Closed Bug 618778 Opened 14 years ago Closed 7 years ago

Unwrapping a chrome window created by content has wrapped properties

Categories

(Core :: XPConnect, defect)

defect
Not set
normal

Tracking

()

RESOLVED INACTIVE

People

(Reporter: whimboo, Assigned: gal)

References

Details

(Keywords: regression, Whiteboard: [mozmill])

Attachments

(3 files)

Attached patch testcase (patch)Splinter Review
Mozilla/5.0 (Macintosh; Intel Mac OS X 10.6; rv:2.0b8pre) Gecko/20101204 Firefox/4.0b8pre ID:20101204030328 In one of our shared modules for Mozmill which is handling modal dialogs, I can see that after unwrapping a chrome window, which has been opened by content, some window properties are still wrapped. Steps: 1. Install Mozmill (https://developer.mozilla.org/en/Mozmill#Installation) 2. Get our Mozmill tests from http://hg.mozilla.org/qa/mozmill-tests/ 3. Patch the repository with the testcase attached to this bug 4. Run Mozmill with (-t firefox/testSecurity/testEncryptedPageWarning.js) Check the dump messages in the console: *** window before: [object ChromeWindow] *** window after: [object ChromeWindow] *** opener before: [object XrayWrapper [object Window]]
blocking2.0: --- → ?
Assignee: nobody → mrbkap
blocking2.0: ? → betaN+
Keywords: regression
Whiteboard: [mozmill] → [mozmill], softblocker
Whiteboard: [mozmill], softblocker → [mozmill][softblocker]
Assignee: mrbkap → gal
Attached patch patchSplinter Review
This patch makes the XUL document code only use the XUL prototype cache when the XUL document is loaded in a type=chrome docshell. This is needed with Andreas' patch to avoid compartment mismatches when loading XUL scripts since chrome XUL documents can now end up being loaded in more than one compartment. And Andreas, with this patch your testcase loads w/o compartment mismatches, but the test still fails.
jst, I was thinking the other thing we can try is putting extensions into their own compartment. That might be lower-impact. I will try to figure out why the test case still fails.
This is a softblocker. We are very seriously considering not blocking on this. The fixes seem all higher risk that the small amount of breakages this introduces.
Henrik, can you explain how this affects you? Can you easily work around this?
It happens when I call getMostRecentWindow from inside our Mozmill scripts. For now I simply have to unwrap the window a second time, so I'm completely fine for now. So no need from my side for a hardblocker.
I think we should unblock on this until we have some evidence that this really creates a headache for extension developers. jst, please overrule me if you disagree.
blocking2.0: betaN+ → ---
Whiteboard: [mozmill][softblocker] → [mozmill]
jst, can we get an update on this bug please?
Per policy at https://wiki.mozilla.org/Bug_Triage/Projects/Bug_Handling/Bug_Husbandry#Inactive_Bugs. If this bug is not an enhancement request or a bug not present in a supported release of Firefox, then it may be reopened.
Status: NEW → RESOLVED
Closed: 7 years ago
Resolution: --- → INACTIVE
You need to log in before you can comment on or make changes to this bug.

Attachment

General

Created:
Updated:
Size: