Login required to started new threads

Login required to post replies

Prev Next
Re: The Best Marathon Investigation Story Yet [racehd] [ In reply to ]
Quote | Reply
You need more than a hash.

If you just do a hash, you have no guarantee of integrity of the data as I can tweak it and run the hashing algorithm again and you'd have no idea I've done it.

You need a cryptographic hash, something along the lines of a HMAC-SHA256 (although that requires symmetric keys which could be problematic for public verification). What I'd do is implement a public / private key signing pair (Garmin could publish their public key for their devices which would allow anyone to verify that the signed file is valid but unless you crack Garmin's private key no one could tamper with it) signature system. ECDSA would be a good choice as it's pretty robust. RSA would also be a good choice (and probably better for a Garmin as it's easier (read cheaper) to accelerate in hardware).
Quote Reply
Re: The Best Marathon Investigation Story Yet [Thomas Gerlach] [ In reply to ]
Quote | Reply
Now, it's hit the "big time". Reddit front page: https://www.reddit.com/..._exposes_phony_time/
Quote Reply
Re: The Best Marathon Investigation Story Yet [Ai_1] [ In reply to ]
Quote | Reply
Ai_1 wrote:
I'm not sure they are all defending the cheater, I think some simply have reservations about where the line lies between exposing them and harassment/bullying. I don't have a problem with calling attention to that.

I think we are rightly furious when people cheat, but I do think we're too inclined to have absolutely no mercy on wrong-doers, as though that will somehow make it better. It won't. And it's not just about whether they "deserve it", it's about who the rest of us want to be as well. It's not confined to this subject. The same tendencies have always been around and are at the root of most mob mentality. You see the same thing whenever there's a public discussion of prisoner rights, for example. The "mob" tend towards the idea that they are scum and deserve no rights. I disagree. Judge a society by how well it treats its prisoners and outcasts.
Well said.


http://www.jt10000.com/
Quote Reply
Re: The Best Marathon Investigation Story Yet [Kula] [ In reply to ]
Quote | Reply
Quote Reply
Re: The Best Marathon Investigation Story Yet [ilsapproach] [ In reply to ]
Quote | Reply
ilsapproach wrote:
You need more than a hash.

If you just do a hash, you have no guarantee of integrity of the data as I can tweak it and run the hashing algorithm again and you'd have no idea I've done it.

You need a cryptographic hash, something along the lines of a HMAC-SHA256 (although that requires symmetric keys which could be problematic for public verification). What I'd do is implement a public / private key signing pair (Garmin could publish their public key for their devices which would allow anyone to verify that the signed file is valid but unless you crack Garmin's private key no one could tamper with it) signature system. ECDSA would be a good choice as it's pretty robust. RSA would also be a good choice (and probably better for a Garmin as it's easier (read cheaper) to accelerate in hardware).

There's an additional trick that has to be performed: you must secure the trusted private key. It needs to be located on the device to produce the signature, and I think that's enough to invalidate the whole idea. Fortunately, hardware exists to try to protect secrets like keys while allowing their use in some computations, but we can't reasonably expect it to stay secure forever. Additionally, by virtue of how this system remains secure, the key cannot ever be replaced if it is compromised. At some point, you have to let the device containing the key out into the open world, and surely one of the people interested in purchasing it is also motivated to extract the private key. This is an arms race that game console manufacturers and media companies trying to enforce use of DRM have been fighting since day one, and they have yet to truly succeed. It would buy a mild confidence increase until someone leaks the key to the world, and then the feature becomes worthless.
Quote Reply
Re: The Best Marathon Investigation Story Yet [zten] [ In reply to ]
Quote | Reply
zten wrote:
ilsapproach wrote:
You need more than a hash.

If you just do a hash, you have no guarantee of integrity of the data as I can tweak it and run the hashing algorithm again and you'd have no idea I've done it.

You need a cryptographic hash, something along the lines of a HMAC-SHA256 (although that requires symmetric keys which could be problematic for public verification). What I'd do is implement a public / private key signing pair (Garmin could publish their public key for their devices which would allow anyone to verify that the signed file is valid but unless you crack Garmin's private key no one could tamper with it) signature system. ECDSA would be a good choice as it's pretty robust. RSA would also be a good choice (and probably better for a Garmin as it's easier (read cheaper) to accelerate in hardware).


There's an additional trick that has to be performed: you must secure the trusted private key. It needs to be located on the device to produce the signature, and I think that's enough to invalidate the whole idea. Fortunately, hardware exists to try to protect secrets like keys while allowing their use in some computations, but we can't reasonably expect it to stay secure forever. Additionally, by virtue of how this system remains secure, the key cannot ever be replaced if it is compromised. At some point, you have to let the device containing the key out into the open world, and surely one of the people interested in purchasing it is also motivated to extract the private key. This is an arms race that game console manufacturers and media companies trying to enforce use of DRM have been fighting since day one, and they have yet to truly succeed. It would buy a mild confidence increase until someone leaks the key to the world, and then the feature becomes worthless.

Agreed. The private key must be secured on the device as well. There are certainly techniques out there (embedded on die, fuses, and so on). There are some techniques out there that also allow you to update the private key you are using (I'm hesitant to say more as this is a field that I work in and have to dance around NDA and IP issues), but it's tricky for sure to make sure only the authorized person can do so. Yes, it's an arms race for sure. The cost and complexity for doing so may not be something worth dealing with for Garmin (or other manufacturers). Can I secure the secret against a nation-state attack, probably not. I can defend against reasonable attacks on physical extraction by implementing anti-tampering techniques (like if someone cracks open the device it disables the key). No cyber security is ever 100%. If you avoid using the same key on every device then you improve security further - and further drive up the cost and complexity of managing it.

It's a solvable problem, but the cost and complexity of making is robust against attack is (IMHO only) probably not worth it for the device manufacturers. I would hazard a guess than probably well less than 1% of their buyers would care about such a feature.
Quote Reply
Re: The Best Marathon Investigation Story Yet [Thomas Gerlach] [ In reply to ]
Quote | Reply
More time on the couch perhaps?
Quote Reply

Prev Next