Actually the brute-force method doesn't need to test the data. It only needs to test the decrypted key sector by looking for certain strings.
got HW crypto? On the (in)security of a Self-Encrypting Drive series:
http://eprint.iacr.org/2015/1002.pdfIn any case, IIUC, My Passport Ultra's the HDD controller is doing the decryption/encryption, not the USB-SATA bridge.
Quote:
The INIC-3608 microprocessor is based on an ARC 600 CPU to bridge USB to SATA once again. This chip does not have a hardware accelerated AES engine. At first this seemed rather suspicious, as the package of the device advertises with hardware based encryption. If done in software there is no chance with an observed clock speed below 100 Mhz to encrypt at USB3.0 speed. Nevertheless as advertised, the MP-Slim and MP-Ultra do hardware accelerated AES encryption, but not on the bridge. It turns out, the HDD controller is doing the en- and decryption of user data. The USB to SATA is performing the user authentication and supports the standard VSCs from WD.
Quote:
We tried to connect the SATA HDD directly over the PCB [4] for accessing a raw sector, but did not succeed. The data access is restricted by a set ATA-password in the SATA-HDD. The bridge is therefore doing user authentication by setting an ATA password, once the user generates his password. If the user does not supply a password, the user data is inaccessible when connecting directly to the SATA ports.
Finally, we were able to bypass the ATA password with commercial tools. Nevertheless, this is not an off-the-self solution offered, so we worked our way through to the AES protection. We located the location of the ATA password and some (unknown) connection to the AES password in different SAs from the internal 2.5" SATA HDD. After resetting the ATA password, we had complete access to the decrypted user data, as the SATA chip decrypted on-the-fly. Regardless of the user-password, KEK or DEK.
Here is what the researchers have to say about SED encryption for drives based on the JMS569 bridge:
Quote:
Facing a protected HDD is not new problem for HDD forensics. As there are already existing commercial solutions (e.g PC-3000), we analyzed the HDD directly with those tools. Their approach seems to follow a straight pattern, which allows SA access by overwriting the RAM/ROM and bypass security features like ATA passwords and optionally AES keys. By forcing SA access and manipulating the SA area 0x124 and 0x127 we were able to unlock the HDD and disable the SATA AES encryption. Note that this works always, independent of the chosen user password and bridge status.
FAQ
Search
Login
Register
HDDGURU FILES
