On 04/14/2010 03:30 PM, huican wrote:
> Hello Bob,
>
> Thanks for your reply...
> I have more questions inline ;<
>
> On Wed, Apr 14, 2010 at 5:15 PM, Robert Relyea <rrel...@redhat.com>
> wrote:
>   
>> On 04/14/2010 02:58 PM, huican wrote:
>>     
>>> Hello,
>>>
>>> I am new to NSS crypto, I just wonder whether there is any easy way to
>>> use my own crl_callback function for CRL check.
>>>
>>>       
>> No, there isn't a callback, there is a verify function that allows you
>> to control just about every possible semantic of revocation checking. In
>> that case you can turn off CRL checking all together, and do your CRL
>> check later by hand.
>>     
> Could you please give me a little bit more details? I need to know
> what is that verify function name, (or which file in the source tree
> to check)?
>   

CERT_PKIXVerifyCert() See cert.h and certt.h.
> How to turn off CRL checking there?
>   
Revocation checking is controled by *paramsIn. There are a set of
revocation policies set by default. None of them turn off CRL checking
for CRLs already fetched and valid. You'll have to write your own
policy. I suggest modelling it after the predefined
CERT_GetClassicOCSPDisabledPolicy(). The Revocation policies are meant
to be as flexible as possible -- and since revocation itself is rather
complicated (and certainly the choices of semantics are fairly
complicated) the revocation policy structure is fairly complicated. The
idea is that you should probably use one of the predefined policies, but
we could build new policies pretty quickly if the need arose.
> What is your meaning of "do CRL check by hand"?
>   
I mean you would have to do the CRL check yourself since you are trying
to use a custom CRL or whatever. I still can't conceive of a case where
you would *not* want to validate against an existing valid CRL if it's
already available, but that is what you seem to be asking.

bob
> Your reply will be highly appreciated.
>
> Thanks
>
>   


-- 
dev-tech-crypto mailing list
dev-tech-crypto@lists.mozilla.org
https://lists.mozilla.org/listinfo/dev-tech-crypto

Reply via email to