[Libre-soc-dev] daily kan-ban update 08sep2020

[Cole Poirier]

On Tue, Sep 8, 2020 at 12:33 PM <whygee at f-cpu.org> wrote:
>
> On 2020-09-08 17:21, Luke Kenneth Casson Leighton wrote:
> > therefore: rather than route 192+ wires across entire swathes of the
> > chip (to 12 separate pipelines), all you do is *PARTIALLY* decode the
> > instruction (enough to know which pipeline to deliver to), then route
> > the *raw* instruction (only 32-bit) over to the pipeline, and have it
> > *locally* decode *ONLY* the instruction fields that it actually needs.
> >
> > this does mean however that immediates get decoded more than once, for
> > example.  given that it's local, i'm not concerned about that.  it's
> > the number of wires crossing the chip that's got to come down.
> >
> > l.
>
> it seems like a good move to me because this would also decrease power
> and
> other things, through easier routing, lower capacitance load, ...
>
> CDC6600s were hand-wired in 3D but our chips are limited to 2D (2.5D
> with
> the metal layers). It made sense for a CDC6600 to have all the bits
> decoded
> and taken directly to the units, because transistors were expensive.
> Today, routing is a major bottleneck, particularly at that scale...

This may just be due to my noobieness, but reading this earlier this
morning my jaw was slightly agape and I was thinking "Wow that's so
elegant and so freaking cool!".

Congratulations on getting it working Luke!

Luke, I uploaded the CR-XER debug vhdl interface that I  did the new
version of yesterday
https://bugs.libre-soc.org/attachment.cgi?id=103&action=diff
(https://bugs.libre-soc.org/show_bug.cgi?id=484).

Yesterday:

Worked on icache.vhdl about 10% to go.. 15% if I include icache_tb.vhdl.

This morning:

Fixed the remaining issues in the libre-soc vulkan driver setup, there
were necessary files missing from the src/libre-soc/vulkan dir but
copying the files that meson errored on not finding from
src/amd/vulkan to src/libre-soc/vulkan allowed me to compile and
install successfully. Do we have a copy hosted on git.libre-soc.org,
or should I be making a pull request to vivek's branch on the
freedesktop mesa?

Updated the wiki documentation page for this, and updating the
associated script now, will push shortly.

This afternoon:

Finish icache.py first translation pass.

Review luke's comments on D-Cache if I have time.

Cole