310828 - JS error when hitting try again button on an error page

Status: RESOLVED FIXED

(Core :: DOM: Navigation, defect)

Description

[Henrik Skupin [:whimboo][⌚️UTC+2]]

Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.9a1) Gecko/20051001 Firefox/1.6a1

When hitting the Try Again button on an error page following lines appears on my
console:

JavaScript error: , line 0: uncaught exception: [Exception... "Component
returned failure code: 0x804b0013 [nsIDOMNSLocation.reload]"  nsresult:
"0x804b0013 (<unknown>)"  location: "JS frame ::
about:neterror?e=deniedPortAccess&u=http%3A//www.mozilla.org%3A25/&c=UTF-8&d=www.mozilla.org%3A25%20uses%20a%20network%20port%20which%20is%20normally%20used%20for%20purposes%20other%20than%20Web%20browsing.%20Firefox%20has%20canceled%20the%20request%20for%20your%20protection.
:: retryThis :: line 98"  data: no]
###!!! ASSERTION: XPConnect is being called on a scope without a 'Components'
property!

This is pretty much always bad. It usually means that native code is
making a callback to an interface implemented in JavaScript, but the
document where the JS object was created has already been cleared and the
global properties of that document's window are *gone*. Generally this
indicates a problem that should be addressed in the design and use of the
callback code.
: 'Error', file
../../../../../../mozilla/js/src/xpconnect/src/xpcwrappednativescope.cpp, line 564
Break: at file
../../../../../../mozilla/js/src/xpconnect/src/xpcwrappednativescope.cpp, line 564
++DOMWINDOW == 21
WARNING: NS_ENSURE_TRUE(NS_SUCCEEDED(result)) failed, file
../../../../../../mozilla/content/xml/document/src/nsXMLContentSink.cpp, line 1002

Comment 1

[Boris Zbarsky [:bzbarsky]]

Is this a regressio?  What's the regression range, if so?

Comment 2

[Boris Zbarsky [:bzbarsky]]

Oh, and this worksforme with current trunk Firefox (9/30) and Seamonkey (10/2) both.

Comment 3

[Adam Guthrie]

*** Bug 311175 has been marked as a duplicate of this bug. ***

Comment 4

[Andrew Schultz]

I can reproduce the bug using the steps from 311175.  This is not a recent
regression (occurred in trunk build 20050708 when error pages were enabled in
seamonkey).

Comment 5

[Boris Zbarsky [:bzbarsky]]

Oh, this is only for the file channel case?

Comment 6

[Henrik Skupin [:whimboo][⌚️UTC+2]]

(In reply to comment #5)
> Oh, this is only for the file channel case?

No, it happens for me for all "file not found" and "denied port access" error
cases. The error of bug 311175 is another one as mine but both result on the
same action. 

Both assertions were raised somewhere here:
http://lxr.mozilla.org/seamonkey/source/dom/src/base/nsLocation.cpp#825

The additional output which I get is only raised on debug builds:
http://lxr.mozilla.org/seamonkey/source/js/src/xpconnect/src/xpcwrappednativescope.cpp#538

Comment 7

[Boris Zbarsky [:bzbarsky]]

I don't see any assertions.  I do see the exceptions; perhaps we should try/catch around that loadURI call in the error page?  Throwing the exception from loadURI is the right thing, maybe.  Not sure about that.

Comment 8

[Vladimir Vukicevic [:vlad] [:vladv] (needinfo me, slow to respond)]

-'ing for 1.9, maybe need try/catch around loadURI -- only thing this hurts is spam in error console

Comment 9

[Adam Guthrie]

Attachment: try/catch the reload() call

This is what bz suggested. I'm not sure why vlad moved this to Firefox : Error Console...

Comment 10

[Boris Zbarsky [:bzbarsky]]

Comment on attachment 237286 [details] [diff] [review]
try/catch the reload() call

Looks good.  Thanks!

Comment 11

[Adam Guthrie]

Attachment: try/catch the reload() call (for check-in)

The previous patch had some tabs in it. This patch is what needs to be checked in.

Comment 12

[Henrik Skupin [:whimboo][⌚️UTC+2]]

Comment on attachment 237295 [details] [diff] [review]
try/catch the reload() call (for check-in)

Trivial patch.

Comment 13

[Mike Schroepfer]

Comment on attachment 237295 [details] [diff] [review]
try/catch the reload() call (for check-in)

a=schrep for drivers.

Comment 14

[Adam Guthrie]

Patch checked in on trunk and branch by bz. ->FIXED