Forgot your password?
typodupeerror
Encryption Security

Theo De Raadt's Small Rant On OpenSSL 301

Posted by timothy
from the heartbleed-of-the-matter dept.
New submitter raides (881987) writes "Theo De Raadt has been on a better roll as of late. Since his rant about FreeBSD playing catch up, he has something to say about OpenSSL. It is worth the 5 second read because it is how a few thousand of us feel about the whole thing and the stupidity that caused this panic." Update: 04/10 15:20 GMT by U L : Reader badger.foo pointed out Ted Unangst (the Ted in the mailing list post) wrote two posts on the issue: "heartbleed vs malloc.conf and "analysis of openssl freelist reuse" for those seeking more detail.
This discussion has been archived. No new comments can be posted.

Theo De Raadt's Small Rant On OpenSSL

Comments Filter:
  • by alphatel (1450715) * on Thursday April 10, 2014 @10:46AM (#46714171)
    This could get a lot more ugly...

    Once upon a time, SSL certificates were signed against a single root certificate, each SSL cert issuer had a single root certificate authority for each of its product lines. Now all corps issue an SSL certificate that is signed against and INTERMEDIATE certificate, which in turn is signed against the root certificate.

    What happens if a provider's server has this exploit and the intermediate certificate is compromised? EVERY certificate signed against that intermediate must be revoked. Or put another way, the ENTIRE PRODUCT LINE must be tossed into the garbage and all certs reissued.

    So if Verisign or Thawte discover new their intermediate certificate MIGHT have been exploited, would they say anything? The servers implementing those certs are in the hands of a select few - it would be easy to hide the possibility they might have been compromised.
    • by putaro (235078) on Thursday April 10, 2014 @10:58AM (#46714335) Journal

      If your intermediate certificate's signing keys are on the internet facing web servers you're doing it wrong. That intermediate signing key should be treated with the same level of security you would treat a root key with.

    • by Anonymous Coward on Thursday April 10, 2014 @11:00AM (#46714363)

      It is good practice to sign against an intermediate certificate. That way if it is compromised you can reject it and issue a new intermediate certificate signed by your root certificate. You can push the new certificates as updates since they would be validated against the root certificate.

      You need to read up on authenticating the entire chain of certificates.

      • by DarkOx (621550) on Thursday April 10, 2014 @01:46PM (#46716147) Journal

        Far more important is that if the intermediate certificate is compromised, you as the CA have ability to act. You know from your records who your customers are. What you need to do is:

        1) Fix the glitch
        2) Get the media that stores the trusted root certs private key out of the vault
        3) Issue new intermediate certificates
        4) return the root certs private key to vault
        5) Start contacting your certificate customers and issuing them new certs, revoking the old ones along the way as customers reported they have switched, or if there *is* indication of a compromised cert, revoke immediately.
        6). Revoke the old intermediate certificates as soon as 5 is complete.

        If were signing client certificates directly with the trusted root like they once did you (as the CA) would be screwed royally. You would need to somehow get every client device to update their trusted roots. Or you'd have upset customers crying about how their reissued certs are untrusted by 3/4ths of the clients out there that nobody bothers to update and nobody who understands these things manages directly..

    • I don't understand what you are trying to imply by contrasting the old way without intermediate vs the new way with intermediate certs. Surely, its better with the intermediates as the root cert that signs it should be and can be used less often and stored in cold storage somewhere.

    • by Anonymous Coward on Thursday April 10, 2014 @11:27AM (#46714691)

      I don't get what you're saying, and I think that's probably because you don't know what you're talking about. Having certificate chains is only a plus, the flat structure was crap. Here's how it works:

      I have a root certificate that's universally trusted. It is used *only* to sign intermediate certificates. Having the public cert in the public is fine since it only contains the public key part of the asymmetric public/private key pair. The private key, sits on a server which is physically isolated from the world. By that, I mean that the root certificate and private keys are literally on servers with *no* network connections. When you want to generate a new intermediate certificate, you put the CSR on a USB stick, plug it in, and sign it from that machine. In this way you never have to worry about external threats potentially gaining access to your private key (internal are an ever constant threat, and you put in good safeguards to prevent against that).

      Now that you have a chained hierarchy, you can use different intermediates to sign different end user certificates. Remember that both the root and intermediate have their own certificate revocation lists: the root can revoke intermediates (which means anything signed by them is null and void) and intermediates can revoke server or subordinate intermediate certs.

      As a result of my chained hierarchy, if an intermediate is compromised, I can revoke it, without revoking every single end user / server certificate out there. This gives me finer grained control.

      Now, I said that the root isn't even connected to the internet. The intermediate ideally is not either. Ideally the user / server signing intermediate is behind a set of firewalls, and *pulls* signing requests from frontends, that it then processes and posts the resulting signed certs to. That way if you compromise the front end, you have to the compromise the firewall (which ideally is blocking all inbound connection requests / unconnected sockets / anything that is not communication required for the intermediate server(s) to send pulls to the frontends) in order to get to the intermediate.

      Your flat view of the world is draconian, wrong, uneducated, and probably hurts everyone who reads it by making them a little less educated.

    • by VortexCortex (1117377) <VortexCortexNO@S ... t-retrograde.com> on Thursday April 10, 2014 @11:55AM (#46715047)

      Well, what you are pointing out is that a CA is a single point of failure -- Something actual security conscious engineers avoid like the plague. What you may not realize is that collectively the entire CA system is compromised by ANY ONE of those single points of failure because any CA can create a cert for ANY domain without the domain owner's permission. See also: The Diginotar Debacle. [wikipedia.org]

      The thing is, nobody actually checks the the cert chain, AND there's really no way to do so. How do I know if my email provider switched from Verisign to DigiCert? I don't, and there's no way to find out that's not susceptible to the same MITM attack.

      So, let's take a step back for a second. Symmetric stream ciphers need a key. If you have a password as the key then you need to transmit that key back and forth without anyone knowing what it is. You have to transmit the secret, and that's where Public Key Crypto comes in, however it doesn't authenticate the identity of the endpoints, that's what the CA system is supposed to do. Don't you see? All this CA PKI system is just moving the problem of sharing a secret from being the password, to being which cert the endpoint is using -- That becomes the essential "secret" you need to know, and it's far less entropy than a passphrase!

      At this time I would like to point out that if we ONLY used public key crypto between an client and server to establish a shared secret upon account creation, then we could use a minor tweak to the existing HTTP Auth Hashed Message Authentication Code (HMAC) proof of knowledge protocol (whereby one endpoint provides a nonce, then the nonce is HMAC'd with the passphrase and the unique-per-session resultant hash provides proof that the endpoints know the same secret without revealing it) to secure all the connections quite simply: Server and client exchange Nonces & available protocols for negotiation, the nonces are concatenated and HMAC'd with the shared secret stored at both ends, then fed to your key-stretching / key expansion system AND THAT KEYS THE SYMMETRIC STREAM CIPHER SIMULTANEOUSLY AT BOTH ENDS so the connection proceeds immediately with the efficient symmetric encryption without any PKI CA system required.

      PKI doesn't really authenticate the endpoint, it just obfuscates the fact that it doesn't by going through the motions and pretending to do so. It's a security theater. SSL/TLS and PKI are essentially the Emperor's New Secure Clothes. At least with the shared secret model I mention above, there's just that one-time small window of PK crypto for secret exchange at worst (failing to intercept account creation means no MITM) and at best you would actually have the CHANCE to go exchange your secret key out of band -- Visit your bank in person and exchange the passphrase, etc. then NO MITM could intercept the data. HTTP Auth asks for the password in a native browser dialog BEFORE showing you any page to login (and it could remember the PW in a list, or even generate them via hashing the domain name with a master PW and some salt so you could have one password for the entire Internet). That's how ALL security should work, it ALL relies on a shared secret, so you want the MOST entropic keyspace not the least entropic selection (which CA did they use). If you're typing a password into a form field on a web page, it's ALREADY game over.

      Do this: Check the root certs in your browser. For Firefox > Preferences > Advanced > Certificates > View. See that CNNIC one? What about the Hong Kong Post? Those are Known bad actors that your country is probably at cyber war with, and THEY ARE TRUSTED ROOTS IN YOUR FUCKING BROWSER?! Not to mention all the other Russian ones or Turkish, etc. ones that are on the USA's official "enemy" list. Now, ANY of those can pretend to be whatever domain's CA they want, and if your traffic bounces through their neck of the woods they can MITM you and you'll be n

  • by darkain (749283) on Thursday April 10, 2014 @10:51AM (#46714241) Homepage

    When Slashdot comments become full front page stories? This was already posted a few times as comments in the last OpenSSL post.

    • by Anonymous Coward on Thursday April 10, 2014 @10:54AM (#46714283)

      Theo De Raadt is a noteworthy personality, same as Linus Torvalds or Bill Gates. Their comments on important matters are important.

      • by MightyMartian (840721) on Thursday April 10, 2014 @11:04AM (#46714411) Journal

        As much as Theo can be an utter and insufferable prick, on this score he's right. This was an insanely trivial error which has exposed who knows how many systems to potential breaches. Right now I'm starting up a full audit of our systems. We use OpenVPN for our interoffice WAN, as well as for clients; many of them Windows, iOS and Android clients, not to mention reviewing all our *nix clients running SSH daemons. We're only a relatively small operation, and it's still a monumental pain in the ass.

        • Re:When comments... (Score:4, Interesting)

          by Anonymous Coward on Thursday April 10, 2014 @01:01PM (#46715667)

          As much as Theo can be an utter and insufferable prick, on this score he's right.

          Actually, most of the time he's right; it's just the prick bit that is the problem. What's worrying about this is that he sounds completely reasonable. He didn't even call the Open SSL developers any names or anything and everything in the post was reasonable. I hope he isn't sick or something.

          • by MightyMartian (840721) on Thursday April 10, 2014 @01:04PM (#46715703) Journal

            Don't worry. I've more than made up for Theo's unusual calmness this morning, and have called them names that would make the gods blanch.

          • by lgw (121541)

            He didn't even call the Open SSL developers any names or anything and everything in the post was reasonable. I hope he isn't sick or something.

            "OpenSSL is not developed by a responsible team."

            Theo the Raanter is fine. Personal abuse and profanity is just noise. This was a truly cutting comment.

      • Theo De Raadt makes comments about Linux being for losers. Linus Torvalds makes comments about OpenBSD users being masturbating monkeys. You usually have to take some of their comments with a grain of salt.

      • De Raadt wrote "OpenSSL is not developed by a responsible team".

        On the contrary, I believe it was developed by a responsible team, that unfortunately made an error.

        Most everyone have made errors, even if most go unnoticed and are essentially harmless. This one appears different, but I don't think it justifies De Raadt's moronic comment.
        • by tsalmark (1265778) on Thursday April 10, 2014 @11:22AM (#46714641) Homepage
          My understanding is Theo said: Developers on a security product made a conscious decision to turn off last line of defence security for all platforms in the interest of performance on some systems. That does not sound like and unfortunate error to me, it sounds outright irresponsible.
          • by squiggleslash (241428) on Thursday April 10, 2014 @11:30AM (#46714729) Homepage Journal

            He said that, but is that what happened? Were OpenSSL's developers aware that malloc()/free() have special security concerns that OpenBSD's developers had specifically addressed (I assume that's what meant by "a conscious decision to turn off last-line-of-defense-security")

            I understand Theo's point, to a certain degree I kinda understand it, but I'm more inclined to feel the problem is with OpenSSL's developers clearly not understanding the security concerns about malloc(). That is, if they were aware that OpenBSD's malloc() contained code to ensure against data leakage, it would seem to me to be highly probable they would have implemented the same deal in OpenSSL given, you know, their entire point is security. The fact they didn't makes me think they didn't know OpenBSD's malloc() had these measures in the first place.

            Should they have done? And how should they have known? Genuine question, and finger pointing would be inappropriate right now: how do we make sure that certain security strategies and issues are as well known as, say, stack pointer issues are today.

            • Re: (Score:3, Informative)

              by EvanED (569694)

              Were OpenSSL's developers aware that malloc()/free() have special security concerns that OpenBSD's developers had specifically addressed (I assume that's what meant by "a conscious decision to turn off last-line-of-defense-security")

              My impression is OpenBSD's hardened allocator is relatively common knowledge and definitely should be among people writing security software. And that's not even remotely the only such allocator out there that does that sort of thing too, though it's probably the most well-known

            • by Eunuchswear (210685) on Thursday April 10, 2014 @11:48AM (#46714975) Journal

              That is, if they were aware that OpenBSD's malloc() contained code to ensure against data leakage, it would seem to me to be highly probable they would have implemented the same deal in OpenSSL given, you know, their entire point is security. The fact they didn't makes me think they didn't know OpenBSD's malloc() had these measures in the first place.

              Not just OpenBSD's malloc(). glibc can do the same thing if you set MALLOC_PERTURB.

            • by mrchaotica (681592) * on Thursday April 10, 2014 @12:12PM (#46715237)

              Should they have done? And how should they have known? Genuine question, and finger pointing would be inappropriate right now: how do we make sure that certain security strategies and issues are as well known as, say, stack pointer issues are today.

              Hell yes they should have known, because the people responsible for one of the most important security applications in the entire world damn well ought to be experts!

            • I understand Theo's point, to a certain degree I kinda understand it, but I'm more inclined to feel the problem is with OpenSSL's developers clearly not understanding the security concerns about malloc(). That is, if they were aware that OpenBSD's malloc() contained code to ensure against data leakage, it would seem to me to be highly probable they would have implemented the same deal in OpenSSL given, you know, their entire point is security. The fact they didn't makes me think they didn't know OpenBSD's malloc() had these measures in the first place.

              Here is what happens, as far as I understand: a client sends two bytes of data to the server and asks the server to send the identical two bytes back to the client, to check that the server is still alive. That's how it works normally. A client could send 65,000 bytes and ask for the 65,000 bytes to be sent back, except that would be inefficient.

              Instead, an attacker sends two bytes of data to the server and asks for the same 65,000 bytes back. The server stores the two bytes with a bit of overhead into a

        • Sometimes the individuals involved can be responsible while the team acts irresponsibly. For example, why is my passphrase of my id_rsa key protected by only one round of hashing with no option for increased rounds [sys4.de]? I hear there are good things coming, like being able to use bcrypt, but this is a scandal. Only a security ignorant fool would want his passphrase attached to an id_rsa key with no password stretching at all. So... how many fools do we have out there? I surely hope you weren't counting on y

        • Re: (Score:3, Informative)

          by Anonymous Coward

          De Raadt wrote "OpenSSL is not developed by a responsible team".

          On the contrary, I believe it was developed by a responsible team, that unfortunately made an error.

          Most everyone have made errors, even if most go unnoticed and are essentially harmless. This one appears different, but I don't think it justifies De Raadt's moronic comment.

          Not so sure they're responsible.

          Did you read this [tedunangst.com]?

          This bug would have been utterly trivial to detect when introduced had the OpenSSL developers bothered testing with a normal malloc (not even a security focused malloc, just one that frees memory every now and again). Instead, it lay dormant for years until I went looking for a way to disable their Heartbleed accelerating custom allocator.

          Building exploit mitigations isn’t easy. It’s difficult because the attackers are relentlessly clever. And it’s aggravating because there’s so much shitty software that doesn’t run properly even when it’s not under attack, meaning that many mitigations cannot be fully enabled. But it’s absolutely infuriating when developers of security sensitive software are actively thwarting those efforts by using the world’s most exploitable allocation policy and then not even testing that one can disable it.

          The OpenSSL team doesn't fully test their product.

          That's pretty much as good an example of incompetence that you can probably find.

          • Re: (Score:3, Interesting)

            This bug would have been utterly trivial to detect when introduced had the OpenSSL developers bothered testing with a normal malloc (not even a security

            This is simply not true, stop spinning it.

            Even if OpenSSL is using system's malloc, with all its mitigation features, the bug still works. The attacker just has to be more careful, lest he should read freed() and unmapped memory, and so cause a crash and (supposedly) leave some kind of meaningful trail.

            • by chis101 (754167) on Thursday April 10, 2014 @04:20PM (#46718247)

              Even if OpenSSL is using system's malloc, with all its mitigation features, the bug still works. The attacker just has to be more careful, lest he should read freed() and unmapped memory, and so cause a crash and (supposedly) leave some kind of meaningful trail.

              You got it exactly right. He's complaining that because they provided their own malloc() wrapper that the read of freed() memory is NOT causing a crash. If they had used the system malloc() then there would be crashes, the issues would be detected, and they would have been fixed.

        • by OneAhead (1495535) on Thursday April 10, 2014 @02:02PM (#46716367)

          That would have been an appropriate boilerplate response to your typical Theo De Raadt comment, but... did you actually read that "analysis of openssl freelist reuse" link? It's hard for me to imagine that having been done by mistake; it's more of an instance of a coder cleverly cutting a corner. And some might argue "don't we all?", but if one is writing such security-critical software, one should at least have some notion that clever corner-cutting is a gigantic no-no.

          Now, the freelist reuse is not the cause of the heartbleed bug; it merely frustrates what would otherwise have been a relatively straightforward mitigation strategy. But it's a symptom of an attitude that is, well, irresponsible.

    • Because its Theo. HIs rants are usually pretty funny and thought provoking.

      This one, I have to admit, is pretty tame/lame. Good security design thoughts as always, but not much of a rant.

  • by Anonymous Coward on Thursday April 10, 2014 @10:54AM (#46714281)
    Years ago the BSD guys added safeguards to malloc and mmap, but they were disabled for all platforms in OpenSSL only because they caused performance problems on some platforms. He finishes by saying that OpenSSL is not developed by a responsible team.
  • by badger.foo (447981) <peter@bsdly.net> on Thursday April 10, 2014 @10:55AM (#46714297) Homepage
    OpenBSD developer Ted Unangst (mentioned in the article) has gone into the code a bit more in two articles, both very well worth reading:

    heartbleed vs malloc.conf [tedunangst.com]

    and

    analysis of openssl freelist reuse [tedunangst.com]. Short articles with a lot of good information.

    • by OmniGeek (72743) on Thursday April 10, 2014 @12:28PM (#46715395)

      As I read his analysis, OpenSSL relies on releasing a buffer, reallocating it, and getting the PREVIOUS contents of that buffer back -- or else it will abort the connection. (Search for the string "On line 1059, we find a call to ssl3_release_read_buffer after we have read the header, which will free the current buffer." in his article referenced by the parent post).

      Now, IMO, this goes way beyond sloppy. Releasing a buffer before you're done with it, and relying on a wacky LIFO reallocation scheme giving you back that very same buffer so you can process it, is either 1) an utterly incompetent coding blunder that just happened to work when combined with an utterly terrible, insecure custom allocation scheme, or 2) specifically designed to ensure that this insecure combination is widely deployed to provide a custom-made back door, as it works only with the leaky custom allocator.

      If 1), then I must agree with Theo that the OpenSSL team were indeed irresponsible, since at least one of these two cooperating blunders ought to have shown up in a decent security audit of the code, and any decent set of security-oriented coding standards would forbid them both.

      If 2), then it was deliberate, and the tinfoil-hat crowd is right for once.

  • Summary. (Score:5, Insightful)

    by jythie (914043) on Thursday April 10, 2014 @11:01AM (#46714367)
    So as far as I can tell, his rant is essentially that people should not use custom allocators and instead rely on the general purpose one built into libc because they can add system wide tools there.

    I can see the argument for most cases, that is kinda the point of a general purpose allocator, but encryption (esp if you are doing lots of it) really strikes me as a case where you can really benefit from having explicit control over the behavior. I have worked on a number of applications where custom allocators had significant (user facing, not just benchmarks) impacts on performance. Ironically it also meant we were able to do better checking then the general exploit detection kits since we could bake more specific knowledge into the validator.
    • Re:Summary. (Score:5, Insightful)

      by mr_mischief (456295) on Thursday April 10, 2014 @11:16AM (#46714575) Journal

      That's all true and correct. When you do that, though, you need to do at least as good a job as what you're circumventing. In this case OpenSSL didn't.

      • Re: (Score:3, Insightful)

        by jythie (914043)
        True, they did not, but I would put that at the level of mistake rather then being unreasonable.
        • Re:Summary. (Score:5, Insightful)

          by Anonymous Coward on Thursday April 10, 2014 @11:59AM (#46715099)
          but if your explicit reason for overriding the kernel is to do something as-good-or-better than the kernel, and you fail to achieve that (let alone validate that you've achieved it), that's a rather spectacular failure. mistakes are common, sure. but this is sort of like saying the seatbelts that came with your car suck (they very well might), so you replace them with something "better", that ends up being little more than a piece of string that you daintily lay across your lap. kind of "unreasonable", no...?
        • I think this Ted fellow linked from the summary makes a good case against it being "reasonable" at the end of his second post [tedunangst.com]:

          This bug would have been utterly trivial to detect when introduced had the OpenSSL developers bothered testing with a normal malloc (not even a security focused malloc, just one that frees memory every now and again). Instead, it lay dormant for years until I went looking for a way to disable their Heartbleed accelerating custom allocator.

          Building exploit mitigations isn't easy. It's

  • by sinij (911942) on Thursday April 10, 2014 @11:01AM (#46714383) Journal
    Why OpenSSL is so popular? It has FIPS-certified module, and this becomes important for selling your product to the government.

    So what could be done to prevent something like this from happening in the future? People will keep writing bad code, this is unavoidable, but what automated tests could be run to make sure to avoid the worst of it? Someone with direct development experience please educate the rest of us.
    • by swillden (191260) <shawn-ds@willden.org> on Thursday April 10, 2014 @11:21AM (#46714633) Homepage Journal

      People will keep writing bad code, this is unavoidable, but what automated tests could be run to make sure to avoid the worst of it?

      Automated testing for security problems doesn't really work. Oh, you can do fuzzing, but that's hit and miss, and general unit testing can catch a few things, but not much. Mostly, security code just requires very careful implementation and code review. Eyeballs -- smart, experienced eyeballs.

      OpenSSL has terrified me for years. The code is very hard to read and understand, which is exactly the opposite of what's desired for easy review and validation of its security properties. It needs to be cleaned up and made as simple, straightforward and accessible as possible, or replaced with something else that is simple, straightforward and accessible. My theory on why it is the way it is -- and it's far from unusual in the crypto library world -- is that cryptographers tend not to be good software engineers, and software engineers tend not to have the cryptography skills needed.

      I spend some (not very much, lately) of my time working on an open source crypto library called Keyczar that tries to address one part of this problem, by providing well-engineered crypto APIs that are easy to use and hard to misuse. That effort focuses on applying good engineering to the boundary between library and application code, which is another source of rampant problems, but Keyczar uses existing crypto libs to provide the actual implementations of the primitives (the C++ implementation uses openssl, actually). I've long wished we could find crypto libs that were well-engineered, both internally and in their APIs.

      • by Daniel_Staal (609844) <DStaal@usa.net> on Thursday April 10, 2014 @11:57AM (#46715079)

        In this case though, general unit testing should have caught the bug: There's an option at compile time which, if used, caused the affected versions of OpenSSL to crash. (Because it disables the bug, and OpenSSL was relying on it in one location...) So, good unit testing would have helped.

        Basically, unit testing should be able to tell you if you've implemented the algorithm competently. It doesn't say if the algorithm is any good, just that your version of it works to the spec.

      • OpenSSL has terrified me for years. The code is very hard to read and understand

        I would like to second this thought. For such an important piece of code, OpenSSL is overly complex. That, by itself, is a bug.

    • Why OpenSSL is so popular? It has FIPS-certified module, and this becomes important for selling your product to the government.

      FIPS certification is for a specific compiled version of OpenSSL and is not a blanket certification. The only way you are FIPS compliant is if you document that your product uses the exact same compiled version of OpenSSL or you submit your version of OpenSSL to be certified.

    • by jythie (914043)
      Well, it is popular because it is a generally well regarded and vetted package that supports a fairly rich set of cryptography tasks out of the box.

      As for what could be done in the future? Well, automated tests really only cover cases you think about, and stress tests may or may not actually notice something. To a degree, there will always be things that slip through, and most of the time things are fixed and patched. In this case something unusually bad slipped through.
      • by MarcoAtWork (28889) on Thursday April 10, 2014 @11:37AM (#46714835)

        it is a generally well regarded and vetted package that supports a fairly rich set of cryptography tasks out of the box.

        I would see that as a drawback for using it in webservers: if I am writing something internet-facing I want to use the smallest and simplest possible library that does the job, maybe it would be time to fork openssl into openssl-core / openssl-extras and have openssl-core have only the most minimal set of functionality related to securing connections and that's it? I would honestly also only support a few platforms for -core to simplify the code analysis even more (the more ifdefs, the more possible issues)

    • by frank_adrian314159 (469671) on Thursday April 10, 2014 @11:36AM (#46714821) Homepage

      First, make sure that code that must be secure is transparent. That means little (or no) optimizations, standard calls to OS functions, and clearly structured. It's clear that the OpenSSL developers made their code more opaque than was prudent and the many eyes of open source land could not see through the murk. Yes, clearer code would mean that it ran more slowly and some folks would need to run a few more servers, but the security problem might have been uncovered sooner (or not have happened) if someone hadn't thought that performance was a reason to make the code more complex.

      Second, formal independent review would have helped. Most code (especially in volunteer-based open source projects) is only vetted by people directly on the development team. Any piece of software as ubiquitous and critical to the operation of today's internet as OpenSSL cannot have verification and validation mainly by its own developers. For software like this, where security is critical, you should have external review. Start an independent project that vets these things, folks.

      Third, understand the limits of testing vs. design. More unit tests would not have caught this. Simple designs lead to simple and correct implementations. Complex designs (or no designs) lead to seas of unit tests that simply tells you the ways that the code happens not to be broken at the moment. Code like that in OpenSSL ideally should be simple enough to be formally proved correct.

      I think we've known about why these sorts of things happen ever since I entered he field thirty years ago. We have ways to prevent them, but they usually take time, money, or lowered performance. That they are still happening because of performance zealotry, bad process, and "teh web-speed is everything" mentality is a black mark on our profession.

  • Hindsight is 20/20 (Score:5, Insightful)

    by slyborg (524607) on Thursday April 10, 2014 @11:09AM (#46714489)

    So, it's always great fun bashing "obvious" bugs, especially when they have such an impact, but let it be noted that thousands of implementers used openssl to build systems taking the package at face value despite these now "obvious" deficiencies in development process. If you were that concerned about security, they would have done what Google did, and audit the code. There are of course many practical reasons why people can't do that, but regardless, the blame arrow here points both ways.

    • by sinij (911942)

      Google can afford to audit the code, and has reasonable expectation of meaningful results. The rest of us? Not so much.
       
      Are there any decent automated code auditing tools that could be used by smaller shops?

    • by DMUTPeregrine (612791) on Thursday April 10, 2014 @11:28AM (#46714705) Journal
      OpenSSL's code is a mess. Go, read it.

      Now that you're back from your stay in the sanitarium, would you like to consider that rewriting it might be a better choice than auditing? Yes?

      Let's just make sure Nyarlathotep isn't on the dev team this time...
    • by imikem (767509)

      Great power, great responsibility. Adding features to a piece of software as critical to so much infrastructure needs to be taken ever so seriously. Multiple levels of code audits for starters. Testing/fuzzing. I'm not qualified to do those things, but I still get to clean up the mess left by the whole fiasco.

    • by gweihir (88907)

      While generally true, this does not apply in the current disaster: The OpenSSL code, coding guidelines, coding style and project management style actually foster this kind of thing, like they wanted critical bugs. For example to test assumptions such as boundaries always at time-of-use is a fundamental secure coding principle. Apparently, they have never heard of it. To clear memory that may contain secrets when freed is another one. And so on.

  • Bug Looks Deliberate (Score:5, Interesting)

    by Anonymous Coward on Thursday April 10, 2014 @11:17AM (#46714587)

    That code is almost a text book example of material that is submitted to the Underhanded C contest...

    http://en.wikipedia.org/wiki/Underhanded_C_Contest

  • I'm not a developer and don't know the convention for version numbers but shouldn't they make the rev number of the repaired OpenSSL something more distinctive than adding "g" to the third digit of the new code? Maybe change it to version 2.0.0 or something more obvious.
    • Major numbers are often reserved for things that break backwards compatibility, minor numbers often denote new features, and the patch revision denotes patches.

      Personally I would have bumped the patch rev (the third number).

    • I think that might scare people from upgrading. Going from 7.10322 to to 8 is scarier than going from 7.10322 to 7.10322g It suggests that upgrading will break things.
  • Really? (Score:3, Interesting)

    by oneandoneis2 (777721) on Thursday April 10, 2014 @11:30AM (#46714725) Homepage

    "it is how a few thousand of us feel about the whole thing"

    Then maybe you thousands should stop complaining and start contributing to the project, which is so under-resourced problems like this are pretty much inevitable.

  • by Kazoo the Clown (644526) on Thursday April 10, 2014 @11:33AM (#46714765)
    That's pretty scathing. I'd hate to be THOSE guys...
  • by gweihir (88907) on Thursday April 10, 2014 @11:35AM (#46714799)

    In addition, the mitigation countermeasures also prevent memory debuggers like Valgrind from finding the problem (Valgrind find use-before-init for malloc'ed blocks, but not if there is a wrapper in between that re-uses blocks), and may also neutralize code-security scanners like Fortify.

    I have to admit that while my original intuition was "screwup", this looks more and more like some parts of the OpenSSL team have been compromised and did things that make this kind of bug far more likely. Like their own insecure memory allocation. Like not requiring time-of-use boundary checks or having any secure coding guidelines in the first place. Like documenting everything badly so potential reviewers get turned away. Like not having working review for patched or a working fuzz-testing set-up (which would have found bug this easily).

    After all, the NSA does not have to sabotage FOSS crypto software. They just have to make sure the quality is low enough. The bugs they can exploit will follow. And the current mess is just a plain classic. Do enough things wrong and eventually stuff breaks spectacularly.

  • A lot of large performance-sensitive projects implement custom allocators in the form of arenas and freelists. Lots of platforms have a fast malloc implementation these days, but none of them will be as fast as this for the simple reason that the program knows more about its memory usage patterns than any general-purpose allocator ever could.

    Not to say I can't understand Theo's point of view -- if he wants maximum security, then a program which bypasses one of his layers in the name of performance might not be the best for him.

    On the flip side, the standards have no notion of such security layers and I feel it is perfectly reasonable for a team to not throw away performance in the interests of some platform-specific behavior. This was a bug, pure and simple. There's nothing wrong with using custom allocators. To say that "OpenSSL is not developed by a responsible team" is simply nonsense.

    • by JDG1980 (2438906) on Thursday April 10, 2014 @11:51AM (#46715009)

      A lot of large performance-sensitive projects implement custom allocators in the form of arenas and freelists. Lots of platforms have a fast malloc implementation these days, but none of them will be as fast as this for the simple reason that the program knows more about its memory usage patterns than any general-purpose allocator ever could.

      This is security software. You don't sacrifice the library's core functionality to make it run a bit faster on the old Celeron 300 running Windows 98.

      • This is security software. You don't sacrifice the library's core functionality to make it run a bit faster on the old Celeron 300 running Windows 98.

        malloc's core functionality is to allocate memory. Any security additions are platform-specific and irrelevant.

    • by phantomfive (622387) on Thursday April 10, 2014 @12:26PM (#46715373) Journal

      To say that "OpenSSL is not developed by a responsible team" is simply nonsense.

      Unless you look at the code, and notice they are using unvalidated data for......anything. That's a rookie mistake.

  • De Raadt is wrong (Score:2, Interesting)

    by stephc (3611857)
    This is not a problem with OpenSSL, or the C Language or the Malloc implementation, this is a problem because everyone is relying on the same black box they do not understand. Because this is "standard" and common practice to use it. The only long term defense against this kind of vulnerability is software (and hardware?) diversity. Software built on custom SSL implementations may have even worse vulnerabilities, but nobody will discover them, and even if they do, it won't affect everyone on this planet
    • by JDG1980 (2438906)

      This is not a problem with OpenSSL, or the C Language or the Malloc implementation, this is a problem because everyone is relying on the same black box they do not understand.

      That's a cop-out. Any kind of advanced economy needs division of labor. This is no less true of the IT industry than anywhere else. The people building the "black box" need to know what they're doing and it needs to work. Period.

    • by tomhath (637240)
      I doubt "diversity" in this case is a good idea. Instead of one bug in one package you would end up playing bop-a-mole with many bugs in a few packages.
  • by Big Hairy Ian (1155547) on Thursday April 10, 2014 @11:43AM (#46714899)
    is not helping!! What would be useful is a list of popular services affected by this issue the BBC is at least making a start here http://www.bbc.co.uk/news/tech... [bbc.co.uk]
  • Feeding Allo(g)ators (Score:3, Interesting)

    by Anonymous Coward on Thursday April 10, 2014 @12:24PM (#46715351)

    Allocators in this case make no significant difference with regards to severity of the problem.

    What is or is not in process free list makes no difference when you can arbitrarily request any block of memory you please.. only slightly effects chance of success when it becomes necessary to shoot in the dark. Lets not forget most OS provided optimized allocators keep freed memory in their heaps for some time as well and may still not throw anything when referenced.

    Looking at code for this bug I am amazed any of this garbage was accepted in the first place. There is no effort at all to minimize chance of error with redundant + 3's and 1 + 2's sprinkled everywhere complete with unchecked allocation for good measure.

    buffer = OPENSSL_malloc(1 + 2 + payload + padding);
     
    r = dtls1_write_bytes(s, TLS1_RT_HEARTBEAT, buffer, 3 + payload + padding);

    Suppose I should be glad 1 + 2 = 3 today and they have not used signed integers when dealing with lengths.

    unsigned int payload;
    unsigned int padding = 16; /* Use minimum padding */

    ... oh dear god ...

    int dtls1_write_bytes(SSL *s, int type, const void *buf, int len)

    Well at least they learned their lesson and have stopped sprinkling redundant and error prone type + length + padding garbage everywhere... see..

    + buffer = OPENSSL_malloc(write_length);
     
    - buffer, 3 + payload + padding,
    + buffer, write_length,

    and here ..

    + if (1 + 2 + 16 > s->s3->rrec.length)+ return 0; /* silently discard */
     
    + if (1 + 2 + payload + 16 > s->s3->rrec.length)+ return 0; /* silently discard per RFC 6520 sec. 4 */
     
    + if (1 + 2 + 16 > s->s3->rrec.length)+ return 0; /* silently discard */
     
    + if (1 + 2 + payload + 16 > s->s3->rrec.length)+ return 0; /* silently discard per RFC 6520 sec. 4 */

    ... oh well .. Looks like plenty of low hanging fruit to be had for anyone with a little spare time.

  • the RFC is horrible (Score:5, Interesting)

    by DrProton (79239) on Thursday April 10, 2014 @12:52PM (#46715593)
    Let's not miss the opportunity to point a finger of blame at the RFC, which says " to make the extension as versatile as possible, an arbitrary payload and a random padding is preferred, ". https://tools.ietf.org/html/rf... [ietf.org] Arbitrary payload and a random padding for a heartbeat instead of a specified sequence of bits? This is very suspicious.
  • by ledow (319597) on Thursday April 10, 2014 @01:14PM (#46715779) Homepage

    Can't say I'm surprised. OpenSSL is a pile of dung. It's nothing to do with being written in any language, it's just horrible.

    There's not even any documentation. I mean, literally, none. Nothing vaguely useful. How do I programmatically load a certificate into the store, along with a chain of related trusted certificates, and then set my requirements (must be in-date, must be validly signed, etc.) and get out a "It's fine" / "Something's not right" response? The only answers I could ever find were to follow published examples and tweak.

    And when it comes to working out where in the published examples structure X comes from, or how to convert it to structure Y, you're on your own unless you happen to have picked a comprehensive (and almost certainly not OpenSSL-supplied) example.

    It's just that bad. I was writing a pseudo-DRM for a game / Steam-like distribution platform as a hobbyist project. It was literally horrible to even try to self-sign some certificate and then see if it all panned out later from another computer to guarantee integrity. In the end, I had to "imagine" every possible case and find a way to counter it (i.e. client cert expired, client cert invalid, server cert not signed client cert, server cert has bad chain of trust, client cert not signable for that purpose, etc.) - and almost always there was NOTHING to indicate what the recommended way to do it was.

    There is no decent OpenSSL documentation at all. Not even a decent overview of the process of checking certificates. It scared me at the time, knowing how important the library is, and it can only lead to bad code.

    In the end, I'm quite glad I don't have to program against it for a living. If I did, I'd be seriously looking for something else.

The only way to learn a new programming language is by writing programs in it. - Brian Kernighan

Working...