Sanjay Agrawal wrote:
> Thanks all. This has been very helpful.
> 
> One more question:
> Using SCA6000 through SCF wouldn't void the FIPS compliance that SCA6000 
> provides. I am aware of the FIPS/SCF project but am not sure if current 
> SCF's non compliance with FIPS would affect the overall compliance. I 
> guess if SCF is purely pass through with the "crypto boundary" at 
> SCA6000, SCF shouldn't impact anything.

Yes, it's all about the crypto boundary.  If only the algorithms and 
keystore are used on the SCA6000, it would be in compliance.

Tony

> 
> Thanks,
> - sanjay
> 
> Darren J Moffat wrote:
>> Sanjay Agrawal wrote:
>>> Let me see if I get this right:
>>> 1) NSS provides a wrapper for PKCS11 which means any PKCS11 provider 
>>> can "plug-in" into NSS libraries.
>> correct
>>
>>> 2) SCF is also a wrapper ( actually much more than a wrapper but I am 
>>> using the term to draw a parallel between NSS and SCF) for PKCS11.
>> correct
>>
>>> 3) NSS does NOT use SCF. It is a standalone library with its own 
>>> plugin modules. So it can't use SCF enabled cryptos. For SCA6000 h/w 
>>> acceleration to work, SCA6000 needs to provide PKCS11 interfaces that 
>>> directly plugin into NSS.
>> NSS does not use the Solaris libpkcs11 (PLEASE don't use the TLA SCF 
>> it mapps to far to many things in Solaris and libscf has nothing to do 
>> with the crypto framework) by default but can be configured to do so 
>> using modutil(1).
>>
>> For Solaris hardware crypto like the SCA-6000 plugins in to the kernel 
>> part of the crypto framework and appears in userland via 
>> /usr/lib/libpkcs11.so.1
>>
>> So the SCA-6000 CAN be accessed via NSS using PKCS#11 just like the 
>> UltraSPARC T1/T2 on chip crypto.
>>
> 
> _______________________________________________
> crypto-discuss mailing list
> crypto-discuss at opensolaris.org
> http://mail.opensolaris.org/mailman/listinfo/crypto-discuss


Reply via email to