-URL: nodejs#48660 Refs: v8/v8@9327503 Reviewed-By: Chengzhong Wu <legendecas@gmail.com> Reviewed-By: Jiawen Geng <technicalcute@gmail.com>

Original commit message: [cppgc] expose wrapper descriptor on CppHeap This makes it possible for embedders to: 1. Avoid creating wrapper objects that happen to have a layout that leads V8 to consider the object cppgc-managed while it's not. Refs: nodejs#43521 2. Create cppgc-managed wrapper objects when they do not own the CppHeap. Refs: nodejs#45704 Bug: v8:13960 Change-Id: If31f4d56c5ead59dc0d56f937494d23d631f7438 Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/4598833 Reviewed-by: Michael Lippautz <mlippautz@chromium.org> Commit-Queue: Michael Lippautz <mlippautz@chromium.org> Cr-Commit-Position: refs/heads/main@{#88490} Refs: v8/v8@9327503 PR-URL: nodejs#48660 Reviewed-By: Chengzhong Wu <legendecas@gmail.com> Reviewed-By: Jiawen Geng <technicalcute@gmail.com>

Previously we hard-code a wrapper id to be used in BaseObjects to avoid accidentally triggering cppgc on these non-cppgc-managed objects, but hard-coding can be be hacky and result in mismatch when we start to create CppHeap ourselves. This patch makes it more robust by deducing non-cppgc id from the effective cppgc id, if there is one. PR-URL: nodejs#48660 Refs: v8/v8@9327503 Reviewed-By: Chengzhong Wu <legendecas@gmail.com> Reviewed-By: Jiawen Geng <technicalcute@gmail.com>

Original commit message: [cppgc] expose wrapper descriptor on CppHeap This makes it possible for embedders to: 1. Avoid creating wrapper objects that happen to have a layout that leads V8 to consider the object cppgc-managed while it's not. Refs: #43521 2. Create cppgc-managed wrapper objects when they do not own the CppHeap. Refs: #45704 Bug: v8:13960 Change-Id: If31f4d56c5ead59dc0d56f937494d23d631f7438 Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/4598833 Reviewed-by: Michael Lippautz <mlippautz@chromium.org> Commit-Queue: Michael Lippautz <mlippautz@chromium.org> Cr-Commit-Position: refs/heads/main@{#88490} Refs: v8/v8@9327503 PR-URL: #48660 Backport-PR-URL: #49187 Reviewed-By: Chengzhong Wu <legendecas@gmail.com> Reviewed-By: Jiawen Geng <technicalcute@gmail.com> Refs: #40786 Refs: https://docs.google.com/document/d/1ny2Qz_EsUnXGKJRGxoA-FXIE2xpLgaMAN6jD7eAkqFQ/edit

Previously we hard-code a wrapper id to be used in BaseObjects to avoid accidentally triggering cppgc on these non-cppgc-managed objects, but hard-coding can be be hacky and result in mismatch when we start to create CppHeap ourselves. This patch makes it more robust by deducing non-cppgc id from the effective cppgc id, if there is one. PR-URL: #48660 Backport-PR-URL: #49187 Refs: v8/v8@9327503 Reviewed-By: Chengzhong Wu <legendecas@gmail.com> Reviewed-By: Jiawen Geng <technicalcute@gmail.com> Refs: #40786 Refs: https://docs.google.com/document/d/1ny2Qz_EsUnXGKJRGxoA-FXIE2xpLgaMAN6jD7eAkqFQ/edit

This patch: - Initializes cppgc in InitializeOncePerProcess() when kNoInitializeCppgc is not set - Create a CppHeap and attach it to the Isolate when there isn't one already during IsolateData initialization. The CppHeap is detached and terminated when IsolateData is freed. - Publishes the cppgc headers in the tarball. This allows C++ addons to start using cppgc to manage objects. A helper node::SetCppgcReference() is also added to help addons enable cppgc tracing in a user-defined object. Co-authored-by: Joyee Cheung <joyeec9h3@gmail.com> Refs: #40786 PR-URL: #45704 Backport-PR-URL: #49187 Refs: https://docs.google.com/document/d/1ny2Qz_EsUnXGKJRGxoA-FXIE2xpLgaMAN6jD7eAkqFQ/edit Reviewed-By: Stephen Belanger <admin@stephenbelanger.com> Reviewed-By: Rafael Gonzaga <rafael.nunu@hotmail.com> PR-URL: #48660 Refs: v8/v8@9327503

Notable changes: deps: * V8: cherry-pick 93275031284c (Joyee Cheung) #48660 doc: * add new TSC members (Michael Dawson) #48841 * add rluvaton to collaborators (Raz Luvaton) #49215 esm: * unflag import.meta.resolve (Guy Bedford) #49028 * add `initialize` hook, integrate with `register` (Izaak Schroeder) #48842 * unflag `Module.register` and allow nested loader `import()` (Izaak Schroeder) #48559 inspector: * (SEMVER-MINOR) open add `SymbolDispose` (Chemi Atlow) #48765 test_runner: * (SEMVER-MINOR) expose location of tests (Colin Ihrig) #48975 module: * implement `register` utility (João Lenon) #46826 * make CJS load from ESM loader (Antoine du Hamel) #47999 src: * add built-in `.env` file support (Yagiz Nizipli) #48890 PR-URL: #49185

Notable changes: deps: * V8: cherry-pick 93275031284c (Joyee Cheung) #48660 doc: * add new TSC members (Michael Dawson) #48841 * add rluvaton to collaborators (Raz Luvaton) #49215 esm: * unflag import.meta.resolve (Guy Bedford) #49028 * add `initialize` hook, integrate with `register` (Izaak Schroeder) #48842 * unflag `Module.register` and allow nested loader `import()` (Izaak Schroeder) #48559 inspector: * (SEMVER-MINOR) open add `SymbolDispose` (Chemi Atlow) #48765 module: * implement `register` utility (João Lenon) #46826 * make CJS load from ESM loader (Antoine du Hamel) #47999 src: * add built-in `.env` file support (Yagiz Nizipli) #48890 * initialize cppgc (Daryl Haresign and Joyee Cheung) #48660 and #45704 test_runner: * (SEMVER-MINOR) expose location of tests (Colin Ihrig) #48975 PR-URL: #49185

Original commit message: [cppgc] expose wrapper descriptor on CppHeap This makes it possible for embedders to: 1. Avoid creating wrapper objects that happen to have a layout that leads V8 to consider the object cppgc-managed while it's not. Refs: nodejs#43521 2. Create cppgc-managed wrapper objects when they do not own the CppHeap. Refs: nodejs#45704 Bug: v8:13960 Change-Id: If31f4d56c5ead59dc0d56f937494d23d631f7438 Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/4598833 Reviewed-by: Michael Lippautz <mlippautz@chromium.org> Commit-Queue: Michael Lippautz <mlippautz@chromium.org> Cr-Commit-Position: refs/heads/main@{#88490} Refs: v8/v8@9327503 PR-URL: nodejs#48660 Backport-PR-URL: nodejs#49187 Reviewed-By: Chengzhong Wu <legendecas@gmail.com> Reviewed-By: Jiawen Geng <technicalcute@gmail.com> Refs: nodejs#40786 Refs: https://docs.google.com/document/d/1ny2Qz_EsUnXGKJRGxoA-FXIE2xpLgaMAN6jD7eAkqFQ/edit

nodejs-github-bot added lib / src Issues and PRs related to general changes in the lib or src directory. needs-ci PRs that need a full CI run. labels Jul 5, 2023

joyeecheung force-pushed the wrapper branch from b221c47 to b84c5e0Compare July 5, 2023 16:12

joyeecheung mentioned this pull request Jul 5, 2023
src,tools: initialize cppgc #45704
Merged

joyeecheung force-pushed the wrapper branch from b84c5e0 to 168c23aCompare July 5, 2023 18:35

This was referenced Jul 6, 2023
CI Reliability 2023-07-06 nodejs/reliability#609
Open
CI Reliability 2023-07-07 nodejs/reliability#610
Open

legendecas approved these changes Jul 7, 2023
View reviewed changes

gengjiawen approved these changes Jul 7, 2023
View reviewed changes

gengjiawen added the commit-queue Add this label to land a pull request using GitHub Actions. label Jul 7, 2023

joyeecheung removed the commit-queue Add this label to land a pull request using GitHub Actions. label Jul 7, 2023

github-actionsBot mentioned this pull request Jul 8, 2023
CI Reliability 2023-07-08 nodejs/reliability#611
Open
23 tasks

joyeecheung force-pushed the wrapper branch from 168c23a to f5abc09Compare July 8, 2023 13:14

This was referenced Jul 9, 2023
CI Reliability 2023-07-09 nodejs/reliability#612
Open
CI Reliability 2023-07-10 nodejs/reliability#613
Open

richardlau mentioned this pull request Jul 10, 2023
deps: update V8 to 11.6 #48710
Closed

joyeecheung mentioned this pull request Aug 15, 2023
[v20.x backport] cppgc related changes #49187
Closed

UlisesGascon mentioned this pull request Aug 18, 2023
v20.6.0 proposal #49185
Merged

targos added backported-to-v20.x and removed backport-requested-v20.x labels Nov 12, 2023

Uh oh!

Uh oh!

Conversation

joyeecheung commented Jul 5, 2023• edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

deps: V8: cherry-pick 93275031284c

src: use effective cppgc wrapper id to deduce non-cppgc id

Uh oh!

nodejs-github-bot commented Jul 5, 2023

Uh oh!

richardlau commented Jul 5, 2023

Uh oh!

joyeecheung commented Jul 5, 2023

Uh oh!

nodejs-github-bot commented Jul 5, 2023

Uh oh!

nodejs-github-bot commented Jul 5, 2023

Uh oh!

richardlau commented Jul 7, 2023

Uh oh!

joyeecheung commented Jul 7, 2023

Uh oh!

nodejs-github-bot commented Jul 7, 2023

Uh oh!

nodejs-github-bot commented Jul 7, 2023

Uh oh!

nodejs-github-bot commented Jul 7, 2023

Uh oh!

joyeecheung commented Jul 8, 2023

Uh oh!

joyeecheung commented Jul 8, 2023

Uh oh!

nodejs-github-bot commented Jul 8, 2023

Uh oh!

nodejs-github-bot commented Jul 12, 2023

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

9 participants

joyeecheung commented Jul 5, 2023•
edited
Loading