[libre-riscv-dev] KCP53010 Updates

[Luke Kenneth Casson Leighton]

---
crowd-funded eco-conscious hardware: https://www.crowdsupply.com/eoma68

On Sun, Aug 4, 2019 at 5:03 AM Samuel Falvo II <sam.falvo at gmail.com> wrote:
>
> I've decided not to abandon the processor.

 yay!

> Instead, I've decided to restart it from scratch, but this time with a
> monolithic IXU (Integer Execution Unit), built 6502- or Z80-style.  It's
> occurred to me that it's not only easier to get working up front, but if my
> hunch is right, all the explicit X0..Xn and F0..Fn states (which you can
> see in the KCP53000 sources) correspond *almost* one to one to the various
> GO_xyz signals in a scoreboard-driven design.  IOW, what I'm trying to say
> is, it'll be easier to adapt an existing IXU to a scoreboard design later,
> than it will be to build a new one from scratch now.  *IF* my hunch is
> right.

if it's kept simple enough, yes.

> I *have*, however, decided to abandon TileLink.

yehh, i did say: there's just not enough adoption, it makes your life
that much harder.  i'm kinda used to this sort of time/effort-based
decision-making process: it takes other people a lot of pain / time to
learn.  ho hum :)


> Not that I dislike it, but Wishbone has a few advantages over it which I
> think would reduce the burden of building out the processor logic going
> forward.

eeexactly - you don't need the hassle, there's more examples out
there, more things that you can connect to and overall a lot less
work.

l.