[Getdp] Re: getdp Digest, Vol 7, Issue 1

Gyselinck Johan Johan.Gyselinck at ulg.ac.be
Fri Aug 1 16:34:53 CEST 2003


>
>
> ----------------------------------------------------------------------
>
> Message: 1
> Date: Thu, 31 Jul 2003 06:43:00 -0700
> From: Christophe Geuzaine <c.geuzaine at ulg.ac.be>
> Subject: Re: [Getdp] (no subject)
> To: "Brian P. Wilfley" <bwilfley at nexraymedical.com>
> Cc: getdp at geuz.org
> Message-ID: <3F291CE4.6040408 at ulg.ac.be>
> Content-Type: text/plain; charset=us-ascii; format=flowed
>
> Brian P. Wilfley wrote:
>
>> Hi folks,
>>
>> I ran a problem (which is not working for other reasons) with the
>> "Full Matrix_Format" option.  I got the following crash.  If someone
>> would like my input files, I can send them under separate cover.
>>
>> Thanks.
>>
>> bpw
>> -----------------------------------------------------------------
>>
>> $ getdp heat1 -solve resTemp
>> Loading   : Problem definition 'heat1.pro'
>> Loading   : Problem definition 'Jacobian_2DAxi.pro'
>> Loading   : Problem definition 'Integration_Lib.pro'
>> Loading   : Problem definition 'TempAxiTime.pro'
>> Info      : Selected Resolution 'resTemp'
>> Loading   : Geometric data 'heat1.msh'
>> Info      : System 'sysTemp' : Real
>> P r e - P r o c e s s i n g . . .
>> Operation : Treatment Formulation 'frmTemp'
>> Info      :   Generate ExtendedGroup '_CO_Entity_9' (NodesOf)
>> Resources : cpu 0.70000 s / mem 2224 kb
>> E n d   P r e - P r o c e s s i n g
>> P r o c e s s i n g . . .
>> Operation : InitSolution[sysTemp]
>> Operation : SaveSolution[sysTemp]
>> Operation : TimeLoopTheta ...
>> Info      : Theta Time = 1e-08 s (TimeStep 1)
>> Operation : Generate[sysTemp]
>> Solver    : Loading parameter file 'solver.par'
>> Info      : Setting System {A,b} to zero
>> Resources : cpu 0.250000 s / mem 5756 kb
>> Operation : Solve[sysTemp]
>> Warning   : Null right hand side in linear system

Already here it goes wrong. The right hand side vector should not be zero.
GetDP doesn't (always) treat this case properly.

Johan


>> Solver    : Dense to sparse matrix conversion
>> Resources : cpu 0.300000 s / mem 6808 kb
>> Solver    : No scaling of system of equations
>> Solver    : No renumbering
>> Solver    : N: 515, NZ: 265225, BW max/avg: 515/515, SW max: 515
>> Resources : cpu 0.300000 s / mem 6812 kb
>> Solver    : ILUTP (Float, fill-in = 20)
>> Solver    : N: 515, NZ: 14095, BW max/avg: 40/27, SW max: 492
>> Resources : cpu 0.410000 s / mem 6944 kb
>> Solver    : Generalized Minimum RESidual (GMRES)
>>     1  0.0000000e+00  NaN
>
> You already have a "Not a Number" here--don't expect anything after this
> to make sense :-)
>
> Christophe
>
>> Solver    : 1 Iterations / Residual: 0
>> Resources : cpu 0.420000 s / mem 3736 kb
>> Operation : SaveSolution[sysTemp]
>> Info      : Theta Time = 2e-08 s (TimeStep 2)
>> Operation : Generate[sysTemp]
>> Info      : Setting System {A,b} to zero
>> Error     : Segmentation violation (invalid memory reference)
>>             ------------------------------------------------------ You
>> have discovered a bug in GetDP! You may report it by
>> e-mail (together with any helpful data permitting to
>> reproduce it) to <getdp at geuz.org>
>>
>> _______________________________________________