[Libre-soc-dev] NGI POINTER gigabit ethernet router ASIC roadmap

[lkcl]

On Tue, Nov 2, 2021 at 2:16 PM lkcl <luke.leighton at gmail.com> wrote:
>
> On Mon, Nov 1, 2021 at 11:11 AM lkcl <luke.leighton at gmail.com> wrote:
>
> > * there is not enough time to do out-of-order (it was 5 months just
> > doing the 6600 experiment), a FSM is too slow therefore we do an
> > in-order core making sure to plan ahead for OoO (use ReservationStations,
> > use MultiCompUnits)
>
> deep breath: i tested the use of the ReservationStations class
> (developed over 2 years ago), and it is the first time it is "hit"
> by full-on ready/valid Signalling: it didn't work.

it didn't work because of an oversight in the design.

* Reservation Stations need to be latched as actually reserved
   until such time as the same MUX id is seen at the output
   (final stage of the single pipeline).

* this means raising a "busy" flag - *per input* (per RS)
   indefinitely until such time as the MUX id pops out the
   end

* the first design relied on state information known by the
  *sender* to respect the MUX ids.

* thus, what i had written is more of a multi-in multi-out arbitrator / routing
  system, not a ReservationStation, at all.

drat.

this will need a redesign.

l.