481434 - Lack of XOW for |this|

Status: RESOLVED FIXED

(Core :: XPConnect, defect)

Description

[moz_bug_r_a4]

This seems to be a regression from bug 480430.

When accessing a window by using |this|, __parent__ property or valueOf.call(),
the window is not wrapped in XOW.

Comment 4

[Blake Kaplan (:mrbkap) (inactive)]

Attachment: Fix

I haven't compiled or tested this, but, ugh.

Comment 5

[Boris Zbarsky [:bzbarsky]]

Comment on attachment 365484 [details] [diff] [review]
Fix

Please, please add tests for this bug and the bug that caused the regression.

Comment 6

[Blake Kaplan (:mrbkap) (inactive)]

http://hg.mozilla.org/mozilla-central/rev/05401ee5565e and http://hg.mozilla.org/releases/mozilla-1.9.1/rev/e7965660ec6d

Comment 7

[Samuel Sidler (old account; do not CC)]

Blake: I'm guessing this landed on 1.9.0.12 with your roll up patch? Feel free to tell me if not...

Comment 8

[moz_bug_r_a4]

It seems that the 1.9.0 patch does not include the regression fix for this bug.

I can reproduce the testcases in this bug on Mozilla/5.0 (Windows; U; Windows
NT 5.1; en-US; rv:1.9.0.12pre) Gecko/2009062406 GranParadiso/3.0.12pre

Comment 9

[Samuel Sidler (old account; do not CC)]

*sigh*

Removing the fixed keyword. Blake?

Comment 10

[Samuel Sidler (old account; do not CC)]

Comment on attachment 365484 [details] [diff] [review]
Fix

Approved for 1.9.0.12. a=ss

Comment 11

[Blake Kaplan (:mrbkap) (inactive)]

Checking in js/src/xpconnect/src/nsXPConnect.cpp;
/cvsroot/mozilla/js/src/xpconnect/src/nsXPConnect.cpp,v  <--  nsXPConnect.cpp
new revision: 1.174; previous revision: 1.173
done

Comment 12

[Al Billings [:abillings - ex-MoCo]]

This is weird. I tried all through cases, as local files, via http and via https here on bugzilla, and none of them give an alert with cookie information on Firefox 3.0.11 on Windows XP. The same happens on 3.0.12. This doesn't jive with the reporter mentioning the issue as not fixed in a 3.0.12pre build earlier.

I'm baffled.

Comment 13

[Blake Kaplan (:mrbkap) (inactive)]

Al, this was a regression from a bug that landed on trunk first, and then this was fixed on trunk. It's blocking and "fixed" on the 1.9.0 branch because the patch was required there. This bug never actually affected the 1.9.0 branch, though.

Comment 14

[Al Billings [:abillings - ex-MoCo]]

Comment #8 is confusing then:

"It seems that the 1.9.0 patch does not include the regression fix for this bug.

I can reproduce the testcases in this bug on Mozilla/5.0 (Windows; U; Windows
NT 5.1; en-US; rv:1.9.0.12pre) Gecko/2009062406 GranParadiso/3.0.12pre"

If it didn't affect the 1.9.0 branch...

Comment 15

[Blake Kaplan (:mrbkap) (inactive)]

(In reply to comment #14)
> If it didn't affect the 1.9.0 branch...

There was a short window of time between when the fix for bug 479442 landed on the 1.9.0 branch (on 6/23 or so) and when this patch landed on the 1.9.0 branch (on 6/24) where the 1.9.0 branch was affected. However, none of the official releases were affected.

Sorry for the confusion!

Comment 16

[Al Billings [:abillings - ex-MoCo]]

Ok. Nothing for QA to do here then.