Re: A89: AU 2.00 "bug" explanations
[Prev][Next][Index][Thread]
Re: A89: AU 2.00 "bug" explanations
Do you know how the "bug" works?
It isn't like they have hidden something on purpouse.. it is quite an
embarrasing little misstake.
the security is really tight. the lock is in hardware, and you have to make some
reads from and then a write to a certain memoryarea, this executed from a
certain memoryarea (that is writeprotected), then the U8 chip (that is connected
to addressbus) "unlocks" the flash, a bit like a codelock..
the way around this, is that you execute an errorstring that lies in this area,
and just happend to form theese 3 reads from the right area, and then a write to
a selectable area, and then generates an "illiegal instruction"
Very easy not to think about this..
so it isn't a bug in the _software_ and not a on purpuse hidden function, but a
badly formulated errorstring in a bad place.. :)
shit happens...
I too think that ti should be informed, and do something about it.. But I will
not replace/upgrade my calc.. fun to be able to completly change the os :) (with
my own software then ofcourse, wouldn't trust anyone else for that :)
//Olle
Nate Mueller wrote:
> Security based on secrecy is no security at all. It would take less then 5
> minutes for someone to run AU through a decompiler and get the complete
> source code (decompiling works much better in asm then in C). The inner
> workings of your program are very easy to access and if that's all someone
> needs to create a virus, then the issue should be reported to TI, along
> with the example code. This is standard pratice in the computer industry
> and it's that way for a reason.
>
> --Nate
References: