If an exploit was found some how to overcome the hypervisor then just do not update your PS3 until a Custom Firmware has been made
And for my personal opinion, I really think the way to go would be the killing of the hypervisor process (AFAIK it runs on a SPU). That thing can't be perfect, I've done my cracks, others have done them, Steam was supposed to be impossible to crack, so this process, because it's nothing more then that, can be killed or fooled someway. The problem is that "someway"
For this HDD info, it's cool to know what is in the HDD and this kind of information can ultimatly be used to code, for example, a linux filesystem driver to read the hdd directly. Not that this can be used to exploits or something like that, but it can make our lifes easier in some way, like copying files into the ps3.
Anyway, good to know there is progress being made. Keep up the good work guys!
Another problem is that built in firmware creates virtual partitions for both ps3 os and otheros on the hdd. The main partition is in firmware memory , after that the OS in the main memory controls all other partitions and a rule for hardware lvl`s for os`s.
Hi. I don't presume to know to much about anything, but, it it possible to find the encrypted key pair just using brute strength attack using a WHOLE bunch of computers connected with distributed computing? Like seti at home or that PS at home protein software?
I know this question is probably completely obviously unreasonable to most of you. but I don't understand why not.
Thanks for taking the time to help/inform/deal with me and other non-devs like me.
Do a search for the KeyVault project... it was investigated a year or so back, but was determined by the lead PS3 Devs (Gigi, Subdub, StrontiumDog) that it wouldn't end up working.
Depending on the key type and length it may not take hundreds of years... the FBi and private organizations like Guidance Software use rainbow tables to hack even the hardest of encryption keys with "CUDA" google it we would have to get a driver for the rsx in linux to work or develop a program to do it more efficiently.. Plus my favorite Truecrypt isnt so secure anymore as well Google truecrypt security flaw. encryption can be defeated its how efficiently you can do it.. and at the present time it would be sloppy at best.
i was just curious on how the hard drive filsystem is layed out specifically i heard mention that there is no continuity to file placement? if the User content is available could $orny be using multiple file allocation methods to confuse pirates IE binary blobs instead of file headers?
- CUDA has been used to accelerate bruteforcing some alogirthms used in WiFi security (there are implementations of [email protected], [email protected], video encryption, etc. using CUDA as well)
- Rainbow tables are used to accelerate some operations used in breaking some encryptions. It itself is not a technique for breaking any cipher.
- Some of the ciphers are still believed to be strong enough to withstand even attempts of FBI and the likes.
- To effectively employ these techniques you need to know what you're trying to break (i.e. what is the cipher, mode, etc.)
- Continuous file placement? If you mean that one file resides on a series of sectors without gaps that is a thing that no filesystem can guarantee without great performance penalty (just try to think about it a bit harder and you'll see why). That is why this usually doesn't work like this and if you want to to ensure it is so, you need to defragment from time to time.
Well, it depends on the encryption / standalone program.
For starters, its usually not the FBI who goes off cracking things, its the NSA, they have tons of buildings full of "supercomputers" for just that, not to mention it seems anyone who works for them has multiple doctorates in all kinds of crazy math!
However, depending on the person, the algo's weaknesses / program backdoors may and have been disclosed to the NSA and agencies alike.
Just something to think about!