Fix failing Core:ImageLib mochitests with cross-origin and Fission enabled
Categories
(Core :: Graphics: ImageLib, defect, P3)
Tracking
()
Tracking | Status | |
---|---|---|
firefox84 | --- | fixed |
People
(Reporter: neha, Assigned: jesup)
References
Details
Attachments
(1 file)
These tests are marked as skipped/failed for cross-origin and Fission, and need to be fixed
image/test/mochitest/test_bug1325080.html
More info: https://wiki.mozilla.org/Project_Fission/Enabling_Tests_with_Fission#Cross-Origin_Mochitests
Reporter | ||
Updated•4 years ago
|
Reporter | ||
Updated•4 years ago
|
Comment 1•4 years ago
|
||
According to the Fission timeline: https://mana.mozilla.org/wiki/pages/viewpage.action?spaceKey=FIREFOX&title=Fission it looks like M6b's timeline around October, so we should get this on our radar for the next couple of months.
Updated•4 years ago
|
Reporter | ||
Comment 2•4 years ago
|
||
The mana page has been updated for more accurate Fission timelines. M6b should be targeting Fx83.
Andrew/Timothy, can you please review this test failure for Fission and inform us if it is a broken test or a broken feature? Thanks!
Comment 3•4 years ago
|
||
From a quick look I would guess that it is the test needs to be modified. I think it would be easiest for someone who works on xorigin stuff to investigate (it might be trivial for them to spot the problem). I don't think there is anything going wrong in imagelib here.
Reporter | ||
Comment 4•4 years ago
|
||
Kashav, is there any guidance/help we can provide here to get these fixed?
Looks like we throw an NS_NOINTERFACE
when document.open()
is called from an out-of-process frame:
JavaScript error: http://mochi.test:8888/tests/image/test/mochitest/test_bug1325080.html?currentTestURL=image%2Ftest%2Fmochitest%2Ftest_bug1325080.html&closeWhenDone=1&showTestReport=true&expected=pass, line 22: NS_NOINTERFACE:
This is coming from the call to nsIDocShell::UpdateURLAndHistory
(and more specifically, the call to nsDocShell::AddToSessionHistory
), so this will probably get fixed by session history in the parent.
I might be mistaken, but it seems like document.write()
is part of what we're testing here, so I'm not sure if we can rewrite the test to not use that. We might need to block this on session history work.
Updated•4 years ago
|
Reporter | ||
Updated•4 years ago
|
Reporter | ||
Updated•4 years ago
|
Assignee | ||
Comment 6•4 years ago
|
||
Passes with --enable-fission --enable-xorigin
Assignee | ||
Comment 7•4 years ago
|
||
Updated•4 years ago
|
Comment 9•4 years ago
|
||
bugherder |
Description
•