[Wannier] Output abnormally

[庞瑞]

Dear all
I installed wannier90 and tried to run example29 in parallel model. I found the output is wired. The Delta Spread and RMS Gradient are all zero and the code run 40 steps finally output "convergence criteria satisfied". I found such behavior in other systems. So I think this might be a bug of wannier90. My system is Ubuntu 20.04, compiler is ifort 2019 + openmpi3.0, system lib, machine AMD 3990. I also tired to compile wannier90 with default mpi of intel, and found it could only run in serial model, the parallel model end when reading nnkp file without printing error message. If I compile using gfortran, with openmpi, the code can be run, but cannot do anything after print b_k direction. Is there any suggestion to correct this? 


> -----原始邮件-----
> 发件人: wannier-request at lists.quantum-espresso.org
> 发送时间: 2023-07-06 18:00:01 (星期四)
> 收件人: wannier at lists.quantum-espresso.org
> 抄送: 
> 主题: Wannier Digest, Vol 184, Issue 1
> 
> Send Wannier mailing list submissions to
> 	wannier at lists.quantum-espresso.org
> 
> To subscribe or unsubscribe via the World Wide Web, visit
> 	https://lists.quantum-espresso.org/mailman/listinfo/wannier
> or, via email, send a message with subject or body 'help' to
> 	wannier-request at lists.quantum-espresso.org
> 
> You can reach the person managing the list at
> 	wannier-owner at lists.quantum-espresso.org
> 
> When replying, please edit your Subject line so it is more specific
> than "Re: Contents of Wannier digest..."
> 
> 
> Today's Topics:
> 
>    1. executing wannier90.x in parallel: Program received signal
>       SIGSEGV: Segmentation fault - invalid memory reference.
>       (Simon Imanuel Rombauer)
> 
> 
> ----------------------------------------------------------------------
> 
> Message: 1
> Date: Wed, 05 Jul 2023 13:01:29 +0200
> From: "Simon Imanuel Rombauer"
> 	<simon.rombauer at student.uni-augsburg.de>
> To: wannier at lists.quantum-espresso.org
> Subject: [Wannier] executing wannier90.x in parallel: Program received
> 	signal SIGSEGV: Segmentation fault - invalid memory reference.
> Message-ID: <25ab64-64a54d80-eb-9a6d350 at 265464995>
> Content-Type: text/plain; charset="utf-8"
> 
> Dear all,
> 
> I am trying to get wannier90.x to run in parallel for some time now. I have installed it with the gfortran compiler and COMMS  = mpi ,MPIF90 = mpif90 in the make.inc file. All test pass when running in serial, when running it on 6 cores only 39 out of 62 tests pass. The error message is always the same (for instance from testw90_lavo3_dissphere):
> 
> Program received signal SIGSEGV: Segmentation fault - invalid memory reference.
> Backtrace for this error:
> Program received signal SIGSEGV: Segmentation fault - invalid memory reference.
> Backtrace for this error:
> Program received signal SIGSEGV: Segmentation fault - invalid memory reference.
> Backtrace for this error:
> Program received signal SIGSEGV: Segmentation fault - invalid memory reference.
> Backtrace for this error:
> Program received signal SIGSEGV: Segmentation fault - invalid memory reference.
> Backtrace for this error:
> #0  0x7f1409423ad0 in ???
> #1  0x7f1409422c35 in ???
> #2  0x7f140904251f in ???
> 	at ./signal/../sysdeps/unix/sysv/linux/x86_64/libc_sigaction.c:0
> #0  0x7f6f82623ad0 in ???
> #1  0x7f6f82622c35 in ???
> #2  0x7f6f8224251f in ???
> 	at ./signal/../sysdeps/unix/sysv/linux/x86_64/libc_sigaction.c:0
> #0  0x7fb45be23ad0 in ???
> #1  0x7fb45be22c35 in ???
> #2  0x7fb45ba4251f in ???
> 	at ./signal/../sysdeps/unix/sysv/linux/x86_64/libc_sigaction.c:0
> #3  0x56361c9b8470 in __w90_comms_MOD_comms_scatterv_cmplx_4
> 	at ../comms.F90:1247
> #3  0x56366c90f470 in __w90_comms_MOD_comms_scatterv_cmplx_4
> 	at ../comms.F90:1247
> #3  0x56143bd90470 in __w90_comms_MOD_comms_scatterv_cmplx_4
> 	at ../comms.F90:1247
> #4  0x56361c8f73c0 in __w90_overlap_MOD_overlap_read
> 	at ../overlap.F90:203
> #4  0x56366c84e3c0 in __w90_overlap_MOD_overlap_read
> 	at ../overlap.F90:203
> #4  0x56143bccf3c0 in __w90_overlap_MOD_overlap_read
> 	at ../overlap.F90:203
> #5  0x56361c89e80e in wannier
> 	at ../wannier_prog.F90:204
> #6  0x56361c89fed0 in main
> 	at ../wannier_prog.F90:55
> #5  0x56366c7f580e in wannier
> 	at ../wannier_prog.F90:204
> #6  0x56366c7f6ed0 in main
> 	at ../wannier_prog.F90:55
> #5  0x56143bc7680e in wannier
> 	at ../wannier_prog.F90:204
> #6  0x56143bc77ed0 in main
> 	at ../wannier_prog.F90:55
> #0  0x7fe9a9623ad0 in ???
> #1  0x7fe9a9622c35 in ???
> #2  0x7fe9a924251f in ???
> 	at ./signal/../sysdeps/unix/sysv/linux/x86_64/libc_sigaction.c:0
> #3  0x560ec2af1470 in __w90_comms_MOD_comms_scatterv_cmplx_4
> 	at ../comms.F90:1247
> #4  0x560ec2a303c0 in __w90_overlap_MOD_overlap_read
> 	at ../overlap.F90:203
> #5  0x560ec29d780e in wannier
> 	at ../wannier_prog.F90:204
> #6  0x560ec29d8ed0 in main
> 	at ../wannier_prog.F90:55
> #0  0x7fcd60823ad0 in ???
> #1  0x7fcd60822c35 in ???
> #2  0x7fcd6044251f in ???
> 	at ./signal/../sysdeps/unix/sysv/linux/x86_64/libc_sigaction.c:0
> #3  0x55b73e6b0470 in __w90_comms_MOD_comms_scatterv_cmplx_4
> 	at ../comms.F90:1247
> #4  0x55b73e5ef3c0 in __w90_overlap_MOD_overlap_read
> 	at ../overlap.F90:203
> #5  0x55b73e59680e in wannier
> 	at ../wannier_prog.F90:204
> #6  0x55b73e597ed0 in main
> 	at ../wannier_prog.F90:55
> --------------------------------------------------------------------------
> Primary job  terminated normally, but 1 process returned
> a non-zero exit code. Per user-direction, the job has been aborted.
> --------------------------------------------------------------------------
> --------------------------------------------------------------------------
> mpirun noticed that process rank 3 with PID 0 on node simon-ubuntu exited on signal 11 (Segmentation fault).
> --------------------------------------------------------------------------
> 
> This error does not occur when running mpirun -np 6 wannier90.x -pp 'seedname'. The output, wout-files confirm that the calculations started in parallel using 6 cores. 
> Any idea and input is highly appreciated!
> 
> Best,
> Simon Rombauer
> 
> Masters Student Physics
> Experimentalphysik IV
> University Augsburg
> Germany
> 
> 
> 
> ------------------------------
> 
> Subject: Digest Footer
> 
> _______________________________________________
> Wannier mailing list
> Wannier at lists.quantum-espresso.org
> https://lists.quantum-espresso.org/mailman/listinfo/wannier
> 
> 
> ------------------------------
> 
> End of Wannier Digest, Vol 184, Issue 1
> ***************************************


--
RuiPang
lecturer
Cell phone:(+86)18838272616
SchoolofPhysicalScienceandEngineering
ZhengzhouUniversity
Zhengzhou450001, People'sRepublicofChina