[Wien] 4p energy so low and unaligned access$B!!(B(hao@qs.t.u-tokyo.ac.jp)

[Torsten Andersen]

I would guess that "unaligned access" refers to an "odd" memory address, 
and thus a consequence of compiler, options, and libraries. However, 
since I do not know the specific setup from experience I am unable to 
specify it any further. Try to run "mixer" with the debugger and see if 
that can trace the problem. Otherwise, you might try to compile "mixer" 
with -O0 (no optimization) and leave out threading and other fancy 
options -- then it should (hopefully) disappear. You will probably not 
lose much performance by that.

Best regards,
Torsten Andersen.

Peter Blaha wrote:
> You missed an important point:
> 
> It is NOT important what is valence or core in case.inst.
> 
> It is IMPORTANT, however, what separation energy you specified during 
> initialization. Aparently you lowered the standard -6.0 Ry such that the 
> 3p of Ge becomes a valence state.
> 
> I've never seen the message you got in mixer, so I can't help.
> 
> 
>>when I focus on Ge3N4, after I run 
> 
> run_lapw, 
> I find that in the last > iteration of scf file sth like this appears:
> 
>>ATOMIC SPHERE DEPENDENT PARAMETERS FOR ATOM  Ge        
>>
>>          OVERALL ENERGY PARAMETER IS    0.3000
>>
>>          OVERALL BASIS SET ON ATOM IS LAPW
>>
>>          E( 1)=    0.3000
>>
>>             APW+lo
>>
>>          E( 1)=   -7.5200   E(BOTTOM)=   -7.615   E(TOP)=   -7.425
>>
>>             LOCAL ORBITAL
>>
>>          E( 2)=   -1.2250   E(BOTTOM)=   -1.560   E(TOP)=   -0.890
>>
>>             APW+lo
>>
>>          E( 2)=    0.3000
>>
>>             LOCAL ORBITAL
>>
>>          E( 0)=    0.3000
>>
>>             APW+lo
>>
>>In my inst file 3p of Ge is in core state and should not appear in
>>bandenergy, so E(1) must be 4p, is it possible that 4p energy is below 4s so
>>much? And if so, 4p should be input into corestate? I check the help031
>>file, Q(UE) is not greater than Q(U), so there is no ghostband. Would you
>>please help me? 
>>
>>The second when I focus on another material, I run run_lapw and get
>>information like this:
>>
>> 
>>
>>LAPW0 END
>>
>> LAPW1 END
>>
>> LAPW2 END
>>
>> CORE  END
>>
>>mixer(10401): unaligned access to 0x60000000000498c1, ip=0x40000000000361b0
>>
>>mixer(10401): unaligned access to 0x60000000000498c1, ip=0x40000000000361b0
>>
>>mixer(10401): unaligned access to 0x60000000000498c1, ip=0x40000000000361b0
>>
>>mixer(10401): unaligned access to 0x60000000000498c1, ip=0x40000000000361b0
>>
>> MIXER END
>>
>>And program is still continue, what does that mean?
>>
>>My computer is IA 64 tiger2 system with redhat linux and ifort, mkl70 with
>>lib 64, wien2k_04.11. compile file is 
>>
>>        -FR -mp -w -pad -ip -DINTEL_VML -tpp2 -ftz
>>
>>        -L../SRC_lib -L/opt/intel/mkl70/lib/64 -Vaxlib -static
>>-static-libcxa -pthread
>>
>>       -DParallel'
>>
>>       -lmkl_lapack -lmkl_ipf -lguide
>>
>>I have compiled wien2k successfully. Could you please help me?
>>
>> 
>>
>>Thank you 
>>
>> 
>>
>>hao
>>
>>
> 
> 
> 
>                                       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/
> --------------------------------------------------------------------------
> _______________________________________________
> Wien mailing list
> Wien at zeus.theochem.tuwien.ac.at
> http://zeus.theochem.tuwien.ac.at/mailman/listinfo/wien
> 

-- 
Dr. Torsten Andersen        TA-web: http://deep.at/myspace/
AG Hübner, Department of Physics, Kaiserslautern University
http://cmt.physik.uni-kl.de    http://www.physik.uni-kl.de/