From: "Freddie Schwenke" <freddie DOT schwenke2 AT sanlam DOT co DOT za>
Newsgroups: comp.os.msdos.djgpp
Subject: Re: memcpy() vs memmove()
Date: Wed, 9 Dec 1998 14:54:51 +0200
Organization: An Internet Solution Customer
Lines: 47
Message-ID: <74lrv4$k9i$1@hermes.is.co.za>
References: <74ll3i$h7k$1 AT hermes DOT is DOT co DOT za> <74lqta$l3o$1 AT antares DOT lu DOT erisoft DOT se>
NNTP-Posting-Host: gauntlet-cpt.sanlam.co.za
X-Newsreader: Microsoft Outlook Express 4.72.2106.4
X-MimeOLE: Produced By Microsoft MimeOLE V4.72.2106.4
To: djgpp AT delorie DOT com
DJ-Gateway: from newsgroup comp.os.msdos.djgpp
Reply-To: djgpp AT delorie DOT com

I'm quite aware that there is a differrence between the two functions.  I
actually want to abuse the memcpy() function to replicate one ore more bytes
throughout a string.  The problem is that memcpy() only works while the
length is less that 16.  WHY???

Is this a C thing or a DJGPP thing?  In MS Visual C++ apparently there is no
difference betwee the two functions!!!

Freddie
:)
Martin Stromberg wrote in message <74lqta$l3o$1 AT antares DOT lu DOT erisoft DOT se>...
>Freddie Schwenke (freddie DOT schwenke2 AT sanlam DOT co DOT za) wrote:
>: I'm using DJGPP to learn C.  I also want to use this tool to train people
in
>: C.  I came accross a problem when I tried the memcpy() and memmove()
>: functions tho'.
>
>: memmove() seems to do what it is supposed to do.  I have no problem with
>: that.
>
>: memcpy() gives some weird results tho'.
>
>:   char
>:     buffer[11] = "abcdefghij";
>
>:   memcpy(buffer + 1, buffer, 9);
>
>: Should this not result in buffer = "aaaaaaaaaa"???
>
>: Well it does...IF the string contains a maximum of 15 characters
(excluding
>: the null character).  The moment the length exceeds 15 characters I get
>: weird results.
>
>: Any suggestions or comments will be appreciated.
>
>Well, learn to read the man page. memmove's page says that overlapping
>regions will be copied correctly. There are no such statement for memcpy
>although it says the function works as efficiently as possible (i. e.
>indeterminate results if overlapping regions.
>
>
>Right,
>
> MartinS