448861 - In <test_bug356370.js> (and 2 others), "WARNING / mozJSComponentLoader.cpp" and "Exception / nsSuiteGlue.js"

Status: RESOLVED WORKSFORME

(SeaMonkey :: General, defect)

Description

[Serge Gautherie (:sgautherie)]

[Mozilla/5.0 (Windows; U; Windows NT 5.0; en-US; rv:1.9.1a2pre) Gecko/2008080202 SeaMonkey/2.0a1pre] (home, debug default) (W2Ksp4)

|make check| fails in a debug build:

SeaMonkey has bug 428078 too;
but it has this following issue first:
{{
WARNING: NS_ENSURE_SUCCESS(rv, rv) failed with result 0x80040111: file .../mozilla/js/src/xpconnect/loader/mozJSComponentLoader.cpp, line 1460

JS Component Loader: ERROR (null):0
                     uncaught exception: [Exception... "Component returned failure code: 0x80040111 (NS_ERROR_NOT_AVAILABLE) [nsIXPCComponents_Utils.import]"  nsresult: "0x80040111 (NS_ERROR_NOT_AVAILABLE)"  location: "JS frame :: file:///.../mozilla/dist/bin/components/nsSuiteGlue.js :: <TOP_LEVEL> :: line 42"  data: no]
}}

Comment 1

[Serge Gautherie (:sgautherie)]

Attachment: test_bug356370.js.log

Comment 2

[Serge Gautherie (:sgautherie)]

Attachment: test_nsAbManager2.js.log

Comment 3

[Serge Gautherie (:sgautherie)]

Attachment: test_bug_437556.js.log

I haven't filed a bug (yet ?) about
{{
ReferenceError: getImportInterface is not defined
}}

Comment 4

[Serge Gautherie (:sgautherie)]

Oh well, this happens in <test_bug394717.js> too (for example),
but that one is marked "PASS"
...
not sure what to think of this :-/

Comment 5

[Serge Gautherie (:sgautherie)]

Eventually, this is "everywhere", as in <test_default_index_handler.js> which passes.

Then, maybe the only blocking part is actually comment 3 :-<

Comment 6

[Mark Banner (:standard8)]

In this bug you have confused messages in logs with actual failures. That is why test_bug394717.js passes, a warning/exception isn't a failure (only assertions get marked as failures).

I suggest that you clarify what you are trying to fix in this bug. The test failure is already covered by bug 428078, the warnings (which I've also seen elsewhere, not just on these tests) are probably subject to another bug, indeed, Neil may know where.

I'm removing bug 448859 from the depends list and bug 437556 from the blocking list, as they aren't really to do with this bug.

Actually, I think really you should close this bug as invalid, and file bug(s) on what you are really trying to fix i.e. "running make check fails in SeaMonkey debug builds" or "test foo throws error bar but the app doesn't do it". Otherwise we're going to be confused as to what is going on here.

If you do do the make check one, then note I've already documented a lot of these here: http://wiki.mozilla.org/MailNews:Automated_Testing#xpcshell_tests but if they are bugs that fail in Firefox, then they should block your core bug 448802, and only that bug and SM specific ones should block a SM specific bug.

Comment 7

[[e-mail]]

I think the warning might be due to which modules are available to xpcshell (i.e. resource://gre/ might be available but resource://app/ might not.)

Comment 8

[Serge Gautherie (:sgautherie)]

Hit this with <test_nsAbAutoCompleteSearch1.js> (bug 452939) too.

***

(In reply to comment #7)
> I think the warning might be due to which modules are available to xpcshell
> (i.e. resource://gre/ might be available but resource://app/ might not.)

Right.

http://mxr.mozilla.org/comm-central/search?string=%2FSanitizer%5C.jsm&regexp=on&case=on&find=%2Fsuite%2F
[
/suite/common/sanitize.xul
    * line 61 -- Components.utils.import("resource:///modules/Sanitizer.jsm");
/suite/common/src/nsSuiteGlue.js
    * line 42 -- Components.utils.import("resource://app/modules/Sanitizer.jsm");
/suite/modules/test/browser_sanitizer.js
    * line 1 -- Components.utils.import("resource://gre/modules/Sanitizer.jsm");
]

(With this test,)
*'app' fails.
*'gre' or '' succeed.

Neil, which one of the 3 paths should we be using in (each of) these 3 places ?

Comment 9

[[e-mail]]

(In reply to comment #8)
> /suite/common/sanitize.xul
>     * line 61 -- Components.utils.import("resource:///modules/Sanitizer.jsm");
> /suite/common/src/nsSuiteGlue.js
>     * line 42 --
> Components.utils.import("resource://app/modules/Sanitizer.jsm");
These two should surely be importing the same module. For the purposes of libxul this should be resource://app/ but I guess the tests will only work if they use resource://gre/ (so in other words the tests are incompatible with libxul).
> /suite/modules/test/browser_sanitizer.js
>     * line 1 --
> Components.utils.import("resource://gre/modules/Sanitizer.jsm");

Comment 10

[Serge Gautherie (:sgautherie)]

(In reply to comment #8)

> http://mxr.mozilla.org/comm-central/search?string=%2FSanitizer%5C.jsm&regexp=on&case=on&find=%2Fsuite%2F

Eventually,

> /suite/common/sanitize.xul
>     * line 61 -- Components.utils.import("resource:///modules/Sanitizer.jsm");

Correct.

> /suite/common/src/nsSuiteGlue.js
>     * line 42 --
> Components.utils.import("resource://app/modules/Sanitizer.jsm");

Fixed by bug 555715.

> /suite/modules/test/browser_sanitizer.js
>     * line 1 --
> Components.utils.import("resource://gre/modules/Sanitizer.jsm");

Will be fixed by bug 556670.

Comment 11

[Philip Chee]

All dependencies look fixed. So closing as WFM