[Wien] problem with SPAGHETTI !!!

[Peter Blaha]

I suggest you use w2web. It guides a beginner through these problems.

Anyway, you said you used

x lapw1 -band     but did not use "-c".

If this is true, already this step must give an error. Please carefully 
read what it says on the screen. You must not have a file case.in1, but 
only case.in1c ???? Is this true ?

Use the "standard syntax":  x lapw1 -c -band    and not   x lapw1c -band
(although this might still work).

look into spaghetti.def. You will NOT see any *.energy* files, but 
spaghetti reads the output1 files. Does case.output1 look ok ? Does it
end with some timing info ? Are the eigenvalue lists ok, no ***** ????

I don't think that unit -5 is case.insp (this is unit 5), but just for 
safety: Did you put EF into this file (or you still have **** there ??)

The suggested compiler options include   -traceback
With this you should get a bit more info where the problem occurs.

katrusiat schrieb:
> Dear Wien2k creators,
> Some days ago I described a problem with SPAGHETTI. I was getting a message about segmentation fault when trying toexecute it.
> Unfortunately, nobody responded to my question.
> I was thinking about the problem though. The system had no inversion and the scf cycle was performed with lapw1c etc.Naturally, one may think that the eigenvalues along a path in k-space have to be calculated as
> x lapw1c -band.
> However, I was too much used to systems with an inversion and automatically typed
> x lapw1 -band.
> That is when I got the message about the segmentation fault error.
> Now, I tried the lapw1c version, but got an error again. This time the message reads (after x spaghetti):
> forrtl: severe (64): input conversion error, unit -5, file Internal Formatted ReadImage              PC                Routine            Line        Sourcespaghetti          0000000000460F07  Unknown               Unknown  Unknownspaghetti          000000000045F2CE  Unknown               Unknown  Unknownspaghetti          000000000045D9D8  Unknown               Unknown  Unknownspaghetti          000000000043B8EB  Unknown               Unknown  Unknownspaghetti          000000000043B554  Unknown               Unknown  Unknownspaghetti          0000000000424877  Unknown               Unknown  Unknownspaghetti          0000000000423D24  Unknown               Unknown  Unknownspaghetti          000000000040976E  Unknown               Unknown  Unknownspaghetti          0000000000414C33  Unknown               Unknown  Unknownspaghetti          0000000000400316  Unknown               Unknown  Unknownspaghetti          00000000004655FC  Unknown               Unknown  Unknownspaghe
tti          0000000000400229  Unknown               Unknown  Unknown0.004u 0.000s 0:00.00 0.0%      0+0k 0+16io 0pf+0werror: command   /home/foyevtsova/static07_ifort_icc_9.0_mkl_9.1-nmatmax14000/spaghetti spaghetti.def   failed
> I inspected the .insp (unit -5) file. It looks ok (with the same outlook of .insp, I had no problems with simpler systems). But I noticed that in most of the .energy_1 etc files (from parallel calculations) the data are written up to some band (the 70th) for the second k-point. In a few .energy_*, the data are complete.
> Please, help me to understand the root of the problem, which might be very trivial.
> Thanks,Kateryna Foyevtsova
> 

-- 
-----------------------------------------
Peter Blaha
Inst. Materials Chemistry, TU Vienna
Getreidemarkt 9, A-1060 Vienna, Austria
Tel: +43-1-5880115671
Fax: +43-1-5880115698
email: pblaha at theochem.tuwien.ac.at
-----------------------------------------