[AMBER] Segmentation fault occurred when using Amber with Plumed

From: Jiahua Deng via AMBER <amber.ambermd.org>
Date: Mon, 9 Jun 2025 03:44:35 +0000

Dear Amber Users,

I have compiled Amber 24 and Plumed 2.9.3 separately and wanted to run Amber with Plumed for metadynamics simulations.

To test if Amber works well with Plumed, I ran a classical MD without any biases/restraints but only as few distance CVs are defined and printed out by providing the “plumed_test.dat” file. In my md.in file, I also added the line “plumed=1, plumedfile='plumed_test.dat’”. In the job submission script, I exported the plumed kernel. However, the job failed with the following error message: Program received signal SIGSEGV: Segmentation fault - invalid memory reference.

I understand this error message is quite vague and can be triggered by a lot of reasons. The Amber output didn’t provide more specific error messages, but I noticed two lines “PLUMED is on” and “PLUMEDfile is plumed_test.dat” so I assumed plumed was detected and turned on by Amber.

I learned from the mailing list that after Amber 20, Plumed support is natively distributed in Amber so I assume I don’t have to patch plumed with Amber myself. Does anyone know what might be the real problem here or if I missed something before Amber can work with Plumed? Thanks for your help!

Best,
Jiahua

_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Sun Jun 08 2025 - 21:00:02 PDT
Custom Search