Closed Bug 1368927 Opened 7 years ago Closed 7 years ago

Intermittent browser/extensions/pdfjs/test/browser_pdfjs_zoom.js | A promise chain failed to handle a rejection: Transport destroyed - stack: null

Categories

(Firefox :: PDF Viewer, defect)

defect
Not set
normal

Tracking

()

RESOLVED INCOMPLETE
Tracking Status
firefox56 --- affected
firefox57 --- affected

People

(Reporter: intermittent-bug-filer, Unassigned)

References

Details

(Keywords: bulk-close-intermittents, intermittent-failure, Whiteboard: [stockwell disabled])

this has failed 40 times since Sunday, primarily on windows 7 pgo (non e10s)! We will be shutting off our tests which are run in parallel for e10s/non-e10s and only keeping the e10s versions around next week- so this will magically fix itself. :bdahl, do you have any concerns with this test failing?
Flags: needinfo?(bdahl)
Whiteboard: [stockwell needswork]
Summary: Intermittent browser/extensions/pdfjs/test/browser_pdfjs_zoom.js | A promise chain failed to handle a rejection: Transport destroyed - rejection date: Tue May 30 2017 21:19:07 GMT-0700 (PDT) - stack: null - false == true - JS frame :: resource://testing → Intermittent browser/extensions/pdfjs/test/browser_pdfjs_zoom.js | A promise chain failed to handle a rejection: Transport destroyed - stack: null
(In reply to Joel Maher ( :jmaher) (UTC-9) from comment #7) > this has failed 40 times since Sunday, primarily on windows 7 pgo (non > e10s)! We will be shutting off our tests which are run in parallel for > e10s/non-e10s and only keeping the e10s versions around next week- so this > will magically fix itself. > > :bdahl, do you have any concerns with this test failing? I'm not that concerned. All the tests pass, but during the PDFViewerApplication.close a promise appears to be rejected that we don't handle. Calling the PDFViewerApplication.close is not typically done in the regular viewer.
Flags: needinfo?(bdahl)
This is extremely frequent on Beta right now. All on Windows and macOS. Almost all in non-e10s mode (which I guess explains why the frequency is lower on trunk - we only run e10s tests there). Brendan, can you please take a look?
Flags: needinfo?(bdahl)
See my comment above, but this is also hopefully addressed by https://github.com/mozilla/pdf.js/pull/8746. We could try to uplift, but seeing as it doesn't appear to really be a bug in the regular viewer I'm not sure we should bother. @ryan, looks like we just missed that pull request in bug 1387412 by a few hours, so it should apply pretty cleanly if we do want to uplift.
Flags: needinfo?(bdahl)
Indeed, we did just miss it. I'll run it through Try on top of Beta and we can go from there. It's frequent enough that I would indeed like to get something landed on Beta to make it go away.
Whiteboard: [stockwell needswork] → [stockwell disabled]
Depends on: 1389211
Status: NEW → RESOLVED
Closed: 7 years ago
Resolution: --- → INCOMPLETE
You need to log in before you can comment on or make changes to this bug.