ABINIT Discussion Forums

Yes, nsym I meant. It is a 3 layer sqrt(3)xsqrt(3) slab, coordinates are from previous calculations on crystal (the program). It was written so, because ready for geometry relaxation and an adsorption study. The symmetry and multiplicity are recognized correctly, imo --------------------------------...

Putting nsim 1 eliminates the density < -0.05 warnings and gives reasonable results again in version 7.0.5, and equal to results from versions before 5.4.4p. The same does displacing an atom by a physically irrelevant quantity to break symmetry. So compilation options and libraries had nothing to do...

When I compile abinit-7.0.5 with debug>enhanced I get instead Program received signal SIGSEGV: Segmentation fault - invalid memory reference. Backtrace for this error: #0 0x7F122AB0C667 #1 0x7F122AB0CC34 #2 0x7F1229EB04EF #3 0x7F122CF1573F #4 0x7F122CF15708 #5 0x7F122CA7437C #6 0x7F122D16C7CF #7 0x4...

Hallo, a comparison of abinit-7.0.5 and abinit packaged by debian 4.3 evidenced different results. So I ran the tests and found that many of them did not pass. I found the same on abinit-6.12.3 and 4.8.5 and so began varying the compilation options: gfortran4.7 gfortran4.4, with and without openmpi ...