I've successfully built and compiled an application on my PC which will generate an RSA keypair (I used GPK for the long arithmetic). I then tested encryption/decryption on a string "Hello World" (ASCII --> Hex transform first).
The question I have is, is is mathematically possible to only decrypt "Hello" from my cipher-text string "Hello World", or would I have to encrypt "Hello" first, then encrypt "World", and then decrypt them separately?
EDIT: My desired use-case:
I have a chip (Zynq SoC from Xilinx), which only has 256K of on chip memory, and 512K of L2 cache.
I can prefill and lock the cache with some basic instructions/data, and then use the 256K for my stack, heap, and .data/.bss sections of code (variables).
I need more space, so I was thinking I could use the 256K as "swap" space, and fetch the rest of the .text (the instruction code) from the external, encrypted flash, but I need to decrypt it on-chip to maintain the security.
I was wondering if I could encrypt the entire image, then "fetch, decrypt, decode, execute". It looks like I might be restricted to encrypting my image in 256K chunks, and decrypting it in that size of chunks as well.