[sdiy] LFSR using lookup tables

[John Ames]

On 2/20/20, Ben Stuyts <ben at stuyts.nl> wrote:
> So lfsr4() using the lookup table gave a 4 times improvement over lfsr1().
> But interestingly, the Galois LFSRs lfsr2l and lfsr2r (for left and right
> shifting versions) were even faster. This makes sense if you look at the
> source code, it is very simple.
Yeah, absolutely. The non-Galois method is probably more
"correct"-looking to crypto math types, but there's absolutely no
reason to break up a very simple operation into a pile of separate
operations - just unnecessary duplication of effort on the part of
both the programmer and the CPU.