1527126 - Land crashtest for bug 1517025

Status: RESOLVED FIXED

(Core :: Networking, enhancement, P5)

Description

[Valentin Gosu [:valentin] (he/him)]

See https://phabricator.services.mozilla.com//D16817

The crashtest triggers some assertions on OSX
[Child 1950, Main Thread] ###!!! ASSERTION: bad inline size: 'reflowOutput.ISize(lineWM) >= 0', file /builds/worker/workspace/build/src/layout/generic/nsLineLayout.cpp, line 962
https://treeherder.mozilla.org/logviewer.html#/jobs?job_id=223031965&repo=autoland&lineNumber=24354

Comment 1

[Valentin Gosu [:valentin] (he/him)]

Attachment: Bug 1527126 - Add crashtest with invalid URL r=kershaw!

Comment 2

[Valentin Gosu [:valentin] (he/him)]

It would seem there are still assertions here.
https://treeherder.mozilla.org/#/jobs?repo=try&revision=36707cd9601f8488063caf7df1e93a9c29af0908&selectedTaskRun=RyHHgzjdQGOxfljxGXuyFg.0

The unit tests I landed for bug 1517025 are enough to prevent regressions of the security bug.
We have this problem in a bunch of other tests (see bug 1305241).
Maybe we should just remove the assertions at some point? Or are they actually useful?

Comment 3

[Emilio Cobos Álvarez (:emilio)]

So those assertions are useful in the sense that they should only fire with degenerate test-case. I'm very confused about how your test-case could trigger it, they likely come from another crashtest? It has nsTableWrapperFrame in the stack and there's no table in your test-case at all... However that shouldn't stop you from landing it, just add an asserts(0-1) annotation to the crashtest list and file a bug so that we can investigate?

Comment 4

[Valentin Gosu [:valentin] (he/him)]

Thanks for the suggestion. Really useful.

Comment 5

[Pulsebot]

Pushed by [e-mail]:
https://hg.mozilla.org/integration/autoland/rev/c872928d14fb
Add crashtest with invalid URL r=kershaw

Comment 6

[Dorel Luca [:dluca]]

https://hg.mozilla.org/mozilla-central/rev/c872928d14fb