Month: December 2010

You may have recently read our blog post regarding Exposed Private Keys and how you may create secure private keys for SSL-enabled devices using wolfSSL. If your application or framework needs to be secured with SSL/TLS but you are uncertain about how the optimal design of a secured system would be structured, we can help!

We offer design consulting for building SSL/TLS security into devices using wolfSSL. Our consultants can provide you with the following services:

1. Assessment:  An evaluation of your current SSL/TLS implementation. We can give you advice on your current setup and how we think you could improve upon this by using wolfSSL.

2. Design:  Looking at your system requirements and parameters we`ll work closely with you to make recommendations on how to implement wolfSSL into your application such that it provides you with optimal security.

If you would like to learn more about design consulting for building SSL into your application or device, please contact info@yassl.com for more information.

So what made it into the wolfSSL Holiday Release?  We think a few items that users will appreciate.  Some concern increased portability through the os_settings.h file and the new C Standard Library Abstraction layer.  Others lower memory use through configurable input/output buffer sizes and less dynamic memory use.  For the readers out there we now have a complete manual available on the website as well as in the release.  wolfSSL can now create CA signed x509v3 certificates too.  There`s also the ability to use new cipher suites in conjunction with a library we`ll be talking about soon.  Happy Holidays from the yaSSL team!

We saw a post on a mailing list recently that questioned the maturity of wolfSSL.  The post came as a surprise.  We think wolfSSL is pretty mature by most standards!  Here’s our key points on wolfSSL’s maturity:
 
1. It is on its 29th release.
2. It has been vetted out in the open source community for 6 years. (View our community page)
3. It is distributed in over 20 million devices, based on our estimates.  That includes commercial distributions such as printers, home monitoring, VoIP phones, hubs, routers, sensors, telecom infrastructure, as well as open source projects.
 
Let us know what you think!  If you think wolfSSL is mature/not mature, email us at info@yassl.com, or post to our forums.

A recent blog about “Breaking SSL on Embedded Devices” has been making the rounds recently.  It`s not an attack on SSL, it`s an attack on weak private key security. Specifically, embedding a private key into firmware allows anyone to extract the key and turns an otherwise secure connection into something nothing more secure than TCP.

We have a few ideas about creating private keys for SSL enabled devices.

1. Each device acting as a server should have a unique private key, just like in the non-embedded world.

2. If the key can`t be placed onto the device before delivery, have it generated during setup.

3. If the device lacks the power to generate it`s own key during setup have the client setting up the device generate the key and send it to the device.

4. If the client lacks the ability to generate a private key have the client retrieve a unique private key over an SSL connection from the devices known website for example.

wolfSSL can be used in all of these steps to help ensure an embedded device has a secure unique private key.  That will go a long ways towards securing the SSL connection itself.

Will it be a new release with awesome new embedded ssl features?  What could those features possibly be?  Will it be even faster and smaller?  More portable?  New ciphers?  Do you think you know what the yaSSL community is getting for Christmas this year?  Send us your guess to info@yassl.com!

Here’s a nice article on Android’s security flaws and latent bugs from the CTO of Green Hills Software, Dave Kleidermacher :  https://www.embedded.com/electronics-blogs/davek-security-blog/4211444/Angry-Bird-droppings.
 
Looks like there is some work to be done to help properly secure Android!

wolfSSL is modular. We’ve got two key modules: wolfSSL handles all SSL needs while CTaoCrypt handles all cryptographic needs including block ciphers, stream ciphers, message digests, hashing, public key cryptography, certificates, and various helper utilities.  Our topic for this post is our crypto module CTaoCrypt. 
 
CTaoCrypt is our implementation of cryptography components including various block and stream ciphers that can be used underneath the SSL protocol. A complete description of our ciphers is available here:

https://www.wolfssl.com/docs/wolfssl-manual/ch10/. 

Not all ciphers are created equally. Some are better than others, depending on the task at hand. CTaoCrypt includes the traditional and most popular block ciphers, which are DES, 3DES, and AES, and the popular stream cipher ARC4. In recent years, the European Union funded the eStream project to identify next generation stream ciphers. The project concluded in 2008. A description of the eStream project is found here:  http://en.wikipedia.org/wiki/ESTREAM. The eStream project was looking to identify ciphers that could handle high throughput requirements and/or operate well in resource constrained environments. 
 
High throughput and resource constrained environments (like VoIP on devices) is where wolfSSL maintains a technology lead over the pack of SSL libraries, so it was natural for us to maintain a keen interest in the results from eStream. The net result is our implementation of the HC-128 and RABBIT ciphers in wolfSSL. If you’d like to see a comparison of cipher performance, see slide 16 of our OSCON presentation http://www.yassl.com/files/ToddOuska_SecuringMemcache.pdf, showing relative performance of the cipher suites in specific implementation of Secure Memcache. 
 
Other ciphers can be added to CTaoCrypt for additional purposes, and we’re happy to do it if we see a need from our customers. Watch this space for further announcements! 

This article from Embedded Internet Design https://www.eetimes.com/author.asp?section_id=36&doc_id=1265936 gives excellent coverage on the recent and upcoming technologies in the Mobile Internet Tsunami.  We’d like to extrapolate from the article and think about some of the less visible, less glamorous segments of the market that will change dramatically due to the innovation and investment brought about by the Tsunami. 
 
Here’s our thinking:
 
1. We agree with the article.  The waves in this Tsunami are just getting bigger and bigger.
2. The visible market driver is consumer adoption.
3. Chips, boards, and systems costs get driven down and horsepower goes up, really fast…of course.  We’re all used to that.
4. The less visible segments of the market get driven along by the investment and innovation in the Tsunami.
5. This means dramatic changes in invisible computing, and of course the security for these invisible devices.
 
One of the less visible markets we’ve noticed changing fast is sensors in the industrial automation, physical security, and process driven industries.  These users deploy a lot of chips, so they absolutely, positively have to be low cost designs.  That means resource constraints, low memory usage, and tight code. 
 
In the sensor space, 1.0 designs just did the sensing, collection and storage of data.  2.0 designs added wireless connectivity.  The 3.0 designs we’ve been involved with add security and data transmission straight to the cloud!  Not only do we think that’s cool, we think it is effective and efficient design. 
 
Us mobile and embedded guys live in exciting times that are changing fast!  Do you have thoughts on how the Tsunami affects the less visible parts of the embedded computing infrastructure?  Let us know what you think at info@yassl.com.

As stated in the TLS 1.1 and 1.2 protocol definitions (RFC 4346, RFC 5246), “The primary goal of the TLS protocol is to provide privacy and data integrity between two communicating applications.” TLS 1.2 is an improvement to the TLS 1.1 standard, but how exactly do they differ? What was changed in TLS 1.2 to warrant a new version of the protocol?

Listed below are the changes made to both version 1.1 and 1.2 of the TLS protocol. TLS 1.2 support is slowly making it’s way into existing projects. wolfSSL embedded SSL/TLS fully supports SSL 3.0, TLS 1.0, TLS 1.1, and TLS 1.2.

A. TLS 1.1
This protocol was defined in RFC 4346 in April of 2006, and is an update to TLS 1.0. The major changes are:
– The Implicit Initialization Vector (IV) is replaced with an explicit IV to protect against Cipher block chaining (CBC) attacks.
– Handling of padded errors is changed to use the bad_record_mac alert rather than the decryption_failed alert to protect against CBC attacks.
– IANA registries are defined for protocol parameters
– Premature closes no longer cause a session to be non-resumable.

RFC 4346: http://tools.ietf.org/html/rfc4346#section-1.1

B. TLS 1.2
This protocol was defined in RFC 5246 in August of 2008. Based on TLS 1.1, TLS 1.2 contains improved flexibility. One of the primary goals of the TLS 1.2 revision was to remove the protocol’s dependency on the MD5 and SHA-1 digest algorithms. The major differences include:
– The MD5/SHA-1 combination in the pseudorandom function (PRF) was replaced with cipher-suite-specified PRFs.
– The MD5/SHA-1 combination in the digitally-signed element was replaced with a single hash. Signed elements include a field explicitly specifying the hash algorithm used.
– There was substantial cleanup to the client`s and server`s ability to specify which hash and signature algorithms they will accept.
– Addition of support for authenticated encryption with additional data modes.
– TLS Extensions definition and AES Cipher Suites were merged in.
– Tighter checking of EncryptedPreMasterSecret version numbers.
– Many of the requirements were tightened
– Verify_data length depends on the cipher suite
– Description of Bleichenbacher/Dlima attack defenses cleaned up.
– Alerts must be sent in many cases
– After a certificate_request, if no certificates are available, clients now MUST send an empty certificate list.
– TLS_RSA_WITH_AES_128_CBC_SHA is now the mandatory to implement cipher suite.
– Added HMAC-SHA256 cipher suites.
– Removed IDEA and DES cipher suites. They are now deprecated.

RFC 5246: http://tools.ietf.org/html/rfc5246

C. Goals of the TLS Protocol

– Cryptographic security: TLS should be used to establish a secure connection between two parties.
– Interoperability: Independent programmers should be able to develop applications utilizing TLS that can successfully exchange cryptographic parameters without knowledge of one another`s code.
– Extensibility: TLS seeks to provide a framework into which new public key and bulk encryption methods can be incorporated as necessary. This will also accomplish two sub-goals: preventing the need to create a new protocol (and risking the introduction of possible new weaknesses) and avoiding the need to implement an entire new security library.
– Relative efficiency: Cryptographic operations tend to be highly CPU intensive, particularly public key operations. For this reason, the TLS protocol has incorporated an optional session caching scheme to reduce the number of connections that need to be established from scratch. Additionally, care has been taken to reduce network activity.

Resources:
If you would like to read more about SSL or TLS, here are several resources that might be helpful:
TLS – Wikipedia (http://en.wikipedia.org/wiki/Transport_Layer_Security)
SSL versus TLS – What`s the Difference? (http://luxsci.com/blog/ssl-versus-tls-whats-the-difference.html)
Cisco – SSL: Foundation for Web Security (http://www.cisco.com/web/about/ac123/ac147/archived_issues/ipj_1-1/ssl.html)

As always, if you have any questions or would like to talk to the wolfSSL team about more information, please contact facts@wolfssl.com.

As participants in securing the mobile internet tsunami, we’ve been keeping a keen eye on the future security needs of the next generation of devices.  We view it from both project involvement and product management for our embedded SSL and embedded web server products.    This article https://www.eetimes.com/author.asp?section_id=36&doc_id=1265933 at Embedded Internet Design gives a view of what’s coming next in smartphones. 
 
If you’re also guy or a gal out there tinkering with this stuff and testing your ideas with an eye on security, take a look at the beagleboard, pandaboard, and mbed.  wolfSSL and the yaSSL Embedded Web server will be there for you!