[Wien] A bug in "runsp_c_lapw" calculations with the option of "-orb -p" in WIEN2k_11.1 (Release 14/6/2011)

Peter Blaha pblaha at theochem.tuwien.ac.at
Tue Jan 24 12:02:00 CET 2012


Thank's for the report.

In order to speed up the runsp_c  calculations, in the new version lapw2 -dn is not
executed, but the necessary files are copied from the corresponding "up" files.

Since lapwdm is usually so fast, this was not done for this step. However, in parallel
mode it would not work this way.

Therefore one has to make the same "trick" as with lapw2:   comment the spin-dn execution and
add two "cp" lines:

total_exec	lapwdm -up $para
# total_exec	lapwdm -dn $para
cp  $file.scfdmup $file.scfdmdn
cp  $file.dmatup $file.dmatdn


Am 24.01.2012 10:37, schrieb Zhiyong Zhu:
> Dear wien2k users and authors,
>
> I have found a bug in the "runsp_c_lapw" calculations with the option
> of "-orb -p" in WIEN2k_11.1 (Release 14/6/2011):
>
> An error in sumpara calculations just after x lapwdm -p -dn is caused
> because of the bug: The case.scf2dn_* files, which are required in the
> "x lapwdm -dn -p; sumpara -dn -d" calculations, are not produced,
> because the absence of  "(x) lapw2 -dn -p; x sumpara -dn -d" in the
> scf cycle in WIEN2k_11.1.
>
>
> Below is the detail.
>
> I have encountered an error when I run a calculation with the command:
> -----------------------------------------------------------------------------
> runsp_c_lapw  -cc 0.0001 -ec 0.00001 -fc 1.0 -i 999 -orb -p -NI
> ----------------------------------------------------------------------------
>
> The :log file is:
> --------------------------------------------------------------------------
>>    (runsp_c_lapw) options: -cc 0.0001 -ec 0.00001 -fc 1.0 -i 999 -orb -p -NI
> Tue Jan 24 11:17:47 AST 2012>  (x) lapw0 -p
> Tue Jan 24 11:17:57 AST 2012>  (x) lapw1 -up -p -orb
> Tue Jan 24 11:19:02 AST 2012>  (x) lapw2 -up -p
> Tue Jan 24 11:19:45 AST 2012>  (x) sumpara -up -d
> Tue Jan 24 11:19:45 AST 2012>  (x) lapwdm -up -p
> Tue Jan 24 11:19:53 AST 2012>  (x) sumpara -up -d
> Tue Jan 24 11:19:54 AST 2012>  (x) lapwdm -dn -p
> Tue Jan 24 11:20:01 AST 2012>  (x) sumpara -dn -d
> --------------------------------------------------------------------------
>
> The error message is:
> ----------------------------------------------------------------------------
>   LAPW0 END
>   LAPW1 END
>   LAPW1 END
>   LAPW1 END
>   LAPW1 END
>   LAPW1 END
>   LAPW1 END
> LAPW2 - FERMI; weighs written
>   LAPW2 END
>   LAPW2 END
>   LAPW2 END
>   LAPW2 END
>   LAPW2 END
>   LAPW2 END
>   SUMPARA END
> LAPWDM END
> LAPWDM END
> LAPWDM END
> LAPWDM END
> LAPWDM END
> LAPWDM END
>   SUMPARA END
> LAPWDM END
> LAPWDM END
> LAPWDM END
> LAPWDM END
> forrtl: No such file or directory
> forrtl: severe (29): file not found, unit 21, file
> /scratch/zhuz/calc/33ACRO/02/01/acro/acro.scf2dn_1
> Image              PC                Routine            Line        Source
> sumpara            00000000004A24AD  Unknown               Unknown  Unknown
> sumpara            00000000004A0FB5  Unknown               Unknown  Unknown
> sumpara            0000000000451F39  Unknown               Unknown  Unknown
> sumpara            0000000000413A8A  Unknown               Unknown  Unknown
> sumpara            0000000000413280  Unknown               Unknown  Unknown
> sumpara            000000000041EE6D  Unknown               Unknown  Unknown
> sumpara            000000000040EB5B  scfsum_                   121  scfsum.F
> sumpara            000000000040AD98  MAIN__                    217  sumpara.f
> sumpara            00000000004033DC  Unknown               Unknown  Unknown
> libc.so.6          00000035AAA1D994  Unknown               Unknown  Unknown
> sumpara            00000000004032E9  Unknown               Unknown  Unknown
> ----------------------------------------------------------------------------
>
>
> There is no such kind of error,
> 1) if I use WIEN2k_09.1 (Release 5/2/2009)
> 2) if I do calculations with the command "runsp_lapw" instead of "runsp_c_lapw"
>
>
> I find the reason as follows:
>
> First, I compare the calculation sequence of WIEN2k_09.1 and WIEN2k_11.1:
> For WIEN2k_09.1 :
> --------------------------------------------------------------------------
>>    (runsp_lapw) options: -cc 0.0001 -ec 0.00001 -fc 1.0 -i 999 -orb -p -NI
> Tue Jan 24 11:44:56 AST 2012>  (x) lapw0 -p
> Tue Jan 24 11:45:05 AST 2012>  (x) lapw1 -up -p -orb
> Tue Jan 24 11:46:04 AST 2012>  (x) lapw1 -dn -p -orb
> Tue Jan 24 11:47:05 AST 2012>  (x) lapw2 -up -p
> Tue Jan 24 11:47:40 AST 2012>  (x) sumpara -up -d
> Tue Jan 24 11:47:40 AST 2012>  (x) lapw2 -dn -p
> Tue Jan 24 11:48:13 AST 2012>  (x) sumpara -dn -d
> Tue Jan 24 11:48:13 AST 2012>  (x) lapwdm -up -p
> Tue Jan 24 11:48:21 AST 2012>  (x) sumpara -up -d
> Tue Jan 24 11:48:21 AST 2012>  (x) lapwdm -dn -p
> Tue Jan 24 11:48:29 AST 2012>  (x) sumpara -dn -d
> Tue Jan 24 11:48:30 AST 2012>  (x) lcore -up
> Tue Jan 24 11:48:30 AST 2012>  (x) lcore -dn
> Tue Jan 24 11:48:30 AST 2012>  (x) mixer
> --------------------------------------------------------------------------
>
> For WIEN2k_11.1:
> --------------------------------------------------------------------------
>>    (runsp_c_lapw) options: -cc 0.0001 -ec 0.00001 -fc 1.0 -i 999 -orb -p -NI
> Tue Jan 24 11:17:47 AST 2012>  (x) lapw0 -p
> Tue Jan 24 11:17:57 AST 2012>  (x) lapw1 -up -p -orb
> Tue Jan 24 11:19:02 AST 2012>  (x) lapw2 -up -p
> Tue Jan 24 11:19:45 AST 2012>  (x) sumpara -up -d
> Tue Jan 24 11:19:45 AST 2012>  (x) lapwdm -up -p
> Tue Jan 24 11:19:53 AST 2012>  (x) sumpara -up -d
> Tue Jan 24 11:19:54 AST 2012>  (x) lapwdm -dn -p
> Tue Jan 24 11:20:01 AST 2012>  (x) sumpara -dn -d
> --------------------------------------------------------------------------
>
> It can be seen that "(x) lapw2 -dn -p; x sumpara -dn -d" is missed in
> WIEN2k_11.1.
>
> Therefore, the above error occurs because:
>
> The case.scf2dn_* files, which are required in the "x lapwdm -dn -p;
> sumpara -dn -d" calculations, are not produced, because the absence of
>   "(x) lapw2 -dn -p; x sumpara -dn -d" in the scf cycle in WIEN2k_11.1.
>
>
> Best Regards,
> Zhiyong Zhu
> _______________________________________________
> Wien mailing list
> Wien at zeus.theochem.tuwien.ac.at
> http://zeus.theochem.tuwien.ac.at/mailman/listinfo/wien

-- 

                                       P.Blaha
--------------------------------------------------------------------------
Peter BLAHA, Inst.f. Materials Chemistry, TU Vienna, A-1060 Vienna
Phone: +43-1-58801-165300             FAX: +43-1-58801-165982
Email: blaha at theochem.tuwien.ac.at    WWW: http://info.tuwien.ac.at/theochem/
--------------------------------------------------------------------------


More information about the Wien mailing list