[Libre-soc-bugs] [Bug 665] very basic nmigen-to-c compiler needed

[bugzilla-daemon at libre-soc.org]

https://bugs.libre-soc.org/show_bug.cgi?id=665

--- Comment #14 from Luke Kenneth Casson Leighton <lkcl at lkcl.net> ---
i'm looking at the output, and it's full of these:

    set(&slots[1280], next_1280);

slots is obviously a global array of all of the Signals(), which get
updated via repeated calls to run() functions, until such time as
no "slot" makes a change.  at that point, you know everything's "settled".

it will have to be passed in as the argument to "run()"

if keeping these "slots", can i make a recommendation to create a suite
of #defines for these indices?

#define {name_of_signal}_1280 1280

then

    set(&slots[{name_of_signal}_1280], next_{name_of_signal}_1280);

that way you can see the names, but in the c code it still uses integer
indexing?

unlike in python where it really really matters that slots[] is accessed
as fast as possible (so integers is a damn good idea), as this will be
compiled it's useful to see what the heck is going on by having actual
signal names.

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