[Webkit-unassigned] [Bug 93661] [JSC] MutationObservers should not create circular, leaky references

bugzilla-daemon at webkit.org bugzilla-daemon at webkit.org
Mon Jan 7 16:06:59 PST 2013


--- Comment #46 from Adam Klein <adamk at chromium.org>  2013-01-07 16:08:52 PST ---
Note that in http://trac.webkit.org/changeset/138841 Chromium stopped using the IDL file to generate V8MutationCallback, since its implementation had diverged almost completely from other callbacks. I suspect that a similar change on the JSC side would be the easiest way forward. The two bits of behavior are:

1. JSMutationCallback should only hold a weak handle to the callback function.
2. A PrivateName should be added on the JSMutationObserver wrapper to the callback function.

(1) is problematic while JSMutationCallback is generated from an IDL, since a new weak version of JSCallbackData would be required. I suspect (though I'm not positive) that a completely custom version won't need the same thread-handling code in JSCallbackData, since mutation observers are only used on the main thread.

Configure bugmail: https://bugs.webkit.org/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are the assignee for the bug.

More information about the webkit-unassigned mailing list