[v6.x backport] tls: fix writeQueueSize prop, long write timeouts by apapirovski · Pull Request #16420 · nodejs/node

apapirovski · 2017-10-23T22:41:08Z

This is a backport of #15791 as requested.

/cc @MylesBorins

Checklist

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

Affected core subsystem(s)

net, tls, test

MylesBorins · 2017-10-23T22:59:38Z

@apapirovski a bit confused... doesn't seem like I requested this backport in the issue. Did we cross wires?

apapirovski · 2017-10-23T23:01:34Z

@MylesBorins You did in our conversation earlier. The name of the issue might just be a tad confusing. See #13391

Either way, even if you hadn't, this should be back-ported given that it's a serious bug. But if you don't want this back-ported for v4.x let me know as I'm almost done.

MylesBorins · 2017-10-23T23:07:10Z

@apapirovski AHHHHHH... thanks ☺️

This is slated to land in 8.8.0. Which means that we can likely land it in the following 6.x release, and a future 4.x maintenance release (assuming we get signoff on the backport)

apapirovski · 2017-10-24T22:41:21Z

CI: https://ci.nodejs.org/job/node-test-pull-request/10939/

apapirovski · 2017-10-25T12:38:00Z

CI is all green (failure unrelated) and I've rebased.

apapirovski · 2017-10-25T12:53:04Z

This is blocked because of #16484 — working on a PR which will then need to be applied here.

apapirovski · 2017-10-25T19:50:42Z

Last commit is technically from a different PR but this shouldn't land without it.

CI: https://ci.nodejs.org/job/node-test-pull-request/10978/

mcollina · 2017-10-26T07:34:03Z

I think this should wait for some more time before backporting (if at all).

Make writeQueueSize represent the actual size of the write queue within the TLS socket. Add tls test to confirm that bufferSize works as expected. PR-URL: nodejs#15791 Fixes: nodejs#15005 Refs: nodejs#15006 Reviewed-By: Anna Henningsen <anna@addaleax.net> Reviewed-By: Matteo Collina <matteo.collina@gmail.com> Reviewed-By: Refael Ackermann <refack@gmail.com> Reviewed-By: Fedor Indutny <fedor.indutny@gmail.com>

Add updateWriteQueueSize which updates and returns queue size (net & tls). Make _onTimeout check whether an active write is ongoing and if so, call _unrefTimer rather than emitting a timeout event. Add http & https test that checks whether long-lasting (but active) writes timeout or can finish writing as expected. PR-URL: nodejs#15791 Fixes: nodejs#15082 Reviewed-By: Anna Henningsen <anna@addaleax.net> Reviewed-By: Matteo Collina <matteo.collina@gmail.com> Reviewed-By: Refael Ackermann <refack@gmail.com> Reviewed-By: Fedor Indutny <fedor.indutny@gmail.com>

This commit handles the case where _onTimeout is called with a null handle. Refs: nodejs#15791 Fixes: nodejs#16484 PR-URL: nodejs#16489 Reviewed-By: Anna Henningsen <anna@addaleax.net> Reviewed-By: James M Snell <jasnell@gmail.com> Reviewed-By: Colin Ihrig <cjihrig@gmail.com> Reviewed-By: Refael Ackermann <refack@gmail.com>

MylesBorins · 2017-12-19T18:46:16Z

@mcollina is this ready to land?

mcollina · 2017-12-20T10:39:31Z

@MylesBorins no. This PR introduced #16484 which was fixed in #16489.
If they are backported, they are backported together.
What's the process for this? I'd be inclined to have two commits in this same PR.

apapirovski · 2017-12-20T11:49:31Z

@mcollina the commit from that second PR is in here

mcollina

LGTM

Make writeQueueSize represent the actual size of the write queue within the TLS socket. Add tls test to confirm that bufferSize works as expected. Backport-PR-URL: #16420 PR-URL: #15791 Fixes: #15005 Refs: #15006 Reviewed-By: Anna Henningsen <anna@addaleax.net> Reviewed-By: Matteo Collina <matteo.collina@gmail.com> Reviewed-By: Refael Ackermann <refack@gmail.com> Reviewed-By: Fedor Indutny <fedor.indutny@gmail.com>

Add updateWriteQueueSize which updates and returns queue size (net & tls). Make _onTimeout check whether an active write is ongoing and if so, call _unrefTimer rather than emitting a timeout event. Add http & https test that checks whether long-lasting (but active) writes timeout or can finish writing as expected. Backport-PR-URL: #16420 PR-URL: #15791 Fixes: #15082 Reviewed-By: Anna Henningsen <anna@addaleax.net> Reviewed-By: Matteo Collina <matteo.collina@gmail.com> Reviewed-By: Refael Ackermann <refack@gmail.com> Reviewed-By: Fedor Indutny <fedor.indutny@gmail.com>

This commit handles the case where _onTimeout is called with a null handle. Backport-PR-URL: #16420 Refs: #15791 Fixes: #16484 PR-URL: #16489 Reviewed-By: Anna Henningsen <anna@addaleax.net> Reviewed-By: James M Snell <jasnell@gmail.com> Reviewed-By: Colin Ihrig <cjihrig@gmail.com> Reviewed-By: Refael Ackermann <refack@gmail.com>

MylesBorins · 2017-12-20T17:01:53Z

landed in 612baca...15d855d

Make writeQueueSize represent the actual size of the write queue within the TLS socket. Add tls test to confirm that bufferSize works as expected. Backport-PR-URL: #16420 PR-URL: #15791 Fixes: #15005 Refs: #15006 Reviewed-By: Anna Henningsen <anna@addaleax.net> Reviewed-By: Matteo Collina <matteo.collina@gmail.com> Reviewed-By: Refael Ackermann <refack@gmail.com> Reviewed-By: Fedor Indutny <fedor.indutny@gmail.com>

Add updateWriteQueueSize which updates and returns queue size (net & tls). Make _onTimeout check whether an active write is ongoing and if so, call _unrefTimer rather than emitting a timeout event. Add http & https test that checks whether long-lasting (but active) writes timeout or can finish writing as expected. Backport-PR-URL: #16420 PR-URL: #15791 Fixes: #15082 Reviewed-By: Anna Henningsen <anna@addaleax.net> Reviewed-By: Matteo Collina <matteo.collina@gmail.com> Reviewed-By: Refael Ackermann <refack@gmail.com> Reviewed-By: Fedor Indutny <fedor.indutny@gmail.com>

This commit handles the case where _onTimeout is called with a null handle. Backport-PR-URL: #16420 Refs: #15791 Fixes: #16484 PR-URL: #16489 Reviewed-By: Anna Henningsen <anna@addaleax.net> Reviewed-By: James M Snell <jasnell@gmail.com> Reviewed-By: Colin Ihrig <cjihrig@gmail.com> Reviewed-By: Refael Ackermann <refack@gmail.com>

apapirovski added net Issues and PRs related to the net subsystem. tls Issues and PRs related to the tls subsystem. labels Oct 23, 2017

nodejs-github-bot added build Issues and PRs related to build files or the CI. doc Issues and PRs related to the documentations. meta Issues and PRs related to the general management of the project. tools Issues and PRs related to the tools directory. labels Oct 23, 2017

apapirovski changed the base branch from master to v6.x-staging October 23, 2017 22:41

apapirovski removed build Issues and PRs related to build files or the CI. doc Issues and PRs related to the documentations. meta Issues and PRs related to the general management of the project. tools Issues and PRs related to the tools directory. labels Oct 23, 2017

apapirovski added c++ Issues and PRs that require attention from people who are familiar with C++. lib / src Issues and PRs related to general changes in the lib or src directory. v6.x labels Oct 23, 2017

apapirovski requested review from addaleax, mcollina and refack October 24, 2017 19:10

refack approved these changes Oct 24, 2017

View reviewed changes

MylesBorins changed the base branch from v6.x-staging to v6.x October 25, 2017 08:08

MylesBorins changed the base branch from v6.x to v6.x-staging October 25, 2017 08:08

MylesBorins force-pushed the v6.x-staging branch from 54286db to 3e96c85 Compare October 25, 2017 08:09

apapirovski force-pushed the backport-15791-to-v6.x branch from b6ed49b to 7feac5c Compare October 25, 2017 12:35

apapirovski added the blocked PRs that are blocked by other issues or PRs. label Oct 25, 2017

apapirovski removed the blocked PRs that are blocked by other issues or PRs. label Oct 25, 2017

MylesBorins force-pushed the v6.x-staging branch from 9219283 to b0fadbe Compare October 26, 2017 04:22

mcollina added the baking-for-lts PRs that need to wait before landing in a LTS release. label Oct 26, 2017

MylesBorins force-pushed the v6.x-staging branch from fa922b9 to 50c3dab Compare November 28, 2017 04:15

apapirovski added 3 commits December 12, 2017 16:54

apapirovski force-pushed the backport-15791-to-v6.x branch from 67ca82f to 920f2ff Compare December 12, 2017 21:54

mcollina approved these changes Dec 20, 2017

View reviewed changes

MylesBorins closed this Dec 20, 2017

apapirovski deleted the backport-15791-to-v6.x branch January 2, 2018 16:26

MylesBorins removed the baking-for-lts PRs that need to wait before landing in a LTS release. label Aug 17, 2018

This was referenced May 13, 2024

[Snyk] Security upgrade eslint from 8.22.0 to 8.26.0 leonardoadame/node#818

Open

[Snyk] Fix for 2 vulnerabilities leonardoadame/node#839

Open

Uh oh!

Conversation

apapirovski commented Oct 23, 2017

Checklist

Affected core subsystem(s)

Uh oh!

MylesBorins commented Oct 23, 2017

Uh oh!

apapirovski commented Oct 23, 2017 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

MylesBorins commented Oct 23, 2017

Uh oh!

apapirovski commented Oct 24, 2017

Uh oh!

apapirovski commented Oct 25, 2017

Uh oh!

apapirovski commented Oct 25, 2017

Uh oh!

apapirovski commented Oct 25, 2017

Uh oh!

mcollina commented Oct 26, 2017

Uh oh!

MylesBorins commented Dec 19, 2017

Uh oh!

mcollina commented Dec 20, 2017

Uh oh!

apapirovski commented Dec 20, 2017

Uh oh!

mcollina left a comment

Choose a reason for hiding this comment

Uh oh!

MylesBorins commented Dec 20, 2017

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

5 participants

apapirovski commented Oct 23, 2017 •

edited

Loading