Date: Thu, 23 Mar 1995 23:20:40 +0900
From: raraki AT human DOT waseda DOT ac DOT jp (Ryuichiro Araki)
To: DJGPP AT sun DOT soe DOT clarkson DOT edu
Subject: Re: A quick way to copy n bytes

>>>>> Mat Hostetter <mat AT ardi DOT com> writes:

> 1) The added overhead of seeing if the size is < 16 bytes and then
>    doing movsb instead is negligible.  When that happens you can punt
>    the (also neglible) overhead associated with long moves and cleanup.
> 2) Many small moves will be constant sized (e.g. sizeof (struct foo)),
>    and will get inlined into gcc.

I've gotten the snapshot of V2 alpha as of 950320 from DJ's server, and
looked into src/libc/ansi/string/ and src/libc/pc_hw/mem/.  I've done some
simple tests to see the degree of the overhead on my 486- and Pentium-based
machines, then realized that you're correct.

I also appreciate your lucid comments on the cache line issue.

Regards,

   ---
   raraki(Ryuichiro Araki)
   raraki AT human DOT waseda DOT ac DOT jp

p.s. linux/asm/string.h impressed me with Linus's smart implementation as well.