Hi, could the release manager(s) please take a
decision on mentioned bug/feature request?
Thanks a lot,
Mike
Hi!
Hi, could the release manager(s) please take a
decision on mentioned bug/feature request?
Could you give a quick summary of what the decision point is just so I
won't miss anything scanning through the discussion and these bugs?
--
Stanislav Malyshev, Software Architect
SugarCRM: http://www.sugarcrm.com/
(408)454-6900 ext. 227
Hi Stas,
Could you give a quick summary of what the decision point is just so I
won't miss anything scanning through the discussion and these bugs?
In r299770 [1] I introduced a global var_hash to serialize() so that
recursive calls to serialize()/unserialize() can know about the same
objects/variable references.
This was IMO a good thing to do (tm), but it obviously broke
serializ behaviour when called in __sleep/__wakeup, because of
the order of calls:
serialize(obj)
-> obj->__sleep does serialize() (in user code)
-> then internally serialize(obj->prop) happens
unserialize(obj)
-> internally unserialize(obj->prop) is done
-> obj->__wakeup is called which does unserialize() (user code)
As one can see the IDs of the referenced objects when unserializing
cannot match
the IDs at serialization time, because of the mixed up call order.
So I can only see two solutions:
- either disallow serialize/unserialize in __sleep/__wakeup
- or revert r299770
... unless someone else has a better idea.
[1] http://svn.php.net/viewvc/?view=revision&revision=299770
Mike
Hi!
So I can only see two solutions:
- either disallow serialize/unserialize in __sleep/__wakeup
- or revert r299770
So, what is the status of this? I think we better revert it for 5.4 and
look for solution that does not mess up existing code.
Stanislav Malyshev, Software Architect
SugarCRM: http://www.sugarcrm.com/
(408)454-6900 ext. 227
Hi!
So I can only see two solutions:
- either disallow serialize/unserialize in __sleep/__wakeup
- or revert r299770
So, what is the status of this? I think we better revert it for 5.4 and
look for solution that does not mess up existing code.
I wrote Michael, that I think we should revert this.
Hi!
So, what is the status of this? I think we better revert it for 5.4 and
look for solution that does not mess up existing code.
Hi there!
Yes, I received your mails, sorry for being quiet!
I cooked up a patch which uses clean var_hashs or (un)serialize
within __wakeup and __sleep.
Tests pass, my test case in the bug produces a stack overflow,
which is IMO expected. I'm just waiting for feedback of the
bug reporter.
Thanks,
Mike