Chrome for iOS: TypeError: undefined is not an object (evaluating '__gCrWeb.autofill.extractForm

2,849 views
Skip to first unread message

Edward Sargisson

unread,
Feb 26, 2016, 5:51:58 PM2/26/16
to Chromium-discuss
Hi,
We have a catch all JavaScript error handler which forwards JS errors to a web endpoint that we log. We are seeing these errors logged on a daily basis. When they happen we get a large burst all at once - up to 1000.
We also know that they are associated with the user failing to use our website - this failure mode is blocking them.

Sadly, however, we have not managed to reproduce this ourselves. We're pretty sure we're looking at a defect in Chrome for iOS or something in our site that it kicking off this behaviour in Chrome.

I have captured the requests and responses from one failure and built a mock app server using them. However, even that did not reproduce the failure.

My current supposition is that the autofill_controller.js is attempting to load into the JS context but either __gCrWeb or __gCrWeb.autofill are not there.

We've seen this on both v 47 and v 48 of Chrome for iOS on iOS 9.2.1 - there are probably other combinations that exhibit this defect.

I realize that this report is fairly useless without a reproduction but I've run out of things to try. We're quite willing to try things in an attempt to find this.

Does anybody have any ideas they might suggest we try.

Many thanks in advance.

Cheers,
Edward

PhistucK

unread,
Feb 26, 2016, 6:03:24 PM2/26/16
to ejs...@gmail.com, Chromium-discuss
This is an internal object of Chrome for iOS, it seems. If you can filter such errors (a hard coded if (e.message.indexOf("__gCrWeb") !== -1) return;), just do it.

Anyway, this is a Chrome bug.
You can search crbug.com for an existing issue and star it. If you cannot find one, file a new issue using the "New issue" link on the same page.
Please, do not add a "+1" or "Me too" or "Confirmed" (or similar) comment. It just wastes the time of Chrome engineers and sends unnecessary e-mails to all of the people who starred the issue.

You can reply with a link to the found or created issue and might get triaged (and fixed) faster.

Thank you.



PhistucK

--
--
Chromium Discussion mailing list: chromium...@chromium.org
View archives, change email options, or unsubscribe:
http://groups.google.com/a/chromium.org/group/chromium-discuss

---
You received this message because you are subscribed to the Google Groups "Chromium-discuss" group.
To unsubscribe from this group and stop receiving emails from it, send an email to chromium-discu...@chromium.org.

Edward Sargisson

unread,
Feb 26, 2016, 6:26:26 PM2/26/16
to Chromium-discuss, ejs...@gmail.com
Hi PhistucK,
Sure - thanks for the link.
I've posted it to Issue 590375.

At this point, I'm unwilling to filter in the error handler. I believe this is causing issues for the users that hit it and I'd quite like to keep a track of that.

Cheers,
Edward

PhistucK

unread,
Feb 26, 2016, 6:33:13 PM2/26/16
to ejs...@gmail.com, Chromium-discuss
But it is not your issue and you may never be able to fix it, it is an internal Chrome issue that is accidentally exposed to your error handler just because the webview API does not let you inject "internal" scripts that are not exposed to the page.


PhistucK

Rhys Gareth

unread,
Mar 22, 2016, 12:26:37 AM3/22/16
to Chromium-discuss, ejs...@gmail.com
I'm seeing the exact same problem as Edward. i.e. we have a JS error catcher, and regularly we get a burst of these errors from a user. And finally, I prefer to not suppress the error in case I lose visibility of its frequency/impact on users.

Rhys
Reply all
Reply to author
Forward
0 new messages