792594 - comp.utils.import("resource://gre/modules/ContactService.jsm") throws NS_ERROR_XPC_GS_RETURNED_FAILURE error in mochitest b2g

Status: RESOLVED FIXED

(Core :: DOM: Device Interfaces, defect)

Description

[Martijn Wargers (dead)]

While doing this piece of code in mochitest b2g:
var comp = SpecialPowers.wrap(Components);
comp.utils.import("resource://gre/modules/ContactService.jsm");

I got this error:
3 ERROR TEST-UNEXPECTED-FAIL |
/tests/dom/tests/mochitest/webapi-smoke-sprint/test_contacts_basics_simple.html
| an unexpected uncaught JS exception reported through window.onerror
- NS_ERROR_XPC_GS_RETURNED_FAILURE: Component returned failure code:
0x80570016 (NS_ERROR_XPC_GS_RETURNED_FAILURE) [nsIJSCID.getService] at
resource://gre/modules/XPCOMUtils.jsm:198

This ContactService.jsm component import is being done by the contacts mochitests, so these tests would fail with this error in mochitest b2g.
http://mxr.mozilla.org/mozilla-central/source/dom/contacts/tests/

Comment 1

[Martijn Wargers (dead)]

Btw, on b2g this ContactService.jsm component import isn't necessary, because that is already done (I guess here: http://mxr.mozilla.org/mozilla-central/source/b2g/chrome/content/shell.js#14 ).
But this component import is needed for mochitests on Firefox desktop.

Comment 2

[Gregor Wagner [:gwagner]]

(In reply to Martijn Wargers [:mw22] (QA - IRC nick: mw22) from comment #1)
> Btw, on b2g this ContactService.jsm component import isn't necessary,
> because that is already done (I guess here:
> http://mxr.mozilla.org/mozilla-central/source/b2g/chrome/content/shell.js#14
> ).
> But this component import is needed for mochitests on Firefox desktop.

Right, it is only needed for desktop mochitests. Don't import it for b2g builds.

Comment 3

[Martijn Wargers (dead)]

I can do that, but from what I understood, it shouldn't cause it to throw js errors anyhow.

Comment 4

[Martijn Wargers (dead)]

For now, I use myself 'if (!SpecialPowers.getBoolPref("dom.mozContacts.enabled"))' to workaround this issue.

Gregor on irc:
[18:46]	<gwagner_>	mw22: it might be because it is already loaded in the parent process
[18:47]	<gwagner_>	mw22: and it should only be loaded in the parent. but the message manager might get confused if you try to load it in the child as well

Comment 5

[David Chan [:dchan]]

This may be related to bug 795275. I'm also getting similar errors for my b2g mochitests

Comment 6

[Martijn Wargers (dead)]

Attachment: fix for now?

This would fix the failures in b2g, but I don't know if it would be acceptable.

Comment 7

[Gregor Wagner [:gwagner]]

Comment on attachment 671516 [details] [diff] [review]
fix for now?


>diff --git a/dom/tests/mochitest/Makefile.in b/dom/tests/mochitest/Makefile.in
>index 5de9321..f49fa15 100644
>--- a/dom/tests/mochitest/Makefile.in
>+++ b/dom/tests/mochitest/Makefile.in
>@@ -27,6 +27,7 @@ DIRS	+= \
> 	storageevent \
> 	pointerlock \
> 	webapps \
>+	webapi-smoke-sprint \

Remove?

I don't really like this fix but I can't think of anything better
right now. Lets try it.

Comment 8

[Martijn Wargers (dead)]

Attachment: fix for now

Oops, yes, sorry I added that.

Comment 9

[Ryan VanderMeulen [:RyanVM]]

https://hg.mozilla.org/integration/mozilla-inbound/rev/0fb868c93955

Should this be [leave open] for a real fix?

Comment 10

[Martijn Wargers (dead)]

Yes, I think so, I mentioned this bug number in the patch.

Comment 11

[Ryan VanderMeulen [:RyanVM]]

you can add it yourself, you know ;)

Comment 12

[(no longer active)]

https://hg.mozilla.org/mozilla-central/rev/0fb868c93955