[Pw_forum] Serious PW91 bug (Was: Bug in fhi2upf.x?)

Paolo Giannozzi giannozz at democritos.it
Wed Feb 15 10:27:20 CET 2012

On Tue, 2012-02-14 at 18:29 +0100, Henning Glawe wrote:

> fhi2upf.x has a hard-coded special case for this, and writes 'PW' in the
> resulting UPF files.

fixed. Ok, this is a problem that occurred only in an exotic case,
and was quite easy to spot ...

>     ! special case : PW91 = GGX + GGC
>     else if ('PW91'.EQ. TRIM(dftout) ) then
>        call set_dft_value (iexch,1) !Default   HG: slater exchange
>        call set_dft_value (icorr,1) !Default 

... while this is WRONG, affects all PW91 calculations since v.4.3,
and may easily go unnoticed. The correct line is

       call set_dft_value (icorr,4)

Thank you for reporting this. 

Paolo Giannozzi, IOM-Democritos and University of Udine, Italy

More information about the users mailing list