Is it bug?

[navneet roshan]

Dear Deal.II community, 

I have been using deal.ii for more than a year. I am solving the reaction-diffusion problem, I am getting a bizarre runtime error in the parallel solver. I am using PETSc CG solver. The error appeared after parallelization.  After 3-4 days of debugging, I found the error appears only if I am using 'repetition' (GridGenerator::subdivided_hyper_rectangle(triangulation,repetitions, a, b,false)) in the rectangular (GridGenerator::subdivided_hyper_rectangle) domain mesh creation. I am attaching a  minimal code showing the issue(minimal_diff_code.cc). 

     I am calling it bizarre because the code runs perfectly with few combinations of 'repetitions' and Global mesh Refinement numbers, but fails with some other combinations. A table(error_Table.pdf) is attached showing how the error depends on the number of repetitions and global_refinement, and sometimes also on the number of processes used. I have tested it in deal.ii 8.5.1 and 9.0.0 also.

I will be thankful if someone could look into this.

Thank you

Navneet

[Daniel Arndt]

Navneet,

What is the error you are observing?

Best,

Daniel

--
The deal.II project is located at http://www.dealii.org/
For mailing list/forum options, see https://groups.google.com/d/forum/dealii?hl=en
---
You received this message because you are subscribed to the Google Groups "deal.II User Group" group.
To unsubscribe from this group and stop receiving emails from it, send an email to dealii+un...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/dealii/CAK9McD15f4B4e94-pFfdj84HmWAOhe0Btd41Hi0uBAozE4ma9A%40mail.gmail.com.

[navneet roshan]

Hi Daniel, 

Thank you for the quick response, If I change line number 190 in the code, to 1 from 4. I am getting following error but no further information even in the debug mode.

make[3]: *** [CMakeFiles/run.dir/build.make:58: CMakeFiles/run] Segmentation fault (core dumped)
make[2]: *** [CMakeFiles/Makefile2:265: CMakeFiles/run.dir/all] Error 2
make[1]: *** [CMakeFiles/Makefile2:272: CMakeFiles/run.dir/rule] Error 2
make: *** [Makefile:196: run] Error 2

Navneet

To view this discussion on the web visit https://groups.google.com/d/msgid/dealii/CAOYDWbKhrrKQ3GqwiyVuOTAAxSaugAb%3Du0%3DFsFXb8n7M%3DuiBzA%40mail.gmail.com.

[Daniel Arndt]

Navneet, what happens if you run your program in gdb or some other debugger?

To view this discussion on the web visit https://groups.google.com/d/msgid/dealii/CAK9McD1%3D6tNDTjaR3OYiggEE-MHkLp6st9bmeXPqhvXEhLkUWw%40mail.gmail.com.

[navneet roshan]

Hi Daniel, 

Thank you, I have not used debugger before, I will try tomorrow, but I am skeptical how much of it I will understand .

Navneet

To view this discussion on the web visit https://groups.google.com/d/msgid/dealii/CAOYDWbK-WRv5O6WvTERw1x%2Bda8P1NH4CwVtB0aW7BS5rjJnLAA%40mail.gmail.com.

[Wolfgang Bangerth]

On 12/13/19 12:41 PM, navneet roshan wrote:
>
> Thank you, I have not used debugger before, I will try tomorrow, but I
> am skeptical how much of it I will understand .

Learn to use a debugger. You will be amazed at how much information you
can get from a debugger about what exactly is going wrong in a program,
and how much faster it makes your debugging process.

It will of course take some time to learn how to use a debugger
efficiently, but it is well worth the effort. There are also a number of
video lectures about using the debugger in eclipse.

Best
W.


--
------------------------------------------------------------------------
Wolfgang Bangerth email: bang...@colostate.edu
www: http://www.math.colostate.edu/~bangerth/