Re: [PATCH] libm/complex: add clog10, clog10f

[Craig Howland <howland at LGSInnovations dot com>]

On 02/22/2016 12:09 PM, Corinna Vinschen wrote:
> On Feb 22 10:14, Yaakov Selkowitz wrote:
> > On 2016-02-22 03:21, Corinna Vinschen wrote:
> > > Did you write these functions by yourself or did you fetch them from
> > > some upstream?  If so, where from?
> > They are tweaks of our existing clog() and clogf().
> >
> > > My complex math is next to inexistent, but I took a look into the glibc
> > > man page.  The man page says "clog10(z) is equivalent to clog(z)/log(10)".
> > > The above is equivalent, I take it?
> > https://sourceware.org/bugzilla/show_bug.cgi?id=19673
> Erm... sorry but I don't see any mention of atan2 in this BZ.
> When I said "next to inexistent", I mean it :}
>
> Also, what about Craig's point?
>
>
> Corinna
Comparing to the associated Wikipedia article, the atan2 part of the equation is 
correct.  I also agree with how it is converted from base-e to base-10 with 
respect to the correct method reported in the referenced bug 19673.  (So with 
the minor tweak of the cast in the float version to avoid any double 
calculations, all appears fine.)
Craig