<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN">
<HTML><HEAD>
<META http-equiv=Content-Type content="text/html; charset=iso-8859-1">
<META content="MSHTML 6.00.6000.16674" name=GENERATOR>
<STYLE></STYLE>
</HEAD>
<BODY bgColor=#ffffff>
<DIV><FONT face=Arial size=2><FONT face="Times New Roman" size=3>Dear
Stefano,<BR> In Example02 I compared the
reference file to the results file outputs as
below:<BR>
reference
results<BR> si.phX.out
23065
16206<BR> si.phXsingle.out
25830
22505</FONT></FONT></DIV>
<DIV><FONT face=Arial size=2><FONT face="Times New Roman" size=3>At the end of
the results file in each case was the message "from broyden file factorization
error". Using grep in the PH directory I found matches for broyden noted
in the dynmat.x, matdyn.x, ph.x, and q2r.x binary files.<BR>
All of the remaining listed files were either exactly or
close to the same length. This agrees with the error message. The
problem is that I don't know where the problem is in the PH directory. Is
there any way to debug the directory? Nothing is said in the
user_guide. I did note in the phq_readin.f90 file line 154 it is possible
to reset lnscf to true but I am not sure what this will tell
me.<BR> Your comments would be appreciated.
<BR>
Amos Leffler<BR>
unaffiliated<BR><BR>Stefano de Gironcoli wrote: </FONT></DIV>
<BLOCKQUOTE cite=mid:4896B6E8.8090000@sissa.it type="cite">Dear Amos Leffler,
<BR> what you are showing us is the error found by the
script, which says that ...there was an error. <BR> have you
verified the content of the several output files that are created by the codes
in the result directory? they might be more informative and you may discover
that you not even need our help... <BR> best regards,
<BR> stefano <BR><BR>Amos Leffler wrote: <BR>
<BLOCKQUOTE type="cite">Dear Forum,
<BR> On May 12 Dong Hee Lim
reported a problem with Example 02 when running espresso-4.0.1. Paolo
G. asked for more details but I have not seen any since then. I am
reporting a similar problem with some of the output. These were run
using espresso-4.0.1 in serial mode using a fortran g95 fortran compiler.
<BR> <BR>/home/amos/Desktop/espresso-4.0.1/examples/example02 <BR>env
LD_LIBRARY_PATH=/opt/intel/mkl/10.0.3.020/lib/em64t ./run_example <BR>
<BR>This example shows how to use pw.x and ph.x to calculate phonon
frequencies at Gamma and X for Si and C <BR>in the diamond structure and for
fcc-Ni <BR> <BR> executables directory:
/home/amos/Desktop/espresso-4.0.1/bin <BR> pseudo
directory:
/home/amos/Desktop/espresso-4.0.1/pseudo <BR> temporary
directory /home/amos/tmp <BR> <BR>
checking that needed directories and files exist... done <BR>
<BR> running pw.x as :
/home/amos/Desktop/espresso-4.0.1/bin/pw.x <BR> running ph.x as
:
/home/amos/Desktop/espresso-4.0.1/bin/ph.x <BR> <BR> cleaning
/home/amos/tmp... done <BR> running the scf calculation...
done <BR> running the phonon calculation at Gamma... done
<BR> running the nscf calculation at X for a single mode... done
<BR> running the phonon calculation at X for a single mode ...
STOP 2 <BR>Error condition encountered during test: exit
status = 2 <BR>Aborting <BR> <BR>Example 07 <BR> <BR>This gives
the same sort of sequence through running pw.x, ph.x, q2r.x, matdyn.x.
It then goes through: cleaning /home/amos/tmp. <BR> <BR> running
the scf calculation with dense k-point grid... done <BR> running
the scf calculation... done <BR> running the el-ph
calculation... STOP 2 <BR>Error condtion encountered during
test: exit status = 2 <BR>Aborting <BR> <BR>Both examples stop in the
same portion of the run_example and probably have the same cause. Some
other examples complete properly but I have not tested all of them
<BR>
<BR>
Amos Leffler. <BR>
unaffiliated
<BR><BR><BR></BLOCKQUOTE></BLOCKQUOTE></FONT></BODY></HTML>