To get rid of the underflow flag output (which is essentially a debugging output) compile with -ffpe-summary=none
See https://gcc.gnu.org/onlinedocs/gfortran/Debugging-Options.html ________________________________ From: users <users-boun...@lists.quantum-espresso.org> on behalf of tanghong...@outlook.com <tanghong...@gmail.com> Sent: Tuesday, 24 August 2021 2:18 PM To: users@lists.quantum-espresso.org <users@lists.quantum-espresso.org>; p.gianno...@gmail.com <p.gianno...@gmail.com> Subject: Re: [QE-users] QE6.8 and QE6.8 with Libxc.5.1.5 install and run errors Dear Paolo, Thank you for heping. I found the problems. Gcc4.8 is not good for compiling QE 6.8. It will meet some errors. 1 . I used gcc 10.2 to re-compiled mpi/openmp 2. I use gcc 10.2 to compiled libxc, 3. I use gcc 10.2 and intel mkl libs to compile QE6.8 (with libxc) Then QE6.8 looks good. However, when I use SCAN via “ input_dft = “ wth SCAN pseudopotential, it will have “ Note: The following floating-point exceptions are signalling: IEEE_UNDERFLOW_FLAG” in the end. But the calculations are correct. However, if I use PBE or TPSS (with PBE pseudopotential or tpss pps), it is pretty good, has no such “Note”. Is there any way I can go around this issue? Thank you very much. Best, Hong
_______________________________________________ Quantum ESPRESSO is supported by MaX (www.max-centre.eu) users mailing list users@lists.quantum-espresso.org https://lists.quantum-espresso.org/mailman/listinfo/users