[PATCH] Fix detection of odd memory configurations on sunxi

[Rudi Horn]

Hi Jernej,

thanks for the comments on the patch.

 > I don't see a benefit for public function here, since it's unlikely 
to be > used anywhere else. Just mark it static. Compiler can optimize 
it better.

My motivation for this was that the implementation relies on subtracting
`sizeof(u32)` from the address. This is an implementation detail that relies
on the internals of `mctl_mem_matches_base`, and so I wanted to place
it in the same file as `mctl_mem_matches_base`, preventing it from being
static. It will only be called twice, so I don't think optimization is 
too much
of an issue. However I don't feel strongly on this if you would still prefer
it in the same file.

Once this is resolved, I will submit another patch with all of the other
comments addressed.

Thanks,
Rudi