http://redmine.emweb.be/http://redmine.emweb.be/favicon.ico?16934085252012-12-15T21:33:07ZRedmineWt - Bug #1581: WResource::handleRequest: pure virtual method calledhttp://redmine.emweb.be/issues/1581?journal_id=44862012-12-15T21:33:07ZKoen Deforchekoen@emweb.be
<ul><li><strong>Status</strong> changed from <i>New</i> to <i>Feedback</i></li></ul><p>Hey,</p>
<p>I guess you still need to implement B's destructor to call beingDeleted(). Have you tried that and did it not solve the problem ?</p>
<p>Regards,<br><br>
koen</p>
Wt - Bug #1581: WResource::handleRequest: pure virtual method calledhttp://redmine.emweb.be/issues/1581?journal_id=44962012-12-16T00:48:09ZBoris Nagaev
<ul></ul><p>Hello!</p>
<p>I have not tried to implement B's destructor to call beingDeleted(). Why is it needed? beingDeleted() is called by A's destructor before vpointer will be set to WResource's vtable. Is it possible to exclude calls to beingDeleted() from derived classes?</p>
<p>I can't reproduce this as test case, because it is very rare.</p>
Wt - Bug #1581: WResource::handleRequest: pure virtual method calledhttp://redmine.emweb.be/issues/1581?journal_id=45452013-01-08T12:44:02ZKoen Deforchekoen@emweb.be
<ul></ul><blockquote>
<p>I have not tried to implement B's destructor to call beingDeleted(). Why is it needed? beingDeleted() is called by A's destructor before vpointer will be set to WResource's vtable. Is it possible to exclude calls to beingDeleted() from derived classes?</p>
</blockquote>
<p>I'm not entirely sure that this is what happens: is it not simply undefined behavior of calling a virtual method during destruction: I'm not sure that the vtable pointer is being reset to the parent's type as destruction moves from one class to its parent.</p>
<p>Regards,<br><br>
koen</p>