Problems with the MgB2 example

[qe_beginner]

Hi!

I'm trying to run the MgB2 example, but when epw.x is executed, I get the following error message:

forrtl: severe (174): SIGSEGV, segmentation fault occurred
Image PC Routine Line Source
epw.x 0000000000CF1711 Unknown Unknown Unknown
epw.x 0000000000CEFE67 Unknown Unknown Unknown
epw.x 0000000000C92A64 Unknown Unknown Unknown
epw.x 0000000000C92876 Unknown Unknown Unknown
epw.x 0000000000C2743F Unknown Unknown Unknown
epw.x 0000000000C2E68D Unknown Unknown Unknown
libpthread.so.0 00002B9D570247E0 Unknown Unknown Unknown
libblas.so.3 00002B9D55CEE510 Unknown Unknown Unknown
epw.x 00000000005B8C36 find_mode_sym_new 148 find_mode_sym.f90
epw.x 000000000058007A set_irr_new_ 143 set_irr.f90
epw.x 000000000054921D find_irrep_ 32 find_irrep.f90
epw.x 00000000004625B0 epw_setup_ 329 epw_setup.f90
epw.x 0000000000406E19 MAIN__ 102 epw.f90
epw.x 000000000040657E Unknown Unknown Unknown
libc.so.6 00002B9D574D4D1D Unknown Unknown Unknown
epw.x 0000000000406489 Unknown Unknown Unknown

I have checked the epw.out file, but it doesn't say more than "mpprun info: Job terminated with error". I suspect that this error might have to do with how I may have set up my "outdir", but the solutions that I've tried so far haven't been successful. Also, I would expect the error message to be a bit clearer in that case (usually when a file or a directory cannot be found, the slurm file says so very clearly).

[sponce]

Dear qe_beginner,

You can try by using all outdir = './' (so in the same folder as the one your are running).

Also make sure that EPW can correctly access the "save" folder from the phonon calculations.

This could also be a compilation error.
Please make sure you are using the latest EPW 4.1 version as some bugs have been fixed.

To help you more I would need to see your input files.

Best,

Samuel

[qe_beginner]

Dear Samuel,

thank you very much for your quick reply. I now have QE 6.0 and EPW 4.1 installed, and I will test if it works better this time. It is nice to see that in this version you have already prepared an almost complete RUN script, and put all the necessary input files in one directory!

[nu2epw]

I'm having trouble with this example as well.

I'm using QE 6 and epw 4.1

when I get to the point of running epw.x
I get the following errors to stderr

namelist read: misplaced = sign
Cannot match namelist object name .true.
Cannot match namelist object name phinterp
namelist read: misplaced = sign
Cannot match namelist object name .true.
Cannot match namelist object name tshuffle2
namelist read: misplaced = sign
Cannot match namelist object name .true.
Cannot match namelist object name tphases
namelist read: misplaced = sign
Cannot match namelist object name .false.

The code executes and appears to finish the call to wannier90 but the crashes with a segfault at

Calculating kmap and kgmap
Progress kmap: ########################################
Progress kgmap: ########################################
kmaps : 1.09s CPU 1.17s WALL ( 1 calls)
Symmetries of bravais lattice: 24
Symmetries of crystal: 24


===================================================================
irreducible q point # 1
===================================================================

any ideas what Im doing wrong?

[sponce]

Hello,

Some input variables have been removed.

Try removing "phinterp, tshuffle2 and tphases" from your input file.


PS: A list of removed input variables is available at the bottom of http://epw.org.uk/Main/About
Best,

Samuel

[nu2epw]

I removed the deprecated input variables but I'm seeing a crash at the same point. are there recommended compiler versions and options?

[sponce]

Hello,

Can you post the segfault message? Which compiler are you using?

If QE compiles and run correctly then it should be the same for epw in principle.

Best,
Samuel

[nu2epw]

I'm using GNU Fortran (GCC) 4.4.7 20120313 (Red Hat 4.4.7-17)

I've tried both the mgb2 and diamond examples

everything seems to work ok up to the point when I run epw

for the diamond example I get output like this

../../../qe-test/qe-6.0/bin/epw.x < epw.in

--snip--

Dipole matrix elements calculated


Calculating kmap and kgmap
Progress kmap: ########################################
Progress kgmap: ########################################
kmaps : 0.69s CPU 0.76s WALL ( 1 calls)
Symmetries of bravais lattice: 48
Symmetries of crystal: 48


===================================================================
irreducible q point # 1
===================================================================

Symmetries of small group of q: 48
in addition sym. q -> -q+G:

Number of q in the star = 1
List of q in the star:
1 0.000000000 0.000000000 0.000000000
Imposing acoustic sum rule on the dynamical matrix

q( 1 ) = ( 0.0000000 0.0000000 0.0000000 )


epw.x:243529 terminated with signal 11 at PC=2aaaacbc4338 SP=7fffffff1690. Backtrace:


-------

I've tried running serial and parallel etc but the end result is the same --

[nu2epw]

still running but it looks like its working with gcc-3.6.1