assembly - Why are branch instructions faster than jump instructions? -
looking @ amtel avr instruction set, noticed branch instruction takes 2 cycles if executed , 1 if not, while jump instruction takes 3 cycles.
why this?
it seems comes op-code length. jmp instruction 32bit instruction because has read address separate word. reading costs additional clock cycle.
branches , relative jump rjmp , indirect jump ijmp 16bit instructions executed in 2 clock cycles.
assembly avr
No comments:
Post a Comment