1390434 - check layout/base/tests/chrome/test_leaf_layers_partition_browser_window.xul

Status: RESOLVED WORKSFORME

(Core :: DOM: Security, defect, P1)

Description

[Yoshi Cheng-Hao Huang [:allstars.chh][:allstarschh][:yoshi]]

This test contains data: URI in it, and is disabled on linux and macos, however it always timeout on Windows,
bug 992311.

I've checked the timeout, the function leafLayersPartitionWindow() always fail 

http://searchfox.org/mozilla-central/rev/e5b13e6224dbe3182050cf442608c4cb6a8c5c55/layout/base/tests/chrome/test_leaf_layers_partition_browser_window.xul#48

Comment 1

[Christoph Kerschbaumer [:ckerschb]]

fwiw, this test did not show up as failing on the latest TRY run:
https://treeherder.mozilla.org/#/jobs?repo=try&revision=37554efc3fb036d426fed8b8aca58148fcb8a4da

Probably still worth rewriting to avoid any unexpected test failures.

Comment 2

[Christoph Kerschbaumer [:ckerschb]]

It seems the failure is back:
https://treeherder.mozilla.org/#/jobs?repo=try&revision=d9cfb1584d74b0f4afaae06a52076ecd83583434

Also, it seems the test is disabled on mac and linux:
> [test_leaf_layers_partition_browser_window.xul]
> skip-if = (!debug) || (toolkit == "cocoa") || (os == "linux") # Disabled on Mac and Linux because of Bug 992311

Probably we should just rewrite it.

Comment 3

[Christoph Kerschbaumer [:ckerschb]]

On the other hand it's a chrome mochitest, so it shouldn't fail because of any data: URIs. There is also this intermittent bug on file [1] which indicates that this test fails quite often (27 out of 901 runs).

[1] https://bugzilla.mozilla.org/show_bug.cgi?id=992311

Comment 4

[Christoph Kerschbaumer [:ckerschb]]

Attachment: bug_1390434_test_leaf_layers_partition.patch

This tests sometimes shows up as failing and sometimes not (please note it's only enabled on windows [1]). Anyway, I don't know why it's failing, it's a chrome test and I thought that SystemPrincipal always has access to data: windows. Anyway, let's rewrite that test as well to not run into any issues.

[1] https://treeherder.mozilla.org/#/jobs?repo=try&revision=7ca5a6b268ccc2559cf5bf0246965631caf05e84

Comment 5

[Olli Pettay [:smaug][[e-mail]]]

Comment on attachment 8899268 [details] [diff] [review]
bug_1390434_test_leaf_layers_partition.patch

We can try this

Comment 6

[Christoph Kerschbaumer [:ckerschb]]

It seems it's not fixing the issue, in fact a11y on Linux x64 is now also failing because of that, see:
https://treeherder.mozilla.org/#/jobs?repo=try&revision=21dc6a91ebc407326ecf42c1cd77ba7d0dcddb54

I can't reproduce locally, probably worth trying on a windows machine.

Comment 7

[Henry Chang [:hchang]]

(In reply to Christoph Kerschbaumer [:ckerschb] from comment #6)
> It seems it's not fixing the issue, in fact a11y on Linux x64 is now also

What failed in a11y on Linux x64 is 'test_zoom_tree.xul' but not 
'test_leaf_layers_partition_browser_window.xul' so we probably still
only need to focus on windows.

https://treeherder-manifest.herokuapp.com/?repo=try&revision=21dc6a91ebc407326ecf42c1cd77ba7d0dcddb54

> failing because of that, see:
> https://treeherder.mozilla.org/#/
> jobs?repo=try&revision=21dc6a91ebc407326ecf42c1cd77ba7d0dcddb54
> 
> I can't reproduce locally, probably worth trying on a windows machine.

Comment 8

[Christoph Kerschbaumer [:ckerschb]]

(In reply to Henry Chang [:hchang] from comment #7)
> 
> What failed in a11y on Linux x64 is 'test_zoom_tree.xul' but not 
> 'test_leaf_layers_partition_browser_window.xul' so we probably still
> only need to focus on windows.

Indeed, I guess I didn't have my coffee this morning before I looked at TRY results - thanks.

Comment 9

[Christoph Kerschbaumer [:ckerschb]]

I retriggered a11y a couple of times, turned green on subsequent runs:
https://treeherder.mozilla.org/#/jobs?repo=try&revision=21dc6a91ebc407326ecf42c1cd77ba7d0dcddb54

So this bug should only focus on test_leaf_layers_partition_browser_window.xul on Windows 7 debug (c2).

Comment 10

[Henry Chang [:hchang]]

It seems now a permanfail:

https://treeherder.mozilla.org/#/jobs?repo=try&revision=4894d45b740a8acc1d05a733708e105c094db67b

so resolve the bug with WORKSFORME

Comment 11

[Henry Chang [:hchang]]

(In reply to Henry Chang [:hchang] from comment #10)
> It seems now a permanfail:

To be more clear: it's now a permanent failure even without the pref being flipped
so the "opaque data:uri origin" has nothing to do with this permanfail.

> 
> https://treeherder.mozilla.org/#/
> jobs?repo=try&revision=4894d45b740a8acc1d05a733708e105c094db67b
> 
> so resolve the bug with WORKSFORME