Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]


Groups > comp.arch > #61311 > unrolled thread

Re: Bit swizzling

Started byHans-Peter Diettrich <DrDiettrich1@netscape.net>
First post2020-09-05 19:38 +0200
Last post2020-09-05 19:38 +0200
Articles 1 — 1 participant

Back to article view | Back to comp.arch

This discussion starts older than the indexed window; earlier articles aren't shown. The article labeled Started by below is the oldest one visible, not the original post.


Contents

  Re: Bit swizzling Hans-Peter Diettrich <DrDiettrich1@netscape.net> - 2020-09-05 19:38 +0200

#61311 — Re: Bit swizzling

FromHans-Peter Diettrich <DrDiettrich1@netscape.net>
Date2020-09-05 19:38 +0200
SubjectRe: Bit swizzling
Message-ID<20-09-015@comp.compilers>
Am 05.09.2020 um 18:05 schrieb Rick C. Hodgin:
> Are there any algorithms which take a known-at-compile-time sequence
> of bitwise operations on an 8-bit to 64-bit quantity, and optimize
> them down to their minimal set of operations?
>
> For example, if I have an 8-bit byte and I want to swizzle the bits
> thusly:
>
>      Input:   07 06 05 04 03 02 01 00
>     Output:   05 04 07 02 01 03 00 06

In your 8-bit case an array of outputs is sufficient, indexed by the input.


> In addition, given the bit operator abilities that exist on various
> CPUs there are potentially other combinations that exist behind an
> operation, such as bitswap, where the order of bits flips or mirrors
> across the center position.

Somebody must build the tables for those operations, distinct for each
CPU, then write code to make use of these tables. I don't think that it
has been done yet, except perhaps for the basic operations (AND, OR...)

> Are there any existing algorithms which examine the operations that
> must be conducted and then create an optimized / minimal sequence of
> mechanical steps to conduct it given a constrained set of features
> (such as those present on a given CPU)?

I often used Quine-McCluskey to minimize my logic circuits.

In most applications a toggled bit in the outputs invalidates all prior
computation optimization attempts, and everything has to be analyzed and
constructed again.

In general a hash code of the inputs could be used to index an array of
outputs. That code is insensitive to changes of output bits, only the
affected array element has to be uptdated then.

DoDi

[toc] | [standalone]


Back to top | Article view | comp.arch


csiph-web