New issue
Advanced search Search tips
Note: Color blocks (like or ) mean that a user may not be available. Tooltip shows the reason.

Issue 831567 link

Starred by 1 user

Issue metadata

Status: Duplicate
Owner:
Closed: Apr 2018
Cc:
Components:
EstimatedDays: ----
NextAction: ----
OS: Linux , Android , Windows , Chrome , Mac , Fuchsia
Pri: 2
Type: Bug



Sign in to add a comment

SitePerProcess policy causes css:hover in parent window when traversing iFrame

Reported by apsche...@gmail.com, Apr 11 2018

Issue description

UserAgent: Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/65.0.3325.181 Safari/537.36

Steps to reproduce the problem:
1. Navigate to http://jsfiddle.net/nickleaf/tKsyw/1/ (not my fiddle, just found one that shows the problem)
2. From above the youtube iframe starting at top left, move mouse down into the iframe and back up out of the iframe (see attached image)
3. Do this from left to right across the top of the iframe

What is the expected behavior?
No change to the top navigation of the parent (fiddle) site

What went wrong?
Hover effects are shown on the top navigation (underlines) when the cursor is nowhere near the top navigation. This can be especially disruptive if there are dropdown navigations based on hover effects (top menus pop down unexpectedly)

Did this work before? N/A 

Chrome version: 65.0.3325.181  Channel: stable
OS Version: 10.0
Flash Version: 

Enabling and Disabling SitePerProcess policy will cause behavior to show or not show.
 
screenGrabBugReport.png
406 KB View Download
Cc: pastarmovj@chromium.org
Labels: Enterprise-Triaged
Owner: palmer@chromium.org
Cc: kkaluri@chromium.org
Labels: Needs-Feedback
Tested this issue on Debian Rodete with chrome #65.0.3325.181

Observations:
Even though Site-Per-Process flag is enabled or disabled, the issue is still exists in both scenarios.

Attaching the screen-cast for reference.

apschenck@ Could you please look into it and confirm this issue is related to site-per-process flag.


Thank You...

831567.mp4
2.8 MB View Download

Comment 3 by apsche...@gmail.com, Apr 12 2018

When I go to flip the flag on Site-Per-Process I notice it is disabled. However, when I go to chrome://policy I see SitePerProcess set to True. When I set the registry key to 0 the behavior goes away. When I set it to 1 the behavior returns. Changing these keys does not change the Site-Per-Process flag you show. https://www.chromium.org/administrators/policy-list-3#SitePerProcess 
Project Member

Comment 4 by sheriffbot@chromium.org, Apr 12 2018

Labels: -Needs-Feedback
Thank you for providing more feedback. Adding the requester to the cc list.

For more details visit https://www.chromium.org/issue-tracking/autotriage - Your friendly Sheriffbot

Comment 5 by palmer@chromium.org, Apr 12 2018

Cc: creis@chromium.org nasko@chromium.org
Components: Internals>Sandbox>SiteIsolation
Labels: OS-Android OS-Chrome OS-Fuchsia OS-Linux OS-Mac
Owner: ----
This is very unlikely to be related to the enterprise policy setting per se, but rather to the out-of-process iframe functionality (however you enable it). I assume it would also affect all platforms where we have Site Isolation.

Comment 6 by creis@chromium.org, Apr 12 2018

Cc: wjmaclean@chromium.org mcnee@chromium.org
Components: Blink>Input
Labels: FoundIn-65 Needs-Bisect
Owner: kenrb@chromium.org
Yes, this looks like a legitimate bug in Chrome 65, and I can repro it on ChromeOS 65.0.3325.209 and Windows 65.0.3325.181 with Site Isolation enabled.  Thanks for the report!

The bug does not repro for me on Windows 66.0.3359.66 or 67.0.3395.0, so I suspect it was fixed in M66.

Ken, James, or Kevin, could you check to see which CL fixed this? A bisect (with Site Isolation enabled) would probably help.

Comment 7 by nasko@chromium.org, Apr 12 2018

This sounds vaguely similar to issue 827229 where mouse enter/leave events are involved. It was fixed in M66, so if it does not repro there, I'm almost certain it is the root cause.

Comment 8 by creis@chromium.org, Apr 12 2018

Cc: kenrb@chromium.org
Owner: sadrul@chromium.org
Comment 7: Nice.  That would be r544744 by sadrul@.  We can close this if we confirm it.

Comment 9 by creis@chromium.org, Apr 12 2018

Labels: -Needs-Bisect
Mergedinto: 822905
Status: Duplicate (was: Unconfirmed)
Confirmed.   Reverting r544744 locally re-introduces the bug.  That means this is fixed in M66, which will likely be released next week.  Thanks Sadrul (and Nasko)!

Sign in to add a comment