[libre-riscv-dev] [Bug 329] New: coriolis2 experiment layout for Dependency Matrices

[bugzilla-daemon at libre-soc.org]

https://bugs.libre-soc.org/show_bug.cgi?id=329

            Bug ID: 329
           Summary: coriolis2 experiment layout for Dependency Matrices
           Product: Libre-SOC's first SoC
           Version: unspecified
          Hardware: PC
                OS: Linux
            Status: CONFIRMED
          Severity: enhancement
          Priority: ---
         Component: Hardware Layout
          Assignee: lkcl at lkcl.net
          Reporter: lkcl at lkcl.net
                CC: libre-riscv-dev at lists.libre-riscv.org
   NLnet milestone: ---

Created attachment 57
  --> https://bugs.libre-soc.org/attachment.cgi?id=57&action=edit
quick preliminary test ilang from soc/scoreboard/test_mem_fu_matrix.py

we need a quick example to demonstrate how the Dependency Matrices would
look so that coriolis2 can do the layout in a regular fashion.  this
even though they may convert from DFF to SRNAND, so the layout needs to
be dynamic, rather than explicit naming of cells.

also, some of the cells may be augmented by additional "setting" logic
that needs to go into every row (not yet written), so any explicit
manually-created layout that hunts for explicit cell names is going to
have to be redone, and not only redone, duplicated manual work several
times.

this is why the idea exists to identify sub-groups of cells by way of
netlist path analysis, and have them placed by Etesian in regular grid
fashion.

-- 
You are receiving this mail because:
You are on the CC list for the bug.