[Webkit-unassigned] [Bug 85138] New: Plugins are killed after 45 seconds if showing keychain "Allow/Deny" dialog

bugzilla-daemon at webkit.org bugzilla-daemon at webkit.org
Sat Apr 28 09:24:54 PDT 2012


https://bugs.webkit.org/show_bug.cgi?id=85138

           Summary: Plugins are killed after 45 seconds if showing
                    keychain "Allow/Deny" dialog
           Product: WebKit
           Version: 528+ (Nightly build)
          Platform: Macintosh Intel
        OS/Version: All
            Status: UNCONFIRMED
          Severity: Major
          Priority: P2
         Component: Plug-ins
        AssignedTo: webkit-unassigned at lists.webkit.org
        ReportedBy: rsherry at adobe.com


I'm developing a mac plugin for WebKit, and on occasion it wants to use the keychain.

This might possibly bring up the keychain's Deny/Allow/Always-Allow dialog.  I've noticed that if that is left up there for more than about 45 seconds, our plugin process is killed.  This is evidently the "hang-detection" timeout.

WebKit suspends the hang-timeout, allowing modal dialogs to stay up essentially forever, using a swizzled runModalForDialog: function; but the keychain dialog is actually a separate process, called within any of a very large number of low-level calls so it would be impractical to swizzle or override them.

Is there a way Safari/WebKit can detect that the plugin process is waiting for a keychain dialog and suspend the hang-timeout?

Right now I have our plug-in creating a 1x1 modal dialog around calls that could possibly show a keychain, but frankly I don't know how many calls we have that could do it or where they are (I'm working with a large body of inherited code).  It would be best to detect the situation rather than try to prevent it.

I can create a plug-in that does this if needed, but will wait for comments about the feasibility of this.

-- 
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