[OMPI devel] OpenMPI error with error codes. The proposed solution.

[Дмитрий Донцов]

Hello!



I've found error in OpenMPI with error codes.

According to MPI standard (2.2 and later):

MPI_ERR_LASTCODE must be equals to last error code and MPI_LASTUSEDCODE
must be greater than or equal to MPI_ERR_LASTCODE.



In OpenMPI MPI_ERR_LASTCODE is greater than real last error code
(MPI_ERR_WIN) for 1 and MPI_LASTUSEDCODE hardcoded linked with MPI_ERR_WIN.
So, possible situation when MPI_LASTUSEDCODE (equals to MPI_ERR_WIN) is
less than MPI_ERR_LASTCODE.



I've investigated it and prepared patch (based on non-configured
openmpi-1.7rc8 code) which corrects it and makes appending of new
predefined error codes more easy. Please find it in attachments.





Best regards,

Dmitry



/**

* mobile: +7 (920) 111-2882

* Skype : dontsov.dmitry

**/


openmpi.patch
Description: Binary data

Re: [OMPI devel] Linux memory hooks testing

[Joshua Ladd]

Brian,

I can confirm your latest fix (in the trunk) works on our SB system. I can 
compile and run, and I get the bandwidth I expect when "leave_pinned" is 
enabled.

Best,

Josh

-Original Message-
From: devel-boun...@open-mpi.org [mailto:devel-boun...@open-mpi.org] On Behalf 
Of Barrett, Brian W
Sent: Friday, March 22, 2013 1:11 PM
To: Open MPI Developers
Subject: [OMPI devel] Linux memory hooks testing

All -

In an effort to get 1.7 out the door, I'm asking that everyone who cares tries 
the memory hooks on Linux to make sure they work as expected on the trunk.  
That is, if you rely on leave pinned, make sure you get the bandwidth you 
expect.  If you don't, make sure things compile in all your compile modes.

Thaks!

Brian

--
  Brian W. Barrett
  Scalable System Software Group
  Sandia National Laboratories





___
devel mailing list
de...@open-mpi.org
http://www.open-mpi.org/mailman/listinfo.cgi/devel