[lkml]   [1996]   [Sep]   [14]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
Messages in this thread
SubjectRe: PPro compile
Robert L. Krawitz writes:
> Date: Thu, 12 Sep 1996 11:49:28 -0400 (EDT)
> From: "Andrew E. Mileski" <>
> > Can I ask this dumb question. If , as Intel claims, the PPro performs
> > better on 386 code then 486-Pentium code, why does the kernel compile
> > with m486 rather than m386?
> I've never seen such a claim.
> Intel has claimed that 32-bit unsegmented code runs faster though.
> One reason that 486-optimized code might be less efficient is that
> just about everything is aligned on a 16 byte boundary, so the code's
> bloated.

This business went around back when people were wondering whether to
compile with -m386 or -m486. -m486 basically does two things: it uses a
few different two-instruction sequences that are faster on the 486 than
on the 386 while executing identically on both, and it aligns branch
targets and subroutine entry points on 16-byte boundaries (not "just
about everything") where cache lines start. -m486 code is therefore up
to about 5% larger and 5% faster.

 \ /
  Last update: 2005-03-22 13:38    [W:0.053 / U:26.672 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site