search

crest-lab / crest

CREST - A program for the automated exploration of low-energy molecular chemical space.
https://crest-lab.github.io/crest-docs/
GNU Lesser General Public License v3.0
183 stars 42 forks source link

SIGSEGV: Segmentation fault - invalid memory reference: possibly from GC algorithm #203

Closed kevmooreiii closed 5 months ago

kevmooreiii commented 1 year ago

I attempted to run conformer sampling for Tamiflu with CREST on default settings. Sometime prior to (or at the beginning of) the GC algorithm, I experience seg faults consistently, and on multiple machines. I have only experiences this issue on this molecule. I have run ~30 or so other structures without issue.

From other Issues (https://github.com/crest-lab/crest/issues/145), I tried turning off GC with -nocross. That does allow this calculation to complete. However, I am concerned that with the size of the molecule I am looking at, I would be missing conformers.

I attached the failed and successful run, inclusion input, submission script, and some output with the error. I am running CREST 2.12, installed via Conda.

Ensemble optimization with tight thresholds

Optimizing all 5624 structures from file "crest_rotamers_1.xyz" ... ... running RMSDs... done.
E lowest : -70.71822. 5184 structures remain within 6.00 kcal/mol window. Program received signal SIGSEGV: Segmentation fault - invalid memory reference.

Backtrace for this error:

0 0x1471f88df3ff in ???

1 0x558e39d8e561 in ???

..

tamiflu_failure.tar.gz

pprcht commented 1 year ago

Hi, This is still the same segfault as in the previous issues. If you absolutely have to use the GC part, as there hasn't been a conda update yet, try the 3.0dev branch and compile from source, that version should work. Otherwise -nocross is your best option. Or set a longer sampling time for the metadynamics, if you are concerned about incomplete ensembles.

github-actions[bot] commented 5 months ago

This issue had no activity for 6 months. It will be closed in 1 week unless there is some new activity.