On Wed, Feb 23, 2011 at 02:29:07PM +0100, Paolo Giannozzi wrote: > On Feb 23, 2011, at 10:25 , Henning Glawe wrote: > > > - my statement 'Affected quantum espresso versions: >=4.2' was > > referring to > > released versions, so the problem may be fixed in CVS (I saw a > > 4.3a version > > string in your case) > > it looks like there is a problem with v.4.2.x that has disappeared in > the > soon-to-be-released 4.3 version. The funny thing is that very little has > changed in the affected part of the code
Another funny thing about the 'wrong' results is: they depend on the number of MPI tasks; could the error be related to read-access to uninitialized memory? in any case, I'll give current CVS a try. Thank you for helping! -- Mit freundlichen Gr??en Henning Glawe Max-Planck-Institut f?r Mikrostrukturphysik Weinberg 2, 06120 Halle (Saale), Germany http://www.mpi-halle.de/~theory Phone: +49-345-5582-613 Fax: +49-345-5511223 Email: glawe at mpi-halle.de
