Closed Bug 420480 Opened 16 years ago Closed 16 years ago

regressions in ecma/ObjectObjects/15.2.4.2.js - browser only

Categories

(Core :: JavaScript Engine, defect, P2)

x86
All
defect

Tracking

()

VERIFIED FIXED
mozilla1.9beta5

People

(Reporter: bc, Assigned: mrbkap)

References

Details

(Keywords: regression, testcase)

Attachments

(1 file)

regression in last 24 hours. I'll try to narrow.

ecma/ObjectObjects/15.2.4.2.js
myvar = this;  myvar.toString = Object.prototype.toString; myvar.toString() expected: [object Window] actual: [object XPCCrossOriginWrapper] reason: wrong value
=var MYVAR = new Object( this ); MYVAR.toString() expected: [object Window] actual: [object XPCCrossOriginWrapper] reason: wrong value

js1_5/Regress/regress-68498-003.js
Testing calling obj.eval(str); currently at expect[0] within test - expected: 43 actual: undefined reason: Type mismatch, expected type number, actual type undefined Expected value '43', Actual value 'undefined'
Testing calling obj.eval(str); currently at expect[1] within test - expected: 43 actual: undefined reason: Type mismatch, expected type number, actual type undefined Expected value '43', Actual value 'undefined'
Flags: in-testsuite+
Flags: in-litmus-
Flags: blocking1.9?
Almost certainly not a JS engine bug, but I'll let mrbkap fix.

/be
Assignee: general → mrbkap
Flags: blocking1.9? → blocking1.9+
js1_5/Regress/regress-68498-003.js regressed between 2008-02-29 16:02 and 16:18
Blocks: 418069
(In reply to comment #0)
> ecma/ObjectObjects/15.2.4.2.js

This is expected (though I could fix it now using wrapperObject). You're calling Object.prototype.toString on a XOW.

> js1_5/Regress/regress-68498-003.js

This looks real and serious... In fact, it is fixed by my patch in bug 418565.
Attached patch Like thisSplinter Review
So, the big problem with this patch is that it is now impossible to verify from JavaScript (e.g. mochitests) that a given object is a wrapper (SJOW, XOW, etc).
Attachment #306943 - Flags: review?(brendan)
(In reply to comment #3)
> (In reply to comment #0)
> > ecma/ObjectObjects/15.2.4.2.js
> 
> This is expected (though I could fix it now using wrapperObject). You're
> calling Object.prototype.toString on a XOW.
> 
 
This used to fail but was fixed up until bug 418069 landed. I can add it back as a known failure if that is what is needed but why was it passing before?
It started passing when we stopped outerizing (and thus wrapping) |this|. Fixing that (which fixed a bunch of other stuff made it come back.
Comment on attachment 306943 [details] [diff] [review]
Like this

Followup bug to consolidate the (with this patch) five places that do this little dance?

I think wrappers should be hard to detect, as transparent as possible. They're not supposed to break toString.

/be
Attachment #306943 - Flags: review?(brendan)
Attachment #306943 - Flags: review+
Attachment #306943 - Flags: approval1.9+
(In reply to comment #3)
> 
> > js1_5/Regress/regress-68498-003.js
> 
> This looks real and serious... In fact, it is fixed by my patch in bug 418565.
> 

verified windows|mac that this no longer occurs now that bug 418565 is fixed.
/cvsroot/mozilla/js/tests/public-failures.txt,v  <--  public-failures.txt
new revision: 1.45; previous revision: 1.44
P2 according to bkap.
Priority: -- → P2
Checked into trunk.
Status: NEW → RESOLVED
Closed: 16 years ago
Resolution: --- → FIXED
Target Milestone: --- → mozilla1.9beta5
narrowing this bug to ecma/ObjectObjects/15.2.4.2.js
js1_5/Regress/regress-68498-003.js new failure is bug 421515.

v
Status: RESOLVED → VERIFIED
Summary: regressions in ecma/ObjectObjects/15.2.4.2.js, js1_5/Regress/regress-68498-003.js - browser only → regressions in ecma/ObjectObjects/15.2.4.2.js - browser only
You need to log in before you can comment on or make changes to this bug.

Attachment

General

Created:
Updated:
Size: