Skip to content

net: fix ambiguity in EOF handling - #9066

Closed
indutny wants to merge 1 commit into
nodejs:masterfrom
indutny:fix/ambiguity-in-net
Closed

net: fix ambiguity in EOF handling#9066
indutny wants to merge 1 commit into
nodejs:masterfrom
indutny:fix/ambiguity-in-net

Conversation

@indutny

@indutnyindutny commented Oct 12, 2016•
edited
Loading

Copy link
Copy Markdown
Member
Checklist
  • make -j8 test (UNIX), or vcbuild test nosign (Windows) passes
  • tests and/or benchmarks are included
  • documentation is changed or added
  • commit message follows commit guidelines
Affected core subsystem(s)

net

Description of change

end MUST always be emitted beforeclose. However, if a handle
will invoke uv_close_cb immediately, or in the same JS tick - close
may be emitted first.

cc @nodejs/collaborators

@nodejs-github-botnodejs-github-bot added the net Issues and PRs related to the net subsystem. label Oct 12, 2016
@mcollina

mcollina commented Oct 12, 2016

Copy link
Copy Markdown
Member

Nice one. I imagine there is very little chance to write a unit test for this, right?

if not, LGTM.

@indutny

indutny commented Oct 12, 2016

Copy link
Copy Markdown
MemberAuthor

@mcollina I actually just got an idea on how to do this, but wanted to gather some preliminary feedback first.

@indutnyindutny mentioned this pull request Oct 12, 2016
Closed
@indutny

indutny commented Oct 12, 2016

Copy link
Copy Markdown
MemberAuthor

Added test.

@Trott

Trott commented Oct 12, 2016

Copy link
Copy Markdown
Member

@indutny Test not showing up here... (Forgot to push maybe? Or need to force-push maybe?)

@indutny
net: fix ambiguity in EOF handling
6e38e76
`end` MUST always be emitted **before** `close`. However, if a handle
will invoke `uv_close_cb` immediately, or in the same JS tick - `close`
may be emitted first.
@indutny
indutnyforce-pushed the fix/ambiguity-in-net branch from 9920f17 to 6e38e76CompareOctober 12, 2016 22:36
@indutny

indutny commented Oct 12, 2016

Copy link
Copy Markdown
MemberAuthor

@Trott sorry, just pushed it!

@indutny

indutny commented Oct 12, 2016

Copy link
Copy Markdown
MemberAuthor

CI: https://ci.nodejs.org/job/node-test-pull-request/4495/

@Trott

Trott commented Oct 12, 2016

Copy link
Copy Markdown
Member

Stress test to make sure there's no unexpected flakiness hidden in the test somewhere:

https://ci.nodejs.org/job/node-stress-single-test/994/

(Failures on the two win2008* jobs can be ignored, I think.)

@Trott

Trott commented Oct 12, 2016

Copy link
Copy Markdown
Member

LGTM if no CI oddities arise.

santigimeno
santigimeno reviewed Oct 12, 2016
View reviewed changes
Comment threadtest/parallel/test-net-end-close.js

process.on('exit', () => {
assert.deepStrictEqual(events, [ 'end', 'close' ]);
});

@santigimenosantigimenoOct 12, 2016

Copy link
Copy Markdown
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Would something like this work?

s.on('end',common.mustCall(()=>{s.on('close',common.mustCall(()=>{}));}));

... so you can get rid of events and the exit listener.

@indutnyindutnyOct 12, 2016

Copy link
Copy Markdown
MemberAuthor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

On one hand - yes, on the other - if this test will ever error, the error message will be more informative.

@mcollina

mcollina commented Oct 13, 2016

Copy link
Copy Markdown
Member

LGTM

jasnell
jasnell approved these changes Oct 13, 2016
View reviewed changes

@jasnelljasnell left a comment

Copy link
Copy Markdown
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

LGTM with green CI

imyller
imyller approved these changes Oct 13, 2016
View reviewed changes

@imyllerimyller left a comment

Copy link
Copy Markdown
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

LGTM

@indutny

indutny commented Oct 14, 2016

Copy link
Copy Markdown
MemberAuthor

Landed in 31196ea, thank you!

@indutnyindutny closed this Oct 14, 2016
@indutny
indutny deleted the fix/ambiguity-in-net branch October 14, 2016 13:40
indutny added a commit that referenced this pull request Oct 14, 2016
@indutny
net: fix ambiguity in EOF handling
31196ea
`end` MUST always be emitted **before** `close`. However, if a handle
will invoke `uv_close_cb` immediately, or in the same JS tick - `close`
may be emitted first.
PR-URL: #9066
Reviewed-By: Matteo Collina <matteo.collina@gmail.com>
Reviewed-By: Rich Trott <rtrott@gmail.com>
Reviewed-By: James M Snell <jasnell@gmail.com>
Reviewed-By: Colin Ihrig <cjihrig@gmail.com>
Reviewed-By: Ilkka Myller <ilkka.myller@nodefield.com>
jasnell pushed a commit that referenced this pull request Oct 14, 2016
@indutny@jasnell
net: fix ambiguity in EOF handling
6f1cae7
`end` MUST always be emitted **before** `close`. However, if a handle
will invoke `uv_close_cb` immediately, or in the same JS tick - `close`
may be emitted first.
PR-URL: #9066
Reviewed-By: Matteo Collina <matteo.collina@gmail.com>
Reviewed-By: Rich Trott <rtrott@gmail.com>
Reviewed-By: James M Snell <jasnell@gmail.com>
Reviewed-By: Colin Ihrig <cjihrig@gmail.com>
Reviewed-By: Ilkka Myller <ilkka.myller@nodefield.com>
@MylesBorins

MylesBorins commented Nov 11, 2016

Copy link
Copy Markdown
Contributor

@indutny this lands cleanly on v4 and v6. Is this something we should consider backporting?

@indutny

indutny commented Nov 11, 2016

Copy link
Copy Markdown
MemberAuthor

@thealphanerd yep

MylesBorins pushed a commit that referenced this pull request Nov 11, 2016
@indutnyMyles Borins
net: fix ambiguity in EOF handling
9a192a9
`end` MUST always be emitted **before** `close`. However, if a handle
will invoke `uv_close_cb` immediately, or in the same JS tick - `close`
may be emitted first.
PR-URL: #9066
Reviewed-By: Matteo Collina <matteo.collina@gmail.com>
Reviewed-By: Rich Trott <rtrott@gmail.com>
Reviewed-By: James M Snell <jasnell@gmail.com>
Reviewed-By: Colin Ihrig <cjihrig@gmail.com>
Reviewed-By: Ilkka Myller <ilkka.myller@nodefield.com>
MylesBorins pushed a commit that referenced this pull request Nov 11, 2016
@indutnyMyles Borins
net: fix ambiguity in EOF handling
0f1a22d
`end` MUST always be emitted **before** `close`. However, if a handle
will invoke `uv_close_cb` immediately, or in the same JS tick - `close`
may be emitted first.
PR-URL: #9066
Reviewed-By: Matteo Collina <matteo.collina@gmail.com>
Reviewed-By: Rich Trott <rtrott@gmail.com>
Reviewed-By: James M Snell <jasnell@gmail.com>
Reviewed-By: Colin Ihrig <cjihrig@gmail.com>
Reviewed-By: Ilkka Myller <ilkka.myller@nodefield.com>
This was referenced Nov 22, 2016
Merged
Merged
@alisccoaliscco mentioned this pull request Nov 29, 2023
Open
@leonardoadameleonardoadame mentioned this pull request Nov 29, 2023
Open
This was referenced Nov 30, 2023
Open
Open
Sign up for freeto join this conversation on GitHub. Already have an account? Sign in to comment

Reviewers

@santigimenosantigimenosantigimeno left review comments
@mcollinamcollinamcollina approved these changes
@jasnelljasnelljasnell approved these changes
@cjihrigcjihrigcjihrig approved these changes
+1 more reviewer
@imyllerimyllerimyller approved these changes
Reviewers whose approvals may not affect merge requirements

Assignees

No one assigned

Labels

netIssues and PRs related to the net subsystem.

Projects

None yet

Milestone

No milestone

Development

Successfully merging this pull request may close these issues.

9 participants

@indutny@mcollina@Trott@MylesBorins@jasnell@santigimeno@imyller@cjihrig@nodejs-github-bot