Closed Bug 1237807 Opened 5 years ago Closed 4 years ago

Intermittent e10s test_websocket.html | Test timed out.

Categories

(Core :: DOM: Workers, defect)

defect
Not set
normal

Tracking

()

RESOLVED FIXED
mozilla47
Tracking Status
e10s + ---
firefox47 --- fixed

People

(Reporter: KWierso, Assigned: baku)

References

(Blocks 1 open bug)

Details

(Keywords: intermittent-failure)

Attachments

(5 files)

baku, any thoughts?
Flags: needinfo?(amarchesini)
It happened just once. And that is a big test, maybe something else was slowing down the mochitests.
I want to see some of these before working on it.
Flags: needinfo?(amarchesini)
Status: NEW → RESOLVED
Closed: 5 years ago
Resolution: --- → DUPLICATE
Duplicate of bug: 1237470
Intermittent failures on opt builds really can't be a duplicate of the bug which disabled the test on debug builds, and baku isn't likely to see any more of them if it's resolved as a duplicate, and thus hidden away in the junk suggestions.

https://treeherder.mozilla.org/logviewer.html#?job_id=3720144&repo=b2g-inbound

There, there's another.
Status: RESOLVED → REOPENED
Resolution: DUPLICATE → ---
Summary: Intermittent test_websocket.html | Test timed out. → Intermittent e10s test_websocket.html | Test timed out.
Version: 45 Branch → Trunk
Blocks: e10s-tests
tracking-e10s: --- → +
Looks like OSX is also affected by this bug. Can we try giving this the rr chaos-mode treatment and see if anything turns up?
Flags: needinfo?(amarchesini)
Also leaks when it fails on OSX:
TEST-UNEXPECTED-FAIL | leakcheck | tab process: 6760 bytes leaked (ChannelEvent, ChannelEventQueue, ListenerAndContextContainer, LoadInfo, Mutex, ...)
See Also: → 1244558
See Also: → 1239237
Assignee: nobody → amarchesini
Flags: needinfo?(amarchesini)
Attached patch ws1.patchSplinter Review
The current test_websocket.html is very complex: it contains 47 tests, often running together. What I propose here is to split that test in multiple mochitests (10 tests per mochitest), using a promise chain to manage them.
Attachment #8726762 - Flags: review?(ehsan)
Attached patch ws2.patchSplinter Review
Attachment #8726764 - Flags: review?(ehsan)
Attached patch ws3.patchSplinter Review
Attachment #8726765 - Flags: review?(ehsan)
Attached patch ws4.patchSplinter Review
Attachment #8726766 - Flags: review?(ehsan)
Attached patch ws5.patchSplinter Review
Attachment #8726767 - Flags: review?(ehsan)
Attachment #8726762 - Flags: review?(ehsan) → review+
Attachment #8726764 - Flags: review?(ehsan) → review+
Attachment #8726765 - Flags: review?(ehsan) → review+
Attachment #8726766 - Flags: review?(ehsan) → review+
Attachment #8726767 - Flags: review?(ehsan) → review+
Comment on attachment 8726762 [details] [diff] [review]
ws1.patch

Review of attachment 8726762 [details] [diff] [review]:
-----------------------------------------------------------------

::: dom/base/test/websocket_helpers.js
@@ +33,5 @@
> +      ok(false, "onerror called on test " + e.target._testNumber + "!");
> +    }
> +
> +    ws.addEventListener("close", function(e) {
> +      ws._receivedCloseEvent = true;

Seems like this flag is not being used after your patches any more...
You need to log in before you can comment on or make changes to this bug.