Glenn Brunette wrote:
>
> Given the strong push by U.S. and other governments, financial
> services organizations, etc. (inside and outside of the U.S.) to
> use FIPS approved algorithms, has there been any consideration
> to make FIPS-140 mode enabled by default? I realize that in a
> global marketplace, this is likely a touchy issue, but I at least
> wanted to put the question on the table and hear from the project
> team and the community.
I think enabling FIPS by default is not necessarily a good thing. While
I've not looked at this particular implementation, I know a few things
about FIPS in general:
1) some perfectly good and useful algorithms will be unavailable in
FIPS mode (e.g. RC5, Blowfish, etc.)
2) some algorithms that *should* be avoided, but are still widely
used (MD5) will be disabled in FIPS mode
3) there may be performance implications for FIPS mode (e.g. extra
... um... pointless... key wrapping operations might be performed
depending on the location of the crypto boundary, extra verification
tests, etc.)
Unless a piece of equipment/software is being sold primarily as a FIPS
solution, I think FIPS should not be enabled by default.
Note also that FIPS 140-2 is also endorsed by a number of other
countries, and is not just a US standard.
- Garrett
>
> g
>
> On 6/9/09 6:17 PM, Krishna Yenduri wrote:
>> I am sponsoring this fast track for Hai-May Chao. The timer
>> is set for 06/17/2009. Micro/patch binding is requested.
>>
>>
>> Template Version: @(#)sac_nextcase 1.68 02/23/09 SMI
>> This information is Copyright 2009 Sun Microsystems
>> 1. Introduction
>> 1.1. Project/Component Working Name:
>> cryptoadm(1M) enhancement for FIPS-140 mode
>> 1.2. Name of Document Author/Supplier:
>> Author: Hai-May Chao
>> Valerie Fenwick
>> Tony Scarpino
>> 1.3 Date of This Document:
>> 09 June, 2009
>>
>> 4. Technical Description
>>
>> 4.1 Proposal:
>>
>> Enhance cryptoadm interface to provide for enabling and disabling
>> of the FIPS-140 mode of operations in the Cryptographic Framework.
>>
>> 4.2 Description:
>>
>> The Cryptographic Framework team is planning on obtaining FIPS 140-2
>> certification. The cryptoadm command is the administrative front-end
>> interface to the framework. This case is intended to add new features
>> to cryptoadm(1M) that allow administrators to enable and disable the
>> FIPS-140 mode in the Cryptographic Framework. Hence, this case
>> represents the first set of changes to get prepared toward the FIPS
>> 140-2 evaluation process.
>>
>> There will be two FIPS-140 modes of operations in the framework: enabled
>> and disabled. The default FIPS-140 mode is disabled.
>>
>> When FIPS-140 mode is enabled, the Cryptographic Framework is put into
>> FIPS-140 mode of operations. The non-approved FIPS algorithms
>> provided by
>> the user-level pkcs11_softtoken provider and the kernel software
>> providers
>> will not be disabled. It is up to the consumers of the framework to be
>> responsible for using only FIPS approved algorithms and that will be
>> documented in the Security Policy. This meets FIPS 140 level 2
>> requirements.
>>
>> As we start working with the certification lab, we anticipate there may
>> be additional changes needed and those changes should be internal to the
>> framework. The cryptoadm interface changes should stand by itself.
>>
>> The cryptoadm command will also be modified to display the active
>> FIPS-140 mode setting.
>>
>> 4.3 Interfaces:
>>
>> The following new options are added to cryptoadm(1M) sub-commands
>> cryptoadm list fips-140
>> cryptoadm enable fips-140
>> cryptoadm disable fips-140
>>
>> Stability level is "committed".
>> Release binding is Micro/Patch.
>>
>>
>> 4.4 Doc Impact:
>>
>> The diff-marked cryptoadm(1M) man page is in the case directory.
>>
>> 5. Reference
>>
>> FIPS 140-2 Spec can be located at:
>> http://csrc.nist.gov/publications/PubsFIPS.html
>>
>> 6. Resources and Schedule
>> 6.4. Steering Committee requested information
>> 6.4.1. Consolidation C-team Name:
>> ON
>> 6.5. ARC review type: FastTrack
>> 6.6. ARC Exposure: open
>>
