Hikvision Feeds a Troll

It’s possible to turn someone towards the light, and eventually lead them to salvation.

A prominent member of the Westboro Baptist Church, Megan Phelps-Roper made a TED speech about it. What saved Megan was not someone yelling in her face. She was conditioned to expect exactly that from the misguided heathens of the world. Instead, someone approached her with curiosity, warmth and civility and lead her out of the congregations grasp. The “enemy” is rarely a mindless drone out to do evil. Although, our leaders would prefer we see things that way.

Our industry has a variant of the WBC, and Hikvision has chosen a different approach to liberate the members of the sect.

a site that has always trafficked in nefarious insults and innuendo. Hiding behind a keyboard, the tabloid’s staff takes unfounded potshots at our entire industry, bullying one company at a time.

and

Instead, he chooses to distract manufacturers with his pursuit of financial gain and efforts to fulfill his delusions of grandeur.

The problem with this sort of message is that the hardcore members are expecting exactly this sort of rhetoric, thereby further entrenching them in their beliefs. Ultimately the blogger will surely capitalize of the increased attention being paid. I thought it was common knowledge that trolls have an insatiable appetite for the kind of copy Hikvision just released.

More please!

Members of the sect can attend a “university” (not at all like this one) and even make the “dean’s list“. This is impressive stuff, and these people are not going to be swayed by a manufacturer having a breakdown in their public relations department. Furthermore, I suspect Hikvision has several active subscriptions, thereby directly funding the site.

I think Hikvision is correct in calling it cyber-bullying. It has all the traits of schoolyard terrorism; the ring-leader points out an arbitrary enemy, then manipulates the enemy to react. Steps in to protect the flock from the aggressor. If it gets too hot, the ring-leader can count on his 3 or 4 lackeys to do the dirty work.

In this case, the sin of the “enemy” is that the company is partially owned by the Chinese government. Therefore, every vulnerability found in a Hikvision camera is proof positive that the Chinese government is spying on us. I don’t buy that. Governments don’t have to own a company to assert influence over it.

You might remember Stuxnet a vulnerability in SCADA equipment was exploitable by governments and for-lulz hackers alike. Vulnerabilities will continue to exists as long a fallible humans write the code. As long as fallible humans install and (fail to) maintain the equipment we will continue to see flaws and problems. Unfortunately, a lot of companies have deployed small time-bombs with terrible security in place, not just Hikvision.

When I was in the army, we had padlocks on our lockers. On the first day, we were instructed to get a hair-cut “to not look like faggots” (I kid you not, that’s what he said), and then to make sure our lockers were safely locked. The reasoning (for locking up) is that you can’t really trust anyone, and giving the bad apples the opportunity to steal was almost as bad as the guy stealing. At a company I worked for a long time ago (starts with an M), someone shat on the toilet seat in the offices restroom. Someone we had lunch with, talked about code, movies, politics and music with, went to the bathroom, and shat on the seat, leaving it there for some poor soul to find.

Same thing goes for your IP camera. Sticking that thing on the internet, REGARDLESS of manufacturer ownership is like leaving your locker unlocked. You are tempting the swines of the world to mess around, and when they do, we all lose.

Listening to Customers

In 2011, BlackBerry peaked with a little more than 50 million devices sold. The trajectory had an impressive ~50% CAGR from 2007 where the sales were around 10 million devices. I am sure the board and chiefs were pleased and expected this trend to continue. One might expect that ~250 million devices were to be sold in 2016 if the CAGR could be sustained. Even linear growth would be fairly impressive.

Today, in 2017, BlackBerry commands a somewhat unimpressive 0.0% of the smartphone market.

There was also Nokia. The Finnish toilet-paper manufacturer pretty much shared the market with Ericsson in Scandinavia and was incredibly popular in many other regions. If I recall correctly, they sold more devices than any other manufacturer in the world. But they were the McDonalds of mobile phones: Cheap and simple (nothing wrong with that per se). They did have some premium phones, but perhaps they were just too expensive, too clumsy or maybe too nerdy?

Talking on a Nokia N-Gage phone

Nokia cleverly tricked Microsoft into buying their phone business, and soon after the Microsoft gave up on that too (having been a contender in the early years with Windows CE/Mobile).

I am confident that BlackBerry was “listening to their customers”. But perhaps they didn’t listen to the market. Every single customer at BlackBerry would state that they preferred the physical keyboard and the naive UI that BlackBerry offered. So why do things differently? Listen to your customers!

If BlackBerry was a consulting agency, then sure, do whatever the customer asks you to. If you’re selling hot-dogs, and the customer asks for more sauerkraut, then add more sauerkraut, even if it seems revolting to you. But BlackBerry is not selling hotdogs or tailoring each device to each customer. They are making a commodity that goes in a box and is pulled off a shelf by someone in a nice shirt.

As the marginally attached customers are exposed to better choices (for them), they will opt for those, and in time, as the user base dwindles, you’re left with “fans”. Fans love the way you do things, but unless your fan base is growing, you’re faced with the very challenging task of adding things your fans may not like. Employees that may be prostrate bowed but not believing, will leave and eventually you’ll have a group of flat-earth preachers evangelizing to their dwindling flock.

It might work as a small, cooky company that makes an outsider device, but it sure cannot sustain the amount of junk that you tag on over the years. Eventually that junk will drag the company under.

Or, perhaps BlackBerry was a popular hotdog stand, in a town where people just lost the appetite for hotdogs and had a craving for juicy burgers and pizza (or strange hotdogs)

Clickbaiting Copycat Caught

It’s pretty damn hard to make secure software. Years ago I commented on Shodan and worried that the IP video industry was next.

Run of the mill ignorance, carelessness, greed what have you, is so common that we scarcely care to click the link. Recently (or not) and old bug was discovered in Intel products that allowed remote control.

Now if you are commercial blogger (or “analyst” if you prefer), you’re not going to try to shed light on the issue. That just doesn’t trigger enough clicks and drama. It’s better to make some unsubstantiated claim that an “Intel backdoor is confirmed”.

I can guarantee that someone is now looking up the word “backdoor”, I’ll save you the trouble (it’s in the link above too)

A backdoor is a method, often secret, of bypassing normal authentication in a product, computer system, cryptosystem or algorithm etc. Backdoors are often used for securing unauthorized remote access to a computer, or obtaining access to plaintext in cryptographic systems.

Wikipedia

So, yes, it is probably not a lie to use the word “backdoor”, but it sure is manipulative, something people with a certain mental defect excel at.

For l33t hackers, finding back-doors is sometimes a fun pastime. The purpose can be to cause extensive damage for lulz or filthy lucre, sometimes for companies, sometimes for governments. Usually, it’s a challenge to find vulnerabilities and defects that let’s you crawl into systems that should be locked down. But to the n00b, a backdoor might suggests that it was intentionally put there. After all, you don’t “accidentally” install a backdoor in your house.

Backdoors in code, however, come in various flavours,

• Deliberate backdoor intended to give an unknown user remote access after the user has deployed the device/software, thereby granting the attacker access. These can be baked into the device, or installed later as a trojan.
• Accidental backdoor caused by unexpected side-effects of the code. In the olden days, you could mess around IIS servers by using unicode strings in the URL.
• Accidental backdoor caused by gross negligence/incompetence on the manufacturers side. Hardcoded credentials is an example of such foolishness.

Today you are not going to get away with #1 and #3 for very long. The hackers at blackhat are not like mortal programmers, they understand assembly code, and will locate a hardcoded password or a backdoor in a few days.

But it’s a gradual scale from #2 to #3. For example, HTTP used to have something called “basic authentication“. It used Base64 encoding to hide the credentials in flight, and plenty of cameras and VMSs would use it. 15 years ago, basic authentication would probably have been considered a #2 issue, but today it’s clearly a #3 (a certain unmentionable blog used it not long ago).

You can make up your own mind if CWE-287 is a #1, #2 or #3. It could, conceivably, be a #1. But it will be very difficult to prove, unless you have network captures showing malicious activity initiated by someone associated to the manufacturer (US tech companies and NSA for example).

Another company was notified of a vulnerability on March 5th 2017, on the 12th a security bulletin is released, and the hacker then states :

“I have been communicating with Hikvision since I notified them and they have actually been been quite responsive.”

Quite responsive indeed.

Eventually we will have software in IP cameras that is safe enough that you can expose it to the internet. But for now, I would be extremely careful about opening my CCTV system to the internet.

In Hikvisions case, I think one of the issues is that to reset the cameras password you need to send a specially crafted payload to the device. This causes a lot of issues for lots of users and it strikes me as a potential attack vector. And rest assured that this is not the only issue in the cameras.

As time passes hackers find ways into older cameras that have long been discontinued, but have been deployed and are still operational, they may get more sophisticated in their attacks and find more complex ways of breaching the software.

I guess this was not as exciting a post as you had expected. I’m sorry. You will have to go somewhere else for BREAKING NEWS about the evil Chinese shell companies set up only to spy on you.

 

 

 

LinkedIn is Worse Than Facebook

I suddenly realized I spent too much time on LinkedIn, and it dawned on me that LinkedIn is even worse than Facebook.

From time to time, people post virtue signalling memes that tell other people to not let LinkedIn turn into Facebook. The want to keep LinkedIn “professional”. That makes me wonder: If your primary interaction with business partners is through LinkedIn, are you really a professional?

The feed that LinkedIn thinks I should look has a few types of posts: Politically correct trivialities, annoying riddles, links to wise words written by someone else, and outright ads and appraisal of yourself or the company you work for.

The ads (not paid ads, but companies hawking something via LinkedIn) are tolerable from my standpoint. It’s pretty easy to filter those out, and move on to something with a little more substance. When I see someone saying “See why widget XYZ from SomeCompany is leading/helping/solving…. ” then you kinda know you don’t need to continue reading. If I see a post that starts with “visit us at …” I just move on. It’s not that I would recommend the company (I still work) for to not post these things, but I wonder who is genuinely impressed by this. It seems to me that this is a lot of choir preaching, with people – who most likely already know what you’re releasing – hitting “like” on a post that tells them nothing new.

I get pointers to good copy from Twitter, co-workers and friends, and from time to time there’s a good read on LinkedIn, but to find those, it feels like an online version of walking through a large bazar looking like a gullible tourist, red-faced from too much sun, complete with selfie stick and tasteless clothing. Every single vendor grabbing your arm, telling you about their wonderfully crafted pieces of shit. If you are willing to endure this torture, you might eventually find something worthwhile, but the chances are slim, and I am getting weary of wandering aimlessly around this crazy market.

Because LinkedIn is considered a “professional” network, i.e. a network between people who only want to engage with others if there’s money to be made. That means that the posts are even more self-censored and manipulative than on Facebook, Instagram, SnapChat or what have you. Every word is carefully chosen, you remember to “like” posts, not because of their content, but because of who wrote them. You might even make a positive comment, like a quick kiss on the old sphincter: “Well done”, someone will say, when a CEO praises his own ability to turn an advantage in currency exchange into revenue growth.

Maybe, just maybe, it’s the business that I am in that is fouling up my LinkedIn feed. In any event, the remedy is quite simple. I really shouldn’t go there..

 

 

 

Magical “GPU” Based Video Decoder

I was recently alerted to an article that described a magical video decoding engine. The site has a history of making odd conclusions based on their observations, so naturally, I was a bit skeptical about the claims that were relayed to me by a colleague. Basically, the CPU load dropped dramatically, and the GPU load stayed the same. This sounded almost too good to be true, so I did some casual tests here (again).

Test setup

I am not thrilled about downloading a 2 GB installer that messes up my PC when I uninstall it, and running things in a VM would not be an honest test. Nor am I about to buy a new Intel PC to test this out (my next PC will be a Ryzen based system), so all tests are done with readily available tools: FFMpeg and GPU-Z. I believe that Intel wrote the QSV version of the h264 decoder, so I guess it’s as good as it gets.

Tests were done on an old 3770K, 32 GB RAM, Windows 7 with a GeForce 670 dedicated GPU. The 3770K comes with the Intel HD Graphics 4000 integrated graphics solution that supports Quick Sync.

Terminology

In the nerd-world, a GPU usually means a discrete GPU; a NVidia GeForce or AMD Radeon dedicated graphics card. Using the term “GPU support” is too vague, because different vendors have different support for different things. E.g. NVidia has CUDA and their NVEC codecs, and some things can be done with pixel shaders that work on all GPUs. (our decoding pipeline uses this approach and works on integrated as well as discrete GPU, so that’s why I use the term GPU accelerated decoding without embarrassment).

However, when you rely on (or are testing) something very specific, like Intel Quick Sync, then that’s the term you should use. If you say GPU support then the reader might be lead to believe that a faster NVidia card will get a performance boost (since the NVidia card is much, much faster than the integrated GPU that hosts Quick Sync). This would not be the case. A newer generation of Intel CPU would offer better performance, and it would not work at all on AMD chips with a dedicated GPU (or AMD’s APU solution). Same if you use CUDA in OpenCV, then say “CUDA support” to avoid confusion.

Results

Usually, when I benchmark stuff, I run the item under test at full capacity. E.g. if I want to test, say the CPU based H264 decoder in FFMpeg against the Intel Quick Sync based decoder, I will ask the system to decode the exact same clip as fast as possible.

So, let’s decode a 720p clip using the CPU only, and see what we get.

The clip only takes a few seconds to decode, but if you look at the task manager, you can see that the CPU went to 100%. That means that we are pushing the 3770K to it’s capacity.

Now, let’s test Quick Sync

Not as fast as the CPU only, but we could run CPU decoding at the same time, and in aggregate get more…. but we got ~580 fps

So we are getting ~200 fps less than the CPU-only method. Fortunately, the CPU is not being taxed to 100% anymore. We’re only at 10% CPU use when the QSV decoder is doing its thing:

Magic!!!

But surprisingly, neither is the GPU. In fact, the GPU load is at 0%

However, if you look at the GPU Power, you can see that there is an increased power-draw on the GPU at a few places (it’s drawing 2.6W at those spikes). Those are the places where the test is being run. You can also see that the GPU clock increases to meet the demand for processing power.

If there is no load on the GPU, why does it “only” deliver ~600 fps? Why is the load not at 100%? I think the reason is that the GPU load in GPU-Z does not show the stress on the dedicated Quick Sync circuitry that is running at full capacity. I can make the GPU graph increase, by moving a window onto the screen that is driven by the Intel HD Graphics 4000 “GPU”, so the GPU-Z tool is working as intended.

I should say that I was able to increase performance by running 2 concurrent decoding sessions, getting to ~800 fps, but from then on, more sessions just lowers the frame rate, and eventually, the CPU is saturated as well.

Grief

To enable Quick Sync on my workstation which has a dedicated NVidia GeForce 670 card on Windows 7, I have to enable a “virtual” screen and allow windows to extend the display to this screen (that I can’t see because I only have one 4K monitor). I also had to enable it in the BIOS, so it was not exactly plug and play.

Conclusion

I stand by my persuasion: yes, add GPU decoding to the mix, but the user should rely on edge-based detection combined with dedicated sensors (any integrator worth their salt will be able to install a PIR detector and hook it up in just a few minutes). This allows you to run your VMS on extremely low-end hardware and the scalability is much better than moving a bottleneck to a place where it’s harder to see.