Closed Bug 1313972 Opened 8 years ago Closed 8 years ago

Intermittent dom/base/test/test_intersectionobservers.html | rootBounds should is set to null for cross-origin observations [observe subframe]

Categories

(Core :: DOM: Core & HTML, defect)

defect
Not set
normal

Tracking

()

RESOLVED FIXED
mozilla55
Tracking Status
firefox52 --- unaffected
firefox-esr52 --- unaffected
firefox53 --- unaffected
firefox54 --- fixed
firefox55 --- fixed

People

(Reporter: intermittent-bug-filer, Assigned: tschneider)

References

Details

(Keywords: intermittent-failure, Whiteboard: [stockwell fixed])

Attachments

(3 files, 3 obsolete files)

New test, added in bug 1243846.
Flags: needinfo?(tschneider)
See Also: → 1313970, 1313927
No recent failures.
Flags: needinfo?(tschneider)
This is back again now that bug 1321865 has re-landed.
Flags: needinfo?(tschneider)
Back with a vengeance. I would say that this needs to be fixed before we can consider uplifting to Aurora.
Fixing intermittent failures due to a racing condition.
Flags: needinfo?(tschneider)
Keywords: checkin-needed
Pushed by ryanvm@gmail.com: https://hg.mozilla.org/integration/mozilla-inbound/rev/45d3d0f2d62a Intermittent dom/base/test/test_intersectionobservers.html | rootBounds should is set to null for cross-origin observations [observe subframe]. r=me
Keywords: checkin-needed
seems that doesn;t fix this - https://treeherder.mozilla.org/logviewer.html#?job_id=89177187&repo=mozilla-inbound after landing of this patch :(
Flags: needinfo?(tschneider)
Status: NEW → RESOLVED
Closed: 8 years ago
Resolution: --- → FIXED
Target Milestone: --- → mozilla55
Bah, leave-open should have been set on this given comment 19.
Assignee: nobody → tschneider
Status: RESOLVED → REOPENED
Resolution: FIXED → ---
Target Milestone: mozilla55 → ---
See Also: → 1324135
with 126 failures in 1 day, I would really like to see this fixed ASAP- In fact this is a new test turned on, so I am going to back this out until we get this looked at. In bug 1353461 where we turned this on there is no mention of a try push or other data to indicate that this was even tested.
Pushed by jmaher@mozilla.com: https://hg.mozilla.org/integration/mozilla-inbound/rev/35af33f46d94 backout bug 1321865 for extremely high test failure rate. r=me
Whiteboard: [stockwell disabled]
I was finally able to reproduce this reliably. The problem was that all our iframe tests load the same iframe content, which calls postMessage on onLoad. Not all of those tests consume the postMessage event tho. That leads to a racing conditions where the test that actually does consume the event receives messages from previous tests.
Flags: needinfo?(tschneider)
That will happen if we re-land bug 1321865.
Keywords: checkin-needed
Comment on attachment 8857572 [details] [diff] [review] Avoid racing condition by making sure postMessage events are consumed by the correct tests. LGTM. Thx!
Attachment #8857572 - Flags: review+
Pushed by ryanvm@gmail.com: https://hg.mozilla.org/integration/mozilla-inbound/rev/05af3a08fb6f Avoid racing condition by making sure postMessage events are consumed by the correct tests. r=smaug
Keywords: checkin-needed
Status: REOPENED → RESOLVED
Closed: 8 years ago8 years ago
Resolution: --- → FIXED
Target Milestone: --- → mozilla55
Status: RESOLVED → REOPENED
Resolution: FIXED → ---
We forgot to list intersectionobserver_cross_domain_iframe.html under the support-files section in mochitest.ini leading to timeout issues when the iframe can't be loaded (not on all platforms tho...).
Should also fix a typo while on it.
Attachment #8858432 - Attachment is obsolete: true
Attachment #8858433 - Flags: review?(bugs)
Attachment #8858433 - Flags: review?(bugs) → review+
Pushed by philringnalda@gmail.com: https://hg.mozilla.org/integration/mozilla-inbound/rev/0e09a3206a83 List intersectionobserver_cross_domain_iframe.html as support file in mochitest.ini, r=jet
Keywords: checkin-needed
Status: REOPENED → RESOLVED
Closed: 8 years ago8 years ago
Resolution: --- → FIXED
Comment on attachment 8857572 [details] [diff] [review] Avoid racing condition by making sure postMessage events are consumed by the correct tests. Approval Request Comment [Feature/Bug causing the regression]: 1321865 [User impact if declined]: No test coverage [Is this code covered by automated tests?]: It is one [Has the fix been verified in Nightly?]: Yes [Is the change risky?]: No
Attachment #8857572 - Flags: approval-mozilla-beta?
Comment on attachment 8857572 [details] [diff] [review] Avoid racing condition by making sure postMessage events are consumed by the correct tests. not part of the build
Attachment #8857572 - Flags: approval-mozilla-beta?
Attached patch Patch for beta uplift (obsolete) — Splinter Review
Blocks: 1362168
Whiteboard: [stockwell disabled] → [stockwell disabled] [checkin-needed-beta]
Flags: in-testsuite+
Whiteboard: [stockwell disabled] [checkin-needed-beta] → [stockwell fixed]
Had an issue when merging back the try-only changes from https://treeherder.mozilla.org/#/jobs?repo=try&revision=95acba62d571c397d08bed74cf963d9f700f9e7e. Should be fixed now.
Attachment #8864625 - Attachment is obsolete: true
Component: DOM → DOM: Core & HTML
You need to log in before you can comment on or make changes to this bug.

Attachment

General

Created:
Updated:
Size: