Convert the funcs to do the conversion inline so that we can do the copy
all at once with memcpy.  This let's us push out an weird arch-specific
issue with accessing different regions of memory to the memcpy function
like the MMRs on Blackfin systems, and it should be a bit faster.

Signed-off-by: Robin Getz <robin.getz@analog.com>
Signed-off-by: Mike Frysinger <vapier@gentoo.org>