Subject: Re: copystr
To: David Laight <>
From: Nathan J. Williams <>
List: tech-kern
Date: 09/03/2002 12:08:02
David Laight <> writes:

> FWIW memmove() check for overlapped copies and memcpy()
> should not (and would be expected to copy 'forwards').

The overlap bit is fine, but there's absolutely no reason to constrain
memcpy() to work in the forwards direction. It should be fast, and if
that's forwards, then fine, but it should also be allowed to copy
backwards, sideways, and split down the middle based on last year's
Super Bowl point spread. If your code doesn't cope with that, it's
your bug, not memcpy's.

        - Nathan