From mboxrd@z Thu Jan  1 00:00:00 1970
From: Michael Snyder <msnyder@cygnus.com>
To: Andrew Cagney <ac131313@cygnus.com>
Cc: gdb-patches@sources.redhat.com
Subject: Re: [RFA] Factor out some functions in remote.c
Date: Thu, 10 May 2001 12:08:00 -0000
Message-id: <3AFAE707.67003BF5@cygnus.com>
References: <3AF07FD4.62C23D84@cygnus.com> <3AFA1EC2.3020902@cygnus.com> <3AFAE1DD.917D4FBE@cygnus.com> <3AFAE4F0.6010209@cygnus.com>
X-SW-Source: 2001-05/msg00174.html

Andrew Cagney wrote:
> 
> > No, I need it.  I just haven't submitted the new code that needs it yet.
> > I think fixed length buffers and null terminated strings are both reasonable
> > input.  If you want me to code two separate functions, one for fixed length
> > buffers and one for null terminated strings, I will (but I'll be sad).
> 
> This relates bach to strncpy() vs strlcpy() and things like asprintf()
> vs sprintf().  If you've not heard of it strlcpy() takes a parameter
> that determines the size of the _destionation_ buffer and not the source
> buffer.  This dramatically reduced the chance of buffer overruns.
> 
> I would prefer that these conversion functions always took a parameter
> that determined the size of the destination buffer.  If nothing else,
> could you code call it as something like hex2bin(src,dest,strlen(src))?

Yes, OK, I'll make that change.