[Wien] Wien2k_16.1
Fecher, Gerhard
fecher at uni-mainz.de
Wed Feb 15 09:47:18 CET 2017
Note that some segmentation faults (SIGSEGV error) are not from the compiler but from the pthread library (not compiler but linux specific)
usually this should not appear in dymamic linking
it appears also that dynamically (!!) linking leads to problems if the libpthread on the computer where you compiled and the computer where you run the program are different.
Ciao
Gerhard
DEEP THOUGHT in D. Adams; Hitchhikers Guide to the Galaxy:
"I think the problem, to be quite honest with you,
is that you have never actually known what the question is."
====================================
Dr. Gerhard H. Fecher
Institut of Inorganic and Analytical Chemistry
Johannes Gutenberg - University
55099 Mainz
and
Max Planck Institute for Chemical Physics of Solids
01187 Dresden
________________________________________
Von: Wien [wien-bounces at zeus.theochem.tuwien.ac.at] im Auftrag von Gavin Abo [gsabo at crimson.ua.edu]
Gesendet: Mittwoch, 15. Februar 2017 07:39
An: A Mailing list for WIEN2k users
Betreff: Re: [Wien] Wien2k_16.1
To me, it seems that the error could likely be due to a buggy Intel compiler version. Which is something the WIEN2k community knows about from past experience. For example, the post [1].
What ifort version did you use to compile WIEN2k 16.1 (e.g., the version should be given by terminal command: ifort -v)?
As mentioned before, recently [2], you don't want to use ifort version 16.0.3 as it is known to have a serious bug that could cause such an error, and Intel recommends according to DPD200411901 [3] to use ifort version 16.0.4 (update 4) [4] or the initial releases of 17.0.0 or 17.0.1 [5].
If you happen to be using an ifort version 15.x just before 16.0.0, perhaps there is also segmentation fault and reading and writing file issues as in the fixes list [6] there are things like:
DPD200363367, Fortran, SIGSEGV at runtime when compiling with -ipo
DPD200363710, Fortran, Errors doing INQUIRE, READ and WRITE in parallel threads
DPD200364274, Fortran, UDIO direct, unformatted read/write don't call child routine, don't transfer data
The following is a list of some of the things that I try when I encounter such a problem.
a) Backup the effected executable (lapw2c) in your $WIENROOT directory that you compiled and try replacing it with the one from the executables package (WIEN2k_16.1_executables.tar.gz).
b) Try different compiler settings.
- Try different optimization levels like -O0 and -O1. The compiler can sometimes over-optimize at a high optimization level like -O3.
- Try to compile against static libraries instead of dynamic libraries, or vice versa. Intel link advisor [7] can help you change the compiler settings between static and dynamic linking.
c) Try a different compiler
- Try different versions of the ifort compiler. Sometimes one version works and another does not.
- Try a different compiler (gfortran)
- Try 32 bit compiler instead of 64 bit. Not recommended to do this on 64 bit system unless you have to. I have a 64 bit laptop, and I could only get Wien2k working on it by creating a 32 bit executable instead of 64 bit. However, this is likely only because it has an unusual processor (a PlayStation processor instead of a typical Intel one like a Core 2 or Intel i8).
d) Try a different operating system (preferably one that is officially supported by the compiler)
- If you see the System Requirements section in the Intel release notes [8] for your ifort version, you will likely see that Intel tests and official supports their ifort compiler on operating systems such as Ubuntu and Red Hat. For example, Scientific Linux is not in the list, but others have Wien2k working with it [9]. Though, there is always the slight chance that the operating system and ifort compiler might not work well together when they are not official tested together.
Hope this helps and good luck.
[1] http://www.mail-archive.com/wien@zeus.theochem.tuwien.ac.at/msg03925.html
[2] http://www.mail-archive.com/wien@zeus.theochem.tuwien.ac.at/msg14954.html
[3] https://software.intel.com/en-us/articles/read-failure-unformatted-file-io-psxe-16-update-3
[4] https://software.intel.com/en-us/articles/intel-fortran-compiler-160-for-linux-release-notes-for-intel-parallel-studio-xe-2016
[5] https://software.intel.com/en-us/articles/intel-fortran-compiler-170-for-linux-release-notes-for-intel-parallel-studio-xe-2017
[6] https://software.intel.com/en-us/articles/intel-parallel-studio-xe-2016-composer-edition-compilers-fixes-list
[7] https://software.intel.com/en-us/articles/intel-mkl-link-line-advisor
[8] https://software.intel.com/en-us/articles/intel-composer-release-notes-by-version
[9] http://www.mail-archive.com/wien%40zeus.theochem.tuwien.ac.at/msg06328.html
On 2/14/2017 5:27 AM, Rajneesh Chaurasiya wrote:
Dear Wien2k users,
I have optimized one perovskite based material in the past using wien2k_13 version and also computed properties but when i ran the same case.struct file using wien2k_16.1 version under the same environment then it gives error like
LAPW0 END
LAPW1 END
forrtl: severe (39): error during read, unit 10, file /home/IITJHOME/ambeshst/WIEN2k/GGA.vector
Image PC Routine Line Source
lapw2c 000000000051C01A Unknown Unknown Unknown
lapw2c 000000000051AB95 Unknown Unknown Unknown
lapw2c 00000000004C96B6 Unknown Unknown Unknown
lapw2c 000000000047BE86 Unknown Unknown Unknown
lapw2c 000000000047B5F9 Unknown Unknown Unknown
lapw2c 000000000049FBF4 Unknown Unknown Unknown
lapw2c 0000000000466C74 read_vec_ 164 read_vec_tmp_.F
lapw2c 0000000000440443 l2main_ 661 l2main_tmp_.F
lapw2c 00000000004598DF MAIN__ 716 lapw2_tmp_.F
lapw2c 00000000004040AC Unknown Unknown Unknown
libc.so.6 000000344221D9C4 Unknown Unknown Unknown
lapw2c 0000000000403FB9 Unknown Unknown Unknown
Can you suggest me a possible reason behind it.
--
Thanks & Regards
Rajneesh Chaurasiya
Research Scholar
IIT Jodhpur, India
Mob. No. +91-9584499697
+91-7610950803
More information about the Wien
mailing list