The fact that it knows the passwords match mean one of: there is no hash, the hash is not salted, or the salts are the same.
The first and third seem pretty unlikely, so I'm assuming that the hash is unsalted -> fail.
Err, no?
When you salt a password, they're stored together. For example, Linux passwords are storedl ike this:
ron:$1$C8i1C6/t$d.SI5o5dcBuLh5rF2DMU90:14153:0:99999:7:::
The first part ("C8i1C6/t") is the sale, and the second part ("d.SI5o5dcBuLh5rF2DMU90") is the hash. You can easily verify whether or not a password matches that salt. If you can't verify that an arbitrary password matches the hash, then how is it supposed to be used?
And for what it's worth, this is full disk encryption software, which means that the harddrive is encrypted with AES256 (or similar). The symmetric key for AES is encrypted using my password, so at the very least it'd be possible to attempt a decryption with the password that was presented and see if that works. But I even doubt that that happens -- it likely stores the hashed password for quick verification.
(Note: that's the actual line from my /etc/shadow file on my laptop -- if somebody can PM me my actual password, I'll send them a prize. I'll be *extremely* surprised if that happens, though, so it'll be an exceptionally nice prize.
)
