Uh oh!
There was an error while loading. Please reload this page.
stream: drop per-chunk Promise alloc in pipeTo - #63572
Merged
nodejs-github-bot merged 1 commit intoJun 2, 2026
Merged
Conversation
Codecov Report✅ All modified and coverable lines are covered by tests. Additional details and impacted files@@ Coverage Diff @@## main #63572 +/- ##
==========================================
+ Coverage 90.33% 90.34% +0.01%
==========================================
Files 732 732 Lines 236454 236458 +4 Branches 44540 44530 -10 ==========================================
+ Hits 213589 213635 +46 + Misses 14581 14524 -57 - Partials 8284 8299 +15
🚀 New features to boost your workflow:
|
aduh95
reviewed
May 26, 2026
Comment on lines
+1574
to
+1582
| // | ||
| // setPromiseHandled's job here is purely to silence the per-chunk | ||
| // unhandled-rejection event — the chain Promise + noop closure it | ||
| // would allocate are unobserved by anything else (errors propagate | ||
| // through writer.closed, which pipeTo monitors via watchErrored). | ||
| // markPromiseAsHandled sets V8's MarkAsHandled + MarkAsSilent flags | ||
| // directly, with no per-chunk allocation. The LAST state.currentWrite | ||
| // is awaited in waitForCurrentWrite during shutdown — markAsHandled | ||
| // does not affect that subsequent .then chain. |
Contributor
There was a problem hiding this comment.
I don't think this comment will make much sense after this PR is merged
Suggested change
| // | |
| // setPromiseHandled's job here is purely to silence the per-chunk | |
| // unhandled-rejection event — the chain Promise + noop closure it | |
| // would allocate are unobserved by anything else (errors propagate | |
| // through writer.closed, which pipeTo monitors via watchErrored). | |
| // markPromiseAsHandled sets V8's MarkAsHandled + MarkAsSilent flags | |
| // directly, with no per-chunk allocation. The LAST state.currentWrite | |
| // is awaited in waitForCurrentWrite during shutdown — markAsHandled | |
| // does not affect that subsequent .then chain. |
aduh95
reviewed
May 26, 2026
| @@ -35,6 +35,7 @@ const { | |||
| kPending, | |||
| }, | |||
| getPromiseDetails, | |||
| markPromiseAsHandled, | |||
Contributor
There was a problem hiding this comment.
Does it really make sense to import it just to re-export it? I don't like it, it makes it harder to realize it crosses the C++ boundary
aduh95
approved these changes
May 28, 2026
mcollinaforce-pushed
the
stream/pipeto-mark-as-handled
branch
from
May 28, 2026 08:58
5c8488d to
7932aaeComparenodejs-github-bot
commented
May 28, 2026
Collaborator
mcollinaforce-pushed
the
stream/pipeto-mark-as-handled
branch
from
May 28, 2026 15:15
7932aae to
80e3a93Comparenodejs-github-bot
commented
May 28, 2026
Collaborator
mcollinaforce-pushed
the
stream/pipeto-mark-as-handled
branch
from
May 30, 2026 16:59
80e3a93 to
9783b9dCompareSigned-off-by: Matteo Collina <hello@matteocollina.com>
mcollinaforce-pushed
the
stream/pipeto-mark-as-handled
branch
from
May 30, 2026 17:02
9783b9d to
8e65650Comparenodejs-github-bot
commented
Jun 2, 2026
Collaborator
Uh oh!
There was an error while loading. Please reload this page.
nodejs-github-bot
commented
Jun 2, 2026
Collaborator
Landed in 6af9026 |
aduh95 pushed a commit
that referenced
this pull request
Jun 18, 2026
Signed-off-by: Matteo Collina <hello@matteocollina.com> PR-URL: #63572 Reviewed-By: Antoine du Hamel <duhamelantoine1995@gmail.com>
4 tasks
aduh95 pushed a commit
that referenced
this pull request
Jul 20, 2026
Signed-off-by: Matteo Collina <hello@matteocollina.com> PR-URL: #63572 Reviewed-By: Antoine du Hamel <duhamelantoine1995@gmail.com>
aduh95 pushed a commit
that referenced
this pull request
Jul 21, 2026
Signed-off-by: Matteo Collina <hello@matteocollina.com> PR-URL: #63572 Reviewed-By: Antoine du Hamel <duhamelantoine1995@gmail.com>
aduh95 pushed a commit
that referenced
this pull request
Jul 22, 2026
Signed-off-by: Matteo Collina <hello@matteocollina.com> PR-URL: #63572 Reviewed-By: Antoine du Hamel <duhamelantoine1995@gmail.com>
aduh95 pushed a commit
that referenced
this pull request
Jul 30, 2026
Signed-off-by: Matteo Collina <hello@matteocollina.com> PR-URL: #63572 Reviewed-By: Antoine du Hamel <duhamelantoine1995@gmail.com>
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for freeto join this conversation on GitHub.
Already have an account?
Sign in to comment
3 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Replace
setPromiseHandledwithmarkPromiseAsHandledat the two pipeTo per-chunk call sites inreadablestream.js. The.then(undefined, () => {})chain thatsetPromiseHandledallocates per chunk has no observer in pipeTo: errors already propagate throughwriter.closed(whichwatchErroredmonitors) and the laststate.currentWriteis awaited inwaitForCurrentWriteduring shutdown (markAsHandled does not affect that subsequent.then). The per-chunk chain Promise + noop closure exist solely to silence the unhandled-rejection event:markPromiseAsHandledsets V8'sMarkAsHandled+MarkAsSilentflags directly with no allocation.benchmark/webstreams/pipe-to.jsat HWM=1024: ~5% throughput improvement (728K vs 694K ops/s mean, 6 alternating samples), with substantially tighter run-to-run variance.WPT streams parity preserved: 1403 subtests passing, 0 unexpected failures.