[libre-riscv-dev] [Bug 178] first coriolis2 tutorial, workflow and "test project" page

[bugzilla-daemon at libre-riscv.org]

http://bugs.libre-riscv.org/show_bug.cgi?id=178

--- Comment #121 from Luke Kenneth Casson Leighton <lkcl at lkcl.net> ---
ok so i cut/paste the code which creates snx_core.ap and got it to
create a *sub* cell... add.ap.  size: 2000 x 800.

then do the same thing for sub.ap

all good so far.

then in the same code, i cut/paste the same section, except of course
creating a new cell "alu_hier" which loads alu_hier.vst, and
setting up the inputs and outputs, and also a new size: 3000 x 2000.

then, exact same thing with alu_hier.vst (Logical on addCell) and
we get this:

     + ./alu_hier.vst [models]
        + /home/lkcl/alliance/install/cells/sxlib/nmx2_x1.vbe [behavioral]
        + /home/lkcl/alliance/install/cells/sxlib/nmx2_x1.ap
        + /home/lkcl/alliance/install/cells/sxlib/sff1_x4.vbe [behavioral]
        + /home/lkcl/alliance/install/cells/sxlib/sff1_x4.ap
     + ./alu_hier.vst [structural]
  o  Creating ToolEngine<Etesian> for Cell <alu_hier>
     - Initial memory ..................................................
307.7Mb
  o  Configuration of ToolEngine<Etesian> for Cell <alu_hier>
     - Cell Gauge ......................................................
<sxlib>
     - Place Effort ..........................................................
2
     - Update Conf ...........................................................
2
     - Spreading Conf ........................................................
1
     - Routing driven ....................................................
false
     - Space Margin ........................................................
20%
     - Aspect Ratio .......................................................
100%
     - Bloat model ....................................................
disabled
  o  Converting <alu_hier> into Coloquinte.
     - H-pitch ..............................................................
5l
     - V-pitch ..............................................................
5l
     - Converting 6191 instances
Traceback (most recent call last):
  File "doAlu16.py", line 344, in <module>
    success      = alu_hier()
  File "doAlu16.py", line 320, in alu_hier
    etesian.place()
hurricane.HurricaneError: [ERROR] EtesianEngine::toColoquinte(): Non-leaf
instance "subckt_49_sub" of "sub" has an abutment box but is *not* placed.

so... errr... what's going on?  the whole point is: i want that subckt_49_sub
(which is the cell created previously, to be treated as "something to place".

how do we get both add.ap and sub.ap to be added to the list of cells
to be "placed"?

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