[Libre-soc-isa] [Bug 817] Big Integer Math (sv.adde, sv.subfe, sv.madded, 128 by 64-bit -> 64-bit div/rem, maybe more...)

bugzilla-daemon at libre-soc.org bugzilla-daemon at libre-soc.org
Thu Apr 21 21:16:31 BST 2022


--- Comment #4 from Luke Kenneth Casson Leighton <lkcl at lkcl.net> ---
(In reply to Jacob Lifshay from comment #3)

> > yeah, as stated on irc, sorry i haven't yet fixed it.

no problem. haven't been upstairs yet to acknowledge, sorry.

> except that i only came up with madded, msubed is something that you
> apparently interpolated out of msubx and madded 

it seemed logical at the time! but if not needed that's great, one less
instruction and it entirely frees up the right half of EXT04 space.

    maddld maddhd MADDED maddhdu -- -- -- --

> -- msubed isn't needed for
> the div algorithm since the required negation is already done by subfe.

nggggh ok makes sense.

mul... mul... arg yes of course, RC is used as the 64-bit carry on the
*product*, nothing to do with the partial result, yes, that goes into

so of course no mul-subtract.

rright, some hatchet-jobs and writeup to do...

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

More information about the Libre-SOC-ISA mailing list