<DIV>&gt;Message:&nbsp;2<BR>&gt;Date:&nbsp;Wed,&nbsp;23&nbsp;Sep&nbsp;2009&nbsp;15:06:07&nbsp;+0200<BR>&gt;From:&nbsp;"Lorenzo&nbsp;Paulatto"&nbsp;&lt;paulatto@sissa.it&gt;<BR>&gt;Subject:&nbsp;Re:&nbsp;[Pw_forum]&nbsp;Pw_forum&nbsp;Digest,&nbsp;Vol&nbsp;27,&nbsp;Issue&nbsp;74<BR>&gt;To:&nbsp;"PWSCF&nbsp;Forum"&nbsp;&lt;pw_forum@pwscf.org&gt;<BR>&gt;Message-ID:&nbsp;&lt;op.u0pn0hl7a8x26q@paulax&gt;<BR>&gt;Content-Type:&nbsp;text/plain;&nbsp;charset=utf-8;&nbsp;format=flowed;&nbsp;delsp=yes<BR>&gt;<BR>&gt;In&nbsp;data&nbsp;23&nbsp;settembre&nbsp;2009&nbsp;alle&nbsp;ore&nbsp;14:12:53,&nbsp;wangqj1&nbsp;&lt;wangqj1@126.com&gt;&nbsp;ha&nbsp;&nbsp;<BR>&gt;scritto:<BR>&gt;&gt;&gt;&nbsp;/home/wang/bin/pw.x&nbsp;-npool&nbsp;8&nbsp;-in&nbsp;ZnO.pw.inp&gt;ZnO.pw.out<BR>&gt;&gt;&nbsp;It&nbsp;only&nbsp;print&nbsp;the&nbsp;output&nbsp;heander&nbsp;and&nbsp;not&nbsp;have&nbsp;iteration&nbsp;.</DIV><PRE>
&gt;Maybe&nbsp;it's&nbsp;not&nbsp;reading&nbsp;the&nbsp;input&nbsp;at&nbsp;all!&nbsp;You&nbsp;should&nbsp;specify&nbsp;the&nbsp;full&nbsp;path&nbsp;&nbsp;
&gt;to&nbsp;you&nbsp;input&nbsp;file,&nbsp;e.g.
&gt;&nbsp;&nbsp;&nbsp;mpirun&nbsp;...&nbsp;bin/pw.x&nbsp;-in&nbsp;/where/is/is/ZnO.pw.inp
&gt;just&nbsp;to&nbsp;be&nbsp;sure.
</PRE><PRE>The output file is as following:
 Program PWSCF     v.4.0.1  starts ...</PRE><PRE>     Today is  24Sep2009 at  8:49:30
      Parallel version (MPI)
      Number of processors in use:       8
     R &amp; G space division:  proc/pool =    8
      For Norm-Conserving or Ultrasoft (Vanderbilt) Pseudopotentials or PAW
 .....................................................................
      Initial potential from superposition of free atoms
      starting charge  435.99565, renormalised to  436.00000
      Starting wfc are  254 atomic +    8 random wfc</PRE><PRE>It seems that it had read the input file  .When use K-point parallel ,it runs well.

&gt;
&gt;&gt;&nbsp;I&nbsp;am&nbsp;not&nbsp;the&nbsp;supperuser,I&nbsp;don't&nbsp;know&nbsp;how&nbsp;to&nbsp;Set&nbsp;the&nbsp;environment&nbsp;variable&nbsp;&nbsp;
&gt;&gt;&nbsp;OPEN_MP_THREADS&nbsp;to&nbsp;1,I&nbsp;can't&nbsp;find&nbsp;where&nbsp;is&nbsp;OPEN_MP_THREADS&nbsp;?
&gt;
&gt;You&nbsp;don't&nbsp;need&nbsp;to&nbsp;be&nbsp;super-user&nbsp;(nor&nbsp;supper-user&nbsp;;-)&nbsp;to&nbsp;set&nbsp;an&nbsp;environment&nbsp;&nbsp;
&gt;variable,&nbsp;you&nbsp;only&nbsp;have&nbsp;to&nbsp;add
&gt;&nbsp;&nbsp;export&nbsp;OMP_NUM_THREADS=1
&gt;in&nbsp;your&nbsp;PBS&nbsp;script,&nbsp;before&nbsp;calling&nbsp;mpirun.&nbsp;To&nbsp;be&nbsp;sure&nbsp;it's&nbsp;propagated&nbsp;to&nbsp;&nbsp;
&gt;all&nbsp;the&nbsp;processors&nbsp;add&nbsp;the&nbsp;option
&gt;&nbsp;&nbsp;-x&nbsp;OMP_NUM_THREAD
&gt;to&nbsp;mpirun&nbsp;arguments&nbsp;(BEFORE&nbsp;pw.x).
I want to know whether it is necessary to set environment&nbsp;variable OPEN_MP_THREADS&nbsp;(I installed MPICH2 not OPENMPI) ?</PRE><PRE>Thanks very much for your help !</PRE><PRE>Best regards</PRE><PRE>&nbsp;</PRE><PRE>&nbsp;</PRE><br><br><span title="neteasefooter"/><hr/>
<a href="http://news.163.com/madeinchina/index.html?from=mailfooter">"中国制造",讲述中国60年往事</a>
</span>