[Wien] Problems with self-compiled lapw2
Peter Blaha
pblaha at theochem.tuwien.ac.at
Fri Feb 23 18:21:25 CET 2007
The first item you cite had a .machines file like:
1:localhost
1:localhost
1:localhost
1:localhost
granularity:1
Most likely this does not work. You should substitute "localhost" by the
actual name of your system.
k-point parallelization is done by ss calls.
Test it by: ssh localhost "cd $PWD;lapw1 lapw1_1.def"
Does it work ? Do you need a password,... ??
2nd item you mentioned concerns lapw2: for a parallel lapw2 the "direct
calling" would need another parameter like:
lapw2 lapw2_1.def 1
Nils Huntemann schrieb:
> Dear WIEN2k-users,
> I tried to run run_lapw -p and I got the same result as described in here:
>
> http://zeus.theochem.tuwien.ac.at/pipermail/wien/2006-October/008200.html
>
> I also got an error in lapw2. But if I start the same calculation without -p everything works fine. I tried to run lapw2 lapw2_1.def as I found in here:
>
> http://zeus.theochem.tuwien.ac.at/pipermail/wien/2006-October/008210.html
>
> But I got a segmentation fault. If I use the precompiled version instead, everything works, but it is very slow.
> To compile I used ifort 9.1.041 and mkl 8.1.1. on a Intel E6400 Suse Linux 10.2 (Kernel 2.6.18.2-34-default x86_64), 1 GB ram. I tried many different options (also -O0) and all I found in here: http://www.ghfecher.de/Fecher_CompileIntel.pdf
> Now I used:
>
> Compiler options:-FR -mp1 -w -prec_div -pc80 -pad -ip -DINTEL_VML -O3 -xP
> Linker Flags:-L/opt/intel/fce/9.1.041/lib/ -lguide -lpthread
> R_LIB (LAPACK+BLAS):-L/opt/intel/mkl/8.1.1/lib/em64t/ -lmkl_lapack -lmkl_em64t -lguide -lpthread
>
> My Version is WIEN2k_07.2 (Release 20/2/2007), but I get the same result with WIEN2k_06.4.
> If I use the self compiled Version in non-parallel case it works. Where is the difference causing the segmentation fault, if I use lapw2 in parallel?
> I hope you can help me,
> best regards,
> Nils
>
> lapw2_1.def:
> 2,'wien.nsh', 'unknown','formatted',0
> 3,'wien.in1', 'unknown','formatted',0
> 4,'wien.inso', 'unknown','formatted',0
> 5,'wien.in2', 'old', 'formatted',0
> 6,'wien.output2_1','unknown','formatted',0
> 8,'wien.clmval_1','unknown','formatted',0
> 10,'wien.vector_1', 'unknown','unformatted',9000
> 11,'wien.weight', 'unknown','formatted',0
> 13,'wien.recprlist', 'unknown','unformatted',9000
> 14,'wien.kgen', 'unknown','formatted',0
> 15,'wien.tmp', 'unknown','formatted',0
> 16,'wien.qtl', 'unknown','formatted',0
> 17,'wien.weightaver','unknown','formatted',0
> 18,'wien.vsp', 'old', 'formatted',0
> 19,'wien.vns', 'unknown','formatted',0
> 20,'wien.struct', 'old', 'formatted',0
> 21,'wien.scf2_1', 'unknown','formatted',0
> 22,'wien.rotlm', 'unknown', 'formatted',0
> 23,'wien.radwf', 'unknown', 'formatted',0
> 24,'wien.almblm', 'unknown', 'formatted',0
> 26,'wien.weigh_1', 'unknown','unformatted',0
> 27,'wien.weighdn_1', 'unknown','unformatted',0
> 28,'wien.vrespval_1', 'unknown','formatted',0
> 29,'wien.energydn_1','unknown','formatted',0
> 30,'wien.energy_1', 'unknown','formatted',0
> 31,'wien.help_1', 'unknown','formatted',0
>
--
P.Blaha
--------------------------------------------------------------------------
Peter BLAHA, Inst.f. Materials Chemistry, TU Vienna, A-1060 Vienna
Phone: +43-1-58801-15671 FAX: +43-1-58801-15698
Email: blaha at theochem.tuwien.ac.at WWW:
http://info.tuwien.ac.at/theochem/
--------------------------------------------------------------------------
More information about the Wien
mailing list