From our company's experience of sha256 signing on atmel Sam7 chips there are as many variants of padding and signing as bootloaders. It must be a multiple of 32 bytes. There is a field for pad bits added at the end. And the actual hash signature isn't part of the data to hash. But as I found out from adventures in squashfs with Linux hacking, there's a period where it deviated from standard to get it working from patches. And they are not all the same. Not sure if this is what has happened here though. Does anyone have nor flash dumps of 5.1 and 4.9?
Hi Ed,
> he says that the signature is SHA256 and the key is the first two or
> three sentences of the King James Bible
In an old private email to me, he said he *thought* it was the `first
few chapters of the King James Bible'.
> the first 6 sentences make exactly 392 bytes:
392 B is 3,136 b and that's only slightly bigger than 3,072 that is
3 Ki. 3,072 b is 384 B, suggesting 8 B of that 392 B aren't the digest?
--
Cheers, Ralph.
_______________________________________________
e3-hacking mailing list
e3-hacking@earth.li
https://www.earth.li/mailman/listinfo/e3-hacking