http: don't emit end after aborted

[ronag]

See, #27981

• 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

[ronag]

I realized we don't need to call destroy since all it does is destroy the socket and since we are inside the socket's close listener it is already destroying...

[ronag]

Another option is to set readable to false on aborted.

[ronag]

Actually, I think it might be best if as much data as possible is read before aborted is emitted?

[ronag]

I kind of want something like this. But I'm not sure about the prependListener hack... is there a better way?

    req.emit('close');
    if (!res.complete) {
      // Socket closed before we emitted 'end' below.
      // Read as much as possible before emitting.
      res.prependListener('end', function() {
        res.off('end');
        this.emit('aborted');
        this.emit('close');
      });
      res.push(null);
    } else if (res.readable) {
      res.on('end', function() {
        this.emit('close');
      });
      res.push(null);
    } else {
      res.emit('close');
    }

[nodejs-github-bot]

CI: https://ci.nodejs.org/job/node-test-pull-request/23547/
CI: https://ci.nodejs.org/job/node-test-pull-request/23839/
CI: https://ci.nodejs.org/job/node-test-pull-request/23861/

[Trott]

Test failures in CI--http tests now timing out.

[ronag]

@Trott trying to fix but I'm no longer able to build master?

/Users/ronagy/GitHub/public/node/out/Release/obj.target/v8_snapshot/geni/embedded.cc:5:10: fatal error: 'src/snapshot/macros.h' file not found

[Trott]

@Trott trying to fix but I'm no longer able to build master?

/Users/ronagy/GitHub/public/node/out/Release/obj.target/v8_snapshot/geni/embedded.cc:5:10: fatal error: 'src/snapshot/macros.h' file not found

@nodejs/build

[lpinca]

It seems a spurious test.js file was commited.

[lpinca]

SMGT with the spurious test.js removed and commit title fixed.
I'm not sure if it should be marked as semver-major though.

[ronag]

what's wrong with the commit title?

[lpinca]

"Fix" should be "fix".

[Trott]

I ran locally against pumpify and it failed there too. I'm going to remove the "Request Changes" and bisect to find out when it broke....

broken CITGM stuff was already broken, will troubleshoot elsewhere....

[Trott]

(csv-parser isn't broken for me locally though.... once I'm done bisecting for pumpify, I'll apply this change to confirm it doesn't break csv-parser for me locally, out of extreme caution... )

[Trott]

CITGM on master shows pumpify and csv-parser issues. sigh

[Trott]

@Trott: I can try and help out. Is there a chat or something so we can communicate?

I'm often in nodeslackers and in Freenode IRC. Twitter DMs are pretty good for me too.

[ronag]

@Trott: How are we doing here?

[Trott]

CITGM: https://ci.nodejs.org/view/Node.js-citgm/job/citgm-smoker/2033/

Recent CITGM run on master for comparison: https://ci.nodejs.org/view/Node.js-citgm/job/citgm-smoker/2031/

[Trott]

CITGM results look worse than they are because AIX ran out of memory. If regular CI is green, I think this is good to land.

[Trott]

(I'd prefer some details in the commit message body but I wouldn't consider that blocking.)

[nodejs-github-bot]

CI: https://ci.nodejs.org/job/node-test-pull-request/26008/ ✔️

[Trott]

Landed in 5f80df8