[Bug 21568] advapi32/crypt tests crash with +heap

[wine-bugs at winehq.org]

http://bugs.winehq.org/show_bug.cgi?id=21568





--- Comment #6 from Austin Lund <austin.lund at gmail.com>  2010-07-12 21:44:27 ---
(In reply to comment #5)
> (In reply to comment #4)
> > OK. So wine needs to be fixed.  Is there a way to know if a pointer has been
> > freed?  Just looking through MSDN it seems there isn't.  If that's the case
> > then a list of valid handles will need to be kept to avoid referencing freed
> > memory.  Unless there is another option.
> 
> Wine doesn't need to be fixed, the tests need to be fixed not to crash when
> +heap is enabled.  The app itself is buggy, there's no guarantee it wouldn't
> crash with a different heap implementation under Windows.

That doesn't make sense to me.  Under the current test windows doesn't crash
and an app depends on that.  Wine has the possibility of crashing and does with
+heap.  So either the test is removed, or wine is changed to not crash like
windows.  I don't understand how the test can be fixed without being removed
whilst still testing the required behaviour.

-- 
Configure bugmail: http://bugs.winehq.org/userprefs.cgi?tab=email
Do not reply to this email, post in Bugzilla using the
above URL to reply.
------- You are receiving this mail because: -------
You are watching all bug changes.