5 byte seed key algo

E38 E92 and many others. Approximately 2007 and newer
nmstec
Posts: 6
Joined: Wed Nov 29, 2023 3:03 pm
cars: 2021 F350, 11 Q7, 11 q5

5 byte seed key algo

Post by nmstec »

Basically I have an unhealthy obsession with figuring out everything, even when not needed or not in scope. And this one is driving me up the fucking wall.
Usual known information I wont go over in depth, we take the aes key hash, ignore first 2 chars, decode, and then use first 32 out of 44 bytes for the actual key calc. Great.

Question here being, does anyone know what the last 8 bytes are for?

Known factors:
after 32 bytes, we have 3 bytes with some form of marker, always 00 f5 00, fine, not really interesting. Is also the same throughout all 256 keys.
Next byte after that being the algo ID itself.
Theres plenty of references in the sa015 that is is a sha256hash of the password. However that seems to include ALL the data after the first 32 bytes

Gone through DPS source as much as I can, as well as sa015, and still fuck all reference to it.
Last edited by nmstec on Wed Jun 25, 2025 4:45 pm, edited 1 time in total.
nmstec
Posts: 6
Joined: Wed Nov 29, 2023 3:03 pm
cars: 2021 F350, 11 Q7, 11 q5

Re: 5 byte seed key algo

Post by nmstec »

Update: Boring. Its a SHA256 of the first portion of the data. Just a verification. I failed to reload the project and recheck that I did from like 2 years ago.
User avatar
antus
Site Admin
Posts: 10014
Joined: Sat Feb 28, 2009 10:34 am
cars: TX Gemini 2L Twincam 8psi
TX Gemini SR20 18psi
Datsun 1200 Ute
Subaru Blitzen '06 EZ30 4th gen, 3.0R Spec B
Subaru WRX 2007

Re: 5 byte seed key algo

Post by antus »

I deleted your image attachment as it contained keys. I am not comfortable with sharing the encryption keys on the site, but I will allow the discussion of how it works.
Have you read the FAQ? For lots of information and links to significant threads see here: http://pcmhacking.net/forums/viewtopic.php?f=7&t=1396
nmstec
Posts: 6
Joined: Wed Nov 29, 2023 3:03 pm
cars: 2021 F350, 11 Q7, 11 q5

Re: 5 byte seed key algo

Post by nmstec »

To be fair those keys are not used, which is why I selected them. I can generate a few fake ones if you prefer?
nailgg
Posts: 2
Joined: Thu Jan 05, 2023 11:22 pm
cars: Passat B8

Re: 5 byte seed key algo

Post by nailgg »

nmstec wrote: Wed Jun 25, 2025 4:40 pm Update: Boring. Its a SHA256 of the first portion of the data. Just a verification. I failed to reload the project and recheck that I did from like 2 years ago.
Correct, but the count of the SHA256 operations will differ from seed to seed.
User avatar
andbel
Posts: 7
Joined: Mon Nov 04, 2024 1:39 am
cars: 2010 Tahoe
2011 HHR
2018 ATS
Location: JC, NJ, USA

Re: 5 byte seed key algo

Post by andbel »

Saying 'hi' and subscribing :)
nmstec
Posts: 6
Joined: Wed Nov 29, 2023 3:03 pm
cars: 2021 F350, 11 Q7, 11 q5

Re: 5 byte seed key algo

Post by nmstec »

nailgg wrote: Wed Jul 30, 2025 2:28 pm
nmstec wrote: Wed Jun 25, 2025 4:40 pm Update: Boring. Its a SHA256 of the first portion of the data. Just a verification. I failed to reload the project and recheck that I did from like 2 years ago.
Correct, but the count of the SHA256 operations will differ from seed to seed.
Yeah the calculation I know of ages ago, I just wanted to figure out what the rest was for.