[openib-general] segfault on openib mvapich

[Dhabaleswar Panda]

Federico, 

>     Federico> I might have done something wrong, but tried to build
>     Federico> using a plain source from the openib gen2 svn tree and
>     Federico> Pete's patches (those that were not rejected).
>  
> For whatever it's worth, basic MVAPICH tests like osu_bw work fine for
> me with two and even four processes on two x86_64 machines.

FYI, we are also running the latest version successfully on multiple
platforms (IA32, Opetron and EM64T) of different sizes.  We are also
able to run applications successfully.

To the best of our knowledge, many other organizations are also
running mvapich-gen2 successfully on their platforms.

>     Federico> Adding the -debug flag to mpirun_rsh does not help (the
>     Federico> xterms flash on then dissapear). The ssh connections are
>     Federico> started fine, but the segfault happens early on.
> 
> Without more data like a traceback from a core file or something like
> that, it's going to be very difficult for anyone to debug this.

As Roland indicates, could you please provide more details on the
platform, OpenIB version (kernel, userlib), and the errors you are
getting. This will help to debug the problem further and faster.

> Also, it might be worth contacting the MVAPICH developers by emailing
> mvapich_request -- they are much more likely to be able to help than
> the openib-general community.

We at OSU are monitoring the OpenIB list for mvapich-gen2 related
questions and are answering them. In addition, if you can send a copy
to mvapich-help at cse.ohio-state.edu (not mvapich_request), we will be
able to respond even faster.

Thanks, 

DK

> - R.  >
_______________________________________________ > openib-general
mailing list > openib-general at openib.org >
http://openib.org/mailman/listinfo/openib-general > > To unsubscribe,
please visit http://openib.org/mailman/listinfo/openib-general >