??? 08/14/07 10:34 Read: times Msg Score: +1 +1 Good Answer/Helpful |
#143248 - Instruction decoding Responding to: ???'s previous message |
As has been previously suggested, look at the instruction set carefully! You can divide the instructions up into groups that share similar encoding as I demonstrated in my last post for two instructions. With this information, you can devise some logic that given the first opcode will tell you whether to fetch another byte, another byte after that or simply execute the instruction that you just fetched. There's no magic, just sequential logic.
There is one fundamental assumption - that the stream of instructions actually make sense to the processor, otherwise it's like a train without a track - it will crash. Please don't expect us to explain the fundamentals of processor design on a forum - I learnt it from a book as did most (if not all) of us. Your question is like asking us how to design a nuclear bomb - if you don't understand the fundamentals, you've got no hope as each question you ask will be followed by another and another.... |
Topic | Author | Date |
decoding | 01/01/70 00:00 | |
It's implicit | 01/01/70 00:00 | |
design | 01/01/70 00:00 | |
A little ambitious? | 01/01/70 00:00 | |
First stop - instruction set | 01/01/70 00:00 | |
no intelligence | 01/01/70 00:00 | |
Back to basics! | 01/01/70 00:00 | |
Instruction decoding | 01/01/70 00:00 | |
decoder unit | 01/01/70 00:00 | |
State Machine | 01/01/70 00:00 | |
more detail | 01/01/70 00:00 | |
IF you've learned the "basics" of FPGA design ... | 01/01/70 00:00 | |
Decoding | 01/01/70 00:00 | |
in a traditionally anthropomorphic metaphor? | 01/01/70 00:00 | |
My favourite anthropomorphic metaphor... | 01/01/70 00:00 | |
My favorite ... | 01/01/70 00:00 | |
Blither | 01/01/70 00:00 | |
re: blither | 01/01/70 00:00 | |
I always think that a blithering idiot.. | 01/01/70 00:00 | |
anthropomorphic metaphor | 01/01/70 00:00 | |
The same way ... | 01/01/70 00:00 | |
Dictionary | 01/01/70 00:00 | |
If anyone is still listening...![]() | 01/01/70 00:00 |