Re: [PATCH 16 of 16] free intermediates in error path
On Sun, Mar 15, 2009 at 5:59 AM, Rocco Rutte <pdmef@xxxxxxx> wrote:
> Hi,
>
> * Erik Hovland wrote:
>
>> # HG changeset patch
>> # User Erik Hovland <erik@xxxxxxxxxxx>
>> # Date 1236899608 25200
>> # Branch HEAD
>> # Node ID b195a368657041f8ad2a537dfeccf7b876b299c2
>> # Parent dfcfb5a6d00dec73c1ffa91e46f7fe37fe2aea50
>> free intermediates in error path.
>
> Doesn't that function contain more error checks where this would leak? I
> didn't really read the code, but it looks like it always leaks in case
> it's not set to SmimeDefaultKey.
This is where static analysis checkers shine over humans eyes. According
to it, there are no more links in that function. And frankly, it is
less likely to
fail then I am in checking for all of the conditions. But I will
attempt to figure
out if I can develop a second opinion of that functions.
E
--
Erik Hovland
erik@xxxxxxxxxxx
http://hovland.org/