test: fix flaky test-watch-file

[Trott]

Example of the failure this will avoid is https://ci.nodejs.org/job/node-test-commit-linux/36156/nodes=alpine-latest-x64/testReport/junit/(root)/test/pummel_test_watch_file/:

assert.js:103
  throw new AssertionError(obj);
  ^

AssertionError [ERR_ASSERTION]: Expected "actual" not to be strictly deep-equal to:

2020-07-17T16:00:38.099Z
    at StatWatcher.<anonymous> (/home/iojs/build/workspace/node-test-commit-linux/nodes/alpine-latest-x64/test/pummel/test-watch-file.js:38:12)
    at StatWatcher.emit (events.js:314:20)
    at StatWatcher.onchange (internal/fs/watchers.js:63:8) {
  generatedMessage: true,
  code: 'ERR_ASSERTION',
  actual: 2020-07-17T16:00:38.099Z,
  expected: 2020-07-17T16:00:38.099Z,
  operator: 'notDeepStrictEqual'
}

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

[Trott]

Stress test that shows the failures on master: https://ci.nodejs.org/job/node-stress-single-test/122/

Stress test with this PR showing the test passing: https://ci.nodejs.org/job/node-stress-single-test/137 ✅

[Trott]

@nodejs/testing

[addaleax]

Just for my understanding, this would never really be a strictly-less-than relation, and only ever larger or equal, but if we check for equality we might as well also check for less-than?

[Trott]

For the most part, yes. It's a coherence check. Super unlikely edge case: It is possible for the time on a host to be adjusted backwards (to correct for it having been set incorrectly in the first place, for example) but that would be highly unusual of course.

[Trott]

Another stress CI with another fixup commit: https://ci.nodejs.org/job/node-stress-single-test/139/ ✅

[Trott]

@nodejs/fs

[MylesBorins]

RSLGTM

[Trott]

Landed in b0b52b2