[Wien] parallel task stops at the end of LAPW1

Laurence Marks L-marks at northwestern.edu
Thu Jul 30 13:57:59 CEST 2009


I missed part of this discussion, I think using "& for background
tasks" is full of danger; you need to redirect both stdout but also
stderr using something like >& for csh or tcsh, see
http://www.faqs.org/faqs/unix-faq/faq/part2/section-9.html  and
http://www.is.ed.ac.uk/itus/sce/linuxunix/unix_redirection.html

On Thu, Jul 30, 2009 at 4:52 AM, Lyudmila Dobysheva<lyu at otf.pti.udm.ru> wrote:
> On 23 July 2009 Pawel Lesniak wrote:
>> If your login shell is tcsh, try running anything (eg. date &) in
>> background and check if it behaves different from lapw1.
>
> I have made a simple bat-file (analog of lapw1para):
> #!/bin/csh -f
> date > l-output &
> wait
> -----------------------------
> it works o'k.
>
> Using simplified fortran programs I have found that there is a strange
> behavior with output of the STOP operator:
> when there is a program in background (&) with STOP 'something to write',
> it cannot finish by itself and unload itself from the execution. The lapw1para
> waits for it and hangs.
> I have changed the last operators in lapw1.f and lapw2.f
> from
>      STOP ' LAPW2 END'
> to
>      STOP
>  and this works, but this is not a good solution.
>
> I checked the same with other compiler f77 --- the same strangeness.
>
> Who knows what happens?
>
> Build Operating System: Linux 2.6.24-15-server x86_64 Ubuntu
> Current Operating System: Linux 2.6.27-14-generic
> Ubuntu 8.10 (Oct. 2008)
>
> Best regards
> Lyudmila Dobysheva
>
> On Tuesday 21 July 2009 Lyudmila Dobysheva wrote:
>> > Non-parallel calculation works well, but the parallel over k-points stops
>> > in lapw1.
>> > More precisely in lapw1para_lapw: the execution does not return from the
>> > line: (cd $PWD;$t $exe ${def}_$loop.def;rm -f .lock_$lockfile[$p])
>> > >>.time1_$loop &
>> >
>> > It performs entirely lapw1_N (up to      STOP ' LAPW1 END'),
>> > cleans the uplapw1_N.error files,
>> > and stops.
>> > top shows that there are 8*lapw1,
>> > 1+8* lapw1_para's
>> > and one reappearing sleep. All with zero CPU %.
>> >
>> > .lockfiles exist in the directory.
> ------------------------------------------------------------------
> Phys.-Techn. Institute of Ural Br. of Russian Ac. of Sci.
> 426001 Izhevsk, ul.Kirova 132
> RUSSIA
> ------------------------------------------------------------------
> Tel.:7(3412) 442118 (home), 218988(office), 250614(Fax)
> E-mail: lyu at otf.fti.udmurtia.su, lyu at otf.pti.udm.ru
> http://fti.udm.ru/content/view/25/103/lang,english/
> ------------------------------------------------------------------
> _______________________________________________
> Wien mailing list
> Wien at zeus.theochem.tuwien.ac.at
> http://zeus.theochem.tuwien.ac.at/mailman/listinfo/wien
>



-- 
Laurence Marks
Department of Materials Science and Engineering
MSE Rm 2036 Cook Hall
2220 N Campus Drive
Northwestern University
Evanston, IL 60208, USA
Tel: (847) 491-3996 Fax: (847) 491-7820
email: L-marks at northwestern dot edu
Web: www.numis.northwestern.edu
Chair, Commission on Electron Crystallography of IUCR
www.numis.northwestern.edu/
Electron crystallography is the branch of science that uses electron
scattering to study the structure of matter.


More information about the Wien mailing list