[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [f-cpu] Relative branch



> huh ... because in FC0, the IP is "virtual".
> there is no such register implemented in HW.

ehh .. I didn't realized it yet :)

> yup but jumping across pages are not possible.
> And that behaviour would change depending on the page size

yes. It was meand as short (near) jump only with
regular jump for other cases.

> >This would help small conditional branches, save
> >code size and register usage ...
>[...]
> the goal is to spread the jump code, because it
> is simply impossible to jump in a single cycle.

hmm seems interesting.

> Imagine that you have a cjump instruction with a new
> address' LSB, you have to
>  - fetch the address (several cycles)

I was thinking about near jumps where the code would
be probably in cache already prefetched.
Or by looking forward in instruction stream ? :-] Joking..

But yes, you are right at the moment. I hope to find time
to finish both emulator and gcc, measure some cases and
then it is the correct time to say: hey this or that instruction
would improve speed by 20%, let's implement it ...

regards, devik

*************************************************************
To unsubscribe, send an e-mail to majordomo@seul.org with
unsubscribe f-cpu       in the body. http://f-cpu.seul.org/