[Patch] fix some warnings in ldb

[Simo]

On Fri, 2013-08-16 at 21:41 -0700, Matthieu Patou wrote:
> On 08/16/2013 05:06 AM, Volker Lendecke wrote:
> > On Thu, Aug 15, 2013 at 11:26:37PM -0700, Matthieu Patou wrote:
> >> Please review & push.
> > Could the first one also be fixed with the attached patch? I
> > don't think the const is strictly required here. I'm taking
> > a look at this code for the first time how, so it might be
> > perfectly possible that I'm missing something. But I think
> > if we can avoid a discard_const, we should do so.
> >
> > Volker
> >
> I think it's ok but as there is more or less a dickat on not changing 
> existing API for ldb I was not keen on changing the signature even 
> thought it's only from const to non const, also I have to check that now 
> it didn't generate more warnings for callers as if we pass const 
> variable it will then generate a warning elsewhere.

Matthieu,
there is no dictat on internal private interfaces, just common sense.

Simo.