Closed Bug 535699 Opened 15 years ago Closed 6 years ago

Permission denied for <https://getfirebug.com> to get property Function.__parent__ from <>

Categories

(Core :: XPConnect, defect)

x86
All
defect
Not set
normal

Tracking

()

RESOLVED INACTIVE
Tracking Status
blocking2.0 --- -

People

(Reporter: johnjbarton, Unassigned)

Details

(Whiteboard: [firebug-p2])

1. Install Firebug 1.5b7, http://getfirebug.com/releases/firebug/1.5X
2. https://getfirebug.com/tests/content/net/1275/issue1275.htm
3. Error in Console panel

Fails on:
Mozilla/5.0 (Windows; U; Windows NT 5.1; en-US; rv:1.9.3a1pre) Gecko/20091217 Minefield/3.7a1pre

Works on:
Mozilla/5.0 (Windows; U; Windows NT 5.1; en-US; rv:1.9.2b6pre) Gecko/20091217 Namoroka/3.6b6pre
blocking2.0: --- → ?
Whiteboard: [firebug-p1]
Causes another Firebug FBTest to fail:
https://getfirebug.com/tests/content/net/1862/issue1862.html
I just tried following the steps in comment 0 and don't get any error messages.  I even tried clicking the button on the test page; still no error messages.

So worksforme as far as I can tell.  Using a 2009-12-20 build here.
Our test net/1275 still fails for me, and I still get the error message. Maybe the test harness is the cause. 

I don't  get the message if I run the test manually. 

Honza, according to your post to the FWG, this test passes for you?
Yes, the test passes for me.
Honza
WFM Mozilla/5.0 (Windows; U; Windows NT 5.1; en-US; rv:1.9.3a1pre) Gecko/20100107 Minefield/3.7a1pre
In Firebug's FBTest window we run this code:
   win.wrappedJSObject.getXMLResponse(function(request)
   {
... operation on FBTest window objects
   }
The function getXMLResponse() is a Web page function with a call back, we are passing the callback as chrome space function and calling it from chrome.

This isn't unsafe in fact because its our web page, but I guess it should cause a security exception? 

If no, then this is a bug.

If yes, then why is the exception so opaque?

The test driver code is 
https://fbug.googlecode.com/svn/tests/content/net/1862/issue1862.js

The error message points to line 44 in 
https://fbug.googlecode.com/svn/tests/content/net/1862/issue1862.html

Honza, I think we need to change these tests to click the button by an event and then monitor the net panel for the updating DOM mutations.
We do not see these messages under FF 3.5
Not blocking here until we get a better picture of what's actually wrong here.
blocking2.0: ? → -
I am finally able to reproduce this reliably on 
Mozilla/5.0 (Windows; U; Windows NT 6.1; en-US; rv:1.9.2.6pre) Gecko/20100625 Namoroka/3.6.6pre

I suspect that all the previous comments about WFM/fails are confused because the issue is somehow dependent on some state that changes. In the particular example below, the existence of another tab with the same url triggers the error message.

1. install Firebug 1.6a14, http://getfirebug.com/releases/firebug/1.6X/firebug-1.6X.0a14.xpi
2. install FBTest 1.6, from SVN at https://fbug.googlecode.com/svn/fbtest/branches/fbtest1.6
3. Open Firebug on home page
4. Open FBTest (shift-T or Firebug Icon Menu > Open test console)
5. Use https://getfirebug.com/tests/content/testlists/firebug1.6.html as the test list in FBTest
6. Open Net tests, right click on  net/1456/issue1456.js and use Open Test Page
You should see https://getfirebug.com/tests/content/net/1456/issue1456.htm in a tab.
7. left click on the same test. 

We get a bunch of the subject error messages and the FBTest fails.

I don't know how this test explains the failure in the testbot case, since there we don't have step 6. But perhaps the error message is trying to tell us about an action that crosses two windows that have the same URL. In the STR above we have two complete tabs, but in the testbot case we have the "same" window being touched at a time when its identity is in flux. Because the testbot is running the page very fast via javascript it may give different results than we get by manually pushing the buttons.

Perhaps if we can decode the error message in the STR here we can get a clue. Honza, is there URL-based lookup code in the net traffic analyzer that can get confused by two windows with the same URL?
I can't reproduce the problem even if having two tabs (the same URL) opened.

> Honza, is there URL-based lookup code in the net traffic analyzer that can get
> confused by two windows with the same URL?
The URL is not used as a key so I don't see how this could be a problem.

> Honza, I think we need to change these tests to click the button by an event
> and then monitor the net panel for the updating DOM mutations.
The test is refactoring (clicking the test button now).

I have updated test-bot config at:
http://getfirebug.com/releases/firebug/test-bot.config

So, 1.6a15 and the patch for the test is used.
The test-bot will use the new configuration in 24 hours.

Honza
Whiteboard: [firebug-p1] → [firebug-p2]
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: 6 years ago
Resolution: --- → INACTIVE
You need to log in before you can comment on or make changes to this bug.