Archive for the Hacking Category
Posted at 7:00 pm by RobOHara in Hacking
If you think you don’t need to read this post, you definitely need to read this post.
Heartbleed is a security vulnerability that was discovered this week. It probably affects you. First, the five W’s:
Who: Anyone who uses the web and uses https links. That’s probably you.
What: Heartbleed is a vulnerability that allows people to see the information you send to some websites that use OpenSSL. It’s a lot of them.
Where: Gmail, Yahoo, Tumblr, Flickr, Facebook…
When: The problem has been around for two years now, but nobody noticed it until this week.
Why: Honest human error.
You’ve probably noticed the letters “HTTP” preceding most web links. HTTP stands for “hypertext transfer protocol,” and by putting that in front of a web link you’re telling your web browser “Hey, what comes next is going to be a web page.” It’s kind of like saying, “the following message will be in English.”
Sometimes, you’ll see HTTPS instead. The S stands for “secure sockets layer” (or SSL for short), but you can think of that S as simply meaning “secure”. When you use HTTP, the things you read and send across the internet are sent in plain text. That means anyone with the means to do so who is looking and listening for your message can read what you are sending and receiving. With HTTPS, what you send and receive to and from websites is secure and encrypted. Even if someone were to intercept your message, if you are using HTTPS, the information would look scrambled and no one would be able to read it. This is why websites like Gmail and Facebook and your bank’s website default to HTTPS — because it’s secure.
Or, so we thought. Turns out, back in early 2012, someone made a mistake while updating OpenSSL. A big one. Well known security expert Bruce Schneier said on his website this week, “on a scale of 1 to 10, this is an 11.” This bug, which again was introduced in 2012, allows/allowed hackers to read information in certain HTTPS transfers. One frustrating thing about this bug is that there’s no way for servers owners to know if people were hacking them or not; all they can tell is if they were vulnerable or not. And it turns out, a lot of websites were vulnerable.
The good news is Heartbleed only lets attackers view a small portion of memory at a time, so there’s a chance nobody ever saw your password. The bad news is, this vulnerability has been around for two years now, so there’s no telling if you were affected or not.
Several sites including this link at Mashable.com are compiling lists of websites that were affected and have been patched. You’ll want to change your password on those sites. Some of the ones on that list currently include: Facebook, Instagram, Pinterest, Tumblr, Flickr, Google/Gmail, Yahoo/Yahoo Mail/AIM, YouTube, Etsy, GoDaddy, Netflix, Soundcloud, TurboTax, USAA, Box, DropBox, Github, and IFTTT.
Oh, and Minecraft.
This is a good time to remind you that if you use the same password on any other site that you also use on those sites, you should change that password too. Also, stop doing that.
So what about your bank or some other SSL page you want to test? Several “Heartbleed Testers” have been stood up online. Here’s one. Simply click the link and cut/paste the URL to your bank (or any other HTTPS web link) and the website will let you know if they are currently using a safe version of OpenSSL. Of course it doesn’t tell you if they had the bad version last week…
I spent a couple of hours last night changing my passwords on a bevy of services including Facebook, Twitter, Gmail, and more. You should to. It’s a pain in the butt, especially when you have multiple devices (phones, tablets, laptops) that will all need the news passwords, but you’ll thank me in the morning.
Posted at 6:40 am by RobOHara in Hacking
A lot of things just happened when you clicked on this article. Your computer connected to my computer, and each of these words I wrote zipped across the internet to their destination. Since this article contains words like encryption, NSA, and secret codes, it probably flagged something for the NSA along the way — you for reading about it, and me for writing about it. In some giant, government data warehouse, there’s now a record that you were here. We’re probably both on a watch list now. Welcome to the machine, and all that.
About five years ago I wrote a silly little program called eCoder Ring. eCoder Ring is a small program that allows you to encrypt and decrypt secret codes. It does this by using any text file, web page, or graphic file as a key for a one-time pad encryption. Here’s what Wikipedia has to say about one-time pad encryption:
In cryptography, the one-time pad (OTP) is a type of encryption which is impossible to crack if used correctly. Each bit or character from the plaintext is encrypted by a modular addition with a bit or character from a secret random key (or pad) of the same length as the plaintext, resulting in a ciphertext. If the key is truly random, as large as or greater than the plaintext, never reused in whole or part, and kept secret, the ciphertext will be impossible to decrypt or break without knowing the key. It has also been proven that any cipher with the perfect secrecy property must use keys with effectively the same requirements as OTP keys. However, practical problems have prevented one-time pads from being widely used.
The key to breaking most codes lies in discovering patterns, and in a properly implemented one-time pad there are none. Not to delve too far into details, but the point of eCoder Ring is that it plucks letters out of a keyfile and uses the numerical position of those letters to represent the letters of your message. eCoder Ring lets you use things like digital pictures (which it converts to ASCII numbers and characters) as keyfiles. It also allows you to skew the code by adding variables to start your code further down in the keyfile, or skip numbers, and do all other sorts of random files. Even if you had eCoder Ring and the keyfile used to generate a message, it would be practically impossible to crack a code generated by it without the proper variables inserted into the program.
It is my belief now as it was when I wrote it that the codes generated with eCoder Ring are impervious to brute force attacks. To prove my point, when I released eCoder Ring I included a code and offered a reward for cracking it. At first I was offering a hundred bucks; later I upped it to two hundred, and I think I may have raised it to five hundred at one point. The reward for cracking the code is moot because without the keyfile or the skew variables, the code is unbreakable. In theory I feel confident about offering a million dollars, but I wouldn’t do that for two reasons, the second of which exposes the weakness of eCoder Ring. The first reason is quite simply that I don’t have a million dollars. The second reason, the scarier reason, and the weakness that plagues all implementations of one-time pads is that both the sender and the receiver have to know what the keyfile is. I know what the keyfile is for the message I encoded. For a hundred dollars I am hoping someone does not kick in my front door, hold a gun to my head and demand access to the keyfile. For a million dollars, someone might. When I wrote that original readme file five years ago that contained the code, I specifically made it clear that the keyfile does not exist on any computer I have control over (not my laptop or my desktop and not my server) and no one else knows what the keyfile is, so bribing my kid with candy or PlayStation games won’t work.
But yes, as I joked in the program’s readme file, any codes generated with eCoder Ring will stand thousands of years of brute force attacks, but will fail in seconds when someone shows up to your house and begins to peel your children’s fingernails off as you watch. As a human being who knows the keyfile, you are eCoder Ring’s weakest link. If the keyfile is stored improperly or transferred improperly, the code can be compromised. When some mug shows up and decides to squeeze the cider out of your Adam’s apple for the keyfile, look out.
So why am I writing about eCoder Ring again after all these years?
From 2007 (when I released it) to 2012, eCoder Ring was downloaded approximately 2,000 times.
In the past two months, eCoder Ring has been downloaded an additional 3,000 times.
In the last two months we have learned that the NSA either gathers or simply pilfers through pretty much everything we do on the Internet. They store records of what websites you visit. They keep track of who you e-mail, and how many times you do so. Most signs point to the fact that the NSA has direct connections to some of the largest content providers in the world and pull data pre-encryption, making the phrases “HTTPS” and “SSL” mean almost nothing. The latest NSA-related leak tells us the NSA pays 35,000 people to break codes and crypto. I hope one of those 35,000 guys runs across a code generated with eCoder Ring someday. That would make me chuckle. There are also rumors that the NSA can effectively either crack or circumvent some/most/all encryption methods being employed today.
Based on the increase in downloads, do I think eCoder Ring is the answer?
No, obviously. It’s too cumbersome to be used on any mass scale and too difficult to properly implement. (What I had always imagined implementing (but is beyond my skills) is an API or something that could be used in chat programs, so instead of sending clear text back and forth across the internet, people could send random-looking encoded text.) What these recent downloads tell me based on current events is that normal people are interested in security. Normal people are interested in learning about codes, and keeping their messages away from prying eyes. Normal people are hitting search engines and looking for ways to regain their privacy. eCoder Ring probably isn’t the answer, but maybe it’ll inspire someone else to create the answer.
Link: eCoder Ring
Posted at 6:00 pm by RobOHara in Hacking
A few months ago I spun up a new website, SpriteCastle.com. There’s no real content there yet — it’s more of a proof of concept site at this point. Last night after finishing up the latest episode of You Don’t Know Flack I decided to do some tweaking to the Sprite Castle. When I opened the site in Google Chrome, I got the following message:
Crap. I know WordPress has been under attack lately, so my first assumption was that the site had been compromised. Bypassing Chrome’s warning, I opened the site and searched for any sign of malware. I couldn’t find any. I then clicked “View Source Code” and quickly found the problem — links to a “posh laptop bag” website. While viewing the page itself I couldn’t see the link, but while viewing the code there it was, plain as day. A quick Google search shows that I’m not the only person running WordPress with the issue.
After a few minutes of research I tracked the problem back to the free WordPress theme I had downloaded. The theme was injecting links to sites hosting malware in the theme’s footer, and the links were encrypted (technically, obfuscated) making them difficult to find while sifting through the code.
There are lots of websites out there like this one that will help you remove encrypted footer links. Even with those removed, I was still seeing links in my source to malware sites. By using Windows’ FINDSTR command (similar to GREP) I was able to find more encrypted sections (hint: search your PHP files for “EVAL”). Each time I tried dinking with the code, the website would stop loading. Someone spent a lot of time putting those encrypted links into this particular theme.
So, I spent a lot of time getting rid of them.
The simplest branching point in any programming language is the IF…THEN clause, which does exactly what it sounds like:
IF (this) THEN (do this)
One baby step beyond that is IF…THEN…ELSE logic. Even if you are not a programmer you can see that this is used in every single program.
IF PASSWORD IS CORRECT
– ALLOW USER TO LOG IN TO E-MAIL
– PRINT “Denied!”
This was also, in its simplest form, the basis for most early forms of copy protection. Consider the old paper-based protection schemes that required gamers to enter a code to play a game.
HAVE USER ENTER CODE
IF CODE IS CORRECT
– RUN GAME
– DO NOT RUN GAME
Once you understand this logic you can see that with a minor change, programs could be re-programmed to always load. Or, “cracked.”
HAVE USER ENTER CODE
IF CODE IS CORRECT
– RUN GAME
DO NOT RUN GAME
Again, simple. No matter what the user enters at the prompt, the game loads. There are other ways to do it, of course. Another simple way would be to tell the program that no matter what the user enters, it’s correct.
HAVE USER ENTER CODE
CODE IS CORRECT
IF CODE IS CORRECT
– RUN GAME
– DO NOT RUN GAME
In this instance, no matter what the player enters, we tell the code that it was correct and the program continues down that path.
This is essentially how I removed the malware from the theme. The theme checks to see if a particular file exists on the computer. If it is, it reads a serial number from the file. If the serial number checks out, the malware links are removed from the footer.
CHECK TO SEE IF LICENSE FILE EXISTS
TELL PROGRAM FILE EXISTS
IF FILE EXISTS
– DO NOT INJECT MALWARE LINKS
– INJECT MALWARE LINKS
A quick check of the theme’s output showed that the technique worked and the malware links had been removed. With that part fixed I began systematically removing all the malware-seeking code. It took a couple of hours, but I think the entire theme is now clean.
Unfortunately, once Google detects malware on a site it removes the URL from its search engine (SpriteCastle.com no longer shows up in Google searches) and Google Chrome still flags the site as one that hosts malware, even though the links have been removed. To get re-added, a request has to be submitted to Google and a scan of the site has to be performed. That ball’s already started rolling, so hopefully in the next day or two I’ll be back in business.
Posted at 8:00 am by RobOHara in Hacking, Podcast
Another week, another episode.
Episode 119 of You Don’t Know Flack is about Hohocon — specifically Hohocon ’94, the last Hohocon and the only one I attended. Hohocon was a hacker conference that ran for 5 years in a row, from 1990 to 1994. It was put on by dFx, the Cult of the Dead Cow, and Phrack Magazine.
This was a tough episode to complete. During the time slot I set aside to record, my sister inconveniently and inconsiderately had a baby. Don’t you hate it when other people schedule things when you already have plans? Sheesh! All kidding aside, I spent a few hours at the hospital yesterday and a few hours watching the NFL playoffs yesterday, just enough to set me back half a day. On top of that I spent 90 minutes recording and another 3 hours editing my own babble.
Listen to me ramble. I sound like Jodie Foster’s award speech from last night, except I’m not coming out in this post. Unless it’ll increase my number of subscribers.
Link: YDKF Episode 119: Hohocon ’94
Facebook: You Don’t Know Flack
Posted at 6:00 am by RobOHara in Hacking, Videogames
Last week at the 27th annual Chaos Communication Congress (CCC), a group calling themselves “fail0verflow” displayed the single-most important PlayStation 3 hack to date. A few months from now, when everybody who wants one has a modified PS3, you’ll be able to point your finger back to fail0verflow’s CCC presentation and say, “that is where is all began.”
Just like the original Xbox, the PlayStation 3’s defenses didn’t fall to pirates, but to Linux experts. The quickest way to have your security precautions ripped out of your device, run up the flagpole and laughed at is to prevent people from running Linux on it. In fact, the general consensus has been all along that since the PlayStation 3 allowed users to install Linux on an unmodified console, Linux hackers have had no incentive to tinker with the console’s security measures. As a result, the PS3 has remain “unbroken” for over four years, the longest of any modern console. However in the late spring of 2009, Sony removed the OtherOS feature from PlayStation 3’s through a mandatory (if you want to play online and/or new games) BIOS upgrade. While this made a lot of PlayStation 3 owners mad, it apparently made fail0verflow really mad.
The reason your PS3 (or any game console) won’t play a copied disc is because games must be digitally signed. As with any encryption, this digital handshake requires a private key and a public key. A PlayStation 3, using its private key, examines public keys and, based on its findings, determines whether or not to execute the code. This is why games you buy off the shelf will run on your PS3, but a copy of that same game will not.
(Old mod chips for the original PlayStation used to trick consoles by returning the right answer, regardless of what the question was. The PS1 was looking for region codes instead of digitally encrypted signatures, but the concept was the same. When a backup copy was inserted into the original PlayStation, the console would ask, “should I play this game?” The console checked for the region code and, when it could not be found, would reply with “no.” That response was sent back through the modchip, who slyly changed it to “yes!”)
While digging through the PlayStation 3, fail0verflow didn’t just find a private key — they found the private key. The master root encryption key. Using this key, hackers can generate working public keys. With valid public keys, hackers can boot anything they want on the PS3. There are two important things to note here. One, is that this key is included in the PlayStation 3’s hardware. It does not appear that a BIOS upgrade can change the master key. And two, changing the key could cause all PlayStation 3 games to stop working — so that’s not very likely. fail0verflow went looking for this key in the name of Linux. Other folks may not be so kind.
You know how there’s that one guy that takes things to another level? In the hacking world, that guy is GeoHot. GeoHot perfected the iPhone jailbreak; if your iPhone is jailbroken, you owe it to GeoHot. The PlayStation 3 has been a thorn in GeoHot’s side for quite some time now. He’s picked at it, poked at it, and even released a couple of hacks that were eventually closed up by Sony. fail0verflow announced that within the next month, they plan on releasing some tools that will allow the homebrew and hacking communities to start looking at the PS3. GeoHot said to hell with that, and posted the master key on his website.
Click to Enlarge
Right now, this kid’s house is probably surrounded by lawyers. Or assassins. Or both.
Now, I don’t know what to do with that number, and chances are you don’t either, but you can get your booty there are people that do, people that have been waiting four long years for those numbers. The PS3’s homebrew and hacking scenes are about to light up. I can’t wait to see what happens next.