[v9.x backport] 17738 and 17841 - #18488
Conversation
Sorry, something went wrong.
Do not share unnecessary information about nextTick state between JS & C++, instead only track whether a nextTick is scheduled or not. Turn nextTickQueue into an Object instead of a class since multiple instances are never created. Other assorted refinements and refactoring. PR-URL: nodejs#17738 Reviewed-By: Anna Henningsen <anna@addaleax.net>
When a process encounters a _fatalException that is caught, it should schedule execution of nextTicks but not in an arbitrary place of the next Immediates queue. Instead, add a no-op function to the queue that will ensure processImmediate runs, which will then ensure that nextTicks are processed at the end. PR-URL: nodejs#17841 Reviewed-By: Anna Henningsen <anna@addaleax.net> Reviewed-By: James M Snell <jasnell@gmail.com>
|
Once this lands, #17736 #17881 #17879 #18139 should land and not fail the tests if applied in the right order. |
Sorry, something went wrong.
|
@apapirovski after landing this PR I am still seeing #17736 fail as below. Would it be possible for you to add all of those other blocked PRs to this backport PR and ensure everything is working === release test-timer-immediate ===
Path: parallel/test-timer-immediate
module.js:704
process._tickCallback();
^
TypeError: process._tickCallback is not a function
at Function.Module.runMain (module.js:704:11)
at startup (bootstrap_node.js:190:16)
at bootstrap_node.js:657:3
As an aside, #18139 was not landing cleanly |
Sorry, something went wrong.
|
@MylesBorins working on this now. Should have an update shortly. Just an FYI, the unref Immediates in #18139 are semver-minor so will need to wait until 9.7.0 now. |
Sorry, something went wrong.
|
I'm going to open a new PR for this given the expanded scope. |
Sorry, something went wrong.
Just required commits applied in the right order.
CI: https://ci.nodejs.org/job/node-test-pull-request/12857/
Checklist
Affected core subsystem(s)
process