[molpro-user] Error prior to EOM-CCSD iterations (failure in dsyev , The problem occurs in diag2)

Andy May MayAJ1 at cardiff.ac.uk
Thu Aug 1 15:17:30 BST 2013


Evgeniy,

Just open src/util/util.F and add after line 792 the extra 1 line:

           nw=min(nw,maxwork)

so it reads like:

           if(nw.lt.minwork) goto 100
           nw=min(nw,maxwork)
           iwork=icorr(nw)

Best wishes,

Andy

On 01/08/13 15:02, Evgeniy Gromov wrote:
> Hello Andy,
>
> Many thanks for your response. I will apply your patch and inform you
> about the result. Btw, to apply the patch I guess I should run:
>
> patch -p1 < your_patch
>
> Right?
>
> Best regards,
> Evgeniy
>
> Andy May wrote:
>> Evgeniy,
>>
>> There has been one small change since 2012.1.3 which might help, it's
>> included in the nightly tarball but since it's just 1 line I include it
>> here too:
>>
>> diff --git a/src/util/util.F b/src/util/util.F
>> index cc7326d..5bbaa4b 100644
>> --- a/src/util/util.F
>> +++ b/src/util/util.F
>> @@ -790,6 +790,7 @@ c
>> nw=icorrm()
>> if(nw.gt.nwork.and.nwork.lt.maxwork) then
>> if(nw.lt.minwork) goto 100
>> + nw=min(nw,maxwork)
>> iwork=icorr(nw)
>> call dsyev_x('V','L',n,x,m,d,q(iwork),nw,info)
>> call corlsr(iwork)
>>
>> Best wishes,
>>
>> Andy
>>
>> On 31/07/13 17:15, Evgeniy Gromov wrote:
>>> Hi Jeff,
>>>
>>> Thanks for your response. I attach the CONFIG file for the build info.
>>> Regarding BLAS, MKL was used. To run with Netlib should I recompile
>>> the code with -noblas option?
>>>
>>> Best regards,
>>> Evgeniy
>>>
>>> Jeff Hammond wrote:
>>>> You probably want to provide build info and BLAS details. I don't
>>>> know what info=4 means for dsyev but such errors are frequently caused
>>>> by build misconfiguration. Barring that, it might be a bug in an
>>>> optimized library. Running with pure Netlib is a useful debugging
>>>> exercise.
>>>>
>>>> Jeff
>>>>
>>>> On Wed, Jul 31, 2013 at 4:59 AM, Evgeniy Gromov
>>>> <Evgeniy.Gromov at pci.uni-heidelberg.de> wrote:
>>>>> Dear Developers and Users of Molpro,
>>>>>
>>>>> I encountered a problem when trying to do a EOM-CCSD calculation.
>>>>> The calculation crashes just prior to starting EOM-CCSD iterations:
>>>>>
>>>>> EOM-CCSD FOR 5 EXCITED STATES OF SYMMETRY 4 STATE(S) 1 2
>>>>> 3 4 5
>>>>>
>>>>>
>>>>> Number of N-1 electron functions: 121
>>>>> Number of N-2 electron functions: 7381
>>>>> Number of singly external CSFs: 11659
>>>>> Number of doubly external CSFs: 548182490
>>>>> Total number of CSFs: 548194150
>>>>>
>>>>> Iter. Vector |HC-EC| E_new-E_old Excitation energy Total
>>>>> energy
>>>>>
>>>>> info= 4
>>>>> ? Error
>>>>> ? failure in dsyev
>>>>> ? The problem occurs in diag2
>>>>>
>>>>> I would appreciate very much any help/ideas how to get it over.
>>>>> Many thanks!
>>>>>
>>>>>
>>>>> Best regards,
>>>>> Evgeniy
>>>>>
>>>>> --
>>>>> _______________________________________
>>>>> Dr. Evgeniy Gromov
>>>>> Theoretische Chemie
>>>>> Physikalisch-Chemisches Institut
>>>>> Im Neuenheimer Feld 229
>>>>> D-69120 Heidelberg
>>>>> Germany
>>>>>
>>>>> Telefon: +49/(0)6221/545263
>>>>> Fax: +49/(0)6221/545221
>>>>> E-mail: evgeniy at pci.uni-heidelberg.de
>>>>> _______________________________________
>>>>>
>>>>>
>>>>>
>>>>> _______________________________________________
>>>>> Molpro-user mailing list
>>>>> Molpro-user at molpro.net
>>>>> http://www.molpro.net/mailman/listinfo/molpro-user
>>>>
>>>>
>>>>
>>>
>>>
>>>
>>>
>>> _______________________________________________
>>> Molpro-user mailing list
>>> Molpro-user at molpro.net
>>> http://www.molpro.net/mailman/listinfo/molpro-user
>>>
>>
>
>



More information about the Molpro-user mailing list