Problem with periodic boundry conditions when using ILP potentials

Dear all,

I would like to report an issue which I faced now. I was using ILP potentials with other force fields. I haven’t had any problem with it since I updated my LAMMPS version 23 Jun 2023 to the last one (2 Aug 2023). When the molecule inside the box intersects with box borders, program will stop calculation without printing any error message. I tried many times to find the origin of this problem, but there is no any comment about that. Reinstallling didn’t help. Then I again installed “my old version”, then it works. I think this is a bug related to the last version.

Best Regards,

The only change in the entire INTERLAYER potentials folder between 23 Jun 2023 and 2 Aug 2023 is this one:

Author: Axel Kohlmeyer <[email protected]>
  Date:   Tue Jul 11 10:31:16 2023 -0400
      use blanks on both sides of comparison operators not only one, remove bla>
  diff --git a/src/INTERLAYER/pair_ilp_tmd.cpp b/src/INTERLAYER/pair_ilp_tmd.cpp
  index fd3a2f8c6f..8b08de39c0 100644
  --- a/src/INTERLAYER/pair_ilp_tmd.cpp
  +++ b/src/INTERLAYER/pair_ilp_tmd.cpp
  @@ -481,11 +481,11 @@ void PairILPTMD::calc_normal()
             jH2 = atom->map(tag[i] - 2);
             iH1 = map[type[jH1]];
             iH2 = map[type[jH2]];
  -          if (strcmp(elements[iH1], "Ow") == 0 ) {
  +          if (strcmp(elements[iH1], "Ow") == 0) {
               vect[0][0] = x[jH1][0] - xtp;
               vect[0][1] = x[jH1][1] - ytp;
               vect[0][2] = x[jH1][2] - ztp;

There is next to nothing that can be done about this report since there are no specifics, no input examples, no way to reproduce what you are seeing. and There should always be output indicating what the issue is, but it may not become visible when running in parallel due to buffering. This can be turned off with the “-nb” command line flag. It should not happen, if you run in serial.

The first step of debugging would be to construct the simplest and smallest input example that can reproduce the situation. The more you can take out, the easier it will be to track down problems.