January 3, 1999 Corruptio Optimi Pessima "corruption of the best is worst of all" Vulnerability: Automountd Operating System: SUN Solaris Versions affected: 2.5, 2.5.1, 2.6, 2.7 (X86 and SPARC architectures) [ Rant -------------------------------------------------------------- ---- You pay for this? Why is it vendors can't get the hint? Network users demand secure software, not poor attempts at patching longstanding holes in the current revisions. Sun was warned about this bug years ago, and over time all they could come up with is a single broken patch that reminds us of the first Microsoft fix for the port 139 OOB attack. Why does it take vendors like Sun over three MONTHS to patch this kind of hole? They claim that it takes that long because of the level of testing they must put the software through. Right. So not only do they supply a broken patch, they test it for three months and never notice it is still vulnerable, then release it to their paying customers. You pay for this? How many security vendors knew about this and said nothing? Why did CERT never report this? Why have hackers been passing these exploits around for two years? Why didn't the ueber-spies of the ICSA get their hands on this and report it? Why didn't Sun's own "security analysis team" of 70 people find this bug (or any others)? Could it be job security? We think so. ---- ------------------------------------------------------------------- ] [ Details ----------------------------------------------------------- ---- Back in 1995, a vulnerability was discovered and a new world of buggy software born. This particular bug, would be conceived when Sun Microsystems made the foolish decision to ship Solaris 2.5 to customers. Since then, it has been a downhill roller coaster for the boys and girls at SUN. Now, not only is 2.5 affected, but 2.5.1 stock, 2.5.1 patched, 2.6 and 2.7 are all affected as well. This bug is particularly important, because it shows the incompetence of Sun Microsystems. Not only do they sit on patching bugs for months at a time, they cant even patch bugs right! When exploit code for the local automountd problem was leaked a few YEARS ago, SUN developed a patch and labeled it #104654-05. However, intrestingly enough, even after the patch was released, this bug could still be exploited quite trivially. What SUN decided todo, was remove the 'popen' call and put in its place an 'execve' call. Absolutely brilliant. Did they forget that the variable holding the program and arguments to be execued could still be set by issuing a AUTOFS_MOUNT call to automountd? Or, is this a case of job security by one of Sun's engineers? We certainly feel it is the latter. Either way, not only is this bug local, but it is remote too! And, not only is it remote, but it is started by default as well! Yes, thats right boys and girls, you can now execute any program you like on any default installation of SUN remotely! When exploiting this vulnerability, be sure to send SUN your warmest compliments. Now, getting back to the bug at hand, a few things have to be established before this can be exploited. First of all, to remotely exploit this bug, it becomes two-pronged. Meaning, we are now going to introduce another remote bug in the SUN Solaris operating system that when combined with the previous, create a lethal combination. This new bug, located in rpc.statd which is also started by default (imagine that!), allows for remote packets to be bounced to the local Operating System. Looking back at the details of automountd, we remember that it cannot accept packets on UDP or TCP protocols. But, it can accept packets on the TLI protocol, which rpc.statd happens to forward to, allowing us to remotely execute commands by sending our packets to rpc.statd. So, with all this in mind, if we send two RPC packets to rpc.statd on Solaris 2.5, 2.5.1 stock, 2.5.1 patched, 2.6, or 2.7, with the SM_MON and SM_NOTIFY commands, rpc.statd will happily forward the packet on to the local operating system's automountd daemon. Variables we will have to setup will be the following: struct mon monr; struct status stat; memset(&monr, 0, sizeof(struct mon)); monr.mon_id.my_id.my_name = cache; monr.mon_id.my_id.my_prog = (unsigned long)AUTOFS_PROG; monr.mon_id.my_id.my_proc = (unsigned long)AUTOFS_MOUNT; mon_id.my_id.my_name will be our cache, which we will be using to tell rpc.statd where to forward the packets to. The cache of the remote system can be found by establishing a connection to port 23, or port 21 of the system. The cache will usually appear inside parentheseis, and will be the hostname of the machine. Another example of obtaining the cache, would be to lookup the ip address of the machine and use the hostname that comes back. For example, for pathetic.sun.com, you could try 'pathetic' or 'pathetic.sun.com', and it will be a pretty safe bet that you will pick the correct hostname. mon_id.my_id.my_prog will be the variable AUTOFS_PROG telling rpc.statd which RPC service to connect to. mon_id.my_id.my_proc will be the variable AUTOFS_MOUNT telling rpc.statd to tell automountd that we want a AUTOFS_MOUNT request. Further variables to setup: monr.mon_id.my_id.my_vers = 1; /* For Solaris 2.5, 2.5.1 */ or monr.mon_id.my_id.my_vers = 2; /* For Solaris 2.6, and 2.7 */ monr.mon_id.mon_name = command; stat.mon_name = command; monr.mon_id.mon_name and stat.mon_name will both contain the program to execute on the remote system. Now, with those variables setup, we can send off our RPC packets to the remote system. However, to complicate things more, we must modify our tekneeq of exploiting this bug, for it to succeed on 2.6 and 2.7. Because SUN implements a new function called SMHASH in rpc.statd on Solaris 2.6 and 2.7, we will now have to involve DNS spoofing to acomplish executing our command. Because of the way SMHASH works and the way RPC arguments are encoded, our command is what SMHASH attempts to lookup in its address tables once rpc.statd receives our packet. If SMHASH cannot detect if our command is a valid address, it will not forward the packet. However, if we DNS spoof our command off the Primary and Secondary nameservers of the remote system, SMHASH will return true for finding the command as an address and our packet will be forwarded. If you wish to test this bug without DNS spoofing, you can also modify the /etc/hosts file on the Solaris system and put in the command you want to execute (without arguments) as a hostname with an ip address, and it will accomplish the same affect. For the actual exploit code to do this, I will let the reader go on to the next section to find it. Examples: If pathetic.sun.com were a Solaris 2.7 machine with pathetic as its hostname, and a vulnerable Primary name server, an exploit attempt would look like this: Execute commands to spoof reboot off Primary NS here ./amountdexp pathetic.sun.com pathetic reboot 1 If pathetic.sun.com were a Solaris 2.5.1 machine with pathetic as its hostname, an exploit attempt would look like this: ./amountdexp pathetic.sun.com pathetic reboot 0 ---- ------------------------------------------------------------------- ] [ Exploit ---- ------------------------------------------------------ ---- Exploit code for this vulnerability can be easily obtained from the following URL: http://www.attrition.org/hosted/cop/index.html ---- ------------------------------------------------------------------- ] ----------------------------------------------------------------------------- ----------------------------------------------------------------------------- Date: Mon, 4 Jan 1999 23:23:48 -0600 From: Scott To: BUGTRAQ@netspace.org Subject: Re: SUN almost has a clue! (automountd) This is in response to Mr. Corruptio's post of: Corruptio Optimi Pessima "corruption of the best is worst of all" the title of which seems to imply that Sun is "on the take" to wreck it's own OS. Or I suppose it could just be something cute sounding that he used his PC mouse to drag out of someone's signature :) I realize that the primary focus of this list is serious security (or the lack thereof), but I'm getting a little winded by the unnecessary commentary that accompanies some of these reports. I'm not fixing to launch into an attack of my own, instead I'd like to try a somewhat more constructive approach and attempt to actually answer some of the questions raised below. First: [] "Why is it vendors can't get the hint? Network [] users demand secure software, not poor attempts at patching [] long-standing holes in the current revisions." Not exactly true. The primary Solaris market is large corporations, database operations, ISP's, etc. These are very sophisticated, and multifaceted users with a whole range of concerns that compete with (or oppose) security concerns for priority. Ease of use and interoperability are two huge concerns at present - and both are generally regarded as somewhat of a contradiction to strong security. All of these issues compete with each other constantly. Certainly just after a break-in that caused some terrible problem - yes, at that moment in time security is a top concern. Next week, when the newest version of some compiler fails at some task...well that issue takes center stage. The REAL question that needs to be asked is - is a user willing to pay MORE for the OS to be more secure? GENERALLY speaking...the overwhelming answer is no. Don't scream and jump up and down just yet...keep reading. []Why does it take vendors like Sun over []three MONTHS to patch this kind of hole? [] []They claim that it takes that long because of the level of testing []they must put the software through. [] I've been dealing with Sun for 12 years. I've never heard anyone make this claim. And in truth, as we all know, it doesn't take 3 months to test a patch. I have actually asked Sun this very question. The answer I got was much more believable. It took 3 months because it was queued with a lot of other "to-do's" to compete with each of them for a valuable programmers time. We are in a security oriented list so security takes precedence here. Go to the DNS groups, or sendmail, or nfs, or whatever. Whatever the topic, rest assured, it will be the priority within that group. All of them filter into Sun (or any other vendor) who must weigh each on it's own merits and prioritize them. Sure Sun could hire more programmers and reduce the turnaround time...but that cost would be passed directly to the customer...remember, the customer that is unwilling to bare any additional cost. Say for example that there are 100 current problems that need to be patched, and further say the cost of Solaris is $100, and it would raise the cost per unit by 2% to commit new programming resources to fix one patch immediately. If you arrange to fix all 100 problems immediately (which you would have to do to please everyone) then the 'improved' Solaris would cost would now cost THREE TIMES AS MUCH. Do you think folks would be please to trade 90 days of response to an obscure exploit if it was going to triple the cost of the product. Absolutely not. Why doesn't everyone pay for 24/7 1 hour on-site repair response? Same reason. Sure they would LIKE to have it, but they don't want it badly enough to justify the additional cost. At this point you should start to be humming the tune I'm playing - Cost/Benefit Analysis. There are only two motivations for Sun to disregard the problem queue and commit immediate resources to a problem (1) it is a widespread problem that is CURRENTLY having an immediate impact on users (customers), (2) a large percentage of Solaris users indicate to Sun that they are so concerned about the POTENTIAL problems that they are willing to accept the marginal costs of committing additional resources to fix it immediately. That's about it. If you didn't notice, I just answered the very first question asked : "You pay for this?" Answer is No, you aren't paying for it. And most are content to wait for some period to AVOID marginal costs. In a sense cost/benefit is a form of security. Exercising it is a form of providing the security that you will still be solvent next year. Of course a shorter answer to the question "You pay for this?" is NO, because Solaris is now FREE to most types of users :) Sorry, I couldn't resist a little humor. Next question: [] How many security vendors knew about this and said nothing? [] Why did CERT never report this? Why have hackers [] been passing these exploits around for two years? [] Why didn't the ueber-spies of the ICSA get their [] hands on this and report it? I'm starting to suspect that Mr. Corruptio is not taking his Lithium (or is self-prescribing something else). Do I smell a a suggestion of a coordinated CONSPIRACY?? No! Real answer: None! Security vendors make there living by being the first to find problems. Rest assured that if someone knew of it...you would have heard it here first. Why don't they know about it? I would suggest because it is not nearly the catastrophic malady that Mr. Corruptio makes it out to be. If memory serves from a quick `grep -i if corruptio.letter| wc -l` I think there were around 7 or 8 conditions that had to be met to even render a machine vulnerable. rpc.statd has to be running (the first thing most security aware folks comment out). Automountd has to be available. OK so afar I'm concerned...but wait, here comes the hook. OOPS, the 2.6 2.7 rpc.statd has changed so now you not only have to spoof your IP, but 'bounce' it off the target's pri/sec nameserver - the details of which are "left as an exercise to the reader" - something that Caspar Dik addressed recently. Of course if you want to short cut this "remote" exploit, you can log into the machine, su to root and modify the /etc/hosts file with the actual code - not a very useful exploit. I just picked 20 Sun's at random, spread over several actual production locations. I received permission from the local owners to do this. I didn't look to see if rpc.statd was up or not, I just picked them at random. This attack failed 100% of the time. Now, I'm not saying the details are in error. I am saying that if it doesn't affect any of 20 random machines it isn't too much of a problem in the real world. Certainly not a problem that - cost/benefit again - customers would be willing to pay a premium for to get fixed overnight. I'd also note here that Sun has comments in the inetd.conf file warning users that security aware sites often disable the 'r' stuff, etc. []Or, is this a case of job security by one of Sun's engineers? [] We certainly feel it is the latter. Well this one is simple to answer. What job security is there in making a bad patch? To assume that such practices promote your longevity at a job is to say that Sun likes and rewards poor programming as the preferred practice. I'm fairly certain that no one believes that. [] The cache of the remote system can be found by establishing a []connection to port 23, or port 21 of the system. The cache will []usually appear inside parentheses, and will be the hostname of the []machine. [] Anyone want to explain this one. I've logged into a lot of Solaris machines and I'm not at all clear what this parenthesis's refers to. Let's try a stock out to the shrink-wrap machine: [/aaa]198#: telnet bloomer.xxx.org 23 Trying 10.10.10.9... Connected to bloomer.xxx.org. Escape character is '^]'. Hmm...no parentheses ...ok keep going login: johnqp Password: Last login: Sun Jan 3 20:10:56 from bufu.xxx.org Sun Microsystems Inc. SunOS 5.7 Generic October 1998 [bloomer:~]2%: Enough...I know what he's talking about...but I'll be darned if I'm sure whether or not he does. [] Since then, it has been a downhill roller coaster for the boys and []girls at SUN. Now, not only is 2.5 affected, but 2.5.1 stock, 2.5.1 []patched, 2.6 and 2.7 are all affected as well. My only comment is that Mr. Corruptio has quite a flair for the dramatic. Downhill? How so? Has anyone actually exploited this? No security companies or CERT knew of it (by Mr.Corruptio's own statements). If downhill is defined as gaining market share, being the dominant UNIX OS, having your stock triple in price, your CEO favorably protrayed on 60 Minutes as the only company with the guts to stand up against a REALLY bad OS like Windows, shipping more hardware that ever, bringing UNIX workstation pricing smack into the middle of the PC price range, inventing Java, and Jini - if all that is downhill then I'm getting my skis on. [] Why didn't Sun's own "security analysis team" of 70 people find []this bug (or any others)? Could it be job security? We think so. Again I'm unclear how a security team that fails to report security problems can be perceived to be a positive career move. And "Security Team" ?? I have this mental picture of a bunch of "computer firemen" sitting in a room...a bell rings and they all quit watching Oprah and jump into their freebie vendor T-shirts, slide down a poll, and run into a specially equipped secure clean-room to attack some bizarre new threat to man-kind. I would suggest that Sun's Security Team is more likely a bunch of folks with specific specialties (and many other job functions) that are consulted once some problem is reported. They are unlikely to be sitting around constantly trying to hack into their own OS. If I was a Sun stockholder and I found out Sun was paying 70 senior level programmers to sit around and try breaking into it's own OS, I'd be very upset. I would consider that to be very irresponsible corporate behavior. Mr. Corruptio - wake up - this isn't some academic exercise. This is a corporation trying to survive and flourish in the most inhospitable environment there ever was - a highly volitle computing industry that is donimated by a true piece of crap company. There is certainly more to Mr. Corruptio's letter to poke holes in, but I've already gone on too much. So let me leave you with this thought. Suppose I pull out the one suit I own (assuming it still fits), and show up at your business with only two products to sell. One is the most used and respected UNIX OS in the world. I tell you it runs a VERY high percentage of commercial and university servers in the world. It has an excellent suite of tools, and more software ported to it than any other UNIX. And, I tell you, it is secure against 99.9% of all types of exploits. And the price of this fine product is -- FREE! The other product I sell, oh well, it is the same product but it has undergone a no-holds-barred, no-expense-spared process to render it 100% safe against any and all types of exploits. The price? Oh well...it's $28 million per instance. This scenario is just about accurate. That last .1% would eat up more resources than it took to accomplish the entire first 99.9%. THIS is why cost/benefit was born. To STOP someone from doing something so foolish. Because every dollar you spend attacking that last 0.1% is a dollar you CANNOT spend on something else - because it is gone. Dollars are finite resources that must be spent where they can be the most effective - something that is not generally dictated by some script-kiddie pointing to an obscure (and highly conditionally dependant) exploit that by his own admission has not exactly been a hot topic among the security conscience. Should Sun address this problem - certainly. And anyone who is concerned about it should call Sun and file a bug / security report. Should Sun drop everything and commit all it's programming resources to this problem? I think the prudent individual would agree to the fallacy of that approach. I'll close by suggesting to Mr. Corruptio Optimi Pessima that when bashing companies, or anyone else for that matter, it is generally regarded as good manners to at least have the fortitude to stand behind what you say by signing your real name ( and thus your reputation ) to the words. Like this: Scott Stubbs bernadette.net Thanks for your time - and repeat after me just one more time - COST/BENEFIT ! ----------------------------------------------------------------------------- ----------------------------------------------------------------------------- Date: Tue, 5 Jan 1999 10:04:06 +0100 From: Casper Dik To: BUGTRAQ@netspace.org Subject: Re: SUN almost has a clue! (automountd) > If pathetic.sun.com were a Solaris 2.7 machine with pathetic > as its hostname, and a vulnerable Primary name server, > an exploit attempt would look like this: > > Execute commands to spoof reboot off Primary NS here > ./amountdexp pathetic.sun.com pathetic reboot 1 > > If pathetic.sun.com were a Solaris 2.5.1 machine with pathetic > as its hostname, an exploit attempt would look like this: > > ./amountdexp pathetic.sun.com pathetic reboot 0 Since tehre's no such thing as Solaris 2.7, I'm surprised it works tehre. Did you perhaps try it on the beta? My Solaris 7 system complains: Jan 5 09:47:31 room101 automountd[222]: Illegal access attempt by uid=1 - reque st ignored Jan 5 09:47:46 room101 statd[217]: statd: cannot talk to lockd at room101, RPC: Timed out(5) Statd doesn't run as root in Solaris 7 so the automounter will ignore its requests. This change was made late in Solaris 7 development and did not make it into any external release. The easiest way to work around this problem quickly is runnign statd as a user other than root, to this end change in /etc/init.d/nfs.client as follows (but not on Solaris 7, where such a change may break statd) 28c28 < /usr/lib/nfs/statd > /dev/console 2>&1 --- > su daemon -c /usr/lib/nfs/statd > /dev/console 2>&1 (make sure you keep the links in /etc/rc?.d/[SK]*nfs.client pointing to /etc/init.d/nfs.client) and run: chown -R daemon /var/statmon chmod -R og-w /var/statmon Then stop and start lockd & statd. Casper