[petsc-users] code hangs inside SUPERLU DIST

[Matthew Knepley]

On Sun, Apr 1, 2012 at 2:18 PM, Wen Jiang <jiangwen84 at gmail.com> wrote:

> Sorry for the last incomplete email. Please forget about it. Let me finish
> it up here.
>
>
> Hello Everyone,
>
> I am using KSPPRECONLY type ksp with SUPERLU DIST. However, I found my
> code hangs inside superlu_dist solver and never finishes.
>
> The problem I am solving is a time-dependent nonlinear pde and I use fully
> Newton-Raphson for each time step. So my code will look like,
>
> For Time STEP = From start to end
>
>     For Newton-Raphson Iterations = From 1 to max iterations
>
>         Calculate Matrix A; (A is different for each Newton-Raphson
> iteration)
>
>         Create KSP
>         KSPSetOperator(A,A,same_nonzero_pattern)
>         KSPSolve;
>
>        KSPDestroy;
>
>
>     END Newton-Raphson Iterations
>
> End Time STEP
>
> Could you give me any hints what the problem might be? I am using Petsc
> 3.1 p8 and superlu dist v2.4. And my code works fine with KSP iterative
> solver like GMRES.
>

There are a list of things I would try:

1) Upgrading to the latest release

2) Trying MUMPS instead of SuperLU_dist

3) Get a stack trace from gdb to determine where it is hanging.

I would do whichever one is easiest.

    Matt


> Regards,
> Wen
>
>
> On Sun, Apr 1, 2012 at 3:12 PM, Wen Jiang <jiangwen84 at gmail.com> wrote:
>
>> Hello Everyone,
>>
>> I am using KSPPRECONLY type ksp with SUPERLU DIST. However, I found my
>> code hangs inside superlu_dist solver and never finishes.
>>
>> The problem I am solving is a time-dependent nonlinear pde and I use
>> fully Newton-Raphson for each time step. So my code will look like,
>>
>> For Time STEP = start to end
>>
>>     For Newton-Raphson Iterations = 1 to max iterations
>>
>>         Calculate Matrix A;
>>
>>
>>
>>     END Newton-Raphson Iterations
>>
>> End Time STEP
>>
>>
>> Could you give me any hints what the problem might be? I am using Petsc
>> 3.1 p8 and superlu dist v2.4.
>>
>
>


-- 
What most experimenters take for granted before they begin their
experiments is infinitely more interesting than any results to which their
experiments lead.
-- Norbert Wiener
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.mcs.anl.gov/pipermail/petsc-users/attachments/20120401/a1e5611e/attachment.htm>