Stephen Gallagher wrote:
On 03/17/2010 03:46 PM, Dmitri Pal wrote:
> Stephen Gallagher wrote:
>> On 03/17/2010 03:20 PM, Stephen Gallagher wrote:
>>> On 03/17/2010 03:13 PM, Stephen Gallagher wrote:
>>>> I introduced some memory leaks with my patch for handling
disconnected
>>>> DBUS sends. This patch fixes it.
>>
>>> Pushed to master and 1-1-0
>>
>> Sorry, I sent this reply to the wrong email. This has not yet been
>> reviewed or pushed upstream.
>>
> Yeah, got me puzzled for a second.
> Looking at the patch it seems pretty straightforward.
> Was the messages dereferenced before your original patch and just the
> dereferencing got lost or they never were dereferenced and leak was
> actually always there?
Prior to my patch, they were dereferenced (which will free the memory
once the references reach zero). When I converted the functions to use
the new routine sbus_conn_send(), I forgot to retain the dereference,
and thus caused a memory leak.
In the case of the monitor, this was a serious problem, as it pings the
other components constantly.
I would say ACK then but I reviewed just he diff and have not applied it.
I suspect this is not enough to give a full ACK so count as 1/2.
_______________________________________________
sssd-devel mailing list
sssd-devel(a)lists.fedorahosted.org
https://fedorahosted.org/mailman/listinfo/sssd-devel
--
Thank you,
Dmitri Pal
Engineering Manager IPA project,
Red Hat Inc.
-------------------------------
Looking to carve out IT costs?
www.redhat.com/carveoutcosts/