[Webkit-unassigned] [Bug 52819] Crash in WebCore::HistoryController::itemsAreClones

Wed Feb 2 18:01:41 PST 2011

https://bugs.webkit.org/show_bug.cgi?id=52819

--- Comment #17 from Charles Reis <creis at chromium.org>  2011-02-02 18:01:41 PST ---
(In reply to comment #16)
> In Chromium (in render_view.cc), it takes care to load the error page with the "replace" parameter set to true when the error page should be shown in place of a back/forward entry.  It only sets replace to false when we need to show an error page for a new navigation that failed to complete.
> 
> In other words, I believe that in the back/forward case, the error page should take the place of the page you were trying to navigate back to.

Thanks-- that helps.  I was able to get the error page to commit by changing HistoryController::updateForCommit.  (It needed to commit for FrameLoadTypeReplace in addition to the back/forward load types.)

After that, Chrome seems to work properly.  I needed another change to test_shell, so that it doesn't forget the page ID when we do the load (of the error page).  test_shell is now working.

Unfortunately, DumpRenderTree is also still behaving incorrectly-- it doesn't commit when we go back or forward to an error page.  I'll probably need to update that as well.  I'll upload the patch of what I have so far for feedback, and I'll work on DumpRenderTree in the mean time.

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