Category Archives: Applications

CISA Gives Us A Starting Point

Posted on by

Everyone struggles with identifying the “right” vulnerabilities to patch. It’s a near universal problem. We are all wondering, “Which vulnerabilities in my long queue are going to get me?”

Most organizations calculate a CVSS (base score or amended) which a solid body of research starting with Alloddi & Massacci, 2014, demonstrates is inadequate to the task.

Exploit Prediction Scoring System (EPSS) is based in the above research, so it could provide a solution. But the web calculator cannot score each open vulnerability in our queue over time: we need to watch the deltas, not a point in time. There’s unfortunately, no working EPSS solution, despite the promise.

CISA have listed 300 “vulnerabilities -currently- being exploited in the wild” in Binding operational directive 22-01.

Finally, CISA have given us a list that isn’t confined to a “Top 10”: Start with these!

Top 10 lists provide some guidance, but what if attackers exploit your #13?

300 is an addressable number in my experience. Besides, you probably don’t have all of them in your apps. > 300. We all can do that much.

The CISA list provides a baseline, a “fix these, now” starting point of actively exploited issues that cuts through the morass of CVSS, EPSS, your security engineer’s discomfort, total vulnerabilities ==> organization “risk”. (If you’re running a programme based upon the total number of open vulnerabilities, you’re being misled.)

https://thehackernews.com/2021/12/why-everyone-needs-to-take-latest-cisa.html

Trojan Source

Posted on by

Trojan Source!

I first read about Trojan Source this morning (ugh, Yet Another Branded Vulnerability: YABV).

Yes, there is a continuing fire hose of vulnerability announcements. But, new techniques are actually fairly rare: 1-3/year, in my experience.

There is nothing new about hiding attack code in source. These are the “backdoors” every software security policy forbids. Hiding attack code in unicode strings by abusing the “bidi override” IS new and important (IMVHO). Brilliant research for which I’m grateful.

I’m a bit unclear about how attack code can supposedly be hidden through bidi override in comments. Every decent compiler strips comments during the transformation from source to whatever form the compiler generates, byte code (java), object, etc. Comments are NOT part of any executable format, so far as I know. And mustn’t be (take up needed space).

Still, attacks buried in string data is bad enough to take notice: every language MUST include string data; modern languages are required to handle unicode, the many language scripts that must be represented, and methods for presenting each script (i.e., bidi override). Which means exactly what the researchers found: nearly every piece of software that parses source into something executable (19 compilers and equivalent) is going to have this problem.

The other interesting research finding is how vendors reacted to being told about Trojan Source and how quickly they are fixing, if at all.

Only 9 vendors have committed to a fix. That is really surprising to me. If you want people to trust your software security, you’ve got to be rigorously transparent and responsive. Doh! What’s wrong with the other 10 vendors/makers? Huh? Let’s get with the program. Refusing to deal, dragging feet, or worse, ignoring appears as though you don’t care about the security of your product(s).

It’ll be interesting to see whether Trojan Source technique actually gets used by attackers and how fast that occurs. I hope that one or more researchers pay attention to that piece of the puzzle. Depending upon the research findings, at least (if not more) 75% of reported vulnerabilities never get exploited. Will Trojan Source? Only time will tell.

Many thanks to Brian Krebs for the excellent write up (below)

https://krebsonsecurity.com/2021/11/trojan-source-bug-threatens-the-security-of-all-code/

Cheers,

/brook

Is API Authentication Enough?

Posted on by

Reading a recent article, Five Common Cloud Security Threats and Data Breaches“, posted by Andreas Dann I came across some advice for APIs that is worth exploring.

The table stakes (basic) security measures that Mr. Dann suggests are a must. I absolutely support each of the recommendations given in the article. In addition to Mr. Dann’s recommendations, I want to explore the security needs of highly exposed inputs like internet-available APIs.

Again and again, I see recommendations like Mr. Dann’s: “…all APIs must be protected by adequate means of authentication and authorization”. Furthermore, I have been to countless threat modelling sessions where the developers place great reliance on authorization, often sole reliance (unfortunately, as we shall see.)

For API’s with large open source, public access, or which involve relatively diverse customer bases, authentication and authorization will be insufficient.

As True Positives‘ AppSec Assurance Strategist, I’d be remiss to misunderstand just what protection, if any, controls like authentication and authorization provide in each particular context.

For API’s essential context can be found in something else Mr. Dann states, “APIs…act as the front door of the system, they are attacked and scanned continuously.” One of the only certainties we have in digital security is that any IP address routable from the internet will be poked and prodded by attackers. Period.

That’s because, as Mr. Dann notes, the entire internet address space is continuously scanned for weaknesses by adversaries. Many adversaries, constantly. This has been true for a very long time.

So, we authenticate to ensure that entities who have no business accessing our systems don’t. Then, we authorize to ensure that entities can only perform those actions that they should and cannot perform actions which they mustn’t. That should do it, right?

Only if you can convince yourself that your authorized user base doesn’t include adversaries. There lies “the rub”, as it were. Only systems that can verify the trust level of authorized entities have that luxury. There actually aren’t very many cases involving relatively assured user trust.

As I explain in Securing Systems, Secrets Of A Cyber Security Architect, and Building In Security At Agile Speed, many common user bases will include attackers. That means that we must go further than authentication and authorization, as Mr. Dann implies with, “…the API of each system must follow established security guidelines”. Mr. Dann does not expand on his statement, so I will attempt to provide some detail.

How do attackers gain access?

For social media and fremium APIs (and services, for that matter), accounts are given away to any legal email address. One has only to glance at the number of cats, dogs, and birds who have Facebook accounts to see how trivial it is to get an account. I’ve never met a dog who actually uses, or is even interested in Facebook. Maybe there’s a (brilliant?) parrot somewhere? (I somehow doubt it.)

Retail browsing is essentially the same. A shop needs customers; all parties are invited. Even if a credit card is required for opening an account, which used to be the norm, but isn’t so much anymore, stolen credit cards are cheaply and readily available for account creation.

Businesses which require payment still don’t get a pass. The following scenario is from a real-world system for which I was responsible.

Customers paid for cloud processing. There were millions of customers, among which were numerous nation-state governments whose processing often included sensitive data. It would be wrong to assume that nation-states with spy agencies could not afford to purchase a processing account in order to poke and prod just in case a bug should allow that state to get at the sensitive data items of one of its adversaries. I believe (strongly) that it is a mistake to assume that well-funded adversaries won’t simply purchase a license or subscription if the resulting payoff is sufficient. spy agencies often work very slowly and tenaciously, often waiting years for an opportunity.

Hence, we must assume that our adversaries are authenticated and authorized. That should be enough, right? Authorization should prevent misuse. But does it?

All software fails. Rule #1, including our hard won and carefully built authorization schemes.

“[E]stablished security guidelines” should expand to a robust and rigorous Security Development Lifecyle (SDL or S-SDLC), as James Ransome and I lay out in Building In Security At Agile Speed. There’s a lot of work involved in “established security”, or perhaps better termed, “consensus, industry standard software security”. Among software security experts, there is a tremendous amount of agreement in what “robust and rigorous” software security entails.

(Please see NIST’s Secure Software Development Framework (SSDF) which is quite close to and validates the industry standard SDL in our latest book. I’ll note that our SDL includes operational security. I would guess that Mr. Dann was including typical administrative and operational controls.)

Most importantly, for exposed interfaces, I continue to recommend regular, or better, continual fuzz testing.

The art of software is constraining software behaviour to the desired and expected. Any unexpected behaviour is probably unintended and thus, will be considered a bug. Fuzzing is an excellent way to find unintended programme behaviour. Fuzzing tools, both open source and commercial, have gained maturity while at the same time, commercial tools’ price has been decreasing. The barriers to fuzzing are finally disappearing.

Rest assured that you probably have authorized adversaries amongst your users. That doesn’t imply that we abandon authentication and authorization. These controls continue to provide useful functions, if only to tie bad behaviour to entities.

Along with authentication and authorization, we must monitor for anomalous and abusive behavour for the inevitable failures of our protective software. At the same time, we must do all we can to remove issues before release, while at the same time, using techniques like fuzzing to identify those unknown, unknowns that will leak through.
 

Cheers,

/brook

Continuing Commentary on NIST AppSec Guidance

Posted on by

Continuing with a commentary on NIST #appsecurity guidance, again, NIST reiterates what we recommended in Core Software Security and then updated (for Agile, CI/CD, DevOps) in Building In Security At Agile Speed:

 

Threat Model

Automate as much as possible

Static analysis (or equivalent) SAST/IAST

DAST (web app/api scan)

Functional tests

Negative testing

Fuzz

Check included 3rd party software

 

NIST guidance call’s out heuristic hard coded secrets discovery. Definitely a good idea. I maintain that it is best to design in protections for secrets through threat modelling. Secrets should never be hardcoded without a deep and rigorous risk analysis. (there are a few cases I’ve encountered where there was no better alternative. these are few and very far between.)

NIST guidance is essentially the same as our Industry Standard, Generic Security Development Life cycle (SDL).

There has been an implicit industry consensus on what constitutes effective software security for quite a while. Let’s stop arguing about it, ok? Call the activities whatever you like. but do them.

Especially, fuzz all inputs that will be exposed to heavy use (authenticated or not!) I’ve been saying this for years. I hope that fuzzing tools are finally coming up to the our needs?

https://www.nist.gov/itl/executive-order-improving-nations-cybersecurity/recommended-minimum-standard-vendor-or-developer

MITRE D3FEND Is Your Friend

Posted on by

Defenders have to find the correct set of defences for each threat. Many attacks have no direct prevention making a many-to-to-many relationship. Threat:defence == M:N 

Certainly some vulnerabilities can be prevented or mitigated with a single action. But overall, there are other threats require an understanding of the issue and which defences might be applied.

While prevention is the holy grail, often we must:

— Limit access (in various ways and at various levels in the stack)
— Mitigate effects
— Make exploitation more difficult (attacker cost)

MITREcorp D3FEND maps defences to attacks

“…tailor defenses against specific cyber threats…”

In my threat modelling classes, the 2 most difficult areas for participants are finding all the relevant attack scenarios and then putting together a reasonable set of defences for each credible attack scenario. Typically, newbie threat modellers will identify only a partial defence. Or, misapply a defence.

MITRE ATT&CK helps to find all the steps an attacker might take from initial contact to compromise.

MITRE D3FEND and the ontology between ATT&CK and D3FEND give defenders what we need to then build a decent defence.

https://www.csoonline.com/article/3625470/mitre-d3fend-explained-a-new-knowledge-graph-for-cybersecurity-defenders.html#tk.rss_all

Security Research Is Threat Modeling’s Constructive Criticism

Posted on by

(the following is partially excerpted from my next book)

Adam Shostack, author of Threat Modeling: Designing for Security has succinctly described a process of constructive critique within engineering:

“The back and forth of design and critique is not only a critical part of how an individual design gets better, but fields in which such criticism is the norm advance faster.”

The Spectre/Meltdown issues are the result of a design critique such as Shostack describes in his pithy quote given above.  In fact, one of the fundamental functions that I believe security research can play is providing designers with a constructive critique.

Using Spectre and Meltdown as an example of engineering critique, let’s look at some of the headlines from before the official issue announcement by the researchers:

“Kernel-memory-leaking Intel processor design flaw forces Linux, Windows redesign”, John Leyden and Chris Williams 2 Jan 2018 at 19:29, The Register, https://www.theregister.co.uk/2018/01/02/intel_cpu_design_flaw/

“A CRITICAL INTEL FLAW BREAKS BASIC SECURITY FOR MOST COMPUTERS”, Andy Greenberg, January 3, 2018, Wired Magazine, https://www.wired.com/story/critical-intel-flaw-breaks-basic-security-for-most-computers/

There were dozens of similar headlines (many merely repeating the first few, especially, The Register’s), all declaiming a “flaw” in CPUs. I want to draw the reader’s attention to the word, “flaw”. Are these issues “flaws”? Or, are they the result of something else?

The Register headline and that first article were based upon speculation that had been occurring amongst the open source community supporting the Linux Kernel. A couple of apparently odd changes that had been made to kernel code. But, the issues to which these changes responded were “embargoed”, that is, the reasoning behind the changes was known only to those making the changes.

Unlike typical open source changes, whose reasoning is public and often discussed by members of the community, these kernel changes had been made in opaquely without public comment, which of course, set concerned kernel community members wondering.

To observers not familiar with the reasoning behind the changes, it was clear that something was amiss and likely in relation to CPU functions; anxious observers were guessing what might be the motivation for those code changes.

Within the digital security universe, there exists an important dialog between security research aimed at discovering new attack techniques and the designers of the systems and protocols upon which that research is carried out. As Adam noted so very wryly, achieving solid designs, even great ones, and most importantly, resilient designs in the face of omnipresent attack requires an interchange of constructive critique. That is how Spectre and Meltdown were discovered and presented.

Neither of this collection of (at the time of announcement) new techniques involved exercising a flaw, that is, a design error – in other words, the headlines quoted just above were erroneous and rather misleading[2].

Speculative execution and the use of kernel mapped user memory pages by operating systems were intentional design choices that had been working as designed for more than 10 years. Taken together, at least some of the increases in CPU performance over that period can directly be tied to speculative execution design.

Furthermore, and quite importantly to this discussion, these design choices were made within the context of a rather different threat landscape. Some of today’s very active threat actors didn’t exist, or at least, were not nearly as active and certainly not as technically sophisticated circa 2005 as they are today, May 2018.

If I recall correctly (and I should be able to remember, since I was the technical lead for Cisco’s Web infrastructure and application security team at that time), in 2005, network attacks were being eclipsed by application focused attack methods, especially, web attack methods.

Today, web attacks are very “ho, hum”, very run of the ordinary, garden variety. But in 2005, when the first CPU speculative execution pipelines were being released, web applications were targets of choice at the cutting edge of digital security. Endpoint worms and gaining entrance through poor network ingress controls had been security’s focus up until the web application attack boom (if I may title it so?). At about that time, targeting web applications was fast displacing network periphery concerns. Attackers were in the process of shifting to targets that used a standard protocol (HTTP) which was guaranteed to pass through organizational firewalls. Many of the new targets’ were becoming always  available via the Public Internet.

Since the web application attack boom, attacks and targets have continued to evolve. The threat landscape changed dramatically over the years since the initial design of speculative execution CPUs. Alongside the changes in types of attackers as well as their targets, attacker and researcher sophistication has grown, as has an attackers’ toolbox. 2018 is a different security world than 2005. I see no end to this curve of technical growth in my crystal ball.

The problem is, when threat modeling, whether in 2005 or 2018, one considers the attacks of the past, those of moment, and then one must try one’s best to project from current understanding to attacks that might arise within the foreseeable future. Ten or twelve years seems an awfully long horizon of prescience, especially when considering the rate at which technical change continues to take place.

As new research begins to chew at the edges of any design, I believe that the wise and diligent practitioner revisits their existing threat models in light of developments.

If I were to fault the CPU and operating system makers whose products are subject to Spectre or Meltdown, it would be for a failure to anticipate where research might lead, as research has unfolded. CPU threat modelers could have taken into account advances in research indicating unexpected uses of cache memory.

Speculative execution leaves remnants of a speculated execution branch in cache memory when a branch has not been taken. It is those remnants that lie at the heart of this line of research.

A close examination of the unfolding research might very well have led those responsible for updating CPU threat models to consider the potential for something like Spectre and Meltdown. (Or, perhaps the threat models were updated, but other challenges prevented updates to CPU designs? CPU threat modelers, please tell us all what the real story is)

I’ve found a publication chain during the 3 years previous that, to me, points towards the new techniques. Spectre and Meltdown are not stand-alone discoveries, but lie on a body of CPU research that had been published regularly for several years.

As I wrote for McAfee’s Security Matters blog in January of 2018 (as a member of McAfee’s Advanced Threat Research Team),

“Meltdown and Spectre are new techniques that build upon previous work, such as “KASLR”  and other papers that discuss practical side-channel attacks. The current disclosures build upon such side-channels attacks through the innovative use of speculative execution….An earlier example of side-channel based upon memory caches was posted to Github in 2016 by one of the Spectre/Meltdown researchers, Daniel Gruss.” Daniel Gruss is one of the Spectre and Meltdown paper authors.

Reading these earlier papers, it appears to me that some of the parent techniques that would be used for the Spectre and Meltdown breakthroughs could have been read (should have been read?) by CPU security architects in order to re-evaluate the CPU’s threat model. That previously published research was most certainly available.

Of course, hindsight is always 20/20; I had the Spectre and Meltdown papers in hand as I reviewed previous research. Going the other way might be more difficult?

Spectre and Meltdown did not just spring miraculously from the head of Zeus, as it were. They are the results of a fairly long and concerted effort to discover problems with and thus, hopefully, improve the designs of modern processors. Indeed, the researchers engaged in responsible disclosure, not wishing to publish until fixes could be made available.

To complete our story, the driver that tipped the researchers to an early, zero-day disclosure (that is, disclosure without available mitigations or repairs) were the numerous speculative (if you’ve forgive the pun?) journalism (headlines quoted above) that gained traction based upon misleading (at best) or wrong conclusions. Claiming a major design “flaw” in millions of processors is certainly a reader catching headline. But, unfortunately, these claims were vastly off the mark since no flaw existed in the CPU or operating system designs.

While it may be more “interesting” to imagine a multi-year conspiracy to cover up known design issues by evil CPU makers, no such conspiracy appears to have taken place.

Rather, in the spirit of responsible disclosure, the researchers were waiting for mitigations to be made available to customers; CPU manufacturers and operating system coders were heads down at work figuring out what appropriate mitigations might be, and just how to implement these with the least amount of disruption. None of these parties was publicly discussing just why changes were being made, especially to the open source Linux kernel.

Which is precisely what one would expect in order to protect millions of CPU users: embargo the technical details to foil attackers. There is actually nothing unusual about such a process; it’s all very normal and typical, and unfortunately for news media, quite banal[3].

What we see through the foregoing example about Spectre and Meltdown is precisely the sort of rich dialog that should occur between designers and critics (researchers, in this case).

Designs are built against the backdrop and within the context of their security “moment”. Our designs cannot improve without collective critique amongst the designers; such dialog internal to an organization or at least, a development team is essential. I have spoken about this process repeatedly at conferences: “It takes a village to threat model,” (to misquote a famous USA politician.)

But, there’s another level, if you will, that can reach for a greater constructive critique.

Once a design is made available to independent critics, that is, security researchers, research discoveries can and I believe, should become part of an ongoing re-evaluation of the threat model, that is, the security of the design. In this way, we can, as an industry, reach for the constructive critique called for by Adam Shostack.

[1]In my humble experience, Adam is particularly good at expressing complex processes briefly and clearly. One of his many gifts as a technologist and leader in the security architecture space.

[2]Though salacious headlines apparently increase readership and thus advertising revenue. Hence, the misleading but emotion plucking headlines.

[3]Disclosure: I’ve been involved in numerous embargoed issues over the years.

Why no “S” in IoT?

Posted on by

My friend, Chris Romeo, a security architect and innovator for whom I have deep respect and from whom I learn a great deal just posted a blog about the lack of security (“S”) in Internet of Things (IoT), “The S In IoT Stands For Security“.

Of course, I agree with Chris about his three problem areas:

  1. Lack of industry knowledge
  2. Developer’s lack of security knowledge and understanding
  3. Time to market pressure, particularly as it exists for startups

I don’t think one can argue with these; in fact, these problems have been and continue to be pervasive, well beyond and before the advent of IoT. Still, I would like to add two more points to Chris’ explanation of the problem space, if I may add to what I hope will be an ongoing and lively conversation within the security industry?

First, IoT products are not just being produced at startups; big players are building IoT, too. Where we place sensors and for what purposes is a universe of different solutions, all the way from moisture sensors in soil (for farming), through networkable household gadgets like light bulbs and thermostats, to the activity monitor on a human’s wrist, to wired up athletes, to the by now famous, mirai botnet camera.

Differing use cases involve varying types and levels (postures) of security. It is a mistake to lump these all together.

When at Intel, I personally reviewed any number of IoT projects that involved significant security involvement, analysis, and implementation. I’m going to guess that most of the major activity monitoring server-side components should have had quite a lot of basic web and cloud security in-built? At least for some products, interaction between the device (IoT sensor) and any intermediary device, like a smart phone or personal computer, has also been given some significant security thought – and it turns out, that this is an area where security has been improving as device capabilities have grown.

It’s hard to make an all out statement on the state of IoT security, though I believe that there are troublesome areas, as Chris points out, most especially for startups. Another problem area continues to be medical devices, which have tended to lag badly because they’ve been produced with more open architectures, as has been presented at security conferences time and again. Then, there’s the Jeep Wrangler remote control hack.

Autonomous autos are going to push the envelop for security, since it’s a nexus between machine learning, big (Big BIG) data, consumer convenience, and physical safety. The companies that put the cars together are generally larger companies, though many of these do not lead in the cyber security space. Naturally, the technology that goes into the cars comes from a plethora of concerns, huge, mid-sized, tiny, and startup. Security consciousness and delivery will be a range from clueful to completely clueless across that range. A repeating problem area for security continues to be complex integrations where differing security postures are pasted together without sufficient thought about how interactions are changing the overall security of the integrated system.

Given the foregoing, I believe that it’s important to qualify IoT somewhat, since context, use case, and semantics/structure matter.

Next, and perhaps more important (save the most important for last?) are the economics of development and delivery. This is highlighted best by the mirai camera: it’s not just startups who have pressure. In the camera manufacturer’s case (as I understand it) they have near zero economic incentive to deliver a secure product. And this is after the mirai attack!

What’s going on here? Importantly, the camaras are still being sold. The access point with a well-known, default password is presumably still on the market, though new cameras may have been remediated. Security people may tear their hair out with an emphatic, “Don’t ever do that!” But, consider the following, please.

  • Product occurs distant (China, in this case) from the consumption of the product
  • Production and use occur within different (vastly different, in this case) legal jurisdictions
  • The incentive is to produce product as cheaply as possible
  • Eventual users do not buy “security” (whatever that means to the customer) from the manufacturer
  • The eventual user buys from manufacturer’s customer or that customer’s customer (multiple entities between consumer and producer)

Where’s the incentive to even consider security in the above picture? I don’t see it.

Instead, I see, as a Korean partner of a company that I once worked for said about acquiring a TCP/IP stack, “Get Internet for free?” – meaning of course, use an open source network stack that could be downloaded from the Internet with no cost.

The economics dictate that, manufacturer download an operating system, usually some Linux variant. Add an working driver, perhaps based upon an existing one? Maybe there’s a configuration thingy tied to a web server that perhaps, came with the Linux. Cheap. Fast. Let’s make some money.

We can try to regulate this up, down, sideways, but I do not believe that this will change much. Big companies will do the right thing, and charge more proportionately. Startups will try to work around the regulations. Companies operating under different jurisdictions or from places where adherence is not well policed or can be bribed away will continue to deliver default passwords to an open source operating system which delivers a host ripe for misuse (which is what mirai turns out to be).

Until we shift the economics of the situation, nothing will change. In the case of mirai, since the consumer of the botnet camera was likely not affected during that attack, she or he will not be applying pressure to the camera manufacturer, either. The manufacturer has little incentive to change; the consumer has little pressure to enforce via buying choices.

By the way, I don’t see consumers becoming more security knowledgeable. Concerned about digital security? Sure. Willing to change the default password on a camera or a wifi router? Many consumers don’t even go that far.

We’re in a bind here; the outlook does not look rosy to me. I open to suggestions.

Thanks, Chris, for furthering this discussion.

cheers,

/brook

Heartbleed Exposure, What Is It Really?

Posted on by

Heartbleed Exposure, what is it really?

“Heap allocation patterns make private key exposure unlikely” Neel Mehta, discoverer of HeartBleed” 

In the media, there’s been a lot of discussion about what might be exposed from the heartbleed OpenSSL attack. It is certainly true that very sensitive items can be exposed. And over thousands of test runs, sensitive items like private keying materials and the like have been returned by the heartbleed buffer overread.

A very strong case can be made for doing exactly as industry due diligence suggests. Teams should replace private keys on servers that had been vulnerable, once these are patched. But should every person on the Internet change every password? Let’s examine that problems by digging into the details of exactly how heartbleed works.

First, heartbleed has been characterized as an “overflow” error: “Heartbleed is basically a buffer-overflow vulnerability”. This unfortunately is a poor descriptor and somewhat inaccurate. It may make better media copy, but calling heartbleed an “overflow” is a poor technical description upon which to base a measured response.

Heartbleed is not a classic buffer overflow. No flow control or executable code may be injected via heartbleed. A read of attacker chosen memory locations is not possible, as I will explain, below. A better descriptor of heartbleed is a “buffer over-read”. Unintentionally, some data from memory is returned to the attacker. To be precise, heartbleed is a data leak, not a flow control error.

In order to understand what’s possible to disclose, it’s key to understand program “heap” memory. The heap is an area of memory that programs use to store data. Generally speaking, well-written programs (like OpenSSL) do not to put executable code into heap (that is, data) memory[1]. Because data and execution are separated, the attacker has no way through this vulnerability to execute code. And that is key, as we shall see.

As a program runs, bits of data, large and small, temporary and more or less permanent for the run, are put into the heap[2]. Typically, data are put wherever is convenient at the moment of allocation, depending upon what memory is available.

Memory that’s been deallocated gets reused. If an available piece of memory happens to be larger than a requested size, the new sized piece will be filled with the new data, while adjacent to the new data will remain bits and pieces of whatever was there previously.

In other words, while not entirely random, the heap is filled with bits and pieces of data, a little from here, a little from there, a nice big chunk from this session, with a bit left over from some other session, all helter-skelter amongst each other. The heap is a jumble; taking random bits from the heap may be considered to be like attending a jumble sale.

Now, let’s return to heartbleed. The heartbleed bug returns whatever happens to be on the heap just above the 16 bytes that are required for the TLS heartbeat packet. The attacker may request as much as 64K bytes. That’s a nice big chunk of stuff from the heap; make no mistake about it. Anything might be in there. At the very least, decrypted  data intended for application processing will be returned to the attacker[3]. That’s certainly bad! It breaks the confidentiality supposedly gained through the TLS encryption. But getting a random bit is different than requesting an arbitrary memory location at the discretion of the attacker. And that is a very important statement to hold in mind as we respond to this very serious situation.

An analogy to Heartbleed might be a bit like going fishing. Sometimes, we fish where we can clearly see the fish (mountain streams) or signs of fish (clearer lakes), or with a “fish finder” appliance, that identifies fish  under the surface when the fish aren’t visible.

Heartbleed is a lot more like fishing for fish that are deep in a turbulent lake with no fish finding capability. The fisher is guessing. If she or he guesses correctly, fish for dinner. If not, it’s a long day holding onto the fishing rod.

In the same manner, the attacker, the “fisher” as it were, doesn’t know where the “fish”, the goodies are. The bait (the heartbleed request) is cast upon the “lake” (the program heap) in the hopes that a big fish will “bite” (secret “bytes” will get returned).

The attacker can heartbleed to her or his heart’s content (pun intended). That is, if left undiscovered, an attacker can continuously pound the other side of the connection with heartbleeds, perhaps thousands of times. Which means multiple chunks of memory will be returned to the attacker, as the heap allocates, deallocates, and moves data around.

Lots of different heap chunks will get returned. There will likely also be overlap between the chunks that are returned to the attacker. Somewhere within those memory chunks are likely to be some sensitive data. If the private key for a session happens to be in one of those chunks, it will be exposed to the attacker. If any particular session open through the OpenSSL library happens to a contain a password that had been transmitted, it’s been exposed. It won’t take an engineering genius to do an ASCII dump of returned chunks of memory in order to go poking about to find interesting bits.

Still, and nonetheless, this is hunting for goodies in a bit of a haystack. Some people are quite good at that. Let’s acknowledge that outright. But that’s very different than a directed attack.

And should a wise and prepared security team, making good use of appropriate security tools, notice a heartbleed attack, they will most likely kill the connection before thousands of buffers can be read. Heartbleed over any particular connection is a linear process, one packet retrieved at a time. Retrieving lots of data takes some time. Time to respond. Of course, an unprotected and unaware site could allow many sessions to get opened by an attacker, each linearly heartbled, thus revealing far more of what’s on the heap than a single session might. Wouldn’t you notice such anomalous behaviour?

It’s important to note that the returns in the heartbleed packets are not necessarily tied to the attackers’ session. Again, it’s whatever happens to be on the heap, which will contain parts of other sessions. And any particular heartbleed packet is not necessarily connected to the data in a previous or subsequent packet. Which means that there’s no continuity of session nor any linearity between heartbleed retrievals. All session continuity must be pieced together by the attacker. That’s not rocket science. But it’s also work, perhaps significant work.

I’ll reiterate in closing, that this is a dangerous bug to which we must respond in an orderly fashion.

On the other hand, this bug does not give attackers free reign to go after all the juicy targets that may be available on any host, server, or endpoint that happens to have OpenSSL installed. Whatever happens to be on the heap of the process using the OpenSSL library and that is adjacent to the heartbeat buffer will be returned. And that attack may only occur during a TLS session. Simply including the vulnerable library poses no risk, at all. Many programs make use of OpenSSL for other functionality beyond TLS sessions.

This bug is not the unfettered keys to the kingdom, unless a “key to the kingdom” just happens to be on the heap and happens to get returned in the over-read. What gets returned is entirely due to the distribution of the heap at the moment of that particular heartbeat.

Cheers,

/brook

These assertions have been demonstrated in the lab through numerous runs of the heartbleed attack by a  team who cannot be named here. My thanks to them for confirming this assessment. Sorry for not disclosing.

[1] There are plenty of specialized cases that break this rule. But typically, code doesn’t run from the heap; data goes onto the heap. And generally speaking, programs refrain from executing on the heap because it’s a poor security practice. Let’s make that assumption about OpenSSL (and there’s nothing to indicate that this is NOT true in this case), in order to make clear what’s going on with heartbleed.

[2] The libraries that support programs developed with the major development tools and running on the major operating systems have sophisticated heap management services that are consumed by the running application as it allocates and deallocates memory. While care must be exercised in languages like C/C++, the location of where data end up on the heap is controlled by these low-level services.

[3] That is, intended for the application that is using OpenSSL for TLS services.

The “Real World” of Developer-centric Security

Posted on by

My friend and colleague, Dr. James Ransome, invited me late last Winter to write a chapter for his 10th book on computer security, Core Software Security, (with co-author, Anmol Misra)  published by CRC Press. My chapter is “The SDL In The Real World”, SDL = “Secure Development Lifecycle”. The book was released December 9, 2013. You can get copies from the usual sources (no adverts here, as always).

It was an exciting process. James and I spent hours white boarding possible SDL approaches, which was very fun, indeed*. We collectively challenged ourselves to uncover current SDL assumptions, poke at the validity of these, and find better approaches, if possible.

Many of you already know that I’ve been working towards a different approach to the very difficult, multi-dimensional and multi-variate problem of designing and implementing secure software for a rather long time. Some of my earlier work has been presented to the industry on a regular basis.

Specifically, during the period of 2007-9, I talked about a new (then) approach to security verification that would be easy for developers to integrate into their workflow and which wouldn’t require a deep understanding of security vulnerabilities nor of security testing. At the time, this approach was a radical departure.

The proving ground for these ideas was my program at Cisco, Baseline Application Vulnerability Assessment, or BAVA, for short (“my” here does not exclude the many people who contributed greatly to BAVA’s structure and success. But it was more or less my idea and I was the technical leader for the program).

But, is ease and simplicity all that’s necessary? By now, many vendors have jumped on the bandwagon; BAVA’s tenets are hardly even newsworthy at this point**. Still, the dream has not been realized, as far as I can see. Vulnerability scanning still suffers from a slew of impediments from a developer’s view:

  • Results count vulnerabilities not software errors
  • Results are noisy, often many variations of a single error are reported uniquely
  • Tools are hard to set up
  • Tools require considerable tool  knowledge and experience, too much for developers’ highly over-subscribed days
  • Qualification of results requires more in-depth security knowledge than even senior developers generally have (much less an average developer)

And that’s just the tool side of the problem. What about architecture and design? What about building security in during iterative, fast paced, and fast changing agile development practices? How about continuous integration?

As I was writing my chapter, something crystalized. I named it, “developer-centric security”, which then managed to get wrapped into the press release and marketing materials of the book. Think about this:  how does the security picture change if we re-shape what we do by taking the developer’s perspective rather than a security person’s?

Developer-centric software security then reduced to single, pointed question:

What am I doing to enable developers to innovate securely while they are designing and writing software?

Software development remains a creative and innovative activity. But so often, we on the security side try to put the brakes on innovation in favour of security. Policies, standards, etc., all try to set out the rules by which software should be produced. From an innovator’s view at least some of the time, developers are iterating through solutions to a new problem while searching for the best way to solve it. How might security folk enable that process? That’s the question I started to ask myself.

Enabling creativity, thinking like a developer, while integrating into her or his workflow is the essence of developer-centric security. Trust and verify. (I think we have to get rid of that old “but”)

Like all published works, the book represents a point-in-time. My thinking has accelerated since the chapter was completed. Write me if you’re intrigued, if you’d like more about developer-centric security.

Have a great day wherever you happen to be on this spinning orb we call home, Earth.

cheers,

/brook

*Several of the intermediate diagrams boggle in complexity and their busy quality. Like much software development, we had to work iteratively. Intermediate ideas grew and shifted as we worked. a creative process?

**At the time, after hearing BAVA’s requirements, one vendor told me, “I’ll call you back next year.”. Six months later on a vendor webcast, that same vendor was extolling the very tenets that I’d given them earlier. Sea change?

Of Flaws, Requirements, And Templates

Posted on by

November 13th, 2013, I spoke at the annual BSIMM community get together in Reston, Virginia, USA. I had been asked to co-present on Architecture Risk Assessment (ARA) and threat modeling along side Jim Delgrosso, long time Cigital threat modeler. We gave two talks:

  1. Introduction to  Architecture Risk Assessment And Threat Modeling
  2. Scaling Architecture Risk Assessment And Threat Modeling To An Enterprise

Thanks much, BSIMM folk, for letting me share the tidbits that I’ve managed to pick up along my way. I hope that we delivered on some of  the promises of our talks? One of my personal values from my participation in the conference was interacting with other experienced practitioners.

Make no mistake! ARA-threat modeling is and will remain an art. There is science involved, of course. But people who can do this well learn through experience and the inevitable mistakes and misses. It is a truism that it takes a fair amount of background knowledge (not easily gained):

  • Threat agents
  • Attack methods and goals used by each threat agent type
  • Local systems and infrastructure into which a system under analysis will go
  • Some form of fairly sophisticated risk rating methodology.

These specific knowledge sets sit on top of significant design ability and experience. The assessor has to be able to understand a system architecture and then to decompose it to an appropriate level.

The knowledge domains listed above are pre-requisite to an actual assessment. That is, there are usually years of experience in system and/or software design, in computer architectures, in attack patterns, threat agents, security controls, etc., that the assessor  brings to an ARA. One way of thinking about it is that ARA/threat modeling is applied computer security, computer security applied to the system under analysis.

Because ARA is a learned skill with many local variations, I find it fascinating to match my practice to practices that have been developed  independent of mine. What is the same? Where do we have consensus? And, importantly, what’s different? What have  I missed? How can I improve my practice based upon others’? These co-presentations and conversations are priceless. Interestingly, Jim and I agreed about the basic process we employ:

  • Understand the system architecture, especially the logical/functional representation
  • Uncover intended risk posture of the system and the risk tolerance of the organization
  • Understand the system’s communication flows, to the protocol interaction level
  • Get the data sensitivity of each flow and for each storage. Highest sensitivity rules any resulting security needs
  • Enumerate attack surfaces
  • Apply relevant active threat agents and their methodologies to attack surfaces
  • Filter out protected, insignificant, or unlikely attack vectors and scenarios
  • Output the security that the system or the proposed architecture and design are missing in order to fulfill the intended security posture

There doesn’t seem to be much disagreement on what we do. That’s good. It means that this practice is coalescing.The places were we disagree or approach the problem differently I think are pretty interesting.

Gary McGraw calls security architecture misses, “flaws”. Flaws as opposed to software bugs. bugs can be described as errors in implementation, usually, coding. Flaws would then be those security items which didn’t get considered during architecture and design, or which were not designed correctly  (like poorly implemented authentication, authorization, or encryption). I would agree that implementing some sort of no entropy scramble and then believing that you’ve built “encryption” is, indeed both a design flaw and an implementation error*.

I respect Gary’s opinion greatly. So I carefully considered his argument. My personal “hit”, not really an opinion so much as a possible rationale, is that “flaw” gets more attention than say, “requirement”? This may especially be true at the senior management level? Gary, feel free to comment…

Why do I prefer the term “requirement”? Because I’m typically attempting to fit security into an existing architecture practice. Architecture takes “requirements” and turns these into architecture “views” that will fulfill the requirements. So naturally, if I want security to get implemented, I will have to deliver requirements into that process.

Further, if I name security items that the other architects may have missed, as “flaws”, I’m not likely to make any friends amongst my peers, the other architects working on a system. They may take umbrage in my describing their work as flawed? They bring me into analysis in order to get a security view on the system, to uncover through my expertise security requirements that they don’t have the expertise to discover.

In other words, I have very good reasons, just as Gary does, for using the language of my peers so that my security work fits as seamlessly as possible into an existing architecture practice.

The same goes for architecture diagram preferences. Jim Delgrosso likes to proffer a diagram template for use. That’s a great idea! I could do more with this, absolutely.

But once again, I’m faced with an integration problem. If my peers prefer Data Flow Diagrams (DFD), then I’d better be able to analyze a system based upon its DFD. If architects use UML, I’d better be able to understand UML. Ultimately, if I prize integration, unless there’s no existing architecture approach with which to work, my best integration strategy is to make use of whatever is typical and normal, whatever is expected. Otherwise, if I demand that my peers use my diagram, they may see me as obstructive, not collaborative. I have (so far) focused on integration with existing practices and teams.

As I spend more time teaching (and writing a book about ARA), I’m finding that having accepted whatever I’ve been given in an effort to integrate well, I haven’t created a definitive system  ARA diagram template from which to work (though I have lots of samples). That may be a miss on my part? (architectural miss?)

Some of the different practices I encountered may be due to differing organizational roles? Gary and Jim are hired as consultants. Because they are hired experts, perhaps they can prescribe more? Or, indeed, customers expect a prescriptive, “do it this way” approach? Since I’ve only consulted sparingly and have mostly been hired to seed and mentor security architecture practices from the inside, perhaps I don’t have enough experience to understand consultative demands and expectations? I do know that I’ve had far more success through integration than through prescription. Maybe that’s just my style?

You, my readers, can, of course, choose whatever works for you, depending upon role, maturity of your organization, and such.

Thanks for the ideas, Jim (and Gary). It was truly a great experience to kick practices around with you two.

cheers,

/brook

*We should be long past the point where anyone believes that a proprietary scramble protects much. (Except, of course, I occasionally do come across exactly this!).