[Libre-soc-bugs] [Bug 734] add Partitioned SimdSignal support for elwidth-based layouts (currently ElwidPartType)

[bugzilla-daemon at libre-soc.org]

https://bugs.libre-soc.org/show_bug.cgi?id=734

--- Comment #15 from Luke Kenneth Casson Leighton <lkcl at lkcl.net> ---
done:

commit 6edd0e1e0732b3e9781d3c4d8d3830d475cd4a50
Author: Luke Kenneth Casson Leighton <lkcl at lkcl.net>
Date:   Mon Oct 25 14:35:57 2021 +0100

    use explicit SimdShape for minitest example rather than
    fixed_width parameter

+            shape = SimdShape(s, fixed_width=width)
+            shape2 = SimdShape(s, fixed_width=width*2)
+            shape3 = SimdShape(s, fixed_width=width*3)
+            self.a = s.Signal(shape)
+            self.b = s.Signal(shape2) # TODO: shape*2
+            self.o = s.Signal(shape3) # TODO: shape*3

TODO: the multiply part.  basically, SimdShape should (multiply-inherit)
derive from SimdMap, *or* (probably better) the functions *of* SimdMap
go into SimdShape.

jacob you had probably best do that, but *please do not go overboard*.

also please do not make SimdShape critically depend on any types.
do not alter it so that it critically depends on the enum Elwid.

if necessary start again with the absolute basics, ignoring SimdMap
entirely and *only* add in the *absolute* minimum rudimentary
functions needed, this time.

remember: write code *on-demand* - NOT "i think this might be needed
so am going to write 1,000 lines of code that has no unit tests
and no feet-on-the-ground justification for its existence"

show the need *first*, create code *second*, ok?

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