Shouldn't cache entries be stored using a weak reference as well?
--
Sent from my mobile device.
Brad Wetmore wrote:
>Tony,
>
>I knew there was another potential issue but couldn't recall it at the
>time.
>
>In JCE verification code, we check to see if the provider in the jar
>file has been si
Tony,
I knew there was another potential issue but couldn't recall it at the time.
In JCE verification code, we check to see if the provider in the jar
file has been signed by an appropriate key. To optimize later provider
checks, we add the signed provider to a "verified" provider cache.
S
Update webrev after a talk about getInstance(), along with comment
embedded below.
http://cr.openjdk.java.net/~mullan/webrevs/ascarpin/7171982/webrev.00/
On 03/28/2013 06:00 PM, Anthony Scarpino wrote:
On 03/28/2013 02:34 PM, Brad Wetmore wrote:
(Vinnie, what do you think about the SunJCE it
I overlooked that potential race condition when creating the SunJCE singleton.
Both proposed solutions risk the construction of superfluous SunJCE objects.
Wouldn't it be better to use the Enum idiom to ensure that multiple SunJCE
objects are not constructed?
On 29 Mar 2013, at 01:00, Anthony
On 03/28/2013 02:34 PM, Brad Wetmore wrote:
(Vinnie, what do you think about the SunJCE item below?)
On 3/22/2013 11:57 AM, Anthony Scarpino wrote:
Hi all,
I need a code review for below webrev. The changes are to have SunJCE
call itself, using it's current instance, for checking such things
(Whoops, was working on two reviews with two related comments, and
reversed the emails).
Just realized, there are no regression tests here.
Simplest is to probably do as much setup as you can, then
java.security.Security.removeProvider("SunJCE"), then issue the calls
that call into these chan
(Vinnie, what do you think about the SunJCE item below?)
On 3/22/2013 11:57 AM, Anthony Scarpino wrote:
Hi all,
I need a code review for below webrev. The changes are to have SunJCE
call itself, using it's current instance, for checking such things as
parameters, instead of searching through t
Hello Tony,
Your changes look fine.
Thanks.
On 22 Mar 2013, at 18:57, Anthony Scarpino wrote:
> Hi all,
>
> I need a code review for below webrev. The changes are to have SunJCE call
> itself, using it's current instance, for checking such things as parameters,
> instead of searching throug
Hi all,
I need a code review for below webrev. The changes are to have SunJCE
call itself, using it's current instance, for checking such things as
parameters, instead of searching through the provider list or creating a
one time instance.
http://cr.openjdk.java.net/~mullan/webrevs/ascarpin