429864 - assertion at startup with venkman

Status: RESOLVED FIXED

(Other Applications Graveyard :: Venkman JS Debugger, defect)

Description

[Andrew Schultz]

Attachment: stack

With current trunk, I get an assertion at startup,

###!!! ASSERTION: This is not supposed to fail!: 'Error', file /build/andrew/old/mozilla/js/src/xpconnect/src/nsXPConnect.cpp, line 1057

except for the first startup (when it does registration).  According to bonsai, the checkin that triggered this is bug 428336, but that's clearly a typo.

http://bonsai.mozilla.org/cvsquery.cgi?treeid=default&module=all&branch=HEAD&branchtype=match&dir=&file=%2Fmozilla%2Fjs%2Fsrc%2F&filetype=match&who=crowder%25fiverocks.com&whotype=match&sortby=Date&hours=2&date=explicit&mindate=2008-04-14&maxdate=2008-04-15&cvsroot=%2Fcvsroot

stacktrace attached

Comment 1

[:Gijs (he/him)]

Actual bug is bug 428366, CC-ing mrbkap and brendan. Does anyone know how bad this is?

Comment 2

[Brendan Eich [:brendan]]

That assertion has nothing to do with bug 428366 or its patch. Was the wrong bug # cited, or could you have the wrong blame window?

/be

Comment 3

[:Gijs (he/him)]

(In reply to comment #2)
> That assertion has nothing to do with bug 428366 or its patch. Was the wrong
> bug # cited, or could you have the wrong blame window?
> 
> /be
> 

The bug number in the blame window from comment #0 is definitely typoed (bug 428336 does not have patches and is a duplicate); the summary in it indicates bug 428366 (or bug 416834, which has the same summary). If none of those bugs apply, I would presume that the blame window is wrong. Andrew?

Comment 4

[Andrew Schultz]

I actually backed out the patch in the blame window (which is bug 428366) and the assertion went away.  The full checkin window between tinderbox cycles is

http://bonsai.mozilla.org/cvsquery.cgi?branch=HEAD&date=explicit&mindate=2008-04-14+16%3A13&maxdate=2008-04-14+17%3A47&cvsroot=%2Fcvsroot
(among those checkins, 428366 looks most relevant to me)

Comment 5

[Brendan Eich [:brendan]]

Ok, mysterious. I'm going to rely on mrbkap here.

/be

Comment 6

[Brian Crowder]

Guessing:  XPCWrapper::FindEval is somehow failing here?

Comment 7

[Blake Kaplan (:mrbkap) (inactive)]

Attachment: Proposed fix

My patch broke using 'eval' for eager standard class users.

Comment 8

[Brian Crowder]

Should block 1.9?  This would be bad for embedders.

Comment 9

[Mike Beltzner [:beltzner, not reading bugmail]]

Don't think this blocks, but we'd take a reviewed, safe fix.

Comment 10

[Brendan Eich [:brendan]]

Comment on attachment 318905 [details] [diff] [review]
Proposed fix

Whoops! This really ought to be taken for 1.9.

/be

Comment 11

[Mike Beltzner [:beltzner, not reading bugmail]]

Comment on attachment 318905 [details] [diff] [review]
Proposed fix

a1.9=beltzner

Comment 12

[Brian Crowder]

jsapi.c: 3.446