version 5.1 Authorisation issue

[TerryO]

Hi. I've only recently joined this forum after I became a D8B owner a few weeks ago. This being my first post, I hope I'm doing it correctly. In keeping with most people on this forum I'm having my share of faults, problems and teething troubles with the D8B although I'm a great admirer of it. But to stick with the topic, re the question of the location of the ENS on the Brain Board I thought I'd contribute my 2 cents worth. I can say that it is not resident in the eprom (U38 on the schematic). To test this I made a copy of the eprom and installed it on another Brain Board (I managed to get a range of spares with the mixer). The system worked with this arrangement but the ENS reported was that of the substitute board. Naturally OS 5.1 authorisation was being sought. If one looks at the relevant schematics and the pdf called "chipper2" (a list of all the chips used throughout the D8B) you'll see reference to U16 (a DS2401) which is a so-called Silicon Serial Number.
http://pdfserv.maximintegrated.com/en/ds/DS2401.pdf
for more info. It connects to U25, the large DSP chip. I'm at a loss as to what can be done with this knowledge, except perhaps to transfer the chip to another Brain Board should the original go faulty - desoldering/soldering skills permitting!
Terry.

[anyhorizon]

Plan C, then.

Peter

[FrankH]

A little news on this subject.

After reading through this thread, I called the guy who wrote the copy protection scheme for the D8B (who has long since left Mackie) and asked him if he had any idea for a remedy. He said that the code generation software must still be up on a machine at his old non-Woodinville office.

So I wrote one of the engineers who still works there who'd probably know about this and asked. This was his answer:

"I'm on the case. Actually, I've been on the case for a month or so now. I
thought I had solved the problem, but I guess not. I came up with a solution
and sent it down to Woodinville and never heard back. I'll talk with them
and see what the status is."

I'll keep you all informed if I get any further info.

As a side story....I guess you could hold me responsible for the copy protection scheme. A long time ago....before the D8B's development, we were working on an automation system call UltraMix. It used a control surface called UltraPilot. When it came around to figuring out how to protect the software, they were going to use PACE. I hated that company and the miserable crap that was (and still is) their software. So I asked: "why can't we use the hardware like a giant dongle?" Looking to avoid third party product support hassles and the licensing fees, they bought the idea.

The lead engineer said there was a chip installed on the board that had been purposed for a job that had since been scrubbed from the featureset. It was still in place and could be programmed to work as a dongle inside the machine.

He figured it out and that's what was done. The same idea was used in the D8B.

[TerryO]

"It was still in place and could be programmed to work as a dongle inside the machine" seems to imply that there is a way of programming the Ds2401 installed on the board. It would be wonderful if it was so.

However, reading the info in the Maxim Integrated pdf for the DS2401 where it says "Unique, Factory-Lasered and Tested 64-Bit Registration Number(8-Bit Family Code + 48-Bit Serial Number + 8-Bit CRC Tester); Guaranteed No Two Parts Alike", my take on it is that the chip is programmed at the manufacturing stage and is fixed.

Terry.

[Recording Drummer]

See next response please

[Recording Drummer]

I just got off the phone with Mackie. As of now unit further notice Loud Technologies is going to take the Mackie d8b 5.1 software down off their website because of authorization problems. Any suggestion to get to the bottom of this without Mackie's help?

[doktor1360]

I just got off the phone with Mackie. As of now unit further notice Loud Technologies is going to take the Mackie d8b 5.1 software down off their website because of authorization problems. Any suggestion to get to the bottom of this without Mackie's help?

This has now reached the level of Absurdity; DAFUQ, Mackie?!?

TerryO nailed it, the DS2401 is THE heart of the matter. Looking over the data sheet, it's a pretty simple & elegant method of implementing a 'dongle' on the hardware. You get an ESN 'baked in' at the chip factory, that can then be written to for simple storage (ie. Authorization Code or similar) by using the Write Cycle of the ADSP-2181 chip (U25) it's wired to. The kicker is, when the OS 5.1 image is loaded and booted, there must be a snippet of code that performs this authorization; checking the entered authorization code against the ESN read from the DS2401. Spidey Sense is telling me that there seems to be some sort of hashing algorithm in play here; my thoughts are as such:

1) OS 5.1 boots and requests an authorization code.
2) Code provided at input is quickly checked to see if the entered character string entered by the user matches a particular bit mask/pattern [XXXX-XXXX-XXXX], an example being something like 0000-61EF-D8B2
3) If it matches, writes the entered code to the DS2401 memory register, and requests a reboot.
4) Upon reboot, the authorization routine then executes; reading the ESN & entered Auth Code, running them against a hashing algorithm and then checked (some kind of CRC?!?) - if it doesn't match, Rinse/Repeat from Step #1 onward

So, is there a way around this?!? UNLESS someone has a Universal 'Back Door' type code, I don't see a really simplistic, easy solution. More on this as I digest this latest bit of info from 'Recording Drummer', I'm currently at the office...

Peace

[RonnR]

doktor PLEASE dont give up!!!
Your our only hope obiDok☺

[doktor1360]

1) OS 5.1 boots and requests an authorization code.
2) Code provided at input is quickly checked to see if the entered character string entered by the user matches a particular bit mask/pattern [XXXX-XXXX-XXXX], an example being something like 0000-61EF-D8B2
3) If it matches, writes the entered code to the DS2401 memory register, and requests a reboot.
4) Upon reboot, the authorization routine then executes; reading the ESN & entered Auth Code, running them against a hashing algorithm and then checked (some kind of CRC?!?) - if it doesn't match, Rinse/Repeat from Step #1 onward

Nice, huh... this whole OS authorization Side Show has me so annoyed that I'm replying to my own friggin' posts now...

Meanwhile, back at the D8B 5.1 Ranch...

Well, I don't think what I've postulated above is completely correct by any means. Upon further review, I'm thinking the Mackie engineers didn't use that DS2401 for anything other than to provide a mutually exclusive (immutable) identifier to each surface unit - the Brain Board when you get down to the Brass Tacks of it all. Correct me if I'm wrong, but if you successfully install AND authorize the Whole Enchilada and then your boot media (HDD, CF) melts down, you obviously need to re-install the operating system. Upon subsequent reboot, you WILL be prompted for an OS 5.1 authorization code, is that not correct? If so, the authorization response (code, some sort of data, whatever it is) would almost assuredly be stored on the boot media; same as the plugin authorization codes, but the plugin authorizations don't require reading the ESN when they're loaded as the OS boots. That would make sense in regards to when you hook another desk into the CPU, the plugins would be loaded but the OS would be 'looking' for ESN match to the stored authorization data (again, whatever the fuck that is) generated by the Brain Board of the original desk.Does this make sense?!? I'm just postulating, shooting from the hip is more accurate, and would greatly appreciate any/all input on this. I'm sure someone will have some valuable input of some kind...c'mon, y'all...

Maybe that damn code generation computer that Mackie is lamenting about really IS the Fly in the Ointment... I'd personally like to know, that's for damn sure...

I'm not done yet, this nonsense has just really pissed me off more than it's done anything else...

Resistance is NOT Futile... stay tuned...

\m/ \m/

[anyhorizon]

Do you think if that IC were removed the brain board might end up being generic?

Peter