Some errors in using epw.x to calculate 2D materials

Post here questions linked with issue while running the EPW code

Moderator: stiwari

Post Reply
Rapunzel
Posts: 4
Joined: Tue Aug 25, 2020 8:01 am
Affiliation: 1050623944@qq.com

Some errors in using epw.x to calculate 2D materials

Post by Rapunzel »

Dear all,

I had some problems when I was running epw.x ,it always reports this error, as shown below. After querying the data, I know that it is a data storage group error, but I don't know how to solve it, and I don't know what kind of output file should be after epw. x runs. I post the error here.

*** glibc detected *** /home/../../EPW/bin/epw.x: free(): invalid pointer: 0x0000000007658b40 ***
*** glibc detected ***/home/../../EPW/bin/epw.x: free(): invalid pointer: 0x00000000048b3660 ***
======= Backtrace: =========
/lib64/libc.so.6[0x3dc5a75e5e]
/lib64/libc.so.6[0x3dc5a78cad]
/home/../../EPW/bin/epw.x(for_dealloc_allocatable+0x124)[0x107ae64]
/home/../../EPW/bin/epw.x[0x413ddd]
/home/../../EPW/bin/epw.x[0x413ddd]
/home/../../EPW/bin/epw.x[0x412329]
/home/../../EPW/bin/epw.x[0x412329]
/home/../../EPW/bin/epw.x[0x4115ae]
/home/../../EPW/bin/epw.x[0x4115ae]
/lib64/libc.so.6(__libc_start_main+0x100/lib64/libc.so.6(__libc_start_main+0x100)[0x/home/wangleixi/qe641-impi/q-e-qe-6.4.1/EPW/bin/epw.x[0x4114b9]
======= Memory map: ========
======= Memory map: ========
*** glibc detected *** /home/../../EPW/bin/epw.x: free(): invalid pointer: 0x0000000005078090 ***


My *.wout file shows " All done: wannier90 exiting" and all of the value of WF smaller than 4. But my epw.out had a error:APPLICATION TERMINATED WITH THE EXIT STRING: Aborted (signal 6)

If anyone could answer my questions, I would appreciate very much, since I have been confused for a long time.

Best wishes,
Rapunzel
Rapunzel
Posts: 4
Joined: Tue Aug 25, 2020 8:01 am
Affiliation: 1050623944@qq.com

Re: Some errors in using epw.x to calculate 2D materials

Post by Rapunzel »

This is the last small part of the document of my epw.out file:
We only need to compute 1600 q-points

Progression iq (fine) = 100/ 1600
Progression iq (fine) = 200/ 1600
Progression iq (fine) = 300/ 1600
Progression iq (fine) = 400/ 1600
Progression iq (fine) = 500/ 1600
Progression iq (fine) = 600/ 1600
Progression iq (fine) = 700/ 1600
Progression iq (fine) = 800/ 1600
Progression iq (fine) = 900/ 1600
Progression iq (fine) = 1000/ 1600
Progression iq (fine) = 1100/ 1600
Progression iq (fine) = 1200/ 1600
Progression iq (fine) = 1300/ 1600
Progression iq (fine) = 1400/ 1600
Progression iq (fine) = 1500/ 1600
Progression iq (fine) = 1600/ 1600
===================================================================
Memory usage: VmHWM = 254Mb
VmPeak = 668Mb
===================================================================

APPLICATION TERMINATED WITH THE EXIT STRING: Aborted (signal 6)

Thanks!
hpaudya1
Posts: 194
Joined: Tue Mar 21, 2017 7:11 pm
Affiliation:

Re: Some errors in using epw.x to calculate 2D materials

Post by hpaudya1 »

Hi Rapunzel,

It is not easy to tell what happened with so few details.

Looking your output, it looks like you do not have enough memory for the further calculation. I guess (because I do not know the input), for example if you are doing superconductivity, your el-ph matrix elements are written for 1600 q-points in 'prefix.ephmatXX' files, and now the code has crashed in reading and saving them in the memory.

I recommend,
1. increasing the number of processors, and see what happens
2. I see you are using qe6.4.1, please use the latest version of the code (qe6.6)

Please try to provide more input/output information (or files) so that it will be easy to understand what is going on.

Best,
Hari Paudyal
Post Reply