[Webkit-unassigned] [Bug 168375] REGRESSION(r212345): [GTK] ASSERT(cookieChangeCallbackMap().contains(jar)) failed in WebCore::stopObservingCookieChanges:54

bugzilla-daemon at webkit.org bugzilla-daemon at webkit.org
Thu Feb 16 00:25:04 PST 2017


https://bugs.webkit.org/show_bug.cgi?id=168375

Carlos Garcia Campos <cgarcia at igalia.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |achristensen at apple.com

--- Comment #5 from Carlos Garcia Campos <cgarcia at igalia.com> ---
I can't reproduce the issue, but I think there are several things with the session management that could leave to problems. This crash could be because because cookie manager sends the stop observing cookie changes relaunching the network process, so if we call this after the network process finishes, we launch a new network process for which start observing changes was not called. But there are more situations this can fail in one way or another. The main problem is that the website data store creates the session ID, but it's web page on creation the one ensuring the session in the web and network processes. So, if you setup your cookie manager before the first web page is created, the start message will not work for ephemeral sessions, because it doesn't exist yet in the network process. But it will exist when we call stop. Were you using a private session when it crashed?

-- 
You are receiving this mail because:
You are the assignee for the bug.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.webkit.org/pipermail/webkit-unassigned/attachments/20170216/1d7820ec/attachment.html>


More information about the webkit-unassigned mailing list