CVS update: samba/source

[Andrew Tridgell]

> Do we want to do something like the __XXX_DONT_USE_STRCPY() macros
> for bzero to stop it creeping back in?

no.

I really don't see what the fuss is all about. We already have:

#if !defined(HAVE_BZERO) && defined(HAVE_MEMSET)
#define bzero(a,b) memset((a),'\0',(b))
#endif

which means bzero is fine. bzero() also has the big advantage that
programmers don't screw up the order of the arguments. That happens
all the time with memset.