async_hooks,process: remove internalNextTick #19147

apapirovski · 2018-03-05T12:15:04Z

Instead of having mostly duplicate code in form of internalNextTick, instead use the existing defaultAsyncTriggerIdScope with a slight modification which allows undefined triggerAsyncId to be passed in, which then just triggers the callback with the provided arguments.

(The ability to pass in undefined removes the need for a bunch of conditional checks for whether the asyncId is set on a given resource.)

Refs: #19104

Checklist

make -j4 test (UNIX), or vcbuild test (Windows) passes
tests and/or benchmarks are included
documentation is changed or added
commit message follows commit guidelines

Affected core subsystem(s)

async_hooks, http, dgram, net, process, test

apapirovski · 2018-03-05T12:15:52Z

lib/_http_agent.js

@@ -342,10 +341,7 @@ Agent.prototype.destroy = function destroy() {
 function handleSocketCreation(request, informRequest) {
  return function handleSocketCreation_Inner(err, socket) {
    if (err) {
-      const asyncId = (socket && socket._handle && socket._handle.getAsyncId) ?


After checking all the call sites, it's not possible to receive a socket here if err is truthy. (I've also double-checked this by adding an assert during debugging.)

apapirovski · 2018-03-05T12:17:30Z

CI: https://ci.nodejs.org/job/node-test-pull-request/13501/
CitGM: https://ci.nodejs.org/view/Node.js-citgm/job/citgm-smoker/1321/

AndreasMadsen · 2018-03-05T17:00:54Z

lib/_http_outgoing.js

-        if (socketAsyncId === undefined) socketAsyncId = null;
-
-        nextTick(socketAsyncId, callback);
+        defaultTriggerAsyncIdScope(conn[async_id_symbol],


Is conn[async_id_symbol] === this.socket[async_id_symbol]? Also, this changes the socketAsyncId so if this is correct the comment should be updated too.

conn === this.socket so it's equivalent. Not quite clear on the second part of the feedback — the comment above it still applies.

AndreasMadsen · 2018-03-05T17:01:58Z

I think we should benchmark the http throughput before landing this.

edit: also, it is kinda an implementation detail that undefined works as expected. The function actually "asserts" that it isn't undefined. https://github.com/nodejs/node/blob/master/lib/internal/async_hooks.js#L284

apapirovski · 2018-03-05T17:10:33Z

Benchmark CI: https://ci.nodejs.org/view/Node.js%20benchmark/job/benchmark-node-micro-benchmarks/129/ (only http/simple and will go from there, don't want to tie up the CI for 40 hours)

apapirovski · 2018-03-05T17:27:22Z

edit: also, it is kinda an implementation detail that undefined works as expected. The function actually "asserts" that it isn't undefined. https://github.com/nodejs/node/blob/master/lib/internal/async_hooks.js#L284

I've adjusted the function to special handle undefined. It would be incorrect to use it with the current version.

apapirovski · 2018-03-05T21:58:14Z

Nothing meaningful from the benchmark. I think these changes are very unlikely to make a difference.

                                                                                       confidence improvement accuracy (*)   (**)   (***)
 http/simple.js chunkedEnc=0 c=500 chunks=1 len=102400 type='buffer' benchmarker='wrk'                -1.20 %       ±6.91% ±9.19% ±11.96%
 http/simple.js chunkedEnc=0 c=500 chunks=1 len=102400 type='bytes' benchmarker='wrk'                  1.41 %       ±4.84% ±6.45%  ±8.41%
 http/simple.js chunkedEnc=0 c=500 chunks=1 len=1024 type='buffer' benchmarker='wrk'                   2.26 %       ±3.52% ±4.69%  ±6.10%
 http/simple.js chunkedEnc=0 c=500 chunks=1 len=1024 type='bytes' benchmarker='wrk'                    0.89 %       ±4.21% ±5.61%  ±7.33%
 http/simple.js chunkedEnc=0 c=500 chunks=1 len=4 type='buffer' benchmarker='wrk'                     -2.00 %       ±3.70% ±4.92%  ±6.41%
 http/simple.js chunkedEnc=0 c=500 chunks=1 len=4 type='bytes' benchmarker='wrk'               **     -5.23 %       ±3.53% ±4.70%  ±6.11%
 http/simple.js chunkedEnc=0 c=500 chunks=4 len=102400 type='buffer' benchmarker='wrk'                -1.02 %       ±5.79% ±7.70% ±10.02%
 http/simple.js chunkedEnc=0 c=500 chunks=4 len=102400 type='bytes' benchmarker='wrk'                  0.21 %       ±3.06% ±4.08%  ±5.31%
 http/simple.js chunkedEnc=0 c=500 chunks=4 len=1024 type='buffer' benchmarker='wrk'                   2.30 %       ±3.86% ±5.13%  ±6.68%
 http/simple.js chunkedEnc=0 c=500 chunks=4 len=1024 type='bytes' benchmarker='wrk'                    0.78 %       ±3.72% ±4.95%  ±6.46%
 http/simple.js chunkedEnc=0 c=500 chunks=4 len=4 type='buffer' benchmarker='wrk'                     -0.21 %       ±3.52% ±4.68%  ±6.09%
 http/simple.js chunkedEnc=0 c=500 chunks=4 len=4 type='bytes' benchmarker='wrk'                       1.51 %       ±3.27% ±4.36%  ±5.68%
 http/simple.js chunkedEnc=0 c=50 chunks=1 len=102400 type='buffer' benchmarker='wrk'                 -0.59 %       ±5.48% ±7.29%  ±9.49%
 http/simple.js chunkedEnc=0 c=50 chunks=1 len=102400 type='bytes' benchmarker='wrk'                   2.26 %       ±3.39% ±4.52%  ±5.89%
 http/simple.js chunkedEnc=0 c=50 chunks=1 len=1024 type='buffer' benchmarker='wrk'                    2.56 %       ±2.76% ±3.67%  ±4.78%
 http/simple.js chunkedEnc=0 c=50 chunks=1 len=1024 type='bytes' benchmarker='wrk'                    -1.84 %       ±3.73% ±4.96%  ±6.45%
 http/simple.js chunkedEnc=0 c=50 chunks=1 len=4 type='buffer' benchmarker='wrk'               **      3.48 %       ±2.50% ±3.34%  ±4.35%
 http/simple.js chunkedEnc=0 c=50 chunks=1 len=4 type='bytes' benchmarker='wrk'                       -1.75 %       ±3.37% ±4.48%  ±5.83%
 http/simple.js chunkedEnc=0 c=50 chunks=4 len=102400 type='buffer' benchmarker='wrk'                 -0.66 %       ±5.72% ±7.61%  ±9.90%
 http/simple.js chunkedEnc=0 c=50 chunks=4 len=102400 type='bytes' benchmarker='wrk'                  -0.44 %       ±2.71% ±3.61%  ±4.69%
 http/simple.js chunkedEnc=0 c=50 chunks=4 len=1024 type='buffer' benchmarker='wrk'                   -0.31 %       ±3.58% ±4.77%  ±6.21%
 http/simple.js chunkedEnc=0 c=50 chunks=4 len=1024 type='bytes' benchmarker='wrk'                     0.35 %       ±3.26% ±4.34%  ±5.66%
 http/simple.js chunkedEnc=0 c=50 chunks=4 len=4 type='buffer' benchmarker='wrk'                       2.11 %       ±3.29% ±4.38%  ±5.71%
 http/simple.js chunkedEnc=0 c=50 chunks=4 len=4 type='bytes' benchmarker='wrk'                        1.13 %       ±3.31% ±4.41%  ±5.74%
 http/simple.js chunkedEnc=1 c=500 chunks=1 len=102400 type='buffer' benchmarker='wrk'                 0.54 %       ±6.20% ±8.25% ±10.74%
 http/simple.js chunkedEnc=1 c=500 chunks=1 len=102400 type='bytes' benchmarker='wrk'                  1.06 %       ±3.15% ±4.19%  ±5.45%
 http/simple.js chunkedEnc=1 c=500 chunks=1 len=1024 type='buffer' benchmarker='wrk'                   1.66 %       ±4.81% ±6.40%  ±8.33%
 http/simple.js chunkedEnc=1 c=500 chunks=1 len=1024 type='bytes' benchmarker='wrk'                    1.11 %       ±4.78% ±6.37%  ±8.32%
 http/simple.js chunkedEnc=1 c=500 chunks=1 len=4 type='buffer' benchmarker='wrk'                      3.30 %       ±3.86% ±5.14%  ±6.71%
 http/simple.js chunkedEnc=1 c=500 chunks=1 len=4 type='bytes' benchmarker='wrk'                       2.30 %       ±4.43% ±5.89%  ±7.67%
 http/simple.js chunkedEnc=1 c=500 chunks=4 len=102400 type='buffer' benchmarker='wrk'                 3.67 %       ±5.22% ±6.95%  ±9.05%
 http/simple.js chunkedEnc=1 c=500 chunks=4 len=102400 type='bytes' benchmarker='wrk'                 -0.62 %       ±2.63% ±3.50%  ±4.56%
 http/simple.js chunkedEnc=1 c=500 chunks=4 len=1024 type='buffer' benchmarker='wrk'                  -0.85 %       ±4.55% ±6.06%  ±7.88%
 http/simple.js chunkedEnc=1 c=500 chunks=4 len=1024 type='bytes' benchmarker='wrk'                   -3.76 %       ±4.87% ±6.48%  ±8.45%
 http/simple.js chunkedEnc=1 c=500 chunks=4 len=4 type='buffer' benchmarker='wrk'                      1.97 %       ±5.12% ±6.82%  ±8.87%
 http/simple.js chunkedEnc=1 c=500 chunks=4 len=4 type='bytes' benchmarker='wrk'                       2.05 %       ±4.80% ±6.39%  ±8.31%
 http/simple.js chunkedEnc=1 c=50 chunks=1 len=102400 type='buffer' benchmarker='wrk'                  1.59 %       ±6.32% ±8.42% ±10.96%
 http/simple.js chunkedEnc=1 c=50 chunks=1 len=102400 type='bytes' benchmarker='wrk'                   1.36 %       ±3.22% ±4.28%  ±5.57%
 http/simple.js chunkedEnc=1 c=50 chunks=1 len=1024 type='buffer' benchmarker='wrk'                   -1.28 %       ±3.73% ±4.96%  ±6.46%
 http/simple.js chunkedEnc=1 c=50 chunks=1 len=1024 type='bytes' benchmarker='wrk'                    -1.58 %       ±2.66% ±3.54%  ±4.61%
 http/simple.js chunkedEnc=1 c=50 chunks=1 len=4 type='buffer' benchmarker='wrk'                      -0.04 %       ±3.53% ±4.70%  ±6.12%
 http/simple.js chunkedEnc=1 c=50 chunks=1 len=4 type='bytes' benchmarker='wrk'                       -0.27 %       ±3.74% ±4.98%  ±6.48%
 http/simple.js chunkedEnc=1 c=50 chunks=4 len=102400 type='buffer' benchmarker='wrk'                  0.35 %       ±5.72% ±7.61%  ±9.91%
 http/simple.js chunkedEnc=1 c=50 chunks=4 len=102400 type='bytes' benchmarker='wrk'                   0.06 %       ±2.60% ±3.47%  ±4.51%
 http/simple.js chunkedEnc=1 c=50 chunks=4 len=1024 type='buffer' benchmarker='wrk'                   -1.18 %       ±2.71% ±3.60%  ±4.69%
 http/simple.js chunkedEnc=1 c=50 chunks=4 len=1024 type='bytes' benchmarker='wrk'                     2.16 %       ±2.54% ±3.38%  ±4.41%
 http/simple.js chunkedEnc=1 c=50 chunks=4 len=4 type='buffer' benchmarker='wrk'                      -2.10 %       ±2.41% ±3.21%  ±4.18%
 http/simple.js chunkedEnc=1 c=50 chunks=4 len=4 type='bytes' benchmarker='wrk'                       -2.06 %       ±2.74% ±3.64%  ±4.74%

AndreasMadsen

LGTM

Instead of having mostly duplicate code in form of internalNextTick, instead use the existing defaultAsyncTriggerIdScope with a slight modification which allows undefined triggerAsyncId to be passed in, which then just triggers the callback with the provided arguments.

apapirovski · 2018-03-08T12:08:48Z

Lite CI (after rebase): https://ci.nodejs.org/job/node-test-pull-request-lite/237/

apapirovski · 2018-03-08T12:24:29Z

Landed in 4d07434

Instead of having mostly duplicate code in form of internalNextTick, instead use the existing defaultAsyncTriggerIdScope with a slight modification which allows undefined triggerAsyncId to be passed in, which then just triggers the callback with the provided arguments. PR-URL: #19147 Refs: #19104 Reviewed-By: Anna Henningsen <anna@addaleax.net> Reviewed-By: Andreas Madsen <amwebdk@gmail.com> Reviewed-By: James M Snell <jasnell@gmail.com> Reviewed-By: Colin Ihrig <cjihrig@gmail.com> Reviewed-By: Ruben Bridgewater <ruben@bridgewater.de>

targos · 2018-03-17T13:48:29Z

Should this be backported to v9.x-staging? If yes please follow the guide and raise a backport PR, if not let me know or add the dont-land-on label.

AndreasMadsen · 2018-03-17T16:08:35Z

Yes, APM providers will appreciate if this were backported.

Instead of having mostly duplicate code in form of internalNextTick, instead use the existing defaultAsyncTriggerIdScope with a slight modification which allows undefined triggerAsyncId to be passed in, which then just triggers the callback with the provided arguments. PR-URL: nodejs#19147 Refs: nodejs#19104 Reviewed-By: Anna Henningsen <anna@addaleax.net> Reviewed-By: Andreas Madsen <amwebdk@gmail.com> Reviewed-By: James M Snell <jasnell@gmail.com> Reviewed-By: Colin Ihrig <cjihrig@gmail.com> Reviewed-By: Ruben Bridgewater <ruben@bridgewater.de>

jasnell · 2018-08-17T23:26:44Z

Should this be backported to 8.x? If so, we need a separate backport PR.

apapirovski added process Issues and PRs related to the process subsystem. async_hooks Issues and PRs related to the async hooks subsystem. labels Mar 5, 2018

nodejs-github-bot added the lib / src Issues and PRs related to general changes in the lib or src directory. label Mar 5, 2018

apapirovski commented Mar 5, 2018

View reviewed changes

jasnell requested review from AndreasMadsen, trevnorris and addaleax March 5, 2018 16:31

addaleax approved these changes Mar 5, 2018

View reviewed changes

AndreasMadsen reviewed Mar 5, 2018

View reviewed changes

apapirovski added a commit to apapirovski/node that referenced this pull request Mar 5, 2018

this commit will need to be rebased out once nodejs#19147 lands

6595cac

apapirovski mentioned this pull request Mar 5, 2018

process: small nextTick cleanup, including removing manual args copying #19153

Closed

3 tasks

AndreasMadsen approved these changes Mar 5, 2018

View reviewed changes

jasnell approved these changes Mar 6, 2018

View reviewed changes

cjihrig approved these changes Mar 6, 2018

View reviewed changes

BridgeAR approved these changes Mar 6, 2018

View reviewed changes

BridgeAR added the author ready PRs that have at least one approval, no pending requests for changes, and a CI started. label Mar 6, 2018

apapirovski force-pushed the patch-remove-internal-nexttick branch from 7958cea to 37f6ef0 Compare March 8, 2018 12:08

apapirovski closed this Mar 8, 2018

apapirovski deleted the patch-remove-internal-nexttick branch March 8, 2018 12:24

targos removed the author ready PRs that have at least one approval, no pending requests for changes, and a CI started. label Mar 17, 2018

targos added the backport-requested-v9.x label Mar 17, 2018

jasnell added the backport-requested-v8.x label Aug 17, 2018

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

async_hooks,process: remove internalNextTick #19147

async_hooks,process: remove internalNextTick #19147

apapirovski commented Mar 5, 2018

apapirovski Mar 5, 2018 •

edited

Loading

apapirovski commented Mar 5, 2018

AndreasMadsen Mar 5, 2018

apapirovski Mar 5, 2018

AndreasMadsen commented Mar 5, 2018 •

edited

Loading

apapirovski commented Mar 5, 2018 •

edited

Loading

apapirovski commented Mar 5, 2018

apapirovski commented Mar 5, 2018

AndreasMadsen left a comment

apapirovski commented Mar 8, 2018

apapirovski commented Mar 8, 2018

targos commented Mar 17, 2018

AndreasMadsen commented Mar 17, 2018

jasnell commented Aug 17, 2018

async_hooks,process: remove internalNextTick #19147

async_hooks,process: remove internalNextTick #19147

Conversation

apapirovski commented Mar 5, 2018

Checklist

Affected core subsystem(s)

apapirovski Mar 5, 2018 • edited Loading

Choose a reason for hiding this comment

apapirovski commented Mar 5, 2018

AndreasMadsen Mar 5, 2018

Choose a reason for hiding this comment

apapirovski Mar 5, 2018

Choose a reason for hiding this comment

AndreasMadsen commented Mar 5, 2018 • edited Loading

apapirovski commented Mar 5, 2018 • edited Loading

apapirovski commented Mar 5, 2018

apapirovski commented Mar 5, 2018

AndreasMadsen left a comment

Choose a reason for hiding this comment

apapirovski commented Mar 8, 2018

apapirovski commented Mar 8, 2018

targos commented Mar 17, 2018

AndreasMadsen commented Mar 17, 2018

jasnell commented Aug 17, 2018

apapirovski Mar 5, 2018 •

edited

Loading

AndreasMadsen commented Mar 5, 2018 •

edited

Loading

apapirovski commented Mar 5, 2018 •

edited

Loading