[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: feff8.2 rmax vs ratx check
Dear Peter,
Thank you very much for the bug report.
That WARNING message should not appear. Interestingly nobody complained
about that before, while it should always appear. Fortunately, this bug
does not affect the final results of calculations. But for sure should be
annoying, since it is wrong message.
The 'trap' can be renamed, or should be declared as an external
in all subroutines where it is called.
I will fix both bugs in the update of feff8.2.
Yours sincerely,
Alexei Ankudinov
On Mon, 6 Oct 2003, Peter Pfalzer wrote:
> Hi all,
>
> I upgraded from feff800 to feff82 and found some strange behaviour (with
> the precompiled DOS version as well as with the version I generated
> myself from the downloaded source code on an AIX system):
> During my EXAFS calculations I always got the warning message
>
> WARNING: rmax > distance to most distant atom.
> Some paths may be missing.
> rmax, ratx 1.00000E+01 0.00000E+00
>
> although the most distant atoms in the input file were at 15A.
> I compared the source code of both feff versions and found the line
>
> if (rtmp .gt. ratx) ratx = rtmp
>
> (subroutine paths, line 18933 in feff800.f) but not in feff82.f. After
> inserting this line after line 19032 into the respective feff82.f
> subroutine, the behaviour of the warning message appears correct.
>
> As I don't know fortran well and since I'm not familiar with the
> detailed structure of feff, I'd like to know whether my "bugfix" is
> complete or if the omitted setting of ratx affects other parts of the
> program as well.
>
> Thanks,
> Peter
>
> PS: Somewhere I read that reports on portability problems are welcome.
> Here is a minor one: To compile feff on our AIX 4.3 machine I had to
> rename the subroutine "trap". Otherwise I got an error message
> complaining about the wrong number of parameters for the INTRINSIC
> subroutine trap.
>