g_chi bus error with -p option
The use of the -p option with g_chi (both version 4.0.7 and git version as of 6/9/2010) exits with a Bus error, preventing a complete chi.log file from being written (although order.xvg) is written. The problem is reproducible with both .gro and .xtc files. I can upload a test case if necessary, but everything I've thrown at g_chi produces the error faithfully.
#1 Updated by Justin Lemkul almost 10 years ago
I should also add that this bug is reproducible on a Linux machine (AMD64) running Gromacs-4.0.7. I took a quick look into the code and it seems as though the -p flag is never used, and thus no attempt is ever made to open a file or write anything to it. I could be missing something, though.
#4 Updated by Justin Lemkul almost 10 years ago
The attachment contains a .gro and .tpr file necessary to reproduce the problem. With the command:
g_chi -s md_0_10.tpr -f md_40_50.tpr -p
the program exits with the following:
94 residues with dihedrals found
453 dihedrals found
Reading frames from gro file 'DEP domain of Dvl2 with 4:1 DHPA in water', 15359 atoms.
Last frame 0 time 0.000
j after resetting (nr. active dihedrals) = 281
WARNING: not all dihedrals found in topology (only 281 out of 453)!
Now calculating transitions...
Total number of transitions: 0
I get an incomplete chi.log file, and an intact order.xvg, but no order.pdb. The error is a segmentation fault on my AMD64 machine, but I don't know if that's significant at all.
#5 Updated by Justin Lemkul almost 10 years ago
I get an incomplete chi.log file, and an intact order.xvg, but no order.pdb.
The error is a segmentation fault on my AMD64 machine, but I don't know if
that's significant at all.
I can reproduce the error with an .xtc trajectory as well, in case that's an issue. The latest git code I tried was from early yesterday (6/9/10), so if I should try a newer version, please let me know.