Possible back door in registers for passing data?
mklrobo
Posts: 420
Hello!
I would like to pose a question, where permissible.
I was playing with assembly with the propeller, and have run across some
registers that do not use some upper bits of their register(reserved for future use).
that is not used in regular programming.
Would it be possible to pass information in the upper registers, and then
retrieve them when shifting right? If so, things could get complex in
a dual way.(no other part of regular programming would corrupt the data)
If this phenomena is in every processor, could hackers
exploit this to pass malware? (in a very small way, of course.)
Just a thought.
I would like to pose a question, where permissible.
I was playing with assembly with the propeller, and have run across some
registers that do not use some upper bits of their register(reserved for future use).
that is not used in regular programming.
Would it be possible to pass information in the upper registers, and then
retrieve them when shifting right? If so, things could get complex in
a dual way.(no other part of regular programming would corrupt the data)
If this phenomena is in every processor, could hackers
exploit this to pass malware? (in a very small way, of course.)
Just a thought.
Comments
Kind of hard to later read it, maybe Test instruction as mask vs data is interchangeable on this instruction.
Not worth it if you are wasting a extra long in the code to handle it that could be used for data further down and have room for 32 bit flags.
I was thinking of using that unused space for storing, to cut down
on clock cycles for RWLONG commands.(8.23 clock cycles)
I know that may sound nit - picking, but hey, a cycle saved is
a cycle earned! (Ha!)