Mshmet segmentation fault error

Hi,

When running mshmet I get a Segmentation fault following the Phase 1. I have compiled the olderVersionThatWorks and I am testing with the 3D example found in this post. The files are temp.mesh and temp.sol. Below is the complete output.

Any idea what could be wrong?

Thanks!

$ mshmet temp.mesh  -eps 0.00001 -hmin 0.01 -iso
  -- MSHMET, Release 2.3c (Aug. 25, 2009) 
     Copyright (c) LJLL, 2007-09
    2023-02-27 20:41:35

  -- INPUT DATA
  %% temp.mesh OPENED
  -- READING DATA FILE temp.mesh
  %% temp.sol OPENED
  -- READING DATA FILE temp.sol
     NUMBER OF GIVEN VERTICES       4139
     NUMBER OF GIVEN TETRAHEDRA    21547
     NUMBER OF GIVEN DATA           4139
  -- DATA READING COMPLETED.     0.04 sec.

  &&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&
   MODULE MSHMET-LJLL : 2.3c (Aug. 25, 2009)
  &&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&

  -- PHASE 1 : ANALYSIS
  ** SETTING ADJACENCIES

 Unexpected error:  Segmentation fault

  -- CPU REQUIREMENTS
  in/out       50.66 %      1. calls,      0.04 sec/call
  analysis     49.34 %      1. calls,      0.04 sec/call
  metric        0.00 %      0. calls,      -nan sec/call
  total       100.00 %      1. calls,      0.08 sec/call

   ELAPSED TIME  0.08 SEC.  (0.08)

Hi, I have the same problem. Did you solve?

Hi,

I think I have actually just found it. In file hash.c, subtract 1 to inival, so in the end replace lines 27 and 145 by inival = 2147483647;

Let me know if this works for you.

Marc

It may seem strange but I solved it by adding to lines 71 and 72 (in the for loop after /* set adjacency */), the lines:

 char str[] = "";
 fprintf(stdout, str);

I’ve only tried it for the 3D case and it seems to work fine.

Anyway, your solution might make more sense and I’ll try it in the next few days.