[Pkg-openmpi-maintainers] Bug#592326: Failure of AZTEC test case run.

[Rachel Gordon]

Dear Manuel, Ralf and Jeff,

Unfortunately there are no news on this; The test case runs on 1 and 2 
processors (with the -pthread parameter) but it does not run on more 
processors.
I think it is somehow related to the link stage when using both fortran 
(compiled with the OPENMPI wrapper) and C codes.

The Aztec code works well (with its inner wrapper routines from fortran to 
c) on my old linux cluster using MPICH 1.2.4, gcc and g77.

Thanks again for your effort,

Sincerely,
Rachel

   Dr.  Rachel Gordon
   Senior Research Fellow   		Phone: +972-4-8293811
   Dept. of Aerospace Eng.		Fax:   +972 - 4 - 8292030
   The Technion, Haifa 32000, Israel     email: rgordon at tx.technion.ac.il


On Sat, 2 Oct 2010, Manuel Prinz wrote:

> Hi Rachel!
>
> On Fri, Sep 03, 2010 at 10:05:17AM +0300, Rachel Gordon wrote:
>> There are some good news,
>> I added -pthread  to both the compilation and link for running
>> az_tutorial_with_MPI.f, and I also compiled aztec with -pthread
>> Now the code runs O.K for np=1,2.
>>
>> Now bad news: when I try running with 3,4 or more processors I get a
>> similar error message:
>> [â&]
>
> Any news on this? As it seems to be a problem in aztec and not really
> related to Open MPI --at least not the Debian package-- I'd like to
> close this bug. Please let me know about the status, or I'll close this
> bug in a few days. Thanks in advance!
>
> In case you'd like to discuss this problem further, I think the user
> mailing list of Open MPI or aztec are the better place for that.
>
> Best regards,
> Manuel
>