[nzie]

and i was liking his post.lol

[mrsoczi]

Originally Posted by Mathieulh Just so you know, the first exploit from my post, which I like to call the "meta exploit" which allows to decrypt the metadata directly in the shared LS works on the bootloader.

??

Quote from PS3Crunch

[hackeyking]

There is most likely a bootrom or a crypto engine that decrypts and load the bootloader/metldr.

another from math

[medi01]

So the new thing in this was:
a) how geohot COULD have cracked metldr
b) a hint of "if you're smart enough, you could use this approach to get boot leader keys", something that Matieulh has already said a while ago on a different occasion

Originally Posted by hackeyking Lets hope we still have talented people (who havent been pushed away by stupid comments and small children) who can get us somewhere....

Let's hope we get enough of "you can't say 'movie sucks' unless you create a better movie" enlightened cretins, that would kiss drama queen's butt long enough for something to come out of it....

[mcmrc1]

Original Post from Asure:
There are two exploits released by Math. One is in source code, and dumps metldr. The keys to old-style 3.55 & 3.56 stuff are inside mtldr. If Sony did not go the 'lv0 vault' route, this exploit would still hold true and allow us to decrypt 3.56+. Except, they didn't, and they moved all into LV0. So the first exploit is cool to know how, but fail for us, Sony changed stuff.

From your pictures, did you deduct LV0 replace metldr? If so, do so now. Then, our goal becomes on 3.6x and 3.7x to decrypt lv0.

Since lv0 is encrypted with the same key0 used to encrypt metldr in 3.56+.. we need a second trick to get that private key they used on metldr.
This second trick i don't see in math's .c examples. But he explains it..

STEP II)
Contrary to popular beliefs, you do not need to know the public key to calculate the private key, you just need two decrypted signature, you now know how to dump these, so let's assume you just did, now all you have to do is to bruteforce the curve type by constantly reloading a self to metldr, the curve type being only 1 byte, that would be 64 possibilities.

The explanation Math made before, about how there are two exploits in one, that's about the meta-sploit as he calls it, and the one where you bruteforce a private key.
I was merely pointing out the need for skeet to get bootloader dumped. If we can dump bootloader already by software, i must have missed that. If you can dump it, do so now

You see, if both bootldr and metldr are same key, what prevents us from say, using the same exploit with a dumped bootldr?
We need the two meta results back, to start calculation/brute force. So the original exploit should be modified to also print out the meta data.

If all we need is meta results, we don't even need to run the bootldr. Math explained it all allready, but nobody is reading and i lack C skills
Just get load the ldr, and dump the meta results. All we need to do recycle the exploit example code for metldr. We don't even need execute the code, the bug is in the loading routing which does not check the length.

Once we have dumped metadata for bootldr and metldr, it seems Sony fail in the key0 too, right Math? Otherwise it would not be possible to bruteforce?

[itriedbutifailed]

this is worthless like my penis. im going to find 3.6+ keys by myself.

[TizzyT]

Originally Posted by itriedbutifailed this is worthless like my penis. im going to find 3.6+ keys by myself.

good luck.

[japsander]

Originally Posted by itriedbutifailed this is worthless like my penis. im going to find 3.6+ keys by myself.

As worthless as your post then? I think your username describes how you will fare in your quest.

No information is worthless, there are only people who cannot use said information.

[H3avyRa1n]

there's new intel floating aroung provided by @mathieulh that surely is extremely helpful..

[baileyscream]

from ps3crunch HERE

Mathieulh post #233
Crunching Developers!

.........................................................................................................................
Originally Posted by TheFrogEater

math do you think ur exploit can be use to get the keys for cfw ?


salut mat il est impossible de t'envoyer un mp donc je te le dis ici, je trouve que t'es vraiment fort et je te remercie pour ton travail aussi bien que récent que passé (psp etc).
tu penses que ca pourrait apporter quoi a plus ou moins long terme ?
.........................................................................................................................

The first one I published can be used in conjunction to a hardware hack to get the bootloader to decrypt lv0 for you, you can also calculate lv0's private key through this, this means you can decrypt lv0, encrypt lv0 using the existing decrypted metdata keys, and sign lv0, you just need to use a ps3 to go through the process of decrypting each lv0 revision though. Know that if you sign lv0 you can also use yet another exploit to gain code execution on the bootloader and dump the keys and be able to decrypt lv0 straight on pc.


Mathieulh post #234
Crunching Developers!

..........................................................................................................................
Originally Posted by Beharii

Math if all the loaders are loaded by lv0 since 3.56+ what's the use of metldr anymore, and is the per_console_key that is obtained from metldr the same as the one from bootldr since in the wiki it is only stated as per console key. I hope you don't mind answering these two questions since I couldn't figure them out in the wiki.
..........................................................................................................................


They are not loaded by lv0 they are simply embeded inside one of the lv0 segment in self format, lv0 copies them to the lv1 loaders' table on the xdr and then lv1 fetches them from ram and loads them through metldr just like on earlier firmwares.
There is a new distinction though, each ldr has its .2 such as appldr.2 which is a new metadata for the loader, if your console happens to have metldr.2, lv1 will feed it the .2 metadata generated with the metldr.2 keysets and private key with the original ldr data sections, metldr.2 will decrypt the *ldr.2 metadata and will use it to decrypt the matching *ldr data sections (so the data section keys match the ones from the original *ldr metadata)

Note that because the signature fail from Segher doesn't work for metldr.2 you cannot use the donut fail to pwn it because you cannot sign your own loader, making the new metldr.2 extreamly hard to exploit (though not impossible, to cut with the suspense it's just the good old metldr with a new keyset, a hardcoded 3.60 min version, compiled with a newer gcc, not much worth exploiting in the first place).

Last edited by Mathieulh; Today at 12:27 PM.

(((((DONT THANK THIS POST AS ITS NOT MY POST. I'M JUST PASSING IT ON)))))